Possible Limitation of Growth in Human Fibroblast Cultures by Diffusion JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES Depurtnzeiits o f Medicine and Biochemistry, Queen's University, Kingston, Ontario, C a n a d a
ABSTRACT Secondary cultures of human diploid fibroblasts which demonstrated density dependent inhibition of cell growth (001) were used to study the possible limitation of growth in cell cultures by diffusion. An oscillating platform system is described which insures constant mixing of the medium during the culturing period. Using this system, i t was found that a greater number of cells in density inhibited cultures, grown to confluence for four days after initial seeding, could be stimulated to resume growth by a fresh medium change if the cultures were incubated on the oscillating platform than if the cultures were left undisturbed. This greater stimulation on the platform was probably not due to mechanical alterations on the surface of the cells due to motion of the medium as judged by TCA preripitable material released into the medium from cells prelabeled with glu~osamine-~H. In spite of this greater stimulation after a single treatment with fresh medium, refeeding the cells on the platform every other day over a 12-day period did not affect the final saturation density achieved in the cultures. The results indicate that diffusion limitation of growth might occur under certain circumstances but that it cannot account entirely for the phenomenon of DDI.
IJsing a pump mechanism, Stoker ( ' 7 3 ) demonstrated that the stimulation of density inhibited 3T3 cells by fresh medium could be greatly increased by causing the medium to be in continuous motion over the cultures. This suggested that the phenomenon of density dependent inhibition of growth (DDI), normally seen in fibrcblast cultures without medium disturbance, might be explained strictly on the basis of either the limitation by diffusion of the availability of essential growth nutrients or by the inability of growth inhibitors produced by the cells to diffuse away fast enough. I n its strict form Stoker's hypothesis does away with the notion of cell to cell interactions as being necessary for DDI. The results of the experiments presented here are consistent with the notion that the formation of diffusion gradients may play a role in determining the rate at which cells attain their first saturation density but not in determining the magnitude of that saturation density. The implications of the results are discussed i n terms of how growth limitation by a diffuJ. CELL PHYSIOL, 86: 567 580.
sion process could affect interpretations of experiments i n this system. MATERIALS A N D METHODS
Cell cultures The strain of human fibroblasts used in this study was obtained a s a secondary culture from Montreal Children's Hospital, being originally grown from human foreskin. The cells were grown in Eagle's Minimal Essential Medium (Eagle, '59) containing either 5 or 10% fetal calf serum (Grand Island Biological Co., Grand Island, N.Y.) in 35 to 60 mm Falcon Tissue Culture Petri dishes (Falcon Plastics, Division B-D Laboratories Inc. Los Angeles, Cd.) in a CO, incubator (2.5% COz, 97.5% air) at 37°C. The pH of the medium was maintained at pH 7.7, as described previously (Froehlich and Anastassiades, '74), by making the medium 28 mM in NaHC03. Cultures grown in the 35 and 60 mm dishes were fed with 2.0 ml and 4.0 ml of medium, respectively. Attempts at culturing mycoplasma from the cells used here __ -
Received Oct. 22, '74. Accepted Apr. 18, '75.
567
568
JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
were negative under conditions in which mycoplasma could be cultured from cell strains known to contain these organisms. Cell counts were performed on duplicate or triplicate cultures after removal of the cells from the dishes with 1-1.5 ml of 0.25% trypsin (Grand Island Biological Co., Grand Island, N.Y.) dissolved in phosphate buffered saline (PBS). After trypsinization, a n equal volume of serum was added to stop trypsin action and three or four cell counts were performed on each culture in a Neubauer hemacytometer. The three to four cell counts on the cells trypsinized from a single culture never varied by more than 5% from each other and each cell count involved counting 150-200 cells.
