|Home | About | Journals | Submit | Contact Us | Français|
Accumulating evidence indicates that Lymphotoxin (LT)-β related cytokines directly contribute to the phenotype of cancer cells and alter the tumor microenvironment. Lymphotoxins are part of a cytokine network well known in controlling the development and homeostasis of secondary lymphoid organs. In the adult, the LT network takes on the responsibility of generating inflammatory microenvironments that control innate and adaptive immune responses involved in host defense. This review provides a perspective of the emerging evidence implicating the LT Network in the development and progression of various cancers including lymphoma. Redirecting the LT Network to alter tumor microenvironments may provide a specific approach to therapeutically target tumor-permissive microenvironments and cancer progression.
Substantial evidence indicates that the development and progression of cancer is aided by inflammation in the tumor microenvironment [1,2]. This transformative hypothesis leads to the prediction that therapeutic targeting of specific inflammation-associated phenotypes may improve therapy and ultimately patient survival. Indeed, associations between autoimmune diseases and lymphomagenesis are well recognized, but with limited mechanistic understanding. However, the complexity of “inflammation” in the cancer setting is not well understood, limiting the number of therapeutic targets and the efficacy of existing approaches. The cellular chaos of lymphoid cancers is reflected by variable numbers of malignant cells intermixed in a matrix of connective tissue cells, endothelium with nascent blood and lymphatic vessel formation, and often prominent lymphocyte and macrophage infiltration . These cellular features are strikingly similar to the formation of nascent lymphoid tissue observed at sites of persistent inflammation. Expressed genes encoding cytokines, chemokines and growth factors can be detected in tumor microenvironments, but the cellular infiltrates are not typically considered “organized”, and vary substantially when compared to normal secondary lymphoid tissues. It is a widely accepted concept that intercellular communication pathways between tumor cells and the surrounding stroma have evolved with the tumor to create permissive or selective microenvironments allowing malignant cells to mutate, grow, and resist immune recognition and destruction, or survive therapeutic assaults [4,5]. In particular, Hodgkin lymphoma can serve as a paradigm how the malignant cells shape their inflammatory microenvironments to their selective advantage (reviewed in ).
Cancer inflammation represents discontinuous phases of new lymphoid tissue formation, lymphoid neogenesis . This perspective of cancer inflammation and neogenesis points directly to the cytokine network formed by Lymphotoxin (LT)-αβ, LIGHT (TNFSF14) and their receptors . The LTαβ-LIGHT cytokine network directs innate and adaptive host defense mechanisms providing strong selective pressures on viral pathogens to evolve countermeasures. Recent evidence identifies the LTαβ-LIGHT network as a contributing factor in several chronic inflammatory diseases (Table 1). Importantly, emerging evidence indicates that solid tumors and hematopoietic malignancies manipulate components in the LTαβ-LIGHT network to modify their tumor microenvironment, suggesting these cytokines provide strong selective pressure for tumor evolution and tumor immune evasion.
The LT-related cytokines include TNFα, lymphotoxin (LT)-α, LT-β, LIGHT (TNFSF14), and their cognate superfamily of receptors, TNF receptor (TNFR)-1, TNFR2, LTβR, herpesvirus entry mediator (HVEM; TNFRSF14) and decoy receptor-3 (DcR3) (Figure 1). The LTαβ-LIGHT Network provides communication between lymphoid cells and surrounding tissue cells. The different ligand-receptor systems can activate proinflammatory pathways (eg., TNFα-TNFR1), homeostatic signaling (eg., LTαβ-LTβR) or inhibitory signaling responses (eg., HVEM-BTLA). The LTαβ-LTβR system serves as a developmental pathway required for the formation of secondary lymphoid organs and maintenance of lymphoid tissue architecture. In the neonate, B lymphocytes assume responsibility for the maturation and homeostasis of secondary lymphoid organs including the spleen through the LTαβ-LTβR pathway . In the adult, LTαβ in B cells provide key differentiation signals to LTβR expressing stromal cells and macrophages to initiate type 1 interferon production to viral pathogens [10,11]. Additionally, LTαβ orchestrates formation of tertiary lymphoid tissues at extra lymphatic sites of chronic inflammation. Importantly, components of the TNFRSF signaling pathway are frequently mutated in cancer including components of the ubiquitin E3 ligase (TRAF3, TRAF2, and cIAP) that regulates expression of the NFκB inducing kinase, NIK .
