Sequence Variants in the Transforming Growth β-Induced Factor (TGIF) Gene Are Not Associated with High Myopia

Abstract

purpose. High myopia is a common complex-trait eye disorder, with implications for blindness due to increased risk of retinal detachment, macular degeneration, premature cataracts, and glaucoma. Mapping studies have identified at least four loci for nonsyndromic autosomal dominant high myopia at 18p11.31, 12q22-q23, 17q21-q23, and 7q36. The smallest haplotyped interval for these loci is that of the MYP2 locus on 18p11.31. Recently, the transforming growth β-induced factor (TGIF) gene was reported to be a candidate gene for MYP2-associated high myopia in single-nucleotide polymorphism studies. The purpose of this study was to determine whether DNA sequence variants in the human TGIF gene are causally related to MYP2-associated high myopia.

methods. The protein coding regions and intron-exon boundaries of the human TGIF gene were sequenced using genomic DNA samples from MYP2 individuals (affected, unaffected) and external control subjects. The TGIF model used was the April 20, 2003, human genome National Center for Biotechnology Information (NCBI) build 33, which has 10 exons and encodes eight transcript variants. Polymorphic sequence changes were compared to those in the previous report. Reverse-transcription polymerase chain reaction (RT-PCR) was performed to validate TGIF gene expression in ocular tissues.

results. A total of 21 polymorphisms of TGIF were found by direct sequencing: 3 were missense, 2 were silent, 10 were not translated, 4 were intronic, and 2 were homozygous deletions. The 3 missense allelic variants were localized to exon 10 at positions 236C→T(Pro→Leu), 244C→T(Pro→Ser), and 245C→T(Pro→Leu). Silent mutations were observed in exon 10 at positions 177A→G, 333C→T. Ten polymorphisms were novel. No sequence alterations were exclusively associated with the affected disease phenotype. RT-PCR results confirmed expression of TGIF in RNA samples derived from human sclera, cornea, optic nerve, and retina.

conclusions. TGIF is a known candidate gene for MYP2-associated high myopia, based on its mapped location within the MYP2 interval. Mutation analysis of the encoded TGIF gene for MYP2 autosomal dominant high myopia did not identify sequence alterations associated with the disease phenotype. Further studies of MYP2 candidate genes are needed to determine the gene that causes of this potentially blinding disorder.

Myopia affects approximately 25% of the adult population of the United States ^{1 2 3 4 5} and is a significant public health problem, especially in Asian populations such as Chinese or Indians, as it is associated with increased risk for visual loss. ^{1 6 7 8 9 10} Myopic chorioretinal degeneration due to high myopia is the fourth most frequent cause of blindness, leading to registration for visual services and disability and accounting for 8.8% of all causes. ¹¹ It has been estimated that 5.6% of blindness among school children in the United States is attributable to high myopia. ¹¹ Substantial resources are required for optical correction of myopia with spectacles, contact lenses, and, more recently, surgical procedures such as photorefractive keratectomy. The market for optical aids in the United States was estimated to exceed $8 billion in annual sales in 1990; most dollars were spent for the correction of myopia. ¹¹ The development of methods for preventing the onset or limiting the progression of myopia would be of considerable importance. 

Our laboratory identified the MYP2 locus in seven families with nonsyndromic autosomal dominant (AD) high myopia of −6.00 D or greater. We demonstrated significant linkage to the short arm of chromosome 18, region 11.31, with a maximum cumulative LOD score of 9.59 at θ = 0.0. ¹² The 7.6 cM recombinant interval was defined distally by marker D18S59 and proximally by marker D18S1138, with recombinants in pedigrees 1, 4, and 5 (Table 1) . The genetic boundaries of the MYP2 region are currently defined by linkage analysis of these seven existing MYP2 pedigrees, which represent the group of MYP2-affected families we have screened for mutations at the MYP2 locus. 

