Responses in Blood Pressure and Kidney Function to Soluble Guanylyl Cyclase Stimulation or Activation in Normal and Diabetic Rats

Abstract

Introduction:

Agonists of soluble guanylate cyclase (sGC) are being developed as treatment for cardiovascular disease. Most effects of nitric oxide (NO) on glomerular and tubular function are mediated through sGC but whether sGC agonists mimic these effects is unknown.

Methods:

Renal clearance and micropuncture studies were performed in Wistar-Froemter rats (WF), with or without streptozotocin diabetes (STZ-WF), and in Goto-Kakizaki rats (GK) with mild type-2 diabetes to test for acute effects of the sGC “stimulator” BAY 41-2272, which synergizes with endogenous NO, and the “activator” runcaciguat, which generates cGMP independent of NO.

Results:

Both sGC agonists reduced arterial blood pressure (MAP). For MAP reductions <10% the drugs increased GFR in WF and STZ-WF but not in GK. Larger MAP reductions outweighed this effect and GFR declined, with better preserved GFR in STZ-WF. Changes in GFR could not be accounted for by changes in RBF, suggesting parallel changes in ultrafiltration pressure and/or ultrafiltration coefficient. The doses chosen for micropuncture in WF and GK reduced MAP by 2–10% and the net effect on single nephron GFR and ultrafiltration pressure was neutral. Effects of the drugs on tubular reabsorption were dominated by declining MAP and no natriuretic effect observed at any dose.

Discussion/Conclusion:

sGC agonists impact kidney function directly and because they reduce MAP. The direct tendency to increase GFR is most apparent for MAP reductions <10%. The direct effect is otherwise subtle and overridden when MAP declines more. Effects of sGC agonists on tubular reabsorption are dominated by effects on MAP.

INTRODUCTION

Soluble guanylyl cyclase (sGC) is an enzyme that promotes vasodilation by facilitating or mediating the effects of nitric oxide (NO) and thereby has large significance for the health of the cardiovascular system [1]. NO’s binding to the heme moiety of sGC stimulates the biosynthesis of the second messenger cyclic guanosine monophosphate (cGMP) from guanosine triphosphate (GTP)[2]. NO-sGC-cGMP signaling induces vascular relaxation and lowers blood pressure [1]. This involves cGMP-induced activation of protein kinase G (PKG), which phosphorylates various proteins, including potassium channels, the IP3 receptor and the myosin light chain phosphatase thereby lowering cytosolic calcium and decreasing the sensitivity of contractile filaments in smooth muscle cells, all leading to vasodilation [3]. The latter can reduce blood pressure with potential beneficial effects on the cardiovascular and renal system [4, 5]. In addition, the NO-sGC-cGMP pathway has been proposed to have antiproliferative, antiinflammatory, and antifibrotic effects [1]. Furthermore, sGC expression is widely distributed in the kidney itself, including the renal vascular system and glomeruli as well as the renal tubular system, consistent with the concept that local sGC activity and cGMP have direct effects on glomerular function, renal blood flow, tubular transport processes and renin secretion [6, 7, 5].

NO and cGMP deficiency have been implicated in the pathogenesis of cardiovascular and renal diseases, including hypertension and diabetic nephropathy [8, 9, 7]. As a new strategy for the treatment of endothelial dysfunction and cardiovascular and renal diseases, synthetic compounds like sGC stimulators and sGC activators are being studied as pharmacological approaches to restore cGMP levels with promising results [7, 10, 5, 1]. sGC activators can restore damaged heme moiety and place the enzyme in an activated form, while sGC stimulators sensitize sGC to NO and directly bind to sGC and trigger cGMP production [11]. sGC stimulation lowered blood pressure and improved renal tissue remodeling in a hypertensive rat model [12] and attenuated kidney fibrosis induced by ureteral obstruction [13]. In renin transgenic rats, blood pressure was reduced or normalized and kidney function improved with sGC stimulators riociguat and vericiguat [14, 15]. In the Dahl salt-sensitive hypertension rat model, the sGC stimulator IW-1973 reduced blood pressure, inflammatory cytokine levels, and renal disease markers, including proteinuria and renal fibrotic gene expression [16]. Moreover, sGC stimulation increased NO-induced glomerular cGMP production and reduced mean arterial pressure, fibrosis, and proteinuria in anti-thy1 glomerulonephritis [17].

In diabetes, the heme moiety of sGC can be oxidized due to oxidative stress, making sGC unavailable to bind with NO [18]. Reduced NO-sGC-cGMP signaling contributes to endothelial dysfunction and thereby to diabetic nephropathy. As a consequence, the effects of sGC agonists are being studied to activate oxidized sGC (stimulators) and heme-free sGC (activators)[19]. In ZSF-1 rats, which are diabetic, obese and have hypertension, the sGC stimulator IW-1973 reduced blood pressure (BP) and proteinuria [20]. Currently IW-1973 is also being studied in patients with diabetic nephropathy (NCT03217591).

The current study aimed to compare acute effects of a systemically administered sGC stimulator and an sGC activator on renal hemodynamics, tubular function, and BP in non-diabetic rats and in rat models of type 1 and type 2 diabetes. The study also characterized the acute renal sodium retention in response to high doses of the compounds as well as the oxygenation of kidney regions when high doses of the compounds were given over multiple days.

METHODS

Compounds:

The studies tested the soluble guanylyl cyclase (sGC) stimulator BAY 41-2272 and the sGC activator runcaciguat. Previous studies established IC50 values for contraction inhibition in isolated aortic rings for BAY 41-2272 (tested in rabbit) and runcaciguat (tested in pig) of 0.3 μM [21] and 0.137 μM [22], respectively. In anesthetized Wistar rats, BAY 41-2272 induced a dose-dependent reduction in blood pressure after i.v. and oral administration and the compound had an oral bioavailability of <50% and a half-life of 0.5 h [21]. BAY 41-2272 is quickly oxidized at its 5-pyrimidinyl-cyclopropyl residue and the concentration of this metabolite was about twice as high as BAY 41-2272 and showed comparable in vivo blood pressure-lowering activity [21]. Runcaciguat has an oral bioavailability of > 90% and a half-life of 7.5 h in rat [22]. Both compounds were provided by Bayer, Germany.

Animals:

All animal experimentation was conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD) and was approved by the local Institutional Animal Care and Use Committee. All animals that entered the study, also completed the study and their data was included in analysis.

Male Wistar Froemter rats (WF) were derived from lab’s breeding colony. In randomly selected rats, a model of type 1 DM (T1DM) was induced by intraperitoneal injection of streptozotocin (65 mg/kg)(STZ-WF). Non-diabetic controls (WF) were injected with vehicle (citrate buffer). Insulin (Recombinant human insulin, Boehringer Ingelheim Vetmedica Inc., St. Joseph, MO) was injected as needed to maintain blood glucose levels in STZ-WF at approximately 350 mg/dl. Animals were fed with standard rodent chow (TD.7001; Envigo, Indianapolis, IN).

