A State of Natriuretic Peptide Deficiency

Abstract

Measurement of natriuretic peptides (NPs) has proven its clinical value as biomarker, especially in the context of heart failure (HF). In contrast, a state of partial NP deficiency appears integral to several conditions in which lower NP concentrations in plasma presage overt cardiometabolic disease. Here, obesity and type 2 diabetes have attracted considerable attention. Other factors—including age, sex, race, genetics, and diurnal regulation—affect the NP “armory” and may leave some individuals more prone to development of cardiovascular disease. The molecular maturation of NPs has also proven complex, with highly variable O-glycosylation within the biosynthetic precursors. The relevance of this regulatory step in post-translational propeptide maturation has recently become recognized in biomarker measurement/interpretation and cardiovascular pathophysiology. An important proportion of people appear to have reduced effective net NP bioactivity in terms of receptor activation and physiological effects. The state of NP deficiency both entails a potential for further biomarker development and could also offer novel pharmacological possibilities. Alleviating the state of NP deficiency before development of overt cardiometabolic disease in selected patients could be a future path for improving precision medicine.

Graphical Abstract

Essential Points.

• Natriuretic peptides constitute a family of evolutionary conserved bioactive peptides with specific receptors

• An important proportion of people appear to have reduced natriuretic peptide bioactivity in terms of receptor activation and physiological effects

• Factors such as age, gender, race, genetics, body mass index, diabetes, and diurnal regulation can markedly reduce the natriuretic peptide concentrations in circulation

• Alleviating a state of natriuretic peptide deficiency before overt cardiometabolic disease in selected patients could be a future path for improving precision medicine

Natriuretic peptides (NPs) constitute a family of evolutionary conserved bioactive peptides with specific receptors (Fig. 1). Atrial NPs (ANPs) and B-type NPs (BNPs) are mainly produced in the heart. This review will focus on the NP system in relation to hormonal function reflected in circulating concentrations in the mammalian circulation, in other words, extracerebral physiology only (Fig. 2A). C-type NP (CNP) also belongs to the NP family and is, among other tissues, released from the vascular endothelium. It represents a different biology and possibly also a different pathophysiology. In mammals, the dominant NP hormone in circulation is ANP (1). ANP is produced and released by atrial (and ventricular) cardiomyocytes during more persistent stress conditions in response to transmural distending pressures, with additional contributions from an assortment of other chemical and physical stimuli. For instance, low oxygen tension and/or supply, such as cardiac hypoxia, is a stimulus for increased gene expression and peptide release to the circulation (2). In circulation, ANP and BNP exert overlapping biological effects via stimulation of the natriuretic peptide receptor (NPR) A. In contrast, CNP is released from the vasculature in response to laminar shear stress and proinflammatory mediators, and triggers signaling mainly through NPR-B. All 3 peptides bind to both NPR-A and NPR-B, albeit with affinities that essentially render mediation of the bioactivity of ANP and BNP specific to NPR-A and that of CNP to NPR-B (3). This has led to emerging drug development (vide infra). The third NPR-C binds and clears all 3 peptides. This receptor has also been shown to be involved in downstream signaling relevant for cardiovascular function.

Figure 1.

The 3 NPs (human sequences) and their receptors.

Figure 2.

(A) Cardiac ANP and BNP in mammalian physiology. Modified from Levin et al., New Engl. J. Med. (1998). (B) ProBNP expression in atrial and ventricular tissue. Note that proBNP is stored in normal atrial tissue while absent in ventricular myocytes. Taken from Goetze et al., FASEB J. (2004).

All 3 peptides are translated from mRNA into prepropeptides, from which the signal sequence is removed during translation. Thus, unprocessed prepropeptides do not occur as independent structures. The propeptides are then translocated from the Golgi network to the immature granules within the cells. Propeptides are then cleaved into several fragments by endoproteolysis, producing biologically active compounds. Hence, the terms ANP, BNP, and CNP refer to the bioactive, in other words, receptor binding, hormones. The unprocessed propeptides are, however, released from cardiac cells in a manner analogous to proinsulin release from pancreatic beta cells. This increasingly occurs in states of increased gene expression (disease), and the propeptides possess reduced bioactivity compared with the mature hormones. Other fragments from endoproteolytic maturation are N-terminal fragments. Finally, all 3 natriuretic propeptides undergo post-translational modifications, mainly O-glycosylation, prior to release (4, 5). The impact of these modifications on the validity of the assays is usually not reported in respect to clinical methods and often goes unmentioned. However, modifications will impact immunological measurement with a variable bias, usually as underreporting of the true circulating concentrations in vivo.

