Hypercholesterolemia Attenuates Angiogenesis but Does Not Preclude Augmentation by Angiogenic Cytokines

Animal Model and Study Design

All protocols were approved by St Elizabeth’s Institutional Animal Care and Use Committee. Weight-matched (3.0- to 3.5-kg) 6-month-old male homozygous WHHL (Camm Research) and NZW (Pine Acre Rabbitry) rabbits were used for the experiments. Representative angiograms performed in nonoperated NZW and WHHL rabbits are illustrated in Fig 1⇓. Surgical procedures were performed with the animals under anesthesia with a mixture of ketamine (50 mg/kg) and acepromazine (0.8 mg/kg) after premedication with xylazine (2 mg/). Animals underwent surgical removal of one femoral artery according to a previously described technique.^{7 14} Excision of the femoral artery results in occlusion of the external iliac artery; blood flow to the ischemic limb is consequently dependent on collateral vessels originating from the internal iliac artery.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Representative angiograms performed before surgical intervention designed to establish unilateral hindlimb ischemia in NZW and WHHL rabbits.

Two sets of pilot experiments were performed. The first was designed to document the severity of angiographic, hemodynamic, and blood flow deficits measured 1 day after surgical creation of unilateral hindlimb ischemia in WHHL rabbits (n=5) (see “Methods” for a description of these analyses). Angiography disclosed angiographic scores of 0.19 to 0.24; blood pressure in the ischemic limb was 0 mm Hg in all rabbits, and blood flow in the ischemic limb was reduced to 5.19±0.62 mm Hg. These findings established evidence of severely compromised hindlimb perfusion in WHHL rabbits.

A second set of pilot experiments were performed to evaluate the ability of WHHL rabbits to tolerate unilateral hindlimb ischemia past day 1. These experiments disclosed reproducible limb necrosis in WHHL rabbits (in the absence of exogenous cytokine therapy) if these animals were permitted to survive beyond 10 days after surgery. (In contrast, limb necrosis has not been observed as a consequence of this surgery in NZW rabbits.) In conjunction with our animal care and use committee, we therefore designed the current protocol to perform functional and anatomic analyses of the animals at 10 days after surgery, after which all animals were electively killed.

At 1 day after surgery (day 1), WHHL rabbits were randomly assigned to receive daily injections of rhVEGF (500 μg in 0.1% PBS; WHHL_vegf, n=9; Genentech) or vehicle alone (WHHL_ctrl, n=9) from day 1 to day 6. NZW rabbits (n=9) received daily injections of vehicle alone. Although previous studies¹⁵ have indicated that this dose of rhVEGF leads to a transient reduction in mean arterial pressure (20.5±1.4%), this was typically tolerated by hypercholesterolemic as well as normal NZW rabbits without sequela.

Plasma levels of cholesterol, triglycerides, and CK were evaluated before surgery and 10 days later. Collateral vessel formation, blood flow, and the extent of tissue damage were evaluated at day 10 as described below.

After surgery, all animals were closely monitored. Analgesia (levorphanol tartrate [60 mg/kg]; Roche Laboratories) was administered subcutaneously as required for evidence of discomfort throughout the duration of the experiment. Prophylactic antibiotics (sulfamethoxazole [15 mg/kg] and trimethoprim [3 mg/kg]; Elkins-Sinn) were also administered subcutaneously for a total of 5 days after surgery.

Lower Limb Calf Blood Pressure Index

Calf blood pressure was measured noninvasively in both hindlimbs using a Doppler flowmeter (model 1050; Parks Medical Electronics) and a cuff connected to a pressure manometer.^{7 14} The calf blood pressure index was defined for each rabbit as the ratio of systolic pressure in the ischemic limb to systolic pressure in the normal limb (×100).

Angiography and Doppler Guide Wire Measurements

Angiography and intra-arterial Doppler guide wire measurement of blood flow velocity were performed on day 10 as previously described.^{7 14}

A 3F infusion catheter (Tracker-18TM; Target Therapeutics) was used for angiography and drug infusion, and an 0.018-in Doppler guide wire (Cardiometrics) was used for the measurement of flow velocity. The catheter and wire were introduced into the right common carotid artery through a small cutdown and advanced through the aorta to the proximal segment of the common iliac artery (catheter) and the proximal segment of the ipsilateral external or internal iliac artery (Doppler wire).

Measurements were performed in the external and internal iliac arteries in the nonischemic limb and in the internal iliac artery in the ischemic limb. In addition to measurements performed at rest, endothelium-dependent and -independent flows were measured in each vessel after intra-arterial administration of acetylcholine chloride and nitroprusside, respectively (Sigma Chemical Co) over 2 minutes via a constant infusion pump (1 mL/min). Each drug was administered at a dose of 1.5 μg · min⁻¹ · kg⁻¹.

