Regulation of DNA-Dependent Protein Kinase Activity in Leukemic Cells

Regulation of DNA-dependent protein kinase activity in leukemic cells

Catherine Muller1,2 and Bernard Salles2

1Service d'heÂmatologie, CHU Purpan, 31300 Toulouse, France and 2Institut de Pharmacologie et de Biologie Structurale, CNRS UPR 9062, 205 route de Narbonne, 31077 Toulouse Cedex, France

The DNA-dependent protein kinase (DNA-PK) complex in DNA-PK activity measured in extracts from various is composed of a catalytic (DNA-PKcs), and a regulatory human cell lines were observed leading to the idea that subunit (Ku70/Ku86 heterodimer). The expression and DNA-PK expression was constitutively expressed in function of DNA-PK subunits was investigated in cells. In a ®rst step to explore this issue, we have puri®ed blood lymphocytes obtained from patients with chosen to study DNA-PK activity in lymphocytes chronic lymphocytic leukemia (CLL) either refractory to obtained from patients with chronic lymphocytic chemotherapy or untreated. Variations in DNA-PK leukemia (CLL), a disease state characterized by the activity were found amongst CLL samples by compar- proliferation of abnormal, developmentally immature ison to human cell lines. It was noticeable that the low B cells or, exceptionally T lymphocytes (Kipps, 1995). DNA-PK activity was associated with samples from CLL seemed to be an interesting tumor model for this untreated patients that exhibited a sensitivity phenotype, study since blood samples usually provide large determined in vitro, to the radiomimetic agent neocarci- quantities of cells that can be easily puri®ed leading nostatin by comparison to samples from refractory to a pure tumoral cell population. In the present study, patients. The regulation in DNA-PK activity was we have examined the expression and function of associated with Ku heterodimer expression while DNA- DNA-PK subunits in lymphocytes from CLL patients

PKcs was unaected. Moreover, the presence of an in order to characterize DNA-PK activity in fresh altered form of the Ku86 subunit was identi®ed in leukemic cells. samples with low DNA-PK activity. These results suggest a regulation process of the DNA-PK activity in fresh human cells. Results Keywords: DNA-PK; Ku autoantigen; CLL lympho- DNA-PK activity in lymphocytes from CLL patients cytes The DNA-PK activity was determined by using a highly sensitive assay (Finnie et al., 1995) in whole cell extracts from 9 CLL lymphocyte samples. Extracts Introduction from AHH1, MRC5 and HeLa cells were used to compare the DNA-PK activity measured in cell lines The DNA-dependent protein kinase (DNA-PK) plays versus fresh cells. DNA-PK activity was present in all an important role in the repair of DNA double-strand CLL extracts analysed, but its level varied between breaks (DSB) and in the V(D)J recombination (Finnie samples (Figure 1). In ®ve samples, the DNA-PK et al., 1996; Jeggo et al., 1995; Weaver, 1995). DNA- activity appears similar to that observed in the three PK is a nuclear protein serine/threonine kinase that cell line extracts (representative of the level of activity comprises a large catalytic subunit (460 kDa), DNA- observed in a panel of cell lines of human origin, data

PKcs, and a DNA binding sub-unit, the Ku autoanti- not shown). In contrast, in four samples, the level of gen. The Ku autoantigen, a heterodimer of the Ku 70 DNA-PK activity was signi®cantly lowered to about and Ku 86 proteins, binds to DNA double strand ends 2.6-fold (3107+1033 versus 8458+1180 c.p.m. incor- and other discontinuities in the DNA (Blier et al., porated in the substrate peptide, respectively, 1993; Falzon et al., 1993; Minori and Hardin, 1986) P50.001). and recruits the catalytic subunit of the complex (Gottlieb and Jackson, 1993). The active DNA-PK Sensitivity of CLL lymphocytes to neocarcinostatin and complex then acquires the capacity, at least in vitro,to alkylating agents phosphorylate many DNA-bound proteins in the vicinity (Anderson et al., 1994). Several DNA-binding In order to examine a putative relation between protein have been shown to be substrates for DNA- variations in DNA-PK activity and in vitro cell

