Increased Immune Gene Expression and Immune Cell Infiltration in High-Grade Astrocytoma Distinguish Long-Term from Short-Term Survivors This information is current as of September 30, 2021. Andrew M. Donson, Diane K. Birks, Stephanie A. Schittone, Bette K. Kleinschmidt-DeMasters, Derrick Y. Sun, Molly F. Hemenway, Michael H. Handler, Allen E. Waziri, Michael Wang and Nicholas K. Foreman

J Immunol 2012; 189:1920-1927; Prepublished online 16 Downloaded from July 2012; doi: 10.4049/jimmunol.1103373 http://www.jimmunol.org/content/189/4/1920 http://www.jimmunol.org/

Supplementary http://www.jimmunol.org/content/suppl/2012/07/16/jimmunol.110337 Material 3.DC1 References This article cites 41 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/189/4/1920.full#ref-list-1 by guest on September 30, 2021 Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Increased Immune Gene Expression and Immune Cell Infiltration in High-Grade Astrocytoma Distinguish Long-Term from Short-Term Survivors

Andrew M. Donson,*,† Diane K. Birks,†,‡ Stephanie A. Schittone,*,† Bette K. Kleinschmidt-DeMasters,x,{ Derrick Y. Sun,‡ Molly F. Hemenway,*,† Michael H. Handler,†,‡ Allen E. Waziri,‡ Michael Wang,*,† and Nicholas K. Foreman*,†,‡

Survival in the majority of high-grade astrocytoma (HGA) patients is very poor, with only a rare population of long-term survivors. A better understanding of the biological factors associated with long-term survival in HGAwould aid development of more effective therapy and survival prediction. Factors associated with long-term survival have not been extensively studied using unbiased genome-wide expression analyses. In the current study, gene expression microarray profiles of HGA from long-term survivors were Downloaded from interrogated for discovery of survival-associated biological factors. Ontology analyses revealed that increased expression of immune function-related genes was the predominant biological factor that positively correlated with longer survival. A notable T cell sig- nature was present within this prognostic immune gene set. Using immune cell-specific gene classifiers, both T cell-associated and myeloid linage-associated genes were shown to be enriched in HGA from long-term versus short-term survivors. Association of immune function and cell-specific genes with survival was confirmed independently in a larger publicly available glioblastoma gene expression microarray data set. Histology was used to validate the results of microarray analyses in a larger cohort of long- http://www.jimmunol.org/ term survivors of HGA. Multivariate analyses demonstrated that increased immune cell infiltration was a significant independent variable contributing to longer survival, as was Karnofsky/Lansky performance score. These data provide evidence of a prognostic anti-tumor adaptive immune response and rationale for future development of immunotherapy in HGA. The Journal of Immu- nology, 2012, 189: 1920–1927.

edian survival in high-grade astrocytoma (HGA), con- mutational status have been established as molecular prognostic sisting predominantly of glioblastoma (GBM) and an- factors in adult HGA (4, 5). M aplastic astrocytoma (AA), is 15 mo and 3 y, respectively Gene expression microarray analyses have been used to identify by guest on September 30, 2021 (1, 2). Few robust prognostic factors have been identified in HGA, novel prognostic biomarkers in HGA. This unbiased genome-wide hindering patient care and stratification in clinical trials. Currently approach has the additional benefit of providing insight into the established clinical risk factors include Karnofsky/Lansky perfor- biological mechanisms of tumorigenesis that can be exploited for mance score, a measure of patient well-being that is widely used in the development of more effective therapies. Despite numerous oncology, and age (3). O-6-Methylguanine-DNA methyltransfer- studies that have identified prognostic gene signatures in HGA ase promoter methylation status and isocitrate dehydrogenase 1 using microarray technology, there remains no predictor of survival that has proved robustly reproducible from study to study (6–10). This may be due to the effects of biological heterogeneity inherent *Department of Pediatrics, University of Colorado Denver, Aurora, CO 80045; in HGA combined with the typically limited duration and range of †The Children’s Hospital, Aurora, CO 80045; ‡Department of Neurosurgery, Univer- survival. sity of Colorado Denver, Aurora, CO 80045; xDepartment of Pathology, University of Colorado Denver, Aurora, CO 80045; and {Department of Neurology, University of Although prognosis is poor for the majority of HGA, a small Colorado Denver, Aurora, CO 80045 but discrete subgroup of long-term survivors exists, with 3–5% of Received for publication November 23, 2011. Accepted for publication June 17, GBM patients surviving longer than 3 y. The driving hypothesis 2012. for this study is that a more pronounced and biologically infor- This work was supported by the Morgan Adams Foundation and by National Insti- mative prognostic gene signature could be obtained by gene ex- tutes of Health Grant R01 CA140614-01A1. pression microarray analysis of long-term survivors of HGA. Using The sequences presented in this article have been submitted to the Gene Expression this approach, the current study identified immune function as the Omnibus (http://www.ncbi.nlm.nih.gov/geo/) database under accession number GSE33331. predominant ontology associated with long-term survival in pe- Address correspondence and reprint requests to Andrew M. Donson, Department of diatric and adult HGA. Histological validation in a larger cohort Pediatrics, University of Colorado Denver, Mail Stop 8302, Building RC1, 12800 E. of long-term survivors demonstrated that increased infiltration of 19th Avenue, Aurora, CO 80045. E-mail address: [email protected] immune cells was prognostically favorable. The online version of this article contains supplemental material. Abbreviations used in this article: AA, anaplastic astrocytoma; AIF1, allograft inhibi- tory factor-1; DAVID, Database for Annotation, Visualization, and Integrated Discov- Materials and Methods ery; EPN, ependymoma; FFPE, formalin-fixed paraffin-embedded; GBM, glioblastoma; Patient cohort and sample collection gcRMA, guanine cytosine Robust Multiarray Average; GO, Gene Ontology Project; GSEA, Gene Set Enrichment Analysis; HGA, high-grade astrocytoma; IHC, immuno- Both adult and pediatric HGA samples, including GBM and AA, were histochemistry. included in this study cohort. Using broad age and diagnostic categories substantially increased the number of long-term survivors available to the Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 study. This inclusive approach also has the potential to identify broad www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103373 The Journal of Immunology 1921 prognostic factors in HGA. Multivariate analyses were used to address the variable. GSEA creates a ranked list of genes based on user-selected cri- confounding effect of age, tumor grade, and a number of other potentially teria such as signal-to-noise, t tests, and so forth. For the purposes of this prognostic factors in this cohort. study, genes were ranked on their correlation with survival time. Thus, For the initial discovery cohort (gene expression microarray analysis), unlike DAVID, GSEA takes into account the strength of association rather surgical tumor samples were obtained from 26 patients who presented than simply whether a gene is significantly associated or not. The output between 1990 and 2008 for treatment of HGA at the University Hospital or for both ontology analysis tools is an enrichment score with associated The Children’s Hospital (Aurora, CO) who were diagnosed with GBM or Student t test p values for each GO term. AA according to World Health Organization guidelines (11). Tumor was either snap frozen in liquid nitrogen or placed in RNAlater storage solu- Enrichment analysis of specific immune cell lineage genes tion (Qiagen, Valencia, CA) at the time of resection. Survival data were Gene expression microarray data derived from surgical brain tumor speci- available for all patients in this study, which was conducted in compliance mens contain genes expressed by nontumor cells of the tumor microenvi- with institutional review board regulations (COMIRB 95-500 and 05- ronment. To address this, a technique was developed to identify and predict 0149). Included in this study were three long-term HGA survivors (two specific expression patterns of immune-related cells, based on data from adult GBM and one pediatric GBM), with a median survival of 7.0 y previously isolated cell populations, should they be found within the tumor (range, 6.0–8.5 y). Clinical details are provided in Supplemental Table IA, microenvironment of HGA. Classifier gene lists for discrimination of specific including previously identified survival-associated factors of age at diag- immune cell lineages were created for this study from publicly available nosis, Karnofsky/Lansky performance score, diagnosis, tumor location, Affymetrix HG-U133A gene expression cell and tissue profiles at BioGPS extent of surgery, and therapy received. The microarray analysis control (17). This data set consists of previously acquired gene expression data for cohort consisted of 23 standard survival HGA samples (21 GBM and 2 isolated cell types including multiple immune cell lineages and brain tis- AA) of which 18 were from pediatric patients. The median survival for this sues. The full BioGPS data set was filtered to create a data set consisting of cohort was 12 mo (range, 1–40 mo). two samples each from a range of immune cell lineages: dendritic cells Validation of the results of gene expression microarray analyses was (BDCA4), monocytes (CD14), B cells (CD19), myeloid cells (CD33), Th performed in a publicly available Affymetrix genechip data set used by Downloaded from cells (CD4), NK cells (CD56), and cytotoxic T cells (CD8). Samples from Phillips et al. (10) to identify molecular subclasses of high-grade glioma normal brain tissues were also included as controls and consisted of two (GSE4271). Gene expression profiles were available for 50 primary sam- each of the following: amygdala, caudate nucleus, cerebellum, cerebellum ples of GBM from deceased patients. The median survival of this cohort peduncles, cingulate cortex, fetal brain, globus pallidus, hypothalamus, was 14.7 mo (range, 0.7–74 mo), of which 8 survived longer than 3 y. medulla oblongata, occipital lobe, olfactory lobe, parietal lobe, pituitary, In the secondary validation cohort, formalin-fixed paraffin-embedded pons, prefrontal cortex, spinal cord, subthalamic nucleus, superior cervical (FFPE) material from 33 patients was obtained from archival diagnostic ganglion, temporal lobe, thalamus, trigeminal ganglion, and whole brain. specimen banks of the pathology departments of the University Hospital or Immune lineage classifier gene lists were thus created for each immune cell http://www.jimmunol.org/ The Children’s Hospital. Included in this study were 14 long-term HGA lineage listed above. Fold-change gene expression and associated p values survivors (12 GBM and 2 AA) of which 4 were pediatric. For the purposes (Student two-tailed t test) were generated for each immune cell type profile of this study, long-term survivors were defined as anyone who survived (n = 2) versus all other immune cell and brain tissue profiles combined (n = longer than 5 y with a diagnosis of GBM and 15 y with a diagnosis of AA. 58). In addition, a combined T cell lineage classifier was created by pooling This long-term survival cohort included the same three long-term survivors both helper (CD4) and cytotoxic (CD8) T cell samples and a combined used in the microarray study. The median survival for the HGA long-term myeloid lineage classifier by pooling monocyte (CD14) and myeloid cell survivor cohort was 8.2 y (range, 5.5–16.6 y). The control cohort consisted (CD33) samples. Classifier genes were selected as those with p , 0.05, an of 19 standard survival HGA samples (13 GBM, 6 AA) of which 10 were increase over controls of greater than 15-fold, and filtered to remove du- pediatric. The median survival for this cohort was 10 mo (range, 1–31 mo). plicate gene symbols (Supplemental Table II). The ability of classifier gene lists to distinguish discrete immune cell lineages was strengthened by using