Goto-Kakizaki rats (GK) are a non-obese model of type 2 DM (T2DM)[23]. Breeding pairs were obtained from the Medical College of Wisconsin (Milwaukee, WI, USA). Starting at 4 weeks of age, experimental male GK rats were fed a 45% kcal Fat Diet (TD.08811; Envigo, Indianapolis, IN, USA) to enhance the hyperglycemic phenotype.

All animals were maintained at 23 ± 2°C under a cycle of 12:12-h light-dark cycle and allowed free access to food and water. There were 4 series of experiments.

Series 1 included renal clearance experiments to determine the acute effects of BAY 41-2272 and runcaciguat on mean arterial blood pressure (MAP), GFR, RBF, and fractional excretion of fluid and electrolytes in WF, STZ-WF, and GK rats under general anesthesia.

Rats were anesthetized with thiobutabarbital (100 mg/kg i.p., Sigma)[24, 25] and placed on a servo-controlled heating table to maintain body temperature at 37 °C. A tracheostomy was performed to secure the airway. A catheter was placed in the left femoral artery for collecting blood samples and for monitoring mean arterial pressure (MAP) using a blood pressure analyzer (BPA Model 300, Micro-Med. Inc., Louisville, Kentucky). Another catheter was placed into a jugular vein for infusion of maintenance fluid, a GFR marker, and test compounds. Total infusion rates were 0.7 ml hr⁻¹ 100g body wt⁻¹ in non-diabetic animals and 1.2 ml hr⁻¹ 100g body wt⁻¹ in diabetic rats to compensate for higher urine flow rates and maintain hematocrit constant during the experiment. Hyperglycemia was maintained in STZ-WF and GK rats during anesthesia by adding glucose (20 mM) to the infusion. The left kidney was exposed by a flank incision. The left ureter was cannulated for urine collection. The left renal artery and vein were separated by blunt dissection and 1-mm transit-time flow probe (model BLF 21D, Transonic System Inc., Ithaca, NY) was placed on the renal artery to measure renal blood flow (RBF) of the left kidney. GFR in the left kidney was determined by the urinary clearance of FITC-sinistrin (continuous infusion of 0.2% FITC-sinistrin [FTC-FS001; Fresenius Kabi, Bad Homburg, Germany] in 3% BSA and 0.5% DMSO 0.01N NaOH in 0.9% NaCl solution). Renal vascular resistance (RVR) was estimated as [MAP/RBF]. Renal plasma flow was calculated as [RBF × (1−hematocrit)]. Renal filtration fraction (FF) was calculated as [GFR/RPF]. Fractional excretion (FE) of fluid was calculated as [urine flow rate/GFR]. FE of electrolytes (Na⁺, K⁺, Cl⁻) was calculated as [plasma electrolyte concentration × GFR / (urine electrolyte concentration × urine flow rate)].

Data were gathered from each animal in 4 or 5 experimental periods. Test compounds were administered by constant intravenous infusion. During the first period of each experiment, animals received vehicle (0.5% DMSO/0.01N NaOH in 0.9% NaCl). In subsequent periods, animals received test compounds at increasing doses. Test compounds included vehicle, BAY 41-2272, or runcaciguat. Each period included 10 minutes equilibration time at the new dose followed by a 20-minute urine collection. Blood samples were taken at the beginning and end of each urine collection for FITC-sinistrin assay. Drug delivery rates were chosen to have either no effect on MAP, a small MAP-lowering effect, or a more pronounced MAP-lowering effect (BAY 41-2272: 0.02, 0.2 and 2 mg kg⁻¹ h⁻¹; runcaciguat: 0.002, 0.02, 0.2 and 2 mg kg⁻¹ h⁻¹). Experiments where vehicle was the test substance served as time-controls. For both BAY 41-2272 and runcaciguat, a dose of 0.2 mg kg⁻¹ h⁻¹ induced a small reduction in blood pressure that did not translate into major changes in GFR and RBF; for this dose we also determined urine excretion of albumin and cGMP.

Series 2 included renal micropuncture experiments to determine acute effects of BAY 41-2272 and runcaciguat on single nephron function in WF and GK. Measurements were made of single nephron GFR (SNGFR), proximal tubular fluid and chloride reabsorption, and of hydrostatic pressures in glomerular capillaries, peritubular capillaries, and proximal tubules.

Rats were prepared for renal micropuncture under inactin anesthesia as described [26–29]. Canulae were placed in femoral artery, jugular vein and left ureter as described above for Series 1. The left kidney was secured in a lucite cup and superfused with warm Ringer saline. ³H-inulin (80 μCi/h) was administered as a marker of GFR. Total volume of fluid administered was the same as for Study 1.

Measurements were made while infusing vehicle (Period 1) and repeated while infusing test compounds (Period 2). Test compounds included vehicle (time control), BAY 41-2272 (0.6 mg kg⁻¹ h⁻¹), or runcaciguat (0.6 mg kg⁻¹ h⁻¹). The doses of BAY 41-2272 and runcaciguat were targeted to reduce MAP by a small amount (5–15%) based on Series 1. Infusion of test compounds was initiated after completing Period 1 and 15 min prior to starting Period 2. Each experimental period included several timed tubular fluid collections made from the last accessible segment of proximal tubule and earliest accessible segment of distal tubule. The volume of a tubular fluid collection was determined by transfer to a constant bore glass capillary. SNGFR was determined by ³H-inulin clearance in each tubular fluid collection. Chloride concentration of tubular fluid was measured with a micro adaptation of the electrometric titration method [27, 29]. Fractional delivery of fluid and chloride to the late proximal tubule, early distal tubule and urine were calculated according to standard formulae. WF rats have glomeruli visible on the kidney surface with early distal tubules accessible for micropuncture. GK rats do not have surface glomeruli so the earliest accessible distal tubule tends to be farther downstream, thus making early distal chloride delivery indeterminate for GK rats. Therefore, data reported for GK rats do not include measurements made on distal tubular fluid. Pressures in proximal tubules under free-flow (P_FF) and stop-flow (P_SF) conditions and in peritubular capillaries (P_PTC) were measured with a servo-null micropressure system (Vista Electronics, San Diego CA) [28]. Glomerular capillary pressure (P_GC) was computed as the sum of P_SF and plasma colloid oncotic pressure, which was computed from the protein concentration in arterial plasma according to the Landis Pappenheimer equation [28].

Series 3 was done in mice to determine whether high dose BAY 41-2272 or runcaciguat affects sodium transport by ENaC in the distal nephron.

Male mice with a conditional knockdown of the alpha-subunit of ENaC in connecting tubule and collecting duct segments (using AQP2-driven Cre expression) were compared with their male wild-type littermates [30]. BAY 41-2272 (3 mg/kg), runcaciguat (3 mg/kg) or vehicle were given by oral gavage to conscious mice along with a water load (25 μl/g body wt). Mice were then transferred to metabolic cages for the duration of a 3-hour urine collection. Mice remained without access to food or water during the urine collection [31, 32].

