METHODS, COMPUTER-READABLE MEDIA, AND SYSTEMS FOR ASSESSING SAMPLES AND WOUNDS, PREDICTING WHETHER A WOUND WILL HEAL, AND MONITORING EFFECTIVENESS OF A TREATMENT
20170247758 · 2017-08-31
Assignee
Inventors
- Kara SPILLER (Glenside, PA, US)
- Sina NASSIRI (Philadelphia, PA, US)
- Michael WEINGARTEN (Penn Valley, PA, US)
Cpc classification
C12Q1/6881
CHEMISTRY; METALLURGY
C12Q1/6883
CHEMISTRY; METALLURGY
G01N2800/52
PHYSICS
International classification
Abstract
One aspect of the invention provides a method of predicting whether a wound will heal. The method includes: obtaining a first measurement of a first macrophage phenotype population within a first sample obtained from a wound; obtaining a second measurement of a second macrophage phenotype population from the wound, wherein the second measurement of the second macrophage phenotype population is either: a different macrophage phenotype obtained from the first sample or the same macrophage phenotype obtained from a second, later sample from the wound; comparing the first measurement to the second measurement; and predicting whether the wound will heal based on a result of the comparing step.
Claims
1. A method of predicting whether a wound will heal, the method comprising: obtaining a first measurement of a first macrophage phenotype population within a first sample obtained from a wound; obtaining a second measurement of a second macrophage phenotype population from the wound, wherein the second measurement of the second macrophage phenotype population is either: a different macrophage phenotype obtained from the first sample; or the same macrophage phenotype obtained from a second, later sample from the wound; comparing the first measurement to the second measurement; and predicting whether the wound will heal based on a result of the comparing step.
2.-20. (canceled)
21. A method of assessing a sample, the method comprising: calculating a first ratio of M1 macrophages to M2 macrophages in a first sample based on gene expression values for at least one marker associated with M1 macrophage activity and at least one marker associated with M2 macrophage activity.
22.-27. (canceled)
28. The method of claim 21, wherein the calculating step includes: calculating a first function of gene expression values of each of a first plurality of markers associated with M1 macrophages; and calculating a second function of gene expression values of each of a second plurality of markers associated with M2 macrophages.
29.-37. (canceled)
38. The method of claim 21, further comprising: calculating a second ratio of M1 macrophages to M2 macrophages in a second sample based on gene expression values for at least one marker associated with M1 macrophage activity and at least one marker associated with M2 macrophage activity, the second sample obtained from a same source as the first sample after passage of a period of time; and comparing the second ratio to the first ratio.
39.-51. (canceled)
52. A non-transitory computer readable medium containing computer-readable program code including instructions for performing the method of claim 1.
53. A system comprising: a gene expression device; and a processor programmed to implement the method of claim 1.
54. (canceled)
55. A method of assessing a wound, the method comprising: extracting RNA from debrided wound tissue; measuring expression of one or more genes within the RNA; and calculating a ratio of M1 macrophages to M2 macrophages based on the measured gene expression.
56. The method of claim 55, wherein the debrided wound tissue was removed from a dressing previously applied a wound.
57. The method of claim 55, wherein the debrided wound tissue is from one or more selected from the group consisting of: a diabetic ulcer, a pressure ulcer, a chronic venous ulcer, a burn, a wound caused by an autoimmune disease, a wound caused by Crohn's disease, a wound caused by atherosclerosis, a tumor, a medical implant insertion point, a surgical wound, a bone fracture, a tissue tear, and a tissue rupture.
58. The method of claim 55, wherein the measuring expression step includes using one or more tools or techniques selected from the group consisting of: cDNA synthesis, quantitative PCR (qPCR), microarrays, and RNA Sequencing (RNA-seq).
59. A high-throughput screening system comprising: a measurement device; and a data processor programmed to implement the method of claim 1.
60. A method of monitoring effectiveness of a treatment of a non-healing wound or a tumor, the method comprising: administering to a patient a therapeutic agent designed to treat a non-healing wound or a tumor; obtaining a first measurement of a first macrophage phenotype population within a first sample obtained from the non-healing wound or the tumor; obtaining a second measurement of second macrophage phenotype population from the non-healing wound or the tumor, wherein the second measurement of the second macrophage phenotype population is either: a different macrophage phenotype obtained from the first sample or the tumor; or the same macrophage phenotype obtained from a second, later sample from the non-healing wound or the tumor; comparing the first measurement to the second measurement; and assessing whether the treatment of the non-healing wound or the tumor is effective based on a result of comparing the measurements.
61. The method of claim 60, wherein the therapeutic agent is selected from the group consisting of an L-arginine, hyperbaric oxygen, a moist saline dressing, an isotonic sodium chloride gel, a hydroactive paste, a polyvinyl film dressing, a hydrocolloid dressing, a calcium alginate dressing, and a hydrofiber dressing.
62. The method of claim 60, wherein the treatment is low-intensity ultrasound treatment.
63. The method of claim 60, further comprising comparing an M1/M2 ratio with a threshold value that discriminates between (i) wound healing and non healing or (ii) tumor progression and non-progression; and adjusting the treatment based on the M1/M2 ratio, wherein: if the M1/M2 ratio is at or below the threshold value, the administration of therapeutic agent is increased, and if the M1/M2 ratio is above the threshold value, the administration of the therapeutic agent is not increased.
64. The method of claim 63, wherein if the level is at or below the threshold value, the therapeutic agent is replaced by a different therapeutic agent.
65. A method of treating a wound comprising: administering an effective amount of interferon gamma (IFNg) to the wound.
66. A non-transitory computer readable medium containing computer-readable program code including instructions for performing the method of claim 21.
67. A system comprising: a gene expression device; and a processor programmed to implement the method of claim 21.
68. A high-throughput screening system comprising: a measurement device; and a data processor programmed to implement the method of claim 21.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0027] For a fuller understanding of the nature and desired objects of the present invention, reference is made to the following detailed description taken in conjunction with the accompanying drawing figures wherein like reference characters denote corresponding parts throughout the several views.
[0028]
[0029]
[0030]
[0031]
[0032]
[0033]
[0034]
[0035]
[0036]
[0037]
[0038]
[0039]
[0040]
[0041]
[0042]
[0043]
[0044]
[0045]
[0046]
[0047]
[0048]
[0049]
[0050]
[0051]
[0052]
[0053]
[0054]
[0055]
[0056]
[0057]
[0058]
[0059]
[0060]
[0061]
DEFINITIONS
[0062] The instant invention is most clearly understood with reference to the following definitions.
[0063] As used herein, the singular form “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.
[0064] Unless specifically stated or obvious from context, as used herein, the term “about” is understood as within a range of normal tolerance in the art, for example within 2 standard deviations of the mean. “About” can be understood as within 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05%, or 0.01% of the stated value. Unless otherwise clear from context, all numerical values provided herein are modified by the term about.
[0065] As used herein, the terms “comprises,” “comprising,” “containing,” “having,” and the like can have the meaning ascribed to them in U.S. patent law and can mean “includes,” “including,” and the like.
[0066] As used herein, the term “healing” refers to the process by which a body repairs itself after injury. The healing process can include several stages such as hemostasis (blood clotting), inflammation, proliferation (growth of new tissue), and maturation (remodeling). Embodiments of the invention can be used to make predictions regarding whether the wound will progress through all or the rest of the healing process without the need for enhanced techniques or can be utilized to make predictions regarding whether wound will progress to a particular stage of healing (e.g., proliferation) without the need for enhanced techniques.
[0067] As used herein, the term “high-throughput screening” refers to a screening method or system that allows analysis of a large number of samples by analyzing the presence, absence, relative levels, or response in one or more measurements including, but not limited to, nucleic acid makeup, gene expression, protein levels, functional activity, response to a stimulus, etc.
[0068] The terms “conversion,” and “converting” refer to the change in macrophage phenotype from one macrophage phenotype to another macrophage phenotype.
[0069] The terms “induce,” and “induction” refer to the promoting a change in macrophage phenotype from one macrophage phenotype to another macrophage phenotype.
[0070] The terms “isolated,” “purified,” or “biologically pure” refer to material that is free to varying degrees from components which normally accompany it as found in its native state. “Isolated” denotes a degree of separation from original source or surroundings. “Purified” denotes a degree of separation that is higher than isolation. A “purified” or “biologically pure” protein is sufficiently free of other materials such that any impurities do not materially affect the biological properties of the protein or cause other adverse consequences. That is, a nucleic acid or peptide is purified if it is substantially free of cellular material, viral material, or culture medium when produced by recombinant DNA techniques, or chemical precursors or other chemicals when chemically synthesized. Purity and homogeneity are typically determined using analytical chemistry techniques, for example, polyacrylamide gel electrophoresis or high performance liquid chromatography. The term “purified” can denote that a nucleic acid or protein gives rise to essentially one band in an electrophoretic gel. For a protein that can be subjected to modifications, for example, phosphorylation or glycosylation, different modifications may give rise to different isolated proteins, which can be separately purified. “Purified” can also refer to a molecule separated after a bioconjugation technique from those molecules that were not efficiently conjugated.
[0071] The phrase “macrophage conversion” as used herein refers to the sequential change in macrophage phenotype, e.g., a macrophage transitioning from pro-inflammatory (M1) to pro-healing (M2a) to pro-remodeling (M2c) phenotypes.
[0072] The term “wound macrophage” as used herein refers to a hybrid population of macrophages in a wound including a spectrum of macrophage phenotypes and subtypes that include, but are not limited to, M0, M1, and M2 (including M2a and M2c) macrophages.
[0073] The term “M1 macrophage” as used herein refers to a macrophage phenotype. M1 macrophage are classically activated or exhibit an inflammatory macrophage phenotype.
[0074] The term “M2” broadly refers to macrophages that function in constructive processes, like wound healing and tissue repair. Major differences between M2a and M2c macrophages exist in wound healing.
[0075] The term “M2a macrophage” as used herein refers to a macrophage subtype of pro-healing macrophages. M2a macrophages are involved in immunoregulation.
[0076] The term “M2c macrophage” as used herein refers to a macrophage subtype of pro-remodeling macrophages. M2c macrophages are involved in matrix and vascular remodeling and tissue repair.
[0077] Unless specifically stated or obvious from context, the term “or,” as used herein, is understood to be inclusive.
[0078] Ranges provided herein are understood to be shorthand for all of the values within the range. For example, a range of 1 to 50 is understood to include any number, combination of numbers, or sub-range from the group consisting 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 (as well as fractions thereof unless the context clearly dictates otherwise).
[0079] As used herein, the term “ratio” refers to a relationship between two numbers (e.g., scores, summations, and the like). Although, ratios can be expressed in a particular order (e.g., a to b or a:b), one of ordinary skill in the art will recognize that the underlying relationship between the numbers can be expressed in any order without losing the significance of the underlying relationship, although observation and correlation of trends based on the ration may need to be reversed. For example, if the values of a over time are (4, 10) and the values of b over time are (2, 4), the ratio a:b will equal (2, 2.5), while the ratio b:a will be (0.5, 0.4). Although the values of a and b are the same in both ratios, the ratios a:b and b:a are inverse and increase and decrease, respectively, over the time period.
