Victoria L Halperin Kuhns, Ava Zapf, Owen M Woodward
Affiliation(s):
University of Maryland School of Medicine, Department of Physiology
Hyperuricemia (HU) contributes to the development of gout, kidney stones, and kidney disease. Many studies exploring the effects of HU on renal disease progression have yielded inconsistent results based on assumptions that all types of HU have the same pathophysiological outcomes. Here, we compare the effects of underexcretion (UX) vs overproduction (OP) types of HU on renal gene expression in each nephron segment to provide insight into how HU may impact kidney disease. Both types of HU result in similar serum urate levels, but critically important differences in renal urate handling with disparate tubular urate levels. Our ABCG2 knock-in UX model (Q140K) has decreased urate secretion, with increased serum urate in male mice only and our OP model is a novel inducible knock-out of the urate metabolizing gene uricase (UOX-iKO), which renders mice unable to metabolize urate, increasing circulating urate levels. UOX-iKO male mice were induced at 9 weeks old. Mice had elevated serum urate levels similar to age matched Q140K male mice from 2 weeks to 6 months after induction. RNA-seq was performed on kidneys harvested from HU male animals from both models and controls, followed by DESeq2 differential expression analysis. Using published RNA-seq data from microdissected nephron segments, gene lists were compiled to create renal segment specific transcriptional profiles. Differentially expressed genes (DEGs) from both models were compared to segment specific profiles to determine which regions of the kidney were most enriched. UOX-iKO mice had increased overall urinary urate excretion (UUE) and increased fractional excretion of urate (FEU), consistent with urate OP. In contrast, Q140K mice showed no change in UUE and decreased FEU, consistent with urate UX. The two models of HU showed significantly divergent transcription profiles. We used an expanded subset of DEGs (DESeq2, p < 0.05) with 2012 DEGs in Q140K and 1107 DEGs in UOX-iKO, to capture changes in potentially all segments. The descending thin limb had the most DEGs in both models (766 and 479 respectively), whereas the proximal tubule showed the greatest differences (339 DEGs in Q140K and 112 DEGs in UOX-iKO). Although overlap between DEGs in both models was low, where it did occur, many of the genes were altered in opposite directions, mirroring the differences in tubular urate. To better understand renal transcriptional and physiological responses to alternative tubular urate levels, we focused on collecting duct water channels AQP2 and AQP3, necessary for water reabsorption to produce concentrated urine. Aqp2 and Aqp3 were up-regulated in Q140K and down-regulated in UOX-iKO. Following 24-hour urine collection, we found Q140K mice have increased urinary osmolarity and lower urine volume, while UOX-iKO mice have decreased urinary osmolarity and higher urine volume. These finding suggest tubular urate levels may be directly influencing urine concentrating mechanisms at the transcriptional level, altering the probability of urate precipitation. Understanding pathological effects of HU on kidney function and disease requires knowledge of the underlying cause. Here, we show that UX and OP result in significantly altered transcriptional responses to similar levels of elevated circulating urate, implying patients with UX or OP types of HU may respond differently to treatments. Thus, trials evaluating a potential causal role for urate in human kidney disease must take this into account.