As mentioned, ANP and BNP are produced in the mammalian heart. In normal heart tissue, the dominant chambers of expression are the atria (1, 6). This is true for both ANP and BNP (Fig. 2B). In ventricular disease, however, the ventricular myocytes increase both ANP and BNP gene expression and become a major site of cardiac peptide production and release. For CNP, the endothelial cells lining the vasculature are one of the main sources of synthesis and release (7). Small amounts of CNP are also produced in the heart (8), but whether this expression is within the cardiac myocytes, local fibroblasts, or coronary vasculature remains unclear. Finally, the CNP system is expressed in both male and female reproductive organs, where the peptide in males is secreted to seminal plasma (9, 10).

In the circulation, NPs are eliminated by proteolytic degradation and/or NPR-C–mediated clearance. One major “elimination” enzyme is neprilysin (neutral endopeptidase EC 3.4.24), which is expressed in most organs. The kidneys contribute to NP clearance via high local degradation by neprilysin (which is found at its highest concentration in the kidney) and to a minor extent via filtration into the urine. Renal dysfunction comes with markedly elevated plasma concentrations of NPs, both proforms and bioactive forms reflecting both the volume expansion and subsequent increased cardiac release of NPs in addition to any accompanying loss of renal peptide clearance capacity. The half-life of the biologically active peptides is generally short, usually around 2 to 5 minutes (11, 12). For the proforms, stringent pharmacokinetic studies have not yet been performed. Thus, estimations of the half-lives are derived from stimulation experiments and thus include some degree of gene expression and cellular release patterns (13). For the N-terminal proBNP form (proBNP 1-76), the half-life is around 70 minutes; for the N-terminal proANP and proCNP fragments, the elimination phase is still unknown.

In the following, we review physiological features and factors affecting normal concentrations of NPs in circulation. Notably, the underlying theme for this review will be regulation in absence of overt cardiac disease where the peptides in general have been extensively characterized as clinical markers. In perspective, a state of NP deficiency driven by noncardiac factors may play a hitherto largely overlooked role in the early development of cardiac disease.

Molecular Forms of Natriuretic Peptides

NPs in plasma are quantitated by immunoassays. This entails that measurement is based on recognition of 1 or 2 well-defined epitopes within the primary peptide structures. For the sake of simplicity, the measurements are often named by the complete structure in question. For instance, it is often reported that the N-terminal fragment of the precursors is listed as “Nt-pro-NP,” such as, Nt-proANP, Nt-proBNP, or Nt-proCNP. However, it is often the case that the N-terminal fragment is not intact but rather processed into smaller forms. In addition, the intact proform of the hormone is also comeasured with these “N-terminal” assays, namely intact proANP, proBNP, or proCNP. For proBNP, this is of major relevance, as the fraction of unprocessed proBNP increases with the severity of the heart disease (14-16). Thus, the dominant form in plasma from patients with heart failure (HF) is not “Nt-proBNP” but, rather, intact proBNP (Fig. 3A). The plasma concentration is, however, still calculated from the molecular size of the calibrator peptide (for Nt-proBNP it is proBNP fragment 1-76), which causes analytical bias. Some of the peptides are also cleaved in circulation, where for instance N-terminal trimming by the enzyme DPP-IV has been shown for BNP (17, 18). The same is the case with the N-terminus for proCNP (19). It is thus always relevant to consider the method(s) used for measurement, as the different methods measuring different epitopes cannot be simplified to just the molecule used as calibrator. To complicate matters further, the propeptides are all highly glycosylated and to various degrees, which will affect detection.

Figure 3.

(A) General presentation of BNP gene expression (humans). The final products are considerably more heterogenous via post-translational modifications than shown here. (B) Comparison of proinsulin and proBNP as prostructures.

While measurement of the different forms is useful in the assessment of possible HF, it has been proposed that the ratio between the bioactive hormone vs the unprocessed proform might provide further information. This idea is largely in parallel with insulin production and secretion, where proinsulin is released in increasing amounts during beta-cell dysfunction and oncoming type 2 diabetes (20, 21). A diminished insulin/proinsulin ratio is thus an early quantitative measure of beta-cell dysfunction. For NPs, the same endocrine feature seems to occur, as HF with a high expression of NPs shifts the release from bioactive peptides to proforms with reduced, if any, activity, a feature that also partly explains the apparent paradoxical lack of effect of the immunoreactivity-detected peptides in HF (22) (Fig. 3B). In that context, detection of bioactive BNP 1 to 32 by use of quantitative mass spectrometry has also demonstrated very low concentrations in patients with symptomatic HF that only accounted for a small fraction of that measured by immunoassays (23-26). As mass spectrometry is not widely available for clinical use, development of clinical assays enabling detection of bioactive NPs will be important for providing essential insight into the extent of NP deficiency and help guide clinical decisions.

