The genomic structure, chromosomal localization, and analysis of SIL as a candidate gene for holoprosencephaly

Holoprosencephaly (HPE) is the most common congenital malformation of the brain and face in humans. In this study we report the analysis of SIL (SCL interrupting locus) as a candidate gene for HPE. Fluorescent in situ hybridization (FISH) analysis using a BAC 246e16 confirmed the assignment of SIL to 1p32. Computational analysis of SIL at the protein level revealed a 73% overall identity between the human and murine proteins. Denaturing high performance liquid chromatography (dHPLC) techniques were used to screen for mutations and these studies identified several common polymorphisms but no disease-associated mutations, suggesting that SIL is not a common factor in HPE pathogenesis in humans.

Abstract. Holoprosencephaly (HPE) is the most common congenital malformation of the brain and face in humans. In this study we report the analysis of SIL (SCL interrupting locus) as a candidate gene for HPE. Fluorescent in situ hybridization (FISH) analysis using a BAC 246e16 confirmed the assignment of SIL to 1p32. Computational analysis of SIL at the protein level revealed a 73 % overall identity between the human and murine proteins. Denaturing high performance liquid chromatography (dHPLC) techniques were used to screen for mutations and these studies identified several common polymorphisms but no disease-associated mutations, suggesting that SIL is not a common factor in HPE pathogenesis in humans.

Copyright © 2002 S. Karger AG, Basel
Holoprosencephaly (HPE) results from an incomplete separation of the primordial single eye field and ventral forebrain into distinct left and right halves. HPE manifests a highly variable phenotype from the severe malformation that leads to cyclopia at one extreme, through a nearly continuous spectrum to simply closely spaced eyes (Muenke and Beachy, 2000;Roessler and Muenke, 2000). The incidence of HPE is 1:10,000-20,000 live births (Roach et al., 1977); however, these estimates do not reflect intrauterine lethality since the true incidence of HPE in utero is 1 in 250 (Matsunaga and Shiota, 1977), making HPE the most common malformation of the brain in humans. The etiology of HPE is incompletely understood, but based on recent studies it is clear that there are many important factors that contribute including human genes and the environment. Mutations in several genes have been identified which can result in HPE in humans, including Sonic Hedgehog (SHH), SIX3, ZIC2, and probably others such as TGIF and PTCH (Roessler and Muenke, 2000;Ming et al., 2002). Thus far, mutations in the human SHH gene are among the most common putative causes of familial HPE, which is consistent with the central role that Shh plays as a ventralizing factor produced by the anterior midline prechordal plate. Consistent with our understanding of the pathogenesis of HPE, all of these genes are active during the early embryological stages of eye field and anterior brain development. Thus genes implicated in developmental programs initiated by Shh in model organisms, such as Sil, themselves become excellent candidates for HPE.
Previous genetic studies have suggested that two transmembrane proteins, namely Patched (Ptc) and Smoothened (Smo) are required for Shh signaling (Chen and Struhl, 1996). When Shh binds to Ptc, Smo accumulates in the membrane in a highly phosphorylated state and activates Shh target genes. Likewise, in the absence of the ligand, Smo is less phosphorylated thereby resulting in inactivation of Shh target genes (Chen and Struhl, 1996;Quirk et al., 1997). Nodal, a putative downstream target of Shh signaling is required for the establishment of the axial midline mesendoderm and later is expressed asymmetrically in the lateral plate mesoderm. Either loss of appropriate Nodal expression on the left side of the lateral plate mesoderm, or its bilateral expression, results in a randomization of heart looping (Meyers and Martin 1999;Tsukui et al., 1999). There is presently no clear conserved mechanism described in vertebrates to actually break symmetry, however, factors such as Nodal are thought help to establish left-right asymmetry; furthermore, as a factor correlated with "leftness" its asymmetric expression ensures that the process is nonrandom and associated with characteristic organ positioning (Meno et al., 1998;Meyers and Martin, 1999;Tsukui et al., 1999). In the murine Shh -/or Smo -/mutants, Nodal is expressed in the vicinity of the node but is not effectively expressed in the lateral plate mesoderm. This absence of left-sided gene expression is correlated with randomization of heart looping in Shh -/mutants and is even more severe in Smo -/mutants where the heart fails to loop in either direction and remains as a straight linear tube. These results suggest an early role of Shh signaling in the control of left-right asymmetry either directly or indirectly through actions impacting on the axial midline (Zhang et al., 2001).
We investigated the human SIL (SCL interrupting locus) gene (Aplan et al., 1991) because the murine Sil -/mutants displayed features compatible with the HPE spectrum including arrest of neural tube closure and lack of midline separation at the anterior end of the cranial folds leading to holoprosencephaly-like defects. In addition, these mutants displayed left-right developmental abnormalities manifested as abnormal cardiac looping. In Sil -/mutant embryos the marker Shh displays discontinuous expression accompanied by the lack of Hfh3 expression in the neural tube fold. Heterozygous Sil +/mice are normal, while homozygous Sil -/mice die in utero at E10.5. Izraeli et. al., also found reduced expression of multiple Shh target genes, despite continued Shh expression in the node and the notochord and suggested a role for Sil in the propagation of Shh signals (Izraeli et al., 1999(Izraeli et al., , 2001. The exact biochemical role of Sil is still unclear due to the absence of homology to any previously known protein. Interestingly, in Ptc -/and Ptc -/double murine mutants the expression of Shh target genes, such as Gli, were markedly reduced suggesting that Sil might be required downstream of Smo in the Shh response pathway (Izraeli et al., 2001).
We hypothesized that a mutation that perturbs the normal function of human SIL might be a risk factor for the development of HPE. Therefore, we decided to study the human SIL gene architecture in detail, to confirm its chromosomal location, and to screen the gene for mutations using denaturing high performance liquid chromatography (dHPLC) in our panel of HPE patients.