Doppler-Derived Blood Flow in Nonischemic and Ischemic Hindlimbs

Angiographic luminal diameter was measured in each vessel at the site of Doppler sample volume recording, at rest and after drug infusion, with the use of an automated edge-detection system (Quantum 2000I; QCS) as previously described.⁷ Doppler-derived flow was calculated as Q_D=(πd2/4)(0.5×APV), where Q_D is Doppler-derived time average flow, d is vessel diameter, and APV is time average of the spectral peak velocity.¹⁶ Mean velocity was estimated as 0.5×APV by assuming a time-averaged parabolic velocity profile across the vessel. Doppler-derived flow calculated in this fashion has been previously validated in vivo.¹⁶ Blood flow in the nonischemic hindlimb was calculated as the sum of the flows recorded in the external and internal iliac arteries. In the ischemic limb, flow measured in the internal iliac artery represented the flow to the entire hindlimb.

Angiographic Analysis of Collateral Vessels

Angiographic analysis of collateral vessels supplying the ischemic limb was performed using angiograms recorded 4 seconds after the injection of contrast media into the internal iliac artery. A grid overlay composed of 2.5-mm-diameter circles arranged in rows spaced 5 mm apart was placed over the angiogram at the level of the medial thigh. The number of contrast-opacified arteries crossing over circles as well as the total number of circles encompassing the medial thigh area were counted in a single blinded fashion. An angiographic score was calculated as the ratio of circles crossed by opacified arteries divided by the total number of circles in the ischemic thigh. This angiographic score reflects vascular density in the medial thigh.^{14 17}

Measurement of Muscle Perfusion

Perfusion of hindlimb skeletal muscles was evaluated using 15-μm-diameter colored polystyrene microspheres.^{14 17} After completion of the invasive measurements described above, two sets of 3×10⁶ Dye-Trak colored microspheres (Triton Technology) were injected through a 3F Teflon catheter into the left ventricle. The first injection was performed under resting conditions, and the second was performed during stimulation with an infusion of nitroprusside (1.5 μg · min⁻¹ · kg⁻¹) in the lower aorta. Each time, a reference blood sample was withdrawn using a syringe pump to collect microspheres at a constant rate of 1.2 mL/min through a peripheral artery (central ear artery). The animals were then killed, and 14 tissue samples (weight, ≈2 g each) were retrieved from seven different muscles (tensor fascia latae, vastus lateralis, vastus medialis, adductor, semimembranosus, gastrocnemius, tibialis cranialis) in each hindlimb (ischemic and nonischemic). Samples from the right and left kidneys were also collected and used as controls to determine homogeneity of the blood content of microspheres. After tissue and blood sample digestion using potassium hydroxide, sphere-filtering extraction, and dimethylformamide dye removal, each sample was analyzed with a conventional spectrophotometer (model 8452A; Hewlett Packard).^{14 17} On the basis of the OD measurements, muscle perfusion expressed in mL · min⁻¹ · 100 g⁻¹ was calculated with the following equation: Perfusion of Muscle Sample x (mL · min⁻¹ · g⁻¹)=(OD of Tissue Sample x/OD of Reference Blood Sample)×[Withdrawal Rate of Reference Blood Sample (mL/min)/Weight of Tissue Sample x (g)].

Muscle perfusion in each hindlimb was expressed as the mean of 14 samples. Perfusion in the ischemic hindlimb was also expressed as percent of perfusion in the nonischemic hindlimb, or “muscle perfusion index.”

Capillary Density

Collateral vessel formation was further examined by measuring the number of capillaries in light microscopic sections taken from ischemic and nonischemic hindlimbs.¹⁴ Tissue specimens were obtained from the adductor muscle and semimembranosus muscle of both hindlimbs of each animal at the time they were killed (day 10). These two muscles were chosen because they are the two major muscles of the medial thigh. Samples were embedded in O.C.T. Compound (Miles) and snap-frozen in liquid nitrogen. Frozen sections (5 μm in thickness) were stained for alkaline phosphatase using an indoxyl-tetrazolium method to detect capillary ECs and then counterstained with eosin.¹⁸ Because method it is based on histochemical staining for an enzyme within capillary endothelium, it shows all capillaries present within the tissue, regardless of the extent of tissue perfusion. Capillaries were counted under a 20× objective to determine capillary density (mean number of capillaries/mm²). A total of 20 different fields from the two muscles were randomly selected, and the number of capillaries were counted. To ensure that analysis of capillary density was not overestimated due to muscle atrophy, capillary density was also evaluated as a function of the number of muscle fibers in the histological section (capillary-to-myocyte ratio).

