(SANITIZED)UNCLASSIFIED SOVIET BLOC PAPERS ON MICROBIOLOGY, 1958(SANITIZED)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R002800080004-5
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 118
• Document Creation Date: December 27, 2016
• Document Release Date: March 5, 2014
• Sequence Number: 4
• Publication Date: November 4, 1958
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
R
50X1 -HUM
Next 20 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
THE PHYSIOLOGICAL STATE OF MICROORGANISMS
DURING CONTINUOUS CULTURE
IvAN 'MA LEK
In the course of the last 10 years extraordinary attention has been paid
to continuous culture. We can mark an increase in the number of papers which
show that it is about to become a new method enabling U3 to deepen our know-
ledge about the multiplication of microorganisms and at the same time gives
us a stable living material for experimental studies of variability and muta-
bility and for biochemical analyses. It also points the way towards a substantial
increase of productivity for a number of fermentations.
Up to 1945 there existed but a few papers describing the use of this method.
Moreover these papers usually did not consider this method as a new and basic
one and one which corresponds to the reproductive capacity of microorganisms
better than the commonly used static methods. This is true for the work of
Rogers and Whittier (1930) who tried to draw an analogy between a bacterial
culture and a multicellular organism, and cultivated bacteria (Strepococcus
lactis) at a rather slow rate of flow of the nutrient medium on the one hand,
and using a microbial filter under which the nutrient medium was repeatedly
renewed on the other hand. Their most striking results deal with the lactic
acid production in continuous culture. Cleary, Beard and Clifton (1935) used
the continuous method to study the basis of the stationary phase of common
batch culture methods. The papers of Jordan and Jacobs (1944,.1947, 1948)
also deal with an analysis of the stationary phase. Moyer (1929) used the conti-
nuous method only to increase the bacterial mass for chemical analysis.
Among the theoretical papers from this period, only those of Utenkov (1929,
1942) and Malek (1943) considered continuous culture as a new experimental
method from the very beginning. Utenkov, who was probably the first to take
up the continuous flow method systematically, proceeded in 16 years of experi-
mental work (from 1922) from the assumption that the batch culture method
does not sufficiently reveal the true characteristics of microorganisms, and
does not permit satisfactory regulation of the development of cultures. It is
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
11
STAT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
only to be regretted that Utenkov's paper remained unknown to the world micro-
biological literature. It was from a similar of point view that Mo.lek studied the
continuous flow method (1943) and stressed its advantages in studying multi-
plication, variribility and pathogenicity of microorganisms.
Practical exploitation of the method was only slightly developed. Lebedev
(1936) developed the method of continuous alcoholic fermentation. In Ger-
many, fermentations have been patented which are based on continuous flow
fermentation (Lupinit, Norddeutsche Hefeindustric, 1934); later on, a group
of American research workers - Unger, Stark, Sealf, Kolachov (1942) and
others - worked out procedures for continuous production of yeast and alcohol,
and pointed out their advantages and greater productivity as compared with
the static "batch" methods.
Later papers (after 1945), however, almost without exception, consider
continuous flow culture as a new method. It has been studied technically both
from the experimental-laboratory viewpoint (Castor (1947), Bactogen-Monod
(1950), Chemostat-Novick and Szilard (1950), Anderson (1953), Malek (1943,
1952), Northrop (1954), Kubitschek (1954), Graziosi (1956, 1957), Davies
(1956), Karush, Iacocca, Harris (1956), Formal, Baron, Spilman (1956), Perret
(1956)) and from the laboratory-production viewpoint (Malmgren and Heiden
(1952), Me.lek (1955), Elsworth et al. (1956), Pirt (1957)) as well as applied
to microcultures (Pirfilev - personal communication, Rosenberg (1956)).
A methematical treatment has been given for the theory of reactions in conti-
nuous flow systems in general (Denbigh (1944), Pasynskij (1957)) and in parti-
cular to the multiplication of microorganisms (Monod (1950), Northrop (1954),
Herbert et al. (1956), Maxon (1955), Jerusalimskij (1958). The method has
been considered from the point of view of experimental possibilities and
perspectives (Monod (1950), Mack (1943, 1952), Maxon (1955), Novick (1955),
Powell (1956)). The following theoretical problems have been studied by this
method: induction of antibiotic properties (Sevage and Florey (1950)), the
formation of mutants and other problems of genetics (Novick and Szilard
(1950), Bryson (1953), Zelle (1955), Lee (1953), Moser (1954)) microbial adap-
tation (Verbina (1955), Graziosi (1956, 1957)), growth of fungi (Duche and Nett
(1953), Hofsten et al. (1953)), bacterial multiplication (Brucella: Gerhardt
(1946); Aerobacter: _Herbert et al. (1956), Pirt (1957); Escherichia coli: Mtilek
(1943, 1950); Streptococcus haem. A.: Karush et al. (1956); Salmonella typhi:
Formal et al. (1956); Mycobacterium lb.: gvachulova and KuAka (1956)), and
development of bacterial cultures (Mitlek et al. (1952, 1953, 1953b, 1955),
Macura and Kotkovi (1953), '8eveik (1952), Jerusalimskij (1956, 1958)).
Other papers deal with the industrial application of the method (Harris et al.
(1948),'Viptorero (1948); Adams and Hungate (1950), garkov and his school
(1950)IElsiv-OrOt-et al. (1956), Malek et al. (1955, 1957)).
Experience haS been acquired not only with the continuous culture of bacte-
12
? ? - ? ? ? ? ? ? - ws ?
_
.ve4.1.1.11,1
ria, yeasts and molds but also of some algae (Ketchum, Bostwick, and Red-
field (193S), Myers et al. (1944) cf. Novick, Tamiya (1957)) and protozoa
(Browning and Lockinger (1953), Vtivra (195S)).
In this review we intend to consider only theoretical papers and results parti-
cularly with regard to the physiological state of microorganisms in continuous
flow cultures. Therefore its mathematical aspect and the kinetics of growth
of microorganisms will not be covered, since this question has been sufficiently
studied in earlier papers (Monod, Northrop, Maxon, Herbert et al.).
Let us begin with a few remarks on t erminology. When referring to "continuous
cultures" we shall usually have in mind continuous flow cultures, as compared with
continuous culture in the broader sense of the word, where culture with periodic renewal
of nutrient medium is included (for this typo of culture sometimes the word "semi-
continuous" according to :Maxon is use(l). Continuous flow culture is thus the extreme
case of periodic continuous culture; the technical difference between these two is usually
only in the length of the interval between additions of medium, because even in a con-
tinuous flow culture the nutrient medium is usually added intermittently. This difference
is at times of great importance for the process. In contrast to continuous culture we shall
consider the static single-run culture i. e. the common batch culture. Sometimes the term
"dynamic culture" is used even for a batch culture when stirred. It is felt that this usage
is incorrect because this type of culture remains basically static in character. It is typical
of a true dynamic culture that a dynamic steady state is established.
What do we actually mean by the physiological state of microor-
ganisms in continuous culture?
A lot of experience obtained in studying common static cultures have shown
that the number of microorganisms follows the typical growth curve. However,
only the exponential part of this curve is significant for multiplication of
microorganisms in common continuous cultures, because only then is uniform
multiplication taking place. Therefore, this part of the curve is usually treated
as a whole quite uniformly, as no striking change of culture can be observed
there. A number of papers (Malmgren and 1-16don (195 I ), B nshclwood (1947),
Valyi-Nagy (1955) and many others) have shown that microorganisms are
undergoing changes even in this exponential part of the curve. Thus, for
instance, the ribonucleic acid content of cells does not remain constant during
the exponential period, i. e. the period of uniform cell division, but usually
drops very rapidly at the very beginning of the curve. On the other hand there
are a number of reasons for the assumption that the beginning of multiplication
coincides with the end of the lag-phase, i. e. with the late lag-phase, because
then a great production of microbial material can be observed as a consequence
of regular microbial proteosynthesis, as shown by an increase in enzymic
activity, by a rise in the ribonucleic acid content, and by an increased sensiti-
vity towards external conditions (Malingren and Redon (1951 ), Winslow,
Walker et al. (1939) and other authors). All this indicates that the physiological
state of microorganisms, as manifested by enzymic activity, sensitivity, proteo-
13
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
_
synthesis, RNA content etc., is undergoing changes in static cultures even
during the phase of full multiplication. The conditions, however, which lead
. -
towards this change have not been sufficiently elucidated.
Now, to which part of the curve which has been thus divided can we compare
the physiological state of microorganisms under conditions of continuous
multiplication? A logical consideration leads to the conclusion that under
conditions fully ensuring the multiplication of microorganisms, the physiologic-
al state will most likely correspond to the state of a culture in the late lag-phase
or at the beginning of the log-phase. But what is the case when microorganisms
are grown under conditions which are below this level, and when slower multi-
plication is taking place; does the physiological state of cultures then change
in a similar sense as can be observed in static cultures?
This question appears to be rather important, since the physiological state
can considerably influence multiplication itself and the kinetics thereof.
Only when we have fully answered this question can we purposefully regulate
continuous processes and exploit them for theoretical or practical ends.
It is particularly important in cases when we intend to grow microorganisms
over a long period with a maximum activity of multiplication, or when we wish
to obtain products which in static cultures are associated with changes in the
state of the culture, and appear therefore either at the end of the exponential,
or during the stationary phase. The question arises, however, whether we
can draw any parallel between continuous and static cultures in terms of the
individual phases of development.
It is most probable that laws governing the multiplication of microorganisms
in static cultures during the phase when the nutrient medium contains suffi-
cient amounts of all necessary components, and when it is not affected by
metabolites produced, are similar to those of continuous cultures. But even
then there is an important difference between these two types of cultures:
In static cultures the concentration of nutrients is usually excessive at the
moment of beginning growth, when it is intended to last for several generations;
in continuous cultures, on the contrary, the concentration of added nutrients
is diluted into the whole volume of nutrient medium with the microbial culture,
which min thus assimilate it immediately. When the basic nutrient factors or the
source of energy are optimally balanced with respect to the requirements
of a multiplying culture, we reach the stage where nutrients are assimilated
on addition and cannot be determined in the culture fluid although constantly
replenished. This is another aspect of the steady state of continuous cultures,
and very important from the point of view of optimal exploitation of nutrients
?a definite requirement in practical application. Is this difference in any way
reflected in the physiological state of microorganisms in culture, as to the
manner of absorbing nutrients and in the manner of their ,utilisation? Such
a possibility cannot be excluded. What is the case when microorganisms are
14
cultivated in such a way that the rate of continuous feed is slower than would
correspond to the growth rate as required by Monod for cultures in which self-
regulation is operating? Is the answer given by merely slowing down multi-
plication, or is there a concurrent change of physiological state as is known
in the latter stages of the exponential curve of a static culture? It cannot be
excluded that some change could take place together with a different response
of cultivated microorganisms to external conditions. Evidence on this would
be of great experimental importance.
These are some of the questions that should help us to show the relation
between the basic problem of the physiological state of cultures under condit-
ions of continuous multiplication, and the application of this method in theory
and practice. These questions have thus far been accorded little experimental
attention, despite the importance for evaluation of obtained and obtainable
results.
The attitude toward this question is connected with another problem of
particular importance during the first period of study of continuous cultures,
and of general biological importance: whether conditions of static or dyna-
mic cultures better meet the physiological requirements of microorganisms.
By the term "requirement" we have in mind the relation of microorganisms
to their environment as developed in the course of their exposure to physio-
logical conditions in nature.
Older papers ? with the exception of those of Utenkov and Malek ?
assumed that the growth-curve characteristics derived from static cultures
have an absolute validity, and that they reflect the natural development
of the physiological state of microorganisms and their requirements. Cleary
et al. (1935) therefore did not study continuous multiplication from the point
of view of microbial multiplication, but rather in order to learn more about
the stationary phase of static cultures. This attitude was apparently an
expression of the fact that in that period a diagnostic raison d'6tre still pre-
vailed in microbiology, for which static cultures are most convenient. Parti-
cular attention has been given to the study of the so called "maximum concen-
tration" according to Bail (1929). There were however more profound biological
reasons for this view: under natural conditions (e. g. in the soil etc.) micro-
organisms do not as a rule multiply in a homogeneous solution with a constant
afflux of nutrients, but rather on structures where they behave similarly
as in artificial cultures on solid media ? they form colonies which arc analogous
to the well-known growth-curves of liquid static cultures (e. g. Vinogradskij
(cf. Volodin 1952), Novogrudskij (1950)). From this the conclusion is usually
drawn that microorganisms have developed a fixed form of multiplication
which corresponds to the conditions of static cultures, and that even when they
can multiply without limitation, cultures undergo developmental changes
that correspond to the typical growth-curve of static cultures. The practical
Declassified in Part -Sanitized Copy Approved for Release @50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
15
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
-
consequence of such assumptions was that microorganisms were cultivated
in continuous cultures with an afflux of nutrients at a rate considerably lower
than would correspond to the growth rate. Another consequence of this
thinking was that it was not generally believed possible to cultivate micro-
organisms in continuous cultures ad infinitum without 801110 degeneration
or change that would correspond to the stationary and decrease phases of
static cultures. This opinion constituted an obstacle in accepting continuous
culture methods and developing them in fermentation. This opinion has been
shown to be wrong by work describing continuous cultures of microorganisms
over long periods of time without signs of degeneration (Herbert et al. (1956),
Malek (1955) and others).
A logical conclusion from the above findings would appear to be that static
culture and the developmental changes observed with it are artefacts. The
lag-phase appears as an artefact because, in the course of it, a culture changed
by previous static cultivation must adapt itself to regular multiplication; the
decrease of. biosynthesis in the latter part of the exponential phage also appears
as an artefact, as well as the whole of the stationary phase. The natural condit-
ions which most fully correspond to the dynamics of microbial multiplication
exist only in continuous culture. Therefore, only by using this method can
we produce cultures which are really physiological, and correspond to the real
physiological and biochemical characteristics of microorganisms. This view
is supported by the results of continuous cultures which have shown that
microorganisms can be grown under such conditions for indefinite periods in
a vegetative form, with optimal results as to the production of living matter
and optimal enzymic activity, etc. From the above the conclusion can be
drawn that in static cultures only the initial, the late lag, and the exponential
phases, can be considered as physiological, because only then can the synthesis
of living matter, and all phenomena connected with it, proceed in an uninhi-
bited manner. Everything else in the static growth curve appears to be an
artefact caused by conditions inadequate for the multiplication of microorga-
nisms.
This conclusion does not take into account the fact that some of the deve-
lopmental changes in static cultures, at least in some microorganisms, have
the character of fixed traits: above all the sporulation of bacilli (and, apparent-
ly, also of actinomycetes) and probably also the formation of resting forms
of bacteria in the stationary phase. These traits are a biological fact which
shows that in the course of development bacilli had to adapt to conditions
somewhat. analogous to those of static cultures. But the same microorganisms
can be kept constantly in a vegetative state under suitable conditions. Further-
more a ,number oflimportant products, which are certainly not a laboratory
artefact, and which are of great practical importance, are formed only during
tbe'latter 'phase Of -static culture (antibiotics etc.). It cannot be affirmed
16
Ii
therefore that continuous flow cultures in their commonly used
simple form correspond to all the physiological variables and
requirements of microorganisms as which have developed. They
meet the optimal requirements of fast vegetative multiplication of micro-
organisms; therefore, all the phenomena which arc associated with the multi-
plication of microorganisms, be it the production of living matter, some basic
enzymic processes, the influence of the environmental conditions on actively
multiplying cells, or spontaneous mutability, can be conveniently studied
only in continuous cultures. Only these cultures, when well set up technically
and under constant conditions, can produce stable material for such investig-
ation. On the other hand, in order to study phenomena caused by changes
in environmental conditions due to the activity of microorganisms themselves,
it is necessary to resort to a static culture, or to modify suitably a continuous
culture method.
The majority of research workers using the continuous culture method base
their views on the second assumption and do not take into account the possi-
bility of changes of the physiological state under changed conditions. Theore-
tical papers devoted to the mathematical basis of multiplication (Monod,
Herbert, Maxon and others) presume the existence of an ideal state in which
the rate of growth and the activity of metabolic processes represent values
dependent only very simply on the flow rate (dilution). This ideal state is
hypothetically reached by assuming a system in equilibrium when the dilution
rate (i. e. the ratio of emptied volume per unit of time to the volume of the
culture) is equal to the multiplication rate (i. e. the number of divisions per
unit of time) multiplied by 0,69 (= logc 2). Should some change in the physio-
logical state of the culture occur, a new equilibrium would be formed, on the
basis of which the experiment proceeds. Therefore no qualitative change in the
physiological state of microorganisms is considered; it is taken to be constant
in the course of a given experiment. The steady state which exists at
different rates of flow is usually considered to differ only quantitatively. This
abstraction is necessary as a basis for experimentation and it can be neglected
within certain limits of multiplication rate. This assumption served as the
basis for methods of the type of Monod's Bactogen or Novick's and Szilard's
Chemostat, which choose an afflux of nutrients which remains below the level
ensuring a maximum growth rate, and limit one important nutrition com-
ponent. Other methods proceeding along the same line are those of the Tur-
bidostat type, as used by Northrop, Bryson, Anderson and others, when a pho-
to-cell helps to keep the density constant and the afflux changes in intensity.
By means of both of the above-mentioned methods, the rate of growth can
be automatically controlled and kept constant ? in the first case by means
of the self-regulating ability of the continuous flow system kept below the
level of maximum growth rate, in the second case by means of an external
2?Symposium
17
Declassified in Part -Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
mechanism. Even workers using these methods to study important physio-
logical processes (Monod, Duche and Neu, Karush et al.) or genetic processes
(Novick and Szilard and others) work under conditions that disregard the
physiological state and its.changes. Thus they have the possibility of studying
a number of important biochemical, physiological and genetic problems
considerably more exactly, and with a more constant and more physiological
microbial material than is possible in static cultures, but they do not fully
exhaust the advantages of continuous methods for studying more thoroughly
the biological factor itself and applying the results in practical fermentations.
It follows from the conclusion of Monod's paper that he is aware of this simpli-
fication. NorthrOp stresses in the introduction of his paper that "any change
inthe concentration of any substance indicates a change in organism" and
Maxon in his paper points out that a rigorous mathematical treatment
of all factors is noteven possible, and so far not useful, because a fermentation
is a highly complex living-system.
Itseems particularly important to investigate to what extent we are
justifiedin -drawing a parallel between the slow growth achieved
in ,the. Chemostat by limiting one of the important nutrient
factors (cf. e. g. Novick and Szilard 1950), and the faster growth
in the physiological state. It remains a question what role a similar
limitation might play in a continuous culture. This question is discussed by
Powell (1956) from the mathematical point of view and in relation to the gener-
ation time of individual microorganisms. He proves that a continuous flow
culture "discriminates heavily-against organisms of unusually long generation
time". Such long generation time certainly reflects a definite physiological
state. Powell:assumes in his conclusion that a continuous culture will stabilize
itself: within_ a certain range of physiological activity of microorganisms.
But the question appears to be even more complicated: it is assumed, on the
basis of a statistical treatment .of the whole population, that the culture is
homogeneous, but such is not the case even in a. well stirred continuous culture.
Let ,us, therefore consider the effect of7limiting one of, the important nutrient
_
factors: will the result of this procedure be that all the microorganisms present
in the population .eonsume,.an..equal minimum amount of' the limited
factor and therefore multiply themselves. more slowly (and in turn influence
_ - ? _ ,
their phyiliological, state); or that microorganisms metabolically more active
consume the limited factor, depriving less active bacteria? In the latter
_ _ _
case the culture can be divided into a_portion uth a normal generation time
_
corresponding to temperature etc, and into ,a portion with an abnormally long
generation time, or not dividing at all The resulting generation time would
reflect the statistical difference ' of ' these cases, but would -net-. give a' true
_ _
picture: of the_physiological. state -of,rthe_microOrganisnia present. From this
- _ z
point of view:it is interesting to notethe remark of Karush et al..(1956),that
- _
-
"the efficiency of utilisation of glucose increases with increasing growth rate".
The above is intended to point out the importance of a more thorough investi-
gation of growth limitation in continuous cultures achieved by limiting one
important source. It is particularly important because under the conditions
of limited growth, mutability has been studied. The relation of this phenomen
to the physiological state deserves attention.
The assumption that only continuous culture is physiological is a starting
point for practical applications in simple fermentative productions, as for
instance, the production of yeast. Even with these simple processes, the ques-
tion of physiological state can play an important role. For instance, in yeast
cultivation the question arises of the behaviour of glycolytic activity when
aeration is prolonged, or what will happen with maltase activity during
prolonged culture on molasses. Similar questions arise as to physiological
state when growth takes place on complex media, e. g. wood hydrolysates
or sulphite liquors, and the capacity to utilise a mixture of carbohydrates, the
role-of diauxia, etc.
We have now compiled sufficient proof to show that the question of the
physiological state of microorganisms is important for the knowledge of condi-
tions and possibilities of continuous culture. But so far little attention has been
given to this question. Utenkov proceeds from it rather systematically by his
method, which he has called "microgeneration". He worked out in detail
the possibilities of combination of the continuous flow method in various
applications (e. g. aerobic culture, the possibility of combining media without
interrupting the experiment, and the like) with static methods. His broad
experimental work is available only in a brief summary of his doctorate
thesis from 1942, where the principles of this method are presented together
with a summary of results from 16 years of work (from 1922) concerning the
development and variability of 40 different species of microorganisms under
continuous flow conditions as compared with static conditions. As mentioned
above, all his work is motivated by the endeavour to influence the physiolo-
gical state of microorganisms through a suitable manipulation of cultivation.
Utenkov is convinced that through a suitable choice of cultivation method
it would be possible to keep microorganisms constantly at the different stages
of their development. The continuous culture method represents for him the
method for maintaining constant the stage of active vegetative multiplication
and its practical exploitation. He studied with particular attention the first
stage Of continuous growth immediately after inoculation, as well as the occur-
renee_and_gignificance of atypical bacterial forms: He hoticed that growth rate
is higher in a continuous culture.than_1n a static one, that the cultures reach
a steady-state, that under 'the conditions of continuous cultivation it, is pbSsible
?
to ,preserve- certain characteristics- of microorganisms (e g. virulence) 'Which
is of importance for the preparation of -vaccines- and for the possibility of
_ _ _
2'
-
Declassified in Part - Sanitized Copy Approved for Release
_
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
19
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
producing long-term (permanent) modifications under changed conditions.
He furthermore treated theoretically, and partly experimentally, the question
of mutants. In conclusion he stressed the advantage of the method for stu-
dying microbial associations, and for practical tasks, because it forms a basis
for the application of automatisation of microbial growth etc... Even if these
experiments are mostly descriptive and not sufficiently analytical, they repre-
sent the basis of continuous cultures, and what is even more important, they
always concentrate upon the microorganism itself and on its physiological
state and its relation to the environment.
Powell (1957) also touches on the physiological state of microorganisms, and
considers the growth rate and generation time of bacteria in relation to the
conditions of continuous culture.
The physiological state of yeasts under continuous culture conditions have
been taken up by Plevako, who studied the dynamics of subcultures from
individual yeast cells (the material of this symposium (1958)). Jerusalimskij
and his coworkers are studying it in detail with cultivation of Cl. acetobutylicum
(1958).
The question of the physiological state of microorganism under continuous
flow conditions has formed also the basis of our studies (Mitlek et al. 1943 und
other papers). We assumed that static batch cultures are in antagonism with
the dynamics of a multiplying culture, and do not comply with that part of
physiology of microorganisms which is manifested by multiplication. From
this antagonism arise those complicated population and physiological changes
of microorganisms as are known from common batch cultures.
What are the possibilities of studying the physiological state of microorga-
nisms? For this purpose we must bear in mind the analogy with experiences
obtained in work with static batch cultures, because only here have those
parameters been studied which reflect the physiological state, although care
must be .taken in making such comparisons. The physiological state manifests
itself above all in the qualitative aspect of multiplication, and in the proteo-
synthesis associated with it. If we permit an analogy with static cultures it is
necessary to determine -Whether the culture and its characteristics correspond
to the culture from the first part of the exponential curve, or from a later
stage. It appears very convenient therefore to estimate the ribonucleic acid
content,- because- this reflects the most marked changes in this period. The
physiological state is also reflected in the qualitative changes in enzymic
systems, cell resistance, etc.
As -far as the first, purely quantitative, aspect is concerned, it is important
to study multiplication and its Utilization of living sources. By comparing
-
the data obtained with mathematically predicted results, we can estimate
-
20
whether the system corresponds to the optimal capacity of vegetative repro-
duction. The mathematical basis has been studied in a number of papers, and
can well be used for the treatment of simple continuous systems.
That is of course possible only in a perfectly homogeneous stirred culture.
But the conditions of such a culture need not necessarily meet all the aspects
of a continuous culture. Thus, for instance, a common problem that had to be
solved by some authors is that even in fully stirred cultures, some bacteria
settle preferently on the glass of the culture flask, and grow there. This pheno-
menon is particularly striking when cultures are grown in small flasks where the
only movement is provided for by the nutrient medium dropping into the flask,
and by its removal. Under such conditions the culture separates into two
distinct parts even at a very fast rate of flow, as we have seen in a long-term
culture of Edcherichia coli; bacteria multiply in the nutrient liquid on
the one hand ? there they resemble the actively multiplying organisms
as to shape and size ? and grow in a thick film along the glass of the flask
on the other hand ? where they have the shape of small rods, multiplying
apparently at a slower rate. The former are like those from the first part of the
exponential curve, the latter like those from the stationary phase. But of
course such a simple type of cultivation cannot provide us with any more
exact data than those presented by a mere description.
Therefore we have developed an experimental method based on multi-
stage cultivation, where several flasks are connected, one to the other, and the
microorganisms pass with the flow of the nutrient medium from one into the
other. In a system thus arranged it is possible to study the influence of different
degrees of nutrient exhaustion in the medium, as well as the influence of parti-
cular metabolites on the change of the physiological state of cultures. The
culture then contains several steady states at different levels. By changing
the number and size of flasks, and the rate of flow, their ratio can be affected
as desired, this makes it possible to study the conditions of origin of some
stages of development of microorganism, c. g. sporulation, etc. It also gives
us the possibility to study the influence of the physiological state of the culture
on the production of various metabolites, such as antibiotics, etc. Finally, by
using this method, we can study in detail the influence of various factors
on the physiological state of microorganisms by adding them to the second,
or some other flask where they are in contact with a fully multiplying culture.
This opens new approaches to the knowledge of the adaptive ability to various
factors. With actively multiplying cultures we may investigate their capacity
to respond more actively to environmental influences, and mutability caused
and influenced by external factors. Jerusalimskij Presents a mathematical
treatment of such a multi-stage system.
To Study the physiological state of microorganisms during continuous culture
in a multi-stage system we first chose Azotobacter chroococcum (ALMA (1952),
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
21
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Macura and Kotkovii (1953)). This microorganism undergoes ? when cultivated
statically ? a complicated form of development from rods to chroococcus bundles,
and from these to cystoid forms. In a number of experiments in a glass appar-
atus it could be observed that with a suitable and sufficient rate of flow
of the medium, ensuring fast multiplication, it was possible to maintain it
permanently in the form of motile rods with undifferentiated plasma, while
at a slower rate of flow the rode were shortened, encysted, and chroococcus
forms appeared. Particularly striking was the difference between microorgan-
iamii in individual stages. Even if the steady state at the various stages did
not differ very niiich as to the number of microorganisms there were striking
differences in shape: in the first stage there is a predominance of actively
multiplying rods with undifferentiated plasma, in the second and third stage
the chrooceceus forms predominate. This difference could be shifted by chan-
ging the rate offlow.
Our second model was Eacherichia coli and Salmonella enteritidia and we took
the sensitivity towards temperature and towards formaldehyde (Malek et al.
1953) as.the criteria of physiological state.
We-proceeded from the assumption that microorganisms during the stage
of "physiological youth" (Walker, Winslow et al. 1933, 1939) exhibit an
increased sensitivity to temperature, 111- ions, phenol etc. (Sherman and Albua
1923).together_ with an increased metabolic activity. This sensitivity enables
us to diatinguish ;clearly- between microorganisms from the beginning of the
exponential phase and those from.,the latter part of this phase, as well as from
the stationary phase. It has been shown experimentally that microorganisms
cultivated in a continuous culture approach in sensitivity those from the begin-
ning of the exponential phase. They are very much more sensitive than micro-
organisms obtained by means of a common static cultivation, and their sensiti-
vity decreases in further stages even if it remains well above the level of static
cultures. At, the same time it was interesting to-note that the culture always
contained a certain number of resistant microorganisms. These experiments
again .stippcirteck the. view that with continuous flow techniques such cultures
are develciped-rwhich in their physiological characteristics resemble
micro-
organisms,from the beginning of the exponentialphase.
'.;Further studiesir Were carried out using an aerobic sporulating bacilluirfrom
the-group-,4 aubtslsa (Sev6lk .(1952); Malek et al.- (1953)). These experiments
showed above all, that ,under the conditions of an aerated continuous culture
the iiiiciiiorganisins studied formed more robust reds or fibers that reproduced
i'regetaliyely;'williont:,_sporullation. Sev6ik observed that the capacity to form
- _ _ - , . _
'these rOliiiitrOilalvaii preserved by the bacilli even in the first static aubculture.
en there a sufficient source Of nitrogen present; together with a limited
_ _
source Of ;-car (glucose), so :that '1.11..caibeini Was -utilised during the _ _
e, 'duringth1e nex.t stages
sporulation, ias a- rule- took plaee. Aside- from
_
sporulation, we have studied the content of RNA and DNA, and the meta-
bolic activity. We observed that during the stage when these robust fibers
started to be formed, the content of RNA increased nearly to values at the
beginning of the exponential phase in a static culture, and remained there
for several days. During the second and third stage, after an analogous initial
phase, a decrease could be observed to values lower than in the first stage,
and an increase started only when sporulation set in. The metabolic activity,
as measured by the comsumption of 02 per nig of dry weight per hour, did
not produce unambiguous results, probably because a non-uniform cellular
development takes place. This inequality of individual cells in the population
was a new, important finding; even cells in individual fibers behaved differently.
There were fully live cells directly adjacent to dying ones. The same could
be observed in the setting-in of sporulation in individual cells of chains and
fibers. We are of the opinion that this inequality of individual cells in the
population during a continuous culture constitutes a fact that must be taken
into account much more seriously.
Further work concerning the physiology of microorganisms in continuous
fermentation was carried out using the yeast Saccharomyces cerevisiae, baker's
type. Because these experiments are to be discussed in another report of this
symposium (Beran), I shall limit myself only to the conclusions. As is well-
known, the growth of this yeast is a manifestation of mixed metabolism.
During cultivation a certain amount of alcohol is always formed, qualitatively
dependent on the rate of flow. If a high yield of yeast is desired it is necessary
to prevent as much as possible the formation of alcohol, which can be done
only by limiting the carbohydrate source. Furthermore, the fermentation is
connected with certain characteristics which are of importance for the quality
of the yeast, e. g. raising power in dough, and durability. We are dealing then
with a complicated system, in which the physiological state of the culture
is of importance for practical applications.
We carried out our experiments ? for practical reasons using molasses ? at
different rates of flow. We worked most often at a rate corresponding either
to a three-hour generation time ? which was found to be optimal for our
experimental purposes ? or a four-hour generation time which is most often
used in yeast manufacturing plants.
We tried to find an answer to the following questions: In which physiological
phase is a continuous flow culture in respect to its development in a batch
culture? To what extent do aerobic system s undergo changes caused by utili-
zation ofingars, ethanol and7acetic- acid, as well As anaerobic systems respons-
ible for anaerobic fermentation of glucose and maltose? For 'practical reasons
we also studied the autolytic rate of these cultures and the relation of data
obtained to the quality of iirodneed'yeast,in the raisingof dough'.
_
On the basis of the RNA content, QZ;r, values on ethanol and acetic acid,
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
23
Declassified in Part - Sanitized Copy Ap roved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
and Qical , values on maltose, we reached the conclusion that at a flow rate
corresponding to a 3-4 hour generation time, we obtain a continuous flow
culture corresponding to the terminal part of the exponential phase of the
growth curve.
The values of QV: on Maltose, glucose, ethanol and acetic acid did not
change for the whole duration of the experiment (120 hours) and did not
depend on the rate of flow. On the other hand Q. values on glucose were
dependent on the rate of flow. At a rate corresponding to the calculated 2,6-
hour generation time, the QZ6. value did not change, at a rate corresponding
to a three-hour generation time it fell, but remained at a rather high value.
At a four-hour generation time, however, a sharp drop could be observed
during the 96 hours of cultivation. We consider it important that for the entire
duration of experiments the CO, from maltose retains an adaptive character
under anaerobic conditions, which fact is in keeping with the physiological
state toward the end of the exponential phase of the growth curve in static
fermentations. We feel that this phenomenon can be of practical value. The
autolytic rate did not exhibit any characteristic changes during incubation.
On the other hand we observed a change in shape of the yeast cells, manifested
by an elongation without changes in volume, in the course of a short-term
continuous cultivation.
From these experiments it can be concluded that yeasts grown in continuous
cultures do not change in the course of the cultivation as to their aerobic
systems, and are in good physiological condition, corresponding to the terminal
stage of the exponential phase of the growth curve.
In orientation experiments we have shown that by a suitable modification
of the multi-stage construction, it is possible to reach a continuous formation
of such evolutionarily complex metabolites as penicilin Meek 1955).
In conclusion to these experiments, to a certain extent orientational
in character, it can be said that the question of physiological state of micro-
organisms in continuous cultures is a very important one, which deserves
considerable attention. With bacteria which undergo, in static cultures,
a morphologically demonstrable development (Azotobacter, sporulating
bacilli) it was shown that even in a continuous culture they can undergo an
analogous development when suitable conditions are preserved, particularly
at slower rates Of flow. The multi-stage continuous culture has proved to be
a convenient method not only for the study of the physiological state of con-
tinuous flow cultures, but also for other, questions. It will be necessary to
develop it both technically and experimentally to such perfection as has been
reached by the one-stage batch method of , the chemostat or turbidostat
type.
24
,
a -
_
SUMMARY
1. A review of results obtained in developing the continuous culture method
is given.
2. From an analysis of the world literature it has been concluded that not
enough attention is being given to the question of the physiological state
of microorganisms cultivated under the conditions of continuous flow culture,
although it is an important question both theoretically and practically.
3. The following important questions arise:
a) Does the continuous flow culture optimally ensure all the necessary condi-
tions of growth and development of microbial cultures?
b) Is it possible to draw a parallel between the physiological state of micro-
organism in a continuous culture and their development in static cultures?
c) In what range of flow rates does the physiological state of the culture
remain constant, and corresponds to the optimal state existing, as a rule,
at the beginning of static culture development?
4. We have mentioned possibilities that exist for the study of physiological
state in a multi-stage continuous flow culture, and have mentioned results
obtained by its application to the study of the development of Azdobacter
cultures, to the determination of sensitivity toward temperature and disin-
fectants with Eacherichia coli and Salmonella enteritidis, to the study of aerobic
bacillus sporulation, and to the production of penicillin. We have reached
the conclusion that the physiological state of continuous flow cultures of
bacteria corresponds, under our conditions, to static cultures at the beginning
of the exponential phase. We have called attention to the practical importance
of the study of the physiological state of baker's yeast grown in continuous
culture; in this case cultures have been compared rather with the terminal
part of the static exponential curve on the basis of their RNA content and
of the analysis of aerobic as well as anaerobic fermentation.
5. It is concluded that it is necessary to pay more attention to the physiolo-
gical state of continuous flow cultures, particularly from the methodological
and experimental point of view.
REFERENCES
ANDERSON P. A., 1953: Automatic Recording of the Growth Rates of Continuously Cul-
tured Microorganisms. J. of Gen. Phys., 36: 733.
BRYSON V., 1952: The Turbidostatic Selector ? a Device for Automatic Isolation of
Bacterial Variants. Science 116: 48.
