A multi-omics analysis of CSF reveals the differential immunopathology molecular mechanisms underlying viral encephalitis and autoimmune encephalitis

doi:10.21203/rs.3.rs-4939988/v1

Download PDF

Research Article

A multi-omics analysis of CSF reveals the differential immunopathology molecular mechanisms underlying viral encephalitis and autoimmune encephalitis

https://doi.org/10.21203/rs.3.rs-4939988/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background: This study explores viral encephalitis (VE) and autoimmune encephalitis (AE), similar in symptoms but requiring distinct treatments. Early diagnosis is crucial for effective intervention. The research utilizes early cerebrospinal fluid (CSF) analysis, employing transcriptomics, proteomics, and metabolomics to understand the immunological aspects of VE and AE.

Methods: Participants from the IMPROVE clinical trial (ClinicalTrial.gov, NCT04946682, July 1, 2021) between April 2018 and November 2022 were included. CSF samples at disease onset were analyzed for VE and AE immunological profiles.

Results: CSF analysis from 34 VE patients, 29 AE patients, and 6 controls unveiled shared and distinct immune mechanisms. Compared to controls, VE and AE patients showed reduced neuronal transporter functions and increased T and B cell activation. VE exhibited heightened interferon responses, complement pathways, and CD8⁺ T cell functions. AE displayed unique modulations in CD4⁺ T cell and Treg cell activities, potentially reducing autoimmune responses. Both conditions induced damage to critical brain areas.

Conclusion: The study identified immunological differences and similarities, emphasizing specific CSF molecular changes for differential diagnosis. Findings highlighted complex immune interactions, with both diseases involving suppressed neuronal functions and heightened T and B cell activities. VE featured distinct interferon and CD8⁺ T cell activations, while AE showed specialized CD4⁺ T and Treg cell modulations. Immune balance played a crucial role in encephalitis pathogenesis. Further studies are crucial to validate biomarkers for accurate diagnosis, paving the way for targeted therapies and improved management of these neurological disorders.

viral encephalitis

autoimmune encephalitis

transcriptomics

proteomics

metabolomics

cerebrospinal fluid

Encephalitis is an inflammatory disorder of the brain parenchyma, with neurologic dysfunction and symptoms including a change in behavior, seizures, abnormal movements, dysautonomia, and altered mental status^[1]. Encephalitis can be life-threatening and represents a challenging medical condition as it contributes to high morbidity worldwide with heavy burden to patients and society. The estimated prevalence of global encephalitis incidence is between 1.9 and 14.3 per 100, 000 inhabitants each year^[2]. And the reported expenses for encephalitis-associated hospitalization strike to $2.0 billion in the US in 2010^[2], while the annual cost for viral encephalitis reached to $350 million to $540 million, excluding post-discharge care, family caregiver costs, and income loss.

Encephalitis is categorized into infectious encephalitis and AE depending on its underlying causes. The infectious encephalitis results from microbial infections, with viruses being the predominant pathogens identified in most cases^[3–6]. While AE is resulted from immune dysregulation, which is associated with antibodies binding to extracellular or intracellular antigens pathologically, such as synaptic receptors and ion channels^{[7, 8]}. VE has many clinical manifestations overlap with AE, including symptoms like fever, seizures, and changes in consciousness^{[3, 9–11]}. Meanwhile, both AE and VE can lead to immune inflammation in the central nervous system. Inflammatory cells would be recruited during central nervous system infection to control the infections that may also contribute to life-threatening pathology^[12]. And AE is also characterized by infiltration of the CNS by immune cells, which would further result in axonal damage and neurodegeneration.

Therefore, we aim to explore the shared and distinct immune pathogenic mechanisms underlying VE and AE through transcriptomes and proteomics, shedding light on the mechanisms behind their similar and different clinical presentations. It may contribute to distinguishing between the two types of encephalitis, providing more precise diagnostic and therapeutic methods, and establishing a foundation for intervention strategies targeting specific mechanisms. Simultaneously, contrasting the immunopathology between AE and VE will enhance our comprehensive understanding the innate and adaptive mechanisms that balance protective and pathological inflammation in the central nervous system, thereby benefiting future exploration in therapeutic options and improving the prognosis of encephalitis.

Patient enrollment and sample collection

We recruited patients from the IMPROVE (ClinicalTrial.gov, NCT04946682), a multicenter, open-label, randomized controlled clinical trial conducted between April 2018 and November 2022. A total of 69 patients were selected. This included 34 patients diagnosed with viral encephalitis who were grouped into the VE Group, 29 patients diagnosed with autoimmune encephalitis who were grouped into the AE Group, and 6 with encephalitis-like symptoms but negative for viral or autoimmune etiologies as the Control Group. CSF collected at disease onset was analyzed by transcriptomics, proteomics, and metabolomics to elucidate the commonalities and differences in immune mechanisms between viral and autoimmune encephalitis. The study protocol and informed consent were approved by the Ethics Committee of Huashan Hospital (protocol number: 2021-KY451).

