Abstract

   The Treponema pallidum membrane protein Tp0136 facilitates Treponema
   pallidum dissemination, and the permeability of vasculature is
   intricately linked to the density of the vascular endothelial barrier,
   which is strongly associated with the morphology of vascular
   endothelial cells. In this study, utilizing the approach of inoculating
   Tp0136 recombinant protein into the skin lesions of rabbits infected
   with Treponema pallidum, it was observed that the Tp0136 recombinant
   protein induced spheroidization of vascular endothelial cells and
   enlargement of intercellular junctions. By high‐throughput RNA
   sequencing, the upregulation of the RNA‐binding protein cysteine‐ and
   glycine‐rich protein 1 (CSRP1) is identified, which modulated the
   alternative splicing of exon 19 of myosin X (MYO10), which in turn
   downregulated the expression of MYO10, ultimately inducing
   morphological spheroidization in vascular endothelial cells. Using
   CSRP1‐specific shRNA to knock down CSRP1 or using the alternative
   splicing inhibitor, the spheroidized vascular endothelial cells revert
   to a flattened state, suggesting that Tp0136 regulates the alternative
   splicing of MYO10 through CSRP1, leading to a downregulation of MYO10,
   followed by the spheroidization of vascular endothelial cells and an
   enlargement of intercellular junctions. These findings contribute to
   elucidating a mechanism underlying the dissemination of Treponema
   pallidum.

   Keywords: alternative splicing, cell spheroidization, permeability,
   Treponema pallidum protein Tp0136 , vascular endothelial cells
     __________________________________________________________________

   Treponema pallidum protein Tp0136 enhances vascular permeability by
   upregulating CSRP1, which induces MYO10 exon 19 skipping, reducing
   MYO10 expression and triggering endothelial spheroidization. This study
   reveals a novel RNA splicing‐mediated dissemination mechanism,
   proposing CSRP1/MYO10 as therapeutic targets to limit bacterial
   dissemination in syphilis.

   graphic file with name SMSC-5-2500046-g004.jpg

1. Introduction

   The clandestine dissemination of Treponema pallidum, which is the
   causative agent of syphilis, takes place covertly within the human
   organism.^[ [34]^1 ^] Prior research has revealed that Treponema
   pallidum promote vascular permeability, thereby facilitating the
   dissemination of the pathogen throughout the bloodstream.^[ [35]^2 ,
   [36]^3 , [37]^4 ^] Our previous researches have provided evidences that
   the Treponema pallidum membrane protein Tp0136 plays a crucial role in
   the infection process of Treponema pallidum. Specifically, Tp0136 has
   been found to activate the PI3K‐AKT pathway in endothelial cells,
   thereby facilitating cell migration and promoting angiogenesis. This
   mechanism ultimately aids in the dissemination of Treponema pallidum.^[
   [38]^5 ^] However, whether Tp0136 directly induces morphological
   alterations in vascular endothelial cells—such as spheroidization and
   intercellular junction widening—through dysregulation of cytoskeletal
   regulatory proteins (e.g., myosins) and the underlying molecular
   mechanisms driving these changes remain unexplored. Nevertheless, the
   precise mechanism responsible for its propagation remains elusive.

   Numerous studies have demonstrated a correlation between alterations in
   vascular permeability and the density of endothelial cell arrangement
   comprising the blood‐tissue barrier.^[ [39]^6 , [40]^7 , [41]^8 ^] The
   morphology of vascular endothelial cells plays a direct role in
   regulating the density of the blood‐tissue barrier.^[ [42]^9 , [43]^10
   ^] The regulation of endothelial barrier permeability during microbial
   infection and dissemination is significantly impacted by
   morpho‐maintenance proteins, which assume a pivotal role.^[ [44]^11 ,
   [45]^12 , [46]^13 ^] Myosin plays an irreplaceable role in maintaining
   cell morphology by regulating the contraction and extension of
   cytoskeletal proteins.^[ [47]^14 ^] A multitude of pathogenic microbes
   have been observed to disrupt the cytoskeletal organization of host
   vascular endothelial cells via obstructing the myosin synthesis, such
   as Toxoplasma gondii and Shigella flexneri,^[ [48]^15 , [49]^16 ^]
   thereby facilitating their dissemination throughout the organism.^[
   [50]^17 , [51]^18 , [52]^19 ^] Treponema pallidum protein Tp0136 has
   been observed to stimulate the migration of fibroblast cells,^[ [53]^20
   ^] which is accompanied by notable alterations in cell morphology.
   However, it remains unresolved regarding whether Tp0136 is responsible
   for morphological alterations through interfering the synthesis of
   myosin(s) in vascular endothelial cells, as well as the underlying
   mechanism that drives these changes.

   In the present study, the vascular structure of syphilis rabbit skin
   lesions was scanned in the presence of the Tp0136 recombinant protein.
   The impact of Tp0136 protein on parameters of vascular endothelial cell
   line human microvascular endothelial cells was measured using an
   in vitro 3D blood vessel model, and the permeability of these vessels
   was assessed. High‐throughput RNA‐seq was utilized to analyze the
   alteration of the transcriptional level in mRNA expression and
   alternative splicing of vascular endothelial cells after exposure to
   Tp0136 protein. The following findings confirmed that Tp0136 protein
   regulates the alternative splicing of MYO10 through CSRP1, leading to a
   downregulation of MYO10, and induces spheroidization of vascular
   endothelial cells, resulting in widened intracellular junctions and
   enhanced vascular permeability. This finding provides a fresh
   perspective for further exploration of the dissemination mechanisms of
   Treponema pallidum.

2. Results

2.1. Tp0136 Spheroidized the Vascular Endothelial Cells, Enlarged the
Intercellular Junctions, and Facilitated the Permeability

   To investigate the role of Tp0136 in the blood vessels, the electron
   microscope photographs of blood vessels in the skin lesions of
   syphilitic rabbit model were collected, referencing our previous
   study.^[ [54]^5 ^] Briefly, we collected the lesions at the peak of
   ulceration (day 18 of infection with Treponema pallidum) and inoculated
   the remaining lesions with Tp0136 protein (marked as Tp0136‐day 0). The
   Tp0136 protein was then inoculated every three days with PBS as a
   control. Then, skin lesions were collected on the 36th (marked as
   Tp0136‐day 18) and 42nd (marked as Tp0136‐day 24) days, and along with
   the lesions from the Tp0136‐day 0, these lesions were scanned using
   electron microscopy. It showed that, on Tp0136‐day 0, the perivascular
   morphologies in both the PBS group and the 10 μg Tp0136 protein group
   were indistinct, with zoomed‐in electron microscopy scanning revealing
   large gaps in the junctions of endothelial cells, which was consistent
   with the morphological characteristics of blood vessels around the
   wound of other infectious diseases.^[ [55]^21 ^] Subsequently, as the
   duration of Tp0136 protein stimulation increased, the clarity of blood
   vessel boundaries in both groups improved by the Tp0136‐day 18 and
   Tp0136‐day 24, relative to observations made on the Tp0136‐day 0.
   High‐magnification scanning electron micrographs revealed a progressive
   reduction in the gaps between vascular endothelial cells in both
   groups. However, the intracellular gaps between endothelial cells in
   the Tp0136 protein‐stimulated group remained unclosed, whereas the
   endothelial connections in the PBS control group had fully closed by
   the Tp0136‐day 18 and Tp0136‐day 24 (the intracellular junction gaps
   were point‐out by purple arrows). Notably, it was observed that under
   the exposure of Tp0136 protein, the morphology of vascular endothelial
   cells seemed more spherical compared to the vascular endothelial cells
   in the PBS group at the same time points (Figure [56] 1A). These
   in vivo vascular results suggested that Tp0136 protein delayed the
   closure of gaps between vascular endothelial cells and may promote the
   spheroidization of vascular endothelial cells.

