Growth of Cultured Cells from Patients with Fanconi Anemia EUGENE ELMORE AND MICHAEL SWIFT lkpartment of Medicine, Genetics Curriculum, and the Biological Sciences Research Center, University of North Carolina, Chapel Hill, North Carolina 27514
ABSTRACT Fibroblast cultures derived from skin biopsies of patients with Fanconi anemia had doubling times (mean of five lines: 30.3 f 0.2 hours) significantly longer than randomly selected normal controls (mean of nine lines: 22.9 f 0.4hours). Control cultures grew more slowly in the enriched media RPMI 1640 and McCoy's 5A than in MEM, while a culture from a patient with Fanconi anemia grew more slowly only in McCoy's 5A. Differences in growth characteristics between Fanconi anemia and normal cell cultures may be useful in analyzing the metabolic error determined by the Fanconi anemia gene. Skin fibroblast cultures have been used al., '72; Olmsted, '67), the incubation temto analyze the biochemical defects associ- perature, and the genetic makeup of the ated with a large number of inherited dis- biopsied subject (Martin et al., ' 7 0 ; Epstein orders. In all these investigations, the cell et al., '66; Segal and McCoy, '74). The esculture experiments were based on a pre- tablished or accepted cell doubling time for viously observed clinical abnormality asso- normal controls in a particular laboratory ciated with the individual disorders. There may also be influenced by the tendency to are numerous single gene syndromes for choose as standard normal lines only those which no characteristic biochemical abnor- cell cultures whose growth is vigorous. For the autosomal recessive disorder, mality has been described. The usefulness of fibroblast cultures in genetic research Werner's syndrome, there is strong evidence would be greatly extended if cultures from that skin fibroblast cultures from homozyindividuals with such syndromes could gous affected individuals grow exceptionalthemselves provide the metabolic clues to ly poorly in culture (Martin et a]., '70; Ep guide further investigation of the biochem- stein et a]., '66). There are also data which ical abnormality associated with the dis- show that fibroblast cultures derived from ease-producing gene. The problem is to individuals trisomic for chromosome 21 choose a cell culture characteristic which grow more slowly than control cultures would allow for a specific metabolic anal- (Segal and McCoy, '74). ysis of the difference between normal conFanconi anemia (FA) is an autosomal retrol cells and the mutant line under inves- cessive syndrome which is of interest betigation. cause of evidence that the gene predisposes In bacterial systems differential growth both the homozygote and the relatively in various culture media has long been used common heterozygote to an increased risk in the analysis of nutritional mutants. Anal- of cancer or leukemia (Swift, '71). Among yses of the growth characteristics of human the cellular abnormalities which have been diploid fibroblast cultures are difficult be- described in cells from patients homozygous cause the growth of individual cultures for the FA gene are a high prevalence of may vary according to the site of the biop chromosomal breakage (Swift and Hirschsy (Martin et al., '70), the age of the biop horn, '66) and an increased susceptibility sied subject (Martin et al., '70), the length to transformation by the oncogenic virus of time in culture (Merz and Ross, '69; SV40 (Todaro et al., '66). Cells derived from Macieira-Coelho et al., '66), the chemical FA patients are said to grow poorly in culcomposition of the medium including poorly Received Apr. 21, '75. Accepted July23, '75 defined factors in the serum (Fedoroff et J. CELL. PHYSIOL., 87: 22!+234.
