Morphological and Pharmacological Characterization of the Porcine Popliteal Artery: A Novel Model for Study of Lower Limb Arterial Disease

Abstract

Objective:

This study was undertaken to characterize structural and pharmacological properties of the pig popliteal artery in order to develop a novel system for the examination of lower limb blood flow regulation in a variety of cardiovascular pathologies, such as diabetes-induced peripheral artery disease.

Methods:

Popliteal arteries were isolated from streptozocin-induced diabetic pigs or age-matched saline-injected control pigs for morphological study using transmission electron microscopy and for examination of vasoreactivity to pharmacological agents using wire myography.

Results:

Transmission electron microscopy of the porcine popliteal artery wall revealed the presence of endothelial cell-smooth muscle cell interactions (myoendothelial junctions) and smooth muscle cell-smooth muscle cell interactions, for which we have coined the term “myo-myo junctions.” These myo-myo junctions were shown to feature plaques indicative of connexin expression. Further, the pig popliteal artery was highly responsive to a variety of vasoconstrictors including norepinephrine, phenylephrine, and U46619, and vasodilators including acetylcholine, ADPβS, and bradykinin. Finally, 2 weeks after streptozocin-induced diabetes, the normalized vasoconstriction of the pig popliteal artery to norepinephrine was unaltered compared to control.

Conclusions:

The pig popliteal artery displays structural and pharmacological properties that might prove useful in future studies of diabetes-associated peripheral artery disease and other lower-limb cardiovascular diseases.

Introduction

Peripheral artery disease (PAD), especially of the lower limb, is a major pathology associated with diabetes mellitus and contributes to the high rates of foot and lower leg amputation in diabetic patients.^1–3 Arterial occlusion induced by PAD, a manifestation of atherosclerosis, reduces blood flow to tissues and generates a variety of symptoms.^1–3 Critical limb ischemia, tissue necrosis, and gangrene are severe, advanced-stage side effects induced by PAD that generally require limb amputation.^1,2,4 A less severe but still debilitating symptom of PAD is intermittent claudication, an exercise-induced cramping or aching in the legs that occurs because blood flow distal to arterial obstruction is insufficient to match metabolic demand.^1,3,5,6 However, this symptom often is not experienced by diabetic patients because of the concurrent peripheral neuropathy that occurs with the disease; the lack of pain sensation prevents early diagnosis of PAD in these patients.^1–3,6 Because PAD remains largely asymptomatic until it progresses to a severe manifestation, its prevalence in the diabetic population is likely underestimated.^1–3 PAD also occurs with increasing frequency in both diabetic and non-diabetic aging individuals.³

Since it is responsible for supplying blood flow to the entire lower leg, the popliteal artery plays a critical role in maintaining the health of a diabetic limb and in assessing treatment options.^1,7,8 Anatomically, the popliteal artery branches from the femoral artery, runs behind the kneecap, and terminates by branching into the anterior tibial artery and tibioperoneal trunk.⁹ PAD can present anywhere from the iliac artery to the plantar artery in the foot, but diabetic patients are more often afflicted in the distal part of the leg,^3,4,8 underlining the importance of popliteal artery blood flow. Upon diagnosis of PAD, bypass grafts or endovascular procedures are commonly used treatments that aim to restore blood flow to tissues downstream of the occlusion.^4,8,10 However, in severe cases of diabetic PAD with significant necrosis and gangrene, amputation must occur.^1,2,11 In those severe cases, the extent of the amputation (i.e. above- or below-the-knee) depends predominantly on blood flow to and oxygenation of the tissue.^1,3,4 Above-the-knee amputations, which occur in conditions of poor popliteal blood flow, are associated with increased mortality and an overall lower quality of life and are considered a high-risk surgery.^11,12 Thus, maintaining sufficient popliteal artery blood flow to the lower leg is vital for maintaining the health and well-being of diabetic patients with PAD.

Due to the growing prevalence of diabetes in developed countries^3,13 and the importance of lower limb blood flow in diagnosing and treating PAD, an animal model of the popliteal artery is an important tool for researchers of diabetic PAD and other lower limb ischemic pathologies. The physical and pharmacological properties of rat^14–16 and mouse^17,18 popliteal and infrapopliteal arteries^14,16 have been studied, but may not exactly mimic the properties of the human disease because of size and species differences.¹⁹ Pigs, whose coronary and retinal arteries are routinely studied because of their similarity to human vessels,^20–23 may provide a more relevant model for the examination of lower limb blood flow,^24–26 especially in conditions of diabetic PAD.