Figure 1.

   Figure 1
   [57]Open in a new tab

   Tp0136 induces spheroidization of vascular endothelial cells, expands
   intercellular junctions, and enhances permeability. A) Effect of Tp0136
   protein on the morphologies and intercellular junctions of vascular
   endothelial cells in the skin lesions of syphilitic rabbits with
   Scanning Electrical Microscopy. The gray bars represent the days after
   infection with Treponema pallidum. The green bars represent the days
   since Tp0136 protein injection. The intracellular junction gaps were
   indicated by purple arrows. Scale bars = 30 μm. B) Schematic drawing
   for 3D microfluidic angiogenesis system of angiogenesis processed by
   Tp0136. C) Growth of HMEC‐1 cells in a 3D microfluidic angiogenesis
   system. Upper: Photographs of the cells in the perfusion channel with
   different concentrations of Tp0136 in X‐Y panel and X‐Z panel. PBS was
   the negative control. Scale bars = 100 μm. Lower: The ratio of cell
   coverages in X‐Y panel, average of cell thickness, and average of cell
   width in X‐Z panel along with the rising concentration of Tp0136, and
   cell number calculated by DAPI number. The values represent the
   mean ± SD of triplicates and are representative of three independent
   experiments. ***, P < 0.001. D) Effect of Tp0136 on vascular
   permeability detected using 3D angiogenesis analysis. The red
   fluorescent images after 0 and 8 min after the addition of the
   TRITC‐Dextran solutions. Time is indicated in hour:minute:second. E)
   Cell number counted by flow cytometry, and viability of HMEC‐1
   evaluated by CCK‐8 assay in 3D‐microfluid system.

   To further evaluate the impact of Tp0136 on cell morphology, we
   utilized a 3D microfluidic blood vessel system^[ [58]^22 ^] to culture
   human microvascular endothelial cells (HMEC‐1) (Figure [59]1B), aiming
   to replicate the blood vascular structure observed in vivo. After the
   cells covered the lumen for 4 days, PBS and Tp0136 protein were exposed
   to the cells for 6 h. In the Tp0136 group, a gradual decrease in the
   cell coverage rate was observed as the concentration of Tp0136 protein
   increased (P < 0.001), which was 75% of cell coverage rate in cells
   treated with 5 μg Tp0136 protein, and declined to 50% in 10 μg of
   Tp0136 in X‐Y panel. Conversely, in the PBS group, the cell arrangement
   exhibited a dense pattern, with a cell coverage reaching 100%.
   Furthermore, in the X‐Z plane, the lumen diameter decreased as the
   Tp0136 concentration increased, with the vascular endothelial cells on
   the lumen wall growing as a monolayer, indicating an increase in the
   thickness of the vascular endothelial cells, which is a characteristic
   of cell spheroidization (P < 0.001). Concurrently, with the elevation
   of Tp0136 concentration, the intercellular space between vascular
   endothelial cells widened (P < 0.001, Figure [60]1C). Then, the
   permeability was measured with TRITC‐Dextran,^[ [61]^23 ^] which was
   labeled in red under the confocal microscope. It showed that the
   diffusion rate of TRTC‐Dextran in the Tp0136 group was significantly
   higher compared to the PBS group. Furthermore, treatment of
   vasculature‐like structures with 10 μg of Tp0136 protein resulted in a
   diffusion distance of 0.4 mm within 8 min, whereas a Tp0136
   concentration of 5 μg limited the diffusion to 0.22 mm. These findings
   suggested a positive correlation between Tp0136 concentration and
   diffusion rate, indicating that higher concentrations of Tp0136
   enhanced the diffusion rate (Figure [62]1D). To rule out the
   possibility that the increased permeability observed in the 3D
   microfluidic system following Tp0136 treatment is attributable to a
   decrease in the quantity or viability of vascular endothelial cells, we
   prepared a single‐cell suspension from the cellular enzymes within the
   3D microfluidic system. We then quantified the number of HMEC‐1 cells
   under both PBS and Tp0136 treatment conditions using flow cytometry.
   The results indicated no significant difference in cell numbers between
   the Tp0136 treatment group and the PBS control group (P > 0.05,
   Figure [63]1E). This finding was corroborated by fluorescence
   microscopy, which also revealed no significant difference in the number
   of DAPI‐stained cells within the same field of view (P > 0.05,
   Figure [64]1C). Furthermore, the viability of HMEC‐1 cells was assessed
   via the cell counting kit (CCK‐8) assay in the 3D microfluidic system.
   We observed no significant difference in cell viability between the
   Tp0136 and PBS groups (P > 0.05, Figure [65]1E), suggesting that the
   increase in vascular permeability induced by the Tp0136 protein is not
   due to a reduction in cell number or viability. Above all, these
   results indicated that Tp0136 contributes to spheroidizing vascular
   endothelial cells, expanding intercellular junctions, and facilitating
   the permeability both in vivo and in vitro.

   Student's t test or nonparametric rank‐sum test was used to compare
   data between two groups. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001.

2.2. Tp0136 Upregulated the Expression of Cysteine and CSRP1 to Enlarge the
Intercellular Junctions in HMEC‐1 Cell Line

   In order to examine the effects of Tp0136 treatment on gene expression
   patterns of vascular endothelial cells, HMEC‐1 cells were treated with
   10 μg of Tp0136 protein and PBS for a duration of 6 h; then, RNAs were
   extracted and subjected to high‐throughput transcriptome RNA
   sequencing. Subsequently, we conducted calculations from RNA sequencing
   data to ascertain the normalized deviations of expression for a total
   of 5,457 genes that exhibited significant differential expression
   (enriched by P < 0.001). The subsequent heatmaps effectively
   illustrated that the transcriptome from the cells in Tp0136 group
   displayed discernible patterns in comparison to the cells in PBS group
   for 6 h (Figure [66] 2A). To obtain a thorough comprehension of the
   functional consequences of these genes that were expressed
   differentially in response to Tp0136, we classified them into Gene
   Ontology (GO) categories utilizing the DAVID database
   (http://david.abcc.ncifcrf.gov/).^[ [67]^24 ^] The analysis revealed
   that these genes exhibited significant enrichment in Biological Process
   terms pertaining to RNA splicing and maintenance of cell morphology, as
   well as in Molecular Function terms associated with RNA binding and
   support structure for cell morphology maintenance (Figure [68]2B).

Figure 2.

