Multiple myeloma (MM) is a hematological malignancy of plasma cells, the mature B lymphocytes producing immunoglobulins. Proteasome inhibitors (PIs) and immunomodulatory drugs (IMiDs) such as Bz and Len are the backbone agents of a MM combination therapy significantly prolonging patient survival (1). However, newly diagnosed patients can display primary refractory MM and most of those initially treated will eventually develop drug-resistant MM, with a poor prognosis (2, 3).

The anti-MM activity of Bz primarily relies on its ability to induce a terminal unfolded protein response (UPR) via toxic accumulation of misfolded immunoglobulins in the endoplasmic reticulum (ER) (4). For its part, Len targets the Cereblon (CRBN) component of an E3 ubiquitination complex, provoking the selective degradation of the tumor pro-survival transcription factors IKZF1s, and leading to MM cell cycle arrest and apoptosis (5). While the molecular basis of MM resistance to PIs is complex, that of IMiD is typically associated with altered CRBN expression (6).

Using mycolactone (Myco) as a model inhibitor, we recently reported that Sec61 - the channel mediating secretory protein import into the ER - is a therapeutic vulnerability in MM (7). By preventing the translocation of newly synthesized secreted and transmembrane proteins into the ER, Myco provokes their cytosolic degradation by the proteasome (8–11). In MM cell lines and patient-derived tumors, Sec61 blockade by Myco triggered an atypical, pro-apoptotic ER stress response synergizing with Bz for induction of MM cell death in vitro and in vivo (7). Using a Bz-resistant version of the model MM cell line MM1S (12), we showed that Sec61 blockade overrides MM resistance to PIs. Here, we investigated the therapeutic interest of combining Sec61 inhibition to Len and Len+Bz combinations, in both chemo-naïve and -resistant MM.

Len-resistant (LenR) and double, Len- and Bz-resistant (BzR LenR) cell lines were generated by CRISPR/Cas9 knockout of CRBN in MM1S and BzR daughter (7), respectively. Both were fully defective for production of CRBN ( Figure 1A ) and protected against Len-induced degradation of IKZF1.1 and IKZF1.2 ( Figure 1B ). In both the parental MM1S and BzR cell lines, Len had a marked anti-proliferative effect after 24h of treatment ( Figure 1C ). On the opposite, the growth of LenR and BzR LenR MM1S was unaffected by Len ( Figure 1C ). While 100% mortality was achieved in MM1S cells exposed to 80 nM Bz for 48h, the cytotoxicity of Myco only manifested after 72h at concentrations ≥ 50 nM ( Figure 1D and 7). The anti-MM activity of Len was even slower, in accordance with previous reports (5). A maximal viability loss of 50% was obtained after a 120h treatment with 80 µM Len in parental and BzR MM1S cells, from which CRBN ko cells were protected ( Figure 1D , left). In contrast, MM1S and its LenR daughter were equally sensitive to Bz and Myco ( Figure 1D , middle and right). It is interesting to note that BzR LenR and BzR MM1S cells displayed an increased sensitivity to Myco at the highest tested concentrations, compared to parental MM1S cells ( Figure 1D , right). Blocking Sec61 thus overcomes MM cell resistance to IMiDs, and even more if it is combined with resistance to PIs.

The clinical efficacy of PIs primarily relies on their capacity to trigger transition from adaptive to terminal UPR in MM, through induction of the Activating Transcription Factor 4 (ATF4) and its pro-apoptotic target C/EBP homology protein (CHOP). Other hallmarks of Bz activity include elevated splicing of X-box binding protein 1 (XBP1) mRNA into transcriptionally active sXBP1 and induction of the ER-resident chaperone BiP, a master regulator of the UPR controlling the threshold of apoptosis induction (18). We previously reported that Myco synergizes with Bz by inducing an atypical UPR marked by hyper-activation of ATF4/CHOP signaling and XBP1 splicing, in the absence of BiP upregulation (7). Figure 1E shows that the distinctive ER stress signatures of Bz, Myco and Bz+Myco were maintained in LenR MM1S cells ( Figure 1E ). In comparison, BzR LenR MM1S displayed a relatively more important induction of ATF4 gene expression in response to Myco treatment ( Figure 1F ). In none of the two CRBN ko cell lines did a co-treatment with Len interfere with these stress marks. Therefore, both the proteotoxic effects of Sec61 blockade and synergy with proteasome inhibition withstand MM resistance to Bz and Len.

To determine if Myco interferes with the anti-MM activity of Len, we next assessed the cytotoxicity of Len/Myco combinations in MM1S and its BzR version ( Figure 2A ). After 96h of exposure, Len marginally affected MM cell viability while Myco displayed a clear dose-dependent cytotoxicity. Notably, in both cell types the two drugs synergized to induce MM cell death ( Figure 2B ), highlighting the interest of combining Sec61 blockers with IMiDs in MM treatments.