BRYSON V., 1953: Applications of the Turbidistat to Microbiological Problems, I: 396.
CASTOR J. G. B., 1947: Apparatus for Continuous Yeast Culture, Science, 166: 23.
CLEARY'I. P., BEARD P. J., Curros C. E., 1935: Studies of Certain Factors Influencing
- the Size of Bacterial Populations. J. of Bact., 29: 205.
npriaccifien in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
25
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
DAviss A., 1956: Invertase Formation in Saccharomyces Iragilia. J. gen. Mier., 14: 109.
DENBIGH K. G., 1944: Velocity and Yield in continuous Reaction Systems. Trans. Fara-
day Soc., _40: 352.
DucatI., Natr J., 1950: Cultures des microorganisrnes en milieu confirm. Acad. des
&Wires,
ELSWOWPWR., MAKIN L. R. P., PIRT S. J., CAPELL G. H., 1956: A Two Litre Scale Con-
tititious Culture Apparatus for Microorganisms. J. of Applied Bact., 19: 264.
Foinstbeliifttingsverfahren fiir die Giirungsindustrie. Lupinges. Mannsheim, Pat. L. 10,
No 10.
FORMAL S. B., BARON L. S., Srlumbr IV., 1956: The Virulence and Immunogenicity of
typhOso Grown in Continuous Culture. J. of Bact., 72: 168.
Gisalf.tabi P., 1946: Brucella suit in Aerated Broth Cultures. Continuous Culture Stu-
dies. J. of Bact., 52: 283.
GRAZIOSI F., 1956:- Metodo per hi coltura continua dei batten i mediante un apparato
turbid istatico. Giorri. di Microb., 1: 491.
GRAZIOSI F., 1957: Studio quantitative dell'adattamento di Micrococcus pyogenes e Prokus
vulgaris cilia tiovebiocina. Rend. dell'Ac. Naz dei Lincei ser. VIII, vol. XXII: 3.
HARRIS E. E., SALMAN I. P., MARQUARETT R. R., HANNAN M. L., ROGERS S. C., 1948:
Fermentation of Wood Hydrolysates by Torula utilia. Ind. Eng. Chem., 40: 1216.
HARRIS T. N.a, 1956i Growth of 'Group a Hemolytic Streptococcua in the Steady State, J. of
Bact., 72: 283.
HADiN C. G.,.Horanc.T.,MALIMIGREN B., 1955: An Improved Method for the Cultivation
of tho Microorganisms by the Continuous Technique. Acta path. et micr. Scand.,
'37: 42.
HERBERT D., ELSWORTH R., TELLING R. C., 1956: Tho Continuous Culture of Baaeria;
TheoWtical and Experithental Study. J. gen. Mier., 14: 601.
iforersti B:, Hoisriru A., FRIES N., 1953: Tho Technique of Continuous Culture as
Applied,to a Fungus Opliioatoma multiannulatum. I: 406.
JERUSALIMSKIJ. N. D., 1952: Fiziologija razvitija 6istych bakterialnych kultur (Auto-
referate) _MoskVa.
JERUSALIMSKIJ N. D., RUKINA E. A., 1956: Issledovanije uslovij sporoobrazovanija
u-masljanokislych Vakterij s pomoil6m kolloidnych gilTz. Mikrob. (Russian) 25: 649.
JORDAN R C., J,te-Oas S. E., 1944: The Growth of Bacteria with a Constant Supply. I.
Preliminary Observations of Bact. coli. J. of Bact., 48: 579. II. The Effect of Tem-
perature ..., J. Gen. Micr., 1: 121, 1947. III. Tho Effect of pH at Different Temper-
, &tures ... J. Gen, -Mier., 2: 150948.
,I4glits T. N., 1956:, Growth of Group A-hemolytic Strepto-
. -emetic, in the Steady State. J. of Bact., 72: 283.
KUBFFSCHBK IL E. 1954: Modifications of the Chemostat. J. of Bact., 17: 254.
Liz HH., 1953: The Mutation of B. coli to Resistance,to BacteriophageTb.Arch:-Bioch.
- et Biophys.; 47i, 438.,. dle NpVicka-1955.'
_ _
MA:CURA KOTKOVI M, 1953: lUvoj azotobaktera v proudicim prostfedi. NI. biol., 2: 41.
MILK I, 1943 Pstov? mikrobet s prOudicith prostfedi, as lek-.6es., 82: 576.
MAtiic I..; 1910: Mn'Of.eni E coli-y pro?dicim proettedl Biol list.Y- 31: 93.- '
'MALIIK I, 1952 Kultivace bakterii ye;vicestupliovenf pioudiciM prOstledi esl. biol., I:
.,MAisx,I-.; 1952: Kiiitivacaazotobakterav pioudicim prostfedi. NI. biol. I: 91.
I'414.1eic ? I.; -- 1953: Sporulation Of, Bacilli, VI congressO' interriazionale dCrniciObiologia;
_ - ? - ,
Roma; 1: 345:','
Miutt I, 1955 0 mnOleni a p?stovnf rnikroorganisiml zvl?bakteriLINC'S,AV, Praha.
MALE): I., 1956: Proto6nyj metod razinnolenija rnikrobov. Mikrobiologija (Ilus.), 25: 659.
MALE): I., 13uHoEit M., liEJmovA L, kil:IcA J., FEscL Z.. BERAN K., 1957: Asimilace
cukrti a kyselin pti kontinu?im zclroidovAni netediWeli still. vyitihu. Cl. mikrob.,
2: 203.
MALEK J., VosvicovA L. a spol., 1953: Resistenee bakterii p6stov1iiVeli v proudiciin
prosttedi. 'Cs'. biol. 2: 68.
MALEK I., CHALOUTKA J., VosvizovA L., 1953: Sporulaco baeili, Csl. biol. 2: 323.
MALEK I., VosvKovA L., 1954: Priitokovii kultivaco kvasinek. Cis!. biol.
3: 261.
MALMGREN B., Hi:DEN C. 0., 1947: Studies of the Nucleotide Metabolism of Bacteria
I-X, Acta Path. et Mier. Scand., XXIV, 412.
MALmoREN B., Hignini C. G., 1952: Studies on the Cultivation of Micro-organisms on
a Semi-industrial Scale. I. General Aspects of the Problem. Acta path. microb. Scand.,
30: 223.
MAxoN IV. D., 1955: Continuous Fermentation. A Discussion of its Principles and Applica-
tion. J. of Appl. Microbiol., 3: 110.
MONOD 1950: La technique de culture continue. Tlikirie et applications. Ann. Inst.
Past., 79: 390.
MoYEa H. V., 1929: A Continuous Method of Culturing Bacteria for Chemical Study. J.
of Bact. 18: 59.
NORTHROP J. H., 1954: Apparatus for Maintaining Bacterial CuILmir-;p. he tcady State.
J. Gen. Phys., 38: 105.
NOVICK A., 1955: Growth of Bacteria, Ann. Rev. of Microbiol., 9: 97.
NOVICK A., SZILARD L., 1950: Experiment with the Chomostat on Spontaneous Mutation
of Bacteria. Proc. Nat. Acad. Sci., Wash., 36: 708.
NOVICK A., SZILARD L., 1951: Genetic Mechanisms in Bacterial Viruses. Experiments
on Spontaneous and Chemically Induced Mutations of Bacteria Growing in Chemos-
tat. Cold Spring Harbour Symposia Quant. Biol., 16: 337.
NovoGauDsKLI D. M., 1950: K voprosu o vnutrividovych i meividovych vzaimootno-
Aenijach po6vennych mikroorganizmov. Agrobiologija 48: 5.
.PASYNSKIJ A. G., 1957: Teorija otkrytych sistcm i jejo znaZenije dlja biochemii. Usp.
Sovr. Biol., 43: 263.
Patent Norddeutsche Hefeind. St. 48881, K1. 66 16/01, 1934.
PERRET C. J., 1957: An Apparatus for the Continuous Culture of Bacteria at Constant
Population Density. J. gen. Mier., 16: 250.
PIRFILEV, Leningrad: Personal Information.
PIRT S. J., 1957: Tho Oxygen Requirement of Growing Cultures of a Bacterium Species
Determined by Means of the Continuous Culture Technique. J. gen. Microb., 16: 59.
POWELL E. 0., 1956: Growth Rate and Generation Time of Bacteria, with Special Refer-
ence to Continuous Culture. J. Gen. Mierob., 15: 492.
ROGERS L. A., WiirrnEa E. 0., 1930: Tho Growth of Bacteria in Continuous Flow of
Broth. J. of Bact., 20: 127.
ROSENBERG M., 1956: Dynamics of the Breaking-down of Lysogenic Cells Irradiated by
Ultra-violet Light. Fol. biol., Praha 2: 206.
SARKOV I., 1950: ?didroliznoje proii-vodstvo. III. Goslesbumizdat.
SEVAGE_M, C.; Florey H. W, 1950: Induced Bacterial Antagonism. Brit. 3. of Exp. Path.,
31: -17.
8Ev6IFV1;.; 1952: TVorbavelkYch forem u Bac. subtilis v proud icim prostfecli. esl. biol., 1: 93.
SHERMANN I. M:,-ALaus W. R., 1924: The Function of Lag in Bacterial Cultures. J. of
--Bact, 9.' 303.-
_
fla
27
Declassified in Part - Sanitized Copy Approved for Release
_
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
avAcittn.ovA J., KuAKA J., 1956: Studium metabolismu mykobakterii v proudicim pro-
attedi. Rozhl. v tb., 16: 488-491.
TAMTIA IL, 1957: Maas Culture of Algae. Ann. Rev. of Pl. Phys., 8: 309.
UNGER E. D., STARK W. H., SCALY R. E., KOLACROV P. J., 1942: Continuous Aerobic
Process for Distiller's Yeast Ind. Eng. Chem., 34: 1402.
UTENKOV M. D., 1941: Mikrogenerirovanio. Cos. izdat. ?Soy. nauka", Moscow.
VALYI-NAGY, T., CSOBAN C., ZABOS P., 1954: Effect of Penicillin on the Nucleid Acid
Metabolism of Staphylococcus aureus. Acta Microbiol. Acad. Sci. Hung, 2: 79-89.
VIvitA J., 1958: Zaiizeni pro preitokovou kultivaci prvokii. csl. biol., In press.
VERBINA N. M., 1955: Priti6enije drolioj k antiseptikam raz1i6nyini motodami. Trudy
Inst. mikrob., IV: 54.
VICTORERD F. A., 1948: Apparatus for Continuous Fermentation. US Patent, 2, 450, 218.
VOLODIN A. P., 1952: K voprosu o prirodo bakterialnych kolonij, Agrobiologija 2: 138.
WiNsLow C. E. A., WALKER H. IL, 1939: The Earlier Phases of the Bacterial Cultures
Cycle. Bact. Rev., 3: 147.
Um= M. R., 1955. Genetics of Microorganisms. Ann. Rev. Microbiol., 9: 1-20.
28
GENETIC AND PHYSIOLOGICAL STUDIES
WITH THE CHEMOSTAT*)
AARON NoVICK
It is exciting to participate in this Symposium on the Continuous Cultivation
of Microorganisms because it provides a unique opportunity for the exchange
of ideas and information workers from many countries. Such exchanges, be-
sides personally rewarding, should help increase the rate of development and
the applications of our science.
The introduction of new experimental techniques has played an important
role in the history of many sciences. The introduction of agar medium for the
cultivation of bacteria on a fixed surface made possible the development of
modern microbiology. Not only did it permit the convenient isolation of pure
cultures of bacteria, but it also permitted the simultaneous examination of
many different bacteria in a single dish. The great rewards resulting from the
introduction of the agar plate offer encouragement to examine other techniques
for the study of micro-organisms. It is hoped that this Symposium will call
wide spread attention to the advantages of continuous culture methods for
growing bacteria. These methods permit experimentation in areas hitherto
not conveniently accessible. This paper will attempt to illustrate some of these
advantages in the study of bacterial genetics and physiology.
I would like in this paper to review the work done at the University of
Chicago mostly by Professor Leo Szilard and myself on the development and
application of techniques of continuous culture. After a discussion of the prin-
ciples and techniques employed I will illustrate the kinds of experiments that
can be performed with such methods.
9 This investigation was suppported-in part by a research grant (E960) from the National
Microbiological Institute, U. S. Public Health Service, and in part by a research grant from the
National Science Foundation (U. S.).
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
29
STAT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
-
PRINCIPLES AND TECHNIQUES
Although microbiologists have employed continuous flow methods for the
culturing of bacteria for some years (Novick, 1955) the useful application of
such methods was made possible by the development of an understanding
of the principles underlying such techniques, the theory being described simul-
taneously by Monod (1950) and by Novick and Szilard (1950a, b).
Typically a continuous culture apparatus consists of a growth tube in which
a growing bacterial population is maintained at constant size by continuous
dilution. The dilution is accomplished by the admission to the growth tube
of sterile nutrient liquid and the removal of bacterial suspension from the growth
tube by means of an overflow that is set to keep the volume in the growth
tube constant. In such a system, the number, N, of bacteria per ml will change
with time at a rate, dN ' given by
dt
dN
dt
aN ? ?VN
(1)
where a is the growth rate constant, w is the flow rate (ml/hr), and V is the
1
growth volume (m1). The generation time, T, is equal to ? . For the size of the
a
dN
population in the growth tube to remain oonstant, ?dt must equal zero, i. e.,
a must equal w/V.
Equality of growth rate and dilution rate can be established in two funda-
mentally, different ways. In one case the bacteria grow at a fixed rate which
is determine&by the choice of medium and temperature. The dilution rate is
adjusted to equal this growth rate by employing some mechanism, e. g. a light
source and photocell, to, observe the population size and to increase the flow
whenever,the, population density exceeds a chosen value, N., and to decrease
the flow when the density is less than No. As a result, the dilution rate will equal
the bacterial growth rate, and measurement of the dilution rate gives the value
of.the'grciiit_ brate constant. In such a case?the growth rate constant is probably
determined by? the low concentration of some substance made by the bacterium
, -
at the lowest ratiy,relative to' theother substances made by the cell. Since the
_ - _
growth tate- here is dependent on the concentration of a nutrilite supplied
by synthesis Within the cell_ the bacteria. under these conditions can be said
tO be'. under internal 'control (Novick, 1954
_ _
Suchan apparatus _was ' constructed and successfully operated. in Chicago
yy_Fof.and_Szilard (1955), and by a number of Other workers elsewhere. All
of these internally controlled systems, hoWever,.have_shown 'a serious technical
hindieiip. -After' the: syideM has been: in operation for a time, a large number
of. gro-ibacteria are fo?ndth' adhere to _the-Wallsief the grOWth tubTe:-.Th. is
? -
=
_
"wall growth", being inefficiently diluted by the flow system, increases in size
and begins to supply large numbers of bacteria to the medium. As a result
there is an apparent dramatic increase in growth rate because the number
of bacteria in the growth tube is increased, not only by the reproduction of
the bacteria suspended in the liquid, but also by the additional supply sloughed
off from the bacteria growing on the walls. Despite elaborate measures taken
A
0.8
0.6
0.4
0.2
^
1
0
2
3 4
5
Fig. I. Growth rate of a tryptophan-requiring strain of E. coli (B/14) as a function of the con-
centration of tryptophan.
A ? Growth rate, a (hr-4); B ? Tryptophan concentration (y/1).
to prevent wall growth, it has not been eliminated for long. Thus internally
controlled systems can be operated for only relatively short periods of, time.
A second method for establishing the equality of dilution rate and growth
rate is found in the principle employed in the chemostat (Novick, Szilard,
1950a, b) and in the bactogen (1950). The present discussion refers only to the
chemostat and experiments performed with it. All of the remarks concerning
principles of operation hold equally for the bactogen. These systems, based
on a principle of external control, operate in the following way. The dilution
rate is set at some value less than the maximum growth rate, and the nutrient
liquid is composed of a large excess of all required nutrilites but one, called
the controlling growth factor. The density of bacteria increases with time as
long as the growth rate exceeds the dilution rate. However, an increase in
number of bacteria means a descrease in the concentration of the controlling
growth factor in the' growth tube. As the concentration of the controlling growth
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
31
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
factor continues to fall, it eventually will cause a decrease in the bacterial
growth rate. It can be shown that the growth rate will fall until it becomes
exactly equal to the dilution rate. As a result, the density of bacteria in the
growth tube remains constant from then on.
An example of the dependence of growth rate on the concentration of a re-
quired growth factor can be seen in Figure 1. Here we sect that, for a trypto-
phan requiring mutant (B/1,t) of the B strain of E. coli, at higher concentra-
tions of tryptophan the growth rate is independent of the concentration
of tryptophan in the nutrient medium and at lower concentrations the growth
rate depends on the concentration of tryptophan. One can expect, therefore,
that in the chemostat or bactogen the rising density of bacteria must eventually
reduce the concentration of tryptophan in the growth tube to such low values
as to cause a reduction in the bacterial growth rate.
Since the bacterial growth rate in the chemostat is a function of the con-
centration of the controlling growth factor in the growth tube,equation 1 should
be rewritten as
dN
ta(c)N ? N
d
In the steady state that results, we have
d/V
= 0 and a(c) -p-
dt
(2)
(3)
The system is evidently self-stabilizing since any perturbation, (e. g., too high
or low a value of N) causes a change which reduces the perturbation.
Moreover, it should be noted that since the growth rate constant a(c) is
a function only of ? the dilution rate, the concentration of the controlling
V
growth factor in the growth tube in the steady state is determined only by the
dilution rate. That is, the concentration of the controlling growth factor
in the growth tube in the steady-state is independent of its concentration in the
incoming nutrient.
The density, N, of bacteria maintained in the growth tube in the steady
state is given by
N=Q
a ? c
(4)
where a is the input concentration of the controlling growth factor and Q
is the yield constint (the amount of the factor required for one bacterium).
This equation assumes that the yield constant, Q, is independent of c, the
concentration of the controlling growth factor in the medium. Often c is found ?
,
a
to be small compared to a, and N is well approximated by (To .
32
In these systems, therefore, the bacterial growth rate is determined by the
concentration c of an externally supplied nutrient. For this reason the chemo-
stat can be said to employ external control. For the controlling growth factor,
any one of a wide variety of required nutrient can be used.
An interesting question concerns the range of growth rates which can be
employed with the chemostat. Obviously the maximum growth rate is that
observed at high concentration of the controlling factor. Little information,
however, is available regarding how slowly bacteria may be grown. Experi-
ments with a tryptophan-requiring mutant (B/1,t) of E. coli indicate that
a lower limit is reached at generation time of 15 hours at 37?C. Lower growth
rates at this temperature do not seem possible since at lower dilution rates
the population seems not to grow a certain fraction of the time. Probably
bacteria cease growing and go into the lag phase, or sporulate if they can,
whenever the growth rate is less than some limiting value. If the dilution
rate is too low, the growth tube contains growing bacteria part of the time
and non-growing bacteria part of the time. Nevertheless, it is important to
note that there is a wide range over which bacteria can regulate their growth
rates. Apparently they are able to regulate all of their metabolic processes
in this range so that they do not make a large excess of any amino acid (Novick,
Szilard, 1954).
APPLICATIONS
Although the present discussion is concerned with the application of conti-
nuous culture methods only for experimental purposes, it should be noted
that such methods offer great technological promise. Clearly, a continuous
process has a higher production rate than does a batch process since it operates
all of the time at maximum population levels. Another advantage, which
should prove to be highly important, is that a continuous variation in the
physiological state can be produced by variation of the growth rate. Quite
likely there are many technologically important products made by micro-
organisms (e. g., antibiotics, amino-acids, vitamins, etc.,) which will be made
in some cases at higher rates at lower growth rates. In this event, both the
rate of production and the concentration of the product can be increased by
selection of the growth rate optimum for these purposes.
For experimental purposes, the principle advantage of an apparatus such
as the chemostat is that a population of bacteria of chosen fixed sic can be
grown under constant conditions for many generations at a selected growth
rate. Not only are the bacteria growing in some well-defined state, but also
the concentrations of all chemical substances remain constant. Even very
low concentrations of a chemical substance can be maintained constant for
long periods of time.
3?Symposlum
33
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
The first group of studies presented below takes advantages of the fact
that the chemostat permits the study of a population for many generations,
and includes studies on rates of mutation from observations of the accumula-
tion of mutants over many generations, on "evolutionary" changes in bacterial
populations, and on the slow adaptation of a bacterial population at low
concentrations of a specific enzyme inducer. The second group of studies is
based on the fact that the chemostat permits experiments at exceedingly low,
yet constant, concentrations of a nutrilite. Here the experiments are concerned
with the study of the regulation by bacteria of the rate at which certain amino
acids are made.
MUTATION STUDIES
The chemostat offers a method for the accurate measurement of mutation
rates in bacteria, whereas the usual methods encounter statistical difficulties
leading to serious lack of precision. The results of a series of studies (Novick,
Szilard, 1950b; 1951; Novick, 1956; Lee, 1953) of spontaneous and induced
mutations in the B strain of E. coil are summarized below.
The principle employed to measure mutation rates with the chemostat
is based on the fact that under certain conditions the fraction of mutants
in a population, or the number per ml in the chemostat, rises linearly with
a slope that is determined by the mutation rate. If the number per ml in the
chemostat of a given mutant is m, if back mutations can be neglected, and if
is small compared to N, the total number of bacteria per ml, the rate of change
in m with time is given by
dm
= AaN ? m(a ? am) (5)
di
where am is the growth rate of the mutant, a the growth rate of the wild type
equals the flow rate, and A is the rate of mutation expressed per bacterium
per generation. If, a = a?, then
and
dm
N
dt 2a
vi = mo AaNt
(6)
(7)
Therefore, under the assumptions made, the number per ml of a given mutant
rises linearly with time with a slope equal to AaN, thus permitting an accurate
measure of A. Conversely, whenever a linear rise in the number of a given
mutant per ml, is observed, it may be concluded that the assumptions above
are Valid and that the slope gives the correct value of the mutation rate.
It is obvious that this method is not disturbed by such things as "plate muta-
34
tions" (mutations which occur during the plating of an aliquot of the popula-
tion to measure in); since the plate mutants are constant in number they do not
affect the slope of the rise in number of mutants. Likewise, it is certain selection
for or against the mutant plays no role since the linear rise shows that a =a,,,
within experimental error.
If a should not equal am, i. e., should there be selection for or against the
mutant, no linear rise will be found. Should an, > a, the number of mutants
will rise exponentially .and eventually displace the wild-type organisms.
Should am < a, the number of mutants will rise toward a limiting value
given by
M
a
?
a
(8)
This method of measuring mutation rates was used for several imitations
in the B strain of E. coli, where linear rises were observed for mutations to
resistance to phage T5 and to phage TG. It was found for these mutations that
over a wide range of generation times the rate of mutation is a constant per
unit time, i. e., proportional to the generation time (Novick, Szilard, 1950b;
Lee, 1953). If the mutation rate per hour is given by i, then A = IIT. Moreover,
it was observed that the rate of mutation depends on the controlling growth
factor used. For instance, control with an amino acid gives higher rates than
does control with the source of energy, or nitrogen, or phosphorus.
Since the chemostat can be used for accurate measurement of mutation
rates, a research program was undertaken to examine the mutagenicity of
a wide range of chemical substances. It was hoped that the precision of the
method would permit us to test the mutagenicity of compounds that might
be only slightly mutagenic but yet would have more specific chemical proper-
ties than most of the reagents and conditions customarily employed. A survey
was made of a wide range of chemical (Novick, Szilard, 1951; Novick, I 956).
A large number of purines and purine analogs were found to be mutagenic,
the methylated xanthines being the most active. Caffeine, (trimethylxanthine)
at a concentration of 150 mg/1 raises the rate of mutation about I 2-fold
to T5 resistance and three fold to TO resistance. Even the normally occuring
base adenine is slightly mutagenic, although none of the pyrimidines or py-
rimidine analogs studies were found to be effective.
Certain of the purine ribosides were found to .bestiongly anti-mutagenic,
i. e., they antagonized the mutagenicity of the purinc mutagens. Fairly low
concentrations of adenosine completely suppress the effect of much higher
concentrations of caffeine. In addition, it was found that these nucleosides
reduce the spontaneous rate of mutation to about one third of the usual value.
None of these anti-mutagens, however, have any effect on mutagencsis by
30
35
nprdassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
..
axe
ultra-violet light or gamma radiation. Curiously enough, the desoxyribosides
are much less effective anti-mutagens than the ribosides.
By providing a method for the accurate measurement of rates of mutation,
the chemostat establishes a beginning for the study of the biochemistry of
mutation. It is not yet understood how caffeine and adenosine produce their
contrary effects; but perhaps further examination with the chemostat of
compounds connected with nucleic acid metabolism, in the light of increased
knowledge of this metabolism, will provide an insight into these problems and
thereby furnish a better understading of the functioning of the genetic sub-
stance.
BACTERIAL EVOLUTION
When a chemostat is operated for many generations the initial strain of
bacteria eventually becomes displaced by a strain which can grow more
rapidly than the first strain at the low concentrations of the controlling growth
factor which occur in the growth tube of the chemostat. This transition can
be detected by observing the change in number of a given mutant the (115-
was... resistant, for example). When the initial strain is displaced by the faster
growing strain, all of the T5-resistant mutants that had been accumulating
in the initial strain are also displaced. This produces a discontinuity in the
linear rise of TS-resistant mutants in the initial population. Once the faster
strain is established in the growth tube and if the rate of mutation to T5-
resistance is the same as in the initial strain, a linear rise in the number of
Tb-resistant mutants should again be observed. The total population of bacteria
is unchanged by such a transition from a slower to a faster growing strain,
as long as the quantity of tryptophan (Q) required per bacterium remains
unchanged. Therefore, observation of the number of T5-resistant mutants
provides a way of detecting Such population changes when observation of the
total population give no indication.
Operation of a chemostat for more than five hundred generations led to the
detection of some ten or eleven transitions from slower-growing strains to
faster strains. In each case the advantage in growth rate is observed only at
low concentrations of the controlling growth factor. At high concentrations
there is no such difference in growth rates; if anything, the later strains grow
more slowly.
ENZYME INDUCTION
There are some enzymes, called inducible enzymes, which are made by
bacteria Only in the preseeice of certain specific inducers in the medium. For
- example, E o/i bacteria form /3-gilactosidase only when certain galactosides
36
Declassified in Part - Sanitized Copy Approved for Release
a
are present in the medium. The kinetics of formation of this enzyme are readily
observable with the chemostat since it can be used for experimentation under
constant conditions for extended periods of time.
No difficulty was encountered with conventional methods in studying the
kinetics of induction at high concentrations of inducer where, within a very
short time following the addition of inducer to the medium, all of the bacteria
form enzyme at maximum rate. At low concentrations of inducer, however,
the rate of enzyme formation per bacterium rises slowly for many generations.
Explanation of the nature of this rise was made possible by using the chemostat
(Novick, Weiner, 1957).
Typically an induction experiment is performed with the chemostat by
adding inducer in the desired concentration to the growth tube and to the
reservoir at a specified time. From then on, of course, the concentration of
inducer in the growth tube is automatically maintained at the desired value.
It was found that at low concentrations of inducer, starting from the time
of addition of inducer, there is a linear rise in the rate at which ?galactosidase
is formed, the slope of the rise depending on the concentration of inducer. The
rise in the rate of enzyme synthesis continues up to some limiting value. This
value, called intermediate saturation, for lower values of the concentrations
of inducer, is less than the maximum found at high concentrations; but it beco-
mes equal to the maximum value at higher values of the inducer concentration.
An explanation of these linear rises was found in the discovery that at low
concentrations of inducer the bacterial population is heterogeneous with respect
to the rate at which individual bacteria form 13-galactosidase. In fact, the popu-
lation consists of two kinds of cells, uninduced bacteria which make essentially
no enzyme and fully induced bacteria which make enzyme at the maximum
rate. At these and at higher concentrations of inducer all of the progeny of
induced cells are also induced. The linear rise in rate of synthesis of p-galacto-
sidase results from a linear increase in the fraction of the population in the
induced state, the transition from the uninduced to the induced state occurring
at a constant rate.
We can see here a resemblance between enzyme induction and mutation.
In both cases transitions occur from one state to another at constant rate;
and in both cases, selection being negligible, there is a linear rise in the fraction
of the population in the altered state. The resemblance to mutation was made
more striking when it was shown that the phenomenon of intermediate satura-
tion resembles the phenomenon of selection equilibrium in mutation. Selection
equilibrium occurs when the mutant type grows more slowly than the wild
type; as a result, the fraction of mutants in the population rises to a con-
stant value. Likewise, in the case of enzyme induction, intermediate saturation
represents a steady state where a constant fraction of the population is indu-
ced because of the fact that induced bacteria grow more slowly than uninduced.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
37
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
The existence at low concentrations of inducer of two possible states of
induction (uninduced or fully induced) and of the inheritability of the state
of induction can be explained on the basis of the presence in the bacteria of
an inducible specific transport mechanism, called "permease" by its discoverers
(Rickenberg et al., 1956) for the concentration of inducer inside the bacteria.
The transition from the inunduced state to the induced could result from the
chance appearance in a bacterium of a treshold level of permease. Upon cell
division both daughters would obtain sufficient permease to remain induced.
The growth rate difference between uninduced and induced bacteria could
be explained, among other ways, by the extra work the concentrating cell
must perform.
The use of the chemostat in physiological experimentation led to the disco-
very of the effect of carbon dioxide on the induction process. It was observed
that the rate of rise of induced cells in the population increases at higher bacte-
rial densities, and it could be shown that this result from the-higher concentra-
tion of carbon dioxide in the medium at higher bacterial densities. The mecha-
nism of action of carbon dioxide is still unknown, but the discovery of the effect
followed directly from the use of the chemostat. Moreover, the effect of rising
carbon dioxide concentration with rising bacterial density obscures in con-
ventional growth apparatus the simple kinetics observable at constant bacte-
rial density in the chemostat.
REGULATORY MECHANISMS
The chemostat is particularly useful for the study of the regulation of the
rate at which a bacterium synthesizes an amino acid. Apparantly, the presence
of an amino acid in the medium suppresses the synthesis by the bacteria of
this amino acid. This suppression seems to occur even at very low concentra-
tions of the amino acid, thereby making it difficult to investigate with con-
ventional methods the relationship between the concentration of the amino
acid and its rate of synthesis. However, such studies become easily possible
with the chemostat because one can maintain a low concentration of an amino
acid at a constant value in the presence of a large bacterial population.
In one case (Novick, Szilard, 1954) a study was made of the formation of
a compound (now thought to be indo1-3-glycerol phosphate; Yanofsky, 1957) pre-
sumde to be a tryptophan precursor, by a coli mutant (B/1,t) unable to synthesize
tryptophan. This compound is not usually formed by these bacteria when they
are grown in the presence of tryptophan, but is made only when the bacteria
are grown at very low concentrations of tryptophan, such as occur at long
generation times with tryptophan control in the chemostat. It was observed
that at generation -times longer than three hours the rate of production per
hinii is at a constant maximum value. As a result, the concentration of the
38
5
compound in the medium is proportional to the generation Mine in the range
of three to twelve hours. At twelve-hours generation time the concentration
of the compound in the growth is over one hundred times that in the incoming
nutrient liquid. At generation times shorter than three hours there is a decrease
in the rate of production of the compound. It could be shown that this decrease
is caused by the higher tryptophan concentrations in the medium at shorter
generation times. Moreover, it was possible to demonstrate that, when the
bacteria are making this compound at the maximum rate, suddenly raising
the tryptophan concentration leads to an immediate cessation in the rate of
compound production. It was also observed that if the bacteria that are
producing the compound are suddenly deprived of all tryptophan (stopping the
flow of incoming nutrient in the chemostat) so that protein synthesis stops,
Production of the compound continues at the maximum rate for at least four
hours.
The regulation of the rate of synthesis of an amino acid was further illustrated
by a series of experiments designed to examine the effect of exogenously
supplied arginine on the rate at which bacteria make arginine. In these experi-
ments two strains of bacteria were employed, one unable to synthesize trypto-
phan (B/1,t) and one unable to synthesize arginine (D84/6). Since the first
is resistant to phage T1 and the second to phage Ta, the numbers of either in
a mixture could be measured by plating with phage T1 or with phage T6.
For convenience, the concentrations of arginine and tryptophan in the
medilum are sometimes given in units proportional to the relative amount
of each of these amino acids in the bacterial protein. One unit of concentration
of tryptophan is taken as 500 yil, a concentration in the incoming medium
which supports a density of 2.5 x 108 of the B/1,t bacteria per ml in the chemo-
stat. Since the ratio of arginine to tryptophan in these bacteria is 4.6, one
unit of concentration of arginine equals 2300 y/l. This concentration of arginine
in the incoming nutrient in a chemostat would support a density of 2.5 x 108
of the D84/6 bacteria per ml.
The first experiment demonstrated that not only can B/1,t bacteria assimilate
arginine at low concentrations of arginine but also that these bacteria can take
arginine at concentrations too low to support the arginine-requiring strain,
D84/6, at a high growth rate. In this experiment a chemostat containing one
unit of tryptophan and one-fifth unit of arginine in the incoming nutrient
was inoculated with both the Wit and D84/6 strains. If each consumed only
its required amino acid and none of that required by the other strain, the
population density would be 3.0 x 108 (1.2 x 2.5 x 108). However, if one
strain consumed the amino acid required by the other, the density would be
less; and if one strain made an excess of the amino acid and secreted it, the
density would be higher.
When an experiment of this kind was performed at a generation time of two
39
npr.lassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Co .y Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
hours, it was observed that the density became constant at 2.5 X 10' and that
the D84/6 bacteria did not remain in the growth tube. This shows that the
B/1,t bacteria assimilate arginine so efficiently that the concentration of argi-
20
40
60
Fig. 2. The strain B/1,t was grown at ? generation time of eight hours at a density of 2.5 x 101
with tryptophan control in four ?hemostats. Each had in addition the indicated concentration
of arginine in the incoming nutrient, representing the amount given in parentheses of the arginine
in this number of B/1,t-bacteria. At zero time the oltemostate were inoculated with the arginine
requiring strain (D84/8), and the number of these bacteria were measured at various later time
by plating with excess phage T8. T = 8 hrs.
A ? Argizaneless bacteria per oc;/0 ? Hours; 1 ? 1150 y/1 arginine (0.5); 2 ? 690 y/1 arginine
(0.3); 3 ? 230 y/1 arginine (04); 4 ? 0 y/1 arginine (0).
nine in the growth tube falls below that required to maintain the D84/6
bacteria growing at a generation time of two hours. The B/1,t bacteria obtain
one-fifth Of their arginine from, the mectipm and synthesize four-fifths. In order
to do this they need to assimilate arginine from the medium only one-fifth
as fast as the D84/6 bacteria, Since the D84/6 bacteria mint secure all of their
arginine from the medium. Thus if the B/1,t bacteria take up arginine at a given
40?
concentration more rapidly than one fifth the rate that the D84/6 bacteria
take it up, the concentration of arginine will fall below that required for the
same growth rate by the 1)84/6 bacteria. In conclusion, this experiment
demonstrates not only that bacteria can efficiently assimilate arginine at low
concentrations, but also that a low concentration of arginine in the medium
is sufficient to reduce by twenty per cent the rate at which the bacteria make
arginine.
A
0.6
0.4
0.2
2
1
3
Fig. 3. Growth rate of an arginine?requiring strain of E. eoli, (D84/6) as a function of the
concentration of arginine.
A ? Growth rate (hr.-1); B ? Arginine concentration (y/1).