   Figure 2
   [69]Open in a new tab

   Tp0136 upregulated the expression of CSRP1 to enlarge the intercellular
   junctions in HMEC‐1 cell line. A) Heatmap of the normalized deviations
   of expression for a total of 5,457 differentially expressed genes
   (enriched by P < 0.001). B) GO analysis (GOTERM_BP_DIRECT category and
   GOTERM_MF_DIRECT category) for the overall differentially expressed
   genes. C) Gene ontology analysis of RBPs with different expression
   levels after Tp0136 protein treatment. Left: GO circle for the function
   annotation of the 146 RBPs; Right: List of the information involved in
   GO:0 060 501. D) Fold changes and qPCR validation of 9 cell morphology
   maintenance RBPs. E) The protein level of CSRP1 during the Tp0136
   treatment in HMEC‐1 cell line. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001. F) The protein level of CSRP1 after CSRP1 targeted
   shRNA treatment. The values represent the mean ± SD of triplicates and
   are representative of three independent experiments. ***, P < 0.001. G)
   Immunofluorescent staining of cytoskeleton by Phalloidin to calculate
   the cell coverage area in X‐Y panel, and average cell thickness or cell
   width in X‐Z panel after shRNA treatment with or without Tp0136
   addition. The values represent the mean ± SD of triplicates and are
   representative of three independent experiments. ***, P < 0.001, ###,
   P < 0.001, ns, P > 0.05. H) Effect of CSRP1 on Tp0136‐induced vascular
   permeability in 3D microfluidic angiogenesis system. Time is indicated
   in hour:minute:second.

   RNA‐binding proteins (RBPs) play a pivotal role in regulating
   alternative splicing events and are intricately linked to the
   preservation of cellular morphology, serving as crucial regulators of
   vascular barrier function.^[ [70]^13 , [71]^25 , [72]^26 , [73]^27 ,
   [74]^28 ^] Following a 6‐h treatment of HMEC‐1 cells with 10 μg Tp0136
   protein, a total of 146 RBPs with P < 0.001 and fold change ≥1.5 were
   identified. In order to enhance the precision of the functional
   clustering of the 146 RBPs that were sorted based on a significance
   level of P < 0.001 and a fold change exceeding 1.5 in the expression
   data, the utilization of the GO circle was used to analyze their
   distinct functions (Figure [75]2C, Left). Among these RBPs, 9 RBPs
   (CSRP1, PTBP2, MYEF2, SRSF6, SNTB2, SUAU2, PAN3, KIF1C, and MACF1)
   exhibited a high frequency of clustering specifically related to the
   maintenance of cell morphology (Figure [76]2C, Right). To identify the
   most functionally relevant RBPs, a sorting process based on fold
   changes was conducted on the 9 RBPs. Among the 9 RBPs, CSRP1, SRSF6,
   and KIF1C were found to be upregulated, while the remaining 6 RBPs were
   downregulated after Tp0136 treatment. Furthermore, only CSRP1, PTBP2,
   and MYEF2 exhibited expression fold changes greater than 2. CSRP1,
   which was considered to regulate the permeability of blood vessels,^[
   [77]^29 ^] was demonstrated with the highest fold change (fold
   change = 2.27) among these 9 RBPs (Figure [78]2D), was then followed
   with the confirmation of the protein level, which showed a significant
   increasing along the 10 μg Tp0136 treatment for 6 h (Figure [79]2E,
   P < 0.001), and CSRP1 was subsequently selected for further
   investigations.

   To clarify the relationship between CSRP1 and cell morphology, CSRP1
   sequence specific shRNA was used to knock down CSRP1 in HMEC‐1 cells to
   generate the sh‐CSRP1 HMEC‐1 cell line, with the mock shRNA was used to
   generate sh‐Ctl HMEC‐1 cell line (Figure [80]2F). These cell lines were
   seeded in the 3D microfluidic angiogenesis system with Tp0136 protein
   for 4 days (PBS as control), and stained with phalloidin for cell
   morphology checking (Figure [81]2G). It showed that following the
   elimination of CSRP1, the cell coverage area in the X‐Y panel returned
   to 100%, despite the presence of Tp0136 protein, aligning with the
   coverage observed in the sh‐Ctl group. In contrast, the cell coverage
   area in the sh‐Ctl + Tp0136 group was limited to 50%. Additionally, in
   the X‐Z panel, compared to the spheroidized and loosely arranged cells
   in the sh‐Ctl + Tp0136 group, we observed that upon the knocked out
   CSRP1, the cells exhibited a flat and dense arrangement, even in the
   presence of Tp0136. Subsequent permeability assessments indicated a
   significant reduction (P < 0.001) in the permeability of HMEC‐1 cells
   following CSRP1 depletion, irrespective of the presence of the Tp0136
   protein, when compared to sh‐Ctl HMEC‐1 cells treated with 10 μg of
   Tp0136 protein, and the permeability of HMEC‐1 cells post‐CSRP1
   knockout reverted to levels comparable to those observed in sh‐Ctl
   HMEC‐1 cells (Figure [82]2H). These outcomes implied that Tp0136
   upregulates the expression of CSRP1 to induces bulging of cellular
   morphology and results in enlarged intercellular gaps, ultimately
   leading to enhanced vascular permeability.

   One‐way ANOVA was used for comparisons between multiple groups and for
   comparisons between two groups. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001, ###, P < 0.001, ns, P > 0.05.

2.3. Tp0136 Mediated Alternative Splicing of Myosin X (MYO10) Through CSRP1

   To investigate potential alternative splicing events associated with
   CSRP1, the proportion of alternative events influenced by CSRP1 was
   calculated from the RNA sequencing data firstly. Among the alternative
   splicing data for the skipping‐exon event, 881 genes were enriched by
   CSRP1. A comprehensive analysis revealed the identification of 881
   genes, accounting for 28.2% of the total alternative splicing events.
   Afterwards, we conducted a GO analysis on 881 genes enriched with CSRP1
   that were associated with skipping‐exon events. In the Biological
   Process category, these genes primarily cluster around terms pertaining
   to cell morphological support, such as cytoskeletal organization, and
   regulatory protein expression, including protein stabilization. These
   suggested that these genes with alternative splicing may be linked to
   the regulation of cell morphology through cytoskeleton protein
   expression. In the Cellular Component category, these genes were mainly
   enriched in terms like cell–cell junction and microtubule organization,
   indicating that these alternative splicing genes may aid in supporting
   cell morphology within cells through the internal skeleton and
   stretched to maintain cell morphology. The aggregation of terms such as
   cytoskeletal protein binding and other related terms in Molecular
   Function implies that genes with alternative splicing may be necessary
   for regulating cell morphology by binding to skeleton proteins in
   molecular biological functions (Figure [83] 3A). To clarify the
   expression levels of these genes enriched by CSRP1 that undergo
   skipping‐exon event and are closely related to cell morphology and
   cytoskeleton, a Venn diagram analysis was performed to compare the 881
   genes (in pink) with those showing significant expression alterations
   in response to Tp0136 treatment. The genes downregulated by Tp0136
   (downregulated genes after Tp0136 treatment: 1151 genes) were shown in
   green, and the genes upregulated (up regulated genes after Tp0136
   treatment: 451 genes) were shown in blue. It revealed that 192 genes
   fulfilled the criteria of being enriched by CSRP1 and exhibiting a
   decrease in expression following Tp0136 treatment, whereas only 3 genes
   met the conditions of CSRP1 enrichment and demonstrated an increase in
   expression upon Tp0136 treatment (Figure [84]3B). Then, a GO analysis
   was conducted on the 192 genes, and 42 genes exhibiting
   cytoskeleton‐related terms were chosen for further Sankey analysis, and
   MYO10 was identified as the predominant protein present in the majority
   of terms associated with cytoskeletal organization (Figure [85]3C).

