{"paper_id":"2e02234e-2858-4dd5-ae93-5dd561f28b9b","body_text":"IL-8 Instructs Macrophage Identity  \nin Lateral Ventricle Contacting Glioblastoma \n \nStephanie Medina1,2,3, Asa A. Brockman 1, Claire E. Cross 1,2,3, Madeline J. Hayes 1,2,4, Bret C Mobley 2,4, \nAkshitkumar M. Mistry 4,5, Silky Chotai 4,5, Kyle D Weaver 4,5, Reid C Thompson 4,5, Lola B Chambless 4,5, \nRebecca A. Ihrie1,4,5,6, Jonathan M. Irish1,2,3,4 \n \n1 Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN, USA. \n2 Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, \nNashville, TN, USA. \n3 Vanderbilt Center for Immunobiology, Vanderbilt University Medical Center, Nashville, TN, USA. \n4 Vanderbilt-Ingram Cancer Center, Vanderbilt University Medical Center, Nashville, TN, USA. \n5 Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA. \n6 Vanderbilt Brain Institute, Vanderbilt University, Nashville, TN, USA. \n \nCorresponding author email: jonathan.irish@vanderbilt.edu (J.M.I.) \n \nRunning Title: IL-8 macrophages predominate in aggressive GBM tumor microenvironments \n- IL-8 is expressed by GBM cells and enriched in lateral ventricle-contacting tumors \n- M_IL-8 macrophages are CD32 + HLA-DR++ CD163+ CD206+ CD86- PD-L1-  \n- M_IL-8 macrophages instructed with IL-8 or GBM conditioned medium match human glioblastoma \nassociated macrophages \n- IL-8 is necessary for GBM tumor cells  to generate M_IL-8 macrophages  \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nAbstract \nAdult IDH-wildtype glioblastoma (GBM) is a highly aggressive brain tumor with no established \nimmunotherapy or targeted therapy.  Recently, CD32 + HLA-DR hi macrophages were shown to have \ndisplaced resident microglia in GBM tumors that contact the lateral ventricle stem cell niche.  Since these \nlateral ventricle contacting GBM tumors have especially poor outcomes, identifying the origin and role of \nthese CD32 + macrophages is likely critical to developing successful GBM immunotherapies.  Here, we \nidentify these CD32+ cells as M_IL-8 macrophages and establish that IL-8 is sufficient and necessary for \ntumor cells to instruct healthy macrophages into CD32+ M_IL-8 M2 macrophages. In ex vivo experiments \nwith conditioned medium from primary human tumor cells, inhibitory antibodies to IL-8 blocked the \ngeneration of CD32 + M_IL-8 cells.  Finally, using a set of 73 GBM tumors, IL-8 protein is shown to be \npresent in GBM tumor cells in vivo and especially common in tumors contacting the lateral ventricle. These \nresults provide a mechanistic origin for CD32+ macrophages that predominate in the microenvironment of \nthe most aggressive GBM tumors.  IL-8 and CD32 + macrophages should now be explored as targets in \ncombination with GBM immunotherapies, especially for patients whose tumors present with radiographic \ncontact with the ventricular-subventricular zone stem cell niche.   \n \nKeywords \nGlioblastoma, IL-8, CD32, M2, macrophages, V-SVZ stem cell niche, cytometry \n  \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nIntroduction \nGlioblastoma (GBM) is the most aggressive and most common primary brain tumor among adults, \nwith a poor prognosis of less than 2 years 1-3. While immunotherapies have revolutionized the treatment \nof many solid tumors including melanoma, lung, and kidney cancer, GBM tumors remain resistant to \nimmunotherapies, in part due to an immune tumor microenvironment (TME) that is refractory to these \nagents 4,5. \nGlioblastoma-associated macrophages and microglia are the most abundant immune cell type \ninfiltrating tumors and can often represent up to half of all the cells of the tumor mass 6-10. The majority of \nGBM tumor macrophages are presumed to originate from circulating blood monocytes 11. Upon entry into \nthe TME, these tumor macrophages adopt new cellular and molecular identities that are critical for GBM \nprogression. Macrophages can contribute to immune suppression through the expression of surface \nproteins like CD163, a marker of immunosuppressive macrophages that contributes to T cell dysfunction \nand has been associated with poor prognosis in GBM 12-14.  High dimensional immune cell profiling in GBM \nrecently identified a subset of CD32 + HLA-DRhi monocyte derived macrophages that were distinguished \nby a potentiated response to inflammatory cytokine signaling via p-STAT3. The abundance of CD32+ GBM \nassociated macrophages (GAMs) independently stratified patient survival, and these GAMs predominated \nin the microenvironment of tumors that contacted the lateral ventricles, the location of the ventricular-\nsubventricular zone (V-SVZ) stem cell niche 15.  V-SVZ contact by GBM tumors is established as closely \nassociated with poorer clinical outcomes for patients16,17.  Reprograming immunosuppressive GAMs into \nmore inflammatory states has been shown to have potential clinical benefit 18-20.  Thus, understanding the \nmechanisms that generate CD32+ GAMs in V-SVZ-contacting human tumors could provide a strategy to \nrehabilitate the immune microenvironment of the most aggressive subtype of a deadly brain tumor.    \nTumor associated macrophages play diverse roles in cancer development and tumor progression \nleading to poor prognosis of many solid tumors beyond GBM 21, such as breast 22, head and neck 23, \nbladder24,  melanoma 25, and prostate cancer 26. Thus far, murine models and patient derived xenograft \nmodels have been used to understand the heterogeneous phenotypic states of infiltrating microglia and \nmacrophages in GBM 27-31.  However, these models may not reflect all aspects of human immunology and \nmay not reflect structural features of the human brain or GBM tumors.  For example, CXCL8, the gene for \nhuman IL-8 protein, is one of the 20% of human genes that lacks a mouse ortholog. It is vital to \ncontinuously improve preclinical models, and one area for urgent attention is to ensure that the immune \nmicroenvironment of different models closely reflects that observed in human tumors.   \nHistorically, healthy macrophage identity has been presented as aligned to one of two extremes: \nM1-like macrophages that promote an inflammatory immune response or M2-like macrophages that \npromote a suppressive immune response 32.  However, the last decade of research has revealed and \ncharacterized a spectrum of macrophage identities that can be tracked through surface proteins which are \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nclosely linked to diverse functions and activation states 33,34.  Current state of the art approaches define \nmacrophage activation states based on surface immunophenotype and name macrophage states based \non the cytokines leading to their specialization.  For example, M1-like, IFNγ polarized macrophages \n(M_IFNγ) are distinguished by elevated expression of CD86, a surface protein that provides costimulatory \nsignals promoting T cell activation and survival 35, PD-L1 36, the programed cell death ligand receptor, and \nlack of expression of scavenger receptor CD163 and mannose receptor CD206 33,37. In contrast, M2-like \nmacrophages can include those polarized by IL-6 or by IL-4 (M_IL-6 or M_IL-4). These macrophage \nactivation states both express high levels of surface CD163 and CD206 proteins, but M_IL-6 express \nhigher levels of the FCγRII CD32 34,38. In addition, myeloid derived suppressor cells (MDSCs) are \ndistinguished functionally by their ability to suppress T cell proliferation and are characterized by low \nexpression of cell surface HLA-DR protein39. \nThe tumor microenvironment can produce cytokines/chemokines, which are involved in the \nrecruitment of normal cells to promote growth, invasion, angiogenesis, and metastasis of glioblastoma. \nThe exact cytokines that are secreted from glioblastoma tissue may vary depending on the specific case \nand the stage of the disease 40. While it is established that macrophages can respond distinctly to select \nstimuli in their environment 32, other cytokines, such as IL-8, have not been studied as extensively in the \ncontext of macrophage polarization. Therefore, it is not clear whether a monocyte derived M_IL-8 cell \nexists, whether it ‘leans’ to M1 or M2, and whether it is phenotypically or functionally distinct from M_IL-6 \nmacrophages or other subtypes. \nIL-8 was originally described as a chemokine whose main functions are generally reported to be \nattraction of neutrophils via receptors CXCR1 and CXCR2 41,42.  