When experiments of this kind were done at longer generation times it was
found that the 1)84/6 bacteria would remain in the growth tube, the total
density of B/1,t approaching 3.0 x 10'. The fact that the D84/6 is able to
compete successfully with the B/1,t bacteria for arginine at longer generation
times can be understood in the-following way. The steady state concentration
of arginine established by the B/1,t bacteria is that which allows the bacteria
to make eighty per cent of their arginine per generation. At longer generation
times the rate per hour at which arginine is made by the B/1,t bacteria decreases,
and quite possibly this decrease in the rate of arginine synthesis corresponds
to a rise in the concentration of arginine in the medium. Furthermore, at longer
generation times a lower concentration of arginine is sufficient to support
growth of 1)84/6 bacteria.
Experiments like that above can be used to establish the relationship between
the rate of synthesis of arginine by the bacteria and the concentration of argi-
4 1
narlaccifiPrl in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
nine in the medium. The results of a series of experiments demonstrating such
an analysis are given in Figure 2: Here a series of chemostats was set up to as
to contain in the incoming nutrient the following concentrations of tryptophan
and arginine. Each chemostat contained one unit of tryptophan, but the argi-
nine content varied as follows: Case A, 0.5 units of arginine; Case B, 0.3 units;
A
1.0
0.8
0.6
0 4
02
1 1
1
8 02
0z 06
Fig. 4. The relationship between the fraction of arginine taken from the medium and the
concentration of arginine in the medium.
A ? Fraction of arginino content taken from outside; B ? Arginino concentration (y/1).
Case C, 0.1 units and Case D, no arginine. These four chemostats were initially
inoculated with the B/1,t strain of bacteria, and they were permitted to run
at a generation time of eight hours until a steady state was reached. In all
cases the concentration of arginine in the growth tube fell so low as to be chemi-
cally undetectable. It can be concluded that in Case A the bacteria make fifty
per cent of their arginine requirement and take up fifty per cent from the
medium, that in Case B they make seventy per cent and take up thirty per
cent, that in Case C the make ninety per cent and take up ten per cent, and
that in 'Case D they make one hundred per cent of their arginine requirement.
The next step is to determine in each case how the concentration of arginine
in the Medium- -corresponds to the rate of synthesis. The concentrations of
42
arginine, being too low for chemical measurement, were measured microbiolo-
gically. This was done by observing the effect of these concentrations on the
growth rate of an arginine-requireing strain. A number of D81!6 bacteria, very
small in comparison with the number of B/1 ,t, was added directly to the growth
tube of each of the four chemostats once the steady state was reached. 13y
plating with excess phage TO, the number of D84/6 bacteria could be measured
as a function of time, and it is these numbers that are given in the semi-log
plot in Figure 2. From the rate of rise, or fall, of the number of DS4/6 bacteria
per ml, the growth rate can be calculated, using the relationship
I dn.
-Vc = or
it
1 cl )1 71,
A =_-
I (it
(9)
1 dn
where a is the growth rate of the D84/6 bacteria, ? ?a-t the slope of the rise
d
or fall in the semi-log plot, and 7 the dilution rate. The observed growth rate
of the D84/6 strain can be used to obtain the concentration of arginine in the
medium, given the relationship shown in Figure 3 between the growth rate
of D84/6 and the concentration of arginine established in independent growth
studies. In this way the results shows in Figure 2 were used to ascertain for
the Bil,t strain the relationship shown in Figure 4 between the rate of arginine
uptake (or synthesis) and the concentration of arginine in the medium.
SUMMAR Y
Continuous culture methods offer many advantages to the microbiologist.
The chemostat can be used to maintain a population of microorganisms
growing indefinitely under constant conditions. The size of the population,
the growth rate, and the controlling growth factor can be selected within wide
ranges according to the experimentor's requirements.
The chemostat is especially useful in two kinds of experiments, where attack
by conventional methods is not feasible. For instance, the study of the accu-
mulation of mutants in a population demands an environment where growth
may be observed over many generations under constant conditions. The
second kind of experiment involves an environment in which a metabolite,
that is being rapidly consumed, must be kept at a low constant concentration
in the medium. Applications of both kinds of experiment have been illustrated
above.
It is expected that the chemostat will be used to obtain optimum production
of industrially important biochemicals produced by microorganisms. Because
it permits a wide variation of growth rate and choice of controlling growth
factor, the chemostat can be used to determine the dependence of the rate
43
fl,,-Inecifiarl in Part - Sanitized Com/ Approved for Release @ 50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
nine in the medium. The results of a series of experiments demonstrating such
an analysis are given in Figure 2. Here a series of chemostats was set up to as
to contain in the incoming nutrient the following concentrations of tryptophan
and arginine. Each chemostat contained one unit of tryptophan, but the argi-
nine content varied as follows: Case A, 0.5 units of arginine; Case B, 0.3 units;
A
1.0
0.8
0.6
0 4
02
02 0 4 06
Fig. 4. The relationship between the fraction of arginino taken from the medium and the
concentration of arginine in the medium.
A ? Fraction of arginine content taken from outside; B ? Arginine concentration (y/l).
Case C, 0.1 units and Case D, no arginine. These four chemostats were initially
inoculated with the B/1,t strain of bacteria, and they were permitted to run
at a generation time of eight hours until a steady state was reached. In all
cases the concentration of arginine in the growth tube fell so low as to be chemi-
cally undetectable. It can be concluded that in Case A the bacteria make fifty
per cent of their arginine requirement and take up fifty per cent from the
medium, that in Casey B they make seventy per cent and take up thirty per
cent, that C they make ninety per cent and take up ten per cent, and
-
that in Case D they make one hundred per cent of their arginine requirement.
The next step -i-it:)-deierMine in each case how the concentration of arginine
in the ineditim -"corresponds to the rate of synthesis. The concentrations of
42
-
arginine, being too low for chemical measurement, were measured microbiolo-
gically. This was done by observing the effect of these concentrations on the
growth rate of an arginine-requireing strain. A number of 1)84/6 bacteria, very
small in comparison with the number of B/i ,t, was added directly to the growth
tube of each of the four chemostats once the steady state was reached. By
plating with excess phage T6, the number of DS4/6 bacteria could be measured
as a function of time, and it is these numbers that arc given in the semi-log
plot in Figure 2. From the rate of rise, or fall, of the number of 1)8.1/6 bacteria
per ml, the growth rate can be calculated, using the relationship
dn
cit
W
? ? 01
di:
A = ?
71, (It V
(9)
du
where a is the growth rate of the D84/6 bacteria, ? ? the slope of the rise
7z di
or fall in the semi-log plot, and 7 the dilution rate. The observed growth rate
of the D84/6 strain can be used to obtain the concentration of arginine in the
medium, given the relationship shown in Figure :3 between the growth rate
of D84/6 and the concentration of arginine established in independent growth
studies. In this way the results shows in Figure 2 were used to ascertain for
the B/1,t strain the relationship shown in Figure 4 between the rate of arginine
uptake (or synthesis) and the concentration of arginine in the medium.
SUMMARY
Continuous culture methods offer many advantages to the microbiologist.
The chemostat can be used to maintain a population of microorganisms
growing indefinitely under constant conditions. The size of the population,
the growth rate, and the controlling growth factor can be selected within wide
ranges according to the experimentor's requirements.
The chemostat is especially useful in two kinds of experiments, where attack
by conventional methods is not feasible. For instance, the study of the accu-
mulation of mutants in a population demands an environment where growth
may be observed over many generations under constant conditions. The
second kind of experiment involves an environment in which a metabolite,
that is being rapidly consumed, must be kept at a low constant concentration
in the medium. Applications of both kinds of experiment have been illustrated
above.
It is expected that the chemostat will be used to obtain optimum production
of industrially important biochemicals produced by microorganisms. Because
it permits a wide variation of growth rate and choice of controlling growth
factor, the chemostat can be used to determine the dependence of the rate
43
?
nn,rimecifiarl in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
at which a given product is made. on changes in these variables. Once the opti-
mum conditions are established, the chemostat can be used for the continuous
synthesis of the product ata high rate.
REFERENCES
Fox M. S., SZ1LARD L., 1955: A Device for Growing Bacterial Populations under Steady
State conditions. J. Gen. Physiol., 39: 261.
Luz H. H., 1953: The Mutation of E. coil to Resistance to Bacteriophage Ts. Arch. Bio-
chem. Biophys., 47: 438.
MONOD J., 1950: La technique do culture continue. ThOorie et applications. Ann. Inst.
Pasteur, 79: 390.
NOWOK A., 1955: Growth of Bacteria. Ann. Rev. Microbiol., 9: 97.
NOVICK A., 1956: Brookhaven Symposium on Mutation, 8: 201.
NOVICK A., SZILARD L., 1950a: Description of the Chemostat. Science, 112: 715.
Novwx A., SZ1LARD L., 1950b: Experiments with Chemoetat on Spontaneous Mutations
of Bacteria. Proc. Natl. Acad. Sol U. S. 36: 708.
Novwx A., SWARD L., 1951: Genetic Mechanisms in Bacteria and Bacterial Viruses.
Experiments on Spontaneous and Chemically induced Mutations of Bacteria Growing
in the Chemostat. Cold Spring Harbour Symp. Quant. Biol. 16: 343.
Novwx A., SZILARD L., 1954: Dynamics of Growth Processes. Princeton Univ. Press,
Princeton, N. J., pp 21-32.
Novwx A., Wanrza M., 1957: Proc. Natl. Acad. U. S., 43: 553.
Rwzzaszza H. V., Comm G. N., Burrni G., MONOD J., 1956: Ann. Inst. Pasteur, 11:
383.
YANOTSKY C., 1957: Enzymatic Studies With a Series of Tryptophan Auxotrophs of
E. coll. J. Biol. Chem. 224: 783.
44
neclassified in Part - Sanitized Copy Approved for Release
CONTINUOUS CULTURE OF MICROORGANISMS;
SOME THEORETICAL ASPECTS
D. HERBERT
I. THEORETICAL
A simple mathematical theory of continuous culture can be derived (Herbert,
Elsworth, Telling, 1956) from two basic features of bacterial growth which
were first established in batch culture experiments (Monod 1942, 1950).
In the simplest case, when bacteria are growing in a medium containing a single
carbon substrate (e. g. glucose-ammonia-salts media), these can be stated as
follows:
1. The growth-rate is a constant fraction, Y, of the rate of utilization of
substrate:
dx v (18
dt = dt
(I)
where x = concentration of bacteria, 8 = concentration of substrate, t = time
and Y is called the yield constant. Over any finite period of growth:
weight of bacteria formed
V (2)
weight of substrate used
2. The specific growth-rate (p) is a function of the concentration of limiting
substrate, being proportional to the substrate concentration when this is
low but reaching a limiting saturation value at high substrate concentrations
according to the equation:
dX
? = =_.- m 8
X dt K, ? s
(3)
where p?, is the growth rate constant (i. e. the maximum possible value of p
in the medium used) and K, is a saturation constant numerically equal
to the substrate concentration at which p = p?,/2. (This is similar to the
Michaelis-Menten equation for the effect of substrate concentration on the
velocity of enzyme action.)
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
45
STAT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
From equations (1) and (3) a mathematical theory can be derived which
allows quantitative prediction of the behaviour of a continuous culture. The
type of apparatus considered is the "Chemostat"; bacteria are grown in a culture
vessel or "fermenter" into.which sterile growth medium is metered at a steady
flow-rate (/) and from which bacterial culture emerges at the same rate, a con-
stant-level device keeping the volume (v) of culture in the fermenter constant. ?
A
4 -J
?
?
3
4
0.5
????
? ? '??? ? ?
?
DH
B C
10
6 8
6
4
4
2
.0 D
Dc
Fig. 1. Steady state relationships in continuous culture (theoretical). The steady-stato values of
substrate concentration and output at different dilution rates are calculated for an organism
with the following growth constants: ? = 1.0 hr.-1, Y = 0.5 and K, = 0.2 g/1.; and a substrate
concentration in the inf lowing medium of dic = 10 g/1.
A ? Output of bacteria, Dx (g/l/lir); steady state bacterial concentration, x (g/l). B ? Doubling
time, td (hr.). C ? Steady-state substrate concentration, a (g/1). D ? Dilution rate, D (hr-1).
Curves: 1 ? bacterial concentration; 2 ? output of bacteria; 3 ? doubling time; 4 ? substrate
concentration.
The culture in the fermenter is vigorously stirred and aerated, and tempera-
ture and pH are automatically controlled.
Residence-times in such a culture vessel will be determined not by the
absolute values of the flow-rate and culture volume but by their ratio which
we call the dilution rate, .D, defined as D = fly, i. e. the riumber of complete
volume-changes per hour. The mean residence-time of an organism in the
culture vessel is evidently equal to 1/D.
Equations can be derived which completely predict the behaviour of such
a system at different flow-rates, medium concentrations, etc. The most import-
ant prediction of the theory is that a continuous culture is an inherently stable
system. For any set dilution rate, the culture automatically adjusts itself to
a steady state in which the concentrations of miero-organisms and nutrients
in the culture remain constant indefinitely, so long as the composition and
46
-
1
?1
flow-rate of the incoming medium remain unaltered. In such a steady state,
the growth-rate of the organisms (,u) must be equal to the dilution rate (D).
The stability of the system is due to the fact that it is essentially substrate-
controlled. A chemostat is in fact a device for controlling the growth rate
through control of the steady-state substrate concentration; each dilution
rate fixes the substrate concentration at that value which makes p equal to D.
In Fig. 1, theoretical curves are drawn from the basic differential equations
of continuous culture, showing how the steady-state concentrations of bacteria
and substrate in the culture may be expected to Vary when the dilution rate
is varied, the inflowing substrate concentration (s?) being held constant.
It will be seen that the existence of an infinite number of steady states is
predicted, with dilution rates ranging from almost zero to a maximum value
corresponding to the maximum possible growth-rate in the medium used
(p,?); above this "critical" dilution rate, "wash-out" of the culture occurs
if the growth rate is further increased. It will also be seen that over most of the
possible range of flow-rates, the steady-state concentration of substrate
in the fermenter- is very low (i. 0. the substrate is almost completely consumed);
only at dilution rates close to the critical does unused substrate appear in the
culture. Fig. 1 also shows the theoretical output of bacteria (i. e. g. cells/litre/
/hour) as a function of dilution rate; it will be seen that this curve goes through
a maximum value at a unique flowrate which is the "optimum" for production
of cells or products. The above features of a continuous culture are of obvious
importance from the practical standpoint.
II. E X PE RIM ENT A L
Descriptions will be given of laboratory and pilot plant types of continuous
culture apparatus used at M. R. E., Porton. These have been operated success-
fully with a number of different bacteria and recently with yeast (Pond('
ittilis) and moulds. In each case, systematic studies were made to see how the
behaviour of the organism in continuous culture compared with the theoretical
predictions. The technique was to set up a steady state at a fixed flow-rate
and allow the apparatus to run for some days, during which samples were
analysed for total and viable cell counts, dry weights of organisms, glucose
and other constituents of the culture supernatant etc; analyses for nucleic
acids and other cell constituents were also made on the cells. The flow rate
was then changed and the determinations repeated at a new steady state;
the aim was to cover as wide a range of flow-rates as possible. A typical run
usually lasts for 2-3 months; no difficulty is now found in maintaining sterility
for such periods and some runs have lasted as long as 200 days. Usually the
respiratory activity of the growing cells was determined by continuous analysis
neclassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
47
Declassified in Part - Sanitized Cop Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
and recording of the oxygen and carbon dioxide content of the exit gas from
the fermenter.
5-
X
10
-8
-6
.4
-2
0
10
-8
6
4
2
0
0.2 0.4 6 0:8 C
X '
Fig. 2. Growth of Aerobacter aerogenes and Torula Willa in continuous culture. Organisms were
grown with aeration at a number of different flow-rates; dry weight of cells and concentration
of substrate determined after at least 2-3 days of steady-state growth at each flow-rate.
A ? Cell coneentration (mg dry weight/ml). B ? Glycerol concentration (mg/ml). C ? Dilution
rate (hr-1). D ? Glucose concentration (mg/ml). X ? Wash-out point. Curves: 1 ? cell con-
centration; 2 ? glycerol concentration; 3 ? glucose concentration.
In general, results agreed quite well with theoretical predictions; in parti-
cular, the stability of a c'ontinuous culture over a wide range of flow-rates
was confirmed With all organisms tested. This self-adjusting property makes
a continuous culture very easy to opeiat; it is neccessary only to set and
_ _
maintain constant the flow-rite; when the systemwill regulate itself.
48
Fig. 2 shows some illustrative quantitative data, obtained with iterobacter
aerogenes and with Torula utilis; similar results were found with other bacteria.
It will be seen that (a) steady states were obtained over a wide range of dilution
rates; the lowest (D = 0.05 hr-1) correspond to a cell division time of c. 14 hours.
It is remarkable that bacteria and yeasts will continue to divide at this very
low rate almost indefinitely; counts showed that the cultures were >90%
A
700
600
500-
400-
300.
200-
100
C.-
Ix
?
0.2
01.4
0.6 0.8
Fig. 3. Respiration of Aerobacter acrogerte8 growing in continuous culture. The organism was
grown with aeration at a number of different flow-rates in a glycerol-N113-salts medium. Oxygen
uptake of steady-state cultures determined with recording paramagnetic oxygen analyzer and
expressed as Q0, (?10,/mg cells/hour).
A ? Q01 of cells (z1/mg dry wt./hr). B ? Dilution rate (hr-1). X ? Endogenous respiration.
viable. (b) The "wash-out" points occurred at dilution rates of 0.585 hr-1
for Torula viii is and 0.85 hr-1 for Aerobacter aerogenes, corresponding to division
times of 71 minutes and 49 minutes respectively; these values agreed well with
measurements of the growth rates during exponential growth in batch culture.
(c) Over most of the range of dilution rates the substrate was almost completely
utilized, being so low as to be almost impossible to estimate; near the "wash-
out" point, however, the substrate concentration rose abruptly as the cell
concentration fell.
In the above respects, agreement with the simple theory is good. With both
organisms, however (and with others we have examined), there is a noticeable
discrepancy at low flow-rates, the yield of organisms being less than at high
4?Symposium
49
npriassifipri in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
flow-rates. In other words, the "yield constant" (Y) is not truly a constant,
but decreases at low growth rates.
The most probable explanation of this is that in addition to the anabolic
metabolism of the organisms (conversion of substrate to cell-substance), they
have also a constant endogenous metabolism, by which cell-substance is
oxidized to CO,. At low growth-rates, when the residence-time of the bacteria
in the fermenter is increased, this basal metabolism becomes proportionally
more important, compared with the anabolic metabolism. Evidence for this
has come from studies of the respiration at different growth-rates (Fig. 3).
In this experiment, which was made with Aerobacter aerogenes, the respiratory
activity of the cells (Q0.) is seen to be a linear function of the growth-rate;
however, the straight line does not pass through the origin but extrapolates
back to a Q0, of about 65 at zero growth rate. Exactly similar results are obtai-
ned if Qc0. values are similarly plotted; moreover a larger proportion of the
substrate-carbon used is converted to CO, and a smaller proportion to cell-
carbon, when the growth-rate is low.
All the above facts support (though they do not definitely prove) the idea
of a constant endogenous metabolism, independent of the growth rate. If an
extra term for the endogenous metabolism is incorporated into the continuous
culture equations, the "theoretical" curves now resemble those found experi-
mentally (Fig. 2), rather than Fig. 1.
2. EFFECT OF GROWTH RATE ON CELL COMPOSITION
AND MORPHOLOGY
During experiments of the type described above, cells grown in the steady
state at different growth rates were analyzed for total protein, ribonucleic
acid (RNA) and desoxyribonucleic acid (DNA). The mean cell mass (i. e. the
average dry weight of a single cell) was also determined as the ratio of the
dry weight of cells/ml. to the total count/ml. Results for Aerobacter aerogenes
ares shown in Fig. 4 (similar results have been obtained with Bacillus megaterium
and Staphylococcus aureus).
It will be seen that rapidly-dividing cells have a much greater dry weight
(and .microscopically are much larger) than slowly-dividing cells. In addition,
the-RNA content of the cells increases markedly with growth-rate, while the
DNA-content decreases somewhat.
The writer prefers not to speculate at present on the meaning of these
?
rather striking effects. It may be pointed out, however, that if the extreme
ends of ,the'reiirve are considered, the cells grown at the maximum rate correspond
in chemvotry.tAndAnotphology to "logarithmic phase" cells in a batch culture,
' hue those grown at the slowest rate correspond very closely to "resting cells".
50
The advantage of continuous culture techniques is that both these extreme
types of cell, and an infinite number of intermediate graduations between them,
may be isolated at will for study and grown indefinitely under steady-state
A
20-
15-
10-
5-
80
60
40
20
4
D
0.2 0:4
0.6
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Fig. 4. Effect of growth rate on mean cell mass ond nucleic acid content of Acrobader aerogenes.
The organism was grown in a glycerol-Nilo-salts medium in continuous culture apparatus at
a number of different flow-rates and the cells analyzed after at least 2-3 clays of steady-state
growth.
A ? % of RNA or DNA. B ? % Protein. C ? lean cell mass (picograms). D ? Dilution rate
(= division rate) ? hours-1.
Curves: 1 ? % protein; 2 ? % RNA; 3 ? mean cell mass; 4 ? % DNA.
conditions, simply by appropriate adjustment of the flow-rate. This may
serve as an illustration that continuous culture techniques are not only of
considerable industrial importance, but also can. be a powerful research tool.
4 ?
51
npriassifipri in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
REFERENCES
HERBERT D., ELSWORTH 11., TELLING 11. C., 1956:
a Theoretical and Experimental Study. .1. Con
Motion 1942: Recherches sur la, croissanco des
Cio, Paris.
Mowou 1950: La technique do culture continue
Pasteur, 79, 390.
52
Tho Continuous Culture of Bacteria;
. Microbiol., 14, 601.
cultures bactdriennes. Herman and
Tlitorio et applications. Ann. Inst.
noriaccifiimr1 in Part - Sanitized Coov Approved for Release
THE CONDITIONS OF GROWTH OF MICROORGANISMS
IN CONTINUOUS FLOW CULTURES
N. D. j ERUSALIMSKIJ
The method of cultivating microorganisms in continuous flow culture is
receiving ever increasing application both in research work and industry.
There exist threc. basic modifications of this method, not to mention various
intermediates.
The first one in a cyclic-flow system in which fresh medium is continually
added into the apparatus for the growth of the culture (the cultivator), while
at the same time an equal volume of liquid medium is removed from it; micro-
bial cells remain inside the apparatus all the time and multiply.
The number of living cells in 1 ml of medium gradually increases until it
reaches the maximum level possible under the given conditions of the medium.
Thus this method appears to be cyclic but not uninterrupted. It is applied
industrially for the growth of yeast.
The second and the third methods are truly continuous. In the second
method microorganisms grow in a fixed state on the surface of a solid phase
while the liquid medium continually flows by. Some of the cells are washed
away by the flow of the liquid and therefore the number of cells on the surface
of the solid phase remains more or less constant even in the course of rather
long experiments. A typical example of this method is represented by the
so called fast method of vinegar production (Selmellessigverfahren) introduced
in Germany at the beginning of the last century.
In the third type of cultivation cells develop in a liquid medium. The amount
of incoming fresh medium is equal to the amount of cell-containing culture
liquid removed from the apparatus. If the rate of washing away of cells equals
the rate of their multiplication the density of the cell population inside the
apparatus remains constant.
During the last few years this last modification of the method has received
wide application in laboratory research work. It exists again in several minor
modifications which differ in the type of regulation of fresh medium delivery,
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
53
STAT
Declassified in Part - Sanitized Co .y Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
of stirring the culture and of aeration. In some of them the medium is delivered
towards the bottom of the apparatus while the excess liquid escapes on top;
in others the flow of the liquid is reversed. Sometimes the apparatus consists
of one "cultivator" only, in some cases, however, several of them are connected
like a battery (the multistage, continuous culture). As long ago as in 1915
S. V. Lebedev proposed a similar battery consisting of several units for alcohol
production. In the laboratory of one of the organizers of this symposium,
Prof. Welt, the multi-stage culture is being fruitfully applied in the study
of various problems.
In our laboratory we have studied the conditions governing bacterial growth
when the first method and partly the second and third method are used. We
used glass flasks of 25 to 250 ml volume according to K. I. Draganov (1957)
as well as those according to S. V. Lebedev (1936). The rate of flow of the
medium was regulated by changing the hydrostatic pressure and, in addition,
by means of a screw clamp. The object of out study was mainly Propionibacte-
rium sherrnanii (the work was carried out by N. M. Neronova) and Clogridium
acelobulylicum (the work was carried out by G. V. Pinaeva).
Monod (1950) and Novick and Szilard (1950) derived mathematical formulae
for the growth of culture during single-stage continuous cultivation. In other
papers (e. g. Maxon, 1955) formulae are presented also for multi-stage cultiva-
tion. We should like to present the following universal formula characterizing
the density of population in any culture system:
= X1e(e-ot ? Xo[e(c-ot ?1]
c r
(1)
The rate of growth of population in a cultivator is given by the following
equation:
dx
dt = cX rX0 ? rX
(2)
In the given formulae X and X, are the density of population (the number
of cells per 1 ml) at the beginning and at the end of period t respectively; X. is
the density of population in the liquid entering the cultivator; e = 2.718, i. e.
the base of natural logarithms; r = y is the rate of dilution of the culture,
i. e. the ratio between the rate of flow of the medium (F ml/hour) and the
dm 1 d(ln m)
volume of the cultivator (V ml); c = ? -- and indicates the re-
dt m dt
lative rate of growth of microorganisms, i. e. the increase of one unit of
living matter_ (iii) per unit of time.
The .rate of grov.:th.of microorganisms depends on the composition of the
nutrient -Modiuni. When the nutrient contained in the medium in minimum
54
^
narlaccifiPrl in Part - Sanitized Copy Approved for Release
concentration is used up, the rate is slowed down. On this principle the regula-
tion of growth of a culture in "Chemostat" (Novick and Szilard) and in
"Bactogen" (Monod) is based. The accumulation of metabolic products can
be another reason for the slowing down of growth; particularly, for example,
of acids lowering the pH to undesirable values. In their turn, processes of
utilization of nutrients and of accumulation of metabolic products are connected
with the density of population. Consequently the rate of growth (c) depends
also on the density of population (X): With increasing density the rate of
growth decreases. Simultaneously the rate of increase of density of population
dX
decreases (?dt as follows from equation (2). When in this equation (cX + rX0)
dX
becomes equal to rX, then 7?It is equal to zero. This means that the density
of population does not increase any more and remains constant. The existing
equilibrium is expressed by the following equation:
C _X?X0
?
X
(3)
If, then, the rate of dilution (r) is purposely changed, the equilibrium is
affected. The density of population (X) and the rate of growth (c) start to
change in one or the other direction until an equilibrium ? at a different
level to be sure ? is reached. Thus the growth of a continuous flow culture is
actually a self-regulating process.
The medium entering the first cultivator contains no bacteria (X0 = 0).
In this case equations (1) and (2) may be simplified to:
X = X1 . e(c-Ot (4)
dX
cX ? r X
di
(5)
In the first cultivator equilibrium between the dilution and growth of the
culture is established when
(6)
It follows that at equilibrium (when the density of population remains
constant) the rate of growth of microorganisms in the first cultivator (c) is
equal to the rate of dilution (r).
A study of continuous cultures Of any microorganisms should start with
measurement of the maximum rate of growth attainable under the given
experimental conditions. This measurement should be carried out in the first
cultivator.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
55
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
- _
In doing this we gradually increase the rate of flow of the medium and thus
the rate of dilution. The highest rate of dilution at which the density of popula-
tion still manages to remain constant represents the maximum rate of growth
(co) : r = co [cf. eq. (6)].
Let us give an example of the measurement of the maximum rate of growth
in the case of Propionibacterium shermanii (Tab. 1).
Table /
Time
in hours
Rate of dilution
(r = irP)
Density
of population in
104/m1 (X)
()
274
16
280
18
253
20
0.38-0.42
22
240
24
230
40
229
302
46
177
48
0.60-0.66
154
50
150
64
165
66
157
68
166
70
160
116
15
118
1.0
13
120
11
121
12
136
0.9
22
138
1.0
20
140
1-2
9
As follows from Tab. 1 in the given case the limiting rate of dilution is appro-
ximately equal to 1.0. Consequently the maximum rate of growth is co = 1.
The generation time (the time for a two-fold increase of the living matter)
is then:
hi 2
te = ? = 0.69 hours
co
In the case if a continuous culture of Cl. acetobutylicum on maize mash with
sugar added. themaximum growth rate c0 was equal to 0.8 and the generation
time was correspondingly 0.87 hr. Under suitable conditions a continuous
culture Can proceed for any length of time. In our experiments, for instance,
a eUlture' of Cl. acetobutylicum multiplied at a constant rate for 200 days.
56
norlaccifiimr1 in Part - Sanitized Copy Approved for Release
KL1
The density of population in the cultivator lay between 350 to 450 x 106 cells/ml.
The rate of delivery of medium to the .cultivator (F) was 20-23 ml/hr.
while the volume of the cultivator was V = :33 ml. It. follows hence that the
rate of dilution (r) and the growth rate (c) was equal to 0.6-0.7. Proceeding
from the relation between the growth rate and the generation time it is not
difficult to deduce that in the course of the mentioned 200 days 4,360 genera-
tions were formed.
A battery consisting of several cultivating units can be used in order to
obtain a denser cell population. But the density of population (X) is not the
only criterion of the qUalities of a given system of cultivation. With this in
mind it is necessary to take into account the yield of living matter per hour:
yield of living matter = XE (7)
Finally it is also necessary to know the productivity value, i. e. the amount
of living matter produced per hour per unit of volume of cultivator:
productivity = = rX
(8)
Tab. 2 shows the density of population values of Propionibact. shermanii
in various systems of cultivation; in Tab. 3 the values of productivity of dif-
ferent systems are presented.
At a low flow rate (F ? 10 ml/hr.) the maximum density of population
is reached when the volume of the second cultivator is equal to 150 ml (Tab. 2).
Table 2
Flow rate
of medium F
in mi/hr.
Mean population of cells of Propionibact. shermanii
in cultivators ? in 106/m1
Volume of cultivator in ml
first
25
1)th
70
ISO
2.10
10
15
25
730
520
25
2000
1170
395
5390
3.160
3800
:5380
5050
5100
At higher flow rates (15 and 25 ml/hr.) the increase of volume of the 1?th
cultivator from 150 to 240 ml causes a supplementary increase of the density
of population.
As far as productivity is concerned "(Tab. 3) at a flow rate F = 10-15 ml/hr.
it reaches its maximum value already in the first cultivator. Therefore at such
a flow rate it is necessary to connect another cultivator if it is required to
obtain a higher concentration of cells in 1 nil.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
57
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
At a flow rate of 25 ml/hr. the highest degree of productivity is reached
when the volume of the PI" incubator is 150 ml (Tab. 3). The density of popula-
tion is at that time also sufficiently high. Therefore the given variant is the
optimal one of all 12 studied in this experiment. In the second place lies the
variant with a volume of the second cultivator equal to 240 ml and with a flow
rate of 25 ml/hr.
Table 3
Flow rate
of medium F
in mi/hr.
-
Productivity
the number
of a system of cultivators
of cells (in 101) produced per hour
of volume of cultivators
X F XI
expressed as
per 1 ml
and
171 V 'V
1 -1-- 2
Arolume of cultivator in ml
first
25
pm
70
150
240
10
15
25
292
312
28
210
185
104
308
296
543
203
286
481
In order to determine in advance the optimal number of cultivator units,
the relation of their volumes and the rate of delivery of medium it is necessary
to know the law according to which the growth rate (c) changes. As mentioned
above c is the reciprocal function of the density of population X.
As the first approximation it is possible to accept the following relationship
between these values:
Co ? C Co
X ? X. M? X. ? const.
In this equation co is the maximum rate attainable in a given medium, X?,
is the density of population which, when exceeded, causes the growth rate c to
decrease, M is the limiting density of population which, when reached, stops
the growth of the culture.
For each strain of microorganisms at a given composition of the medium
the values of 4, X. and M are constant.
Providing the density of population is not too high the equation (9) satis-
factorily corresponds to the obtained results (Tab. 4).
It we substitute the value of c from equation (9) in equations (3) and (6)
we get,formulae by means of which we can calculate in advance the density
of microbial population in various types of cultivation.
, ,
Examples of such calculations are shown in Tab. 5. We assumed that the
volumes of all,"miltivafors" in the battery are equal (V1 = Vs -= Vs etc.); the
maximuni growth, rate c = 1.0 and --- = 0.95.
-
(9)
x.
58
1
The highest productivity is reached ? if single-stage culture is employed ?
when the flow rate is equal to 0.4-0.6 (Tab. 5). But then the density of popul-
ation represents only 43 to 62% of the maximum value. It it is desired to have
Table 4
Density
of population
-V in lOgiml
Growth
rate
c
Ratio
Co ? C
X ? Xm
13k)
1.0)
?
143
OGS
0.5
152
0.62
2.7
159
0.61
2.7
169
0-57
2.8
230
0-42
2.7
247
0-40
2.6
277
0-35
2-5
371
025
2.2
447
0-26
1.7
805
0.07
1.2
?) The value 13 x 104 is taken for X,;
?*) The value 1-0 is taken for cn.
Table 5
Ratio of density of population to
Productivity in arbitrary units at
Ratio of flow
maximum density
various numbers of cultivators in
rate to volume
of Pt cultivator
x
AI
battery
F
Cultivators
Number of cultivators
V
pa
ond
3rd
4th
1
0
-
3
4
0 1
0-91
0 99
1-0
1-0
0-09
0.05
0-03
0-02
0-2
0-81
0.97
0-99
1 0
0-16
0 10
0-07
0-05
0-3
0.72
0.93
0-99
1-0
0-21
0.14
0-10
0-07
0-4
0.62
0-89
0 97
0-99
0-25
0-18
0-13
0-10
0-5
0-53
080
0.94
0 98
0-26
0-21
0-16
0-12
0-6
0-43
0-76
0-91
0-97
0-26
0-23
0-18
0.14
0-7
0-34
0-67
0 86
0-94
0-23
0-23
0-20
0-16
0 8
0-24
0-56
0-79
0-90
0-19
0-23
0-21
0 - 18
0-9
0-15
0-43
0-69
0-84
0-13
0-19
0-21
0-19
1-0
005
0 24
0 51
0-72
0-05
0-12
0-17
0-18
a density of population higher than 0-9, then it is necessary to use a battery
of 2 or 3 cultivator units. In this case it would be uneconomical to use only one
cultivator because of low productivity. Batteries consisting of 4 or more
cultivator units are also uneconomical, but they have the advantage of enabling
the simultaneous study of cultures at different stages of development.
59
npriassifipri in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
..,?????
At a flow rate of 25 ml/hr. the highest degree of productivity is reached
when the volume of the Ph incubator is 150 ml (Tab. 3). The density of popula-
tion is at that time also sufficiently high. Therefore the given variant is the
optimal one of all 12 studied in this experiment. In the second place lies the
variant with a volume of the second cultivator equal to 240 ml and with a flow
rate of 25 ml/hr.
Table :5
Flow rate
of medium F
in ml/hr.
-
Productivity
the number
of a system of cultivators
of cells (in 106) produced per hour
of volume of cultivators
xp XF
expressed as
per 1 ml
and
VI 171 + 12
Volume of cultivator in ml
first
25
Pth
70
150
240
10
15
25
292
312
28
210
185
104
308
296
543
203
286
481
In order to determine in advance the optimal number of cultivator units,
the relation of their volumes and the rate of delivery of medium it is necessary
to know the law according to which the growth rate (c) changes. As mentioned
above c is the reciprocal function of the density of population X.
As the first approximation it is possible to accept the following relationship
between these values:
co -c co
X - X.- - const.
(9)
In this equation c, is the maximum rate attainable in a given medium, X?,
is the density of population which, when exceeded, causes the growth rate c to
decrease, M is the limiting density of population which, when reached, stops
the growth of the culture.
For each strain of microorganisms at a given composition of the medium
the values of co, X. and M are constant.
