Methods for detecting allergic skin reactions

Abstract

Provided herein are methods of detecting an allergic skin reaction in a subject in need thereof. Also provided herein are methods of distinguishing an allergic skin reaction from an irritant skin reaction in a subject in need thereof.

Description

METHODS FOR DETECTING ALLERGIC SKIN REACTIONSRELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 560,431, filed on March 1, 2024, and U.S. Provisional Application No. 63 / 554,790, filed on February 16, 2024. The entire teachings of the above applications are incorporated herein by reference.BACKGROUND

[0002] Allergic contact dermatitis is one of the leading causes for dermatologic clinic visits worldwide, affecting over 1 million people each year within the United States alone (Lim et al. 2017; Ravishankar et al. 2022). Contact allergy occurs when susceptible individuals are exposed to environmental chemicals that permeate the skin barrier and cause an itchy, eczematous skin reaction. Environmental sources of contact allergy commonly include plants, household products, cosmetics, and occupational chemical exposures — the latter of which can cause loss of worker productivity, absences, and even require a change of profession (Kalboussi et al. 2019). Recurrent episodes of contact allergy can significantly decrease patient quality of life, and avoidance of environmental sources requires clear identification of specific causative chemical allergens.

[0003] Accordingly, there is a need for effective and reliable methods for detecting allergic contact dermatitis and other allergic skin reactions.SUMMARY

[0004] Provided herein are methods of detecting an allergic skin reaction in a subject in need thereof. The methods of detecting generally comprise the steps of a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining an expression level for one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL- 17 A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2,MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof; and c) identifying the subject as having an allergic skin reaction based on the level (e.g., expression level) of the one or more biomarkers.

[0005] Also provided herein are methods of diagnosing an allergic skin reaction in a subject in need thereof. The methods of diagnosing generally comprise the steps of a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining an expression level for one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof; and c) identifying the subject as having an allergic skin reaction based on the level (e.g., expression level) of the one or more biomarkers.

[0006] Further provided herein are methods of distinguishing an allergic skin reaction from an irritant skin reaction in a subject in need thereof. The methods of distinguishing generally comprise the steps of a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining an expression level for one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1,IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, C ASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof; and c) identifying the subject as having an allergic skin reaction or an irritant skin reaction based on the level (e.g., expression level) of the one or more biomarkers.

[0007] Also provided herein are methods of preparing a skin sample that is useful for detecting an allergic skin reaction in a subject. The methods of preparing generally comprise the steps of a) applying a potential allergen to an area of skin of a subject; b) obtaining a skin sample from the area of skin to which the allergen has been applied; and c) combining the sample with one or more reagents for detecting one or more protein biomarkers selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, C ASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof.

[0008] Also provided herein are methods of identifying a substance as a skin allergen or a skin irritant for a subject. The methods of identifying generally comprise the steps of a providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a test substance; b) determining an expression level for one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1,CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof; and c) identifying the test substance as an allergen or an irritant based on the level (e.g., expression level) of the one or more biomarkers.

[0009] In some embodiments of the above methods, the one or more protein biomarkers are selected from CCL20, CXCL1, IL8, TSLP, ARTN, IL4, CXCL10, OPG, IL5, CXCL9, TNFSF14, TNFB, TRAIL, IL13, IFNG, CXCL11, and TNF.

[0010] In some embodiments of the above methods, the one or more protein biomarkers are selected from CCL20, CXCL1, IL8, TSLP, ARTN, IL4, CXCL10, OPG, IL5, CXCL9, TNFSF14, TNFB, IL13, IFNG, CXCL11, and TNF.

[0011] In some embodiments of the above methods, the one or more protein biomarkers are selected from IFNG, CXCL9, CXCL10, CXCL11, IL4, IL5, and IL13.

[0012] In some embodiments of the above methods, the one or more protein biomarkers are selected from CCL20 and CXCL1.

[0013] In some embodiments of the above methods, the one or more protein biomarkers are selected from CXCL10, IL4, OPG, CCL19, CCL20, TSLP, IL-1 alpha, and NRTN.

[0014] In some embodiments of the above methods, the one or more protein biomarkers are selected from CXCL10, IL4, OPG, CCL19, CCL20, TSLP, and IL-1 alpha.BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0016] The foregoing will be apparent from the following more particular description of example embodiments, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating embodiments.

[0017] FIGs. 1A-1H: SADBE as a model for allergic contact dermatitis. Contact dermatitis was induced in SADBE-sensitized healthy volunteers using patch applications ofSADBE as a model allergen and SLS as a model irritant. FIG. 1A is a schematic showing contact dermatitis induced in healthy volunteers using patch applications of SADBE as a model allergen and SLS as a model irritant. Patches were applied 2 days and 4 days prior to skin sampling by suction blister biopsy, and skin samples were also taken from acetone patches, as an allergen vehicle control, as well as nonlesional skin that received no patch. Skin reactions were photographed, and the severity of patch reactions was graded using the standard ICDRG clinical patch test scoring system. FIG. IB shows skin reaction photographs, with circled outlines in the location of nonobvious patch placement. Reaction photos, including those from acetone patches and repeated patient participation, are shown in FIG. 1G. ICDRG, International Contact Dermatitis Research Group; ID, identification; SADBE, squaric acid dibutyl ester; SLS, sodium lauryl sulfate. FIG. 1C shows scRNA-seq data from suction blister skin samples of 9 patch test patients and 12 healthy volunteers clustered into 4 broad categories consisting of keratinocytes, melanocytes, antigen presenting immune cells, and lymphocytes. FIG. ID shows cell composition analysis where 25 distinguishable cell clusters were identified that each contained unique patterns of marker genes. FIGs. IE and IF show comparisons of allergic to irritant lesion samples, several cell population changes were common to both forms of contact dermatitis, yet 4 of the 25 cell populations were enriched in allergic samples at greater than 2-fold relative to irritant samples. FIG. IE shows shared contact dermatitis cellular changes. FIG. IF shows allergic- enriched cell population. FIG. 1G shows patch test reaction photos. Photos of patient patch test reactions and sites nonlesional skin sampling. Conditions not tested are labeled null. FIG. 1H shows patient and sample details. Information on all included patient samples are provided, including patient demographics, skin sampling dates, and which samples were used for single cell RNA sequencing analysis and Olink proteomic analysis. Note that CB-pattern patient IDs represent healthy control suction blister samples, and P-pattern patient IDs represent patch test patients. NULL patch test scores represent conditions that were not tested in that patient on that date.

[0018] FIGs. 2A-2C show cells were grouped broadly into 9 categories for gene expression analysis, and each skin condition was compared to nonlesional skin to find all differentially expressed (DE) genes. DE genes from each cell type were grouped by k-means clustering, and then arranged into heatmaps based on specific patterns of expression across skin conditions: 1) nonlesional genes commonly downregulated in patch samples; 2) up- regulated genes common to irritant and allergy; 3) up-regulated allergy-specific genes. FIG.2A shows gene ontology enrichment analysis was performed for each group and the top biologic process term for each section was listed along with associated genes. FIG. 2B shows the top 5 enriched ontology terms for allergy-specific DE genes show common biologic processes occur within multiple cell types, including anti-viral response, IFNG response, and unfolded protein cell stress response. FIG. 2C shows single-cell gene expression plots show IFNG production by CD8+ T cells and NK cells, and IL4 production by CD4+ T cells are distinguishing features of day 2 allergy samples relative to irritant samples.

[0019] FIGs. 3A and 3B show cell-specific expression of ligand and receptor genes that were used to create a cell signaling network. Differentially expressed (DE) ligands and receptors, as well as non-DE ligands and receptors, were clustered by k means. All DE ligand / receptor data was plotted by heatmap, as well as non-DE ligand / receptor data that appeared to have convincing expression trends. FIG. 3A shows ligands. FIG. 3B shows receptors. FIG. 3C shows a representative plot for LYZ+ myeloid cells in Day 2 Allergy data relative to nonlesional skin, and is an example of how the nichenetr package was used to estimate top ligands and ligand-induced target genes. FIG. 3D shows expression data for predicted IL4 and IL13 target genes in LYZ+ myeloid cells, which shows greater expression of target genes in allergy samples relative to nonlesional samples. FIG. 3E shows a directed acyclic graph that was created to compare estimated cell signaling at day 2 of irritant versus allergic contact dermatitis, which was filtered to depict the portion of this network related to IL4 and IL13. FIG. 3F shows the full signaling network pertaining to all ligands that can be used to compare estimated cell signaling at day 2 of irritant versus allergic contact dermatitis. FIG. 3G shows expression data as pseudo-bulk bar plots for selected ligands within the estimated signaling network with strong allergy-associated patterns of expression. FIG. 3H shows expression data as pseudo-bulk bar plots for selected ligands within the estimated signaling network with strong irritant-associated patterns of expression. Ligand prediction with Nichenetr for Irritant versus nonlesional samples. Nichenetr was used to predict ligands that might account for the gene expression changes within all cell types in day-2 allergic samples and irritant samples relative to nonlesional samples. FIG. 31 shows data plotted for LYZ+ myeloid cells in day-2 allergy samples. FIG. 3J shows data plotted for Langerhans cells in day-2 allergy samples. FIG. 3K shows data plotted forZKZ+ myeloid cells in day- 2 irritant samples. FIG. 3L shows data plotted for Langerhans cells in day-2 irritant samples. Data not shown for other cell types and skin conditions are analyzed by nichenetr.

[0020] FIG. 4A shows data from Olink proteomic proximity extension analysis that was used to quantify levels of 92 inflammatory proteins within acellular supernatant from suction blister skin. FIG. 4B shows a volcano plot that depicts differential expression analysis of day 2 allergic versus day 2 irritant suction blister protein data. Suction blister samples from day-2 ACD and day-2 ICD were analyzed, and differential expressions were plotted on a volcano plot. PC analysis of suction blister proteins also separates most allergy, irritant, and healthy samples. FIG. 4C shows that principle component analysis of suction blister proteomic data produces a plot in which most samples separate into appropriate groups enriched for allergy, irritant, and healthy samples, respectively. FIG. 4D shows the principle component protein loadings along with day 2 differential expression status. FIG. 4E shows receiver operator curves for regularized logistic LASSO regression models trained on suction blister proteomic data. Increased detection power is achieved by models using greater numbers of protein biomarkers. Black dot in FIG. 4E corresponds to a classifier threshold yielding 93% sensitivity and 93% specificity. FIG. 4F shows protein quantification data for proteins used in logistic LASSO regression models (see FIG. 16 for model coefficients). Suction blister fluid proteins distinguish ACD from ICD and nonlesional control skin. Acellular supernatant from suction blister skin samples were analyzed by Olink’ s Target 96 inflammation panel to assess 92 inflammatory proteins. FIG. 4G shows a heatmap created from protein row- normalized data. FIG. 4H shows results of differential expression analysis comparing proteins in day-2 allergic and day-2 irritant contact dermatitis were relatively unchanged by exclusion of patch reactions with less than a 1+ score because results were highly correlated to those using all samples. ACD, allergic contact dermatitis; ICD, irritant contact dermatitis.

[0021] FIG. 5A shows doublets filtered by a new method of data subsetting, reclustering, and definition of doublets as clusters containing marker genes which we believed represented mutually exclusive cell types. Comparison of previously labeled doublets to data processed by other published doublet identification methods is shown above. FIG. 5B shows data also contained batch effects associated with sequencer instrument, which we mitigated for cell clustering using standard integration functions included in the Seurat R package. FIG. 5C shows that patient samples mix well across most cell clusters after integration.

[0022] FIG. 6 shows single cell heatmap showing cluster marker gene expression. Cluster subsets were often named using a highly expressed marker gene. Other cluster name abbreviations include: KRT-b, basal keratinocyte; KRT-sp, spinous keratinocyte; KRT-g, granular keratinocyte; KRT-wr, wound response keratinocyte; LC, Langerhans cell; MAC,macrophage; cDCl, conventional dendritic cell 1; pDC, plasmacytoid dendritic cell; Treg, T regulatory cell; NK, natural killer cell; CD4-CD8-active, CD4+CD8+T cell with activation marker gene expression.

[0023] FIG. 7 shows treg identity was validated using published Treg gene sets for GSEA analysis. The Treg upregulated gene set is increased and the Treg downregulated gene set is decreased relative to other CD4+ T cells. Treg gene sets were gathered from 2 published sources (Feuerer et al. 2009; Pfoertner et al. 2006) and merged to a single reference gene set for GSEA analysis.

[0024] FIG. 8 shows Multi-linear models were created for each cell type to predict cell proportion of day 2 contact dermatitis samples using the type of lesion (allergy / irritant) and the lesion severity grade as two covariates as shown by the design formula below: Proportion. estimate = intercept + pi*lesion.type + P2*lesion. severityThe estimated P coefficients as well as 95% confidence intervals are plotted.

[0025] FIG. 9 shows UMAP embeddings for the 25-group cell type clusters used for cell population analysis, as well as for 9-group cell type clusters used for gene expression analyses. LC, Langerhans cell; pDC, plasmacytoid dendritic cell; Treg, T regulatory cell; UMAP, Uniform Manifold Approximation and Projection.

[0026] FIG. 10A shows unfiltered DE gene heatmaps that show all DE genes from FIG. 2 along with DE genes which were not shown in FIG. 2 for having unclear / complex trends. FIG. 10B shows several statistics for each cell type in a table, including lesion cell numbers, total expressed gene number, total DE gene number, and categorization of DE genes.

[0027] FIG. 11A and 11B show violin bar plots for all CCL, and CXCL-pattern chemokines expressed within the data, as well as all interferon and interleukins. FIG. 11A shows all chemokines in data set. FIG. 11B shows IFNs and interleukins in data. Some additional genes like IL6 are found within fewer than 5 cells, and were filtered out of the gene set during normalization. These filtered genes found in fewer than 5 cells are not plotted. FIG. 11C shows violin bar plots for miscellaneous other genes referenced in the text.

[0028] FIG. 12A shows CCL22 pseudobulk gene expression for LYZ+ myeloid cells, plotted for each biologic replicate across skin lesions. FIG. 12A shows decreased CCL22 expression in both irritant and allergic contact dermatitis relative to nonlesional samples. FIG. 12B shows pseudobulk CCL22 gene expression for each subcluster of LYZ+ myeloid cells (Myeloid-1, Myeloid-ccl22, MAC-fcnl, cDCl), which shows relatively stable CCL22 expression across lesion types within each cell type subcluster. FIG. 12C shows cellproportion analysis of LYZ+ myeloid subclusters across skin lesions, which shows decreased proportions of the Myeloid-ccl22 cluster (named for their abundant expression of CCL22) within contact dermatitis lesions that likely accounts for pattern of decreased CCL22 expression within the merged LYZ+ myeloid cell cluster in contact dermatitis lesions. Biologic replicates within these plots represent suction blister cells from a single patient skin lesion on a single date.

