Abnormal Microvascular Architecture, Fibrosis, and Pericyte Characteristics in the Calf Muscle of Peripheral Artery Disease Patients with Claudication and Critical Limb Ischemia

Work from our laboratory documents pathological events, including myofiber oxidative damage and degeneration, myofibrosis, micro-vessel (diameter = 50–150 μm) remodeling, and collagenous investment of terminal micro-vessels (diameter ≤ 15 µm) in the calf muscle of patients with Peripheral Artery Disease (PAD). In this study, we evaluate the hypothesis that the vascular pathology associated with the legs of PAD patients encompasses pathologic changes to the smallest micro-vessels in calf muscle. Biopsies were collected from the calf muscle of control subjects and patients with Fontaine Stage II and Stage IV PAD. Slide specimens were evaluated by Quantitative Multi-Spectral and Fluorescence Microscopy. Inter-myofiber collagen, stained with Masson Trichrome (MT), was increased in Stage II patients, and more substantially in Stage IV patients in association with collagenous thickening of terminal micro-vessel walls. Evaluation of the Basement Membrane (BM) of these vessels reveals increased thickness in Stage II patients, and increased thickness, diameter, and Collagen I deposition in Stage IV patients. Coverage of these micro-vessels with pericytes, key contributors to fibrosis and BM remodeling, was increased in Stage II patients, and was greatest in Stage IV patients. Vascular pathology of the legs of PAD patients extends beyond atherosclerotic main inflow arteries and affects the entire vascular tree—including the smallest micro-vessels.


Introduction
Peripheral artery disease (PAD) is a severe atherosclerotic condition, primarily of the elderly, which afflicts more than 200 million individuals worldwide [1]. PAD patients carry an increased risk of cardiovascular morbidity and mortality [2]. Patients with PAD have decreased calf muscle strength and exhibit a wide range of lower leg symptoms and signs [3,4]. Approximately one-third of PAD patients

PAD Groups
We recruited 15 patients with claudication (Fontaine Stage II) and 16 patients presenting with tissue loss (Fontaine Stage IV) who were scheduled for lower extremity operations for symptomatic PAD. Medical history, physical examination, decreased ABI (<0.9), and computerized or standard arteriography that revealed stenotic and/or occluded arteries supplying the lower extremities were evaluated to establish the diagnosis for each PAD patient. Stage II PAD patients presented with intermittent claudication, but no rest pain or tissue loss. Stage IV PAD patients presented with non-healing ulcers and/or gangrene.

Biopsy
Calf muscle samples weighing approximately 250 mg were obtained from the anteromedial aspect of the muscle belly, 10 cm distal to the tibial tuberosity. All biopsies were obtained with a 6 mm Bergstrom needle. Samples were placed immediately into cold Methacarn fixative, transferred at 48 h to cold 50% ethanol, and 5 to 10 days later embedded in paraffin.
Paraffin-embedded biopsies were sectioned at 4 µm and mounted to glass slides. Slide specimens were deparaffinized with xylene and rehydrated via a series of ethanol washes and distilled water. The rehydrated slide specimens were heated at 92 • C in either 10 mM citrate buffer (pH 6.0) or 10 mM tris with EDTA buffer (pH 9.0) for 15 min and then allowed to cool at room temperature for 20 min. Subsequently, slide specimens were washed in Super-Sensitive Wash buffer (BioGenex Laboratories, Freemont, CA, USA) for 30 min at room temperature, followed by a 10 min wash in Dulbecco's Phosphate Buffered Saline. Then, the specimens were transferred to a programmable robotic autostainer (BioGenex i6000, BioGenex Laboratories) where they were washed with Super Sensitive Wash Buffer and blocked with normal sera. Primary antibodies and corresponding immunoglobulin isotypes were pipetted manually onto the appropriate slides and held overnight at 4 • C. On the following day, slide specimens were processed at room temperature with the programmable autostainer. They were washed with Super-Sensitive Wash buffer, labeled with secondary Ab for one hour and washed with Dulbecco's Phosphate Buffered Saline. The specimens were mounted in Prolong Diamond Mounting Media that contained nuclear label (DAPI: 4 ,6-diamidino-2-phenylindole; Invitrogen, Carlsbad, CA, USA). For each patient specimen, duplicate slides were treated with primary Ab, and a third slide was treated with corresponding immunoglobulin isotype.
HCImage was programmed for automated acquisition of the fluorescence signals from the slide-mounted muscle specimens. The entirety of each slide specimen was captured in the 647 nm channel (Collagen I or Collagen IV) with a 10× objective, producing 70-150 microscope frames (10× fields). The acquired fields were montaged into one image representing the entire slide specimen. Twelve regions of interest (ROIs) were identified from a review of the montage, and the ROIs were located on the original slide specimen. For micro-vessel collagen measurements, four contiguous frames (2 × 2) in each ROI were captured with a 40× objective, in three fluorescence channels corresponding to nuclei (DAPI: 358 nm excitation max and 461 nm emission max), cell membrane and extracellular matrix (WGA-Alexa Fluor 488: 488 nm excitation max and 520 nm emission max) and Collagen I (Alexa Fluor 647: 650 nm excitation max and 668 nm emission max) or Collagen IV (Alexa Fluor 647: 650 nm excitation max and 668 nm emission max). Fluorescence intensities produced by Ab treatment were corrected for fluorescence intensities produced by the corresponding isotype controls.
The labeled collagen ring associated with each micro-vessel was partitioned in the Image Pro ® Plus software environment and both ring area (µm 2 ) and mean fluorescence intensity (greyscale units; gsu) were captured. We placed a circular area of interest (AOI) around each collagen ring to restrict partitioning to the collagen ring ( Figure 1D). Collagen abundance for each micro-vessel was calculated as the product of ring area and mean fluorescence intensity (gsu·µm 2 ) corrected for isotype background fluorescence. Mean collagen abundance per micro-vessel was determined for the 12 ROIs of each slide specimen, and the abundance per patient was determined as the average of two slides.