It is now known to play key roles in wound \nhealing, angiogenesis, inflammation, and tumor growth 43,44, and it has been most studied in cancer in the \ncontext of epithelial origin tumors45.  In GBM, IL-8 was reported to be highly expressed and to play a role \nin regulating GBM stem cells, promoting tumor growth, and promoting angiogenesis 46-49.  Previous work \nfocused on IL-8’s ability to activate vascular mimicry in tumor cells, including after treatment with alkylating \nchemotherapy temozolomide 50.  Il-8 has been negatively correlated with glioma patient survival 51,52, but \nhas not previously been linked to macrophages in GBM or to the aggressive subtype of GBM that presents \nin contact with the V-SVZ.   \nHere, we use ex vivo culture of primary GBM tumor cells and healthy blood derived macrophages to model \nthe generation of suppressive CD32+ GAMs.  IL-8 is identified as the primary factor directing macrophage \nidentity in human GBM tumors, establishing a novel role for IL-8 in the tumor microenvironment and in \nGBM.  \nResults  \nHealthy blood macrophages polarized ex vivo by IL-6 contrast with macrophages in human GBM \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nIn this study, GBM tumor conditioned medium (GBM_TCM) was added to the field-standard macrophage \npolarization assay used to generate inflammatory and suppressive macrophages from healthy monocytes \n37,53.  The goal of this experiment was to establish whether macrophages polarized with GBM_TCM \n(M_GBM_TCM) closely resemble those observed in vivo  in human GBM tumors (GBM-associated \nmacrophages, GAMs).  GBM_TCM for macrophage polarization was created by culturing dissociated \nsingle cells from primary human glioblastoma tumors for 3 days (see Methods and 37 for additional detail).   \nAfter 3 days of culturing monocytes in the presence of macrophage colony stimulating factor (M-CSF), the \nresulting macrophages were stimulated with GBM TCM for an additional 3 days, analyzed using spectral \nflow cytometry, and their phenotype compared to GAMs or to canonical healthy macrophage subtypes \nrepresenting M1 (M_IFNγ) and M2 (M_IL-6).  \nTo determine whether M_GBM_TCM \nmacrophages model the generation of \npreviously described GAMs, we performed \nMarker Enrichment Modeling (MEM) analysis \n37,54 using published cytometry data from \nprimary human GBM macrophages 15 and \nnewly generated data from macrophages \nsubjected to different stimulatory conditions, \nincluding TCM, M2 cytokines IL-6 and IL-4, \nand M1 control cytokine IFNγ 33. Following \nMEM, a ΔMEM analysis 55, which subtracts two \nMEM labels to identify differentially enriched \nfeatures, was performed to quantify and \ncompare changes in protein expression \nbetween macrophages. Published cytometry \npanels and newly generated data panels \nshared 4 features that we compared in this \nanalysis: CD64, CD32, HLA-DR/MHC II, and \nPD-L1.  Of these, CD32 and HLA-DR were signature features that distinguished C-GBM GAMs, and the \nclassic M1 feature PD-L1 was observed to be missing on C-GBM GAMs. The newly generated data also \nmeasured CD86 (M1), CD206 (M2), and CD163 (M2) macrophage markers.  The ΔMEM analysis showed \nthat TCM produces cells that differ from IL-6 and other conditions but were a close match for GAMs \nobserved in C-GBMs.  While M_GBM_TCM were similar to GAMs in expression of HLA-DR, CD64, and \nCD32 proteins, GAMs generally had lower PD-L1 expression compared to M_GBM_TCM ( Box 1).  In \ncontrast, IL-6 or IFNγ stimulations did not lead to as much CD163 or CD206 expression as TCM. IL-6 did \nnot trigger expression of CD32, HLA-DR, or CD64 and IFNγ led to much higher expression of CD86 than \nTCM (Box 1).  In addition to the markers shared between panels, MEM was used to evaluate features \nBox 1 – Quantitative comparison of proteins expressed on GAMs, \nM_IL-6, or M_IFNγ to M_GBM_TCM macrophages using MEM. \nProteins differing between GAMs vs. M_GBM_TCM:  \n  ΔPD-L1-4 HLA-DR-2 CD64-2 CD32+2  \n[GAM proteins measured in a single panel: CD45+9 CD45RO+7 \nCD206-7 CD163 -7 CD44 +5 CCR4 +4 CD14 +4 CD69 +4 CD86 -4 \nCD11b+3 CXCR3 +3 CD43 +2 CD33 +2 CD28 +2 CD56 +2 TIM3 +1 \nCD4+1 CD38+1 TCRgd+1 CD16+1 CD3+1 CD57+1 \nProteins differing between M_IL-6 vs. M_GBM_TCM:  \n  ΔCD163-7 CD206-6 HLA-DR-5 CD64-4 CD32-4  \n[M_IL-6 proteins measured in a single panel:  CD11c+8 CD45+7 \nCD43+6 CD33 +6 S100A9APC +6 CD36 +5  PD-L1 -5 CD164 +4 \nCD13+4 CD9 +4 CD11b +3 Tim3 +2 Slan +1 CCR2 +1 CD123 +1 \nCD14+1 CD19 +0 MerTK +0 CD45RA +0 CD27 +0 CD16 +0 \nCD120a+0 CCR7 +0 CD8 +0 CD25 +0 CD3 +0 CD68 +0 PD1 +0 \nCD274+0 CD127+0 \nProteins differing between M_IFNγ vs. M_GBM_TCM:  \n  ΔCD163-7 CD206-6 CD86+3 CD64-2 CD32-2 HLA-DR-2 \n[M_IFNγ proteins measured in a single panel:  CD11c +7 \nS100A9APC+6 CD45 +6 CD43 +5 CD33 +5 PD-L1 -5 CD164 +4 \nCD13+4 CD11b +3 CD36 +2 CCR2 +2 CD14 +2 CD9 +2 TIM3 +1 \nSLAN+1 CD123+1 CD19+0 MerTK+0 CD45RA+0 CD27+0 CD16+0 \nCD120a+0 CCR7 +0 CD8 +0 CD25 +0 CD3 +0 CD68 +0 PD1 +0 \nCD274+0 CD127+0 \nFeatures with a significant difference (>3) between TCM polarized \nmacrophages and GAMs, M_IL-6, or M_IFNγ are shown in bold. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nenriched or missing from these macrophage populations (Box 1, proteins measured in a single panel).  \nTaken together, these results indicated that IL-6 and other tested cytokines alone were not sufficient to \ngenerate macrophages with the phenotype observed in vivo in GBM tumors.  \nHealthy blood macrophages polarized ex vivo with tumor conditioned media are CD32 + M2 cells \ncomparable to primary GBM macrophages \nHaving established key features expressed by macrophages generated with GBM TCM, the goal was next \nto understand percent positivity for different proteins using a traditional gating strategy.  Using manual \nexpert gating, polarized macrophages were analyzed to quantify the amount of CD206 and CD163 positive \ncells, CD32 and CD163 positive cells, CD163 positive and CD86lo expressing cells, and CD163 negative \nand PD-L1hi expressing cells (Figure 1A). Unpolarized monocytes expressed low levels of all markers \nand contained less than 5% of cells positive for any markers assessed. Less than 12% of M_IFNγ \nmacrophages were found to be CD163 +CD32+CD206+ and CD86 lo. However, over 80% of M_ IFNγ \nmacrophages were gated as CD163 - PD-L1hi cells, which is expected for M1 macrophages. M_IL-6 and \nM_GBM_TCM macrophages consistently expressed M2 phenotype markers: over 39% of cells were \nCD206+ CD163+, over 34% of cells were CD163+CD32+, over 26% of cells were CD163+ CD86lo, and less \nthan 1% were CD163 - PD-L1 hi for all conditions (Figure 1B) . This established that in addition to \nphenocopying GAMs, M_GBM_TCM macrophages exhibit a suppressive M2 like phenotype.  \nIL-8 was produced by cells from all tested primary human GBM tumors. \nTo learn more about potential mechanisms by which M_GBM_TCM macrophages are polarized into a \nsuppressive phenotype it was important to dissect the factors that are secreted by tumor cells during ex \nvivo culture. The goal of this experiment was to identify any potential candidate mediators of macrophage \npolarization in GBM. To identify secreted proteins present in GBM TCM, an array analysis testing for 105 \ndifferent soluble factors and cytokines was performed on TCM from 7 primary human tumor samples \n(Figure 2 ). Unconditioned media was used as a negative control (Figure 2A) . Quantification of dot \nintensity density, which is relative to the quantity of protein present in any sample, revealed 6 potential \ncandidates whose median secretion was above the calculated threshold of significance across all 7 \ntumors: Emmprin, CXCL8/IL-8, Macrophage migration inhibitory factor (MIF), Matrix metalloproteinase-9 \n(MMP9), Osteopontin, and Serpin E1 (Figure 2B) Only IL-8 was consistently secreted at a high level by \ncells from all tested GBM tumors (Figure 2B, dark blue dots, N=7).  \nBlocking IL-8 abrogates M2 macrophage polarization by GBM tumor cells. \nAfter identifying IL-8 as the most abundant and reproducible cytokine secreted into TCM across tumors, \nwe hypothesized that IL-8 is necessary and sufficient to generate CD32+ M_GBM_TCM macrophages ex \nvivo.  To test this hypothesis, macrophages were polarized over 3 days in culture in the presence of either \nrecombinant IL-8 or GBM TCM and with or without a monoclonal blocking antibody against IL-8 (α-IL-8), \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nwhich remained present throughout macrophage polarization. Using spectral flow cytometry, the \nexpression of signature M_GBM_TCM surface markers was measured.  \nTo gain a visual understanding of how macrophage identity shifts across different polarization conditions, \na t-SNE analysis of all polarized macrophages based on all measured surface marker features was \nperformed. t-SNE plots displaying the cellular density across the map revealed that both M_IL-8 and \nM_GBM_TCM macrophage conditions lacking α-IL-8 predominantly clustered towards the top left of the \nt-SNE map (>57%). Macrophages polarized in the presence of an IL-8 blocking antibody (+ α-IL-8) were \nfound to predominately cluster towards the bottom right of the t-SNE map (>58%) (Figure 3A).  \nTo quantify differences in phenotypes, a T-REX analysis 56 was performed to compare the macrophage \nphenotypes that were generated in the presence or the absence of α-IL-8 (Figure 3B) . T-REX plots \nrevealed condition-specific clusters of macrophages enriched when macrophages were polarized in either \nthe presence (colored in blue) or absence (colored in red) of α-IL-8. MEM analysis of these condition \nspecific clusters revealed that cells in, cluster 1 which were enriched in conditions lacking α-IL-8 expressed \nhigher levels of HLA-DR, CD206, CD163, and CD32 but lower levels of CD86 in comparison to cells in \ncluster 2 which were enriched in conditions containing α-IL-8.  \nTo get a closer look at specific expression of GAM signature proteins, expert gating was used to quantify \nCD163+ CD32+ CD86lo cells in conditions with or without the addition of α-IL-8 (Figure 3C). M_IL-8 and \nM_GBM_TCM macrophages displayed over 35% positivity CD163 + and CD32 + cells, and over 26% \npositivity for CD163 +CD86lo cells. Upon addition of α-IL-8 blocking antibody, the percentage of CD163 + \nand CD32+ and CD163 +CD86lo was reduced to under 23% for M_IL-8 cells and under 17% of cells for \nM_GBM_TCM conditions.  \nFinally, histograms were used to view the effects of α-IL-8 on the expression of individual protein markers \n(Figure 3D). This highlighted how blocking IL-8 in TCM prevents the expression of M2/GAM markers \n(CD32 and CD163) in polarized macrophages. In addition, blocking IL-8 promotes a higher expression of \nM1 marker CD86, but has little to no effect on PD-L1 expression. These results support two major findings: \n1) IL-8 on its own is sufficient to polarize macrophages towards a phenotype that is similar to \nM_GBM_TCM, and IL-8 in GBM TCM is necessary for the polarization of M_GBM_TCM macrophages.  \nIL-8 is common in primary tumors and more highly expressed in C-GBM tumors. \nTo confirm that IL-8 secretion also occurs in vivo in primary human GBM tumors, immunohistochemical \n(IHC) staining for CXCL8 (IL-8) was performed on a tumor microarray (TMA) composed of primary tumor \nsamples from 73 IDH-wt GBM patients. To evaluate IL-8 expression across primary GBM tumor samples, \nthe percentage of IL-8 positive pixels per cell and per core was quantified for every patient core (3 \ncores/patient, 73 patients total). A wide range varying from 0%-50% positive pixels of IL-8 signal were \ndetected across cells and patients (Figure 4 A-B). Next, to evaluate IL-8 expression patterns across tumor \ntissue, cellularity per core was quantified and plotted against % positive pixels. In addition, visual validation \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nconfirmed IL-8 staining patterns were predominantly extracellular as opposed to nuclear or cellular (Figure \n4C).  This confirmed that IL-8 secretion is a feature that was conserved from in vivo tumors in the ex vivo \nGBM model system. \nPresuming that M_GBM_TCM macrophages model CD32+ GAMs previously described as a \ndistinguishing immune cell subset for tumors that contact the V-SVZ (C-GBM), we hypothesized that if IL-\n8 is necessary for the generation of M_GBM_TCM macrophages, then IL-8 would be more abundantly \nexpressed in C-GBM. To test this hypothesis, every tumor sample included in the TMA was \nradiographically scored and classified as contacting (C-GBM) or non-contacting (NC-GBM). The \npercentage of positive pixels per core was quantified and compared between C-GBM cores and NC-GBM \ncores, revealing that IL-8 was more significantly expressed in C-GBM tumors (p<0.001, N=73) (Figure \n4C).  \nImages of representative cores were then chosen to determine whether IL-8 protein expression was \nprimarily overlapping with non-immune tumor cells or immune cells. Strikingly, when IL-8 was expressed \nat high levels, it was apparently present in non-immune tumor cells ( Figure 4).  Taken together, these \nresults indicate that in human GBM tumors that contact the V-SVZ, GBM cells produce IL-8 which shifts \nincoming macrophages into CD32 + M2 macrophages highly similar to M_IL-8s produced ex vivo by IL-8 \nand TCM. \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nDiscussion \nImmune cells are known to play a critical role in the development and the progression of tumors. \nThus, tumor immune evasion has become recognized as a hallmark of cancer 57,58. The modulation of \nimmune cells represents one of the main driving features of GBM59, and GBM subtypes have been shown \nto modulate their TME through the aberrant secretion of multiple factors60, especially including those that \ncontribute to immunosuppression 61. Myeloid derived macrophages, brain resident microglia, dendritic \ncells, and myeloid derived suppressor cells (MDSCs) are the main components of the GBM TME60. Since \nmacrophages are known to respond to stimuli in their microenvironment, we hypothesized that a GBM \nsecreted soluble factor might play a role in instructing the aggressive macrophages that have been shown \nto drive the immunosuppressive microenvironment of GBM. We predict that these suppressive \nmacrophages may be functioning at the end of the cancer immunity cycle by suppressing T cell activities \nthat would ultimately lead to cancer cell death 58. The focus of this study was to understand how GBM \ntumor secreted factors instruct the functional identity of previously described, survival stratifying, \nimmunosuppressive GAMs. In this study, an ex vivo cell culture model was established that phenocopied \nGAMs that predominate in tumors that contact the V-SVZ neural stem cell niche.  These GAMs were \noriginally described to have M2-like suppressive features, a finding that was confirmed here.  Assessment \nof GBM secreted proteins identified IL-8 as a key factor necessary for instructing this macrophage identity \nin GBM.  \nUpon initial assessment of marker expression of polarized macrophages, we anticipated that a \nfactor such as IL-6 or IL-4 would be secreted across GBM samples, as these cytokines have been well \ncharacterized for their role in instructing macrophages towards M2-like suppressive identities 62,63. In \naddition, when assessing the protein expression of markers in the M_IL-6 and M_GBM_TCM macrophage \nphenotypes, we noticed that these two classes shared the expression of some key surface markers (Figure \n1A). However, we unexpectedly identified IL-8 as the most robustly secreted molecule among the 105 \ndifferent soluble proteins that were tested (Figure 2), and we observed that IL-6 was no better than IFNγ \nand far worse that IL-8 and TCM at producing macrophages like those observed in vivo. \nCytokines like IL-6 and IL-8 have been identified as a part of the molecular signature of cytokines \nsecreted by GBM tumors 64,65. Yet so far, literature surrounding IL-8 in cancer has focused on its role in \nneutrophil recruitment 66 and angiogenesis 67 during tissue remodeling and inflammation 68. In a healthy \nwound healing response setting, activated macrophages, endothelial cells, and epithelial cells can \nproduce IL-8 in response to infections or tissue injury. IL-8 can function as a chemoattractant for \nneutrophils who may form extracellular traps to kill invading microbes 69, and in endothelial cells IL-8 \nsignaling can induce cell proliferation, survival and migration leading to angiogenesis 70. IL-8 expression \nhas been previously detected and linked to tumor progression in several types of cancers including breast, \ncolon, ovarian, bladder, and prostate cancers, as well as in melanoma 71-74. In GBM especially, studies \nhave focused on describing the role of IL-8 in promoting glioblastoma stem cells48, which are also thought \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nto be major drivers of aggressive GBM tumors, cell migration signaling 75, and angiogenesis 46,50,76.  \nPreclinical studies have suggested that blocking antibody treatments against IL-8 could potentiate greater \nefficacy of immunotherapies in GBM 77. While GBM secreted IL-8 has been shown to be an important \nplayer driving GBM cancer cell behavior52, a major gap in the field remained when it came to understanding \nhow its presence in the TME could be affecting infiltrating macrophages. In fact, when looking further to \nunderstand how IL-8 affects macrophages in a healthy tissue remodeling setting, we found that thus far \nIL-8 has solely been described as a cytokine that is secreted by macrophages, and the features of a \nmacrophage response to this cytokine remained largely unknown until this study. Here we show that IL-8 \nis secreted both in vivo and ex vivo by GBM tumor cells, and we show that it is necessary for polarization \nof M_GBM_TCM macrophages. Ultimately, these results indicate that IL-8 should be further studied as a \nkey determinant of the microenvironment in GBM and other cancer types, as IL-8 and its receptors may \nrepresent suitable targets for therapies. \nWhile this study describes the surface protein expression of key markers activated by IL-8 and \nTCM in polarized macrophages, further studies are necessary to continue to fully describe the impact of \nIL-8 on polarization of healthy microglia, monocytes and macrophages. Future studies of M_IL-8 \npolarization should assess functional tests such as antigen presentation, T cell activation/suppression in \na mixed leukocyte reaction, and cytokine secretion. In addition, phospho-flow cytometry could be used to \ndissect downstream signaling pathways that are mechanistically involved in M_IL-8 polarization. \nIt is well established that immortalized cancer cell lines, which are often used as surrogates for \nhuman tumors, provide a poor reflection of the diverse profiles of human patients’ tumors78. GBM tumors \nare especially notorious for their extensive genetic, epigenetic, intratumoral and intertumoral \nheterogeneity79 which is often depicted as one of the major challenges for understanding the cellular and \nmolecular underpinnings driving GBM. It has been shown that the use of patient derived primary tumor \ncells can provide a model that more closely reproduces the in vivo human tumor microenvironment80,81. In \nattempt to understand GBM heterogeneity, transcriptomic analysis of primary tumor samples was recently \nused to describe 4 distinct transcriptional states: neural progenitor (NPC)-, oligodendrocyte progenitor \n(OPC)-, astrocytic (AC)-, and mesenchymal (MES)-like 82. Other models such as the use of organoids and \npatient derived xenograft models have also been very useful tools for dissecting GBM tumor heterogeneity \nand are increasingly utilized in neuro-oncology research for preclinical studies 83,84. In this study, the use \nof primary human GBM tumor cells and healthy blood monocytes provided a model that yielded results \nwith in vivo human relevance. Given that even immune competent mouse models don’t express an IL-8 \nequivalent, the ex-vivo culture of primary cells provided an appropriate and representative model in this \nstudy.  \nThis work suggests a new mechanism for IL-8 in glioblastoma and implicates CD32+ macrophages \nas a key feature of the aggressive, immunosuppressive immune microenvironment in GBM.  Clinical \nstudies should now explore targeting of IL-8, especially for patients with V-SVZ contacting GBM.   \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nMethods \nTissue Collection of Human Specimens and Processing  \nSurgical resection specimens of IDH- wild type glioblastomas that were collected at Vanderbilt University \nMedical Center between 2014 and 2023 were processed into single cell suspensions following an \nestablished dissociation protocol 85. All samples were collected with patient informed consent in \ncompliance with the Vanderbilt Institutional Review Board (IRB #131870), and in accordance with the \ndeclaration of Helsinki. \nPatients were adults (≥18 years of age) at the time of surgical resection. Resections were classified as \ngross or subtotal resections by a neurosurgeon and a neuroradiologist. Tumor contact status in relation to \nthe lateral ventricle (V-SVZ) was determined by a neurosurgeon and a radiologist based on pre-operative \nradiographic magnetic resonance imaging (MRI) of the brain, as detailed in Mistry et al. 86,87.  \nEx vivo culture of GBM tumor cells and generation of tumor conditioned media \nCryopreserved samples of single cell dissociated GBM tumor cells were cultured for 3 days in ultra-low \nattachment plates at a density of 2×10 5 cells/ml in a humidified atmosphere at 37°C, 5% CO2 in \nmacrophage polarization medium (RPMI 1640 enriched with FBS 10% and supplemented with 1% \nPenStrep solution 37,88). The resulting tumor conditioned medium was removed from the cells via \ncentrifugation and stored in 500 µl aliquots at -20 °C for future experiments. \nMacrophage polarization \nPeripheral blood mononuclear cells (PBMCs) from healthy donors were obtained commercially (Stem Cell \nTechnologies). For in vitro macrophage polarization experiments, healthy macrophages were obtained by \ndifferentiating healthy monocytes. For ex vivo differentiation of monocytes, cells were cultured in 6 ‐well \nplates at 2 × 106 cells/ml in a humidified atmosphere at 37°C, 5% CO2 in RPMI 1640 enriched with FBS \n10% and supplemented with 1% Pen/Strep solution. To purify monocytes from PBMCs via plate adhesion, \nPBMCs were allowed to adhere to the plate for 3 hours to enrich monocytes; the cells in suspension were \ndiscarded. To activate macrophage differentiation, monocytes were stimulated with macrophage colony \nstimulating factor (M‐CSF) (50 ng/ml) for 3 d in standard tissue culture, as previously described34,37. Then, \nmacrophages were further polarized for 3 d by IL ‐6 (10 ng/ml), IFN‐g (10 ng/ml), or a mixture composed \nof 50% tumor conditioned media (GBM_TCM) (see TCM section) and 50% fresh media supplemented \nwith M-CSF. At the end of the polarization, wells were treated with Accutase (Sigma ‐Aldrich) prewarmed \nat 37°C for 30 s before collection of cells for further analysis. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nIL-8 blockade in ex vivo studies \nTo block IL-8 during macrophage polarization, a monoclonal antibody against IL-8(ɑ-IL-8) (clone #: 6217) \nwas used. Aliquots of tumor conditioned media were thawed and incubated with 0.8µg/mL of ɑ-IL-8 for 30 \nminutes at 37°C prior macrophage stimulation. TCM was then combined with complete macrophage media \nand added to differentiated macrophages to allow for polarization in culture over the next 3 days. \nMacrophages were then collected accordingly with other conditions for analysis via spectral flow \ncytometry.  \nSpectral flow cytometry \nPolarized macrophages were collected post Accutase treatment and transferred to FACS tubes for live \nstaining. Cells were washed once with PBS and once with PBS BSA. Cells were stained for 30 minutes \nwith an antibody cocktail which included CD86 (IT2.2), CD64 (10.1),CD206 (EPR6828(B)), HLA-DR \n(I243), PD-L1 (29E.2A3), CD163 (GHI/61), CD32 (FUN-2). During the last 5 minutes of staining, the \nviability dye Alexa 700 SE dye (NHS) was added. Cells were washed once with PBS and once with PBS \nBSA. After staining was complete, cells were analyzed on a CyTEK spectral flow cytometer within 30-60 \nminutes.   \nCytokine arrays \nThe Proteome Profiler Human XL Cytokine Array Kit (R&D Systems) was used to measure the secretion \nof 105 different soluble factors and cytokines in GBM tumor conditioned media. 500µl of tumor conditioned \nmedia from each patient sample were analyzed per the manufacturer’s instructions on a nitrocellulose \nmembrane and then visualized using chemiluminescent detection reagents and the iBright imaging \nsystem. \nHuman tumor microarray \nGeneration of a tumor microarray (TMA) of formalin fixed paraffin embedded (FFPE) glioblastoma \nspecimens was described in Leelatian et. al. 89. Briefly, three 1mm areas were selected from each tumor \nsample by a neuropathologist. Blocks were delivered to the Vanderbilt University Medical Center TPSR \n(Translational Pathology Shared Resource), where cores were extracted from the encircled areas using \nthe Tissue Microarray Grandmaster (3DHistech). IHC of serial sections of the two resulting TMA blocks \n(<10 μm thick) were stained with primary antibodies conjugated to HRP and 3,3′-Diaminobenzidine (DAB) \ndetection for CXCL8 (clone), and counter stained with hematoxylin by the TPSR. Digital images were \nobtained with an Ariol SL-50 automated scanning microscope and the Leica SCN400 Slide Scanner from \nVUMC Digital Histology Shared Resource. \nData processing \nAll flow cytometry data collected was uploaded to Cytobank for further analysis.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nTMA IHC data was de-arrayed using QuPath (version 0.5.0) and per pixel DAB signal was quantified in \npython using the separate_stains function in scikit-image (version 0.22.0).  Individual core images were \nprocessed for display using ImageJ. \nFor cytokine arrays, ImageJ Dot blot analyzer macro 90 was used to quantify intensity density of dots \npresent on each dot blot. \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nData Availability  \nDatasets analyzed in this manuscript are online at FlowRepository91 made available for reviewers \n(see link in submission materials), and will be made public upon acceptance.  Transparent analysis scripts \nfor datasets in this manuscript are available on the CytoLab Github page (https://github.com/cytolab/) with \nopen-source code and commented Rmarkdown analysis walkthroughs.   \nAcknowledgements \nWe thank Vanderbilt’s Cancer and Immunology Core and Flow Cytometry Shared Resource \nfacilities as well as all the surgeons, patients, and families that supported this work.  Research was \nsupported by the following funding resources: NIH/NCI grants R01 NS096238 (RAI, JMI), R01 CA226833 \n(JMI, SM, CEC, MJH), R01 NS118580 (RAI, AAB), U01 AI125056 (JMI), U54 CA217450 (JMI, MJH), T32 \nGM139800 (SM), T32 AI138932 (SM), the Vanderbilt-Ingram Cancer Center (VICC, P30 CA68485), the \nMichael David Greene Brain Cancer Fund (RAI, JMI), the Southeastern Brain Tumor Foundation (RAI, \nJMI), a gift from Daniel F Hewins (RAI), and the Ben & Catherine Ivy Foundation (RAI, JMI). Translational \nPathology Shared Resource (TPSR) is supported by NCI/NIH Cancer Center Support Grant \nP30CA068485. We thank outstanding undergraduate students Alejandra Rosario-Crespo, Amanda \nKouaho, and Niraj Rama for their work on projects exploring cancer and immune cell identity.   \nAuthor Contributions \nSM, RAI, and JMI designed and conceptualized the study. RAI, and JMI, provided intellectual \nsupport in assembling datasets. SM, MJH, and AAB collected data.  SM, CEC, and AAB developed data \nanalysis scripts. SM and JMI performed flow cytometry data analysis and interpretation.  AAM and SC \nscored MRI images for tumor contact with the lateral ventricle and provided patients’ clinical \ncharacteristics.  BCM confirmed tissue pathology. LBC, RCT, and KDW provided freshly resected tissue \nspecimens.  SM and JMI wrote the manuscript. RAI and JMI provided financial support.  All authors \ncontributed in reviewing and editing the manuscript. \nDeclaration of interests \nAll authors declare no competing interests. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nReferences \n1. Davis, M. (2016). Glioblastoma: Overview of Disease and Treatment. Clinical Journal of Oncology \nNursing 20, S2-S8. 10.1188/16.cjon.s1.2-8. \n2. Ostrom, Q.T., Cioffi, G., Gittleman, H., Patil, N., Waite, K., Kruchko, C., and Barnholtz-Sloan, J.S. \n(2019). CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors \nDiagnosed in the United States in 2012–2016. Neuro-Oncology 21, v1-v100. \n10.1093/neuonc/noz150. \n3. Sheikh, S., Radivoyevitch, T., Barnholtz-Sloan, J.S., and Vogelbaum, M. (2019). Long-term trends in \nglioblastoma survival: implications for historical control groups in clinical trials. Neuro-Oncology \nPractice. 10.1093/nop/npz046. \n4. Zhang, Y., and Zhang, Z. (2020). The history and advances in cancer immunotherapy: understanding \nthe characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cellular & \nMolecular Immunology 17, 807-821. 10.1038/s41423-020-0488-6. \n5. Fecci, P.E., and Sampson, J.H. (2019). The current state of immunotherapy for gliomas: an eye \ntoward the future. Journal of Neurosurgery 131, 657-666. 10.3171/2019.5.jns181762. \n6. Fu, W., Wang, W., Li, H., Jiao, Y., Huo, R., Yan, Z., Wang, J., Wang, S., Chen, D., Cao, Y., and Zhao, J. \nSingle-Cell Atlas Reveals Complexity of the Immunosuppressive Microenvironment of Initial and \nRecurrent Glioblastoma. \n7. Gutmann, D.H., and Kettenmann, H. (2019). Microglia/Brain Macrophages as Central Drivers of Brain \nTumor Pathobiology. Neuron 104, 442-449. 10.1016/j.neuron.2019.08.028. \n8. Yao, Y., Ye, H., Qi, Z., Mo, L., Yue, Q., Baral, A., Hoon, D.S.B., Vera, J.C., Heiss, J.D., Chen, C.C., et al. \n(2016). B7-H4(B7x)-Mediated Cross-talk between Glioma-Initiating Cells and Macrophages via the \nIL6/JAK/STAT3 Pathway Lead to Poor Prognosis in Glioma Patients. Clin Cancer Res 22, 2778-2790. \n10.1158/1078-0432.ccr-15-0858. \n9. Khan, F., Pang, L., Dunterman, M., Lesniak, M.S., Heimberger, A.B., and Chen, P. (2023). \nMacrophages and microglia in glioblastoma: heterogeneity, plasticity, and therapy. Journal of \nClinical Investigation 133. 10.1172/jci163446. \n10. Hambardzumyan, D., Gutmann, D.H., and Kettenmann, H. (2016). The role of microglia and \nmacrophages in glioma maintenance and progression. Nat Neurosci 19, 20-27. 10.1038/nn.4185. \n11. Chen, Z., Feng, X., Herting, C.J., Garcia, V.A., Nie, K., Pong, W.W., Rasmussen, R., Dwivedi, B., Seby, \nS., Wolf, S.A., et al. Cellular and Molecular Identity of Tumor-Associated Macrophages in \nGlioblastoma. \n12. Mignogna, C., Signorelli, F., Vismara, M.F., Zeppa, P., Camastra, C., Barni, T., Donato, G., and Di Vito, \nA. (2016). A reappraisal of macrophage polarization in glioblastoma: Histopathological and \nimmunohistochemical findings and review of the literature. Pathol Res Pract 212, 491-499. \n10.1016/j.prp.2016.02.020. \n13. Walentynowicz, K.A., Engelhardt, D., Cristea, S., Yadav, S., Onubogu, U., Salatino, R., Maerken, M., \nVincentelli, C., Jhaveri, A., Geisberg, J., et al. Single-cell heterogeneity of EGFR and CDK4 co-\namplification is linked to immune infiltration in glioblastoma. \n14. Liu, S., Zhang, C., Maimela, N.R., Yang, L., Zhang, Z., Ping, Y., Huang, L., and Zhang, Y. Molecular and \nclinical characterization of CD163 expression via large-scale analysis in glioma. \n15. Bartkowiak, T., Lima, S.M., Hayes, M.J., Mistry, A.M., Brockman, A.A., Sinnaeve, J., Leelatian, N., Roe, \nC.E., Mobley, B.C., Chotai, S., et al. (2023). An immunosuppressed microenvironment distinguishes \nlateral ventricle–contacting glioblastomas. JCI Insight 8. 10.1172/jci.insight.160652. \n16. Mistry, A.M., Hale, A.T., Chambless, L.B., Weaver, K.D., Thompson, R.C., and Ihrie, R.A. (2017). \nInfluence of glioblastoma contact with the lateral ventricle on survival: a meta-analysis. J \nNeurooncol 131, 125-133. 10.1007/s11060-016-2278-7. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\n17. Mistry, A.M., Dewan, M.C., White-Dzuro, G.A., Brinson, P.R., Weaver, K.D., Thompson, R.C., Ihrie, \nR.A., and Chambless, L.B. Decreased survival in glioblastomas is specific to contact with the \nventricular-subventricular zone, not subgranular zone or corpus callosum. \n18. Poon, C.C., Sarkar, S., Yong, V.W., and Kelly, J.J.P. (2017). Glioblastoma-associated microglia and \nmacrophages: targets for therapies to improve prognosis. Brain 140, 1548-1560. \n10.1093/brain/aww355. \n19. Ye, Z., Ai, X., Yang, K., Yang, Z., Fei, F., Liao, X., Qiu, Z., Gimple, R.C., Yuan, H., Huang, H., et al. (2023). \nTargeting Microglial Metabolic Rewiring Synergizes with Immune-Checkpoint Blockade Therapy for \nGlioblastoma. Cancer Discov 13, 974-1001. 10.1158/2159-8290.Cd-22-0455. \n20. Fermi, V., Warta, R., Wöllner, A., Lotsch, C., Jassowicz, L., Rapp, C., Knoll, M., Jungwirth, G., Jungk, C., \nDao Trong, P., et al. (2023). Effective Reprogramming of Patient-Derived M2-Polarized Glioblastoma-\nAssociated Microglia/Macrophages by Treatment with GW2580. Clin Cancer Res 29, 4685-4697. \n10.1158/1078-0432.Ccr-23-0576. \n21. Quail, D.F., and Joyce, J.A. Microenvironmental regulation of tumor progression and metastasis. \n22. Zhang, Y., Cheng, S., Zhang, M., Zhen, L., Pang, D., Zhang, Q., and Li, Z. (2013). High-infiltration of \ntumor-associated macrophages predicts unfavorable clinical outcome for node-negative breast \ncancer. PLoS One 8, e76147. 10.1371/journal.pone.0076147. \n23. Kumar, A.T., Knops, A., Swendseid, B., Martinez-Outschoom, U., Harshyne, L., Philp, N., Rodeck, U., \nLuginbuhl, A., Cognetti, D., Johnson, J., and Curry, J. (2019). Prognostic Significance of Tumor-\nAssociated Macrophage Content in Head and Neck Squamous Cell Carcinoma: A Meta-Analysis. \nFront Oncol 9, 656. 10.3389/fonc.2019.00656. \n24. Koll, F.J., Banek, S., Kluth, L., Köllermann, J., Bankov, K., Chun, F.K., Wild, P.J., Weigert, A., and Reis, \nH. (2023). Tumor-associated macrophages and Tregs influence and represent immune cell \ninfiltration of muscle-invasive bladder cancer and predict prognosis. J Transl Med 21, 124. \n10.1186/s12967-023-03949-3. \n25. Wang, H., Yang, L., Wang, D., Zhang, Q., and Zhang, L. (2017). Pro-tumor activities of macrophages in \nthe progression of melanoma. Hum Vaccin Immunother 13, 1556-1562. \n10.1080/21645515.2017.1312043. \n26. Erlandsson, A., Carlsson, J., Lundholm, M., Fält, A., Andersson, S.-O., Andrén, O., and Davidsson, S. \n(2019). M2 macrophages and regulatory T cells in lethal prostate cancer. The Prostate 79, 363-369. \nhttps://doi.org/10.1002/pros.23742. \n27. Szulzewsky, F., Pelz, A., Feng, X., Synowitz, M., Markovic, D., Langmann, T., Holtman, I.R., Wang, X., \nEggen, B.J., Boddeke, H.W., et al. (2015). Glioma-associated microglia/macrophages display an \nexpression profile different from M1 and M2 polarization and highly express Gpnmb and Spp1. PLoS \nOne 10, e0116644. 10.1371/journal.pone.0116644. \n28. Chen, D., Varanasi, S.K., Hara, T., Traina, K., Sun, M., McDonald, B., Farsakoglu, Y., Clanton, J., Xu, S., \nGarcia-Rivera, L., et al. CTLA-4 blockade induces a microglia-Th1 cell partnership that stimulates \nmicroglia phagocytosis and anti-tumor function in glioblastoma. \n29. Gangoso, E., Southgate, B., Bradley, L., Rus, S., Galvez-Cancino, F., McGivern, N., Güç, E., Kapourani, \nC.-A., Byron, A., Ferguson, K.M., et al. (2021). Glioblastomas acquire myeloid-affiliated \ntranscriptional programs via epigenetic immunoediting to elicit immune evasion. Cell 184, 2454-\n2470.e2426. 10.1016/j.cell.2021.03.023. \n30. Yang, F.A.-O., Akhtar, M.A.-O., Zhang, D.A.-O., El-Mayta, R.A.-O.X., Shin, J.A.-O., Dorsey, J.F., Zhang, \nL., Xu, X.A.-O., Guo, W., Bagley, S.A.-O., et al. An immunosuppressive vascular niche drives \nmacrophage polarization and immunotherapy resistance in glioblastoma. \n31. Chryplewicz, A., Scotton, J., Tichet, M., Zomer, A., Shchors, K., Joyce, J.A., Homicsko, K., and \nHanahan, D. Cancer cell autophagy, reprogrammed macrophages, and remodeled vasculature in \nglioblastoma triggers tumor immunity. \n32. Murray, J., Peter, Allen, E., Judith, Biswas, K., Subhra, Fisher, A., Edward, Gilroy, W., Derek, Goerdt, \nS., Gordon, S., Hamilton, A., John, Ivashkiv, B., Lionel, Lawrence, T., et al. (2014). Macrophage \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nActivation and Polarization: Nomenclature and Experimental Guidelines. Immunity 41, 14-20. \n10.1016/j.immuni.2014.06.008. \n33. Roussel, M., Ferrell, P.B., Greenplate, A.R., Lhomme, F., Le Gallou, S., Diggins, K.E., Johnson, D.B., \nand Irish, J.M. (2017). Mass cytometry deep phenotyping of human mononuclear phagocytes and \nmyeloid-derived suppressor cells from human blood and bone marrow. Journal of Leukocyte Biology \n102, 437-447. 10.1189/jlb.5ma1116-457r. \n34. Xue, J., Schmidt, V., Susanne, Sander, J., Draffehn, A., Krebs, W., Quester, I., Nardo, D., Dominic, \nGohel, D., Trupti, Emde, M., Schmidleithner, L., et al. (2014). Transcriptome-Based Network Analysis \nReveals a Spectrum Model of Human Macrophage Activation. Immunity 40, 274-288. \n10.1016/j.immuni.2014.01.006. \n35. Chen, R.A.-O., Yang, D., Shen, L., Fang, J., Khan, R., and Liu, D.A.-O. Overexpression of CD86 \nenhances the ability of THP-1 macrophages to defend against Talaromyces marneffei. \n36. Schulz, D., Severin, Y., Zanotelli, V.R.T., and Bodenmiller, B. In-Depth Characterization of Monocyte-\nDerived Macrophages using a Mass Cytometry-Based Phagocytosis Assay. \n37. Roussel, M., Bartkowiak, T., and Irish, J.M. (2019). Picturing Polarized Myeloid Phagocytes and \nRegulatory Cells by Mass Cytometry. Methods in molecular biology 1989, 217-226. 10.1007/978-1-\n4939-9454-0_14. \n38. Peter, Judith, Subhra, Edward, Derek, Goerdt, S., Gordon, S., John, Lionel, Lawrence, T., et al. (2014). \nMacrophage Activation and Polarization: Nomenclature and Experimental Guidelines. Immunity 41, \n14-20. 10.1016/j.immuni.2014.06.008. \n39. Bronte, V., Brandau, S., Chen, S.-H., Colombo, M.P., Frey, A.B., Greten, T.F., Mandruzzato, S., \nMurray, P.J., Ochoa, A., Ostrand-Rosenberg, S., et al. (2016). Recommendations for myeloid-derived \nsuppressor cell nomenclature and characterization standards. Nature Communications 7, 12150. \n10.1038/ncomms12150. \n40. Sharma, I., Singh, A., Sharma, K., and Saxena, S. (2017). Gene Expression Profiling of Chemokines and \nTheir Receptors in Low and High Grade Astrocytoma. Asian Pac J Cancer Prev 18, 1307-1313. \n10.22034/apjcp.2017.18.5.1307. \n41. Raghuwanshi, S.K., Su, Y., Singh, V., Haynes, K., Richmond, A., and Richardson, R.M. (2012). The \nChemokine Receptors CXCR1 and CXCR2 Couple to Distinct G Protein-Coupled Receptor Kinases To \nMediate and Regulate Leukocyte Functions. The Journal of Immunology 189, 2824-2832. \n10.4049/jimmunol.1201114. \n42. Holmes, W.E., Lee, J., Kuang, W.J., Rice, G.C., and Wood, W.I. (1991). Structure and functional \nexpression of a human interleukin-8 receptor. Science 253, 1278-1280. 10.1126/science.1840701. \n43. Waugh, D.J.J., and Wilson, C. (2008). The Interleukin-8 Pathway in Cancer. Clinical Cancer Research \n14, 6735-6741. 10.1158/1078-0432.Ccr-07-4843. \n44. Heidemann, J., Ogawa, H., Dwinell, M.B., Rafiee, P., Maaser, C., Gockel, H.R., Otterson, M.F., Ota, \nD.M., Lugering, N., Domschke, W., and Binion, D.G. (2003). Angiogenic effects of interleukin 8 \n(CXCL8) in human intestinal microvascular endothelial cells are mediated by CXCR2. J Biol Chem 278, \n8508-8515. 10.1074/jbc.M208231200. \n45. Prince, E.W., Apps, J.R., Jeang, J., Chee, K., Medlin, S., Jackson, E.M., Dudley, R., Limbrick, D., Naftel, \nR., Johnston, J., et al. (2024). Unraveling the Complexity of the Senescence-Associated Secretory \nPhenotype in Adamantinomatous Craniopharyngioma Using Multi-Modal Machine Learning \nAnalysis. Neuro-Oncology, noae015. 10.1093/neuonc/noae015. \n46. Chen, Z., Will, R., Kim, S.N., Busch, M.A., Dünker, N., Dammann, P., Sure, U., and Zhu, Y. (2023). \nNovel Function of Cancer Stem Cell Marker ALDH1A3 in Glioblastoma: Pro-Angiogenesis through \nParacrine PAI-1 and IL-8. Cancers 15, 4422. 10.3390/cancers15174422. \n47. Brat, D.J., Bellail, A.C., and Van Meir, E.G. (2005). The role of interleukin-8 and its receptors in \ngliomagenesis and tumoral angiogenesis. Neuro Oncol 7, 122-133. 