BBB dysfunction drives neurotoxic myeloid cells in glioblastoma and Alzheimer’s disease

doi:10.21203/rs.3.rs-3610020/v1

Download PDF

Research Article

BBB dysfunction drives neurotoxic myeloid cells in glioblastoma and Alzheimer’s disease

https://doi.org/10.21203/rs.3.rs-3610020/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

The blood-brain barrier (BBB) defines the physiological function of the brain and is disrupted in multiple neuropathological diseases. Although vascular abnormalities and BBB leakage characterize glioma progression and occurs in the early stages of Alzheimer's disease (AD), their correlation with the neurodegeneration observed in the brain parenchyma is unclear.

Methods

Using RNAseq, bioinformatics tools and a cohort of patients with Glioma and Alzheimer's disease for validation of results, we have established an analysis of BBB dysfunction and neuron loss. A mouse model for GBM pathology was also used that reversed BBB disruption and neuron loss, with the incorporation of the IDH mutation. Finally, to establish a characterization of the relevant immune populations with an IHC analysis and transcriptional profile.

Results

Molecular analysis of the brain’s ecosystem showed that gene expression signatures linked to BBB dysfunction and neuronal synapses correlate directly and inversely, respectively, with the progression of AD and gliomas. Using human samples, patient´s-derived xenografts and mice with orthotopically implanted glioma cells, we observed that neuronal loss is linked to BBB breakdown, particularly in mesenchymal tumors. This involves the extravasation of monocyte-derived macrophages, specifically those with a suppressive phenotype, along with an increase in inflammatory cytokines. Moreover, our data demonstrate that vascular normalization can revert the neuronal loss and aggressiveness in these gliomas.

Conclusions

We propose that BBB dysfunction is a common brain process that drives chronic inflammation, governed by suppressive myeloid cells. These results could help us to apply common therapies to the progression of central nervous system pathologies associated with BBB dysfunction, since it functions as an enhancer through toxic neuroinflammation and myeloid dysfunction.

Diseases of the Central nervous system (CNS) present special characteristics compared to other pathologies, due to specific attributes of this tissue, such as an abundant microvasculature, a rich extracellular matrix and a differential immune composition [1]. Furthermore, blood-brain barrier (BBB) disruption has recently been shown to occur in CNS diseases of different etiologies, such as stroke, multiple sclerosis, traumatic brain injury and epilepsy, associated with characteristic gene expression signatures that define them [2]. In addition, BBB disruption occurs during the brain aging process and precedes the onset of neurodegeneration in Alzheimer´s disease (AD) [3]. The disruption of the brain barrier generates a strong uncoupling of metabolites and ions homeostasis, associated with the entry of a large content of plasma molecules, as well as immune cells [4]. Although these processes are well established, the toxic and pro-pathogenic consequences have not yet been fully elucidated. Brain tumors such as gliomas or metastases induce a major rearrangement of the brain microenvironment, generating a profound change in vascular and immune phenotypes [5]. Adult-type diffuse gliomas, in particular, are defined as three distinct entities: IDH-mutant and 1p/19q codeleted oligodendroglioma, IDH-mutant astrocytoma (grade 2 (G2), G3 and G4) and IDH-wt Glioblastoma- (GBM) (G4)[6]. At the molecular level, these tumor types have been widely studied and the genomic modifications and transcriptional profiles that govern them are known in detail [7]., it remains an incurable tumor, linked to severe neurological deterioration without an effective clinical approach. More recently, several of our studies and those of others have re-examined the glioma microenvironment to understand the complex biology of this type of tumor, helping to understand the presence of fewer vascular and immune alterations in IDH-mut gliomas as compared to wild-type ones [1, 8–12]. However, the complex processes leading to BBB dysfunction and how it is associated with brain disease are largely unknown [1]. To begin to address if there are common mechanisms driving neurodegeneration in different pathologies, we have performed a cross-cutting analysis comparing glioma and AD samples from patients. We have been able to establish a strong connection between BBB disruption generated in gliomas, particularly in those with a mesenchymal profile, with myeloid cell entry into the brain and severe neuronal loss and neurological deterioration. Moreover, we have confirmed a similar pattern during Alzheimer´s progression, suggesting that the brain microenvironment determines its pathological evolution, linked to the dysfunction of the myeloid cells entering the brain. The most surprising finding is that this process occurs to a similar degree in AD and gliomas associated with BBB dysfunction. Thus, we propose that BBB dysfunction is a promoter of alterations in the myeloid cell ecosystem, which could generate a favorable brain microenvironment for the development of different pathologies.

Patient’s samples

Glioma tissues, fresh frozen or embedded in paraffin, belong to two different cohorts, the first one contains 26 patients and the second one 62 patients (supplementary TABLE 1). Tissues were obtained from patients treated at the “Hospital 12 de Octubre” (Madrid, Spain). In addition, an AD cohort has been used with 43 patient samples obtained from “Fundación CIEN” (Madrid, Spain). Samples were obtained after patient's written consent and with the approval of the Ethical Committee at “Hospital 12 de Octubre” and “Fundación CIEN” (CEI_ 21/551 and CEI PI-74_2019 respectively)

Public data cohort, curation and signatures discovery

The array of normal brain and AD data was obtained from the Hisayama study (Hokama et al 2014, GSE36980 and GSE36981)[13] and from Berchtold et al. 2008 (GSE48350) [14]. Glioma and glioblastoma cohort were obtained from TCGA consortium xena browser y gliovis (CNIO, ISCIII). The scRNAseq analyzes were obtained from the Single Cell Portal for the datasets (GSE131928) and (GSE182109).

Characterization of immune cells by flow cytometry

Tumors suspensions were obtained after mechanical and enzymatic disaggregation (Accumax (Merck Millipore) (15 min, room temperature (RT)). Erythrocytes were lysed with Quicklysis buffer (Cytognos) and cells were incubated with hFcR Blocking (Miltenyi), before the antibody (supplementary TABLA 1) incubation (20 min at 4°C in PBS 1% fetal bovine serum (FBS)). Viable cells were labelled with a Fixable Viability Stain (Becton Dickinson) (20 min, RT). The analysis was conducted in a Macsquant10 flow Cytometry (Miltenyi). Subset definition was: CD45⁺, lymphoid (CD45⁺CD11b⁻SSC^lo), myeloid (CD45⁺CD11b⁺SSC^lo or SSC^hi). The staining panels were published previously [15].

