Funding: This work was supported by grants from the Fondo de Investigación Sanitaria (PI051318 and PI070548).
Competing interests: None declared.
Statistics from Altmetric.com
Three metabolic defects in catabolism of histidine have been recognised: histidase (OMIM 235800), urocanase (OMIM 276860) and glutamate formiminotransferase (OMIM 229100) deficiencies.1 2 Urocanase deficiency was first reported in 1971,3 and mental retardation was described as the main neurological sign in a single case. Later on, two sisters were reported presenting mental retardation plus ataxia and dysarthria.4 In these cases, the defect could be found by indirect analysis of some metabolites of the histidine pathway in blood and urine and by the quantification of deficient urocanase activity in liver biopsies. In contrast, it has been suggested that urocanase deficiency might be a benign condition, as urocanic aciduria was found by screening analysis as in some healthy cases.1 Interestingly, glutamate formiminotransferase deficiency, the following biochemical step in histidine catabolism, presents some neurological signs common to urocanic aciduria, and mutations explaining severe and mild phenotypes have been reported.2 We describe a patient presenting with urocanic aciduria, mental retardation and intermittent ataxia associated with two base-pair changes in the UROC1 gene that produce two amino acid substitutions. We suggest that these molecular changes may be the genetic cause of the clinical and biochemical picture in this patient.
The patient is a woman born to healthy non-consanguineous parents. She had normal development in the first year of life and started to walk unaided at 16 months, but the parents observed clumsy performance. At 4 years of age, she had three episodes of severe ataxia that lasted 20–25 days, which coincided with recurrent infections. These episodes resolved slowly over the following weeks. At 5 years of age, she was referred for neurological evaluation to our hospital.
Clinical examination found truncal ataxia with broad base, ataxic gait, asymmetry that was more evident on the left arm, action tremor, dysartric speech, and normal ocular movements with slight nystagmus on horizontal gaze. Normal muscle strength with brisk tendon reflexes and moderate mental retardation were seen. Cranial MRI and neurophysiological studies, including electroencephalography, electroretinography, visual and brainstem evoked potentials, electromyography, nerve conduction velocities, fundus oculi and cardiological examinations, gave normal results. At 9 years of age, the child again had crises of severe ataxia of 20 days duration. At that time, she was attending normal school with a special programme of integration for those with moderate mental retardation. After this point, she began to show improvement in some neurological signs, such as tremor and ataxia. A second cranial MRI was normal. She is currently 19 years of age. She has slight tremor and nystagmus, walks with broad base and has very clumsy performance. Cognitive evaluation rated her IQ at 54 (Wechsler Intelligence Scale for Children-Revised: verbal 57 and manipulation 60).
Routine laboratory analyses including serum amino transferases, creatinine, urea and other parameters were normal (no liver or renal dysfunction was detected). Screening for inborn errors of metabolism, including plasma and urine amino acids (no histidine accumulation), blood lactate, sialotransferrin isoelectricfocusing, creatine deficiency syndromes, purine and pyrimidine defects, and organic acidurias, gave normal results,. Hypothyroidism and coeliac disease were also excluded. Biochemical and/or genetic screening for early onset forms of inherited ataxia (Friedreich ataxia, ataxia with isolated vitamin E deficiency, ataxia teleangiectasia, abetalipoproteinaemia, and ataxia of mitochondrial origin) gave normal results.
Informed written consent was obtained for biochemical and genetic investigations, which were performed after diagnostic protocols approved by the ethics committee of Hospital Sant Joan de Déu, Barcelona.