In an effort to contract the MYP2 interval, transmission disequilibrium test (TDT) statistics ¹³ were obtained with the Statistical Analysis for Genetic Epidemiology Transmission Disequilibrium Test (SAGE-TDTEX) ¹⁴ and Genehunter-TDT (GH2-TDT) ¹⁵ programs. Both programs examine each allele separately to look for increased frequency of disequilibrium or nonrecombination events on disease-bearing chromosomes over normal chromosomes, using a standard one-sided test (Fisher exact test). The SAGE program also calculates a summary χ² for each marker, as it examines the degree of linkage disequilibrium at the marker. TDT analysis was focused on eleven 18p markers used for fine mapping in the original study. ¹⁶ The significance values determined by both programs are listed in Table 1 for each marker locus in marker order for the 18p11.31 region. Markers D18S52 and D18S1138 show the strongest statistical association with the disease phenotype. These data suggest that the MYP2 gene is likely within a 2.2-cM interval between D18S52 and D18S481. 

Critically important are the recent independent confirmations of the MYP2 locus with an Italian patient population with AD high myopia by Heath et al. ¹⁷ and six families of Hong Kong Chinese descent by Lam et al. ¹⁸ The mapping studies of both laboratories support directing further gene identification efforts to the centromeric region of the initial 7.6-cM recombinant interval. These results, combined with our studies, provide a basis for focused positional candidate gene analysis at the MYP2 locus, as the interval of interest has likely contracted significantly from the initial 7.6 cM. 

We constructed a physical bacterial artificial chromosome (BAC) contig map across the MYP2 critical region, shown in Figure 1 , by taking advantage of the multiple databases available in conjunction with the Human Genome Project (HGP). Integration was obtained by mapping markers of different types (monomorphic, polymorphic, genes and expressed sequence tags [ESTs]) from different sources (e.g., NCBI, Genethon, Whitehead Institute, UCSC- “Golden Path”, Celera; see listing at end of article). The core region extends from marker D18S481 to D18S52. It ranges in depth from 1 to 9 BACs, with an average depth of −4 BACs and requires 19 overlapping BACs, averaging 150 to 200 kb, to span the MYP2 region. The MYP2 critical region on the short arm of chromosome 18 is now fully sequenced and is a 1.2-Mb region on contig NT_010859.13. There are six known and nine hypothetical genes that map within the MYP2 interval. All the sequences in this region are now labeled “finished” sequences. 

One gene that maps within the 18p11.3 interval is the transforming growth β-induced factor (TGIF) gene. TGIF is a DNA-binding homeo-domain protein that belongs to the TALE homeobox family. ^{19 20} It is a transcription repressor with multiple actions, including a role in retinoid-responsive transcription. ²¹ TGIF mutations are associated with holoprosencephaly, a congenital craniofacial and brain anomaly disorder. ^{22 23 24 25}  

The direct analysis of sequence within a critical region can be the most accurate, precise, and efficient approach to disease gene identification. This is particularly true for instances where the “perfect” candidate gene (based on function or expression) does not exist within a defined critical region. It is also true for a disorder such as myopia, in which the temporal and spatial expression of the disease gene is not known, and could be restricted to early development and to any eye component. All genes that map within the MYP2 critical region are candidate disease genes based on position. TGIF for example, therefore, is a candidate gene for the MYP2-associated high myopia based on map position alone. Sequence variants must be uncovered only in affected subjects compared to unaffected subjects for a fully penetrant dominant disorder such as MYP2-linked high myopia. 

A recent report by Lam et al. describes a TGIF sequence variation study of the 3-exon transcript variant 4 using conformation specific gel-electrophoresis. ²⁶ They found 25 single-nucleotide polymorphism (SNPs) on exon 3 (exon 10 in our study). Six SNPs showed significant high myopia association with univariate analysis, and one showed significance with multivariate analysis. 

We sought to determine whether the TGIF gene is causally related to MYP2-associated high myopia by direct DNA sequencing, using DNA samples from the original MYP2 pedigrees. One consideration is that the TGIF genetic structure studied by Lam et al. ²⁶ had 3 exons—the current sequence build is a 10-exon gene structure. Exons 1, 2, and 3 are now exons 5, 9, and 10, respectively, according to the reference sequence build 33 (http://www.ncbi.nlm.nih.gov/genome/guide/human/HsStats.html) of TGIF, which corresponds to transcript variant 4. 

Methods

Patients

Probands, and affected subject representatives of the seven MYP2-affected families with an AD form of high myopia were studied (Table 2) . Each of the affected individuals had high myopia of −6.00 D sphere or more with elongated axial lengths. Clinical details regarding the complete pedigrees have been published. ¹² Controls were obtained from family marry-ins, nonmyopic family members, and unrelated subjects. Table 2 displays the family and member number of each individual, as well as controls with refractive error. Twenty subjects were studied, of which 10 had high myopia and 10 were nonmyopic. 