Series 4 examined the effect of high doses of BAY 41-2272 and runcaciguat on a marker of kidney oxygenation in GK rats.

Vehicle (ethanol/kolliphor HS15^®/ tap water (10/40/50 v/v/v), 5 ml/kg), BAY 41-2272 (3 mg/kg bw), or runcaciguat (3 mg/kg bw) were administered by oral gavage twice daily for 9 doses. The final dose was given 3–4 hrs prior to blood collection and kidney harvest. Renal hypoxia was evaluated immunohistochemically using the hypoxic probe-1 system of pimonidazole (Chemicon, Temecula, CA) [33]. Briefly, pimonidazole (60 mg/kg) was injected i.p. into rats 2 hours before kidney harvesting. During this 2 hour period, pimonidazole is reductively activated and protein adducts are generated that are detected with a specific antibody. Rats were anesthetized with thiobutabarbital (100 mg/kg i.p., Sigma) and blood collected by puncturing the descending aorta (~150 μl) for measurement of the compounds in plasma. Kidneys were harvested by perfusion fixation with 4% paraformaldehyde and prepared for pimonidazole immunostaining according to established protocols [33]. Levels of hypoxia were determined in kidney regions of interest using digital HSCORE [34]. The investigator was blinded to the treatment.

Sample analysis:

Urine was analyzed for cGMP (RE29071; IBL International, Hamburg, Germany) and albumin (ab108790; Abcam, Cambridge, MA) using commercial assays. Na⁺, K⁺, and Cl⁻ in plasma and urine were determined by ion-selective electrodes (Easy Electrolyte, Medica Corporation, Bedford, MA). Plasma concentrations of BAY 41-2272 and runcaciguat were determined by LC-MS.

Data analysis:

All data are presented as mean±SE. P<0.05 was considered statistically significant. Series 1: Basal values obtained during the initial vehicle period for WF and GK were compared with STZ-WF by 1-way ANOVA. With regard to drug effects, fractional changes (in %) in a given parameter in response to drug or vehicle time control were calculated for each period versus the initial vehicle period. Effects of BAY 41-2272 and runcaciguat were analyzed by 2-way ANOVA for the factors drug and period and the interaction term. If the P-value for drug or the interaction term were significant, then pairwise multiple comparison procedures (Holm-Sidak method) were performed to determine for which dose drug effects were different from time control vehicle. In addition, a linear regression analysis was performed for effects on filtration fraction versus GFR. Series 2: For paired data (MAP, GFR, fractional excretion into urine) the change in period 2 vs period 1 in response to BAY 41-2272, runcaciguat or vehicle was determined and the responses compared by 1-way ANOVA. Unpaired individual micropuncture data were analyzed by 2-way ANOVA with the cross term (group × period) indicating whether drug is different from vehicle, followed by pairwise multiple comparison procedures (Holm-Sidak method) to determine which groups were different in period 1 or period 2. Series 3: 2-way ANOVA for factors drug and ENaC knockout was performed followed by pairwise multiple comparisons (Holm Sidak method). Series 4: 2-way ANOVA for factors drug and kidney region was performed followed by multiple pairwise comparisons (Holm Sidak method).

RESULTS

Series 1 - Baseline values in WF, STZ-WF and GK.

See Figure 1 and Table 1. STZ-WF had higher blood glucose levels, slightly lower MAP, lower body weight, higher kidney weight, higher GFR, higher RBF, more albuminuria, and higher urinary cGMP excretion than their non-diabetic WF counterparts. Excretion of fluid and sodium were also higher in STZ-WF, reflecting the osmotic diuresis, polyphagia and polydipsia, which are stereotypical of diabetes.

Figure 1: Basal values in WF, STZ-WF and GK.

Basal values obtained for WF and GK during the initial vehicle period were compared with STZ-WF by 1-way ANOVA. *P<0.05 vs WF; # P<0.05 vs STZ-WF. N=18–22 rats/group. FF, filtration fraction; GFR, glomerular filtration rate; MAP, mean arterial pressure; RBF, renal blood flow; RVR, renal vascular resistance. See text for result details.

Table 1:

Basal characterization of 3 rat groups

	WF	STZ-WF	GK
Plasma [Na+] (mM)	144.5±0.3	146.8±0.5*	144.7±0.4#
Plasma [Cl−] (mM)	109.9±0.6	110.0±0.6	108.9±0.5
Plasma [K+] (mM)	4.9±0.1	4.9±0.1	4.8±0.1
UV (μl/min)	3.8±0.5	13.8±1.3*	3.3±0.4#
UNaV (nmol/min)	267±46	630±106*	122±22#
UClV (nmol/min)	1204±145	1233±133	426±73#
UKV (nmol/min)	995±126	1187±133	612±93#
FE fluid (%)	0.21±0.04	0.53±0.05*	0.09±0.01#
FE Na+ (%)	0.10±0.02	0.17±0.03*	0.02±0.01#
FE Cl− (%)	0.56±0.08	0.43±0.05	0.10±0.02#
FE K+ (%)	11±2	9±1	4±1#

Data are mean±SE; n=18–22/group

^*P<0.05 vs WF

^#P<0.05 vs STZ-WF.

1-way ANOVA. UV, urinary flow rate; UNaV, urinay sodium excretion; UCIV, urinary chloride excretion; UKV, urinary potassium excretion; FE, fractional urinary excretion.

Diabetes was milder in GK (non-obese type 2 diabetes) than in STZ-WF. In awake GK, blood glucose levels were only modestly higher (~150 mg/dl) than in WF (~100 mg/dl). In GK rats, body weight, MAP, heart rate, RBF and GFR were higher than in STZ-WF, whereas cGMP excretion, kidney weight and albuminuria were less. Lower kidney weight and lesser albuminuria in GK may reflect the milder diabetes in GC vs STZ-WF [35]. Blood glucose levels increased in GK during clearance studies (~340 mg/dl), possibly reflecting their impaired glucose homeostasis, which was challenged by surgical stress and glucose infusion (Figure 1).

Baseline excretion of fluid and electrolytes was lower in GK than STZ-WF while GFR was higher in GK than STZ-WF. It is possible that the higher GFR in GK was partially due to the increase in glucose levels induced during the clearance studies, which is expected to enhance Na-glucose cotransport and increase GFR by signaling through the macula densa [36]. Differences in baseline fluid and electrolyte excretion could reflect differences in food and fluid intake leading up to the experiment because the more severe diabetes in STZ-WF makes them polydipsic and polyphagic.

Series 1 - Acute effects of BAY 41-2272 and runcaciguat on MAP, GFR, and RBF and in WF, STZ-WF and GK.

Blood glucose levels were not significantly affected by BAY 41-2272 or runcaciguat versus vehicle (data not shown).

