Methods

The Institutional Animal Care and Use Committee approved all the procedures (A17807). Twenty-one domestic pigs (weight, 55 to 65 kg) were studied after 6 weeks and again after 10 to 12 weeks of observation. In 14 pigs (Pork Partners, Stewartville, Minn), a local-irritant coil was placed in the main renal artery at baseline and induced gradual development of unilateral RAS. We have shown previously that by 4 to 5 weeks after coil implantation, this model already exhibits a chronic decrease in renal function and that significant stenoses induce pathological renal alterations that closely resemble chronic renovascular disease in humans, including inflammation, fibrosis, and mild glomerulosclerosis.^10,11,15,16 The pigs had renal angiography 6 weeks after induction of RAS and were then randomized into 2 groups that were not further treated (RAS group; n=7) or treated with an intrarenal infusion of autologous EPCs (RAS+EPC group; n=7). To generate EPCs, peripheral mononuclear cells were isolated from each pig 3 and 5 weeks after induction of RAS, expanded in vitro, and delivered to the same RAS+EPC pig during the 6-week renal angiography. Four weeks later, all animals underwent repeated renal angiography, as well as in vivo studies. Throughout these 10 weeks, blood pressure was monitored continuously with a telemetry system (PhysioTel, Data Sciences) implanted at baseline in the left femoral artery. Mean arterial pressure (MAP) was recorded with telemetry at 5-minute intervals and averaged for each 24-hour period^10,11,17 and was also measured during in vivo studies with a side arm of the arterial catheter. The other 7 pigs were used as controls (normal group; n=7).

By 6 weeks after induction of RAS, all of the pigs underwent renal angiography, as mentioned above. For this, all the pigs were anesthetized with intramuscular Telazol (5 mg/kg) and xylazine (2 mg/kg), intubated, and mechanically ventilated with room air. Anesthesia was maintained with a mixture of ketamine (0.2 mg/kg per minute) and xylazine (0.03 mg/kg per minute) in normal saline, administered via an ear vein cannula (0.05 mL/kg per minute). Under sterile conditions and fluoroscopic guidance, an 8F arterial catheter was advanced to the stenotic renal artery, proximal to the stenosis. Short bolus injections (4 to 6 mL) of low-osmolar nonionic contrast media (iopamidol, Isovue-370; Squibb Diagnostics, Princeton, NJ) were used to visualize the lumen of the renal artery with a fluoroscopy system (Siemens Siremobil Compact), and images were then recorded and later analyzed offline to determine the degree of RAS, as described previously.¹⁷ After angiography, in the RAS+EPC animals, EPCs (10⁶ cells/mL suspended in 10 mL of saline) were delivered into the stenotic renal artery (for details, please see the online-only Data Supplement).

Four weeks later, all the animals were again anesthetized similarly for repeated renal angiography, which was followed by in vivo functional studies. After angiography, the catheter was positioned in the superior vena cava, and in vivo helical multidetector computed tomography (CT) flow studies were performed for assessment of basal regional renal perfusion, renal blood flow (RBF), and glomerular filtration rate (GFR), as detailed previously.^9–11,15,18 Briefly, this involved sequential acquisition of 160 consecutive scans after a central venous injection of iopamidol (0.5 mL/kg per 2 seconds), repeated during suprarenal infusion of acetylcholine (5 μg/kg per minute) to test endothelium-dependent responses. Blood samples were collected from the inferior vena cava and both renal veins for measurement of plasma renin activity (PRA) (radioimmunoassay) and systemic asymmetrical dimethylarginine (ADMA) levels (Euroimmun US LLC, Boonton Twp, NJ). Urine samples were collected by suprapubic bladder puncture, and protein content was measured by spectrophotometry with the Bradford method.

After completion of all studies, the pigs were allowed to recover for a few days (to allow for contrast media washout) and were then euthanized with a lethal intravenous dose of sodium pentobarbital (100 mg/kg; Sleepaway, Fort Dodge Laboratories, Inc, Fort Dodge, Iowa). Both kidneys were removed from each pig with the use of a retroperitoneal incision and immersed in 4°C Krebs’ solution containing heparin. A lobe of tissue was immersed in 10% buffered formalin (Sigma, St Louis, Mo), and a segmental artery perfusing the intact end of the stenotic kidney was cannulated and prepared for micro-CT. Other lobes were shock-frozen in liquid nitrogen and stored at −80°C or preserved in formalin.^10,11,15

