RIG-I/MAVS and STING signaling promote gut integrity during irradiation- and immune-mediated tissue injury

The molecular pathways that regulate the tissue repair function of type I interferon (IFN-I) during acute tissue damage are poorly understood. We describe a protective role for IFN-I and the RIG-I/MAVS signaling pathway during acute tissue damage in mice. Mice lacking mitochondrial antiviral-signaling protein (MAVS) were more sensitive to total body irradiation– and chemotherapy-induced intestinal barrier damage. These mice developed worse graft-versus-host disease (GVHD) in a preclinical model of allogeneic hematopoietic stem cell transplantation (allo-HSCT) than did wild-type mice. This phenotype was not associated with changes in the intestinal microbiota but was associated with reduced gut epithelial integrity. Conversely, targeted activation of the RIG-I pathway during tissue injury promoted gut barrier integrity and reduced GVHD. Recombinant IFN-I or IFN-I expression induced by RIG-I promoted growth of intestinal organoids in vitro and production of the antimicrobial peptide regenerating islet–derived protein 3 γ (RegIIIγ). Our findings were not confined to RIG-I/MAVS signaling because targeted engagement of the STING (stimulator of interferon genes) pathway also protected gut barrier function and reduced GVHD. Consistent with this, STING-deficient mice suffered worse GVHD after allo-HSCT than did wild-type mice. Overall, our data suggest that activation of either RIG-I/MAVS or STING pathways during acute intestinal tissue injury in mice resulted in IFN-I signaling that maintained gut epithelial barrier integrity and reduced GVHD severity. Targeting these pathways may help to prevent acute intestinal injury and GVHD during allogeneic transplantation.


Crypt isolation
Isolation of intestinal epithelial crypts was performed as previously described (26). Briefly, after harvesting small intestines, the organs were opened longitudinally and washed. Small intestine was incubated in 10 mM ethylenediamine-tetraacetatic acid (EDTA) for 25 min (4°C) to dissociate the crypts. The supernatant containing crypts was collected.

Immunohistochemistry
Intestines of mice 8 days after allo-HSCT were harvested, formalin-fixed, paraffin embedded. The immunohistochemical detection of Lysozyme was performed using Discovery XT processor (Ventana Medical Systems). The tissue sections were deparaffinized with EZPrep buffer (Ventana Medical Systems), antigen retrieval was performed with CC1 buffer (Ventana Medical Systems) and sections were blocked for 30 minutes with Background Buster solution (Innovex). Slides were incubated with anti-Lysozyme antibodies (DAKO; cat# A099; 2ug/ml) for 5h, followed by 60 minutes incubation with biotinylated goat anti-rabbit IgG (Vector labs, cat#PK6101) at 1:200 dilution. The detection was performed with DAB detection kit (Ventana Medical Systems) according to manufacturer´s instruction. Slides were counterstained with hematoxylin (Ventana Medical Systems) and coverslipped with Permount (Fisher Scientific). To quantify Lysozyme + Paneth cells, the number of positive cells per crypt was evaluated over a 5000 µm length of intestinal mucosa. Lysozyme + Paneth cells are depicted as mean Paneth cell number / crypt. Quantification was performed in a blinded fashion by S.M.

Detection of bioluminescence and in vivo imaging
Ifn-β ∆β-luc mice were injected i.v. with 100 μl luciferin (30 mg/ml in PBS)/20 g mouse weight and anesthesized using isoflurane. Within 10 min after luciferin injection, mice or isolated organs were analyzed with an in vivo imaging instrument (IVIS 200;PerkinElmer). The acquired images were analyzed using Living Image 4.4. software.

