Analytical Methods and Arrays for Use in the Same

Abstract

The present invention relates to a method for identifying agents capable of inducing respiratory sensitization in a mammal and arrays and diagnostic kits for use in such methods. In particular, the methods include measurement of the expression of the biomarkers listed in Table A(i), Table A(ii) and/or Table A(iii) in cells exposed to a test agent.

Claims

1. A method for identifying agents capable of inducing respiratory sensitization in a mammal comprising or consisting of the steps of: a) exposing a population of dendritic cells or a population of dendritic-like cells to a test agent; and b) measuring in the cells the expression of one or more biomarker(s) selected from the group defined in Table A(i) and/or Table A(ii); wherein the expression in the cells of the one or more biomarkers measured in step (b) is indicative of the respiratory sensitizing effect of the test agent.

2. The method according to claim 1 further comprising: c) exposing a separate population of the dendritic cells or dendritic-like cells to one or more negative control agent that is not a respiratory sensitizer in a mammal; and d) measuring in the cells the expression of the one or more biomarker(s) measured in step (b) wherein the test agent is identified as a respiratory sensitizer in the event that the presence and/or amount in the test sample of the one or more biomarker measured in step (d) differs from the presence and/or amount in the control sample of the one or more biomarker measured in step (b).

3. The method according to claim 1 or 2 further comprising: e) exposing a separate population of the dendritic cells or dendritic-like cells to one or more positive control agent that is a respiratory sensitizer in a mammal; and f) measuring in the cells the expression of the one or more biomarker(s) measured in step (b) wherein the test agent is identified as a respiratory sensitizer in the event that the presence and/or amount in the test sample of the one or more biomarker measured in step (f) corresponds to the presence and/or amount in the positive control sample of the one or more biomarker measured in step (b).

4. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of one or more biomarkers defined in Table A(ii) for example, at least 2 or 3 of the biomarkers defined in Table 1A.

5. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of TNFRSF19.

6. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of SNORA74A.

7. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of SPAM1.

8. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of TNFRSF19, SNORA74A and SPAM1.

9. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring in step (b) the expression of one or more biomarkers defined in Table A(ii), for example, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287, 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 340, 341 or 342 of the biomarkers defined in Table A(ii).

10. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of all of the biomarkers defined in Table A(ii).

11. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of one or more of the biomarkers defined in Table A(iii), for example, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43 or 44 of the biomarkers defined in Table A(iii).

12. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of all of the biomarkers defined in Table A(iii).

13. The method according to any one of the preceding claims wherein step (b) comprises or consists of measuring the expression of all of the biomarkers defined in Table A.

14. The method according to any one of the preceding claims wherein step (b) comprises measuring the expression of a nucleic acid molecule encoding the one or more biomarker(s).

15. The method according to claim 14 wherein the nucleic acid molecule is a cDNA molecule or an mRNA molecule.

16. The method according to claim 14 wherein the nucleic acid molecule is an mRNA molecule.

17. The method according to claim 14 wherein the nucleic acid molecule is a cDNA molecule.

18. The method according to any one of claims 14 to 17 wherein measuring the expression of the one or more biomarker(s) in step (b) is performed using a method selected from the group consisting of Southern hybridisation, Northern hybridisation, polymerase chain reaction (PCR), reverse transcriptase PCR (RT-PCR), quantitative real-time PCR (qRT-PCR), nanoarray, microarray, macroarray, autoradiography and in situ hybridisation.

19. The method according to any one of claims 14 to 18 wherein measuring the expression of the one or more biomarker(s) in step (b) is determined using a DNA microarray.

20. The method according to any one of the preceding claims wherein measuring the expression of the one or more biomarker(s) in step (b) is performed using one or more binding moieties, each capable of binding selectively to a nucleic acid molecule encoding one of the biomarkers identified in Table A.

21. The method according to claim 20 wherein the one or more binding moieties each comprise or consist of a nucleic acid molecule.

22. The method according to claim 21 wherein the one or more binding moieties each comprise or consist of DNA, RNA, PNA, LNA, GNA, TNA or PMO.

23. The method according to claim 20 or 22 wherein the one or more binding moieties each comprise or consist of DNA.

24. The method according to any one of claims 21 to 24 wherein the one or more binding moieties are 5 to 100 nucleotides in length.

25. The method according to any one of claims 21 to 25 wherein the one or more nucleic acid molecules are 15 to 35 nucleotides in length.

26. The method according to any one of claims 21 to 26 wherein the binding moiety comprises a detectable moiety.

27. The method according to claim 26 wherein the detectable moiety is selected from the group consisting of: a fluorescent moiety; a luminescent moiety; a chemiluminescent moiety; a radioactive moiety (for example, a radioactive atom); or an enzymatic moiety.

28. The method according to claim 26 wherein the detectable moiety comprises or consists of a radioactive atom.

29. The method according to claim 28 wherein the radioactive atom is selected from the group consisting of technetium-99m, iodine-123, iodine-125, iodine-131, indium-111, fluorine-19, carbon-13, nitrogen-15, oxygen-17, phosphorus-32, sulphur-35, deuterium, tritium, rhenium-186, rhenium-188 and yttrium-90.

30. The method according to claim 27 wherein the detectable moiety of the binding moiety is a fluorescent moiety.

31. The method according to any one of claims 1 to 22 wherein step (b) comprises or consists of measuring the expression of the protein of the one or more biomarker defined in step (b).

32. The method according to claim 31 wherein measuring the expression of the one or more biomarker(s) in step (b) is performed using one or more binding moieties each capable of binding selectively to one of the biomarkers identified in Table A.

33. The method according to claim 32 wherein the one or more binding moieties comprise or consist of an antibody or an antigen-binding fragment thereof.

34. The method according to claim 33 wherein the antibody or fragment thereof is a monoclonal antibody or fragment thereof.

35. The method according to claim 33 or 34 wherein the antibody or antigen-binding fragment is selected from the group consisting of intact antibodies, Fv fragments (e.g. single chain Fv and disulphide-bonded Fv), Fab-like fragments (e.g. Fab fragments, Fab′ fragments and F(ab)2 fragments), single variable domains (e.g. V.sub.H and V.sub.L domains) and domain antibodies (dAbs, including single and dual formats [i.e. dAb-linker-dAb]).

36. The method according to claim 35 wherein the antibody or antigen-binding fragment is a single chain Fv (scFv).

37. The method according to claim 32 wherein the one or more binding moieties comprise or consist of an antibody-like binding agent, for example an affibody or aptamer.

38. The method according to any one of claims 32 to 37 wherein the one or more binding moieties comprise a detectable moiety.

39. The method according to claim 38 wherein the detectable moiety is selected from the group consisting of a fluorescent moiety, a luminescent moiety, a chemiluminescent moiety, a radioactive moiety and an enzymatic moiety.

40. The method according to any one of the preceding claims wherein step (b) is performed using an array.

41. The method according to claim 40 wherein the array is a bead-based array.

42. The method according to claim 41 wherein the array is a surface-based array.

43. The method according to any one of claims 40 to 42 wherein the array is selected from the group consisting of: macroarray; microarray; nanoarray.

44. An array for use in a method according any one of the preceding claims, the array comprising one or more first binding agents as defined in any one of claims 20 to 30 and 32 to 39.

45. An array according to claim 44 comprising binding agents which are collectively capable of binding to all of the biomarkers defined in Table 1.

46. An array according to claim 44 or 45 wherein the first binding agents are immobilised.

47. The method according to any one of the preceding claims for identifying agents capable of inducing a respiratory hypersensitivity response.

48. The method according to any one of the preceding claims wherein the hypersensitivity response is a humoral hypersensitivity response.

49. The method according to claim 47 or 48 wherein the hypersensitivity response is a type I hypersensitivity response.

50. The method according to any one of the preceding claims for identifying agents capable of inducing respiratory allergy.

51. The method according to any one of the preceding claims wherein the population of dendritic cells or population of dendritic-like cells is a population of dendritic-like cells.

52. The method according to claim 51 wherein the dendritic-like cells are myeloid dendritic-like cells.

53. The method according to claim 52 wherein the myeloid dendritic-like cells are derived from myeloid dendritic cells.

54. The method according to claim 53 wherein the cells derived from myeloid dendritic cells are myeloid leukaemia-derived cells.

55. The method according to claim 54 wherein the myeloid leukaemia-derived cells are selected from the group consisting of KG-1, THP-1, U-937, HL-60, Monomac-6, AML-193 and MUTZ-3.

56. The method according to any one of the preceding claims wherein the dendritic-like cells are MUTZ-3 cells.

57. The method according to any one of the claims 2 to 56 wherein the one or more negative control agent provided in step (c) is selected from the group consisting of 1-Butanol; 2-Aminophenol; 2-Hydroxyethyl acrylate; 2-nitro-1,4-Phenylenediamine; 4-Aminobenzoic acid; Chlorobenzene; Dimethyl formamide; Ethyl vanillin; Formaldehyde; Geraniol; Hexylcinnamic aldehyde; Isopropanol; Kathon CG*; Methyl salicylate; Penicillin G; Propylene glycol; Potassium Dichromate; Potassium permanganate; Tween 80; and Zinc sulphate.

58. The method according to claim 57 wherein at least 2 control non-sensitizing agents are provided, for example, at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or at least 20 control non-sensitizing agents.

59. The method according to any one of claims 3 to 58 wherein the one or more positive control agent provided in step (e) comprises or consists of one or more agent selected from the group consisting of ammonium hexachloroplatinate, ammonium persulfate, glutaraldehyde, hexamethylen diisocyanate, maleic anhydride, methylene diphenol diisocyanate, phtalic anhydride, toluendiisocyanate and trimellitic anhydride.

60. The method according to claim 59 wherein at least 2 control sensitizing agents are provided, for example, at least 3, 4, 5, 6, 7, 8, 9 or at least 10 control sensitizing agents.

61. The method according to any one of the preceding claims wherein the method is indicative of the sensitizing potency of the sample to be tested.

62. An array for use in a method according any one of the preceding claims, the array comprising one or more binding moieties as defined in any one of claims 20 to 30 and 32 to 39.

63. An array according to claim 62 wherein the binding moieties are capable of binding to all of the biomarkers defined in Table A(i).

64. An array according to claim 62 or 63 wherein the binding moieties are capable of binding to all of the biomarkers defined in Table A(ii).

65. An array according to claim 62, 63 or 64 wherein the binding moieties are capable of binding to all of the biomarkers defined in Table A(iii).

66. An array according to any one of claims 62 to 65 wherein the binding moieties are capable of binding to all of the biomarkers defined in Table A.

67. An array according to any on of claims 62 to 65 wherein the binding moieties are immobilised.

68. Use of two or more biomarkers selected from the group defined in Table A in combination for identifying respiratory hypersensitivity response sensitising agents.

69. The use according to claim 68 wherein all of the biomarkers defined in Table A are used collectively for identifying hypersensitivity response sensitising agents.

70. Use of one or more binding moiety as defined in any one of claims 20 to 30 or 32 to 39 for identifying respiratory hypersensitivity response sensitising agents.

71. The use according to claim 70 wherein all of the biomarkers defined in Table A are used collectively for identifying hypersensitivity response sensitising agents

72. An analytical kit for use in a method according any one of claims 1 to 61 comprising: A) an array according to any one of claims 62 to 67 and/or one or more binding moiety as defined in any one of claims 20 to 30 or 32 to 39; and B) instructions for performing the method as defined in any one of claims 1 to 60 (optional).

73. An analytical kit according to claim 72 further comprising one or more control samples.

74. An analytical kit according to claim 73 comprising one or more non-sensitizing agent(s).

75. An analytical kit according to claim 72, 73 or 74 comprising one or more sensitizing agent(s).

76. A method of treating or preventing a respiratory type I hypersensitivity reaction (such as respiratory asthma) in a patient comprising the steps of: (a) providing one or more test agent that the patient is or has been exposed to; (b) determining whether the one or more test agent provided in step (a) is a respiratory sensitizer using a method provided in the first aspect of the present invention; and (c) where one or more test agent is identified as a respiratory sensitizer, reducing or preventing exposure of the patient to the one or more test agent identified as a respiratory sensitizer and/or providing appropriate treatment for the symptoms of sensitization.

77. The method according to claim 76 wherein the treatment of the symptoms of sensitization is selected from the group consisting of short-acting beta2-adrenoceptor agonists (SABA), such as salbutamol; anticholinergic medications, such as ipratropium bromide; other adrenergic agonists, such as inhaled epinephrine; Corticosteroids such as beclomethasone; long-acting beta-adrenoceptor agonists (LABA) such as salmeterol and formoterol; leukotriene antagonists such as montelukast and zafirlukast; and/or mast cell stabilizers (such as cromolyn sodium).

