Distal Venous Arterialization Modulates Autophagy-Related Markers in Skeletal Muscle in a Rat Hindlimb Ischemia Model

Introduction

Peripheral arterial disease (PAD) affects over 113 million people worldwide and is a major cause of disability and limb loss.¹ Also known as lower extremity arterial disease (LEAD), PAD is characterized by atherosclerotic narrowing or occlusion of lower extremity arteries, resulting in insufficient perfusion and progressive ischemia.^1–3 Clinical manifestations range from intermittent claudication to ischemic rest pain and gangrene.⁴ Management depends on severity and includes lifestyle modification, pharmacotherapy, surgical bypass, or endovascular revascularization.³ In patients without viable distal runoff, conventional revascularization is not feasible, and major amputation remains the only option—accounting for approximately 51–93% of all lower extremity amputations.⁵ Distal venous arterialization (DVA) has emerged as an alternative revascularization strategy for such inoperable limbs.⁶ By creating an arterial–venous anastomosis, DVA redirects blood flow into the venous system to reperfuse ischemic tissue. While clinical studies and meta-analyses have demonstrated promising outcomes,⁷ the mechanisms by which DVA influences skeletal muscle physiology remain poorly defined. Autophagy, a cellular process responsible for degrading and recycling damaged proteins and organelles, plays a critical role in skeletal muscle adaptation.^8,9 During hypoxia, autophagy is regulated by hypoxia-inducible factor 1-alpha (HIF-1α) and AMP-activated protein kinase (AMPK).^10–13 HIF-1α promotes autophagy through BNIP3-mediated pathways, while AMPK functions as an energy sensor that activates autophagy during metabolic stress.¹⁴ Although autophagy is protective under moderate stress, excessive activation has detrimental effects, potentially leading to cell death.^8,12,13,15 To date, no experimental studies have specifically examined the effects of DVA on ischemic skeletal muscle and its regulation of autophagy.⁷ We therefore hypothesized that DVA would improve skeletal muscle function in a rat hindlimb ischemia (HLI) model and would be associated with modulation of autophagy-related signaling. This study was conducted in accordance with the ARRIVE 2.0 guidelines. Although this femoral artery ligation model does not fully replicate the chronic, multifactorial progression of clinical chronic limb-threatening ischemia, it reliably produces sustained reductions in perfusion, functional impairment, and hypoxia-related molecular responses. Therefore, it serves as a controlled experimental platform to investigate mechanistic adaptations to ischemia and the physiological effects of revascularization strategies such as distal venous arterialization.

Materials and Methods

Animals

Thirty healthy male Wistar rats (Rattus norvegicus), aged 18–20 weeks and weighing 300–350 g, were obtained from PT. Bio Farma (LLC, Indonesia). All animals were housed under a 12-hour light/12-hour dark cycle at a controlled room temperature of 26–29 °C, with ad libitum access to drinking water and standard laboratory feed. A 4-week acclimatization period was conducted at the Animal Husbandry Laboratory, Universitas Padjadjaran.

Grouping

This was an experimental microsurgical study. On day 0, all rats underwent ligation of the right common femoral artery to establish a hindlimb ischemia (HLI) model. On day 7, animals were allocated into three groups: (1) the distal venous arterialization (DVA) group, (2) the opened ligation (OL) group, and (3) the persistent ligation (HLI) group serving as ischemic controls. All groups were terminated on day 14 for skeletal muscle sampling. Untreated contralateral (left) limbs were used as sham controls to minimize animal use.

Sample size was determined using the Federer formula, requiring a minimum of six animals per group plus 1–2 reserve animals. To reduce the risk of technical failure, the DVA group—considered the most challenging microsurgical procedure—included 15 animals, while the OL and HLI groups consisted of 8 and 7 animals, respectively.

Randomization

Randomization was performed prior to HLI induction using permuted blocks of unequal size. Block allocations followed a 15:8:7 ratio (DVA:OL:HLI), consisting of blocks of six (3 DVA, 2 OL, 1 HLI) and blocks of twelve (6 DVA, 2 OL, 4 HLI). Three blocks of six and one block of twelve were employed. Allocation concealment and blinding were implemented by laboratory staff independent of the surgical and assessment teams.

Exclusion Criteria and Disqualification

Animals showing signs of illness or gait abnormalities before treatment were excluded. Disqualification criteria included unintended injury, perioperative death, surgical wound infection, foot gangrene, or technical failure of DVA or OL procedures. DVA failure was defined as thrombus formation at the arterial–venous anastomosis site, while OL failure was defined as thrombosis in the previously ligated artery upon tissue harvesting. Only healthy animals meeting predefined inclusion criteria were enrolled in the study. Animals with signs of illness, injury, or abnormal baseline conditions were not included prior to randomization. No animals were excluded after study initiation or from the final analysis.