found after labeling with glucosamine reside in the glycoprotein fractions of the cell (Onodera and Sheinin, '70; Buck et al., '70; Baig and Roberts, '73) with as much as 80% of the label capable of being released from the cells by trypsinization (Onodera and Sheinin, ' 7 0 ) , indicating that many of the labeled glycoproteins are in the surface membrane. After the cells were prelabeled, the total cell TCA precipitable cpm were determined i n a manner similar to that described above for TdR-3H incorporation except that the final TCA concentration was 7 % . The total TCA precipitable cpm that could be released by trypsin was determined by first washing the cells three times in prewarmed PBS (37"C), trypsinizing the cultures for 10-15 minutes in 0.05% trypsin dissolved Incorporation studies in PBS, centrifuging out the cells at 500 g for ten minutes and precipitating the maUptake of thymidine ( m e t h ~ l - ~ H(TdR) 'H) into acid precipitable material was terial in the supernatant with TCA at a used to measure the rate of DNA syn- final concentration of 7 % . Finally, the thesis in cell cultures. Cultures were in- rate of release of TCA precipitable cpm cubated for either 30 minutes or four into the medium by prelabeled cells was hours in 2.0 ml of culture medium con- determined by washing the prelabeled taining 2.0 $2 of TdR-3H (New England cells with prewarmed PBS, refeeding them Nuclear Corp. (Boston, Mass. ) ; sp. activ- with the appropriate medium without label tiy 20 Ci/mMole). After this time the cul- and then collecting that medium at varitures were washed twice with cold (OOC) ous times after refeeding and precipitating 0.15 M NaC1, dissolved in 1.0 ml of 1.0 with TCA, final concentration 7 % . I n all NaOH, neutralized with 1.0 ml of 1.0 N the above cases, preparation of samples for HC1, and precipitated with trichloracetic scintillation counting was similar to that acid (TCA) at a final concentration of for TdR incorporation except that all pre5 % . The precipitate was washed three cipitations with glucosamine labeling were times with 5% TCA and the final precipi- allowed to proceed for at least four to five tate was dissolved in 2.0 ml of 0.05 N hours. NaOH and 0.1 ml was counted in Aquasol As mentioned i n the RESULTS, the total in a Nuclear Chicago Liquid Scintillation trypsin releasable cpm precipitable by TCA Spectrometer at a n efficiency of 15% for represented about 50% of the total cell 'H. The rate of DNA synthesis was ex- cpm. In the case of release of TCA precippressed as total TCA precipitable cpm/106 itable material into the medium, centrifugcells, where the cell counts were deter- ing the medium at 500 g for ten minutes mined in separate, duplicate cultures. did not alter the radioactivity recovered in Incorporation of D-glu~osamine-G-~H the medium, indicating that the rate of (gluc~samine-~H; New England Nuclear; cell detachment into the medium was sp. activity 2.4 Ci/mMole) into TCA pre- negligible . cipitable material was used to estimate Determination of mitoses the synthesis and release of cell associated glycoproteins into the medium. Cultures The number of mitoses occurring durwere prelabeled during exponential growth ing the interval 24-48 hours after stimufor a 4-day period prior to experimental lation of density inhibited cultures was use in fresh medium containing 1 & determined by adding colcemid (gift of gluc~samine-~H/ml. Several studies have Dr. R. Mankovitz) to the cultures, final shown that the TCA precipitable counts concentration 0.06 pg/ml (Dulbecco, ' 7 0 ) ,
569
DIFFUSION AND CELL GROWTH
at 24 hours and then allowing the cultures to incubate for another 24 hours. The mitoses occurring during this 24hour colcemid mitotic block, were then enumerated either directly under phase microscopy by counting the average number of rounded mitotic cells per high power field and expressing this as a percentage of the average number of cells per field or by regular microscopy after first staining the cultures with orcein, made up in 50% glacial acetic acid, in order to stain the condensed chromatin of the mitotic cells. The two procedures gave essentially the same results with the orcein procedure sometimes producing lower values for percent mitoses because the process of staining tended to remove some of the rounded and therefore tenuously attached mitotic cells from the dishes. In all cases, at least 200 mitotic cells were counted in determining the fraction of cells blocked in mitosis. Increasing the length of the mitotic block beyond 24 hours did not further increase the percentage of cells blocked in mitosis, indicating that this amount of time was sufficient to trap all the cells stimulated to progress into metaphase. Similarly, beginning the colcemid block prior to 24 hours after stimulation of the cultures did not increase the percent of mitoses,