Materials and methods
Patients and DNA preparation DNA samples of patients with HPE were collected by informed consent according to the guidelines of the Institutional Review Board of the NHGRI. DNA was isolated from blood or lymhoblastoid cell lines by standard protocols. A total of 83 probands with familial HPE were chosen for dHPLC screening; this panel is known to be representative of the full spectrum of HPE severity. No patients with laterality phenotypes were available for a similar analysis.

Computational analysis of DNA and protein sequences
Unpublished data (I.K.) pertaining to the genomic organization of SIL was further confirmed by nucleotide homology searches in the public database using the BLASTN program (http://www.ncbi.nlm.nih.gov/blast). The compiled data reflecting the gene structure of all 18 exons were then submitted to GenBank. MultiAlign software (http://www.toulouse.inra.fr/multalin.html) was used for protein homology searches between human and murine SIL. Putative structural motifs were analyzed by PFAM (http://www.sanger.ac.uk/Software/Pfam/search.shtml) and ProfileScan (http://www.isrec. isb-sib.ch/cgi-bin/).

Fluorescent in situ hybridization (FISH) analysis
Slides with normal chromosome metaphase spreads were incubated for 1 h at 37°C in 2× SSC (0.3 M NaCl and 0.3 M sodium citrate) and then dehydrated sequentially in 70 %, 80 %, and 90 % ethanol. Chromosomal DNA was denatured in 70 % formamide, 2× SSC for 2 min at 72°C followed by dehydration in ethanol washes of 70 %, 80 %, 90 %, and 100 %. FISH was performed with the BAC probe 246e16 (Incyte Genomics, MO) labeled with digoxigenin-11-dUTP by nick translation (Boehringer Mannheim, IN) and ethanol precipitated in the presence of 50× Cot-1 human DNA (Lichter et al., 1988). The DNA pellet was resuspended to a final concentration of 25 ng/ml. The hybridization and the post-hybridization step were performed as described earlier (Pack et. al., 1999). Slides were counterstained with DAPI, 250 ng/Ìl (Boehringer Mannheim, IN) with Antifade. Hybridization signals were scored using a Zeiss Axiophot epifluorescence microscope (Zeiss, NY) and the two-color image was captured on a Photometric charge-coupled device camera (Photometrics, AZ) using IP Lab image software (Signal Analytics, VA).

