Cytotechnology 7: 179-186, 1991. 9 1991 KluwerAcademic Publishers.Printedin the Netherlands.
Cell cycle, cell size and mitochondrial activity of hybridoma cells during batch cultivation M. A1-Rubeai 1, S. Chalder 1, R. Bird 2 and A.N. Emery 1 1SERC Centre for Biochemical Engineering, School of Chemical Engineering and 2Department of Immunology, University of Birmingham, PO Box 363, Birmingham, B15 2TT, UK Received 12 July 1991; acceptedin revised form 15 November 1991
Key words: hybridoma cell culture, cell cycle, flow cytometry, cell size, mitochondria. Abstract Cell cycle, cell size and rhodamine 123 fluorescence in cell populations of two batch cultures were analysed and quantified with a fluorescence-activated cell sorter (FACS). Two cultures derived from either exponential or stationary phase innocula were investigated in order to demonstrate the dependency of the subsequent cell growth on innoculum condition. The results demonstrated that the level of activity of ceils in the innoculum culture could have a significant effect on cellular activity during the initial phase of the inoculated culture, as it advances through its growth cycle. Positive correlation was found between the cell size and mitochondrial activity (as measured by rhodamine 123 uptake) with S and G2 fractions as the cell progressed through the cell cycle. The enumeration of the fractions of cell cycle phases has helped in prediction of the changes in cell numbers following perturbation of the culture condition.
Introduction Increase in the cell biomass in culture can be achieved by two ways, growth in the number of cells and growth in cell size. During the growth cycle of batch culture cells are constantly produced by division, and lost from the population by death. However, the net gain or loss in biomass is determined by the physical and chemical environment. The changes in growth rate with time in the culture reflect the ever changing environment during batch cultivation. An optimum environment can lead to increased growth rate by reducing cell cycle duration and cell death and by enhancing biosynthetic activity and cell division. Since cell size doubles prior to
cell division, it follows that the distribution of cell sizes in the population should reflect the progression of cells through the cell cycle. Progression through the cell cycle and the associated cell volume changes provide important insights into the state of the cells during a bioreaction process (A1-Rubeai et al., 1991a). Moreover changes in energy metabolism which are exhibited in changes of mitochondrial structure, activity and biogenesis (Ernster and Schatz, 1981; Benel et al., 1985) represent another important aspect of cell growth and differentiation. In this work ~ve report about the utility of the above three parameters: viz., cell cycle, cell volume and mitochondrial metabolism in analysing and quantifying the cellular changes during batch cultiva-
180 tion through the application of flow cytometric analyses.
on a Data System Design computer using Consort 40 software.
Materials and methods
Results and discussion
TB/C3 murine hybridomas producing antibodies against human IgG were maintained in RPMI 1640 medium (Sigma, Poole, UK) supplemented with 5% (v/v) Newborn Bovine Serum (NBS; Advanced Protein Products Ltd, Brierley Hill, UK). Two batch cultures were set up using innocula from cultures of different age. Cells were grown in magnetically stirred 500 ml flasks. Viability and viable cell number of random samples were counted using trypan blue dye exclusion in a haemacytometer. Cell samples were collected daily during batch cultivation, centrifuged and resuspended at 5 • 105 cells/ml in RPMI 1640 medium containing 0.5 ~tg/ml Rhodamine 123 (R123; Sigma, Poole, UK). R123 is a vital stain specific for mitochondria (Johnson et al., 1980) which allows the estimation of organelle numbers and/or activities in various cell states (Ronot et al., 1986). After an incubation period of 30 min at 37~ the ceils were collected by centrifugation and resuspended in Phosphate Buffer Saline (PBS). 50 ~tg/ml Propidium Iodide (PI; Sigma, Poole, UK) was added 5 min prior to flow cytometric analysis of the ceils in order to stain dead cells and nuclear fragments. Daily cell samples were fixed in 70% ethanol and stored at 4~ and then the fixed cells were washed in PBS and DNA was specifically stained with 50 ].tg/ml PI for 15 min prior to flow cytometric analysis of cells.
