Inventors: Palleja; Xavier Estivill (Provenca 132, 08029 Barcelona, ES); Fuentes; Juan Jose (Barcelona, ES); Pritchard; Melanie (Barcelona, ES) Assignee: Palleja; Xavier Estivill (Barcelona, ES) Appl. No.: 665040 Filed: June 7, 1996

Primary Examiner: Arthur; Lisa B. Attorney, Agent or Firm: Ladas & Parry United States Patent 5,869,318 February 9, 1999

Claims
We claim:

1. An isolated DNA sequence having the sequence identified as SEQ ID NO:1.

2. An expression vector comprising the DNA sequence of claim 1.

3. A host cell comprising the DNA sequence of claim 1.

4. An isolated DNA sequence having the sequence identified as SEQ ID NO:1, said sequence coding for a protein having the peptide sequence of SEQ ID NO:2.

5. An isolated DNA sequence having the sequence identified as SEQ ID NO:1 isolated by the process comprising amplifying, cloning and characterizing coding DNA sequences by employing an Alu-splicing PCR technique and using primers corresponding to the cutting and splicing 5'spl or 3'spl acceptor and donor consensus sequences, in combination with consensus primers for 5' or 3' sequences of Alu repeats.

6. The DNA sequence of claim 5 wherein the process for isolating the DNA sequence further comprises primers to amplify the restriction targets NotI, SacII and SalI, said primers having the sequences:

NotI-5' spl(5'-CGCGCGGCCGCACWYACCW-3'(5'-spl));

SacII-3' spl(5'-CGCCCGCGGTCNCAGGT-3'(3'-spl)),

where W represents A or T, Y represents T or C, and N represents any nucleotide; and

Alu primers:

SalI-A33(5'-CGCGTCGACCACTGCACTCCAGCCTGGGCG-3'); and

SalI-A44(5'-CCGTCGACGGGATTACAGGCGTGAGCCAC-3').

7. The DNA sequence of claim 5 isolated by a process that further comprises a digestion of the coding DNA sequences, amplified, cloned and characterized by employing an Alu-splicing PCR technique according to claim 5, with restriction enzymes and further subcloning into a vector.

8. The DNA sequence of claim 5 characterized by means of sequencing with a primer complementary to T3 exon and by means of a primer complementary to T7 sequences for an Alu-sequence.

Description
FIELD OF THE INVENTION

Down syndrome is one of the main causes of mental retardation and of congenital heart defects, which is predominantly the result of three copies of chromosome 21. Chromosome studies of rare patients with partial chromosome 21 trisomy have defined a minimum region for the Down syndrome phenotype which includes approximately 3 megabases (Mb) around the D21S55 marker. The overexpression of a new gene sequence, DSCR1 (SEQ ID NO:1), isolated from this region may be involved in pathogenic abnormalities of mental retardation and/or heart defects in patients with Down syndrome.

PRIOR ART

Down syndrome (DS) is the main cause of mental retardation and or congenital heart defects, which affects one in seven hundred newborn babies.sup.1. DS is associated with abnormalities in the gastrointestinal tract, an increase in the risk of contracting leukemia defects of the immune and endocrine systems, premature onset of Alzheimer's dementia and characteristic facial and physical features.

The relationship between DS and chromosome 21 trisomy was established 35 years ago.sup.3,4. Since then, a total of 43 new genes in chromosome 21 have been identified and cloned, but little progress has been made in our understanding of the disorder.sup.5. The presence of three copies of chromosome 21 should give rise to the overexpression of various genes, some of which would be responsible for several of the phenotypic features of DS. Although various genes may contribute to each specific abnormality observed in DS, it is possible to speculate that one gene alone may be responsible for each clinical characteristic. In fact, studies of cases with partial chromosome 21 trisomy suggest that particular regions of the chromosome contain the genes responsible for specific features of DS.sup.6. Although a minimum region has been proposed for the DS phenotype around the D21S55 marker, other studies suggest that genes located outside the D21S55 region contribute significantly to the main phenotype of DS.sup.7,8. Elucidation of the contribution of each of the individual genes to the main phenotypic features will be possible only through the isolation of each gene and the analysis of its expression and function. Of the whole of chromosome 21, the q22 region is the main target in the search for genes potentially involved in DS.