78. A computer program for operating the method defined in the first aspect of the invention.

79. The computer program according to claim 78 wherein the computer program is recorded on a computer-readable carrier.

80. A method or use substantially as described herein.

81. An array, kit or computer program substantially as described herein.

Description

[0187] Preferred, non-limiting examples which embody certain aspects of the invention will now be described, with reference to the following figures:

[0188] FIG. 1. CD86 expression of MUTZ-3 cells following chemical stimulations. Data shown is an average of chemical stimulations, (n=3), 4-Aminobenzoic acid, DMSO and unstimulated cells (n=6) and potassium permanganate, 2-aminophenol, Hexylcinnamic aldehyde and 2-Hydroxyethyl acrylate (n=2), with error bars showing standard deviation. Statistical significance was determined by student's t-test, comparing each stimulation with its corresponding vehicle, with p<0.05 indicated by *.

[0189] FIG. 2. Establishment of a predictive biomarker signature for prediction of respiratory sensitization. (A) Unsupervised learning was used to construct the representation of the dataset. The method was visualized using PCA based on 999 transcripts identified by one-way ANOVA p-value filtering between respiratory sensitizers (blue, n=29) and non-respiratory sensitizers (green, n=74). (B) The 999 transcripts identified by p-value filtering were used as input into an algorithm for backward elimination. A breakpoint in Kullback-Leibler divergence was observed after removal of 610 transcripts. (C) The remaining 389 transcripts were used as input variables into a PCA. As illustrated in the figure, a complete seperation between respiratory sensitizers and non-respiratory sensitizers was achieved in the training data.

[0190] FIG. 3. Visual classification of independent test compounds using GARD

[0191] Respiratory Prediction Signature, GRPS. (A) The PCA space was constructed from the three first PCA components from the panel of reference chemicals (n=103) used for biomarker signature identification, using the 389 genes of GRPS as input into the unsupervised representation. Each of the chemicals in the test dataset (n=92) were plotted into the PCA space without allowing the compounds to influence PCA components. (B) Samples in the test dataset was colored according to sensitizing properties as either respiratory sensitizers (dark blue) or non-respiratory sensitizers (dark green). A separation between respiratory sensitizers and non-respiratory sensitizers can be seen along the first PCA component for both the training data and the test data. (C) The training dataset has been removed in order to obtain a clear view of the training dataset.

[0192] FIG. 4. Support Vector Machine (SVM) classifications of the test dataset. The predictor performance of GRPS was validated using SVM for supervised machine learning. The SVM algorithm was inductively learned by experience to the compounds in the training dataset (n=103) and subsequently applied to predict each individual sample in the test dataset (n=70, vehicle controls excluded). The predictive performance was evaluated by ROC curve analysis and estimated to an Area Under the Curve (AUC) of 0.97.

[0193] FIG. 5. Classification and gene expression of respiratory sensitizers in the independent test dataset. The SVM algorithm was once again trained on the samples in the training dataset (n=103) and subsequently applied in order to classify samples in the independent test dataset (n=70, vehicle controls were excluded). SVM decision values for each individual sample in the independent test dataset were plotted in decreasing order and coloured according to sensitizing capacity (Respiratory sensitizers=purple, non-respiratory sensitizers=dark green). The dotted line in the scatterplot represents the threshold level for classifications as respiratory sensitizers (SVM decision value>0) or non-respiratory sensitizers (SVM decision value<0). Relative expression of transcripts within GRPS is shown in the heat map.

EXAMPLES

[0194] Introduction

[0195] Background: Repeated exposure to certain low molecular weight (LMW) chemical compounds may result in development of allergic reactions in the skin or in the respiratory tract. In most cases, a certain LMW compound selectively sensitize the skin, giving rise to allergic contact dermatitis (ACD), or the respiratory tract, giving rise to occupational asthma (OA). To limit occurrence of allergic diseases, efforts are currently being made to develop predictive assays that accurately identify chemicals capable of inducing such reactions. However, while a few promising methods for prediction of skin sensitization have been described, to date no validated method, in vitro or in vivo, exists that is able to accurately classify chemicals as respiratory sensitizers.

[0196] Results: Recently, we presented the in vitro based Genomic Allergen Rapid Detection (GARD) assay as a novel testing strategy for classification of skin sensitizing chemicals based on measurement of a genomic biomarker signature. We have expanded the applicability domain of the GARD assay to classify also respiratory sensitizers by identifying a separate biomarker signature containing 389 differentially regulated genes for respiratory sensitizers in comparison to non-respiratory sensitizers. By using an independent data set in combination with supervised machine learning, we validated the assay, showing that the identified genomic biomarker is able to accurately classify respiratory sensitizers.

[0197] Conclusions: We have identified a genomic biomarker signature for classification of respiratory sensitizers. Combining this newly identified biomarker signature with our previously identified biomarker signature for classification of skin sensitizers, we have developed a novel in vitro testing strategy with a potent ability to predict both skin and respiratory sensitization in the same sample.

[0198] Materials and Methods

[0199] Chemicals

[0200] A panel of 32 reference chemicals comprising a selection of 10 well characterized respiratory sensitizers and 22 non-respiratory sensitizers, collectively termed the training dataset, were used for cell stimulations in order to establish the predictive genomic biomarker signature. The respiratory sensitizers were ammonium hexachloroplatinate, ammonium persulfate, ethylenediamine, glutaraldehyde, hexamethylene diisocyanate, maleic anhydride, methylene diphenyl diisocyanate, phtalic anhydride, toluene diisocyanate and trimellitic anhydride. The non-respiratory sensitizers were 1-butanol, 2-aminophenol, 2-hydroxyethyl acrylate, 2-nitro-1,4-phenylenediamine, 4-aminobenzoic acid, chlorobenzene, dimethyl formamide, ethyl vanillin, formaldehyde, geraniol, hexylcinnamic aldehyde, isopropanol, Kathon CG, methyl salicylate, penicillin G, potassium dichromate, potassium permanganate, propylene glycol, Tween 80, zinc sulfate and the vehicle controls dimethyl sulfoxide and water. Additionally, a panel of 25 chemicals, including 6 respiratory sensitizers and 19 non-respiratory sensitizers, collectively termed the independent test dataset, were used for cell stimulations in order to form an independent testset for validation of the identified predictive genomic biomarker signature. The independent training dataset comprised both control chemicals, included during the training of the model, as well as chemicals previously unseen during training of the model. The respiratory sensitizers were chloramine T, ethylenediamine, Isophorone diisocyanate, phtalic anhydride, piperazine and reactive orange. The non-respiratory sensitizers were 1-butanol, 2,4-dinitrochlorobenzene, 2-mercaptobenzothiazole, benzaldehyde, chlorobenzene, cinnamyl alcohol, diethyl phthalate, eugenol, glycerol, glyoxal, isoeugenol, lactic acid, octanoic acid, phenol, p-hydroxybenzoic acid, p-phenylendiamine, resorcinol, salicylic acid and sodium dodecyl sulfate. All chemicals were purchased from Sigma-Aldrich (St. Louis, Mo., USA). Chemicals were dissolved and diluted into GARD input concentration in either water or DMSO prior to stimulation of cells. For chemicals dissolved in DMSO, the in-well concentration of DMSO was 0.1%. Monitoring of chemical cytotoxicity and establishment of GARD input concentration for each chemical compound was performed as previously described [41,42]. In short, GARD input concentration was determined according to the following decision schedule: Non-toxic and freely soluble compounds were used at a concentration corresponding to 500 uM. Non-toxic and poorly soluble compounds, insoluble at 500 uM, were used at highest soluble concentration. Toxic compounds were used at a concentration yielding 90% relative viability (Rv90). The criterion that was first met determined the GARD input concentration for each compound. The GARD input concentration, sensitizing potency and solvent are presented in Table 1 for compounds used to establish the predictive genomic biomarker signature, and in Table 2 for compounds used to validate the predictive genomic biomarker signature.

[0201] Cell Cultures, Phenotypic Analysis, Chemical Exposure, Cell Harvest and Mrna Isolation

[0202] The human acute myelomonocytic leukemia cell line MUTZ-3[68,69] was obtained from Leibniz-lnstitut DSMZ-Deutsche Sammlung von Mikroorganismen and Zellkulturen (DSMZ, Braunschweig, Germany). Maintenance of cells, chemical stimulation of MUTZ-3 and all subsequent isolation of mRNA and preparation of cDNA were performed as previously described [41,42]. In short, a phenotypic control of MUTZ-3 was performed using flow cytometry prior to chemical stimulation to ensure cells were in an immature state. The following FITC-conjugated mouse monoclonal antibodies (mAbs) were used: CD1a (DakoCytomation, Glostrup, Denmark), CD34, CD86, HLA-DR, IgG1 (BD Biosciences, Franklin Lakes, N.J.). The following PE-conjugated mouse monoclonal antibodies were used: CD14 (DakoCytomation), CD54, CD80, IgG1 (BD Biosciences). Cell viability was determined using Propidium Iodide (BD Biosciences) staining. Samples were run on FACSCanto II instrument. Data was acquired using FACS Diva software (BD Biosciences) and analyzed using FCS Express V4 (De Novo Software, Los Angeles, Calif.). Gating was performed to exclude cell debris and non-viable cells based on forward- and side-scattering properties and quadrants established using isotype-controls. During chemical exposure, cells were seeded at 200.000 cells/ml in 24-well plates and exposed to chemical compound at the GARD input concentration. Stimulated cells were harvested after 24 h incubation at 37° C., 5% CO.sub.2 and RNA was isolated with TRIzol® reagent (Life Technologies, Carlsbad, Calif.) using standardized protocols provided by the manufacturer. In parallel, a control of the maturity state of the cells was performed by flow cytometric analysis of CD86. Preparation of cDNA and hybridization, washing and scanning of the Human Gene 1.0 ST Arrays (Affymetrix, Santa Clara, Calif., USA) was performed, according to standardized protocols provided by the manufacturer (Affymetrix).

[0203] Microarray Data Analysis and Statistical Methods

[0204] Gene expression data obtained from the Human Gene 1.0 ST Arrays were normalized using the Single-Channel array normalization (SCAN) algorithm [70], and potential batch effects between different set of experiments were adjusted using the ComBat [71] empirical bayes method. Normalizations and batch adjustments were performed in R statistical software [72] using the open software Bioconductor v2.14 [73] with the additional software packages SCAN.UPC [70] and sva [74] . Normalized data was imported into Qlucore Omics Explorer 3.0 (Qlucore AB, Lund, Sweden) and visualized using Principal Component Analysis (PCA) [75]. Predictors were selected from a one-way ANOVA p-value filtration, using false discovery rate (FDR) [76] to adjust for multiple hypothesis testing, comparing respiratory sensitizers and non-respiratory sensitizers. A wrapper algorithm for Backward Elimination [41,52] was applied on the top 999 predictors, to further reduce and refine the biomarker signature size. The Backward Elimination algorithm was modified to minimize the Kullback-Leibler error [53] rather than maximizing the Area Under the Receiver Operating Characteristic (AUC ROC) [77], in order to enable signature optimization in cases where the AUC ROC reaches 1.0. The selected top 389 predictors after backward elimination were collectively designated “GARD Respiratory Prediction Signature”. The script for Backwards Elimination was programmed in R, with the additional package e1071 [78]. The method by which the predictive genomic biomarker signature was established was validated using cross-validation based on Support Vector Machines (SVM) [79], based on a linear kernel, as described previously [41]. In short, the training dataset was randomly divided into a new cross-validation training dataset comprising 70% of the stimulations, and a cross-validation test dataset comprising 30% of the stimulations. In addition, care was taken to maintain the same proportion between respiratory sensitizers and non-respiratory sensitizers as in the complete training dataset. A new predictive genomic biomarker signature was identified from the cross-validation training dataset using one-way ANOVA p-value filtration as described above. The identified predictive biomarker signature was used to train a SVM based on the information in the cross-validation training dataset. SVMs were compiled in R statistical software with the additional package e1071. The SVM model was subsequently used to predict the samples of the cross-validation test dataset. The process of biomarker identification was repeated 20 times and the robustness of the feature selection process was evaluated by calculating the frequency (referred to as the Validation call frequency, of VCF) by which each individual transcript was included in the 20 training datasets. The predictive performance of the GRPS in terms of prediction of unknown samples was estimated using the independent test dataset as described in [46]. In short, a SVM was trained on the training dataset, using the GRPS as variable input. Subsequently, the SVM was then used to predict the samples in the training dataset in the same way as described for the cross-validation above, and the predictive performance of the model was evaluated using AUC ROC, determined in R statistical environment using the additional package ROCR [80]. Classification of samples as respiratory sensitizers or non-respiratory sensitizers were based on SVM decision values on replicate level. Hence, a chemical was classified as a respiratory sensitizer if any of the replicate stimulations from a certain chemical stimulation had an SVM decision value >0. The accuracy, sensitivity and selectivity of the assay was determined using cooper statistics [81]. The biological relevance of the GRPS was explored using MetaCore™ (Thomson Reuters, New York, N.Y.) by performing a functional enrichment. The top 999 predictors from a p-value filtering were used as input into the MetaCore™ algorithm and biological relevance was established by exploring the Canonical Pathways associated with input molecules. The array data has been uploaded to ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession number E-MEXP-3773.

