Abstract We performed studies to further elucidate the mechanisms of angiotensin II (Ang II)–induced angiogenesis of the microvasculature. Rats were placed on a high salt diet (4% NaCl), and Ang II was infused at a subpressor rate (5 ng/kg per minute) for 3 days. Blood pressure was measured daily for 2 control and 3 infusion days. Microvessel density in the cremaster muscle was measured at the end of the infusion. Vessel density in rats that received subpressor Ang II infusion increased by 12.6% compared with rats that received vehicle infusion. When the angiotensin type 2 (AT2) receptor antagonist PD 123319 was coinfused with Ang II, blood pressure was elevated and vessel density increased above that observed with Ang II infusion alone (23% increase). When the AT1 receptor antagonist losartan was coinfused with Ang II, blood pressure was lower than control and vessel density was reduced compared with the Ang II group but was still greater than control (7.8% increase). In this study, Ang II stimulated angiogenesis in the rat cremaster muscle; this effect was enhanced by AT2 antagonism and inhibited by AT1 antagonism. Ang II infusion at a subpressor dose resulted in a pressor response with AT2 antagonism and a depressor response with AT1 antagonism. This suggests that in the microvasculature, the AT1 receptor mediates angiogenesis and vasoconstriction, and the AT2 receptor mediates an inhibition of angiogenesis and vasodilation.
Angiotensin II has been shown to cause cell hypertrophy and hyperplasia in many tissues and cell types.1 2 3 4 5 The AT1 receptor has been implicated in the mediation of cardiac myocyte hypertrophy6 7 and vascular smooth muscle cell hyperplasia in large vessels.8 Less is known about the receptor mechanisms of the trophic actions of Ang II in the microvasculature.
In a study by Wang and Prewitt,9 microvascular density was reduced in hypertensive and normotensive rats treated with captopril for 4 weeks. Furthermore, Hernandez and colleagues10 showed that rats fed a high salt diet (4% NaCl) for 4 weeks also exhibited a reduction in microvascular density, which they attributed to suppression of the renin-angiotensin system caused by elevated salt intake. Chronic infusion of Ang II to maintain constant circulating Ang II levels blocked this reduction in microvessel density. Ang II has also been shown to stimulate angiogenesis in the corneal circulation and the developing chick chorioallantoic membrane model.3 4 11
Although Ang II has been shown to have trophic and angiogenic actions in the microcirculation, receptor mechanisms mediating these effects are not clear. This study confirms the vasoconstrictor and trophic activity of the AT1 receptor, which has previously been documented in larger vessels and other tissues, and indicates vasodilator and growth-inhibitory activity of the AT2 receptor.
All procedures involving instrumentation and handling of animals were done in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by the animal care committee at the Medical College of Wisconsin.
Surgical Implantation of Catheters
Seven- to 8-week-old (200 to 275 g) Sprague-Dawley rats were placed on a high salt diet (4% NaCl, Dyets) 1 day before surgery. Rats were anesthetized with ketamine (82 mg/kg IM) and acepromazine (1.8 mg/kg IM). Polyvinyl chloride catheters were placed in the left femoral artery and vein. Catheters were tunneled subcutaneously and exited through the back between the scapulae. Groin incisions were closed in layers with polyester suture. Catheters were housed in a flexible spring that attached to a swivel device at the top of the rat’s cage. Dacron felt was affixed to a plate at the end of the spring, which was placed subcutaneously. The incision was closed over the plate with a polyester purse-string suture. Rats were allowed to recover for 3 to 4 days.
BP was measured for at least 1.5 hours at the same time of day on 2 control and 3 infusion days. Arterial BP and heart rate were measured with Statham P23id pressure transducers connected to a four-channel BP display unit (Stemtech). The analog signal was low pass–filtered at 100 Hz, sampled at 300 Hz (model RTS-132, Significat), and processed with software of our own design. Acquired data were averaged in 1-minute intervals throughout the measurement period.