Gene expression microarray analysis by guest on September 30, 2021 only genes that were 15-fold higher than controls. Five micrograms of RNA that had been extracted from tumor was amplified, biotin-labeled (ENZO, Farmingdale, NY), and hybridized to Affymetrix Enrichment of HGA survival-associated genes in gene HG-U133 Plus 2 microarray chips (Affymetrix, Santa Clara, CA). Analysis expression microarray validation data set of gene expression microarray data was performed using Bioconductor HGA gene expression microarray profiles were filtered to generate gene lists functions written in the R programming language (http://www.bioconductor. , org). Microarray data CEL files were background corrected and normalized that were either positively or negatively correlated with survival (p 0.05). using the guanine cytosine Robust Multiarray Average (gcRMA) algorithm Genes that were positively correlated with survival were further filtered to contain only genes that were annotated with “immune response” GO term (12), resulting in log2 expression values. The Affymetrix HG-U133 Plus 2 microarray contains 54,675 probe sets including multiple probe sets for the (6955) (Supplemental Table II). The three HGA survival-associated gene same gene. To reduce error associated with multiple testing and to avoid bias sets were combined with the specific immune cell lineage classifier gene in enrichment statistics from genes containing multiple probe sets, a filtered sets described earlier for GSEA analysis of the Phillips validation data set. list containing a single probe set for each gene that possessed the highest Immunohistochemistry gcRMA expression level across all samples used was created (20,722 genes). The microarray data discussed in this publication have been deposited in the Immunohistochemistry (IHC) was performed on 5-mm FFPE tumor tissue National Center for Biotechnology Information Gene Expression Omnibus sections. Slides were deparaffinized and then subjected to optimal Ag re- (GEO) database (13) and are publicly accessible through GEO Series trieval protocols. Subsequent steps were performed using the EnVision- accession number GSE33331 (http://www.ncbi.nlm.nih.gov/geo/query/ HRP kit (Dako, Glostrup, Denmark) on a Dako autostainer according to acc.cgi?acc=GSE33331). standard protocol. Incubation with primary Ab was performed for 2 h. The GBM patient samples from the Phillips cohort (GSE4271) were analyzed following dilutions of primary Ab were used and applied to the sections for on both Affymetrix HG-U133A and HG-U133B platforms. CEL files specific 1 h: 1:250 allograft inhibitory factor-1 (AIF1) (cat. no.01-1974) from to the two platforms were normalized separately using gcRMA normali- Waco Pure Chemicals (Richmond, VA); 1:2 Prediluted CD4 (SP35; cat. no. zation. These data sets were then combined and filtered to obtain single probe 104R-18) from Cell Marque (Rocklin, CA); and 1:100 CD8 (C8/144B; cat. sets for each gene as described earlier. no. M7103) from Dako. Each of these Abs stained a discrete subpopulation of cells that were distributed throughout the parenchyma of the tumor. Gene ontology analyses Sections were counterstained with hematoxylin. Slides were analyzed with the Olympus BX40 microscope, 340 objective lens and 310 eyepiece Two computer-based ontology analysis tools were used to measure en- (Olympus, Center Valley, CA). Images were captured using an Optronics richment of specific gene functions: Database for Annotation, Visualization, MicroFire 1600 3 1200 camera and PictureFrame 2.3 imaging software and Integrated Discovery (DAVID; http://david.abcc.ncifcrf.gov) (14) and (Optronics, Goleta, CA). Slides were scored in a blinded fashion, with Gene Set Enrichment Analysis (GSEA; http://www.broad.mit.edu/gsea) infiltrating cell abundances measured as the mean number of positive (15). Both analyses were used to assess gene lists for enrichment of genes staining cells per multiple fields of view (number dependent on sample annotated with specific Gene Ontology Project (GO) biological process size) at 3400 magnification. terms (http://www.geneontology.org) (16). Enrichment is defined as more genes that are associated with a particular variable than would be expected Statistical analysis by chance. Briefly, DAVID is a Web-based resource that provides GO term fold-enrichment scores and p values for lists of genes that have already The Kaplan–Meier method was used to estimate the probability of survival been identified by the user as significantly associated with a particular as a function of time. Survival was calculated from the date of initial di- 1922 PROGNOSTIC IMMUNE FACTORS IN HIGH-GRADE ASTROCYTOMA agnosis to the date of death from any cause; patients alive at time of anal- correlated with survival were immune-related (Table I). Cell cycle ysis were censored. Differences between survival curves were analyzed for and M-phase–related ontologies were found to be enriched in genes significance using the log-rank test. Multivariate analysis of the relative that were negatively correlated with survival. These cell cycle on- importance of factors to survival was performed using the Cox propor- tional hazards method. tologies reached greater statistical significance than the positive survival correlate enriched ontologies. GSEA was used as a sec- ondary analysis of functional enrichment in survival-associated gene Results lists. Similar to the DAVID results, GSEA identified three of the Genes positively correlated with long-term survival in HGA highest 10 ontologies in genes that positively correlated with sur- are predominantly immune-related vival as immune-related (Table I). In the reverse analysis, among In the initial gene ontology analysis, DAVID was used to identify those genes that were negatively correlated with survival, the highest enriched biological functions in genes associated with survival as enriched GO terms were cell cycle-related. a continuous variable. As input for DAVID ontology analysis, a list Ontological analysis using GO is dependent on curator-driven of 1106 genes that positively correlated (p , 0.05 estimated by annotation of gene functions based on traceable author statements Pearson correlation test) with survival in HGA (n = 26) was created and/or inferences from electronic annotations. As a consequence, from all 20,722 genes. Using the same approach, a list of 469 genes genes that have not yet been annotated are ignored by the computer- that were negatively correlated with survival was also created. based ontological analyses described above. Therefore, the 1106 DAVID showed that the predominant ontology in genes positively genes positively correlated with survival genes used above were Downloaded from Table I. Ontologic analyses of genes associated with survival in HGA

Enrichment GO Term ID Rank GO Term Annotation Number Fold p Value DAVID ontology analysis Positively correlated with survival http://www.jimmunol.org/ 1 Immune system process 2376 2.69 2.16 3 10232 2 Immune response 6955 3.02 1.61 3 10229 3 Regulation of immune system process 2682 2.84 4.27 3 10214 4 Defense response 6952 2.35 1.50 3 10213 5 Regulation of immune response 50776 3.42 4.76 3 10213 6 Positive regulation of immune system process 2684 3.32 8.00 3 10213 7 Cell activation 1775 3.04 1.34 3 10212 8 Response to external stimulus 9605 1.99 5.26 3 10212 9 Response to wounding 9611 2.35 1.04 3 10211 10 Inflammatory response 6954 2.79 1.51 3 10211 by guest on September 30, 2021 Negatively correlated with survival 1 M phase 279 8.54 1.00 3 10248 2 Cell cycle phase 22403 7.23 3.05 3 10246 3 Cell cycle 7049 5.05 6.54 3 01246 4 Cell cycle process 22402 5.82 4.76 3 01243 5 M phase of mitotic cell cycle 87 9.73 6.47 3 10241 6 Nuclear division 280 9.74 3.17 3 01240 7 Mitosis 7067 9.74 3.17 3 10240 8 Organelle fission 48285 9.36 3.55 3 10239 9 Mitotic cell cycle 278 6.79 1.03 3 01236 10 Cell division 51301 6.51 5.58 3 01227 GSEA ontology analysis Positively correlated with survival 1 Generation of precursor metabolites and energy 6091 1.94 0.0112 2 Immune system process 2376 1.92 0.0421 3 Immune response 6955 1.90 0.0521 4 Response to drug 42493 1.89 0.0137 5 Protein oligomerization 51295 1.88 0.0056 6 Humoral immune response 6959 1.84 0.0264 7 Regulation of protein secretion 50708 1.83 0.0041 8 Positive regulation of MAP kinase activity 43407 1.78 0.0161 9 Monocarboxylic acid metabolic process 32787 1.74 0.0372 10 Response to chemical stimulus 42221 1.73 0.0609 Negatively correlated with survival 1 Protein amino acid autophosphorylation 46777 2.07 0.0020 2 Protein autoprocessing 16540 1.99 0.0040 3 Establishment and/or maintenance of chromatin 6325 1.75 0.0565 architecture 4 Microtubule cytoskeleton organization and biogenesis 226 1.72 0.0362 5 Cell cycle process 22402 1.68 0.0971 6 Cell cycle phase 22403 1.67 0.0937 7 Chromosome organization and biogenesis 51276 1.65 0.0917 8 mRNA processing 6397 1.65 0.0891 9 DNA replication 6260 1.63 0.0886 10 DNA recombination 6310 1.62 0.0779 The top 10 enriched ontologies for positive and negative correlates of survival using DAVID and GSEA ranked according to p value and normalized enrichment score, respectively. The Journal of Immunology 1923 manually reviewed to identify those with a documented predom- cance (p , 0.1). B cells (CD19) and dendritic cells (BDCA4) were inant role in specific immune mechanisms. This process identified the least enriched cell types. Conversely, none of these immune 19% (205 of 1106) genes that were related to immune function. The cell lineage classifiers were enriched in the genes associated with results of this analysis, with genes listed and categorized into shorter survival in HGA. subgroups according to their documented role in specific immune mechanisms, are provided in Supplemental Table IB. Notably, 99 Validation of survival-associated immune response gene immune-related genes (48%) were found that did not have an expression signature annotation of “immune response” (GO term 6955), including a After primary analysis of the in-house HGA data set, a publicly number of genes of key CD immune markers such as CD2, CD3, available gene expression microarray data set (Phillips; GSE4271) CD33, and CD40, underscoring the importance of manual review was used to confirm our findings (10). The patient cohort represented of gene lists as performed here. by this data set contained 50 primary GBM that were from deceased A number of notable genes associated with the adaptive immune patients of which 8 survived longer than 3 y (range, 3.5–6.2 y). system were identified in those genes positively correlated with This data set was subjected to the same ontology and immune longer survival in HGA. In particular a large number of genes cell lineage analyses as described earlier. Expression of all 19,110 known to be expressed by T cells (CD3D, CD3E, CD3G, CD8B, HG-U133A and U133B best-expressed probe sets in the Phillips TRAC, TRAT1, VAV1, and ZAP70) were shown to be associated data set were correlated (Pearson) with survival as a continual with long-term survival. Additionally, multiple components of the variable. This approach identified 1391 genes that were positively innate immune system, including genes associated with microglia/ correlated with survival (p , 0.05). Ontology analysis of these macrophages (AIF1, CD68, CD86, CIITA,HLA-DOA, HLA-DQB2, genes using DAVID identified immune response as the most highly Downloaded from HLA-DRB1, HLA-DRB6, NOD2), were present in this prognostic enriched biological process associated with long-term survivor immune gene set. Several TLRs were also associated with long-term GBM samples (Table III). Similar to the DAVID results, GSEA survival (TLR 2, 3, 5, 6, 7, and 8). identified two of the highest 10 ontologies in genes that positively Known negative regulators of immune function in gliomas were correlated with survival as inflammation and immune-related examined to identify any association with shorter survival in HGA. (Table III). These results match the results of the in-house analy-

IL-10 was found to be associated with longer survival (p = 0.0127), sis thus supporting the hypothesis that host immunity contributes http://www.jimmunol.org/ and no significant survival association was observed for IL-13, significantly to long-term survival in HGA. In the converse anal- FOXP3, STAT3, TGFB1, or TGFB2 gene expression. ysis, 1181 genes were significantly correlated with shorter survival. DAVID and GSEA analyses demonstrated that the predominant Lymphoid lineage-specific genes are enriched in HGA ontologies associated with short-term survival were mitosis-related, long-term survivors again matching the results of the in-house data set analysis (Table To clarify which immune cell types might account for the immune III). Enrichment of genes associated with specific immune cell function enrichment identified earlier, genes uniquely specific to lineages in the Phillips data set was measured using GSEA as de- each immune cell lineage were identified using existing publicly scribed earlier, but with the addition of HGA positive and nega- available gene expression data. Gene expression microarray data tive correlate gene sets generated from our in-house data set (n = by guest on September 30, 2021 obtained from purified immune cell lineages were used to created 1106 and n = 469, respectively). A third gene set was created by classifier gene sets. These classifier gene sets were then used in using only “immune response” (GO term 6955) annotated genes combination with GSEA to indicate the presence of specific types that were significantly positively correlated with survival (n = 117). of infiltrating immune cells in HGA from long-term survivors. This Seven of the nine immune cell lineages were shown to be associ- analysis revealed that lymphoid cell lineage classifiers Th cell ated with long-term survival (Supplemental Table IC). In contrast (CD4), combined T cell (CD4 and CD8), and NK cell (CD56) were to the HGA in-house data set, monocyte and myeloid linage- statistically significantly enriched (p , 0.05) in the long-term associated genes were significantly enriched in long-terms survi- survivor-associated genes (Table II). Cytotoxic T cells (CD8) and vors in the Phillips data set. HGA positive and negative survival- myeloid lineages (combined myeloid lineage [CD14 and CD33], correlate genes were significantly enriched in the corresponding myeloid cell [CD33], and monocyte [CD14]) approached signifi- Phillips data set survival-correlated genes. Of note, a higher en- richment was seen in “immune response” annotated HGA positive correlates than in the full HGA positive correlate gene set. Table II. Immune cell lineage enrichment analysis of genes correlated Immune-related genes positively correlated with survival in the with survival in HGA Phillips were compared with those identified in the HGA data set. Fifteen “immune response” GO term annotated genes were shared Enrichment between the two data sets. These included myeloid-lineage ex- pressed genes GPR183, IRF8, IL6R, KYNU, NCF4, STXBP2, Rank Immune Cell Lineage NES p Value TNFAIP8L2, and TNFSF13 (APRIL) based on BioGPS, suggest- Positively correlated with survival ing a common myeloid-lineage enrichment in long-term survivors 1 CD4 Th cell 2.02 0.0197 from the two data sets. Those immune-related genes that distin- 2 T cell lineage (CD4 and CD8 combined) 1.96 0.0430 guished the Phillips from the HGA data set included a number of 3 CD56 NK cell 1.95 0.0406 4 Myeloid lineage (CD14 and CD33 1.74 0.0559 Fc-receptors (CD16A and B, CD32, CD64) and MHC class I HLA- combined) A. MHC class II and T cell-associated genes that had been iden- 5 CD8 cytotoxic T cell 1.65 0.0986 tified in the HGA gene set were absent from the Phillips data set. 6 CD33 myeloid cell 1.65 0.0824 Negative regulators of immune function were studied to identify 7 CD14 monocyte 1.61 0.0936 8 CD19 B cell 1.41 0.1684 any association with shorter survival in the Phillips data sets. IL-4 9 BDCA4 dendritic cell 1.32 0.2119 was found to be associated with shorter survival (p = 0.0175) Negatively correlated with survival supporting a link between TH2 immune response polarization and No gene sets were enriched a shorter survival. Conversely, FOXP3 and TGFB1 were signifi- NES, Normalized enrichment score. cantly associated with long-term survival ( p = 0.0203 and p = 1924 PROGNOSTIC IMMUNE FACTORS IN HIGH-GRADE ASTROCYTOMA

Table III. Ontologic analyses of genes associated with survival in GBM in the Phillips (GSE4271) data set