Procedure for continuous mixing of m e d i u m The apparatus used to insure mixing of the medium is a n oscillating platform similar to the Bellco model-6000 rockcr platform (Bellco, Vineland, N.J.) except that the apparatus was designed so that the angle of oscillation could be varied. As pictured in figure 1, this platform on which the cultures were incubated was allowed to oscillate between 8 = 12" and 8 = - 12", as measured in the plane of oslcillation, driven at a frequency 15-20/ min by a variable speed electric motor. AS discussed in the RESULTS, this range of frequencies of oscillation did not mechanically dislodge cells from the culture plates and was chosen because it gave the most reproducible results over a large number of experiments. Furthermore, the rate of oscillation could not be increased over 30-35/min without having some of the
+
4 ml of medium start to slosh aver the top of the 60 m m dishes. In this frequency range of oscillation (15-20/min), 4 ml of medium in 60 m m petri dishes alternately flows back and forth with very little turbulence. When 8 = & 12" about one-sixth of the culture is not covered with medium for about onethird of a second, except for a thin layer of medium which does not flow away from the cells with the bulk af the medium. The maximum velocity attained by the medium as it passes over the cells in the center of the dish at 6 = 0" is 20-25 mm/sec. This velocity is within the range reported by Stoker ('73) for the velocity of the medium attained with his Riddle pump. I n spite of some medium turbulence at the outside rim of the plates which occurs during the oscillation of the platform, the resulting mitoses were found to be uniformly distributed throughout the dishes. I n all experiments involving comparison of cultures incubated on or off the platform, the time of incubation on the platform was never more than 20 hours. In experiments which carried on beyond 20 hours, cultures on the platform were simply taken off it and were left to incubate without agitation for the remainder of the experiment. Furthermore, all radiolabeling was done off the platform, regardless of the previous history of the cultures to be labeled. RESULTS
When normal fibroblasts become confluent in culture, a reversible GI arrest of growth occurs (Nilausen and Green '65) which can be overcome by refeeding the cells with fresh medium containing fresh serum. When such refeeding is done, thymidine incorporation experiments demonstrate a lag period af about ten hours preceding a partially synchronous wave of DNA synthesis in the interval 10-35 hours after treatment (Froehlich and Rachmeler, '72, '74; Todaro et al., '65). This, i n turn, is followed by a wave of mitoses occurring in the interval 25-40 hours after feeding (Todaro et al., '65; Wiebel and Baserga, '69). I n the first experiment described here, human fibroblast cultures were seeded at 1 X lo4 cells,/cmz in EMEM containing
570
JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
Fig. 1 Diagram oscillating platform described -htext. 9, the angle the platform makes with the horizontal, was allowcd to vary between 2 12" with a frequency of 15-20/min. In the figure,
e = + 12", and the level of the medium at that angle in a petri dish is shown.
10% serum and were allowed to grow for four days until growth stopped at a confiuent density of 6 X 104/cm2.At this time, the cultures were stimulated with fresh medium containing 5% serum. The medium removed from the cultures after the four days growth prior to stimulation was still able to support nonconfluent cell growth, though at only two-thirds the rate of fresh medium (data not shown). The effect of continuous medium agitation on the DNA synthetic pattern of these confluent cultures released from growth arrest by refeeding with medium containing 5% serum is shown in figure 2. This experiment demonstrates that the wave of increased TdR-3H incorporation into TCA precipitable material, during 30 minutes pulses, occurred during approximately the same interval whether or not the cells were incubated on the platform but that the amount of incorporation in the cultures stimulated with 5% serum on the platform was three times as great a s the incorporation in the similarly stimulated cultures not on the platform. Since the peak of synthesis occurred at the same time whether the cells were on or off the platform, several experiments identical to that in figure 1 were performed in which the amount of incorporation was measured with a continuous 4-hour pulse during the interval 20-24 hours after stimulation, with all cultures being pulsed off the platform. The results in table 1
