Biallelic mutations in CAD, impair de novo pyrimidine biosynthesis and decrease glycosylation precursors

Abstract

In mitochondria, carbamoyl-phosphate synthetase 1 activity produces carbamoyl phosphate for urea synthesis, and deficiency results in hyperammonemia. Cytoplasmic carbamoyl-phosphate synthetase 2, however, is part of a tri-functional enzyme encoded by CAD; no human disease has been attributed to this gene. The tri-functional enzyme contains carbamoyl-phosphate synthetase 2 (CPS2), aspartate transcarbamylase (ATCase) and dihydroorotase (DHOase) activities, which comprise the first three of six reactions required for de novo pyrimidine biosynthesis. Here we characterize an individual who is compound heterozygous for mutations in different domains of CAD. One mutation, c.1843-1G>A, results in an in-frame deletion of exon 13. The other, c.6071G>A, causes a missense mutation (p.Arg2024Gln) in a highly conserved residue that is essential for carbamoyl-phosphate binding. Metabolic flux studies showed impaired aspartate incorporation into RNA and DNA through the de novo synthesis pathway. In addition, CTP, UTP and nearly all UDP-activated sugars that serve as donors for glycosylation were decreased. Uridine supplementation rescued these abnormalities, suggesting a potential therapy for this new glycosylation disorder.

Figure 1.

Schematic of the de novo pyrimidine biosynthesis pathway. UDP4003 Pedigree and characterization of CDAII/HEMPAS like phenotype. (A) Diagram showing the de novo pyrimidine biosynthesis pathway with known disorders highlighted with an (X) over the arrow. (B) Peripheral blood smear showing abnormal erythrocytes with characteristics of varying sizes and abnormal shapes (anisopoikilocytosis), spiked cell membrane (acanthocytes) and fragmented with pointed ends (schistocytes) resulting in a dyserythropoietic anemia phenotype. (C) Western blot analysis of RBC proteins, RhAG and Band-3, showing slower migration consistent with abnormal glycosylation.

Figure 2.

Characterization of CAD Mutations identified in UDP4003. (A) Schematic for Human CAD (UniProt - P27708) with location of mutations identified by exome sequencing. The c.1843-1G>A results in an in-frame deletion of 63-amino acid in the CPSase A portion of Carbamoyl-phosphate synthase 2. (B) Abnormal splicing caused by the c.1843-1G>A was analyzed from two controls, GM00038 and GM03348 and UDP4003 using two different sets of primers. In both controls, a minor portion (4–6%) of CAD transcript lacks exon-13, whereas in UDP4003 there is substantially more (52–61%). (C) Conservation of human Arg2024 across other organisms.

Figure 3.

Analysis of purified nucleotide sugars by HPLC. (A) Nucleotide sugars were extracted and the purified material from CHO-K1 (wild-type) or CHO-G9C (CAD defective) cells with or without 30 µm uridine treatment were run on HPLC. (B) Nucleotide sugars were extracted in the same manner as above and the purified material from GM00038, GM03348 and UDP4003 with or without 30 µm uridine treatment was run on HPLC. The control sample represents an average of both GM00038 and GM03348. In both (A) and (B), the bar graph represents the peak area for each metabolite normalized to ATP levels, since ATP were consistently equal in multiple experiments. Error bars show standard deviation.

Figure 4.

Aspartate Flux. (A) Cells were serum starved for 24 h prior to stimulation with 100 nm recombinant Human Insulin for 7 h in the presence of 2uCi/ml U-¹⁴C aspartate and 1uCi/ml [5,6-³H] uridine. The radiolabeled U-¹⁴C aspartate incorporated into total RNA was normalized to the amount of [5,6-³H] uridine incorporated into RNA. Controls represent the averages of two separate fibroblast lines, GM00038 and GM03348. (B) Cells were serum starved for 24 h prior to stimulation with 100 nm recombinant Human Insulin for 7 h in the presence of 2 mm ¹³C,¹⁵N aspartate. Hydrolyzed samples were analyzed via GCMS. No difference was observed for incorporation of ¹³C aspartate into adenine. Controls represent the averages of two separate control fibroblast lines GM00038 and GM03348.

Figure 5.

Expression and complementation assays. (A) Western blot analysis of HEK 293T cells transfected with FLAG tagged CAD expression constructs showing reduced expression of the exon-13 deleted CAD and normal expression of the pR2024Q CAD, when compared with wild-type CAD. (B) CHO-G9C cells transfected with FLAG tagged CAD constructs showing their ability to rescue the growth defect in the absence of uridine. In the non-rescued samples there is minimal growth, with any detectable growth is likely due to low levels of uridine with in the serum.