Abstract

   The risk of diabetic cardiovascular complications is closely linked to
   the length of hyperglycemia exposure. Mitophagy plays a significant
   role in vascular endothelial injury. However, the specific mechanisms
   by which mitophagy contributes to endothelial injury during sustained
   hyperglycemia remain unclear. In diabetic ApoE^−/− mice and human
   umbilical vein endothelial cell (HUVEC) models, mitophagy is enhanced
   following short‐term and long‐term high‐glucose exposure. Short‐term
   high‐glucose exposure promotes Parkin‐mediated mitophagy and
   upregulates mitochondrial fission protein 1 (Fis1) expression, whereas
   long‐term high‐glucose exposure suppresses Parkin‐mediated mitophagy
   and downregulates Fis1. With prolonged high‐glucose exposure, Syntaxin
   17 (STX17) translocates from the endoplasmic reticulum to the
   mitochondria, activating STX17‐mediated mitophagy. Silencing STX17
   alleviates mitochondrial degradation, decreases reactive oxygen species
   (ROS) levels, enhances endothelial nitric oxide synthase (eNOS)
   phosphorylation, and reduces apoptosis. Silencing Fis1 accelerates the
   switching to STX17‐mediated mitophagy, worsening endothelial
   dysfunction, whereas Fis1 overexpression prevents this switching,
   reducing ROS and apoptosis and enhancing eNOS phosphorylation. In
   summary, these findings suggest that the switching from Parkin‐mediated
   to STX17‐mediated mitophagy drives vascular endothelial injury
   following long‐term hyperglycemic exposure, providing valuable insights
   into therapeutic strategies for diabetic cardiovascular complications.

   Keywords: (diabetes, Fis1, mitophagy, Syntaxin 17), vascular
   endothelial injury
     __________________________________________________________________

   Mitophagy switching drives high‐glucose‐induced vascular endothelial
   injury. Short‐term high‐glucose exposure upregulates Fis1 expression
   and enhances Fis1‐TBC1D15 interaction, activating Parkin‐mediated
   mitophagy. Long‐term high‐glucose exposure reduces Fis1 expression,
   inhibiting Parkin‐mediated mitophagy and triggering mitophagy switching
   from Parkin‐mediated to STX17‐mediated mitophagy, leading to vascular
   endothelial injury. Preventing mitophagy switching may be a promising
   approach to alleviate high‐glucose‐induced vascular endothelial injury.

   graphic file with name ADVS-12-2414960-g002.jpg

1. Introduction

   Diabetes mellitus (DM) prevalence is rising rapidly around the world,^[
   [44]^1 ^] with cardiovascular complications emerging as the primary
   cause of mortality among diabetic individuals.^[ [45]^2 ^] Accumulating
   evidence suggests that the risk of cardiovascular complications
   increases with prolonged glycemic exposure.^[ [46]^3 ^] Moreover,
   hyperglycemia has been shown to induce long‐lasting effects on vascular
   injury, as demonstrated in both experimental models^[ [47]^4 ^] and
   clinical trials.^[ [48]^5 ^] However, the mechanisms underlying the
   long‐lasting impacts of hyperglycemic exposure on complication risk
   have not been fully elucidated.

   As the innermost layer of the vessel wall, endothelial cells (ECs) are
   the front line in sensing and responding to hyperglycemia.^[ [49]^6 ^]
   Endothelial dysfunction is a significant early event in the onset of
   diabetes vascular complications.^[ [50]^7 ^] Numerous studies have
   demonstrated that mitophagy plays a vital role in endothelial
   dysfunction.^[ [51]^8 ^] Mitophagy is a critical quality control
   mechanism, efficiently eliminating damaged mitochondria, maintaining
   intracellular homeostasis, supporting signal transduction, and
   mitigating ROS‐induced oxidative stress.^[ [52]^9 ^] Dysregulated
   mitophagy in ECs exposed to high glucose leads to the accumulation of
   dysfunctional mitochondria, triggering excessive ROS production and
   ultimately causing endothelial dysfunction.^[ [53]^10 ^] For example,
   high‐glucose exposure significantly inhibited mitophagy in human
   umbilical vein endothelial cells (HUVECs), resulting in mitochondrial
   dysfunction and apoptosis.^[ [54]^11 ^] Conversely, another study found
   that parkin‐mediated mitophagy increased in the vascular walls of obese
   mice and diabetic mice.^[ [55]^12 ^] Additionally, when exposed to high
   glucose, HUVECs treated with particulate matter significantly increased
   mitophagy and apoptosis.^[ [56]^13 ^] These discrepancies among studies
   may be attributed to different exposure times of high glucose.
   Different mitophagy mechanisms might be involved in developing
   endothelial dysfunction following short‐term or long‐term exposure to
   hyperglycemia. Thus, additional studies are required to explore and
   clarify the distinct mitophagy responses in ECs exposed to short‐term
   and long‐term hyperglycemia.

   Following high‐glucose exposure, the classical Parkin‐mediated
   mitophagy pathway in ECs is receiving more attention.^[ [57]^8 ,
   [58]^14 ^] A seminal finding of Parkin‐mediated mitophagy was that
   mitochondrial fission events at the periphery separate unhealthy
   daughter organelles from healthy mitochondria, with the damaged ones
   subsequently removed via Parkin‐mediated mitophagy.^[ [59]^15 ^] During
   this process, the enrichment of mitochondrial fission protein 1 (Fis1)
   regulates peripheral fission by recruiting lysosomes through its
   interaction with TBC1D15.^[ [60]^16 ^] Following peripheral fission,
   the membrane depolarization of daughter mitochondria recruits the E3
   ubiquitin ligase parkin, leading to their clearance through
   Parkin‐mediated mitophagy.^[ [61]^15 , [62]^17 ^] However, this
   mitophagy mechanism's function in advancing high‐glucose‐induced
   endothelial dysfunction requires further investigation. Besides
   Parkin‐mediated mitophagy, Parkin‐independent pathways contribute to
   mitochondrial elimination, involving many proteins, including BNIP3,
   NIX, FUNDC1 and Fis1.^[ [63]^18 ^] Among these proteins, the absence of
   Fis1 permits STX17 to trigger mitophagy via the non‐selective autophagy
   machinery, independent of ectopic stimulations like mitochondrial
   depolarization. Under normal conditions, Fis1 functions as a
   gatekeeper, inhibiting the exposure of STX17's N‐terminus. The loss of
   Fis1 leads to the exposure of STX17's N‐terminus, thereby triggering
   mitophagy. Fis1‐STX17‐induced mitophagy leads to the degradation of
   overall mitochondrial proteins.^[ [64]^19 ^] Nevertheless, whether
   Fis1‐STX17‐mediated mitophagy is relevant to vascular endothelial
   injury induced by long‐term hyperglycemia exposure remains unclear.

   The present study aims to elucidate the distinct impacts of short‐term
   and long‐term hyperglycemia exposure on vascular endothelial cells,
   focusing on whether distinct mitophagy mechanisms are involved.
   Following short‐term and long‐term high‐glucose exposure, we
   investigated vascular endothelial injury and its underlying mitophagy
   mechanisms in ApoE^−/− mice and HUVECs. Our findings suggest that
   mitophagy switching may be a critical mechanism driving
   high‐glucose‐induced vascular endothelial injury, and Fis1 appears to
   play a crucial role in mitophagy switching.

2. Results

2.1. High‐Glucose‐Induced Increase in Mitophagy Precedes Vascular Endothelial
Injury

   To fully understand the mechanism of hyperglycemia‐induced vascular
   injury, STZ was injected into male ApoE^−/− mice (6 weeks old) to
   induce hyperglycemia. We performed a 4D‐Label free quantitative
   proteomic analysis by mass spectrometry of the whole aortas from three
   groups of mice: HG 0 W (0 weeks of hyperglycemia), HG 6 W (6 weeks of
   hyperglycemia), and HG 20 W (20 weeks of hyperglycemia). Through strict
   quality control, 4443 quantifiable proteins were obtained (Figure
   [65]S1A, Supporting Information). The Venn diagrams (Figure [66]S1B,
   Supporting Information) ensured the robustness and reproducibility of
   our data. Specifically, between HG 0 W and HG 6 W, 517 proteins showed
   significant differential expression, including 177 proteins that were
   downregulated and 340 that were upregulated. Between HG 20 and HG 0 W,
   350 proteins showed significant differential expression, including 262
   proteins that were downregulated and 88 that were upregulated. Between
   HG 20 and HG 6 W, 372 proteins showed significant differential
   expression, including 196 proteins that were downregulated and 176 that
   were upregulated (Figure [67]1A). The heatmaps (Figure [68]S1C,
   Supporting Information) illustrated distinct expression profiles of the
   differentially expressed proteins across the different groups. GO
   functional annotation and KEGG pathway enrichment analysis
   (Figure [69]1B; Figure [70]S1D, Supporting Information) elucidated
   critical biological processes and pathways. In the comparison between
   HG 6 and HG 0 W, significant enrichment was observed in the
   mitochondrion, lysosome, and autophagy pathways. In the comparison
   between HG 20 and HG 0 W, there was marked enrichment in processes
   involving apoptosis, mitochondrion, lysosome, and autophagosome
   pathways. Comparing HG 6 and HG 20 W, alterations were evident in
   mitochondrion, lysosome, and metabolic pathways. Compared to HG 0 W,
   the differentially expressed proteins in the HG 6 W group differed from
   those in the HG 20 W group, but both were enriched in “mitochondrion”
   and ″lysosome′ pathways, which suggest that mitochondrial and
   autophagy‐related mechanisms may play crucial roles in
   hyperglycemia‐induced vascular injury. Due to limitations in observing
   mitochondria and lysosomes in vivo, HUVECs were stained with
   MitoTracker and LysoTracker in vitro, compared with the normal glucose
   (NG) group, 12 h (12HG) and 72 h (72HG) of high‐glucose exposure
   significantly increased the interaction between mitochondria and
   lysosomes (Figure [71]1C). These findings indicate mitophagy is closely
   linked to the mechanisms driving hyperglycemia‐induced vascular injury.

