New insights into the molecular regulation of urine concentration

the ability to excrete a concentrated urine was a critical future in evolutionary development, and it continues to be necessary for humans to maintain health. Urine concentration is a complex process, and requires interaction between multiple renal epithelial cell segments and the appropriate function of multiple proteins. Critical aspects of the roles of the thin descending and thick ascending limb of the loop of Henle, the connecting segment, and the collecting duct have been recognized and intensively studied over the past several decades. The inability to concentrate the urine appropriately is termed diabetes insipidus, and when it is due to kidney-specific defects it is termed nephrogenic diabetes insipidus (NDI) (1).

The recent study by Klein and colleagues (5) in an article in the American Journal of Physiology-Renal Physiology is an important addition to our understanding of the molecular mechanisms regulating urine concentration and has the potential for rapid clinical translation. Metformin is a medication widely used in clinical medicine for its glucose-lowering effects in patients with diabetes, and it is believed to work by activating the adenosine monophosphate-activated kinase (AMPK). Klein and colleagues show that AMPK is present in the renal inner medulla and that it phosphorylates two key proteins involved in urine concentration, aquaporin-2 (AQP2) and urea transporter (UT)-A1. In vitro microperfusion studies show that the AMPK agonist metformin increases inner medullary collecting duct (IMCD) water and urea permeability and that an alternative AMPK agonist, AICAR, has similar effects on urea permeability. Finally, they show that in a model of NDI, genetic deletion of the V2 receptor, that metformin increases urine concentration.

The molecular mechanisms through which metformin is having these effects are likely to involve activation of the cytosolic energy sensor, AMPK. AMPK is a heterotrimeric protein, composed of one α subunit, one β subunit, and one γ subunit (2), and Klein and colleagues show that both known α subunits, α-1 and α-2, are expressed in the inner medulla (5). AMPK is activated by AMP and ADP binding to the β subunit, and inhibited by ATP, thereby enabling it to function as a “cytosolic energy sensor.” However, AMPK is also regulated by a variety of other mechanisms, including intracellular Ca²⁺, liver kinase B (LKB), calcium-calmodulin-dependent kinase kinase (CaMKKβ), and transforming growth factor β-activated kinase (TAK1) (2). Although metformin is widely recognized to activate AMPK, its effects also include AMPK-independent mechanisms, including interaction with mTOR, PRAS40, and RAPTOR (7). Klein and colleagues do show the alternative AMPK activator, AICAR, has similar effects as metformin on IMCD urea permeability, supporting the likelihood that metformin is active through AMPK. However, it is important to recognize that metformin and AICAR can have similar biological effects, but through distinct, but differing, mechanisms that do not involve AMPK (7).

The studies add to a growing list of evidence that AMPK has important biological roles other than serving solely as a cytosolic “energy sensor.” AMPK has been shown to regulate multiple and transport processes in the kidney, including H-ATPase, NaPi2a, ROMK, ENaC, and KCNQ1 (8). It may be involved in the progression of chronic kidney disease, in renal cystogenesis in autosomal dominant polycystic kidney disease, and in acute kidney injury (3, 6, 9). Finally, it may be involved in renal cytokine responses (4). Of important concern, particularly as clinical implications of AMPK activation are examined, is the report that AMPK activation increases renal medullary cell apoptosis in both normally hydrated and dehydrated mice with type II diabetes mellitus (10).

Where do we go from here? First, Klein and colleagues are to be congratulated for identifying a novel, and potentially rapidly translatable, mechanism regulating AQP2, UT-A1, and urine concentration. Second, several important questions remain to be examined. For example, does metformin act through AMPK activation or through AMPK-independent effects? If AMPK is involved, which of the α subunits is necessary? Is metformin's ability to stimulate urine concentration specific for V2 receptor deletion-specific models of NDI or does it have the same effect in other causes of NDI? Why does metformin not routinely cause excessive urine concentration and lead to hyponatremia in patients who do not have diabetes insipidus? Finally, given that the most common form of NDI is due to lithium therapy, is not reversible with lithium discontinuation, and places patients at increased risk of hypernatremia, it is hoped that carefully performed preclinical and clinical trials of metformin therapy in first animals and then humans with lithium-induced NDI will be forthcoming.

GRANTS

Preparation of this editorial review was supported by funds from the National Institutes of Health (R01-DK045788 and R01-DK107798). This publication does not reflect official policy of the Department of Veterans Affairs.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

I.D.W. drafted manuscript; I.D.W. edited and revised manuscript; I.D.W. approved final version of manuscript.