indicate that the incorporation during this period ranged from 2.5-4 times greater for the cells preLiously incubated on the platform. Cell counts performed 20 hou-s after stimulation demonstrated no loss of cells in the cultures during incubation on the platform; thus the increased synthesis was not due to medium movement causing cell detachment thereby relieving adjoining cells from the growth constraining effects of confluence. As mentioned in legend to table 1, incubation of control unstimulated cultures on the platform had very little effect on DNA synthesis. The increased incorporation of TdR-T-1 synthesis observed in figure 2 and table 1 could have been due to either increased DNA synthesis or to increased uptake of TdR-JH into the soluble precursor pools. The following considerations, however, argue against merely increased uptake. If the cultures stimulated off the platform were put onto the platform for six to eight hours just prior to the actual DNA synthesis measurements as in table 1, no further increase in synthesis was observed over that observed in cultures stimulated entirely off the platform. However, the experiment shown in figure 3 indicates that the duration of incubation on the platform immediately after refeeding needed to increase TdR-W incorporation in the interval 20-24 hours after stimulation was only two to four hours. Thus, if changes in TdR-3H incorporation into soluble precursor pools occurred after only two to four hours initial incubation on the platform, they should also have occurred when the cells were incubated on the platform during the 6 to 8-hour period just prior to measuring DNA synthesis and an increased incorporation should have resulted. In spite of the above arguments, the number of cells entering mitosis after stimulation was measured to insure that the increased TdR-'H incorporation actually represented increased cellular proliferation. Cultures were grown to confluence as before and stimulated on or off the platform. Colcemid was added to the cultures at 24 hours and the percentage of mitotic cells accumulating over the next 24 hours was determined as described in MATERIALS AND METHODS. The results, reported in
57 1
DIFFUSION AND CELL GROWTH
I0
5
0
I
10
I
1
i
20
30
40
HOURS Fig. 2 Effect of medium mixing on DNA synthesis in stimulated confluent cultures. Cultures were seeded at l x 104/cm2in 10% serum and allowed to grow for four days until growth stopped at 6 x 10'/cm2. At this point, corresponding to t = 0 in the figure, one group of cultures was stimulated with EMEM containing 5% serum, while another control group remained unstimulated. DNA synthesis, measured by the incorporation of TdRJH into TCA precipitable material during a 30-minute pulse, was determined at the times indicated in the figure on duplicate cultures i n both the stimulated and control cultures. In the stimulated group, some of the cultures were incubated on the oscillating platform described in figure 1 for periods up to a maximum of 20 hours (0-0). In these cultures incubated on the platform, points in the figure prior to 20 hours were derived from cultures which were removed from the platform at the times indicated and immediately pulsed. Points after 20 hours represent cultures removed at 20 hours and pulsed at the points indicated and all contUP to 40 hours. The remaining cultures in the stimulated group (0-0) trol cultures (.----a) were incubated off the platform for the entire experiment.
table 2, demonstrate that the number of mitoses occurring in this interval was 2.34 times a s great in the cdtures incubated for 20 hours on the platform. Thus, the increased TdR-3H incorporation in these studies did indicate greater proliferative activity. The studies up to this point essentially confirm the findings of Stoker ('73), who
was able to greatly increase the percentage of confluent 3T3 cells to resume growth by using a pump mechanism to keep the medium flowing over the cells, and who suggested that diffusion limitation may play a role in the inhibition of growth in confluent cultures of normal fibroblasts. An alternative explanation for our re-
57 2
.JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
4.
3
2
I
I
I
5
0
10
I
1
15
20
HOURS ON PLATFORM Fig. 3 Time of incubation 0x1 platform required for increased DNA synthesis in stimulated confluent cultures. Following the same protocol as in figure 2, cultures were stimulated at t = 0 with EMEM containing 5 % serum and were then allowed to incubate o n the oscillating platform for various periods of time up to 20 hours before being removed from the platform. Regardless of the time at which the cultures were removed from the platform, they were continuously incubated with TdRJH during the interval 20-24 hours after initial stimulation. The DNA synthesis determinations obtained on duplicate cultures are plotted in the figurc I S a function of the time of the cultures were allowed to incubate initially on the platform. The closed triangle (A) at t = 0 represents the synthesis in control unstimulated cultures. TABLE 1 Experiment no.
1 2 3
4
D N A synthesis (cpm/l06 cells) Controls
A
B
Ratio(B/A)
180 300 250 340
1100 1950 1500 950
2800 8100 4250 3350
2.5 4.1 2.8 3.5
D N A synthesis in stimulated confluent cultures. Following the protocol in the legend to figure 2, D N A synthesis was determined a s described in MATERIALS AND METHODS, by a 4-hour pulse with TdR-3H 20-24 hours after initial stimulation with EMEM containing 5 % serum in stimulated cultures either on (Column B ) or off (Column A ) the platform. Incorporation of TdR-3H in control unstimulated cultures is also given. Each figure in the table represents the average of two or three cultures. Control cultures incubated on the platform without medium change gave essentially the same results a s those off the platform (control column).
