Alopecia areata (AA) is a highly prevalent autoimmune disease (AD) leading to hair loss in affected individuals, in which both genetic and environmental factors likely play a role. Associations between changes in the microbiome composition and many ADs have been reported, however, a characterization of the role of gut microbiota in the development of AA has not been undertaken. To investigate the functional relevance of gut microbiota to AA development in vivo, we depleted the gut microbiome in the C3H/HeJ mouse model of AA with broad-spectrum antibiotics. We found that antibiotic-treated mice were protected from AA induction. Additionally, using 16S rRNA gene sequencing (16SrRNA Seq) we found that AA-affected mice show an overrepresentation of a single species of Ligilactobacillus, Ligilactobacillus murinus (L. murinus), that preceded the development of the first clinical signs of AA. Further, using L. murinus specific primers, we showed that co-housed C3H/HeJ mice protected from AA show reduced levels of L. murinus in comparison to mice affected with AA, suggesting that L. murinus is associated with the development of AA. Examining the effects of the microbiome on T cell composition, we found that gut microbiota-depleted mice had a decreased percentage of effector CD8+ T cells in skin draining lymph nodes, which we previously showed to be required to drive the pathogenesis of AA. To establish the relevance of these observations to human AA, we performed 16SrRNA Seq sequencing and metagenomics and compared gut microbiome composition in stool samples between AA patients and healthy controls (HCs). We discovered a striking gut dysbiosis in AA patients compared to HCs and identified species of Lachnospiraceae and Bacteroidaceae to be overrepresented in AA. Taken together, these findings suggest that gut microbiome dysbiosis contributes to the pathogenesis of AA and offer new avenues of treatment for this disease.

Primary cicatricial alopecia (PCA), often referred as scarring alopecia, is a group of disorders that result in permanent hair loss, and for which no FDA approved treatments exist. Common forms of PCA include Lichen Planopilaris (LPP), Frontal Fibrosing Alopecia (FFA) and Central Centrifugal Cicatricial Alopecia (CCCA), with LPP predominating in middle aged white women, FFA in the postmenopausal group, and CCCA in women of African descent. FFA is considered to be a variant of LPP, and CCCA shares several pathological features with advanced LPP, however, whether PCAs are separate diseases or share common pathomechanisms remains an unresolved question. To determine if PCAs are molecularly distinct, we performed RNAseq of scalp biopsies from 28 LPP, 30 FFA, and 9 CCCA patients compared to 12 normal controls. This revealed a core set of dysregulated pathways shared among all PCA subtypes: 1) downregulated cholesterogenic genes (CYP51A1, NSDHL), 2) upregulated fibrosis and scarring genes (COL1A1, BMP10, CLDN5), and 3) a striking enrichment of mast cell genes (TPSAB1, MS4A2, CMA1). Notably, positive staining for mast cells (MC tryptase) was detected near the sebaceous gland of PCAs and not in controls. We also found that each subtype was associated with unique pathways, such as JAK/STAT signaling in CCCA, indicating potential molecular signatures within each subtype. To investigate the clinical significance of mast cells in PCAs, we treated 37 patients with the oral mast cell stabilizer drug, cromolyn. Five patients reported improvement within 6 weeks and 3 continued to report transient improvement of symptoms after 3 months. While clinical improvement was transient, optimization in formulation and delivery may improve durability, and many additional drugs to target mast cells are available for investigation. Our findings revealed common gene expression pathways among PCAs, and demonstrated targeting of mast cells as a novel treatment strategy.

Alopecia Areata (AA) is a highly prevalent autoimmune disease leading to hair loss in affected individuals. We and others have demonstrated a strong genetic component in the development of AA, however, emerging evidence suggests that environmental factors clearly contribute to AA pathogenesis. To determine the role of the microbiome as an environmental factor in AA, we performed 16S sequencing of skin, hair follicle (HF), and feces from 34 AA patients and 12 healthy controls (HCs). We found no significant differences in the composition of the skin or HF microbiome, but we discovered striking gut dysbiosis in AA patients compared with HCs. We detected a pronounced dysbiosis characterized by an increased relative abundance of members of the Firmicutes phylum and decreased relative abundance of the Bacteroides phylum in the gut microbiome of AA patients compared with HCs. Utilizing metagenomics sequencing, we identified perturbations in the relative abundance of different microorganisms of the Firmicutes and Bacteroides phyla, such as Ruminococcus sp., Alistipes shahii, and some lactobacillales in the gut microbiome of AA patients. To investigate the causal role of the gut microbiome in AA development, we conducted antibiotic-mediated depletion of the gut microbiome in C3H/HeJ mice, and found that treated mice were largely protected from hair loss, concomitant with decreased numbers of CD8+ T cells and an increase in the Treg/CD8+ ratio. Taken together with recent reports of reversal of AA in patients following fecal microbiota transplant (FMT), our findings suggest that restoring homeostasis of the gut microbiome may represent a new therapeutic approach for AA. We recently initiated a clinical trial of FMT for treatment of AA, in which we will monitor the normalization of gut microbiome dysbiosis relative to the restoration of hair growth.

