Hypertension is a leading preventable risk factor for cardiovascular, cerebrovascular, and kidney diseases [1] and in turn a major contributor to premature death, disability, and significant health care costs [2]. In Japan, ~50% of people treated for hypertension do not achieve their blood pressure (BP) goal. Furthermore, if untreated patients are included, ~70% of people with hypertension have persistently poorly controlled BP [3]. This information highlights the multifactorial nature of this debilitating condition, as current medication regimens are not sufficient to target the myriad of pathways leading to hypertension.

The problem with sodium

While sodium is a critically important cation for the physiological regulation of fluid volume, tissue perfusion, and cell metabolism, excess dietary sodium, mainly as sodium chloride (NaCl), has long been thought to be an important trigger for hypertension [4,5,6,7]. Consistent with this, even modest reductions in salt intake can significantly reduce BP, establishing NaCl or salt as a key modifiable risk factor for hypertension [8].

The population level significance of salt is underscored by the fact that in most countries, the average daily sodium intake is 100–200 mmol/day (~2–4 g/day). This amount greatly exceeds the daily physiological requirement of 10–20 mmol/day and the World Health Organization’s recommended daily intake of <2 g of dietary sodium (equivalent to 5 g of salt or NaCl) per day [9,10,11].

However, despite population-level evidence, individual susceptibility to the hypertensive ramifications of salt is heterogenous. This first led Kawasaki et al. [12] to classify individuals as either salt-sensitive (defined as a 10 mmHg rise in arterial BP with increased sodium intake) or salt-resistant [13, 14]. Such definitions are nevertheless arbitrary, and the phenotypes of salt sensitivity and salt resistance may be a continuous spectrum rather than strictly bimodal [15]. Although long-term observational studies have demonstrated an increased prevalence of cardiovascular events and mortality in a subset of salt-sensitive individuals compared to their salt-resistant counterparts [13, 14], the specific mechanisms underlying salt-induced hypertension are being redefined.

Classical Guytonian theory posits that the kidneys’ pressure-natriuresis system is core to body sodium and water regulation in response to alterations in renal arterial pressure and is hence centrally responsible for BP control [16]. Increased sodium intake and ensuing extracellular fluid volume expansion drive a subsequent rise in the perfusion pressure of the kidneys, triggering pressure natriuresis water excretion to correct the transient rise in circulating volume and prevent an increase in BP. Hypertension is a result of an abnormality in this pressure-natriuresis system [17]. This established dogma of sodium balance is not only overly simplistic but also not reflective of the complex physiology of body sodium and water regulation. The paradigm of sodium, fluid, and BP regulation is no longer considered kidney-centric but has shifted to include a network of novel extrarenal mechanisms for influencing body sodium and the management of hypertension.

In this review, we provide a holistic view of this pathophysiology using a whole-body understanding of sodium regulation involving the skin. We provide a mechanistic context for the steps linking dietary sodium and clinical outcomes. In doing so, we also seek to highlight current knowledge gaps.

Dietary sodium handling and physiology

Dietary sodium is absorbed across the apical membrane of intestinal enterocytes, facilitated by sodium/hydrogen (Na+/H+) exchangers, cotransport with sugars and phosphates, and diffusion through epithelial sodium channels (ENaC). From here, sodium is actively pumped across the intestinal basal membrane by sodium-potassium ATPase (Na+-K+ ATPase) [18] before entering the intestinal capillaries (Fig. 1).

Fig. 1
figure 1

Schematic representation of the relationship between dietary sodium (Na+) intake, circulation, and storage. Mϕ: monocyte-macrophages; Ton-EBP/NFAT5: tonicity-enhancer binding protein; VEGF-C: vascular endothelial growth factor-C

Full size image

Despite wide variations in dietary sodium intake, acute and chronic adaptive neural (e.g., sympathetic outflow) and hormonal (e.g., renin-angiotensin-aldosterone system [RAAS] and atrial natriuretic peptide) [16] responses maintain relatively constant sodium plasma concentrations. A short-term high-salt diet (HSD) seems to protect against increases in BP, while long-term high salt intake increases BP. Why this is the case is unclear. A HSD over the short term (<7 days) raises serum osmolality and arginine vasopressin (AVP)-mediated vasoconstriction, promptly initiating potent aortic baroreceptor GABAergic inhibition of AVP [19]. Conversely, in chronic HSD situations, there is increased AVP production and subsequent vasoconstriction via activation of the brain-derived neurotrophic factor/tropomyosin-related kinase B receptor pathway, and downregulation of the potassium-chloride-cotransporter-2 channel in central neurons prevents inhibitory GABAergic signaling [20, 21].

