sign in sign up
<<
Home , Nitrosourea

(CANCER RESEARCH 54, 4524-4531, August 15. 1994] Capacity of Individual Chronic Lymphatic Leukemia Lymphocytes and Leukemic Blast Cells for Repair of O6-Ethylguanine in DNA: Relation to Chemosensitivity in Vitro and Treatment Outcome1

Mark R. Müller,2Frank Seiler, JürgenThomale, Claudia Buschfort, Manfred F. Rajewsky, and Siegfried Seeber

Department of Internal Medicine (Cancer Research) ¡M.R. M., S. S.¡and Institute of Cell Biology (Cancer Research) [F. S.,./. T., C. B., M. F. R./. West German Cancer Center Essen, University of Essen Medical Sclutol, D-45122 Essen, Germany

ABSTRACT BCNU or MAF react further to form DNA intra- or interstrand cross-links which are considered to be the major cytotoxic lesion (8). The elimination kinetics of the alkylation product O6-ethylguanine Alkyl groups covalently attached to the O'' atom of guanine are (06eGua) from nuclear DNA were determined in individual lymphocytes eliminated from DNA by the repair protein AT (EC 2.1.1.63) (4, 9). or blast cells isolated from 27 patients with chronic lymphatic leukemia Different types of human or rodent cells vary widely in AT activity, (CLL) and 26 patients with de novo acute myeloid leukemia i \MI i. A monoclonal antibody-based immunocytological assay was used for quan and in addition, there is considerable interindividual variability (10, tification of OVGua in DNA of individual cells after pulse exposure of cells 11). In cultured cell lines, elevated AT levels have been linked to to \-rili\l- \-nliioMHin-.i 11INI i. In cell specimens from a given patient, increased resistance to methylating and chloroethylating agents (12, no major subpopulations with significantly different capacities for repair 13). Depletion of the cellular AT pool by nontoxic bases such as of O'eGua were observed. The time required to remove 50% of induced O''-BG increased the cytotoxicity of BCNU in human cell lines (14) O6eGua residues varied interindividually between 0.5 and 8.4 h in CLL and leukemic blasts derived from patients (15). lymphocytes and between 0.8 and 6.3 h in leukemic blast cells. An inverse Other studies, however, did not obtain evidence for AT activity as relationship was found between the rate of removal of O6eGua from DNA a major cause of resistance to alkylating agents (16, 17). Recent and the Chemosensitivity of cells to EtNU, Ir3-bis(2-chloroethyl)-l-nitro- sourea or in vitro. High rates of 06eGua repair and pro findings suggest that base and nucleotide excision repair pathways may also contribute to the repair of O'VAlkGua in DNA. Thus, the nounced resistance to mafosfamide, l,3-bis(2-chloroethyl)-l-nitrosourea, bacterial UV-excision repair complex efficiently eliminates O6- and EtNU in vitro were found in samples from 8 CLL patients nonrespon methyl and -ethylguanine from DNA in vivo, and enhanced cytotox sive to with alkylating agents. In VMI patients treated with and l-ß-i>-arabinofuranosylcytosine, no relation was found icity of EtNU has been observed in cell lines deficient in excision between DNA repair capacity and treatment outcome. However, increased repair enzymes (18, 19). Excision repair pathways are also involved in P-glycoprotein expression was observed between specimens derived from the removal of DNA monoadducts and interstrand cross-links induced \MI patients who had failed to reach complete remission (n = 12) after by "bulky" alkylating agents, e.g., MAF (1). Enhanced repair of DNA chemotherapy versus responsive patients (n = 14). DNA repair rate was cross-links, for example, was found in human cell lines resistant to not related to Chemosensitivity to Adriamycin and 1-ß-u-arabinofurano- (20). As for AT activity, considerable interindividual vari sylcytosine in vitro, nor were cellular glutathione content, glutathione ations have been reported for the expression of excision repair genes S-transferases activity, or P-glycoprotein expression. in bone marrow cells from cancer patients (21). Little is known about the contribution of different DNA repair pathways to clinical drug resistance. This may be partly due to INTRODUCTION methodological difficulties of measuring DNA repair in individual The development of resistance to alkylating agents remains a major cells of heterogeneous biopsy material. In cell extracts, increased obstacle in cancer chemotherapy. In cell lines, resistance may result expression of DNA excision repair genes was found in lymphocytes from increased detoxification by GSH1 or cellular capacity to repair from CLL patients nonresponsive to treatment (22). cytotoxic DNA adducts and cross-links (1-3). Various reports have Enhanced repair of DNA cross-links has been observed in lympho indicated that O''-AlkGua is of particular importance for both muta- cytes from therapy-resistant CLL patients (23). However, no relation genicity and cytotoxicity of alkylating agents (1,3, 4). It has been between repair of cross-links or AT level and treatment outcome in shown that unrepaired Oh-AlkGua in DNA leads to excess strand leukemia patients was found in other studies (24, 25). Thus, the breaks considered to be mainly responsible for cytotoxicity (5-7). relevance of DNA repair as a major mechanism of cytotoxic drug DNA adducts produced by bifunctional alkylating agents such as resistance remains an open question. Assays to study DNA repair mechanisms in individual cells have Received 3/1/94; accepted 6/8/94. been developed recently (26, 27). These techniques have been applied The costs of publication of this article were defrayed in part by the payment of page to the analysis of cell lines but not yet to a larger number of clinical charges. This article must therefore he hereby marked advertisement in accordance with specimens. As a functional test for the repair of O''eGua in DNA of 18 U.S.C. Section 1734 solely to indicate this fact. ' This work was supported by a grant from the Dr. Mildred Scheel Stiftung (W69/91/ isolated leukemic cells, we have used a newly established immun- Muí). ofluorescence assay and a monoclonal antibody specific for O6-alkyl- 2 To whom requests for reprints should be addressed, at Innere Klinik und Poliklinik 2'-deoxyguanosine. A unique feature of this method is its ability to (Tumorforschung). Westdeutsches Tumorzenlrum Essen. Universitätsklinikum Essen, Hufclandstrasse 55, D-45122 Essen, Germany. %The abbreviations used are: GSH. glutathione; CLL, chronic lymphatic leukemia; quantify the initial level and the kinetics of repair of a specific type of AML, acute myeloid leukemia; EtNU, A^-ethyl-N-nitrosourea; C/'eGua. O^-ethyìguanine; DNA damage (e.g., O6eGua) in individual cells of a given cell O''-AlkGua, O"-alkylguanine; O''-BG. W-benzylguanine; BCNU, l,3-bis(2-chloroethyl)- population. The advantage of using EtNU, although not clinically 1-nitrosourea; MAF, mafosfamide; ADM, Adriamycin; ARA-C, l-ß-D-arabinofuranosyl- cytosine; GST, glulathione .S'-transferascs; PGP. P-glycoprotein; PBS, phosphate-buffered used as a chemotherapeutic agent, is that all major DNA alkylation saline; DAPI, 4,6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide; MDR, multiple products are chemically well characterized and their relative rates of drug resistance; MTT, 3-(4,5-dimethylthiazol-2-yl)-2.5-diphenyltetrazolium bromide; PI. propidium iodide; FITC, fluorcscein ¡sothiocyanate; AT, (/"-alkylguanine-DNA alkyl- formation in DNA are known. transferase; BSA, bovine serum albumin; IDM),drug concentration required to reduce ^54,, It is probable that multiple cellular mechanisms involved in drug to 50% of control; (wra (/,««,),periodrequired to remove 50% (90%) of O*eGua residues. resistance operate in a given tumor cell population. Therefore, we 4524

