Differences in the Incorporation of Thymidine into DNA of Normal and Cystic Fibrosis Fibroblasts In Vitro S . C. BARRANC0,2 W. E. BOLTON,2 B. R. HAENELT,2 c. W. ABELL3 Division of Cell Biology, Division of Biochemistry, Department of Human Biological Chemistry and Genetics, The University of Texas Medical Branch, Galueston, Texas 77550 AND
ABSTRACT Although similar fractions of cells were in the S phase of the cell cycle, normal human skin fibroblasts were shown to incorporate more than twice the JHTdR into their DNA in vitro than did cells obtained from individuals with cystic fibrosis (CF). Obligate heterozygotes incorporated a n intermediate amount of the DNA precursor. Studies were initiated to determine the basis of the differential incorporation of 3HTdR among the genotypes. A n analog of thymidine, BUdR, produced varied effects on the growth kinetics of the three genotypes. The growth of cells in BUdR resulted in a 50% increase i n the population doubling times of all three genotypes, and caused the cell morphology to change from a spindle shape to one in which the cells became broadened and flat, with numerous cytoplasmic projections extending for distances of several cell diameters. The activities of thymidine kinase and the participation of the exogenous and de nouo pathways in the synthesis of TMP were found to be approximately the same in all three genotypes. The data suggest that a n alteration in the transport of thymidine into the cells may account for the differences in TdR incorporation into DNA, and this may be associated with other changes i n cystic fibrosis that are apparently membrane associated.
During the characterization of the in vitro cell kinetics and growth properties of human skin diploid fibroblasts (Bolton and Barranco, '75), we observed that normal cells incorporated more than twice the amount of tritiated thymidine (3HTdR) into their DNA than did cells derived from individuals with cystic fibrosis (CF). Furthermore, cells derived from parents of CF patients (obligate heterozygotes) incorporated an intermediate amount of the DNA precursor. In contrast, when cell numbers and growth fractions were examined, no differences in cell kinetics or growth properties could be detected among the genotypes, in experiments where the CF, the heterozygote, and normal cells were matched according to their in vitro culture age (Bolton and Barranco, '75). Studies were initiated to determine the basis of the differential incorporation of 3HTdR among the genotypes. In this study, we compared the effects of bromodeoxyuridine, the activities of thymidine kinase, and the participation of the exogenous and de nono pathways in the synthesis of TMP in the three different types of cells. J.
CELL. PHYSIOL.,
88: 3 3 4 2
METHODS
Cell and culture techniques Skin fibroblasts derived from punch biopsies were grown in Ham's F-10 medium (2) supplemented with 10% fetal calf serum, 100,000 U streptomycin (Pfizer Laboratories, New York City), and 100,000 U polymixin B sulfate (Nutritional Biochemicals Corp., Cleveland, Ohio). All experiments utilized the same lot of fetal calf serum (Grand Island Biological Co., Grand Island, New York). Cells were maintained at 37°C in an atmosphere of 5% COz and 95% air. Cells are routinely tested for the presence of PPLO and have never been found to be contaminated.
Cell kinetics All cell lines have been previously characterized with respect to age, subculture number, population doubling time, growth fraction and cell cycle time (Bolton and Barranco, '75). The population doubling times presented in this report were obtained by plating 1 0 5 cells in petri dishes Received Aug. 6, '75. Accepted Sept. 26, '75. 1 Grant support DHEW 5P01 AM 17040(02) (Program Projects B and F).
33
34
S. BARRANCO, W. BOLTON. B. HAENELT AND C. ABELL
(100 x 20 mm) containing 20 ml of growth medium (controls) or with growth medium supplemented with 25 pg 5’-bromo2’-deoxyuridine (BUdR). Cells were counted on an inverted microscope with the aid of a reticule located in one ocular. The reticule was aligned with a counting grid attached to the petri dish, and the doubling time was obtained from the slope of the plot of increasing cell numbers versus time. The doubling times were determined during each experiment reported in this paper, but are presented only in figures 1 and 2.
lncorporation of “HTdRand “CdUR Cells of each line (3 X lo5) were placed in replicate plates containing 5 m l of growth medium. All plates were incubated in 37”C until sampled. At each sample time two plates of each cell line were pulse labeled for ten minutes with 2 pCi/ml of 3HTdR (specific activity 1.9 Ci/mM) or 0.5 pCi/ml of l4CdUR(specific activity 54.5 mCi/mM). All radiochemicals were purchased from Schwarz/Mann, Orangeburg, New York. Following the pulse label, the cells were rinsed twice with Pucks solution A and removed from the plates by trypsinization. The cells were rinsed with 5 ml physiological saline (0.85 g % ) and centrifuged at 1200g (4 minutes). The cell pellets were resuspended in saline, a cell count made and lo5 cells from each sample placed into tubes containing 7 ml of ice cold 10% trichloroacetic acid (TCA) for ten minutes. The TCA precipitates were collected on Gelman filters (0.45 p ) , placed into vials containing the liquid scintillation cocktail and counts were recorded using the Packard liquid scintillation spectrometer. Separate aliquotes of cells from the same samples were prepared for autoradiography using liquid emulsion (Ilford K5) au tor adiogr aphic techniques (Bolton and Barranco, ’75). The autoradiographic slides were maintained in a lightproof box at 4°C for seven days and developed in Kodak D19 developer. The labeling index (LI) was determined for each sample.
Thymidine hinase assay Fibroblasts (2.5 X 105 cells) in exponential and plateau phases of growth were used for each assay. All procedures were carried out at 5” unless otherwise noted.