Total RNA extraction and meta-transcriptome sequencing

For metatranscriptomic sequencing, total RNA from CSF samples was extracted using QIAamp® Viral RNA Kit (Qiagen, Valencia, CA, USA), then subjected to human rRNA depletion by application of Ribo-Zero rRNA Removal Kit (Illumina, San Diego, CA, USA), following the manufacturer’s instructions. After removing the contaminating DNA via DNase I, cDNA was generated through reverse transcriptase and dNTPs. Nextera XT DNA Library Prep Kit (Illumina) was then applied for the construction of cDNA library, which was subsequently sequenced on an Illumina NextSeq 550 System (Illumina) using a single-end 75bp sequencing strategy.

Meta-transcriptomic data analysis

Before bioinformatics analysis, the raw sequencing data were quality controlled by fastp^[13] to remove low-quality reads (containing adapter, containing more than 10% N, base numbers with Q ≤ 20 account for more than 50%), and the quality of generated clean reads was assessed by FastQC (https://github.com/s-andrews/FastQC).

All clean data were mapped to the human genome hg38 using HISAT2^[14] with default parameters, and then FeatureCounts^[15] was used to quantify expression at the gene level. Differentially expressed genes (DEGs) were determined with an adjusted P value of < 0.05 and absolute log fold change (log2FC) of ≥ 0.585. With the application of clusterProfiler package^[16], gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Reactome pathway enrichment analysis of DEGs, as well as Gene Set Enrichment Analysis (GSEA) were performed, with an adjusted P value of < 0.05 showed significant enrichment.

The immune-related gene (IRG) list was downloaded from the ImmPortDB^[17] database (https://immport.org/shared/home), and interferon (IFN)-stimulated genes (ISGs) were retrieved from a recognized study^[18]. According to the gene expression profile of IRGs and ISGs, the Wilcoxon Rank Sum test was used to select genes that were significantly different between two groups. To infer the composition of immune cells, the CIBERSORT^[19] algorithm was applied and significant differences of proportions between two groups were judged by Wilcoxon Rank Sum test.

Analysis of gene co-expression network was performed by weighted gene co-expression network analysis (WGCNA) algorithm^[20]. Comparison of differences in the global transcriptional profiles between two groups in Principal Component Analysis (PCA) plot were assessed by Permutational multivariate analysis of variance (PERMANOVA). STRING database^[21] was employed for predicting Protein-Protein Interaction (PPI) networks, which were then trimmed and visualized through Cytoscape^[22].

To construct clinical prediction model, sequential forward selection algorithm and random forest classifier were used to screen out the gene subset that could distinguish the two groups to the greatest extent. The performance of each model was assessed by AUC values, which were produced by ten-fold cross-validation.

Proteome analysis

Sample Depletion, Digestion and TMT labeling

Initially, samples were thawed from − 80℃. High-abundance proteins were removed using specific beads in a centrifugation column. This involved shaking and incubating plasma samples with the beads, followed by centrifugation to separate the depleted sample. Subsequent steps included ultrafiltration for glycerol removal and transfer of the filtered liquid into a new tube. The samples underwent a lysis process using a urea and triethylammonium bicarbonate solution. This was followed by reduction with tris(2-carboxymethyl) phosphine and alkylation with iodoacetamide in a dark room. Protein digestion was performed using lysc and trypsin. The reaction was quenched using trifluoroacetic acid, and the peptides were then desalted. TMTpro 16plex labeling reagent was used for peptide labeling, combining all samples into a single mixture for analysis.

High pH Fractionation

Fractionation was carried out using a Waters XBridge Peptide BEH C18 column under a DIONEX UltiMate 3000 Liquid Chromatogram system. The mobile phases included ammonium hydroxide and acetonitrile. Peptides were collected at intervals across a gradient and combined into 30 distinct fractions. These were then dried and reconstituted for LC-MS analysis.

MS analysis

The LC-MS/MS analysis was performed using a DIONEX UltiMate 3000 RSLCnano System coupled to an Orbitrap Exploris 480 mass spectrometer. This setup included a FAIMS Pro™ interface, operating in data-dependent acquisition mode. Buffers for the analysis consisted of acetonitrile and water with formic acid. Peptides were loaded onto a precolumn and then separated using a specific LC gradient. The mass spectrometry parameters were set to capture a defined range of m/z values with specific resolutions and AGC targets.

Database search and statistical analysis

MS data were analyzed using the Proteome Discoverer (version 2.5.0.400, Thermo Fisher) search engine against the human protein database downloaded from SwissProt (version 15/07/2020; 20368), with a precursor ion mass tolerance of 10 ppm, and fragment ion mass tolerance of 0.02 Da. Briefly, TMT pro-plex labels at lysine residues and the N-terminus, and carbamidomethylation of cysteine residues were set as static modifications. A cut-off criterion of a q-value of 0.01, corresponding to a 1% FDR, was set for filtering-identified peptides with highly confident peptide hits. After filtering proteins with 80% missing rate, 1967 proteins were used for differential expression analysis.

Several pathway analysis tools were used to perform functional analysis of the significant DEPs. Enrichment analyses based on Gene Ontology processes, KEGG pathways, reactome gene sets were performed using the web-based platform Metascape^[23]. In the proteomics dataset, missing values were imputed with the minimum value and zero, respectively. Corrected p-values were calculated using the Benjamini & Hochberg correction method. Proteins with statistically significant changes were identified using a threshold of an adjusted p-value less than 0.05. Figure 6 shows the brain created with BioRender.com.

