Summary

Urea cycle defects (UCDs) are inborn errors of metabolism that cause dysfunction in ammonia detoxification and endogenous arginine synthesis. Argininosuccinic aciduria (ASA) (OMIM 207900) is the second most common UCD, occurring ~1 in every 100,000 live births. ASA is caused by deficiency in argininosuccinate lyase (ASL), a cytosolic urea cycle enzyme, which catalyses the conversion of argininosuccinate into arginine and fumarate, thereby enabling the removal of excess nitrogen. ASL is also involved in the citrulline-nitric oxide (NO) cycle to produce NO through the channelling of extracellular L-arginine to nitric oxide synthase (NOS).
Patients with ASA display acute hyperammonaemia and a chronic phenotype of neurocognitive impairment and liver disease. The aims of the current therapeutic guidelines for ASA are to normalise ammonia and arginine concentrations through a low-protein diet, ammonia scavenging drugs, and arginine supplementation. Liver transplantation is used in cases of progressive liver disease or recurrent hyperammonaemic crises that occur despite conventional treatment. Proposed experimental treatments include antioxidants, autophagy enhancers, creatinine supplementation, and gene therapies.
Chronic liver dysfunction causes morbidity in all UCD subtypes but is reported with higher frequency and severity in ASA. This liver disease commonly manifests with hepatomegaly and transaminitis and can progress to liver failure and, eventually, hepatocellular carcinoma. Liver pathology progresses despite appropriate ammonia control, suggesting hyperammonaemia is not the sole cause. Other suggested mechanisms include arginine deficiency, argininosuccinate toxicity, NO deficiency, and oxidative stress. There are no reliable biomarkers that predict the degree of liver disease in ASA and the underlying processes that trigger liver disease are unclear. More detailed mechanistic insight into liver pathophysiology will be crucial to identifying optimal diagnostic markers for better assessment of disease severity, prediction of disease progression, and assessment of response to therapy.
The Asl^Neo/Neo mouse model recapitulates much of human ASA, with reports of hepatomegaly, elevated transaminases, aberrant hepatic glycogen accumulation, and variable fibrosis. Here, the authors studied the dysregulation of liver glutathione metabolism and its role in the chronic liver disease observed in both patients with ASA and Asl^Neo/Neo mice. We assessed pathways of glutathione biosynthesis, which requires the rate-limiting biosynthetic enzyme glutamate cysteine ligase (GCL), glutathione recycling with γ-glutamyltranspeptidase (GGT) activity and metabolic fluxes through the hepatic antiporter system x_c⁻, which promotes glutathione synthesis to counteract oxidative stress in health and disease. System x_c⁻ activity was monitored with the positron emission tomography (PET) radiotracer (S)-4-(3-¹⁸F-fluoropropyl)-L-glutamate ([¹⁸F]FSPG) used both as a diagnostic tool and to assess the progression of liver disease in ASA. mRNA therapy was tested as a treatment for both neonatal and adult Asl^Neo/Neo mice to restore both glutathione metabolism and ureagenesis in vivo.

Results from the nanoScan PET/CT

For the imaging studies, naesthetized (1.5-2% isoflurane in O₂) 2-3 week old Asl^Neo/Neo mice with age-matched WT littermates were imaged after IV injection of 1-3 MBq of radiotracer. For rescue experiment with mRNA therapy, Asl^Neo/Neo mice were given 1 mg/kg IV dose of hASL mRNA dose at day 1 of birth followed with weekly 2 mg/kg IP injection. Age-matched control untreated Asl^Neo/Neo and WT littermates were included. For all animals, dynamic PET scans were acquired between 40 and 90 min post-injection on the Mediso nanoScan PET/CT system (1-5 coincidence mode; 3D reconstruction; CT attenuation-corrected; scatter corrected) using a four-bed mouse hotel. CT images were acquired for anatomical visualization (360 projections; helical acquisition; 55 kVp; 600 ms exposure time). A dynamic iterative reconstruction algorithm, Tera-Tomo 3D (0.4 × 0.4 × 0.4 mm³ voxel size), was used with attenuation, scatter, and random coincidences correction. Radiotracer concentration was quantified using VivoQuant software, with volumes of interest drawn manually using the CT image as reference. Data were expressed as percent injected dose per gram of tissue (%ID/g). After [¹⁸F]FSPG PET, mice were culled by cervical dislocation and liver, skin and other tissue was collected, snap frozen and moved to a -80°C freezer for later ex vivo analysis.
In all mice, typical healthy tissue [¹⁸F]FSPG retention was observed in the salivary glands, thymus, and pancreas, accompanied by renal elimination (Fibure 2B). Liver [¹⁸F]FSPG retention for Asl^Neo/Neo mice was 14 ± 4% injected dose (ID)/g, which was 3-fold higher (p=0.002) than that of WT mice (5.2 ± 1.5% ID/g). In PET images of WT mice, [¹⁸F]FSPG retention in the liver was just above background, with images dominated by radiotracer retention in the pancreas and kidney. Conversely, it was challenging to distinguish between pancreatic and liver [¹⁸F]FSPG retention in Asl^Neo/Neo mice (Figures 2B, 2C, FigureS4A).

To investigate the potential of [¹⁸F]FSPG PET as a non-invasive tool of therapeutic efficacy, [¹⁸F]FSPG was administered IV to 2 weeks-old untreated and hASL mRNA-treated Asl^Neo/Neo mice (IV administration of 1 mg/kg hASL mRNA at birth followed by weekly IP administration of 2 mg/kg mRNA before imaging at 2 weeks of age). Supporting our functional and metabolic data, [¹⁸F]FSPG retention was halved in hASL mRNA-treated (11 ± 2.0% ID/g) versus untreated Asl^Neo/Neo mice (22 ± 2.3% ID/g; p = 0.026; Figure 6A, 6B, Figure S11A). At the treatment time-point, however, [¹⁸F]FSPG retention was not completely restored to baseline levels seen in WT livers (5.0 ± 2.8% ID/g). In line with lowered [¹⁸F]FSPG retention, glutathione metabolism was corrected, with restoration of hepatic liver glutathione in both neonatal and adult treated Asl^Neo/Neo mice to concentrations similar to those of WT mice (Figure 6C). This restoration of glutathione was associated with a significant reduction of total homocysteine ratio compared to WT in livers from hASL mRNA-treated versus untreated Asl^Neo/Neo mice (Figure 6D). In these animals, the expression of cystine/glutamate antiporter system x_c⁻ was greatly reduced in livers from hASL mRNA-treated versus Asl^Neo/Neo mice (Figure 6E). Conversely, skin [¹⁸F]FSPG retention was not affected by mRNA therapy. [¹⁸F]FSPG skin retention was 4.2 ± 3.4% ID/g in WT mice, compared to 15 ± 3.7% ID/g and 15 ± 4.2% ID/g in untreated and hASL mRNA treated Asl^Neo/Neo mice, respectively (Figures S11B, S11C).

• In conclusion, the authors' study shows dysfunction of glutathione metabolism in both ASL-deficient patients and an Asl^Neo/Neo mouse model, whilst mRNA-LNP therapy corrected both glutathione metabolism and ureagenesis in vivo.
•  Preliminary data suggests that glutathione biosynthesis in the liver is regulated by NO availability.
• Furthermore, the authors demonstrated the potential of [¹⁸F]FSPG-PET as a companion diagnostic to assess liver disease and therapeutic efficacy in ASA.
• These insights into the liver pathophysiology of ASA provide perspectives for targeted therapies, which could change the outcome of patients affected by this rare disease with currently high unmet needs and limited therapeutic options.

Full article on science.org and PMC