We went on to assess the impact of bi- and tri-therapies in both parental MM1S and its single- and double-drug resistant daughters. Because Len takes 120h to display cytotoxicity, cells were treated with Len alone for 96h, then Myco and Bz were added for the next 24h. In all cell lines, the Myco/Bz combination was highly cytotoxic ( Figure 2C ), confirming our previous findings obtained with parental and BzR MM1S (7), and extending them to their LenR and BzR LenR counterparts. Notably, adding Len to the Myco/Bz combination further increased the induction of apoptosis in the two Len-susceptible MM1S cell lines. Altogether, these results consolidated the potential of blocking Sec61 in drug-resistant MM.

To evaluate the interest of adding Myco to the Bz/Len bi-therapy in vivo, we engrafted MM1S on the one hand, and its BzR LenR counterpart on the other hand, in immunodeficient NOD/SCID/IL2rγnull (NSG) mice as described in Domenger et al. (7). Mice were randomized in eight groups receiving Len, Myco and Bz as single, double, or triple drug combinations twice weekly. In the conditions used, none of the monotherapies did influence the growth of MM1S cells ( Figure 3A ). In contrast, the Bz/Myco combination significantly delayed the progression of both MM1S and its BzR LenR version. In both systems, the tri-therapy was significantly superior to the clinically used Bz/Len combo. The Myco/Bz/Len treatment, and Myco/Bz to a lower extent, nevertheless induced toxic effects, with elevated levels of serum ALT suggesting hepatotoxicity ( Figure 3B ).

Using the Sec61 inhibitor Myco and a combination of MM cell lines, patient-derived tumors and xenograft mouse models of disease, we previously demonstrated that pharmacological blockade of Sec61 triggers MM cell-selective apoptosis through induction of ER stress (7). Inhibiting Sec61 also impaired MM cell secretion of immunoglobulins and surface expression of pro-survival receptors. Moreover, it synergized with Bz for MM cell killing and this synergy extended to B cell acute lymphoblastic leukemia. Collectively, these data established the translational potential of Sec61 blockers as novel anti-MM agents. Beyond MM, they uncovered the interest of targeting both the translocon and the proteasome in proteostasis-addicted tumors (7). This study also revealed that Sec61 blockade kills MM cells irrespective of their innate or acquired resistance to Bz (7). In addition to MM1S, Myco treatment induced apoptosis in the JIM3 and KMS11 cell lines, which display a relatively higher resistance to Bz. Myco also showed potent cytotoxicity in MM cells isolated from 6 patients, with variable resistance to Bz. By generating a Bz-resistant daughter of MM1S, we could formally demonstrate that the anti-MM activity of Myco overrides resistance to proteasome inhibition.

Since the current standard of MM care combines PIs with IMiDs, it was important to determine how Sec61 blockade interferes with- and whether it overrides resistance to- IMiDs. We show in the present study that both the toxicity of Myco and synergy with Bz operate in LenR and BzR LenR MM1S cells. Interestingly, at concentrations ≥ 200 nM Myco was more effective in BzR cells, and even more in BzR LenR cells, than in the parental cell line. In addition to activating the UPR, we reported previously that Myco triggers pro-apoptotic oxidative stress responses through depletion of intracellular glutathione pools (19). Interestingly, recent proteomic studies of MM cells have associated acquired resistance to PIs with the generation of important oxidative stress (20, 21). This may render BzR cells more susceptible to the lethal oxidative activity of Myco. Furthermore, deletion of CRBN was shown to promote PERK signaling (22), providing a possible explanation for the superior susceptibility of BzR LenR cells, compared to BzR cells, to Myco-mediated induction of terminal UPR.

Notably, we also observed a synergy between Myco and Len in parental and BzR MM1S. Furthermore, Myco enhanced the efficacy of the PI/IMiD bitherapy in mice engrafted with MM1S cells, irrespective of their resistance to PIs and IMIDs. While confirmatory studies with more MM cell lines will be needed, these data support the interest of further evaluating Sec61 blockers in MM drug combinations. They strongly suggest that MM patients developing resistance to PIs and/or IMiDs will respond to Sec61 blockade therapy.

While a useful tool to study the consequences of Sec61 blockade in vitro and in vivo, Myco is a complex natural product whose structure and physicochemical properties are incompatible with drug development (23, 24). Furthermore, its broad-spectrum inhibitory activity on Sec61 substrates confers Myco with toxic effects on non-cancerous cells that limit its therapeutic window (8, 10, 25). Importantly, newly generated small molecule inhibitors of Sec61 are currently being evaluated in Oncology phase 1 trials, paving the way to the first generation of drugs targeting protein translocation (26). The work presented here suggests that blocking Sec61 is a promising therapeutic avenue for the treatment of refractory or relapsed MM.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by the CETEA ethics committee number 0068 (Institut Pasteur, Paris, France, DAP210001) and received the approval of the French Ministry of Higher Education and Research.

AD, DR and VM performed the experiments. LM generated the lentiviral particles for CRISPR/Cas9 knock out of CRBN. AD, CM, GD, and CD designed the experiments. Data interpretation and writing of the manuscript were performed by AD and CD. All authors contributed to the article and approved the submitted version.