[0029] FIG. 13A shows pseudobulk plots for total expression of IFNG, IL4, and IL13. FIG. 13B shows faceted pseudobulk plots, depicting cellular sources of IFNG, IL4, and IL13.

[0030] FIG. 14A shows full heatmaps of all ligands in data set, separated by DE status, clustered by k means, and categorized by patterns of expression. FIG. 14B shows full heatmaps of all receptors in data set, separated by DE status, clustered by k means, and categorized by patterns of expression.

[0031] FIG. 15A shows density plots of ligand gene expression levels, which demonstrates a difference in expression levels across gene categories. FIG. 15B shows density plots of receptor gene expression levels, which demonstrates a difference in expression levels across gene categories.

[0032] FIG. 16 shows Lasso regression formula parameters for logistic models that were derived from suction blister proteomic data, and for which ROC curves were plotted in FIG. 4E. LASSO, least absolute shrinkage and selection operator; ROC, receiver operating characteristic curve.

[0033] FIG. 17 shows pseudobulk gene expression data grouped by patient, lesion, and cell type were used to create violin plots for all CCL- and CXCL-pattem chemokines, IFNs, and ILs. Genes such as IL6 that were found in fewer than 5 cells were filtered out of the gene set during normalization and also not plotted.

[0034] FIGs. 18A-18C: IL-4 and IL- 13 are among top predicted ligands based on differential expression data from LYZ+ myeloid cells and Langerhans cells inACD. Nichenetr was used to predict ligands most likely to affect gene expression in ACD and ICD. FIG. 18A shows top predicted ligands for differentially expressed gene sets from LYZ+ myeloid cells in day-2 allergy relative to those in nonlesional skin. FIG. 18B shows top predicted ligands for differentially expressed gene sets from Langerhans cells in day-2 allergy relative to those in nonlesional skin. IFNG, IL-4, and IL- 13 were among the top predicted ligands for differentially expressed gene sets from LYZ+ myeloid cells and Langerhans cells in day-2 allergy relative to those in nonlesional skin. FIG. 18Cshows pseudobulk expression data grouped by patient, lesion, and cell type for IL-4 / IL-13 target genes show upregulation in LYZ+ myeloid cells and Langerhans cells in both ACD and ICD. ACD, allergic contact dermatitis; ICD, irritant contact dermatitis.DETAILED DESCRIPTION

[0035] Several aspects of the disclosure are described below, with reference to examples for illustrative purposes only. It should be understood that numerous specific details, relationships, and methods are set forth to provide a full understanding of the disclosure. One having ordinary skill in the relevant art, however, will readily recognize that the disclosure can be practiced without one or more of the specific details or practiced with other methods, protocols, reagents, cell lines, and animals. The present disclosure is not limited by the illustrated ordering of acts or events, as some acts may occur in different orders and / or concurrently with other acts or events. Furthermore, not all illustrated acts, steps, or events are required to implement a methodology in accordance with the present disclosure. Many of the techniques and procedures described, or referenced herein, are well understood and commonly employed using conventional methodology by those skilled in the art.

[0036] Unless otherwise defined, all terms of art, notations, and other scientific terms or terminology used herein are intended to have the meanings commonly understood by those of skill in the art to which this disclosure pertains. In some cases, terms with commonly understood meanings are defined herein for clarity and / or for ready reference, and the inclusion of such definitions herein should not necessarily be construed to represent a substantial difference over what is generally understood in the art. It will be further understood that terms, such as those defined in commonly-used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and / or as otherwise defined herein.

[0037] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.

[0038] As used herein, the indefinite articles “a,” “an,” and “the” should be understood to include plural reference unless the context clearly indicates otherwise.

[0039] Throughout this specification and the claims which follow, unless the context requires otherwise, the word “comprise,” and variations such as “comprises” and “comprising,” will be understood to imply the inclusion of, e.g., a stated integer or step or group of integers or steps, but not the exclusion of any other integer or step or group ofintegers or steps. When used herein, the term “comprising” can be substituted with the term “containing” or “including.”

[0040] As used herein, “consisting of’ excludes any element, step, or ingredient not specified in the claim element. When used herein, “consisting essentially of’ does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim. Any of the terms “comprising,” “containing,” “including,” and “having,” whenever used herein in the context of an aspect or embodiment of the disclosure, can in some embodiments, be replaced with the term “consisting of,” or “consisting essentially of’ to vary the scope of the disclosure.

[0041] As used herein, the conjunctive term “and / or” between multiple recited elements is understood as encompassing both individual and combined options. For instance, where two elements are conjoined by “and / or,” a first option refers to the applicability of the first element without the second. A second option refers to the applicability of the second element without the first. A third option refers to the applicability of the first and second elements together. Any one of these options is understood to fall within the meaning, and, therefore, satisfy the requirement of the term “and / or” as used herein. Concurrent applicability of more than one of the options is also understood to fall within the meaning, and, therefore, satisfy the requirement of the term “and / or.”

[0042] When a list is presented, unless stated otherwise, it is to be understood that each individual element of that list, and every combination of that list, is a separate embodiment. For example, a list of embodiments presented as “A, B, or C” is to be interpreted as including the embodiments, “A,” “B,” “C,” “A or B,” “A or C,” “B or C,” or “A, B, or C.”

[0043] A description of example embodiments follows.Methods of the Detecting, Diagnosing, and / or Distinguishing a Skin Reaction

[0044] Allergic contact dermatitis is a common pruritic skin disease caused by environmental chemicals that induce cell-mediated skin inflammation within susceptible individuals. Diagnosis of patient-specific allergen sensitivities can be difficult, as it relies heavily on clinical intuition. Development of an objective, molecular-based test has been hindered by the striking similarity between allergic and irritant contact skin reactions.

[0045] Contact dermatitis is typically categorized as either allergic contact dermatitis (ACD) or irritant contact dermatitis (ICD). ICD is believed to be caused by chemical erosion of the stratum corneum, the outermost hydrophobic portion of the skin barrier, and manychemicals can cause this non-allergic form of skin inflammation when applied under patch test conditions. “Handwashing dermatitis” due to overuse of detergents is a common form of environmental ICD, and it is believed that without the protective stratum comeum in place, damage occurs to keratinocytes that leads to non-allergic inflammation.

[0046] Allergic contact dermatitis is a “type 4” delayed hypersensitivity reaction whereby the immune system becomes increasingly sensitive to a specific chemical trigger, resulting in an increasingly stronger immune response with each subsequent exposure (Inagaki and Nagai 2009). The process of increasing future immunoreactivity to a specific allergen is known as sensitization, and subsequent induction of allergic immune response is referred to as allergy elicitation. Sensitization occurs when a host is exposed to an allergen at a sufficiently stimulating dose. The allergen crosses through imperfections of the skin barrier, activating innate immune pattern recognition receptors while also covalently binding to host proteins. This combination of events induces Langerhans cell egress from the skin to nearby lymph nodes, where the allergen-modified host peptides are presented to T cells. T cells specific to allergen-modified peptides become activated, expand in numbers, and circulate thereafter permanently within the host as memory T cells. Allergen-specific T cells, and in some cases natural killer cells, provide host memory of prior allergen exposures, driving stronger immune responses upon allergen re-exposure and heightened sensitivity to low exposure doses (Engeman et al. 2004; Fyhrquist et al. 2012; Rouzaire et al. 2012; van den Boorn et al. 2016). Since this memory immune response does not occur in non-allergic, irritant contact dermatitis, characterization of this adaptive immune response can provide the basis for a future molecular-based diagnostic test that is specific for chemical-induced allergic-type inflammation

[0047] Provided herein are methods of detecting an allergic skin reaction in a subject in need thereof. Also provided herein are methods of diagnosing an allergic skin reaction in a subject in need thereof. Further provided herein are methods of distinguishing an allergic skin reaction from an irritant skin reaction in a subject in need thereof.

[0048] The methods of detecting, diagnosing, or distinguishing generally comprise the steps of a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining an expression level for one or more (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, or more) protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF,IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL 18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof (see Table 1); and c) identifying the subject as having an allergic skin reaction and / or, in some embodiments, an irritant skin reaction, based on the level (e.g., expression level) of the one or more biomarkers.

[0049] Table 1. Protein Biomarker Information

[0050] In some embodiments, the methods comprise comparing the level (e.g., expression level) of one or more (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, or more) protein biomarkers in the sample from the subject to a control, wherein an increase or decrease in the level of the biomarker compared to the control is indicative of either an allergic skin reaction or an irritant skin reaction. The control can be a reference standard or unaffected skin of the subject, such as an area of skin of the subject that has not been contacted with the potential allergen.

[0051] The expression level of a protein biomarker in a skin sample can be determined by a person of skill in the art using any of a variety of known techniques and reagents. In some embodiments, the expression level of a protein biomarker can be determined using an antibody-based assay that utilizes antibodies (e.g., detectably-labeled antibodies) that specifically bind to the biomarker. Such assays include, for example, an enzyme-linked immunosorbent assay (ELISA), OLINK® assays, radioimmunoassays, and proteomics assays.

[0052] In some embodiments, the methods comprise determining the relative amount of one or more of the protein biomarkers, wherein the relative amount(s) are indicative of an allergic skin reaction or an irritant skin reaction. In some embodiments, the relative amount of a protein biomarker is determined by calculating a ratio of the expression levels of two or more of the protein biomarkers. In particular embodiments, the ratio is a ratio wherein the numerator is the expression level(s) of one or more allergen-specific protein biomarkers, such as IL4, CXCL10, or IFNG for example, and the denominator is the expression level(s) of one or more irritant-enriched protein biomarkers, such as CCL20 or TSLP for example. In particular embodiments, the ratio is a ration wherein the numerator is the expression level(s) of one or more allergen-specific protein biomarkers, such as IL4, CXCL10, CXCL11, or IFNG for example, and the denominator is the expression level(s) of one or more shared allergen-irritant common upregulated protein biomarkers, such as, e.g., CCL19 or IL8.

[0053] In some embodiments, the relative amount of a protein biomarker is determined by a multi-linear logistic regression analysis. In particular embodiments, the multi-linear logistic regression analysis uses a formula that for example assigns a positive coefficient multiplier to the expression level of an allergen-specific protein biomarker, and a negative coefficient multiplier to the expression level of an irritant-specific protein biomarker.

[0054] In some embodiments, the methods disclosed herein are performed using a skin sample that comprises skin interstitial fluid. For example, the skin sample can be a suction blister sample, an absorptive microneedle skin sample, a tape strip sample, a shave biopsy sample, or a punch biopsy skin sample.

[0055] A person of ordinary skill in the art will appreciate that a variety of sampling methods and devices can be employed to obtain a suitable skin sample for use in the methods disclosed herein. For example, the skin sample can be obtained by suction blister sampling, absorptive microneedle skin sampling, or punch biopsy skin sampling.

[0056] Generally, the skin sample will be of an area of skin that has been contacted with or otherwise exposed to a potential allergen. For example, a potential allergen can be applied to the skin of a subject using any of a variety of application methods including, but not limited to, an allergen patch test, a tape strip, or a microneedle device.

[0057] In some embodiments, the skin sample is obtained from the subject within about 7 days, for example, about 1, about 2, about 3, about 4, about 5, about 6, or about 7 days, after contacting the subject’s skin with a potential allergen(s). In some embodiments, the skin sample is obtained in the range about two to about four days after contacting the skin with the potential allergen. In some embodiments, the skin sample is obtained less than 24 hours after contacting the skin with the potential allergen.

[0058] In some embodiments, the subject is identified as having an allergic skin reaction. The allergic skin reaction can be characterized by itch, redness, scale, or blisters, or any combination thereof, which can represent varying degrees of severity. In some embodiments, the subject is identified as having allergic contact dermatitis. In some embodiments, the subject receives a diagnosis of having an allergic skin reaction. In some embodiments, the subject is treated for an allergic skin reaction. Suitable treatments for allergic skin reactions are known to those of skill in the art. Such treatments include, for example, limiting or avoiding exposure to the allergen that caused the reaction, for example, to prevent subsequent future allergic reactions, and / or treatment with immunosuppressive modalities such as with corticosteroid medications.

[0059] In some embodiments, the subject is identified as having an irritant skin reaction. The irritant skin reaction is typically characterized by redness and a glazed appearance, but sometimes also by itch, scale and blisters. In some embodiments, the subject is identified as having irritant contact dermatitis. In some embodiments, the subject receives a diagnosis of having an irritant skin reaction. In some embodiments, the subject is treated for an irritant skin reaction. Suitable treatments for allergic skin reactions are known to those of skill in the art. Such treatments include, for example, limiting or avoiding exposure to the irritant that caused the reaction, and / or use of products that enhance skin barrier protection such as personal protective equipment, including gloves, or topical moisturizer ointments or creams, including for example petrolatum, silicone, and dimethicone based products.Methods of Preparing a Skin Sample

[0060] Also provided herein are methods of preparing a skin sample that is useful for detecting an allergic skin reaction in a subject. The methods of preparing generally comprise the steps of a) applying a potential allergen to an area of skin of a subject; b) obtaining a skin sample from the area of skin to which the allergen has been applied; and c) combining the sample with one or more reagents for detecting one or more protein biomarkers selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, C ASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof.

[0061] The “applying” and “obtaining” steps in the methods of preparing a skin sample disclosed herein can be practiced as described herein for the methods of detecting, diagnosing, or distinguishing disclosed herein.

[0062] The methods of preparing disclosed herein further comprise combining the sample (e.g., all or a portion of the sample) with one or more reagents for detecting one or more protein biomarkers selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof.

[0063] The one or more reagents for detecting protein biomarkers can be any agent that is capable of specific detection of at least one of the protein biomarkers disclosed herein. Forexample, the reagents can be antibodies (e.g., antibodies that specifically bind a protein biomarker disclosed herein), nucleic acids (e.g., aptamers that specifically bind a protein biomarker disclosed herein), or small molecules (e.g., small molecules that specifically bind a protein biomarker disclosed herein), or a combination of any of the foregoing.

[0064] In some embodiments, the one or more reagents are immobilized, such as on beads or a chromatography resin. In some embodiments, the one or more reagents comprise a detectable label, such as a fluorescent dye.Methods of Identifying or Classifying a Substance

[0065] Also provided herein are methods of identifying a substance as a skin allergen or a skin irritant for a subject. The methods of identifying generally comprise the steps of a providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a test substance; b) determining an expression level for one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, CASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof; and c) identifying the test substance as an allergen or an irritant based on the level (e.g., expression level) of the one or more biomarkers.

