Potent T cell responses against infections and malignancies depend on the release of effector molecules, such as pro-inflammatory cytokines. Because effector molecules can be toxic, their production is tightly regulated through post-transcriptional events at 3’ Untranslated Regions (3’UTRs). RNA binding proteins (RBPs) were shown to be key regulators herein. With an RNA aptamer-based capture assay from human T cells, we identified >130 RBPs interacting with the IFNG, TNF and IL2 3’UTRs in human T cells. T cell activation altered RBP-RNA interactions, revealing that RBP-target mRNA interactions rapidly respond to stimulation. Furthermore, we uncovered the intricate and time-dependent regulation of cytokine production by RBPs: whereas HuR supports early cytokine production, ZFP36L1, ATXN2L and ZC3HAV1 dampen and shorten the production duration, each at different time points. Strikingly, even though ZFP36L1 deletion did not phenotypically rescue T cell dysfunction in tumors, the increased production of cytokines and cytotoxic molecules resulted in superior anti-tumoral T cell responses in vivo. Our findings thus show that identifying RBP-RNA interactions reveals key modulators of T cell responses in health and disease.
T cells are critical players in our defense against infections and malignancies. Their production of effector molecules such as granzymes and proinflammatory cytokines is key. Interferon gamma (IFN-γ) and Tumor necrosis factor (TNF) are major contributors to anti-microbial and anti-tumoral T cell responses, with the most potent T cells co-producing the survival-inducing cytokine IL-2. Whereas the activity of cytokines on target cells is well characterized, the molecular switches that dictate their production is not well understood. Recently, post-transcriptional regulation (PTR) was found to dictate the cytokine production levels. We showed that the strength of T cell receptor signaling together with co-stimulation defines the synthesis and degradation rate of cytokine mRNA in T cells, as well as their translation efficiency.
The 3’Untranslated region (3’UTR) of the mRNA is a major contributor in PTR. For instance, germ-line deletion of cis-elements such as AU-rich sequences (AREs) from the Tnf and Ifng 3’UTR results in hyperinflammation and immunopathology. Conversely, tumor-infiltrating T cells (TILs) fail to produce IFN-γ protein despite their continuous high expression of Ifng mRNA. Germ-line deletion of AREs from the Ifng 3’UTR restored IFN-γ protein production in murine TILs, and boosted their anti-tumoral potency. Notably, this augmented protein production was conserved in human T cells.
RNA binding proteins (RBPs) are critical mediators of PTR that define the fate of mRNA. For instance, ZC3H12A (Regnase-1) and Rc3h1 (Roquin-1) prevent naïve CD4+ T cells from exiting quiescence by destabilizing mRNAs encoding regulators of CD4+ T cell differentiation and function. In contrast, Arid5a promotes T cell differentiation by stabilizing Stat3 and Ox40 mRNA. We found that ZFP36L2 blocks translation of pre-formed cytokine mRNAs in resting memory T cells, thereby preventing aberrant protein production from these ready-to-deploy mRNAs. Importantly, reactivating memory T cells releases the Tnf and Ifng mRNA from ZFP36L2, thereby licensing the immediate cytokine production of memory T cells. Thus, RBPs substantially impact the acquisition and execution of T cell effector function. Which RBPs interact with cytokine 3’UTRs upon T cell activation and define their protein production levels is however not well understood. A comprehensive study on RBP interactions with target mRNAs in T cells is lacking.
More than 2000 RBPs have been annotated in mammalian cells. RNA-RBP interaction maps have been generated for 150 RBPs with Cross-linked immunoprecipitation (CLIP) methods. However, both mRNA expression and RBP expression is cell-type specific, thereby prohibiting direct translation of the RBP interaction maps from epithelial cell lines to human T cells. Furthermore, CLIP depends on the availability of suitable antibodies or on tagging endogenous RBPs. To identify the RBPs that interact with cytokine mRNAs in an unbiased manner, an RNA-centric approach is required. With an RNA aptamer pull-down approach from primary human T cell lysates using full-length cytokine 3’UTRs, we present the first comprehensive analysis of RBP-mediated regulation of T cell effector function, and we reveal the potential of RBP modulation in defining T cell responses to target cells.
To determine how cytokine 3’UTRs contribute to protein production, we retrovirally transduced peripheral blood-derived human T cells with GFP reporter constructs containing the full length 3’UTR of the human GZMB, IFNG, TNF or IL2 mRNA. The empty GFP construct served as control (GFP control). Whereas GZMB 3’UTR only slightly reduced the GFP expression levels in nonactivated CD8+ T cells compared to GFP control, the presence of cytokine 3’UTRs conferred a substantial loss of the GFP signal. T cell activation with PMA/Ionomycin for 6 h and 16 h marginally increased the protein expression for GFP control and GFP-GZMB- 3’UTR. In contrast, the GFP expression levels for cytokine 3’UTR-containing constructs increased by a 2-to 8-fold. No loss of regulation was observed for the endogenous cytokine production. However, the increased GFP levels for cytokine 3’UTRs in activated T cells did not reach the levels of GFP control, indicating further regulation of cytokine production in activated T cells. Similar results were obtained with human CD4+ T cells, and with murine OT-I T cell receptor (TCR) transgenic CD8+ T cells expressing the murine Gzmb, Gzma and Prf1 3’UTR, compared to the Ifng, Tnf or Il2 3’UTRs. The profound regulation of protein expression by cytokine 3’UTRs is thus conserved.
Fig. 1.Conserved regulation of protein production by cytokine 3’UTRs.
We next questioned which RBPs can interact with cytokine 3’UTRs. We assessed putative RBP binding sites in silico with the ATtRACT database, which contains 2297 consensus motifs for >140 RBPs. All three cytokine 3’UTRs contain A-rich, AU-rich, CU-rich, U-rich and G-rich motifs, and several poly(A) sites (PAS), which is also observed for the GZMB 3’UTR. These motifs are potential RBP binding hubs for ZFP36 family members, ELAVL1, KHSRP, PPIE, CSTF2, NUDT21 and other RBPs such as members of the protein families PCBP, PTBP, HNRNP, SRP, SRSF, PABP, RBM and YTHDC. In conclusion, the cytokine 3’UTRs contain a wide range of putative RBP binding sites, which warrants the identification of their actual RBP interactors.
To experimentally identify the RBPs that interact with cytokine 3’UTRs in human T cells, we generated in vitro transcribed streptavidin-binding 4xS1m RNA aptamers with full length 3’UTRs of IFNG, TNF, and IL2. The empty 4xS1m RNA aptamer served as control. The 4xS1m aptamer system efficiently purifies RNPs, because its improved affinity to streptavidin allows for stringent washes to reduce background, without losing bona fide RBP interactions. Because RBP expression is cell-type specific, we performed the RBP pull-down with full cell lysates from human primary CD3+ T cells, allowing to interrogate both nuclear and cytosolic RBP-3’UTR interactions. T cells were isolated from 3 pools of 40 donors as RBP source and activated with α-CD3/α-CD28 prior to culture for 5 days in low IL-2 to generate nonactivated effector T cells. At this time point CD69 expression was low and cytokine production was undetectable. Upon capture of RNA-RBP complexes with streptavidin beads, RNA-interacting proteins were identified by mass spectrometry (MS).