The Janus kinase/signal transducers and activators of transcription (JAK/STAT) are key intracellular mediators in the signal transduction of many cytokines and growth factors. Common γ chain cytokines and interferon-γ that use the JAK/STAT pathway to induce biological responses have been implicated in the pathogenesis of alopecia areata (AA), a T cell–mediated autoimmune disease of the hair follicle. We previously showed that therapeutic targeting of JAK/STAT pathways using the first-generation JAK1/2 inhibitor, ruxolitinib, and the pan-JAK inhibitor, tofacitinib, was highly effective in the treatment of human AA, as well as prevention and reversal of AA in the C3H/HeJ mouse model. To better define the role of individual JAKs in the pathogenesis of AA, in this study, we tested and compared the efficacy of several next-generation JAK-selective inhibitors in the C3H/HeJ mouse model of AA, using both systemic and topical delivery. We found that JAK1-selective inhibitors as well as JAK3-selective inhibitors robustly induced hair regrowth and decreased AA-associated inflammation, whereas several JAK2-selective inhibitors failed to restore hair growth in treated C3H/HeJ mice with AA. Unlike JAK1, which is broadly expressed in many tissues, JAK3 expression is largely restricted to hematopoietic cells. Our study demonstrates inhibiting JAK3 signaling is sufficient to prevent and reverse disease in the preclinical model of AA.

Alopecia areata (AA) is an autoimmune disease characterized by the immune-mediated destruction of hair follicles in scalp skin, resulting in partial or total hair loss. Ongoing research has identified JAK/STAT and the T cell costimulatory signaling pathways as key druggable targets, and several clinical trials have been conducted. Although efficacy signals have been impressive, there are subsets of patients in each trial that do not respond to treatment. This variable response correlates with distinct molecular mechanisms of action (MoA) that define each tested JAK/STAT inhibitor. Using reverse-engineered regulatory networks (ARACNe networks) and machine learning algorithms, we defined molecular predictors of drug response for four compounds used in AA clinical trials: tofacitinib (pan-JAK inhibitor), ruxolitinib (JAK1/2 inhibitor), abatacept (CTLA4-Ig), and intralesional triamcinolone, through the transcriptional activity of ten candidate master regulators including RREB1, CAMKK2, DACH1, and HLF. The activity of these candidate regulators produced four non-overlapping (FDR<0.05) MoA gene signatures. In patients, non-responder status was defined as a lack of improvement in SALT score (delta<20%) at the end of a corresponding clinical trial in combination with lack of resolution of the ALADIN score. Patients were binned into nonresponder and nonresponder status based on these criteria for analysis. For each nonresponder patient, these signatures were percentile-ranked for overlap with each of the drug MoAs to predict the overall efficacy of each compound, producing a preliminary tool for assessing overall treatment efficacy prior to administering treatment.

