β2SP/TET2 complex regulates gene 5hmC modification after cerebral ischemia

Abstract βII spectrin (β2SP) is encoded by Sptbn1 and is involved in the regulation of various cell functions. β2SP contributes to the formation of the myelin sheath, which may be related to the mechanism of neuropathy caused by demyelination. As one of the main features of cerebral ischemia, demyelination plays a key role in the mechanism of cerebral ischemia injury. Here, we showed that β2SP levels were increased, and this molecule interacted with TET2 after ischemic injury. Furthermore, we found that the level of TET2 was decreased in the nucleus when β2SP was knocked out after oxygen and glucose deprivation (OGD), and the level of 5hmC was reduced in the OGD+β2SP KO group. In contrast, the expression of β2SP did not change in TET2 KO mice. In addition, the 5hmC sequencing results revealed that β2SP can affect the level of 5hmC, the differentially hydroxymethylated region (DhMR) mainly related with the Calcium signalling pathway, cGMP‐PKG signalling pathway, Wnt signalling pathway and Hippo signalling pathway. In summary, our results suggest that β2SP could regulate the gene 5hmC by interacted with TET2 and will become a potential therapeutic target for ischemic stroke.

Ten eleven translocation (TET) protein families were discovered by Rao and colleagues in 2009 17 and include TET1, TET2 and TET3. TET1 is expressed in differentiated stem cells and the nervous system; TET2 is widely distributed, mainly in the hematopoietic system; and TET3 is mainly expressed in the colon and muscle, with little expression in brain tissue. 18 The main functions of the three TET enzymes are to convert 5mC-5hmC. The 5hmC modification has been reported to affect the mechanism of many diseases, including stroke. The latest research shows that focal ischemia can increase the activity of the TET enzyme and catalyse the formation of 5hmC, and ascorbic acid (TET enzyme activator) treatment had a neuroprotective effect and improved the recovery of motor function after cerebral ischemia injury in mice. 19 TET2 regulates changes in 5hmC in the promoter region of the Bdnf gene to affect the recovery of neurological function after cerebral ischemia injury, and loss of TET2 significantly increased the volume of cerebral infarction. 20 Moreover, TET2 can regulate the expression of mitochondrial genes by catalysing the modification of mitochondrial DNA 5hmC, thereby damaging mitochondrial function upon ischemia injury. 21 In summary, TET2 plays a critical role in the process of ischemic injury. However, the specific mechanism still needs more in-depth research. In this article, we studied the molecular importance of the interaction between TET2 and β2SP in the brain after ischemic stroke.

| Mice
Male ICR mice (23-25 g) were acquired from the SLAC Company of

| Cell culture and oxygen and glucose deprivation (OGD) cell model
The PC12 cells used in this study were purchased from the American Type Culture Collection (ATCC). The cells were maintained in DMEM with 10% foetal bovine serum (HyClone) at 37°C in a 5% CO 2 incubator. PC12 cells were incubated in sugar-free medium and then placed in an airtight chamber (Billups Rothenberg) flushed with a mixed gas of 95% N2/5% CO 2 for 15 min. The chamber was kept at 37°C for 1 h. Control PC12 cells were incubated with sugar-free medium and placed in the incubator. The cells are restored to their normal culture conditions after the hypoxia is over.

| Mouse model of middle cerebral artery occlusion (MCAO)
We refer to the previous method to prepare the MCAO models. 22 At first, a 6-0 nylon filament was inserted into the internal carotid artery about 9-11 mm. Then, the filament was removed after 45 min, and recover reperfusion. Mice in the sham group underwent the same experimental procedures, but the nylon filament was not inserted. The mouse was placed on an electric blanket to keep the body temperature at 36.5°C-37.5°C, and monitored the brain blood flow. 23

| Western blot
Mouse brain tissues and PC12 cells were solubilized and desaturated. We previously described our analysis of protein expression. 20

| Dot blot
Genomic DNA was extracted from brain tissues and cells for dot blot analysis by the phenol chloroform method. We accordance the previous method of dot blot to test 5hmC levels. 24 Anti-5hmC rabbit antibody (1:10000, Active Motif, 39769, RRID:AB_10013602) was used.

| Methylene blue staining
After dot blot, the membrane was incubated with methylene blue (0.02% in 0.3 M sodium acetate, Sigma-Aldrich Company) for 10 min. 25 Photograph was acquired after washing, and the results were analysed with Alpha Ease Image Analysis Software.

| Coimmunoprecipitation (Co-IP)
Brain tissues and PC12 cells were dissolved in lysis buffer. The samples were centrifuged to collect the supernatant. The supernatant was incubated with anti-TET2 overnight at 4°C. Then, the supernatant was precipitated with protein A/G-agarose beads (Santa Cruz, sc-2003, RRID:AB_10201400) for 4 h at 4°C. Beads were collected, and protein samples were obtained for western blotting.
The Guide RNAs are listed in Table 1

| 5hmC-specific capture and analysis
We follow the previous method to mark 5hmC, 26 coupled with high-throughput sequencing. The sequencing data were processed using previous method. 27 Briefly, Bowtie 2 28 and Samtools 29 softwares were used to deal the FASTQ sequence data, and mapping the data with Mus musculus reference genome (mm10). There were not allowed to appear more than two mismatches in the first 25 bp, and retained one non-repetitive genome. We counted the unique and non-repetitive reads of enriched sample genomic DNA in the 100, 1,000 and 10,000 bp and non-enriched input genomic DNA in whole genome bins, and normalized these data to the total number of repeated reads. This study used ngsplot 30 software to analyse the coverage of each sample reads in the genomes, and the Model-based Analysis of ChIP-Seq (MACS2) 31 was used to determine the true regions or "peaks" of 5hmC-enriched.