Enrichment GO Term ID Rank GO Term Annotation Number Fold p Value DAVID ontology analysis Positively correlated with survival 1 Immune response 6955 1.57 2.45 3 1024 2 Defense response 6952 1.58 4.79 3 1024 3 Regulation of transcription from RNA polymerase II promoter 6357 1.51 6.12 3 1024 4 Enzyme linked receptor protein signaling pathway 7167 1.70 0.00230 5 Vascular development 1944 1.80 0.00352 6 Cellular defense response 6968 2.91 0.00392 7 Taxis 42330 2.02 0.00464 8 Chemotaxis 6935 2.02 0.00464 9 Blood vessel development 1568 1.78 0.00494 10 Positive regulation of protein kinase activity 45860 1.81 0.00569 Negatively correlated with survival 1 DNA metabolic process 6259 3.10 6.67 3 10223 2 Cell cycle 7049 2.43 6.14 3 10219 3 DNA replication 6260 4.17 7.75 3 10217 216 4 Cell cycle process 22402 2.59 2.95 3 10 Downloaded from 5 Response to DNA damage stimulus 6974 2.98 2.14 3 10215 6 Cell cycle phase 22403 2.77 3.34 3 10214 7 DNA repair 6281 3.20 5.34 3 10213 8 M phase 279 2.92 5.66 3 10213 9 Mitotic cell cycle 278 2.73 2.33 3 10212 10 Cell division 51301 2.86 4.11 3 10211 GSEA ontology analysis Positively correlated with survival http://www.jimmunol.org/ 1 Peptidyl tyrosine modification 18212 2.20 ,0.001 2 Peptidyl tyrosine phosphorylation 18108 2.11 ,0.001 3 Regulation of peptidyl tyrosine phosphorylation 50730 2.08 ,0.001 4 Regulation of protein amino acid phosphorylation 1932 1.98 ,0.001 5 Inflammatory response 6954 1.93 ,0.001 6 Positive regulation of protein amino acid 1934 1.86 ,0.001 phosphorylation 7 JAK–STAT cascade 7259 1.81 0.00361 8 Defense response 6952 1.80 ,0.001 , 9 Small GTPase mediated signal transduction 7264 1.78 0.001 by guest on September 30, 2021 10 Ras protein signal transduction 7265 1.77 0.00167 Negatively correlated with survival 1 Cell cycle process 22402 2.75 ,0.001 2 Cell cycle phase 22403 2.74 ,0.001 3 M phase 279 2.64 ,0.001 4 Mitotic cell cycle 278 2.60 ,0.001 5 DNA-dependent DNA replication 6261 2.59 ,0.001 6 Cell cycle checkpoint 75 2.47 ,0.001 7 Cell cycle 7049 2.46 ,0.001 8 Meiotic cell cycle 51321 2.45 ,0.001 9 Chromosome segregation 7059 2.41 ,0.001 10 Mitosis 7067 2.39 ,0.001 The top 10 enriched ontologies for positive and negative correlates of survival using DAVID and GSEA ranked according to p value and normalized enrichment score, respectively.

0.0143, respectively). No significant survival association was ob- myeloid/monocyte function and lineage-specific genes with long- served for STAT3, IL-10, IL-13, or TGFB2 gene expression. term survival in HGA also implied the presence of infiltrating microglia/macrophages in these tumors. To confirm the presence In a histological validation cohort of HGA, immune cell of microglia/macrophages, the most abundant myeloid-linage cells infiltration and Karnofsky/Lansky performance score in the CNS, immunostaining for AIF1 was performed (Fig. 1B). contribute to long-term survival AIF1, also known as Iba1, is a specific marker for the microglia/ On the basis of results of microarray analysis of long-term survivors macrophage population in the CNS (18). Frequency of infiltration of HGA, it was hypothesized that increased tumor infiltration of of specific immune cells was scored as number of stained cells per microglia and/or T cells is associated with longer survival. To val- 3400 field of vision. Immunostaining representing high and low idate this hypothesis, IHC was used to measure immune cell infil- T cell and microglia/macrophage infiltration is provided in Sup- tration in a larger cohort of long-term survivors of HGA. The patient plemental Fig. 1. Additional known survival-associated factors— cohort for this study included 14 long-term survivors (median sur- age at diagnosis, Karnofsky/Lansky performance score, tumor grade, vival, 8.5 y; range, 5.5–17 y) and 19 patients whose survival was in location, extent of surgery, and therapy received—were included in the typical range (median survival, 9 mo; range, 1–31 mo). Immu- the survival analyses to address any potential confounding factors. nostaining of FFPE diagnostic samples was performed for CD8 and Univariate Kaplan–Meier analysis was first used to identify factors CD4, specific markers for cytotoxic T cells and Th cells, respec- that were significantly associated with survival. Factors addressed tively (Fig. 1A). The association of a significant number of putative in this study consisted of noncontinuous variables (tumor grade, The Journal of Immunology 1925

FIGURE 1. Representative histology of (A) greater than median tumor infiltration of cytotoxic T cells (CD8; red) and Th cells (CD4; brown) in long-term survivor HGA11, and (B) .75th percentile tumor in- filtration of microglia/macrophages (AIF1; brown) in long-term survivor HGA12. IHC was performed using FFPE tumor sections with hematoxylin counterstaining (original magnification 3400). (C) Kaplan–Meier sur- vival analysis of combined immune cell infiltration. High combined immune cell infiltration was defined as greater than median cytotoxic T cell or Th cell or .75th percentile microglia/macrophage infiltration. Thirty-three HGA samples were used. Downloaded from http://www.jimmunol.org/ adult/pediatric, radiation treatment, temozolomide treatment, supra- a significant correlation with each other by either Pearson or tentorial location, thalamic location, and gross total resection) and Spearman approaches. Because these factors were not therefore continuous variables that were divided at the mean value into high/ independent, it was considered appropriate also to combine them low [Th cell (CD4) infiltration, cytotoxic T cell (CD8) infiltration, (microglia/macrophage .75th percentile or cytotoxic T cell greater microglia/macrophage (AIF1) infiltration, age at diagnosis, and than mean or Th cell greater than mean) into one measure as a Karnofsky/Lansky performance score]. Factors significantly asso- single variable. By Kaplan–Meier analysis, the combined immune ciated with survival were performance score greater than the mean of cell variable was significantly associated with survival (p =0.0066) by guest on September 30, 2021 80 (p = 0.00387), cytotoxic T cell infiltration greater than the mean (Fig. 1C). (p = 0.0154), and Th cell infiltration greater than mean (p = 0.0272). Combined immune cell and individual immune cell variables Mean cytotoxic and Th cell infiltration was 1.26 (range, 8.2–0) and were modeled with Karnofsky/Lansky performance score and 1.98 (range, 37–0) cells per 3400 field of view, respectively. As- analyzed using Cox proportional hazard (Table IV). This analysis sociation of survival with greater than mean microglia/macrophage showed that when combined with performance score, each of the infiltration showed a trend toward significance (p = 0.089). To in- immune variables apart from Th cell infiltration were independent vestigate further the influence of microglia/macrophage infiltration variables for survival. The best model appeared to consist of per- on survival, the cutoff for high infiltration in this cell population was formance and combined immune cell infiltration, with both factors made more stringent by using .75th percentile, which showed a contributing significantly to longer survival (p = 0.00268 and significant association with survival (p = 0.0146). The 75th per- 0.00772, respectively) (Table IV). centile score for microglia/macrophage infiltration was 100 (range, 220–0) cells per 3400 field of view. None of the remaining factors Discussion showed a significant association with survival. HGAs are heterogeneous but almost uniformly fatal with only Significant variables were next subjected to multivariate anal- a small percentage of long-term survivors. An unbiased genome- ysis. Each of the three immune cell population infiltrations showed wide microarray approach, with validation in an independent data

Table IV. Multivariate survival model of Karnofsky/Lansky performance score and immune cell infiltration using Cox’s proportional hazards regression analysis

Model Variable Hazard Ratio 95% Confidence Interval p Value Karnofsky/Lansky performance and microglia/macrophage infiltration (AIF1) Performance . mean 0.216 0.086–0.538 0.00101 AIF1 . 75th percentile 0.138 0.031–0.618 0.00963 Karnofsky/Lansky performance and Th cell infiltration (CD4) Performance . mean 0.402 0.169–0.957 0.0394 CD4 . mean 0.202 0.026–1.59 0.1282 Karnofsky/Lansky performance and cytotoxic T cell (CD8) Performance . mean 0.317 0.132–0.762 0.0102 CD8 . mean 0.206 0.047–0.913 0.0375 Karnofsky/Lansky performance and combined immune (AIF1 . 75th percentile or CD4 . mean or CD8 . mean) Performance . mean 0.253 0.103–0.620 0.00268 Combined immune 0.183 0.0523–0.638 0.00772 1926 PROGNOSTIC IMMUNE FACTORS IN HIGH-GRADE ASTROCYTOMA set, was used to analyze this broad diagnostic class to determine This report potentially illustrates just such a response, underscoring whether increased enrichment of immune genes and cells was the value and potential impact of these findings. associated with better patient survival as had been seen in a similar The data described in the current study can provide some study of another brain tumor, ependymoma (19). The results of this mechanistic detail of a putative antitumor immune response in the study showed that this was indeed the case, with immune function human CNS. Biological inferences can be gained by examining ontologies, immune cell lineage-specific genes, and infiltrating the specific immune function genes and cell types that are asso- immune cell frequency being significantly associated with long- ciated with HGA long-term survivors and by comparison with the term survival in HGA. These data provide additional support to good-outcome-associated immune genes found in other CNS tu- the theory that host immunity can control tumor growth in a sub- mors. An earlier study by our group identified a similar association set of HGA and provide a potential explanation for instances of of immune gene and cell enrichment in ependymoma (EPN), a long-term survival in this otherwise uniformly and rapidly fatal glial tumor of childhood (19). EPN commonly has a less aggres- disease. sive phenotype than HGA, with a recurrence rate of ∼50%. A A number of large-scale gene expression clustering studies have number of T cell-specific genes are seen in the HGA survival-asso- identified sets of genes reportedly predictive of prognosis in HGA. ciated genes that were not identified in EPN outcome-associated Rich et al. (6) identified a migration/invasion molecular signature immune genes. This includes multiple components of the T cell that was associated with a worse survival in GBM. Using an ag- immune synapse (CD2, CD3D, CD3E, CD3G, CD8B, TCRGC2, glomerative approach, Phillips et al. (10) defined three molecular TRBC1, TARP, and TRAT1), cytotoxic mediators (granzyme B, subgroups of HGA (proliferative, proneural and mesenchymal), H, K, and M), and key signaling molecules restricted to activated one of which conveyed a better prognosis. However, the gene sets T cells (CD69, ZAP70, CARD11, and VAV1). The presence of Downloaded from identified in these studies share few genes in common. To address CD8B and cytotoxic mediators, in addition to the above-listed this problem, Zhang et al. (9) combined a number of independent T cell-specific genes, implies that cytotoxic T cells are an im- data sets and identified a 23-gene classifier that outperformed portant cellular correlate of survival in HGA. This theory is sup- previously reported classifiers from the independent cohorts. Of ported by multivariate analysis of a larger cohort of HGA, which note, 2 of 23 genes were immune-related, the remaining 21 being showed that frequency of tumor-infiltrating cytotoxic T cells

related to cellular proliferation. (CD8) was significantly associated with longer survival. http://www.jimmunol.org/ The majority of studies aimed at identifying prognostic factors in Some of the immune genes upregulated in long-term survival HGA have used standard survival patient cohorts that contain few HGA are associated with specific T cell functions. Prior studies long-term survivors. The narrow survival range associated with have demonstrated that the majority of T cells within most human these cohorts presents difficulty in assigning good and bad out- GBM are Th2-biased; however, these data were not generated from comes. The inclusion of a large proportion of long-term survivors in long-term survivors (22). Polarization of infiltrating T cells in the current study, although not representative of natural survival long-term survivors to the Th1 phenotype is implied by the pres- distribution in HGA, provides a wider survival range and thus more ence of GIMAP4, HAVCR2 (TIM3), and STAT4 (23-25). Together, clearly delineated good and bad outcomes. These more clearly these data provide preliminary evidence that, beyond the simple defined survival cohorts allow for stronger inferences to be made presence of an immune infiltrate, the phenotype and function of by guest on September 30, 2021 from associated gene expression analyses. To explore this hy- that infiltrate may influence survival in HGA. This conclusion is pothesis, one other study, in addition to the current study, com- consistent with the report by Galon et al. (26) demonstrating that pared primary GBM from seven long-term survivors (.2 y) with the type (specifically Th1), density, and location of immune cells primary GBM from 13 short-term survivors (,9 mo) (20). Using within human colorectal tumors predict clinical outcome better a genomic approach, they identified a prognostic fingerprint of 43 than current staging criteria. A number of histological studies have genes. Consistent with the current study, six survival-associated been performed to assess the role of the host immune system in immune genes were identified (CD34, IGHG1, IL13RA1, IL17, control of HGA growth and patient outcome. Three studies iden- IL22, and SERPING1) of which two were also associated with tified a positive correlation of lymphocyte infiltration with outcome longer survival in the current analysis (IL17 and SERPING1). (27–29). The results of three other studies, however, showed no Data from the current study, which identified apparently re- correlation (30, 31) or a negative correlation (32). It should be ciprocal prognostic cell cycle and immune functions, support the noted that these studies relied on morphometry to identify immune theory that an equilibrium may exist between immune system and infiltrates and unlike the current study were not able distinguish tumor, known as immunoediting (21). A shift in this equilibrium subtypes of T cells nor identify microglia/macrophages. can result in either complete elimination of tumor by the immune Much attention in the study of the interaction of HGA with the system or alternatively escape from immune control leading to host immune system has been devoted to immunosuppressive fac- tumor progression. In long-term survivors of HGA, therapeutic tors that have shown to contribute to a poor prognosis. A number intervention (surgery, radiation, and/or chemotherapy) may have of soluble factors released by HGA, most notably TGF-b, IL-4, skewed this equilibrium in favor of the host immune system, IL-10, and IL-13, have been shown to be immunosuppressive (33– resulting in clinically significant tumor control. Opposing prog- 35). The current study demonstrated that IL-4 expression corre- nostic cell cycle and immune function gene sets have previously lated with shorter survival, but no other immunosuppressive fac- been observed in GBM and ependymoma (9, 19). tors were shown to be associated with a worse outcome. More That genes related to immune functions are associated with long- recently, attention has been focused on a number of specific im- term survivors of HGA implies that the host immune system may be mune cells shown to contribute to an immunosuppressive milieu in involved with control of tumor growth in these cases. Further the tumor. These include regulatory T cells, myeloid-derived sup- details of the specific immunological mechanism through which pressive cells, and, most recently, neutrophils (36–40). STAT3- this occurs is of significant interest, potentially providing a basis for inducedsignalingingliomashasbeenimplicatedininduction the design of more effective immunotherapeutic strategies. A ra- of inflammation and mesenchymal transformation resulting in a tional approach to improvement of immunotherapeutic approaches poor prognosis (41, 42). No correlation of these immunosup- would be to design strategies based on data taken from direct pressive cells and signaling pathways with a poor prognosis was clinical studies of human host anti-CNS tumor immune responses. observed in either data set in the current analysis. The Journal of Immunology 1927

