Natriuretic peptide analogues with distinct vasodilatory or renal activity: integrated effects in health and experimental heart failure

Abstract

Aims

Management of acute decompensated heart failure (ADHF) requires disparate treatments depending on the state of systemic/peripheral perfusion and the presence/absence of expanded body–fluid volumes. There is an unmet need for therapeutics that differentially treat each aspect. Atrial natriuretic peptide (ANP) plays an important role in blood pressure and volume regulation. We investigate for the first time the integrated haemodynamic, endocrine and renal effects of human ANP analogues, modified for exclusive vasodilatory (ANP-DRD) or diuretic (ANP-DGD) activities, in normal health and experimental ADHF.

Methods and results

We compared the effects of incremental infusions of ANP analogues ANP-DRD and ANP-DGD with native ANP, in normal (n = 8) and ADHF (n = 8) sheep. ANP-DRD administration increased plasma cyclic guanosine monophosphate (cGMP) in association with dose-dependent reductions in arterial pressure in normal and heart failure (HF) sheep similarly to ANP responses. In contrast to ANP, which in HF produced a diuresis/natriuresis, this analogue was without significant renal effect. Conversely, ANP-DGD induced marked stepwise increases in urinary cGMP, urine volume, and sodium excretion in HF comparable to ANP, but without accompanying vasodilatory effects. All peptides increased packed cell volume relative to control in both states, and in HF, decreased left atrial pressure. In response to ANP-DRD-induced blood pressure reductions, plasma renin activity rose compared to control only during the high dose in normals, and not at all in HF—suggesting relative renin inhibition, with no increase in aldosterone in either state, whereas renin and aldosterone were both significantly reduced by ANP-DGD in HF.

Conclusion

These ANP analogues exhibit distinct vasodilatory (ANP-DRD) and diuretic/natriuretic (ANP-DGD) activities, and therefore have the potential to provide precision therapy for ADHF patients with differing pathophysiological derangement of pressure–volume homeostasis.

Open in new tabDownload slide

1. Introduction

Heart failure (HF) is a common, costly and potentially fatal condition. It affects ∼26 million people worldwide,¹ and in addition to being the leading cause of one in four deaths, is the foremost cause of hospitalization in people aged >65 years.² The course of chronic HF is punctuated by recurrent episodes of acute decompensation [termed acute decompensated heart failure, ADHF)]—an abrupt deterioration characterized by haemodynamic instability and pulmonary oedema often associated with decline in renal function. Based on symptoms, ADHF patients are classified according to position on ‘cold/warm’ and ‘dry/wet’ spectrums depending on the state of peripheral perfusion (a surrogate reflecting perfusion of vital organs) and absence/presence of expanded circulating and body–fluid volumes.³ Therapy must be tailored to the individual patient contingent upon the presenting profile of haemodynamic and/or volume derangement. Initial treatment of ADHF is aimed at resolution of life-threatening pulmonary oedema and/or tissue hypoperfusion and symptom relief. Management usually includes some combination of a loop diuretic such as furosemide with a vasodilator, inotrope, opiate, and/or ultra-filtration.⁴ However, there is no evidence-based consensus on the optimal management of ADHF.³ Treatments may ameliorate one aspect of pressure/volume derangement whilst impacting adversely on another. Indeed, loop diuretic use in ADHF relieves congestion at the expense of unfavourable neurohumoral activation and renal dysfunction,⁵ induces unwanted potassium loss, has some direct vasodilatory effects⁶ and is associated with conflicting data regarding mortality and morbidity benefit.⁷ Furthermore, many ADHF therapies may trigger or exacerbate hypotension which is strongly related to poor clinical outcomes. Positive inotropes may rescue patients from impending shock but are pro-arrhythmic and are associated with an overall increase in mortality. Thus, there is an unmet need for a menu of therapeutic agents allowing controlled vasodilation without excess diuresis at one end of the spectrum and effective diuresis without excess lowering of blood pressure at the other.