Western Blot

Proteins from human and mice samples were obtained by manual disaggregation with lysis buffer (50 mM Tris-pH 7.5, 300 mM NaCl, 0.5% SDS, and 1% Triton X-100) prior to heating 15 min with agitation at 95ºC. Protein quantification was made using a BCA Protein Assay Kit (Thermo-Fisher-Scientific) and 20–40 µg of proteins were resolved by 10% or 12% SDS-PAGE and then transferred to a nitrocellulose membrane (Amersham Protran). The membranes were blocked (2 h, RT in PBS and 0.1% Tween-20 with 5% skimmed milk) and then incubated with the primary (TABLA 2) (overnight 4ºC) and the secondary (HRP-conjugated anti-mouse or -rabbit, Cell Signaling) (2h, RT) antibodies diluted in PBS-T. Proteins were visualized with ECL (Bio-RAD) using the Imager 680 (Amersham).

Immunofluorescent and immunohistochemistry staining

Samples were fixed in 4% formalin overnight, dehydrated through a series of graded ethanol baths and then embedded in paraffin. The 5 µm thick sections obtained in a microtome were heated at 60ºC for 1 hour followed by deparaffination, hydration, and antigen retrieval with sodium citrate pH 6.0. After that, the slides were permeabilized with 1% Triton X-100 (Sigma-Aldrich) in PBS and blocked for 1 hour with 5% BSA (Sigma), 10% FBS (Sigma) and 0.1% Triton X-100 (Sigma). Primary antibodies (Supplementary TABLA 2) were incubated O/N at 4ºC. Prior to secondary antibody incubation (1:200 dilution) for 1h at room temperature, the sections were washed 3 times, 5 minutes each. Sections were then washed with PBS 5 minutes and then nuclei were counterstained with DAPI. Imaging was done with Leica SP-5 confocal microscope and THUNDER Imaging Systems. Cells were counted using the ImageJ software. For immunohistochemical analysis, after primary antibodies incubation the sections were incubated with HRP conjugated antibodies (1:200 dilution) (Supplementary TABLA 2). Target proteins were detected with the ABC Kit and the DAB kit (Vector Laboratories)

qRT-PCR assay

RNA isolation was performed using the RNA isolation Kit (Roche). Total RNA (1µg) was reverse transcribed with PrimeScript RT Reagent Kit (Takara). Quantitative real-time PCR was performed using the Light Cycler 1.5 (Roche) with the SYBR Premix Ex Taq (Takara) and specific primers for each gene (Supplementary TABLA 3). Gene expression was quantified by the delta-delta Ct method.

Cell culture

The GBM1 human GBM cells were maintained in DMEM-F12 plus 2,5% N2, 5% Neurocult SM1 neuronal supplement® (Stem cell technologies); GlutaMAX (1:100) (Invitrogen); penicillin-streptomycin (1:100) (Lonza); 0.4% heparin (Sigma-Aldrich); and 40 ng/ml EGF and 20 ng/ml bFGF2 (Peprotech). 2mM L-glutamine, 0.1% penicillin (100 U/ml) and streptomycin (100 µg/ml) as previously described from Segura-Collar et al 2021[16].

Lentivirus preparation and infection

Pseudotyped lentivectors were produced using reagents and protocols from those described [17–20] and Addgene lentiviral protocols, https://www.addgene.org/guides/lentivirus/, using Lentiviral and retroviral particles were produced in 293T cells with packaging plasmid pCMVdR8.74 (Addgene #Plasmid 22036) and VSV-G envelope protein plasmid pMD2G (Addgene #Plasmid 12259) using Lipofectamine and Plus reagent (Invitrogen). After the production of the LV-GFP, LV-IDH1 wt and LV-IDH1 mut lentiviruses, GBM1 cells were infected for 48 h adding the lentiviral supernatant and 4 mg/ml of polybrene. These GBM1 cells were implanted into nude mice.

Mouse model study

Animal experiments were reviewed and approved by the Research Ethics and Animal Welfare Committee at “Instituto de Salud Carlos III” (PROEX 02/16), in agreement with the European Union and national directives. Intracranial transplantations to establish orthotopic xenografts were performed injecting 50.000 to 100000 cells (resuspended in 2 µl of culture medium) with a Hamilton syringe into the striatum of Nude mice (A–P, -0.5 mm; M–L, + 2 mm, D–V, -3 mm; related to Bregma) using a Stoelting Stereotaxic device [10].

Rota-Rod test

Mice were placed in 47650 Rota-Rod NG (Ugo Basile) with a ramp/accelerating program with a starting speed of 5 rpm, and maximum speed 40 rpm in 1 minute. Four repetitions in cycles between both groups of mice were performed, so that they could rest between each analysis. The latency time to fall was recorded, eliminating measurements below 10 seconds.

Bioinformatic studies

For studies of gene expression and gene profiling, the merge TCGA dataset (diffuse Glioma) was used with 1153 enrolled patients and a set of 702 patients with Glioma tumors with RNAseq values (IlluminaHiSeq). The different signatures were obtained from MSigDB (The Molecular Signatures Database). For the GSEA (Gene Set Enrichment Analysis) study, the TREM2 expression was continuously computed through the Glioma data-set using the expression by RNAseq, then we used the method continuous class label and gene-sets from the “CGP: chemical and genetic perturbations” and “CP: Hallmark gene sets”. For correlation studies, the expression values were obtained from http://gliovis.bioinfo.cnio.es/. For gene ontology analysis, the DAVID gene ontology 6.9 and Panther program was used with a cluster of 1000 genes.

RNAseq

Quantification of gene expression

STAR was used to obtain the number of reads associated to each gene in the Gencode v31 annotation (restricted to protein-coding genes, antisense and lincRNAs). Raw counts for each sample were imported into R statistical software. Extracted count matrix was normalized for library size to compute CPM expression levels.