Measurement of Micro-Vessel Basement Membrane Thickness, Inner Diameter, and Overall Diameter
Greyscale images from 12 ROIs of each slide specimen labeled with anti-Collagen IV Ab and captured with a 40× objective as a montage of 2 × 2 contiguous microscope fields per ROI were imported into the Image Pro ® Plus software (Media Cybernetics, Warrendale, PA, USA) environment and corrected for isotype fluorescence intensities. Subsequently, the corrected montages were imported into the AutoQuant ® software (Media Cybernetics) environment for deconvolution that improved image resolution and contrast. Measurements of micro-vessel architecture in the deconvoluted images were implemented with a custom Matlab program (MathWorks, Inc., Natick, MA, USA). Briefly, two perpendicular lines were drawn such that they transected a micro-vessel. For vessels that were not perfectly circular, one of the perpendicular lines transected the vessel at its longest diameter. The program extracted and plotted Collagen IV fluorescence intensities along each line and fitted the raw fluorescence data to Gaussian distributions ( Figure 2). The thickness of the Collagen IV ring (basement membrane) for each micro-vessel was determined as the average distance at the 95% confidence interval of each Gaussian curve. The inner diameter of the basement membrane was determined as the mean of the distances between the pairs of Gaussian curves at the locations of their 95% confidence intervals. Total micro-vessel diameter was calculated as the mean inner basement membrane diameter plus twice the mean basement membrane thickness for each micro-vessel ( Figure 2). To ensure uniform cross-sectional measurements, we measured only micro-vessels whose total micro-vessel diameter at each of the perpendicular lines exhibited a ratio of 0.75 to 1.00 (corresponding to a perfect circle). Approximately 50-250 micro-vessels were assessed per slide specimen. The number of micro-vessels evaluated per ROI varied, reflecting the variation of the micro-vessel density per specimen. Greyscale images from 12 ROIs of each slide specimen labeled with anti-Collagen IV Ab and captured with a 40× objective as a montage of 2 × 2 contiguous microscope fields per ROI were imported into the Image Pro ® Plus software (Media Cybernetics, Warrendale, PA, USA) environment and corrected for isotype fluorescence intensities. Subsequently, the corrected montages were imported into the AutoQuant ® software (Media Cybernetics) environment for deconvolution that improved image resolution and contrast. Measurements of micro-vessel architecture in the deconvoluted images were implemented with a custom Matlab program (MathWorks, Inc., Natick, MA, USA). Briefly, two perpendicular lines were drawn such that they transected a micro-vessel. For vessels that were not perfectly circular, one of the perpendicular lines transected the vessel at its longest diameter. The program extracted and plotted Collagen IV fluorescence intensities along each line and fitted the raw fluorescence data to Gaussian distributions ( Figure 2). The thickness of the Collagen IV ring (basement membrane) for each micro-vessel was determined as the average distance at the 95% confidence interval of each Gaussian curve. The inner diameter of the basement membrane was determined as the mean of the distances between the pairs of Gaussian curves at the locations of their 95% confidence intervals. Total micro-vessel diameter was calculated as the mean inner basement membrane diameter plus twice the mean basement membrane thickness for each microvessel ( Figure 2). To ensure uniform cross-sectional measurements, we measured only micro-vessels whose total micro-vessel diameter at each of the perpendicular lines exhibited a ratio of 0.75 to 1.00 (corresponding to a perfect circle). Approximately 50-250 micro-vessels were assessed per slide specimen. The number of micro-vessels evaluated per ROI varied, reflecting the variation of the micro-vessel density per specimen. Measurements of Micro-vessel Architecture. Slide specimens labeled with anti-Collagen IV Ab were imaged with a 40× objective and deconvoluted with AutoQuant ® software for reduced light scatter and improved micro-vessel delineation. For each micro-vessel (overall diameter ≤15μm), two perpendicular line profiles of Collagen IV fluorescence intensity were generated with a custom MatLab program. Gaussian distributions (red) were fitted to the raw fluorescence intensity data (blue). The thickness of the Collagen IV ring (basement membrane) for each micro-vessel was determined as the average distance at the 95% confidence interval for each of the Gaussian curves. The inner diameter of the basement membrane was determined as the mean of the distances between the pairs of Gaussian curves at the locations of their 95% confidence intervals.

Identification of Micro-Vessel Associated Pericytes
We identified pericytes as α-smooth muscle actin (αSMA) positive cells (1) located abluminal to CD31 positive endothelial cells lining micro-vessels (overall diameter ≤ 15 microns) and (2) intimately associated with the Collagen IV positive basement membrane of these micro-vessels.
Slide specimens were blocked with 10% donkey serum (cat# ab7475, Abcam, Cambridge, MA, USA) and then treated with a mixture of mouse monoclonal anti-collagen IV Ab (10 μg/mL, clone Measurements of Micro-vessel Architecture. Slide specimens labeled with anti-Collagen IV Ab were imaged with a 40× objective and deconvoluted with AutoQuant ® software for reduced light scatter and improved micro-vessel delineation. For each micro-vessel (overall diameter ≤15µm), two perpendicular line profiles of Collagen IV fluorescence intensity were generated with a custom MatLab program. Gaussian distributions (red) were fitted to the raw fluorescence intensity data (blue). The thickness of the Collagen IV ring (basement membrane) for each micro-vessel was determined as the average distance at the 95% confidence interval for each of the Gaussian curves. The inner diameter of the basement membrane was determined as the mean of the distances between the pairs of Gaussian curves at the locations of their 95% confidence intervals.

Identification of Micro-Vessel Associated Pericytes
We identified pericytes as α-smooth muscle actin (αSMA) positive cells (1) located abluminal to CD31 positive endothelial cells lining micro-vessels (overall diameter ≤ 15 microns) and (2) intimately associated with the Collagen IV positive basement membrane of these micro-vessels.
Five microscope fields per duplicate slide specimen were captured with a 40× objective, in four fluorescence channels corresponding to (1) DAPI (nuclei), (2) Alexa Fluor 488 (Collagen IV) with an excitation maximum at 495 nm and emission maximum at 519 nm, (3) Alexa Fluor 555 (αSMA), and (4) Alexa Fluor 647 (CD31). Fluorescence intensities produced by Ab treatment were corrected for fluorescence intensities produced by the corresponding isotype controls.

Micro-Vessel Density
Greyscale images obtained from Collagen I and IV labeling were transferred into the Image Pro ® Plus environment. Collagen IV label uniformly identified micro-vessel structure and was used to create partitions that yielded accurate vessel identification. The number of micro-vessels and total tissue area were determined. A ratio of micro-vessels per total tissue area was calculated to determine micro-vessel density per each ROI assessed for micro-vessel architecture.

Statistical Analysis
Demographic variables were evaluated for significant differences across study groups (Fontaine Stage II and Stage IV PAD patients and controls) by the non-parametric Kruskal-Wallis Test, for continuous variables, or by the Yates' Continuity Corrected Chi-Square Test, for categorical variables. All other between-group differences were evaluated by the two-tailed Student's t-test for normally distributed data and by the non-parametric Mann-Whitney U test or randomization test for data that were not distributed normally. Selected clinical parameters were analyzed as covariates with a General Linear Model for fixed factors. All statistical analyses were completed with NCSS 2020 Statistical Software (NCSS, LLC. Kaysville, UT, USA).

Patient Demographics
The number of subjects in each cohort (Control N = 14; Stage II PAD N = 15; Stage IV PAD N = 16) was limited by the labor-intensive measurements of micro-vessels in slide specimens of calf muscle. Previous studies of micro-vessel structure in lower leg muscles of PAD patients and control subjects documented significant differences with 8 to 14 subjects in each cohort; consequently, the number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses. The number of subjects in each cohort (Control N = 14; Stage II PAD N = 15; Stage IV PAD N = 16) was limited by the labor-intensive measurements of micro-vessels in slide specimens of calf muscle. Previous studies of micro-vessel structure in lower leg muscles of PAD patients and control subjects documented significant differences with 8 to 14 subjects in each cohort; consequently, the number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses.

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of intermyofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented were not distributed normally. Selected clinical parameters were analyzed as covariates with a General Linear Model for fixed factors. All statistical analyses were completed with NCSS 2020 Statistical Software (NCSS, LLC. Kaysville, UT, USA).