[0066] The methods of identifying a substance as a skin allergen or a skin irritant can be practiced as described herein for the methods of detecting, diagnosing, or distinguishing disclosed herein. The substance can be any test substance, such as a chemical substance (e.g., a chemical compound). The substance can be a naturally occurring substance or an artificial or synthetic substance. The substance can be a known allergen and / or a known irritant, or a substance that has not previously been characterized as an allergen and / or irritant.Exemplification

[0067] To discover specific biomarkers of allergic contact dermatitis, transcriptomic and proteomic changes that occur within the skin during each type of contact dermatitis were analyzed. Allergic and irritant contact dermatitis were induced in healthy human volunteers and skin was sampled using a non-scarring suction blister biopsy method to analyze skin reactions. Single cell RNA-sequencing analysis identified differences in cell populations between irritant and allergic dermatitis, however, these differences were heavily influenced by reaction severity, with significant heterogeneity from subject to subject. Cell-specific transcriptome analysis identified a wound response-like signature common to both irritant and allergic contact dermatitis, which was characterized by upregulation of keratinocyte genes associated with re-epithelization (KRT17,KRT16,KRT6A,KRT6B) as well as upregulation of certain inflammatory genes such as IL8 and CCL17 by myeloid cells. Allergy-specific transcriptomic signatures included upregulation of interferon-induced genes such as CCL2, CXCL9, and CXCL10 by keratinocytes and LYZ+ myeloid cells, and well as upregulation of cell stress and unfolded protein response genes within CD4+ and CD8+ T cells. Cell signaling network analysis also estimated that several ligands such as IL4, IL1, TNF, and several gpl30-family ligands, which were not found by DE analysis (possibly due to low gene expression), may significantly influence cellular gene expression in contact dermatitis. Proteomic analysis of suction blister samples confirmed that IFNG, the IFNG- induced chemokines CXCL9 / CXCL10 / CXC11, and IL4 are specifically upregulated in allergic lesions, and revealed that CCL20 is comparatively higher expressed in irritant contact dermatitis.Example 1.Materials and Methods

[0068] Study design: Healthy individuals without a history of autoimmune or inflammatory skin disease were recruited at the University of Massachusetts Chan Medical School under several Institutional Review Board-approved protocols (H-3947, H-14848, STUDY-0321). Under protocol H-14848, 12 volunteers donated healthy suction blister skin biopsy samples, which were analyzed by single cell RNA sequencing to form a baseline of healthy skin data. Under protocol H-3947, 10 volunteers were recruited for chemical patch testing followed by skin sampling by suction blister skin biopsy.

[0069] Chemical patch test induction of irritant and allergic contact dermatitis: Irritant contact dermatitis was induced by a solution of 2% sodium lauryl sulfate (SLS), obtained from Sigma Aldrich (product #1614363), which was dissolved in sterile cell culture grade water. 35 pL of 2% SLS was placed onto 8mm SmartPractice Finn Chambers® containing paper discs, applied to volunteer skin, and occluded by 3M Tegaderm™ dressing. Allergic contact dermatitis was induced using squaric acid dibutyl ester (SADBE). SADBE was obtained in numerous concentrations in an acetone base from Boulevard Compounding Pharmacy in Worcester, MA. Volunteers were first sensitized to SADBE by application of 100 pL of 2% SADBE to the inner arm, which was covered by Tegaderm™ dressing for 24 hours. To identify each participant’s individual ideal elicitation dose, SADBE doses ranging between 0.0001% and 2% were applied to participant skin in a volume of 20 pL under Tegaderm™ dressing for 24 hours to find the minimal dose that induced a mild erythematous skin reaction. Allergic contact dermatitis was then induced by application of 20 pL of each participant’s individual elicitation dose under Tegaderm™ dressing for 2 days.

[0070] Suction blister skin sampling: Suction blisters were induced using the Negative Pressure Instrument Model NP-4 (Electronic Diversities, Finksburg, MD), with a negative pressure between 7-10 mmHg until blisters formed. Blister fluid was aspirated using a syringe, transferred to a collection tube, and centrifuged at 4°C at 350 x g for 10 minutes. Blister fluid supernatant was frozen at -80°C until proteomic analysis. Pelleted blister fluid cells were resuspended in 100 pL of PBS with 15% Axis-shield OptiPrep™ density gradient medium for single cell RNA sequencing by InDrop.

[0071] Single cell RNA sequencing: Single cell RNA sequencing was performed using the inDrop method as described by Zilionis et al. and Gellatly et al. (Gellatly et al. 2021; Zilionis et al. 2017). RNA transcriptome libraries were sequenced on Illumina NextSeq or NovaSeq platforms, and BCL files were converted to fastq files using bcl2fastq or bcl2fastq2. Fastq files were processed using the DolphinNext pipeline to generate final data tables containing unique molecular identifier (UMI) gene counts for each barcoded droplet that could be analyzed in R as single cell RNAseq data (Yukselen et al. 2020).

[0072] Single cell RNA sequencing data analysis: UMI tables for all suction blister skin samples in this study were filtered to droplets containing high quality cell data by only including droplets containing 300 or more UMIs, 150 or more unique genes, and mitochondrial UMIs amounting to less than 7.5% of total droplet UMIs. Data was then processed in R using the Seurat R package to normalize, scale, and initially cluster cells.Five broad subsets of cells were observed corresponding to keratinocytes, melanocytes, antigen presenting cells, lymphocytes, and erythrocytes; and initial cell clusters were assigned to each category by top cluster marker genes. Mutually exclusive marker genes for each broad subset were defined as follows: keratinocyte (KRT1 / 2 / 5 / 6A / 6B / 10 / 14 / 15 / 17 / 77, KRTDAP, DMKN, MUCL1), melanocyte (TYR, ML ANA, PMEL, DCT, MITF), antigen presenting cell (HLA.DR pattern, LYZ, CD207, CD86, CD83, CLEC4A, CLEC4C, CD163, MRC1, NRP1, HM0X1, JCHAIN), lymphocyte (TRAC, TRDC, NCAM1), erythrocyte (HBB, HBA2). Any unclassified cell clusters were assigned to the closest proximity cell type based on UMAP distance.

[0073] Erythrocyte clusters were filtered from further analysis, and each other broad cell subset was reprocessed as an individual dataset in Seurat. New cell clusters within each dataset that were found to possess marker genes from other broad cell categories were labeled as doublets and filtered from future analyses. After filtration of doublets, all remaining cells were re-merged and processed in Seurat for a data set of pure singlets (FIG. 5). Several clusters were observed to correlate with cell cycle associated genes such as TOP2A and CDK1, so a list of 2151 genes that were identified by DE analysis to be significantly upregulated in cell clusters with high TOP2A and CDK1 expression were excluded. Batch effects were observed that were associated with the 5 different sequencers, so Seurat’s IntegrateData function was used to integrate across sequencer datasets prior to PCA analysis that was used for final cell clustering and LTMAP embedding (FIG. 5). Top cluster marker genes with highly specific expression were identified for each final cell cluster by DE analysis and weight of genes by average log2(fold change) / (percent of cells outside of cluster with positive gene expression). Treg cluster identity was also validated using GSEA analysis of published FACS-sorted Treg gene sets (FIG. 7).

[0074] Statistical analysis of differential gene expression across lesional skin conditions was performed using EdgeR R package, and gene set enrichment analysis as well as gene ontology enrichment analysis were performed using ClusterProfiler R package (Robinson and Oshiack 2010; Yu et al. 2012). Cell signaling network analysis was performed using a modified method that merged outputs of nichenetr and cellphoneDB analyses (Browaeys et al. 2020; Efremova et al. 2020). All ligand and receptor data used for signaling network analysis is plotted as heatmaps in FIG. 11. For trend classification of non-differentially expressed ligands within signaling network plots, ligand cell-specific pseudobulk data was filtered to the cell type containing maximum gene expression among nonlesional, day 2allergy, and day 2 irritant lesions; and then scaled across lesions by percent of maximum expression. Ligands were classified as having an increasing trend if nonlesional percent of maximum expression was less than 50%, and lesions contained greater than 75%. Several gene expression plots, including violin plots and heatmaps, were created by modified functions of the SignalingSingleCell R package (Gellatly et al. 2021). All R scripts used to analyze scRNAseq data are available at: Link here.

[0075] Skin interstitial fluid proteomic analysis: Suction blister fluid supernatant was analyzed by Olink® proteomic proximity extension assay using the 92-marker Inflammation panel, and reported Olink Normalized Protein Expression (NPX) values were analyzed in R. Differential expression analysis for protein concentrations was performed using the pairwise t test function from the rstatix package with assumed unequal variances and the Benjamini-Hochberg correction for multiple hypothesis testing. PCA was performed using all NPX protein values. LASSO regression analysis with 10-fold cross validation was performed using a logistic model to classify samples as allergic or non-allergic by protein NPX values. Multiple lambda penalty values were tested, and models were assessed by the area under the ROC curve. The model with the least misclassification error had 6 non-zero protein coefficients, and for comparison other models with fewer non-zero protein coefficients that were derived from larger lambda penalty values were also evaluated by ROC curve. Data for each LASSO model is provided in FIG. 16.Results

[0076] Protocol to systematically profile contact dermatitis

[0077] To minimize variability between allergic and irritant reactions, we induced allergic contact dermatitis in healthy human volunteers using a single allergen, squaric acid dibutyl ester (SADBE), and irritant contact dermatitis using a single irritant, sodium lauryl sulfate (SLS). We applied each chemical under occlusion for 2 days, similar to the standard clinical patch testing protocol, as well as acetone vehicle alone. As diagramed in FIG. 1A, skin was sampled using a suction blister biopsy approach. The cellular portion of suction blister fluid was analyzed by single cell RNA-sequencing using a custom made indrop device as described previously (10.1038 / nprot.2016.154, 10.1126 / scitranslmed.abd8995) while the non-cellular fluid was analyzed by Olink® proteomic proximity extension assay. To longitudinally sample allergic inflammation, 2 separate SADBE patches were placed 2 days apart, on day -4 and day -2, and all sites were sampled on day 0. Skin reactions werephotographed and scored according to the standard clinical patch testing method as recommended by the ICDRG (FIG. IB). Notably, strong (vesicular) allergic contact reactions could not be sampled by suction blisters due to impaired epidermal integrity, so we aimed to induce mild reactions that were characterized by erythema with minimal vesicles, which we found could be more reliably sampled by suction blistering. Two participants in this study also developed an allergic-like reaction to the acetone vehicle patch, which may indicate contamination or impurity of the acetone. Data from visible reactions to acetone vehicle patches were excluded from all analyses.

[0078] Cell proportions are similar between allergic and irritant contact dermatitis

[0079] We used single cell RNA-seq (scRNAseq) to investigate if differences in cell proportionality distinguish allergic from irritant contact dermatitis, as well as to identify cellular sources of inflammatory signals. Data clustered into distinct groups representing all epidermal cell types, and we categorized cells into 25 populations based on Louvain clustering results and patterns of marker gene expression (FIG. 1C, FIG. ID, FIG. 6). Several cell populations, including myeloid- 1 and many CD4+ T cell populations, were enriched in both irritant and allergic contact lesions relative to negative control samples (healthy, nonlesional, and acetone vehicle) (FIG. IE). The proportions of spinous keratinocytes (KRT-sp) were decreased in both irritant and allergic compared to control samples, likely due to inflammatory dilution. Wound response keratinocytes (KRT-wr) were identified by their expression of keratins 6, 16, and 17, and were found in increased proportions within all day 2 patch samples. Langerhans cells and melanocytes were decreased in all patch samples.

[0080] Four cell populations were increased by 2-fold or greater in allergic contact samples relative to irritant (FIG. IF). Consistent with findings by Fortino et al. that BATF3 and HSPA6 were 2 of the top 10 highest allergy-enriched genes relative to irritant reactions, we found that BATF3+ conventional dendritic cells (cDCl) and a subpopulation of HSPA6+ CD4+ T cells were both enriched in allergy lesions (Fortino et al. 2020). Notably, allergic reactions tended to be more severe than irritant reactions, so it is possible that the reaction severity also influenced recruitment of these cell populations. To minimize the confounding effect of reaction severity, we attempted to model cell proportionality by multi-linear regression using both reaction type (irritant or allergic) and reaction severity (visual score); however, we were unable to draw confident conclusions due to limited sample numbers and high data variability (FIG. 8).

[0081] Patch application induces downregulation of many homeostatic genes

[0082] For cell type-specific transcriptomic analysis, we merged the 25 cell type clusters into 9 groups to reach greater statistical power while preserving major distinctions among cell types (FIG. 9). Day 4 SADBE, Day 2 SADBE, and SLS were each compared to healthy / nonlesional skin, and differentially expressed (DE) genes were then clustered by k- means to identify groups of genes with similar expression patterns across the 5 skin conditions. Across multiple cell types, we noticed many DE genes followed 3 patterns of expression corresponding to downregulated homeostatic genes (FIG. 2A top panels), common upregulated genes induced within both irritant and allergic lesions (FIG. 2A middle panels), and allergy-specific upregulated genes (FIG. 2A bottom panels). The number of DE genes across cell types did not correlate with total cell numbers or transcripts detected per cell, suggesting that cell numbers and sequencing depth did not substantially influence this DE analysis (FIG. 10B). Among downregulated homeostatic genes, we found melanocytes downregulated the growth receptor KIT and the adhesion gene LI CAM, suggesting that a biological effect of patch application may contribute to the decreased melanocyte number in all patch test samples (FIG. 11C). CXCL14, which is known to antagonize Langerhans cell exit from the skin, was also downregulated in several cell types consistent with the observed decrease in Langerhans cell proportionality within patch test samples (FIG. 11 A). Aside from these examples, several downregulated homeostatic genes were difficult to interpret due to a decision to merge closely related cell type clusters for greater statistical power in DE analysis. For example, CCL22 first appeared to be a downregulated DE gene among LYZ+ myeloid cells, yet this was due to a dilution of the CCL22-producing cellular subpopulation within the merged LYZ+ myeloid cluster rather than a significant change in CCL22 expression on a per cell basis (FIG. 12). Therefore, further analysis was primarily on upregulated DE genes.

