Enhanced expression and function of renal SGLT2 (sodium-glucose cotransporter 2) in heart failure: Role of renal nerves

Abstract

Background:

Recent clinical studies demonstrate that sodium-glucose cotransporter 2 (SGLT2) inhibitors ameliorate heart failure (HF). The present study was conducted to assess the expression and function of renal SGLT2 and the influence of enhanced renal sympathetic tone in HF.

Methods:

Four weeks after coronary artery ligation surgery to induce HF, surgical bilateral renal denervation (RDN) was performed in rats. Four groups of rats (Sham-operated control [Sham], Sham+RDN, HF and HF+RDN; n=6/group) were used. Immunohistochemistry and western blot analysis were performed to evaluate the renal SGLT2 expression. One week after RDN (five weeks after induction of HF), intravenous injection of SGLT2 inhibitor dapagliflozin were performed to assess renal excretory responses. In vitro, human embryonic kidney cells were used to investigate the fractionation of SGLT2 after norepinephrine treatment.

Results:

In rats with HF, 1) SGLT2 expression in the proximal tubule of the kidney was increased, 2) the response of increases in urine flow, sodium excretion and glucose excretion to dapagliflozin were greater, and 3) RDN attenuated renal SGLT2 expression and normalized renal functional responses to dapagliflozin. In vitro, norepinephrine promoted translocation of SGLT2 to the cell membrane.

Conclusions:

These results indicate that the enhanced tonic renal sympathetic nerve activation in HF increases the expression and functional activity of renal SGLT2. Potentiated trafficking of SGLT2 to cell surface in renal proximal tubules mediated by norepinephrine may contribute to this functional activation of SGLT2 in HF. These findings provide critical insight into the underlying mechanisms for the beneficial effects of SGLT2 inhibitors on HF reported in the clinical studies.

INTRODUCTION

Sodium-glucose cotransporter 2 (SGLT2) is localized in the proximal convoluted tubules of the kidney and responsible for 90% of glucose reabsorption.^{1, 2} It has been reported that the expression and activity of SGLT2 are increased in diabetes.^{1, 2} Based on this property SGLT2 inhibitors promoting glucose excretion are widely used to treat type 2 diabetic patients. To date, clinical studies suggest that SGLT2 inhibitors suppress the risk for hospitalization of heart failure (HF) in type 2 diabetic patients.^{3, 4} Further, recent clinical trials in both diabetic and non-diabetic HF patients with reduced ejection fraction demonstrated that SGLT2 inhibitors prevent worsening the HF condition.^{5, 6} Following these results, the clinical use of SGLT2 inhibitors for chronic HF patients, in addition to diabetic patients, has been approved.

Several possible mechanisms for the beneficial effects of SGLT2 inhibitors on HF have been postulated. First potential mechanism is mediated by the effects of SGLT2 inhibitor mediated osmotic diuresis inducing the reduction in the interstitial fluid volume to a greater extent than blood volume, providing better control of congestion without reducing arterial filling and organ perfusion in the HF condition.^7–10 Second potential mechanism is mediated by the effects of SGLT2 inhibitors improving cardiac function by promoting ketone body production and utilization to ameliorate myocardial energy supply^11–13, by reducing myocardial sodium and calcium overload^14–16, by promoting autophagy in the cardiomyocyte accompanied with the reduction in oxidative stress and normalization of mitochondrial function^{17, 18} and by alleviating the activation of the inflammasome in the cardiomyocyte.¹⁹ A third possible mechanism suggest that since SGLT2 inhibitors alleviate stress-induced hypoxia in the kidney to stimulate erythropoietin production, this in turn improves oxygen delivery in the HF condition.^{10, 20, 21}

These multiple mechanisms listed above probably contribute to the beneficial effects of SGLT2 inhibitors during HF condition. However, the pathophysiology of HF has been characterized by enhanced sympathetic drive and specifically renal sympathetic nerve activity.^22–24 Renal sympathetic nerve activation causes norepinephrine release to promote renin secretion, sodium absorption and increased renal vascular resistance, inducing sodium and water retention associated with HF.^25–27 There are some recent reports demonstrating a possible interaction between sympathetic nervous system and SGLT2, showing that norepinephrine upregulates SGLT2 expression in vitro.^{28, 29} However, the expression levels of renal SGLT2 in HF and its relevance to enhanced renal nerve activation remains to be explored. Such an examination would provide critical new insight into the possible mechanisms for the use of SGLT2 inhibitors specifically in the HF condition.

