LIGHTing CAR T in the tumor microenvironment

Main text

Adoptive cell therapy using T cells expressing chimeric antigen receptors (CAR T) has demonstrated remarkable efficacy in hematological malignancies. However, CAR T therapy has shown limited responses in solid tumors, which may be due to inefficient CAR T cell recruitment, activation, and survival inside the suppressive tumor microenvironment (TME). The expression of the tumor necrosis factor superfamily member 14 (TNFSF14 or LIGHT) inside the TME promotes the expression of chemokines to form tertiary lymphoid structures (TLSs), which improves the recruitment and expansion of T cells as well as resistance to the suppressive TME.¹ In this issue, Zhang et al. engineered LIGHT-expressing CAR T cells, resulting in the accumulation of the LIGHT-CAR T cells and potent tumor killing activities in both syngeneic and xenograft mouse models.²

Most immunotherapy with antibodies or cytokines enhances anti-tumor T cell immunity through either blocking immune inhibitory signaling or boosting immune stimulatory signaling in pre-existing T cells. However, these approaches are generally ineffective in treating cold solid tumors, which are characterized by a lack of T cell infiltration. Adoptive CAR T cell transfer offers a promising solution by providing exogenous tumor-specific T cells. Nevertheless, several factors limit the efficacy of CAR T cells in solid tumors, including (1) inadequate adaptation to in vivo conditions, (2) inefficient trafficking, penetration, and expansion inside the TME, and (3) a more suppressive TME facilitated by immune-suppressive molecules and cells. To overcome these obstacles, researchers have conducted extensive studies focused on improving the fitness, survival, and cytotoxicity of CAR T cells. For example, strategies such as the expansion of CAR T cells using vaccines or engineered cytokines have been employed.³ Additionally, gene modulation of molecules or transcription factors related to T cell differentiation have been explored to expand or promote CAR T cell survival and enhance their effector functions. Meanwhile, a few limited approaches have demonstrated the ability to actively recruit and expand CAR T cells into the TME. For instance, CAR T cells with transgenic expression of chemokine receptors such as CXCR2 or CCR2 have shown improved migration to tumor sites.⁴ Engineering CAR T cells to express chemokines favors infiltration of T and dendritic cells (DCs) but not expansion. Additional robust expression of IL-7 is required to improve the survival of CAR T cells.^5,6 Nevertheless, efforts to identify a single molecule for both the recruitment and expansion of CAR T cells inside the suppressive TME have not been easy.

LIGHT binds to two different receptors, herpesvirus entry mediator (HVEM, a member of the TNFR superfamily) and lymphotoxin β receptor (LTβR). Interaction between LIGHT and HVEM delivers costimulatory signals that activate and expand T cells, even in the presence of regulatory T cells (Tregs). LTβR is commonly expressed on non-lymphoid cells and is critical for the formation of secondary and tertiary lymphoid structures.^7,8 Signaling downstream of LTβR involves various chemokines and adhesion molecules that control the migration and positioning of lymphocytes and DCs. Consequently, LIGHT signaling might create a proper niche for effective CAR T infiltration and expansion inside the TME.

In their bioinformatic analysis of prostate cancer (PRAD), Zhang et al. found a correlation between the expression of LIGHT and TLS signature genes, all of which were highly expressed in the relatively higher tumor immune infiltration sample group. Motivated by this finding, the authors expressed LIGHT in CAR T cells to overcome the limitations of CAR T cell therapy for solid tumors. LIGHT can function as both a soluble and membrane protein. As the secretion of the soluble truncated form (LIGHTt66) but not the full-length transmembrane form (LIGHT-TM) was much higher in vitro, the authors chose LIGHTt66. During the construction process, a vascular targeting peptide RGR was fused to LIGHT to facilitate the delivery of LIGHT protein into the TME. This design aims to maximize the effect of LIGHT protein secreted from CAR T cells inside the TME, recruiting more peripheral T cells for infiltration into the tumor. In an indirect paracrine manner, LIGHT promoted T cell migration by inducing the expression of chemokines CCL19, CCL21, CXCL13, as well as the adhesion molecule VCAM-1 in stromal, tumor, and bone marrow cells. However, LIGHT protein alone did not enhance the migration of T cells. RNA-seq analysis of LIGHT-CAR T cells in vitro revealed upregulated expression of activation-related cytokines (IL13, IFNγ, IL2) and genes that promote T cell survival and proliferation (IL9, IL12A, IL18, BCL6), suggesting direct modulation of CAR T cell status by LIGHT through its costimulatory HVEM receptor. In the xenograft PRAD mouse model, mice that received adoptive transfer of LIGHT-CAR T cells showed more T cell infiltration, and approximately 80% of mice became tumor-free, while only 20% of mice treated with control CAR T became tumor-free.

In the syngeneic melanoma tumor model, LIGHT-CAR T treatment induced more TLS-like structures indicated by IHC staining of markers of T cells, DCs, B cells, myeloid cells, and vascular high endothelial venules in the tumors. Moreover, the dose of LIGHT-CAR T administered in vivo induced mild inflammatory cytokines in serum and no detectable lesions in normal tissue sections, suggesting its potential safety. Nonetheless, more sensitive mouse strains and older mice need to be extensively evaluated for efficacy and potential toxicity. It is also unclear how long LIGHT-CAR T cells will survive and function upon rechallenge. CAR T cells can survive for years in some patients and may be critical to prevent cancer recurrence. Their long-term persistence may reflect the memory or stemness status of T cells. However, further studies are needed examine the differentiation status and duration of LIGHT-CAR T before and after tumor clearance. These efforts will help better understand LIGHT-CAR T for clinical use.

Solid tumor cells are typically heterogeneous and lack a unique tumor-specific epitope, which may result into tumor relapse after a single type of CAR T treatment. Therefore, recruiting and activating endogenous tumor-specific T cells may be crucial for overcoming the loss of epitopes. Although this study did not investigate this aspect, it is worth exploring whether LIGHT can recruit endogenous T cells, contribute to final tumor clearance, and respond after rechallenge.

The members of TNFSF family typically function as a homotrimer, posing challenges for their use as proteins in cancer therapy. LIGHT or fusion proteins with tumor-targeting antibodies have been reported to facilitate T cell infiltration into solid tumors and overcome the limitations of PD-L1 blockade therapy.⁹ Despite these promising findings, the clinical application of these therapies has been hindered, in part due to difficulties in producing high-quality proteins with proper tumor targeting. Gene-engineered cell therapy offers an avenue for the clinical application of TNFSF proteins, and this study serves as a good example of its potential. Interestingly, T cells do not naturally express LTβR, but the introduction of LTβR into CAR T cells significantly increased cell effector functions and resistance to exhaustion by constitutively activating the canonical NF-κB pathway.¹⁰ LTβR and HVEM also have alternative ligands, LTα1β2 and LTα3, respectively. As the development of gene engineering and cell-based therapy techniques advances, there is a great potential to explore these new molecules alone or in combinations to empower and enhance the survival, differentiation, tumor migration, and effector functions of CAR T cells to achieve better tumor control.