Atrial natriuretic peptide (ANP), one of a family of three structurally related hormones identified in mammals, plays an important role in pressure–volume homeostasis via its well-defined natriuretic/diuretic and vasodilating actions.⁸ The peptide is released in response to cardiac distension and other stimuli, and its actions oppose the deleterious fluid/sodium retention and vasoconstriction central to the pathophysiology of ADHF. The 28-amino acid (aa) form of mature native human ANP contains an intra-molecular 17-residue ring structure (formed by a disulfide bond between two cysteine residues at positions 7 and 23) that is required for bioactivity. ANP exerts most of its actions via the transmembrane guanylate cyclase-coupled natriuretic peptide receptor-A (NPR-A).⁸

Recently, we developed analogues of ANP which have the potential to provide precision therapy for ADHF patients with differing pathophysiology.⁹ Based on evaluation of structure–function relationships of a range of ANP mutants in anaesthetized rats,⁹ the specific aa residues within the ring structure responsible for vasodilatory and diuretic functions of ANP were identified. Several constructs were then engineered based on the human ANP scaffold to possess distinct pharmacological profiles—one with exclusive vasodilatory effects without diuresis (designated ANP-DRD) and another with diuretic effects without vasodilation (ANP-DGD). As with ANP, both analogues were shown to activate the NPR-A receptor and generate the intracellular second messenger cyclic guanosine monophosphate (cGMP).⁹ In the present study, we investigate for the first time the integrated haemodynamic, endocrine and renal effects of these two ANP analogues, ANP-DRD (vasodilatory) and ANP-DGD (diuretic), in both normal and ADHF settings using our established ovine model.^10,11 The parent ANP molecule was used as a comparison.

2. Methods

2.1 Peptide design and synthesis

Peptides ANP and analogues ANP-DRD and ANP-DGD, engineered to express exclusive vasodilatory and diuretic effects, respectively,⁹ were administered in the present study. ANP-DRD incorporated 3aa residue replacements within the conserved (17aa) ring structure of human ANP, comprising G9D, G10R, and G20D substitutions, whereas analogue ANP-DGD contained 2aa residue exchanges comprising G9D and G20D substitutions (Figure 1).

All peptides (ANP, ANP-DRD, ANP-DGD) were synthesized by Fmoc-based solid-phase peptide synthesis using an automated microwave peptide synthesizer (CEM, Matthews, NC, USA). N,N-dimethylformide (DMF) was the general solvent used throughout synthesis. C-terminal aa was loaded to Cl-MPA ProTide resin (CEM, Matthews, NC, USA) in the presence of 1 M N,N-diisopropylethylamine (DIEA) and 0.125 M potassium iodide. Subsequent coupling of aas (0.2 M) was performed using 0.5 M N,N-diisopropylcarbodiimide (DIC) as activator and 0.1 M DIEA in 2 M Oxyma as activator base. N-Fmoc deprotection was achieved with 10% w/v piperazine in ethanol: N-methyl-2-pyrrolidone (NMP) (1:9), and 0.1 M Oxyma added to minimize aspartimide formation during synthesis. Fmoc-Asp(OtBu)-(Dmb)Gly-OH dipeptide was also used where Asp-Gly was present in the sequence to avoid aspartimide formation.

Cleavage of the synthesized peptides from resin was performed at room temperature for 3 h in a cocktail of trifluoroacetic acid (TFA)/triisopropylsilane (TIS)/water/dioxa-1,8-octane-dithiol (DODT) (92.5/2.5/2.5/2.5). After cleavage, the peptides were precipitated using cold diethyl ether and resuspended in 0.1% v/v TFA solution for subsequent purification.

Purification of peptides was performed on Jupiter^® Proteo (250 × 21.2 mm) reverse-phase column (Phenomenex, Torrence, CA, USA) with gradient adjusted using the following solutions: (i) 0.1% v/v TFA and (ii) 0.1% v/v TFA in 80% v/v acetonitrile (CH₃CN). Purified peptides were lyophilized and disulfide formation induced through air oxidation at room temperature for 30 h at 25 µM purified peptide in 20 mM Tris–HCl pH 8.5, 10% v/v CH₃CN solution. The successful formation of a disulfide bridge was ascertained through the change in elution volume on reverse-phase column and the reduction of 2 Da of observed mass by mass spectrometry. Peptides were dissolved and reconstituted in 0.9% saline prior to administration.

