Sorting Nexins: Role in The Regulation of Blood Pressure

Abstract

Sorting nexins (SNXs) are a family of proteins that regulate cellular cargo sorting and trafficking, maintain intracellular protein homeostasis, and participate in intracellular signaling. SNXs are also important in the regulation of blood pressure, via several mechanisms. Aberrant expression and dysfunction of SNXs participate in the dysregulation of blood pressure. Genetic studies show correlation between SNX gene variants and response to antihypertensive drugs. In this review, we summarize the progress in SNX-mediated regulation of blood pressure, discuss the potential role of SNXs in the pathophysiology and treatment of hypertension, and propose novel strategies for the medical therapy of hypertension.

Introduction

Hypertension has become a serious public health problem [1]. It was predicted in 2005 that globally, in 2025, 29% of adult subjects (1.56 billion) may develop hypertension [2]. However, the incidence of hypertension has greatly exceeded this prediction. The global incidence of hypertension was 31.1%, or 1.39 billion people in 2010 [3]. The high prevalence of hypertension and its complications cause a huge economic burden worldwide [3,4].

The pathogenesis of essential hypertension is complex, involving inherited and external factors [5–9]. Many organs and systems including arteries, heart, kidneys, and nervous system play important roles in the development of hypertension. Numerous reports have demonstrated that lifestyle, including diet, exercise, sleep, and external factors, such as environmental pollution, affect the development, prevention, and treatment of hypertension [6,7]. Genetics and epigenetics are intricately associated with the pathogenesis and treatment of hypertension [8,9]. However, current studies have not totally uncovered the pathogenesis of hypertension, and thus, continuing investigation into the involvement of new factors and mechanisms are needed.

Sorting nexins (SNXs) are a family of proteins that participate in the trafficking of proteins in the cytoplasm and plasma membrane [10]. SNXs have been identified to modulate the entire endocytic process [10–12]. Many reports have demonstrated that the deficiency of SNXs is associated with several diseases, including hypertension, in which protein sorting and trafficking are disturbed [13–15]. In this review, we summarize the progress in SNX-mediated regulation of blood pressure, discuss the potential role of SNXs in the pathophysiology and treatment of hypertension, and propose novel strategies for the medical therapy of hypertension.

The subfamilies of SNXs

Ten subtypes in yeast and 33 subtypes in mammals have been identified as members of the SNX family [16]. The SNX members are categorized into five subfamilies based on their structural domains [12,17]: SNX-PX subfamily, SNX-BAR (Bin/Amphiphysin/Rvs) subfamily, SNX-FERM (protein 4.1/ezrin/radixin/moesin) subfamily, SNX-PXA (PX-associated domain A)-RGS (regulator of G-protein signaling domain)-PXC (PX-associated domain C) subfamily, and others. Basically, all SNXs share a phagocyte oxidase (phox) homology (PX) domain, containing the evolutionarily canonical 100–130 amino acids [17]. (1) The SNX-PX subfamily has ten members (SNX3/10/11/12/16/20/21/22/24/29). Their common PX domain interacts with phosphatidylinositols (PIs), such as phosphatidylinositol-3-phosphate (PI3P), which is well expressed in organelle membranes [18,19]. (2) The SNX-BAR subfamily has twelve members (SNX1/2/4/5/6/7/8/9/18/30/32/33). They possess a special BAR domain in the C-terminal and common PX domains [20]. (3) The SNX-FERM subfamily has three members (SNX17/27/31). They have a FERM domain, with an atypical tertiary structure in the C-terminal and PX domains [21,22]. Specifically, SNX27 has an N-terminal postsynaptic density 95/discs large/zonula occludens (PDZ) domain [23]. (4) The SNX-PXA-RGS-PXC subfamily has four members (SNX13/14/19/25). They have a PXA-RGS-PX-PXC or PXA-PX-PXC domain in the C-terminal domain [24,25]. The other unique SNX subfamily includes SNX15, SNX23, SNX26, and SNX28 [17,26,27].

Biological functions of SNXs

The most basic role of SNXs is being a vital constituent of the retromer [28–30]. The retromer is a heteropentameric complex composed of two associated but distinct subcomplexes: a membrane-associated SNX dimeric subcomplex and a heterotrimer that includes VPS26, VPS29, and VPS35 [28]. The retromer core complex is an endosomal scaffold, that mediates the recycling or trafficking of intracellular proteins, such as G protein-coupled receptors (GPCRs) and other cargos from endosomes to either the trans-Golgi network or the plasma membrane [29,30]. The retromer interacts with three SNX subfamily members, including SNX3 from the SNX-PX subfamily, SNX-BAR subfamily members, and SNX27 from the SNX-FERM subfamily. The retromer VPS35 and a heterodimer, such as SNX1/5 and SNX2/6, participate in the insulin-like growth factor 2 receptor retrieval [31]. The SNX27-retromer complex participates in the regulation of transmembrane cargos, including the parathyroid hormone receptor; the SNX27 and parathyroid receptor interaction occurs with their the PDZ domains. However, the overall rate of recycling of the parathyroid hormone does not involve the PDZ domain but rather through a direct interaction of the retromer [32]. The SNX3-retromer regulates the membrane distribution and function of proteins associated with polycystic kidney disease [33]. However, SNX3 also mediates retromer-independent functions, such as facilitating tubular endosomal recycling of intracellular cargoes [34]. These reports indicate that the retromer complex and associated SNXs comprise a vital cellular trafficking machinery, which plays an important role in endosomal protein sorting.

SNXs are important in the regulation of the routing of cargoes, such as mitochondria, endoplasmic reticulum, peroxisomes, and cytoplasmic protein aggregates [35], via endocytosis, autophagy, and other trafficking pathways [36]. Endocytosis mediates the cellular uptake of extracellular components through membrane invagination [37]. SNX9 promotes podocin endocytosis associated with the chemotherapeutic drug adriamycin-induced podocyte injury and participates in clathrin-dependent and -independent endocytosis and cargo trafficking [38,39]. SNX33 modulates the endocytosis and alpha-secretase cleavage of amyloid precursor proteins [40]. SNXs also modulate the activity and/or function of cargoes by regulating varying trafficking pathways, but not their expression. For example, SNX1 is important for the trafficking of group 1 metabotropic glutamate receptors by directing them into a slow recycling pathway which allows receptor resensitization. The absence of SNX1 results in faster recycling of group 1 metabotropic receptors, preventing their resensitization [41]. In intestinal epithelial cells, SNX27 regulates the targeting of sodium-hydrogen exchanger type 3 (NHE3) at the plasma membrane by promoting the exocytosis of NHE3 from early endosome to the plasma membrane; SNX27 silencing decreases the plasma membrane activity and expression of NHE3 activity but not its total expression [42].

