RNA Splicing of the Abi1 Gene by MBNL1 contributes to macrophage‐like phenotype modulation of vascular smooth muscle cell during atherogenesis

Abstract Background Vascular smooth muscle cells (VSMC) switch to macrophage‐like cells after cholesterol loading, and this change may play an important role in atherogenesis. Muscleblind‐like splicing regulator 1 (MBNL1) is a well‐known splicing factor that has been implicated in many cellular processes. However, the role of MBNL1 in VSMC macrophage‐like transdifferentiation is largely unknown. In this study, we aim to characterize the role of MBNL1‐induced gene splicing during atherogenesis. Methods The expression of MBNL1 and Abelson interactor 1 (Abi1) splice variants (Abi1‐e10 and Abi1‐Δe10) was compared between artery tissues from healthy donors and atherosclerosis patients. Regulatory mechanisms of MBNL1‐induced Abi1 gene splicing were studied, and the signal pathways mediated by Abi1 splice variants were investigated in VSMC. Results Loss of MBNL1 was found in the macrophage‐like VSMC (VSMC‐M) in artery wall from atherosclerosis patients. In vitro and in vivo evidence confirmed that Abi1 is one of the MBNL1 target genes. Loss of MBNL1 significantly induces the Abi1‐Δe10 isoform expression. Compared to the known actin organization activities of the Abi1 gene, we discovered a novel action of Abi1‐Δe10, whereby Abi1‐Δe10 activates Rac1 independent of upstream stimulation and triggers the Rac1‐NOX1‐ROS pathway, which results in increased expression of transcription factor Kruppel‐like factor 4 (KLF4). While Abi1‐Δe10 inhibits contractile VSMC biomarkers expression and cell contraction, it stimulates VSMC proliferation, migration and macrophage‐like transdifferentiation. Conclusion Loss‐of‐function of MBNL1 activates VSMC‐M transdifferentiation to promote atherogenesis through regulating Abi1 RNA splicing.


| BACKG ROU N D
Atherosclerosis is a chronic inflammatory condition that results from complex interactions of modified lipoproteins and various cell types of the vessel wall. 1 The detail that each particular cell type contributes to atherogenesis is not completely understood. Vascular smooth muscle cells (VSMC) are the main cell type of vascular wall and a major component of atherosclerosis. 1 It is generally accepted that VSMC is not terminally differentiated and can undergo phenotypic transdifferentiation during atherosclerosis development. [1][2][3] Accumulating evidence has suggested that VSMC can convert to macrophage-like cells that have lost classical contractile ability and make up a major component of atherosclerotic lesions. Several mechanisms about VSMC-M transdifferentiation have been purposed; however, knowledge remains to be elucidated.
The development of VSMC-M is regulated by cell lineage plasticity, whereby VSMC cells acquire a stem-like phenotype followed by dedifferentiation to VSMC-M. 7,8 The most well-defined core 'stemness' genes in human cells are OCT4, SOX2, c-Myc and KLF4 which are powerful enough to reprogram terminally differentiated cells into induced pluripotent stem cells. 9 It has been reported that KLF4 is not expressed in differentiated VSMC, but can be induced by oxidized phospholipids and inflammatory cytokines, downregulating the expression of VSMC contractile biomarkers while inducing expression of macrophage biomarkers. 8,10 Apoe−/− mice with VSMCspecific conditional knockout of KLF4 had a decreased number of VSMC-M cells. 11 These findings suggest that upregulation KLF4 may regulate a stem-like gene network to confer VSMC cell lineage plasticity for VSMC-M development.
Abelson interactor 1 (Abi1) is a key component of several intrinsic complexes that regulate actin cytoskeletal remodelling upon upstream stimulation. [12][13][14] Alteration of Abi1 has been reported to associate with cell proliferation, migration and macropinocytosis. [15][16][17] Through alternative splicing, numerous structurally distinct variants of Abi1 exist in human cells, 17,18 providing a potential diversity of Abi1 regulatory signalling, suggesting the possible existence of potential mechanisms through which Abi1 regulates VSMC-M development.
Recently research has focused on identifying the genomic and transcriptomic alteration profile of VSMC-M, and studies of alternative splicing have lagged. Through analysing RNA-seq data, we found that an Abi1 splicing variant that lacking exon10 (Abi1-Δe10) is upregulated in VSMC-M. Abi1-Δe10 splicing is downregulated by splicing factor MBNL1 which expression was decreased in VSMC-M. This splicing event activates Rac1 in absence of upstream stimulating and promotes KLF4 expression through a NADPH oxidase 1 (NOX1)-dependent pathway. Our research provides a novel insight into the splicing regulatory mechanism in VSMC phenotype modulation during atherogenesis.

