Methods of determining the clinical phenotype of patients suffering from hidradenitis suppurativa

Abstract

The present invention relates to methods of determining the clinical phenotype of patients suffering from hidradenitis suppurativa (HS), a chronic inflammatory skin disease characterized by recurrent nodules, abscesses, and sinus tracts in the intertriginous areas. The invention is based on the discovery of a set of biomarkers that are differentially expressed in the hair-follicles of HS patients. The invention provides methods of measuring the levels of these biomarkers in biological samples from HS patients, and using them to classify the patients into different phenotypic groups. The invention also provides methods of selecting or adjusting the treatment for HS patients based on their phenotypic group. The invention thus offers a novel and useful tool for the personalized diagnosis and management of HS.

Description

[0001]METHODS OF DETERMINING THE CLINICAL PHENOTYPE OF PATIENTS SUFFERING FROM HIDRADENITIS SUPPURATIVA FIELD OF THE INVENTION: The present invention is in the field of medicine, in particular immunology. BACKGROUND OF THE INVENTION: Hidradenitis suppurativa (HS) is a chronic, debilitating, inflammatory skin disorder with a prevalence of around 1% and a profound impact on patients’ quality of life1. Characteristic lesions such as inflammatory nodules, abscesses, and sinus tracts develop in the axillae,inguinal, and gluteal areas, typically during or after puberty. HS is a complex disease withcontributory genetic and epigenetic changes, and hormonal, mechanical, microbial and lifestylefactors such as obesity and smoking. Multiple immunological pathways are described,encompassing innate immune cells such as macrophages, immune soluble mediators and, inparticular the complement and the Th1 / Th17 pathways2. Despite the incapacitating nature of HS, the treatment landscape is limited, leading to inadequate management of the disease for many patients. This is mainly due to the poor characterization of the events causing HS. HS is recognized as an autoinflammatory keratinization disease because the pathomechanisms of HS are intimately associated with aberrantly activated keratinization and autoinflammation3.In unaffected skin from predilection sites of patients with HS, the infundibular outer root sheathshows hyperplasia and hyperkeratosis likely driven by an intrinsic dysregulation of the hairfollicle stem cell (HF-SC) compartment. The outer root sheath cells exhibit replication stressstimulating type I interferon responses4. Combined with factors such as obesity and smoking,this leads to follicle occlusion and stasis of keratin and bacteria within the dilating hair follicle.Subsequent cyst formation and rupture induce acute inflammation, characterized by a mixed immune infiltrate of neutrophils, macrophages, dendritic cells, and T and B cells and increased expression of a battery of proinflammatory cytokines including IL-1β, IL-17, and TNF-α5. Treatment targeting inflammation, such as adalimumab, achieved clinical response in only 40% to 60% of patients. Response rates among newer, repurposed, biological therapies in large clinical trials have so far failed to exceed this. A major barrier to the identification of treatment targets and successful clinical translation is our lack of understanding of the interplay between keratinization and inflammation. SUMMARY OF THE INVENTION: The present invention is defined by the claims. In particular, the present invention relates to methods of determining the clinical phenotype of patients suffering from hidradenitissuppurativa and selecting the most appropriate therapies.DETAILED DESCRIPTION OF THE INVENTION:The inventors took advantage of single-cell RNA sequencing (scRNA-Seq) to define thecellular composition of hair follicle cells, the primary cells involved in HS. The results providean unprecedented view of HS pathology, demonstrating HF-SCs differentiate into two cell typeswith very specific characteristics: one involved in keratinization, the second enriched ininterferon response genes. In a cohort of 49 patients, the inventors demonstrate that these twoHF cell types are associated with distinct clinical phenotypes. This could serve as a potentialmarker for determining personalized medicine for patients. The present invention relates to a method of determining the clinical phenotype of a patient suffering from hidradenitis suppurativa comprising the steps of i) providing a hair-follicle sample from the patient and ii) determining the expression level of least 3 genes GORS, GIFE and GTwherein -the gene GORS is selected from the group consisting of KRT16, KRT6B, KRT6A,NDUFA4L2, TM4SF1 and LYPD3,- the gene GIFE is selected from the group consisting of MOXD1, ALDH3A1, GPX2,AHNAK2, CDH13, LAMB3, S100A8, NBL1, CCDC3, and SPARC, -the gene GT is selected from the group consisting of CD3e, CD3g, CD3z and CD3z,and concluding that the patient belongs to: -a profile wherein inflammation is predominant (group 1) when high expression levelsof GIFE and GT are detected in the sample -a profile wherein keratinisation occurs without inflammation (group 2) when a highexpression level of GORSis detected -a profile wherein inflammation coexists with keratinisation when high levels of GORS,GIFE and GT are detected (group 3). As used herein the term “hidradenitis suppurativa” has its general meaning in the art and refers to a chronic skin disease characterized by clusters of abscesses and / or cysts that most commonly affects apocrine sweat gland bearing areas. Hidradenitis suppurativa is also called acne inversa or Verneuil’s disease. As used herein, the term "hair-follicle sample" refers to a biological sample obtained from a hair follicle, which is a part of the skin that produces and contains hair. A hair-follicle sample can be collected by plucking a hair with its root or by using a punch biopsy to obtain a small piece of skin tissue containing one or more hair follicles. Typically, the sample is prepared asdescribed in the EXAMPLE. The hair-follicle sample can be prepared for gene expression / In the present specification, the name of each of the genes of interest refers to the internationally recognised name of the corresponding gene, as found in internationally recognised gene sequences and protein sequences databases, in particular in the database from the HUGO Gene Nomenclature Committee, that is available notably at the following Internet address : http: / / www.gene.ucl.ac.uk / nomenclature / index.html. In the present specification, the name of each of the various biological markers of interest may also refer to the internationallyrecognised name of the corresponding gene, as found in the internationally recognised genesequences and protein sequences databases ENTRE ID, Genbank, TrEMBL or ENSEMBL. Through these internationally recognised sequence databases, the nucleic acid sequences corresponding to each of the gene of interest described herein may be retrieved by the one skilled in the art. In some embodiments, the expression level of one or more GORSgenes selected from the groupconsisting of KRT16, KRT6B, KRT6A, NDUFA4L2, TM4SF1 and LYPD3 is determined. Insome embodiments, the expression level of KRT16 is determined. In some embodiments, the expression level of one or more GIFEgenes selected from the group consisting of MOXD1, ALDH3A1, GPX2, AHNAK2, CDH13, LAMB3, S100A8, NBL1,CCDC3, and SPARC is determined. In some embodiments, the expression level of GPX2 isdetermined. In some embodiments, the method of the present invention comprises determining the expression level of KRT16, GPX2 and CD3e. Methods for determining the expression level of a gene product such as nucleic acid (e.g. RNA) are also well known in the art. Conventional methods typically involve polymerase chain reaction (PCR). For instance, U.S. Pat. Nos. 4,683,202, 4,683,195, 4,800,159, and 4,965,188 disclose conventional PCR techniques. PCR typically employs two oligonucleotide primers that bind to a selected target nucleic acid sequence. Primers useful in the present invention include oligonucleotides capable of acting as a point of initiation of nucleic acid synthesis within the target nucleic acid sequence. A primer can be purified from a restriction digest by conventionalmethods, or it can be produced synthetically. PCR involves use of a thermostable polymerase.The term “thermostable polymerase” refers to a polymerase enzyme that is heat stable, i.e., the enzyme catalyzes the formation of primer extension products complementary to a template anddoes not irreversibly denature when subjected to the elevated temperatures for the timenecessary to effect denaturation of double-stranded template nucleic acids. Thermostable polymerases have been isolated from Thermus fiavus, T. ruber, T. thermophilus, T. aquaticus, T. lacteus, T. rubens, Bacillus stearothermophilus, and Methanothermus fervidus. Nonetheless, polymerases that are not thermostable also can be employed in PCR assays provided the enzyme is replenished. Typically, the polymerase is a Taq polymerase (i.e. Thermus aquaticus polymerase). Quantitative PCR is typically carried out in a thermal cycler with the capacity to illuminate each sample with a beam of light of a specified wavelength and detect the fluorescence emitted by the excited fluorophore. The thermal cycler is also able to rapidly heat and chill samples, thereby taking advantage of the physicochemical properties of the nucleic acids and thermal polymerase. In order to detect and measure the amount of amplicon (i.e. amplified target nucleic acid sequence) in the sample, a measurable signal has to be generated, which is proportional to the amount of amplified