Figure 1.

   Figure 1
   [72]Open in a new tab

   Mitophagy activation precedes hyperglycemia‐induced vascular
   endothelial injury. A,B, D–J) Male ApoE^−/− mice (6 weeks old) were
   injected with STZ to induce hyperglycemia. Mouse aortas were collected
   at 0 weeks (HG 0 W), 6 weeks (HG 6 W), and 20 weeks (HG 20 W) following
   hyperglycemia exposure. A,B) 4‐D label‐free quantitative proteomic
   analysis of aortas in mice. n = 3. A) Differentially expressed protein
   counts for the following comparisons: HG 0 W versus HG 6 W, HG 20 W
   versus HG 0 W, and HG 20 W versus HG 6W. B) Bar and bubble charts
   showing the GO and KEGG pathway enrichment analyses of differentially
   expressed proteins (HG 6 W vs HG 0 W, HG 20 W vs HG 0 W). C)
   Co‐localization of mitochondria (MitoTracker, red) and lysosomes
   (LysoTracker, green) in HUVECs exposed to normal glucose (NG group), 12
   h of high glucose (12HG group), and 72 h of high glucose (72HG group),
   with Pearson's coefficient for mitochondria‐lysosome co‐localization.
   Scale bar = 10 µm. n = 6. D) Representative in situ images of
   atherosclerotic plaque in the aortic arch with statistical analysis.
   Scale bar = 1 mm. n = 6. E) Oil Red O staining of cross‐sections at the
   aortic root with statistical analysis. Scale bar = 100 µm. n = 6. F–G)
   Representative images and statistical analysis of macrophage
   infiltration (F4/80) F) and α‐smooth muscle actin (α‐SMA)
   immunostaining G) in aortic root plaque. Scale bar = 100 µm. n = 6. H)
   Representative images and statistical analysis of TUNEL^+ cells of
   aortal vascular endothelium. Scale bar = 50 µm. n = 3. I,J) En face
   immunofluorescence staining of the mice abdominal aorta with
   immunofluorescent density analysis. n = 6. I) Red, green, and blue
   represent LC3, VWF (an endothelial cell marker) and DAPI (a nuclear
   marker), respectively. J) Red, green, and blue represent p62, VWF (an
   endothelial cell marker) and DAPI (a nuclear marker), respectively.
   Scale bar = 20 µm. Data are expressed as mean ± SEM. ^* p < 0.05; ^**
   p < 0.01.

   To analyze the temporal changes in mitophagy during the progression of
   hyperglycemia‐induced vascular injury, we measured vascular endothelial
   damage and the expression of autophagy marker proteins in mice at HG
   0 , HG 6 , and HG 20W. The atherosclerotic lesion area was
   significantly increased in the HG 20 W group compared to the HG 0 W
   group, as shown by in situ images (Figure [73]1D) and Oil Red O
   staining (Figure [74]1E). This increase is accompanied by more
   extensive macrophage content (Figure [75]1F) and smooth muscle cell
   content (Figure [76]1G) in atherosclerotic lesions. Increased apoptosis
   in the vascular endothelium of HG 20 W group mice was detected by TUNEL
   staining (Figure [77]1H). The oxidative stress also displayed parallel
   changes as assessed by Immunohistochemical analysis of oxidative stress
   markers 3‐NT and 8‐OHdG (Figure [78]S2, Supporting Information). In
   summary, the HG 6 W group showed a minor degree of vascular endothelial
   damage compared to the HG 0 W group, but the damage was significantly
   worsened in the HG 20 W group. Intriguingly, vascular en face
   fluorescence staining shows a significant increase in autophagy markers
   p62 and LC3 expression, starting at 6 weeks and persisting through 20
   weeks (Figure [79]1I,J). These findings indicate that changes in
   mitophagy may occur before the onset of hyperglycemia‐induced vascular
   endothelial damage and oxidative stress.

   The level of mitophagy was also investigated during a
   high‐glucose‐induced endothelial injury in HUVECs, compared with normal
   glucose (NG group), 72 h of high‐glucose exposure (72HG group)
   increased the BAX/BCL2 ratio and the levels of cleaved caspase 3
   expression, while reducing eNOS phosphorylation at serine 1177 (Figure
   [80]2A). These changes were accompanied by increased ROS and
   mitochondrial ROS production (Figure [81]2B,C). In contrast, oxidative
   stress and cell apoptosis were not induced by 12 h of high‐glucose
   exposure (12HG group). The co‐localization of LC3 and TOM20 and
   ratiometric measurements of mt‐Keima (a mitochondrial‐targeted form of
   the fluorescent reporter Keima) were analyzed to validate mitophagy.
   These observations indicate that 12 h of high‐glucose exposure
   activated mitophagy (Figure [82]2D,E). Electron microscopy (EM) of
   HUVECs revealed a significantly increased number of mitochondria within
   autophagosomes in the 12HG and 72HG groups compared with the NG group
   (Figure [83]2F). This finding was further supported by increased levels
   of LC3 cleavage in the 12HG and 72HG groups compared with the NG group
   (Figure [84]2A). These results suggest that mitophagy is activated
   preceding high‐glucose‐induced vascular endothelial injury.

Figure 2.

   Figure 2
   [85]Open in a new tab

   Mitophagy is activated prior to high‐glucose‐induced injury in HUVECs.
   A–E) HUVECs were exposed to normal glucose (NG group), 12 h of high
   glucose (12HG group), and 72 h of high glucose (72HG group). A)
   Representative Western blot images and corresponding relative
   intensities of BAX, BCL2, cleaved caspase 3, p‐eNOS, eNOS, LC3, and p62
   bands in HUVECs. n = 6. B‐C) ROS B) and mitochondrial ROS C) production
   were determined by DCF and MitoSOX staining, respectively. Scale bar =
   50 µm. n = 6. D) Representative IF images of TOM20 (red), LC3 (green),
   and DAPI (blue) in HUVECs from each group, with Pearson's coefficient
   for TOM20‐LC3 co‐localization. Scale bar = 10 µm. n = 6. E) HUVECs
   stably expressing mitochondrial‐targeted Keima (mt‐Keima) were imaged
   by confocal microscopy with excitation at 440 nm (red) and 590 nm
   (green). Scale bar = 10 µm. n = 6. F) HUVECs were analyzed by
   transmission electron microscopy. Blue arrows indicate autophagic
   structures enclosing mitochondria. Scale bar = 2 µm. n = 6. Data are
   expressed as mean ± SEM. ^* p < 0.05; ^** p < 0.01.

2.2. Parkin‐Mediated Mitophagy Is Activated by Short‐Term High‐Glucose
Exposure but Suppressed by Long‐Term High‐Glucose Exposure

   Parkin‐mediated mitophagy has been considered the principal mechanism
   mediating mitophagy. To investigate the involvement of parkin‐mediated
   mitophagy in vascular endothelial injury in vitro, we monitored the
   co‐localization of Parkin, mitochondria, and LC3. Compared with normal
   glucose (NG group), 12 h of high‐glucose exposure induced the
   mitochondrial translocation of Parkin. However, no Parkin translocation
   was detected in the 72HG group (Figure [86]3A). HUVECs treated with HG
   were transfected with or without Prkn‐targeting siRNA (si‐Parkin)
   (Figure [87]S3A,B, Supporting Information). Oxidative stress and
   apoptosis were elevated, while eNOS phosphorylation was decreased in
   12‐h HG‐treated si‐Parkin HUVECs compared to 12‐h HG‐treated control
   HUVECs (Figure [88]3B–D). Moreover, apoptosis and oxidative stress were
   further exacerbated in 72 h HG‐treated si‐Parkin HUVECs, as reflected
   by an increased BAX/BCL2 ratio, ROS, and mitochondrial ROS production
   (Figure [89]3B–D). These observations suggest that high‐glucose‐induced
   vascular endothelial injury was accelerated and aggravated by
   si‐Parkin, indicating that Parkin‐mediated mitophagy plays a prominent
   role in protecting against endothelial injury. Furthermore, long‐term
   high‐glucose exposure inhibited Parkin‐mediated mitophagy, potentially
   leading to ROS overproduction, endothelial dysfunction, and apoptosis
   in HUVECs.

Figure 3.

