WO2007106507A2 - Detection of gene expression in mixed sample or tissue - Google Patents

Abstract

An algorithmic approach can be used for calculating the expression profile of a tissue/sample of interest that consists of at least two types of cells. Specifically, the approach electronically subtracts the expression profile of one component of a sample from the expression profile of the total sample, thus revealing the profiles of the other component.

Description

DETECTION OF GENE EXPRESSION IN MIXED SAMPLE OR TISSUE

CROSS-REFERENCE TO A RELATED APPLICATION

[0001] This application claims the benefit of U.S. Provisional Application No. 60/782,124, filed March 14, 2006, and is incorporated by reference in its entirety.

GOVERNMENT INTERESTS

[0002] This invention was made with U.S. Government support under Grant Nos. 1R01AI33940 and 1R21AI67453, awarded by the National Institutes of Health. The U.S. Federal Government may have certain rights in the invention.

BACKGROUND

[0003] RNA profiling is a well established technique for identifying global expression patterns within cells, and is used for purposes ranging from the identification of disease biomarkers to basic understanding of cellular processes. Tavazoie S, et al. Nat Genet 22(3):281-285 (1999); Yewdell JW and Bennink JR, Annu Rev CellDev Biol, 15:579-606 (1999). Unfortunately, many biological samples contain mixtures of cell-types. For example, viruses infect only a proportion of the cells in a tissue, organs contain numerous tissue types (Shiki T, Cell Tissue Res, 244(2):285-298 (1986); Woods GL, Walker DH: CHn Microbiol Rev, 9(3):382-404 (1996)) and cancer cells make up only part of a biopsy sample. Cleatoref al. recently demonstrated that the non-cancerous portion of breast cancer samples can significantly affect expression profiles, and that factoring in the amount of cancerous material in samples can improve the accuracy of response prediction. Cleator S.J., et al. Breast Cancer Res, 8(3):R32 (2006). In all these examples, the cell type of interest (infected cells, cancer cells, or specific components of an organ) will be only a subset of cells in the sample. This severely limits the conclusions one can make about the specificity of gene expression in the cell type of interest.

[0004] Techniques such as laser capture microdissection (LCM) allow isolation of regions of a biological sample that are separated by as little as a few cell widths. Such samples can then be analyzed by expression analysis. However, LCM requires the tissue of interest to be morphologically distinguishable and physically separated from the other tissues in the sample. Finally, it can be very time consuming and requires specialized equipment to obtain a sufficient quantity of tissue to perform expression profiling. RNA amplification procedures can be used but these introduce artifacts of amplification. Mills LC, et al. Nat Cell Biol, 3(8):E175-178 (2001). Cell-sorting can also be used to isolate cells of interest; however this technique requires a suitable biomarker for the tissue to have been previously identified. In addition, both of these techniques are limited in that LCM is only practical with solid tissues, while flow-sorting is only applicable for tissues which can be put into suspension. Finally, the act of separation itself can result in the alteration of expression patterns.

[0005] There is an urgent need in the art for identification of gene expression in cells in order to understand cellular processes, cell maturation, cell differentiation, abnormal cell growth, the role of cellular interactions and growth in disease states. This will allow for the understanding of physiology, disease processes, therapeutic targeting and in identifying new therapeutic agents.

SUMMARY OF THE INVENTION

[0006] Gene expression by specific cell types is identified among mixed populations of cells in tissues. The method comprises a) laser microdissection or other methods of microdissection of the tissue regions of interest; b) microarray analysis of RNA isolated from the microdissected regions; c) microarray analysis of purified individual cellular components from the tissue; d) subtraction of the results from V from the results from 'b.'

[0007] The method can be used to assess gene expression in tissues and cells that are difficult to isolate. Examples, include, but not limited to: isolating thymic stromal cells from functionally-defined sub-regions of the tissue. This method has allowed for the characterization of regionally discrete patterns of gene expression in thymic stroma, as well as characterizing the stroma in general. Other examples, include, identifying changes (in stromal cells) that lead to degeneration (atrophy) of the thymus during aging. This method will subsequently be used to assess the specific genes expressed by stromal cells in other components of the immune system, and may ultimately be used for study of non-immune tissues as well. Still further examples, include pancreatic islets (diabetes) or metastatic tumors (cancer). [0008] One major advantage is that this method allows characterization of gene expressing in cells that cannot be isolated by conventional methods (enzymatic digestion, density centrifiigation, tissue culture expansion, cell sorting). Another advantage of this method is that it allows characterization of regional differences in gene expression at the microscopic level, as well as in specific cell types.

[0009] In a preferred embodiment, a method of identifying gene expression by specific cell types in mixed cell populations comprises microdissecting tissues; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in mixed cell populations. Preferably, the microdissected tissues are morphologically and functionally distinct tissue regions.

[0010] In another preferred embodiment, the microdissected tissues are from normal and diseased tissues and organs. The tissues comprise epithelial tissue, connective tissues, muscle tissues or nervous tissues and tissues are also microdissected from organs, for example, skin, digestive, muscular, nervous, respiratory circulatory, excretory, endocrine, lymphatic, and reproductive organs.

[0011] In a preferred embodiment, the tissues are microdissected by laser capture. Preferably, the microdissected tissues have a thickness of between 10 μm to 80 μm, more preferably, the microdissected tissues have a thickness of about 40 μm. Preferably, the microdissected tissues have a length of between about 100 μm to 700μm and a depth of between about 5μm to 50 μm.

[0012] In another preferred embodiment, nucleic acids isolated from the microdissected tissues are subjected to low cycle (limited) high-fidelity amplification. Preferably, the cycles are about two to 8 cycles of high fidelity amplification, more preferably, the cycles are about four to six cycles of high fidelity amplification.

[0013] In another preferred embodiment, the isolated nucleic acids are analyzed by microarray analysis and nucleic acid profiles are generated, The nucleic acid profiles of microdissected tissue regions are compared to nucleic acid profiles of specific cell types and/or nucleic acid profiles of known genes. [0014] In a preferred embodiment, the tissues comprise mixed cell populations, for examples, tissues from the thymus comprise mixed cell populations of thymocytes, tissues from tumors comprise mixed populations of tumor cells, normal cells, stem cells and the like.

[0015] In another preferred embodiment, a method of identifying gene expression in tumors, said method comprising: microdissecting tissues comprising tumors; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in tumors. Preferably, tumor cell pre-metastatic, metastatic and post-metastatic gene expression is identified.

[0016] In another preferred embodiment, a method of identifying gene expression in autoimmune diseases, said method comprising: microdissecting tissues comprising autoimmune cells; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in autoimmune diseases.

[0017] In another preferred embodiment, a method of identifying gene expression in diabetes, said method comprising: microdissecting pancreatic tissues; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in diabetes.

[0018] In another preferred embodiment, a method of identifying gene expression in neural diseases, said method comprising: microdissecting neural tissues; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in neural diseases.

[0019] In another preferred embodiment, a method of identifying pathogen induced gene expression in diseases, said method comprising: microdissecting tissues comprising a pathogen; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in diseases. Preferably, the pathogen comprises virus, bacteria, fungal or parasitic organisms.

[0020] In another preferred embodiment, a method of identifying candidate therapeutic agents for treatment of disease, said method comprising: administering a candidate agent; microdissecting control tissues and tissues treated with the candidate agent; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissues; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in control and treated tissues.

[0021] In one aspect, the present invention comprises a method for identifying gene expression by specific cell types in a mixed cell population within a tissue, said method comprising:

(a) obtaining a tissue region;

(b) quantifying the amount and relative level of gene a expression product (such as mRNA or polypeptides) isolated from said microdissected tissue region to obtain the level of gene expression for the tissue region;

(c) quantifying the amount and relative level of gene expression products in from an isolated cell or from a group of cells to obtain the level of gene expression for an isolated cell or isolated group of cells (by, for example, microarray or MALDl-TOF);

(d) comparing the level of gene expression from said tissue region against the level of gene expression from said isolated cell or isolated group of cells, thereby

(e) obtaining the level of gene expression for said isolated cell or isolated group of cells and cells not isolated.

10022] The presently inventive method is useful for examining expression in tissue from any sources such as skin, digestive, muscular, nervous, respiratory circulatory, excretory, endocrine, lymphatic, and reproductive organs. In a related embodiment, the method is performed on epithelial tissue, connective tissues, muscle tissues or nervous tissues, for example. In further related embodiments, the method is used for examining expression in a mixed cell population comprising stromal cells of the thymic cortex or thymic medulla; tumor cells; or stem cells.

[0023] The method of the invention may be advantageously used on microdissected tissues from morphologically and functionally distinct tissue regions, such from normal and diseased tissues and organs. In one embodiment, tissue is microdissected by laser capture. The may be practiced on different thicknesses, cross sections and like of tissue. In some embodiments, microdissected tissues have a thickness of between about 10 μm to 80 μm; in others the microdissected tissues have a thickness of about 40 μm; in others, the microdissected tissues have a length of between about 100 μm to 700μm and a depth of between about 5μm to 50 μm; or from about 100 μm to about 500 μm thick.

[0024] The sensitivity of the present method may be increased by low cycle (limited) high-fidelity amplification on microdissected tissue.

[0025] In the method of the invention, the data generated may be compared not only to cells from the same tissue region, but from other regions, or even against expression profiles of specific cell types, or expression profiles of known genes or regulatory pathways.

[0026] The present invention also includes a method of identifying gene expression in cancer-containing tissue, said method comprising:(a) microdissecting tissues comprising tumors; (b) analyzing gene expression products from the microdissected tissues; (c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue, and (e) identifying gene expression by specific cell types in cancer-containing tissue. Such a method can be used to identify pre-metastatic, metastatic and post-metastatic gene expression.

[0027] In another embodiment, the invention includes a method of identifying gene expression in autoimmune diseases, said method comprising:(a) microdissecting tissues comprising autoimmune cells; (b) analyzing gene expression products from the microdissected tissues; (c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue, (e) dentifying gene expression by specific cell types in autoimmune diseases.

[0028] In another embodiment, the invention includes a method of identifying gene expression in diabetes, said method comprising: (a) microdissecting pancreatic tissue; (b) analyzing gene expression products from the microdissected tissue; (c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue, (e) identifying gene expression by specific cell types in diabetes. In a related embodiment, the effect of diabetes on gene expression can be performed on other tissues in the body.

[0029] In another embodiment, the invention includes a method of identifying gene expression in neural diseases, said method comprising: (a) microdissecting neural tissues; (b) analyzing gene expression products from the microdissected tissue; (c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue; (e) identifying gene expression by specific cell types in neural diseases.

[0030] In another embodiment, the invention includes a method of identifying pathogen induced gene expression in diseases, said method comprising: (a) microdissecting tissue infected with a pathogen; (b) analyzing gene expression products from the microdissected tissue; (c)analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d)comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue; (e) identifying gene expression by specific cell types in diseases. Pathogens can include a virus, bacteria, fungal or parasitic organism.

[0031] In another embodiment, the invention includes a method of identifying candidate therapeutic agents for treatment of disease, said method comprising: administering a candidate agent; microdissecting control tissues and tissues treated with the candidate agent; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissues; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and identifying gene expression by specific cell types in control and treated tissues.

[0032] Related embodiments include the use of the methods of the invention for determining a change in cell expression in response to a chemical compound, infectious agent, or cellular signal; for determining whether a compound of interest affects a given cell type in situ.

[0033] The present invention has identified genes expressed in thymic stromal cells, as presented in Tables 4 and 5, and therefore also includes a number of novel gene expression identified in the thymus and the use of such knowledge for therapeutic and drug design purposes. Thus, the invention includes a method of detecting thymic cortical stromal cells comprising use of a probe against a gene identified in Table 4, or an antibody against a polypeptide gene product identified in Table 4; and a method of detecting thymic medullary stromal cells comprising use of a probe against a gene identified in Table 5, or an antibody against a polypeptide gene product identified in Table 5. In related embodiments, the invention includes a method of treating cancer of the thymus, comprising use of an antibody against a gene product identified in Table 4 or 5.

[0034] Further embodiments of the invention include a method for diagnosing a human disease or disorder, comprising: (a) detecting gene expression in a sample, wherein said sample is subjected to microdissection and microarray analysis and (b) correlating the nucleic acid molecule expression profile with a disease or disorder. This and the other method of the invention are useful for examing gene expression in diseases including cancer, Parkinson's disease, Alzheimer's disease, Huntington's chorea, amyotrophic lateral sclerosis (ALS), nutritional diseases, diabetes, Bell's palsy, systemic lupus erythematosus, multiple sclerosis, human immunodeficiency virus-associated myelopathy, transverse myelopathy or various etiologies, progressive multifocal leukoencephalopathy, and central pontine myelinolysis.

[0035] The invention also includes a method for determining the expression profile of a sample of interest having at least two types of cells, comprising subtracting the expression profile of one component of said sample from the expression profile of the total sample. In another embodiment, the invention includes a method of diagnosing a disease or disorder, comprising: (a) detecting gene expression in a sample, wherein said sample is subjected to microdissection; (b) subjecting the mRNA to microarray analysis; and (c) correlating the nucleic acid molecule expression profile with a disease or disorder.

[0036] In a further embodiment, the invention includes a method for calculating the expression profile of a sample of interest having at least two types of cells, comprising subtracting the expression profile of one component of said sample from the expression profile of the total sample.

BRIEF DESCRIPTION OF THE DRAWINGS

[0037] FIGURE 1: Graphically shows changes in gene expression patterns during thymocyte developmental progression.

[0038] FIGURE 2A-2D: Graphically shows the differential expression of Notch and related genes during progenitor thymocyte differentiation.

[0039] FIGURE 3A-3B: show specific expression of α6β4 integrin on DN2 and DN3 thymocyte progenitors.

[0040] FIGURE 4A-4C: Graphically shows deviations in cell size and DN proportion in mice lacking distal signaling motifs in the β4 integrin tail.

[0041] FIGURE 5A-5B: show the differential expression of lymphoblastic leukemia- 1 (LyIl) during thymocyte progenitor differentiation.

[0042] FIGURE 6: Schematic representation showing two models for regulation of lymphocyte differentiation by distinct microenvironments in the post-natal thymus.

[0043] FIGURE 7: scan of a gel showing the differential expression of Notch ligands in different thymic microenvironments.

[0044] FIGURE 8A-8C: scans of photograph showing microdissection of functionally defined tissue regions from the thymus.

[0045] FIGURE 9: Schematic representation showing the methods for identifying stromally-expressed genes from laser captured regions. [0046] FIGURE 10: Sorted Ratios (R/_mWtaM>wn) for Genes from Simulated Datasets at a Range of pA Values and from a Mix of Seed Experiments.

[0047] FIGURE 11: Comparison of Actual Fractional Percent pA Versus the Ratio Values from Different Sorted List Positions.

[0048] FIGURE 12: Comparison of Actual Fractional Percent (pA) Versus Calculated Value.

[0049] FIGURE 13: Sensitivity and Specificity to Detect Genes Over- expressing in Unknown Samples.

[0050] FIGURE 14: Hierarchical clustering of microarray results from dissected tissues and sorted lymphocytes.

[0051] FIGURE 15: The antigen processing pathway, of which 30 genes are activated in cortical stromal cells.

[0052] FIGURE 16: Neurodegenerative disease pathways, of which 12 genes are activated in cortical stromal cells.

[0053] FIGURE 17: Leukocyte trans-endothelial migration pathways, of which 50 genes were activated in medullary stromal cells.

[0054] FIGURE 18: TGFβ pathway, of which 32 genes were activated in medullary stromal cells.

[0055] FIGURE 19 : Hierarchical clustering of microarray results from AIRE mutant or wild-type medulla.

[0056] FIGURE 20 : Tissue specificity score: published vs DGEM AIRE target lists, showing relative range.

[0057] FIGURE 21 : Tissue specificity score: published vs DGEM AIRE target lists, showing intensity skew.

[0058] FIGURE 22 : Presence of consensus AIRE binding sites in promoter for published versus DGEM AIRE target lists. DETAILED DESCRIPTION

[0059] Biological samples frequently contain multiple cell types that each play a crucial role in the development and / or regulation of adjacent cells or tissues. The search for biomarkers of, or expression patterns for, one cell-type in those samples can be a complex and time-consuming process. Ordinarily, extensive laboratory bench work must be performed to separate the desired tissue into its sub-components, such that each can be accurately characterized.

[0060] The present inventors have developed a methodology to electronically subtract gene expression in one or more components of a tissue from a mixture, to reveal the expression patterns of other minor or difficult to isolate components. This methodology can reliably determine the expression patterns in cell types that contribute as little as 5% of the total expression in a tissue, and can be used in a wide range of therapeutic and diagnostic applications for numerous diseases and physiological conditions.

[0061] AU technical terms in this description are commonly used in biochemistry, molecular biology and immunology, respectively, and can be understood by those skilled in the field of this invention. Those technical terms can be found in: MOLECULAR CLONING: A LABORATORY MANUAL, 3rd ed., vol. 1-3, ed. Sambrook and Russel, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y., 2001; CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, ed. Ausubel et al., Greene Publishing Associates and Wiley- Interscience, New York, 1988 (with periodic updates); SHORT PROTOCOLS IN MOLECULAR BIOLOGY: A COMPENDIUM OF METHODS FROM CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, 5^th ed., vol. 1 -2, ed. Ausubel et al., John Wiley & Sons, Inc., 2002; GENOME ANALYSIS: A LABORATORY MANUAL, vol. 1-2, ed. Green et al., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y., 1997; CELLULAR AND MOLECULAR IMMUNOLOGY, 4^th ed. Abbas et al., WB Saunders, 1994.

DEFINITIONS

[0062] As used herein, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise.

[0063] An array is "addressable" in that it has multiple regions of different moieties (for example, different polynucleotide sequences) such that a region (a "feature" or "spot" of the array) at a particular predetermined location (an "address") on the array will detect a particular target or class of targets (although a feature may incidentally detect non- targets of that feature). Array features are typically, but need not be, separated by intervening spaces. In the case of an array, the "target" will be referenced as a moiety in a mobile phase (typically fluid), to be detected by probes ("target probes") which are bound to the substrate at the various regions. However, either of the "target" or "target probes" may be the one which is to be evaluated by the other (thus, either one could be an unknown mixture of polynucleotides to be evaluated by binding with the other). An "array layout" refers collectively to one or more characteristics of the features, such as feature positioning, one or more feature dimensions, and some indication of a moiety at a given location. "Hybridizing" and "binding", with respect to polynucleotides, are used interchangeably.

[0064] AIRE: Autoimmune regulator (AIRE) is a transcription factor that controls the self-reactivity of the T cell repertoire

[0065] "Allogeneic" refers to immune cells derived from non-self major histocompatibility complex donors. HLA haplotypes/allotypes vary from individual to individual and it is often helpful to determine the individual's HLA type. The HLA type may be determined via standard typing procedures.

[0066] The terms "bioarray", "biochip", "biochip array", and "microarray" refer to an ordered spatial arrangement of immobilized biomolecular probes arrayed on a solid supporting substrate. Preferably, the biomolecular probes are immobilized on second linker moieties in contact with polymeric beads, wherein the polymeric beads are immobilized on first linker moieties in contact with the solid supporting substrate. Biochips, as used in the art, encompass substrates containing arrays or microarrays, preferably ordered arrays and most preferably ordered, addressable arrays, of biological molecules that comprise one member of a biological binding pair. Typically, such arrays are oligonucleotide arrays comprising a nucleotide sequence that is complementary to at least one sequence that may be or is expected to be present in a biological sample. Alternatively, proteins, peptides or other small molecules can be arrayed in such biochips for performing, inter alia, immunological analyses (wherein the arrayed molecules are antigens) or assaying biological receptors (wherein the arrayed molecules are ligands, agonists or antagonists of said receptors). Useful microarrays are also commercially available, inter alia, from Affymetrix (Santa Clara, CA). An example of a commercially available biochip, but not meant to be limiting, is the Affymetrix GeneChip® MOE430A.

[0067] A "biopolymer" is a polymer of one or more types of repeating units. Biopolymers are typically found in biological systems (although they may be made synthetically) and particularly include peptides or polynucleotides, as well as such compounds composed of or containing amino acid analogs or non-amino acid groups, or nucleotide analogs or non-nucleotide groups. This includes polynucleotides in which the conventional backbone has been replaced with a non-naturally occurring or synthetic backbone, and nucleic acids (or synthetic or naturally occurring analogs) in which one or more of the conventional bases has been replaced with a group (natural or synthetic) capable of participating in Watson- Crick type hydrogen bonding interactions. Polynucleotides include single or multiple stranded configurations, where one or more of the strands may or may not be completely aligned with another.

[0068] "Cancer" refers to all types of cancer or neoplasm or malignant tumors found in mammals, including, but not limited to: leukemias, lymphomas, melanomas, carcinomas and sarcomas. Examples of cancers are cancer of the brain, breast, pancreas, cervix, colon, head and neck, kidney, lung, non-small cell lung, melanoma, mesothelioma, ovary, sarcoma, stomach, uterus and Medulloblastoma. As used herein, the terms "cancer," "neoplasm," and "tumor," are used interchangeably and in either the singular or plural form, refer to cells that have undergone a malignant transformation that makes them pathological to the host organism. Primary cancer cells (that is, cells obtained from near the site of malignant transformation) can be readily distinguished from non-cancerous cells by well-established techniques, particularly histological examination. The definition of a cancer cell, as used herein, includes not only a primary cancer cell, but any cell derived from a cancer cell ancestor. This includes metastasized cancer cells, and in vitro cultures and cell lines derived from cancer cells. When referring to a type of cancer that normally manifests as a solid tumor, a "clinically detectable" tumor is one that is detectable on the basis of tumor mass; e.g., by procedures such as CAT scan, MR imaging, X-ray, ultrasound or palpation, and/or which is detectable because of the expression of one or more cancer-specific antigens in a sample obtainable from a patient.