Bolus Ang II Infusion
An Ang II bolus (25 ng/kg) was infused through the venous line on the first control day and last treatment day. BP responses were recorded and analyzed for determination of the peak change.
Chronic Agonist and Antagonist Infusions
Saline vehicle (0.9% NaCl), Ang II (5 and 10 ng/kg per minute), Ang II (5 ng/kg per minute) plus the selective AT1 receptor antagonist losartan (a gift from DuPont, 2 and 20 μg/kg per minute) or the selective AT2 receptor antagonist PD 123319 (a gift from Parke Davis, 0.5 and 5 μg/kg per minute), or antagonists alone were infused through the venous catheters for 3 days with a syringe infusion pump (Harvard Apparatus) at a rate of 1 mL/h. Solutions were exchanged daily to prevent degradation of compounds.
Plasma Ang II Measurement
Plasma samples were drawn on the first control day and last treatment day into chilled tubes containing 0.125 mol/L Na2EDTA and 0.025 mol/L phenanthroline. Samples were centrifuged within 15 minutes of collection, and plasma was separated and frozen at −70°C until extraction. Ang II concentration in plasma was measured by the method of Nussberger et al.12 Briefly, angiotensins were extracted from plasma with a C18 column (Waters Associates) and eluted with methanol. Ang II was separated from other angiotensin metabolites by high-performance liquid chromatography (HPLC) and measured by radioimmunoassay with an antibody (kindly provided by Dr Charles Wood, University of Florida) and [125I]iodotyrosyl4-Ang II (No. IM177, Amersham). Recovery from C18 columns was 97% to 99%. Recovery of radiolabeled Ang II from the HPLC column was 78%, and recovery of unlabeled Ang II was 75%. Within-assay variability was less than 8% and between-assay variability was 26% to 30% on three pools followed over 12 assays in 4 months. Variability of the concentration at 50% of maximum binding was 6.29%.
Microvascular Density Measurement
At the end of the third day, rats were killed by pentobarbital overdose. Cremaster muscles were immediately harvested and cut into quarters for analysis of vessel density. These sections were fixed in 2% formalin for 3 days. The sections were immersed in 30 mg/mL rhodamine-labeled Griffonia simplicifolia I lectin for 2 hours, rinsed several times in physiological salt solution, and whole-mounted on a microscope slide with a water-soluble mountant. Samples were then visualized at ×200 with a computerized video fluorescent microscope system with epi-illumination (Olympus). Five to seven fields from each section were randomly selected and saved for subsequent vessel density measurement. We have previously shown that this technique preferentially identifies small arterioles and capillaries13 by binding to the basement membranes of third- and fourth-order arterioles and capillaries. This renders the technique insensitive to alterations in perfusion status and merely identifies vessels that are intact.
Vessel density was determined by our previously described technique.9 Briefly, each microvascular image was enhanced by a series of digital image processing techniques with a commercial software package (Image-1, Universal Imaging Corp) and program of our own design.14 From this, a resultant line image of the network was created, and the intersections of the network and a computer-generated square grid overlay were automatically identified. Intersections within a 0.224-mm2 area field were automatically counted and provided a quantitative estimate of vessel density. Approximately 20 fields were counted and averaged from each rat.
All results are presented as mean±SE. Comparisons between groups were made with a two-factor ANOVA with one repeat (time). A value of P<.05 was considered significant. Significant differences in BP between groups were further analyzed with Bonferroni’s method for multiple comparisons. Significant differences in vessel density between groups were further analyzed with Dunnett’s method for comparison with a control group (vehicle or Ang II, 5 ng/kg per minute). BP responses to bolus Ang II were analyzed with a paired t test, as were differences in plasma Ang II concentrations.
