EXPERIMENTAL RESULTS ON MAMMALIAN CELLS GROWING IN VITRO IN DEUTERATED MEDIUM FOR NEUTRON-SCATTERING STUDIES

Transcription

1 J. Cell Sci. 25, (i977) 87 Printed in Great Britain EXPERIMENTAL RESULTS ON MAMMALIAN CELLS GROWING IN VITRO IN DEUTERATED MEDIUM FOR NEUTRON-SCATTERING STUDIES J. MURPHY, C. DESAIVE, W. GIARETTI, F. KENDALL AND C. NICOLINI Division of Biophysics, Department of Physiology & Biophysics and Department of Pathology, Temple University Health Sciences Center, Philadelphia, Pennsylvania 19140, U.S.A. SUMMARY SV-40 virus-transformed human diploid fibroblasts (2RA) were grown in a monolayer on plastic Pctri dishes in an aqueous medium deuterated to different concentrations of deuterium oxide: 10, 20, 30, up to 60%. The cells must be acclimatized to concentrations higher than 20 % D 2 O by stepping them from a lower initial concentration during their exponential growth. The increase of cell doubling time with increasing deuterium concentrations seems to correlate, at least at 20 % D 2 O, with an initial period of suspended cell growth (lag-phase), and is qualitatively similar to that previously reported for Escherichia coli. INTRODUCTION Quantitative studies of the quaternary structure of macromolecules by neutron scattering at low angle have been recently reported in the study of ribosomes (Moore, Engelman & Schoenborn, 1974) and chromatin (Baldwin, Bosely, Bradbury & Ibel, 1975; Baserga & Nicolini, 1976). The work reported here shows that SV-40 virus-transformed human diploid fibroblasts (2RA cells) grow in media prepared with aqueous solvents containing up to 60% deuterium, thereby incorporating deuterium atoms instead of hydrogen atoms into their subunits. The results we report are preliminary requirements for the study by low-angle neutron scattering of the quaternary structure of large biomolecules obtained from mammalian cells. Incorporation of deuterium atoms in non-exchangeable positions in the subunits of macromolecules is essential for neutron scattering. It can be shown in fact that a molecule will not coherently scatter neutrons at low angles (coherent scattering produces the neutron interference effect that gives information on molecular structure) if its scattering density equals that of its solvent. Because we can utilize a mixture of D 2 O (~ pure coherent scatterer) and H 2 O (~ pure incoherent scatterer) in any Please direct all correspondence to: Professor C. Nicolini, Head, Division of Biophysics, Temple University Health Sciences Center, Philadelphia, Pennsylvania 19140, U.S.A.