229
230
EUGENE ELMORE A N D MICHAEL SWIFT
ture (Young, '71) but there are no quantitative data supporting this. Because a reproducible difference in growth rate between FA and normal cells would provide an opportunity for the semisystematic irvestigation of the metabolic abnormality determined by the FA gene, we have compared the growth of fivecultures from homozygous FA patients to nine control lines grown at the same time in our laboratory. METHODS
Cell culture Fibroblast cultures were established from biopsies of forearm skin, abdominal skin of deceased patients, or from the forearms or hands of patients undergoing plastic surgery. The biopsies from surgical patients were pieces of normal skin removed in the course of repairing nongenetic conditions. Cultures were maintained and most growth experiments were done in minimal Eagle's medium (MEM) with 10% fetal calf serum (Grand Island Biological Company) in 5 % COz at 37°C. Only two lots of serum were used for all the experiments; both sera had been carefully selected for their support of rapid growth of normal human fibroblast cultures. All media were purchased in powdered form from Grand Island Biological Company and prepared in our laboratory. The cultures were grown initially in 25 om2 Falcon flasks and were subcultured using 0.005% purified trypsin (Worthington TR TCPK) in 75 cm2 flasks twice weekly (at 1:2 or 1:3) before reaching confluency. Each cell line was frozen at the second, third, or fourth passage in MEM with fetal calf serum and glycerin. Tests for mycoplasma (broth culture and by measurement of uridine phosphorylase activity (Levine, '72)) were done at the time of freezing each cell line and at every third week subsequent to thawing when a line was being carried for experiments. Cell counts were done on trypsinized cells suspended in phosphate-buffered saline, using a model B Coulter counter. Cells from the fifth through tenth subculture were used in all experiments. All nine normal control cultures initiated in our laboratory from 4/9/73 through 3/25/74 were used, in addition to five lines derived from unrelated patients with Fanconi anemia and two cultures obtained from parents of two unrelated FA patients.
Measurement of cell doubling time In each experiment at least one FA cell line and one normal control were studied simultaneously, to control for undetected variation of the medium, serum, or other factors affecting the rate of cell growth. Cells from actively growing sub-confluent cultures were plated into 20 cm2 Falcon tissue culture dishes at two or three different densities between 2 X 103 and 3.5 X 103 cells/cmZ, with at least six replicate plates for each cell line at each density. Care was taken to disperse the cells evenly on the plates, and plates that showed an uneven distribution of cells after attachment were discarded. Preliminary experiments were done in which normal or FA cell lines were plated onto 24-36 replicate plates at a single density and counts done at two different times each day beginning 18 hours after subcultivation, in order to determine two time points between which exponential cell growth was always observed. As a result of these preliminary experiments, measurements of cell doubling time were based on initial cell counts done 2G28 hours after plating and final cell counts done 4 2 4 3 hours later. For each cell density the initial and final cell numbers (the mean of three or four plates) were used to calculate the mean doubling time which is given by the formula DT (doubling time) = (log 2) (time lapse in hours) log (final cell count) -log (initial cell count).
We studied the variation in cell doubling time that occurs when control or FA cell lines, grown previously in MEM, are transferred to an enriched medium (McCoy's 5A, RPMI 1640) during the growth experiment. Sixteen replicate plates were prepared at a single cell density and three hours after subculturing the MEM was removed. Half the plates were refed with MEM and the remainder with the enriched medium. Initial and final counts were done as described above. Autor adiograp hy Cells were plated on coverslips at an initial density of 5 X 103 cells/cmZ. Thymidine (methyl-3H; 20 Cilmmole; New England Nuclear) was added at 1 pCi/ml for one hour for pulse labeling at 2, 22, 34 and 45 hours after subcultivation, and at 0.1 pCi/
23 1
GROWTH OF CULTURED FANCONI ANEMIA CELLS
ml immediately after cell attachment for continuous labeling. At the specified times (up to 99 hours for continuous labeling) cells were rinsed with phosphate buffered saline and fixed with Carnoy's. The air-dried coverslips were stored at 4 "C for three days after dipping in Kodak NTB-2 emulsion, developed in Kodak D19 and stained with Giemsa stain for one minute. At each time point 500 cells were counted for each cell line. RESULTS
For Fanconi anemia cultures 24 t 2.2 days were required from the time the biopsy was initiated to the first subculture, while the cultures from the parents required 19.7 t 2.7 days and the normal controls 20.1 -e 0.9 days. In all experiments exponential cell growth had begun 24 hours after subcultivation. The logarithm of the cell number increased linearily with time over the interval from 24 to 72 hours after subcultivation. Growth curves for one normal and one FA cell line are shown in figure 1. The mean cell doubling times for nine normal controls, five FA homozgous and two FA heterozygous cell lines are shown in table 1. In no experiment did an FA cell
