The mechanisms responsible for cardiac comorbidities in psoriasis and rosacea are unknown. Prior clinical observations have shown the human cathelicidin peptide LL37 accumulates in atherosclerotic plaques, and we hypothesized that LL37 may contribute to these comorbidities. We first found that LL37 induced LDL aggregation and a large, 6-fold increase of LDL entry into THP-1 macrophages. This occurred at concentrations as low as 78 nM, similar to the elevated serum concentrations of LL37 in psoriasis and rosacea. Proximity ligation assays showed that LDL binding to the cell surface was promoted by the capacity of LL37 to bind LDL to LDL receptors (LDLR, SR-B1 and CD36). Blockade with endocytosis inhibitors or use of specific blocking antibodies for LDLR, SR-B1 and CD36 showed that this increased binding enabled cell uptake. Synchrotron x-ray scattering measurements of LL37-LDL complexes, and structure-function studies of LL37 with sequential substitutions of alanine at each aa position, further defined the mechanism of uptake and identified that the hydrophobic alpha-helical face of LL37 is important to promote LDL aggregation and described the unique structure of the aggregate generated by LL37 that is required for LDL uptake. RNA sequencing revealed that LDL-LL37 promoted transcriptional responses in macrophages consistent with elevated cytosolic LDL including altered expression of lipid metabolism genes such as SREBF2 and HMGCR. Since the specific structure defined to promote uptake was unique to LL37, and not found in the mouse cathelicidin ortholog, we next developed a humanized LL37-transgenic mouse on an apolipoprotein E null background to examine the importance of these observations in vivo and found that LL37 mice fed a high fat diet developed larger aortic plaques than control mice. Overall, our study implicates LL37 as a contributing factor for development of cardiovascular comorbidities.

Diseases of the skin and gut overlap in inflammatory disorders such as atopic dermatitis and food allergy or psoriasis and inflammatory bowel disease. Accumulating evidence suggests gut microbes affect distant organs such as in the gut-brain axis, but little recognition has been given to the contribution of the skin to the function of the gut or understanding of the mechanisms responsible for associations between skin and gut disease. We previously reported that the release of hyaluronan fragments after dermal injury triggers susceptibility to colitis (Dokoshiet al. JCI 2021). To understand the mechanisms for this response, mouse intestine was studied by both spatial and single-cell RNASeq after K14 driven expression of hyaluronidase (K14/HYAL1). Skin-specific interventions increased the expression of mRNA for host defense genes Muc2 and Reg3in the proximal colon. This was confirmed by immunostaining. Furthermore, this change in intestinal host defense led to an altered gut microbiome as metagenomic analysis of feces showed that skin injury resulted in significant differences in Shannon alpha diversity (p=0.008) and robust Aitchison beta diversity (p=0.02) in the fecal microbiome. Skin injured mice also had more dead bacteria and greater penetration of bacteria into the colonic epithelium. The increased colitis after skin wounds or K14-Hyal1 was rescued in germ-free mice or SPF mice given oral vancomycin. Furthermore, fecal microbiome transplantation from mice with skin injury to gnotobiotic mice enhanced inflammatory response in DSS (2-fold induction of IL6: p<.05, shorter colon). These findings suggest exacerbation of intestinal disease after skin-specific intervention was due to a change in gut bacteria and show how the skin and release of hyaluronan fragments from the dermis can drive effects on systemic health.