573
DIFFUSION AND CELL GROWTH
~-
TABLE 2
~ - _ _ _ _ _ _ ~
Experiment no
-_________~
_
_
~
~
~
~~~~
Controls
1
-
ABSTRACT Secondary cultures of human diploid fibroblasts which demonstrated density dependent inhibition of cell growth (001) were used to study the possible limitation of growth in cell cultures by diffusion. An oscillating platform system is described which insures constant mixing of the medium during the culturing period. Using this system, i t was found that a greater number of cells in density inhibited cultures, grown to confluence for four days after initial seeding, could be stimulated to resume growth by a fresh medium change if the cultures were incubated on the oscillating platform than if the cultures were left undisturbed. This greater stimulation on the platform was probably not due to mechanical alterations on the surface of the cells due to motion of the medium as judged by TCA preripitable material released into the medium from cells prelabeled with glu~osamine-~H. In spite of this greater stimulation after a single treatment with fresh medium, refeeding the cells on the platform every other day over a 12-day period did not affect the final saturation density achieved in the cultures. The results indicate that diffusion limitation of growth might occur under certain circumstances but that it cannot account entirely for the phenomenon of DDI.
IJsing a pump mechanism, Stoker ( ' 7 3 ) demonstrated that the stimulation of density inhibited 3T3 cells by fresh medium could be greatly increased by causing the medium to be in continuous motion over the cultures. This suggested that the phenomenon of density dependent inhibition of growth (DDI), normally seen in fibrcblast cultures without medium disturbance, might be explained strictly on the basis of either the limitation by diffusion of the availability of essential growth nutrients or by the inability of growth inhibitors produced by the cells to diffuse away fast enough. I n its strict form Stoker's hypothesis does away with the notion of cell to cell interactions as being necessary for DDI. The results of the experiments presented here are consistent with the notion that the formation of diffusion gradients may play a role in determining the rate at which cells attain their first saturation density but not in determining the magnitude of that saturation density. The implications of the results are discussed i n terms of how growth limitation by a diffuJ. CELL PHYSIOL, 86: 567 580.
sion process could affect interpretations of experiments i n this system. MATERIALS A N D METHODS
Cell cultures The strain of human fibroblasts used in this study was obtained a s a secondary culture from Montreal Children's Hospital, being originally grown from human foreskin. The cells were grown in Eagle's Minimal Essential Medium (Eagle, '59) containing either 5 or 10% fetal calf serum (Grand Island Biological Co., Grand Island, N.Y.) in 35 to 60 mm Falcon Tissue Culture Petri dishes (Falcon Plastics, Division B-D Laboratories Inc. Los Angeles, Cd.) in a CO, incubator (2.5% COz, 97.5% air) at 37°C. The pH of the medium was maintained at pH 7.7, as described previously (Froehlich and Anastassiades, '74), by making the medium 28 mM in NaHC03. Cultures grown in the 35 and 60 mm dishes were fed with 2.0 ml and 4.0 ml of medium, respectively. Attempts at culturing mycoplasma from the cells used here __ -
Received Oct. 22, '74. Accepted Apr. 18, '75.
567
568
JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
were negative under conditions in which mycoplasma could be cultured from cell strains known to contain these organisms. Cell counts were performed on duplicate or triplicate cultures after removal of the cells from the dishes with 1-1.5 ml of 0.25% trypsin (Grand Island Biological Co., Grand Island, N.Y.) dissolved in phosphate buffered saline (PBS). After trypsinization, a n equal volume of serum was added to stop trypsin action and three or four cell counts were performed on each culture in a Neubauer hemacytometer. The three to four cell counts on the cells trypsinized from a single culture never varied by more than 5% from each other and each cell count involved counting 150-200 cells.
