We aimed at identifying early cerebral vascular responses to ischemic stroke in female rats and identifying potential new therapeutic targets. We also aimed at comparing the differential gene expression patterns observed in MCAs from female rats with those seen in males. The study was designed as a case-control study with the occluded and non-occluded MCAs (after tMCAO) as case groups and the non-stroke MCAs (non-operated) as control group. In this study, the experimental unit was individual animal, only animals surviving the operation until the 3 hours’ time point were included and the confounding factor sex was controlled for.
12-week old female (n = 36) and male (n = 24) Wistar rats were obtained from Charles River (Charles River laboratories, Sulzfeld, Germany) were used. Animals were housed in a 12 hour light/dark cycle (lights on 7am-7pm), climate- and humidity- controlled environment with free access to food and water in the University of Lund animal facility. All experimental procedures were performed in accordance with the ARRIVE guidelines, the ethical guidelines of the International Association for the Study of Pain regulations on animal welfare and the National Institutes of Health guidelines for the care and use of laboratory animals. The experimental procedures have been previously approved by the Institutional Animal Care and Use Committee of the University of Lund. The rats were randomly divided into control and tMCAO experiments. The female rats were monitored daily with vaginal smears for a minimum of two consecutive cycles prior to sacrifice to determine the estrous phase. Samples were collected with a saline soaked cotton swab, transferred onto microscope slides, air-dried and stained with hematoxylin-eosin. The samples were collected at the same time-point each day. Animals that were in the high estrogen phase, proestrus, were not included to reduce variation due to hormone fluctuations. The methodology has been further described by Goldman et al. .
Transient middle cerebral artery occlusion
MCA occlusion was performed by using the intraluminal filament technique . Anesthesia was induced by 3.5% isoflurane in N20:02 (70:30) and maintained by continuous inhalation of 2.5% isoflurane in N20:02 (70:30). An arterial tail catheter was inserted to monitor blood gases and mean arterial blood pressure during the occlusion. A rectal thermometer connected to a homoeothermic blanket was used to maintain body temperature at 37°C during the procedure. A laser Doppler probe (Perimed, Järfälla, Sweden) was fixed to the skull 6 mm laterally of the midline and 1 mm posterior to bregma. An incision was made in midline of the neck, exposing the carotid artery. The external carotid and common carotid were permanently ligated with sutures. A rubber-coated monofilament (Doccol Corporation, Redlands, CA, USA) was inserted through an incision in the common carotid artery through the internal carotid artery until a sudden drop in cerebral blood flow was observed in the area supplied by the MCA (as measured with Laser Doppler flowmetry). Animals with a minimum flow reduction of 50% and subsequent flow increase of 30% were included in the study. The filament was secured by sutures and the surgical areas closed while the anesthesia was discontinued, and the animal allowed recovering. The method has been described in details before [21, 22].
Prior to reperfusion, the rats were evaluated with the 6-point neuroscore [39, 40]. After 120 minutes, the animals were re-anesthetized and the filament removed, which resulted in an increase in cerebral blood flow (as measured with Laser Doppler flowmetry). The animals were then allowed to recover with free access to food and water for 3 hours.
Middle cerebral artery isolation
Three hours post reperfusion; the animals were euthanized with carbon dioxide and decapitation. The brain was immediately removed and both the right and left MCA (distal part; length = 5 mm, diameter 0.2 mm) were carefully dissected out. The arteries were carefully cleaned from surrounding connective tissue and blood, frozen on dry ice and stored in -80°C. Inclusion criteria are successful tMCAO operation as judged from the laser Doppler analysis of rCBF. All animals fulfilling this survived the 3 h until sacrifice. The subsequent handling and analysis of the removed MCA segments were blinded and tissue and groups unknown to the analyst.
The extraction of RNA was performed with the same method for both the microarray and PCR. The RNA was isolated with the Nucleospin miRNA isolation kit (Machery-Nagel, Düren, Germany), following the manufactures instructions for extraction of total RNA. The artery-samples were first homogenized in in Lysing matrix D tubes containing 1.4 mm ceramic spheres (MP Biomedicals, CA, USA) and lysis buffer (ML buffer) from the NucleoSpin kit on dry ice in a FastPrep-24Ô 5G instrument (MP Biomedicals, USA) with 3x20sec cycles.
After RNA extraction, the amount of RNA was quantified using a NanoDrop 2000 UV-Vis spectrophotometer (ThermoFisher Scientific, MA, USA). A ratio of sample absorbance at 260 nm and 280 nm in the range of 1.7 to 2.1 was accepted. The RNA which was to be used for microarray analysis was concentrated with a Scan Speed 32 speed vacuum concentrator (Labogene, Denmark). The concentration and quality of the concentrated RNA was determined with a NanoDrop ND1000 spectrophotometer (ThermoFisher Scientific, MA, USA).
Whole genome microarray
Affymetrix whole-transcriptome expression profiling was processed by Swegene centre for integrative biology (SCIBLU) genomics, Affymetrix unit at Lund University, Sweden.
The integrity of the RNA was measured with the Agilent Bioanalyzer (Agilent Technologies, CA, USA). Three samples were excluded at this stage due to poor RNA integrity values. From a total of 100 ng RNA, single stranded complimentary DNA (cDNA) was synthesized using primers containing a T7 promoter sequence. The single stranded cDNA was converted to double stranded DNA and used as a template for in vitro transcription, producing complimentary RNA (cRNA) . At this step, one sample was excluded due to poor amplification. After purification, sense-strand cDNA was synthesized and purified. The sense strand cDNA was fragmented and labeled and loaded onto Affymetrix GeneChip rat gene 2.0 ST arrays (ThermoFisher Scientific, MA, USA). This was followed by hybridization for 16 h at 45 °C in an Affymetrix Gene Chip Hybridization 645 oven. The array was scanned using the Affymetrix GeneChip scanner 3000 7G.