product. All current detection systems use fluorescent technologies. Some of them are non-specific techniques, and consequently only allow the detection of one target at a time. Alternatively, specific detection chemistries candistinguish between non- specific amplification and target amplification. These specifictechniques can be used to multiplex the assay, i.e. detecting several different targets in the same assay. For example, SYBR® Green I probes, High Resolution Melting probes, TaqMan® probes, LNA® probes and Molecular Beacon probes can be suitable. TaqMan® probes are the most widely used type of probes. They were developed by Roche (Basel, Switzerland) and ABI (Foster City, USA) from an assay that originally used a radio-labelled probe (Holland et al. 1991), which consisted of a single-stranded probe sequence that was complementary to one ofthe strands of the amplicon. A fluorophore is attached to the 5’ end of the probe and a quencherto the 3’ end. The fluorophore is excited by the machine and passes its energy, via FRET (Fluorescence Resonance Energy Transfer) to the quencher. Traditionally, the FRET pair has been conjugated to FAM as the fluorophore and TAMRA as the quencher. In a well-designed probe, FAM does not fluoresce as it passes its energy onto TAMRA. As TAMRA fluorescence is detected at a different wavelength to FAM, the background level of FAM is low. The probe binds to the amplicon during each annealing step of the PCR. When the Taq polymerase extends from the primer which is bound to the amplicon, it displaces the 5’ end of the probe, which is then degraded by the 5’-3’ exonuclease activity of the Taq polymerase. Cleavage continuesuntil the remaining probe melts off the amplicon. This process releases the fluorophore andquencher into solution, spatially separating them (compared to when they were held together by the probe). This leads to an irreversible increase in fluorescence from the FAM and a decrease in the TAMRA. In some embodiments, the expression level is determined by RNA-seq. As used, the term "RNA-Seq" or "transcriptome sequencing" refers to sequencing performed on RNA (or cDNA) instead of DNA, where typically, the primary goal is to measure expression levels, detect fusion transcripts, alternative splicing, and other genomic alterations that can be better assessed from RNA. RNA-Seq typically includes whole transcriptome sequencing. As used herein, the term “whole transcriptome sequencing” refers to the use of high throughput sequencing technologies to sequence the entire transcriptome in order to get information about a sample's RNA content. Whole transcriptome sequencing can be done with a variety of platforms for example, the Genome Analyzer (Illumina, Inc., San Diego, Calif.) and theSOLiD™ Sequencing System (Life Technologies, Carlsbad, Calif.), However, any platformuseful for whole transcriptome sequencing may be used. Typically, the RNA is extracted, and ribosomal RNA may be deleted as described in U.S. Pub, No.2011 / 0111409. cDNA sequencing libraries may be prepared that are directional and single or paired-end using commerciallyavailable kits such as the ScriptSeq™ M mRNA-Seq Library Preparation Kit (EpicenterBiotechnologies, Madison, Wis.). The libraries may also be barcoded for multiplex sequencing using commercially available barcode primers such as the RNA-Seq Barcode Primers from Epicenter Biotechnologies (Madison, Wis.). PCR is then carried out to generate the second strand of cDNA to incorporate the barcodes and to amplify the libraries. After the libraries are quantified, the sequencing libraries may be sequenced. Nucleic acid sequencing technologies are suitable methods for expression analysis. The principle underlying these methods is that thenumber of times a (DNA sequence is detected in a sample is directly related to the relative RNAlevels corresponding to that sequence. These methods are sometimes referred to by the term Digital Gene Expression (DOE) to reflect the discrete numeric property of the resulting data. Early methods applying this principle were Serial Analysis of Gene Expression (SAGE) andMassively Parallel Signature Sequencing (MPSS). See, e.g., S. Brenner, et al., NatureBiotechnology 18(6):630-634 (2000). Typically RNA-seq uses Next Generation Sequencing or NGS. As used herein, the term "Next Generation Sequencing" (NGS) refers to a relatively new sequencing technique as compared to the traditional Sanger sequencing technique. For review, see Shendure et al., Nature Biotech., 26(10): 1135-45 (2008), which is hereby incorporated by reference into this disclosure. For purpose of this disclosure, NGS may include cyclic array sequencing, microelectrophoretic sequencing, sequencing by hybridization, among others. By way of example, in a typical NGS using cyclic-array methods, genomic DNA or cDNA libraryis first prepared, and common adaptors may then be ligated to the fragmented genomic DNAor cDNA. Different protocols may be used to generate jumping libraries of mate-paired tags with controllable distance distribution. An array of millions of spatially immobilized PCR colonies or "polonies" is generated with each polonies consisting of many copies of a single shotgun library fragment. Because the polonies are tethered to a planar array, a singlemicroliter- scale reagent volume can be applied to manipulate the array features in parallel, forexample, for primer hybridization or for enzymatic extension reactions. Imaging-based detection of fluorescent labels incorporated with each extension may be used to acquire sequencing data on all features in parallel. Successive iterations of enzymatic interrogation and imaging may also be used to build up a contiguous sequencing read for each array feature. Typically, the level of a gene may be expressed as absolute level or normalized level. For instance, levels are normalized by correcting the absolute level of a gene by comparing its expression to the expression of a gene that is not a relevant for determining the risk. Thisnormalization allows the comparison of the level in one sample, e.g., a subject sample, toanother sample, or between samples from different sources. In some embodiments, the method of the present invention further comprises comparing theexpression level with a predetermined reference value wherein detecting a difference betweenthe expression level and the predetermined reference value indicates whether the subject is oris not at risk of having cancer.As used herein, the term “high” refers to an expression level that is greater than normal, greaterthan a standard such as a predetermined reference value or a subgroup measure or that isrelatively greater than another subgroup measure. For example, a high expression level refersto an expression level that is greater than a normal expression level. A normal expression levelmay be determined according to any method available to one skilled in the art. High expressionlevel may also refer to a level that is equal to or greater than a predetermined reference value.Thus, in some embodiments, the expression level of the gene is compared to a predetermined reference value that is a relative to a number or value derived from population studies, including without limitation, subjects of the same or similar age range, subjects in the same or similar ethnic group, and subjects having the same severity of disease. Such predetermined reference values can be derived from statistical analyses and / or risk prediction data of populations obtained from mathematical algorithms and computed indices. In some embodiments, retrospective measurement of the expression level in properly banked historical subject samples may be used in establishing these predetermined reference values. Accordingly, in some embodiments, the predetermined reference value is a threshold value or a cut-off value. The threshold value has to be determined in order to obtain the optimal sensitivity and specificity according to the function of the test and the benefit / risk balance (clinical consequences of false positive and false negative). Typically, the optimal sensitivity and specificity (and so the threshold value) can be determined using a Receiver Operating Characteristic (ROC) curve based on experimental data. For example, after determining the expression level in a group of reference, one can use algorithmic analysis for the statistic treatment of the measured expression levels in samples to be tested, and thus obtain a classification standard having significance for sample classification. The full name of ROC curve is receiver operator characteristic curve, which is also known as receiver operation characteristic curve. It is mainly used for clinicalbiochemical diagnostic tests. ROC curve is a comprehensive indicator that reflects thecontinuous variables of true positive rate (sensitivity) and false positive rate (1-specificity). It reveals the relationship between sensitivity and specificity with the image composition method. A series of different cut-off values (thresholds or critical values, boundary values between normal and abnormal results of diagnostic test) are set as continuous variables to calculate a series of sensitivity and specificity values. Then sensitivity is used as the vertical coordinate and specificity is used as the horizontal coordinate to draw a curve. The higher the area under the curve (AUC), the higher the accuracy of diagnosis. On the ROC curve, the point closest to the far upper left of the coordinate diagram is a critical point having both high sensitivity and high specificity values. The AUC value of the ROC curve is between 1.0 and 0.5. When AUC>0.5, the diagnostic result gets better and better as AUC approaches 1. When AUC