Tsnp-Based Identification of Allelic Loss of Gene Expression in a Patient with a Balanced Chromosomal Rearrangement ⁎ Gregory F

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Genomics 89 (2007) 562–565 www.elsevier.com/locate/ygeno

Short Communication tSNP-based identification of allelic loss of gene expression in a patient with a balanced chromosomal rearrangement ⁎ Gregory F. Guzauskas, Kennedy Ukadike, Lynn Rimsky, Anand K. Srivastava

J.C. Self Research Institute of Human Genetics, Greenwood Genetic Center, 113 Gregor Mendel Circle, Greenwood, SC 29646, USA Received 23 August 2006; accepted 12 December 2006 Available online 22 January 2007

Abstract

Identification of genes affected by disease-associated rare chromosomal rearrangements has led to the cloning of several disease genes. Here we have used a simple approach involving allele-specific RT-PCR-based detection of gene expression to identify a gene affected by a balanced autosome;autosome translocation. We identified a transcribed SNP (tSNP), c.68G→A, present in a novel untranslated exon of the CLDN14 gene in a male patient with mental retardation who had a balanced t(13;21) chromosomal translocation. We determined an allelic loss of expression of the CLDN14 gene isoform at the 21q22.1 chromosomal breakpoint. Although additional work is necessary to explore a possible function of the novel CLDN14 isoform in brain development and function and the potential pathogenic consequences of its disruption in this patient, the result clearly demonstrates the utility of a tSNP-based detection of allelic loss of gene expression in studies involving chromosomal rearrangements. © 2007 Elsevier Inc. All rights reserved.

Keywords: Chromosome translocation; tSNP; RT-PCR; Allele-specific expression

The identification of disease genes through positional region and identification of candidate genes from the breakpoint cloning approaches has proven to be one of the most powerful critical region [5–8]. In most of these cases, the causative genes tools in unraveling gene function. Autosomal dominant diseases have been found to be physically disrupted by the chromosomal are normally caused by a defect in one of two copies of an breaks. However, in a number of cases, a long-range effect of autosomal gene. Such defects lead to either haploinsufficiency the chromosomal break on genes located at a distance has also or a gain in gene function. Identification of causative autosomal been identified [9]. Although productive, there are limitations in genes in diseases of a heterogeneous nature such as mental the use of traditional approaches for detecting genes disrupted retardation and autism has proven to be very difficult due to the or affected by a chromosomal break. lack of large families for linkage analysis [1]. Furthermore, Real-time RT-PCR expression analysis to determine expres- phenotypic variability among small families, often consisting of sion variations in genes affected by chromosomal rearrange- single sibships, precludes the pooling of families for linkage ments has also been utilized in a limited number of studies [10]. analysis. Thus, progress in the identification of autosomal However, this has not been a method of choice due to the wide mental retardation (MR) genes has been very limited [2–4]. range of observed expression variability of reference and Studies involving disease-associated rare chromosomal experimental genes among test or control samples [11,12]. rearrangements, such as balanced translocations, inversions, Allele-specific gene expression has been extensively utilized in and duplications, have been very productive in identifying the studies of gene silencing related to X inactivation and causative genes [5,6]. However, mapping of the chromosomal imprinting studies [13–15]. Here we have used a transcribed breakpoints is a labor-intensive and multistep process and SNP (tSNP)-based RT-PCR approach and show its utility in the requires delineation of the critical chromosomal breakpoint identification of a gene affected by a chromosomal breakpoint in a 19-year-old patient with a balanced t(13;21) (q22;q22.1) ⁎ Corresponding author. Fax: +1 864 388 1808. translocation, mild MR (IQ 69), macrocephaly (OFC>98th E-mail address: [email protected] (A.K. Srivastava). percentile), and aggressive behavior.