[0069] "CD4" is a cell surface protein important for recognition by the T cell receptor of antigenic peptides bound to MHC class II molecules on the surface of an APC. Upon activation, naive CD4 T cells differentiate into one of at least two cell types, ThI cells and Th2 cells, each type being characterized by the cytokines it produces. "ThI cells" are primarily involved in activating macrophages with respect to cellular immunity and the inflammatory response, whereas "Th2 cells" or "helper T cells" are primarily involved in stimulating B cells to produce antibodies (humoral immunity). CD4 is the receptor for the human immunodeficiency virus (HIV). Effector molecules for ThI cells include, but are not limited to, IFN-γ, GM-CSF, TNF-α, CD40 ligand, Fas ligand, IL-3, TNF-β, and IL-2. Effector molecules for Th.2 cells include, but are not limited to, IL-4, IL-5, CD40 ligand, IL-3, GS-CSF, IL-IO, TGF-β, and eotaxin. Activation of the ThI type cytokine response can suppress the Th2 type cytokine response, and reciprocally, activation of the Th2 type cytokine response can suppress the ThI type response.

[0070] A "chemokine" is a small cytokine involved in the migration and activation of cells, including phagocytes and lymphocytes, and plays a role in inflammatory responses.

[0071] A "cytokine" is a protein made by a cell that affect the behavior of other cells through a "cytokine receptor" on the surface of the cells the cytokine effects. Cytokines manufactured by lymphocytes are sometimes termed "lymphokines." Cytokines are also characterized as Type I (e.g. IL-2 and IFN-γ) and Type II (e.g. IL-4 and IL-10).

[0072] "Detecting the level of expression" includes methods that quantitate expression levels as well as methods that determine whether a gene of interest is expressed at all. Thus, an assay which provides a yes or no result without necessarily providing quantification of an amount of expression is an assay that requires "detecting the level of expression" as that phrase is used herein.

[0073] "Cells of the immune system" or "immune cells" as used herein, is meant to include any cells of the immune system that may be assayed, including, but not limited to, B lymphocytes, also called B cells, T lymphocytes, also called T cells, natural killer (NK) cells, natural killer T (NK) cells, lymphokine-activated killer (LAK) cells, monocytes, macrophages, neutrophils, granulocytes, mast cells, platelets, Langerhans cells, stem cells, dendritic cells, peripheral blood mononuclear cells, tumor-infiltrating (TIL) cells, gene modified immune cells including hybridomas, drug modified immune cells, and derivatives, precursors or progenitors of the above cell types. [0074J "Immune effector cells" refers to cells capable of binding an antigen and which mediate an immune response selective for the antigen. These cells include, but are not limited to, T cells (T lymphocytes), B cells (B lymphocytes), monocytes, macrophages, natural killer (NK) cells and cytotoxic T lymphocytes (CTLs)₁ for example CTL lines, CTL clones, and CTLs from tumor, inflammatory, or other infiltrates.

[0075] "Immune related molecules" refers to any molecule identified in any immune cell, whether in a resting ("non-stimulated") or activated state, and includes any receptor, ligand, cell surface molecules, nucleic acid molecules, polypeptides, variants and fragments thereof.

[0076] "T cells" or "T lymphocytes" are a subset of lymphocytes originating in the thymus and having heterodimeric receptors associated with proteins of the CD3 complex (e.g., a rearranged T cell receptor, the heterodimeric protein on the T cell surfaces responsible for antigen/MHC specificity of the cells). T cell responses may be detected by assays for their effects on other cells (e.g., target cell killing, activation of other immune cells, such as B- cells) or for the cytokines they produce.

[0077] As are recognized by those in the art, the term "host compatible" or "autologous" cells means cells that are of the same or similar haplotype as that of the subject or "host" to which the cells are administered, such that no significant immune response against these cells occurs when they are transplanted into a host.

[0078] "Substrate" refers to any rigid or semi-rigid support to which nucleic acid molecules or proteins are bound and includes membranes, filters, chips, slides, wafers, fibers, magnetic or nonmagnetic beads, gels, capillaries or other tubing, plates, polymers, and microparticles with a variety of surface forms including wells, trenches, pins, channels and pores.

[0079] "Immunoassay" is an assay that uses an antibody to specifically bind an antigen (e.g., a marker). The immunoassay is characterized by the use of specific binding properties of a particular antibody to isolate, target, and/or quantify the antigen.

[0080] "Neural (neuronal) defects, disorders or diseases" as used herein refers to any neurological disorder, including but not limited to neurodegenerative disorders (Parkinson's; Alzheimer's) or autoimmune disorders (multiple sclerosis) of the central nervous system; memory loss; long term and short term memory disorders; learning disorders; autism, depression, benign forgetfulness, childhood learning disorders, close head injury, and attention deficit disorder; autoimmune disorders of the brain, neuronal reaction to viral infection; brain damage; depression; psychiatric disorders such as bi-polarism, schizophrenia and the like; narcolepsy/sleep disorders(including circadian rhythm disorders, insomnia and narcolepsy); severance of nerves or nerve damage; severance of the cerebrospinal nerve cord (CNS) and any damage to brain or nerve cells; neurological deficits associated with AIDS; tics (e.g. Giles de Ia Tourette's syndrome); Huntington's chorea, schizophrenia, traumatic brain injury, tinnitus, neuralgia, especially trigeminal neuralgia, neuropathic pain, inappropriate neuronal activity resulting in neurodysthesias in diseases such as diabetes, MS and motor neurone disease, ataxias, muscular rigidity (spasticity) and temporomandibular joint dysfunction; Reward Deficiency Syndrome (RDS) behaviors in a subject.

[0081] "Autoimmunity" is the failure of an organism to recognise its own constituent parts (down to the sub-molecular levels) as "Self, as a result of which it attempts to mount an immune response against its own cells and tissues. Any disease that results from such an aberrant immune response is termed an "autoimmune disease" the prominent examples being Systemic Lupus Erythematosus (SLE), Sjogren's syndrome and Rheumatoid Arthritis. Other examples of autoimmune diseases include but not limited to graft immune diseases (chronic GVHD), ulcerative colitis, myasthenia gravis, systemic progressive scleroderma, interstitial cystitis, Hashimoto's diseases, Basedow's diseases, autoimmune hemolytic anemia, idiopathic thrombocytopenic purpura, Goodpasture's syndrome, atrophic gastritis, pernicious anemia, Addison diseases, pemphigus, pemphigoid, lenticular uveitis, sympathetic ophthalmia, primary biliary cirrhosis, active chronic hepatitis, multiple myositis, dermatomyositis, polyarteritis nodosa, rheumatic fever, glomerular nephritis (lupus nephritis, IgA nephropathy, and the like), allergic encephalitis, atopic allergic diseases (for example, bronchial asthma, allergic rhinitis, allergic dermatitis, allergic conjunctivitis, pollinosis, urticaria, food allergy and the like), Omenn's syndrome, vernal conjunctivitis and hypereosinophilic syndrome.

[0082] As used herein, "pathogens" refer to any organism that causes diseases, Examples include, viral, bacterial, parasitic, fungal and the like. Non-limiting examples include, human disease-causing organisms (and the diseases caused by them)such as Neisseria gonorrhoeae (gonorrhoea); Chlamydia trachomatis (chlamydia, lymphogranuloma venereum); Treponema pallidum (syphilis); Haemophilus ducrei (chancroid); Donovania granulomatis (donovanosis); Mycoplasma pneumoniae, M. homm^' is, M. genitalium, Ureaplasma urealyticum (mycoplasmas); Shigella flexneri (shigella); Salmonella typhi, S. choleraesuis, S. enteritidis (salmonella); Campylobacter fetus, C. jejuni (Campylobacter); human immunodeficiency virus HIV-I and HIV-2 (HlV, AIDS); HTLV-I (T-lymphotropic virus type 1); herpes simplex virus type 1 and type 2 (HSV-I and HSV-2); Epstein-Barr virus; cytomegalovirus; human herpesvirus 6; varicella-zoster virus; human papillomaviruses (many types) (genital warts); Molluscum contagiosum (MSV); hepatitis A virus, hepatitis B virus (viral hepatitis); Trichomoniasis vaginalis (trichomoniasis); yeasts such as Candida albicans (vulvovaginal candidiasis).

[0083] A "peptide" is used to refer to an amino acid multimer of any length (for example, more than 10, 10 to 100, or more amino acid units). A biomonomer fluid or biopolymer fluid reference a liquid containing either a biomonomer or biopolymer, respectively (typically in solution). As used herein, the terms "polypeptide" or "peptide" encompasses amino acid chains of any length, including full length proteins recited herein.

[0084] As used herein, "peptides or epitopes with longer amino sequences" encompasses amino acid chains of any length, including full length proteins recited herein.

[0085] As used herein, "variant" or "derivative" of polypeptides refers to an amino acid sequence that is altered by one or more amino acid residues. The variant may have "conservative" changes, wherein a substituted amino acid has similar structural or chemical properties (e.g., replacement of leucine with isoleucine). More rarely, a variant may have "nonconservative" changes (e.g., replacement of glycine with tryptophan). Analogous minor variations may also include amino acid deletions or insertions, or both. Guidance in determining which amino acid residues may be substituted, inserted, or deleted without abolishing biological activity may be found using computer programs well known in the art.

[0086] The resulting polypeptides generally will have significant amino acid identity relative to each other. A polymorphic variant is a variation in the polynucleotide sequence of a particular gene between individuals of a given species. Polymorphic variants also may encompass "single nucleotide polymorphisms" (SNPs) or single base mutations in which the polynucleotide sequence varies by one base.

[0087] "Stringency" is meant the combination of conditions to which nucleic acids are subject that cause the duplex to dissociate, such as temperature, ionic strength, and concentration of additives such as formamide. Conditions that are more likely to cause the duplex to dissociate are called "higher stringency", e.g. higher temperature, lower ionic strength and higher concentration of formamide.

[0088] For applications requiring high selectivity, one will typically desire to employ relatively stringent conditions to form the hybrids, e.g., one will select relatively low salt and/or high temperature conditions, such as provided by about 0.02 M to about 0.10 M NaCl at temperatures of about 50° C. to about 70° C.

[0089] For certain applications, it is appreciated that lower stringency conditions are required. Under these conditions, hybridization may occur even though the sequences of probe and target strand are not perfectly complementary, but are mismatched at one or more positions. Conditions may be rendered less stringent by increasing salt concentration and decreasing temperature. For example, a medium stringency condition could be provided by about 0.1 to 0.25 M NaCl at temperatures of about 37° C. to about 55° C, while a low stringency condition could be provided by about 0.15 M to about 0.9 M salt, at temperatures ranging from about 20° C. to about 55° C. Thus, hybridization conditions can be readily manipulated depending on the desired results.

[0090] The phrase "hybridizing conditions" and its grammatical equivalents, when used with a maintenance time period, indicates subjecting the hybridization reaction admixture, in context of the concentration of the reactants and accompanying reagents in the admixture, to time, temperature, pH conditions sufficient to allow the polynucleotide probe to anneal with the target sequence, typically to form the nucleic acid duplex. Such time, temperature and pH conditions required to accomplish the hybridization depend, as is well known in the art on the length of the polynucleotide probe to be hybridized, the degree of complementarity between the polynucleotide probe and the target, the guanidine and cytosine content of the polynucleotide, the stringency of the hybridization desired, and the presence of salts or additional reagents in the hybridization reaction admixture as may affect the kinetics of hybridization. Methods for optimizing hybridization conditions for a given hybridization reaction admixture are well known in the art.

General Methodology

A. Methods for Identifying Gene Expression [0091] Any method known in the art for identifying gene expression may be used. For example, gene expression may be determined by Northern, Southern, PCR, sequencing, mass spectrometry, array technology, or any other method known in the art.

[0092] As an illustrative example which is not meant to limit or construe the invention in any way, a method of identifying gene expression is provided. RNA is obtained from isolated tissue regions (prepared by microdissection) and from the corresponding lymphoid constituents (prepared by cell sorting). The stratified distribution of thymocyte developmental stages in discontinuous tissue regions indicates that different regions each deliver relatively distinct sets of signals to developing T cells. Tissues are microdissected from six defined regions of the thymus, as shown in Figure 8. These regions were selected because they each represent a unique signaling environment, as defined by the functions of individual lymphoid progenitor species within them. As shown in Figure 8, only tissue regions displaying relatively concentric cortical/medullary organization and broad tissue depth are used, in order to minimize cross-contamination between regions. Strips of tissue 40 μm in width (approximately 6-8 cell diameters in width) are dissected until approximately 1 mm² of tissue has been collected (about 50 strips 500 μm long); preliminary studies show that this will yield approximately 50 μg of RNA, thus requiring minimal amplification to prepare sufficient template for gene chip analysis. Samples dissected from these regions are stored in individual microiϊige tubes until the post-dissection tissue is mounted and examined, and only those samples that are appropriately located are utilized.

[0093] The dissected strips of tissue contain both lymphoid and non-lymphoid cells. In order to distinguish genes expressed by the stromal components, those genes expressed by the lymphoid constituents of these regions were identified and filtered out. To filter out the lymphoid genes, all major conventional T lymphoid stages were purified and screened. These include DNl, DN2, DN3, preDP, DP, CD4SP, and CD8SP for the TCRαβ lineage, as well as CD3⁺TCRγδ⁺ for this alternate thymic lineage. These populations are identified and purified by cell sorting.

High-fidelity amplification, labeling, and hybridization to gene chip tnicroarrays;

[0094] Microdissection of tissue regions 25μm wide, 200-500 μm long, and 10 μm deep can reliably yield 50ng of tissue. High-fidelity amplification can be used to generate the additional cRNA needed for microarray. Briefly, microdissected RNA are used for cDNA synthesis using oligo-dT(T7) primers (Affymetrix) and MessageAmp RNA kits (Ambion). This cDNA are used for reverse transcription (MessageAmp), and the process of reverse transcription//« vitro transcription are repeated 4-6 times until sufficient linearly amplified cRNA is obtained. During the last in vitro transcription cycle, cRNA are labeled by the addition of biotinylated nucleotides, and 1.5-2.0μg are hybridized to MOE430 2.0 arrays.

Proteomic profiling

[0095] An additional means of determining gene expression is by proteomic profiling, examining the proteins directly. This may be achieved by immobilized on a "Protein Chip" array and analysis by SELDI-TOF mass spectrometry. Nakagawa et al. "Proteomic profiling of primary breast cancer predicts axillary lymph node metastasis," Cancer Res, 2006, 66: 11825-11830.

Laser Capture Microdissection

[0096] Laser Capture Microdissection (LCM) enables separation of clusters of cells or even individual cells of interest from a background of millions of other cells. The collected cells can be directly visualized to verify their identity and purity. LCM is used to select small clusters of cells of interest from frozen sections of tissue by embedding them in a transfer film, e.g., a thermoplastic polymer. An example of a suitable thermoplastic polymer is ethylene vinyl acetate (EVA). The general methods of LCM are well known. See, e.g., U.S. Pat. Nos. 5,985,085; 5,859,699; and 5,843,657; as well as Suarez-Quian et al., "Laser Capture Microdissection of Single Cells from Complex Tissues," BioTechniques, Vol. 26, pages 328- 335 (1999); Simone et al., "Laser-capture microdissection: opening the microscopic frontier to molecular analysis," TIG, Vol. 14, pages 272-276 (1998); and Bonner et al., "Laser Capture Microdissection: Molecular Analysis of Tissue," Science, Vol. 278, pages 1481-1483 (1997).

[0097] LCM is a process by which cells and portions of biological tissue samples are acquired directly from tissue sections mounted on glass slides or other solid surfaces. The process involves placing a Capsure™ device, containing a thin-film polymer, onto the tissue section. Once the cells or tissue portions of interest (tissue targets) are located in the sample, a laser is focused over the tissue targets. When the laser is fired, the thin-film located directly above the tissue targets melts, flows down and adheres to the tissue targets. The Capsure™ device, holding the adhered tissue targets, is then removed from the tissue sample. The tissue targets are now stabilized on the Capsure™ device and ready for molecular analysis. [0098] Alternatively, another method of microdissecting tissue is the use of enzyme treatment. Various enzyme treatments used to microdissect tissue are well known in the art. In other embodiments, enzyme treatment may increase overall cell yield. Accordingly, enzyme treatment may be used alone or in combination with microdissection methods. A wide variety of cell-sustaining media that can be used to keep the pH of the liquid in a range that promotes survival of cells and to provide additional volume of liquid within which the enzymatic digestion can occur. Non-limiting examples include F12/DMEM, Ham's FlO (Sigma), CMRL-1066, Minimal essential medium (MEM, Sigma), RPMI-1640 (Sigma), Dulbecco's Modified Eagle's Medium (DMEM, Sigma), and Iscove's Modified Eagle's Medium (IMEM). In addition, any of the nutrient media described in Ham and Wallace (1979) Meth. Em., 58:44, Barnes and Sato (1980) Anal. Biochem., 102:255, or Mather, J. P. and Roberts, P. E. (1998) "Introduction to Cell and Tissue Culture", Plenum Press, New York can also be used.

[0099] If these isolated cells contain, or are suspected to contain, one or more DNA or RNA of interest, the extracted sample may be subjected to polymerase chain reaction (PCR) amplification, followed by, for example, microarray analysis, hybridization, strand conformational polymorphism, and southern and northern blotting, sequencing, etc. as desired. Other techniques for analysis of DlSiA and RNA are known to those skilled in the art and encompassed by the spirit and scope of the invention.

[0100] If the extracted cells contain, or are suspected to contain proteins or polypeptides of interest, the extracted sample can be subjected to enzyme zymography, for example using one or more labeled substrates, an immunoassay utilizing, for example, labeled antibodies or functional fragments thereof, a biochemical assay, and the like.

[0101] Selective extraction or microdissection of frozen tissue sections according to the present invention allows for recovery and analysis of both active enzymes and mRNA. Additionally, the DNA recovered from these sections is in the native condition and can be used for studies such as DNA fingerprinting. Microdissection of paraffin embedded tissues according to the present invention allows for PCR amplification of DNA, for example, from pure cell populations representing less than one high powered field, or a single layer of cells lining [0102] For general preparation of samples for frozen section microdissection according to the present invention, microdissection slides can be prepared by placing 1% agarose on a standard histology slide and cover slipping. After a short period of time, e.g., about 5 minutes, the cover slip is removed leaving a thin gel on the slide. A small frozen tissue section, e.g. about 25 micron thick, is placed on the agarose gel and briefly stained with eosin. The tissue may also be treated with agents to denature or otherwise inhibit RNase depending on the subsequent extraction method.

[0103] For enzyme analysis the procured tissue specimen can be placed in an appropriate buffer depending on the enzyme of interest, as known to the person skilled in the art. The enzyme levels can be measured by several methods including zymography and the use of specific substrates, including fluorometric, colorometric and radioactive substrates. The precise levels of enzyme expression in a specific, predefined cell population can be thus determined and, where desired, compared to that of another, independently isolated sample from the tissue sample.

[0104] For mRNA analysis the tissue specimen can be placed on agarose and treated with agents to denature or otherwise inhibit RNase, if desired. The procured tissue specimen is immediately frozen in liquid nitrogen. The tissue can be used immediately or stored at -70⁰C for several months. The mRNA can be extracted using, for example, column chromatography on oligo-dT (Micro-FastTrack mRNA Isolation Kit, Invitrogen Co.). The recovered mRNA of the pure cell populations can also be amplified and investigated using polymerase chain reaction (PCR) technology, such as, for example, by RT-PCR as known to those skilled in the art.

[0105] For DNA analysis the tissue specimen can be placed in a single step extraction buffer solution of 50 mM Tris, pH 8.5, 1 mM EDTA, 0.5% Tween 20, and 0.2 mg/ml proteinase K, incubated for four hours at about 37°C, followed by ten minutes incubation at about 95°C. The recovered DNA can also be amplified and analyzed using PCR technology in combination with analysis techniques, such as microarray analysis, blotting, sequencing, etc., known in the art. If native DNA is required for DNA fingerprinting analysis, the proteinase K can be added after DNase in the fingerprinting protocol.

[0106] For paraffin section microdissection routine, formalin fixed, paraffin embedded tissue sections are microdissected after de-paraffinization and brief staining with eosin. Tissue sections are visualized by direct microscopy and cell populations or subpopulations of interest are procured using a modified glass pipette with the adhesive coated tip discussed above. Tissue specimens as small as one cell can be procured with this method. The specificity of dissection represents a significant improvement over currently known techniques.

[0107] For DNA analysis of paraffin embedded tissue, the glass pipette with the dissected tissue specimen is placed in a single step extraction buffer solution of 50 mM Tris, pH 8.5, 1 mM EDTA, 0.5% Tween 20, and 0.2 mg/ml proteinase K, which removes the tissue from the pipette tip. The sample is incubated, depending on sample size, from two to twenty-four hours at about 37°C, followed by a ten minute incubation at about 95⁰C. The glass pipette tip can then be sterilized and reused, although this is not generally recommended in the case of PCR-based analysis due to the potential amplification of cross-contaminating materials.