is between 0.5 and 0.7, the accuracy is low. When AUC is between 0.7 and 0.9, the accuracy is moderate. When AUC is higher than 0.9, the accuracy is quite high. This algorithmic method is preferably done with a computer. Existing software or systems in the art may be used for the drawing of the ROC curve, such as: MedCalc 9.2.0.1 medical statistical software, SPSS 9.0, ROCPOWER.SAS, DESIGNROC.FOR, MULTIREADER POWER.SAS, CREATE- ROC.SAS, GB STAT VI0.0 (Dynamic Microsystems, Inc. Silver Spring, Md., USA), etc.As used herein, the term “low” refers to an expression level that is less than normal, less than astandard such as a predetermined reference value or a subgroup measure that is relatively lessthan another subgroup level. For example, a low expression level means an expression levelthat is less than a normal expression level. A normal expression level may be determinedaccording to any method available to one skilled in the art. Low expression level may also meana level that is less than a predetermined reference value as recited above.According to the present invention, group 1 is characterized by high expression levels of genes GIFEand GTand low expression levels of gene GORS.According to the present invention, group 2 is characterized by high expression levels of genesGORSand low expression levels of gene GIFEand GT. According to the present invention, group 3 is characterized by high expression levels of GORS,GIFE and intermediate expression level of GT.In some embodiments, the method of the present invention comprises the steps of a) assessingat least 3 parameters that are the expression levels of GORS, GIFE and GT, b) implementing analgorithm on data comprising or consisting of the parameter assessed at step a) as to obtain an algorithm output, the implementing step being computer-implemented; and c) determining towhich group the patient belongs to from the algorithm output obtained at step b).As used herein, the term “algorithm” is any mathematical equation, algorithmic, analytical or programmed process, or statistical technique that takes one or more continuous parameters and calculates an output value, sometimes referred to as an “index” or “index value”. As used herein, the term “parameter” refers to any characteristic assessed when carrying out the method according to the invention. As used herein, the term “parameter value” refers to a value (a number for instance) associated to a parameter. In some embodiments, the algorithm implements one or more additional parameters. Typically, the additional parameters are selected from the group consisting of age, sex, BMI, physical activity, disease onset and smoking status. Non other limiting examples of algorithms include sums, ratios, and regression operators, such as coefficients or exponents, biomarker value transformations and normalizations (including, without limitation, those normalization schemes based on clinical parameters, such as gender, age, or ethnicity), rules and guidelines, statistical classification models, and neural networkstrained on historical populations. Non- limiting examples of algorithms thus include logisticregression, linear regression, random forests, classification and regression trees (C&RT), boosted trees, neural networks (NN), artificial neural networks (ANN), neuro fuzzy networks (NFN), network structures, perceptrons such as multi-layer perceptrons, multi-layer feed- forward networks, support vector machines (e.g., Kernel methods), multivariate adaptive regression splines (MARS), Levenberg-Marquardt algorithms, Gauss-Newton algorithms, mixtures of Gaussians, gradient descent algorithms, learning vector quantization (LVQ), and combinations thereof. Of particular use in combining parameters are linear and non-linear equations and statistical classification analyses to determine the relationship between levels of said parameters and the objective response to the preoperative adjuvant therapy. Of particular interest are structural and syntactic statistical classification algorithms, and methods of risk index construction, utilizing pattern recognition features, including established techniques such as cross-correlation, Principal Components Analysis (PCA), factor rotation, Logistic Regression (LogReg), Linear Discriminant Analysis (LDA), Eigengene Linear Discriminant Analysis (ELDA), Support Vector Machines (SVM), Random Forest (RF), Recursive Partitioning Tree (RPART), as well as other related decision tree classification techniques, Shrunken Centroids (SC), StepAIC, Kth-Nearest Neighbor, Boosting, Decision Trees, Neural Networks, Bayesian Networks, Support Vector Machines, and Hidden Markov Models, among others. Other techniques may be used in survival and time to event hazard analysis, includingCox, Weibull, Kaplan-Meier and Greenwood models well known to those of skill in the art.In some embodiments, the method of the present invention comprises the use of a machine learning algorithm. The machine learning algorithm may comprise a supervised learning algorithm. Examples of supervised learning algorithms may include Average One-DependenceEstimators (AODE), Artificial neural network (e.g., Backpropagation), Bayesian statistics (e.g.,Naive Bayes classifier, Bayesian network, Bayesian knowledge base), Case-based reasoning, Decision trees, Inductive logic programming, Gaussian process regression, Group method of data handling (GMDH), Learning Automata, Learning Vector Quantization, Minimum message length (decision trees, decision graphs, etc.), Lazy learning, Instance-based learning Nearest Neighbor Algorithm, Analogical modeling, Probably approximately correct learning (PAC) learning, Ripple down rules, a knowledge acquisition methodology, Symbolic machine learning algorithms, Subsymbolic machine learning algorithms, Support vector machines, Random Forests, Ensembles of classifiers, Bootstrap aggregating (bagging), and Boosting. Supervised learning may comprise ordinal classification such as regression analysis and Information fuzzy networks (IFN). Alternatively, supervised learning methods may comprise statistical classification, such as AODE, Linear classifiers (e.g., Fisher's linear discriminant, Logistic regression, Naive Bayes classifier, Perceptron, and Support vector machine), quadratic classifiers, k-nearest neighbor, Boosting, Decision trees (e.g., C4.5, Random forests), Bayesian networks, and Hidden Markov models. The machine learning algorithms may also comprise an unsupervised learning algorithm. Examples of unsupervised learning algorithms may include artificial neural network, Data clustering, Expectation-maximization algorithm, Self-organizing map, Radial basis function network, Vector Quantization, Generative topographic map, Information bottleneck method, and IBSEAD. Unsupervised learning may also comprise association rule learning algorithms such as Apriori algorithm, Eclat algorithm and FP-growth algorithm. Hierarchical clustering, such as Single-linkage clustering and Conceptual clustering, may also be used. Alternatively, unsupervised learning may comprise partitional clustering such as K-means algorithm and Fuzzy clustering. In some embodiments, the machine learning algorithms comprise a reinforcement learning algorithm Examples of reinforcement learning algorithms include, but are not limited to, temporal difference learning, Q-learning and Learning Automata. Alternatively, the machine learning algorithm may comprise Data Pre- processing.The present invention also provides a method for selecting a therapeutic regimen or determiningif a certain therapeutic regimen is more appropriate for a patient identified as belonging to the group 1, 2 or 3. The method of the present invention is particularly suited to determining which patients will be responsive or experience a positive treatment outcome to a treatment.For instance, a patient who belongs to group 1 is eligible and / or will achieve a response with atherapy that consists in administering to the patient a therapeutically effective amount of an anti-inflammatory drug. As used herein, the term “anti-inflammatory drug” has its generalmeaning in the art and refers to a medication that reduces inflammation.Some examples of anti-inflammatory drugs include antibiotics having anti-inflammatoryproperties such as:- Clindamycin: this is a lincosamide antibiotic that inhibits bacterial protein synthesis andalso modulates the production of inflammatory cytokines and chemokines, -Doxycycline: This is a tetracycline antibiotic that inhibits bacterial growth and also hasanti-inflammatory effects by reducing the expression of matrix metalloproteinases, interleukin-8, and tumor necrosis factor-alpha, -Rifampicin: This is a rifamycin antibiotic that inhibits bacterial RNA synthesis and alsohas immunomodulatory properties by affecting the activation and differentiation of T cells and macrophages, and -Dapsone: This is a sulfone antibiotic that inhibits bacterial folic acid synthesis and alsohas anti-inflammatory and immunosuppressive effects by inhibiting the generation of reactive oxygen species, the release of lysosomal enzymes, and the function of neutrophils and monocytes. Some examples of anti-inflammatory drugs include jak / stat inhibitors. As used herein, the term “jak / stat inhibitor” refers to a drug that inhibits the janus kinase / signal transducer and activator of transcription (jak / stat) pathway, which is involved in the regulation of immune andinflammatory responses, as well as cell proliferation and differentiation. Jak / stat inhibitors canmodulate the activity of cytokines such as interleukin-6 (IL-6), interleukin-23 (IL-23), andinterferons (IFNs). Jak / stat inhibitors include but are not limited to