   Figure 3
   [90]Open in a new tab

   Parkin‐mediated mitophagy is activated by short‐term high‐glucose
   exposure while silencing Parkin accelerates and aggravates the
   high‐glucose‐induced vascular endothelial injury. A) HUVECs following
   normal glucose (NG group), 12 h of high glucose (12HG group), and 72 h
   of high glucose (72HG group) exposure were stained with MitoTracker
   (red) and analyzed by immunofluorescence microscopy for Parkin (green)
   and LC3 (cyan). Nuclei were stained with Hoechst (blue). HUVECs in the
   NG group, treated with 10 µm CCCP for 4 h, were used as a positive
   control. Scale bar = 10 µm. n = 6. B–E) HUVECs were transfected with
   small interfering RNA (siRNA) against control (si‐Con) or Parkin
   (si‐Parkin) and then exposed to normal glucose (NG group), 12 h of high
   glucose (12HG group), and 72 h of high glucose (72HG group). B‐C) ROS
   and mitochondrial ROS production were measured by DCF and MitoSOX
   staining, respectively. Scale bar = 50 µm. n = 6. D) Representative
   Western blot images and corresponding relative intensities of BAX,
   BCL2, cleaved caspase 3, p‐eNOS, eNOS, LC3, and p62 bands in HUVECs. n
   = 6. E) HUVECs stably expressing mitochondrial‐targeted Keima
   (mt‐Keima) were imaged by confocal microscopy with excitation at 440 nm
   (red) and 590 nm (green). Scale bar = 10 µm. n = 6. Data are expressed
   as mean ± SEM. A) ^* p < 0.05; ^** p < 0.01. B–E) ^* p < 0.05; ^**
   p < 0.01 versus NG + si‐Con group; ^# p < 0.05, ^## p < 0.01 versus
   12HG + si‐Con group; ^° p < 0.05, ^°° p < 0.01 versus 72HG + si‐Con
   group.

   The elevated overall mitophagy was reduced by si‐Parkin following 12 h
   of high‐glucose exposure in HUVECs, indicating that Parkin‐mediated
   mitophagy is the predominant mitophagy pathway activated by short‐term
   high‐glucose exposure. However, it was not suppressed by si‐Parkin in
   the 72HG group, as reflected by mt‐Keima (ex:590 nm) fluorescence ratio
   and levels of LC3 cleavage (Figure [91]3D,E). The results suggest that
   endothelial cells exposed to long‐term high glucose inhibit
   Parkin‐mediated mitophagy and activate additional Parkin‐independent
   mitophagy pathways.

2.3. Long‐Term High‐Glucose Exposure Initiates STX17‐Mediated Mitophagy

   Short‐term and long‐term high‐glucose exposure may activate distinct
   mitophagy pathways. To test this, daughter mitochondria resulting from
   mitochondrial fission were first tracked, and their distinct mitophagy
   fates were observed. During mitophagy, dysfunctional portions of
   mitochondria separate from healthy portions through mitochondrial
   peripheral fission (occurring less than 25% from a tip). Daughter
   mitochondria, resulting from peripheral fission and loss of
   mitochondrial membrane potential (membrane depolarization), are
   subsequently targeted for mitophagy.^[ [92]^15 , [93]^20 ^] HUVECs were
   imaged following 12 h of high‐glucose exposure, and elevated peripheral
   fission rates were observed compared to the NG group, whereas the
   midzone fission rates were unaffected. However, 72 h of high‐glucose
   exposure reduced the rates of peripheral fission, while the midzone
   fission rate remained constant (Figure [94]4D,E). HUVECs exposed to
   long‐term high‐glucose showed inhibited peripheral fission and did not
   produce daughter mitochondria with membrane depolarization
   (Figure [95]4G). It seems inconsistent with the mitophagy activation
   observed in the 72HG group. Additionally, mitochondria in the 72HG
   group were entirely enveloped by LysoTracker‐labeled lysosomes, leading
   to their subsequent degradation (Figure [96]4F). This process may
   contribute to mitochondrial fragmentation and loss (Figure [97]4A–C).
   Therefore, we speculated that endothelial cells exposed to long‐term
   high‐glucose may activate an additional, non‐selective mitophagy
   pathway independent of membrane depolarization.

Figure 4.

   Figure 4
   [98]Open in a new tab

   Long‐term high‐glucose exposure activates a mitophagy pathway
   independent of mitochondrial depolarization. A–F) HUVECs were exposed
   to normal glucose (NG group), 12 h of high glucose (12HG group), and 72
   h of high glucose (72HG group). A) HUVECs were stained with MitoTracker
   Red. Mitochondrial length B) and total number of fragmented, tubular,
   and elongated mitochondrial structures C) were quantitatively analyzed
   in each group. Scale bar = 10 µm. n = 6. D) The mitochondrial
   morphology of live HUVECs stained with MitoTracker Red was recorded
   using time‐lapse confocal microscopy, capturing images at 10s intervals
   over 10 min. Scale bar = 10 µm. E) Quantification of mitochondrial
   fission events in live HUVECs was performed, with peripheral fission
   events (orange, 0–25%) occurring near the tip and midzone fission
   events (green, 25%–50%) near the center. n = 20. F) Positioning of
   mitochondria (MitoTracker, red) and lysosomes (LysoTracker, green) in
   HUVECs exposed to 72 h of high glucose. Scale bar = 1 µm. G) The
   mitochondrial membrane potential was assessed during peripheral and
   midzone fission in TMRE‐stained mitochondria in the 12HG and 72HG
   groups. Scale bar = 10 µm. Data are expressed as mean ± SEM. ^*
   p < 0.05; ^** p < 0.01.

   Parkin‐independent and non‐selective mitophagy pathways mediated by
   proteins such as Nix, FUNDC1, Bcl‐2‐L‐13, FKBP8, and Fis1 have been
   reported.^[ [99]^18 , [100]^21 ^] Among these, 4D‐label‐free
   quantitative proteomic analysis revealed that Fis1 was the only protein
   with significantly downregulated expression in the aorta of mice
   exposed to 20 weeks of hyperglycemia (Figure [101]5A). Previous studies
   suggested that loss of Fis1 facilitates STX17 translocation from the
   endoplasmic reticulum (ER) to mitochondria, which allows STX17 to
   trigger mitophagy via the non‐selective mitophagy machinery.^[ [102]^19
   , [103]^21 ^] To verify whether long‐term high‐glucose exposure
   triggers STX17‐mediated mitophagy, the percentage of mitochondria
   co‐localized with STX17 and LC3 in HUVECs was evaluated. Compared to
   normal glucose (NG group), 72 h of high‐glucose exposure significantly
   increased Fis1‐STX17‐induced mitophagy in HUVECs. Fragmented
   mitochondria co‐localized with STX17 were observed to be enclosed
   within LC3‐labeled autophagosomes (Figure [104]5B). Additionally, to
   monitor temporal changes in the subcellular distribution of STX17
   during high‐glucose‐induced endothelial injury, mitochondria and ER
   were extracted and identified using organelle markers (COXIV for
   mitochondria and Calnexin for ER). STX17 expression was continuously
   measured in HUVECs, showing that following 72 h of high‐glucose
   exposure, STX17 levels decreased in the ER while increasing in the
   mitochondria, indicating its translocation from the ER to mitochondria
   (Figure [105]5D). These results suggest that long‐term high‐glucose
   exposure initiates STX17‐mediated mitophagy in HUVECs, causing a
   significant reduction in overall mitochondrial protein levels, as
   detected by immunoblotting (Figure [106]5C).

Figure 5.

   Figure 5
   [107]Open in a new tab

   Long‐term high‐glucose exposure triggers STX17‐mediated mitophagy in
   HUVECs. A) Male ApoE^−/− mice (6 weeks old) were injected with STZ to
   induce hyperglycemia. Mouse aortas were collected at 0 weeks (HG 0 W)
   and 20 weeks (HG 20 W) following hyperglycemia exposure. Volcano plot
   (HG 20 W versus HG 0 W) showing differentially expressed proteins, with
   downregulated Fis1 labeled. B) HUVECs were exposed to normal glucose
   (NG group), 12 h of high glucose (12HG group), and 72 h of high glucose
   (72HG group). HUVECs were labeled with MitoTracker (red) and analyzed
   using immunofluorescence microscopy to detect STX17 (green) and LC3
   (cyan), with nuclei stained with Hoechst (blue). Scale bar = 10 µm. An
   enlarged image of STX17‐mediated mitophagy in the 72HG group is shown.
   Scale bar = 1 µm. n = 6. C) Representative Western blot images and
   corresponding relative intensities of TOM20, Cyto C, TIM23, and HSP60
   bands, representing proteins from the OMM, IMS, IMM, and matrix. OMM is
   the outer mitochondrial membrane, IMS is the intermembrane space, and
   IMM is the inner mitochondrial membrane. n = 6. D) Representative
   Western blot images and corresponding relative intensities showing
   temporal changes in STX17 expression in mitochondria and the
   endoplasmic reticulum (ER) following high‐glucose exposure, with COXIV
   used as a mitochondrial marker and Calnexin as an ER marker. n = 6.
   Data are expressed as mean ± SEM. ^* p < 0.05; ^** p < 0.01.