Metabolic sample preparation and data analysis

To prepare the samples, 100 µL of serum (equilibrated to room temperature) was mixed with 20 µL of L-2-chlorophenylalanine (Shanghai HC Biotech Co., Ltd., concentration: 0.06mg/mL) and vortexed for 10 seconds. A 300 µL methanol-acetonitrile solution (V/V = 2/1) was then added and the mixture was agitated for 1 minute. This was followed by an ultrasonic extraction in an ice water bath for 10 minutes, after which the mixture was allowed to rest at -20℃ for 30 minutes. Centrifugation was carried out at 13,000 RPM at 4℃ for 10 minutes, with 200 µL of the supernatant being transferred to an LC-MS vial and dried. The residue was reconstituted with 300 µL of a methanol-water solution (V/V = 1/4) and kept at -20℃ for 2 hours. Following another round of centrifugation under the same conditions, 150 µL of the resulting supernatant was filtered through a 0.22 µm organic phase pinhole filter and introduced into an LC-MS vial for subsequent analysis. A Quality Control (QC) sample was prepared by combining equal volumes of extracts from all the samples.

The analysis was conducted using an ACQUITY UPLC I-Class PLUS system (Waters) coupled with a Q Exactive mass spectrometer (Thermo Scientific) equipped with a heated electrospray ionization (ESI) source, facilitating both ESI positive and ESI negative ion modes. An ACQUITY UPLC HSS T3 column was employed for both modes.

LC-MS raw data was processed and identified using Progenesis QI software (Waters Corporation, Milford, USA), referencing public databases such as http://www.hmdb.ca/ and http://www.lipidmaps.org/, along with custom-built databases. Metabolic compound identification was rooted in databases including Human Metabolome Database (HMDB), LIpidmaps (V2.3), Metlin, EMDB, PMDB, and proprietary databases. Principal Component Analysis (PCA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) were applied to elucidate the metabolic differences between experimental groups, post mean centering (Ctr) and Pareto variance (Par) scaling. Visual representations, including heatmaps and metabolite correlations, were rendered using the ggplot2 package in R.

The transcriptome, proteome and metabolism atlases of CSF from viral and autoimmune encephalitis patients

To investigate the global transcriptomic, proteomic and metabolomic profiles of CSF cells in different types of encephalitis, we conducted a cohort of patients including 34 patients with VE, 29 patients with AE, and 6 patients with non-infectious and non-autoimmune encephalitis with clinical features similar to VE/AE. CSF samples were collected from each patient within a few days of admission. Transcriptomic, proteomic and metabolomic profiling of the CSF were performed, respectively.

Compared with the Control Group, 214 DEGs with common changing trends were identified in both VE and AE groups (Fig. 1C-D). GO pathway enrichment analysis of the DEGs revealed common immune responses in viral and autoimmune encephalitis patients, including nonspecific immune responses with inhibited neuronal transporter functions, as well as activation of T cells, promotion of T cell cytotoxic functions, and adaptive immune responses with B cell activation (Fig. 1E-F, 2A-B). Comparing the VE and AE groups, 1421 DEGs were identified (Fig. 3A). GO pathway enrichment analysis of the up-regulated DEGs in the VE group revealed that interferon responses were specifically activated in viral encephalitis compared to autoimmune encephalitis (Fig. 3B). GO pathway enrichment analysis of the up-regulated DEPs in the VE group revealed that the complement pathway was specifically upregulated in viral encephalitis compared to autoimmune encephalitis (Fig. 4B).

In metabolomic analysis, as shown in the volcano plot, 123 differential metabolites were identified and quantified between the VE group and the AE group (Fig. 3A). The heatmaps of differential metabolites and orthogonal partial least squares discriminant analysis (OPLS-DA) plot showed significant differences between VE and AE groups, and the Z-score plot also supported this view (Fig. 5B-D). The metabolite creatine associated with CD8⁺ T cell activation was upregulated in the VE group^[24]. The metabolites L-dopa and gabapentin associated with CD4⁺ T cell and Treg cell modulation were upregulated in the AE group^[25].

Both VE and AE exhibited inhibition of neuronal transporter function, activation of T cells, enhanced T cell cytotoxicity, and activation of B cells

Inhibition of neuronal transporter functions was observed in both VE and AE patients. Compared with the Control Group, a group of genes (KIF1C, KCTD3, ANKRD34A, SH3PXD2A) in AE and VE groups showed strong interactions and were down-regulated. It has been reported that KIF1C is a plus-end directed motor protein that can bind and transport the exon junction complex (EJC), participating in neuronal RNA localization and regulating neural development and synaptic plasticity^[26]. The KCTD gene family encodes potassium channel regulating proteins involved in regulating neural development and conduction, and their mutations or dysregulated expression are associated with neural developmental defects and neuropsychiatric disorders^[26]. ANKRD34A is specifically expressed in brain tissues. The downregulation of these genes may lead to inhibited neuronal transporter function.