tofacitinib, baricitinib,ruxolitinib, Upadacitinib, fedratinib, abrocitinib and ritlecitinib. Some examples of anti-inflammatory drugs are anti-IL1beta drugs. As used herein, the term "anti-IL1beta drug" refers to a drug that inhibits the activity of interleukin-1 beta (IL-1 beta), a pro-inflammatory cytokine that plays a key role in inflammation. Anti-IL1beta drugs canblock the binding of IL-1 beta to its receptor, prevent the maturation and release of IL-1 betafrom cells, or neutralize the biological effects of IL-1 beta. Anti-IL1beta drugs include but are not limited to anakinra, canakinumab, gevokizumab, and rilonacept. Some examples of anti-inflammatory drugs are PDE4 inhibitors. As used herein, the term "PDE4 inhibitor" refers to a drug that inhibits the activity of phosphodiesterase 4 (PDE4), an enzyme that degrades cyclic adenosine monophosphate (cAMP), a second messenger thatregulates various cellular functions, including inflammation. PDE4 inhibitors can increase thelevels of cAMP and thereby suppress the production of pro-inflammatory cytokines such as IL-1 beta, IL-6, TNF-alpha, and IL-17. PDE4 inhibitors include but are not limited to apremilast,roflumilast, cilomilast, and crisaborole.Some examples of anti-inflammatory drugs are anti-TNFα drugs. As used herein, the term“anti-TNFα drug” is intended to encompass agents including proteins, antibodies, antibody fragments, fusion proteins (e.g., Ig fusion proteins or Fc fusion proteins), multivalent bindingproteins (e.g., DVD Ig), small molecule TNFα antagonists and similar naturally- ornormaturally-occurring molecules, and / or recombinant and / or engineered forms thereof, that,directly or indirectly, inhibit TNFα activity, such as by inhibiting interaction of TNFα with a cell surface receptor for TNFα, inhibiting TNFα protein production, inhibiting TNFα gene expression, inhibiting TNFα secretion from cells, inhibiting TNFα receptor signaling or any other means resulting in decreased TNFα activity in a subject. The term “anti-TNFα drug” preferably includes agents which interfere with TNFα activity. Examples of anti-TNFα drugs include, without limitation, infliximab (REMICADE™, Johnson and Johnson), human anti- TNF monoclonal antibody adalimumab (D2E7 / HUMIRA™, Abbott Laboratories), etanercept (ENBREL™, Amgen), certolizumab pegol (CIMZIA®, UCB, Inc.), golimumab (SIMPONI®; CNTO 148), CDP 571 (Celltech), CDP 870 (Celltech), as well as other compounds which inhibit TNFα activity, such that when administered to a subject in which TNFα activity is detrimental, the disorder (i.e. acute severe colitis) could be treated.Some examples of anti-inflammatory drugs include anti-Th17 drugs. As used herein, the term“anti-Th17 drug” refers to a drug that targets the interleukin-17 (IL-17) pathway, which isinvolved in the production of Th17 cells, a type of inflammatory T cell that contributes to inflammation. Anti-Th17 drugs include secukinumab, ixekizumab, brodalumab, and bimekizumab.For instance, a patient who belongs to group 2 is eligible and / or will achieve a response with atherapy that consists in administering to the patient a therapeutically effective amount of aretinoid. As used herein, the term “retinoid” is a compound that is derived from vitamin A orhas a similar chemical structure and biological activity. A retinoid would be suitable for treatingkeratinization that can cause dryness, scaling, thickening, and inflammation of the skin.Retinoids can modulate the proliferation and differentiation of keratinocytes, the cells that produce keratin, and reduce the formation of comedones, which are clogged pores caused by excess keratin and sebum. Retinoids can be classified into different generations based on theirmolecular structure and binding affinity to nuclear receptors called retinoic acid receptors(RARs) and retinoid X receptors (RXRs). The first generation includes sotretinoin, acitretin,and alitretinoin, which have a high affinity for both retinoic acid receptors (RARs) and retinoidX receptors (RXRs). The second generation includes topical retinoids, such as tretinoin,adapalene, tazarotene, and trifarotene, which preferentially bind to RARs and have more selective effects on keratinocytes and sebocytes. The third generation includes alitretinoin and bexarotene, which are selective for RXRs.For instance, a patient who belongs to group 3 is eligible and / or will achieve a response with atherapy that consists in administering to the patient a therapeutically effective amount of an aSTING inhibitor optionally in combination with an anti-inflammatory drug and a retinoid. Asused herein, the term “STING” has its general meaning in the art and refers to the adaptor protein STING (Stimulator of Interferon Genes), also known as TMEM 173, MPYS, MITA and ERIS, that has been identified as a central signalling molecule in the innate immune response to cytosolic nucleic acids (Ishikawa H and Barber G N, Nature, 2008: 455, 674-678; WO2013 / 1666000). In some embodiments, the inhibitor is a small molecule such as a small organic molecule, which typically has a molecular weight less than 5,000 kDa. Examples ofSTING inhibitors are described in WO2015185565 as well as in US9549944B2. In someembodiments, STING inhibitors are selected from the compounds described in Haag S.M. etal., 2018. Targeting STING with covalent small-molecule inhibitors. Nature 559:269-73. In particular, the STING inhibitor is N-(4-iodophenyl)-5-nitrofuran-2-carboxamide, also known as C-176. In some the STING inhibitor is N-(4-Ethylphenyl)-N’-1H-indol-3-yl-urea also known as H-51 that has the formula of: As used herein, the expression "therapeutically effective amount" refers to an amount effective, at dosages and for periods of time necessary, to achieve a desired therapeutic result.A therapeutically effective amount of drug may vary according to factors such as the diseasestate, age, sex, and weight of the individual, and the ability of drug to elicit a desired response in the individual. A therapeutically effective amount is also one in which any toxic ordetrimental effects of the antibody or antibody portion are outweighed by the therapeuticallybeneficial effects. The efficient dosages and dosage regimens for drug depend on the disease or condition to be treated and may be determined by the persons skilled in the art. A physician having ordinary skill in the art may readily determine and prescribe the effective amount of the pharmaceutical composition required. For example, the physician could start doses of drug employed in the pharmaceutical composition at levels lower than that required in order to achieve the desired therapeutic effect and gradually increase the dosage until the desired effect is achieved. In general, a suitable dose of a composition of the present invention will be that amount of the compound, which is the lowest dose effective to produce a therapeutic effectaccording to a particular dosage regimen. Such an effective dose will generally depend uponthe factors described above. For example, a therapeutically effective amount for therapeutic usemay be measured by its ability to stabilize the progression of disease. A therapeutically effective amount of a therapeutic compound may decrease tumour size, or otherwise ameliorate symptoms in a subject. One of ordinary skill in the art would be able to determine such amountsbased on such factors as the subject's size, the severity of the subject's symptoms, and theparticular composition or route of administration selected. An exemplary, non-limiting range for a therapeutically effective amount of drug is about 0.1-100 mg / kg, such as about 0.1-50 mg / kg, for example about 0.1-20 mg / kg, such as about 0.1-10 mg / kg, for instance about 0.5, about such as 0.3, about 1, about 3 mg / kg, about 5 mg / kg or about 8 mg / kg. An exemplary, non-limiting range for a therapeutically effective amount of an antibody of the present invention is 0.02-100 mg / kg, such as about 0.02-30 mg / kg, such as about 0.05-10 mg / kg or 0.1-3 mg / kg, for example about 0.5-2 mg / kg. The invention will be further illustrated by the following figures and examples. However, these examples and figures should not be interpreted in any way as limiting the scope of the present invention. FIGURES:Figure 1. ORS and IFE represent two branches of differentiation from HF-SCs, withdistinct transcriptomic activities.(A) tSNE plot showing 3,532 ORS and IFE cells from the atlas. Cells are colored based on acell type annotation.(B) Representation of differentially expression genes between IFE and ORS. X- and Y-axisrespectively corresponds to the proportion of cells within ORS and IFE with positive expressionof the gene. Dashed line corresponds to genes expressed in same proportion of cells in ORS and IFE. One dot, corresponding to one gene, is colored according to average log fold change(avg_logFC) between IFE and ORS.(C) GSEA plots made from average fold change of all genes, between IFE and ORS.(D) Diffusion map of merged HF-SCs, ORS and IFE cells from HD and HS patient samples,colored by cell type annotation. Right panel depicts the location of the three cell types on the atlas.(E) Diffusion map of HF-SCs, ORS and IFE colored based on sample types, showingoverlapping between HD and HS patient cells.