Ruxolitinib Cream Has Dual Efficacy on Pruritus and Inflammation in Experimental Dermatitis

The goal of this study was to elucidate the anti-pruritic and anti-inflammatory efficacy of ruxolitinib cream in experimentally-induced dermatitis. Atopic dermatitis (AD), the most common chronic relapsing inflammatory skin disease, significantly impairs patients’ quality of life, with pruritus being a common complaint. The sensation of itch results from the interplay between epidermal barrier dysfunction, upregulated immune signaling and the activation of the central nervous system. The Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathway plays a central role in pro-inflammatory cytokine signaling in AD. Ruxolitinib cream is a potent and selective JAK1/2 inhibitor currently undergoing clinical evaluation in adults with mild-to-moderate AD (NCT03745638, NCT03920852 and NCT03745651). The efficacy of ruxolitinib cream was tested in murine models of acute and chronic dermatitis and was also characterized in an ex vivo human skin dermatitis model. Ruxolitinib cream was highly effective at ameliorating disease symptoms in multiple murine dermatitis models through downregulation of T helper (Th)2-driven inflammation, resulting in reduced skin thickening and decreased itch. Pathway analysis of mouse ear tissue and human skin explants underscored the role for ruxolitinib in ameliorating inflammation and reducing itch via modulation of the JAK-STAT pathway. Together, the data offer a strong rationale for the use of ruxolitinib cream as a potent therapeutic agent for the clinical management of atopic dermatitis.


INTRODUCTION
Atopic dermatitis (AD) is the most common inflammatory skin disease, with intense pruritus being the major and most burdensome symptom (1,2). Intractable pruritus has a significant impact on patients' quality of life and constitutes one of the unmet medical needs (3,4).
The Janus kinase (JAK) signal transducer and activator of transcription (STAT) pathway, a classical signal transduction pathway for numerous cytokines and growth factors, has been shown to play an important role in the dysregulation of immune responses in AD (5). The JAK family is comprised of four types of cytoplasmic tyrosine kinases: JAK1, JAK2, JAK3, and TYK2. The STAT family contains seven members: STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6 (6). Multiple immune mediators present within the inflamed skin utilize class I/II cytokine receptors and are dependent on the JAK-STAT pathway for signal transduction (7)(8)(9)(10).
Eczematous lesions are characterized by T helper (Th)2 cells leading to eosinophil recruitment and immunoglobulin production via the secretion of distinct cytokines, including interleukins (IL) IL-4, IL-5 and IL-13 (11). Epithelial mediators, such as IL-33 and thymic stromal lymphopoietin (TSLP), also play an important role in the type 2 innate immune response. IL-33, constitutively produced by skin epithelial cells, binds to the ST2 receptor on Th2 and other innate immune cells, and utilizes JAK1/2 kinase activity for downstream signal transduction (12,13). In AD patients, IL-33 overexpression in the epidermis, infiltration of ST2-positive cells and elevated serum IL-33 levels have been reported (14,15). Transgenic mice with constitutive epidermalspecific IL-33 expression (IL-33tg) spontaneously develop a progressive, AD-like skin inflammation and pruritus (16). Moreover, the epithelial cell-derived cytokine TSLP also promotes Th2 cytokine-expressing cells (17). TSLP, signaling through JAK, acts as a dual mediator of inflammation and pruritus (18,19).
The knowledge of JAK-STAT pathway involvement in inflammatory skin diseases has led to the development of oral and topical JAK inhibitors (5,9,10,(20)(21)(22) (23,24). Novel topical selective JAK inhibitors represent a promising option in the treatment of AD (10,25,26), and a topical pan-JAK inhibitor was recently approved in Japan for the treatment of atopic dermatitis (27). The focus of our study was ruxolitinib cream, a potent, selective JAK1/2 inhibitor that demonstrated significant clinical benefit in a phase 2b trial in adults with AD (NCT03011892) (28,29) and is currently being evaluated for the treatment of mild-to-moderate AD (NCT03745638, NCT03920852 and NCT03745651).
The goal of the current study was to characterize the dual anti-inflammatory and anti-pruritic potential of ruxolitinib cream in mouse models of experimentally-induced dermal inflammation. In addition to murine models of AD, the dual efficacy of ruxolitinib cream on pruritus and inflammation was assessed ex vivo using human skin explants.