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CU. LYMPHOCYTES AND AML BLAST CELLS determined, as additional parameters, the cellular GSH content, GST air dried, fixed for 30 s in acetone at -20°C, evaporated at 4°C.and stored at activity and PGP expression. GSH can inactivate several alkylating -80°C until stained. agents by a conjugation reaction catalyzed by GST (2). Elevated Immunofluorescence Staining. Immunofluorescence staining of O"eGua cellular GSH content and enhanced activity of GST have frequently in nuclear DNA was performed essentially as previously described (37, 38). been observed in cell lines resistant to (28) and Briefly, cells on slides were thawed, fixed in mcthanol for 15 min at room temperature, rehydrated in 2x standard sodium citrate (300 mw NaCI, 30 mM (29). Elevated GST activity was also observed in lymphocytes from sodium citrate), and treated with RNases (RNase A, 200 (ig/ml; RNase Tl, 50 CLL patients nonresponsive to chlorambucil treatment (30). Further units/ml) for l h at 37°C.Cells were then washed in 0.14 M NaCI, and DNA more, overexpression of the drug efflux pump PGP has been identified was partly denatured by alkali treatment (NaOH, 70 mM in 0.14 M NaCL, 5 as one mechanism underlying the MDR phenotype (31). MDR cell min. 0°C).After two rinses with PBS and preincubation with PBS containing lines are simultaneously resistant to several unrelated cytotoxic agents 20% BSA (PBS/BSA) for 20 min at room temperalure, cells were incubated such as anthracyclines and Vinca alkaloids but not to alkylating agents with the monoclonal antibody ER-17 at a concentration of 0.2 tig/ml PBS/BSA for 16 h at 4°C.Cells were then washed twice with PBS and stained with goat (31). PGP expression in leukemic blasts of newly diagnosed AML anti-rat IgG F(ah)2 fragment conjugated with rhodamine isothiocyanate (Di- patients appeared to be closely related to resistance to induction anova, 2 (xg/ml PBS/BSA) for 45 min at 37°C.Nuclear DNA was counter- chemotherapy with anthracyclines (32, 33). The relationship between stained with the fluorescent dye DAPI (3 x 10~7 Min PBS, 10 min). To reduce these mechanisms of resistance and cellular DNA repair capacity in fading of fluorescence, slides were then mounted in PBS containing 0.05 M malignant cells derived from patients remains to be elucidated. Tris-HCI, 0.033 M dithioerytritol, 30% glyccrol, and 10% Elvanol (pH 8.2). The aim of this study was to determine the repair kinetics of Quantitation of O'eGua in the Nuclear DNA of Individual Cells. A O''eGua in the DNA of individual human leukemic cells and to Zeiss fluorescence photomicroscope III equipped with an HBO 100-W mercury correlate cellular repair capacity with chemosensitivity profiles in arc lamp and Zciss standard filter combinations 02 (for DAPI) and 14 (for vitro, GSH content, GST activity, MDR expression, and treatment rhodaminc) were used. Fluorescence signals were amplified by an electronic outcome. The chemosensitivity pattern of isolated leukemic cells was intensifier (Proxitronic Proxifier BV2532). recorded by a video camera (Hamamatsu Vidicon C 1000-12 SIT), and fed into a multiparameter image studied using the MTT assay which has provided good clinical cor analysis program (ACAS Cytometry Analysis System; Ahrens Electronics, relations (34, 35). Since different alkylating agents vary with respect Bargteheide, Germany). This program enables integration of pictures with low to the relative frequencies of different alkylation products, we evalu signal to noise ratios and separate quantification of both antibody fluorescence and ated the cytotoxic effects of EtNU, BCNU, and MAP, a direct acting DNA fluorescence of the same cell. Thresholds were set to discriminate back derivative of 4-hydroxycyclophosphamide. In addition, chemosensi ground from DNA-staining signals and to determine the image points to be tivity to ADM and ARA-C was measured. included in the evaluation. Signal intensities for both DNA staining and i/'eGua immunostaining of the selected pixels were expressed as the integrated values (average signal X number of selected pixels) for each nucleus. C/'eGua signals MATERIALS AND METHODS were corrected for nuclear DNA content. Average staining intensities were ex pressed as the mean fluorescence intensity from 100-2(X) nuclei. Drugs. The drugs used and the manufacturers are as follows: MAP, Asta Cellular GSH Content and GST Activity. Cellular GSH content was Medica (Bielefeld, Germany); EtNU, Serva (Heidelberg, Germany); ARA-C, determined by measuring the formation of 2-nitro-5-mcrcaptobenzoic acid Mack (Jllertissen, Germany); ADM, Sigma (Deisenhofen, Germany); and from 5,5'-dithiobis(2-nitrobenzoic acid) in the presence of glutathione reduc BCNU, Bristol-Myers (Bergisch Gladhach, Germany). Stock solutions of íase,as previously described (39). drugs were prepared in PBS, cthanol, or DMSO and stored at -20°. GST activity was measured in the supernatants of cell sonicates by a Preparation of Lymphocytes or Blast Cells and Clinical Data. Blood previously described procedure using l-chloro-2.4-dinitrobenzene as substrate samples obtained from AML or CLL patients were obtained before treatment (40). Activity values were recorded in nmol/mg protein/min. as monitored by and analyzed on the same day. Heparinized blood (10 ml) was layered onto 10 the change in /\,4I, at 25°C. ml Ficoll-Hypaque and centrifuged for 25 min at 200 x g at room temperature. Staining for Flow Cytometry. The flow cytometric determination of PGP Cells at the interface were removed, washed twice, and resuspended in PBS. expression was carried out using the monoclonal antibody MRKI6 (kindly The resulting cell suspension contained >90% leukemic blast cells or lympho provided by Dr. T. Tsuruo. Tokyo, Japan) as previously described (33). cytes, as confirmed by light microscopy. Specimens were obtained from 26 Briefly, after preparation of single-cell suspensions, 2 X l()h cells were AML patients at the time of initial diagnosis. Treatment of newly diagnosed suspended in 50 /nl of MRK16 antibody diluted to 50 jig/ml in PBS containing patients included combination chemotherapy with either , 10% PBS/BSA. As a control for nonspecific staining, cells were incubated ARA-C. and thioguanine or and ARA-C. Complete remission was with mouse myeloma protein-IgG2Ax (Sigma) diluted to 50 /xg/ml. After defined as normal hematopoiesis in the marrow with <5% blast cells after one incubation for 45 min at 4°C,cells were washed twice with PBS. Cells were course of induction therapy. Treatment failure (resistant disease) was defined then resuspended in 50 /il of FITC-conjugated rabbit anti-mouse IgG serum as persistence of >5% blasts after induction therapy. In the case of CLL, 9 (Sigma) diluted 1:32 in PBS/BSA and incubated for 45 min at 4°C.Thereafter, patients did not receive any chemotherapy and were classified as having stage 5 JAM PI was added for staining of cell debris. The custom-made flow 0 or I disease (36). Ten specimens were derived from treated, sensitive CLL cytometer was instructed to record forward light scatter, FITC. and PI fluo patients (stage I-Ili) whose peripheral lymphocyte count returned to rescence. F1TC and PI fluorescence were excited at 488 nm simultaneously and < 15,000/ixl after intermittent treatment with chlorambucil. In 8 treated, resis recorded at 530 ±15 and 620 ±20 nm, respectively. Calculation of mean tant patients (stage II/III), therapy with chlorambucil or intermittent high-dose fluorescence was carried out after conversion of logarithmically amplified , , and prednisone failed to reduce peripheral signals into values on a linear scale and expressed as relative fluorescence lymphocyte counts by 25-30%. All specimens were derived from B-CLL units. Twenty thousand cells per sample were analyzed. Total FITC fluores cases, and no transformation into Richter's syndrome was noted. cence was determined after exclusion of cell debris by a gating procedure, and Pulse Exposure to EtNU. Lymphocytes or blast cells were exposed to the fraction of specifically stained cells was calculated by histogram subtrac EtNU in PBS supplemented with 900 /XMCa2+, 490 JAMMg2*, and 25 mM tion analysis (33). 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid at pH 7.25 and 37°C. MTT Assay. For the MTT assay, cells were seeded into 96-well microtiter Stock solutions of 100 mg EtNU/ml DMSO were prediluted in serum-free plates using 200 /xl/well of a cell suspension containing 5 x IO5 blast cells/ml "acidic" RPMI medium (adjusted to pH 6.2 with CO,) and added to the cells or 10'' lymphocytes/ml of RPMI medium supplemented with 10% fetal calf at a final concentration of 100 /xg/ml. After 20 min of incubation at 37°C,cells serum. Drugs dissolved in PBS (20 /xl) were added. On day 4, 20 jxl of a were washed twice in PBS and resuspended in fresh, prewarmed medium. solution of 5 mg MTT/ml of PBS were added to each well, and the plates were Samples were taken immediately or 1.5, 3, 6, 9, and 24 h after EtNU treatment returned to the incubator for 5 h. Thereafter, the plates were centrifuged for 10 and washed with PBS, and cells were seeded onto glass slides. Slides were then min at 100 x g. The medium was then removed from each well, 200 /xl of 4525