Atopic dermatitis (AD) is susceptible to S. aureus colonization which exacerbates this disease. AD has reduced expression of cathelicidin (Camp) and b-defensin (Defb) antimicrobial peptides (AMPs) compared to other inflammatory disorders but the expression of these AMPs is still higher in AD than on normal skin. To understand why increased AMPs in AD compared to normal skin would enable increased S. aureus we analyzed the skin of Il4ra-/- or WT mice with or without MC903-induced Th2 inflammation. scRNA-Seq showed that topical application of S. aureus to WT skin increased frequency of Il17a+ Th17 subsets but enhanced frequency of Gata3+ Th2 subsets when pretreated with MC903. In contrast, Il4ra-/- skin shifted back toward Th17-predominant response to S. aureus despite MC903 pretreatment. In keratinocytes, scRNAseq identified diminished induction of multiple AMP genes such as S1008/9, Lnc2 and Camp from WT mice treated with MC903 and this was associated with an increase in S. aureus survival. Il4ra-/- mice exposed to MC903 had greatly increased AMP expression in keratinocytes relative to WT mice and a 3-log (P<.01) decrease in S. aureus survival, thus supporting a central role of IL-4Ra in suppressing the innate immune response to S. aureus but not explaining why S. aureus was increased relative to normal skin. We next asked if Th2 inflammation inhibited survival of coagulase-negative Staphylococcus (CoNS) strains producing bacteriocins (CoNS-Bac+) that kill S. aureus. CoNS-Bac+were 2-8 times more susceptible in vitro to human AMPs, LL-37, DEFB3 and DEFB2, than CoNS-Bac- (P<.001) and these protective bacteria were killed by low levels of AMPs that could still be produced by the Th2-inflamed skin. Notably, the selective killing of CoNS-Bac+ was rescued on the skin of Camp-/- mice with Th2 inflammation. These results show that the combination of both a partial loss of host AMPs and elimination of CoNS-Bac+during Th2 inflammation can explain why S. aureus expands on AD skin.

Neutrophil-associated skin diseases such as bacterial infection, psoriasis, and other dermatoses involve type 17 inflammation. Advancements in treatment may come from a more complete understanding of the cell types that participate in the pathogenesis of disease. To address this, we first performed an unbiased investigation of cell-cell communication during type 17 skin inflammation using a machine learning approach with single-cell RNA sequencing (scRNA-Seq) data from several different human skin diseases and mouse models.

Immune fibroblastic cell (IFC) subpopulations of dermal fibroblasts differentiate into adipocytes and provide skin immune defense in response to infection. To better understand the role of these fibroblast cell subsets, we have undertaken a systematic analysis of dermal fibroblasts in models of skin inflammation. First, communication from fibroblasts to mesenchymal, epithelial, and bone-marrow derived cells in the skin was investigated using unbiased machine learning of single-cell RNA sequencing (scRNAseq) data from mice infected with Cutibacterium acnes. This network analysis revealed that fibroblasts communicate more with myeloid cells than any other cell lineage. Further analysis of scRNAseq data showed that lipocalin 2 (LCN2/NGAL)—a myeloid cell chemoattractant and antimicrobial protein—was a major gene upregulated by IFCs (17-fold change, p<0.0001), and LCN2/NGAL was most highly expressed by an IFC subset that expressed genes related to fat cell differentiation. Next, LCN2/NGAL+ IFCs were visualized in situ in the skin of mice infected with Staphylococcus aureus and modeled in vitro with the 3T3-L1 preadipocyte IFC cell line. 3T3-L1 IFCs showed massive induction of LCN2/NGAL mRNA following initiation of adipogenesis by bulk RNA sequencing (340-fold change, p<0.0001) and qPCR (59-fold change, p<0.0001). This response was further increased by TLR2 stimulation (5-fold change, p<0.0001). Functionally, conditioned media from TLR2 activated 3T3-L1 IFCs demonstrated robust in vitro chemotactic activity for neutrophils compared to unstimulated 3T3-L1 IFCs (4-fold change, p<0.0001), and LCN2/NGAL−/− mice showed reduced C. acnes induced inflammation. These findings show for the first time that dermal adipocyte precursor IFCs express LCN2/NGAL and likely promote host defense through interaction with myeloid cells.