PKcs mediated phosphorylation in vitro, including the sensitivity, survival curves were determined by the transcription factor Sp1, c-jun,c-fos and Oct-1, the MTT assay, already validated with these cells tumor suppressor gene p53 and the carboxyl-terminal (Bramson et al., 1995a). Four samples exhibited a domain of RNA polymerase II (Anderson et al., 1994). sensitive phenotype to the radiomimetic agent neocar- Despite many studies with established cell lines little cinostatin (about ®vefold) by comparison with the ®ve is known about DNA-PK activity and its regulation in other samples (Figure 2). Based upon these results, we fresh human cells and tissues. No signi®cant variations de®ned two groups, group I (samples 1 ± 5) and group II (samples 6 ± 9) resistant and sensitive to neocarci-

nostatin, respectively. The mean IC50 to neocarcinosta- Correspondence: B Salles tin was 56.8+11.7 nM versus 9.8+2.3 nM for group I Received 12 May 1997; revised 8 July 1997; accepted 9 July 1997 and group II, respectively (see Table 1). Interestingly, DNA-PK activity in leukemic cells C Muller and B Salles 2344 samples in group I were obtained from refractory Basis of the dierence in DNA-PK activity in patients, while samples in group II were from untreated lymphocytes from CLL patients patients. Despite the small number of samples in each group and intra-group variation, sensitivity to neocar- Variations in DNA-PK activity might stem from cinostatin of group II samples was in accordance to various mechanisms among them dierential Ku/

low DNA-PK activity, by comparison with group I. In DNA binding and DNA-PKcs expression. First, it

addition, as shown in Table 1, lymphocytes from the was shown that the DNA-PKcs (460 000 m.w) poly- ®ve treated-resistant patients (group I) were resistant to peptide was detected in all extracts at almost similar chlorambucil in vitro (about fourfold) by comparison amounts between the dierent samples (Figure 3a). with those from the four untreated patients (group II), Second, the DNA-end binding (DEB) activity of which illustrates an accordance between in vitro cell Ku 70/Ku 86 was determined by electrophoretic sensitivity and clinical data, although it is unknown mobility shift assays (EMSA) since Ku/DNA binding whether the group II patients will be clinically represents the early limiting step for the formation of responsive to chlorambucil. This is not surprising an active DNA-PK complex. The major Ku/DNA since clinical studies have established that 60 to 80% complex was present in cell extracts from AHH1 and of the patients will respond to chlorambucil adminis- HeLa cell lines (complex A) (Figure 3b). Additional tered as ®rst line therapy (Kipps, 1995). To further slower migrating complexes (A' and A'') were observed con®rm that in our series resistance to chlorambucil is that appeared to depend upon Ku concentration and associated with resistance to other bifunctional are likely to correspond to the binding of multiple Ku alkylating agents as previously described (Bramson et heterodimers to the same DNA-probe ( deVries et al., al., 1995a) the sensitivity of CLL lymphocytes to 1989; Yaneva and Jhiang, 1991). When the CLL mechlorethamine was tested. The group I samples were extracts were considered the results obtained consis- found to be 3.4-fold more resistant to mechlorethamine tently showed two features (Figure 3b): (i) the Ku DEB than those from group II (Table 1). activity was higher in group I than in group II when

Figure 1 DNA-PK activity in CLL lymphocytes. Extracts (50 mg) were analysed by the DNA-PK pull-down assay in the presence of wild-type or mutated peptide as indicated in Material Figure 2 Survival to neocarcinostatin treatment. Survival was and methods. The established cell lines corresponded to AHH1, determined as reported in the Materials and methods section by HeLa and SKOV3 cells. Samples from nine patients were tested the MTT procedure. Lymphocytes from the nine patients (1 ± 9) (1 ± 9). Each value was the mean of at least two independent cell were tested. Each value was the mean of at least two independent extracts per sample experiments per sample

Table 1 Sensitivity of CLL lymphocytes to cytotoxic drugs Group I samples Group II samples a a Drug IC50 mean+s.e. Range IC50 mean+s.e. Range

Neocarcinostatin 56.8+11.7 nM 42.5772.3 mM 9.8+2.3 nM 6.7716.8 nM Chlorambucil 49.0+15.1 mM 26.2766.5 mM 12.4+4.3 mM 6.5716.2 mM Mechlorethamine 40.8+10.1 mM 26.6754.4 mM 12.0+5.0 mM 5.0717 mM aConcentration of drug that inhibited the cell survival by 50%. The ®ve group I samples and the four group II samples were tested. All values are the mean of at least two independant experiments DNA-PK activity in leukemic cells C Muller and B Salles 2345 CLL 1 CLL 6 a NE PNE NE PNE CLL samples 1 2 3 4 5 7 8 9