[0083] Irritant and allergic contact dermatitis both induce an epidermal wound response

[0084] Many DE genes were upregulated within both irritant and allergic contact dermatitis, and one clear signature in this category resembles an epidermal wound response. Cells classified as wound response keratinocytes were found in increased proportion within both Day 2 irritant and Day 2 allergic patch reactions, and genes associated with wound response were significantly upregulated by keratinocytes (KRT6A, KRT6B, KRT16, and SPRR1B) (FIG. 2A). Differentially expressed genes within LYZ+ myeloid cells alsoincluded CCL17 and CXCL8 (IL8) (FIG. 11 A), which have been associated with epidermal injury (Raymondi Silva et al. 2022; Rennekampff et al. 2000).

[0085] Allergic-specific transcriptomic responses are associated with IFNG response

[0086] Allergy-specific induced genes were found in every major cell type except for Langerhans cells, and we first used gene ontology enrichment analysis to characterize these genes for each cell type (FIG. 2A). Several different ontology terms were enriched by different cell types. To characterize the degree of cell specificity, we performed gene set enrichment analysis for each ontology term using data from day 2 allergic and irritant lesions for all cell types (FIG. 2B). In allergic relative to irritant lesions, gene sets related to cellular stress were elevated in several immune cell types; the response to interferon gamma gene set was enriched in most cell types; and cytokine-associated gene sets were enriched in keratinocytes and LYZ+ myeloid cells. Consistent with gene set analysis, IFNg-induced chemokines (CXCL9, CXCL10, and CXCL11) were specifically upregulated in keratinocytes and LYZ+ myeloid cells within allergy samples, and pseudobulk total IFNG was elevated in allergy samples (FIG. 11, FIG. 13A). All lymphocyte populations were sources of IFNG within allergic lesions, yet CD8+ T cells and NK cells contributed the most towards total pseudobulk IFNG (FIG. 2C, FIG. 13B).

[0087] IL 4 and IL 13 exhibit different expression patterns in allergic and irritant contact dermatitis

[0088] Since the type 2 cytokines IL4 and IL13 have also been associated with allergy, we also investigated the pattern of cellular expression for these cytokines (Bitton et al. 2020; Dieli et al. 1999). IL4 was found in very few cells, yet lymphocyte expression of IL4 was a highly specific quality of allergy samples (FIG. 2C). IL13 was more broadly detected, including within CD4+ T cells from nonlesional skin, and both CD8+ T cells and NK cells were found to upregulate IL13 in both irritant and allergic lesions (FIG. 2C). Thus, IL13 expression was increased in both irritant and allergic lesions, while IL4 expression was specifically increased in allergic lesions (FIG. 13).

[0089] Receptor-ligand cell-to-cell signaling network analysis estimates that several lowly expressed ligands contribute to contact dermatitis transcriptional response

[0090] The low expression level of many cytokines such as IL4 complicated direct differential expression analysis, nevertheless manual inspection of their expression suggested that many may be involved in both allergic and irritant dermatitis (FIG. 3A, FIG. 3B, FIG. 14, FIG. 15). To investigate this possibility, we sought to use reported cytokine signalingcascades to identify cytokines upstream of observed transcriptional changes. We performed signaling analysis using nichenetr and cellphonedb R packages to create a consensus network of ligand-receptor interactions that each method agreed upon, and that were estimated to influence surrounding cellular transcription.

[0091] Using this method, we found that IFNG, IL4, and IL 13 were all predicted by nichenetr to be highly active ligands capable of inducing observed transcriptional changes that occur within LYZ+ myeloid cells of allergic contact lesions relative to those of healthy skin (FIG. 3C). Consistent with signaling analysis prediction, several putative IL4 / IL13 target genes were significantly increased within LYZ+ myeloid cells in allergic lesions (FIG. 3D). Although IL4 and IL 13 induce many common target genes, ABL2 is reportedly more strongly regulated by IL4 than IL13. Consistent with the common upregulation of IL13 within both irritant and allergic lesions and the allergy-specific upregulation of IL4, many IL13 / IL4 shared target genes were upregulated by LYZ+ myeloid cells within both allergic and irritant lesions while ABL2 expression was greater in allergic relative to irritant lesions (FIG. 3D). This suggests that although IL4 expression is too low for statistically confident differential expression analysis, it may be functionally influential within allergic contact lesions.

[0092] To directly compare irritant and allergic contact dermatitis using signaling analysis, we created directed acyclic graphs to visualize these pathways and weighted signals by their comparative magnitude (FIG. 3E). Application of this comparative network method using all ligand and receptor data produced a highly complex network showing that many signals are upregulated within allergic relative to irritant lesions (FIG. 3F). The network includes many signals based on sparse data, yet some high magnitude cell-specific expression differences were identified by subsequent manual analysis. CD8+ T cells and Tregs both appear to specifically upregulate LIF and OSM (two IL6 family ligands) within allergic lesions; and Tregs also upregulate LTA within allergic lesions (FIG. 3G). In addition to IFNG-associated chemokines CXCL9, CXCL10, CXCL11, and CCL2, several other chemokines were upregulated in allergic lesions relative to irritant lesions, including CCL7, CCL8, and CXCL1. Irritant-associated trends included upregulation of SEMA7A by pDCs and LYZ+ myeloid cell expression of EREG, AREG, CXCL2, and CXCL3 (FIG. 3H). While some trends in lowly expressed genes could be affected by technical issues such as sequencing dropout rate, this analysis method possesses the capacity to highlight lesionalsignals of potential functional significance that otherwise could be missed by simple differential expression analysis.

[0093] Skin interstitial fluid proteins distinguish allergic from irritant contact dermatitis

[0094] To test whether allergy-specific gene expression can also be detected in proteinbased assays, we analyzed suction blister fluid non-cellular supernatant by Olink® proteomic proximity extension assay to quantify 92 inflammatory proteins. Most of these proteins were detectable in suction blister fluid samples, and many proteins appeared to have increased concentrations within contact dermatitis samples (FIG. 4A). DE analysis found that IFNG, IFNG-associated chemokines (CXCL9, CXCL10, CXCL11), IL4, IL5, and IL13 were all significantly increased in allergy samples relative to irritant (FIG. 4B). CCL20 and CXCL1 were significantly increased in irritant samples relative to allergy, and this is consistent with a previous report that SLS more strongly induces CCL20 than allergic patch reactions (Koppes et al. 2017). PCA analysis of suction blister fluid proteins separated allergic, irritant, and healthy samples with moderate success, indicating that linear combinations of interstitial fluid proteins can distinguish contact dermatitis samples similar to bulk RNA captured from skin punch biopsy samples as shown by Fortino et al. (Fortino et al. 2020). PC2 appeared to distinguish between allergic and irritant samples, and proteins with the largest PC2 loading were mostly also identified by DE analysis (FIG. 4C, FIG. 4D). To build a numeric classifier that could be used in a protein-based diagnostic device, we performed LASSO regularized logistic regression. Using 10-fold cross validation we found that several linear models using combinations of 7 proteins (CXCL10, IL4, OPG, CCL19, CCL20, TSLP, and ILlalpha) could reliably distinguish both day 2 and day 4 allergic samples from nonlesional and irritant samples with as high as 93% sensitivity and 93% specificity (FIG. 4E, FIG. 4F).

[0095] Discussion

[0096] Classification of contact dermatitis using soluble skin interstitial fluid proteins obtained using minimally invasive techniques could transform clinical diagnosis of patient environmental contact sensitivities. The noninvasive skin sampling technique disclosed herein in conjunction with proteomic quantification disclosed herein could aid or replace current chemical patch test methods of reaction assessment.

[0097] Example 1 References

[0098] Arita K, South AP, Hans-Filho G, Sakuma TH, Lai-Cheong J, Clements S, et al.Oncostatin M Receptor-P Mutations Underlie Familial Primary Localized Cutaneous Amyloidosis. Am. J. Hum. Genet. 2008;82(l):73-80

[0099] Bitton A, Avlas S, Reichman H, Itan M, Karo-Atar D, Azouz NP, et al. A key role for IL-13 signaling via the type 2 IL-4 receptor in experimental atopic dermatitis. Sci. Immunol. 2020;5(44):eaaw2938

[0100] Browaeys R, Saelens W, Saeys Y. NicheNet: modeling intercellular communication by linking ligands to target genes. Nat. Methods. 2020; 17(2): 159-62

[0101] Dequeant M-L, Sagert J, Kalaitzidis D, Yu H, Porras A, McEwan B, et al. CD70 knockout: A novel approach to augment CAR-T cell function.

[0102] Dieli, Sireci, Scire, Salerno, Bellavia. Impaired contact hypersensitivity to trinitrochlorobenzene in interleukin-4-deficient mice: Reduced CH in IL-4-deficient mice.Immunology. 1999;98(l):71-9

[0103] Efremova M, Vento-Tormo M, Teichmann SA, Vento-Tormo R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligandreceptor complexes. Nat. Protoc. 2020; 15(4): 1484-506

[0104] Engeman T, Gorbachev AV, Kish DD, Fairchild RL. The intensity of neutrophil infiltration controls the number of antigen-primed CD8 T cells recruited into cutaneous antigen challenge sites. J. Leukoc. Biol. 2004;76(5):941-9

[0105] Feuerer M, Hill JA, Mathis D, Benoist C. Foxp3+ regulatory T cells: differentiation, specification, subphenotypes. Nat. Immunol. 2009;10(7):689-95

[0106] Fortino V, Wisgrill L, Werner P, Suomela S, Linder N, Jalonen E, et al. Machinelearning-driven biomarker discovery for the discrimination between allergic and irritant contact dermatitis. Proc. Natl. Acad. Sci. 2020;117(52):33474-85

[0107] Fyhrquist N, Wolff H, Lauerma A, Alenius H. CD8+ T Cell Migration to the Skin Requires CD4+ Help in a Murine Model of Contact Hypersensitivity. Kaul R, editor. PLoS ONE. 2012;7(8):e41038

[0108] Gellatly KJ, Strassner JP, Essien K, Refat MA, Murphy RL, Coffin-Schmitt A, et al. scRNA-seq of human vitiligo reveals complex networks of subclinical immune activation and a role for CCR5 in T reg function. Sci. Transl. Med. 2021;13(610):eabd8995

[0109] Grootjans J, Kaser A, Kaufman RJ, Blumberg RS. The unfolded protein response in immunity and inflammation. Nat. Rev. Immunol. 2016;16(8):469-84

[0110] Inagaki N, Nagai H. Analysis of the Mechanism for the Development of Allergic Skin Inflammation and the Application for Its Treatment: Mouse Models for the Development of Remedies for Human Allergic Dermatitis. J. Pharmacol. Sci.2009; 110(3):251-9

[0111] Kalboussi H, Kacem I, Aroui H, El Maalel O, Maoua M, Brahem A, et al. Impact of Allergic Contact Dermatitis on the Quality of Life and Work Productivity. Dermatol. Res. Pract. 2019;2019: 1-8

[0112] Koppes SA, Ljubojevic Hadzavdic S, Jakasa I, Franceschi N, Jurakic Toncic R, Marinovic B, et al. Stratum comeum profiles of inflammatory mediators in patch test reactions to common contact allergens and sodium lauryl sulfate. Br. J. Dermatol.2017;176(6): 1533-40

[0113] Lim HW, Collins SAB, Resneck JS, Bolognia JL, Hodge JA, Rohrer TA, et al. The burden of skin disease in the United States. J. Am. Acad. Dermatol. 2017;76(5):958- 972. e2

[0114] Meller S, Lauerma Al, Kopp FM, Winterberg F, Anthoni M, Muller A, et al.Chemokine responses distinguish chemical-induced allergic from irritant skin inflammation: Memory T cells make the difference. J. Allergy Clin. Immunol. 2007; 119(6): 1470-80

[0115] Mendel I, Shevach EM. Activated T cells express the 0X40 ligand: requirements for induction and costimulatory function. Immunology. 2006; 117(2): 196-204

[0116] Nethercott JR. Practical problems in the use of patch testing in the evaluation of patients with contact dermatitis. Curr. Probl. Dermatol. 1990. p. 97-123

[0117] Pfoertner S, Jeron A, Probst-Kepper M, Guzman CA, Hansen W, Westendorf AM, et al. Signatures of human regulatory T cells: an encounter with old friends and new players. Genome Biol. 2006;7(7):R54

[0118] Ravishankar A, Freese RL, Parsons HM, Warshaw EM, Goldfarb N. Trends in Patch Testing in the Medicare Part B Fee-for-Service Population. Dermatitis.2022;33(2): 129-34

[0119] Raymondi Silva J, Iftinca M, Fernandes Gomes FI, Segal JP, Smith OMA, Bannerman CA, et al. Skin-resident dendritic cells mediate postoperative pain via CCR4 on sensory neurons. Proc. Natl. Acad. Sci. 2022;l 19(4):e2118238119

[0120] Robinson MD, Oshiack A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010;l l(3):R25

[0121] Rouzaire P, Luci C, Blasco E, Bienvenu J, Walzer T, Nicolas J-F, et al. Natural killer cells and T cells induce different types of skin reactions during recall responses to haptens: Cellular immune response. Eur. J. Immunol. 2012;42(l):80-8

[0122] Sofen H, Bissonnette R, Yosipovitch G, Silverberg JI, Tyring S, Loo WJ, et al. Efficacy and safety of vixarelimab, a human monoclonal oncostatin M receptor P antibody, in moderate-to-severe prurigo nodularis: a randomised, double-blind, placebo-controlled, phase 2a study. eClinicalMedicine. 2023;57: 101826

[0123] Takamori A, Nambu A, Sato K, Yamaguchi S, Matsuda K, Numata T, et al. IL-31 is crucial for induction of pruritus, but not inflammation, in contact hypersensitivity. Sci.Rep. 2018;8(l):6639

[0124] Ulriksdotter J, Svedman C, Bruze M, Mowitz M. Allergic contact dermatitis caused by dipropylene glycol diacrylate in the Omnipod® insulin pump*. Br. J. Dermatol. 2022;186(2):334-40

[0125] van den Boom JG, Jakobs C, Hagen C, Renn M, Luiten RM, Melief CJM, et al. Inflammasome-Dependent Induction of Adaptive NK Cell Memory. Immunity.2016;44(6): 1406-21

[0126] Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. OMICS J. Integr. Biol. 2012;16(5):284-7

[0127] Yukselen O, Turkyilmaz O, Ozturk AR, Garber M, Kucukural A. DolphinNext: a distributed data processing platform for high throughput genomics. BMC Genomics.2020;21(l):310

[0128] Zilionis R, Nainys J, Veres A, Savova V, Zemmour D, Klein AM, et al. Singlecell barcoding and sequencing using droplet microfluidics. Nat. Protoc. 2017;12(l):44-73Example 2. Single-Cell RNA Sequencing Reveals Molecular Signatures that Distinguish Allergic from Irritant Contact Dermatitis

[0129] The study described in this Example 2 is further described in Frisoli, M. L., Ko, W. C. C., Martinez, N., Afshari, K., Wang, Y., Garber, M., & Harris, J. E. (2024), Single-Cell RNA Sequencing Reveals Molecular Signatures that Distinguish Allergic from Irritant Contact Dermatitis, Journal of Investigative Dermatology, the entire contents of which are incorporated herein by reference.