Animal Experiments
Animal studies were approved by the Institutional Animal Care and Use Committee (IACUC) and performed in Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited facilities.
Female BALB/c mice were purchased from the Jackson Laboratories (USA). IL-33 transgenic (IL-33tg) mice were produced by TransGenic Inc. (Japan). All animals were housed under specific pathogen-free conditions and reared in line with standardized methods at 22 ± 1°C on a 12-h light/dark cycle with free access to food and water.
Acute TSLP-Induced Dermatitis BALB/c mice were randomized to the following groups; 1) sham injected untreated, 2) vehicle cream b.i.d., 3) 1.5% w/w ruxolitinib cream b.i.d. or 4) 0.05% w/w clobetasol cream q.d. For groups 2-4, murine TSLP (Invitrogen, USA) in sterile saline (3 µg in 20 µl) was injected intradermally into the outer pinna of the right ear on days 0, 2, 4, and 7. Topical cream (20 mg) was applied to the right ear from day 0 to 9. Ear swelling was measured with a thickness gauge (Mitutoyo, Japan) at 24, 48, and 72 h post day 7 injection. At study termination, 6 mm ear punch biopsies were collected, weighed, and fixed for histopathology. RNA isolation was performed on the remaining ear skin.
In a separate TSLP-induced dermatitis study, the spontaneous activity of vehicle and 1.5% w/w ruxolitinib b.i.d. treated mice was quantified using continuous home cage video recording (Vium, USA). The Vium platform provides continuous realtime measurement of activity.

Chronic FITC-Induced Dermatitis
BALB/c mice were randomized to the following groups: vehicle cream b.i.d., 1.5% w/w ruxolitinib cream q.d., 1.5% w/w ruxolitinib cream b.i.d. and 0.05% w/w betamethasone cream q.d. and dosed on both ears with 20 mg cream per ear for the duration of the study.
On day 0, mice were sensitized on the shaved abdomen with 100 µl of 0.5% w/v fluorescein isothiocyanate (FITC) in acetonedibutyl phthalate 1:1 (v:v) solution (Sigma Aldrich, USA). Sensitization was repeated on days 1 and 2. On day 7 mice were challenged with 20 µl of FITC solution on the right ear. Ear re-challenge was repeated once per week for 4 additional weeks to induce chronic skin inflammation. Ear swelling was measured at 24 and 72 h post FITC challenge. At termination, ear biopsies were weighed and RNA was isolated. Right draining (auricular) lymph nodes were excised for immunophenotyping.

IL-33-Induced Spontaneous Dermatitis
IL-33 transgenic (IL-33tg) mice spontaneously develop itchy dermatitis symptoms after 8 weeks of age, with prominent skin lesions around the eyes and the base of the tail (16,30). For a prophylactic treatment study, 5-week-old IL-33tg mice were randomized to the following groups: vehicle cream, 1.5% w/w ruxolitinib cream, 0.05% w/w clobetasol cream and non-transgenic littermate control treated with vehicle cream. Mice were dosed twice daily for 7 weeks on the skin around the eyes and tail base.
In a therapeutic treatment paradigm, 12 to 14-week-old IL-33tg mice, with confirmed dermatitis, were randomized to; vehicle cream b.i.d., 1.5% w/w ruxolitinib cream b.i.d., 0.05% w/w betamethasone cream twice a week, and non-transgenic littermate control treated with vehicle cream b.i.d., and treated to 24 weeks of age.
Mice were recorded by video camera for 30 min to quantify excessive scratching and grooming behavior. Hind limb scratching frequency and fore limb grooming were calculated as a cumulative count in 5 min from the video recording.
At study end, skin from the right eyelid was formalin fixed for histology. Skin from the left eyelid was collected for RNA analysis.