The purpose of this study was to perform ex vivo studies of pharmacological and structural features of the pig popliteal artery. The popliteal artery seemed an ideal choice to examine because it plays a major role in human diabetic PAD, and the pig displays important similarities to humans.^24–26 We employed wire myography, a high-throughput method of examining isolated vessel function, to characterize the pharmacological properties of the pig popliteal artery, and we employed electron microscopy to characterize structures of the vascular wall relevant for signaling. Overall, in this study, we developed a novel model to examine lower limb blood flow regulation mechanisms in the pig.

Materials and Methods

Animals.

Domestic (Yorkshire) male pigs (6–12 weeks old) were purchased from Real Farms (San Antonio, TX, USA) and employed as part of a tissue sharing scheme. These animals were also used for examination of retinal and coronary arteries, in addition to this study of the popliteal artery. Two subsets of animals were utilized in this study. The first subset (15 pigs, 14.4–26.5 kg at sacrifice) was used to characterize the popliteal artery. These pigs received either an intravenous saline injection, intraocular siRNA injection, or no injection and were considered to be untreated. In the second subset of animals, diabetes was induced in some pigs (n=6) through intravenous ear injections of streptozocin (STZ; Zanosar®, 200 mg/kg in saline), which selectively ablates pancreatic β-cells; control pigs (n=5) received matched saline injections.^22,23 Following injection, the animals had free access to water and were fed hog chow mixed with syrup to temporarily prevent hypoglycemia for 24 hours after injection.^22,23 Thereafter, the animals had free access to water and were fed commercial hog chow. The animals were maintained for approximately 2 weeks with close monitoring of general condition, body weight and glucose levels. STZ-injected animals were considered diabetic and included in the study only if fasting blood glucose levels exceeded 250 mg/dL.

Pigs were anesthetized with Telazol® (4–8 mg/kg, intramuscular injection) and maintained at a surgical plane with 2–5% isoflurane inhalation.^20–23 After euthanasia by heart excision, the posterior hindlimb skin was removed and superficial muscles were excised to reveal the gastrocnemius. Proximal portions of both medial and lateral heads of the gastrocnemius muscle were collected from both legs, ensuring that the popliteal artery was undamaged (Fig 1). All animal procedures were approved by the Baylor Scott & White Institutional Animal Care and Use Committee.

Figure 1. Schematic outline of the procedure to examine porcine popliteal artery function.

After the gastrocnemius muscle (pink) was isolated from the animal, a bundle of vasculature, nerve fibers, connective tissue, and fat was dissected from the proximal region between the two popliteal muscle heads (blue outline). Extraneous tissue was cleared away from the artery, which was then either left intact or cut into 2 mm segments (pink outline). Segments left intact were fixed for electron microscopy, while 2 mm segments were mounted in a DMT wire myograph using two 40-μm-diameter stainless steel wires (yellow outline).

Dissection of the Popliteal Artery.

Immediately after removal from the animal, the gastrocnemius muscle was placed into a tissue container filled with chilled MOPS-buffered physiological salt solution (PSS) containing (in mM): 145 NaCl, 4.7 KCl, 2 CaCl₂, 1.17 MgSO₄·7H₂O, 3 MOPS, 1.2 NaH₂PO₄·H₂O, 5 glucose, 2 sodium pyruvate, 0.02 EDTA, 2.75 NaOH (pH 7.40±0.02). The gastrocnemius was then washed 2–3 times in chilled MOPS-buffered PSS, and placed in a large glass dissection dish filled with chilled MOPS-buffered PSS. The popliteal artery runs between the medial and lateral heads of the gastrocnemius in a “bundle” with two flanking veins, a nerve fiber, connective tissue, and fat. The entire bundle was removed from the gastrocnemius muscle and moved to a Sylgard-bottomed glass dish filled with chilled MOPS-buffered PSS. All dissection was performed using a stereomicroscope (Amscope SM-4T, Amscope, Irvine, CA, USA). After pinning out the bundle, surrounding tissues were dissected away from the popliteal artery. For anatomical consistency, we isolated the same 8 mm region of popliteal artery immediately upstream of the branch point from which the anterior tibial artery originates.⁹ The isolated region of the popliteal artery was either cut into 2 mm segments for wire myography experiments or left intact for TEM studies. These four segments were not statistically different in vessel diameter or pharmacological response to norepinephrine (NE, data not shown).

Transmission Electron Microscopy.

After excision of the popliteal artery, as described above, it was fixed in 4% PFA + 2.5% glutaraldehyde for 30 minutes, then shipped overnight on ice in PSS to the University of Virginia School of Medicine’s Advanced Microscopy Facility. After arrival, the vessels were fixed with 1% osmium tetraoxide, followed by dehydration in a gradient of alcohol and embedding in Epon.^27,28 Ultrathin sections (75 nm) were cut, carbon coated, and imaged using a JEOL 1230 transmission electron microscope (JEOL, Peabody, MA, USA) equipped with a 4Kx4K CCD camera from SIA (SIA, Duluth, GA, USA) as previously described.^27,28

Wire Myography.