2.2 In vivo studies

Studies were approved by the Animal Ethics Committee of the University of Otago-Christchurch (#C8/17A) (in accordance with the New Zealand Animal Welfare Act 1999, and Amendment No. 2 [2015]) and conform to current NIH guidelines (8th Edition).

2.2.1 Surgical preparation

Sixteen Coopdale ewes (4–6 years; 45–62 kg; Canterbury, New Zealand) were instrumented under general anaesthesia [induced by intravenous diazepam 0.5 mg/kg/ketamine 4 mg/kg and maintained with isoflurane (2%) inhalation] and received approved peri-/post-operative antibiotics (intravenous cephazolin 20 mg/kg and oxytetracycline 6.6–20 mg/kg) and analgesia (intravenous carprofen 4 mg/kg; subcutaneous buprenorphine 0.005–0.01 mg/kg; intercostal bupivacaine 0.5%/lignocaine 2% during thoracotomy). Adequacy of anaesthesia was monitored via serial measurement of heart rate (HR), blood pressure, capillary refill time, mucous membrane colour, palpable blink response, jaw tone, anaesthetic gas analysis, oxygen saturation, and end-tidal CO² levels.

2.2.1.1 Normal group

Eight sheep were instrumented via a 5 cm neck incision. The left carotid artery was cannulated (16G Angiocath: Becton Dickinson, USA) for measurement of mean arterial pressure (MAP) and HR, and a polyethylene catheter was placed in the jugular vein for blood sampling and recording of right atrial pressure (RAP). A Swan-Ganz thermodilution catheter (Edwards Life Sciences, USA) was placed in the pulmonary artery via the jugular vein for the measurement of cardiac output (CO).¹² Animals recovered for at least 5 days before commencing the study protocol.

2.2.1.2 Heart failure group

Eight sheep were instrumented via a left lateral thoracotomy.¹³ Polyvinyl chloride catheters were inserted in the left atrium for blood sampling and left atrial pressure (LAP) determination; Konigsberg pressure-tip transducers were inserted in the aorta to record MAP and into the apex of the left ventricle (LV) to measure HR, LV and dP/dt(max); a Swan-Ganz thermodilution catheter was placed in the pulmonary artery to measure CO, and a 7-Fr His-bundle electrode stitched sub-epicardially to the left ventricular wall for subsequent pacing. Animals recovered for at least 14 days before induction of HF by 7 days of rapid left ventricular pacing (∼220 b.p.m.)¹⁰ followed by commencement of the study protocol (with pacing maintained throughout).

In both groups, a bladder catheter was inserted per urethra for timed urine collections. During experiments, all animals were housed in metabolic cages in air-conditioned, light-controlled rooms and received a diet of Lucerne chaff and food pellets (providing ∼75 mmol sodium and 150 mmol potassium/day) with free access to water.

2.2.2 Study protocol

On four separate study days, each sheep received in balanced random order, intravenous infusions (via the Swan Ganz catheter) of increasing doses of ANP (2.5, 5, and 10 pmol/kg/min, for 90 min at each dose), ANP-DRD (2.5, 5, and 10 pmol/kg/min, for 90 min/dose), ANP-DGD (25, 50, and 100 pmol/kg/min, for 90 min/dose), and a vehicle control (0.9% saline). Doses were based on previous experience administering ANP in our ovine models^14,15 and the relative bioactivity of the analogues to ANP in anaesthetized rats.⁹

Haemodynamic measurements [HR, MAP, RAP/LAP, CO, dP/dt(max), and calculated total peripheral resistance (CTPR = MAP/CO)] were recorded using an on-line data acquisition system (PowerLab Systems; ADInstruments, Dunedin, New Zealand) at 15-min intervals in the hour preceding the first infusion (baseline), and at 15, 30, 60, and 90 min after commencement of each infusion and following cessation of the highest dose. All measurements were made with the sheep standing quietly in the metabolic cage.