In vitro studies were then performed to assess renal histology and expression of angiogenic and fibrotic factors. Western blotting and immunohistochemistry were used to probe expression of the proangiogenic factors phosphorylated (p)-Akt, p-endothelial nitric oxide synthase (p-eNOS), vascular endothelial growth factor (VEGF), hypoxia-inducible factor (HIF)-α, angiopoietin-1, and integrin β₃. The expression of markers and mediators of renal fibrosis such as transforming growth factor (TGF)-β, tissue inhibitor of metalloproteinases (TIMP)-1, α-smooth muscle actin (α-SMA), and matrix metalloproteinases (MMP)-2 and -9 was also investigated. Furthermore, microvascular and renal tissue remodeling were assessed in 5-μm midhilar renal paraffin-embedded slices stained with trichrome, and the presence of resident progenitor cell in the kidney was assessed by immunoreactivity of Oct-4.¹⁹ Double immunofluorescence for DiI and CD31 or cytokeratin was used to localize the EPCs in renal vessels or tubules, respectively.

Results

Renal Function

MAP and the angiographic degree of stenosis were similarly and significantly greater in RAS and RAS+EPC animals compared with normal animals, whereas systemic PRA and ADMA levels were similar among the groups, as was renal vein PRA (Table and Figure 1, top). Basal RBF, cortical perfusion, and GFR were all diminished in RAS, but RBF and GFR significantly improved after EPC treatment (ANOVA P<0.03 for RBF and GFR, ANOVA P=NS for perfusion). Similarly, RBF and GFR responses to the endothelium-dependent vasodilator acetylcholine that were blunted in RAS were restored in RAS+EPC, suggesting improved renovascular endothelial function (Figure 1, bottom). These were accompanied by a significant increase in renal expression of p-eNOS in RAS+EPC, implying greater potential for NO availability.

View this table:

• View inline
• View popup

Table. Characteristics and Basal Stenotic Kidney Function in Normal, RAS, and RAS Pigs Treated With Autologous EPCs (RAS+EPC)

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Top, Changes in blood pressure over time (n=3 to 4 animals per group). Bottom, RBF (A), GFR (B), and cortical perfusion (C) at baseline and in response to acetylcholine in normal, RAS, and RAS treated with autologous EPCs (RAS+EPC) animals. EPC infusion did not decrease blood pressure in RAS. The blunted basal and challenged renal hemodynamics and function in RAS were normalized after EPC treatment, suggesting restoration of renovascular endothelial function. *P<0.05 vs normal, †P<0.05 vs baseline, **P<0.05 vs normal (2-way repeated-measures ANOVA).

Angiogenic Factors

The blunted expression of proangiogenic HIF-1α and VEGF in RAS was increased after EPC treatment (ANOVA P=0.004 and P=0.01, respectively), suggesting a proangiogenic milieu in the treated RAS kidney. Furthermore, renal expression of p-Akt and p-eNOS, key mediators of VEGF, were significantly augmented in RAS+EPC compared with untreated RAS. This was accompanied in RAS+EPC by increased renal expression of integrin β₃ (19.7±3.3 versus 10.6±2.8 and 12.9±1.5 positive vessels per field in RAS and normal, respectively; ANOVA P=0.02) and angiopoietin-1 (ANOVA P=0.03; Figure 2), suggesting that EPCs not only promoted angiogenesis but also favored the maturation of the new vessels.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Representative immunoblots (top) and densitometric quantification (bottom) demonstrating renal protein expression of p-Akt, p-eNOS, VEGF, HIF-1α, and angiopoietin-1 in the following groups: normal, RAS, and RAS treated with autologous EPCs (RAS+EPC). Replenishment of EPC in RAS augmented the expression of vascular growth factors in the stenotic kidney, suggesting a proangiogenic stimulus. *P<0.05 vs normal, †P<0.05 vs RAS, #P=0.079 vs normal.

Microvascular 3-Dimensional Architecture

Microvascular density was diminished in RAS across the renal cortex (inner, middle, and outer cortex). Notably, RAS+EPC substantially increased cortical microvascular density in all the cortical regions (although it remained lower than normal), resulting in improved vascular volume fraction (ANOVA P=0.0002; Figure 3). Microvascular tortuosity was significantly increased in RAS+EPC kidneys compared with both RAS and normal controls (1.87±0.09 versus 1.30±0.02 and 1.39±0.09, respectively; ANOVA P=0.0007), supporting the notion of abundant angiogenic vessels. Furthermore, CD31 expression on capillaries was significantly reduced in RAS compared with normal (0.29±0.04% versus 1.91±0.6%; P=0.01) but improved in RAS+EPC (1.13±0.6%; P=0.3 versus normal and P=0.05 versus RAS), suggesting augmented capillary proliferation in the treated kidney.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Representative 3-dimensional tomographic images of renal cortex and medulla (top) and quantification (bottom) from normal, RAS, and RAS treated with autologous EPCs (RAS+EPC) kidneys. EPCs in RAS augmented intrarenal microvascular density throughout the cortex, which consequently restored renal vascular volume fraction. *P<0.05 vs normal, †P<0.05 vs RAS.