Providing the density of population is not too high the equation (9) satis-
factorily corresponds to the obtained results (Tab. 4).
It we substitute the value of c from equation (9) in equations (3) and (6)
we get formulae by means of which we can calculate in advance the density
of microbial Population in various types of cultivation.
.Examples, of such Calculations are shown in Tab. 5. We assumed that the
voli?f-all "Cultivators" in the battery are equal (V1 = Vs = V, etc.); the
= -
X?,
maximum growth rate co = 1.0 and - = 0.95.
58
The highest productivity- is reached - if single-stage culture is employed -
when the flow rate is equal to 04-0.6 (Tab. 5). But then the density of popul-
ation represents only 43 to 62% of the maximum value. It it is desired to have
Table 4
Density
of population
X in 106/m1
Growth
rate
c
Ratio
co - c
X - XIll
136)
1.0")
-
143
0.68
2.5
152
0.02
2.7
159
0.61
2.7
169
0.57
2.8
230
0.42
2.7
247
0.40
2.6
277
0.35
2.5
371
0.25
2.2
447
0.26
1.7
805
0.07
1.2
*) The value 13 x 106 is taken for X?,;
**) The value 1 0 is taken for co.
Table 5
Ratio of density of population to
Productivity in arbitrary units at
Ratio of flow
maximum density
various numbers of cultivators in
rate to volume
of lot cultivator
X
ilf
battery
F
Cultivators
Number of cultivators
V
1st
2nd
3rd
4th
1
')
-
3
4
0.1
0.91
0.99
1.0
1.0
0000>00005)0
0 .. .., I:D 1.7 1:2 I.D t?D '? ,i,
.? ca ca
0.05
0.03
0.02
0.2
0.81
0.97
0.99
1.0
0 10
0.07
0.05
0.3
0-72
0 93
0.99
1.0
0.14
0-10
0.07
04
0.62
0.89
0.97
0.99
0.18
0-13
0.10
0.5
053
0.80
0.94
0.98
0.21
0.l6
0.12
0.6
0.43
0.76
0.91
0 97
0.23
0.18
0.14
0.7
0.34
0.67
0.86
0.94
0.23
0.20
0.16
0 8
0.24
0.56
0.79
0.90
0 23
0.21
0.18
0-9
0.15
0.43
069
0.84
0.19
0.21
0.19
1.0
0.05
024
0.51
0.72
0.12
0.17
0.18
a density of population higher than 0.9, then it is necessary to use a battery
of 2 or 3 cultivator units. In this case it would be uneconomical to use only one
cultivator because of low productivity. Batteries consisting of 4 or more
cultivator units are also uneconomical, but they have the advantage of enabling
the simultaneous study of cultures at different stages of development.
59
npriassifipri in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
N, ?
It is possible to make calculations in a similar way also in the case of batteries
with other ratios of cultivator volumes and at different microorganism growth
rates.
It is necessary, however,, to point out that at high densities of microbial
population the growth rate equation (9) is no more valid and therefore the
above-mentioned calculations are incorrect. In dense populations in continuous
flow cultures only a part of the total number of cells continues to multiply
while the rest gradually lose their living activity. In the end an equilibrium
between the multiplication and loss of activity of the cells is established.
In our experiments the culture of Cl. acetobutylicum grew in collodion sacs
submerged in the continuous flow medium. The density of living cells remained
constant the whole time at a level of 1000-1200 x 104/ml, while the total
density of cells increased continually along a straight line until it reached
about 7000 x 104/m1 or more. Similar data for E. coli can be found in litera-
ture.
The cultivation in collodion sacs in a flow medium can be compared with
the first of the three types of cultivation desctibed in the first paragraphs
of this article. As stated there, in the second type of cultivation microorganisms
grow on the surface of a solid phase along which liquid medium flows continu-
ously. It has been mentioned in literature that the walls of cultivators are
sometimes covered with a film of bacteria. This is considered as undesirable
as it affects experimental findings.
We tried to add glass wool to a cultivating flask containing Pro pionibact.
shermanii. At the beginning bacteria grew mostly on the surface of the wool
and then spread through the liquid medium, where they started to multiply.
The cell concentration in the escaping liquid reached a constant value. For
instance, in one of the experiments their density in 1 ml remained within the
limits of 3000 to 4000 x 106 in the course of 19 hours. The dilution rate
in this experiment was equal to 1.0. Consequently the medium in the cultivator
was actually renewed 19 times. At the same flow rate in a cultivator without
any solid phase the density did not exceed 20 to 30 x 10./ml. Thus it appears
that the presence of a solid phase considerably increases the productivity
of the cultivator. Unfortunately the bacterial film is easily torn away from the
surface Of the solid phase and the continuity of the process is thus broken.
In addition the amount of cellular material on the surface of the solid phase
cannot be counted and it is impossible to determine the growth rate.
In the submitted paper we are dealing only with the problem of growth of
bacterial mass in continuous flow media. But the method of continuous flow
cultures can be conveniently applied to a number if different tasks, e. g. the
obtaining of fermentation products, the study of metabolism and phenomena
of 'Mutability. lu particular in our laboratory continuous flow cultures of
Propionibact:-81terminii were used to Prepare vitamin B12. Under the conditions
60
of continuous vegetative reproduction the culture of Cl. ocelot titylicum was
artificially adapted to higher concentrations of butylalkohol. A knowledge
of the laws governing the growth of continuous flow cultures helps to draw
a boundary between accumulation of individual resistant types on one hand,
and the massive adaptive capacity of cells on the other hand. ln our experiments
with adaptation to butylalcohol this last phenomenon seems to play an import-
ant role.
61
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Ap roved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
A STUDY OF THE PROCESS OF 'DEVELOPMENT
OF MICROORGANISMS BY THE CONTINUOUS FLOW
AND EXCHANGE OF MEDIA METHOD
N. 1). JEKUSALIMSKIJ
There exist two basic points of view concerning the causes of development
and ageing of microorganisms. Whilst some authors attribute a decisive role
to intracellular processes, others maintain that neither progressive develop-
ment nor ageing would be possible without corresponding changes of external
conditions.
It is rather difficult to differentiate between the effects of inner and external
factors of development on usual media, which are continually changing due
to influence of the activities of microorganisms. Thus special cultivation methods
have to be used, among others also methods of continuous flow and exchange
of media. While the first method enables microorganisms to be maintained for
long periods under stable media conditions, the second method makes it possible
to change the medium as required and to regulate its composition.
Studies regarding this problem have been carried out in our laboratory
(E. A. Rukina and G. V. Pinaeva) with spore-forming bacteria: anaerobic
(Cl. saccharobutyricum and Cl. acetobutylicum), aerobic (Bac. megatherium) and
facultative aerobes (Bac. acetoethylicus). Spore-forming bacteria are advanta-
geous because with them the developmental cycle can be more clearly demon-
strated than with asporogenous bacteria. The sporulation process is accom-
panied by extensive intracellular transformations with lysis of a considerable
portion of the cytoplasm. It may be assumed that only stable hereditary
properties can be transmitted either through spores or through reproductive
cells, but not reversibly adaptive changes; or changes due to ageing of the cells,
both occurring in the course of the individual life of the cell. On the other
hand, during multiplication by vegetative division of cells, these changes
are entirely transmitted to daughter cells. We have demonstrated these
differences on the synthesis of adaptive enzymes required for the fermentation
of iylose by Cl. acetobutylicum.
62
After inoculation of the culture from a glucose medium on xylose a lag-phase
can be observed, related to the synthesis of these enzymes. During the following
transfers by vegetative cells the culture grows on xylose medium without a lag
phase, the capacity for the formation of adaptive enzymes being preserved.
However, after sporulation this capacity is lost, as may be judged from the
fact that the lag-phase again reappears.
Also changes of size are not transferred through spores. There are data
in literature showing that as the result of prolonged vegetative multiplication
spore-forming bacteria weaken and degenerate, but after sporulation their
activity is restored. The studied culture of Cl. acetobutylicum degenerated
after several dozen transfers.
However, such degeneration due to ageing is not unavoidable. We cultivated
Cl. acetobutylicum in a continuous flow of stable and suitable medium for some
200 days. During that period more than 4000 vegetative generations were
formed, corresponding to 400-600 transfers on normal, non-exchanged media.
In spite of this prolonged vegetative multiplication the bacteria fully maintain-
ed their capacity of normally fermenting and of sporulation. Therefore the
degeneration of a culture under normal conditions of cultivation can be explain-
ed by unsuitable changes of the media.
Under conditions of continuous cultivation in the presence of increasing
amounts of butanol a culture of Cl. acetobutylicum changed, becoming more
resistant to this toxic substance. The original culture withstood only 0.8%
of butanol: after 21 days of growth in the presence of 0.6% butanol, it with-
stood up to 1% of butanol. This increased resistance to butanol was unstable
and ceased after the cells had changed to spores, and these proliferated in the
absence of butanol. In the course of further adaption the resistance to butanol
increased still further and became constant. In the course of 200 days the
bacteria were able to withstand more than 2.5% of butanol and this level of
resistance was then transferred through the spores to further generations.
Thus the resistance began to show a hereditary character.
It follows that the same bacterial culture may under certain conditions
weaken and degenerate, while under different conditions it may remain
unchanged or may even change its hereditary character.
The process of sporulation is also possible only under defined conditions
of the medium, or, more exactly, under defined changes of these conditions.
This problem was studied in detail in Cl. saccharobelyricum.
The culture was grown in collodium sacs, immersed in different media which
could be exchanged for other media when required. The following variant
proved optimal: at first bacteria were grown for several hours on a rich medium
containing yeast hydrolysate; the medium was then exchanged for phosphate
buffer containing 0-3-0.5% of glucose and 0.2% of acetate, and finally, this
second solution was exchanged for tap water. Under these conditions about
63
STAT
npriaccifien in Part - Sanitized Copy Approved for Release
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
4
- - ? - - _ - - - ? .-
75% of vegetative cells were transformed into pro-spores (clostridia), and 75%
of the latter transformed into mature spores. The spore concentration amounted
to 600 x 104 per ml medium. Such a high concentration of spores could not
be obtained with unexchanged media, natural substrates included.
With the use of other variants spore formation proceeds less intensively,
e. g. when the second medium (glucose-acetate-phosphate buffer) is enriched
either by the addition of a source of nitrogen or when acetate is omitted, or
when the culture is transferred directly from the rich medium to water with
omission of the second medium. On cultivation of the culture in collodium
sacs with a continuous flow of a rich medium practically no spores are formed;
the total number of cells reaches, however, an unusually high number, i. e.
7000x 104 cells per ml medium. It should be mentioned that of the above
number of cells only 100-1200 x 106 cells per ml are viable.
The spore formation in Cl. acelobutylicunt, grown in collodium sacs, is also
stimulated by a lack of nitrogeneous substances. Butanol in a concentration
of 0-3% exerts an additional favourable effect.
Thus the above experiments confirm the view that spore formation is,
generally speaking, correlated with a lack of nutrients and to some extent
also with the accumulation of products of metabolism in the medium.
It is clear that on media poor in nutrients spore formation will proceed
more intensively than on rich media. On agar media, where in the vicinity
of the cells a lack of nutrient substances may easily occur, spore formation
proceeds more rapidly than in liquid media. This spore formation is still
further increased by placing underneath the layer of nutrient agar a layer
of agar without nutrients; the nutritive components then gradually diffuse
into this layer from the one placed above.
Thus, a 5-day culture of Bac. inegatherium on a synthetic medium with agar
contained 50% spores when cultivated in the presence on inorganic nitrogen,
15% spores on a medium containing also hydrolysate of casein and only 5%
spores on a meat-peptone medium.
When the same media were underlayered by a solution of agar in water
only, the number of spores was increased to 70%, 70% and 10% respectively.
The culture of Cl. saccharobutyricum forms spores intensively on a rich
nutrient agar, which is underlayered by an agar solution in water. On a rich
nutrient agar this culture, however, forms only prospores (clostridial forms),
which consequently lyse due to the absence of conditions required for the
conversion into mature spores.
The developmental process is irreversible. This conclusion is borne out by the
fact that clostridial forms may either change into mature spores or die off,
but cannot revert to the vegetative mode of multiplication. This can also be
demonstrated by following the changes of the number of cells of a developing
culture of Cl. saccharolnayricum using two methods: direct counting of cells
64
npriassifipri in Part - Sanitized Copy Approved for Release
and counting of colonies on Petri dishes. When the culture consists of young
vegetative cells, both methods give identical results. When the vegetative
cells begin to lose their viability or change to clostridial forms, the number
of colonies is considerably smaller than the total number of cells. At a further
stage unviable vegetative cells lyse and disappear; the total number of the
population correspondingly decreases. However, the number of colonies grown
on Petri dishes somewhat increases. This observation may be explained as
follows: clostridia, themselves unable tu multiply, are converted to spores
which, after being transferred on to agar, can germinate and form colonies.
The above mentioned experiments were carried out with populations of
different history: some of them divided, others converted to spores, and still
others died off. It is known that the progeny of even one cell may greatly
differ within a short period of time. The formation of these differences may
depend first on the fact that individual cells may develop under non-identical
microzonal conditions of the medium; secondly, in the course of the division
two physiologically different cells with different properties may form. Some
authors suggest that differences in the qualities of cells appear to be accidental
and not of determined character. Notably the appearance of spontaneous
mutations is explained in this way. Others maintain that such qualitative
differences are a determined and necessary phenomenon, arising from unequal
cellular division. According to this view of two just divided cells one always
appears to be the mother cell, the other the daughter cell. (Kolbmiiller, Bisset,
Pennington, Malek, Streginskij and others).
In order to be able to follow the development and ageing of individual cells
a special microvessel was used. This consists of a flat flask with two tubes
attached for the inflow and outflow of the medium. The vessel is placed on the
microscope stage. The vessel is closed by a glass cover on the lower surface
of which bacterial spores are placed; the spores are fixed to the surface of the
glass cover by a very thin layer of cellulose in the same way as is in the case of
electron microscopy. The vessel is filled to the top with nutrient medium which
may be changed, if required, for a medium of the same or different composition.
Using these vessels, cultures of Cl. saccharobulyricum and Bac. acetoethylicus
were studied. The experiments confirmed that bacterial spores are hetero-
geneous. This follows from their morphology, resistance to heat and rate of
germination in the microvessel. Differences in the vegetative cells grown from
these spores were also found. Not seldom was it found that of two vicinal
cells one changed into a prospore, while the other remained in the vegetative
state and later lysed. Differences in the fate of the cells might possibly be
explained by the assumption that some of them were mother-cells, while others
were daughter-cells. Cells in groups convert more easily to spores than isolated
cells. It appears to be clear that the microzonal conditions surrounding the
cells are not identical and also affect the development of the cells.
5?Symposium
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
65
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
Simultaneously with this unexpected individual variability of the cells,
determined changes of the whole mass were found, depending on the composi-
tion of the nutrient medium. Thus, when filling the microvessel with a rich
nutrient medium, vegetative multiplication occurred until the whole micro-
scopic field was practically full of cells; under these conditions no spore form-
ation was observed.
If, however, at the correct time the rich nutrient medium is exchanged
for a medium without nitrogeneous substances the majority of cells are con-
verted to prospores (clostridia). If now the vessel is again filled with a rich
nutrient medium, the remaining vegetative cells start to multiply again, whilst
the clostridia die off and undergoe lysis. On the other hand, after sufficiently
long maintenance on nitrogen-free medium and its subsequent exchange for
water, the undamaged vegetative cells weaken and undergo lysis whilst
clostridia convert to mature spores. Under appropriate conditions these spores
may again germinate and change to vegetative cells if the vessel is once again
filled with the rich nutrient medium. In this way the developmental cycle
of the bacteria can be repeated.
The results of the experiments described above show that the physiological
inequality of cells, due to inequality of division, does not determinate their
further fate. Together with the physiological state of growth of the micro-
organisms the conditions of the external medium play an important role.
Neither changes of the life cycle nor the weakening and death of cells appear
to be processes occurring spontaneously and autonomously.
66
- - -
GLYCOGEN FORMATION IN CONTINUOUS CULTURE
OF ESCHERICHIA COO B
T. HOLME
Many industrial fermentations are based on specific nutritional deficiencies
(Foster, 1949, Underkofler and Hickey, 1954). These processes represent a type
of fermentation which depends on the existence of an inverse relationship
between the growth rate of a microorganism and the synthetic rate of a certain
product of its metabolism. Thus, an essential substance in a nutrient medium
may be supplied in a concentration which limits the growth of a microorganism
at a point, where optimal conditions for the synthesis of a desired product
still are maintained. This production may proceed for a long time after growth
has stopped.
It seems not to be generally realized, however, that a fermentation of this
type may be well suited for a chemostatic continuous process. In the experi-
ments to be presented here, a system of the above-mentioned type, namely
glycogen synthesis in Escherichia coli, has been investigated with the continuous
culture technique. In addition, the formation of some extracellular substances
in the continuous cultures has been studied. The continuous culture system
used was based on the principles given by Monod (1950) and Novick and Szilard
(1950).
GLYCOGEN METABOLISM IN ESCHERICHIA COL1
Escherichia coli contains an alkali-stable polysaccharide which belongs
to the class of glycogens (Palmstierna, 1956). In rapidly growing cells the
amount of glycogen usually will not exceed 2% of the dry weight. However,
a rapid accumulation of glycogen occurs in cells which are subjected to nitro-
gen starvation (Holme and Palmstierna, 1956). Figure 1 illustrates an experi-
ment, where cells of E. coli B were inoculated into a synthetic medium which
contained a limiting concentration of the nitrogen source. After an initial lag
and a short multiplication period, exhaustion of the limiting factor prevented
5.
?
67
STAT
narlaccifiinri in Part - Sanitized Com/ Approved for Release
z
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
further synthesis of nitrigenous compounds. During this period of constant
population density, glycogen accumulated in the cells.
In similar experiments it could be shown, that during this period of constant
population density (measured as bacterial nitrogen per unit volume of culture)
25
20
15
10
1 2 3 4 5
1
6 7 8 8
Fig. 1. Glycogen accumulation and utilization during growth in a culture of E. coli B. Cells were
inoculated into a nitrogen-deficient medium with sodium lactate as the carbon source and ammo-
nium chloride as the sole nitrogen source. The initial concentration of ammonium nitrogen was
10.5 mg per litre. At the moment indicated by the arrows, ammonium chloride was added in
excess.
A ? Protein N or glycogen (mg per litre of culture); B ? Time (hours); Curves: 1 ? Protein N;
2 ? Glycogen.
glycogen reached a final value of 20-25% of the dry weight of cells. The
accumulated glycogen was responsible for at least 90% of the increase in the
dry weight of _cells during this period.
In the experiment recorded in Figure 1, an amount of the nitrogen source
was added immediately before the maximum content of glycogen was attained,
, _., -
to give the Sallie concentration of this factor as in the complete medium. The
addition caused an immediate resumption-of growth. During this growth period
the glycogen decieftied rapidly.
_
68
npriassifipri in Part - Sanitized Copy Approved for Release
?
CONTINUOUS CULTURE EXPERIMENTS
EQUIPMENT
A culture vessel consisting of a 2-litre Erlenmeyer flask was given a working
volume of 1 litre by means of a small exit tube on the wall for removing the
emergent bacterial culture. Mixing and aeration was achieved by placing the
flask on a rotary shaker and letting the air stream enter the culture fluid through
a sintered glass disc as close to the wall of the flask as possible. The neck of
the flask was sealed with a rubber stopper, which was perforated by glass
tubes for the supply of air and fresh medium and for inoculation. Effluent
air left the flask only through the exit tube mentioned, making the transfer
of the emergent bacterial suspension to a cooling flask a matter of seconds.
A hose pressure pump was used for the continuous supply of fresh medium.
MEDIA
Two media were employed: one with sodium lactate (Friedlein, 1928) and
the other with glucose (Hook et al., 1946) as the carbon source. Ammonium
chloride was the sole nitrogen source in both of the media. This nutrilite was
used as the limiting factor, and supplied in concentrations ranging between
50-200 mg per litre in different experiments, corresponding to a nitrogen
content of 13-52 mg per litre.
RESULTS
In nitrogen-limited continuous cultures of E. coli 13 steady-state growth was
established at dilution rates of 0.13-0.94 hr-1. (Holme, 1957 a and b). The
glycogen and nitrogen content of the cells and the dry weight were determined
under these steady-state growth conditions. The results are summarized in
Figure 2, where the amount of glycogen formed in one hour by a continuously
growing population corresponding to 50 mg bacterial nitrogen is recorded.
It is seen that the amount of glycogen formed in one hour increased when the
dilution rate was reduced. This means that there exists a negative correlation
between the rate of glycogen synthesis and the rate of synthesis of nitrogenous
compounds in the cells. The production of glycogen in the continuous cultures
could thus be increased about 3-fold in the lactate medium and about 1.5-fold
in the glucose medium by a reduction in the dilution rate from 0.8 to 0.2 hr-'.
The glycogen content of the cells increased in the same interval from 2-3%
to about 20% of the dry weight of cells.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
69
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
Pi
Figure 2 also shows that rapidly growing cells have higher glycogen content
in a glucose medium than they do in a lactate medium.
As a rough method of estimating the "overflow metabolism" the ultra-violet
absorption of the cell-free culture medium was determined. The absorption
increased considerably with decreasing dilution rate. This absorption in the
ultraviolet region mainly had the character of an unspecific end absorption.
It did not depend on accumulation of nucleic acids in the culture fluid, but
A
25
15
10
"0
0.2
0.4
0.6
Fig. 2. Glyoogen formation in nitrogen-limited continuous cultures of Z. coli B. The values are
recalculated on basis of a population density corresponding to 50 mg bacterial nitrogen per litre
of culture. The culture volume was 1 litre.
A.? Glycogen (ng/hr.); B ? Dilution rate (hr-1); Curves: 1? Carbon source: glucose; 2? Carbon
source: lactate.
was probably the result of an accumulation of a variety of compounds, the
nature of which is unknown. These compounds were dialyzable to at least 90%
as measured by the ultra-violet absorption.
The culture fluid in these nitrogen-limited cultures contained high levels
of keto-acids, as determined by the method of Friedmann and Haugen (1943).
Paper chromatography of 2,4-dinitrophenylhydrazones of the keto-acids
according to Civallilli et al. (1954) revealed that the main portion of the isolated
- , -
cO400c18. "consisted of tz-ketoglutaric aeid. Small amounts of pyruvic acid
were'aliio present. The lorthation of these substances has only been studied
in culture grown with lactate' as the carbon ciiurce.
In the 'upper graph of FigUre 3 is 811-0*-1c1 the formation of keto acids, calcul-
70
narlaccifii=r1 in Part - Sanitized Com/ Approved for Release
ated as cc-ketoglutaric acid, in the continuous cultures. It can be Seen, that the
rate of synthesis of these substances Seems to be constant at dilution rates
ranging between 0.2 and 0.7 hr-', decreasing at dilution rates lower than
0.2 hr-1. A similar calculation may be made on basis of the determinations
of the ultraviolet absorption of the culture fluid at different dilution rates.
A
IS
10
5
01
0.10
0.05
Ime
?
?
0.2
0.4
0.6
Imm
?
?
a
?
?
?
0.2
0.6
Fig. 3. Formation of extra-cellular substances in nitrogen limited continuous cultures of E. coli B.
Population density corresponding to 50 mg bacterial nitrogen per litre of culture. Tho culture
volume was 1 litre. Upper graph: keto acids calculated as cc-ketoglutarie acid. Lower graph:
substances causing absorption in the ultra-violet region
A ? Keto acids (mg per hr); B ? Dilution rate (hr-1); C ? Optical density (at 265 mix).
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
71
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
?
If it is assumed that the absorption is caused by a variety of overflow inter-
mediates in a constant proportion, and that the optical density is proportional
to the concentration of the metabolic products, the optical density at a given
wave-length may be taken.as a rough estimate of the concentration of these
products.
The values. obtained if the optical density at 265 mil is taken as a basis for
the calculations are recorded in the lower graph of Figure 3. The similarity
between the two curves' presented in Figure 3 suggests that this type of pro-
duction is not confined to the keto-acids but may also be true for other over-
flow intermediates. At the observed concentration levels, cc-ketoglutaric acid
did not contribute to the absorption at 265 Inv, to any measurable degree.
DISCUSSION
In general, this discussion will be limited to the experimental system pre-
sented in this paper, but it may be justifiable to speak of this system in rather
general terms, so that applications to other systems present themselves more
easily. The intracellular glycogen and extracellular materials formed by E. coli
in nitrogen-limited cultures will be referred to as "nitrogen-free" substances.
It appears that the formation of nitrogen-free substances in nitrogen-limited
cultures of E. coli proceeds without being inhibited by the limitation. These
compounds may be considered as "overflow metabolites" accumulating because
of the reduced rate of the synthesis of nitrogenous compounds in the cells.
Sinix, under conditions of nitrogen starvation the production of the nitrogen-
free substances is directly proportional to the cell material producing them,
a maximal productivity in the continuous culture might be expected at popul-
ation -densities approximating the maximal obtained in non-limited media.
However, two general limitations must be considered. Firstly, if the carbon
source in a nutrient medium is not in great excess, too much of the available
carbon may be used for the synthesis of cell material, and consequently too
little left for an optimal product formation. Secondly, at population densities
near the maximal in non-limited media, the synthesis of the nitrogen-free
substances is strongly inhibited, even if the carbon source remains in excess.
Unknown factors, presumably metabolic products formed during the synthesis
of cell material, may be responsible for this inhibition. Deficiency in additional
factors, such as oxygen, may also be responsible for the inhibition if too large
population densities are used.
This means, that the nitrogen limitation provides an effective tool for limit-
ing the production of. cell material at the highest population density, where
optimal.conditions.for the synthesis of a desired product still are maintained.
From 'an experimental point of view, the continuous culture technique seems
72
-
r.nr1V Approved for Release
?
to be well suited for studies on the influence of various environmental factors,
especially the nutritional factors, on product formation.
It is seen in Figure 3, that the output of the extracellular substances decreases
rapidly at dilution rates below 0.2 hr-1. The explanation may be, that there
is a certain minimal growth rate, i. c. a minimal rate of renewal of cellular
material, which is necessary to maintain maximum activity of certain enzyme
systems. This minimal growth rate may vary for different microorganisms.
At dilution rates between 0.2 and 0.7 hr-1. in the continuous cultures the
rate of glycogen synthesis increases significantly with decreasing dilution
rates (Figure 2). The rate of synthesis of the extracellular materials seems to be
independent of the dilution rate in the same interval (Figure 3). This means
that these extracellular materials are poured out to the culture fluid at a cons-
tant rate, irrespective of the growth rate of the organisms. This may depend
on an inhibition of their synthesis because of the accumulation of this new,
low-molecular weight end product. When a high molecular compound, such
as glycogen, is the end product, its influence on the equilibrium of the reaction
responsible for its.synthesis is very limited, and it may thus attain high con-
centration before the rate of its formation is inhibited.
In Figure 2 it is seen that there is a difference in the rate of glycogen form-
ation depending on the nature of the carbon source. This difference is greatest
at the high dilution rates, diminishing with decreasing dilution rates. If the
straight lines obtained from the values of glycogen output represented in
Figure 2 are extrapolated to zero, a value of the synthetic rate is obtained,
comparable with the initial rate of glycogen synthesis in batch cultures under
conditions of complete nitrogen starvation. It may be concluded that the
maximal rate of glycogen synthesis is the same irrespective of whether the
glycogen is synthetized from lactate or glucose.
A continuous fermantation has certain advantages over the corresponding
batch process, e. g. reduced size of unit at the same production rate, and greater
uniformity of the product. The conditions for the synthesis of nitrogen-free
substances in chemostatie continuous cultures of Escherichia coli now seem
well established but it should perhaps be pointed out that before general
conclusions can be drawn, investigations of other systems must of course
be carried out. In this connection seems particularly important to study if the
concentration of a given substance formed in a chemostatic continuous culture
may attain the same values as in a batch culture. At maximum output in the
continuous cultures of E. coli the concentration of the nitrogen-free substances
seemed to be somewhat smaller than. that obtained in the batch cultures.
However, the advantage of having a system in equilibrium, where optimal
conditions for the formation of a given product have been established, should
be obvious.
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
73
Declassified in Part - Sanitized Co .y Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
REFERENCES
CAvALLANI D., et al., 1949: Determination of Keto Acids by Partion Chromatography on
Filter Paper. Nature, 163: 568.
Fotirrma J. W., 1949: Chemical Activities of Fungi. Academie Press, New York.
FRIED/MANN T. E., HAUGES 0., 1943: Pyruvic Acid. II. Determination of Kato Acids
in Blood and Urine. J. Biol. Chem., 147: 415.
Fiukoimbr F., 1928: Quatitative Metabolism of Paratyphoid B, Colon and Phyocyancus
Bacilli. Biochem. Z., 164: 273.
HoLitk T., I957a: Continuous Culture Studies on Glycogen Synthesis in E8cherichia coli B.
Acta Chem., Scand., 11: 763.
Haulm T., 1957b: Bacterial Synthesis During Limited Growth. M. D. thesis, Almqvist
och Wikselbi Boktryckeri AB, Uppsala.
Emma T., PALMETIERNA H., 1958: Changes in Glycogen and Nitrogen-containing
Compounds in Escherichia coli B During Growth in Deficient Media. Acta Chem.
Scand., 10: 578.
Hook A. E. et al., 1946: Isolation and Characterisation ot the T, Bacteriophage of
Escherichia coll. J. Biol. Chem., 131: 211.
Morton J., 1950: La technique do culture continuo. Theorie et applications. Ann. Inst.
Past., 79: 390.
NORTHROP J. H., Anus L. H., MORGAN R. R., 1919: Tho Fermentation Process for the
Production of Acetone and Ethyl Alcohol. J. Ind and Eng. Chem., II: 723.
NovIek A., Szmutn L., 1950: Experiments with the Chemostat on Spontaneous Mutations
of Bacteria. Proc. Natl. Acad. Sci. N. Y., 36: 708.
PALEISTIERNA H., 1956: Glycogen-like Polyglucose in Escherichia call B During the
First Hours of Growth. Acta Chem. Scand., 10: 567.
UNDERKOPLER L. A., HICKEY R. J., 1954: Industrial Fermentations. Volumes I and II.
Chemical Publishing Co., New York.
74
CONTINUOUS CULTURE TECHNIQUES
J. ItIICA
In the course of development of culture techniques enabling us to study
the life activity of microorganisms ever increasing stress is being laid on the
perfect culture technique that can make use of all possibilities afforded by the
progress of improvement of apparatus and equipment and thus fulfil to the
utmost the requirements dictated by the present progress of microbiology.
Every type of microbiological work requires a corresponding culture technique,
be it a diagnostical work, a genetic, selection or growth study, or an investiga-
tion of physiological and biochemical problems, or the synthesis and transform-
ation of some product. The reproducibility of work on biological material is
undergoing improvement owing to mechanization and automatization of
culture apparatus. Its accuracy has been further increased by introducing
physical and physico-chemical aspects and by applying them to culture
techniques.
We shall omit the conventional static culture methods. An important
improvement has been achieved here by the introduction of stirring. Movement
ensures a better supply of nutrients to the cell. Organisms respond to it by
intensified metabolism and growth. Therefore ingenious shaking machines
have been developed based on the rotational or reciprocal principle. Duo to
disturbing of the liquid surface a more active gas absorption is achieved and
thus growth and metabolism are further stimulated. To produce a similar
effect tall and slender cylinders are used occasionally in which the stream
of entering gas is dispersed by some sort of atomizer at the bottom of the vessel
and passes upward in the form of tiny bubbles mixing with the liquid and
saturating at the same time the medium with gas.
For some special purposes rotating drums have been constructucd in which
the liquid is stirred by rotating the whole vessel along its longitudinal axis
either vertically or inclined to a certain angle. As the volume of medium is
small in comparison with the whole vessel, the liquid can spread over the large
surface of the vessel walls when rotating. Thus effective contact with the enter-
75
STAT
narlaccifiPrl in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
ing gas can be achieved; in most cases with oxygen from the air. An increase
of partial pressure sometimes aids its dissolving.
For highly aerobic processes and in order to utilize better the space of the
vessel fermentors have been developed in which stirrers are used dispersing
the under pressure entering gas into minute bubbles. Due to the turbulent
motion of the liquid in the whole volume oxygen is dissolved more easily and
transported rapidly to the powerfully respiring cells of the aerobic micro-
organisms. The volume of passing gas is measured by flowmeters of various
types.
Experiments have been carried out in which oxygen is supplied by electro-
lysis of the aqueous nutrient medium (Sadoff, Halvorson and Finn, 1956).
To determine the parameters controlling or reflecting the course of the pro-
cess, culture vessels are supplemented with physicochemical analysers by
means of which the content of 02 and of CO, can be determined either in the
escaping gas or directly in the culture medium. Also electronic equipment
can be provided which automatically registrates and regulates the pH, the tur-
bidity of cultures, flows of gases and liquids, temperature, etc., and controls
further auxiliary apparatus.
Cultivation of microorganisms can proceed either in the form of a batch
culture or continuously ? in identical culture vessels, the only difference
being that in the continuous type equipment is provided by means of which
the nutrientmedium is continuously added and withdrawn.
In batch, culture the substrate is changed by the actively metabolizing
microorganisms and the slightly modified medium thus affects reciprocally
the microorganisms themselves. The rate of the whole process is determined
by innumerable biochemical and biophysical reactions, limited first of all by
the concentration of available nutrients in the substrate and by the size of the
vessel. The process is a dynamic one in a statically closed and unchanging
spat* and changes with time. It is extremely difficult under these conditions
to study the individual reactions, particularly the irreversible ones, and to
apply the laws of thermodynamics to them. As the nutrients are exhausted
and metabolites ccumulates the rate of the process is slowed down until it
reaches the point when no free energy is transformed (dF = 0); then the pro-
cess stops and an equilibrium is established. By applying this characteristic
we can consider the batch process as a closed system.
The living cell, however, belongs fundamentally among open systems.
If we wish to study the physiology of the fully developed cell and to follow
the kinetics of one-step and of multi-step complicated reactions it is necessary
to cultivate inierOorganisnis under the conditions of open systems. These
conditions are met only in continuous flow culture processes which range among
the ? open _systems.
.
The theory Of Idynamics and of yields of irreversible processes in open
76
narlaccifii=r1 in Part - Sanitized Com/ Approved for Release
?
chemical systems in a steady state, as determined by the reaction rate either
for one or more reactors connected in series, has been well developed and stu-
died and its results have aroused the great interest of many biologists.
So Pasynskij (1957) deals with the question of the application of classical
thermodynamic equilibria of closed systems as well as of the physicochemical
relationships of simple reactions of open chemical systems to the cell. Fie
maintains that they cannot fully interpret the kinetics of chain reactions
forming the basis of complicated physiological functions. Pasynskij states
further that the simplest form of life (the biological open system) differs from
the most complicated chemical reaction of an open system in having the
capacity of self-preservation and auto-reproduction and in directing all the
complicated chemical reactions to recreate the given system. He claims that
it is impossible to design a chemical apparatus as a result of chemical changes
in which a similar apparatus or its part would function. Pasynskij, however,
does not stress sufficiently the fact that into a continuous culture vessel
(a chemically open system) a living microbial cell (a biologically open system)
can be introduced:
This type of open system is characterized by the fresh nutrient medium
being added to the culture vessel and by removing at the same rate the medium
modified by the metabolic activity of the organism, together usually with
a part of the propagated organisms. The medium, the volume of which is
kept constant, must be stirred thoroughly in order that the concentration
of reacting factors (i. e. nutrients and cells) may be identical in the whole
vessel. Considering the auto-reproductive capacity of microorganisms, the
whole process can be compared to an auto-catalytic reaction proceeding at
a constant rate. If the rate of inflow of fresh nutrients is equal to the total
rate of their utilization the concentration of nutrients becomes stationary.
