Analysis of COVID-19-Associated Proteins Identifies TANK-Binding Kinase 1 As an Immunotherapeutic Target in Multiple Myeloma

Background and Rationale. Patients with multiple myeloma (MM) are at a higher risk of viral infection due to immune deficiency, as recently highlighted by the severity of COVID-19 in MM patients. However, the mechanism(s) underlying the lack of anti-viral immune response in MM are unclear. We have previously shown that there are more plasmacytoid dendritic cells (pDCs) in the bone marrow of MM patients than in the bone marrow of healthy individuals and that these dysfunctional patient pDCs suppress T and NK cell–mediated immunity. Importantly, pDCs play a key role in both recognizing viral nucleic acid and initiating anti-viral immunity via type 1 interferon (IFN) response signaling. Therefore, we analyzed the immune-pathway proteins implicated in COVID-19-host interactions [COVID-19 protein interaction map, Gordon et al, Nature 2020; 583, 459-568] to assess whether pDC-MM interactions modulate these pathways to confer immune suppression and susceptibility to COVID-19.

Results. We first assessed whether the expression of human genes involved in the COVID-19-host interaction changes when ex vivo MM cells are co-cultured with patient pDCs (compared to MM cells alone). We found that 41 of these genes were differentially expressed in MM cells when exposed to pDCs (log2Fold change: ± 3.5-fold; p < 0.00001), and this expression varied widely (median (log10): -0.13 to 4.5; adj p < 0.001). Three of these genes are linked to innate immunity and are essential for IFN production in MM cells – TBK1, IRF3 (Interferon regulatory factor 3), and RAE1 (interferon-inducible mRNA nuclear export factor). Importantly, pDCs downregulated the expression of TBK1 (Log2FC: -0.5) and RAE1 (Log2FC: -0.3) and upregulated the expression of IRF3 (Log2FC: 1.5) in MM cells after coculture (compared to MM cells alone, p < 0.0001). Concordantly, analysis of publicly available gene profiling datasets on relapsed MM patients showed low levels of TBK1 and RAE1 and higher levels of IRF3 (n = 50) (Log2FC: TBK1: -0.208; RAE1: -0.286; IRF3: 0.273; vs normal; p < 0.05). Of note, TBK1 expression was positively correlated with prognosis (n = 350, p = 0.026).

We found that most of the 41 genes were also expressed in pDCs (NCBI GEO Datasets based on pDC cell line CAL1 and human peripheral blood pDCs). In unstimulated pDCs, TBK1 expression was significantly lower than IRF3 and RAE1 expression (2–3-fold, p < 0.05). Yet treating pDCs with CpG-ODN type-A, which induces IFN-α expression, increased both TBK1 (adj P = 0.004) and RAE1 (adj P = 0.043) expression, without significantly altering IRF3 expression.

Finally, COVID-19 relies on the host ubiquitin-proteasome system for propagation, and we found that targeting ubiquitin receptor Rpn13 with the specific inhibitor RA190 restored TBK1 expression in MM cell lines, indicating the potential clinical use of targeting Rpn13 in restoring anti-viral immune responses in MM.

Conclusion. Taken together, we show that TBK1 is downregulated in pDCs, and pDCs interacting with MM cells further decrease TBK1 expression in MM cells, thereby attenuating IFN-mediated anti-viral immune response signaling. Our findings suggest that: 1) low levels of TBK1 may confer increased susceptibility of MM patients to COVID-19 and 2) targeting TBK1 may restore the anti-viral immune response and reduce COVID-19 severity in MM.