Intestinal and skin microflora can influence several systemic health behaviors. In this study, we hypothesized that skin and gut tissue microbiota are linked, and that skin health can control the gut microbiota. To test this, we compared intestinal responses in control mice to co-housed littermates with full-thickness incisions on back skin or K14Cre-targeted expression of HYAL-1 in the epidermis. The latter mice represent a skin-specific model of the hyaluronidase activity produced by wounding but without a systemic inflammatory or stress response. Both skin-specific interventions greatly exacerbated intestinal disease when mice were subsequently challenged by oral Dextran sulfate sodium salt(DSS ) (60% vs 0% mortality, 11-fold induction of IL6 in survivors, p<.001). The DSS response was dependent on the gut microbiome as Germ-free mice or SPF mice given oral Vancomycin were rescued from the DSS effects exacerbated by skin injury or K14HYAL1. 16S rDNA sequencing of feces confirmed that the skin interventions altered the gut microbiome and showed a relative decrease in gram-positive bacteria including protective bifidobacterium. Mechanistically, this response could be explained by the results of spatial transcriptomics using 10X Visium sequencing that showed the colon of wounded or K14HYAL1 mice had greatly increased expression of Mucin 2(Muc2) and Antimicrobial Peptide Regenerating Family Member 3(Reg3). qPCR and immunochemistry validated this increased Muc2 and Reg3b in the transverse colon. Furthermore, this response was modeled by addition of 6.8kDa hyaluronan fragments to gut epithelial cells (HT-29), showing these HA fragments produced by skin injury or K14HYAL1 expression resulted in a 1.7 fold greater Muc2 and 2 fold greater Reg3a expression (p<.01) in vitro. Since intestinal mucin and Reg3 are known to regulate gram-positive bacteria in the intestine, these findings demonstrate how skin injury alters the gut microbiome and susceptibility to disease.

LL37 is increased in sera of obese patients and in patients with psoriasis and rosacea. These diseases all have significant cardiac comorbidities. Since LL37 is a multifunctional molecule that also promotes inflammation through its capacity to enhance the uptake of nucleic acids into the cytosol, we hypothesized that it may also contribute to development of cardiovascular disease by influencing cellular lipid accumulation. To test this, we first studied lipoprotein uptake and delivery into macrophages in the presence of LL37 or a C-terminal 3aa truncated peptide (LL34) by using Phrodo-labeled fluorescent LDL. LL37 and LL34 induced a 6-fold increase of LDL entry into THP-1 (P=0.017 and P < 0.001, respectively) or mouse peritoneal macrophages. LL34 also promoted a 12X increase in formation of LDL aggregates (P< 0.0001), a process that promotes LDL entry. These aggregates enlarged over time (6hr) and could be induced at LL37 concentrations as low as 313 nM, similar to LL37 serum concentrations in psoriasis and rosacea. LL34-induced LDL uptake was blocked by endocytosis inhibitors and partially decreased by targeting of SR-B1 (P < 0.01). RNA-Seq analysis, confirmed by qPCR, revealed transcriptional responses consistent with elevated cytosolic LDL including decreased expression of lipid metabolism genes such as SREBF2, HMGCR and LDLR. FACS analysis of peritoneal macrophages from transgenic mice expressing human LL37 showed elevated LDL uptake. Structure-function studies with sequential substitutions of alanine at each aa position of LL34 showed the hydrophobic alpha-helical face of LL37 is important to promote LDL uptake as well as inflammatory cytokine responses. These structural determinants are being defined by synchrotron x-ray scattering measurements of LL37 LDL complexes. Overall, our observations implicate LL37 as a contributing factor for cardiovascular comorbidities observed in some inflammatory skin disorders.