Total genomic DNA was extracted from 10 to 15 mL of venous blood from all participants after informed consent was obtained. DNA was purified from lymphocyte pellets according to standard procedures using a kit (Puregene⁴ kit; Gentra Systems, Minneapolis, MN) or the phenol-chloroform extraction method. The study protocol was approved by the Children’s Hospital of Philadelphia Institutional Review Board on Human Subjects Research and adhered to the tenets of the Declaration of Helsinki. 

DNA Amplification and Mutation Screening

The genomic structure of TGIF, as reported in MapViewer (build 33) of the reference human genome sequence is outlined in Figure 2 . The genomic structure of TGIF contains 10 exons spanning 46 kb and has eight transcript variants encoding four proteins of 402 residues (variant 1), 287 residues (variant 2), 273 residues (variants 3 and 4), and 253 residues (variants 5 to 8). All participant DNA samples were also screened for sequence variants on exon 7, although it has no continuous open reading frame with the conserved region of exons 9 and 10, and only one known corresponding expressed sequence tag (EST). 

Twelve oligonucleotide primer pairs were designed to amplify the exonic sequences with 50 to 200 bp extensions beyond the intron-exon boundary (Table 3) . Polymerase chain reactions were performed on 150 ng genomic DNA with standard methods. Amplified products were separated by agarose gel electrophoresis and visualized by staining with ethidium bromide. Amplicons were purified in purification columns (QIAquick; Qiagen, Valencia, CA) and sequenced using dye terminator chemistry (BigDye Terminator ver. 3.1 on a model 3700 Genetic Analyzer; Applied Biosystems, Foster City, CA). 

Sequences were trimmed for quality, BLASTed and aligned (Sequencher; Gene Codes, Ann Arbor, MI. or Gap4; Staden, Cambridge, UK). Sequence alterations were annotated and compared between normal and affected individual DNA samples. 

Reverse Transcription-Polymerase Chain Reaction

Total RNA from retina, cornea, optic nerve, and sclera was isolated from four pooled human donor eyes from the Pennsylvania Lions Eye Bank (Philadelphia) using extraction reagent (TRIzol; Invitrogen-Gibco, Grand Island, NY). The eyes were treated by submersion in RNA stabilization solution (RNALater; Ambion Inc., Austin, TX) within 2 to 12 hours after death. Reverse transcription-polymerase chain reaction (RT-PCR) was performed with random hexamers using standard methods to synthesize cDNA (SuperScript II; Invitrogen Corp., Carlsbad, CA). One microgram of total RNA from the sclera, optic nerve, retina, and cornea, as well as commercially prepared poly-A RNA (BD Biosciences-Clontech Inc., Palo Alto, CA, and Ambion Inc.) from various human organs were used as templates for a 20 μL first-strand cDNA synthesis reaction. Gene-specific PCR was performed using platinum Taq polymerase according to recommended conditions, using 2 μL of each cDNA sample and 50 pM of primer in a final reaction volume of 50 μL. The PCR cycling conditions included an initial denaturation for 120 seconds at 95°C, followed by 34 cycles of denaturation for 15 seconds at 95°C, annealing for 30 seconds at 54°C, extension for 45 seconds at 68°C, and a final extension for 4 minutes at 68°C. The sense (5′GGGAGAGAGTTGGGCGAGGGA-3′) base pair 59-79 on NM_003244 and antisense (5′-TGCCTGAGCCAGCGGATGA-3′) base pair 419-401 on NM_003244 PCR primer pairs amplify a 360-bp product spanning exons 5 (sense) and 9 (antisense), detecting transcripts 3 and 4. The RT-PCR products, along with the amplicon products of the housekeeping gene β-actin were visualized on 2% agarose gels (Fig. 3) after electrophoresis and staining with ethidium bromide. 

