Loss of m6A demethylase ALKBH5 promotes post‐ischemic angiogenesis via post‐transcriptional stabilization of WNT5A

CMECs were cultured as previously described.28, 29, 30, 31 Briefly, 2‐week‐old male standard deviation (SD) rats were cervical dislocated, and the hearts were removed. Next, the epicardial and endocardial tissues were removed by ophthalmic forceps, the ventricular tissues were split into 1 mm³, plated in dishes pre‐coated with 1 mL FBS (BI, #04‐001‐1A), and cultured for 4 h. Then high‐glucose DMEC (Gibco, #1791920) with 10% FBS was added for another 48 h of incubation. After the cells reached 80% confluency, the pieces were removed, and the cells were passaged. The second‐passage was used for subsequent experiments. The purity of cultured CMECs was identified by immunofluorescence and flow cytometry, as described in the supplemental materials. The antibodies for immunofluorescence staining and flow cytometry analysis used in this study are listed in Table S1. For hypoxic treatment, CMECs were divided into two groups: hypoxic group, CMECs were placed in a 37°C incubator containing 1% O₂, 5% CO₂, and 94% N₂ for 24 h and normoxic group, CMECs were placed in a 37°C incubator containing 5% CO₂ and 21% O₂ for 24 h.

The equal amounts of protein extracts were separated by SDS‐PAGE and then transferred to poly vinylidene fluoride (PVDF) membranes (Merck, #ISEQ00010). The membranes were then blocked with 5% Bovine serum albumin (BSA) and individually incubated with the independent primary antibodies at 4°C overnight. After subsequent washes with Tris Buffered Saline with Tween 20 (TBST) three times, the PVDF membranes were then incubated with the Horseradish Peroxidase (HRP)‐linked secondary antibodies at room temperature for 1 h. The Electrochemiluminescence (ECL) substrate (Thermo, #32132) was added, and the proteins were visualized using the ChemiDoc Imaging System (Bio‐Rad, CA, USA). The protein bands' density was determined using NIH ImageJ software (1.50i, US) after normalization to β‐actin. The antibodies for Western blot used in this study are listed in Table. S1

Total RNA was extracted with TRIzol (Invitrogen, #15,596,026) according to the manufacturer's instruction. Afterwards, 1 μg of total RNAs was reverse‐transcribed using the PrimeScript RT kit (TaKaRa, #RR036A) and then amplified with SYBR Green dye (TaKaRa, #RR420A) on a CFX96 real‐time Polymerase chain reaction (PCR) System (Bio‐Rad Laboratories, Inc., CA, USA) for amplification. A total reaction system of 10 μl was used, which included 5 μl SYBR Green, 1 μl template DNA, 0.5 μl forward primers, 0.5 μl reverse primers, and 3 μl ddH₂O. The two‐step PCR amplification method was performed: 30 s at 95°C, 5 s at 95°C, and 30 s at 60°C for 39 cycles. Relative gene expression was normalized to β‐actin using the standard 2^−△△Ct quantification method. The primer sequences are listed in Table S2.

Total RNAs were extracted, and the concentrations were adjusted by serial dilution to 50 ng/μl and 100 ng/μl for one assay. The diluted total RNAs were denatured at 95°C for 3 min to disrupt any secondary structures. Then, 2 μl of serially diluted RNAs was dropped onto a Hybond‐N⁺ membrane (GE Healthcare, #RPN203B) and cross‐linked with a Stratalinker 2400 UV Crosslinker (1,200 μJ [ × 100]; 25–50 s). The membrane was then washed and blocked with 5% BSA for 1 h at room temperature and subsequently incubated overnight at 4°C with anti‐m6A antibody (Table S1). After extensive washing, the Hybond‐N⁺ membrane was incubated with HRP‐linked secondary antibody at room temperature for 1 h and exposed to ECL substrate (Thermo, #32132). The same amount of total RNAs was stained with 0.02% methylene blue in ddH₂O for 2 h at room temperature and then photographed to verify the equal loading of RNA.