While the concept of using a ratio such as BNP/proBNP or Nt-proBNP/proBNP is appealing, it has not been pursued in the clinical setting, which to some extent may be driven by the relatively high cost per analysis and the degree of variation of the peptides and lack of standardized sampling conditions. Nevertheless, a more detailed molecular mapping of the peptide forms in health and disease could increase the information level for measurement beyond what is known today—and that includes advanced disease where the markers tend to lose their diagnostic power, for instance, critically ill patients just have “very high” plasma concentrations. Better prognostication might be achieved by a deeper analysis and characterization of the NP forms released from the heart according to differing type and severity of disease (27).

Briefly on Methods

Since NPs and their molecular precursors circulate in low picomolar concentrations, quantification in clinical practice is based on immunoassays. Immunoassay performance is critically defined by the characteristics of antibodies used, including their physical and chemical binding properties and their molecular specificity. Most antibodies, whether monoclonal or polyclonal, will bind to an epitope within the primary sequence consisting of 4 to 6 amino acids. Thus, the true measurement reports only on that epitope and not necessarily the larger structure for which the assay is calibrated. While this works well for clinical use of many hormone measurements, it is still important to emphasize that immunological measurement does not give a qualitative measure of potential changes in molecular maturation prior to release into circulation, nor does it really reflect the elimination phase. Crucially, an immunobased measure cannot readily be translated into bioactivity. This has been standard knowledge in endocrinology for a very long time but only more recently for the endocrine heart. As for most other endocrine systems, there seems to be a shift in the molecular forms toward larger and less processed forms when gene expression increases. Also, it is to be expected that the larger molecular forms will be less bioactive; for instance, intact proANP, proBNP, and proCNP will bind to the native receptors with reduced affinity or maybe with no binding at all. Finally, the elimination phase for larger forms is largely unknown. The extent of in vivo proteolytic proform activation in the circulation is also an open question.

A Few Words About HF

Measurement of NPs in the diagnosis of HF has been solidly established and is recommended by both the European Society of Cardiology and the American Heart Association (28, 29). The topic has been covered in multiple original research papers and reviews (30-33). In brief, measurement of NP and/or their proforms excluding CNP and proCNP is useful in early assessment of possible HF (Fig. 4A). The markers can efficiently exclude a diagnosis of HF but are less specific as an inclusive marker. However, as previously described, a subset of patients with confirmed HF or cardiac dysfunction have been reported to present with normal (34) or even low BNP concentrations (35-37), thus suggesting a state of NP deficiency that may leave these patients at risk for an incorrect diagnosis.

Figure 4.

(A) The first published data on ANP as a novel plasma biomarker in cardiovascular disease. Modified from Burnett Jr et al., Science (1986). (B) A simplified presentation of the 2 main hormonal systems in HF. Inhibition of the RAAS system is today a hallmark of HF therapy, whereas stimulation or replacement of NPs is still an area for further exploration. (C) Presentation of diagnostic performance for a given biomarker. Note that the so-called “gray” zone lies between A and B. For NPs, clinical measurement may reside in this zone up to 40% of all prescribed tests. (D) Genetic mutation in human ANP gene leading to a C-terminal elongated peptide form.

Some debate is still ongoing as to which patients should be screened. Also, the question of how to handle “confounding” comorbidities, such as renal disease and obesity, as well as other factors like age and pharmacotherapy, has still not been answered. The best proven diagnostic application for BNP and Nt-proBNP is in symptomatic (dyspneic) acute HF (38, 39). Dyspnea can be caused by numerous diseases that are often common comorbidities in the HF population (eg, chronic obstructive lung disease); plasma BNP and proBNP exhibit high diagnostic discrimination for acute HF even when such comorbidities are present (40). This underpins their utility and widespread adoption as diagnostic aids in the setting of unexplained dyspnea.