The effects of innoculum condition on cell growth In Fig. 1, results for five batch cultures grown under the same physical and chemical conditions are presented. It can be seen that there are clear differences in growth rates and growth extent between these cultures. Such differences in growth patterns have been repeatedly observed and can only be explained as due to the differences which exist in the levels of metabolic activity of cells in the innoculum. It was for this reason that we suggested (Oh et al., 1989) that for the evaluation of the effect of hydrodynamic conditions on hybridoma growth andmonoclonal antibody production a rigorous preparation of the cells should be carried out so that on inoculation they are all in the same condition, otherwise less reproducible results are obtained even between the controls. Based on such findings we decided to further investigate the effect of the innoculum condition on cell growth using cytometric parameters to assess such effects. Figure 2 illustrates the variation in the growth patterns of two cultures of hybridomas where
Batch
Cultures
Inoeulum
Condition
Cell no. (*lOe5/ml)
.......... 4
Flow cytomewic analysis The flow cytometric analysis was carried out using a Becton-Dickinson FACS 440 fitted with an argon ion laser operating at 200 mW light power with excitation at 488 nm. Emission was collected using a 530 nm band pass filter for R123 and 620 nm band pass filter for PI. 20,000 particles were recorded and results were analysed
0
20
40
80
80
100
120
Time (hour) "-~-"
Culture 1
-4-
Culture 8
-e--
Culture 4
~
Culture 5
'*--
Culture 3
Fig. 1. Growth of TB/C3 hybridomas in five suspension batch cultures under the same chemical and physical conditions. Innocula were obtained form different cultures.
181
Growth C u r v e s Cell no, ( * l O e S / m l )
10
8
6
4
2
Culture 1
-}-
0 --
1
2
3
Culture 2
4
Time (day)
Fig. 2. Growth of TB/C3 hybridomas in two suspension batch cultures. Cultures 1 and 2 are derived form stationary and exponential phase innocuhm, respectively.
cells were obtained ~from different phases of the innoculum culture. The culture derived from exponential phase innoculum (culture 2) exhibited a higher extent of growth (in terms of maximum cell density) than the culture derived from stationary phase innoculum (culture 1). The dependency of the subsequent cell growth on innoculum condition supports the concept that the cellular activity of daughter cells is influenced by the level of cellular activity of the parent cells of continuously dividing cell lines. As the increase in number of cells is the most essential property of animal cell culture, the expression of such a property is indeed, as shown here, influenced by the individuality of each cell (i.e., the physiological state of the cell). To investigate further the characteristics of cells in the two different growth pattems, changes in the distribution of cells in the growth cycle (Gb S, G2 and M), and in cell size and metabolic activity were compared. The different fractions of dells in G1, S and G2 + M phases of the cell cycle were determined from the DNA histograms (Ormerod, 1990). The area under the first peak represented the fraction of G1 cells while the area under the second peak represented the fraction of G2 cells. The fraction of cells in S phase was represented by the area between the two peaks.
Growth rates (which are determined from the slope of the linear regression of the log of the cell number vs time) for culture 1 and 2 were 0.013 and 0.022 respectively. The difference in the growth patterns between the two cultures can be unmistakably demonstrated when it is evaluated by the viability index (VI) (Luan et al., 1987). VI is the integral of viable cells over the cultivation period. It measures the overall differences throughout the various growth phases. The VIs for culture 1 and 2 were 210 • 105 and 421 • 105 cells -h/ml respectively. The difference was also manifested in the cell cycle kinetics of the two cultures (Fig. 3). Whereas the innoculum of culture 2 showed a relatively high fraction of S-cells which had yet further increased in day 1, culture 1 innoculum showed a lower fraction of S-cells and a small but a steady increase in day 1. A similar but opposite trend was also observed in the fraction of Gl-cells. In both cultures at day 1 the fraction of Gl-cells showed a decrease, as might be expected due to the enhancement of DNA synthesis in optimal environmental conditions. The rapid decline in cell number of culture 2 can be explained in terms of cell cycle duration. The rapidly growing large (see below) S-G 2 cells which have led to the higher maximum cell number presumably contributed more to nutrient exhaustion and to release of metabolic