Role of the Na+/H+ exchanger 3 in angiotensin II-induced hypertension in NHE3-deficient mice with transgenic rescue of NHE3 in small intestines

The role of Na+/H+ exchanger 3 (NHE3) in the kidney in angiotensin II (ANG II)-induced hypertension remains unknown. The present study used global NHE3-deficient mice with transgenic rescue of the Nhe3 gene in small intestines (tgNhe3−/−) to test the hypothesis that genetic deletion of NHE3 selectively in the kidney attenuates ANG II-induced hypertension. Six groups of wild-type (tgNhe3+/+) and tgNhe3−/− mice were infused with either vehicle or ANG II (1.5 mg/kg/day, i.p., 2 weeks, or 10 nmol/min, i.v., 30 min), treated with or without losartan (20 mg/kg/day, p.o.) for 2 weeks. Basal systolic blood pressure (SBP) and mean intra-arterial blood pressure (MAP) were significantly lower in tgNhe3−/− mice (P < 0.01). Basal glomerular filtration rate, 24 h urine excretion, urinary Na+ excretion, urinary K+ excretion, and urinary Cl− excretion were significantly lower in tgNhe3−/− mice (P < 0.01). These responses were associated with significantly elevated plasma ANG II and aldosterone levels, and marked upregulation in aquaporin 1, the Na+/HCO3 cotransporter, the α1 subunit isoform of Na+/K+-ATPase, protein kinase Cα, MAP kinases ERK1/2, and glycogen synthase kinase 3 α/β in the renal cortex of tgNhe3−/− mice (P < 0.01). ANG II infusion markedly increased SBP and MAP and renal cortical transporter and signaling proteins in tgNhe3+/+, as expected, but all of these responses to ANG II were attenuated in tgNhe3−/− mice (P < 0.01). These results suggest that NHE3 in the kidney is necessary for maintaining normal blood pressure and fully developing ANG II-dependent hypertension.


Introduction
Hypertension is well recognized as a chronic medical condition, affecting approximately one in three adults worldwide (Kearney et al. 2005;Lloyd-Jones et al. 2010). It has been estimated that up to 90-95% of patients with this medical condition have essential hypertension without clearly identified causes, whereas the remaining 5-10% have secondary hypertension with known genetic or endocrine disorders (Kearney et al. 2005;Lloyd-Jones et al. 2010). Although the etiologies and underlying mechanisms of essential hypertension remain incompletely understood, genetic, endocrine, dietary and environmental, and central neural factors have all been extensively investigated. For example, several genome-wide association studies (GWAS) have recently identified over 50 common genetic variants or loci associated with blood pressure or hypertension in humans, however the contribution of each of these common genetic variants to the development of hypertension is much smaller than expected (Levy et al. 2009;Newton-Cheh et al. 2009;Ehret 2010;Kato et al. 2011). Likewise, the role of high salt intake in the development hypertension has also been extensively investigated in salt-sensitive animal models, but 8% salt and deoxycorticosterone acetate (DOCA) are too often administered to induce hypertension. The relevancy of this model to human essential hypertension remains unknown, given that the U.S. Dietary Guidelines for Americans recommend eating a diet of 100 mmol of sodium a day or lower (Sacks et al. 2001;Chobanian et al. 2003;McCarron 2003). Furthermore, the role of increased renal sympathetic nerve activity in hypertension has also been extensively studied, but one of the most rigorously designed and randomized studies, the SYMPLICITY HTN-3 (Renal Denervation in Patients With Uncontrolled Hypertension) trial, has failed to show significant antihypertensive effect with catheter-based renal artery denervation (Bakris et al. 2015). Clearly, more studies are required to further uncover the mechanisms of essential hypertension.