BAY 41-2272 and runcaciguat, both dose-dependently lowered MAP in WF, STZ-WF and GK. The highest doses (2 mg kg⁻¹ h⁻¹) reduced MAP by 20–30% below levels recorded in vehicle controls. Runcaciguat was slightly more potent than BAY 41-2272 at reducing MAP (28.0±1.3 vs 22.4±1.3 %, P<0.005). See Figures 2 and 3.

Figure 2: Acute effect of BAY 41-2272 on MAP and renal hemodynamics in WF, STZ-WF and GK.

Fractional changes (in %) in a given parameter in response to BAY 41-2272 or vehicle time control were calculated for each period versus the initial vehicle period. Data were analyzed by 2-way ANOVA for the factors BAY 41-2272 (BAY) and dose and for the interaction term (inter). If the P-value for BAY or the interaction term was significant, then pairwise multiple comparison procedures (Holm-Sidak method) were performed to determine for which dose drug effects were different from time control vehicle. N=6–8/group; * P<0.05 vs veh control. See text for result details.

Figure 3: Acute effect of runcaciguat on MAP and renal hemodynamics in WF, STZ-WF and GK.

Fractional changes (in %) in a given parameter in response to runcaciguat or vehicle time control were calculated for each period versus the initial vehicle period. Data were analyzed by 2-way ANOVA for the factors runcaciguat and dose and for the interaction term (inter). If the P-value for runcaciguat or the interaction term was significant, then pairwise multiple comparison procedures (Holm-Sidak method) were performed to determine for which dose drug effects were different from time control vehicle. N=6–8/group; * P<0.05 vs veh control. See text for result details.

In WF rats, RBF was unaffected by BAY 41-2272 despite declining MAP (Figure 2). By contrast, RVR remained constant and RBF declined along with MAP in WF rats during runcaciguat infusion (Figure 3).

In diabetic STZ-WF, RBF was preserved as MAP declined during infusion of either BAY 41-2272 or runcaciguat. Notably, GFR and FF modestly increased during 0.2 mg kg⁻¹ h⁻¹ of runcaciguat relative to vehicle (see below for GFR discussion).

In GK rats, RBF was also unaffected even though MAP declined by as much as 20–30% during BAY 41-2272 or runcaciguat infusions.

The renal vasodilation that prevented RBF from declining in the face of declining MAP, as observed in all groups except in WF receiving runcaciguat, could have resulted from reflex RBF autoregulation or from direct vasodilatory effects of test compounds on both renal and extra-renal vasculature.

The effect of BAY 41-2272 and runcaciguat on MAP may give the best indication of their effective modulation of sGC activity. Therefore, and to take a closer look at GFR responses, Figure 4 depicts changes in renal hemodynamics relative to changes in MAP in all groups.

Figure 4: Relationship between MAP and renal hemodynamic responses to BAY 41-2272 and runcaciguat.

Each symbol reflects a rat that received BAY 41-2272 or runcaciguat; changes in renal hemodynamics in response to BAY 41-2272 or runcaciguat relative to vehicle time control are depicted versus changes in MAP. Rats with a reduction in MAP <10% or >10% were pooled (mean values ± SEM shown) and probed for significant differences in RVR, RBF, GFR and FF versus vehicle control by Welch’s t-test where equal variance is not assumed. * P<0.05 vs veh control; “*” positioned to the left or right of “BAY” or “runc” refer to significant changes in rats with a reduction in MAP <10% and >10%, respectively. See text for further details.

In non-diabetic WF, the data suggest a biphasic effect on GFR for both compounds (Figure 4). For drug-induced MAP reductions <10%, mean GFR values were modestly higher than vehicle time controls, whereas for MAP reduction ≥10% mean GFR values were lower. The modestly higher GFR values accompanying small MAP reductions were associated with slightly lower RVR and higher RBF for runcaciguat while FF remained unchanged for both compounds. In comparison, the lower GFR values in response to MAP reductions ≥10% were associated with lower values of FF in response to both compounds, and in response to runcaciguat also with reduced RBF, while the latter was preserved in response to BAY 41-2272 associated with lower RVR.

As in non-diabetic WF, both compounds slightly increased GFR in STZ-WF for MAP reductions <10% (Figure 4). This increase in GFR was associated with higher FF. For MAP reductions ≥10%, GFR was better maintained in STZ-WF than in WF, associated with better preserved FF.

In GK, neither compound increased GFR or FF for MAP reductions <10% (Figure 4) whereas MAP reductions ≥10% were accompanied by lower GFR and reduced FF for both compounds, similar to the effects in non-diabetic WF.

Thus, BAY 41-2272 and runcaciguat appear to slightly increase GFR for MAP reductions <10% in non-diabetic and STZ-diabetic WF but not in GK. On the other hand, BAY 41-2272 and runcaciguat reduced GFR for MAP reductions ≥10% in non-diabetic WF and GK. The GFR reduction in non-diabetic WF in response to runcaciguat seemed to be associated with less efficient autoregulation of RBF. In comparison, renal autoregulation, FF and GFR were better preserved in response to both compounds in STZ-WF. Figure 5 uses linear regression analysis to illustrate the significant positive association between changes in FF and GFR in response to both sGC agonists using data of all 3 groups of rats.

Figure 5: Changes in FF as a potential determinant of changes in GFR in response to BAY 41-2272 and runcaciguat.

Each symbol reflects a rat that received BAY 41-2272 or runcaciguat. Linear regression analysis was performed for each compound across the combined 3 groups of rats. See text for further details.

Series 1 – Does acute application of BAY 41-2272 or runcaciguat at a dose that induces only a small MAP reduction alter albuminuria or urinary cGMP excretion in WF, STZ-WF and GK?

For both BAY 41-2272 and runcaciguat, a dose of 0.2 mg kg⁻¹ h⁻¹ reduced MAP by 5–10% (Figures 2 & 3). Urinary cGMP tended to increase during BAY 41-2272 (increase by 54±25 vs 3±8% in vehicle control; P=0.083), but no effect of either compound on urinary cGMP or albuminuria met the P=0.05 threshold for statistical significance versus vehicle control (change in urinary cGMP: BAY 41-2272: listed as drug vs. vehicle; WF: −17±21 vs. 15±18%; STZ-WF: 5±12 vs. 21±12%; GK: 54±25 vs. 3±8%; runcaciguat: WF: 6±10 vs. 34±31%; STZ-WF: 25±13 vs. 44±18%; GK: −12±8 vs. 3±6%, all NS)(change in albuminuria: BAY 41-2272: listed as drug vs. vehicle; WF: −10±8 vs. −16±9%; STZ-WF: 7±6 vs. −6±13%; GK: −8±12 vs. 11±21%; runcaciguat: WF: −8±11 vs. −5±10%; STZ-WF: −5±7 vs. −18±6%; GK: 8±22 vs. 32±15%, all NS).

Series 1 - Acute effects of BAY 41-2272 and runcaciguat on kidney transport in WF, STZ-WF and GK.