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES AND AML BLAST CELLS

DMSO were added to dissolve the crystals, and the plates were shaken for 10 cytes. Similarly, the range of im,% values varied by factors of 19 in min on a plate shaker. A54()and Ahm were read on a dual-wavelength Dynatech blast cells and 17 in CLL lymphocytes. The ranking of f50,{ and t

en z UJ

S .;•'..-.! ••••' en .":!:..- •: u Œ O

a _j o

O 20 40 60 80 100 120 140

DIGITIZED IMAGE ANALYSIS RED FLUORESCENCE INTENSITY (REL. UNITS)

Z 20-- = 20 8 8 i

BLUE FLUORESCENCE INTENSITY (REL. UNITS) RED FLUORESCENCE INTENSITY (REL. UNITS) Fig. 1. Immunofluorescence staining of O6eGua in nuclear DNA of individual CLL lymphocytes 1.5 h after pulse exposure to EtNU. Top, left, digitized image analyses of electronically amplified fluorescence signals representing O6eGua in DNA. Top, right, bivariate histograms of monoclonal antibody derived (abscissa) and DNA fluorescence (ordinate; relative fluorescence units, linear scale). Bottom, monodimensional histograms of red and blue fluorescence representing O6eGua (red: right) and DNA content (blue; left). 4526

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES AND AML BLAST CELLS

150 PGP-expressing cells, or ID,,, values for MAP, BCNU, EtNU, ADM, or ARA-C (P > 0.05). PGP-expressing cells were found in 12 of 27 CLL specimen. The fraction of PGP-expressing CLL lymphocytes varied from 0-66.4%

100- of all cells. PGP expression was also detected in 17 of 26 AML samples by flow cytometry, ranging from 0-84.2% of all cells in individual specimens. Relation to Clinical Status. Tables 1 and 2 display the parameters determined in CLL lymphocytes and AML blast cells in relation to 50- clinical status. The mean /5()%and il)()%values for repair of O6eGua and the cytotoxicity values for EtNU, BCNU, and MAP were not significantly different between specimens derived from untreated or treated sensitive CLL patients. However, high rates of repair were 0- observed in specimens from 8 CLL patients resistant to treatment with alkylating agents (Fig. 6, Table 1). Mean r,(ra, and ?w% values were 0 5 10 15 20 25 significantly lower in comparison to untreated or treated sensitive CLL patients, and the kinetics of O6eGua elimination exhibited linear TIME AFTER ETNU PULSE (h) Fig. 2. Kinetics of elimination of C/'eGua from nuclear DNA in two specimens of CLL characteristics. In this group of specimens, the mean ID,,, values for lymphocytes after in vitro pulse exposure to EtNU. Elimination curves were determined EtNU, BCNU, and MAP were also elevated in comparison to un in specimens obtained from 2 CLL patients who were cither sensitive (closed symbols) or treated or treated sensitive CLL patients. For all other parameters resistant (open symbols) to treatment with alkylaling agents. Each point represents the mean value of KM)cells. Antibody-derived fluorescence signals were corrected for cellular tested, no significant differences were observed between groups of DNA content for each point, i,,,,, and i,A,a were determined as a measure for DNA repair CLL patients. DNA repair capacity, GSH content, GST activity, and capacity. chemosensitivity to alkylating agents in vitro were not different in specimens derived from AML patients when compared for treatment outcome. However, increased MDR expression was found in samples rs = 0.81; BCNU/MAF, rs = 0.63; P < 0.001, n = 26). No relation derived from resistant AML patients. ship was observed between DNA repair rate and ID5()values for ADM and ARA-C in either cell types (AML blast cells: fso% values/ARA-C, DISCUSSION rs = 0.18; /50%values/ADM, rs = 0.16, P > 0.05; CLL lymphocytes: tso% values/ARA-C, rs = 0.12; fso% values/ADM, rs = 0.15; The aim of this study was to evaluate the cellular capacity for repair P > 0.05). Furthermore, IDS()values for alkylating agents, ADM and of a critical DNA alkylation product (OheGua) in individual CLL ARA-C, were not correlated in CLL lymphocytes or blast cells lymphocytes and AML blasts in relation to chemosensitivity in vitro (P > 0.05). and to treatment outcome. A recently developed immunofluorescence Cellular GSH content was determined in CLL lymphocytes (in assay was applied for the quantification of O''eGua in the nuclear nmol/mg protein; median, 23.2; range, 15.2-31.2; mean, 24.9; SD, DNA of individual cells. The data demonstrate a broad range of 4.9; n = 27) and leukemic blast cells (in nmol/mg protein; median, intervals required for repair of O''eGua among specimens derived 51.2; range, 35.7-66.2; mean, 50.2; SD, 6.7; n = 26). Moreover, GST from CLL or AML patients, suggesting large interindividual differ activity was measured in CLL lymphocytes (in nmol/mg protein/min; ences in the activities of DNA repair proteins in human cells (10, 11, median, 49.2; range, 35.6-58.2; mean, 50.2, SD, 8.4; n = 27) and 21). The relative contribution of distinct DNA repair pathways, such blast cells (in nmol/mg protein/min; median, 118.2; range, 80.2- as base and nucleotide excision repair or single-step repair by AT, to 128.4; mean, 120.4; SD, 15.2; n = 26). Cellular GSH content and the elimination of OheGua is not yet clear. The important role of AT GST activity were slightly elevated in leukemic blast cells compared in restoring DNA integrity by transfer of the ethyl group from guanine to CLL lymphocytes (P > 0.02). No correlation was observed between to one of the cysteine residues of AT is well established (4, 9). GSH content or GST activity and f,0% or tw„/cvalues, fraction of However, recent reports suggest that excision repair mechanisms may

< z 5z Q Z Q 15- Fig. 3. Interindividual variations of cellular ca 5 5 pacities for repair of O^eGua (O''-ETGUA) in nu LU rr clear DNA in specimens of CLL lymphocytes and leukemic blast cells from different patients. (5OT, and rw% are shown on left and right, respectively. S P l Horizontal lines, means of the distributions. °J ~r ai O

CLL AML CLL AML 4527

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES AND AML BLAST CELLS

500 500

01 a. l 10(H 100-

*•'*•. i- UÕ V •• e o n io O Q

1 0 I ! 1 0 024 6 8 10 1 0

100 1 00

E S 1 0- t

o 1 O- m u 1- CD IO O o io O

0.1 024 8 10 1 O

100 1 00

10- O! io-i a. u.