The present study was conducted to elucidate the expression and function of renal SGLT2 and its modification by an enhanced renal sympathetic tone in HF. First, we assessed the expression levels of renal SGLT2 and the effect of SGLT2 inhibitor on urine flow, sodium excretion and glucose excretion in rats with HF. Secondly, we investigated the changes in the expression and function of renal SGLT2 in the presence and absence of renal nerves in rats with HF. Finally, to gain further insight into the potential mechanisms involved in the regulation of SGLT2 by sympathetic nervous system we assessed the changes in expression and localization of SGLT2 in human embryonic kidney cells treated with norepinephrine directly, in vitro.

METHODS

The authors declare that all supporting data and detailed methods are available within the article and the Data Supplement.

Animals

All procedures used for this study were approved by University of Nebraska Medical Center Institutional Animal Care and Use Committee and conducted according to the National Institutes of Health guiding principles for the research involving animals. Male Sprague-Dawley rats weighing 220 to 250 g were purchased from Sasco Breeding laboratories (Omaha, NE). Animals were housed with a 12-hour light-dark cycle at ambient 22°C 30–40% relative humidity. Laboratory chow and tap water were available ad libitum.

Model of chronic HF in rats

Rats were randomly assigned to either a Sham-operated Control group or a HF group. HF was produced by left coronary artery ligation, as previously described.^{25, 26} The degree of left ventricular dysfunction and HF was determined by using both hemodynamic and anatomic criteria at the end of the terminal experiments conducted 5 weeks after ligation surgery. Rats with both left ventricular end-diastolic pressure (LVEDP) > 15 mmHg and infarct size > 30% were considered to be in chronic HF.^{22, 23, 30, 31} Sham-operated Control group underwent similar surgical process without coronary artery ligation. Extended additional details of the methods are available in the Data Supplement.

Chronic bilateral renal denervation (RDN)

Four weeks after ligation surgery, a cohort of the Sham and HF rats underwent RDN under anesthesia (2 to 2.5% isoflurane, gas vaporizer) as described previously.^{22, 23, 30} Briefly, the kidneys were exposed through a midline incision in the abdomen. Total bilateral RDN was achieved by cutting the visible renal nerves around the renal artery and vein, followed by painting the vessels with 95% ethanol bilaterally. The completeness of RDN was confirmed by measurement of renal tissue norepinephrine content. All acute experiments were performed 1 week after RDN (5 weeks after the coronary artery ligation surgery).

Measurement of renal cortex norepinephrine content

Renal cortex was homogenized using a solution of 1 mM ethylenediaminetetraacetic acid and 4 mM sodium metabisulfite in 0.01 N HCl. After centrifugation of the homogenates, norepinephrine concentration in the supernatants was measured by using a commercially available ELISA kit (Labor Diagnostika Nord, Nordhorn, Germany), following the manufacturer’s instructions. The sensitivity and limit of detection of the assay was 2.0 pg/ml.

Statistical analysis

The data are expressed as mean ± SD. Differences between groups and among groups were assessed by t-test, one or two-way ANOVA followed by Bonferroni multiple comparisons test for post hoc analysis of significance (Prism 7; GraphPad Software) as appropriate. P values <0.05 were considered indicative of statistical significance.

