TGFβ: Protecting PD-1 from mRNA Decay

TGFβ has pleiotropic effects in tumors, including induction of epithelial–mesenchymal transition, activation of fibroblasts, and inhibition of CD8⁺ T cells (1, 2). How exactly TGFβ inhibits CD8⁺ T-cell function has been unclear, with both direct and indirect effects proposed. Conditional deletion of TGFBR2 in mature T cells leads to increased proliferation and cytokine production, indicating a direct effect of TGFβ signaling in restraining CD8⁺ T-cell effector functions (1). In tumors, activated fibroblasts are a major source of both TGFβ and chemokines that attract immunosuppressive cells, ultimately leading to exclusion of T cells from the tumor nests (1). TGFβR small-molecule inhibitors are in clinical trials, as are TGFβ-blocking antibodies and agents that prevent release of active TGFβ from its latency-associated protein cage (1). Understanding how TGFβ interfaces with CD8⁺ T cells will help inform which therapies might be included in combination trials including a TGFβ-targeting agent.

mRNA stability is controlled by numerous factors, including association of miRNAs and binding of regulatory proteins to the 3′ untranslated region (UTR). For the most part, longer 3′UTRs have been associated with decreased mRNA stability due to nonsense-mediated decay; however, certain 3′UTR-binding proteins can stabilize transcripts (3). Here, the authors used actinomycin D to inhibit new transcription in activated CD8⁺ T cells, and then quantified particular mRNAs present at given time points (4). TGFβ increases the half-life of PDCD1 transcript, but importantly does not affect mRNA stability of housekeeping genes (ACTB and TUBA1A) or genes associated with activated CD8⁺ T cells (IL10 and TNF). These additional mRNAs are key, as they indicate that TGFβ in CD8⁺ T cells is not globally affecting mRNA stability, but rather has a specific effect on PDCD1 and possibly other coregulated transcripts. To elucidate the connection between TGFβ and PDCD1 stabilization, the authors used an elegant combination of overexpression, siRNA-mediated knockdown, and small-molecule inhibitors. TGFβ usually signals via SMAD family transcription factors; however, TGFβ can also cross-activate the MAPK and other signaling pathways (1). Wu and colleagues found TGFβ exposure during CD8⁺ T-cell activation induces phosphorylation of p38MAPK, leading to translocation of NF-κB p65 to the nucleus, where it activates transcription of the RNA-binding proteins SRSF3 and SRSF5, resulting in increased stability of PDCD1 transcripts (4).

Although originally described as a splicing factor, SRSF3 is now appreciated to have prominent roles in extension of 3′UTRs and in mRNA nuclear export via its association with the export protein, NXF1 (3). Wu and colleagues used RNA coprecipitation to show that both SRSF3 and NXF1 associate with PDCD1. Deletion of SRSF3-binding sites from the PDCD1 3′UTR eliminated these interactions, trapped PDCD1 in the nucleus, and abolished the TGFβ-mediated increase in mRNA half-life (4). The biochemistry here is elegant, and reveals a causal link between TGFβ, SRSF3, and PDCD1 nuclear export and stabilization in CD8⁺ T cells. Whether this is related to the in vivo synergy of TGFβ and PD-1 coblockade is unknown, although the authors nicely showed that loss of TGFβ in a model of renal cell carcinoma leads to decreased SRSF3 in tumor-infiltrating CD8⁺ T cells (4). How many transcripts are regulated by SRSF3/5 is unclear, but PDCD1 is likely one gene of a small module of genes upregulated by TGFβ. Understanding the gene modules targeted by TGFβ through specific mRNA-binding proteins may offer a new approach to globally influencing cell fate.

S.K. Dougan reports grants from Novartis, Eli Lilly, and Bristol-Myers Squibb and personal fees from GlaxoSmithKline and Kojin outside the submitted work.