The ALKBH5 overexpression and control adenoviruses were purchased from the Oobio Co., Ltd. (Shanghai, China). Adenoviral infection was performed according to the recommended protocol. Briefly, 1 × 10⁸ pfu/ml of control and overexpression adenovirus were quantified and diluted in serum‐free DMEC medium. Subsequently, the adenovirus mixtures were added to the cultured plates containing CMECs. The supernatants were discarded after 12 h and replaced with standard DMEM containing 10% FBS for another 12 h. Real‐time quantitative PCR (RT‐qPCR) and Western blot assays were used to detect ALKBH5 expression after 24 h of adenovirus infection.

CMECs were cultured and plated in a six‐well plate overnight. Then, transfected with ALKBH5 siRNA (siALKBH5) (#siG170714021452‐1‐5) or negative control siRNA (siCtrl) (#siN0000001‐1‐5) purchased from RiboBio Co., Ltd. (Guangzhou, China) using the Lipofectamine RNAiMAX Transfection Reagent (Thermo‐Fisher Scientific, #13778075), according to the recommended instruction. Briefly, 2.5 μl siRNA and 7.5 μl transfection reagents were diluted in 250 μl DMEM. The transfection mixture was added to 1.725 ml DMEM containing 10% FBS, which was then supplemented to each well and incubated at 37°C for 24 h. ALKBH5 siRNAs sequences are listed in Table. S3

The Cell counting kit‐8 (CCK‐8) (Biotechwel, #WH1199) and EdU (RiboBio, #C10310↗ ‐1) assays were performed according to the manufacturers' instructions. Briefly, after transfected with ALKBH5 siRNA or infected with the ALKBH5 adenovirus, CMECs were reseeded into a 96‐well plate (5000 cells/ well), followed by the normoxic or hypoxic treatment for 24 h. Then, the CMECs were incubated with 10 μl CCK‐8 solution per well for another 2 h at 37°C, and the absorbance value (OD) was detected at 450 nm using a microplate reader (Synergy H4; BioTek Instruments, Inc., USA). In EdU assay, CMECs were incubated with an EdU mixture (50 μM) for 2 h, fixed with 4% paraformaldehyde, decolorized with 2 mg/ml glycine, permeabilized with 0.5% TritonX‐100, and washed three times with PBS. Afterward, the CMECs were incubated with 100 μl 1 × Apollo for 30 min and stained with 100 μl 1 × Hoechst 33342 (1:100 dilution) for another 10 min. The EdU positive CMCEs were visualized in a fluorescence microscope.

CMECs were seeded in a 24‐well plate after being transfected with ALKBH5 siRNA or infected with the ALKBH5 adenovirus. Once the cells reached a ∼70%–80% confluency, a straight scratch was slowly made in the monolayer across the center of each well with a 10 μl pipette tip. The plate was then gently washed to remove detached CMECs. The CMECs were cultured for additional 24 h under normoxic or hypoxic condition. The scratch widths were visualized and quantified using ImageJ software. The transwell assay was also performed with a 24‐well Boyden chamber with porous polycarbonate membrane (8 μm pore size; Corning, NY, USA). Briefly, CMECs (10,000 cells/well) transfected with ALKBH5 siRNA or infected with the ALKBH5 adenovirus were suspended in DMEM supplemented with 0.1% FBS (200 μl, upper chamber) and filled with 500 μl DMEM supplemented with 10% FBS (lower chamber). After incubation for 24 h in normoxic or hypoxic condition, the cells that migrated across the filter were washed, stained (0.1% crystal violet), fixed (4% paraformaldehyde), and photographed using a microscope. Finally, the number of migratory cells was determined with NIH ImageJ software.

After transfected with ALKBH5 siRNA or infection with the ALKBH5 adenovirus and exposure to normoxic or hypoxic condition for 24 h, 15,000 CMECs were resuspended in 50 μl endothelial cell growth medium (PromoCell, #C‐22010) and seeded onto a 10 μl angiogenesis slide (Ibidi, #81506) pre‐coated with matrigel (BD Biosciences, #356231). After 3 h, the slides were visualized for tube formation. Floating cells were first removed by washing, and 50 μl of calcein (Invitrogen, #2049068) diluted in serum‐free medium (1:160 dilution) was added and incubated for another 30 min in the dark. Then, excess calcein was removed and washed with PBS. Images were taken with a fluorescence microscope, and tube formation was measured using the Angiogenesis Analyzer by NIH ImageJ software.