If the specific growth rate is equal to the dilution rate and if both the concen-
tration of nutrients and the dilution rate which must not exceed a certain
critical value, remain constant, a steady state is established which is, under
certain conditions, capable, of auto-regulation. This is a dynamic steady state
and not an equilibrium because the transformation of free energy proceeds
at a constant rate (dF = const.).
In a precisely defined steady state of any chosen quality the application of
results from chemical open systems is facilitated and a mathematical inter-
pretation of the kinetics of the individual reactions is simpler.
The application of the fully developed dynamics of propagation of micro-
organisms and of the formation of products is not only valuable for theoretical
studies, but is of farreaching practical importance.
The continuous system has the following advantages as compared with the
batch process:
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
77
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
1. The course of reactions can be studied in a certain chosen phase of the
process under constant steady state conditions in which the time factor is
eliminated. The influence of nutrients, temperature, pH, stirring, aeration
and other factors on the course of reactions can be determined more easily
and without complications clue to the incessantly changing conditions.
2. The mathematical treatment of the course of the process is simpler.
3. The product is more homogeneous.
4. There is a greater possibility of automatizing the whole process and of
regulating it by a suitable application of physicochemical methods.
5. The production is more economical due to two factors: the dimensions
of the given apparatus can be decreased while the same production is preserved,
the cultivation being more physiological and thus more efficient; the processing
time can be decreased while the dimensions of the apparatus remain unchanged.
6. The cost of construction of a well-operating equipment is somewhat
higher than in the case of the batch process, but as soon as the continuous
process is in operation, the cost of production per unit weight of product is far
lower.
7. The preparation of the medium and obtaining of the product are more
efficient and more economical.
The idea of utilizing the continuous processes in various microbiological
fields is not a new one. We have now very extensive literature on this subject
which cannot be grasped in its whole. As early as at the beginning of this
century the possibility of producing ethylalcohol continuously by means of
fermentation was indicated. Later on a number of papers and patents appeared
dealing with this fermentation process and it is actually the only process that
has found wide application in industry. To a small -extent the continuous
production and baker's yeast has been used. Somewhat later also other branches
of microbiology started to use continuous culture for experimental purposes.
But since then the majority of continuous culture processes has been developed
empirically and not always with full comprehension of the principles of open
systems, and therefore the experimental results have not been applied signi-
ficantly in practice. It may also be due to the fact that at that time the culture
technique in industry had not reached its present level, particularly with
respect to the ensuring of sterility, continuous preparation of the nutrient
medium and automatization of the whole process.
It was only in the middle of this century (Monod, 1950; Novick and Szillard,
1950a, b) that the continuous methods were laid on a solid theoretical basis,
particularly due to the mathematical treatment of the chief growth prin-
ciples in open systems.
Further_ theoretical aspects were published by Adams and Hungate (1950),
Gole (1953), Finn and Wilson (1954) and Northrop (1954); these authors,
78
however, did not stress sufficiently the mutual relationship between the sub-
strate concentration, the rate of dilution and the specific growth rate. Novick
(1954, 1955), Spicer (1955), Malek (1955) published extensive papers of a more
general character denoted to the questions of continuous culture methods.
Danckwerts (1954) treated the main principles of chemical flow systems and
limited the possibilities of individual types of flow reactors for certain types
of reactions. Some of his conclusions can be applied also to continuous flow
cultivation. Saeman, Locke and Dickerman (1946), Maxon (1954), Katsume
(1956) studied the problems of industrial continuous fermentations mostly
with yeast organisms and Tamiya (1957) with algae.
The theory has been developed in a number of other papers. Herbert,
Elsworth and Telling (1956) undertook a mathematical analysis of the kinetics
of bacterial growth and changes in the concentration of microorganisms on the
basis of a minimum number of very simple postulates. They treated again the
steady state problem and the effect of various rates of dilution at various
substrate concentrations in the inflowing medium. They supported their
theoretical conclusions by practical results. Powell (1956) treated mathemati-
cally the relationship between the rate of growth and the percentage occurrence
of cells of different ages and of different generation times in cultures growing
under continuous flow conditions. He discussed the various factors leading
to the establishment and preservation of a steady state. Pirt (1957) derived
formulae for calculation of the oxygen intake in a continuous culture.
The perfection of the theory resulted in the production of new and better
laboratory equipment of various types (1-lean, Holme and Malmgren, 1955;
Fox, Szilard, 1955; Elsworth, Meakin, Pirt and Capell, 1956; Anderson, 1956;
Perret, 1957).
The methods employing the continuously operating culture apparatus
can be divided into four groups.
I. METHODS BASED ON THE PRINCIPLE OF CIRCULAT1 ON
The medium flowing out of the stock vessel and returning to it again circul-
ates constantly through the cultivation vessel. Circulation of the liquid is
effected by a pump or, with most types of apparatus, by an air lift, which at the
same time aerates the liquid. Through the metabolic activity of the organisms
the nutrient level decreases, the number of cells increases and metabolites
accumulate. In spite of the fact that the apparatus operates continuously the
process takes place batchwise and corresponds in character to the closed
systems.
It is known in two basic types:
a) The organism grows on a fixed carrier and some of the free cells circulate
with the liquid.
79
noriaccifiarl in Part - Sanitized Copy Approved for Release
Declassified in Part - Sanitized Cop Approved for Release
A typical example of this type is the production of vinegar in vinegar vats.
The same principle was used by Hermann and Neuschul (1935) who studied
the formation of gluconic acid by bacteria growing on wood shavings. Audus
(1946), Lees (1949), Temple (1951) and Dubash (1956), who designed several
types of perfusion apparatus studied the metabolism of microorganisms
growing on soil particles by means of the percolation technique. Pasynskij
and Nejtnark (1952) caused the microorganisms to adhere to pieces of sintered
glass in their circulation fermentor. Di tnopoullosos and Pritham (1951) cultivated
animal tissues fixed in a percolator. Darlington and Quastel (1953) used a double
perfusion apparatus for studying the passage of various substances through
the wall of a still living piece of intestine.
b) The organism grows dispersed in the medium. Lundgren and Jennison
(1955) (Beesch, Shull, 1956) used the air lift pump principle in their laboratory
fermentor and Ashton and Holgate (1951) patented a circulation equipment
for the production of streptomycin. Tamiya (1957) mentions several methods
of cultivation by circulation for the propagation of algae. Krauss and Thomas
(1954) harvested algae by means of a continuously operating centrifuge and
returned the supernatant to the cultivation trays. Boeckeler (1948) carefully
separated alcohol from the circulating fermentation liquid in a stripping tower
and returned the liquid into the fermentor together with non-damaged cells.
IL FEEDING METHODS
The substrate is fed at a certain rate, either periodically or continuously, to
the growing culture. Only those types are of importance which start with
a small volume of medium containing a certain concentration of microorganisms
in the cultivation vessel and maintaining during the period of feeding a station-
ary concentration of cells and nutrients per unit of volume in harmony with
the rate of growth. In this phase the rate of the process does not change in
time and resembles the steady state of open systems but washing out factor
is not represented here. Only the space and the amount of added nutrients
change. As soon as the desired volume is reached feeding is stopped. From
this moment on the cultivation process becomes a closed system changing
with time in a constant space and ends like any other batch process.
The feeding types of cultivation are most widely employed in the ferment-
ation industry. But most of the feeding schemes used to-day for yeast propaga-
tion
-
have been designed, on an empirical basis and do not always fully cor-
respond to the specific growth rate.
The various feeding systems are in extensive use also for the propagation
of a great- mass of algae (Krauss, 1955; Tamiya, 1957). Olson and Johnson
80
-
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Dmr+ - RflifI7d CODV Aooroved for Release
(1948) fed slowly acid hydrolysed wheat mash to a culture of A. aerogenes
producing 2,3-butylenglycol so that the sugar concentration in the medium
was kept constant.
We know of experimental cases when the substrate was added periodically
or continuously to a considerably grown culture. The inflow was started only
when a certain phase of the cultivation process had been reached. These cases
are, however, only modified examples of the batch system.
A similar type was used for the cultivation of fodder yeast by Tyler and Maske
(1948) (Lee, 1949) who added a constant amount of nutrients into the recircul-
ating liquid and thus combined the circulation and feeding systems. Davey and
Johnson (1953) and Hosier and Johnson (1953) studied in their laboratory
fermentor the effect of several saccharides on the production of-penicillin
by adding them periodically or continuously in certain constant amounts to
a mold culture in which they kept the pH constant. Soltero and Johnson
(1954) studied the formation of penicillin in shake flask fermentations to which
they added glucose continuously at a constant rate. Jacob (1953) decreased
the rate of growth of E. coli by adding a small amount of urease upon exhaus-
tion of the nutrients, thus ensuring slow liberation of the ammonia from the
urea, or by adding small doses of lactase, thus ensuring liberation of the
glucose from the lactose.
III. SEMI-CONTINUOUS, PERIODIC METHODS
These processes are based on the principle of repeated and mutually depend-
ent batch cultures placed in one or more vessels connected in series. The
purpose of this process is long-term exploitation of the culture and its conti-
nuity is considered strictly from the industrial production point of view. Accord-
ing to the characteristics of the organism and the type of cultivation several
modifications have been employed so far:
a) repeated use of a grown culture. When molds are grown in surface tray
cultures the mycelium cakes are washed underneath with fresh substrate
after the fermented medium has been removed. This type was used in the
production of citric acid, but its application in surface fermentation of peni-
cillin on trays or in a horizontal fermentor Abraham (1941), Haller (1950)
(Brinberg, 1953) was not particularly successful in practice.
b) A part of the culture can be used as inoculum after the completion of
fermentation. This type is widely employed in the fermentation industry where
a part of the separated yeast cells are returned as the inoculum of a new fer-
mentation process. Garibaldi and Feeney (1949) left a part of the culture
in the fermentor on completion of the fermentation process as inoculum for
a further fermentation process when producing subtihn. They repeated this
6-S y m p 031 u m
50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
81
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
process several times, observing no decrease in the production of the anti-
biotic.
c) A part of the culture is removed at a given growth phase and the rest
is supplemented with fresh.substrate to the original volume. The removed
part of the culture continues to ferment in one or several other flasks.
To meet the needs of the fermentation industry several modifications were
developed, e. g. for the initial propagation of yeast (Castor and Stier, 1947),
for yeast production (Meyer, 1929; Daranyi, 1936; Pig, 1956), for gluconic acid
production (Porges, Clark, Gastrock 1940), and for fodder yeast production
(Leopold and Fend, 1955). Wakaki, Ishida, Yamagushi, Mizuhara and Masuda
(1952), Foster and McDaniel (1952), and the Schenley Ind. Co. (1953) tried to
produce penicillin by a semi-continuous method and the Distillers Co. (Perl-
man et at., 1952) streptomycin. An analogous culture technique can be applied
in the production of algae as shown by Tamyia (1957) in his general report.
IV. CONTINUOUS FLOW METHODS
The apparatus for continuous culture techniques consists usually of a reser-
voir of sterile nutrient medium, of a device effecting the inflow of the nutrient
medium into the culture flask and an overflow device which ensures a constant
volume of the culture in the growth tube by enabling the microbial suspension
to escape. According to the type of process taking place in the described
apparatus, the continuous flow techniques can be divided as follows:
1. HETEROCONTINUOUS METHODS
The nutrient medium flows continuously as a unified stream through the
static space in which the microorganism grows. The composition of the medium
in the flask changes along the direction of flow according to a certain rate
gradient. If the conditions of flow rate are constant then the composition
of the medium at a certain point of the culture space approaches a constant
and time-independent value. The system as a whole does not change with time
and is characterized by changes in the arrangement of spatial coordinates.
In other words, it is spatially variant but temporally invariant.
Two modifications are applied:
a) The Organism Grows on a Fixed Carrier
Some examples belonging to this group form an exception as with them
no mechanism for keeping the volume of culture constant is used. In order
to observe Cellular division under the microscope Harris and Powell (1951)
82
t.,
constructed a small chamber in which bacteria grew on a cellophane membrane
under which the nutrient medium constantly passed. Rosenberg (1956)
observed under the microscope a lysis of a colony of St. aureus (LS2) which
grew on an agar plate covered with a cover glass. The medium was hrought
in and carried away by two strips of filter paper. Also Utenkov (1941) used
agar slant as culture carrier in one of his modifications of a microgenerator.
Savage and Florey (1950) studied the induced antagonism in a microculture
of .13. proteus which grew on a small metal spiral and was washed very slowly
by the inflowing nutrient medium containing the suspension of the microbe
against which the antagonism was to be formed. Kautsky and Kautsky (1951)
cultivated bacteria, algae and Rhodotorula on a filter paper strip in the vertical
position which was at its upper end continuously saturated with a substrate
solution. Pasynskij and Nej mark (1952) used pieces of sintered glass as a carrier
for organisms in a cylindrical flask. Northrop, Ashe and Morgan (1919) divided
a cylindrical fermentor into several zones by means of perforated plates.
On the corn cobs with which the individual zone were filled they grew B. aceto-
ethylicum which fermented cane molasses flowing in the upward direction
to acetone and ethylalcohol. 8arkov (1950) described a method of alcoholic
fermentation ofsulphite liquors in which yeasts grew in wood or metal baskets
filled with cellulose fibres and submerged into the medium in the fermentor
at a continuous flow of liquid. Kaljuinyj (1955) used yeast cells deposited
on cellulose fibres to ferment sulphite waste liquors. Clifton (1943) tried to
produce penicillin continuously in an modified vinegar vat. The mold grew
on wood shavings in a tall and narrow glass cylinder. The stream of liquid
and of air was directed downward. Swiss patent*) authors tried to use the
method for industrial production of penicillin but without practical success.
The mold was to grow on the surface of tubes, grids or various small objects
washed by the substrate solution. Jeffreys (1948) underwashed continuously
mold mats growing on the surface of a liquid and studied the production of
mold enzymes. The same principle was used for the production of penicillin
by Stice and Pratt (1946) who placed several converted trays one above the
other into a cylindrical fermentor filled with medium. Under every tray they
formed a bubble from the entering air. Thus the mold grew on the interface
of a liquid and a gaseous phase. The flow of liquid was directed downward.
8vachu1ova and Ku?ka (1956) described a simple apparatus arranged in such
a way that Mycobacterium tbc could be cultivated in it on the surface of the
nutrient medium undisturbed by the continuously passing liquid.
Moyer (1929) cultivated Bad. aerogenes and Moor (1945) cultivated some
penicillia and yeasts under highly aerobic conditions in variously adjusted
long glass tubes placed horizontally. Organisms grew on the lower side of the
*) Verfahren und Einrichtung zur Herstellung von Schirnmelpilzprodukten, insbesondere
Penicillin, Swiss Patent 279 098, 1948.
6?
n.,,ineeifiarl in Part - Sanitized Coov Approved for Release @ 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5'
83
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
tubes in a slow current of liquid and flowed out from them into a collecting
vesiel, or else they were let out periodically from the sedimentation cells at the
outflow end of the tube.
A typical example of the "piston flow" or of the "tubular flow" is the culti-
vation in which the liquid flows slowly inside a permeable cellophane tube
and the organism grows on its external surface. Lewis and Lucas (1945) grew
P. ?totalling first on the surface of a cellophane tube, but as the mold attacked
the cellophane they used a porcelain tube, in some cases covered with a cellu-
lose nitrate film. A ceramic cylinder was used for the same purpose by Beatty
(1946) (Brinberg 1953). Harmsen and Kolff (1947) obtained considerable
amounts of microbial mass on the same principle. They grew bacteria on the
surface of variously prepared cellophane tubes.
b) The Organism Grows Diffusely Dispersed in an Unstirred
Medium
This culture method has found wide application in the fermentation industry.
Most methods are based on the empirical process of substrate utilization in
a series of mutually connected fermentors. Long ago Lebedev (1915) (Lebedev,
1936) pointed to the possibility of fermentative production of ethylalcohol in this
way; later on ho developed it himself on an industrial scale. A similar type
of process of fermentation of very varied substrate to ethylalcohol is described
e. g. by Becze and Rosenblatt (1943), Altsheller, Monet, Brown, Stark and
Smith (1947), Sav6enko (1957); to acetone, butylalcohol, ethylalcohol by
Lagodkin (1939); to protein and fat yeast by arkov (1950). Seeman, Locke,
Dickertnan (1946) mentioned several types of industrial continuous ferment-
ation processes of wood hydrolysates for alcohol and yeast production. The
system of three fermentors of different sizes and in various combinations was
patented by Sak (1932). arkov (1950) described also a continuous alcoholic
fermentation process in which he increased the yield of alcohol per fermented
sugar by returning a part of the separated yeast cells into the process.
The principle of vertical cylindrical fermentor divided by variously adjusted
partitions into several zones was employed by Victorcro (1948) and by Owen
(1948) for continuous alcoholic fermentation. The former used the up-flow
of mash, the latter the down-flow when yeast cells accumulate on the upper
plates.
Further examples of continuous alcoholic fermentation were given by garkov
(1950), Maxon (1954) Gaden (1956), Katsume (1956) and Ueda (1956).
Simple laboratory apparatus were described by Saenko (1950) who tried
to adapt wine yeast to a higher concentration of alcohol in an unreliably oper-
ating mechanism, and by Rudakov (1936), and Verbina (1955) who studied the
adaptation of yeast to antiseptics.
84
nn,rimecifiarl in Part - Sanitized Copy Approved for Release
Nkan (1939) prepared bread leavening in a continuous multistage apparatus
in which two different alcoholic fermentation and lactic acid fermentation
took place.
Cotass et al. (1951), Oswald et al. (1953) (Tamiya, 1957) used specially
adapted sewage lagoons and oxidation ponds which they named "symbiocon".
They used them for con-
tinuous cultivation of
algae together with aero-
bic bacteria without any
supply of atmospheric
oxygen. The bacteria
utilized the oxygen libera-
ted by algae during pho-
tosynthesis and oxidized
the waste organic mate-
rial. The algae in their
turn photosynthesized or-
ganic matter from the
CO2 and NH3 produced
by the bacteria.
Also several bacterio-
logical papers have appe-
ared in which the possi-
bility of maintaining con-
stant culture conditions
is discussed, but these
same conditions are de-
termined by the authors
mostly empirically.
Utenkov, patented in
1922 (Utenkov, 1941) his
microgenerator in which
he studied two stages of
development of microbes,
the subcellular and the cellular one, and in his division theory he assumed
even the existence of a sexual cycle. He presumed that each stage required
special suitable conditions and arrived at the conclusion that the constancy
of these specific conditions is ensured only by a flowing nutrient medium.
In the course of later years Utenkov developed several types of microge-
nerators, one of which is shown in Fig. 1, as well as several culture techni-
ques as a part of a broader method which he called "mikrogenerirovanie"
(Utenkov, 1941, 1944). Felton and Dougherty (1924) constructed an auto-
Fig. 1. Microgenerator; Utenkov (1941).
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
85
Declassified in Part - Sanitized Co .y Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
matic transferring device in which they studied the conservation of the viru-
lence and the influence of the pH on the virulence of one strain of Pneumo-
COCCUS grown on skimmed milk. The new supply of food from the storage
bottle was regulated by an electromagnetic clamp controlled by a time-relay
so that active growth of organisms was made possible. The cell suspension
flowed out of a U-shaped growth
receptacle over on overflow si-
phon. Haddon (1928) (Novick,
1955) described a simple device
for continuous culture of micro-
organisms.
Rogers and Whittier (1930) pro-
ceeded from the analogy of a bac-
terial culture with a multicellular
organism and in order to form
analogous conditions they con-
structed a continuously operating
apparatus in which they culti-
vated E. coli and Sir. locus either
independently or in a mixed cul-
ture. The supply of nutrients was
very slow so that the content of
the culture flask was renewed once
in 24-30 hours. They realized the
importance of this method for in-
dustrial purposes and tried to
produce lactic acid in open fer-
mentors (Whittier and Rogers
1931).
The Rogers method was employed by Cleary, Beard and Clifton (1935) who
limited growth by the concentration of various nutrients and tried to investigate
the principle of the stationary phase at a very slow rate of flow of nutrients.
In the years 1934-1942 Weir (1955) formed a working hypothesis as to the
indirect proportionality between the virulence and ease with which bacteria
can propagate. He tried to solve this problem as well as that of the propagation,
M-concentration and variability of bacteria using E. coil in a simple flow culture
tube (Malek 1943) (Fig. 2).
Jordan and Jacobs (1944) proceeded from a criticism of papers concerning
the M-concentration. Their starting point and comparison standard was the
concept of static cultivation. Therefore they used a very slow rate of flow
of substrate added automatically at a constant rate by means of a special
pipette.
Fig. 2. Simple continuous culture apparatus; Malt*
(1943).
86
Barnes and Dewey (1947) described a simple laboratory device for continuous
cultivation of Si. paradysenteriae, and Duch6 and Neu (1950) for cultivation
of Dermatophyta.
2. HOMOCONTINUOUS 311:T1101)8
The process takes place in a thoroughly stirred culture flask to which fresh
substrate is continuously added at a fixed rate. in order to keep the volume
of the culture constant the fermented medium flows out through an overflow
together with an amount of propagated organisms. The medium composition
is the same at all points of the culture space as well as in the overflow. Although
a number of reactions take place here the system as a whole does not change;
the fact that some reactions must have taken place is evident from the differ-
ence in the composition of the inflowing and outflowing medium. Under con-
stant conditions of flow a complete stationary state of the culture is established
which is both temporally and spatially invariant. The growth process can be
compared to an autocatalytic reaction for which a thoroughly stirred and in
certain cases also aerated flow fermentor is the most suitable one, a fact which
is not sufficiently appreciated in cases when the fermentation liquid is not
stirred sufficiently merely by passing air.
There are two types of culture methods enabling the achievement of a con-
stant density of microorganisms.
a) Methods Based on the Principle of the Turbid istat
It is assumed that in a turbidistat microorganisms grow at their maximum
growth rate while their density is kept at a certain selected value of a controlled
flow rate.
The density of microorganisms in a turbidistat is observed directly in the
culture flask by means of physical or physicochemical methods. Most widely
used are photocells by means of which the intensity of light beam is measured
after passing through the cell suspension. Sometimes also indirect measurement
is employed, e. g. measuring the pH or concentration changes of some substance
resulting from the changes in the number of microorganisms. The device used
to measure the density automatically controls the rate of flow of the nutrient
medium. The constancy of flow and the density depend on the sensitiveness
of the detection and supply mechanism. If the organism is to grow at a maximal
rate its density must be chosen within such limits that even a small change
can be optically registered and at the same time that the nutrient capacity
of the culture medium is not exceeded. The experimenter can change the density
depending on external conditions, within certain limits, preserving at the same
time the maximal growth rate which in a turbidistat is given by the chemical
and physical quality of the medium.
87
narlaccifiPrl in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
11
Myers and Clark (1944) were among the first to employ the method of
photoelectric control of density of unicellular algae. They studied the effi-
ciency of inorganic and organic nutrient media for growth at a given rate
of flow of the inflowing medium. They controlled the dilution of growing
cultures by means of an overflow making it possible to change at will the volume
of medium and thus also the dilution rate. Cook (1951) cultivated algae (Morella
pyrenoidosa) on a laboratory scale in a tall glass cylinder, aerating the suspen-
sion by a mixture of air and CO2, and studied the influence of intensity and time
of illumination. He measured density in a side circuit through which part of
the culture passed. The fresh medium supply was controlled automatically
by a solenoid valve connected in the photoelectric circuit. He also proposed
developing industrial production in long horizontal tanks into which the medium
would enter through holes located at definite distances along the tank. Phillips
and Myers (1954) described a small apparatus in which they measured the
growth rate of unicellular algae as a function of intensity and itermittency
of illumination under conditions of constant density of population and of
constant volume. Krauss (1955) discussed the physical conditions and nutri-
tion requirements in a largo continuous production, Tamiya (1957) quoted
in his report a number of further papers dealing with the propagation of
a considerable amount of algae by means of this culture technique.
Bryson (1952, 1953) tried to obtain forms resistant to antibiotics in his
turbidistatic selector. He measured the turbidity of the culture by means of
a photoelectric system into which a device was also introduced which controlled
automatically the supply of nutrient solution when the culture reached a certain
constant turbidity. Also in this circuit a proportional-feed system was placed
which made it possible to increase geometrically the concentration of the
toxic substance in the medium in proportion to the propagation of cells.
Graziosi (1956) described an automatically operating turbidistat in which he
registered under certain conditions the rate of adaptation of Micrococcus 7)yo-
genes and Proleus vulgaris to the antibiotic novobiocin (Graziosi and Puntoni,
1957).
Northrop (1954) adapted his automatically operating turbidistat for use
with a commercial-type photoelectric colorimeter in order to study the growth
of a lysogenic strain of B. megatherium. Bacteria growing on the walls of the
culture cell in which the measurements were carried out were removed by
windshield wipers fixed to a stainless stirring rod which moved vertically,
driven by a motor. In his turbidistat he registered changes of the growth rate
under the influence of oxytetracycline and studied the rate of adaptation
(Northrop, 1957a); he also endeavoured to treat mathematically the formation
of resistant mutants (Northrop, 1957b).
Fox and Szilard (1955) constructed a turbidistatic cultivation device with
a photoeleCtric system which they called a "breeder". They kept the density
88
A
of exponentially growing bacteria at a certain va-
lue by controlling the influx of fresh nutrient me-
dium, and maintained the volume of the culture
at a constant value by means of an overflow siphon.
The operation of the breeder was controlled by
a time-relay. Every 4 minutes stirring was stop-
ped, the light source turned on, a density reading
carried out, the light source turned off, stirring
started again and the influx of the medium ad-
justed according to a pre-set density value. The
whole operation did not last more than 30 seconds.
200 V
)4
a b c 115V
? ? ?
6
II
1.5 V
Fig. 3. 3. Microbial auxanometer (diagrammatic); Anderson (1956).
A ? Growth tube; B ? Circuit of complete devicewith turbidity recording; 1 ? light, source; 2 ?
mirror; 3 ? growth tube; 4 ? phototube; 5 ? helipot; 6 ? balancing motor; 7 ? relay; 8 ? drop
recorder; 9 ? thyratron relay; 10 ? turbidity recorder; 11 ? galvanometer.
They broke up foam with a red-glowing platinum filament which was stretched
across the flask above tho surface of the liquid. Fox (1955) measured the rate
of formation of mutations in a breeder and in a chemostat with two strains
of E. coli B under a variety of steady state growth conditions.
Anderson (1953, 1956) described a turbidistatic device called a "microbial
auxanometer" (Fig. 3) which, by means of a system of photocells and relays,
registers automatically the course of the proceza and controls the process
itself according to preselected values under turbidistatic and chemostatic
89
norlaccifiPrl in Part - Sanitized Coov Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release
conditions. The inflow of the liquid into the culture flask is controlled auto-
matically and registered by mean ? of a magnetic valve and a drop counter.
The culture flask rotates while the wipers, fixed magnetically in a certain
position, clean the walls and stirr the culture.
b) Methods Based on the Principle of the ?hemostat
In a chemostat the medium flows continuously at a given rate. The volume
of the thoroughly stirred culture is kept constant by means of an overflow
system. The growth of organisms is controlled by the concentration of substrate
and each dilution or flow rate fixes the substrate concentration at a value
required for the specific growth rate to be equal to the dilution rate. Under
these conditions the specific growth rate is a function of the substrate concentra-
tion or rather of the concentration of an important nutrient which is often
called the limiting or growth controlling factor. Of decisive importance are
the quantitative chemical conditions from which the method derives its name.
The chemostatic continuous cultivation of microorganisms makes it possible
not only to select-at will any growth rate with the density unchanged, but also
to select different values of density under certain conditions.
Before the chemostatic conditions were mathematically analysed and the
controlling influence of the concentration of the substrate on the growth
of microorganisms was .classified and stressed, the continuous processes were
used in industry as well as sometimes in the laboratory rather with the aim
of full utilization of fermentable sugars. It was the metabolic activity of the
organism that was considered the leading component of the process of ferment-
ation. The mutual dependence of the rate of growth, rate of flow and concent-
ration of the substrate was not understood to its full extent. For this reason
the production processes, even though appreciating the advantages of conti-
nuous processes, were based on empiric experience. The first traces concerning
fermentation of this type appeared in the fermentation industry in the process
of production of ethanol and yeast cells from various substrates containing
saccharides. A minor part of the work, mainly of an experimental character,
was concerned with the possibilities of full development of the dynamics of
multiplication under continuous flow conditions and utilized or followed the
influence Of a purposefully chosen limiting factor on the growth or some
other physiological function.
Hayduck (1923) patented a method of yeast cultivation with aeration and
at a slow rate of,flow of concentrated molasses in order to utilize also the pro-
duced alcohol. .Imray (1928, 1929) directed the rate of inflow and outflow
of wort : according t the sugar exhaustion. The regulation was performed
by means Of lowering the inflow orby returning the separated yeast. He carried
out-aet* feinientatiefis?both in the presence and in the absence of alcohol.
90
?
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Biihrig (1929) produced yeast by the continuous-addition-withdrawal process.
The diluted substrate flowed into the suspension of yeast cells and was with-
drawn at an approximately identical rate into a reserve vessel where the ripen-
ing was completed. Both vessels were aerated. Harrison (1930) used three
cylinders inserted into one another for the production of baker's yeast. The
medium was continuously added to the central cylinder from which it flowed
down through a number of overflow openings into the adjacent annulus with
a lower level and then identically into the second annulus with the smallest
volume of culture. An three stages were aerated thoroughly. Olsen (1930) used
a sezies of mutually connected fermentors in which he kept a constant rate
of flow. If the nutrients were not utilized in one fermentor he added another
one. Bilford, Scalf, Stark and Kolachov (1942) used a continuous fermentation
process for the production of ethylalcohol; this took place in one tormentor
stirred by CO, in the course of a 5-7 hour cycle. Seidel (1943) patented two
procedures for the production of yeasts and other microorganisms. In one case
he used three intensively aerated wide fermentation vats arranged in a cascade.
Yeasts separated from the last vat were returned to the first one. In the second
case he divided concentrically a wide cylindrical vessel into a number of
round compartments linked together alternately by overflows and openings
near the bottom. The medium added continuously or periodically into the
non-aerated centre, flowed through the openings at the bottom into the
neighbouring compartment whence it was carried by air to the overflow and
into the next nonaerated compartment from which the yeast suspension was
withdrawn for further treatment. Unger, Stark, Scall and Kolachov (1942)
described a continuous aerobic propagation process of alcohol yeasts which
were used only after separation as a starter in a proper alcoholic fermentation.
Ruf, Stark, Smith and Allen (1948) were able to ferment both in the laboratory
and in the pilot-plant acid-hydrolyzed grain mash in the course of a 12 hour
cycle.
At that time a number of papers appeared studying fodder yeast production
on molasses and on sulphite liquors, as shown by Lee (1949) in his general
report. Harris, Hannan, Marquardt and Bubl (1948) cultivated T. ?dais on
wood hydrolyzates using a battery of interconnected tormentors. Later Harris,
Saeman, Marquardt, Hannan and Rogers (1948) and Harris, Hannan and
Marquardt (1948) used a fermenter of the Waldhof type in which they studied
the problem of supplementation with inorganic salts when cultivating T.
on non-fermentable sugars remaining after the alcoholic fermentation of wood
hydrolyzates.
Stier, Scalf and Brockman (1950) .were able to obtain yeasts with constant
fermentation properties after having grown them continuously under anaerobic
conditions in their glass apparatus. They stirred the culture by means of
passing nitrogen through it. Adams and Hungate (1950) tried to calculate the
91
noriaccifiarl in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05 CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
characteristics for a yeast flow culture using the growth curve and the curve
of sugar decrease obtained in a batch culture. Maxon and Johnson (1953)
discussed the problems of oxygen transport and of material balance in a conti-
nuous fermentation process in one fermentor. A device measuring the pH of the
outflowing liquid controlled automatically the addition of ammonia to the
culture. Owen and Johnson (1955) designed a continuous shaker propagator
for yeasts and bacteria. Schulze (1956) studied the influence of various concen-
trations of ammonium dihydrogen phosphate on the growth rate of yeasts
cultivated continuously on sulhpite liquors from beechwood. Milek, Burger,
Hejmovi, It Mica, Fend and Beran (1957) used a multi-stage apparatus to study
hexose and pentose utilization by T. ntilis and lionilia murmanica growing
on non-diluted sulphite waste liquors. The organisms became adapted during
the flow culture to the toxic substances contained in the liquor but no adap-
tation could be observed to those sugars to which the organisms had not been
adapted previously. Borzani and Aquarone (1957) fermented continuously
blackstrap molasses to ethylalcohol on a semi-production scale and studied
the influence of the sugar concentration, feed rate, agitation speed and for-
mentor capacity. The application of penicillin to suppress contamination
had a beneficient influence on the course of fermentation.
In the field of industrial production of antibiotics it was Donowick (1952)
(Perlman et al., 1953) who tried to produce streptomycin in the continuous fer-
mentation way, and Kolachov- and Schneider (1952) who tried it with penicillin.
Liebmann and Becze (1950) patented a method for the production of anti-
biotics and especially of penicillin; they start to add substrate at the time of
maximum multiplication. The overflowing liquid was collected in another
vessel where the final phase of fermentation took place.
Nowrey and Finn (1955) (Boesch, Shull, 1956) studied the continuous fermenta-
tion production of acetone and butylalcohol by Cl.acetobutylicum keeping a very
low concentration of cells in the propagator. But a practically applicable con-
tinuous fermentation method for these products has not been discovered as
yet. Protiva and Dyr (1958) and Dyr and Protiva (1958) used a glass apparatus
with several flasks arranged in a cascade for the same purpose.
Ketchum and Redfield (1938) cultivated the marine diatom Nitzschia
closterium under chemostatic conditions. The rate of growth of this photo-
synthetic organism changed according to the intensity of light and to the
concentration of COI. Browning and Lockinger (1953) and Vivra (1958)
constructed a simple apparatus for cultivation of the infusorium Tetrahymena
geld under stabilized chemical and physical conditions; Vivra also studied
the morphology of Euglena gracilis (Klebs).
Most exPetiMental and theoreitcal papers dealing with stabilized conditions
under continuous flo.w oftthe medium have appeared in bacteriology, but only
. _ _
arsMall,nuthber-baie been treated from the point of view of practical appli-
92
in Dart - Raniti7ed Coov Approved for Release
?
cation. The greateSt attention has been given to genetic and growth problems
and only recently has the study of metabolite production and of various
enzymatic systems stepped into the foreground.
Fig. 4. Bactogen; Monod (1950).
Gerhardt (1946) designed a laboratory apparatus in which the addition of
fresh medium and the withdrawal of the product were accomplished continuously
and simultaneously. He used an aerated culture of Brucella suis at various
generation times without selecting the limiting growth factor.
Mack and his co-workers used a cultivating device consisting of several
aerated growth vessels arranged in a cascade connected by overflows. They
50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
93
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
studied the problem of growth cycles of bacteria, particularly of Azolobacter
(Wok 1952a, b; Macura, Kotkovi 1953), of the resistance (Wok, Vosykovii,
Wolf 1953) and the vegetative forms and spore formation of bacilli (geveik
1952; Wick, Chaloupka, Vosyjtovi, Vinter, Wolf, Hlavato, 1953). Wel/ (1955)
compiled his and his co-workers' results and laid the main stress on the import-
ance of studying metabolism and growth conditions from the point of view
of the optimal physiological state of the cultivated microorganism.
Monod (1950) studied the growth principles of bacteria and the role of the
concentration of important substrate on the rate of growth. By his mathe-
matical treatment of the principal steady-state factors a basis for the theory
of continuous flow cultures was laid. He called his apparatus "Le Bactogene"
(Fig. 4). The culture flask rotates along its longitudinal axis inclined by 3-40
.