Other novel physiological regulatory responses have now been identified. Contrary to expectations, Rakova et al. [22] found that increased salt intake was accompanied by reduced water consumption in human subjects. Here, in the setting of a chronic high-salt diet, increased glucocorticoid release directly increases surplus sodium excretion without the need for further water ingestion. Additionally, normal mice on a high-salt diet accumulate urea from hepatic ketogenesis and glucocorticoid-derived catabolic muscle loss. This urea is an osmotic driver that increases both urinary sodium excretion and endogenous water creation [23]. Together, these water-conserving mechanisms of sodium excretion provide the means to excrete excess osmolytes without jeopardizing extracellular volume control and blood osmolality.

In a transgenic mouse model of psoriasis, transepidermal water loss occurs across psoriatic skin lesions with cutaneous vasoconstriction and metabolic adaptations (that enhance urea and organic osmolyte production) that conserve water and maintain body hydration but subsequently increase arterial hypertension and catabolic muscle wasting [24].

In the kidney, an integrated system of nephron ion channels, exchangers (Na+/H+ exchange in the proximal tubule) and transporters (sodium-potassium-chloride [Na++K+ + 2Cl−] cotransporter in the thick ascending loop of Henle, and sodium-chloride [Na+-Cl−] cotransporter in the distal convoluted tubule) function in concert to reabsorb 23 moles of NaCl a day. A small but critical final 2% of the reabsorbed sodium is controlled by ENaC in the cortical collecting tubule [25]. Thus, multiple hypotheses for salt sensitivity have emerged, including defects in renal sodium excretion, abnormal systemic vascular resistance, and impaired vasodilatory responses [26] as well as genetic components [27].

What we increasingly appreciate is that the extrarenal regulation of sodium has a pivotal role in whole-body sodium homeostasis and hence BP control (Fig. 1).

Sodium storage in tissue interstitium

Conventionally, we recognize two storage compartments of body sodium: one circulating and systemically active in the plasma [28] and one slowly exchangeable pool in bone [29]. We now know of a third store of sodium in the interstitial spaces of the vascular endothelium, muscle, and skin [30].

The interstitium comprises three dynamic phases: free fluid (with ions, albumin, and amino acids), a collagen matrix for structural integrity, and a gel phase rich in glycosaminoglycans (GAGs) [31]. Negatively charged GAGs fixed in confined interstitial fluid spaces can bind to cations, providing a protective buffering role for positively charged sodium [32].

The endothelium in blood pressure regulation

Dietary sodium must first pass through the vascular system into the extracellular space. The endothelial surface layer [33], also known as the endothelial glycocalyx, coats the luminal side of the endothelium and is composed mainly of GAGs [32]. The ESL, therefore, acts as an intravascular sodium buffer, although the sodium binding capacity of the ESL is as yet unknown [34]. While the significance of interstitial sodium is widely appreciated, why and how sodium transverses from the vascular lumen into the interstitium is poorly understood. Current thinking is that damage to the ESL results in paracellular leakage of sodium into the interstitial space [34].

Excess plasma sodium can also directly damage the underlying endothelial cells, leading to cellular entry of sodium. Exposure to sodium increases the ENaC density on endothelial cell membranes. This in turn leads to a cellular influx of sodium and an increase in mechanical stiffness in neighboring vascular smooth muscle cells through reduced endothelial cell generation of vasodilatory endothelial nitric oxide synthase (eNOS) [32, 35]. In addition, high plasma sodium levels trigger endothelial release of the proinflammatory cytokines IL-1β and TNFα [36].