SNXs are also important in the regulation of the expression of intracellular proteins, via the proteasomal and lysosomal pathways [35]. In colorectal cancer cells, SNX16 activates c-Myc signaling by suppressing ubiquitin-mediated proteasomal degradation of the eukaryotic translation elongation factor 1A2 [43]. SNX1 modulates the protein decay of epidermal growth factor receptor (EGFR) by sorting it to lysosomes [44,45]. SNX11 facilitates the transport of transient receptor potential cation channel subfamily V member 3 into lysosomes for its decay [46]. SNX5 regulates the degradation of vesicular acetylcholine transporter, via the lysosomal pathway [47]. However, there are inconsistent reports on the effects of SNXs in the regulation of protein decay. For example, SNXs are also involved in the prevention of protein degradation. SNX10 restrains the development of colorectal cancer by inhibiting the trafficking of SRC, a non-receptor tyrosine kinase, to lysosomes where it would be degraded [48]. SNX17 by interacting with β₁-integrin, causes its trafficking to the membrane, preventing its lysosomal degradation [49]. SNX27 interacts with the retromer VPS26, leading to the retardation of the transport of cargos to lysosomes [50]. These suggest that SNXs not only promote intracellular protein degradation but also are involved in the prevention of protein degradation.

SNXs can directly determine the destination of cells by regulating vital signaling molecules or pathways. For example, SNX9 inhibits the proliferation of immortalized autosomal dominant polycystic kidney disease cells by activating the Hippo-yes-associated protein signaling pathway [51]. SNX7 exerts its anti-apoptotic effect via Fas-associated death domain-like IL-1 converting enzyme-like inhibitory protein/caspase 8 pathway during the embryonic liver development [52]. SNX5 is essential for virus-induced autophagy by interacting with beclin 1 and autophagy-related gene (ATG) 14-containing class III phosphatidylinositol-3-kinase complex 1 [53]. SNX4 mediates the recycling of ATG9A, a membrane spanning ATG protein, from endolysosomes and autolysosomes to early endosomes, which is essential for autophagy [54]. SNX5 facilitates ferroptosis, a programmed cell death that is dependent on iron, in Parkinson’s disease models. Conversely, SNX5 depletion reduces dopaminergic neuronal ferroptosis in 6-OHDA-lesioned PC12 cells, a rat pheochromocytoma cell line [55]. Thus, current studies demonstrate that SNXs are involved in multiple cellular biological activities, including cell proliferation, apoptosis, autophagy, and ferroptosis.

SNXs have other physiological functions. SNXs directly participate in the intracellular signaling cascade. SNX10 controls mTOR activation, via the modulation of amino acid metabolism [56]. SNX9 makes a distinction between transforming growth factor β (TGF-β)-activated Smad3 from Smad2 and Smad4 to promote rapid nuclear delivery [57,58]. Our studies also found that SNX5 siRNA treatment reduces the amount of insulin receptors in human renal proximal tubule cells (hRPTCs), decreasing the levels of insulin receptor substrate-1 (IRS-1) and phospho-protein kinase B, two vital molecules in the insulin receptor-mediated signaling cascade [59]. SNXs are also involved in other intracellular functions, including restricting endolysosome motility, regulating the saturated fatty acid metabolism and endoplasmic reticulum homeostasis, and modulating cellular lipid deposition, among others [60–62].

Regulation of blood pressure by SNXs

SNXs regulate blood pressure by three mechanisms: renal sodium excretion, vasoconstriction and vascular remodeling, and insulin sensitivity. SNXs and their intracellular functions associated with the regulation of blood pressure are presented in Table 1.

Table 1.

Summary of SNXs and their intracellular functions associated with the regulation of blood pressure