toxic products than did slower growing small G1 cells. In culture 1 the fewer S-G2 and smaller but more abundant GI cells were able to utilise the nutrients present at the lower levels of possible cytotoxic products to extend the decline phase thereby extending the culture duration. During the decline phase of batch culture the cells do not leave the cycle and become nondividing, by being terminally arrested in G1 phase as is the case with density-inhibited cells when the monolayer is completed. In the latter case the quiescent cells can remain viable for several days. However, hybridoma ceils cannot survive for a long period if they are inhibited from cycling by arrest in G1 (A1-Rubeai and Emery, 1990). Moreover, the cells continue during the decline phase to synthesize D N A at appreciable but decreasing rates, and the cells leave the cycle to die at any point
182 regardless of whether they are in G1 or in the midst of DNA synthesis. It is therefore plausible to say that at any time during batch cultivation the cell population is a mixture of dividing cells and dying cells. The pattern of the growth curve will therefore be the result of the cell cycle time, growth fraction and rate of cell loss (Baserga, 1985). It has been suggested that amino acid deprivation rather than serum limitation is responsible for the reduced rate of cell division during the late growth phase (Ramirez and Mutharasan, 1990). It is clear from the present work that such differences in growth rate, death rate and maximum cell number which are due to the difference in the rate of cell division are in turn predictable from measurement of the fraction of S-cells in the population. This interesting feature of cell cycle kinetics can therefore provide an early indicator of growth long before it can be assessed by cell counting. The fraction of S-cells can help to clarify and reveal more fully the growth potential and the changes in cell number with time during batch growth, thereby allowing sufficient time to manipulate the growth conditions in order to extend the growth cycle.
Cell cycle analysis as a performance predicts The effectiveness and usefulness of this type of analysis in monitoring of bioreaction processes is demonstrated in Fig. 4. A hybridoma batch culture was split into 2 during the exponential phase at a time when the culture showed a 7% drop in the fraction of S-phase cells. 2.5g/1 meat peptone was added to one culture while the other served as a control. The improvement in cell number, which was only demonstrated 21 h later, could actually be predicted by the cell cycle analysis just 5 h after the supplementation of peptone, even though at this time there was no observable difference in cell number between the supplemented and the unsupplemented cultures; however the fraction of ceils in the S-phase of the supplemented culture was clearly 5% greater than the other. At the time the cultures were terminated (90 h) the differences in cell number and in
ceU cycle phase
70
(~)
6O
culture I 5O 4O 30"
2O 10 ~ ,0 0
I
I
I
t
1
2
$
4
Time (days) Cell 70
cycle phue
(Z)
60 cult~ 2
60 40. -..................................................... q.
50
q.
20
0
I
i
I
1
I
2
3
4
Time GI
,
(days)
-+- S
--x~-- G2+M
Fig. 3. Percentagesof cells in the G1, S and G2 + M phases of the cell cycle of TB/C3 hybridomasgrown in two different suspension cultures (as in Fig. 2). the proportion of S-phase cells between the supplemented and unsupplemented cultures had extended to 3.8 x 105/ml and 15% respectively. These data confirm the claims of Kloppinger et al. (1989) and A1-Rubeai et al. (1991a) that cell cycle analysis can provide advanced and reliable information on proliferation dynamics in cell culture.
Cell size chances during the growth cycle Figure 5 demonstrates the changes in mean cell size during the same two batch cultivations 1 and 2 referred to above. Both cultures show similar increases in mean cell size at the start of the exponential phase followed by a gradual decrease toward the end of this phase. However there are no obvious changes in mean cell size during the decline phase. Such changing patterns can be seen to be positively correlated with the fractions
183
Growth Curve & S -Phase 12
Cell no,*lOe5/ml
Cell S i z e
Fraction S -phase
fraction
F o r w a r d Angle L i g h t S c a t t e r 60 50
l02485 1
///+',,,
~z
130
40
120
30
110
20
Forward angle light sea~ter
100
10 0 -0
J 20 Control -q-"Peptone
i i 40 60 Time (hours) suppl.
--)g'- Control
90
i 80 - ~ - Peptone
0 IO0 suppl.
80
r 1
i i 2 3 Time (day)
i 4
Fig. 4. Growthcurve and percentage of cells in S phase of the cell cycle in suspensionbatch culture before and after supplementationwithmeatpeptone.Arrowindicatesthe timeat which peptone is added.