RESULTS

General morphological and hemodynamic characteristics

The general morphological and hemodynamic characteristics of the four groups of rats used in this study are summarized in Table 1. The heart weight and heart weight/body weight ratio were significantly higher in HF group compared with Sham group (1.39 g vs. 1.04 g [95% CI, 0.14 to 0.55], n=8, P=0.0005 and 3.94 vs. 3.05 [95% CI, 0.26 to 1.51], n=8, P=0.0029, respectively). RDN had no significant effect on these parameters in both Sham and HF groups. HF group had an average myocardial infarct of 33% of the left ventricle, while Sham group had no visible myocardial damage. RDN did not significantly change infarct size in HF group. HF group had significantly higher LVEDP compared with Sham groups (19.13 mmHg vs. 3.29 mmHg [95% CI, 9.73 to 21.95], n=8, P<0.0001), which was partially mitigated by RDN. HF group had significantly reduced +dP/dt and −dP/dt compared with Sham groups (5173.71 mmHg/sec vs. 7033.57 mmHg/sec [95% CI, −3464.25 to −255.46], n=8, P=0.0198 and −4025.57 mmHg/sec vs. −6190.71 mmHg/sec [95% CI, 949.21 to 3381.08], n=8, P=0.0004, respectively), which were partially alleviated by RDN. These data confirmed that rats in HF groups were experiencing cardiac dysfunction and that RDN may contribute to partial mitigation of this cardiac dysfunction. There was no significant difference in the weight of kidneys between groups. Kidney norepinephrine content was significantly greater in HF rats compared to Sham operated controls (333.75 ng/g vs. 153.18 ng/g [95% CI, 133.47 to 227.67], n=8, P<0.0001). RDN reduced the kidney content of norepinephrine to very low level in both Sham and HF rats, which confirms the completeness of the RDN procedure (Table 1).

Table 1.

Characteristics of Sham and HF rats

	Sham (n=8)	Sham+RDN (n=8)	HF (n=8)	HF+RDN (n=8)
Body weight: BW (g)	343.43±14.50	365.57±31.23	351.67±25.63	344.10±22.96
Heart weight: HW (g)	1.04±0.08	1.12±0.03	1.39±0.10‡	1.28±0.17†
HW/BW *1000	3.04±0.12	3.09±0.29	3.94±0.25†	3.71±0.51*
Infarct size (%)	0	0	33.33±3.04§	33.30±4.08§
LVEDP (mmHg)	3.29±1.03	3.60±0.80	19.13±5.53§	12.88±4.31‡‖
+dP/dt (mmHg/sec)	7033.57±1565.92	7348.80±577.32	5172.71±1004.93†	6336.17±585.65†‖
−dP/dt (mmHg/sec)	−6190.71±1262.91	−6409.80±684.89	−4024.57±307.28‡	−4836.33±560.91‡‖
Kidney weight (g)	1.18±0.11	1.30±0.24	1.08±0.13	1.08±0.08
Kidney NE content (ng/g)	153.18±27.30	9.03±7.66#	333.75±54.49§	17.51±8.94#

Values are means ± SD; n=8 for each group of rats. LVEDP, left ventricular end-diastolic pressure. +dP/dt, maximal slope of systolic pressure increment.

-dP/dt, maximal slope of diastolic pressure decrement. NE, norepinephrine.

^*P < 0.05,

^†P < 0.01,

^‡P < 0.001,

^§P < 0.0001 compared with Sham.

^‖P < 0.05,

^#P < 0.0001 compared with the group without denervation.

Increased immunostaining for SGLT2 in the kidneys from rats with HF

Figures 1 A and 1B show the representative images of immunohistochemistry for SGLT2 in the kidneys of the four groups of rats. The intensity of immunofluorescent staining for SGLT2 in the proximal convoluted tubule segments was significantly higher in HF compared to Sham rats (Figure 1C; 10.54 vs. 6.07 [95% CI, 0.95 to 7.99], n=6, P=0.0104). RDN reduced the intensity of immunofluorescent staining for SGLT2 in the HF group, but did not significantly alter it in the Sham group (Figure 1C).

Figure 1.

