Hydrochlorothiazide attenuates lithium-induced nephrogenic diabetes insipidus independently of the sodium-chloride cotransporter.

Abstract

Lithium is the most common cause of nephrogenic diabetes insipidus (Li-NDI). Hydrochlorothiazide (HCTZ) combined with amiloride is the mainstay treatment in Li-NDI. The paradoxical antidiuretic action of HCTZ in Li-NDI is generally attributed to increased sodium and water uptake in proximal tubules as a compensation for increased volume loss due to HCTZ inhibition of the Na-Cl cotransporter (NCC), but alternative actions for HCTZ have been suggested. Here, we investigated whether HCTZ exerted an NCC-independent effect in Li-NDI. In polarized mouse cortical collecting duct (mpkCCD) cells, HCTZ treatment attenuated the Li-induced downregulation of aquaporin-2 (AQP2) water channel abundance. In these cells, amiloride reduces cellular Li influx through the epithelial sodium channel (ENaC). HCTZ also reduced Li influx, but to a lower extent. HCTZ increased AQP2 abundance on top of that of amiloride and did not affect the ENaC-mediated transcellular voltage. MpkCCD cells did not express NCC mRNA or protein. These data indicated that in mpkCCD cells, HCTZ attenuated lithium-induced downregulation of AQP2 independently of NCC and ENaC. Treatment of Li-NDI NCC knockout mice with HCTZ revealed a significantly reduced urine volume, unchanged urine osmolality, and increased cortical AQP2 abundance compared with Li-treated NCC knockout mice. HCTZ treatment further resulted in reduced blood Li levels, creatinine clearance, and alkalinized urinary pH. Our in vitro and in vivo data indicate that part of the antidiuretic effect of HCTZ in Li-NDI is NCC independent and may involve a tubuloglomerular feedback response-mediated reduction in glomerular filtration rate due to proximal tubular carbonic anhydrase inhibition.