BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 717-721
Vol. 176, No. 2, 1991 April 30, 1991
A DNA Crosslinking Drug Alters Synthesis of Several Low Molecular Weight Proteins in Human Lymphoma Cells Richard L. Widstrom and Jonathan M. Ducore Department of Pediatrics, School of Medicine, University of California, Davis, California 95616 Received March 15, 1991
The cytotoxicity of bifunctionat alkylating agents is generally attributed to DNA damage, especially DNA-DNA crosslinking activity. It is unclear how crosslinks or other cellular damage result in cell death. Studies of drug effects at the level of expression of specific gene products may help elucidate the mechanism of cell killing. We examined proteins synthesized in L-phenylalanine mustard treated human lymphoma cells by [35S]methionine labeling and SDS-PAGE. Drug-treated cells showed decreased labeling of proteins in two molecular weight bands of 17 kDa (a doublet) and 12 kDa at 6, 18 and 24 hours after drug removal. One of the components of the 17 kDa doublet has been identified as calmodulin, a calcium binding protein essential to cell cycle progression and survival. ~ 1991Ao~demioP..... ~no.
L-phenylalanine mustard (L-PAM) is a member of the bifunctional alkylating agent group of anti-cancer drugs. The cytotoxicity of these drugs is attributed to their crosslinking action on cellular DNA, forming covalent crosslinks between and within DNA strands. The mechanism by which crosslinks result in cell death is unknown. Most investigators have focused on changes in DNA synthesis, cell cycling or poly(ADP-ribose)-induced damage as mediators of cytotoxicity (1). We are studying perturbations of expression of specific gene products which could result in cell death. We found that L-PAM inhibits the synthesis of several proteins, one of which may be calmodulin, a calcium receptor essential to cell survival.
Materials and Methods The human Burkitt's lymphoma cell line, BHM was established by Dr. lan Magrath at the National Cancer Institute and graciously made available to us. BHM cells were maintained as suspension cultures in RPMI (Roswell Park Memorial Institute) 1640 containing 10% heat inactivated fetal bovine serum (HIFBS, Hazelton), penicillin, streptomycin and glutamine (100 units/ml, 100/Jg/ml and 2 mM respectively). L-PAM (Drug Synthesis and Chemistry Branch, National Cancer Institute) was dissolved in 0.1 N HCI at a concentration of 10 mM and stored at -20°C. BHM cells in mid-log phase growth were pelleted, resuspended in low serum medium (RPMI 1 640 with 1% HIFBS) at 1 x 106 cells/ml and treated with 50 pM LPAM or equivalent amount of solvent for one hour. Cells were then pelleted, washed several times and returned to fresh medium with 10% serum. At 0, 6, 18 and 24 hours after drug removal, 1 x 106 cells were placed in one ml of medium lacking methionine and pulse labeled with 150 pCi of [35S]methionine (Amersham, 1224
717
0006-291X/91 $I.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 176, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Ci/mmol) for one hour. Cells were washed several times with cold phosphate buffered saline and pellets resuspended in lysis buffer (50 mM Tris, 4 % SDS, 2% 2-mercaptoethanol, 12% glycerol, pH 6.8) with protease inhibitors (1 mM EDTA and 1 mM PMSF). Equivalent trichloroacetic acid insoluble cpm were loaded onto 16% SDS-PAGE minigels (Hoefer) or 14 x 16 cm gels. Electrophoresis was performed according to the modified Laemmli procedure of Schagger and von J a g o w (2). Gels were dried and exposed to Kodak X-AR5 film. Autoradiographs were scanned with a transmission densitometer (Heeler GS300) connected to a Macintosh computer. Data analysis software (Hoefer) integrated peak areas from gel lane tracings and ratios of peak areas (treated cells) to peak areas (control cells) were calculated for the gel lanes of interest. Confidence limits were determined using statistical software (Abstat, AndersonBell).