After the popliteal artery was cut into 2 mm segments, two 40-μm-diameter stainless steel wires (Goodfellow, Huntingdon, UK) were advanced through the lumen of each vessel segment. The segments could then be mounted to the jaws of a Mulvany-Halpern DMT wire myograph (model 610A, Danish Myo Technology, Aarhus, Denmark): one wire was fixed to a force transducer, while the other was fixed to an adjustable micrometer.^29–32 The temperature of the myograph was then increased to and maintained at 37ºC for the duration of the experiment. Each vessel was normalized to a resting tension approximating the vessel’s wall tension at 80 mmHg in vivo, as previously described.^30–32 Solutions were changed every 30 minutes. Before beginning protocols, increasing doses of NE followed by increasing doses of acetylcholine (ACh) were used to assess the viability of the smooth muscle cells (SMCs) and endothelial cells (ECs), respectively. Vessels generating less than ~15 mN of raw tension after preconstriction with NE (100, 300 nM, and 1 μM) were excluded from further experimentation, as were vessels that failed to relax by more than ~90% to ACh (30, 100, and 300 nM).

Pharmacological Agents.

Constrictors used to examine SMC function consisted of: 1) increasing concentrations of isosmotic KCl (Fisher Chemical, P217), a nonspecific depolarizing agent, 2) thromboxane A₂ mimetic U46619 (U4; Cayman, NC9336776), a thromboxane prostanoid (TP) receptor agonist, 3) NE (Sigma-Aldrich, A9512), an α₁ and α₂ adrenergic receptor agonist, and 4) phenylephrine (PE; Sigma-Aldrich, P6126), an α₁ adrenergic receptor agonist. Diethylamine NONOate (DEA; Sigma-Aldrich, D184), a source of exogenous NO, served as an SMC-dependent vasodilator.

Endothelium-dependent vasodilators consisted of: 1) ACh (Sigma-Aldrich, A6625), a muscarinic receptor agonist, 2) carbachol (CCh; Sigma-Aldrich, PHR1511), a more stable analog of ACh that also activates muscarinic receptors, 3) ADPβS (Sigma-Aldrich, A8016), a purinergic P2Y₁ receptor agonist, 4) bradykinin (BK; Sigma-Aldrich, B3259), a bradykinin receptor agonist, and 5) substance P (SP, Sigma-Aldrich, S6883), a neurokinin (NK) receptor agonist.

All pharmacological agents were dissolved in water, with the exception of U4 (DMSO), BK (0.1 mM acetic acid), and SP (0.1 mM acetic acid). The maximum DMSO concentrations used to solubilize U4 in this study, 0.3 μL/mL, are below concentrations that have been shown to affect vascular function.³³ Acetic acid control curves, represented by grey triangles throughout the results, were performed in parallel with BK and SP concentration response curves to ensure that the solvent did not influence vascular tone. All stock solutions were prepared at 10⁻² M, except for BK and SP (5×10⁻⁴ M), and subsequently diluted in MOPS buffer. During concentration response curves, vessels were incubated with NE, PE, and U4 for 2 minutes per concentration, KCl for 10 minutes per concentration, and all vasodilators for 1 minute per concentration. Preconstriction was achieved by adding NE to the myograph chamber for 2–4 minutes. The dose of NE used to preconstrict was established by first performing a full NE concentration response curve for each artery to determine the maximum constriction, then adding a dose of NE that produced 70% of the maximum constriction.

Data Analysis.

Vasoconstrictor data are presented as baseline subtracted and normalized to the length of the vessel, as previously described.^30–32 Vasodilator data are presented as a percent of preconstricted tension. If multiple segments of popliteal artery from the same animal were tested with the same drug, as in the case of NE and ACh, the data from each animal were averaged together and considered n=1. Data from diabetic animals are represented by black circles throughout the results, while control or untreated animals are represented as white circles throughout the results. Pressure estimations of diabetic and control popliteal arteries were performed based on the Law of LaPlace,^30–32 which states: P = T / r, where P = pressure, T = active tension, and r = radius. The radius was obtained from the normalization process, which estimates the vessel diameter, while active tension was derived from raw traces according to the relationship: T = (F – F₀) /( 2 * l), where F = force, F₀ = force at baseline, and l = vessel length. Pressure estimations were performed on each individual vessel segment; multiple pressure-diameter curves from the same animal were then averaged together and considered n=1.