Blood samples were drawn at 30 min and immediately preceding the first infusion (baseline), and at 30 and 90 min after commencement of each infusion and following cessation of the highest dose. Samples were taken into tubes on ice, centrifuged at 4°C, and stored at −20°C before assay for ANP, B-type natriuretic peptide (BNP), cGMP, plasma renin activity (PRA), aldosterone, and endothelin-1.^10–15 Analogues ANP-DRD and ANP-DGD were measured as per ANP. For each hormone, all samples from individual animals were measured in the same assay to avoid inter-assay variability. Packed cell volume (PCV) was measured with every blood sample taken. Note, PCV levels exhibited by these sheep in this laboratory setting tend to be in the lower ranges due to a combination of blood depletion (via serial blood sampling for hormone/biochemical assay) and haemodilution (via administration of intravenous fluids during line flushing, drug administration and measurement of CO by bolus thermodilution).

The natriuretic peptide concentration of each infusate was measured to calculate the infusion rate of the peptide, and metabolic clearance rate (MCR) of each peptide was calculated as: MCR = measured infusion rate/plateau-baseline level.

Blood samples for measurement of plasma sodium, potassium, and creatinine concentrations were taken at baseline and at 90-min intervals throughout the study. Urine volume and samples were collected every 90 min for the measurement of urine cGMP, sodium, potassium, and creatinine excretion.

At study end, animals were euthanased via IV pentobarbital overdose (Pentobarb 300; 100 mg/kg).

2.3 Statistical analysis

Results are expressed as mean ± SEM. To test for baseline differences between normal and HF sheep (effect of pacing), baseline data from each state (mean of measurements within the hour pre-treatment) were compared using unpaired Student’s t-tests.

To test for effects of treatment, control, ANP, ANP-DRD, and ANP-DGD study limbs (in normal and HF states separately) were compared with repeated-measures analysis of variance (ANOVA) (SPSS version 25; SPSS Inc., Chicago, USA). Significance was assumed at P < 0.05.

3. Results

Rapid left ventricular pacing for 7 days induced the haemodynamic, endocrine, and sodium-retaining hallmarks of congestive HF,^10,11 with significant declines in MAP and CO, increases in atrial pressure (all P < 0.001 vs. normal sheep), activation of multiple hormone systems [plasma ANP, BNP, cGMP (all P < 0.001), PRA, aldosterone (both P < 0.01), urine cGMP (P < 0.001), and reduced sodium excretion (P < 0.05)] (Figures 2–5).

As expected, incremental infusions of ANP and analogues ANP-DRD and ANP-DGD dose-dependently increased plasma concentrations of each peptide (all P < 0.001 in both states) (Figure 2). Achieved plasma levels of ANP-DRD tended to be greater than those of ANP at each of the three doses in both normal (ANP 41 ± 2, 70 ± 11, and 140 ± 18 pmol/L; ANP-DRD 87 ± 16, 133 ± 18, and 238 ± 29 pmol/L; P < 0.001) and HF settings (ANP 346 ± 23, 397 ± 41, and 528 ± 50 pmol/L; ANP-DRD 430 ± 47, 560 ± 78, and 733 ± 103 pmol/L; not significant) despite comparable measured infusate concentrations. Accordingly, the calculated MCR was lower for ANP-DRD than ANP (in normal sheep 2.63 vs. 5.19 L/min, P < 0.05; in HF 1.79 vs. 3.20 L/min, 0.1>P > 0.05). The MCRs of both ANP and ANP-DRD tended to be lower in HF than in normals. The calculated MRC for analogue ANP-DGD (normal 3.22 L/min; HF 1.79 L/min), administered at a 10-fold higher dose, was also reduced compared to ANP (P < 0.05) and diminished in the HF state (P < 0.05).

Plasma levels of second messenger cGMP were significantly and dose-dependently raised by all three peptides (normal all P < 0.001; HF ANP and ANP-DGD P < 0.001, ANP-DRD P < 0.05) (Figure 2). Notably, however, the amount of cGMP generated by each peptide (nmol cGMP/pmol peptide) differed—with the cGMP/NP ratio for ANP-DRD significantly reduced compared to that of ANP (normal P < 0.01; HF P < 0.05), and reduced further still with ANP-DGD (both states P < 0.001) (Figure 2). Plasma concentrations of BNP, which shares receptors and degradative pathways with ANP, tended to increase following infusion of the ANP peptides (normal ANP P < 0.01, ANP-DRD P < 0.05, ANP-DGD P < 0.001; HF ANP-DGD P < 0.05) (Figure 2), especially with the 10-fold higher ANP-DGD doses.

