Amiloride and the diabetic kidney

Diabetes is often associated with hypertension and alterations in sodium handling by the kidney. Multiple mechanisms may contribute to the reduced ability of the diabetic kidney to excrete sodium, including activation of the intrarenal renin–angiotensin system, particularly the Ang II-AT1 receptor axis, other peptidergic systems and stimulation of the sympathetic system. An important pathological consequence of diabetes is the progressive loss of glomerular integrity that leads to excessive filtration of a number of circulating proteins into the tubular fluid that contribute to or exacerbate tubular injury. These filtered proteins encompass various proteases such as plasmin that may potentially activate sodium channels on the apical face of the renal epithelium to facilitate sodium reabsorption. In the current issue of the Journal of Hypertension, Andersen et al. [1] sought to establish whether increased filtration of the protease plasmin in type 1 diabetic patients contributes to excessive sodium retention by activation of the epithelial sodium channel (ENaC). The authors assessed two diabetic cohorts with established nephropathy and those without nephropathy regarding urinary plasmin levels, as well as the blood pressure (BP) and sodium responses to the ENaC-inhibitor amiloride. The rationale for the current investigation is that excessive filtration of plasmin by the nephropathy patients would subsequently lead to greater activation of ENaC versus the non-nephropathy group with structurally intact glomeruli. These studies extend previously published work by the authors and others regarding the inappropriate filtration of proteases in various disorders that may lead to activation of ENaC [2–5]. Indeed, the present study established that significantly higher levels of plasmin protein are evident in the urine of the nephropathy patients. Both patient groups were maintained on a defined sodium diet of 200 mmol/day for 4 days followed by a 2-day treatment with amiloride. Urinary levels of amiloride were similar for both groups as determined at the end of the treatment period. Amiloride induced a comparable increase in the total and fractional excretion of sodium in both the nephropathy and non-nephropathy group; the nephropathy group exhibited mild hyperkalemia. The diuretic significantly reduced BP in both cohorts; however, the nephropathy group exhibited a larger decrease in pressure, as well as increased levels of plasma renin and aldosterone.

The regulation of sodium excretion by the kidney is facilitated by an array of sodium transporters and exchangers situated along the renal nephron [6]. The ENaC transporter is localized exclusively to the apical face of the principal cells of the collecting duct and is considered a key component of the tubule system that may discretely influence the final level of sodium output. Regulation of ENaC itself is quite complex and encompasses both genomic and various posttranslational mechanisms, as well as the overall coupling to the sodium, potassium-ATPase to facilitate the basolateral transport of sodium [7]. Polymorphisms were identified in the ENaC gene that prevented the ubiquination, internalization and subsequent degradation of the protein [8,9]. The inability to downregulate the ENaC protein leads to an inappropriate increase in sodium reabsorption and subsequent hypertension that is ameliorated by potassium-sparing diuretics such as amiloride and triamterene. These studies provided the molecular mechanism underlying the increase in BP in patients with Liddle’s disease, as well as emphasized the importance of ENaC in sodium handling [10].

The ENaC protein comprises three subunits (α–β–γ) that traverse the luminal membrane and form the sodium channel. The extracellular loops of both the β and γ subunits undergo limited proteolysis to activate the channel. Several proteases, including kallikrein, prostasin and plasmin, are reported to metabolize specific sequences on ENaC that may subsequently stimulate channel activity [11,12]. Plasmin typically circulates in an inactive proform termed plasminogen, and it is likely that plasminogen is primarily filtered in the diabetic nephropathy group [13]. However, plasminogen is processed to plasmin by another protease, the urokinase-type plasminogen activator (uPA), that also circulates in plasma and is found in the urine suggesting that it may be filtered as well [13]. The activation of plasmin by uPA is a key step in this pathway as amiloride directly inhibits uPA to block the processing of plasminogen, as well as blocks the channel to attenuate sodium entry. In the current study, the authors attempted to directly assess the extent of ENaC proteolysis prior to and following amiloride using an antibody to the inhibitory peptide sequence of the γ-subunit in exosomes that stained positive for the aquaporin-2 channel; however, both intact and the 37-kDa processed form of the ENaC γ-subunit were absent in the exosomes from the nephropathy and non-nephropathy patients.

The inhibition of uPA to block the conversion of plasminogen to its active form is a property not shared by other ENaC-directed diuretics such as triamterene. Thus, the surprising finding in the Andersen study that amiloride treatment significantly reduced the extent of proteinuria in the diabetic nephropathy patients may potentially reflect the glomerular effects of uPA and its interaction with the uPA receptor (uPAR). It should be noted that in the current study, amiloride was administered for only 2 days, and the reduction in proteinuria in the nephropathy groups may likely reflect a hemodynamic change rather than a marked influence on podocyte function or structure. However, apart from binding to uPA, uPAR apparently interacts with other proteins including β-3 integrin on podocytes, and activation of the uPAR-β-3 complex may contribute to podocyte effacement and proteinuria [14]. In experimental models of glomerular damage, amiloride, but not triamterene, attenuated the extent of glomerulosclerosis and proteinuria in the absence of a change in BP [15,16]. Moreover, amiloride treatment reduced the expression of uPAR and activation of β-3 integrin in glomerular podocytes in vivo and in isolated podocytes [15,16]. Finally, in a case study, Trimarchi et al. [17] report that amiloride reduced the urinary podocyte level in a patient with Alport syndrome, a disease with mutations in the collagen IV gene that leads to podocyte loss and early end-stage renal disease.

The present study demonstrates a significant effect of amiloride to reduce BP and stimulate sodium excretion in type 1 diabetic patients. Although the effects of amiloride were evident in both nephropathy and non-nephropathy groups, the study encompassed a small number of patients, and the treatment period for the diuretic was relatively brief. In lieu of these limitations, it may be premature to conclude that exposure of circulating proteases in the tubular fluid does not necessarily contribute to the activation of ENaC. Moreover, the potential for amiloride to attenuate proteinuria/albuminuria and glomerular injury in the diabetic kidney through inhibition of the uPAR-β-3 pathway should be a focus of continuing investigation for the novel actions of this diuretic.