[0205] Results

[0206] Phenotypic Analysis of Unstimulated and Chemically Stimulated MUTZ-3 Cells

[0207] Prior to chemical challenge, cells were quality controlled by measuring the cellular expression of common myeloid and dendritic cell markers using flow cytometry. These markers included CD1a, CD14, CD34, CD54, CD80, CD86 and HLA-DR. Results correlated with previously published phenotypic profiles [41], ensuring that cells were successfully maintained in an immature state. Following chemical stimulation, using a panel of reference chemicals comprising 10 respiratory sensitizers and 20 non-respiratory sensitizers, as well as vehicle controls (Table 1), the general maturity state of the cells was again verified by measuring levels of cell surface expression of the co-stimulatory marker CD86, with results presented in FIG. 1. Chemically induced up-regulation of CD86 for each stimulation in comparison to unstimulated cells was evident in cells after a number of stimulations. However, due to large standard deviation between replicate stimulations, only regulation induced by the respiratory sensitizers ammonium hexachloroplatinate and glutaraldehyde could be confirmed as statistically significant (students t-test, p<0.05). Consequently, an assay using CD86 as a single biomarker for classification of respiratory sensitization would result in a sensitivity of only 20%. Additionally, up-regulation of CD86 was also observed in the non-respiratory sensitizing stimulations 2-aminophenol and kathon CG. Thus, we concluded that CD86 could not be used as a single biomarker in the MUTZ-3 cell line to classify respiratory chemical sensitizers. However, as many of the respiratory sensitizers have a poor solubility in cell media, and can thus not be used in sufficient concentration to induce cytotoxicity, the expression of CD86 may act as a complementary quality control to ensure bioavailability of chemicals stimulations.

[0208] Identification of a Predictive Genomic Biomarker Signature by Transcriptional Profiling of Chemically Stimulated MUTZ-3

[0209] Chemically induced changes in the MUTZ-3 cells were investigated on transcriptional-wide basis in order to identify the most discriminatory transcripts between respiratory sensitizers and non-respiratory sensitizers. Following 24 h of cellular stimulation with a panel of reference chemicals, mRNA was collected for transcriptional profiling. The stimulations included 10 different respiratory sensitizers, 20 non-respiratory sensitizers (negative controls) and vehicle controls (DMSO, distilled water). All stimulations were performed in biological triplicates except for 4-aminobenzoic acid, which was analyzed in 6 replicates due to internal controls, and potassium permanganate, which was analyzed in only 2 replicates due to a faulty array. In addition, vehicle controls (DMSO and distilled water) were analyzed in 6 replicates each. Quality control of samples revealed that one of the replicate stimulations of ammonium persulfate was a significant outlier and had to be removed in order not to interfere with biomarker discovery. Summarized, the data set ready for analysis consisted of 103 arrays, each with measurements of 29,141 transcripts.

[0210] Gene expression data was imported into Qlucore Omics Explorer and visualized using principal component analysis (PCA). We applied a tiered approach for feature selection, combining a filtering method in order to reduce the noise in the dataset and select predictors based on intrinsic properties, and a wrapper method based on Backward Elimination in order to reduce the number of genes in the signature by taking into account how each individual predictor performs collectively with the entire signature. The filtering method was based entirely on p-values, as determined by one-way ANOVA analysis, comparing respiratory sensitizers and non-respiratory sensitizers. The wrapper method was based on repeated supervised learning. The algorithm for Backward Elimination was developed in house [52] and iteratively extracted a subset of transcripts which were subsequently evaluated by training and testing of a Support Vector machine (SVM), using a leave-one-out cross validation procedure. The least informative variables were removed, and the process was repeated until the highest performance of classification model was achieved. Due to computational limits, approximately 1000 genes is an appropriate amount of potential predictors to use as an input in the algorithm for Backward Elimination. In the present data set, this pre-selection of predictor candidates resulted in 999 genes, with a p-value of 0.024 or lower. As illustrated in FIG. 2A, these genes were collectively able to achieve a partition between respiratory sensitizers and non-respiratory sensitizers, although some overlap occurred between the two groups. Reducing the number of predictors further, by the ranking given by their p-value, did not achieve a clear separation, even though the data contained predictor candidates with p-values down to 10.sup.−6. The in house developed algorithm for Backward Elimination was then applied, removing the predictors (genes) that contribute the least information. A local minimum in Kullbach-Liebler Divergence (KLD) [53] was observed when 610 predictors were eliminated (FIG. 2B). The remaining 389 genes are collectively termed the “GARD Respiratory Prediction Signature” (GRPS), and their ability to differentiate between respiratory chemical sensitizers and non-respiratory sensitizers in the training dataset are illustrated in FIG. 2C. The identities of the genes are listed in Table A. In order to validate the method by which the biomarker signature was established, we performed a cross validation procedure where we randomly divided the samples used during biomarker discovery into a training set and a test set as described in methods. The process was iterated 20 times and the frequency of each predictor transcript among the 20 new biomarker signatures was used as a measurement of robustness. Results from this exercise were tabulated as Validation Call Frequencies (VCF) and are summarized in table A.

[0211] Visual Classification of Independent Test Compounds Using GARD Respiratory Prediction Signature as an Estimate of Predictive Performance

[0212] The predictive performance of GRPS was validated using an independent test dataset comprising both respiratory sensitizers as well as non-respiratory sensitizers in order to illustrate the relevance of the genomic biomarker signature as a predictive assay for respiratory sensitizers. The chemical compounds included in the independent test dataset are described in Table 2. Some of the compounds used during the training of the model, including 1-butanol, chlorobenzene, ethylenediamine, phtalic anhydride, and the vehicle controls (DMSO, distilled water) were included also in the independent test dataset to be used as controls. The remaining chemicals were unseen by the model prior to classification of the samples. All chemicals in the independent test dataset were based on additional stimulations, separated in time from the stimulations comprising the training dataset. Therefore, both the unseen compounds, as well as the compounds seen by the model during the identification of the GRPS could be classified without the risk of over fitting the model. Following 24 h of cellular stimulation, mRNA was collected, converted to cDNA and hybridized to the microarrays. The stimulations included 6 respiratory sensitizers, 19 non-respiratory sensitizers and vehicle controls (DMSO, distilled water). The non-respiratory sensitizing stimulations were reused from a previous set of experiments [41] and performed in biological triplicates. Stimulations performed with the respiratory sensitizers, together with the non-respiratory sensitizer 1-butanol, comprised a novel round of stimulations, and were performed in biological duplicates. The chemical chlorobenzene was included in both sets due to internal controls, hence comprised a total of 5 stimulations. In addition, vehicle controls DMSO and distilled water were analyzed in 13 and 9 replicates, respectively. In summary, the independent test dataset comprised 92 arrays. The process of performing visualized classifications of unknown samples is sequentially illustrated in FIG. 3, using the test dataset to highlight the methodology. In an initial step, the training data set, i.e. the panel of reference chemicals used to identify the predictive GRPS genomic biomarker signature, was first used to generate the PCA space, using the 389 genes in the GRPS as variable input. The PCA was then frozen in space, and each of the compounds in the test dataset was plotted into this space without allowing them to influence the PCA components (Fig.3A). As demonstrated in FIG. 3B, upon identification of true identities of the samples in the test dataset, a clear separation between respiratory sensitizers and non-respiratory sensitizers can be seen along the first PCA component for both the training data and the test data, indicating that similarities and differences in structure of gene expression in the GPRS between respiratory sensitizers and non-respiratory sensitizers was present also in the test dataset. FIG. 3C illustrates the test dataset plotted into the frozen PCA space generated by the training dataset, but where the training dataset have been removed in order to facilitate interpretation. As seen in the figure, respiratory sensitizers and non-respiratory sensitizers appears to be separated by a hyper plane generated by the 2.sup.nd and 3.sup.rd principal components, indicating that the GRPS clearly contain information relevant for achieving discrimination between respiratory sensitizers and non-respiratory sensitizers in the previously unseen samples in the independent test dataset.

[0213] SVM Classifications and Predictive Performance of GRPS

[0214] In a consecutive step to classify the samples in the independent test dataset, the visual classifications were challenged with a binary classification model, using an SVM for supervised machine learning. The SVM was trained on the training data set to recognize differences in gene expression structure between respiratory sensitizers and non-respiratory sensitizers within the GRPS. The trained SVM model was applied to classify each sample in the independent test data set, on the level of each individual replicate, as either a respiratory sensitizer or a non-respiratory sensitizer. The output from the SVM, the SVM decision values, were compared to true identities of samples in the test dataset, and the performance of the predictor was evaluated using ROC AUC analysis with results illustrated in FIG. 4. SVM classifications were based on linear kernels with an unbiased maximum-margin hyperplane separating the two groups, hence threshold for classifications as respiratory sensitizers corresponded to a SVM decision value >0. As shown in the figure, predictor performance of the GRPS on the level of individual stimulations was estimated to an area under the ROC curve of 0.97. Decision values obtained from the SVM classifications for each chemical compound in the test dataset are presented in Table 3 (n=70, vehicle controls excluded). Data presented in this table is further summarized in FIG. 5. In the figure, samples are sorted in a descending scale from highest to lowest SVM value, and the SVM decision value for each compound in the test dataset are correlated to the individual expression profile of the 389 transcripts in the GRPS. In order to facilitate interpretation, samples in the figure were colored according to sensitizing capacity (Respiratory sensitizers=purple, non-respiratory sensitizers=dark green) and the threshold value for classification is illustrated by the dotted line. As illustrated in the figure, although some overlap was observed, the majority of the respiratory sensitizers were assigned SVM decision values that were higher than corresponding values assigned to non-respiratory sensitizers, which also correlated with differences in the expression profiles between the majorities of compounds within each group. For decision making and classification on sensitizing capacity for each chemical, we chose to use the cut-off stating that any given sample in the test dataset should be classified as a respiratory sensitizer if any of the replicate stimulations has an SVM decision value >0, as determined in previously published protocols [42]. Based on this criterion, the accuracy, sensitivity and specificity of the GRPS was estimated using cooper statistics to 84%, 67%, and 89%, respectively.

[0215] Canonical Pathways Associated with Respiratory Sensitizers and GARD Prediction Signature

[0216] Aiming to investigate the biologic response initiated by respiratory chemical sensitizers in MUTZ-3 cells, the data was analyzed using functional enrichment analysis in Metacore™. The top 999 genes, selected with p-value filtering, were used as input into Metacore™. Of the 999 genes, Metacore.sup.TM was able to map 948 to unique IDs. Significantly regulated pathways (p<0.01) are listed in Table 4, ranked by −log (p-Value) and sorted in order of statistical significance. Genes present in GRPS are indicated in bold. A clear majority of these identified and significantly regulated pathways are mainly driven by a limited set of molecules. The most highly populated pathways included oxidative phosphorylation (26 molecules) and Ubiquinone metabolism (19 molecules), showing that cellular events such as oxidation-reduction processes and the respiratory electron transport chain function is highly affected by the studied chemicals. In addition, several of the less significantly regulated pathways, including Inhibitory PD-1 signaling in T cells, Antigen presentation by MHC class I and MHC class II, Generation of memory CD4+ T cells, IL-33 signaling pathway are relevant from an immunological point of view. Of note, central for many of these pathways is the bridge between innate and adaptive immunity, and the engagement of innate immune responses initiated by recognition of foreign substances, leading to dendritic cell maturation and activation of specific T-cell responses. Key aspects of this process is well monitored and significantly regulated in the MUTZ-3 cell line, including upregulation of antigen presentation-associated molecules, such as MHC class I and MHC class II complex, upregulation of co-stimulatory molecules, such as CD80 and CD86, and cross-talk with key players such as T-cells through initiation and coordination of pathways responsible for driving the immune response. Of note, activated pathways are only to a very limiting extent overlapping with pathways activated by skin sensitizers in MUTZ-3 [54] (Granzyme B and Granzyme A signaling), indicating that respiratory sensitizers and skin sensitizers are involved in engagement of different signaling pathways.

[0217] Discussion

[0218] A variety of chemicals are able of inducing allergic hypersensitivity reactions in both skin and respiratory tract, eventually giving rise to clinical symptoms of Allergic Contact Dermatitis (ACD) or Occupation Asthma (OA). Although the numbers of chemicals able of inducing respiratory sensitization are far fewer in comparison to those causing skin sensitization, identification and hazard classification of respiratory chemical sensitizers remains an area of great importance due to the severe impact on health and quality of life associated with acquired OA. Development of reliable assays that accurately identifies respiratory sensitizers as well as distinguishing those from skin sensitizers have proven challenging. In previous studies, we described the development and application of the Genomic Allergen Rapid Detection (GARD) assay as an in vitro alternative to animal testing for identification and risk assessment of skin sensitizing chemicals. In the GARD assay, unknown test chemicals are classified based on readout from a pre-determined genomic biomarker signature, measured by genome-wide transcriptional profiling. Utilizing the great versatility that comes with analyzing the complete transcriptome, we hypothesized that the applicability domain of GARD could be broadened to also cover hazard classification of respiratory sensitizers through the identification of an alternative genomic biomarker signature.