As anticipated, infusion of ANP dose-dependently decreased arterial pressure, with the fall in normal sheep (P < 0.001; high dose −20 mmHg) twice that observed in HF (P < 0.001; −9.5 mmHg) where baseline pre-infusion pressures are significantly lower (Figure 3). Analogue ANP-DRD produced vasodilatory effects very similar to those of ANP in both normal (P < 0.001; −21.4 mmHg) and HF (P < 0.001; −10.3 mmHg) settings. Blood pressure falls induced by both peptides promptly returned to control levels over the 15–30 min following cessation of infusion. In stark contrast to both ANP and ANP-DRD, ANP-DGD had negligible effect on arterial pressure, despite being administered at a 10-fold higher dose (Figure 3).

In HF, a state characterized by global vasoconstriction, the hypotensive actions of ANP and ANP-DRD occurred in conjunction with significant declines in CTPR (ANP P < 0.05, ANP-DRD P < 0.001), together with reductions in LAP, with the latter sustained over the 90 min post-infusion period (both P < 0.001; Figure 3). Analogue ANP-DGD similarly reduced cardiac preload in these HF sheep (P < 0.001) but had no effect on CTPR. In normal animals, ANP-DGD appeared to reduce CO levels relative to control (P < 0.001) (Figure 3), resulting in an apparent increase in CTPR (P < 0.001) given blood pressure was unchanged. In HF, the ANP peptides had a neutral effect on CO and contractility (dP/dt(max)), and in normals, had no effect on HR (Table 1).

In normal sheep, infusion of ANP and ANP-DRD were associated with rises in PRA during the high dose only (ANP P < 0.05, ANP-DRD P < 0.01) (Figure 4) at which time vasodilatory effects were greatest (with falls in MAP of ±20 mmHg), with no significant changes in PRA compared to control noted in HF. During ANP-DGD administration, where blood pressure was unaffected, PRA concentrations remained stable for the duration of the treatment period but, interestingly, rose sharply post-infusion (P < 0.01) (Figure 4). Plasma aldosterone concentrations were not significantly altered by any of the ANP peptides in the normal state (despite PRA elevations), but tended to be reduced compared to control in HF, significantly so by ANP-DGD (P < 0.05) (Figure 4). No significant changes in plasma endothelin or electrolytes were noted for any treatments in either state (Table 1).

Packed cell volume, a sensitive indicator of NP-induced shift of fluid volume from intra- to extra-vascular spaces, was similarly raised by all three peptides in both normal and HF sheep relative to a gradual fall seen over the control days (Figure 4).

The renal effects of ANP included significant graded increases in urine cGMP concentrations in both normal (P < 0.01) and HF (P < 0.01) sheep compared to control (Figure 5), which in the latter setting marked by avid sodium retention, was associated with a ∼4-fold rise in urine output (P < 0.01) and 5.5-fold elevation in sodium excretion (P < 0.01) at the top dose. Analogue ANP-DGD produced comparable increments in urinary cGMP in both states (normal P < 0.001; HF P < 0.01) and in urine volume (∼4-fold rise; P < 0.05) and sodium excretion in HF (∼6.8-fold rise; P < 0.01) (Figure 5). Neither ANP nor ANP-DGD significantly altered urinary potassium. In contrast to ANP-DGD (and ANP), analogue ANP-DRD produced no significant renal effects in either normal or HF states (Figure 5; Table 1), apart from a small rise in urinary cGMP noted in normal sheep relative to control (P < 0.05). While there were trends for creatinine clearance to be raised by ANP and ANP-DGD (but not ANP-DRD) relative to control in both states (Table 1), this did not achieve statistical significance. Water intake was unaffected by any active treatment compared to control (data not shown).