10.1215/s1152851704001061. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\n48. Infanger, D.W., Cho, Y., Lopez, B.S., Mohanan, S., Liu, S.C., Gursel, D., Boockvar, J.A., and Fischbach, \nC. (2013). Glioblastoma stem cells are regulated by interleukin-8 signaling in a tumoral perivascular \nniche. Cancer Res 73, 7079-7089. 10.1158/0008-5472.Can-13-1355. \n49. Sharma, I., Singh, A., Siraj, F., and Saxena, S. (2018). IL-8/CXCR1/2 signalling promotes tumor cell \nproliferation, invasion and vascular mimicry in glioblastoma. J Biomed Sci 25, 62. 10.1186/s12929-\n018-0464-y. \n50. Hasan, T., Caragher, S.P., Shireman, J.M., Park, C.H., Atashi, F., Baisiwala, S., Lee, G., Guo, D., Wang, \nJ.Y., Dey, M., et al. (2019). Interleukin-8/CXCR2 signaling regulates therapy-induced plasticity and \nenhances tumorigenicity in glioblastoma. Cell Death & Disease 10, 292. 10.1038/s41419-019-1387-6. \n51. Hasan, T., Caragher, S.P., Shireman, J.M., Park, C.H., Atashi, F., Baisiwala, S., Lee, G., Guo, D., Wang, \nJ.Y., Dey, M., et al. (2019). Interleukin-8/CXCR2 signaling regulates therapy-induced plasticity and \nenhances tumorigenicity in glioblastoma. Cell Death & Disease 10. 10.1038/s41419-019-1387-6. \n52. Holst, C.B., Christensen, I.J., Vitting-Seerup, K., Skjøth-Rasmussen, J., Hamerlik, P., Poulsen, H.S., and \nJohansen, J.S. (2021). Plasma IL-8 and ICOSLG as prognostic biomarkers in glioblastoma. Neurooncol \nAdv 3, vdab072. 10.1093/noajnl/vdab072. \n53. Benner, B., Scarberry, L., Suarez-Kelly, L.P., Duggan, M.C., Campbell, A.R., Smith, E., Lapurga, G., \nJiang, K., Butchar, J.P., Tridandapani, S., et al. (2019). Generation of monocyte-derived tumor-\nassociated macrophages using tumor-conditioned media provides a novel method to study tumor-\nassociated macrophages in vitro. Journal for ImmunoTherapy of Cancer 7. 10.1186/s40425-019-\n0622-0. \n54. Diggins, K.E., Greenplate, A.R., Leelatian, N., Wogsland, C.E., and Irish, J.M. (2017). Characterizing \ncell subsets using marker enrichment modeling. Nature Methods 14, 275-278. 10.1038/nmeth.4149. \n55. Greenplate, A.R., McClanahan, D.D., Oberholtzer, B.K., Doxie, D.B., Roe, C.E., Diggins, K.E., Leelatian, \nN., Rasmussen, M.L., Kelley, M.C., Gama, V., et al. (2019). Computational Immune Monitoring \nReveals Abnormal Double-Negative T Cells Present across Human Tumor Types. Cancer Immunol Res \n7, 86-99. 10.1158/2326-6066.CIR-17-0692. \n56. Barone, S.M., Paul, A.G., Muehling, L.M., Lannigan, J.A., Kwok, W.W., Turner, R.B., Woodfolk, J.A., \nand Irish, J.M. (2021). Unsupervised machine learning reveals key immune cell subsets in COVID-19, \nrhinovirus infection, and cancer therapy. eLife 10. 10.7554/elife.64653. \n57. Hanahan, D., and Robert (2011). Hallmarks of Cancer: The Next Generation. Cell 144, 646-674. \n10.1016/j.cell.2011.02.013. \n58. Chen, Daniel S., and Mellman, I. (2013). Oncology Meets Immunology: The Cancer-Immunity Cycle. \nImmunity 39, 1-10. https://doi.org/10.1016/j.immuni.2013.07.012. \n59. Nørøxe, D.S., Poulsen, H.S., and Lassen, U. (2016). Hallmarks of glioblastoma: a systematic review. \nESMO open. 1, e000144. 10.1136/esmoopen-2016-000144. \n60. Martinez-Lage, M., Lynch, T.M., Bi, Y., Cocito, C., Way, G.P., Pal, S., Haller, J., Yan, R.E., Ziober, A., \nNguyen, A., et al. (2019). Immune landscapes associated with different glioblastoma molecular \nsubtypes. Acta Neuropathologica Communications 7. 10.1186/s40478-019-0803-6. \n61. Razavi, S.M., Lee, K.E., Jin, B.E., Aujla, P.S., Gholamin, S., and Li, G. (2016). Immune Evasion \nStrategies of Glioblastoma. Front Surg 3, 11. 10.3389/fsurg.2016.00011. \n62. Wang, H.W., and Joyce, J.A. (2010). Alternative activation of tumor-associated macrophages by IL-4: \npriming for protumoral functions. Cell Cycle 9, 4824-4835. 10.4161/cc.9.24.14322. \n63. Shapouri-Moghaddam, A., Mohammadian, S., Vazini, H., Taghadosi, M., Esmaeili, S.A., Mardani, F., \nSeifi, B., Mohammadi, A., Afshari, J.T., and Sahebkar, A. (2018). Macrophage plasticity, polarization, \nand function in health and disease. J Cell Physiol 233, 6425-6440. 10.1002/jcp.26429. \n64. Tobin, R.P., Jordan, K.R., Kapoor, P., Spongberg, E., Davis, D., Vorwald, V.M., Couts, K.L., Gao, D., \nSmith, D.E., Borgers, J.S.W., et al. (2019). IL-6 and IL-8 Are Linked With Myeloid-Derived Suppressor \nCell Accumulation and Correlate With Poor Clinical Outcomes in Melanoma Patients. Frontiers in \nOncology 9. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\n65. Jarmuzek, P., Defort, P., Kot, M., Wawrzyniak-Gramacka, E., Morawin, B., and Zembron-Lacny, A. \n(2023). Cytokine Profile in Development of Glioblastoma in Relation to Healthy Individuals. Int J Mol \nSci 24. 10.3390/ijms242216206. \n66. Baggiolini, M., Walz, A., and Kunkel, S.L. (1989). Neutrophil-activating peptide-1/interleukin 8, a \nnovel cytokine that activates neutrophils. J Clin Invest 84, 1045-1049. 10.1172/jci114265. \n67. Li, A., Dubey, S., Varney, M.L., Dave, B.J., and Singh, R.K. (2003). IL-8 directly enhanced endothelial \ncell survival, proliferation, and matrix metalloproteinases production and regulated angiogenesis. J \nImmunol 170, 3369-3376. 10.4049/jimmunol.170.6.3369. \n68. David, J.M., Dominguez, C., Hamilton, D.H., and Palena, C. (2016). The IL-8/IL-8R Axis: A Double \nAgent in Tumor Immune Resistance. Vaccines (Basel) 4. 10.3390/vaccines4030022. \n69. Brinkmann, V., Reichard, U., Goosmann, C., Fauler, B., Uhlemann, Y., Weiss, D.S., Weinrauch, Y., and \nZychlinsky, A. (2004). Neutrophil extracellular traps kill bacteria. Science 303, 1532-1535. \n10.1126/science.1092385. \n70. Lattanzio, L., Tonissi, F., Torta, I., Gianello, L., Russi, E., Milano, G., Merlano, M., and Lo Nigro, C. \n(2013). Role of IL-8 induced angiogenesis in uveal melanoma. Invest New Drugs 31, 1107-1114. \n10.1007/s10637-013-0005-1. \n71. Xie, K. (2001). Interleukin-8 and human cancer biology. Cytokine & growth factor reviews. 12, 375-\n391. 10.1016/S1359-6101(01)00016-8. \n72. Filimon, A., Preda, I.A., Boloca, A.F., and Negroiu, G. (2021). Interleukin-8 in Melanoma \nPathogenesis, Prognosis and Therapy-An Integrated View into Other Neoplasms and Chemokine \nNetworks. Cells 11. 10.3390/cells11010120. \n73. Zhang, W., Yang, F., Zheng, Z., Li, C., Mao, S., Wu, Y., Wang, R., Zhang, J., Zhang, Y., Wang, H., et al. \n(2022). Sulfatase 2 Affects Polarization of M2 Macrophages through the IL-8/JAK2/STAT3 Pathway in \nBladder Cancer. Cancers 15, 131. 10.3390/cancers15010131. \n74. Aalinkeel, R., Nair, B., Chen, C.K., Mahajan, S.D., Reynolds, J.L., Zhang, H., Sun, H., Sykes, D.E., \nChadha, K.C., Turowski, S.G., et al. (2016). Nanotherapy silencing the interleukin-8 gene produces \nregression of prostate cancer by inhibition of angiogenesis. Immunology 148, 387-406. \n10.1111/imm.12618. \n75. Zhang, B., Shi, L., Lu, S., Sun, X., Liu, Y., Li, H., Wang, X., Zhao, C., Zhang, H., and Wang, Y. (2015). \nAutocrine IL-8 promotes F-actin polymerization and mediate mesenchymal transition via ELMO1-NF-\nκB-Snail signaling in glioma. Cancer Biol Ther 16, 898-911. 10.1080/15384047.2015.1028702. \n76. Dumitru, C.A., Schröder, H., Schäfer, F.T.A., Aust, J.F., Kreße, N., Siebert, C.L.R., Stein, K.-P., Haghikia, \nA., Wilkens, L., Mawrin, C., and Sandalcioglu, I.E. (2023). Progesterone Receptor Membrane \nComponent 1 (PGRMC1) Modulates Tumour Progression, the Immune Microenvironment and the \nResponse to Therapy in Glioblastoma. Cells 12, 2498. 10.3390/cells12202498. \n77. Liu, H., Zhao, Q., Tan, L., Wu, X., Huang, R., Zuo, Y., Chen, L., Yang, J., Zhang, Z.-X., Ruan, W., et al. \n(2023). Neutralizing IL-8 potentiates immune checkpoint blockade efficacy for glioma. Cancer cell. \n41, 693-710.e698. 10.1016/j.ccell.2023.03.004. \n78. Ross, D.T., and Perou, C.M. (2001). A comparison of gene expression signatures from breast tumors \nand breast tissue derived cell lines. Dis Markers 17, 99-109. 10.1155/2001/850531. \n79. Vartanian, A., Singh, S.K., Agnihotri, S., Jalali, S., Burrell, K., Aldape, K.D., and Zadeh, G. (2014). \nGBM's multifaceted landscape: highlighting regional and microenvironmental heterogeneity. Neuro-\nOncology 16, 1167-1175. 