As shown in Figure 6 for BAY 51-2272 and in Figure 7 for runcaciguat, the clearance studies did not provide strong evidence that either compound directly affects fluid or electrolyte transport by the kidney in comparison to vehicle control. The observed time-dependent increase in FE of Na⁺ and Cl⁻ in response to vehicle application may reflect a partial relaxation of the anti-natriuretic influence of anesthesia and surgical fluid loss in response to the infused saline. Lower fractional excretions during high dose BAY 41-2272 or runcaciguat may have reflected lesser pressure natriuresis and/or neurohumoral responses to lower blood pressure.

Figure 6: Acute effect of BAY 41-2272 on renal reabsorption in WF, STZ-WF and GK.

Fractional changes (in %) in a given parameter in response to BAY 41-2272 or vehicle time control were calculated for each period versus the initial vehicle period. Data were analyzed by 2-way ANOVA for the factors BAY 41-2272 (BAY) and dose and for the interaction term (inter). If the P-value for BAY or the interaction term was significant, then pairwise multiple comparison procedures (Holm-Sidak method) were performed to determine for which dose drug effects were different from time control vehicle. N=6–8/group; * P<0.05 vs veh control. See text for result details.

Figure 7: Acute effect of runcaciguat on renal reabsorption in WF, STZ-WF and GK.

Fractional changes (in %) in a given parameter in response to runcaciguat or vehicle time control were calculated for each period versus the initial vehicle period. Data were analyzed by 2-way ANOVA for the factors runcaciguat and dose and for the interaction term (inter). If the P-value for runcaciguat or the interaction term was significant, then pairwise multiple comparison procedures (Holm-Sidak method) were performed to determine for which dose drug effects were different from time control vehicle. N=6–8/group; * P<0.05 vs veh control. See text for result details.

Series 2 – Does acute application of BAY 41-2272 or runcaciguat at a dose that has only a small effect on MAP affect glomerular or peritubular capillary pressure in non-diabetic WF or GK?

A dose with only a small effect on MAP was chosen based on results from Series 1. In two-period micropuncture experiments in non-diabetic WF and GK, infusion of vehicle, BAY 41-2272 or runcaciguat (both 0.6 mg kg⁻¹ h⁻¹) was initiated after completion of period 1 and ~15 minutes prior to initiating period 2.

MAP declined by 2–10% during BAY 41-2272 or runcaciguat as planned. Compared to vehicle controls, the effect on MAP was statistically significant for GK but not for WF (Figure 8). GFR of the micropuncture kidney was not significantly affected by either drug in this series of micropuncture experiments.

Figure 8: Acute effect of BAY 41-2272 or runcaciguat on single nephron GFR and hydrostatic pressures in WF and GK – using a dose with only a small effect on blood pressure.

Two period (p1, p2) renal micropuncture studies were performed with infusion of BAY 41-2272, runcaciguat (both 0.6 mg kg⁻¹ h⁻¹) or vehicle initiated prior to period 2. For paired data (MAP, GFR) the change in period 2 vs period 1 in response to BAY 41-2272, runcaciguat or vehicle was calculated and the responses compared by 1-way ANOVA. Unpaired individual micropuncture data (single nephron GFR (SNGFR); hydrostatic pressures in glomerular capillaries (P_GC), proximal tubules (P_PT), and peritubular capillaries (P_PTC)) were analyzed by 2-way ANOVA with the cross term (group × period) indicating whether the drug was different from vehicle, followed by pairwise multiple comparison procedures (Holm-Sidak method) to determine which groups were different in p1 or p2. N=7–8 MF rats /group; with n=12–18 single nephron collections per period and group for late proximal and early distal collections, respectively; and individual pressure measurements per period and group: n=34–50 for P_GC, 33–41 for P_PTC, 31–42 for P_PT. N=5 GK rats/group; with n=22–30 single nephron collections from late proximal tubules per period and group; and individual pressure measurements per period and group: n=16–26 for P_GC, 13–20 for P_PTC, 20–26 for P_PT. * P<0.05 vs vehicle. See text for result details.

SNGFR and hydrostatic pressure in glomerular capillaries (P_GC) and proximal tubules (P_PT) were unaffected by either drug compared to vehicle (Figure 8). Data shown for SNGFR were derived from late proximal tubule fluid collections. For WF (but not GK – see Methods), SNGFR measurements were also made in fluid collected from early distal tubules. SNGFR measured by collecting from proximal tubules is expected to be greater than SNGFR as measured by collecting from distal tubules because tubuloglomerular feedback (TGF) is interrupted during collection from proximal tubules [28]. Neither proximal or distal SNGFR were significantly affected by either test compound. Neither was the difference between proximal and distal SNGFR affected. Runcaciguat did not significantly alter hydrostatic pressure in peritubular capillaries (P_PTC) in WF or GK. BAY 41-2272 did not significantly alter P_PTC in WF, but caused a small increase among GK (Figure 8).

Whereas GK rats showed glomerular hyperfiltration in Series 1 (Figure 1), assessment of SNGFR in micropuncture studies did not indicate hyperfiltration and the SNGFR values were even modestly lower compared with WF rats (Figure 8). In the micropuncture studies, GFR of the micropunctured kidney during period 1 was likewise modestly lower in GK versus WF (989±37 vs 1387±70 μl/min), indicating that single nephron and whole kidney GFR was lower in GK rats subjected to micropuncture than in the GK rats subjected to renal clearance studies in Series 1. Blood glucose levels in GK were somewhat less during micropuncture compared with the clearance studies in Series 1 (296±14 versus 350±16 mg/dl). While this may contribute to a lower GFR during micropuncture, it is clearly not expected to fully explain the difference. Alternative explanations include a greater sensitivity of the GK kidney to the procedure needed for embedding the kidney for micropuncture or, as sometimes observed, significant changes in the GFR of a rat line over time (micropuncture was performed 6–10 months after the clearance studies).

Series 2 – Does BAY 41-2272 or runcaciguat affect tubular transport in WF or GK?

The analysis of renal transport on the whole kidney level, as shown in Figures 6 and 7, argued against major effects of the 2 compounds. To exclude segment specific and potentially opposing effects, transport was also assessed by micropuncture. As shown in Figure 9, in non-diabetic WF, the effect of BAY 41-2272 or runcaciguat on the fractional reabsorption or delivery of fluid up to the late proximal tubule, early distal tubule and urine was not significantly different from vehicle. However, both BAY 41-2272 and runcaciguat slightly reduced fractional delivery of chloride to early distal tubules, while the delivery to late proximal tubules and urine was not significantly different from vehicle. Thus, in non-diabetic WF, BAY 41-2272 and runcaciguat may have shifted a fraction of chloride reabsorption from the distal nephron and collecting duct to the upstream loop of Henle.

Figure 9: Acute effect of BAY 41-2272 or runcaciguat on fractional fluid and chloride delivery up to late proximal tubule, early distal tubule and into urine in WF or GK – using a dose with only a small effect on blood pressure.