In the rotating flask the liquid spreads on the flask walls and thus the culture
is stirred and aerated. The same principle was employed by Hofsten, Hofsten
and Fries (1953) for cultivation of the Ophiostonza multiannulatum mold; they
controlled the rats of growth by supplying new nutrients. Cohn ,and Torriani
(1953) studied the kinetics of the formation of a certain protein which is
related to beta-galactosidase while cultivating E. coli on maltose, lactose and
galactose in a synthetic medium. Perret (1953) added small doses of penicillin
to a culture of B. cereus growing in steady state at a constant density and mea-
sured the subsequent changes in the kinetics of penicillinase formation. He
used the disappearance of 32 to measure exactly the rate of dilution. Heden,
Holme and Malmgren (1955) modified somewhat Monod's bactogen and used
it to study the possible relationship between the growth rate and nucleic acid
content in E. coli. Rogers (1957) studied the suppression of hyaluronidase
formation in St. aurenus cultivated in the bastogen. Ferret (1957) described
an autoregulating apparatus for continuous cultivation applying the rotating
flask principle; its axis was inclined at an angle of 60? with the horizontal.
He described the function of an auxiliary device which kept the rate of flow
of medium and of air constant. In the apparatus described by Anderson (1953,
1956) a cylindrical vessel rotates along the vertical axis, the culture being
stirred by magnetically fixed wipers. He equipped his apparatus with a very
ingenious photoelectric attachment which automatically registers and regulates
the cultivation process.
At the same time as Monod, Novick and Szilard (1950a, b) described their
Chemostat (Fig. 5) and endeavoured to analyse mathematically the steady
state conditions of culture. They treated systematically the influence of the
concentration of various limiting factors, particularly of purines, on the rate
of formation of mutants of E. coli resistant to phage in relation to the absolute
time and presented a theoretical treatment of this process (Novick and Szilard
1950b, 1951, 1952, 1953, 1954). Using a chemostat Lee (1953) found that
, ?
the rite of formation of resistant mutants is independent of the rats of growth.
94
nerdassified in Part - Sanitized Copy Approved for Release
The limiting factors were tryptophane and theophylline, which, under certain
conditions, differed in their influence on the rate of formation of resistance
toward phages T5 and T6. Labrum (1953) (Novick, 1955) studied the rela-
tionship between the generation time and the time necessary for expression of
induced mutations. Various chemostats for use in continuous culture of bacte-
ria were described by Kubitschek (1954),*1-faan and Winkler (1955), and Rotman
(1955). The sonic lysis of cells of Azotobacter vinelandii and of E. coli grown
Fig. 5. Chemostat; Novick, Szilard (1950).
on synthetic media in the chemostat was studied by Rotman (1956), Formal,
Baron and Spilman (1956) studied the influence of continuous culture on
virulence and immunogenicity with mice of two strains of Salmonella typhosa.
Karush, Iacocca and Harris (1956) described an apparatus which they used
to keep a culture of haemolytic streptococcus of group A in a steady state
at different rates of growth. They studied the influence of the pH, glucose
and tryptophane as factors limiting growth. Holme (1957) carried out continuous
cultivation of E. coli in an Erlenmeyer flask placed in a rotation shaker. Ho
studied the relationship between the rates of synthesis of nitrogen-containing
components and of no-nitrogen compounds such as glycogen, using the source
of nitrogen as the limiting factor.
A technically acceptable and modernly equipped apparatus of larger type
for use in bacterial culture Nvas described by Elsworth and Meakin (1954) and
Herbert, Elsworth and Telling (1956). A very advanced laboratory fermentor
together with a reliably and constantly operating auxiliary equipment was
constructed by Elsworth, Meakin, Pirt and Capell (1956) (Fig. 6). Their appara-
-^
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
95
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
- ' -
96
gel
1
Fig. 7. Two litre scale continuous culture apparatus.
(Photo .1 Fiala )
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05 : CIA-RDP81-01043R002800080004-5
Fig. Sii. Twenty litre scale continuous culture apparatus. Preparation of inoculum.
(Photo Jug. J liospodka
Fig. Sb. Twenty litre scale continumIS culture apparatus Two stage baker's yeast
continuous culture with unequal volume of !kink!.
(Photo I sg J II ospodka
a
neclassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
tus was then equipped with an automatic pH registration and regulation device
designed by Callow and Pirt (1956) who also developed a method for chemical
sterilization of glass electrodes.
Fig. 7 shows a laboratory fermentation apparatus which can be used for
three simultaneous batch fermentations or for three simultaneous continuous
processes under indentical or different conditions. As fermentors are fixed on
a common carrier it is possible to connect them in series for some special
purposes (e. g. mixed fermentations). According to the rate of reaction taking
place in the individual vessels the rate of dilution can be adjusted by changing
the volume of culture. The feed rate of the substrate can be regulated by a ver-
tical movement of the outflow end of the thermostatic resistance capillary
tube with respect to the constant level in Mariotte's bottle. The amount of
liquid passing through can be measured either by a flowmeter, by the number
of drops, or by reading off the difference in the graduated storage bottle level.
Both the liquid and the air are thoroughly mixed by a centrifugal glass stirrer
(rti6ica, Gr?nwald 1954). Any registration or regulation device can be connected
with this apparatus:
Fig. 8 a and b show a glass flow apparatus with a working capacity of 20
litres which was used for continuous cultivation of S. cerevisiae on molasses.
At first the inoculum was prepared by a batch process in the first fermentor.
The continuous process took place in two other fermentors the second of which
had a smaller operating capacity so that the period of delay corresponded to
the time of complete utilization of the formed alcohol. The complex of problems
presented by this process as well as the qualitative properties of baker's yeast
prepared in this way are discussed by Beran in another report in this sym-
posium.
AUXILIARY DEVICES
The device by means of which the substrate solution is continuously added
to the culture flask must be simple, reliable and constant in operation as well
as easily manageable. It must be easy to sterilize and keep sterile while operat-
ing. It should be made from corrosion-proof material.
Most authors use the principle of Mariotte's bottle. The rate of feed is
regulated by a screw clamp, by an electromagnetic valve, or in most chemostatic
cultivations by a thermstatic resistance capillary tube. The limiting factor
is either the gas pressure above the level in the stock bottle or the difference
between the level and the outflow end of the capillary tube. Elsworth et al.
(1956) and Perret (1957) have recently described a very good apparatus of
this type. Lundsted, Ash and Koslin (1950), Michaeli (1951), Maude (1952) and
others used the principle of a rod or plunger being inserted into the tube in
7?Symposium
97
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
order to regulate the flow. Geankoplis and Hixson (1952) constructed a glass
needle valve used for adding small amounts of liquids under pressure. Jordan
and Jacobs (1944) added the fresh nutrient medium by means of an automatic
special pipette. Rosenberg (1956) added substrate to a microculture by drawing
it in by means of a filter paper strip. Stier et at. (1950). displaced the liquid
from the stock bottle by means of nitrogen. Softer? and Johnson (1954) used
the gas evolved by electrolysis of water, the amount of gas being regulated
by the intensity of the current.
Browning and Lockinger (1953) transferred small amounts of liquid into
the cultivation flask by means of a small rotating motor-driven glass dipper.
But its rotating parts could be kept sterile only with difficulty. The constant
speed cam principle was employed by Andreev (another report in this sym-
posium) and by Main, Cole, Bryant and Morris (1957); by raising and lowering
the dosing flask the cam regulates the rate of flow. Sims and Jordan (1942),
Savage and Florey (1950), Dale, Amsz, Ping Shu, Peppier and Rudert (1953),
Karush et al. (1956), Formal et al. (1956) used variously adjusted piston
pumps mostly adapted from syringes. As with most piston and membrane
pumps the weak point is represented by valves and by the interrupted flow
of the liquid a number of authors preferred pumps by means of which they
could add continuously solutions even of suspension character. The rotary
rubber tubing pumps proved to be the best ones (Weigl and Stallings 1950;
Apolcin 1953; Maxon and Johnson 1953; Hosier and Johnson- 1953; Malek
1955; and others); also the keyboard and hose pressure Sigmamotor pumps
(Holme, 1957) proved to be very reliable and exact.
The volume of passing liquid is usually measured in graduated vessels,
flowmeters (Herbert et al., 1956), or by automatic drop counters (Anderson,
1953, 1956). Catheron and Hainsworth (1956) presented a review of flow
measurement methods mentioning a diaphragm with an electric differential
transfer system to the electric registration device, a diaphragm with a pneu-
matic differential transfer system to the pneumatic registration device,
a magnetic flowmeter with electronic registration, a controlled regulation valve
with electronic registration etc. The rate of flow as measured by radioactive
indicators was studied, for instance, by James (1951), who designed an induc-
tion flowmeter, and Perret (1953). In his review of various dosing mechanisms,
particularly those for dosing of liquids, Henke (1955) considered also the prob-
lem of corrosion. Fuld and Dunn (1957) described an apparatus for continuous
refraction index measurement by means of which they control sugar concen-
tration during yeast propagation and simultaneously automatically control
the pH, temperature and the anti-foaming agent. Bartholomew and Kozlov
(1957) designed an apparatus for automatic control of the addition of the
anti-foaming agent and of the nutrient medium to the battery of fermentors.
One of the most important problems of continuous fermentations is presented
98
.4
by continuous preparation and sterilization of the substrate. A solution to this
problem was sought, for instance, by Unger et at. (1942), Gallagher et al. (1942),
Stark et al. (1943), Pfeifer and Vojnovich (1952), Whitmarsh (1954), Tomikk
(1956) as well as by others.
REFERENCES
ADAMS S. L., HUNGATE R. E., 1950: Continuous Fermentation Cycle Times. lnd. Eng.
Chem., 42: 1815.
ALTSHELLER W. B., Mow H. W., BROWN E. H. C., STARK W. H., SMITH L. A., 1947:
Design of Two-bushel per Day Continuous Alcohol Unit. Chem. Eng. Progress, 43:
467.
ANDERSON P. A., 1953: Automatic Recording of the Growth Rates of Continuously
Cultured Microorganisms. J. Gen. Physiol., 36: 733.
ANDERSON P. A., 1956: Continuous Recording of the Growth of Microorganism under
Turbidostatic and Chemostatic Control. Rew. Sci. Instr., 27: 48.
APOLCIN B. I., 1953: Novaja konstrukcija IIRSOS/1. V6Stnik Mainostroj., 33: 43.
ASHTON W. H., HOLGATE I., 1951: Fermentation Process. Brit. Patent. 660145.
AUDUS L. J., 1946: A New Soil Perfusion Apparatus. Nature, 158: 419.
BARNES F. W. jr., DEWEY M., 1947: Chemical Detoxification of Flexner Dysentery Antigen.
J. Bact., 53: 759.
BARTHOLOMEW W. H., KozLow D., 1957: Automatic Antifoam and Nutrient Feed
Controll for Bench Scale Fermentation. Ind. Eng. Chem., 49: 1221.
BECZE DE G., ROSENBLATT M., 1943: Continuous Fermentation. Am. Brewer, 76: 11-16,
30, 32, 34.
BEESH S. C., SKULL C. M., 1955: Fermentation. Incl. Eng. Chem., 47: 18.57.
BEESH S. C., Sum", G. M., 1956: Fermentation. Incl. Eng. Chem., 48: 1585.
BEESH S. C., SKULL G. M., 1957: Fermentation. Ind. Eng. Chem., 49: 1491.
Buxom) H. R., SCALY R. E., STARK W. H., KOLACHOV P. I., 1942: Alcoholic Fermentation
of Molasses. Rapid Continuous Fermentation Process. Incl. Eng. Chem., 34: 1406.
BOECKELER B. C., 1948: Fermenting Method. U. S. Patent 2, 440, 925,
BORZANT W., AQUARONE E.: 1957: Continuous Fermentation. Alcoholic Fermentation
of Blackstrap Molasses. Penicillin as Contamination Control Agent. J. Agr. Food
Chem., 5: 610.
BRINBERG S. L., 1953: Technologia procesov fermentacii. Antibiotiki. 36: 3.
Bnow-xiso I., LOCKINGER L. S., 1953: Apparatus for Stabilising the Chemical and Physical
Enviroment of Certain Free-living Cells. Texas. Repts. Biol. Med., 2: 200.
BRYSON V., 1953: Applications of the Turbidostat to Microbiological Problems. Atti VT.
congr. intern. microbiol. (Roma, Italia), I. 804.
BwrsoN V., SZYTIALSKI W., 1952: Microbial Selection. IL The Turbidostatic Selector ?
a Device for Automatic Isolation of Bacterial Variants. Science, 116: 45.
BOHRIO W. H. F., 1929: Process for the Manufacture of Yeast. U. S. Patent 1,730,876.
CALLOW D. S., PERT S. J., 1956: Automatic Control of pH Value in Cultures of Micro-
organisms. J. Gen. Microbiol., 14: 661,
CASTOR J. G. B., STIER T. J. B., 1947: Apparatus for Continuous Yeast Culture. Science,
106: 43.
CATHERON A. R., HAINSWORT B. D., 1956: Dynamics of Liquid Flow Control. Incl. Eng.
Chem., 48: 1042.
7.
99
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
- - -
CLEARY J. P., Brcartn P. J., CurroN C. E., 1935: Studies on Certain Factors Influencing
the Size of Bacterial Populations. j. Bact., 29: 205.
Currox C. E., 1943: Large-scale Production of Penicillin. Science, 98: 69.
Coax M., TORRIANI A. M., 1953: The Relationships in Biosynthesis of the ci-galactosidase
and Pi.-proteins in Escherichia coli. Biochim. et Biophys. Acta., 10: 280.
Cooic P. M., 1951: Chemical Engineering Problems in Large Scale Culture of Algae. Ind.
Eng. Chem., 43: 2385.
C'EKAN L. I., 1939: Priinenenie nepreryvnovo motocla pri kombinirovanom
Mikrobiologia, 8: 484.
DAL]: R. F., Amsz J., PING SHU, PEPPLER H. J., RUDERT F. I., 1953: A Continuous
Feeding Assembly for Laboratory Fermentations. Appl. Microbiol., 1: 68.
DANCKWERTS P. V., 1954: Continuous Flow of Materials through Processing Units. Ind.
Chem. Mfr., 30: 102.
DARANYI S. C., 1936: Manufacture of Yeast. U. S. Patent 2,035,048.
Dartrano.rom W. A., QUASTEL J. H., 1953: Absorption of Sugars from Isolated Surviving
Intestine. Arch. Biochim. Biophys., 43: 194.
DAVEY V. F., JOHNSON M. J., 1953: Penicillin Production in Corn-steep Media with
Continuous Carbohydrate 'Addition. Appl. Microbiol., 1: 208.
DIMOPOULLOS G. T., PRITHAM G. H., 1951: A Simplified, Continuous-flow Apparatus for
Use in Tissue Culture. J. Lab. Clin. Med., 37: 162.
DUBARCH P. J., 1956: A Simple Soil Perfusion Apparatus. Agron. Jour., 48: 138.
Ducirk J., Nair J., 1950: Culture des mieroorganismes on milieu continu. C. R. Acad.
Sc., 231: 83.
DYR J., PacrrIva J., 1958: Tvorba neutralnich rozpouat&lel u Clostridium acetobutylicum
pi kontinualni fermentaci. Cal. mikrobiol., (in press).
ELSWORTH R., MEAKIN L. R. P., 1954: Laboratory and Pilot Plant Equipment for the
Continuous Culture of Bacteria with Examples of its Use. Chem. Ind. July 24., 926.
ELSWORTH R., lkixamirr L. R. P., PIRT S. J., CAPELL G. H., 1956: A Two Litre Scale Con-
tinuous Culture Apparatus for Microorganisms. J. Appl. Bacteriol., 19: 264.
Fzurox L. D., DOUGHERTY K. M., 1924: Studies on Virulence. IT. The Increase in Viru-
lence in Vitro of a Strain of Pneum( -:occus. J. Exp. Med. 39: 137.
FINN R. K., WILsorr R. E., 1954: Population Dynamics of a Continuous Propagator
for Microorganisins. J. Agr. Food Chem., 2: 66.
FORMAL S. B., BARON L. S., SPILMAN W., 1956: The Virulence and Immunogenicity of
Salmonella typhosa Grown in Continuous Culture. J. Bact., 72: 168.
FOSTER J. W., MoDANizr, L. E., 1952: Fermentation Process. U. S. Patent 2,584,009.
Fox M. S., 1955: Mutation Rates of Bacteria in Steady-state Populations. J. Gen. Physiol.,
39: 267.
Fox M. S., SZILARD L., 1955: A Device for Growing Bacterial Populations under Steady-
state Conditions, J. Gen. Physiol., 39: 261.
FULD J. G., MINN C. G., 1957: Now Process Control Applications in Fermentation. Ind.
Eng. Chem., 49: 1215.
GADEN E., 1956: Chemical Technology of Fermentation. Chem. Eng., 63: 159.
GALLAGHER F. H., BILFORD H. R., STARK W. H., KOLACHOV P. J., 1942: Fast Con-
version of Distillery Mash for Use in a Continuous Process. Ind. Eng. Chem., 34: 1395.
GARIBALDI J. A., FEENEY R. E., 1949: Subtilin Production. Ind. Eng. Chem., 41: 432.
GEANKOPLIS C. J., Hixsox A. N., 1952: Control of Small Liquid Flows Using Glass
Valves. Ind. Eng. Chem., 44: 589.
GERHARDT P., 1946: Brucella Inds in Aerated Broth Culture: III. Continuous Culture
Studies. J. Bact., 52: 283.
2
GOLLE H. A., 1953: Theoretical Considerations of a Continuous Culture System. J. Agr.
Food. Chem., 1: 789.
Guaziost F., 1956: Metodo per in coltura continua dei batten i mediante un apparato
turbidostatico. Giorn. Microbiol., 1: 491.
GRAZIOSI F., PUNTONI S. V., 1957: Studio quantitativo (lelPadattamento di Mkrococcus
pyogenes o Proteus vulgaris alla novobiocina in coltura continua. R. C. Accad. Naz.
Lined, S. vm., 22: 369.
HAAN 1)5 P. C., WINKLER K. C., 1955: An Apparatus for Continuous Culture of Bacteria
at Constant, Generation Times. J. Microb. Serol., 21: 33.
HARMSEN G. W., KOLFF W. J., 1947: Cultivation of Microorganisms with the Aid of
Cellophane Membranes. Science, 105: 582.
HAmus E. E., HANNAN M. L., ManQuAitoT R. 11., 1948: Production of Food Yeast from
IVood Hydrolysates. Nutrient Requirements. Ind. Eng. Chem.. 40: 2068.
HARRIS E. E., HANNAN M. L., MARQUARDT R. R., Bum, J. L., 1948: Fermentation of
Wood Hydrolysates by Pomba Wills. Ind. Eng. Chem., 40: 1216.
HARRIS E. E., SAEMAN J. F., MARQUARDT R. R., HANNAN M. L., RocrEas S. C., 1948:
Fodder Yeast from Wood Hydrolyzates and Still Residues. Ind. Eng. Chem., 40: 1220.
HARRIS N. K., PowEt,t, E. 0., 1951: A Culture Chamber for the Microscopical Study of
Living Bacteria, with Some Observations on the Spore-bearing Aerobes. j. R. Mier.
Soc., 71: 407.
HARRISON A. P., 1930: Process for the Manufacture of Yeast. U. S. Patent 1,761,789.
HAYDUCK F., 1923: Low Alcohol Yeast Process. U. S. Patent 1,449,107, 1,448,108.
HEIAN C. G., HOLME T.. MALMOREN B., 1955a): An Improved Method for the Cultivation
of Microorganisms by the Continuous Technique. Acta Pathol. Microbiol. Scand.,
37: 42.
HEDEN C. G., HoumE T., MALMGREN B., 1955b): Application of the Continuous Cultiva-
tion Technique to Problems Connected with Growth Rate and Nucleic Acid Synthesis.
Acta Pathol. Microbiol. Scand., 37: 50.
HENKE R. IV., 1955: Chemical Feeding and Proportioning Equipment. Ind. Eng. Chem.,
47: 684.
HERBERT D., &swami'!" R., TELLING R. C., 1956: The Continuous Culture of Bacteria;
a Theoretical and Experimental Study. J. Gen. Microbiol., 14: 601.
HERMANN S., NEUSCIFUL P., 1953: Kontinuierliche Olukonsiiiire Darstellung mittels
Beet. gluconicum (Hermann). Ztbl. Bad,. Parasit. Infr. IT., 93: 25.
HOFSTEN VON B., HOFSTEN VON A., FRIES N., 1953: The Technique of Continuous Culture
as Applied to a Fungus Ophiostoma multiannulatum. Exp. Cell. Res., 5: 530.
HorztE T., 1957: Continuous Culture Studies on Glycogen Synthesis in Eschcrichia coli B.
Acta Chem. Scand., 11: 763.
Hosr,Eit P., JOHNSON M. J., 1953: Penicillin from Chemically Defined Media. hid. Eng.
Chem., 45: 871.
IMRAY A., 1928: Improved Process of Producing Yeast. Brit. Patent. 294134.
InitAY A., 1929: Process of Producing Yeast. Brit,. Patent 308324.
JACOB F., 1953: Influence du regime carbone sur le development des bacteriophages.
Ann. Inst. Pasteur, 84: 254.
JAMES W. G., 1951: An Induction Flowmeter Design Suitable for Radioactive Liquids.
Rev. Sci. Instr., 22: 989.
JEFFREYS G. A., 1948: Mold Enzymes of Microorganisms Produced by Continuous Tray
Method. Food Industries, 20: 688.
JORDAN R. C., JACOBS S. E., 1944: The Growth of Bacteria with a Constant, Food Supply.
I. Preliminary Observations on Bacterium coli. J. Bact., 48: 578.
101
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
KALJUiNYJ M. J., 1955: Nopreryvnoo sbrafivanie sulfitnych kcolokov droff.ami sorbiro-
vannymi voloknami celljulozy, Milcrobiologia, 24: 462.
KAausa F., Ltcocew V. F., HARRIS T. N., 1956: Growth of Group a Hemolytic Strepto-
coccus in the Steady-state. J. Bact., 72: 283.
HATSUME E., 1956: Symposium on Continuous Process in Fermentation Industry. J.
Fermentation Assoc., 14: 224.
KAUTSKY H., KAUTSKY H. jr., 1951: Apparatur zur Erziolung stationiiier Bedingungen
boi der Kultur von Microorganismen. Z. Naturforsch. 6b: 190.
Kr=Hum B. H., REDFIELD A. C., 1938: A Method for Maintaining a Continuous Supply
of Marino Diatoms by Culture. Biol. Bull., 75: 169.
Kor,mmov P., SCHNEIDER W. C., 1952: Continuous Process for Penicillin Production.
U. S. Patent 2,609,327.
Kamiss R. W., 1955a: Nutritional Requirements and Yields of Algae in Mass Culture.
Conf. Solar Energy: Tho Scientific Basis, Tucson, Arizona, October 31, November 1.
KRAUSS R. IV., 1955b: Nutrient Supply for Largo Scale Algal Cultures. Sci. Monthly
80: 21.
KRAUSS R. W., THOMAS W. H., 1954: The Growth and Inorganic Nutrition of Scenedeamus
obliquua in Mass Culture. Plant Physiol., 29: 205.
Kuarrsemax H. E., 1954: Modifications of the Chemostat. J. Bact., 67: 254.
LAacerKtbr I. S., 1939: Novyj metal batarejnovo brolenia v proirvodstve acotona-
-butanols-otanola. Mikrobiologia VIII, 486.
LEBEDEV S. V., 1936: Motody noproryvnovo spirtovovo broienija. Paopromizdat Moskva
150 pp.
Lax H. H., 1953: The Mutation of E. coli to Resistance to Bacteriophage T. Arch.
Biochem. Iliophys., 47: 438.
LEK S. B., 1949: Fermentation. Ind. Eng. Chem., 41: 1868.
Lsa S. B., 1950: Fermentation. Ind. Eng. Chem., 42: 1672.
Lax S. B., 1951: Fermentation. Ind. Eng. Chem., 43: 1948.
Lam H., 1949: The Soil Percolation Technique. Plant and Soil, 1: 221.
LEOPOLD J., Fatict, Z., 1955: P6stovimi kvasinek Torulopaia utilia v odpaclnich louzich
z citronovolto kvaenI.es. biol., 4: 212.
Law's R. IV., LUCAS E. H., 1945: Apparatus for Growing Microorganisms on a flowing
Medium. Science 101: 304.
L/EBMANN A. J., BECZE DE G. T., 1950: Continuous Method of Conducting Microbiological
Processes. U. S. Patent 2,524,200.
LETNDSTED L. G., Asa A. B., Kosurt N. L., 1950: Constant Rate Feed Device. Anal. Chem.,
22: 626.
MACLTRA J., KontovA M., 1953: Vy'voj azotobactera v proudicitn prostfedi. s. biol.,
2: 41.
MAnt R. K., COLE L. J., BRYANT L. M., MORRIS S. K., 1957: Constant Rate Flow Device
for Electrolyte Elements in Column Chromatography. Anal. Chem., 29: 1558.
MAixx. I., 1943: P6stovani mikrobit v proudicim prostfedi. as. lek. Us., 82: 576.
MALEK I., 1952: Kultivace azotobactera v proudicim prostiedi. Cs. biol., 1: 91.
MALEK I., 1952: Ktiltivace bakterii ve vicestuptiovem proudicim prostfedi. Os. biol., 1: 18.
MALEK I., 1955: 0 nutoioni a piistovani mikroorganismet, zvlit5t6 bakterii. Nakl. (SAV,
Praha, 170 pp.
Mlaxic I., BURGER M., lizmovi L., nidica J., FENCL Z., BERAN K., 1957: Asimilace
cukrit a lcyselin pi kontinutilnim zdroidovaini nefediinich sulfitovYch Vluhti. Cs.
mikr., 2: 203.
102
e
MALEK L, CHALOUPKA J., VOSYKOVA L., VINTKR V., ?1701.r A., ILLAvATA D., 1953: Spo-
rulace bacilS.Cs. biol., 2: 323.
MALEK I., VosvicovA L., WOLF A., 1953: Resistenco bakterii p6stovanftli v proudichn
prosttedi. C's. biol., 2: 68.
MAuDE A. H., 1952: Constant Liquid Feed Device, Ind. Eng. Chem., 44: 585.
Maxos W. D., 1954: Continuous Fermentation. A Discussion of its Principles and Applica-
tions. App!. Mier., 3: 110.
MAxos W. D., JOHNSON M. J., 1953: Aeration Studies on Propagation of Baker's Yeast.
Ind. Eng. Chem., 45: 2554.
MEYER E. A., 1929: Manufacture of Yeast. U. S. Patent 1,724,952.
MicuAELt 3. 1951: A Constant Liquid Feed Apparatus. Chem. Ind. Lond., 8: 123.
Moson J., 1950: La technique do culture continue. Theorie et applications. Ann.
Inst. Pasteur, 79: 390.
MOOR W. A., 1945: The Continuous Cultivation of Microorganisms. Science, 102; 594.
MorEa H. V., 1929: A Continuous Method of Culturing Bacteria for Chemical Study. J.
Bact., 18: 59.
MrElts j., CLARK L. B., 1944: Culture Conditions and the Development of the Photo-
synthetic Mechanism. II. An Apparatus for the Continuous Culture of Chloral".
J. Gen. Physiol., 28: 103.
l`ouTuitor H. J., 1654: Apparatus for Maintaining Bacterial Cultures in the Steady
State. j. Gen. Physiol., 38: 105.
NORTHROP J. H., 1957a: Adaptation of Bacillua megatlicrium to Terramycin (Oxytetra-
cycline). J. Gen. Physiol., 40: 547.
NORTHROP J. H., 1957b: The Proportion of Terramycin-Resistant Mutants in 13. mega-
therium Cultures. J. Gen. Physiol., 41: 131.
NORTHROP J. H., ASHE L. H., MottcAN R. R., 1919: Tho Fermentation Process for the
Production of Aceton and Ethyl-Alcohol. Incl. Eng. Chem., Il: 723.
NovicK A., 1955: Growth of Bacteria. Ann. Rev. Microbiol., 9: 97.
NOVICK A., SZILARD L., 1950: Description of the Chemostat. Science, 112: 715.
NOVICK A., SZILARD L., 1950: Experiments with the Chemostat on Spontaneous Mutations
of Bacteria. Proc. Natl. Acad. Sci., 36: 708.
NovicK A., SZILARD L., 1951: Genetic Mechanisms in Bacteria and Bacterial \grilses.
Experiments on Spontaneous and Chemically Induced Mutations of Bacteria Growing
in the Chemostat. Cold Spring Harbor Symposia Quant. Biol., 16: 337.
NOVICK A., SZILARD L., 1952: Anti-mutagens. Nature 570: 926.
NOVICK A. SZILARD L., 1953: Studies of Mutation Rates in Bacteria. (Atti VI. congr.
intern. microlnol. (Roma, Italia), 1, 658.
NOVICK A., SziLkun L., 1954: Dynamics of Growth Processes. Princeton Univ. Press,
Princeton, N. J. 304 pp.
OLSEN A. J. C., 1930: Perfectionnements relatifs aux procedes do preparation do lovu-
re. Francais Pat. 676,835.
OLsos B. H., JOHNSON J. J., 1948: The Production. of 2,3-Butylene Glycol by Arcobacter
aerogenes 199. J. Bact., 55: 209.
OWEN S. P., JOHNSON M. J., 1955: Continuous Shake-Flask Propagator for Yeast and
Bacteria. J. Agr. Food Chem., 3: 606.
0 WEN W. L., 1948: Continuous Fermentation. Sugar 43: 36.
PASYNSKIJ A. G., 1957: Teorija otkrytych sistem i cc zna6enie dlja biochimii. Usp. Sovr.
Biol., 43: 263.
PASYNSKIJ A. G., NEJMARK A. M., 1952: St6k1jannyj laboratornyj fermenter. Atikrobiolo-
gia 21: 594.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
103
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
PiRLMAN D., BROWN W. E., Lza S. B., 1952: Fermentation. Ind. Eng. Chem., 44: 1996.
PERLMAN D., KROLL C. L., 1954: Fermentation. Ind. Eng. Chem., 46: 1809.
PzaLmAND.,TEmpicL A. E., jr., BROWN W. E., 1953:Fermentation. Ind. Eng. Chem., 45:1944.
PERRET C. J., 1953: Penicillinase Adaptation in Multiplying Cultures of Bacillus cercus
maintained at Constant Population Density. Atti VI. congr. intern. microbiol., (Roma,
Italia), I, 153.
FERRET C. J., 1957: An Apparatus for the Continuous Culture of Bacteria at Constant
Population Density. J. Gen. Nicrobiol., 16: 250.
Prxinca V. E., VOJNOVICH C., 1952: Continuous Sterilization of Media in Biochemical
Processes. Ind. Eng. Chem., 44: 1940.
PRILLIPs J. N. jr., MYERS J., 1954: Measurement of Algae Growth under Controlled Steady
State Conditions. Plant Physiol., 29: 148.
PiaT S. J., 1957: The Oxygen Requirement of Growing Cultures of an Aerobacter Species
Determined by Means of the Continuous dulture Technique. J. Gen. Mierobiol., 16: 59.
PM E., 195e Vfroba poluirskoho drof.dia polokontinuitnjm a kontinuitnjm vedenim
kvasenia. Kvaa. prism., 2: 2.
PORGE8 N., CLARK T. F., GASTROCK E. A., 1940: Gluconic Acid Production. Repeated.
Use Submerged Aspergillus niger for Semicontinuous Production. Ind. Eng. Chem.,
32: 107.
POWELL E. 0., 1955: Some Features of the Generation Times of Individual Bacteria.
Biometrika, 42: 16.
PowaLL E. 0., 1956: Growth Rate and Generation Time of Bacteria, with Special Refer-
ence to Continuous Culture. J. Gen. Microbiol., 15: 492.
PitonvA J., DYR J., 1958: Dlouhodobe vegetativni ptevaid6ni a pcxstovani Clostridium
aceiobutylicum v prtitokove kultufe. sl. milcrobiol., (in press).
Rocacits H. J., 1957: The Preferential Suppression of Hyaluronidase Formation in Cultures
of Staphylococcus Aureus. J. Gen. Microbiol., 16: 22.
Rocizas L. A., WEirriza E. 0., 1930: The Growth of Bacteria in a Continuous Flow
of Broth. J. Bact., 20: 127.
ROSENBERG M., 1956: The Dynamics of the Breaking-Down of Lysogenic Cells Irradiated
by Ultra-Violet Light. Folia biol., Praha, 2: 206.
RoTNAN B.?1955: A Simplified Device for Continuous Growth of Microorganisms. J.
Bact., 70: 485.
ROTMAN B., 1956: On the Mechanism of Sonic Lysis of Bacteria. J. Bact., 72; 6.
RUDAKOV K. I., 1936: Biologi6eskijo i fiziko-chimi6eskijo faktory aktivnosti mikroorga-
nizmov. Tr. vsos. inst. s.-ch. mikr., 8: 123.
Bur E. W., STARK W. H., SMITH L. A., ALLEN E. E., 1948: Alcoholic Fermentation of
Acid-Hydrolyzed Grain Mashes. Continuous Process. Ind. Eng. Chem., 40: 1154.
ittinicA GRUNWALD H., 1954: Odstfedivo laboratorni inichadlo. Chem. listy, 48: 1250.
SADOPF H. L., HALVORSON H. 0., FINN R. K., 1956: Electrolysis as a Means of Aerating
Submerged Cultures of Microorganisms. Appl. Microbiol., 4: 164.
SAEMAN J. F., LOCKE E. G., DICKERMAN G. K., 1946: The Production of Wood Sugar in
Germany and its Conversion to Yeast and Alcohol. Paper Trade J., 123: 38.
SAENKO N. F., 1950: Povygenie spirtoustogivosti chereznych droliej putem ich napra-
vlennovo vospitanija. Vincxlelie i vinogradstvo SSSR, 2: 22.
SAX S., 1932: Method on Controlling Yeast Propagation. U. S. Patent 1,884,272.
SAVAGE M. C., FLOREY H. W., 1950: Induced Bacterial Antagonism. Brit. j. Exp. Path.,
31: 17.
SAvozNxo N. J., 1957: Ob opyte raboty zavoda po adnoproto6noj scheme. Spirtovaja
PromyAlennosf, 23: 41.
104
SCHEN'LEY IND. Co.; 1953: Perfectionnements apportes aux proc6d6 do pric.paration de
produits microbiologiques. Francais Patent 1,018,459.
SCHULZE K. L., 1956: The Effect of Phosphate Supply on the Rate of Growth and Fat
Formation in Yeast. Appl. Microbiol., 4: 207.
SEIDEL M., 1943: Proced6 do preparation do levures et autres microorganisms. Francais
Patent 878,985.
SILCOX H. E., LEE S. B., 1948: Fermentation, Ind. Eng. Chem., 40: 1602.
Sims A. L., JORDAN R. C., 1942: An Accurate Automatic Syringe Mechanism. J. Sci.
Instruments, 19: 58.
SOLTERO F. V., JonNsox M. J., 1954: Continuous Addition of Glucose for Evaluation of
Penicillin Producing Cultures. Appl. Microbiol.. 2: 41.
SPICER C. C., 1955: The Theory of Bacterial Constant Growth Apparatus. Biometrics
11: 225.
STARK W. H., KoLAcnov P., WILLI:1E H. F., 1943: Wheat as a Raw Material for Alcohol
Production. Ind. Eng. Chem., 35: 133.
STICE E., Pavrr R., 1946: Production of Penicillin X in "Submerged" Surface Cultures.
Science, 103: 2678.
STIER T. J. B., SCALP R. E., BROCKMAN M. C., 1950: An All Class Apparatus for the
Continuous Cultivation of Yeast under Anaerobic Conditions. J. Bact., 59: 45.
,gARKOV V. I., 1950: Gidroliznoe proizvodstvo. III. Goslesbumizdat.
Szveix V., 1952: Tvorba velkjrch foretn u Bacillus subtilis v proudicim prostfedi. Cs.
biol., 1: 93.
