Angiotensin II augments renal vascular smooth muscle sGC expression via an AT₁R - FoxO transcription factor signaling axis

Abstract

Background and Purpose:

Reduced renal blood flow triggers activation of the renin-angiotensin-aldosterone system (RAAS) leading to renovascular hypertension. Renal vascular smooth muscle expression of the nitric oxide (NO) receptor, soluble guanylyl cyclase (sGC), modulates the vasodilatory response needed to control renal vascular tone and blood flow. Here, we tested if angiotensin II (Ang II) impacts sGC expression via an Ang II type 1 receptor (AT₁R)-forkhead box subclass O (FoxO) transcription factor dependent mechanism.

Experimental Approach:

Using a murine two-kidney-one-clip (2K1C) renovascular hypertension model, we measured renal artery vasodilatory function and sGC expression. Additionally, we conducted cell culture studies using rat renal pre-glomerular smooth muscle cells (RPGSMCs) to test the in vitro mechanistic effects of Ang II treatment on sGC expression and downstream function.

Key Results:

Contralateral, unclipped renal arteries in 2K1C mice showed increased NO-dependent vasorelaxation compared to sham control mice. Immunofluorescence studies revealed increased sGC protein expression in 2K1C contralateral renal arteries over sham controls. RPGSMCs treated with Ang II caused a significant upregulation of sGC mRNA and protein expression as well as downstream sGC-dependent signaling. Ang II signaling effects on sGC expression occurred through an AT₁R and FoxO transcription factor-dependent mechanism at both the mRNA and protein expression levels.

Conclusion and Implications:

Renal artery smooth muscle, in vivo and in vitro, upregulate expression of sGC following RAAS activity. In both cases, upregulation of sGC leads to elevated downstream cGMP signaling, suggesting a previously unrecognized protective mechanism to improve renal blood flow in the uninjured contralateral renal artery.

Graphical Abstract:

Renal artery stenosis causes elevated circulating Angiotensin II. This, in turn, leads to contraction and hypertrophy of smooth muscle and elevated blood pressure. Additionally, soluble guanylyl cyclase (sGC) expression and cGMP in the non-stenotic renal artery increases allowing for increased blood flow to the healthy kidney.

INTRODUCTION:

Renovascular hypertensive patients constitute 24.2% of all patients with drug resistant hypertension (Benjamin, et al., 2014), a condition characterized by uncontrolled hypertension despite treatment with three or more adequately dosed anti-hypertensive therapies (Carey, et al., 2019). While the prevalence in the general population is low (1–2%) (Derkx and Schalekamp, 1994), renovascular hypertension is more common in elderly patients over age 65 (6.8%) and is present in nearly 40% of individuals with established peripheral or coronary artery disease (Goldfarb, 2003), (Iglesias, et al., 2000). Approximately 90% of renovascular hypertension stems from atherosclerotic renal artery stenosis (ARAS) (Sawicki, et al., 1991), (Tollefson and Ernst, 1991). ARAS leads to obstruction of renal artery blood flow, resulting in renin-angiotensin-aldosterone-system (RAAS) activation and subsequent elevation of circulating blood plasma angiotensin II (Ang II) (Goldblatt, et al., 1934). In response, the non-stenosed renal artery is subjected to increased blood flow leading to augmented sodium and water excretion by the kidney. This process, known as pressure natriuresis, helps to mitigate increased fluid retention, volume overload, and systemic blood pressure (Selkurt, 1951).

A main contributor to pressure natriuresis is endothelial-derived nitric oxide (NO), which has been shown to play a critical role in the dilation of the renal vasculature (Dautzenberg, et al., 2011), (Majid and Navar 2001), (Majid, et al., 1998), (O’Connor and Cowley 2010). NO diffuses to vascular smooth muscle cells (VSMCs) where it binds its cognate receptor, soluble guanylyl cyclase (sGC), which produces cGMP to elicit vasorelaxation (Arnold, et al., 1977), (Furchgott and Zawadzki, 1980), (Ignarro, et al., 1987). Of clinical importance, sGC modulating compounds, which enhance cGMP production, have been approved for treatment of pathologies such as pulmonary arterial hypertension and heart failure with reduced ejection fraction (Ghofrani, et al., 2013), (Armstrong, et al., 2020), and many are currently under investigation for treatment of renal and cardiovascular diseases (Stasch, et al., 2015). In addition, we have recently shown that basal sGC expression is regulated by the forkhead box subclass O (FoxO) transcription factors in aortic VSMCs (Galley, et al., 2019).

Based on this evidence, we hypothesized that renal artery smooth muscle responds to elevated RAAS signaling with amplified sGC-mediated production of cGMP. In this study, we used a two-kidney-one-clip (2K1C) hypertension model, wherein blood flow to one renal artery is reduced (Goldblatt, et al., 1934), as a model of RAAS activation and renal hypertension. We find that renal smooth muscle responds to increased levels of Ang II by increasing the expression of sGC. This increased sGC expression occurs in an Ang II type 1 receptor (AT₁R) and FoxO transcription factor-dependent manner. Consequently, this results in enhanced downstream cGMP signaling and increased smooth muscle relaxation. These studies are the first to show that exposure of renal smooth muscle to elevated Ang II results in a protective mechanism whereby sGC expression is increased, leading to elevated cGMP production and vasorelaxation.

METHODS:

Statement on Ethical Use of Animals:

Rodent models were used in this study, as they have physiological mechanisms similar to the mechanisms of human blood pressure control. Mechanistic studies to understand the changes in multi-organ signaling and physiology were then studied in more detail at the molecular level in cell culture studies. These two main approaches complement one another and reduce the number of animals used. All animals were used in ethical compliance with the University of Pittsburgh’s Institutional Animal Care and Use Committee (Protocol #IS00016317 and IS00015180). For all experiments, a priori power analyses were performed to determine the number of animals necessary to determine whether differences observed were significant. For ex vivo procedures, blinding occurred post tissue harvesting and kidney removal so that the identity of clip vs. sham groups were unknown for myography to avoid bias.