[0080] As used herein, the term “initial medical encounter” encompasses one or more related interactions with one or more medical professionals. For example, if a subject visits her primary care provider's office regarding a wound, her interactions with a medical assistant, nurse, physician's assistant, and/or physician would constitute a single “medical encounter.” Likewise, a subject's interactions with a plurality of medical professionals during an emergency department visit would also constitute an “initial medical encounter.” The term “initial medical encounter” also encompasses the first interaction with a medical professional specializing in wound care. For example, a subject's first appointment with a wound clinic could be considered an “initial medical encounter.” The “initial medical encounter” can be the actual first or subsequent encounter with a medical professional. For example, a medical professional may not obtain a first sample until after the wound persists from a first appointment to a second appointment.
[0081] As used herein, the term “sample” includes biological samples of materials such as organs, tissues, cells, fluids, and the like. In one embodiment, the sample can be obtained from a wound. In other embodiments, the sample can be obtained from inflamed tissue such as tissue afflicted with Inflammatory Bowel Syndrome, Crohn's disease, and the like. In still another embodiment, the tissue can be cancerous tissue (in which an increase in M1/M2 ratio would be desired for inhibition of tumor progression and a low or decreasing M1/M2 ratio would be indicative of tumor progression and metastasis). In still another embodiment, the sample can be obtained from an in vivo or in vitro testing platform such as a culture dish, a scaffold, an artificial organ, a laboratory animal, and the like.
[0082] As used herein, the term “wound” includes injuries in which the skin (particularly, the dermis) is torn, cut, or punctured. Examples of types of wounds that can be assessed using embodiments of the invention described herein include external wounds, internal wounds, clean wounds (e.g., those made in the course of a medical procedure such as surgery), contaminated wounds, infected wounds, colonized wounds, incisions, lacerations, abrasions, avulsions, puncture wounds, penetration wounds, gunshot wounds, and the like. Specific wound examples include diabetic ulcers, pressure ulcers (also known as decubitus ulcers or bedsores), chronic venous ulcers, burns, and medical implant insertion points. Embodiments of the invention are particularly useful in identifying nonhealing wounds that are prevalent in diabetic and/or elderly subjects.
DETAILED DESCRIPTION OF THE INVENTION
[0083] Previously proposed indicators of healing outcome biomarkers for diagnosis of nonhealing wounds suffer from high variability between wounds, technical difficulties in detection methods, and impose burdens both on the patient and the care provider because the methods of detection are not a normal part of the wound care regimen.
[0084] Aspects of the invention utilize genetic information about macrophage behavior to identify differences between healing and nonhealing in diabetic chronic wounds. Macrophages are the central cell of the inflammatory response and are recognized as primary regulators of wound healing, with their phenotype orchestrating events specific to the stage of repair. Macrophages exist on a spectrum of phenotypes ranging from pro-inflammatory or “M1” to anti-inflammatory and pro-healing or “M2.” M2 macrophages can be further categorized as M2a, M2b, or M2c macrophages. In early stages of wound healing (1-3 days), M1 macrophages secrete pro-inflammatory cytokines and clear the wound of debris. In later stages (4-7 days), macrophages switch to the M2 phenotype and promote extracellular matrix (ECM) synthesis, matrix remodeling, and tissue repair. If the M1-to-M2 transition is disrupted, depicted by persistent numbers of M1 macrophages, the wound suffers from chronic inflammation and impaired healing.
[0085] While abnormal macrophage activation in diabetic wounds has been thoroughly described in animal models of diabetes, it has not yet been assessed in human diabetic wounds.
[0086] Applicant proposes that absolute, relative, and proportional counts of M1, M2, M2a, M2b, and/or M2c macrophages as well as surrogates thereof can be utilized to predict whether a wound will heal.
[0087] In one embodiment of the invention, Applicant investigated differential expression of M1 and M2 genes over time in human diabetic wounds, hypothesizing that healing wounds would exhibit a decrease in the relative proportion of M1 to M2 macrophages. Furthermore, Applicant investigated if gene expression signatures of M1 and M2 macrophages cultured in vitro could be used to quantify wound healing progression, and found that this method may hold potential as a novel noninvasive or minimally invasive diagnostic assay.
Methods of Assessing a Sample and/or Predicting Whether a Wound Will Heal
[0088] Referring now to
[0089] In step S1802, a first measurement of a first macrophage phenotype population within a first sample obtained from a wound is obtained.
[0090] Exemplary techniques for obtaining a sample from a wound are discussed herein.
[0091] The first measurement of a first macrophage phenotype population can be any measurement of the number of macrophages within a sample or a volumetric or mass unit thereof or a surrogate for the same. For example, the number of macrophages can be measured using microscopy or one or more measurements correlated with a population of macrophages can be measured using one or more techniques that measure the amount of a substance produced or expressed by the population of macrophages.
[0092] Suitable techniques for measuring a surrogate of macrophage population include, but are not limited to, flow cytometry, immunostaining, and other techniques for measuring gene expression, protein expression, cytokines, and/or other metabolomics byproducts associated with particular macrophage phenotypes.
[0093] Gene expression data can be processed or analyzed using a sets of individual expression values as discuss herein (e.g., through linear sums and other algorithms). Additionally or alternatively, gene expression data can be presented using a variety of gene set enrichment analysis algorithms that assess activation of a family of genes that are associated with a biological pathway or functionality (often referred to as a “gene set”), as opposed to individual genes. Exemplary gene set enrichment analysis algorithms include but are not limited to the GSEA method as described in A. Subramanian et al., “Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles,” 102(43) PNAS 15545-50 (2005) and V. Mootha et al., “PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes,” 34 Nature Genetics 267-73 (2003) and the QUSAGE method as described in G. Yaari et al., “Quantitative set analysis for gene expression: a method to quantify gene set differential expression including gene-gene correlations,” 41(18) Nucleic Acids Res. e170 (October 2013) and available at http://clip.med.yale.edu/qusage/. The GSEA and QUSAGE methods both yield a score that can be used by itself or in a ratio with scores reflective of other macrophage populations to make the comparisons discussed herein.
[0094] Various techniques can be utilized to determine which data (e.g., genes and corresponding functions combining particular genes) are particularly relevant in assessing a macrophage population. These techniques can be divided into two broad categories. The first category includes methods that preserve all features (in this case, genes) and may or may not include weighting strategies to give more weight to more important features or based on the correlation of a feature with a certain outcome. For example, statistical hypothesis testing such as a t-test can be used to weight features as described herein, or correlation coefficient of a feature with a certain outcome can be used to weight features. The second category includes methods that use a subset of features. This subset can be obtained through a variety of methods known as dimensionality reduction methods. Dimensionality reduction methods can be either linear such as principal component analysis (PCA), independent component analysis (ICA), singular value decomposition (SVD), and non-negative matrix factorization, or non-linear such as kernel PCA and graph-based methods (also known as Laplacian eigenmaps). The new combinatorial features, which number far less than the number of features all together, are then treated as new variables. Alternative methods for feature subset selection include use of discrimination properties of features. In this regard, if features are treated individually, a variety of class separablility measures such as the receiver operating characteristics (ROC) curves, Fisher's discriminant ratio, and one-dimensional divergence can be used to select a subset of features. These methods, however, do not take into account the correlation that may exist among features and as a result their influence on the classification capabilities of the selected subset of features. To address this limitation, techniques measuring classification capabilities of feature vectors are applied. Neural networks can also be applied for feature generation and selection.
[0095] In step S1804, a second measurement of second macrophage phenotype population from the wound is obtained. The second measurement of the second macrophage phenotype population can be a different macrophage phenotype obtained from the first sample or the same macrophage phenotype obtained from a second, later sample from the wound. For example, if a single sample is used, the first measurement can relate to the M1 macrophage population and the second measurement can relate to the M2 macrophage population (e.g., all M2 macrophages or one or more of M2a, M2b, and/or M2c macrophages). If the second measurement is obtained from a second, chronologically later sample from the wound, the first and the second measurement can relate to the same macrophage phenotype in both measurements (e.g., a first measurement of M1 macrophages and a second measurement of M1 macrophages, a first measurement of M2 macrophages and a second measurement of M2 macrophages, a first measurement of M2a macrophages and a second measurement of M2a macrophages, a first measurement of M2b macrophages and a second measurement of M2b macrophages, a first measurement of M2c macrophages and a second measurement of M2c macrophages, and the like, including ratios of measurements).
[0096] In step S1806, the first measurement is compared to the second measurement. In one embodiment, this comparison is expressed as a ratio as discussed herein.
[0097] In step S1808, a prediction of whether the wound will heal is made based on a result of the comparing step. Without being bound by theory, it is believed that ratios exceeding the thresholds specified in Tables 1 and 2 herein are indicative of wounds that will heal without the need for enhanced techniques such as the use of synthetic skin substitutes, hyperbaric oxygen therapy, or negative-pressure wound therapy.
TABLE-US-00001 TABLE 1 Exemplary thresholds for wound healing predictions based on single sample, where the single sample constitutes the genes and methods described in FIG. 9 (“linear sum method” for M1/M2a and IL1B/CD163 for M1/M2c) First Measurement Second Measurement Threshold (1st:2nd) M1(t.sub.0) M2a(t.sub.0) 320 M1(t.sub.0) M2c(t.sub.0) 126
TABLE-US-00002 TABLE 2 Exemplary thresholds for wound healing predictions based on two samples First Measurement Second Measurement Threshold (1st:2nd) M1(t.sub.0)/M2a(t.sub.0) M1(t.sub.1)/M2a(t.sub.1) 4.6 M2a(t.sub.0)/M2c(t.sub.0) M2a(t.sub.1)/M2c(t.sub.1) 4
[0098] Referring now to
[0099] In step S102, a biological sample can be obtained (e.g., from a wound of a subject). In one embodiment, the biological sample is debrided tissue, which can include, but is not limited to, dead, damaged, or infected tissue. A variety of debriding techniques can be applied.
[0100] In one embodiment, mechanical debridement is used in which removal of a dressing from a wound that proceeded from moist to dry will non-selectively remove tissue adjacent to the dressing. This removed tissue can then be separated from the dressing (e.g., by scraping, rinsing, and the like). Advantageously, harvesting of debrided tissue from removed dressings avoids the challenges associated with more invasive approaches and provides sufficient quantities of human wound tissues for quantitative analyses of the cellular content using tissue that would otherwise be discarded.
[0101] In another embodiment, surgical debridement can be performed using various surgical tools such as a scalpel, a laser, and the like. Advantageously, harvesting of debrided tissue avoids the challenges associated with more invasive approaches such as using punch biopsies while providing sufficient quantities of human wound tissues for quantitative analyses of the cellular content using tissue that would otherwise be discarded. Although relatively non-invasive procedures can be used, the samples used herein can also be obtained through invasive procedures such as punch biopsies, shave biopsies, incisional biopsies, excisional biopsies, curettage biopsies, saucerization biopsies, fine needle aspiration, and. the like.
[0102] In step S104, the sample can be preserved and/or stabilized until further analysis can be performed. For example, the sample can be immersed in a stabilization reagent such as RNALATER® stabilization reagent available from QIAGEN of Venlo, Netherlands.
[0103] In step S106, RNA can be extracted from the sample, for example by using a lysing agent such as the TRIZOL® Plus RNA Purification Kit available from Life Technologies of Grand Island, N.Y.
