P. Taverna et al., Characterization of MLH1 and MSH2 DNA mismatch repair proteins in cell lines of the NCI anticancer drug screen, CANC CHEMOT, 46(6), 2000, pp. 507-516
Purpose and methods: The lack of a functional DNA mismatch repair (MMR) pat
hway has been recognized as a common characteristic of several different ty
pes of human cancers due to mutation affecting one of the MMR genes or due
to promoter methylation gene silencing. These MMR-deficient cancers are fre
quently resistant to alkylating agent chemotherapy such as DNA-methylating
or platinum-containing, compounds. To correlate drug resistance with MMR st
atus in a large panel of human tumor cell lines, we evaluated by Western bl
ot the cellular levels of the two MMR proteins most commonly mutated in hum
an cancers, MLH1 and MSH2, in the NCI human tumor cell line panel. This pan
el consists of 60 cell lines distributed among nine different neoplastic di
seases. Results: We found that in most of these cell lines both MLH1 and MS
H2 were expressed, although at variable levels. Five cell lines (leukemia C
CRF-CEM, colon HCT 116 and KM12 and ovarian cancers SK-OV-3 and IGROV-1) sh
owed complete deficiency in MLH1 protein. MSH2 protein was detected in all
57 cell lines studied. Absence of MLH1 protein was always linked to resista
nce to the methylating chemotherapeutic agent temozolomide. This resistance
was independent of cellular levels of O-6-alkylguanine DNA alkyltransferas
e. Based on data available for review in the NCI COMPARE database, cellular
levels of MLH1 and MSH2 did not correlate significantly with sensitivity t
o any standard anticancer drug or with any characterized molecular target a
lready tested against the same panel of cell lines. Conclusion: Based on ev
aluation of 60 tumor cell lines in the NCI anticancer drug screen, MLH1 def
iciency was more common than MSH2 deficiency and was always associated with
a high degree of temozolomide resistance. These data will enable correlati
ons with other drug sensitivities and molecular targets in the COMPARE data
base to evaluate linked processes in tumor drug resistance.