2K1C Renal Stenosis Model:

Renovascular hypertension through a 2K1C model was produced as described previously (Ong, et al., 2019), (DeLalio, et al., 2020). C57B6/J male mice (Jackson Laboratories) were fed a 0.1% NaCl diet (D17020) for 1 week prior to 2K1C surgery and thereafter. Mice were anesthetized with 2–3% isoflurane in 100% O₂. Through a retroperitoneal incision, the right renal artery was carefully isolated from the renal nerve. A 0.5-millimeter polytetrafluoroethylene catheter (ID: 0.008 X OD: 0.014; Braintree Scientific, SUBL140) was cut longitudinally and placed around the renal artery, distal to the adrenal artery. The catheter was then secured in place with two 10–0 sutures to attenuate blood flow and induce renal stenosis. The mice that received this procedure are the renal clip group of mice. For sham control procedures, renal arteries were isolated from the renal nerve, but a catheter was not placed. Post-surgery, animals were treated by subcutaneous injection of 0.03 mg/kg buprenorphine twice/day for 48 hours (Henry Schein Inc.) and 2 mg/kg enrofloxacin (Norbrook Laboratories) as previously published. Mice were then sacrificed 19–21 days following renal clip or sham surgery. Vascular reactivity for the aorta and mesenteric arteries using these mice was reported previously (DeLalio, et al., 2020). The renal arteries from these same animals are analyzed herein. A separate, second cohort of animals using this model (n=6 for each group) were used for immunohistochemical analyses in this study.

Animal harvesting for immunohistochemical analysis:

Male C57BL6/J mice with or without renal stenosis were sacrificed via CO₂ asphyxiation followed by cervical dislocation. The artery contralateral to renal clip (left renal artery) was excised and placed in 4% paraformaldehyde in PBS for 24 hours then placed in 100% ethanol for processing (International Consortium for Blood Pressure Genome-Wide Association et al., 2011). Tissues were embedded in paraffin and sectioned at 8 micron thickness. Immunohistochemical analysis was performed as previously described (Durgin, et al., 2019). Tissue sections were deparaffinized with xylenes and rehydrated by sequentially decreased concentrations of ethanol (100%−70%) followed by deionized distilled water. Heat-mediated antigen retrieval was then performed using citric acid-buffer (Vector Laboratories, H-3300) for 20 minutes, then sections cooled for 30 minutes at 4°C. Sections were then blocked in 10% horse serum (Sigma H1138) in PBS (MilliporeSigma, H1270) at room temperature for 1 hour. For Figure 2, primary antibodies (See Table 2) for sGCβ (Abcam, ab154841, 1:100) and von Willebrand Factor (vWF; Abcam, ab11713, 1:250) were incubated on sections in PBS containing 10% horse serum overnight at 4°C in a humidity chamber. One section per slide was stained with rabbit (Vector Laboratories, I-1000) IgG control to match the corresponding sGCβ antibody concentration. Tissue sections were washed thrice for 5 minutes with PBS. For Supplementary Figure 1, incubation for sGCα (Cayman, 160895, 1:80) and corresponding IgG control were performed on tissue sections in 10% horse serum in PBS overnight at 4°C in a humidity chamber. Tissue sections were then washed thrice for 5 minutes in PBS and then incubated for 1 hour at room temperature using PBS containing 10% horse serum and vWF primary antibody. These samples were washed thrice again in PBS for 5 minutes before the next steps. For all tissues used in Figure 2 and Supplementary Figure 1, sections were then incubated in PBS containing 10% horse serum with smooth muscle α-actin (ACTA2) primary antibody pre-conjugated to FITC fluorophore (MilliporeSigma, F3777 clone 1A4, 1:500), 4′,6-diamidino-2-phenylindole (DAPI, D3571, Thermo Fisher Scientific, 1:100) and secondary antibodies (See Table 2) donkey anti-rabbit AlexaFluor 594 (Invitrogen, A-21207, 1:250) and donkey anti-sheep AlexaFluor 647 (Invitrogen, A-21447, 1:250) for 1 hour at room temperature in a humidity chamber. Tissue sections were then washed thrice in PBS for 5 minutes before being mounted on coverslips using Prolong Gold Antifade mounting medium with DAPI reagent (Invitrogen, P36931). Immunohistochemistry staining of renal arteries were imaged using a Nikon A1 Confocal Laser Microscope at the University of Pittsburgh Center for Biological Imaging. Images were taken with 40X objective magnification with 1024 × 1024 pixel resolution. Increments for Z-stacks of 1 μm were applied for stained and IgG controls. In ImageJ, a region of interest was drawn around ACTA2+ areas representing the smooth muscle cell tunica media then superimposed on sGCβ images for quantification of medial smooth muscle sGCα and sGCβ expression per medial area.

Table 2: Catalogue of primary and secondary antibodies.

Soluble guanylyl cyclase α-subunit (sGCα), soluble guanylyl cyclase β-subunit (sGCβ), smooth muscle α-actin (ACTA2), vasodilator-stimulated protein (VASP) antibodies which were used for western blots (WB), immunohistochemistry (IHC), and/or immunocytochemistry (ICC).

Antibody	Species	Application	Concentration	Company	Cat. Number
sGCα	rabbit	WB, IHC	1:200, 1:80	Cayman	160895
sGCβ	rabbit	WB, ICC,	1:1000, 1:200,	Cayman	160897
sGCβ	rabbit	IHC	1:100	Abcam	ab154841
β-actin	mouse	WB	1:500	Santa Cruz	sc-47778
α-tubulin	mouse	WB	1:10,000	Sigma	T6074
Anti-rabbit Alexafluor-594	donkey	IHC, ICC	1:250, 1:250	Life Technologies	A21207
Anti-goat Alexafluor-647	donkey	IHC	1:250	Life Technologies	A21447
ACTA2 conjugated Alexafluor-488	mouse	IHC	1:500	Sigma	F3777
Von Willebrand Factor	sheep	IHC	1:250	Abcam	ab11713
AlexaFluor 488 Phalloidin	N/A	ICC	1:100	ThermoFisher	A12379
Phospho-VASP (S239)	rabbit	WB	1:1000	Cell Signaling	3114S
VASP	rabbit	WB	1:500	Proteintech	13472–1-AP
Rabbit IgG	rabbit	IHC	concentration matched	Vector Laboratories	I-1000

Treatment of Renal Artery Rings and Myography:

The following treatment method was performed as previously described (Durgin, et al., 2019). In brief, murine renal arteries were rapidly cleaned, excised and placed in room temperature physiological salt solution (PSS) which contains: 119 mM NaCl, 4.7 mM KCl, 1.17mM MgSO₄, 1.18 mM KH₂PO4, 5.5 mM D-glucose, 25 mM NaHCO₃, 0.027 mM EDTA, and 2.5 mM CaCl₂. Arteries were cut into 2-millimeter rings, then placed on a small vessel wire myograph (DMT 620M) filled with PSS (pH 7.4 when bubbled with 95% O₂ 5% CO₂ at 37°). Following a 30 minute rest, arteries were gradually stretched to a tension corresponding to a transmural pressure of 80mmHg. Arteries were then constricted with a dose response of phenylephrine (50nM-50mM). Blood vessels that failed to constrict in response to phenylephrine were excluded from the experiments on the grounds that they could not produce any contractile response. Rings were washed 3 times with PSS and allowed to rest for 30 minutes. A final wash was performed, and arteries were rested for an additional 10 minutes. Following the final 10 minute rest period, arteries were constricted with a single dose of phenylephrine (1 mM) wherein all vessels reached at least 50% of maximal contraction. After reaching a plateau, increasing concentrations of acetylcholine (ACh; 10 nM - 1 μM, Sigma, MA6625) or sodium nitroprusside (SNP; 10 nM - 10 μM, Sigma, 71778) were administered to assess endothelium-dependent and NO-dependent relaxation, respectively. Subsequently, 100 μM SNP in Ca²⁺-free PSS was added to the vessels to determine their maximal dilatory responses. The percentage relaxation reported represents the data normalized to the difference between maximal dilation and maximal constriction. For the cohort of animals where renal arteries were treated with ACh vasodilator, n=7 animals for sham surgery and n=8 animals for 2K1C (renal clip) surgery. For the cohort of animals where renal arteries were treated with SNP vasodilator, n=5 animals for sham surgery and n=8 animals for 2K1C (renal clip) surgery.

Cell culture, drug, and peptide treatments:

Renal pre-glomerular smooth muscle cells (RPGSMCs) were isolated from Wistar-Kyoto rats as previously described (Zhu and Jackson, 2017) and cultured at 37°C in SmGm-2 fully supplemented growth medium (Lonza, CC-3181) containing 5% FBS and SmGm-2 SingleQuot (Lonza, CC-3182) reagents and passaged using 1X trypsin–EDTA (Gibco, 10779413) dissolved in 1X PBS. RPGSMCs were used between passages 2–7 for all experiments, after which they were discarded. Cells were cultured to approximately 90% confluency (approximately 48–72 hours) prior to any drug treatment. During drug treatments, RPGSMCs were washed twice with 1X PBS and cultured in serum and growth factor starved Dulbecco’s Modified Eagle Medium/Ham’s F12 (DMEM/F12, Sigma, D6421) media containing: 100 U/mL penicillin/streptomycin (Gibco, 15140-122), 1.6 mM L-glutamine (Gibco, 25030-081), 200 μM L-ascorbic acid (Sigma, 50-81-7), 5 μg/mL apo-transferrin (Sigma, 11096-37-0), and 6.25 ng/mL sodium-selenite (Sigma, 10102-18-8). Losartan (Cayman, 124750-99-8), PD123319 (Sigma, 136676-91-0), and AS18428456 FoxO inhibitor (Cayman, A15871) were dissolved in dimethyl sulfoxide (DMSO, D8418), while Angiotensin II (Ang II, Sigma, A9525) peptide was dissolved in sterile deionized distilled water for stock solutions prior to treatment. Treatment concentrations were 100 nM Losartan, 100 nM PD123319, 1 μM AS1842856, and 1 μM Ang II. Control treatments involved 0.1% DMSO treatment for 48 hours prior to harvesting. For NO stimulation experiments, cells were pretreated with 10 μM sildenafil citrate (Sigma, PZ0003) for 45 minutes to inhibit cGMP-specific phosphodiesterase 5 activity, and then stimulated with the NO-donor, diethylammonium (Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate (DEA-NONOate, Cayman, 82100) for 15 minutes.

qRT-PCR:

RPGSMCs were cultured in 6-well culture plates until approximately 90% confluent before being washed and switched to serum and growth factor starved media. Cells were then subjected to 48-hour drug and/or peptide treatment before lysis in TRIzol reagent (ThermoFisher, 15596026). The Direct-zol RNA miniprep plus (Zymo, R2051) manufacturer’s protocol was used to isolate RNA from cells. For cDNA synthesis, the SuperScript IV First Strand Synthesis (ThermoFisher, 18091050) kit manufacturer’s protocol was used. For quantitative real time PCR analysis, the PowerUp SYBR Green Master Mix (ThermoFisher, A25742) and 1 μM target primer (Table 3) were mixed according to manufacturer’s protocol with settings for 40 PCR cycles, 95°C melting temperature, 58°C annealing temperature, and 72°C extension temperature set on a QuantStudio 5 Real-Time 384-well PCR System (ThermoFisher A28140) for amplification. The Δ-Δ-ct value fold change in expression was used in order to control for cell number and RNA quality with values normalized to an 18S housekeeping gene transcript.

Table 3: Catalogue of PCR primer sequences used.

Primers for soluble guanylyl cyclase α-subunit type 1 (sGCα₁), soluble guanylyl cyclase β-subunit type 1 (sGCβ₁), 18S small subunit ribosomal RNA (18S), cGMP-specific phosphodiesterase type 5 (PDE5), cGMP-activated protein kinase type 1 (PGK1) and glucose-6-phosphatase (G6Pase), listed here were used in our analyses. For primers, F denotes forward primer and R denotes reverse primer with sequences reported in 5’ – 3’ direction.

