Patient sample collection

For the ischemia stroke group, the brain tissues were collected from patients conformed to surgical decompression due to intracranial hypertension after ischemia stroke. For the controlled group, the normal brain tissues were collected from patients who underwent surgical decompression due to intracranial hypertension without ischemia stroke, such as traumatic brain injury. During the operation, only the brain tissue at ischemic sites flowing from the incision or adhering to the meninges was collected. Informed consent was obtained from patients and/or their authorized agent before obtaining clinical samples. This study was approved by Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology.

Construction of RNF13-KO mouse

Forecasting a sequence of guide-RNA (sgRNA: GGATGCTCATGCTGTCCGCCACAC) target to the exon 2 of the RNF13 gene throughout online CRISPR design tools (http://chopchop.cbu.uib.no/). The pUC57-sgRNA (Addgene, 51132) was used as the skeleton vector to construct the RNF13-sgRNA expression vector. Firstly, mixing the purified products of Cas9 expression vector pST1374-Cas9 (Addgene, 44758) and sgRNA expression vector in vitro. Subsequently, the mixture was injected into single cell fertilized ovum of C57BL/6 mouse by FemtoJet 5247 microinjection system. Ultimately, the injected fertilized eggs were transplanted into surrogate female mice. After a period of 19–21 days of gestation, the F0 generation mice were obtained. To confirm the genetic identification, genomic DNA was extracted from the toe tissue of mice at 2 weeks after birth. The identification primers were:

F: GAAAGGTCCGGTGATTTGAA

R: GCATTAGGGGACGTTTTCAG

Male RNF13-KO mice and their littermate negative mice (C57BL/6 background) aged 10–12 weeks and weighing 26–28 g were selected for the further experiments. All mice were housed in SPF Laboratory Animal Center of Wuhan University. The mice were kept at room temperature 22–24℃, a humidity 40–70%, alternating lighting time for 12 h, and were free to drink and eat.

Construction of cerebral ischemia-reperfusion injury model in mice

Before operation, the mice were anesthetized with 2.0% isoflurane and oxygen/nitrous oxide mixture. Primarily, a longitudinal incision on the left cranial roof skin was made to expose the skull and peel the connective tissue from the skull surface. To record the cerebral blood flow, the regional cerebral blood flow (rCBF) was measured using a laser-Doppler flowmetry instrument with a flexible probe affixed to the skull (1.5 mm posterior and 3–4 mm lateral to the bregma). The rectal temperature was maintained at 37 ± 0.5 °C using a homoeothermic blanket. To establish MCAO (middle cerebral artery occlusion), a 8 − 0 silicon-coated monofilament surgical suture was inserted into the left external carotid artery, advanced into the internal carotid artery and wedged into the cerebral arterial circle to obstruct the origin of the middle cerebral artery. A decline in rCBF > 75% confirmed the interruption of blood flow in mice. 45 min after ischemia, the suture was withdrawn, and a return to > 70% of basal cerebral blood flow within 10 min of suture withdrawal confirmed reperfusion of the ischemia territory. On the contrary, the mice in which the suture was withdrawn immediately after the decline in rCBF were used as sham-operated group.

Neurological function score and sampling of mice

After 24 h (or indicated time points) reperfusion, neurological function score was evaluated according to Modified Berderson scoring (9- point scale) [24]:

0

score: no symptoms of nerve damage;

1

score: the opposite forelimb curls up when lifting the tail, or cannot reach the affected forelimb completely;

2

points: the opposite shoulder is adducted when the tail is lifted;

3

points: flat push: resistance decreases when pushing to the opposite side;

4

points: move spontaneously in all directions, but only turn to the opposite side when taking off the tail;

5

points: turning in a circle or only turning to the opposite side when moving spontaneously;

6

points: involuntary movement, only when stimulated;

7

points: involuntary movement, no movement when stimulated;

8

points: death related to cerebral ischemia.

After the evaluation was completed, the mice were anesthetized with tribromoethanol (1.25%, 20 ml/kg, intraperitoneal injection) and sacrificed. Brain tissues were taken for subsequent experiments.

