The authors declare that all supporting data are available within this article and its Data Supplement. The RNA-seq data and the list of differentially expressed genes (DEGs) can be viewed at the Gene Expression Omnibus (GEO) under accession number GSE167110 (29). Scripts and code are uploaded to [Regnier-Golanov AS, (29)]. Other datasets generated for this study are available upon request to the corresponding author.

Animal experimental procedures were conducted in accord with the U.S. National Institutes of Health “Guide for the care and use of laboratory animals” and in compliance with the ARRIVE guidelines. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Houston Methodist Research Institute, Houston, TX (Protocol #: AUP-0220-0009). Animals were housed in the institutional animal facilities on a 12 h day/night cycle with ad libitum access to food and water.

Studies were performed in 11 male C56BL/6J mice, 10–14 weeks old (Jackson Labs). To minimize the variability of the RNA-seq experiments, the present study was limited to the male animals. Animals from the same batch were randomly assigned to one of three groups: SAH (n = 4), Sham (n = 3), and Naïve (n = 3). Experimental SAH was induced by the unilateral perforation of the Circle of Willis as previously described (30). Anesthesia was induced with 3% isoflurane in air and maintained at 1.5–1.75% isoflurane in 80% N₂/20% O₂. For regional cerebral blood flow (rCBF) measurement, needle probe (1.5 mm diameter, Moor Instruments, Wilmington, DE) fixed to the stereotaxic holder was placed over the intact skull over the parietal cortex through the appropriately prepped (alcohol and chlorhexidine 2%) skin cut on the side of perforation, after animal anesthetization and placement in stereotaxic frame. Drop of mineral oil was applied to the skull for optical contact between the probe and the skull. Animal placed on a homeothermic blanket with the body temperature maintained at 37°C by the rectal probe feedback (Harvard apparatus) was rotated in a supine position. The heart rate was continuously monitored (LabChart, AD Instruments). Neck skin was shaved and prepped with alcohol and chlorhexidine (2%). In aseptic conditions through the midline neck incision, common, external, and internal carotid arteries were exposed. All three arteries were mobilized using silk thread (7–0). The external carotid artery was cut between two ligatures, and a stump was formed. To minimize the bleeding and avoid mortality, 6-0 prolene filament (length 12 mm) was advanced through the stump into the internal carotid and moved forward for ~0.5 mm beyond the point of slight resistance. The filament was withdrawn and stump ligated, and the wound was closed. In sham-operated mice, the filament was advanced to the resistance point and withdrawn without puncture (30). Wounds were closed, and animals were allowed to recover before returning to the home cage. Following the surgery, mice were monitored daily for 96 h. We experienced no procedural or post-surgery mortality when performing experiments in this study.

Only SAH mice showing a rCBF drop by ≥50% (average 80 ± 15%, P < 0.001) immediately after perforation were selected for the study. A representative photograph of an SAH experimental animal done for the study is shown (Supplementary Figure 5).

Neurological assessment was done with the Garcia scale as previously reported (31) daily. No significant difference was observed in the baseline scores for the animals randomly assigned to either group [H₍₃₎ = 3.6133, P = 0.1581, n = 3–4/group] at 24–96 h after surgery (Supplementary Table 2). Due to the limited number of animals, Garcia scale, initially developed for stroke neurological consequences evaluation (31), was not sufficiently sensitive to detect differences of neurological outcomes between Sham and SAH in our experiments, but to rather detect effects of surgery. Further neurobehavioral studies (N = 18 with n = 10 for SAH and n = 8 for Sham) in our laboratory detected no neurological impairments with the Garcia scale but instead showed significant neurocognitive deficits at 4 days after SAH with Y-maze, Social interaction, and Open-field test (in preparation).

Four days following surgery, mice were euthanized with CO₂ followed by thoracotomy and decapitation. Immediately after bilateral whole hippocampus of SAH (n = 4), Sham (n = 3), and Naïve (n = 3) group was manually collected by an experimenter blinded to the specific group and total RNA was extracted immediately after using QIAzol reagent (RNeasy Mini Kit, Qiagen Inc., Hilden, Germany). RNA was quantified using the Nanodrop 1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). RNA-seq was performed by RNA core of Weill Cornell Medicine (New York, NY). Total RNA integrity was evaluated with the Agilent Bioanalyzer 2100 (Agilent Biotechnologies, Santa Clara, CA), and only samples with an RNA integrity number (RIN) >9.0 were further processed for RNA-seq. All RNA-seq libraries were prepared using 100 ng of total RNA with the TruSeq Stranded Total RNA sample preparation kit (ribosomal RNA depletion and stranded RNA-Seq) with Ribo-Zero according to the manufacturer's specified protocol (Illumina, San Diego, CA). Samples were multiplexed and sequenced across multiple lanes of a HiSeq 4000 Sequencing System (Illumina) using 50-base paired-end reads to achieve a minimum depth of 30 million aligned read-pairs, which is sufficient to evaluate the similarity between transcriptional profiles according to ENCODE guidelines (32).