4. Discussion

Treatment strategies in ADHF need to be individually tailored to haemodynamic and volume status. With current therapies limited through their inability to improve one aspect without influencing another, there is a demand for medications that differentially influence circulatory pressure and volume. While the NPs, with their actions to counter the harmful vasoconstriction and congestion often characteristic of ADHF, have obvious therapeutic potential in this setting, their use to date has been restricted by the frequent onset of symptomatic hypotension.¹⁶ However, the recent development of function-based ANP analogues allows differentiation and maximization of the beneficial properties of the NPs. The current study investigates for the first time the effects of two distinct functional analogues ANP-DRD and ANP-DGD, designed to exert exclusive vasodilatory and diuretic activities, respectively, in a large animal model of normal health and experimental HF. The results demonstrate that ANP-DRD produces reductions in MAP very similar to the hypotensive effects produced by ANP, but with little urinary response (in contrast to ANP), whereas ANP-DGD is associated with a marked diuresis/natriuresis in the setting of HF comparable to ANP, without the accompanying vasodilatory effects.

The NPs reduce blood pressure via several mechanisms including direct dilation of arterial resistance blood vessels (via cGMP-induced smooth muscle relaxation), increased vascular endothelial permeability (resulting in the shifting of intravascular fluid into the extravascular compartment) and through natriuretic/diuretic actions that reduce blood volume.¹⁷ In the current study, the falls in blood pressure induced by analogue ANP-DRD were comparable to those produced by the parent ANP peptide in both normal and HF states and occurred in conjunction with similar declines in total peripheral resistance (in HF) and rises in PCV (in both states). Reductions in cardiac preload in HF (whether induced by dilation of capacitance vessels or direct lusitropic actions),¹⁸ which may also have contributed to MAP decreases, were also equivalent between ANP and ANP-DRD. Notably, however, while ANP significantly increased excretion of sodium and water, ANP-DRD had negligible renal effect. Another notable difference was the lower plasma cGMP/peptide ratio exhibited by ANP-DRD relative to ANP—suggesting impaired cGMP generation by this analogue. These findings concur with early reports by Sridharan and Kini⁹ following equimolar doses of both peptides (ANP and ANP-DRD) in anaesthetised rats, where ANP-DRD displayed equipotent vascular effects yet evoked lesser cGMP rises and no change in urine volume. Structure/function studies led the authors to conclude that replacement of Gly (G) by Asp (D) residues at positions 9 and 20 lowered the analogue’s activation kinetics and potency at the NPR-A compared with ANP, while the G10R substitution (in the presence of D9 and D20) appeared to enhance the effect on MAP and reverse function selectivity from renal>vasoactivity to vasoactivity>renal.

Importantly, the hypotensive actions of ANP-DRD occurred without the significant reflex tachycardia often induced by other pharmacological vasodilator agents, indicating this peptide, like ANP, has sympatho-inhibitory actions and augments cardiac parasympathetic effects on HR¹⁹ (ultimately reducing cardiomyocyte oxygen demand). In addition, the mechanism of NP systemic vasodilation is reported to involve attenuation of sympathetic tone in the peripheral vasculature.

The NPs are also reported to play a crucial role in counterbalancing the renin–angiotensin–aldosterone system (RAAS),²⁰ and while PRA was increased relative to control in normals during the top ANP and ANP-DRD doses—likely in response to overwhelming afferent arteriolar baroreceptor activation following the prominent drops in arterial pressure at this dose (20–21 mmHg), plasma renin was not significantly elevated by the mid doses in normals (where MAP still fell markedly by ∼14 mmHg), and not at all during infusion of these peptides in HF, suggesting relative inhibition of renin release.²¹ Furthermore, plasma aldosterone levels were either unchanged (despite PRA rises at top dose in normals) or relatively suppressed compared to control (HF)—suggesting ANP-DRD, like ANP, directly inhibits aldosterone secretion at the adrenal gland.²²

Also of note, while ANP-DRD was not associated with any significant urinary response, it was, on the other hand, not accompanied by any deterioration in renal function. This was despite prominent concomitant falls in blood pressure (and therefore presumably renal perfusion pressure) evident with the top doses. This reno-protective effect of ANP-DRD might reflect unique actions of the NPs on the renal vasculature to sustain glomerular filtration pressure by both dilating preglomerular vessels and constricting efferent arterioles.²³