The force reading of the vessels was converted from analog to digital using an iWorx interface and recorded using LabScribe (iWorx, Dover, NH, USA). An unpaired t-test or a two-way ANOVA followed by a Bonferroni post-hoc test was used to compare data, as appropriate. To compare agonist responses, we used nonlinear regression analysis to calculate logEC₅₀, the logarithm of the molar concentration required to achieve half of the maximal response. Thus, logEC₅₀ values throughout are expressed in units of log(molarity). A value of P<0.05 was considered statistically significant. All statistical tests were performed in Prism (GraphPad, La Jolla, CA, USA). Data are presented as mean ± SEM.

Results

Morphological Characterization.

Dissection of the pig popliteal artery revealed that the vessel had a thick medial layer. Subsequent TEM confirmed this initial observation, as it was found that the vessels displayed an EC layer separated from multiple layers of SMCs by an internal elastic lamina (IEL). Electron micrographs also displayed structures that facilitate intercellular communication in the vessel wall of the porcine popliteal artery. The myoendothelial junction (MEJ), a distinct heterocellular signaling domain characterized by the physical interaction of EC and SMC projections through the IEL,^34,35 is present in the vessel wall (Fig 2A). In addition to the EC-SMC interactions at the MEJ, SMC-SMC interactions, which we have termed “myo-myo junctions” (MMJs) are present in the artery as well. MMJs occur when adjacent SMC layers project through intervening collagen layers to make direct contact with one another (Fig 2B-E). Plaques (dark spot indicated by an arrow) were shown to exist at the border of the MMJs (Fig 2C), suggesting connexin expression.^36–39

Figure 2. Transmission electron micrographs of structures involved in intercellular communication within the popliteal artery wall.

The myoendothelial junction, a distinct heterocellular signaling microdomain where an endothelial cell (labeled EC) sends a projection (indicated by the white arrowhead) through the internal elastic lamina (IEL; light gray region) to make contact with an adjacent smooth muscle cell (SMC), is present in the pig popliteal artery (A). Black streaks indicate folding of the fragile IEL layer during tissue preparation. Homocellular MMJs are also present in the pig popliteal artery (B-E), which has multiple layers of smooth muscle cells separated by layers of collagen and other connective tissue. An inset (C) displays a plaque at the MMJ (indicated by a white arrow), which classically indicates the presence of connexins.^36,37

Pharmacological Characterization.

In this study, the porcine popliteal artery was exposed to increasing doses of four different vasoconstrictors, all of which induced a dose-dependent increase in tension. KCl, a nonspecific depolarizing agent, produced the most robust constriction (22.92±1.67 mN/mm at the 90 mM dose) and plateaued at 60 mM KCl (Fig 3A, Table 1). The EC₅₀ of KCl occurred at 29.88±2.59 mM. NE, an α₁ and α₂ adrenergic receptor agonist, produced the next highest peak constriction (17.77±1.63 nM/mm at the 30 μM dose), and displayed a logEC₅₀ of −6.18±0.10 (Fig 3C, Table 1). In contrast, PE, an α₁ receptor agonist, produced a peak constriction of only 9.30±0.91 mN/mm at 10 μM. However, the logEC₅₀ value was similar to NE at −6.25±0.08 (Fig 3D, Table 1). U4, a TP receptor agonist, produced a peak constriction of 12.37±0.69 mN/mm at 1 μM with a logEC₅₀ of −7.29±0.06, a noticeably lower value than NE or PE (Fig 3B, Table 1).

Figure 3. Response of the porcine popliteal artery to constricting agents.

Constriction of the popliteal artery was elicited by treatment with KCl (A, n=5 vessel segments from 5 different pigs), U4 (B, n=5 vessel segments from 5 different pigs), NE (C, n=11; 34 vessel segments from 11 different pigs), and PE (D, n=5 vessel segments from 5 different pigs). Each panel includes a representative trace from the respective vasoconstrictor.

Table 1.

Maximum responses and EC₅₀/logEC₅₀ values for vasoconstrictors. Data summarizing popliteal artery responses to KCl, U4, NE, and PE are displayed in this table. The logEC₅₀ value represents the logarithm of the molar concentration needed to produce 50% of the maximum constriction. E_max represents the maximum constrictor response to each agonist. EC₅₀ for KCl, logEC₅₀ for all others, and E_max were calculated using the cumulative concentration response curves shown in Figure 3. All data are comprised of n=5 vessel segments from 5 different pigs, except for NE (n=34 vessel segments from 11 different pigs).