Unsupervised analysis: The Bioconductor edgeR package was used to import raw counts into R statistical software and compute normalized log2 CPM (counts per millions of mapped reads) using the TMM (weighted trimmed mean of M-values) as normalization procedure. The normalized expression matrix from the 1000 most variant genes (based on standard deviation) was used to classify the samples according to their gene expression patterns using principal component analysis (PCA), hierarchical clustering and consensus clustering. PCA was performed by FactoMineR::PCA function with “ncp = 10, scale.unit = FALSE” parameters. Hierarchical clustering was performed by stats::hclust function (with euclidean distance and ward.D method). Consensus clustering was performed by ConsensusClusterPlus::ConsensusClusterPlus function to examine the stability of the clusters. We established consensus partitions of the data set in K clusters (for K = 2, 3,. . ., 8), on the basis of 1,000 resampling iterations (80% of genes, 80% of sample) of hierarchical clustering, with Euclidean distance and ward.D method. Then, the cumulative distribution functions (CDFs) of the consensus matrices were used to determine the optimal number of clusters (K = 3 for instance), considering both the shape of the functions and the area under the CDF curves. tSNE analysis was performed with the Bioconductor Rtsne package applied to the PCA object (theta = 0.0, perplexity=, max_iter = 1000).

Differential expression analysis

The Bioconductor edgeR package was used to import raw counts into R statistical software. Differential expression analysis was performed using the Bioconductor limma package and the voom transformation. To improve the statistical power of the analysis, only genes expressed in at least one sample (CPM >= ) were considered. A qval threshold of < = 0.05 and a minimum fold change of 1.2 were used to define differentially expressed genes.

Pathway enrichment analysis – GSEA: Gene list from the differential analysis was ordered by decreasing log2 fold change. Gene set enrichment analysis was performed by clusterProfiler::GSEA function using the fgsea algorithm. Gene sets from MSigDB v7.4 database were selected among the C2_curated, C5_GO, C6_oncogenic, Hallmark classes, keeping only gene sets defined by 10–500 genes.

Statistical Analysis

No specific statistical method was used to predetermine the sample size, but the sizes of our cohort are like or greater than those reported in previous publications ^8,9,15,32,49. There was also no randomization or blinding due to the exploratory nature of the study. The normality of the data distribution was formally tested and, therefore, non-parametric or parametric statistical analyzes were used. Variance was also analyzed for the patient samples. Statistical significance of the results was analyzed using specific statistic tests with Prism version 8.0 using the test indicated within the corresponding figure caption or sections. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 were considered significant. SEM stands for Standard Error of the Mean. For ROC curve analysis signature score were considered responsive to treatment when D% patient > 10%. For Kaplan-Meier survival curves, the level of significance was determined by the two-tailed log-rank test. For correlation analysis between each gene, expression data were tested by Pearson’s correlation coefficient and Spearman´s correlation coefficient using GraphPad Prism 8.0 software.

Analysis of the composition of the adult diffuse glioma microenvironment by gene profiling

To begin our study, we re-evaluated the glioma microenvironment First, we improved the definition of the different areas by generating several gene expression signatures. Profiling of all the ecosystem-related signatures allowed clustering of the TCGA glioma cohort (IDH1-wt and mut) (Fig. 1A). To obtain a list of genes associated with synaptic function, we selected pure neuronal genes, such as MAP2, that were down-regulated on AD. This signature was named “Synapse” and we observed that it is highly expressed in the leading edge of the tumors (Fig. 1B). Similarly, we generated a transcriptomic signature associated with cell division rate markers (such as Ki67 and PCNA) in histological sections, which we termed "Proliferation rate". It accumulates in areas of microvascular proliferation in gliomas (Fig. 1C). Next, we defined two other transcriptomic signatures associated with glioma progression and the microenvironment, such as macrophage infiltration or hypoxia, termed “TAM inflammatory response” and “Hypoxia-induced angiogenesis”, respectively. The former accumulates in areas of microvascular proliferation (Fig. 1D), whereas the latter is enriched in pseudopalisading zones (Fig. 1E). Finally, we generated a “BBB-dysfunction” signature using a transcriptomic analysis of endothelial cells of the brain from four different neurological pathologies: stroke, traumatic brain injury (TBI), multiple sclerosis (EAE) and Seizure. The BBB-dysfunction signature is also enriched in pseudopalisading zones of gliomas (Fig. 1F). Moreover, we observed that the expression of most of these signatures increases with pathology progression, from normal tissue (NT) to GBM (Fig. 1G and Supplementary Fig. S1). Only the Synapses signature follows a reverse pattern (Fig. 1G and Supplementary Fig. S1). Figure 1H illustrates the clear inverse correlation between the BBB dysfunction and the Synapse profiles. Notably, the expression of these signatures showed a similar trend in AD patients compared to NT (Fig. 1I). Besides, we found a consistent increase in BBB dysfunction in those tumors that had progressed to a more aggressive phenotype (Fig. 2A), which was not observed when there was no change in histological classification of recurrent gliomas (Fig. 2B). In addition, high expression of the BBB dysfunction signature gene was associated with a worse prognosis in diffuse gliomas (Fig. 2C and Supplementary Fig. S2A-B) and with the detection of motor and cognitive dysfunction in these patients (Fig. 2E). By contrast, the Synapse signature was higher in the less aggressive tumors (Fig. 2D and Supplementary Fig. S2C-D) and in those glioma patients without strong neurological symptoms (Fig. 2F). These results suggest and inverse distribution of vascular alterations and neurodegeneration in gliomas and AD patients.

Defining small gene signatures of BBB-dysfunction and Synapse predictive of glioma and Alzheimer's disease aggressiveness