Isoform	Cell types	Cargos	Intracellular distribution	Intracellular physiological functions	References
SNX1	hRPTCs	D5R	Distributed to a greater extent in the cytoplasm than at the cell membrane in hRPTCs	Required for D5R endocytosis upon agonist stimulation; Snx1 silencing (siRNA) in hRPTCs impairs D5R internalization, GTP binding, and function (stimulation of cAMP production and inhibition of sodium transport)	[89]
	VSMCs	AT1R	Distributed in the cytoplasm and at the cell membrane	Snx1 knockdown increases AT1R protein expression and its mediation of Ca2+ signaling	[97]
SNX3	HEK-293 cells, mCCD cells	ENaC	—	Increases ENaC cell surface expression	[124]
SNX4	COS-7 cells	ENaC	Distributed in early endosomes and recycling endosomes	Promotes the endocytosis of ENaC	[121]
SNX5	HEK-293 cells	D1R	Expressed at the plasma membrane and in the cytoplasm	Snx5 silencing (siRNA) in hRPTCs increases agonist-activated D1R phosphorylation, prevents D1R internalization and cAMP response, and delays D1R receptor recycling to the plasma membrane	[77]
	hRPTCs	IR		Snx5 silencing (siRNA) in hRPTCs decreases IR expression and abundance of p-IRS and p-PKB	[59]
	hRPTCs	IDE		Snx5 silencing (siRNA) in hRPTCs decreases IDE expression and activity	[134]
SNX9	HUVECs	β1-integrins	—	Required for tube formation in HUVECs and regulates recycling of β1-integrins from endosomes to the plasma membrane	[103]
	3T3L1 adipocytes	IR GLUT4	Expressed to a greater extent in the cytosol than at the plasma membrane	Regulates the IR trafficking and GLUT4 translocation to the plasma membrane induced by insulin	[139]
SNX10	Mouse BMDMs, human PBMCs	CD36	—	SNX10 deficiency results in the polarization of macrophages towards the anti-inflammatory M2 phenotype, and inhibition of foam cell formation by interrupting the internalization of CD36	[110,111]
SNX11	HEK-293T cells, HeLa cells	TRPV3	Distributed throughout the cytoplasm	Promotes TRPV3 trafficking from the plasma membrane to lysosomes for degradation	[46]
SNX17	mouse fibroblasts	β1-integrins	Distributed in early endosomes and recycling endosomes	Required for β1-integrin recycling and stabilization, and regulation of integrin-mediated cell function	[49]
	HUVECs	P-selectin	Expressed in the cytosol	Promotes the endocytosis of P-selectin, and retards its movement into lysosomes	[112]
SNX19	hRPTCs	D1R	Distributed in the cytoplasm and at the plasma cell membrane	Silencing Snx19 (siRNA) in hRPTCs impairs D1R-mediated increase in cAMP production and inhibition of basolateral sodium transport	[80]
	pancreatic β cells	DCVs	—	Silencing Snx19 (siRNA) in mouse pancreatic cells decreases insulin secretion and the number of DCVs; SNX19 interacts with IA-2 to initiate the pre-apoptotic state of pancreatic β cells	[127–129]
SNX25	HEK-293T cells	D1R, D2R	Localized in distinct clusters near the plasma membrane	Overexpression of SNX25 increases the expression of D1R and D2R, enhances receptor-mediated signaling, and perturbs both endocytosis and recycling of D2R; Silencing Snx25 (siRNA) decreases D1R and D2R expression	[83,84]
SNX27	Caco2 cells	DRA	Localized in early endosomes	Mediates DRA recycling to facilitate its activity in lipid rafts at the cell plasma membrane	[123]
	SK-CO15 cells	NHE3	Mainly localized intracellularly	Promotes cell surface expression of NHE3	[42]
	Rat IMCD cells	AQP2	Mainly localized intracellularly	Regulates the autophagy-lysosomal degradation of AQP2 protein	[125]
	mouse and human β cells	GLP-1R	Distributed to a greater extent in the cytoplasm than at the plasma membrane	Regulates GLP-1R trafficking and signaling	[130]
	HeLa cells	GLUT1	—	Modulates GLUT1 lysosomal degradation and plasma membrane levels; Snx27 silencing (shRNA) decreases GLUT1 recycling to the cell plasma membrane	[50]
	human SGBS cells	GLUT4	Distributed to a greater extent in the cytoplasm than at the cell membrane, but insulin recruits SNX27 to the plasma membrane	Snx27 silencing (shRNA) decreases GLUT4 expression	[141]
SNX31	mouse fibroblasts	β1-integrins	Distributed in early endosomes and recycling endosomes	Regulates β1-integrins surface levels and stability by binding to its tails	[105]

Abbreviations: AQP2, aquaporin-2 protein; AT₁R, ang II type 1 receptor; BMDMs, bone marrow derived macrophage; D₁R, dopamine D₁ receptor; D₂R, dopamine D₂ receptor; D₅R, dopamine D₅ receptor; DCVs, dense core vesicles; DRA, down-regulation adenoma protein (aka SLC26A3 and chloride anion exchanger); ENaC, epithelial Na⁺ channel; GLP-1R, glucagon-like peptide type 1 receptor; GLUT1, glucose transporter 1; GLUT4, glucose transporter type 4; hRPTCs, human renal proximal tubule cells; HUVECs, human umbilical vein endothelial cells; IA-2, islet antigen-2; IDE, insulin-degrading enzyme; IMCD, inner medullary collecting duct; IR, insulin receptor; mCCD, mouse cortical collecting duct; PBMCs, peripheral blood mononuclear cells; p-IRS, phosphorylated insulin receptor substrate 1; p-PKB, phosphorylated protein kinase B; SGBS, Simpson-Golabi-Behmel syndrome; SNX, sorting nexin; TRPV3, transient receptor potential cation channel subfamily V member, Usp10, ubiquitin-specific protease 10; VSMCs, vascular smooth muscle cells.

Renal SNXs and blood pressure

The kidney is the main organ involved in the regulation of sodium balance and blood pressure [63]. Essential hypertension may be initiated by an increase in renal sodium reabsorption and an inappropriate decrease in renal sodium reabsorption and increase in sodium excretion with the increase in blood pressure. This impaired pressure natriuresis in hypertension is, in part, caused by a predominance of the autocrine, paracrine, and endocrine factors that increase renal sodium transport, such as angiotensin II, over the autocrine, paracrine, and endocrine factors that decrease renal sodium transport, such as dopamine [63–66].

Dopamine exerts its natriuretic effects through the occupation of renal several receptor subtypes, which are categorized into D₁-like (D₁R/D₅R) and D₂-like receptor (D₂R/D₃R/D₄R) subtypes [67–69]. In hypertensive states, the expressions of renal dopamine receptors are reduced, which cause impaired sodium excretion, and thus resulting in a positive sodium balance [70–72]. SNXs and dopamine receptors in the kidney interact to regulate blood pressure.

SNX-mediated regulation of renal D₁R

There are abundant D₁Rs in the kidney, which participate in the regulation of sodium excretion [73]. During normal or moderately increased NaCl intake, urinary sodium excretion is increased due, in part, to increased dopamine generation and enhanced D₁-like receptor activation in the kidney [74]. The knockout of the Drd1 gene in mice promotes the development of hypertension [75,76].

The D₁R is regulated by SNX5. Our previous studies showed that SNX5 and D₁R colocalize in several segments of the renal tubule, including the proximal and distal tubules, of humans and rats [77]. We also found their localization in the brain, showing that the interaction between SNX5 and D₁R is not restricted to the kidney [77]. Treatment of hRPTCs with fenoldopam, an agonist for both D₁R and D₅R, promotes the endocytosis of D₁R and changes its intracellular distribution, as well as SNX5 [77]. Further studies showed that SNX5 not only regulates the endocytosis and cyclic adenosine monophosphate (cAMP) response induced by stimulation of D₁R, but also the transport of D₁R back to cell membrane, involving its C-terminus [78,79]. SNX5 also regulates renal D₁R function in hypertensive animals. Renal-selective depletion of Snx5 aggravates the already elevated blood pressure in spontaneously hypertensive rats (SHRs), which is not observed in vehicle-infused SHRs. The further increase in blood pressure in SHRs with Snx5 knockdown is accompanied by impaired natriuresis [77].