To improve understanding of acne we evaluated host cell networks in the pilosebaceous unit and variables associated with C. acnes that promote skin inflammation. Single-cell RNA sequencing uncovered unsuspected activation of specific dermal fibroblast subsets in both human acne and mice challenged by C. acnes. This transcriptional response was characterized by initiation of fat cell differentiation and expression of cathelicidin (Camp) by fibroblasts; responses that were validated by immunostaining of PREF1 and CAMP in these specialized perifollicular immune fibroblastic cells(IFCs). mRNA interactome analysis showed that these novel IFCs share a major inflammatory communication network with myeloid cells in mice and human acne. 3T3L1 fibroblasts were then used as a model for IFCs. Analysis of protein and mRNA expression by these cells and in mice injected with C. acnes showed bacterial strain-specific proinflammatory triggers associated with bacterial genomic sequence, growth conditions and plasmid incorporation, not only phylotype. Next, functional relevance of IFCs was demonstrated in C. acnes-injected mice treated with retinoic acid(RA). After RA treatment, epidermal lipid synthesis decreased while Camp expression in PREF1+ fibroblasts significantly increased (p>0.0001), and this associated with a decrease in acneiform lesions (p>0.004). Skin biopsies of human acne patients treated with isotretinoin also showed increased cathelicidin in dermal fibroblasts compared to pre-treatment. However, RA treated Camp-/- mice showed no significant decrease in C. acnes acneiform lesions despite inhibition of lipid synthesis. These observations reveal previously unsuspected functions for dermal IFCs, and novel variables associated with C. acnes, that are involved in the pathophysiology of acne.

We recently reported that short-chain fatty acids (SCFA) promote an inflammatory response in keratinocytes by suppression of HDAC8 or HDAC9, specific histone deacetylases whose activity increases tolerance of the skin to inflammatory signals. Upon silencing of HDAC8 or 9 in keratinocytes, subsequent exposure to TLR2/6, TLR3, or TLR7 ligands enhances inflammatory cytokine production in keratinocytes, but this effect does not occur in bone-marrow derived cells, thus demonstrating epidermal specificity of this mechanism. Chip-Seq and signal pathway analysis by RNA-Seq identified MAP2K3 as a key intermediate in this process, with increased acetylation at H3K9 and H3K27 in the MAP2K3 promoter after silencing HDAC8 and HDAC9 or inhibition of HDAC activity by SCFA butyrate. Antibody pull-down and mass spec analysis showed that HDAC8 and HDAC9 bind the FACT complex to drive gene elongation. These responses were ablated with MAP2K3 knock down. Furthermore, HDAC8 and HDAC9 silencing in keratinocytes lead to IFN-β-dependent activation of antigen presentation ability in cultured dendritic cells and enhanced T cell proliferation in culture. Increased immune reactivity of keratinocytes was also seen in K14Cre-HDAC8/9flox mice in response to UV radiation or imiquimod application, thus validating the critical role of this epigenetic mechanism in the skin. To exploit a potential benefit of HDAC8 and HDAC9 inhibition, we evaluated the impact of HDAC inhibition by topical application of SCFAs on survival of S. aureus. Topical treatment of mice with butyrate upregulated antimicrobial peptide production (Camp and mBD4) and subsequently inhibited S. aureus in mice despite elevated Th2 cytokines generated in an MC903-induced AD mouse model. These observations show a novel approach to enhance host defense against pathogens on human skin.

Inflammatory bowel diseases (IBD) associate with skin inflammatory diseases but why this occurs is unknown. Inflammation or wounding of the skin induces Cemip hyaluronidase in the dermal extracellular matrix (ECM) and production of HA fragments that have been shown to be recognized by TLR4. We hypothesized that such HA fragments may enable the skin to promote inflammation in the gut. To test this, mice expressing hyaluronidase in the skin (K14/Hyal1 mice), or mice with skin wounds (Wd), were compared to their littermate controls. Both groups showed HA digestion in the dermis but expression of Hyal1 did not induce skin inflammation. Remarkably, both skin interventions enhanced disease in the colon when mice were fed DSS, seen histologically, by increased weight loss (p<0.0001), lower survival rates (Control 100% survived, K14/Hyal1: 20%, Wd; 80%), and FACS. Even in the absence of DSS challenge, scRNA Seq of colons from K14/Hyal1 revealed large changes in the abundance of stromal fibroblast subsets; cluster 5 increased from 1.21% to 43.2%, clusters 0, 2 and 7 decreased from 29.8 to 4.16%, 18.9 to 0% and 4.5 to 0%, respectively). Pseudotime analysis distinguished three lineages within these populations with HA-induced shift from Cluster 2 toward 5 in lineage 3 most associated with fat cell differentiation. Genes altered in these subsets were validated by whole tissue RNA Seq and qRT-PCR. Colon fibroblasts from TLR4-/- mice failed to respond to HA fragments. DSS challenge in mice further induced genes related to adipogenesis. Reanalysis of scRNA-Seq data from healthy human subjects and patients with newly diagnosed IBD was consistent with our observations in mice of increased fat cell differentiation. Taken together, these results show the role of fibroblasts and reactive adipogenesis in tissue inflammation and directly demonstrate how the skin can control intestinal inflammation.