The cells were rinsed twice in Pucks solution A and once in sonication fluid (0.03 M Tris, 2.50mM mercaptoethanol and 0.04 M thymidine). After each rinse, the cells were centrifuged at 400g for ten minutes. The pellet was finally resuspended in 0.2 ml sonication fluid and exposed to ultrasonic vibrations for five seconds. The sonicate was centrifuged at 50,000 rpm for 30 minutes in a Beckman model L ultracentrifuge with a 50.1 rotor. The supernatant contained all of the thymidine kinase (TK) activity. Fifty p1 of the substrate mixture (5 x M ATP, 1 X M MgC12, 6 X 10-3 M 3-PGA and 0.20 mls of 100 pCi/ml 3HTdR, specific activity 1.9 CilmM) were added to conical centrifuge tubes containing 50p1 of the enzyme solution. The tubes were incubated at 37°C for ten minutes and the enzyme reaction was terminated by boiling for two minutes. The enzyme reaction mixture was centrifuged (400 g, 10 minutes) to remove the precipitated protein, and 20p1 of the supernatant was spotted on each of two DE81 cellulose discs. One disc was immediately placed into a vial containing lOml fluor. After the second disc was dried, it was placed into a beaker containing 1Omls of 95% ethyl alcohol and swirled intermittently for five minutes to remove excess 3HTdR. The disc was removed and placed into another beaker of ethanol and treated in the same manner. Each disc was removed from the alcohol, allowed to air dry, placed into fluor and counted in a liquid scintillation counter. The percent conversion of TdR to TMP was calculated from the ratio of counts in the washed and unwashed samples. The TK activity was determined using a modification of the procedures reported by Furlong (‘63) and Stubblefield et al. (’67). RESULTS
Growth kinetics The data reported here were obtained from four normal human skin fibroblast cell lines (no previous history of CF), five CF homozygous recessives (diagnosed as having CF) and three obligate heterozygate cell lines (parents of CFs). The experiments were performed on culture-age matched lines ranging from the fourth to the seventh subculture periods. The p o p -
DIFFERENCES IN THYMIDINE INCORPORATION
lation doubling times under these conditions for the three genotypes averaged 28 hours (normals), 29 hours (heterozygotes), and 31 hours (CF). The growth fractions were 9 2 % , 8 9 % and 85% for the normals, heterozygotes and CFs, respectively. The cell cycle time for all genotypes ranged from 18 to 20 hours. Differential incorporation of 3HTdR It can be seen in figure 1 A that all three genotypes had similar population doubling times. After an initial lag, the cell number began increasing exponentially, reaching a plateau in 90 to 100 hours. The fraction of 3HTdR-labeled cells obtained from autoradiographs (fig. 1B) was low during the initial lag period of growth, but increased sharply to 45% as the cells began exponential growth. As the cells approached plateau phase, the fraction of 3HTdR-labeled cells began decreasing and reached a minimum of 4 % at 70 hours, and remained low through plateau phase. Although no differences in fractions of labeled cells were observed between genotypes, there was a noticeable reduction in the grain counts over labeled nuclei of cells from CF and heterozygous individuals when compared to normals. The decreased grain counts suggested that the incorporation of 3HTdR into D N A was reduced and this interpretation is supported by the data shown in figure 1C. During the lag period and in plateau phase the 3HTdR counts were low and no significant differences in 3HTdR incorporation was noted between genotypes. As the cells entered exponential growth, however, the counts increased sharply and the incorporation of 3HTdR into cells from CF and heterozygotes was 1.5-2 times less than in normal cells. There was a significant difference between the cpmlcell in CF and in normal lines from the twentieth hour through the forty-fifth hour; and between CF and heterozygous lines from the twentieth hour through the fourtieth hour. The incorporation of 3HTdR in both the CF and heterozygous lines was significantly less than in normals in the thirty-eighth and forty-fifth hour samples. Effects of B U d R on cell growth Since a differential incorpoation of TdR was observed between the three geno-
35
Differentiol Uptake of H’TdR in Normal, Heterozygote and CF Humon Fibrobiosts
10o’i
J
$
50-
zo
40-
,
5
:302010-
0 10
30
50
70
90
110
130
HOURS
Fig. 1 Differences in the incorporation of3HTdR in normal, 0 ; obligate heterozygotes, X; a n d CF fibroblasts, 0 during lag, exponential and plateau phases of growth i n vitro. A, growth kinetics; B, labeling indices; C, incorporation of 3HTdR into DNA.
types during exponential growth (fig. lC), the cells were treated with BUdR, an analog of TdR, to determine and compare the effects on cell growth. It can be seen in figures 2A,B that cells from CF and hetero-
S . BARRANCO, W. BOLTON, B. HAENELT AND C . ABELL
36 500
Photographs taken of the same cells after the tenth day in either BUdR or regular growth medium are shown in figure 3. The same numbers of cells from each genotype were plated at the beginning of the experiment and since the cell lines had similar doubling times, the cell densities seen in figure 3A, B, and C were essentially equaI. However, the ceII densities in figures 3D, E, and F varied considerably; the BUdR treated CF population (fig. 3D) contained more cells than the heterozygous population (fig. 3E), and very few cells remained in the normal population (fig. 3F) after a ten day treatment with BUdR (25 kg/ml). In addition to BUdR induced changes in population densities the cell morphology became broadened and flat, and contained numerous cytoplasmic projections stretching distances of several cell diameters. When the cells from CF patients were refed with fresh growth medium they began to grow again, and the morphology reverted to the original spindle shape.
~
100.
70.
50
.
30 -
,
zoo< LL W
m
I00 -
3
z
J
J
50
I
w
0
30
~
, -
500
200
-
100
.
70.
50.
30 I
0
20
40
60
80
100
I20
HOURS
Fig. 2 Effects of BUdR on the doubling times of CF (A) heteros (B) or normal fibroblasts (C) growing i n vitro i n Ham's F10 Medium.
'YgoUs required 1'6 and 2'3 times longer to double in the presence of BUdR than in fresh growth medium. Normal cells, which incorporated more TdR than did cells from CF's and heterogygates, were affected the most by the pres. Of BUdR in the medium and were able to double O d y once before cell division stopped (fig. 2C).