Renal EPCs and Morphology

An average of 13.3±0.8% of the total injected EPCs was detected 4 weeks later; the cells were evident at the tubular and vascular compartments of the treated stenotic kidneys. Some of the injected EPCs were detected incorporated in very small (capillaries and vasa vasorum) microvessels (CD31⁺) and in tubules (cytokeratin⁺) and costained with these markers, suggesting that they assumed endothelial and tubular characteristics (see Figure V in the online-only Data Supplement). Interestingly, the number of Oct-4⁺ cells in renal tubules was significantly increased after EPC treatment, many of which costained with DiI fluorescence, indicating that they originated from injected EPCs. Nevertheless, only EPC-treated stenotic kidneys also showed Oct-4⁺ cells unlabeled with DiI (Figure 4A), suggesting activation and mobilization of circulating or resident stem cells (Figure 4B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. A, Representative double fluorescence of CM-DiI (red) and immunoreactivity of Oct-4 (green, ×40) in frozen renal sections. B, Representative staining (×40), Western blotting bands, and densitometry (bottom) of renal immunoreactivity to Oct-4 in normal, RAS, and RAS treated with autologous EPCs (RAS+EPC) kidneys. Some Oct-4⁺ cells are of EPC origin (dashed arrow), but others are not (white arrow), suggesting recruitment of endogenous progenitor cells. *P<0.05 vs normal, †P<0.05 vs RAS.

RAS kidneys showed increased expression of TGF-β and α-SMA and decreased MMP-2 (Figure 5A), which were accompanied by increased glomerulosclerosis and perivascular and tubulointerstitial fibrosis compared with normal controls (Figure 5B), overall suggesting renal remodeling. Importantly, EPCs improved the expression of those factors and decreased fibrosis in the stenotic kidney, without fully normalizing it (ANOVA P=0.001). Conversely, EPCs did not affect apoptosis, as the fraction of apoptotic cells was significantly and similarly elevated in RAS and RAS+EPC compared with normal (0.59±0.1%, 0.52±0.2%, and 0.09±0.01%, respectively; P<0.04).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. A, Representative immunoblots and densitometric quantification demonstrating renal protein expression of TGF-β, MMP-2, MMP-9, and TIMP-1, and pictures showing renal immunoreactivity of α-SMA⁺ cells (arrows), likely myofibroblasts. B, Renal trichrome staining (×40) and quantification of renal fibrosis and glomerulosclerosis in normal, RAS, and RAS treated with autologous EPCs (RAS+EPC) kidneys. EPCs significantly decreased profibrotic activity and attenuated renal remodeling in RAS. *P<0.05 vs normal, †P<0.05 vs RAS.

Contralateral Kidney

The presence of labeled cells and changes in morphology in the contralateral kidney were determined. Interestingly, although few EPCs were observed in the contralateral kidney of RAS+EPC animals (1 to 2 cells per slide), VEGF expression was increased compared with both normal and untreated RAS pigs. However, unlike the RAS+EPC kidney, the contralateral kidney of RAS and RAS+EPC animals did not show any changes in tubulointerstitial fibrosis (2.09±0.1% and 1.83±0.05%, respectively; P<0.01 versus normal), blunted number of capillaries (0.64±0.04% and 0.70±0.04%, respectively; P<0.05 versus normal), or renal expression of TGF-β (elevated compared with normal), MMP-2 and -9 (unchanged), and TIMP-1 (attenuated compared with normal), implying lack of effect of these cells in remodeling of the contralateral kidney (Figure 6). Apoptosis also remained evident (1.22±0.47 and 1.07±0.30, respectively; P<0.05 versus normal).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6. Left, Representative immunoblots and densitometry demonstrating protein expression of TGF-β, MMP-2, MMP-9, TIMP-1, and VEGF in the contralateral kidney (clk). Right, Renal trichrome staining (×40) of normal kidneys and contralateral kidney of RAS and RAS treated with EPCs (RAS+EPC) animals. Intrarenal infusion of EPCs in the stenotic kidney only augmented VEGF in the contralateral kidney but did not have any effect in attenuating renal fibrosis or regulating the expression of its mediators. *P<0.05 vs normal.