Figure 3.

   Figure 3
   [86]Open in a new tab

   Tp0136 mediates alternative splicing of MYO10 through CSRP1. A) GO
   analysis of the 881 genes that enriched in both CSRP1 binding RNAs and
   skipping‐exon events. B) Venn diagram of CSRP1 binding RNAs in pink,
   downregulated genes after Tp0136 treatment in green, and upregulated
   genes after Tp0136 treatment in blue. C) Sankey map and GO bubble map
   depict the downregulated genes which were actin cytoskeleton related
   and were CSRP1 bind after Tp0136 treatment. D) CSRP1 protein‐MYO10 mRNA
   immunoprecipitation. The values represent the mean ± SD of triplicates
   and are representative of three independent experiments. ***,
   P < 0.001, ###, P < 0.001, ns, P > 0.05. E) MISO algorithm analysis of
   the MYO10 alternative splicing. F) Validation of the alternative
   splicing variants switch in MYO10 exon19 after Tp0136 protein
   treatment. Upper: Alternative splicing variants detection of MYO10
   exon19 in high resolution agarose gel; Lower: Alternative splicing
   ratio of the MYO10 exon 19 calculated from high resolution agarose gel
   band grayscale. The values represent the mean ± SD of triplicates and
   are representative of three independent experiments. ***, P < 0.001. G)
   Validation of the alternative splicing variants switch in MYO10 exon19
   in sh‐Ctl HMEC‐1 and sh‐CSRP1 HMEC‐1 cell lines with/without Tp0136
   protein treatment. Upper: Alternative splicing variants detection of
   MYO10 exon19 in high resolution agarose gel; Lower: Alternative
   splicing ratio of the MYO10 exon 19 calculated from high resolution
   agarose gel band grayscale.

   To confirm the interaction between the CSRP1 protein and MYO10 mRNA,
   RNA‐protein immunoprecipitation was used firstly. The findings
   demonstrated that CSRP1 protein was pulled down with beads covered with
   anti‐CSRP1 antibody and the level of enrichment increasing in
   correlation with the duration of Tp0136 protein exposure. Concurrently,
   we assessed the MYO10 mRNA released from CSRP1 protein by qPCR, and the
   results showed that MYO10 mRNA was pulled down in conjunction with the
   enrichment of CSRP1 protein (P < 0.001), with the enrichment levels of
   both MYO10 mRNA and CSRP1 protein exhibiting a corresponding increase
   as the duration of Tp0136 protein exposure was extended (P < 0.01,
   Figure [87]3D), confirming the CSRP1 protein binds to MYO10 mRNA. To
   elucidate the alternative splicing site of MYO10, a multiple‐input
   single‐output analysis was conducted on the alternative splicing data
   of MYO10, revealing that exon 19 underwent skipping following a 6‐h
   treatment with Tp0136 (Figure [88]3E). Subsequently, exon‐19 specific
   alternative splicing primers were used to confirm the alternative
   splicing pattern of MYO10 in the presence of Tp0136. The results
   demonstrated that with prolonged treatment of the Tp0136 protein, the
   inclusion rate of exon 19 included variant of MYO10 (included isoform:
   PCR product length 242 bp) among all alternative splicing variants
   (included+skipped) progressively declined from nearly 100% to 50%.
   Conversely, the exclusion rate of exon 19 in the MYO10 gene (skipped
   isoform: PCR product length 160 bp) correspondingly increased from less
   than 1% to ≈50% (P < 0.001, Figure [89]3F). To investigate the role of
   CSRP1 in the alternative splicing of MYO10 exon 19, an analysis was
   undertaken focusing on the splicing pattern of MYO10 exon 19 within the
   sh‐CSRP1 cell line. In comparison to the sh‐Ctl + Tp0136 group, where
   the exon 19 included variant of MYO10 constituted 50% of the total, the
   elimination of CSRP1 resulted in a significant increase in the
   proportion of the exon 19 included variant of MYO10 (P < 0.001), which
   approached or exceeded 80%, irrespective of the presence of the Tp0136
   protein. This finding suggests that the CSRP1 elimination leads to a
   substantial enhancement of the exon 19 included variant of MYO10, while
   concurrently abrogating the Tp0136‐induced exon 19 skipped variant
   (Figure [90]3G). These results suggested that Tp0136 mediated the MYO10
   exon 19 skipping through CSRP1.

   One‐way ANOVA was used for comparisons between multiple groups and for
   comparisons between two groups. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001, ###, P < 0.001, ns, P > 0.05.

2.4. Tp0136 Downregulated the Protein Level of MYO10 Through CSRP1‐Mediated
Alternative Splicing to Enlarge the Intercellular Junctions, Resulting in the
Enhancing Permeability

   To corroborate the regulation of MYO10 expression by Tp0136 protein,
   the correlation between MYO10 protein expression and Tp0136 treatment
   duration was investigated in HMEC‐1 cell line. The expression of MYO10
   was observed to decrease with the prolongation of Tp0136 protein
   treatment time (P < 0.001, Figure [91] 4A). Furthermore, the
   association between MYO10 protein expression patterns and those of
   CSRP1 protein was analyzed. It showed that upon silencing CSRP1, the
   expression level of MYO10 remained unaffected by Tp0136 (P > 0.05), and
   MYO10 expression reverted to the level observed in the sh‐Ctl group
   when compared to the sh‐Ctl+Tp0136 group (P < 0.001), suggesting that
   the expression pattern of MYO10 was inversely correlated with that of
   CSRP1 (Figure [92]4B). Immunofluorescence staining with MYO10 in a 3D
   microfluidic angiogenesis system revealed an enhancement in the
   expression of MYO10 (P < 0.001) accompanied with cell coverage area of
   HMEC‐1 cells (P < 0.001) upon the knockdown of CSRP1 compared to those
   in sh‐Ctl + Tp0136 group (Figure [93]4C), providing further evidence
   that the expression of MYO10 and cell coverage area were modulated by
   the expression of CSRP1.

Figure 4.

   Figure 4
   [94]Open in a new tab

   Tp0136 switched the alternative splicing pattern of MYO10 to
   downregulate the protein level through CSRP1, and enlarges the
   intercellular junctions. A) Protein level of MYO10 treated with Tp0136
   protein. The values represent the mean ± SD of triplicates and are
   representative of three independent experiments. ***, P < 0.001. B)
   Protein levels of MYO10 and CSRP1 after CSRP1 knocked down and Tp0136
   treatment. The values represent the mean ± SD of triplicates and are
   representative of three independent experiments. ***, P < 0.001. C)
   Photographs of the cells in the 3D microfluidic angiogenesis system to
   calculate the cell coverage and MYO10 fluorescence. The values
   represent the mean ± SD of triplicates and are representative of three
   independent experiments. ***, P < 0.001, ###, P < 0.001, ns, P > 0.05.
   D) Switch of MYO10 exon19 alternative splicing variants after
   alternative splicing inhibitor SSA treatment, Tp0136 protein treatment,
   and CSRP1 knock‐down. The values represent the mean ± SD of triplicates
   and are representative of three independent experiments. ***,
   P < 0.001, ###, P < 0.001, ns, P > 0.05. E) Protein levels of MYO10 and
   CSRP1 after SSA treatment, Tp0136 protein treatment, and CSRP1
   knock‐down. The values represent the mean ± SD of triplicates and are
   representative of three independent experiments. ***, P < 0.001, ###,
   P < 0.001, ns, P > 0.05. F) Immunofluorescent staining of cytoskeleton
   by Phalloidin in the 3D microfluidic angiogenesis system to calculate
   the cell coverage area in X‐Y panel, and average cell thickness or cell
   width in X‐Z panel after SSA treatment with or without Tp0136
   treatment. Scale bars = 100 μm. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001, ###, P < 0.001, ns, P > 0.05. G) Effect of SSA on
   Tp0136‐induced vascular permeability in 3D microfluidic angiogenesis
   system. Time is indicated in hour:minute:second.