Comparison of incorporation of " T d R and '4Cd U R into DNA Since the differences in incorporation of 3HTdR in the three genotypes may simply reflect a shift in the utilization of the exogenous pathway to the de novo pathway for the synthesis of TMP, a comparison between the utilization of 3HTdR and 14CdUR was obtained. It can be seen in figure 4 that the incorporation of "TdR increased in all three genotypes as the cells progressed into exponential growth (cell kinetics data not shown) and decreased as the cells approached and entered plateau phase (open symbols in figs. 4A,B,C). The same general Dattern oFincorwration was obtainGd with 14CdUR. The coints rose as the cells began expotentid growth and decreased as the cells entered plateau phase (closed symbols in figs. 4A,B,C). The I4CdUR counts were slightly higher and remained elevated longer in the normal cells (fig. 4A) than in cells from heterozygotes or CF's (figs. 4B,C), Fig. 3 The effects of a 10-day treatment of BUdR on cell population density. A, CF cells; 6 heteros; and C, normal fibroblasts grown i n Ham's F10 medium. D, CF cells; E, heteros; and F, normal fibroblasts maintained i n 25 gg/ml BUdR. Magnification X 50.
DIFFERENCES IN THYMIDINE INCORPORATION
Figure 3
37
38
S. BARRANCO, W. BOLTON, B. HAENELT AND C. ABELL 40
1
30 -
20 10
~
* 10
y o 0
*
*
l
30
. 50
'
'
.
70
1 . 110
l
90
al
t
40
*
I
-7
B. HETERO
w
_1
-6
:
I a
H3
u 10-
2
Q I
c'4
-I W
\ z
-3 -2 u -1
no1
0
' 10
'
'
30
'
'
'
50
'
'
70
90
'
1
t x
1
%
110
2
I R
40 t
-7 -6 -5
c. CF
-4
-3 -2 -1
0
10
30
50
70
110
90
D. RATIO
OF
8
s
v 6
R 0 .c
a"
4 2 0
0 10
30
70
50
90
110
HOURS Fig. 4 A comparison of the incorporation of 3HTdR and '4CdUR into DNA of normal, A, Hetero, B; or CF, C, fibroblasts i n vitro, and the ratio of the 3Hp4C counts, D.
The ratios of the "TdR to '4CdUR cpmlcell are plotted in figure 4D. The 3H/W ratio increased as the cells moved into exponential growth and remained elevated until the cells progressed into plateau phase. The fact that the ratio in-
creases indicates that the exogenous pathway is utilized more extensively than the de novo pathway during the growth phase. The only difference in the values of the ratios occurs at 65 hours where the control ratio is slightly, but significantly, high-
DIFFERENCES IN THYMIDINE INCORPORATION
er than that for cells from CFs and heterozygotes.
T hy m i d i n e kinase assay TK activity in cells from each genotype was measured by the ability to convert 3HTdR to TMP. The kinase activities for each cell type were low (fig. 5 ) while the cells were in the lag phase (cell kinetics data not shown). The entry of the cells into exponential growth at 25 hours was accompanied by approximately a 6-fold increase in TK activity. The enzyme activity decreased gradually over the next 20 hours as the cells approached plateau phase. No statistical differences in enzyme activities were observed between the genotypes except at the thirtieth hour where the activity in cells from CF patients was slightly lower. DISCUSSION
Cells normally acquire the ability to incorporate TdR as they enter the DNA synthesis or S phase of the cell cycle (Howard and Pelc, '53). In asynchronous populations of cells, such as those studied here, the fraction of cells in S phase (LI) at any one time can be determined easily by counting labeled cells (autoradiographs) after exposure to 3HTdR. Any cell having
04 0
I 10
20
30
40
50
HOURS Fig. 5 The percent of conversion of TdR to TMP by thymidine kinase i n normal, 0 ; hetero, X; and CF fibroblasts, 0 in vitro.
39
at least ten grainslnucleus was considered labeled in our studies. During the lag phase of growth, occurring shortly after the cells were plated, the LI was low (fig. 1B). However, as the cells began exponential growth, the LI increased and was the same in each genotype. If the cells are not refed, the media becomes depleted of nutrients essential for growth (Tobey and Ley, '70), and in the case of the nontransformed fibroblasts, the cells also become contact inhibited. Under these "unfed' conditions, the cells progress through the cell cycle and eventually become blocked in a nondividing state resembling the Go compartment in vivo (Barranco et al., '73; Hahn et al., '68; Mendelsohn, '62; Tobey and Ley, '70). The LI decreases (fig. 1B) as the cells enter plateau phase and the cells are characterized as having a G I quantity of DNA (Tobey and Ley, '70). The rise and fall in the LI as cells progress from lag to exponential growth and plateau phase is a reproducible phenomenon and has been reported in a number of in vitro studies (Barranco et al., '73; Hahn et al., '68; Tobey and Ley, '70). Although the LI were similar for the three genotypes, (indicating similar fractions of cells were in S phase) the grain counts in the hetero and CF cells were much lower than in normal cells suggesting less incorporation of :'HTdR. This is supported by the observation that :3HTdR incorporation into DNA is substantially decreased (fig. 1C). Although no differences in incorporation of :{HTdRare seen in cells from each of the genotypes during the lag or plateau phases, as the cells enter exponential growth the cpmlcell are approximately two times higher in normals than in the CF lines. To further explore this phenomenon, an analog of thymidine, BUdR, was used to determine whether it too would be incorporated differentially and produce varied effects on the growth kinetics of the three genotypes. Cells from all three genotypes grown continuously in the presence of BUdR demonstrated a 50% increase in the lengths of the population doubling times (fig. 2). The cells from CFs and heterozygotes continued to divide at the reduced rate and doubled at least twice during the experiment; however, the normal cells doubled only one time before cessa-
40
S. BARRANCO, W . BOLTON, B. HAENELT AND C. ABELL