[0130] A protocol to systematically profile contact dermatitis

[0131] To minimize variability between allergic and irritant reactions, we induced ACD in healthy human volunteers using squaric acid dibutyl ester (SADBE) after prior sensitization (Noster et al., 1976) and ICD using sodium lauryl sulfate (Geier et al., 2003) (Materials and Methods provide the details). We applied each chemical under occlusion for 2 days, in accordance with standard clinical patch testing, as well as placement of acetone as a vehicle control. Skin was sampled using a suction blister biopsy approach (FIG. 1 A). The cellular portion of suction blister fluid was analyzed by scRNAseq using a custom made inDrop device as described previously (Gellatly et al., 2021; Zilionis et al., 2017), whereas the noncellular fluid was analyzed by Olink proximity extension assay (Assarsson et al., 2014) for protein quantification. To longitudinally sample allergic inflammation, 2 separate SADBE patches were placed 2 days apart, on day-4 and day-2, and all sites were sampled on day 0. Skin reactions were photographed and scored according to the International Contact Dermatitis Research Group’s clinical patch test scoring metric (FIGs. IB, 1G, and 1H). Notably, strong (vesicular) allergic contact reactions could not be sampled by suction blisters owing to impaired epidermal integrity, so we aimed to induce mild reactions that were characterized by erythema with minimal vesicles that could be reliably sampled.

[0132] Skin interstitial fluid proteins distinguish ACD from ICD

[0133] We first measured 92 inflammatory proteins in the suction blister fluid by Olink proteomic proximity extension assay (Table 1). Most of these proteins were detectable in suction blister fluid samples, and several were upregulated in both allergy and irritant samples relative to those in healthy control (FIG. 4G). We observed mixed inflammation in ACD, which was characterized by type 1 immunity (T helper 1) (IFNG, CXCL9, CXCL10, CXCL1 1), and type 2 immunity (T helper 2) (IL-4, IL- 5, and IL-13). Markers of both pathways were significantly increased in ACD relative to those in ICD (FIG. 4B). CCL20 and CXCL1 were significantly increased in ICD relative to the expression in ACD, consistent with previous literature (Koppes et al., 2017). Notably, protein analysis of ACD relative to that of ICD also produced similar results when samples were limited to those with positive patch test scores (FIG. 4H). Principal component analysis of suction blister fluid proteins separated allergic, irritant, and healthy samples, indicating that linear combinations of interstitial fluid proteins can distinguish contact dermatitis samples in similar fashion as previously reported with microarray RNA from skin punch biopsy samples (Fortino et al., 2020). Principal component 2 distinguished between allergic and irritant samples, and proteins with the largest principal component 2 loading included type 1 immune markers forACD and CCL20 and CXCL1 for ICD (FIGs. 4C and D). We then performed LASSO (least absolute shrinkage and selection operator) regularized logistic regression (Friedman etal., 2010) to build a protein-based classifier of ACD. Using 7-fold cross-validation, we found that several linear models using combinations of 8 proteins (CXCL10, IL- 4, OPG, CCL19, CCL20, TSLP, IL- la, and NRTN) could distinguish both day-2 and day-4 allergic samples from nonlesional and irritant samples with as high as 93% sensitivity and 93% specificity (FIGs. 4E-F and 16).

[0134] Cell proportions are similar between ACD and ICD

[0135] We then used scRNAseq to corroborate the protein expression data, to investigate whether differences in cell proportionality distinguish ACD from ICD, and to identify cellular sources of inflammatory signals. Data clustered into distinct groups representing all epidermal cell types, and we identified 25 cell populations (FIG. 1C-D, 6 and 9). Several cell populations, including myeloid- 1 and many CD4+ T-cell populations, were enriched in both irritant and allergic contact lesions relative to those in negative control samples (healthy, nonlesional, and acetone vehicle) (FIG. IE). The proportions of spinous keratinocytes (denoted as KRT-sp) were decreased in both irritant and allergic compared with those in control samples, likely owing to dilution by the presence of more inflammatory cells in inflamed, lesional states. Activated keratinocytes (denoted as KRT-a, which were identified by their expression of keratins 6, 16, and 17) (Freedberg et al.. 2001) were found in increased proportions in day-2 allergy, irritant, and acetone vehicle samples, which suggests an acute response during chemical patch occlusion. Langerhans cells and melanocytes were decreased in day-2 allergy, day-4 allergy, irritant, and acetone vehicle samples, which could be a result of inflammatory dilution or possibly hint at how hypopigmentation could be related to fewer active melanocytes after skin inflammation (Zaynoun etal., 1983).

[0136] Two cell populations (CD8 T cells and cytotoxic NK cells) were significantly increased by 2-fold or greater in ACD compared with those in ICD, and both HSPA6- expressing CD4 T cells (CD4-hspa6) and conventional dendritic cell 1 cells (cDCl) displayed nonsignificant trends toward allergy enrichment (FIG. IF). Notably, allergic reactions tended to be more severe (ie, more erythematous and eczematous) (FIG. IB) than irritant reactions, so it is possible that the reaction severity also influenced recruitment of cell populations.

[0137] Patch application induces downregulation of many homeostatic genes

[0138] For cell-type-specific transcriptomic analysis, we merged the 25 cell-type clusters into 9 clusters on the basis of cell lineage (eg, CD4+ T-cell subclusters were merged into 1CD4+ T-cell cluster) to reach greater statistical power while preserving major distinctions among cell types (FIG. 9). Day-4 allergy, day-2 allergy, and irritant samples were each compared with healthy / nonlesional skin, and differentially expressed genes (DEGs) were then clustered by k-means to identify groups of genes with similar expression patterns across the 5 skin conditions. Across multiple cell types, we noticed that many DEGs followed 3 patterns of expression corresponding to downregulated homeostatic genes, common upregulated genes induced within both irritant and allergic lesions, and allergy-specific upregulated genes (FIG. 2A). Full DEGs heatmaps also show these 3 patterns and 1 unclear / complex pattern across irritant and allergic states (FIG. 10A). The number of DEGs across cell types did not correlate with total cell numbers or transcripts detected per cell, suggesting that cell numbers and sequencing depth did not substantially influence this DE analysis (FIG. 10B). Among downregulated homeostatic genes, we found that melanocytes downregulated the growth receptor KIT and the adhesion gene LI CAM, suggesting that a biological effect of patch application may contribute to the decreased melanocyte number in day-2 allergy, day-4 allergy, irritant, and acetone samples, which represent all samples that underwent patch application (FIG. 11C). CXCL14, which is known to antagonize Langerhans cell exit from the skin, was also downregulated in several cell types, consistent with the observed decrease in Langerhans cell proportion within patch application samples (FIG. 17). Other downregulated homeostatic genes were difficult to interpret because we merged cell subclusters for DE analysis. For example, CCL22 appeared to be downregulated in LYZ+ myeloid cells, yet this was due to dilution of the CCL22-producing cellular subpopulation within the merged LYZ+ myeloid cluster rather than a change in CCL22 expression on a percell basis (FIGs. 12A-C). Therefore, the analysis initially focused primarily on upregulated DEGs.

[0139] ICD and ACD both induce an activated keratinocyte response

[0140] Many DEGs were upregulated in both ICD and ACD, and 1 clear signature within this category resembled an epidermal activated response. Cells classified as activated keratinocytes were found in increased proportion within both day-2 irritant and day-2 allergic patch reactions, and keratinization genes associated with an activated response were significantly upregulated by keratinocytes (KRT6A, KRT6B, KRT16, and SPRR1B) (FIG. 2A). DEGs within LYZ+ myeloid cells also included CCL17 and CXCL8 (IL-8) (FIG. 17), which have been associated with epidermal injury (Rennekampff et al., 2000; Silva et al.,Allergic-specific transcriptomic responses are associated with IFNG response

[0141] Allergy-specific induced genes were found in every major cell type we annotated except for Langerhans cells. We used gene ontology enrichment analysis to characterize these genes for each cell type (FIG. 2A), which enriched for several different ontology terms. To characterize the degree of cell specificity, we performed gene set enrichment analysis for each allergy-associated ontology term using data from day -2 allergic and irritant lesions for all cell types (FIG. 2B). In day-2 allergic lesions relative to irritant lesions, gene sets related to cellular stress were elevated in several immune cell types; the response to IFNg gene set was enriched in most cell types; and cytokine-associated gene sets were enriched in keratinocytes and LYZ+ myeloid cells. Consistent with enrichment of the response to IFNg gene set within allergy samples, we observed allergy-specific upregulation of CXCL9, 10, and 11 by keratinocytes and LYZ+ myeloid cells. Furthermore, pseudobulk total IFNG was elevated in allergy samples (FIG. 13 A). All lymphocyte populations were sources of IFNG within allergic lesions, yet CD8+ T cells and NK cells contributed the most toward total pseudobulk IFNG (FIG. 2C and FIG. 13B).

[0142] IL-4 and IL- 13 exhibit different expression patterns in ACD and ICD

[0143] We then focused on the cellular expression of the type 2 immunity (T helper 2) cytokines IL4 and IL 13 because these were observed in the protein data and have also been associated with allergy (Bitton et al., 2020; Dieli et al., 1999). IL4 was found in very few cells, hindering statistically significant conclusions, yet it was specifically upregulated in allergy lesions (FIG. 2C and FIG. 13 A). IL4 was detected predominantly within CD4+ lymphocytes, including T regulatory cells, at day 2 of allergy as well as within CD8+ lymphocytes at day 4 of allergy. IL13 was expressed at higher levels than IL4, and in contrast to IL4, it was upregulated in both irritant and allergic samples with broad cellular sources that included CD4+ T cells, CD8+ T cells, and NK cells (FIG. 2C and FIG. 13B). Nevertheless, the detection levels of both cytokines were too low to attain statistical significance. The question remained as to whether the observed upregulation in both protein and RNA levels results in downstream signaling in cells expressing their receptors.

[0144] Predicting IL-4 and IL-13 presence using downstream gene expression analysis of contact dermatitis

[0145] To investigate the potential effect of IL-4 and IL-13 on skin cells, we sought to measure the expression of genes downstream of their known cytokine signaling cascades. We used nichenetr R package (Browaeys et al., 2020) to predict ligands most likely to affect geneexpression in ACD and ICD, and we found that IFNG, IL-4, and IL- 13 were all among top predicted ligands for LYZ+ myeloid cells and Langerhans cells (FIGs. 18A-B and FIGs. 31- L). Nichenet uses a prior model of ligand-target signaling pathways and combines it with observed cell-type-specific gene expression data to predict ligand-receptor interactions that may be associated with observed differential gene expression data. Consistent with signaling predictions, patient pseudobulk analysis found that several putative IL-4 / IL-13 target genes were significantly increased within LYZ+ myeloid cells and Langerhans cells in allergic lesions (FIG. 18C). Taken together, the increased protein concentrations of IL-4 and IL- 13 in ACD blister fluid, the increased detection of their mRNA in ACD, and the significant upregulation of their downstream target genes in ACD suggest that these cytokines contribute to ACD pathogenesis along with IFNG.

[0146] Discussion

[0147] Because it is currently challenging to differentiate between ACD and ICD on the basis of physical examination and histopathologic analysis, a molecular test to classify skin patch test reactions as ACD, ICD, or nonreactive on the basis of biomarker expression patterns would potentially transform clinical diagnostics. The strength of this study lies in the use of a well-known allergen in dermatology, SADBE, along with a prototypical irritant, sodium lauryl sulfate. We induced suction blisters, of which the collected blister fluid represents interstitial fluid in the skin. We discovered that unique signatures exist for ACD and ICD among soluble interstitial fluid proteins, and we used scRNAseq to corroborate proteomic findings and identify cellular sources. Because there is no further digestion step needed, in contrast to analysis of punch biopsy samples, the suction blistering protocol described herein is beneficial in research settings where unperturbed cells and blister fluid can be directly used in downstream assays.

[0148] Because we were able to differentiate ACD from ICD through RNA and protein signatures, this opens the possibility of diagnosing ACD and ICD through molecular diagnostics. Although people will not inadvertently encounter SADBE in the real world (which is a strength of the use of SADBE in research settings), there is likely considerable overlap of signatures across allergens that are encountered in daily living. Evidence suggests that many different allergens, including metals, fragrances, and rubber accelerators, induce immune responses with shared features as well as unique allergen-dependent features (Dhingra et al.. 2014; Fortino et al.. 2020). A classifier trained on the proteomics results in this study was able to identify SADBE-induced ACD with 93% sensitivity and 93%specificity. In addition to diagnostic improvements, this study has also revealed DEGs not only across diseases states but also at single-cell resolution. This aids in understanding the mechanisms of contact dermatitis with greater detail, which could inspire targeted treatments for ACD or ICD in the way that targeted biologies have become standard of care for atopic dermatitis or psoriasis. Currently, the only cure for ACD is allergen avoidance. However, this may not be possible owing to widespread allergen abundance, occupational hazard, or medical necessity. Many inflammatory skin diseases have been successfully treated by targeted blockade of selective inflammatory pathways; however, given the scRNAseq and protein quantification data, this strategy may be exceptionally difficult for prevention or treatment of ACD. Dupilumab (anti-IL-4 receptor antagonist) has recently been tested by several groups as a treatment for ACD, and whereas some report that dupilumab treatment accelerates resolution of contact allergy skin lesions, others report that allergic patch test reactions still occur in patients taking dupilumab (Bhatia et al., 2020; De Wijs et al., 2021; Goldminz and Scheinman, 2018; Joshi and Khan, 2018; Maehler et al., 2019). This may be due to the presence of multiple immune pathways in SADBE-induced ACD, which include type 1 (CXCL10, CXCL11, IFNG), type 2 (IL4, IL13), and innate (IL8, CCL17, CCL19, CCL20) immune cytokines as the scRNAseq data show (FIG. 2A), and thus blockade of only 1 pathway may be insufficient and perhaps may even elicit a compensatory mechanism of other pathways. Through the use of SADBE, scRNAseq, and functional perturbation of specific immune pathways, we can explore how these multiple pathways interact and discover which pathways or combinations of pathways might be exploitable as therapeutic treatments for contact dermatitis.MATERIALS AND METHODSStudy design