Plasma Ang II concentration in all groups (n=38) was 11.3±1.4 pg/mL before the start of infusion. After a 3-day infusion of Ang II with or without antagonists, plasma Ang II levels in paired samples increased significantly to 19.7±3.5 pg/mL. Paired Ang II levels did not change after a 3-day infusion of vehicle, PD 123319 alone, or Ang II plus PD 123319. In rats infused with Ang II alone (5 ng/kg per minute), plasma Ang II levels increased slightly (+3.4±2.2 pg/mL) but did not reach significance. Ang II levels increased significantly after treatment with Ang II plus losartan (18.5±4.3 pg/mL, n=15), Ang II plus losartan plus PD 123319 (26.3±7.1 pg/mL, n=4), and losartan alone (25.4±3.9 pg/mL, n=4).
Neither vehicle infusion nor Ang II infusion at 5 ng/kg per minute resulted in a change in MAP over the infusion period. However, Ang II infusion at a higher dose (10 ng/kg per minute) resulted in a large increase in MAP (Fig 1⇓, top). Because we were interested in the direct effects of Ang II on microvascular structure without the confounding effects of elevated BP, Ang II was infused at 5 ng/kg per minute in all groups coinfused with receptor antagonists. In some of these groups, coinfusion did result in alterations in BP.
When the selective AT2 receptor antagonist PD 123319 was coinfused with the subpressor dose of Ang II (5 ng/kg per minute), a low dose of the antagonist (0.5 μg/kg per minute) did not alter MAP (Fig 1⇑, middle). When a higher dose of antagonist was coinfused (5 μg/kg per minute), a large increase in MAP resulted which was similar to that seen with the pressor dose of Ang II. Increasing concentrations of the selective AT1 receptor antagonist losartan coinfused with Ang II produced graded decreases in MAP (Fig 1⇑, bottom). Heart rate did not change over time and was not different in any of the infusion groups. Fig 2⇓ shows the effects of infusion of losartan alone (2 μg/kg per minute, n=4), PD 123319 alone (5 μg/kg per minute, n=6), and both antagonists infused with Ang II (n=4). None of these groups showed a change in BP over time compared with vehicle infusion alone.
BP response to a 25 ng/kg bolus infusion of Ang II was determined before the start of and after the chronic infusion treatment period to confirm blockade of the receptors by the antagonists (Table⇓). Vehicle infusion had no effect, whereas the Ang II–infused rats had an exaggerated bolus response after chronic infusion. The group infused with the highest dose of PD 123319 (5 μg/kg per minute) showed no difference in bolus response after chronic infusion. The group infused with the lower dose of PD 123319 (0.5 μg/kg per minute) showed a reduced bolus response after infusion. However, in this group, preinfusion responses were abnormally high compared with all other groups, and postinfusion responses in that group were not different from those in the vehicle-infused group. The groups receiving 2 and 20 μg/kg per minute losartan exhibited a significantly blunted bolus response, suggesting adequate blockade of the AT1 receptor.
Measurement of microvascular density in the cremaster muscle showed an increase in the number of capillaries and small arterioles in rats infused with the subpressor dose of Ang II (5 ng/kg per minute) compared with rats infused with vehicle (Fig 3⇓, top). Infusion of Ang II at the higher, pressor dose (10 ng/kg per minute) resulted in no alteration in microvascular density. This agrees with previous studies in which we have shown an effect of BP that opposes angiogenesis.13 Coinfusion of PD 123319 with Ang II produced a larger angiogenic response than with Ang II alone (Fig 3⇓, middle). Coinfusion of losartan with Ang II inhibited the angiogenesis induced by Ang II at the lowest concentration of losartan (Fig 3⇓, bottom). Coinfusion of both antagonists with Ang II resulted in a similar inhibition of the angiogenic action of Ang II to AT1 antagonism (153±1.9 intersections, n=4). Infusion of either losartan or PD 123319 without coinfusion of Ang II resulted in no alteration in vessel density compared with vehicle-infused controls (losartan, −0.63±0.04%, n=4; PD 123319, −1.26±0.11%, n=4).
Fig 4⇓ shows a comparison of the structural alterations in the microvasculature among groups. Systemic infusion of Ang II at a low concentration for 3 days resulted in marked angiogenesis (center) compared with vehicle infusion (left). Selective blockade at the AT2 receptor produced an even greater angiogenic response (right).