tion of division. The effects of BUdR on normal human fibroblasts are in excellent agreement with the effects of BUdR reported on HeLa cells (Toliver et al., '68). Treatment with BUdR results in replacement of thymidine by 5-bromouracil, and the rate of cell division is reduced, due primarily to a n increase in length of S phase (Toliver et al., '68). Photographs taken of these cells after ten days in BUdR (figs. 3D,E,F) show the dramatic effects that BUdR had on the growth and morphology of the cells. Few of the normal cells remained in culture (fig. 3F), whereas cultures derived from CF and heterozygotes contained considerably more cells. Although the CF cells were affected least by BUdR (fig. 3D), some reduction in population density was obtained when compared to cells maintained in regular medium for the same length of time (fig. 3A). The reduction in cell population densities is probably the result of the BUdR influence on doubling times (fig. 2), and cell viability (at least in the case of the normal cells). The morphology of the cells treated with BUdR changed from a spindle shape to one in which the cells became broadened and flat with numerous cytoplasmic projections extending for distances of several cell diameters. However, once the BUdR-treated cells obtained from CF patients were refed with fresh growth medium, they began to grow again, and the morphology reverted to the original spindle shape. Since TMP can be synthesized by either the exogenous or de novo route, the possibility that the participation of these pathways varies in the different gentotypes was considered. When WdUR was used as a precursor, however, incorporation into DNA was approximately the same in all three genotypes (fig. 4). Thus, the lowered incorporation of :'HTdR in CF cells was not accompanied by a corresponding increase in 14CdURincorporation into DNA. These observations, in view of the similar growth rates of the cell types, suggested that the basis of the difference in TdR incorporation resided in either alterations of thymidine kinase activities or in the transport of thymidine into the cells. Thymidine enters the nucleotide pool by the exogenous pathway after being converted to TMP by thymidine kinase. Consequently T K activity may be lower in
hetero and CF cells than in normal fibroblasts. However, upon examination the kinase activities were found to be equal in all three genotypes with the exception of the 30-hour sample point. The TK activity was low during the initial lag phase of growth and increased to a high level with the onset of rapid growth (fig. 5). Bukovsky and Roth ('65) have presented similar data correlating variability in TK activities with the growth state in several animal hepatomas; the highest activity being found in tumors with the fastest growth rate. However, by matching the cells used in our studies according to culture age, growth fraction, cell cycle times and doubling times, we have eliminated such variability in kinase activity due to growth kinetics. In addition, all efforts to synchronize the CF fibroblasts by the excess thymidine technique (Humphrey et al., '70) were unsuccessful, probably because the TdR did not enter the cells readily; whereas, the normal cells were easily synchronized (data not shown). The fact that the ability to convert TdR to TMP is normal, and that the ratios between the de novo and exogenous pathways are the same in all of the cell lines tested in our study suggests that the lower acid insoluble SH counts observed in the cells from CF's and heterozygotes were the result of impairment of the thymidine transport system. A reduction in thymidine transport may be associated with other changes in cystic fibrosis, which apparently are membrane associated, such as abnormal electrolyte balance (Shwachman and Mahmoodian, '67) and increased mucopolysaccharide content (Danes et al., '73; Matalon and Dorfman, '69) and altered glycoprotein synthesis.4 Moreover it has recently been observed5 that the membrane bound Na+/K+/Mg++ and Ca++ ATPase of normal erythrocytes are profoundly depressed when exposed to media from CF cells, implying that another transport system is depressed. Studies on the relationship between thymidine transport and these parameters are continuing in our laboratories. ACKNOWLEDGMENTS
We wish to thank Drs. Barbara H. Bow4
5
R. A. Novak and C. W. Abell (in preparation). G. M. Fuller, F. Baglia and B. H. Bowman ( i n pre-
paration).
DIFFERENCES IN THYMIDINE INCORPORATION
man, B. R. Brinkley and G . M. Fuller for helpful discussions; and Mrs. Peggy Anderson and Mr. David Hom for their assistance in preparing the manuscript. LITERATURE CITED Barranco, S. C., J. K. Novak and R. M. Humphrey 1973 Response of Mammalian Cells Following Treatment with Bleomycin and 1,3-Bis(2-chloroethyl)-1-nitrosourea during Plateau Phase. Cancer Res., 3 3 : 6 9 1 4 9 4 . Bolton, W. E., and S. C. Barranco 1975 Characterization of the Cell Kinetics and Growth Properties of Cystic Fibrosis Diploid Fibroblasts In Vitro. Am. J. Hum. Genet., 27: 3944409. Bukovsky, J., and J. S. Roth 1965 Some Factors Affecting the Phosphorylation of Thymidine by Transplantable Rat Hepatomas. Cancer Res., 25: 358-364. Danes, B . S., S. D. Holley and J. F. Watkins 1973 Cystic Fibrosis-a hybrid model. Am. J. Hum. Genet., 25: 323-327. Furlong, N. Burr 1963 A Rapid Assay for Nucleotide Kinases Using CI4or H3-Labeled Nucleotides. Analytical Biochem., 5: 515522. Hahn, G. M., J. R. Stewart, S. J. Yang and V. Parker 1968 Changes in Cell Dynamics and Modifications of the Cell Cycle with the Period of Growth. Exptl. Cell Res., 49: 285-292. Howard, A,, and S. R. Pelc. 1953 Synthesis of deoxyribonucleic acid in normal and irradiated
41
cells and its relation to chromosome breakage. Heredity, 6(suppl.) :261-273. Humphrey, R. M., D. Steward and B. Sedita 1970 DNA Strand Scission and Rejoining in Mammali a n Cells. In: Genetic Concepts in Neoplasia. The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston. The Williams and Wilkins Co., Baltimore, pp, 5 7 0 5 9 2 , Matalon, R., and A. Dorfman 1969 Acid mucopolusaccharides in cultured human fibroblasts. Lancet, 2: 838441. Mendelsohn, M. L. 1962 Autoradiographic analysis of cell proliferation i n spontaneous breast cancer of the C3H mouse. 111. The growth fraction. J. Natl. Cancer Inst., 2 8 : 1015-1029. Schwachman, H., and A. Mahmoodian 1967 Pilocarpine iontophoresis sweat testing results of seven years experience. 4th International Conference on Cystic Fibrosis of the Pancreas (mucoviscidosis). Mod. Probl. Pediat., 10: 158-161. Stubblefield E.,and S. Murphree 1967 Synchronized Mammalian Cell Cultures. 11. Thymidine Kinase Activity in Colcemid Synchronized Fibroblasts. Exptl. Cell Res., 48: 6 5 2 4 5 6 . Tobey, R. A,, and K. D. Ley 1970 Regulation of Initiation of DNA Synthesis of Chinese Hamster Cells. I. Production of Stable, Reversible, G1-arrested Populations i n Suspension Culture. J. Cell Biol., 46: 151-157. Toliver, A., E. H. Simon and P. T. Gilham 1968 On the Mechanism of 5-Bromouracil Inhibition of DNA Synthesis and Cell Division. Exptl. Cell Res., 53: 506-518.