Results

A total of 21 polymorphisms were found in the 10 exons screened (Table 4) . Of these, 3 were missense variances, 2 were silent, 10 were not translated, 4 were intronic, and 2 were homozygous deletions. The three missense allelic variants were observed at exon 10 at positions 236C→T (Pro→Leu), 244C→T(Pro→Ser), and 245C→T(Pro→Leu). Silent mutations were observed on exon 10 at positions 177A→G and 333C→T. The two deletions causing frameshift mutations were observed in exon 6 at positions 3442216 and 3442223 on NT_010859.13. Both deletions are predicted to cause early termination, yielding proteins of 132 and 141 residues, respectively. Ten polymorphisms were novel and have been submitted to the dbSNP database (http://www.ncbi.nlm.nih.gov/SNP/) provided in the public domain by NCBI. Eleven polymorphisms corresponded with previously reported SNPs in public databases. None of the sequence variants cosegregated with the affected phenotype. No heterozygous or homozygous polymorphisms were specific to affected individuals in any MYP2 pedigree. 

Notably, 18 of the 25 polymorphisms published by Lam et al. ²⁶ did not match with the published wild-type sequence (accession no. NM_003244). Of these 25 previously published SNPs, only one at position 804A→G was observed in our investigation. We observed a C→T polymorphism at bp 245 on exon 10, which presumptively corresponds to the 804 A→G base pair change. 

RT-PCR results confirmed TGIF expression in all ocular and nonocular tissues (Fig. 3) . 

Discussion

Nonsyndromic high myopia is a common, complex disorder that is likely to result from alterations of multiple genetic factors. Indeed, several loci have been mapped for the disorder. 

We sequenced the full TGIF gene in our patient samples of individuals from pedigrees with MYP2-associated high myopia. No DNA sequence variants were noted that implicated TGIF as the causative gene. TGIF exon 10 (exon 3 in the initial build of this gene) did not show the same level of polymorphic variants in our cohort, as we observed 8 variants rather than the 25 reported by Lam et al. ²⁶ This may be due to the ethnic differences in our two sample sets, although family 1 of the MYP2 pedigrees studied was of Chinese descent. All other families were of Northern European descent. 

In conclusion, TGIF is a known candidate gene for MYP2-associated high myopia based on its mapped location within the MYP2 interval. Mutation analysis of the encoded TGIF gene for MYP2 AD high myopia did not identify sequence alterations associated with the disease phenotype. Further studies of MYP2 candidate genes are needed to determine the molecular genetic factors that cause this potentially blinding disorder. 

Electronic Database Addresses

Electronic databases (listed below) were used for developing a physical map of the MYP2 critical region. The Genethon, Whitehead, and NCBI websites were queried for microsatellite marker data. The NCBI, UCSC, and Celera websites were used to align BACs and close gap regions. The hypothetical genes were determined by the GENSCAN and OTTO websites. All listed links are free to the public except for OTTO and Celera.

•  
  Celera: http://cds.celera.com/cds
•  
  Genethon: http://www.genethon.fr/php/index_us.php
•  
  GENSCAN: http://genes.mit.edu/GENSCAN.htm
•  
  MapViewer: http://www.ncbi.nlm.nih.gov/mapview/map_search.cgi
•  
  NCBI: http://www.ncbi.nlm.nih.gov/
•  
  OTTO: http://cds.celera.com/biolib/info
•  
  Public SNP repository: http://www.ncbi.nlm.nih.gov/SNP
•  
  UCSC “Golden Path”: http://genome.ucsc.edu/
•  
  Whitehead Institute: http://www-genome.wi.mit.edu/

 Supported by National Eye Institute Grant EY00376-03, Research to Prevent Blindness, Mabel E. Leslie Research Endowment Funds, and the Lions Eye Bank of Delaware Valley.

 Submitted for publication August 25, 2003; revised January 7, 2004; accepted February 10, 2004.

 Disclosure: G.S. Scavello, None; P.C. Paluru, None; W.R. Ganter, None; T.L. Young, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

MYP2 Locus Marker Recombinants and TDT Allelic Association Analysis

Marker Distance (cM)	Pedigree Marker (Telomere)	1	4	5	SAGE-TDTEX P	GH2-TDT P
	D18S1140	+	−	−	0.036	0.083
0.1<
	D18S59	+	−	−	0.013	0.317
1.5<
	D18S476	−	−	−	0.007	0.045
0.1<
	D18S1146	−	−	−	0.227	0.083
4.5<
	D18S481	−	−	−	0.001	0.108
1.4<
	D18S63	−	−	−	0.062	0.034
0.1<
	D18S1138	−	−	+	3.9 × 104	0.011
0.7<
	D18S52	−	−	+	1.79 × 106	0.007
9.4<
	D18S62	−	+	+	0.141	0.479
18.6<
	D18S1150	+	+	+	0.018	0.096
4.1<
	D18S1116	+	+	+	0.214	0.683

 Dark shaded area, region excluded by recombinants; boxed light shaded area, the most significant marker associations with the phenotype by transmission disequilibrium test analysis. +, recombinant marker; −, nonrecombinant marker.