Fig. 5. Meanrelativecell size (forwardlight scatter) of TB/C3 hybridomas grown in two differentsuspension cultures (as in Fig. 2).
of S and G2 + M cells (Fig 6; r = 0.78, p < 0.01). An increase in size is due to the duplication of all the main components of the cells, i.e., DNA, RNA and proteins (Baserga, 1985). However, growth in size is reversible, the cells halving their size at mitosis. Therefore in a batch culture the increase in cell number is preceded by an increase in the fraction of S-phase cells as well as in cell size. The unchanged mean cell size during the decline phase probably resulted from the increase in cell size heterogeneity (e.g., in culture 2, the coefficients of variation are 25% and 32% at day 1 and 4 respectively). It has also been observed by Darzynkiewicz et al. (1982) that cell size is more heterogeneous in G] than in G2 + M cell populations. These authors explained such increases in heterogeneity as a result of the unequal distribution into the daughter cells of cytoplasmic constituents during cytokinesis and the fact that unequal cytokinesis generates intracellular metabolic variability during the cycle. The relationship between growth in batch culture of hybridoma cells as determined by cell count and cytometric parameters such as DNA content, proportion of S-phase cells and cell volume have been repeatedly observed in our laboratory, and is demonstrated in Table 1. As expected the cytometric parameters show similar and con-
sistent trends of change over time. They are found to be related and the relationship appears to be positive and substantially rectilinear (Table 1). The temporal changes in cell cycle phases of the cell population during chemostat and perfusion cultures of hybridoma cells have been effectively and reliably determined using flow cytometry (A1-Rubeai et al., 1991b). Moreover we have successfully used cell cycle analysis for the regulation of medium flow rate in continuous perfusion culture of hybridoma cells. The measurement of cell size may well have the same importance as that of DNA in monitoring growth simply because growth in cellular size is understandably just as critical for cell division as DNA replication.
Chances in mitochondrial activity during the growth cycle As a means of monitoring mitochondrial activity, the cellular uptake of R123 was monitored during the same two batch cultures 1 and 2. Results are shown in Fig. 7. In both cultures the uptake of R123 increased during the early growth phase but then fell throughout the course of late exponential and decline phases. However there was a qualitative difference between the two cultures. The
184 Table 1. Growth Kinetics of hybridoma batch culture.
Time (h)
Cell no. (x 105/ml)
Rdl. mean DNA contenta
Percentage of S + G2/M cells
Re1. mean of cell sizea
1 22 32 46 51 55 70
2.0 6.2 10.3 11.0 7.1 6.1 1.9
93.6 94.8 83.0 66.8 57.6 63.1 63.8
70.2 67.5 57.2 29.0 30.8 22.4 --
80.1 82.3 74.3 64.1 68.5 65.3 67.6
DNA content vs %S + G2/M cells: r = 0.99, p < 0.001 DNA content vs cell size: r = 0.97, p < 0.001 %S + G2/M ceils vs cell size: r = 0.97, p < 0.001 a Flow cytometric analysis was carried out using a Becton-DickinsonFACStar PLUS (see A1-Rubeaiet al. 1991a).
increase in R123 uptake in culture 1 was slow and steady during the initial stages of growth in contrast to that in culture 2 which was initially sharp and striking. Such a difference in metabolic patterns between the two cultures during the lag and exponential phases was reflected in the proliferative capacity o f each culture. Culture 2 with a higher R 123 uptake rate showed higher growth rate and growth extent. Correlation was found between R123 uptake and the progression of cells through the cell cycle (r value for R123 uptake vs fraction o f S, G2 + M cells = 0.6, p < 0.05). During the rapid flow of cells into the S-phase of the cell cycle, which characterized the lag phase of batch culture, there was a surge in metabolic activity as indicated by the increase in R123
uptake. It is conceivable that the increased metabolic activity is a response to a requirement for more D N A and cellular biomass. Subsequently, R123 uptake declined during the death phase in both cultures. The conclusion is that mitochondrial activity is higher for S and G2 cells than G 1 cells. The difference in R123 uptake during the exponential phase of the two cultures is consistent with the above conclusion. Further computerbased analysis of small and large cells subpopulations suggests the existence of a relationship between cell size and R 123 uptake. Thus, the cell cycle dependence of the R123 uptake m a y be