The characterization of chromosome 21 has advanced considerably over the last five years. With the development of new markers, the genetic map of this chromosome has achieved a resolution of better than 3 centimorgans (cM).sup.9,10. Two physical maps of this chromosome have been obtained using overlapping YACS (yeast artificial ;chromosomes) and cosmid clones ordered along the long arm of the chromosome.sup.11,12, and a complete NotI restriction map has been generated with its corresponding binding clones.sup.13. Recently, expression maps of chromosome 21 have been started, which will assist in identifying all the genes of this chromosome.sup.14,15.

Various methods have been developed for the identification of coding sequences of specific regions of the genomic DNA. These methods include the hybridization of cDNAs on filters using genomic DNA as probe.sup.16,17, the trapping.sup.18,19, and the direct selection of cDNA clones (cDNA selection).sup.20,21.

The identification of genes of chromosome 21 is crucial for understanding the mechanisms involved in the various phenotypic manifestations of DS. Transcription maps of chromosome 21 and of the DS critical region have been developed, and work is continuing on obtaining them.sup.14,15,24. This effort will provide a catalogue of coding genes for this chromosome, which will enable us to study diseases linked to human chromosome 21. A region critical for DS (about 3 Mb around D21S55) covers the majority of the clinical features of DS, including mental retardation.sup.7,25-27. Another study has demonstrated that DS is a syndrome of contiguous genes and that the genes outside D21S55 also contribute to the DS phenotype.sup.8. Although these studies are based on a very small number of patients, the region around D21S55 is the main target for the identification of genes which are probably involved in DS. In this connection, a large number of patients have employed the presence of some polypeptide linked to genes present on chromosome 21 as instruments for diagnosis of the disease, especially in the gestation period. An example of these methods is that relating to the precursor protein of the amyloid .beta.-peptide (EP 576,152). Similarly, probes specific for chromosome 21 employing Alu sequences (TC65 and 517) as specific initiators (FR 2,690,460) or complex cosmids which map in the 21q22 zone (WO93/18184) are described in the literature.

The screening of a gene library of cDNA from foetal brain (1.times.10.sup.6 lfu (lysis-forming units)) gave several positives. The DSCR1 protein deduced from the clones BC-17.8-1 and BC-17.8-2 has a 41% identity with the protein product deduced from F54E7.7 of C. elegans. The main difference between the two amino acid sequences lies in their length, since C. elegans protein has 20 amino acids more at the carboxy-terminal end and possesses 45 additional amino acids at the amino-terminal end. A search in the database for EST (expressed tagged sequences) showed that the human sequences 21ES87, 21ES127 and 21ES165 were almost identical to BC-17.8 in nucleotides 392 to 761, 724 to 1012 and 1200 to 1392, respectively, the last two corresponding to the untranslated 3' end of BC-17.8, and only 21ES87 contains part of the coding sequence.sup.14. Similarly, DSCR1 has a 65.4% identity at nucleotide level with the human brain EST EST07037 (T09144).sup.52.

Generally speaking, the homologies found between both nucleotide and polypeptide sequences, though high in some cases, correspond to only a part of the sequence described here, sometimes corresponding to very few residues and, except in the case mentioned above, not to coding residues. A search in the EMBL, Gen Bank, STRAND (European Patent Office) and SWISS PROT databases with respect to BC-17.8 reveals similarities only with deduced proteins of unknown function, F54E7.7 in chromosome III of C. elegans.sup.22 and a protein of 24.1 kD (YKL159c) from the intergenic region PIR3-APE2 in chromosome XI of S. cerevisiae.sup.30. A more detailed analysis of smaller regions of the DSCR1 protein shows similarity with a large number of known proteins and peptides, identifying regions with some possible functional importance.

SUMMARY OF THE INVENTION

By means of a new molecular cloning technique designated "Alu-splicing PCR", we have isolated a new gene sequence, pSCR1 (SEQ ID NO:1), located in the 21q22.1-q22.2 region. DSCR1 (SEQ ID NO:1) displays a high expression in brain and heart, coding for a new protein with a proline-rich region and with some structural characteristics typical of a protein involved in transcription and/or in protein-protein interactions. The increase in the transient expression of DSCR1 (SEQ ID NO:1) mRNA in brains of young rats compared with the brains of adult rats suggests an important role of DSCR1 during the development of the central nervous system. The overexpression of DSCR1 (SEQ ID NO:1) may be involved in pathogenic abnormalities of mental retardation and/or heart defects in patients with Down syndrome.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A-1C show the identification of a putative exon of YAC 72H9 of chromosome 21. FIG. 1A is a diagrammatic representation of the cloning strategy by "Alu-splicing PCR" for the isolation of possible exons. FIG. 1B shows PCR products stained with ethidium bromide and analyzed by electrophoresis on an agarose gel. FIG. 1C is the DNA sequence of clone 72H9-17.8.