[0104] In step S108, complementary DNA ((DNA) can be synthesized from the extracted RNA by using, for example, an APPLIED BIOSYSTEMS® High-Capacity cDNA Reverse Transcription Kit available from Life Technologies.
[0105] In step S110, expression of one or more markers can be measured, for example, using quantitative polymerase chain reaction (qPCR). Exemplary approaches to steps S108 and S110 are described in K. L. Spiller et al., “The role of macrophage phenotype in vascularization of tissue engineering scaffolds,” 35(15) Biomaterials 4477-88 (May 2014) (hereinafter “Spiller 2014”).
[0106] Exemplary markers associated with M1 macrophage activity include YEGF, CCR7, CD80, and IL1B. Exemplary makers associated with M2 macrophage activity include CCL18, CD206, MDC, PDGF, and TIMP3. Exemplary makers associated with M2c macrophage activity include MMP7, CD163, TIMP1, Marco, VCAN, SH3PXD2B, MMP8, PLAU, PROS1, SRPX2, NAIP, and F5. Sequences for these markers are provided in Tables 3-6 below.
TABLE-US-00003 TABLE 3 Sequences for exemplary markers of M1 activity Gene Forward Sequence Reverse Sequence CCR7 TGAGGTCACGGACGATTACAT (SEQ ID NO: 1) GTAGGCCCACGAAACAAATGAT (SEQ ID NO: 2) CD80 AAACTCGCATCTACTGGCAAA (SEQ ID NO: 3) GGTTCTTGTACTCGGGCCATA (SEQ ID NO: 4) IL1B ATGATGGCTTATTACAGTGGCAA (SEQ ID NO: 5) GTCGGAGATTCGTAGCTGGA (SEQ ID NO: 6) (IL1P)
TABLE-US-00004 TABLE 4 Sequences for exemplary markers of M2 activity Gene Forward Sequence Reverse Sequence CCL18 GCTCTCTGCCCGTCTATACC (SEQ ID NO: 7) GGGCTGGTTTCAGAATAGTCAACT (SEQ ID NO: 8) CD163 TTTGTCAACTTGAGTCCCTTCAC (SEQ ID NO: 9) TCCCGCTACACTTGTTTTCAC (SEQ ID NO: 10) CD206 AAGGCGGTGACCTCACAAG (SEQ ID NO: 11) AAAGTCCAATTCCTCGATGGTG (SEQ ID NO: 12) (MRC1 MDC GCGTGGTGTTGCTAACCTTCA (SEQ ID NO: 13) AAGGCCACGGTCATCAGAGT (SEQ ID NO: 14) (CCL22 PDGFB CTCGATCCGCTCCTTTGATGA (SEQ ID NO: 15) CGTTGGTGCGGTCTATGAG (SEQ ID NO: 16) TIMP3 ACCGAGGCTTCACCAAGATG (SEQ ID NO: 17) CATCATAGACGCGACCTGTCA (SEQ ID NO: 18)
TABLE-US-00005 TABLE 5 Sequences for exemplary markers of M2a activity Gene Forward Sequence Reverse Sequence MDC GCGTGGTGTTGCTAACCTTCA (SEQ ID NO: 19) AAGGCCACGGTCATCAGAGT (SEQ ID NO: 20) (CCL22) CD206 AAGGCGGTGACCTCACAAG (SEQ ID NO: 21) AAAGTCCAATTCCTCGATGGTG (SEQ ID NO: 22) (MRC1)
TABLE-US-00006 TABLE 6 Sequences for exemplary markers of M2c activity Gene Forward Sequence Reverse Sequence CD163 TTTGTCAACTTGAGTCCCTTCAC (SEQ ID NO: 23) TCCCGCTACACTTGTTTTCAC (SEQ ID NO: 24)
[0107] Lists of the most expressed genes for M1, M2a, and M2c macrophage populations are provided in Tables 7, 8, and 9, respectively. The genes are arranged in descending order by rows and then by columns. HUGO Gene Nomenclature Committee (HGNC) symbols are provided for each gene. Corresponding ENSEMBL IDs and EntrezGene IDs are provided in the files incorporated by reference herein and are also available through publicly available databases.
TABLE-US-00007 TABLE 7 Most expressed genes for M1 macrophages B2M C6orf48 CPEB3 C1RL-AS1 RPL7P18 HLA-B RSAD2 C22orf46 EBI3 PPP1R17 HLA-A ZC3H12A OASL ZNF165 MORN3 SOD2 C1RL RASGRP1 APOBEC3B ADORA2A HLA-C H1F0 ZFAND4 CCL8 GPRC5B HLA-E PSMB10 C21orf91 NKG7 DSCAML1 WARS PSMB9 STAT4 NEURL3 MT-TM SAT1 TNIP2 PDE4D LAMA5 CXCL10 CD74 LMTK2 MICA MFSD4 CKMT1B RNF213 MT1G IL15RA A4GALT CTRL PLAUR SERTAD1 LAMA3 IGFLR1 PLLP ACSL1 TRAF3IP2 GADD45G MMP25 ASAP3 NCOA4 NFKBIZ IL15 HOXB2 PLIN5 MCL1 PSTPIP2 WEE1 SPOCK2 UPK3A SNX10 HLA-H CLEC2D ADAM19 NTNG2 CYP27B1 C1QC MDGA1 HAPLN3 MTND1P11 MMP14 PELI1 ERO1LB TWF1P1 NFE4 STAT1 TRIM21 VAMP5 HOXA1 CES1P1 HLA-DRA XAF1 TUBE1 MIR4519 OR2A7 CFLAR DYNLT1 RAB43 ITGA9 GK-IT1 TNFAIP2 MSRB1 CMKLR1 IFITM1 SBK3 SLAMF7 ADAM28 GVINP1 FBLIM1 SRD5A3-AS1 GNA13 OTUD1 RMI2 TSPAN5 PROB1 CD83 APOO C4orf32 SH2D3A KCNE1L BTG1 TMEM170A ITK ANKRD22 EXOC3L1 LAP3 SEMA4A WWC3 CLIC3 RAB44 IFI6 BTN3A3 MAMDC4 LNX1 TNK2-AS1 ZNFX1 MYO1G TMOD2 RAPSN KCNN1 IL8 TOR1B FAM26F MAGI2-AS3 HLA-DRB9 ATF3 DPP4 ZNF702P LMTK3 C9orf50 TXN FBXO32 CHST3 GATA2 BAIAP2L1 PTPRJ PVR CCDC149 MEFV TMEM54 PARP14 HIVEP2 IRF9 NFIX TINCR TAP1 IFIT2 DFNB31 MT1L GBP7 ICAM1 PRPF3 OSM UNC13A SCG3 TYMP APOL1 OSBP2 SCN4B DMGDH NFE2L1 APOL3 GCH1 PTGES3P1 NKX3-1 CLCN7 ARL5B HLA-DRB6 ARHGEF35 LAX1 SCPEP1 RELB CRISPLD2 IGFBP4 PRPH MTHFD2 ST8SIA4 PLEKHG3 DNAJC3-AS1 KRT7 NFKBIA N4BP2L1 MAP1LC3A KIAA1045 TRPA1 STOM RGL1 DTNB GBP1P1 RAB39B STAT2 FCHSD1 NCF1C NOTCH4 CFH PIM1 CCL5 KSR1 PLEKHN1 C11orf96 HLA-DPA1 ENDOD1 CFB FAM46C IL1RL1 KYNU MT1E LYSMD2 IL32 HLA-V TSC22D1 PPA1 MIR155HG BEGAIN LPAR4 NAMPT TDRD7 MEIS3 KCNA3 ANXA3 KIAA0247 PLA2G4C RTP4 DOCK9 HSD17B13 HLA-DRB1 PHF1 RAB3IP IL27 PBX1 NINJ1 C19orf12 RARG GRHL1 S100A3 PTAFR FADS3 HCAR3 FAM71A DTX3 PNRC1 SLC37A1 EDN1 TFAP2E DMBX1 C15orf48 DHX58 BBS12 TXNP6 LINC00482 IL13RA1 GPR157 CLEC4E HIP1R MST1R TNFAIP3 RHBDD2 IRF6 CD6 RGS11 B4GALT5 TAPBPL STAP2 HLA-DQB2 FAM71E1 APOL6 TAP2 BATF2 CCDC80 PRSS8 C3 AP1AR TNF CAV2 SOX5 PILRA ACVRL1 DIRAS2 PNPLA1 RPL32P1 SLC31A2 NAB1 ITGA2 AOC2 RGS9 PSME2 IFI35 RARRES3 GFPT2 ICOS GBP2 USP11 CEBPE VWA7 SERPINB7 CD48 HMCES CCDC154 BEAN1 KIT RDX HLA-DQA1 CXCL1 CXCL9 CPA4 ZCCHC6 SEMA4D ASPHD2 GALNT3 ARSI MT2A PROCR PERP P2RY10 CCL15 LILRA3 NUPR1 PARD6B FAM177B ABCC11 RHBDF2 RUNX3 VMO1 KL FAAH2 SMAP2 TIFA CD38 CCL20 C9orf172 TBC1D9 INSIG2 AMZ1 ANKRD1 EXOC3L4 CXCL5 TBKBP1 PIPOX CLEC6A EPN3 ALDH2 MAP3K8 SPAG1 HDC MTND6P5 CD274 ULK2 ZEB1 PRRG2 PRF1 UBE2L6 ZMYND15 TSHZ3 RHEBL1 FGF2 METTL9 GIMAP2 BRIP1 PRDM8 CCDC73 DUSP5 IDO1 TMEM154 LINC00189 SAPCD1-AS1 B4GALT1 BCL9L LRRC32 PTPRH NT5C3AP1 TANK MESDC1 VILL UPB1 ORM2 DOCK4 ARNTL2 GRAMD3 TMEM25 TMEM92 SDC4 IL1B HLA-DOA UPK3B FAM3B RAP1B LAMB3 CREB5 CEACAM4 TSPAN1 LY6E CSRNP2 HCAR2 SPAG5-AS1 NBEAP1 HLA-F PSMA6 ARHGEF5 PLEKHA7 SMIM5 CDC42SE2 PDP1 DDR1 C15orf62 NRG2 SLFN5 DTX2 STOML1 LINC00996 SPTA1 ADAMDEC1 ASAP2 NCF1B SKOR1 CYP4F3 TMEM140 MICB