Discussion

The present study shows, for the first time, the feasibility of a cell-based approach to treat the ischemic kidney distal to the stenosis. A single intrarenal infusion of autologous EPCs during the evolution of RAS restored the hemodynamics and function of the ischemic kidney, preserved microvascular architecture, and attenuated renal remodeling. The capability of EPCs to restore vascular integrity and preserve the stenotic kidney may therefore enable development of novel therapeutic renoprotective strategies in chronic renovascular disease.

Renovascular disease is often associated with atherosclerosis but constitutes a strong predictor for increasing morbidity and mortality independent of other cardiovascular risk factors,⁸ as is a decrease in GFR.³³ Renal disease is often characterized by decreased microvascular density as well as tubulointerstitial fibrosis, which are determinants of renal outcomes.³⁴ Regression of intrarenal microvasculature accompanies many forms of renal disease such as diabetes and aging, and microvascular remodeling correlates with development of renal scarring.³⁵ Indeed, we have previously shown that renal injury is evident by 1 month after development of the RAS^16,31 and that deterioration of renal hemodynamics and function is paralleled by significant fibrosis,^10,11 microvascular rarefaction, and remodeling.^9,12 Microvascular and parenchymal damage distal to the obstruction likely contributes to renal dysfunction observed in humans with chronic renovascular disease despite revascularization, but few therapeutic options are available to restore renal viability and vascular integrity.

In many ischemic and injured organs, mobilization, homing, and transdifferentiation of EPCs play an important role in augmenting neovascularization³⁶ and endothelial replacement after vascular injury.^22,24 EPCs augment angiogenesis both by stimulating the secretion of angiogenic growth factors and by providing a source of progenitor cells that can differentiate into mature vascular endothelial cells. They appear to confer their beneficial effect not only by their long-term engraftment and local retention in the injured tissue but also by transient secretion of vascular growth factors within this region.³⁷ Yet, despite the promise of EPC delivery in treating diseases associated with blood vessel disorders, the potential of this strategy to salvage the ischemic kidney in RAS has not been explored. The present study indicates that this strategy is feasible and effective. In the present study, we delivered into the stenotic kidney a combination of early EPCs, which retain monocytic characteristics and avid angiogenic activity, and late EPCs, which exhibit more mature endothelial-like features. Previous studies have demonstrated that these cell types synergistically enhance angiogenesis more effectively than each cell type alone.²⁰ This intervention elicited new vessel formation and reversed most of the functional and structural deterioration of the stenotic kidney. We observed that some of the injected cells assumed endothelial and tubular phenotypes and incorporated into renal structures. However, when the relatively small number of EPCs retained in the tissue is considered, their autocrine and paracrine activities were likely key for the increase in protein expression and improvement in renal function.

Angiogenesis involves a sequence of events regulated by numerous factors that results in development of new vessels. Among those factors, VEGF is crucial for preservation of the microvasculature and, in concert with other factors, stimulates processes responsible for cell division, migration, and survival, extracellular matrix degradation, and tube formation that generate, repair, and maintain microvascular networks. We have previously shown both HIF-1α and VEGF to be paradoxically downregulated in chronic RAS,^9,12 attended by microvascular rarefaction and renal fibrosis. HIF-1α is the most important transcription factor driving VEGF mRNA expression and production and is considered a crucial primary defense mechanism for the adaptive response to ischemia in the kidney.³⁸ Remarkably, intrarenal administration of EPCs in the stenotic kidney restored HIF-1α, VEGF, eNOS, Akt, and angiopoietin-1, all of which stimulate or mediate angiogenesis. Angiopoietin-1 is an endothelial cell survival factor that in concert with VEGF can promote angiogenesis, maturation of the new vessels, and vascular repair,³⁹ and the concurrent improvement of both may have had an additive effect on vascular proliferation and maturation in RAS+EPC.⁴⁰ Augmented neovascularization was also reflected by the increases in renal expression of integrin β₃, peritubular CD31⁺ capillaries, microvascular density, and tortuosity, which are all indices of angiogenic vessels. Therefore, EPCs not only attenuated microvascular dysfunction and rarefaction in RAS but also improved maturation and stabilization of the new vessels, thereby improving the overall hemodynamics and function of the stenotic kidney. Although apoptosis remained elevated in RAS+EPC, this therapeutic approach significantly decreased the scarring in the stenotic kidney, possibly as a result of increased blood supply and NO availability after EPC treatment. Furthermore, it may have also been related to a decreased fibrogenic activity and improved matrix turnover, as suggested by the improved expression of TGF-β and MMP-2 and decreased fibrosis and glomerulosclerosis in RAS+EPC, which in turn may have facilitated restoration of the microvascular network as well.