Denaturing high performance liquid chromatography (dHPLC) screening
Amplification of genomic DNA was performed in a 30 Ìl reaction volume, using 60-100 ng DNA template, 50 ÌM dNTP, 0.25 ÌM of each primer, 3 Ìl of 10× PCR Amplification buffer (Gibco, MD), 1.5 Ìl 10× Enhancer buffer (Invitrogen, CA), 0.9 Ìl of 50 mM MgSO 4 , and 1 U AmpliTaq (Perkin Elmer, CA). All reactions were performed using a PTC-225 thermocycler (MJ Research, MA). Primer pairs and their respective annealing temperatures are described in Table 1. Typical PCR cycling parameters were 95°C for 4 min followed by 30 cycles at 95°C for 30 s, annealing at the indicated temperatures for 30 s, extension at 72°C for 1 min, culminating in a final step of 72°C for 5 min. One half of the PCR product was used for dHPLC analysis and the other half was retained for direct DNA sequencing.
The dHPLC was performed on a ProStar helix dHPLC system (Varian, CA). In order to enhance heteroduplex formation, the PCR products were denatured at 95°C for 5 min followed by gradual cooling to 60°C over 15 min (1°C/30 sec). The PCR product was automatically injected onto a Helix column and eluted with a linear gradient of buffers A and B (Varian, A: 100 mM triethylamine acetate [pH 7], 0.1 mM EDTA, and B: 100 mM triethylamine acetate [pH 7], 0.1 mM EDTA, 25 % (v/v) acetonitrile) at a constant rate of 0.9 ml/min. Each sample was analyzed at the melting temperature (T m ) determined by using the dHPLCMelt software (http://insertion.stanford.edu/melt.html). In some instances, two different temperatures were used to enhance the sensitivity of screening. Heterozygous profiles were identified by visual inspection of the chromatograms on the basis of the appearance of additional earlier eluting peaks. Corresponding profiles for the same amplicons, which exhibited only a single peak, could be demonstrated to be homozygous and normal in sequence (data not shown).

Direct DNA sequencing
Amplicons displaying heterozygous profiles were purified using a Qiagen PCR purification kit (Qiagen, CA) and bi-directionally sequenced using the BigDye TM terminator cycle sequencing kit according to the manufacturer's protocol (Applied Biosystems, CA). Sequencing reactions were analyzed on an ABI 377 automated sequencer.

Nomenclature
Gene mutation nomenclature used in this study follows the recommendations of den Dunnen and Antonarakis (2001). Gene symbols are those following the recommendations of the HUGO Gene Nomenclature Committee (http://www.gene.ucl.ac.uk).

Results
The BLAST program (NCBI) was used to confirm the gene architecture and establish additional sequence information from the limited initial intron-exon sequence information available (characterized by S.I. and I. K.) facilitating the primer designing process (Table 1). SIL consists of 18 exons wherein the coding region starts from the third exon, with a predicted open reading frame of 3861 nucleotides that encodes 1287 amino acid residues. The nucleotide sequence analysis of the coding region between the human and mouse SIL using MultiAlign software revealed an overall homology of 81 %. The amino acid sequence alignment of the human and the mouse SIL displayed an overall identity of 73 %. The protein pattern and domain prediction software detected the presence of a putative nuclear localization signal motif KKKTH (residue no. 979-983) and the presence of a cysteine-terminal domain (residue no. 601-702) interestingly sharing a weak homology (N-score F6.7) to the carboxyl terminal domain in TGF-ß (Fig. 1). The chromosome location of the SIL gene was determined to be chromosome 1p32 by FISH analysis (Fig. 2). No signal was seen on any other chromosome in the 20 normal metaphase cells examined.
The dHPLC analysis was performed at the dHPLCmelt recommended temperature(s) as shown in Table 1. All 83 samples with homozygous wild type sequences were correctly identified at the recommended temperature for all the exons. The GC content of amplicons spanning exon 2, 14, and 16 were high at one particular region and thus the dHPLC analysis was repeated at a raised temperature to make it more sensitive in detecting heteroduplexes.    Four different heterozygous sequence variations were identified in human SIL (Table. 2); however, none were unique to HPE patients. The first homozygous change in exon 5, 257C → T, predicts an A86V mutation and this sequence change introduces an AccI restriction site. The two intron variations detected were present in exon 7 (IVS 7-21T → A) and exon 13 (IVS 13-53A → T). Examples of heteroduplex detection in exons 7 and 13A are shown in Fig. 3. The sequence variation in exon 7 produced a consistent elution profile but in other cases (not shown) distinct sequence variations with a given amplicon produced different characteristic elution profiles. The fourth heterozygous sequence variation, 2874G → A, predicts a substitution P1015R encoded by exon 17 which introduces an AflIII restriction site. Over 200 normal chromosomes were screened either by restriction fragment length polymorphisim or by dHPLC for each of the mutations detected in SIL. All the nucleotide variations detected in SIL were also found to be present in the normal control samples.

