Identification of novel pro-alpha2(IX) collagen gene mutations in two families with distinctive oligo-epiphyseal forms of multiple epiphyseal dysplasia.

Multiple epiphyseal dysplasia (MED) is a genetically heterogeneous disorder with marked clinical and radiographic variability. Traditionally, the mild "Ribbing" and severe "Fairbank" types have been used to define a broad phenotypic spectrum. Mutations in the gene encoding cartilage oligomeric-matrix protein have been shown to result in several types of MED, whereas mutations in the gene encoding the alpha2 chain of type IX collagen (COL9A2) have so far been found only in two families with the Fairbank type of MED. Type IX collagen is a heterotrimer of pro-alpha chains derived from three distinct genes-COL9A1, COL9A2, and COL9A3. In this article, we describe two families with distinctive oligo-epiphyseal forms of MED, which are heterozygous for different mutations in the COL9A2 exon 3/intron 3 splice-donor site. Both of these mutations result in the skipping of exon 3 from COL9A2 mRNA, but the position of the mutation in the splice-donor site determines the stability of the mRNA produced from the mutant COL9A2 allele.     

Congenital end-plate acetylcholinesterase deficiency caused by a nonsense mutation and an A-->G splice-donor-site mutation at position +3 of the collagenlike-tail-subunit gene (COLQ): how does G at position +3 result in aberrant splicing?

Congenital end-plate acetylcholinesterase (AChE) deficiency (CEAD), the cause of a disabling myasthenic syndrome, arises from defects in the COLQ gene, which encodes the AChE triple-helical collagenlike-tail subunit that anchors catalytic subunits of AChE to the synaptic basal lamina. Here we describe a patient with CEAD with a nonsense mutation (R315X) and a splice-donor-site mutation at position +3 of intron 16 (IVS16+3A-->G) of COLQ. Because both A and G are consensus nucleotides at the +3 position of splice-donor sites, we constructed a minigene that spans exons 15-17 and harbors IVS16+3A-->G for expression in COS cells. We found that the mutation causes skipping of exon 16. The mutant splice-donor site of intron 16 harbors five discordant nucleotides (at -3, -2, +3, +4, and +6) that do not base-pair with U1 small-nuclear RNA (snRNA), the molecule responsible for splice-donor-site recognition. Versions of the minigene harboring, at either +4 or +6, nucleotides complementary to U1 snRNA restore normal splicing. Analysis of 1,801 native splice-donor sites reveals that presence of a G nucleotide at +3 is associated with preferential usage, at positions +4 to +6, of nucleotides concordant to U1 snRNA. Analysis of 11 disease-associated IVS+3A-->G mutations indicates that, on average, two of three nucleotides at positions +4 to +6 fail to base-pair, and that the nucleotide at +4 never base-pairs, with U1 snRNA. We conclude that, with G at +3, normal splicing generally depends on the concordance that residues at +4 to +6 have with U1 snRNA, but other cis-acting elements may also be important in assuring the fidelity of splicing.     

Cloning and nucleotide sequence of placental hGH-V cDNA

We have previously demonstrated the presence in human placenta and maternal serum of a GH variant, called human placental growth hormone (hPGH). We have also shown that the hGH-V gene is expressed at the placental level thus possibly coding for hPGH. The hGH-V cDNA has now been isolated from a lambda gt 11 human placenta cDNA library. Its sequence has been determined which firmly establishes the GH-V gene mode of splicing as well as the GH-V protein structure. Our data give final evidence of placental hGH-V gene expression and reinforce the hypothesis of identity between the hGH-V protein and hPGH.     

Splice-site mutations: a novel genetic mechanism of Crigler-Najjar syndrome type 1.

Crigler-Najjar syndrome type 1 (CN-1) is a recessively inherited, potentially lethal disorder characterized by severe unconjugated hyperbilirubinemia resulting from deficiency of the hepatic enzyme bilirubin-UDP-glucuronosyltransferase. In all CN-1 patients studied, structural mutations in one of the five exons of the gene (UGT1A1) encoding the uridinediphosphoglucuronate glucuronosyltransferase (UGT) isoform bilirubin-UGT1 were implicated in the absence or inactivation of the enzyme. We report two patients in whom CN-1 is caused, instead, by mutations in the noncoding intronic region of the UGT1A1 gene. One patient (A) was homozygous for a G-->C mutation at the splice-donor site in the intron, between exon 1 and exon 2. The other patient (B) was heterozygous for an A-->G shift at the splice-acceptor site in intron 3, and in the second allele a premature translation-termination codon in exon 1 was identified. Bilirubin-UGT1 mRNA is difficult to obtain, since it is expressed in the liver only. To determine the effects of these splice-junction mutations, we amplified genomic DNA of the relevant splice junctions. The amplicons were expressed in COS-7 cells, and the expressed mRNAs were analyzed. In both cases, splice-site mutations led to the use of cryptic splice sites, with consequent deletions in the processed mRNA. This is the first report of intronic mutations causing CN-1 and of the determination of the consequences of these mutations on mRNA structure, by ex vivo expression.     