BBB dysfunction can be an early marker of AD and this correlates with the presence of genetic alterations of this pathology such as APOE4 [3, 21]. To validate the data presented above, we characterized the transcription of a reduced set of genes from the BBB-dysfunction and the Synapse signatures in our own cohorts of AD and glioma samples corresponding to different clinical stages. Tau (monomers and oligomers) (Fig. 3A-B) and pThr205-Tau (Fig. 3A and C) levels were increased in fully developed AD (Braak IV-VI) compared with incipient AD (Braak I-III) and normal brain. The subdivision into high-Braak (IV-VI) and Low-Braak (I-III) in the AD cohort was performed using the aggregation of Tau as an internal marker due to its involvement in the development of pathology (Fig. 3B-C and Supplementary Fig. S3A-B). Thus, it can also be observed that phosphorylation in Thr205-Tau, PHF1 and AT8 show an increase in Braak IV-VI (High) compared to healthy patients, with only a small increase between Braak I-III (Low) compared to healthy tissue in the case of AT8 (Fig. 3A-C and Supplementary Fig. S3C-E). To analyze the relevance of Synapse and BBB dysfunction genes, we selected those ones with higher fold change and p value to generate smaller signatures and evaluated their expression by qRT-PCR in our cohorts. This analysis showed a strong induction of the expression of BBB-dysfunction genes as Alzheimer’s disease progresses (Fig. 3D and Supplementary Fig. S4A), whereas the Synapse signature followed an inverse pattern (Fig. 3E and Supplementary Fig. S4B). Vascular and neuronal changes were already observed in early cases of AD. In the glioma cohort, the BBB-dysfunction signature was induced (Fig. 3F and Supplementary Fig. S4C), whereas the Synapse signature was reduced (Fig. 3G and Supplementary Fig. S4D) in GBM-IDH wt compared to astrocytoma IDH mut or normal tissue. Notably, no significant changes in the expression of these genes were observed when comparing astrocytoma IDH mut and normal tissue (Supplementary Fig. S4C-D). Our data suggest that reduced synapse expression could be due to neuronal loss in the context of BBB disruption and that this occurs only in the most aggressive tumors. Accordingly, immunofluorescence (IF) analysis of vascular leakage (IgG extravasation) and healthy neurons (NeuN positive cells) in human diffuse glioma samples (Fig. 3H), confirmed the direct correlation between the BBB dysfunction and the loss of neurons (Fig. 3I). Moreover, the data showed that tumors with a leakier BBB were associated with a reduced overall survival (Fig. 3J). These small signatures showed significant predictive value (specificity and sensitivity) in the glioma and the AD cohorts (Supplementary Fig. S4E-H). Thus, with all these data we demonstrate that the progression of AD and glioma can be characterized with small genetic signatures that define the degree of BBB dysfunction and the loss of neuronal synapses.

Mesenchymal GBM promote BBB dysfunction and neuronal loss

To address the potential contribution of BBB dysfunction to neuronal loss, we applied RNAseq of eight gliomas with Low versus High BBB dysfunction. We found genes in two up- and down-regulated groups (Fig. 4A). The first were enriched in the mesenchymal profile (MES) of GBM (Fig. 4B, left) and those down-regulated in the proneural profile (PN) (Fig. 4B, right), the neural system, and the synapse (Fig. 4C). We then stratified GBM into subtypes and found that expression of the BBB dysfunction signature preferentially accumulated within the MES subtype (Fig. 4D). Accordingly, the expression of MES markers in patient´s derived xenografts (PDXs) was higher in tumors with high levels of BBB disruption (Fig. 4E and Supplementary Fig. S5A), while PN transcripts accumulated in gliomas with low BBB-dysfunction (Fig. 4F and Supplementary Fig. S5A). To validate these observations, we performed IF characterization of PDXs, which represent the three glioma subtypes. Vascular leakage (measured by IgG and albumin extravasation) (Fig. 4G-H and Supplementary Fig. S5B) and neuronal loss (Fig. 4G and I) were clearly higher in MES, compared to PN or classical (CL) intracranial mouse tumors. In addition, we were also able to correlate MES gliomas with reduced motor abilities in mice implanted with this type of tumors (Supplementary Fig. S5C). On the other hand, when we compared MES (GBM1) and PN (GBM2) PDXs we observed that the former was much more aggressive (Fig. 4J) while having a greater expression of the BBB-dysfunction signature (Fig. 4K and Supplementary Fig. S5D) and lower expression of the Synapse signature (Fig. 4L and Supplementary Fig. S5E). These data suggest that BBB-disruption correlates with neuronal loss and motor dysfunction in MES gliomas. Indeed, we observed an accumulation of IDH1/mutations in tumors with lower expression of the BBB dysfunction signature and higher expression of the Synapse signature (Fig. 5A and Supplementary Fig. S6 A-B). Mutations in IDH1/2 generate a hypermethylation profile in a large group of genes (Fig. 5B), a phenotype called G-CIMP that is associated with a better prognosis of glioma patients (Fig. 5C). Taken together, these data allow us to speculate that gliomas with IDH mutated (PN subtype) maintains normal BBB permeability and neuronal functionality. To confirm this, we overexpressed IDH1 mutant (IDH1 R132H) in a MES glioma cell line (GBM1-IDHmut) (Supplementary Fig. S6C). Animals carrying GBM1-IDHmut cells showed increased survival than animals injected with control cells (Fig. 5D). Expression of IDH1mut reduced vessel leakage in tumors (Fig. 5E-F) and this vascular normalization was accompanied by an increase in the number of neurons in the peritumoral area (Fig. 5E and G). Accordingly, the BBB-dysfunction signature decreased (Fig. 5H and Supplementary Fig. S6D), whereas the Synapses signature increased (Fig. 5I and Supplementary Fig. S6E) after IDH1mut overexpression. Simultaneously, GBM1-IDHmut tumors showed a decrease in their motor performance in Rota-rod tests as the tumors grew (Fig. 5J). Notably, there was no decrease in the size of IDHmut compared to control tumors (Supplementary Fig. S6F), suggesting that the aggressiveness of the glioma is not only associated to tumor growth, but also with the changes that the tumor induces in the brain microenvironment.

Recently, we have begun to understand how epigenetic changes induce chromatin remodeling and generate gene expression patterns[22]. In this sense, it has been possible to establish that structural proteins such as CTCF (the chromatin insulator CCCTC-binding factor) generate DNA loops that isolate the promoter from regulatory sequences, which can be affected by methylation[23]. Considering the function of CTCF, we analyze how CTCF binding to DNA is affected by the hyper-methylation phenotype generated by IDH as seen in the Fig. 5B-F and Supplementary Fig. S6C. For this, we used the ChiP-seq data of CTCF in glioma samples with IDH wild-type or mutant previously published, this shows that among the methylated and over-expressed genes, in tumors with IDH-mutant. There is a large group of genes losing CTCF binding to CpG island sequences (Fig. 6A) and this loss of CTCF binding does not occur at random sequences (Fig. 6B). To understand how CTCF binds to DNA and how this process is regulated by the presence of mutant HDI, we analyzed the pattern of CTCF around the TSS transcription start site. Similar to what was observed before, in these areas of transcriptional regulation CTCF binding is also reduced at the CpG-islands of DNA (Fig. 6C-D). This change in the pattern in the binding of CTCF affects a group of genes associated with intercellular junction genes that could allow the establishment of a more stable BBB and less susceptible of being permeabilized by the presence of perivascular tumor cells (Fig. 6E)[8, 10]. This occurs due to loss of the TAD structures in CpG areas which correlates with greater transcriptional activation, for example the CIPC/ZDHHC22 gene pair (Fig. 6F).

