Association of TALS Developmental Disorder with Defect in Minor Splicing Component U4atac snRNA

See allHide authors and affiliations

Science  08 Apr 2011:
Vol. 332, Issue 6026, pp. 240-243
DOI: 10.1126/science.1202205


The spliceosome, a ribonucleoprotein complex that includes proteins and small nuclear RNAs (snRNAs), catalyzes RNA splicing through intron excision and exon ligation to produce mature messenger RNAs, which, in turn serve as templates for protein translation. We identified four point mutations in the U4atac snRNA component of the minor spliceosome in patients with brain and bone malformations and unexplained postnatal death [microcephalic osteodysplastic primordial dwarfism type 1 (MOPD 1) or Taybi-Linder syndrome (TALS); Mendelian Inheritance in Man ID no. 210710]. Expression of a subgroup of genes, possibly linked to the disease phenotype, and minor intron splicing were affected in cell lines derived from TALS patients. Our findings demonstrate a crucial role of the minor spliceosome component U4atac snRNA in early human development and postnatal survival.

Taybi-Linder syndrome (TALS) is a rare autosomal recessive syndrome of as yet unknown etiology. Originally described in 1967 (1), a total of about 30 TALS cases have been reported (2). The TALS phenotype includes marked intrauterine and postnatal growth retardation; short, bowed long bones with severe delay in epiphyseal maturation; severe microcephaly; brain malformations, including pachygyria or agyria; characteristic dysmorphic features; dry skin; and sparse hair [supporting online material (SOM) text and fig. S1] and unexpected death within the first 3 years of life.

We originally mapped TALS to a 13-cM region (D2S2254 to D2S2215, 10 Mb) on chromosome 2q14 (3) in three consanguineous families (families F1 to F3, fig. S2) from the Mediterranean basin using a new homozygosity mapping approach. This approach relied on individual genome-based inbreeding estimates (4) rather than genealogical information, which is often limited. In the present study, we refined the TALS interval to 3.19 Mb by genotyping additional unaffected individuals from families F1 to F3 and a new consanguineous Moroccan family (family F4, fig. S2). No common haplotype appeared to be shared by TALS patients from families F1 to F4, ruling out a single founding effect.

Classical Sanger sequencing of candidate genes located within the newly refined 3.19-Mb TALS interval revealed no mutation (see SOM methods). This prompted us to perform targeted high-throughput sequencing of the region, rather than exome sequencing, in 10 selected individuals from TALS families F1 to F4. We selected one affected individual from each of the four TALS families and, to facilitate discrimination between rare variants and causative mutations by studying their segregation patterns, either both parents (families F2 and F4, fig. S2) or one healthy sibling (sib) not sharing the 2q14 haplotype of the affected sib(s) (families F1 and F3, fig. S2). For each individual, more than 15 million 75-base high-quality reads were obtained, 75% of which were specific to the TALS 3.19-Mb genomic interval. 100% of the nonrepetitive 1.6-Mb region was captured at least once, with a mean sequencing depth of 300x. Mean sequencing depth was >30x for over 80% of this region. Among loci, 4356 bases differed from the reference DNA sequence, but only one substitution, a G→A change, segregated in TALS families F1 to F4 and met all criteria for a putative causative mutation; that is, homozygous in probands, heterozygous in parents, and absent from healthy sibs not sharing the 2q14 haplotype of their affected sib(s) (table S1). The genomic 51 (g.51) G→A substitution is located within the U4atac small nuclear RNA (snRNA) gene, transcribed into a noncoding RNA that is unique to the minor spliceosome (5, 6). Sanger sequencing of all available individuals in TALS families F1 to F4 confirmed the presence of the g.51 G→A U4atac snRNA gene mutation. This mutation cosegregated with the disease and was concordant with linkage data (families F1 to F4, fig. S3).

Four additional TALS families (families F5 to F8) were then screened for U4atac snRNA mutations (fig. S3). Patient 8 (family F5) from India, without known consanguinity, was homozygous for the g.51 G→A U4atac snRNA gene mutation. Two unrelated white American nonconsanguineous patients (patient 9, family F6; and patient 10, family F7) were compound heterozygotes, combining the g.51 G→A mutation with g.50 G→A or g.50 G→C mutation, respectively. Finally, compound heterozygosity for g.51 G→A and g.53 C→G mutations was found in a Norwegian nonconsanguineous patient (patient 11, family F8). Mutations cosegregated with the disease.

