IN VITRO Volume13, No. 12, 1977 Allrightsreserved9

GROWTH

OF HUMAN SKIN FIBROBLASTS FETAL BOVINE SERUM

IN DIALYZED

EUGENE ELMORE ANDMICHAEL SWIFT

Biological Sciences Research Center 220H, Genetics Curriculum, Department of Medicine, University o/North Carolina, Chapel Hill, North Carolina

SUMMARY Human diploid fibroblast cultures plated at or below a density of 2 x 103 cells per cm 2 grew very slowly or not at all in M E n supplemented with 10% fetal bovine serum that had been dialyzed for 24 hr. Adding serine (0.2 mM) or pyruvate (1.0 mM) to M E M and 10% dialyzed serum restored growth to the level observed with 10% nondialyzed serum. Serine and pyruvate also were able to overcome partially the growth arrest induced by a reduced serum concentration (1 or 2%). Human fibroblast cultures grew very well in 100% fetal bovine serum that had been dialyzed against M E M . For cells grown in dialyzed serum, the final number increased with increasing serum concentration, in contrast to the well established toxic effects of high concentrations of nondialyzed serum.

K e y words: human fibroblasts; serum; toxicity; dialysis. INTRODUCTION

Biological Co. Sodium pyruvate was obtained from Sigma Chemical Co. (>99% purity}, Grand Island Biological Co. (>99% purity}, and Nutritional Biochemicals (reagent grade). Seven cell lines were used in these experiments. For C34, C35, C39, C40, C41 and C47, the age, sex, and race of the biopsy donor and the mean population doubling time are given in Table 1 of Elmore and Swift (8). C53 was from a normal white female, aged 11 yr, and has a doubling time of 21.7 hr. All experiments were done between the 4th and l l t h passages, with each subculture at a 1:3 or 1:4 dilution. Cells were counted with a Biophysics Systems (Mahopak, N.Y.) Model 6300A Cytograf. The cell counts with the Cytograf were compared for accuracy to counts made with a Model B Coulter Counter and with a hemacytometer. Falcon culture flasks and dishes were used. Sera were dialyzed in cellulose tubing (8-667-B, Fisher Scientific Co. } that had been boiled for 3 hr in distilled water (the water was removed and replaced with fresh distilled water twice during the 3 hr) to remove contaminants. In preliminary exMATERIALS AND METHODS periments, the growth-supporting ability of serum was measured after 2, 4, 6, 8, 10, 12, 18 and 24 hr Human fibroblast cultures were initiated from biopsies of forearm skin, tested for mycoplasma of dialysis against distilled water or M E M . In all and maintained as described previously (7). All later experiments, one volume of serum was diamedia, fetal bovine sera (FBS) and L-amino acids lyzed for 24 hr at 4~ against 25 volumes of (>99% purity) were purchased from Grand Island M E M with one change of the dialyzate. Dialyzed 837 Cell culture media made with dialyzed serum are valuable for analyzing the nutritional requirements of cell cultures and for a variety of genetic studies of human diploid fibroblast cultures. For the growth of certain cell lines, Eagle (1) pointed out that serine, asparagine, glycine and pyruvate must be added to a dialyzed-serum medium. While the growth requirements of human diploid fibroblasts in dialyzed-serum medium have not been systematically studied, cells of this type have been grown in the presence of dialyzed serum in a complex medium I2, 3), and in minimal essential medium (MEM) (1) or basal medium I4} at a high cell density (5, 6). We have studied the growth at low cell density, the response to supplementation with pyruvate and several of the nonessential amino acids, and the relationship between growth rate and serum concentration for several human diploid fibroblast cultures in M E M with dialyzed or nondialyzed serum.

838

ELMORE AND SWIFT

and nondialyzed sera were filtered through a 0.22/~m Millipore filter. To attempt to restore growth in 10% dialyzed FBS, serine, glycine, aspartic acid, asparagine, proline and alanine at 0.1 mM, and pyruvate at 1.0 mM were added individually to M E M . Those amino acids whose inclusion in the culture medium (at 0.1 mM) led to a greater cell number than that observed in M E M with only dialyzed serum were tested then at various concentrations (0.05 mM to 1.6 mM). After individual nutrients were tested, the o p t i m u m concentrations for aspartic acid and asparagine were determined in the presence of 0.2 mM serine. For each experiment, ceils were plated at 1 to 2 x 103 cells per cm ~ in 35-mm dishes with 2 ml M E M with 10% FBS. After a 3to 4-hr attachment period, the medium was removed and the plates were refed randomly with the various test media. Each experiment also ineluded a set of plates which was reled with 10% F B S in M E M and a set which was refed with 10% dialyzed F B S in M E M . The o p t i m u m pyruvate concentration was determined in M E M with 10% dialyzed FBS, both with and without 0.2 mM serine and 0.2 mM aspartic acid, using seven different pyruvate concentrations between 0.01 mM and 1.6 raM. For experiments in which concentrations of nondialyzed serum above 10% were used, the concentration of M E M - n u t r i e n t s was adjusted to equal that in 10% nondialyzed s e r u m - M E M . In adjusting the concentration of small molecularweight compounds, the contribution of nondialyzed serum was disregarded.