   Alternative splicing exhibits a close correlation with protein
   expression levels.^[ [95]^30 , [96]^31 , [97]^32 , [98]^33 ^] To verify
   that the downregulation of MYO10 expression under Tp0136 treatment is
   attributed to Tp0136‐mediated alternative splicing of MYO10 through
   CSRP1, alternative splicing inhibitor spliceostatin A (SSA,
   HY‐16 466)^[ [99]^34 ^] was used to attenuate the alternative splicing
   of MYO10. It revealed that the inhibitory effect of SSA on alternative
   splicing in HMEC‐1 mirrored that of knocked‐down CSRP1 in HMEC‐1
   (P > 0.05), and Tp0136‐induced alternative splicing of MYO10 exon 19
   was significantly suppressed (P < 0.001, Figure [100]4D). Then, the
   protein expression level was quantified, revealing an elevation in the
   expression of MYO10 protein following SSA treatment. This observation
   aligned with the expression pattern of MYO10 observed after CSRP1
   knockdown, approaching the levels observed in the sh‐Ctl HMEC‐1
   (Figure [101]4E). Then, phalloidin staining was used to visualize the
   morphological characteristics of HMEC‐1 cells treated with SSA. The
   results revealed that in sh‐Ctl HMIC‐1 cells, the application of SSA
   counteracted the spheroidization of cell morphologies and loose
   connections induced by Tp0136, which were analogous to those observed
   in sh‐Ctl, sh‐CSRP1, and sh‐CSRP1 + Tp0136 cells (Figure [102]4F).
   Furthermore, the cell coverage rate of sh‐Ctl HMEC‐1 cells following
   SSA treatment was significantly elevated compared to that following
   Tp0136 treatment (P < 0.001), approaching the level observed in sh‐Ctl
   HMEC‐1 cells, and the thickness of sh‐Ctl cells attached to the lumen,
   after concurrent treatment with Tp0136 and SSA, recovered to the level
   of untreated sh‐Ctl cells, resembling the cell thickness observed after
   CSRP1 knockout (P > 0.05, Figure [103]4F). Additionally, the
   permeability of the sh‐Ctl + Tp0136 + SSA group was significantly
   decreased when compared to that in the sh‐Ctl + Tp0136 group
   (P < 0.001) and return to the levels of sh‐Ctl group and the CSRP1
   knockout group (P > 0.05, Figure [104]4G). This result further
   suggested that inhibiting MYO10 alternative splicing may aid in
   repairing the large cell gap induced by Tp0136 protein.

   In summary, these findings suggest that Tp0136 elicited a reduction in
   MYO10 expression via CSRP1‐mediated alternative splicing, which
   subsequently results in the spheroidization of cell morphology and the
   enlargement of cell junction gaps.

   One‐way ANOVA was used for comparisons between multiple groups and for
   comparisons between two groups. The values represent the mean ± SD of
   triplicates and are representative of three independent experiments.
   ***, P < 0.001, ###, P < 0.001, ns, P > 0.05.

3. Discussion

   Investigating the mechanism underlying the dissemination of Treponema
   pallidum from the wounds to other organs serves as the foundation for
   comprehending its pathogenic mechanism and subsequently blocking its
   dissemination. Our previous study utilizing the syphilis rabbit model
   has demonstrated that Tp0136 facilitated the dissemination of Treponema
   pallidum from the primary chancres to distal organs. Furthermore,
   investigations using HMEC‐1 cells have revealed that Tp0136 promotes
   angiogenesis by activating the PI3K‐AKT signaling pathway in vascular
   endothelial cells, ultimately leading to increased vascular
   permeability.^[ [105]^5 ^] However, the precise mechanism through which
   Tp0136 facilitates angiogenesis, ultimately resulting in increased
   vascular permeability, remains elusive. Utilizing the established
   in vivo study system and Tp0136 protein treatment protocol,^[ [106]^5
   ^] we observed the primarily characterized by blood vessel endothelial
   cell spheroidization and enlargement of intercellular spaces. In the
   in vitro experiments, high‐throughput transcriptional sequencing in
   HMEC‐1 cell line analysis revealed a downregulation of MYO10 expression
   and an upregulation of CSRP1 upon Tp0136 treatment. Additionally,
   alternative splicing analysis demonstrated an increased skipping
   frequency of MYO10 exon 19 ruled by CSRP1 under the treatment of
   Tp0136. Furthermore, by inhibiting the alternative splicing of MYO10
   exon 19 using splicing inhibitors, or knocking down CSRP1, we observed
   an increase in MYO10 expression, leading to the restoration of the
   morphology and intercellular connectivity of vascular endothelial cells
   to that observed in the negative control groups. These findings suggest
   that Tp0136 regulates MYO10 alternative splicing through CSRP1,
   resulting in the downregulation of MYO10 expression, which in turn
   leads to spheroidization in vascular endothelial cells, enlargement of
   intercellular spaces, and enhanced vascular permeability (Figure [107]
   5 ).

Figure 5.

   Figure 5
   [108]Open in a new tab

   Treponema pallidum protein Tp0136 inhibits the expression of MYO10,
   leading to spheroidization of vascular endothelial cells and
   enlargement of intercellular gaps.

   Blood barrier endothelial cell morphological alterations have been
   reported as a supporting channel with enhanced permeability,
   facilitating the dissemination of pathogens and tumor cells.^[ [109]^22
   , [110]^35 , [111]^36 ^] The cytoskeleton–myosin system serves as a
   pivotal factor in maintaining the morphology and barrier function of
   vascular endothelial cells, and both are indispensable.^[ [112]^37 ,
   [113]^38 ^] Myosin, as a cellular component responsible for regulating
   cytoskeletal dynamics, is highly susceptible to alterations during the
   pathogenesis of various pathogens.^[ [114]^39 ^] It is reported that
   Streptococcus pyogenes, Mycobacterium tuberculosis, and Mycobacterium
   ulcerans regulate myosin assembly to alter cell morphology, thereby
   assisting pathogens to enter the bloodstream.^[ [115]^40 , [116]^41 ,
   [117]^42 ^] Given the analogous pathogenicity between Treponema
   pallidum and the aforementioned bacteria, Treponema pallidum has been
   proven to promote the migration of cells in the blood vessels,^[
   [118]^43 , [119]^44 ^] which indirectly indicates that Treponema
   pallidum also has the effect of regulating the cell morphology. The
   findings in this study, along with those of our previous article,^[
   [120]^5 ^] collectively demonstrate the diverse morphological
   manifestations of vascular endothelial cytopathia induced by Treponema
   pallidum infection, from the perspective of the membrane protein
   Tp0136. These manifestations include excessive angiogenesis and
   cellular morphological changes characterized by spheroidization,
   leading to enlarged intercellular junctions. Our findings provide
   deepen insights into the mechanism by which Treponema pallidum
   facilitates its dissemination via the membrane protein Tp0136.