Since the first recommendations of measuring NPs in the clinical setting, increasing attention has focused on the spectrum of the HF phenotype defined by left ventricular ejection fraction with recent categorization into reduced, mildly reduced, and preserved ejection fraction. While NP measurement is used also in HF with preserved ejection fraction, the diagnostic cut-off concentrations may differ from those used for HF with reduced ejection fraction, at least in the subacute setting (41-43). It has also been suggested to monitor HF status and therapeutic efficacy by serial measurement of plasma NP concentrations. While this is already practiced in specialized settings, the formal evidence base for this practice remains to be fully established. In this context, therapy titration guided by serial measurements has been explored in multiple trials that have generated several strongly positive meta-analyses. The practice has been hard to implement, and some trials have failed to trigger differential pharmacotherapy deployment and therefore shown no difference between treatment strategies. This is likely to reflect the difficulty in achieving consistent clinical adherence to peptide-driven dosing algorithms. One major drawback is that certain drugs per se alter peptide concentrations yet still improve longevity, as exemplified by reports on patients treated with angiotensin-converting enzyme inhibitors (44) (Fig. 4B). Thus, as in all competent clinical care, temporal use of peptide measurement during disease progression cannot be a sole guide.

The “Gray” Zone

Reference intervals for NPs have been established, which most often include a correction for the mere effects of age and gender. From a diagnostic point of view, however, there is still a troublesome large “gray” zone for concentrations, where they are elevated when compared with reference intervals but still not diagnostic for HF (45). A substantial proportion of all ordered tests may yield results in this diagnostic gray zone when including patients seen in the primary health care setting. Other studies focusing on prognosis show that even in asymptomatic community-dwelling populations there is a gradient of risk for incident cardiovascular disease over the subsequent decade or longer, even for lower to higher sections of the reference range of plasma NP concentrations. Therefore, in the cardiological setting (and as for all clinical application of tests), a clear definition of when to measure is needed. Screening without indication can lead to concerns and time spent on trying to explain a concentration in the gray zone without a clear focus on the primary recommendation today, namely, to measure when suspecting HF. On the other hand, there may still be analytical issues that could improve the diagnostic usefulness of NP measurement, specifically what to measure. Given the complexity of NPs, evaluation of the molecular pattern could help to improve the clinical use of peptide measurement in the gray zone (Fig. 4C).

Genetics

Several mutations have been reported in humans for the ANP and BNP gene. For the ANP gene, mutations generally affect the processing of the peptide and lead to elongated forms of ANP. Mutations can also be located outside the exome and thus lead to altered gene expression. For either type of mutation, it is important to note that changes should be associated with clinical phenotype and disease; for the ANP gene, this will mostly be on blood pressure regulation, cardiovascular disease, and stroke. One mutation in the human ANP gene is located within exon 3 and changes the native stop codon, leading to an elongated ANP peptide by 2 C-terminal arginyl residues (46). The mutation frequency is estimated to be high, perhaps as high as 14% in the general population. The mutation is associated with increased risk of myocardial infarction and stroke (47-50) but seemingly not to hypertension (51, 52). From a pharmacological point of view, this mutation has attracted some attention. The mechanism behind the clinical findings has been suggested to be via the NPR-C, which as stated previously also acts as a clearance receptor (45). In addition, this mechanism may affect platelet aggregation and thus be a direct biological stimulus behind increased risk of myocardial infarction and stroke. Furthermore, a genetic variant that leads to slightly higher concentrations of circulating ANP is associated with lower blood pressure and protects against development of hypertension and cardiac remodeling (53, 54). Another ANP mutation has been reported in rare families with lone atrial fibrillation (55). This mutation comprises a frameshift mutation and leads to a C-terminally elongated peptide that is resistant to proteolysis and elimination (56) (Fig. 4D).

In relation to ANP processing, a mutation in an enzyme named Corin has been reported (57). This mutation causes reduced hormonal maturation of ANP and can be considered a “loss-of-function” mutation (58). Reduced activity of Corin is associated with hypertension and is prevalent in African Americans (59, 60). This mutation underscores the relevance of having functional ANP in the cardiovascular system and that changes in this hormone axis can lead to disease over time. Mutations in the human BNP gene have also been reported. In contrast to the ANP gene, mutations within the BNP gene are mostly located outside the exons, mainly in the promoter region. These mutations lead to altered gene expression rather than changes in peptide structure and receptor binding. One mutation in the promoter region is associated to increased gene expression and thus increased BNP release (61). While the BNP gene in normal physiology seems to be of less relevance, this promoter variant is still associated with lowered susceptibility to developing type 2 diabetes (62). This highlights the relevance of NPs in pathologies even outside the cardiovascular system (Fig. 5).

Figure 5.

Overview of factors involved in the state of NP deficiency.

For the CNP gene, knockout mice models display distinct phenotypes in bone formation (63, 64). In extension, a genotype of increased CNP expression in humans is associated with overgrowth and bone anomalies (65), whereas a loss-of-function mutation in the NPR-B gene leads to extreme short stature (66) These findings, including observations of short stature in humans displaying mutations within the highly conserved CNP ring (67), have led to the development of CNP as a novel therapy regimen for achondroplasia, the most common form of dwarfism (68, 69). This brings new hope to genetic disease in bone formation, but it also excludes mutational analysis as a model for possible effects on the cardiovascular system, as gross changes in stature and body composition complicate interpretation of whether changes in the CNP gene and its expression also contain cardiovascular phenotypes. Interestingly, a primary association between a genetically determined shorter height and an increased risk of coronary artery disease has been established (70).

