If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Communications should be addressed to: Dr. Heussinger; Division of Neuropediatrics; Department of Pediatrics; Friedrich-Alexander University of Erlangen-Nuremberg; Loschgestraße 15; 91054 Erlangen, Germany.
The X-linked creatine transporter deficiency (CRTD) caused by an SLC6A8 mutation represents the second most common cause of X-linked intellectual disability. The clinical phenotype ranges from mild to severe intellectual disability, epilepsy, short stature, poor language skills, and autism spectrum disorders. The objective of this study was to investigate phenotypic variability in the context of genotype, cerebral creatine concentration, and volumetric analysis in a family with CRTD.
Patients and Methods
The clinical phenotype and manifestations of epilepsy were assessed in a Caucasian family with CRTD. DNA sequencing and creatine metabolism analysis confirmed the diagnosis. Cerebral magnetic resonance imaging (cMRI) with voxel-based morphometry and magnetic resonance spectroscopy was performed in all family members.
An SLC6A8 missense mutation (c.1169C>T; p.Pro390Leu, exon 8) was detected in four of five individuals. Both male siblings were hemizygous, the mother and the affected sister heterozygous for the mutation. Structural cMRI was normal, whereas voxel-based morphometry analysis showed reduced white matter volume below the first percentile of the reference population of 290 subjects in the more severely affected boy compared with family members and controls. Normalized creatine concentration differed significantly between the individuals (P < 0.005).
There is a broad phenotypic variability in CRTD even in family members with the same mutation. Differences in mental development could be related to atrophy of the subcortical white matter.
The X-linked creatine transporter deficiency (CRTD) caused by a mutation in the creatine transporter gene SLC6A8 (MIM # 300036) represents the second most common cause of X-linked intellectual disability.
Intellectual disability associated with speech and behavioral disorders are clinical hallmarks of inherited defects of creatine (Cr) metabolism. To date three gene mutations have been described: two defects of creatine synthesis (arginine:glycine amidinotransferase and guanidinoacetate N-methyltransferase deficiencies) and one defect of creatine transport.
The clinical phenotype of CRTD ranges from mild to severe intellectual disability along with speech delay, seizures, short stature, poor language skills, and autism spectrum disorders. Female carriers may exhibit learning disabilities of varying degree and behavioral problems.
The question arises whether cognitive deficiencies are related to brain volumes. Especially subcortical white matter is known to play a crucial role in linking the different components of cortical processing networks necessary for cognitive functions.
Here we describe five members of a family with X-linked CRTD presenting with variable phenotypes and cerebral creatine concentrations. This is the first analysis of brain structure and genotype–phenotype relations in CRTD.
Patients and Methods
Two Caucasian brothers were evaluated for developmental delay, short stature, and epilepsy. Family history revealed short stature in both nonconsanguineous young parents (mother: 158 cm; father: 164 cm) as well as mild learning disabilities in the mother. There was no family history of seizures or other neurological disorders. Consent to neuroradiological procedures was obtained except from the father living apart from the family.
Physical and neurological examinations were performed by an experienced pediatric neurologist. Height was measured with a Harpenden Stadiometer; standard deviation score values for height and body mass index were calculated by applying German reference data.
The mother showed impaired intellectual functioning with reading comprehension difficulties. Nevertheless, she completed secondary school and qualified as childcare assistant. Clinical examination of Child 1 (female; age 18 years) was without abnormal findings except for short stature (height 158 cm; height SDS −1.28; Tanner stage 5).
In contrast, two boys and one girl showed a more complex phenotype. Child 2 (male; age 17 years) was first evaluated at age five years for global developmental delay, short stature (height 97.5 cm; height SDS −4.48), and delayed speech development. Pregnancy and delivery had been uneventful. Motor milestones were normal, but language and cognitive development had been markedly impaired since age two years. Hyperactive and impulsive behavior worsened over time without evidence of regression of cognitive and motor functions. Laboratory investigations for metabolic diseases, molecular genetic studies (karyotype, and molecular analysis of FMR1, SCNA1A, and Array-comparative genomic hybridization), and brain MRI did not identify an etiology of growth delay and intellectual disability. At age 6.1 years the boy presented with generalized epileptic seizures. Electroencephalography showed primary generalized irregular spike and waves and polyspikes. Afebrile generalized tonic-clonic seizures increased in frequency to two or more per day.