To identify the potential candidate gene(s) for the observed CLDN14 gene (variants 3 and 4; Fig. 1). These findings also clinical features in the patient, we mapped the 21q22.1 break- complemented recently identified novel isoforms of the point to an approximate 50-kb genomic region flanked by two CLDN14 gene [20]. Altogether, identification of novel previously characterized genes, CLDN14 and SIM2 (cen– CLDN14 isoforms extended the 5′ end of the CLDN14 gene CLDN14–21q breakpoint–SIM2–ter) using fluorescence in situ further within the 21q breakpoint region and suggested a likely hybridization (data not shown). It has been suggested that an physical disruption of the CLDN14 gene (Fig. 1) by the 21q22.1 extra copy of SIM2 may be important for the pathogenesis of breakpoint. We hypothesized that the chromosomal break at Down syndrome, especially mental retardation (MIM 600892) 21q22.1 should lead to a loss of expression of a copy of a [16]. However, a role for the Down syndrome critical region CLDN14 gene isoform located on the derivative chromosome genes in specific Down syndrome phenotypes has been debated 21. [17]. The SIM2 gene mapped approximately 200 kb telomeric To check for an effect of the 21q chromosomal break on the to the breakpoint and was apparently not affected by the trans- expression of a novel CLDN14 isoform, we performed location breakpoint. quantitative RT-PCR analysis. However, results were incon- The CLDN14 gene comprises three exons spanning (tel→ clusive. Therefore, we devised an alternative approach to look cen) about 19.4 kb of genomic DNA. The first two exons of the for the expression of the CLDN14 gene using allele-specific RT- CLDN14 gene encode a 5′ untranslated region; exon 3 contains PCR. First, we identified a heterozygous nucleotide variation the complete 720-bp open reading frame, encoding a 239- (c.68G→A) in the novel untranslated exon 1a of the CLDN14 amino-acid protein of the claudin gene family. An isoform of the gene in the patient (Figs. 1 and 2A, top). We used this tSNP to CLDN14 gene has been previously shown to be mutated in some examine specifically the expression of one of the two isoforms cases of nonsyndromic deafness (DFNB29) (MIM 605608) [18– of the CLDN14 gene (Fig. 1, variant 3) that was predicted to be 20]. No deafness was noted in the present patient. disrupted by the translocation breakpoint. We performed allele- To determine if one of the two genes or a new previously specific RT-PCR analysis on total RNA extracted from the unidentified expressed sequence is affected by the 21q break- patient’s EBV-transformed lymphoblasts as described pre- point in the patient, we performed computer-assisted sequence viously [21]. First-strand synthesis was performed using the analysis of the breakpoint critical region. A limited number of SuperScript first-strand synthesis system for RT-PCR (Invitro- EST sequences were identified within the breakpoint interval. gen). A cDNA amplicon was amplified using primers specific Further EST sequence analysis and analysis of the correspond- for CLDN14 exon 1a (5′-TGGCTGGGCTGGGTGGTCTC-3′) ing mouse sequences led to identification of three novel 5′ and exon 1b (5′-AGCGGCCATGAGCTGGTGTG-3′)ina untranslated exons and two new transcript isoforms of the reaction volume of 25 μl that contained 10 μM each primer,

Fig. 1. A schematic representation of the 21q22.1 chromosomal breakpoint region in the patient with the t(13;21) translocation. Arrows indicate the direction of the transcription. The entire open reading frame of the CLDN14 gene is present in exon 3. The novel alternative exons (1a, 1b, and 1c) and transcript variants/isoforms of the CLDN14 gene are shown. The respective GenBank accession numbers of the variants are in parentheses. Three tSNPs, c.68G/A in alternative exon 1a (bottom strand) and c.11C/A and c.63G/A in coding exon 3 (top strand), are indicated. 564 G.F. Guzauskas et al. / Genomics 89 (2007) 562–565

Fig. 2. Detection of an allele-specific loss of CLDN14 gene expression in the patient with t(13;21). Automated sequence traces (antisense strand) showing the CLDN14 tSNP in (A) the patient and in (B) one of two control individuals. The exon 1a-specific amplicon was amplified using genomic DNA and the gel-purified products were sequenced for both stands. The patient was heterozygous for c.68G/A (top). The corresponding RT-PCR product, amplified using exon 1a- and exon 1b-specific primers, showed only the c.68G allele in the patient total RNA (A, bottom), but both alleles, c.68G and c.68A, in the control (B, bottom). The results show that both copies of the gene are expressed in normal control individuals but only one copy of the gene is expressed in the patient.