Inflammatory and immunosuppressive processes associated with myeloid cells define the progression of CNS pathologies

Neuroinflammation processes are strongly associated with multiple CNS pathologies and with loss of BBB integrity, but the connection between the two processes is not fully understood [1, 24]. The subclassification of GBM into PN, MES and CL subtypes has allowed the establishment of a positive feedback interaction between mesenchymal tumor cells and immune cells [25, 26]. Using flow cytometry analysis, we were able to observe a strong increase in the number of CD45 + cells in the high-BBB dysfunction compared to the low-BBB dysfunction glioma group (from our own stratified glioma cohort), which corresponded to myeloid cells (macrophages, MDSC (myeloid-derived-suppressor-cells) and neutrophils) and not to lymphoid cells (Fig. 7A-D and Supplementary Fig. S7A-B). The myeloid population of gliomas is composed of brain-resident microglia and monocytes-derived macrophages (MDM), infiltrated from the periphery [27]. Immunohistochemical (IHC) analysis showed an increase in CD68 + myeloid cells without differences in the amount of IBA-1 + cells, or other markers such as SALL1, TMEM119 and P2RY12; normally used to detect microglia (Supplementary Fig. S7C-E). Furthermore, we observed a higher number of CD68 + cells expressing CD49d, whose expression has been previously used to distinguish MDM and microglia [28], in those tumors with a higher expression of the BBB-dysfunction signature (Fig. 7E). These two populations can also be differentiated based on the expression of a defined genetic signature [27]. The TCGA glioma dataset stratified into High and Low BBB-dysfunction groups corroborated that the MDM signature is increased in the former compared to the latter (Supplementary Fig. S7F), whereas no significant changes in the microglia signature were observed between both groups (Supplementary Fig. S7G). Based on these results, we derived a representative MDM signature, consisting of 9 genes, to measure their expression in our own glioma and AD cohorts. The results showed a correlation between MDM and BBB-dysfunction signatures in the glioma (Fig. 6F) and the AD cohort (Fig. 7G). Our results suggest that CD68/CD49d cells may be critical for the development of a chronic inflammation process associated with BBB-dysfunction [29–31]. To consolidate this hypothesis, we evaluated altered myeloid markers that are specifically associated with AD and cancer, such as TREM2 [32, 33]. We observed an increase in CD68/CD49d/TREM2 cells in gliomas with High BBB-dysfunction (Fig. 7H). Then, we included immunosuppressive markers that are a hallmark of the glioma microenvironment. Thus, we generated an immunossupressive signature in glioma-associated macrophages using scRNAseq, further curated on TREM2-based expression (Fig. 7I) and an inflammatory signature associated with the production of the following cytokines IL1A, IL1B, IL1R1, IL6, CCL2, CCL7, CCL8, TNF and ICAM1. We observed that both signatures were increased simultaneously in glioma and AD samples with high levels of vascular disruption (Fig. 7J-K). Overall, we propose that the brain microenvironment, when permeable to immune cells, establishes a chronic inflammatory profile with a process of inflammation and myeloid suppression associated with the presence of CD68/CD49D cells.

Diseases of the central nervous system continue to be a pending issue in global health systems, and pathologies such as gliomas and AD have shown little or no clinical improvement in recent decades. Aging of the blood-brain barrier is recognized as one of the main risk factors in the development of brain pathologies such as cancer and neurodegeneration, but the mechanisms are not fully understood [34, 35]. Association studies between these two almost antagonistic diseases has shown common features, which has aroused great interest in public health systems [36]. We have recently begun to understand the molecular relationship between cancer and neuroscience, where a series of interactions play a role in favor of both tumor pathology and cognitive impairment [37–39]. We and others have demonstrated the important role of proteins associated to neurodegeneration pathologies, such as Tau or APP, in the development of gliomas or brain metastases [10, 40]. The results presented here reinforce pre-existing ideas about the function of the BBB as a direct regulator of immune disfunction in the brain [35, 41]. The molecular mechanism of glioma neurotoxicity was incipiently studied by Savaskan et at 2008 and our results show a specific neurotoxicity in GBM with a MES profile, due to the alteration of the BBB that generates a severe neuroinflammatory process directly associated with myeloid cell dysfunction [42].

Regarding AD, we have observed that the dysfunction of the brain barrier allows the entry of myeloid cells, which establishes a chronic pattern of inflammatory cytokines and immunosuppressive signals. This correlate is in with what was demonstrated in AD by Chen et al. to the. 2023 (cita): where myeloid cell dysfunction leads to exhausted T-cell infiltration. Besides, it has recently been observed in AD that myeloid cells stimulate T cell entry which determines neurotoxicity associated with interferon production and exhausted cell generation[43]., These immune infiltration processes are highly neurotoxic in Taupathies and we suggest here that they also generate neuronal loss in GBM. This axis pattern of BBB-rupture/Neuroinflammation/Myeloid-lymphocyte dysfunction/cognitive impairment is common to many pathologies of the central nervous system. It has even been observed in the respiratory failure of COVID-19 that myeloid cells in the brain can generate neurotoxicity[44]. In these patients, a strong immune alteration in the brain has been observe in relation to a population of myeloid cells (CD14, C1QC and FTH1) like those observed in neurodegeneration such as AD[44]. We speculate that neuroinflammatory processes are a common booster of multiple pathologies and are defined by brain aging. Other authors have demonstrated a varied infiltration of immune cells in the brain in Alzheimer's disease but its pathology implication is still unknown[45].

