The NeST long ncRNA controls microbial susceptibility and epigenetic activation of the interferon-γ locus

Abstract

Long noncoding RNAs (lncRNAs) are increasingly appreciated as regulators of cell-specific gene expression. Here, an enhancer-like lncRNA termed NeST (nettoie Salmonella pas Theiler's [cleanup Salmonella not Theiler's]) is shown to be causal for all phenotypes conferred by murine viral susceptibility locus Tmevp3. This locus was defined by crosses between SJL/J and B10.S mice and contains several candidate genes, including NeST. The SJL/J-derived locus confers higher lncRNA expression, increased interferon-γ (IFN-γ) abundance in activated CD8(+) T cells, increased Theiler's virus persistence, and decreased Salmonella enterica pathogenesis. Transgenic expression of NeST lncRNA alone was sufficient to confer all phenotypes of the SJL/J locus. NeST RNA was found to bind WDR5, a component of the histone H3 lysine 4 methyltransferase complex, and to alter histone 3 methylation at the IFN-γ locus. Thus, this lncRNA regulates epigenetic marking of IFN-γ-encoding chromatin, expression of IFN-γ, and susceptibility to a viral and a bacterial pathogen.

Figure 1. Genotypes of parental and congenic strains used to investigate NeST RNA and the Tmevp3 locus on murine chromosome 10

(A) Schematic of the NeST-encoding genes in mouse chromosome 10 and human chromosome 12. The bars represent exons, arrows indicate the direction of transcription. NeST, previously termed Tmevpg1, is adjacent to both murine Ifng and human IFNG (Vigneau et al., 2003). The major transcript, shown in red, encodes six exons. In both mice and humans, the NeST and IFN-γ-encoding transcripts are convergently synthesized; in humans the transcribed regions overlap. (B) A diagram of the Tmevp3 locus on murine chromosome 10, as defined by the differential ability to clear persistent infection by Theiler’s virus observed between SJL/J and B10.S mice. The SJL/J.Tmevp3^B10.S line, previously termed ‘C2’ (Vigneau et al., 2003), is congenic with SJL/J except for the region shown, from microsatellite marker D10Mit74 to the interval between D10Mit180 and D10Mit233. The B10.S.Tmevp3^SJL/J strain is congenic with B10.S except for the region shown, between the D10Mit151/D10Mit271 interval and the D10Mit233/D10Mit73 interval. The Theiler’s virus (TMEV) persistence and clearance phenotypes and Nramp1 alleles for all four strains are listed. (C) Finer mapping of the polymorphic regions of the congenic lines. The x-axis indicates nucleotide number on mouse chromosome 10. The introgressed region of SJL/J in B10.S.Tmevp3^SJL/J is up to 1.6×10⁷ bp (top) whereas the introgression in SJL/J.Tmevp3^B10.S is approximately 5.5×10⁵ bp (middle and bottom). Each bar displays the number of single-nucleotide polymorphisms (SNPs) in the window size indicated. The most polymorphic region maps to the Tmevp3 locus and coincides with the region of introgression in SJL/J.Tmevp3^B10.S; see Table S1 for lists of all genetic differences between the two Tmepv3 alleles. The physical locations and direction of transcription of the murine NeST, Ifng, Il22 and Mdm1 genes are indicated by arrows. (D) NeST RNA expression in CD3⁺ T cells. The abundance of NeST RNA in CD3⁺ splenocytes from B10.S mice and B10.S.Tmevp3^SJL/J was determined by preparing total cellular RNA and determining the amount of RNA per cell using quantitative RT-PCR and standard curves of transcribed RNAs. The threshold of detection was 0.005 molecules of NeST RNA per cell. Mean values are shown with standard error.