(F) Diffusion map of HF-SCs, ORS and IFE colored based on pseudotime, resulting fromtrajectory inference using TInGa, starting in HF-SCs. EXAMPLE: Material & Methods: Ethic of human samples HS patients were recruited from FolHYDRA study and enrolled from the dermatology department of the Henri Mondor Teaching hospital between June 2022 and December 2023.The Clinical Investigation Center performed skin biopsies. This monocentric, prospective studywas conducted in accordance with the declaration of Helsinki and was approved by the appropriate ethics committee (CPP Noth-West IV: 2021-A02352-39, 13 / 01 / 2022). All patients gave written informed consent before study enrolment. For scRNA-Seq analysis, hair follicle (HFs) samples were from discarded plastic surgery specimens from two healthy individuals and five HS patients’ localization (not shown). The main inclusion criteria were 1) history of HS according to the judgement of the investigator and 2) aged 18 years or older. Human HF separation Human scalp samples were digested in 4 mg / ml Dispase (Stem Cell) overnight. Individual HFs were then pulled out one by one using forceps. To obtain single-cell suspension for scRNAseq, the isolated HFs were digested in 0.05% Trypsin-EDTA (Gibco) for 30 mins at room temperature, washed in DMEM + 10 % FBS, and filtered through a 100 μm cell strainer. Single-cell RNA sequencing:HF cells were sorted (BD InfluxTM Cell Sorter) using a nuclear labeling DRAQ7TM Far-RedFluorescent Live-Cell Impermeant DNA Dye (Abcam, ref: ab109202). Cells were collected in 20% FBS-RPMI solution. Around 25,000 cells were loaded into one channel of the Chromium system using the V3.1 single-cell reagent kit (10X Genomics) to generate single-cell GEMs. Following capture and lysis, cDNAs were synthesized, and then amplified by PCR for 12 cycles as per the manufacturer’s protocol (10X Genomics). The amplified cDNAs were used to generate Illumina sequencing libraries that were each sequenced on one flow cell Nextseq500 Illumina. Bioinformatic analysis BCL files were processed using 10x Genomics Cell Ranger 3. Reads were mapped on GRCh37 (hg19) transcriptome. Raw count matrices were processed up to the figures using RStudio, with Rmarkdown to generate fully traceable notebooks. To ensure version stability, a Singularity container containing R version 3.6.3 and all packages of interest was developed and used to compile notebooks. Individual datasets, associated with a single patient sample, were analyzed individually from its raw count matrix using Seurat V3 package . Cells having less than 6 log number of uniquemolecular identifiers (UMI) or less than 500 genes were filtered out. Doublet cells wereremoved using scDblFinder tool and scds in hybrid mode . Then, cells having more than 20 % of UMI related to mitochondrial genes or more than 50 % related to ribosomal genes were filtered out. UMI count matrix for remaining cells was normalized using LogNormalize method implemented in Seurat::NormalizeData function. Single cells were annotated for cell type using Seurat::AddModuleScore. According to previous publications with scRNA-Seq from skin, they encompass CD4+ T cells, CD8+ T cells, Langerhans cells, macrophages, B cells, cuticle, cortex, medulla, IRS (inner root sheath), proliferative cells, ORS, IFE, IFE (interfollicular epidermis), HF-SC (hair follicle stem cell), sebocytes and melanocytes. To smoothen eventual mis-annotation, single cell level annotation was grouped by cluster. Clustering was generated as follows. First, 3000 highly variable features (HVFs) were identified using Seurat::FindVariableFeatures. Then, we generated a PCA with 100 components, from those 3000 HVFs, using Seurat::RunPCA. Distances between cells were computed using top 20 components of the PCA, with Seurat::FindNeighbors. Finally, a clustering was generated by Seurat::FindClusters with resolution 2. Cluster annotation was defined as the most represented cell type by cluster. To build the main “atlas” dataset, melanocytes clusters were removed in individual datasets prior to combination, using base::merge function. For the populations-specific datasetsassociated with immune cells, HF-SCs (1) or ORS and IFE (2) clusters of interest were selectedin the atlas dataset, then merged. For the HFSCs + ORS + IFE dataset, datasets (1) and (2) weremerged. For all 5 combined datasets, the same procedure was applied. Genes expressed in less than 5 cells were removed. Merged count matrices were normalized using LogNormalize method and a 100 components PCA was made on 2000 HVFs. Sample-specific effect wasremoved using harmony::RunHarmony on the PCA. On the PCA, n dimensions were selectedsuch that they summarize 60 %, for the atlas, or 35 %, for populations-specific datasets, of the variability between cells. This number of dimensions and the harmonized PCA were used as input to generate both clustering and tSNE. Clustering resolution was manually adjusted such that no clusters were shared across populations visible on the tSNE. The tSNE was built using Seurat::RunTSNE function. Dataset Cell typeAtlas All exceptmelanocytes Immune cells Immune clustersHF-SCs (1) HF-SC clustersORS + IFE (2) ORS and IFE clustersHF-SCs + ORS +(1) + (2) IFEFor the dataset containing HF-SCs, ORS and IFE, diffusion map was used to generate aprojection representative of transcriptomic changes between cells, using implementation from destiny package. Trajectory inference was made using two distinct methods: slingshot andTInGa, using cells in HF-SC cluster as trajectory root, and the harmonized PCA as input.Slingshot output 4 lineages: two stopping in HF-SC population, one HF-SC-ORS and one HF- SC-IFE. TInGa is able to infer branching trajectory, thus was used for visualization purpose. Default parameters, except max_nodes set to 10, were used.Differential expression was conducted using Seurat::FindMarkers. Heatmaps were made using ComplexHeatmap package. Functional enrichment analyses were conducted using msigdbr package for the gene sets database, and clusterProfiler package for the analyses.Similarly, to validate results from ORS and IFE in our data, we reanalyzed data from Wu et al.(2022). Briefly, FASTQ files were downloaded at accession number OEP002321 on biosino.org portal. Read mapping, individual samples analysis from raw count matrices,construction of the combined ORS + IFE dataset and its analysis were conducted using theexact same code as for our data. To visualize the whole atlas dataset (all samples, all cell types) coherently with original publication, melanocytes were not removed. RNA extraction and Real time Quantitative PCR RNA extraction was performed according to the manufacturer’s protocol (RNeasy Micro Kit, QIAGEN Inc.). RNA was converted to cDNA with High-Capacity RNA-to-cDNA Kit (Applied Biosystems). Quantitative PCRs were performed using the Brilliant II SYBR GREEN QPCR Master Mix kit (Agilent Technologies) for the expression of CD3, MSX2 and KRT16. OAZ-1 mRNA was used as a control for sample normalization. Expression of the gene OAZ-1 was used as a reference and the relative levels of each gene were calculated using the 2ΔΔCT method. Clustering analysis of patients Clustering analyses consisted of using the self-organizing maps (SOM) algorithm developed by Kohonen. In a nutshell, SOM algorithms assign each individual a specific area on the map based on their characteristics, placing similar individuals in close proximity and distinct ones in remote locations, thus allowing to draw visual comparisons of unique or overlapping patient characteristics and disease subtypes. The SOMs were constructed by applying the approach developed within the Numero package framework for the R statistical platform) building the SOMs with statistical verification of the robustness of the contrasts observed by permutation tests and 2) determining suitable groupings based on the direct visualization of data patterns and key characteristics of the dataset. For illustrative purposes, Gabriel’s biplots were plotted to project the patients along the principal components axes from mixed principal component analysis according to their individual characteristics, colouring patients according to their final diagnosis or cluster and thus allowing for direct visual assessment of the discriminative ability of each subgrouping. For descriptive statistics, categorical variables are expressed as count (%), quantitative variables as means (±standard deviation [SD]) or medians (interquartile range [IQR]), as appropriate. Biological parameters were log-transformed due to the non-normality of theirdistribution. We compared groups from clustering by means of one-way ANOVA or Kruskal-Wallis tests for continuous variables and chi-square tests or Fisher’s exact tests for categorical variables. A p-value < 0.05 was considered significant. Descriptive statistics and between- clusters comparisons were realized using Stata software 15.1 (StataCorp, Tx, USA), and R 3.4.3 (R Foundation, Vienna, Austria; pca2d, Numero packages) for clustering analyses and visualizations. Results: Single cell RNA Seq has revealed immune cell infiltration in HF in HS patients We obtained discarded human skin samples from plastic surgery and isolated HFs from them by dispase digestion followed by individually pulling the HFs out of dermis following the protocol described by Wu and colleagues6. Samples were collected from 5 HS patients inperilesional lesions in the axillary area and 2 healthy donors (HD) in the same region. Weperformed single-cell transcriptomic profiling of hair follicle cells using 10x Chromium platform (see Materials and Methods) and obtained 12,111 cells passing quality control steps. Cells were annotated for cell types based on known hair differentiation markers previouslypublished with original scRNA-Seq from hair follicles and skin6–8 (data not shown), enrichedwith best markers from differential expression analysis. Overall, all cells were assigned to threemain families of cells: i) matrix (Mx) ii) non matrix (noMx) and iii) immune cells (data notshown). The family of immune cells expressed PTPRC (data not shown). We identified CD4+T cells and CD8+T cells based on CD3, CD4 or CD8A expression. B