Figure 1.

 

Physical map of the 18p11.31 critical region. The horizontal scale in megabases (M) is at the top. Polymorphic microsatellite markers are labeled above the scale in grey. Below the scale, finished (phase 3) BAC clones are labeled in black lines, working draft (phase 2 or 1) BAC clones are in grey lines, known genes are in dark grey boxes, and in silico predicted genes by GENSCAN (http://genes.mit.edu/GENSCAN.htm) are in light grey boxes.

Table 2.

 

Subject DNA Samples Used in the Study

Original MYP2 Pedigree-Individual	Refractive Error
		OD	OS
1-15	−10.50 + 2.50 × 82	−11.50 + 2.75 × 97
1-16	−11.00 + 0.75 × 60	−12.25 + 1.00 × 70
1-19	−6.00 + 1.50 × 50	−6.50 + 0.75 × 80
1-21	Plano	Plano
1-23	Plano	Plano
2-6	−16.00 sph	−16.00 sph
2-10	Plano	Plano
3-6	−19.25 + 2.25 × 135	−21.00 + 3.00 × 60
4-7	−6.25 sph	−6.00 sph
4-8	Plano	Plano
5-6	−10.00 + 1.50 × 46	−8.25 + 1.75 × 105
5-9	+0.25 + 0.50 × 170	+0.25 + 0.25 × 170
6-4	−7.25 + 1.25 × 180	−7.25 + 1.25 × 20
6-6	Plano	Plano
7-5	−10.75 + 2.00 × 10	−10.50 + 2.50 × 70
External control-1 (C1)	Plano	Plano
External control-2 (C2)	Plano	Plano
External control-3 (C3)	Plano	Plano
External control-4 (C4)	Plano	Plano
External Myopic control-5 (C5)	−9.00 sph	−9.00 sph

 Reference to the individual’s family identification relates back to the study by Young et al. ¹² , who first described the MYP2 myopia locus. The refractive errors of all control participants are also provided. C, control; sph, sphere.

Figure 2.

 

The associated ∼47.6-kb region of NT_010859 on 18p11.31 of TGIF, showing 10 exons with alternative start sites and splicing that generate eight transcript variants. Boxes: exons; vertical line with arrowhead: initiation codons; vertical line with square: stop codons.

Table 3.

 

TGIF Gene Primers Designed for Mutation Screening

Exon	Size (bp)	Primer Sequence 5′–3′	Primer
1	754	CAGGAAACACAGACCAGCTTATCTT	TGIF1F
		AATGCAGTCATGCCACTCGATATA	TGIF1R
2	249	CCCAAATTGTCTATCGGTGA	TGIF2F
		ATGACTAGGTTCAAGCCAATG	TGIF2R
3	449	TTTTGAGGCTGTTCCCTTTGTTACG	TGIF3F
		TGGACTTGGCTAATGACTGCGATT	TGIF3R
4	632	GAGGCGGCTGTCTCGTGCGGCTAGA	TGIF4F
		GATCCCAGGCGCCCGCTCCTT	TGIF4R
5	662	CCGCCCCCGAGGGACGAGT	TGIF5F
		AGCCGCTGCCGTTTCAGACGC	TGIF5R
6	1061	GTCCCCCGAGTGTTCCGCTGT	TGIF6F
		CCCACCCCAGACGACTCGC	TGIF6R
7	711	CCTCTGAGAAGGTGGGATTCACG	TGIF7F
		CCTTTGAGCTGGGAGCAATGTCTCT	TGIF7R
	796	GAATTCCTGCTTTGGGATTGC	TGIF7AF
		CACAAGCTTCTTTCACGCTATCCAC	TGIF7AR
8	329	AGAACGTGCAGGAACAAGGTCG	TGIF8F
		CACTGCAGTAAAACCACGTTTGCTG	TGIF8R
9	361	CTCGCTCTCAGTTGTTGG	TGIF9F
		TCACGCTCTCTTTCTTTA	TGIF9R
	562	TCTTCGGGATTGGCTGTATGA	TGIF9AF
		TCCCCATTTTAACCTATCACCTACT	TGIF9AR
10	427	TCAATTAGGTACCCCATAGAACAT	TGIF10F
		TCCCACACCGACCGACTG	TGIF10R
	385	TCTGCCATACCACTGTGACTG	TGIF10AF
		AATATATAAAATAATGCAATTCATCTCTTG	TGIF10AR
	284	GGGAATGAGAAGATGCTAT	TGIF10BF
		GGTGAAGGCAAGAGATGAAT	TGIF10BR