	KCl (mM)	U4	NE	PE
EC50/logEC50	29.88±2.59	−7.29±0.06	−6.18±0.10	−6.25±0.08
Emax (mN/mm)	22.92±1.67	12.37±0.69	17.77±1.63	9.30±0.91

ACh, a muscarinic receptor agonist, induced a robust dilation (92.02±2.10% at 30 μM) with a relatively low logEC₅₀ value (−7.79±0.10) (Fig 4A, Table 2). CCh, a more stable analog of ACh, displayed similar results, with robust dilation (95.82±1.53% at 3 μM), albeit with a slightly higher logEC₅₀ (−7.07±0.08) (Fig 4C, Table 2). Although ADPβS induced maximal dilation (91.84±1.65%) at 10 μM, a dose of 30 μM ADPβS induced constriction. ADPβS also displayed a high logEC₅₀ value of −5.82±0.16 (Fig 4D, Table 2). DEA produced a maximum dilation of 87.51±3.00% at 10 μM and a logEC₅₀ of −7.14±0.09 (Fig 4B, Table 2). In contrast to the prior four agents, BK and SP were more potent vasodilators, so concentration response curves were initiated at lower doses. BK induced robust vasodilation (98.14±0.81% at 300 nM), with a relatively low logEC₅₀ value (−8.66±0.13) (Fig 5A, Table 2). SP also induced maximum dilation at an even lower dose than BK (91.61±1.77% at 30 nM), displaying a logEC₅₀ of −8.75±0.15 (Fig 5B, Table 2). Like ADPβS, SP induced vasoconstriction at high doses, possibly due to off-target receptor activation.

Figure 4. Response of the porcine popliteal artery to dilating agents.

Dilation of the popliteal artery was elicited by treatment with ACh (A, n=8 vessel segments from 5 different pigs, except at 30 μM, at which point n=4 segments from different 4 pigs), DEA (B, n=5 vessel segments from 5 different pigs), CCh (C, n=5 vessel segments from 5 different pigs), and ADPβS (D, n=5 vessel segments from 5 different pigs). In all cases, preconstriction was achieved by adding a dose of NE that produced a tension equal to 70% of the maximum constriction elicited by a previously performed NE dose response curve. Representative traces of each vasodilator are displayed.

Table 2.

Maximum responses and logEC₅₀ values for vasodilators. Data summarizing popliteal artery responses to ACh, CCh, ADPβS, DEA, BK, and SP are displayed in this table. The logEC₅₀ value represents the logarithm of the molar concentration needed to produce 50% of the maximum dilation. E_max is expressed as the percentage of preconstricted tension. logEC₅₀ and E_max were calculated using the cumulative concentration response curves shown in Figures 4 & 5. All data are comprised of n=5 vessel segments from 5 different pigs, except for ACh (n=8 vessel segments from 5 different pigs).

	ACh	CCh	ADPβS	DEA	BK	SP
logEC50	−7.79±0.10	−7.07±0.08	−5.82±0.16	−7.14±0.09	−8.66±0.13	−8.75±0.15
Emax (%)	92.02±2.10	95.82±1.53	91.84±1.65	87.51±3.00	98.14±0.81	91.61±1.77

Figure 5. Response of the porcine popliteal artery to dilating agents.

Dilation of the popliteal artery was elicited by BK (A, n=5 vessel segments from 5 different pigs) and SP (B, n=5 vessel segments from 5 different pigs). Both BK and SP were dissolved in a solution of 0.1 mM acetic acid, so an acetic acid control curve was performed to verify that the solvent did not affect the results. In the control curve, the volume of acetic acid added at each time point was equivalent to the volume added at respective time points of the BK and SP concentration response curves. Preconstriction was achieved by adding a dose of NE that produced a tension equal to 70% of the maximum constriction elicited by a previously performed NE dose response curve. Representative traces of each vasodilator are displayed.

Vascular Changes Resulting from Diabetes.

Two weeks after STZ injection, diabetic pigs displayed lower body weight (15.8±1.9 kg) and increased blood glucose (411.7±48.1 mg/dL) compared to control counterparts (22.9±2.0 kg and 83.2±7.6 mg/dL, respectively) (Fig 6). Without accounting for the diameter of the vessels being examined, diabetic pigs displayed significantly decreased constriction to NE compared to control animals (Fig 7A). However, the diameter of the popliteal artery from diabetic pigs was significantly smaller than that of control pigs (710.0±27.3 μm vs 910.2±35.5 μm) (.Fig 7B). As a result, when the NE response was normalized to vessel diameter by converting to estimated pressure (as described in the methods), the statistical difference between groups was abolished (Fig 7C).

Figure 6. Characteristics of STZ-treated and control animals.

STZ-treated animals (n=6 pigs) displayed significantly higher blood glucose levels and significantly decreased body weight compared to control animals (n=5 pigs). The blood glucose reading for one STZ-treated pig exceeded 600 mg/dL, the maximum detectable level with the meter that was used. This animal’s blood glucose level was recorded as 600 mg/dL. *P < 0.05.

Figure 7. Constriction to NE is not attenuated in the popliteal artery of diabetic pigs.