Methylated RNA immunoprecipitation sequencing (MeRIP‐seq) service was provided by Cloudseq Biotech Inc. (Shanghai, China). Briefly, >100 ug total RNAs from CMECs were extracted after transfected with control siRNA or ALKBH5 siRNA, followed by exposure to hypoxic condition. Then, total RNAs were isolated for polyA⁺ RNAs (Invitrogen) and quantified. Note that 500 ng pre‐fragmented mRNA was reserved as the input control for RNA‐seq, and 5 μg fragmented mRNA was then incubated with 5 μg anti‐m6A antibody in an immunoprecipitation buffer for 2 h at 4°C. Protein‐A/G beads (Thermo‐Fisher) were then added and incubated for 2 h at 4°C. Then, bound mRNA was eluted and extracted with TRIzol reagent. Purified mRNA was used for library generation; the input and the immunoprecipitation samples were used for RNA‐seq. The peaks were identified, and the differentially methylated peaks were counted and analized by the gene ontology (GO). MeRIP‐qPCR and subsequent RNA immunoprecipitation qPCR (RIP‐qPCR) were performed using the RIP Kit (Millipore, Bedford, MA, USA) with antibodies for m6A (MeRIP‐qPCR) or ALKBH5 (RIP‐qPCR), according to the recommended instructions. Briefly, CMECs were lysed with RIP buffer and immunoprecipitated with specific or control IgG antibodies at 4°C overnight, followed by RNA purification. Finally, the immunoprecipitated RNA was analyzed by RT‐qPCR. The primer sequences are shown in Table S4.

CMECs (2 × 10⁵/ well) were seeded in a six‐well plate. After transfected with ALKBH5 siRNA or infected with the ALKBH5 adenovirus and exposure to the normoxic or hypoxic condition, 20 μl actinomycin D stock (1 mg/ml) diluted in 100 μl medium was added dropwise into each well. Samples were collected at 0, 1, 2, 4, 6, and 8 h after actinomycin D addition, and the cell pellets were harvested in 1 ml TRIzol reagent. Total RNAs were extracted and reverse‐transcribed for RT‐qPCR analysis. The average Ct value for each time point was normalized to 0 h to calculate the mRNA's abundance. The relative mRNA decay rate and half‐life were determined by non‐linear regression curve fitting (one phase decay) using GraphPad Prism 8.0 (San Diego, CA, USA).

ALKBH5 global knockout (KO) mice were constructed by GemPharmatech (Nanjing, China). The littermate wild‐type (WT) and ALKBH5 KO mice were cervical dislocated, and the thoracic aortas were removed. The aortic rings (∼1 mm) were cut, embedded in 1 mg/ml matrigel (BD Biosciences), and incubated for 30 min at 37°C. DMEM containing 10% FBS was added to the wells and replaced every 2 days. Finally, the images of sprouting vessels were obtained after 5 days in a microscope. The total sprout lengths were calculated using NIH ImageJ software. Primer sequences for genotyping of the ALKBH5 KO and WT mice are listed in Table. S5

Before surgery, mice were anesthetized with 2% isoflurane, and the hair of the hind‐limb was removed. After preparation, ∼10 mm incision was made, and the femoral vein was separated from the nerve. The femoral artery segment proximal to the outlet of the profundal femoris artery and distal to the outlet of the saphenous artery was ligated then excised. Single‐layer closure using 4–0 prolene sutures was performed. The blood flow was detected using a laser Doppler ultrasound scanning system (PeriScan PIM 3 system, Perimed, Sweden) before and after hind‐limb ischemia at days 0, 3, 7, 14, and 21. The perfusion ratio of the ischemic hind‐limb versus the non‐ischemic hind‐limb was quantified by the relative average flux units.

Adeno‐associated virus (AAV) and adenovirus were purchased from Oobio Co., Ltd. (Shanghai, China). In AAV‐mediated ALKBH5 overexpression, 1 × 10¹¹ v.g./ml of control AAV (OE‐Ctrl), and overexpression AAV (OE‐ALKBH5) diluted in 10 μl PBS were injected into the WT mice gastrocnemius 4 weeks before hind‐limb ischemia. The RT‐qPCR and Western blot were then performed to detect ALKBH5 expression. In adenoviral injection, the WT mice were subjected to hind‐limb ischemia firstly. Then, 1 × 10¹⁰ pfu/ml of control adenovirus (shCtrl) or ALKBH5‐knockdown adenovirus (shALKBH5) was injected into the gastrocnemius immediately after hind‐limb ischemia as well as 1 and 2 weeks later. Every 1 week later, the RT‐qPCR and Western blot were used to detect the silencing efficacy after injection.