Central centrifugal cicatricial alopecia (CCCA) is a primary lymphocytic scarring alopecia that presents as inflammation, pain and scarring hair loss at the scalp vertex primarily in women of African descent. It is estimated that between 6-8% of individuals of African descent have CCCA with the incidence rising in populations over 50 years of age, although CCCA is thought to be grossly underreported. Its onset has been associated with the use of hair straightening practices, suggesting mechanical factors in the hair shaft may predispose in part to CCCA. Current standard treatments are largely guided by expert opinion and include topical and injection corticosteroid therapies as well as oral tetracycline. To initiate a molecular analysis of CCCA, RNAseq analysis was performed on scalp biopsies of 5 CCCA patients compared to healthy control subjects. We identified three major pathways associated with CCCA: 1) downregulation of steroid/cholesterol/fatty acid pathways (ACAA1 and FASN); 2) upregulation of fibrosis and hypertrophic scarring pathways (COL1A1, BMP10, and CLDN5); and 3) upregulation of a mast cell signature (TPSAB1, MS4A2, and CMA1). Consistent with the molecular analysis, the histology of CCCA lesions displayed infiltration of mast cells, especially around the fibrotic tissue surrounding the remaining hair follicles. Current treatments involving immunomodulatory modalities are not clinically efficacious, consistent with the lack of immune and inflammatory pathway dysregulation in our analyses. Since the same three core pathways are shared among all cicatricial alopecias, rational development of therapies focusing on restoring cholesterol and fatty acid metabolism, and downregulation of fibrosis and mast cells may hold promise for improved clinical outcomes. Leveraging this core set of pathways to inform new therapies, these findings provide a paradigm in which to develop new treatments for CCCA as well as other primary cicatricial alopecias.

Lymphocytic primary cicatricial alopecias (PCA) have been considered different clinical entities with diverse treatment paradigms. Although Lichen Planopilaris (LPP), Frontal Fibrosing Alopecia (FFA) and Central Centrifugal Cicatricial Alopecia (CCCA) all have female predominance, affected patient groups of each subtype are distinct, with LPP predominating in middle aged Caucasian women, FFA in the postmenopausal group, and CCCA in women of African descent. Moreover, the anatomical distribution of each subtype is unique. LPP presents in a patchy distribution across the scalp, FFA in a band like frontal temporal distribution and CCCA at the vertex expanding centrifugally. Anticipating that these disease entities would be molecularly distinct, we performed a detailed RNAseq transcriptomic analysis of lesional and non-lesional scalp biopsies from 29 LPP, 18 FFA, and 5 CCCA patients compared with controls. Unexpectedly, our computational analysis revealed a core set of dysregulated pathways was shared among all three subtypes: 1) downregulated steroid/cholesterol/fatty acid biosynthesis genes (CYP51A1 and NSDHL), 2) upregulated fibrosis and hypertrophic scarring genes (COL1A1, BMP10, and CLDN5), and 3) striking enrichment of mast cell signatures (TPSAB1, MS4A2, and CMA1). The mast cell infiltrate was corroborated by positive staining of mast cells in lesional scalp. In addition, we found that each subtype displayed 1-3 unique dysregulated pathways, indicating that these disorders are far more pathophysiologically similar to one another than they are different. The core pathways form the basis of a biomarker known as SCALDIN (Scarring Alopecia Disease Activity Index) being developed to monitor response to treatments that could be used across all patients, such as LXR agonists, antifibrotic agents and/or mast cell inhibitors. Our study uncovers new targetable gene expression pathways that lay the groundwork for developing novel treatment strategies for PCAs.

Alopecia Areata (AA) is one of the most prevalent autoimmune disorders in humans leading to patchy or total loss hair with about a 2% lifetime prevalence. AA has a significant impact on patients’ quality of life, and associations with other autoimmune diseases. The development of AA is influenced by genetic, immunological, and environmental factors, though these are not completely defined. The gut microbiome has an immunomodulatory effect capable of eliciting pathologic immune responses beyond the gut. Our recent studies in the C3H/HeJ mouse model of AA showed that oral broad spectrum antibiotics prevented onset of AA, suggesting the gut microbiota is required for AA onset. Thus, to determine the microbiome composition of patients with AA, we collected skin swabs, hair follicle samples, and stool samples from a cohort of 26 AA patients. Analysis of 16S rRNA sequencing on stool samples revealed significant differential representation of bacterial taxa, between AA patients and healthy subject specifically, members of the firmicutes and bacteroides phyla, similar to our mouse model findings. When AA patients were compared to healthy controls we found under representation the bacteroides phyla and over representation of the firmicutes phyla in AA gut microbiome, similar to changes reported in other autoimmune disorders. Importantly, there was no difference in the skin or hair follicle microbiome in AA patients as compared to healthy controls, underscoring the importance of gut microbiota dysbiosis in AA patients. The presence of gut microbiota dysbiosis in human AA patients provides a rationale for the development of novel therapeutic strategies for AA patients, including Fecal Microbiota Transfer (FMT) and targeted microbial therapy, since restoring the gut microbiota composition to a healthy state has been suggested as an approach to improve the course of autoimmune diseases, such as AA.