2.4. Silencing STX17 Alleviates High‐Glucose‐Induced Vascular Endothelial
Injury

   To investigate the involvement of STX17‐mediated mitophagy in vascular
   endothelial injury following high‐glucose exposure, we treated HUVECs
   with high glucose for 72 h, followed by transfection with
   STX17‐targeting siRNA (si‐STX17) to block STX17‐mediated mitophagy
   (Figure [108]S4A,B, Supporting Information). Mitochondrial
   fragmentation and loss induced by 72 h of high‐glucose exposure were
   significantly reduced by si‐STX17 transfection (Figure [109]6A–C).
   Consistent with these morphological changes, Western blot analysis
   revealed a significant increase in overall mitochondrial proteins in
   si‐STX17‐transfected HUVECs following 72 h high‐glucose exposure,
   including TOM20 (outer mitochondrial membrane protein), Cyto C
   (intermembrane space protein), TIM23 (inner membrane mitochondrial
   protein) and HSP60 (mitochondrial matrix protein), compared with
   si‐Control transfected HUVECs following 72 h high‐glucose exposure
   (Figure [110]6G). Then, the role of STX17‐mediated mitophagy in
   apoptosis was investigated. Notably, transfection with si‐STX17 reduced
   apoptosis induced by 72 h of high‐glucose exposure, indicated by
   decreased cleaved caspase 3 expression and a reduced Bax/Bcl‐2 ratio
   (Figure [111]6F). Moreover, si‐STX17 increased eNOS phosphorylation,
   improving endothelial function (Figure [112]6F). DCF and MitoSOX
   staining showed that blocking STX17‐mediated mitophagy suppressed ROS
   and mitochondrial ROS overproduction (Figure [113]6D,E). The elevated
   overall mitophagy was reduced by si‐STX17 following 72 h of
   high‐glucose exposure in HUVECs, indicating that STX17‐mediated
   mitophagy is the predominant mitophagy pathway activated by long‐term
   high‐glucose exposure (Figure [114]6F,H). These findings suggest
   long‐term high‐glucose exposure induces STX17‐mediated mitophagy,
   potentially leading to vascular endothelial injury. Silencing STX17
   protects against vascular endothelial injury caused by high glucose.

Figure 6.

   Figure 6
   [115]Open in a new tab

   Silencing STX17 alleviated high‐glucose‐induced vascular endothelial
   injury. HUVECs were transfected with siRNA against control (si‐Con) or
   STX17 (si‐STX17) and then exposed to normal glucose (NG group) and 72 h
   of high glucose (72HG group). A) HUVECs were stained with MitoTracker
   Red. Mitochondrial length B) and total number of fragmented, tubular,
   and elongated mitochondrial structures C) were quantitatively analyzed
   in each group. n = 6. D,E) ROS and mitochondrial ROS production were
   measured by DCF and MitoSOX staining, respectively. n = 6. F)
   Representative Western blot images and corresponding relative
   intensities of BAX, BCL2, cleaved caspase 3, p‐eNOS, eNOS, LC3, and p62
   bands in HUVECs. n = 6. G) Representative Western blot images and
   corresponding relative intensities of TOM20, Cyto C, TIM23, and HSP60,
   representing proteins from the OMM, IMS, IMM, and matrix. n = 6. H)
   HUVECs stably expressing mitochondrial‐targeted Keima (mt‐Keima) were
   imaged by confocal microscopy with excitation at 440 nm (red) and
   590 nm (green). Scale bar = 10 µm. n = 6. Data are expressed as mean ±
   SEM. ^* p < 0.05; ^** p < 0.01 versus NG + si‐Con group; ^# p < 0.05,
   ^## p < 0.01 versus 72HG + si‐Con group.

2.5. Fis1 Is Involved in the Switch Between Parkin‐Mediated Mitophagy and
STX17‐Mediated Mitophagy

   Our findings indicate that inhibition of Parkin‐mediated mitophagy,
   combined with activation of STX17‐mediated mitophagy in endothelial
   cells exposed to long‐term high‐glucose, might lead to endothelial
   injury. Given that Fis1‐recruited TBC1D15 regulates Parkin‐mediated
   mitophagy,^[ [116]^22 ^] while Fis1 loss promotes STX17‐initiated
   mitophagy,^[ [117]^19 ^] the involvement of Fis1 in modulating both
   Parkin‐mediated and STX17‐mediated mitophagy following high‐glucose
   exposure was explored. The expression of Fis1 was continuously measured
   in HUVECs exposed to high glucose. Notably, Fis1 expression initially
   increased and decreased over time (Figure [118]7A). Similarly, en‐face
   fluorescence staining of mouse vasculature showed a significant
   upregulation of Fis1 in the HG 6 W group, followed by a decrease in the
   HG 20 W group (Figure [119]7B). The co‐immunoprecipitation analysis
   indicated that the interaction of Fis1 with TBC1D15 was significantly
   enhanced in the 12HG group compared with the NG group. As anticipated,
   long‐term high‐glucose exposure reduced the interaction of Fis1 with
   TBC1D15 and the combination of Fis1‐STX17 in HUVECs (Figure [120]7K,L).

Figure 7.

   Figure 7
   [121]Open in a new tab

   Fis1 modulates Parkin‐mediated mitophagy and STX17‐mediated mitophagy
   in high‐glucose conditions. A) Representative Western blot bands and
   relative band intensities showing temporal changes in Fis1 expression
   following high‐glucose exposure. n = 6. B) Male ApoE^−/− mice (6 weeks
   old) were injected with STZ to induce hyperglycemia. Mouse aortas were
   collected at 0 weeks (HG 0 W), 6 weeks (HG 6 W), and 20 weeks (HG 20 W)
   following hyperglycemia exposure. En face immunofluorescence staining
   of the mice abdominal aorta with immunofluorescent density analysis.
   Red, green, and blue represent Fis1, VWF (an endothelial cell marker)
   and DAPI (a nuclear marker). Scale bar = 20 µm. n = 6. C–J, M–P) HUVEC
   cell lines with Fis1 overexpression (OE‐Fis1) and silencing (sh‐Fis1)
   were constructed using lentivirus‐mediated transduction and then
   exposed to normal glucose (NG group), 12 h of high glucose (12HG
   group), and 72 h of high glucose (72HG group). C,D) Fis1‐silencing
   HUVECs were stained with MitoTracker (red) and analyzed by
   immunofluorescence microscopy for Parkin (green) and LC3 (cyan). Nuclei
   were stained with Hoechst (blue). n = 6. E,F) Fis1‐silencing HUVECs
   were labeled with MitoTracker (red) and analyzed using
   immunofluorescence microscopy to detect STX17 (green) and LC3 (cyan),
   with nuclei stained with Hoechst (blue). n = 6. G,H)
   Fis1‐overexpression HUVECs were stained with MitoTracker (red) and
   analyzed by immunofluorescence microscopy for Parkin (green) and LC3
   (cyan). Nuclei were stained with Hoechst (blue). n = 6. I,J)
   Fis1‐overexpression HUVECs were labeled with MitoTracker (red) and
   analyzed using immunofluorescence microscopy to detect STX17 (green)
   and LC3 (cyan), with nuclei stained with Hoechst (blue). Scale bar =
   10 µm. n = 6. K–L) Co‐immunoprecipitation analysis of the interaction
   of K) Fis1 and TBC1D15, L) Fis1 and STX17 in HUVECs exposed to normal
   glucose (NG group), 12 h of high glucose (12HG group), and 72 h of high
   glucose (72HG group). n = 3. M,N) Co‐immunoprecipitation analysis of M)
   Fis1 and TBC1D15 interaction, N) Fis1 and STX17 in Fis1‐overexpression
   HUVECs. n = 3. O,P) Representative Western blot bands and relative band
   intensities showing STX17 expression in mitochondria and the
   endoplasmic reticulum (ER) of O) Fis1‐overexpression and P)
   Fis1‐silencing HUVECs, with COXIV used as a mitochondrial marker and
   Calnexin as an ER marker. n = 6. Data are expressed as mean ± SEM. A)
   ^* p < 0.05; ^** p < 0.01 versus 0‐h high‐glucose exposure group. B) ^*
   p < 0.05; ^** p < 0.01. C‐P) ^* p < 0.05; ^** p < 0.01 versus NG + Vec
   group; ^# p < 0.05, ^## p < 0.01 versus 12HG + Vec group; ^° p < 0.05,
   ^°° p < 0.01 versus 72HG + Vec group. Vec, vector.

   To investigate the involvement of Fis1 in modulating Parkin‐mediated
   and STX17‐mediated mitophagy under high‐glucose conditions, we
   constructed lentivirus‐mediated stable HUVEC cell lines with Fis1
   overexpression and silencing (Figure [122]S5A,B, Supporting
   Information). Fis1 silencing inhibited Parkin‐mediated mitophagy
   (Figure [123]7C,D) and promoted STX17‐mediated mitophagy
   (Figure [124]7E,F) in the 12‐hour HG‐treated groups. Further Western
   blot analysis of STX17 confirmed that silencing of Fis1 in the 12 h of
   HG‐treated groups significantly stimulated the translocation of STX17
   from the ER to mitochondria (Figure [125]7O,P) compared with 12 h of
   HG‐treated VEC groups, accelerating the initiation of STX17‐mediated
   mitophagy. In contrast, Fis1 overexpression in 72‐h HG‐treated groups
   significantly restored Parkin‐mediated mitophagy (Figure [126]7G,H).
   They prevented the initiation of STX17‐mediated mitophagy
   (Figure [127]7I,J) compared with 72‐h HG‐treated VEC groups. Moreover,
   silencing Fis1 disrupted the association between Fis1 and STX17,
   starting at 12 h of high‐glucose exposure in HUVECs. Conversely, Fis1
   overexpression maintained the interaction between Fis1 and TBC1D15 and
   the association between Fis1 and STX17, even after 72 h of high‐glucose
   exposure (Figure [128]7M,N). These findings suggest that short‐term
   high‐glucose exposure upregulates Fis1 expression, enhancing its
   interaction with TBC1D15 and promoting Parkin‐mediated mitophagy. In
   contrast, long‐term high‐glucose exposure reduces Fis1 expression,
   inhibiting Parkin‐mediated mitophagy and promoting the translocation of
   STX17 from the ER to mitochondria, shifting mitophagy to an
   STX17‐mediated pathway. Furthermore, silencing Fis1 accelerates the
   switching from Parkin‐mediated to STX17‐mediated mitophagy.
   Overexpression of Fis1 prevents the switching from Parkin‐mediated to
   STX17‐mediated mitophagy following long‐term high‐glucose exposure.