FIGS. 2A-2C show the physical localization of the DSCR1 gene of human chromosome 21q22. FIG. 2A shows the subchromosomal localization of 72H9-17.8 on 21q22.1-22.2 using hybrid cell lines. FIG. 2B shows the subregional localization of the DSCR1 gene sequence in the YACs originating from the 21q22.1-22.2 region. FIG. 2C is a diagrammatic representation of the 21q22.1-22.2 region.

FIGS. 3A-3D show the analysis of expression of the DSCR1 gene. FIG. 3A is a northern analysis of adult mRNA. FIG. 3B is a northern analysis of fetal mRNA. FIG. 3C shows the detection by RT-PCR of the DSCR1 mRNA in human lymphocytes. FIG. 3D is a northern analysis of total mRNA originating human brain and rat brain, heart and kidney.

FIGS. 4A-4C show the expression of DSCR1 mRNA. FIG. 4A is expression in adult rat brain. FIG. 4B is expression in adult rat heart and FIG. 4C is expression in adult rat liver.

DETAILED DESCRIPTION OF THE INVENTION

New Method for the identification of gene sequences

The method we describe here is based on the proximity of Alu repeat elements to one or more exons of a given gene. The system employs primers corresponding to the cutting and splicing (5'spl or 3'spl) acceptor and donor consensus sequences, in combinations with consensus primers for 5' or 3' sequences of Alu repeats. The sequences amplified using these combinations can include exons form any gene localized in the genomic sequence used as substrate for the amplification (FIG. 1a). This new method, designated "Alu-splicing PCR", has the limitation that not all exons are near to Alu repeats (A33 or A44) and that not all 5' or 3' splicing points are consensus, but the genes with a high number of exons will be readily identified by this method. Moreover, the Alu-splicing method has the advantage that mRNA (messenger RNA) or cDNA (complementary DNA) is not necessary for the detection of possible genes, and that whole YACs may be employed for the identification of exons. Since mRNA is not employed, the method does not depend on the levels of gene expression. Finally, as in the other methods, screening in cDNA libraries is needed in order to confirm the coding sequences and to obtain the complete length of the clones.

Among several possible exons of the q22.1-22.2 region identified by this technique, we have characterized one completely, which corresponds to the new gene sequence DSCR1 (SEQ ID NO:1). Amplifications using different combinations of 5'spl or 3' spl primers and A33 or A44 primers provide several fragments of the YACs 72H9 and 860G11. Since these YACs overlap on mpas of STSs (sequence tagged sites), the amplified products should also reflect this overlapping. FIG. 1b shows the common amplification of products detected between the two YACs using two combinations of primers. The subclones obtained from each of the four PCR reactions were sequenced, and possible homologies were sought in the public databases using the BLAST e-mail service at the NCBI. One of the clones, 72H9-17.8 (FIG. 1c), showed a 46% identity with the product deduced from the P54E7.7 gene of C. elegans.sup.22. Confirmation that the 72H9-17.8 sequence was obtained by means of a 5'spl-A33 amplification was carried out by means of hybridization with an internal oligonucleotide of the new sequence 72H9-17.8 on a filter which contained products of all the combinations of primers. YAC 860G11 was also positive for 72H9-17.8 (data not shown).

Identification of DSCR1 cDNA

The overlapping of two clones (BC-17.8-1 of 1.3 kb, and BC-17.8-2 of 2.0 kb) resulted in a cDNA of 2174 bp (FIG. 2). The sequence has an open reading frame of 513 nucleotides and encodes 171 amino acids; a putative ATG initiation codon, with the consensus sequence of Kozak.sup.23 at the nucleotide at position 49, in frame with a termination codon, localized 9 bp and 42 bp upstream. In this way, the untranslated 5' region is 48 bp in length. The untranslated 3' region of DSCR1 is 163 bp. Two consensus polyadenylation signals were detected within the untranslated 3' region. The new gene sequence was designated Down syndrome critical region 1 (DSCR1) (SEQ ID NO:1).