Primer	Sequence (5’→3’)
Rat sGCα1 F	CTC CCG TGA CCG CAT CAT
Rat sGCα1 R	CCG GTG TTG ATG TTG ACT GA
Rat sGCβ1 F	AAT TAC GGT CCC GAG GTG TG
Rat sGCβ1 R	GCA GCA GCC ACC AAG TCA TA
Mammalian 18S F	ACG GAC AGG ATT GAC AGA TTG
Mammalian 18S R	TTA GCA TGC CAG AGT CTC GTT
Rat PDE5 F	GCC GAT CTG GGC TGA ACT AAC
Rat PDE5 R	GCT CAC GGT TCC CTC AGA AT
Rat PKG1 F	ATGAGCGAACTGGAGGAAGAC
Rat PKG1 R	GTCGATCAATGGCCCAGAGT
Rat G6Pase F	GGC TCA CTT TCC CCA TCA GG
Rat G6Pase R	ATC CAA GTG CGA AAC CAA ACA G

Western blot:

RPGSMCs were cultured in 12-well culture dishes until approximately 90% confluent before being switched to serum and growth factor starved media for 48 hours. Cells were then washed with PBS and 1X Cell Lysis Buffer (Cell Signaling, 9803) containing: (pH 7.5) 20 mM Tris-HCl, 1 mM Na₂EDTA, 1 mM EGTA, 1% Triton, 1 mM β-glycerophosphate, 1 mM Na₃VO₄, 1 μg/mL leupeptin, 2.5 mM sodium pyrophosphate, and additional 1X protease (MilliporeSigma, P8340) and phosphatase inhibitors (MilliporeSigma, P5726) at 4°C. A bicinchoninic acid kit (ThermoFisher, 23225) was used according to the manufacturer’s protocol to quantify lysate protein concentration and approximately 15 μg of protein was used for each western blot lane. Lysates were boiled at 100°C for 10 minutes and Laemmli buffer was added such that final lysates contained: (pH 6.8) 31.5 mM Tris-HCl, 10% glycerol, 1% SDS, 2.5% β-mercaptoethanol and 0.005% Bromophenol Blue before being loaded onto 4–12% gradient BisTris polyacrylamide gels (Invitrogen Life Technologies, NP0335BOX). Proteins were transferred from polyacrylamide gels to nitrocellulose membranes (LiCor, 926–31092) and blocked for approximately 30 minutes at room temperature with 1% bovine serum albumin (BSA) in PBS. Membranes were incubated in primary antibody (Table 2) solution containing 1% BSA in PBS with 0.1% Tween 20 overnight at 4°C. Blots were then washed with 1X PBS with 0.05% Tween 20. Membranes were then incubated for one hour at room temperature in secondary antibody solutions containing 1:1 LiCor Intercept Buffer (LiCor, 927–70001): 1X PBS with 0.2% Tween 20 and corresponding secondary antibodies (Table 2). An Odyssey CLx Imager (LiCor, 9140) was used for fluorescence visualization and semi-quantitative analysis was performed using Image Studio software.

Immunocytochemical analysis of hypertrophy:

RPGSMCs were cultured on a single 24 × 50 millimeter cover glass (VWR, 16004–322) to approximately 90% confluency and then serum starved for 48 hours in the DMEM/F12 starvation media described above. Media was then gently aspirated by hand and the cover glass washed with PBS containing 0.1% Triton X-100 for 5 minutes. Cells were then fixed in 4% paraformaldehyde in PBS for 30 minutes at room temperature and then gently washed twice with PBS containing 0.1% Triton X-100. RPGSMCs blocking was carried out in PBS with 10% horse serum for 1 hour at room temperature. Primary antibody incubation for rabbit-sGCβ (Cayman, 160897, 1:200) antibody (See Table 2) was carried out in PBS with 10% horse serum overnight at 4°C in a humidified chamber. Cells were then gently washed thrice with PBS for 5 minutes each time before being incubated in donkey anti-rabbit (AlexaFluor, A21207, 1:250) secondary antibody (See Table 2) for 1 hour at room temperature in PBS containing 10% horse serum. PBS was used to wash cells twice for 5 minutes and AlexaFluor 488-conjugated phalloidin (ThermoFisher, A12379, 1:100) in PBS with 10% horse serum was added for 20 minutes to stain for filamentous (F)-actin. Cells were gently washed twice for 5 minutes in PBS before applying Prolong Gold Antifade mounting medium with DAPI reagent (Invitrogen, P36931) with coverslips applied to stained cells. Immunocytochemistry images of RPGSMCs were taken using a Leica DM1000 microscope at 40X objective with 2X zoom applied and cell area was quantified using ImageJ software.

Statistics:

Statistical analyses were performed using Graphpad Prism Software 8.0d. For wire myography, two-way analysis of variance (ANOVA) and an unpaired two-tailed t-test for each treatment concentration. This allowed for determination of significance between groups overall and at specific treatment concentrations. For Figure 1, listed (numerical) P-values represent significance by two-way ANOVA test, while * is indicative of a P < 0.05 by unpaired two-tailed t-test to assess differences between groups at each vasodilator concentration. Data normality was assessed via Shapiro-Wilk test. The P-values reported for the qPCR, Western blot, immunostaining analysis, and calculated EC₅₀ and E_max values were assessed using an unpaired two-tailed t-test for data that was normally distributed, an unpaired two-tailed t-test with Welch’s correction to account for data normally distributed but with unequal variance, or Mann-Whitney U-test was applied when data was non-normal and/or bimodally distributed. Symbols were consistent throughout wherein * denotes P < 0.05.

RESULTS:

To determine the effects of RAAS activation on the renal vasculature, we used a two-kidney-one-clip (2K1C) model (Figure 1A), which involves placing a surgical cuff around a single renal artery which clips the artery to restrict blood flow and induce unilateral renal stenosis. To assess the vasoreactivity responses of the 2K1C unclipped contralateral renal arteries compared to their sham counterparts, wire myography was performed on the renal arteries contralateral to the surgical procedures of clipped and sham animals to compare the two groups. We observed that the 2K1C contralateral arteries contracted significantly less to 1 mM phenylephrine (PE) than their sham counterparts (Figure 1B). We also observed an increase in endothelium-dependent vasodilation in response to 10⁻⁷ M acetylcholine (ACh) treatment (Figure 1C). A significant difference in NO-dependent vasodilation in response to the NO-donor, sodium nitroprusside (SNP), was observed between sham and clip groups with significant increases in vasorelaxation in the clip groups following 10⁻⁷ M and 10⁻⁶ M SNP treatment (Figure 1D). E_max and EC₅₀ values for ACh and SNP responses for sham control and renal clip groups are listed in Table 1. Together, the findings of the responses to both ACh and SNP indicate an improvement in the vasorelaxation responses of the contralateral arteries from 2K1C animals over sham controls.

Figure 1: 2K1C mouse model contralateral renal arteries have improved vasodilation compared to sham controls.