Patient Demographics
The number of subjects in each cohort (Control N = 14; Stage II PAD N = 15; Stage IV PAD N = 16) was limited by the labor-intensive measurements of micro-vessels in slide specimens of calf muscle. Previous studies of micro-vessel structure in lower leg muscles of PAD patients and control subjects documented significant differences with 8 to 14 subjects in each cohort; consequently, the number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses.

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of intermyofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented All other between-group differences were evaluated by the two-tailed Student's t-test for normally distributed data and by the non-parametric Mann-Whitney U test or randomization test for data that were not distributed normally. Selected clinical parameters were analyzed as covariates with a General Linear Model for fixed factors. All statistical analyses were completed with NCSS 2020 Statistical Software (NCSS, LLC. Kaysville, UT, USA).

Patient Demographics
The number of subjects in each cohort (Control N = 14; Stage II PAD N = 15; Stage IV PAD N = 16) was limited by the labor-intensive measurements of micro-vessels in slide specimens of calf muscle. Previous studies of micro-vessel structure in lower leg muscles of PAD patients and control subjects documented significant differences with 8 to 14 subjects in each cohort; consequently, the number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses.

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of intermyofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented . A Transient Ischemic Attack. B Body Mass Index > 30 kg/m 2 . C Creatinine Clearance < 60 mL/min/1.73 m 2 .

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of inter-myofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented myofibers embedded in a fibrotic matrix, as well as large swollen fibers, likely undergoing necrosis, with abundant inter-myofiber collagen. Inter-myofiber collagen abundances in all specimens were determined from greyscale images (illustrated in Figure 3D-F) extracted from Multi-Spectral images of Masson Trichome stained specimens. Collagen abundances are presented as mean ± s.e.m. For each study group. Stage IV collagen abundance (1114 ± 130 gsu) was nearly three to four times greater than collagen abundances of Stage II patients (409 ± 39 gsu at p < 0.001) and control subjects (306 ± 29 gsu at p < 0.001). Collagen abundance of Stage II patients was greater than that of control subjects at p = 0.041 ( Figure 4). Notably, micro-vessels were circumscribed with well-defined collagen rings ( Figure 3G-I) that were larger and exhibited increased wall thickness, in Stage IV PAD patients.
Since the prevalence of DM, RI and FH were greater among Stage IV PAD patients (Table 1), these clinical parameters were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV). For response variable Collagen Abundance, the predictive value (adjusted R 2 ) of the model was 0.545 (p < 0.001) where Group (p < 0.001) was a significant predictor and DM (p = 0.367), RI (p = 0.613), and FH (0.750) were not significant predictors.
( Figure 3G-I) that were larger and exhibited increased wall thickness, in Stage IV PAD patients.
Since the prevalence of DM, RI and FH were greater among Stage IV PAD patients (Table 1), these clinical parameters were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV). For response variable Collagen Abundance, the predictive value (adjusted R 2 ) of the model was 0.545 (p < 0.001) where Group (p < 0.001) was a significant predictor and DM (p = 0.367), RI (p = 0.613), and FH (0.750) were not significant predictors. is uniform among closely associated myofibers (RED) of relatively homogeneous polygonal shape and cross-sectional area. Collagen staining is increased in PAD Stage II (Panel (B)) (N = 15) and Stage IV (Panel (C)) (N = 15) with the latter presenting remarkable departures in myofiber geometry and collagen deposition compared to both control and Stage II muscle. Stage IV calf muscle exhibited many small degenerating and fragmented myofibers embedded in a large fibrotic matrix, as well as enlarged rounded myofibers, likely necrotic, surrounded by more abundant collagen. Greyscale images of deposited collagen (Panels D-F) were extracted from Multi-Spectral images of slide specimens stained with MT and captured with the Nuance System (20× objective). A magnified region (Panels G-I) of each greyscale image revealed well-defined collagenous rings around the micro-vessels (Arrows). Scale bars (Panels A-F) represent 200 microns.  . Microscopic images were captured with a Multi Spectral imaging system. Quantitative greyscale images of inter-myofiber collagen were extracted from the Multi-Spectral images and quantified in the Image Pro ® Premier environment. Collage abundance (gsu) was computed as the sum of the products (gsu•micron2) of mean pixel intensity and event area divided by the total area (micron2) of myofibers plus total area of inter-myofiber collagen per microscopic field. The average of five fields per slide specimen was taken as the collage abundance per patient. Data are presented as Fence Box Plots. Statistical significance was determined with the non-parametric randomization test. a-Significantly different at p = 0.041. b-Significantl different at p < 0.001 c-Significantly different at p < 0.001

icro-Vessel Basement Membrane Architecture Is Altered in Calf Muscle of Patients with PAD
Masson Trichrome labeling of collagen in calf muscle specimens of control subjects and pa PAD revealed well-defined ring structures delineating calf muscle micro-vessels of all y groups (Figure 3). The high definition and location of these ring structures suggested tha sented basement membrane which appeared to be greater in diameter and wall thickn e IV patients compared to control subjects and Stage II PAD patients. To investigate the base brane architecture of micro-vessels, we labeled BMs of calf muscle slide specimens with gen IV antibody and measured (μm) BM thickness and inner and overall BM diameter ( ll three parameters of BM architecture were increased uniformly by 23% to 24% (p < 0.0 o-vessels of calf muscle specimens from patients with Stage IV PAD compared to co cts. Compared to controls, BM thickness in micro-vessels of Stage II patients was increas (p < 0.001), inner diameter was not different and overall diameter was increased by 6% bu . Microscopic images were captured with a Multi-Spectral imaging system. Quantitative greyscale images of inter-myofiber collagen were extracted from the Multi-Spectral images and quantified in the Image Pro ® Premier environment. Collagen abundance (gsu) was computed as the sum of the products (gsu·micron2) of mean pixel intensity and event area divided by the total area (micron2) of myofibers plus total area of inter-myofiber collagen, per microscopic field. The average of five fields per slide specimen was taken as the collagen abundance per patient. Data are presented as Fence Box Plots. Statistical significance was determined with the non-parametric randomization test. a-Significantly different at p = 0.041. b-Significantly different at p < 0.001 c-Significantly different at p < 0.001

Micro-Vessel Basement Membrane Architecture Is Altered in Calf Muscle of Patients with PAD
Masson Trichrome labeling of collagen in calf muscle specimens of control subjects and patients with PAD revealed well-defined ring structures delineating calf muscle micro-vessels of all three study groups (Figure 3). The high definition and location of these ring structures suggested that they represented basement membrane which appeared to be greater in diameter and wall thickness in Stage IV patients compared to control subjects and Stage II PAD patients. To investigate the basement membrane architecture of micro-vessels, we labeled BMs of calf muscle slide specimens with anti-Collagen IV antibody and measured (µm) BM thickness and inner and overall BM diameter ( Table 2). All three parameters of BM architecture were increased uniformly by 23% to 24% (p < 0.001) in micro-vessels of calf muscle specimens from patients with Stage IV PAD compared to control subjects. Compared to controls, BM thickness in micro-vessels of Stage II patients was increased by 12% (p < 0.001), inner diameter was not different and overall diameter was increased by 6% but did not reach statistical significance. The uniform increase of 23% to 24% in all BM architectural parameters in micro-vessels of Stage IV calf muscle compared to control muscle suggested that the dimensions of these BM parameters are coordinated. This was supported by analysis of the correlations of BM overall diameter (R = 0.876; p = 0.0001) and BM inner diameter (R = 0.640; p = 0.0001) and with BM thickness across all control subjects and PAD patients ( Figure 5). number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses.