The D₁R in the kidney is also regulated by other SNXs. In hRPTCs, SNX19 colocalizes with D₁R in the cytoplasm and at the plasma membrane; activation of D₁Rs with fenoldopam causes their endocytosis [80]. In the basal state, in the mouse kidney, as in hRPTCs, the colocalization between SNX19 and D₁R is also observed at the plasma membrane of renal proximal tubules (RPTs); the intravenous infusion of fenoldopam increased the colocalization of SNX19 and D₁R in the cytosol of RPTs. The SNX19 interaction with D₁R is specific because the D₅R does co-immunoprecipitate with SNX19 in RPTs [80]. SNX19 participates in the palmitoylation of D₁R, needed for its targeting to lipid rafts, which are specialized plasma membrane microdomains enriched by sterols and sphingolipids [80–82]. The SNX19-mediated regulation of cellular D₁R localization is important in the function of renal D₁R in cell and animal experiments. Snx19 depletion in hRPTCs decreases the fenoldopam-mediated cAMP generation and inhibition of sodium pump activity. Snx19 depletion in normal salt diet-fed mice decreases renal D₁R expression, leading to an increase in systolic blood pressure. Renal-selective overexpression of Drd1 normalizes the systolic blood pressure in these mice with kidneys depleted of Snx19, which is not observed in mice overexpressed with mutant D₁R [80]. SNX25 also regulates D₁R in the kidney. Overexpression of Snx25 in HEK 293 cells heterologously expressing D₁R and D₂R increases the expression of D₁R and D₂R, activates the downstream molecules, and disturbs the endocytosis and recycling of D₂R. Conversely, the expressions of D₁R and D₂R are reduced in cells lacking Snx25 [83,84]. Thus, current studies show that renal D₁R is regulated by some SNXs, specifically SNX5, SNX19, and SNX25. In addition, Rao et al reported that SNX13 may interact with D₁R in the modulation of renal albumin excretion [85], which may be predictive of the risk of cardiovascular events in patients with essential hypertension [86].

SNX1-mediated regulation of renal D₅R.

D₅R, the other dopamine D₁-like receptor subtype in humans, is also involved in the regulation of blood pressure [87]. The D₅R is constitutively active and has a higher affinity for dopamine than the D_lR [88]. Our previous studies showed that SNX1 regulates D₅R trafficking and its functions in human and mouse kidney cells [89,90]. SNX1, as well as the retromer, including SNX1/2, participates in intracellular cargo transport [91,92].

D₁R and D₅R can be clearly distinguished from each other by SNX1, which binds strongly to the D₅R but not D₁R C-terminal tail [93]. In the human kidney, SNX1 is well expressed in the RPT and glomerulus. In hRPTCs, SNX1 colocalizes with D₅R that is mostly evident at the cell membrane [89]. Treatment of hRPTCs with fenoldopam causes D₅R endocytosis, and subsequently accelerates its translocation to the membrane. Snx1 depletion also impairs D₅R-mediated physiological functions, including inhibition of renal sodium pump activity and impairment of cAMP generation [89]. These interactions in vitro are reflected in in vivo studies. Snx1 knockout (Snx1^−/−) or renal Snx1 knockdown, using Snx1-specific siRNA, in mice increases systolic blood pressure and blunts the natriuretic response to fenoldopam [89,90]. The impaired natriuretic response to dopamine when Snx1 is silenced is related to the increased renal expression of sodium exchangers, pump, and cotransporters such as NHE3, Na⁺-K⁺-ATPase, and the thiazide-sensitive sodium-chloride cotransporter (NCC), which are normally inhibited by D₅R [94]. Renal-selective depletion of Snx1 also leads to increased renal expression of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX) components, such as NOX1, NOX2, p47phox, resulting in oxidative stress [89,90]. Renal-selective Snx1 depletion also increases the renal expression of angiotensin II (Ang II) type 1 receptor (AT₁R) and contributes to the increase in blood pressure which cannot be counteracted by the dysfunctional D₅R [89,95,96]. These results indicate that SNX1-mediated regulation of D₅R expression and function is needed for normal renal sodium handling and blood pressure.

Vascular SNXs and blood pressure

It cannot be overstated that abnormalities in the vasculature causes hypertension. In addition to the effects of SNXs in the kidney, SNXs regulate blood pressure by actions in the vasculature. We have reported that SNX1 and AT₁R colocalize in the tunica media of the aorta and vascular smooth muscle cells (VSMCs) [97]. Snx1^−/− mice, which as indicated earlier are hypertensive, have increased mesenteric arterial reactivity to Ang II, which is due to increased AT₁R expression, without affecting the expression of the vasorelaxant AT₂R. Snx1 knockdown with siRNA in VSMCs increases AT₁R expression and decreases in the colocalization of AT₁R and an endosome marker Rab5 [97]; the elevation in cytoplasmic Ca²⁺ level may explain the augmented VSMC reactivity in this situation [98,99]. Other studies in hRPTCs found that the elevation in AT₁R protein may be a result of its reduced degradation in the proteasome pathway [96]. In Snx1^−/− mice, the impaired function of D₅R may be partly responsible for the increased expression of renal and vascular AT₁R that is due to impaired SNX1-mediated AT₁R sorting and degradation [89,90,97].

Vascular remodeling is both a cause and a consequence of hypertension [100]. Integrins, participating in the pathogenesis of vascular remodeling [101,102], are regulated by SNXs via different mechanisms. In endothelial cells, SNX9 determines the membrane levels of β₁-integrins, an essential adhesion molecule; SNX9 depletion delays the β₁-integrin cellular recycling and reduces the cell surface levels of β₁-integrins, which are involved in the adhesion of VSMCs and endothelial cells [103,104]. SNX17 also enables β₁-integrins to recycle to the cell surface and attenuates its transport to the lysosome for its decay by interacting with the β₁-integrin’s tail [49]. SNX31, an SNX sharing with SNX17 their obligate phox domains, also modulates β₁-integrin level at the membrane by binding to its tail [105]. Thus, SNXs, including SNX9, SNX17, and SNX31, can directly regulate the trafficking and expression of integrins, which may be involved in the modulation of vascular remodeling. However, more studies are needed to determine the role of SNX-regulated cellular localization of integrin in the pathogenesis of hypertension.