Histological Analysis
Paraffin sections were stained with hematoxylin-eosin and scored by a pathologist blinded to the treatment groups. Histopathology of ear biopsies was graded based on the extent of inflammation, inflammatory cell infiltration, hyperplasticity, and presence of crusting, ulcers, and erosions. Skin samples from IL-33tg mice were graded based on the condition of the epidermis, keratinization, inflammation, and mast cell activation.

Ex Vivo Human Skin
Healthy human skin explant cultures from 4 unrelated donors were dermatomed to a thickness of 750 µm and sectioned into approximately 1 cm 2 pieces. The explants were mounted onto 0.6 cm 2 static Franz cells for topical drug dosing at 40 h and 16 h prior to stimulation. The skin samples were positioned with the epidermis facing up, with the donor compartment of the Franz cell clamped in place to prevent lateral migration of ruxolitinib cream around the tissue. Following tissue pre-treatment with ruxolitinib cream, the Th2 stimulation was performed once at time 0, as described (31). The treatment groups included; 1) unstimulated and untreated, 2) Th2 stimulation and untreated, 3) Th2 stimulation treated with vehicle cream and 4) Th2 stimulation treated with 1.5% w/w ruxolitinib cream. The explants were harvested at 6, 24, and 48 h post-stimulation.

Transcriptomic Pathway Analysis
Total mouse (50 ng) and total human (100 ng) RNA were processed using the nCounter autoimmune profiling codeset (770 genes) or the neuropathology profiling codeset (770 genes) (Nanostring, USA), according to the manufacturer's protocol. After an 18 h hybridization, the samples were run on an nCounter SPRINT Profiler (Nanostring, USA). Data was analyzed using nSolver 4.0 Advanced Analysis software (Nanostring, USA). P-values were adjusted using the Benjamini-Yekutieli false discovery rate method.

Statistical Analysis
Statistical analysis was performed using Graphpad Prism 7.04 (Graphpad Software Inc, USA). Differences between groups were assessed by Kruskal-Wallis with Dunn's post-hoc test for nonparametric data, or ANOVA with Dunnett's post-hoc test for parametric data sets.

Ruxolitinib Cream Ameliorates Chronic FITC-induced Dermatitis Via Modulation Inflammatory T-Cell Subsets
Ruxolitinib cream was not associated with drug-induced cachexia (p<0.001) (Figure 2A). FITC challenge resulted in chronic, non-resolving ear swelling in the vehicle group, while ruxolitinib cream dose-dependently ameliorated the disease. At study termination, the ruxolitinib cream b.i.d. and q.d. treated groups exhibited 1.4% and 5.5% ear swelling only, which was in contrast to 40% ear swelling in the vehicle group (p<0.0001) ( Figure 2B). Ruxolitinib cream also significantly (p<0.01) decreased ear biopsy weights ( Figure 2C).
FITC-induced dermatitis was associated with lymphocyte expansion in the auricular lymph nodes of the vehicle-treated mice, while ruxolitinib cream dose-dependently inhibited this response (p<0.0001) ( Figure 2D). The proportion of Th1 cells (p<0.05) ( Figure 2E) and Th2 cells (p<0.01) ( Figure 2F) was significantly reduced in the ruxolitinib cream treated animals.   No significant difference in the proportion of Th17 cells between groups was observed ( Figure 2G). Cells from the steroid-treated group were unable to be analyzed due to low cell count and viability.

