Down syndrome (DS), a major cause of mental retardation, is characterized by subtle abnormalities of cortical neuroanatomy, neurochemistry and function. Recent work has shown that chromosome band 21q22 is critical for many of the neurological phenotypes of DS. A gene, DSCAM (Down syndrome cell adhesion molecule), has now been isolated from chromosome band 21q22.2-22.3. Homology searches indicate that the putative DSCAM protein is a novel member of the immunoglobulin (Ig) superfamily that represents a new class of neural cell adhesion molecules. The sequence of cDNAs indicates alternative splicing and predicts two protein isoforms, both containing 10 Ig-C2 domains, with nine at the N-terminus and the tenth located between domains 4 and 5 of the following array of six fibronectin III domains, with or without the following transmembrane and intracellular domains. Northern analyses reveals the transcripts of 9.7, 8.5 and 7.6 kb primarily in brain. These transcripts are differentially expressed in substructures of the adult brain. Tissue in situ hybridization analyses of a mouse homolog of the DSCAM gene revealed broad expression within the nervous system at the time of neuronal differentiation in the neural tube, cortex, hippocampus, medulla, spinal cord and most neural crest-derived tissues. Given its location on chromosome 21, its specific expression in the central nervous system and neural crest, and the homologies to molecules involved in neural migration, differentiation, and synaptic function, we propose that DSCAM is involved in neural differentiation and contributes to the central and peripheral nervous system defects in DS.
Down syndrome (DS) is the most common cause of genetic mental retardation (1), and is associated with characteristic facial and physical features, congenital heart and gut disease, defects of the eye, of the immune and endocrine systems, and an increased risk of leukemia and Alzheimer's disease. While most traits are of variable penetrance, mental retardation and neonatal hypotonia are seen in most DS patients (2). DS is generally caused by trisomy for human chromosome 21 (3,4), and the phenotype of DS is due ultimately to an increase in the expression of some of the genes on this chromosome. The phenotypes of DS have been correlated with particular regions of chromosome 21, providing significant evidence of the contribution of genes on chromosome band 21q22 distal to D21S267 to DS phenotypes including mental retardation, microcephaly and congenital heart disease as well as some of the physical features (5-7).
Although the syndrome is associated with prenatal defects of cortical lamination, and a vertical superior temporal gyrus, further gross brain pathology emerges during early postnatal development (8). The dendrites become relatively atrophic 4 months after birth, concurrent with a progressive microcephaly (9). Adults with DS versus normal controls, have substantially smaller cerebral and cerebellar hemispheres, ventral pons, mammillary bodies and hippocampal formations, but a larger parahippocampal gyrus (10). The histopathological abnormalities seen in adult DS brain include decreased cell density and number in cortical layers 2 and 4.
The congenital gut diseases of DS include Hirschsprung disease (HSCR) (11,12), a congenital disorder characterized by absence of enteric ganglia along a variable length of the intestine. The pathology of HSCR is thought to be due to failure of neural crest cell migration, differentiation or colonization, during embryonic weeks 6-12. Non-DS HSCR is caused by mutations in the RET oncogene (13), the endothelin receptor type B gene (14) and the endothelin 3 gene (15,16). It is interesting in relation to DS that preliminary evidence for a genetic modifier of HSCR on chromosome 21q22 has been described in one kindred (17).
Taken together, the neural and gut features of DS have implicated the processes of migration, proliferation and differentiation as being in part responsible for the anomalies, suggesting overexpression of genes encoding cell adhesion molecules located in band 21q22 as potential candidates.
Here we report the isolation and characterization of a novel gene, DSCAM (Down syndrome cell adhesion molecule), a member of the immunoglobulin superfamily that represents a new class of neural cell adhesion molecules. It is expressed largely in the developing nervous system, is located in chromosome band 21q22.2-q22.3 and may be responsible for a part of the mental retardation and visceral anomalies of DS.
In order to isolate genes that are responsible for the phenotypes of DS, cDNAs have been isolated by using direct cDNA selection and exon trapping techniques (see Materials and Methods) and BAC (bacterial artificial chromosome) and PAC (P1-derived artificial chromosome) contigs on 21q22.2-q22.3. These contigs cover a minimum of 3.5 Mb of the 4-5 Mb interval extending from D21S55 to MX1 (18). A modified direct cDNA selection technique (19,20) was applied to BAC-423A5, BAC-430F1, BAC-628H2, BAC-371H8 and PAC-31P10 (Fig. 1) by using cDNA from trisomy 21 human fetal brain, and the selected fragments were then subcloned into a plasmid vector. The results of Southern hybridization of 400 selected cDNA clones, probed with the BAC DNAs, showed that 320 out of 400 clones (80%) were positive for the BAC and PAC DNAs. DNA sequencing of one of those clones (E64) showed homologies to known genes including the human titin gene (53.5% identity in 228 bp overlap) in the region of the fibronectin type III domains. To isolate larger clones, E64 was used as a probe to screen the trisomy 21 fetal brain cDNA library (19). We identified 62 clones out of 2y106 clones in the original library. Eighteen of these positive phage clones were converted to plasmids, and their DNAs were isolated. These cDNAs were named as DSCAM. The length of the inserts of these clones ranged from 2.4 to 6.6 kb. Exon trapping (21,22) was also used to isolate cDNAs in the BAC and PAC contig. With this approach, three exons identified from BAC-539E7 and one from BAC-430F1 were found to identify the same sequences as those isolated by cDNA selection (data not shown).