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of intermyofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented Overall Diameter 6.058 ± 0.111 6.448 ± 0.112 7.466 ± 0.224 * , muscle. Previous studies of micro-vessel structure in lower leg muscles of PAD patients and control subjects documented significant differences with 8 to 14 subjects in each cohort; consequently, the number of subjects in each of our cohorts was considered adequate for this study. By design, mean ABI of controls was significantly different from both Stage II and Stage IV PAD patients (Table 1). It is of interest that ABI and prevalence of Diabetes Mellitus (DM), Renal Insufficiency (RI), and Family history of cardiovascular disease (FH) of Stage IV PAD patients differed significantly from both Stage II PAD patients and control subjects. These variables were treated as potential covariates in subsequent analyses.

Fibrosis Is Increased in Calf Muscle of Patients with PAD
Fibrosis in calf muscle was determined as collagen abundance, from measurements of intermyofiber collagen exclusive of large vessel collagen. In control specimens (N = 14), collagen was uniformly distributed between closely associated myofibers of relatively homogeneous polygonal shape and cross-sectional area ( Figure 3A). Stage II PAD specimens (N = 1 5) were characterized by increased variation in myofiber shape, reduced cross-sectional area and the appearance of more deeply stained inter-myofiber collagen ( Figure 3B). Both myofiber geometry and inter-myofiber collagen deposition in Stage IV specimens (N = 15) ( Figure 3C) presented remarkable departures from these features of control and Stage II specimens, exhibiting regions of degenerating, fragmented l N = 14; Stage II PAD N = 15; Stage IV PAD N = ents of micro-vessels in slide specimens of calf in lower leg muscles of PAD patients and control to 14 subjects in each cohort; consequently, the sidered adequate for this study. By design, mean h Stage II and Stage IV PAD patients (Table 1). It ellitus (DM), Renal Insufficiency (RI), and Family AD patients differed significantly from both Stage riables were treated as potential covariates in demographics. h PAD llagen abundance, from measurements of interen. In control specimens (N = 14), collagen was myofibers of relatively homogeneous polygonal I PAD specimens (N = 1 5) were characterized by ross-sectional area and the appearance of more B). Both myofiber geometry and inter-myofiber Figure 3C) presented remarkable departures from exhibiting regions of degenerating, fragmented

Control
Significantly different from Stage II PAD (p < 0.001).

The Abundance of Collagen IV in Calf Muscle Micro-Vessels across Control Subjects and PAD Patients Reflects the Architectural Measurements of Micro-Vessel Basement Membranes (BM)
Slide specimens of calf muscle biopsies collected from control subjects and patients with PAD were labeled with the antibody specific for Collagen IV, a marker for BM. For both control subjects and PAD patients, Collagen IV labeling delineated individual myofibers and presented as welldefined rings of intense labeling associated with the micro-vessels ( Figure 6A-C). Apart from control specimens, micro-vessel Collagen IV abundance was somewhat heterogeneous, but generally was greater in specimens from PAD patients with Stage IV disease compared to patients with Stage II disease ( Figure 6D-F). Using Quantitative Fluorescence Microscopy, we quantified micro-vessel Collagen IV abundance in slide specimens labeled with anti-Collagen IV antibody ( Figure 6; Fence Box Plot). For each micro-vessel, abundance was determined as the product of mean fluorescence intensity (pixel) and area occupied by Collagen IV. Medians (M) and interquartile ranges (IQR) were 16,670 gsu•μm 2 , 12,778 gsu•μm 2 ; 22,378 μm 2 , 14,860 μm 2 ; and 29,165 μm 2 , 27,286 μm 2 for control, Stage II PAD and Stage IV PAD specimens, respectively. Collagen IV abundance of Stage IV patients was 75% greater than that of control subjects (p = 0.029) and 30% greater than Stage II patients (p = 0.042). Abundances in Stage II compared to control specimens was increased by 34% but were not statistically significant (p = 0.581). Micro-vessel counts per unit area of slide specimen increased with advancing stage of the disease and reflected increased Collagen IV accumulation in micro-vessels with advancing disease, suggesting a coordinated response to the intermittent ischemia (hypoxia) of PAD. Micro-vessel counts in the calf muscle of Stage IV patients (5.9 micro-vessels per 10 4 μm 2 ) were Since the prevalence of DM, RI and FH were greater among Stage IV PAD patients (Table 1), these clinical parameters were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV). For response variable BM Thickness, the predictive value (adjusted R 2 ) of the model was 0.359 (p < 0.001) where Group (p < 0.001) was a significant predictor and DM (p = 0.960), RI (p = 0.741), and FH (0.543) were not significant predictors. For response variable BM Inner Diameter, the predictive value (adjusted R 2 ) of the model was 0.407 (p < 0.001) where Group (p < 0.001) was a significant predictor and DM (p = 0.288), RI (p = 0.344), and FH (0.962) were not significant predictors. For response variable BM Overall Diameter, the predictive value (adjusted R 2 ) of the model was 0.432 (p < 0.001) where Group (p < 0.0001) was a significant predictor and DM (p = 0.542), RI (p = 0.460), and FH (0.728) were not significant predictors.

The Abundance of Collagen IV in Calf Muscle Micro-Vessels across Control Subjects and PAD Patients Reflects the Architectural Measurements of Micro-Vessel Basement Membranes (BM)
Slide specimens of calf muscle biopsies collected from control subjects and patients with PAD were labeled with the antibody specific for Collagen IV, a marker for BM. For both control subjects and PAD patients, Collagen IV labeling delineated individual myofibers and presented as well-defined rings of intense labeling associated with the micro-vessels ( Figure 6A-C). Apart from control specimens, micro-vessel Collagen IV abundance was somewhat heterogeneous, but generally was greater in specimens from PAD patients with Stage IV disease compared to patients with Stage II disease ( Figure 6D-F). Using Quantitative Fluorescence Microscopy, we quantified micro-vessel Collagen IV abundance in slide specimens labeled with anti-Collagen IV antibody ( Figure 6; Fence Box Plot). For each micro-vessel, abundance was determined as the product of mean fluorescence intensity (pixel) and area occupied by Collagen IV. Medians (M) and interquartile ranges (IQR) were 16,670 gsu·µm 2 , 12,778 gsu·µm 2 ; 22,378 µm 2 , 14,860 µm 2 ; and 29,165 µm 2 , 27,286 µm 2 for control, Stage II PAD and Stage IV PAD specimens, respectively. Collagen IV abundance of Stage IV patients was 75% greater than that of control subjects (p = 0.029) and 30% greater than Stage II patients (p = 0.042). Abundances in Stage II compared to control specimens was increased by 34% but were not statistically significant (p = 0.581). Micro-vessel counts per unit area of slide specimen increased with advancing stage of the disease and reflected increased Collagen IV accumulation in micro-vessels with advancing disease, suggesting a coordinated response to the intermittent ischemia (hypoxia) of PAD. Micro-vessel counts in the calf muscle of Stage IV patients (5.9 micro-vessels per 10 4 µm 2 ) were 23% greater (p = 0.011) than controls (4.8 micro-vessels per 10 4 µm 2 ) and counts for Stage II patients (5.6 micro-vessels per 10 4 µm 2 ) were 17% greater (p = 0.017) than controls. The difference between Stage IV and Stage II patients did not reach significance (p = 0.256). were labeled with anti-Collagen IV antibody, and fluorescence images were captured with a widefield microscope (40× objective). Selected regions in Panels (A-C) were enlarged (Panels (D-F)), providing a more detailed view of the micro-vessel labeling (Arrows). Both inner and overall diameters of the BM rings appear to be larger and more intense in the specimen from the patient with Stage IV disease, compared to specimens from the control subject and the patient with Stage II disease. Scale bars represent 100 μm. Median Collagen IV abundances per micro-vessel were determined by Quantitative Fluorescence Microscopy, computed as the product of ring area and mean labeling intensity, and are presented as Fence Box Plots. Collagen IV abundance was significantly increased in patients with Stage IV PAD compared to control subjects and Stage II patients. Data were analyzed by the randomization t-test where: a-significantly different at p = 0.029. b-significantly different at p = 0.044.