| Tissue collection
Artery samples were collected from peripheral artery atherosclerosis patients who received thigh amputation, and control arteries were obtained from healthy people who underwent amputation by accidents. 19 Demographic information of the patients was provided in Supplemental Table S1. Vascular segments with atherosclerotic lesions and control tissues were collected for further analysis. All specimens were obtained from Renji Hospital, School of Medicine, Shanghai Jiaotong University, and were processed in the Biobank of Renji Hospital. All procedures were approved by the research ethics committee of Renji Hospital (RA-2020-071).

| Immunofluorescence assays
Tissues were fixed in 4% paraformaldehyde, permeabilized in 0.25% Triton X-100 and blocked with 1% BSA for 1 hour at room temperature. Cells were then incubated with the primary antibody, washed with PBS with 0.1% Triton X-100 and incubated with FITC-conjugated secondary antibody (1:1000 in PBST containing 1% BSA). Tissue imaging was captured by Zeiss fluorescent microscope (Carl Zeiss).

| RNA in situ hybridization analyses
An Abi1-e10-specific RISH probe targeting the bp of Abi1-e10 mRNA, an Abi1-Δe10-specific RISH probe targeting the bp of Abi1-Δe10 mRNA and a negative control probe (targeting the dapB gene from bacteria) were designed. RISH assays were performed by using the QuantiGene ViewRNA ISH Assay Kit (Panomics) following the manufacture's protocol.

| Flow cytometry
Tissues were digested by collagenase and cells were suspended with 1X TrypLE Express (Gibco) individually, blocked in PBS (Sigma) containing 0.1% FBS (Gibco) and stained with fluorescent antibodies against CD68 or CD45. PBS-treated cells served as a control.

| Cell lines and transfection
Human primary aortic smooth muscle cells were purchased from American Type Medium (ATCC PCS-100-030) with Vascular Smooth Muscle Cell Growth Kit (ATCC PCS-100-04). Lipofectamine 3000 (Invitrogen) and SuperFect Transfection Reagents (Qiagen) were used for plasmid and siRNA transfection.

| PCR and Immunoblotting assays
Real-time qPCR assays were performed as described. 19 All real-time qPCR assays were carried out using three technical replicates and three independent cDNA syntheses. Primer information was listed in Supplementary Table S2. Western blotting assays were performed as reported. 19 For nuclear fraction assay, cytosol and nuclear protein lyse were extracted using NE-PER nuclear and cytoplasmic extraction kit (Life Technology). Antibody information was listed in Supplementary   Table S3. Experiments were repeated in three independent experiments, and one of the representative blots was shown.

| Cell proliferation and migration assays
Cell proliferation and migration assays were described previously. 19,20 Briefly, the cell proliferation assay was performed using the MTS (Promega) reagent according to the manufacture's protocol. Cell proliferation rates were calculated as relative fold change of OD450. For migration assays, a monolayer wound was created when cells reached 100% confluence. Cell migration was subsequently captured at time point 0 and 24 hours after wound scratch. The migration ability of cells was calculated as the migration distance from 0h to 24 hours.
Immunoblotting assays follow the standard protocol as reported. 20 Information on antibodies is listed in materials. When performing co-IP assays, cell lysates were extracted by NETN buffer containing 0.5% NP40, 1 mM of EDTA, 50mM of Tris, and 150mM of NaCl plus proteinase and phosphatase inhibitor (Roche). Pre-cleared lysates were incubated with indicated antibodies, and the associated proteins were immunoblotted by antibodies as indicated. Experiments were repeated at least three times, and one set of the representative blots was shown.