DSCR1 is localized on human chromosome 21q22.1-q22.2

The localization of DSCR1 (SEQ ID NO:1) on human chromosome 21 was confirmed by PCR with the cell line WAV17, which contains only human chromosome 21 in mouse cells. Subchromosomal localization using a panel of somatic hybrids showed that all the lines except 1x18 and 8q.sup.-a were positive for the new gene sequence, defining the localization of the latter in 21q22.1-q22.2 (FIG. 3a).

A finger localization of 72H9-17.8 was carried out using YACs located in the subregion 21q22.1-q22.2 as templates for amplification with PCR. The following YACs were positive: 73D10, 72H9, 812F11, 860G11 and 916H12; while 520C10, 529C8, A222A12, 972C3, 212A5 and 230E8 were negative (FIG. 3b). Analysis of the cell lines of somatic hybrids with STSs of this region showed that 21q.sup.+ (21pter.fwdarw.21q22) and 8q.sup.-a (21q.fwdarw.22gter) were both positive for AML1 and D21S65, but only 21q.sup.+ contained the sequence of 72H9-17.8. A restriction map for PPGE (pulsed-field gel electrophoresis) was carried out using BC-17.8 as probe and the PCR products derived from markers present in the YACs 72H9 and 73D10, localizing gene sequence DSCR1 at approximately 400 kb from the locus of AML1 in the centromers direction (FIG. 3c).

Analysis of the expression of the gene sequence DSCR1

The analysis of expression of DSCR1 (SEQ ID NO:1) in human tissues by means of the Northern transfer method identified a transcript of about 2.2 kb. The highest levels of expression were detected in foetal brain and adult heart. Lower levels of expression were detected in adult brain, lung, liver, skeletal muscle, kidney and pancreas, and in the foetal period in lung, liver, kidney and placenta. A second transcript of 2.0 kb was detected in foetal and adult liver (FIGS. 4a and b). Levels of DSCR1 mRNA in peripheral blood lymphocytes, in fibroblasts of the cell line WAV17 and in brain were determined by RT-PCR (reverse transcription-PCR). Amplification was observed in lymphocytes and fibroblasts after carrying out two PCRs, the second with internal primers (nested PCR), but only one PCR amplification was needed in brain (FIG. 4c). A Northern transfer was also carried out with samples of rat brain, heart and kidney using BC-17.8 as probe, and the pattern of expression was similar to that for human tissue, also with more expression in brain and heart (FIG. 4d).

In situ hybridization studies were carried out in rats with a 40-bp antisense oligonucleotide derived form the DSCR1 human transcript. DSCR1 mRNA is expressed in adult and newborn rat tissues. In adult rats, the in situ hybridization signal was larger in particular regions of the brain and heart than in the liver (FIG. 4). The in situ hybridization signals form heart and from liver showed a fairly homogeneous distribution pattern, but hybridization signals with different intensities were found in some regions of adult brain, such as the neocortex, striate bodies, hippocampal formation, hypothalamus, thalamus, olfactory bulb and cerebellar cortex, and no signal at all was detected in the white matter. The largest in situ hybridization signal was localized in the olfactory bulb, piriform cortex pyramidal cell layer of the hippocampus, striate bodies and cerebellar cortex. Similarly, rats on days 2, 7 and 16 after birth expressed DSCR1 mRNA, although the signal in the neocortex and hypothalamus was larger in young rats than in adult rats. Comparatively, the largest in situ hybridization signal was detected on days P2 and P7, and decreased in the majority of the regions of the brain from day P16.

DSCR1 is a new gene sequence of chromosome 21

We describe here the isolation, characterization and analysis of expression of a new gene sequence (DSCR1) (SEQ ID NO:1) of the q22.1-22.2 region of human chromosome 21.