ABSTRACT Although similar fractions of cells were in the S phase of the cell cycle, normal human skin fibroblasts were shown to incorporate more than twice the JHTdR into their DNA in vitro than did cells obtained from individuals with cystic fibrosis (CF). Obligate heterozygotes incorporated a n intermediate amount of the DNA precursor. Studies were initiated to determine the basis of the differential incorporation of 3HTdR among the genotypes. A n analog of thymidine, BUdR, produced varied effects on the growth kinetics of the three genotypes. The growth of cells in BUdR resulted in a 50% increase i n the population doubling times of all three genotypes, and caused the cell morphology to change from a spindle shape to one in which the cells became broadened and flat, with numerous cytoplasmic projections extending for distances of several cell diameters. The activities of thymidine kinase and the participation of the exogenous and de nouo pathways in the synthesis of TMP were found to be approximately the same in all three genotypes. The data suggest that a n alteration in the transport of thymidine into the cells may account for the differences in TdR incorporation into DNA, and this may be associated with other changes i n cystic fibrosis that are apparently membrane associated.
During the characterization of the in vitro cell kinetics and growth properties of human skin diploid fibroblasts (Bolton and Barranco, '75), we observed that normal cells incorporated more than twice the amount of tritiated thymidine (3HTdR) into their DNA than did cells derived from individuals with cystic fibrosis (CF). Furthermore, cells derived from parents of CF patients (obligate heterozygotes) incorporated an intermediate amount of the DNA precursor. In contrast, when cell numbers and growth fractions were examined, no differences in cell kinetics or growth properties could be detected among the genotypes, in experiments where the CF, the heterozygote, and normal cells were matched according to their in vitro culture age (Bolton and Barranco, '75). Studies were initiated to determine the basis of the differential incorporation of 3HTdR among the genotypes. In this study, we compared the effects of bromodeoxyuridine, the activities of thymidine kinase, and the participation of the exogenous and de nono pathways in the synthesis of TMP in the three different types of cells. J.
CELL. PHYSIOL.,
88: 3 3 4 2
METHODS
Cell and culture techniques Skin fibroblasts derived from punch biopsies were grown in Ham's F-10 medium (2) supplemented with 10% fetal calf serum, 100,000 U streptomycin (Pfizer Laboratories, New York City), and 100,000 U polymixin B sulfate (Nutritional Biochemicals Corp., Cleveland, Ohio). All experiments utilized the same lot of fetal calf serum (Grand Island Biological Co., Grand Island, New York). Cells were maintained at 37°C in an atmosphere of 5% COz and 95% air. Cells are routinely tested for the presence of PPLO and have never been found to be contaminated.
Cell kinetics All cell lines have been previously characterized with respect to age, subculture number, population doubling time, growth fraction and cell cycle time (Bolton and Barranco, '75). The population doubling times presented in this report were obtained by plating 1 0 5 cells in petri dishes Received Aug. 6, '75. Accepted Sept. 26, '75. 1 Grant support DHEW 5P01 AM 17040(02) (Program Projects B and F).
33
34
S. BARRANCO, W. BOLTON. B. HAENELT AND C. ABELL
(100 x 20 mm) containing 20 ml of growth medium (controls) or with growth medium supplemented with 25 pg 5’-bromo2’-deoxyuridine (BUdR). Cells were counted on an inverted microscope with the aid of a reticule located in one ocular. The reticule was aligned with a counting grid attached to the petri dish, and the doubling time was obtained from the slope of the plot of increasing cell numbers versus time. The doubling times were determined during each experiment reported in this paper, but are presented only in figures 1 and 2.
lncorporation of “HTdRand “CdUR Cells of each line (3 X lo5) were placed in replicate plates containing 5 m l of growth medium. All plates were incubated in 37”C until sampled. At each sample time two plates of each cell line were pulse labeled for ten minutes with 2 pCi/ml of 3HTdR (specific activity 1.9 Ci/mM) or 0.5 pCi/ml of l4CdUR(specific activity 54.5 mCi/mM). All radiochemicals were purchased from Schwarz/Mann, Orangeburg, New York. Following the pulse label, the cells were rinsed twice with Pucks solution A and removed from the plates by trypsinization. The cells were rinsed with 5 ml physiological saline (0.85 g % ) and centrifuged at 1200g (4 minutes). The cell pellets were resuspended in saline, a cell count made and lo5 cells from each sample placed into tubes containing 7 ml of ice cold 10% trichloroacetic acid (TCA) for ten minutes. The TCA precipitates were collected on Gelman filters (0.45 p ) , placed into vials containing the liquid scintillation cocktail and counts were recorded using the Packard liquid scintillation spectrometer. Separate aliquotes of cells from the same samples were prepared for autoradiography using liquid emulsion (Ilford K5) au tor adiogr aphic techniques (Bolton and Barranco, ’75). The autoradiographic slides were maintained in a lightproof box at 4°C for seven days and developed in Kodak D19 developer. The labeling index (LI) was determined for each sample.
Thymidine hinase assay Fibroblasts (2.5 X 105 cells) in exponential and plateau phases of growth were used for each assay. All procedures were carried out at 5” unless otherwise noted.