Surgical Technique

Preoperative Preparation

Prior to surgery, body weight was recorded, gait footprints were obtained for subsequent analysis,¹⁶ and thermal images of both hindlimbs were taken. Preoperative fasting was not required. Anesthesia was induced with an intraperitoneal injection of ketamine (Ket-A-100^®, 200mg/kg) and xylazine (Xyla^®, 28mg/kg) prepared in a 1 mL syringe. Unconsciousness was achieved within approximately 5 minutes and lasted for ~45 minutes. Intubation, oxygen supplementation, and continuous monitoring were not required.

Hindlimb Ischemia (HLI) Model Creation

The groin and medial thigh were shaved, and the rat was fixed in a supine position on a wooden board covered with an underpad sheet. The skin was prepared with 70% alcohol and 10% povidone-iodine solution. The board was placed in a tray positioned under an electronic surgical microscope (Eakins^® digital microscope). A 7.0× surgical loupe was also used to aid magnification.

A longitudinal incision was made in the medial thigh using a No. 11 scalpel blade. The common femoral artery (CFA) and its branches, the superficial femoral artery (SFA) and deep femoral artery (DFA), as well as the femoral vein (FV), were carefully dissected. The CFA was isolated from the FV and secured with 7–0 polypropylene sutures (Optilene^®, C3090545). Small arterial branches were identified but not ligated. The CFA was then ligated with a reef knot using the same suture. The incision was closed with 4–0 absorbable monofilament glyconate sutures (Monosyn^®). While still anesthetized, a second thermal image was obtained (Figure 1A–D). Postoperatively, oral paracetamol syrup (120 mg/5 mL; 0.1 mL = 4.5 mg, two drops by pipette) was administered. Rats were returned to warmed cages and monitored daily for complications.

Figure 1 (A) CFA identification prior ligation; (B) Ligation with 7/0 Polypropylene; (C and D) Thermal imagin before and after ligation, the right hind limb temperature is significantly lower, marked with blue color (D); (E–H) Microsurgical technique of Deep Venous Arterialization (DVA). Note, the vein was cut as proximal as possible (blue arrow (F)), while the artery was cut as distal as possible (red arrow (F)); old artery ligation that was not remove (yellow arrow (F)).

Postoperative Clinical Assessment

On postoperative days 2 and 7, animals were observed for clinical signs of ischemia, including skin color changes, coldness, muscle atrophy, wound infection or dehiscence, temperature alteration, and gait impairment. Observations were documented photographically using a smartphone. Functional assessment was performed using the Tarlov score, ischemia score, and modified ischemia score.^17,18 Gait analysis was performed by manual footprint analysis as described by Wertman et al¹⁶ Rats showing infection, wound dehiscence, or death were excluded.

Microsurgical Technique for Distal Venous Arterialization (DVA)

A re-incision was made along the previous surgical line. The ligated artery was identified by locating the polypropylene suture knots. The CFA and FV were carefully dissected and controlled using bulldog clamps placed proximally and distally on both vessels. The artery was transected proximal to the ligation knot, and the vein was transected just distal to the proximal clamp. The proximal venous stump was ligated with 7–0 polypropylene sutures. The distal vein was cannulated with a 0.014” soft cardiac angioplasty wire to disrupt venous valves. The operative field was flushed with 1250 units of heparin in 10 mL saline to prevent coagulation.¹⁹

An end-to-end arterial–venous anastomosis was created with 10–0 nylon sutures (Ethilon^®, BVH100-3, Ethicon) using interrupted sutures (Figure 1E–H). Following anastomosis, distal clamps were released first, followed by proximal clamps. Patency was confirmed by visible pulsation in both artery and vein. Any leakage was corrected with additional sutures. The wound was closed as previously described, followed by immediate postoperative thermal imaging.

Microsurgical Technique for Open Ligation (OL)

For the OL group, the previously ligated CFA was identified by tracing the polypropylene loop suture. The knot was carefully cut using microsurgical scissors to avoid arterial injury. Once the ligature was released, the wound was closed using the same technique, and thermal imaging was performed. The OL group served as an experimental positive control to model restoration of arterial inflow and reperfusion, rather than as a clinically analogous intervention.