   Alternative splicing serves as a crucial regulatory mechanism in
   governing the morphology and permeability of vascular endothelial
   cells,^[ [121]^12 , [122]^45 , [123]^46 , [124]^47 ^] and it has been
   identified as a crucial factor in the dissemination of infectious
   diseases.^[ [125]^48 , [126]^49 ^] In the investigation of the
   pathogenic mechanisms of Treponema pallidum, the alternative isoforms
   of XBP‐1 have been observed in response to stimuli from Treponema
   pallidum and facilitated the immune evasion.^[ [127]^50 ^] However, the
   regulatory mechanisms governing these alterations remain primarily
   confined to the IRE1α pathway linked to the endoplasmic reticulum
   stress response.^[ [128]^50 , [129]^51 ^] To the best of our knowledge,
   in the field of Treponema pallidum infection, there is no report on the
   alternative splicing that regulates the structure of vascular
   endothelial cells to participate in the infection and dissemination
   process. In this study, we investigated the effect of Treponema
   pallidum on vascular endothelial cells and its underlying molecular
   mechanism, with particular emphasis on elucidating the mechanism by
   which Treponema pallidum protein regulates the expression of myosin
   proteins through alternative splicing, ultimately enhancing the
   permeability of the vascular endothelial barrier.

   CSRP1 localized in various cell types. Previous research has
   demonstrated that CSRP1 operates in conjunction with the cytoskeleton,
   particularly in the formation of mature focal adhesions between cells,
   where a reduction in its abundance is often necessary.^[ [130]^52 ^]
   Furthermore, studies have indicated that during focal inflammation,
   tissue cells at the wound site, including vascular endothelial cells,
   require substantial recruitment of myosin to facilitate the
   re‐establishment of strong adhesions. CSRP1 is implicated throughout
   this entire process.^[ [131]^52 , [132]^53 ^] In the process of
   infection and spread of pathogenic microorganisms, the regulatory
   effect of CSRP1 has not been found in mammalian models, but in the
   immunological studies of model animals such as fish and shellfish, it
   has been found that high expression of CSRP1 promotes the infection of
   pathogenic microorganisms,^[ [133]^54 ^] which provides a way to study
   the function of mammalian CSRP1 protein. As an RNA‐binding protein,
   CSRP1 is postulated to play a role in the process of alternative
   splicing.^[ [134]^55 ^] However, aside from our study demonstrating
   that CSRP1 regulates cell morphology through the modulation of myosin
   alternative splicing, there appears to be a paucity of research
   investigating the role of CSRP1 in regulating cellular function via
   alternative splicing.

   As a member of the myosin family, MYO10 is widely recognized, similar
   to other myosins, for its role in regulating the mobility of
   cytoskeletal proteins and consequently influencing cell morphology.^[
   [135]^56 ^] Deficiency in MYO10 can significantly impair the
   maintenance and function of cell morphology, potentially resulting in
   cell death.^[ [136]^57 , [137]^58 ^] In the context of infectious
   diseases, numerous studies have demonstrated that the integrity of the
   MYO10 gene and its protein expression can be compromised by pathogenic
   bacteria, facilitating the dissemination of pathogens such as the
   Shigella flexneri ^[ [138]^59 ^] and Marburg virus.^[ [139]^60 ^] In
   our study, we discovered that Treponema pallidum facilitates the
   isoform switch of MYO10 via the Tp0136 protein. Specifically, the
   isoform lacking exon 19 fails to encode the functional protein, leading
   to a downregulation of MYO10 protein expression, which consequently
   results in a spherical cell morphology and assists Treponema pallidum
   dissemination.

   Our findings collectively demonstrate that Tp0136‐induced endothelial
   spheroidization and junctional widening are mediated through the
   CSRP1‐MYO10 axis, providing a mechanistic basis for Treponema pallidum
   dissemination. Notably, the reversal of these phenotypes by CSRP1
   knockdown or splicing inhibition suggests that targeting CSRP1 activity
   (e.g., via RNA interference or small‐molecule inhibitors) or
   stabilizing MYO10 expression (e.g., through mRNA delivery or
   proteasomal blockade) may serve as potential therapeutic strategies to
   reinforce vascular barrier integrity and limit bacterial spread. These
   approaches could complement existing anti‐Tp0136 antibody‐based
   interventions,^[ [140]^5 ^] offering a multi‐faceted strategy to
   counteract syphilis progression. Future studies exploring the synergy
   between CSRP1/MYO10‐targeted therapies and conventional antibiotics may
   further optimize clinical outcomes.

   This study has several limitations. Firstly, Treponema pallidum is
   widely regarded as an extracellular bacterium, yet how it regulates
   intracellular CSRP1 through the outer membrane protein Tp0136 remains
   unexplored in our study. Our previous research has demonstrated that
   Treponema pallidum can be phagocytosed into immune cells, which allows
   the release of bacterial proteins into the intracellular space to exert
   their functions.^[ [141]^61 ^] However, in the early‐stage electron
   microscopy scans of this study, we did not observe Treponema pallidum
   entering the cells, although this does not necessarily mean that Tp0136
   is not present intracellularly. Secondly, in our previous study, we
   observed that Treponema pallidum within the anti‐Tp0136 group was
   unable to disseminate from cutaneous lesions to the testes,
   attributable to the block effects of anti‐Tp0136 antibodies to Tp0136
   protein on angiogenesis.^[ [142]^5 ^] Through a series of histochemical
   analyses and complementary methodologies, we consistently encountered
   challenges in detecting traces of blood vessels within the lesions of
   the anti‐Tp0136 group. This limitation hindered our ability to
   investigate the alterations in vascular endothelial cell transcripts
   associated with this group. Finally, based on our prior research
   indicating that Tp0136 facilitates angiogenesis and enhances vascular
   permeability, the precise location of Tp0136‐mediated MYO10 alternative
   splicing remains uncertain, whether it occurs within the newly formed
   blood vessels or within the preexisting vasculature at this juncture.
   Additionally, we acknowledge that the lack of in vivo genetic knockout
   models limits direct validation of the CSRP1‐MYO10 axis in live
   animals. While our in vitro knockdown and splicing inhibition
   experiments (Figure [143]2G,H and [144]4D–G) strongly support the
   functional role of CSRP1 and MYO10, the technical challenges of
   implementing CRISPR/Cas9 or shRNA‐based knockout approaches in the
   rabbit model—currently, the only syphilis pathology‐recapitulating
   system—precluded direct in vivo mechanistic verification. Future
   studies using localized shRNA delivery in rabbits will be critical to
   bridging this gap.

   In summary, the present study elucidated the mechanism underlying the
   Tp0136 protein's facilitation of MYO10 exon 19 skipping via the
   upregulation of vascular endothelial cell CSRP1. This upregulation
   subsequently led to downregulation of MYO10 expression, inducing
   morphological spheroidization in vascular endothelial cells and
   ultimately resulting in loosened cell connectivity. These findings have
   the potential to further clarify the mechanism underlying the
   dissemination of Treponema pallidum.