Vasodilators are one of the cornerstones of ADHF management, particularly when accompanied by raised blood pressure and are primarily used to ameliorate elevated filling pressures and/or left ventricular afterload.²⁴ While nitrovasodilators are most commonly employed for this purpose in ADHF, these agents have not been shown to significantly diminish mortality or hospital readmissions.²⁴ The development of the exclusive vasodilator ANP-DRD, with its ability to reduce vascular resistance and blood pressure without reflex increases in HR or declines in renal function (Graphical Abstract), may potentially offer a more attractive therapeutic profile over conventional vasodilators. It is also possible that this NP variant, like nesiritde,²⁵ may present tolerability and safety advantages such as fewer reported headaches than nitroglycerine²⁶ and fewer concerns relating to coronary perfusion and possible myocardial damage.²⁷

In contrast to ANP-DRD, analogue ANP-DGD exhibited no hypotensive activity in either normal or HF states, even at 10-fold higher concentrations. However, in HF on a background of fluid and sodium retention, ANP-DGD induced a significant natriuresis/diuresis similar to ANP. In the kidney, the NPs have direct effects via guanylyl cyclase-coupled receptors²⁸ to inhibit sodium/water reabsorption within the proximal tubule and collecting ducts,^29,30 and the renal actions of ANP-DGD in the present study occurred in conjunction with significant elevations in urinary cGMP (equal to that produced by ANP). The NPs also impact renal haemodynamics, and while we observed a tendency for glomerular filtration rate (GFR) to increase during infusion of ANP (as evidenced by the rise in measured CRCL—baseline vs. treatment mean: 85.3 ± 6.4 to 92.2 ± 7.2), a similar trend was also evident with ANP-DGD (84.9 ± 3.3 to 90.5 ± 5.0; with no rise seen during ANP-DRD 90.7 ± 6.4 to 88.4 ± 5.0). While this might be considered surprising given the lack of effect of ANP-DGD on the systemic vasculature, and therefore presumably also the renal vessels, the analogue might have modulated GFR through other NP-related actions to enhance glomerular permeability and mesangial cell relaxation (resulting in greater filter load of sodium and water),^29,31 to alter the distribution of intrarenal blood flow (thereby changing medullary haemodynamics and promoting natriuresis),³² and to counter the constricting effects of norepinephrine on the afferent arteriole³³ and angiotensin II (AngII) in mesangial cells.³⁴ Indeed, with regards to the latter, ANP-DGD was the only peptide to significantly reduce PRA in HF (relative to control), and therefore presumably AngII levels—a peptide that has a marked vasoconstrictor effect in the kidney (reducing renal blood flow and, to a lesser extent, GFR),³⁵ as well as direct effects in the proximal tubules to increase sodium reabsorption.²⁹ Furthermore, ANP-DGD was also the only peptide to significantly attenuate plasma aldosterone levels in HF (although ANP and ANP-DRD showed similar trends),²² countering the mineralocorticoid’s effects in the kidney to retain sodium and lose potassium.

While ANP-DGD induced an appreciable natriuresis in HF, there was no corresponding increase in urinary potassium seen with administration of the analogue. This potassium-sparing effect of the NPs, also exhibited by ANP-DGD, is unlike most diuretic drugs and is of clinical relevance given the risk of hypokalaemia with loop diuretics and the attendant proarrhythmic effects.³⁶ Other complications associated with current diuretics used in ADHF (especially during aggressive therapy) including reduced GFR and deleterious neurohormonal activation (especially of the RAAS),⁵ were also absent with administration of ANP-DGD.

Along with vasodilators, diuretics are a mainstay in the management of ADHF. Indeed, the majority of ADHF hospital admissions exhibit volume overload/congestion and will require and receive diuretic therapy. However, on a backdrop where approximately one-third of ADHF patients have pre-existing kidney disease³⁷ and a quarter incur significant worsening of kidney function during admission,³⁸ and where current diuretics may in fact worsen renal function and even promote acute kidney injury,³⁹ ANP-DGD, with its exclusive volume-eliminating effects unaccompanied by potassium depletion in combination with its RAAS-inhibitory actions and absence of any hypotensive activity (Graphical Abstract), has the potential to provide a safer and more effective alternative to those affected.