10.1093/neuonc/nou035. \n80. Leblanc, V.G., Trinh, D.L., Aslanpour, S., Hughes, M., Livingstone, D., Jin, D., Ahn, B.Y., Blough, M.D., \nCairncross, J.G., Chan, J.A., et al. (2022). Single-cell landscapes of primary glioblastomas and \nmatched explants and cell lines show variable retention of inter- and intratumor heterogeneity. \nCancer Cell 40, 379-392.e379. 10.1016/j.ccell.2022.02.016. \n81. Lum, D.H., Matsen, C., Welm, A.L., and Welm, B.E. (2012). Overview of human primary tumorgraft \nmodels: comparisons with traditional oncology preclinical models and the clinical relevance and \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nutility of primary tumorgrafts in basic and translational oncology research. Curr Protoc Pharmacol \nChapter 14, Unit 14.22. 10.1002/0471141755.ph1422s59. \n82. Neftel, C., Laffy, J., Filbin, M.G., Hara, T., Shore, M.E., Rahme, G.J., Richman, A.R., Silverbush, D., \nShaw, M.L., Hebert, C.M., et al. (2019). An Integrative Model of Cellular States, Plasticity, and \nGenetics for Glioblastoma. Cell 178, 835-849.e821. 10.1016/j.cell.2019.06.024. \n83. Jacob, F., Salinas, R.D., Zhang, D.Y., Nguyen, P.T.T., Schnoll, J.G., Wong, S.Z.H., Thokala, R., Sheikh, S., \nSaxena, D., Prokop, S., et al. (2020). A Patient-Derived Glioblastoma Organoid Model and Biobank \nRecapitulates Inter- and Intra-tumoral Heterogeneity. Cell 180, 188-204.e122. \n10.1016/j.cell.2019.11.036. \n84. Vaubel, R.A., Tian, S., Remonde, D., Schroeder, M.A., Mladek, A.C., Kitange, G.J., Caron, A., \nKollmeyer, T.M., Grove, R., Peng, S., et al. (2020). Genomic and Phenotypic Characterization of a \nBroad Panel of Patient-Derived Xenografts Reflects the Diversity of Glioblastoma. Clin Cancer Res 26, \n1094-1104. 10.1158/1078-0432.Ccr-19-0909. \n85. Leelatian, N., Doxie, D.B., Greenplate, A.R., Sinnaeve, J., Ihrie, R.A., and Irish, J.M. Preparing Viable \nSingle Cells from Human Tissue and Tumors for Cytomic Analysis. \n86. Mistry, A.M., Dewan, M.C., White-Dzuro, G.A., Brinson, P.R., Weaver, K.D., Thompson, R.C., Ihrie, \nR.A., and Chambless, L.B. (2017). Decreased survival in glioblastomas is specific to contact with the \nventricular-subventricular zone, not subgranular zone or corpus callosum. Journal of Neuro-\nOncology 132, 341-349. 10.1007/s11060-017-2374-3. \n87. Mistry, A.M., Wooten, D.J., Davis, L.T., Mobley, B.C., Quaranta, V., and Ihrie, R.A. (2019). Ventricular-\nSubventricular Zone Contact by Glioblastoma is Not Associated with Molecular Signatures in Bulk \nTumor Data. Scientific Reports 9, 1842. 10.1038/s41598-018-37734-w. \n88. Davies, J.Q., and Gordon, S. Isolation and culture of human macrophages. \n89. Leelatian, N., Sinnaeve, J., Mistry, A.M., Barone, S.M., Brockman, A.A., Diggins, K.E., Greenplate, \nA.R., Weaver, K.D., Thompson, R.C., Chambless, L.B., et al. (2020). Unsupervised machine learning \nreveals risk stratifying glioblastoma tumor cells. eLife 9. 10.7554/elife.56879. \n90. Klemm, A.H. (2020). Semi-automated analysis of dot blots using ImageJ/Fiji. F1000Research 9, 1385. \n10.12688/f1000research.27179.1. \n91. Spidlen, J., Breuer, K., Rosenberg, C., Kotecha, N., and Brinkman, R.R. (2012). FlowRepository: A \nresource of annotated flow cytometry datasets associated with peer-reviewed publications. \nCytometry Part A 81A, 727-731. 10.1002/cyto.a.22106. \n \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nMedina et al. – Figure 1 \n \nFigure 1 – M_GBM_TCM are similar to an M2 phenotype and distinct from other M2 macrophages, \nincluding M_IL-6. A) 2D contour plots show expression of signature surface proteins CD163, CD206, \nCD86, and PD-L1 for each condition. Cells were either unstimulated monocytes, macrophages stimulated \nwith IFNγ (a control for M1-like macrophages), macrophages stimulated with IL-6 (a control for M2-like \nmacrophages), or macrophages stimulated with primary tumor conditioned media collected after 3 days \nof ex vivo culture (sample ID: TCM1, TCM2).  B) Bar graphs display the percentage of cells included in \nindicated gates across conditions.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nMedina et al. – Figure 2 \n \n \nFigure 2 – IL-8 is secreted by glioblastoma tumor cells ex vivo.  A) Cytokine arrays are shown from \nmedia negative control (RPMI) and media conditioned by a representative glioblastoma tumor (RPMI + \nLC-26).  Darkness of the spot indicates increasing presence of one of the 105 tested cytokines.  Positive \nand negative control spots are located on each corner of the blot.  B) Box plot shows quantification of the \nintensity density of proteins measured on the cytokine arrays of GBM tumor conditioned media (N = 7 \ntumors).  A significance threshold was calculated based on three standard deviations above the level of \nbackground observed in negative control wells (red line).  Cytokines present in at least 1 of 7 tumors are \nlabeled across the X axis. Significant cytokines where the median expression across tumors exceeded \nthe calculated threshold are colored in dark blue. Non-significant cytokines are colored light blue. Stacked \ndots above the plot indicate the number of tumor samples that exceeded the threshold for any cytokine \nmeasured. The most significant cytokine identified is labeled by a red box. The full dataset is available \nonline (Supplementary Information).   \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nMedina et al. – Figure 3 \n \nFigure 3 – GBM secreted IL-8 mediates  ex vivo  polarization of macrophages to express a \nsuppressive signature.  A) t-SNE plots display cell density of polarized macrophages across different \nconditions including recombinant IL-8 (M_IL-8) and GBM tumor conditioned media (M_GBM_TCM), in the \npresence or absence of α-IL-8 blocking antibody.  B) T-REX analysis comparing polarization conditions in \nthe presence or absence of α-IL-8 where cells that are >80% enriched in macrophages polarized in the \nabsence of α-IL-8 are colored in dark red, >60% enriched in light red,  cells enriched >80% in macrophages \npolarized in the presence of α-IL-8 are colored in dark blue and >60% enriched in light blue. Cells colored \nin gray are similarly enriched in both conditions C) 2D contour plots show expression of signature surface \nproteins CD163, CD32, CD86, and PD-L1 for each condition. Gates colored in blue indicate the conditions \nwhere macrophages were polarized in the presence of α-IL-8. D) Histogram plots display individual protein \nexpression of markers CD32, CD163, CD206, HLA-DR, CD86 and PD-L1. Overlaid histograms represent \nmacrophages polarized in the presence (blue) or absence (gray) of α-IL-8 blocking antibody. A blue dotted \nline indicates a threshold for positive expression.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint \n\nMedina et al. – Figure 4 \n \nFigure 4 – IL-8 is expressed in primary contacting tumors and less abundantly expressed in non-\ncontacting tumors. A) Dot plot comparing the percentage of positive pixels per core stained for IL-8 in \ncontacting (N=114) and non-contacting (N=78) GBM tumor cores. 3 cores of each patient sample were \nincluded for a total of 192 cores. A Mann Whitney statistical test was used to analyze the difference in \nimmunohistochemical staining of IL-8 between the two groups. *** indicates p < 0.001. B) Dot plot graph \ncomparing the average percentage of positive pixels per patient (N=73) stained for IL-8 in contacting \n(N=44) and non-contacting (N=29) GBM tumors. A Mann Whitney statistical test was used to analyze the \nstatistical difference in immunohistochemical staining of IL-8 between the two groups. * Indicates p < 0.05. \nC) Dot plot comparing number of cells per mm2 (on an arcsinh scale) to percentage of positive IL-8 pixels \nfor each patient. Contacting tumors are displayed in red and non-contacting tumors are displayed in blue.  \nD) Representative immunohistochemistry images of CXCL8 (IL-8) expression on glioblastoma tumor \nmicroarray (TMA) (12 Contacting tumor examples shown with red labels and 12 Non-Contacting tumor \nexamples shown with blue labels). Top right corner of each image contains 3x zoomed in view (scale \nbar=100µm). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 30, 2024. ; https://doi.org/10.1101/2024.03.29.587030doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}