Vascular microarray profiling in two models of hypertension identifies caveolin-1, Rgs2 and Rgs5 as antihypertensive targets

Hypertension is a complex condition with many contributory genetic and environmental factors [1], which have constrained identification of the genes, biochemical pathways and pathophysiological processes that accompany its development. Hypertension is a critical risk factor for stroke, renal failure and cardiovascular disease and has been associated with a range of well recognized complications, including congestive heart failure, left ventricular hypertrophy, and atherosclerosis, and with more subtle structural and functional changes to target organs and tissues occurring as a consequence of enhanced salt sensitivity [2], enhanced angiotensin II sensitivity [3] and structural modifications to the arterial wall [4, 5]. Genetic factors are thought to partly account for structural changes by directly influencing the quantity, density and organization of the extracellular matrix and also indirectly through association with risk factors such as age, gender, blood pressure, total and high-density lipoprotein cholesterol, smoking and diabetes [6].

A variety of both genetic, such as the spontaneously hypertensive rat (SHR), and nongenetic, such as glucocorticoid-induced hypertension, animal models of hypertension have been developed [7]. Administration of adrenocorticotropic hormone (ACTH) to rats provides a quick and reproducible means of inducing hypertension by stimulating the adrenal production of corticosterone [8]. ACTH-induced hypertension was originally thought to involve altered kidney function and loss of control of blood volume [9] but is now known to involve disruption of the nitric oxide synthase (NOS) pathway at various points, affecting NO bioavailability, oxidative stress and function of the endothelium [10, 11]. The genetically hypertensive SHR is a renin-angiotensin system-dependent model of hypertension that also exhibits oxidative stress and disrupted endothelial NOS (eNOS) activity [12, 13]. Such convergence in effectors holds the promise of some overlap in causative factors.

Hypertension is generally characterized by remodeling of arteries, such as thickening of the media and intima, reduced lumen diameter and reduced distensibility [14]. Endothelial dysfunction is also a characteristic of hypertension and alterations to endothelial morphology and density have been described [15–17]. Resistance vessels have a primary role in autoregulation of the blood supply to the vascular bed and changes to their structure and function are crucial to the onset of target organ damage [4]. Angiotensin II has a prominent role in affecting resistance vessels as a pro-inflammatory stimulator of signaling pathways promoting vascular growth, remodeling and oxidative stress [18, 19]. Both hypertrophic and eutrophic remodeling have been documented in arteries of SHR, perhaps related to the size of the vessel and intraluminal pressures experienced [20–23]. Whether similar changes to resistance vessels accompany ACTH-induced hypertension is not known. Microarray expression profiling offers a means of studying the modifications to global gene expression that accompany these events [24] and hence may provide novel insights into potential targets for common therapeutic intervention. The aim of the present study was therefore to compare gene expression and morphology of a resistance artery from adult normotensive Wistar-Kyoto (WKY) rats, SHR and ACTH-induced hypertensive rats; the latter two as representatives of renin-angiotensin-dependent and steroid-dependent hypertension, respectively. Although the majority of genetic changes differed between the two hypertensive models, changes in common were found to caveolin-1 expression and the regulators of G protein signalling, Rgs2 and Rgs5.

Comparative gene expression profiling of a resistance artery taken from rats with either renin-angiotensin-dependent or glucocorticoid-induced hypertension has identified over 800 genes showing significant changes. Statistical analysis of the microarrays showed consistency in gene expression amongst the replicates within each experimental group but evidence of substantial transcriptional differences between the two hypertensive models and the normotensive controls. The majority of these differences reflected differential vascular remodeling, as confirmed by our anatomical studies, as well as activation of specific intracellular pathways. Nevertheless, arteries from both models exhibited significantly increased expression of caveolin-1, which binds a number of important signalling molecules including endothelial nitric oxide synthase [27], and significantly decreased expression of Rgs2 and Rgs5, which terminate the signalling of the G proteins Gq and Gi/o [28]. Together these common effects would be expected to elevate blood pressure through augmentation of vasoconstrictor mechanisms.

Saphenous arteries from SHRs were characterised by hypertrophic remodeling, in contrast to the arteries from the ACTH rats in which no alterations were found in the number of smooth muscle cell layers. Similar changes have been described in other arteries of SHR [20, 22, 23], typical of remodeling of similarly sized resistance vessels during angiotensin II-induced hypertension [19]. The absence of hypertrophy in the saphenous artery of ACTH-treated rats suggests that inward eutrophic remodeling [14] may be more common to this form of hypertension, perhaps related to the lower blood pressures experienced in comparison with those of the SHR. In the endothelium of SHR arteries, there was a significant reduction in the area of endothelial cells, consistent with changes observed in other muscular arteries of SHR [29–31]. Surprisingly, the area of endothelial cells was increased during ACTH-treatment resulting in a decrease in overall endothelial cell numbers. Changes in gene expression were consistent with cell proliferation in angiotensin II-dependent hypertension but not in glucocorticoid-induced hypertension. Thus, expression of growth promoters was upregulated and growth inhibitors suppressed in SHR arteries, while effects on both growth promoters and inhibitors were more diverse in the ACTH arteries. A lack of change in myoendothelial gap junctions suggests that these structures are not involved in the remodeling processes in saphenous arteries in either form of hypertension.