[0149] Healthy individuals without a history of autoimmune or inflammatory skin disease were recruited with written, informed consent at the University of Massachusetts Chan Medical School under institutional review board-approved protocols (H-3947, H-14848). Under protocol H-14848, 12 volunteers donated healthy suction blister skin biopsy, and under protocol H-3947, 10 volunteers were recruited for chemical patch induction of contact dermatitis (as described below) with subsequent suction blister biopsy. Suction blister biopsy samples were processed for scRNAseq and Olink proteomic analysis. Two participants developed a visible allergic-like reaction to the acetone vehicle patch, which may indicatecross-contamination with SADBE patch preparation. Data from these visible reactions to acetone vehicle patches were excluded from all analyses. Table 1H provides patient and sample details.Chemical patch induction ofICD and AC D

[0150] ICD was induced using 2% sodium lauryl sulfate (product number 1614363, Sigma- Aldrich, Burlington, MA) in sterile cell culture grade water. A total of 35 ml of 2% sodium lauryl sulfate was placed onto 8-mm Finn Chambers with filter paper (SmartPractice Canada, Calgary, Canada), applied to volunteer skin, and occluded by Tegaderm dressing (3M, Maplewood, MN). ACD was induced using multiple concentrations of SADBE in acetone (Boulevard Compounding Pharmacy, Worcester, MA), which was applied directly to patient skin using a micropipette with subsequent occlusion under by Tegaderm dressing. Volunteers were first sensitized to SADBE by application of 100 ml of 2% SADBE to the inner arm, which was covered by Tegaderm dressing for 24 hours. To identify each participant’s optimal elicitation dose, SADBE doses ranging between 0.0001 and 2% were applied to participant skin in a volume of 20 ml under Tegaderm dressing for 24 hours to identify the minimal dose that induced a mild erythematous skin reaction. After a rest period >14 days from any previous exposures to SADBE, ACD was induced by application of 20 ml of each participant’s optimal SADBE dose under Tegaderm dressing for 2 days. The only exception to the 14-day SADBE rest period was for serial induction of contact dermatitis as depicted in FIG. 1 A by application of SADBE patches at day -4 and day -2.Suction blister skin sampling

[0151] Suction blisters were induced using the Negative Pressure Instrument Model NP-4 (Electronic Diversities, Finksburg, MD), with a negative pressure between 7 and 10 mmHg until blisters formed. Blister fluid was aspirated using a 1.0-ml insulin syringe, transferred to a collection tube, and centrifuged at 4°C at 350g for 10 minutes to pellet cells. Blister fluid supernatant was snap frozen at -80°C and stored for proteomic analysis. Pelleted blister fluid cells were resuspended in 100 ml of PBS with 15% Axis-shield OptiPrep density gradient medium for scRNAseq by InDrop. scRNAseq

[0152] scRNAseq was performed using the inDrop method as previously described (Gellatly et al., 2021; Zilionis et al., 2017). Single-cell RNA libraries were sequenced onIllumina NextSeq550, NextSeq2000, and NovaSeq platforms. Raw FASTQ files were processed using the DolphinNext pipeline (Yukselen et al., 2020) using GRCh38 as the reference genome, to generate final count matrices containing unique molecular identifier gene counts for each barcoded droplet (Table 2) that was then analyzed in R (version 4.2.1) (R Core Team, 2021) and Seurat R package (version 4.3.0) (Hao et al., 2021).Table 2. Summary scRNAseq Statistics per Patient.Summary information for single cell RNA-sequencing data for each patient, including the number of cells within the final filtered dataset, total unique molecular identifiers (UMIs), total gene count, mean UMIs per cell, and mean genes per cell. Note that CB-pattem patient IDs represent healthy control suction blister samples, and P-pattern patient IDs represent patch test patients. scRNAseq data analysis

[0153] Count matrices for all suction blister skin samples were filtered for barcodes containing 300 or more unique molecular identifiers and 150 or more unique genes and whose mitochondrial gene expression did not exceed 7.5% of total droplet unique molecular identifiers. Data were then processed using Seurat’s SCTransform normalization workflow, followed by dimension reduction, clustering, and 2-dimensional Uniform Manifold Approximation and Projection reduction of the dataset for further visual analysis. Five broad subsets of cells were detected with the FindCluster function (Louvain algorithm) at resolution 0.5, which corresponded to keratinocytes, melanocytes, antigen-presenting cells,lymphocytes, and erythrocytes on the basis of marker genes generated by the Find AllMarkers function. Marker genes for each broad subset were defined as follows: keratinocyte (KRT1 / 2 / 5 / 6A / 6B / 10 / 14 / 15 / 17 / 77, KRTDAP, DMKN, MUCL1), melanocyte (TYR, MLANA, PMEL, DCT, MITF), antigen-presenting cell (HLA.DR genes, LYZ, CD207, CD86, CD83, CLEC4A, CLEC4C, CD163, MRC1, NRP1, HM0X1, JCHAIN), lymphocyte (TRAC, TRDC, NCAM1), and erythrocyte (HBB, HBA2). Erythrocytes were then excluded, and the remaining unclassified clusters were subclustered for further identification. During subclustering, clusters that were found to possess marker genes from previously identified clusters were labeled as doublets and subsequently filtered. The final dataset after processing is illustrated in FIG. 5A-C. Several clusters were characterized by cell cycle genes such as T0P2A and CDK1, so we excluded 2151 genes that were identified by differential expression analysis to be significantly upregulated in cell clusters with high T0P2A and CDK1 expression. Batch effects were observed in association with 5 different sequencing runs, so we used Seurat’s IntegrateData function prior to principal component analysis followed by clustering and Uniform Manifold Approximation and Projection embedding (FIGs. 5A-C). Top cluster marker genes with highly specific expression were identified for each final cell cluster by differential expression analysis and weight of genes by average log2(fold change) / (percentage of cells outside of cluster with positive gene expression) (FIG. 6). T regulatory cell cluster identity was also validated using gene set enrichment analysis of published FACS-sorted T regulatory cell gene sets (Feuerer et al., 2009; Pfoertner et al., 2006) (FIG. 7).

[0154] Statistical analysis of differential gene expression across lesional skin conditions was performed using EdgeR R package, and gene set enrichment analysis as well as gene ontology enrichment analysis were performed using ClusterProfiler R package (Robinson and Oshiack, 2010; Yu et al., 2012). Ligand prediction analysis was performed with nichenetr (Browaeys et al., 2020). Several gene expression plots, including violin plots and heatmaps, were created by modified functions of the SignallingSingleCell R package (Gellatly et al., 2021). Cell proportion analysis was conducted using samples with >100 cells in the final scRNAseq dataset to minimize variance associated with small cellular sample sizes. Skin interstitial fluid proteomic analysis

[0155] Suction blister fluid supernatant was analyzed with Olink Target 96 Inflammation panel (Table 1), and the resulting Olink normalized protein expression values were analyzed in R. Differential expression analysis for proteomic readout across conditions was performedusing the pairwise t-test function from the rstatix R package (Kassambara, 2023) with assumed unequal variances and Benjamini -Hochberg correction for multiple hypothesis testing. Principal component analysis was performed using all normalized protein expression protein values. LASSO regression analysis using glmnet R package (Friedman et cd.. 2010) with 7-fold cross-validation was performed using a logistic model to classify samples as allergic or nonallergic by protein normalized protein expression values. Multiple lambda penalty values were tested, and models were assessed by the area under the receiver operating characteristic curve. The model with the least misclassification error had 7 nonzero protein coefficients, and for comparison, other models with fewer nonzero protein coefficients that were derived from larger lambda penalty values were also evaluated (FIG. 16).

[0156] Example 2 References

[0157] Assarsson E, Lundberg M, Holmquist G, Bjo "rkesten J, Thorsen SB, Ekman D, et al. Homogenous 96-plex PEA immunoassay exhibiting high sensitivity, specificity, and excellent scalability. PLoS One 2014;9:e95192.

[0158] Bahamon I, Strassner JP, Ko WCC. Toward a molecular diagnosis: looking under the skin at allergic contact dermatitis. Curr Derm Rep 2023;12:245e53.

[0159] Bhatia J, Sarin A, Wollina U, Lotti T, Navarini AA, Mueller SM, et al. Review of biologies in allergic contact dermatitis. Contact Dermatitis 2020;83: 179e81.

[0160] Bitton A, Avlas S, Reichman H, Itan M, Karo-Atar D, Azouz NP, et al. A key role for IL- 13 signaling via the type 2 IL-4 receptor in experimental atopic dermatitis. Sci Immunol 2020;5:eaaw2938.

[0161] Browaeys R, Saelens W, Saeys Y. NicheNet: modeling intercellular communication by linking ligands to target genes. Nat Methods 2020;17: 159e62.

[0162] De Wijs LEM, Van Der Waa JD, Nijsten T, Silverberg JI, Kunkeler ACM, Hijnen DJ. Effects of dupilumab treatment on patch test reactions: a retrospective evaluation. Clin Exp Allergy 2021;51:959e67.

[0163] Dhingra N, Shemer A, Correa Da Rosa J, Rozenblit M, Fuente s-Ducul an J, Gittler JK, et al. Molecular profiling of contact dermatitis skin identifies allergen-dependent differences in immune response. J Allergy Clin Immunol 2014;134:362e72.

[0164] Dieli F, Sireci G, Scire' E, Salerno A, Bellavia A. Impaired contact hypersensitivity to trinitrochlorobenzene in interleukin-4-deficient mice. Immunology 1999;98:71e9.

[0165] Fartasch M. Ultrastructure of the epidermal barrier after irritation. Microsc Res Tech 1997;37: 193e9.

[0166] Feuerer M, Hill JA, Mathis D, Benoist C. Foxp3+ regulatory T cells: differentiation, specification, subphenotypes. Nat Immunol 2009;10:689e95.

[0167] Fortino V, Wisgrill L, Werner P, Suomela S, Linder N, Jalonen E, et al. Machinelearning-driven biomarker discovery for the discrimination between allergic and irritant contact dermatitis. Proc Natl Acad Sci USA 2020;l 17: 33474e85.

[0168] Freedberg IM, Tomic-Canic M, Komine M, Blumenberg M. Keratins and the keratinocyte activation cycle. J Invest Dermatol 2001;116:633e40.

[0169] Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw 2010;33: le22.

[0170] Geier J, Uter W, Pirker C, Frosch PJ. Patch testing with the irritant sodium lauryl sulfate (SLS) is useful in interpreting weak reactions to contact allergens as allergic or irritant. Contact Dermatitis 2003;48:99el07.

[0171] Gellatly KJ, Strassner JP, Essien K, Refat MA, Murphy RL, Coffin-Schmitt A, et al. scRNA-seq of human vitiligo reveals complex networks of subclinical immune activation and a role for CCR5 in T reg function. Sci Transl Med 2021;13:eabd8995.

[0172] Goldminz AM, Scheinman PL. A case series of dupilumab-treated allergic contact dermatitis patients. Dermatol Ther 2018;31 :el2701.

[0173] Goulden V, Wilkinson SM. Evaluation. Clin Exp Dermatol 2000;25:67e70.

[0174] Hao Y, Hao S, Andersen-Nissen E, Mauck WM, Zheng S, Butler A, et al. Integrated analysis of multimodal single-cell data. Cell 2021;184:3573e87.e29.

[0175] Joshi SR, Khan DA. Effective use of dupilumab in managing systemic allergic contact dermatitis. Dermatitis 2018;29:282e4.

[0176] Kaplan DH, Igya'rto' BZ, Gaspari AA. Early immune events in the induction of allergic contact dermatitis. Nat Rev Immunol 2012; 12: 114e24.

[0177] Kassambara A. rstatix: pipe-friendly framework for basic statistical tests, rstatix, https: / / rpkgs.datanovia.com / rstatix / . (accessed May 1, 2023).

[0178] Koppes SA, Ljubojevic Hadzavdic S, Jakasa I, Franceschi N, Juraki c Ton ci c R,

[0179] Marinovi c B, et al. Stratum comeum profiles of inflammatory mediators in patch test reactions to common contact allergens and sodium lauryl sulfate. Br J Dermatol 2017;176: 1533e40.

[0180] Lefevre MA, Nosbaum A, Rozieres A, Lenief V, Mosnier A, Cortial A, et al.Unique molecular signatures typify skin inflammation induced by chemical allergens and irritants. Allergy 2021;76:3697e712.

[0181] Lim HW, Collins SAB, Resneck JS, Bolognia JL, Hodge JA, Rohrer TA, et al. The burden of skin disease in the United States. J Am Acad Dermatol 2017;76:958e72.e2.

[0182] Maehler BC, Sung CT, Darwin E, Jacob SE. Dupilumab use in allergic contact dermatitis. J Am Acad Dermatol 2019;80:280el.el.

[0183] Meller S, Lauerma Al, Kopp FM, Winterberg F, Anthoni M, Mu ' Iler A, et al.Chemokine responses distinguish chemical-induced allergic from irritant skin inflammation: memory T cells make the difference. J Allergy Clin Immunol 2007;l 19: 1470e80.

[0184] Nethercott JR. Practical problems in the use of patch testing in the evaluation of patients with contact dermatitis. Curr Probl Dermatol 1990;2: 97el23.

[0185] Noster U, Hausen BM, Krische B, Schulz KH. Squaric-acid-diethylester - a strong sensitizer. Contact Dermatitis 1976;2:99el04.

[0186] Pfoertner S, Jeron A, Probst-Kepper M, Guzman CA, Hansen W,

[0187] Westendorf AM, et al. Signatures of human regulatory T cells: an encounter with old friends and new players. Genome Biol 2006;7: R54.

[0188] R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2021. r-project.org. (accessed June 23, 2022).

[0189] Rennekampff HO, Hansbrough JF, Kiessig V, Dore' C, Sticherling M,

[0190] Schro' der JM. Bioactive interleukin-8 is expressed in wounds and enhances wound healing. J Surg Res 2000;93:41e54.

[0191] Robinson MD, Oshiack A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol 2010; 11 : R25.

[0192] Scheinman PL, Vocanson M, Thyssen JP, Johansen JD, Nixon RL, Dear K, et al. Contact dermatitis. Nat Rev Dis Primers 2021;7:38.

[0193] Silva JR, Iftinca M, Fernandes Gomes FI, Segal JP, Smith OMA, Bannerman CA, et al. Skin-resident dendritic cells mediate postoperative pain via CCR4 on sensory neurons. Proc Natl Acad Sci USA 2022; 119 : e2118238119.

[0194] Strassner JP, Rashighi M, Ahmed Refat M, Richmond JM, Harris JE. Suction blistering the lesional skin of vitiligo patients reveals useful biomarkers of disease activity. J Am Acad Dermatol 2017;76:847e55.e5.