B. Preparation of Nucleic Acids

[0108] As is apparent to one of ordinary skill in the art, nucleic acid samples used in the methods and assays of the invention may be prepared by any available method or process. Methods of isolating total mRNA are also well known to those of skill in the art. For example, methods of isolation and purification of nucleic acids are described in detail in Chapter 3 of Laboratory Techniques in Biochemistry and Molecular Biology: Hybridization With Nucleic Acid Probes, Part I Theory and Nucleic Acid Preparation, Tijssen, (1993) (editor) Elsevier Press. Such samples include RNA samples, but also include cDNA synthesized from a mRNA sample isolated from a cell or tissue of interest. Such samples also include DNA amplified from the cDNA, and an RNA transcribed from the amplified DNA. One of skill in the art would appreciate that it is desirable to inhibit or destroy RNase present in homogenates before homogenates can be used.

[0109] Biological samples may be of any biological tissue or fluid or cells from any organism as well as cells raised in vitro, such as cell lines and tissue culture cells. Frequently the sample will be a "clinical sample" which is a sample derived from a patient. Typical clinical samples include, but are not limited to, sputum, blood, blood-cells (e.g., white cells), tissue or fine needle biopsy samples, urine, peritoneal fluid, and pleural fluid, or cells therefrom. Biological samples may also include sections of tissues, such as frozen sections or formalin fixed sections taken for histological purposes.

C. Microarray Analysis

[0110] Identification of a nucleic acid sequence capable of binding to a biomolecule of interest can be achieved by immobilizing a library of nucleic acids onto the substrate surface so that each unique nucleic acid was located at a defined position to form an array, The array is then exposed to the biomolecule under conditions which favored binding of the biomolecule to the nucleic acids. Non-specifically binding biomolecules are washed away using mild to stringent buffer conditions depending on the level of specificity of binding desired. The nucleic acid array is then analyzed to determine which nucleic acid sequences bound to the biomolecule. Preferably the biomolecules would carry a fluorescent tag for use in detection of the location of the bound nucleic acids.

[0111] An assay using an immobilized array of nucleic acid sequences can be used for determining the sequence of an unknown nucleic acid; single nucleotide polymorphism (SNP) analysis; analysis of gene expression patterns from a particular species, tissue, cell type, etc.; gene identification; etc.

[0112] Microarrays can be purchased from commercial sources, for example Affymetrix. However, microarrays may be prepared, used, and analyzed using methods known in the art if desired (see, e.g., Brennan et al., 1995, U.S. Pat. No, 5,474,796; Schena et al., 1996, Proc. Natl. Acad. Sci. U.S.A. 93: 10614-10619; Baldeschweiler et al., 1995, PCT application WO95/251116; Shalon, et al., 1995, PCT application WO95/35505; Heller et al., 1997, Proc. Natl. Acad Sci. U.S.A. 94: 2150-2155; and Heller et al., 1997, U.S. Pat. No. 5,605,662).

[0113] Any hybridization assay format may be used, including solution-based and solid support-based assay formats. Solid supports containing oligonucleotide probes for differentially expressed genes of the invention can be filters, polyvinyl chloride dishes, silicon or glass based chips, etc. Such wafers and hybridization methods are widely available, for example, those disclosed by Beattie (WO 95/11755). Any solid surface to which oligonucleotides can be bound, either directly or indirectly, either covalently or non- covalently, can be used. A preferred solid support is a high density array or DNA chip. These contain a particular oligonucleotide probe in a predetermined location on the array. Each predetermined location may contain more than one molecule of the probe, but each molecule within the predetermined location has an identical sequence. Such predetermined locations are termed features. There may be, for example, about 2, 10, 100, 1000 to 10,000; 100,000 or 400,000 of such features on a single solid support. The solid support, or the area within which the probes are attached may be on the order of a square centimeter.

[0114] Oligonucleotide probe arrays for expression monitoring can be made and used according to any techniques known in the art (see for example, Lockhart et al., (1996) Nat. Biotechnol. 14, 1675-1680; McGaIl et al, (1996) Proc. Nat. Acad. Sci. USA 93, 13555-13460). Such probe arrays may contain at least two or more oligonucleotides that are complementary to or hybridize to two or more of the genes described herein. Such arrays may also contain oligonucleotides that are complementary or hybridize to at least about 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 50, 70, 100 or more genes. Examples, include, but not limited to those described herein.

[0115] The genes which are assayed are typically in the form of mRNA or reverse transcribed mRNA. The genes may be cloned or not and the genes may be amplified or not. The cloning itself does not appear to bias the representation of genes within a population. However, it may be preferable to use poly A⁺RNA as a source, as it can be used with less processing steps.

[0116] Probes based on the sequences of the genes may be prepared by any commonly available method. Oligonucleotide probes for assaying the tissue or cell sample are preferably of sufficient length to specifically hybridize only to appropriate, complementary genes or transcripts. Typically the oligonucleotide probes will be at least 10, 12, 14, 16, 18, 20 or 25 nucleotides in length. In some cases longer probes of at least 30, 40, or 50 nucleotides will be desirable.

[0117] As used herein, oligonucleotide sequences that are complementary to one or more of the genes described herein, refers to oligonucleotides that are capable of hybridizing under stringent conditions to at least part of the nucleotide sequence of said genes. Such hybridizable oligonucleotides will typically exhibit at least about 75% sequence identity at the nucleotide level to said genes, preferably about 80% or 85% sequence identity or more preferably about 90% or 95% or more sequence identity to said genes. [0118] "Bind(s) substantially" refers to complementary hybridization between a probe nucleic acid and a target nucleic acid and embraces minor mismatches that can be accommodated by reducing the stringency of the hybridization media to achieve the desired detection of the target polynucleotide sequence.

[0119] "Background" or "background signal intensity" refer to hybridization signals resulting from non-specific binding, or other interactions, between the labeled target nucleic acids and components of the oligonucleotide array (e.g., the oligonucleotide probes, control probes, the array substrate, etc.). Background signals may also be produced by intrinsic fluorescence of the array components themselves. A single background signal can be calculated for the entire array, or a different background signal may be calculated for each target nucleic acid. In a preferred embodiment, background is calculated as the average hybridization signal intensity for the lowest 5% to 10% of the probes in the array, or, where a different background signal is calculated for each target gene, for the lowest 5% to 10% of the probes for each gene. Of course, one of skill in the art will appreciate that where the probes to a particular gene hybridize well and thus appear to be specifically binding to a target sequence, they should not be used in a background signal calculation.

[0120] Alternatively, background may be calculated as the average hybridization signal intensity produced by hybridization to probes that are not complementary to any sequence found in the sample (e.g., probes directed to nucleic acids of the opposite sense or to genes not found in the sample such as bacterial genes where the sample is mammalian nucleic acids). Background can also be calculated as the average signal intensity produced by regions of the array that lack any probes at all. The phrase "hybridizing specifically to" refers to the binding, duplexing or hybridizing of a molecule substantially to or only to a particular nucleotide sequence or sequences under stringent conditions when that sequence is present in a complex mixture (e.g., total cellular) DNA or RNA.

[0121] Assays and methods of the invention can utilize available formats to simultaneously screen at least about 100, preferably about 1000, more preferably about 10,000 and most preferably about 1,000,000 or more different nucleic acid hybridizations.

[0122] "Mismatch control" or "mismatch probe" refer to a probe whose sequence is deliberately selected not to be perfectly complementary to a particular target sequence. For each mismatch (MM) control in a high-density array there typically exists a corresponding perfect match (PM) probe that is perfectly complementary to the same particular target sequence. The mismatch may comprise one or more bases.

[0123] While a mismatch(s) may be located anywhere in the mismatch probe, terminal mismatches are less desirable as a terminal mismatch is less likely to prevent hybridization of the target sequence. In a particularly preferred embodiment, the mismatch is located at or near the center of the probe such that the mismatch is most likely to destabilize the duplex with the target sequence under the test hybridization conditions.

[0124] The term "perfect match probe" refers to a probe that has a sequence that is perfectly complementary to a particular target sequence. The test probe is typically perfectly complementary to a portion (subsequence) of the target sequence. The perfect match (PM) probe can be a "test probe", a "normalization control" probe, an expression level control probe and the like. A perfect match control or perfect match probe is, however, distinguished from a "mismatch control" or "mismatch probe."

[0125] A "probe" is defined as a nucleic acid, capable of binding to a target nucleic acid of complementary sequence through one or more types of chemical bonds, usually through complementary base pairing, usually through hydrogen bond formation. As used herein, a probe may include natural (i.e., A, G, U, C or T) or modified bases (7- deazaguanosine, inosine, etc.). In addition, the bases in probes may be joined by a linkage other than a phosphodiester bond, so long as it does not interfere with hybridization. Thus, probes may be peptide nucleic acids in which the constituent bases are joined by peptide bonds rather than phosphodiester linkages.

B. Nucleic Acid Analysis

[0126] The "percentage of sequence identity" or "sequence identity" is determined by comparing two optimally aligned sequences or subsequences over a comparison window or span, wherein the portion of the polynucleotide sequence in the comparison window may optionally comprise additions or deletions (i.e., gaps) as compared to the reference sequence (which does not comprise additions or deletions) for optimal alignment of the two sequences. The percentage is calculated by determining the number of positions at which the identical monomer unit (e.g., nucleic acid base or amino acid residue) occurs in both sequences to yield the number of matched positions, dividing the number of matched positions by the total number of positions in the window of comparison and multiplying the result by 100 to yield the percentage of sequence identity. Percentage sequence identity when calculated using the programs GAP or BESTFIT is calculated using default gap weights.

[0127] Homology or identity may be determined by BLAST (Basic Local Alignment Search Tool) analysis using the algorithm employed by the programs blastp, blastn, blastx, tblastn and tblastx (Karlin et al., (1990) Proc. Natl. Acad ScL USA 87, 2264- 2268 and Altschul, (1993) J. MoI. Evol. 36, 290-300, fully incorporated by reference) which are tailored for sequence similarity searching. The approach used by the BLAST program is to first consider similar segments between a query sequence and a database sequence, then to evaluate the statistical significance of all matches that are identified and finally to summarize only those matches which satisfy a preselected threshold of significance. For a discussion of basic issues in similarity searching of sequence databases, see Altschul et al., (1994) Nature Genet. 6, 119-129) which is filly incorporated by reference. The search parameters for histogram, descriptions, alignments, expect (i.e., the statistical significance threshold for reporting matches against database sequences), cutoff, matrix and filter are at the default settings. The default scoring matrix used by blastp, blastx, tblastn, and tblastx is the BLOSUM62 matrix (Henikoff et al., (1992) Proc. Natl. Acad. Sci. USA 89, 10915-10919, fully incorporated by reference). Four blastn parameters were adjusted as follows: Q=IO (gap creation penalty); R=I 0 (gap extension penalty); wink=l (generates word hits at every wink™ position along the query); and gapw=16 (sets the window width within which gapped alignments are generated). The equivalent Blastp parameter settings were Q=9; R=2; wink=l ; and gapw=32. A Bestfit comparison between sequences, available in the GCG package version 10.0, uses DNA parameters GAP=50 (gap creation penalty) and LEN=3 (gap extension penalty) and the equivalent settings in protein comparisons are GAP=8 and LEN=2.

[0128] One of skill in the art will appreciate that an enormous number of array designs are suitable for the practice of this invention. The high density array will typically include a number of probes that specifically hybridize to the sequences of interest. See WO 99/32660 for methods of producing probes for a given gene or genes. In addition, in a preferred embodiment, the array will include one or more control probes.

[0129] High density array chips of the invention include "test probes." Test probes may be oligonucleotides that range from about 5 to about 500 or about 5 to about 50 nucleotides, more preferably from about 10 to about 40 nucleotides and most preferably from about 15 to about 40 nucleotides in length. In other particularly preferred embodiments the probes are about 20 to 25 nucleotides in length. In another preferred embodiment, test probes are double or single strand DNA sequences. DNA sequences are isolated or cloned from natural sources or amplified from natural sources using natural nucleic acid as templates. These probes have sequences complementary to particular subsequences of the genes whose expression they are designed to detect. Thus, the test probes are capable of specifically hybridizing to the target nucleic acid they are to detect.

[0130] In addition to test probes that bind the target nucleic acid(s) of interest, the high density array can contain a number of control probes. The control probes fall into three categories referred to herein as (1) normalization controls; (2) expression level controls; and (3) mismatch controls.

[0131] Normalization controls are oligonucleotide or other nucleic acid probes that are complementary to labeled reference oligonucleotides or other nucleic acid sequences that are added to the nucleic acid sample. The signals obtained from the normalization controls after hybridization provide a control for variations in hybridization conditions, label intensity, "reading" efficiency and other factors that may cause the signal of a perfect hybridization to vary between arrays. In a preferred embodiment, signals (e.g., fluorescence intensity) read from all other probes in the array are divided by the signal (e.g., fluorescence intensity) from the control probes thereby normalizing the measurements.

[0132] Virtually any probe may serve as a normalization control. However, it is recognized that hybridization efficiency varies with base composition and probe length. Preferred normalization probes are selected to reflect the average length of the other probes present in the array, however, they can be selected to cover a range of lengths. The normalization control(s) can also be selected to reflect the (average) base composition of the other probes in the array, however in a preferred embodiment, only one or a few probes are used and they are selected such that they hybridize well (i.e., no secondary structure) and do not match any target-specific probes.

[0133] Expression level controls are probes that hybridize specifically with constitutively expressed genes in the biological sample. Virtually any constitutively expressed gene provides a suitable target for expression level controls. Typical expression level controJ probes have sequences complementary to subsequences.of constitutiveJy expressed "housekeeping genes" including, but not limited to the β-actin gene, the transferrin receptor gene, the GAPDH gene, and the like.

[0134] Mismatch controls may also be provided for the probes to the target genes, for expression Jevel controls or for normalization controls. Mismatch controls are oligonucleotide probes or other nucleic acid probes identical to their corresponding test or control probes except for the presence of one or more mismatched bases. A mismatched base is a base selected so that it is not complementary to the corresponding base in the target sequence to which the probe would otherwise specifically hybridize. One or more mismatches are selected such that under appropriate hybridization conditions (e.g., stringent conditions) the test or control probe would be expected to hybridize with its target sequence, but the mismatch probe would not hybridize (or would hybridize to a significantly lesser extent). Preferred mismatch probes contain a central mismatch. Thus, for example, where a probe is a twenty-mer, a corresponding mismatch probe will have the identical sequence except for a single base mismatch (e.g., substituting a G, a C or a T for an A) at any of positions 6 through 14 (the central mismatch).

[0135] Mismatch probes thus provide a control for non-specific binding or cross hybridization to a nucleic acid in the sample other than the target to which the probe is directed. Mismatch probes also indicate whether a hybridization is specific or not. For example, if the target is present the perfect match probes should be consistently brighter than the mismatch probes. In addition, if all central mismatches are present, the mismatch probes can be used to detect a mutation. The difference in intensity between the perfect match and the mismatch probe (I_PM-I_MM) provides a good measure of the concentration of the hybridized material.

E. Nucleic Acid Hybridization

10136] Nucleic acid hybridization simply involves contacting a probe and target nucleic acid under conditions where the probe and its complementary target can form stable hybrid duplexes through complementary base pairing (see Lockhart et al., (1999) WO 99/32660). The nucleic acids that do not form hybrid duplexes are then washed away leaving the hybridized nucleic acids to be detected, typically through detection of an attached detectable label. It is generally recognized that nucleic acids are denatured by increasing the temperature or decreasing the salt concentration of the buffer containing the nucleic acids.

[0137] Under low stringency conditions (e.g., low temperature and/or high salt) hybrid duplexes (e.g., DNA-DNA, EtNA-RNA or RNA-DNA) will form even where the annealed sequences are not perfectly complementary.

[0138] Thus specificity of hybridization is reduced at lower stringency. Conversely, at higher stringency (e.g., higher temperature or lower salt) successful hybridization requires fewer mismatches. One of skill in the art will appreciate that hybridization conditions may be selected to provide any degree of stringency. Hybridization can be performed at low stringency, for example, in 6 * SSPE-T at 37° C. (0.005% Triton x- 100) to ensure hybridization and then subsequent washes are performed at higher stringency (e.g., 1 x SSPE-T at 37° C.) to eliminate mismatched hybrid duplexes. Successive washes may be performed at increasingly higher stringency (e.g., down to as low as 0.25 * SSPET at 37° C. to 50° C.) until a desired level of hybridization specificity is obtained. Stringency can also be increased by addition of agents such as formamide. Hybridization specificity may be evaluated by comparison of hybridization to the test probes with hybridization to the various controls that can be present (e.g., expression level control, normalization control, mismatch controls, etc.).

[0139] In general, there is a tradeoff between hybridization specificity (stringency) and signal intensity. Thus, in a preferred embodiment, the wash is performed at the highest stringency that produces consistent results and that provides a signal intensity greater than approximately 10% of the background intensity. Thus, in a preferred embodiment, the hybridized array may be washed at successively higher stringency solutions and read between each wash. Analysis of the data sets thus produced will reveal a wash stringency above which the hybridization pattern is not appreciably altered and which provides adequate signal for the particular oligonucleotide probes of interest.

[0140] The hybridized nucleic acids are typically detected by detecting one or more labels attached to the sample nucleic acids. The labels may be incorporated by any of a number of means well known to those of skill in the art (see Lockhart et al., (1999) WO 99/32660). F. Identifying Genes

[0141] Identity of genes, or variants thereof, can be verified using techniques well known in the art. Examples include but are not limited to, nucleic acid sequencing of amplified genes, hybridization techniques such as single nucleic acid polymorphism analysis (SNP), microarrays wherein the molecule of interest is immobilized on a biochip. Overlapping cDNA clones can be sequenced by the dideoxy chain reaction using fluorescent dye terminators and an ABI sequencer (Applied Biosystems, Foster City, Calif.).

[0142] Any type of assay wherein one component is immobilized may be carried out using the substrate platforms of the invention. Bioassays utilizing an immobilized component are well known in the art. Examples of assays utilizing an immobilized component include for example, immunoassays, analysis of protein-protein interactions, analysis of protein-nucleic acid interactions, analysis of nucleic acid-nucleic acid interactions, receptor binding assays, enzyme assays, phosphorylation assays, diagnostic assays for determination of disease state, genetic profiling for drug compatibility analysis, SNP detection, etc.

Databases for identifying genes

[0143] Databases may also contain information associated with a given sequence or tissue sample such as descriptive information about the gene associated with the sequence information, or descriptive information concerning the clinical status of the tissue sample, or the patient from which the sample was derived. The database may be designed to include different parts, for instance a sequences database and a gene expression database. Methods for the configuration and construction of such databases are widely available, for instance, see Akerblom et al., (1999) U.S. Pat. No. 5,953,727, which is herein incorporated by reference in its entirety

[0144] For example, and in no way limiting the invention, the data generated by scanning microarrays probed with cRNA from microdissected or purified lymphoid tissues are imported into GREX software (Affymetrix). The PLIER algorithm are used to generate relative RNA signal values. PLIER parameters are those established for similarity to robust multichip analysis, namely quantile normalization, use of perfect matched oligonucleotides only, percentile background, and quick signal optimization. The Microarray Suite 5.0 statistical algorithm, including both matched and mismatched oligonucleotides, are used to calculate the probability that differences between each pair is not due to chance. The calculated probabilities for each probe set (gene) for the replicate gene chips of each microdissected region are averaged. An absolute detection call (present, marginal, or absent) for each gene are determined based on this pooled probability, using default levels: p<0.04, 0.04-0.06, and >0.06, respectively. A list of all genes designated as present in the each microdissected region are prepared, while genes found to be marginal or absent are filtered out.

[0145] Expression of the genes in this list is analyzed in each lymphoid microarray, to establish a list of stromal-specific genes. This is accomplished by identifying those genes present in microdissected regions, but not expressed in any of the lymphoid constituents of those regions. Such analysis of the sub-capsular cortex yields 420 genes expressed in this region that are not found in cortical thymocytes. Each list of stromal- specific genes can then be further filtered by sorting the results based on highest expression (signal) levels, or by those that differ most substantially from the mean of all genes expressed on the chip. This process is reiterated for each microdissected region, and for each lymphoid subtype, until a full accounting of non-lymphoid (i.e., stromal) gene expression is mapped for the entire thymus.

[0146] A database can be linked to an outside or external database. For example, the external database is GenBank and the associated databases maintained by the National Center for Biotechnology Information (NCBI).

[0147] Any appropriate computer platform may be used to perform the necessary comparisons between sequence information, gene expression information and any other information in the database or provided as an input. For example, a large number of computer workstations are available from a variety of manufacturers, such has those available from Silicon Graphics. Client-server environments, database servers and networks are also widely available and appropriate platforms for the databases of the invention.

[0148] A database can be used to produce, among other things, electronic Northerns to allow the user to determine the cell type or tissue in which a given gene is expressed and to allow determination of the abundance or expression level of a given gene in a particular tissue or cell.

[0149] A database also be used to present information identifying the expression level in a tissue or cell of a set of genes. Such methods may be used to predict the physiological state of a given tissue by comparing the level of expression of a gene or genes from a sample to the expression levels found in tissue from normal, malignant or carcinoma. Such methods may also be used in the drug or agent screening assays as described below.

[0150] For example, and in no way limiting the invention, the inventive methodology can be used to identify genes and/or variants and correlate the effects of the protein encoded by these genes, when a patient is diagnosed with cancer. The identification of genes which can distinguish between susceptible and resistant individuals is important for distinguishing which nucleic acid sequences render individuals susceptible to cancers.