2 88 J. Murphy and others proportion as a solvent, it is possible to contrast-match the solvent to the scattering density of the single components (e.g. protein or DNA) of any complex macromolecule (e.g. chromatin) which has incorporated deuterium atoms in non-exchangeable positions in a given quantity. The level of incorporation of deuterium atoms into cells can be assessed by photodisintegration (Haigh, 1953), by neutron transmission experiments or by CsCl gradient (Moore & Engelman, 1976). MATERIALS AND METHODS Suspensions of SV-40 virus-transformed human fibroblast (2RA) cells (generally cells in 5 ml) were distributed on plastic Petri dishes to grow in a monolayer in a constant temperature and humidity incubatoi" with CO 2 supplied. The nutrient medium consisted f 5'35 S f GIBCO powdered Minimum Essential Medium (with L-glutamine and Hanks' salts) and 3-5 g of sodium bicarbonate dissolved in 500 ml water or water-and-deuterium oxide, and sterilized by nitration through Nalge 0-20 micron-pore filters. To this were added 1 ml penicillin, 1 ml streptomycin, 10 ml BME vitamins, and 25 ml foetal bovine serum, using sterile technique. In some experiments, the cells were resuspended with trypsin, divided and replated at higher concentrations of deuterium oxide. In later experiments, the deuterium oxide concentration was stepped up before the monolayer reached confluence, that is, while the cells were in log-phase of rapid division. Before the cells were counted, the supernatant containing some dead cells was removed; the monolayer was washed 3 times with 3-ml volumes of cold Ca- Mg-free Hanks' solution. The cells were then suspended by incubation with 2 ml of trypsin in Hanks' solution for 10 min at 37 C. The plates were scraped with a rubber policeman, the trypsin suspension was removed, and the plates were washed with an additional 1 ml each of Ca-Mg-free Hanks', which was added to each sample of trypsinized suspension for counting in a haemocytometer or Coulter Counter. Statistical comparisons between the means of the samples drawn from the control and from the various deuterated suspensions were made by using the Cochran-Cox test (Snedecor & Cochran, 1972), which does not require any assumption about the total number of cells and their variances. The significance level required to reject the null hypothesis was set at 1 % in advance. The mean numbers of cells at each day for all experiments were first plotted on a semilog scale versus days to determine the validity of an assumption of a log-model growth. The doubling times T d were calculated by fitting the number of cells, Y, versus days T (after plating) with an exponential of the form Y = A e BT, such that T d = In 2/B (see Table 1). Alternatively, for deuterium concentrations of 20 % or larger, the data were better fitted (e.g. larger regression coefficient), with a model Y = A e B ' tt ~^' F \ where AT represents the lag-time of suspended growth; the newly computed doubling times were then T' d = In 2/B' (see Table 3). RESULTS Fig. 1 shows the results obtained for 2RA cells growing in deuterated solutions with concentrations of deuterium oxide at 10, 20, and 50 % (O, and 4-, respectively) after replating directly from aqueous medium. No significant statistical differences (see Materials and methods) could be inferred to exist between the means of both control ( ), and 10% D 2 O samples up to the fourth day. The curve at 20% deuterium concentration shows a lag-time of suspended growth of 1 day with respect to the aqueous control; thereafter the cells begin a similar log-phase growth. Furthermore, the deuterated cells reach a plateau phase at a lower cell concentration per

3 Cell growth in deuterated medium 89 square centimetre: specifically, the cell 'plateau concentration' decreases with increasing concentration of deuterium from 10 to 20%. The computed doubling times of 2RA cells consistently increase withthe increasing concentration of deuterium in the aqueous medium, showing a dramatic increase at 30 % D 2 O and higher concentrations (see Fig. 2 (A A) and Table 1). All the doubling times, computed relative to controls running simultaneously, are normalized to the mean value for 20 o x 10 Time (days) Fig. 1. Typical example of 2RA growth in aqueous (# %), 10% (O O)> 20% ( ) and 50% deuterated medium (H (-). The data for each experiment were obtained from cells that were grown in deuterated and aqueous media, prepared simultaneously, and sterilized in parallel (see Materials and methods). The lines represent a least square fit of the data up to 7 days, starting the first day, to an exponential growth model, without introducing any lag time. After 7 days the deuterated cells reach a plateau phase. the control doubling time. At 50 % D 2 O, the cells continue to grow extremely slowly for a week and longer at population values fairly close to the initial values (see Fig. 1, H h). There is morphological evidence, however, of cellular enlargement, as previously reported for lower organisms (Katz & Crespi, 1966). Above 50% D 2 O concentrations this experimental approach is unprofitable, because of the greatly increased dose-related cell loss. Because of this, different experiments have been performed, exposing 2RA cells at deuterium concentrations higher than 20% by stepping up samples from lower initial D 2 O concentrations and at different stages of 12

4 90 J. Murphy and others their growth processes. We found that for cells to continue growing at deuterium concentrations of 30, 40, 50, and 60% (in a few cases), after having been grown previously at 20%, they must be exposed to the higher concentration during their growth spurt. Cells may persist in higher D 2 0 concentrations for a week or longer, 20 o D J_ I D 2 O(%) Fig. 2. Doubling times (T d in days) of 2RA cells versus the percentage ( >) of deuterium oxide in the (H 2 O + D 2 O) medium, when transferred directly from aqueous medium (A A) and from a lower concentration (see Table 2) of deuterated medium ( ). The lines represent the actual fit of the data to a model of the form T d = A* e kji. For other details see the text and Tables 1-3. All T a values were normalized to the same average doubling time of 2RA in aqueous medium (T d = 1-77 days). but do not increase in numbers if transferred to the higher concentration in the plateau phase. Fig. 3 shows the possibility of growing 2RA cells at up to 50 % D 2 O concentration if they are stepped up from 20 % D 2 O at the second day of log-phase. DNA content was increasing in direct proportionality with cell numbers (not shown). Table 2 shows the doubling times calculated for 2RA cells at various high deuterium concentrations after the transfer of the cells from lower D 2 O concentration. The correlation coefficients of the least square regression of each data set was always above 0-96 and the number of experimental points was between 4 and 8. The extrapolated values (see Table 2 and Fig. 2) obtained for the cells suspended in up to 60 % deuterium oxide show that mammalian cells may grow under the proper protocols (stepping from 90