SNXs regulate vascular remodeling via other mechanisms. For example, SNX16 is also involved in the development of other types of hypertension, including pre-eclampsia [106]. Inflammation and immune mechanisms are involved in the pathogenesis of pre-eclampsia. SNX10 has been reported to regulate the inflammatory response. Snx10 deficient mice have increased basal and lipopolysaccharide-induced increase in anti-inflammatory M2 macrophages but decrease in M1 pro-inflammatory macrophages, which were verified by in vitro studies [107]. However, the role of different monocyte/macrophage populations in the pathogenesis of hypertension is complicated [108,109]. Thus, the M1 phenotype may predominate early while the M2 phenotype may predominate late in the disease process; the increase in M2 macrophages may then lower the blood pressure level [109]. SNX10 also modulates diet-induced atherogenesis by the Lyn-dependent transcription factor EB pathway [110]. Compared with wild-type (WT) mice, Snx10^−/− mice have attenuated atherosclerotic plaque progression, as well as reduced foam cell amount [111]. In addition, SNX17 interacts with an adhesion protein P-selectin via its FERM domain, leading to the acceleration of cargo internalization and inhibition of the transport of P-selectin to lysosomes for its decay [112–114]. However, the roles of SNX10 and SNX17 on blood pressure regulation have not been determined.

Regulation of SNXs on cellular sodium and water handling

As aforementioned, essential hypertension may be initiated by an increase in basal renal tubular sodium reabsorption but an appropriate decrease in renal tubular sodium reabsorption and increase in sodium excretion does not occur with the increase in blood pressure [63,115,116]. SNXs can directly affect cellular renal sodium handling. For example, SNX4 is required for the trafficking of the epithelial sodium channel (ENaC), which is involved in sodium reabsorption in the late distal convoluted tubule, connecting tubule, and collecting duct of the kidney [117–120]. SNX4-mediated trafficking of ENaC may involve the kidney and brain protein, KIBRA [121]. Depletion of Snx27, a retromer-associated protein, does not alter ENaC activity [121]. However, SNX27 can directly interact with the C-terminus of NHE3 to regulate its trafficking to the renal brush border membrane [42]. SLC26A3, aka down-regulated adenoma protein (DRA) and chloride anion exchanger, with NHE3 mediate intestinal sodium absorption [122]. SNX27 regulates the activity of DRA by accelerating its recycling, but not its expression, to the membrane in intestinal cells [123].

There are other SNXs participating in the modulation of sodium and water handling. SNX11 facilitates the movement of transient receptor potential cation channel subfamily V member 3 (TRPV3) to lysosomes for its degradation [46]. SNX3 increases the cell surface expression of ENaC in HEK-293 cells and is required for the increase in ENaC cell surface expression mediated by vasopressin in mouse renal cortical collecting duct cells [124]. There is a co-localization of SNX27 and the water channel aquaporin 2 protein (AQP2) in the kidneys of rats. Further studies showed that in the rat renal collecting duct, the autophagy-lysosomal degradation of AQP2 is prevented by SNX27 [125]. Current studies have shown that SNXs, including SNX3, SNX4, SNX11, SNX27, can directly regulate the expression and/or activity of cellular sodium and water handling proteins by modulating their trafficking. However, the role, if any, of these SNXs in the regulation of blood pressure needs further studies.

SNXs can also indirectly regulate renal sodium handling. As discussed above, the depletion of renal Snx1 or Snx19 leads to increased activity of sodium pump, exchanger, and cotransporters by decreasing the D₅R or D₁R function [80,89,90].

SNXs and insulin resistance

Insulin resistance, a major metabolic risk factor, is involved in the pathogenesis of salt-induced hypertension [126]. Insulin production is regulated by SNXs, such as SNX1, SNX19, and SNX27. Both the amount of dense core vesicles (DCVs) and insulin production in β cells are reduced after Snx19 silencing, which are reversed with the re-expression of Snx19. These may be associated with the lysosome pathway because Snx19 silencing, in a mouse pancreatic β-cell line, increases the number of lysosomes, as well as the activity of cathepsin D, a lysosome enzyme [127]. SNXs also indirectly regulate insulin secretion. For example, SNX19 interacts with insulin antigen 2, part of DCVs that regulate insulin production, in part, by initiating the pre-apoptotic state of pancreatic β cells [128,129]. In mouse pancreatic β-cells, incretin-mediated insulin production is also increased after SNX27 or SNX1 silencing [130].

SNXs, including SNX1, SNX2, SNX4, SNX5, SNX9 and SNX27, can also regulate insulin-mediated signaling by regulating insulin receptor expression and its downstream signaling. SNX5, similar to SNX19, also regulates circulating insulin levels. The kidney is the main organ involved in the elimination of circulating insulin [131,132]. Our studies, and those of others, found that renal SNX5 positively regulates the expression and function of insulin-degrading enzyme (IDE), which is thought to be responsible for insulin clearance [133,134]. Co-localization of SNX5 and IDE is observed in RPTs of humans and mice, and at the plasma membrane and perinuclear area of hRPTCs, which are increased after insulin activation [134]. Depletion of renal Snx5 decreases IDE expression and activity, impairs insulin excretion from the urine, and responses to insulin and glucose, leading to the occurrence of insulin resistance. The levels of both SNX5 and IDE are also reduced in RPTCs from SHRs and hypertensive subjects [134].

Our previous studies also found that the expression of the insulin receptor is regulated by SNX5. There is a co-localization of SNX5 and insulin receptor throughout the RPTs of the human kidney, but their co-localization is only observed in brush borders and apical membranes of rat RPTs [59]. Snx5 depletion in hRPTCs reduces the amounts of insulin receptor and phosphorylated IRS-1 and PKB [59], the downstream signaling molecules of the insulin receptor [135,136]. Immuno-coprecipitation studies also showed that other SNXs, including SNX1, SNX2, and SNX4, can interact with the insulin receptor [137,138]. SNX9 is also involved in insulin action by stimulating insulin-stimulated glucose transport [139]. Proteomic analysis of SNX27 and retromer cargo specificity found that the glucose transporter, glucose transporter 1 (GLUT1), a critical signaling molecule of insulin, interacts with SNX27 and requires SNX27-retromer to prevent its decay in the lysosomes and preserve plasma cell membrane levels [50,140]. Glucose transporter 4 (GLUT4) is also modulated by SNXs, such as SNX27 and SNX9. In 3T3L1 adipocytes, Snx27 silencing with Snx27-shRNA decreases the expression of GLUT4 and blocking SNX9 with its antibody also inhibits insulin-mediated translocation of GLUT4 [139,141].