A) Schematic of 2K1C renovascular hypertension model and experimental design showing wire myography protocol. Black circles represent sham animals while red bars and squares represent contralateral renal arteries from 2K1C (clip) animals. B) Maximal constriction of sham (white bars) or clip (red bars) animal contralateral renal arteries following treatment with 10⁻³ M phenylephrine (PE). The * indicates P < 0.05 by unpaired two-tailed t-test. C) Vasodilatory responses to acetylcholine (ACh; 10⁻⁸ −10⁻⁶ M) in isolated contralateral renal arteries from mice subjected to 2K1C (n=8) or sham (n=7) surgery. D) Vasodilatory responses to sodium nitroprusside (SNP; 10⁻⁸ −10⁻⁵ M) in isolated contralateral renal arteries from mice subjected to 2K1C (n=8) or sham surgeries (n=5). Numerical P-values represent statistical differences between contralateral renal artery response curves from 2K1C versus sham animals by two-way ANOVA with the * indicative of P < 0.05 by unpaired two-tailed t-test at individual doses between groups. Error bars are ± SEM.

Table 1: Calculations for vasodilatory responses of renal arteries to ACh and SNP.

Vasodilator drugs acetylcholine (ACh) and sodium nitroprusside (SNP) were used to induce vasorelaxation of contralateral renal arteries for animals receiving 2K1C (renal clip) or sham surgery. Effective maximum dilation (E_max), standard error of the mean (SEM), and effective concentration to produce 50% of maximal response (EC₅₀) are shown above. P-values are calculated from unpaired t-test or unpaired t-test with Welch’s correction to account for unequal variances.

Vasodilator Drug	Treatment Group	Emax (%)	SEM (%)	P-value	EC50 (M)	SEM (M)	P-value
ACh	Sham	66.1	4.06	0.2951	1.04×10−7	±3.94×10−8	0.2881
	Clip	73.4	5.37		5.63×10−8	±1.45×10−8
SNP	Sham	44.2	6.16	0.0592	7.77×10−7	±3.12×10−7	0.1943
	Clip	63.0	5.90		2.70×10−7	±1.52×10−7

We then quantified renal artery protein expression to explore the causes of the increased NO-sensitivity of 2K1C contralateral renal arteries (Figure 2A–E; Supplementary Figure 1A–F). No significant changes in vessel nuclei staining (Figure 2A,F; Supplementary Figure 1A,F), smooth muscle α-actin (ACTA2; Figure 2B,G; Supplementary Figure 1B,G), or the endothelial marker, Von Willebrand Factor (vWF; Figure 2D,I; Supplementary Figure 1D,I) expression were observed. We did, however, observe a significant increase in sGCβ protein expression (Figure 2C,H) and sGCα protein expression (Supplementary Figure 1C,H) in the contralateral arteries of 2K1C animals over controls. In addition, no changes were observed in medial area (Figure 2J; Supplementary Figure 1J). These data indicate the increases in sGC expression in contralateral renal arteries compared to sham controls likely drive the increased NO-sensitive vasodilation response.

Figure 2: Immunohistochemistry of contralateral renal arteries from 2K1C (renal clip) animals show an increase in sGCβ expression over sham controls.

Immunofluorescent staining of representative maximum intensity projection images and quantification of sham control (n=6, white bars) or renal clip (n=6, red bars) animals for A,F) DAPI (nuclei, blue), B,G) ACTA2 (vascular smooth muscle, green), C,H) sGCβ (red) and D, I) vWF (endothelium, grey). E) Merge image from all channels. J) Quantification of medial area (μm²). The * indicates a P < 0.05 statistically significant differences between renal clip and sham groups by unpaired two-tailed t-test. Error bars are ± SEM.

Next, we sought to determine what drives sGC expression changes in renal artery smooth muscle. It is well established that reduced renal blood flow increases angiotensin II (Ang II) in models of 2K1C (Murphy, et al., 1984),(Sadjadi, et al., 2002). Therefore, we treated rat renal pre-glomerular smooth muscle cells (RPGSMCs) with vehicle or 10⁻⁶ M Ang II to test if Ang II increased sGC expression (Figure 3A–D). Ang II treatment led to no differences in cell number (Figure 3A,E) but increased filamentous (F)-actin expression (Figure 3B, F), and cell area (Figure 3D,H) indicating RPGSMC are hypertrophic (Stephenson, et al., 1998). Additionally, Ang-II resulted in a 1.7 fold increase in sGCβ protein expression via immunofluorescence (Figure 3B,F). We also observed an increase in sGCα by 1.9-fold (Supplementary Figure 2A,C) and a 4-fold increase in sGCβ protein expression via western blot analyses (Figure 3I, J). To test if increased sGC expression impacted cGMP production and PKG activity, RPGSMCs were treated with Ang II or vehicle and then subjected to treatment with the NO donor, DEA-NONOate, for 15 minutes prior to harvest to induce sGC-mediated cGMP production. Quantification of vasodilator stimulated protein (VASP) phosphorylated at the serine 239 position, a surrogate indicator of cGMP-dependent protein kinase activity (Smolenski, et al., 1998), showed an 8-fold increase in VASP serine 239 phosphorylation (pVASP) in Ang II-treated cells stimulated with DEA-NONOate compared to vehicle controls (Figure 3I,K). Taken together, these data show that Ang II in vitro augments sGC expression and cGMP signaling, indicating that elevated RAAS activity increases sGC expression and downstream signaling in vivo.

Figure 3: Immunocytochemistry of Ang II treated renal pre-glomerular smooth muscle cells show an increase in F-actin, sGC protein expression, and VASP phosphorylation compared to control-treated cells.

Immunocytochemistry staining and quantification resulting from control (n=10, white) or Ang II (10⁻⁶ M, n=10, light blue) treatment for A, E) DAPI (blue, nuclei), B, F) F-actin (green), C, G) sGCβ (red), D) Merged image of channels, or H) area/cell analyzed. I-K) Image and quantification of sGCβ band density/β-actin density, phosphorylated serine 239 vasodilator stimulated protein (pVASP) band density/β-actin density from Western blot for control (n=5 samples, white) or Ang II (10⁻⁶ M, n=5 samples, light blue)-treated RPGSMCs following DEA-NONOate treatment (10⁻⁶ M). NS denotes a non-specific band. The * indicates a P < 0.05 statistically significant difference between control and Ang II-treated cells by unpaired two-tailed t-test. Error bars are ± SEM.