LTβR signaling in stromal cells forms the microarchitecture of the lymphoid organ. Scaffolds of highly differentiated stromal cells, reticular fibrocytes, lymphatic endothelium, and follicular dendritic cells, all responsive to LTβR signaling, partition the lymphoid organ into specialized niches where T and B lymphocytes congregate . Distinct lineages of myeloid cells require LTβR signaling for positioning and differentiation to populate these niches . LTαβ expressing marginal zone B cells form a critical barrier to pathogens by inducing differentiation of macrophages and stromal cells that produce type 1 interferons (IFNαβ) in response to viral pathogens [10,11]. These LT-differentiated cells produce high levels of IFNαβ, but allow the pathogen to replicate, amplifying antigens available to stimulate effective T and B cell responses . T and B cells segregate into separate niches created by LTαβ differentiated stromal cells secreting distinct patterns of chemokines. CXCL13 is the chemoattractant forming the B cell-rich follicles, and CCL21 recruits T cells around the arteriole. CXCL13 through its cognate receptor CXCR5 stimulates LTαβ expression in B cells forming a reciprocal signaling pathway. Inflammatory cells organize at extra lymphatic sites during persistent inflammation into tertiary lymphoid tissues through the expression of LTαβ expressed in a variety of activated cells, such as T and B cells, NK cells, DC and LTi cells . T and B cells, and LTβR signature chemokines (CCL21 and CXCL13) and integrins are detected at these ectopic sites [17–19]. The LTαβ-LIGHT Network provides an essential component for generating inflammatory environments, from the initial innate response to full blown adaptive immune responses.
The LIGHT-HVEM-BTLA component of the Network provides inhibitory and proinflammatory signaling initiated by dual action of HVEM [20,21]. HVEM-mediated signaling is an important mechanism suppressing inflammation in mucosal tissue . For example, experimental colitis is dramatically accelerated in Hvem-deficient mice. This immune regulation mechanism involves trans signaling between HVEM expressed in intestinal epithelia cells and BTLA in T cells to activate inhibitory signaling. In lymphoid cells, intrinsic HVEM expression in T cells is important for the survival of effector memory T cells during chronic lung inflammation  and graft vs host disease [24–26]. BTLA expression in B cells is thought to play a similar role in limiting cellular activation . The loss or amplification of HVEM may act intrinsically in cancer cells or modulate the microenvironment.
Emerging evidence in animal models and human genetic data implicates the LTαβ-LIGHT Network as a key mechanism creating microenvironments suitable for cancer development and progression. Two recent studies showed that the LTαβ pathway can drive carcinogenesis, with LTαβ-LTβR linking malignant cells to the inflammatory, tumor permissive environment. In the mouse, transgenic expression of LTα and LTβ under control of the liver-specific albumin promoter caused chronic hepatitis in all transgenic animals . By 18 months of age, around 35% of these mice developed liver tumors closely resembling human hepatocellular carcinoma (HCC). The development of HCC was dependent on lymphocytes and NFκB signaling (RelA dependent) in hepatocytes. As expected, administration of LTβR-Fc decoy to block LTαβ and LIGHT also prevented tumor formation. LTαβ-dependent carcinoma failed to develop in mice lacking T and B cells (Rag1−/−). This result suggested that the LTαβ-LTβR system did not have direct oncogenic properties in hepatocytes, but rather LTβR signaling reshapes the hepatic microenvironment through secretion of chemokines resulting in recruitment of inflammatory cells that drive oncogenesis. Robust upregulation of LTα, LTβ, and LTβR transcripts were found in the livers of HCC patients consistent with an influx of inflammatory cells during cancer progression. The link between LTβR and HCC is anticipated from data showing that stimulation of LTβR in hepatocytes is critical for liver regeneration after partial hepatectomy, and this process depends on production of IL-6 induced by LTβR signaling . Intriguingly, hepatic LT production can be induced by IL-6 , suggesting that hepatocyte LTβR forms the center of an IL-6–LTαβ inflammatory loop that contributes to both liver regeneration and liver cancer. The connection between IL-6/Stat3 and LTβR/NFκB are particularly intriguing in view of the cooperation between Stat3 and NFκB in cancer .
Expression of LTαβ in B cells drives the rapid re-emergence of prostate cancer (CaP) following androgen deprivation . In this model, CaP cells transplanted into mice, which are then castrated, causing progression to androgen independence. However, androgen-independent tumor progression was significantly delayed by treatment with LTβR-Fc decoy or in B cell specific deletion of LTαβ. The authors demonstrate that expression of LTβ in B cells depends on the expression of IKKβ in those cells, while expression of LTβR and IKKα in CaP cells was required for rapid androgen-independent tumor emergence. The dependence of prostate cancer progression on LTβ in B cells is remarkably similar to the innate B cell host defense mechanism to viral pathogens.