Figure
To confirm that the DSCAM cDNA originated from the BACs and PACs in the DS region and to determine the genomic size of DSCAM, the longest DSCAM cDNA clones (CHD2-42; 6.1 kb, CHD2-18; 6.5 kb, CHD2-52; 6.6 kb) were hybridized to Southern blots containing the BAC and PAC clone contig (Fig. 1). CHD2-42, 18 and 52 hybridized to BACs 423A5, 430F1, 628H2, 539E7, 371H8, 825E1, 593D1, 261F12, 30E4, 385B7 and 388F4, and to PACs 31P10, 58D10. BACs 816F6, 116E8, 720G4 and 619H8 were only positive for CHD2-18 and 52 but negative for CHD2-42. All other BACs shown in Figure 1 were negative. These results indicate that the DSCAM gene spans 900-1200 kb genomic DNA and covers a gap in this BAC and PAC contig, indicated by an arrowhead, as well as in the available YAC contigs (23,24). DSCAM cDNA sequences were also confirmed to originate from these BACs and PACs by direct sequencing of the BACs and PACs as templates using cDNA sequence-specific primers.
The map position of DSCAM on chromosome 21q22.2-22.3 was confirmed by using CHD2-42 as a probe for fluorescence in situ hybridization (Fig. 2). Two independent experiments were performed, and >100 metaphase cells were evaluated. Signals were clearly seen on two chromatids of at least one chromosome in 85% of cells. There were no other double signal sites seen in >1% of cells.
Figure
Sequence analysis of clone CHD2-42 revealed a 6110 bp DNA sequence (GenBank accession no. AF023449) which contained a large open reading frame (ORF) (5687 bp) as well as a 34-untranslated region (UTR) sequence (423 bp), but the 54-UTR and start codon were not found. In order better to define the 54 end, two further clones, CHD2-18 of 6.5 kb and CHD2-52 of 6.6 kb (GenBank accession no.AF023450) were characterized. Sequence analyses of these clones close to the 54 end overlap with sequence at the 54 end of CHD2-42. However, CHD2-18 extends 416 bp farther 54, and CHD2-52 extends 494 bp farther 54 than CHD2-42. The extra 494 bp sequence extends the ORF by 43 bp at the 54 end and contains a start codon. Two stop codons occur 330 and 427 bp upstream of the start codon. The 494 bp of additional 54 sequence found in CHD2-52 combined with CHD2-42 (6604 bp) yield a consensus cDNA containing an ORF of 5730 bp (1910 amino acids; mol. wt 211 407 Da) flanked by 451 bp of 54-UTR and 423 bp of 34-UTR. The presumptive 54-UTR is highly GC rich (81% GC over 451 bp) and contains 13 MspI sites, as well as 72 CG and 93 GC dinucleotide pairs. A homology search of the predicted amino acid sequence of this ORF to genes registered in the GenBank and the EMBL database was conducted by using the BLAST-P program (25). The putative gene product revealed homologies to multiple proteins including CAM-L1 (26), BIG-1 [brain-derived immunoglobulin (Ig) superfamily molecule-1] (27), BIG-2 (28) and DCC (deleted in colon cancer) (29), and revealed DSCAM as defining a novel class of the Ig superfamily. The closest member of the Ig superfamily, BIG-2, showed 40% nucleotide identity, 27% amino acid identity and 47% amino acid similarity with DSCAM.
Homology searches with sequences of Ig type C2 domains and fibronectin type III domains of the most highly related Ig superfamily members (CAM-L1, DCC and axonin-1) were conducted by using the FASTA program (30) and revealed the following structure. The putative DSCAM protein contains an extracellular component at the N-terminus consisting of nine tandemly repeated Ig-like C2-type domains and a tenth Ig-C2 domain located between domains 4 and 5 of the following array of six repeated fibronectin type III domains (Fig. 3). Each Ig-like C2 domain consists of ~100 amino acids, with a pair of conserved cysteines separated by 49-56 residues. A single transmembrane domain of 22 amino acids was defined by using the TMBASE program (31). The remaining 294 amino acids at the C-terminus corresponding to the cytoplasmic domain have partial homologies to the mouse M-phase inducer phosphatase 2 (32) in two regions, one with 34% identity and 52% similarity over 46 bp and a second with 38% identity and 52% similarity over 21 bp, the homolog of the Drosophila glass gene product (33) with 30% identity and 52% similarity over 42 bp, and the mouse delta opioid receptor (34) with 43% identity and 60% similarity over 30 bp. The putative protein contains 16 potential N-glycosylation sites.