Regardless of the precise and/or direct cause or factor, the development, progression, and maintenance of most, if not all, forms of hypertension appear to converge on a final common pathway, increased salt reabsorption or retention due to abnormal renal sodium (Na + ) handling involved with or without ANG II (Guyton 1991;Guyton et al. 1995). In the kidney, over 70% of the filtered Na + load is reabsorbed by the proximal nephron (Rector 1983;Weinstein 2008;Zhuo and Li 2013). Among all Na + transporters or cotransporters in different nephron segments, the sodium and hydrogen exchanger 3, NHE3, is considered to be the most important Na + transporter in apical membranes of the proximal tubule and loop of Henle (Lorenz et al. 1999;Wang et al. 1999;Vallon et al. 2000;McDonough 2010). NHE3 acts to secrete H + from the cells in exchange for luminal Na + entry, directly and indirectly contributing to up to 75% of Na + and 90% of HCO À 3 reabsorption in the proximal tubule Na + reabsorption and body acid-base balance (Aronson 1983;Rector 1983;Boron and Boulpaep 1989). It is therefore not difficult to extrapolate from the function of NHE3 that it may play a potential role in mediating sodium retention and hypertension in humans. However, the role of NHE3 in the kidney, especially in the proximal tubule, in the development of angiotensin II (ANG II)-dependent hypertension has not been investigated previously. Given that ANG II plays an important role in increasing the expression of NHE3 or activity in the proximal tubule (Houillier et al. 1996; Wang et al. 1999;Banday and Lokhandwala 2011;Li and Zhuo 2011), the present study test the hypotheses that NHE3 in the kidney is necessary for maintaining longterm blood pressure homeostasis, and that genetic deletion of NHE3 in the kidney attenuates the development of ANG II-induced hypertension in mice.

Animals
Heterozygous breeding pairs of NHE3 mutant mice with transgenic rescue of the Nhe3 gene in small intestines (tgNhe3 +/À ) were generously provided by Dr. Gary E. Shull of the University of Cincinnati College of Medicine (Woo et al. 2003;Noonan et al. 2005). Homozygous mutant tgNhe3 À/À mice were generated by breeding tgNhe3 +/À mice and genotyped in this laboratory according to Woo et al. (2003) and Noonan et al. (2005), respectively. Specifically, genotyping was carried out using the 5 0 -oligonucleotide primer sequence from the intestinal fatty acid binding protein (IFABP) promoter sequence (5 0 -CTGCCAGGTTATCTCTTGAAC-3 0 ), and the 3 0 reverse primer sequence from the NHE3 cDNA sequence (5 0 -CTGTTCGGTTCCTCCTCAATG-3 0 ). PCR conditions for genotyping were set as: 94°C for 3 min, then 35 cycles at 94°C for 30 sec, 60°C for 30 sec, and 72°C for 30 sec, followed by 72°C for 10 min (Woo et al. 2003;Noonan et al. 2005). All tgNhe3 À/À mice used in the present study carried the IFABP/NHE3 transgene in small intestines and had deletion of the Nhe3 gene primarily in the kidney (Woo et al. 2003;Noonan et al. 2005). Because very little NHE3 is expressed outside the digestive system and the kidney, tgNhe3 À/À mice were used as an alternative kidney-selective NHE3-KO mouse model, whereas agematched tgNhe3 +/+ littlemates were used as wild-type controls. The experiments as described in this study were approved by the Institutional Animal Care and Use Committee of the University of Mississippi Medical Center.
Angiotensin II-induced hypertension in tgNhe3 +/+ and tgNhe3 À/À mice Induction of ANG II-induced hypertension in tgNhe3 +/+ and tgNhe3 À/À mice was performed as we described previously (Zhuo et al. 2002;Li et al. 2007;Zhuo 2011, 2013). Briefly, three groups of adult male tgNhe3 +/+ and tgNhe3 À/À mice (n = 10-15 each group) were anesthetized with pentobarbital (50 mg/kg, i.p.) and implanted with an osmotic minipump (Model 2002) to infuse vehicle saline or a high pressor dose of ANG II (Bachem., Torrance, CA, 1.5 mg/kg/day, i.p.). Mice in Group 1 received saline and served as time controls. Mice in Group 2 were infused with ANG II for 2 weeks (Li et al. 2007;Li and Zhuo 2013). Mice in Group 3 were infused with ANG II as in Group 2 and concurrently treated with the AT 1 receptor blocker losartan (20 mg/kg/day, p.o.) for 2 weeks. Our previous studies have shown that at this dose of ANG II, blood pressure rose robustly and progressively in mice over the 2week period (Li et al. 2007;Zhuo 2008a, 2013). Thus, the tail-cuff technique was used to determine basal and weekly systolic blood pressure responses to ANG II or losartan (Zhuo et al. 2002;Li et al. 2007;Li and Zhuo 2008a). Additionally, 24 h fecal and urine samples were collected using metabolic cages to determine fecal and urinary Na + excretory responses in tgNhe3 +/+ and tgNhe3 À/À mice before and after ANG II was infused (Zhuo et al. 2002;Li et al. 2007;Li and Zhuo 2008a). Plasma, urinary, and fecal Na + , K + , and Cl À concentrations were determined by NOVA 13 Electrolyte Analyzer (Nova Biomedical, Waltham, MA).
