CRISPR/Cas9 Mediated Deletion of the Uox Gene Generates a Mouse Model of Hyperuricemia with Multiple Complications

Hyperuricemia, characterized by increased concentration of serum uric acid (SUA) above 420 μmol/L, has been an increasingly common metabolic disorder with a precipitous increase particularly among young adults [1, 2]. It is the major causal risk factor of gout and urate-nephropathy, and often entangles with a cluster of cardiometabolic disorders including hypertension, diabetes mellites, dyslipidemia and nonalcoholic fatty liver disease (NAFLD), which could jointly contribute to the development of cardiovascular disease [3]. However, the independent causal effect of hyperuricemia on these metabolic comorbidities and cardiovascular disease remains largely obscure [4]. Therefore, murine models to replicate early-onset hyperuricemia in humans is of paramount importance to better understand the progression trajectories and underlying mechanisms, and to develop effective medications.

Uric acid (UA) is an end product of purine metabolism that is specific to humans due to silencing of the gene encoding urate oxidase (Uox) or uricase during primate evolution. Unlike human, most mammals have active Uox that further degrades UA to allantoin, and therefore challenges exist in the establishment of efficient and sustainable animal model of hyperuricemia using pharmaceutical or dietary-induced approaches [5]. Given the major difference in Uox gene expression between human and mice, genetic modifications that target Uox either by homologous recombination in embryonic stem cells [6] or by using transcription activator-like effector nuclease mediated deletion have been used to generate Uox knockout (Uox^-/-) mouse model with spontaneously high SUA concentrations [7]. However, both embryonic lethality and postnatal mortality are the major hurdles of these Uox deficient mice. Alternatively, CRISPR/Cas9 is an advanced technology that facilitates more precise genetic engineering. The uricase-deficient “Kunming-DY rats” with stable hyperuricemia and better survival is an example of such an approach [8]. Herein, we established a Uox^-/- mouse model on C57BL/6J genetic background by deleting the exon 2-4 of Uox using CRISPR/Cas9 system. In this study, we present the hyperuricemic phenotypes regarding magnitude of serum UA elevation, urate nephropathy, cardiometabolic abnormalities, as well as cardiovascular and hepatic complications.

Despite many efforts to model hyperuricemia in human, a suitable mouse model that have stable hyperuricemia and normal lifespan has been particularly lack [9]. The present study introduced a novel Uox deficient mice with spontaneous hyperuricemia generated by CRISPR/Cas9 system on a pure C57BL/6J background. To improve postnatal survival, we also applied a strategy that combined genetic manipulation with allopurinol intervention, which enabled the SUA concentration not lethally high but maintain valid that mimics “human like” hyperuricemia. Moreover, this allopurinol-rescued Uox^-/- mice presented various complications including kidney disease, metabolic syndrome, cardiovascular disease and NAFLD. The present mouse model could provide a candidate tool to study early-onset hyperuricemia and associated comorbidities.

Hyperuricemia is caused by hepatic overproduction and (or) renal and (or) intestinal underexcretion of SUA, which are regulated by genetic and environmental factors and their interactions [10]. Obviously, owing to the evolutionary inactivation of Uox in humans, genetic modification of rodent Uox is an essential and most effective way to replicate human hyperuricemia and urate biology, as compared to genetically engineered mice targeting other loci (for example, SLC2A9 and ABCG2), as well as environmentally induced models [9, 11‐13].

The first Uox knockout model was generated on hybrid genetic background mouse (C57BL/6J*129Sv) by Neomycin-cassette insertion into exon 3 of the Uox that resulted in resulted in an elevated SUA concentration of 650 μmol/L, 12 times higher than that in the WT controls [6]. Lately, the second Uox^-/- mice was established on a C57BL/6J background by deleting 28 bp of exon 3 using TALEN technique, and the SUA levels were at 420 to 520 μmol/L, two to three times higher than in WT mice [7]. At present, ours represents the third Uox knockout mouse model so far, which has been generated on C57BL/6J background using CRISPR/Cas9 system for the first time. The exon 2 to exon 4 were targeted, which account for about 46% of the entire Uox protien coding gene. By deleting a bigger region (3296 bp) than the other models [6, 7], a frameshift mutation was expected with more significant phenotypic effect. In addition, deleting these three exons would result in less N-terminal amino acid residue. Beside, targeting other exons could affect the Dnase2b, since they coincide with the exons of Dnase2b. What’s more, off-target effect was highly unlikely given that their offspring presented the expected and stable biological features. Consequently, our Uox^-/- mice reached an extremely high SUA concentration of 1351.04±276.58 μmol/L, which was 5.5-fold higher than that in WT mice (248.19±100.59μmol/L), while allopurinol-treated Uox^-/- mice had moderately elevated SUA (612.55±146.98μmol/L), 2.5-time higher than that in WT mice. Due to lack of standard protocol for blood sampling and urate measurement, these reported SUA values varied widely without a clear definition of normal range in mice, and therefore were not directly comparable. Even though, the proportional increase in SUA suggested that present Uox^-/- mice yield a strong phenotypic effect on SUA level.