Activation of T cells and promotion of T cell cytotoxic functions were observed in both viral and autoimmune patients. Compared with the Control Group, AE and VE groups showed differential expression of genes including BTN3A1, BTN3A2, BTN3A3, MAP4K1, ZAP70 and LCK, which were enriched in T cell activation. A group of genes (BTN3A1, BTN3A2, BTN3A3) exhibited strong interactions with each other, and they are isoforms^[27]. It has been reported that BTN3A serves as a ligand for γδTCR, regulating human γδ T cell activation and effector functions by inducing conformational changes to expose activation-related sites on Vγ9Vδ2 T cells^{[28, 29]}. In addition, the interaction between MAP4K1, ZAP70 and LCK is notable. ZAP70 and LCK are upstream kinases in the TCR signaling pathway that activate the MAPK pathway. MAP4K1, as an upstream kinase in the MAPK pathway, is a negative regulator of T cell activation. Together they lead to immunosuppression of T cells, especially CD8⁺ T cells^[30]. The downregulated DEGs CAV1, VANGL1 and THEM216 in AE and VE groups compared to the Control Group can also participate in the key regulation of T cell functions by modulating the intensity of TCR signaling. In T cells, CAV1 interacts with the TCR complex to inhibit downstream TCR signaling, thereby tuning the strength of TCR signals. CAV1 can also interact with Foxp3 to stabilize the Treg cell phenotype. Through these two mechanisms, CAV1 controls the differentiation of Treg cells into effector T cells^[31]. Its downregulation may facilitate TCR activation.

Activation of B cells is observed in both viral and autoimmune patients. Our data revealed that VE and AE Groups have differential expression of multiple genes compared with the Control Group, including CBL, PLCG1, INPP5D, PIk3CD, and CD79A genes with strong interactions, and most of them are related to the activation of B cells. Compared to the Control Group, these genes were upregulated in both the VE and AE groups^[32–34]. It is speculated that both AE and VE have enhanced B cell activation functions compared with the Control Group (Fig. 2A). It has been reported that the ubiquitin ligase activity of Cbl proteins is required for B cell antigen processing and presentation to T cells, serving as a key regulator for germinal center formation and antibody responses^[35]. PLCγ promotes cell proliferation and survival mainly by activating the mTOR signaling pathway, thus facilitating early B cell development^[36]. INPP5D plays a role in regulating microglial barrier against Aβ toxicity, thereby regulating tau pathology in an Aβ-dependent manner relevant to Alzheimer's disease^[37]. Activated PIK3CD drives intrinsic B cell clonal expansion but restricts intrinsic immune responses in B cells^[34]. CD79 is an essential component of the B cell receptor (BCR) complex, and mutations in this gene result in failure of developing B lymphocytes to express premature BCR and to develop from pre-B to naive B cell stage^[38]. The up-regulation of these genes may facilitate B cell activation. In addition, we found that CAV1, VANGL1 and THEM216, which exhibited strong interactions among the DEGs of the AE Group and VE Group compared to the Control Group, were downregulated. CAV1 can participate in the key regulation of B cell functions by modulating the intensity of BCR signaling. In B cells, CAV1 interacts with BCR to maintain BCR nanoclusters and inhibit BCR signaling by regulating Csk localization. CAV1 also stabilizes ordered lipid domains in the plasma membrane involved in tuning BCR signals. Through these mechanisms, CAV1 tunes the intensity of BCR signals to maintain B cell self-tolerance to autoantigens^[39]. Its downregulation may facilitate BCR activation. VANGL1 can form complexes with SCRIB and NOS1AP to regulate cell polarity and migration, and its expression deficiency can lead to immunoglobulin deposition and tumor progression^[40]. THEM216, also known as VANGL2, exerts its role by participating in the Wnt/PCP pathway, and can also regulate cell polarity and migration^[41]. By performing GO enrichment on the DEPs of VE and AE compared with the Control Group, it can be found that B cell mediated immunity is up regulated in both groups.

Compared with the Control Group, the expression of genes related to B cell activation was up-regulated in both the VE and AE groups, while the expression of genes related to B cell inhibition was down-regulated. DEPs enriched in B-cell immune-related pathways were up-regulated in both the VE and AE groups, suggesting B cell activation in the VE and AE patients. This may be an important factor resulting in immune imbalance in VE and AE.

Compared with AE, the interferon response, complement pathways, and CD8 ⁺ T cell functions were more specifically activated in VE

Interferon response genes were specifically upregulated in the VE Group, compared to that in the AE Group. GO pathway enrichment analysis of upregulated DEGs in VE group indicated significantly upregulated pathways related to interferon production and response (Fig. 3B). Additionally, the results of immune cell abundances, calculating by CIBERSORT, indicated increased innate immune cells in VE group compared to the AE group, with significant differences observed in M1 macrophages and neutrophils (Fig. 3D). Compared to the AE Group, the core cluster of the PPI network analysis of the upregulated DEGs in the VE Group can be enriched in the ISG15-EGFR pathway (Fig. 3C). ISG15 is an interferon (IFN) α/β-inducible ubiquitin-like intracellular protein and also acts as an IFN-γ induced secreted molecule. ISG15 modifies Rab GDP-Dissociation Inhibitor Beta (GDI2) through ISGylation, consequently altering the endocytic recycling of EGFR. This implies that ISG15 can influence the intracellular localization of EGFR, further impacting its signaling function. In this manner, ISG15 can indirectly regulate EGFR-mediated signaling, affecting cell growth and survival^[42].Moreover, the interaction of IDO1 with other DEGs is also notable. Given IDO1 is a key regulatory factor limiting excessive immune responses, it is likely playing a role in immune homeostasis in the VE group^[43]. Among the upregulated DEGs in the VE group, a group of genes (H2AC21, H2BC17, H2BC6, H2BC9, H3C12, H4C2) exhibited strong interactions, which may participate in virus replication-related deacetylation while inhibiting IFN-γ associated responses^[44] (Fig. 3C). In summary, the results indicate that in comparison to AE, the interferon response regulation in VE is specific.