The Abundance of Collagen I, a Principal Contributor to Fibrosis, Is Increased in Terminal Micro-Vessels in Calf Muscle of PAD Patients with More Advanced Fibrosis
Collagen I production is increased substantially in models of pathological fibrosis [16], in association with increased expression of TGFβ1 [17]. Type I pericytes, micro-vessel subendothelial cells that produce Collagen I and have been implicated as the principal source of myofibroblasts in models of fibrosis [16]. In a previous study [15], our laboratory established the presence of hyperplastic VSMC, laden with TGFβ1, in small vessels (50 to 150 μm diameter) of the calf muscle of   D-F)), providing a more detailed view of the micro-vessel labeling (Arrows). Both inner and overall diameters of the BM rings appear to be larger and more intense in the specimen from the patient with Stage IV disease, compared to specimens from the control subject and the patient with Stage II disease. Scale bars represent 100 µm. Median Collagen IV abundances per micro-vessel were determined by Quantitative Fluorescence Microscopy, computed as the product of ring area and mean labeling intensity, and are presented as Fence Box Plots. Collagen IV abundance was significantly increased in patients with Stage IV PAD compared to control subjects and Stage II patients. Data were analyzed by the randomization test where: a-significantly different at p = 0.029. b-significantly different at p = 0.044. DM, RI and FH were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV patients). For response variable Collagen IV abundance, the predictive value (adjusted R 2 ) of the model was 0.103 (p = 0.099) where Group (p = 0.099) was not a significant predictor and DM (p = 0.965), RI (p = 0.513), and FH (p = 0.270) were not significant predictors. With DM, RI and FH excluded, the model had a predictive value of 0.120 (p = 0.026) and Group (p = 0.0258) was a significant predictor. For response variable Micro-vessel Density, the predictive value (adjusted R 2 ) of the model was 0.150 (p = 0.043) where Group (p = 0.010) was a significant predictor and DM (p = 0.128), RI (p = 0.547), and FH (p = 0.916) were not significant predictors.

The Abundance of Collagen I, a Principal Contributor to Fibrosis, Is Increased in Terminal Micro-Vessels in Calf Muscle of PAD Patients with More Advanced Fibrosis
Collagen I production is increased substantially in models of pathological fibrosis [16], in association with increased expression of TGFβ1 [17]. Type I pericytes, micro-vessel subendothelial cells that produce Collagen I and have been implicated as the principal source of myofibroblasts in models of fibrosis [16]. In a previous study [15], our laboratory established the presence of hyperplastic VSMC, laden with TGFβ1, in small vessels (50 to 150 µm diameter) of the calf muscle of patients with PAD. Patients with Stage IV disease were characterized by advanced fibrosis. In this study, we evaluated the possibility of micro-vessel peri-endothelial deposition of Collagen I in the calf muscle of control subjects and patients with Stage II and Stage IV PAD. Slide specimens of calf muscle biopsies collected from control subjects and patients with PAD were labeled with anti-Collagen I antibody. Collagen I deposition presented as well-defined micro-vessel rings (Figure 7) that were generally larger in the calf muscle of Stage IV patients. Using Quantitative Fluorescence Microscopy, we quantified micro-vessel Collagen I abundance in slide specimens labeled with anti-Collagen I antibody (Figure 7; Fence Box Plot). For each micro-vessel, abundance was determined as the product of mean fluorescence intensity (mean pixel) and area occupied by Collagen I. Medians (M) and interquartile ranges (IQR) were 23,058 gsu·m 2 , 15,877 gsu·µm 2 ; 21,272 gsu·µm 2 , 13,369 gsu·µm 2 ; and 42,660 gsu·µm 2 , 26,412 gsu·µm 2 for control subjects and Stage II and Stage IV PAD patients, respectively. In a randomization test, Collagen I abundance of Stage IV patients was greater than that of control subjects (p < 0.002) and Stage II patients (p < 0.001). Collagen I abundances in specimens of control subjects and Stage II PAD patients were not significantly different (p = 0.305).
DM, RI and FH were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV patients). For Collagen I Abundance as response variable, the predictive value (adjusted R 2 ) of the model was 0.316 (p < 0.002) where Group (p < 0.007) was a significant predictor and DM (p = 0.738), RI (p = 0.689), and FH (p = 0.987) were not significant predictors.