AMER1 RND1 CTLA4 PSME1 LONRF1 PGAP1 ACTRT3 TREML4 SNHG16 IFIT5 KLF5 GP1BA SERINC4 SLC39A8 NMI TLCD2 MAP3K9 MTHFD2P7 ARID5B FBXO6 CARD16 FCGR1C KRT23 CSF2RB GADD45B RNF207 CCDC157 ITM2A DTX3L MDK REC8 HES4 C19orf84 PCNX CREM THNSL2 SIK3-IT1 CCL19 TRIM25 DDX58 PDGFRB IL6 CCR10 SAMD9L LAMP3 TNFRSF18 FGF13 FEZ1 OPTN SCO2 USP18 ANKRD33B GPR174 BAZ2A PLEKHM3 SEMA4C F2RL1 RDH16 XRN1 SLC2A6 SERPINI1 FAM227B IL22RA1 INHBA PMAIP1 C12orf79 ALG1L13P ELOVL3 MET FAM135A PTCRA OR2A1-AS1 DMC1 APOL2 FAS ABCC6 TNIP3 TTC22 SOCS3 C9orf91 IGF2BP3 MYCBPAP LINC00337 LGALS3BP MEI1 TMEM216 NKAPL SMCO2 PVRL2 SASH1 ZG16B EYA4 RNU7-45P TRIM22 TMEM150B MAMLD1 LYPD3 C6orf100 MKNK2 IL18BP TMEM229B TRIM31 FAM169A WTAP MPZL3 ABTB2 RLTPR HMGB1P3 ACOT9 ITPR3 ASCL2 MUC1 MAPK8IP2 RIPK2 PLAGL2 C5orf56 HLA-DOB IL12B SAMD9 NCF1 ZNF462 SIX5 ADRB1 ZBTB38 LILRB2 APOBEC3F SH3BP4 ANKRD2 AARS SMPD2 SLC51B ILDR1 ACHE SBNO2 IRF7 RNF144A OLIG2 SPAG6 P2RX7 FOXO4 PARP15 CYP4F22 TRIM55 P2RX4 VPS9D1 STX1A TMPRSS4 OR2A20P CD40 JMY GLIS3 GCNT4 CDC42BPG NLRC5 NT5C3A NAGS ATP8A2 PADI4 ANKRD13A APOBEC3G RPS6KA5 APOBEC3H CEACAM1 ERAP2 PPP3CC LRRCC1 BANK1 CAMK1G TLR2 C1orf132 SDCBP2 SRSF12 TULP2 RNF144B TBX21 TPT1-AS1 ADAMTSL4-AS1 MIR3945 NBN NOD2 CBLN3 CEACAM19 DNAJC22 IFNLR1 CCR7 FZD4 CCR3 MT1A RICTOR GBP4 HLA-K BAIAP3 GFAP C1QA CCNA1 KCNG1 ITPKA RPL21P44 GPBP1 GBP3 S100P IGHEP2 LIPH CALCOCO2 HIP1 NRCAM IL31RA AJUBA TMEM38B ISG15 XIRP1 CIB2 ISLR CUL1 SIGLEC1 IL12RB1 CCDC96 CPXM1 CLU ISOC1 SLC2A12 TMEM8B RNF43 LSS CBX4 PTGES EXPH5 DRD4 GBP1 RHOU PTK7 PLA2G16 EBF4 LCP2 SLC25A28 FSTL1 HESX1 GREB1L NUB1 MOB3C TAF1A PAX5 NELL2 MXD1 PARP3 YES1 ADAMTS2 RSPO3 RNF114 GIMAP6 HMGN2P46 SLC38A5 B3GNT3 AP5B1 MLKL FBN1 SPRY4 HCG4P11 IRF1 C15orf39 C6orf223 SYNPO2 SNAP25 ECE1 FAM177A1 CASZ1 MT-TW CYP2E1 MARCKS OVOL1 TUBD1 AGBL2 IGFBP3 ITGB7 NFE2L3 VNN3 HCG4P7 SLC8A2 PPP4R2 EREG RRAD CNTNAP1 RN7SL124P HLA-DPB1 KREMEN1 FAM225A IDO2 TCEA3 BTN3A1 PTGS2 ATHL1 MN1 PLXNA4 BIRC3 GOLM1 IKZF4 SPATC1 LINC01093 PARP9 ITPKC PSME2P2 SYNGR3 KCTD14 LRRK2 C3AR1 GRIN3A CACTIN-AS1 CXCR3 CCSER2 IFI44 CEACAM21 DNAJC19P5 SLC12A3 STX11 PRRG4 AFAP1 FBXO2 NNMT SAMD8 CSF2RA IFI44L TEAD4 SEZ6L RAPGEF2 CIITA CEP19 EPB41L5 CTHRC1 OAS3 DGAT2 RHOH LYPD5 PDE9A FEM1C HEBP2 MEP1A CA13 BCL6B OSGIN2 C2 SGPP2 LAP3P2 IL36G JAK2 CXCL2 APOL4 TMEM110-MUSTN1 ALMS1P ITPRIP SLC25A37 FAM47E-STBD1 EPHX3 C15orf26 ELMO2 RINL HLA-DQA2 ORM1 LINC00243 NUP50 APBA3 SYT12 C4A GBP6 TRANK1 TRIM5 CD80 EPHA2 VSIG2 CNP MB21D1 GPBAR1 TECTA PPP1R1A KDM7A HSBP1L1 GPR133 NTN1 LINC00158 CREBRF LRRC61 GIMAP7 CR1L CAND2 DAG1 BPGM N4BP3 DKK2 ABCC6P1 RCN1 GIMAP8 ARHGEF34P TSPAN9 KLF15 RILPL2 PRKAR2B SHISA4 RN7SL834P SPNS2 KLHL21 LMNB1 ZNF425 HIC1 OR7E140P ATG2A PIM2 NID1 IRG1 FDPSP3 VEGFA TAGAP TJP3 PLA1A FAT3 IFIH1 ETS2 FAAH BACH2 MS4A2 RAP2C SIAH1 JMJD7-PLA2G4B SGSM1 PRRT2 TGIF1 ITGA7 TNFRSF8 RN7SL473P KLC3 TRAFD1 LSR HCG4P5 ETV7 GRIN1 MSC ST6GALNAC2 CDC42EP2 MYO7B IL17C EIF2AK2 SIK1 CD69 S100A12 FGF7 CASP4 CHI3L2 PCDH12 CLEC4D VCAM1 GRAMD1A TP53INP2 IL3RA AURKC GGT5 RELA FER1L6 PKD2L1 RDH5 CGNL1 ADM CXCL3 TYW1B CCL1 INSL3 SLC6A12 ZSWIM4 FAM122C FAM35DP ADAMTS7 HLA-DQB1 ZC3H3 GALNT4 ARRDC5 GPR97 RNF24 KCP LYSMD1 TMPRSS9 HNRNPA1P27 IFIT3 IRAK2 ZMYM6NB RN7SL600P CDIPT-AS1 CCDC115 EPS8L2 GIPR CD70 F2RL2 SDS IFITM3 KLK4 LAG3 ARHGAP40 TNFAIP8 HERC5 PTGIR C1orf61 CASP5 GBP5 PLEKHG2 PSMD6-AS2 TNK1 MT-TE C5orf15 IFITM2 ADRB2 CTF1 KIAA1644 TBK1 MT1H KLHDC7B SLFN12L RN7SL559P TMEM41B CMPK2 SLC9B2 BEST4 STEAP4 DIXDC1 EPSTI1 MEIS3P1 ITIH1 ARTN G0S2 FPR2 APOBEC3A ELF3 C17orf66 SESN2 TNFSF10 HLA-J C1QTNF1 TXNP4 PARP10 BMP6 CDKN2D HTR2B GZMA ZC3H12C ABLIM1 ADM2 WHAMMP3 MT1DP MIER1 GPR132 HSH2D NYNRIN LINC00944 ORAI2 NMRK1 FCGR1B CSF3 MT1JP DUSP10 RPLP0P2 APOBEC3D AASS RHCG MX1 CASP1 JHDM1D-AS1 SNX15 CXCR6 CMTR1 SLC35E4 NSUN7 DNAAF1 SERPINB13 PML RALGPS2 BTBD11 MYO5B HHIPL2 SCARF1 PDE4B NLGN2 LINC00426 GRIP2 TSC22D3 SMPDL3A CHAC1 BLK VEGFC GYPC ELOVL7 ISG20 UBXN10 DOK5 SIPA1L1 BCL3 SSPN DPYS ATP1B2 FAM126B SP140 EGR3 ZBP1 WNT5A-AS1 CNNM4 FAM124A EBF1 FAM185BP CCM2L KIAA0040 ARHGAP24 C1R FFAR2 SPRN TMCC3 TBL1X PPIL6 HPD HOXA10 CARS NUAK2 TMEM158 MYH11 ITIH5 CCDC50 MIR29A LIPG PRRX1 GPR171 BAK1 MIR29B1 NAMPTL GAPDHP14 KCNG2 ITGB8 PRKD2 AIM2 CDC42EP5 GAL3ST2 C1QB STARD10 CMYA5 RIMS2 LTK OCSTAMP HS3ST3B1 IFI27 MCC LTA PRDM1 SCYL3 SLCO5A1 KLK10 FAM160A1 SIK3 CLCF1 TMC4 CPA3 SAA2 DCP1A BBC3 EPHA1 FLT3LG KRT8P31 CSRNP1 KCNJ2 CDKN1C FOLR1 WFDC2 COQ10B CLDN7 TPBG TWIST1 IL12RB2 CA12 KIF3C NHS STRA6 GJA3 GCC1 XKR8 CPT1B JMJD7 UNC5C HCP5 TFPI2 TESPA1 ROBO1 PTCHD3P3 CASP10 HSPBAP1 DEGS2 PRICKLE1 KCNQ3 RFFL CCDC88C IL2RA LRTM2 ESRP2 HELZ2 ANGPTL4 EMR1 S100A14 TARM1 ARID5A SLC9A7P1 PIK3R3 HS3ST3A1 IL18RAP OAS2 ABCG1 TNRC18P1 BCL2L14 HCAR1 ERN1 MARCKSL1 PRSS22 C22orf42 MIR4451 RNF19B SECTM1 PDCD1 HPN NRN1 CCDC69 LINC01137 LAD1 NPPA-AS1 KCNJ10 ARID3A A1BG-AS1 PLAC8 MTND4P14 VWA3B RBCK1 ASNS GPR64 BSPRY CTAGE8 TMEM176B ODF3B TICAM2 ADC CXCL11 MX2 CIDECP CYB5R2 ANO7P1 TMEM171 ENPP4 CYB561 TRPC2 ITIH4 RORB DDIT4 FAM65B BEX2 UBD ROR2 DSP MAP3K7CL ZDBF2 NLRC3 SAA1 RAB12 MAP3K10 ANXA2R ADAM11 SERPINE3 HLA-DRB5 MYRF 41886 C2orf62 LINC00322 JADE2 PIGR ITGA1 DND1 DES GSDMD SP4 USP30-AS1 CD7 ART5 CP HERC6 MARVELD2 SVEP1 CRB3 OTUD7B ZHX2 PANX2 CPNE5 CRABP1 KIAA0226 LINC-PINT RCN1P2 MYO1A TCF7L1 GIMAP4 ETS1 INHBE HEATR4 L1TD1 ATXN7 TRIM9 AVIL LINC00528 SHROOM3 FAM20A GIMAP5 GJA5 ITGA10 LINC00336 RETSAT SLC39A14 CADM4 OCLN CSF2 TMEM189 PPP2R5B PPAP2A WEE2-AS1 CHRNA1 CEBPG ABTB1 IL23A DEFB1 HMSD TNFSF13B ARID3B AXL RXFP1 C1QL1 IER3 MDM1 SERPINB2 GPR113 MKX IGFBP6 MVB12B C17orf107 MTND5P14 C1orf210 DDX60 IFIT1 PKN3 SLC6A9 SCARA3 SAMSN1 IL1A POU6F1 FOXP2 ANKRD33 UXS1 SUSD3 C8G WNT3 OR6D1P CES1 USP42 SEC61A2 LRRC43 SLC51A ARMC9 TRAF3IP3 ZDHHC23 KCNMB3 TMEM212-AS1 GTPBP1 MTMR11 MPZL2 BMX PLA2G2D SP110 RAB39A SNAI1 CD96 DMKN DAPP1 LRWD1 CLUHP3 B4GALNT3 FAM26E ICAM3 NCAM1 MT1M JAG2 UBXN10-AS1 TMEM194A WNT5A GOLGA2P5 PTPRS CLIC5 HIVEP1 GPR84 AKAP2 CA15P1 FBXO39 IFRD1 MT1F TMEM255B TIGD3 GRB7 CDCP1 HLA-L LINC00937 TNNT2 MEP1B RFX5 HELB DSG2 P2RY6 GUCY1B2 ZMIZ2 TMEM45A PRLR FAM187B2P FUT2 GDF15 GRASP SLC9A3R2 SUSD2 MTNR1B SERPING1 SWT1 CLC STAP1 LINC00487 TNFAIP6 ZNF688 TTC39A KCNJ2-AS1 SLC44A4 ZFP36 C1S SCN1B SEMA3F C22orf31 PLSCR1 MYEOV ERMN PLA2G4B SLC35F1 LIMK2 SMARCD3 KIAA1211 ODF3L1 CCL14 PRMT5 HLA-G TNFRSF4 CD8A SAA2-SAA4 LATS2 SNHG15 CDC14B C4B FOXD3 TMEM132A LAYN SOCS2 STK4-AS1 IFNG