5 Cell growth in deuterated medium 91 lower concentrations during log-phase growth) at deuterium concentrations higher than previously supposed. There is some evidence of a general dose-related inhibition of growth by D 2 O, but a more striking and reliable observation was of a lag time at the beginning of exposure to D 2 O concentrations of 20% and higher, during-which the numbers of cells remained constant. Table 1. Doubling time (7\i in days) of 2RA cells replated directly from aqueous to deuterated media of various concentrations (% D 2 O) D 2 O (%) Doubling time (T d, days) o-o IOO of 5 - t 70-of i-73 I-8I All the doubling times, computed relatively to a control running simultaneously, are normalized to the mean value for the control doubling time. Fitting doubling times versus deuterium concentration ( )) with a model of the form, T& = A* e ld as shown in Fig. 2 we can see that T d increases asymptotically to infinity in the concentration range of D 2 O %, i.e. mammalian 2RA cells do not practically increase in number any more and start dying at about 70 % D 2 O. The negative value at 70 % D 2 O represents the death rate, i.e. the time necessary for reducing to half the original cell number. f At these concentrations we followed the cells up to 7-9 days. The doubling times were computed indirectly by fitting the data with the model Y = A* exp (BT). At 50 % D 2 O the cells were becoming quite large. Table 3 summarizes the results of 5 independent experiments with 2RA cells growing in 20 % D 2 O and an aqueous control medium, prepared at the same time and by the same person as described in Materials and methods. The results consistently show a lag-time of 1 day in 20% D 2 O concentration, during which cell growth is arrested, followed by the onset of a growth spurt similar to that in aqueous solution. An increased lag period was found for increasing D 2 O concentration (not shown). Similar initial lag-phases have been reported in the tobacco seed (Lewis, 1933), Escherichia coli (Rothstein, Manson, Hartzell & Kritchevsky, 1959), ectromelia virus (French, Hughes & Siegel, 1961), and Arbacia punctulata (Gross & Spindel, i960). DISCUSSION These results indicate that mammalian cells (2RA) may grow at deuterium concentrations at least up to 60 % if stepped from a lower concentration during log-phase growth. A striking effect in the dose-related inhibition of the growth process by various increasing concentrations of D 2 O seems to be the appearance of a variable lag time during which cell division ceased but other cell machinery such as protein synthesis, indicated by cell enlargement, may continue. In the case of 20 % D 2 O, the

6 J. Murphy and others 20 o X o Z Time (dnys) Fig. 3. Growth of 2RA cells in aqueous medium ( ), 20% D 2 O ( ), 50 % D 2 O after transfer at the 4th day from 20 % D 2 O ( D) and 70 % D 2 O after transfer at the 6th day from 50% (A A)- During these experiments the old medium at lower concentration was discarded and replaced with the new medium at higher concentration. The continuous curves represent the actual fits with the exponential growth model. Table 2. Doubling times (Ta in days) during step-up experiments of 2RA cells replated from a lower to a higher D 2 O concentration (D) and directly from aqueous control running simultaneously (C). The ratios Ta (D)jT<x (C) are reported Step-up experiments I. 30 % D 2 O from 10 % D 2 O II. 5 % D 2 O from 20 % D 2 O III. 50 % D 2 O from 20 % D 2 O IV. 60 % D 2 O from 50 % D 2 O The correlation coefficients of the least-square regression of each data set to the same model as in Table 3 was constantly above 0-96 and the number of experimental points was between 4 and 8. See also Fig. 2 for doubling time dependence at various D 2 O concentrations when stepping up from lower D 2 O concentration. The doubling times T d were computed from T a = In2/.B, where B is obtained by fitting the data to a model Y = Ae BT d (C)