Gastrocnemius tissues were frozen in liquid nitrogen and embedded with optimal cutting temperature compound and sectioned at a thickness of 6–8 μm. Sections were permeated with 0.5% TritonX‐100 for 15 min, blocked with 5% BSA for 1 h at room temperature and then incubated with pre‐diluted CD31 and alpha‐smooth muscle actin (α‐SMA) antibodies at 4°C overnight. Next, the sections were incubated with the fluorescent secondary antibodies at room temperature for 1 h, and the cell nuclei were stained with DAPI. Images were visualized in a fluorescence microscope.

All data are presented as the mean ± SD. The two‐tailed, unpaired Student's t‐test was used to determine the statistical significance for comparisons between two groups. One‐way ANOVA followed by Bonferroni's post hoc test was used to determine statistical significance for comparisons among multiple groups. All experiments were performed and repeated three to six times. A p‐value less than 0.05 was considered to be statistically significant. Statistical analyses were performed with GraphPad Prism 8.0 (San Diego, CA, USA).

To elucidate the role of m6A in post‐ischemia angiogenesis, we isolated identified CMECs (Figures S1A‐S1C) and examined the effects of the hypoxic injury on the angiogenic phenotypes of CMECs, including changes in cell proliferation, migration, and tube formation. The CCK‐8 assay revealed that CMECs viability increased 3 h after hypoxia, peaked at 12 h, and then decreased significantly after 24 h (Figure 1A).

The EdU assay was conducted to evaluate cell proliferation, and the proportion of EdU‐positive cells was significantly decreased after 24 h of hypoxia (Figures 1B and 1C). The migratory capacity of CMECs was evaluated by scratch and migration assays. Hypoxia significantly reduced the healing area (Figures 1B and 1D) and the number of migratory CMECs (Figures 1B and 1E). Furthermore, in vitro tube formation ability of CMECs was significantly impaired by hypoxia (Figures 1B and 1F). Taken together, these results indicate that 24 h of hypoxia inhibits proliferation, migration, and angiogenesis of CMECs.

To determine whether m6A was involved in this process, we performed a dot blot assay, and the data showed that the m6A abundance was significantly reduced in hypoxic CMECs (Figure 1G). Considering m6A is a dynamic process and can be added by m6A writers, such as methyltransferase‐like 3 (METTL3), methyltransferase‐like 14 (METTL14),32 Wilms tumor 1‐associated protein (WTAP),33 and methyltransferase‐like 16 (METTL16),34 or removed by m6A erasers, such as fat‐mass and obesity‐associated protein (FTO)35 and AlkB homolog 5 (ALKBH5).36 We performed RT‐qPCR to evaluate the changes in the mRNA expression levels of these enzymes in hypoxic CMECs and found that methyltransferase WTAP and demethylase ALKBH5 were significantly upregulated after hypoxic injury (Figure 1H). Western blot analysis further verified that only m6A demethylase ALKBH5 expression was significantly upregulated in hypoxic CMECs (Figures 1I and 1J). The expression of the AlkB family members in CMECs was also explored. The results showed that while ALKBH1, ALKBH4, ALKBH7, and ALKBH9 (FTO) remained unchanged after 24 h of hypoxia, the expression of ALKBH2, ALKBH3, ALKBH6, and ALKBH8 was downregulated after 24 h of hypoxia. Only ALKBH5 was upregulated in response to hypoxia among these AlkB homologs (Figure S1D).