Immunohistochemistry by a diaminobenzidine staining (A) and a fluorescent staining (B) for sodium-glucose cotransporter 2 (SGLT2) in renal cortex of Sham, Sham+RDN, HF and HF+RDN rats. (Magnification, 200X in A and 400X in B) C: mean values of SGLT2 staining intensity in the kidneys from each group. n=6. *p=0.0104 vs. Sham., †p=0.0346 vs. without RDN.

Increased SGLT2 protein expression in the kidneys from rats with HF

Western blot analysis revealed that protein levels of SGLT2 in the renal cortex were significantly increased in the HF group compared to the Sham group (Figure 2; 0.83 vs. 0.58 [95% CI, 0.003 to 0.507], n=6, P=0.0468). RDN significantly reduced protein levels of SGLT2 in the renal cortex of the HF group (0.57 vs. 0.83 [95% CI, −0.511 to −0.007], n=6, P=0.0433), while did not significantly change that of the Sham group (Figure 2).

Figure 2.

Western blot analysis for sodium-glucose cotransporter 2 (SGLT2) in renal cortex of Sham, Sham+RDN, HF and HF+RDN rats. n=6. *p=0.0468 vs. Sham., †p=0.0433 vs. without RDN.

Enhanced sodium and glucose excretory responses to SGLT2 inhibitor in rats with HF

To assess the functional contribution of SGLT2 before and after RDN, urine flow, sodium excretion and glucose excretion responses to SGLT2 inhibitor dapagliflozin were measured. There were no significant differences in kidney weights among the groups (Table 1). The basal urine flow and sodium excretion before dapagliflozin injection were significantly lower in HF compared to Sham groups (cumulative urine flow at baseline; 44.59 μl/gkw vs. 74.10 μl/gkw [95% CI, −57.65 to −1.37], n=6, P=0.0379, cumulative sodium excretion at baseline; 8.68 μEq/gkw vs. 15.17 μEq/gkw [95% CI, −11.80 to −1.18], n=6, P=0.0137) (Figures 3 and 4). The basal glucose excretion was undetectable before dapagliflozin injection in both groups of rats (Figure 5). Dapagliflozin injection produced increases in urine flow, sodium excretion and glucose excretion in both groups of rats. These were significantly greater in HF compared to Sham rats after dapagliflozin injection (cumulative urine flow at 120 min; 1782.13 μl/gkw vs. 1137.01 μl/gkw [95% CI, 102.96 to 1187.29], n=6, P=0.0166, cumulative sodium excretion at 120 min; 226.99 μEq/gkw vs. 133.24 μEq/gkw [95% CI, 6.96 to 180.54], n=6, P=0.0316, and cumulative glucose excretion at 120 min; 114.44 mg/gkw vs. 69.39 mg/gkw [95% CI, 10.53 to 79.56], n=6, P=0.0083) (Figures 3, 4 and 5). These results indicated an enhanced functional SGLT2 activity in rats with HF. RDN significantly reduced dapagliflozin-induced urine flow, sodium excretion and glucose excretion in rats with HF (cumulative urine flow at 120 min; 1218.99 μl/gkw vs. 1782.13 μl/gkw [95% CI, −20.97 to −1105.30], n=6, P=0.0401, sodium excretion; 137.72 μEq/gkw vs. 226.99 μEq/gkw [95% CI, −2.49 to −176.06], n=6, P=0.0425, and glucose excretion; 77.78 mg/gkw vs. 114.44 mg/gkw [95% CI, −2.14 to −71.17], n=6, P=0.0351) (Figures 3, 4 and 5). RDN did not significantly change the SGLT2 inhibitor-induced urine flow, sodium excretion and glucose excretion in the Sham group.

Figure 3.

A. Urine flow in response to sodium-glucose cotransporter 2 (SGLT2) inhibitor dapagliflozin injection in Sham, Sham+RDN, HF and HF+RDN rats. n=6. *p<0.05, **p<0.01, ***p<0.001 vs. Sham., †p<0.05, ††p<0.01 vs. without RDN. B. Cumulative urine flow at baseline and 120 min after dapagliflozin injection in each group. DAP indicates intravenous injection of dapagliflozin. n=6. *p=0.0379 vs. Sham., †p=0.0431 vs. without RDN., ‡p=0.0166 vs. Sham., §p=0.0401 vs. without RDN.