| Abi1-minigene construction and in vivo splicing assays
The human genomic BAC clone (RP11-75E16) was used as the template for PCR to amplify exon9, exon10, exon11, and their flanking intron regions (~300-400 base pairs) of Abi1 gene (NM_001012751) by Platinum Taq DNA Polymerase High Fidelity (Invitrogen). The Abi1-minigene with mutant was generated using the original Abi1minigene as the template. The integrity of the final construct was confirmed by DNA sequence.

| RNA pulldown assay
RNA pulldown assay was performed as previously described. Then the beads were washed with binding buffer DG and suspended in SDS loading buffer. Eluted proteins were analysed by Western blot and detected by Flag-tag antibody.

| RNA immunoprecipitation assays
RNA immunoprecipitation assay was performed as previous described. 20 Total RNA of VSMC cell transfected with Flag-MBNL1 plasmid was cross-linked with formaldehyde (10 mL PBS + 270 μL 1% formaldehyde) for 20 minutes at 37°C and sonicated in 300 μL buffer I (1% SDS, 10 mM EDTA, and 50 mM Tris, pH 8.0, plus protease inhibitor cocktail). After centrifugation, the supernatants were added to 2.7ml buffer II, and 30 μL of the mixture was saved as input.
For immunoprecipitation, 2 μg of Flag antibody or control IgG antibody was added to the purified chromatin sample and incubated overnight at 4°C. Immunocomplexes were precipitated by adding 50 μL of protein A/G agarose beads for 2 hours at 4°C with agitation.
Beads were washed sequentially for 5 minutes each in 1 mL of buffers III-VI, as described previously. 20 Immunocomplexes were eluted by adding 1600 μL of elution buffer (1% SDS and 0.1 M NaHCO3) and 50 μL RNase inhibitor to beads. 500 μL of eluted immunocomplexes was added in 10 μL 5 M NaCl and subsequently heated for 2 hours at 64°C to reverse formaldehyde-induced cross-links. RNA segments were isolated and collected by 1.5 mL lysis buffer with 15 μL 2-mercaptoenthanol using Purelink RNA Mini Kit (Ambion) according to the manufacture's instruction and subsequence to reverse transcription and analysed by real-time qPCR as described above. Data were calculated as a percentage of input.

| Lentivirus production, infection and generation of stable cell lines
To generate VSMC lines stably expressing Abi1-e10 and1Abi1-Δe10, the Abi1-e10 and Abi1-Δe10 cDNA were cloned into the FU(GW) BW-blasticidin-resistant lentivirus vector (purchased from Addgene) using Gateway Technology (Invitrogen) according to the manufacture's instruction. 293T packaging cells were co-transfected with viral vectors, VSV-G-encoding plasmid and pCMV R8.9. The supernatants of transfected 293T cells containing viruses were used to infect VSMC cells in the presence of 10%FBS. The infected LNCaP cells were selected in the presence of 5μg/ml blasticidin. Expression of Abi1-e10 and Abi1-Δe10 was confirmed by both Western blot and real-time qPCR.

| ROS detection
In cultured VSMC, the ROS levels were determined immediately after sample collection. Cellular ROS levels were assessed by measuring CM-H2DCFDA (Thermo Fisher) fluorescence. All values were normalized to VSMC protein concentration.

| Statistics
Statistical analysis was carried out by the GraphPad Prism 8.0 software. Differences among groups were compared by Student's t test.
The level of significance was set at P <.05 as *, P <.01 as **.