Results
Protein synthesis following drug removal Figure 1 shows an autoradiograph of [35S]methionine pulse-labeled proteins from L-PAM treated and control cells. There were no qualitative differences in the overall pattern of proteins synthesized in the drug-treated and control cells. Drug treatment did not produce proteolytic fragments which would have smeared the background of the sample lanes. Quantitative differences were, however, evident in t w o low molecular w e i g h t bands. Bands at 17 kDa (a doublet) and 12 kDa decreased in intensity with drug treatment at 6 (gel not shown), 18 and 24 hours after L-PAM removal while no differences were seen in cells harvested immediately after drug removal (0 hours). We quantitated changes in the 17 kDa and the 12 kDa bands and actin (43 kDa) by densitometric scanning of the autoradiographs. As figure 2 shows, for drugtreated cells, the 9 5 % confidence intervals for label incorporation into actin overlap
0 hr C
Figure 1.
I
L-PAM
18 hr C
I
L-PAM
24 hr C
I
L-PAM
Autoradioeraohs of [35$]methionine labeled oroteins seoarated bv SDSPAGE. BHM cells were treated with 50 pM L-PAM for one hour, washed, returned to growth media and pulse labeled with [35S]methionine immediately after drug treatment (0 hours), 18, and 24 hours later. Cell lysates of untreated cells (C) or treated (L-PAM) were analyzed on 16% SDS gels. Arrows indicate actin (43 kDa) and two low molecular weight sizes (17 kDa and 12 kDa) where consistent band intensity decreases were apparent in treated lysates. 0 hr samples were run on 14 x 16 cm gels; 18, 24 hr on minigels.
718
Vol. 176, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
H o u r s A l t e r Drug R e m o v a l
H o u r s After Drug R e m o v a l
0 i
0 i
6 i
18 I
24 i
6 I
120"
18 i
H o u r s A f t e r Drug R e m o v a l
24 0 i
6 I
18 i
24 I
12 kDa
,~o 1005E So o o
--~o ¢
8060" 4£
Figure 2.
Synthesis of pr0"~eins in L-PAM treated cells as oercent of control at different times following drug removal. Relative synthesis was measured by densitometric scanning of autoradiographs of labeled proteins separated on SDS gels. Means (u) and 95% confidence limits (bars) are from three separate drug treatments. Dashed line is 100% of control.
the control synthesis (100%) line at 0, 6, 18 and 24 hours after L-PAM removal. The synthesis of proteins in the 17 kDa and 12 kDa bands was unaffected immediately following the one-hour drug treatment, but by 6 hours, synthesis was inhibited 20-30% in the drug-treated cells. At 18 and 24 hours, label incorporation into both bands remained decreased with synthesis at 65% of control by 24 hours after drug treatment. Measurements of the 12 kDa band have wider confidence intervals due to its faint intensity. Detection o f calmodulin in the 17 kDa doublet. Calmodulin, (17 kDa molecular
weight), comprises up to 1% of total cellular protein and labels intensely with [35S]methionine in intact cells (3). Because these properties matched those of the 17 kDa doublet, we tested whether calmodulin was a component of those bands. The electrophoretic mobility of calmodulin on SDS gels is affected by the presence of calcium (4). Calmodulin migrates slightly faster when Ca 2+ is present in the sample or gel buffer. Figure 3 shows SDS gels of purified calmodulin, molecular weight standards and cell lysates from control and drug-treated cells. In the Coomassie blue stained gels (3A and 3B), the protein bands from the cell lysates stain only faintly. In the gel run with Ca 2+ (figure 3A), the calmodulin band is seen to run slightly ahead of the 1 7 kDa molecular weight marker while it runs slightly behind the marker in the gel with EDTA (no Ca 2+) (figure 3B). The autoradiographs of these gels are shown in 3C and 3D. Figure 3C is of the gel with Ca 2+ and 3D is of the gel with EDTA. The dots in the lanes mark the positions of calmodulin and the molecular weight standards on the autoradiographs. The upper band of the doublet co-migrates with the 17 kDa size marker. In the control lanes, there is a marked decrease in the intensity of the lower band of the 17 kDa doublet when the gels are run without Ca 2÷ and a corresponding increase in the intensity of the upper band. This parallels the shift of calmodulin from the faster to the slower migrating form. The shift is also seen in material from L-PAM treated cells, but in this latter case, there is near disappearance of the lower band of the doublet. As was measured in the quantitative 719
Vol. 176, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCHCOMMUNICATIONS
+ Ca 2+
+ EDTA
I
l Std "w
C
I L-PAM
CaM
II
CaM
,,w
Std
C
L-PAM
*--17 kDa
17 kDa -~
Vol. 176, No. 2, 1991 April 30, 1991
A DNA Crosslinking Drug Alters Synthesis of Several Low Molecular Weight Proteins in Human Lymphoma Cells Richard L. Widstrom and Jonathan M. Ducore Department of Pediatrics, School of Medicine, University of California, Davis, California 95616 Received March 15, 1991
The cytotoxicity of bifunctionat alkylating agents is generally attributed to DNA damage, especially DNA-DNA crosslinking activity. It is unclear how crosslinks or other cellular damage result in cell death. Studies of drug effects at the level of expression of specific gene products may help elucidate the mechanism of cell killing. We examined proteins synthesized in L-phenylalanine mustard treated human lymphoma cells by [35S]methionine labeling and SDS-PAGE. Drug-treated cells showed decreased labeling of proteins in two molecular weight bands of 17 kDa (a doublet) and 12 kDa at 6, 18 and 24 hours after drug removal. One of the components of the 17 kDa doublet has been identified as calmodulin, a calcium binding protein essential to cell cycle progression and survival. ~ 1991Ao~demioP..... ~no.