Child 3 (female; age 14 years) first presented at age four years with short stature (height 93.5 cm; height SDS −2.4), learning difficulties, and language delay. She developed a complex focal epilepsy at age seven years associated with electrographic spikes and irregular spike-wave complexes and a course refractory to pharmacologic treatment. Child 4 (male; age ten years) presented at age three years with global developmental delay, short stature (height 79 cm, height SDS −4.78), and retarded speech development. Like in his elder brother (Child 2), his motor milestones were normal, whereas language and cognitive development had been impaired from age two years. The severity of his hyperactive and impulsive behavior increased without regression in cognitive and motor functions. Laboratory tests for metabolic diseases, molecular genetic studies, and brain MRI were normal. At age seven years, he developed primary generalized epilepsy with generalized tonic-clonic seizures associated with irregular spike-waves and polyspikes. His generalized tonic-clonic seizures increased to two per day despite pharmacologic treatment. Of note, anticonvulsive therapy with valproate and lamotrigine at adequate dosages was not effective for either of the boys (Children 2 and 4) or the girl (Child 3).
None of the patients exhibited facial dysmorphic signs. Body height at the initial presentation varied from −4.78 to −1.28 SDS, head circumference from −1 SD to +1.7 SD, and body mass index varied from −2.79 to +1.2 SDS. The two affected boys exhibited dysdiadochokinesis, dysmetria in the point-to-point tests, and slight gait ataxia without progression. The three females did not exhibit cerebellar signs during their neurological examination. Muscle tone and strength, deep tendon reflexes, and sensory tests were normal in all patients.
The following instruments were used to estimate general intelligence in all subjects: the standardized intelligence tests Kaufmann Assessment Battery for Children was used for Patient 3; the Hamburger-Wechsler Intelligence Assessments for Adults (HAWIE) was used to assess the mother, and the Hamburger-Wechsler Assessment for Children (HAWIK)
The X-inactivation pattern was determined by polymerase chain reaction analysis of a polymorphic (CAG) n repeat in the first exon of the androgen receptor gene with and without digestion of the DNA with the methylation-sensitive enzyme HhaI.
All samples were analyzed in triplicate. A male control was included in each run.
Magnetic resonance imaging and spectroscopy
MRI and MRS were performed in all family members using a 3-Tesla MR scanner (Siemens Verio) and a 32-channel head coil (both Siemens Healthcare, Erlangen, Germany). For exclusion of anatomic abnormalities, voxel-based morphometry (VBM) and planning of the MRS, three-dimensional T1 magnetization-prepared rapid gradient-echo sequence was acquired (MPRAGE; repetition time 1900 ms, inversion time 900 ms, echo time 2.48 ms). A chemical-shift MRS point-resolved spectroscopy box (PRESS, matrix 16 × 16, voxel size 10 × 10 × 20 mm3, repetition time 1700 ms, echo time 135 ms) was placed over the frontoparietal cortex and the bilateral semioval center as illustrated by the white box in Fig 1. Corresponding spectra with and without water suppression were acquired. The mother and Child 1 kindly provided themselves for comparative measurements.
MRS data were evaluated using LCModel Version 6.3 (LCModel Inc, Drums, USA). This software analyzes in vivo spectra as a linear combination of modeled in vitro spectra from individual metabolite solutions and calculates concentrations for all metabolites assessed. Concentrations of the cerebral metabolites total creatine (total Creatine = Creatine + Phosphocreatine) and total N-acetylaspartate (tNAA = N-acetylaspartate + N-acetylaspartylglutamate) were quantified by using the water scaling method.