EXPERIMENTAL RESULTS In M E M with 10% serum, final cell counts decreased with increasing serum dialysis time over an 8-hr period. If dialysis was continued and the dialyzate not changed, there was no further change in the growth-supporting characteristics of the serum. With a change of dialyzate at 8 hr, there was a further decline in the growth-promoting quality of the serum. T h e r e was no difference in final cell counts using serum which had been dialyzed for 18 or 24 hr. As expected, dialysis against distilled water produced some precipitate in the serum. R e m o v a l of this precipitate by filtration did not affect the ability of dialyzed serum to support growth in experiments similar to those described (see legend, Fig. 2). The precipitate also could be removed by dialysis against M E M . Serum prepared in this

way was indistinguishable in growth experiments from that dialyzed against two changes of M E M . Fig. 1 shows that increasing the initial cell density in M E M with 10% dialyzed serum increased the number of cell doublings. T h e number of doublings in 72 hr at the highest density was less than that I1.7 to 2.7) in 10% nondialyzed serum. T h e differences between cell lines in their growth rate at varying cell densities were much smaller when cell protein rather than cell number per ml of culture medium was taken as the independent variable. The greater the age of the subject at the time of the biopsy, the greater the protein content per cell ~see legend, Fig. 1). At the lowest density tested, three of six lines showed some increase, although small, in cell number in 10% dialyzed serum.

i

.I

.5 CELLULAR

PROTEIN

I.O (,lJlilcrn2)

i

i

1,5

FIG. I. The number of doublings iLrter72 hr in 10%

dialyzed FBS in MEM is shown as a function of the initial population density expressed as cellular protein per cm2. The curve for each line represents one experiment in which eight 35-ram dishes were plated for each point. Four of these were counted 3 hr after plating finitial cell counts), and the remaining four were refed with 2 ml 10% dialyzed FBS in MEM and counted after 3 days {final cell counts). For each cell line, two 60-mm dishes were plated at the highest density for cellular protein determinations after 3 hr of incubation. The age of the biopsy donor, the passage number, the protein per cell and the symbol used in this figure for each cell line are: C34 - - 5 yr, passage 8, 0.13 ng, B; C35 - - 4 yr, passage 7, 0.16 ng, i ; C39 - - 26 yr, passage 8, 0.55 ng, O; C40 23 yr, passage 5, 0.39 ng, 0 ; C40 - - 23 yr, passage 9, 0.41 ng, , ; C41 - - 23 yr, passage 6, 0.19 ng; ~; C47 - 13 yr, passage 5, 0.09 ng, ~ . At each density, the number of doublings during the 3 days was determined from the formula: ffinal cell counts) log (initial cell counts) Number of doublings = log 2 -

-

When the number of doublings was below 0, there was cell loss at that density.

839

GROWTH IN DIALYZED SERUM

FIG. 2. A representative experiment showing the final cell number as a function of serum concentration for two lots of dialyzed FBS. Httrnan skin fibroblasts i l 1-yr-old female} were plated at 8 x 103 cells per 35-ram dish, refed with 2 ml various concentrations of dialyzed or nondialyzed serum of Lot 1 t[]~ or Lot 2 ([i~, and counted after 72 hr. Each point is the mean of four plates _+S.D. The final cell number increased with increasing concentrations of dialyzed F B S IFig. 2}. Cells in dialyzed serum required a serum concentration of 60 to 70~ to equal the growth rate in 10% nondialyzed serum. There were small differences between the two serum lots studied in the four experiments of this type using four different h u m a n cell lines. Note the excellent growth rate with no apparent toxicity at high concentrations of dialyzed serum. Dialyzed F B S supplemented with Eagle's i l) mixture of nonessential amino acids supported vigorous growth ITable 1L Serine alone was equally effective in restoring growth in dialyzed serum. Glycine was effective only partially, and