Without a gradual adaption to the evolutionary changes in urate mediated biological system [9], acute disruption in the Uox gene may largely explain the substantial embryonic and postnatal mortality in mice. The percentages of the Uox^-/- mice born from heterozygous mating were reported at 7.1% [6] and 15.9% [7], respectively, both far below the expected Mendelian frequency. By contrast, the birth rate in present Uox^-/- mice is 21.4%, approximated to the expected rate of 25%, suggesting the advantage of CRISPR/Cas9 technique in reducing embryonic lethality. However, neonatal mortality at four weeks after birth was extremely higher (92%) in the present model than the previous reports (40%~65%) [6, 7]. The same genetic approach targeting at exon 2-4 of Uox on Sprague Dawley rats resulted in a 95% survival up to one year, but only mild hyperuricemia, suggesting different responses to Uox inactivation between animal species [8]. Clearly, the extremely high SUA is the fundamental cause of premature death in the present Uox^-/-mice. With a 55% reduction in SUA by allopurinol intervention, they all could survive to eight weeks to be sexually matured.

The kidney is one of the important organs of urate metabolism, and studies have shown that patients with hyperuricemia are at increased risk of renal disease [3, 13]. As previous hyperuricemic mouse models [6, 7, 14, 15], our experimental results demonstrated severe renal insufficiency evidenced by remarkably increased levels of blood urea and serum creatinine, significant glomerular and tubular deformations accompanied by corticomedullar inflammation, interstitial fibrosis and uric acid crystal deposition. Given the fact that SUA concentration was even doubled in the Uox^-/- mouse prototype, we assume that severe urate nephropathy and renal failure in the first four weeks of life may account for their poor survival.

Epidemiological and experimental studies have shown that hyperuricemia is associated with hypertension, coronary atherosclerosis and heart failure [16‐23]. However, the causal relationship between UA and cardiovascular disease remains controversial. Our experimental mouse model of hyperuricemia manifested hypertension along with distinct ventricular remodeling, reduced cardiac output and significant aortic endothelial dysfunction, which are consistent with the literature [21‐23]. Owing to the renal failure, the independent role of hyperuricemia in the development of hypertension remains uncertain. Likewise, both kidney disease and hypertension may contribute to cardiac dysfunction and atherosclerosis, reiterating the issue of inconclusive causal relationship between hyperuricemia and cardiovascular disease. Notwithstanding, such a complex links enables our Uox^-/- mice as an appropriate model to study the progression trajectories of these comorbidities from birth to adulthood. Another appealing use is to detect whether new urate lowering drugs in addition to allopurinol could further ameliorate early-onset hypertension and cardiovascular complications as compared to allopurinol alone.

In addition to hypertension, our experimental mouse model of hyperuricemia manifested distinct hyperglycemia and dyslipidemia, all of which are typical to metabolic syndrome [18]. Onset sequence study showed that hyperuricemia is an earlier-onset metabolic disorder in relation to hypertension, hypertriglyceridemia, and diabetes mellitus, indicating a potential upstream role of hyperuricemia in the development of metabolic syndrome [1]. It was reported that high SUA could directly cause pancreatic β-cell apoptosis and dysfunction [24, 25]. The previous results from Uox^-/- mice also supported hyperuricemia was probably a causal factor of islet dysfunction and therefore diabetes [7]. Whether hyperuricemia also induces insulin resistance in Uox^-/- mice warrants further investigation.

Regarding lipid metabolism, our Uox^-/- mice displayed so-called atherogenic lipid triad: markedly increased levels of TC, TG and LDL-C, as well as decreased levels of HDL-C, although not reaching a statistical significance except for TC. The underlying mechanism is unclear. One possible explanation is that knocking out of Uox may cause hepatocellular injury, as evidenced in our Uox^-/- mice, thereby disturbing hepatic lipid metabolism [25]. On the other hand, excessive influxes of lipids could promote fat accumulation and NAFLD, which has been widely recognized as a cardiometabolic disorder [26]. NAFLD is speculated to cause hyperuricemia by inducing the expression of hepatic xanthine oxidase without changing Uox expression [27, 28]. Interestingly, an opposite trend was observed in our Uox^-/- mice, in which NAFLD was evidenced by hepatic steatosis and hepatocellular ballooning, suggesting hyperuricemia as a potential driving cause of NAFLD. This was supported by previous experimental studies showing UA can induce hepatic steatosis in HepG 2 cells [29].

The present model has important shortfalls. To begin, we lack direct autopsy evidence for major cause of premature death in prototype Uox^-/- mice. Moreover, only 50% of allopurinol-rescued Uox^-/- mice survived to 12 weeks, limiting the use of current model in the long-term studies. However, the spontaneous high levels of SUA and multiple organ injuries actually replicate the clinical patients of early-onset hyperuricemia with severe complications. Another major limitation is that we failed to exam phenotypic sex disparities due to limited sample size [9]. Confounding factors due to sex bias should be considered when interpreting these study results. Most importantly, the use of allopurinol should be titrated in a dose and time dependent manner and dynamic observations of biological properties of our modified Uox^-/- mice is necessary in the future.

In summary, this is a novel mouse model of hyperuricemia established in two steps: knocking out of Uox by CRISPR/Cas9 system (prototypic Uox^-/- mice) and allopurinol administration to improve their survivals up to 16 weeks (modified Uox^-/- mice). Given the significantly elevated SUA along with urate nephropathy, cardiovascular disease and metabolic syndrome, the present modified Uox^-/- mice could be suitable candidate for short and mid-term studies on early-onset hyperuricemia with severe complications. The model is useful to improve our understanding the hyperuricemia mediated cardiovascular and cardiometabolic disorders, as well as to develop more efficient novel therapies.