A large number of other genetic differences in the arteries from the two hypertensive models emerged from the study. These related to lipid metabolism, mitochondrial activity, calcium handling, ion transport, the adenylate cyclase pathway and protein kinase C. Not surprisingly, saphenous arteries from SHR showed upregulation of genes consistent with renin-angiotensin hypertension, and angiotensin II signaling pathways through Raf and MEK1/2 [19]. Upregulation of growth promoters and down regulation of growth suppressors as found in the SHR samples would be expected to stimulate cell growth and proliferation, mobilise intracellular Ca²⁺ and reactive oxygen species [18, 32]. In contrast, the saphenous arteries from ACTH-treated animals showed upregulation of the p38MAP kinase pathway which is involved in inflammation and apoptosis, as well as growth and differentiation [33], accentuating mechanistic differences between the two models of hypertension. While alterations to the extracellular matrix were found in both hypertensive models, substantial changes involving multiple collagen genes were found predominantly in the ACTH model. Such changes have been linked with hypertension [34] and could impact on cytoskeletal signalling through integrin receptors.

Downstream mechanisms underlying smooth muscle contraction and calcium sensitization also varied between the two hypertensive models. Calcium/calmodulin-dependent protein kinase II (CaMkII) has an important role in regulating smooth muscle contractility by inhibiting myosin light chain kinase (MLCK) and the phosphorylation of myosin light chain [35]. In SHR, upregulation of CaMkII may shift smooth muscle contraction toward Rho-dependent mechanisms and decreased phosphatase activity. In contrast, in ACTH, CaMkII was downregulated suggesting that smooth muscle contraction could be more influenced by MLCK, in addition to Rho-dependent mechanisms.

In the face of these substantial differences in gene expression between the two hypertensive models, we found coordinate upregulation in gene expression of caveolin-1 and increased caveolin-1 protein in both endothelial and smooth muscle cells. Our ultrastructural observations of increased incidence of caveolae in these sites provides strong confirmation of the immunohistochemical findings since caveolae are absent in caveolin-1 knockout mice [36]. While some of the genetic changes in SHR relative to WKY might have arisen through decades of isolation from the parent WKY strain following inbreeding, coordinate changes in gene expression between the SHR and the ACTH-treated WKY rats cannot be attributed to such a cause. Caveolae are now recognized as important sites for colocalisation of signaling molecules, including receptor tyrosine kinases, G-protein-coupled receptors and G-proteins, and hence are critical for vascular function [37]. The inactivation of eNOS through binding to caveolin-1 is the best characterized function for caveolae in endothelial cells [37], highlighting their crucial role in vascular homeostasis. Alterations to caveolae in smooth muscle cells will also impact on cell signalling and recent studies of idiopathic pulmonary arterial hypertension in humans have concluded that increased smooth muscle expression of caveolin-1 contributes to the pathophysiology of the disease through increased capacitative calcium entry [38]. Whether upregulation of caveolin-1 in smooth muscle cells and in endothelial cells could serve to induce hypertension warrants further investigation.

In contrast to caveolin-1, the G protein regulators, Rgs2 and Rgs5, were coordinately downregulated in both SHR and ACTH saphenous arteries. These molecules terminate G protein signalling through Gq and Gi/o by activating the GTPase activity of the Gα subunit. Downregulation thus potentiates the effect of vasoconstrictors like endothelin and angiotensin II [39]. A role in blood pressure regulation has been demonstrated for Rgs2, since knockout mice are severely hypertensive [40]. Activation of Rgs2 occurs in resistance and conduit arteries through binding of protein kinase G, the downstream mediator of nitric oxide action [40, 41]. Interestingly, Rgs2 was also decreased in aortic smooth muscle during L-NAME-induced hypertension [42]. Upregulation of caveolin-1 and decreased eNOS function may therefore play a role in the downregulation of Rgs2.

Less is known about a role in blood pressure regulation for Rgs5, which has been described as a marker of arterial but not venous smooth muscle cells [43] and is downregulated during proliferation to form atherosclerotic plaques [44]. In contrast, upregulation of Rgs5 has been reported during neovascularisation of tumours [45]. Rgs5 has also been identified in endothelial cells of cerebral capillaries where it was downregulated in stroke-prone SHR compared to SHR [46]. Our results demonstrating downregulation in arteries from both SHR and ACTH rats would suggest a role in hypertension rather than proliferation since smooth muscle hypertrophy was not a feature of the ACTH model.

Study of changes in gene expression in two different models of hypertension have highlighted numerous differences which can be linked to variation in the form of vascular remodeling and the unique signalling pathways affecting contractility and cell growth. In contrast, we have identified common changes to upstream targets crucial for the alteration in the balance between vasoconstriction and vasodilation as occurs in hypertension. Common increases in expression of caveolin-1 are likely to increase vasoconstriction through direct effects on calcium influx in smooth muscle cells and indirect inhibitory effects on eNOS in endothelial cells. Decreases in Rgs2 and Rgs5 will augment these effects by prolonging the actions of Gq and Gi/o and their downstream vasoconstrictor pathways. Further studies examining the functional implications of these data and aimed at reversing these gene changes during the different forms of hypertension are needed to determine whether these molecules could be considered as therapeutic drug targets.

Sixteen-week-old male WKY and age-matched SHR were used. These rats were originally from stock derived from the National Institutes of Health, Bethesda, MD. A group of 12-week-old male WKY rats was administered ACTH (0.2 mg/kg/day) subcutaneously for 4 weeks. Systolic blood pressure was measured using tail-cuff plethysmography. All experiments were approved by the Australian National University Animal Experimentation Ethics Committee under guidelines of the National Health and Medical Research Council of Australia.

All analyses were made on the first lateral branch of the saphenous artery. This vessel was defined as a resistance vessel due to its small diameter (<250 μm), since feed arteries and the microcirculation contribute to blood pressure regulation in rats [47].