| GO and KEGG analyses
GO analyses were to deal with the Database for Annotation, Visualization and Integrated Discovery (DAVID) analysis tools.
KEGG analyses were to deal with Kyoto Encyclopedia of Genes and Genomes (KEGG) software.

| Statistics
All data are expressed as the mean ±SEM. GraphPad Prism 6.0 (GraphPad Software, Inc.) was used for statistical analysis. Student's t-test and one-way analysis were used to determine the differences in the groups. p < 0.05 was considered statistically significant.

| β2SP is increased after ischemic stroke
At present, the level of β2SP is still unknown after cerebral ischemic injury. To explore this, we used the MCAO model and western blot method to detect the β2SP levels in ischemic tissues.
The different time points (6, 12, 24, 48 h and 7 days) were set, and the results showed that the β2SP levels were increased and reached peak at 48 h ( Figure 1A), which was confirmed by quantitative analysis (p < 0.05, Figure 1B). This result was consistent with our previous findings on 5hmC. 20 Moreover, fluorescence staining demonstrated an increase in β2SP at 48 h, and β2SP may have transferring to nucleus ( Figure 1C). These results indicated that the expression and distribution of β2SP were changed after ischemic injury.

| β2SP is transferred to the nucleus and increases 5hmC levels after OGD
Previous articles have shown that the adaptor protein β2SP plays an essential role in the nucleus to drive TGFβ-mediated tumour suppression, 32 and the nuclear accumulation of β2SP was decreased in cirrhotic liver tissue. 33 These results indicated that β2SP can enter the nucleus and play a key role in cell signalling.
To further study the mechanism by which β2SP enters the nucleus after ischemic injury, we used an OGD cell model. First, we detected the expression of β2SP in the cytoplasm and nucleus at different time points (1, 2, 4, 8, 12 and 24 h) after OGD. β2SP levels were increased at 8 h in the cytoplasm (Figure 2A), which was confirmed by quantitative analysis (p < 0.05, Figure 2B). In the nucleus, β2SP expression appeared at 4 h, reached its peak at 8 h and disappeared at 24 h ( Figure 2C); this finding was confirmed by quantitative analysis (p < 0.05, Figure 2D). To explore whether β2SP affects 5hmC modification after entering the nucleus, we detected the levels of 5hmC at different time points (1, 2, 4

| β2SP interacts with TET2 after OGD
Our previous results suggested that the increase of 5hmC was correlated with TET2 after ischemic injury. 20 Figure 3A). Furthermore, Co-IP demonstrated that TET2 interact with β2SP ( Figure 3B).

| β2SP knockout decreases TET2 and 5hmC levels after OGD
From the above experimental results, we know that β2SP increases significantly in the nucleus and interacts with TET2 after ischemic injury. How TET2 change when β2SP is deleted? To explore this topic, we detected changes in the TET2 and 5hmC levels when β2SP was knocked out after OGD. First, we designed 5 sgRNA (Table 1)  These results confirmed that β2SP can affect TET2 translocation to the nucleus to convert 5mC to 5hmC. But the levels of β2SP did not change in the nucleus or cytoplasm in the Tet2 WT and KO mice after MCAO ( Figure 4I-N). These results indicated that TET2 was not necessary for β2SP nuclear transfer and that its expression increased after ischemic injury.

| The 5hmC was changed after β2SP knockout
To explore the effect of deletion of β2SP on 5hmC modification, we analyse the 5hmC distribution of genes in the β2SP KO-PC12 cell line.
The analysis results indicated that the overall 5hmC levels were lower in OGD +β2SP KO group than in OGD group ( Figure 5). Figure 5A showed that the 5hmC levels in the different chromosome. The 5hmC levels significantly decreased in OGD +β2SP KO group ( Figure 5A).
The result of 5hmC reads distribution indicated that the levels were reduced in exon, intron and 3′UTR after β2SP KO ( Figure 5B).
Then, we visually analysed the read count Per Million mapped reads in Genomic Region (5′ −> 3′), and the result showed that the level was decreased in OGD +β2SP KO group ( Figure 5C). These results showed that the 5hmC abundance was decreased after β2SP KO.