Two period (p1, p2) renal micropuncture studies were performed with infusion of BAY 41-2272, runcaciguat or vehicle initiated prior to period 2 – same studies/doses as shown in Figure 8. Unpaired individual micropuncture data were analyzed by 2-way ANOVA with the cross term (group × period) indicating whether drug is different from vehicle, followed by pairwise multiple comparison procedures (Holm-Sidak method) to determine which groups were different in p1 or p2. N=7–8 MF rats/group; with n=12–18 single nephron collections per period and group for late proximal and early distal collections, respectively. N=5 GK rats/group; with n=22–30 single nephron collections from late proximal tubules per period and group. * P<0.05 vs vehicle. See text for result details.

As shown in Figure 9, in GK, BAY 41-2272 had no detectable effect vs vehicle on fluid or chloride transport/delivery. Runcaciguat induced a slightly larger reduction in MAP and this was associated with reduced chloride and fluid delivery to the late proximal tubule, but not to the urine. Thus, the applied dose of runcaciguat may have shifted a fraction of chloride and fluid reabsorption upstream, i.e. from the distal nephron/collecting duct to the proximal convoluted tubules, possibly in response to lowering MAP.

Series 3 – High dose runcaciguat induces ENaC-dependent Na⁺ retention in mice.

Metabolic cage studies were performed in mice with a knockdown of ENaC in the connecting tubule and collecting duct (ENaC-KO) and their wild-type littermates (WT). As shown in Figure 10, runcaciguat (3 mg/kg by oral gavage) lowered urinary Na⁺ excretion in WT vs vehicle, an effect not detected in ENaC-KO. Runcaciguat tended to lower urine Cl⁻ excretion in both genotypes and did not significantly change fluid or K⁺ excretion. The same dose of BAY 41-2272 had no significant effect on urinary excretion of fluid or electrolytes vs vehicle; although it numerically lowered Na⁺ excretion in WT but not ENaC-KO. An increase in ENaC-mediated Na⁺-reabsorption in response to a reduction in blood pressure can enhance K⁺ secretion in the same principal cells, but the resulting somewhat lesser flow rates are expected to reduce flow-dependent K⁺ secretion by intercalated cells [37], such that net K⁺ excretion remains unchanged.

Figure 10: Acute effect of high dose BAY 41-2272 or runcaciguat on renal ENaC-dependent Na⁺ excretion.

In non-fasted mice with renal knockdown of ENaC (ENaC-KO) and littermate wild type mice (WT), high dose BAY 41-2272, runcaciguat (each at 3 mg/kg) or vehicle were applied by oral gavage combined with a water load (25 μl/g body wt) and the urine quantitatively collected over 3 hrs in metabolic cages without access to food or water. Data were analyzed by 2-way ANOVA for factors drug and ENaC-KO (ENaC) followed by pairwise multiple comparisons (Holm Sidak method) to determine which groups were different. N=5–7 mice/group. * P<0.05 vs vehicle. # P<0.05 vs wild-type (WT). See text for result details.

Series 4 – Effects of high dose BAY 41-2272 and runcaciguat on a marker of kidney oxygenation in GK rats.

BAY 41-2272 or runcaciguat (each at 3 mg/kg by gavage twice daily) were administered to GK rats for 5 days and compared with vehicle application. The last dose was given 3–4 hrs before kidney harvest. The hypoxia marker pimonidazole was injected i.p. 2 hours prior to kidney harvesting. Plasma collected just before kidney harvest revealed mean concentrations of BAY 41-2272 and runcaciguat of 0.048±0.014 μM and 4.17±0.11 μM, respectively. The 2-way ANOVA indicated that both factors, drug and region, had significant effects on pimonidazole staining, while the interaction term was not significant. As shown in Figure 11, multiple pairwise comparisons (Holm Sidak method) indicated that reductively activated pimonidazole staining was higher in outer medulla vs inner cortex vs outer cortex, consistent with the expected lower oxygen tension in the former. Moreover, pimonidazole staining was significantly higher in runcaciguat vs vehicle (P<0.05). Mean values for pimonidazole staining were numerically higher in BAY 41-2272 versus vehicle treated rats, but this did not reach statistical significance (P=0.078).

Figure 11: Effect of high dose BAY 41-2272 and runcaciguat on marker of kidney oxygenation.

BAY 41-2272 or runcaciguat (each at 3 mg/kg by gavage twice daily) were administered to GK rats for 4 days plus in the morning of the next day, which was 3–4 hrs before kidney harvest, and compared with vehicle application. Pimonidazole was injected 2 hrs before kidney harvest, and reductively activated pimonidazole staining was used as a hypoxia marker. A) Representative images of kidney sections, including outer cortex (OC), inner cortex (IC), outer medulla (OM) and inner medulla (IM). B) Quantitative analysis of pimonidazole staining. 2-way ANOVA (drug and kidney region) was performed followed by multiple pairwise comparisons (Holm Sidak method). Across all 3 groups, the following order of staining intensity was observed: OC<IC<OM=IM; significance indicated by *P<0.05. Across all regions (interaction term >0.05), staining intensity was higher in runcaciguat vs vehicle (# P<0.05). N=5 rats/group. See text for result details.

DISCUSSION

The second messenger, cGMP, influences blood pressure (BP) and renal hemodynamics. It is locally formed by sGC, the activity of which can be enhanced by sGC “stimulators” or “activators”, which target the enzyme in two different redox states, the nitric oxide (NO)-sensitive reduced enzyme and the NO-insensitive oxidized enzyme, respectively [1]. Little is known about the relative effects of these agents on BP versus renal hemodynamics, particularly in diabetes mellitus. The current study shows that BAY 41-2272 and runcaciguat induced similar dose-dependent reductions in MAP in non-diabetic WF as well as in STZ-WF, a model of T1DM, and in GK, a model of non-obese T2DM. Across the 3 strains, the same highest dose of BAY 41-2272 and runcaciguat lowered MAP by 20–30%, with responses being modestly greater in response to runcaciguat vs BAY 41-2272 (28.0±1.3 vs 22.4±1.3 %, P<0.005). In non-diabetic WF, BAY 41-2272 and runcaciguat increased GFR for doses that caused only small MAP reductions (<10%). In comparison, for MAP reductions ≥10%, BAY 41-2272 and runcaciguat reduced GFR in non-diabetic WF. Thus, the data indicated a potential biphasic effect of both compounds on GFR in non-diabetic WF. While the GFR increasing effect of the 2 compounds at doses with small effects on MAP was preserved in STZ-WF, GK did not show such a GFR-increasing effect. Moreover, while in WT and GK a greater reduction in MAP lowered GFR, the latter was better preserved in response to both compounds in STZ-WF. As a consequence, higher doses of both compounds lowered glomerular hyperfiltration in GK, but less so in STZ-WF. These results indicated that the GFR response to both compounds can be biphasic relative to MAP response and affected by the rat and diabetes model.