cells are enriched in CD79. Among innate immune cells which express AIF1, Langerhans cells express CD207 and macrophages TREM2 and MSR1 (Fig 1B). The second family is enriched for HF matrix markerMSX2, indicating that they represent the HF matrix lineage (data not shown). This Mx groupcan be further divided into five populations: medulla (ALDH1A3, BAMBI), hair cortex (KRT31), hair fiber cuticle (KRT32, KRT35), inner root sheath (IRS) (KRT71, KRT73) and proliferative cells, enriched in TOP2A and MCM5 genes (data not shown). The third family (noMx) lacks MSX2 and hair differentiation markers but is enriched for KRT14 (data not shown). It also regroups five populations. HF-SCs are highly enriched for multiple known markers, including KRT15, COL17A1, and CXCL14, and also lack all the major differentiationmarkers examined. IFE cells are KRT14 enriched group with reminiscent KRT15 expressionand is closely connected to the HF-SCs. ORS cells which were described by Wu et al.6 expressKRT16 and KRT6C. IFE are enriched in DIO2 and TCEAL2 whereas sebocytes express CLMP and PPARG (data not shown).The fine annotation of the three main cell families yielded in 15 transcriptionally distinctpopulations, which are shared between all seven samples, suggesting that they are robustly present across different samples (data not shown). As expected, HF cells from HS patients have clearly more immune cells than HF cells from HD, implicating elevated immune cell infiltration into HFs during HS disease (data not shown). To explore whether differences could be observed between HF cells from HD and HS patients, cells were clustered independently from their annotation (data not shown). We then compared the abundance of cells within each cluster between the different samples. Unsupervised clustering of abundances distinguished HD from HS patients and further identified two groups of patients (data not shown). The most striking differences were observed in non-matrix cells. The first group of HS patients is enriched in cluster 1 (ORS), while the second is enriched in clusters 0 (IFE) and cluster 2 (T cells). It is worth nothing that HF cells from HD are devoid of T cells. The results indicate that HF cells from HS patients exhibit differences in cell composition when compared to those from healthy donors, the most notable being the infiltration of immune cells and the type of non-matrix cells. The molecular characteristics of immune cells from patients with HS depicted activation compared to those from healthy donors. We further interrogated potential heterogeneity among immune cells between HS patients and HD. All immune populations were shared across all samples, suggesting similar cellular composition. However, samples differ from the abundance of immune cell types. We discriminated two groups of HS patients: one enriched in T cells (HS1, HS2 and HS4) and the second enriched in macrophages population (HS3 and HS5) (data not shown). To investigate the differences in each T cells or macrophages population between HD and all HS patients, we performed differential expression analysis. Results in macrophages showed an increase in MHC class I and MHC class II expression in HS patients compared to HD. Moreover, complement pathway is enriched in HS patients (data not shown). Because IL1β is suspected to be the dominant pathway in HS disease9, we investigated IL1β expression within macrophages. As expected, IL1β levels are elevated in HS patients, although there is some variability in its expression among them (data not shown). Additionally, the only FDA-approved treatment for HS is adalimumab, targeting TNF pathway10. Therefore, we focused on TNF expression and, similar to IL-1β, we observed an increase in its expression among HS patients, although with some heterogeneity (data not shown). Note that IL-6 expression was not observed in the macrophages of HS patients (data not shown). Overall, these results indicate a pro-inflammatory phenotype of macrophages located around the hair follicle in the perilesional areas of HS patients. We then focused on CD4+T cells. Regarding IFN-γ and IL-17 expressing CD4+T cells, a small number of these cells was observed in HS patients and even fewer in HD patients (data not shown). Interestingly, CD4+T cells have a cytotoxic phenotype with unusual high expression of granzyme A (GZMA) and KLRB1 (data not shown). The presence of cytotoxic CD4+T cells was validated by immunofluorescence on skin biopsies from HS patients with GZMA staining (data not shown). Since type I interferon and IL-2 signaling are required for the generation of cytotoxic CD4+T cells in influenza model11, our results emphasize the crucial role of interferon in HS pathogenesis. HF-SCs from HS patients exhibit molecular features of inflammation and activation We have previously shown a disturbance in the homeostasis of HF-SC in HS disease4. To drill down on the heterogeneity within the HF-SC cell types, we performed a clustering and investigated differentially expressed genes between clusters. We observed nine clusters among HF-SCs (data not shown). Unsupervised clustering based on the abundance of HF-SCs in each cluster revealed two groups of samples (data not shown). Strikingly, it correlated with the overall proportion of HF-SCs in each sample, with approximately 3% for HS-1, HS-2, and HD-1, compared to more than 10% for HS-3, HS-4, HS-5, and HD-2 (data not shown). The group characterized by very low proportion of HF-SC was enriched in subcluster 2 (S2) compared to the other one. Differential expression analysis revealed that S2 was characterized by the expression of angiopoietin like 7 (ANGPTL7) (data not shown). This molecule is expressed during the telogen phase of the HF cycle12, suggesting that the dichotomy observed in the seven samples is linked to the stage of the HF cycle, with telogen being associated with the first group and anagen with the second. Nevertheless, an enrichment in subcluster 0, which expresses the highest level of TGFβ2, was observed in HS patients (data not shown). Differential expression analysis within this S0 subpopulation identified fundamental transcriptomic changes between HD and HS patients (data not shown). S0 from HS patients have an inflammatory phenotype, as evidenced by an enrichment in HLA class I molecule, MIF and IFITM3 (data not shown). In addition, S0 was enriched for ribosome genes in HS patients suggesting that it represented HF-SCs poised for activation. In contrast, S0 from HD patients showed an enrichment in genes involved in DNA repair, such as DDIT4, and the apoptosis pathway (data not shown). Our results indicate HF- SCs from HS patients are highly enriched in inflammatory and activation markers and are more susceptible to genomic instability. These results highlight the significance of HF-SCs in the pathogenesis of HS.ORS and IFE exhibit distinct molecular characteristics and function.We previously identified two distinct groups of HS patients based on respective enrichment inORS (cluster 1) and IFE (cluster 0), the latter being enriched in T cells (data not shown). Wethus extracted ORS and IFE from the complete dataset (Fig 1A) and investigated differencesbetween both populations, in all HS and HD samples. Notably, both populations displayedfundamental transcriptomic changes (Fig 1B). To further identify signalling pathwaysunderlying differentially expressed genes, we used Gene Set Enrichment Analysis (GSEA). Noteworthy, GSEA revealed significant enrichment for genes involved in keratinization inORS, whereas IFE showed enrichment for interferon response genes (Fig 1C). To assess thatenrichment analysis reflects the physiological phenotype of ORS and IFE, rather than aspecificity of HS disease, we attributed expression score for both gene sets to all cells. Splitting cells by condition showed that keratinization is a feature of both HS and HD ORS, andconversely for the inflammatory profile of IFE (data not shown). To generalize our findings,we re-analyzed a previously published scRNA-Seq data from hair follicles6, using same methodology as for our data (data not shown). The fine annotation of the three main cell families yielded in 11 transcriptionally distinct populations, which are shared between all 6 samples. Compared to our data, this dataset lacks sebocytes, macrophages, B cells and CD8+ T cells, which may results from distinct biological processing. Enrichment analysis betweenORS and IFE indicate that they exhibit similar profiles with regard to keratinization andinflammation (data not shown). To decipher whether ORS and IFE represent successive ordistinct differentiation branches from HF-SCs, we merged both datasets and generate a diffusion map. Diffusion map is used to better conserve continuous transcriptomic changes incell populations on the representation, thus eventual differentiation branches. Figure 1D indeedshows that ORS and IFE form two branches from HF-SCs, suggesting distinct differentiationprocesses rather than successive progression from one to the other. The representation does not highlight differences between HD and HS patients, suggesting that global structure of the three populations is conserved in HS (Fig 1E). We further used trajectory inference starting from HF-SC population to evaluate their fate. Slingshot tool, which is better suited to infer linear branches, indeed identified two distinct differentiation branches, respectively HF-SCs to ORS,and HF-SCs to IFE (data not shown). TInGa tool recapitulates the two lineages in one branchingtrajectory (Fig 1F). Overall, results from scRNA-Seq of HFs suggest HF-SCs differentiate in 2cell types with very specific characteristics: i) ORS with keratinizing phenotype and ii) IFEwith an inflammatory