F I G U R E 4 βII spectrin knockout decreased the TET2 and 5hmC levels after OGD. (A and D)
The protein levels of β2SP and TET2 in the cytoplasm and nucleus when β2SP was knocked out after OGD. Actin and Histone 3 were used as a control. The statistical analysis of the relative intensity is shown in (B, C, E and F). (G) The level of 5hmC was detected when β2SP was knocked out after OGD, and statistical analysis of the relative intensity is shown in H. (L and M) The protein levels of β2SP in the cytoplasm and nucleus in the mice after OGD. Actin and Histone 3 were used as controls, and the statistical analysis of the relative intensity is shown in I and N * was compared with the control group, # was compared with the OGD group, * p and # p < 0.05 To further study the different genes with significant changes in 5hmC distribution after β2SP KO, we used the 5hmC sequence data to analyse the differentially hydroxymethylated regions (DhMRs).
The results showed that there are 2,234 genes were down-regulated and 1,275 genes were upregulated after β2SP KO ( Figure 5D).
Then, we used DAVID analysis tools to perform Gene Ontology KEGG analysis found that the DhMRs were highly related with cell connection signalling pathway, such as Calcium signalling pathway, cGMP-PKG signalling pathway, Wnt signalling pathway and Hippo signalling pathway ( Figure 5F). GO analysis showed that the DhMRs were related with axonogenesis, positive regulation of neuron differentiation, ameboidal-type cell migration and cell junction assembly ( Figure 5E). interacts with Smad3 via the N-terminal fragment, 37 binds to ankyrin with the C-terminus 38 and so on. β2SP exerts its specific functions through these interactions, including regulation of the cell cycle, apoptosis and transcription. Our results showed that β2SP interacts with TET2 to affect the levels of 5hmC after ischemic injury, but the domain to which they bind has not been identified, and more work is needed to explore the specific domain. Some studies found that full-length αII-spectrin protein was decreased and αII-spectrin breakdown products (αII-SBDPs) were significantly increased in brain tissue after ischemic injury 39,40 ; these SBDPs play a key role in ischemic injury. 41,42 Currently, there is no related report of β2SP in ischemic stroke, and only one article showed that full β2SP(260 kDa)

| DISCUSS ION
is breakdown to 110 kDa, 108 kDa, 85 kDa and 80 kDa fragments in the rat brain (hippocampus and cortex) after traumatic brain injury. 43 In our studies, full-length β2SP was increased significantly, but β2SP breakdown products (βsBDPs) were not significantly changed ( Figure 1A) after I/R injury.
The roles and functions of 5hmC have been studied in many diseases since 2010, when the TET protein family was identified.
Different TET proteins play various functions in different diseases, but TET2 has been reported to play a more important role in ischemic injury. Our previous study showed that the levels of 5hmC were increased, and TET2 protein deletion could increase the Infarct volume after ischemic brain injury. 20 The mtDNA 5hmC abundance is increased after ischemic brain injury and may be associated with the expression of mitochondrial genes and cellular ATP levels. 21 In addition, TET2 was reported to be dispensable for kidney development and function under baseline conditions while protecting against renal IR injury, possibly by repressing the inflammatory response, 44 and TET2 can affect the levels of cell death-related genes by regulating the 5hmC levels after spinal cord injury. 45 However, there are other reports that overall 5hmC abundance in the cortex was decreased significantly, and the reduced expression of Tet1 and Tet2 enzymes might be responsible for this change after hypoxic-ischemic injury 46 ; moreover, Kahlilia C et al. 19 thought that TET3 regulated 5hmC to provide endogenous neuroprotection against cerebral ischemia. Although 5hmC plays a key role in cerebral ischemic injury, the specific mechanism is not yet clear. Our results confirmed that β2SP can regulate the levels of TET2 in the nucleus and affect the levels of 5hmC after ischemic injury.
βII spectrin as an adaptor protein formed by the Smad3/Smad4 complex in the process of TGFβ signal transduction, 47 and participates in the regulation of multiple signal pathways. Wnt signalling was activated after SPTBN1 lose, 48 βIV-spectrin can mediate function and excitability through to bind and recruit Ca 2+ /calmodulin kinase II 49 and Spectrin can regulate the Hippo signalling pathway. 7 Our results showed that the 5hmC modifications of 2,234 genes have a significant decrease after β2SP KO in OGD cell model. KEGG analysis result showed that the DhMRs were highly related with Calcium signalling pathway, cGMP-PKG signalling pathway, Wnt signalling pathway and Hippo signalling pathway. These results were basically the same as previously reported in the literature. These results suggest that β2SP may regulate the 5hmC modify of genes to affect the corresponding signalling pathways through bind with TET2.
In summary, we show that the levels of β2SP were increased after ischemic injury, and these increases were mainly in the nucleus.
Immunofluorescence staining and Co-IP analyses suggested an interaction between β2SP and TET2. β2SP KO reduced the levels of TET2, but TET2 KO did not affect the expression of β2SP after ischemic injury, highlighting the key role of β2SP in ischemic injury. Our data suggest that β2SP could regulate the gene 5hmC by interacting with TET2 and will become a potential therapeutic target for ischemic stroke.

ACK N OWLED G EM ENTS
This work was supported the National Science Foundation of China (81601154, 81601147), the Colleges and Universities Natural Science Foundation of Jiangsu Province (21KJB310020) and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

CO N FLI C T O F I NTE R E S T
We have no conflicts of interest to declare.