Various mutations in the genes encoding the 3 receptors have provided further insight into the physiological role of NP signaling on cardiovascular and bone homeostasis. In accord with the well-described vasodilator and natriuretic effect of ANP and BNP, a functional deletion mutation of the 5′-flanking region of NPR-A, which reduces transcriptional activity, is associated with increased susceptibility to developing essential hypertension and left ventricular hypertrophy (71). Moreover, a missense mutation in exon 3 of NPR-A is associated with increased risk of essential hypertension and myocardial infarction (72). Bi-allelic loss-of-function mutations in the NPR-C gene are associated with a reduced Nt-proNP/NP ratio and higher plasma cyclic guanosine monophosphate, suggesting reduced clearance by the defective NPR-C. Increased signaling activity was evident from a phenotype characterized by tall stature, long digits, and variable connective tissue alterations (73). In addition to its clearance function, signal transduction through NPR-C also leads to cellular responses. One example of this is a nonsynonymous single nucleotide polymorphism in NPR-C that results in a substitution in the intracellular domain. This single nucleotide polymorphism is independently associated with diastolic dysfunction in the absence of altered circulating concentrations of NP (74).

Race, Age, and Sex

The impact of race on NPs in plasma has mainly been examined for the BNP system, as proBNP-derived peptides have become the global biomarkers of choice for cardiac disease. Most studies have evaluated Caucasians in health and disease, and the current “reference intervals” are largely based on such populations. However, it has been shown that African American people have lower circulating concentrations of Nt-proBNP than Caucasians (75-79) (Table 1). Hispanic people also display low concentrations, whereas the pattern in people of Asian origin is like that in Caucasians (77, 80, 81). It is still not known whether these differences reflect a causal effect of race per se or rather reflects acquired differences in metabolic regulation, body mass composition, or differences in lifestyle. However, it does raise the question of whether people of African descent may be more prone to a state of NP deficiency and accordingly cardiovascular disease.

Table 1.

Racial differences in NT-proBNP concentrations and use of antihypertensive medication

	Caucasian			African American
Study	Subjects (n)	NT-proBNP (pg/mL)	Antihypertensive medication (%)	Subjects (n)	NT-proBNP (pg/mL)	Antihypertensive medication (%)
Patel et al (79)	2101	80 (41-178)	50.8	2005	57 (27-134)	71.8
Yang et al (76)	993	63 (33-126)	—	584	39 (15-85)	—
Gupta et al (77)	7164	68 (36-124)	33	1973	43 (18-88)	54
Gupta et al (78)	969	32 (16-61)	16	1606	24 (10-52)	21

Values are median (25th to 75th percentiles). P < .001 for all between group comparisons, except Gupta et al (78) (P < .0001).

The associations of lower age and male sex with lower NP concentrations have been consistently documented in numerous publications. This has come about from the need for adequate data to define reference intervals for “normality” and thus better discrimination between health and disease. One major issue from these studies is the fact that the prevalence of cardiac disease is strongly correlated to age, which includes unrecognized disease. One example is diastolic dysfunction with stiffer cardiac chambers. Thus, whether shifts in plasma NP concentrations may be solely related to age is not trivial. Nevertheless, proANP-derived peptides in plasma increase with age while the effect of sex is less pronounced (82). For the BNP system, the increase as a function of age is well proven, as is the sex difference in people <50 years of age (83). Notably, sex is one of the most important determinants of NP concentrations in individuals without HF (84), which may—at least in part—be related to hormone status as a suppressive effect of testosterone on the NP system (85, 86). Unraveling the mechanisms underlying this NP deficiency in males is clinically relevant but also very challenging in humans, given the experimental limitations. Hence, preclinical studies may be utilized with the notion that a high degree of translation may be expected, given the very high conservation of the NP system across species. Interestingly, concentrations in women after menopause become more similar to those of men of the same age group. For the CNP system, plasma concentrations of Nt-proCNP are slightly higher in males than in females and increase in both genders after the age of 50 (87).