[0219] In the current study, we present a further development of GARD, allowing for classification of respiratory sensitizing chemicals, using a different biomarker signature termed the GARD Respiratory Prediction Signature (GRPS). The intended use of the defined GRPS will thus be in a novel combined in vitro assay, in which MUTZ-3 cells are stimulated with the unknown compounds to be classified. Of note, using the two distinct biomarker signatures, the compound can be classified as either a skin sensitizer, a respiratory sensitizer or a non-sensitizer. Chemicals that are able to induce both respiratory and skin sensitization will also be specifically classified as such.

[0220] The GRPS was identified, using a set of reference chemicals known to be either respiratory sensitizers or non-respiratory sensitizers. Differentially regulated genes in these two groups were then identified by an ANOVA p-value filtering and further optimized, using an in house developed wrapper algorithm for backward elimination. We suggest that the 389 genes in the GRPS can function as a genomic biomarker signature to discriminate between respiratory sensitizing chemicals and non-respiratory sensitizing chemicals. Assessment of the predictive performance of GRPS is important in order to establish the reliability of the genomic biomarker signature for identification of respiratory sensitizers. In this study, we used a Support Vector Machine (SVM) algorithm for supervised machine learning. We trained the model to recognize structures and similarities in gene expression data in the identified GRPS genomic biomarker signature, and challenged the model with an independent test set comprising chemicals previously unseen by the model.

[0221] Subsequently, we used the model to binary classify the unseen chemical compounds as either respiratory sensitizers or non-respiratory sensitizers. Performing this exercise, we demonstrated the potential of GRPS to achieve accurate predictions. The predictive performance of GRPS was estimated, using ROC AUC analysis and cooper statistics, achieving an area under the ROC curve of 0.97 and sensitivity, specificity and accuracy of 67%, 89% and 84%, respectively. This is the highest reported accuracy for hazard classification of chemicals inducing respiratory sensitization.

[0222] To date, we can only speculate on possible explanations to why the GRPS does not reach the same high sensitivity in predictions as the GPS for skin sensitizers [46]. It could partly be due to the smaller number of reference compounds used during assay development of GRPS in comparison to GPS, but another possible explanation could perhaps be found on the molecular level, i.e. that skin sensitizers are more potent regulators of gene expression in MUTZ-3 cells. Irrespectively, the use of whole genome arrays as readout for classifications still makes the GRPS highly flexible. As more samples are analyzed, additional information can easily be implemented into the assay to improve sensitivity, specificity and accuracy and to fine-tune the methodology to reflect the diversity of available chemical compounds.

[0223] To further explore the biological effects of sensitizing chemicals on MUTZ-3, an enrichment analysis was performed. In order to achieve sufficient significance in the data, the top 999 genes from p-value filtering were used as input in the Metacore™ software, rather than the top 389 genes of the GRPS. Without doubt, the most highly populated pathways initiated by the respiratory sensitizers were involved in cellular events such as oxidation-reduction processes and respiratory electron transport chain (see table 4). These molecules were among the top genes from the p-value filtering procedure, and not present in the GRPS signature. In this respect, it is important to distinguish between functionality, in this case aiming at describing the biological relevance of the transcripts, and the GRPS prediction profile, aiming at performing accurate classification of independent samples. Several of the molecules involved in the oxidative phosphorylation and ubiquinone metabolism pathways are subunits of protein complexes, and thus spatially and temporally linked. The Backward Elimination procedure applied during feature selection in this study is based on orthogonal selection of variables, thus, features that did not contribute to orthogonal information were removed during this process. Therefore, it is not surprising, but rather expected, that some of the significantly regulated pathways did not contain, or only contained a few transcripts from the GRPS signature as e.g. subunits in a molecular complex will likely have a similar expression pattern. Based on several of the less activated pathways, the biological response in MUTZ-3 to chemical respiratory allergens involves also regulation of innate immune response signalling pathways that ultimately results in cell maturation, leading to enhanced antigen presentation and interaction with other immune cells. Furthermore, novel findings of usage of signalling pathways that has previously been associated with respiratory sensitization to protein allergens will shed light on the biological process leading to sensitization of the respiratory tract in response to chemical allergens. Thus, the GRPS is indeed relevant in an immunologically mechanistic perspective, and provides measurement of transcripts that monitor the biologic events leading to respiratory sensitization.

[0224] Further, results from enrichment analysis along with the results presented for the GARD assay, demonstrates that MUTZ-3 is a suitable model for prediction of both skin- and respiratory sensitizers. Despite some similarities in immunobiological mechanisms, important mechanistic differences exist between skin- and respiratory sensitization. Skin sensitization is primary associated with induction of Th1 cells, promoting a cytotoxic CD8+ T-cell response and secretion of IL-2 and interferon (IFN)-γ, while respiratory sensitization generally involve CD4+ Th2 cells and are characterized by high levels of IL-4, IL-10 and IL-13. Although respiratory sensitization to protein antigens are driven by the production of specific IgE antibody, it is still unclear what role the IgE antibody has during the development of respiratory allergy to chemical allergens, and whether there are mechanisms through which respiratory sensitization can be achieved that are independent of IgE antibody production [55,56]. It has been suggested that it may be sufficient with an induced Th2 response, without the need of IgE, to support the development of respiratory sensitization [57]. Although clear differences in T-cell responses, activation of dendritic cells (DCs) is common for both skin- and respiratory sensitization. Consequently, DCs are natural targets for assay development in terms of both skin and respiratory sensitization due to their physiological roles during initiation, modulation and polarization of immune responses in response to xenobiotic compounds. The MUTZ-3 cell line resembles primary dendritic cells (DCs) in terms of expression profile and ability to activate specific T-cell responses [45]. In comparison to primary DCs, MUTZ-3 are easy to grow using standardized protocols and provides a sustainable source of cells, offering an opportunity to scale up the assay to a high-throughput format.

[0225] In the context of developing an assay for both skin- and respiratory sensitization, it is important to acknowledge the formal semantics behind the nomenclature. Analogous to others [24,58-60], we use the terminology to indicate the local site of the immunological response and not the route of initial exposure in this study. For example, it has been shown that sensitization of the respiratory tract can arise also after dermal exposure [61-63] to relevant chemicals. In general, a certain chemical compound selectively sensitizes either the skin or the respiratory tract. However, during certain circumstances and in immunologically susceptible individuals, some chemicals have been shown to give rise to both type of sensitization. For example, the chemical triglycidylisocyanurate (TGIC) has been shown to cause both OA and ACD [64].

[0226] Finally, the approach of the GARD assay has several advantages, in comparison to other alternative methods. Using our data driven methodology, we were able to circumvent problems associated with the current shortage in knowledge regarding the exact mechanisms by how respiratory sensitizers provoke immunological responses in susceptible individuals. Secondly, the large amount of information obtained by the transcriptome-wide approach provides an additional opportunity to elucidate molecular mechanisms, such as specific signalling or metabolic pathways involved in the process of respiratory sensitization.

[0227] The major aim of this study was to develop an in vitro method in accordance with the three Rs principle on reduction, refinement and replacement of animal experiments for prediction of respiratory sensitization. Having trained a model with a set of reference chemicals, we present a tool to determine whether an unknown chemical is likely to behave as a non-respiratory sensitizer or a respiratory sensitizer. In the future, as the gaps in the current knowledge of how chemicals cause sensitization in the respiratory tract continues to be filled in, a consensus similar to the formulation of Adverse Outcome Pathways (AOP) for skin sensitization [67] may be a reality also for testing of respiratory sensitizers. The GRPS will then be an appealing part of an Integrated Testing Strategy (ITS), useful for assessment of DC maturation.

[0228] In conclusion, this study presents a predictive biomarker signature for classification of respiratory chemical sensitizers in MUTZ-3 cells that complement the previously described GARD assay for assessment of skin sensitizers. The ability to test for two different endpoints in the same sample provides an attractive and hitherto unique assay for safety assessment of chemicals in an in vitro testing strategy that comply with the three R principle on reduction, refinement and replacement of animal experiments.

REFERENCES

[0229] 1. Boverhof D R, Billington R, Gollapudi B B, Hotchkiss J A, Krieger S M, et al. (2008) Respiratory sensitization and allergy: current research approaches and needs. Toxicol Appl Pharmacol 226: 1-13.

[0230] 2. Banks D E, Tarlo S M (2000) Important issues in occupational asthma. Curr Opin Pulm Med 6: 37-42.

[0231] 3. Sastre J, Vandenplas O, Park H S (2003) Pathogenesis of occupational asthma. Eur Respir J 22: 364-373.

[0232] 4. Zammit-Tabona M, Sherkin M, Kijek K, Chan H, Chan-Yeung M (1983) Asthma caused by diphenylmethane diisocyanate in foundry workers. Clinical, bronchial provocation, and immunologic studies. Am Rev Respir Dis 128: 226-230.

[0233] 5. Bernstein D I, Patterson R, Zeiss C R (1982) Clinical and immunologic evaluation of trimellitic anhydride-and phthalic anhydride-exposed workers using a questionnaire with comparative analysis of enzyme-linked immunosorbent and radioimmunoassay studies. J Allergy Clin Immunol 69: 311-318.

[0234] 6. Murdoch R D, Pepys J, Hughes E G (1986) IgE antibody responses to platinum group metals: a large scale refinery survey. Br J Ind Med 43: 37-43.

[0235] 7. Docker A, Wattie J M, Topping M D, Luczynska C M, Newman Taylor A J, et al. (1987) Clinical and immunological investigations of respiratory disease in workers using reactive dyes. Br J Ind Med 44: 534-541.

[0236] 8. Bourne M S, Flindt M L, Walker J M (1979) Asthma due to industrial use of chloramine. Br Med J 2: 10-12.

[0237] 9. Tarlo S M, Lemiere C (2014) Occupational asthma. N Engl J Med 370: 640-649.

[0238] 10. EU (2008) Regulation (EC) no 1272/2008 of the European Parliament and of the Council on classification, labeling and packaging of substances and mixtures, amending and repealing Directives 67/768/EEC and 1999/45/EC, and amending Regulation (EC) no 1907/2006. Annex VI with tables 3.1 and 3.2. Official Journal of the European Union L353/1.

[0239] 11. Bharadwaj A, Agrawal D K (2004) Immunomodulation in asthma: a distant dream or a close reality? Int Immunopharmacol 4: 495-511.

[0240] 12. Elias J A, Zhu Z, Chupp G, Homer R J (1999) Airway remodeling in asthma. J Clin Invest 104: 1001-1006.

[0241] 13. Kopferschmitt-Kubler M C, Ameille J, Popin E, Calastreng-Crinquand A, Vervloet D, et al. (2002) Occupational asthma in France: a 1-yr report of the observatoire National de Asthmes Professionnels project. Eur Respir J 19: 84-89.

[0242] 14. Munoz X, Cruz M J, Orriols R, Bravo C, Espuga M, et al. (2003) Occupational asthma due to persulfate salts: diagnosis and follow-up. Chest 123: 2124-2129.

[0243] 15. Orriols R, Costa R, Albanell M, Alberti C, Castejon J, et al. (2006) Reported occupational respiratory diseases in Catalonia. Occup Environ Med 63: 255-260.

[0244] 16. Verstraelen S, Bloemen K, Nelissen I, Witters H, Schoeters G, et al. (2008) Cell types involved in allergic asthma and their use in in vitro models to assess respiratory sensitization. Toxicol In Vitro 22: 1419-1431.

[0245] 17. Pauluhn J, Eidmann P, Mohr U (2002) Respiratory hypersensitivity in guinea pigs sensitized to 1,6-hexamethylene diisocyanate (HDI): comparison of results obtained with the monomer and homopolymers of HDI. Toxicology 171: 147-160.

[0246] 18. Hilton J, Dearman R J, Boylett M S, Fielding I, Basketter D A, et al. (1996) The mouse IgE test for the identification of potential chemical respiratory allergens: considerations of stability and controls. J Appl Toxicol 16: 165-170.

[0247] 19. Dearman R J, Basketter D A, Kimber I (1992) Variable effects of chemical allergens on serum IgE concentration in mice. Preliminary evaluation of a novel approach to the identification of respiratory sensitizers. J Appl Toxicol 12: 317-323.

[0248] 20. Arts J H, Droge S C, Spanhaak S, Bloksma N, Penninks A H, et al. (1997) Local lymph node activation and IgE responses in brown Norway and Wistar rats after dermal application of sensitizing and non-sensitizing chemicals. Toxicology 117: 229-234.

[0249] 21. Warbrick E V, Dearman R J, Kimber I (2002) Induced changes in total serum IgE concentration in the Brown Norway rat: potential for identification of chemical respiratory allergens. J Appl Toxicol 22: 1-11.