Animal Termination and Tissue Harvesting

On day 14, endpoint assessments were performed. Animals first underwent gait analysis, followed by induction of general anesthesia with ketamine–xylazine (intraperitoneal). Infrared thermography was then conducted, after which terminal re-exploration was performed to confirm vascular patency in each group: (i) DVA anastomosis, (ii) arterial patency in the OL group, and (iii) persistence of ligation in the HLI group. While under deep anesthesia, animals were euthanized by intraperitoneal overdose of ketamine–xylazine, in accordance with institutional guidelines.²⁰ Death was confirmed by loss of reflexes, cessation of spontaneous respiration and heartbeat, and verified by thoracotomy. After a 10-minute postmortem interval, the right gastrocnemius muscle (and contralateral muscle for sham controls) was harvested. Each sample was divided: one portion was snap-frozen in liquid nitrogen and stored at −80 °C, and the other was fixed in 4% neutral-buffered formalin and embedded in paraffin.

Gait Analysis

Gait analysis was performed using manual footprint analysis¹⁶ to assess hindlimb locomotion. Immediately before testing, the plantar surfaces of both hind paws were coated with non-toxic, water-washable paint in contrasting colors (Flomo^® washable paint). Rats traversed a transparent tunnel lined with white paper toward a darkened goal box. Paper was replaced between runs. Footprints were digitized and analyzed by a blinded assessor. Primary outcomes were (i) mean stride length for the right and left hindlimbs, and (ii) mean base-of-support (stride width).

Hindlimb Temperature Analysis with Infrared Thermography (IRT)

Intraoperative patency of the DVA anastomosis was verified by direct observation of arterial pulsation within the venous segment and confirmed after wound closure using infrared thermography (IRT; HIKMICRO^® Mini1 smartphone module, model HM-TJ11-3AMF-Mini1; Hangzhou Microimage Software). Longitudinal IRT was performed at three time points: baseline (day 1, pre-treatment), day 7 (post-HLI induction), and day 14 (pre-termination). All measurements were performed under standardized room conditions (26°C, no artificial lighting). The examiner handled only the upper body of rats to avoid heat transfer to the hindlimbs. Thermal images were analyzed in ImageJ v1.54g (NIH). Images were converted to RGB stacks, standardized regions of interest were applied over the hindlimbs, and pixel intensities were calibrated against the embedded scale to yield quantitative temperature values (°C).

Immunohistochemical Analysis for Capillary Density

Capillary density was assessed as a histologic surrogate of angiogenesis.^21–23 Formalin-fixed, paraffin-embedded gastrocnemius sections were stained with a monoclonal anti-CD31 (PECAM-1) antibody (GeneTex, USA; GTX34489). Capillary density was defined as the number of CD31-positive capillaries per mm² of muscle fiber area. Five randomly selected fields were analyzed using an Olympus CX21LED microscope and imaged with an Olympus BX51 microscope equipped with a digital camera.

Immunoblotting of HIF-1α, AMPK, LC3, and p62

Western blot analysis was performed to evaluate hypoxia signaling (HIF-1α), energy-sensing pathways (AMPK), and autophagy markers (LC3 and p62/SQSTM1).^10,11,14 Gastrocnemius muscle was homogenized in lysis buffer (10 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% NP-40, plus protease inhibitors). Lysates were centrifuged, and supernatants were heat-denatured at 96 °C for 5 minutes. Equal amounts of protein (10 µg/lane) were separated by SDS–PAGE and transferred to nitrocellulose membranes (GE Healthcare). Membranes were blocked overnight at 4 °C in 2% blocking solution (GE Healthcare) in TBS-T.

Primary antibodies (Cell Signaling Technology) were used at 1:1000 dilution: anti-HIF-1α (#3716), anti-AMPKα (#2603), anti-LC3 (#12741), anti-p62/SQSTM1 (#8025), and anti-β-actin (#4967). After HRP-conjugated secondary antibody incubation, signals were visualized with enhanced chemiluminescence (GE Healthcare) and imaged using a LI-COR C-DiGit scanner. When required, membranes were stripped (Thermo Scientific) and reprobed for β-actin as a loading control. Densitometry was performed in ImageJ, and protein expression was normalized to β-actin.

Statistical Analysis

Analyses were performed using JASP v0.19 (University of Amsterdam) and Microsoft Excel. Distribution was assessed with the Shapiro–Wilk test (normality for paired data assessed on within-animal differences). Paired, two-tailed t-tests were used for left–right comparisons in gait and IRT (background-corrected). Independent two-sample t-tests were used for immunohistochemistry and Western blot, with variance homogeneity confirmed by Levene’s test. If variance was unequal, Welch’s t-test was applied. Data are expressed as mean ± SD with 95% confidence intervals. Significance was defined as p<0.05 (two-tailed). Data distribution was assessed using the Shapiro–Wilk test. Results are presented as mean ± standard deviation with corresponding effect sizes to enhance interpretability. For molecular analyses (immunohistochemistry and Western blot), tissue samples were randomly selected from each group and were not paired with contralateral limbs on an individual-animal basis. Therefore, these datasets were analyzed as independent samples using unpaired statistical tests.