Figure 2. Effect of the Tmevp3 locus on Salmonella pathogenesis

Strains SJL/J and SJL/J.Tmevp3^B10.S were inoculated via oral (A) and intraperitoneal (B) routes with S. enterica Typhimurium. The Nramp1^+/+ alleles expressed by SJL/J and SJL/J.Tmevp3^B10.S mice render them relatively resistant to Salmonella infection. Strains B10.S and B10.S.Tmevp3^SJL/J were also inoculated via oral (C) and intraperitoneal (D) routes with S. enterica Typhimurium at dosages indicated and mortality was monitored. The Nramp1^{169Asp/169Asp} alleles render these mice highly sensitive to Salmonella pathogenesis. In both backgrounds, the SJL/J allele of the Tmevp3 locus reduced mortality after oral inoculation. Statistical significance was determined by the logrank test. (E) B10.S and B10.S.Tmevp3^SJL/J were orally inoculated with S. Typhimurium at 10⁶ CFU/mouse. Bacteria were monitored in spleen and feces at the indicated days. (F) Intracellular bacterial growth was monitored ex vivo in bone marrow derived macrophages from B10.S and B10.S.Tmevp3^SJL/J mice. Lines represent mean of triplicate experiments and statistical significance was determined using the Student t-test. See also Figure S1.

Figure 3. Effect of transgenically expressed NeST RNA on Salmonella pathogenesis

(A) Schematic of transgenes introduced into B10.S mice. SJL/J NeST cDNA (Sea green) and B10.S NeST cDNA (Blue) were cloned downstream of a CD4⁺ and CD8⁺ T-cell specific promoter. The promoter-NeST transgene fragments were used for the construction of transgenic mouse lines in the B10.S background. (B) The abundance of NeST RNA was measured in CD8⁺ splenocytes from B10.S mice congenic for the SJL/J-derived Tmevp3 locus (B10.S.Tmevp3^SJL/J), B10.S mice, B10.S mice containing the SJL/J NeST transgene (B10.S.NeST^SJL/J) and B10.S mice containing the B10.S NeST transgene (B10.S.NeST^B10.S). The amount of RNA per cell was determined using quantitative RT-PCR; in vitro transcribed NeST RNA was used to construct standard curves. (C) B10.S, B10.S.Tmevp3^SJL/J, B10.S.NeST^SJL/J and B10.S.NeST^B10.S mice were orally inoculated with S. Typhimurium at dosages indicated and mortality was monitored. All panels with the 10⁷ CFU/mouse were performed at the same time; the B10.S control is shown in these panels for clarity. Statistical significance was determined by the logrank test. See also Figure S1.

Figure 4. Effect of NeST RNA on Theiler’s virus persistence

B10.S mice, B10.S mice congenic for the Tmevp3 locus from SJL/J mice (B10.S.Tmevp3^SJL/J) and B10.S mice containing the B10.S NeST transgene (B10.S.NeST^B10.S) were inoculated by intracranial injections of 10⁷ PFU of Theiler’s virus. Spinal cords were harvested at 7 days (A) and 57 days (B) post-inoculation and viral load was measured by plaque assay on BHK-21 cell monolayers. (C) The abundance of viral RNA in spinal cord from B10.S, B10.S.Tmevp3^SJL/J and B10.S.NeST^B10.S mice was determined by preparing total cellular RNA from homogenized tissue and determining the amount per gram of tissue using quantitative RT-PCR. TMEV RNA was transcribed from cDNA-containing plasmid to construct standard curves. Means and standard error are shown.

Figure 5. Effect of Tmevp3 locus and transgenically expressed NeST RNA on cytokine expression by T cell subsets

Splenic (A) CD4⁺ and (B) CD8⁺ T cells were isolated from three B10.S (black circles) and three B10.S.Tmevp3^SJL/J (white circles) mice and stimulated ex vivo with PMA and ionomycin. The abundance of IFN-γ and IL-22 protein secreted was determined by ELISA from supernatants collected at the indicated times. Means and standard error are indicated for each time point. Statistical significance was determined using a two-way ANOVA test; asterisks denote those values that differ significantly between T cells derived from B10.S and T cells derived from B10.S.Tmevp3^SJL/J mice. (C) Splenic CD8⁺ T cells were isolated from B10.S (black), B10.S.NeST^SJL/J (sea green) and B10.S.NeST^B10.S (blue) mice and stimulated ex vivo with PMA and ionomycin. The abundance of secreted IFN-γ was determined by ELISA. Asterisks and p values refer to the comparisons between T cells derived from B10.S and T cells derived from each transgenic line. See also Figure S2.