The number of CFUs that we found is comparable to previous studies in which similar techniques were used.²⁶ We observed that CFUs were similar in normal and RAS animals, suggesting that circulating EPCs were not depleted in RAS, yet RAS EPCs showed increased proliferation potential and better angiogenic function (tube formation) compared with controls. These characteristics differ from EPC function in humans with chronic refractory hypertension⁴¹ but are more consistent with other clinical studies on EPCs in essential or pregnancy-induced hypertension^42,43 or with recent studies in comparable models.⁴⁴ It is possible that duration and nature of the disease, as well as species differences, may have contributed to these differences. For example, a preferential increase in PRA at the early stage (4 to 5 weeks) of the disease³¹ might favor angiotensin II–mediated angiogenic activity⁴⁵ that would gradually dissipate as the disease progresses and oxidative stress increases.³¹ Interestingly, the present study shows that delivery of autologous EPCs into the stenotic kidney also increased number of cells positive for Oct-4, a stem cell transcription factor and marker suggestive of pluripotency.¹⁹ Although many of these were likely the Oct-4⁺–injected EPCs, at least some of them were not labeled with the EPC markers, suggesting that this intervention also promoted mobilization of resident or homing of endogenous circulating progenitor cells. Either by incorporating and differentiating in the tissue host and/or by autocrine and paracrine activity, EPCs are capable of stimulating the function and proliferation of surrounding progenitor and mature cells.²⁰ Importantly, some of the injected EPCs were detected engrafted into blood vessels and tubules and assumed at least some endothelial and tubular features (CD31 and cytokeratin expression), although the level of functionality that the engrafted cells achieved remains to be determined. Interestingly, EPCs were observed incorporated only in very small microvascular vessels, likely peritubular capillaries and vasa vasorum, but not in larger vessels. Presumably, longer transit times and greater surface area facilitate their contact, adherence, retention, and migration into tissue structures.

Importantly, our study showed that EPCs restored the blunted RBF and GFR of the RAS kidney, which might be a result of the preserved microvasculature but also is likely due to augmented availability of eNOS-derived NO, as implied by the increased renal expression of activated eNOS.⁵ In turn, NO is indispensable for microvascular sprouting by maintaining vasodilation during the early steps of angiogenesis⁴⁶ and by promoting VEGF-induced capillary proliferation.⁴⁷ Furthermore, eNOS-derived NO might have contributed to enhance Oct-4 expression in resident progenitor cells and promoted endothelial differentiation⁴⁸ in the RAS+EPC kidney. Hence, upregulated eNOS not only favored renal microvascular endothelial function but may have also contributed to angiogenesis in the RAS+EPC kidney. Moreover, VEGF is also a potent vasodilator that may have contributed to recovering endothelial function in RAS+EPC.⁴⁹ It is interesting that although the degree of stenosis and hypertension in RAS remained unchanged by EPCs, renal function improved with intrarenal EPC therapy. Although the remaining obstruction of the renal artery and decrease in renal perfusion pressure might have been sufficient to activate the intrarenal renin-angiotensin system, renal vein PRA did not lateralize to the stenotic side. Alternatively, an increase in oxidative stress may also sustain hypertension at the chronic phase of untreated RAS when PRA declines.³¹ In addition, we cannot rule out the possibility that disease in the contralateral kidney results in volume retention and thereby mediates hypertension. Indeed, the sustained fibrosis and decreased capillary density in both the stenotic and contralateral kidneys indicate residual injury that might have also contributed to the persistent hypertension.

In summary, the present study shows the renoprotective effects of EPCs in a model of chronic RAS. A targeted intervention using autologous EPCs during the evolution of the disease reversed most of the functional and structural deterioration of the stenotic kidney in this otherwise progressive disease. It is likely that restoration of the angiogenic cascade by autologous EPCs involved not only generation of new vessels but also acceleration of their maturation and stabilization. This contributed to preserving the blood supply, hemodynamics, and function of the RAS kidney and thereby decreased renal remodeling. Future studies are needed to examine the feasibility of this approach after a longer duration of chronic renal ischemia in humans and in the presence of additional cardiovascular risk factors.