Collagen I and Collagen IV Were Co-Localized in the Basement Membrane of Calf Muscle Biopsies from Control Subjects and Patients with PAD
The patterns of deposition of Collagen IV ( Figure 6) and Collagen I (Figure 7) in micro-vessels of calf muscle biopsies of control subjects and patients with PAD are similar and suggest that both may be expressed in the basement membrane of the micro-vessels. To test this possibility, we evaluated co-localization of the fluorescence signals of these collagens in calf muscle micro-vessels. Slide specimens of calf muscle biopsies from control subjects and patients with PAD were labeled concurrently with a rabbit antibody specific for Collagen I and a mouse monoclonal antibody specific for Collagen IV. Subsequently, the specimens were treated with a mixture of AlexaFluor 555 conjugated goat anti-rabbit IgG and AlexaFluor 647 conjugated goat anti-mouse IgG. Each microscopic field (40× objective) was captured in two fluorescence channels corresponding to the secondary fluorescent antibodies. Fluorescence images of Collagen I deposition pseudo-colored red and images of Collagen IV deposition pseudo-colored green are presented as independent images and as merged images of control subjects, Stage II PAD patients and Stage IV PAD patients (Figure 8). Fibrosis in specimens of Stage II and Stage IV patients is evident as inter-myofiber labeling with anti-Collagen I antibody in both the separate and merged images. Fibrosis is more pronounced in the Stage IV specimens. The merged images demonstrate co-localization of the Collagen I and Collagen IV signals (yellow) in the basement membrane of micro-vessels in control, Stage II and Stage IV specimens (Figure 8). Selected regions in Panels (A-C) were enlarged (Panels (D-F)), providing a more detailed view of micro-vessel labeling (Arrows). The inner and overall diameter of the micro-vessel rings appears to be larger and more intense in the specimen from the patient with Stage IV disease, compared to specimens from the control subject and the patient with Stage II disease. Scale bars = 100 µ m. Median Collagen I abundances per micro-vessel were determined by Quantitative Fluorescence Microscopy, computed as the product of ring area and mean labeling intensity, and are presented as Fence Box Plots. Collagen I was significantly increased in patients with Stage IV PAD compared to control subjects and Stage II patients. Data were analyzed by the randomization test where: a-significantly different at p < 0.002. b-significantly different at p < 0.001. PAD were labeled with anti-Collagen I antibody, and fluorescence images were captured with a widefield microscope (40× objective). Selected regions in Panels (A-C) were enlarged (Panels (D-F)), providing a more detailed view of micro-vessel labeling (Arrows). The inner and overall diameter of the micro-vessel rings appears to be larger and more intense in the specimen from the patient with Stage IV disease, compared to specimens from the control subject and the patient with Stage II disease. Scale bars = 100 µm. Median Collagen I abundances per micro-vessel were determined by Quantitative Fluorescence Microscopy, computed as the product of ring area and mean labeling intensity, and are presented as Fence Box Plots. Collagen I was significantly increased in patients with Stage IV PAD compared to control subjects and Stage II patients. Data were analyzed by the randomization test where: a-significantly different at p < 0.002. b-significantly different at p < 0.001.

Pericyte Coverage of Micro-Vessels in Calf Muscle Specimens Was Greater in Patients with Stage IV Disease Compared to Controls and Patients with Stage II Disease
Pericytes, along with endothelial cells, are required for basement membrane assembly and remodeling [18]. In the present study, we documented basement membrane remodeling in the calf muscle of patients with Stage II and Stage IV PAD. Late-stage disease is characterized by increased basement membrane thickening and diameter in association with fibrosis and greater abundances of basement membrane Collagen IV and Collagen I. Consequently, we evaluated pericyte coverage of micro-vessels in relation to progression of PAD. Slide specimens of calf muscle biopsies were labeled concurrently with antibodies specific for Collagen IV, αSmooth Muscle Actin (αSMA), and endothelial cell CD-31 ( Figure 9). Pericytes were identified as α-Smooth Muscle Actin (α-SMA) fluorescence labeling intimately associated with Collagen IV labeling and endothelial cell labeling (CD-31) of micro-vessels (overall diameter < 15 µm) [19].
J. Clin. Med. 2020, 9, x FOR PEER REVIEW 14 of 21 Figure 8. Co-localization of Collagen I and IV deposited as distinctive rings associated with microvessels in the calf muscle of control subjects and patients with PAD. Biopsy specimens embedded in paraffin were sectioned at four microns and mounted to glass slides. Slide specimens were labeled simultaneously with anti-Collagen I and anti-Collagen IV antibodies, and subsequently, with goat anti-rabbit and goat anti-mouse secondary antibodies (respectively) coupled with fluorophores. Images of each slide specimen were taken in separate fluorescence channels corresponding to each collagen label, with a widefield microscope (40× objective). Distinctive rings of Collagen I (Red) and Collagen IV (Green) delineate micro-vessels (Arrows) and are co-localized (Yellow) in the basement membrane. Myofibers are identified as large black areas delineated by both Collagen I and Collagen IV. Labeling of inter-myofiber Collagen I (Red) is readily apparent in the merged images of the Stage II and Stage IV specimens. Scale bars represent 50 microns.

Pericyte Coverage of Micro-Vessels in Calf Muscle Specimens Was Greater in Patients with Stage IV Disease Compared to Controls and Patients with Stage II Disease
Pericytes, along with endothelial cells, are required for basement membrane assembly and remodeling [18]. In the present study, we documented basement membrane remodeling in the calf muscle of patients with Stage II and Stage IV PAD. Late-stage disease is characterized by increased basement membrane thickening and diameter in association with fibrosis and greater abundances of basement membrane Collagen IV and Collagen I. Consequently, we evaluated pericyte coverage of micro-vessels in relation to progression of PAD. Slide specimens of calf muscle biopsies were labeled concurrently with antibodies specific for Collagen IV, αSmooth Muscle Actin (αSMA), and endothelial cell CD-31 ( Figure 9). Pericytes were identified as α-Smooth Muscle Actin (α-SMA) fluorescence labeling intimately associated with Collagen IV labeling and endothelial cell labeling (CD-31) of micro-vessels (overall diameter < 15 μm) [19].

Figure 8.
Co-localization of Collagen I and IV deposited as distinctive rings associated with micro-vessels in the calf muscle of control subjects and patients with PAD. Biopsy specimens embedded in paraffin were sectioned at four microns and mounted to glass slides. Slide specimens were labeled simultaneously with anti-Collagen I and anti-Collagen IV antibodies, and subsequently, with goat anti-rabbit and goat anti-mouse secondary antibodies (respectively) coupled with fluorophores. Images of each slide specimen were taken in separate fluorescence channels corresponding to each collagen label, with a widefield microscope (40× objective). Distinctive rings of Collagen I (Red) and Collagen IV (Green) delineate micro-vessels (Arrows) and are co-localized (Yellow) in the basement membrane. Myofibers are identified as large black areas delineated by both Collagen I and Collagen IV. Labeling of inter-myofiber Collagen I (Red) is readily apparent in the merged images of the Stage II and Stage IV specimens. Scale bars represent 50 microns.
Because of the very labor-intensive work required for partitioning and quantification of pericyte events in each tissue section, we evaluated pericyte coverage in duplicate slide specimens of calf muscle biopsies from five control subjects, five Stage II patients and five Stage IV patients. Using Quantitative Fluorescence Microscopy, we quantified pericyte coverage as the sum of the areas occupied by pericyte segments per 40× microscopic field. We analyzed five fields of each duplicate slide per control and patient and present the data as medians per group (Figure 9; Fence Box Plot). Medians (M) and interquartile ranges (IQR) were 126 µm 2 , 135µm 2 ; 446 µm 2 , 121 µm 2 ; and 1077 µm 2 , 564 µm 2 for control subjects and Stage II and Stage IV PAD patients, respectively. In a randomization test, pericyte coverage of calf muscle micro-vessels of Stage IV patients, was greater than that of control subjects (p = 0.006) and Stage II patients (p = 0.010). Pericyte coverage in Stage II specimens was greater than that of control specimens (p = 0.008). To determine whether differences in coverage across control subjects and PAD patients was due simply to increased micro-vessel density (cf. Results, Section 3.4), we determined pericyte coverage as the area of pericyte events per micro-vessel. Coverage per micro-vessel was greater in the calf muscle of Stage IV patients (Median = 16.0 µm 2 ; IQR = 11.0) compared to control subjects (Median = 6.0 µm 2 ; IQR = 1.5) and Stage II patients (Median = 10.0 µm 2 ; IQR = 3.0) at p = 0.008 and 0.009, respectively. The difference between control subjects and Stage II patients was significant (p = 0.015).
DM Because of the very labor-intensive work required for partitioning and quantification of pericyte events in each tissue section, we evaluated pericyte coverage in duplicate slide specimens of calf Stage IV (N = 5) PAD were labeled with antibodies specific for Collagen IV (Blue), αSmooth Muscle Actin (αSMA) (Red), endothelial cell CD-31 (Green) and nuclei (Grey/White). Fluorescence images were captured with a widefield microscope (40× objective). Pericyte segments were identified as αSMA positive labeling intimately associated with the micro-vessel endothelium. Micro-vessels of control muscle (A) had few pericyte segments, and these occupied very small areas. The number and size of pericyte labeling events increased in patients with Stage II PAD (B) and were greater in patients with Stage IV PAD (C). Scale bar = 25 µm. Pericyte coverage per microscopic field was determined by Quantitative Fluorescence Microscopy as the sum of the areas occupied by pericyte segments. We analyzed five 40× fields in duplicate slide specimens per patient and control. Coverage for controls and patients with Stage II and Stage IV disease is presented as Fence Box Plots, and differences were analyzed by the randomization test where: a-significantly different at p = 0.009; b-significantly different at p < 0.009; c-significantly different at p < 0.007.