The new gene sequence DSCR1 has been localized by somatic hybrids, PFGE and maps of YACs in 21q22.1-22.2, at 400 kb from AML1, around 2 Mb of the D21S55 locus. Thus, DSCR1 is in a region considered to contribute significantly to the main phenotype of DS. The pattern of expression of DSCR1 mRNA and its distribution in tissues is consistent with the view that this gene is important in brain and heart, and consequently its overexpression in these tissues may cause some of the phenotypic features of DS. The absence of DSCR1 mRNA in the white matter of the central nervous system is compatible with neuronal expression. Nevertheless, this pattern does not reflect the neuronal density, since regions such as the striatum, in which the neuronal density is not particularly high, show, comparatively, high levels of DSCR1 mRNA. Moreover, the level of expression is probably unrelated to cell size, since regions with a similar degree of cell size have different signal intensities.sup.28. This implies that some neuronal subpopulations show a larger degree of expression of DSCR1 than others. Transient increases in the expression of DSCR1 mRNA in brains of young rats in comparison with those of adults suggest an especially important role of DSCR1 during the development of the central nervous system. Consequently, it is possible that abnormalities in the development of the brain may be due to the overexpression of DSCR1. Finally, the distribution of DSCR1 mRNA does not conform to any pattern of the phylogenetic scale.sup.29. Thus, DSCR1 may have a critical function in particular states of development and in specific neuronal subpopulations, regions of the brain which express, comparatively, high levels of DSCR1 mRNA. The high levels of expression of DSCR1 in heart suggest that it is also important for the function and/or development of the heart. The determination of the importance of DSCR1 in brain and in heart and its involvement in DS is carried out by means of functional studies, as the use of mice without any copy of the homologue of this DSCR1 human gene sequence in mouse (knock-out) and of transgenic mice which overexpress the sequence.

Protein deduced from DSCR1 and its function

The deduced protein sequence (SEQ ID NO:2) provides some keys to the possible function of DSCR1. The amino-terminal region of DSCR1 is extremely rich in leucines and phenylalanines (25%), residues 17 to 56 could form an alpha-helix. Although this region cannot be considered to be a leucine-zipper domain, a helix-loop-helix motif or a homeo domain, this region has the hydrophobicity and the amphipathic motifs characteristic of these types of regions and could be involved in the binding of DNA and/or in multimerization.sup.31-33.

There are two regions rich in proline residues at residues 69 to 88. 80% of these residues, including 6 prolines, are identical to the product deduced from the F54E7.7 gene of C. elegans. Similarly, a 68% identity was observed with the hypothetical protein of 24.1 kD of yeast. The first proline-rich sequence in DSCR1 (position 69 to 75, HLAPPNPDK) shows great similarity with ligands of both classes I and II of the SH3 domains. The consensus sequence of these SH3 domain ligands is Xp.alpha.PpXP, where P represents a proline, p is generally a proline, .alpha. is a hydrophobic residue and X corresponds to an unconserved residue.sup.34,35 l. SH3 domains are conserved sequences present in various signal proteins and/or proteins of the cytoskeleton, some of which are involved in tyrosine kinase-dependent signal transduction.sup.36. If this proline-rich motif in DSCR1 is really an SH3 domain ligand, it should interact with proteins which contain SH3 domains.

The second proline-rich sequence is DSCR1 shows identity with a large number of proline-rich proteins, including some transcription factors. Clues to the possible function of the motif SPPASPP may be found in the cellular transcription factor E2F-1.sup.37. The phosphorylation of two serines in the motif SPPPSSPPSS of E2F-1 inhibits its interaction with the product of the retinoblastoma gene and stimulates the interaction with the E4 protein of adenovirus.sup.38. It would appear that the phosphylation of E2F-1 facilitates the protein-protein interactions which may be essential for the transcription of E2A and/or the expression of other target cellular genes of E2F.sup.39. In DSCR1, the serines at positions 82 and 86 are in each case followed by a proline, and may consequently be regulated by protein kinases.

Seven of the nine residues between positions 140 and 148 of DSCR1 correspond to glutamic acids. The sequences with capacity to activate transcription are frequently negatively charged and increase the capacity of basal transcription factors to initiate transcription. Various types of transcriptional activators have been identified, which may be acid, glutamine-rich or proline-rich domains.sup.41. Many proteins with various functions, including some transcription factors, have a glutamic acid-rich domain. The glutamic acid domain of DSCR1 could interact with other components of the transcriptional apparatus.