L-phenylalanine mustard (L-PAM) is a member of the bifunctional alkylating agent group of anti-cancer drugs. The cytotoxicity of these drugs is attributed to their crosslinking action on cellular DNA, forming covalent crosslinks between and within DNA strands. The mechanism by which crosslinks result in cell death is unknown. Most investigators have focused on changes in DNA synthesis, cell cycling or poly(ADP-ribose)-induced damage as mediators of cytotoxicity (1). We are studying perturbations of expression of specific gene products which could result in cell death. We found that L-PAM inhibits the synthesis of several proteins, one of which may be calmodulin, a calcium receptor essential to cell survival.
Materials and Methods The human Burkitt's lymphoma cell line, BHM was established by Dr. lan Magrath at the National Cancer Institute and graciously made available to us. BHM cells were maintained as suspension cultures in RPMI (Roswell Park Memorial Institute) 1640 containing 10% heat inactivated fetal bovine serum (HIFBS, Hazelton), penicillin, streptomycin and glutamine (100 units/ml, 100/Jg/ml and 2 mM respectively). L-PAM (Drug Synthesis and Chemistry Branch, National Cancer Institute) was dissolved in 0.1 N HCI at a concentration of 10 mM and stored at -20°C. BHM cells in mid-log phase growth were pelleted, resuspended in low serum medium (RPMI 1 640 with 1% HIFBS) at 1 x 106 cells/ml and treated with 50 pM LPAM or equivalent amount of solvent for one hour. Cells were then pelleted, washed several times and returned to fresh medium with 10% serum. At 0, 6, 18 and 24 hours after drug removal, 1 x 106 cells were placed in one ml of medium lacking methionine and pulse labeled with 150 pCi of [35S]methionine (Amersham, 1224
717
0006-291X/91 $I.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 176, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Ci/mmol) for one hour. Cells were washed several times with cold phosphate buffered saline and pellets resuspended in lysis buffer (50 mM Tris, 4 % SDS, 2% 2-mercaptoethanol, 12% glycerol, pH 6.8) with protease inhibitors (1 mM EDTA and 1 mM PMSF). Equivalent trichloroacetic acid insoluble cpm were loaded onto 16% SDS-PAGE minigels (Hoefer) or 14 x 16 cm gels. Electrophoresis was performed according to the modified Laemmli procedure of Schagger and von J a g o w (2). Gels were dried and exposed to Kodak X-AR5 film. Autoradiographs were scanned with a transmission densitometer (Heeler GS300) connected to a Macintosh computer. Data analysis software (Hoefer) integrated peak areas from gel lane tracings and ratios of peak areas (treated cells) to peak areas (control cells) were calculated for the gel lanes of interest. Confidence limits were determined using statistical software (Abstat, AndersonBell).