Biological Variation

NPs are subjected to diurnal variation. In animal studies, the variation is circadian and relates closely to the expression of local clock genes within the cardiomyocytes (88). In humans, this variation is measurable for proANP and proBNP, whereas there is still no report on the CNP system in circulation (89-93). In fact, this variation corresponds well to what some refer to as biological variation (usually around 20%), which is a circumstance that clinicians must consider when using the markers in a 24/7 clinical setting. However, the variation also allows for optimized diagnostics, as reference intervals could be established as a function of time of day. Chronobiology also has an impact on potential pharmacological intervention. Should certain drugs targeting the NP system be taken in the morning or the evening? In 2 studies, proANP concentrations are increased in the night while proBNP mostly peaks during the day (94, 95). If this not only reflects postural changes but also clock-related biology in humans, this can be used to target either system. For patients with HF, there are no valid data on circadian regulation of NPs. However difficult, this may prove worthwhile to pursue when optimizing medical treatment in HF.

Obesity

In the pursuit of defining reference values for NPs in blood and thereby optimizing their clinical use as cardiovascular biomarkers, it was noted early that body mass per se had a marked impact on their concentrations (80, 96). One of the more recent reports on the phenomenon came from the Dallas Heart study, where plasma concentrations were correlated with lean mass (97). This observation has been documented in most clinical studies, and body mass index, together with age and gender, remain cardinal factors to correct for in studies dealing with the diagnostic use of the peptides, albeit the mechanism behind this association is still debated (98). One facet seems to be changes in the molecular forms, as changes in glycosylation as a function of obesity (mainly close to the cleavage site for BNP) have been reported (99). Another feature is changes in metabolic clearance of the peptides (100). As obesity is a multifactorial measure, including lifestyle, metabolic disturbances, and hemodynamic changes, several other hypotheses have been proposed. The diabetes connection will be discussed separately. One explanation for the lower concentrations could also relate to factors relating to a sedentary lifestyle. This suggestion is supported by physiological experiments in healthy individuals, as mere bed rest for a relatively short period of time (days) markedly downregulates NPs in circulation (101). One feature of bed rest is lying down. In that position, venous return is, at least acutely, increased, which leads to increased secretion of ANP from the right atrium (102). However, if this stimulus is maintained, it is not clear whether an increased secretion of peptides is maintained or, perhaps, downregulated.

Another feature of obesity is the size of the heart vs the size of the body. As body mass index increases, the size of the heart largely remains constant. Thus, the heart as a pump must increase its output as a function of body size. In the earlier phases of obesity, the heart can perform through the Frank–Starling law of contraction, but over time the heart becomes more fibrotic (103). Diastolic dysfunction develops and the cardiac chambers fill less readily. Thus, a principal regulatory mechanism for NP release may be hampered, which could lead to decreased concentrations in circulation. Diastolic dysfunction, however, is also related to diabetes and lifestyle. To follow this hypothesis, arteriovenous gradients across the heart and achieved peripheral plasma concentrations while volume loading across the spectrum of body mass index may provide further insights.

One organ-related feature of obesity has attracted considerable interest, as NPRs, including the degradation receptor (NPR-C), are expressed in adipose tissue. As the total mass of white fat increases, so could the overall degradation of the peptides in circulation (104). If this suggestion holds true, the association of decreased NPs should mostly correlate with fat mass and to a lesser extent lean mass. This does not seem to be the case, and the above-mentioned report from the Dallas Heart Study in fact already then argued against this mechanism (94). In addition, as well as reductions in plasma ANP and BNP, obesity is associated with decreased concentrations of the amino-terminal peptides (Nt-proANP and Nt-proBNP), which do not depend on NPRs for clearance. Since then, several studies on body composition and associations with peptide concentrations have begun to attract pharmacological interest, for instance, as a potential need for substitution in obesity (103, 104).

Obesity comes with hormonal changes. Obesity is associated with dysregulation of gut peptides and the hormones involved in glucose regulation. Hormones specifically related to satiety and hunger could be actively involved in the obesity-related regulation of cardiac NPs. Leptin, a hormone released from adipocytes, is a satiety hormone. The connection of leptin with NPs started with the observation that ANP seemed to inhibit the release of leptin (105). It could therefore be argued that decreased ANP concentrations may favor increased food intake. However, leptin infusion in rats downregulates circulating ANP (106). Moreover, food intake per se reduces circulating concentrations of proANP (107, 108), a mechanism that has been shown to be mainly glucose mediated (108). Hence, any leptin connection to decreased NPs in circulation seems less certain or relevant.