In both STZ-injected (n=32 vessel segments from 6 different pigs) and control pigs (n=23 vessel segments from 5 different pigs), increasing concentrations of NE elicited constriction. Without normalizing responses to vessel diameter, STZ-injected diabetic pigs displayed attenuated constriction compared to control pigs (A). However, the popliteal artery diameter of STZ animals was significantly smaller than that of control animals (B). When the data from (A) was re-plotted, but normalized to vessel diameter, all significant differences between control and STZ groups were abolished (C). *P < 0.05 as compared to control.

Discussion

Importance of the Model System.

In this study, we developed a technique for the isolation of the pig popliteal artery that could be later employed to study diabetes-associated PAD as well as other lower limb cardiovascular pathologies. Because the popliteal artery plays a central role in diabetic PAD, we chose to examine this artery; further, the porcine model better approximates the human cardiovascular system than a rat or mouse model.^19,24–26 Morphological studies provided a basic description of vessel characteristics, including size and signaling-associated structures in the vascular wall. Pharmacological studies identified agents to which the vessel responds, indicating the functional receptor profile of the pig popliteal artery. The effect of STZ-induced diabetes was evaluated through the use of NE concentration response curves and the observation of morphological features. Overall, our characterization of the porcine popliteal artery indicates its suitability as a model of diabetes-associated PAD and other lower limb vascular diseases.

Vessel Morphology.

We first characterized the morphology and mechanical properties of the vessel using electron microscopy and wire myography. The pig popliteal artery is a large vessel, averaging 910.2±35.5 μm diameter in non-diabetic animals, with a layer of ECs surrounded by multiple layers of SMCs. Electron micrographs of the vessel displayed the MEJ and MMJ, which may play important roles in heterocellular^40,41 and homocellular^42,43 signaling in the vascular wall, respectively (Fig 2). The MEJ is more commonly found in smaller resistance arteries^39,44,45 and has been shown to play a profound role in controlling vascular tone and thus blood flow.^35,40,46,47 The MEJ serves as a crucial site for the regulation of fundamental vasodilator pathways including endothelium-derived hyperpolarization (EDH)-^34,40,46,47 and nitric oxide (NO)-mediated^48,49 signaling. Recent work has shown that a rise in SMC intracellular calcium concentration leads to calcium⁴⁰ and inositol 1,4,5-trisphosphate⁵⁰ diffusion into and hyperpolarization of ECs, which in turn feeds back on SMCs to oppose the initial constriction. This result builds upon earlier studies that showed increased SMC calcium leads to increased EC calcium, inducing EC NO generation.⁴⁸ In effect, constriction of an artery is actively opposed by an endothelial feedback mechanism, whether EDH- or NO-mediated.⁴⁰ The acetic acid control curve (.Fig 5C) displays a gradual decrease in tension after preconstriction with NE, providing some evidence that feedback dilation does occur in the pig popliteal artery. The gradual decrease in tension observed after preconstriction may indicate that a healthy endothelium is feeding back on preconstricted SMCs. To eliminate the possibility that acetic acid causes vasodilation, an additional curve was performed in which we exposed the vessel to a preconstricting dose of NE and observed the feedback over a 9 minute interval (Supplemental Fig 1). The gradual decrease in tension to approximately 50% of the initial value in the absence of a vasodilator or vehicle control provides further evidence of an active endothelial feedback dilation.

We also detected a homocellular structure that we have termed the MMJ, a morphologically similar junction to the MEJ, present between smooth muscle layers of the popliteal artery (Fig 2B-E). Previous electron microscopy studies of porcine coronary arteries did not show the presence of MMJs,⁵¹ so it is possible that their presence only occurs in specific anatomical locations. In MMJs, adjacent SMCs project through intervening collagen layers to make contact with each other. Fig 2B-C appears to display a plaque, a region traditionally indicative of connexin protein expression,^36–39 at the border of adjacent SMCs. Evidence from vascular co-culture models, in which gap junctions were formed by plating murine SMCs on opposite sides of a Transwell™ membrane, indicates that Cx37 and Cx43 may be present at in vitro MMJs.⁴¹ A recent study showed that inhibition of vascular gap junctions with 18 β-glycyrrhetinic acid prevents the propagation of calcium waves between rat mesenteric SMCs.⁴³ This finding indicates that electrochemical signals may be transmitted through MMJs of SMCs to coordinate vascular constriction.⁴³ Importantly, the majority of the reports on SMC gap junctions have examined mesenteric vessels with 1–2 layers of SMC coverage.^41–43 In contrast, the pig popliteal artery has many SMC layers and thus may be more reliant on electrical coupling of SMCs than smaller arteries. Further, in this artery, electrical signals may need to travel not only to cells in the same SMC layer, but also to cells located several layers closer to or farther from the lumen. Overall, we believe that the presence of MMJs and plaques in the pig popliteal artery supports the idea that the many concentric layers of SMCs are directly electrically coupled by gap junctions, offering a mechanism by which multiple SMC layers in a large artery coordinate uniform constriction and dilation.