Figure
Two DSCAM cDNA clones (CHD2-18 and CHD2-52), however, contain identical deletions of 191 bp that occur in neighboring exons and delete base pairs 4639-4829 of the CHD2-42 sequence (Fig. 4). The resulting transcript is deleted for the entire transmembrane domain that is encoded by the more 34 of these exons. Further, the deletion changes the reading frame and creates a stop codon 36 bp downstream of the deletion. The distal border of the resulting deletion contains the canonical AG of the RNA splicing consensus acceptor site. The proximal border contains a potential variant of the donor splice site consensus sequence (35). To investigate whether the deleted cDNA clones reflected bona fide expressed transcripts, cDNAs from 13 independent fetal and adult sources were analyzed by PCR using primer pairs that flanked the deleted region. These were designed to generate products of different sizes from the two transcripts; 536 bp from the non-deleted CHD2-42 transcript and 345 bp from the deleted CHD2-18 and CHD2-52 transcripts. The analyses included adult samples from amygdala (24 years), skeletal muscle (36 years) and three independent lymphoblastoid cell lines. Fetal samples included whole brain of a trisomy 21 fetus (14 weeks), four from whole brain (4.5-13 weeks), one from temporal lobe (28 weeks) and two from heart (4.5 and 13 weeks). The results indicated that all fetal and adult samples produced two bands of the predicted sizes, supporting the expression of both the non-deleted and deleted transcripts (data not shown). A schematic comparison of CHD2-18, CHD2-42 and CHD2-52 is described in Figure 5.
Figure
Figure
Northern blot analysis of human fetal tissues using an insert of CHD2-42 reveals 8.5 and 7.6 kb transcripts expressed in the fetal brain and not expressed in fetal lung, liver or kidney (Fig. 6A). In adult human tissues, three transcripts of 9.7, 8.5 and 7.6 kb are seen mainly in the brain. Placenta shows faint bands, and the sizes are similar to those in brain. In skeletal muscle, a faint smaller band (6.5 kb) is detected (Fig. 6B). In the multiple parts of adult brain, transcripts of 9.7, 8.5 and 7.6 kb appear to be differentially expressed. The 9.7 kb transcript appears highly expressed in the substantia nigra, moderately expressed in amygdala and hippocampus, and less expressed in the whole brain. (Fig. 6C). A similar pattern is obtained by using a PCR product which spans the 191 bp deletion found in CHD2-18 and 52 (data not shown).
Figure
To investigate the expression of DSCAM further, we isolated mouse homologs of DSCAM using a 3.5 kb human cDNA clone (CHD2-36) for the 34 end of CHD2-42 as a probe to screen a mouse brain cDNA library (see Materials and Methods). Forty positive clones out of 2y106 clones were identified. A partial sequence analysis of clone (CHD2M-14; 3.0 kb) reveals 86% homology to CHD2-42 (base pairs 3051-3356) at the 54 end of CHD2M-14 and 89% homology in 372 bp at the 34 end (base pairs 5739-6110 of CHD2-42). Clone CHD2M-21 (6 kb) revealed 93% in 650 bp at the 54 end (base pairs 126-775 of CHD2-42), 87% over 799 bp (base pairs 2340-3138) and 87% over the terminal 1585 bp (base pairs 4526-6110) terminating at the poly(A) tail (data not shown). CHD2M-14 is identical to the 34 end of CHD2M-21. A schematic comparison of human and mouse cDNA clones is shown in Figure 5.