Pressor responses to ANG II in anesthetized tgNhe3 +/+ and tgNhe3 À/À mice To further confirm the differences in basal systolic blood pressure and its responses to ANG II in conscious tgNhe3 +/+ and tgNhe3 À/À mice, three additional groups of tgNhe3 +/+ and tgNhe3 À/À mice (n = 6-8 each group) were anesthetized with Inactin (50 mg/kg wt., i.p.). The left carotid artery and jugular vein were cannulated with a polyethylene tube (PE50) with a fine fabricated tip to measure mean intra-arterial blood pressure (MAP) using a PowerLab Data Acquisition System (ADInstruments, Colorado Springs, CO) and to infuse saline containing 3% FITC-inulin to measure the whole kidney glomerular filtration rate (GFR) and 2% BSA (10 lL/min, i.v.) for 30 min before ANG II (10 pmol/min, i.v.) was infused for additional 30 min, as we described previously (Li and Zhuo 2008a;Li et al. 2009Li et al. , 2012b. Mice in Group 1 served as time controls. Mice in Group 2 received ANG II infusion only, whereas mice in Group 3 were pretreated with the AT 1 receptor blocker losartan (20 mg/kg, p.o, 3 days) before ANG II was infused to determine the role of the AT 1 receptor.
Measurement of plasma ANG II and aldosterone levels in tgNhe3 +/+ and tgNhe3 À/À mice Upon the completion of the experiment, blood samples were collected from tgNhe3 +/+ and tgNhe3 À/À mice upon decapitation in an inhibitor cocktail solution for measurement of plasma ANG II and aldosterone concentrations, as described previously (Zhuo et al. 2002;Li et al. 2007;Li and Zhuo 2011). Plasma samples were immediately extracted for measurement of plasma ANG II and aldos-terone levels using a sensitive aldosterone (Cat. 501090; Cayman, Ann Arbor, MI) or ANG II enzyme immunoassay kit (Cat. S-1133; Bachem.) (Li et al. 2007;Zhuo 2008a, 2011), respectively. Aldosterone and ANG II concentrations are expressed as pg/mL. Molecular, morphological, and reabsorptive phenotypes of the intestines and the kidney in tgNhe3 +/+ and tgNhe3 À/À mice To determine the intestinal and kidney phenotypes, the entire digestive system and the kidneys were collected from age-and body weight-matched male tgNhe3 +/+ and tgNhe3 À/À mice for comparisons of NHE3 mRNA and protein expression in the renal cortex, net weights, histology, and absorptive function, as described previously (Woo et al. 2003;Noonan et al. 2005). Briefly, the guts and kidneys were blotted dry and weighed, whereas fluid accumulation within the intestines, largely in the cecum, was collected from tgNhe3 +/+ and tgNhe3 À/À mice and Na + , K + , and Cl À concentrations measured accordingly (Woo et al. 2003;Noonan et al. 2005). The kidneys were sectioned and stained by Masson Trichrome for histological examination of the glomeruli and proximal tubule structures (Li et al. 2009;Li and Zhuo 2011).

Data analysis and statistics
All data are presented as mean AE SEM. The differences between tgNhe3 +/+ and tgNhe3 À/À mice in conscious systolic blood pressure, mean intra-arterial blood pressure, plasma aldosterone, plasma and kidney ANG II, and other intestinal and renal excretory responses were analyzed using Student's unpaired t-test. The differences between basal responses and weekly responses to ANG II in the same group of tgNhe3 +/+ and tgNhe3 À/À mice were first analyzed using one-way ANOVA followed by Dunnett's comparisons between experimental treatments. All statistically different levels in or between responses were set P < 0.05.
Morphological and basic excretory phenotypes of the kidney in tgNhe3 +/+ and tgNhe3 À/À mice Upon visual inspection, the kidneys looked not significantly different between tgNhe3 À/À mice and tgNhe3 +/+ mice, with similar kidney weight to body weight ratios (Table 1). At the light microscopic level, the size and structure of the glomeruli were also similar between tgNhe3 À/À and  (Li et al. 2009). U Na V, urinary Na + excretion. U K V, urinary K + excretion. U Cl V, urinary Cl À excretion. CF NA þ , cecum fluid Na + concentration. CF K þ , cecum fluid K + concentration. CF Cl À , cecum fluid Cl À concentration. ** P < 0.01 versus tgNhe3 +/+ mice. ND, not determined because no fluid accumulation in the cecum segment of tgNhe3 +/+ mice.