1 -I o 1- o IO •t IO o Q

0.1 0.1 i 0246810 4 6 10

DNA REPAIR TIME (t50%; [h]) DNA REPAIR TIME (tso%; [h])

Fig. 4. Correlation between chemosensitivity in vitro and DNA repair time in CLL Fig. 5. Correlation between chemosensitivityinvitro and DNA repairtime in blast cells lymphocytes. Ordinales. 1DW values for ElNU. BCNU, and MAP; Abcissas. I5in values. derivedfrompatientswithAML Ordinales,ID,,,valuesforElNU,BCNU,and MAP;abscissas, Spearman rank correlation coefficients (rs) for DNA repair time and cytotoxicity were values.Spearmanrank correlationcoefficients(rj for DNA repair time and cytotoxicity -0.71 for EtNU, -0.54 for BCNU, and -0.69 for MAP (P < 0.001). were -0.75 for EtNU. -0.63 for BCNU.and -0.72 for MAP(P < 0.001). 4528

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES AND AML BLAST CELLS also contribute to the repair of O6eGua in DNA (18, 19). We recently showed that the removal of ethyl groups from the O6 position of guanine in DNA of human lymphocytes was decelerated but not completely abolished after depletion of the cellular pool of active AT by O6-BG (38). Repair of O6eGua could, therefore, result from the 20- combined action of different DNA repair systems with overlapping z £. substrate specificity. Furthermore, the curves of the kinetics of o O6eGua elimination from nuclear DNA exhibited at least two com ¡I ponents in most specimens: an initial period of fast repair, followed by = E a phase of slower removal. Whether different components of the Ãœ O curves represent the predominant action of distinct DNA repair sys tems remains to be elucidated. Experiments addressing this question should use AT-blocking agents such as O''-BG and determine the §3 KZ elimination of Oh-guanine adducts (i.e., larger size alkyl residues) not O ai repaired by AT. Heterogeneity of DNA damage and repair has been observed among different tissues, which stresses the need for a reliable quan • 5- titative assay to monitor DNA repair capacity in individual cells of S í heterogeneous patient material (41, 42). Although the present ap -t proach did not allow the analysis of single cells over time, it permitted the evaluation of cell-to-cell variation of O6eGua repair in individual

Table 1 O^eGua repair, PGP expression, GSH content, GST activity, and chemosensitivity profiles in vitro of isolated lymphocytes derived from CLL patients: relation to treatment outcome RESISTANT Fig. 6. Cellular capacity for repair of O^eGua in nuclear DNA of CLL lymphocytes in sensitive resistant (n =10)2.9 (n =8)1.1 relation to treatment outcome, t^n repair values were determined in specimens obtained from CLL patients who were either untreated or sensitive or resistant to treatment with ±1.9°12.5 ±0.6*3.1 DNA(h)'50%PGP repair time ±1.310.2 alkylating agents. Horizontal lines, means of the distributions. ±8.45.6 ±8.323.1 ±1.1*35.2 expression(% ±4.922.4 ±23.928.4 ±18.926.4 cells)GSHof total cells from a given population. Distinct major subpopulations express content(nmol/mg ±2.249.2 ±3.452.4 ±5.251.2 ing a DNA repair phenotype significantly different from the average protein)GST value were not observed among the specimens examined. However, activity(nmol/mg ±7.956.2 ±9.849.9 ±9.184.1 protein/min)IDsoEtNU larger numbers of cells analyzed per specimen will be required to detect small fractions of cells with distinct DNA repair capacities. In ±19.2e4.4 ((ig/ml)BCNUMAPADMARA-CUntreated3.6±24.92.3 ±21.33.2 order to evaluate cellular heterogeneity in the expression of DNA ±1.41.4 ±2.21.5 ±1.2e4.6 ±1.10.27 ±1.20.42 ±2.0e0.56 repair proteins such as AT, recently developed antibodies should also ±0.210.52 ±0.170.28 ±0.210.62 be used (43). Furthermore, since differences in DNA damage and ±0.34Treated ±0.14Treated ±0.31 " Mean ±SD. repair have been found between specific gene sequences, these types b Significantly different from Ihe DNA repair rate of untreated and treated sensitive of studies should be complemented by the measurement of gene- CLL patients (P < 0.005). specific repair (44, 45). Moreover, this approach should be extended c Significantly different from ID51, values of untreated and treated sensitive CLL to the selective measurement of DNA repair capacity in cancer cells patients (P < 0.005). in heterogeneous clinical biopsy material such as bone marrow or solid tumors. An inverse correlation has been observed between AT activity and Table 2 CfeGua repair, GSH content, GST activity, PGP expression, and the cytotoxicity of alkylating agents in cell lines (12, 14, 15). In this chemosensitivity profiles in vitro of isolated leukemic blast cells from AML patients: relation to treatment outcome study, a significant ranking correlation was found in leukemic cells between DNA repair time and the degree of chemoresistance to remission disease (n =14)2.1 (n =12)3.1 alkylating agents in vitro. This observation indicates the clinical