found after labeling with glucosamine reside in the glycoprotein fractions of the cell (Onodera and Sheinin, '70; Buck et al., '70; Baig and Roberts, '73) with as much as 80% of the label capable of being released from the cells by trypsinization (Onodera and Sheinin, ' 7 0 ) , indicating that many of the labeled glycoproteins are in the surface membrane. After the cells were prelabeled, the total cell TCA precipitable cpm were determined i n a manner similar to that described above for TdR-3H incorporation except that the final TCA concentration was 7 % . The total TCA precipitable cpm that could be released by trypsin was determined by first washing the cells three times in prewarmed PBS (37"C), trypsinizing the cultures for 10-15 minutes in 0.05% trypsin dissolved Incorporation studies in PBS, centrifuging out the cells at 500 g for ten minutes and precipitating the maUptake of thymidine ( m e t h ~ l - ~ H(TdR) 'H) into acid precipitable material was terial in the supernatant with TCA at a used to measure the rate of DNA syn- final concentration of 7 % . Finally, the thesis in cell cultures. Cultures were in- rate of release of TCA precipitable cpm cubated for either 30 minutes or four into the medium by prelabeled cells was hours in 2.0 ml of culture medium con- determined by washing the prelabeled taining 2.0 $2 of TdR-3H (New England cells with prewarmed PBS, refeeding them Nuclear Corp. (Boston, Mass. ) ; sp. activ- with the appropriate medium without label tiy 20 Ci/mMole). After this time the cul- and then collecting that medium at varitures were washed twice with cold (OOC) ous times after refeeding and precipitating 0.15 M NaC1, dissolved in 1.0 ml of 1.0 with TCA, final concentration 7 % . I n all NaOH, neutralized with 1.0 ml of 1.0 N the above cases, preparation of samples for HC1, and precipitated with trichloracetic scintillation counting was similar to that acid (TCA) at a final concentration of for TdR incorporation except that all pre5 % . The precipitate was washed three cipitations with glucosamine labeling were times with 5% TCA and the final precipi- allowed to proceed for at least four to five tate was dissolved in 2.0 ml of 0.05 N hours. NaOH and 0.1 ml was counted in Aquasol As mentioned i n the RESULTS, the total in a Nuclear Chicago Liquid Scintillation trypsin releasable cpm precipitable by TCA Spectrometer at a n efficiency of 15% for represented about 50% of the total cell 'H. The rate of DNA synthesis was ex- cpm. In the case of release of TCA precippressed as total TCA precipitable cpm/106 itable material into the medium, centrifugcells, where the cell counts were deter- ing the medium at 500 g for ten minutes mined in separate, duplicate cultures. did not alter the radioactivity recovered in Incorporation of D-glu~osamine-G-~H the medium, indicating that the rate of (gluc~samine-~H; New England Nuclear; cell detachment into the medium was sp. activity 2.4 Ci/mMole) into TCA pre- negligible . cipitable material was used to estimate Determination of mitoses the synthesis and release of cell associated glycoproteins into the medium. Cultures The number of mitoses occurring durwere prelabeled during exponential growth ing the interval 24-48 hours after stimufor a 4-day period prior to experimental lation of density inhibited cultures was use in fresh medium containing 1 & determined by adding colcemid (gift of gluc~samine-~H/ml. Several studies have Dr. R. Mankovitz) to the cultures, final shown that the TCA precipitable counts concentration 0.06 pg/ml (Dulbecco, ' 7 0 ) ,
569
DIFFUSION AND CELL GROWTH
at 24 hours and then allowing the cultures to incubate for another 24 hours. The mitoses occurring during this 24hour colcemid mitotic block, were then enumerated either directly under phase microscopy by counting the average number of rounded mitotic cells per high power field and expressing this as a percentage of the average number of cells per field or by regular microscopy after first staining the cultures with orcein, made up in 50% glacial acetic acid, in order to stain the condensed chromatin of the mitotic cells. The two procedures gave essentially the same results with the orcein procedure sometimes producing lower values for percent mitoses because the process of staining tended to remove some of the rounded and therefore tenuously attached mitotic cells from the dishes. In all cases, at least 200 mitotic cells were counted in determining the fraction of cells blocked in mitosis. Increasing the length of the mitotic block beyond 24 hours did not further increase the percentage of cells blocked in mitosis, indicating that this amount of time was sufficient to trap all the cells stimulated to progress into metaphase. Similarly, beginning the colcemid block prior to 24 hours after stimulation of the cultures did not increase the percent of mitoses,