We next tested whether the angiotensin II type 1 (AT₁R) or type 2 receptor (AT₂R) was responsible for increasing sGC levels. Consistent with increased sGC protein expression, we found that Ang II treatment also caused a 3.6-fold increase in sGCα mRNA (Figure 4A) and a 4.4-fold increase in sGCβ mRNA (Figure 4B). RPGSMCs co-treated Ang II and Losartan, an AT₁R antagonist (Timmermans, et al., 1995), showed inhibition of Ang II – induced increases in sGCα mRNA (Figure 4A), sGCβ mRNA (Figure 4B), sGCα protein expression (Supplementary Figure 2A,C) and sGCβ protein expression (Figure 4C, D). Conversely, RPGSMCs co-treated with Ang II and AT₂R antagonist PD123319 (Blankley, et al., 1991) showed no significant impact on the Ang II-induced increases in sGCα mRNA (Figure 4A), sGCβ mRNA (Figure 4B), sGCα protein expression (Supplementary Figure 2A,C) or sGCβ protein expression (Figure 4C, D). Combined, these data indicate that the Ang II mediated increase in sGC expression occurs through activation of AT₁R.

Figure 4: Losartan, but not PD123319, prevents Ang II-mediated increase of sGC expression in renal smooth muscle cells.

RPGSMCs were subjected to treatment with control (DMSO), Ang II (10⁻⁶ M), Losartan (10⁻⁷ M), and/or PD123319 (10⁻⁷ M). A) sGCα₁ mRNA (n=6 samples per group), B) sGCβ₁ mRNA (n=6 samples per group) and C, D) sGCβ protein expression (n=5 samples per group) was quantified relative to α-tubulin expression. NS denotes a non-specific band. The * indicates a P < 0.05 significant statistical difference between identified sample groups by unpaired two-tailed t-test. Error bars are ± SEM.

Recently, we published evidence that the FoxO family of transcription factors regulate the mRNA expression of sGC in aortic smooth muscle cells (Galley, et al., 2019). To determine whether the FoxO family of transcription factors can also influence the function of renal smooth muscle, we treated RPGSMCs with 10⁻⁶ M of the FoxO transcription factor inhibitor, AS1842856 (Nagashima, et al., 2010), alone and in conjunction with Ang II. When AS1842856 was administered alone to RPGSMCs, a significant reduction in sGCα mRNA (Figure 5A), sGCβ mRNA (Figure 5B), and sGCβ protein expression (Figure 5C,D) was observed. When co-administered with Ang II, AS1842856 was significantly reduced compared to Ang II treatment, but produced no significantly different effect on either sGCα mRNA (Figure 5A), sGCβ mRNA (Figure 5B), sGCα protein expression (Supplementary Figure 2A,C) or sGCβ protein expression (Figure 5C,D) when compared to controls. Treatment with AS1842856 also significantly blunted the expression of the downstream FoxO protein target, glucose-6-phosphatase (G6Pase), indicating that FoxO protein activity is potently suppressed (Supplementary Figure 3A). Additionally, this experiment showed that Ang II, Losartan and PD123319 had no significant effect on the expression of G6Pase. This suggests that Ang II-mediated changes do not universally impact FoxO activity (Supplementary Figure 3A). These data indicate the FoxO transcription factors are necessary for the Ang II-mediated sGC expression increases in renal smooth muscle.

Figure 5: AS1842856 treatment causes a decrease in sGC expression and blunts Ang II-mediated increases in sGC expression in renal smooth muscle cells.

RPGSMCs were treated with control (DMSO), Ang II (10⁻⁶ M), and/or AS1842856 (10⁻⁶ M) and expression of A) sGCα₁ mRNA (n=6 samples per group), B) sGCβ₁ mRNA (n=6 samples per group) and C, D) sGCβ protein expression (n=5 samples per group) quantified relative to α-tubulin protein expression. NS denotes a non-specific band. The * indicates a P < 0.05 statistically significant difference between identified sample groups by unpaired two tailed t-test or Mann-Whitney U-test. Error bars are ± SEM.

Next, we tested how cGMP production in RPGSMCs was influenced by Ang II receptor antagonists and FoxO inhibitors in the absence and presence of NO-stimulation with DEA-NONOate. In the absence of DEA-NONOate-stimulation, the vehicle-treated cells showed no significant differences in cGMP production or VASP phosphorylation were observed between the control, Ang II, AS1842856, Ang II + AS1842856, Ang II + Losartan, or Ang II + PD123319 treatment groups (Figure 6A–C). Similar to the observed effect in sGC expression, Ang II and PD123319 co-treatment followed by DEA-NONate stimulation produced significant increases in downstream sGC function via cGMP production (Figure 6A) and VASP phosphorylation (Figure 6B,C) compared to solely DEA-NONOate-stimulated controls. Treatment with AS1842856 or Ang II + Losartan followed by DEA-NONOate stimulation caused a decrease in cGMP production (Figure 6A) and showed no significant differences in VASP phosphorylation (Figure 6B,C) or total VASP expression (Supplementary Figure 2B, D) compared to control cells stimulated solely with DEA-NONOate. We also examined the expression of PDE5 and PKG1 and found that there were no significant changes in mRNA expression, suggesting that the changes in cGMP production and downstream signaling are due solely to changes in sGC expression Supplementary (Figure 3B,C). Combined, these data show that PD123319 has no significant effect on Ang II-mediated sGC signaling and that the blunting effects of Losartan or AS1842856 on the Ang II-mediated sGC expression also inhibited downstream cGMP signaling following NO-dependent stimulation.

Figure 6: Losartan or AS1842856 blunt Ang II-mediated increases in downstream cGMP signaling in renal smooth muscle cells following NO-stimulation.

A) Quantification of cGMP production by ELISA following treatment of RPGSMCs with control, Ang II (10⁻⁶ M), AS1842856 (10⁻⁶ M), Losartan (10⁻⁷ M), PD123319 (10⁻⁷ M) and then stimulated with vehicle or DEA-NONOate for 15 minutes (10⁻⁶ M). B, C) Quantification and representative western blot of phosphorylated serine 239 VASP (pVASP) protein following the same treatment as in A (n=5 samples per group). The * indicates a P < 0.05 statistically significant differences between identified sample groups by unpaired two-tailed t-test. Error bars are ± SEM.