TTC staining and infarct volume statistics

The brain tissue was frozen at -20℃ for 30 min and cut into 1 mm coronal sections. Generally, 7 slices were cut (4 slices anterior to the fontanel and 3 slices posterior). The slices were immediately placed in 2% TTC staining solution and incubated at 37℃ for 10 min. Turn the brain slices over from time to time so that the tissue is stained evenly. Normal brain tissue is bright red when stained, while the infarcted area is pale. After fixation with 4% neutral paraformaldehyde solution, the infarct volume and proportion of edema were calculated using Image Pro Plus software (version 6.0).

Pathological analysis

To prepare the frozen section, the brain tissues were refrigerated and completely precipitated in 40 ml of 20% (W/V) sucrose solution. Subsequently, the brains were transferred to 40 ml of 30% (W/V) sucrose solution. Then, they were immersed in the embedding frame with OTC, and frozen sections were performed using freezing microtome (Thermo, NX50).

For hematoxylin and eosin (H&E) staining, after placing the paraffin slices in oven at 60℃ for 60 min, the paraffin slice was added hematoxylin (Sinopharm Group, G1004) and eosin (BASO, BA-4024) dye for 30s–5 min, Then, the slices were washed with distilled water for 3 times. Finally, the tablet was sealed with ultra-clean sealer (BASO, BA-7004) and photographed under light microscope (Nikon, ECLIPSE 80i).

For IBA1 (Ionized Calcium-binding Adapter Molecule 1), and GFAP (Glial fibrillary acidic protein) immunofluorescence staining, frozen sections were incubated with anti-IBA1 (Abcam, ab5076) and anti-GFAP (HUABIO, ET1601-23) primary antibody overnight at 4 °C. After washing with PBS, sections were incubated with the second antibody for 1 h. The nucleus was stained with DAPI. The images were observed and photographed under a fluoroscope (OLYMPUS, BX51), and counting of positive cells was analyzed by Image J software (version 6.0).

For TUNEL (Terminal Deoxynucleotidyl Transferase-mediated dUTP Nick-End Labeling) staining, the frozen sections of brain tissue were stained with a TUNEL staining kit (Roche, 11684817910). The brain frozen sections were incubated with anti-NeuN primary antibody (Proteintech, 26975-1-AP) overnight at 4 °C. The Alexa Fluor® 568 goat anti-Rabbit IgG (H + L) (Invitrogen, A11036) was used as the second antibody. The nucleus was labeled with DAPI (SouthernBiotech, 0100 − 20). The photos were taken by a fluoroscope (OLYMPUS, BX51), and Image Pro Plus (version 6.0) was used for image analysis.

For F4/80 immunofluorescence staining, frozen sections were incubated with F4/80 antibody (Serotec, MCA497) overnight at 4 °C. After washing with PBS, sections were incubated with the second antibody (Alexa Fluor® 555 Conjugate Anti-rat IgG (H + L) (CST, 4417)) for 1 h. The nucleus was stained with DAPI. The images were observed and photographed under a fluoroscope (OLYMPUS, BX51), and counting of positive cells was analyzed by Image Pro Plus software (version 6.0).

For p-p65, RIP3 (Receptor interacting serine/threonine kinase 3) and RNF13 immunohistochemical staining, frozen sections were incubated with anti-p-p65 antibody (CST, 3033), anti-RNF13 antibody (Abclonal, A8363) and anti-RIP3 antibody (Abcam, ab72106) overnight at 4 °C. After incubation, the sections were washed with PBS. Then, the Rabbit Two-step Detection Kit (Rabbit Enhanced Polymer Detection System) (ZSGB-BIO, PV-9001) was incubated according to the instructions. DAB (ZSGB-BIO, ZLI-9018) was used for color development and hematoxylin (Servicebio, G1004) was used to stain the nucleus. The images were observed and photographed under a bright field microscope (Nikon, ECLIPSE 80i), and analyzed with Image Pro Plus software (version 6.0).

Isolation, culture and OGD treatment of primary neuron cells

Sprague-Dawley rats aged 1–2 days were used to collect cerebral cortices. Then, the cerebral cortices were cut into small pieces and digested with 10 ml of 0.125% trypsin (GIBCO, Grand Island, NY, USA) for 15 min at 37℃. The digestion reaction was terminated with DMEM-F12 medium (GIBCO) containing 10% fetal bovine serum and DNA enzyme (Roche). Clumps of cells and undigested tissue blocks were removed with a 100 μm cell strainer. Subsequently, cells were centrifuged at 1500 rmp for 5 min at 4℃. The cell pellet was resuspended in DMEM-F12 medium containing 10% fetal bovine serum and 1% double antibody (GIBCO). After cells were counting and viability were determined, the cells were plated on inoculated on poly-lysine (10 mg/ml, Sigma) coated dishes and cultured in a 5% CO2 incubator at 37℃ for 3 h. Then, the medium was replaced with neurobasal medium containing 2% B27 when the cells began to develop synapses and showed a tadpole shape. The medium should be protected from light and changed every 48 h. The experiments were conducted after 7 days of culture.