Figure 4.

A. Sodium excretion in response to sodium-glucose cotransporter 2 (SGLT2) inhibitor dapagliflozin injection in Sham, Sham+RDN, HF and HF+RDN rats. n=6. *p<0.05, **p<0.01 vs. Sham., †p<0.05, ††p<0.01 vs. without RDN. B. Cumulative sodium excretion at baseline and 120 min after dapagliflozin injection in each group. DAP indicates intravenous injection of dapagliflozin. n=6. *p=0.0137 vs. Sham., †p=0.0198 vs. without RDN., ‡p=0.0316 vs. Sham., §p=0.0425 vs. without RDN.

Figure 5.

A. Glucose excretion in response to sodium-glucose cotransporter 2 (SGLT2) inhibitor dapagliflozin injection in Sham, Sham+RDN, HF and HF+RDN rats. n=6. *p=0.0332, †p=0.0102 vs. Sham., ‡p=0.0429, §p=0.0283 vs. without RDN. B. Cumulative glucose excretion at baseline and 120 min after dapagliflozin injection in each group. DAP indicates intravenous injection of dapagliflozin. n=6. **p=0.0083 vs. Sham., †p=0.0351 vs. without RDN.

Changes in blood pressure and HR responses to SGLT2 inhibitor in Sham and HF groups

SBP, DBP, MAP and HR data before and after administration of dapagliflozin in four groups of rats are displayed in Supplemental Table I. Although SBP, DBP and MAP in both Sham and HF groups were numerically decreased after injection of dapagliflozin, these differences between, “at baseline” and “after SGLT2 inhibitor” were not statistically significant. HR was not significantly changed after injection of dapagliflozin in both groups. SBP, DBP, MAP and HR responses to dapagliflozin were not significantly different between Sham and HF group. Further, RDN did not significantly affect SBP, DBP, MAP or HR at baseline, or affect SBP, DBP, MAP and HR responses to dapagliflozin in both Sham and HF groups.

Enhanced trafficking of SGLT2 after norepinephrine treatment in HEK293 cells

Western blot analysis demonstrated that norepinephrine treatment decreased the cytoplasmic levels of SGLT2 while increasing membranous levels of SGLT2, promoting the idea of translocation to the cell surface from the cytosol (Figure 6A). Consistent with these observations SGLT2 protein immunostaining, expressed as punctate patterns on the cell membrane (Figure 6B) was increased after norepinephrine treatment. There data using confocal imaging of HEK293 cells in vitro further support the concept of increased tendency of SGLT2 to migrate from the cytosol to membrane after treatment with norepinephrine. This trafficking of SGLT2 to the cell membrane might be a critical step for the activation of SGLT2 to promote sodium and glucose reabsorptions in rats with HF.

Figure 6.

A. Western blot analysis for sodium-glucose cotransporter 2 (SGLT2) protein in cytosolic and membranous fraction of human embryonic kidney 293 (HEK293) cells with or without 10 μmol/L norepinephrine (NE) treatment. n=6. **p=0.0059 vs. Control., ††p=0.0012 vs. Control. B. Immunofluorescent staining for SGLT2 in HEK293 cells with or without 10 μmol/L NE treatment. Top: Scale bars = 50 μm, bottom: magnified images of the boxed portion of top images. Scale bars = 100 μm. C. Western blot analysis for sodium-glucose cotransporter 2 (SGLT2) protein in cytosolic and membranous fraction of renal cortex in HF and HF+RDN rats. n=6. **p=0.0011 vs. without RDN., ††††p<0.0001 vs. without RDN.

Diminished translocation of SGLT2 to the cell surface after RDN in the kidneys of rats with HF

Western blot analysis demonstrated that RDN increased the cytoplasmic levels of SGLT2 while decreasing membranous levels of SGLT2 in the kidneys of rats with HF, suggesting that renal nerves contribute to promote the translocation of SGLT2 to the cell surface from the cytosol in HF (Figure 6C).