DISCUSSION:

Renal artery stenosis remains a pervasive cause of secondary hypertension and a condition significantly correlated with high morbidity and mortality (Textor, 2003), (Kalra, et al., 2005), (de Mast and Beutler, 2009), (Kalra, et al., 2010). NO plays an important role in maintaining renal blood flow and glomerular filtration rate following single renal artery stenosis (Granger, et al., 2002), (Majid and Navar, 1997), (Majid, et al., 1998). In addition, there is emerging pre-clinical evidence that sGC stimulator drugs which have had notable anti-fibrotic effects, in conjunction with RAAS blockade confer resistance to end stage renal disease and chronic kidney disease (Beyer, et al., 2014), (Sandner and Stasch, 2017). Such therapies have demonstrated an ability to elevate blood flow and/or improve cardiac outcomes as a result of decreased vascular tone and decreased blood pressure (Stasch, et al., 2001), (Stasch, et al., 2011), (Evgenov, et al., 2006). Our previous study, in accordance with previous 2K1C models, showed that our 2K1C renal artery stenosis model causes increased blood pressure without altering body weight or plasma electrolyte concentrations (DeLalio, et al., 2020), (Ong, et al., 2019). Here we provide the first evidence that sGC expression increases in the vascular smooth muscle of the renal artery contralateral to stenosis to preserve renal blood flow in the 2K1C model.

In this study, we observed a significantly increased vasodilation in the contralateral renal arteries of 2K1C animals. Interestingly, vasodilatory function of the aorta and mesenteric arteries was not different between these sham versus 2K1C mice (DeLalio, et al., 2020). This increased vasodilation of the contralateral renal artery was likely due to the significant increase in sGC expression observed in unobstructed renal artery smooth muscle from 2K1C animals compared to their sham controls. We also observed an increase vasodilation of renal clip animals over the sham controls when treated with 10⁻⁷ M ACh, suggesting some contribution of the endothelium in this response. However, as indicated by the SNP results, the response of the smooth muscle appears to be the predominant change in the renal vasculature. This response may indicate that NO-dependent signaling, which has been established to be a pivotal player in promoting regulation of renovascular homeostasis of blood pressure and fluid retention (O’Connor and Cowley, 2010), (Dautzenberg, et al., 2011), is enhanced in the contralateral (non-stenotic) renal arteries compared with other vascular beds following elevated RAAS activity. Others have noticed the significance of endothelial-derived NO, specifically in the context of the 2K1C model, showing that inhibition of endothelial NO signaling exacerbates the high blood pressure and reduced renal blood flow of the 2K1C model (Sigmon and Beierwaltes, 1993), (Choi, et al., 1994). Additionally, this pressor response occurs irrespective to RAAS blockade, suggesting that while the regulation of these two pathways remain distinct, the interaction between them has vital implications (Sigmon and Beierwaltes, 1993). Moreover, studies of the 2K1C model suggest that Losartan-mediated RAAS blockade following NO-inhibition further diminishes blood flow to the contralateral renal artery instead of the expected increase in renal blood flow following RAAS inhibition (Sigmon and Beierwaltes, 1998). Our study builds upon this data, suggesting that the reason for the worsening of the phenotype when blocking Ang II signaling is due to an inability of the renal smooth muscle to mobilize a compensatory increase in sGC expression.

In the 2K1C model, it is possible that there are multiple factors which contribute to the changes observed in sGC expression, including RAAS signaling peptides such as Ang II. Therefore, we treated isolated renal pre-glomerular smooth muscle cells (RPGSMCs) with Ang II, to test the response of the renal vasculature to RAAS activation. Following treatment of cultured (RPGSMCs) with Ang II for 48 hours, the increase observed in sGC mRNA and protein expression suggests that RPGSMCs respond differently from aortic smooth muscle. Aortic smooth muscle and endothelial cells exhibit decreased functional NO signaling with excess Ang II exposure, via pathological overproduction of reactive oxygen species (ROS) (Griendling, et al., 1994), (Doughan, et al., 2008). Moreover, aortic sGC protein expression decreases with Ang II (Mollnau, et al., 2002), (Rippe, et al., 2017), and Ang II impairs aortic smooth muscle sGC function (Rippe, et al., 2017), (Crassous, et al., 2012). Furthermore, sGC activator therapy has been shown to rescue heart function in patients with acute decompensated heart failure, many of whom were on RAAS pathway drugs to treat their heart failure, hypertension and other vasculopathies (Erdmann, et al., 2013). However, the trial was halted because the dosage used for this study resulted in hypotension in patients. It may be possible that other signaling molecules like catecholamines, prostacyclins, and others may contribute to the physiological responses we observe following 2K1C, however, further research is needed to test these possibilities. Our studies are the first to show that treatment with Ang II in conjunction with NO-stimulation caused elevated cGMP signaling in RPGSMCs, as indicated by increased cGMP production and VASP phosphorylation. This indicates uniquely enhanced renal smooth muscle sGC-cGMP signaling following Ang II treatment in renal vascular smooth muscle.

Remarkably, other known responses to Ang II treatment were noted in RPGSMCs, such as elevated F-actin expression and increased cell size. These patterns have been observed in aortic smooth muscle both in vivo following infusion with Ang II and in vitro following Ang II treatment in culture (Geisterfer, et al., 1988), (Zhang, et al., 2005). This suggests that while the increases in sGC expression responses are unique to renal smooth muscle, the hypertrophic responses to Ang II conform to the patterns that have been observed by others.