phenotype.Specific genes for ORS are depicted in Table A.Specific genes for IFE are depicted in Table B.Table A: genes specific for ORS gene_name avg_logFC pct.1 pct.2 p_val_adj pct.1_cut pct.2_cutKRT16 3,892906 0,905 0,155 0 (0.905,0.995] (0.102,0.2]KRT6B 3,166639 0,921 0,275 0 (0.905,0.995] (0.2,0.297]KRT6A 2,86001 0,752 0,26 0 (0.726,0.815] (0.2,0.297]NDUFA4L2 1,721714 0,68 0,212 0 (0.637,0.726] (0.2,0.297]TM4SF1 1,597284 0,801 0,243 0 (0.726,0.815] (0.2,0.297]LYPD3 1,408274 0,732 0,289 0 (0.726,0.815] (0.2,0.297] Table B: genes specific for IFE gene_name avg_logFC pct.1 pct.2 p_val_adj pct.1_cut pct.2_cutMOXD1 0,815775 0,764 0,126 0 (0.738,0.825] (0.112,0.21]ALDH3A1 0,762003 0,607 0,087 0 (0.566,0.652] (0.013,0.112]GPX2 0,746862 0,633 0,112 0 (0.566,0.652] (0.013,0.112]AHNAK2 0,711334 0,749 0,184 0 (0.738,0.825] (0.112,0.21]CDH13 0,607103 0,741 0,158 0 (0.738,0.825] (0.112,0.21]LAMB3 0,602659 0,74 0,163 0 (0.738,0.825] (0.112,0.21]S100A8 0,577242 0,659 0,196 0 (0.652,0.738] (0.112,0.21]NBL1 0,480192 0,627 0,159 0 (0.566,0.652] (0.112,0.21]CCDC3 0,435511 0,601 0,182 0 (0.566,0.652] (0.112,0.21]SPARC 0,418014 0,607 0,18 0 (0.566,0.652] (0.112,0.21]IFE and ORS depicted inflammatory molecular signature in HS patientsWe then investigated differences in IFE and ORS, between HS and HD. Splitting therepresentation showed good overlapping between HS and HD, except that a subpopulation ofIFE was enriched in HS samples (20 % of IFE in HS patients versus 9 % in HD) (data notshown). This population is characterized by IL1R2 and WNT3 expression (data not shown). Their absence in HD patients was confirmed in an already published dataset (data not shown). Transcriptomic profile of this subpopulation indicates that these cells are inflammatory (IFI27, IL17 and CXCL14), responsive to IL-1 and highly proliferative (data not shown). Bycomparing proportions of cells between each sample, we observed enrichment of IFE in 3patients (HS1, HS2, HS4) while ORS population was enriched in 2 patients (HS3, HS5) (datanot shown). Noteworthy, all HS patients who have IFE enrichment exhibit T cell infiltrationaround the hair follicle and high proportion of IL1R2+cells (data not shown).Finally, we interrogated differences in the molecular characteristics of ORS and IFE betweenHD and HS patients. Our analysis revealed that ORS from HS patients exhibit an inflammatoryphenotype with overexpression of MIF, S100A7 and display an active differentiation process with overexpression of galectin-7 (LGALS7), a lectin that is almost exclusively expressed instratified epithelia 13 and ARF5, a small GTPase that is involved in splicing complex, ribosomesynthesis and function, intracellular material transport 14. In contrast, ORS from HD patientsshowed up-regulation of DUSP1 and DDIT4, which play a crucial role in cell cycle control (data not shown). For IFE, we observed an identical profile with an increase in the inflammatory profile and a type 1 IFN signature (IFI27 and IFITM3) and a decrease in genes involved in the cell cycle control and in genes involved in cell adhesion (CTGF and CLDN1)for HS patients (data not shown). These results suggest IFE and ORS from HS patients exhibitan inflammatory phenotype associated with less susceptibility to apoptosis. Distinct physiopathology processes observed in HS patients The heterogeneity observed in scRNA-Seq data from HS patients suggests the existence ofdistinct profiles: i) one characterized by an abundance of inflammatory IFE and T cells infiltrateii) the second characterized by an abundance of ORS. We wondered whether HS patients presented a distinct clinical phenotype according to the HF cellular composition. Therefore, weconducted a qPCR analysis of one specific gene for ORS, namely KRT16 and on gene specificfor IFE, namely GPX2, in fresh HF cells from an independent cohort of HS patients (Fol-HYDRA). Given the considerable clinical heterogeneity observed, we initiated our analysis by characterizing the patients according to their clinical characteristics. To do so, we used with the SOM methodology to perform the clustering analysis displaying polarized distributions of patient characteristics across the maps, as evidenced by the contrasting colorings from blue (lowest values) to red (highest values) (data not shown). From these results, 3 clusters of HSpatients with homogenous features were built. The clinical characteristics of said clusters arepresented in Table 1. The first group (n=14) is characterized by the presence of young, non- smoking women with a high BMI and an early onset of disease. 11 out of 14 had an LC1 phenotype and 12 had inflammatory nodules. The hair follicles were mainly enriched of T cells and IFE. The second group (n=16) is also characterized by the presence of women with LC1 phenotype, but they are older, smokers and have a normal BMI and a later onset of disease.Most of the nodules are non-inflammatory. In this group, the HF cells are enriched of ORS withno T cells infiltrate. The final group was predominantly male (11 out of 19), smokers, with a later age at diagnosis and little family history (1 out of 19). They differed from the others in the presence of patchy scarring, fistulas and continuous involvement. The hair follicle cells areenriched in ORS and IFE, but there is no T cell infiltrate. These results suggest the existence ofdistinct pathophysiological processes: i) one in which inflammation is predominant (group 1),ii) one in which keratinisation occurs without inflammation (group 2), and iii) one in which thetwo mechanisms are intertwined (group 3). Table 1: clinical characteristics of the identified clusters of patients:Characteristic N Overall2bis N=191Subgroup 1 Subgroup 3 p- N=491N=161N=141value2Clusterting featuresAge 49 29,0 [23,0;38,0 [29,0; 31,0 [23,0; 23,5 [20,3; <0,00 42,0] 45,0] 40,8] 25,8] 1femal gender 49 29 (59,2%) 8 (42,1 %) 12 (75,0 %) 9 (64,3 %) 0,13age at diagnosis of Verneuil48 24,0 [19,8;26,0 [20,3; 27,5 [22,0; 19,0 [17,3; 0,001 disease 29,3] 32,8] 36,0] 28,8]age at onset of symptoms 49 18,025,0 [17,5; 21,0 [17,0; 14,0 [14,0; <0,00 [15,0;25,0] 26,5] 23,8] 15,8] 1 family history of Verneuil38 13 (34,2%) 1 (6,3 %) 5 (55,6 %) 7 (53,8 %) 0,006diseasecurrent tobacco 35 24 (68,6%) 12 (75,0 %) 9 (81,8 %) 3 (37,5 %) 0,12BMI 49 26,0 [23,0;27,0 [25,0; 23,5 [21,0; 25,5 [24,3; 0,22 32,0] 32,5] 30,5] 31,3]LC1 phenotype 49 35 (71,4 %) 13 (68,4 %) 11 (68,8 %) 11 (78,6 %) 0,85LC2 phenotype 49 10 (20,4 %) 4 (21,1 %) 3 (18,8 %) 3 (21,4%) >0,99LC3 phenotype 49 10 (20,4 %) 6 (31,6 %) 2 (12,5 %) 2 (14,3%) 0,41pilonidal sinus 49 12 (24,5 %) 7 (36,8 %) 3 (18,8 %) 2 (14,3%) 0,29inflammatory nodule 49 28 (57,1 %) 15 (78,9 %) 1 (6,3 %) 12 (85,7 %) <0,001non-inflammatory nodule 49 28 (57,1 %) 12 (63,2 %) 8 (50,0 %) 8 (51,1 %) 0,74abscess 49 7 (14,3 %) 5 (26,3 %) 1 (6,3 %) 1 (7,1 %) 0,23open comedones 49 25 (51,0 %) 13 (68,4 %) 4 (25,0 %) 8 (51,1 %) 0,033bridge scar 49 13 (26,5 %) 10 (52,6 %) 0 (0,0 %) 3 (21,4 %) <0,001cookie cutter scar 49 21 (42,9 %) 9 (47,4 %) 5 (31,3 %) 7 (50,0 %) 0,51ice pick scar 49 19 (38,8 %) 10 (52,6 %) 3 (18,8 %) 6 (42,9 %) 0,11draining fistula 49 12 (24,5 %) 11 (57,9 %) 1 (6,3 %) 0 (0,0 %) <0,001non draining fistula 49 15 (30,6%) 12 (63,2 %) 1 (6,3 %) 2 ( 14,3 %) <0,001papules 49 13 (26,5 %) 6 (31,6 %) 5 (31,3 %) 2 ( 14,3 %) 0,51folliculitis 49 13 (26,5 %) 7 (36,8 %) 5 (31,3 %) 1 (7,1 %) 0,14continued infringement 49 27 (55,1 %) 15 (78,9 %) 4 (25,0 %) 8 (57,1 %) 0,006HS PGA score 49 2,00 [1,00;3,00 [2,00; 1,00 [0,00; 2,00 [2,00; <0,00 3,00] 3,00] 1,00] 3,00] 1dissecting cellulitis 49 1 (2,0 %) 1 (5,3 %) 0 (0,0 %) 0 (0,0 %) >0,99decalcifying folliculitis 49 2 (4,1 %) 2 ( 10,5 %) 0 (0,0 %) 0 (0,0 %) 0,32keloidal acne 49 1 (2,0 %) 0 (0,00 %) 0 (0,0 %) 1 (7,1 %) 0,29Illustrative FeaturesKRT16 49 376 [268;398 [321;416 [317; 464] 261 [91,4; 392] 0,027472] 612]GPX2 49 50,2 [32,5;66,2 [480; 30,5 [23,8; 61,1 [46,3; 0,015 68,9] 78,4] 43,8] 67,7]CD3e 49 2,36 [1,62;252 [1,60; 2,06 [1,04;4,24 [2,35; 7,0] 0,0165,1] 5,2] 2,50]exclusive LC1 phenotype 49 29 (59,2 %) 9 (47,4 %) 11 (68,8 %) 9 (64,3 %) 0,4exclusive LC2 phenotype 49 10 (20,4 %) 4 (21,1 %) 3 (18,8 %) 3 (21,4 %) >0,99exclusive LC3 phenotype 49 4 (8,2 %) 2 (10,5 %) 2 (12,5 %) 0 (0,0%) 0,54mixed phenotype 49 6 (12,2 %) 4 (21,1 %) 0 (0,0 %) 2 (14,3 %) 0,181Median [25%; 75%]; n(%)2Kruskal-Wallis rank sum test; Pearson's Chi-squared test; Fisher's exact test Discussion: Studies aimed at elucidating the primary events that trigger follicular hyperplasia in HS remain extremely difficult in the absence of appropriate skin or animal models. In contrast to published studies on the cells involved in HS lesions, our research focused on the cells that comprise the hair follicle, as we have previously demonstrated their involvement in the early stages of the disease. Using scRNA-Seq analysis of hair follicle cells, we provided several critical insights into the pathogenesis of HS. We establish that two populations derived from HF-SC play two major roles, one in keratinization process and the other associated with inflammation. We revealthree endotypes: i) young non-smoking women with a high BMI whose hair follicle cellsexhibited an enrichment in IFE and T cells; ii) smoking women with a normal BMI and a lateonset disease whose ORS are predominant in the hair follicle cells; iii) smoking men with fistulaand continuous involvement whose hair follicle cells exhibited an enrichment in ORS and IFEwithout a T cell infiltration. Our data suggest that both inflammatory and keratinization mechanisms are involved in follicular hyperplasia suggesting that they could serve as potential markers for determining personalized medicine for patients. REFERENCES: Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.