All voxels were assessed individually and metabolite calculations that did not fulfill the following quality criteria in one patient were excluded from further analysis for all patients. Spectra were excluded which exhibited full width at half maximum >0.067, signal-to-noise-ratio < 10, for the individual metabolites standard deviation > 20%, chemical shift artifacts on the edges of the MRS—PRESS–Box or LCModel error message. The remaining 20 voxels are indicated by the black rectangle in Fig 1. In this example, voxels of column 1 and 4 are merely white matter; voxels of column 2, 3, and 5 are merely gray matter. To compensate for potential differences in brain water concentration between the individuals and voxels, receiver gain instabilities and relaxation time corrections, tCr was normalized to tNAA (nCr = tCr/tNAA). To account for tNAA differences in gray and white matter, gray and white matter have been analyzed separately.
Volumes of white and gray brain matter were calculated from T1-weighted three-dimensional MRI datasets (magnetization-prepared rapid gradient-echo) using VBM toolbox (Structural Brain Mapping Group, University of Jena, Germany) and compared with an institutional control database of 290 individuals.
Separately for gray and white matter, normalized total creatine (nCr) was analyzed for differences between the family members by analysis of variance (ANOVA) with post hoc corrections (Bonferroni, Tukey B) using SPSS (IBM Corp., New York, USA). A P-value <0.005 was assumed to indicate a significant difference.
Clinical and laboratory findings of the five family members are summarized in Table.
TableClinical, Biochemical, Molecular, and MRS Data of the Patients
Intelligent quotient (IQ) scores of the females varied from 80 to 94. The mother scored 84 and Child 1 scored 94 on the IQ test. Child 3 scored 80 in the Kaufmann Assessment Battery for Children. Their expressive as well as receptive speech and language development were severely delayed: Child 2 was not able to speak single words, could not follow simple orders, and needed help for all daily life tasks.
The active vocabulary of Child 4 consisted of approximately ten words. The patient was able to understand simple requests and orders and was able to do some daily life tasks without help, such as putting on a coat. Functional developmental age was estimated 12 months for Child 2, and 16 months for Child 4 in terms of communication, 22 and 30 months in terms of daily living skills, and 26 and 32 months in terms of socialization, respectively (Fig 2).
In all females GAA, Cr, and Cr/Crn in urine were within normal range; in both boys the GAA levels were normal, whereas Cr/Crn ratios were increased (Child 2: 3.7; Child 4: 3.5; normal range, 0-2), and plasma Cr levels were reduced (2.2 and 1.84, respectively; normal range, 5-19.5 mmol/L).
Analysis of the SLC6A8 gene
DNA sequence analysis of the SLC6A8 gene identified a hemizygous missense mutation (c.1169C>T; p.Pro390Leu, exon 8) in Children 2 and 4 and the same heterozygous mutation in the mother and the affected girl (Child 3). Rosenberg et al.
previously described this mutation in 2004. In the unaffected girl (Child 1) no mutation in the SLC6A8 gene was found.
In all females X-inactivation studies were performed. The mother showed a nonrandom X-inactivation of 74%, Child 1 of 42%, and Child 3 of 52%.
Magnetic resonance imaging
MRI showed prominent ventricles and external cerebral fluid spaces in Child 2 (Fig 3), which suggests an ex vacuo compensation for brain substance atrophy. No structural abnormalities were found in the other family members.
VBM analysis showed a white matter volume below the first percentile in Child 2 compared with the other family members and the control database of 290 subjects. Gray matter volume was between fifth and 95th percentile in all family members (Fig 4).
Magnetic resonance spectroscopy
Example spectra of voxels with predominantly gray and white matter of the five family members are shown in Fig 5. Children 2 and 4 show smaller Cr peaks. Child 3 shows a weaker decrease of the Cr peak in the displayed spectrum that is best recognized in the gray matter spectrum (Fig 5). Statistical analysis of all voxels confirmed a significant difference of the nCr concentrations between the individuals (P < 0.005). The highest concentrations were found in the healthy girl (Child 1) and the mother, and the lowest in Children 2 and 4. Tukey B test grouped the family members in three groups with significantly different nCr concentrations in between (first group: mother and Child 1; second group: Child 3; third group: Children 2 and 4) (Fig 6).