Discussion
Despite extensive investigations over several decades, the mechanisms of ANG II-induced or ANG II-dependent hypertension remain incompletely understood. Although the involvement of ANG II-induced increases in sympathetic neural activity, cardiac hypertrophy, and vascular constriction, renal vasoconstriction and salt retention has been extensively investigated, their respective contributions remain difficult to be defined. Recently, we investigated the role of global NHE3 in ANG II-induced hypertension using mutant mice with global deletion of NHE3 in target tissues including the kidney and the digestive system, Nhe3 À/À mice (Li et al. 2015). The results of that particular study demonstrated two key findings with potentially physiological as well as clinical implications. The first conclusion is that NHE3 is absolutely necessary for maintaining basal blood pressure homeostasis due to its physiological action on promoting Na + absorption from the small intestines and Na + reabsorption from the proximal tubule of the kidney. Genetic removal of NHE3 in all tissues led to a decrease in basal blood pressure by~13 AE 3 mmHg (Schultheis et al. 1998;Li et al. 2015). The second conclusion is that NHE3 is also required for the full development of ANG its responses to ANG II in conscious adult male tgNhe3 +/+ and tgNhe3 À/À mice. (A) Note that basal SBP were significantly lower in conscious tgNhe3 À/À mice (n = 12) than tgNhe3 +/+ mice (n = 15), and that SBP responses to ANG II (1.5 mg/kg/day, i.p., 2 weeks) were markedly attenuated in tgNhe3 À/À mice. **P < 0.01 versus their basal SBP. ++ P < 0.01 versus tgNhe3 +/+ during the same treatment. Losartan completely blocked ANG II-increased SBP in tgNhe3 +/+ mice (n = 8), but markedly reduced SBP significantly below their basal SBP in tgNhe3 À/À mice (n = 6). (B) Net increase in SBP in response to ANG II in tgNhe3 +/+ and tgNhe3 À/À mice, respectively. II-dependent hypertension. That conclusion is supported by the findings that ANG II-induced hypertension was significantly attenuated in global Nhe3 À/À mice. Thus, those results strongly suggest that NHE3 indeed plays a key role in the physiological regulation of blood pressure, and is involved in the full development of ANG IIinduced hypertension. However, limitations of that study are that global Nhe3 À/À mice were used to test the hypothesis. All Nhe3 À/À mice had moderate to severe intestinal phenotypes including diarrhea and fluid accumulation in the cecum segment between small and large intestines, resulting in marked salt wasting from the intestines (Schultheis et al. 1998;Li et al. 2015) and compensatory salt retention from the kidney (Lorenz et al. 1999;Wang et al. 1999;Li et al. 2015). Because Nhe3 À/À mice are a global NHE3 knockout model, previous studies were unable to separate the contributions between intestinal versus kidney NHE3 in blood pressure responses to endogenous or exogenous ANG II. To further test our hypothesis, we used a better genetic mouse model, namely global Nhe3 À/À mice with transgenic rescue of the Nhe3 gene selectively in small intestines, tgNhe3 À/À , using the intestinal fatty acid binding protein (IFABP) promoter in the present study (Woo et al. 2003). The rationale is that there is very little NHE3 expression beyond the small intestines of the digestive system and the kidney, so that  Figure 6. Comparisons of basal mean intra-arterial blood pressure (MAP) and its responses to acute ANG II infusion, 10 pmol/min, i.v., in anesthetized tgNhe3 +/+ (n = 7) and tgNhe3 À/À (n = 8) mice. (A) Basal MAP was significantly lower in tgNhe3 À/À mice (**P < 0.01).