2.6. Overexpression of Fis1 Alleviated High‐Glucose‐Induced Vascular
Endothelial Injury

   Mitochondrial fragmentation and a reduction in overall mitochondrial
   protein levels were observed in HUVECs treated with 72 h of high
   glucose compared to the NG group; both were reversed by Fis1
   overexpression (Figure [129]8A–C,G). Furthermore, Western blot analysis
   demonstrated that Fis1 overexpression alleviated ROS and mitochondrial
   ROS production (Figure [130]8D,E) and apoptosis (Figure [131]8F) caused
   by 72 h of high‐glucose exposure, as evidenced by decreased ROS and
   mitochondrial ROS production, a reduced BAX/BCL2 ratio, and lower
   cleaved caspase 3 expression. Additionally, Fis1 overexpression
   significantly increased eNOS phosphorylation following 72 h of
   high‐glucose exposure, indicating the role of Fis1 in enhancing
   endothelial function (Figure [132]8F). These observations suggested
   that Fis1 overexpression reduced ROS production and apoptosis and
   improved endothelial dysfunction in vascular endothelial cells exposed
   to long‐term high glucose.

Figure 8.

   Figure 8
   [133]Open in a new tab

   Fis1 overexpression alleviated high‐glucose‐induced vascular
   endothelial injury in HUVECs exposed to long‐term high‐glucose. HUVEC
   cell lines with Fis1 overexpression (OE‐Fis1) were constructed using
   lentivirus‐mediated transduction and then exposed to normal glucose (NG
   group), 12 h of high glucose (12HG group), and 72 h of high glucose
   (72HG group). A) HUVECs were stained with MitoTracker Red.
   Mitochondrial length B) and total number of fragmented, tubular, and
   elongated mitochondrial structures C) were quantitatively analyzed in
   each group. Scale bar = 10 µm. n = 6. D‐E) ROS and mitochondrial ROS
   production were measured by DCF and MitoSOX staining, respectively.
   Scale bar = 50 µm. n = 6. F) Representative Western blot images and
   corresponding relative intensities of BAX, BCL2, cleaved caspase 3,
   p‐eNOS and eNOS bands in HUVECs. n = 6. G) Representative Western blot
   images and corresponding relative intensities of TOM20, Cyto C, TIM23,
   and HSP60, representing proteins from the OMM, IMS, IMM, and matrix. n
   = 6. Data are expressed as mean ± SEM. ^* p < 0.05; ^** p < 0.01 versus
   NG + Vec group; ^# p < 0.05, ^## p < 0.01 versus 12HG + Vec group; ^°
   p < 0.05, ^°° p < 0.01 versus 72HG + Vec group. Vec, vector.

2.7. EC‐Specific Fis1 Silencing Exacerbated Hyperglycemia‐Induced Vascular
Endothelial Injury

   To investigate the role of Fis1 in hyperglycemia‐induced vascular
   endothelial injury, ApoE^−/− mice were injected via the tail vein with
   AAV9 (recombinant adeno‐associated virus 9) containing Fis1 shRNA,
   driven by the endothelial cell‐specific Tie promoter. Hyperglycemia was
   then induced in these mice by injecting STZ. Fis1 expression in the
   aortic intima of the mice injected with Fis1 shRNA was markedly reduced
   compared to mice injected with control shRNA, as confirmed by western
   blot (Figure [134]9B) and vascular en face fluorescence staining
   (Figure [135]9A). In situ images (Figure [136]9C,D) and Oil Red O
   staining (Figure [137]9E,F) of mouse aortas showed that Fis1 silencing
   in ECs accelerated and aggravated atherosclerosis lesions compared with
   control shRNA mice. Macrophage content (Figure [138]9G,H) and the
   smooth muscle cell content (Figure [139]9I,J) in atherosclerotic
   lesions were higher in mice transduced with Fis1 shRNA starting from 6
   weeks of hyperglycemia exposure and further aggravated following 20
   weeks of hyperglycemia exposure, compared to the control shRNA group.

Figure 9.

   Figure 9
   [140]Open in a new tab

   EC‐specific Fis1 silence exacerbated hyperglycemia‐induced vascular
   endothelial injury in STZ‐induced diabetic mice. ApoE^−/− mice were
   injected via the tail vein with recombinant adeno‐associated virus 9
   (AAV9) containing Fis1 shRNA or control shRNA, driven by the
   endothelial cell‐specific Tie promoter. The mice were then exposed to
   hyperglycemia for 0 weeks (HG 0 W), 6 weeks (HG 6 W), or 20 weeks (HG
   20 W). A) En face immunofluorescence staining of the abdominal aorta
   with immunofluorescent density analysis from the Con shRNA and Fis1
   shRNA groups. Red, green, and blue represent Fis1, VWF (an endothelial
   cell marker) and DAPI (a nuclear marker). Scale bar = 20 µm. n = 6. B)
   Representative Western blot images and corresponding relative
   intensities of Fis1 expression in the aorta from Con shRNA and Fis1
   shRNA groups. n = 6. C,D) Representative in situ images of
   atherosclerotic plaque in the aortic arch (C) with statistical analysis
   (D), expressed as % of total arcus aortae area. Scale bar = 1 mm. n =
   6. E,F) Oil Red O staining of cross‐sections at the aortic root (E)
   with statistical analysis (F), expressed as % of the total aortic root
   area. Scale bar = 100 µm. n = 6. G‐H) Representative images (G) and
   statistical analysis (H) of macrophage infiltration (F4/80)
   immunostaining in aortic root plaque, expressed as % of DAPI^+ lesion
   area. Scale bar = 100 µm. n = 6. I,J) Representative images (I) and
   statistical analysis (J) of α‐smooth muscle actin (α‐SMA)
   immunostaining in aortic root plaque, expressed as % of DAPI^+ lesion
   area. Scale bar = 100 µm. n = 6. Data are expressed as mean ± SEM. A,B)
   ^* p < 0.05; ^** p < 0.01. C–J) ^* p < 0.05; ^** p < 0.01 versus HG 0 W
   + Con shRNA group; ^# p < 0.05, ^## p < 0.01 versus HG 6 W + Con shRNA
   group; °p < 0.05, °°p < 0.01 versus HG 20 W + Con shRNA group.

   Consistent with in vivo data, oxidative stress (Figure [141]S5D,E,
   Supporting Information) and cell apoptosis (Figure [142]S5F, Supporting
   Information) were higher in 12‐h HG‐treated Fis1‐silenced HUVECs than
   in 12‐h HG‐treated control cells. Apoptosis and oxidative stress were
   further exacerbated in 72 h HG‐treated Fis1‐silenced HUVECs, as
   reflected by increased BAX/BCL2 ratio, elevated levels of cleaved
   caspase 3, and heightened levels of ROS and mitochondrial ROS
   production. Additionally, eNOS phosphorylation was markedly decreased
   in Fis1‐silencing HUVECs treated with high glucose for 12 h compared to
   the control group (Figure [143]S5F, Supporting Information). Fis1
   silencing also significantly increased mitochondrial fragmentation in
   12 h of HG‐treated HUVECs compared with 12 h of HG‐treated controls
   (Figure [144]S5C, Supporting Information). Consistent with these
   morphological changes, Western blot analysis showed a significant
   reduction in overall mitochondrial proteins in 12‐h HG‐treated
   Fis1‐silenced HUVECs compared with controls (Figure [145]S5G,
   Supporting Information). Both in vivo and in vitro experiments suggest
   that Fis1 may play a critical role in hyperglycemia‐induced vascular
   endothelial injury.

3. Discussion

   In the present study, we demonstrated that mitophagy switching drives
   vascular injury following prolonged high‐glucose exposure. Long‐term
   high‐glucose exposure reduces Fis1 expression, inhibiting
   Parkin‐mediated mitophagy and triggering the translocation of STX17
   from the endoplasmic reticulum to mitochondria in vascular endothelial
   cells. This switching from Parkin‐mediated to STX17‐mediated mitophagy
   drives endothelial dysfunction and vascular injury (Figure [146]10 ).
   Our study has identified mitophagy switching as an important factor in
   vascular endothelial injury, offering potential insights into
   therapeutic strategies for diabetic cardiovascular complications.

Figure 10.