Another noteworthy feature of the carboxyl end of DSCR1 is a motif (PYTPI) which includes amino acids 163 to 168 which could bind to SH2 domains. This type of domain binds with different regions with phosphotyrosines (pTyr), with a specificity determined by the next three residues from the carboxyl end to the pTyr.sup.43. The domain of the hypothetical SH2 ligand of DSCR1 coincides with the consensus and is similar to other SH2 ligands described in other proteins. The signal transduction induced by the possible phosphorylation on the tyrosine of this domain could involve it in processes of dimerization and/or binding to DNA.sup.44.

In summary, the invention consists in the isolation and characterization of a new gene sequence (pSCR1) localized on human chromosome 21, in the Down syndrome critical region, which codes for a proline-rich protein, with structural characteristics of protein involved in transcription and/or in protein-protein interactions. The said protein displays a high expression in brain and in heart. The larger expression of DSCR1 in brains of young rats suggests an important role in the development of the central nervous system. Four features of the protein deduced from DSCR1 lead us to the view that it is involved in transcription processes and/or in protein-protein interactions: possible domain of binding to DNA, multimerization domain, two potential domains of transcriptional activation (one acid and the other proline-rich) and putative ligand domains for SH2 and SH3. All these features are among the characteristics of transcription factors and transcription enhancers.sup.41. Thus, it is possible to speculate that the overexpression of a transcription factor in the central nervous system may have an effect on several genes, whose overexpression could be directly responsible for the abnormal development of the brain in DS.

Methodology Used in the Implementation of the Invention

Characterization and cloning by "Alu-splicing" PCR.

The putative exons of the YACs 72H9 and 860G11 were amplified by PCR by virtue of their proximity to human Alu sequences. The splicing site primers are: NotI-5'spl, 5'-CGCGCGGCCGCACWYACCW-3' (SEQ ID NO:3) (5'-splice); SacII-3'spl: 5'-CGCCCGCGGTCNCAGGT-3' (SEQ ID NO:4) (3'-splice), where W represents A or T, Y represents T or C, and N represents any nucleotide. The Alu primers are:

SalI-A33: 5'-CGCGTCGACCACTGCACTCCAGCCTGGGCG-3'(SEQ ID NO:5); and

SalI-A44: 5'-CGCGTCGACGGGATTAGGCGTGAGCCAC-3'(SEQ ID NO:6).sup.45,46.

The primers were designed with a restriction target to permit directional cloning of the amplified fragments. The final volume of the "Alu-splicing PC" reactions was 25 .mu.l with the following components: 40 mM KCl, 8 mM Tris-HCl (pH 8.3), 2.7-3.3 mM MgCl.sub.2, depending on the purity of the sample of YAC DNA, 0.01% gelatin, 0.15 mM each deoxynucleotide, 20 pmol of each primer and 300 ng of template DNA. The PCR conditions were 94.degree. C. for 20 s, 55.degree. C. for 40 s and 74.degree. C. for 2 min, in a total of 30 cycles. After the amplification, the PCR products were purified with phenol/chloroform, precipitating the DNA with ethanol, the latter being digested with the appropriate restriction enzymes and the products being subcloned into the vector pBluescript SK.sup.+ (Stratagene). The positive clones selected were sequenced using T3 for the sequence of the exon (splice side) or with T7 for the Alu sequence (Alu side), using an ABI373A automatic DNA sequencer and DyeDeoxy Terminators (Applied Biosystems). The human origin and the localization of the clones were determined using a panel of human and rodent hybrid cells containing whole chromosome 21 or parts of this chromosome. A panel of clones of YACs, including 72H9 and 860G11 and other YACs from other regions of chromosome 21, was used to confirm the localization of the Alu-splice clones.

Screening and analysis of the cDNA.