None of these U4atac snRNA gene mutations has been described as a polymorphism (dbSNP build 130 database, table S2). In addition, in an ethnically diverse control sample of 138 individuals, none of the three unique U4atac snRNA gene mutations was encountered, although one North African individual was heterozygous for the recurrent mutation. Thus, from this control population, disease incidence appears to be low. Estimation of TALS frequency as a function of ethnic origin will require sequencing of a larger control sample with diverse ethnic origins.

The U4atac snRNA is located within intron 2 of the CLASP1 gene, –682 base pairs (bp) to –556 bp upstream of exon 3. Because of this, TALS mutations could, in theory, alter CLASP1 splicing and/or expression. In silico splice site predictions, real time–quantitative polymerase chain reaction (RT-qPCR) analysis of CLASP1 mRNA levels, and an RT-PCR study of CLASP1 exon 3 splicing in TALS patients revealed no effects of TALS mutations on CLASP1 expression or splicing (SOM methods and fig. S4).

U4atac forms a base-paired duplex with U6atac, required to activate minor splicing (Fig. 1). In the duplex, two intermolecular base-paired regions, stem I and stem II, are separated by an intramolecular base-paired 5′ stem-loop (7). We investigated how the four TALS substitutions affect the structure of the U4atac snRNA molecule, and found that they disrupt the 5′ stem-loop (fig. S5). In silico modeling of the U4atac/U6atac bimolecular structure predicts that these mutations drastically modify a site critical to 15.5-K protein binding and subsequent Prp31 protein binding (811). Binding of these proteins stabilizes the U4atac/U6atac.U5 tri small nuclear ribonucleoprotein (snRNP) and is critical to splicing activity (12, 13). Accordingly, U4atac snRNA expression plasmids containing various point mutations or deletions within the 5′ stem-loop were either inactive or showed partial splicing activity (10). It is therefore likely that TALS mutations within U4atac snRNA molecules affect minor splicing efficiency. Indeed, U4atac snRNA 50G, 51G, and 53C nucleotides are highly conserved across several species, including mouse, dog, and opossum (fig. S6). Several species such as zebrafish contain a g.51 G→A mutation, but in these cases there are compensatory mutations restoring the 5′ stem-loop structure. The fact that the structure is tightly controlled across species suggests that the sites mutated in TALS patients are structurally important and affect spliceosomal function.

Fig. 1

Two-dimensional model of wild-type human U4atac/U6atac snRNAs, adapted from Padgett and Shukla (7) and from Liu et al. (12). The 5′ stem-loop results from U4atac snRNA intramolecular base pairing. Independent TALS mutations are indicated on the U4atac snRNA sequence by red arrowheads, and the 15.5-K protein binding site is highlighted in blue.

The minor U12-dependent spliceosome is a ribonucleoprotein complex comprising U11, U12, U4atac, U5, and U6atac snRNAs. It is both structurally and functionally related to the U1, U2, U4, U5, and U6 snRNAs of the major U2-dependent spliceosome; U5 snRNA is shared by both spliceosomes. The human genome contains ~700 U12-type introns removed by the minor spliceosome (14, 15). U12-type introns are characterized by their consensus splice recognition sequences, combining a nearly invariant 5′ splice site, either GTATCCT or ATATCCT, and a highly conserved branch site (16). Most genes containing U12-type introns (hereafter U12 genes) are either involved in key cellular functions such as DNA replication and repair, transcription, RNA processing and transport, translation, and cytoskeletal organization, or belong to a group of cellular ion channels. In TALS patients, U4atac snRNA gene mutations are expected to result in defects in minor splicing. This could produce abnormal transcripts containing either unspliced or alternatively spliced U12-dependent introns, often leading to the introduction of premature termination codons. In some cases, these aberrant mRNA molecules may be destroyed by nonsense-mediated mRNA decay as part of proofreading of nascent mRNA transcripts (17, 18). The effect of the homozygous recurrent g.51 G→A U4atac snRNA gene mutation on the expression of a subset of 23 genes spliced by the minor U12-dependent spliceosome was investigated. As expected in TALS fibroblasts, expression of several U12 genes was reduced (Fig. 2 and fig. S7). This could be the result of either rapid destruction of abnormal transcripts or down-regulated transcription. However, expression of a large group of U12 genes was normal or even slightly increased in TALS fibroblasts. In Drosophila, inactivation of U6atac snRNA led to mild perturbation of gene expression and splicing for most U12 genes, although the metabolic pathway downstream of mitochondrial prohibitin is particularly affected (19, 20). Expression of the human homolog of Drosophila prohibitin, the U12 gene PHB2, is also markedly reduced in TALS fibroblast cell lines (Fig. 2), suggesting that this gene or its downstream effectors may play a role in the TALS phenotype.