Pharmacological Response.

We used four vasoconstrictors and six vasodilators to examine the pharmacological properties of the porcine popliteal artery. The agents were selected for this characterization because previous studies have implicated their involvement in the progression of diabetic PAD in some capacity.^52–57 The pharmacological profiles of these relevant agents may help to guide future studies of the pig popliteal artery. NE, which acts on α₁ (G_q11-coupled) and α₂ (G_s-coupled) adrenoreceptors,⁵⁸ elicited a much greater maximum constriction than PE (17.77±1.63 mN/mm vs. 9.30±0.91 mN/mm, respectively), which only acts on α₁ adrenoreceptors (Fig 3C-D, Table 1).⁵⁸ Importantly, the logEC₅₀ values of NE and PE are nearly identical (−6.18±0.10 vs −6.25±0.08, respectively), showing that sensitivity of the vessels to each agent is not different (Fig 3C-D, Table 1). Overall, this indicates that in the porcine popliteal artery, α₂ adrenergic receptors contribute significantly to vessel constriction. Prior work has shown that adrenoreceptor distribution varies with anatomical location, vessel size, and branch order. Studies of both the rat⁵⁹ and mouse⁵⁸ cremaster vasculature showed that both α₁ and α₂ adrenoreceptors mediate the vasoconstriction of larger arterioles, but only α₂ adrenoreceptor mediate the constriction of smaller precapillary arterioles.^58,59 A study of the gluteus maximus vasculature of C57BL/6J mice indicated a different trend: in this particular skeletal muscle, α₂ adrenoreceptors mediate greater constriction in 1A arterioles, while α₁ adrenoreceptors mediate greater constriction in 3A arterioles.⁵⁸ Additionally, studies of rabbit hindlimb arteries indicate a predominance of α₂ adrenoreceptors in the popliteal region of the hindlimb.^60,61 However, constriction of the thoracodorsal²⁷ and mesenteric⁶² arteries of the mouse are mediated predominantly by α₁ adrenoreceptors. Since adrenoreceptor distribution varies widely by vascular bed, it is valuable to know that constriction of the pig popliteal artery relies primarily upon α₂ adrenoreceptors, with some α₁ adrenoreceptor contribution.

The pig popliteal artery also constricted to U4 (Fig 3B), showing the functional presence of TP receptors on SMCs. Robust dilation to ACh and CCh (Fig 4A,C) indicated the presence and function of endothelial muscarinic receptors. Although ADPβS initially caused dilation of the artery, a 30 μM dose of ADPβS induced vasoconstriction (Fig 4D), which indicates that in addition to acting upon P2Y₁ receptors on the vascular endothelium,^63,64 high concentrations of the drug may nonspecifically activate receptors on SMCs which are also known to express P2Y receptors.^63,65 BK and SP were both shown to be potent vasodilators of the pig popliteal artery, producing maximal dilation at approximately 30 nM (Fig 5). However, high concentrations of SP produced a similar trend as the ADPβS curve: after maximal dilation at 30 nM, SP began to induce constriction. This constriction may indicate that NK1 receptors are present on both ECs and SMCs of the pig popliteal artery, a receptor distribution that has been observed previously in the rabbit jugular vein.⁶⁶

Diabetes Model.

The STZ-treated pig has frequently been used to model diabetes and complications arising from diabetes.^{22,23,67–71} As previous studies^22,23,70 and this present study (Fig 7) have shown, STZ treatment recapitulates the symptoms of human Type I diabetes mellitus, including hyperglycemia and reduced weight gain.⁷² Importantly, the STZ-injected pig has also been used to model the vascular complications of diabetes in retinal^22,23 and coronary^70,71 arteries. However, the STZ pig model has never before been used to study skeletal muscle arteries like the popliteal artery, which is known to be adversely affected in human diabetic patients.^4,8 It has been hypothesized that the vascular response to NE, a potent endogenous vasoconstrictor, may increase in diabetic arteries, although studies have reached conflicting results.^22,70,71 Since reduced blood flow to the lower limb is a prominent side effect of diabetes,^3,4,8 the role of NE-induced vasoconstriction in the popliteal artery of STZ-injected pigs was investigated in this study.