Complimentary DNA synthesis and quantitative PCR
After RNA extraction, the amount of RNA was quantified using a NanoDrop 2000 UV-Vis spectrophotometer (ThermoFisher Scientific, MA, USA). Five samples were excluded at this stage due to low RNA-concentration. cDNA was synthesized using the RT2 First Strand Kit (Qiagen, Germany) according to the manufacturer’s protocol.
The QuantStudio 12 K Flex real-time PCR system (ThermoFisher Scientific, MA, USA) was used for the qPCR. Taqman gene expression assays for Ccl2 (Rn00580555_m1), Olr1 (Rn00591116_m1), Adamts4 (Rn02103282_s1), Serpine1 (Rn01481341_m1), S1pr3 (Rn01757498_m1), Socs3 (Rn01470502_g1), JunB (Rn00572994_s1) and Fosl1 (Rn00564119_m1) were purchased from ThermoFisher Scientific, MA, USA.
The qPCR was performed in a 10-µl reaction volume containing TaqMan 2× universal PCR master mix (ThermoFisher Scientific, MA, USA), 20× TaqMan gene expression assay, RNase-free water and 2 µl cDNA using the QuantStudio 12 K Flex real-time PCR system (ThermoFisher Scientific, MA, USA) with ROX as a passive reference. A no-template control with RNase-free water instead of cDNA was used as negative control for all TaqMan gene expression assays. An inter-plate control for all TaqMan gene expression assays was used to control the thermal cycling between plates. ActB (Rn00667869_m1) and Gapdh (Rn01775763_g1) acted as housekeeping genes. All TaqMan gene expression assays were pipetted in triplicates for each sample.
Analysis and Statistics
Analysis of microarray data
Basic Affymetrix GeneChip analysis and experimental quality control were performed using the Expression Console Software (v1.1.2), and the Robust Multi-array analysis method was used for probe summarization and data normalization (quantile normalization and log transformation). Data filtration was done for probe sets having a value less than the median values of the negative control in 80% of total samples.
Significance analysis of microarray was performed using the TMEV software (v4.0). Differentially expressed genes with a false detection rate (q-value) of zero were selected and single probe sets with more than one annotation were excluded. For downstream analyses, we performed enrichment analyses using the PANTHER Classification System [42–44]. Enriched PANTHER pathways and protein classes (v15), Reactome pathways (v65) and GO biological process terms (released 2020-02-21) were identified using Fisher’s Exact Test and the p-values were adjusted for multiple testing using the conservative Bonferroni correction. P < 0.05 was considered statistically significant.
For the biological process GO terms, only the most specialized term within each hierarchical group was included and GO terms with less than 30 annotated genes from the list of differentially expressed genes were excluded. For the PANTHER protein classes and Reactome pathways, only the least specialized protein class/pathway within each hierarchical group was included (cut-off: at least 10 annotated genes). On the contrary, only the most specialized pathway within each hierarchical group was included for the PANTHER pathways (cut-off: at least 5 annotated genes). Subsequently, overlap between overrepresented GO terms for the differentially expressed genes in the occluded MCA and non-occluded MCA both compared with non-stroke controls were identified.
To explore interactions between selected gene products, we utilized the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database v11.0 . Only experimentally determined and database curated Rattus Norvegicus protein-protein interactions were used to create the networks. We defined a cluster as a network formed by at least 5 interacting proteins.
Gene selection and GO term categorization
Based on the results from the microarray, eight genes were selected for validation with qPCR. Ccl2, S1pr3, Socs3, Serpine1, JunB and Fosl1 were selected due to occurrence in biological processes that were significantly enriched in the GO analysis in combination with analysis of fold change and review of literature. Adamts4 was selected due to a very high relative expression, and Olr1 was selected due to the availability of previous data in the same model in hypertensive and normotensive males. In addition to qPCR, we categorized the selected genes into preselected GO biological process terms using the PANTHER Classification System [42–44] and the statistical software R (v4.0.2) to examine their involvement in specific biological processes.
In the PCR-validation, the mean computed cycle threshold (CT) for the triplicates of each sample was calculated. The results were normalized against the inter-plate control. The mean CT-value from the housekeeping genes ActB and Gapdh were subtracted from the mean CT value of the target gene to be able to compare samples that contained a differing amount of total-RNA. This relative value is hereafter referred to as delta-CT (dCT). With each gene, three groups were compared against each other (control, non-occluded and occluded) using the Kruskal-Wallis test. Statistical analyses used in these studies were calculated using GraphPad Prism 7 (GraphPad Software, La Jolla, CA). These data are presented as mean ± SEM and statistical significance was set at p < 0.05. If statistically significant differences (p < 0.05) between the groups were detected, post hoc Mann-Whitney was performed against the control. Due to lack of complete pairs, a pairwise comparison between the occluded artery and its internal non-occluded control was not performed. Since a higher dCT value corresponds with a lower expression, the axes in the figures were reversed to offer a more intuitive reading of the figures.
Cross-analysis with findings from Grell et al. 
To reveal sex similarities between males and females at a larger scale, we compared the findings from the current study with findings obtained by Grell et al. . Here, they investigated differential expression of genes in the occluded MCA compared with the non-occluded MCA in WKY male rats. They used the same stroke model (tMCAO), the rats were the same age (12-weeks) and the samples were processed by the same center using the same method (SCIBLU). Even though the groups are similar in many ways, some of the variations observed might be due to strain differences rather than sex differences. Therefore, we decided to focus solely on the overlapping differentially expressed genes.