TTC Staining

The application of the oxidation-reduction indicator TTC stains tissue with normal levels of dehydrogenase red, whereas ischemic and infarcted tissue remained unstained due to loss of the enzyme. TTC staining was performed immediately after the animals were killed.¹⁹ The remaining portions of the tensor fascia latae, vastus lateralis, vastus medialis, adductor, semimembranosus, gastrocnemius, and tibialis cranialis that were not used for the microsphere evaluation were further sectioned and incubated in TTC for 30 min at 37°C. Percent unstained/total muscle weight was used as an index of tissue damage. The healthy limb was used as a control for the quality of the staining (absence of unstained muscles in the healthy limb).

Pathological Analysis of Atherosclerotic Lesions

The presence and extent of atherosclerotic lesions were evaluated in tissue samples retrieved from the lower aorta, right and left iliac arteries, and the remaining femoral artery of each rabbit (six sections from each vessel for a total of 24 sections per animal). Tissues were fixed by immersion in 100% methanol, embedded in paraffin, stained by hematoxylin and eosin or elastic-trichrome, and examined by light microscopy.

Statistical Analysis

Results are expressed as mean±SEM. Comparisons were performed using ANOVA followed by Scheffé’s procedure or, for categorical data, contingency table analysis. Statistical significance was assumed at P<.05.

Lipid Levels and Atherosclerotic Lesions

Both WHHL_ctrl and WHHL_vegf rabbits had higher levels of total plasma cholesterol than NZW rabbits (WHHL_ctrl, 686±96; WHHL_vegf, 846±115; NZW, 34±3 mg/dL; P<.0001). Similar results were observed for triglyceride levels (WHHL_ctrl, 527±65; WHHL_vegf, 405±60; NZW, 86±6 mg/dL; P<.0001). No significant differences were observed between the values at day 0 and day 10 or between the two WHHL groups.

Atherosclerotic lesions were identified in 18 of 18 WHHL rabbits. Among 432 tissue sections retrieved from WHHL rabbits, atherosclerotic lesions were observed in 102 (24%). All lesions were focal, raised, and eccentric, and none resulted in significant luminal compromise, which is consistent with what has been described previously in WHHL rabbits of this age.¹¹ The distribution, number, and anatomic extent of lesions were not different for WHHL_ctrl (47 of 216, 22%) versus WHHL_vegf (55 of 216, 25%) rabbits (P=.36). No atherosclerotic lesions were observed in any of the tissue sections (n=216) retrieved from NZW rabbits.

Survival Rate

Two of 9 WHHL_ctrl (8.22%) rabbits died before completion of the follow-up period (day 7 and day 8, respectively). In both cases, examination of the ischemic limb disclosed extensive necrosis. No other cause of death was found. No deaths occurred among NZW or WHHL_vegf rabbits (χ²=4.32; P=. 11). The 2 rabbits who died were excluded from the subsequent analyses.

Angiography

In WHHL_ctrl, collateral artery development in the medial thigh was attenuated compared with that observed for NZW (Fig 2⇓). The administration of rhVEGF, however, markedly improved angiographically visible collateral vessels in WHHL rabbits. An analysis of angiographic score quantitatively confirmed the reduced collateral vessel network in WHHL_ctrl rabbits compared with NZW and WHHL_vegf rabbits (WHHL_ctrl, 0.19±0.02; NZW, 0.35±0.03; WHHL_vegf, 0.47±0.04; P<.001; Fig 3A⇓). The angiographic score for WHHL_vegf was significantly higher than that measured for NZW rabbits (P=.04).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Analysis of angiogenesis by selective internal iliac angiography (A through C) and capillary density (A′ through C′). Compared with NZW rabbits (A and A′), angiographic and histological evidence of angiogenesis is attenuated in ischemic limb of WHHL rabbits receiving vehicle control (B and B′). WHHL rabbits were nevertheless responsive to rhVEGF administration, as illustrated by the increase in angiographically visible collateral vessels and capillary density (C′ and C) compared with animals receiving vehicle control. (Scale is in centimeters.)

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Angiographic score (A) and the capillary density (B) (see “Methods”) were significantly lower in WHHL rabbits (WHHL_ctrl) than in NZW. Administration of rhVEGF (WHHL_vegf) resulted in significant improvement of both indices.

Capillary Density

In normal limb muscles, capillary density was not significantly different among the three groups (P=NS). In contrast, in the ischemic limb muscles (Fig 2⇑), capillary density in sections of muscle harvested from WHHL_ctrl was less than half of that found in sections from NZW but was markedly increased among WHHL_vegf (WHHL_ctrl, 46.4±4.1; NZW, 78.9±4.6; WHHL_vegf, 98.7±4.4/mm²; P<.0002; Fig 3⇑). As was the case for angiographically detected vessels, capillary density was also higher for WHHL_vegf than for NZW (P=.009). An analysis of capillary-to-myocyte ratio yielded similar results (WHHL_ctrl, 0.14±0.01; NZW, 0.23±0.02; WHHL_vegf, 0.28±0.01; P<.0004).