Urocanic acid in urine was analysed by high-performance liquid chromatography (HPLC) with ultraviolet and diode array detection as previously reported.5 Blood count, serum folate, and plasma total homocysteine were determined by standard automated procedures. Cerebrospinal fluid was collected according to a standardised protocol, and 5-methyltetrahydrofolate, neopterin, biopterin and biogenic amines (5-hydroxyindoleacetic (5-HIAA) and homovanillic acids (HVA)), metabolites of serotonin and dopamine, respectively) were determined by HPLC with electrochemical and fluorescence detection, as previously reported.6 7
Genomic DNA was obtained by standard methods from peripheral white blood cells from the propositus and her father. Mutation analysis of the UROC1 gene was performed by direct sequencing of PCR products in an automated analyser ( ABI Prism 3130×l; Applied Biosystems, Foster City, California, USA), using primers designed according to the reference sequence NM_144639, which amplify 20 exons of the UROC1 gene and their intronic flanking sequences (table 1). We computed the biological relevance of the mutated residues (in silico analysis). Conservation of residues was analysed by alignment of related sequences using the BLAST programme.8 Secondary structure predictions were calculated by four different software tools (GOR, Jpred3, APSSP, and nnPredict, available from the ExPASy Proteomics Server: http://www.expasy.ch). Sequence-based predictions of the phenotypic consequences of the UROC1 p.R450C mutation were assessed using the SIFT programme.9 We also predicted the effects of the mutation on the basis of the 3D structure using the quantitative rules of Wang and Moult,10 and the PolyPhen server.11 Visualisation of the structure and of the consequences of the UROC1 p.R450C mutation was carried out using the Coot programme.12 We used the structure 1UWK13 (RCBS Protein Data Bank), which corresponds to the urocanase from Pseudomona putida in complex with its substrate, urocanate.
Protein production and enzyme activity
The UROC1 gene (cDNA clone MGC:135008, IMAGE: 40076086, Nottingham, UK) was cloned into the expression vector pET100/D-TOPO (Invitrogen, Carlsbad, CA, USA), which allows expression of recombinant proteins with an N-terminal tag containing the Xpress epitope and a 6×Histag. The enzyme was produced in Escherichia coli BL21Start (D3) (Invitrogen). The cells were induced with 1 mmol/l isopropyl-beta-D-thiogalactopyranoside (IPTG) and grown as previously described.13 The pellet was resuspended in 5 mL of lysis buffer (50 mmol/l potassium phosphate pH 7.8, 400 mmol/l NaCl, 100 mmol/l KCl, 10% glycerol, 0.5% Triton X-100, 10 mmol/l imidazole) containing 1 mmol/l phenylmethylsulphonylfluoride (AppliChem, Darmstadt, Germany) and sonicated. After removing the cells debris by centrifugation, the supernatant was applied to a His SpinTrap column (GE Healthcare, Amersham, Buckinghamshire, UK) and eluted according to the manufacturer’s instructions. Human UROC1 p.L70P and p.R450C missense mutations were generated by PCR with specific primers containing the nucleotide changes and mutated proteins were purified as described above for the native urocanase. The fractions containing the urocanase (native or any of the mutated variants) were identified by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE), and the overexpressed bands of the desired proteins were confirmed by western blotting using a mouse anti-His6 monoclonal antibody (Roche Applied Science, Penzberg, Germany).
Urocanase was assayed by a previously described spectrophotometric method, in which the disappearance of urocanic acid is measured as decrease in absorbance at 277 nm.14 The reaction mixture containing 0.5 mL of 0.2 mol/l potassium phosphate buffer, pH 7.55, 0.10 mL of 0.01 mol/l EDTA, adjusted to pH 8.0, 0.01 mL of 0.01 mol/l urocanic acid in 0.01 N HCl, and water in a total volume of 0.84 mL, was preincubated at 37°C for 5 minutes. The reaction was started by adding 0.160 mL of enzyme solution, and the decrease in absorbance at 277 nm was monitored for about 15 minutes. The molar extinction coefficient of urocanate under these conditions is 18 800.15 One unit of enzyme is defined as the amount that causes the disappearance of 1 nmol/L of urocanic acid per min.
Urocanic aciduria was measured at different ages (12, 14 and 18 years) in this patient, and urocanis acid levels ranged between 158 and 202 mmol/mol creatinine (normal value <10 ). Urocanic acid excretion was normal in the father. Serum folate was at the lower limit of our reference interval (10 nmol/L: reference values 10–26), although neither hyperhomocysteinaemia nor megaloblastic anaemia was seen at any point during the evolution of the disease. Cerebrospinal fluid results are reported in table 2. Deficiencies in 5-methyltetrahydrofolate, neopterin and biopterin deficiencies were seen, together with an evident reduction in homovanillic acid (HVA) and the HVA:5-hydroxyindolacetic acid ratio. Other parameters determined in cerebrospinal fluid (CSF), such as glucose, proteins, cells and amino acids were normal.