Ethical Approval

This study complied with the principles of the 3Rs (Replacement, Reduction, Refinement) and the 5Fs (freedoms from hunger and thirst, discomfort, pain, injury, disease, fear/distress, and freedom to express natural behavior). Ethical approval was obtained from the Universitas Padjadjaran Research Ethics Committee (No. 768/UN6.KEP/EC/2022) and the study was registered with AnimalStudyRegistry.org (DOI: 10.17590/asr.0000353).

Results

All animals remained healthy without illness or peri-procedural complications throughout the study period. Mean body weight was 313.7 ± 47 g at baseline (day 0), 310.2 ± 42 g on day 7, and 301.1 ± 23 g on day 14, with no significant change over time (p = 0.674). No physical or functional deficits were observed apart from reduced temperature in the right hind limb. Functional assessments yielded Tarlov, ischemia, and modified ischemic scores of 6, 5, and 7, respectively, on day 7 and day 14.^17,18

Gait Analysis

Measurements were performed on days 0, 7, and 14 (Figure 2C). The parameters assessed included the mean stride length (SL; right versus left) and stride width (SW; right versus left), analyzed separately. All data were normally distributed.

Figure 2 Gait analysis and background-correction. Bar graphs show mean stride length (A) and stride width (B) differences between the right (ischemic) and left limbs at Day 0, Day 7, and Day 14 across Deep Venous Arterialization (DVA), Open Ligation (OL), and Ligation groups (L groups). The right limb displayed shorter stride length at Days 7 and 14 in the HLI model, whereas no significant asymmetry was observed in DVA or OL groups (C). Baseline symmetry (D and E) was confirmed (p=0.119); significant asymmetry developed at Day 7 (p=0.003) and persisted in L (p=0.004). Bars represent mean ± SD; asterisks indicate p<0.05. For background correction, the contralateral limb of each group is used as the baseline.

Paired t-test analysis (Figure 2A) showed no significant difference in stride length between the left and right limbs on day 0. However, on day 7 a significant reduction was observed (left: 12.9±1.3cm vs. right: 12.2±1.6cm; t=3.252, p=0.003). Similarly, on day 14, the HLI model (control) continued to demonstrate a shorter stride length in the right limb (left: 12.3±1.1cm vs. right: 11.0±1.8cm; t=2.729, p=0.034). In contrast, no significant differences were detected in the DVA model (left: 12.1±1.4cm vs. right: 12.0±1.5cm; t=0.257, p=0.802), nor in the OL model (left: 11.9±1.7cm vs. right: 11.0±1.7cm; t=1.365, p=0.214).

For stride width (Figure 2B), paired t-test results indicated no significant differences between the left and right limbs across all models at days 0, 7, and 14 (p≥0.05).

Background Correction

Baseline gait symmetry (Figure 2D) was confirmed at Day 0 (p=0.119). A significant asymmetry emerged at Day 7 (t=−3.201, p=0.003), indicating ischemic gait disturbance. By Day 14, the DVA group achieved full restoration of symmetry (p=0.805), while the OL group showed mild, non-significant improvement (p=0.227). In contrast, the L group remained significantly asymmetric (t=4.452, p=0.004), reflecting persistent ischemic dysfunction. Paired-sample analysis of background-corrected stride width (Figure 2E) revealed no significant right–left differences at any time point (all p>0.05). Mean ΔStride values were minimal (≤0.1 cm), indicating overall gait symmetry, with only the Day 14 L group showing a slight, non-significant tendency toward longer right-limb strides (p=0.130).

Infrared Thermography Analysis

Prior to surgery, temperatures of the left and right hind limbs were measured (Figure 3C). The Wilcoxon test (non-normally distributed data) showed no significant difference between limbs (left: 33.8±1.3°C vs. right: 33.6±1.0°C; z=1.727, p=0.086). However, on day 7 after the establishment of the HLI model (Figure 3A), a significant temperature difference was observed (left: 33.8±1.9°C vs. right: 28.9±2.0°C; z=4.782, p<0.001).