Moreover, pathways involving IFN-γ suppression and adaptive immune inhibition are also activated, indicating a complex interplay among immune modulations.

The complement activation pathway was specifically activated in the VE group, when compared to the AE group. GO pathway enrichment analysis of DEPs upregulated in VE group showed activated pathways related to complement and humoral immune responses in VE group (Fig. 4B).

PPI analysis of all DEPs exhibit strong interactions of C4BPB, C1QC, C4BPA, FCN3, C1QA, and C1R, which enriched in compement activation pathway(Fig. 4A).

Therefore, we inferred that immune-mediated injury in VE patients could be attributed to complement activation.

The expression of creatine was elevated in CD8⁺ T cell function activation in the VE group. Notably, as shown in the heatmap and boxplot, after normalization using the z-score, the creatine levels in the VE group were consistently higher (P < 0.01) (Fig. 5C-E). Creatine is an essential nutrient required for normal CD8⁺ T cell function^[24]. Creatine can promote proliferation and activation of CD8⁺ T cells, and maintain secretion of cytokines and cytotoxic molecules. Creatine can increase intracellular ATP levels in CD8⁺ T cells, satisfying their rapid proliferation and immune function exertion^[45].

AE exhibited specific modulation of CD4⁺ T cell and Treg cell activity

In the process of autoimmune encephalitis, CD4⁺ T cells and Treg cells may be adaptively activated to balance the autoimmune response, thus protecting nervous system from severe damage. Based on metabolomic data, we speculate that the specific immune mechanism of AE was mainly focused on the regulation of CD4⁺ T cells and Treg cells. Levodopa and gabapentin were expressed at higher levels in AE group(Fig. 5F-G). Levodopa can be converted to dopamine, which depending on the stimulation of specific dopamine receptors (DARs) on DCs and T cells, can differentially promote CD4⁺ T cell differentiation into either Th1 or Th17 inflammatory cells^[25]. Regulatory T cells also release high levels of dopamine, which in turn can negatively feedback to inhibit T cell suppressive activity, thereby exerting immune regulation^[25]. Box plots also showed higher levels of levodopa (P < 0.001) and gabapentin (P < 0.0001) in AE group compared to VE group (Fig. 5F-G). The levels of levodopa and gabapentin were markedly higher in AE group than VE group. After standardization with z-score, the observed elevations in both (Fig. 3D) indicate CD4⁺ T cell differentiation and to some extent suppressed autoimmune responses, conferring neural protection in AE patients. Unfortunately, we did not identify any immunological pathways unique to AE from the transcriptomic and proteomic data.

VE and AE both induce damage to the cerebral cortex, basal ganglia, hippocampus, and cerebellum

We compared the CSF proteomes between VE Group and Control Group and found 26 proteins highly expressed in VE which are specifically expressed in brain. 10 brain-specific proteins are found differentially high expressed in AE Group compared to Control Group in CSF. VE Group and AE Group shared 9 brain-specific expressed proteins differentially high expressed compared to Control Group, including GFAP, NEFL, COTL1, CORO1A, SCRN1, EFHD2, KCNAB2, MERTK, NEFM (Fig. 6). Based on the distribution of these proteins, we speculate both VE and AE induce damage in cerebral cortex, basal ganglia, hippocampus, and cerebellum.

Multi-omics analysis of CSF revealed that both VE and AE exhibited inhibition of neuronal transporter function, activation of T cells, enhanced T cell cytotoxicity, and activation of B cells. Neuronal transport related genes (KIF1C, KCTD3, ANKRD34A) were decreased in both VE and AE groups, indicating inflammatory factors generated during viral infection and autoimmune responses may interfere with the expression of these genes and compromise neuronal structural and functional integrity^[44]. Additionally, in encephalitis, neuronal transport proteins can promote spreading of certain viruses and proinflammatory factors between neurons to regulate inflammatory responses. They can also provide neuroprotection by controlling molecular concentrations in the neuronal microenvironment, exerting roles in neural repair^[46]. Downregulation of neuronal transport proteins may represent an inhibitory regulatory response of the immune system. Activation of γδT cells related genes (BTN3A1, BTN3A2, BTN3A3) were concurrently upregulated in both VE and AE. We speculated that in VE, innate immune cells including γδT cells, dendritic cells, and monocytes can be activated at the early stage of infection, rapidly recruited to infection sites, directly killing infected cells while secreting cytokines like IL-17, IFN-γ. IFN-γ can further upregulate BTN3A1 on monocytes via positive feedback, amplifying γδ T cell activation^[47], and promoting continued release of proinflammatory factors that participate in neuroinflammation of viral encephalitis to facilitate viral clearance^[48]. In AE, γδT cells and pDC cells expressing MHC-II can capture and present self-antigens upon brain tissue damage, while releasing large amounts of proinflammatory factors like IFN-γ, IL-17, and amplifying γδT cell activation via the BTN3A1 related mechanisms. These proinflammatory factors can also stimulate B cell proliferation and antibody production, promoting autoantibody formation^[49]. Meanwhile, aggregation of γδT cells may enhance blood-brain barrier permeability, allowing more autoreactive lymphocytes to enter brain tissues, thus playing an important role in AE^[50]. Compared to the Control Group, both VE and AE Groups exhibited upregulation of genes associated with BCR activation (CBL, PLCG1, INPP5D, PIk3CD, CD79A) as well as downregulation of genes regulating BCR signaling (CAV1, VANGL1 and THEM216). This indicates B cell activation may have played an important immune role in both AE and VE. Many studies have reported that AE can follow VE^[51], the latter called biphasic encephalitis induced by viral infection. Through multi-omics studies of cerebrospinal fluid from patients with VE and AE, we discovered activation of both T cells and B cells in VE and AE. Therefore, we speculate that viruses directly infect and damage brain tissues, eliciting inflammatory responses, while at the same time, innate and adaptive immunity are activated. Activation and proliferation of T cells directly attack neuronal cells, and activation of B cells leads to production of autoantibodies and inflammatory factors, thereby triggering autoimmune reactions that mediate secondary damage to the nervous system.