Appropriate immune function of the epidermis requires recognition of danger while also tolerating diverse environmental and microbial exposures. Previously, we observed tolerance of the skin to TLR2 and 3 ligands is inhibited by metabolic products of the microbiome that act to inhibit HDAC8 and HDAC9 in keratinocytes (Science Immunol 2016). In this study, we sought to gain a detailed mechanistic understanding of this newly recognized system of innate immune tolerance. We observed that the promoters of inflammatory cytokines suppressed by HDAC8/9 were not direct targets of these deacetylases, but RNA-seq analysis after siRNA silencing of HDAC8 or 9 identified MAP2K3 as a gene target that could indirectly control expression of multiple inflammatory cytokines. Western blot confirmed HDAC8 and 9 were regulators of MAP2K3 as chemical inhibition or gene silencing of HDAC8 or 9 increased MAP2K3 protein expression and increased phosphorylation of p38MAPK. Chip-qPCR showed direct binding of HDAC8 and 9 to the MAP2K3 gene (P value=0.008). H3K9 and H3K27 acetylation marks were also increased in the MAP2K3 promoter after HDAC8 or 9 silencing (P value=0.0162 and 0.0323). Immunoprecipitation and mass spec identified both HDAC8 and 9 were complexed with FACT proteins SSRP1 and SPT16H; genes that act in transcriptional elongation. Silencing of these FACT proteins in keratinocytes abolished the effect of HDAC8/9 inhibition to induce inflammatory cytokines. HDAC8/9 silencing in keratinocytes also resulted in IFN-β-dependent activation of dendritic cell function leading to increased T cell proliferation (P value=0.010). Finally, following UVB radiation, HDAC8-/- or 9-/- mice showed a visible increase in skin inflammation and increased IFN-β mRNA (increase of 223.8% and 271.5%, respectively, P value=0.02). A similar increased inflammation occurred after imiquimod treatment of HDAC8-/- or 9-/- mice (P value=0.04). Taken together, we provide a detailed understanding of a novel innate immune tolerance mechanism in the skin.

Inflammatory bowel diseases (IBD) are associated with several skin inflammatory diseases but the mechanism responsible for communication between organs is unknown. Local immune activation occurs in part by the action of induced hyaluronidase in the extracellular matrix (ECM) and subsequent recognition of soluble HA fragments. We hypothesized that such HA fragments may also act to enable organ crosstalk between skin and gut. To test this, the intestine was examined in mice expressing hyaluronidase in the skin (K14/Hyal1 mice). K14/Hyal1 mice do not show spontaneous skin inflammation. These mice were then compared with littermate controls or mice with incisional skin wounds (Wd) that induce the endogenous dermal hyaluronidase Cemip. Both groups showed HA digestion in the dermis. Remarkably, all skin specific interventions enhanced DSS-induced inflammation in the colon as seen by greater weight loss (p<0.0001) and lower survival rates (Control 100% survived, K14/Hyal1: 20%, Wd; 80%). Transcriptional profiling by single cell RNA Seq revealed that expression of hyaluronidase in skin promoted large changes in the abundance of colon fibroblast subsets; cluster 5 of 9 increased from 1.21% to 43.2%, and clusters 0, 2 and 7 decreased from 29.8 to 4.16%, 18.9 to 0% and 4.5 to 0%, respectively). Genes altered in these subsets were validated by whole tissue RNA Seq and qRT-PCR, and after DSS challenge showed activation of genes related to reactive adipogenesis in colon. FACS analysis also showed increase in PDGFRa and Thy-1 positive cell populations (K14/Hyal1: p=0.0001, Wd; p=0.07) as well as a shift in resident colon RORgt Tregs (vs K14/Hyal1: p=0.009, Wd; p=0.013). The fecal microbiome also significantly changed in K14-Hyal1 mice compared to cohoused littermate controls(vs K14/Hyal1: p=0.05). Taken together, these data show how disruption of skin ECM HA may explain associations between skin and gut disorders.