Quantitative PCR
RNA was isolated from cells lysed in TRIzol (ambion) or from whole tissue homogenates. Tissue homogenates were prepared as follows: 1 cm large or small intestine was flushed and longitudinally opened pieces were frozen in 500ul TRIzol reagent using liquid nitrogen. After thawing, samples were supplemented with stainless stell beads 5mm (Qiagen) and homogenized using a Tissuelyser II (Qiagen) 1 min with 30Hz (1800 oscillations/minute). Total RNA was isolated and transcribed using standard methods and kits according to manufacturer´s protocols (RNeasy Mini Kit, Qiagen; SuperScript III Reverse Transcriptase, invitrogen). The specific primer pairs were as follows: mRegIIIγ fwd TTCCTGTCCTCCATGATCAAAA, rev CATCCACCTCTGTTGGGTTCA; mActin fwd CACACCCGCCACCAGTTCG, rev CACCATCACACCCTGGTGC; mLgr5 fwd ACCCGCCAGTCTCCTACATC rev GCATCTAGGCGCAGGGATTG; mLysozymeP fwd CAG GCCAAGGTCTACAATCG, rev TTGATCCCACAGGCATTCTT.; mItgb6 fwd ATTGTCATTCCCAATGATGG, rev CATAGTTCTCATACAGATGGAC. The qPCR Core kit for SYBR Green I (Eurogentec) and a LightCycler 480 II (Roche) Real-Time PCR System were used as indicated by the manufacturer. The relative transcript level of each gene was calculated according to the 2−Ct, for unnormalized genes, and the 2−ΔΔCt method, for the genes normalized to β-Actin. Alternatively, the following Taqman Expression Assay IDs were used: BETA-ACTIN Mm01205647_g1; IFNB1 Mm00439552_s1; REG3G Mm00441127_m1; Measurement of cytokines TNF and IL-6 were analyzed using the Cytometric Bead Array Enhanced Sensitivity Flex Set System (BD) according to manufacturer's instructions. IFNα and IFNβ were analyzed by ELISA (PBL Assay Science) according to manufacturer's instructions.

Assessment of epithelial regeneration in intestinal organoid cultures
To determine the effect of 3pRNA / interferon stimulatory DNA / rmIFN-β on organoid size and morphology, bright-field microscopy images were taken using a Zeiss Axiovision Observer microscope with a 5x objective lens after 5 or 7 days in culture. 2D area and perimeter were analyzed using border perimeter tracing of organoids found in four representative fields of each well using Image J software. For assessment of gene expression by quantitative (q) PCR, organoids were subjected to RNA extraction 24 hours after culture using Trizol reagent (Invitrogen) according to manufacturer's protocol. Isolated RNA was reverse-transcribed using the Quantitect Reverse Transcription Kit (Qiagen). Gene expression was assessed by quantitative real-time PCR using Taqman Expression Assay pre-designed probes (Applied Biosystems). Signals were normalized to β-Actin. mRNA expression. Normalized values were used to calculate relative expression by ΔΔCt analysis or absolute expression by ΔCt. Taqman IDs are depicted below (qPCR).

16S RNA gene Sequencing
Stool specimens were stored at -80°C. DNA was purified using a phenol-chloroform extraction technique with mechanical disruption (bead-beating) based on a previously described protocol (48) and analyzed using the Illumina MiSeq platform to sequence the V4-V5 region of the 16S rRNA gene. Sequence data were compiled and processed using mothur version 1.34(49), screened and filtered for quality (50), then classified to the species level (51) using a modified form of the Greengenes reference database (52), screened and filtered for quality (50), then classified to the species level (51) using a modified form of the Greengenes reference database (52).