Tissue in situ hybridization analysis was performed by using the mouse cDNA CHD2M-14 as a probe on sections of normal mouse embryos from days 8.5-17.5 post-coitum (p.c.) as well as from newborn, 2-week-old and adult brains. There is no detectable expression of dscam at 8.5 days p.c. Expression was detected in placenta for human DSCAM (data not shown). At 9.5 days p.c., expression can be seen in the neuroepithelium (Fig. 7A). There appear to be low levels of expression within the branchial arches, suggestive of migrating neural crest cells (Fig. 7B). At 10.5 days p.c., the trigeminal ganglia (neural crest derived) are beginning to express the transcript (data not shown), and expression within the branchial arches is more evident. Expression at 11.5 days p.c. is comparatively abundant throughout the brain. The transcript is found within the regions of the nervous system that differentiate earliest during development (Fig. 7C and E) (36). In the brain, this includes the ventral-most regions, such as the thalamus and medulla. Some expression is also seen within the olfactory epithelium (Fig. 7C). Expression within the neural tube begins in two areas. The first is ventrolateral, corresponding to the areas in which the motor neurons differentiate. The second is within the lateral gray columns that later in development form commissural neurons (37) (see Fig. 6F at 12.5 days p.c.). The dorsal root ganglia (neural crest derived) express the transcript at 11.5 days p.c. (Fig. 7C and E). The trigeminal ganglia show higher levels at 11.5 days p.c. than they did at 10.5 days (Fig. 7E). Putatively migrating neural crest can also be seen within the maxilla, the mandibular arch and in the developing gut. Interestingly, signal is also apparent within the mesenchyme surrounding the umbilical vein and artery (Fig. 7C). At 12.5 days p.c. (Fig. 7G), expression is more extensive than at 11.5 days p.c. More of the nervous system exhibits expression of the transcript, including a larger portion of midbrain, the pontine areas, the basal ganglia and the outermost layer of cortex. Neurons in this layer have undergone mitosis in the subependymal layer of the cortex and migrated into the mantle layer of the cerebral cortex as differentiated cells (38). At 13.5 days p.c., expression is seen throughout most of the brain (Fig. 7H). The outermost layer of the gut also appears to be expressing at this stage (Fig. 7I); these cells are neural crest derived and will form the myenteric ganglia. At 15.5 and 16.5 days p.c., most if not all of the neural crest-derived neural structures have some expression. For example, the regions of the snout that will develop into the sensory structures at the base of the vibrissae (Fig. 7L), the pancreatic ganglia (data not shown), the heart ganglion (Fig. 7J), the enteric nervous system and the sympathetic trunk (data not shown) all express the transcript to some extent. Interestingly, there is no expression within the umbilicus at this stage. Also at this stage, two non-neuronal structures express this gene, the gonad (Fig. 7K) and the annulus fibrosus of the intervertebral disk (data not shown). The olfactory bulb exhibits signal both in the granule cells and within the tufted mitral cells (data not shown). Within the newborn brain, the transcript is expressed most extensively within the differentiating regions such as the septal area, olfactory bulb, inferior colliculus and hippocampus (data not shown). In the adult brain, the gene continues to be expressed in many areas including amygdala, cortex, hippocampus and thalamus (data not shown). In the adult cerebellum, the transcripts are detected in the Purkinje cell layer and in the deep cerebellar nuclei (Fig. 7M).
Figure
The DSCAM gene and its putative protein product define a novel subclass of the Ig superfamily, with highest homologies to BIG-1 (27), BIG-2 (28), L1CAM (26), DCC (29), neogenin (39) and contactin (40). The structure of the predicted DSCAM protein is unique within the neural Ig superfamily (Figs 3 and 8). This derives both from the distinctive number of Ig type C2 and fibronectin III domains (10 and 6 respectively) and from the interruption of the fourth and fifth fibronectin domains by the 10th C2 domain. There is substantial evidence that the neural Ig superfamily members play critical roles in neural development and function and they have been implicated in cell migration and sorting, axon guidance and fasciculation, the formation of neural connections and, most recently, in synaptic plasticity (41
Figure
Alternative splicing products of CAMs have distinct roles in different parts of the brain, as has been demonstrated for closely related Ig superfamily members such as NCAM (52,53) and L1CAM (54). For DSCAM, this is suggested by the differential expression of transcripts seen in various parts of the human adult brain (Fig. 6C). Further, in DSCAM, some clones appear to contain a small deletion which includes the transmembrane domain (Fig. 4) and results in a stop codon 36 bp downstream. The results of RT-PCR indicated that all RNAs tested from various human tissues contain the deleted and non-deleted transcripts. The proximal and distal borders of the deletion are located within neighboring exons and reveal variant consensus splice site sequences (35) with further surrounding imperfect homology to the U1 spliceosome RNA. These results suggest that the deleted transcripts may be products of variant alternative splicing, and these generate a putative transmembrane adhesion molecule and a molecule without a transmembrane domain that is transported to the extracellular matrix. This mode of generating extracellular and membrane-bound forms of CAMs is in contrast to the glycosylphosphatidylinositol (GPI) linkage used by most CAMs (the consensus for which is not yet found in DSCAM) and would provide a way of generating longer range homophilic interactions between cells and the extracellular matrix, of possible significance in cell migration.