The HVEM-BTLA pathway is intimately linked with the LTβR pathway  and emerging evidence indicates mutations in these components may alter the tumor and its microenvironment. The GISTIC somatic gene copy number dataset  (www.broadinstitute.org/tumorscape) revealed deletions in BTLA are frequent in human hematologic malignancies. A frequent (5.1%) and narrow deletion peak at chr3q13.2 encompassing human BTLA is associated with acute lymphoblastic leukemia and in all hematologic malignancies. HVEM also shows significant and frequent deletion in epithelia tumors including lung, breast and renal tumors, and is also found amplified in epithelia tumors, perhaps revealing dual action of HVEM in promoting and inhibiting inflammation. Moderate to high HVEM expression was observed in melanoma, perhaps serving as a potential suppressive mechanism for anti-tumor CD8+ T cells. In contrast to virus-specific CD8+ effector cells, which downregulate BTLA upon activation, tumorspecific CD8+ effectors maintain high BTLA levels and remain susceptible to inhibition through binding of HVEM [35,36]. CD160, in contrast to BTLA, was unexpectedly identified as a tumor specific marker with significant amplification in acute lymphoblastic leukemia [37,38]. The frequency of mutations in HVEM, CD160 and BTLA in epithelial and hematologic malignancies adds substantial significance to the idea that selective pressures promote clonal dominance of these mutations in cancer.
The most common B-cell lymphoma subtypes are diffuse large B-cell lymphoma (DLCBL) and follicular lymphoma (FL). Classical Hodgkin lymphoma (CHL) is less common, but the most frequent lymphoma in younger patients and thus represents a significant healthcare issue. The role of non-neoplastic immune cells in the microenvironment is increasingly recognized as an important contributor to lymphoma pathogenesis . Importantly, the abundance of non-cancerous cells infiltrating the tumor correlates with the survival of patients with FL, DLBCL and CHL [40–42]. Experimental models of lymphoma reinforce the tumor-stroma dialog via LTαβ-LTβRregulated chemokine pathways . However, the cellular interactions between immune cells and B cell lymphoma are complex, underscored in a recent study that revealed a tumor-promoting role for innate NKT cells through inhibition of the CD8 cytotoxic T cell response to the B cell lymphoma by opposing the effects of type II NKT cells .
HVEM has emerged as a recognized mutational target in lymphoma, which is not surprising in that HVEM is subject to strong selective pressures as target of viral immune evasion strategies . These data raise the possibility that HVEM-dependent pathways are critical for immune reactions against tumor cells. The dominance of somatic HVEM mutations was revealed in FL and DLBCL [46,47] and recurrent deletions of HVEM in Reed-Sternberg cells microdissected from Hodgkin lymphoma . Together these results implicate HVEM as a universal tumor suppressor gene across a broad range of lymphoma subtypes.
The incidence of non-synonymous mutations in the HVEM gene is high. Cheung et. al.  identified 47 cases (18%) with mutations affecting HVEM in a cohort of 251 FL patients. These mutations were associated with inferior outcome and high-risk clinical features in addition to an increased risk of transformation to DLBCL. Another study showed similar results with HVEM mutations in FL, but with an even higher frequency (44%). However, the latter cohort was not associated with diminished survival . Frequent deletion of chromosome 1p36.22–36.33 involving HVEM occurred in 25% of FL patients and were associated with poor outcome . More recently, deep-sequencing approaches in DLBCL uncovered recurrent mutations in HVEM, and of interest, all of the reported mutations were exclusive to the germinal center B cell type [49,50]. Since the cell-of-origin in follicular lymphoma and Hodgkin Reed Sternberg cells are believed to be a (post-)germinal center B cell, HVEM inactivation might be exclusively associated with germinal center-driven lymphomagenesis.