250 μM dNTPs, 1× PCR buffer, and 1 unit of Taq polymerase role for the smaller CLDN14 isoform containing exon 1, exon 2, (Sigma) with 0.02 μM TaqStart antibody (Clontech). The RT- and exon 3 in autosomal recessive deafness has been reported PCR conditions included an initial denaturation at 95°C for [18–20]. The patient studied here reportedly has no deafness. 300 s, followed by 40 cycles of 95°C for 30 s, 69°C for 30 s, and Although a likely role for one of these isoforms in human brain 72°C for 30 s. The PCR product (140 bp) was gel purified and function is predicted, the present data are not sufficient to draw both strands were sequenced using the DYEnamic ET Dye any such conclusion and would require additional functional Terminator Cycle Sequencing Kit (GE Healthcare) and the studies. Nonetheless, the method described here was clearly automated MegaBACE 1000 DNA analysis system (GE able to show the utility of allele-specific RT-PCR in determining Healthcare). Sequences were analyzed using the DNASTAR the effect of the translocation breakpoint on the transcription of program (Madison, WI, USA). Sequence analysis revealed that one copy of the CLDN14 gene in the patient. the patient’s cDNA carried only a “G” allele at nucleotide 68 Polymorphic genomic markers and single-nucleotide poly- (Fig. 2A). Taken together, the results indicate that only one copy morphic alterations have been extensively used in genotype of CLDN14 was transcribed and suggested a loss of expression analysis using genomic DNA. SNPs as well as other of the second copy of CLDN14 located on the derivative polymorphic markers have also been used extensively to infer chromosome 21. genomic deletions, in studies involving uniparental disomy and To exclude a possibility of an imprinting effect on the loss of heterozygosity, and in imprinting studies [13,22,23]. The CLDN14 gene expression that may have caused the monoallelic approach described here utilizes the presence of a heterozygous expression of the CLDN14 gene in the patient, we analyzed nucleotide alteration in transcribed regions to identify genes both genomic DNA and cDNA from two control individuals affected by the translocation breakpoints. who were also heterozygous for the c.68G/A variation (Fig. 2B) In most autosomal dominant cases, a chromosomal and apparently carried two normal chromosomes 21. A normal rearrangement would likely result in the loss of expression biallelic expression of the CLDN14 gene was determined in of one copy of a gene either by physical disruption of the gene both controls (Fig. 2B and data not shown). These findings or due to its presence in close vicinity of the breakpoint. suggest that both copies of the CLDN14 gene are expressed in However, in the latter case, an overexpression of the gene due the control individuals. Altogether, the results would point to to physical separation of a regulatory element by the the loss of allele “A” in the patient being due to a disrupted copy breakpoint cannot be excluded. Detection of such long-range of the CLDN14 gene located on the derivative chromosome 21. effects of the autosomal breakpoints on a gene(s) located at As expected, the results were further confirmed by ampli- some distance from the breakpoint remains problematic [24]. fying the cDNA products using a primer from exon 1b and exon Utility of real-time expression analysis using quantitative RT- 3 (variant 3). Furthermore, two additional tSNPs, c.11C/T and PCR has also been very limited [10]. Thus, the tSNP-based c.63G/A, located in coding exon 3 were also examined in the RT-PCR approach described here provides a quick and patient and in two controls (Fig. 1 and data not shown). efficient screening method to analyze a large number of The functional significance of the CLDN14 isoforms genes from the breakpoint critical regions or regions flanking containing untranslated exons remains to be determined. A chromosomal breakpoints. A similar approach has been G.F. Guzauskas et al. / Genomics 89 (2007) 562–565 565 successfully applied in some cases with unbalanced X; [6] J.F. Abelson, et al., Sequence variants in SLITRK1 are associated with ’ – autosome translocations to examine spreading of X inactiva- Tourette s syndrome, Science 310 (2005) 317 320. [7] I. Borg, et al., Disruption of Netrin G1 by a balanced chromosome tion in the translocated segment of the autosomes [25]. The translocation in a girl with Rett syndrome, Eur. J. Hum. Genet. 13 (2005) approach can be easily adapted to determine quantitative 921–927. expression difference due to difference in copy number of the [8] S. Birgitt, Molecular breakpoint cloning and gene expression studies of a gene utilizing any tSNP [26,27] via information available novel translocation t(14;15) (q27;q11.2) associated with Prader–Willi through the SNP database (http://www.ncbi.nlm.nih.gov/). It is syndrome, BMC Med. Genet. 6 (2005) 18. [9] L. Crisponi, et al., The putative forkhead transcription factor FOXL2 is important to note that this approach is increasingly feasible mutated in blepharophimosis/ptosis/epicanthus inversus syndrome, Nat. because there are now over 5,000,000 SNPs available for Genet. 