The findings of the current study provide evidence that despite 18. Yamada, M., K. Ohsawa, Y. Imai, S. Kohsaka, and S. Kamitori. 2006. X-ray structures of the microglia/macrophage-specific protein Iba1 from human and the immunosuppressive nature of HGA, immunologically mediated mouse demonstrate novel molecular conformation change induced by calcium control of tumor growth may arise in some instances, resulting in binding. J. Mol. Biol. 364: 449–457. prolonged survival. Further characterization of the mechanism of 19. Donson, A. M., D. K. Birks, V. N. Barton, Q. Wei, B. K. Kleinschmidt- Demasters, M. H. Handler, A. E. Waziri, M. Wang, and N. K. Foreman. 2009. immune control of HGA in long-term survivors is warranted and Immune gene and cell enrichment is associated with a good prognosis in may aid in the design of more effective immunotherapies. ependymoma. J. Immunol. 183: 7428–7440. 20. Marko, N. F., S. A. Toms, G. H. Barnett, and R. Weil. 2008. Genomic expression patterns distinguish long-term from short-term glioblastoma survivors: a pre- Acknowledgments liminary feasibility study. Genomics 91: 395–406. We thank Patsy Ruegg at IHCtech for assistance with IHC and Liza 21. Swann, J. B., and M. J. Smyth. 2007. Immune surveillance of tumors. J. Clin. Litzenberger for photographic expertise. Invest. 117: 1137–1146. 22. Waziri, A., B. Killory, A. T. Ogden, III, P. Canoll, R. C. Anderson, S. C. Kent, D. E. Anderson, and J. N. Bruce. 2008. Preferential in situ CD4+CD56+ T cell Disclosures activation and expansion within human glioblastoma. J. Immunol. 180: 7673– The authors have no financial conflicts of interest. 7680. 23. Monney, L., C. A. Sabatos, J. L. Gaglia, A. Ryu, H. Waldner, T. Chernova, S. Manning, E. A. Greenfield, A. J. Coyle, R. A. Sobel, et al. 2002. Th1-specific cell surface protein Tim-3 regulates macrophage activation and severity of an References autoimmune disease. Nature 415: 536–541. 1. Stupp, R., W. P. Mason, M. J. van den Bent, M. Weller, B. Fisher, 24. File´n, J. J., S. File´n, R. Moulder, S. Tuomela, H. Ahlfors, A. West, P. Kouvonen, M. J. Taphoorn, K. Belanger, A. A. Brandes, C. Marosi, U. Bogdahn, et al; S. Kantola, M. Bjo¨rkman, M. Katajamaa, et al. 2009. Quantitative proteomics European Organisation for Research and Treatment of Cancer Brain Tumor reveals GIMAP family proteins 1 and 4 to be differentially regulated during and Radiotherapy Groups; National Cancer Institute of Canada Clinical Trials human T helper cell differentiation. Mol. Cell. Proteomics 8: 32–44. Group. 2005. Radiotherapy plus concomitant and adjuvant temozolomide for 25. Nishikomori, R., T. Usui, C. Y. Wu, A. Morinobu, J. J. O’Shea, and W. Strober. Downloaded from glioblastoma. N. Engl. J. Med. 352: 987–996. 2002. Activated STAT4 has an essential role in Th1 differentiation and prolif- 2. Prados, M. D., C. Scott, W. J. Curran, Jr., D. F. Nelson, S. Leibel, and S. Kramer. eration that is independent of its role in the maintenance of IL-12R beta 2 chain 1999. Procarbazine, lomustine, and vincristine (PCV) chemotherapy for ana- expression and signaling. J. Immunol. 169: 4388–4398. plastic astrocytoma: A retrospective review of radiation therapy oncology group 26. Galon, J., A. Costes, F. Sanchez-Cabo, A. Kirilovsky, B. Mlecnik, C. Lagorce- protocols comparing survival with carmustine or PCV adjuvant chemotherapy. Page`s, M. Tosolini, M. Camus, A. Berger, P. Wind, et al. 2006. Type, density, and J. Clin. Oncol. 17: 3389–3395. location of immune cells within human colorectal tumors predict clinical out- 3. Curran, W. J., Jr., C. B. Scott, J. Horton, J. S. Nelson, A. S. Weinstein, come. Science 313: 1960–1964.

A. J. Fischbach, C. H. Chang, M. Rotman, S. O. Asbell, R. E. Krisch, et al. 1993. 27. Brooks, W. H., W. R. Markesbery, G. D. Gupta, and T. L. Roszman. 1978. http://www.jimmunol.org/ Recursive partitioning analysis of prognostic factors in three Radiation Therapy Relationship of lymphocyte invasion and survival of brain tumor patients. Ann. Oncology Group malignant glioma trials. J. Natl. Cancer Inst. 85: 704–710. Neurol. 4: 219–224. 4. Hegi, M. E., A. C. Diserens, T. Gorlia, M. F. Hamou, N. de Tribolet, M. Weller, 28. Palma, L., N. Di Lorenzo, and B. Guidetti. 1978. Lymphocytic infiltrates in J. M. Kros, J. A. Hainfellner, W. Mason, L. Mariani, et al. 2005. MGMT gene primary glioblastomas and recidivous gliomas. Incidence, fate, and relevance to silencing and benefit from temozolomide in glioblastoma. N. Engl. J. Med. 352: prognosis in 228 operated cases. J. Neurosurg. 49: 854–861. 997–1003. 29. Bo¨ker, D. K., R. Kalff, F. Gullotta, S. Weekes-Seifert, and U. Mo¨hrer. 1984. 5. Yan, H., D. W. Parsons, G. Jin, R. McLendon, B. A. Rasheed, W. Yuan, I. Kos, Mononuclear infiltrates in human intracranial tumors as a prognostic factor. I. Batinic-Haberle, S. Jones, G. J. Riggins, et al. 2009. IDH1 and IDH2 mutations Influence of preoperative steroid treatment. I. Glioblastoma. Clin. Neuropathol. in gliomas. N. Engl. J. Med. 360: 765–773. 3: 143–147. 6. Rich, J. N., C. Hans, B. Jones, E. S. Iversen, R. E. McLendon, B. K. Rasheed, 30. Schiffer, D., D. Cavicchioli, M. T. Giordana, L. Palmucci, and A. Piazza. 1979. A. Dobra, H. K. Dressman, D. D. Bigner, J. R. Nevins, and M. West. 2005. Gene Analysis of some factors effecting survival in malignant gliomas. Tumori 65:

expression profiling and genetic markers in glioblastoma survival. Cancer Res. 119–125. by guest on September 30, 2021 65: 4051–4058. 31. Rossi, M. L., N. R. Jones, E. Candy, J. A. Nicoll, J. S. Compton, J. T. Hughes, 7. Liang, Y., M. Diehn, N. Watson, A. W. Bollen, K. D. Aldape, M. K. Nicholas, M. M. Esiri, T. H. Moss, F. F. Cruz-Sanchez, and H. B. Coakham. 1989. The K. R. Lamborn, M. S. Berger, D. Botstein, P. O. Brown, and M. A. Israel. 2005. mononuclear cell infiltrate compared with survival in high-grade astrocytomas. Gene expression profiling reveals molecularly and clinically distinct subtypes of Acta Neuropathol. 78: 189–193. glioblastoma multiforme. Proc. Natl. Acad. Sci. USA 102: 5814–5819. 32. Safdari, H., F. H. Hochberg, and E. P. Richardson, Jr. 1979. Histological cor- 8. Sera˜o, N. V., K. R. Delfino, B. R. Southey, J. E. Beever, and S. L. Rodriguez-Zas. relations with survival in malignant gliomas. Acta Neurochir. Suppl. (Wien) 28: 2011. Cell cycle and aging, morphogenesis, and response to stimuli genes are 485–488. individualized biomarkers of glioblastoma progression and survival. BMC Med. 33. Wrann, M., S. Bodmer, R. de Martin, C. Siepl, R. Hofer-Warbinek, K. Frei, Genomics 4: 49. E. Hofer, and A. Fontana. 1987. T cell suppressor factor from human glioblas- 9. Zhang, J., B. Liu, X. Jiang, H. Zhao, M. Fan, Z. Fan, J. J. Lee, T. Jiang, T. Jiang, toma cells is a 12.5-kd protein closely related to transforming growth factor-beta. and S. W. Song. 2009. A systems biology-based gene expression classifier of EMBO J. 6: 1633–1636. glioblastoma predicts survival with solid tumors. PLoS ONE 4: e6274. 34. Hishii, M., T. Nitta, H. Ishida, M. Ebato, A. Kurosu, H. Yagita, K. Sato, and 10. Phillips, H. S., S. Kharbanda, R. Chen, W. F. Forrest, R. H. Soriano, T. D. Wu, K. Okumura. 1995. Human glioma-derived interleukin-10 inhibits antitumor A. Misra, J. M. Nigro, H. Colman, L. Soroceanu, et al. 2006. Molecular sub- immune responses in vitro. Neurosurgery 37: 1160–1166, discussion 1166–1167. classes of high-grade glioma predict prognosis, delineate a pattern of disease 35. Doherty, T. M., R. Kastelein, S. Menon, S. Andrade, and R. L. Coffman. 1993. progression, and resemble stages in neurogenesis. Cancer Cell 9: 157–173. Modulation of murine macrophage function by IL-13. J. Immunol. 151: 7151– 11. McLendon, R. E., O. Wiestler, J. M. Kros, A. Korshunov, and H. K. Ng. 2007. 7160. Ependymoma. In Pathology of Tumours of the Nervous System. World Health 36. Bettinger, I., S. Thanos, and W. Paulus. 2002. Microglia promote glioma mi- Organization Classification of Brain Tumors. W. W. Cavanee and P. Kleihues, gration. Acta Neuropathol. 103: 351–355. eds. International Agency for Research on Cancer, Lyon, France, p. 74–78. 37. Hussain, S. F., D. Yang, D. Suki, K. Aldape, E. Grimm, and A. B. Heimberger. 12. Wu, Z., I. R. Gentleman, R. Martinez, F. Murillo, and F. Spencer. 2004. A model- 2006. The role of human glioma-infiltrating microglia/macrophages in mediating based background adjustment for oligonucleotide expression arrays. J. Am. Stat. antitumor immune responses. Neuro-oncol. 8: 261–279. Assoc. 99: 909–917. 38. Fecci, P. E., D. A. Mitchell, J. F. Whitesides, W. Xie, A. H. Friedman, 13. Edgar, R., M. Domrachev, and A. E. Lash. 2002. Gene Expression Omnibus: G. E. Archer, J. E. Herndon, II, D. D. Bigner, G. Dranoff, and J. H. Sampson. NCBI gene expression and hybridization array data repository. Nucleic Acids 2006. Increased regulatory T-cell fraction amidst a diminished CD4 compart- Res. 30: 207–210. ment explains cellular immune defects in patients with malignant glioma. 14. Dennis, G., Jr., B. T. Sherman, D. A. Hosack, J. Yang, W. Gao, H. C. Lane, and Cancer Res. 66: 3294–3302. R. A. Lempicki. 2003. DAVID: Database for Annotation, Visualization, and 39. Heimberger, A. B., M. Abou-Ghazal, C. Reina-Ortiz, D. S. Yang, W. Sun, Integrated Discovery. Genome Biol. 4: 3. W. Qiao, N. Hiraoka, and G. N. Fuller. 2008. Incidence and prognostic impact of 15. Mootha, V. K., C. M. Lindgren, K. F. Eriksson, A. Subramanian, S. Sihag, FoxP3+ regulatory T cells in human gliomas. Clin. Cancer Res. 14: 5166–5172. J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstra˚le, E. Laurila, et al. 2003. 40. Sippel, T. R., J. T. White, K. Nag, V. Tsvankin, M. K. Klaassen, PGC-1alpha-responsive genes involved in oxidative phosphorylation are coor- B. K. Kleinschmidt-Demasters, and A. Waziri. 2011. Neutrophil degranulation dinately downregulated in human diabetes. Nat. Genet. 34: 267–273. and immunosuppression in patients with GBM: restoration of cellular immune 16. Ashburner, M., C. A. Ball, J. A. Blake, D. Botstein, H. Butler, J. M. Cherry, function by targeting arginase I. Clin. Cancer Res. 17: 6992–7002. A. P. Davis, K. Dolinski, S. S. Dwight, J. T. Eppig, et al; The Gene Ontology 41. Justicia, C., C. Gabriel, and A. M. Planas. 2000. Activation of the JAK/STAT Consortium. 2000. Gene ontology: tool for the unification of biology. Nat. Genet. pathway following transient focal cerebral ischemia: signaling through Jak1 and 25: 25–29. Stat3 in astrocytes. Glia 30: 253–270. 17. Wu, C., C. Orozco, J. Boyer, M. Leglise, J. Goodale, S. Batalov, C. L. Hodge, 42. Carro, M. S., W. K. Lim, M. J. Alvarez, R. J. Bollo, X. Zhao, E. Y. Snyder, J. Haase, J. Janes, J. W. Huss, III, and A. I. Su. 2009. BioGPS: an extensible and E. P. Sulman, S. L. Anne, F. Doetsch, H. Colman, et al. 2010. The transcriptional customizable portal for querying and organizing gene annotation resources. network for mesenchymal transformation of brain tumours. Nature 463: 318– Genome Biol. 10: R130. 325. SUPPLEMENTAL FIGURE LEGEND

Supplemental figure 1. Representative HGA histology of (A) greater than median, and (B) less than median tumor infiltration of helper T-cells (CD4; brown), (C) greater than median, and (D) less than median tumor infiltration of cytotoxic T-cells (CD8; red), (E) greater than 75th percentile, and (F) less than 75% percentile tumor infiltration of microglia/macrophages (AIF1; brown). Immunohistochemistry was performed using FFPE tumor sections and with hematoxylin counterstaining (400x magnification).