Alopecia areata (AA) is a heritable autoimmune form of hair loss. The discordance of AA susceptibility between identical twins and spontaneous hair loss in the inbred mouse mode of AA suggests epigenetic modulation of AA risk. Through a reverse-engineered regulatory network analysis (DeMAND algorithm), we predicted that Vorinostat, an HDAC1/3 inhibitor FDA-approved for treatment of CTCL, would target over 50% of the known molecular pathways dysregulated in AA. While widely used as an anticancer drug, isoform-specific HDAC inhibitors have anti-inflammatory and immune-suppressive functions. To evaluate the efficacy of Vorinostat in reversing AA, daily topical application of Vorinostat (2% in DMSO) on C3H/HeJ mice with established AA markedly induced hair regrowth within 5 weeks. To investigate the effect of Vorinostat on activated immune cells observed in AA, we treated C3H/HeJ mouse cultured lymph node cells that were activated by phytohemagglutinin. Vorinostat inhibited the production of IFNg in vitro in a dose-dependent manner and decreased proliferation. To test the ability of other Class I HDAC inhibitors to reverse AA, we tested topical Entinostat (HDAC1/3 inhibitor, 2%) on AA affected mice, which was more effective than Vorinostat (2% and 5%), suggesting that other HDAC targets may be inhibited. To establish a profile of the chromatin landscape of T cells in mice with and without AA, we performed Assay for Transposase Accessible Chromatin sequencing (ATAC-seq) analysis of lymph node CD4 and CD8 T cells from AA-affected and non-alopecic mice. ATAC-seq revealed significantly different accessibility of transcription factor (TF) binding, sites such as RUNX1, is involved in T cell development. Computationally driven drug-repurposing strategies has aided in the discovery of new classes of epigenome-altering therapies, such as HDAC inhibitors, which could be a novel therapeutic approach for patients with AA.

Publisher's copyright statement: This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Lichen planopilaris (LPP) is a primary lymphocytic scarring alopecia whose incidence and prevalence are poorly defined. The literature reports LPP patients comprise <1%-8% of annual new referrals to tertiary hair centers, with a reported 2:1 to 4:1 female predominance. Clinically, LPP is characterized by burning, inflammation and scarring hair loss in a patchy distribution over the scalp. Expert opinion currently guides accepted therapies, therefore most patients opting for treatment are given individually or a combination of topical or injection corticosteroids, antibiotics, and hydroxychloroquine. To delineate the molecular pathways underlying LPP, we performed RNAseq on lesional and non-lesional scalp biopsies of 29 LPP subjects compared to controls. We identified three major core pathways associated with LPP: 1) downregulation of steroid/cholesterol/fatty acid pathways (CYP51A1 and NSDHL); 2) upregulation of fibrosis and hypertrophic scarring genes (COL6A1, COL1A1, TIMP1 MMPs, and FGF7); and 3) upregulation of mast cell signatures (TPSAB1, MS4A2, and CMA1). Steroid, cholesterol biosynthesis and fatty acid degradation pathways were most significantly downregulated in LPP biopsies (P<10-8, P<10-6, and P<10-4). Computational deconvolution of immune infiltrates revealed significant enrichment of marker genes associated with mast cells, with surprisingly little evidence for T cell signatures, and histology was strongly correlated with enriched presence of mast cells. Unexpectedly, the same three core pathways are shared among the other subtypes of scarring alopecias. Other pathways, such as ABC transporter pathways involved in drug uptake and metabolism, were strongly associated with LPP (P<10-5), but not other scarring alopecias. These findings suggest that a set of therapeutic modalities aimed at these core pathways may be applicable across scarring alopecia patients, whereas unique pathways such as ABC transporters found in LPP may be targeted as additional individualized therapies to maximize clinical efficacy.