Figure 6. NeST RNA localization and IFN-γ trans activation

(A) Nuclear and cytoplasmic RNA from CD8⁺ T cells from B10.S.Tmevp3^SJL/J, B10.S.NeST^B10.S and B10.S.NeST^SJL/J mice were fractionated by differential centrifugation (Huarte et al., 2010). NeST RNA, unspliced actin RNA (nuclear) and spliced actin RNA (cytoplasmic) from the nuclear and cytoplasmic fractions were assessed by RT-PCR and gel electrophoresis. (B) Quantitation of expression ratios of IFN-γ mRNA from the B10.S and the B10.S.Tmevp3^SJL/J alleles. A natural SNP in the IFN-γ mRNA (coordinate 117882772, see Table S1) was amplified by RT-PCR (top and left panel). cDNAs from B10.S and B10.S.Tmevp3^SJL/J were subjected to a B10.S allele-specific TaqI restriction digest (bottom, left) and fragments were analyzed by gel electrophoresis. (C) Splenic CD8⁺ T cells were isolated from two B10.S×B10.S.Tmevp3^SJL/J heterozygous mice and stimulated with PMA and ionomycin. The proportion of B10.S and B10.S.Tmevp3^SJL/J-derived IFN-γ mRNA was determined by densitometry of the allele-specific restriction fragments. Mixtures of in vitro transcribed RNAs at 1:10 and 1:1 ratios were used as controls.

Figure 7. NeST RNA physical association with WDR5 protein and effect on Histone 3 Lysine 4 trimethylation at the Ifng locus

(A) RNA preparations from 293T cells that were co-transfected with FLAG-tagged WDR5 cDNA and either B10.S-derived NeST cDNA (blue) or SJL/J-derived NeST cDNA (sea green) were analyzed after immunoprecipitation with either anti-FLAG antibodies or anti-IgG control antibodies. NeST RNA retrieval was determined by measuring RNA input levels normalized to GAPDH (bottom left panel). Specific RNA retrieval was determined by subtracting NeST RNA retrieval with anti-IgG antibodies from the retrieval with anti-FLAG antibodies, followed by normalization to the amount of input RNA. Immunoblot analysis (bottom right panel) confirmed FLAG-WDR5 expression following transfection and the specificity of the anti-FLAG and anti-IgG antibodies. (B) IFN-γ production and H3K4me3 occupancy in spleen following immune challenge. B10.S, B10.S.NeST^B10.S and B10.S.NeST^SJL/J mice were injected intraperitoneally with 50 µg of lipopoylsaccharide (LPS); four and six hours later spleens were dissected. The abundance of IFN-γ protein was determined by ELISA in tissue homogenates (top panel) and the occupancy of Histone 3 Lysine 4 trimethylation at the Ifng gene was determined by ChIP-qPCR analysis (bottom panel). A schematic diagram of the positions of primers used for H3K4me3 is shown. Specific DNA retrieval was measured by normalization to the amount of input DNA and GAPDH. (C) ChIP-qPCR analysis of H3K4me3 at the Ifng locus in CD8⁺ T cells from B10.S and B10.S.NeST^SJL/J transgenic mice. CD8⁺ T cells were isolated from four B10.S and four B10.S.NeST^SJL/J mice and stimulated ex vivo with PMA and ionomycin. Occupancy of H3K4me3 at the Ifng gene was assayed 24 hours after stimulation by ChIP-qPCR at four different regions. For all pooled data, means and standard error are shown.