(B) MAP responses to ANG II were significantly attenuated in tgNhe3 À/À mice. **P < 0.01 versus tgNhe3 À/À mice at the same time point. . Comparisons of the responses to ANG II (1.5 mg/kg/day, i.p., 2 weeks) in the expression of Na + /HCO 3 À , Na + /K + -ATPase a1, and AQP1 proteins in the superficial cortex of the kidney in tgNhe3 +/+ and tgNhe3 À/À mice (n = 6 for each strain). (A) Representative Western blots of selective transporter protein responses to ANG II. (B) Semi-quantitated data on the sodium and bicarbonate cotransporter. (C) Semi-quantitated data on the sodium and potassium ATPase a1 isoform. (D) Semi-quantitated data on Aquaporin 1. **P < 0.01 versus basal control tgNhe3 +/+ mice. ++ P < 0.01 versus control or ANG II-infused tgNhe3 +/+ mice. Note the complete lack of responses to ANG II in tgNhe3 À/À mice. . Comparisons of major signaling protein levels and their responses to ANG II (1.5 mg/kg/day, i.p., 2 weeks) in the renal cortex, primarily in the proximal tubules of tgNhe3 +/+ and tgNhe3 À/À mice. (A) Representative Western blots of selective signaling protein responses. Note that under basal conditions, phosphorylated protein kinase Ca isoform (B, p-PKCa), MAP kinases ERK1/2 (C, p-ERK1/2), and glycogen synthase kinase 3a/b (D, p-GSK3a/b) protein expression were all increased in the superficial cortex of the kidney, primarily the proximal tubules in tgNhe3 +/+ mice (n = 6 for each strain). While these signaling proteins were significantly increased by ANG II in tgNhe3 +/+ mice, there were no differences in these signaling proteins in tgNhe3 À/À mice with or without ANG II infusion. **P < 0.01 versus basal control tgNhe3 +/+ mice. ++ P < 0.01 versus basal control or ANG II-infused tgNhe3 +/+ mice. tgNhe3 À/À mice may be considered alternatively to be the kidney-selective NHE3 knockout mice (Woo et al. 2003;Noonan et al. 2005).
The key objective of the current study was not to evaluate basic intestinal and renal phenotypes such as the body salt and fluid status and tolerance of high salt intake of tgNhe3 À/À mice as reported previously by Woo et al. (2003) and Noonan et al. (2005). Instead, we specifically determined whether the development of ANG II-induced hypertension would be attenuated in tgNhe3 À/À mice with transgenic rescue of the NHE3 gene selectively in small intestines. The hypothesis tested was that NHE3 in the kidney plays a key role in ANG II-induced hypertensive responses. A direct interaction or relationship between ANG II and NHE3 in the kidney has been studied previously in cultured proximal tubule cells or the proximal tubules of the kidney. We and others have previously showed that ANG II significantly increases NHE3 expression and activity in cultured proximal tubule cells (Geibel et al. 1990;du et al. 2003;Li and Zhuo 2008b;Li et al. 2012a) or in the proximal tubule of the kidney (Banday and Lokhandwala 2011;Zhuo 2011, 2013). NHE3 is the most important Na + transporter in apical membranes of the proximal tubules (Lorenz et al. 1999;Wang et al. 1999;Vallon et al. 2000;McDonough 2010). Indeed, NHE3 not only directly contributes to about 25% of Na + reabsorption (Wilcox et al. 1992;McDonough 2010), but also acts indirectly to drive passive reabsorption of additional 50% of the filtered Na + load in the proximal tubules of the kidney (Rector 1983;Schafer and Robert 1984;Berry and Rector 1991). Hypertensive effects of nonpressor or slow pressor doses, but not acute high pressor doses, of ANG II are mediated at least in part by stimulating NHE3 expression/activity in the proximal tubules of the kidney (Banday and Lokhandwala 2011;Zhuo 2011, 2013). The results of the present study clearly confirmed our hypothesis by demonstrating that both systolic blood pressure responses to ANG II in conscious tgNhe3 À/À mice (Fig. 5) and intra-arterial MAP responses to ANG II in anesthetized tgNhe3 À/À mice (Fig. 6) were markedly attenuated.