Dysregulation of the Janus kinase/signal transducers and activators of transcription (JAK/STAT) pathway has pathological implications in autoimmune diseases. We found that therapeutic targeting of JAK/STAT pathways using first-generation JAK1/2 inhibitor (ruxolitinib) or pan-JAK inhibitor (tofacitinib) were effective in the treatment of human AA and in the C3H/HeJ mouse model. To further understand the role of individual JAK in AA and to produce the desired anti-inflammatory effects without unnecessary inhibition of other JAK pathways, we used JAK-selective inhibitors (JAK1-selective inhibitor Itacitinib, JAK2-selective inhibitor CEP33779 and JAK3-selective inhibitor PF06651600) to treat C3H/HeJ mice with AA. We demonstrate that JAK1-selective inhibitors as well as JAK3-selective inhibitors are effective in reversal of alopecia in C3H/HeJ mice with AA. In contrast, JAK2-selective inhibitor failed to restore hair regrowth. JAK1-seletive inhibition and JAK3-seletive inhibition was associated with decreased skin immune cells infiltration, reduced pathogenic cell responses and analysis of RNAseq data of whole skin biopsies taken from the JAK inhibitors treated-mice revealed statistically independent (orthogonal at FDR<0.05) molecular response profiles reflecting the variable efficacy of each JAK inhibitor type. JAK1 and JAK3 inhibitors (but not JAK2 inhibitors or controls) exhibited normalized gene expression patterns similar to unaffected C3H/HeJ mice, indicating JAK1 and JAK3 inhibition suppressed the dermal inflammatory/ cytotoxic T cell signature for AA reversal. Our results contribute to a more refined understanding of JAK1 or JAK3 inhibition as a therapeutic target in AA treatment, and suggest that inhibition of JAK2 is not required for treatment efficacy in AA.

The gut microbiome has emerged as a potential immunomodulatory factor capable of eliciting pathologic immune responses beyond the gut. Here, we studied the composition of the gut microbiome and its role in alopecia areata (AA) development in the C3H/HeJ mouse model. 8-wk old female C3H/HeJ mice were grafted with skin from AA mice and treated with a broad-spectrum antibiotic cocktail (ampicillin, metronidazole, neomycin and vancomycin), and followed for development of AA. We also tested antibiotic pretreatment in young animals (3wk old) prior to grafting. Strikingly, antibiotic treatment resulted in significant protection from hair loss. To determine the characteristics of the microbiome associated with the development of AA in the C3H mouse model, DNA was extracted from skin swabs and fecal samples and the 16S-rRNA was sequenced. Microbial diversity in the gut was significantly reduced in antibiotic treated mice, but no changes in the skin microbiome were observed. Principal component analysis of the gut microbiome showed distinct pretreatment, untreated and treated clusters, with dysbiosis characterized by over representation of Lactobacillus and under representation of Bacteroides in AA animals, similar to our findings in human AA. In C3H mice that spontaneously developed AA, the levels of butyric acid were significantly reduced when compared with control mice. Flow cytometry analysis of gut- and skin-draining lymph nodes showed a population of pathogenic T cells in the gut and a decrease in these cells infiltrating the skin in antibiotic treated animals. AMNV treated mice showed an increase in the Treg/CD8+ ratio and a decrease in CD4+ T cells in the spleen, as well as reduced numbers of CD4 and CD8 cells in the Peyer's patches. Together, these data indicate a potential role of the gut microbiome in T cell priming in AA and offer new therapeutic opportunities in AA focused on restoring gut microbial homeostasis.

Each year, more than one million patients are hospitalized in the U.S. for significant skin loss due to thermal and pressure injuries, chronic diabetic ulcers or genetic skin diseases. The ability to generate engineered human skin constructs (HSCs) has provided a promising therapy for these patients. However, it remained an unsolved challenge to incorporate hair follicles (HFs) into HSCs in order to improve wound healing and skin function. We initially employed a 3D-printing strategy to induce human HF formation within HSCs through guiding physiological 3D organization of cells in the HF microenvironment. Grafting our HF-bearing HSCs onto immunodeficient mice after their vascularization in vitro resulted in efficient growth of human HFs. However, when HSCs were maintained in vitro instead of grafting, the efficiency of HF induction was as low as 19%, due in part to incomplete restoration of DPC hair inductive gene signature achieved solely through 3D spheroid formation. To further enhance hair inductivity, we leveraged two master regulator (MR) genes, Lef1 and Fli1, which we previously identified as the key regulators of the intact DPC gene signature, as well as extrinsic factors targeting Wnt-signaling. Using RNA-sequencing, we confirmed that Lef1 overexpression and 3D-spheroid culture synergistically restored the intact DPC gene signature. Lef1 overexpression in DPCs markedly increased the efficiency of HF induction from 19% to 70%. Treatment of HSCs with recombinant Wnt10b and Wnt-activator small molecule, CHIR99021, also increased the efficiency to 50%, albeit lower than that was achieved with Lef1 overexpression. Overall, this data suggested that overexpressing Lef1 in DPCs or extrinsically targeting Wnt-signaling in HSCs can be an effective strategy to promote the generation of engineered human HFs from cultured cells.