With the establishment of immunotherapy as one of the therapeutic promises against cancer, numerous clinical trials have shown us the way to advance the knowledge of the relevant immune populations to awaken intrinsic defenses against the tumor. Unfortunately, immunotherapy-based trials (NCT02017717 and NCT02617589) against GBM have failed, [17–20]. Thus, we need to better understand how these tumors evade the immune system. In this sense, some recent studies have provided a better characterization of the myeloid component in GBM [46], or proposed myeloid markers such as CD73 as a therapeutic target [47]. But more information is necessary about the processes of generation of myeloid dysfunction or how these cells prevent the functioning of the immune checkpoint blockade [48]. Our data indicate that BBB disruption perturbs the immune landscape of the brain by increasing inflammatory processes, including inflammatory cytokines and myeloid suppressor cells. This profile of chronic brain inflammation promotes myeloid cell dysfunction, characterized by the expression of CD68+/CD49d+/TREM2 + markers. Currently, there is a great interest in identifying molecular patterns that better define the dysfunction of myeloid populations in brain pathologies [46]. Despite the large number of characterization articles, we know little about how to block the function this type of myeloid cells. These are master regulators of immune suppression that ultimately leads to the generation of exhausted T cells [49]. Our work sheds light on one of the most complex scenarios, common vascular/immune regulation in different CNS pathologies, and opens a new therapeutic window for a wide variety of pathologies.

Acknowledgments

We thank all members of the Neuro-oncology Group and UNMO, M. Caamaño-Moreno, Sara Hiller-Vallina, Blanca Comitre, Dennise Alcivar and Marina I. Garin for critical discussion of the manuscript and the Pathological Anatomy Service and IMAS12 Core Facilities and the Animal Facilities Service (ISCIII) for their excellent assistance.

Author contributions

B.S-C: Conceptualization; Data curation; Formal analysis; Validation; Investigation; Visualization; Methodology; Writing-original draft; Writing-review and editing. L.M-R., M.G-A., P.M.: Formal analysis; Investigation; Visualization; Methodology. R J.A.; A.H-L; J.M.S.; A.P.N: Human samples and clinical evaluation; Supervision; Visualization. P.S-G.: Resources, Formal analysis; Supervision; Funding acquisition; Writing-review and editing. B.S-C; L.M-R; M.G-A; A.H-L; J.M.S.; A.P.N: provided and processed the human samples and collected clinical data; R.G.: Conceptualization; Resources; Formal analysis; Supervision; Funding acquisition; Validation; Investigation; Visualization; Methodology; Writing-original draft; Writing-review and editing.

Competing of interests

The authors declare no conflict of interest.

Ethical Approval

The samples from glioma patients were collected from surgeries performed at Hospital Universitario 12 de Octubre, (Madrid, Spain). The samples from Alzheimer patient were obtained from “Fundación CIEN” (Madrid, Spain). Samples were obtained after patient's written consent and with the approval of the Ethical Committee at “Hospital 12 de Octubre” and “Fundación CIEN” (CEI_ 21/551 and CEI PI-74_2019 respectively). They are not samples derived from a clinical trial. All mice used in this study were produced under specific pathogen free (SPF) conditions at the animal facility of the Insituto de Salud Carlos III, Madrid (Spain). Animal experiments were reviewed and approved by the Research Ethics and Animal Welfare Committee at “Instituto de Salud Carlos III” (PROEX 02/16), in agreement with the European Union and national directives.

Funding

This study was funded by ISCIII and FEDER funds (CP21/00116 and PI22/01171) (R.G.), PI21CIII/00002 (P.S.G), PI21/01406 (J.M.S), PI21/01168 (A.P.N), and by FC-AECC (Investigator AECC 2017 INVES192GARG) (RG) (IDEAS20095SÁNC) (R.G. and P.S-G.).

Availability of data and materials

The RNA-seq datasets for the expression counts will be published in the Sequence Read Archive (SRA). Due to the protection of patient data, the raw RNA-seq data will be available upon request. For the study, the gene signatures are published in MsigDB (http://www.gsea-msigdb.org/gsea/msigdb) and the combinations attached in an excel in the extended data.