The OGD/R model was constructed to mimic cerebral I/R in vitro. In brief, the neurons were cultured in the DMEM medium (Gibco, 11966025) with serum-free, glucose-free and sodium pyruvate-free medium under hypoxic conditions (94% N2, 5% CO2 and 1% O2) for 3 h. Subsequently, the medium was replaced with normal medium and continued cultured under normal oxygen conditions for 6 h. The cells in the control group were not treated with hypoxia.

RT-PCR

The total RNA of tissues and cells was extracted using TRIzol (Sigma, T9424). The RNA was reverse-transcribed into cDNA using a Transcriptor First Strand cDNA Synthesis Kit (Vazyme Biotech Co., R323-01). RT-PCR was carried out using a ChamQ SYBR Master Mix (Vazyme Biotech Co., Q311-03) and a LightCycler 480 qPCR System (Roche Holding AG). β-Actin was used as the internal reference gene. The primer sequences used are shown in Supplementary Table 1.

Western blot

The cells or tissues were lysed using of RIPA lysate (65 mMTris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate and 0.1% SDS) with protease (Roche, 04693132001) and phosphatase inhibitors (Roche, 4906837001). Then, the BCA protein quantification kit (Thermo, 23225) was taken to determine protein concentration. Subsequently, the protein samples of the equal concentration were mixed with the loading buffer and boiled for 15 min at 95℃. The protein samples were separated by 10% SDS-PAGE electrophoresis and subsequently transferred to a 0.45 μm PVDF membrane (IPVH00010, Millipore). After sealed with 5% skim milk powder at room temperature for 1 h, primary antibody was added and incubated overnight at 4℃. After washing in TBST thrice, the membrane was incubated with the corresponding secondary antibodies. The ECL luminescent substrate (Bio-Rad, 1705062) was used for color development, and a Bio-Rad imaging system (ChemiDoc XRS+) was used for visualization. All antibodies used are listed in Supplementary Table 2.

Luciferase

The 45 pathway reporter plasmids, sea cucumber luciferase plasmids and RNF13 overexpression or control plasmids were co-transfected into 293T cells. 24 h after transfection, cells were dissolved with 1×Passive Lysis Buffer (Dual-Luciferase Reporter Assay System, Promega). Then, LAR II and Stop & Glo® Reagent were added to collect the fluorescence signals by multi-function reagent and perform further statistical analysis.

Vector construction

We designed PCR primers to obtain the CDs (Coding sequences) of RNF13 and p62 genes according to human cDNA as template, then reconstructed them into mammalian overexpression vectors pcDNA5-FLAG, pcDNA5-HA and pcDNA5-GST-HA, respectively. Finally, pcDNA5-FLAG-RNF13, pcDNA5-HA-RNF13, pcDNA5-HA-p62, pcDNA5-GST-HA-RNF13, pcDNA5-GST-HA-p62, pcDNA5-Myc-ub, and pcDNA5-FLAG-Nrf2 were obtained for further experiments.

Adenovirus construction and infection

The adenovirus of RNF13-knockdown in rats was accomplished by constructing replication-deficient adenovirus vectors carrying short hairpin RNA targeting RNF13, with the AdshRNA adenovirus as the control. Similarly, the adenovirus of RNF13-overexpression in mouses was constructed with GFP expression as the control. Primary neuronal cells were infected with adenovirus according to the multiplicity of infection (MOI) of 100 particles/cells, and then identified by Western blot or RT-PCR. The primers were shown in supplementary materials.