Figure 3 Infrared Thermographic Assessment of Hindlimb Ischemia and Temperature Recovery Following Deep Venous Arterialization. (A) Infrared thermography of hindlimb temperature. In the Hind Limb Ishcemia (HLI) model, the right limb showed a significant reduction in temperature compared with the left at days 7 and 14, while no asymmetry was observed in the Deep Venous Arterialization (DVA) or Open Ligation (OL) models. Figure 1C and D are the Representative images of infrared thermography. (B) Background correction Box plot showing mean right–left limb temperature differences at Day 0, Day 7, and Day 14. Ische-mic hypothermia appeared at Day 7, recovered in Deep Venous Arterialization (DVA), partially improved in Open Ligation (OL), and persisted in Ligation (L) Hind Limb Ischemia (HLI). The contralateral limb of each group is used as the baseline. (C) The HLI model creation. The right hind limb temperature is significantly reduced. (D) The temperature on day 14. Deep Venous Arterialization (DVA) and Open Ligation (OL) resumed to its normal temperature, while the Hind Limb Ischemia (HLI) remain cold. Note: *p < 0.05 vs contralateral limb at the same time point.

On day 14 (Figure 3A and D), the DVA model did not demonstrate any significant temperature differences between the left and right limbs (left: 32.7±2.4°C vs. right: 32.4±2.1°C; z=0.157, p=0.910), nor did the OL model (left: 32.9±2.2°C vs. right: 32.1±2.6°C; t=1.447, p=0.191). In contrast, the HLI model exhibited a significant difference (left: 34.9±0.5°C vs. right: 30.4±1.8°C; t=7.623, p<0.001).

Background Correction

Paired-sample and Wilcoxon signed-rank analysis revealed no significant thermal asymmetry at baseline (contralateral limb - Day 0, p=0.227 (Figure 3B)). Following ischemia induction, a marked right–left temperature difference was observed at Day 7 (t=11.599, p<0.001), indicating significant hypothermia of the ischemic limb. By Day 14, thermal symmetry was restored in the DVA group (p=0.681), while the OL group showed mild, non-significant improvement (p=0.191). In contrast, the L/HLI group maintained a pronounced temperature deficit (t=7.623, p<0.001), consistent with persistent ischemic hypoperfusion.

Immunohistochemical Analysis for Capillary Density

Muscle samples were obtained on day 14 (Figure 4A–D). All data were normally distributed, except for the sham group. The HLI (L) model demonstrated a significant increase in capillary density compared with sham (U=10.000, p=0.031). In contrast, the DVA group showed a significant reduction compared with HLI (t=3.184, p=0.004), as did the OL group (t=2.951, p=0.007) (Figure 4A).

Figure 4 Capillary Density Analysis Capillary density was significantly higher in the Hind Limb Ischemia (HLI) (L) group compared with sham (p=0.031). Both Deep Venous Arteliazation (DVA) (p=0.004) and Open Ligation (OL) (p=0.007) groups showed significant reductions compared with HLI. Capillary density immunostaining (CD31) in a skeletal muscle on day 14. Representative images from Hind Limb Ischemia (HLI) (B), Deep Venous Arterialization (DVA) (C), and Sham (D) are shown. Note: (A) *p < 0.05 (statistically significant difference between the indicated groups). Arrows in (B–D) Arrows indicate CD31-positive capillary structures used for density quantification.

Western Blot Analysis

Western blot analysis was performed to assess the relative expression ratios of HIF-1α, AMPK, LC3-II, and p62 proteins in skeletal muscle samples on day 14.

HIF-1α (Figure 5A). The HLI group exhibited the highest HIF-1α protein expression compared with all other groups, with a significant increase relative to the sham group (0.730±0.3 vs. 0.371±0.1; U=2.000, p=0.016). Neither the DVA group (0.488±0.2; U=8.000, p=0.210) nor the OL group (0.517±0.2; U=9.000, p=0.274) differed significantly from sham. Likewise, the decreases observed in the DVA and OL groups relative to HLI did not reach statistical significance (U=20.000, p=0.075 and U=19.000, p=0.111, respectively).