DISCUSSION

We have shown that SGLT2 expression in the proximal tubule of the kidney was significantly increased in HF. Consistent with these observations the response of increases in urine flow, sodium excretion and glucose excretion to SGLT2 inhibitor, dapagliflozin were greater in HF as a consequence of the enhanced presence of renal SGLT2. Further, RDN reduced the expression of renal SGLT2 as well as normalized the renal excretory responses to SGLT2 inhibition in HF. Direct application of norepinephrine on renal HEK293 cells, in vitro, resulted in an increased expression of SGLT2 and specifically promoted the translocation of SGLT2 from the cytosol to the cell surface. Moreover, RDN increased the cytoplasmic levels of SGLT2 while decreasing membranous levels of SGLT2 in the kidneys of rats with HF. Taken together these findings indicate that the exaggerated renal sympathetic nerve activation in HF^{22, 23, 26, 30} enhances the expression of SGLT2 protein trafficked to the proximal cell luminal membrane resulting in boosted functional activity of renal SGLT2 causing greater sodium retention in the HF condition (Figure 7).

Figure 7.

Proposed model for neural mechanisms in regulating renal sodium-glucose cotransporter 2 (SGLT2) expression and functional activity in heart failure (HF). Intrarenal sympathoexcitation induces norepinephrine release to enhance the expression of SGLT2 protein trafficking to the proximal cell luminal membrane resulting in boosted functional activity of renal SGLT2 causing greater sodium retention in the HF condition. Renal denervation (RDN) inhibits the renal sympathetic nerve mediated norepinephrine releases to normalize the renal SGLT2 expression and functional activity resulting in ameliorated sodium and fluid retention in HF. Red arrows indicate changes during HF and green arrows indicate changes after RDN. RSNA, renal sympathetic nerve activity.

The current study reports for the first time that renal SGLT2 protein expression is enhanced in rats with HF, induced by coronary artery ligation. There are several studies addressing the changes in SGLT2 expressions in various disease conditions.^{29, 32–36} The renal SGLT2 mRNA and protein expressions are increased in both db/db mice and Akita mice, as models of type 2 and type 1 diabetes, respectively.^{32, 33} In high fat diet-induced metabolic mice, renal SGLT2 protein expression is enhanced.^{29, 34} The renal SGLT2 mRNA and protein expressions are also increased in the spontaneously hypertensive/NIH -corpulent rat (SHR/NDmcr-cp), a genetic model of the metabolic syndrome.³⁵ Further, in uninephrectomised diabetic Otsuka Long-Evans Tokushima Fatty rats, a model of the obese type 2 diabetes, renal SGLT2 mRNA and protein expressions are upregulated and these are decreased by RDN to increase glycosuria.³⁶ These previous observations taken together with the current results, suggest that renal SGLT2 protein expression is possibly enhanced by renal sympathetic hyperactivity commonly associated with HF as well as metabolic disorders induced by high glucose and high fat diet. To address the direct interaction between sympathetic neurotransmitter norepinephrine and SGLT2 expression, we performed cell culture experiments using HEK293 cells. Direct treatment with norepinephrine enhanced protein expression and promoted translocation of SGLT2 to the cell membrane which is detected by western blot analysis and immunohistological staining. This finding is consistent with a previous report using the human renal proximal tubule cell line and an analysis by ELISA.²⁹

One previous study using the human renal proximal tubule cell line demonstrated that treatment with high glucose upregulated SGLT2 mRNA expression, which was modestly enhanced by exposure to norepinephrine.³⁶ These findings lead to a possibility that the production of SGLT2 is mainly governed by glucose levels, and secondarily that the localization of SGLT2 is modulated by norepinephrine levels. The adrenergic receptors are present on the proximal tubule where they are concentrated in the basolateral membranes.²⁷ Intrarenal sympathetic activation-induced norepinephrine release could stabilize the renal SGLT2 protein expression on the apical membrane, via the adrenergic receptor followed by the adrenergic signaling, affecting the posttranslational intracellular translocation, ubiquitination, internalization and degradation of SGLT2. To address the contribution of each of these various processes remains to be investigated.