TABLE-US-00008 TABLE 8 Most expressed genes for M2a macrophages CCL22 TMEM55A DTNA CYP7B1 ST8SIA6 LIPA FCGR2B THBS1 KTN1-AS1 MACROD2 TGM2 CHCHD7 GOLGA8B TAL1 VSTM1 MGAT1 SUOX PDE6G NFATC2 FAM95B1 ANPEP GPD1 ZNF620 SNX32 HLA-DPB2 ANXA11 CR1 SLC6A7 LCA5L PLAT QSOX1 XYLT1 MAP2K6 LINC00526 FAM212B-AS1 PICALM PDE1B CCL17 DNAH7 LRRC46 SEL1L NHSL1 NIPAL1 TMEM169 GABRA4 HADHA GADD45A CCL23 BEX1 KCNK3 H6PD FHL1 C2orf71 ATOH8 FOXC1 ABCC3 TNFRSF11A TTC9 CHDH CYYR1 CTSC TTN-AS1 C9orf9 GLI1 EPPK1 KTN1 COL7A1 ENHO ADAMTS15 NTRK1 HIPK2 IL21R SLC24A4 CACNB4 CA5A G6PD PLA2G4A CXCR2 METAP1D C2orf91 PCM1 C17orf58 PTPRF LRRC1 ARHGAP26-IT1 TBC1D8 TRAF5 TDRKH HS3ST2 SIGLEC8 PFKP CD22 SYT17 ROR1 SEC14L5 SASH3 ACE ADORA3 ANKRD13B POU3F1 SLC7A8 MYC CD1B ALOX15 IL22RA2 SLA AUH ERRFI1 NAT8L SLC9A2 PAM TNFSF13 LINC00607 GUCA1A RHO MAN2A1 PTPN4 ABCC2 VSTM4 AP3B2 ADD3 COL5A3 SMTNL2 ESPNL LINC00484 PTGS1 NDFIP2 COL11A2 FAM198A HCRTR1 PALLD QPRT LHFP CALD1 SRL AKIRIN1 ARL4C SMARCA1 PALD1 SEMG1 AMPD2 CAMK1D LINC01160 CACNA1D ERVFRD-1 PSME4 B3GALTL PRRT3 SYT6 GPC6 HSPH1 ABCG2 SLC25A15 OSBPL10 FST SLC27A3 MIR4435-1HG PTPRU COL4A1 KIAA1462 GALNT12 ECHDC3 EHF PCSK1 LINC01114 CD300LB NUDT16P1 OLFML3 PLCL2-AS1 GRIK1-AS1 EEA1 CISH CCDC85C UCN2 WDR86-AS1 FLT1 CCL4 PBX4 C17orf64 CTTNBP2 FPR3 CLEC4A TREML1 EMBP1 ELFN1 SPOCD1 PDGFC LRRC4 SNCAIP CCL13 FAR2 C3orf18 ZNF827 BACE2 CCL26 SPINT2 MRC1 CD1C MIR621 ZNF705A OSBPL1A CD1D CENPV CH25H MPL LIMA1 SLC47A1 SETBP1 RORC CCDC85A LILRB1 OTUD6B KIAA1024 CR2 ROBO4 C1orf162 EGLN3 PLXNA2 CD1E FHL2 PIK3R1 ADAMTSL4 KIAA1161 LINC00639 TENM4 ARHGAP26 CCDC85B WDR66 DUOXA1 RDH8 FABP4 PPP1R16B CLEC10A LINC00941 DUOX2 SIGLEC10 TPTEP1 MGAT3 SH3BGRL2 HSD3BP5 SUCNR1 C10orf128 ARHGEF28 WNT5B SFRP4 HSPG2 DDO RASAL1 CXCR2P1 GADL1 NR4A3 IFFO2 CELSR2 TINAGL1 SLC39A12 PKD2 NMB HS3ST1 LY86-AS1 RAB3C RCAN1 PVRL1 GAS6 SLC18A2 MAOB EMB CCRN4L CUBN DNASE1L3 DTHD1 ETV3 TMEM130 ITGA11 NEO1 CST2 ACOT7 GATM TSPAN12 UNC80 ELF5 SNX8 EGFL7 MORC4 ANKEF1 FAM27A TTC39B CBR3 FCER2 RGMB FOXD4L1 WFS1 ANKS6 ZBTB8A SLC14A2 FNDC5 RAB32 PECR RGL3 XKR3 RAMP2-AS1 SIDT2 GGTA1P SEMA3G DMD RNU6-853P OPN3 FCGR2C SLC22A16 SLC25A48 SRRM3 NUDT16 FAM212B PHOSPHO1 IL17RB SEMG2 CAMK2D TGFA NEK10 TSPAN7 SERPINB4 SH3PXD2A ADPGK-AS1 PLCB1 NKD1 RCOR2 SCIMP ADAM12 GPRC5C CLEC4G GPR143 SLC26A6 MTUS1 PLEKHA6 BIK C19orf33 TTYH2 CD209 CCL28 DLG3 STK32A MAF HOMER2 UCP3 SIGLEC6 RAMP2 CCNH FAM110B TMEM26 LINC00885 ANKRD20A1 SOWAHC RAMP1 TTN IGHE WISP1 BCL2L11 SLC37A3 SCD5 FOXQ1 LINC01122 EPB41L2 RPIA ZNF365 GLP2R CDX1 ALDH1A2 DACT1 SIGLEC12 PRKCQ MB HN1 HES6 B3GALNT1 GATA3 TBX2 PPFIBP1 PTGFRN SDPR ANKRD55 TMEM200A TLE1 PODXL GCNT3 SNORD125 P2RY12 KMO NAPSB PARM1 HRH4 GAL3ST1 BPNT1 CACNA1G SLC30A4 DNAH3 B3GNT6 PPM1L BCL7A CRB2 DNAI1 HHLA2 MAOA AKAP12 PNPLA3 PLXDC1 TMC3 SIGLEC14 XXYLT1 CDH1 NPAS2 IGHEP1 CTNNAL1 AKAP5 LINC00475 RBM11 PLA2G5 PCSK5 TMTC4 STAMBPL1 SCUBE1 GABRG2 FRMD4A GPR141 SEBOX DUOX1 S1PR5 PELP1 CHN2 VTN PRKCQ-AS1 DUXAP2 NIPA1 MANEAL GALNT18 LIMS2 ABCC13 PPARG MS4A6E ENO1-IT1 DAAM2 CRH KCNK6 MEX3B ASTN2 GPT OR8G3P ODF2 S100A1 NME8 SULT1C2P1 FGL1 HOPX LRP5 LPO TRIM71 SPTSSB BACE1 FARP1 CAPN14 C9orf24 PPP1R14A DHRS11 DIP2C SOGA2 TRPM1 CIB4 FAM126A CD180 MOCOS CHRNA3 KLRG2 CARD9 AIG1 CABLES1 S100B SLC16A9 SYNJ2 GPR35 RUFY4 CCDC151 KRT3 PPFIBP2 NEB CD200R1 FAM19A4 HSD3B1 RABEPK CFP MEST NAALADL2 CFHR1 FGD2 GPR146 ZNF711 SH3TC2 FAM170B SNAI3 RAB33A KCNK5 IL1RL2 SFRP1 MAP1A TAGLN LEPREL2 BAI2 ZNF366 VWCE ACSM5 SLC7A2 GAS2L3 RRS1 STON2 CNGA1 SLC45A4 XPNPEP2 F13A1 TMEM236
TABLE-US-00009 TABLE 9 Most expressed genes for M2c macrophages MMP7 LIN7A CXADR KIAA1211L CHRNA6 TIMP1 GLDN GXYLT2 PCDHGA11 BCYRN1 CD163 NAIP WASF1 PHEX ZFPM2 MARCO MMP8 NPDC1 CRYAB PRL VCAN CD226 DNAH17 AR CHGA SEMA6B PTPRN SPINK1 PVALB LRRC2 SH3PXD2B TSPAN13 PARVA NMNAT2 DNAH17-AS1 PLAU PCOLCE2 CLEC1A SLC16A2 OR13A1 SLC25A19 LIMCH1 TDO2 FAP PRG3 COL22A1 PLOD2 LAMC2 C10orf55 RNF175 SLC12A8 CD300E CCR2 BNIP3P1 PROK2 PROS1 F5 GRPR DDAH1 AWAT2 FPR1 CASC15 CD163L1 BICC1 SNCB PDPN LGI2 FGD1 SPATA20P1 KCNK15 SRPX2 SH2D4A EDNRB C7orf63
[0108] Other suitable markers are described in Marc Beyer et al., “High-Resolution Transcriptome of Human Macrophages,” 7(9) PLOS ONE e45466 (2012) and Fernando O. Martinez et al., “Transcriptional Profiling of the Human Monocyte-to-Macrophage Differentiation and Polarization: New Molecules and Patterns of Gene Expression,” 177 J. Immunol. 7303-11 (2006).
[0109] Although steps S106, S108, and S110 were described in the context of cDNA synthesis and quantitative PCR, one of ordinary skill in the art will recognize that gene expression can be measured using other tools and techniques such as microarrays, RNA Sequencing (RNA-seq), and the like.
[0110] In step S112, a function of one or more of the expression levels of the measured markers is calculated. Various methods are described herein, including in the context of step S1802 in
[0111] For example, gene expression values for five M1 and five M2 markers can be combined into a single number using a linear sum of M1 markers divided by a linear sum of M2 markers, after multiplication of each expression value by a coefficient chosen to enhance or diminish the contribution of its corresponding gene according the following formula.