Hyaluronan (HA) has essential functions in control of immune reactions. Digestion of HA into small fragments occurs following tissue injury and in other inflammatory conditions, resulting in loss of high molecular weight HA (HMW-HA) from the dermis. The mechanism responsible for this loss of HMW-HA has been unknown since expression of classical hyaluronidases does not change after injury. In this study, we investigated a newly discovered hyaluronidase (Cemip), and its role in the innate immune response to S. aureus infection. Transgenic mice lacking Cemip (Cemip-/-) failed to digest dermal HMW-HA after S. aureus infection as measured by ELISA that detected a 2X more abundant amount of total HA, or electrophoresis of dermal extracts that revealed HA to have larger size of 1000 to 500KDa in Cemip-/- vs less than 50KDa in controls. This increased HMW-HA due to the loss of Cemip promoted excessive reactive adipogenesis as seen histologically and by elevated mRNA expression of PPARg (P=0.04) and by increased PDGFRα+ preadipocytes measured by FACS from cells in the dermis. Persistence of HMW-HA also resulted in more CD11c+/MHCII+ DCs (Control: 7.4%, KO: 9.2%, P=0.01) and Ly6-G+/CD11b+ Neutrophils (Control: 8.5%, KO: 9.9%, P=0.04) in uninjured skin. After infection, Cemip-/- mice had increased mRNA expression of IL-6 (Control: 4.9 vs KO:31, 4.3 fold change, P<0.01) and Camp (Control: 1.9 vs KO: 9.4, 5.7 fold change, P<0.01), increased cathelicidin protein staining and enhanced recruitment of CD11b+/F4/80+/Ly6-C+ monocytes (Control: 37%, KO: 57%, P=0.04). This increased inflammatory state protected against S. aureusinfection as challenge of Cemip-/-mice with bioluminescent S. aureus and subsequent in vivo imaging by IVIS revealed Cemip-/- mice had a superior capacity to kill bacteria. Taken together, these findings show for the first time that Cemip is important for digestion of HA in the dermis and is an important regulator of skin inflammation and antimicrobial defense

Dermal white adipose tissue is activated upon skin infection or injury and this process of reactive local adipogenesis is an essential part of innate immune defense of the skin against bacterial infection. We recently showed that reactive adipogenesis also takes place in the colon in response to injury stimulated by dextran sodium sulfate (DSS). In the epithelial setting of the intestine, reactive adipogenesis is important to resist bacterial translocation but also promotes inflammation and colitis. Since adipogenesis is dependent on hyaluronan (HA) in the surrounding extracellular matrix (ECM), we investigated mice over-expressing human-Hyaluronidase 1 (hHyal1) during embryogenesis (EIIa/Hyal1) or only in skin (K14/Hyal1 mice). EIIa/Hyal1 mice showed less adipogenesis of skin and colon, less inflammation in both organs, and increased bacterial invasion at both sites. K14/Hyal1 mice had normal HA in their colon and less HA in skin as expected after K14 targeted Hyal1 expression. Unexpectedly however, K14/Hyal1 mice with DSS colitis had increased adipogenesis in the submucosal layer of the colon when compared to controls with DSS. Skin targeting of Hyal1 led to less anaerobic bacteria in colonic fat tissue (Control:49.3 CFU/mg, K14/hyal1: 1.4 CFU/mg, P=0.04). An explanation for the capacity of the skin to influence colon adipogenesis and immune defense was seen with the observation that the K14/hyal1 mouse had more CLA positive lymphocytes in the colonic submucosa (Control:1.25/HPF, K14/Hyal1: 13/HPF, P=0.03). These results demonstrate two important processes; 1) adipogenesis is important for immune defense of both skin and colon, and 2) targeting the ECM of the skin can have significant influence on lymphocyte trafficking to the colon. Such an observation invites speculation that topical therapies on the skin could be used to influence inflammatory bowel disease or inflammation at other organs.