Immunoprecipitation and mass spectrometry assay

Cultured cells were transfected with the indicated plasmids or adenovirus supernatant for 24 h and lysed in ice-cold IP buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 1mM EDTA and 1% Triton X-100) containing phosphatase inhibitor tablets (4906837001; Roche) and protease inhibitor cocktail tablets (04693132001, Roche) for 30 min. After an ultrasonic bath (5s on/5s off, power 20%) and centrifugation (12,000 g for 10 min), supernatants were divided into the input group and the IP group. For the input group, the 4× loading buffer was added in supernatants. For the IP group, the supernatants were incubated with protein A/G agarose beads (catalog no.: AA104307; Bestchrom, Shanghai, China) and anti-tag antibody overnight at 4 °C. Finally, the IP group were washed 4 times with 300 mM and 150 mM NaCl buffers respectively, and Western blot analysis was performed after 2× loading buffer was added. For mass spectrometry (MS) analysis, the eluate was separated by 10% SDS-PAGE gels and stained according to the instructions of the Pierce Silver Stain Kit (24600; Thermo Fisher, Rockford, IL). The bands were excised and the proteins were subjected to liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis.

GST pull-down

The fusion proteins Flag-p62, GST-HA-RNF13, Flag-RNF13 and GST-HA-p62 were overexpressed in 293T cells, which were then lysed by lysis buffer (50 mM Na2HPO4, pH 8.0; 300 mM NaCl; 1% TritonX-100; cocktail). After purified by GST beads, Flag-p62 and GST-HA-RNF13, or Flag-RNF13 and GST-HA-p62 protein were mixed and incubated overnight at 4 °C. After washed about 3 times with the buffers (20 mM Tris-HCl; 150 mM NaCl; 0.2% TritonX-100), and resuspended in 2× SDS loading buffer, the immunocomplex was boiled at 95 °C for 5–10 min.

RNA sequencing and analysis

For the RNA-sequencing (RNA‐seq) analysis, total RNA was extracted from WT and RNF13‐KO mice brain tissues after 45 min of ischemia and 24 h of reperfusion. Then, cDNA (complementary DNA) libraries were constructed using a kit provided by MGI Tech. After the library was constructed, the raw reads were extracted using a sequenator (MGISEQ-2000 platform). Subsequently, clean reads were obtained by deleting reads of low quality, contamination and high content of unknown base N using SOAPnuke software (version 2.0.7). Clean reads were matched to reference genome sequences using HISAT2 software (version 2.1.0), and using StringTie software (version 1.3.3b) to calculate the expression level of genes. Ultimately, DESeq2 software (version 1.2.10) was used to calculate differential gene expression according to the following two criteria: [1] folding change > 2 and [2] adjusted p value < 0.05.

Hierarchical clustering analysis

The hierarchical clustering analysis was performed to analyze sample distribution profiles by constructing a clustering tree based on data from RNA-seq by the hclust function of R software.

KEGG pathway enrichment analysis

A KEGG (Kyoto encyclopedia of genes and genomes) pathway enrichment analysis was performed to analyze the differential expression genes through phyper function in R software. KEGG pathway annotations were downloaded from the KEGG database, and p values < 0.05 were considered statistically significant enrichment pathways.

Gene set enrichment analysis

The Java gene set enrichment analysis (GSEA, version 3.0) platform was performed to analyze gene set enrichment based on the degree of differential expression of genes. Gene sets with p values < 0.05 and false discovery rate values < 0.25 were considered statistically significant.

Molecular docking

The crystal structure of the protein was obtained from the PDB (Protein Data Bank) database. The protein models used for docking were RNF13_HUMAN (PDB ID: 5ZC4), p62_HUMAN (PDB ID: 6MJ7), and TRIM21_HUMAN (PDB ID: 7BBD). HDOCK SERVER (http://hdock.phys.hust.edu.cn/) was used for protein-protein molecular docking. The protein model was pre-treated with PyMol 2.4 (removing water molecules and excess ligands, adding hydrogen atoms). Docking Score, Confidence Score and Ligand RMSD are used as docking evaluation criteria. The model with the highest score is selected as the best docking model. Finally, we use Pymol 2.4 software to visualize the docking results.

Statistical analysis

Data were analyzed using SPSS 21.0 and GraphPad Prism 8.0 was used to draw the graphs. Data were expressed as mean ± SD. Shapiro-Wilk’s test was used to determine data normality. For normally distributed data, differences between multiple groups were analyzed using one-way or two-way ANOVA. The differences between two groups were calculated using the Student’s t-test. A Kruskal-Wallis nonparametric statistical test was used when the data showed non‐normal distribution. The probability level accepted for significance was p < 0.05.