Introduction

X-linked gene defects have long been considered important causes of mental retardation on the basis of the observation that mental retardation is significantly more common in male subjects than in female subjects (Lehrke 1972, 1974). X-linked mental retardation (XLMR) is subdivided into syndromic (S-XLMR) and nonsyndromic (NS-XLMR) forms depending on whether further abnormalities (in addition to mental retardation) are found on physical examination, laboratory investigation, and brain imaging. NS-XLMR accounts for approximately two thirds of all cases, and as many as 100 different genes on the X chromosome may be involved in this condition (Gecz and Mulley 2000). Studying genes of NS-XLMR has been one of the hotspots in this field.

FTSJ1 is a human homolog of the Escherichia coli 2′-O-rRNA methyltransferase FtsJ/RrmJ gene (Ogura et al. 1991; Caldas et al. 2000) and is functionally unrelated to all previously identified NS-XLMR genes. It is located on chromosome Xp11.23 and harbors 13 exons, of which exons 2–9 encode a highly conserved FtsJ domain—S-adenosylmethionine (SAMe)-binding domain (Ramser et al. 2004). It is expressed widely in fetal tissue, including brain, lung, liver, and kidney. Remarkably, expression is highest in fetal brain. In the adult brain, it is expressed in amygdala, caudate nucleus, corpus callosum, hippocampus, and thalamus (Freude et al. 2004). Human FTSJ1 protein is presumed to function as a ribonucleic acid (RNA) methyltransferase modifying untranslated RNAs and thereby plays a critical role in protein translation, that is, participate in gene expression regulation (Freude et al. 2004). According to Renieri et al. (2005), gene expression regulation is an important cellular process that is altered in patients with MR.

Different mutations in FTSJ1 were found in family MRX44 (Hamel et al. 1999) and two other families when Freude et al. (2004) performed a systematic mutation screen of brain-expressed genes at Xp11. Northern blot hybridizations suggest that the phenotype in these families most likely results from functional loss of FTSJ1 protein due to these mutations. Ramser et al. (2004) also identified a splice-site mutation in the FTSJ1 gene in a large Belgian family denominated as MRX9 (Willems et al. 1993). Reports about the association of the FTSJ1 gene with NS-XLMR are all family-based studies using linkage analysis. Most of the mutations identified in the FTSJ1 gene are single-site mutations (Freude et al. 2004; Ramser et al. 2004; Froyen et al. 2007). The patients of these families are all male subjects and are European. Association between FTSJ1 gene variations and NS-XLMR in the Chinese Han population has not been reported. In this study, three SNPs were chosen as genetic markers, and a case–control study was conducted to investigate the association of the FTSJ1 gene with NS-XLMR of children in the Qinba mountain region and whether there is a gender-specific effect of this gene.

Materials and methods

Subjects

All 556 Chinese Han subjects were recruited in the Qinba mountain region. There was a high prevalence of MR in this relatively isolated and mountainous area. The incidence ratio of 0- to 14-year-old children was about 3.19%(Guo et al. 2004). The epidemiological survey revealed family aggregation of MR in this region (Zhang et al. 2005, 2006), suggesting that genetic factors might play an important role in the etiology of MR. The intelligence of each child was screened using the Chinese Wechsler Young Children Scale of Intelligence (C-WYCSI) (Gong and Dai 1992) for 4- to 5-year-old children, and the Chinese Wechsler Intelligence Scale for Children (C-WISC) (Gong and Cai 1993) was used for 6- to 14-year-old children. The social disability (SD) scores were assessed using the adaptive scale for infants and children revised by Zuo et al. (1988). The children whose IQs were <70 combining SD scores of 8 or lower were classified as MR, and the children whose IQs were 70–79 with SD scores of 9 were classified as a borderline. The definition and the diagnosis criteria of MR and borderline were based on the Chinese Classification of Mental Disorders, Second Revision (CCMD-2-R; Psychiatry 1995) and the classification of mental and behavioral disorders from the World Health Organization (WHO) (T.W.H.Organization 1992). Subsequently, clinical examination and laboratory investigation were carried out by a group of psychologists, pediatricians, and neurologists to evaluate the children with an IQ < 80 and SD scores of 9 or lower. Cases of MR affected by pregnancy infection, toxicity, caecotrophy, birth asphyxia, cretinism, chromosomal syndrome, or other diseases such as nervous system diseases and metabolic diseases were excluded. Subjects who belonged to the genetic basis of NS-MR were included in the cases group. Controls came from the same areas and were randomly selected from families without history of MR, and there was a similar age and gender distribution to cases. In total, there were 118 cases of MR (56 boys and 62 girls), 116 borderline MR (52 boys and 64 girls), and 322 controls (167 boys and 155 girls) recruited for the study according to age, gender, and habitations.