Here, G.sub.i and G.sub.j are genes associated with M1 and M2 macrophages cultured in vitro, respectively, and α.sub.i and β.sub.j are coefficients obtained using the following methods summarized in Table 10.
TABLE-US-00010 TABLE 10 Summary of methods used for the conversion of gene expression data into a combinatorial M1/M2 score. Name of method Purpose Approach PCA To capture maximum variance between Principal Component Analysis M1 and M2 by magnifying differences of the most important genes Weighted scaling To give greater weight to those genes that Using t-statistics are expressed at very different levels by M1 and M2 Greedy To maximize the difference between M1 Non-linear Optimization and M2 as two distinct populations Mean-centering To equalize contribution of all genes Each gene was normalized to its in vitro expression Linear sum To account for natural differences in the All coefficients set to one level of expression by M1 and M2 IL1B/CD206 To determine the major contributors and Correlation Matrix to reduce the number of genes IL1B/CD163 Utilizing newly discovered importance of The expression of IL1B was M2c in wound healing. normalized to the expression of CD163
[0112] In the first method, α.sub.i and β.sub.j were obtained from principal component analysis (PCA) performed on gene expression data of M1 and M2 macrophages cultured in vitro (“PCA method”). PCA is a mathematical algorithm that is frequently used in gene expression studies for dimensionality reduction and data visualization as discussed in M. Ringner, “What is principal component analysis?” 26(3) Nature Biotechnology 303-04 (March 2008) and M. Parka et al., “Several biplot methods applied to gene expression data,” 138 J. Statistical Planning and Inference 500-15 (2008). In brief, PCA finds new directions in dataset, referred to as principal components (PCs), by capturing most of the variation in dataset. PCs are defined as linear combinations of the original variables. Therefore, the original variables and the transformed data can be visualized in a 2D or 3D vector space built upon the first two or three PCs, respectively.
[0113] In a “weighted scaling” method, αa.sub.i and β.sub.j are chosen to be t statistics obtained from a Student's t-test performed to compare expression of the corresponding gene between M1 and M2 macrophages cultured in vitro. A higher t-statistic indicates a greater degree of difference between M1 and M2 macrophages. Thus, the weighted scaling approach aims to give more weight to those genes with higher levels of significance. Use of t statistics has been reported previously in formulation of linear predictor scores from gene expression data in G. Wright et al., “A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma,” 100(17) P.N.A.S. 9991-96 (2003).
[0114] Alternatively, the “greedy method” seeks α.sub.i and β.sub.j such that the p value of a t-test performed on the combinatorial score of M1 and M2 macrophages cultured in vitro was minimized (“Greedy method”). The greedy method iteratively solves for coefficients such that the difference between M1 and M2 macrophages cultured in vitro was maximized; i.e., the p value of the t test on the combinatorial score of M1 and M2 macrophages cultured in vitro was minimized. For example, one could first set up a t-test comparing the scores, with coefficients of α.sub.i and β.sub.j, of in vitro-derived M1 and M2 macrophage populations, and with an output of the p-value. Any optimization method can then be employed (such as the Solver add-in in MICROSOFT® EXCEL®, available from Microsoft Corporation of Redmond, Wash.) to find α.sub.i and β.sub.j such that the p-value is as small as possible, or that the difference between the scores for the M1 and M2 populations is as large as possible. These optimized coefficients α.sub.i and β.sub.j could then be used in the calculation of the scores for wound data.
[0115] In the “mean-centering” method, the inverse of the mean in vitro expression of each gene is used as its coefficient in the M1/M2 score to equalize contribution of all genes. This approach seeks to account for inherent differences between expression values of different genes and to prevent those genes that are naturally expressed at higher levels from possible masking of the expression of the rest of the genes. This approach was used to scale the expression values for genes, which are expressed at very different levels, to the same level so that one highly expressing gene would not mask all the others, for example.
[0116] For example, CD206 and CCL18 are both M2 markers, meaning their expression is significantly higher in M2 macrophages comparing to M1 macrophages, yet their expression values differ several orders of magnitude. On average, CD206 is expressed 162.84 and 2.25 times relative to house keeping gene GAPDH in M2 and M1 macrophages, respectively. CCL18, however, is expressed 1.07 and 0.02 times relative to house keeping gene GAPDH in M2 and M1 macrophages, respectively. In addition, for example, CCR7 and IL1B are both M1 markers, meaning their expression is significantly higher in M1 macrophages comparing to M2 macrophages, yet their expression values differ several orders of magnitude. On average, CCR7 is expressed 0.33 and 0.02 times relative to housekeeping gene GAPDH in M1 and M2 macrophages, respectively. IL1B, however, is expressed 0.04 and 0.0004 times relative to housekeeping gene GAPDH in M1 and M2 macrophages, respectively.
[0117] Therefore, the following steps can be utilized to process a typical sample (from a wound or other tissue) with exemplary expression values of [CCR7, IL1B, CD206, CCL18]=[0.16, 0.026, 1.99, 0.05] under the mean-centering approach. First, expression values of M1 markers are normalized to the average expression value of those markers in in vitro polarized M1 macrophages, i.e. [0.16/0.33=0.48, 0.026/0.04=0.65]. Second, expression values of M2 markers are normalized to the average expression value of those markers in in vitro polarized M2 macrophages, i.e. [1.99/162.84=0.001, 0.05/1.07=0.046]. Third, the M1/M2 score of the mean-centered values is calculated, i.e. M1/M2=(0.48+0.65)/(0.001+0.046)=24.04.
[0118] In a “linear sum” method, all of the coefficients were set to 1.
[0119] Steps S102-S112 can be repeated again after a period of time in order to assess the change in the ratio of M1 to M2 macrophage activity over time.
[0120] In step S114, the outputs of the functions (e.g., ratios) can be compared. The comparison can be a simple, absolute comparison of calculated ratios, a calculation of the linear rate of change, or can utilize a fold change to measure a ratio of the second ratio to the first ratio. Generally speaking, if the ratios remain substantially steady over the period of time, a transition from M1 to M2 macrophage activity has not occurred and the wound is not healing. If the ratio decreases (i.e., the M2 weighted sum increases relative to the M1 weighted sum or the M1 weighted sum decreases relative to the M2 weighted sum), the transition from M1 to M2 macrophage activity is occurring and the wound will likely heal. Although the degree of change associated with healing and nonhealing wounds will vary between the functions applied to generate the M1 and M2 scores, healing wounds and nonhealing wounds scored using the six functions listed in Table 6 exhibited a MEAN+/−SEM fold changes of 0.29+/−0.07 and 4.09+/−0.83, respectively. Without being bound by theory, it is believed that, regardless of the method used to generate the M1 and M2 scores, the fold change of the M1:M2 ratio over time will be between 0 and 1 for healing wounds and greater than 1 (e.g., between 1 and 20, between 1 and 25, between 1 and 30, and the like) for nonhealing wounds.
[0121] The diagnostic threshold for a particular function can be computed using tools and techniques such as receiver operating characteristic (ROC) curves.
Implementation in Computer-Readable Media and/or Hardware
[0122] The methods described herein can be readily implemented in software that can be stored in computer-readable media for execution by a computer processor. For example, the computer-readable media can be volatile memory (e.g., random access memory and the like) and/or non-volatile memory (e.g., read-only memory, hard disks, floppy disks, magnetic tape, optical discs, paper tape, punch cards, and the like).
[0123] Additionally or alternatively, the methods described herein can be implemented in computer hardware such as an application-specific integrated circuit (ASIC).
[0124] Referring now to
High-Throughput Screening for Identification of M1, M2a, and M2c Macrophages
[0125] Another aspect of the invention provides a high-throughput (HTP) screening assay and system for analyzing healing and wound healing properties, such as identifying macrophage phenotype, predicting healing progression, analyzing response to a stimulus, etc. The HTP assay allows screening of expression transcripts, proteins, protein activity, functional response to a stimulus, etc. of multiple samples.
[0126] The HTP screening assay refers to the analysis of at least two samples simultaneously, iteratively, concurrently, or consecutively. In one embodiment, the number of samples assayed simultaneously is in the range of 1-10,000 samples. In another embodiment, the following ranges of sample number are assayed in the HTP screen: 1-5,000, 1-2,500, 1-1,250, 1-1,000, 1-500, 1-250, 1-100, 1-50, 1-25, 1-10, 1-5, 7,500-10,000, 5,000-10,000, 4,000-10,000, 3,000-10,000, 2,000-10,000, 1,000-10,000, 500-10,000, 100-1,000, 200-1,000, 300-1,000, 400-1,000, 500-1,000, and any other number of samples therebetween.
[0127] The HTP system can include, but is not limited to, measurement devices, robotic pippettors, robotic samplers, robotic shakers, data processors and storage devices, data processing and control software, liquid handling devices, incubators, detectors, hand-held detectors, and the like. For the purposes of automation, the number of samples tested at one time can correspond to the number of wells in a standard plate (e.g., 6-well plate, 12-well plate, 96-well plate, 384-well plate, and the like). The samples can be obtained from a plurality of cells, tissues, individuals, or from a plurality of samples obtained from a single individual.
[0128] In one embodiment, the HTP screening assay permits the analysis and/or prediction of healing or wound healing properties. In another embodiment, the HTP screening assay permits the identification of macrophage phenotype, such as M0, M1, M2a, M2b, M2c macrophage. In yet another embodiment, the HTP screening assay allows for the analysis and/or identification of response to a stimulus, such as a titration of a therapeutic, sensitivity or response to a library of therapeutics, or other agents. In still another embodiment, the HTP screening assay allows for comparison of gene expression signatures.
[0129] In one embodiment, the method includes obtaining one or more measurements as described elsewhere herein and comparing the measurements to analyze and/or predict healing or wound healing properties in the wound. The measurements can be obtained from one or more macrophage phenotype populations. In another embodiment, the method includes obtaining one or more measurements from a wound, a non-wound, different wounds, a healing wound, a non-healing wound, and any combination thereof. In yet another embodiment, multiple measurements are taken from the same sample for comparison. The measurements can be taken in a time course over a defined period of time, seconds, minutes, hours, days, weeks, etc.
[0130] In another embodiment, the method includes obtaining one or more samples and/or preparing the samples for analysis. The HTP screening assay as described herein can utilize techniques previously used in the art to obtain and prepare the samples for analysis. The preparation of the samples can depend on the measurement(s) to be obtained, the type of sample, and any other property dependent on the HTP screen.
[0131] In another embodiment, the method includes analyzing the phenotype of macrophages cultivated in vitro.
[0132] In another embodiment, the method includes comparing the measurements as described elsewhere herein. The HTP screening assay allows for the comparison and output analysis of multiple measurements of the same property, multiple properties, or a combination thereof.
[0133] In one aspect, the HTP screening includes a method of analyzing and/or predicting healing or wound healing properties. In one embodiment, the method includes obtaining one or more measurements of one or more macrophage phenotype populations in a wound and comparing the measurements to analyze and/or predict a healing or wound healing property.
[0134] In another aspect, the HTP screening includes a method of identifying macrophage phenotype in a wound. In one embodiment, the method includes obtaining one or more measurements of one or more macrophage phenotype populations and comparing the measurements to identify macrophage phenotype. In one embodiment, comparing the measurements identifies a primary or predominant macrophage phenotype in the wound, such as M0, M1, M2a, M2b, M2c macrophage.
