Abstract The treatment landscape of diffuse large B-cell lymphoma (DLBCL) has progressed with the emergence of immune-cell therapies (ICTs), such as chimeric antigen receptor (CAR) T-cell therapy and bispecific antibodies. However, many patients still remain incurable despite these innovative options. Treatment resistance in DLBCL is associated with the immunosuppressive tumor microenvironment (TME), and myeloid-derived suppressor cells (MDSCs) play a crucial role in contributing to it. Furthermore, an increase in MDSCs has been linked to disease progression, poor prognosis, and reduced effectiveness of ICTs. Therefore, understanding how DLBCL cells promote MDSC development is essential for overcoming treatment resistance to ICTs. To address this issue, in this study, we investigated how DLBCL cells induce monocytic (M)-MDSCs from normal peripheral blood mononuclear cells (PBMCs) using an indirect co-culture system and traced the molecular changes involved. Four human DLBCL-derived cell lines (HDBCLs), including two double-expressor HDBCLs (KPUM-UH1 and KPUM-MS3) (Sasaki N, Exp Hematol 2011) and two activated B-cell-like HDBCLs (HBL1 and DLBCL2), were used. PBMCs from healthy donors were indirectly co-cultured with HDBCLs for 96 hours, and flow cytometry was used to analyze PBMCs for the induction of M-MDSCs displaying the CD14+/HLA-DR-/low phenotype. Total RNA was extracted from the CD33+ myeloid fraction of PBMCs after co-culture, and gene expression profiles (GEP) were analyzed using microarray analysis. Consequently, we first demonstrated that all four HDBCLs tested were effective in inducing CD14+/HLA-DR-/low myeloid cells from normal PBMCs within 96 hours of co-culture. The induced CD14+/HLA-DR-/low myeloid cells were also shown to be potent in inhibiting T cell proliferation, confirming these cells as M-MDSCs both phenotypically and functionally. The ability to induce M-MDSCs was highest in DLBCL2, followed by HBL1, while KPUM-UH1 and KPUM-MS3 showed modest M-MDSC-inducing capabilities. Based on the finding that M-MDSCs were induced through indirect co-culture with HDBCLs, we screened soluble factors in the conditioned medium (CM) of HDBCLs secreted by these cells using cytokine arrays, and found that all four HDBCLs secreted MIF. Additionally, the results showed that DLBCL2 cells produced high levels of interleukin-10 (IL-10), while HBL1 cells secreted IL-10 at low levels, and KPUM-UH1 and KPUM-MS3 cells did not produce detectable IL-10. We then examined the functional roles of MIF and IL-10 in the induction of M-MDSCs by HDBCLs. Interestingly, the MIF inhibitor 4-IPP greatly reduced M-MDSC induction by all four HDBCLs, and exposure to recombinant (r) MIF alone had minimal ability to induce M-MDSCs from normal PBMCs, indicating that MIF plays an essential but preparatory role in M-MDSC induction. Conversely, blocking IL-10 with a neutralizing antibody significantly decreased M-MDSC induction from PBMCs by IL-10-secreting HDBCLs, and adding rIL-10 increased the M-MDSC-inducing effect of IL-10-non-secreting HDBCLs, indicating that IL-10 plays a more facilitative role in inducing M-MDSCs. To understand the molecular regulation in PBMCs during the induction of M-MDSC by the co-existence with HDBCLs, we examined gene expression changes in CD33+ myeloid cells at 24, 48, and 96 hours after co-culturing with HDBCLs. We also examined the molecular dynamics in co-culture with non-IL-10-secreting cell lines that occurred after 96 hours of rIL-10 treatment. Results from gene set enrichment analysis revealed that gene sets associated with inflammatory response and tumor necrosis factor-alpha (TNF-α) signaling, as well as inflammatory molecules such as IL-1α, were upregulated 24 hours after co-culture. However, these upregulated gene sets became less expressed after 48 hours and eventually were downregulated after 96 hours. Additionally, the downregulation of inflammatory response and TNF-α signaling was further amplified by rIL-10. In conclusion, our study demonstrated that the multistep process of M-MDSC induction from PBMCs, triggered by HDBCLs, begins with an initial transient hyper-inflammatory phase induced by MIF and transitions into a post-inflammatory immunosuppressive phase accelerated by IL-10. Our findings suggest that targeting phase-specific cytokines and related molecules could reduce the induction, maturation, and proliferation of M-MDSCs, ultimately enhancing therapeutic outcomes, especially in ICTs.
Shimura et al. (Mon,) studied this question.