Segura-Collar B, Mata-Martínez P, Hernández-Laín A, Sánchez-Gómez P, Gargini R: Blood-Brain Barrier Disruption: A Common Driver of Central Nervous System Diseases. Neuroscientist 2022, 28:222-237.
Munji RN, Soung AL, Weiner GA, Sohet F, Semple BD, Trivedi A, Gimlin K, Kotoda M, Korai M, Aydin S, et al: Profiling the mouse brain endothelial transcriptome in health and disease models reveals a core blood-brain barrier dysfunction module. Nat Neurosci 2019, 22:1892-1902.
Nation DA, Sweeney MD, Montagne A, Sagare AP, D'Orazio LM, Pachicano M, Sepehrband F, Nelson AR, Buennagel DP, Harrington MG, et al: Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. NatMed 2019.
Sweeney MD, Zhao Z, Montagne A, Nelson AR, Zlokovic BV: Blood-Brain Barrier: From Physiology to Disease and Back. Physiol Rev 2019, 99:21-78.
Gargini R, Segura-Collar B, Sánchez-Gómez P: Cellular Plasticity and Tumor Microenvironment in Gliomas: The Struggle to Hit a Moving Target. Cancers (Basel) 2020, 12.
Louis DN, Perry A, Wesseling P, Brat DJ, Cree IA, Figarella-Branger D, Hawkins C, Ng HK, Pfister SM, Reifenberger G, et al: The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro Oncol 2021, 23:1231-1251.
Ceccarelli M, Barthel FP, Malta TM, Sabedot TS, Salama SR, Murray BA, Morozova O, Newton Y, Radenbaugh A, Pagnotta SM, et al: Molecular Profiling Reveals Biologically Discrete Subsets and Pathways of Progression in Diffuse Glioma. Cell 2016, 164:550-563.
Segura-Collar B, Garranzo-Asensio M, Herranz B, Hernández-SanMiguel E, Cejalvo T, Casas BS, Matheu A, Pérez-Núñez Á, Sepúlveda-Sánchez JM, Hernández-Laín A, et al: Tumor-derived pericytes driven by EGFR mutations govern the vascular and immune microenvironment of gliomas. Cancer Res 2021.
Cejalvo T, Gargini R, Segura-Collar B, Mata-Martínez P, Herranz B, Cantero D, Ruano Y, García-Pérez D, Pérez-Núñez Á, Ramos A, et al: Immune Profiling of Gliomas Reveals a Connection with IDH1/2 Mutations, Tau Function and the Vascular Phenotype. Cancers (Basel) 2020, 12.
Gargini R, Segura-Collar B, Herranz B, Garcia-Escudero V, Romero-Bravo A, Nunez FJ, Garcia-Perez D, Gutierrez-Guaman J, Ayuso-Sacido A, Seoane J, et al: The IDH-TAU-EGFR triad defines the neovascular landscape of diffuse gliomas. SciTranslMed 2020, 12.
Amankulor NM, Kim Y, Arora S, Kargl J, Szulzewsky F, Hanke M, Margineantu DH, Rao A, Bolouri H, Delrow J, et al: Mutant IDH1 regulates the tumor-associated immune system in gliomas. Genes Dev 2017, 31:774-786.
Alghamri MS, McClellan BL, Avvari RP, Thalla R, Carney S, Hartlage CS, Haase S, Ventosa M, Taher A, Kamran N, et al: G-CSF secreted by mutant IDH1 glioma stem cells abolishes myeloid cell immunosuppression and enhances the efficacy of immunotherapy. Sci Adv 2021, 7:eabh3243.
Hokama M, Oka S, Leon J, Ninomiya T, Honda H, Sasaki K, Iwaki T, Ohara T, Sasaki T, LaFerla FM, et al: Altered expression of diabetes-related genes in Alzheimer's disease brains: the Hisayama study. Cereb Cortex 2014, 24:2476-2488.
Berchtold NC, Cribbs DH, Coleman PD, Rogers J, Head E, Kim R, Beach T, Miller C, Troncoso J, Trojanowski JQ, et al: Gene expression changes in the course of normal brain aging are sexually dimorphic. Proc Natl Acad Sci U S A 2008, 105:15605-15610.
Cejalvo T, Gargini R, Segura-Collar B, Mata-Martínez P, Herranz B, Cantero D, Ruano Y, García-Pérez D, Pérez-Núñez Á, Ramos A, et al: Immune profiling of gliomas reveals a connection with Tau function and the tumor vasculature. bioRxiv 2020:2020.2007.2017.208165.
Segura-Collar B, Garranzo-Asensio M, Herranz B, Hernández-SanMiguel E, Cejalvo T, Casas BS, Matheu A, Pérez-Núñez Á, Sepúlveda-Sánchez JM, Hernández-Laín A, et al: Tumor-Derived Pericytes Driven by EGFR Mutations Govern the Vascular and Immune Microenvironment of Gliomas. Cancer Res 2021, 81:2142-2156.
Reardon DA, Brandes AA, Omuro A, Mulholland P, Lim M, Wick A, Baehring J, Ahluwalia MS, Roth P, Bähr O, et al: Effect of Nivolumab vs Bevacizumab in Patients With Recurrent Glioblastoma: The CheckMate 143 Phase 3 Randomized Clinical Trial. JAMA Oncol 2020, 6:1003-1010.
Zhao J, Chen AX, Gartrell RD, Silverman AM, Aparicio L, Chu T, Bordbar D, Shan D, Samanamud J, Mahajan A, et al: Immune and genomic correlates of response to anti-PD-1 immunotherapy in glioblastoma. Nat Med 2019, 25:462-469.
Cloughesy TF, Mochizuki AY, Orpilla JR, Hugo W, Lee AH, Davidson TB, Wang AC, Ellingson BM, Rytlewski JA, Sanders CM, et al: Neoadjuvant anti-PD-1 immunotherapy promotes a survival benefit with intratumoral and systemic immune responses in recurrent glioblastoma. Nat Med 2019, 25:477-486.
Schalper KA, Rodriguez-Ruiz ME, Diez-Valle R, Lopez-Janeiro A, Porciuncula A, Idoate MA, Inoges S, de AC, Lopez-Diaz de CA, Tejada S, et al: Neoadjuvant nivolumab modifies the tumor immune microenvironment in resectable glioblastoma. NatMed 2019, 25:470-476.
Montagne A, Nation DA, Sagare AP, Barisano G, Sweeney MD, Chakhoyan A, Pachicano M, Joe E, Nelson AR, D'Orazio LM, et al: APOE4 leads to blood-brain barrier dysfunction predicting cognitive decline. Nature 2020, 581:71-76.
Phillips RE, Soshnev AA, Allis CD: Epigenomic Reprogramming as a Driver of Malignant Glioma. Cancer Cell 2020, 38:647-660.
Nora EP, Goloborodko A, Valton AL, Gibcus JH, Uebersohn A, Abdennur N, Dekker J, Mirny LA, Bruneau BG: Targeted Degradation of CTCF Decouples Local Insulation of Chromosome Domains from Genomic Compartmentalization. Cell 2017, 169:930-944.e922.
Smith BC, Tinkey RA, Shaw BC, Williams JL: Targetability of the neurovascular unit in inflammatory diseases of the central nervous system. Immunol Rev 2022, 311:39-49.
Hara T, Chanoch-Myers R, Mathewson ND, Myskiw C, Atta L, Bussema L, Eichhorn SW, Greenwald AC, Kinker GS, Rodman C, et al: Interactions between cancer cells and immune cells drive transitions to mesenchymal-like states in glioblastoma. Cancer Cell 2021, 39:779-792.e711.