[0250] 22. Dearman R J, Kimber I (2001) Cytokine fingerprinting and hazard assessment of chemical respiratory allergy. J Appl Toxicol 21: 153-163.

[0251] 23. Dearman R J, Kimber I (1999) Cytokine fingerprinting: characterization of chemical allergens. Methods 19: 56-63.

[0252] 24. Vanoirbeek J A, Tarkowski M, Vanhooren H M, De Vooght V, Nemery B, et al. (2006) Validation of a mouse model of chemical-induced asthma using trimellitic anhydride, a respiratory sensitizer, and dinitrochlorobenzene, a dermal sensitizer. J Allergy Clin Immunol 117: 1090-1097.

[0253] 25. Scheerens H, Buckley T L, Davidse E M, Garssen J, Nijkamp F P, et al. (1996) Toluene diisocyanate-induced in vitro tracheal hyperreactivity in the mouse. Am J Respir Crit Care Med 154: 858-865.

[0254] 26. (EU) D (2010) DIRECTIVE 2010/63/EU OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 22 Sep. 2010 on the protection of animals used for scientific purposes Official Journal of the European Union

[0255] 27. Verstraelen S, Nelissen I, Hooyberghs J, Witters H, Schoeters G, et al. (2009) Gene profiles of THP-1 macrophages after in vitro exposure to respiratory (non-) sensitizing chemicals: identification of discriminating genetic markers and pathway analysis. Toxicol In Vitro 23: 1151-1162.

[0256] 28. Verstraelen S, Nelissen I, Hooyberghs J, Witters H, Schoeters G, et al. (2009) Gene profiles of a human bronchial epithelial cell line after in vitro exposure to respiratory (non-)sensitizing chemicals: identification of discriminating genetic markers and pathway analysis. Toxicology 255: 151-159.

[0257] 29. Verstraelen S, Nelissen I, Hooyberghs J, Witters H, Schoeters G, et al. (2009) Gene profiles of a human alveolar epithelial cell line after in vitro exposure to respiratory (non-)sensitizing chemicals: identification of discriminating genetic markers and pathway analysis. Toxicol Lett 185: 16-22.

[0258] 30. Huang S, Wiszniewski L, Constant S, Roggen E (2013) Potential of in vitro reconstituted 3D human airway epithelia (MucilAir) to assess respiratory sensitizers. Toxicol In Vitro 27: 1151-1156.

[0259] 31. Lalko J F, Kimber I, Dearman R J, Gerberick G F, Sarlo K, et al. (2011) Chemical reactivity measurements: potential for characterization of respiratory chemical allergens. Toxicol In Vitro 25: 433-445.

[0260] 32. Isola D, Kimber I, Sarlo K, Lalko J, Sipes I G (2008) Chemical respiratory allergy and occupational asthma: what are the key areas of uncertainty? J Appl Toxicol 28: 249-253.

[0261] 33. Kimber I, Agius R, Basketter D A, Corsini E, Cullinan P, et al. (2007) Chemical respiratory allergy: opportunities for hazard identification and characterisation. The report and recommendations of ECVAM workshop 60. Altern Lab Anim 35: 243-265.

[0262] 34. Mehling A, Eriksson T, Eltze T, Kolle S, Ramirez T, et al. (2012) Non-animal test methods for predicting skin sensitization potentials. Arch Toxicol 86: 1273-1295.

[0263] 35. Basketter D A, Evans P, Fielder R J, Gerberick G F, Dearman R J, et al. (2002) Local lymph node assay—validation, conduct and use in practice. Food Chem Toxicol 40: 593-598.

[0264] 36. Ashikaga T, Yoshida Y, Hirota M, Yoneyama K, Itagaki H, et al. (2006) Development of an in vitro skin sensitization test using human cell lines: the human Cell Line Activation Test (h-CLAT). I. Optimization of the h-CLAT protocol. Toxicol In Vitro 20: 767-773.

[0265] 37. Nukada Y, Ashikaga T, Sakaguchi H, Sono S, Mugita N, et al. (2011) Predictive performance for human skin sensitizing potential of the human cell line activation test (h-CLAT). Contact Dermatitis 65: 343-353.

[0266] 38. Gerberick G F, Vassallo J D, Bailey R E, Chaney J G, Morrell S W, et al. (2004) Development of a peptide reactivity assay for screening contact allergens. Toxicol Sci 81: 332-343.

[0267] 39. Natsch A (2010) The Nrf2-Keap1-ARE toxicity pathway as a cellular sensor for skin sensitizers—functional relevance and a hypothesis on innate reactions to skin sensitizers. Toxicol Sci 113: 284-292.

[0268] 40. Andreas N, Caroline B, Leslie F, Frank G, Kimberly N, et al. (2011) The intra- and inter-laboratory reproducibility and predictivity of the KeratinoSens assay to predict skin sensitizers in vitro: results of a ring-study in five laboratories. Toxicol In Vitro 25: 733-744.

[0269] 41. Johansson H, Lindstedt M, Albrekt A S, Borrebaeck C A (2011) A genomic biomarker signature can predict skin sensitizers using a cell-based in vitro alternative to animal tests. BMC Genomics 12: 399.

[0270] 42. Johansson H, Albrekt A S, Borrebaeck C A, Lindstedt M (2013) The GARD assay for assessment of chemical skin sensitizers. Toxicol In Vitro 27: 1163-1169.

[0271] 43. Santegoets S J, Masterson A J, van der Sluis P C, Lougheed S M, Fluitsma D M, et al. (2006) A CD34(+) human cell line model of myeloid dendritic cell differentiation: evidence for a CD14(+)CD11b(+) Langerhans cell precursor. J Leukoc Biol 80: 1337-1344.

[0272] 44. Masterson AJ, Sombroek C C, De Gruijl T D, Graus Y M, van der Vliet H J, et al. (2002) MUTZ-3, a human cell line model for the cytokine-induced differentiation of dendritic cells from CD34+ precursors. Blood 100: 701-703.

[0273] 45. Larsson K, Lindstedt M, Borrebaeck C A (2006) Functional and transcriptional profiling of MUTZ-3, a myeloid cell line acting as a model for dendritic cells. Immunology 117: 156-166.

[0274] 46. Johansson H, Rydnert F, Kuhnl J, Schepky A, Borrebaeck C, et al. (2014) Genomic allergen rapid detection in-house validation—a proof of concept. Toxicol Sci 139: 362-370.

[0275] 47. Haneke K E, Tice R R, Carson B L, Margolin B H, Stokes W S (2001) ICCVAM evaluation of the murine local lymph node assay. Data analyses completed by the National Toxicology Program Interagency Center for the Evaluation of Alternative Toxicological Methods. Regul Toxicol Pharmacol 34: 274-286.

[0276] 48. Vanoirbeek J A, Mandervelt C, Cunningham A R, Hoet P H, Xu H, et al. (2003) Validity of methods to predict the respiratory sensitizing potential of chemicals: A study with a piperidinyl chlorotriazine derivative that caused an outbreak of occupational asthma. Toxicol Sci 76: 338-346.

[0277] 49. Holsapple M P, Jones D, Kawabata T T, Kimber I, Sarlo K, et al. (2006) Assessing the potential to induce respiratory hypersensitivity. Toxicol Sci 91: 4-13.

[0278] 50. Kimber I, Basketter D A, Butler M, Gamer A, Garrigue J L, et al. (2003) Classification of contact allergens according to potency: proposals. Food Chem Toxicol 41: 1799-1809.

[0279] 51. OECD (2002) Guideline 429: Skin Sensitization: Local Lymph Node Assay (2002).

[0280] 52. Carlsson A, Wingren C, Kristensson M, Rose C, Ferno M, et al. (2011) Molecular serum portraits in patients with primary breast cancer predict the development of distant metastases. Proc Natl Acad Sci USA 108: 14252-14257.

[0281] 53. Kullback S, Leibler RA (1951) On Information and Sufficiency. Annals of Mathematical Statistics 22: 79-86.

[0282] 54. Albrekt A S, Johansson H, Borje A, Borrebaeck C, Lindstedt M (2014) Skin sensitizers differentially regulate signaling pathways in MUTZ-3 cells in relation to their individual potency. BMC Pharmacol Toxicol 15: 5.

[0283] 55. Cartier A, Grammer L, Malo J L, Lagier F, Ghezzo H, et al. (1989) Specific serum antibodies against isocyanates: association with occupational asthma. J Allergy Clin Immunol 84: 507-514.

[0284] 56. Cullinan P (1998) Occupational asthma, IgE and IgG. Clin Exp Allergy 28: 668-670.

[0285] 57. Kimber I, Dearman R J, Basketter D A, Boverhof D R (2014) Chemical respiratory allergy: reverse engineering an adverse outcome pathway. Toxicology 318: 32-39.

[0286] 58. Vanoirbeek J A, Tarkowski M, Ceuppens J L, Verbeken E K, Nemery B, et al. (2004) Respiratory response to toluene diisocyanate depends on prior frequency and concentration of dermal sensitization in mice. Toxicol Sci 80: 310-321.

[0287] 59. Sailstad D M, Ward M D, Boykin E H, Selgrade M K (2003) A murine model for low molecular weight chemicals: differentiation of respiratory sensitizers (TMA) from contact sensitizers (DNFB). Toxicology 194: 147-161.

[0288] 60. Pauluhn J (2003) Respiratory hypersensitivity to trimellitic anhydride in Brown Norway rats: analysis of dose-response following topical induction and time course following repeated inhalation challenge. Toxicology 194: 1-17.

[0289] 61. Karol M H, Hauth B A, Riley E J, Magreni C M (1981) Dermal contact with toluene diisocyanate (TDI) produces respiratory tract hypersensitivity in guinea pigs. Toxicol Appl Pharmacol 58: 221-230.

[0290] 62. Kimber I, Dearman R J (2002) Chemical respiratory allergy: role of IgE antibody and relevance of route of exposure. Toxicology 181-182: 311-315.

[0291] 63. Bello D, Herrick C A, Smith T J, Woskie S R, Streicher R P, et al. (2007) Skin exposure to isocyanates: reasons for concern. Environ Health Perspect 115: 328-335.

[0292] 64. Meuleman L, Goossens A, Linders C, Rochette F, Nemery B (1999) Sensitization to triglycidylisocyanurate (TGIC) with cutaneous and respiratory manifestations. Allergy 54: 752-756.

[0293] 65. Cary R, Dobson S, Delic J (1999) Concise International Chemical Assessment Document 15 (CICAD): 1,2-DIAMINOETHANE (ETHYLENEDIAMINE). World Health Organization.

[0294] 66. Commission E (2005) European Union Risk Assessment Report Piperazine. European Chemicals Bureau: European comission. 159 p.

[0295] 67. OECD (2012) The Adverse Outcome Pathway for Skin Sensitisation Initiated by Covalent Binding to Proteins, Part 1: Scientific Evidence. Environment, Health and Safety Publications, Series on Testing and Assessment No. 168.

[0296] 68. Hu Z B, Ma W, Zaborski M, MacLeod R, Quentmeier H, et al. (1996) Establishment and characterization of two novel cytokine-responsive acute myeloid and monocytic leukemia cell lines, MUTZ-2 and MUTZ-3. Leukemia 10: 1025-1040.

[0297] 69. Quentmeier H, Duschl A, Hu Z B, Schnarr B, Zaborski M, et al. (1996) MUTZ-3, a monocytic model cell line for interleukin-4 and lipopolysaccharide studies. Immunology 89: 606-612.

[0298] 70. Piccolo S R, Sun Y, Campbell J D, Lenburg M E, Bild A H, et al. (2012) A single-sample microarray normalization method to facilitate personalized-medicine workflows. Genomics 100: 337-344.

[0299] 71. Johnson W E, Li C, Rabinovic A (2007) Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 8: 118-127.

[0300] 72. R Development Core Team (2014) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria.

[0301] 73. Gentleman R C, Carey V J, Bates D M, Bolstad B, Dettling M, et al. (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80.

[0302] 74. Leek J T, Johnson W E, Parker H S, Jaffe A E, Storey J D sva: Surrogate Variable Analysis. R package version 3.6.0.

[0303] 75. Ringner M (2008) What is principal component analysis? Nat Biotechnol 26: 303-304.

[0304] 76. Benjamini Y H Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B (Methodological) 57: 12.

[0305] 77. Lasko T A, Bhagwat J G, Zou K H, Ohno-Machado L (2005) The use of receiver operating characteristic curves in biomedical informatics. J Biomed Inform 38: 404-415.

[0306] 78. Dimitriadou E, Hornik, K., Leisch, F., Meyer, D., Weingessel, A. (2011) e1071: Misc functions of the Department of statistics (q1071). TU Wien R package version 16 http://CRAN.R-project.org/package=e1071.

[0307] 79. Cortes C V, V. (1995) Support-Vector Networks. Machine Learning 20: 273-297.

[0308] 80. Sing T, Sander 0, Beerenwinkel N, Lengauer T (2005) ROCR: visualizing classifier performance in R. Bioinformatics 21: 3940-3941.