[0135] In still another aspect, the HTP screening includes a method of differentiating a macrophage phenotype from another macrophage phenotype. In one embodiment, the method includes obtaining one or more measurements and comparing the measurements to differentiate M0, M1, M2a, M2b, or M2c macrophages from the other phenotypes. In this embodiment, an expression profile/signature and/or protein levels are measured and compared to differentiate the macrophage phenotypes. For example, an expression profile/signature includes expression or protein levels of one or more of CD163, MMP7, MMP8, MMP9, MMP12, TIMP1, VCAN, PLAU, PROS1, SRPX2, NAIP, and F5 to differentiate M2c macrophage from one or more other phenotypes. In another example, an expression profile/signature includes expression of SOD2 to differentiate M1 from one or more other phenotypes or expression of CCL22 to differentiate M2a from one or more other phenotypes. Analysis of the expression profile/signature and/or protein levels can also predict a healing or wound healing property, response of macrophage to a stimulus, or other property described herein.
[0136] In yet another aspect, the HTP screening includes a method of analyzing and/or identifying a response of macrophage in a wound or from macrophages cultivated in vitro to a stimulus. In one embodiment, the method includes screening a library of therapeutics or small molecules by analyzing a response of the macrophage exposed to a stimulus, such as therapeutic or small molecule. One or more of the samples can be exposed to stimulus before, during or after measurement. Additional measurements may be obtained on the same samples any time after exposure.
Methods of Predicting Tumor Progression
[0137] Referring now to
[0138] In step S2502, a sample is obtained from the tumor. This sample can be obtained before, during, or after removal of the tumor using various biopsy, surgical, and/or laboratory tools and techniques.
[0139] In step S2504, one or more measurements of macrophage phenotype population are obtained, e.g., using the methods described herein. In one embodiment, measurements of the M1 and M2 macrophage populations (e.g., M1 and M2a, M1 and M2c, and the like) are obtained.
[0140] In step S2506, the measurements are compared to each other. In one embodiment, this comparison is expressed as a ratio as discussed herein.
[0141] In step S2508, a prediction of whether the tumor will metastasize is made based on a result of the comparing step S2506. Without being bound by theory, it is believed that an M1:M2, M1:M2a, or M1:M2c ratio exceeding a threshold that can be determined through analysis of data obtained using a particular panel of biomarkers can be indicative of a tumor that has a low likelihood of metastasis. This prediction can be used to inform clinical decisions regarding what prophylactic measures should be undertaken (if any).
WORKING EXAMPLES
Materials and Methods
Experimental Design
[0142] A panel of genes were selected that were highly indicative of macrophage phenotype using macrophages cultivated and polarized in vitro towards the M1 and M2 phenotypes. Next, a number of algorithms for converting expression data of 10 different genes into a combinatorial score were evaluated. These algorithms were applied to debrided wound tissue obtained from human diabetic foot ulcers over the course of 30 days from the initial visit in order to describe differences in macrophage behavior between healing and nonhealing diabetic wounds and in comparison to healing acute wounds. A publicly available dataset from a longitudinal study of wound healing in acute burn wounds in humans provided in Greco was used as the healing acute wound data.
Preparation and Characterization of Polarized Macrophages in Vitro
[0143] Freshly isolated primary monocytes, purified via negative selection from human peripheral blood mononuclear cells, were purchased from University of Pennsylvania Immunology Core. Monocytes were cultured and polarized in vitro into M1 or M2 macrophages as previously described in Spiller 2014. In brief, monocytes were cultured with monocyte colony stimulating factor (MCSF; 20 ng/ml) for 5 days to differentiate them into macrophages. Then, M1 or M2 polarization was achieved by addition of interferon-gamma (IFNγ; 100 ng/ml) and lipopolysaccharide (LPS; 100 ng/ml) for M1 or Interleukin-4 (IL-4; 40 ng/ml) and Interleukin-13 (IL13; 20 ng/ml) for M2. After 2 days of polarization, RNA was extracted for gene expression analysis of M1 and M2 markers by real time quantitative reverse transcription polymerase chain reaction (qRT-PCR) as in Spiller 2014.
Patient Enrollment
[0144] Thirteen patients with chronic diabetic foot ulcers were recruited from the Drexel University Wound Healing Center in compliance with the study protocol reviewed and approved by the Drexel University Institutional Review Board. Participants were between 50-70 years of age and had at least one open wound on either a foot or lower extremity that had not healed for 8 weeks at the time of enrollment. Patients were excluded if they presented with signs and symptoms of a major infection, abscess, or untreated osteomyelitis. During the study, participants underwent standard wound care procedures determined by the physician, including weekly or biweekly wound debridement, standard length-times-width ruler measurement of wound size, and prescribed topical dressings. Participants were divided into two groups, healing and nonhealing, based on whether their wound was completely healed within 70 day from the initial visit. Only patients who returned for follow-up visits were included in this study in order to facilitate a longitudinal analysis of wound healing. Of these seven patients, three had wounds that completely closed over the course of 70 days and thus were designated “healing,” and four had wounds that did not heal and thus were designated “nonhealing.” Wound Sample Collection
[0145] Participants underwent wound debridement as part of standard wound care regimen during each visit to the clinic. Debrided tissue was immediately collected in RNALATER® solution to stabilize and protect the RNA content of the tissue. Samples were stored in RNALATER® solution at 4° C. overnight as per the manufacturer's suggestion, and were subsequently moved to −80° C. until further analysis by qRT-PCR.
RNA Extraction, Complementary DNA Synthesis, and qRT-PCR
[0146] Wound samples were thawed at room temperature and processed for RNA extraction using TRIZOL® Plus RNA purification kit according to the manufacturer's instructions. Extracted RNA was eluted in 30 μL of RNAse-free water and stored at −80° C. until synthesis of complementary DNA (cDNA) using the APPLIED BIOSYSTEMS® High-Capacity cDNA Reverse Transcription Kit available from Life Technologies. Lastly, quantitative analysis of expression of multiple markers of macrophage phenotype was performed using qRT-PCR with GAPDH as a reference gene, as previously described in Spiller 2014.
Identification of M2c Macrophage Biomarkers
[0147] Referring now to
[0148] Referring now to
[0149] Referring now to
[0150] Referring now to
[0151] Referring now to
[0152] Referring now to
Algorithms
[0153] Using the algorithms described herein, a score was calculated for each sample and plotted over time as fold change over the initial visit. The mean fold change for healing vs. nonhealing wounds was compared at 4 weeks after the initial visit, which is the amount of time recommended for assessment of the effectiveness of therapy and likelihood of healing by the guidelines provided by the Wound Healing Society for the treatment of diabetic ulcers, was assessed.
Conversion of a Panel of Macrophage Markers into a Single Score
[0154] Macrophages are complex and can exist as hybrid phenotypes exhibiting properties of both M1 and M2 macrophages and even other subtypes. Thus, a large number of genes may be required to accurately depict changes in their behavior. Applicant selected 9 genes and compared their expression levels in M1 and M2 macrophages cultivated in vitro as depicted in
[0155] Applicant next explored methods to convert the panel of 9 genes into a single score indicative of the relative M1-M2 character of the macrophages. To accomplish this, Applicant defined, for example, an “M1 over M2 score” as the linear sum of the expression of M1 genes divided by the linear sum of expression of M2 genes, resulting in higher scores for the M1 macrophages and lower scores for the M2 macrophages as depicted in
Macrophage Gene Expression Profile in Human Healing and Nonhealing Diabetic Wounds
[0156] In order to investigate the accuracy of the M1 over M2 score in describing macrophage behavior over time in wounds, Applicant used the Greco burn data set, representing acute or “normal” healing wounds. Conversion of the raw data into the M1 over M2 score allowed for a single number that reflects the macrophage character of the tissue, while simultaneously normalizing the gene expression in such a way that the number would not be sensitive to wound heterogeneity. As depicted in
[0157] Interestingly, when a linearly-summed M1-M2 score was calculated using only those genes that were significantly different between M1 and M2 populations in vitro based on the data depicted in
Profile Analysis and the Confusion Matrix
[0158] To assess the potential utility of the proposed algorithms in a diagnostic assay and to compare their relative performance, profile analysis was performed by fitting a linear curve through the data points to obtain the score for each patient as a function of time. The average fold changes were then compared between healing and nonhealing wounds over time for as long as 4 weeks after the initial visit. To test the hypothesis that healing wounds would show a decrease in the relative proportion of M1/M2 macrophages and to assess the predictive functionality of each method over time, the threshold of the fold change was set to 1. To explore the possibility of predicting healing outcomes earlier than 4 weeks, which is the current clinical standard based on wound size, the true positive rate, true negative rate, positive predictive value, negative predictive value, and accuracy were calculated over time based on the confusion matrix for each method, and the true positive rate was plotted versus the false positive rate (defined as one minus true negative rate) over 1-4 weeks.
Statistical Analysis
[0159] MATLAB® software (available from The MathWorks, Inc. of Natick, Mass.) was used for PCA and curve fitting. The Greedy method was executed in MICROSOFT® EXCEL® using the GRG nonlinear solver. A correlation matrix was plotted using the corrplot package in R software. All other statistical analyses were performed in GRAPHPAD™ PRISM™ 6 (available from GraphPad Software, Inc. of La Jolla, Calif.). Data are shown as mean±SEM and p<0.05 was considered significant. Student's t-test was used to compare M1 and M2 populations in vitro, as well as healing and nonhealing wounds at each time point. Grubb's test was used to identify the outlier in M2 macrophages polarized in vitro, as indicated.
Results
[0160] According to the guidelines provided by the Wound Healing Society, a 40% reduction in wound size after 4 weeks is suggested as a predictor of healing in patients with diabetic ulcers. In order to compare changes in wound size between healing and nonhealing chronic diabetic ulcers, fold changes of wound size relative to the initial visit for 30 days after the first visit were compared as depicted in
[0161] Ten genes were selected and their expression levels compared between the two phenotypes as depicted in
[0162] P-values of the difference between healing and nonhealing wounds at 4 weeks for ratios of single markers of M1 macrophage activity to single markers of M2 macrophage activity are presented in Table 11.
TABLE-US-00011 TABLE 11 P-value of the Difference Between Healing and Nonhealing at 4 Weeks for Ratios of Single Markers P-value of the Difference Between Healing M1 Gene M2 Gene and Nonhealing at 4 Weeks VEGF CCL18 0.368625445 VEGF CD206 0.327864313 VEGF MDC 0.381608604 VEGF PDGF 0.216295875 VEGF TIMP3 0.78106261 VEGF CD163 0.353863894 CCR7 CCL18 0.370353945 CCR7 CD206 0.371951557 CCR7 MDC 0.298055089 CCR7 PDGF 0.32543361 CCR7 TIMP3 0.744078158 CCR7 CD163 0.85056864 CD80 CCL18 0.456329225 CD80 CD206 0.932151157 CD80 MDC 0.163398836 CD80 PDGF 0.657799104 CD80 TIMP3 0.304039946 IL1B CD163 0.40178575 IL1B CCL18 0.252829412 IL1B CD206 0.046251377 IL1B MDC 0.13204204 IL1B PDGF 0.139303456 IL1B TIMP3 0.081109122 IL1B CD163 0.011682275
[0163] P-values of the difference between healing and nonhealing wounds at 4 weeks for ratios of linear summations of one or more markers of M1 macrophage activity to linear summations of a plurality of markers of M2 macrophage activity are presented in Table 12.