The relative specificity of DSCAM expression for the central nervous system and the timing of its expression to the period of neurite outgrowth in both the central and peripheral nervous systems support a role for this gene in early development and differentiation (Figs 6 and 7). This is seen from early in development when, with the exception of neural crest precursors, expression is clearly absent from regions that contain dividing neuroepithelial precursors such as the ependymal layer of the neural tube and the ventricular zone of the brain(36). At least in the embryo, differentiated neurons express DSCAM when they have finished migrating to their proper positions within the neuroepithelium, during neurite outgrowth. Neural crest cells may express DSCAM while they are migrating. This is inferred from the observation that at 15.5 and 16.5 days p.c., most of the neural crest-derived tissues have some expression, although not all have finished migration (Fig. 7J, K and L). This may be related to the differentiation of some neural crest sub-lineages during migration, for example the sympathoadreanal. The continued expression of DSCAM in the myenteric plexus after 15.5-16.5 days p.c. may be due to the portion of neural crest cells that have stopped dividing, although others are still in the cell cycle. Approximately 50% of myenteric ganglia neurons arise after birth, and DSCAM may be expressed later in this subset. At later stages, DSCAM is down-regulated in the neural crest derivatives such as the myenteric ganglia and ganglia of the pancreas. It will be necessary to study younger postnatal gut to determine the latest stage at which DSCAM can be detected. The observation of DSCAM expression in tissues that are derived from the neural crest is of interest with respect to the high level seen in the umbilical cord (Fig. 7C). This suggests that, although previously unsuspected, the tissue surrounding the umbilical artery and vein may also be derived from the neural crest and may function to aid in coordinating the cardiovascular changes occurring at birth. It will be of interest to determine if the expression seen in the fetal liver and branchial arches is also derived from neural crest related to the ductus venosus and ultimately the ductus arteriosus and cardiac outflow tracts respectively.
The expression patterns of the neural Ig superfamily members are variable and some of them are overlapping or complementary. These are thought to contribute to the formation and maintenance of specific neuronal networks in the brain. For example, the expression of neogenin is not restricted to the brain, but is also expressed in kidney, pancreas and skeletal muscle at the high levels (39). In contrast, the expression of BIG-1 (27), BIG-2 (28), PANG (55) and contactin (38) is largely restricted in subsets of central and peripheral neurons. The expression of BIG-1 and BIG-2 in embryonic brains is relatively low and the levels increase after birth. The expression level of F3 is fairly constant in both embryonic and adult brains (28). In the cereberal neocortex, BIG-1 is expressed in layer II, BIG-2 in layers II-V, whereas TAG-1 is found in the white matter (28). DSCAM is expressed in all layers of differentiated neurons in the cortex, as well as other regions including olfactory bulb and thalamus. The specific expression patterns of DSCAM in the brain suggests important roles in the formation of neural networks.
Similarly to DSCAM, L1CAM (26) is expressed mainly in the brain at early developmental stages. Mutations in L1CAM have well established roles in human disease. These result in X-linked hydrocephalus (56), type 1 X-linked spastic paraplegia and the MASA syndrome which includes mental retardation, aphasia, shuffling gait, adducted thumb and agenesis of the corpus callosum (57). The perturbation of development by the mutations of L1 also supports a possible role for the aneuploid expression of DSCAM in the causation of developmental and neurological abnormalities.
The DSCAM gene was isolated by using the BAC contig on 21q22.2-q22.3 covering the region between D21S55 and MX1 (Fig. 1). This gene spans a minimum of 900 kb, estimated by summing the size of BACs and PACs that are non-overlapping and covered by the DSCAM gene (Fig. 1). The DSCAM gene also covers a gap in all physical maps of this region and may provide clues to the basis of this gap. The location of DSCAM suggests a possible role in the DS phenotype, as supported by the studies of patients with partial trisomy 21. Early studies suggested that a subset of the DS features including the typical facial appearance and mental retardation might be produced by duplication of band 21q22 only (58). Other studies mapped those features and congenital heart disease to the region 21q22.2-q22.3 and between D21S267 and MX1/MX2 (5,6), a region of ~4 Mb that contains DSCAM. The Ts65Dn mouse model of DS contains the region of MMU16 (Pgk1-ps1 to MX1/2) that includes DSCAM and reveals some of the neurobehavioral features of DS (59,60). This model will provide an opportunity to investigate the role of DSCAM in normal and DS development.
Of particular interest is the association of chromosome 21 with HSCR. Close to 6% of DS individuals have HSCR (61), and >10% of all HSCR is associated with DS (62). Further, even in non-DS HSCR, a modifier region of HSCR on chromosome 21q22 (D21S259-D21S156) reported (17), and the DSCAM gene maps within this small region. The expression of DSCAM in the neural crest-derived enteric plexus of the gut was shown by mouse tissue in situ hybridization (Fig. 6E and I). Taken together with the expected function of the DSCAM protein and the association of this region of chromosome 21 with HSCR, it is suggested that DSCAM is responsible both for the chromosome 21 association seen in non-DS HSCR and for the HSCR seen in DS.
Why and how might three copies of the DSCAM gene lead to mental retardation and HSCR? First, not only do different neural Ig superfamily members share similar competitive binding sites, but a subset of the functional effects of Ig superfamily members is highly dosage sensitive. For example, aggregation analyses have revealed that a 2-fold increase of NCAM and Ng-CAM resulted in a >30-fold increase both in binding rates and in the rate of cell aggregation (63-65). Three copies of DSCAM may also alter the adhesive properties of neural cells, and lead to the abnormalities of lamination and gross architecture seen in DS. In addition, studies in Aplysia californica and Drosophila melanogaster indicate that the increased production of neural proteins of the Ig superfamily inhibits synapse growth or migration under some conditions (44,66), and these suggest that the overexpression of DSCAM may also affect synapse growth or migration and lead to synaptic abnormalities seen in DS.
