The Study of Carbamoyl Phosphate Synthetase 1 Deficiency Sheds Light on the Mechanism for Switching On/Off the Urea Cycle

Abstract

Carbamoyl phosphate synthetase 1 (CPS1) deficiency (CPS1D) is an inborn error of the urea cycle having autosomal (2q34) recessive inheritance that can cause hyperammonemia and neonatal death or mental retardation. We analyzed the effects on CPS1 activity, kinetic parameters and enzyme stability of missense mutations reported in patients with CPS1 deficiency that map in the 20-kDa C-terminal domain of the enzyme. This domain turns on or off the enzyme depending on whether the essential allosteric activator of CPS1, N-acetyl-L-glutamate (NAG), is bound or is not bound to it. To carry out the present studies, we exploited a novel system that allows the expression in vitro and the purification of human CPS1, thus permitting site-directed mutagenesis. These studies have clarified disease causation by individual mutations, identifying functionally important residues, and revealing that a number of mutations decrease the affinity of the enzyme for NAG. Patients with NAG affinity-decreasing mutations might benefit from NAG site saturation therapy with N-carbamyl-L-glutamate (a registered drug, the analog of NAG). Our results, together with additional present and prior site-directed mutagenesis data for other residues mapping in this domain, suggest an NAG-triggered conformational change in the β4-α4 loop of the C-terminal domain of this enzyme. This change might be an early event in the NAG activation process. Molecular dynamics simulations that were restrained according to the observed effects of the mutations are consistent with this hypothesis, providing further backing for this structurally plausible signaling mechanism by which NAG could trigger urea cycle activation via CPS1.

Introduction

Carbamoyl phosphate synthetase 1 (CPS1) deficiency (CPS1D, OMIM #237300), a recessively inherited autosomal (2q34) (McReynolds et al., 1981) inborn error of the urea cycle (Freeman et al., 1964, Gelehrter and Snodgrass, 1974), has an estimated incidence of 1/50000 to 1/300000 (Uchino et al., 1998, Summar et al., 2013). CPS1 is the entry point of ammonia, the nitrogenous waste product of protein catabolism, into the urea cycle (Fig. 1A). Therefore, CPS1D causes pure hyperammonemia (Häberle and Rubio, 2014), leading to encephalopathy and even death (Brusilow and Horwich, 2001), and to depletion of downstream urea cycle intermediates, particularly of citrulline (Häberle and Rubio, 2014).

A large repertory of mutations affecting the CPS1 gene (OMIM #608307; 201,425 nucleotides; start/end chromosome 2 coordinates, 211,342,405/211,543,830, plus strand; http://www.genecards.org/cgi-bin/carddisp.pl?gene=CPS1) has been compiled from patients with CPS1D (Häberle et al., 2011). Over 50% of these mutations are missense changes spreading over the entire 1462-residue mature CPS1 polypeptide (Nyunoya et al., 1985, Haraguchi et al., 1991). The CPS1 gene encompasses 4500 coding nucleotides over 38 exons (Funghini et al., 2003, Häberle et al., 2003, Summar et al., 2003). It may be difficult to ascertain the responsibility of a given CPS1 missense mutation in causing CPS1D, particularly for mutations mapping outside the two catalytic domains of the enzyme (the two phosphorylation domains, Fig. 1B) which bind the substrates and catalyze the three-step CPS1 reaction (Alonso et al., 1992, Alonso and Rubio, 1995).

Our present work dealt with the analysis of the effects of CPS1D-associated mutations (called here clinical mutations) that affect a non-catalytic domain of human CPS1, the C-terminal domain of 20 kDa (Häberle et al., 2011). This domain is called the allosteric domain (abbreviated ASD) (Fig. 1B) because it binds N-acetyl-L-glutamate (NAG) (Rodriguez-Aparicio et al., 1989, Pekkala et al., 2009), the essential allosteric activator of CPS1. Without NAG, CPS1 is inactive (Rubio et al., 1981, Rubio et al., 1983), possibly reflecting the need to stop catalysis by the enzyme (Shigesada et al., 1978, Stewart and Walser, 1980) before the ammonia level is too low (Fig. 1A). Too much decrease in the ammonia level would lead to depletion of ammonia-derived amino acids such as glycine, glutamate and glutamine (Bender, 2012), and possibly to protein catabolism. NAG is a proper effector of the CPS1 switch because its level reflects the nitrogen burden manifested in the glutamate level (Fig. 1A). This is so because NAG has a short half-life (Morita et al., 1982) and it is made from glutamate by an enzyme (NAG synthase) exhibiting a high K_m for glutamate (Sonoda and Tatibana, 1983).

To explore the effects of ASD-mapping missense clinical mutations (abbreviated ASD clinical mutations), we utilized a novel system for production and mutagenesis of recombinant CPS1 that uses baculovirus and insect cells (Díez-Fernández et al., 2013). We had already applied this system to analyze the effects of some ASD missense mutations (Pekkala et al., 2010, Díez-Fernández et al., 2013). We now extend this analysis to all the reported ASD clinical mutations, as well as to some mutations designed to test the role of the ASD (Fig. 1B, vertical lines, and Table 1, Table 2). Analysis of the effects of these mutations has helped assess disease causality, opening the way to improved genetic counselling and even to individualized therapy. Furthermore, we now shed some light on the as yet unclarified NAG activation process, and define better the NAG site. We had previously localized the NAG site (Fig. 1C) (Pekkala et al., 2009) by photoaffinity labeling and by in silico docking in the deposited (but unpublished) crystal structure of the isolated human ASD free from NAG [Protein Databank (PDB; www.rcsb.org) file 2YVQ; Xie et al., 2007]. We now have found that some ASD mutations have effects that are inconsistent with the previously proposed NAG site. With the help of molecular dynamics (MD), applying restraints based on the site-directed mutagenesis results, we now propose a refined NAG site structure where a conformational change in the β4-α4 loop could be the initial signal in NAG activation.