27 (2001) 159–166. potential use for screening of genes and isoforms, and the [10] D.A. Kwasnicka-Crawford, et al., Characterization of a novel cation annotation of the genome map is very informative. The transporter ATPase gene (ATP13A4) interrupted by 3q25–q29 candidate genes from the breakpoint regions could be easily inversion in an individual with language delay, Genomics 86 (2005) 182–194. monitored for the presence of informative heterozygous tSNPs. [11] J. Vandesompele, et al., Accurate normalization of real-time quantitative In the present study, we tested and provide a simple and RT-PCR data by geometric averaging of multiple internal control genes, efficient approach for identifying genes affected by chromoso- Genome Biol. 3 (2002) R34.1–R34.11. mal breakpoints. We have been applying this tSNP-based [12] C.L. Andersen, J.L. Jensen, T.F. Orntoft, Normalization of real-time approach to screen and identify genes affected by the quantitative reverse transcription-PCR data: a model-based variance estimation, Cancer Res. 64 (2004) 5245–5250. translocation breakpoint in other cases with chromosomal [13] K.W. Brown, et al., Imprinting mutation in the Beckwith–Wiedemann rearrangements. We have successfully analyzed a candidate syndrome leads to biallelic IGF2 expression through an H19-independent gene in a female patient with a t(3;9) translocation. Furthermore, pathway, Hum. Mol. Genet. 5 (1996) 2027–2032. we were able to screen genes present at a distance and flanking [14] J.C. Knight, Allele-specific gene expression uncovered, Trends Genet. 20 – the breakpoints in a patient with a t(2;11) translocation and (2004) 113 116. [15] L. Carrel, H.F. Willard, X-inactivation profile reveals extensive variability determined a normal biallelic expression. This suggests that the in X-linked gene expression in females, Nature 434 (2005) 400–404. approach can also be efficiently used to screen a number of [16] X. Meng, et al., Effects of overexpression of Sim2 on spatial memory and genes from the chromosomal breakpoint critical regions to expression of synapsin I in rat hippocampus, Cell Biol. Int. 30 (2006) determine a long-range effect of the chromosomal break on the 841–847. gene expression when the breakpoint does not necessarily [17] L.E. Olson, et al., A chromosome 21 critical region does not cause specific Down syndrome phenotypes, Science 306 (2004) 687–690. disrupt a gene. [18] Online Mendelian Inheritance in Man: http://www.ncbi.nlm.nih.gov. [19] E.R. Wilcox, et al., Mutations in the gene encoding tight junction claudin- Acknowledgments 14 cause autosomal recessive deafness DFNB29, Cell 104 (2001) 165–172. We thank K.B. Clarkson and C. Skinner for assistance in [20] M. Wattenhofer, et al., Different mechanisms preclude mutant CLDN14 ’ proteins from forming tight junctions in vitro, Hum. Mutat. 25 (2005) obtaining the patient s sample, D. Schultz for assistance in 543–549. sequencing, Ada King for technical assistance, and C.E. [21] V.S. Vervoort, et al., AGTR2 mutations in X-linked mental retardation, Schwartz for helpful discussion. This work was supported by Science 296 (2002) 2401–2403. a grant from the National Institutes of Child Health and [22] V.M. Miller, et al., Allele-specific silencing of dominant disease genes, – Development (RO1-HD39331). Proc. Natl. Acad. Sci. USA 100 (2003) 7195 7200. [23] K. Kittiniyom, et al., Allele-specific loss of heterozygosity at the DAL-1/ 4.1B (EPB41L3) tumor-suppressor gene locus in the absence of mutation, References Genes Chromosomes Cancer 40 (2004) 190–203. [24] L. Crisponi, et al., FOXL2 inactivation by a translocation 171 kb away: [1] J.K. Inlow, L.L. Restifo, Molecular and comparative genetics of mental analysis of 500 kb of chromosome 3 for candidate long-range regulatory retardation, Genetics 166 (2004) 835–881. sequences, Genomics 83 (2004) 757–764. [2] F. Molinari, et al., Truncating neurotrypsin mutation in autosomal recessive [25] A.J. Sharp, H.T. Spotswood, D.O. Robinson, B.M. Turner, P.A. Jacobs, nonsyndromic mental retardation, Science 298 (2002) 1779–1781. Molecular and cytogenetic analysis of the spreading of X inactivation in X; [3] J.J. Higgins, et al., A mutation in a novel ATP-dependent Lon protease gene autosome translocations, Hum. Mol. Genet. 11 (2002) 3145–3156. in a kindred with mild mental retardation, Neurology 63 (2004) 1927–1931. [26] I. Tournier, et al., Analysis of the allele-specific expression of the mismatch [4] L. Basel-Vanagaite, et al., The CC2D1A, a member of a new gene family repair gene MLH1 using a simple DHPLC-based method, Hum. Mutat. 23 with C2 domains, is involved in autosomal recessive nonsyndromic mental (2004) 379–384. retardation, J. Med. Genet. 43 (2006) 203–210. [27] R. Kimura, T. Nishioka, A. Soemantri, T. Ishida, Allele-specific transcript [5] C.S. Lai, et al., A forkhead-domain gene is mutated in a severe speech and quantification detects haplotypic variation in the levels of the SDF-1 language disorder, Nature 413 (2001) 519–523. transcripts, Hum. Mol. Genet. 14 (2005) 1579–1585.