Supplemental table IA. High grade astrocytoma long-term survivor patient details.

Sample Survival Dead Age Karnofsky/ Dx detail location of Extent of Therapy received Used in

ID (yrs) at Lansky tumor surgery microarray

Dx performance study

(yrs) score

HGA1 17.0 N 34 100 AA pons biopsy radiation and BCNU

HGA2 16.5 N 8 80 AA cerebrum GTR carboplatin, etoposide, CCNU, vincristine and radiation

HGA3 16.0 N 47 90 GBM 1o cerebrum STR radiation and temozolmide

HGA4 11.0 N 45 60 GBM, giant cell cerebrum GTR none

HGA5 10.5 N 27 100 GBM 1o, cerebrum GTR radiation, CCNU, irinotecan, and temozolmide

epithelioid

HGA6 8.5 Y 21 100 RIG cerebellum STR radiation and temozolmide

HGA7 8.5 N 2 80 GBM 1o thalamus/ STR cisplatin, vincristine, cytoxan, etoposide, carboplatin,

cerebrum thiotepa

HGA8 8.5 N 42 100 GBM 2o cerebrum GTR radiation, CCNU, irinotecan and temozolmide

HGA9 8.5 N 46 80 GBM 1o cerebrum GTR radiation, CCNU, erlotinib and temozolmide yes

HGA10 7.0 N 10 80 GBM 1o thalamus biopsy radiation and temozolmide

HGA11 7.0 Y 24 80 GBM, prior cerebrum STR radiation, vincristine, iphosphamide, cisplatin and yes

PNET etoposide

HGA12 6.5 N 29 70 GBM 1o cerebrum STR radiation and temozolmide

HGA13 6.0 Y 9 90 GBM 1o midbrain STR radiation, etoposide, high dose chemos and temozolmide yes

HGA14 5.5 N 49 90 GBM 1o cerebrum GTR radiation, BCNU, irinotecan, taxol, thalidomide,

navelbine, erlotinib and temozolmide

Abbreviations: HGA, high grade astrocytoma; AA, anaplastic astrocytoma; GBM glioblastoma; PNET, primitive neuroectodermal tumor; GTR, gross total resection; STR, subtotal resection; BCNU, bis-chloroethylnitrosourea; CCNU, N-(2-chloroethyl)-N'- cyclohexyl-N-nitrosourea. Supplemental table IB. Manually curated list of immune-related genes associated with long-term survival in high grade astrocytoma

Gene gene name Correlation p-value

(Pearson r)

Adaptive immunity

BLNK B-cell linker 0.418 0.033801

BTK Bruton agammaglobulinemia tyrosine kinase 0.497 0.009814

CARD11 caspase recruitment domain family, member 11 0.557 0.003093

CD1D CD1d molecule 0.687 0.000104

CD274 CD274 molecule 0.415 0.035022

CD2AP CD2-associated protein 0.527 0.005663

CD3D CD3d molecule, delta (CD3-TCR complex) 0.600 0.00119

CD3E CD3e molecule, epsilon (CD3-TCR complex) 0.493 0.010433

CD3G CD3g molecule, gamma (CD3-TCR complex) 0.453 0.020257

CD52 CD52 molecule 0.429 0.028914

CD69 CD69 molecule 0.570 0.002361

CD7 CD7 molecule 0.475 0.014181

CD72 CD72 molecule 0.394 0.046511

CD8B CD8b molecule 0.500 0.009346 CLECL1 C-type lectin-like 1 0.494 0.01036

CTSW cathepsin W 0.454 0.019709

DEF6 differentially expressed in FDCP 6 homolog (mouse) 0.607 0.001006

FAIM3 Fas apoptotic inhibitory molecule 3 0.575 0.002113

FCER1A Fc fragment of IgE, high affinity I, receptor for; alpha polypeptide 0.708 5.18x10-05

FYB FYN binding protein 0.396 0.045248

GBP5 guanylate binding protein 5 0.390 0.049072

GIMAP1 GTPase, IMAP family member 1 0.479 0.013327

GIMAP4 GTPase, IMAP family member 4 0.497 0.009817

GIMAP5 GTPase, IMAP family member 5 0.607 0.000998

GPR183 G protein-coupled receptor 183 0.474 0.014393

HAVCR2 hepatitis A virus cellular receptor 2 0.446 0.022472

ICOS inducible T-cell co-stimulator 0.412 0.036551

IGHM immunoglobulin heavy constant mu 0.430 0.02829

IKZF1 IKAROS family zinc finger 1 (Ikaros) 0.564 0.002667

ITGAL integrin, alpha L (antigen CD11A (p180), lymphocyte function- 0.469 0.015709

associated antigen 1; alpha polypeptide)

LCK lymphocyte-specific protein tyrosine kinase 0.558 0.003042

LCP2 lymphocyte cytosolic protein 2 (SH2 domain containing 0.416 0.03445 leukocyte protein of 76kDa)

NLRC3 NLR family, CARD domain containing 3 0.476 0.014085

NLRC5 NLR family, CARD domain containing 5 0.468 0.015878

PIK3R5 phosphoinositide-3-kinase, regulatory subunit 5 0.540 0.00442

PLCG2 phospholipase C, gamma 2 (phosphatidylinositol-specific) 0.627 0.000612

PLEKHA1 pleckstrin homology domain containing, family A 0.454 0.019918

(phosphoinositide binding specific) member 1

PLSCR1 phospholipid scramblase 1 0.392 0.047903

RHOH ras homolog gene family, member H 0.648 0.000349

SASH3 SAM and SH3 domain containing 3 0.523 0.006086

SIT1 signaling threshold regulating transmembrane adaptor 1 0.471 0.015095

SYK spleen tyrosine kinase 0.452 0.020383

TAGAP T-cell activation RhoGTPase activating protein 0.501 0.009084

TAL1 T-cell acute lymphocytic leukemia 1 0.603 0.001112

TARP TCR gamma alternate reading frame protein; T cell receptor 0.502 0.008982

gamma variable 9; T cell receptor gamma constant 1

TARP /// T cell receptor gamma locus; T cell receptor gamma constant 2 0.493 0.010533

TRGC2

THEMIS thymocyte selection pathway associated 0.492 0.010715 TRAC T cell receptor alpha constant; T cell receptor alpha locus; T cell 0.514 0.007217

receptor alpha variable 20; T cell receptor delta locus; T cell

receptor delta variable 2

TRAT1 T cell receptor associated transmembrane adaptor 1 0.499 0.009414

TRBC1 T cell receptor beta variable 19; T cell receptor beta constant 1 0.552 0.003456

VAV1 vav 1 guanine nucleotide exchange factor 0.560 0.002923

ZAP70 zeta-chain (TCR) associated protein kinase 70kDa 0.450 0.021116

Cytotoxic effectors

ALOX5 arachidonate 5-lipoxygenase 0.466 0.016301

CYP4F11 cytochrome P450, family 4, subfamily F, polypeptide 11 0.602 0.001144

GZMA granzyme A (granzyme 1, cytotoxic T-lymphocyte-associated 0.473 0.014678

serine esterase 3)

GZMB granzyme B (granzyme 2, cytotoxic T-lymphocyte-associated 0.728 2.54E-05

serine esterase 1)

GZMH granzyme H (cathepsin G-like 2, protein h-CCPX) 0.623 0.000671

GZMK granzyme K (granzyme 3; tryptase II) 0.605 0.001068

GZMM granzyme M (lymphocyte met-ase 1) 0.416 0.034328

Innate immunity

AIF1 allograft inhibitory factor 1 0.458 0.018674 AOAH acyloxyacyl hydrolase (neutrophil) 0.500 0.009271

APOL1 apolipoprotein L, 1 0.400 0.042917

AXL AXL receptor tyrosine kinase 0.455 0.019627

C1R complement component 1, r subcomponent 0.421 0.032257

CARD9 caspase recruitment domain family, member 9 0.542 0.00424

CD180 CD180 molecule 0.587 0.001633

CD300LF CD300 molecule-like family member f 0.509 0.007904

CD33 CD33 molecule 0.514 0.007171

CD55 CD55 molecule, decay accelerating factor for complement 0.404 0.040668

(Cromer blood group)

CD59 CD59 molecule 0.462 0.017369

CD68 CD68 molecule 0.585 0.001697

CD86 CD86 molecule 0.422 0.031677

CFB complement factor B 0.472 0.014956

CFD complement factor D (adipsin) 0.481 0.012853

CFH complement factor H 0.554 0.003319

CHUK conserved helix-loop-helix ubiquitous kinase 0.394 0.046662

CIITA class II, major histocompatibility complex, transactivator 0.568 0.002448

COLEC12 collectin sub-family member 12 0.517 0.006783 FCN1 ficolin (collagen/fibrinogen domain containing) 1 0.608 0.000989

GBP1 guanylate binding protein 1, interferon-inducible, 67kDa 0.493 0.010586

GBP2 guanylate binding protein 2, interferon-inducible 0.563 0.002754

HCP5 HLA complex P5 0.595 0.001343

HLA-DOA major histocompatibility complex, class II, DO alpha 0.522 0.006194

HLA-DQB2 major histocompatibility complex, class II, DQ beta 2 0.483 0.012521

HLA-DRB1 major histocompatibility complex, class II, DR beta 1 0.693 8.76x10-05

HLA-DRB6 major histocompatibility complex, class II, DR beta 6 0.427 0.029693

(pseudogene)

HLA-E major histocompatibility complex, class I, E 0.451 0.020895

HMHA1 histocompatibility (minor) HA-1 0.488 0.011434

IGSF6 immunoglobulin superfamily, member 6 0.535 0.004842

INPP5D inositol polyphosphate-5-phosphatase, 145kDa 0.518 0.006683

KLRB1 killer cell lectin-like receptor subfamily B, member 1 0.734 1.97x10-05

LY96 lymphocyte antigen 96 0.391 0.048467

LYN v-yes-1 Yamaguchi sarcoma viral related oncogene homolog 0.398 0.043881

MERTK c-mer proto-oncogene tyrosine kinase 0.427 0.02964

MNDA myeloid cell nuclear differentiation antigen 0.392 0.047515

MYO1F myosin IF 0.427 0.029689 NCF4 neutrophil cytosolic factor 4, 40kDa 0.506 0.008301

NFKBIA nuclear factor of kappa light polypeptide gene enhancer in B-cells 0.526 0.005751

inhibitor, alpha

NOD2 nucleotide-binding oligomerization domain containing 2 0.710 4.87x10-05

PLA2G7 phospholipase A2, group VII (platelet-activating factor 0.464 0.017041

acetylhydrolase, plasma)

POU2F2 POU class 2 homeobox 2 0.467 0.016088

PRAM1 PML-RARA regulated adaptor molecule 1 0.661 0.000238

RNF125 ring finger protein 125 0.397 0.044679

SAA1 /// serum amyloid A1 0.607 0.001011

SAA2

SERPING1 serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 0.441 0.024195

SH2D1A SH2 domain protein 1A 0.508 0.008027

SP140 SP140 nuclear body protein 0.597 0.001293

TLR2 toll-like receptor 2 0.480 0.013066

TLR3 toll-like receptor 3 0.484 0.012162

TLR5 toll-like receptor 5 0.476 0.013997

TLR6 toll-like receptor 6 0.471 0.015179

TLR7 toll-like receptor 7 0.442 0.023944 TLR8 toll-like receptor 8 0.388 0.049933