SNXs and hypertension

The deficiency of specific SNXs leads to the hypertensive phenotype (Table 2). Both systolic and diastolic blood pressures are higher in anesthetized Snx1^−/− mice than WT mice [90]. Moreover, both nocturnal and diurnal systolic blood pressures are also elevated in conscious Snx1 knockout mice. The elevated blood pressure in Snx1^−/− mice is not salt-sensitive because there are no differences of blood pressure in Snx1^−/− mice fed with diets containing varying salt content [90]. The silencing of renal Snx1 also increases blood pressure in different strains of mice, e.g., C57Bl.6J and 129S1/SvlmJ mice [89]. The increased blood pressures in Snx1^−/− mice or renal Snx1-silenced mice are accompanied by impaired D₅R-dependent sodium excretion and increased AT₁R-mediated reactivity of mesenteric artery [89,90]. Our previous studies also showed that silencing of renal Snx5 elevates the already increased systolic blood pressure in SHRs, accompanied by a decrease in 24-hour sodium excretion [77]. Moreover, knockdown of renal Snx5, by reducing the amount of IDE in RPTs, augments the blood levels of both insulin and glucose, leading to insulin resistance and hypertension [134]. We also found that the silencing of renal Snx19 elevates the blood pressure in mice fed with normal salt diet [80].

Table 2.

Summary of the role of SNX isoforms in hypertension

SNX isoform	SNX variants or modifications	Effects of SNX Modification on the regulation of blood pressure or related function	SNX expression and function in human hypertension	References
SNX1	Renal-selective Snx1 silencing (siRNA)	Increases systolic and diastolic blood pressures, in salt-sensitive C57Bl/6J and salt-resistant BALB/cJ mice and impairs the natriuretic response to fenoldopam (D1R and D5R agonist) in salt-resistant BALB/cJ mice; Increases renal AT1R protein expression in C57Bl/6J and BALB/cJ mice	Snx1 silencing (siRNA) in hRPTCs impairs D5R trafficking and function (stimulation of cAMP production and inhibition of sodium transport). Decreased mRNA and protein expressions of SNX1, impaired SNX1-mediated D5R endocytosis, and increased ROS in RPTCs from hypertensive Euro-American males, relative to normotensive males	[89,90]
	Global Snx1 knockout	Increases systolic and diastolic blood pressures and impairs the natriuretic response to fenoldopam in 129S1/SvlmJ mice; Increases AT1R protein expression in the kidney and aorta and reactivity of mesenteric artery to Ang II
	Snx1 variants	Not studied in rodents, as related to BP	Snx1 variants, rs11635627, rs11854249 and rs12591927, are associated with impaired ability of HCTZ to decrease systolic blood pressure in African-Americans	[90]
SNX5	Renal-selective Snx5 silencing (siRNA)	Increases further the increased systolic blood pressure and decreased sodium excretion in SHRs; Decreases renal IDE protein expression and urinary insulin excretion, and causes insulin resistance in C57Bl/6J mice; Increases non-fasting serum insulin and glucose levels in normotensive WKY rats	Decreased SNX5 protein expression in RPTCs from hypertensive Euro-American males, relative to normotensive subjects	[77,134]
SNX9	Snx9 variant	Not studied in rodents, as related to BP	Snx9 variant rs2364349 is associated with the heart rate lowering effect of β-blockers	[150]
SNX10	Myeloid cell-specific SNX10 deficiency	Snx10 depletion polarizes macrophages towards the anti-inflammatory M2 phenotype	None	[110,111]
SNX19	Renal-selective Snx19 silencing (siRNA)	Increases the systolic blood pressure and decreases renal D1R expression of C57BL/6J mice fed a normal salt diet	None	[80]

Abbreviations: AT₁R, angiotensin II type 1 receptor; β-blockers, β-adrenergic receptor blockers; BP, blood pressure; D₁R, dopamine D₁ receptor; D₅R, dopamine D₅ receptor; HCTZ, hydrochlorothiazide; IDE, insulin-degrading enzyme; ROS, reactive oxygen species; SHRs, spontaneously hypertensive rats; WKY, Wistar-Kyoto.

The renal expression of several SNXs is decreased in hypertensive states. Compared with normotensive subjects, the SNX5 protein expression is lower in RPTCs from hypertensive than normotensive subjects. This is consistent with the results in hypertensive animals; SNX5 expression is also lesser in the RPTCs and kidneys from SHRs than Wistar-Kyoto (WKY) rats [134]. Decreased renal SNX5 expression in SHRs may be a reason for their impaired ability to excrete a salt load. Indeed, silencing of renal Snx5 decreases sodium excretion and further elevates the already increased blood pressure of SHRs [77]. This is associated with impaired renal tubule cellular recycling of D₁R, increased levels of phosphorylated D₁R, and impaired D₁R function, for example, impaired D₁R-mediated increase in cAMP production [77]. We and others have reported that the renal D₁R is important in the natriuresis with salt loading [5,142,143]. We also found that both mRNA and protein expressions of renal SNX1 are lower in hypertensive than normotensive Euro-American males [90]. The decreased amounts of renal SNXs may be a cause in the impaired dopamine receptor-mediated renal sodium handling in hypertension. The decreased levels of SNX1 in RPTCs from hypertensive humans blunts D₅R trafficking, which leads to impaired renal D₅R-mediated cAMP production and suppression of Na⁺-K⁺-ATPase activity. Silencing of renal Snx1 in mice increases blood pressure that is related to the impairment of a D₁R and D₅R agonist fenoldopam-mediated natriuretic effect [90], similar to that found with Snx5 silencing [77].

The Snx5 variant, rs6045116, and the five Snx19 variants, rs3751037, rs3190345, rs681982, rs4414223, rs2298566, are not associated with hypertension, regardless of salt sensitivity status in Americans of European ancestry [144]. However, there are a few studies showing the association between SNXs and hypertension or metabolic syndrome in humans. A GWAS showed the association of Snx19 and proliferation and apoptosis of pancreatic β-cells, which are involved in insulin production [145]. In addition, some Snx1 variants are associated with the response to antihypertensive drugs (see below).