These questions will be investigated by the analysis of the DSCAM gene in both cellular and organismal models. Moreover, it will be important to analyze nervous system structure and function with respect to DSCAM expression in the trisomy 16 mouse models of DS such as Ts65Dn (59). Understanding the possible role of DSCAM in the pathways that lead to the abnormalities of cortical lamination, of the dendritic tree and of the enteric neural crest derivatives in DS may provide critical insights into the origin of neurocognitive defects in DS and the development and function of the human nervous system.
Construction of BAC contigs in the DS region is described elsewhere (18). Briefly, BAC libraries (67,68) were screened with DNA of YACs 750F, 68E7, 75D12, 152F7, 152G11, 179B7 and 761B5 (23,69). A PAC library (70) was also screened with identified BAC clone DNAs. The location and the order of each clone was determined by clone-to-clone Southern hybridization, fluorescence in situ hybridization and sequence-tagged site (STS) analysis in which primers of 59 STSs were used. These were obtained either through the Research Genetics Human Chromosome 21 Panel, from the Genome Data Base (GDB) or generated from end-clone sequences of BACs and PACs.
The method for direct cDNA selection was described in detail elsewhere (19,20). Briefly, RNA was isolated from a 14 week trisomy 21 fetal brain, and cDNA was synthesized. Sau3AI linkers were attached to the cDNA followed by digestion with Sau3AI. Adaptors were attached to the digested cDNA and the products were amplified by PCR using one strand of the adaptor as a primer. On the other hand, BAC and PAC DNAs from the DS region were biotinylated by nick translation and biotin-16-dUTP. Denatured PCR-amplified cDNA were pre-annealed with Cot-1 DNA and line-1 DNA. After pre-hybridization, biotinylated and heat-denatured BAC and PAC DNA was added to the pre-hybridized cDNA mixture and incubated. The cDNA-genomic DNA hybrids were captured on magnetic beads. The beads were washed under stringent conditions, and the cDNAs were eluted. The eluted cDNAs were amplified by PCR and subcloned into a plasmid vector (BlueScript KS+; Stratagene). Insert DNAs were isolated from the subclones, and were analyzed by Southern hybridization and DNA sequencing. Exon trapping was performed by using a commercially available kit, the Exon Trapping System (Gibco BRL).
Screening of the trisomy 21 fetal brain cDNA library (19) was performed by using the cDNA fragments that were isolated by the direct cDNA selection. Phages were plated to an average density of 1y105 per 175 cm2 plate. Plaque lifts of 20 plates (2y106 phages) were made using duplicate nylon membranes (Hybond-N+; Amersham). Hybridized membranes were washed to a final stringency of 0.2y SSC, 0.1% SDS at 65_C. Filters were exposed to X-ray film overnight. Positive phages were transformed into plasmids by M13-mediated excision according to the manufacturer's protocol of the cDNA library construction using Uni-ZAP XR vector (Stratagene). The sizes of inserts were measured by digesting with EcoRI and XhoI followed by separation on 0.8% agarose gels.
The insert of one of the in vivo excised clones, CHD2-42, was cut out from the Bluescript vector by digesting with XhoI and EcoRI. After labeling by the random priming method (RadPrime Labeling System; Gibco BRL) followed by purification using G-50 Sephadex columns (Quick Spin Columns; Boehringer Mannheim), the fragments were used as probes for northern hybridization using a Multiple Tissue Northern Blot (Clontech). northern hybridization was performed by following the manufacturer's instructions. Hybridized membrane was washed at a final stringency of 0.1y SSC and 0.1% SDS at 50_C. The filter was exposed to X-ray film (Kodak X-OMAT AR) at -70_C for 1-5 days.
The nucleotide sequences of both the coding and non-coding strand were determined in their entirety by the dideoxy chain termination method using the Sequenase DNA sequencing kit (USB) and T3 and T7 sequencing primers as well as custom-made primers.
The DNA of CHD2-42 was labeled with biotin-14-dATP (Gibco, BRL) by using the nick translation method and was hybridized to metaphase chromosomes prepared from normal male peripheral blood lymphocytes by the bromodeoxyuridine synchronization method. In situ hybridization was performed according to the method described by Korenberg and Chen (71).
For isolation of the mouse DSCAM homolog, a mouse brain cDNA library Uni-ZAP XR vector (Stratagene; cat. #937314) was used. This library was made from C57 Black/6 mice, adult females (19 weeks old). The cDNAs were oligo(dT) primed and cloned unidirectionally into the EcoRI and XhoI sites of the vector. The average insert size is 1.0 kb. The method for screening is described above.