-
30
-
0
0 X I
K W
m 10-
5= i w
0
7-
6: 5 4-
TABLE 1
Mean doubling times for cell cultures derived from normal controls, from patients with Fanconi anemia, and from parents of patients with Fanconi anemia Doubling time Cell line
Controls c2 c3 c4 c5 C8 C18 c20 c22 C23
Age at time of biopsy
32 23 25 2 24 21 3 29 4
Fanconi anemia homozygotes c12 6 8 C17 c19 6 C25 14 B68 6 Fanconi anemia heterozygotes C13 31 C24 52 1
Sex
F M F M
Race
W
B B
Skin site
Abdomen Wrist Wrist Hand Hand Forearm Abdomen Hand Hand
22.2 f 0.6 (4) I 19.7 f 0.5 (6) 24.7k0.5 (9) 22.2 f 0.7 (3) 24.2 % 1.0 (4) 24.8 f2.0 (2) 20.0 f 0.7 (4) 24.0 f 0.7 (6) 22.0 f 0.5 (3) 30.6 f 0.8 (7) 30.7 -t 1.6 (6) 32.1 & 1.3 (8) 30.9 f 1.2 (4) 26.2 k 1.2 (3) 25.5 % 1.0 (4) 26.3 f 1.7 (5)
F
W W W W B W
M M
B
F F
W W
M
B
Fore arm Forearm Abdomen Forearm Interscapula
M M
W W
Forearm Forearm
M M M M
W
mean f S.E.
The number in parentheses i s the number of experiments done with each cell line
232
EUGENE ELMORE AND MICHAEL SWIFT
growth (table 2) than normal controls. With continuous labeling the two control cultures exhibited 89 and 93 percent labeled cells at 45 hours, and 97 percent at 70 or 99 hours. The Fanconi anemia cultures had 72 and 75 percent labeled cells after 45 hours of continuous labeling, and 85 to 92 percent at 70 and 99 hours. The mean cell doubling time for the FA heterozygous cells (26.0 f 1.0 hours), was less than that of the FA homozygous cells (p
ABSTRACT Fibroblast cultures derived from skin biopsies of patients with Fanconi anemia had doubling times (mean of five lines: 30.3 f 0.2 hours) significantly longer than randomly selected normal controls (mean of nine lines: 22.9 f 0.4hours). Control cultures grew more slowly in the enriched media RPMI 1640 and McCoy's 5A than in MEM, while a culture from a patient with Fanconi anemia grew more slowly only in McCoy's 5A. Differences in growth characteristics between Fanconi anemia and normal cell cultures may be useful in analyzing the metabolic error determined by the Fanconi anemia gene. Skin fibroblast cultures have been used al., '72; Olmsted, '67), the incubation temto analyze the biochemical defects associ- perature, and the genetic makeup of the ated with a large number of inherited dis- biopsied subject (Martin et al., ' 7 0 ; Epstein orders. In all these investigations, the cell et al., '66; Segal and McCoy, '74). The esculture experiments were based on a pre- tablished or accepted cell doubling time for viously observed clinical abnormality asso- normal controls in a particular laboratory ciated with the individual disorders. There may also be influenced by the tendency to are numerous single gene syndromes for choose as standard normal lines only those which no characteristic biochemical abnor- cell cultures whose growth is vigorous. For the autosomal recessive disorder, mality has been described. The usefulness of fibroblast cultures in genetic research Werner's syndrome, there is strong evidence would be greatly extended if cultures from that skin fibroblast cultures from homozyindividuals with such syndromes could gous affected individuals grow exceptionalthemselves provide the metabolic clues to ly poorly in culture (Martin et a]., '70; Ep guide further investigation of the biochem- stein et a]., '66). There are also data which ical abnormality associated with the dis- show that fibroblast cultures derived from ease-producing gene. The problem is to individuals trisomic for chromosome 21 choose a cell culture characteristic which grow more slowly than control cultures would allow for a specific metabolic anal- (Segal and McCoy, '74). ysis of the difference between normal conFanconi anemia (FA) is an autosomal retrol cells and the mutant line under inves- cessive syndrome which is of interest betigation. cause of evidence that the gene predisposes In bacterial systems differential growth both the homozygote and the relatively in various culture media has long been used common heterozygote to an increased risk in the analysis of nutritional mutants. Anal- of cancer or leukemia (Swift, '71). Among yses of the growth characteristics of human the cellular abnormalities which have been diploid fibroblast cultures are difficult be- described in cells from patients homozygous cause the growth of individual cultures for the FA gene are a high prevalence of may vary according to the site of the biop chromosomal breakage (Swift and Hirschsy (Martin et al., '70), the age of the biop horn, '66) and an increased susceptibility sied subject (Martin et al., '70), the length to transformation by the oncogenic virus of time in culture (Merz and Ross, '69; SV40 (Todaro et al., '66). Cells derived from Macieira-Coelho et al., '66), the chemical FA patients are said to grow poorly in culcomposition of the medium including poorly Received Apr. 21, '75. Accepted July23, '75 defined factors in the serum (Fedoroff et J. CELL. PHYSIOL., 87: 22!+234.