Persistent urocanic aciduria and clinical data, mainly mental retardation, pointed towards a diagnosis of urocanase deficiency. Analysis of the UROC1 gene showed that the patient harboured two nucleotide changes, c.209T→C in exon 2 and c.1348C→T in exon 14 (fig 1). These base-pair changes are translated into two amino acid changes, p.L70P and p.R450C, respectively. Moreover, sequencing analysis showed the presence of the neutral change c.1056C→T (p.G352G) in exon 11, previously described as a single-nucleotide polymorphism (SNP), rs34025926, (http://www.ncbi.nlm.nih.gov/SNP). The proband’s father was a heterozygous carrier of the p.R450C mutation (no DNA was available from the mother). We searched for both nucleotide change in 200 chromosomes from healthy controls of Spanish ancestry by automated sequencing, and found neither, suggesting that they could be pathogenic mutations. We then proceeded to perform a phylogenetic analysis of the L70 and R450 residues. Leucine at position 70 was present in several mammals besides Homo sapiens (Canis familiaris, Equus caballus, Monodelphis domestica, Bos taurus and Macaca mulatta) and in Gallus gallus. The secondary structure predicted for the human urocanase showed that an α-helix was always predicted from residue 63 to residue 70.With respect to arginine 450, the alignment of the urocanase sequences showed that R450 is extremely conserved and in fact no variation was found in >100 aligned sequences from different organisms (data available on request). The p.R450C mutation was always predicted as a probable pathological change by four software programmes. In P putida, the orthologous residue of human R450 is R362, which is extremely important for the enzymatic activity of urocanase. Urocanate is tightly bound to urocanase by means of its carboxylate group, forming a buried salt bridge with R362.13 If this arginine is mutated to cysteine, urocanase would lose this interaction and would therefore be unable to bind to urocanate (fig 2).
The protein expression analysis by SDS-PAGE, after sonication and before purification, showed that both the native and the p.R450C urocanase were present in the supernatant and the pellet fractions (figs 3A and 3B). However, the p.L70P form was only detected in the pellet (fig 3C). To optimise protein expression in bacteria, we performed experiments of the p.L70P mutant under different conditions of IPTG concentration, temperature and inducing time. We could not obtain the p.L70P protein in the supernatant under any of these conditions (fig 3C). Thus, purification and enzymatic assays were only performed for the native protein and the p.R450C mutant. The spectrophotometric enzymatic assay showed that the native urocanase had a specific activity of approximately 70 U/mg of protein. In contrast, the p.R450C urocanase had no detectable activity.
We investigate the underlying molecular defect in a patient with from urocanic aciduria presenting with mental retardation and intermittent ataxia. The biochemical findings were consistent with a defect in urocanase.
We found two putative mutations in the urocanase gene UROC1 that produce two amino acid substitutions, p.L70P and p.R450C. Computed predictions, protein expression studies and enzyme activity assays suggest that none of the mutations produce a fully functional enzyme. The p.L70P substitution, which probably imply the disruption of an α-helix in the N-terminus, would alter its properties and therefore, its function. The p.R450C change would render impossible any interaction between urocanase and its substrate and would loss its enzyme activity. We therefore suggest that mutations in the UROC1 gene might be the cause of the urocanic aciduria in our patient.
We found a partial cerebral folate deficiency in this case, that probably lead to a impaired pterin and biogenic amines status in CSF. These findings might be explained by the fact that urocanase participates in the synthesis of folate metabolites through histidine catabolic pathway. However, low CSF 5-MTHF concentrations have been related to several genetic and non-genetic conditions and consequently, this finding might be also considered unspecific.
