TNF-α Confers Resistance of Myeloma Cells to IMiDs through TRAF2 Degradation

The development of novel agents including immunomodulatory drugs (IMiDs) lenalidomide (Len) and pomalidomide (Pom) has led to improved patient outcome in multiple myeloma (MM); however, acquired resistance to IMiDs commonly underlies relapse during the course of treatment. Previous studies show that IMiDs bind to the CRL4^CRBN ubiquitin ligase cereblon (CRBN) and promote proteasomal degradation of IKZF1 and IKZF3 followed by downregulation of c-Myc and IRF4, resulting in MM cell growth inhibition. Therefore, CRBN is the primary binding target and master regulator of IMiDs sensitivity; however, the molecular mechanisms regulating resistance to IMiDs have not been fully defined. Importantly, some MM cells show resistance to IMiDs despite harboring high CRBN expression levels.

To delineate the molecular mechanisms underlying IMiDs resistance, we first performed genome-wide knockout screening in IMiDs-sensitive MM.1S cells using a CRISPR-Cas9 GeCKOv2 library containing 6 unique sgRNAs against each of 19,050 genes and 4 sgRNAs against each of 1,864 miRNAs. We observed that knockout (KO) of twenty-eight genes and one miRNA were associated with resistance to IMiDs. Of note, all six sgRNAs targeting CRBN were identified as a positive regulator of IMiDs sensitivity, consistent with previous studies. Among these genes, we found that three different sgRNAs targeting TRAF2 were enriched after IMiDs treatment. Therefore, we next individually cloned the sgRNAs of TRAF2 into the Cas9 lentiviral vector, and then re-introduced them into MM.1S cells. Importantly, TRAF2-KO MM.1S cells acquire significant resistance to Pom and Len treatment. To examine whether TRAF2 KO induced IMiDs resistance was CRBN-pathway dependent, we assessed CRBN and its downstream interacting protein levels. TRAF2 KO showed no effect on CRBN expression; moreover, IMiDs can still trigger IKZF1 and IKZF3 degradation, associated with downregulation of IRF4. Taken together, these data suggest that TRAF2 mediates sensitivity to IMiDs in a mechanism independent of CRBN-IKZF1/3 axis.

Since TRAF2 is a member of the TNF receptor associated factor (TRAF) protein family required for activation of several signaling pathways including NF-ĸB and JNK, we next examined the biologic impact of TRAF2 KO. Importantly, TRAF2 KO cells show significantly increased processing of precursor p100 to p52 (NF-ĸB2), resulting in hyperactivation of the non-canonical NF-κB pathway; conversely, p52 KO re-sensitizes the cells to IMiDs treatment. The activity of the canonical NF-κB pathway was not similarly altered in TRAF2 KO cells. These results suggest that TRAF2 predominantly activates non-canonical NF-κB pathway, associated with resistance to IMiDs in MM cells.

We next examined the relevance of TRAF2 downregulation in the context of the bone marrow (BM) microenvironment. Co-culture of MM cells with either bone marrow stromal cells (BMSCs) or culture supernatants (BM-CS) confers resistance to IMiDs. Importantly, TRAF2 is downregulated in BMSCs and BM-CS, suggesting that IMiDs resistance in the BM microenvironment may be mediated by TRAF2 downregulation induced by soluble factors. Cytokine array assay confirmed detectable TNF-α in BM-CS. Indeed, MM cells treated with TNF-α showed decreased TRAF2 expression, associated with resistance to IMiDs. These data demonstrate that TNF-α secreted by BMSCs induces TRAF2 downregulation, thereby inducing IMiDs resistance.

We further analyzed the mechanism of TNF-α-mediated TRAF2 downregulation. Since previous studies have shown that TRAF2 is a substrate of the proteasome, we treated MM cells with TNF-α in the presence and absence of proteasome inhibitor. TRAF2 downregulation by TNF-α was partially abrogated by proteasome inhibitor, associated with accumulation of ubiquitinated TRAF2. These data confirm that TNF-α induces TRAF2 downregulation is due, at least in part to proteasomal degradation.

In conclusion, our data show that TRAF2 is a CRBN-independent regulator of IMiDs sensitivity and provide the preclinical rationale for combining IMiDs with inhibitors of non-canonical NF-κB or TNF-α signaling to overcome IMiDs resistance and improve patient outcome.