 Exon 7, 9, and 10 primers are divided into 2, 2, and 3 segments, respectively. The primers for TGIF9F/R were initially published by Gripp et al. ²⁴ , and primers for TGIF9AR, TGIF10F/R, and TGIF10AF/R were originally published by Chen et al. ²⁵

Figure 3.

 

Polymerase chain reaction β-actin and TGIF amplicons of reverse-transcribed human RNA from ocular tissues and commercially available poly-A RNA tissue types. Lane 1: sclera; lane 2: cornea; lane 3: optic nerve; lane 4: retina; lane 5: lung; lane 6: skeletal muscle; lane 7: heart; lane 8: trachea; lane 9: kidney (lanes 5–9 from BD Biosciences-Clontech, Inc.); lane 10: brain (from Ambion, Inc.).

Table 4.

 

Observed Sequence Polymorphisms in the TGIF Gene

NT_010859.13 Position	Wild-Type Nucleotide	Base Pair Change Observed	SNP rs	Original MYP2 Family-Individual	Exon Position	AA Change
3402167	C	C→T, T→T	Novel	1-23, -15, -16, and -19, 2-10, 5-6, C2 and 3	bp96 of exon 1	5′ UTR
3402386	C	C→T	rs8092903	C2	Intron 1
3402523	C	C→T, T→T	Novel	5-6, 7-5, C2	Intron 1
3408219	T	C→T	Novel	7-5	Intron 2
3437725	C	C→T	238137	1-16, 1-21, C2	bp118 of exon 3	5′ UTR
3437871	T	C→T	rs151472	1-15, 1-19, 4-7, 5-6, 7-5, 1-21, 1-23, 2-10, C2	Intron 3
3440455	C	C→A	rs238132	4-8, 6-4, C2, -3, and -4	bp284 of exon 5	5′ UTR
3441762	C	C→T, T→T	rs238533	4-7, 6-4, 4-8, C4	bp172 of exon 6	5′ UTR
3441896	G	G→A, A→A	Novel	4-7, 6-4, 4-8	bp306 of exon 6	5′ UTR
3442978	C	C→T, T→T	2238536	4-7, 6-4, 4-8	bp374 of exon 7	UTR
3442216	C	C→C Deletion	Novel	C1	bp626 of exon 6	frameshift
3442223	T	T→T Deletion	Novel	3-6, C1	bp633 of exon 6	frameshift
3443789	T	G→T	Novel	1-16 and 21, 2-6	bp18 of exon 8	5′ UTR
3447539	A	G→A, G→G	rs2229337	4-7, 6-4, C2, -4, and C5	bp177 of exon 10	NONE
3447598	C	C→T	Novel	6-6, C1	bp236 of exon 10	PRO→LEU
3447606	C	C→T	rs4468717	3-6	bp244 of exon 10	PRO→SER
3447607	C	C→T, T→T	rs2229333	4-7, 6-4, 4-8	bp245 of exon 10	PRO→LEU
3447695	C	C→T	rs2229335	C1 and -4	bp333 of exon 10	NONE
3447776	T	T→G	rs2229336	C1 and -4	bp414 of exon 10	3′ UTR
3448161	G	G→A	Novel	5-9	bp799 of exon 10	3′ UTR
3448260	G	G→A	Novel	4-7, 5-6 and -9, 6-4, 7-5, C2	bp898 of exon 10	3′ UTR

 The wild-type sequence is derived from the scaffold sequence NT_010859 of 18p. Amino acid changes are for relevant splice variants. Affected individual sample numbers are in bold. C, control, UTR, untranslated region, rs, public reference SNP number from the dbSNP database.

The authors thank the families for their participation.