SNXs and antihypertensive treatment

Methods of targeting SNXs

Targeting of SNXs, using several methods, has led to the identification of their roles in different diseases. Gene deletion or depletion, via siRNA or shRNA, and gene insertion/overexpression are widely used in the study of the association of diseases and genes, including SNXs. For example, Snx10 deficiency decreases foam cell abundance, alleviates atherosclerotic plaque progression, and reduces inflammation and pathological damage induced by dextran sulfate sodium [111,146]. By contrast, gene overexpression via plasmid or adenovirus transfection is used to enhance SNXs-mediated physiological effects. For example, SNX8 overexpression in human liver cells markedly blocks hepatocyte lipid deposition [147]. SNX1 gene-rescue in hRPTCs from hypertensive subjects restores the impaired D₅R-mediated cAMP production and inhibition of renal sodium transport [90]. In addition, some novel methods, including nanoparticle-encapsulated Snx-shRNA plasmids with polylactide-polyglycolide or small-molecule inhibitor targeting of SNXs, have also been proved to be effective in altering gene expression [146,148]. However, it should be noted that while the biochemical role of SNXs appears largely conserved, the phenotypic consequences of loss of function may vary among species [149]. Thus, it is probably important to target the pathophysiological functions of aberrant SNXs in cells from hypertensive subjects, and not speculate or modify SNXs based entirely on the results of animal experimentations. Nevertheless, the associations between single nucleotide polymorphisms of SNXs and development of hypertension, as well as effects of antihypertensive treatment in several ethnic groups, have been reported (vide infra) [90,150].

SNX single nucleotide polymorphisms and antihypertensive effects

The potential role of SNXs in the treatment of hypertension has been demonstrated. Shahin et al. used GWAS to search genetic variants associated with heart rate during treatment with β-adrenergic receptor blockers (β-blockers) in Americans of European and African ancestry. They found that the Snx9 variant rs2364349 was associated with a decrease in heart rate following treatment with β-blockers. The decrease in heart rate was greater in subjects carrying Snx9 rs2364349 G/G and A/G genotypes than non-carriers [150].

We have also reported the association of SNX1 variants and response to antihypertensive drugs. We found twelve SNX1 variants in subjects with African- American ancestry and ten in subjects with European-American ancestry. In African-Americans, the SNX1 variants rs11635627and rs11854249 were associated with a decrease in systolic blood pressure with treatment with the diuretic hydrochlorothiazide [90]. We also found a trend of a decline in blood pressure with hydrochlorothiazide treatment in hypertensive patients with the SNX1 variant allele rs12591927 [90]. These data indicate that the SNX genotype may be predictive of the effect of anti-hypertensive treatment, a case of pharmacogenomics.

SNXs as potential antihypertensive targets

As discussed above, SNXs, such as SNX5 and SNX19, play roles in the development of hypertension in animal models. SNX27 may be a potential target of antihypertensive drugs because the recycling of β₂-adrenergic receptors back to the membrane is blunted after Snx27 silencing [151]. β-adrenergic blockers are used in the treatment of hypertension [152]. Other SNXs, such as SNX3, SNX10, SNX13 and SNX17, reported to be associated with cardiovascular diseases, may also be antihypertensive targets [14,110,153–155]. Indeed, SNXs are associated with risk factors, e.g., inflammation, involved the development of hypertension. For example, impaired SNX9 expression has been reported in different animal models of chronic inflammation, including chronic infection (Leishmania donovani), and autoimmune disease (rheumatoid arthritis) [156]. Hypertension is associated with chronic inflammation [157,158]. Thus, SNXs associated with chronic inflammation may be also involved in the pathogenesis of hypertension. However, the direct correlation between the above-mentioned SNXs and hypertension or antihypertensive effects needs to be clarified in the future.

SNXs and anti-cancer drug-induced hypertension

Angiogenesis, the formation of new blood vessels, is involved in the development and progression of many diseases, including cancers [159]. Some anti-cancer drugs, including monoclonal antibodies that inhibit angiogenesis or proliferation, can increase blood pressure [160,161]. Thus, it is important to find new strategies to decrease the anti-angiogenesis drug-induced hypertension with anti-cancer therapy. We suggest that targeting SNXs may be a potential strategy to reduce the cancer-associated hypertension for three reasons:

1. SNXs are associated with carcinoma progression. For example, SNX1 serves as a tumor suppressor and potential prognostic marker in colorectal cancer and gastric cancer [162,163]. Snx1 depletion in colon cancer cells increases cell proliferation and decreases apoptosis, which would promote colon tumorigenesis [164]. SNX5 expression is increased in hepatocellular carcinoma, which promotes vascular invasion and intrahepatic metastasis [165]. Snx5 depletion decreases cell proliferation, migration, and invasion [165].

2. SNXs regulate the expression and trafficking of some growth factor receptors, which are involved in the regulation of both cancerogenesis and blood pressure. For example, EGFR, a transmembrane glycoprotein belonging to the ERBB family of tyrosine kinase receptors, plays a vital role in signal transduction and oncogenesis [166,167]. Abnormal endocytosis and trafficking of EGFR contribute to aberrant EGFR signaling in cancer [168]. Overexpression of Snx1 promotes EGFR degradation in lysosomes and decreases the amount of EGFR on the cell surface [44,45]. SNX2, which has a redundant function with SNX1, can also participate in the lysosomal sorting of EGFRs [169,170]. SNX5 interacts with EGFR in hepatocellular carcinoma cells; SNX5 promotes cell proliferation, migration, and invasion, via the activation of EGFR by blocking EGF-mediated EGFR internalization, which is reversed by the knockdown of EGFR [165]. Other SNXs, including SNX3, SNX6 and SNX16, directly regulate the intracellular trafficking and sorting of EGFRs [171–173].

3. SNXs sensitize cancer cells to anti-angiogenesis drugs by regulating the activity and trafficking of EGFRs. For example, gefitinib, a specific inhibitor of EGFR tyrosine kinase, is known to repress the activation of EGFR signaling required for cell growth in non-small cell lung cancer (NSCLC) cell lines. In the gefitinib-sensitive NSCLC cell line PC9, there is efficient endocytosis of phosphorylated EGFR which leads to its degradation. However, in gefitinib-resistant NSCLC cell lines, the internalized phosphorylated EGFR accumulates in early endosomes, not in late endosomes/lysosomes for degradation; this process can be normalized by Snx1 depletion [174–176]. However, other studies have reported that SNX1 actually enhances the degradation of EGFR [44,177]. Snx2 silencing in human lung cancer cells enhances the sensitivity to gefitinib and erlotinib, another anti-angiogenic drug that targets EGFR [178]. Snx5 overexpression promotes while its silencing inhibits hepatocellular carcinoma cell proliferation, migration, and invasion that are related to activation of EGFR [165]. As is the case with Snx2 silencing, Snx5 silencing increases the sensitivity to EGFR inhibitors, e.g., erlotinib and sorafenib [165]. These findings suggest that targeting SNXs may be useful in overcoming drug resistance to EGFR-targeted drugs in cancer cells. Moreover, the SNXs-mediated increase in sensitivity of cancer cells to anti-angiogenesis drugs may lower the dosage of these anti-cancer drugs and reduce the risk of drug-induced high blood pressure. However, there is lack of direct evidence to show the effect of targeting SNXs on the regulation of anti-angiogenesis drug-induced hypertension, which remains to be determined.