[0309] 81. Cooper J A, 2nd, Saracci R, Cole P (1979) Describing the validity of carcinogen screening tests. Br J Cancer 39: 87-89.

TABLE-US-00001 TABLE A Biomarkers from the GPRS Prediction Signature. Validation Affymetrix call Gene ID Ensembl Transcript ID Probe set ID frequency.sup.1 Table A(i) - Core biomarkers 1. TNFRSF19 ENST00000403372 7968015 100 2. SNORA74A NR_002915 8108420 100 3. SPAM1 ENST00000340011 8135835 100 Table A(ii) - Preferred biomarkers 4. ETID: ENST00000364621 ENST00000364621 7917972 95 5. HOMER3 ENST00000392351 8035566 95 6. CD1C ENST00000368169 7906348 90 7. IGHD /// IGHM ENST00000390538 7981601 90 8. SNRPN /// SNORD116-26 NR_003340 7982000 90 9. ETID: ENST00000364678 ENST00000364678 7934896 85 10. STRAP ENST00000025399 7954173 85 11. DIABLO ENST00000267169 7967230 85 12. ETID: ENST00000411349 ENST00000411349 8151989 85 13. ETID: ENST00000385497 ENST00000385497 7923037 80 14. OR51A2 ENST00000380371 7946017 80 15. MRPL21 ENST00000362034 7949995 80 16. PPP1R14A ENST00000301242 8036473 80 17. DEFB127 ENST00000382388 8060314 80 18. C9orf130 ENST00000375268 8162562 80 19. PRO2012 BC019830 7924817 75 20. LOC399898 AK128188 7940116 75 21. ETID: ENST00000387701 ENST00000387701 7969914 75 22. WDR68 ENST00000310827 8009164 75 23. NEU2 ENST00000233840 8049243 75 24. ETID: ENST00000386677 ENST00000386677 8072575 75 25. SPARC ENST00000231061 8115327 75 26. ETID: ENST00000390342 ENST00000390342 8139107 75 27. CRNN ENST00000271835 7920178 70 28. MMP12 ENST00000326227 7951297 70 29. ACVRL1 ENST00000267008 7955562 70 30. EIF4E2 ENST00000258416 8049180 70 31. RP11-191L9.1 ENST00000380990 8076819 70 32. PDCD6 /// AHRR ENST00000264933 8104180 70 33. ARRDC3 ENST00000265138 8113073 70 34. VWDE ENST00000275358 8138258 70 35. ZBTB34 ENST00000319119 8157945 70 36. ITGB1BP2 ENST00000373829 8168291 70 37. OR10K2 ENST00000392265 7921356 65 38. FLJ22596 AK026249 7950442 65 39. ETID: ENST00000306515 ENST00000306515 8043572 65 40. ACVR2A ENST00000404590 8045587 65 41. ETID: ENST00000385690 ENST00000385690 8092312 65 42. ETID: ENST00000386018 ENST00000386018 8097945 65 43. C6orf201 ENST00000360378 8116696 65 44. ETID: ENST00000385583 ENST00000385583 8136932 65 45. ETID: ENST00000385719 ENST00000385719 8148515 65 46. GPR20 ENST00000377741 8153269 65 47. ETID: ENST00000364357 ENST00000364357 8163084 65 48. ZCCHC13 ENST00000339534 8168420 65 49. GPR64 ENST00000356606 8171624 65 50. CD1D ENST00000368171 7906330 60 51. DUSP12 ENST00000367943 7906810 60 52. KLHL33 ENST00000344581 7977567 60 53. PSMB6 ENST00000270586 8003953 60 54. TMEM95 ENST00000396580 8004364 60 55. C1QBP ENST00000225698 8011850 60 56. EMILIN2 ENST00000254528 8019912 60 57. CD8A ENST00000352580 8053584 60 58. C20orf152 ENST00000349339 8062237 60 59. KCNJ4 ENST00000303592 8076072 60 60. ETID: ENST00000364163 ENST00000364163 8078310 60 61. FAM19A1 ENST00000327941 8080918 60 62. ETID: ENST00000384601 ENST00000384601 8081233 60 63. POLR2H ENST00000296223 8084488 60 64. AK000420 AK000420 8110706 60 65. ETID: ENST00000363354 ENST00000363354 8120360 60 66. APID: 8121483 — 8121483 60 67. EGFL6 ENST00000361306 8166079 60 68. POU3F4 ENST00000373200 8168567 60 69. ETID: ENST00000385841 ENST00000385841 7905629 55 70. OR52A5 ENST00000307388 7946023 55 71. TIMM8B ENST00000280354 7951679 55 72. PEBP1 ENST00000261313 7959070 55 73. OR4F6 ENST00000328882 7986530 55 74. CDH15 ENST00000289746 7997880 55 75. TMEM199 ENST00000292114 8005857 55 76. ABI3 ENST00000225941 8008185 55 77. FLJ42842 AK124832 8008540 55 78. MC4R ENST00000299766 8023593 55 79. ETID: ENST00000410673 ENST00000410673 8045931 55 80. ISM1 ENST00000262487 8061013 55 81. LOC440957 ENST00000307106 8080416 55 82. KLB ENST00000257408 8094679 55 83. GM2A ENST00000357164 8109344 55 84. ANXA6 ENST00000354546 8115234 55 85. TAS2R40 ENST00000408947 8136846 55 86. APID: 8142880 — 8142880 55 87. RARRES2 ENST00000223271 8143772 55 88. SH2D4A ENST00000265807 8144880 55 89. PLP1 ENST00000361621 8169061 55 90. ATP1A2 ENST00000392233 7906501 50 91. ETID: ENST00000386800 ENST00000386800 7932610 50 92. MAT1A ENST00000372206 7934755 50 93. TSGA10IP ENST00000312452 7941469 50 94. PRDM7 ENST00000325921 8003571 50 95. ETID: ENST00000390847 ENST00000390847 8015739 50 96. ETID: ENST00000255183 ENST00000255183 8066444 50 97. MRPL39 ENST00000307301 8069620 50 98. ETID: ENST00000386327 ENST00000386327 8074884 50 99. TIPARP ENST00000295924 8083569 50 100. HES1 ENST00000232424 8084880 50 101. ETID: ENST00000363502 ENST00000363502 8089727 50 102. PRDM9 ENST00000253473 8104634 50 103. ETID: ENST00000390917 ENST00000390917 8137433 50 104. KIAA1688 ENST00000377307 8153876 50 105. ETID: ENST00000391219 ENST00000391219 8156759 50 106. ETID: ENST00000387973 ENST00000387973 8160782 50 107. LOC100129534 — 7911718 45 108. SLC2A1 ENST00000397019 7915472 45 109. AF116714 AF116714 7935359 45 110. EPS8L2 ENST00000318562 7937443 45 111. MGC3196 ENST00000307366 7948836 45 112. 7952733 — 7952733 45 113. ETID: ENST00000384391 ENST00000384391 7990031 45 114. EME2 ENST00000307394 7992379 45 115. NETO1 ENST00000299430 8023828 45 116. NPHS1 ENST00000353632 8036176 45 117. ETID: ENST00000384109 ENST00000384109 8047215 45 118. ETID: ENST00000364143 ENST00000364143 8059799 45 119. ISX ENST00000404699 8072636 45 120. IL17RB ENST00000288167 8080562 45 121. PCOLCE2 ENST00000295992 8091243 45 122. LRIT3 ENST00000409621 8096839 45 123. ETID: ENST00000330110 ENST00000330110 8104615 45 124. ZNF354C ENST00000315475 8110491 45 125. ETID: ENST00000386444 ENST00000386444 8162927 45 126. OR2G3 ENST00000320002 7911209 40 127. GLUL ENST00000331872 7922689 40 128. CCKBR ENST00000334619 7938090 40 129. OR1S2 ENST00000302592 7948312 40 130. DCUN1D5 ENST00000260247 7951325 40 131. ETID: ENST00000388291 ENST00000388291 7951420 40 132. EMG1 ENST00000261406 7953594 40 133. PTHLH ENST00000395868 7962000 40 134. PTGES3 ENST00000262033 7964250 40 135. CIDEB ENST00000258807 7978272 40 136. ETID: ENST00000383863 ENST00000383863 7985918 40 137. ATP10A ENST00000356865 7986789 40 138. MYO5C ENST00000261839 7988876 40 139. ETID: ENST00000380078 ENST00000380078 7989951 40 140. PLA2G10 ENST00000261659 7999588 40 141. HSPE1 ENST00000409729 8047223 40 142. ETID: ENST00000388324 ENST00000388324 8096249 40 143. MYO6 ENST00000369977 8120783 40 144. C7orf30 ENST00000287543 8131860 40 145. ETID: ENST00000340779 ENST00000340779 8139828 40 146. LOC441245 AK090474 8139887 40 147. CRIM2 ENST00000297801 8142821 40 148. XKR4 ENST00000327381 8146475 40 149. FAM110B ENST00000361488 8146533 40 150. PEBP4 ENST00000256404 8149725 40 151. LOC644714 BC047037 8161943 40 152. PAPPAS AY623011 /// AY623012 8163672 40 153. BEX4 ENST00000372691 8169009 40 154. HMGB4 ENST00000323936 7899905 35 155. ETID: BC028413 /// BC128516 BC028413 /// BC128516 7911676 35 156. ETID: ENST00000363919 ENST00000363919 7928750 35 157. ETID: ENST00000335621 ENST00000335621 7958942 35 158. SOX1 ENST00000330949 7970146 35 159. CTSG ENST00000216336 7978351 35 160. ETID: ENST00000362344 ENST00000362344 7982100 35 161. FLJ37464 ENST00000398354 7996377 35 162. RAX ENST00000334889 8023549 35 163. IL29 ENST00000333625 8028613 35 164. CEACAM20 ENST00000316962 8037482 35 165. ETID: ETID: ENST00000365557 ENST00000365557 8044684 35 166. SEC14L3 ENST00000403066 8075375 35 167. C3orf52 ENST00000264848 8081645 35 168. FETUB ENST00000265029 8084657 35 169. PIGY ENST00000273968 8101718 35 170. CDH12 ENST00000284308 8111234 35 171. LGSN ENST00000370657 8127380 35 172. ETID: ENST00000391031 ENST00000391031 8129067 35 173. HGC6.3 AB016902 8130824 35 174. tcag7.873 NM_001126493 8138797 35 175. T1560 ENST00000379496 8146527 35 176. EXOSC4 ENST00000316052 8148710 35 177. TRAM1 ENST00000262213 8151281 35 178. APID: 8159371 — 8159371 35 179. OR13C2 ENST00000318797 8162936 35 180. PLS3 ENST00000289290 8169473 35 181. TMEM53 ENST00000372244 7915578 30 182. CD1B ENST00000368168 7921346 30 183. SORCS3 ENST00000393176 7930341 30 184. OR52E8 ENST00000329322 7946111 30 185. FAM160A2 ENST00000265978 7946128 30 186. LOC649946 BC017930 7952126 30 187. FAM158A ENST00000216799 7978114 30 188. APID: 7986637 — 7986637 30 189. MYO1E ENST00000288235 7989277 30 190. NUPR1 ENST00000395641 8000574 30 191. APID: 8005433 — 8005433 30 192. SIGLEC15 ENST00000389474 8021091 30 193. 2-Mar ENST00000393944 8025421 30 194. LOC100131554 — 8041886 30 195. GGTLC1 ENST00000335694 8065427 30 196. PSMA7 ENST00000395567 8067382 30 197. SLC25A18 ENST00000399813 8071107 30 198. C3orf14 ENST00000232519 8080847 30 199. CDX1 ENST00000377812 8109226 30 200. ETID: ENST00000386433 ENST00000386433 8121249 30 201. RRAGD ENST00000359203 8128123 30 202. SDK1 ENST00000389531 8131205 30 203. LOC168474 NR_002789 8139826 30 204. ETID: ENST00000384125 ENST00000384125 8146120 30 205. TRHR ENST00000311762 8147877 30 206. IL11RA ENST00000378817 8154934 30 207. MGC21881 /// LOC554249 ENST00000377616 8155393 30 208. ZNF483 ENST00000358151 8157193 30 209. C9orf169 ENST00000400709 8159624 30 210. MGC21881 /// LOC554249 ENST00000377616 8161451 30 211. ETID: ENST00000364507 ENST00000364507 8168161 30 212. ETID: ENST00000387003 ENST00000387003 7914137 25 213. ETID: ENST00000388083 ENST00000388083 7929614 25 214. ETID: ENST00000365084 ENST00000365084 7934568 25 215. FRG2 /// FRG2B /// FRG2C ENST00000368515 7937251 25 216. C14orf53 ENST00000389594 7975154 25 217. ODF3L1 ENST00000332145 7985025 25 218. FAM18A ENST00000299866 7999412 25 219. PRTN3 ENST00000234347 8024048 25 220. CFD ENST00000327726 8024062 25 221. TMED1 ENST00000214869 8034101 25 222. ETID: ENST00000387150 ENST00000387150 8035937 25 223. HSD17B14 ENST00000263278 8038213 25 224. BOK ENST00000318407 8049876 25 225. ETID: ENST00000365609 ENST00000365609 8050801 25 226. SNRPB ENST00000381342 8064502 25 227. EPHA6 ENST00000338994 8081138 25 228. SCARNA22 NR_003004 8093576 25 229. FLJ35424 ENST00000404649 8093821 25 230. ETID: ENST00000387555 ENST00000387555 8104723 25 231. ETID: ENST00000388664 ENST00000388664 8107115 25 232. ETID: ENST00000363365 ENST00000363365 8108566 25 233. ETID: ENST00000362861 ENST00000362861 8111358 25 234. ETID: ENST00000363181 ENST00000363181 8114581 25 235. GRM6 ENST00000319065 8116253 25 236. LOC646093 — 8116400 25 237. HIST1H1E ENST00000304218 8117377 25 238. TIAM2 ENST00000367174 8122933 25 239. ETID: ENST00000363074 ENST00000363074 8128712 25 240. ETID: ENST00000385777 ENST00000385777 8148331 25 241. MTUS1 ENST00000400046 8149500 25 242. MUC21 ENST00000383351 8177931 25 243. WDR8 ENST00000378322 7911839 20 244. LOC100131195 AK097743 7933190 20 245. OR4D10 ENST00000378245 7940182 20 246. C12orf63 ENST00000342887 7957688 20 247. ELA1 ENST00000293636 7963304 20 248. DNAJC14 /// CIP29 ENST00000317269 7963935 20 249. FLJ40176 ENST00000322527 7972670 20 250. ETID: ENST00000410207 ENST00000410207 7985308 20 251. PSME3 ENST00000293362 8007397 20 252. ETID: ENST00000405656 ENST00000405656 8009515 20 253. HN1 ENST00000356033 8018305 20 254. ETID: ENST00000335523 ENST00000335523 8027385 20 255. CYP2A7 /// CYP2A7P1 ENST00000301146 8036981 20 256. ATXN10 ENST00000252934 8073799 20 257. ZMAT5 ENST00000397779 8075276 20 258. ETID: ENST00000362493 ENST00000362493 8084215 20 259. FHIT ENST00000341848 8088458 20 260. FRG2 /// FRG2B /// FRG2C ENST00000368515 8104124 20 261. SNX18 ENST00000381410 8105328 20 262. ETID: ENST00000362433 ENST00000362433 8128445 20 263. DTX2 ENST00000307569 8133736 20 264. ASB4 ENST00000325885 8134376 20 265. ETID: ENST00000365242 ENST00000365242 8147445 20 266. ETID: ENST00000364204 ENST00000364204 8156450 20 267. COL5A1 ENST00000355306 8159142 20 268. LCAP ENST00000357566 8170786 20 269. APOO ENST00000379226 8171823 20 270. PTPRU ENST00000373779 7899562 15 271. IL28RA ENST00000327535 7913776 15 272. NEUROG3 ENST00000242462 7934083 15 273. VAX1 ENST00000277905 7936552 15 274. LOC440131 ENST00000400540 7968323 15 275. C13orf31 ENST00000325686 7968883 15 276. ADAMTS7 ENST00000388820 7990736 15 277. SMTNL2 ENST00000338859 8003892 15 278. LOC284112 AK098506 8012004 15 279. ETV2 ENST00000402764 8027920 15 280. FUT2 ENST00000391876 8030094 15 281. C2orf39 ENST00000288710 8040672 15 282. LOC200383 /// DNAH6 ENST00000237449 8043071 15 283. ETID: ENST00000385676 ENST00000385676 8055204 15 284. CCDC108 ENST00000341552 8059028 15 285. APID: 8065011 — 8065011 15 286. C22orf27 BC042980 8072400 15 287. ETID: ENST00000364444 ENST00000364444 8103041 15 288. PDLIM3 ENST00000284767 8104022 15 289. ETID: ENST00000330110 ENST00000330110 8104613 15 290. ETID: ENST00000384539 ENST00000384539 8107125 15 291. ETID: ENST00000390214 ENST00000390214 8130372 15 292. MGC72080 BC029615 8141169 15 293. C9orf128 ENST00000377984 8161154 15 294. RGAG4 NM_001024455 8173503 15 295. PIP5K1A ENST00000409426 7905365 10 296. GPR161 ENST00000367838 7922108 10 297. ETID: ENST00000385353 ENST00000385353 7925434 10 298. OR56A3 ENST00000329564 7938066 10 299. OR5A2 ENST00000302040 7948377 10 300. WNT11 ENST00000322563 7950534 10 301. APID: 7960259 — 7960259 10 302. RAB37 ENST00000340415 8009666 10 303. LAIR1 ENST00000391742 8039257 10 304. ETID: ENST00000388385 ENST00000388385 8041420 10 305. CHAC2 ENST00000295304 8041961 10 306. ETID: ENST00000387574 ENST00000387574 8062337 10 307. ETID: ENST00000387884 ENST00000387884 8062962 10 308. BCL2L1 ENST00000376062 8065569 10 309. KDELR3 ENST00000409006 8073015 10 310. TMEM108 ENST00000321871 8082767 10 311. SPATA16 ENST00000351008 8092187 10 312. BTC ENST00000395743 8101002 10 313. SUPT3H ENST00000371460 8126710 10 314. EIF4B ENST00000262056 8135268 10 315. CHMP4C ENST00000297265 8147057 10 316. H2BFM ENST00000243297 8169080 10 317. APID: 8180392 — 8180392 10 318. NR5A2 ENST00000367362 7908597 5 319. TRIM49 ENST00000332682 7939884 5 320. MS4A6A ENST00000323961 7948455 5 321. C11orf10 ENST00000257262 7948606 5 322. HSPC152 ENST00000308774 7949075 5 323. RASAL1 ENST00000261729 7966542 5 324. ETID: ENST00000387531 ENST00000387531 7975694 5 325. PLDN ENST00000220531 7983502 5 326. PER1 ENST00000354903 8012349 5 327. ALS2CR12 ENST00000286190 8058203 5 328. C20orf142 ENST00000396825 8066407 5 329. ETID: ENST00000386848 ENST00000386848 8073680 5 330. LOC100129113 AK094477 8074307 5 331. CERK ENST00000216264 8076792 5 332. ETID: ENST00000385783 ENST00000385783 8083937 5 333. PROS1 ENST00000407433 8089015 5 334. PCDHGA ENST00000378105 8108757 5 335. MUC3B /// MUC3A ENST00000332750 8135015 5 336. ETID: ENST00000365355 ENST00000365355 8142534 5 337. APID: 8156969 — 8156969 5 338. ETID: ENST00000358047 ENST00000410626 8163013 5 339. FAM47C ENST00000358047 8166703 5 340. NXF4 ENST00000360035 8168940 5 341. PIWIL4 ENST00000299001 7943240 0 342. ETID: ENST00000384727 ENST00000384727 7968732 0 343. ALDH6A1 ENST00000350259 7980098 0 344. TNAENA64 ENST00000324979 8151747 0 345. ETID: ENST00000364816 ENST00000364816 8168079 0 Table A(iii) - Optional biomarkers 346. C11orf73 ENST00000278483 7942932 100 347. OR5B21 ENST00000278483 7948330 100 348. NOX5 /// SPESP1 ENST00000395421 7984488 100 349. AMICA1 ENST00000356289 7952022 95 350. ETID: ENST00000387422 ENST00000387422 8159963 90 351. SERPINB1 ENST00000380739 8123598 85 352. ETID: ENST00000387396 ENST00000387396 8065752 80 353. CD1A ENST00000289429 7906339 75 354. RAB9A ENST00000243325 8166098 75 355. C10orf90 ENST00000356858 7936996 70 356. LPXN ENST00000263845 7948332 65 357. GGTLC2 ENST00000215938 8071662 65 358. ETID: ENST00000384680 ENST00000384680 8051862 60 359. PNPLA4 ENST00000381042 8171229 60 360. CAMK1D ENST00000378845 7926223 55 361. ETID: ENST00000410754 ENST00000410754 8120979 55 362. CDC123 ENST00000281141 7926207 50 363. WDFY1 ENST00000233055 8059361 50 364. hCG_1749005 — 8167640 50 365. CD48 ENST00000368046 7921667 45 366. MED19 ENST00000337672 7948293 45 367. DRD5 ENST00000304374 8053725 45 368. APID: 7967586 — 7967586 40 369. VAPA ENST00000340541 8020129 40 370. FAM71F1 ENST00000315184 8135945 40 371. APID: 8141421 — 8141421 35 372. HCCS ENST00000321143 8165995 35 373. CNR2 ENST00000374472 7913705 25 374. OIT3 ENST00000334011 7928330 25 375. BMP2K ENST00000335016 8096004 25 376. ZNF366 ENST00000318442 8112584 25 377. SYT17 ENST00000396244 7993624 20 378. CALM12 ENST00000272298 8052010 20 379. XK ENST00000378616 8166723 20 380. ART4 ENST00000228936 7961507 15 381. ETID: ENST00000332418 ENST00000332418 7997907 15 382. ZFP36L2 ENST00000282388 8051814 15 383. GSTA3 ENST00000370968 8127087 15 384. COL21A1 ENST00000370817 8127201 15 385. ETID: ENST00000332418 ENST00000332418 8170322 15 386. FUCA1 ENST00000374479 7913694 5 387. ETID: ENST00000386628 ENST00000386628 7925821 5 388. AZU1 ENST00000334630 8024038 5 389. IL7R ENST00000303115 8104901 5