The correlation of MAP with FF (Figure 4) has implications for the determinants of ΔGFR, which, as defined by Brenner-Deen, include the filtration pressure ΔP, RBF, the ultrafiltration coefficient Kf, and plasma oncotic pressure. ΔGFR and ΔFF each vary inversely with ΔMAP. Therefore, ΔGFR and ΔFF are positively correlated. When ΔRBF is the only determinant of ΔGFR, there cannot be a positively correlation between ΔGFR and ΔFF. Therefore, ΔRBF cannot be the only determinant of ΔGFR in these experiments and changes in ΔP and/or Kf are also required to fully account for non-zero ΔGFR. sGC activation or stimulation may directly dilate renal resistance vessels, like the afferent arteriole, which would be consistent with the established vasodilator effects on the afferent arteriole of atrial natriuretic peptide (ANP), which acts through cGMP generated by particulate GC [38–40]. Indirect effects on the afferent arteriole could also include a proposed role of cGMP through tubuloglomerular feedback [41]. ANP has also been proposed to constrict the efferent arteriole [38, 40, 42], which may increase FF, but not all studies found constrictor effects of ANP on the efferent arteriole [39] and the role of cGMP in this effect has not been established. Moreover, higher doses of ANP slightly dilated the efferent arteriole [38]. Kf has been shown to be increased in response to ANP [43] and reduced during NOS inhibition [44–47], consistent with a potential increase in Kf by cGMP or sGC agonist-induced cGMP. In the micropuncture studies, both compounds lowered MAP, as intended, but had no effect on GFR, SNGFR, PGC, or PT. These studies exclude reciprocal changes in ΔP and Kf that might nullify each other’s effect on GFR such as occurs with mild pressor doses of Ang II [48]. These studies provided no information about the determinants of ΔGFR in situations where ΔGFR is non-zero.

Notably, the two compounds only increased GFR in response to smaller doses and MAP reductions <10% in non-diabetic WF and STZ-WF, which both had lower basal RBF and GFR values than GK rats. Further studies are needed to test the hypothesis that GK rats may have a higher basal sGC or cGMP tone compared with the other two models, which could contribute to higher basal RBF and GFR but blunt the response to additional stimulation/activation of sGC. On the other hand, at least one other stimulator of vasoactive sGC in the kidney (NO derived from NOS 1) is constitutively over-active in hyperfiltering STZ-diabetic rats [49]. Adding another agonist would not normally have greater effect in a system with heightened levels of endogenous agonist to begin with but the findings could be reconciled if the STZ kidney were primed with excess capacity to make sGC rather than excess NO.

On the other hand, higher doses of BAY 41-2272 or runcaciguat induced a stronger MAP reduction, which forces the kidney to autoregulate. In non-diabetic WF rats, the lower GFR values in response to MAP reductions ≥10% were associated with lower values of FF in response to both compounds and, in response to runcaciguat, also with unchanged RVR and reduced RBF, indicating that runcaciguat, but not BAY 41-2272, may have impaired renal autoregulation. While changes in NO availability have been shown to induce parallel changes in afferent and efferent arteriolar resistances, predominating effects on the efferent resistance have been reported [45, 50, 47], which may contribute to the observed reduction in FF in response to both sGC agonists. A new study found in the presence of oxidative stress (induced by ODQ) that a sGC activator induced glomerular cGMP levels, dilated isolated glomerular arterioles, and increased cGMP and RBF in the isolated-perfused kidney [51]. Further studies are needed to follow up on the hypotheses that sGS activation or stimulation can directly dilate the afferent arteriole, but that this can impair renal autoregulation when blood pressure drops more strongly (more prominent for sGS activator?), while the efferent vasodilation induced by higher doses of BAY 41-2272 or runcaciguat lower FF and further reduce GFR.

The current results are similar to previous studies using the sGC stimulator cinaciguat in Wistar rats, in which a small increase in RBF was reported at doses that only slightly reduced blood pressure, whereas with larger reductions in blood pressure RBF was overall maintained, although it markedly fell in ~25% of animals [52]. In studies conducted in dogs with heart failure, cinaciguat [53] and BAY 41-2272 [54] also induced some level of renal vasodilation. Moreover, a reduction in FF was also observed with higher doses of cinaciguat in Wistar rats and the authors proposed an efferent preponderance for cinaciguat [52].

In all 3 groups of rats, a dose of BAY 41-2722 or runcaciguat that reduced MAP by ~10% was close to neutral with regard to the GFR response. This was observed in whole kidney inulin clearance studies in non-diabetic WF, STZ-WF and in GK, and confirmed by micropuncture with SNGFR measurements on the single nephron level as performed in non-diabetic WF and GK (GFR appeared a bit more sensitive to runcaciguat in Series 1 vs Series 2). Moreover, at doses that only had small effects on MAP, both compounds were neutral with regard to glomerular capillary pressure and albuminuria. Neutral responses were also observed with regard to the hydrostatic pressure in peritubular capillaries (P_PTC) in response to both compounds in WF, and in response to BAY 41-2272 in GK. The results of the clearance studies depicted in Figure 4 also indicated that MAP reductions in response to BAY 41-2272 of about 10% marked the point in non-diabetic WF at which any further fall in MAP was associated with a decline in FF, potentially reflecting drug-induced lessening of efferent arteriolar resistance.

cGMP can modulate transport processes on the level of the tubule or collecting duct segment. In general, NO and cGMP are expected to lower NaCl reabsorption, as particularly proposed for the proximal tubule, thick ascending limb and the collecting duct [55–60]. However, the performed whole kidney clearance studies did not provide evidence that either of the compounds induced a robust increase in fractional excretion of fluid, sodium, chloride or potassium on the whole kidney level at any of the doses in any of the 3 rat groups. A reduction in fractional excretion in response to BAY 41-2272 or runcaciguat, mostly observed with application of the highest doses, could be secondary to the blood pressure effect. In micropuncture studies, doses of BAY 41-2272 or runcaciguat that reduced mean MAP by only 2–10%, did not significantly alter whole kidney fractional excretion of fluid or chloride.

A closer look at the segmental tubular transport processes, however, indicated potential subtle effects. Both BAY 41-2272 and runcaciguat may have enhanced chloride reabsorption in segments upstream of the early distal tubule. In GK, this may include runcaciguat-induced stimulation of fluid and chloride reabsorption upstream of the late proximal tubule. In WF this may include runcaciguat and BAY 41-2272 induced stimulation of chloride reabsorption somewhere between the late proximal and early distal tubule, potentially in the water impermeable thick ascending limb. This could in part reflect the influence of a small reduction in blood pressure and the activation of neurohumoral adaptations and the reversal of pressure natriuresis effects on the proximal tubule and loop of Henle [61]. In mice, the inhibitory effect of high concentrations of angiotensin II on proximal tubular reabsorption was mediated by the NO/cGMP pathway, i.e., direct effects of sGC agonists are unlikely to explain a stimulation of proximal tubular reabsorption [62]. In contrast, in human proximal tubules, the NO/cGMP pathway has been implicated in the angiotensin II-induced stimulation of reabsorption [62].

