mTORC2, a Novel “Twist” on the Epithelial Sodium Channel (ENaC)

Recent randomized clinical trials have demonstrated the benefits of modifying the ratio of sodium to potassium in the human diet, not just on blood pressure but also on “hard” outcomes, such as stroke and mortality.^1,2 These studies raise the potential for a ceasefire in the decades-long “salt wars” because they show how modest, and palatable, dietary changes can have profound effects on longevity. It is becoming clear that part, if not much, of the effect of a potassium-enriched diet to reduce blood pressure requires the kidney.

During the past 15 years, remarkable progress has also been made in understanding how the distal convoluted tubule (DCT) participates in potassium homeostasis and how aldosterone can be either kaliuretic or sodium-retentive, depending on the underlying stimulus (the “aldosterone paradox”). Work from several groups has identified a renal potassium switch, which involves DCT cells detecting a decrease in plasma [K⁺] through basolateral channels, secondary changes in intracellular chloride concentration, phosphorylation of specific serine and threonine kinases, and ultimately activation of the thiazide-sensitive NaCl cotransporter (NCC).³ The reverse takes place when potassium intake is high; in this case, NCC is dephosphorylated rapidly, by a complex mechanism involving protein phosphatase 1 and inhibitor 1.⁴ The relevance of these phenomena to human health and disease is evident because patients with Gitelman and Gordon (familial hyperkalemic hypertension) syndromes, diseases of the DCT, present with hypokalemia and hyperkalemia, respectively. Potassium-induced dephosphorylation also contributes to the natriuresis consequent to high potassium loading.

Yet, the DCT itself does not, for the most part, transport potassium; so, why does it have such a large effect on potassium excretion? It does so through a unique coupling with the connecting tubule and collecting duct, where potassium is secreted in exchange for sodium reabsorption through the epithelial sodium channel (ENaC). Aldosterone is the canonical ENaC-activating hormone, and both stimulates ENaC trafficking to the apical membrane and ENaC proteolytic cleavage to increase conductance. Aldosterone secretion itself is stimulated by a rise in plasma [K⁺], so hyperkalemia both inhibits NCC and stimulates ENaC. This leads to increased sodium delivery to a connecting tubule primed to secrete potassium. Part of this priming depends on serum and glucocorticoid-induced kinase 1 (SGK1), which particularly promotes ENaC trafficking to the apical membrane.

During the past several years, Pearce and colleagues have introduced a new “twist” on this scenario, one that involves the mammalian (or mechanistic) target of rapamycin (mTOR).⁵ mTOR can form a complex called mTORC2, which, unlike mTORC1, is NOT sensitive to the immunosuppressive drug rapamycin. In the kidney, mTORC2 seems to be necessary to phosphorylate and, therefore, fully activate SGK1. The Pearce group showed previously that inhibiting mTORC2 with two structurally distinct competitive inhibitors caused a natriuresis and reduced ENaC activity. In addition, they showed that mTORC2 inhibition affected the phosphorylation of SGK1, thereby placing mTORC2 within the aldosterone-SGK1-ENaC signaling pathway.

This view became controversial, however, when Huber and colleagues later generated mice in which mTORC2 activity was reduced, not pharmacologically, but rather by genetically deleting the rapamycin-insensitive companion of mTOR (RICTOR), another component of the mTORC2 complex, along the distal nephron.⁶ These mice exhibited elevated aldosterone and nearly absent phosphorylated SGK1, but they otherwise appeared normal. When challenged with a high potassium diet, however, profound hyperkalemia developed. Yet, these mice were able to reduce urinary sodium concentration when challenged with a low salt diet, suggesting that ENaC activity was intact, and patch clamp experiments showed low renal outer medullary K⁺ (ROMK) activity.