   Figure 10
   [147]Open in a new tab

   The model depicts the distinct mitophagy pathways activated following
   short‐term and long‐term high‐glucose exposure in vascular endothelial
   cells. Short‐term high‐glucose exposure (left) elevates Fis1 expression
   and enhances the interaction between Fis1 and TBC1D15, promoting
   Parkin‐mediated mitophagy. Meanwhile, Fis1 inhibits the exposure of
   STX17's N‐terminus and prevents its translocation from the ER to the
   mitochondria. In contrast, long‐term high‐glucose exposure (right)
   reduces Fis1 expression, inhibiting Parkin‐mediated mitophagy. The
   reduction in Fis1 expression exposes the N‐terminus of STX17, promoting
   STX17 translocation from the ER to the mitochondria and leading to its
   accumulation in the mitochondria. This process initiates STX17‐mediated
   mitophagy, ultimately leading to vascular endothelial injury.

   Previous clinical studies suggest that there remains a continuing
   increased risk of death from diabetic cardiovascular complications
   despite improvements in glycemic control, a phenomenon known as
   “metabolic memory”.^[ [148]^3 , [149]^5 , [150]^23 ^] However, the
   critical mechanisms by which hyperglycemia triggers persistent vascular
   injury have not yet been elucidated fully. Our previous results showed
   that mitochondrial dynamics undergoes three stages during
   high‐glucose‐induced vascular injury: the compensation stage, balance
   shifting stage, and decompensation stage.^[ [151]^24 ^] During this
   progression, Fis1 expression significantly increased at the initial
   compensation and balance‐shifting stages. Once Fis1 expression
   decreased below normal levels and maintained at a low level, this
   change signified the transition into the decompensation stage. In the
   present study, “short‐term” and “long‐term” high‐glucose exposure were
   defined based on the changes in Fis1 expression levels. During the Fis1
   upregulation stage, persistent vascular damage had not yet developed,
   which was defined as short‐term high‐glucose exposure. Both 12 h of
   high‐glucose exposure in vitro and 6 weeks of hyperglycemia exposure in
   vivo showed no endothelial damage, which was classified as short‐term
   high‐glucose exposure. As Fis1 expression declined below normal levels,
   vascular endothelial cells underwent persistent apoptosis, defining
   this stage as long‐term high‐glucose exposure. Both 72 h of
   high‐glucose exposure in vitro and 20 weeks of hyperglycemia exposure
   in vivo led to endothelial cell apoptosis, which was classified as
   long‐term high‐glucose exposure. In diabetic patients, “short‐term” and
   “long‐term” high‐glucose exposure may also be defined based on changes
   in Fis1 expression levels. Vita et al. reported that in diabetic
   patients ≈50 years old, vascular function remained unchanged, while
   their Fis1 expression was increased in endothelial cells freshly
   isolated from these patients compared to controls.^[ [152]^25 ^] This
   finding is consistent with our in vivo and in vitro results of
   short‐term high‐glucose exposure. The reduced Fis1 expression in
   endothelial cells of diabetic patients has not yet been reported.
   However, Houzelle et al. reported that Fis1 protein levels were
   significantly decreased in the muscles of individuals with type 2
   diabetes compared to athletes, and these diabetic patients exhibited
   impaired insulin sensitivity.^[ [153]^26 ^] We speculate that their
   vascular endothelial function may also be impaired at this stage, which
   merits further investigation in future clinical studies. Although our
   in vivo and in vitro experiments do not fully reflect clinical
   conditions, future clinical studies may define “short‐term” and
   “long‐term” high‐glucose exposure based on changes in Fis1 levels in
   endothelial cells of diabetic patients.

   To investigate the mechanisms underlying hyperglycemia‐induced vascular
   injury, we conducted a 4D‐label‐free quantitative proteomic analysis of
   whole aortas from three groups of diabetic ApoE^−/− mice: HG 0 W, HG
   6 W, and HG 20W. Compared to HG 0 W, the differentially expressed
   proteins in the HG 6 W group differed from those in the HG 20 W group,
   but both were enriched in “mitochondrion” and “lysosome” pathways,
   which suggest that mitochondrial and autophagy‐related mechanisms may
   play crucial roles in hyperglycemia‐induced vascular injury. Due to
   limitations in observing mitochondria and lysosomes in vivo,
   mitochondrial‐lysosome interactions, co‐localization of LC3 and TOM20,
   and mt‐Keima measurements were observed in HUVECs. These findings
   demonstrate that mitophagy is activated following both short‐term and
   long‐term high‐glucose exposure. Classic Parkin‐mediated mitophagy is
   activated following short‐term high‐glucose exposure, whereas long‐term
   exposure suppresses it. Interestingly, silencing Parkin did not reduce
   the overall mitophagy levels induced by long‐term high‐glucose
   exposure, suggesting that alternative mitophagy pathways are activated
   under these conditions. Long‐term high‐glucose exposure triggered a
   Parkin‐independent, membrane depolarization‐independent mitophagy
   pathway, as evidenced by tracking daughter mitochondria resulting from
   mitochondrial fission and observing their subsequent outcomes. Further
   quantitative proteomic analysis and experiments demonstrated that
   long‐term high‐glucose exposure reduces Fis1 expression, thereby
   inducing STX17‐mediated mitophagy. This process promotes the
   translocation of STX17 from the ER to the mitochondria, ultimately
   leading to excessive mitochondrial clearance and cell apoptosis. In
   summary, long‐term exposure triggers a switch in mitophagy from
   Parkin‐mediated to STX17‐mediated mitophagy. Additionally, we found
   that Fis1 plays a key role in this switching process.

   ECs are essential in the onset and progression of vascular
   complications in diabetes.^[ [154]^27 ^] Although mitochondria comprise
   just 2% to 6% of the cytoplasmic volume in ECs and have a limited
   energy supply,^[ [155]^28 ^] they are considered vital signaling hubs
   in responding to oxidative stress, calcium buffering, and apoptosis.^[
   [156]^29 ^] Mitophagy, as a regulatory mechanism for maintaining
   mitochondrial homeostasis, selectively removes damaged and depolarized
   mitochondria, thereby protecting ECs from hyperglycemia‐induced
   injury.^[ [157]^30 ^] Several drugs have been found to promote
   mitophagy, mitigating high‐glucose‐induced EC damage.^[ [158]^31 ^]
   However, it is worth noting that mitophagy in ECs is not always
   beneficial and may also result in cell damage^[ [159]^32 ^] and
   death.^[ [160]^33 ^] For instance, vitamin D and salvianolic acid B
   have been shown to inhibit mitophagy, protecting ECs from apoptosis.^[
   [161]^13 ^] Additionally, dynamic changes in mitophagy occur during
   disease progression. Increased mitophagy is an early response to
   mitochondrial damage, but stalled mitophagy can exacerbate symptoms as
   the condition progresses.^[ [162]^34 ^] Acute lipid exposure promotes
   cellular autophagy as a protective mechanism, and prolonged lipid
   stimulation inhibits autophagy by disrupting autophagosome‐lysosome
   fusion.^[ [163]^35 ^] Thus, mitophagy acts as a double‐edged sword for
   endothelial cells, as it can cause different results depending on
   disease progression. In this study, short‐term high‐glucose exposure
   activated Parkin‐mediated mitophagy, while long‐term high‐glucose
   exposure activated STX17‐mediated mitophagy, leading to different
   outcomes. Parkin‐mediated mitophagy is highly selective, targeting
   damaged mitochondria with membrane depolarization. This pathway
   efficiently eliminates damaged mitochondria, protecting cells from
   oxidative stress and apoptosis.^[ [164]^14 , [165]^15 ^]
   High‐glucose‐induced vascular endothelial injury was accelerated and
   aggravated by si‐Parkin, suggesting that Parkin‐mediated mitophagy
   plays a critical role in protecting against endothelial injury during
   short‐term high‐glucose exposure. In contrast, STX17‐mediated mitophagy
   represents a Parkin‐independent pathway targeting damaged and healthy
   mitochondria.^[ [166]^36 ^] In this pathway, the loss of Fis1 causes
   STX17 to translocate from the endoplasmic reticulum (ER) to the
   mitochondria, where it recruits essential autophagic components such as
   ATG14, facilitating mitophagosome formation. This process triggers
   non‐selective mitophagy, potentially leading to excessive mitochondrial
   clearance. Xian et al. reported that HeLa cells exhibited a
   significantly increased population of dead cells with permeable plasma
   membranes during STX17‐mediated mitophagy, substantiating the role of
   STX17‐mediated mitophagy in cell death.^[ [167]^19 ^] Our results
   indicate that endothelial cells exposed to long‐term high glucose may
   activate STX17‐mediated mitophagy, leading to excessive mitochondrial
   degradation, ultimately resulting in apoptosis. Silencing of STX17
   alleviated high‐glucose‐induced endothelial cell apoptosis and vascular
   endothelial injury, highlighting the detrimental role of STX17‐mediated
   mitophagy in maintaining vascular endothelial homeostasis. As
   STX17‐mediated mitophagy is a relatively newly discovered pathway, the
   specific mechanism by which STX17‐mediated mitophagy causes cell injury
   or dysfunction remains to be further studied. Based on these findings,
   this work suggests a potential therapeutic strategy: selectively
   activating Parkin‐mediated mitophagy while inhibiting STX17‐mediated
   mitophagy to protect ECs from high‐glucose‐induced injury rather than
   focusing solely on the activation or inhibition of overall mitophagy
   levels.