Ruxolitinib Cream Abrogates Pruritus-Induced Behaviors and Reduces Dermatitis Pathology
To evaluate the effectiveness of ruxolitinib cream on an acute inflammatory lesion formation, treatment was initiated prior to appearance of overt dermatitis symptoms in the spontaneous, non-remitting, IL-33tg model. Ruxolitinib cream treatment was not associated with any adverse effects and significantly ameliorated body weight loss compared to vehicle cream (p<0.05). In contrast, clobetasol cream was poorly tolerated and caused rapid wasting symptoms including significant body weight loss that required early termination due to humane endpoints ( Figure 3A). Vehicle-treated IL-33tg mice exhibited abnormal scratching and grooming behavior that worsened over time. Prophylactic treatment with ruxolitinib cream normalized  scratching and grooming to baseline levels observed in healthy wild-type littermate controls (p<0.001) (Figures 3B, C). Prophylactic application of ruxolitinib cream abrogated all histological signs of skin inflammation, including lymphocytic cell infiltrates (Figures 3D, E, G), and reduced mast cell frequencies to normal levels (p<0.0001) ( Figures 3F, H). The improvement of skin condition was underscored by a significant decrease in cumulative histology score (p<0.0001) ( Figure 3G). The IL-33tg mouse model was also used to evaluate a therapeutic treatment of established, progressing, dermatitis. Vehicle cream treated IL-33tg mice exhibited time-dependent worsening of dermatitis symptoms (p<0.0001). In contrast, ruxolitinib cream application rapidly reduced dermatitis score after one week of treatment, and after 3 weeks of treatment the scoring was similar to wild-type healthy animals. Steroid cream prevented dermatitis symptoms from progressing, but could not reverse established disease ( Figure 4A). Similar to the dermatitis score, spontaneous scratching behavior in the vehicle treated animals worsened over time. Again, ruxolitinib cream administration normalized scratch counts (p<0.0001) from the second week of treatment onwards to levels observed in healthy wild-type mice. Steroid cream also ameliorated abnormal scratching behavior (p<0.01), but was numerically inferior to ruxolitinib cream ( Figure 4B). Spontaneous forelimb grooming behavior was increased in vehicle-treated IL-33tg mice compared to healthy wild-type littermates. Topical ruxolitinib cream administration prevented worsening of this pathological behavior and was statistically significant compared to vehicle treated animals (p<0.05). Notably, steroid treatment did not appear to ameliorate dermatitis-induced grooming behavior ( Figure 4C). Consistent with the in-life dermatitis and pruritus scoring, histological analysis revealed ruxolitinib cream  was highly effective at reducing skin inflammation (p<0.01) and was numerically superior to steroid cream. (Figures 4D-G). Ruxolitinib cream also significantly reduced skin tissue mast cell frequency (p<0.05) compared to vehicle treatment ( Figures 4G-H).

Ruxolitinib Cream Downregulated the Inflammatory Transcriptome in Th2-Stimulated Human Skin
Differential transcriptomic analysis of Th2-stimulated human skin explants showed statistically significant downregulation of Th2 associated transcripts and decreased JAK-STAT signaling markers in the ruxolitinib cream treatment group compared to vehicle. There was also a downregulation in interferon signaling, lymphocyte trafficking, and TNF family signaling. Inflammasome nuclear factor-kappa-B subunit 2 (NFKB2) was also downregulated in response to ruxolitinib ( Figures 6A, B).