Mitochondrial Activity Rhodamine
123
Rhodamine 123
Fraction of S,G2+M c e l l s
150
75
7O 145 60 55 50
140
45 Culture I
40
-]-
35
30 80
rffi+O,78, P
Cell cycle, cell size and mitochondrial activity of hybridoma cells during batch cultivation M. A1-Rubeai 1, S. Chalder 1, R. Bird 2 and A.N. Emery 1 1SERC Centre for Biochemical Engineering, School of Chemical Engineering and 2Department of Immunology, University of Birmingham, PO Box 363, Birmingham, B15 2TT, UK Received 12 July 1991; acceptedin revised form 15 November 1991
Key words: hybridoma cell culture, cell cycle, flow cytometry, cell size, mitochondria. Abstract Cell cycle, cell size and rhodamine 123 fluorescence in cell populations of two batch cultures were analysed and quantified with a fluorescence-activated cell sorter (FACS). Two cultures derived from either exponential or stationary phase innocula were investigated in order to demonstrate the dependency of the subsequent cell growth on innoculum condition. The results demonstrated that the level of activity of ceils in the innoculum culture could have a significant effect on cellular activity during the initial phase of the inoculated culture, as it advances through its growth cycle. Positive correlation was found between the cell size and mitochondrial activity (as measured by rhodamine 123 uptake) with S and G2 fractions as the cell progressed through the cell cycle. The enumeration of the fractions of cell cycle phases has helped in prediction of the changes in cell numbers following perturbation of the culture condition.
Introduction Increase in the cell biomass in culture can be achieved by two ways, growth in the number of cells and growth in cell size. During the growth cycle of batch culture cells are constantly produced by division, and lost from the population by death. However, the net gain or loss in biomass is determined by the physical and chemical environment. The changes in growth rate with time in the culture reflect the ever changing environment during batch cultivation. An optimum environment can lead to increased growth rate by reducing cell cycle duration and cell death and by enhancing biosynthetic activity and cell division. Since cell size doubles prior to
cell division, it follows that the distribution of cell sizes in the population should reflect the progression of cells through the cell cycle. Progression through the cell cycle and the associated cell volume changes provide important insights into the state of the cells during a bioreaction process (A1-Rubeai et al., 1991a). Moreover changes in energy metabolism which are exhibited in changes of mitochondrial structure, activity and biogenesis (Ernster and Schatz, 1981; Benel et al., 1985) represent another important aspect of cell growth and differentiation. In this work ~ve report about the utility of the above three parameters: viz., cell cycle, cell volume and mitochondrial metabolism in analysing and quantifying the cellular changes during batch cultiva-
180 tion through the application of flow cytometric analyses.
on a Data System Design computer using Consort 40 software.
Materials and methods
Results and discussion
TB/C3 murine hybridomas producing antibodies against human IgG were maintained in RPMI 1640 medium (Sigma, Poole, UK) supplemented with 5% (v/v) Newborn Bovine Serum (NBS; Advanced Protein Products Ltd, Brierley Hill, UK). Two batch cultures were set up using innocula from cultures of different age. Cells were grown in magnetically stirred 500 ml flasks. Viability and viable cell number of random samples were counted using trypan blue dye exclusion in a haemacytometer. Cell samples were collected daily during batch cultivation, centrifuged and resuspended at 5 • 105 cells/ml in RPMI 1640 medium containing 0.5 ~tg/ml Rhodamine 123 (R123; Sigma, Poole, UK). R123 is a vital stain specific for mitochondria (Johnson et al., 1980) which allows the estimation of organelle numbers and/or activities in various cell states (Ronot et al., 1986). After an incubation period of 30 min at 37~ the ceils were collected by centrifugation and resuspended in Phosphate Buffer Saline (PBS). 50 ~tg/ml Propidium Iodide (PI; Sigma, Poole, UK) was added 5 min prior to flow cytometric analysis of the ceils in order to stain dead cells and nuclear fragments. Daily cell samples were fixed in 70% ethanol and stored at 4~ and then the fixed cells were washed in PBS and DNA was specifically stained with 50 ].tg/ml PI for 15 min prior to flow cytometric analysis of cells.