229
230
EUGENE ELMORE A N D MICHAEL SWIFT
ture (Young, '71) but there are no quantitative data supporting this. Because a reproducible difference in growth rate between FA and normal cells would provide an opportunity for the semisystematic irvestigation of the metabolic abnormality determined by the FA gene, we have compared the growth of fivecultures from homozygous FA patients to nine control lines grown at the same time in our laboratory. METHODS
Cell culture Fibroblast cultures were established from biopsies of forearm skin, abdominal skin of deceased patients, or from the forearms or hands of patients undergoing plastic surgery. The biopsies from surgical patients were pieces of normal skin removed in the course of repairing nongenetic conditions. Cultures were maintained and most growth experiments were done in minimal Eagle's medium (MEM) with 10% fetal calf serum (Grand Island Biological Company) in 5 % COz at 37°C. Only two lots of serum were used for all the experiments; both sera had been carefully selected for their support of rapid growth of normal human fibroblast cultures. All media were purchased in powdered form from Grand Island Biological Company and prepared in our laboratory. The cultures were grown initially in 25 om2 Falcon flasks and were subcultured using 0.005% purified trypsin (Worthington TR TCPK) in 75 cm2 flasks twice weekly (at 1:2 or 1:3) before reaching confluency. Each cell line was frozen at the second, third, or fourth passage in MEM with fetal calf serum and glycerin. Tests for mycoplasma (broth culture and by measurement of uridine phosphorylase activity (Levine, '72)) were done at the time of freezing each cell line and at every third week subsequent to thawing when a line was being carried for experiments. Cell counts were done on trypsinized cells suspended in phosphate-buffered saline, using a model B Coulter counter. Cells from the fifth through tenth subculture were used in all experiments. All nine normal control cultures initiated in our laboratory from 4/9/73 through 3/25/74 were used, in addition to five lines derived from unrelated patients with Fanconi anemia and two cultures obtained from parents of two unrelated FA patients.
Measurement of cell doubling time In each experiment at least one FA cell line and one normal control were studied simultaneously, to control for undetected variation of the medium, serum, or other factors affecting the rate of cell growth. Cells from actively growing sub-confluent cultures were plated into 20 cm2 Falcon tissue culture dishes at two or three different densities between 2 X 103 and 3.5 X 103 cells/cmZ, with at least six replicate plates for each cell line at each density. Care was taken to disperse the cells evenly on the plates, and plates that showed an uneven distribution of cells after attachment were discarded. Preliminary experiments were done in which normal or FA cell lines were plated onto 24-36 replicate plates at a single density and counts done at two different times each day beginning 18 hours after subcultivation, in order to determine two time points between which exponential cell growth was always observed. As a result of these preliminary experiments, measurements of cell doubling time were based on initial cell counts done 2G28 hours after plating and final cell counts done 4 2 4 3 hours later. For each cell density the initial and final cell numbers (the mean of three or four plates) were used to calculate the mean doubling time which is given by the formula DT (doubling time) = (log 2) (time lapse in hours) log (final cell count) -log (initial cell count).
