Abstract
Mantle cell lymphoma (MCL), characterized by a t(11;14) translocation that results in up-regulation of cyclin D1, is incurable with standard chemotherapy. Recent phase II studies have shown that bortezomib (BZM), an inhibitor of the proteasome, can induce responses in about 50% of pre-treated MCL patients. However, the molecular mechanisms that mediate either chemosensitivity or resistance to BZM in MCL remain largely unknown. In this study, we used a panel of MCL cell lines to investigate molecular mechanisms of response to BZM. In 11 MCL cell lines we found a bimodal pattern of chemosensitivity to BZM; the resistant group, REC-1, Mino, and NCEB-1, had an IC50>10nM (median 12.9 nM), while the sensitive group, Granta-519, JVM-2, Jeko-1, HBL-2, UPN1, SP-53, SP-49, and Z-138, had an IC50<10nM (median 5.9nM). No correlation between BZM resistance and p53 mutations was apparent, arguing against a role for this common chemotherapy resistance mechanism. To test whether the differences in sensitivity to BZM might be mediated by drug export mechanisms we measured P-gp activity using the rhodamine efflux assay. We found that more than 50% of the REC-1 (resistant), SP-53, and SP-49 cells (both sensitive) excluded the rhodamine dye, while all other cell lines showed only minimal or no activity, arguing against a role of P-gp in BZM resistance. To determine whether high proteasome activity or reduced sensitivity of the proteasome to inhibition could cause resistance to BZM we quantified proteasome activity by measuring the cleavage of the labeled substrate LLVY-AMC. Basal proteasome activity was comparable except in the sensitive cell line JVM-2, which had an activity 71% higher than the second highest; dose dependent inhibition was similar in all cell lines. BZM can interfere with components of the NFkB pathway, and this effect has been proposed to mediate cytotoxicity in MCL. We measured the relative activation of the NFkB pathway by quantifying p50, p52, p65, c-Rel and Rel-B nuclear factors with an ELISA assay. Cell lines with high and low expressions of NFkB nuclear factors were equally sensitive, arguing against a major role for this pathway in determining sensitivity to BZM. Heat shock proteins (Hsp) have been reported to confer resistance to BZM in lymphoid cell lines. However, in our hands, Hsp27, Hsp70 and Hsp90 were equally expressed between resistant and sensitive cells, and there was no consistent pattern of regulation of these proteins in response to BZM. Recently, NOXA has been reported to mediate BZM induced apoptosis. Indeed, all cell lines showed up-regulation of protein levels of NOXA when exposed to BZM in excess of their IC50 concentrations, suggesting that determinants of resistance are upstream of NOXA. Given the cardinal role of cyclin D1 in MCL, we hypothesized that effects on cell cycle control could be responsible for chemosensitivity. Upon BZM exposure sensitive cells were arrested in G2/M, whereas resistant cell lines either accumulated in G1 (Mino) or, at higher concentrations, underwent apoptosis without arresting in any specific phase. We conclude that BZM can overcome conventional mechanisms of drug resistance and that an effect on cell cycle control may determine BZM activity in MCL.
Nucleolar stress causes the entry into replicative senescence in budding yeast