Using a “dialectical approach,” where disagreements are solved through rational dialog and experiment, Pearce and colleagues have now created a new mouse model in which RICTOR is deleted along the entire tubule and used this model to resolve the differences.⁷ Several aspects of the new approach differ from those taken by Huber and colleagues. First, these authors use the Pax8-Cre system, which effectively deletes proteins all along the renal tubule but only when induced by doxycycline, whereas Huber and colleagues used a system that deletes genes along most of the distal nephron constitutively. Second, Pearce and colleagues studied the effects of potassium loading acutely (after 4 hours) and after 48 hours, whereas Huber and colleagues studied the mice at 5 days. Finally, Pearce and colleagues do not recapitulate the dietary salt restriction experiments. Yet, many of the observations are concordant; the mice in both studies were relatively normal at baseline, with only elevated aldosterone indicating a secretory defect. Both models developed profound hyperkalemia after a potassium chloride challenge (in the first article, when combined with low NaCl intake), and both found that the abundance of phosphorylated SGK1 was quite low.

Yet, there were differences in both results and interpretation. Pearce and colleagues showed that ENaC inhibitable sodium excretion was not as large in the knock out (KO) animals as in controls. However, they note correctly that this measure is complicated by the linear structure of kidney tubules, so that upstream and possibly downstream effects complicate interpretation. Hence, they turned to patch clamp electrophysiology to study ENaC directly; although KCl gavage increased ENaC activity in control mice, it did not in the knockout. Interestingly, ROMK activity was not stimulated by KCl gavage in either group.

The results of the acute gavage experiments paint a clear picture of a signaling pathway involving ENaC, but by 48 hours of potassium loading, things are not as straightforward. At this time, although ENaC activity is still reduced, protein kinase C (PKC) and ROMK activity are lower in the KO mice, similar to what Huber reported at 4 days. Synthesizing the current results with previous results from the Pearce laboratory and results from Huber and colleagues, there seems to be an important role for mTORC2 in mediating rapid SGK1 phosphorylation in response to high potassium challenge. This is necessary for ENaC trafficking and the response to potassium and likely explains the clear inability of mice with inactive mTORC2 to tolerate potassium loading. Over the longer term, however, although ENaC activity still requires mTORC2 signaling, defects in PKC and ROMK activation develop, leading to a mixed picture.

The current results reflect rigorous efforts to solve the mysteries of aldosterone and potassium actions on the kidney distal tubule. One is left, however, hoping that this work continues, as several questions remain. First, the baseline phenotype of the RICTOR KO mice does not recapitulate that observed when the mineralocorticoid receptor is deleted, using the same Pax8-Cre system. In that case, basal plasma [K⁺] is elevated, and there is salt wasting. Furthermore, the abundance of cleaved αENaC is increased in the RICTOR KO mice, whereas both total and cleaved αENaC are reduced substantially in mice lacking mineralocorticoid receptors, a finding recapitulated by SGK1 deletion. Third, although it is clear that SGK1 plays an important role in activating ENaC, especially over the short term, deletion of SGK1 itself leads to a milder phenotype than mineralocorticoid receptor deletion, suggesting that other signaling molecules are important. Finally, and most importantly, it would be very important to know how these mice respond to low salt intake, and whether like the mice generated by Huber and colleagues, they can activate ENaC to prevent sodium loss. Although great progress has been made recently unraveling the aldosterone paradox, there is much more to learn. For example, the effects of mineralocorticoid receptor activation on ENaC might be different when aldosterone secretion is stimulated by potassium or when it is induced by angiotensin II. Nevertheless, it is now clear that mTORC2 plays an important role in signaling to ENaC in the distal tubule.

Disclosures

D.H. Ellison reports Honoraria: Renaissance School of Medicine and Boston University School of Medicine; Patents or Royalties: Author for UpToDate; and Advisory or Leadership Role: Consulting Editor for Hypertension, Editorial Board American Journal of Physiology-Renal Physiology, and Editorial Board of JASN. J.A. McCormick reports Advisory or Leadership Role: Editorial boards member of American Journal of Physiology-Renal Physiology, Frontiers in Physiology: Renal and Epithelial Physiology, and Kidney360.

Funding

None.

Author Contributions

Conceptualization: David H. Ellison.

Writing – original draft: David H. Ellison, James A. McCormick.

Writing – review & editing: David H. Ellison, James A. McCormick.