DISCUSSION
The JAK-STAT pathway has been implicated as a key driver of several inflammatory skin diseases, including AD (20). Targeted pathway modulation via the development of JAK inhibitors has uncovered a novel therapeutic opportunity for disease treatment.
Ruxolitinib is a potent and selective JAK1/2 inhibitor. Oral ruxolitinib is currently approved for the treatment of patients with myeloproliferative neoplasms (32) and has shown efficacy in treating steroid-refractory graft-versus-host disease (GvHD) (33,34). A phase 3 study is on-going for the use of oral ruxolitinib in steroid-refractory GvHD (NCT03112603). Ruxolitinib cream is currently being evaluated in clinical trial evaluation for the treatment of mild-tomoderate AD (NCT03745638, NCT03920852 and NCT03745651). Additionally, ruxolitinib cream efficacy has been evaluated in lichen planus, an inflammatory skin condition marked by an itchy rash (NCT03697460).
Ruxolitinib cream was evaluated in multiple experimentally induced models of dermatitis. In one model, dermal inflammation was evoked by TSLP injections. Aberrant TSLP signaling through JAK1/2 has been associated with AD (35,36). Ruxolitinib cream ameliorated dermatitis symptoms by modulating the transcription of genes directly involved in TSLP signaling, such as IL-7 receptor (Il7r) and JAK-STATs (37)(38)(39). Levels of interferons are known to increase during the progression of AD from acute to chronic form (40,41), and elevated transcript levels of IFNg receptor 1 (Ifngr1) and interferon regulatory factor 9 (Irf9) were observed in the TSLP-challenged vehicle group. Activation of inflammasomes can lead to an exacerbation of AD (42,43).
The itch-scratch cycle of AD promotes cutaneous lesion formation and mechanical damage to the epithelial barrier, further enhancing skin inflammation. It is conceivable that by inhibiting TSLP downstream signaling, ruxolitinib cream breaks the itch-scratch cycle allowing for restoration of epithelial barrier, integrity, consequently lowering the expression of JAK-dependent interferon signaling and JAK-independent inflammasome pathway. Notably, ruxolitinib treatment normalized mouse sleep pattern in parallel to resolving skin dermatitis. Sleep disturbance due to pruritus is a debilitating AD symptom (44,45). Ruxolitinib cream  inhibited pruritus without adversely affecting normal voluntary activity, such as wheel running (46,47). Repeated FITC challenge resulted in unresolvable ear swelling, consistent with previously published observations (48). Increased infiltration of innate lymphoid cells, a hallmark of AD (49,50), was also observed. Administration of ruxolitinib cream dose-dependently reduced inflammatory swelling. Efficacy was associated with a significant reduction in total immune cell infiltrates and Th2 and Th1 lymphocytes within the draining auricular lymph node.
Elevated IL-33 expression is observed in AD lesions and blood (10,51). Following IL-33 binding to its cognate receptor, downstream signaling is dependent on the JAK1/2 heterodimer (12,13). The activation of IL-33/ST2 signaling and its interaction with primary sensory neurons is a critical component of AD pruritus (15). In mice, overexpression of skin-specific IL-33 leads to AD-like inflammation through activation of Th2 cells, innate lymphoid cells (ILC)2, and mast cells (16,30,52). We employed the IL-33tg murine model to test ruxolitinib cream efficacy using both prophylactic and therapeutic dosing regimens. Anti-pruritic efficacy of ruxolitinib was consistent with previous results observed with tofacitinib, a pan-JAK inhibitor that decreased scratching bouts in NC/Nga mice sensitized with dust mite allergen (53). Similarly, ruxolitinib cream provided clinically meaningful reduction in itch in the phase 2 clinical trial in patients with AD (29). In addition to ILC2s, IL-33 promotes mast cell activation, adhesion, migration and maturation (15,51). IL-33 also indirectly impacts the skin barrier integrity (54). Ruxolitinib cream significantly reduced mast cell frequency as compared to vehicle-only cream and ameliorated skin histopathology, while steroid treatment did not significantly improve cumulative histology score. Ruxolitinib treatment also downregulated Mmp19 expression, a mechanism which facilitates migration of lymphocytes into inflamed epidermis (55).
Analysis of Th2-stimulated human skin explants corroborated the murine models of dermatitis findings. Inflamed human skin treated with ruxolitinib cream exhibited patterns of differential transcriptomic expression similar to those observed in the mouse dermatitis models, including modulation of the JAK-STAT pathway, interferon signaling, and inflammasome. Consistency of transcriptomic data suggest the translational validity of the animal models.
Atopic dermatitis pathophysiology encompasses skin barrier dysfunction, acute and chronic inflammation, as well as pruritus. In multiple preclinical models of dermatitis, topically applied ruxolitinib cream significantly ameliorated both pathogenic itch, as evaluated in the IL-33tg mice, and dermal inflammation. Taken together, this data strengthens the scientific rationale for topical JAK-STAT pathway inhibition for the treatment of AD.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal studies were reviewed and approved by The Institutional Animal Care and Use Committee (IACUC), Incyte.

AUTHOR CONTRIBUTIONS
MS and BF contributed equally to performance of the experiments and manuscript drafting. AC performed the Vium data analysis and wrote sections of the manuscript. MP contributed to the development of the TSLP-induced dermatitis model. MS and PS contributed to conception and design of the studies. All authors contributed to the article and approved the submitted version.

FUNDING
This study was funded by Incyte Corporation (Wilmington, DE). The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.