We next examined whether hypoxia‐induced ALKBH5 was involved in angiogenesis. Adenovirus‐mediated ALKBH5 overexpression was evaluated under both normoxic and hypoxic condition, and the overexpression efficiency was confirmed by RT‐qPCR (Figure S2A) and Western blot (Figures 2A and S2B). Because m6A possesses a reversible and dynamic feature that involves various enzymes, we also examined the expression of the remaining enzyme, including METTL3, METTL14, METTL16, WTAP, and FTO. RT‐qPCR showed that ALKBH5 overexpression did not alter the expression of the remaining m6A‐related enzymes (Figure S2C). Next, we examined whether ALKBH5 overexpression affected m6A levels. Dot blot assay revealed reduced m6A abundance after ALKBH5 overexpression in CMECs under both normoxic and hypoxic conditions (Figures 2B, S2D, and S2E).

The EdU and CCK‐8 assays results indicated that ALKBH5 overexpression impaired CMECS proliferation (Figures 2C and 2D) and cell viability (Figure 2E) under hypoxic conditions. However, these effects were not observed under normoxic condition. Scratch and transwell assays also showed that ALKBH5 overexpression inhibited the migratory capacity of CMECs in hypoxic but not in normoxic condition (Figures 2F and 2I). Additionally, increased ALKBH5 expression impaired hypoxic tube formation, but not the normoxic CMECs (Figures 2J and 2K). These findings demonstrate that although ALKBH5 upregulation may not significantly affect CMECS function under normoxic condition, it substantially exacerbates CMECS dysfunction under hypoxic condition, thus resulting in impaired angiogenesis.

To further examine the role of ALKBH5 in angiogenesis, small interfering RNA (siRNA) was used to knock down the expression of ALKBH5, and the impact of ALKBH5 knockdown was tested in both normoxic and hypoxic CMECs. After transfection, RT‐qPCR and Western blot assays confirmed that ALKBH5 expression was significantly downregulated under both normoxic and hypoxic condition (Figures 3A, S3A, and S3B). Meanwhile, the expression of the remaining m6A‐related enzymes was not observed to be affected by ALKBH5 knockdown (Figure S3C). Additionally, the dot blot assay further demonstrated that the global m6A levels were significantly elevated after ALKBH5 knockdown under both normoxic and hypoxic condition (Figures 3B, S3D, and S3E).

We next examined the effect of ALKBH5 knockdown on angiogenesis in vitro. The EdU and CCK‐8 assays showed that ALKBH5 knockdown improved the proliferation and viability of CMECs under hypoxic condition (Figures 3C‐3E). Moreover, knocking down ALKBH5 enhances the migration of CMECs under hypoxic condition but not normoxic condition (Figures 3F‐3I). Similarly, silencing ALKBH5 promoted the tube formation of CMECs during hypoxia but not normoxia (Figures 3J and 3K). Together, these data collectively suggest that ALKBH5 knockdown plays a pivotal role in preventing hypoxia‐induced impairment of endothelial angiogenesis.

To explore the underlying mechanism of hypoxia‐induced inhibition of endothelial angiogenesis by ALKBH5, an integrated approach combining MeRIP‐seq and RNA‐seq was applied in CMECs after transfected with control or ALKBH5 siRNA under hypoxic condition. MeRIP‐seq revealed a total of 12,783 common peaks, 477 unique peaks in control, and 2358 unique peaks in the ALKBH5‐knockdown CMECs. Additionally, 771 hypo‐methylation and 1072 hyper‐methylation peaks were found (Figure 4A). MeRIP‐seq further revealed the distribution characteristics (Figure 4B) and percentages (Figure 4C) of differentially methylated mRNA peaks. Representative motif analysis of the two CMECs groups was performed based on these data (Figure 4D).

The differential expression analysis of the enriched genes revealed that 1352 genes were downregulated, whereas 2914 genes were upregulated (Figure 4E). In combination with the MeRIP‐seq analysis, natriuretic peptide C, and F‐box and WD repeat domain containing 5 (FBXW5) were identified as representative genes with hypo‐methylation and hyper‐methylation, respectively, and were viewed using an integrated genome browser (Figure 4F). GO analysis was also performed for these genes and revealed enrichment of hypo‐methylated transcripts mainly involved in processes such as the cell cycle and RNA metabolism. While the hyper‐methylated transcripts were mainly enriched in the angiogenic sprouting and regulation of cell proliferation (Figure 4G). These results further corroborate the critical role of ALKBH5 knockdown‐induced m6A hyper‐methylation for enhanced angiogenesis. The POSTAR (http://lulab.life.tsinghua.edu.cn/postar/index.php↗), angiogenic database (http://angiogenes.uni‐frankfurt.de/↗), our MeRIP‐seq data, and RNA‐seq data of differentially hypermethylated and upregulated genes were combined respectively to identify the ALKBH5 target genes that regulate angiogenesis. Obtained results revealed that S‐phase kinase‐associated protein 2 (SKP2), WNT family member 5A (WNT5A), and fibroblast growth factor 18 (FGF18) were the potential pro‐angiogenic ALKBH5 target genes for further investigation (Figure 4H).

