Poly(ADP-ribose) polymerase-1 gene in human tumor cell lines: Its expression and structural alteration

Poly(ADP-ribose) polymerase-1 (Parp-1) is involved in DNA repair and cell-death induction after DNA damage. Parp-1−/− mice show higher susceptibility to the carcinogenic effects of nitrosamine and azoxymethane. To elucidate the role of alterations of the PARP-1 gene in human carcinogenesis, we examined the expression level of PARP-1 gene in various human tumor cell lines. The presence of gross rearrangement of PARP-1 gene in these cell lines was also examined by Southern blot hybridization analysis. The expression levels of PARP-1 gene in several cell lines, including T-cell leukemia cell lines (Molt-4 and CCRF-CEM), colon cancer cell line (WiDr), and gastric cancer cell lines (KATOIII, OKAJIMA, and MKN45) was substantially lower than in other cancer cell lines. Among the 85 analyzed cell lines, structural alteration of PARP-1 gene was detected in a gastric cancer cell line, MKN28. A low level of PARP-1 expression in human cancer could potentially influence cancer cell growth, differentiation and cancer development by affecting genomic instability, as well as the response of tumors to chemo- and radiotherapy.

[Vol. 80, Introduction. Poly(ADP-ribose) polymerase-1 (PARP-1) catalyzes polyADP-ribosylation of proteins using NAD as a substrate after activation by DNA strand breaks. 1)-3) PARP-1 has a regulatory role in base excision repair (BER) and DNA strand break repair by interacting with proteins involved in these repair pathways and modifying them. 4) In Parp-1 -/mice, a higher frequency of deletion mutations was observed than in Parp-1 +/+ mice after treatment with carcinogenic Nnitrosobis(2-hydroxypropyl)amine (BHP), 5) suggesting that enhanced genomic instability contributes to higher susceptibility to the carcinogenic effects of BHP in Parp-1 deficiency. 6) PARP-1 is also activated by DNA damages, its activation induces depletion of cellular NAD level 7) and consequently stimulates apoptosisinducing factor-dependent cell death. 8) Although the evidence suggests that PARP-1 is likely to be involved in human carcinogenesis, alteration of the PARP-1 gene in human cancer has not been fully studied yet, but several reports described the level of PARP-1 gene expression 9) and alteration of the structure of PARP-1 gene in various malignancies. 10)-12) For example, Bhatia et al. 10)-12) reported that the allele frequency of a polymorphism in the processed pseudogene of PARP-1 on chromosome 13q33-qter is increased in Burkitt lymphoma, multiple myeloma, colon and prostate cancers in African American populations. Furthermore, Bieche et al. 13) reported that low PARP-1 gene expression correlated with higher genomic instability in breast cancer. Prasad et al. 9) reported that in comparison to other tumor cell lines, Ewing's sarcoma cell lines exhibited a high constitutive level of PARP-1 activity and an increased level of PARP-1 mRNA expression.
Results and discussion. PARP-1 mRNA expression was analyzed by northern blot analysis in 9 colon tumor cell lines, 7 liver cancer cell lines, 6 leukemia cell lines, and 4 sarcoma cell lines (Fig. 1A & B). Colon cancer cell line (WiDr), 25)  2) PARP-1 activity was measured by incorporation of radioactivity derived from 32 P-NAD into an acid-insoluble fraction. Data are mean ± SE. and MKN45, 29) also demonstrated low expression levels of PARP-1 gene compared to six other gastric cancer cell lines ( Fig. 2A, upper panel). The low level of PARP-1 gene expression in these tumor cell lines could be due to either inactivation of PARP-1 promotor by DNA hypermethylation or mutation, or by downregulation of some transcription factors involved in the regulation of PARP-1 gene expression. Furthermore, MKN28 29) was found to harbor two extra fragments of PARP-1 as indicated by Southern blot hybridization analysis after EcoRI digestion of genomic DNA ( Fig. 2A, lower panel, Arrows "a" & "b"). We observed the presence of an extra fragment of PARP-1 also after digestion with HindIII, BamHI, and XbaI (data not shown). These results suggest that MKN28 contains the structural alteration in the PARP-1 gene. When the PARP-1 cDNA probe, encompassing the entire coding region, was divided into N-terminal, automodification, and C-terminal domains, the extra fragment "a" hybridized only to the C-terminal probe, whereas the fragment "b" hybridized to both automodification and C-terminal probes. This suggests that the structural alteration spans the auto-modification to C-terminal domains of the PARP-1 gene. The probe used for Southern blot hybridization analysis is a cDNA encompassing the whole coding region of the PARP-1 gene and it detects PARP-1 gene on chromosome 1q41-q42 and pseudogenes on chromosome 13q33-qter and on chromosome 14q22. 11) Further studies are necessary to clarify whether the rearrangement also involves the pseudogenes. Little or no information is available on whether the rearrangement is continuous or not and whether it is caused by intragenic or intergenic recombination. Since the densities of fragments "a" and "b" were approximately half compared to other bands, the rearrangement was considered as a monoallelic event.
We measured PARP-1 activity using cell extracts of MKN28 and MKN45, as shown in Fig. 2C, and found that the former had a four-fold higher activity of PARP-1 than the latter. That the activity pattern correlated with PARP-1 mRNA expression level implies that the structural alteration in MNK28 did not affect the basal activity of PARP-1. It is yet to be clarified whether the enzymatic activity or the function of PARP-1 under DNA damaging conditions is influenced by this structural alteration of the PARP-1 gene in MKN28.
To determine the role of PARP-1 dysfunction in carcinogenesis, further analysis of PARP-1 gene mutation and polymorphism should be conducted at the sequence level by refined methods, including PCR-SSCP (single-strand conformation polymorphism) and direct sequencing. Reduced polyADP-ribosylation activity was demonstrated in patients with Werner Syndrome after treatment with an alkylating agent, methylmethanesulfonate, but was not observed without any treatment or after treatment with bleomycin, 30) indicating that aberration of PARP-1 activity should be examined after treatment with various types of DNA damaging agents as well.
In conclusion, we identified in this study those cancer cell lines with low level of expression and structural alteration of PARP-1 gene. A low level of PARP-1 expression in human cancer could have a substantial impact on cancer cell growth, differentiation and cancer development by affecting genomic instability, as well as on the response of tumors to chemo-and radiotherapy.