Specifically, our data shows that the AT₁R, but not the AT₂R, is responsible for the elevated expression of sGC observed in renal smooth muscle in response to Ang II. Indeed, co-treatment with Losartan and Ang II was sufficient to reverse all of the Ang II-induced phenotypes we observed in RPGSMCs. In contrast, Ang II co-treatment with PD123319 did not impact any of the phenotypes facilitated by Ang II treatment alone. This response may be due to the high density of AT₁Rs that have been observed in the adventitia of the renal vasculature (Doughan, et al., 2008), (Harrison-Bernard, et al., 1997), and the increased constriction of renal vasculature and, to a smaller extent, gut vasculature following acute Ang II infusion (Jackson and Herzer, 2001). Curiously, this contrasts with the role for the AT₂R in cardiac function, which has been shown to improve outcomes following treatment with AT₂R-specific agonists following myocardial infarction (Kaschina, et al., 2008). Our findings suggest that the observed effect on sGC expression and function are largely independent of AT₂R activation. It is also possible that truncation products such as angiotensin 1–7 or angiotensin IV may also play role, albeit minor, in this phenomenon (Savergnini, et al., 2010), (Esteban, et al., 2005). Taken together, the data herein suggest that renal smooth muscle responds uniquely to Ang II via the AT₁R to promote increased sGC expression and sGC-cGMP induced vasodilation while maintaining the canonical hypertrophic responses associated with elevated Ang II exposure.

Consistent with our previous work in aortic SMCs (Galley, et al., 2019), Ang II studies in RPGSMCs showed that inhibition of the forkhead box subclass O (FoxO) transcription factors significantly impair sGC expression. This finding indicates FoxO regulation of sGC expression likely applies to multiple branches of the vascular tree. The FoxO protein(s) responsible for sGC expression regulation are not yet known and the specific role of the FoxO transcription factors in the development and pathology of renal artery stenosis requires further study to assess their diverse functions in vascular physiology. It is nevertheless clear that the Ang II-mediated increases in sGC function cannot occur without functional FoxO transcription factor activity. Ang II can activate Akt (Li and Malik, 2005), and Akt-mediated phosphorylation is a common regulatory mechanism known to modulate FoxO transcriptional activity (Biggs, et al., 1999), (Brunet, at al., 1999), (Kops, et al., 1999). In addition, Ang II has been shown to cause increases in ROS, and oxidative stress is known to impact FoxO transcription factor activity through acetylation/deacetylation (Salminen, et al., 2013), (Ichiki, et al., 2003), (Motta, et al., 2004), (van der Horst, et al., 2004). These findings suggest that there could be an indirect regulatory mechanism between the AT₁R and FoxO transcription factors. Future research in this area should investigate the potential mechanistic links between agonism of the AT₁R and activation of the FoxO transcription factors. Moreover, our research has shown that oxidation or loss of sGC heme iron leads to NO insensitivity, making the protein more responsive to sGC-activating compounds which target oxidized or heme-deficient sGC to produce cGMP (Rahaman, et al., 2017), (Durgin, et al., 2019). Investigation of Ang II-mediated ROS production may reveal a novel therapeutic target for sGC activating drugs under conditions where high RAAS activity promotes oxidative stress.

Collectively, we show for the first time that in response to elevated RAAS activity, renal smooth muscle responds through an AT₁R and FoxO transcription factor-dependent mechanism to increase sGC expression and cGMP signaling. These responses likely constitute a compensatory response to allow for maintenance of homeostatic blood volume and electrolyte balance to counteract Ang II-dependent increases in systemic blood pressure, and a means of preserving homeostatic fluid volume and electrolyte balance. These findings will have important clinical implications for the use of NO, FoxO transcription factor activating compounds, or sGC modulators for therapeutic treatment of renal hypertension. Our data suggest that these methods may offer a targeted approach to improve renal blood flow in RAAS-mediated renal stenosis. Combined, this study marks an important discovery of how the renal vasculature responds uniquely to elevated circulating plasma Ang II, advancing our understanding of renal vascular hypertension and the regulation of cGMP signaling within the renal vascular wall.

Supplementary Material

1

Supplementary Figure 1: Immunohistochemistry of contralateral renal arteries from 2K1C (renal clip) animals show an increase in sGCα expression over sham controls. Immunofluorescent staining of representative images and quantification of sham control (n=6, white bars) or renal clip (n=6, red bars) contralateral arteries for A, F) DAPI (nuclei, blue) B, G) ACTA2 (vascular smooth muscle, green), C, H) sGCα (red), and D, I) vWF (endothelium, grey). E) Merge image of maximum intensity projections from all channels. J) Quantification of medial area (μm²). The * indicates a P < 0.05 statistically significant difference between renal clip and sham by unpaired two-tailed t-test. Error bars are ± SEM.

Supplementary Figure 2: sGCα protein expression is AT₁R and FoxO-dependent, hypertrophy only requires AT₁R agonism, and tVASP expression remains unaffected in renal smooth muscle. Western blot analysis and total protein expression in RPGSMCs following treatment with Control, Ang II (10⁻⁶ M), AS1842856 (10⁻⁶ M), Losartan (10⁻⁷ M), and/or PD123319 (10⁻⁷ M). A, C) Representative western blot and quantification of sGCα protein expression relative to β-actin expression (n=5 samples per group). B, D) Representative western blot and quantification of total vasodilator stimulated protein (tVASP) protein expression relative to β-actin expression (n=5 samples per group). E) Total protein quantified using a bicinchoninic acid protein assay (n=5 samples per group). The * indicates a P < 0.05 statistically significant difference between indicated groups by unpaired two-tailed t-test with or without Welch’s correction. Error bars are ± SEM.

Supplementary Figure 3: mRNA expression of sGC pathway and FoxO pathway-related genes. qRT-PCR of genes expressed by RPGSMCs following treatment with Control, Ang II (10⁻⁶ M), AS1842856 (10⁻⁶ M), Losartan (10⁻⁷ M), and/or PD123319 (10⁻⁷ M). Graphs represent mRNA expression of A) glucose-6-phosphatase (G6Pase), B) cGMP-specific phosphodiesterase type 5 (PDE5), and C) cGMP-activated protein kinase 1 (PKG1). For each group there are n=6 samples. The * indicates a P < 0.05 statistically significant difference between control and treated groups by unpaired two-tailed t-test with Welch’s correction or Mann-Whitney U-test. Error bars are ± SEM.

SUMMARY:

• What is already known: Renovascular hypertension via renal artery stenosis activates RAAS signaling.

• What this study adds: Angiotensin II stimulates renovascular sGC expression through AT₁R and FoxO-dependent mechanisms to improve vascular function.

• Clinical significance: Enhanced RAAS-mediated signaling promotes higher sGC expression and signaling through AT₁R and FoxO transcription factors.

Data Availability Statement:

The data that support the findings of this study are available from the corresponding author upon reasonable request. Some data may not be made available because of privacy or ethical restrictions.