Claims

CLAIMS:

1. A method of determining the clinical phenotype of a patient suffering from hidradenitissuppurativa comprising the steps of i) providing a hair-follicle sample from the patient and ii) determining the expression level of least 3 genes GORS, GIFE and GT wherein- the gene GORS is selected from the group consisting of KRT16, KRT6B, KRT6A,NDUFA4L2, TM4SF1 and LYPD3,- the gene GIFE is selected from the group consisting of MOXD1, ALDH3A1, GPX2,AHNAK2, CDH13, LAMB3, S100A8, NBL1, CCDC3, and SPARC,- the gene GT is selected from the group consisting of CD3e, CD3g, CD3z and CD3z,and concluding that the patient belongs to:- a profile wherein inflammation is predominant (group 1) when high expression levelsof GIFE and GT are detected in the sample- a profile wherein keratinisation occurs without inflammation (group 2) when a highexpression level of GORSis detected- a profile wherein inflammation coexists with keratinisation when high levels of GORS,GIFEand GTare detected (group 3).

2. The method according to claim 1 wherein the expression level of one or more GORSgenes selected from the group consisting of KRT16, KRT6B, KRT6A, NDUFA4L2, TM4SF1 and LYPD3 is determined.

3. The method according to claim 2 wherein the expression level of KRT16 is determined.

4. The method according to any one of claims 1 to 3 wherein the expression level of oneor more GIFE genes selected from the group consisting of MOXD1, ALDH3A1, GPX2,AHNAK2, CDH13, LAMB3, S100A8, NBL1, CCDC3, and SPARC is determined.

5. The method according to claim 4 wherein expression level of GPX2 is determined.

6. The method according to any one of claims 1 to 5 that comprises determining theexpression level of KRT16, GPX2 and CD3e.

7. The method according to any one of claims 1 to 6 that comprises the steps of a) assessingat least 3 parameters that are the expression levels of GORS, GIFE and GT, b)implementing an algorithm on data comprising or consisting of the parameter assessed at step a) as to obtain an algorithm output, the implementing step being computer- implemented; and c) determining to which group the patient belongs to from the algorithm output obtained at step b).

8. The method according to claim 7 wherein the algorithm implements one or moreadditional parameters selected from the group consisting of age, sex, BMI, physicalactivity, disease onset and smoking status.

9. The method according to any one of claims 1 to 8 wherein when it is concluded that thea patient belongs to group 1, then the patient is eligible and / or will achieve a responsewith a therapy that consists in administering to the patient a therapeutically effective amount of an anti-inflammatory drug.

10. The method of claim 9 wherein the anti-inflammatory drug is selected from the groupconsisting of antibiotics, jak / stat inhibitors, anti-IL1beta drugs; PDE4 inhibitors, anti- TNFa drugs and anti-Th17 drugs.

11. The method according to any one of claims 1 to 8 wherein when it is concluded that thea patient belongs to group 2, then the patient is eligible and / or will achieve a response with a therapy that consists in administering to the patient a therapeutically effective amount of a retinoid.

12. The method according to any one of claims 1 to 8 wherein when it is concluded that thea patient belongs to group 3 then the patient is eligible and / or will achieve a responsewith a therapy that consists in administering to the patient a therapeutically effective amount of STING inhibitor optionally in combination with an anti-inflammatory drugand a retinoid.

Images