Fig. 2

A restricted number of U12 genes show decreased expression in fibroblasts derived from TALS patients. RT-qPCR determination of relative expression levels for U12 genes in fibroblasts derived from TALS2 and TALS6 patients, both harboring the g.51 G→A mutation, and fibroblasts derived from two control individuals matched for age and gender (six replicates for each experiment) are shown. Results within patient and control groups were similar; therefore expression levels are shown as mean values (12 replicates) with a 1 SD error bar (see SOM methods). Asterisks indicate statistical significance at 5%. The genes studied are indicated on the x axis: CLASP1, located within the refined TALS genomic interval; the U4atac snRNA gene; and 12 U12 genes. The y axis indicates relative expression ratios. mRNA expression was reduced for a subset of U12 genes (DIAPH3, E2F2, GPAA1, and PHB2) in TALS-derived fibroblasts. U4atac snRNA expression was normal.

Splicing defects may lead to intron retention. We therefore used RT-qPCR experiments to examine the accumulation of transcripts retaining U12-type introns in TALS fibroblast cell lines. Statistically significant inhibition of minor intron splicing was detected in almost all U12-type introns tested (Fig. 3). Unspliced U12-type introns were increased up to sixfold in TALS versus control fibroblast cell lines, indicating a defect in minor spliceosome efficiency. Semiquantitative RT-PCR studies with gel visualization confirmed the defect in U12-type intron splicing in TALS fibroblast cells but showed the prevailing presence of normally spliced mRNA in these cells (fig. S8). Thus, the homozygous g.51 G→A U4atac snRNA recurrent mutation is “hypomorphic,” reducing the activity of the minor U12-dependent spliceosome rather than completely abolishing it. Alternatively, an as yet unknown biological mechanism may partly compensate for the minor spliceosome deficiency resulting from mutations in U4atac snRNA.

Fig. 3

Amounts of unspliced U12-type introns are markedly increased in mRNAs from TALS-derived fibroblasts. U12 introns from six U12 genes were quantified using RT-qPCR in TALS2- and TALS6-derived fibroblast cell lines. Results are expressed as relative values, as in Fig. 2. Asterisks indicate statistical significance at 5%. Unspliced U12-type introns detected in TALS-derived fibroblasts were up to six times more abundant than in control fibroblasts. For IPO9, only a slight increase in the unspliced U12-type intron was observed in TALS-derived fibroblasts. For GPAA1, the splicing defect for the U12 intron may be partially obscured by this gene's low expression level (Fig. 2).

This paper shows compelling evidence that mutations in the U4atac snRNA gene are responsible for the early postnatal sudden death and severe brain and bone malformations seen in TALS, through minor spliceosome deficiency. Decreased expression of a restricted number of U12 genes and subsequent effects on downstream metabolic pathways may explain the TALS phenotype.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S8

Tables S1 to S4


References and Notes

  1. We express our heartfelt thanks to the patients and families who participated in this study. We thank M. Gallagher-Gambarelli for her editorial assistance and R. Hennekam for his assistance during the project; M. Letexier for her assistance in analyzing high-throughput sequencing data; T. Rio Frio for helpful discussions on splicing analyses; M. de Tayrac and C. Depienne for helpful discussions on expression data analysis; I. Rouvet and M.-T. Zabot for fibroblast banking; and R. Padgett for providing us with the hand-built U4atac/U6atac bimolecular structure reflecting mfold software predictions. This work was supported by the Fondation Yves Cotrel-Institut de France, the Hospices Civils de Lyon, the French Ministry of Health (Direction de l’Hospitalisation et de l’Organisation des Soins, Plan Maladies Rares 2004), Lyon 1 University, the INSERM, the Association GEN’HOM, and the Texas Scottish Rite Hospital for Children Research Fund.

Stay Connected to Science

Navigate This Article