[0151] Tissue samples from patients are microdissected, nucleic acid molecules isolated and subjected to microarray analysis of nucleic acids, such as for example, RNA. This is followed by subtraction of genes expressed in the tumors versus the normal cells, tumors at different stages, pre-metastatic tumors and the like. The genes identified from individuals can also be amplified by PCR and sequenced if desired, by methods well known in the art. As more gene sequences and their amino acid sequences are identified, allows for a correlation between the effects of tumor progression expression and different gene sequences.

[0152] As more genes or variants thereof, are identified, oligonucleotide sequences are generated, or fragments thereof, may be employed as probes in the purification, isolation and detection of genes with similar sequences. Identification of a nucleic acid sequence capable of binding to a biomolecule of interest can be achieved by immobilizing a library of nucleic acids onto the substrate surface so that each unique nucleic acid was located at a defined position to form an array. The array would then be exposed to the biomolecule under conditions which favored binding of the biomolecule to the nucleic acids. Non- specifically binding biomolecules could be washed away using mild to stringent buffer conditions depending on the level of specificity of binding desired. The nucleic acid array would then be analyzed to determine which nucleic acid sequences bound to the biomolecule. Preferably the biomolecules would carry a fluorescent tag for use in detection of the location of the bound nucleic acids. Assays using an immobilized array of nucleic acid sequences may be used for determining the sequence of an unknown nucleic acid; single nucleotide polymorphism (SNP) analysis; analysis of gene expression patterns from a particular species, tissue, cell type, etc.; gene identification; etc. Any sequence can then be tested in macrophage viability assays described infra, or any other physical phenotypic criteria such as localization, MAP kinase 3 cleavage patterns and the like. [0153] Other methods to determine the contributions of individual genes and or variants thereof, and their expression products. Genes or variants, thereof, can be isolated. Techniques are available to inactivate or alter any genetic region to any mutation desired by using targeted homologous recombination to insert specific changes into chromosomal variants. One approach for detecting homologous alteration events uses the polymerase chain reaction (PCR) to screen pools of transformant cells for homologous insertion, followed by screening individual clones (Kim et al., Nucleic Acids Res. 16:8887-8903 (1988); Kim et al, Gene 103:227-233 (1991)). Alternatively, a positive genetic selection approach has been developed in which a marker gene is constructed which will only be active if homologous insertion occurs, allowing these recombinants to be selected directly (Sedivy et al., Proc. Natl. Acad. Sci. USA 86:227-231 (1989)). One of the most general approaches developed for selecting homologous recombinants is the positive-negative selection (PNS) method developed for genes for which no direct selection of the alteration exists (Mansour et al., Nature 336:348-352: (1988); Capecchi, Science 244:1288-1292, (1989); Capecchi, Trends in Genet. 5:70-76 (1989)). The PNS method is more efficient for targeting genes that are not expressed at high levels because the marker gene has its own promoter. Nonhomologous recombinants are selected against by using the Herpes Simplex virus thymidine kinase (HSV- TK) gene and selecting against its nonhomologous insertion with the herpes drugs such as gancyclovir (GANC) or FIAU (l-(2-deoxy 2-fhioro-B-D-arabinofluranosyl)-5-iodouracil). By this counter-selection, the number of homologous recombinants in the surviving transformants can be enriched. Such transformants can be correlated with phenotypes as described infra.

G. Identifying Genes Associated with a Disease or Condition

[0154] Many biological functions are accomplished by altering the expression of various genes through transcriptional (e.g., through control of initiation, provision of RNA precursors, RNA processing, etc.) and/or translational control. For example, fundamental biological processes such as cell cycle, cell differentiation, and cell death, are often characterized by the variations in the expression levels of groups of genes.

[0155] Changes in gene expression also are associated with pathogenesis. For example, the lack of sufficient expression of functional tumor suppressor genes and/or the over expression of oncogene/protooncogenes could lead to tumorgenesis or hyperplastic growth of cells (Marshall, (1991) Cell, 64,313-326; Weinberg, (1991) Science, 254, 1138- 1146). Thus, changes in the expression levels of particular genes, such as oncogenes or tumor suppressors, serve as signposts for the presence and progression of various diseases.

[0156] Monitoring changes in gene expression may also provide certain advantages during drug screening development. Often drugs are screened and prescreened for the ability to interact with a major target without regard to other effects the drugs have on cells. Often such other effects cause toxicity in the whole animal, which prevent the development and use of the potential drug.

[0157] Accordingly, the present invention contemplates, for example, identifying genetic differences between stromal cells from various microenvironmental compartments of the post-natal thymus. Briefly, RNA is isolated from microdissected regions corresponding to functionally-defined thymic microenvironments, as well as from the thymocytes contained therein. High-fidelity linear amplification, labeling, and microarray screening is performed and expressed genes are then identified, for example, in stromal cells in each microdissected region. These expressed genes are then subjected to gene profiling by characterizing and making distinctions between the gene expression profiles of, for example, stromal cells in that region.

[0158] In another aspect, the invention provides methodology for aiding a human tumor and/or tumor disorder diagnosis and/or whether the tumor has metastatic potential by identifying gene expression of specific tumor cell types. The methods used herein, identify, for example, expression of genes at different tumor stages and will provide a profile of genes expressed at each stage, pre-metastatic stage and during metastasis. The identified genes can be used singularly or in combination with other markers of tumors in any set, for example, CEA, Her2⁺. Many tumor antigens are well known in the art. See for example, Van den Eynde BJ, van der Bruggen P. Curr Opin Immunol 1997; 9: 684-93; Houghton AN, Gold JS, Blachere NE. Curr Opin Immunol 2001 ; 13: 134-140; van der Bruggen P, Zhang Y, Chaux P, Stroobant V, Panichelli C, Schultz ES, Chapiro J, Van den Eynde BJ, Brasseur F, Boon T. Immunol Rev 2002; 188: 51-64, which are herein incorporated by reference.

[0159] Non-limiting examples of tumor antigens, include, tumor antigens resulting from mutations, such as: alpha-actinin-4 (lung carcinoma); BCR-ABL fusion protein (b3a2) (chronic myeloid leukemia); CASP-8 (head and neck squamous cell carcinoma); beta- catenin (melanoma); Cdc27 (melanoma); CDK4 (melanoma); dek-can fusion protein (myeloid leukemia); Elongation factor 2 (lung squamous carcinoa); ETV6-AML1 fusion protein (acute lymphoblastic leukemia); LDLR-fucosyltransferaseAS fusion protein (melanoma); overexpression of HLA-A2^d (renal cell carcinoma); hsp70-2 (renal cell carcinoma); KIAAO205 (bladder tumor); MART2 (melanoma); MUM-If (melanoma); MUM-2 (melanoma); MUM-3 (melanoma); neo-PAP (melanoma); Myosin class I (melanoma); 0S-9g (melanoma); pml-RARalpha fusion protein (promyelocytic leukemia); PTPRK (melanoma); K-ras (pancreatic adenocarcinoma); N-ras (melanoma). Examples of differentiation tumor antigens include, but not limited to: CEA (gut carcinoma); gplOO / PmeI17 (melanoma); Kallikrein 4 (prostate); mammaglobin-A (breast cancer); Melan-A / MART-I (melanoma); PSA (prostate carcinoma); TRP-I / gp75 (melanoma); TRP-2 (melanoma); tyrosinase (melanoma). Over or under-expressed tumor antigens include but are not limited to: CPSF (ubiquitous); EphA3 ; G250 / MN / CAIX (stomach, liver, pancreas); HER-2/neu; Intestinal carboxyl esterase (liver, intestine, kidney); alpha-foetoprotein (liver ); M-CSF (liver, kidney); MUCl (glandular epithelia); p53 (ubiquitous); PRAME (testis, ovary, endometrium, adrenals); PSMA (prostate, CNS, liver); RAGE-I (retina); RU2AS (testis, kidney, bladder); survivin (ubiquitous); Telomerase (testis, thymus, bone marrow, lymph nodes); WTl (testis, ovary, bone marrow, spleen); CA 125 (ovarian).

[0160J The gene expression profiles of tumor cells are differentially present in samples of a human patient, for example a cancer patient. For example, some are expressed at an elevated level and/or are present at a higher frequency in human patients with tumor and/or cancer related disorders than in normal subjects. Therefore, detection of one or more of these genes or nucleic acids in a person would provide useful information regarding the probability that the person may have tumor and/or cancer related disorder.

[0161] Accordingly, embodiments of the invention include methods for diagnosing human tumor and/or cancer related disorders, by (a) detecting gene expression in a sample, wherein the sample is subjected to microdissection, and (b) subjecting the nucleic acids or genes to microarray analysis and correlating the detection of the genes or nucleic acid molecules with a probable diagnosis of human tumor and/or cancer related disorder. The correlation takes into account the different types of expressed molecules, the degree of expression and the like, in the sample compared to a control genetic profile (e.g., in normal subjects in whom human tumor is undetectable). The correlation takes into account the presence or absence of the genes in a test sample and the frequency of detection of the same genes in a control. That is, comparing nucleic acid profiles from microdissected tissue regions and from purified individual cellular nucleic acids and, identifying gene expression by specific cell types in mixed cell populations. The correlation may take into account both of such factors to facilitate determination of whether a subject has tumor, the degree of severity of the tumor, and subcellular location of the injury, or not.

[0162] Other diseases which can be diagnosed with the compositions of the invention (e.g., polypeptides, polynucleotides, and/or agonists or antagonists), include, but are not limited to, diabetes by identifying gene expression in pancreatic islets, for example, nervous system injuries, and diseases, disorders, and/or conditions which result in either a disconnection of axons, a diminution or degeneration of neurons, or demyelination. Nervous system lesions which may be treated, prevented, and/or diagnosed in a patient (including human and non-human mammalian patients) according to the invention, include but are not limited to, the following lesions of either the central (including spinal cord, brain) or peripheral nervous systems: (1) ischemic lesions, in which a lack of oxygen in a portion of the nervous system results in neuronal injury or death, including cerebral infarction or ischemia, or spinal cord infarction or ischemia; (2) traumatic lesions, including lesions caused by physical injury or associated with surgery, for example, lesions which sever a portion of the nervous system, or compression injuries; (3) malignant lesions, in which a portion of the nervous system is destroyed or injured by malignant tissue which is either a nervous system associated malignancy or a malignancy derived from non-nervous system tissue; (4) infectious lesions, in which a portion of the nervous system is destroyed or injured as a result of infection, for example, by an abscess or associated with infection by human immunodeficiency virus, herpes zoster, or herpes simplex virus or with Lyme disease, tuberculosis, syphilis; (5) degenerative lesions, in which a portion of the nervous system is destroyed or injured as a result of a degenerative process including but not limited to degeneration associated with Parkinson's disease, Alzheimer's disease, Huntington's chorea, or amyotrophic lateral sclerosis (ALS); (6) lesions associated with nutritional diseases, disorders, and/or conditions, in which a portion of the nervous system is destroyed or injured by a nutritional disorder or disorder of metabolism including but not limited to, vitamin B12 deficiency, folic acid deficiency, Wernicke disease, tobacco^alcohol amblyopia, Marchiafava- Bignami disease (primary degeneration of the corpus callosum), and alcoholic cerebellar degeneration; (7) neurological lesions associated with systemic diseases including, but not limited to, diabetes (diabetic neuropathy, Bell's palsy), systemic lupus erythematosus, carcinoma, or sarcoidosis; (8) lesions caused by toxic substances including alcohol, lead, or particular neurotoxins; and (9) demyelinated lesions in which a portion of the nervous system is destroyed or injured by a demyelinating disease including, but not limited to, multiple sclerosis, human immunodeficiency virus-associated myelopathy, transverse myelopathy or various etiologies, progressive multifocal leukoencephalopathy, and central pontine myelinolysis.

Autoimmunity

[0163] Similarly, gene expression in autoimmunity can be identified. The identification of genes that are expressed in autoimmune diseases can provide the necessary information relevant to mechanisms and development of drugs to treat these diseases.

[0164] The role of the thymus in immunity is discussed in detail in the examples which follow. CD4+ helper T cells (hereinafter referred to as Th cells) involved in the onset of allergic diseases or autoimmune diseases are classified based on the type of the cytokines they produce into two types, namely, type I helper T cells (hereinafter referred to as ThI cells) and type II helper T cells (hereinafter referred to as Th2 cells). ThI cells produce IL-2, IFN-γ, TNF-β and the like, whereby inducing a cellular immunity. On the other hand, Th2 cells produce IL-4, IL-5, IL-6, IL-10, IL- 13 and the like, whereby inducing a humoral immunity.

[0165] ThO cells which are common precursors for ThI cells and Th2 cells are differentiated into either ThI cells or Th2 cells in response to an antigenic stimulation and then becomes mature. For example, a bacterium such as Bacillus tuberculosis and a virus such as an influenza virus are known to induce the differentiation to ThI cells, while allergens such as a mite and a pollen are known to induce the differentiation to Th2 cells.

[0166] Recently, it has been reported that a polarized existence of ThI cells and Th2 cells in a body is involved greatly in a prevention of infection and induction of allergic diseases or autoimmune diseases, and it is expected that inhibiting an excessive differentiation to Th2 cells serve to give a therapeutic effect against allergic diseases or autoimmune diseases induced by Th2 cells.

[0167] Identification of gene expression in the triggering and maintenance of an autoimmune disease is important in understanding the mechanisms and development of candidate therapeutic agents for developing a treatment. [0168] Accordingly, in a preferred embodiment, a method of identifying gene expression in autoimmune diseases, said method comprising: microdissecting tissues comprising autoimmune cells; analyzing nucleic acids isolated from the microdissected tissues; analyzing purified individual cellular nucleic acids from the tissue; comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and, identifying gene expression by specific cell types in autoimmune diseases. The microdissected tissues can be from any source, such as for example, the thymus, or in the case of diabetes (e.g. Type I), tissues are microdissected from the pancreas.

Cell Signaling

[0169] Gene expression in cell signaling can also be identified. Cell-signaling is important in cell differentiation of, for example, stem cells, tumors, etc; maturation; activation e.g., cells of the immune system; secretion; and the like. Identification of gene expression as a result of cell signaling is important in understanding, diagnosing, and developing therapies.

G. Pharmaceutical Compositions, Kits, and Diagnostic Reagents

[0170] The invention includes kits combining, in different combinations, high- density oligonucleotide arrays, reagents for use with the arrays, signal detection and array- processing instruments, gene expression databases, and analysis and database management software described above. The kits may be used, for example, to predict or model the toxic response of a test compound, to monitor the progression of disease states, to identify genes that show promise as new drug targets and to screen known and newly designed drugs as discussed above.

[0171] Databases packaged with the kits are a compilation of expression patterns from human or laboratory animal genes and gene fragments. Data is collected from a repository, of both normal and diseased animal tissues and provides reproducible, quantitative results, i.e., the degree to which a gene is up-regulated or down-regulated under a given condition.

[0172] The kits can be used in the pharmaceutical industry, where the need for early drug testing is strong due to the high costs associated with drug development, but where bioinformatics, in particular gene expression informatics, is still lacking. These kits will reduce the costs, time and risks associated with traditional new drug screening using cell cultures and laboratory animals. The results of large-scale drug screening of pre-grouped patient populations, pharmacogenomics testing, can also be applied to select drugs with greater efficacy and fewer side-effects. The kits can also be used by smaller biotechnology companies and research institutes who do not have the facilities for performing such large- scale testing themselves.

[0173] Databases and software designed for use with use with microarrays is discussed in the examples which follow. Other alternatives, for example are discussed in Balaban et ah, U.S. Pat. No. Nos.6,229,911, a computer-implemented method for managing information, stored as indexed tables, collected from small or large numbers of microarrays, and U.S. Pat. No. 6,185,561, a computer-based method with data mining capability for collecting gene expression level data, adding additional attributes and reformatting the data to produce answers to various queries. Chee et al., U.S. Pat. No. 5,974,164, disclose a software- based method for identifying mutations in a nucleic acid sequence based on differences in probe fluorescence intensities between wild type and mutant sequences that hybridize to reference sequences.

Diagnostic markers

[0174] As described above, the genes and gene expression information may be used as diagnostic markers for the prediction or identification of the malignant state of a tissue, for example, the liver tissue. For instance, a liver tissue sample or other sample from a patient may be assayed by any of the methods described above, and the expression levels from a gene or genes may be compared to the expression levels found in normal liver tissue, tissue from metastatic liver cancer or hepatocellular carcinoma tissue. Expression profiles generated from the tissue or other sample that substantially resemble an expression profile from normal or diseased liver tissue may be used, for instance, to aid in disease diagnosis. Comparison of the expression data, as well as available sequence or other information may be done by researcher or diagnostician or may be done with the aid of a computer and databases as described above.

Candidate Drugs:

[0175] According to the present invention, the genes identified using the methods of the invention may be used as markers to evaluate the effects of a candidate drug or agent on a cell, particularly a cell undergoing malignant transformation, for instance, a liver cancer cell or tissue sample. A candidate drug or agent can be screened for the ability to simulate the transcription or expression of a given marker or markers (drug targets) or to down-regulate or counteract the transcription or expression of a marker or markers. According to the present invention, one can also compare the specificity of drugs' effects by looking at the number of markers which the drugs have and comparing them. More specific drugs will have fewer transcriptional targets. Similar sets of markers identified for two drugs indicates a similarity of effects. As used herein, an agent is said to modulate the expression of a nucleic acid of the invention if it is capable of up- or down-regulating expression of the nucleic acid in a cell.

[0176] In one assay format, gene chips containing probes to at least two genes may be used to directly monitor or detect changes in gene expression in the treated or exposed cell as described in more detail above. In another format, cell lines that contain reporter gene fusions between the open reading frame and/or the 3' or 5' regulatory regions of a gene and any assayable fusion partner may be prepared. Numerous assayable fusion partners are known and readily available including the firefly luciferase gene and the gene encoding chloramphenicol acetyltransferase (Alam et al., (1990) Anal. Biochem. 188, 245-254). Cell lines containing the reporter gene fusions are then exposed to the agent to be tested under appropriate conditions and time. Differential expression of the reporter gene between samples exposed to the agent and control samples identifies agents which modulate the expression of the nucleic acid.

[0177] Additional assay formats may be used to monitor the ability of the agent to modulate the expression of a gene identified by the methods of the invention. For instance, as described above, mRNA expression may be monitored directly by hybridization of probes to the nucleic acids of the invention. Cell lines are exposed to the agent to be tested under appropriate conditions and time and total RNA or mRNA is isolated by standard procedures such those disclosed in Sambrook et al., (1989) Molecular Cloning — A Laboratory Manual, Cold Spring Harbor Laboratory Press).

[0178] In another assay format, cells or cell lines are first identified using the methods of the invention which express the gene products of the invention physiologically. Cell and/or cell lines so identified would be expected to comprise the necessary cellular machinery such that the fidelity of modulation of the transcriptional apparatus is maintained with regard to exogenous contact of agent with appropriate surface transduction mechanisms and/or the cytosolic cascades. Further, such cells or cell lines may be transduced or transfected with an expression vehicle (e.g., a plasmid or viral vector) construct comprising an operable non-translated 5'-promoter containing end of the structural gene encoding the instant gene products fused to one or more antigenic fragments, which are peculiar to the instant gene products, wherein said fragments are under the transcriptional control of said promoter and are expressed as polypeptides whose molecular weight can be distinguished from the naturally occurring polypeptides or may further comprise an immunologically distinct tag. Such a process is well known in the art (see Sambrook et al., (1989) Molecular Cloning — A Laboratory Manual, Cold Spring Harbor Laboratory Press).

[0179] Cells or cell lines transduced or transfected as outlined above are then contacted with agents under appropriate conditions; for example, the agent comprises a pharmaceutically acceptable excipient and is contacted with cells comprised in an aqueous physiological buffer such as phosphate buffered saline (PBS) at physiological pH, Eagles balanced salt solution (BSS) at physiological pH, PBS or BSS comprising serum or conditioned media comprising PBS or BSS and serum incubated at 37° C. Said conditions may be modulated as deemed necessary by one of skill in the art. Subsequent to contacting the cells with the agent, said cells will be disrupted and the polypeptides of the lysate are fractionated such that a polypeptide fraction is pooled and contacted with an antibody to be further processed by immunological assay (e.g., ELlSA, immunoprecipitation or Western blot). The pool of proteins isolated from the "agent-contacted" sample will be compared with a control sample where only the excipient is contacted with the cells and an increase or decrease in the immunologically generated signal from the "agent-contacted" sample compared to the control will be used to distinguish the effectiveness of the agent.

[0180] Another embodiment of the present invention provides methods for identifying agents that modulate the levels, concentration or at least one activity of a protein(s) encoded by the genes identified using the methods of the invention. Such methods or assays may utilize any means of monitoring or detecting the desired activity.

[0181] In one format, the relative amounts of a protein of the invention between a cell population that has been exposed to the agent to be tested compared to an unexposed control cell population may be assayed. In this format, probes such as specific antibodies are used to monitor the differential expression of the protein in the different cell populations. Cell lines or populations are exposed to the agent to be tested under appropriate conditions and time. Cellular lysates may be prepared from the exposed cell line or population and a control, unexposed cell line or population. The cellular lysates are then analyzed with the probe, such as a specific antibody.