Figure 5 Deep Venous Arterialization Modulates Hypoxia and Autophagy-Related Protein Expression Following Hindlimb Ischemia. (A) Western Blot HIF-1α: Hind Limb Ischemia (HLI) (0.730±0.3) was significantly increased compared with sham (0.371±0.1; p=0.016). Deep Venous Arterialization (DVA) (0.488±0.2) and Open Ligation (OL) (0.517±0.2) did not differ from sham (p>0.05). (B) Western Blot. AMPK: Hind Limb Ischemia (HLI) (0.855±0.3) was significantly higher than sham (0.457±0.2; p=0.011). Deep Venous Arterialization (DVA) (0.549±0.3) and and Open Ligation (OL) (0.452±0.3) were not different from sham (p>0.05), but both were significantly lower than HLI (p<0.05). (C) Western Blot. p62: No significant differences were observed among groups (p>0.05), with lowest expression in Hind Limb Ischemia (HLI) (0.556±0.1) versus Deep Venous Arterialization (DVA) (0.619±0.2), sham (0.740±0.2), and Open Ligation (OL) (0.745±0.3). (D) Western Blot. LC3-II: Deep Venous Arterialization (DVA) (0.301±0.1) showed the lowest expression, with significant reductions versus sham (0.673±0.2; p=0.032) and Hind Limb Ischemia (HLI) (0.801±0.5; p=0.002). OL (0.570±0.3) showed nonsignificant decreases compared with sham and HLI (p>0.05).

AMPK (Figure 5B). The HLI group also showed the highest AMPK protein expression, which was significantly increased compared with sham (0.855±0.3 vs. 0.457±0.2; t=2.728, p=0.011). In contrast, DVA (0.549±0.3; t=−0.691, p=0.505) and OL (0.452±0.3; t=0.036, p= 0.972) did not differ significantly from sham. However, both groups demonstrated significantly lower AMPK expression compared with HLI (DVA: t=1.884, p=0.044; OL: t=2.429, p=0.018).

p62 (Figure 5C). Protein expression of p62 did not differ significantly among groups (p>0.05). The HLI group demonstrated the lowest expression (0.556±0.1) compared with DVA (0.619±0.2) and sham (0.740±0.2), while the OL group (0.745±0.3) showed a slight, nonsignificant increase relative to sham.

LC3-II (Figure 5D). The DVA group demonstrated the lowest LC3-II expression (0.301±0.1) compared with sham (0.673±0.2) and HLI (0.801±0.5), with statistically significant reductions (t =2.116, p=0.032 and t=3.819, p=0.002, respectively). In contrast, the OL group (0.570±0.3) showed nonsignificant decreases compared with sham (t_Welch=0.650, p=0.267) and HLI (t=0.924, p=0.189).

Discussion

This study evaluated functional, vascular, and molecular adaptations of ischemic skeletal muscle following distal venous arterialization (DVA) in a rat hindlimb ischemia model, with particular focus on autophagy-related signaling.

Gait Function

Following the induction of HLI, none of the extremities exhibited overt physical damage such as necrosis, gangrene, or autoamputation, despite ligation of the right common femoral artery. This confirms that the model successfully produced ischemia without severe limb destruction, thereby serving as an appropriate experimental HLI model.^17,22,24,25

Generally, hindlimb muscle injury (non-ischemic) reduces stride length on day 3, which begins to recover by day 7.²⁶ Consistent with other studies, HLI induction in Wistar rats via femoral artery ligation and excision leads to altered gait, including prolonged swing phase and shortened stride length (SL).^27–29 Interestingly, human studies on peripheral artery disease (PAD) gait parameters remain inconsistent. Human gait studies in PAD show heterogeneous findings, likely due to confounding factors such as fitness and disease severity.^30,31

In the present study, the right hindlimb displayed a significantly shorter stride length than the left on day 7 after HLI induction (p=0.029). By day 14, after DVA or OL procedures, this difference was no longer significant (p=0.422 and p=0.140, respectively). Conversely, the HLI group continued to show a significant reduction in stride length (p=0.041). These findings suggest that HLI induces gait impairment manifested as shortened stride length, which improves following revascularization (DVA and OL) but persists in the ligation-only model. In contrast, stride width did not differ significantly in any group, either at day 7 or day 14. To our knowledge, no prior studies have specifically investigated the impact of DVA on gait in rodents (PubMed, Google Scholar).

Infrared Thermography Analysis

Infrared thermography (IRT) provides a non-invasive estimate of surface temperature and tissue perfusion and has been used to monitor ischemia and reperfusion in both clinical and experimental settings.^32–36

IRT has been increasingly applied in lower-limb ischemia for both diagnostic purposes and monitoring revascularization outcomes in acute limb ischemia (ALI) and chronic limb-threatening ischemia (CLTI).^32–35 In rodent models, IRT effectively detects ischemia: femoral artery clamping induces rapid limb cooling, followed by temperature recovery upon reperfusion.³⁷ Other mouse studies confirm sharp temperature drops after ligation with gradual, non-significant recovery up to day 28.^35,37–39 Smartphone-based IRT studies similarly show persistent ischemia at day 7 in controls, with improved perfusion in treated groups.^40,41