This study suggests a role for the AT2 receptor in opposing the actions of the AT1 receptor. Evidence suggests that the AT1 receptor mediates vasoconstriction15 and vascular growth in the aorta and large arteries.8 Our results agree with these findings and show for the first time that Ang II induces angiogenesis in the microcirculation through the AT1 receptor. Our results also indicate that the AT2 receptor mediates vasodilation and inhibition of growth and tends to buffer the action of the more dominant AT1 receptor. The actions of the AT1 receptor may dominate in vivo because it may be the most prevalent angiotensin receptor in the vasculature. It has been shown to be expressed at a higher level than the AT2 receptor in the adult rat aorta16 ; however, the Ang II receptor populations in the microvasculature have not yet been characterized.
The current study showed that angiogenesis occurs after as few as 3 days of Ang II infusion and is the result of opposing forces of the AT1 and AT2 receptors. Previous studies have also shown a role for Ang II in microvascular proliferation.3 4 9 10 However, these studies were done using longer treatment periods and did not determine the receptor subtypes responsible for mediating this growth. LeNoble et al11 showed that CGP 42112, a then-putative AT2 antagonist, blocked Ang II–induced angiogenesis in the developing chick chorioallantoic membrane in a 7-day study. This compound is now considered to be a full AT2 agonist in many systems.17 If CGP 42112 was acting as an AT2 agonist in that study, then the results of LeNoble et al agree with ours, which suggest that the AT2 receptor has antiangiogenic activity. Unlike our findings, however, LeNoble et al did not observe any effect of losartan or PD 123319 on microvascular density. In receptor binding studies, the chick chorioallantoic membrane was shown to have a single class of receptor.11 CGP 42112 had moderate binding affinity for this class of receptor, whereas losartan and PD 123319 had extremely low affinity. This suggests that the avian receptor is markedly different from mammalian Ang II receptors, making determination of the relevance to mammalian systems difficult.
A result similar to our finding that stimulation of the AT2 receptor mediates a growth-inhibitory response in the microvasculature was shown in a recent study with cultured cells.18 That study showed that treatment of quiescent coronary artery endothelial cells with PD 123177 (an AT2 antagonist) and Ang II resulted in proliferation, whereas treatment with Ang II alone had no effect. This is further evidence that the stimulation of AT2 receptor acts as a brake to suppress cell growth during Ang II stimulation.
Another interesting result of the current study was the effect of the AT2 receptor on BP. Our results show that chronic infusion of PD 123319 causes an enhanced BP response to chronic but not acute Ang II infusion, suggesting that the AT2 receptor mediates vasodilation. This lack of an acute response of AT2 blockade as measured by bolus Ang II injections may explain the difference in results between this study and others aimed at illuminating the functions of the AT2 receptor. In a separate study, aortic rings stimulated with Ang II and PD 123319 produced a greater constriction than rings stimulated with Ang II alone.19 Another study showed a similar result with regard to the biphasic pressor response observed with bolus Ang III infusions.20 In that study, pretreatment of rats with PD 123319 resulted in an enhanced pressor response to Ang III, whereas pretreatment with losartan abolished the pressor response and exhibited an enhanced depressor response. Additional indirect evidence supporting our finding is that rat cerebral arteries, which express only the AT2 receptor subtype,21 vasodilate in response to Ang II stimulation.22 Although these findings do not rule out any role of central or renal effects of AT2 receptor inhibition on the alterations in BP we observed, they strongly suggest that Ang II causes vasodilation through stimulation of the AT2 receptor. In the present study, the effect of chronic Ang II and losartan coinfusion on BP was a losartan dose–dependent depressor response. The AT1 receptor is known to mediate vasoconstriction and so this response may be partly due to blockade of the AT1 receptor directly. However, since the Ang II dose being infused was already subpressor, this suggests that selective blockade of the AT1 receptor unmasked the vasodilator activity of the AT2 receptor, which resulted in an MAP below control (no Ang II infusion).