The cells were rinsed twice in Pucks solution A and once in sonication fluid (0.03 M Tris, 2.50mM mercaptoethanol and 0.04 M thymidine). After each rinse, the cells were centrifuged at 400g for ten minutes. The pellet was finally resuspended in 0.2 ml sonication fluid and exposed to ultrasonic vibrations for five seconds. The sonicate was centrifuged at 50,000 rpm for 30 minutes in a Beckman model L ultracentrifuge with a 50.1 rotor. The supernatant contained all of the thymidine kinase (TK) activity. Fifty p1 of the substrate mixture (5 x M ATP, 1 X M MgC12, 6 X 10-3 M 3-PGA and 0.20 mls of 100 pCi/ml 3HTdR, specific activity 1.9 CilmM) were added to conical centrifuge tubes containing 50p1 of the enzyme solution. The tubes were incubated at 37°C for ten minutes and the enzyme reaction was terminated by boiling for two minutes. The enzyme reaction mixture was centrifuged (400 g, 10 minutes) to remove the precipitated protein, and 20p1 of the supernatant was spotted on each of two DE81 cellulose discs. One disc was immediately placed into a vial containing lOml fluor. After the second disc was dried, it was placed into a beaker containing 1Omls of 95% ethyl alcohol and swirled intermittently for five minutes to remove excess 3HTdR. The disc was removed and placed into another beaker of ethanol and treated in the same manner. Each disc was removed from the alcohol, allowed to air dry, placed into fluor and counted in a liquid scintillation counter. The percent conversion of TdR to TMP was calculated from the ratio of counts in the washed and unwashed samples. The TK activity was determined using a modification of the procedures reported by Furlong (‘63) and Stubblefield et al. (’67). RESULTS
Growth kinetics The data reported here were obtained from four normal human skin fibroblast cell lines (no previous history of CF), five CF homozygous recessives (diagnosed as having CF) and three obligate heterozygate cell lines (parents of CFs). The experiments were performed on culture-age matched lines ranging from the fourth to the seventh subculture periods. The p o p -
DIFFERENCES IN THYMIDINE INCORPORATION
lation doubling times under these conditions for the three genotypes averaged 28 hours (normals), 29 hours (heterozygotes), and 31 hours (CF). The growth fractions were 9 2 % , 8 9 % and 85% for the normals, heterozygotes and CFs, respectively. The cell cycle time for all genotypes ranged from 18 to 20 hours. Differential incorporation of 3HTdR It can be seen in figure 1 A that all three genotypes had similar population doubling times. After an initial lag, the cell number began increasing exponentially, reaching a plateau in 90 to 100 hours. The fraction of 3HTdR-labeled cells obtained from autoradiographs (fig. 1B) was low during the initial lag period of growth, but increased sharply to 45% as the cells began exponential growth. As the cells approached plateau phase, the fraction of 3HTdR-labeled cells began decreasing and reached a minimum of 4 % at 70 hours, and remained low through plateau phase. Although no differences in fractions of labeled cells were observed between genotypes, there was a noticeable reduction in the grain counts over labeled nuclei of cells from CF and heterozygous individuals when compared to normals. The decreased grain counts suggested that the incorporation of 3HTdR into D N A was reduced and this interpretation is supported by the data shown in figure 1C. During the lag period and in plateau phase the 3HTdR counts were low and no significant differences in 3HTdR incorporation was noted between genotypes. As the cells entered exponential growth, however, the counts increased sharply and the incorporation of 3HTdR into cells from CF and heterozygotes was 1.5-2 times less than in normal cells. There was a significant difference between the cpmlcell in CF and in normal lines from the twentieth hour through the forty-fifth hour; and between CF and heterozygous lines from the twentieth hour through the fourtieth hour. The incorporation of 3HTdR in both the CF and heterozygous lines was significantly less than in normals in the thirty-eighth and forty-fifth hour samples. Effects of B U d R on cell growth Since a differential incorpoation of TdR was observed between the three geno-
35
Differentiol Uptake of H’TdR in Normal, Heterozygote and CF Humon Fibrobiosts
10o’i
J
$
50-
zo
40-
,
5
:302010-
0 10
30
50
70
90
110
130
HOURS
Fig. 1 Differences in the incorporation of3HTdR in normal, 0 ; obligate heterozygotes, X; a n d CF fibroblasts, 0 during lag, exponential and plateau phases of growth i n vitro. A, growth kinetics; B, labeling indices; C, incorporation of 3HTdR into DNA.
types during exponential growth (fig. lC), the cells were treated with BUdR, an analog of TdR, to determine and compare the effects on cell growth. It can be seen in figures 2A,B that cells from CF and hetero-
S . BARRANCO, W. BOLTON, B. HAENELT AND C . ABELL
36 500
Photographs taken of the same cells after the tenth day in either BUdR or regular growth medium are shown in figure 3. The same numbers of cells from each genotype were plated at the beginning of the experiment and since the cell lines had similar doubling times, the cell densities seen in figure 3A, B, and C were essentially equaI. However, the ceII densities in figures 3D, E, and F varied considerably; the BUdR treated CF population (fig. 3D) contained more cells than the heterozygous population (fig. 3E), and very few cells remained in the normal population (fig. 3F) after a ten day treatment with BUdR (25 kg/ml). In addition to BUdR induced changes in population densities the cell morphology became broadened and flat, and contained numerous cytoplasmic projections stretching distances of several cell diameters. When the cells from CF patients were refed with fresh growth medium they began to grow again, and the morphology reverted to the original spindle shape.
~
100.
70.
50
.
30 -
,
zoo< LL W
m
I00 -
3
z
J
J
50
I
w
0
30
~
, -
500
200
-
100
.
70.
50.
30 I
0
20
40
60
80
100
I20
HOURS
Fig. 2 Effects of BUdR on the doubling times of CF (A) heteros (B) or normal fibroblasts (C) growing i n vitro i n Ham's F10 Medium.
'YgoUs required 1'6 and 2'3 times longer to double in the presence of BUdR than in fresh growth medium. Normal cells, which incorporated more TdR than did cells from CF's and heterogygates, were affected the most by the pres. Of BUdR in the medium and were able to double O d y once before cell division stopped (fig. 2C).
Comparison of incorporation of " T d R and '4Cd U R into DNA Since the differences in incorporation of 3HTdR in the three genotypes may simply reflect a shift in the utilization of the exogenous pathway to the de novo pathway for the synthesis of TMP, a comparison between the utilization of 3HTdR and 14CdUR was obtained. It can be seen in figure 4 that the incorporation of "TdR increased in all three genotypes as the cells progressed into exponential growth (cell kinetics data not shown) and decreased as the cells approached and entered plateau phase (open symbols in figs. 4A,B,C). The same general Dattern oFincorwration was obtainGd with 14CdUR. The coints rose as the cells began expotentid growth and decreased as the cells entered plateau phase (closed symbols in figs. 4A,B,C). The I4CdUR counts were slightly higher and remained elevated longer in the normal cells (fig. 4A) than in cells from heterozygotes or CF's (figs. 4B,C), Fig. 3 The effects of a 10-day treatment of BUdR on cell population density. A, CF cells; 6 heteros; and C, normal fibroblasts grown i n Ham's F10 medium. D, CF cells; E, heteros; and F, normal fibroblasts maintained i n 25 gg/ml BUdR. Magnification X 50.