[0182] Agents that are assayed in the above methods can be randomly selected or rationally selected or designed. As used herein, an agent is said to be randomly selected when the agent is chosen randomly without considering the specific sequences involved in the association of the a protein of the invention alone or with its associated substrates, binding partners, etc. An example of randomly selected agents is the use a chemical library or a peptide combinatorial library, or a growth broth of an organism.

[0183] As used herein, an agent is said to be rationally selected or designed when the agent is chosen on a nonrandom basis which takes into account the sequence of the target site and/or its conformation in connection with the agents action. Agents can be rationally selected or rationally designed by utilizing the peptide sequences that make up these sites. For example, a rationally selected peptide agent can be a peptide whose amino acid sequence is identical to or a derivative of any functional consensus site.

[0184] The agents of the present invention can be, as examples, peptides, small molecules, vitamin derivatives, as well as carbohydrates. Dominant negative proteins, DNA encoding these proteins, antibodies to these proteins, peptide fragments of these proteins or mimics of these proteins may be introduced into cells to affect function. "Mimic" as used herein refers to the modification of a region or several regions of a peptide molecule to provide a structure chemically different from the parent peptide but topographically and functionally similar to the parent peptide (see Grant, (1995) in Molecular Biology and Biotechnology Meyers (editor) VCH Publishers). A skilled artisan can readily recognize that there is no limit as to the structural nature of the agents of the present invention.

*********************

[0185] The following examples are offered by way of illustration, not by way of limitation. While specific examples have been provided, the above description is illustrative and not restrictive.

Example 1: Subtraction of Gene Expression in Tissue or Mixed Sample.

[0186] Microarray analysis is an extremely powerful technique for examining the expression of the majority of genes within a cell. In the case of cell cultures, the results of microarray analysis permit the researcher to understand the processes taking place within that cell type. Unfortunately samples isolated from higher organisms usually contain a number of different cell types. Each underlying cell type may play a distinct role in the function of that sample and have distinct patterns of gene expression. Microarray analysis on the whole sample will yield an average expression profile and could easily miss important genes specific to one cell type. If one is interested in the function of one of these underlying cell types, it is critical that the cells of interest be isolatable. Unfortunately, this can be a complex process and it is not always possible to physically isolate every cell type within a sample.

[0187] The present inventors have developed an algorithmic approach to calculate the expression profile of a tissue/sample of interest that consists of at least two types of cells. This technique electronically subtracts the expression profile of one component of a sample from the expression profile of the total sample, and thus revealing the profiles of the other component. By eliminating the need to identify procedures to purify a tissue from a complex mixture, this process can achieve significant cost savings and significantly earlier discoveries.

Algorithm Development

(01881 The expression for any one gene in mixture of two cell types (A + B) is equal to the sum of the expression of that gene in each cell type. The expression of any gene / in the mixture can be calculated from the equation:

E/_AB = pA * Ei_A + (1 - pA) * Ef₈ (1)

[0189] E/_AB is the expression of that gene as measured in the mixed sample, E/'_A is the expression of gene i in cell type A and E/_B is the expression of gene i in cell type B. The fractional percent of expression in the mix due to cell type A is given by pA. If this percentage and the expression in the mix and one of the cell types are known, then it is relatively straight forward to subtract the known expression of the first sample from known expression of the mix to yield the unknown expression pattern in the second cell type. Unfortunately, for most samples, the percentage of expression coming from each cell type is also unknown making the calculation of the expression pattern of the second cell type impossible to determine. In a few cases, a set of genes may be known to express exclusively in one of the cell types. These genes may then be used to determine the value of pA . This approach was also reported by Lu et al, Proc Natl Acad Sci USA 100(18): 10370- 10375 (2003), with their work on yeast cell profiling.

[0190] This method was used initially, but during analysis, it was observed that this fractional percent, pA, could also be approximated from the lowest ratio value of the intensity from the mixed sample genes divided by the intensity from the known sample genes. This relationship was explored using a number of simulated expression datasets. Further analysis with simulated noiseless expression data indicated that the minimal ratio value always approached the pA value. For noiseless expression data the equivalency of the minimal ratio value and pA can be demonstrated mathematically.

[0191] A mixed to pure ratio, Ri_miχyknown_> is calculated for each gene / from:

R'mix/pure ⁼ E/AB / EJ^'A (2)

[0192] By substituting in values from equation 1, equation 2 can be rearranged to

RlmixΛcnown = (pA * E/_A + (1 - pA) * E/_B) / El_A (3a)

Or

Rl^'mbΛnowπ = 0>A * Ei_A) / El_A + (( 1 - pA) * E/_B) / E/_A (3b)

[0193] For a constant expression level of gene / in cell type A, E/A, as the expression of gene / in cell type B increases, Rw_pure becomes very large. However, for genes with significant expression in cell type A and little or no expression in cell type B (EfB = 0), equation 3b simplifies to:

R/mix/known = (pA * E_A)/ E_A + 0 / E_A (4a)

Or

R'mix_/known = pA * (E_A/ E_A) (4b)

Or

R'mix/known ⁼ pA (4C)

[0194] Therefore, the minimum value for Rimis/_kn_own. where Ei_Λ is significant, can be assumed to be equivalent to pA. However, in non-simulated datasets, experimental noise results in some genes from either the mixed sample or the cell type A sample which have aberrantly high or low values. In addition, the expression of genes under one condition is never completely independent from the expression of those genes under different conditions. Finally, any differences between the expression in the cell type B sample used and the cell type B in the mixed sample may introduce further error. This noise results in a number of genes with aberrant R/_miχ/known values.

[0195] A number of different combinations of expression pattern sets were analyzed with simulated noisy data. These simulated noisy expression datasets were generated from pairs of expression samples from NCBI's GEO database where one sample represented the known cell type (A) and the other represented the unknown sample (B). These expression profiles were used as expression value seeds to generate a mixed (AB) expression profile. Noisy profiles were generated from the seeds and the mixed profiles by adding simulated noise. A comparison of the generated noisy samples was indistinguishable from real data sets. Lists of sorted Rimix_Λn_own values, from smallest to highest, were generated from each of these datasets. Analysis of the R/_mix_/kn_own values generated from simulated datasets at series of fractional percent (pA) values is shown in Figure IA. The data indicate that there is a dependence of ratio list to the fractional percent (pA) of the "purifiable" cell type in the mixed sample. The effects of using different seed sets were examined at a constant fractional percent. The results indicate that the ratio lists varied based on the pair of experimental seeds used (Figure IB).

[0196] To develop an equation to calculate pA from the ratio data, ratio data sets were generated from 10 different combinations of seed experiments over a series of pA values ranging from 0.05 to 0.95. Ratio values were extracted from each sorted ratio data set at a series of positions within each list. The extracted values from a given position within the lists were plotted against the pA values used to generate the data. For example, the ratio values extracted at 5% and 20% index of the ranked list plotted against pA value are shown in Figure 2. Regression analysis was performed using a variety of techniques using values collected at different positions within the list. It was found that the pA value could best be calculated from a second order polynomial of the 5% value of the ranked lists. This polynomial resulted in a correlation coefficient at 0.977.

pA = -0.01831 + 0.3485 * R/_mixtoown s% + 1.182 * (RWkπown s% ^Λ 2) (5)

[0197] The calculated values correlated well over most values of pA with the most deviation at the extreme values. a. Analysis of Simulated Data

[0198] Equation 5 was used to calculate pA values from simulated data generated from one expression seed pairing (GSMl 8977 & GSMl 8979) over a range of actual pA values. The actual fractional percent was plotted against the calculated value in Figure 3. Linear regression indicates that the slope of the line was 0.998 with a maximal deviation at the ratio extreme values. The correlation coefficient for the calculated ratio versus the actual ratio was 0,996 indicating that the electronic-subtraction method accurately calculates fractional percent (pA) values of the 2 cell types. Other pairings produced similar results (data not shown).

[0199] Once the pA value was determined, the expression of each gene in the profile set could then be calculated from Equation 1. A variety of methods have been employed to determine which genes are differentially expressed between two samples. Jeffery, I. et al. BMC BioinformaticsJ :359 (2006). Fold-change was used in this analysis. Specifically, genes which were more than 2-fold elevated in the unknown sample as compared to the known sample over a range of mixture fractions were identified. Figure 4 demonstrates the sensitivity and specificity of the electronic-subtraction method. The results indicate that the predictive ability of this method remains strong until the fraction of the unknown in the sample falls below 5% of the total. It was also observed that many of the genes which had mixed sample-to-known sample ratios (R'_mi_x/_know_n) below the value calculated for the factional percent had aberrant fold-change values. Where those genes were removed, there was a substantial improvement in specificity and sensitivity (Figure 4B).

b. Analysis øfSev infected cell data

[0200] Two expression datasets from NCBPs GEO database were re-examined to determine additional relevant information could be identified. The datasets were chosen to represent different kinds of mixed samples commonly seen in experimental data. In the first set, the expression patterns were examined in virally infected macrophages of Tyner et al., Nat Med 11(11):1180-1187 (2005). The naϊve method was used to identify the fractional percent of expression from the infected cells as 0.667. This fraction approximates the rate of infected cells as visualized by Sev immunostain versus DAPI, Id. The electronic-subtraction method was used to calculate the expression patterns in the infected cells alone and two groups of genes specific to virally infected cells were identified. [0201] The first group consisted of genes identified by straight 2-fold or more expression in the Sev-infected macrophage cells versus the mock infected cells. A minimal intensity value was set in the infected sample as > 13 (the median intensity value) to eliminate the noise associated with low expressors. Under these criteria, 154 probesets are identified as specific to virally infected macrophages. These elevated probesets include the Ccl5 gene which was reported by the authors. Id. The second group of identified genes were those meeting the same criteria but using the electronic subtraction calculated value for the infected cells. Using the same criteria as above, 527 probesets are identified as specific to virally infected macrophages. The hypergeometric distribution was used and implemented in "Gene Ontology Browser" tool in SpotFire's "DecisionSite for Functional Genomics" to identify Gene Ontology (GO) categories which were over-represented in the genes selected by the two methods (Table 1). The hypergeometric distribution method is described by Tavazoie et al (1999). In their macrophage infection paper, Tyner et al. (2005) demonstrate that the elevated Ccl5 gene has a role in the regulation of apoptosis through the G_αi-PI3K-AKT and G_αj-MEK- ERK pathways. More genes in the GO category 'regulation of apoptosis' are identified using the subtraction method (Table 1). The addition of these genes results in an improvement in the confidence for this Biologic Function from 1.6e-3 to 5.2e-4.

Analysis of Regulatory T-cell data

[0202] In the second dataset, expression data from T-cells was re-examined, one dataset was from a wild-type and one dataset was from a Foxp3 deletion mouse. Fontenot, et al., Immunity 22(3):329-341 (2005). Wild-type T-cells contain a mixture of regulatory T-cells (T-reg) and noπ-T-reg cells. The fractional percent of expression due to the T-reg cells in the wild-type mouse was calculated to be 0.548 by the naive method. This value was used with the electronic-subtraction method to calculate the expression of the T-reg cells in the T-cells from the w.t. mouse. The fractional percent value compares favorably with the values of 0.537, 0.547, and 0.534 for genes I12ra (CD25), Tnfrsfl8 (Gitr), and Ctla4. These genes are specific to T-reg cells. Damoiseaux, J., Neth J Med 64(1 ):4-9 (2006); Maggi, et al. Autoimmun Rev, 4(8):579-586 (2005). Two groups of genes were identified which were 2- fold or more over-expressed using either the electronic subtraction calculated values for the T- reg cells or the mixed / Foxp3+ cells. The "Gene Ontology Browser" tool in SpotFire's "DecisionSite for Functional Genomics" was used to identify gene ontology categories which were over-represented in the genes selected by the 2 methods (Table 2). A number of processes which have been implicated in the function of T-reg cells were significantly over- represented in both sets of genes; however in all categories greater numbers of genes where identified by the electronic subtraction method. In all but one instance, these categories were identified with greater confidence as evidenced by the lower p-values.

10203) Accordingly, this electronic subtraction method can be used to extract the expression profile of underlying cell types without the need for time-consuming and costly physical purification of these important cell types. The methodology has potential applications in basic research as well as biomarker discovery. For example, the method could be applied to biopsy samples to remove the normal tissue profile from that of the cancerous cell profile. The electronic subtraction methodology also finds applicability to other expression technologies, such as but not limited to proteomics profiling.

Example 2: Application to Thymus.

[0204] The invention is applicable to all instances wherein the identity of genes expressed in cells or the identification of types of cells in tissues can be identified. For example, post-natal production of naive T lymphocytes is a complex process, the efficient execution of which requires conditions specific to the thymus. The thymic microenvironment is responsible for inducing a large number of discrete but interrelated functions, the first of which is the periodic induction of new progenitor recruitment from the blood, since the thymus contains no self-renewing potential.

[0205] The thymus produces multiple T lineages, including CD4 and CD8 cells as well as NK-T cells, T-regulatory cells, and γδ T cells, among others. Thus, the thymic environment continues to play a role in determining lineage fate long after the importation process is complete. The thymus also supports extensive progenitor proliferation, provides conditions that limit that proliferation, and enacts numerous other vary varied functions. However, remarkably few of the signals the thymus provides to enable these processes are known, the primarily ones being represented by ligands for Notch 1, IL7R, c-kit, and the TCR

[0206] Using the inventive methodology, the microenvironments that induce and/or support specific developmental events have been functionally mapped. For instance, the peri-medullary cortex is the region where blood-borne progenitors enter the thymus, and where the first ten rounds of proliferative expansion occur. Thus, stromal cells from this region must liberate factors that induce homing and retention of blood-borne progenitors, as well their proliferative expansion. Signals found in the inner cortex must result in further upregulation of mitotic activity, induction of RAG gene expression, and loss of most NK progenitor potential. The environment of the outer cortex must signal responses such as chromatin remodeling and TCRβ locus accessibility, and down-regulation of mitotic activity. The immediate sub-capsular region must provide signals that induce the DN/DP transition, as well as mitogenic and growth factors for the following wave of proliferation. Unlike outward- migrating DN cells, which depend on stromal cells to provide the matrix for migration as well as for developmental signals, DP cells moving inward do not appear to require constant contact with stromal cells. However, non-hematopoietic stroma clearly provide signals to DP cells, including MHC/peptides for positive selection, although one would suspect that other, as yet undefined, signals are also provided to DP by cortical stroma. The outer medulla is densely populated with dendritic cells that ensure stringent negative selection prior to entry into the medulla proper. However, medullary thymocytes only become functional after spending a prolonged period within the medulla, probably reflecting a period for additional antigen receptor screening and the induction of tolerance, although other maturational processes clearly occur.

[0207] Identification of the anatomic distribution of each progenitor stage, together with knowledge of their status and potentials, allowed us to functionally map stromal microenvironments within the thymic cortex, and within the medulla. By performing microarray analysis on purified thymocyte progenitor stages, their signal requirements were predicted by identifying which surface receptors were expressed. Further, during each stage of differentiation, new surface receptors would be expressed in anticipation of subsequent need and by analyzing changes in receptor expression in the context of where cells were in the thymus, the location of regionally restricted microenvironmental signals was predicted. This allowed for identifying a number of candidate receptors that are currently being evaluated.

[0208] Laser microdissection is used to isolate the entire region of tissue adjacent to the capsule, and to perform gene expression analysis (by high-density microarray) on the entire tissue region. The results of such analysis include genes expressed by subcapsular stromal cells, and also include genes expressed by the developing thymocytes in that region (primarily DN3 and preDP cells). However, the latter results could be filtered out on the basis of microarray results from purified lymphoid progenitors. [0209] The present methodology allowed the characterization of the transcriptional profile of stromal cells in the sub-capsular cortex. This methodology permitted the definition, for the first time, not just which specific stromal products define each microenvironment by signaling to lymphoid progenitors, but also how stromal cells in each region differ.

[0210] Cell signaling: The relationship between the location of progenitor cells in the post-natal thymus, and their differentiation is established. The signals that mediate directional migration of early progenitors are determined and the consequences of failed migration on early precursor differentiation in the thymus are evaluated.

[0211] The continuous production of new T lymphocytes is essential for effective immune function. Although thymus mass begins to progressively decline after puberty, the thymus continues to produce new T cells even late in life, although the levels of new T cell production decrease relative to thymus size. Homeostatic expansion can supplement the production of new T cells later in life, but such cells do not provide the same spectrum of immune surveillance, and this ultimately leads to increased susceptibility to infectious disease and/or autoimmunity with age. Understanding the biology of post-natal T cell production, therefore, represents an essential step in devising strategies that can reduce morbidity in individuals with secondary immunodeficiencies resulting from age and other insults such as chemotherapy.

[0212] Intravenous injection of stem cells to irradiated mice leads to long-term T cell reconstitution, but intrathymic injection leads only to a transient wave of T cell production. This and other related observations have been used to show that the thymus contains no self-renewing progenitors, but instead relies on marrow-derived progenitors that circulate in the bloodstream. The recruitment process is periodic, not continuous, suggesting that the biochemical signals responsible for new progenitor recruitment fluctuate in response to presently undefined stimuli. One function of the thymus is to expand these marrow-derived progenitors to generate a large pool of cells for MHC/TCR-mediated selection; during the period of intrathymic residence, each newly arrived progenitor undergoes approximately 20 serial cell divisions leading to the production of about a million CD4⁺8⁺ (T lineage double positive, DP) cells. About 12 of these 20 cell divisions occur at the CD4T (double negative, DN) stage, which includes three sub-stages designated DNl, DN2, and DN3. The remaining 7-8 divisions occur immediately after the transition to the DP stage in cells that express low surface levels of CD3, CD4, and CD8. We refer to this final lymphopoietic stage as preDP cells, consistent with expression of these lineage markers, as well as other hallmarks of DP cells such as in-frame TCRβ rearrangements and the ability to spontaneously acquire high levels of CD4 and CD8 in vitro.

[0213] In addition to extensive proliferative expansion, the thymus must provide the signals that instruct each new progenitor to adopt specific lineage fates. The thymus contains progenitors for multiple distinct hematopoietic lineages, including γδ T cells, γδ T cells, dendritic cells, and NK cells, as well as various sub-lineages of these, including CD4 or CD8 single positive (SP) αβ cells, NK-T cells, regulatory T cells, and others. The microenvironmental signals that induce differentiation into these various lineages are quite poorly understood, and very few of the differential signal requirements have been defined. This is at least partly because the different types of stromal cells that induce various aspects of these processes are not known.

Example 3: Stromal Cell Expression Profile.

[0214] Stromal cells are connective tissue cells of an organ found in the loose connective tissue. They are most often associated with the uterine mucosa, prostate, bone marrow precursor cells, and the ovary, as well as the hematopoietic system and elsewhere. Because stromal cells are difficult to isolate, little information is known concerning stromal gene expression patterns. However, because the present inventors have developed a methodology to electronically subtract gene expression in one or more components of a tissue from a mixture, such method can be used for analyzing stromal gene expression.

Identification of genes expressed specifically by stromal cells,

10215] KNA from microdissected tissue include lymphoid as well as stromal cells. The data generated by scanning microarrays probed with cRNA from microdissected or purified lymphoid tissues are imported into GREX software (Affymetrix). The PLIER algorithm are used to generate relative RNA signal values. PLIER parameters are those established for similarity to robust multichip analysis, namely quantile normalization, use of perfect matched oligonucleotides only, percentile background, and quick signal optimization. The Microarray Suite 5.0 statistical algorithm, including both matched and mismatched oligonucleotides, are used to calculate the probability that differences between each pair is not due to chance. The calculated probabilities for each probe set (gene) for the replicate gene chips of each microdissected region are averaged. An absolute detection call (present, marginal, or absent) for each gene are determined based on this pooled probability, using default levels: p<0.04, 0.04-0.06, and >0.06, respectively. A list of all genes designated as present in the each microdissected region are prepared, while genes found to be marginal or absent are filtered out. Expression of the genes in this list will then be analyzed in each lymphoid microarray, to establish a list of stromal-specific genes.

[0216] This is accomplished by identifying those genes present in microdissected regions, but not expressed in any of the lymphoid constituents of those regions. Such analysis of the sub-capsular cortex yields 420 genes expressed in this region that are not found in cortical thymocytes. Each list of stromal-specific genes can then be further filtered by sorting the results based on highest expression (signal) levels, or by those that differ most substantially from the mean of all genes expressed on the chip. This process are reiterated for each microdissected region, and for each lymphoid subtype, until a full accounting of non-lymphoid (ie, stromal) gene expression is mapped for the entire thymus.

[0217] Genetic profiles of stromal cells in each region: The stromal gene list for each region are compared to the lists for other regions, based on absolute detection calls (present, absent, marginal), and a refined list of those genes exclusive to each region are extracted. Individual analysis of functionally-defined regions can give detailed information corresponding to specific developmental events. This list will reveal genes that may be targeted for deletion in stromal cells as assessed using resources such as the GNF Gene Expression Atlas database, (http://expression.gnf.org/cgi-bin/index.cgi), and thus allow us to examine how disruption of certain regions affects organ development and/or thymocyte differentiation. Another use would be to identify promoters that may be used for site-specific transgene expression in the thymus, for instance, to probe the effects of mis-expression of factors such as chemokines, cytokines, or morphogens in inappropriate thymic regions.