The unique immune mechanisms of VE and AE may provide insights into their precise diagnosis and treatment. For VE, interferon stimulants, macrophage and neutrophil cell therapies, IDO1 agonists etc. can be used to enhance antiviral immunity. T cell function can also be improved by inhibiting MAP4K1 or modulating creatine metabolism pathways^[52]. For AE, levodopa inhibitors or gabapentin agonists could be considered to control overactive autoimmunity. Neuroprotective therapies targeting neuronal transport proteins^[53]could also be applied, as combining immunotherapies with neuroprotection will aid prognosis in encephalitis patients^[54]. Additionally, the shared activation of immunosuppressive pathways in VE and AE may serve as common therapeutic targets to restore immune function. But risks of overactive immunity must also bewared. In summary, this study has provided several potential immunotherapeutic targets for VE and AE. However, further research is still needed to elucidate the precise mechanisms of the various targets, and strike a balance between augmenting therapeutic effects and minimizing adverse reactions. And personalized immunotherapies should be designed based on individual immune profiles. This will help optimize immunotherapy for VE and AE, enabling precision medicine.

We found that immune processes involving both promotion and inhibition of immune responses were shared and distinct between VE and AE groups. Immune inhibitory pathways were activated in both VE and AE, which may be negative feedback regulation against hyperactive immune responses to maintain immune homeostasis. The reduced expression of neuronal transporters in both VE and AE may be a protective mechanism of the immune system to control inflammation. The shared activation of immunosuppressive pathways, such as negative regulation of T and B cells, may be feedback inhibition against excessive immune response. Interestingly, the significant activation of innate immunity in VE, particularly increased neutrophil and macrophage infiltration, promoted rapid viral clearance but may also result in a "cytokine storm." Thus, negative regulators like IDO1 were upregulated to suppress proinflammatory factors like interferons to maintain immune homeostasis. In AE, activation of autoimmunity-related pathways was more prominent, with increased expression of levodopa and GABA, which may be negative feedback by the host through dopamine to inhibit T cells and GABA to inhibit inflammation against overactive autoimmunity. In summary, both VE and AE exhibit a dynamic balance between immune activation and inhibition, but through differential regulatory mechanisms that may relate to immune homeostasis^[55]. Immune equilibrium is critical for the pathogenesis and prognosis of encephalitis^[56]. Further elucidation of the roles of different pathways in regulating immune equilibrium will facilitate optimization of therapeutic strategies to achieve a balance between pathogen elimination and prevention of immune-mediated damage. However, such regulation may also lead to insufficient immunity against pathogens. Further studies are needed to elucidate the roles and impacts of these immunosuppressive pathways in the pathogenesis of encephalitis.

As a tertiary hospital, our hospital has more standardized diagnosis and treatment capabilities than other hospitals. Therefore, this study may have included more moderate to severe encephalitis patients. These patients often have medical histories at other hospitals and may have empirically used hormones, antibiotics, antivirals etc., which could have impacted the results^[57].

In summary, this study conducted systematic transcriptomic, proteomic and metabolomic analyses on cerebrospinal fluid samples from multiple patients with VE, AE and Control Groups, revealing shared and unique molecular signatures between VE and AE. Both VE and AE exhibited inhibition of neuronal transporter function, and activation of T cells and B cells. Compared to AE, VE may induce a more robust interferon response, complement activation and CD8⁺ T cell activation. AE demonstrated specific modulation of CD4⁺ T cell and Treg cell activity. Our data provides a landscape view of molecular alterations in cerebrospinal fluid of patients with VE, AE and Control Groups, which may offer useful diagnostic clues for VE and AE, and lays the groundwork for developing targeted therapeutic strategies.

Ethical Approval and Consent to participate. The study protocol and informed consent forms have been approved by the Huashan Hospital Ethical committee (protocol ID: 2021-KY451). All participants signed an informed consent form. All methods are executed in accordance with the Declaration of Helsinki.

Consent for publication. Not applicable

Availability of data and materials. Not applicable.

Competing interest. All authors have no competing interest.