DNA(h)'.sir;''Mi'1;PGP repair time ±1.2°8.1 ±1.88.8 importance of DNA repair as a mechanism contributing to cellular ±7.43.2 ±7.144.3 resistance to alkylating agents. Furthermore, CLL lymphocytes and leukemic blasts were cross-resistant to both mono- and bifunctional ±24.2*50.2 expression(% ±5.352.4 alkylating agents that produce a broad spectrum of potentially cyto- cells)GSHof total content(nmol/mg ±5.3116.2 ±4.9122.4 toxic DNA adducts. This finding suggests the existence of either protein)GST common mechanisms or different, but coregulated, pathways mediat activity(nmol/mg ±13.468.9 ±15.470.2 protein/min)EtNU ing resistance to structurally unrelated alkylating agents. It is not yet clear whether DNA repair pathways involved in the removal of (ng/ml)BCNUMAPADMARA-CComplète ±42.24.2 ±39.83.9 O6eGua from DNA do indeed play a critical role for resistance to ±3.74.4 ±3.53.9 ±4.30.14 ±3.70.41 alkylating agents. In addition, single steps of specific DNA repair ±0.160.21 ±0.450.36 pathways, such as the processing of abasic sites or single-strand ±0.40Resistant ±0.20 breaks, repair synthesis, or ligation, are not measured by the adduct- " Mean ±SD. '"Significantly different from the fraction of PGP-expressing blasts in specimens from specific assay applied. Partial or complete coregulation of those responsive AML patients. mechanisms with repair activities directly involved in the removal of 4529