We studied the variation in cell doubling time that occurs when control or FA cell lines, grown previously in MEM, are transferred to an enriched medium (McCoy's 5A, RPMI 1640) during the growth experiment. Sixteen replicate plates were prepared at a single cell density and three hours after subculturing the MEM was removed. Half the plates were refed with MEM and the remainder with the enriched medium. Initial and final counts were done as described above. Autor adiograp hy Cells were plated on coverslips at an initial density of 5 X 103 cells/cmZ. Thymidine (methyl-3H; 20 Cilmmole; New England Nuclear) was added at 1 pCi/ml for one hour for pulse labeling at 2, 22, 34 and 45 hours after subcultivation, and at 0.1 pCi/
23 1
GROWTH OF CULTURED FANCONI ANEMIA CELLS
ml immediately after cell attachment for continuous labeling. At the specified times (up to 99 hours for continuous labeling) cells were rinsed with phosphate buffered saline and fixed with Carnoy's. The air-dried coverslips were stored at 4 "C for three days after dipping in Kodak NTB-2 emulsion, developed in Kodak D19 and stained with Giemsa stain for one minute. At each time point 500 cells were counted for each cell line. RESULTS
For Fanconi anemia cultures 24 t 2.2 days were required from the time the biopsy was initiated to the first subculture, while the cultures from the parents required 19.7 t 2.7 days and the normal controls 20.1 -e 0.9 days. In all experiments exponential cell growth had begun 24 hours after subcultivation. The logarithm of the cell number increased linearily with time over the interval from 24 to 72 hours after subcultivation. Growth curves for one normal and one FA cell line are shown in figure 1. The mean cell doubling times for nine normal controls, five FA homozgous and two FA heterozygous cell lines are shown in table 1. In no experiment did an FA cell
-
30
-
0
0 X I
K W
m 10-
5= i w
0
7-
6: 5 4-
TABLE 1
Mean doubling times for cell cultures derived from normal controls, from patients with Fanconi anemia, and from parents of patients with Fanconi anemia Doubling time Cell line
Controls c2 c3 c4 c5 C8 C18 c20 c22 C23
Age at time of biopsy
32 23 25 2 24 21 3 29 4
Fanconi anemia homozygotes c12 6 8 C17 c19 6 C25 14 B68 6 Fanconi anemia heterozygotes C13 31 C24 52 1
Sex
F M F M
Race
W
B B
Skin site
Abdomen Wrist Wrist Hand Hand Forearm Abdomen Hand Hand
22.2 f 0.6 (4) I 19.7 f 0.5 (6) 24.7k0.5 (9) 22.2 f 0.7 (3) 24.2 % 1.0 (4) 24.8 f2.0 (2) 20.0 f 0.7 (4) 24.0 f 0.7 (6) 22.0 f 0.5 (3) 30.6 f 0.8 (7) 30.7 -t 1.6 (6) 32.1 & 1.3 (8) 30.9 f 1.2 (4) 26.2 k 1.2 (3) 25.5 % 1.0 (4) 26.3 f 1.7 (5)
F
W W W W B W
M M
B
F F
W W
M
B
Fore arm Forearm Abdomen Forearm Interscapula
M M
W W
Forearm Forearm
M M M M
W
mean f S.E.
The number in parentheses i s the number of experiments done with each cell line
232
EUGENE ELMORE AND MICHAEL SWIFT
growth (table 2) than normal controls. With continuous labeling the two control cultures exhibited 89 and 93 percent labeled cells at 45 hours, and 97 percent at 70 or 99 hours. The Fanconi anemia cultures had 72 and 75 percent labeled cells after 45 hours of continuous labeling, and 85 to 92 percent at 70 and 99 hours. The mean cell doubling time for the FA heterozygous cells (26.0 f 1.0 hours), was less than that of the FA homozygous cells (p