Procedure for continuous mixing of m e d i u m The apparatus used to insure mixing of the medium is a n oscillating platform similar to the Bellco model-6000 rockcr platform (Bellco, Vineland, N.J.) except that the apparatus was designed so that the angle of oscillation could be varied. As pictured in figure 1, this platform on which the cultures were incubated was allowed to oscillate between 8 = 12" and 8 = - 12", as measured in the plane of oslcillation, driven at a frequency 15-20/ min by a variable speed electric motor. AS discussed in the RESULTS, this range of frequencies of oscillation did not mechanically dislodge cells from the culture plates and was chosen because it gave the most reproducible results over a large number of experiments. Furthermore, the rate of oscillation could not be increased over 30-35/min without having some of the
+
4 ml of medium start to slosh aver the top of the 60 m m dishes. In this frequency range of oscillation (15-20/min), 4 ml of medium in 60 m m petri dishes alternately flows back and forth with very little turbulence. When 8 = & 12" about one-sixth of the culture is not covered with medium for about onethird of a second, except for a thin layer of medium which does not flow away from the cells with the bulk af the medium. The maximum velocity attained by the medium as it passes over the cells in the center of the dish at 6 = 0" is 20-25 mm/sec. This velocity is within the range reported by Stoker ('73) for the velocity of the medium attained with his Riddle pump. I n spite of some medium turbulence at the outside rim of the plates which occurs during the oscillation of the platform, the resulting mitoses were found to be uniformly distributed throughout the dishes. I n all experiments involving comparison of cultures incubated on or off the platform, the time of incubation on the platform was never more than 20 hours. In experiments which carried on beyond 20 hours, cultures on the platform were simply taken off it and were left to incubate without agitation for the remainder of the experiment. Furthermore, all radiolabeling was done off the platform, regardless of the previous history of the cultures to be labeled. RESULTS
When normal fibroblasts become confluent in culture, a reversible GI arrest of growth occurs (Nilausen and Green '65) which can be overcome by refeeding the cells with fresh medium containing fresh serum. When such refeeding is done, thymidine incorporation experiments demonstrate a lag period af about ten hours preceding a partially synchronous wave of DNA synthesis in the interval 10-35 hours after treatment (Froehlich and Rachmeler, '72, '74; Todaro et al., '65). This, i n turn, is followed by a wave of mitoses occurring in the interval 25-40 hours after feeding (Todaro et al., '65; Wiebel and Baserga, '69). I n the first experiment described here, human fibroblast cultures were seeded at 1 X lo4 cells,/cmz in EMEM containing
570
JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
Fig. 1 Diagram oscillating platform described -htext. 9, the angle the platform makes with the horizontal, was allowcd to vary between 2 12" with a frequency of 15-20/min. In the figure,
e = + 12", and the level of the medium at that angle in a petri dish is shown.
10% serum and were allowed to grow for four days until growth stopped at a confiuent density of 6 X 104/cm2.At this time, the cultures were stimulated with fresh medium containing 5% serum. The medium removed from the cultures after the four days growth prior to stimulation was still able to support nonconfluent cell growth, though at only two-thirds the rate of fresh medium (data not shown). The effect of continuous medium agitation on the DNA synthetic pattern of these confluent cultures released from growth arrest by refeeding with medium containing 5% serum is shown in figure 2. This experiment demonstrates that the wave of increased TdR-3H incorporation into TCA precipitable material, during 30 minutes pulses, occurred during approximately the same interval whether or not the cells were incubated on the platform but that the amount of incorporation in the cultures stimulated with 5% serum on the platform was three times as great a s the incorporation in the similarly stimulated cultures not on the platform. Since the peak of synthesis occurred at the same time whether the cells were on or off the platform, several experiments identical to that in figure 1 were performed in which the amount of incorporation was measured with a continuous 4-hour pulse during the interval 20-24 hours after stimulation, with all cultures being pulsed off the platform. The results in table 1
indicate that the incorporation during this period ranged from 2.5-4 times greater for the cells preLiously incubated on the platform. Cell counts performed 20 hou-s after stimulation demonstrated no loss of cells in the cultures during incubation on the platform; thus the increased synthesis was not due to medium movement causing cell detachment thereby relieving adjoining cells from the growth constraining effects of confluence. As mentioned in legend to table 1, incubation of control unstimulated cultures on the platform had very little effect on DNA synthesis. The increased incorporation of TdR-T-1 synthesis observed in figure 2 and table 1 could have been due to either increased DNA synthesis or to increased uptake of TdR-JH into the soluble precursor pools. The following considerations, however, argue