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES ANI) AML BLAST CELLS the O''-ethyl groups could also explain the observed relationship methylguanine-DNA methyltransferase-defective human cell mutant: O^-methylgua- nine, DNA strand breaks and cytotoxicity. Carcinogcnesis (Lond.), 9: 1749-1753, between DNA repair time and in vitro resistance. Combining the 1988. measurement of elimination kinetics for specific DNA adducts with Lefebvre, P.. and Laval, F. A human cell line proficient in C/'-methylguanine-DNA- other single-cell assays for DNA repair (e.g., the "comet assay"; 26) methyltransferase and hypersensitive to alkylating agents. Carcinogenesis (Lond.), 14: 1671-1675, 1993. could further elucidate the contribution of distinct repair pathways to 8. Kohn. K. W. Interstrand crosslinking of DNA by l,3-bis(2-chlorocthyl)-l-nitrosourea eliminate specific DNA lesions. As expected, the DNA repair rate for and other l-(2-haloethyl)-l-nitrosoureas. Cancer Res.. 37: 1450-1454. 1977. 0''eGua was not correlated with the cytotoxicity of ADM or ARA-C, 9. Sedgwick, B.. and Lindahl. T. A common mechanism for repair of (/'-methylguanine and O"-cthylguaninc in DNA. J. Mol. Biol., 154: 169-175, 1982. indicating no significant contribution of the DNA repair pathways 10. Waldstcin, E., Cao, E., Bender, M., and Setlow, R. Abilities of extracts of human involved to chemoresistance to these substances. lymphocytes to remove O6 methylguanine from DNA. Mutât.Res.. °5:406-416. No relationship was observed in the present study between cellular 1982. Myrnes, B.. Giercksky, K., and Krokan. H. Intcrindividual variation in the activity of GSH content or GST activity and any other parameter determined. C/'-methylguanine-DNA methyltransferase und uracil-DNA glycosylase and its ap Since single-cell assays are presently not available to measure reliably plication to human neoplastic tissues. Carcinogenesis (Lond.). •/:1565-1568, 1983. Dolan, M. E., Norbcck, L., Clyde. C.. Hora, N. K., Erickson, L. C., and Pegg, A. E. GSH or GST (46), we have applied biochemical assays. Interestingly, Expression of mammalian O"-alkylguanine-DNA alkyltransferase in a celi line sen slightly increased GSH content and GST activity were observed in sitive lo alkylating agents. Carcinogenesis (Lond.), 10: 1613-1619. 1989. leukemic blast cells compared to CLL lymphocytes, possibly contrib 13. Pieper, R. O., Futscher. B. W., Dong, O., and Erickson, L. C. Effects of streptozo- tocin/bis-chloroethylnitrosourea combination therapy on (/'-methylguanine DNA uting to increased chemoresistance to alkylating agents in leukemic methyltransferase activity and mRNA levels in HT-29 cells in vitro. Cancer Res., 51: blasts in vitro. 1581-1585. 1991. It is not yet known whether the wide range of interindividual DNA Dolan, M. E., Mitchell, R. B.. Mummert. C., Moschel. R. C., and Pegg, A. E. Effect of O"-benzylguanine analogues on sensitivity of human tumor cells to the cytotoxic repair capacities observed may be clinically exploited in cases of drug effects of alkylating agents. Cancer Res., Si: 3367-3372, 1991. resistance to alkylating agents. In AML specimens, no obvious clin Gerson, S. L., and Trey, J. E. Modulation of nilrosourea resistance in myeloid leukemias. Blood, 71: 1487-1494, 1988. ical value for DNA repair time was found in relation to treatment Walker, M. C., Masters, J. R. W., and Margison, G. P. O"-alkylguanine-DNA outcome. This is probably due to the fact that MDR-related drugs but alkyltransfcrase activity and nitrosourea sensitivity in human cancer cell lines. Br. J. not alkylating agents were contained in treatment regimens for these Cancer. 66: 840-843, 1992. Silber, J. R.. Bobola. M. S., Ewers, T. G., Muramoto, M.. and Berger, M. S. patients. Other mechanisms, such as PGP expression, might be more O^-alkylguanine-DNA alkyltransferase is not a major determinant of sensitivity to important in mediating resistance to specific classes of anticancer 1.3-bis(2-chloroethyl)-l-nitrosourea in four medulloblasloma cell lines. Oncol. Res., drugs used in the treatment of AML. The higher incidence of PGP 4: 241-248, 1992. Samson. L.. Thomalc, J.. and Rajewsky, M. F. Alternative pathways for the in vivo expression observed in nonresponsive AML patients would seem to repair of C/'-alkylguanine and O4-alkylthymine in E. coli: the adaptive response and strengthen this assumption. However, cell samples derived from CLL nucleotide excision repair. EMBO J., 7: 2261-2267, 1988. patients clinically resistant to treatment with alkylating agents dis Bronstein, S. M.. Skopek, T. R., and Swenberg, J. A. Efficient repair of (/'- played a higher rate of O''eGua elimination from nuclear DNA in cthylguanine. but not (/"-ethylthyminc or O2-ethyllhymine, is dependent upon (/'- alkylguanine-DNA alkyltransfcrase and nucleotide excision repair activities in human comparison to the mean value for the responsive patients. This ob cells. Cancer Res.. 52: 2008-2011, 