Quantification of Plasma DNA levels
Mouse plasma was collected from peripheral blood (8800rcf, 10min). Plasma samples of 3-4 mice were combined to a final volume of 400-500μl and DNA extracted using the QIAamp Circulating Nucleic Acids Kit (Qiagen). dsDNA was quantified using a Qubit 2.0 Fluorometer with the Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific).
GVT model and bioluminescence imaging A20-TGL (H-2 d ), a BALB/c B-cell lymphoma, were generated as described previously (53). A20-TGL tumor cells were inoculated via separate intravenous injection on the day of allo-BMT (54). To visualize and quantify tumor burden, A20-TGL inoculated mice were administered D-luciferin (Goldbio), anesthetized, and imaged using in vivo bioluminescence imaging systems (Caliper Life Sciences) Cell lines, culture and RNA transfection, feces RNA isolation Mode-K cells were purchased from Dominique Kaiserlian (French Institute of Health and Medical Research, Unit of Immunity Infection Vaccination, France) and cultured as previously described (31). Cell lines were tested as mycoplasma negative. Where indicated, MODE-K cells were transfected with mouse RIG-I siRNA (100μM, Eurofins Genomics,) or control siRNA (Qiagen) using Lipofectamine 2000 (Life Technologies) according to manufacturer's instructions. After 48 h, cells were transfected with 3pRNA (0.8μg/mL) or mouse feces-derived RNA complexed to Lipofectamine 2000. Supernatants were collected and RNA was extracted 18h after transfection followed by IFN-β measurement with ELISA (PBL Assay Science) or by assessment of IFN-β mRNA by qPCR. Mouse feces from healthy WT mice was diluted (RNAprotect Reagent, Quiagen) and homogenized with Glass beads (Sigma) and a Tissuelyser II (Qiagen). After centrifugation, supernatant was subtracted and total feces RNA was isolated using standard methods and kits according to manufacterers' protocols.

Gene Expression Profiling Analysis
For gene expression profiling analysis, (i) Balb/c mice were solely irradiated (9Gy) (n=3), (ii) pretreated with 3pRNA prior (d-1) to irradiation (n=3) or (iii) pre-treated with 3pRNA (d-1) + α-IFNaR1 blocking antibody (d-2) prior to irradiation (n=3). RNA from small intestines was isolated 12 h after irradiation and used for RNA sequencing. Poly(A) RNA sequencing was performed with three biological replicates for each group and analyzed with an Illumina HiSeq2500 platform. The heatmap depicted in Fig. S4D shows all genes listed in the interferome database (55) that show significantly changed gene expression of 3pRNA pretreated and irradiated mice compared to both the other groups simultaneously.

Data Analyis
The output data (FASTQ files) were mapped to the target genome using the rnaStar aligner that maps reads genomically and resolves reads across splice junctions. We used the 2 pass mapping method in which the reads are mapped twice. The first mapping pass used a list of known annotated junctions from Ensemble. Novel junctions found in the first pass were then added to the known junctions and a second mapping pass was done. After mapping we computed the expression count matrix from the mapped reads using HTSeq (wwwhuber.embl.de/users/anders/HTSeq) and one of several possible gene model databases. The raw count matrix generated by HTSeq was then processed using the R/Bioconductor package DESeq (www-huber.embl.de/users/anders/DESeq) which was used to both normalize the full dataset and analyze differential expression between sample groups.
A heatmap was generated using the heatmap.2 function from the gplots R package. The data plot was the mean centered normalized log2 expression of the top 100 significant genes. For simple hierarchical clustering the correlation metric was used (Dij = 1cor(Xi,Xj)) with the Pearson correlation on the normalized log2 expression values.

Statistics
Animal numbers per group (n) are depicted in the figure legends. We never used technical replicates. GraphPad Prism version 6 was used for statistical analysis. Survival was analyzed using the Log-rank test. Differences between means of experimental groups were analyzed using two-tailed unpaired t test or ordinary one-way Anova correspondingly to the distribution shape of our observations. We used ordinary one-way Anova for multiple comparisons and always performed Dunnett's test for Multiple-test corrections. Applied statistical tests are indicated in the figure legends. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). Data are presented as mean ± S.E.M. Ly-6G/Ly-6C (Gr-1) RB6-8C5 1DB-001-0000839101  Animal numbers per group (n) are depicted. (I) Survival of BALB/c mice that received 8.5Gy TBI + 5x10 6 BM alone or 5x10 6 BM and 0.5x10 6 T cells (donor C57BL/6 into recipient BALB/c) and that were inoculated with 0.25x10 6 A20 tumor cells. Pooled data of 2 independent experiments. Animal numbers per group (n) are depicted. (J) Allo-HSCT recipients were inoculated with A20-TGL and in vivo bioluminescence imaging was conducted to determine tumor burden. Bioluminescence of one representative experiment on d21 after allo-HSCT is shown. All experiments were analyzed using twotailed unpaired t test or ordinary one-way Anova for multiple. Survival was analyzed using the Log-rank test. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). If not otherwise indicated, significance was calculated compared to untreated groups. Data are presented as mean ± S.E.M.  The lower panel shows mice that received combination treatment of 3pRNA and α-IFNaR1 blocking Ab at indicated time points. Animal numbers per group (n) are depicted. All experiments were analyzed using two-tailed unpaired t test or ordinary one-way Anova for multiple comparisons. Survival was analyzed using the Log-rank test. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). If not otherwise indicated, significance was calculated compared to untreated groups. Data are presented as mean ± S.E.M.