All subjects were of the Chinese Han population and were randomly collected. Written informed consent was obtained from either the participant or the participant’s guardian after the procedure had been fully explained. The protocol was reviewed and approved by the Ethical Committee of the National Human Genome Center.

Variants identification and genotyping

FTSJ1 spans ∼10 kb in the human genome. According to the LDView (http://www.ensembl.org/Homo_sapiens/ldview), of the total six tagged single nucleotide polymorphisms (SNPs) within the FTSJ1 gene, there are four tagged SNPs (rs2268954, rs2070991, rs7066831, and rs5905692) with minor allele frequency >5%. We tested allele frequencies in 48 individuals randomly chosen from subjects by using polymerase chain reaction single-strand confirmation polymorphism (PCR–SSCP) method. The rs7066831 was then excluded from the study because of no polymorphism. Finally, three tagged SNPs with minor allele frequency >5% (rs2268954, rs2070991, and rs5905692) were selected. The three tagged SNPs nearly cover the entire gene: rs2268954 in intron 1, rs2070991 in intron 8, and rs5905692 in intron 12. The PCR–SSCP method was used in genotyping the three SNPs, and the detailed information is described in Table 1.

Table 1 Single nucleotide polymorphisms (SNPs) of FTSJ1 and their corresponding primers
Full size table

Statistical analysis

Demographic data (including the age and gender) and alleles and genotype frequencies for the three SNPs were prepared using the SPSS 10.0 software (SPSS, Chicago, IL, USA) and Microsoft Visual Studio 6.0 package (Microsoft Corporation, Washington, DC, USA). For each polymorphism, Hardy–Weinberg equilibrium (HWE) was tested using the Haploview program (http://www.broad.mit.edu/mpg/haploview/; Barrett et al. 2005). Differences in allele and genotype distributions were assessed by the Monte Carlo method with the CLUMP 2.3 with 10,000 simulations (Stephens et al. 2001). Haploview was also used to estimate pairwise linkage disequilibrium (LD) of all possible pairs of the three polymorphisms. Odds ratio (OR) and 95% confidence interval (CI) was measured with CI calculator (Newcombe 1998). Haplotypes were inferred by Bayesian methods (Stephens et al. 2001) and implemented in the UNPHASED package version 3.0.6 (http://www.mrc-bsu.cam.ac.uk/personal/frank/software/unphased/; Dudbridge 2003). Statistical significance was defined as P < 0.05. Statistical power analysis was performed using the G*Power program (Erdfelder et al. 1996).

Results

We analyzed the rs2268954, rs2070991, and rs5905692 SNPs in 322 controls and 118 MR (MR) and 116 borderline MR (border) children. Genotype frequencies of all three markers showed no deviations from Hardy–Weinberg equilibrium (HWE). A breakdown comparison of FTSJ1, an X-linked gene, was conducted between cases and controls within different gender groups.

Single locus analysis

In the male group, results of single locus analysis using CLUMP showed that there were statistical differences in allele frequencies between cases and controls for all three SNPs. The T-allele frequency of rs2268954 was higher in cases (66.7%) than in controls (49.3%, P = 0.036, OR = 2.053, 95% CI 1.041–4.05). The G-allele frequency of rs2070991 was higher in cases (72.0%) than in controls (55.9%, P = 0.043, OR = 2.029, 95% CI 1.016–4.05). The C-allele frequency of rs5905692 was higher in cases (70.8%) than in controls (50.7%, P = 0.014, OR = 2.365, 95% CI 1.174–4.761). No differences of allele or genotype frequencies were observed among the compared groups (MR cases vs. controls, or borders vs. controls) at these three markers in the female group (Table 2).

Table 2 Genotype and allele frequencies of FTSJ1 single nucleotide polymorphisms (SNPs)
Full size table

Haplotype analysis

To identify the possible haplotypes for the three target SNPs, pairwise LD analysis was performed for each pair of the three SNPs. The D′ value and r 2 value of each pair of SNPs demonstrated that the three SNPs were in strong LD (all D′ > 80%) (Table 3). Therefore, haplotype analysis was performed subsequently for haplotypes with probabilities >5%, which represented the majority of haplotype diversity.

