The emerging roles of B cells in cancer development

B cells arise from the bone marrow and differentiate into various functional subsets in the peripheral lymphoid organs and inflammatory sites. Whereas T cells have been recognized as the dominant immune population in the tumor microenvironment, increasing evidence has shown the involvement of B cells in cancer development [1]. Within tertiary lymphoid structures (TLSs), which consist of aggregates of spatially organized immune cells in tumors, B cells actively interact with other populations and promote immunotherapy responses [1, 2]. However, the cellular heterogeneity, functional diversity, and underlying mechanisms of B-cell subsets across diverse cancer types remain major challenges for dissecting the roles of B cells in tumor development and immunotherapies.

In newly published papers in Science and Cell, two research groups independently established pancancer landscapes of tumor-infiltrating B cells across major human cancer types by integrating single-cell transcriptomes, B-cell-receptor (BCR) repertoires, chromatin accessibility data, and clinical prognostic information [3, 4]. Using public datasets and their generated single-cell RNA sequencing (scRNA-seq) data derived from cancer patients, Ma et al. identified 15 distinct B-cell subsets, whereas Yang et al. discovered a total of 20 fine-grained B-cell subpopulations, both of which showed significant tissue and cancer-type preferences. These two studies successfully characterized major B-cell subsets during different maturation stages, including naïve B cells, pre-germinal center and germinal center (GC)B cells, memory B cells, and plasma cells, on the basis of their hallmark gene expression signatures. Notably, they revealed the molecular phenotypes and functional features of novel B-cell subsets, including stressed B cells and tumor-associated atypical B (TAAB) cells, across cancer types. These novel findings revealed the existence of diverse functional B-cell subsets at the pancancer level and their associations with the clinical prognosis of cancer patients.

Antibody-secreting cells (ASCs), including plasmablasts and plasma cells, have long been recognized as terminally differentiated B cells with effector functions of producing antibodies. High-resolution single-cell analysis revealed a high diversity of tumor-infiltrating ASCs, which exhibited clonal expansion and homogeneity [3]. Interestingly, the frequencies and isotypes of tumor-infiltrating ASCs are strongly affected by cancer type, suggesting that cancer origin and the tumor microenvironment significantly affect ASC differentiation [3, 4]. ASCs are generated mainly from GCs, where B cells undergo somatic hypermutation, class-switch recombination, and clonal expansion. Recent findings have shown that extrafollicular B-cell responses are important for the generation of short-lived and polyreactive ASCs. Using BCR clonal sharing strategies, both Ma et al. and Yang et al. recapitulated GC reactions in diverse tumor types and highlighted their important roles in the formation of ASCs [3, 4]. Many ASCs share BCR clones with GC B cells, suggesting their dependence on GC reactions in tumors [3, 4]. Additionally, Ma et al. reported that tumor-infiltrating ASCs were also generated via extrafollicular pathways since extrafollicular atypical memory B cells clonally shared with ASCs and efficiently differentiated into ASCs in culture [3]. GC-dependent and extrafollicular pathway-derived ASCs were observed to have great variance and preferences in different cancer types. Extrafollicular-derived ASCs presented significantly higher frequencies than did GC-derived ASCs in tumors at the pancancer level, while these cells presented low somatic hypermutation and diverse BCR clonotypes [3]. Isotype analysis revealed a shift from IgA to IgG in tumor ASCs [4]. The extrafollicular-derived ASCs presented high interferon-induced IgG isotypes, whereas the GC-derived ASCs presented abundant IgA isotypes, which may reflect different stages of B-cell responses during cancer development. Likewise, the overexpression of IGHV genes in GC-derived ASCs correlated with strong antigen-specific affinities, whereas extrafollicular-derived ASCs implied a polyreactive BCR repertoire, suggesting their distinctive roles in the antitumor response [3].