Frontal Fibrosing Alopecia (FFA) is a lymphocytic cicatricial alopecia that predominantly occurs in post-menopausal women. It is characterized by a band-like loss of hair and follicular ostia of the frontal temporal scalp, and commonly involves bilateral eyebrow loss. Over the last 15 years the incidence of FFA has shown a noticeable, albeit unexplained rise. Current accepted treatment is based on expert opinion and includes steroids, hormone blockers, and hydroxychloroquine. To delineate the transcriptional landscape of FFA, we performed RNAseq on scalp biopsies of 18 FFA patients compared with controls. We found three major pathways associated with FFA: 1) downregulation of steroid/cholesterol/fatty acid pathways (NSDHL and FADS); 2) upregulation of fibrosis and hypertrophic scarring pathways (COL1A1, COL1A1, TIMP1 and MMPs); and 3) upregulation of mast cell genes. Histology of FFA lesions showed increased presence of mast cells. Similar downregulation of cholesterol synthesis pathways was found in non-lesional FFA scalp, suggesting that molecular changes can be detected even before obvious hair loss. FFA lesional scalp also showed upregulation of T cell activation and antigen presentation genes such JAK3, TAP2, and ICOS, which were not observed in other forms of cicatricial alopecia. To our surprise, we found that cicatricial alopecias share a core set dysregulated gene expression pathways consisting of downregulation of steroid and cholesterol metabolism and upregulation of fibrotic and mast cell signatures. Additionally, FFA displays immune response pathway dysregulation, suggesting that a set of drugs may be useful in targeting these pathways in all cicatricial alopecia patients, potentially combined with immunomodulatory drugs, as indicated by the unique pathways in FFA. These findings provide a molecular framework in which to further investigate the pathomechanisms of FFA in the broader context of cicatricial alopecias.

Immunotherapies are emerging as promising avenues for cancer treatment. The treatment paradigm of immune checkpoint inhibition specifically restores a host's ability to recognize and destroy tumors. However, despite their overall promise, a majority of patients do not benefit from treatments. Patients with immunologically "cold" tumors bear mutations rendering immunotherapies ineffective. We previously derived an immune infiltrate recruitment transcriptional signature from the hair follicle in alopecia areata (AA), that was controlled by a single network-inferred master regulator (MR), IKZF1. We postulated that cold tumors have gone cold by inactivating an immune infiltrate MR like IKZF1 could be converted into an immunologically "hot" susceptible state by reactivating it. We screened TCGA cancer cohorts for patient subsets that had perturbations of IKZF1 in their genetic/genomic networks. This analysis identified 25-40% of patients across six cancer types(melanoma, lung, thyroid, head-and-neck, prostate, and bladder) that would be amenable immune enhancement. Overexpression of IKZF1 in representative cell lines resulted in activation of immune infiltrate recruitment signatures and increased immune-mediated cytotoxicity. These results translated into a sygenic, immunocompetent mouse model where we showed that cutaneous melanomas were suppressed when IKZF1 was expressed in the tumors on the same order as anti-PD1 treatment. Furthermore, combining IKZF1 with anti-PD1 + anti-CTLA4 treatment provided synergistic enhancement, and the triple combination ablated tumor growth completely. Finally, we investigated IKZF1 in independent melanoma cohorts, and found that IKZF1 disruption was strongly correlated with poor outcome (p<1e-10) and recurrence. Our findings promise of enhancing immunotherapies by identifying key immune recruitment networks, as well as to identify cancer cohorts and individual patients that would be amenable to immunotherapy enhancement through specific, therapeutically targetable master regulators.