The present study shows that two weeks after STZ injection, there is a decrease in popliteal artery diameter but no change in normalized vasoconstriction to NE (Fig 7). Results from a study of retinal arteries, albeit with a different receptor-mediated vasoconstrictor, show similar results; 2 weeks after STZ injection, vasoconstriction to endothelin-1 (ET-1) was unchanged.²² Though arterial diameter was not reported, further examination showed no change in retinal artery vasoconstriction to ET-1 after 6 and 12 weeks.²² Conduit coronary arteries of pigs subjected to either a high-fat diet (HFD) or STZ + HFD treatment for 2.5 months also displayed unchanged vasoconstriction to ET-1.⁷⁰ However, the same study examined small coronary arteries and showed that 2.5 months after treatment, vasoconstriction to ET-1 was diminished in the STZ + HFD group compared to both control and HFD treatment groups.⁷⁰ In contrast, in a second study of small coronary arteries 15 months after treatment, vasoconstriction of both HFD and STZ + HFD groups to ET-1 was shown to be increased compared to control.⁷¹ Though studies of small coronary arteries show conflicting results and may be confounded by HFD treatment, the data overall seem to support the finding of the present study that arterial vasoconstriction is largely unaltered in diabetic conditions.

Comparisons between the few prior studies of diabetic porcine vascular function and the present study should be made cautiously for several reasons. First of all, arterial function differs based on anatomical location, and retinal and coronary vessels constitute some of the more unique vascular beds in the body.^20,73,74 Hypotheses about skeletal muscle artery function cannot be made solely on the basis of results from specialized circulations. Secondly, arterial size influences function; for example, vasoconstriction in large and small porcine coronary arteries were shown to respond quite differently after 2.5 months of diabetes.⁷⁰ The popliteal artery, a conduit artery,⁷⁵ is not likely to function similarly to microvessels. Third, treatment duration varies highly in these studies of pig arteries; studies range from 2 weeks²² to 15 months,⁷¹ and differences in at least small coronary artery function appear to be tied to treatment duration.^70,71 It is conceivable that STZ treatment longer than 2 weeks might eventually alter the constrictor response in pig popliteal arteries, although a longer treatment did not alter the function of retinal arterioles.²² A final conflicting factor is the fact that a few studies of aortae⁷⁶ and mesenteric⁵² and coronary arteries⁷⁷ from STZ-injected rats have reported an increase in maximum constriction to NE^52,76 and ET-1.⁷⁷ These findings could be due to species difference, tissue difference, or normalization technique, as the authors did not normalize constrictor response to vessel diameter, instead normalizing to tissue wet weight⁷⁶ or not normalizing at all.⁵² It is worth noting that many other studies of various arteries from STZ-injected rats found no change in maximum vasoconstriction to NE.^78–81 Though these factors should be taken into account, it seems reasonable to conclude that STZ treatment does not significantly change the vasoconstrictor response of the porcine popliteal artery.

The recapitulation of experimental results in human studies strengthens the validity of a disease model. Two studies of isolated arteries from diabetic human gluteal biopsies have been performed, and results show that receptor-mediated constriction to NE is unaltered in diabetic arteries.^82,83 An additional in vivo examination of human forearm blood flow indicates that diabetes does not alter arterial vasoconstriction to NE.⁸⁴ These studies provide evidence that human arteries, just like pig popliteal (Fig 7), retinal,²² and large coronary arteries,⁷⁰ do not seem to display altered vasoconstriction to receptor-mediated agents. Though the human vessels that were studied differ in size and anatomical location from the porcine vessels, the similar pharmacological response supports the idea that the STZ-treated pig provides a relevant model for diabetic complications in humans.

Perspectives

This study sought to characterize the pig popliteal artery from a structural and pharmacological perspective and to examine the effect of STZ-induced diabetes on its structural and functional properties. Our results show that this artery is reactive to a number of vasoactive agents, that it contains notable signaling-associated structures in the vascular wall, and that it can be an effective system to model lower-limb vascular pathologies. Overall, we have established a novel system that may serve as a suitable model of lower limb blood flow regulation in conditions of diabetes-associated PAD and other vascular pathologies.

Abbreviations:

ACh

Acetylcholine

ADPβS

Adenosine 5’-[β-thio] diphosphate

BK

Bradykinin

CCh

Carbachol

Cx37/40

Connexin 37/40

DEA

Diethylamine-NONOate

EC

Endothelial cell

EDH

Endothelium-derived hyperpolarization

ET-1

Endothelin-1

IEL

Internal elastic lamina

KCl

Potassium chloride

MEJ

Myoendothelial junction

MMJ

Myo-myo junction

NE

Norepinephrine

NO

Nitric oxide

PAD

Peripheral artery disease

PE

Phenylephrine

PSS

physiological salt solution

SMC

Smooth muscle cell

STZ

Streptozocin

SP

Substance P

TEM

Transmission electron microscopy

TP

Thromboxane prostanoid

U4

U46619