TREML1 triggering receptor expressed on myeloid cells-like 1 0.482 0.012659

Cytokines/chemokines

CCL19 chemokine (C-C motif) ligand 19 0.456 0.019194

CCL2 chemokine (C-C motif) ligand 2 0.411 0.037176

CCL5 chemokine (C-C motif) ligand 5 0.545 0.004012

CXCL13 chemokine (C-X-C motif) ligand 13 0.397 0.044455

CXCL9 chemokine (C-X-C motif) ligand 9 0.515 0.00716

IL10 interleukin 10 0.482 0.012653

IL15 interleukin 15 0.416 0.034579

IL16 interleukin 16 0.460 0.018056

IL17B interleukin 17B 0.585 0.001699

IL18 interleukin 18 (interferon-gamma-inducing factor) 0.475 0.014108

IL32 interleukin 32 0.495 0.010217

IL34 interleukin 34 0.447 0.021999

TNF tumor necrosis factor 0.392 0.047899

TNFSF10 tumor necrosis factor (ligand) superfamily, member 10 0.472 0.014838

TNFSF12 tumor necrosis factor (ligand) superfamily, member 12 0.469 0.015709

TNFSF13 tumor necrosis factor (ligand) superfamily, member 13 0.429 0.028592 XCL1 chemokine (C motif) ligand 1 0.469 0.015612

Cytokine/chemokine receptors

CD40 CD40 molecule, TNF receptor superfamily member 5 0.461 0.017777

CSF2RB colony stimulating factor 2 receptor, beta, low-affinity 0.500 0.009313

(granulocyte-macrophage)

CXCR6 chemokine (C-X-C motif) receptor 6 0.512 0.00745

IFNGR1 interferon gamma receptor 1 0.498 0.009675

IL10RA interleukin 10 receptor, alpha 0.411 0.037138

IL15RA interleukin 15 receptor, alpha 0.520 0.006424

IL18R1 interleukin 18 receptor 1 0.434 0.026918

IL1R1 interleukin 1 receptor, type I 0.577 0.002022

IL28RA interleukin 28 receptor, alpha (interferon, lambda receptor) 0.458 0.018551

IL4R interleukin 4 receptor 0.455 0.01955

IL6R interleukin 6 receptor 0.516 0.006917

IL7R interleukin 7 receptor 0.487 0.011605

LTB4R leukotriene B4 receptor 0.513 0.00734

PTAFR platelet-activating factor receptor 0.512 0.007546

TGFBR2 transforming growth factor, beta receptor II (70/80kDa) 0.408 0.038313

TGFBR3 transforming growth factor, beta receptor III 0.427 0.029792 TNFRSF14 tumor necrosis factor receptor superfamily, member 14 0.444 0.023131

(herpesvirus entry mediator)

Immune cell intracellular signaling

CMTM7 CKLF-like MARVEL transmembrane domain containing 7 0.429 0.028688

EBI3 Epstein-Barr virus induced 3 0.507 0.008255

IFI27 interferon, alpha-inducible protein 27 0.391 0.048321

IFI35 interferon-induced protein 35 0.432 0.027491

IFIH1 interferon induced with helicase C domain 1 0.508 0.008014

IFIT2 interferon-induced protein with tetratricopeptide repeats 2 0.545 0.004004

IFIT3 interferon-induced protein with tetratricopeptide repeats 3 0.416 0.034776

IL18BP interleukin 18 binding protein 0.594 0.001391

IRF1 interferon regulatory factor 1 0.453 0.020196

IRF8 interferon regulatory factor 8 0.454 0.019847

ITK IL2-inducible T-cell kinase 0.492 0.010746

JAK1 Janus kinase 1 0.453 0.019994

STAT1 signal transducer and activator of transcription 1, 91kDa 0.390 0.048831

STAT4 signal transducer and activator of transcription 4 0.563 0.00276

Migration, tethering and rolling

ARHGAP4 Rho GTPase activating protein 4 0.395 0.045818 CD2 CD2 molecule 0.528 0.005568

CEACAM8 carcinoembryonic antigen-related cell adhesion molecule 8 0.578 0.002

CORO1A coronin, actin binding protein, 1A 0.431 0.02775

DOCK2 dedicator of cytokinesis 2 0.463 0.017224

ICAM2 intercellular adhesion molecule 2 0.392 0.047513

PECAM1 platelet/endothelial cell adhesion molecule 0.404 0.040405

SELE selectin E 0.392 0.047926

SELPLG selectin P ligand 0.406 0.039731

Miscellaneous immune function

CST7 cystatin F (leukocystatin) 0.449 0.021445

CTSC cathepsin C 0.470 0.015388

ELF1 E74-like factor 1 (ets domain transcription factor) 0.434 0.026729

FERMT3 fermitin family member 3 0.414 0.035552

FGR Gardner-Rasheed feline sarcoma viral (v-fgr) oncogene homolog 0.414 0.035458

HCST hematopoietic cell signal transducer 0.416 0.03439

IDO1 indoleamine 2,3-dioxygenase 1 0.502 0.009025

IKBKB inhibitor of kappa light polypeptide gene enhancer in B-cells, 0.494 0.010245

kinase beta

ITGB7 integrin, beta 7 0.426 0.030069 JAG2 jagged 2 0.508 0.008108

KYNU kynureninase 0.436 0.026063

LILRB4 leukocyte immunoglobulin-like receptor, subfamily B (with TM 0.441 0.024106

and ITIM domains), member 4

LY75 lymphocyte antigen 75 0.422 0.031654

LY86 lymphocyte antigen 86 0.401 0.042552

PLSCR4 phospholipid scramblase 4 0.413 0.036207

PML promyelocytic leukemia; similar to promyelocytic leukemia 0.484 0.012282

protein isoform 1

PODXL podocalyxin-like 0.457 0.018982

PROCR protein C receptor, endothelial 0.448 0.021594

PTPN22 protein tyrosine phosphatase, non-receptor type 22 (lymphoid) 0.444 0.02311

PTPN6 protein tyrosine phosphatase, non-receptor type 6 0.421 0.032079

PYHIN1 pyrin and HIN domain family, member 1 0.487 0.01165

RPS6KA1 ribosomal protein S6 kinase, 90kDa, polypeptide 1 0.438 0.025356

RPS6KA2 ribosomal protein S6 kinase, 90kDa, polypeptide 2 0.430 0.028358

SAMSN1 SAM domain, SH3 domain and nuclear localization signals 1 0.453 0.020169

SEMA4D sema domain, immunoglobulin domain (Ig), transmembrane 0.499 0.009409

domain (TM) and short cytoplasmic domain, (semaphorin) 4D STXBP2 syntaxin binding protein 2 0.438 0.02514

TAP2 transporter 2, ATP-binding cassette, sub-family B (MDR/TAP) 0.407 0.038823

TAPBPL TAP binding protein-like 0.429 0.028878

TMEM173 transmembrane protein 173 0.393 0.047116

TNFAIP3 tumor necrosis factor, alpha-induced protein 3 0.389 0.049809

TNFAIP8L2 tumor necrosis factor, alpha-induced protein 8-like 2 0.489 0.011267

TRIM22 tripartite motif containing 22 0.424 0.030926

XAF1 XIAP associated factor 1 0.393 0.047236

Supplemental table IC. Immune cell lineage and HGA survival associated gene enrichment in glioblastoma survival correlated genes in a publicly available gene expression dataset (Phillips; GSE4271).