   Treponema pallidum protein Tp0136 upregulates the expression of CSRP1
   to enhance the alternative splicing of MYO10 exon 19. Consequently,
   this process leads to a reduction in the level of MYO10 protein and
   mediates the widening of the intercellular space within blood vessel
   endothelial cells.

4. Conclusion

   In this study, we elucidated a novel mechanism by which Treponema
   pallidum membrane protein Tp0136 facilitates bacterial dissemination
   through vascular endothelial barrier disruption. Combining in vivo
   syphilitic rabbit models and in vitro 3D microfluidic systems, we
   demonstrated that Tp0136 induces endothelial cell spheroidization and
   intercellular junction widening, leading to enhanced vascular
   permeability. High‐throughput RNA sequencing revealed that Tp0136
   upregulates the RNA‐binding protein CSRP1, which modulates the
   alternative splicing of MYO10 exon 19, resulting in MYO10
   downregulation. Functional validation confirmed that CSRP1 knockdown or
   inhibition of MYO10 splicing (via spliceostatin A) restores endothelial
   morphology and barrier integrity, directly linking the CSRP1‐MYO10 axis
   to Tp0136‐mediated vascular pathology. These findings provide the first
   evidence that Treponema pallidum exploits host RNA splicing machinery
   to manipulate cytoskeletal dynamics, offering a paradigm shift in
   understanding bacterial dissemination mechanisms. Therapeutically,
   targeting CSRP1 activity or restoring MYO10 expression presents a
   promising strategy to counteract vascular leakage and limit systemic
   spread of Treponema pallidum.

5. Experimental Section

5.1.

5.1.1.

Preparation of the Recombinant T. pallidum Membrane Protein Tp0136

   Recombinant T. pallidum membrane protein Tp0136 was purified, and
   endotoxins were removed as described in a previous study.^[ [145]^20 ^]
   Briefly, the full‐length Tp0136 was directly cloned into the
   pEXP‐5‐CT‐TOPO vector; then, the recombined plasmid was inserted into
   E. coli BL21 strains and grown in Luri–Bertani medium supplied with
   60 μg mL^−1 ampicillin. Protein expression was added 0.5 mmol L^−1
   isopropyl‐B‐Dthiogalactopyranoside (IPTG) at 25 °C for 6 h. The E. coli
   inserted with Tp0136 genes were harvested by centrifugation (6000 g,
   15 min, and 4 °C) and lysed by a sonicator with PBS augmented with
   10 μg mL^−1 lysozyme (Solarbio, Beijing, China). The lysate was
   collected by centrifugation (12 000 g, 20 min, and 4 °C). The pellet at
   the bottom of the tube was solubilized overnight in a buffer containing
   8 mol L^−1 urea and 0.1 mol L^−1 Tris‐HCl (pH = 8.0) at 4 °C. Tp0136
   recombinant proteins were purified from the pellet and eluted with PBS.
   Endotoxin contamination was removed using trace dynamic chromogenic
   limulus reagent (Bioendo, Xiamen, China) and was maintained at less
   than 0.05 endotoxin units (EUs)/mL. A bicinchoninic acid (BCA) (Takara
   Bio, Inc., Kusatu City, Japan) assay kit was used to estimate the
   concentrations.

Sample Preparation for Electron Microscopy

   Samples were embedded using a protocol for scanning electron
   microscopy.^[ [146]^62 , [147]^63 ^] Briefly, after fixation, the
   samples were post‐fixed in 2% osmium tetroxide and 1.5% potassium
   ferricyanide v v^−1 for 1 h on ice, incubated in 1% thiocarbohydrazide
   in dH[2]O w v^−1 for 20 min, followed by 2% osmium tetroxide in dH[2]O
   w v^−1 for 30 min, and then washed in dH[2]O and incubated overnight in
   1% aqueous uranyl acetate at 4 °C. Cells were then stained with
   Walton's lead aspartate for 30 min at 60 °C. The coverslips were
   removed from the dishes after submerging the bottom in methanol for
   20 min to soften the glue. The cells were then dehydrated stepwise
   through an ethanol series on ice, incubated in a tin foil container in
   a 1:1 propylene oxide and Durcupan resin mixture, and embedded in
   Durcupan ACM resin according to the manufacturer's instructions
   (Sigma‐Aldrich).

Cell Culture

   Human microvascular endothelial cell line‐1 (HMEC‐1) (Procell Life
   Science & Technology, China) was maintained in HMEC‐1 cell‐specific
   medium (Procell Life Science & Technology, China) in a humidified
   incubator at 37 °C and 5% CO[2] for 3 days, and the medium was changed
   every 24 h.

   For the HMEC1 CSRP1 knock‐down cell line construction, 5 × 10^6 HMEC1
   were infected with lentivirus expressing short hairpin RNA (shRNA)
   targeting human CSRP1 gene (Santa Cruz Biotechnology, Dallas, TX,
   http://www.scbt.com) and selected with puromycin for 4 days;
   nontargeting viral particles were used as a control. The alternative
   splicing inhibitor SSA was added as described previously.^[ [148]^64 ^]
   Briefly, the final concentration of 100 ng mL^−1 SSA was prewarmed with
   the HMEC1 culture medium and added to the 6‐well plate during medium
   change for a duration of 6 h.

   To ensure the uniformity of cell fusion across experimental batches, a
   standardized cell inoculum (2 × 10^7 cells/mL) was utilized, and a
   single pore cell from each batch was subjected to CD31 staining for
   cell‐edge determination. The boundaries of each cell were subsequently
   analyzed using the ImageJ software, with each cell being color‐coded.
   The extent of cell fusion was quantified by determining the ratio of
   the colored area within a given visual field to the total area of that
   field.

Immunocytochemistry

   Cells were fixed in phosphate‐buffered saline (PBS) containing 4%
   paraformaldehyde (PFA) for 30 min at room temperature. Thereafter, all
   cells were blocked with 5% FBS and Triton X‐100 and incubated with the
   MYO10 primary antibody (Santa Cruz, USA). The cells were then rinsed
   with PBS and incubated with species‐specific Alexa Fluor 488‐conjugated
   secondary antibodies (Thermo Fisher, USA), followed by staining with
   DAPI (Sigma‐Aldrich, USA) to counterstain the nuclei.

3D Microfluidic Angiogenesis System and Detection of Vascular Permeability

   As previously described, angiogenesis analysis was based on a
   microfluidic model and 3D cell culture to simulate angiogenesis in vivo
   (ref. to LW). HMEC‐1 were dissociated, pelleted, and suspended in a
   concentration of 2 × 10^7 cells mL^−1. Two microliters of the cell
   suspension was dispensed into the perfusion inlet and incubated for
   10 days. When the cells formed a stable monolayer, stimulation factors
   were added to the bottom perfusion channel, which was then placed in a
   shaker for continuous perfusion. On day 10 after stimulation, 50 mL of
   150 kDa TRITC‐Dextran solution (Sigma‐Aldrich, USA) was added to the
   perfusion channel inlet and time‐lapse images were acquired at 30‐s
   intervals using a confocal microscope. The cytoskeleton and nuclei were
   stained via immunofluorescence, using a Phalloidin‐iFluor 555 reagent
   (Abcam, UK) and an antifluorescence quenching blocking solution
   containing DAPI (Beyotime, China), respectively. A laser confocal
   scanning microscope (LSM780, Zeiss, Germany) was used to obtain 3D
   scanning images.