[0310] The table shows predictor genes in GRPS, identified by one-way ANOVA p-value filtering and Backward elimination. When possible, the Ensembl transcript ID was used as gene identifier. The Affymetrix Probe Set ID for the Human ST 1.0 Array are provided.

[0311] .sup.1Validation call frequency (%) describes the occurrence of each predictor transcript among the 20 biomarker signatures obtained by cross validation.

TABLE-US-00002 TABLE B GRPS predictor gene expression trends upon sensitizer exposure Probe ID Up/down 7899905 down 7905365 up 7905629 up 7906330 up 7906339 up 7906348 up 7906501 down 7906810 up 7908597 up 7911209 down 7911676 up 7911718 up 7911839 up 7913694 up 7913705 up 7913776 down 7914137 up 7915472 up 7915578 up 7917972 down 7920178 down 7921346 up 7921356 up 7921667 up 7922108 down 7922689 down 7923037 up 7924817 down 7925434 down 7925821 up 7926207 up 7926223 up 7928330 up 7928750 up 7929614 up 7930341 down 7932610 up 7933190 up 7934083 down 7934568 up 7934755 up 7934896 up 7935359 down 7936552 down 7936996 up 7937251 up 7937443 up 7938066 up 7938090 down 7939884 up 7940116 down 7940182 up 7941469 down 7942932 up 7943240 down 7946017 up 7946023 down 7946111 down 7946128 down 7948293 up 7948312 down 7948330 up 7948332 up 7948377 up 7948455 up 7948606 up 7948836 up 7949075 up 7949995 up 7950442 down 7950534 up 7951297 up 7951325 up 7951420 up 7951679 up 7952022 up 7952126 up 7952733 up 7953594 up 7954173 up 7955562 up 7957688 up 7958942 up 7959070 up 7960259 up 7961507 up 7962000 down 7963304 down 7963935 up 7964250 up 7966542 up 7967230 up 7967586 down 7968015 down 7968323 up 7968732 up 7968883 up 7969914 up 7970146 down 7972670 up 7975154 up 7975694 down 7977567 down 7978114 up 7978272 up 7978351 down 7980098 down 7981601 down 7982000 down 7982100 up 7983502 down 7984488 up 7985025 up 7985308 up 7985918 down 7986530 down 7986637 down 7986789 up 7988876 up 7989277 up 7989951 down 7990031 down 7990736 down 7992379 down 7993624 down 7996377 up 7997880 down 7997907 up 7999412 up 7999588 up 8000574 down 8003571 up 8003892 up 8003953 up 8004364 up 8005433 up 8005857 up 8007397 up 8008185 up 8008540 down 8009164 up 8009515 up 8009666 down 8011850 up 8012004 up 8012349 down 8015739 down 8018305 up 8019912 up 8020129 up 8021091 down 8023549 down 8023593 down 8023828 up 8024038 down 8024048 down 8024062 down 8025421 down 8027385 down 8027920 down 8028613 up 8030094 up 8034101 up 8035566 down 8035937 down 8036176 down 8036473 down 8036981 down 8037482 down 8038213 down 8039257 down 8040672 up 8041420 down 8041886 down 8041961 up 8043071 up 8043572 up 8044684 up 8045587 up 8045931 down 8047215 down 8047223 up 8049180 up 8049243 down 8049876 up 8050801 down 8051814 down 8051862 up 8052010 up 8053584 down 8053725 up 8055204 down 8058203 up 8059028 down 8059361 down 8059799 up 8060314 up 8061013 up 8062237 up 8062337 down 8062962 down 8064502 up 8065011 up 8065427 up 8065569 down 8065752 down 8066407 up 8066444 down 8067382 up 8069620 up 8071107 up 8071662 down 8072400 down 8072575 down 8072636 up 8073015 down 8073680 up 8073799 up 8074307 down 8074884 up 8075276 up 8075375 down 8076072 down 8076792 down 8076819 up 8078310 up 8080416 up 8080562 up 8080847 down 8080918 down 8081138 up 8081233 down 8081645 up 8082767 up 8083569 up 8083937 up 8084215 up 8084488 up 8084657 up 8084880 down 8088458 down 8089015 down 8089727 down 8091243 down 8092187 up 8092312 down 8093576 up 8093821 down 8094679 up 8096004 down 8096249 up 8096839 down 8097945 down 8101002 down 8101718 up 8103041 down 8104022 down 8104124 up 8104180 up 8104613 up 8104615 up 8104634 up 8104723 up 8104901 down 8105328 down 8107115 down 8107125 down 8108420 up 8108566 down 8108757 up 8109226 down 8109344 down 8110491 up 8110706 up 8111234 down 8111358 up 8112584 up 8113073 down 8114581 down 8115234 down 8115327 down 8116253 up 8116400 up 8116696 down 8117377 up 8120360 up 8120783 down 8120979 up 8121249 down 8121483 up 8122933 down 8123598 down 8127087 down 8127201 up 8127380 up 8128123 down 8128445 up 8128712 down 8129067 down 8130372 down 8130824 down 8131205 down 8131860 up 8133736 up 8134376 up 8135015 up 8135268 down 8135835 down 8135945 up 8136846 up 8136932 down 8137433 up 8138258 up 8138797 up 8139107 down 8139826 down 8139828 down 8139887 up 8141169 up 8141421 down 8142534 down 8142821 down 8142880 down 8143772 down 8144880 down 8146120 up 8146475 down 8146527 down 8146533 up 8147057 up 8147445 up 8147877 up 8148331 up 8148515 up 8148710 up 8149500 down 8149725 up 8151281 down 8151747 down 8151989 up 8153269 up 8153876 down 8154934 down 8155393 up 8156450 down 8156759 down 8156969 up 8157193 down 8157945 down 8159142 down 8159371 up 8159624 up 8159963 down 8160782 down 8161154 down 8161451 up 8161943 up 8162562 down 8162927 down 8162936 up 8163013 up 8163084 down 8163672 down 8165995 up 8166079 down 8166098 up 8166703 down 8166723 down 8167640 down 8168079 up 8168161 down 8168291 up 8168420 up 8168567 down 8168940 up 8169009 down 8169061 up 8169080 up 8169473 up 8170322 up 8170786 down 8171229 up 8171624 down 8171823 up 8173503 up 8177931 up 8180392 up