Wang L, Jung J, Babikir H, Shamardani K, Jain S, Feng X, Gupta N, Rosi S, Chang S, Raleigh D, et al: A single-cell atlas of glioblastoma evolution under therapy reveals cell-intrinsic and cell-extrinsic therapeutic targets. Nat Cancer 2022, 3:1534-1552.
Klemm F, Maas RR, Bowman RL, Kornete M, Soukup K, Nassiri S, Brouland JP, Iacobuzio-Donahue CA, Brennan C, Tabar V, et al: Interrogation of the Microenvironmental Landscape in Brain Tumors Reveals Disease-Specific Alterations of Immune Cells. Cell 2020, 181:1643-1660.e1617.
Bowman RL, Klemm F, Akkari L, Pyonteck SM, Sevenich L, Quail DF, Dhara S, Simpson K, Gardner EE, Iacobuzio-Donahue CA, et al: Macrophage Ontogeny Underlies Differences in Tumor-Specific Education in Brain Malignancies. Cell Rep 2016, 17:2445-2459.
Zhang J, Liu Y, Zheng Y, Luo Y, Du Y, Zhao Y, Guan J, Zhang X, Fu J: TREM-2-p38 MAPK signaling regulates neuroinflammation during chronic cerebral hypoperfusion combined with diabetes mellitus. J Neuroinflammation 2020, 17:2.
Wu K, Byers DE, Jin X, Agapov E, Alexander-Brett J, Patel AC, Cella M, Gilfilan S, Colonna M, Kober DL, et al: TREM-2 promotes macrophage survival and lung disease after respiratory viral infection. J Exp Med 2015, 212:681-697.
Nazmi A, Field RH, Griffin EW, Haugh O, Hennessy E, Cox D, Reis R, Tortorelli L, Murray CL, Lopez-Rodriguez AB, et al: Chronic neurodegeneration induces type I interferon synthesis via STING, shaping microglial phenotype and accelerating disease progression. Glia 2019, 67:1254-1276.
Xuan W, Lesniak MS, James CD, Heimberger AB, Chen P: Context-Dependent Glioblastoma-Macrophage/Microglia Symbiosis and Associated Mechanisms. Trends Immunol 2021, 42:280-292.
Leng F, Edison P: Neuroinflammation and microglial activation in Alzheimer disease: where do we go from here? Nat Rev Neurol 2021, 17:157-172.
López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G: Hallmarks of aging: An expanding universe. Cell 2023, 186:243-278.
Banks WA, Reed MJ, Logsdon AF, Rhea EM, Erickson MA: Healthy aging and the blood-brain barrier. Nat Aging 2021, 1:243-254.
Sánchez-Valle J, Tejero H, Fernández JM, Juan D, Urda-García B, Capella-Gutiérrez S, Al-Shahrour F, Tabarés-Seisdedos R, Baudot A, Pancaldi V, Valencia A: Interpreting molecular similarity between patients as a determinant of disease comorbidity relationships. Nat Commun 2020, 11:2854.
Gargini R, Segura-Collar B, Sanchez-Gomez P: Novel Functions of the Neurodegenerative-Related Gene Tau in Cancer. Front Aging Neurosci 2019, 11:231.
Pan C, Winkler F: Insights and opportunities at the crossroads of cancer and neuroscience. Nat Cell Biol 2022, 24:1454-1460.
Winkler F, Venkatesh HS, Amit M, Batchelor T, Demir IE, Deneen B, Gutmann DH, Hervey-Jumper S, Kuner T, Mabbott D, et al: Cancer neuroscience: State of the field, emerging directions. Cell 2023, 186:1689-1707.
Kleffman K, Levinson G, Rose IVL, Blumenberg LM, Shadaloey SAA, Dhabaria A, Wong E, Galán-Echevarría F, Karz A, Argibay D, et al: Melanoma-Secreted Amyloid Beta Suppresses Neuroinflammation and Promotes Brain Metastasis. Cancer Discov 2022, 12:1314-1335.
Rustenhoven J, Kipnis J: Brain borders at the central stage of neuroimmunology. Nature 2022, 612:417-429.
Savaskan NE, Heckel A, Hahnen E, Engelhorn T, Doerfler A, Ganslandt O, Nimsky C, Buchfelder M, Eyüpoglu IY: Small interfering RNA-mediated xCT silencing in gliomas inhibits neurodegeneration and alleviates brain edema. Nat Med 2008, 14:629-632.
Chen X, Firulyova M, Manis M, Herz J, Smirnov I, Aladyeva E, Wang C, Bao X, Finn MB, Hu H, et al: Microglia-mediated T cell infiltration drives neurodegeneration in tauopathy. Nature 2023, 615:668-677.
Yang AC, Kern F, Losada PM, Agam MR, Maat CA, Schmartz GP, Fehlmann T, Stein JA, Schaum N, Lee DP, et al: Dysregulation of brain and choroid plexus cell types in severe COVID-19. Nature 2021, 595:565-571.
Zhang H, Cao S, Xu Y, Sun X, Fei M, Jing Q, Xu X, Tang J, Niu B, Li C: Landscape of immune infiltration in entorhinal cortex of patients with Alzheimer's disease. Front Pharmacol 2022, 13:941656.
Pombo Antunes AR, Scheyltjens I, Lodi F, Messiaen J, Antoranz A, Duerinck J, Kancheva D, Martens L, De Vlaminck K, Van Hove H, et al: Single-cell profiling of myeloid cells in glioblastoma across species and disease stage reveals macrophage competition and specialization. Nat Neurosci 2021, 24:595-610.
Goswami S, Walle T, Cornish AE, Basu S, Anandhan S, Fernandez I, Vence L, Blando J, Zhao H, Yadav SS, et al: Immune profiling of human tumors identifies CD73 as a combinatorial target in glioblastoma. Nat Med 2020, 26:39-46.
Jackson CM, Choi J, Lim M: Mechanisms of immunotherapy resistance: lessons from glioblastoma. Nat Immunol 2019, 20:1100-1109.
Kersten K, Hu KH, Combes AJ, Samad B, Harwin T, Ray A, Rao AA, Cai E, Marchuk K, Artichoker J, et al: Spatiotemporal co-dependency between macrophages and exhausted CD8. Cancer Cell 2022, 40:624-638.e629.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

BBB dysfunction drives neurotoxic myeloid cells in glioblastoma and Alzheimer’s disease

Status:

Version 1

Abstract

Figures

Background

Materials and Methods

Patient’s samples

Public data cohort, curation and signatures discovery

Characterization of immune cells by flow cytometry

Western Blot

Immunofluorescent and immunohistochemistry staining

qRT-PCR assay

Cell culture

Lentivirus preparation and infection

Mouse model study

Rota-Rod test

Bioinformatic studies

RNAseq

Quantification of gene expression

Statistical Analysis

Results

Analysis of the composition of the adult diffuse glioma microenvironment by gene profiling

Defining small gene signatures of BBB-dysfunction and Synapse predictive of glioma and Alzheimer's disease aggressiveness

Mesenchymal GBM promote BBB dysfunction and neuronal loss

Inflammatory and immunosuppressive processes associated with myeloid cells define the progression of CNS pathologies

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1