[0195] Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16: 284e7.

[0196] Yukselen O, Turkyilmaz O, Ozturk AR, Garber M, Kucukural A. DolphinNext: a distributed data processing platform for high throughput genomics. BMC Genomics 2020;21:310.

[0197] Zaynoun ST, Aftimos BG, Tenekjian KK, Bahuth N, Kurban AK. Extensive pityriasis alba: a histological histochemical and ultrastructural study. Br J Dermatol 1983;108:83e90.

[0198] Zilionis R, Nainys J, Veres A, Savova V, Zemmour D, Klein AM, et al. Singlecell barcoding and sequencing using droplet microfluidics. Nat Protoc 2017;12:44e73.

[0199] Feuerer M, Hill JA, Mathis D, Benoist C. Foxp3+ regulatory T cells: differentiation, specification, subphenotypes. Nat Immunol 2009;10:689e95.

[0200] Pfoertner S, Jeron A, Probst-Kepper M, Guzman CA, Hansen W, Westendorf AM, et al. Signatures of human regulatory T cells: an encounter with old friends and new players. Genome Biol 2006;7:R54.

[0201] The teachings of all patents, published applications and references cited herein are incorporated by reference in their entirety.

[0202] While example embodiments have been particularly shown and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the embodiments encompassed by the appended claims.

Claims

CLAIMSWhat is claimed is:

1. A method of detecting an allergic skin reaction in a subject in need thereof, comprising: a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining a level of one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from Fractalkine (CX3CL1), Sulfotransferase 1A1 (ST1A1), Stem cell factor (SCF), C-C motif chemokine 25 (CCL25), Eotaxin (CCL11), Interleukin-1 alpha (IL-1 alpha), Glial cell line-derived neurotrophic factor (GDNF), Interleukin-7 (IL7), Hepatocyte growth factor (HGF), Tumor necrosis factor (Ligand) superfamily, member 12(TWEAK), Interleukin-2 receptor subunit beta (IL-2RB), Interleukin-20 receptor subunit alpha (IL-20RA), Interleukin- 22 receptor subunit alpha-1 (IL-22 RAI), Interleukin-33 (IL33), Cystatin D (CST5), Adenosine Deaminase (ADA), Fibroblast growth factor 23 (FGF-23), Fibroblast growth factor 21 (FGF-21), C-C motif chemokine 28 (CCL28), CUB domain-containing protein 1 (CDCP1), Axin-1 (AXIN1), Interleukin- 18 (IL 18), Eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), SIR2-like protein 2 (SIRT2), STAM-binding protein (STAMBP), Beta-nerve growth factor (Beta-NGF), Interleukin- 17C (IL-17C), Fms-related tyrosine kinase 3 ligand (Flt3L), C-C motif chemokine 20 (CCL20), Thymic stromal lymphopoietin (TSLP), Artemin (ARTN), Interleukin- 17A (IL- 17 A), Fibroblast growth factor 19 (FGF-19), Macrophage colony-stimulating factor 1 (CSF-1), Monocyte chemotactic protein 1 (MCP-1), C-C motif chemokine 19 (CCL19), Monocyte chemotactic protein 2 (MCP-2), Monocyte chemotactic protein 3 (MCP-3), Interleukin-6 (IL6), Oncostatin-M (OSM), Tumor necrosis factor (TNF), Interleukin-20 (IL-20), Leukemia inhibitory factor (LIF), Interleukin-8 (IL8), C-C motif chemokine 4 (CCL4), C-C motif chemokine 3 (CCL3), Neurotrophin-3 (NT-3), Interleukin-2 (IL2), Interleukin-24 (IL-24), Interleukin- 10 receptor subunit beta (IL- 1 ORB),Interleukin- 18 receptor 1 (IL-18R1), Leukemia inhibitory factor receptor (LIF-R), C-X-C motif chemokine 5 (CXCL5), Delta and Notch-like epidermal growth factor-related receptor (DNER), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 1 (CXCL1), Protein SI 00-Al 2 (EN-RAGE), Latency-associated peptide transforming growth factor beta-1 (LAP TGF-beta-1), CD40L receptor (CD40), Transforming growth factor alpha (TGF-alpha), Caspase-8 (CASP-8), Vascular endothelial growth factor A (VEGFA), Urokinase-type plasminogen activator (uPA), Matrix metalloproteinase- 1 (MMP-1), Matrix metalloproteinase- 10 (MMP-10), Fibroblast growth factor 5 (FGF-5), Interleukin- 10 receptor subunit alpha (IL- 1 ORA), Interleukin- 15 receptor subunit alpha (IL-15RA), Neurturin (NRTN), Interleukin- 12 subunit beta (IL-12B), C-C motif chemokine 23 (CCL23), T-cell surface glycoprotein CD8 alpha chain (CD8A), Natural killer cell receptor 2B4 (CD244), Osteoprotegerin (OPG), TNF-related apoptosisinducing ligand (TRAIL), T cell surface glycoprotein CD6 isoform (CD6), Programmed cell death 1 ligand 1 (PD-L1), TNF-related activation-induced cytokine (TRANCE), T-cell surface glycoprotein CD5 (CD5), Tumor necrosis factor receptor superfamily member 9 (TNFRSF9), TNF-beta (TNFB), C-X-C motif chemokine 9 (CXCL9), Monocyte chemotactic protein 4 (MCP-4), C- X-C motif chemokine 10 (CXCL10), C-X-C motif chemokine 11 (CXCL11), Tumor necrosis factor ligand superfamily member 14 (TNFSF14), Interleukin- 10 (IL10), Interferon gamma (IFNG), Interleukin-4 (IL4), Interleukin- 13 (IL 13), Interleukin-5 (IL5), Signaling lymphocytic activation molecule (SLAMF1), or any combination thereof; and c) identifying the subject as having an allergic skin reaction based on the level of the one or more biomarkers.

2. A method of diagnosing an allergic skin reaction in a subject in need thereof, comprising: a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen;b) determining a level of one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from Fractalkine (CX3CL1), Sulfotransferase 1A1 (ST1A1), Stem cell factor (SCF), C-C motif chemokine 25 (CCL25), Eotaxin (CCL11), Interleukin-1 alpha (IL-1 alpha), Glial cell line-derived neurotrophic factor (GDNF), Interleukin-7 (IL7), Hepatocyte growth factor (HGF), Tumor necrosis factor (Ligand) superfamily, member 12(TWEAK), Interleukin-2 receptor subunit beta (IL-2RB), Interleukin-20 receptor subunit alpha (IL-20RA), Interleukin- 22 receptor subunit alpha-1 (IL-22 RAI), Interleukin-33 (IL33), Cystatin D (CST5), Adenosine Deaminase (ADA), Fibroblast growth factor 23 (FGF-23), Fibroblast growth factor 21 (FGF-21), C-C motif chemokine 28 (CCL28), CUB domain-containing protein 1 (CDCP1), Axin-1 (AXIN1), Interleukin- 18 (IL 18), Eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), SIR2-like protein 2 (SIRT2), STAM-binding protein (STAMBP), Beta-nerve growth factor (Beta-NGF), Interleukin- 17C (IL-17C), Fms-related tyrosine kinase 3 ligand (Flt3L), C-C motif chemokine 20 (CCL20), Thymic stromal lymphopoietin (TSLP), Artemin (ARTN), Interleukin- 17A (IL- 17 A), Fibroblast growth factor 19 (FGF-19), Macrophage colony-stimulating factor 1 (CSF-1), Monocyte chemotactic protein 1 (MCP-1), C-C motif chemokine 19 (CCL19), Monocyte chemotactic protein 2 (MCP-2), Monocyte chemotactic protein 3 (MCP-3), Interleukin-6 (IL6), Oncostatin-M (OSM), Tumor necrosis factor (TNF), Interleukin-20 (IL-20), Leukemia inhibitory factor (LIF), Interleukin-8 (IL8), C-C motif chemokine 4 (CCL4), C-C motif chemokine 3 (CCL3), Neurotrophin-3 (NT-3), Interleukin-2 (IL2), Interleukin-24 (IL-24), Interleukin- 10 receptor subunit beta (IL- 1 ORB), Interleukin- 18 receptor 1 (IL-18R1), Leukemia inhibitory factor receptor (LIF-R), C-X-C motif chemokine 5 (CXCL5), Delta and Notch-like epidermal growth factor-related receptor (DNER), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 1 (CXCL1), Protein SI 00-Al 2 (EN-RAGE), Latency-associated peptide transforming growth factor beta-1 (LAP TGF-beta-1), CD40L receptor (CD40), Transforming growth factor alpha (TGF-alpha), Caspase-8 (CASP-8), Vascular endothelial growth factor A (VEGFA), Urokinase-type plasminogen activator (uPA), Matrixmetalloproteinase- 1 (MMP-1), Matrix metalloproteinase- 10 (MMP-10), Fibroblast growth factor 5 (FGF-5), Interleukin- 10 receptor subunit alpha (IL- 1 ORA), Interleukin- 15 receptor subunit alpha (IL-15RA), Neurturin (NRTN), Interleukin- 12 subunit beta (IL-12B), C-C motif chemokine 23 (CCL23), T-cell surface glycoprotein CD8 alpha chain (CD8A), Natural killer cell receptor 2B4 (CD244), Osteoprotegerin (OPG), TNF-related apoptosisinducing ligand (TRAIL), T cell surface glycoprotein CD6 isoform (CD6), Programmed cell death 1 ligand 1 (PD-L1), TNF-related activation-induced cytokine (TRANCE), T-cell surface glycoprotein CD5 (CD5), Tumor necrosis factor receptor superfamily member 9 (TNFRSF9), TNF-beta (TNFB), C-X-C motif chemokine 9 (CXCL9), Monocyte chemotactic protein 4 (MCP-4), C- X-C motif chemokine 10 (CXCL10), C-X-C motif chemokine 11 (CXCL11), Tumor necrosis factor ligand superfamily member 14 (TNFSF14), Interleukin- 10 (IL10), Interferon gamma (IFNG), Interleukin-4 (IL4), Interleukin- 13 (IL 13), Interleukin-5 (IL5), Signaling lymphocytic activation molecule (SLAMF1), or any combination thereof; and c) identifying the subject as having an allergic skin reaction based on the level of the one or more biomarkers.

3. A method of distinguishing an allergic skin reaction from an irritant skin reaction in a subject in need thereof, comprising: a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a potential allergen; b) determining a level of one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from Fractalkine (CX3CL1), Sulfotransferase 1A1 (ST1A1), Stem cell factor (SCF), C-C motif chemokine 25 (CCL25), Eotaxin (CCL11), Interleukin-1 alpha (IL-1 alpha), Glial cell line-derived neurotrophic factor (GDNF), Interleukin-7 (IL7), Hepatocyte growth factor (HGF), Tumor necrosis factor (Ligand) superfamily, member 12(TWEAK), Interleukin-2 receptor subunit beta (IL-2RB), Interleukin-20 receptor subunit alpha (IL-20RA), Interleukin- 22 receptor subunit alpha-1 (IL-22 RAI), Interleukin-33 (IL33), Cystatin D(CST5), Adenosine Deaminase (ADA), Fibroblast growth factor 23 (FGF-23), Fibroblast growth factor 21 (FGF-21), C-C motif chemokine 28 (CCL28), CUB domain-containing protein 1 (CDCP1), Axin-1 (AXIN1), Interleukin- 18 (IL 18), Eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), SIR2-like protein 2 (SIRT2), STAM-binding protein (STAMBP), Beta-nerve growth factor (Beta-NGF), Interleukin- 17C (IL-17C), Fms-related tyrosine kinase 3 ligand (Flt3L), C-C motif chemokine 20 (CCL20), Thymic stromal lymphopoietin (TSLP), Artemin (ARTN), Interleukin- 17A (IL- 17 A), Fibroblast growth factor 19 (FGF-19), Macrophage colony-stimulating factor 1 (CSF-1), Monocyte chemotactic protein 1 (MCP-1), C-C motif chemokine 19 (CCL19), Monocyte chemotactic protein 2 (MCP-2), Monocyte chemotactic protein 3 (MCP-3), Interleukin-6 (IL6), Oncostatin-M (OSM), Tumor necrosis factor (TNF), Interleukin-20 (IL-20), Leukemia inhibitory factor (LIF), Interleukin-8 (IL8), C-C motif chemokine 4 (CCL4), C-C motif chemokine 3 (CCL3), Neurotrophin-3 (NT-3), Interleukin-2 (IL2), Interleukin-24 (IL-24), Interleukin- 10 receptor subunit beta (IL- 1 ORB), Interleukin- 18 receptor 1 (IL-18R1), Leukemia inhibitory factor receptor (LIF-R), C-X-C motif chemokine 5 (CXCL5), Delta and Notch-like epidermal growth factor-related receptor (DNER), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 1 (CXCL1), Protein SI 00-Al 2 (EN-RAGE), Latency-associated peptide transforming growth factor beta-1 (LAP TGF-beta-1), CD40L receptor (CD40), Transforming growth factor alpha (TGF-alpha), Caspase-8 (CASP-8), Vascular endothelial growth factor A (VEGFA), Urokinase-type plasminogen activator (uPA), Matrix metalloproteinase- 1 (MMP-1), Matrix metalloproteinase- 10 (MMP-10), Fibroblast growth factor 5 (FGF-5), Interleukin- 10 receptor subunit alpha (IL- 1 ORA), Interleukin- 15 receptor subunit alpha (IL-15RA), Neurturin (NRTN), Interleukin- 12 subunit beta (IL-12B), C-C motif chemokine 23 (CCL23), T-cell surface glycoprotein CD8 alpha chain (CD8A), Natural killer cell receptor 2B4 (CD244), Osteoprotegerin (OPG), TNF-related apoptosisinducing ligand (TRAIL), T cell surface glycoprotein CD6 isoform (CD6), Programmed cell death 1 ligand 1 (PD-L1), TNF-related activation-induced cytokine (TRANCE), T-cell surface glycoprotein CD5 (CD5), Tumor necrosisfactor receptor superfamily member 9 (TNFRSF9), TNF-beta (TNFB), C-X-C motif chemokine 9 (CXCL9), Monocyte chemotactic protein 4 (MCP-4), C- X-C motif chemokine 10 (CXCL10), C-X-C motif chemokine 11 (CXCL11), Tumor necrosis factor ligand superfamily member 14 (TNFSF14), Interleukin- 10 (IL10), Interferon gamma (IFNG), Interleukin-4 (IL4), Interleukin- 13 (IL 13), Interleukin-5 (IL5), Signaling lymphocytic activation molecule (SLAMF1), or any combination thereof; and c) identifying the subject as having an allergic skin reaction or an irritant skin reaction based on the level of the one or more biomarkers.