Although the mechanisms underlying these attenuated hypertensive responses to ANG II in tgNhe3 À/À mice were not fully investigated in the present study, some plausible mechanisms were still uncovered. The roles of nonrenal and nonintestinal NHE3 may probably be minimal, because there is little evidence that a significant level of NHE3 is expressed in the blood vessels, the heart, the adrenal glands, or the brain. Since the functional Nhe3 gene was in theory replaced in small intestines of tgNhe3 À/À mice (Woo et al. 2003;Noonan et al. 2005), and NHE3 is primarily expressed in small intestines in the digestive system, our results may at least in part be attributed to the absence of NHE3 expression and function in the proximal tubule of the kidney in these mice. Indeed, only a small fraction of NHE3 is reportedly expressed in renal tubules beyond the proximal tubules of the kidney, primarily in the loop of Henle (Amemiya et al. 1995;Biemesderfer et al. 1997). The impaired or attenuated ANG II responses in tgNhe3 À/À mice was unlikely to relate to structural abnormalities in the kidney, because tgNhe3 À/À mice show basically similar glomerular and proximal tubular structures of wild-type animals (Fig. 3). By contrast, we recently found that the vascular pole of numerous glomeruli became enlarged or hypertrophic in global Nhe3 À/À mice, consistent with increased expression of kidney renin (Schultheis et al. 1998) and kidney ANG II levels (Li et al. 2015), and reportedly significantly reduced GFR (Ledoussal et al. 2001). However, the vascular pole of the glomeruli in tgNhe3 À/À mice looks similar to wild-type counterparts in the present study (Fig. 3). The third potential mechanism may be due to the incomplete rescue of the Nhe3 gene expression and function in small intestines of tgNhe3 À/À mice. The results of the present study largely replicated the basic intestinal phenotypes of these mice reported previously (Woo et al. 2003;Noonan et al. 2005). Although significantly improved from global Nhe3 À/À mice, small intestines of tgNhe3 À/À mice remained enlarged relative to tgNhe3 +/+ mice, the entire digestive system was still significantly heavier, a large volume of fluid was still accumulated in the cecum segment, whereas 24 h Na + excretion from the feces was still significantly increased in tgNhe3 À/À mice (Fig. 2). These results suggest that the rescue of the Nhe3 gene in small intestines of these mice may be incomplete. This may have led to continuous and significant salt wasting from the digestive system and subsequent causing body salt and fluid contraction in tgNhe3 À/À mice.
The most important mechanism underlying the attenuated hypertensive responses to ANG II in tgNhe3 À/À mice is most likely that the entire circulating and tissue reninangiotensin-aldosterone system (RAAS) is markedly activated in response to salt wasting from the gut in tgNhe3 À/À mice. Indeed, basal plasma ANG II and aldosterone levels were significantly elevated in these mice in the present study (Fig. 4). Our results provide further support to previous studies in which renal renin and serum aldosterone levels were markedly elevated in global Nhe3 À/À (Schultheis et al. 1998;Li et al. 2015) and tgNhe3 À/À mice (Woo et al. 2003). Although the expression of AT 1 receptors in the kidney of tgNhe3 À/À mice was previously found to be similar to that in wild-type mice (Noonan et al. 2005), these receptors in target tissues may be theoretically downregulated by high circulating ANG II. Alternatively, AT 1 receptors may be desensitized by high plasma ANG II, so that the hyperten-sive effects of ANG II were significantly diminished in tgNhe3 À/À mice (Figs. 5 and 6). Consistent with these interpretations are that under basal conditions with high plasma ANG II and aldosterone levels, the expression of several key water (AQP1) and Na + transporter or cotransporter proteins (Na + /HCO À 3 and the Na + /K + -ATPase a1), as well as some key signaling proteins (p-PKCa, p-ERK1/ 2, and p-GSK3a/b), in the proximal tubules of the kidney was markedly upregulated (Figs. 7-9). While the expression of all of these transporter or signaling proteins was responded robustly to ANG II stimulation in wild-type animals, these responses were all lost or attenuated in tgNhe3 À/À mice.
Taken together, the results of the present study strongly support the overall hypothesis that NHE3 is necessary for maintaining basal blood pressure homoeostasis and the full development of ANG II-dependent hypertension. However, it should be recognized that, despite its phenotypic and functional improvements from global Nhe3 À/À mice, our results suggest that tgNhe3 À/À mice remain to be an inadequate mutant mouse model to determine the relative role or contribution of proximal tubule or kidney NHE3 in the physiological regulation of blood pressure and the development of ANG II-dependent hypertension. Major renal functional phenotypes of global Nhe3 À/À mice persist in tgNhe3 À/À mice due to the incomplete rescue of the Nhe3 gene in the digestive system (Woo et al. 2003;Noonan et al. 2005). As salt wasting continues to occur in the gut, the kidney responds to the loss of NHE3, and subsequent activation of the RAAS and/or other salt-retaining peptides or hormones to mobilize other water and Na + transporters or cotransporters in the proximal and distal nephron segments. The marked decreases in basal GFR, 24 h urine excretion, and urinary Na + excretion observed in tgNhe3 À/À mice (Table 1) are consistent with this interpretation. Thus, a newer rodent model with conditional knockout of the Nhe3 gene selectively in the intestines or in the proximal tubules of the kidney using the Cre/LoxP approach ) may be necessary to further dissect the specific role of intestinal versus proximal tubule NHE3 in blood pressure regulation and the development of ANG II-dependent or independent hypertension.