SIR–Creatine transporter (CRTR) deficiency (MIM 300036) is an X-linked inherited defect characterized by the absence or decrease in the 1H-MRS brain creatine signal due to mutations in the SLC6A8 creatine transporter gene.1 Since 1H-MRS is not available in many hospitals, the suspicion of this disease is often based on clinical features and/or on the presence of biochemical markers in physiological fluids.2 It has been proposed that an increase in the ratio 4 urine creatine to creatinine (Cr/Crn) may serve as a useful metabolic marker to identify patients with CRTR deficiency, as well as the mutational analysis of the SLC6A8 gene.3 Indeed, some patients were identified in groups of males with learning disability* of unknown aetiology or with autism by direct genomic DNA sequencing, although further analysis (metabolites determination, creatine uptake, mRNA expression analysis, etc.) could not be carried out on additional samples such as urine or skin biopsy.4–6 Although the prevalence of this disease was estimated to be 2.1/100 in a European X-linked learning difficulty study,4 its true frequency is unknown. Indeed, the poor specificity of the clinical symptoms could be responsible for the underdiagnosis of the CRTR defect within the paediatric population.7 Therefore, prospective biochemical screening of urine from patients with suspected inborn errors of metabolism has been added to the neurometabolic work-up in our laboratory, measuring Cr and guanidinoacetate (GAA) levels.
By applying this policy, urine from more than 4100 patients referred for investigation of possible inborn errors of metabolism and from 180 males with autistic behaviour were studied over a period of 3 years for creatine deficiency syndromes. Urine and plasma Cr and GAA were analysed by High performance liquid chromatography tandem mass Spectrometry,8 adding 13C2-guanidinoacetic acid as an internal standard. About 1 out of 10 of the screened urine samples was found to have increased or decreased urine GAA levels or an altered Cr/Crn ratio. A defect in creatine biosynthesis was ruled out as all cases with persistent altered urinary GAA levels presented a normal GAA plasma concentration. Seven cases with a persistent altered Cr/Crn ratio were further investigated. To confirm a CRTR defect, a novel functional assay for Cr uptake in fibroblasts using [4-14C]creatine9 was developed that enabled the measurement of both the specific uptake and the kinetic constants in contrast to the previously reported tandem mass spectrometry method.1 Mutational analysis was performed by directly sequencing of the entire SLC6A8 gene coding sequence, and confirmed by DNA sequencing of the corresponding exon and its intronic flanking sequences.
We identified a 3-year-old male with CRTR deficiency referred for study owing to mild psychomotor retardation with independent walking acquired at 18 months and expressive language impairment. He never experienced epileptic seizures. Clinical examination revealed clumsiness in his gait and subtly abnormal hand movements. An abnormal Cr/Crn ratio was found in three different urine samples (3.28, 3.05, and 2.09: reference value, 0.51 SD 0.44). Brain 1H-MRS showed a significant decrease in the creatine signal at 3.02ppm (N-Acetylaspartate/Cr: 12.94 and Choline/Cr: 6.89, age-matched controls [n=5] 2.18 and 1.37 respectively), while no abnormality was evident in conventional MRI. An impaired Cr uptake (6% at 25μM Cr) in patient’s fibroblasts was confirmed. The molecular analysis of the SLC6A8 gene revealed a novel nucleotide variation in exon 8, c.1210G>C (p.Ala404Pro). This mutation affects a less strongly conserved residue, Ala 404 located in the putative transmembrane domain VIII (TM8) of CRTR (Swiss-Prot P48029), highly conserved between different species and with other human Na+/Cl− dependent neurotransmitter transporters (Fig. 1).
Although no further functional characterization of the mutation was performed, it was considered to be pathogenic because it was not found in a large cohort of normal chromosomes (100 chromosomes from this study plus 276 controls).4
The mutation could not be detected in the mother by DNA sequence analysis of two independent blood samples, suggesting that the mutation arose de novo.10–12 However, somatic mosaicism cannot be excluded and thus, the risk of repetition is not zero. This is essential for family counselling when a familial mutation is not detected in the mother’s DNA. Indeed, somatic and germline mosaicism for a SLC6A8 mutation was recently demonstrated by denaturing high-performance liquid chromatography in a mother with two affected siblings, illustrating that even if the mutation cannot be detected in the mother by sequence analysis (or other techniques), there is a risk of repetition.13 After reviewing the clinical description of seven cases under the age of 4 years diagnosed previously,7,10,11,14,15 we conclude that our patient has a mild presentation of this condition with late-onset walking and no seizures or autistic behaviour. The identification of this new case encourages the use of prospective screening employing the Cr/Crn ratio for this disorder in the paediatric population.