Adipocytes protect against bacterial infection of skin by producing cathelicidin antimicrobial peptide (AMP) during adipogenesis (Zhang et al. Science (2015)). Inflammatory bowel disease (IBD) has increased bacterial trans-location from the colon and is accompanied by local adipogenesis and hyaluronan (HA) breakdown. In this study, we hypothesized that colon is similar to skin it that the submucosal adipocytes may suppress bacterial trans-location, and further hypothesized that HA may influence this adipogenic response. We first tested this in vitro and observed that 3T3L1 fibroblasts increased high molecular weight HA (HMW-HA), and increased mRNA for Hyaluronan Synthesis 2 (HAS2) (p =0.006) when stimulated to differentiate into adipocytes. Conversely, digestion of HMW-HA by addition of hyaluronidase (Hyal) inhibited adipocyte differentiation seen by decreased mRNA for the markers of adipocyte differentiation zfp423(p=0.01) and adiponectin (p =0.004). Hyal also decreased accumulation of lipid droplets and expression of cathelicidin (Camp). Inhibition of early adipogenesis was not due to the accumulation of increased HA fragments since the addition of HA fragments of had no effect on differentiation. In vivo, colitis induced in mice by DSS resulted in local adipogenesis and increase Camp expression in submucosal fat. In contrast, over-expression of human-hyaluronidase 1 (hHYAL1) resulted in digestion of HA around the colon and inhibited adipogenesis and cathelicidin as seen by immunostaining, loss of fibroblasts expressing PDGFRα, lesser mRNA for zfp423(p=0.05) and PPARγ(p=0.003). These mice showed increased bacterial trans-location out of the colon as measured by culture from mesenteric fat (p=0.01) and bacterial DNA in the spleen detected by qPCR (p=0.001). These results show HA is important for local adipogenesis and plays an essential role in protection against both infection of skin and bacterial trans-location out of the colon.

Skin immune defense and repair is a coordinated response to a variety of danger signals and involves the complex interactions of several different cell types. Dermal adipogenesis is a previously unknown but critical event in this process that participates in antimicrobial defense and inflammation. We have previously shown that high-molecular-weight hyaluronan (HMW-HA) undergoes breakdown into small fragments after injury where it can then activate endothelial cells, macrophages and DCs via TLR4. In this study we investigated if HA may also influence adipogenesis and thus regulate the wound repair process. Maturation of 3T3L1 cells into adipocytes was accompanied by increased HMW-HA seen by Gel Electrophoresis and expression of HAS2 mRNA measured by qPCR (p =0.006). Conversely, digestion of HMW-HA by addition of hyaluronidase inhibited differentiation into mature adipocytes as seen by decreased mRNA for zfp521(p=0.02), zfp423(p=0.01) and Adiponectin (p =0.004), as well as decreased accumulation of lipid droplets and enhanced cell proliferation. The inhibition of early adipogenesis was not due to HA fragments since the addition of HA fragments of 6.4kDa, 50kDa, 150kDa and 1000kD had no effect. In vivo, tissue specific expression of hyaluronidase 1 (HYAL1) in mice prior to injury resulted in less mature dermal fat tissue as seen by immunohistochemistry and lower expression of mRNA for C/EBPα (p=0.002) but not Adiponeqtin(p=0.13) or PPARγ(p=0.27). In normal mice, HA immunostaining was lost 3 days following injury coincident with an increase in expression of HAS2 mRNA (p=0.0008) and mRNA for the hyaluronidase KIAA1199 (p=0.01). The loss of HMW-HA in vivo occurred after the expected increase in adipogenesis immediately after injury. Taken together, these results suggest that HA plays a regulatory role during wound repair by controlling dermal adipogenesis. HMW-HA facilitates fat maturation early in the wound repair process, and hyaluronidase action stops this event later in the wound healing process.