Two nucleotide substitutions were found in the coding region of the UROC1 gene in a girl with urocanic aciduria, who had mental retardation and intermittent ataxia as the main symptoms. She is compound heterozygous for the UROC1 p.L70P/p.R450C mutations. Computed analyses of the putative structure of human urocanase showed that both amino acid substitutions could have a pathological effect. The L70 residue is only conserved in a few mammals and in Gallus gallus. Lower organisms such as P putida lack it, making impossible to analyse the consequences of this mutation from the available structures of the urocanase (RCSB Protein DataBank; www.rcsb.org). However, the predicted secondary structure using the human urocanase sequence as a template suggested that the L70 residue could form part of an α-helix,; the change to proline may disrupt the α-helix, possibly resulting in an alteration of the structure of the N-terminal region. The normal function of the enzyme would therefore be altered. We could not perform an enzyme activity assay for the p.L70P protein because this mutant protein was not obtained in the soluble fraction, only in the precipitate. According to the computed analysis, this mutation would produce a conformational change of α-helix at the N-terminus. The putative abnormal conformation could affect the solubility of the enzyme and the correct enzymatic function. The p.R450C mutation affects an extremely conserved residue and according to the computed analyses, no change can be tolerated in this position. Because of the predictable interaction by a salt-bridge between the R450 in the human urocanase and its substrate, urocanate, according to the structure of P putida,13 the substitution of R450 by a cysteine would imply that urocanase could not interact with urocanate. That is why an enzyme with the p.R450C mutation would not be able to metabolise urocanate, leading to a deficiency in the enzyme activity, and consequent accumulation of urocanic acid, which is then excreted in large amounts in urine. In fact, the enzyme activity was not detectable in the p.R450C urocanase purified extracts from the bacterial expression system whereas that the native urocanase had a specific activity of approximately 70 U/mg of protein. Moreover, neither of these two changes was identified in 200 control chromosomes, providing additional support for the view that these nucleotide changes do not represent polymorphisms and may be the causative pathogenic mutations. Consequently, computed predictions, protein studies and enzyme activity assays would suggest that none of the mutations, p.L70P and p.R450C, in urocanase could produce a fully functional enzyme.
The healthy father of the patient is a heterozygous carrier of the UROC1 p.R450C mutation. This finding would indicate that the urocanase deficiency is a hereditary disease. To date, four symptomatic children with urocanic aciduria have been reported: two isolated cases3 16 and two affected sisters.4 In the sisters’ family, previous generations were also possibly affected, which would suggest autosomal dominant inheritance. The genetic analysis performed in our patient and her father allows us to conclude that the urocanase deficiency seems to be a genetic disorder with autosomal recessive inheritance. Other defects of the histidine catabolism, histidase and glutamate formiminotransferase deficiencies, are also autosomal recessive disorders.
The propositus has urocanic aciduria presenting as mental retardation and ataxia, and also tremor and nystagmus. The latter two neurological signs have not previously been associated with urocanase deficiency.1 3 4 Histidine catabolism comprises several metabolic steps leading to the synthesis of 5–10-methenyltetrahydrofolate (MTHF), a precursor of the purine ring and other folate metabolites. Although the biosynthesis of folate derivatives through histidine catabolism might represent a minor source of single carbon units,17 we found low–normal blood folate and decreased CSF 5-MTHF values in our case. We also found an association between decreased folate values in CSF, pterin deficiencies and decreased values of HVA, the main metabolite of dopamine, similar to results previously reported for other folate metabolism disturbances18 19. In this sense, tremor and probably other neurological signs might be related to dopamine deficiency. Furthermore, symptoms such as ataxia and psychomotor retardation have commonly been seen among patients with cerebral folate deficiency, even those with very slight deficiency.20 However, low CSF 5-MTHF concentrations has been related to several genetic and non-genetic conditions,17 18 and consequently, this finding might be also considered nonspecific.
It has been found that patients with glutamate formiminotransferase deficiency also have mental and physical retardation, which reinforces the hypothesis that a severe metabolic blockade in urocanic acid catabolism might cause the clinical phenotype. An explanation for the differing phenotypes of patients with urocanic aciduria might be similar to that suggested for glutamate formiminotransferase deficiency,2 and factors such as severity of mutations, amount of accumulated urocanic acid, impaired central nervous system folate, pterins and dopamine status, and other unknown factors might be involved. Urocanic acid has been related to skin photoprotection and regulation of immune response,21 22 although no immune alterations or skin abnormalities were seen in our case. The precise implication of urocanic acid accumulation in the clinical phenotype therefore requires further investigation.
In conclusion, we suggest that mutations in the UROC1 gene might be the cause of the urocanic aciduria in our patient. Although we could not show a definite causative relationship between the UROC1 mutations and the mental and neurological syndrome in this patient, we propose that mutational analysis of the urocanase gene is advisable in this rare condition, along with evaluation of folate deficiency consequences in the central nervous system, such as pterin and dopamine status or other biological markers of impaired folate status.
We are grateful to the patient for her kind collaboration. We are indebted to Dr C Marco-Marín for the in silico structural studies. CIBERER is an initiative of the Instituto de Salud Carlos III.
Funding: This work was supported by grants from the Fondo de Investigación Sanitaria (PI051318 and PI070548).
Competing interests: None declared.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.