1992. servation suggests a potential clinical significance of DNA repair Batist. G., Torres-Garcia, S.. Demuys, J. M., Greene, D., Lehnen. S.. Rochon. M., and Panasci, L. Enhanced DNA cross-link removal: the apparent mechanism of resistance pathways in mediating resistance to alkylating agents in CLL lym in clinically relevant melphalan resistant human breast cancer cell line. Mol. Phar- phocytes. However, fast elimination rates were also found in some mol.. 36: 224-230, 1989. Dabholkar. M., Bostick-Bruton. F.. Weber, V., Egwuagu, C.. Bohr. V. A., and Reed. specimens derived from untreated or treated sensitive patients. Due to E. Expression of excision repair genes in non-malignant bone marrow from cancer the relatively small number of patients studied thus far, we cannot patients. Mutât.Res. 29.?: 151-160, 1993. exclude biased selection of patients; therefore, future studies should Geleziunas R.. McQuillan A., Malapetsa A.. Hulchinson. M., and Kopriva, D. Increased DNA synthesis and repair-enzyme expression in lymphocytes from patients encompass a larger number of specimens. It is, however, clear that with chronic lymphocytic leukemia resistant to nitrogen mustards. J. Nati. Cancer measurements of the repair kinetics of specific DNA adducts provide Inst., 83: 557-564, 1991. a sensitive, functional assay to evaluate DNA repair capacity in Torres-Garcia, S. J., Cousineau. L., Caplan, S.. and Panasci. L. Correlation of resistance to nitrogen mustards in chronic lymphocytic leukaemia with enhanced malignant cells derived from patients. Interindividual differences in removal of melphalan-induced cross-links. Biochem. Pharmacol., 38: 3122-3123, DNA repair rates may be exploited for the design of individualized 1989. chemotherapy regimens. Begleiter. A.. Goldenberg. G. J.. Anhalt, C. D., Lee, K., Mowat, M. R., Israels, L. G., and Johnston, J. B. Mechanisms of resistance lo chlorambucil in chronic lymphocytic leukemia. Lcuk. Res.. 15: 1019-1027, 1991. Joncourt. F., Oberli, A., Redmond. S. M. S., Fey, M. F., Tobler, A., Margison, G. P., ACKNOWLEDGMENTS Gratwohl, A.. Buser. K., and Cerny, T. Cytoslatic drug resistance: parallel assessment of glutathione-based detoxifying enzymes, O''-alkylguanine-DNA-alkyltransferase We are grateful to Dipl.Ing. K. Lennartz for advice regarding flow cyto- and P-glycoprotein in adult patients with leukaemia. Br. J. llaemalol., 85: 103-111, metric methods, to Bettina Baumgart for excellent technical assistance, to 1993. 26. Olive, P. L., Wlodek, D., and Banáth,J. P. DNA double-strand breaks measured in M. R. Nowrousian for helpful discussions, and to Anja Arenas for constant individual cells subjected to gel electrophoresis. Cancer Res., 51: 4671-4676, 1991. support. 27. Frankfurt. O. S.. Seckinger, D., and Sugarbaker. E. V. Flow cytometric analysis of DNA damage and repair in the cells resistant to alkylating agents. Cancer Res., 50: 4453-4457, 1990. REFERENCES 28. Müller.M. R., Wright, K. A., and Twentyman, P. R. Differential properties of cisplatin and tetraplatin with respect to cytotoxicity and perturbation of cellular 1. Epstein. R. J. Drug-induced DNA damage and tumor chemosensitivity. J. Clin. glutathione levels. Cancer Chemothcr. Pharmacol., 28: 273-276, 1991. Oneol., 8: 2062-2084, 1990. 29. Yang, W. Z., Begleiter. A., Johnston, J. B., Israels, L. G., and Mowat, M. R. A. Role 2. Morrow, C. S.. and Cowan. K. H. Glutathione S-transferases and drug resistance. of glutathione and glutathione-S-transferase in chlorambucil resistance. Mol. Phar Cancer Cells, 2: 15-22, 1990. macol.. 41: 625-630, 1992. 3. Colvin, M.. and Chabner. B. A. Alkylating agents, in: B. A. Chahner and J. M. 30. Schisselbauer. J. C., Silber, R., Papadopoulos, E., Abrams, K., LaCreta. F. P., and Collins (eds.). Cancer Chemotherapy Principles and Practice, pp. 276-314. Philadel Tew, K. D. Characterization of glutathione 5'-transferase expression in lymphocytes phia: Lippincott Co., 1990. from chronic lymphatic leukemia patients. Cancer Res., 50: 3562-3568, 1990. 4. Pegg. A. E. Mammalian O"-alkylguaninc-DNA alkyltransfcrase: regulation and im 31. Bradley, G.. Juranka. P. F.. and Ling, V. Mechanisms of mullidrug resistance. portance in response lo alkylating carcinogenic and therapeutic agents. Cancer Res., Biochim. Biophys. Acta, 948: 87-128, 1988. 50: 6119-6129, 1990. 32. Pirker, R.. Waliner, J., Geissler, K., Linkesch, W., Haas, O. A., Bettelheim, P., 5. Karran, P., Macpherson, P., Cecotti, S., Dogliotti, E., Griffin, S.. and Bignami, M. Hopfner, M.. Scherrer, R., Valent, P., Havelec, L., Ludwig, H., and Lechner. K. 0(>-Methylguaninc residues elicit DNA repair synthesis by human cell extracts. J. MDR1 gene expression and treatment outcome in acute myeloid leukemia. J. Nail. Biol. Chem., 26«:15878-15886, 1993. Cancer Insl., 83: 708-712, 1991. 6. Kalamegham, R.. Warmels-Rodenhiser. S.. MacDonald, H.. and Ebisuzaki, K. O*- Müller.M. R.. Lennartz, K., Nowrousian, M. R., Dux, R., Tsuruo, T., Rajewsky, 4530