Results
Protein synthesis following drug removal Figure 1 shows an autoradiograph of [35S]methionine pulse-labeled proteins from L-PAM treated and control cells. There were no qualitative differences in the overall pattern of proteins synthesized in the drug-treated and control cells. Drug treatment did not produce proteolytic fragments which would have smeared the background of the sample lanes. Quantitative differences were, however, evident in t w o low molecular w e i g h t bands. Bands at 17 kDa (a doublet) and 12 kDa decreased in intensity with drug treatment at 6 (gel not shown), 18 and 24 hours after L-PAM removal while no differences were seen in cells harvested immediately after drug removal (0 hours). We quantitated changes in the 17 kDa and the 12 kDa bands and actin (43 kDa) by densitometric scanning of the autoradiographs. As figure 2 shows, for drugtreated cells, the 9 5 % confidence intervals for label incorporation into actin overlap
0 hr C
Figure 1.
I
L-PAM
18 hr C
I
L-PAM
24 hr C
I
L-PAM
Autoradioeraohs of [35$]methionine labeled oroteins seoarated bv SDSPAGE. BHM cells were treated with 50 pM L-PAM for one hour, washed, returned to growth media and pulse labeled with [35S]methionine immediately after drug treatment (0 hours), 18, and 24 hours later. Cell lysates of untreated cells (C) or treated (L-PAM) were analyzed on 16% SDS gels. Arrows indicate actin (43 kDa) and two low molecular weight sizes (17 kDa and 12 kDa) where consistent band intensity decreases were apparent in treated lysates. 0 hr samples were run on 14 x 16 cm gels; 18, 24 hr on minigels.
718
Vol. 176, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
H o u r s A l t e r Drug R e m o v a l
H o u r s After Drug R e m o v a l
0 i
0 i
6 i
18 I
24 i
6 I
120"
18 i
H o u r s A f t e r Drug R e m o v a l
24 0 i
6 I
18 i
24 I
12 kDa
,~o 1005E So o o
--~o ¢
8060" 4£
Figure 2.
Synthesis of pr0"~eins in L-PAM treated cells as oercent of control at different times following drug removal. Relative synthesis was measured by densitometric scanning of autoradiographs of labeled proteins separated on SDS gels. Means (u) and 95% confidence limits (bars) are from three separate drug treatments. Dashed line is 100% of control.
the control synthesis (100%) line at 0, 6, 18 and 24 hours after L-PAM removal. The synthesis of proteins in the 17 kDa and 12 kDa bands was unaffected immediately following the one-hour drug treatment, but by 6 hours, synthesis was inhibited 20-30% in the drug-treated cells. At 18 and 24 hours, label incorporation into both bands remained decreased with synthesis at 65% of control by 24 hours after drug treatment. Measurements of the 12 kDa band have wider confidence intervals due to its faint intensity. Detection o f calmodulin in the 17 kDa doublet. Calmodulin, (17 kDa molecular
weight), comprises up to 1% of total cellular protein and labels intensely with [35S]methionine in intact cells (3). Because these properties matched those of the 17 kDa doublet, we tested whether calmodulin was a component of those bands. The electrophoretic mobility of calmodulin on SDS gels is affected by the presence of calcium (4). Calmodulin migrates slightly faster when Ca 2+ is present in the sample or gel buffer. Figure 3 shows SDS gels of purified calmodulin, molecular weight standards and cell lysates from control and drug-treated cells. In the Coomassie blue stained gels (3A and 3B), the protein bands from the cell lysates stain only faintly. In the gel run with Ca 2+ (figure 3A), the calmodulin band is seen to run slightly ahead of the 1 7 kDa molecular weight marker while it runs slightly behind the marker in the gel with EDTA (no Ca 2+) (figure 3B). The autoradiographs of these gels are shown in 3C and 3D. Figure 3C is of the gel with Ca 2+ and 3D is of the gel with EDTA. The dots in the lanes mark the positions of calmodulin and the molecular weight standards on the autoradiographs. The upper band of the doublet co-migrates with the 17 kDa size marker. In the control lanes, there is a marked decrease in the intensity of the lower band of the 17 kDa doublet when the gels are run without Ca 2÷ and a corresponding increase in the intensity of the upper band. This parallels the shift of calmodulin from the faster to the slower migrating form. The shift is also seen in material from L-PAM treated cells, but in this latter case, there is near disappearance of the lower band of the doublet. As was measured in the quantitative 719
Vol. 176, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCHCOMMUNICATIONS
+ Ca 2+
+ EDTA
I
l Std "w
C
I L-PAM
CaM
II
CaM
,,w
Std
C
L-PAM
*--17 kDa
17 kDa -~