Role of the skin in sodium and blood pressure regulation

The skin comprises two tissue layers, the epidermis (external layer of nonstratified epithelial cells) and the dermis (composed mainly of acellular connective tissue of fibroblasts, blood vessels, lymphatics, and nerves in an extracellular matrix) [37]. Filtration of fluid across the interstitium from the capillaries to the lymphatics is driven by hydrostatic pressure and is opposed by osmotic pressure.

Importantly, the abundant extracellular matrix of the skin is rich in GAGs, which can bind and store sodium, in effect acting as a major reservoir for excess sodium in situations where the body is inundated [38] (Fig. 1). In rodent studies, skin sodium concentrations up to 180–190 mmol/L (compared to normal levels of ~140 mmol/L) can be stored without concomitant effects on extracellular fluid volume, body weight or BP [39]. A HSD increases the negative interstitial charge density and skin sodium in rodent and human studies [39, 40]. Conversely, a low-salt diet (LSD) reduces GAG polymerization, decreasing the negative charge density and leading to sodium release [39]. Interestingly, the ratio of interstitial space to intracellular fluid is much higher in skin than in other more tightly packed tissues, such as muscle [31], suggesting that skin has a greater capacity to store sodium than other tissues.

The osmotic activity of accumulated interstitial sodium remains controversial, and the use of the binary terms “osmotically active” and “osmotically inactive” to describe sodium may be misleading. On the one hand, the activation of Ton-EBP/NFAT5 (a known osmotic response factor) in response to the hypertonic skin microenvironment suggests that skin sodium is osmotic in nature (made hypertonic by GAG inactivation) [41]. Conversely, Rossitto et al. posited that a HSD does not lead to skin-specific water-free sodium storage but rather systemic sodium excess. In a spontaneously hypertensive rodent model fed a HSD for 3 weeks, there were similar increases in skin, liver, lung and skeletal muscle sodium contents. Tissue water was similarly increased in these organs apart from skeletal muscle [42]. This almost global effect could be explained by an intracellular shift of sodium in exchange for potassium or other intracellular osmolytes [43]. It is suggested that over time, extracellular and intracellular cation and volume proportions could adapt to accommodate excess sodium and corresponding water retention while preserving osmolality [44]. Salt-sensitive human volunteers gained weight “commensurate to iso-osmolar water retention” during acute salt loading [45]. Studies using 23-sodium magnetic resonance imaging (23Na MRI) have demonstrated increased sodium signals in human skin layers; however, this technology visualizes whole tissues and cannot discriminate between the extracellular and intracellular space [46]. Therefore, in the absence of being able to measure osmotic pressure in situ, there is no consensus on the osmotic nature of stored skin sodium. Likewise, further research is needed to determine whether body sodium accumulation in tissues is skin specific or in fact a more systemic phenomenon and its relationship with concomitant water accrual.

Skin sodium sensors and modulators

Immune cells residing in different layers of skin are critical local sensors of the interstitial sodium concentration and have a role in systemic fluid regulation and BP control. It has been proposed that keratinocytes pump sodium into the dermis and increase the sodium concentration at the superficial dermal papilla, perfused by a vascular countercurrent system [47, 48]. What drives sodium to enter the skin is yet to be clearly elucidated. Skin phagocytes sense the hypertonic microenvironment during salt loading, and an influx of cells of the mononuclear phagocytic system (MPS) activates tonicity-enhancer binding protein (Ton-EBP/NFAT5), a transcription factor that regulates the expression of osmo-protective genes. Ton-EBP binds to vascular endothelial growth factor (VEGF-C), leading to increased VEGF-C secretion into the skin interstitium, which mediates lymphangiogenesis and eNOS expression to enhance the skin’s extrarenal clearance of sodium [49] (Fig. 1).

Disruption of sodium clearance by macrophage depletion in skin is associated with increases in BP [50], suggesting that failure of the TonEBP-VEGF-C axis might be a mechanism of salt sensitivity. MPS depletion in both a HSD group and a deoxycorticosterone acetate-salt-treated rodent model abrogated lymphatic capillary hyperplasia and increased BP [51]. In a rodent model, sunitinib, an antiangiogenic and anticancer agent that inhibits all three types of VEGF-C receptors, decreased lymphangiogenesis, increased skin sodium and chloride content and induced a 15-mmHg rise in BP, independent of changes in plasma sodium concentration and body weight. These responses were further exacerbated by a HSD [52].