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. DNA REPAIR CAPACITY OF CLL LYMPHOCYTES AND AML BLAST CELLS

M. F., and Seeber, S. Improved flow-cytometric detection of low P-glycoprotein enzymatic step in mercapturic acid formation. J. Biol. Chem., 249: 7130-7139, 1974. expression in leukemic blasts by histogram subtraction analysis. Cytometry, /5: 41. Scherer, E., Van den Berg, T., Vermeulen, E., Winterwerp. H. H. K., and Den 64-72, 1994. Engelse, L. Immunocytochemical analysis of O^-alkylguaninc shows tissue specific 34. Lambert. E., Rees, J. K. H., and Twenlyman, P. R. Resistance circumvention strat formation in and removal from esophageal and liver DNA in rats treated with egies tested in clinical leukemia specimens using the MTT colorimetrie assay. methylbenzylnitrosamine, dimcthylnitrosamine, diethylnitrosaminc and ethylnitro- Leukemia (Baltimore), 6: 1063-1071, 1992. sourea. Cancer Lett., 46: 21-29, 1989. 42. Wani, G., Wani. A. A., and D'Ambrosio, S. M. Cell type-specific expression of the 35. Kirkpatrick. D. L., Duke, M., and Goh. T. S. Chemosensitivity testing of fresh human leukemia cells using both a dye exclusion assay and a tetrazolium dye (MTT) assay. O''-alkylguanine-DNA alkyltransferase gene in normal human liver tissues as re Leuk. Res., 14: 459-466, 1991). vealed by in situ hybridization. Carcinogenesis (Lond.). 14: 737-741, 1993. 36. Rai, K. R., Sawilsky, A., Cronkite, E. P., Chanana, A. D., Levy, R. N., and 43. Brent, T. P., von Wronski, M. A., Edwards, C. C, Bromley, M., Margison, G. P., Pasternack, B. S. Clinical staging of chronic lymphocytic leukemia. Blood, 46: Rafferty, J. A., Pegram, C. N., and Bigner, D. D. Identification of nitrosourea- 219-234, 1975. resistant human rhabdomyosarcomas by in situ immunostaining of C^-methylgua- 37. Seiler, F., Kirstein, U., Eberle, G., Hochleitner, K., and Rajewsky, M. F. Quantifi nine-DNA methyltransferase. Oncol. Res., 5: 83-86, 1993. cation of specific DNA O^-alkylation products in individual cells by monoclonal 44. Zhen, M., Link, C. J., O'Connor, P. M., Reed, E., Parker, R., Howell, S. B. and Bohr, antibodies and digital imaging of intensified nuclear fluorescence. Carcinogenesis V. A. Increased gene-specific repair of cisplatin interstrand cross-links in cisplatin- (Lend.). 9: 1907-1913, 1993. resistant human ovarian cancer cell lines. Mol. Cell. Biol., 11: 3095-3104, 1992. 38. Thomale, J., Seiler, F., Müller,M. R., Seeber, S., and Rajewsky. M. F. Repair of 45. Thomale, J.. Hochleitner. K., and Rajewsky. M. F. Differential formation and repair O6-alkylguanines in the nuclear DNA of human lymphocytes and leukaemic cells: of the mutagenic DNA alkylation product O''-cthylguanine in transcribed and non- analysis at the single cell level. Br. J. Cancer, 69: 698-705, 1994. transcribed genes of the rat. J. Biol. Chem., 269.- 1681-1686, 1994. 39. Tietze. F. Enzymic method for quantitative determination of nanogram amounts of 46. Cook, J. A., lype, S. N., and Mitchell, J. B. Differential specificity of monochloro- total and oxidized glutalhione. Anal. Biochem., 27: 502-522, 1969. bimane for isozymes of human and rodent glutathione S-transferases. Cancer Res., 40. Habig, W. H., Pabst, M. J., and Jakoby, W. B. Glutathione 5-transferases: the first 51: 1606-1612, 1991.

4531

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research. Capacity of Individual Chronic Lymphatic Leukemia Lymphocytes and Leukemic Blast Cells for Repair of O6 -Ethylguanine in DNA: Relation to Chemosensitivity in Vitro and Treatment Outcome

Mark R. Müller, Frank Seiler, Jürgen Thomale, et al.

Cancer Res 1994;54:4524-4531.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/54/16/4524

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/54/16/4524. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1994 American Association for Cancer Research.