A gene library of cDNA from foetal brain subcloned into .lambda.gt10 was acquired from Clontech (Palo Alto, Calif.) and was inoculated into 20 plates at an average density of 5.times.10.sup.4 lysis-forming units (lfu) per plate. The phages from each plate were recovered by covering the plates with SM buffer (100 mM NaCl, 10 mM MgSO.sub.4, 50 mM Tris-HCl, pH 7.5, 0.01% gelatin) and, 12 hours later, the buffer was recovered by aliquoting it. The aliquots from each sub-gene library were used in subsequent PCR experiments. Two pairs of primers were synthesized taking as template the sequences of putative exons, identified by the Alu-splicing technique, and were used in the screening of these sub-gene libraries of cDNA described above.sup.47. The primers corresponding to one of the putative exon sequences (17.8-1: 5'-AGGACGTATGACAAGGACATC-3'; (SEQ ID NO:7) corresponding to nucleotides 79-98), and (17.8-2: 5'-TGAGCAAAATATAACTTCATTTCCT-3', (SEQ ID NO:8) complementary to nucleotides 206-230) were used to screen the gene library of cDNA already described using as probe the 17.8-2 primer terminally labelled with T4 polynucleotide kinase.sup.47. The positive phages were purified and the inserts were subcloned into pBluescript SK.sup.+ at the EcoRI target. The clones which contained the desired inserts were sequenced as described above. The complete sequence of the cDNA was obtained by synthesizing new primers on the sequences already analysed (primer waling), and was analysed using the BLAST program of the National Center for Biological Information (NCBI) to look for possible homologies.

Somatic hybrid cells and genomic samples.

The localization of DSCR1 on chromosome 21 was confirmed using DNA of the human-mouse hybrid cell line WAV17, which contains chromosome 21 as the only human chromosome.sup.48. The sublocalization of the clone 72H9-17.8 was determined by amplifying the DNA of the panel of the somatic hybrid cell line specific for human chromosome 21.times.Chinese hamster, composed of the following cell lines: 153E7b, 2Fur-1, ACEM, JC6, 6;21, 4;21, 1;21, 3;21, 10;21, 21;22, R2-10W, 7;21, 6918-8a1, 1x18C9, 3x2S, 8q.sup.-a, 21q.sup.+ and MRC2G.sup.49,50. DNA from whole peripheral human blood, Chinese hamster DNA and mouse DNA were included as controls in the PCR experiments. DNAs of the cell lines were obtained from D. Patterson. The human DNAs were obtained from precipitates of lymphocytes by lysis with SDS, digestion with proteinase K and salt precipitation. The genomic cDNA probes were labelled with .alpha.-.sup.32 P!dCTP by "random priming".

YACs of human chromosome 21 and STSs.

The following YACs of human chromosome 21 were used in this study: 72H9, 860G11, 73D10, 520D10, 529C8, 230E8, 212A5 and 812F11 supplied by CEPH/Genethon.sup.11 and A222A12, 972C3 and 916H12 supplied by D. Patterson. The yeasts which contained the desired YACs were grown in AHC.sup.- selective medium and were encapsulated in agarose beads using a modification of the method described by Overhauser and Radio.sup.51. Sequence specific to these YACs were amplified by the use of PCR with the primers which flank the respective STSs. To produce the restriction map, the agarose beads were digested overnight with restriction enzymes and electrophoresis was carried out on a 1% agarose gel in 0.5.times. TBE. The DNA was transferred to Hybond N.sup.+ membranes in 0.4N NaOH. Hybridizations were carried out in 7% SDS/0.5M NaP.sub.2 O.sub.4 buffer using probes labelled by "random priming".

The STSs which were analyzed are: D21S332, D21S328, D21S65, AML1, D21S393, D21S211 and D21S17. The primer pairs corresponding to the 5' and 3' sequence of DSCR1 were also used for its localization on the YACs and in the panels of somatic hybrids. PCR amplifications were carried out with 10 pmol of each primer, 1 mM MgCl.sub.2 40 mM KCl, 8 mM Tris-HCl (pH 8.3), 150 mM each dNTP, 100 ng of yeast DNA containing the relevant YAC and 0.5 units of Taq DNA polymerase (Boehringer Mannheim) in a final volume of 25 .mu.l, using a Perkin-Elmer 9600 thermocycler. The initial denaturation time was 92.degree. C. for 5 min, amplification proceeded for 30 cycles of 94.degree. C. for 20 s, hybridization at 58.degree. C. for 40 s and extension of 74.degree. C. for 1 min.

RNA analysis.

Two northern blots (Clontech, Pal Alto, Calif.) containing poly(A).sup.+ messenger RNA (mRNA) form adult and foetal human tissues were hybridized with probes labelled according to the protocol from the commercial firms. In addition, total RNA was isolated from lymphocytes and from brain, and poly(A).sup.+ mRNA was prepared from rat brain, liver, heart, kidney and lung using the guanidine isothiocyanate method and the FastTract system (Invitrogen). Samples of human total RNA were also analyzed using reverse-transcription PCR (RT-PCR), employing 1 .mu.g of total RNA and the GeneAmp RNA-PCR kit (Perkin Elmer).