7 Cell growth in deuterated medium 93 increase in doubling time (7\i) is due essentially to the lag time of 1 day, and the total cell-cycle duration seems to remain practically invariant (see Table 2). These findings suggest the need to explore further the concentration-dependence of the lag-time length and of the total cell-cycle duration of mammalian cells suspended in deuterium oxide at higher concentrations. Experiment I II III IV V Table 3. Experiments on 2RA cells growing in aqueous (C) and 20% deuterated medium (D). AT days 1 1 i 1 1 Doubling times A T d (C), days T' d {D), days T' d (D)l T d (C) I-I2 098 I-O9 i-oo I 02 Regression L.UC1111-lCJ I to / ^ (C) (D) 099 (7) 099 (6) 099 (4) 099 (3) 099 (6) 097 (4) 0-98 (s) 098 (6) 097 (4) 086 (4) The data during log growth were fitted by a least square regression to a model Y = ^4e B ' (7 " ^r> where T represents time (days). In the first column the differences ( AT in days) in lag time of the deuterated cell with respect to the control are reported and are equal to 1 for all 5 experiments. In the last column the regression coefficients and the number of experimental points (in parentheses) are also reported. These studies could prove useful for the study of the quaternary structure of complex macromolecules (such as chromatin, membranes, ribosomes) by the lowangle neutron-scattering technique. For this reason we started these experiments and we are carefully computing by various physical means (Haig, 1953; Rothstein et al. 1959; Moore & Engelman, 1976) the extent of deuteration obtained in the non-exchangeable positions in various cell components (proteins, DNA, RNA, etc.). As a final comment, we may say that one of the biggest limitations we encountered in growing mammalian cells in deuterated media has been the economic one. This work was supported by grants numbers CA18258 and CA19293 from the National Institute of Health. Dr Claude Desaive was a recipient of an International Fellowship from the National Cancer Institute. REFERENCES BALDWIN, J. B., BOSELEY, P. G., BRADBURY, E. M. & IBEL, K. (1975). The subunit structure of the eukariotic chromosome. Nature, Lond. 253, BASERGA, R. & NICOLINE, C. (1976). Chromatin structure and function in proliferating cells. Biochim. biophys. Acta 458, FRENCH, S. W., HUGHES, A. M. &SIEGEL, B. V. (1961). Effect of deuterium oxide on ectromelia viral infection. Lab. Invest. 10, GROSS, P. R. & SPINDEL, W. (i960). Heavy water inhibition of cell division: approach to mechanism. Ann. N. Y. Acad. Sci. 90, CEL 25

8 94 J- Murphy and others HAIGH, C. P. (1953). Analysis for deuterium based on photoneutron effect. Nature, Lond. 172, 359- KATZ, J. J. & CRESPI, H. L. (1966). Deuterated organisms: cultivation and uses. Science, N.Y. 151, LEWIS, G. N. (1933). The biochemistry of water containing hydrogen isotopes, jf. Am. chem. Soc O3-35O4- MOORE, P. B. & ENGELMAN, D. M. (1976). The production of deuterated E. coli. In Neutron Scattering for the Analysis of Biological Structure, BNL (ed. B.P. Schoenborn), pp Brookhaven National Laboratory. MOORE, P. B., ENGELMAN, D. M. & SCHOENBORN, B. P. (1974). Asymmetry in the 50s ribosomal subunit of Escherichia colt. Proc. natn. Acad. Sci. U.S.A. 71, ROTHSTEIN, E. L., MANSON, L. A., HARTZELL, R., JR. & KRITCHEVSKY, D. (1959). Effect of deuterium oxide on the synthesis of T-5 and T-7 bacteriophages. Nature, Lond. 184, SNEDECOR, G. & COCHRAN, W. (1972). Statistical Methods. Ames: Iowa University Press. (Received 12 July 1976)