Allogeneic T cell responses are regulated by a specific miRNA-mRNA network

Abstract

Donor T cells that respond to host alloantigens following allogeneic bone marrow transplantation (BMT) induce graft-versus-host (GVH) responses, but their molecular landscape is not well understood. MicroRNAs (miRNAs) regulate gene (mRNA) expression and fine-tune the molecular responses of T cells. We stimulated naive T cells with either allogeneic or nonspecific stimuli and used argonaute cross-linked immunoprecipitation (CLIP) with subsequent ChIP microarray analyses to profile miR responses and their direct mRNA targets. We identified a unique expression pattern of miRs and mRNAs following the allostimulation of T cells and a high correlation between the expression of the identified miRs and a reduction of their mRNA targets. miRs and mRNAs that were predicted to be differentially regulated in allogeneic T cells compared with nonspecifically stimulated T cells were validated in vitro. These analyses identified wings apart-like homolog (Wapal) and synaptojanin 1 (Synj1) as potential regulators of allogeneic T cell responses. The expression of these molecular targets in vivo was confirmed in MHC-mismatched experimental BMT. Targeted silencing of either Wapal or Synj1 prevented the development of GVH response, confirming a role for these regulators in allogeneic T cell responses. Thus, this genome-wide analysis of miRNA-mRNA interactions identifies previously unrecognized molecular regulators of T cell responses.

Figure 1. AGO-CLIP-ChIP procedures and mRNA CLIP-ChIP profiles of Allo T versus CD3/28 T cells.

(A) Schematic illustration of the enrichment profiles of miRNAs and their targets using AGO-CLIP-ChIP. After CLIP, the miRNPs were coimmunoprecipitated with anti-AGO2 antibodies and then bound to the bridge antibody [anti-rabbit IgG F(c)], which links to the protein A agarose beads. The miRNAs and the target mRNAs that associated with the AGO protein were processed using Exiqon miRNA and Affymetrix Gene 430.2 microarrays. This high-throughput assay can screen direct and legitimate miRNA targets in a genome-wide manner and can be used to identify an miRNA-mediated gene regulation network. (B) mRNA CLIP-ChIP profiles of Allo and CD3/28 T cells. Transcript enrichments after CLIP were screened using Affymetrix 430.2 GeneChips. Enrichment values were calculated for each gene using a robust multiarray average. This modeling strategy converts the PM probe values into an enrichment value (log₂ transformed) for each gene. The probe sets with a fold change of 2 or greater were selected (an enrichment value of 2⁶ or greater was used for one of the two samples to prevent the selection of large fold changes based on two small numbers).

Figure 2. miRNA CLIP-ChIP profile.

A total of 44 miRNAs that are annotated in miRBase, version 16.0, were detected after CLIP using the sixth-generation miRCURY LNA Chip. These miRNAs were classified into four groups according to the differential enrichment profiles that were observed in Syn, Allo, and CD3/28 T cells.

Figure 3. GSEA.

To identify the most predominant genes that uniquely respond to allostimulation and exhibit reduced enrichment after CLIP (Table 2), we used GSEA to identify the transcripts that were most significantly overrepresented. The Syn and Allo data were analyzed with the assumption that the number of enriched transcripts in the Syn T cells would be higher than that in Allo T cells. Of the 50 transcripts that exhibited reduced enrichment (Table 1) and were predicted by TargetScan and miRanda to be targets of the group 3 miRNAs (Figure 2 and Supplemental Figure 2), 48 were positively confirmed by GSEA analysis. Of these, 8 transcripts were most significantly enriched, including Wapal, Synj1, and Lpp.

Figure 4. Validation of the selected miRNAs that were enriched in group 3 after AGO-CLIP-ChIP.

Expression patterns of the selected miRNAs in T cells were assessed by TaqMan qPCR. (A) Enrichment of selected group 3 miRNAs in Syn, Allo, and CD3/28 T cells after AGO-CLIP. Data shown are the combined results of three independent experiments (mean ± SEM). (B) Relative miRNA expression levels in whole-cell lysates of Syn, Allo, and CD3/28 T cells. Data shown are the combined results of three independent experiments (mean ± SEM).

Figure 5. Validation of the expression patterns of selected genes targeted by group 3 miRNAs.

(A) mRNA enrichment of selected genes targeted by group 3 miRNAs after AGO-CLIP. Data shown are the combined results of three independent experiments (mean ± SEM). (B) Whole-cell mRNA expression patterns of selected genes targeted by group 3 miRNAs. Data shown are the combined results of three independent experiments (mean ± SEM). (C) Whole-cell protein expression patterns of selected genes (Wapal, Synj1, and Lpp) targeted by group 3 miRNAs. Protein expression was assessed by Western blot. Data shown are representative of three independent experiments. Full, uncut gels are shown in the Supplemental Figure 8.

Figure 6. Use of in vivo BMT to validate selected group 3 miRNAs and the mRNA expression patterns of their predicted targets (Wapal, Synj1, and Lpp).

(A) Expression patterns of selected group 3 miRNAs in isolated naive T cells and donor T cells (Syn T and Allo T cells) on day 14 after BMT. Data shown are the combined results of two independent experiments (mean ± SEM). (B) mRNA expression patterns of selected genes that are predicted targets of group 3 miRNAs. These patterns were determined using isolated naive T cells and donor T cells (Syn T and Allo T cells) 14 days after BMT. Data shown are the combined results of two independent experiments (mean ± SEM).

Figure 7. Effects of Wapal and Synj1 knockdown on allogeneic T cell responses.

Effects of Wapal and/or Synj1 knockdown on the production of (A) Allo T cells, (B) IFN-γ, (C) IL-17, (D) IL-6, and (E) IL-2. (F) Proliferation of T cells stimulated with CD3⁺CD28⁺ antibodies. Data shown are the combined results of three independent experiments (mean ± SEM). (G) Relative mRNA expression of Synj1 and Wapal was significantly increased in OVA-specific OT-II TCR transgenic T cells cocultured with OVA323-339–pulsed DCs. (H) Increased expression of Synj1 and Wapal in OVA-specific TCR transgenic T cells after transplantation. OT-II T cells were purified 7 days after transplantation. mRNA expression of Synj1 and Wapal was assessed by qPCR. The expression of Synj1 and Wapal was significantly higher in OT-II T cells than in naive T cells, but was significantly lower than in B6 T cells following B6→BALB/c BMT. Data shown are the combined results of three independent experiments (mean ± SEM). *P < 0.01; **P < 0.001. KD, knockdown; Scra, scrambled; S, Synj1; W, Wapal.

Figure 8. Knockdown of Synj1 and Wapal in T cells attenuates acute GVHD severity and mortality.

BALB/c mice received 800 cGy of total body irradiation (split dose) and were transplanted with 5 × 10⁶ TCD BM. They also received 0.75 × 10⁶ T cells from allogeneic B6 donors that were transduced prior to infusion with lentiviral particles containing WAPAL- and SYNJ1-specific shRNAs or scramble control ex vivo. Syngeneic B6 mice were given 1,000 cGy of total body irradiation and transplants of 5 × 10⁶ BM and 2 × 10⁶ T cells from B6 Ly 5.2 donors that were transduced with scramble control particles ex vivo. Knockdown of Synj1 and Wapal reduced (A) GVHD clinical severity after allogeneic BMT and (B) increased survival. Data shown are the combined results of two independent experiments (mean ± SEM). **P < 0.03.