The effects of innoculum condition on cell growth In Fig. 1, results for five batch cultures grown under the same physical and chemical conditions are presented. It can be seen that there are clear differences in growth rates and growth extent between these cultures. Such differences in growth patterns have been repeatedly observed and can only be explained as due to the differences which exist in the levels of metabolic activity of cells in the innoculum. It was for this reason that we suggested (Oh et al., 1989) that for the evaluation of the effect of hydrodynamic conditions on hybridoma growth andmonoclonal antibody production a rigorous preparation of the cells should be carried out so that on inoculation they are all in the same condition, otherwise less reproducible results are obtained even between the controls. Based on such findings we decided to further investigate the effect of the innoculum condition on cell growth using cytometric parameters to assess such effects. Figure 2 illustrates the variation in the growth patterns of two cultures of hybridomas where
Batch
Cultures
Inoeulum
Condition
Cell no. (*lOe5/ml)
.......... 4
Flow cytomewic analysis The flow cytometric analysis was carried out using a Becton-Dickinson FACS 440 fitted with an argon ion laser operating at 200 mW light power with excitation at 488 nm. Emission was collected using a 530 nm band pass filter for R123 and 620 nm band pass filter for PI. 20,000 particles were recorded and results were analysed
0
20
40
80
80
100
120
Time (hour) "-~-"
Culture 1
-4-
Culture 8
-e--
Culture 4
~
Culture 5
'*--
Culture 3
Fig. 1. Growth of TB/C3 hybridomas in five suspension batch cultures under the same chemical and physical conditions. Innocula were obtained form different cultures.
181
Growth C u r v e s Cell no, ( * l O e S / m l )
10
8
6
4
2
Culture 1
-}-
0 --
1
2
3
Culture 2
4
Time (day)
Fig. 2. Growth of TB/C3 hybridomas in two suspension batch cultures. Cultures 1 and 2 are derived form stationary and exponential phase innocuhm, respectively.
cells were obtained ~from different phases of the innoculum culture. The culture derived from exponential phase innoculum (culture 2) exhibited a higher extent of growth (in terms of maximum cell density) than the culture derived from stationary phase innoculum (culture 1). The dependency of the subsequent cell growth on innoculum condition supports the concept that the cellular activity of daughter cells is influenced by the level of cellular activity of the parent cells of continuously dividing cell lines. As the increase in number of cells is the most essential property of animal cell culture, the expression of such a property is indeed, as shown here, influenced by the individuality of each cell (i.e., the physiological state of the cell). To investigate further the characteristics of cells in the two different growth pattems, changes in the distribution of cells in the growth cycle (Gb S, G2 and M), and in cell size and metabolic activity were compared. The different fractions of dells in G1, S and G2 + M phases of the cell cycle were determined from the DNA histograms (Ormerod, 1990). The area under the first peak represented the fraction of G1 cells while the area under the second peak represented the fraction of G2 cells. The fraction of cells in S phase was represented by the area between the two peaks.
Growth rates (which are determined from the slope of the linear regression of the log of the cell number vs time) for culture 1 and 2 were 0.013 and 0.022 respectively. The difference in the growth patterns between the two cultures can be unmistakably demonstrated when it is evaluated by the viability index (VI) (Luan et al., 1987). VI is the integral of viable cells over the cultivation period. It measures the overall differences throughout the various growth phases. The VIs for culture 1 and 2 were 210 • 105 and 421 • 105 cells -h/ml respectively. The difference was also manifested in the cell cycle kinetics of the two cultures (Fig. 3). Whereas the innoculum of culture 2 showed a relatively high fraction of S-cells which had yet further increased in day 1, culture 1 innoculum showed a lower fraction of S-cells and a small but a steady increase in day 1. A similar but opposite trend was also observed in the fraction of Gl-cells. In both cultures at day 1 the fraction of Gl-cells showed a decrease, as might be expected due to the enhancement of DNA synthesis in optimal environmental conditions. The rapid decline in cell number of culture 2 can be explained in terms of cell cycle duration. The rapidly growing large (see below) S-G 2 cells which have led to the higher maximum cell number presumably contributed more to nutrient exhaustion and to release of metabolic toxic products than did slower growing small G1 cells. In culture 1 the fewer S-G2 and smaller but more abundant GI cells were able to utilise the nutrients present at the lower levels of possible cytotoxic products to extend the decline phase thereby extending the culture duration. During the decline phase of batch culture the cells do not leave the cycle and become nondividing, by being terminally arrested in G1 phase as is the case with density-inhibited cells when the monolayer is completed. In the latter case the quiescent cells can remain viable for several days. However, hybridoma ceils cannot survive for a long period if they are inhibited from cycling by arrest in G1 (A1-Rubeai and Emery, 1990). Moreover, the cells continue during the decline phase to synthesize D N A at appreciable but decreasing rates, and the cells leave the cycle to die at any point
182 regardless of whether they are in G1 or in the midst of DNA synthesis. It is therefore plausible to say that at any time during batch cultivation the cell population is a mixture of dividing cells and dying cells. The pattern of the growth curve will therefore be the result of the cell cycle time, growth fraction and rate of cell loss (Baserga, 1985). It has been suggested that amino acid deprivation rather than serum limitation is responsible for the reduced rate of cell division during the late growth phase (Ramirez and Mutharasan, 1990). It is clear from the present work that such differences in growth rate, death rate and maximum cell number which are due to the difference in the rate of cell division are in turn predictable from measurement of the fraction of S-cells in the population. This interesting feature of cell cycle kinetics can therefore provide an early indicator of growth long before it can be assessed by cell counting. The fraction of S-cells can help to clarify and reveal more fully the growth potential and the changes in cell number with time during batch growth, thereby allowing sufficient time to manipulate the growth conditions in order to extend the growth cycle.