Figure S5. RIG-I-induced IFN-Is enhance epithelial regeneration through stimulation of the ISC compartment. (A) 90 days after syngeneic bonemarrow transplantation between
Ifnar1 -/and Ifnar1 +/+ mice (C57BL/6), bonemarrow chimera were lethally irradiated and transplanted with 5 × 10 6 BM cells with 1 × 10 6 T cells from donor B10.BR mice and monitored for survival. Indicated mice were either left untreated or treated with 3pRNA on d-1. Pooled data of 2 independent experiments. Animal numbers per group (n) are depicted. (B) Survival of either Ifnar1 fl/fl or Ifnar1 fl/fl CD11cCre allo-HSCT recipients (donor BALB/c into recipient C57BL/6) in the presence or absence of 3pRNA (d-1). Pooled data of 2 independent experiments. Animal numbers per group (n) are depicted (C) Measurement of organoid size (perimeter) of C57BL/6 small intestinal organoids after 5 days in culture. Indicated crypts were treated with 3pRNA (2µg/ml), α-IFNaR1 blocking Ab (10ug/ml) or IgG1 Isotype control (IC). The experiment was performed 3 times and one representative experiment is shown. (D) IFNβ mRNA transcript expression 24 hours after 3pRNA stimulation of Mavs +/+ or Mavs -/small intestinal organoids. The experiment was performed 3 times and resulting data were pooled. (E) Measurement of organoid size (perimeter) of C57BL/6 small intestinal organoids after 5 days in culture. Indicated crypts were treated with rec. IFN-β (20U/ml). The experiment was performed 3 times and one representative experiment is shown. (F) Number of organoids of C57BL/6 small intestinal organoids after 7 days in culture. Indicated Crypts were treated with rec. IFN-β (20U/ml). (G) Number of C57BL/6 small intestinal organoids treated as in (C) after 7 days in culture. Survival was analyzed using the Log-rank test. All other experiments were analyzed using two-tailed unpaired t test or ordinary one-way Anova for multiple comparisons. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). Data are presented as mean ± S.E.M. Determination of Lysozyme + paneth cells per crypt in the ileum of untreated Mavs +/+ and Mavs -/mice using immunohistochemistry (IHC). Pooled data of 2 independent experiments. Animal numbers per group (n) are depicted. (C) RNA of small intestines from Balb/c WT mice isolated 24 hours after irradiation (9Gy). Gene expression was determined by qPCR. Relative transcript levels of Lysozyme P and Lgr5 were normalized to the housekeeping gene β-Actin. Pooled data of 2 independent experiments. Animal numbers per group (n) are depicted. All experiments were analyzed using two-tailed unpaired t test or ordinary one-way Anova for multiple comparisons or. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). Data are presented as mean ± S.E.M.

Figure S6
Mavs were analyzed using two-tailed unpaired t test or ordinary one-way Anova for multiple comparisons. Significance was set at p values < 0.05, p < 0.01 and p < 0.001 and was then indicated with asterisks (*, ** and ***). Data are presented as mean ± S.E.M.