The MeRIP‐seq data revealed significantly increased m6A abundance of SKP2, WNT5A, and FGF18 mRNAs after ALKBH5 knockdown by integrating the genome browser view (Figure 5A). To verify these MeRIP‐seq findings, MeRIP‐qPCR was used to determine m6A modification alterations within these target genes. The results demonstrated a significantly elevated m6A level of SKP2, WNT5A, and FGF18 mRNA after ALKBH5 knockdown (Figure 5B). To further test whether m6A affected the expression of the target genes, RT‐qPCR and Western blot analysis was performed and indicated that silencing ALKBH5 increased WNT5A expression (Figures 5C, 5D, and S4A).

Next, we used RT‐qPCR to examine the stability and decay of WNT5A mRNA in CMECs at different time points after co‐culture with actinomycin D because the internal mRNA m6A modification has different effects on target genes, especially mRNA stability and decay. This transcriptional inhibitor intercalates into DNA and prevents the DNA double‐helix from unwinding, thereby inhibiting gene transcription and mRNA synthesis. The results showed that WNT5A expression was decreased in a time‐dependent manner post the addition of actinomycin D. However, ALKBH5 knockdown significantly delayed the degradation of WNT5A mRNA, thus prolonging its half‐life, compared with the control (Figures 5E and S4B). To further explore the regulating role of ALKBH5 on WNT5A mRNA, we evaluated the expression and stability of WNT5A mRNA after adenoviral ALKBH5 overexpression in hypoxic CMECs. The data showed that ALKBH5 overexpression significantly decreased the expression of WNT5A both in protein and mRNA level (Figures 5F and S4C), shortened half‐life, and diminished stability of WNT5A mRNA (Figures 5G and S4D). To investigate whether ALKBH5 protein could interact with WNT5A mRNA, the RIP‐qPCR experiment was further conducted, and the data showed that ALKBH5 bonded to WNT5A mRNA (Figure 5H). These data collectively demonstrate that ALKBH5 promotes WNT5A mRNA decay and decreases its half‐life via the removal of post‐transcriptional m6A modification.

We further explored whether the ALKBH5‐mediated angiogenesis was dependent on WNT5A. Foxy‐5, a WNT5A agonist, was used after ALKBH5 overexpression in hypoxic CMECs and followed by the detection of angiogenic phenotypes. CCK‐8 and EdU assays showed that Foxy‐5 significantly reversed the inhibitory effect of ALKBH5 overexpression on CMECs proliferation (Figures 5I, 5J, and S4E). Furthermore, Foxy‐5 significantly abrogated the inhibition of migration in CMECs after ALKBH5 overexpression (Figures 5K, S4F, and S4G). Finally, tube formation assay confirmed that Foxy‐5 restored the tube formation capacity of CMECs that was impaired with ALKBH5 overexpression (Figures 5L and S4H). Taken together, these results indicate that WNT5A is an ALKBH5 target mRNA. ALKBH5 reduces WNT5A expression in hypoxic CMECs by destabilizing and decaying the mRNA in a manner that inhibits angiogenic ability.