Cell cycle analysis as a performance predicts The effectiveness and usefulness of this type of analysis in monitoring of bioreaction processes is demonstrated in Fig. 4. A hybridoma batch culture was split into 2 during the exponential phase at a time when the culture showed a 7% drop in the fraction of S-phase cells. 2.5g/1 meat peptone was added to one culture while the other served as a control. The improvement in cell number, which was only demonstrated 21 h later, could actually be predicted by the cell cycle analysis just 5 h after the supplementation of peptone, even though at this time there was no observable difference in cell number between the supplemented and the unsupplemented cultures; however the fraction of ceils in the S-phase of the supplemented culture was clearly 5% greater than the other. At the time the cultures were terminated (90 h) the differences in cell number and in
ceU cycle phase
70
(~)
6O
culture I 5O 4O 30"
2O 10 ~ ,0 0
I
I
I
t
1
2
$
4
Time (days) Cell 70
cycle phue
(Z)
60 cult~ 2
60 40. -..................................................... q.
50
q.
20
0
I
i
I
1
I
2
3
4
Time GI
,
(days)
-+- S
--x~-- G2+M
Fig. 3. Percentagesof cells in the G1, S and G2 + M phases of the cell cycle of TB/C3 hybridomasgrown in two different suspension cultures (as in Fig. 2). the proportion of S-phase cells between the supplemented and unsupplemented cultures had extended to 3.8 x 105/ml and 15% respectively. These data confirm the claims of Kloppinger et al. (1989) and A1-Rubeai et al. (1991a) that cell cycle analysis can provide advanced and reliable information on proliferation dynamics in cell culture.
Cell size chances during the growth cycle Figure 5 demonstrates the changes in mean cell size during the same two batch cultivations 1 and 2 referred to above. Both cultures show similar increases in mean cell size at the start of the exponential phase followed by a gradual decrease toward the end of this phase. However there are no obvious changes in mean cell size during the decline phase. Such changing patterns can be seen to be positively correlated with the fractions
183
Growth Curve & S -Phase 12
Cell no,*lOe5/ml
Cell S i z e
Fraction S -phase
fraction
F o r w a r d Angle L i g h t S c a t t e r 60 50
l02485 1
///+',,,
~z
130
40
120
30
110
20
Forward angle light sea~ter
100
10 0 -0
J 20 Control -q-"Peptone
i i 40 60 Time (hours) suppl.
--)g'- Control
90
i 80 - ~ - Peptone
0 IO0 suppl.
80
r 1
i i 2 3 Time (day)
i 4
Fig. 4. Growthcurve and percentage of cells in S phase of the cell cycle in suspensionbatch culture before and after supplementationwithmeatpeptone.Arrowindicatesthe timeat which peptone is added.