| RNA splicing of Abi1-Δe10 is upregulated in macrophage-like VSMC-derived from atherosclerotic tissue
It has been reported that vascular smooth muscle cells (VSMC) will transdifferentiate into a macrophage-like phenotype (VSMC-M) during atherogenesis. To identify the phenotype of VSMC with the use of cell-specific markers, we first isolated the intima of control and atherosclerotic artery as described in the Material and Methods section. By double immunofluorescence staining for macrophage biomarker CD68 and smooth muscle biomarker αSMA, we found that the amount of CD68+ VSMC was significantly increased in the thickened intima of ASO arteries ( Figure 1A). To evaluate the origin of these CD68+ cells, we also applied double immunofluorescence staining for CD68 and myeloid biomarker CD45. Among tissue cells in the intima of ASO arteries, about 50% of them were CD68+/CD45−, but only 20.8% were CD68+/CD45+, indicating that the majority of CD68+ cells were not originated from monocyte ( Figure S1). These results demonstrated the upregulation of VSMC-M in the hyperplastic intima. We further isolated and se-

| Loss of MBNL1-induced RNA splicing of Abi1-Δe10 in VSMC macrophage-like VSMC
Splicing of Abi1 has been reported to be associated with several splicing factors. We profiled the mRNA level of these factors and found that the change of MBNL1 mRNA level was the most significant between control and atherosclerosis arteries (Figure 2A). This result was validated by immunoblotting showed that MBNL1 protein was downregulated in ASO intima ( Figure 2B). These findings were consistent with both MNBL1 mRNA and protein expression was dramatically decreased in VSMC-M ( Figure 2C). In vivo immunofluorescences results showed decreased MBNL1 level was accompanied with increased CD68 level ( Figure 2D). Together, these results establish a correlation of MBNL1 suppression with Abi1-Δe10 splicing.
To confirm that MBNL1 regulates Abi1 splicing, we constructed an Abi1 minigene reporter in which exon10 and is flanking ~ 300bp nucleotides were inserted between exons 9 and 11 and transfected into VSMC. MBNL1 depletion upregulated Abi1-Δe10 but downregulated Abi1-e10 mRNA derived from the minigene reporter ( Figure 2E-F). In vivo RNA binding assays showed that MBNL1 was recruited to the region next to the 3' splice site of Abi1 intron 9 (P1 region), but not the control P2 region ( Figure 2G). The UGCU motif was predicted to be a consensus MBNL1 recognition site and RNA pulldown assays confirmed that MNBL1 protein from VSMC interacted with the wild type, but not the mutant UGCU motif within the intron 9 ( Figure 2H). Site-directed mutagenesis (UGCU to UAUU) within the Abi1 minigene showed a failure of MBNL1-mediated exon10 inclusion ( Figure 2I). These results confirmed that MBNL1 regulates Abi1-Δe10 splicing.