Losartan treatment resulted in a decrease in microvascular density compared with Ang II infusion alone; however, vessel density was still increased compared with control. In fact, we might have expected a decrease in vessel density to values lower than control because of blockade of the AT1-mediated angiogenic activity and unmasking of the AT2-mediated antiangiogenic activity. The reason for this result is not clear. Our bolus Ang II infusion data after 3 days of losartan infusion suggest complete blockade of the AT1 receptor. One confounding factor is the depressor response we saw in these groups. In the Ang II (10 ng/kg per minute) infusion group, an increased MAP appeared to inhibit the Ang II effect on vessel density. In previous studies, we have shown that in a renal hypertension model, the rise in pressure contributed to the reduction in vessel density.13 In a similar manner, a decrease in MAP such as was caused by losartan infusion may enhance the angiogenic effect. In fact, Hogan and Hirschmann23 found that a reduction in perfusion pressure of 30% in the rat cremaster muscle with normal flow induced arteriolar proliferation. It also is possible that losartan was spilling over and binding to the AT2 receptor as well as the AT1 receptor. This is unlikely because the bolus BP responses show a dose-dependent depressor response, indicating that the levels we are infusing are in the correct operating range for the AT1 receptor. Furthermore, the chronic BP data for the losartan- and PD 123319–infused groups exhibit differential effects by these antagonists, suggesting that the concentrations used are sufficient for selective receptor blockade. Experiments using both AT1 and AT2 receptor antagonists suggest no role for other receptor subtypes involved in the action of Ang II on vessel density or BP regulation in this model.
The effects of Ang II on BP and vessel density observed in this study required a very low Ang II concentration. In fact, Ang II levels in the plasma after infusion of Ang II alone were not significantly different from preinfusion levels. The increase in circulating Ang II was too small to be detected by our assay involving HPLC separation followed by radioimmunoassay yet was sufficient for detection by bioassay. The measured change of 3.4 pg/mL in the Ang II–infused group falls into the range of assay variability and so may be less sensitive than our bioassay. This phenomenon has been documented in other hormone systems as well.24 25 Because high salt intake results in suppression of the intrinsic renin-angiotensin system, Ang II receptor populations may have been upregulated and the system may be more responsive to low levels of Ang II. Also, the most dramatic responses to Ang II stimulation in both BP and vessel density occurred during selective receptor blockade. Because the respective actions of the two receptor types balance each other out to a great extent, blockade of one receptor was necessary to reveal the activity of the opposing receptor. Thus, a subpressor concentration of Ang II becomes pressor with AT2 receptor blockade and depressor with AT1 blockade.
The results from this short-term Ang II infusion study differ from those from an earlier study from this laboratory.10 Although we detected no alteration in MAP after Ang II infusion at 5 ng/kg per minute in the present study, we found Ang II infusion at a rate of 10 ng/kg per minute to be pressor. In contrast, Hernandez et al10 showed Ang II doses of both 5 and 10 ng/kg per minute to be subpressor. That study assessed BP and vessel density changes after 4 weeks of Ang II infusion compared with 3 days in the present study. It is likely that long-term BP regulation compensated for any elevation of BP caused by Ang II infusion in the 4-week study that was not apparent in our 3-day study. This inherent difference between the two studies helps to explain the difference in the angiogenic activity of Ang II observed. In both studies, the 5 ng/kg per minute dose was found to be angiogenic. In the 4-week study, the 10 ng/kg per minute dose was found to be angiogenic, whereas it was ineffective in stimulating vessel growth in the 3-day study. This may be due to the fact that the dose was pressor in the 3-day study, which may have counteracted the angiogenesis. This observation that elevated BP inhibits the angiogenic action of Ang II is consistent throughout the present study, with one exception. Ang II coinfused with PD 123319 (5 μg/kg per minute) caused an increase in both MAP and vessel density. Apparently, in the presence of AT2 receptor blockade, the angiogenic stimulus is so strong it overrides the inhibitory influence of elevated MAP. It is possible that if MAP were artificially held constant in the face of maximal AT2 receptor antagonism, the resulting angiogenesis would be even greater.
