The B2M–LILRB1 axis as a novel myeloid immune checkpoint and therapeutic target in multiple myeloma Free

Cancer immunotherapies have shown significant promise in multiple myeloma (MM). However, their long-term efficacy is limited by immune evasion, and MM remains incurable. A key evasion strategy involves upregulating inhibitory phagocytosis checkpoints (IPCs), which prevent antigen-presenting cells (APCs) from phagocytosing cancer cells. This pathway is critical for the immune effects of standard anti-MM agents, including bortezomib (BTZ), an inducer of immunogenic cell death (ICD), and the monoclonal antibody daratumumab (DARA). Yet, the mechanisms regulating IPCs in MM remain poorly understood.

To address this gap, we screened IPC axes in MM cells and APCs using bone marrow (BM) aspirates from MM patients (n=39) and healthy donors (n=5) via multiparametric flow cytometry. Immunophenotyping included 9 parameters for MM cells and 16 for APCs, covering IPC and lineage markers. We identified the B2 microglobulin (B2M) / leukocyte immunoglobulin-like receptor B1 (LILRB1) axis as the dominantly expressed IPC in MM. B2M, expressed on MM cells, binds LILRB1 on APCs. Notably, B2M is a known adverse prognostic factor in MM.

Computational analysis of flow-cytometry data using a semi-automated pipeline allowed us to define 8 distinct myeloid cell clusters, with consistently high LILRB1 abundance, especially in conventional type 1 dendritic cells (DCs). Single-cell RNA-seq data from 90 MM BM samples confirmed LILRB1 expression in DCs, monocytes (Mo) and macrophages (MΦ). qRT-PCR of BM Mo-derived DCs found that LILRB1 mRNA levels increased from healthy donor to smoldering MM to overt MM, suggesting a link to disease progression.

To functionally assess the B2M/LILRB1 axis, we knocked out B2M in MM cell lines (n=3) and re-expressed a chimeric B2M mutant (B2M^MUT) lacking functional LILRB1-binding site by introducing 16 missense mutations. B2M functions as an MHC-I subunit, so this mutant allowed us to isolate its checkpoint role without the confounding effects from complete B2M loss. Yeast 2-hybrid assays confirmed a direct interaction between B2M and LILRB1, but not for B2M^MUT. Phagocytosis assays found that B2M^KO and B2M^MUT cells were more susceptible (more than 2-fold) than WT to phagocytosis by DCs and MΦ after BTZ or DARA treatment (p<0.01). LILRB1 knockdown (KD) in APCs, using an antisense oligonucleotide, similarly enhanced the phagocytic potential after BTZ treatment (p<0.05).

Since MM cells actively secrete B2M (s-B2M), we analyzed the media from MM cell lines with different B2M statuses: B2M WT, B2M^KO, and B2M^MUT. We found s-B2M in the media from B2M WT cells, but no B2M in the KO cells and s-B2M^MUT in the B2M^MUT. Exposure of DCs to the conditioned media of WT s-B2M, but not s-B2M^MUT, reduced phagocytosis after BTZ or DARA (p<0.05) of B2M^KO cells, suggesting that s-B2M impairs APC function in a LILRB1-dependent manner. Treatment of APCs with His-tagged recombinant B2M (rB2M) confirmed activation of downstream LILRB1 signaling, assessed by modulation of p-SHP1.

Interestingly, APCs internalized rB2M in a LILRB1-dependent manner, as both LILRB1 KD and treatment with rB2M^MUT resulted in reduced internalization. Correlative light and electron microscopy (CLEM) and TEM revealed that internalized rB2M localized to perinuclear vesicles near the microtubule-organizing center (MTOC). Confocal imaging found that rB2M uptake increased microtubule stability, as assessed by a higher acetylated-to-tyrosinated tubulin ratio. Functionally, this correlated with impaired APC maturation following lipopolysaccharide stimulation (assessed by phalloidin staining) and a reduced ability to activate T cells after ICD, as measured shown by reduced T cell-mediated MM cell lysis.

To evaluate the in vivo relevance, we generated a murine MM cell line (5TGM1) expressing a chimeric B2M (B2M^mut), which cannot bind the murine LILRB1 ortholog. Mice injected with B2M^mut cells exhibited reduced tumor engraftment and growth compared to those injected with WT cells, accompanied by increased infiltration of CD3⁺ and CD8⁺ T cells.Ongoing studies are assessing the effect of the B2M/LILRB1 axis on BTZ or DARA efficacy in vivo and will be presented at the meeting.

This work provides fundamental insight into the immune role of B2M in MM, revealing the B2M/LILRB1 axis as a novel immune checkpoint in MM driven by impaired APC function. Targeting this axis represents a new therapeutic vulnerability to restore immune competence and improve MM outcomes.