| Loss of MBNL1 promotes phenotype modulation of VSMC through upregulating Abi1-Δe10 isoform
Because MBNL1 is strongly expressed in VSMC but barely expressed in VSMC-M, we introduced shRNA targeting MBNL1 into VSMC by lentivirus to study functions of MBNL1 ( Figure S2A-B).  Both G-LISA Rac1 activation assay and PAK-CRIB pulldown assay showed that the VSMC(Abi1-e10) had lower levels of activated Rac1 than did the VSMC(Abi1-Δe10) ( Figure 4A-B). Co-IP assays indicated that Abi1-Δe10 had a higher affinity to Rac1 than Abi1-e10 ( Figure 4C). Rac1 is known to localize to the actin-based membrane to exert its function upon upstream stimulation. We applied immunofluorescence microscopy to show that Rac1 was only localized in cytoplasm in VSMC(Abi1-e10), while translocated into the membrane in VSMC(Abi1-Δe10) ( Figure 4D). These results were confirmed by immunoblotting assays using cytoplasm and membrane fractions of VSMC cells transfected with Abi1-e10 and Abi1-Δe10 reporter was co-transfected with siRNA targeting MBNL1 in VSMC cells. Total RNA was extracted to measure Abi1-e10 and Abi1-Δe10 mRNA levels by real-time qPCR. (G) A schematic diagram shows the P1 and P2 regions used in in vivo RNA binding assays. RNA protein complexes in VSMC cells were cross-linked by formaldehyde and immunoprecipitated with control or Flag antibody. Eluted RNA fragments were used as templates for real-time qPCR to amplify the P1 and P2 regions. Signals were calculated as the percentage of input. (H) Biotin-labelled RNA oligos containing wild-type and mutant UGCU motifs were incubated with streptavidin-conjugated beads. They were used to pull down protein extracts from VSMC cells. Proteins associated with streptavidin beads were eluted and immunoblotted with the MBNL1 antibody. (I) VSMC cells were transfected with control, Abi1-minigene or Abi1-minigene with MT1/MT2 mutations in the presence of ∓ siMBNL1 RNA. Total RNA was collected and used to measure Abi1-e10/Abi1-Δe10 mRNA levels by real-time qPCR. All immunoblotting, real-time qPCR and RNA binding assays were repeated in three independent experiments that were performed in triplicate. Data were presented as the mean ± SD ** delegates P <.01 when comparing with controls. One-way ANOVA followed by Tukey test was used in pairwise comparison among different groups that activating the Rac1 signalling to promote VSMC macrophagelike transdifferentiation.

| Abi1-Δe10 activates the Rac1-NOX1-ROS signalling to active KLF4
Recent evidence extends VSMC plasticity to transdifferentiation into macrophage-like cells during atherogenesis that is dependent on KLF4, a transcription factor that induces cell dedifferentiation.

Rac1 may promote VSMC transdifferentiation through modulating
the expression and function of KLF4. Specifically, Rac1 activates the NADPH oxidase catalytic subunits 1 (NOX1) which in turn increases reactive oxygen species (ROS) production to generate oxidized phospholipids and induced KLF4 expression ( Figure 5A). We showed that Abi1-Δe10 stimulated ROS production, which effect was inhibited by Rac1 and NOX1 inhibition ( Figure 5B). Co-immunoprecipitation assays confirmed that Abi1-Δe10 promotes active Rac1 interacted with NOX1 in VSMC ( Figure 5C). Overexpression of Abi1-Δe10 increased both KLF4 expression and nuclear localization ( Figure 5D-E). This function of Abi1-Δe10 could also be attenuated by NOX1 inhibition (Figure 5F-G). These results revealed a novel action of Abi1-Δe10 that activates Rac1 and enhance KLF4 function via the Rac1-NOX1-ROS axis.

| D ISCUSS I ON
In this manuscript, we report a novel action of Abi1 that contrib- Abi1 is a known adapter protein that has been implicated in actin cytoskeletal remodelling, intercellular adhesion, and smooth muscle migration and contraction upon upstream stimulation. 13,14,16 However, no study has reported the function difference among  Expression of KLF4 can be induced by oxidized phospholipids. 10,11 Our results demonstrate that Abi1-Δe10 engages not only ROS production but also lipid intake through macropinocytosis, both of which promote oxidized phospholipids production. The connection between KLF4 and the macrophage marker expression comes from studies showing that it is a required factor in monocyte differentiation. Thus, it is tempting to speculate that the increased expression of KLF4 after Abi1-Δe10 overexpression participates not only in the loss of the contractile phenotype but also in the assumption macrophage features in VSMC-M.

CO N FLI C T O F I NTE R E S T
The authors declared they do not have anything to disclose regarding conflict of interest with respect to this manuscript.

AUTH O R CO NTR I B UTI O N S
Zhang L involved in concept and design. Li Y, Guo X and Xue G in-

CO N S E NT FO R PU B LI C ATI O N
All authors of the manuscript have read and agreed to its content and are accountable for all aspects of the accuracy and integrity of the manuscript in accordance with ICMJE criteria. The manuscript or portions thereof are not under consideration by any other journal that these findings have not been previously published.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.