Can Acetylation of Lysine 282 of Aquaporin 3 Improve Urinary Concentration in Mice with Lithium-Induced Nephrogenic Diabetes Insipidus?

Background: The principal cell basolateral water channel AQP3 is decreased in lithium-induced nephrogenic diabetes insipidus (NDI) and AQP3-null mice have an NDI phenotype with polyuria. We previously identified a novel AQP3 post-translational modification, acetylation of lysine 282 (K282), which we found to affect protein membrane localization. We hypothesized that lysine acetylation of AQP3 may improve the polyuria in NDI. Methods: Mice with homozygous mutations of AQP3 K282R (deacetylated mimetic, R) or K282Q (acetylated mimetic, Q) were compared to wild type K282 (WT) mice. Male and female 11–16-week-old mice were fed a lithium-containing diet or control diet (40 mmol LiCl or NaCl/kg dry food) for 14 days (n = 5-6/group). 24 h food and water intake, urine flow, and plasma osmolality at day 14 were measured. Results: On the control diet, food intake (WT 6.8±0.3, Q 6.3± 0.7, R 6.2±0.8 g/day, p=0.9), water intake (WT 3.6±0.3, Q 3.4±0.3, R 4.0±0.2, p=0.9), and urine flow (WT 1.4±0.4, Q 1.3±0.3, R 1.3±0.3 ml/day, p = 0.9) were similar among the male mice. The control diet females also had similar food intake (WT 6.4±0.4, Q 6.2±0.3, R 6.2±0.7 g/day, p = 0.38), water intake (WT 3.2±0.3, Q 3.4±0.2, R 3.9±0.4, p=0.9), and urine flow (WT 0.2±0.1, Q 0.5±0.1, R 0.5 ±0.2 ml/day, p=0.9). As expected, the lithium-diet fed mice developed polyuria/polydipsia. In male mice, food intake was lower in the WT (5.5±0.3 g/day) than Q and R (7.4±0.6 and 7.0±0.4 g/day, p = 0.02); however, R mutant mice had greater urine flow (12.8 ± 1.2 ml/day) than WT (4.8±0.5 ml/day) and Q mice (7.9±1.8 ml/day, p=0.0004). The females ate similar amounts of lithium diet (WT 6.4±0.3, Q 6.4±0.5, R 5.4±0.5 g/day, p=0.99) but urine flow was enhanced in Q (9.8±1.4 ml/day) and R (8.5±1.3 ml/day) mice compared to WT (4.8±0.5 ml/day, p=0.02). Water intake was significantly increased in all lithium-fed mice regardless of genotype (p<0.001). There were no significant effects of treatment or genotype on plasma osmolality among the groups. Conclusion: Contrary to our hypothesis, AQP3 acetylation- and deacetylation-mimetics had worsened polyuria in lithium-induced NDI, suggesting that K282 is an important regulatory site for AQP3 function perhaps independent of acetylation status. Funding: NIDDK Support