Discussion
The myopathy in the lower extremities of patients with PAD is a consequence of atherosclerotic occlusive disease of the named inflow arteries (macrovascular disease) supplying the legs. This macrovascular disease characteristically affects the aortic bifurcation, the superficial femoral artery at Hunter's hiatus and the popliteal trifurcation [20]. In the present study, we extend (1) our previous work [15] establishing a pathologic remodeling of micro-vessels (50-150 µm diameter) and collagenous investment of the smallest micro-vessels (≤15 µm diameter) in the calf muscle of PAD patients and (2) the work of Baum et al. [21] demonstrating pathologic thickening of the basement membrane (BM) of terminal micro-vessels in lower leg muscles of PAD patients. We document pathological events in terminal micro-vessels (≤15 µm diameter) of the calf muscle of PAD patients, establishing that vascular pathology associated with the legs of PAD patients extends beyond the atherosclerotic main inflow arteries and affects the entire vascular tree down to the smallest micro-vessels. Pathological events included increased BM thickness, diameter and abundances of Collagen I and Collagen IV. Moreover, pericyte coverage of terminal micro-vessels was increased significantly in both Stage II and Stage IV disease. Generally, micro-vessel pathology was most severe in patients with Stage IV disease ( Figure 10). Compared to controls, terminal micro-vessel density in the calf muscle of Stage II and Stage IV patients was increased by 17% and 27%, respectively, suggesting an adaptive response to oxygen and nutrient deficits in the PAD muscle. The nearly universal micro-vessel pathology in the calf muscle specimens suggests that this adaptive response was frustrated. = 10.0 μm 2 ; IQR = 3.0) at p = 0.008 and 0.009, respectively. The difference between control subjects and Stage II patients was significant (p = 0.015).
DM, RI and FH were evaluated as covariates in a General Linear Model for fixed factors (Group: Control, Stage II and Stage IV patients). For response variable Pericyte Coverage, the predictive value (adjusted R 2 ) of the model was 0.745 (p < 0.003) where Group (p = 0.001) was a significant predictor and DM (p = 0.277), RI (p = 0.773), and FH (p = 0.758) were not significant predictors.