Androgenetic alopecia (AGA) is a complex genetic trait that is characterized by regional hair follicle miniaturization in response to androgens. While female-pattern hair loss is characterized by a diffuse thinning of the scalp, male pattern can be induced upon elevation of testosterone levels. What confers regional susceptibility vs. refractivity on different regions of the scalp is unknown. Donor Dominance refers to the phenomenon by which hair follicles retain the characteristics of the donor site when transplanted to a recipient site. This property forms the basis for the success of hair transplantation. Since the calvarium begins to develop shortly prior to hair follicle induction, this suggests that the craniofacial dermis epigenetically/differentially influences hair follicle patterning and development. We noticed that the hair pattern in AGA overlaps precisely with the demarcations of scalp dermis/underlying bones, which have a dual origin (neuroectoderm for parietal bone), vs. mesoderm (for occipital bone). Computational analyses of RNA seq from parietal and occipital scalp of matching control and AGA affected volunteers revealed a striking differential gene expression profile along the cranial-caudal axis defining two distinct biosignatures that reflect: 1) the developmental origins of the skin and 2) the susceptibility to develop AGA. Functional annotation of the differentially expressed genes shows enriched pathways in AGA samples, including genes implicated in cartilage-ECM interaction (ADAMTS4), in immunity (CD300c, FCGR1A), and epigenetic factors. Using the ARACNe algorithm, we identified transcription factors or master regulators (MRs) that govern the molecular mechanisms of AGA. This list of MRs was used to perform functional studies. Altogether, we present novel insights into the genetic, epigenetic, and developmental factors required for temporal specification of the skin and the interdependence of hair follicle, skull and craniofacial development.

Alopecia areata (AA) is a common autoimmune disease with a lifetime risk of ∼2%. In AA, the immune system targets the hair follicle, resulting in clinical hair loss. The prognosis of AA is unpredictable, and currently there is no definitive treatment. Our previous whole genome expression studies identified active immune circuits in AA lesions, including common γ-chain cytokine and IFN pathways. Because these pathways are mediated through JAK kinases, we prioritized clinical exploration of small molecule JAK inhibitors. In preclinical trials in mice, tofacitinib successfully prevented AA development and reversed established disease. In our tofacitinib trial in 12 patients with moderate to severe AA, 11 patients completed a full course of treatment with minimal adverse events. Following limited response to the initial dose (5 mg b.i.d.), the dose was escalated (10 mg b.i.d.) for nonresponding subjects. Eight of 12 patients demonstrated ≥50% hair regrowth, while three patients demonstrated <50% hair regrowth, as measured by Severity in Alopecia Tool scoring. One patient demonstrated no regrowth. Gene expression profiles and Alopecia Areata Disease Activity Index scores correlated with clinical response. Our open-label studies of ruxolitinib and tofacitinib have shown dramatic clinical responses in moderate to severe AA, providing strong rationale for larger clinical trials using JAK inhibitors in AA. ClinicalTrials.gov ID NCT02299297.

Reverse-engineered regulatory networks have recently demonstrated great promise in the analysis of complex diseases, such as cancer and have been used for drug prediction. We recently conducted regulatory modeling of an autoimmune form of hair loss disease, alopecia areata (AA), from gene expression analyses of AA scalp biopsies with the goal of predicting new drug targets. Vorinostat is an HDAC1/3 inhibitor that was predicted by computational mechanism of action analysis (DeMAND algorithm) to target over 50% of the known molecular pathology of AA. Vorinostat was originally FDA-approved for treatment of CTCL where it affected the expression of genes involved in differentiation of T cells. We treated C3H/HeJ mouse cultured lymph node cells that were activated by a mitogen (phytohemagglutinin) with Vorinostat, to investigate its effect on activated immune cells as observed in AA. We found Vorinostat inhibited the production of IFNg in vitro in a dose-dependent manner. To evaluate the efficacy of Vorinostat in treating AA, daily topical application of Vorinostat (2% in DMSO) on C3H/HeJ mice with established AA markedly induced hair regrowth within 5 weeks. Following our previous findings with JAK inhibitors as a potent promoter of anagen initiation in wildtype mice, we tested whether HDAC inhibition can also promote hair regrowth in wildtype mice. We found that topical treatment of C56BL/6 mice at P60 (telogen phase) with 2% Vorinostat or Entinostat (another HDAC1/3 inhibitor) promoted early entrance into anagen phase in normal mouse skin, similar to JAK inhibition. STAT proteins are substrates of HDACs and STAT3 is a known substrate of HDAC1/3. HDAC inhibitors may promote hair regrowth in AA mice via interfering with JAK-STAT signaling that governs both hair cycle and T cell activity. Given that there are no FDA-approved drugs for AA, computationally driven drug-repurposing strategies based on gene regulatory networks offer a new approach in identifying new targeted therapies for patients with AA.