The integrated studies by Ma et al. and Yang et al. identified a unique TAAB subset that resembles atypical memory B cells or age-associated B cells detected in autoimmune diseases, chronic inflammation, and aged mice [5, 6]. The identified atypical memory B cells [3] and TAAB cells [4] shared similar immunophenotypes, such as high expression of ITGAX, FCRL4, TBX21, and ZEB2. Moreover, TAAB cells have been detected across various cancer types and exhibit distinctive transcriptional and epigenetic features, suggesting that these cells represent a unique B-cell population with potential functionality in tumor development. As precursors of extrafollicular-derived ASCs, TAAB cells undergo antigen-driven activation and are less switched with low levels of somatic hypermutation [3, 4]. Moreover, TAAB cells highly expressed genes related to B-cell activation and antigen presentation. Moreover, these cells upregulated the expression of exhaustion-associated transcription factors (TOX, BATF, etc.) and immune checkpoints (PDCD1, CTLA4, etc.) [3, 4], suggesting that TAAB cells are activated and may contain exhausted B-cell subsets. One of the important questions is whether TAAB cells could be derived from GC responses or differentiate into GC B cells. To address this issue, Ma et al. analyzed the BCR repertoire, developmental trajectory, epigenomic regulation, and gene expression features of TAAB and GC B cells. They reported that TAAB cells and GC B cells were completely separated into two independent pathways, leading to extrafollicular- and GC-derived ASCs in tumors [3]. Unlike ASCs, which reside mainly in the perimeter of TLSs, TAAB cells are significantly enriched in the center of immature TLSs because of their high expression levels of homing receptors [3]. Moreover, the presence of TAABs was correlated with TLS formation in tumors. Within TLSs, TAAB cells are located in close proximity to CD4 + T cells and exhibit strong interactions. IL-21 is the key cytokine for the induction of CD11c⁺ atypical memory B cells in the pathogenesis of autoimmune diseases [6]. In tumors, PD1^hiCXCL13⁺ peripheral helper T cells promote the formation of TAAB cells through the IL-21–IL-21R axis [3, 4]. TAAB cells also provide activation signals for CD4 T cells through antigen presentation and the expression of costimulatory molecules [4]. Mechanistically, the high levels of glutamine in tumors shifted B-cell differentiation to TAAB via the regulation of TBX21, which established the epigenetic identity of extrafollicular TAAB cells [3]. In some cancer cohorts, such as colon adenocarcinoma and neck squamous cell carcinoma cohorts, the presence of TAAB cells was correlated with worse survival [3, 4]. Although memory B cells are correlated with improved responses and longer survival upon anti-PD1 treatment, TAAB cells are associated with treatment resistance in melanoma and lung cancer patients [3]. However, the TAAB signature at the pancancer level was strongly associated with favorable prognosis, indicating different functions of TAAB cells across human cancers [4]. Currently, the available data suggest the heterogeneity of TAAB cells and a cancer-type preference for these subsets, which may partially explain the functional discrepancies of TAAB in different cancer types.

Previous studies have reported that tumor-associated stressed T cells are present in the tumor microenvironment across various cancer types [7]. Similarly, both Ma et al. and Yang et al. identified a stressed B-cell subset that expressed high levels of heat shock proteins and hypoxia-related genes [3, 4], resembling the previously characterized stressed T cells in tumors. The stressed B cells were associated with poor survival at the pancancer level and may serve as a predictive parameter for unfavorable outcomes in various cancers [3, 4], suggesting that the stress responses in the tumor microenvironment could represent common molecular characteristics of different cell lineages across diverse cancer types.

Regulatory B cells (Bregs), characterized by their suppressive functions and production of inhibitory cytokines, reportedly exhibit protumor functions and are associated with poor prognosis in cancer patients. Recently, the B-cell-specific checkpoint molecule TIM1 and the immunosuppressive cytokine IL-35 were shown to regulate antitumor immunity [8, 9]. In the studies by Ma et al. and Yang et al., Bregs were not uniquely distinguished in the scRNA-seq datasets. The gene expression of IL10 was dispersed among major B-cell subsets, suggesting that IL-10-producing Bregs potentially arose from different stages of B-cell responses in tumors [4]. However, sorting-purified tumor-infiltrated TAAB cells produce massive amounts of IL-10 and TGF-β [3]. Moreover, glutamine-induced atypical memory B cells that resembled TAAB cells inhibited the proliferation of T cells and inflammatory cytokine production and promoted T-cell differentiation toward regulatory and exhausted phenotypes in a coculture experiment [3]. The functional interactions between TAAB cells and T-cell subsets need further investigation. PD1^hi B cells represent a novel protumorigenic Breg subset that causes T-cell dysfunction and promotes disease progression in advanced-stage hepatocellular carcinoma patients [10]. Although Bregs were not clustered as a separate B-cell lineage, the high expression of immune checkpoints, including PD1, suggested the possible existence of a regulatory B-cell subset within TAAB cells, which might explain the poor TAAB-associated prognosis in some cancer types. Nevertheless, the phenotypes, developmental pathways, and functional roles of Bregs in cancer progression warrant further investigations beyond the current transcriptome-based analysis.

In summary, the studies by Ma et al. and Yang et al. established a blueprint for tumor-infiltrating B cells across human cancers. Extensive transcriptome analysis revealed the abundance and heterogeneity of tumor-infiltrating B cells (Fig. 1), which provides a comprehensive platform for further exploration of the functional commonality and diversity of B cells in cancer. The identification of novel B-cell subsets, including TAAB cells, and their clinical associations with tumor progression at the pancancer level shed new light on the current understanding of the immune microenvironment in tumors, which may facilitate the development of novel B-cell-based immunotherapies.

Fig. 1.

B-cell subsets in the tumor microenvironment. Antibody-secreting cells (ASCs) are derived from the germinal center (GC) pathway and the extrafollicular (EF) pathway. Tumor-associated atypical B (TAAB) cells extensively interact with T cells and exhibit diverse functions. Stressed B cells express high levels of heat shock proteins and hypoxia-related genes. Regulatory B cells (Bregs) produce inhibitory cytokines with immunosuppressive functions. Both stressed B cells and Bregs are associated with poor prognosis in cancer patients

Competing interests

The authors declare no competing interests. Liwei Lu is editorial board member of Cellular & Molecular Immunology, but he has not been involved in the peer review or the decision-making of the article.