Enrichment Immune cell lineage/HGG survival correlate NES p-value

Positively correlated with survival

1 CD14 monocyte 2.02 < 0.001

2 HGA positive correlates of survival: “immune response” 1.90 < 0.001

GOterm only

3 Myeloid lineage (CD14 and CD33 combined) 1.89 < 0.001

4 HGA positive survival correlates 1.87 < 0.001

5 CD33 myeloid cell 1.65 < 0.001

6 CD56 natural killer cell 1.25 0.126

7 BDCA4 dendritic cell 0.96 0.525

8 CD4 helper T-cell 0.85 0.694

9 T-cell lineage (CD4 and CD8 combined) 0.80 0.851

Negatively correlated with survival

1 HGA negative correlates of survival 3.06 < 0.001

2 CD19 B-cell 0.87 0.742

3 CD8 cytotoxic T-cell 0.76 0.868

Supplementary table II. Classifier gene sets for specific immune cell lineages and HGA survival associated genes. helper T-cecytotoxic T-Monocytes CD19 (B-ceMyeloid cel natural kille Dendritic ceT-cell lineagmyeloid linaHGA positivHGA positivHGA negative correlate AQP3 ANKRD36B AIF1 ABCB4 ARL4A ADRB2 ADA ANKRD36B AIF1 ABCA13 anxa11 ABL1 ATHL1 ATHL1 ALDH3B1 ACP5 AZU1 AKR1C3 ALOX5AP AQP3 ALDH3B1 ABCA2 APOL1 ACLY BCL11B ATM ASGR1 ADAM19 BST1 APOBEC3F BCL11A ATHL1 ALOX5 ABCB1 BLNK ADD2 CCR7 BCL11B ASGR2 ADAM28 C17orf91 ARHGEF3 BLNK ATM ANXA1 ABCC1 C1r ADNP CD27 CCR7 C11orf75 ADK CAPG ARL4C C1orf34 ATXN7 AOAH ABCC4 CARD9 AGAP1 CD28 CD27 CD14 BACE2 CARD9 AXIN1 C20orf103 BCL11B AP1S2 ABCC6 CCL14 AGAP10 CD3D CD3D CD163 BACH2 CCL3 CCL4 C9orf127 BTN3A2 ASGR1 ABHD15 CCL19 AKAP8L CD3G CD3E CD1D BANK1 CCR1 CCL5 CBFA2T3 BTN3A3 ASGR2 ABI3 Ccl2 AKT1S1 CYLD CD8A CD36 BIRC3 CD14 CD160 CCDC88A C16orf67 AZU1 ABI3BP CCL5 AMH CYorf15B CD8B CD86 BLK CD163 CD244 CCR2 CASP8 BST1 ACAA1 CD180 ANKRD20A1 DGKA DGKA CFD CCR6 CD36 CD247 CD2AP CCR7 C11orf75 ACACB CD1D ANKRD28 FLT3LG FLT3LG CPVL CD180 CD44 CD300A CD38 CD2 C17orf91 ACADSB CD274 ANKRD50 GIMAP5 GPR171 CSTA CD19 CD93 CD7 CD4 CD27 C20orf27 ACBD4 CD300LF ANKRD54 ICOS GSDML FAM26B CD22 CDA CHST12 CDYL CD28 CAPG ACCS CD302 ANXA6 IL11RA GZMK FBP1 CD24 CDC42EP3 CLIC3 CKS2 CD3D CARD9 ACOT13 Cd55 ARIH2 IL32 HIST1H2BGFGL2 CD37 CEBPD CST7 CLIC3 CD3E CASP1 ACSL1 CD7 ARL4A IL7R IL32 FOLR3 CD40 CENTA2 CTSW COBLL1 CD3G CCL3 ACSL5 CD86 ASF1B ITGA6 IL7R FUCA1 CD72 CES1 CX3CR1 CPNE3 CD6 CCR1 ACSM5 CD8B ASPM ITK ITK HMOX1 CD79A CKAP4 DUSP2 CSF2RA CD8A CCR2 ADAM28 CEACAM8 ASXL1 LAT KLRG1 HNMT CD79B CLEC11A ERBB2 CUEDC1 CD8B CD14 ADAMTS16CFB ATAD2 LCK KLRK1 KYNU CR2 CLEC4A GATA3 CUX2 CDC25B CD163 ADAMTS8 cfd AURKA LDLRAP1 LAT LILRA1 CXCR5 CLEC5A GNLY CXCR3 CENTB1 CD1D ADAMTSL2 CFH AURKB LEF1 LCK LILRA3 EAF2 CXCL2 GPR56 DAB2 CYLD CD302 ADAP1 CHUK AZIN1 PASK LEF1 LILRA6 EZR CYBB GPR65 DCPS DENND2D CD33 ADAP2 Ciita BAHD1 TCF7 N4BP2L2 LILRB1 FAIM3 CYP1B1 GZMA DNASE1L3 DGKA CD36 ADCY7 clu BAT1 TMC6 NUP210 LILRB2 FAM30A DDEF1 GZMB DUSP5 FLT3LG CD86 ADD3 Colec12 BAZ1B TNFRSF25 PASK MS4A6A FAM46C DDX3X GZMH EPHA2 FUSIP1 CD93 ADORA3 Coro1a BFAR TRA@ SEPT6 NUP214 FCER2 DUSP1 GZMM FAM20B GBP1 CDA ADPGK Cst7 BIRC5 TRAC SFI1 OAS1 FCGR2B DUSP6 IL18RAP FCER1A GIMAP5 CDC42EP3 ADRA2A CTSC BUB1 TRAF3IP3 TARP PIK3R2 FCRL2 EGR1 IL2RB FCHSD2 GOLGA8A CEBPA ADRB2 CTSW BUB1B TRAT1 TMC6 PILRA GGA2 ELA2 IQGAP1 FLT3 GPR171 CEBPD ADRBK2 CXCL13 BYSL TRAV20 TRA@ PYCARD GPR18 F5 KIR2DL2 FUT7 GZMK CECR1 AGPAT9 Cxcl9 C11orf92 TRBC1 TRAC RIN3 HHEX FAM118A KIR3DL1 GAS6 HIST1H2BGCENTA2 AGTRAP dhx58 C11orf95 UBASH3A TRGC2 S100A8 HIST1H1C FBN2 KIR3DL2 GGA2 IL11RA CFD AHCYL1 ebi3 C12orf53 SLC7A7 HLA-DOB FCAR KLRB1 GNA15 IL32 CFP AIF1 FAIM3 C15orf23 STAB1 HLA-DPA1 FCGR1A KLRC1 GPX7 IL7R CKAP4 AIFM3 FCN1 C16orf75 SULT1A1 HLA-DQA1 FCGR1B KLRC3 GZMB ITGA6 CLEC4A AIG1 FTH1 C16orf80 TBXAS1 HLA-DQB1 FCN1 KLRD1 HLA-DQA1 ITK CLEC7A AK5 FYB C16orf88 TMEM176AHLA-DRB1 FER1L3 KLRF1 HLA-DRB4 ITPR3 COQ2 AKAP13 GBP1 C17orf51 TMEM176BIFT57 FES LAIR2 HSP90B1 KLRG1 CPVL AKR1A1 GBP2 C17orf76 TNFSF12 IGHA1 FLJ22662 LL22NC03-5IDH3A KLRK1 CSTA AKR1C1 GBP3 C18orf54 TNFSF13 IGHD FOS LPCAT1 IGFBP3 LAT CTSS AKR1C2 GBP4 C19orf29 TYMP IGHG1 FOSB MAF IGH@ LCK CXCL2 AKR1C3 gbp5 C1orf187 VCAN IGHM FPR1 MATK IGHA1 LDLRAP1 CYBB ALG14 Gpr183 C20orf117 VNN1 IGKC FUT4 MYBL1 IGHD LEF1 CYP1B1 ALOX5 GPSM3 C2orf72 IGL@ G0S2 MYOM2 IGHG1 LEPROTL1 DPYD AMICA1 GZMA C3orf19 IGLJ3 GLIPR1 NCR3 IGHM LONP2 DUSP1 AMPD2 Hamp C4orf46 IL4R GNS NKG7 IGJ LY9 DUSP6 AMPD3 HLA-DOA C8orf30A ISG20 GPR109B PDGFD IGK@ N4BP2L2 EGR1 ANK2 HLA-DRB3 CABLES2 KIAA0125 HK2 PLEKHF1 IGKC NLRP1 ELA2 ANKRD34C HLA-E CAMK2N2 LOC100133HK3 PRF1 IGKV4-1 NSUN5B F5 ANXA11 ICOS CASC5 LOC652493HLA-DQB1 PRKCH IGL@ NUP210 FAM118A ANXA3 IFI35 CCDC99 LOC728970HLA-DRB4 PRR5 IGLC2 PASK FBN2 ANXA7 IFIH1 CCNA2 LTB HNMT PRSS23 IGLJ3 PBXIP1 FBP1 AOAH IGH@ CCNB1 MBD4 HP PTGDR IGLV4-60 PIK3IP1 FCAR AP1B1 Igsf6 CCNB2 MS4A1 IER3 PTGER2 IL18R1 PPWD1 FCGR1A APOL1 il10 CDC20 NFATC1 IFI44 PTPN4 IL3RA RAPGEF6 FCGR1B APOL3 IL15 CDC25A NUP88 IL1B RAB7L1 IRF4 RASGRP1 FCN1 ARAP1 IL16 CDC45 ODC1 IL8 RASSF1 IRF7 RBL2 FER1L3 AREG IL17B CDC6 OSBPL10 IRF5 RGS3 IRF8 RHOH FES ARFGEF2 IL18 CDCA2 P2RX5 ITGAM RORA KCNK10 SEPT6 FGL2 ARHGAP15 IL18BP CDCA5 PALM2-AKAJUN RUNX3 KCTD5 SFI1 FLJ22662 ARHGAP24 Il18r1 CDK1 PDLIM1 KIAA0101 S1PR5 KRT5 SIGIRR FOS ARHGAP27 Il1r1 CDK2 PLEKHF2 KLF4 SH2D2A LAIR1 SIRPG FOSB ARHGAP4 IL32 CDK5R1 PNOC LGALS2 SMAD7 LEPREL1 SKAP1 FPR1 ARHGAP9 IL4R CDKN2AIPNL POU2AF1 LYZ SPON2 LILRA4 SLC7A6 FUCA1 ARHGEF37 il6r CDKN3 QRSL1 MAFB SYNGR1 LOC100130STAT4 FUT4 ARL5A IL7R CDT1 RABEP2 MAPK14 TARP LOC339562TARDBP G0S2 ARRB1 Inpp5d CDYL2 RFX5 MCL1 TBX21 LOC652493TARP GLIPR1 ARRDC2 IRF8 CENPA RPS21 MPO TGFBR3 MAN2B1 TCF7 GNS ASMTL ITGAL CENPE RRAS2 MS4A6A TKTL1 MAPKAPK2TMC6 HCK ASPA Kynu CENPF SP100 MTMR11 TOX MTMR1 TMEM204 HK2 ATG10 Lair1 CENPH SP140 NAIP TPST2 MUTED TMEM63A HK3 ATG4C lcp2 CENPK SPIB NFKBIA TRA@ MX1 TNFRSF25 HLA-DRB1 ATHL1 LILRB4 CENPN STAG3 NLRP3 TRD@ MYCL1 TRA@ HNMT ATL3 LTB4R CEP135 STAP1 NR4A2 TRPV2 NPC1 TRAC IER3 ATP2A3 LY86 CEP55 SWAP70 PADI4 XCL1 OPN3 TRAF3IP3 IFI30 ATP6V0E1 LY96 CGGBP1 SYPL1 PLK3 ZAP70 P2RY14 TRAF5 IFI44 ATP8B4 LYN CHAF1A TCL1A PPP1R15A PPP1R14B TRAJ17 IL1B ATPAF1 myo1f CHEK1 VPREB3 PTX3 PSCD4 TRAT1 IL1RN AVPI1 Ncf4 CHML ZNF395 RAB11FIP1 PTCRA TRBC1 IL8 AXL NLRX1 CLIP2 RASSF4 RGS1 TRGC2 IMPA2 BAG3 NOD2 CLPTM1 RETN RRM2 TRPV6 IRAK3 BATF P2RY14 COL9A3 RIPK2 SCARB2 UBASH3A IRF5 BATF2 PLCG2 CPNE5 RNF24 SCT USP34 JUN BCAS1 Pou2f2 CPSF2 RRP12 SEC61A1 KLF10 BCAT2 Procr CSE1L S100A12 SERINC2 KLF4 BCO2 RNF125 CSNK1E S100A8 SERPINF1 KYNU BDH2 SEMA4D CSPG4 SAMHD1 SLC2A8 LGALS2 BEST1 serpinb4 CTDSPL2 SCO2 SLC7A5 LILRA1 BHMT2 SERPING1 CXorf24 SCPEP1 SMC6 LILRA2 BIN1 Sh2d1a DBF4 SERPINB1 SNF1LK LILRA3 BLNK SIT1 DCAF15 SERPINB2 SOX4 LILRA6 BLOC1S2 Snca DCX SGK1 SPIB LILRB1 BMP2K Sp100 DDA1 SLC2A14 ST14 LILRB2 BTBD11 STXBP2 DDX11 SLC2A3 ST6GALNAC4 LMO2 BTG2 STXBP3 DDX39 SLC7A7 STT3A LY86 BTK TAP2 DEPDC1 STAB1 TCF4 LYZ C10orf11 TFEB DEPDC1B STK17B TLR7 MAFB C10orf125 Tgfbr3 DHFR STX11 TMEM149 MANBA C10orf128 THEMIS DHPS TFEC TNFRSF17 MAP3K3 C10orf26 TLR2 DHRS13 TLR2 TNFRSF21 MCL1 C10orf4 TLR3 DIRAS1 TLR4 TPM2 MPO C10orf54 TLR5 DLGAP5 TncRNA TRAF4 MS4A6A C10orf78 TLR7 DMWD TNFAIP2 TRAM1 MTMR11 C10orf81 TLR8 DNMT1 TNFSF12 TXNDC1 NAAA C10orf92 Tmem173 DOK4 TNFSF13 UBE2J1 NAGA C11orf59 TNF DOT1L TRIB1 UGCG NAIP C11orf67 TNFAIP8L2 DRG2 TSPO NCF2 C11orf75 TNFRSF14 DSCAM TYMS NLRP3 C11orf87 TNFSF10 DTD1 VCAN NR4A2 C12orf59 Tnfsf12 DTL NUP214 C13orf31 TRAT1 DVL3 OAS1 C14orf159 TREML1 E2F7 PADI4 C15orf52 TRIM22 ECSIT PPP1R15A C16orf54 VAV1 ECT2 PYCARD C17orf108 XCL1 EHMT2 RAB32 C17orf28 zap70 ELK1 RASSF4 C17orf62 ELOF1 RETN C19orf76 EMILIN3 RNF24 C1orf175 ENC1 S100A12 C1orf190 ENO2 S100A8 C1orf38 EPN2 SAMHD1 C1orf63 ERCC6L SCO2 C1R ERF SCPEP1 C20orf56 ERGIC3 SERPINB1 C21orf130 ESPL1 SGK1 C21orf131 EXO1 SLC2A14 C22orf25 EZH2 SLC7A7 C2CD2 FAM127B SQRDL C3orf54 FAM131B STAB1 C3orf58 FAM64A STX11 C4orf12 FAM83D TBXAS1 C4orf19 FANCD2 TFEC C4orf32 FANCE TGFBI C4orf6 FANCG TLR2 C5orf27 FANCI TLR4 C5orf4 FASN TLR5 C6orf106 FBXL19 TMEM176BC6orf217 FBXO46 TNFAIP2 C6orf70 FBXO5 TNFRSF1B C7orf54 FGD1 TNFSF12 C7orf58 FHDC1 TNFSF13 C8orf12 FIZ1 TRIB1 C8orf4 FKBP4 TSPO C8orf56 FKSG49 TYMP C8orf60 FLJ39632 VCAN C9orf135 FOXM1 VNN1 CABC1 FREM2 CACHD1 FSCN1 CALB2 FSD1 CALML4 FUS CAPN3 G0S2 CAPZA1 GABPB1 CARD11 GAS2L3 CARD16 GGT7 CARD16 GINS1 CARD6 