DNA-PKCS

b cell lines CLL group 1 CLL group 2 A H 1 2 3 4 5 6 7 8 9

Ku 86 A"" A' A B

free probe Ku 70

Figure 3 Measurement of Ku/DNAPKcs expression. (a) Western blots experiments 50 mg of protein extracts obtained from CLL lymphocytes were electrophoresed on a 8% SDS polyacrylamide gel under reducing conditions and blotted onto a nitrocellulose ®lter. The expression of the DNA-PKcs subunit was detected as indicated in Materials and methods. (b) Band shift assay: 4 mgof protein extracts obtained from human cell lines (A, AHH1; H, HeLa) or CLL were incubated with 0.1 ng 32P-labeled DNA p62 probe in the presence of 1 mg of closed circular plasmid DNA, as a non speci®c competitor. The electrophoretic mobility of the protein-DNA complexes were analysed in 5% polyacrylamide gel as described in Materials and methods. F, position of free DNA Figure 4 Western Blot analysis in lymphocyte extracts revealed probe; A , A',A'' and B, positions of protein-DNA complexes the expression of a variant form of the Ku 86 protein. 20 mgof consisting of DNA-end binding activity nuclear (NE) or post-nuclear protein extracts (PNE) obtained from the indicated samples were subjected to electrophoresis on a 8% SDS polyacrylamide gel under reducing conditions. The proteins were then blotted onto a nitrocellulose ®lter and the formation of complex A and of the corresponding expression of the proteins were detected as indicated in Materials slower-migrating complex (A' and A'' were considered; and methods with indicated antibodies (ii) within group II samples, in addition to the major complex (A, A',A''), a new complex (B) with faster electrophoretic mobility was present. This latter result cytoplasmic extracts, although predominantly in the suggests that an altered form of the Ku heterodimer nuclear fraction of the two samples; (iii) the level of the may be expressed in CLL extracts obtained from nuclear expression of full-length Ku 70 and Ku 86 untreated patients. proteins was higher in the extracts from sample 1 than from sample 6; (iv) in addition to the full-length Ku 86 protein, a second immunoreactive band present as a Expression and function of DNA-PK complex in CLL doublet (at approximately 69 and 71 kDa) was nuclear extracts detected in the nuclear extract obtained from the Another way of regulating DNA-PK activity in CLL untreated patient (no 6) whereas Ku 70 protein was might be a dierential distribution of the proteins detected as a unique band in both samples. This involved in the complex. We examined the nuclear and variant form of the Ku 86 protein was also detected by cytoplasmic compartments of two samples: one from Western blot in two other extracts obtained from group I where the usual pattern of Ku DEB activity is untreated patients (no 7 and 9) (data not shown). observed (sample no 1) and one from group II where Finally, in EMSA experiments complex B found in the pattern of Ku DEB activity revealed the presence sample no 6 was only detected after the fractionation of the altered heterodimer (sample no 6). First, nuclear procedure in the nuclear extract (data not shown). (NE) and post-nuclear (PNE) extract fractions were Thus, these experiments suggest that the altered form analysed separately by immunoblotting for p62 of the heterodimer indeed corresponded to the expression, a component of the TFIIH basal transcrip- expression of a variant form of the Ku 86 protein as tion factor (Svejstrup et al., 1996), in order to validate previously described in HL60 cell line (Han et al., the quality of the sub-cellular fractionnement. As 1996) and that the Ku-86 variant is localized in the expected, its expression was restricted to nuclei in nucleus of CLL lymphocytes. both samples (Figure 4). Concerning DNA-PK protein complex expression, the following results were found

(Figure 4): (i) the DNA-PKcs was predominantly Discussion present in the nuclear fraction with little or none in the cytoplasmic fraction; (ii) the expression of Ku 70 The DNA-dependent protein kinase (DNA-PK) com- and Ku 86 proteins was observed in both nuclear and plex, composed of DNA-PKcs and Ku subunits, plays a DNA-PK activity in leukemic cells C Muller and B Salles 2346 major role in cell survival following ionizing radiation among them inter-strand cross-links. Previous studies and in V(D)J recombination. In established human cell have established that the development of chlorambucil