Other signaling mechanisms may also link skin sodium levels and BP regulation. Skin capillary rarefaction, from impaired angiogenesis (functional nonperfusion) or capillary loss (structural), alters peripheral vascular resistance, mediating changes in BP [53]. He et al. showed that a modest reduction in salt intake improves functional and structural dermal capillary density in hypertensive humans [54]. Functional but not structural capillary rarefaction was exhibited in patients with mild hypertension, regardless of antihypertensive treatment status, and was associated with endothelial dysfunction [55]. Alternatively, rodent studies with deletion of HIF-1α (promotes vasodilation) had higher BP, while deletion of HIF-2α (promotes vasoconstriction) resulted in lower BP [56]. Nevertheless, the clinical significance of different signaling mechanisms in the skin is unclear and largely unknown. Therefore, more needs to be done to understand how skin sodium physiology relates to the totality of human sodium balance and BP.

The role of the immune system in the genesis and propagation of hypertension

Available evidence suggests that sodium activates the innate and adaptive immune systems to induce a BP-independent proinflammatory profile in lymphoid or peripheral tissues. Increased renal perfusion pressure in salt-sensitive hypertension drives macrophages into the interstitium, where increased extracellular osmolarity triggers a proinflammatory phenotype [57]. The systemic RAAS activates [58] and a high-salt environment primes monocytes [59]. Enhanced proinflammatory and suppressed anti-inflammatory gene expression [59] results in end-organ damage. Dendritic cells (DCs) produce isolevuglandin (IsoLG)-adduct, activating nicotinamide adenine dinucleotide phosphate (NADPH) oxidase that promotes T-lymphocyte activation and the secretion of proinflammatory cytokines and reactive oxygen species [60]. Healthy human subjects were found to have increased levels of the intermediate subset of CD14++CD16+ monocytes (implicated in peripheral artery disease [61]) after a 7-day HSD, and these levels were fully reversed following a 7-day LSD [62]. Yi et al. found that individuals fed a HSD had a significant increase in the levels of peripheral monocytes and an increase in proinflammatory cytokines (interleukin (IL)-6 and IL-23) along with enhanced anti-inflammatory cytokines (IL-10) compared to those of controls fed a LSD [63]. The salt-induced production of IL-23 polarizes the TH17 cell phenotype via the p38/MAPK pathway involving Ton-EBP/NFAT5 and serum/glucocorticoid-regulated kinase 1 (SGK1), which has been implicated in autoimmune disease [64], hypertension and kidney injury [65].

A high-salt microenvironment in tissues and vessels with an exaggerated proinflammatory and profibrotic response is a causative factor in high BP. When DCs are adoptively transferred into naïve mice, sodium can prime hypertension in response to angiotensin II via amiloride-sensitive channels [66]. Moreover, dendritic cells express mRNA for sodium channels, including sodium-hydrogen exchangers (NHEs), sodium-calcium exchangers (NCXs) and ENaC channels, and when inhibited, prevent salt-induced activation of NADPH oxidase [66]. The role of the innate immune system, including neutrophils [67] and the complement system [68], has also been implicated in the cultivation of hypertension. This inflammatory milieu favors more sodium reabsorption and consequently compounds the BP-induced microvascular remodeling in cardiovascular, kidney and vascular diseases.

The importance of chloride

Interestingly, the hypertensive effect of dietary NaCl appears to require the coupling of both sodium and chloride. In hypertensive rodent models, significant BP increases are seen with a high-NaCl diet but not with a stoichiometrically similar anion load of bicarbonate or nonchloride salts [69, 70]. Pharmacological inhibition of a calcium-activated chloride channel expressed in vascular smooth muscle cells prevents arterial vasoconstriction and reduces BP in spontaneously hypertensive rats [71]. Moreover, increased BP in response to a HSD could not be replicated when sodium was given as sodium citrate [72], sodium bicarbonate [73], or sodium phosphate [74] in human studies. Both urinary chloride and sodium excretion show similar positive correlations with BP [75], while large observational studies have found that a low serum chloride level is independently associated with both cardiovascular mortality, after adjustment for serum sodium levels and traditional cardiovascular risk factors [76], and all-cause mortality, independent of serum sodium and bicarbonate levels [77]. Patients with chronic kidney disease who had low serum chloride concentrations demonstrated an independent correlation with low systolic BP and proteinuria after one month of medical management [78]. In contrast, a large randomized controlled study demonstrated that salt substitutes with potassium chloride significantly reduce cardiovascular events, stroke and mortality [79]. This result suggests that a reduction in dietary sodium and/or potassium supplementation may be a cost-effect, large-scale method to mitigate the negative ramifications of salt intake [80,81,82].