In the current study, examination of renal SGLT2 function by using SGLT2 inhibitor dapagliflozin, showed an enhanced excretory responses of urine flow, sodium excretion and glucose excretion in rats with HF. The changes in urine flow, sodium excretion and glucose excretion responses to SGLT2 inhibitor are functional corollary of the increased expression of SGLT2 in the proximal tubules of rats with HF. It should be noted that this acute inhibition of SGLT2 did not affect arterial blood pressure or HR while showing increases in urine flow and sodium excretion, suggesting a lack of compensatory sympathoexcitation possibly due to an acute fluid reduction. These findings are consistent with previous studies^{2, 37} in humans and animal models of diabetes and chronic kidney disease, and appear to be favorable therapeutic treatment for HF.

RDN normalized renal excretory responses to SGLT2 inhibitor, accompanied with normalized protein expressions of renal SGLT2 in the proximal tubules of rats with HF. These results suggest that tonic renal sympathetic nerve activation contribute to the overexpression and enhanced function of SGLT2 in the kidneys of rats with HF. In Sham-operated control rats, RDN did not show any significant effects on either the expression of SGLT2 or the functional excretory responses to SGLT2 inhibition, suggesting that RDN could affect the expression and function of SGLT2 in disease conditions with tonic sympathoexcitation such as in HF but not in normal healthy condition. These findings of specific renal denervation are congruent with a previous report which shows that global chemical denervation with 6-hydroxydopamine reduces renal SGLT2 expression in neurogenic hypertensive Schlager mice.³⁸

The precise molecular mechanisms in the regulation of renal SGLT2 are not fully elucidated. It has been postulated that glucose levels in the glomerulus modulate SGLT2 expression via intracellular cAMP and PKA signaling pathway.^39–41 Additionally, it is important to note that inflammatory cytokines such as interleukin (IL)-6 and tumor necrosis factor-α (TNF-α) upregulate SGLT2⁴² and that norepinephrine stimulates IL-6 secretion in human kidney proximal tubule cell line.²⁹ It has been shown that renal nerves mediate renal inflammation through activation of T-cells and increase inflammatory cytokines.^{43, 44} The ability of RDN to attenuate renal inflammation and reduce inflammatory cytokines such as IL-6 and TNF-α has been well documented in hypertensive animal models such as Ang II-induced hypertensive mice and DOCA-salt-induced hypertensive rats.^44–46 Renal inflammation is also one of the common pathophysiological properties in HF.^{47, 48} Hence, another potential mechanism may be sympathoexcitation-induced renal inflammation which mediates enhanced SGLT2 expressions which leads to enhanced renal retentive functions that are ameliorated/normalized by RDN. To assess the potential interaction between renal inflammation and SGLT2 in HF remains to be examined.