In the present study, short-term Ang II infusion resulted in an increase in microvascular density. This angiogenesis appeared to be mediated by the AT1 receptor, whereas the AT2 receptor mediated antiangiogenesis, buffering the full AT1 growth response when both receptors were activated. In addition, the AT1 receptor mediated an increase in MAP in this study, and the AT2 receptor mediated a depressor response that similarly acted to buffer the full AT1 pressor response when both receptors were activated. These results fit well with previous findings in large vessels and other cell and tissue types and offer new insight into the mechanisms that may be involved in microvascular remodeling. Potential differential regulation of these receptor subtypes in disease states such as diabetes or hypertension may prove to be a factor in the development of microvascular pathology.
Selected Abbreviations and Acronyms
|Ang II, III||=||angiotensin II, III|
|AT1, AT2||=||angiotensin receptor type 1, type 2|
|MAP||=||mean arterial pressure|
The authors would like to thank Jennifer Clark for expert technical assistance.
Daemen MJAP, Lombardi DM, Bosman FT, Schwartz SM. Angiotensin II induces smooth muscle cell proliferation in the normal and injured rat arterial wall. Circ Res. 1991;68:450-456.
Fernandez LA, Caride VJ, Twickler J, Galardy RE. Renin-angiotensin and the development of collateral circulation after renal ischemia. Am J Physiol. 1982;243:H869-H875.
Geisterfer AAT, Peach MJ, Owens GK. Angiotensin II induces hypertrophy, not hyperplasia, of cultured rat aortic smooth muscle cells. Circ Res. 1988;62:749-756.
Sadoshima J, Izumo S. Molecular characterization of angiotensin II-induced hypertrophy of cardiac myocytes and hyperplasia of cardiac fibroblasts: critical role of the AT1 receptor subtype. Circ Res. 1993;73:413-423.
Wang D-H, Prewitt RL. Captopril reduces aortic and microvascular growth in hypertensive and normotensive rats. Hypertension. 1990;15:68-77.
Hernandez I, Cowley AW Jr, Lombard JH, Greene AS. Salt intake and angiotensin II alter microvessel density in the cremaster muscle of normal rats. Am J Physiol. 1992;263:H8664-H8667.
LeNoble FAC, Schreurs N, van Straaten HWM, Slaaf DW, Smits JFM, Struyker Boudier HAJ. Evidence for a novel angiotensin II receptor involved in angiogenesis in chick embryo chorioallantoic membrane. Am J Physiol. 1993;264:R460-R464.
Nussberger J, Brunner DB, Waeber B, Brunner HR. True versus immunoreactive angiotensin II in human plasma. Hypertension. 1985;7(suppl I):I-1-I-7.
Stoll M, Steckelings UM, Paul M, Bottari SP, Metzger R, Unger T. The angiotensin AT2-receptor mediates inhibition of cell proliferation in coronary endothelial cells. J Clin Invest. 1995;95:651-657.
Munzenmaier DH, Greene AS. Angiotensin II receptor subtypes mediate opposing actions in the control of blood pressure and vascular tone. Hypertension. 1995;26:579. Abstract.
Scheuer DA, Perrone MH. Angiotensin type 2 receptors mediate depressor phase of biphasic response to angiotensin. Am J Physiol. 1993;264:R917-R923.
Tsutsumi K, Saavedra JM. Characterization of AT2 angiotensin II receptors in rat anterior cerebral arteries. Am J Physiol. 1991;261:H667-H670.
Haberl RL, Anneser F, Villringer A, Einhäupl KM. Angiotensin II induces endothelium-dependent vasodilation of rat cerebral arterioles. Am J Physiol. 1990;258:H1840-H1846.