In situ hybridizations of tissues.

Studies of in situ hybridizations in rat brain and in heart and oliver were carried out with an antisense synthetic oligonucleotide of 40 bases (5'-CAAAGGTGATGTCCTTGTCATACGTCCTAAAGAGGGACTC-3') (SEQ ID NO:9) which is complementary to bases 136 to 97 of the human cDNA sequence of DSCR1. Adult rats and 2-, 7- and 16-day-old rats (3 rats per age studied) were sacrificed by decapitation, their tissues were rapidly extracted and frozen in dry ice, and 20 mm sections were cut with a cryostat at -20.degree. C. and were kept at -20.degree. C. coated with gelatin until required for use. The oligonucleotides were labelled with .alpha.-.sup.32 P!dATP (>3,000 Ci/mmol, NEN) with a specific activity of 1-10 .mu.Ci/pmol. Two pmol of each oligonucleotide were added to (50 .mu.Ci (5 ml) of .alpha.-.sup.32 P!dATP in the presence of 10 .mu.l of cutting buffer (which contained 500 mM Na cacodylate pH 7.2, 10 mM CoCl.sub.2 and 1 mM dithiotreitol) and 50 units of deoxynucleotidile terminal transferase (Boehringer Mannheim) and the mixture was incubated at 37.degree. C. for 2 h. The labelled probe was purified using an ion exchange column (Pharmacia). The brain sections were dried and were fixed with 4% paraformaldehyde (in 0.1M phosphate buffered saline), and were then treated with 12 units/ml of predigested pronase (Calbiochem), dehydrated and incubated overnight at 42.degree. C. with a mixture containing 40% deionized formamide, 0.6M NaCl, 1.times. Denhardt's solution (1.times. contains 0.02% Ficoll, 0.02% polyvinylpyrrolidone and 0.02% bovine serum albumin), 10 mM Tris-HCl pH 7.5, 1 mM ethylenediaminetetraacetic acid, 0.5 mg/ml yeast transfer RNA (Gibco BRL), 10% dextran sulphate and 10% of labelled probe. After hybridization, the sections were washed twice in 1.times. SSC (150 mM sodium chloride and 15 mM sodium citrate, pH 7.0) at 52.degree. C. for 1 h each time. Subsequent experiments were carried out with different stringencies in the post-hybridization washes (SSC concentrations within the range 2.times. to 0.1.times.). In these cases, the sections were dehydrated in 70% and subsequently 95% ethanol., both solutions containing 0.3M ammonium acetate pH 7 for 5 min in each case, and then dried in the air. .beta.max films (Amersham) were exposed to the labelled sections for 7 days at -80.degree. C. The sections were then stained with cresyl violet. Specific controls were included in the hybridizations, such as the sense probe and competitive experiments entailing addition of the non-radioactive antisense oligonucleotide (50 times the concentration of the antisense oligonucleotide) in the hybridization solution.

Application in transgenic animals

The gene homologous to DSCR1 in the mouse, Dscr1, has been identified from gene libraries of cDNA from several tissues (liver, brain and embryo). Transgenic mice are being developed which will enable the role of the DSCR1 gene in Down syndrome to be studied. A first transgenic animal would be that which expresses the DSCR1 gene under the control of the promoter of the PDGF-B gene, which directs the expression of the gene functioning effectively and specifically in the brain. The transgenic mice will be obtained by means of the micro-injection of the transgene into B6SJL.times.B6SJL embryos, and the presence of the chimeric gene will be identified by means of PCR and Southern Blotting. Positive animals are analysed from the standpoint of expression, localization and function of the DSCR1 protein. A second transgenic animal will be that which is obtained by means of the use of the antisense sequence Dscr1, with the object of curbing the expression of this gene in the mouse. These experiments also include crosses with mice carrying an extra chromosome which includes the Dscr1 gene. Since these mice have phenotypic characteristics of Down syndrome, the reversion of some of the features in the mouse will enable an advance to be made in our knowledge about the function of Dscr1/DSCR1 and its role in Down syndrome.