In a large family with nonsyndromic X-linked intellectual disability, DNA sequence analysis of the SLC6A8 gene identified a hemizygous missense mutation in Children 2 and 4 and an identical heterozygous mutation in the mother and Child 3. The diagnosis of CRTR deficiency was subsequently confirmed by brain MRS and urinary creatine/creatinine ratio in Children 2, 3, and 4. Comprehensive clinical, genetic, and laboratory, as well as quantitative MRS and MRI, results underline former reports on phenotypic variability in CRTD. Notably, differences in cognitive development could not be explained or predicted by the genotype of the creatine transporter gene SLC6A8.
Children 2 and 4 (both males) presented with severe intellectual disability and major language impairment in contrast to other siblings with comparable genotype. A similar discrepancy has been described in other conditions of intellectual disability such as fragile X and Down syndromes.
that females heterozygous for CRTD are likely to experience intellectual disability and learning difficulties. Nevertheless, this mother and Child 3 (both heterozygous) had significantly different phenotypes, with only one of them (Child 3) having complex focal seizures, learning difficulties, and language delay. For that reason, the diagnosis of CRTD is intricate and cannot be established on clinical and neuroradiological findings alone.
Bearing in mind the broad spectrum of nonspecific clinical signs at the time of manifestation, routine testing for creatine deficiency disorders in patients with intellectual disability and expressive speech and language delay should be considered.
We therefore normalized the LCModel output to tNAA to eliminate the variable water content, receiver gain instabilities, relaxation time corrections, and further artifacts. Regardless, tNAA is known to be different in the gray and white brain matter, as well as various kinds of epilepsies.
To account more precisely for gray and white matter differences of nCr, we separately analyzed voxels with prevalent gray and white matter. Hence, the nCr differences between the individuals correspond.
brain nCr was reduced in both hemizygous and heterozygous individuals with CRTD.
Both hemizygous brothers (Children 2 and 4) showed that nCr brain concentration was significantly reduced by a factor of about 75%. Nevertheless, the clinical course of the disease was more severe in Child 2 in whom the MRI scan showed an atrophy of the cerebral volume. Moreover, the VBM showed that Child 2 had a significantly reduced white matter brain volume compared with all other family members and healthy controls while possessing a normal grey matter volume. In children with epilepsy, clinical studies suggest that lower white matter integrity may result in reduced neuropsychological function.
we could not confirm a progressive atrophy of the brain in our patients within the monitoring timeframe (Fig 3).
In both heterozygous individuals (mother, Child 3) nCr brain concentrations differed significantly. Although maternal nCr concentration was comparable with that of the nonaffected girl (Child 1), nCr of the affected girl was significantly decreased by about 30% (Child 3). We therefore concluded that the decreased nCr is linked to decreased intelligence score and epilepsy in Child 3. Our findings are consistent with observations of van de Kamp et al.
might further explain variable cognitive functions in the mother and the affected girl (Child 3).
In the female family members no relevant nonrandom X-inactivation (74% in mother versus 52% in Child 3) was found. Therefore, further factors like additional genetic mutations and epigenetic regulations might be responsible for phenotypic differences.
In the future, prospective multicenter studies incorporating genotype–phenotype correlations and long-term follow-up intervals are required to gain further insights into brain metabolism of this complex genetic disorder.
Clinical data of the presented family illustrate a wide intrafamililal phenotype–genotype variability. Interestingly, variable expressivity in carriers of SLC6A8 mutations is not only the consequence of the mutation but might be modified by the degree of creatine depletion as well as environmental factors. Our analysis of the brain structure and intrafamilial genotype–phenotype correlations may indicate a significant role for detailed white matter investigations in patients with CRTD.
The authors gratefully appreciate Dr. Sigrun von der Haar for genetic and Dr. Schultis for biochemical analysis, as well as Professor Fusch for valuable comments on the manuscript. Furthermore, we would like to thank our patients for their precious collaboration and participation in this study.
van Dooren S.J.
X-linked creatine-transporter gene (SLC6A8) defect: a new creatine-deficiency syndrome.