4. The method of claim 1, 2 or 3, comprising comparing the level (e.g., expression level) of one or more protein biomarkers in the sample from the subject to a control, wherein an increase or decrease in the level of the biomarker compared to the control is indicative of either an allergic skin reaction or an irritant skin reaction.

5. The method of any one of claims 1-4, comprising determining the relative amount of one or more protein biomarkers selected from CX3CL1, STI Al, SCF, CCL25, CCL11, IL-1 alpha, GDNF, IL7, HGF, TWEAK, IL-2RB, IL-20RA, IL-22 RAI, IL33, CST5, ADA, FGF-23, FGF-21, CCL28, CDCP1, AXIN1, IL18, 4E-BP1, SIRT2, STAMBP, Beta-NGF, IL-17C, Flt3L, CCL20, TSLP, ARTN, IL-17A, FGF-19, CSF-1, MCP-1, CCL19, MCP-2, MCP-3, IL6, OSM, TNF, IL-20, LIF, IL8, CCL4, CCL3, NT-3, IL2, IL-24, IL-10RB, IL-18R1, LIF-R, CXCL5, DNER, CXCL6, CXCL1, EN-RAGE, LAP TGF-beta-1, CD40, TGF-alpha, C ASP-8, VEGFA, uPA, MMP-1, MMP-10, FGF-5, IL-10RA, IL-15RA, NRTN, IL-12B, CCL23, CD8A, CD244, OPG, TRAIL, CD6, PD-L1, TRANCE, CD5, TNFRSF9, TNFB, CXCL9, MCP-4, CXCL10, CXCL11, TNFSF14, IL10, IFNG, IL4, IL13, IL5, SLAMF1, or any combination thereof.

6. The method of claim 5, wherein the relative amount of protein is determined by calculating a ratio of the expression levels of two or more of the protein biomarkers, wherein the ratio is a ratio of the level (e.g., expression level) of one or more allergenspecific protein biomarkers (numerator) to the level (e.g., expression level) of one or more irritant-specific protein biomarkers (denominator).

7. The method of claim 5, wherein the relative amount of protein is determined by a multi-linear logistic regression analysis.

8. The method of claim 7, wherein the multi-linear logistic regression analysis uses a formula that assigns a positive coefficient multiplier to the expression level of an allergen-specific protein biomarker, and a negative coefficient multiplier to the expression level of an irritant-specific protein biomarker.

9. The method of any of the preceding claims, wherein the skin sample comprises skin interstitial fluid.

10. The method of any of the preceding claims, wherein the skin sample is a suction blister sample, an absorptive microneedle skin sample, a tape strip sample, a shave biopsy sample, or a punch biopsy skin sample.

11. The method of any of the preceding claims, wherein the potential allergen was applied to the skin of the subject using an allergen patch test, a tape strip, or a microneedle device.

12. The method of any of the preceding claims, wherein the skin sample was obtained within about seven days after contacting the skin with the potential allergen.

13. The method of claim 12, wherein the skin sample was obtained from about two to about four days after contacting the skin with the potential allergen.

14. The method of any of the preceding claims, wherein the allergic skin reaction is allergic contact dermatitis.

15. The method of claim 3, wherein the irritant skin reaction is irritant contact dermatitis.

16. A method of preparing a skin sample that is useful for detecting an allergic skin reaction in a subject, comprising: a) applying a potential allergen to an area of skin of a subject; b) obtaining a skin sample from the area of skin to which the allergen has been applied; and c) combining the sample with one or more reagents for detecting one or more protein biomarkers selected from Fractalkine (CX3CL1), Sulfotransferase 1 Al(ST1A1), Stem cell factor (SCF), C-C motif chemokine 25 (CCL25), Eotaxin (CCL11), Interleukin-1 alpha (IL-1 alpha), Glial cell line-derived neurotrophic factor (GDNF), Interleukin-7 (IL7), Hepatocyte growth factor (HGF), Tumor necrosis factor (Ligand) superfamily, member 12(TWEAK), Interleukin-2 receptor subunit beta (IL-2RB), Interleukin-20 receptor subunit alpha (IL-2 ORA), Interleukin-22 receptor subunit alpha- 1 (IL-22 RAI), Interleukin-33 (IL33), Cystatin D (CST5), Adenosine Deaminase (ADA), Fibroblast growth factor 23 (FGF-23), Fibroblast growth factor 21 (FGF-21), C-C motif chemokine 28 (CCL28), CUB domain-containing protein 1 (CDCP1), Axin-1 (AXIN1), Interleukin- 18 (IL 18), Eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), SIR2-like protein 2 (SIRT2), STAM-binding protein (STAMBP), Beta-nerve growth factor (Beta-NGF), Interleukin-17C (IL-17C), Fms-related tyrosine kinase 3 ligand (Flt3L), C-C motif chemokine 20 (CCL20), Thymic stromal lymphopoietin (TSLP), Artemin (ARTN), Interleukin- 17A (IL-17A), Fibroblast growth factor 19 (FGF-19), Macrophage colony-stimulating factor 1 (CSF-1), Monocyte chemotactic protein 1 (MCP-1), C-C motif chemokine 19 (CCL19), Monocyte chemotactic protein 2 (MCP-2), Monocyte chemotactic protein 3 (MCP-3), Interleukin-6 (IL6), Oncostatin-M (OSM), Tumor necrosis factor (TNF), Interleukin-20 (IL-20), Leukemia inhibitory factor (LIF), Interleukin-8 (IL8), C-C motif chemokine 4 (CCL4), C-C motif chemokine 3 (CCL3), Neurotrophin-3 (NT-3), Interleukin-2 (IL2), Interleukin-24 (IL-24), Interleukin- 10 receptor subunit beta (IL- 1 ORB), Interleukin- 18 receptor 1 (IL-18R1), Leukemia inhibitory factor receptor (LIF-R), C-X-C motif chemokine 5 (CXCL5), Delta and Notch-like epidermal growth factor-related receptor (DNER), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 1 (CXCL1), Protein S100-A12 (EN-RAGE), Latency-associated peptide transforming growth factor beta-1 (LAP TGF-beta-1), CD40L receptor (CD40), Transforming growth factor alpha (TGF-alpha), Caspase-8 (CASP-8), Vascular endothelial growth factor A (VEGFA), Urokinase-type plasminogen activator (uPA), Matrix metalloproteinase- 1 (MMP-1), Matrix metalloproteinase- 10 (MMP-10), Fibroblast growth factor 5 (FGF-5), Interleukin- 10 receptor subunit alpha (IL- 1 ORA), Interleukin- 15 receptorsubunit alpha (IL-15RA), Neurturin (NRTN), Interleukin- 12 subunit beta (IL-12B), C-C motif chemokine 23 (CCL23), T-cell surface glycoprotein CD8 alpha chain (CD8A), Natural killer cell receptor 2B4 (CD244), Osteoprotegerin (OPG), TNF-related apoptosis-inducing ligand (TRAIL), T cell surface glycoprotein CD6 isoform (CD6), Programmed cell death 1 ligand 1 (PD-L1), TNF-related activation-induced cytokine (TRANCE), T-cell surface glycoprotein CD5 (CD5), Tumor necrosis factor receptor superfamily member 9 (TNFRSF9), TNF-beta (TNFB), C-X-C motif chemokine 9 (CXCL9), Monocyte chemotactic protein 4 (MCP-4), C-X-C motif chemokine 10 (CXCL10), C-X-C motif chemokine 11 (CXCL11), Tumor necrosis factor ligand superfamily member 14 (TNFSF14), Interleukin- 10 (IL10), Interferon gamma (IFNG), Interleukin-4 (IL4), Interleukin- 13 (IL13), Interleukin-5 (IL5), Signaling lymphocytic activation molecule (SLAMF1), or any combination thereof.

17. The method of claim 16, wherein the potential allergen is applied to the skin of the subject using an allergen patch test, a tape strip, or a microneedle device.

18. The method of claim 16 or 17, wherein the skin sample is obtained by suction blister sampling, absorptive microneedle skin sampling, or punch biopsy skin sampling.

19. The method of claim 16, 17, or 18, wherein the skin sample is obtained within about seven days after applying the potential allergen.

20. The method of claim 19, wherein the skin sample was obtained from about two to about four days after applying the potential allergen.

21. The method of any one of claims 16-20, wherein the skin sample comprises skin interstitial fluid.

22. The method of any one of claims 16-21, wherein the allergic skin reaction is allergic contact dermatitis.

23. The method of any one of claims 16-22, wherein the one or more reagents for detecting protein biomarkers is an antibody, a nucleic acid or a small molecule, or a combination thereof.

24. The method of any one of claims 16-23, wherein the one or more reagents are immobilized.

25. A method of identifying a substance as a skin allergen or a skin irritant for a subject, comprising: a) providing a skin sample from the subject, wherein the skin sample has been obtained from an area of skin of the subject that has been contacted with a test substance; b) determining a level of one or more protein biomarkers in the sample from the subject, wherein the one or more protein biomarkers are selected from Fractalkine (CX3CL1), Sulfotransferase 1A1 (ST1A1), Stem cell factor (SCF), C-C motif chemokine 25 (CCL25), Eotaxin (CCL11), Interleukin-1 alpha (IL-1 alpha), Glial cell line-derived neurotrophic factor (GDNF), Interleukin-7 (IL7), Hepatocyte growth factor (HGF), Tumor necrosis factor (Ligand) superfamily, member 12(TWEAK), Interleukin-2 receptor subunit beta (IL-2RB), Interleukin-20 receptor subunit alpha (IL-20RA), Interleukin- 22 receptor subunit alpha-1 (IL-22 RAI), Interleukin-33 (IL33), Cystatin D (CST5), Adenosine Deaminase (ADA), Fibroblast growth factor 23 (FGF-23), Fibroblast growth factor 21 (FGF-21), C-C motif chemokine 28 (CCL28), CUB domain-containing protein 1 (CDCP1), Axin-1 (AXIN1), Interleukin- 18 (IL 18), Eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), SIR2-like protein 2 (SIRT2), STAM-binding protein (STAMBP), Beta-nerve growth factor (Beta-NGF), Interleukin- 17C (IL-17C), Fms-related tyrosine kinase 3 ligand (Flt3L), C-C motif chemokine 20 (CCL20), Thymic stromal lymphopoietin (TSLP), Artemin (ARTN), Interleukin- 17A (IL- 17 A), Fibroblast growth factor 19 (FGF-19), Macrophage colony-stimulating factor 1 (CSF-1), Monocyte chemotactic protein 1 (MCP-1), C-C motif chemokine 19 (CCL19), Monocyte chemotactic protein 2 (MCP-2), Monocyte chemotactic protein 3 (MCP-3), Interleukin-6 (IL6), Oncostatin-M (OSM), Tumor necrosis factor (TNF), Interleukin-20 (IL-20), Leukemia inhibitory factor (LIF), Interleukin-8 (IL8), C-C motif chemokine 4 (CCL4), C-C motif chemokine 3 (CCL3), Neurotrophin-3 (NT-3), Interleukin-2 (IL2), Interleukin-24 (IL-24), Interleukin- 10 receptor subunit beta (IL- 1 ORB),Interleukin- 18 receptor 1 (IL-18R1), Leukemia inhibitory factor receptor (LIF-R), C-X-C motif chemokine 5 (CXCL5), Delta and Notch-like epidermal growth factor-related receptor (DNER), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 1 (CXCL1), Protein SI 00-Al 2 (EN-RAGE), Latency-associated peptide transforming growth factor beta-1 (LAP TGF-beta-1), CD40L receptor (CD40), Transforming growth factor alpha (TGF-alpha), Caspase-8 (CASP-8), Vascular endothelial growth factor A (VEGFA), Urokinase-type plasminogen activator (uPA), Matrix metalloproteinase- 1 (MMP-1), Matrix metalloproteinase- 10 (MMP-10), Fibroblast growth factor 5 (FGF-5), Interleukin- 10 receptor subunit alpha (IL- 1 ORA), Interleukin- 15 receptor subunit alpha (IL-15RA), Neurturin (NRTN), Interleukin- 12 subunit beta (IL-12B), C-C motif chemokine 23 (CCL23), T-cell surface glycoprotein CD8 alpha chain (CD8A), Natural killer cell receptor 2B4 (CD244), Osteoprotegerin (OPG), TNF-related apoptosisinducing ligand (TRAIL), T cell surface glycoprotein CD6 isoform (CD6), Programmed cell death 1 ligand 1 (PD-L1), TNF-related activation-induced cytokine (TRANCE), T-cell surface glycoprotein CD5 (CD5), Tumor necrosis factor receptor superfamily member 9 (TNFRSF9), TNF-beta (TNFB), C-X-C motif chemokine 9 (CXCL9), Monocyte chemotactic protein 4 (MCP-4), C- X-C motif chemokine 10 (CXCL10), C-X-C motif chemokine 11 (CXCL11), Tumor necrosis factor ligand superfamily member 14 (TNFSF14), Interleukin- 10 (IL10), Interferon gamma (IFNG), Interleukin-4 (IL4), Interleukin- 13 (IL 13), Interleukin-5 (IL5), Signaling lymphocytic activation molecule (SLAMF1), or any combination thereof; and c) identifying the test substance as an allergen or an irritant based on the level of the one or more biomarkers.

26. The method of any one of claims 1-25, wherein the one or more protein biomarkers are selected from CCL20, CXCL1, IL8, TSLP, ARTN, IL4, CXCL10, OPG, IL5, CXCL9, TNFSF14, TNFB, TRAIL, IL 13, IFNG, CXCL11, and TNF.

27. The method of any one of claims 1-26, wherein the one or more protein biomarkers are selected from CCL20, CXCL1, IL8, TSLP, ARTN, IL4, CXCL10, OPG, IL5, CXCL9, TNFSF14, TNFB, IL13, IFNG, CXCL11, and TNF.

28. The method of any one of claims 1-27, wherein the one or more protein biomarkers are selected from IFNG, CXCL9, CXCL10, CXCL11, IL4, IL5, and IL13.

29. The method of any one of claims 1-27, wherein the one or more protein biomarkers are selected from CCL20 and CXCL1.

30. The method of any one of claims 1-25, wherein the one or more protein biomarkers are selected from CXCL10, IL4, OPG, CCL19, CCL20, TSLP, IL-1 alpha, and NRTN.

31. The method of any one of claims 1-25 and 30, wherein the one or more protein biomarkers are selected from CXCL10, IL4, OPG, CCL19, CCL20, TSLP, and IL-1 alpha.

Images