DIFFERENCES IN THYMIDINE INCORPORATION
Figure 3
37
38
S. BARRANCO, W. BOLTON, B. HAENELT AND C. ABELL 40
1
30 -
20 10
~
* 10
y o 0
*
*
l
30
. 50
'
'
.
70
1 . 110
l
90
al
t
40
*
I
-7
B. HETERO
w
_1
-6
:
I a
H3
u 10-
2
Q I
c'4
-I W
\ z
-3 -2 u -1
no1
0
' 10
'
'
30
'
'
'
50
'
'
70
90
'
1
t x
1
%
110
2
I R
40 t
-7 -6 -5
c. CF
-4
-3 -2 -1
0
10
30
50
70
110
90
D. RATIO
OF
8
s
v 6
R 0 .c
a"
4 2 0
0 10
30
70
50
90
110
HOURS Fig. 4 A comparison of the incorporation of 3HTdR and '4CdUR into DNA of normal, A, Hetero, B; or CF, C, fibroblasts i n vitro, and the ratio of the 3Hp4C counts, D.
The ratios of the "TdR to '4CdUR cpmlcell are plotted in figure 4D. The 3H/W ratio increased as the cells moved into exponential growth and remained elevated until the cells progressed into plateau phase. The fact that the ratio in-
creases indicates that the exogenous pathway is utilized more extensively than the de novo pathway during the growth phase. The only difference in the values of the ratios occurs at 65 hours where the control ratio is slightly, but significantly, high-
DIFFERENCES IN THYMIDINE INCORPORATION
er than that for cells from CFs and heterozygotes.
T hy m i d i n e kinase assay TK activity in cells from each genotype was measured by the ability to convert 3HTdR to TMP. The kinase activities for each cell type were low (fig. 5 ) while the cells were in the lag phase (cell kinetics data not shown). The entry of the cells into exponential growth at 25 hours was accompanied by approximately a 6-fold increase in TK activity. The enzyme activity decreased gradually over the next 20 hours as the cells approached plateau phase. No statistical differences in enzyme activities were observed between the genotypes except at the thirtieth hour where the activity in cells from CF patients was slightly lower. DISCUSSION
Cells normally acquire the ability to incorporate TdR as they enter the DNA synthesis or S phase of the cell cycle (Howard and Pelc, '53). In asynchronous populations of cells, such as those studied here, the fraction of cells in S phase (LI) at any one time can be determined easily by counting labeled cells (autoradiographs) after exposure to 3HTdR. Any cell having
04 0
I 10
20
30
40
50
HOURS Fig. 5 The percent of conversion of TdR to TMP by thymidine kinase i n normal, 0 ; hetero, X; and CF fibroblasts, 0 in vitro.
39
at least ten grainslnucleus was considered labeled in our studies. During the lag phase of growth, occurring shortly after the cells were plated, the LI was low (fig. 1B). However, as the cells began exponential growth, the LI increased and was the same in each genotype. If the cells are not refed, the media becomes depleted of nutrients essential for growth (Tobey and Ley, '70), and in the case of the nontransformed fibroblasts, the cells also become contact inhibited. Under these "unfed' conditions, the cells progress through the cell cycle and eventually become blocked in a nondividing state resembling the Go compartment in vivo (Barranco et al., '73; Hahn et al., '68; Mendelsohn, '62; Tobey and Ley, '70). The LI decreases (fig. 1B) as the cells enter plateau phase and the cells are characterized as having a G I quantity of DNA (Tobey and Ley, '70). The rise and fall in the LI as cells progress from lag to exponential growth and plateau phase is a reproducible phenomenon and has been reported in a number of in vitro studies (Barranco et al., '73; Hahn et al., '68; Tobey and Ley, '70). Although the LI were similar for the three genotypes, (indicating similar fractions of cells were in S phase) the grain counts in the hetero and CF cells were much lower than in normal cells suggesting less incorporation of :'HTdR. This is supported by the observation that :3HTdR incorporation into DNA is substantially decreased (fig. 1C). Although no differences in incorporation of :{HTdRare seen in cells from each of the genotypes during the lag or plateau phases, as the cells enter exponential growth the cpmlcell are approximately two times higher in normals than in the CF lines. To further explore this phenomenon, an analog of thymidine, BUdR, was used to determine whether it too would be incorporated differentially and produce varied effects on the growth kinetics of the three genotypes. Cells from all three genotypes grown continuously in the presence of BUdR demonstrated a 50% increase in the lengths of the population doubling times (fig. 2). The cells from CFs and heterozygotes continued to divide at the reduced rate and doubled at least twice during the experiment; however, the normal cells doubled only one time before cessa-
40
S. BARRANCO, W . BOLTON, B. HAENELT AND C. ABELL
tion of division. The effects of BUdR on normal human fibroblasts are in excellent agreement with the effects of BUdR reported on HeLa cells (Toliver et al., '68). Treatment with BUdR results in replacement of thymidine by 5-bromouracil, and the rate of cell division is reduced, due primarily to a n increase in length of S phase (Toliver et al., '68). Photographs taken of these cells after ten days in BUdR (figs. 3D,E,F) show the dramatic effects that BUdR had on the growth and morphology of the cells. Few of the normal cells remained in culture (fig. 3F), whereas cultures derived from CF and heterozygotes contained considerably more cells. Although the CF cells were affected least by BUdR (fig. 3D), some reduction in population density was obtained when compared to cells maintained in regular medium for the same length of time (fig. 3A). The reduction in cell population densities is probably the result of the BUdR influence on doubling times (fig. 2), and cell viability (at least in the case of the normal cells). The morphology of the cells treated with BUdR changed from a spindle shape to one in which the cells became broadened and flat with numerous cytoplasmic projections extending for distances of several cell