Chronic graft-versus-host disease (cGVHD) remains a major obstacle impeding successful allogeneic hematopoietic cell transplantation (HCT). MicroRNAs (miRs) play key roles in immune regulation during acute GVHD development. Preclinical studies to identify miRs that affect cGVHD pathogenesis are required to develop these as potential lifesaving interventions. Using oligonucleotide array, we identified miR-31, which was significantly elevated in allogeneic T cells after HCT in mice. Using genetic and pharmacologic approaches, we demonstrated a key role for miR-31 in mediating donor T-cell pathogenicity in cGVHD. Recipients of miR-31–deficient T cells displayed improved cutaneous and pulmonary cGVHD. Deficiency of miR-31 reduced T-cell expansion and T helper 17 (Th17) cell differentiation but increased generation and function of regulatory T cells (Tregs). MiR-31 facilitated neuropilin-1 downregulation, Foxp3 loss, and interferon-γ production in alloantigen-induced Tregs. Mechanistically, miR-31 was required for hypoxia-inducible factor 1α (HIF1α) upregulation in allogeneic T cells. Therefore, miR-31–deficient CD4 T cells displayed impaired activation, survival, Th17 cell differentiation, and glycolytic metabolism under hypoxia. Upregulation of factor-inhibiting HIF1, a direct target of miR-31, in miR-31–deficient T cells was essential for attenuating T-cell pathogenicity. However, miR-31–deficient CD8 T cells maintained intact glucose metabolism, cytolytic activity, and graft-versus-leukemia response. Importantly, systemic administration of a specific inhibitor of miR-31 effectively reduced donor T-cell expansion, improved Treg generation, and attenuated cGVHD. Taken together, miR-31 is a key driver for T-cell pathogenicity in cGVHD but not for antileukemia activity. MiR-31 is essential in driving cGVHD pathogenesis and represents a novel potential therapeutic target for controlling cGVHD.

We previously showed that pericyte-like cells derived from the FoxD1-lineage contribute to myofibroblasts following bleomycin-induced lung injury. However, their functional significance in lung fibrosis remains unknown. In this study, we used a model of lung pericyte-like cell ablation to test the hypothesis that pericyte-like cell ablation attenuates lung fibrosis in bleomycin-induced lung injury. Lung fibrosis was induced by intratracheal instillation of bleomycin. To ablate pericyte-like cells in the lung, diphtheria toxin (DT) was administered to Foxd1-Cre;Rosa26-iDTR mice at two different phases of bleomycin-induced lung injury. For early ablation, we coadministered bleomycin with DT and harvested mice at days 7 and 21. To test the effect of ablation after acute injury, we delivered DT 7 days after bleomycin administration. We assessed fibrosis by lung hydroxyproline content and semiquantitative analysis of picrosirius red staining. We performed bronchoalveolar lavage to determine cell count and differential. We also interrogated mRNA expression of fibrosis-related genes in whole lung RNA. Compared with DT-insensitive littermates where pericyte-like cells were not ablated, DT-sensitive animals exhibited no difference in fibrosis at day 21 both in the early and late pericyte ablation models. However, early ablation of pericytes reduced acute lung inflammation, as indicated by decreased inflammatory cells. Our data confirm a role for pericytes in regulating pulmonary inflammation in early lung injury.

Acute injury of the lung involves damage to the epithelium and its underlying extracellular matrix (ECM), the basement membrane (BM). How BMs contribute to injury resolution is poorly understood. Nephronectin (NPNT) is a high-affinity ligand for integrin α8β1 and, although first identified in the mouse kidney, is prominently expressed in the lung, where it localizes to BMs in the alveoli. To determine if NPNT plays a role in acute injury and inflammation of the lung, we developed a model for postnatal deletion of NPNT using mice with a floxed allele of Npnt in combination with a tamoxifen-inducible Cre recombinase expressed at the ROSA locus. Expression of NPNT was substantially reduced in lungs from tamoxifen-treated Cre+ animals. Cre+ mice and Cre- controls were given E. coli LPS by oropharyngeal aspiration to induce injury and inflammation. In Cre- lungs, although both Npnt and Itga8 (integrin α8) transcripts were downregulated at the peak of inflammation, NPNT protein was still detectable. While the onset of inflammation was similar for Cre+ and Cre-, NPNT-deficient lungs still had thickened alveolar septa and there were increased macrophages in the bronchoalveolar lavage fluid (BALF) in the resolution phase. BALF from Cre+ lungs was more chemotactic for bone marrow-derived macrophages than Cre- in in vitro experiments, but there were no differences in the elaboration of chemokines in vivo. We speculate that absence of NPNT in BMs of the alveoli impairs or delays inflammatory and injury resolution in this model, but further studies are needed to establish the precise role of NPNT in tissue repair.

Nephronectin (NPNT) is a basement membrane (BM) protein and high-affinity ligand of integrin α8β1 that is required for kidney morphogenesis in mice. In the lung, NPNT also localizes to BMs, but its potential role in pulmonary development has not been investigated. Mice with a floxed Npnt allele were used to generate global knockouts (KOs). Staged embryos were obtained by timed matings of heterozygotes and lungs were isolated for analysis. Although primary and secondary lung bud formation was normal in KO embryos, fusion of right lung lobes, primarily the medial and caudal, was first detected at E13.5 and persisted into adulthood. The lung parenchyma of KO mice was indistinguishable from wild-type (WT) and lobe fusion did not alter respiratory mechanics in adult KO mice. Interrogation of an existing single-cell RNA-seq atlas of embryonic and adult mouse lungs identified Npnt transcripts in mesothelial cells at E12.5 and into the early postnatal period, but not in adult lungs. KO embryonic lungs exhibited increased expression of laminin α5 and deposition of collagen IV in the mesothelial BM, accompanied by abnormalities in collagen fibrils in the adjacent stroma. Cranial and accessory lobes extracted from KO embryonic lungs fused ex vivo when cultured in juxtaposition, with the area of fusion showing loss of the mesothelial marker Wilms tumor 1. Because a similar pattern of lobe fusion was previously observed in integrin α8 KO embryos, our results suggest that NPNT signaling through integrin α8, likely in the visceral pleura, maintains right lung lobe separation during embryogenesis.

In this study, we confirmed protein expression differences in only a subset of predicted biomarkers from IPF and control subjects. Most differentially expressed proteins were predicted to be secreted from lung cells. However, a significant minority of the proteins are not known to be transcriptionally expressed by lung cells, suggesting an ongoing systemic response. The contributions of the systemic response in IPF may be important as new therapeutics are developed.