Discussion
Animal models of brain development have been successfully applied to further our understanding the pathogenetic mechanism of HPE in humans. Sil -/mutants display severe anterior midline neural-tube defects, abnormalities in left-right axis development, holoprosencephaly-like features, and are embryonic lethal (Izraeli et al., 1999). Furthermore, discontinuous expression of Shh and lack of Hfh3 in the notochord and reduced expression of multiple Shh target genes suggested that Sil deserved to be examined as an HPE candidate gene. In this study, we investigated the sequence variations of SIL in HPE and found no evidence for its role as a common genetic determinant of this disorder in our panel of samples.
Sil -/murine mutants display a lack of midline separation at the anterior end of the cranial folds resulting in holoprosencephaly-like defects and these embryos ultimately die in utero while the heterozygous mice have no distinct phenotype. Since all of our clinical samples were obtained from live-born infants, it appears likely that our failure to find evidence for a role for SIL in HPE results, in part, from the fact that the phenotype may be autosomal recessive lethal in both mice and humans.
The utilization of a high-throughput approach to mutation screening by dHPLC methodology is extremely sensitive, rapid, and very cost-effective as compared to the traditional singlestrand conformation polymorphism (SSCP). Even though our mutation screening did not suggest loss or gain of function of SIL as a common denominator of HPE in humans, nonetheless, we cannot formally exclude the possibility that rare cases could be caused by this mechanism.
The bilateral expression of Nodal, Leftb, and Pitx2, in the lateral plate mesoderm of Sil -/murine mutants could be attributed to a defective midline barrier, which allows the spread of the left-right signaling cascade across the midline (Izraeli et al., 1999;Bisgrove et al., 2000). Similar observations have been made in zebrafish ntl and flh mutants (Rebagliati et al., 1998) where the generation of the midline barrier and not the generation of left-right signals at the node are affected (Bisgrove et al., 2000). In the zebrafish, cyc (nodal) is critical for the regulation of lft1, lft2, and pitx2, and is required for the establishment of cardiac laterality (Yost, 1999). These findings suggest that the genes involved in establishing or maintaining the anterior mid-line barrier can be associated with the development of brain, heart, and gut asymmetries (Bisgrove et al., 2000).
Targeted mutation of Ebaf (Meno et al., 1998) and Pitx2 (Kitamura et al., 1999) has demonstrated the importance of these genes in left-right axis formation (Casey and Hackett, 2000). Since Ptc -/murine mutants exhibit left-right development abnormalities associated with a lack of Ebaf expression and bilateral expression of Nodal, Leftb, and Pitx2, we propose that SIL might be an excellent candidate gene to study in laterality patients. Furthermore, the defect appears to be related to Shh signaling since Ptch and Gli expression are both reduced in Sil -/embryos. Therefore, the identification of the gene structure and our strategy to analyze the human SIL gene should facilitate the examination of laterality phenotypes.