against merely increased uptake. If the cultures stimulated off the platform were put onto the platform for six to eight hours just prior to the actual DNA synthesis measurements as in table 1, no further increase in synthesis was observed over that observed in cultures stimulated entirely off the platform. However, the experiment shown in figure 3 indicates that the duration of incubation on the platform immediately after refeeding needed to increase TdR-W incorporation in the interval 20-24 hours after stimulation was only two to four hours. Thus, if changes in TdR-3H incorporation into soluble precursor pools occurred after only two to four hours initial incubation on the platform, they should also have occurred when the cells were incubated on the platform during the 6 to 8-hour period just prior to measuring DNA synthesis and an increased incorporation should have resulted. In spite of the above arguments, the number of cells entering mitosis after stimulation was measured to insure that the increased TdR-'H incorporation actually represented increased cellular proliferation. Cultures were grown to confluence as before and stimulated on or off the platform. Colcemid was added to the cultures at 24 hours and the percentage of mitotic cells accumulating over the next 24 hours was determined as described in MATERIALS AND METHODS. The results, reported in
57 1
DIFFUSION AND CELL GROWTH
I0
5
0
I
10
I
1
i
20
30
40
HOURS Fig. 2 Effect of medium mixing on DNA synthesis in stimulated confluent cultures. Cultures were seeded at l x 104/cm2in 10% serum and allowed to grow for four days until growth stopped at 6 x 10'/cm2. At this point, corresponding to t = 0 in the figure, one group of cultures was stimulated with EMEM containing 5% serum, while another control group remained unstimulated. DNA synthesis, measured by the incorporation of TdRJH into TCA precipitable material during a 30-minute pulse, was determined at the times indicated in the figure on duplicate cultures i n both the stimulated and control cultures. In the stimulated group, some of the cultures were incubated on the oscillating platform described in figure 1 for periods up to a maximum of 20 hours (0-0). In these cultures incubated on the platform, points in the figure prior to 20 hours were derived from cultures which were removed from the platform at the times indicated and immediately pulsed. Points after 20 hours represent cultures removed at 20 hours and pulsed at the points indicated and all contUP to 40 hours. The remaining cultures in the stimulated group (0-0) trol cultures (.----a) were incubated off the platform for the entire experiment.
table 2, demonstrate that the number of mitoses occurring in this interval was 2.34 times a s great in the cdtures incubated for 20 hours on the platform. Thus, the increased TdR-3H incorporation in these studies did indicate greater proliferative activity. The studies up to this point essentially confirm the findings of Stoker ('73), who
was able to greatly increase the percentage of confluent 3T3 cells to resume growth by using a pump mechanism to keep the medium flowing over the cells, and who suggested that diffusion limitation may play a role in the inhibition of growth in confluent cultures of normal fibroblasts. An alternative explanation for our re-
57 2
.JEFFREY E. FROEHLICH AND TASSOS P. ANASTASSIADES
4.
3
2
I
I
I
5
0
10
I
1
15
20
HOURS ON PLATFORM Fig. 3 Time of incubation 0x1 platform required for increased DNA synthesis in stimulated confluent cultures. Following the same protocol as in figure 2, cultures were stimulated at t = 0 with EMEM containing 5 % serum and were then allowed to incubate o n the oscillating platform for various periods of time up to 20 hours before being removed from the platform. Regardless of the time at which the cultures were removed from the platform, they were continuously incubated with TdRJH during the interval 20-24 hours after initial stimulation. The DNA synthesis determinations obtained on duplicate cultures are plotted in the figurc I S a function of the time of the cultures were allowed to incubate initially on the platform. The closed triangle (A) at t = 0 represents the synthesis in control unstimulated cultures. TABLE 1 Experiment no.
1 2 3
4
D N A synthesis (cpm/l06 cells) Controls
A
B
Ratio(B/A)
180 300 250 340
1100 1950 1500 950
2800 8100 4250 3350
2.5 4.1 2.8 3.5
D N A synthesis in stimulated confluent cultures. Following the protocol in the legend to figure 2, D N A synthesis was determined a s described in MATERIALS AND METHODS, by a 4-hour pulse with TdR-3H 20-24 hours after initial stimulation with EMEM containing 5 % serum in stimulated cultures either on (Column B ) or off (Column A ) the platform. Incorporation of TdR-3H in control unstimulated cultures is also given. Each figure in the table represents the average of two or three cultures. Control cultures incubated on the platform without medium change gave essentially the same results a s those off the platform (control column).
573
DIFFUSION AND CELL GROWTH
~-
TABLE 2
~ - _ _ _ _ _ _ ~
Experiment no
-_________~
_
_
~
~
~
~~~~
Controls
1
-