lines, DNA-PKcs and Ku are abundant proteins and resistance in CLL is associated with enhanced repair of the level of DNA-PK activity detected appears DNA crosslinks which may involve a recombination relatively homogenous within cell lines from dierent dependent pathway (Bramson et al., 1995b; Panasci et origin ((Finnie et al., 1996) and unpublished results) al., 1988). Consequently, DNA-PK activity is likely to leading to the idea that the kinase activity was play a role in this phenotype. Indeed, we have shown constitutively expressed. We showed here, for the ®rst that the increase in DNA-PK activity resulted in time, that DNA-PK activity could be measured in increased resistance to neocarcinostatin and chloram- primary human tumour cells and more importantly bucil, suggesting that the levels of DNA-PK activity is that its level of activity varied amongst CLL samples. implicated in the regulation of DNA DSBs repair in Interestingly, even in advanced disease, cell cycle these cells. However, further studies will be needed analysis consistently showed that most malignant with a large number of samples directly to correlate CLL cells are in G0/G1 phase and very few in S DNA-PK activity, enhanced crosslinks removal and phase (Orfao et al., 1992). Thus, DNA-PK activity drug resistance in CLL. appears to be present even in non-cycling primary In conclusion, this ®rst study on DNA-PK activity human tumour cells. in primary human tumor cells opens promising In CLL cells, changes in DNA-PK activity appeared perspectives. First, our study suggest that changes in to be regulated through a variation in the Ku DNA-PK may play an important role in regulating heterodimer DEB activity but also potentially through tumor treatment response to radiotherapy and DNA the expression of an altered form of the heterodimer cross-linking agents. Secondly, the mechanism of (detected in all the four samples with low DNA-PK regulation of this activity in primary tumor cells is activity, Figure 3b), due to the appearence of a variant likely to involve the relative expression a variant form form of the Ku 86 protein, that could negatively of the Ku 86. regulate the DNA-PK activity in these cells. Recently, the expression of an altered form of the Ku heterodimer with faster electrophoretic mobility as Materials and methods detected by EMSA has been reported in the HL60 cell line (Han et al., 1996). This altered form of the Ku Patient characteristics heterodimer, that contained Ku 70 and a variant Nine patients (Service d'HeÂ matologie, Hopital Purpan, Ku 86 protein, was unable to recruit DNA-PKcs leading to a loss of DNA-PK activity in these cells. Toulouse, France) were enrolled in the study. Table 2 The variant form of the Ku 86 protein detected in the summarizes their main clinical features. All patients had B- cell derived CLL. Patients were either untreated (no 6 ± 9) CLL samples presents similarities with the protein or treated-resistant (no 1 ± 5). Patients with resistant B- detected in the HL 60 cell line. In both studies, the CLL had received 4 ± 6 mg of chlorambucil daily for 6 mAb S10B1 (which recognizes an epitope de®ned by aa months or more. All the patients with resistant disease had 8 ± 221 on the N terminus of Ku 86) detected a doublet previously responded to chlorambucil and were progressive consisting of two lower forms of the Ku 86 protein at the time at the analysis. with apparent molecular weight of 69 and 71 kDa. Thus, our results show that the expression of this Peripheral blood lymphocytes and transformed cell lines. variant Ku 86 is observed in primary human cells and is likely to be an important mechanism of regulation of Blood (50 ml) from patients who gave informed consent DNA-PK activity in vivo. Preliminary experiments was collected into heparinized tubes. Monuclear cell suspensions were obtained after Hypaque-Ficoll (MSL, performed in our laboratory show that the altered Eurobio, France) gradient centrifugation as described form Ku heterodimer is also present in normal previously (Barret et al., 1996). The monocytes were circulating lymphocytes (C Muller et al., unpublished results). Finally, immunoblotting revealed the presence of the two subunits of the Ku heterodimer in the nuclei and cytoplasm of CLL cell extracts. In contrast, it has Table 2 CLL patients characteristics been previously reported that, in normal human Sex/Age Lymphocytes Previous Binet Stage at Patient no (years) (cell/mm3) treatmenta time of testingb lymphocytes, Ku proteins are strictly con®ned to the nucleus (Reeves, 1987; Yaneva and Jhiang, 1991). Treated patients 1 M/51 42 000 CLB B Thus, a dierent pattern of cellular distribution may be 2 M/75 55 000 CLB B associated with the malignant phenotype of the cells 3 F/63 232 000 CLB, CHOP C and these results may suggest that dierences occur 4 M/49 80 000 CLB C between malignant and normal B-lymphocytes with 5 F/57 54 000 CLB C regards to the regulation of Ku expression and Untreated patients 6 F/72 142 000 No B function. 7 M/81 200 000 No B Taking into account the origin of the samples tested, 8 M/53 82 000 No A the decreased kinase activity was observed only in 9 F/68 44 000 No A lymphocytes from untreated patients by comparison aCLB, chlorambucil; CHOP, cyclophosphamide+doxo-rubicine with those from treated-refractory patients. In the light +vincristine+prednisone. bBinet staging is as follows: (A) hemoglo- of the clinical status of the disease, basically the bin 410 g/dl, platelets 41006109/L, two sites or less of lymph node sites clinically involved; (B) hemoglobin 410 g/dl, plate- lymphocytosis, alykylating agents such as chlorambucil lets 41006109/L, three or more lymph node sites clincally are used in ®rst treatment. These drugs covalently bind involved; (C) anemia (hemoglobin 10 g/dl or less) or thrombocyto- to DNA and induce the formation of various adducts penia (platelets 1006109/L or less) DNA-PK activity in leukemic cells C Muller and B Salles 2347 removed by adherence. After centrifugation of the non- was assayed in the presence of the peptide EPPLSQEA- adherent lymphocytes, cell extracts were immediately FADLLKK, part of the p53 protein that is phosphorylated prepared from the pellets. The AHH1, HeLa and MRC5 by DNA-PK and the mutated peptide EPPLSEQA- cell lines have been described previously (Barret et al., FADLLKK as a negative control. Radioactivity asso- 1996). ciated with the mutated peptide was in the range 6 ± 10% of that with the substrate peptide. DNA-PK activity is expressed as the dierence of radioactivity incorporated by Cytotoxicity assay the substrate minus the mutated peptide for a given B-CLL lymphocytes, to a ®nal concentration of 1.56106 to extract. 36106 cells/mL, were seeded into 96-well microculture dishes in aliquots of 100 ml. The lymphocytes were then Ku DNA-end binding activity exposed to increasing concentrations of neocarcinostatin (gift from Dr V Favaudon, Institut Curie, Orsay, France), The band shift assay was performed as described chlorambucil or mechlorethamine (Sigma, France) for 48 h previously (Zhang and Yaneva, 1992). The 123 bp linear at 378Cunder5%CO2. Five replicates of each drug probe used is obtained by AvaI digestion of a 123 bp DNA concentration were performed. After 48 h, growth inhibi- ladder, followed by puri®cation of 123 bp-monomers. tion was evaluated by the MTT colorimetric method as described previously (Muller et al., 1994). Western blot Proteins were subjected to electrophoresis on a 8% SDS- Cell extracts polyacrylamide gel, then electroblotted onto a nitrocellul- Whole cell extracts (WCE) were prepared as described lose ®lter (Bio-Rad Trans-Blot). Western blots experiments previously (Finnie et al., 1995). Nuclear (NE) and post- were performed as described previously (Muller et al., nuclear (PNE) cell extracts were prepared as follows: cells 1994). The antibodies used were as follows: the polyclonal were washed twice in phosphate-buered saline, and antibody directed against Ku 70 protein (provided by Dr suspended in 0.5 to 2 mL of ice-cold hypotonic buer A Falaschi, UNIDO, Trieste, Italy) (Tuteja et al., 1994) was