Funding. This work was supported by National Natural Science Foundation of China(82002141), Shanghai Science and Technology Committee (21NL2600100, 20dz2260100) and Key Discipline Construction Plan from Shanghai Municipal Health Commission (GWV-10.1-XK01).

Authors' contributions. KZ, KL, QLX wrote the main manuscript and prepared figures. HCZ, YZ, JXG contributed to the data analysis. JWA and WHZ supervised the study and revised the manuscript.

Acknowledgements. We acknowledge Vision Medicals for Transcriptomics data.

Tyler KL. Acute Viral Encephalitis. N Engl J Med. 2018;379(6):557–66.
Wang H, et al. Global magnitude of encephalitis burden and its evolving pattern over the past 30 years. J Infect. 2022;84(6):777–87.
Granerod J, et al. Causes of encephalitis and differences in their clinical presentations in England: a multicentre, population-based prospective study. Lancet Infect Dis. 2010;10(12):835–44.
Glaser CA, et al. Beyond viruses: clinical profiles and etiologies associated with encephalitis. Clin Infect Dis. 2006;43(12):1565–77.
Mailles A, Stahl JP. Infectious encephalitis in france in 2007: a national prospective study. Clin Infect Dis. 2009;49(12):1838–47.
Glaser CA, et al. In search of encephalitis etiologies: diagnostic challenges in the California Encephalitis Project, 1998–2000. Clin Infect Dis. 2003;36(6):731–42.
Dalmau J, Graus F. Antibody-Mediated Encephalitis. N Engl J Med. 2018;378(9):840–51.
Dalmau J, Geis C, Graus F. Autoantibodies to Synaptic Receptors and Neuronal Cell Surface Proteins in Autoimmune Diseases of the Central Nervous System. Physiol Rev. 2017;97(2):839–87.
Armangue T, Leypoldt F, Dalmau J. Autoimmune encephalitis as differential diagnosis of infectious encephalitis. Curr Opin Neurol. 2014;27(3):361–8.
Gable MS, et al. Anti-NMDA receptor encephalitis: report of ten cases and comparison with viral encephalitis. Eur J Clin Microbiol Infect Dis. 2009;28(12):1421–9.
Gable MS, et al. The frequency of autoimmune N-methyl-D-aspartate receptor encephalitis surpasses that of individual viral etiologies in young individuals enrolled in the California Encephalitis Project. Clin Infect Dis. 2012;54(7):899–904.
Harry GJ. Microglia during development and aging. Pharmacol Ther. 2013;139(3):313–26.
Chen S, et al. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90.
Mortazavi A, et al. Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 2008;5(7):621–8.
Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30.
Yu G, et al. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics. 2012;16(5):284–7.
Bhattacharya S, et al. ImmPort: disseminating data to the public for the future of immunology. Immunol Res. 2014;58(2–3):234–9.
Wu X, et al. Intrinsic Immunity Shapes Viral Resistance of Stem Cells. Cell. 2018;172(3):423–e43825.
Newman AM, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453–7.
Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.
von Mering C, et al. STRING: a database of predicted functional associations between proteins. Nucleic Acids Res. 2003;31(1):258–61.
Shannon P, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504.
Zhou Y, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523.
Samborska B, et al. Creatine transport and creatine kinase activity is required for CD8(+) T cell immunity. Cell Rep. 2022;38(9):110446.
Pacheco R, Contreras F, Zouali M. The dopaminergic system in autoimmune diseases. Front Immunol. 2014;5:117.
Nagel M, et al. The kinesin motor KIF1C is a putative transporter of the exon junction complex in neuronal cells. RNA. 2022;29(1):55–68.
Palakodeti A, et al. The molecular basis for modulation of human Vγ9Vδ2 T cell responses by CD277/butyrophilin-3 (BTN3A)-specific antibodies. J Biol Chem. 2012;287(39):32780–90.
Vavassori S, et al. Butyrophilin 3A1 binds phosphorylated antigens and stimulates human γδ T cells. Nat Immunol. 2013;14(9):908–16.
Blazquez JL, et al. New Insights Into the Regulation of γδ T Cells by BTN3A and Other BTN/BTNL in Tumor Immunity. Front Immunol. 2018;9:1601.
Chuang HC, Tan TH. MAP4K Family Kinases and DUSP Family Phosphatases in T-Cell Signaling and Systemic Lupus Erythematosus. Cells, 2019. 8(11).
Schönle A, et al. Caveolin-1 regulates TCR signal strength and regulatory T-cell differentiation into alloreactive T cells. Blood. 2016;127(15):1930–9.
Sohn HW, Gu H, Pierce SK. Cbl-b negatively regulates B cell antigen receptor signaling in mature B cells through ubiquitination of the tyrosine kinase Syk. J Exp Med. 2003;197(11):1511–24.
Yu M, et al. PLCγ-dependent mTOR signalling controls IL-7-mediated early B cell development. Nat Commun. 2017;8(1):1457.
Wray-Dutra MN, et al. Activated PIK3CD drives innate B cell expansion yet limits B cell-intrinsic immune responses. J Exp Med. 2018;215(10):2485–96.
Linterman MA. Cbls boost B cells. J Exp Med, 2020. 217(9).
Yu M, et al. PLCγ-dependent mTOR signalling controls IL-7-mediated early B cell development. Nat Commun. 2017;8(1):1457.
Iguchi A, et al. INPP5D modulates TREM2 loss-of-function phenotypes in a β-amyloidosis mouse model. iScience. 2023;26(4):106375.
Minegishi Y, et al. Mutations in Igalpha (CD79a) result in a complete block in B-cell development. J Clin Invest. 1999;104(8):1115–21.
Minguet S, et al. Caveolin-1-dependent nanoscale organization of the BCR regulates B cell tolerance. Nat Immunol. 2017;18(10):1150–9.
Anastas JN, et al. A protein complex of SCRIB, NOS1AP and VANGL1 regulates cell polarity and migration, and is associated with breast cancer progression. Oncogene. 2012;31(32):3696–708.
Shafer B, et al. Vangl2 promotes Wnt/planar cell polarity-like signaling by antagonizing Dvl1-mediated feedback inhibition in growth cone guidance. Dev Cell. 2011;20(2):177–91.
Bogunovic D, et al. Mycobacterial disease and impaired IFN-γ immunity in humans with inherited ISG15 deficiency. Science. 2012;337(6102):1684–8.
King NJ, Thomas SR. Molecules in focus: indoleamine 2,3-dioxygenase. Int J Biochem Cell Biol. 2007;39(12):2167–72.
Jean Beltran PM, Cristea IM. The life cycle and pathogenesis of human cytomegalovirus infection: lessons from proteomics. Expert Rev Proteom. 2014;11(6):697–711.
Di Biase S, et al. Creatine uptake regulates CD8 T cell antitumor immunity. J Exp Med. 2019;216(12):2869–82.
Gasnier B. The loading of neurotransmitters into synaptic vesicles. Biochimie. 2000;82(4):327–37.
Miyagawa F, et al. Essential requirement of antigen presentation by monocyte lineage cells for the activation of primary human gamma delta T cells by aminobisphosphonate antigen. J Immunol. 2001;166(9):5508–14.
Déchanet J, et al. Human gammadelta T cells and viruses. Microbes Infect. 1999;1(3):213–7.
Chien YH, Meyer C, Bonneville M. γδ T cells: first line of defense and beyond. Annu Rev Immunol. 2014;32:121–55.
Shiromizu CM, Jancic CC. γδ T Lymphocytes: An Effector Cell in Autoimmunity and Infection. Front Immunol. 2018;9:2389.
Esposito S, et al. Autoimmune encephalitis after herpes simplex encephalitis: A still undefined condition. Autoimmun Rev. 2022;21(12):103187.
Song X, Qian Y. The activation and regulation of IL-17 receptor mediated signaling. Cytokine. 2013;62(2):175–82.
Rudy CC, et al. The role of the tripartite glutamatergic synapse in the pathophysiology of Alzheimer's disease. Aging Dis. 2015;6(2):131–48.
Chitnis T, et al. Consensus statement: evaluation of new and existing therapeutics for pediatric multiple sclerosis. Mult Scler. 2012;18(1):116–27.
Koyuncu OO, Hogue IB, Enquist LW. Virus infections in the nervous system. Cell Host Microbe. 2013;13(4):379–93.
Lebon P, et al. Intrathecal synthesis of interferon-alpha in infants with progressive familial encephalopathy. J Neurol Sci. 1988;84(2–3):201–8.
Elenkov IJ. Glucocorticoids and the Th1/Th2 balance. Ann N Y Acad Sci. 2004;1024:138–46.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
21 Aug, 2024
Editor assigned by journal
19 Aug, 2024
Submission checks completed at journal
19 Aug, 2024
First submitted to journal
19 Aug, 2024