In contrast to more proximal segments, moderate doses of BAY 41-2272 and runcaciguat that had only a small effect on blood pressure may have had a minor inhibitory effect on chloride reabsorption downstream of the early distal tubule since urine excretion was not changed. While this would be consistent with the documented inhibition of NaCl reabsorption in the distal nephron by the NO/cGMP pathway [55, 57, 58, 60], the presented evidence is rather indirect.

A robust reduction in blood pressure is expected to cause sodium retention via ENaC in the distal nephron and collecting duct. To further explore this issue, the acute effects of one relatively high dose of BAY 41-2272 or runcaciguat was determined on urinary sodium excretion using metabolic cage studies in mice with a knockdown of ENaC in the connecting tubule and collecting duct (ENaC-KO) and their wild-type littermates (WT). The key finding was that runcaciguat reduced urinary sodium excretion, an effect only observed in WT mice. The authors speculate that runcaciguat lowered blood pressure in both genotypes due to its vascular effect, which would be expected to induce compensatory renal NaCl retention. Based on the rat studies, the effect on blood pressure was possibly stronger for runcaciguat versus BAY 41-2272. Renal Na⁺ retention involves ENaC in the connecting tubule and collecting duct and, as a consequence, this effect induced by runcaciguat was blunted in the absence of ENaC. No clear water retention response was seen with BAY 41-2272 or runcaciguat, probably because the mice were water loaded. Thus, the studies revealed an antinatriuretic effect of high dose runcaciguat that involves activation of renal ENaC and likely is secondary to blood pressure lowering.

The clearance studies in GK indicated that higher doses of BAY 41-2272 and runcaciguat may lower GFR due to a reduction in FF, while RBF is relatively well maintained due to intact renal autoregulation (Figures 2–4). Theoretically, a reduction in GFR and thus transport work in the presence of unchanged blood flow may improve kidney oxygenation. It was observed, however, that pretreatment of GK with relatively high doses of the compounds (3 mg/kg twice daily) for 4.5 days, with the last dose given 3–4 hrs prior to kidney harvest, did not lower reductively activated pimonidazole staining as an indicator of renal hypoxia. In fact, runcaciguat increased the pimonidazole staining, and BAY 41-2272 tended to increase this marker. Based on measurements of plasma concentrations at kidney harvest, it is likely that the applied dose of runcaciguat induced larger BP lowering effects than BAY 41-2272 and larger BP effects than those observed in the clearance studies, thereby overwhelming the capacity for renal autoregulation and dropping RBF. The runcaciguat concentrations at the time of harvest were ~ 4 μM and thus ~25 times higher than the reported IC50 for inhibiting constriction of isolated porcine arteries [22], suggesting close to maximum vasodilation. In comparison, the plasma concentration for BAY 41-2272 was only ~0.05 μM. Considering the similarly active metabolite of BAY 41-2272, which previously was found to have twice the plasma concentrations of BAY 41-2272, the estimated total active concentration at kidney harvest was about 0.15 μM and thus half of the IC50 for inhibiting constriction of isolated rabbit arteries [21]. An oral dose of BAY 41-2272 of 3 mg/kg given to anesthetized Wistar rats has previously been shown to reduce blood pressure by up to ~35%, observed at 40 min after application, and subsequent recovery to ~15% reduction at 2 hrs after giving the compound [21]. Considering that in the current study blood and kidneys were harvest 3–4 hrs after giving the last dose, the data suggest robust blood pressure effects of both compounds in the renal hypoxia experiment that likely were stronger in response to runcaciguat versus BAY 41-2272. The results indicate that caution is warranted when sGC stimulators or activators are given at rather high doses. Further studies are needed to confirm the presented observations.

Importantly in this regard, the sGC stimulator-induced reduction in left ventricular and vascular fibrosis in angiotensin II-induced hypertension [63] and in subtotal nephrectomy [14] occurred in dosages with or without moderate blood pressure lowering efficacy. In hypertensive Dahl salt-sensitive rats, the sGC stimulator praliciguat improved markers of renal damage at a dose that had minimal effect on blood pressure [64]. Also in a clinical trial in HF patients with reduced ejection fraction (SOCRATES reduced: SOluble guanylate Cyclase stimulatoR in heArT failurE Studies), the sGC stimulator vericiguat induced a dose-dependent reduction in the fluid retention marker NT-proBNP and a trend for lesser CV deaths and HF hospitalizations, while systolic and diastolic BP were not significantly reduced [65]. As a consequence, the authors hypothesized that a sGC stimulator may exert favorable cardiac effects without sizeable changes in blood pressure [65]. In accordance, a new study provided evidence that the sGC activator runcaciguat can induce cardio-renal protection at doses which did not reduce blood pressure and is effective in hypertensive as well as diabetic and metabolic CKD models [66]. In other words, there possibly is a therapeutic window for sGC agonists with little treatment-emergent hypotension and its potential negative consequences. Furthermore, a moderate blood pressure lowering effect of sGC agonists in hypertensive states has the potential to protect the kidney in the long term. Moreover, chronic effects of sGC agonists, which have not been tested in the current study, also have the potential to lower albuminuria and protect the kidney by improving inflammation, fibrosis and endothelial dysfunction [1].

In summary, the sGC stimulator BAY 41-2271 and the activator runcaciguat induced qualitatively similar effects on blood pressure and renal hemodynamics and transport, with a greater potency for runcaciguat vs BAY 41-2272 for lowering blood pressure. In non-diabetic WF, both compounds induce a biphasic effect on GFR with a small increase at doses that lowered MAP <10%, while a fall in MAP ≥10% lead to a progressive reduction in GFR. The latter was associated with a fall in FF, potentially reflecting efferent arteriolar dilation, and, in response to runcaciguat, also due to impaired autoregulation of RVR and RBF. These MAP-dependent responses in renal hemodynamics were altered in models of diabetes, such that the highest doses of BAY 41-2272 and runcaciguat, which lowered MAP by 20–30%, reduced glomerular hyperfiltration in the type 2 GK diabetes model but not in the type 1 STZ-diabetes WF model. Overall there was a strong correlation between changes in GFR and FF for both compounds across the 3 groups of rats. Application of one high dose of runcaciguat (3 mg/kg p.o.) in mice induced an ENaC-dependent antinatriuretic effect, and twice daily application of this dose over 5 days in GK rats increased reductively activated pimonidazole staining in the kidney as an indicator of hypoxia, both responses likely reflecting the consequence of a robust blood pressure lowering capacity when given at high doses. Doses of the compounds that reduced MAP by ~10% (0.2–2 mg kg⁻¹ h⁻¹ or ~3–30 μg kg⁻¹ min⁻¹) were relatively neutral with regard to renal hemodynamics, including glomerular capillary pressure, as tested by micropuncture in non-diabetic WF and type 2 GK rats. These doses left fractional excretion of fluid and chloride unaltered on the whole kidney level, while micropuncture studies indicated potential minor shifts in chloride reabsorption along the nephron.

Data Availability Statement

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.