diameters. However, once the BUdR-treated cells obtained from CF patients were refed with fresh growth medium, they began to grow again, and the morphology reverted to the original spindle shape. Since TMP can be synthesized by either the exogenous or de novo route, the possibility that the participation of these pathways varies in the different gentotypes was considered. When WdUR was used as a precursor, however, incorporation into DNA was approximately the same in all three genotypes (fig. 4). Thus, the lowered incorporation of :'HTdR in CF cells was not accompanied by a corresponding increase in 14CdURincorporation into DNA. These observations, in view of the similar growth rates of the cell types, suggested that the basis of the difference in TdR incorporation resided in either alterations of thymidine kinase activities or in the transport of thymidine into the cells. Thymidine enters the nucleotide pool by the exogenous pathway after being converted to TMP by thymidine kinase. Consequently T K activity may be lower in
hetero and CF cells than in normal fibroblasts. However, upon examination the kinase activities were found to be equal in all three genotypes with the exception of the 30-hour sample point. The TK activity was low during the initial lag phase of growth and increased to a high level with the onset of rapid growth (fig. 5). Bukovsky and Roth ('65) have presented similar data correlating variability in TK activities with the growth state in several animal hepatomas; the highest activity being found in tumors with the fastest growth rate. However, by matching the cells used in our studies according to culture age, growth fraction, cell cycle times and doubling times, we have eliminated such variability in kinase activity due to growth kinetics. In addition, all efforts to synchronize the CF fibroblasts by the excess thymidine technique (Humphrey et al., '70) were unsuccessful, probably because the TdR did not enter the cells readily; whereas, the normal cells were easily synchronized (data not shown). The fact that the ability to convert TdR to TMP is normal, and that the ratios between the de novo and exogenous pathways are the same in all of the cell lines tested in our study suggests that the lower acid insoluble SH counts observed in the cells from CF's and heterozygotes were the result of impairment of the thymidine transport system. A reduction in thymidine transport may be associated with other changes in cystic fibrosis, which apparently are membrane associated, such as abnormal electrolyte balance (Shwachman and Mahmoodian, '67) and increased mucopolysaccharide content (Danes et al., '73; Matalon and Dorfman, '69) and altered glycoprotein synthesis.4 Moreover it has recently been observed5 that the membrane bound Na+/K+/Mg++ and Ca++ ATPase of normal erythrocytes are profoundly depressed when exposed to media from CF cells, implying that another transport system is depressed. Studies on the relationship between thymidine transport and these parameters are continuing in our laboratories. ACKNOWLEDGMENTS
We wish to thank Drs. Barbara H. Bow4
5
R. A. Novak and C. W. Abell (in preparation). G. M. Fuller, F. Baglia and B. H. Bowman ( i n pre-
paration).
DIFFERENCES IN THYMIDINE INCORPORATION
man, B. R. Brinkley and G . M. Fuller for helpful discussions; and Mrs. Peggy Anderson and Mr. David Hom for their assistance in preparing the manuscript. LITERATURE CITED Barranco, S. C., J. K. Novak and R. M. Humphrey 1973 Response of Mammalian Cells Following Treatment with Bleomycin and 1,3-Bis(2-chloroethyl)-1-nitrosourea during Plateau Phase. Cancer Res., 3 3 : 6 9 1 4 9 4 . Bolton, W. E., and S. C. Barranco 1975 Characterization of the Cell Kinetics and Growth Properties of Cystic Fibrosis Diploid Fibroblasts In Vitro. Am. J. Hum. Genet., 27: 3944409. Bukovsky, J., and J. S. Roth 1965 Some Factors Affecting the Phosphorylation of Thymidine by Transplantable Rat Hepatomas. Cancer Res., 25: 358-364. Danes, B . S., S. D. Holley and J. F. Watkins 1973 Cystic Fibrosis-a hybrid model. Am. J. Hum. Genet., 25: 323-327. Furlong, N. Burr 1963 A Rapid Assay for Nucleotide Kinases Using CI4or H3-Labeled Nucleotides. Analytical Biochem., 5: 515522. Hahn, G. M., J. R. Stewart, S. J. Yang and V. Parker 1968 Changes in Cell Dynamics and Modifications of the Cell Cycle with the Period of Growth. Exptl. Cell Res., 49: 285-292. Howard, A,, and S. R. Pelc. 1953 Synthesis of deoxyribonucleic acid in normal and irradiated
41
cells and its relation to chromosome breakage. Heredity, 6(suppl.) :261-273. Humphrey, R. M., D. Steward and B. Sedita 1970 DNA Strand Scission and Rejoining in Mammali a n Cells. In: Genetic Concepts in Neoplasia. The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston. The Williams and Wilkins Co., Baltimore, pp, 5 7 0 5 9 2 , Matalon, R., and A. Dorfman 1969 Acid mucopolusaccharides in cultured human fibroblasts. Lancet, 2: 838441. Mendelsohn, M. L. 1962 Autoradiographic analysis of cell proliferation i n spontaneous breast cancer of the C3H mouse. 111. The growth fraction. J. Natl. Cancer Inst., 2 8 : 1015-1029. Schwachman, H., and A. Mahmoodian 1967 Pilocarpine iontophoresis sweat testing results of seven years experience. 4th International Conference on Cystic Fibrosis of the Pancreas (mucoviscidosis). Mod. Probl. Pediat., 10: 158-161. Stubblefield E.,and S. Murphree 1967 Synchronized Mammalian Cell Cultures. 11. Thymidine Kinase Activity in Colcemid Synchronized Fibroblasts. Exptl. Cell Res., 48: 6 5 2 4 5 6 . Tobey, R. A,, and K. D. Ley 1970 Regulation of Initiation of DNA Synthesis of Chinese Hamster Cells. I. Production of Stable, Reversible, G1-arrested Populations i n Suspension Culture. J. Cell Biol., 46: 151-157. Toliver, A., E. H. Simon and P. T. Gilham 1968 On the Mechanism of 5-Bromouracil Inhibition of DNA Synthesis and Cell Division. Exptl. Cell Res., 53: 506-518.