We next evaluated the role of ALKBH5 in post‐ischemic angiogenesis in a mouse model of hind‐limb ischemia. We first examined the expression of ALKBH5 at different time points post‐ischemia. The data showed that ALKBH5 exhibited a dynamic variation with a prolonged ischemia period. ALKBH5 increased in the early stages of ischemia (within 6 h) compared with the non‐ischemia group. However, ALKBH5 was downregulated from day 1 to day 21 post‐ischemia. On day 28 post‐ischemia, ALKBH5 expression recovered to a normal level. These data showed that ALKBH5 exhibited a dynamic variation during the ischemia period (Figure). S6A

To investigate the in vivo effects of ALKBH5 on angiogenesis in response to ischemia, sustained ALKBH5 overexpression was achieved via AAV injection into the gastrocnemius of mice at 4 weeks prior to hind‐limb ischemia. ALKBH5 overexpression efficiency, m6A levels, blood flow rates, and the expression of angiogenic markers were evaluated prior to hind‐limb ischemia as well as over the duration of 21 days after hind‐limb ischemia (Figure 6A). Our data showed a significant upregulation of ALKBH5 mRNA (Figure S6B) and ALKBH5 protein (Figures 6B and S6C) after injection of ALKBH5‐overexpression AAV (OE‐ALKBH5) into gastrocnemius. Additionally, the m6A abundance was significantly reduced after ALKBH5 overexpression (Figures 6C and S6D). Blood flow recovery after hind‐limb ischemia was measured with a laser Doppler ultrasound scanning system. Mice with ALKBH5 overexpression exhibited significantly decreased blood flow recovery rates on days 7–21 after hind‐limb ischemia compared with control mice (Figures 6D and 6E). Additionally, the expression levels of CD31 and α‐SMA, which indicate the densities of capillaries and small arteries, respectively, were also significantly downregulated in the mice with sustained ALKBH5 overexpression compared with control mice (Figures 6F and 6G). Thus, our data suggest that ALKBH5 overexpression attenuates blood flow recovery and post‐ischemic angiogenesis.

To further investigate the role of ALKBH5 on post‐ischemic angiogenesis, we generated the genetically engineered ALKBH5‐KO mice (Figure). The global ALKBH5 KO mice appear to exhibit a relative reduction in body weight compared to the littermate WT mice (Figure). Considering the concern that bodyweight reduction may affect the functional integrity of vital organs, we further examined the morphology in the heart, liver, and kidney organs. The H&E and Masson staining showed no significant morphological differences between the ALKBH5 WT and KO mice (Figure). S7A–S7C S7D S7E

Next, the ex vivo model of vessel sprouting and the in vivo hind‐limb ischemia were used to verify the angiogenic regulatory role of ALKBH5. Dot blot assay showed that the m6A abundance was significantly increased in both aortic rings and the gastrocnemius tissues from ALKBH5 KO mice (Figures 7A, 7C, S7F, and S7G). Furthermore, the aortic ring assay revealed that the aortic rings of ALKBH5 KO mice exhibited more robust vessel sprouting compared with those of littermate WT mice (Figure 7B). Similarly, ALKBH5 KO mice showed significantly enhanced blood flow restoration on days 7–21 following hind‐limb ischemia (Figure 7D). Moreover, the expression of CD31 and α‐SMA of the gastrocnemius in ALKBH5 KO mice were also significantly upregulated compared with WT mice (Figure 7E). These results illustrate that ALKBH5 KO showed a protective role for post‐ischemic angiogenesis.

Having established the importance of ALKBH5 in post‐ischemic angiogenesis, we next investigated the translational value of ALKBH5. We initiated the short‐term and transient downregulation of ALKBH5 via adenoviral injection into the gastrocnemius of WT mice immediately, 7 days, and 14 days after hind‐limb ischemia (Figure 7F). The significantly decreased expression of ALKBH5 mRNA and ALKBH5 protein at 1, 2, and 3 weeks after adenoviral injection were observed (Figures 7G, S7H, and S7I). Furthermore, with ALKBH5 knockdown, the dot blot assay revealed significantly increased levels of gastrocnemius m6A at 3 weeks post‐ischemia (Figures 7H and S7J). After knocking down ALKBH5, we found significantly higher levels of CD31 and ‐SMA expression in ischemic gastrocnemius (Figure 7I). Finally, Laser Doppler ultrasound scanning data showed that the blood flow recovery was improved significantly in the mice with ALKBH5 knockdown (Figure 7J). Collectively, these results elucidated the critical role of ALKBH5 in post‐ischemic angiogenesis and revealed the potential translational benefits of knockdown or inhibiting ALKBH5. Future studies are warranted to validate the translational position of this proposal aiming to promote angiogenesis after ischemic injury in clinical settings.