[0218] All thymuses are from male C57BL6 mice at 4.5 weeks of age, corresponding to maximum size (important for discrimination between regions) and to peak activity for importation of new progenitors from blood. At least 50ng of isolated RNA are subjected to 4-6 rounds of high-fidelity amplification. Microdissection are performed as illustrated in Figure 8, using a Leica AS AMD scope. Sorted lymphoid populations are defined as follows: DNl Gin^'CD24^vlo25^"44⁺l 17⁺), DN2 (lin-CD24⁺25⁺44⁺117⁺), DN3 (Hn^' CD24⁺25⁺44'°l 17'°), preDP (lin^loCD24⁺25-44^lol 17), DP (CD4⁺8⁺, including blast and small non-cycling cells), CD4 (CD3⁺4⁺8^"), CD8 (CD3⁺4^"fr^*), and γδ (CD3⁺TCRγδ⁺). Cells are sorted on a three laser BD FACS DiVa cell sorter. In all cases, RNA are extracted using Qiagen RNeasy Mini Kits. In the case of laser microdissection, samples in lysis buffer (Qiagen) are stored individually until the slides can be mounted and the accuracy of dissection confirmed, and then selected sampled conforming to the standards shown in Figure 8 are pooled.

[0219] Stromally expressed genes encoding cell-surface or secreted proteins: The method of choice utilizes a gene ontology browser display within the Affymetrix NetAffx Analysis Center (https://www.affymetrix.com/analysis/netafrx/). Gene lists (probeset IDs) are uploaded to the website, and are assigned to various Gene Ontology nodes based on biological process, cell component, or molecular function categories. Use of the web-based NetAffx analysis center ensures that when identities are assigned to ESTs or annotations change, the analysis is simultaneously updated. Hierarchical organization is used to visualize the results, and chi square tests are used to determine the significance of the association between the list and each individual node, which is useful for validating and prioritizing outcomes.

[0220] Examination of the upstream nodes allows many of these to be immediately discarded, such as catalytic activity, motor activity, etc. However, others are immediately of interest. For instance, signal transducer activity is a second order node (molecular function > signal transducer activity). This node contains 100 of the original 644 results, and one of the next nodes downstream is receptor binding (21 results, all of which are known genes, Table 1).

[0221] For example, and in no way limiting the invention, stromally-expressed genes that encode cell surface or secreted proteins are identified. This includes, refining and prioritizing these gene lists by various methods (presence of receptor/counter-receptor on lymphoid cells, most restricted expression patterns, highest expression levels). The best candidates are validated for follow-up by in situ analysis, as well as first-order functional validation. For example, RNA was prepared from microdissected tissue and subjected to limited high-fidelity amplification, followed by labeling and hybridization to the gene array. Genes expressed in this region (i.e., found to be present, using absolute detection calls in Affymetrix Microarray Suite software) were then compared to those found to be expressed by lymphoid progenitors in the cortex. Genes expressed in the dissected sub-capsular tissue, but not in cortical thymocytes, were identified. This high-stringency approach not only results in the elimination of lymphoid-specific genes from the microdissection results, but also results in the elimination of the bulk of all known genes, since most of them are related to common metabolic processes. Thus, a manageable and highly relevant list of 420 results was generated, including 371 known genes and 49 genes with no known function. Analysis of genes known to be expressed by cortical stroma, including keratin 8 (Krt2-8), delta-like ligand-1 (DLLl,), interleukin-7 (IL7), stromal-derived chemotactic factor-1 (CXCL12), and Indian hedgehog (Ihh) among many others, shows that they are indeed present in this list. Also present are a large number of genes whose products have known roles in signaling proliferation, differentiation, and/or survival in other cell types, but with no known role in the thymus. There are also a number of genes that encode intracellular products (transcription factors, etc) that are likewise novel in the context of the thymus, and these are potentially of great interest in revealing the nature of the stromal cells themselves.

Example 4: Notch-3 in thymocyte differentiation, and use Notch-3 deficient thymocytes and transcriptional targets for Notch signaling.

[0222] Notch signaling plays a fundamental role in post-natal thymocyte differentiation. However, the transcriptional changes that result from Notch signaling and induce T lineage fate are still somewhat unclear.

[0223] In contrast to Notch- 1, Notch-3 expression is very low in DNl precursors, increased in DN2, and peaks at the DN3 stage. The role of Notch-3 in T cell development is less clear than Notch- 1. However, unlike Notch- 1, Notch-3 deficiency is not embryonically lethal, suggesting that these two molecules probably have different functions. Further, they are differentially expressed in thymocyte subsets (Figures 2A-2D); Notch-1 is high in DNl cells, consistent with the phenotypes of various knockouts, while Notch-3 is very low in DNl but is expressed at high levels in DN3. We also find that Deltex, a canonical mediator of Notch signaling, is highly unregulated at the DN3 stage (Figures 2A-2D).

[0224] Hes-1, mutation, which mimics the phenotype of Notch- 1 mutation is expressed in a manner similar to Notch-1 (Figures 2A-2D), i.e., high levels in very early DN 1 cells, and remaining high through the DN3 stage. Thus, Notch-1, Notch-3, Hes-1, and Deltex are all highly expressed in DN3 cells, with Notch-3 and Deltex being nearly specific for DN3. This data, strongly suggests the presence of a second critical period for Notch signaling at the DN3 stage, and implicate Notch-3 as well as Notch-1 in this process. This represents an opportunity to evaluate the role of Notch signaling in the induction of transcriptional changes in developing lymphocytes, since unlike arrest at the DNl stage, an arrest at the DN3 stage would still allow substantial numbers of thymocytes to be generated (there are approximately two million DN2/DN3 cells in an average-sized thymus of 200 million cells).

[0225] Notch-3 deficient mice are available. The thymic phenotype of these mice are examined, in thymuses from homozygous mutant offspring, and in competitive chimeras generated from the bone marrow of these. DN3 or other thymocytes from homozygous mutants or controls are used to determine the definitive downstream targets of Notch signaling, using comparative microarray analysis. Together these experiments will accomplish several things: an analysis of ^"Notch-3 function in thymocytes, a comparison of Notch- 1 and Notch-3 contributions to T cell development, and definition of the downstream targets of Notch-3 signaling, and potentially of Notch signaling in general.

[0226] Analysis of the requirement for Notch-3 in thymocyte differentiation: Unlike Notch- 1, Notch-3 expression is low during the proximal stages of thymocyte differentiation, but peaks at the DN3 stage (Figures 2A-2D). Forced expression of Notch-3 induces an accumulation of DN3 cells and induction of T cell leukemias with an immature phenotype. Thus, Notch-3 may play a unique role at later stages of thymocyte differentiation than Notch- 1. The accumulation of Deltex (Dtxl, a regulator of Notch signaling) at the DN3 stage (Figures 2A-2D), when Notch- 1 expression is waning, is further evidence for a second period of Notch signaling in T cell development, and Notch-3 represents a very likely candidate for this.

[0227] Notch-3 -deficient mice have already been generated and heterozygous parents have been obtained. The first phase are to generate homozygous mutant offspring and analyze their thymuses for size, cellular composition, and proliferative status. This will include analysis of CD44 x CD25 for early cells, CD4 x CD8 for later cells, as well as TCRγδ, NK, NK-T, and dendritic cell (DC) lineage proportions and absolute numbers. The bone marrow from mutant mice will also be used to generate competitive chimeras, by transplantation into sub-lethally irradiated Ly-5 congenic recipients. We have shown that the phenotype of mutant thymocytes in mutant mice may differ quite substantially from that observed in competitive chimeras (Petrie, H. T. et al. (2000). J Immunol 165, 3094-3098; Plotkin, J., Prockop, S. E., Lepique, A., and Petrie, H. T. (2003). J Immunol 171, 4521-4527). Further, these animals are used to generate paired samples (plus/minus Notch-3) for identification of Notch downstream signals, as described below. [0228] Specific methods: Mice are bred and typed for homozygosity using primers for PCR as described (Krebs, L. T. et al. (2003) Genesis 37, 139-143). Thymic phenotypes, cell numbers, and cell cycle analysis are performed on an LSR2 four laser analyzer as previously described (Petrte, H. T, et al. (2000). J Immunol 165, 3094-3098; Plotkin, J., Prockop, S. E., Lepique, A., and Petrie, H. T. (2003). J Immunol 171, 4521-4527; Gordon, K. M., Duckett, L., Daul, B., and Petrie, H. T. (2003). J Immunol Methods 275, 113- 121). Chimeras are prepared and analyzed as previously described (Petrie et al., 2000; Plotkin et al., 2003; Gordon et al., 2003). Wildtype Ly-5.2⁺ donor cells are used as chimeric controls, as well as wildtype recipient cells (Ly-5.1⁺).

[0229] Downstream targets for Notch signaling: Notch- 1 expression peaks very early in intrathymic differentiation, and its deletion results in a very small thymus with no detectable T cells, thus making it difficult to assess the downstream consequences of Notch signaling. In contrast, Notch-3 is unregulated much later, and appears to have a role later in DN and potentially DP development. Thus, even if differentiation is arrested in Notch-3 mutants, arrest at a later stage will make it possible to compare gene expression patterns in thymocytes lacking Notch signals, versus that of their normal counterparts. If there is an arrested phenotype associated with Notch-3 deficiency, cells representing the terminal stage of differentiation in Notch-3 mutant cells are isolated from Notch-3 mutant (Ly-5.2*) / wildtype (LySA⁺) chimeric thymuses. If there is no phenotype, DN2, DN3, and preDP cells are isolated from both components of the chimera, as these bracket the period of highest expression of Notch-3 in normal cells (Figures 2A-2D). RNA are extracted and used as template for cRNA synthesis and hybridization to high-density microarrays. Gene expression patterns in the presence or absence of Notch-3 signaling are compared. A list of genes that are modulated in the presence of Notch-3 signaling are prepared, and ordered according to the highest fold changes.

[0230] Chimeras are constructed, cells purified, RNA isolated, and microarray analysis performed as previously described (Plotkin et al., 2003). Gene expression analysis are performed using GeneSpring software.

[0231] a6β4 integrin in migration of progenitor thymocytes to the subcapsular region of the cortex, and its roles in differentiation, proliferation, and/or survival at the DN/DP transition: The transition from DN to DP (more specifically, DN3 to preDP) is characterized not only by differentiation (i.e., acquisition of a new transcriptional program), but also by cell death in 44% of cells, and by massive proliferation in those that remain. The complexity of this transition clearly obligates the involvement of multiple signaling interactions mediating proliferation, survival, and differentiation.

[0232] The role of a novel integrin, composed of α6 and β4 chains, in differentiation, proliferation, and/or survival at this developmental transition is characterized. Unlike most integrins, which have very short cytoplasmic tails and interact primarily with the actin cytoskeleton, α6β4 integrin has multiple intracellular domains contributed by the β4 chain. Membrane-proximal domains of the β4 tail interact with intermediate filaments to moderate stable adhesion and migration. However, more distal domains containing regulatory tyrosines signal proliferation through the ERK pathway, as well as cell survival through PI3K/Akt.

[0233] Using these two mutant strains, the requirement for α6β4 integrin in T cell differentiation are assessed and to discriminate between cell migration defects and direct versus indirect defects on proliferation or survival.

[0234] α6β4 in thymocyte differentiation are examined using knock-in mutations of β4 integrin. The first (tailless) completely lacks the cytoplasmic domains of β4. α6β4⁺ cells from this mouse can adhere loosely to laminin-5, but do not form stable contacts, exhibit impaired migration, and cannot signal intracellularly. The second is a truncated mutant that contains membrane proximal domains (interacting with keratin intermediate filaments) and can thus facilitate stable adhesion and migration, but cannot signal through tyrosine kinases.

[0235] Defective signaling through a6β4 on thymocyte differentiation: An arrest may be complete at a specific stage (most predictably, around the DN3/preDP transition, or may be partial (reduced proportions of cells after a given stage). For instance, a partial arrest with failure to localize to the sub-capsular zone would substantiate the importance of the sub-capsular stromal microenvironment in normal DN to DP transition; whether this was a proliferative, survival, or adhesion defect.

[0236] Hematopoietic chimeras, generated from fetal liver (tailless mutant) or bone marrow (truncated mutant or wildtype), are used to assess the ability of mutant donor cells to differentiate in a normal thymus. When allowed to revert to steady state (5-6 weeks post-transplant), this system provides normal tissue architecture and spatial contexts, as well as intact stromal environments. Importantly, a normal thymic microenvironment and competition from wildtype cells is essential in revealing otherwise subtle defects such as those of β4 truncation mutants (Figures 4A-4C). Experiments will involve comparison of various TCRαβ developmental stages (DN stages, DP cells, SP cells) and other thymic lineages (γδ cells, dendritic cells, NK cells, NK-T cells) in mutant or wildtype donor thymocytes (Ly- 5.2*), as well as in wildtype recipient cells (Ly-5.1⁺). Both absolute number and relative proportion are calculated, and stages where defects occur are identified.

[0237] Tailless β4 mutant mice are used as donors for all initial experiments, since these are most likely to reveal a phenotype; subsequently, truncation mutants are used to help distinguish the mechanism. Tailless mutant mice die at birth, and consequently, day 16 fetal liver are used as donor cells to generate hematopoietic chimeras. For truncation mutants, adult bone marrow are used. Hematopoietic chimeras are generated and analyzed.

[0238] Adhesion, proliferation, or survival roles for a6β4 signaling in the DN/DP transition: Without wishing to be bound by theory, the DN3/preDP transition is the most likely point at which α6β4 may play a role in thymocyte differentiation. This transition is complex, and requires migration into the sub-capsular microenvironment, death of those cells with sterile TCRβ rearrangements, and survival, differentiation, and proliferative amplification of those cells that pass the TCRβ selection checkpoint. These are all functions associated with expression of α6β4 integrin in other progenitor cells.

[0239] The approach is multifaceted, and involves ruling out certain functions as much as demonstrating others. Essentially, the primary approach are to generate hematopoietic chimeras using donor cells from β4 mutant mice (or wildtype mice) transplanted into congenic recipients, as described infra. The location of mutant (versus wildtype) cells in thymuses from chimeric mice are assessed by staining in situ, and their proliferative status (DNA content) by flow cytometry. In addition, the ability of mutant or wildtype DN3 thymocytes to survive when plated on laminin-5 coated plates are tested, and their ability to proliferate and differentiate in vitro on stromal cells that support T cell differentiation are assessed. Ultimately, biochemical assays for nuclear translocation of NF- kB or phosphorylation of Akt may be used to determine the mechanism of cell survival, depending on the outcomes of proliferation/survival experiments and the differences between tailless and truncated mutant cells.

[0240] Complete arrest at the DN3 stage in tailless, but not truncation, mutants: The differences between tailless mutants (which exhibit impaired adhesion and migration) and truncation mutants (which migrate normally) will aid in determining that impaired migration is the primary requirement for α6β4. Without wishing to be bound by theory, the results should show a decrease in the frequency of tailless mutant DN3 thymocytes (but not truncation mutants or wildtype DN3) in the sub-capsular region of chimeric thymuses. Here, the OP9/DL1 culture system are important: if the generation of preDP and more mature cells can be rescued by culture of arrested mutant thymocytes on OP9/DL1 (i.e., by their ectopic placement into conditions that provide all necessary stromal signals), this, together with the other results obtained will show that the primary role of α6β4 integrin in T cell development is to facilitate migration into the sub-capsular region.

[0241] Partial arrest at the DN3 stage in both truncation and tailless mutants: Without wishing to be bound by theory, since truncation and tailless forms do not differ, defects in signaling rather than migration are primarily responsible for the phenotype (since truncation mutants adhere and migrate normally). This are confirmed by analysis of mutant cells (of either type) in the sub-capsular region of chimeric mice, and, if necessary, by culture on OP9/DL1. The next step are to discriminate between proliferative (ERK) and survival (PI3K/Akt or NF-kB) functions of α6β4 signaling. The first step are to establish or rule out an effect on proliferation. This are accomplished in two ways. The first is by assessing cell cycle status (DNA content) by flow cytometry in mutant cells versus their wildtype counterparts. It is anticipated that a reduction in proliferative activity are reflected by a reduced proportion of cells in S/G₂/M phases of cell cycle. Since this static measurement has some limitations (e.g., longer cell cycle times can reduce proliferative rate without a corresponding reduction in hyperdiploid cells), a second assessment are made by culturing mutant or wildtype cells on OP9/DL1, and measuring cell proliferation directly by counting. Reduced cell growth together with a decrease in cell cycle are strongly supportive of a role for α6β4 in proliferation, although either of these singly, together with ruling out an effect on survival or migration would also be definitive. A final test for a proliferative effect are analysis of nuclear translocation of phospho-ERK in purified mutant or wildtype cells plated on OP9/DL1 [0242] Regardless of whether α6β4 is shown to play a role in progenitor thymocyte proliferation or not, the effects on cell survival must be tested, since these are dependent on different tyrosine phosphorylation motifs of the β4 integrin tail. The appearance of cells with sub-diploid DNA in cell cycle studies may be one indicator of an increased susceptibility to cell death in β4 mutants, although apoptotic cells appear to be very efficiently cleared from the thymus, and thus may fall below the limits of detection for this assay. Several additional tests are applied. One are to culture purified mutant or wildtype progenitors on culture dishes coated with latninin-5, and to measure the rate of cell death by viability dye exclusion. A second assay are to measure the ability of mutant versus wildtype cells to undergo nuclear translocation of NF-kB in response to plating on laminin-5. In particular, the addition of a second signal, such as cross linking of preTCR using anti-CD3ε antibody, or addition of IL-7, may be informative in revealing whether α6β4 integrin cooperates with preTCR or cytokine receptors in mediating survival at the DN/DP transition. Analysis of phosph-Akt will also be performed to ascertain whether this pathway is activated in response to laminin-5 binding (with or without CD3ε/IL7) in wildtype versus mutant progenitors. The combined results of localization studies, proliferation studies, and cell death studies are expected to give clear results regarding the role of α6β4 in T cell differentiation, and are also anticipated provide novel information regarding non-canonical mediators required to support the complex processes that overlap at the DN/DP transition.

[0243] Characterization of the function of a novel transcription factor, LyIl₁ in the control of differentiation and proliferation at proximal stages of intrathymic development: The numerous complex changes that accompany lymphopoiesis in the post-natal thymus are initiated by signals from stromal cells. Ultimately, however, cellular differentiation requires, and is defined by, changes in gene expression. One of the primary mechanisms to control gene expression is the modulation of transcription, as mediated by nuclear transcription factors. Consequently, the identification of transcriptional regulatory networks present in precursor cells versus their progeny may be used to reveal the mechanisms that drive differentiation. A number of transcription factors have already been shown to be required for proper T cell differentiation, proliferation, or survival. Our results show the presence of a transcription factor, lymphoblastic leukemia 1 (LyIl) with the potential to be very important in regulating early T lymphopoiesis, as well as oncogenic transformation of immature T cells. LyIl is a basic helix-loop-helix transcription factor originally identified because it is disregulated in approximately 20% of human acute T cell leukemias. This disregulation is caused by chromosomal translocation, fusing the coding sequence of LyIl to the TCRβ locus. LyIl is thought to be expressed in all hematopoietic lineages except T cells, although the exact functions of LyIl remain unknown. However, we find that LyIl is expressed at very high levels in DNl and DN2 cells, but is completely repressed at the DN3 stage and onward (Figures 5A-5B). Notably, TCRβ gene rearrangement initiates at the DN3 stage (Petrie, H. T. et al. (1995). J Exp Med 182, 121-127). This raises the possibility that errors in the somatic recombination process, resulting from simultaneous activity at the TCRβ and LyIl loci during the DN2/DN2 transition, may result in inappropriate expression of LyIl, and thus potentiate T cell transformation. Comparison of T-ALL with LyIl translocations to those with other cytogenetic abnormalities shows that LyIl malignancies constitute a very immature T lineage cell, with the CD4^~8^'34⁺ phenotype, i.e., early DN cells.

Example 5: LyIl and for early T cell differentiation in the thymus.

[0244] The role of LyIl in hematopoietic and lymphopoietic differentiation is not known. However, LyIl is one of the most highly expressed, differentially regulated genes found in early DN thymocytes (Figures 5A-5B). Other genes that are regulated similarly to LyIl, and thus emerged from microarray analysis using the same criteria that revealed LyIl, include Bell Ia and PU.l. The function of LyIl are characterized by generating knockout mice. A genomic BAC clone containing the full-length LyIl mouse gene in which the gene has been subcloned into conventional cloning vectors has been constructed. The gene is approximately 4 Kb consisting of four exons. The entire gene are targeted for deletion using standard technology (i.e., generating a neomycin construct that includes flanking regions homologous to those of LyIl, and generating knockout ES cells by homologous recombination). Once homozygous knockout mice are generated, hematopoietic organs are analyzed (note: there is a possibility that deletion of LyIl may cause embryonic lethality due to gross hematopoietic deficiency. Any of the hematopoietic lineages may be affected, and if non-T defects are found they are characterized. Analysis will include thymus, lymph node, and spleen cell size and T cell number, as well as characterization of thymus-derived lineages (CD4, CDS, TCRγδ, NK-T, dendritic cell, etc.), early thymocyte differentiation (CD25/44/117), and proliferation (by DNA staining and flow cytometry).

[0245] A role for LyIl in proliferative expansion of early T cells are addressed by analysis of DNA content and cell cycle distributions. An effect on cell survival/accelerated cell death may be characterized by analysis of DNA strand breaks by flow cytometry, using standard TUNEL assays. A more forma] role in differentiation may be analyzed by coculture of arrested LyI 1 -deficient thymocytes on stromal cells that support T cell differentiation.