TABLE-US-00012 TABLE 12 P-value of the Difference Between Healing and Nonhealing at 4 Weeks for Ratios of Single Markers P-value of the Difference Between Healing and M1 Gene(s) M2 Genes Nonhealing at 4 Weeks IL1B TIMP3 + CD163 0.009924383 VEGF + CCR7 + CD80 CD206 + TIMP3 0.047132339
[0164] In order to further explore the in vitro data and to visualize similarities and differences between M1 and M2 macrophages, principal component analysis (PCA) was performed as depicted in
[0165] With gene expression profile of in vitro polarized M1 and M2 macrophages serving as signature profiles of the two extremes on the spectrum of phenotypes along which macrophages exist, a combinatorial score was defined based on gene expression data of 5 M1 and 5 M2 macrophage genes. The M1/M2 score was then applied to in vitro data of polarized macrophages, healing and nonhealing chronic diabetic ulcers over the course of 4 weeks, and public data from acute healing wounds. In all methods (depicted over
[0166] In order to develop a score that generates a difference between M1 and M2 macrophages by weighing each gene according to its share of the total variation, PCA was used to obtain the linear combinations of M1 and M2 genes. The absolute value of the PC1 coefficients indicates contribution of each gene in capturing most of variance, as well as its ability to classify samples into M1 and M2. The sign of each coefficient, however, is not of interest unless visualization on the PC vector space is intended. Therefore, the absolute values of the PC1 coefficients were used to define the M1/M2 score as depicted in
[0167] Results for the weighted scaling approach are depicted in
[0168] Results for the greedy approach are depicted in
[0169] Results for the mean-centering approach are depicted in
[0170] To address the question of whether those higher-expressed genes are more important contributors to the M1/M2 score, the opposite of mean-centering method was performed by simply summing the contributions from each gene thereby allowing contribution from all the genes analyzed based on their inherent levels of expression as depicted in
[0171] Lastly, with the aim of reducing the number of genes even further and to determine the major M1 and M2 contributors to the predictive M1/M2 score, one M1 and one M2 marker was chosen to define the M1/M2 score. In the case of a large sample size, a number of methodical approaches exist for feature selection. The small sample size prevented implementation of these methods. However, it was hypothesized that out of all possible combinations of M1/M2, those with a highly negatively correlated M1 and M2 genes would most likely yield the best outcome. To this end, a correlation matrix of the in vitro dataset was calculated and is depicted in
[0172] CD163 is another marker for a subtype of M2 macrophages referred to as M2c. IL1B/CD163 was also found to accurately describe healing as depicted in
[0173] In order to assess application of each M1/M2 score in predicting healing outcomes, and to compare the proposed methods to one another, profile analysis was performed on each method and the corresponding true positive rate was plotted versus false positive rate over the course of 4 weeks (
TABLE-US-00013 TABLE 13 Utility of each method as diagnostic compared to wound size. True positive rate, true negative rate, positive predictive value, negative predictive value, and accuracy are reported at 4 weeks after initial visit using 1 as the threshold for M1/M2 score. True Positive True Positive Negative Predictive Negative Rate Rate Value Predictive Accu- (sensitivity) (specificity) (precision) Value racy Wound 66 50 50 66 57 size PCA 100 100 100 100 100 Weighted 100 100 100 100 100 scaling Greedy 100 100 100 100 100 Mean- 100 75 75 100 86 centering Linear 100 100 100 100 100 sum IL1B/ 100 75 75 100 86 CD206 IL1B/ 80 100 100 83 90 CD163
Application of M1/M2 Ratios to in Vitro Testing of Biomaterials
[0174] Referring now to
[0175] To test this hypothesis, undifferentiated macrophages were cultured on the 4 different scaffolds in vitro and analyzed for expression of known M1 and M2 genes after 6 days of culture. Using the proposed “linear sum” method, an M1/M2 score was calculated for each scaffold. Comparison of the M1/M2 score between different materials revealed that in agreement with the hypothesis, IFNg material exhibited an initial increase followed by a decrease in the M1/M2 score between day 2 and day 6, suggesting an effective M1-to-M2 transition of macrophages over time. Without this score, the number of genes analyzed makes the data extremely difficult to interpret as seen in
Application of M1/M2 Ratios to Characterize Macrophage Behavior After Stent Implantation
[0176] Implantation of the stents induces injury at the site of implantation. Considering the key role of macrophages in tissue repair and regeneration, and given the fact that M1 macrophages are dominant in the early inflammatory stages of wound healing and M2 macrophages are dominant in later stages of wound healing such as proliferation and remodeling, it was hypothesized that macrophages would exhibit a natural M1-to-M2 transition after stent implantation in rat arteries. Referring now to
Predictive Power of Initial M1/M2 Ratio
[0177] Referring to
Evaluation of M1/M2 Ratios to Assess Effectiveness of Treatment of Nonhealing Wounds
[0178] Referring now to
Discussion
[0179] Taken together, the findings suggest that healing and nonhealing chronic diabetic ulcers are significantly different with respect to expression of M1 and M2 macrophage markers. Utilizing a number of methods to convert gene expression data into a combinatorial score that reflects the underlying physiology of wound healing, Applicant was able to use gene expression signature of in vitro polarized macrophages to indicate the inflammatory state of the wound. To the best of Applicant's knowledge, this study confirms for the first time that macrophages in human nonhealing diabetic wounds have a persistently elevated M1 character, while diabetic wounds that heal progress through a more natural M1-to-M2 transition. The results demonstrate that M1 and M2 macrophage gene expression signatures have the potential to be used as reference in quantification of wound healing progression, as well as prediction of healing outcomes.
[0180] Wound healing is a complex process and can be divided into several stages: hemostasis, inflammation, proliferation or granulation, and remodeling. Macrophages are key players in the onset and resolution of inflammation and are known to play critical roles in various stages of wound healing. Considering the fundamental role of macrophages in various stages of wound healing, and using the M1-to-M2 transitioning as an indication of tissue regeneration and healing, Applicant aimed to quantify the M1-to-M2 transition in chronic diabetic ulcer over time and to study its association with healing outcomes.
[0181] The conventional method for characterization of macrophage profile in biological tissue is immunohistochemistry (IHC). However, IHC approaches are extremely time-consuming, expensive, and only semi-quantitative. Although some of these limitations have been addressed in flow cytometry methods, practical challenges such as tissue digestion and small sampling volume still remain unresolved.
[0182] Applicant utilized gene expression of the wound tissue. Looking for gene expression enabled Applicant to consider using wound debrided tissue as the source of tissue. Using debrided wound tissue makes embodiments of the invention extremely advantageous over alternative methods that use optical approaches or wound fluid for assessment and quantification of wound healing progression. Such optical or fluid-based methods impose additional burdens both on the patient and on the care provider, whereas wound debridement is a procedure commonly performed as part of the standard wound care regimen. Moreover, optical or fluid-based methods suffer from high variability from patient to patient (not all wounds are exudative especially as they heal) as well as practical challenges such as detection methods. Moreover, such methods are also time consuming and expensive.
[0183] Embodiments of the invention described herein can also be used as a means of quantifying the effectiveness of an experimental therapy, which may be useful in facilitating regulatory approval of novel treatment strategies.
[0184] Applicant set out to convert gene expression data into a combinatorial score based on the underlying biology of M1-to-M2 transitioning of macrophages in the wound, using gene expression profile of in vitro polarized M1 and M2 macrophages. Because of the heterogeneity of debrided wound tissue, the total number of macrophages varies from sample to sample, which necessitates some form of data normalization before raw data can be used. Interestingly, defining a quotient of M1 markers over M2 markers, expression values are essentially normalized as the ratio of genes is independent of total number of cells.
[0185] Applicant then defined an M1/M2 score using six different methods to weigh M1 and M2 genes. In all methods, the M1/M2 score decreases over time in healing chronic diabetic ulcers, whereas it stays constant if not increases in nonhealing chronic diabetic ulcers. Applicant found this difference to be significant at 4 weeks, and already outperform the gold standard of the wound care, which is based on reduction in wound size. Moreover, Applicant found that decreasing trend of M1/M2 score in healing chronic diabetic ulcers resembles the trend observed in acute normal wounds, although with a much slower rate. Unlike wound size, using the M1/M2 score, healing and nonhealing chronic diabetic ulcers were found to be significantly different at 4 weeks, confirming that indeed healing and nonhealing wounds are different with respect to expression of M1 and M2 macrophage markers. Another interesting finding was the results obtained from linear sum method. Despite common belief that without normalization genes with higher expression values would dominate the score and mask the effect of other genes, Applicant found this method to effectively differentiate between healing and nonhealing diabetic wounds. This could be indicative of the importance of the genes with higher expression values in the wound healing process, something that need to be verified in future studies. Similarly, although Applicant found IL1B over CD206 to successfully differentiate between healing and nonhealing wounds at 4 weeks, this finding needs to be verified in independent studies due to possibility of over-fitting because, unlike other methods, definition of the model was based on the data.
[0186] By choosing 9 key genes that describe macrophage phenotype, normalization of M1 to M2 genes, and comparison of the score back to a baseline value, Applicant was able to track macrophage behavior using a method that is insensitive to patient-to-patient variability, wound heterogeneity, and variability in sampling methods. It has been suggested that a more accurate method of assessing wound progression would save an average of $12,600 per patient if ineffective treatments could be discontinued sooner. Remarkably, despite the small sample size of this study, Applicant found highly significant differences between changes in M1 over M2 scores for healing and nonhealing diabetic ulcers, suggesting a potential for its use as a diagnostic.
[0187] Although the sample size did not allow for thorough assessment of the predictive functionality of the proposed methods, Applicant compared the methods over time to one another and to wound size. Given that debrided tissue was used as the tissue source, and since wound debridement is already a standard part of wound care, this approach has great potential to be easily incorporated in wound care regimen. Although preliminary at this point, the results suggest that a small subset of genes can be used to define a macrophage signature, which in return facilitates incorporation of these method as an off-site qRT-PCR based diagnostic assay. Alternatively, with the advent of portable gene sequencing technologies, Applicant envisions this method for real-time measurement of the wound healing progression to complement physician's assessment and discretion in the clinic.
[0188] Taken together, Applicant's results suggest that macrophage gene expression signature may be strongly associated with wound healing progression and has the potential to be used in monitoring wound healing progression and to provide diagnostic information on healing outcomes. Furthermore, these findings shed light on the promise of using macrophage gene expression signatures to explore existing gene expression profiles of wounds, as well as other tissues. Given the importance of macrophages in the function and dysfunction of all tissues, the novel techniques described herein may be useful for the study of macrophage behavior in other disease and injury situations.
Equivalents
[0189] Although preferred embodiments of the invention have been described using specific terms, such description is for illustrative purposes only, and it is to be understood that changes and variations may be made without departing from the spirit or scope of the following claims.
INCORPORATION BY REFERENCE
[0190] The entire contents of all patents, published patent applications, and other references cited herein are hereby expressly incorporated herein in their entireties by reference.