Our study demonstrated a significant temperature decline in created HLI model at day 7 (28.9±2.0°C) compared with pre-ligation (33.6±1.0°C) and controls (33.8±1.9°C). By day 14, HLI group remained significantly cooler (30.4±1.8°C vs. 34.9±0.5°C, p<0.05). In contrast, DVA and OL groups showed temperature restoration approaching contralateral limbs (DVA: 32.4±2.1°C vs. 32.7±2.4°C; OL: 32.1±2.6°C vs. 32.9±2.2°C). These results corroborate previous reports on the diagnostic and follow-up utility of IRT in ischemia and reperfusion models.³⁷ Notably, no prior literature on IRT in DVA models was identified.

Capillary Density

Capillary density, assessed by CD31 immunohistochemistry, reflects angiogenesis. Prior rodent studies reported significant increases in capillary density following femoral artery ligation.^{18,21,42–44} In one unique experiment, capillary density increased in HLI alone but decreased in HLI with denervation or denervation only, suggesting a modulatory role of peripheral nerves in arteriogenesis.⁴⁵ Contrastingly, human studies in lower extremity arterial disease (LEAD/PAD) revealed reduced capillary density in gastrocnemius biopsies,^46–48 although some earlier reports found elevated density, suggesting compensatory mechanisms.^47,49,50

Reperfusion injury introduces additional complexity, as prolonged ischemia and inflammatory mediators can impair tissue and vascular responses.^51–54 Rodent tourniquet studies showed non-significant changes in capillary density after reperfusion.⁵⁵ More recent studies with collateral-dependent reperfusion also reported no density changes up to 90 days.^56,57

In the current study, capillary density increased significantly in HLI compared with sham (p=0.031). However, both DVA and OL groups exhibited significant decreases compared with HLI, though values were not significantly different from sham. This suggests that revascularization normalizes capillary density levels, consistent with previous findings.⁵⁸ The discrepancy between rodent and human data likely reflects differences between acute experimental ischemia and chronic human disease. Importantly, no comparable data for DVA models were identified.

HIF-1α Expression

HIF-1α is a master regulator of hypoxia-mediated cellular adaptation.^10,11,59 Its expression increases under ischemia and reperfusion injury, often triggering autophagy.¹⁰ Prior studies consistently demonstrate HIF-1α upregulation under ischemic conditions.^21,60 In a sildenafil study, HIF-1α increased in HLI controls but remained normal with treatment, despite unchanged capillary density.⁶¹ Mouse models similarly show early increases in HIF-1α (day 3), persisting but reduced by day 14, and correlating with increased capillary density.^62,63

Our study found significantly elevated HIF-1α expression in HLI compared with sham at day 14 (p=0.016), consistent with previous reports.^61–63 DVA and OL groups, however, showed no significant increases compared with sham (p>0.05), indicating that revascularization may inhibit HIF-1α activation. This aligns with our capillary density findings, supporting the role of hypoxia-driven HIF-1α signaling in angiogenesis.^21,62,63

AMPK Expression

AMP-activated protein kinase (AMPK) is a central regulator of cellular metabolism, activated under low intracellular ATP (eg., hypoxia). Increased AMP/ATP ratio triggers AMPK activation independently of HIF-1α.¹⁴ As a “metabolic checkpoint,” AMPK suppresses mTORC1 activity, thereby promoting autophagy.^64,65 Rodent studies show HLI-induced AMPK activation coincides with increased capillary density, highlighting its role in ischemia-induced angiogenesis.^66,67

In this study, AMPK expression increased significantly in HLI compared with sham, while DVA and OL groups demonstrated significant reductions toward sham values. These findings suggest an association between ischemia, AMPK activation, and metabolic stress, with normalization following revascularization.⁶⁷

p62 and LC3-II Expression

Autophagy can be evaluated via p62 (sequestosome-1) and LC3 (microtubule-associated protein 1A/1B light chain 3), both widely recognized autophagosome markers.^8,68–73 LC3-II localizes to autophagosome membranes, while p62 is degraded during autophagic flux.⁷⁴

Rodent models show dynamic expression: LC3-II increases on day 7, while p62 rises on day 3 and declines by day 7.⁷⁵ Other studies report stable LC3-II/LC3-I ratios in chronic HLI unless exercise is introduced,⁷⁴ or reduced LC3-II after reperfusion with concomitant p62 decline.⁷⁶ Additional evidence supports decreased p62 and increased LC3 in ischemia, alongside elevated HIF-1α and CD31.⁷⁰ By day 28, LC3-II levels typically normalize.⁶⁵ Human PAD studies confirm higher LC3 and capillary density, with reduced plasma p62 indicating enhanced autophagy.^54,77,78