   Mitochondrial fission protein 1 (Fis1), located on the outer
   mitochondrial membrane, plays a crucial role in the process of
   mitochondrial fission.^[ [168]^37 ^] Growing evidence indicates that
   Fis1 also has a prominent role in mitophagy.^[ [169]^38 ^] While the
   exact mechanistic details of Fis1 in mitophagy remain unclear.
   Recently, A groundbreaking discovery indicated that mitochondrial
   peripheral fission (occurring within 25% of the tip) divides
   metabolically healthy organelles from unhealthy ones, with the damaged
   daughter mitochondrion subsequently removed through mitophagy, and Fis1
   is essential for regulating peripheral fission.^[ [170]^15 , [171]^20
   ^] Fis1 facilitates peripheral fission to promote cell survival under
   stress by interacting with TBC1D15, a GTPase‐activating protein that
   modulates RAB7 GTP hydrolysis.^[ [172]^22 , [173]^39 ^] This peripheral
   fission process is crucial for facilitating Parkin‐mediated mitophagy,
   a pathway critical for the selective removal of damaged mitochondria
   and maintaining mitochondrial quality control. Peripheral mitochondrial
   fission has risen under different cellular stress conditions, including
   oxidative stress and nutrient deprivation, highlighting its role in
   cellular adaptation and survival.^[ [174]^15 ^] However, the role of
   peripheral fission in mitochondrial dynamics and its potential
   contribution to mitophagy in high‐glucose‐induced endothelial
   dysfunction still need to be explored. Our study revealed that
   short‐term high‐glucose exposure enhanced the Fis1‐TBC1D15 interaction,
   promoting peripheral mitochondrial fission and activating
   Parkin‐mediated mitophagy as a protective mechanism to mitigate the
   acute mitochondrial stress induced by hyperglycemia. However, long‐term
   high‐glucose exposure suppressed this interaction, inhibiting
   Parkin‐mediated mitophagy and shifting toward STX17‐mediated mitophagy.
   Fis1‐mediated STX17 mitophagy represents a distinct, Parkin‐independent
   mitophagy pathway.^[ [175]^18 , [176]^19 ^] In this pathway, the loss
   of Fis1 causes STX17 to translocate from the endoplasmic reticulum (ER)
   to mitochondria, where it recruits essential autophagic components such
   as ATG14, facilitating the formation of mitophagosomes.^[ [177]^19 ,
   [178]^40 ^] This process triggers non‐selective mitophagy, potentially
   leading to excessive mitochondrial clearance.^[ [179]^19 ^] However,
   the role of STX17‐initiated mitophagy in various physiological states
   remains unclear. Our results indicated long‐term high‐glucose exposure
   induced STX17‐mediated mitophagy, leading to excessive mitochondrial
   clearance and cell apoptosis. Silencing STX17 alleviated
   high‐glucose‐induced vascular endothelial injury. Furthermore,
   silencing Fis1 inhibited Parkin‐mediated mitophagy while inducing
   STX17‐mediated mitophagy, which led to endothelial damage following
   short‐term high‐glucose exposure in HUVECs. In ApoE^−/− mice,
   EC‐specific Fis1 silencing exacerbated vascular endothelial injury
   following long‐term high‐glucose exposure. These findings suggest that
   Fis1 is essential in hyperglycemia‐induced vascular injury, indicating
   its potential as a therapeutic target. While Fis1 regulation has shown
   benefits in various disease models, our study revealed that
   overexpression of Fis1 during long‐term high‐glucose exposure reduced
   endothelial cell injury. These findings suggest that enhancing Fis1
   expression might be beneficial to restore mitophagy balance and reduce
   cellular damage following long‐term high‐glucose exposure.

   We have demonstrated that high‐glucose exposure triggers switching from
   Parkin‐mediated to STX17‐mediated mitophagy; whether other pathogenic
   factors can initiate this switching remains unclear. Additionally, the
   precise regulatory mechanisms controlling mitophagy switching require
   further investigation. In summary, our findings reveal that mitophagy
   switching from Parkin‐mediated to STX17‐mediated mitophagy drives
   vascular endothelial injury, providing valuable perspectives on
   therapeutic strategies for diabetic cardiovascular complications.
   Mitophagy switching may serve as an indicator of disease progression in
   high‐glucose‐induced vascular injury in diabetic patients.

4. Experimental Section

Animals and Treatments

   The apoE knockout mice (C57BL/6 background, 5 weeks old) were obtained
   from Cyagen Biosciences. Six‐week‐old male mice were given daily
   intraperitoneal injections of streptozotocin (STZ) at 55 mg kg^−1 for
   five consecutive days to induce hyperglycemia. One week after the last
   injection, mice with fasting blood glucose levels of ≥16.7 mmol L^−1
   were chosen for the study. The hyperglycemic mice were separated into
   two groups: one group was fed for 6 weeks (HG 6 W group) and the second
   group for 20 weeks (HG 20 W group). Aortas were collected from both
   groups at the respective time points. Control mice (HG 0 W group) were
   administered vehicle injections of 0.1 mol L^−1 citrate‐phosphate
   buffer (pH 4.5). They had aortas collected one week after the
   injections.

   Genomeditech Biotechnology (Shanghai, China) constructed the
   adenoviruses. To specifically knock down Fis1 protein expression in the
   endothelium of ApoE^−/− mice, adeno‐associated virus serotype 9 (AAV9)
   carrying Fis1 shRNA driven by the endothelial‐specific Tie2 promoter
   was used. The Fis1 shRNA targeting sequence was
   5′‐CCTGATTGATAAGGCCATGAA‐3′, and the control shRNA sequence was
   5′‐TTCTCCGAACGTGTCACGT‐3′. Mice were injected via the tail vein with
   recombinant adeno‐associated virus 9 (AAV9) containing Fis1 shRNA (1.5
   × 10¹¹ vector genomes per mouse) under the supervision of the Tie2
   promoter. The mice were then divided into HG 0 W, HG 6 W, and HG 20 W
   groups as described above. Aortas were collected for further analysis.
   All mice had free access to standard chow and water and were fasted for
   8 h before taking blood glucose measurements.

Reagents and Antibodies

   The reagents listed below were utilized: MitoTracker Red FM
   (ThermoFisher, [180]M22425), LysoTracker Green (ThermoFisher, L7526),
   MitoSOX Red (ThermoFisher, [181]M36008), DCF (ThermoFisher, D399),
   tetramethylrhodamine ethyl ester perchlorate (Sigma, 87917), DAPI
   (ThermoFisher, 62248), Hoechst 33342 (ThermoFisher, 62249), TUNEL
   Apoptosis Assay Kit (keygentec, KGA1406), Mitochondrial Isolation Kit
   (Invent, MP‐007), ER Enrichment Kit (Invent, ER‐036), The reagents for
   Western blotting were obtained from Yeasen (Shanghai, China).

   The following Antibodies were used: F4/80 (Proteintech, 29414, 1:200),
   α‐SMA (Proteintech, 14395, 1:200), p62 (Cell Signaling Technology,
   23214, 1:5000) (Proteintech, 18420, 1:3000), LC3 (Cell Signaling
   Technology, 3868, 1:4000) (Cell Signaling Technology, 83506, 1:1000),
   vWF (Proteintech, 66682, 1:500), BAX (Proteintech, 60267, 1:5000), BCL2
   (Proteintech, 60178, 1:1000), Caspase 3 (Proteintech, 66470, 1:2000),
   eNOS (Cell Signaling Technology, 9572, 1:2000), Phospho‐eNOS (Ser1177)
   (Cell Signaling Technology, 9570, 1:1500), Beta Actin (Proteintech,
   66009, 1:10000), Tom20 (Proteintech, 11802, 1:5000), Tim23
   (Proteintech, 67535, 1:2000), Cyto C (Proteintech, 66264, 1:5000),
   HSP60 (Proteintech, 66041, 1:10000), Parkin (Abcam, ab15494, 1:2000),
   STX17 (Abcam, ab229646, 1:1000), Fis1 (Abcam, ab229969, 1:3000),
   TBC1D15 (Abcam, ab121396, 1:1000), 3‐NT (EMD Millipore, 06–284,
   1:1000), 8‐OHdG (Abcam, ab62623, 1:1000), COXIV (Proteintech, 11242,
   1:5000), Calnexin (Proteintech, 10427, 1:5000).

4D‐Label Free Proteomics Analysis

   The aortas of mice in the HG 0 , HG 6 , and HG 20 W groups were
   separated and subjected to 4D‐label‐free quantitative proteomic
   analysis. The main experimental steps for this analysis involved
   protein extraction and digestion, LC‐MS/MS analysis, and protein
   identification and quantification, followed by bioinformatic analysis.
   MaxQuant 1.6.14 software was used for protein identification and
   quantitation. Differentially expressed protein sequences were first
   analyzed using NCBI BLAST+ to identify homologous sequences, then GO
   terms were mapped and annotated with Blast2GO. Proteins were also
   compared against the KEGG database ([182]http://geneontology.org/) to
   obtain KEGG orthology IDs. The proteomic analysis was supported by
   APTBIO (Shanghai, China).