[0246] The 'recombineering' method of Copeland et al. Nat Rev Genet 2, 769- 779, (incorporated by reference) are used to generate germline knockout constructs. Although this method has been described as a method to accelerate conditional knockout mice, it can also be used to clone homology arms for germline knockouts. Chimeric founders are bred and knockout strains established by the applicant's laboratory.

[0247] The effects of enforced expression of LyIl in immature thymocytes: Expression of LyIl later in T cell development is associated with abnormal T cell growth and malignant transformation. Given the reduction in proliferation that occurs in conjunction with repression of LyIl at the DN2/DN3 transition (see above), the question of whether enforced LyIl expression will change the proliferative status of early thymocytes, or their ability to differentiate, are investigated.

[0248] The LyIl gene are cloned into the plO17 vector ₅ which utilizes the lck proximal promoter to drive transgene expression in immature T cells (Wildin, R. S. et al. (1991). J Exp Med 173, 383-393). Specifically, the lck promoter becomes activated at the DNl stage, and operates at high levels until just before export of mature cells to the periphery. Thus, the high level of LyIl expression found in DNl and DN2 thymocytes (Figures 5A-5B) can be maintained throughout intrathymic development. Two basic types of experiments are performed. The first are to analyze developmental progression through DN, DP and Sp phases, as well as the differentiation of various thymus-derived lineages, as described. The second are to analyze cell cycle status (DNA content), also by flow cytometry, to determine the effects of LyIl expression on control of thymocyte proliferation.

[0249] Novel lymphostromal signaling interactions: At the DN3 to preDP transition, the TCRβ locus is silenced and the TCRα locus made accessible (Petrie, H. T et al. (1995). J Exp Med \%2, 121-127), while cells that fail to assemble a functional TCRβ gene are deleted. The cells that remain are induced to undergo massive, rapid proliferation leading to the generation of a large pool of DP cells for further selection. Very few genes are dramatically changed when DN3 cells are compared to preDP progeny (Figure 1). This can mean that surface receptors required to support differentiation, proliferation, and/or survival at this transition may be expressed throughout DN development, rather than being differentially regulated in response to exposure to the sub-capsular microenvironment, while differential signaling occurs because cells bearing the receptor must migrate to the sub-capsular zone in order to encounter ligand.

[0250] Stromally-derived signals that are unique to the sub-capsular microenvironment: The sub-capsular zone is a unique region where DN cells undergo differentiation to the DP stage, and where the bulk of progenitor expansion occurs in the thymus. The DN/DP transition occurs in a very defined location corresponding to the immediate sub-capsular zone. Preliminary data (not shown) suggests that direct contact with the single layer of classical type 1 epithelium that forms the inner lining of capsule may, in fact, be required. Alternatively, it is possible that they only need enter the immediate subcapsular region, defined by a dense layer of proliferating cells, to undergo this transition.

[0251] Tissue corresponding to the sub-capsular region (the fibroblast and epithelial layers of the capsule, plus the adjacent region, equal to approximately ten layers of lymphoid cells) are isolated by laser catapulting. RNA are isolated, and high-fidelity linear amplification are performed as described (Iscove, N. N et al. (2002). Nat Biotechnol 20, 940- 943) to make sufficient RNA for microarray analysis. From this, a list of all genes expressed are prepared. The genes in this list will derive from three primary sources: stromal cells that form the capsule (fibroblasts and type 1 simple epithelium), stromal cells that permeate the sub-capsular region (reticular epithelium), and lymphoid cells (DN3 and preDP). Signals produced uniquely by either of the former are of primary interest. To identify these, genes expressed by DN3 or preDP cells (microarray analysis already performed) are subtracted from the total genes found. The remaining genes will represent those expressed exclusively by capsular/sub-capsular stromal cells.

[0252] The list of stromal-specific genes will then be categorized based on Gene Ontology (GO) designations to define molecules that may be secreted, or that are expressed on the cell surface, and that may consequently be used to signal to lymphocytes. For example, GO Molecular Function category of 'receptor binding' (GO:0005102), which has 414 genes under 55 additional categories (terms) that include chemoattractant activity (GO:0042056, with 34 genes including many chemokines), cytokine activity (GO:0005125, with 192 genes in further categories including growth factor receptor, interleukin receptor, and transforming growth factor receptor binding activities among others), hormone activity (GO:0005179, with 102 genes in multiple sub-categories), integrin binding (GO:0005178, with 10 genes), etc. A manual search of genes will also be used.

[0253] The list of factors secreted by or expressed on the surface of stroma! cells from the sub-capsular region can include autocrine signals or signals for other stromal cells, as well as signals specific for developing lymphocytes. The list of genes expressed by stromal cells in the sub-capsular region and encoding secreted or cell-surface proteins will then be compared to surface proteins expressed on DN3 and preDP cells. This list will then be examined and candidates for further evaluation are prioritized.

[0254] Stromal signals that are restricted to the capsular/ sub-capsular zone: Two general approaches are used. The first are to perform RT-PCR analysis on RNA from sub-capsular versus mid-cortical tissue, as previously described (Plotkin, J., Prockop, S. E., Lepique, A., and Petrie, H. T. (2003). J Immunol 171, 4521-4527). This will eliminate those signals that are not restricted to the sub-capsular region. For those that are specific to the subcapsular region, RNA in situ hybridization are performed to confirm sub-capsular expression, and to define the specific cellular sources. Cellular morphology and histologic location are generally sufficient to define various types of stromal cells, but if necessary, two-color RNA in situ I antibody immunofluorescent staining can be performed to identify epithelial (anti- cytokeratin⁺) or fibroblastoid (ER-TR7*) cells.

[0255] Functionally characterize the effects of signals specific to the subcapsular region. Preliminary testing for effects on the DN/DP transition are screened in vitro, using a stromal cell system that supports T cell differentiation. This will include transduction with vectors encoding anti-sense, dominant-negative, or siRNA constructs. For those genes that give a phenotype in vitro (inhibition of differentiation, inhibition of proliferation, increased cell death), knockout mice are constructed, and definitive analysis of function are assessed using competitive in vitro chimeras as previously described (Plotkin et al, 2003).

[0256] Tissue preparation, laser microdissection, and RNA extraction are performed as previously described (Plotkin et al, 2003). Amplification are performed using methods for linear amplification from small samples (Iscove et al., 2002); which we have successfully used to amplify 5 μg of RNA from only 10 ng starting material, which was obtained by microdissection. Microarray analysis will also be performed. Microarray results are annotated using Affymetrix and Gene Ontology Consortium databases. Annotated results are filtered for genes present, and on GO functional categories, using GeneSpring software. Subtraction of genes present in RKA from DN3 or preDP cells from those present in capsular/sub-capsular tissue will also be performed using GeneSpring. RT-PCR and RNA in situ hybridization (with or without antibody staining) are performed.

Example 6: Identification of genes expressed by stromal cells from the sub-capsular cortex.

[0257] Genes expressed by stromal cells in distinct regions of the thymus (are identified by performing microarray analysis on RNA from microdissected tissue, followed by subtraction of genes expressed by lymphocytes in those regions (also determined by microarray).

[0258] RNA was prepared from microdissected tissue and subjected to limited high-fidelity amplification, followed by labeling and hybridization to the gene array. Genes expressed in this region were then compared to those found to be expressed by lymphoid progenitors in the cortex. Genes expressed in the dissected sub-capsular tissue, but not in cortical thymocytes, were identified. This high-stringency approach not only results in the elimination of lymphoid-specific genes from the microdissection results, but also results in the elimination of the bulk of all known genes, since most of them are related to common metabolic processes. Thus, a manageable and highly relevant list of 420 results was generated, including 371 known genes and 49 genes with no known function. Analysis of genes known to be expressed by cortical stroma, including keratin 8 (Krt2-8), delta-like ligand-1 (DLLl), interleukin-7 (IL-7), stromal-derived chemotactic factor- 1 (CXCLl 2), and Indian hedgehog (Ihh) among many others. Also present are a large number of genes whose products have known roles in signaling proliferation, differentiation, and/or survival in other cell types, but with no known role in the thymus. There are also a number of genes that encode intracellular products (transcription factors, etc) that are likewise novel in the context of the thymus, and these are potentially of great interest in revealing the nature of the stromal cells themselves.

Differential expression of Notch ligands Delta-like ligand-1 (DLLl) versus DLL4 in the thymus,

[0259] One of the basic principles underlying this application is that there are differences in the genes expressed by stromal cells from diverse thymic microenvironments. Using microdissection and RT-PCR, we have now found that this is the case for certain ligands for Notch- 1₁ an indispensable signal for T cell differentiation.

[0260] To test any differential requirements for alternate Notch ligands in the thymus, tissue from various regions of the thymus were microdissected, the corresponding RNAs were isolated, and subjected to RT-PCR. DLLl was found to be expressed throughout the cortex (Figure 7) as well as in the medulla. In contrast, DLL4 was restricted to the subcapsular cortex. These data confirm that stromal cells from different microenvironments do differ in the genes they express, and further, that they differentially express genes involved in signaling to developing lymphocytes.

Identification of the genetic differences between stromal cells from various microenvironmental compartments of the post-natal thymus:

[0261] RNA is obtained from isolated tissue regions (prepared by microdissection) and from the corresponding lymphoid constituents (prepared by cell sorting). The stratified distribution of thymocyte developmental stages in discontinuous tissue regions indicates that different regions each deliver relatively distinct sets of signals to developing T cells. Tissues are microdissected from six defined regions of the thymus, as shown in Figures 8A-8C. These regions were selected because they each represent a unique signaling environment, as defined by the functions of individual lymphoid progenitor species within them. As shown in Figures 8A-8C, only tissue regions displaying relatively concentric cortical/medullary organization and broad tissue depth are used, in order to minimize cross- contamination between regions. Strips of tissue 40 μm in width (approximately 6-8 cell diameters in width) are dissected until approximately 1 mm² of tissue has been collected (about 50 strips 500 μm long); preliminary studies show that this will yield approximately 50 μg of RNA, thus requiring minimal amplification to prepare sufficient template for gene chip analysis. Samples dissected from these regions are stored in individual microfuge tubes until the post-dissection tissue is mounted and examined, and only those samples that are appropriately located are utilized.

[0262] The dissected strips of tissue will contain both lymphoid and non- lymphoid cells. In order to distinguish genes expressed by the stromal components, those genes expressed by the lymphoid constituents of these regions are identified and filtered out. To filter out the lymphoid genes, all major conventional T lymphoid stages are purified and screened. These include DNl, DN2, DN3, preDP, DP, CD4SP, and CD8SP for the TCRαβ lineage, as well as CD3⁺TCRγδ⁺ for this alternate thymic lineage. These populations are identified and purified by cell sorting.

High-fidelity amplification, labeling, and hybridization to gene chip microarrays:

[0263] Microdissection of tissue regions 25μm wide, 200-500 μm long, and 10 μm deep can reliably yield 50ng of tissue. High-fidelity amplification can be used to generate the additional cRNA needed for microarray. Briefly, microdissected RNA are used for cDNA synthesis using using oligo-dT(T7) primers (Affymetrix) and MessageAmp RNA kits (Ambion). This cDNA are used for reverse transcription (MessageAmp), and the process of reverse transcription//« vitro transcription are repeated 4-6 times until sufficient linearly amplified cRNA is obtained. During the last in vitro transcription cycle, cRNA are labeled by the addition of biotinylated nucleotides, and 1.5-2.0μg are hybridized to MOE430 2.0 arrays.

[0264] Identification of genes expressed specifically by stromal cells: RNA from microdissected tissue will include lymphoid as well as stromal cells. The data generated by scanning microarrays probed with cRNA from microdissected or purified lymphoid tissues are imported into GREX software (Affymetrix). The PLIER algorithm are used to generate relative RNA signal values. PLIER parameters are those established for similarity to robust multichip analysis, namely quantile normalization, use of perfect matched oligonucleotides only, percentile background, and quick signal optimization. The Microarray Suite 5.0 statistical algorithm, including both matched and mismatched oligonucleotides, are used to calculate the probability that differences between each pair is not due to chance. The calculated probabilities for each probe set (gene) for the replicate gene chips of each microdissected region are averaged. An absolute detection call (present, marginal, or absent) for each gene are determined based on this pooled probability, using default levels: p<0.04, 0.04-0.06, and >0.06, respectively. A list of all genes designated as present in the each microdissected region are prepared, while genes found to be marginal or absent are filtered out. Expression of the genes in this list will then be analyzed in each lymphoid microarray, to establish a list of stromal-specific genes.

[0265] This is accomplished by identifying those genes present in microdissected regions, but not expressed in any of the lymphoid constituents of those regions. Such analysis of the sub-capsular cortex yields 420 genes expressed in this region that are not found in cortical thymocytes. Each list of stromal-specific genes can then be further filtered by sorting the results based on highest expression (signal) levels, or by those that differ most substantially from the mean of all genes expressed on the chip. This process are reiterated for each microdissected region, and for each lymphoid subtype, until a full accounting of non-lymphoid (i.e., stromal) gene expression is mapped for the entire thymus.

[0266] Characterize and make distinctions between the genetic profiles of stromal cells in each region: The stromal gene list for each region will first be compared to the lists for other regions, based on absolute detection calls (present, absent, marginal), and a refined list of those genes exclusive to each region are extracted. Individual analysis of functionally-defined regions (Figures 8A-8C) can give detailed information corresponding to specific developmental events. This list will reveal genes that may be targeted for deletion in stromal cells as assessed using resources such as the GNF Gene Expression Atlas database, (http://expression.gnf.org/cgi-bin/index.cgi), and thus allow us to examine how disruption of certain regions affects organ development and/or thymocyte differentiation. Another use would be to identify promoters that may be used for site-specific transgene expression in the thymus, for instance, to probe the effects of mis-expression of factors such as chemokines, cytokines, or morphogens in inappropriate thymic regions.

[0267] All thymuses are from male C57BL6 mice at 4.5 weeks of age, corresponding to maximum size (important for discrimination between regions) and to peak activity for importation of new progenitors from blood. At least 50ng of isolated RNA are subjected to 4-6 rounds of high-fidelity amplification. Microdissection are performed as illustrated in Figures 8A-8C, using a Leica AS AMD scope. Sorted lymphoid populations are defined as follows: DNl (lin-CD24-^/lo25-44⁺l 17⁺), DN2 (lin CD24⁺25⁺44⁺l H⁺), DN3 (tin^" CD24⁺25⁺44^lol 17^l0), preDP (Hn^loCD24⁺25-44^lol 17^{^}), DP (CD4⁺8⁺, including blast and small non-cycling cells), CD4 (CD3⁺4⁺8^'), CD8 [Cm⁺^i⁺), and γδ (CD3⁺TCRγδ⁺). Cells are sorted on a three laser BD FACS DiVa cell sorter. In all cases, RNA are extracted using Qiagen RNeasy Mini Kits. In the case of laser microdissection, samples in lysis buffer (Qiagen) are stored individually until the slides can be mounted and the accuracy of dissection confirmed, and then selected sampled conforming to the standards shown in Figures 8A-8C are pooled.

{0268] Stromally expressed genes that encode cell-surface or secreted proteins: The method of choice utilizes a gene ontology browser display within the Affymetrix NetAffx Analysis Center (https://www .afiymetrix.com/analysis/netaffx/). Gene lists (probeset IDs) are uploaded to the website, and are assigned to various Gene Ontology nodes based on biological process, cell component, or molecular function categories. Use of the web-based NetAffx analysis center ensures that when identities are assigned to ESTs or annotations change, the analysis is simultaneously updated. Hierarchical organization is used to visualize the results, and chi square tests are used to determine the significance of the association between the list and each individual node, which is useful for validating and prioritizing outcomes.

[0269J Examination of the upstream nodes allows many of these to be immediately discarded, such as catalytic activity, motor activity, etc. However, others are immediately of interest. For instance, signal transducer activity is a second order node (molecular function > signal transducer activity). This node contains 100 of the original 644 results, and one of the next nodes downstream is receptor binding (21 results, all of which are known genes, Table 2).

[0270] This node then splits into three (Figure 10), including cytokine activity (15 results), G protein-coupled receptor binding (6 results), and growth factor activity (12 results: note that some results can be assigned to more than one category, since some cytokines may be growth factors, etc). Among these results are CXCL 12, CCL25, and IL7, all of which are known to have roles in thymocyte differentiation, as well as a number of other well known factors not previously known to have a role in the thymus (including Fgfl, Fgfl4, IL4, IL6, and IL20).

[0271] Refine and prioritize the lists of stromal cell surface/secreted products: Two primary methods will allow 1) identification of those cell surface/secreted stromal gene products that display the most restricted patterns of expression, and 2) identification of those cell surface/secreted stromal gene products whose receptor/counter-receptor can be identified on lymphoid cells. Using the method for the first approach the gene lists, e.g. surface/secreted products are compared to the lists of stromal genes found to be restricted to single microenvironments, and genes common to both (surface/secreted and restricted to a single region) are identified.

[0272] The method for the second type of analysis will import array results for purified lymphoid cells into NetAffx, followed by organization based on functional, cell component, or process-based annotations, and examination of nodes likely to contain receptors/counter-receptors (receptor; integral to plasma membrane; signal transduction component; etc). The second and complementary approach are to manually analyze receptor expression for nodes with obvious relevance, such as was done with the "receptor binding" node in the example provided in the previous section. For example, when the contents of the "receptor binding" node were compared to receptors expressed on lymphoid cells, receptors for CXCLl 2 (CXCR4), CCL25 (CCR9), and IL7 (IL7Rα, IL2Ry) were found, confirming known requirements for these signals in the thymus. Of the other Iigands in this node, receptors for CXCLl 1 (CXCR3), ILl receptor antagonist (ILlRl, IL1R2), and IL4 (IL4Rα, IL2Rγ) were also found on specific lymphoid progenitor subsets, confirming these as potential high priority candidates for further study.

[0273] In an interesting contrast to the above examples of potential lymphostromal signals, while stromal cells in the sub-capsular region were found to express Fgfl and Fgfl4 (Table 3), no receptors were found on lymphoid cells. However, FgfR4 (one of the receptors for Fgfl) was found in RNA from the microdissected tissue itself, suggesting that Fgfl production by stromal cells in the sub-capsular region is likely to signal to other stromal cells. Thus, the methods used will reveal novel signals that stromal cells deliver to developing lymphocytes, but will also reveal the nature of stromalrstromal interactions and the novel signaling requirements thereof.

[0274] Validation of highest priority candidates: RNA in situ hybridization are used to confirm that the Iigands are expressed in the appropriate histologic regions and on which types of stromal cells. RNA in situ staining is preferred to immunohistochemistry. Following this confirmation, experiments such as overexpression or interference are carried out. The first approach are electroporation using the AMAXA system for primary mouse T cells, with either RNAi, antisense oligos, or expression vectors, followed by culture on stromal cells that support T lineage differentiation. If transgenics or knockouts are available, they are analyzed both directly, and in competitive hematopoietic chimeras.

Example 7: Differential Gene Expression Mapping characterization.

[0275] The present method of the invention is useful in any tissue. In an exemplary embodiment, the method of the present invention was used to identify gene expression in stromal cells of the thymus.

[0276] There is a steady state production of T-lymphocytes in the thymus. Uncommitted, blood-borne, marrow-derived progenitors undergo homing / extravasation into the thymus; where they undergo lymphopoiesis, which includes lineage specification and proliferation. The cells are then selected by a successful TCR / MHC interation; endowed with full functional capacity; and selected for export or, if self-reactive or otherwise defective, for disposal. As a result, the lymphoid components of the thymus are transient, and mostly non-functional.

[0277] The transient and nonfunctional nature of the lymphoid component contrast with the crucial functional components of the thymus; the stable "stromal" elements, that establish conditions to induce and/or support steady-state lymphopoiesis. However, little was known about stromal cells in the thymus and how they interact with the lymphoid cells. Stromal cells are very difficult to isolate and the process of isolation affects gene expression. Previous attempts have encountered problems that include that enzymatic digestions are selective; lengthy incubations at 37oC affect gene expression; removal from 3D context affects gene expression; yields are low, repeatability is poor; and there is a lack of markers for sub-regions of cortex or medulla. In addition, the majority of stromal signals are juxtacrine or paracrine and therefore the 3D environment is very important.

[0278] For these reasons, it is most advantageous to examine stromal cells in the tissue. Tissue sections can be isolated by microdissection which requires minimal amounts of tissue; allows sub-region specificity; and, because the tissue is frozen within 3-4 minutes, processing artifacts are minimal. This will obtain stromal cells, but the tissue will also contain lymphoid cells.

[0279] The presently inventive method, Differential Gene Expression Mapping (DGEM) utilizes microdissection, microarray screening, and computational profiling to identify gene expression by stromal cells in situ. As a result, DGEM, a reverse identification approach, solves the previously insurmountable problems, as the lymphoid progenitors can be readily isolated, allowing us to use fluctuations in receptor expression on lymphoid cells(by microarray) to predict stratified stromal signals.

[0280] As a result, the inventive method of the invention can be used identify the signals that stromal cells provide to developing lymphocytes, other than Notch ligands, kit ligand, 1L7, and MHC. The majority of signals are juxtacrine or paracrine. By defining the immediate stromal microenvironment for each developmental stage, new signals can be identified. The present method can also identify where in the thymus does each progenitor stage resides; and the stromal signals, or combinations of signals, that define each region of the thymus. Given that crucial role of the thymus in T-cell development, such knowledge is not of mere academic interest, but is relevant to immune system development, autoimmune diseases, immunodeficiency and more. Accordingly, identification of the relevant markers can be used to prevent or treat immune diseases. Moreover, knowledge of the aspects of stromal cell function and survival can be used to determine if a given compound or condition affects stromal cell function. For example, if a given compound adversely affects stromal cell function, it may cause long-term T-cell dysfunction, such as autoimmune disease.