GINS2 CARD9 GINS3 CARNS1 GPR153 CASP1 GPR173 CASP4 GSK3B CASS4 GSX1 CBX7 H2AFX CC2D1B H2AFY2 CCBE1 H2BFXP CCBL2 HELLS CCDC102A HJURP CCDC69 HMGN1 CCDC85A HNRNPU CCDC88B HOXD3 CCL14 HR CCL19 HRK CCL2 HSP90AB1 CCL5 HUNK CCRL2 IER2 CCS IFT81 CD180 IGFBP2 CD1D IGFBP5 CD2 IGHG1 CD274 ILF3 CD2AP ING5 CD300LF IQGAP3 CD302 IRF2BP1 CD33 KCTD15 CD3D KDM2B CD3E KHDRBS1 CD3G KHSRP CD40 KIAA0101 CD52 KIAA0114 CD55 KIAA0922 CD59 KIAA1524 CD68 KIAA1549 CD69 KIF11 CD7 KIF14 CD72 KIF15 CD86 KIF18B CD8B KIF20A CDADC1 KIF23 CDC42SE2 KIF2C CDH19 KIF4A CDK18 KIFC1 CDKL2 KLHDC3 CDS1 KNTC1 CEACAM8 LEMD2 CEBPD LIG1 CECR1 LIN9 CELF6 LIX1L CES4 LMNB1 CES8 LMNB2 CFB LOC100288490 CFD LOC100288507 CFH LOC283454 CH25H LOC283658 CHI3L2 LOC400027 CHIC1 LOC642869 CHMP4A LOC645323 CHODL LOC729887 CHST12 LRFN4 CHST15 LRP5 CHUK LRRC4B CHURC1 LRRC68 CIDEB LRRN1 CIITA LSM14B CILP LSM7 CLCA4 LTBP4 CLECL1 LUC7L2 CLIC2 MAD2L1 CLIC5 MAP3K1 CLMN MCM10 CLU MCM2 CLYBL MCM4 CMAH MCM7 CMPK2 MCM8 CMTM7 MDFI CNDP1 MED1 CNN1 MELK CNTNAP1 MEX3A CNTNAP4 MEX3B COL4A5 MEX3D COL4A6 MFGE8 COLEC12 MGC2752 COPS4 MIR770 CORO1A MKI67 COX4I2 MLF1IP CPEB1 MLH1 CPNE6 MMP15 CPZ MOCS3 CREBL2 MOSPD3 CREG1 MPDU1 CREM MPHOSPH9 CRYL1 MYBL2 CRYZL1 NCAPG CSF1R NCAPG2 CSF2RA NCRNA00188 CSF2RB NEK2 CST7 NFYA CTAGE5 NONO CTBS NOTCH1 CTSC NOVA1 CTSH NR0B1 CTSW NRBP1 CTSZ NRM CTTN NT5C3 CUTC NT5DC2 CXCL13 NUF2 CXCL9 NUSAP1 CXCR6 NXPH3 CXorf21 OBSL1 CYB5A ODC1 CYB5R1 ODF2 CYBASC3 OIP5 CYFIP1 OLFML2A CYP21A2 ORC6L CYP2J2 PAFAH1B3 CYP4F11 PAICS CYR61 PBK CYSLTR1 PCBP4 CYTH4 PCGF2 CYYR1 PCMTD2 DAAM1 PCNA DAB1 PDAP1 DAGLB PEG10 DAPP1 PGAP1 DARC PHC1 DDHD2 PIK3R2 DDO PKDCC DDR2 PKMYT1 DDX60 PLK4 DDX60L POGK DEF6 POLD1 DENND1C POLE2 DENND2D POLR2J2 DENND3 POLR2J2 DEPDC6 POLR2J4 DGKE POLR3D DHCR24 PPM1G DHRS12 PPP1R10 DHRS4 PPP2R1A DHRS7 PPP5C DHRS9 PPPDE1 DHX58 PRC1 DKFZP586I1420 PRIM2 DLX4 PRKAR1B DNAH1 PRMT1 DNAH5 PRR3 DNAJB4 PTK7 DNAJC3 PTMS DNMBP PTPRG DOCK2 PTTG1 DOCK4 PUM2 DOCK5 PXMP4 DOK7 QTRT1 DPEP2 RAB11B DPEP3 RACGAP1 DPYS RAD51 DTX3L RAD51AP1 DUSP22 RAD54B EAF2 RAF1 EBI3 RANBP1 ECHDC1 RAPGEF1 ECHS1 RAVER1 EDN1 RBL1 EFEMP1 RBM15B EFTUD2 RBM8A EGF RCC1 ELF1 RFC3 ELL2 RFWD3 ELOVL1 RGAG4 ELOVL5 RNASEH2A ELOVL7 RNF144A ELSPBP1 RP9P EMB RPP14 EMR1 RPRM ENPP5 RPS5 ENTPD1 RRM2 EPHX1 RRN3 EPHX2 RRP1 EPS15 SAMD1 ERMAP sept9 ERMP1 SETD5 EVI5 SF3B3 EXTL1 SGOL2 FAAH SH3BP5L FAH SH3GL1 FAHD2A SHMT2 FAIM3 SLC25A16 FAM118A SLC25A17 FAM119A SMARCA4 FAM124A SMARCC1 FAM134B SMARCD1 FAM179A SMEK1 FAM181A SNRPB2 FAM19A4 SOX11 FAM43B SOX12 FAM49B SOX4 FAM69A SPATA6 FAM78A SPC24 FAM78B SPC25 FAM82A1 SPRY4 FAM91A2 SPSB4 FASTKD1 SRFBP1 FBP2 STIP1 FBXO2 STK32B FBXO31 STMN3 FBXO32 STRN4 FBXO44 STX10 FCER1A SYNCRIP FCN1 TAF1L FERMT3 TAF4 FES TFAP2A FGD2 TGFB1I1 FGD4 TIMELESS FGF19 TIMM44 FGF20 TK1 FGL2 TMEM169 FGR TMEM183A FIG4 TMEM97 FILIP1L TMF1 FLI1 TMPO FLJ12120 TMSB15A FLJ41484 TMSB15B FOLH1B TOP1 FOLR2 TOP2A FOSL2 TOP2B FOXA2 TPM4 FOXF1 TPX2 FPR1 TRAF4 FRAT2 TRIB2 FTH1 TRIM24 FUCA1 TRIM28 FYB TRIO GAB3 TRIP13 GABBR1 TROAP GADD45B TTK GAPT TUBB GAS6 TUBB2A GATA2 TUBB2C GBP1 TUBB3 GBP2 TYMS GBP3 U2AF1 GBP4 U2AF2 GBP5 UBE2C GCA UBE2I GCC2 UBE2M GDE1 UBE2S GDF10 UBE2T GGA2 UBP1 GGTA1 UHRF1 GHITM USP10 GIMAP1 USP21 GIMAP4 VASH2 GIMAP5 VAV2 GIMAP6 VEGFA GIMAP7 VGLL4 GIMAP8 VMA21 GJC2 VPRBP GLDN VPS37D GLI1 WDR76 GLUL WDR82 GLYATL1 WEE1 GMFG WHSC1 GNA13 XPO7 GNG13 XRN2 GNGT2 ZBED4 GPR120 ZBTB6 GPR133 ZBTB9 GPR143 ZC3H18 GPR160 ZC3HAV1L GPR171 ZFP64 GPR183 ZIK1 GPR37 ZNF134 GPR62 ZNF135 GPRIN3 ZNF136 GPSM3 ZNF141 GRP ZNF146 GSDMD ZNF175 GTF2H5 ZNF182 GVIN1 ZNF202 GYG1 ZNF207 GZMA ZNF213 GZMB ZNF273 GZMH ZNF282 GZMK ZNF286A GZMM ZNF286A HAMP ZNF300 HAP1 ZNF311 HAPLN2 ZNF316 HAVCR2 ZNF320 HCG26 ZNF384 HCP5 ZNF414 HCST ZNF439 HERPUD1 ZNF496 HGF ZNF501 HHAT ZNF502 HHATL ZNF510 HHIP ZNF551 HHLA3 ZNF568 HIAT1 ZNF586 HIBCH ZNF596 HLA-DOA ZNF627 HLA-DQB2 ZNF653 HLA-DRB1 ZNF670 HLA-DRB6 ZNF709 HLA-E ZNF713 HMHA1 ZNF768 HP ZNF813 HP /// HPR ZNF829 HPD ZNF850P HPGDS ZWINT HPN HPS5 HPSE HS3ST4 HSD17B3 HSD3B7 HSPA1A HSPA2 HSPBAP1 HTATIP2 HTR2B HYDIN ICAM2 ICOS IDE IDO1 IFI27 IFI35 IFIH1 IFIT2 IFIT3 IFNGR1 IGF1 IGFBP1 IGHM IGSF6 IKBKB IKZF1 IL10 IL10RA IL15 IL15RA IL16 IL17B IL18 IL18BP IL18R1 IL1R1 IL28RA IL32 IL34 IL4R IL6R IL7R IMMP2L INMT INPP5D IP6K3 IPCEF1 IRF1 IRF8 ISCU ITGAL ITGB7 ITIH2 ITK ITPKC ITSN2 JAG2 JAK1 JKAMP KCNAB2 KCNC3 KCND1 KCNK15 KCNQ1 KCTD21 KIAA0125 KIAA0247 KIAA0368 KIAA0513 KIAA0562 KIAA1109 KIAA1598 KIF16B KLF2 KLF4 KLF9 KLHL2 KLHL6 KLK6 KLRAQ1 KLRB1 KPNA3 KRT17 KRT222 KYNU LACTB LAIR1 LAP3 LCK LCP2 LDLRAP1 LEPR LEPREL1 LGALS9 LGI3 LGMN LHPP LILRB4 LMAN2L LOC100128844 LOC100129406 LOC100131170 LOC100131434 LOC100131564 LOC100131993 LOC100132741 LOC100233209 LOC100271722 LOC100286925 LOC100287445 LOC100288617 LOC100288643 LOC100289465 LOC100310756 LOC146429 LOC151438 LOC283904 LOC283999 LOC284112 LOC284570 LOC285281 LOC285628 LOC285758 LOC285771 LOC285972 LOC389332 LOC643008 LOC643733 LOC645638 LOC646014 LOC646870 LPAR5 LPIN1 LRCH4 LRIG3 LRMP LRP10 LRRC39 LRRC56 LRRC8C LRRK1 LSR LTA4H LTB4R LTC4S LUZP1 LY75 LY86 LY96 LYL1 LYN MAG MAGT1 MAL MAN2A1 MAN2B2 MANBA MAP1B MAP3K11 MAP3K8 MARVELD2 MAST3 MAT2A MATK MATN3 MB MBNL1 MBNL2 MCCC1 MCOLN2 MED8 MEF2A MERTK METTL12 MFAP5 MFNG MFSD1 MGAT1 MGC24103 MGC45922 MIR30C2 MLKL MLXIPL MMEL1 MMP11 MNDA MOBKL1B MOBKL2C MOBP MOG MPPE1 MRAS MRC1 MRPS18C MRVI1 MS4A14 MSLN MTHFD2 MTIF3 MTMR10 MTO1 MTUS1 MUC1 MUC5AC MUTED MYBPH MYCBP2 MYH7B MYO1F MYOM1 MYOT MYOZ1 MYOZ3 NADSYN1 NAIP NAPSA NAPSB NAT1 NCEH1 NCF4 NCOA7 NCRNA00202 NDST2 NEAT1 NEDD9 NEFH NEIL1 NEK3 NEK9 NEURL2 NEXN NFE2L2 NFKBIA NHS NINJ2 NIPA1 NIPAL4 NKAIN2 NLRC3 NLRC5 NLRX1 NMI NOD2 NOS1AP NOSTRIN NOX1 NPL NPY NQO1 NRG3 NT5DC1 NTF3 NUDT14 NUDT18 NUDT7 NUDT9P1 OBFC2A ODF3B OMA1 OPLAH OPTN OR1I1 OR7A5 OSBPL11 OSBPL9 OSCAR OSTF1 OXTR P2RY12 P2RY13 P2RY14 P2RY8 PACS2 PALMD PAOX PARP10 PARP14 PARP9 PARVG PATL2 PBLD PCK1 PCTP PCYT2 PDCD4 PDE6B PDHB PDK4 PEBP4 PECAM1 PELI3 PER3 PET112L PEX11G PEX14 PEX5L PFKFB2 PGAM2 PGM5 PHF11 PHF15 PICALM PIGV PIK3AP1 PIK3IP1 PIK3R5 PILRA PLA1A PLA2G15 PLA2G4C PLA2G7 PLAU PLCB2 PLCD4 PLCG2 PLCH2 PLCXD3 PLD4 PLEKHA1 PLEKHA7 PLIN3 PLN PLSCR1 PLSCR4 PML PNPLA7 PNRC1 PODXL POLD4 POLN POM121L9P POU2F2 PPA2 PPAP2A PPCS PPFIBP2 PPIL4 PPT1 PRAM1 PRB1 PRB4 PRDX3 PRKCH PROCR PRRG4 PRRT3 PSME1 PSTPIP2 PTAFR PTGFR PTP4A2 PTPLAD2 PTPN2 PTPN22 PTPN6 PTPRB PTPRCAP PUS10 PVRIG PWWP2B PYGM PYHIN1 PZP RAB40B RANBP2 RAP1A RARRES3 RASAL3 RASD1 RASGEF1C RASGRF2 RASGRP3 RASSF4 RBKS RBM41 RBP7 RCAN2 RCAN3 RCBTB1 RDH12 REM1 RETN RFPL1S RFTN1 RGL1 RGS10 RGS18 RGS8 RHBDF2 RHOG RHOH RIPK3 RNASE2 RNASET2 RNF125 RNF130 RNF144B RNF213 RNF6 RNPC3 ROBO3 ROM1 ROPN1 RPE RPS6KA1 RPS6KA2 RRAD RTP4 RYR3 S1PR5 SAA1 SAMD3 SAMSN1 SASH3 SCN1B SCN5A SCN7A SCYL1 SDCCAG1 SDPR SDS SEC14L2 SEC14L5 SELE SELPLG SEMA4D SEPT1 sept4 sept15 SERPINB2 SERPINB4 SERPING1 SESN2 SETDB2 SFN SFRP5 SFXN2 SFXN4 SFXN5 SGK2 SGK3 SGPP1 SH2D1A SH2D5 SH3BP1 SH3GL2 SH3TC1 SIDT1 SIGLEC1 SIGLEC10 SIGLEC8 SIGLEC9 SIPA1 SIRPB1 SIRT5 SIT1 SLA SLC11A2 SLC13A3 SLC15A2 SLC25A18 SLC29A2 SLC29A3 SLC2A12 SLC30A1 SLC31A2 SLC35F4 SLC39A11 SLC44A3 SLC45A3 SLC46A3 SLC48A1 SLC5A11 SLC6A13 SLC9A9 SLCO2B1 SLPI SMPD1 SMPDL3A SMPX SNAI3 SNCA SNCG SNTG2 SNX18 SNX20 SORBS1 SORL1 SORT1 SOSTDC1 SP100 SP140 SPINT1 SPINT2 SPOCD1 SPOCK3 SST SSU72 ST6GALNAC3 STAR STARD13 STARD8 STAT1 STAT4 STK3 STK40 STOM STX11 STXBP2 STXBP3 SUSD3 SYK SYNGR2 SYNPO2 SYT16 SYTL2 SYVN1 TAGAP TAL1 TAP2 TAPBPL TARP TARP TBC1D10C TBC1D12 TBC1D22A TBC1D30 TBC1D4 TBC1D8B TBXAS1 TDGF1 TEP1 TEX2 TFEB TFEC TGFBR2 TGFBR3 TGM2 TGM3 TGOLN2 THEMIS TIPARP TLR2 TLR3 TLR5 TLR6 TLR7 TLR8 TM2D2 TM6SF1 TMC6 TMC8 TMCO4 TMED5 TMEM109 TMEM125 TMEM140 TMEM144 TMEM149 TMEM150A TMEM151A TMEM156 TMEM173 TMEM217 TMEM63A TMEM71 TNF TNFAIP3 TNFAIP8L2 TNFRSF14 TNFSF10 TNFSF12 TNFSF12 TNFSF13 TPD52L1 TPH1 TPRG1L TPST1 TRAC TRAT1 TRBC1 TREML1 TRIM22 TRIM34 TRIM38 TRIM47 TRIM56 TRPM2 TRPV2 TSPAN15 TSPAN32 TST TTLL7 TWSG1 TXNDC11 TXNRD3 TYMP UBA7 UBE2L6 UCP2 UHMK1 UMODL1 USP53 UTF1 UVRAG VAMP1 VAMP3 VAV1 VPS26A VPS4B WARS WBP2 WDFY4 WDR11 WDR37 WDR49 WIPI1 WNT2 WWP1 XAF1 XCL1 /// XCL2 ZAP70 ZBTB16 ZCCHC2 ZCCHC5 ZFHX3 ZFP36 ZFYVE26 ZFYVE28 ZMYM6 ZMYND15 ZNF385B ZNF436 ZNF846 ZRANB2 es