Before differential gene expression analysis, the quality of the sequences was assessed based on several metrics using FastQC and QoRTs (33, 34). All our samples passed the quality control (Supplementary Figures 1A–F). We aligned, with default parameters, the RNA-seq reads to the mouse reference genome (mm9) using STAR (35). Gene coverage estimates were obtained with featureCounts using composite gene models (union of the exons of all transcript isoforms per gene) from UCSC mm9 annotation from Illumina's iGenomes (36, 37). Uniquely mapped reads that unambiguously overlapped with no more than one Gencode composite gene model were counted for that gene model; the remaining reads were discarded. The counts for each gene model were used for the subsequent analyses.

Hippocampal tissue and RNA extraction were obtained identically to RNA-seq experiments. Male C56BL/6J mice, 10–14 weeks old (N = 18 with n = 9 for SAH, n = 9 for Sham), were used for this set of experiments. A preamplification step was simultaneously done with the reversed transcription (RT) step using the iScript Explore One-step kit (Biorad, Hercules, CA). Briefly, 1 μg of DNase-treated total RNA was reverse transcribed for 60 min at 45°C, RT was inactivated by increasing temperature to 95°C for 3 min, and 14 preamplification steps were done with the following thermocycler protocol: 95°C for 15 s and 58°C for 4 min, hold at 4°C. Primers (Supplementary Table 1) (Sigma-Aldrich, St. Louis, MO), designed with NCBI tool (53), and 2.5 μl of a custom preamplification assay pool of the reverse and forward primers, diluted at 200 μM, were added to the reaction. The preamplification samples were diluted 10-fold in TE buffer (Invitrogen, Carlsbad, CA). Quantitative PCR was run using 5 μl of diluted cDNA and SYBR green chemistry with the SsoAdvanced Universal SYBR green supermix on a CFX Connect thermocycler (Biorad, Hercules, CA) and a cycling program of 30 s at 95°C, and 40 cycles of 95°C for 15 s and 60°C for 30 s. Primers were used at a 300 nM final concentration in a 20 μl total volume. After completion of qPCR, a melting curve of amplified products was determined. Pgk1 was selected as our reference gene after comparing several common housekeeping genes (54) and our RNA-seq list of genes showing no difference in expression between the control and SAH group. In short, the mean of the threshold cycle (Ct) of three technical replicates was used as the Ct for the target of a given animal ΔCt(Animal) = Ct(Target)–Ct(Pgk1), and fold changes were calculated using the following equation: 2^−ΔΔCt = 2^{−[MeanΔCt(SAH)−MeanΔCt(Sham)]} (55).

Male C56BL/6J mice, 10–14 weeks old (N = 7 with n = 3 for SAH, n = 4 for Sham), were used for this set of experiments. Mouse bilateral whole hippocampus was homogenized in ice-cold RIPA buffer [25 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1X HALT Protease and Phosphatase Inhibitor Cocktail Solutions (ThermoFisher Scientific)], centrifuged, and stored at −80°C until use. Proteins were quantified by Pierce BCA Protein Assay Kit (ThermoFisher Scientific) and resolved by high resolution Nupage 4–12% Bis-Tris Gel (Thermo Fisher Scientific-Invitrogen). Anti-rabbit MOG (1:1,000; #12690-1-AP, Proteintech) was incubated overnight and diluted in blocking buffer (5% dried milk in Tris-buffered saline, pH 7.4, and 0.2% Tween-20, TBST). After no-stripping of the membrane, anti-rabbit GAPDH linked to HRP (1:2,000; #8884, Cell Signaling Technology) was incubated on the same membrane following MOG labeling overnight and diluted in blocking buffer. Detection was done with Pierce ECL detection reagent (ThermoFisher Scientific). Protein bands were detected with X-ray film (GenHunter.com), and densitometry was calculated after scanning the films at 600 dpi using ImageJ software and the Miller's tutorial (56). Results were expressed as percentage of change from Sham.

Graphs were plotted and statistical analysis was done with GraphPad Prism 9 (GraphPad Software, San Diego, CA). Shapiro–Wilk test was used for test of normality, and F-test for heterogeneity of the variance. Parametric data are represented as mean ± SD, and nonparametric as median with interquartile range. Statistical significance was set at P < 0.05. Pairwise comparisons were analyzed using a two-tailed t-test. For the Garcia scale, a Kruskal–Wallis test was done. Statistics are reported in the text and figure legend.

Validation of RNAseq findings was performed with mRNA expression analysis using RtqPCR. We found 10 overexpressed DEGs genes statistically significant in a different cohort of animals (Figure 1). To explore active ongoing processes in the hippocampus, we analyzed DEGs using REACTOME pathways and GO-biological processes (GO-BP) analyses and searched for clusters of interrelated processes. The five top interacting clusters identified, using both approaches, revealed overexpression of groups of genes related to immune and inflammatory processes and extracellular matrix reorganization (Figures 2, 3). Additional exploration using Ingenuity Pathway Analysis canonical pathways (IPA, QIAGEN, 2020) and GSEA Hallmark collection confirmed the prevalence of these processes (Supplementary Figure 3). Five top upregulated genes included Ccl5, Lcn2, Plin4, Timp1, and Gpr65 related to inflammation (Figure 4; Table 1).