You are reading this latest preprint version

A multi-omics analysis of CSF reveals the differential immunopathology molecular mechanisms underlying viral encephalitis and autoimmune encephalitis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Patient enrollment and sample collection

Total RNA extraction and meta-transcriptome sequencing

Meta-transcriptomic data analysis

Proteome analysis

Sample Depletion, Digestion and TMT labeling

High pH Fractionation

MS analysis

Database search and statistical analysis

Metabolic sample preparation and data analysis

Results

The transcriptome, proteome and metabolism atlases of CSF from viral and autoimmune encephalitis patients

AE exhibited specific modulation of CD4⁺ T cell and Treg cell activity

VE and AE both induce damage to the cerebral cortex, basal ganglia, hippocampus, and cerebellum

Discussion

Declarations

References

Additional Declarations

Status:

Version 1

A multi-omics analysis of CSF reveals the differential immunopathology molecular mechanisms underlying viral encephalitis and autoimmune encephalitis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Patient enrollment and sample collection

Total RNA extraction and meta-transcriptome sequencing

Meta-transcriptomic data analysis

Proteome analysis

Sample Depletion, Digestion and TMT labeling

High pH Fractionation

MS analysis

Database search and statistical analysis

Metabolic sample preparation and data analysis

Results

The transcriptome, proteome and metabolism atlases of CSF from viral and autoimmune encephalitis patients

AE exhibited specific modulation of CD4+ T cell and Treg cell activity

VE and AE both induce damage to the cerebral cortex, basal ganglia, hippocampus, and cerebellum

Discussion

Declarations

References

Additional Declarations

Status:

Version 1

AE exhibited specific modulation of CD4⁺ T cell and Treg cell activity