Table 3 Pairwise linkage disequilibrium
Full size table

In the analysis of the male group, there were significant differences between cases and controls in haplotypes T-G-C and C-A-T in rs2268954 (C/T), rs2070991 (A/G), and rs5905692 (C/T). The frequency of haplotype T-G-C was higher in cases (69.57%) than in controls (45.89%, P = 0.0135, OR = 2.191, 95% CI 1.127–4.261). At the same time, the frequency of haplotype C-A-T was lower in cases (28.26%) than in controls (45.89%, P = 0.01202, OR = 0.576, 95% CI 0.292–1.135). The global haplotype frequency also showed significant differences between cases and controls (P = 0.01236 from UNPHASED software). However, no differences of haplotype frequencies were observed between cases and controls at these three markers in the female group. All P values corresponding to haplotypes are shown in Table 4.

Table 4 Estimated haplotype frequencies (rs2268954–rs2020991–rs5905692) and association significance
Full size table

Power analysis

A power analysis was performed using the G*Power program, which is based on Cohen’s method (Erdfelder et al. 1996). The sample size revealed >95% power for detecting significant association (P < 0.05), even if the tested variations had a weak to moderate gene effect (under a lower effect size index of 0.25).

Discussion

Three SNPs were selected in the FTSJ1 gene, and a case–control study was performed to investigate the relationship between the FTSJ1 gene and MR within our samples. Both single-locus and haplotype analyses showed a positive association between these three SNPs and NS-XLMR in the Chinese Han population of young male subjects of the Qinba mountain region in China.

Though the three markers (rs2268954, rs2070991, rs5905692) selected in this study are all in the intron region, they are in strong LD and nearly span the entire FTSJ1 gene. In general, a haplotype of closely located markers increases the power to detect association with the disease (Collins and Morton 1998). Therefore, though the biological functions of the three SNPs is obscure, they can still be used as a powerful tool to verify the association of FTSJ1 and NS-XLMR. The three SNPs constructed effective haplotype block whose variants reflect, to a great degree, genetic covariance of the region they span. The common disease/common variant (CD/CV) hypothesis (Reich and Lander 2001; Smith and Lusis 2002) proposes that the genetic factors underlying common diseases could be common alleles in the population. Under this hypothesis and LD information, this study suggested that at least one susceptibility locus for NS-XLMR lies within, or very close to, the FTSJ1 gene in young male subjects of the Qinba mountain region in China and that haplotype T-G-C might be a risk haplotype for NS-XLMR in the young male subjects.

FTSJ1 protein belongs to a large phylogenetically conserved family of RNA methyltransferases, and sequence conservation among all members of this protein family is high, especially in the functional domains (Bugl et al. 2000; Feder et al. 2003). Human FTSJ1 protein is presumed to function as an RNA methyltransferase for tRNA and/or for rRNA, possibly involved in the regulation of translation (Feder et al. 2003; Freude et al. 2004). So far, all mutations detected in the FTSJ1 gene of NS-XLMR patients lead to partial or entire loss of functional domain of FTSJ1 protein (FtsJ domain) (Freude et al. 2004; Ramser et al. 2004), therefore impairing its capacity to bind SAMe. As is known, SAMe, which has the ability to cross the blood–brain barrier, serves as the sole methyl donor for methylation processes in the central nervous system. So far, low cerebrospinal fluid levels of SAMe have been observed in several neuropsychiatric and neurological disorders, including depression, brain ischemia, and dementia (Chishty et al. 2002). It is probable that mutations in FTSJ1 lead to its encoded product’s activity changing in some young NS-XLMR male subjects of the Qinba mountain region in China. The impairing of its capacity to bind SAMe can affect the methylation of transfer RNA (tRNA) and/or ribosomal RNA (rRNA) and result in an alteration in the normal pattern of gene expression in a small group of genes that maybe important for correct brain functioning in neuronal cells, thereby causing MR.

Particularly intriguing is the fact that a gender-related association was found in our case–control study, which is in accordance with the observation that NS-XLMR that arises from mutations in this gene occurs in male subjects only (Freude et al. 2004; Ramser et al. 2004; Froyen et al. 2007). There maybe two possible reasons for this phenomenon. FTSJ1 is located on the X chromosome; therefore, mutations will not affect the function of the FTSJ1 protein of heterozygous female subjects. A second hypothesis is that the same X chromosome becomes inactivated in all cells in the obligate carrier in female subjects (Froyen et al. 2007). Further works needs to clarify the molecular mechanism of the gender-specific effect of FTSJ1.

In conclusion, this case–control study has shown a positive association of FTSJ1 gene polymorphisms with NS-XLMR in young male subjects in the Chinese Han population of the Qinba mountain region in China. Further research is needed to validate whether this association still exists in the Chinese Han population of other regions.