The protocol which was used to fix and embed BALB/c and C57BL/6yDBA/2 embryos, fetuses and postnatal brains is described in detail in Lyons et al. (72). Briefly, embryos were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) overnight, dehydrated and infiltrated with paraffin. The 5-7 mm serial sections were mounted on gelatinized slides. Two sections were mounted per slide, deparaffinized in xylene, rehydrated and post-fixed. The sections were digested with proteinase K, post-fixed, treated with tri-ethanolamine/acetic anhydride, washed and dehydrated. cRNA probes were prepared from CHD2M-14. The plasmid was linearized with XbaI, and T7 polymerase was used to generate the antisense cRNA. The plasmid was linearized with KpnI, and T3 polymerase was used to generate the sense control cRNA. The cRNA transcripts were synthesized according to the manufacturer's conditions (Stratagene) and labeled with [35S]UTP (>1000 Ci/mmol; Amersham). cRNA transcripts >100 nucleotides were subjected to alkali hydrolysis to give a mean size of 70 bases for efficient hybridization. Sections were hybridized overnight at 52_C in 50% deionized formamide, 0.3 M NaCl, 20 mM Tris-HCl pH 7.4, 5 mM EDTA, 10 mM NaPO4, 10% dextran sulfate, 1y Denhardt's, 50 mg/ml total yeast RNA and 50 000-75 000 c.p.m./ml 35S-labeled cRNA probe. The tissue was subjected to stringent washing at 65_C in 50% formamide, 2y SSC, 10 mM dithiothreitol and washed in PBS before treatment with 20 mg/ml RNase A at 37_C for 30 min. Following washes in 2y SSC and 0.1y SSC for 10 min at 37_C, the slides were dehydrated and dipped in Kodak NTB-2 nuclear track emulsion and exposed for 2-3 weeks in light-tight boxes with desiccant at 4_C. Photographic development was carried out in Kodak D-19. Slides were counterstained lightly with toluidine blue and analyzed using both light- and darkfield optics of a Zeiss Axiophot microscope. Sense control cRNA probes (identical to the mRNAs) always gave background levels of the hybridization signal (Fig. 5D). Embryonic structures were identified with the help of the following atlases: Rugh (73), Kaufman (74) and Altman and Bayer (36).
This research was supported by grants from the National Institute of Health (NICHD #PO1 HD17449 and NHBLI #RO1 HL50025 to J.R.K. and NICHD #29471 to G.E.L.), and the Department of Energy (#DE-FG03-92ER61402) to J.R.K. J.R.K. holds the Geri & Richard Brawerman Chair in Molecular Genetics. M.A.H. was supported by NIH GMO 7507-18. G.E.L.was supported by a grant from the American Heart Association Wisconsin Affiliate.
Human Molecular Genetics Pages
DSCAM: a novel member of the immunoglobulin superfamily maps in a Down syndrome region and is involved in the development of the nervous system
Introduction
Results
Isolation of the DSCAM cDNA
Characterization of the DSCAM cDNA
Northern analysis of human DSCAM
Isolation of mouse DSCAM (dscam) cDNA
Tissue in situ hybridization analysis of dscam cDNA on mouse tissues
Discussion
Materials And Methods
Isolation of cDNAs from BAC contigs by direct cDNA selection and exon trapping
Northern blot analysis
DNA sequencing of DSCAM
Fluorescence in situ hybridization
Isolation of mouse DSCAM cDNA (dscam)
Tissue in situ hybridization
Acknowledgements
References
DSCAM: a novel member of the immunoglobulin superfamily maps in a Down syndrome region and is involved in the development of the nervous system
INTRODUCTION
RESULTS
Isolation of the DSCAM cDNA
Characterization of the DSCAM cDNA
Northern analysis of human DSCAM
Isolation of mouse DSCAM (dscam) cDNA
Tissue in situ hybridization analysis of dscam cDNA on mouse tissues
DISCUSSION
MATERIALS AND METHODS
Isolation of cDNAs from BAC contigs by direct cDNA selection and exon trapping
Northern blot analysis
DNA sequencing of DSCAM
Fluorescence in situ hybridization
Isolation of mouse DSCAM cDNA (dscam)
Tissue in situ hybridization
ACKNOWLEDGEMENTS
REFERENCES
These authors contributed equally to this work