Limitations

There are several limitations to be considered in the current study. First, we assessed renal SGLT2 function by an acute response to initial bolus injection of SGLT2 inhibitor, dapagliflozin under anesthetized condition. There have been several clinical studies reporting that diuretic effects of SGLT2 inhibitors are intense in acute phase of treatment and attenuated with chronic treatment in type 2 diabetes.^{8, 10} Hence, renal SGLT2 function could be changed after chronic administration of SGLT2 inhibitor. The effects of anesthesia on renal function also cannot be completely eliminated. However, we used Inactin as an anesthetic which was specifically chosen to minimize the effects of changes in arterial blood pressure and thus provide an ideal stable perfusion pressure to the kidneys to perform renal function study in this time frame.⁴⁹ Under this anesthesia a lot of renal function studies are operational^{23, 25, 26, 31, 50} and therefore being used in current study examining effects of SGLT2 inhibitor on renal hemodynamics and excretion. Second, we assessed the contribution of renal SGLT2 function with renal nerves in HF by administration of SGLT2 inhibitor alone. In practical clinical conditions the standard therapy for HF is using a combination of drugs including renin angiotensin aldosterone system (RAS) inhibitor, beta-blocker and/or diuretics. To further assess the additive and/or synergistic effects of SGLT2 inhibitor with conventional therapy for HF should be addressed in the future. Third, we used Sprague-Dawley rats to induce HF by coronary artery ligation. RDN surgery was performed 4 weeks after coronary artery ligation surgery and the acute experiments were performed 1 week after RDN (5 weeks after coronary artery ligation surgery). Strains of rats and time-course of surgeries could affect the severity of HF and the effectiveness of RDN.⁵¹ In our current study, LVEDP of Sprague-Dawley rats with HF is 19.1 mmHg at 5 weeks after coronary ligation. Also, we have previously shown that in Sprague-Dawley rats with HF, left ventricular end-diastolic volume was increased to 653 μl and left ventricular ejection fraction was decreased to 33% at 5–6 weeks post coronary artery ligation.⁵² Hence, HF was substantially severe to emphasize the results of our current study. Regarding the timing of RDN surgery and acute experiments performed in current study, renal nerves are reported to reinnervate the kidney after RDN by 39–47% four weeks after denervation in Sprague-Dawley rats.⁵³ Therefore, to minimize an influence of reinnervation after RDN, we performed renal function study using dapagliflozin at 1 week after RDN surgery, a time frame we were confident that reinnervation does not occur.

Conclusions

In summary, we have shown that the expression and activity of renal SGLT2 were enhanced in HF. Removal of tonic renal nerve activation normalized these changes in HF. Further, norepinephrine as a neurotransmitter of sympathetic nerves promoted the translocation of SGLT2 to the cell surface in HEK293 cells. Consistent with these observations in vitro, RDN increased the cytoplasmic levels of SGLT2 while decreasing membranous levels of SGLT2 in the kidneys of rats with HF. These results suggest a critical contribution for tonic renal sympathetic nerve activation in the regulation of renal SGLT2 and subsequent sodium and water retention associated with HF (Figure 7). Enhanced diuretic and natriuretic responses to SGLT2 inhibitor counteracting the enhanced effects of tonic activated renal nerves in HF may contribute in part to the beneficial effects of SGLT2 inhibitor on the treatment of HF reported in the recent clinical studies.^3–6

WHAT IS NEW?

• Sodium-glucose cotransporter 2 (SGLT2) expression in the proximal tubule in the kidney is increased in rats with chronic heart failure (HF).

• Diuretic and natriuretic responses to intravenous injection of SGLT2 inhibitor are enhanced in rats with chronic HF.

• Norepinephrine as a neurotransmitter of sympathetic nerves promotes the translocation of SGLT2 to the cell surface in human embryonic kidney 293 cells, in vitro.

• Renal denervation increases the cytoplasmic levels of SGLT2 while decreasing membranous levels of SGLT2 in the kidneys of rats with chronic HF.

WHAT ARE THE CLINICAL IMPLICATIONS? 　

• Enhanced diuretic and natriuretic responses to SGLT2 inhibitor are beneficial for the treatment of HF and may underly in part the mechanisms of the effects of SGLT2 inhibitors suppressing the hospitalization of HF reported in clinical studies.

• Renal denervation mitigates the enhanced expression and activity of SGLT2 and ameliorates the subsequent sodium and water retention associated with HF, suggesting the potential therapeutic use of renal denervation for HF patients.

Nonstandard Abbreviations and Acronyms

SGLT2

sodium-glucose cotransporter 2

HF

heart failure

RDN

renal denervation

LV

left ventricle/ventricular

LVEDP

left ventricular end-diastolic pressure

+dP/dt

maximal slope of systolic pressure increment

−dP/dt

maximal slope of diastolic pressure decrement

SBP

systolic blood pressure

DBP

diastolic blood pressure

MAP

mean arterial pressure

HR

heart rate

HEK293 cells

human embryonic kidney 293 cells