8VACHULOVA J., KLIAKA J., 1956: JednoduchS' pfistroj na kultivaci povrehovil rostoucich
kultUr v proudicim prostfedi. Rozhl. Tb., 16: 488.
TAMPA H. 1957: Mass Culture of Algae. Ann. Rev. Plant. Phys., 8: 309.
TEMPLE K. L., 1951: A Modified Design of the Lees Soil Percolation Apparatus. Soil Sci.
71: 209.
TOMgEK J., 1956: Kontinualni ptiprava melasovet ziripary v kvattn6m prtimyslu. KVaS.
prism., 2: 33.
UEDA K., 1956: Theoretical Consideration and Experiments on Continuous Alcoholic
Fermentations. J. Perm. Assoc., 14: 268.
UNGER E. D., STARK W. H., SCALP R. E., KOLACHOV P. J., 1942: Continuous Aerobic
Process for Distiller's Yeast. Engineering and Design Factors. Ind. Eng. Chem., 34:
1402,
UTENKOV M. D., 1941: Thkrogenerirovanie. Gos. Izd. Sov6tskaja nauka, Moskva, 152 pp.
UTENKOV M. D., 1944: Dlitelnoe mikrogenerirovanie. Mikrobiologia 13: 310.
VIVRA j., 1958: Pouittiprittokovtho zarizeni ke studiu morfologie Evulcna gracilis (Klebs).
Cs. biol., On press).
VERDINA N. M , 1955: Priaenie droliej k antiseptikam razlii;nymi metodami. Trudy
Inst. Mikrobiol., 4: 54.
VICTORERO A. F., 1942: New Process of Continuous Fermentation. intern. Sugar 274.
VicronERo A. F., 1945: System of Continuous Fermentation. U. S. Patent 2,871,208.
VieTonEno A. F., 1948: Apparatus for Continuous Fermentation. U. S. Patent 2,450,218.
WARAKI S., ISHIDA T., YAMAGUSHI Y., MuzuttARA K., MAstmA K., 1952: Continuous
Fermentation of Penicillin. J. Antib. (Japan), 5: 622.
WEIGL J. W., STALLINGS D. W., 1950: Rubber Tubing Pump. Rev. Sci. Instr., 21: 395.
WitrotAnsit J. M., 1954. Continuous Sterilization of Fermentation Media. J. Appl.
Bact., 17: 27.
WitirriEli E. G., ROGERS L. A., 1931: Continuous Fermentation in the Production of
Lactic Acid. Ind. Eng. Chem., 23. 532.
105
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
AN IMPROVED CELL FOR MAINTAINING BACTERIAL
CULTURES IN THE STEADY STATE
JOHN H. NORTHROP
If bacteria are grown in a constant volume of culture medium, the compo-
nents of the medium, the growth rate, and the physiological state of the cells
are constantly changing.
' If the concentration of bacteria is kept constant, on the other hand, the
composition of the culture medium and the growth rate and physiological
state of the bacteria are also constant.
The bacteria may be kept at constant concentration by diluting the culture
at a constant rate which is less than the maximum growth rate of the bacteria
in that culture medium. Under these conditions, the concentration of bacteria
will increase, and the growth rate decrease, until the latter exactly equals the
dilution rate. The system is now in stable equilibrium. This principle has been
used by Monod (1950) and Novick and Szilard (1950), Novick (1955), and
Perret (1957) to maintain the culture in continuous growth. It has the advan-
tage of experimental simplicity.
It has the disadvantage that it is not possible to maintain a culture at its
maximum growth rate, since this would be an unstable equilibrium. In order
to maintain the culture at its maximum growth rate, it is necessary to control
the rate of flow of the culture medium in such a way that the cell concentration
remains constant, no matter what the growth rate. Several types of apparatus
for this purpose have been devised (Felton and Dougherty, 1924; Myers and
Clark, 1944; Bryson, 1952; Anderson, 1953; Northrop, 1954). This method is
more complicated experimentally, and is restricted to a range of cell concentra-
tion which may be determined by optical methods. It has an advantage in that
it is possible to make automatic records of growth rate with considerable
,
accuracy for long periods of time.
Under, these conditions the growth of the organism in the cell is exactly
compensated for by the addition of more culture medium so that
106
--
Declassified in Part - Sanitized Copy Approved for Release
B d
K=
= Jr
dt - (It V
growth rate dilution rate
STAT
dB is the number of organisms washed out of the cell in unit time, Be is the
number of organisms in the cell (a constant), V, is the volume of the cell and di'
is the overflow from the cell in unit time.
On integration
V
lid = V ct
since when t = 0, B = 0 and V = 0.
The value of V may be determined by measuring the volume of culture
medium which flows through the cell or calculated from the kymograph
record which shows the fraction of the time during which the culture medium
flowed into the cell. In this case
Kml . hr.-4 time of flow
=
I',1 elapsed time
where ml. hr.--' is the ml. of culture medium which will flow per hour without
interruption. If the drops per ml. of the culture medium from the capillary
tip is known,
K=
K drops culture medium min.-4 60 time of flow
drops culture medium elapsed time
It may be noted that under these conditions the increase in bacteria (assum-
ing that growth occurs only in the cell) is arithmetic, instead of logarithmic,
as under usual conditions, since the increase in bacteria is removed as fast
as it is formed. If this were not the case, the growth rate could not be balanced
by an arithmetic dilution rate.
The time for the number of bacteria to double, therefore, is equal to the
In 2
generation time 1/K, instead of as in growth at constant volume.
A
In case the culture contains 2 organisms growing at different rates, the ratio
of the 2 is
M M0e
IV ?
where 4-9- is the ratio of the 2 cells at t = 0 :and 4.
is the ratio at time e.
V 0
K. and K ,? are the respective growth rate constants. The time required to
reach various values of yr is (Northrop, 1954)
2.3 (log - log ?
M0\
0
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
107
Declassified in Part - Sanitized Cop
EXPERIMENTAL
Approved for Release
The principal experimental difficulties are foaming and sticking of the
organisms to the walls of the cell. For practical purposes this may be an advant-
age, as in the quick vinegar -process, or the continuous acetone-ethyl alcohol
fermentation (Northrop, Ashe, and Morgan, 1919).
For theoretical purposes, however, it is necessary
that the concentration of organisms in the overflow
is the same as that in the cell, since this is assumed
in deriving the equations. If the cell contents foam,
the foam contains fewer bacteria per ml. than the
bulk of the suspension (Northrop and Murphy, 1956).
If the organisms stick to the walls, they interfere
with any optical system and also cause an error in
the growth rate determination. If these organisms
differe qualitatively from those in suspension, ana-
lysis of the overflow may lead to entirely erroneous
conclusions.
11 The cell shown in Fig. 1 has been found to overcome
10
Fig. 1. Cell for maintaining bacterial cultures in the steady state.
1 ? Culture medium inlet, 3 mm I. D. with fine tip; 2 ? Tube,
10 mm I. D.; 3 ? About 2 rnm I. D. 4?. Lactic acid inlet, 3 num
I. D.; 5 ? Tube, 3 mm I. D.; if ? Water outlet, 5 mm I. D.; 7 ?
Water inlet, 5 mm I. D.; 8 ? Vessel, 22 mm 0. D., 85 mm high
(up to the water outlet); 9 ? Black tape, 15 mm; 10 ? Overflow,
5 ;rim I. I).; 11 ? Air, 3 min I. D.; 12 ? Vessel, 35mm 0. D.
both of these difficulties. The air does not bubble through the suspension so
that there is little foam. The solution is kept saturated with air and the cell
walls are kept clean by the rapid motion of the wipers.
The wipers are made from Lusteroid test tubes and are transparent so that
they do not interfere seriously with the optical system. A 10 per cent solution
of lactic acid, containing formaldehyde or any suitable disinfectant, is dropped
continuously (except when a sample is taken) into the collar of the cell where
it mixes with the overflow from the cell. This keeps the overflow clean and
sterile, and prevent? contaminating organisms from growing back up the
overflow tube.
The rate of flow of the culture medium is controlled by a galvanometer
connected to the photo-electric colorimeter in which the cell is placed. The
galvanometer operates a photoelectric realy which in turn activates a solenoid
controlling the flow of culture medium.
108
?
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
In case the growth in a large vessel is to be controlled, the overflow from the
tank is allowed to flow through the cell. The flow of culture medium into the
tank is then controlled in the same way as that of the cell.
In case pathogenic organisms are grown, the cell may be put under slight
negative pressure, instead of positive. The rubber stopper and stirrer guard
tube may be enclosed in a cylinder filled with strong disinfectant.
CONSTRUCTION OF THE APPARATUS
APPARATUS REQUIRE!)
Cell and wipers as in Fig. 1.
Klett-Summerson Industrial Colorimeter, Model 900.3.
Worner Photo-electric Relay, Model 5-100R.
Galvanometer, Leeds and Northrup, No. 2420, Coil P. I. 102 B. S.
Potter and Brumfield Relay, 6V-50-60 C (For kymograph pen).
General Electric Solenoid. CR9503-209C modified as in Fig. 2.
Westinghouse double contact projection lamp, 100W-P11/100T8/108.
Bodine speed reducer motor, NSE-12RH, 115 AC-DC, 1.1 amps., int. duty,
5000 RPM, 1/18 HP, with 10: 1 reducing
gear, with eccentric crank having about
1 cm stroke.
100 Cl resistance.
100 Cl var. resistance.
40 Cl var. resistance.
200 Cl var. resistance.
Bird kymograph (No. 70-060) with pen
attached so as to drop 5 cm every re-
volution of the drum.
6 cm UF Pyrex filter.
Fig. 2. Control of flow of culture medium
by General Electric solenoid.
1 ? Culture medium; 2 ? Pins.
The Cell Wipers
A 15 x 110 mm lusteroid test tube is cut open lengthwise and the closed
end cut off. It is cemented to a No. 16 stainless steel wire by means of a solution
of lusteroid in acetone. The tube is cut about every 1.5 cm as shown in Fig. 3.
Electrical Connections
The wiring diagrams are shown in Fig. 4. It is advisable to have a constant
voltage transformer in the lamp circuit, since the sensitivity varies somewhat
with the light intensity.
npriaccifipri in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
109
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
-
The wires from the photo cells in the Klett-Summerson colorimeter are
disconnected from the colorimeter galvanometer and connected to the Leeds
and Northrup galvanometer. A 40 ohm variable resistance is put in parallel
with the galvanometer.
-4-
Fig. 3. Wiper for culture
cell.
1 ? Rubber tubing, 5 mm
I. D.; 20 cm long; 2 ? Glass
tube; 3 ? Lusteroid with
cuts; 4 ? 16 Stainless steel
wire.
Fig. 4. Wiring diagram.
1 ? 110 V constant voltage trans-
former; 2 ? Colorimeter lamp and
100 CI variable resistance; 3 ?
Galvanometer lamp and 100 Cl re-
sistance; 4 ? Photo-electric relay
? operating circuit; 5 ? Photo-
electric relay circuit, solenoid, ky-
mograph; per magnet; 6 ? Wiper
motor and 200 Cl variable resistance.
it
1
2
I
3
-1-? 4
I
I
I 1
I I
JL 5
Fig. 5. Galvanometer light
and photo-relay.
1 ? Photo-electric relay;
2 ? Screen; 3? Lamp; 4 ?
Lens; 5 ? Galvanometer
Mirror.
The 100 W projection lamp is put in the galvanometer case in place of the
usual lamp. 100 ohms resistance is put in series with the lamp. A screen is put
around the lamp so that light from the lamp cannot strike the photo-cell of the
relay (cf. Fig. 5).
The galvanometer, photo-relay, and light are adjusted so that, when the
galvanometer is it rest, the image of the galvanometer lamp filament is pro-
110
jected through the opening in the screen and on the photo-relay tube so as
to operate the relay.
Rubber to Glass Connections
Coat glass with rubber cement, wet rubber tubing with acetone and slip
over glass.
Connection between Culture Medium Tube and Intake Tube
of Cell
Cf. Fig. 6.
2
5 4
3
Fig. 6. Method of connecting tube from culture medium stock bottle to intake tube of coll. The
connection is immersed in 50 per cont alcohol.
1 ? Clamp; 2 ? Wire binding; 3 ? Culture medium; 4 ? 21 Hypodermic needle; 5 ? Rubber
tubing, 3 atm I. D.
Operating Directions
Connect glass air filter and culture medium inlet tube to cell. A fine glass
capillary tube to regulate the flow of culture medium is put in the culture
medium intake line. The air intake tube is closed with a spring clip to prevent
the filter from becoming wet when the cell is autoclaved. The lip of the cell,
in which the rubber stopper will be inserted, is coated with rubber cement,
and the cell and connections autoclaved for 1/2 hour. The cell stopper and wipers
are inverted and immersed in a cylinder containing 5 per cent formaldehyde
for 24 hours. The stopper and wipers are then inserted in the cell, and fastened
in place with strong rubber bands. The apparatus is then assembled as in
Fig. 7.
The culture medium is allowed to flow, and the motor regulated so as to
operate the wipers as rapidly as possible, without foam. This is usually 200
to 500 strokes per minute. If this is not sufficient to prevent organisms sticking
to the glass, a commutator may be put in the kymograph axle which will
speed the motor up to 700 to 800 strokes per minute for 5 to 10 seconds every
half hour.
The colorimeter is now adjusted so that the relay operates when the colori-
meter dial reads in the range of 0 to .400, and the sensitivity of the system
regulated by means of the intensity of the colorimeter lamp and the galvano-
meter damping resistance so that the relay operates in a range of about ?5
divisions on the colorimeter scale. If the system is too sensitive, the time during
111
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
which the culture medium flows will be too short to read from the kymograph
record. If it is not sensitive enough, the bacteria concentration will vary too
much during the cycle.
The culture medium is allowed to flow for several hours, until all the formal-
dehyde has been washed out of the
cell. The cell is then inoculated by
injecting a few milliliters of culture
through the rubber guard tube of
the wiper, by means of a hypodermic
syringe. The flow of culture medium
is now stopped and the organisms
allowed to grow up in the cell until
the concentration is reached at which
the culture is to be maintained. The
colorimeter dial is adjusted so that
the relay operates at this point. The
system should now operate automa-
tically so as to maintain the bacterial
concentration at this level.
In case a kymograph record is
made*), it is necessary, to adjust the
sensitivity of the galvanometer, the
flow rate of the culture medium, and
Fig. 7. Diagrammatic flow sheet. the speed of the kymograph so that
the time interval during which the
culture medium flows into the cell
may be accurately read from the re-
cord. The best condition is when the culture medium flows into the cell about
half the time. This arrangement allows for quite wise changes in the growth
rate. The -faster the growth rate, the better the automatic control, since the
optical density changes rapidly. Very slow growth rates removal_ of. the cells with the substrate as well as the stability
_ ? _
of their, composition, imit the rate of fermentation. In order to accelerate
the process, it!..was of importance to use repeatedly the yeast separated from
the _spent' .mash, this. yeast representing a great store of enzymes. For this
reason it waanecessary to evaluate the increase of cells during the ferment-
ationproeesa_under conditions, where the amount of yeast starter was 3-6
ticaea:Oeater_than,the number of the natural population.
- The.. yeast culture A3, a rapidly proliferating yeast, and strain R XII,
all cultivated on beer,..wort, were, tested. In order to evaulate their viability
'ini.iiOOd,,hydrelyaite,Wort an inoculum 4.10 g of yeast per litre was used, the
aulistrate:rbeing- renewed -once a ..day. The concentration of reducing sugar
in the wort was 2.85%;and that of fermentable sugar 2.11?A; thusi g of yeast
fermented 2 11 g of--tiugai per da3. It follows from analyses (Table 1) that
-with each culture dead Cells appeared on the second day. On the fourth day
tha.nuMber`,Of living cellsdecreased-,On in average 0110, X 104/nil. In ih-e-fol-
-;:k4iiiiisiiii4;aril'aierage-':of 100 x 10. cells per ml survived and this level
as maintained The 'Culture. A3, Which was isolited-fronithe yeast it a'hydro-
fiat-6: plan Of --higher -activity: This experiment showed- that
? - , ,
_
Declassified in Part - Sanitized Copy Approved for Release
under stationary conditions with reintroduction of yeast and renewal of sub-
strate the number of the culture population considerably increases,
Table 1
Orig.
nuin-
be
Time of sampling (lir)
M ?Hal
Of
ex-
11:=
of
cells
24
48
96
120
144
168
102
216
104/1111
Culture A,
Living cells 104/1111
120.0
124
111.3
110.1
107.6
110.8
121.5
106.5
129.6
Dead cells 104/m1
0
o
8.7
9.9
21.4
18.7
25.5
53.2
71.6
Sugar fermented, %
2.07
-
2.11
2.10
2.11
2.07
2.09
2.08
Yield of alcohol
per 100 kg RS/litres
40.0
-
38.6
38.6
41.0
38.6
37.0
37.7
38.8
Rapidly proliferating yeast culture
Living cells 104/m1
87.0
87.0
91.1
78.0
97.8
87.0
102.5
96.4
95.3
Dead cells, 106/m1
0
0
6.4
14.0
22.2
16.0
26.5
43.1
? 51-9
Sugar fermented, %
2.02
2.07
2-11
1.93
2.02
2.03
2.08
2.10
Yield of alcohol
per 100 kg RS/litres
38.1
37.7
38.6
37.2
37.2
37.7
37.7
37.2
37-7
Strain XII
Living cells, 104/m1
96.0
96.5
85.8
87.0
108.3
120.0
100.2
102-3
105.7
Dead cells, 104/m1
0
0
4.9
6.5
15.4
29.0
48.8
52.4
86.8
Sugar fermented, %
2.02
2.11
2.11
2.07
2.05
2.09
2.10
2.04
Yield of alcohol
per 100 kg RS/litres
40.0
-
38.6
37.2
37.7
37.7
38.1
36.0
37.9
Average number of
living cells from
three cultures ....
101.0
102.4
96.0
91.7
104.5
105.9
108.0
101.7
110.2
In another series
culture L33, which
hydrolysate wort. 20
to 2.44 g of sugar/day
of culture A, and 150
viability and increase
presence of nutrients.
of experiments the culture A, was compared with
was sub-cultivated for many generations in wood
g of yeast were used as inoculum, which corresponded
X g. It was calculated that 1 ml contained 132 x 10' cells
X 104 cells of L?. From these figures it follows that the
of cells depend on the type of culture and mainly on the
VIABILITY AND PRODUCTIVITY OF YEAST UNDER VARYING
CONDITIONS OF INOCULATION
In a second experimental series the effect of the renewal number of sub-
strate on the viability of the culture A3 accumulating in wood hydrolysate
wort was investigated.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
201
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
The fermentation was carried out with varying inocula; the substrate was
renewed 1-5 times a day; thus the daily quantity of sugar per g of yeast
varied. Furthermore, in an .additional flask, inoculated with 50 g yeast per
litre, the wort content of phosphorus and nitrogen was increased by the
addition of 2 g (N114)1111104 per litre. For the first three concentrations the
duration of the experiinent was 10 days and 7 days for the concentration of
50 g Per litre. The initial count of cells, their increase, dying off and alcohol
productivity are shown in Table 2 as average values for the duration of the
experiments.
Table 2
Ferment-
able
sugar in
mash
%
Number
of
renewals
of mediaf
per day
yeast inoculum
Daily
level
of sugar
gig of
yeast
Increase
(-I-) or (10.
crease (-)
of living
ells 105/m1
Yield of alcohol
g/1
10i/ m1
1 per
100 kg
sugar
%
of the theo-
roticalyielcl
2.8
1
10
123
2.8
+ 8.3
53.8
83.8
2.8
1
20
227
1.4
+ 2.2
52.8
82.0
2.8
3
20
240
4.2
+ 6.9
54.0
84.0
2.8
1.
40
460
0.7
-10
51.8
80.5
2.8-
3
40
466
2.1
- 3.8
52.1
81.0
2.4
1
50
750
0.48
-35.0
45.7
71.0
2.4
5
50
754
2.4
0.00
50.0
77.8
24+(NH)2131'04
5
50
754
2.4
+ 2.2
49.3
76.4
_
It follows from these data that in the case of an inoculum of 10 and 20 g
per litre the yeast not only survives, but that proliferation of the cells takes
place corresponding to the quantity of sugar available. Thus at a level of
4.2 g of sugar and a threefold renewal of the wort the yeast weight increases
from the original 20 g to 27 g per litre. The alcohol yield produced by the yeast
is closely related to the proliferation of the cells. Lower values of the alcohol
yield and the state of the yeast were obtained with inocula of 40 an& 50 g
per litre. In ,these variant's a threefold renewal of substrate per day did not
lead to proliferation of the cells. With a sugar content of 2.1 g and less a con-
siderable_ decrease of the cell number in 1 ml substrate can be observed as
compared with the itinial number; at a level of 2.4 g of sugar, both in the usual
wort and in that with an addition of (N114)21004, only the surviving of ino-
culated, resting cells is found. Thus an increased nutrient level of phosphorus
and nitrogen does not produce an increase of yeast.
In these experiments a clear example of self-regulation of the amount of
yeast in the case of insufficient nutrition was observed. On the strength of these
'experimental results it is concluded that a yeast concentration of 20 g per litre
is well, suited for continuous fermentation of wood hydrolysate wort with
yeast ? backf low.
202
FERMENTATION COEFFICIENT AND PRODUCTIVITY
OF THE CULTURE
Experiments in which a small amount of sugar was fermented by a great
inoculum showed that the process is carried out by resting cells, without
proliferation, even if the substrate is repeatedly renewed. However, the fer-
mentation of sugar with small inocula leads to cell increase and also to an
increased alcohol yield. This increase of cells is but small and does not exceed
one half of the cell number of the natural population. It may be asked whether
it is possible to obtain in a concentrated culture an increase of cells greater
than the cells number of the natural population, and how much sugar will
be required to form one g of yeast under these conditions.
In order to solve this problem it was necessary to estimate the fermentation
activity of a young culture in wood hydrolysate wort.
The relationship between the fermentative activity and the yeast concentra-
tion, found earlier (Slator, 1908), enables the coefficient of the rate of ferment-
ation to be calculated. However, for the yeast which is required for the fer-
mentation of wood hydrolysate wort, it was sought to estimate the ferment-
ation coefficient experimentally under industrial conditions.
Yeast A3, separated in the plant, was mixed with wood hydrolysate wort
in such proportion that its final concentration was between 2.3 to 10 g per
litre. The duration of the fermentation process was defined as the time required
from the mixing of the wort with the yeast to the moments when the reducing
sugar in the wort ceased to fall off, the R. S. value being controlled each hour.
The results of these experiments are shown in Table 3. The last column shows
Living yeast
Fermentation
coefficient for
Volume of
Fermentable
Duration of
1 kg of yeast
fermenting
sugar
fermentation
Product
per hr
mixture
D
a
a X t
m3
kg
kg
01
"t" (lir)
rn
K
-
D X t
49
557
115
2.35
16.0
37.6
0.303
48
553
139
2-90
130
37.7
0.306
43
535
149
3.10
12.0
37-2
0.300
47
450
153
3.26
11.0
35.9
0.270
47
451
192
4I0
8.0
32.8
0.296
20
210
137
6-85
5 0
34-2
0-306
19
206
153
8.05
4.5
34.2
0.300
20
220
188
9.4
4.0
37.6
0.294
17
187
175
103
3-5
360
0.305.
Average t 0 300 kg
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
203
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
? A
the average rate of fermentation of sugar by one kg of yeast per hour accord-
ing to the formula, K =where K = fermentation coefficient, m =
?
= the amount- of fermentable sugar, D = living yeast in kg, and t = duration
of fermentation.:
These eiperinients show that 1 kg of yeast ferments on the average 0.3 kg
Of sugar perliour; i e. 7.2 kg in 24 hrs.
From these experiments it; follows that the fermentation coefficient repre-
senti the ratio Of sugar weight to yeast weight and duration of active ferment-
ation, when the cells are not starving and their alcohol producing capacity
thus not being impaired. It might be assumed that with such a level of carbo-
hydrate nutrition conditions Would be established for an increased multi-
plication of cells even in a Ooncentrated culture.
In order to:confirm this assumption the hydrolysate was fermented with
inocula of 10-20: and 40 g per litre at a sixfold renewal of substrate during
24 hrs. The, amount of fermentable sugar per one .1; of yeast was between
1.26 to 10-12 g; and in one 'variant of this experiment it corresponded to the
ferinentitien coefficient; i. e: amounted to 7.59 g per hr. The average results
of these experinients for a periodof 8 days (Table 4) confirmed that on increas-
ing thedlilramount of sugar up to 10.12 g with an inoculum of 10 g per litre,
the additional yeast multiplication above the original amount represented
on the average 37.2 'x 104 e. 7.2 x 10?, more than the cell number
Of the natural, population.
The total number of living cells in 1 ml increases in 8 days from 120 x 104 to
375 x 104. In the initial phases the yeast was not capable of fermenting the
given amount of sugar in time and the separated mash contained a high amount
of nonfermented sugar.
Table 4
Yeast inoculum
10
20
40 -
Number of renewals of media
per day
1
2
4
1 2
4
6
4 6
Sugar per g of yeast
2.53
5.06
10.12
1.26 2.53 5.06
7.59
2.53 3.79
Increase of cells 104/m1 .
8.0
20.0
37.2
0.0 27.5
42.2
56.2
6.0 9.0
Yield of alcohol from fermented
sugar,,%?pf theor. , ' ' -
75.6
77.6
76.8
77.2 77.7
77.5
82.2
76.6 77.4
Nitrogen in yeast, ?f, of dry
weight
6.83
7.27
7.60
6.40 6.81
6.96
6.65
6.71 6.8
?.
____? . _ _ _
Of particular interest is the fermentation process with an inoculum of 20 g
per litre and a sugar amount of -7.59 g, corresponding to the activity coefficient
of the yeast. The cell increase per ml here was 56-2 x 106, thus exceeding
twice the cell number of the natural population. in this experimental variant
the cell population corresponds to a continuous rather than stationary culture.
In addition to increased cell multiplication an increased alcohol yield is found.
Therefore it may be concluded that the capacity for alcohol formation is closely
related to the state of vegetation of the cells. The fermentation coefficient
for an inoculum of 20 g per litre can thus be considered as the main factor
guaranteeing a sufficient increase of cells and their high productivity.
REMOVAL OF WEAKENED YEAST DURING CONTINUOUS
FERMENTATION
It has been shown above that the use of yeast under conditions of insufficient
nutrition leads within a short time to self-regulation of the number of cells in
the population. The viability in this case is not determined by the concentra-
tion of cells, as assumed by Bayle, but by the conditions of the medium, by the
level of nutrition. The last experimental series convinces us that the level
of self-regulation should not serve as criterium for the maintenance of a high
activiby of the culture, as believed in industrial practice, but instead the level
of carbohydrate nutrition, ensuring a preponderance of cell multiplication
ova. their dying off, should be used as criterium. As shown by the experiment,
already on the fifth day a doubling of the cell number can be observed as
compared with the initial count, due to active multiplication of the cells.
It is clear that a double amount of cells requires an increase of the medium
inflow. However, under industrial .conditions, the amount of sugar supplied
for fermentation cannot exceed a certain limit. Therefore the only way of
maintaining a high metabolic activity of the cells is removal of the weakened
cells with normal nutrition of the yeast HMS.
Considering the different activities of suspended and sedimented cells an
attempt was made to carry out the removal of yeast from the sedimented mass.
A suspension of yeast in hydrolysate wort or fermented mash was placed into
a cylider of 9 litres capacity, provided at its lower end with a funnel and a stop-
cock of 25 x 15 mm boring, and left standing for one hour. Then by a rapid
turn of the stopcock one litre of mash was removed together with the sedi-
mented'yeast, which was then estimated by weighing after separation from the
mash. The sedimentation value was calculated according to the,formula
(B - A) x 100
PS -
205
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
where PS = percentage of sedimentation, A and B the concentrations of yeast
in g/1 before sedimentation and in the sediment, 8 the volume of the mash
in litres from which the yeast sedimented in time T.
Table 5
Dependence of sedimentation on the inoculum and concentration of sugar in the mash
Sugar in mash
%
Yeast, g/1
Sedimentation per hour
%
in suspension
in sediment
010.0
4.0
5.0
022.5
13.0
7.2
Mash without
I
025.0
20.0
10.0
fermentable sugar
035.0
35.0
30.0
040.0
121.0
38-0
045.0
181.0
50.0
2.86
20.0
5.0
3.1
0.71
20.0
10.0
6.2
1.30
25.0
6.8
3.4
1.70
25.0
27.0
13.5
1.60
30.0
6.2
2.6
0.61
30.0
85.0
35.3
2..86
400
11.0
3.0
0.65
40.0
17.6
49.0
Experiments showed (Table 5) that in a spent mash not containing ferment-
able sugar; the sedimentation of yeast was lineraly related to its concentration.
Thus at 25 g/1, 10% or even as much as 15% of the yeast sediments per hour;
at a concentration of 35-45 g/1, up to 50% of the yeast sediments per hour.
In a fermenting mash with a rapid evolution of carbon dioxide and at a sugar
concentration higher than 1%, a suspension of 30 or 40 g/1 remains suspended
and there is only a small percentage of sedimentation. However, at a sugar
concentration of less than 1% and a weak CO, formation, the suspension
shows the same picture as .the mash without sugar, i. e. up to 50% of yeast
sediments. This shows that continuous fermentation of wood hydrolysates
with high concentrations of yeast is not feasable_due to the rapid sedimentation
of the yeast. With an inoculum of 20 g/1 the situation is greatly different. Here
the main portion of yeast remains in suspension and only a small amount of
sedimentation _ occurs, irrespective of whether the CO, evolution is high or
low. Therefore the concentration of 'yeast should not exceed 20 g/1 for continu-
ous fermentation of wood hydrolysate with a sugar content of 3.0-3.5%,
since with these yeast concentrations little sedimentation occurs even during
stoppages of work., _
Yeast from a hydrolysate plant, fermenting for a prolonged period at a con-
centration of 18-20 g/1 with a fourfold renewal of medium per day was divided
i - nto two fractions_.- sedimenting and suspended,' - and these were examined
208 _
Declassified in Part - Sanitized Copy Approved for Release
for their alcohol forming capacity. Experiments showed (Table 6) that sedi-
menting yeast ferments more slowly and gives a lower yield of alcohol than
suspended yeast.
Table 6
No.
Yeast
RS in wort
%
Alcohol content of mash
Duration of
fermentation
hr
volume
0,
ii
yield per
10 kg sugar
% of theor.
yield
1
o
-
3
Suspended
Sedimented
Suspended
Sedimented
Suspended
Sedimented
3.26
3.26
:'?18
3.18
3.00
3.00
1.38
1.28
1.39
1.33
1.37
1.31
54.7
504
57.4
54.9
58.5
56.0
85.0
78.2
89.0
850
90.8
86.9
10
18
8
20
12
18
Therefore a fourfold renewal of medium per day does not remove. the inactiv-
ation of the sedimenting yeast. The removal of yeast sediment from the fer-
menting tanks represents thus an indispensable part of the regime of continuous
fermentation with backflow of yeast.
THE DYNAMICS OF FERMENTATION OF WOOD HYDROLYSATE
WORT
Carbon dioxide plays a fundamental role in the intensification of ferment-
ation and maintenance of yeast in suspension by its mixing effect of the con-
tents. In order that this factor be better utilised when working with high
concentrations of yeast, the dynamics of fermentation of wood hydrolysate
was studied in a plant where tanks of 100 n13 capacity were connected in series.
Continuous fermentation was carried out with a backflow of separated yeast
and at a flow rate of 25 m3 of mash per hour. The degree of sugar fermentation
in the individual tanks is shown in Table 7.
Table 7
Plant
consisting
of
RS
in wort
%
yeast
g/1
Fermentation of sugar
in % of initial tank No.
Fermented
total
%
Duration of
fermentation
hr.
1
2
3
4
2 tanks
2.96
20
76.8
3-7
-
-
80-5?)
71.".10"d??
3 tanks
2.92
14.2
69.2
8.6
2.7
-
80.5
4 tanks
2.97
10.2
61.8
16.1
2.0
1.9
80.0
111"..40mia?
,
5) The residual sugars in the mash are pentoses.
207
1.-,-,...lassumaremancon glea
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
-The ?auks- show that under all variations of work the main portion of sugar
is ferrikented in the 'first tank. In the remaining tanks at the most 0.2% sugar
of the Mash is fermented, this requiring, however, more time than the main
This finding can be readily explained by the fact that the end of the fer-
mentation' proceeds under considerably changed conditions, since the CO,
''hiings about mixing of the mash and thus ensures better contact of the
- - ,
yeast
with thesugar, evolves mainly in the first tank. It follows thus that the
fermentation of wood hydrolysates by the continuous method should be carried
dtit in a plant consisting of one or two tanks.
CONCLUSIONS
The-conditions of fermentation of wood'hydrolysates were examined, using
thi-metliiid' 'of renewal of substrate, in order to establish the basis for continuous
fernieptatio'n with yeast backflow. The following results were obtained:
1. Thiringileintentation-yith repeated use of yeast a greater number of cells
stir*ive, int 1'ml:0P:renewed substrate than in substiate without renewal.
Under conditions of renewal of substrate the viability and increase of
y6ast`cells are controlled by the amount of nutrients in the medium.
The carbohydrate nutrition is the decisive determinant of the accumul-
ation of yeaSt and, its productivity.
In' wood one g of yeast (75% water content) ferments in the
course of the fermentation period on an average 0.3 g of sugar per hour, or about
7.5 gper 24hr.11 ,
AAA. dally,leyelpf 7.5 g of sugar per 1. g of yeast, the sugar level being
achieyed either by,renewal or flow of medium, the state of the yeast corre-
sponils more to continuous than,stationary conditions and gives high yields of
alephol,a189,,:with,an:inpulum of 20 g/1.
, 6.11'0r:the, preservation of a,high productivity of yeast continuous ferment-
ation Of Wood hydrolysates should-,be carried out in a plant consisting of two.,
tanks with removal oflhe sedimenting weakened yeast.
REFERENCES
-
A1USTOV81AJA:T. V., 1955: K diskusiipo voprosu po6vennoj mikrobiologii. Mikrobiologija,
VOL- xxiy? ?3: 359.- -
.1',L4i.tii*--ri'rE.-c'1914rHe-fe-tind,Ciirttr:,tg, Stuttgart.
ViE'NK6VA V. A., 1055: Osecittnije droilej pri no
gicirOlizatov.,3iikrobiologija, vol. XXIV, 3: 348.
Ky4sti:ixo,V:.-V4I.Boausi.tvsk..A.717,fit..L;_j956:_,Korreljacia rneidu processaini razmno-
Bjit1;:Aktidllauk. yz4ekiikOj'ssii,;:N?9 5.
- ?
LEBEDEV S. V., 1936: Metod nopreryvnogo alkogolnogo brolenija, Pigepromizdat.
LEONARD R. If., ITAJNY J., 1945: Fermentation of Wood Sugars of the Alcohol. End. and
Eng. Chem., 5: 37.
MA-LEK L, 1956: ProtoZnyj metal razmnacnija mikrobov. 3likrobiologija, XXV, 6: 659.
MONO!) J., 1950: La Technique do Culture Continue, Thtiorie et Applications. Ann. last.
Pasteur, 79: 390.
Nrcovicor V. K., OCHRIMENKO 0. I., 1947: Izaenie kolloidnych vegilestv drovesnych
gidrolizatov i ich koli6estvennoje oprediHenie. Trudy VNIICS, vol. 2: 202.
8ARKOV V. I., 1950: Cidroliznoe proizvodstvo, vol. In.
SLAirott A., 1906, 1908: Studies in fermentation. Part T. The Chemical dynamics of Alco-
holic Fermentation by Yeast. J. Chem. Soc. 83: 128; Part 11'. 93: 217.