Discussion
The myopathy in the lower extremities of patients with PAD is a consequence of atherosclerotic occlusive disease of the named inflow arteries (macrovascular disease) supplying the legs. This macrovascular disease characteristically affects the aortic bifurcation, the superficial femoral artery at Hunter's hiatus and the popliteal trifurcation [20]. In the present study, we extend (1) our previous work [15] establishing a pathologic remodeling of micro-vessels (50-150 μm diameter) and collagenous investment of the smallest micro-vessels (≤15 μm diameter) in the calf muscle of PAD patients and (2) the work of Baum et al. [21] demonstrating pathologic thickening of the basement membrane (BM) of terminal micro-vessels in lower leg muscles of PAD patients. We document pathological events in terminal micro-vessels (≤15 μm diameter) of the calf muscle of PAD patients, establishing that vascular pathology associated with the legs of PAD patients extends beyond the atherosclerotic main inflow arteries and affects the entire vascular tree down to the smallest microvessels. Pathological events included increased BM thickness, diameter and abundances of Collagen I and Collagen IV. Moreover, pericyte coverage of terminal micro-vessels was increased significantly in both Stage II and Stage IV disease. Generally, micro-vessel pathology was most severe in patients with Stage IV disease ( Figure 10). Compared to controls, terminal micro-vessel density in the calf muscle of Stage II and Stage IV patients was increased by 17% and 27%, respectively, suggesting an adaptive response to oxygen and nutrient deficits in the PAD muscle. The nearly universal microvessel pathology in the calf muscle specimens suggests that this adaptive response was frustrated. Micro-vessel architectural changes across stages of PAD are depicted in the cartoon. The number of αSMA + pericytes (red) increases with PAD severity around the endothelium (blue) of micro-vessels less than 15 μm in diameter. The BM (grey) is thickened in association with increased deposition of Collagen I and Collagen IV in the micro-vessels of PAD patients. As the area of αSMA + pericytes increases, the BM becomes thicker and the diameter of the BM annulus increases.
Our work demonstrating increased thickness of the BM of terminal micro-vessels in the calf muscle of patients with PAD confirms the findings of Baum et al. [21]. In this study, we have extended this work by showing coordinated increases in both thickness and diameter of the BM ring of microvessels in association with inter-myofiber and microvascular fibrosis in the calf muscle of PAD patients with Stage IV disease. Recent studies have identified pericytes as the principal precursors of myofibroblasts which produce fibrosis in skeletal muscle and other organ systems [22,23]. In phenotype tracing studies, Birbrair et al. [24] identified Type-1 pericytes as fibrogenic in a mouse Micro-vessel architectural changes across stages of PAD are depicted in the cartoon. The number of αSMA + pericytes (red) increases with PAD severity around the endothelium (blue) of micro-vessels less than 15 µm in diameter. The BM (grey) is thickened in association with increased deposition of Collagen I and Collagen IV in the micro-vessels of PAD patients. As the area of αSMA + pericytes increases, the BM becomes thicker and the diameter of the BM annulus increases.
Our work demonstrating increased thickness of the BM of terminal micro-vessels in the calf muscle of patients with PAD confirms the findings of Baum et al. [21]. In this study, we have extended this work by showing coordinated increases in both thickness and diameter of the BM ring of micro-vessels in association with inter-myofiber and microvascular fibrosis in the calf muscle of PAD patients with Stage IV disease. Recent studies have identified pericytes as the principal precursors of myofibroblasts which produce fibrosis in skeletal muscle and other organ systems [22,23]. In phenotype tracing studies, Birbrair et al. [24] identified Type-1 pericytes as fibrogenic in a mouse model of age-dependent skeletal muscle fibrosis and showed that they became fibrogenic myofibroblasts when exposed to TGF-β1. In a study of a mouse model of kidney fibrosis, Wu et al. [25] found that TGF-β1 signaling from damaged tubular epithelial cells activated micro-vessel pericyte-to-myofibroblast transition, causing fibrosis in the kidney. Thus, TGF-β1 production and secretion in a location other than terminal micro-vessels, signals micro-vessel pericytes to produce Collagen I and transition to myofibroblasts which migrate to the ECM and produce organ fibrosis. Expression of both TGF-β1 and Collagen I were shown to increase in a mouse model of chronic kidney disease in response to ischemia/reperfusion [17]. Ischemia/reperfusion is central to the myopathy of the lower legs of PAD patients [26]. In a previous study, we documented fibrotic remodeling of micro-vessels (50 to 150 µm diameter), in the calf muscle of patients with PAD [15]. The collagenous media of these vessels were thickened and contained hyperplastic VSMC that were laden with TGF-β1. In the same study, we noted "collagenous investment" of micro-vessels (≤15 µm diameter). We propose that micro-vessel remodeling documented in the present study is, at least in part, a response to TGF-β1 originating in the hyperplastic VSMC of neighboring, larger micro-vessels. On this basis, we evaluated Collagen I deposition in the terminal micro-vessels of the calf muscle of PAD patients and found a significant increase in Collagen I abundance in the BM of these micro-vessels in patients with Stage IV PAD. Collagen I deposition co-localized with that of Collagen IV, a marker for BM, in terminal micro-vessels. We conclude that the pathology of micro-vessels in PAD patients includes a BM fibrosis at the level of terminal micro-vessels and is likely mediated by TGF-β1 activated pericytes. We anticipate that these changes in the BM of terminal micro-vessels have an adverse effect on micro-vessel compliance, blood flow and nutrient and oxygen diffusion.
The same features of vascular remodeling we have documented for the calf muscle of patients with PAD [15] have been documented for Pulmonary Arterial Hypertension (PAH) [27,28]. Both pathologies are linked to tissue hypoxia and are characterized by substantial thickening of vascular media caused by extensive proliferation and hypertrophy of VSMC that exhibit increased expression of TGF-β1, and deposition of extracellular matrix; features that extend to normally non-muscularized microvasculature. In a more recent study of PAH, Ricard et al. [29] documented abnormally increased pericyte coverage of micro-vessels, attributing this to pericyte proliferation induced by Fibroblast Growth Factor-2 produced and secreted by dysfunctional endothelial cells. TGF-β1 promoted pericyte transitioning to "smooth muscle-like cells" in agreement with Wu et al. [25] and Birbrair et al. [24] who showed that TGF-β1 stimulated pericyte-myofibroblast transitioning but did not stimulate pericyte proliferation. In the present study, we document increased pericyte coverage of terminal micro-vessels in the calf muscle of PAD patients. The greatest increase occurred in the calf muscle of patients with Stage IV PAD. We propose that fibrotic deposition of Collagen I in the BM of terminal micro-vessels is a response to TGF-β1 signaling from the hyperplastic VSMC of neighboring, larger micro-vessels. Future studies are needed to determine whether endothelial cells are damaged in the calf muscle of PAD patients. However, our study of oxidative damage in the calf muscle of PAD patients identified abnormal accumulation of carbonyl adducts in the micro-vessels of the calf muscle in these patients (Weiss et al. 2013, Figure 1) [14]. Carbonyl adducts may contribute to quenching of endothelial cell nitric oxide, producing endothelial cell dysfunction that drives pericyte-mediated fibrosis of the BM and the ECM [30].
The architectural changes observed in this study are consistent with reports of perfusion deficits in PAD patients. Several studies, using contrast-enhanced ultrasound [31], have documented a delayed time to peak microvascular blood flow in PAD patients relative to controls after exercise [32][33][34] and post-occlusive reactive hyperemia [35]. During moderate contractile exercise, microvascular blood volume was slightly lower than controls, whereas microvascular blood velocity was substantially diminished in PAD patients. Furthermore, Duerschmied et al. [33] demonstrated, using contrast-enhanced ultrasound, that blood traversed the muscle more slowly in PAD patients compared to controls, with the greatest delays in contrast transit time arising within the microvascular compartment, which is consistent with our observation of increased BM thickness and diameter of terminal micro-vessel in association with increased pericyte coverage of these micro-vessels. The BM thickening of PAD micro-vessels described in our study may contribute directly to the slowing of the microcirculation by decreasing compliance of the terminal micro-vessels.
A limitation of this study is the relatively small number of patients analyzed. Although we found no significant contribution of DM, FH, and RI to microvascular pathology, the small number of patients per group limited the statistical power of the analysis. Micro-vessel thickening has been observed in the microvascular beds of several organs in patients with DM and patients with RI. However, in patients analyzed in this study, BM thickening and increased diameter, increased pericyte coverage of micro-vessels, microvascular fibrosis, and myofibrosis did not differ between patients with and without DM or RI, which suggests that PAD alone is enough to induce and propagate microvascular pathology. DM microangiopathy is well characterized in several organs. Death of pericytes, and consequently, microvascular instability and decreased capillary density are hallmark features of diabetic microvascular complications and are well documented features of diabetic retinopathy [36]. Our observations that microvascular density and pericyte coverage are increased in PAD patients with and without DM suggests that the microangiopathy of PAD is different from, and likely dominant to the microangiopathy observed in DM. Additional studies are necessary to better characterize the contribution of insulin resistance and DM to PAD microvascular pathology.
Regarding the microvascular disease several, questions remain. Future research needs to address the contribution of oxidative damage to micro-vessel pathology, including BM fibrosis and increased pericyte coverage. Understanding the relationship between microvascular pathology and micro-perfusion offers a unique opportunity for the development of non-invasive diagnostic evaluation of PAD limbs and may have great utility in the assessment of the efficacy of different therapies. Furthermore, non-invasive imaging modalities, such as contrast-enhanced ultrasonography [31,32,35] and magnetic resonance imaging [37,38], could enhance our understanding of the functional alterations of the microvasculature of PAD limbs. Understanding PAD microvascular pathology may offer opportunities for more individualized therapies, which may include angiotensin-inhibitors, anti-hypertensives, statins, and DM medications, and may aid in the identification of patients who are likely to experience significant disease progression and need more aggressive therapy.

Conclusions
Our study establishes the extension of vascular pathology of PAD beyond macrovascular atherosclerosis to pathologic fibrosis and remodeling of the smallest micro-vessels in ischemic muscles of the lower legs of PAD patients. The microvascular disease of PAD is characterized by increased thickness and diameter of the BM in association with increased Collagen I deposition, an index of fibrosis, in the BM, and increased pericyte coverage of the terminal micro-vessels. Fibrosis of the BM likely is driven, at least in part, by TGF-β1 signaling from hyperplastic VSMC in neighboring, larger micro-vessels, which promotes Collagen I production by pericytes and their transitioning to myofibroblasts. In addition, oxidative damage (which is central to the myopathy of PAD) may serve as a central mechanism of endothelial cell damage, and subsequent signaling that drives pericyte proliferation and increased pericyte coverage of the terminal micro-vessels. Our findings provide insight into the pathophysiology of PAD and direction for future mechanistic studies, and consequently, a basis for improved diagnosis and treatment for patients with PAD.