Conclusion

In summary, SNXs exert their biological functions by regulating cellular cargo sorting and trafficking, maintaining intracellular protein homeostasis, and participating in intracellular signaling cascade. Increasing pieces of evidence demonstrate that SNXs exert important functions to maintain a normal blood pressure, via different mechanisms (Figure 1). Dysfunction of SNXs in the cardiovascular system and kidney participate in the development of hypertension. Moreover, genetic studies have also shown an association of SNXs gene polymorphisms and responses to anti-hypertensive drugs. However, there remain some important questions in the role of SNXs in blood pressure regulation. For example: 1) the role of SNXs in the development and treatment of hypertension needs to be studied further; 2) most studies focus on the regulation of SNXs on the trafficking of intracellular cargos. There are few studies exploring how the expression and function of SNXs are regulated [179]; 3) in addition to the kidney and vasculature, the role of SNXs in other blood pressure-related organs or target organs in hypertension needs to be determined; and 4) further studies on the crystal structure of SNX domains, SNX-mediated membrane remodeling, and interaction with their targets may be helpful in understanding SNX regulatory mechanisms in hypertension and anti-hypertensive treatment [180,181]. Such studies will improve our understanding of the role of SNXs in the regulation of blood pressure and provide novel strategies for the treatment of hypertension.

Figure 1. Regulation of blood pressure by SNXs.

SNXs regulate blood pressure via different mechanisms. In the artery, SNX1 negatively regulates AT₁R-mediated vasoconstriction by increasing AT₁R-mediated proteasomal degradation; SNXs9/17/31 promote the plasma membrane recycling of β₁-integrin; SNX10 promotes the polarization of macrophages towards the M_I pro-inflammatory phenotype. In the kidney, SNXs1/5/19 promote D₁R- and D₅R-induced natriuresis. SNXs regulate cellular sodium and water transport: SNXs3/4 regulate the trafficking of ENaC, increasing its membrane expression and therefore activity; SNX27 increases the activity of NHE3, DRA (aka SLC26A3), and AQP2 by increasing their recycling to the plasma membrane; and SNX11 facilitates the movement of TRPV3 to lysosomes, increasing its degradation. SNXs regulate insulin production and response: SNXs1/19/27 increase insulin secretion; SNXs1/2/4 co-immunoprecipitate with the insulin receptor; and SNX5 negatively regulates the expression of the renal insulin receptor and insulin clearance by increasing the expression and function of IDE. SNXs9/27 increase the expression or plasma membrane translocation of GLUT1/4. Dysfunctions of SNXs participate in the pathogenesis and pharmacogenomics of hypertension. The response to antihypertensive drugs is influenced by variants of some SNXs. For example, the Snx9 variant rs2364349 is associated with the decrease in heart rate following treatment with β-blockers; the decrease in systolic blood pressure, in response to the diuretic HCTZ, is associated with the SNX1 variants rs11635627and rs11854249; Snx2 silencing in human lung cancer cells enhances the sensitivity to the tyrosine kinase inhibitors, gefitinib and erlotinib, but may also cause hypertension.

Abbreviation: AT₁R: angiotensin II (Ang II) type 1 receptor; AQP2: aquaporin-2 protein; β-blockers: β-adrenergic receptor blockers; D₁R: dopamine D₁ receptor; D₅R: dopamine D₅ receptor; DRA, down-regulated adenoma protein (also known as chloride ion exchanger and SLC26A3, solute carrier family 26 member 3); ENaC: epithelial sodium channel; GLUT1: glucose transporter type 1; GLUT4: glucose transporter type 4; HCTZ: hydrochlorothiazide; IDE: insulin-degrading enzyme; NHE3, sodium-hydrogen exchanger type 3; SNX: sorting nexin.

Abbreviations

Ang II

angiotensin II

AT₁R

angiotensin II type 1 receptor

AQP2

aquaporin 2

ATG

autophagy-related gene

BAR

Bin/Amphiphysin/Rvs

β-blockers

β-adrenergic receptor blockers

cAMP

cyclic adenosine monophosphate

DCVs

dense core vesicles

D₁R

dopamine D₁ receptor

D₅R

dopamine D₅ receptor

DRA

down-regulated adenoma protein (also known as chloride ion exchanger and SLC26A3, solute carrier family 26 member 3)

EGFR

epidermal growth factor receptor

ENaC

epithelial sodium channel

FERM

protein 4.1/ezrin/radixin/moesin

GLUT1

glucose transporter type 1

GLUT4

glucose transporter type 4

GPCRs

G protein-coupled receptors

HCTZ

hydrochlorothiazide

hRPTCs

human renal proximal tubule cells

IDE

insulin-degrading enzyme

IR

insulin receptor

IRS-1

insulin receptor substrate-1

NADPH

nicotinamide adenine dinucleotide phosphate

NCC

thiazide-sensitive sodium-chloride cotransporter

NHE3

sodium-hydrogen exchanger type 3

NOX

NADPH oxidase

NSCLC

non-small cell lung cancer

PX

phagocyte oxidase (phox) homology domain

PDZ

postsynaptic density 95/discs large/zonula occludens

PI3P

phosphatidylinositol-3-phosphate

PXA

PX-associated domain A

PXC

PX-associated domain C

RGS

regulator of G-protein signaling

RPTs

renal proximal tubules

SHRs

spontaneously hypertensive rats

SNX

sorting nexin

TGF-β

transforming growth factor β

TRPV3

transient receptor potential cation channel subfamily V member 3

VSMCs

vascular smooth muscle cells

WKY

Wistar-Kyoto

WT

wild-type