Ghrelin is another hormone involved in appetite regulation. The role of ghrelin in obesity has been extensively examined, but the link to the cardiac NPs is poorly understood. Infusion of BNP was reported to reduce ghrelin concentrations, suppress hunger, and increase satiety (109). The authors concluded that this observation may help to explain cardiac cachexia, a metabolic state observed in advanced HF with high concentrations of circulating NPs. On the other hand, ghrelin infusion in rats reduced ANP concentrations in vivo but not in cultured cardiomyocytes (110). As is the case for leptin, it remains difficult to conclude to what extent ghrelin regulates cardiac NPs and/or vice versa.

Adiponectin has also attracted attention in connection with cardiovascular disease and the NP system. NP infusion in man increases adiponectin concentrations, which has been suggested to be a participating mechanism involved in the development of insulin resistance observed in HF (111). In patients with HF, however, a positive association between adiponectin and BNP has also been reported (112). These reports thus seem paradoxical, but it should be noted that infusion experiments are typically performed on a background of normal physiology, whereas clinical studies on HF should be interpreted with knowledge of the gross disturbances in physiology. NP release during infusion of adiponectin has not been reported, as infusion of exogenous adiponectin is not possible.

Major metabolic hormones include pancreatic insulin and glucagon. From the gut, the incretins are also involved in obesity prior to overt metabolic disease such as type 2 diabetes. Overall, the glucose-regulating hormones are dominantly increased in obesity, in other words, a state of hyperinsulinemia, hyperglucagonemia, and, for instance, increased fasting concentrations of GLP-1. All the mentioned hormones could, therefore, conceivably be involved in obesity-related decrease in expression of cardiac NPs. Notably, specific receptors for these hormones are expressed on cardiomyocytes. Several efforts have been pursued to deconvolute the isolated effect of each hormone. This has, however, proven difficult as in vivo interplay between the hormones cannot be accounted for. A murine study reported that exogenous GLP-1 in pharmacological doses could stimulate ANP expression (113). In a canine study, insulin infusion increased CNP expression in plasma and skeletal muscle, indicating that a loss of insulin sensitivity could lead to a state of CNP deficiency (114). Nevertheless, such findings have not been reported in healthy humans or in patients with type 2 diabetes (115-117). Therefore, however elegant the concept, the results from humans do not support such a gut–heart axis. Likewise, neither exogenous insulin nor glucagon seems to affect circulating concentrations of NP (99). As mentioned, the effect of metabolic hormones may not be simple, and some data still point to a possibility of combined hormonal effects. Food intake reduces NP concentrations in both healthy individuals and in patients with type 2 diabetes (107, 108, 118). Whether this effect is driven by postprandial vasodilation and redistribution of intravascular volumes, changes in glucose alone or together with shifts in pancreas/gut hormones, or a combination of these effects remains to be resolved.

Type 1 Diabetes

NPs in circulation have been thoroughly examined in patients with type 1 diabetes. The focus has almost exclusively been on the diagnostic and predictive role of plasma NP measurement in cardiovascular disease (119-123). In contrast, the possible direct impact of diabetes on the NP system without a cardiovascular focus remains unclear with only a few studies (124, 125). As type 1 diabetes affects children and adults, the lack of basal studies may seem understandable; in adult patients alone, type 1 diabetes eventually leads to complications that in themselves affect the NP concentrations in plasma. Renal disease, for instance, is a well-known cause of increased NPs in plasma due to increases in body fluid volume, including circulating volume, and decreased clearance (126). To make matters more complicated, type 1 diabetes cannot go untreated and diagnosis mandates urgent initiation of insulin therapy. A human study on type 1 diabetes prior to therapy is thus very difficult. As decreased NP concentrations are well established in obesity and type 2 diabetes, however, it seems reasonable to question whether the profound insufficiency of insulin production which characterizes type 1 diabetes may regulate cardiac NP production and release prior to the evolution of secondary diabetes–related cardiovascular complications. In addition, other types of diabetes are not well examined, as genetic forms of diabetes, for instance, the maturity-onset diabetes of the young classes.

To elucidate the effect of type 1 diabetes on NP biology, plasma measurement in newly diagnosed children could prove informative. This will require age-matched control children, which is generally a well-known challenge in establishing reference intervals. Only a few studies on the subject have been published (127-129), and the matter has still not been resolved. Finally, children born from mothers with type 1 diabetes could be considered a proxy of type 1 diabetes, at least with respect to the generally elevated concentrations of glucose also in the uterus. But the matter is complicated by the fact that the hearts of newborn children from mothers with type 1 diabetes are often structurally affected (130). Hence, it remains unanswered as to whether the NP system is affected by type 1 diabetes.