Cell Number Counting by Flow Cytometry and DAPI Number Calculation

   To quantify the HMEC‐1 cell population within the microfluidic system,
   the cells were enzymatically dissociated into a single‐cell suspension
   using 50 μL of a Trypsin containing 0.05% EDTA. The enzymatic reaction
   was subsequently halted by the addition of an equal volume of culture
   medium. Following the removal of the supernatant, the cells were
   resuspended in PBS, stained with CD31‐FITC, and enumerated using flow
   cytometry.

   To ascertain the cell count within a 100 × 100 μm area of the
   microfluidic system, we used a visual enumeration method using DAPI
   staining. This procedure involved selecting and analyzing 10 distinct
   fields of view.

Cell Assessment Validation Through CCK‐8 Assay

   Total cell viability was examined using cell counting kit‐8 (CCK‐8)
   assay. The samples were processed according to the protocol of the
   CCK‐8 assay kit (Beyotime, Shanghai, China) before determination of
   absorbance on a reader. After subtracting the absorbance of the blank
   control, the absorbance values of the samples were used to calculate
   the total cell viability rate.

RNA‐Sequencing Data Analysis and Visualization

   RNA sequencing data were analyzed and visualized based on the
   methodology reported in a previous study.^[ [149]^65 ^] Paired‐end
   sequencing reads were trimmed, and quality was checked using Fastqc
   (version 0.11.7). Alignment to the human genome was performed using
   STAR software (version 2.5.2a). Differential gene expression analysis
   was performed using the RStudio software (version 1.0.153). For the
   study of differentially expressed genes exposed to the Tp0136
   recombinant protein, we conducted a comparative analysis between the
   group treated with the Tp0136 recombinant protein and the untreated
   group, identifying a gene subset with a significance level of
   P < 0.001. To identify genes of greater research significance and more
   pronounced changes, those with a P < 0.001 and a gene fold change ratio
   ≥1.5 following treatment with Tp0136 recombinant protein were selected
   as the target gene set. Subsequent sequencing based on the change ratio
   resulted in the selection of the top 100 genes for further analysis
   using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes
   (KEGG) pathways. The GO and enrichment analyses of RNA sequencing data
   were conducted using Panther (http://pantherdb.org/) and Enrichr
   (https://amp.pharm.mssm.edu/Enrichr/) on specified gene lists. To more
   clearly observe the enriched function term of differentially expressed
   genes, we filtered the function term with a significance level of
   P < 0.001 in the GO enrichment analysis. In conducting differentially
   expressed gene pathway enrichment analysis, we used a methodology
   similar to that utilized in GO analysis to identify the genes
   exhibiting the most significant changes. Subsequently, these identified
   genes were subjected to analysis using the KEGG database.

   The splice junctions were identified using TopHat. Uniquely mapped
   reads (75–84% of sequenced reads) were selected for coverage and
   junction read count assembly. Replicate MATS (rMATS) was used to
   analyze alternative splicing events.^[ [150]^66 ^] The raw sequence
   data reported in this article have been deposited in the Genome
   Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in the
   National Genomics Data Center (Nucleic Acids Res 2022), China National
   Center for Bioinformation/Beijing Institute of Genomics, Chinese
   Academy of Sciences (GSA‐Human: HRA005158), which are publicly
   accessible at https://ngdc.cncb.ac.cn/gsa‐human.

Western Blotting

   Western blotting was performed as previously described.^[ [151]^67 ^]
   The information and dilutions of CSRP1, MYO10, and GAPDH antibodies can
   be found in Supporting Information. Bands were visualized using the Gel
   Recording System (Bio‐Rad, USA) or iBright CL750 Imaging System (Thermo
   Fisher Scientific, USA).

RNA‐Protein Immunoprecipitation

   RNA immunoprecipitation and RIP‐PCR was performed following the
   protocol described previously.^[ [152]^68 ^] Briefly, cells cultured in
   10 cm plates were washed twice with ice‐cold PBS and scraped off in
   1 mL PBS. Then, the cell was centrifuged and resuspended in an equal
   pellet volume of complete RIP lysis buffer (Merck Millipore).
   Five micrograms of antibody was prebound to protein A/G magnetic beads
   in an immunoprecipitation buffer (20 mM Tris‐HCl pH 7.5, 140 mM NaCl,
   0.05% TritonX‐100) for 2 h and then incubated with 100 μL cell lysates
   over night at 4 °C with rotation. Then, RNA was eluted from the beads
   by incubating with 400 μL elution buffer for 2 h. The eluted RNA was
   precipitated with ethanol and dissolved with RNase‐free water.
   Enrichment of certain fragments was determined by real‐time PCR. The
   primer used in RIP‐PCR is listed in Supporting Information.

Experimental of Alternative Splicing

   Regarding the screening of alternative splicing candidates associated
   with CSRP1 binding, the RNA‐Protein Interaction Prediction (RPISeq)
   database (Iowa State University, http://pridb.gdcb.iastate.edu/RPISeq/)
   was utilized and followed with TopHat analysis to identify junction
   sites, thereby obtaining candidate genes for CSRP1‐specific alternative
   splicing. Specific primers flanking predicted sites of alternative
   splicing were used for RT‐PCR amplifications. PCR products were
   separated by high‐resolution agarose gel electrophoresis. The primer
   used in alternative splicing detection is listed in Supporting
   Information.

Statistical Analysis

   All experiments were performed at least three times, and the data were
   presented as mean ± standard deviation. GraphPad Prism 8.0 (GraphPad
   Software, La Jolla, CA, USA) was used for all the statistical analyses.
   Student's t test or nonparametric tests were used when two groups were
   compared. One‐way analysis of variance (ANOVA) was used to compare the
   mean values of three or more groups with one independent variable.
   Two‐way ANOVA was used to compare the mean values three or more groups
   with two independent variables. Post hoc comparisons were performed
   using Tukey's test. Statistical significance was defined as a two‐sided
   P value of  <0.05.

Data Availability

   The RNA‐seq sequencing data reported in this article have been
   deposited in the Genome Sequence Archive (Genomics, Proteomics &
   Bioinformatics 2021) in the National Genomics Data Center (Nucleic
   Acids Res 2022), China National Center for Bioinformation/Beijing
   Institute of Genomics, Chinese Academy of Sciences (GSA‐Human:
   HRA005158), which are publicly accessible at
   https://ngdc.cncb.ac.cn/gsa‐human. Source data are provided with this
   article. The raw immunofluorescence and statistical figures can be
   obtained on the figshare website, DOI: 10.6084/m9.figshare.26 197 763.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   Yu Lin: conceptualization (equal); methodology (lead); project
   administration (lead); resources (equal); software (lead); supervision
   (equal); validation (lead); visualization (lead); writing—original
   draft (lead); writing—review editing (lead). Xi Luo: conceptualization
   (equal); methodology (equal); project administration (equal);
   validation (equal); visualization (equal). Xu Shen: resources (equal);
   software (equal); validation (supporting). Xiao‐Lin Fu: methodology
   (supporting); validation (supporting); visualization (supporting).
   Li‐Rong Lin: funding acquisition (equal); resources (equal);
   visualization (equal). Tian‐Ci Yang: conceptualization (lead);
   resources (lead); supervision (lead).

Supporting information

   Supplementary Material
   [153]SMSC-5-2500046-s001.pdf^ (345.5KB, pdf)

Acknowledgements