Fig. 5. Meanrelativecell size (forwardlight scatter) of TB/C3 hybridomas grown in two differentsuspension cultures (as in Fig. 2).
of S and G2 + M cells (Fig 6; r = 0.78, p < 0.01). An increase in size is due to the duplication of all the main components of the cells, i.e., DNA, RNA and proteins (Baserga, 1985). However, growth in size is reversible, the cells halving their size at mitosis. Therefore in a batch culture the increase in cell number is preceded by an increase in the fraction of S-phase cells as well as in cell size. The unchanged mean cell size during the decline phase probably resulted from the increase in cell size heterogeneity (e.g., in culture 2, the coefficients of variation are 25% and 32% at day 1 and 4 respectively). It has also been observed by Darzynkiewicz et al. (1982) that cell size is more heterogeneous in G] than in G2 + M cell populations. These authors explained such increases in heterogeneity as a result of the unequal distribution into the daughter cells of cytoplasmic constituents during cytokinesis and the fact that unequal cytokinesis generates intracellular metabolic variability during the cycle. The relationship between growth in batch culture of hybridoma cells as determined by cell count and cytometric parameters such as DNA content, proportion of S-phase cells and cell volume have been repeatedly observed in our laboratory, and is demonstrated in Table 1. As expected the cytometric parameters show similar and con-
sistent trends of change over time. They are found to be related and the relationship appears to be positive and substantially rectilinear (Table 1). The temporal changes in cell cycle phases of the cell population during chemostat and perfusion cultures of hybridoma cells have been effectively and reliably determined using flow cytometry (A1-Rubeai et al., 1991b). Moreover we have successfully used cell cycle analysis for the regulation of medium flow rate in continuous perfusion culture of hybridoma cells. The measurement of cell size may well have the same importance as that of DNA in monitoring growth simply because growth in cellular size is understandably just as critical for cell division as DNA replication.
Chances in mitochondrial activity during the growth cycle As a means of monitoring mitochondrial activity, the cellular uptake of R123 was monitored during the same two batch cultures 1 and 2. Results are shown in Fig. 7. In both cultures the uptake of R123 increased during the early growth phase but then fell throughout the course of late exponential and decline phases. However there was a qualitative difference between the two cultures. The
184 Table 1. Growth Kinetics of hybridoma batch culture.
Time (h)
Cell no. (x 105/ml)
Rdl. mean DNA contenta
Percentage of S + G2/M cells
Re1. mean of cell sizea
1 22 32 46 51 55 70
2.0 6.2 10.3 11.0 7.1 6.1 1.9
93.6 94.8 83.0 66.8 57.6 63.1 63.8
70.2 67.5 57.2 29.0 30.8 22.4 --
80.1 82.3 74.3 64.1 68.5 65.3 67.6
DNA content vs %S + G2/M cells: r = 0.99, p < 0.001 DNA content vs cell size: r = 0.97, p < 0.001 %S + G2/M ceils vs cell size: r = 0.97, p < 0.001 a Flow cytometric analysis was carried out using a Becton-DickinsonFACStar PLUS (see A1-Rubeaiet al. 1991a).
increase in R123 uptake in culture 1 was slow and steady during the initial stages of growth in contrast to that in culture 2 which was initially sharp and striking. Such a difference in metabolic patterns between the two cultures during the lag and exponential phases was reflected in the proliferative capacity o f each culture. Culture 2 with a higher R 123 uptake rate showed higher growth rate and growth extent. Correlation was found between R123 uptake and the progression of cells through the cell cycle (r value for R123 uptake vs fraction o f S, G2 + M cells = 0.6, p < 0.05). During the rapid flow of cells into the S-phase of the cell cycle, which characterized the lag phase of batch culture, there was a surge in metabolic activity as indicated by the increase in R123
uptake. It is conceivable that the increased metabolic activity is a response to a requirement for more D N A and cellular biomass. Subsequently, R123 uptake declined during the death phase in both cultures. The conclusion is that mitochondrial activity is higher for S and G2 cells than G 1 cells. The difference in R123 uptake during the exponential phase of the two cultures is consistent with the above conclusion. Further computerbased analysis of small and large cells subpopulations suggests the existence of a relationship between cell size and R 123 uptake. Thus, the cell cycle dependence of the R123 uptake m a y be
Mitochondrial Activity Rhodamine
123
Rhodamine 123
Fraction of S,G2+M c e l l s
150
75
7O 145 60 55 50
140
45 Culture I
40
-]-
35
30 80
rffi+O,78, P