[0312] The table shows expression trends (i.e., up-regulation or down-regulation) of GRPS predictor genes in MUTZ-3 cells exposed to respiratory sensitizer. The Affymetrix Probe Set ID for the Human ST 1.0 Array are provided.

TABLE-US-00003 TABLE 1 Concentrations and vehicles used for each reference compound during assay development GARD input Max solubility Rv90 concentratrion Compound Abbreviation Vehicle (μM) (μM) (μM) Respiratory sensitizers Ammonium AH Water 35 — 35 hexachloroplatinate Ammonium persulfate AP DMSO — — 500 Ethylenediamine EDA Water — — 500 Glutaraldehyde GA Water — 10 10 Hexamethylen diisocyanate HDI DMSO 100 — 100 Maleic Anhydride MA DMSO — — 500 Methylene diphenol MDI DMSO 50 — 50 diisocyanate Phtalic Anhydride PA DMSO 200 — 200 Toluendiisocyanate TDI DMSO 40 — 40 Trimellitic anhydride TMA DMSO 150 — 150 Non-Respiratory sensitizers 1-Butanol BUT DMSO — — 500 2-Aminophenol 2AP DMSO — 100 100 2-Hydroxyethyl acrylate 2HA Water — 100 100 2-nitro-1,4-Phenylenediamine NPDA DMSO — 300 300 4-Aminobenzoic acid PABA DMSO — — 500 Chlorobenzene CB DMSO 98 — 98 Dimethyl formamide DF Water — — 500 Ethyl vanillin EV DMSO — — 500 Formaldehyde FA Water — 80 80 Geraniol GER DMSO — — 500 Hexylcinnamic aldehyde HCA DMSO 32.34 — 32.34 Isopropanol IP Water — — 500 Kathon CG* KCG Water — 0.0035% 0.0035% Methyl salicylate MS DMSO — — 500 Penicillin G PEN G Water — — 500 Propylene glycol PG Water — — 500 Potassium Dichromate PD Water 51.02 1.5 1.5 Potassium permanganate PP Water 38 — 38 Tween 80 T80 DMSO — — 500 Zinc sulphate ZS Water 126 — 126

[0313] *The chemical Kathon CG is a mixture of the two compounds MC and MCI. The concentration of the mixture is given in %.

TABLE-US-00004 TABLE 2 Chemicals included in the independent dataset used for validation of GRPS GARD input Max solubility Rv90 concentratrion Compound Abbreviation Vehicle (μM) (μM) (μM) Respiratory sensitizers Chloramine T CH-T Water — — 500 Ethylenediamine EDA Water — — 500 Isophorone diisocyanate IPDI DMSO  25 — 25 Phtalic Anhydride PA DMSO 200 — 200 Piperazine PPZ Water — — 500 Reactive Orange RO Water — 100 100 Non-respiratory sensitizers 1-Butanol BUT DMSO — — 500 2,4-dinitrochlorobenzene DNCB DMSO —  4 4 2-mercaptobenzothiazole MBT DMSO 250 — 250 Benzaldehyde BA DMSO 250 — 250 Chlorobenzene CB DMSO  98 — 98 Cinnamyl alcohol CALC DMSO 500 — 500 Diethyl phthalate DP DMSO  50 — 50 Eugenol EU DMSO 649 300 300 Glycerol GLY Water — — 500 Glyoxal GO Water — 300 300 Isoeugenol IEU DMSO 641 300 300 Lactic acid LA Water — — 500 Octanoic acid OA DMSO 504 — 500 Phenol PHE Water — — 500 p-hydroxybenzoic acid HBA DMSO 250 — 250 p-phenylenediamine PPD DMSO 566  75 75 Resorcinol RC Water — — 500 Salicylic acid SA DMSO — — 500 Sodium dodecyl sulphate SDS Water — 200 200

TABLE-US-00005 TABLE 3 Results from SVM classifications of the independent test dataset Classification.sup.1 SVM decision value Pos if 1 Treatment 1 2 3 4 5 sample >0 Respiratory sensitizers Chloramine T 0.52 0.59 Sensitizer Ethylenediamine −0.32 −0.20 Non-sensitizer Isophorone 0.10 0.17 Sensitizer diisocyanate Phtalic Anhydride 0.20 −0.12 Sensitizer Piperazine −0.05 −0.12 Non-sensitizer Reactive Orange 0.41 0.41 sensitizer Non-respiratory sensitizers 1-Butanol −0.32 0.12 Sensitizer 2,4-dinitrochloro- −1.66 −1.18 −1.90 Non-sensitizer benzene 2-mercaptobenzo- −0.44 −0.43 −0.57 Non-sensitizer thiazole Benzaldehyde −0.79 −0.87 −0.70 Non-sensitizer Chlorobenzene −1.03 −0.76 −1.15 0.24 0.06 Sensitizer Cinnamyl alcohol −0.57 −1.44 −1.26 Non-sensitizer Diethyl phthalate −1.37 −0.96 −1.22 Non-sensitizer Eugenol −1.67 −1.53 −1.51 Non-sensitizer Glycerol −1.05 −1.11 −0.77 Non-sensitizer Glyoxal −1.02 −0.69 −0.56 Non-sensitizer Isoeugenol −1.44 −1.27 −1.32 Non-sensitizer Lactic acid −1.20 −0.81 −0.89 Non-sensitizer Octanoic acid −0.65 −0.79 −1.22 Non-sensitizer Phenol −1.04 −0.38 −0.95 Non-sensitizer p-hydroxybenzoic −0.81 −0.56 −1.09 Non-sensitizer acid p-phenylenediamine −1.38 −1.19 −1.80 Non-sensitizer Resorcinol −1.01 −0.99 −1.40 Non-sensitizer Salicylic acid −0.73 −1.08 −1.13 Non-sensitizer Sodium dodecyl −1.49 −0.80 −1.30 Non-sensitizer sulphate .sup.1Classification on sensitizing properties for each chemical compound was based on the rule stating that any given sample in the test dataset should be classified as a respiratory sensitizer if any of replicate stimulations have a SVM decision value > 0.

TABLE-US-00006 TABLE 4 Canonical pathways associated with the top 999 predictors able to separate respiratory chemical sensitizers from non-respiratory sensitizers −log(p- Canonical Pathway value) Regulated molecules.sup.1 Oxidative phosphorylation 17.62 ATP5E, ATP5I, ATPK, COX Vb, COX VIIa-2, NDUFA1, NDUFA13, NDUFA2, NDUFA3, NDUFA6, NDUFA7, NDUFA9, NDUFAB1, NDUFB10, NDUFB4, NDUFB6, NDUFB8, NDUFB9, NDUFC1, NDUFS4, NDUFS5, NDUFS6, NDUFS8, NDUFV2, UQCR10, UQCRQPC Ubiquinone metabolism 13.29 NDUFA1, NDUFA13, NDUFA2, NDUFA3, NDUFA6, NDUFA7, NDUFA9, NDUFAB1, NDUFB10, NDUFB4, NDUFB6, NDUFB8, NDUFB9, NDUFC1, NDUFS4, NDUFS5, NDUFS6, NDUFS8, NDUFV2 Granzyme B signaling 4.72 Bid, Caspase-2, Lamin A/C, LAMP2, Smac/Diablo, tBid, Tubulin alpha FAS signaling cascades 3.79 Bid, c-FLIP (S), Caspase-2, DAXX, Lamin A/C, Smac/Diablo, tBid Cytoplasmic/mitochondrial transport of proapoptotic 3.57 Bid, DAXX, DLC1 (Dynein LC8a), DLC2 (Dynein LC8b), proteins Bid, Bmf and Bim Smac/Diablo, tBid Inhibitory PD-1 signaling in T cells 3.27 BCL2L1, CD8, CD8 alpha, CD80, CD86, MHC class II, PTEN HSP60 and HSP70/TLR signaling pathway 3.22 CD80, CD86, MD-2, MEK1/2, MHC class II, MyD88, Ubiquitin Astrocyte differentiation from adult stem cells 3.17 HES1, ID1, ID2, ID3, MEK1/2, SOX1 Apoptotic TNF-family pathways 3.06 Apo-2L(TNFSF10), BCL2L1, Bid, Caspase-2, Smac/Diablo, tBid TNFR1 signaling pathway 3.00 Bid, c-FLIP (S), Caspase-2, jBid, Smac/Diablo, tBid Role of Nek in cell cycle regulation 2.80 Histone H1, Ran, Tubulin (in microtubules), Tubulin alpha, Tubulin beta ATP/ITP metabolism 2.70 5′-NTC, ADSL, APRT, POLR2G, POLR2J, PPAP, RPB10, RPB6, RPB8, RRP41 Generation of memory CD4+ T cells 2.51 BCL2L1, CD80, CD86, IL7RA, MHC class II Dynein-dynactin motor complex in axonal transport 2.48 DYNLL, DYNLT, Tctex-1, TMEM108, Tubulin (in microtubules), in neurons Ubiquitin Antigen presentation by MHC class II 2.44 HLA-DM, HLA-DRA1, MHC class II IL-33 signaling pathway 2.36 Histone H2A, Histone H2B, MEK1/2, MyD88, ST2L, Ubiquitin Insulin regulation of translation 2.27 eEF2, eIF4A, eIF4B, eIF4G1/3, MEK1(MAP2K1) TNF-alpha-induced Caspase-8 signaling 2.23 Bid, c-FLIP (S), Caspase-2, PP2A regulatory, tBid Antigen presentation by MHC class I 2.18 CD8, CD8 alpha, PSMB5, PSME3 Main pathways of Schwann cells transformation in 2.18 BCL2L1, Calmodulin, MEK1(MAP2K1), MEK1/2, Neuregulin 1, PTEN neurofibromatosis type 1 Granzyme A signaling 2.08 Histone H1, Histone H2B, Lamin A/C, LAMP2 G-CSF-induced myeloid differentiation 2.08 G-CSF receptor, MEK1/2, Myeloblastin, PERM Substance P mediated membrane blebbing 2.07 MRLC, Tubulin (in microtubules), Tubulin alpha Role of IAP-proteins in apoptosis 2.03 Bid, Smac/Diablo, tBid, Ubiquitin .sup.1Molecules indicated in bold are present in GRPS.