alanine, aspartic acid, asparagine and proline had no measurable effect on growth in M E M with 10% dialyzed serum. In the experiments in which cells were grown in M E M with 10% dialyzed serum and 0.2 mM serine, the addition of 0.1 mM aspartic acid or asparagine led to an even greater final cell number, whereas adding alanine, glycine or proline (0.1 mM) did not. The increase in cell number was approximately the same for serine concentrations between 0.1 and 1.6 mM. T h e m i n i m u m concentration of aspartic acid or asparagine required for m a x i m u m growth enhancement was 0.1 mM. The final cell number decreased if asparagine or proline was present at concentrations above 0.8 raM. When dialyzed serum was supplemented with 0.8 to 1.6 mM pyruvate, the increase in cell number was equal to or greater than that with serine, aspartic acid or glycine singly or in combination. Pyruvate from three commercial sources was tested with identical results. In M E M with 0.2 mM serine and 10% dialyzed serum, the final cell number after 3 days was 1.3 times greater with pyruvate than without (four experiments). Fig. 3 shows the effect of combined supplementation with 0.2 mM serine, 0.2 mM aspartic acid and 1.0 mM pyruvate on growth in M E M with varying concentrations of dialyzed and nondialyzed FBS. In nondialyzed serum without supplementation, there was less growth in low (1 to 5%~ and high (20 to 50%} serum concentrations than in 10%. At any one concentration of nondialyzed serum, the growth rate was consistently greater with serine, aspartic acid and pyruvate supplementation than without. In 1% nondialyzed serum with supplementation, the final cell

TABLE 1 GROWTHIN MEM WITH 10% FETALBOVINESERUMANDWITH VARIOUSADDITIVES Serum

Additivesa

Number of Lines Tested

Mean % Relative Growth b

Nondialyzed 7 100 ~9) Dialyzed 7 0 19; Dialyzed GIBCO NEAA mixture 7 107.0 _+3.6 c {9j Dialyzed Serine 4 108.6_+ 3.6 ~8} Dialyzed Glycine 2 32.4_+ 7.4 {2D Dialyzed Alanine, asparagine, aspartic acid or proline 2 0_+1.0 t8) Dialyzed Serine and aspartic acid, or serine and asparagine 3 123.8_+3.5 {5} Dialyzed Pyruvate 2 130.7_+ 9.2 {3j a The concentration of each amino acid added was 0.1 mM, and pyruvate was 0.8 mM. b Initial cell counts were not done. The net increase in cell number with each supplement or set of supplements was estimated by subtracting the final cell number in 10% dialyzed FBS from the final cell number in the supplemented medium. The percent relative growth was calculated for each condition by dividing the net increase in cell number in that condition by the net increase in cell number in 10% nondialyzed FBS and multiplying by 100. The plating density for these experiments was 1 to 2 x 103 cells per cm 2. The number in parentheses is the number of experiments. c Standard error of the mean.

840

ELMORE AND SWIFT

~P~~

7

TABLE 2 GROWTH IN MEM WIT• VARYING CONCENTRATIONS OF NONDIALYZED SERUM AND OF NONESSENTIAL AMINO

?

ACIDS(NEAA) a 100

i

% Serum

Q- 0 PERCENT SERUM

FIO. 3. The percent relative growth in MEM with various concentrations of dialyzed or nondialyzed FBS with or without the combined addition of 0.2 mM serine, 0.2 mM aspartic acid and 1.0 mM pyruvate. Cells were replicate-plated at 9 to 13 x 102 cells per cm 2 in 35-mm dishes and refed various concentrations of nondialyzed serum t&), nondialyzed serum with additives (11), and dialyzed serum with additives ( 9 The percent relative growth was calculated using 10% dialyzed serum in MEM as the 0% control and 10% nondialyzed serum in MEM as the 100% control. Each point represents the mean - S.E.M. of four experiments in which four plates were counted for each experiment. number after 3 days was approximately 40% of that in M E M with 10% nondialyzed serum without supplementation. Even though serine, aspartic acid and pyruvate supplementation enhanced the growth rate at concentrations of nondialyzed serum between 20 and 50%, the final cell number at these concentrations did not exceed that with 10% nondialyzed serum alone. For all serum concentrations, the final cell number in M E M with dialyzed serum supplemented with serine, aspartic acid and pyruvate exceeded that in M E M with nondialyzed serum. Striking differences were observed at low ~1 to 2%) and high ~20 to 50%) serum concentrations. In contrast to nondialyzed serum, the final cell number in dialyzed serum with supplementation was greater at concentrations of 20 to 50% than at 10%. With supplementation, the percent relative growth (relative to that in 10% nondialyzed serum) in 50% dialyzed serum (159.1 _+ 6.6%) was significantly greater than that in 10% dialyzed serum I124.9 _+ 11.0%; mean of five experiments; p