Defective splicing of thyroglobulin gene transcripts in the congenital goitre of the Afrikander cattle

The structure of thyroglobulin mRNA was analyzed in an inbred herd of Afrikander cattle with hereditary goitre. Northern transfer of RNA from affected animals revealed both a shorter (approximately 7100 bases) and a normal-sized (approximately 8200 bases) thyroglobulin mRNA when hybridized to bovine thyroglobulin cDNA clones. S1 nuclease mapping experiments established that 1100 bases are deleted in the 5' region of the smaller mRNA. Electron microscopy of RNA from animals with goitre hybridized to a bovine genomic DNA clone showed that the region deleted corresponds to exon 9 of the thyroglobulin gene. Southern blot analysis of the exon 9 region revealed differences between affected and control animals with the enzymes PstI and TaqI. Although they could reflect a linkage disequilibrium between the mutation and restriction fragment length polymorphism, it is noteworthy that these differences map in the region of the exon 9/intron 9 junction. Our results show that a genetic lesion in the thyroglobulin gene causes aberrant splicing of the pre-mRNA, and suggest that the responsible mutation is at the exon 9/intron 9 junction.     

Redefinition of exon 7 in the COL1A1 gene of type I collagen by an intron 8 splice-donor-site mutation in a form of osteogenesis imperfecta: influence of intron splice order on outcome of splice-site mutation.

Most splice-site mutations lead to a limited array of products, including exon skipping, use of cryptic splice-acceptor or -donor sites, and intron inclusion. At the intron 8 splice-donor site of the COL1A1 gene, we identified a G+1-->A transition that resulted in the production of several splice products from the mutant allele. These included one in which the upstream exon 7 was extended by 96 nt, others in which either intron 8 or introns 7 and 8 were retained, one in which exon 8 was skipped, and one that used a cryptic donor site in exon 8. To determine the mechanism by which exon-7 redefinition might occur, we examined the order of intron removal in the region of the mutation by using intron/exon primer pairs to amplify regions of the precursor nuclear mRNA between exon 5 and exon 10. Removal of introns 5, 6, and 9 was rapid. Removal of intron 8 usually preceded removal of intron 7 in the normal gene, although, in a small proportion of copies, the order was reversed. The proportion of abnormal products suggested that exon 7 redefinition, intron 7 plus intron 8 inclusion, and exon 8 skipping all represented products of the impaired rapid pathway, whereas the intron-8 inclusion product resulted from use of the slow intron 7-first pathway. The very low-abundance cryptic exon 8 donor site product could have arisen from either pathway. These results suggest that there is commitment of the pre-mRNA to the two pathways, independent of the presence of the mutation, and that the order and rate of intron removal are important determinants of the outcome of splice-site mutations and may explain some unusual alterations.     

Splicing defects in the ataxia-telangiectasia gene, ATM: underlying mutations and consequences.

Mutations resulting in defective splicing constitute a significant proportion (30/62 [48%]) of a new series of mutations in the ATM gene in patients with ataxia-telangiectasia (AT) that were detected by the protein-truncation assay followed by sequence analysis of genomic DNA. Fewer than half of the splicing mutations involved the canonical AG splice-acceptor site or GT splice-donor site. A higher percentage of mutations occurred at less stringently conserved sites, including silent mutations of the last nucleotide of exons, mutations in nucleotides other than the conserved AG and GT in the consensus splice sites, and creation of splice-acceptor or splice-donor sites in either introns or exons. These splicing mutations led to a variety of consequences, including exon skipping and, to a lesser degree, intron retention, activation of cryptic splice sites, or creation of new splice sites. In addition, 5 of 12 nonsense mutations and 1 missense mutation were associated with deletion in the cDNA of the exons in which the mutations occurred. No ATM protein was detected by western blotting in any AT cell line in which splicing mutations were identified. Several cases of exon skipping in both normal controls and patients for whom no underlying defect could be found in genomic DNA were also observed, suggesting caution in the interpretation of exon deletions observed in ATM cDNA when there is no accompanying identification of genomic mutations.     

Diversity of Protein and mRNA Forms of Mammalian Methionine Sulfoxide Reductase B1 Due to Intronization and Protein Processing

Background

Methionine sulfoxide reductases (Msrs) are repair enzymes that protect proteins from oxidative stress by catalyzing stereospecific reduction of oxidized methionine residues. MsrB1 is a selenocysteine-containing cytosolic/nuclear Msr with high expression in liver and kidney.

Principal Findings

Here, we identified differences in MsrB1 gene structure among mammals. Human MsrB1 gene consists of four, whereas the corresponding mouse gene of five exons, due to occurrence of an additional intron that flanks the stop signal and covers a large part of the 3′-UTR. This intron evolved in a subset of rodents through intronization of exonic sequences, whereas the human gene structure represents the ancestral form. In mice, both splice forms were detected in liver, kidney, brain and heart with the five-exon form being the major form. We found that both mRNA forms were translated and supported efficient selenocysteine insertion into MsrB1. In addition, MsrB1 occurs in two protein forms that migrate as 14 and 5 kDa proteins. We found that each mRNA splice form generated both protein forms. The abundance of the 5 kDa form was not influenced by protease inhibitors, replacement of selenocysteine in the active site or mutation of amino acids in the cleavage site. However, mutation of cysteines that coordinate a structural zinc decreased the levels of 5 and 14 kDa forms, suggesting importance of protein structure for biosynthesis and/stability of these forms.