The top five downregulated interacting clusters of GO-BP related to myelin ensheathment and oligodendrocyte activity (Figure 3C). Significantly decreased was the expression of genes specific to the myelin structure: Pllp, Mbp, Mal, Mag, Cntnap1, Plp1, Mog, and Mobp as well as the oligodendrocytes signature genes: Olig1, Sox8, Sox10, and Nkx6.2 (Supplementary Table 3). Decreased levels of the mRNA of Mog, Mag, and Rnf122, oligodendrocytes-related genes by Rt-qPCR and decrease in MOG protein expression by Western blot (Figures 5A,B) confirmed the decrease of oligodendrocyte activity.

Over-expression of DEGs of GO-BP suggested activation of genes related to innate and adaptive immunity (Figure 3A). In addition, IPA and GSEA (Supplementary Figures 3A,B) processes demonstrated significant activation of the complement system with activation of the genes coding for the key elements of the complement system C1, C2, C3, C4, and Cfb and complement system regulators Cfh and C1-inh (Serping1) (Supplementary Table 4). Components of complement receptors CR3 (Cd11b+CD18) and CR4 (Cd11c+CD18), integrins alpha Itga1, 5 and integrins beta Itgb1, 2 (CD18), and 5 were also overexpressed. Expression of the gene comprising “membrane attack complex” (C5-C9) did not change significantly.

Increased expression of genes belonging to IPA canonical pathways “Dendritic cell maturation” and “Th2 pathway” suggested activation of adaptive immunity processes (Supplementary Figure 3A). These pathways share class I major histocompatibility complex (MHCI)-related genes, H2-D1, H2-K1, and β2m (Supplementary Table 5) with the “Immunoregulatory interactions between a Lymphoid and a non-Lymphoid cell” REACTOME (Figures 2A,B) along with “Antigen presentation: folding, assembly, and peptide loading of class I MHC” pathways (Figure 2A). Genes associated with major histocompatibility II (MHCII) Cd74, H2-Aa, H2-Ab1, and H2-Eb1 were also upregulated as well as the costimulatory molecule gene CD86 (Supplementary Table 6). Chemokine genes involved in macrophage/monocyte recruitment Ccl5 (4-log₂-fold increase; Figure 4; Table 1), Ccr2 (2.5-log₂-fold increase; Figure 4; Table 1), and its murine ligand Ccl12 as well as Cxcl10 (Supplementary Table 7) were also increased. Toll-like receptors (TLR) 2 (0.91-log₂-fold increase, P < 0.01), 8 (1.67-log₂-fold increase, P < 0.06), and 13 (0.83-log₂-fold increase, P < 0.003) did significantly increase 4 days after SAH.

Our data showed upregulation of microglial markers of M1-like pro-inflammatory phenotype, particularly microglial signature genes of phagocytotic processes Calr, Itgb2, Cd68, Cxcl16, Scara5, and scavenging receptor gene Scarf1, along with IgG Fc receptors Fcgr2b (CD32), Fcgr4 (CD16), lectin Lgals3, and Trem2 and its ligand Tyrobp (Supplementary Table 8). M2-like anti-inflammatory phenotype Il14ra, Il13ra1, and Il10rb and M0-resting-state microglia along with upregulation of the growth factors supporting damage repairment and tissue remodeling Cxcl10, Igf1, Tgfbr2, Tgfbr3, Pdgfd, and Mmp2 (Supplementary Table 8) were also increased.

IPA upstream regulator analysis (URA) showed activation of five top gene sets regulated by IFNG, IRF7, IRF3, STAT1, and CHUK (Supplementary Table 9A). Also, among significantly activated sets of genes were genes under upstream regulators TNF-α, IL1-β, TLR4, and IFNAR1. Supporting the involvement of type I and II interferons' pathways at 4 days after SAH were the overexpression of interferon-related DEGs Irf7, Irf1, and Irf8 as well as upregulation of Ifit1, Usp18, Isg15, Ifih1, and Unc93b1(Supplementary Table 10) downstream of interferon regulatory genes.

The URA also identified TGFβ1 as a potential activated upstream regulator (Supplementary Table 9A). Supporting TGFβ pathway engagement 4 days after SAH are the overexpression of bone morphogenetic proteins Bmp5, Bmp6, and Bmp7 in DEGs (Supplementary Table 11), Smad6 and Smad7, inhibitory Smads in the TGFβ intracellular pathway, as well as Smad3. Also increased in DEGs were the TGFβ receptors Tgfbr3 and Tgfbr2 (Supplementary Table 11).

Transcription factor binding motif prediction and enrichment analyses (AME, MEME-suite) confirmed URA identification of interferons as upstream regulators, showing enrichment of IRF7 binding motifs and enrichment of STAT1/STAT2 along with binding motif of SPIB (Spi-B Transcription Factor) (Supplementary Figure 4A).

Analysis of enriched motifs independently of their location within the promoter (CENTRIMO, MEME-suite) suggested enrichment of binding sites for the moderately expressed Krüppel-like factor 3 (KLF3) and SP1 transcription factors (Supplementary Figures 3B,C).