Histological Analyses

   Aortic roots from mice were sectioned and stained using Oil Red O
   (Solarbio, China) according to the manufacturer's instructions.
   Cryosections of mouse aortic roots were fixed and incubated with 5%
   BSA. Sections were incubated with F4/80 or α‐SMA primary antibodies.
   After three washes with PBS, fluorescently labeled secondary antibodies
   (ThermoFisher, A‐21245, 1:1000) were applied. DAPI was used to stain
   nuclei. Fluorescent images of the aortic root tissues were captured
   using the Olympus FV3000 microscope.

   The fresh thoracic aortas from mice were fixed in 4% formaldehyde,
   dehydrated, and then embedded in paraffin. Immunohistochemical staining
   was performed to detect 3‐nitrotyrosine (3‐NT) and
   8‐oxo‐29‐deoxyguanosine (8‐OHdG). Hematoxylin kits (Solarbio, China)
   were then used for counterstaining. A semi‐quantitative analysis based
   on staining intensity (0: no staining, 1: faint yellow, 2: light brown,
   3: brown) and the proportion of positive cells (0: 0%–5%, 1: 6%–25%, 2:
   26%–50%, 3: 51%–75%, 4: 76%–100%). Paraffin‐embedded sections of mouse
   abdominal aortas were dewaxed, hydrated, and blocked. The TUNEL
   Apoptosis Assay Kit assessed apoptosis following the manufacturer's
   instructions.

   Mouse abdominal aortas were dissected, opened longitudinally, and fixed
   in 4% paraformaldehyde. The aortas were incubated with primary
   antibodies against LC3, P62, and Fis1. Following three washes in PBS,
   fluorescently labeled secondary antibodies were applied to the aortas.

Cell Culture and Cell Transfection

   HUVECs, obtained from the American Type Culture Collection (Manassas,
   VA, USA), were cultured in 10% FBS. HUVECs were exposed to a normal
   glucose medium (NG, 5.5 mm) for 0, 60, or 72 h and then shifted to a
   high‐glucose medium (HG, 33 mm). All cells were cultured for 72 h
   before harvesting, resulting in exposure to high glucose for 72, 12, or
   0 h, respectively.

   The small interfering RNAs (siRNAs) were synthesized by Invitrogen,
   with the following sequences: Control siRNA: 5′‐UUCUCCGAACGUGUCACGU‐3′;
   si‐Parkin: 5′‐GGAGUGCAGUGCCGUAUUU‐3′; and STX17 siRNA:
   5′‐GAAAGTCCGAAAGGATGACCTAGTA‐3′. Using Lipofectamine 3000 reagent
   (Invitrogen) and following the manufacturer's protocol, HUVECs were
   transfected with the specified siRNAs.

Lentiviral Transfection

   Lentiviruses carrying Fis1 shRNA (sh‐Fis1; 5′‐AGGCCTTAAAGTACGTCCG‐3′)
   and control lentivirus (Vec; 5′‐TTCTTCGAACGTGTCACGT‐3′), lentiviruses
   for Fis1 overexpression (OE‐Fis1) and control (Vec) lentivirus were
   designed by Hanbio Biotechnology (Shanghai, China). HUVECs were seeded
   in 24‐well plates at 30%–50% confluence and infected with the specific
   lentiviruses at a multiplicity of infection (MOI) 30. Transfection
   efficiency was evaluated by fluorescence microscopy 72 h post‐infection
   (Olympus Corporation, Japan). Puromycin was added to select cells
   successfully virus‐infected.

Mitophagy Analysis by Mt‐Keima Transfection

   Mitophagy was monitored using mitochondria‐targeted Keima (mt‐Keima),
   which emits distinct colors in neutral and acidic environments. The
   mt‐Keima adenovirus was transfected following the manufacturer's
   instructions (Hanbio Biotechnology, Shanghai, China). Mt‐Keima signals
   were detected at 440 or 590 nm, where excitation at 440 nm represents
   mitochondria in the cytoplasm (red), and excitation at 590 nm indicates
   mitochondria in acidic lysosomes (green).

Measurement of Oxidative Stress

   ROS levels, indicative of oxidative stress in HUVECs, were assessed
   using DCF staining. MitoSox was used to evaluate mitochondrial ROS
   levels. Fluorescence from both ROS and mt‐ROS was detected using
   confocal microscopy (Olympus FV3000).

Cell Immunofluorescence

   Cells grown in confocal dishes were fixed. After blocking with 3%
   FBS/PBS, the cells were incubated with primary antibodies, followed by
   Alexa Fluor‐conjugated secondary antibodies. Nuclei were stained with
   Hoechst 33 342. Images were captured using a confocal microscope
   (Olympus FV3000), and co‐localization analysis was carried out with
   ImageJ software.

Mitochondrial and Lysosome Imaging

   Cells were treated with 500 nm Mitotracker Red for 5 min of incubation.
   The Opera Phenix High Content Screening System (PerkinElmer, Waltham,
   MA, USA) was used to acquire images, and mitochondrial quantification
   was conducted using Harmony software version 4.9. Live‐cell imaging was
   carried out in a chamber heated to 37 °C with 5% CO₂. Mitochondrial
   fissions were characterized as instances in live cells where a single
   mitochondrion divided into two independently moving daughter
   mitochondria, with the constriction site diameter measuring less than
   180 nm. The mitochondrial skeleton was mapped to calculate the total
   length of the mitochondrion prior to fission. Peripheral fission was
   defined as occurring within 25% of the tip, while midzone fission
   occurred within the central 50% of the mitochondrion. Fission events
   were recorded as the number of fissions per 10 min in a single field of
   view.

   Mitochondrial morphology was classified into three categories according
   to length: fragmented (less than 3 mm), tubular (3–8 mm), and elongated
   (greater than 8 mm). At least 200 mitochondria were analyzed for each
   experimental group.

   Cells were treated with 500 nM MitoTracker Red and 500 nm LysoTracker
   Green. Images were taken using a confocal microscope (Olympus FV3000),
   and co‐localization analysis was performed with ImageJ software.

Mitochondrial Membrane Potential

   Cells received 500 nm TMRE treatment. The fluorescence intensity is
   positively correlated with membrane potential, where the fluorescence
   intensity is represented by color gradients.

Subcellular Fractionation

   Mitochondrial and ER fractions were isolated using the Mitochondrial
   Isolation Kit and ER Enrichment Kit (Invent Biotechnologies), following
   the manufacturer's protocols.

Western Blotting and Co‐Immunoprecipitation

   Cells were lysed in RIPA buffer. Protein samples were run on SDS‐PAGE
   and then incubated with primary antibodies. Following incubation with
   secondary antibodies, signals were detected using the ECL Kit and
   analyzed with ImageJ software.

   Protein complexes were isolated using a co‐immunoprecipitation (Co‐IP)
   kit (Thermo Fisher Scientific). Cells were lysed in IP lysis buffer,
   and protein samples were incubated with the primary antibody. Pierce
   protein A/G beads were used to capture antigen‐antibody complexes.
   After being washed three times with washing buffer, the beads were used
   to elute the antigen‐antibody complexes, which were then analyzed by
   Western blot.

Electron Microscopy

   Cells were fixed with 2.5% glutaraldehyde, washed three times in PBS,
   and embedded in agarose. They were then post‐fixed in 1% osmium
   tetroxide and dehydrated through a graded ethanol series (30%, 50%,
   70%, 80%, 95%, and 100%) for 20 min each, followed by two changes of
   acetone for 15 min. Next, the cells were infiltrated with EMBed 812
   resin and polymerized overnight at 65 °C. Ultra‐thin sections
   (60–80 nm) were cut and collected on copper grids. The sections were
   stained with 2% uranyl acetate and 2.6% lead citrate before imaging
   with a transmission electron microscope (HITACHI, HT7800). Quantitative
   analysis of mitophagosome numbers was performed across randomly
   selected fields of view.

Statistical Analysis

   All data were presented as mean ± SEM, and the numbers of independent
   experiments were indicated. Student's t‐test was used to compare two
   groups, while one‐way ANOVA with Tukey's post‐test was applied to
   compare differences between multiple experimental groups. Data analysis
   was performed using GraphPad Prism software version 6.0 (SanDiego, CA,
   USA). Statistical significance was defined as P < 0.05.

Ethics Approval Statement

   The animal study protocol was approved by the Ethics Committee of
   Hainan Medical University (approval number: HYLL‐2023‐185).

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   A.L. and Y.Z. performed conceptualization. A.L., J.G., and S.W.
   performed methodology. R.W. and Y.J. performed validation. A.L. and
   R.W. performed formal analysis. A.L., J.G., S.W., and Y.J. performed
   investigation. X.L., H.D., H.L., and Y.Z. performed resources. X.L. and
   Y.Z. performed data curation. A.L. and Y.Z. wrote the original draft
   preparation. X.L. and Y.Z. wrote reviewed and edited the final
   manuscript. C.Y. and Z.C. performed visualization. H.D., H.L., and X.L.
   performed supervision. C.Y. and Z.C. performed project administration.
   X.L., H.D., and Y.Z. acquired funding acquisition. All authors have
   read and agreed to the published version of the manuscript.

Supporting information

   Supporting Information
   [183]ADVS-12-2414960-s001.docx^ (8.2MB, docx)

Acknowledgements