Conclusions

This study characterized unexpected diversity of protein and mRNA forms of mammalian selenoprotein MsrB1.     

Characterization of a new splicing mutation in the steroid 21-hydroxylase gene

We identified a novel mutation in the CYP21A2 gene, a C to G substitution in the 7-position of the intron 2 acceptor splice site (c.290-7C>G), which causes a steroid 21-hydroxylase deficiency. The effect of the mutation on splicing was checked in the system of CYP21A minigene expression in cultured mammalian cells. The mutation impairs the use of the intron 2 acceptor splice site, resulting in intron retention in mRNA.     

Human growth hormone-variant demonstrates a receptor binding profile distinct from that of normal pituitary growth hormone

We have recently established that the human growth hormone-variant (hGH-V) gene is functional in vivo by documenting its expression in the placenta. We have subsequently generated transformed murine cell lines stably expressing the genes for normal pituitary growth hormone (hGH-N), hGH-V, and each of two chimeric genes generated by exon 3 exchanges, hGH-NV3 and hGH-VN3. In the present study, we utilize these cell lines as sources of hormone to characterize and compare the receptor binding profiles of hGH-N with hGH-V. hGH-V was found to displace 125I-ovine prolactin bound to rat liver microsomes (lactogen binding) and to displace 125I-hGH bound to rabbit liver microsomes (somatogen binding). Therefore, hGH-V would be predicted to display both somatogenic and lactogenic bioactivity, a dual specificity previously thought to be unique to hGH-N. The concentrations of hormone necessary to displace 50% (IC50) of the 125I-hGH from somatogen receptors and 125I-ovine prolactin from lactogen receptors was expressed as a ratio, IC50 somatogen: IC50 lactogen, for each hormone tested. A 7.4-fold difference in this ratio was observed for hGH-N compared to hGH-V, suggesting significantly greater selectivity by hGH-V in binding to the somatogen receptor. The intermediate binding ratios of the hGH-NV3 and hGH-VN3 chimeric proteins confirmed the distinct receptor binding profiles of the two parent hormones and served to identify three amino acids of potential importance in defining their respective receptor binding specificities.     

Comparative analysis of GDF 8 (myostatin) in Bos indicus and Bos taurus.

We report the myostatin gene sequence of Bos indicus cattle in comparison to Bos taurus. B. indicus genomic sequence was obtained by overlapping PCR amplification of genomic DNA. Exon splice sites were confirmed by mRNA sequencing. There were 5 exonic single nucleotide polymorphisms (SNP) only one of which was a non-synonymous mutation that resulted in a serine to asparagine (S214N) amino acid substitution. The B. indicus gene has two insertions of 16 and 12 bases in the first intron. In addition, SNPs in the 3' UTR and intronic regions are also reported.     

Nature and frequency of mutations in the argininosuccinate synthetase gene that cause classical citrullinemia

Citrullinemia is an autosomal recessive disorder caused by a genetic deficiency of argininosuccinate synthetase (ASS). So far 20 mutations in ASS mRNA have been identified in human classical citrullinemia, including 14 single base changes causing missense mutations in the coding sequence of the enzyme, 4 mutations associated with an absence of exons 5, 6, 7, or 13 in mRNA, 1 mutation with a deletion of the first 7 bases in exon 16 (which is caused by abnormal splicing), and 1 mutation with an insertion of 37 bases between the exon 15 and 16 regions in mRNA. In order to identify the abnormality in the ASS gene causing the exon 7 and 13 deletion mutations and the 37-base insertion mutation between exons 15 and 16 in mRNA, and to establish a DNA diagnostic test, we isolated and sequenced the genomic DNA surrounding each exon. The absence of exon 7 or 13 in ASS mRNA resulted from abnormal splicing caused by a single base change in the intron region: IVS-6^–2 (a transition of A to G at the second nucleotide position within the 3 splice cleavage site of intron 6) and IVS-13⁺⁵ (a transition of G to A at the fifth nucleotide position within the 5 splice cleavage site of intron 13), respectively. The IVS-6^–2 mutation resulted in the creation of an MspI restriction site. DNA diagnostic analysis of 33 Japanese alleles with classical citrullinemia showed that 19 alleles had the IVS-6^–2 mutation (over 50% of the mutated alleles in Japanese patients). It was thus confirmed that one mutation is predominant in Japan. This differs from the situation in the USA where there is far greater heterogeneity. The insertion mutation in mRNA on the other hand resulted from abnormal splicing caused by a 13-bp deletion at the splice-junction between exon 15 and intron 15. The deletion had a short direct repeat (CTCAGG) at the breakpoint junction and presumably resulted from slipped mispairing.