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 6 Jan 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. O. Korbel, T. Tirosh-Wagner, A. E. Urban, X.-N. Chen, M. Kasowski, L. Dai, F. Grubert, C. Erdman, M. C. Gao, K. Lange, et al. The genetic architecture of Down syndrome phenotypes revealed by high-resolution analysis of human segmental trisomies PNAS, July 21, 2009; 106(29): 12031 - 12036. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Corty, B. J. Matthews, and W. B. Grueber Molecules and mechanisms of dendrite development in Drosophila Development, April 1, 2009; 136(7): 1049 - 1061. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Amano, M. Fujii, S. Arata, T. Tojima, M. Ogawa, N. Morita, A. Shimohata, T. Furuichi, S. Itohara, H. Kamiguchi, et al. DSCAM Deficiency Causes Loss of Pre-Inspiratory Neuron Synchroneity and Perinatal Death J. Neurosci., March 4, 2009; 29(9): 2984 - 2996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Liu, W. Li, L. Wang, A. Kar, K.-L. Guan, Y. Rao, and J. Y. Wu DSCAM functions as a netrin receptor in commissural axon pathfinding PNAS, February 24, 2009; 106(8): 2951 - 2956. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Schmucker and B. Chen Dscam and DSCAM: complex genes in simple animals, complex animals yet simple genes Genes & Dev., January 15, 2009; 23(2): 147 - 156. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L. Andrews, S. Tanglao, W. T. Farmer, S. Morin, S. Brotman, M. A. Berberoglu, H. Price, G. C. Fernandez, G. S. Mastick, F. Charron, et al. Dscam guides embryonic axons by Netrin-dependent and -independent functions Development, December 1, 2008; 135(23): 3839 - 3848. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Brites, S. McTaggart, K. Morris, J. Anderson, K. Thomas, I. Colson, T. Fabbro, T. J. Little, D. Ebert, and L. Du Pasquier The Dscam Homologue of the Crustacean Daphnia Is Diversified by Alternative Splicing Like in Insects Mol. Biol. Evol., July 1, 2008; 25(7): 1429 - 1439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. R. Uhl, Q.-R. Liu, T. Drgon, C. Johnson, D. Walther, J. E. Rose, S. P. David, R. Niaura, and C. Lerman Molecular Genetics of Successful Smoking Cessation: Convergent Genome-Wide Association Study Results Arch Gen Psychiatry, June 1, 2008; 65(6): 683 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. R. Uhl, T. Drgon, Q.-R. Liu, C. Johnson, D. Walther, T. Komiyama, M. Harano, Y. Sekine, T. Inada, N. Ozaki, et al. Genome-Wide Association for Methamphetamine Dependence: Convergent Results From 2 Samples Arch Gen Psychiatry, March 1, 2008; 65(3): 345 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Fusaoka, T. Inoue, K. Mineta, K. Agata, and K. Takeuchi Structure and function of primitive immunoglobulin superfamily neural cell adhesion molecules: a lesson from studies on planarian Genes Cells, May 1, 2006; 11(5): 541 - 555. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. O'Doherty, S. Ruf, C. Mulligan, V. Hildreth, M. L. Errington, S. Cooke, A. Sesay, S. Modino, L. Vanes, D. Hernandez, et al. An Aneuploid Mouse Strain Carrying Human Chromosome 21 with Down Syndrome Phenotypes Science, September 23, 2005; 309(5743): 2033 - 2037. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R. GRAVELEY, A. KAUR, D. GUNNING, S. L. ZIPURSKY, L. ROWEN, and J. C. CLEMENS The organization and evolution of the Dipteran and Hymenopteran Down syndrome cell adhesion molecule (Dscam) genes RNA, October 20, 2004; 10(10): 1499 - 1506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Li and K.-L. Guan The Down Syndrome Cell Adhesion Molecule (DSCAM) Interacts with and Activates Pak J. Biol. Chem., July 30, 2004; 279(31): 32824 - 32831. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Kahlem, M. Sultan, R. Herwig, M. Steinfath, D. Balzereit, B. Eppens, N. G. Saran, M. T. Pletcher, S. T. South, G. Stetten, et al. Transcript Level Alterations Reflect Gene Dosage Effects Across Multiple Tissues in a Mouse Model of Down Syndrome Genome Res., July 1, 2004; 14(7): 1258 - 1267. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Hedges, P. A. Callinan, R. Cordaux, J. Xing, E. Barnes, and M. A. Batzer Differential Alu Mobilization and Polymorphism Among the Human and Chimpanzee Lineages Genome Res., June 1, 2004; 14(6): 1068 - 1075. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Celotto and B. R. Graveley Alternative Splicing of the Drosophila Dscam Pre-mRNA Is Both Temporally and Spatially Regulated Genetics, October 1, 2001; 159(2): 599 - 608. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hildmann, X. Kong, J. O'Brien, L. Riesselman, H.-M. Christensen, E. Dagand, H. Lehrach, and M.-L. Yaspo A Contiguous 3-Mb Sequence-Ready Map in the S3-MX Region on 21q22.2 Based on High- Throughput Nonisotopic Library Screenings Genome Res., April 1, 1999; 9(4): 360 - 372. [Abstract] [Full Text]
|
|
|
|
|
|
D. E. Cabin, J. W. McKee-Johnson, L. E. Matesic, T. Wiltshire, E. E. Rue, A. E. Mjaatvedt, Y. K. Huo, J. R. Korenberg, and R. H. Reeves Physical and Comparative Mapping of Distal Mouse Chromosome 16 Genome Res., September 1, 1998; 8(9): 940 - 950. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||