(10 mM Tris-HCl pH 7.4, 10 mM NaCl, 3 mM MgCl2)in used at a 1 : 2500 dilution (secondary antibody mouse anti- the presence of proteinase inhibitors as described above for rabbit at a 1 : 2500 dilution); the puri®ed monoclonal whole cell extracts. Cells were broken in a Dounce antibodies directed against Ku 86 (mAb S10B1, Interchim, homogenizer and the suspension centrifuged for 2 min at France) (Wang et al., 1993), DNA-PKcs (mAb 18-2 5000 r.p.m. at 48C. The PNE was carefully removed from provided by Dr T Carter, St-John's University, New the nuclear pellet, frozen in liquid nitrogen and stored at York, USA) (Carter et al., 1990), p62 (a subunit of the 7708C. The pellet was suspended in buer A and the transcription factor TFIIH; provided by Dr JM Egly, lysate was layered on the top of 5 mL ice-cold sucrose IGBM, Strasbourg, France) were used at 1 : 1000, 1 : 5000 buer (buer A with 30% sucrose) and puri®ed nuclei were and 1 : 5000, respectively (secondary antibody goat anti- obtained at the bottom of the tubes after centrifugation at mouse at a 1 : 2500, 1 : 5000 and 1 : 5000 dilution, 2000 r.p.m. at 48C for 5 min. The pellet was resuspended in respectively). lysis buer A, and the nuclei were puri®ed by centrifugation through sucrose medium twice. Nuclei were then resuspended in extraction buer and processed as described Acknowledgements above for whole cell extracts to obtain the NE. We thank Dr BT Hill for the critical reading of this manuscript. This work was partly granted by Ligue Nationale contre le Cancer. We gratefully aknowledge Dr DNA-dependent protein kinase activity A Huyhn, Dr F Rigal-Huguet and Dr C Nouvel for DNA-PK `pulldown' kinase assay was performed as supplying the clinical samples. CM was supported by a previously described (Finnie et al., 1995). Each sample fellowship from `Fondation Contre la leukeÂ mic'.