An important characteristic evinced by the ANP analogues is their reduced clearance from the circulation compared to ANP, with both analogues demonstrating significantly lower metabolic clearances of between 38% and 49% in both normal and HF states. This is in keeping with preliminary findings from in vitro competitive binding and degradation studies showing that ANP-DRD and ANP-DGD exhibit decreased binding to the natriuretic peptide clearance receptor (NPR-C) in Chinese hamster ovary cells as well as reduced degradation by the enzyme neprilysin compared to the parent ANP peptide (data not shown). While the findings show that these ANP variants demonstrate improved resistance to degradation relative to ANP, it also suggests that the impact of neprilysin inhibition to further augment concentrations of these analogues and their actions in vivo are likely to be attenuated.

These novel function-based ANP analogues (ANP-DRD, ANP-DGD) join a growing catalogue of designer peptides engineered from the native NPs through addition, deletion, or modification of aa residues with a view to optimizimg efficacy, specificity and/or resistance to degradation. These include chimeric ANP constructs such as CNAAC (a 27aa peptide consisting of the C-terminus and ring of ANP and the 5aa N-terminus of CNP) reported to exhibit more stability in blood and more potent and durable hypotensive effects than other NPs in addition to eliciting a diuresis,⁴⁰ and vasonatrin NP (VNP; consisting of CNP-22 plus the 5aa C-terminus of ANP) shown to induce venodilating actions like CNP in association with arterial vasodilating properties and enhanced renal actions.^41,42 In addition, MANP (Mutant-ANP; ZD100), with a 12aa extension to native ANP’s C-terminal which renders resistance to metabolism and enhanced blood pressure lowering, renal-enhancing and aldosterone-suppressing actions,⁴³ is now entering early clinical trials in human hypertension. Other NP chimeras have been synthesized with the goal to enhancing renal activity without inducing excessive hypotension, and include CD-NP (Cenderitide; a fusion between CNP-22 and the 15aa C-terminus of DNP), CRRL269 (an integration of the ring of urodilatin with the C- and N-terminals of BNP),⁴⁴ and ANX-042 [a blend of alternative spliced BNP’s (AS-BNP’s) 16aa C terminal and 26aa of native BNP]^41,42—with the latter currently being tested in patients with a history of HF and kidney dysfunction. While the clinical application of native NPs has been limited to date, the development of designer NP variants such as these and those assessed in the present study, which demonstrate functional selectivity/exclusivity and maximization of the beneficial properties of their donor peptides, present a unique opportunity to provide precision therapy in various cardiovascular/cardiorenal disease states.

4.1 Study limitations

While these in vivo studies have demonstrated the distinct vasodilatory and diuretic actions of the respective ANP-DRD and ANP-DGD synthetic molecules, they do not add any new observations as to the cellular mechanisms involved. Although the acute nature of the study may also be viewed as a limitation, intensive sampling and multiple doses and treatment arms restricts investigation duration with this design.

In conclusion, the present study demonstrates that ANP variants ANP-DRD and ANP-DGD differentially exert vasodilatory and diuretic actions, respectively, in experimental ADHF. The engineering of analogues with decoupled functionality potentially offers a unique treatment strategy which could fill the gap in current ADHF therapy by providing personalized care for ADHF patients with ends-of-the-spectrum clinical characteristics. With this ‘menu’ of NP-modified peptides, ANP-DRD may be useful in treating patients with normal/high blood pressure and acute left ventricular failure (pulmonary oedema) with little true overall volume expansion where volume redistribution is the ideal approach rather than volume depletion, whereas ANP-DGD may be useful in oedematous ADHF patients with low blood pressure and possibly renal compromise where further falls in blood pressure may be catastrophic. Not only may these analogues be less likely to have adverse effects on liver and kidney given their design is based on an endogenous neurohormone, but our findings indicate, that in addition to a reduced clearance rate, they also retain other favourable aspects exhibited by the original ANP molecule, including sympatho- and RAAS-inhibitory actions, and absence of accompanying reflex tachycardia and potassium depletion that is often attendant with current ADHF therapies.

Acknowledgements

The authors acknowledge the staff of the University of Otago-Christchurch Animal Research Area for animal care and the Endocrine Laboratory for hormone assays. The authors also thank Samuel Henry Kurniawan (National University of Singapore) who assisted with peptide synthesis and purification.

Conflict of interest: none declared.

Funding

This work was supported by the National Health Innovation Centre of Singapore [I2D-1612150] and Singapore-MIT Alliance for Research and Technology Centre [ING-000101 BIO].

References

Author notes