In our study, LC3-II expression increased in HLI but not significantly compared with sham (p=0,287). Interestingly, DVA significantly reduced LC3-II compared with both sham and HLI (p=0.032; p=0,002), indicating suppressed autophagy. p62 expression decreased in HLI versus sham, though not significantly (p=0.083), and increased in DVA and OL compared with HLI, again without significance. While not statistically robust, these findings suggest potential modulation of autophagy via p62 signaling in chronic ischemia. Collectively, the observed AMPK elevation in HLI and reduction in DVA further support these mechanistic links. Therefore, the observed changes in LC3-II and p62 should be interpreted as alterations in autophagy-related markers rather than definitive evidence of altered autophagic flux.

Overall Interpretation

Overall, this study demonstrates that prolonged hindlimb ischemia leads to reduced oxygen delivery, metabolic shift to anaerobic respiration, and increased AMP/ATP ratio, manifested as limb cooling, elevated capillary density, impaired gait, and upregulation of HIF-1α and AMPK, with a corresponding increase in autophagy marker (LC3-II). In contrast, distal venous arterialization restored perfusion, normalized temperature and capillary density, and improved gait, accompanied by reductions in HIF-1α, AMPK, and LC3-II toward sham levels. These changes suggest that DVA was associated with normalization of hypoxia-related signaling and reductions in autophagy-related markers. Together, the functional, thermal, vascular, and molecular outcomes highlight both the pathological cascade induced by ischemia and the physiological recovery achieved through DVA, supporting its role as a potential revascularization strategy for ischemic skeletal muscle. From a translational perspective, these findings provide mechanistic support for the clinical use of distal venous arterialization in chronic limb-threatening ischemia with no other option. Demonstrating improvements in functional performance, tissue perfusion, and molecular markers of hypoxia and autophagy helps bridge the gap between observed clinical limb salvage outcomes and their underlying biological mechanisms. These data therefore strengthen the biological plausibility of DVA as a revascularization strategy for ischemic skeletal muscle.

Study Limitations and Future Directions

This study has several important limitations. The HLI model, based on femoral artery ligation without collateral excision, may have permitted residual collateralization and thus does not fully replicate the chronic progression of PAD/CLTI in humans, for which no ideal animal model currently exists.^{17,21,22,24,25,79–81} The use of relatively young, otherwise healthy rats without comorbidities such as diabetes or vascular disease further limits translational relevance, as age and comorbidity strongly influence ischemic recovery.^18,22,24,25

In addition, the marked difference in operative time between DVA and OL could have introduced bias through prolonged anesthesia and surgical stress, while analgesia and stress biomarkers were insufficiently evaluated. Methodological constraints such as tissue sampling limited to day 14, variability in gait analysis, and the lower accuracy of infrared thermography compared with Laser Doppler Perfusion Imaging also reduce the robustness of the findings. Infrared thermography provides an indirect estimate of perfusion and may be influenced by ambient temperature or handling despite standardization. The small sample size and scarcity of comparative literature on autophagy in PAD and on DVA in animal models further restrict generalizability. Because autophagic flux and AMPK activation status were not directly measured, these findings should be interpreted as associations rather than definitive causal mechanisms.

Future studies should aim to employ older animals or models with relevant comorbidities, extend observation periods beyond 14 days to better approximate chronic ischemia, and incorporate more precise tools such as treadmill-based gait systems and Laser Doppler Perfusion Imaging. Standardizing operative times, expanding sample sizes, and integrating stress biomarkers could also strengthen validity. Ultimately, the development of more clinically relevant animal models will be essential to bridge the translational gap between experimental research and human PAD/CLTI. The relatively small and unequal group sizes, particularly the larger DVA group due to technical considerations of microsurgery, may have limited statistical power and could affect the generalizability of the findings. In addition, molecular analyses were performed at a single endpoint (day 14), which may not fully capture the temporal dynamics of autophagy regulation.

Conclusion

In this study, distal venous arterialization (DVA) in a rat hindlimb ischemia model restored gait function and normalized limb temperature without inducing severe complications such as gangrene. DVA was associated with improvement of ischemia-related skeletal muscle changes, including normalization of capillary density and reductions in hypoxia-related markers (HIF-1α, AMPK). Furthermore, DVA was accompanied by lower expression of the autophagy-related marker LC3-II, while p62 remained unchanged. These findings suggest that DVA promotes functional and suggesting molecular recovery in ischemic skeletal muscle and represents a promising revascularization strategy.