Targeted DNA re-sequencing of the TNFRSF14 coding exons found 47 mutations in total from 251 FL patients . Furthermore, genome-wide RNA sequencing revealed additional mutations in FL and DLBCL clinical samples . The mutational spectrum indicated some “hot spot” areas (Figure 2). Several mutations overlap with mutations in HVEM reported by Launay et al , suggesting that these recurrently altered sites might be functionally relevant. Although gene inactivating mutations are predominant in HVEM, the majority of point mutants are almost exclusively located in the extracelluar domain suggesting the potential to alter ligand binding. In FL and DLBCL, where HVEM expression is frequently deleted, this mutational pattern could reflect different selective pressures at work to override the intrinsic (cis) inhibitory HVEM-BTLA signaling pathway. The HVEM-BTLA cancer genotype is consistent with the dual roles of BTLA and HVEM in activation of NFκB survival pathways and as an inhibitory pathway. We also suggest that mutations in HVEM or BTLA, owing to its counter regulatory actions , will likely lead to significant perturbations in the LTαβ-LTβR network, thus altering local microenvironments. Together with human genetic data these results implicate HVEMBTLA as a possible control point in B cell lymphoma development and progression.
Therapeutic rewiring of the LTαβ-LIGHT Network can provide microenvironments suitable to induce immune rejection overriding intrinsic mutations in cancer cells. Most instructive in this regard is the demonstration that LIGHT can induce a strong anti-tumor response in vivo. Ectopic expression of LIGHT in tumor cells leads to a robust and sustained anti-tumor immune response dependent on NK cells, IFNγ, and CD8+ T cells . LIGHT expression in tumor cells led to an increase in LTβR-regulated chemokines and cytokines within the tumor, which may form a more lymphoid-like environment that allows for increased priming of naïve T cells within the tumor site. The anti-tumor effects of LIGHT likely depend on both HVEM and LTβR, as HVEM is necessary for NK cell activation . Further work has demonstrated LIGHT delivered to tumors through adenoviral vectors or engineered mesenchymal stem cells  are effective at inducing meaningful anti-tumor responses and tumor regression, demonstrating that the antitumor effects of LIGHT do not depend on tumor-specific LIGHT expression.
The foregoing discussion provides strong evidence supporting the idea that the LTαβ-LIGHT Network is involved in cancer development and progression. Principles emerging from investigating these cytokines in infectious and inflammatory diseases have provided a framework to understand the selective pressures that are operative in cancer (Figure 3). In the cancer setting, the expression of HVEM in melanoma cells would likely provide a selective advantage by suppressing immune rejection via BTLA, and cell survival signaling via HVEM (bidirectional mode) in the tumor cell. However, the loss of HVEM observed in other solid tumors may lead to enhanced inflammation, perhaps via increased LTαβ-LTβR signaling or loss of lymphocyte inhibition through BTLA, creating tumor-promoting microenvironments, as observed in the hepatocellular and prostate models [28,32]. In FL and DLBCL, HVEM loss suggests a strong selective pressure from environmental interactions.
The linkage of tumor cells with the inflammatory microenvironment through LTαβ-LTβR and LIGHT/HVEM/BTLA pathways offers potential for significant therapeutic intervention. The specificity inherent in receptor and antibody therapeutics provides a direct route to rewiring these pathways to achieve efficacy in controlling cancer progression. Antagonism of the LTβR pathway utilizing LTβR-Fc (baminercept) may deprive the microenvironment of key inflammation-driven survival factors required by the HCC cells. TNFα is a contributing inflammation factor in the tumor milieu suggesting that antagonism of TNFR signaling with anti-TNF may be warranted. In HVEM-deficient FL or DLCBL agonist activation of BTLA could attenuate signals required for continued lymphoma progression. However, the efficacy of these approaches will clearly depend on knowledge of the mutational spectrum contained in an individual patient’s tumor. For example, mutations affecting signaling components downstream of the receptor (such as deletions in TRAF3 leading to constitutively active NFκB mediated survival signaling) could nullify the effects of antagonizing these cytokines produced by inflammatory cells in the tumor microenvironment. Accumulating evidence indicates the LTαβ-LIGHT Network contributes strong selective pressures for tumor evolution and immune evasion in both solid tumor and hematologic malignancies. We suggest that effective therapy utilizing these pathways will need to consider both targeting of tumor cells and the microenvironment to achieve optimal effectiveness.
The authors would like to recognize grant support from the US National Institutes of Health (AI033068, CA164679 and AI48073 to CFW), and support from the Terry Fox Foundation New Frontiers Program Project (Grant 019001) and Genome Canada/BC Grant Competition III (To RDG). CS is supported by a Career Investigator Award from the Michael Smith Foundation for Health Research and RB is supported by an NIH Postdoctoral Fellowship (T32CA121949). The bioinformatics and artwork assistance of Fong Chun Chan and Vasileios Bekiaris is gratefully appreciated.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest Statement: RB, RCG, CS declare no conflicts of interest; CFW has licensed patents to several TNFSF/TNFRSF molecules.