Hemodynamic Measurements

At day 10 after surgery, the blood pressure index was lower among WHHL_ctrl than NZW but improved significantly in response to rhVEGF (WHHL_ctrl, 9.5±3.5%; WHHL_vegf, 46.4±4.6%; NZW, 32.8±2.8%; P<.0001).

Resting blood flow and endothelium-independent blood flow (nitroprusside-induced) were similar in the normal limb of WHHL_ctrl, NZW, and WHHL_vegf rabbits (P=NS; Fig 4A⇓). In contrast, endothelium-dependent blood flow (acetylcholine-induced) was lower in WHHL_ctrl than in NZW rabbits but increased significantly in response to rhVEGF (WHHL_ctrl, 35.9±2.7; WHHL_vegf, 44.9±3.0; NZW, 48.0±3.2 mL/min; P<.05; Fig 4A⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Evaluation of arterial blood flow (A) and muscle perfusion (B) in NZW and WHHL rabbits receiving either vehicle control (WHHL_ctrl) or VEGF (WHHL_vegf). A, Arterial blood flow (mL/min) in each limb (nonischemic and ischemic) was measured at rest, after the endothelium-dependent agonist acetylcholine (Ach), and after the endothelium-independent agonist nitroprusside (NP). B, Muscle perfusion (mL · min⁻¹ · 100 g⁻¹) of each limb (nonischemic and ischemic) measured at rest and after NP.

In the ischemic limb, the resting (WHHL_ctrl, 7.5±0.6; WHHL_vegf, 15.2±0.9; NZW, 13.6±0.7 mL/min; P<.0001), endothelium-dependent (WHHL_ctrl, 8.6±0.7; WHHL_vegf, 25.8±1.7; NZW, 21.8±1.3 mL/min; P<.0001), and endothelium-independent (WHHL_ctrl, 14.0±0.8; WHHL_vegf, 30.2±1.8; NZW, 27.6±1.4 mL/min; P<.0001) blood flows were lower for WHHL_ctrl than for NZW, and all improved to a statistically significant degree after the administration of rhVEGF (Fig 4A⇑). In particular, endothelium-dependent flow was specifically reduced in WHHL_ctrl as demonstrated by the adjustment to endothelium-independent stimulated blood flow; the latter may be considered an index for the size of the vascular bed. Indeed, a reduction in the endothelium-dependent flow/endothelium-independent flow ratio was observed for WHHL_ctrl and improved with rhVEGF (WHHL_ctrl, 0.61±0.03; WHHL_vegf, 0.86±0.03; NZW, 0.79±0.04; P<.002).

Systemic hemodynamic changes accompanying administration of vasoactive agents are summarized in the Table⇓.

Muscle Perfusion

As was the case for measurements performed with the intra-arterial Doppler guide wire, muscle perfusion of the normal limb measured at rest and after nitroprusside provocation did not differ among WHHL_ctrl, NZW, and WHHL_vegf (P=NS).

In the ischemic hindlimb (Fig 4B⇑), however, muscle perfusion at rest was significantly lower in WHHL_ctrl than in NZW but improved significantly after rhVEGF (WHHL_ctrl, 2.99±0.33; NZW, 4.83±0.36; WHHL_vegf, 5.58±0.46 mL · min⁻¹ · 100 g⁻¹; P<.004). Similar findings were recorded in the ischemic hindlimb in response to nitroprusside (WHHL_ctrl, 4.97±0.74; NZW, 9.63±1.06; WHHL_vegf, 11.32±0.92 mL · min⁻¹ · 100 g⁻¹; P<.002). The muscle perfusion index was also lower in WHHL_ctrl than in the two other groups at rest (WHHL_ctrl, 40.09±3.19%; NZW, 65.87±2.03%; WHHL_vegf, 68.32±3.01%; P<.0001) and after nitroprusside provocation (WHHL_ctrl, 21.83±1.80%; NZW, 43.65±5.08%; WHHL_vegf, 49.62±5.96%; P=.001).

Tissue Damage

CK levels did not differ among groups before surgery (P=NS). At day 10, however, the CK level for the WHHL_ctrl group (10 714±1629 U/L) was still more than fivefold that of NZW (2114±325 U/L) and WHHL treated with rhVEGF (1630±475 U/L); P<.0002). Similarly, the percentage of TTC negative (ischemic plus infarcted) muscles was significantly higher for WHHL_ctrl (48.16±5.41%) than for NZW (25.90±3.83%) or WHHL_vegf (29.11±2.90%; P<.004; Fig 5⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Tissue damage, quantified by percentage of TTC-negative muscles in the ischemic hindlimb (see “Methods”), was more severe in WHHL rabbits receiving vehicle control (WHHL_ctrl) than in NZW rabbits or WHHL treated with VEGF (WHHL_vegf).