ZUBKOVA S. R., ICO6DCOVA N. V., ZA0 M. R., 1936: K voprosu polti6enija spirta iz tiro-
vesnych gidrolizatov. Biochimija, vol. 1, No. 1: 63.
14?Symposlum
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
209
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
FORMATION OF NEUTRAL SOLVENTS
IN CONTINUOUS FERMENTATION
BY MEANS OF CLOSTRIDIUM ACETOBUTYLICUM
JOSEF DYR, Jritf PROTIVA and ROMAN PRAM
This paper deals with the formation of neutral solvents in the continuous
fermentation process by means of Cl. acetolndylicum. Continuous fermentation
has been applied with advantage to some classical fermentation processes even
on a pilot plant scale. Attempts have been made to use this continuous method
also for other industrial fermentation processes but investigations carried
out so far have been only on laboratory or semi-pilot plant scale. This applies
also to the acetone ? butanol fermentation process which .is a comparatively
young branch of the fermentation industry. Althought acetone-butanol batch
fermentation has been successfully employed for decades in different countries,
few reports have been published on the continuous cultivation of strains of
Cl. acetobutylicum producing neutral solvents.
The report of Logotkin (loc. cit. by Mitlek, 1955) is one of the available
papers on the continuous cultivation of Cl. acelobutylicum and deals with the
evolution stages of Clostridia under various conditions of a continuously
supplied fresh cultivating medium. However, no data on the course and yields
of the fermentation are given. Another work by Nowrey and Finn (1956)
describes the repeated propagation of a culture of Cl. acetobutylicum. On the,
contrary, a number of papers have been published which seem to reject the
possibility of carrying out acetone-butanol fermentation by the continuous
method.
Many investigators (Dyr, 1952; Kutzenok and Aschner, 1952; Hejnalova,
1955; Nowrey and Finn, 1956) showed that repeated transfers of the culture
of, Cl. acetobutylicum at the stage of acid formation resulted in a diminished
ability. to form neutral solvents in the following stages of the fermentation,
in a longer acid forming Period and after a certain number of transfers it
_ .
brought about physiological degeneration of the culture.
,These-finilings were -elucidated in different ways. Some authors, admitting
210
Declassified in Part - Sanitized Copy Approved for Release
?
the evolution cycle of Cl. acetobutylicum, suppose that the culture, which owing
to repeated transfers .at the acid formation stage did not end its evolution
cycle with spore formation, impairs the process of its evolution and loses its
ability to respond to the outer environmental conditions with the formation
of neutral products. The above mentioned authors, Nowrey and Finn reported,
that the butyric acid bacteria degenerated after a number of transfers owing
to the spontaneous selection of mutants at a low pH, which was reached at the
initial stage of fermentation. Thus acetone-butanol fermentation appeared to
be unsuitable for the continuous method inevitably requiring prolonged
propagation period of the culture without losing their physiological activities
in any way.
In the present paper we report the results of some of our experiments with
the continuous acetone-butanol fermentation technique on laboratory scale,
carried out in 1956-1957.
METHODS
MICROORGANISMS AND CULTIVATION
Cl. acelobutylicum, strain Ca 3, isolated by Dyr in 1946, was used for our
experiments. The strain was maintained by conventional microbiological
practices in potatoe mash and transfererd once in two months, the culture
being subjected to heat shocks. For experimental purposes fresh subcultures
were always used.
Repeated vegetative transfers of the culture Cl. acetobutylicum, strain Ca 3,
were performed in a potatoe mash containing approximately 4% w/v of starch.
The culture was propagated after 12 or 24 hours by inoculating 500 ml of
fresh potatoe mash with 10 ml of fermenting culture without the application
of shocks. These subcultures were allowed to complete their fermentation
for another 48 or 60 hours, respectively, and determinations of the acetone
content and titratable acidity were made.
Continuous cultivation was carried out in a complex liquid medium consisting
of tuber water and 4% w/v of glucose. The pH values of the medium prior
to sterilisation was 6.5, the titratable acidity after sterilisation being 0.5 to
0.9 of 1N NaOH.
CONTINUOUS FLOW SYSTEM
Continuous propagation and cultivation of Cl. acelobutylicum were carried
out in cylindrical glass vessels fitted with side outlets. The contents of each
vessel were continuously agitated by means of stirrers slowly rotating at 200 r.
14?
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
211
- - ,
STAT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
p. m. Fresh medium was fed by means of a slightly modified Mariotte bottle
based on the gravitational flow of the liquid under steady hydrostatic pressure,
*as described by Attila (1955) and further elaborated by rti6ica (1956). This
equipment is illustrated diagrammatically in Fig. 1.
Throughout the experiment the
storage container was connected
with the outer atmosphere only
with the vent pipe A reaching with
its lower contracted end the bot-
tom of the storage container.
This arrangement made the feed
independent from the level of the
liquid surface inside the storage
container, because the theoretic
surface of the liquid was on the
same level with the mouth of the
lower end of the pipe A. The rate
of the feed was controlled by ad-
justing the height of the surface
levels (A) at the end of the over-
flow pipe C (L1) and those of the
storage container (L2). The neces-
sary fine control of the flow was
attained by means of a capillary
of a suitable diameter (R) incor-
porated into the flow system of
the fresh medium. The feed system
and the fermentation vessels were
placed in a thermostatic chamber.
Fig. 1. Feeding equipment (schematic).
A a II ? vent tubes; C ? overflow tube; D ? stor-
age container; R ? capillary resistance; L1 ? sur-
face level of cultivating medium in the overflow
tube; L, ? theoretical level of cultivating medium
in the storage container; A ? value of hydrostatic
pressure.
ANALYTICAL METHODS
. Butyl alcohol and ethyl alcohol were determined by Johnson's method (1934),
acetone according to GOodwin (1920) and glucose according to Shaffer-Hart-
niann (1921). The bacterial dry weight was determined by weighing, the titrat-
able acidity by titrating 10 ml of the centrifuged cultivating medium with
0.1 N NaOH, using phenolphthalein as indicator.
212
EXPERIMENTAL
VEGETATIVE TRANSFER OFA CULTURE OF C L. ACETO 13 UT I'L I U .1!
We started by checking earlier experiments by transferring vegetatively
a culture of Cl. acetobutylicum, strain Ca 3. Forty transfers of the culture at
24 hours intervals were made. From the data in Tab. 1 kreproduced in concise
form) it can be seen that even after forty transfers the culture did not tend
to an increased formation of fatty acids, neither to a lower production of neutral
solvents. In the course of repeated vegetative culture transfers only the appear-
ance of the fermenting mash was changed. Approximately after the 30th
transfer the fermenting liquid lost its characteristic appearance, the solid re-
mainders of the mash did not rise to the surface and the foam formation was
reduced. There is no doubt that together with the vegetative forms also a cer-
tain amount of cells with prospore and spores were transferred. They were
formed owing to the heterogenuity of the cells in the culture already at the
logarithmic phase of growth. This applied even in the case of transferring 24
hour old cultures. The experiment showed that repeated transfers of cultures
without heat shocks and in the stage when most cells have not yet started
sporulation did not necessarily cause lower physiological activity. By this the
problem of the neutral solvent production by continuously propagated cultures
of Cl. acetobutylicum was not entirely solved, since even cultures aged 24 hours,
i. e. at the beginning of the second stage of fermentation, retained their ability
to produce acetone, butanol and ethanol. It should be pointed out that the
biochemical behaviour of the transferred cultures after forty transfers was
studied only in the laboratory, not in the plant.
Table 1
Analysis of fermented liquors of vegetative passages of the culture of Cl. acelobtaylicum, strain Ca 3
PassagePassage
Acetone
ing/m1
Titratable
acidity
Acetone
ing/m1
Titratable
acidity
1.
3-86
3-2
9.
3-86
3-5
2.
1-64
3-3
10.
3-96
3-6
3.
4-83
3-2
15.
5-02
3-0
4.
4-15
30
20.
4-54
3-1
5.
3-86
3-3
? 25.
4-25
2-8
6.
348
3-1
30.
4.64
2-9
7.
3-28
3-2
35.
4-15
2-7
8.
2-70
3-1
40.
4-54
9.7
Cultures wore transferred at 24 hours intervals.
Cultivating medium: potato? mash containing 4% w/v of starch.
Fermentation was completed in 72 hours.
213
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Another series of experiments was made with cultures transferred after
12 hours into a fresh potato? mash in a similar way as previously described.
The yields of acetone and the values for the titratable acidity arc given in
Tab. 2.
Table 2
Analysis of fermented liquors of vegetative passages of tho culture of Cl. acetobtitylicton, strain Ca 3
Passage
Acetone
mg/m1
Titratable
acidity
Passage
Acetone
mg/ml
Titratable
acidity
1.
3.67
2.9
13.
4.25
3.0
IL
3.67
3.0
14.
3.67
2.9
3.
3.67
3.1
15.
3.38
2.9
4.
4.06
3.0
16.
4.06
3.0
a.
3.38
3.0
17.
4.30
3.0
6.
4.06
3.0
18.
4.06
3.0
7.
3.57
3.5
19.
4.74
3.0
8.
3.67
2.9
20.
3.72
3.0
9.
3.96
3.0
21.
3.81
3.2
10.
3.96
3.0
22.
3.52
3.2
11.
4.25
3.0
23.
3.01
2.9
12.
4.35
3.0
24.
3.43
3.0
-
Cultures were transferred at 12 hour intsrvals.
Cultivating medium: potato? mash containing 4% w/v of starch.
Fermentation was completed in 72 hours.
After the 24th transfer the experiment Was stopped, the culture being in
good physiological condition. It maintained its usual appearance until the
20th transfer when after further transfers reduced foaming was observed.
In this experiment with transfers at 12 hour intervals the fresh media were
inoculated with an inoculum in the logarithmic phase and thus the conditions
resembled to a considerable extent those of the continuous flow system.
Only a small number of cells which have acquired the ability to produce acetone
was transferred together with the young propagating cells. This may be assumed
owing to the observation that in batch fermentation the culture entered the
second reducing stage in a short period of time, which implied great biochemic-
al.uniformity of the cells constituting the culture.
CONTINUOUS PROPAGATION OF CL. ACETOBUTYLICUM
With the results obtained in the previous experiments we started to cultivate
Cl. atetobutylica ni, strain Ca 3, under flow system conditions. Laboratory
investigation of the continuous acetone-butanol fermentation required pri-
maSly a proper cultivation medium, So that simple and reliable feeding equip-
214
_ - -
merit might be used. For small operating volumes media of low viscosity,
free of insoluble particles, proved to be the most suitable. A number of them
was tested and the yields of acetone and fatty acids determined. They were
made up of easily available materials, rich in growth factors and desired forms
of assimilable nitrogen, such as corn-steep, malt germ extract, yeast autolysate
and malt wort. Though most of the tested media showed desirable qualities
in batch fermentation, all turned out to be unsuited for continuous culti-
vation. For all experiments described in this paper a liquid medium of tuber
water and glucose was used.
Experiments attempting continuous propagation of the particular strain
of Cl. acetobutylicum were started by following the effect of continuous pro-
pagation of a culture on its ability to form neutral products in the course of
further batch cultivation. The experiment was carried out as follows: the culture
was kept at the stage of intensive growth in the cultivation vessel into which
fresh medium was fed at such a rate that the titratable acidity of the cultivation
medium in the vessel or the cultivation liquid running out of the vessel was
maintained at the same value. For this case, limiting values for titratable
acidity, i. e. ml 1N NaOH per 100 ml, were established. They were
slightly lower than the maximum values obtained in the batch fermentation
process. Preliminary experiments showed that 45-50 ml per hour of fresh
medium had to be Jed to maintain the fixed values for titratable acidity, pro-
vided that the capacity of the fermentation vessel was 250 ml. That indicated
that theoretically once in 5 hours the contents of the vessel had to be changed.
Approximately 50 ml of the medium in the cultivation vessel were inoculated
Table 3
Analysis of fermenting liquor leaving the vessel (I) and after 72 hours of batch fermentation (Ti)
Hours
Glucose
mg/m1
I
II
T. A.
13.
mg/ml
A.
mg/mi
E.
ing/m1
B. C.
mg/m1
Glucose
mg/mi
T. A.
13.
mg/m1
A.
ing/m1
E.
mg/m1
0
38.3
0.8
-
-
-
-
-
-
-
-
-
3
33.2
1.7
0.0
002
0.24
0-52
0.95
2.6
7.75
2.53
2-80
15
292
3.0
00
023
1.32
0.77
0.50
2.2
S.40
2.46
2.96
27
23.0
30
1.90
004
0.60
1.05
0.0
2.1
7.S7
2.0S
2.73
39
21.7
2.9
2.98
0.98
0.49
0.73
00
1?9
S.37
2.24
2.63
51
22.6
2.4
2.46
0-87
0 50
0.73
0.0
2.3
7.62
2.22
2.14
63
23.9
2.9
2.62
0-86
0.70
1?15
0.0
2.4
701
2.14
3.35
75
20.4
2.9
2.86
1.02
0.83
1.20
0 0
2.7
8.04
2.28
2.86
88
21.0
2.7
2.83
1.38
0.93
1.16
00
2.5
S?59
2.66
2.35
120
20.8
2.S
2.75
1.29
0.95
1.08
0.0
2-6
8.33
2.42
2.25
T. A. = titratable acidity; B. = butanol; A. = acetone; E. = ethanol; B. C. = bacterial cells
Cultivating medium: tuber water with 3.80% w/v of glucose.
Flow rate: 80 nil/hr., dilution rate: 0.16 hr.-1, hold-up time: 6.25 hrs.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
215
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
. - -
With 5 ml of a 24 hour old culture of Cl. acelobtaylicum, strain Ca 3, grown
in potatoe magi. Three hours after inoculation the feed of the fresh medium
was ,20 ml/hr, after another five hours 50 ml/hr. During the experiment the
_
appearance and the activity of the culture in the cultivation vessel were follow-
ed and 200 ml samples of the fermenting medium leaving the vessel were
withdrawn and allowed to complete their fermentation in 72 hours. Then
analyses were made, the results of 5 days continuous cultivation being given
in Tab. a.
At a theoretical total volume change of the vessel every 5 hours, the values
of titratable acidity were 2.4-2.9. The bacterial dry weight had an initial
rising- tendency and after reaching the steady state it settled at the value of
about. 0.5 mg/ml of the medium. The adaptation of the culture to the new
conditions was manifested also by the rapid utilization of the glucose and by
the formation of neutral solvents, which had a rising tendency until the 60th
.1ufiur-of the experiment. Similarly, in the first 60 hours also the mobility of the
Clogridia- WM increased: Morphologically the culture was uniform throughout
the whole experiment and constituted by typical short cell forms. No spores
were noticed:- The yield of the experiment (with the exception of the first day)
amounted to about 35% w/v. Rather unusual is the ratio of the solvents pro-
duced; the amount of butanol and ethanol was higher, whereas the acetone
content was slightly lower in comparison with the batch fermentation process.
(Tab. 4.)
Table 4
Comparison of yields and ratio of neutral solvents in batch
and continuous cultivation of Cl. azetobtaylicum
CultivationB.
Yield of neutral
solvents in % w/v
: A. : E.
Batch ?
Continuous
32
33
4.5: 2'7: 2.8
5-2 : 2-7 : 2.1
B. = butanol; A. = acetone; E. = ethanol.
Cultivating medium: tuber water containing 3?90% w/v of glucose, titratablo acidity 0-7.
Bitch fermentation was completed in 100 hours, the total hold-up time of four-vessel fermentation
system being 30 hours.
From-_ the average rate of glucose utilisation determined in the previous
experbitentsmid,amounting to about 0.2% w/v per hour, it was assumed that
40% of glucose in the medium, was, utilized in 20 hours in the continuous
fe-riiientiitiOn'eeis: 'Act-4141y, however, More time was needed owing to the
accumulation of metabolic end products in the medium and their inhibitive
effeetS;
218
Declassified in Part - Sanitized Copy Approved for Release
CONTINUOUS FERMENTATION
For the experiments with continuous fermentation by Cl. acelobulylicunt
three successively connected vessels fitted with slowly rotating stirrers of a total
volume of 1500 ml were used. The flow system of the fermentation is illus-
trated in Fig. 2.
3
Fig. 2. Diagram of the three-vessel flow system.
A ? inflow; B ? effluent; C? C,, C, ? sampling points; D ? vent tube; E? E,, E, ? mechanical
stirrers.
According to the expected fermentation rate the flow rate was 50 ml/hr.,
so that the partial retention time was 10 hours and the total retention time
30 hours. In the first vessel only the growth stage and in the other two
reducing stage should take place. This applied only if the cells in all vessels
were approximately of the same age and if the medium passed through the
whole series of vessels at a constant rate. However, the analytical cross-section
of the continuous three-vessel-cultivation of Cl. acetobidylicum (Fig. 3) showed
that even in the first cultivation stage there was a considerable percentage
of aged cells, producing acetone as well as butanol and ethanol. The data for
bacterial dry weight indicate that the propagation of the cells took place
only in the first cultivation stage.
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
217
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
A considerably increased amount of fermentation products in the second
vessel proved that predominantly reducing processes had set in. A slightly
higher titratable acidity with small differences in the individual stages was
in agreement with this observation. During the first four days of cultivation,
the activity of the culture rose to such an extent that in the first cultivation
1.0
-0.5
0
II III
Fig. 3. Analytical cross-section of continuous acetone-butanol fermentation in the three-vessel
flow system.
Cultivating medium: tuber water containing 3.6% w/v of glucose; titratable acidity 0.5. Flow rate:
BO ml/hr., dilution rate for the first vessel: 04 hr.-1, total hold-up time: 30 hrs. Curves: 1 ? dry
weight of bacteria; 2 ? butanol; 3 ? titratable acidity; 4 ? acetone; 5 ? ethanol; 6 ? glucose.
Ordinate: hold-up time in hours: abscissa; A ? glucose mg/ml; B ? neutral solvents mg/ml;
C ? titratable acidity; D ? dry weight of bacteria mg/ml. Abscissas marked by Roman figures
sinclinate the retention times in individual vessels.
vessel more than half of the final amount of butanol was produced and as
much as 00% of the glucose originally present was utilized. From this it
follows that the throughput rate was lower than the retention time in the first
fermentation vessel, corresponding to the logarithmic phase of growth. Con-
siderably more butanol and acetone were produced, whereas the value for
r_
ethanol slightly varied. Increased butanol formation was evident, especially
in the 'second vessel, during the 4th day of fermentation. At the same time
all the'ghiceSe-present Was utilized On the 5th day, spores appeared also in the
first chltivatiOn-stage'ind hi the third cultivation Stage filamentous bacterial
-
fOrnis.in- all, three ;Vessels were found the following day Simultaneously, the
titratable acidity rose in all is and the formation of' the final products
'218
Declassified in in Part - Sanitized Copy Approved for Release
and the rate of glucose utilization were identified, the glucose being utilized
the 7th day only to 60%. It was supposed that the activity of the culture
weakened owing to the increasing amount of inhibiting products, especially
that of butanol, in the first cultivation stage. Thus the bacteria propagated.
under permanently unfavourable conditions. The dilution rate was, therefor,
50
40
20 ?.
10-
0
10
20
30
I II
III
IV
- 1.0
0
Fig. 4. Analytical cross-section of continuous acetone-butanol fermentation in the four-vessel
flow system.
Cultivating medium: tuber water containing 3.9% w/v of glucose, titratable acidity 0.4. Flow-rate:
50 mi/hr., dilution rate for the first vessel: 0.2 hr.-1, total hold-up time: 30 hrs. Curves: 1 ? dry
weight of bacteria; 2 ? butanol; 3 1- titratable acidity; 4 ? acetone; 5 ? ethanol; 6 ? glucose.
Ordinate: hold-up time in hours; abscissa: A ? glucose mg/nil; B ? neutral solvents mg/nil;
C ? titratable acidity; D ? dry weight of bacteria mg/ml. Abscissas marked by Roman figures
stand for retention times in individual vessels.
increased during the 7th day, so that the total volume of the fermenting system
was changed once in 25 hours, which led to a partially improved state of the
culture. The filamentous forms disappeared and the culture at the first pro-
pagation stage consisted of short motile cells. The yield rose during the 10th
day to 34% w/v, the remaining content of reducing substances being 0.5% w/v.
The average yield of the whole experiment, including the period of degeneration
of the culture, was about 30% w/v of neutral solvents. The time required for
utilization of the glucose present was by about 45% shorter compared with
the time of the periodical method.
Attempting to avoid the difficulties in the continuous cultivation of Cl. aceto-
butylicuin- which may be caused by an intolerable amount of toxic metabolic
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
219
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
products as early as in the first cultivation stage, the cultivation system was
divided into four stages, the first two of which had a capacity 250 ml, the other
two 500 ml each: The time ,for the volume change in the first two stages was
by 50% shorter than the time in the third and fourth stage. With this arrange-
ment the exchange of the medium in the first vessel was doubled, so that
it' could be expected that no accumulation of toxic substances would occur
at the stages with the greatest propagation rate of bacteria. The analytical
cross-section of the experiment in the four-vessel-system is illustrated in
Fig. 4.
? With-a retention time of 5 hours in the first cultivation stage half the time
was required for doubling of the bacterial dry weight compared with the pre-
vious experiment. The bacterial dry weight in the course of a 12 days' experi-
ment was the highest in the first cultivation stage. The enhanced flow was shown
to favour the fermentation activity of the culture. On the 9th day of cultivation,
the medium _flowing out of the third vessel contained as little 88 0.08% of
reducinisubstances, indicating that the fermentation was sufficientlly com-
pleted in 20 hours. In spite of the initial activity the culture degenerated even in
this case. The cells, short -and extremely, motile, changed into extended fila-
mentous forms even in the first two cultivation vessels. These degenerated
forms, reputedly hiving a low fermenting power, could not be removed even
by the enhanced:flow in the last three days of cultivation.
With regards to these unfavourable facts, the media usedin the experiments
were checked. They were subjected to batch fermentation to offer the possi-
bility of comparing both methods of fermentation and active substances
extracted from potatoes were estimated. Analyses of the total nitrogen in the
potatoe tubers and in the extracted tuber water showed that only 29% of the
total- nitrogen was obtained in the preparation of the media, based on the
amount of potatoes used. In this respect the employed medium was very poor
in comparison with the natural fermentation mashes and it can be assumed
that this was the cause of the degeneration processes in the continuous cultiv-
ation.
From the batch-fermentation experiments data for the comparison of the .
various cultivation methods were obtained and changes in the final ratio
of the neutral solvents produced by continuous fermentation were partially
elucidated. The comparison of these experiments is shown in Tab. 4.
The batch fermentation in tuber water, supplemented with 4% w/v of
glucose, proceeded- comparatively slowly and it took 96-100 hours to utilize
the glucose completely. The final- ratio of solvents 4.5% w/v of butanol,
2.7% w/v of acetone and 2.8% w/v of ethanol,was approximately identical to
that, Of the continuous fermentation.-The yields were 32% by weight in both
for the continuous and batch fermentation processes, except for the time, which
was three times as longin the case of batch fermentation.'
220- ,
npriassifipri in Part - Sanitized Copy Approved for Release
?
The results obtained leave no doubt as fo the possibility of employing the
continuous method for acetone-butanol fermentation. Tho difficulties arising
in the course of our experiments can be ascribed to insufficient differentiation
of the growth- and neutral solvent
production stage. In further experi-
ments attention was called to the
output of bacteria in the first vessel
of the fermenting system. It can bo
assumed that the conditions securing
the maximum output of bacteria at
the initial stage of the fermenting
system and shifting the productive 0.6
stage to the final cultivation stage
will have a decisive effect on the
entire course of the continuous ace-
tone-butanol fermentation process.
Theoretical conditions for the
growth of microorganisms in the con-
tinuous cultivation process, which
are mentioned in the works of Monod
(1950), Maxon (1955) and Herbert
et al. (1956) indicate that the con- 0 0
4
centration of bacteria in the culti-
0 0.2 0. 0.6 0.8 to
vating medium and the bacterial
output, based on one volume unit
of the medium in the vessel, are pri-
marily affected by the volume ratio
of the medium flowing through the
fermentation vessel to the fermenting
medium for a certain unit of time.
Designating the volume of the fer-
menting medium as V and the volume of the feed as F, the ratio is F/V and is
termed dilution rateD.Theoretically it may have the value from zero to infinity.
Actually, however, the dilution rate has a limited value. If the feed passes
through the particular fermentation vessel containing a static microbial
culture in the logarithmic phase of growth, the value of the dilution rate is
above zero. If the dilution rate is increased and the steady state established
in the fermentor, the bacterial concentration in the cultivation vessel changes
with the increasing dilution rate only very little, whereas the bacterial output
rises rapidly. If the dilution rate attains a certain value, termed the maximum
dilution rate, the concentration of bacteria in the culture gradually falls and the
bacterial output reaches its maximum. If the dilution rate is further increased
A
Fig. 5. Steady.stato relatioships of bacterial con-
contration and output of bacteria to the increasing
value of dilution rate in the continuous culture
of Cl. acetobutylicum, strain Co.3.
Curves: 1 ? output of bacteria; 2 ? bacterial
concentration. Ordinate: dilution rate I) in hr.-1;
abscissa: A ? bacterial concentration gil of ope-
rating volume; 13 ? output of bacteria gihr./I of
operating volume. D ? maximum dilution
rate; D, ? critical dilution rate.
5 -Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
921
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
to the value termed "Critical dilution rate" both the concentration of bacteria
in the fermentation vessel and the bacterial output falls rapidly and the cells
are washed out of the fermentor.
A series of experiments was conducted with the varying dilution rates in the
range of 0.20-0.60 and the uptake of glucose, concentration of bacteria,
A B C D
60- 12' 6
io ? -5 -1.o
40-
30-
20-
10-
0
10
5
---o
6
0 -0
' 30
I II UI
Fig. 6. Analytical cross-section of continuous actcone-butanol fermentation in the five-vessel
flow system.
Cultivating medium: tuber water containing 4.1% w/v of glucose, titratable acidity: 0.4. Flow
-rate: 79.5 ml/hr.; dilution rate of the first vessel: 0.3 hr.--1; total hold-up time: 26 hours. Curves:
1 ? dry weight of bacteria; 2 ? butanol; 3 ? titratable acidity; 4 ? acetone; 5 ? glucose; 6 ?
ethanol. Ordinate: hold-up time in hours; abscissa: A ? glucose mg/ml; B ? neutral solvents
mg/m1; C ? titratable acidity; D ? dry weight of bacteria mg/ml. Abscissas marked by Roman
figures stand Tor the retention times in individual vessels.
bacterial output, titratable acidity and accumulation of neutral solvents in the
medium leaving the first fermentation vessel were determined. The values
obtained are shown in Fig. 5. The shape of the bacterial concentration curve
and that for the yield of, bacteria dependent on the increasing dilution rate
were in full agreement with the theoretical assumptions.
It was sheivn that the optimum dilution rate for the production of the
cellular matter of Cl. acetobutylicum under the conditions employed was
0.3, -Which is lower than the maximum dilution rate (D31)
-?- -
the Villue, Of Which, is about 0.36. The critical dilution rate (Dc) derived by
extrapolatioliji approximately 0.7.
.--Further eipeFinieas *ere Conducted to follow the neutral solvent production
222..
by continuous fermentation of a five-vessel cultivation system. The cultivating
medium was fed in such a way that in the first vessel, where maximum pro-
pagation should have occured, the dilution rate approached the value of 0.3.
Under those conditions the retention time of the first vessel was 3.3 hours,
the time for the total volume change of the whole fermentation system being
26-0 hours. The experiment was stopped after 1$ days of continuous ferment-
ation. An analytical cross-section of the fermentation system on the 12th day
of continuous fermentation is illustrated in Fig. 6.
DISCUSSION
These experiments showed that the culture of Cl. acdobutylicum kept its
activity after repeated vegetative transfers in the laboratory stage. The results
obtained with the culture of Cl. acetobutylicum, strain Ca 3, did not .agree
with some earlier published data. In the experiments of llejnalovri (1955) the
degeneration of the culture set in after a few vegetative transfers. According
to the statement of Nowrey and Finn (1956) the culture of Cl. dcdobutylicum
lost its characteristics only after four 24 hour transfers, which corresponded
to about 23 generations. They are supposed to have confirmed the well known
theory that butyric acid bacteria degenerate after a number of transfers
owing to the spontaneous selection of mutants at low pH values in the first
stage of fermentation. Kutzenok and Aschner (1952), experimenting with
a selected strain of Cl. but ylicum, observed anomalous behaviour of the cul-
ture already after the sixth or ninth passage without heat shocks. On further
passages the cultures lost their activity till they died which occured after the
eighteenth transfer at the latest. On the basis of available data and his own
experience in acetone-butanol fermentation in the laboratory and plant, Dyr
(1954) came to the conclusion that the culture of Cl. acetobutylicum must pass
through all evolution stapes and close its metabolis cycle if it is to be used
in the plant. The culture not closing its life cycle after a number of generations,
whether owing to unfavourable conditions of cultivation or repeated transfers
of the culture in the logarithmic phase of growth (i. c. in the acid producing
phase), does not accumulate certain enzymatic systems and is incapable of
responding to the changed environmental conditions with neutral solvent,
production. In the experiments described in this paper employing a greatly
active strain of Cl. acetobutylicum, we succeeded in keeping the culture in good
condition even after forty vegetative 24 hour transfers and twenty four 12 hours
passages. Roughly culculated, the culture produced 220 generations and 120 ge-
nerations respectively. The number of transfers conducted was not limited,
as shown by the unchanged activity of the culture at the end of the experiment.
Positive results were obtained also with the continuous propagation of
Cl. acetobutylicum in one vessel only. Samples were taken from the continuous
npriassified in Part - Sanitized Copy Approved for Release
?
'
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
993
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
fermentor in the' course of a five days' experiment and allowed to complete
their fermentation in the usual time with the average yield of solvents of 34%.
The concentration of the 'bacterial dry weight in the cultivation vessel re-
mained at the value of about 0.5 mg/ml, i. e., approximately 60 % of the
maximum amount obtained by batch fermentation. The doubling time of
this experiment was about 3.5 hours, which was considerably longer than
that of the logarithmic phase of growth in the batch fermentation process.
The prolonged, so called "generation time" was not caused by the conti-
nuous cultivation itself, but was the result of the steady states attained
in the culture, which were controlled by the dilution rate in the presence of
excess nutrition sources (Monod 1950, Herbert et al. 19E6). The dilution rate
employed in these experiments for the Cl. acetobtaWairs jar from being the
maximum. Nowrey and Finn (1956), conducting continuous propagation of
Cl. acetobutylicum, worked with a very low concentration of bacterial dry
weight in the vessel. They obtained about 650 generations in a one vessel
continuous propagation after a fortnight, which means that the "generation
time" was a little longer than half an hour. The low yield of such a propagation
is obvious.
In the three-vessel fermentation system, the propagation of microorganisms
occured predominantly in the first stage at the dilution rate of .D = 0.1.
When the steady state has been established, the time needed for doubling the
dry weight (td)was about 7 hours. (td is a constant characteristic only for the
continuous cultivation and can be compared with the "generation time" of
a microbe in the logarithmic phase of growth of a static culture only at a maxi-
mum dilution rate. It is expressed by the following equation (Herbert et al
1956):
1 V
t d = ?Din 2? ?FIn 2) .
The propagation phase even in the four-vessel system, where the first two
vessels contained 250 ml each and where the dilution rate D = 0.2 was used,
took place practically only in the first cultivation stage at a doubling time
of the bacterial dry weight of only 3.5 hours and a volume change time of
about 5 hours.
Compared with the conditions for acetone-butanol batch fermentation in the
same medium one of the advantages of the continuous fermentation was obvious;
the time needed ,for the growth of the total bacterial dry weight in the conti-
nuous fermentation process was by 3/5 shorter when the steady state had
been established than that of batch fermentation process lasting at least
12 hours. However, in spite of employing a doubled dilution rate complete
separation of the propagation stage from the production_ stage could not be
achieved. Owing to the accumulation of a considerable amount of fermentation
- 224
Declassified in Part - Sanitized Copy Approved for Release
?
end products the young proliferating cells were injured even in the first
cultivation stage and lost their physiological activity and gradually degenerated.
According to C'ekan (1934), especially acetone and butanol were very toxic,
butanol having a strong inhibitive effect on Cl. acetobutylicum at, a concentra-
tion as low as 1% w/v. These difficulties with the continuous fermentation
of Cl. acetobutylicum, strain Ca 3, were successfully avoided by increasing the
dilution rate, which resulted in a lower concentration of toxic products in the
medium and the regeneration of active cells. It is supposed that the separation
of the propagation stage from the neutral solvent production stage could be
attained either by adjusting the dilution rate in the vessels, where the pro-
pagation takes place or by inducing such carbohydrate conditions which would
provoke competition for the substrate between the proliferating cells and the
cells producing neutral solvents. Experiments with varying dilution rate
showed that the best separation of the bacterial growth stage from the pro-
duction stage was achieved at a dilution rate approaching the maximum, i. 0.,
at the value of D = 0.3. In continuous fermentation, conducted under these
conditions in a five-vessel apparatus, the maximum bacterial growth occurred
in the first vessel. In the first two vessels acids were formed and the neutral
solvent production was most intensive in the third and further vessels. The
number of generations after eighteen days was 190.
No morphological adaptation to the altered way of cultivation was observed
in any of our experiments. The presence of filamentous forms and prolonged
cells was always accompanied by diminished fermentation activity of the
culture and obviously by a degeneration process. Similar morphological
changes are often observed in degenerative batch fermentations. The formation
of spores was observed only in the final stages of favourable continuous fer-
mentation processes.
REFERENCES
eicKAw L. I., 1934: Vlijanije rastvorit6lej na acetobutilovoje broienijo. Mikrobiologija,
3: 266.
DYR J., 1952: Problemy pfipravy kultur pH butanol-acetonovein kvatleni. Prinnysl
potravin, 3: 262.
DYE J., 1954: Vfroba organick3%ch rozpouilt6de1 kvasnou c,estoti. Sbornik ptednallek
veldecke /conference v Bansk6 8tiavnici: 59.
GOODWIN L. P., 1920: Modification of the Messinger Method for Acetone Determination,
J. Am. Chem. Soc., 42: 39.
HEJNALOVA D., 1955: Diplomovti prace, Karlova universita, Praha.
HERBERT D., ELSWORTH R., TELLING R. C., 1956: The Continuous Culture of Bacteria ?
a Theoretical and Experimental Study. J. Gen. Microbiol., 14: 601.
JERUSALIMSICIJ N. D., 1946: 0 fiziologiiSeskich stadijach v razvitii baktcrij. Milcrobiologija,
15: 405.
15--Symposium
50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
225
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
JONNSOM M. J., 1932: Determination of Small Amounts of Ethyl and Butyl Alcohols.
Ind. ,Eng. Chem. Anal. Ed., 4: 20.
KUTZENOIC A., ASCHNER M., 1952; Degenerative Processes in a Strain of Cl. Butylicum.
J. Baia., 84: 829.
MAusic I., 1956: 0 mnoteni a p6stovaini mikroorganismil, zvligt6 hakterii. Praha: 102.
MAxoN W. D., 1955: Continuous Fermentation ? a Discussion of its Principles and
Applications. Applied Microbiol., 3: 110.
MONOD J., 1950: La technique do culture continuo; theorio ot applications. Annal. Inst.
Pasteur, 79: 390.
NOWREY J. E., FINN R. K., 1955: Division of Agricultural and Food Chemistry, 128th
Meeting, ACS, p. 25A, Minneapolis (cit. Ind. Eng. Chem. 48: 1585 (1956)).
ft Id law J., 1956: Personal communication.
SHAFFER P. A., HARTMANN A. F., 1921: The Iodomotric Determination of Copper and
its Use in Sugar Analysis If. J. Biol. Chem., 45: 365.
- 226,
STAT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5
R
STAT
Next 4 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/03/05: CIA-RDP81-01043R002800080004-5