References

Anderson CW, Connelly MA, Zhang H, Sipley JD, Finnie NJ, Gottlieb TM, Blunt T, Jeggo PA and Jackson SP. Leesmiller SP, Sakaguchi K, Ullrich SJ, Jackson SP and (1996). Phil. Trans. Royal Soc. of London Series B -, 351, Appella E. (1994). J. Protein Chem., 13, 500 ± 501. 173 ± 179. Barret JM, Calsou P, Laurent G and Salles B. (1996). Mol. Gottlieb TM and Jackson SP. (1993). Cell, 72, 131 ± 142. Pharmacol., 49, 766 ± 771. Han ZY, Johnston C, Reeves WH, Carter T, Wyche JH and Blier PR, Grith AJ, Craft J and Hardin JA. (1993). J. Biol. Hendrickson EA. (1996). J. Biol. Chem., 271, 14098 ± Chem., 268, 7594 ± 7601. 14104. Bramson J, Mcquillan A, Aubin R, Alaouijamali M, Batist Jeggo PA, Taccioli GE and Jackson SP. (1995). Bioessays, G, Christodoulopoulos G and Panasci LC. (1995a). 17, 949 ± 957. Mutat. Res., 336, 269 ± 278. Kipps TJ. (1995). Chronic Lymphocytic Leukemia and Bramson J, Mcquillan A and Panasci LC. (1995b). Cancer Related Diseases, Beutler E, Lichtman MA, Coller BS Lett., 90, 139 ± 148. and Kipps TJ (eds.), New-York: McGraw-Hill Inc, Carter TH, Vancurova I, Sun I, Lou W and DeLeon S. pp. 1017 1039. (1990). Mol. Cell Biol., 10, 6460 ± 6471. Minori T and Hardin JA. (1986). J. Biol. Chem., 261, 10375 ± deVries E W. , van Driel WG, Bergsma A, Arnberg C and 10379. Van der Vliet PC. (1989). J. Mol. Biol., 208, 65 ± 78. Muller C, Bailly JD, Goubin F, Laredo J, Jarezou JP, Falzon M, Fewell JW and Ku EL. (1993). J. Biol. Chem., Bordier C and Laurent G. (1994). Int. J. Cancer, 56, 749 ± 268, 10546 ± 10552. 754. Finnie NJ, Gottlieb TM, Blunt T, Jeggo PA and Jackson SP. (1995). Proc. Natl. Acad. Sci. USA, 92, 320 ± 324. DNA-PK activity in leukemic cells C Muller and B Salles 2348 Orfao A, Ciudad J, Gonzalez M, San Miguel JF, Garcia AR, Wang J, Chou CH, Blankson J, Satoh M, Knuth MW, Lopez-Berges MC, Ramos F, Del CanÄ izo MC, Rios A, Eisenberg RA, Pisetsky DS and Reeves WH. (1993). Mol. Sanz M and Lopez Borrasca A. (1992). Leukemia, 6, 47 ± Biol. Rep, 18, 15 ± 28. 51. Weaver DT. (1995). Adv. Immunol., 58, 29 ± 85. Panasci L, Henderson D, Torres-Garcia SJ, Skalski V, Yaneva M and Jhiang S. (1991). Biochim. Biophys. Acta, Caplan S and Hutchinson M. (1988). Cancer Res., 48, 1090, 181 ± 187. 1972 ± 1976. Zhang WW and Yaneva M. (1992). Biochem. Biophys. Res. Reeves W. (1987). J. Rheumatol., 14, 97 ± 105. Comm., 186, 574 ± 579. Svejstrup JQ, Vichi P and Egly JM. (1996). Trends Biochem. Sci., 21, 346 ± 350. Tuteja N, Tuteja R, Ochem A, Taneja P, Huang NW, Simoncsits A, Susic S, Rahman K, Marusic L, Chen JQ, Zhang JW, Wang SG, Pongor S and Falaschi A. (1994). EMBO J., 13, 4991 ± 5001.