Example 8: Differentia] Gene Expression Mapping of whole cortical and medullary regions of mouse thymus.

[0281] The present method was applied to whole cortical and medullary regions of mouse thymus. From microdissected regions, lymphoid cells were isolated/ purified, using known markers for lymphoid cells at each location: cortical: CD3-/lo and CD45+ and [CD90+ or CDl 17+]; medullary: CD3+ and CD45+. Microarray analysis was performed on lymphocytes and tissue, and, by Boolean subtraction, identified stromal genes from the medulla and cortex. The process was repeated with an average of 7 gene chips. Results were then filtered through a hierarchical filtering process as follows:

Cortex

1. The expression of the gene must significantly greater in whole cortex than in cortical thymocytes, at p<0.05.

2. The expression of the gene in the cortical thymocyte must be less than 3x the median gene expression in the whole cortex.

3. The gene must be expressed in the whole cortex at a level greater than NotchL, a known marker for stromal cells.

[0282] The results of this three stage filtering are presented in Table 4. Column 1 lists the title of the gene by name, as used by Affymetrix. Column 2 presents the gene symbol. Column 3 lists probability that the gene is expressed more in whole cortex than in cortical thymocytes. As can be seen in column 3, levels of statistical significance range over more than 8 orders of magnitude. Column 4 lists the probability that the gene is expressed more in whole medulla than in medullary thymocytes, which is useful for comparison with the cortical expression data. Column 5 lists the base 2 log of the average of expression in whole cortex over that in cortical thymocytes, and therefore provides a measure of how much greater the expression is in whole cortex over cortical thymocytes. Column 6 lists the base 2 log of the average of expression in whole medulla over that in medullary thymocytes.

[0283] In a further analytical step, and additional layer of filtering was applied, selecting only for those genes whose expression in the whole cortex was greater than the median expression level. This additional filtering step removed genes that were known to be expressed specifically in stromal cells. This additional filtering step is therefore overly stringent and the results are not shown.

Medulla

[0284] The data from medulla was also filtered by a three step process, as with the cortical data:

1. The expression of the gene must significantly greater in whole medulla than in medullary thymocytes, at p<0.05.

2. The expression of the gene in the medullary thymocyte must be less than 3x the median gene expression in the whole medulla.

3. The gene must be expressed in the whole medulla at a level greater than NotchL, a known marker for stromal cells.

[0285] Again, as with the cortex, further filtering to select only for those genes whose expression in the whole medulla was greater than the median expression level was found to be overly stringent, removing genes known to be expressed specifically.

[0286] The results of three stage filtering of medulla are presented in Table 5. Column 1 lists the title of the gene by name, as used by Affymetrix. Column 2 presents the gene symbol. Column 3 lists probability that the gene is expressed more in whole cortex than in cortical thymocytes. Column 4 lists the probability that the gene is expressed more in whole medulla than in medullary thymocytes. As can be seen in column 4, levels of statistical significance exceed lOE-10, Column 5 lists the base 2 log of the average of expression in whole cortex over that in cortical thymocytes, and therefore provides a measure of how much greater the expression is in whole cortex over cortical thymocytes. Column 6 lists the base 2 log of the average of expression in whole medulla over that in medullary thymocytes.

[0287] To validate the DGEM process as applied to the thymus, several additional experiments and/or analytical steps were conducted. Hierarchical clustering groups gene expression profiles according to cell type

[0288] The data presented in, for example, Tables 4 and 5 above is the result of averages over several samples. To confirm the robustness of the DGEM procedure, the gene expression profiles from each sample of whole medulla, whole cortex, cortical thymocytes and medullary thymocytes was sorted (without identifying the source of the data) based only on the expression profile. Figure 14 presents the hierarchical clustering of this data. As can be observed, the data sorted according to the source of the cells. Therefore, the variation in expression profile due to different samples is less than the variation between different types of cells from the same tissue sample. As a result, it is possible to apply the DGEM process to a single sample of tissue.

[0289] Highly expressed genes are usually more readily identified as gene markers by older technology, and can therefore serve as a validating controls for the DGEM method.

[0290] Table 6 lists the thirty most highly expressed cortical stromal genes. Of these, 11 have been previously identified in cortical stromal cells: RIKEN cDNA 1700021K02 gene; protease, serine, 16 (thymus); keratin 8; chemokihe (C-X-C motif) ligand 12; desmoplakin; CD83 antigen; histocompatibility 2, class II antigen E beta; vascular cell adhesion molecule 1 ; histocompatibility class I] antigen A/E alpha; decorin; and procollagen, type III, alpha 1.

[0291] Table 7 lists the thirty most highly expressed medullary stromal genes. Of these, 16 have been previously identified in cortical stromal cells: chemokine (C-C motif) ligand 21; histocompatibility 2, class Il antigen E beta; histocompatibility class II antigen A/E alpha; histocompatibility class II antigen A beta 1; keratin 8; amyloid beta (A4) precursor protein; Fc fragment of IgG binding protein; complement component 3; hemoglobin, beta adult major chain; macrophage expressed gene 1; casein beta; desmoplakin; periostin, osteoblast specific factor; desmoglein 2; regenerating islet-derived 3 gamma; complement component 1, q subcomponent, alpha polypeptide.

[0292] Seven (7) additional genes known to be expressed in thymic stromal cells were not found in the top 30 results. Table 8 demonstrates that these genes were identified by the DGEM method. Therefore, genes previously shown to be expressed in stromal cells by previous technologies were also successfully identified by the DGEM method, therefore validating the DGEM method by yet another means. Moreover, the DGEM method demonstrates superiority over previous methods in the art: 19 of the 30 (63%) most highly expressed cortical stromal cell genes have not previously been identified as being expressed in cortical stromal cells. Likewise and 14 of the 30 (46%) most highly expressed medullary stromal cell genes have not previously been identified as being expressed in cortical stromal cells. Overall, more than 80% of genes identified by DGEM have never before been identified as being expressed in stromal cells of the thymus. That DGEM was able to identify the expression of genes not previously identified as being expressed in the tissue, including a large number of highly expressed genes, demonstrates the surprising superiority of DGEM over other techniques in the art. Moreover, DGEM was able to provide a far more comprehensive overview of expression patterns in the stroma, enabling the identification of pathways being activated in the thymus.

[0293] The newly identified expression of genes in the thymic stromal cells may be confirmed by well known techniques, and we have further examined the expression of CD40, CD45, panKeratin, and CD80 in the thymus, confirming the results obtained by DGEM.

Example 9: DGEM identifies regulatory pathways activated in stroma) cells.

[0294] Identification of a expression of a gene or genes known to be part of a regulatory network may be used to identify the activation of such a regulatory network in cells. Because DGEM provides a comprehensive list of gene expression in a cell population, it is useful for identifying further extending the understanding of regulatory pathways in a specific cell type. Thus, one could not only map expression patterns onto previously known regulatory networks, but demonstrate how regulation differs in a specific cell type, or identify new regulatory networks.

[0295] Such knowledge is useful in treating a disease, such as cancer, immune disorders, infections and more. A clear understanding of the regulatory networks enables a person of skill in the art to identify compounds, conditions, or therapeutic regimes that can specifically target a given cell type (such as a cancer cell) with minimal effects on other cell types. A reduction in side effects means that the therapy is more effective, with a reduction in cost, morbidity, and other undesireable side effects.

Example 10: Antigen Processing Pathway in the Cortex. [0296] Antigen presentation is an important stromal cell function, especially in the presentation of self antigens and the resultant identifying and inactivating or down- regulating of T-cells that react to such self-antigens. Antigen processing has been studied in much greater detail, however, in other cell types, such as dendritic cells, macrophages and other specialized cells of the immune system. Of 77 genes known to be used in antigen presentation; DGEM identified 30 activated in cortical stromal cells. In the MHCl pathway: INFα; BiP; MHCl; β2m. TAP 1/2; and KIR. In the MHCII pathway:GILT; AEP; CTSB; Ii; MHCII; SLIP; CTSB/L/S; CLIP; HLA-DM. And, in the nucleus, CIITA. Figure 15 presents major features of the antigen presentation pathway.

{0297] These results validate DGEM, but also indicate difference in antigen presentation between stromal cells and other antigen presenting cells.

Example 11: Neurodegenerative Disease

[0298] A surprising result of the DGEM assay was that many genes known to be important in neurodegenerative disease are expressed in cortical stromal cells. Figure 16 presents major neurodegenerative disease pathways, of which 12 out of 33 genes were found to be activated.

[0299] Those activated include genes associated with Parkinson's disease (SNCA; PARK2; PARK3; PARK4; PARK 7; PARKlO; PARKl ljUCHLl; PINKl; LRRK2; and NR4A2); genes associated with Amyotrophic lateral sclerosis (ALS2; ALS4; NEFH; SODl); Huntington's disease (HD); Dentatorubropallidoluysian atrophy (DRPLA); Prion diseases (PrP); Alzheimer's (APP; APOE; PSEN) and other genes associated with more than one disease (MAPT; NGFR; APLPl; HSPA5).

[0300] These data suggest that compounds, agents or conditions which affect thymic function may also affect neurodegenerative diseases.

Example 12: Leukocyte trans-endothelial migration.

[0301] Outside the thymus, leukocyte transendothelial migration is important in immune response, causing circulating leukocytes to translocate across the endothelium into sites of infection, cancer, and the like. In the thymus, leukocyte transendothelial migration would be expected to be important in stimulating the migration of leukocyte progenitor cells from the bloodstream into the thymus, and subsequent development into mature T-cells. 50/115 genes associated with leukocyte transendothelial migration were expressed by medullary stromal cells.

[0302] Figure 17 presents major elements of leukocyte transendothelial migration pathway known from other cell types. Of those shown in Figure 23 the following are activated in medullary stromal cells: ROCK; Vav; PI3K; Gi; CXCR4; SDF-I; ITGBl; VCAMl; ICAMl; PLCγ; αActinin; Vinculin; RhoGAP; FAL; pl30^Cas; PKC; Nox; p67phox; p47phox; p40phox; p38; pl20ctn; CDH5; CAMs α-catenin; β-catenin; JAM2; ITGBl; and JAMl.

[0303] These results both validate the DGEM method and provide new information about functions of the thymus.

Example 13: TGFβ pathway.

[0304] 32/81 genes in the TGFβ pathway were also found to be activated in stomal cells in the medulla. These include: Chordin; BMP; BMPRI; BMPRII; Smadl/5/8; Id; Smad6/7; Rbxl; Cull; Decorin; THBSl; LTPPl ; TGF β; Lefty; FST; Activin; ActiviπRI; ActivinRlI; NodalRII; PP2A; ROCKl; and p 15.

Example 14: Identification of the AIKE regulatory pathway in the thymus.

[0305] The Autoimmune Regulator (AIRE) has been previously identified in medullary stromal cells, and AIRE activation was also observed by DGEM. (Table 8) Although AIRE knockout mice shows relatively normal thymic histology, AIRE is believe to be important in tolerance induction. AIRE is a bHLH transcription factor thought to induce tolerance to non-thymic tissue antigens, via expression of these genes in medullary stromal cells. Therefore, stromal-specific genes in the medulla would be expected to include targets of AIRE, as well as stromal-intrinsic genes. We used DGEM to identify intrinsic genes (expressed in wildtype AND AIRE mutant mice), as well as targets of AIRE (expressed in wildtype but not AIRE mutant mice).

Hierarchical clustering validated the DGEM protocol ( Figure 19)

[0306] Using DGEM, we identified 1564 stromal-specific genes common to wildtype and AIRE-/- (intrinsic) and 358 stromal-specific genes present in wildtype but not AIRE-/-, and therefore targets of the AIRE regulator in the stromal cells. To determine if these AIRE regulated genes are expressed in other tissues, expression analysis was then also performed on other tissues, confirming that most AIRE targets identified by DGEM are restricted to one or a few tissues.

[0307] Simple Boolean analysis of DGEM data (present in tissue, absent in lymphoid) yields validated lists of stromal genes, and has allowed us to refine a target list for AIRE. Computational subtraction was then used to remove lymphoid expression to identify a list of stromal specific, AIRE regulated genes. The 30 most abundant AIRE regulated genes, found in the stromal cells of wild type but not AIRE mutants, is listed in Table 9.

[0308] Next, results obtained by DGEM were compared with those obtained by others using art known methods. Putative AIRE target genes have been previously studied using conventional methods of stromal cell isolation, by Derbinski et al, J Exp Med202:33 (2005) and Johnnidis e/ o/. Proc Natl Acad Sci 102:7233 (2005). Data from these studies is available from publications and from microarray data repositories.

[0309] Figures 20 and 21 show that the list of target genes identified by DGEM have more relative tissue specificity than the list identified in the prior art, demonstrating the superiority of DGEM over previous methods in the art.

[0310] The expression results were also confirmed by analysis of the promoter region in AIRE regulated genes. Compared to a selection of random genes, AIRE regulated genes significantly more likely to contain a consensus AIRE promoter. This was true both for the lists found in the prior art and those generated by DGEM. However, as seen in Figure 21, the presence of AIRE consensus binding sites was more significantly associated with the selection of genes identified by AIRE than those identified by other methods previously used in the art. As the number of randomly-selected control genes increased, significance rapidly decreased, such that only the list of genes identified by DGEM was significantly different from a collection of random genes.

[0311] This result demonstrates that target selection by DGEM is both more comprehensive and yields better quality results than previous methods in the prior art.

Example 15: Cancer treatment

[0312] A biopsy is taken from tissue suspected of comprising malignant or pre- malignant cells. The tissue is subjected to DGEM to obtain the expression profile of malignant cells, potential pre-malignant cells, and surrounding tissue. The resultant profile identifies that the suspected malignant cells are indeed cancerous, and by comparison with databases, that the cells are of a type known to be of highly metastatic. The profile also identifies that the cancer cells (a) highly express 10 genes known to encode surface-expressed antigens and (b) show defects in tumor suppressor pathways. The surrounding tissue does not show evidence of malignancy or pre-malignant changes, but does show altered gene expression in response to the presence of the malignant cells. As a result, cells in the immediate vicinity of the cancer show a different expression profile from normal cells distal from the cancer.

[0313] Treatment of the cancer is then designed based on the DGEM profile of the tissue. First, antibodies targeting the antigens expressed on the surface of the cancer are injected into the vicinity of the tumor. Second, compounds are administered which cause a toxic effect on the cancer cell but do not affect other cells in the body. Finally, other compounds are administered that preferentially target cells in the immediate vicinity of the cancer, such as endothelial cells, thereby disrupting access of the tumor cells to oxygen and nutrients.

06363

TABLE

Claims

What is claimed is:

1. A method for identifying gene expression by specific cell types in a mixed cell population within a tissue, said method comprising:

(a) obtaining a tissue region;

(b) quantifying the amount and relative level of gene a expression product isolated from said microdissected tissue region to obtain the level of gene expression for the tissue region;

(c) quantifying the amount and relative level of gene expression products in from an isolated cell or from a group of cells to obtain the level of gene expression for an isolated cell or isolated group of cells ;

(d) comparing the level of gene expression from said tissue region against the level of gene expression from said isolated cell or isolated group of cells, thereby

(e) obtaining the level of gene expression for said isolated cell or isolated group of cells and cells not isolated.

2. The method of claim 1, wherein said level of gene expression products are a levels of mRNA or levels of polypeptides.

3. The method of claim 2, wherein said levels of mRNA are quantified by microarray analysis.

4. The method of claim 2, wherein said level of polypeptides are quantified by MALDI- TOF.

5. The method of claim 1, wherein the mixed cell population comprise stromal cells of the thymic cortex or thymic medulla.

6. The method of claim 1, wherein the mixed cell population comprise tumor cells.

7. The method of claim 1, wherein the mixed cell population comprise stem cells.

8. The method of claim 1, wherein the microdissected tissues are morphologically and functionally distinct tissue regions.

9. The method of claim 1, wherein the microdissected tissues are from normal and diseased tissues and organs.

10. The method of claim 9, wherein the tissues comprise epithelial tissue, connective tissues, muscle tissues or nervous tissues.

11. The method of claim 9, wherein the tissues are microdissected from organs, said organs are selected from the group consisting of: skin, digestive, muscular, nervous, respiratory circulatory, excretory, endocrine, lymphatic, and reproductive organs.

12. The method of claim 1, wherein tissues are microdissected by laser capture.

13. The method of claim 1, wherein the microdissected tissues have a thickness of between about 10 μm to 80 μm.

14. The method of claim 10, wherein the microdissected tissues have a thickness of about 40 μm.

15. The method of claim 1, wherein the microdissected tissues have a length of between about 100 μm to 700μm and a depth of between about 5μm to 50 μm

16. The method of claim 2, wherein mKNA is isolated from a section of tissue from about 100 μm to about 500 μm thick.

17. The method of claim 2, wherein mRNA is isolated from the microdissected tissues are subjected to low cycle (limited) high-fidelity amplification.

18. The method of claim 1, wherein expression profiles of microdissected tissue regions are compared to expression profiles of specific cell types.

19. The method of claim 2, wherein expression profiles of microdissected tissue regions are compared to expression profiles of known genes.

20. A method of identifying gene expression in cancer-containing tissue, said method comprising:

(a) microdissecting tissues comprising tumors;

(b) analyzing gene expression products from the microdissected tissues;

(c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue;

(d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue,

(e) identifying gene expression by specific cell types in cancer-containing tissue.

21. The method of claim 18, wherein pre-metastatic, metastatic and post-metastatic gene expression is identified.

22. A method of identifying gene expression in autoimmune diseases, said method comprising:

(a) microdissecting tissues comprising autoimmune cells;

(b) analyzing gene expression products from the microdissected tissues;

(c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue;

(d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue;

(e) identifying gene expression by specific cell types in autoimmune diseases.

23. A method of identifying gene expression in diabetes, said method comprising:

(a) microdissecting pancreatic tissue;

(b) analyzing gene expression products from the microdissected tissue;

(c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue; (d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue,

(e) identifying gene expression by specific cell types in diabetes.

24. A method of identifying gene expression in neural diseases, said method comprising:

(a) microdissecting neural tissues

(b) analyzing gene expression products from the microdissected tissue;

(c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue;

(d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue,

(e) identifying gene expression by specific cell types in neural diseases.

25. A method of identifying pathogen induced gene expression in diseases, said method comprising:

(a) microdissecting tissue infected with a pathogen;

(b) analyzing gene expression products from the microdissected tissue;

(c) analyzing gene expression products from purified individual cells or groups of cells from within the tissue;

(d) comparing and subtracting gene expression profiles of microdissected tissue regions from gene expression profiles obtained from purified individual cells or groups of cells from within the tissue,

(e) identifying gene expression by specific cell types in diseases.

26. The method of claim 25, wherein the pathogen comprises virus, bacteria, fungal or parasitic organism.

27. A method of identifying candidate therapeutic agents for treatment of disease, said method comprising:

(a) administering a candidate agent;

(b) microdissecting control tissues and tissues treated with the candidate agent;

(c) analyzing nucleic acids isolated from the microdissected tissues; (d) analyzing purified individual cellular nucleic acids from the tissues;

(e) comparing and subtracting nucleic acid profiles of microdissected tissue regions from purified individual cellular nucleic acids; and

(f) identifying gene expression by specific cell types in control and treated tissues.

A.

28. Use of the method of claim 1, for determining a change in cell expression in response to a chemical compound, infectious agent, or cellular signal.

29. Use of the method of claim 1, for determining whether a compound of interest affects a given cell type in situ.

30. A method of detecting thymic cortical stromal cells comprising use of a probe against a gene identified in Table 4, or an antibody against a polypeptide gene product identified in Table 4.

31. A method of detecting thymic medullary stromal cells comprising use of a probe against a gene identified in Table 5, or an antibody against a polypeptide gene product identified in Table 5.

32. A method of treating cancer of the thymus, comprising use of an antibody against a gene product identified in Table 4 or 5.

33. A method for diagnosing a human disease or disorder, comprising:

(a) detecting gene expression in a sample, wherein said sample is subjected to microdissection and microarray analysis;

(b) correlating the nucleic acid molecule expression profile with a disease or disorder.

34. The method of claim 33, wherein said disease is selected from the group consisting of cancer, Parkinson's disease, Alzheimer's disease, Huntington's chorea, amyotrophic lateral sclerosis (ALS), nutritional diseases, diabetes, Bell's palsy, systemic lupus erythematosus, multiple sclerosis, human immunodeficiency virus-associated myelopathy, transverse myelopathy or various etiologies, progressive multifocal leukoencephalopathy, and central pontine myelinolysis.

35. A method for determining the expression profile of a sample of interest having at least two types of cells, comprising subtracting the expression profile of one component of said sample from the expression profile of the total sample.

36. A method of diagnosing a disease or disorder, comprising:

(a) detecting gene expression in a sample, wherein said sample is subjected to microdissection;

(b) subjecting the mRNA to microarray analysis; and

(c) correlating the nucleic acid molecule expression profile with a disease or disorder.

37. A method for calculating the expression profile of a sample of interest having at least two types of cells, comprising subtracting the expression profile of one component of said sample from the expression profile of the total sample.