We therefore should not neglect the potential role of chloride and its intimate relationship with sodium in influencing BP. It appears that this anion has been overshadowed by its cation counterpart and that there is still much left to understand (reviewed in [83]).

Broader clinical implications of excess sodium

Locally increased sodium concentrations have been implicated in combating pathogens. After Leishmania major infection was reduced with antibiotic treatment [84], high sodium levels increased NO production and activated macrophages in a Ton-EBP/NFAT-5-dependent manner, promoting cutaneous antimicrobial defense [84].

Sodium can be considered a “uremic toxin” [85] that promotes oxidative stress [86,87,88,89]. Sodium influences urea-driven changes in metabolism and muscle mass [23]. Increased plasma sodium concentrations can also have a direct pathological effect on vascular smooth muscle cell hypertrophy [90], and elevated intracellular sodium concentrations can increase vascular smooth muscle tension [91]. These HSD-driven metabolic consequences can lead to sarcopenia with associated insulin resistance and obesity [92]. This effect can be seen as an increased muscle sodium content on 23Na MRI in maintenance hemodialysis patients with increased insulin sensitivity [93]. Furthermore, an increased tissue sodium concentration potentially accelerates the aging process. Tissue sodium levels on 23Na MRI are associated with increasing age [40], while high extracellular NaCl accelerates cellular senescence [94].

A chronic low-grade inflammatory response from sodium may provide the missing link between tissue sodium, increased BP and cardiovascular/metabolic disease. Henceforth, existing and novel therapeutic agents that reduce sodium levels could have the dual potential to mitigate both inflammatory and metabolic disease. Salt-sensitive hypertension in animal models is attenuated by immunosuppression [95]. A reversible reduction in BP was seen in patients with rheumatoid arthritis and psoriasis treated with immunosuppressive mycophenolate mofetil [96]. Increased peripheral CD4+ T lymphocytes were associated with an increased prevalence of hypertension in human immunodeficiency virus-positive patients treated with highly active antiretroviral therapy [97]. Patients with diabetes randomized to sodium-glucose cotransporter (SGLT)-2 inhibitor treatment had a significant decrease in tissue sodium concentrations after 6 weeks, as assessed by 23Na MRI [98]. Acute 2-day dapagliflozin treatment exhibited a direct vascular benefit with improved endothelial function and reduced arterial stiffness and renal resistive indices in patients with type 2 diabetes, independent of BP changes [99]. In contrast, there were no significant differences in the 23Na MRI-quantified skin and muscle sodium contents or the pulse wave velocity among patients with mild hypertension randomized to a LSD, chlorthalidone, spironolactone, or placebo for 8 weeks [100]. However, this again supports the notion that clinical improvement from dietary salt changes and manipulation of the sodium paradigm via medications may only be observed if implemented chronically or if renal sodium excretion is disrupted, as observed in patients with chronic kidney disease (including those on maintenance hemodialysis and peritoneal dialysis) [101].

Conclusion

Excessive salt intake contributes to worsening BP and its ensuing cardiovascular, cerebrovascular, and kidney disease. The skin–sodium–hypertension axis involves the endothelium, lymphatics, and immune cells. Skin sodium accumulation not only causes hypertension but may also be part of a cluster of multifactorial risks that steer pro-inflammatory profiles, disrupt metabolism, boost cardiovascular disease, and hasten aging. We are only starting to grasp an understanding of this complex and intricate interplay (Table 1). There is still much to be discovered about the salt within us.

Table 1 Important knowledge gaps in our understanding of salt and blood pressure
Full size table