Cereblon expression is required for the antimyeloma activity of lenalidomide and pomalidomide

Abstract

The precise molecular mechanism of action and targets through which thalidomide and related immunomodulatory drugs (IMiDs) exert their antitumor effects remains unclear. We investigated the role of cereblon (CRBN), a primary teratogenic target of thalidomide, in the antimyeloma activity of IMiDs. CRBN depletion is initially cytotoxic to human myeloma cells, but surviving cells with stable CRBN depletion become highly resistant to both lenalidomide and pomalidomide, but not to the unrelated drugs bortezomib, dexamethasone, and melphalan. Acquired deletion of CRBN was found to be the primary genetic event differentiating isogenic MM1.S cell lines cultured to be sensitive or resistant to lenalidomide and pomalidomide. Gene expression changes induced by lenalidomide were dramatically suppressed in the presence of CRBN depletion, further demonstrating that CRBN is required for lenalidomide activity. Downstream targets of CRBN include interferon regulatory factor 4 (IRF4) previously reported to also be a target of lenalidomide. Patients exposed to, and putatively resistant to, lenalidomide had lower CRBN levels in paired samples before and after therapy. In summary, CRBN is an essential requirement for IMiD activity and a possible biomarker for the clinical assessment of antimyeloma efficacy.

Figure 1

CRBN knockdown induces myeloma cell cytotoxicity. (A) Five lentiviral CRBN shRNA constructs were cotransfected with Flag-tagged CRBN cDNA expression constructs into 293 cells, and Western blot was performed at day 3 after transfection to measure flag-tagged CRBN level. Lentiviruses were prepared from 3 lentiviral CRBN shRNA constructs (1, 10, and 13) and the corresponding NT constructs and subsequently were used to infect 4 different myeloma cell lines. CRBN knockdown was detected by quantitative PCR at day 3 after infection (B), and the cell viability was measured by MTT assay at day 6 after infection (C). Cell cycle (D) and apoptosis assays (E) were performed on OPM2 cells at day 3 and day 5 after lentiviral infection, respectively.

Figure 2

Suppression of CRBN confers lenalidomide resistance. Three myeloma cell lines were infected with NT and CRBN shRNA-expressing lentivirus. Most CRBN shRNA-expressing myeloma cells died within the first week after infection. An average of 30% of myeloma cells survived, and those cells plus NT control expressing cells were sorted for GFP expression at 3 weeks after infection (A). CRBN expression level in each pair of sorted NT and CRBN shRNA-expressing cells was measured by quantitative PCR (B). (C) Activity of lenalidomide on sorted myeloma cell lines was measured by MTT assay.

Figure 3

Drug resistance mediated by CRBN depletion is specific for IMiDs, not for other unrelated antimyeloma compounds. OPM2 cells stably expressing either NT or CRBN shRNA were seeded and incubated with lenalidomide (A), pomalidomide (B), and other antimyeloma drugs, including bortezomib (C), dexamethasone (D), and melphalan (E) at the indicated concentration, followed by MTT assay at day 3 after adding drugs. Each experimental condition was performed in triplicate and repeated at least once.

Figure 4

Loss of CRBN expression in MM1.S cell line is associated with lenalidomide resistance. (A) Diagram of the karyotype evolution observed from MM1.S (lenalidomide-sensitive) to MM1.Sres cell line (lenalidomide-resistant). Red blocks represent the deletions on chromosome 3p, including CRBN (arrows indicate its location). Although 1 CRBN allele is present in the MM1.Sres cell line, there is a significant reduction of CRBN transcript (B) and lack of protein expression (C). (D) GEP comparison showed CRBN as one of the top 10 underexpressed genes in MM1.Sres compared with MM1.S. The lack of or very low CRBN expression was associated with lenalidomide (E) and pomalidomide (F) resistance.

Figure 5

Low CRBN expression is common in lenalidomide refractory patients. (A) Nine MM patients, whose pretreatment and post-treatment samples were available, were analyzed by real-time PCR for the expression of CRBN in MM cells collected at both diagnosis (D) and relapse (R) stages. The data for each sample represent the mean values of 4 independent experiments (mean ± SD). (B) An analysis of CRBN expression across different differentiation stages of normal B cells and a variety of B-cell tumors, including marginal zone lymphoma (MZL, n = 86), lymphoplasmacytic lymphoma/Waldenström macroglobulinemia (LPL/WM, n = 21), follicular lymphoma (FL, n = 38), diffuse large B-cell lymphoma (DLBCL, n = 73), MM (n = 238), and HMCLs (n = 60), shows a similar expression level across entities without low expression outliers. Arrows indicate the CRBN expression level in 4 HMCLs that show resistance to lenalidomide (OCIMY5, OPM1, SKMM2, and KMS12 PE). (C) The low CRBN gene expression level found in OCIMY5 and OPM1 was correlated with low protein expression level.

Figure 6

GEP identified common changes in OPM2 cells treated either with lenalidomide or with CRBN shRNA. GEP analysis was performed on OPM2 cells either treated with lenalidomide for 48 and 72 hours or transduced with CRBN shRNA (for 48 and 72 hours). (A) The numbers of genes that have expression changed at least 2-fold in both treatments (lenalidomide vs vehicle and CRBN shRNA vs NT shRNA) are summarized at the top. Overall, 123 gene changes were shared between lenalidomide treatment and CRBN knockdown. Those 123 genes were uploaded as the input list for generation of biologic networks using MetaCore pathway analysis. (B) Illustration of one of the top scored networks from active experiments. Green lines indicate activation; red lines, inhibition; and gray lines, unspecified. Red circles represent up-regulated genes; and blue circles, down-regulated genes. (C) Western blot analysis was performed to detect IRF4 expression in HMCLs treated either with lenalidomide or transfected with CRBN shRNA.