Type 2 Diabetes

A state of NP deficiency has been reported in patients with type 2 diabetes (131). Type 2 diabetes is, however, often associated with overweight and obesity, and studies of the NP system in “lean” patients with type 2 diabetes have not been reported. Type 2 diabetes in most population-based studies is considered to be 1 disease, which hampers a more detailed understanding of any potential mechanisms linking this disease to other hormone systems, including the cardiac hormones. Furthermore, type 2 diabetes is treated with various interventions, which should be considered in any causal interpretation between the disease complex and the cardio-endocrine response. One phenomenon linking NPs to type 2 diabetes comes from genetic studies on mutations/polymorphisms within the NP system. As mentioned, it has been shown that a polymorphism in the promoter region of the BNP gene reduces the risk of developing type 2 diabetes (61, 62). This polymorphism is associated with increased BNP concentrations in circulation and can thus be considered a gain-of-function polymorphism. The same role has been suggested for ANP (132, 133), where the peptides in general seem protective against developing type 2 diabetes. The magnitude of the contribution of NPs in the combined risk profile for type 2 diabetes remains undefined.

The state of NP deficiency in type 2 diabetes may have an impact on the combined phenotype seen in patients with an established diagnosis. First and foremost is the hypertensive phenotype, where the NP system is hormonally involved. A state of prolonged decreased circulating NPs leads to hypertension. It has been proposed that hypertension in type 2 diabetes may at least in part be caused by the relative NP deficiency (134, 135), and this deficiency may also be involved in the pathogenesis of type 2 diabetes. If this holds true, replacing NPs might be relevant as a therapy. To this point, replacement therapy could be personal (meaning biomarker driven), as patients with decreased NPs may benefit from such a regimen, while patients with normal concentrations of NPs in circulation may not. Notably, current treatment options are confined to molecules with very low half-life as only synthetic versions of ANP and BNP are available for treatment of HF, whereas an analog of CNP with a half-life of ∼28 minutes is approved for treatment of children with achondroplasia (136). Development of highly selective analogs with improved pharmacokinetic profile would very likely provide basis for an optimized treatment approach that could effectively alleviate a state of NP deficiency and benefit a broad range of patients with cardiometabolic disease.

NP replacement therapy could be performed with simple measurement of their endogenous NP system prior to therapy (and dosing). It seems relevant to not only measure basal NP concentrations but also introduce a stress test. As with most hormonal systems, a test for evaluating the dynamic capacity of hormone release (best known is perhaps the glucose tolerance test) is often more informative than just a singular measurement. Stressing the NP system in this context has not been systematically pursued. However, it is generally accepted that an increased venous return (preload) will cause a rapid release of ANP (and possibly BNP) from the right atrium—an effect that early on was described in connection to water immersion (137). Tilting is another and feasible test for this (100, 138, 139). Hormonal stress can also be used, for instance with angiotensin II or epinephrine administration (140, 141). Finally, a potential test for the NP system could be developed via its response to fluids, sodium load, or antidiuretic hormones.

A Note on Glucose

Plasma glucose is a central molecule in obesity and diabetes. While hormones affecting the glucose metabolism can be examined using either in vivo or in vitro models, it is difficult to exclude concomitant glucose concentrations from any analyses in human beings. To this end, examining the effects of insulin on cardiac NPs without concomitant changes in glucose is to be preferred. Here, the euglycemic, hyperinsulinemic glucose clamp model is a well-established methodology. In physiology and pathophysiology, however, a state of normal glucose concentrations and highly elevated insulin concentrations simply does not occur. Thus, when examining either obesity—with its risk of metabolic changes—and type 1 and type 2 diabetes, glucose per se will always be included as a possible common regulator of hormone expression. In fact, glucose has been suggested to regulate cardiomyocyte ANP expression via a cardio-specific miRNA (108), a feature that could also explain the effects of food intake. In extension of this, studies on glucose tolerance can be used with the acknowledgement that these tests (oral or intravenous) are still best suited for understanding physiology.

Conclusions

NPs have proven their value as meaningful biomarkers in cardiology with an emphasis on acute HF. Prior to HF, there are several conditions and factors that may be considered a state of NP deficiency; obesity and type 2 diabetes have so far attracted the most attention. But other factors such as age, sex, race, genetics, and diurnal regulation can also reduce the NP “armory” and leave some people prone to later developing cardiometabolic disease. The prevalence of, and the mechanisms underlying, relative NP deficiency require ongoing elucidation. Addressing the state of NP deficiency before development of cardiometabolic disease in selected patients could be a future path for improving longevity and quality of life.

Abbreviations

ANP

atrial natriuretic peptide

BNP

B-type natriuretic peptide

CNP

C-type natriuretic peptide

HF

heart failure

NP

natriuretic peptide

NPR

natriuretic peptide receptor