![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
|
Vol. 11, Issue 6, 1053-1070, June 2001
LETTER
|
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
We have sequenced a 1.1-Mb region of human chromosome 22q containing the dosage-sensitive gene(s) responsible for cat eye syndrome (CES) as well as the 450-kb homologous region on mouse chromosome 6. Fourteen putative genes were identified within or adjacent to the human CES critical region (CESCR), including three known genes (IL-17R, ATP6E, and BID) and nine novel genes, based on EST identity. Two putative genes (CECR3 and CECR9) were identified, in the absence of EST hits, by comparing segments of human and mouse genomic sequence around two solitary amplified exons, thus showing the utility of comparative genomic sequence analysis in identifying transcripts. Of the 14 genes, 10 were confirmed to be present in the mouse genomic sequence in the same order and orientation as in human. Absent from the mouse region of conserved synteny are CECR1, a promising CES candidate gene from the center of the contig, neighboring CECR4, and CECR7 and CECR8, which are located in the gene-poor proximal 400 kb of the contig. This latter proximal region, located ~1 Mb from the centromere, shows abundant duplicated gene fragments typical of pericentromeric DNA. The margin of this region also delineates the boundary of conserved synteny between the CESCR and mouse chromosome 6. Because the proximal CESCR appears abundant in duplicated segments and, therefore, is likely to be gene poor, we consider the putative genes identified in the distal CESCR to represent the majority of candidate genes for involvement in CES.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Cat eye syndrome (CES, Online Mendelian Inheritance in Man (OMIM)
no. 115470) is a rare developmental disorder in
humans associated with the presence of three or four copies of a
segment of chromosome 22q11.2, usually in the form of a bisatellited,
isodicentric supernumerary chromosome (Schinzel et al. 1981
; McDermid
et al. 1986). CES is characterized by a variety of congenital defects
including ocular coloboma, anal atresia, preauricular tags/pits, heart
and kidney defects, dysmorphic facial features, and mental retardation
(Schinzel et al. 1981). The penetrance and severity of these phenotypic features are highly variable. The smallest region of duplication required to produce the CES phenotype is the first 2 Mb of 22q (from
the centromere to marker D22S57), as determined by analysis of a
patient (25105) with all major features of the syndrome and an
unusually small supernumerary dicentric ring chromosome (Mears et al.
1995). The 2-Mb CES critical region (CESCR) can be subdivided into
~1-Mb proximal and distal halves (between markers D22S795 and
D22S543) based on the proximal breakpoint of an interstitial duplication in patient "SK" (H. McDermid et al., in prep.). The CES
features present in this patient (facial features, ear pits/tags, heart
and kidney malformations, and mental retardation) (Knoll et al. 1995)
should, therefore, map to the distal half of the CESCR. Absence of
coloboma and anal atresia in this patient may be due to the phenotypic
variability of the syndrome.
To identify candidate genes for CES, the distal half of the CESCR was
cloned in a contiguous array of bacterial and P1-based artificial
chromosomes (BACs/PACs) (Johnson et al. 1999). A minimal overlapping
set of clones from the region now has been sequenced and partially
annotated with various ESTs and the genes IL-17R, ATP6E, and BID (Dunham et al. 1999). Here, we report
the analysis of 14 putative genes within and adjacent to the distal 1.1 Mb of the CESCR. In addition, the sequence of a contiguous array of
orthologous mouse genomic BAC clones revealed that 10 of these genes
are present as a single linkage group on mouse chromosome 6. Sequence
analysis of the proximal 400 kb of the human CESCR region revealed a
complex mosaic of duplicated segments originating from elsewhere in the
genome, similar to the pericentromeric regions of human chromosomes 16 (Horvath et al. 2000) and 10 (Jackson et al. 1999). The discovery of a
22q pericentromeric region transcript (CECR7), in which individual
exons appear to be derived via duplication of gene sequences from
different chromosomes, indicates this region may be the birth site of
novel genes produced by shuffling of existing sequences.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Physical and Transcript Map of the CESCR and Orthologous Mouse Region
Figure 1 outlines our general strategy
for combining standard positional cloning techniques with annotation of
large-scale genomic sequence to establish comprehensive transcript maps
of the distal 1.1 Mb of the CESCR and the 450-kb orthologous region of
mouse chromosome 6. Previous mapping of Bid (Footz et al.
1998), Atp6e (Puech et al. 1997; Footz et al. 1998), and
Il-17r (Yao et al. 1997) to mouse chromosome 6 suggested that
much of the region would show a conservation of synteny.
|
A minimal set of genomic BAC/PAC clones from human chromosome 22q11.2
and mouse chromosome 6 were subjected to shotgun sequencing and the
relative position of putative genes identified are illustrated in
Figure 2. The cloning, sequencing, and
initial analysis of the human BAC/PAC contig for this region was
described previously (Dunham et al. 1999; Johnson et al. 1999). The
mouse contig, consisting of 13 clones (only seven are shown in Fig. 2),
was constructed by probing a mouse BAC library with mouse cDNAs
orthologous to human CESCR genes (identified through
BLASTN dbEST searches) and with a 5' RACE-PCR product of
CECR2. The sequence of the contigs of human and mouse clones
was used to identify putative genes, described in detail below. A
comparison of human and murine sequence using the PipMaker
program (Schwartz et al. 2000) indicated that 10 genes from
IL-17R to BID on 22q11.2 have orthologs in a single
linkage group on mouse chromosome 6, with preservation of both gene
order and orientation, although two central genes are absent (Fig.
3).
|
|
A unique distribution of putative genes was discovered within the
1.1-Mb region of the CESCR. Only 2 genes (CECR7 and
CECR8) localize to the ~400 kb proximal to IL-17R,
whereas the distal ~700 kb contains 11 genes, including
IL-17R (Fig. 2). This gives an average gene density of 1 per
200 kb in the proximal region and 1 per 64 kb in the distal region.
Other features suggest a basic difference between these two regions.
The proximal 400 kb is 41.1% GC with only eight predicted CpG islands
(~1 per 50 kb), whereas the distal 700 kb is 46.4% GC and contains
28 predicted CpG islands (~1 per 25 kb; Fig. 2). Of the eight CpG
islands in the proximal 400 kb, only two are >1 kb in length, and
these correspond to the 5' end of CECR7 and the region just
upstream of IL-17R. The remaining six CpG islands are ~500
bp or less in size and none fall near the other pericentromeric gene,
CECR8. In the distal 700 kb, CpG islands of >1 kb were found
to correspond to the 5' ends of CECR2, CECR6,
ATP6E, and each of the alternate 5' ends of MIL1 and
CECR5. The remaining 20 CpG islands are all <900 bp in length
and appear to be scattered throughout the distal region. The
distribution of interspersed repeats also differs between the proximal
and distal regions (Table 1). The 400 kb
proximal to IL-17R is relatively rich in LINE and LTR elements
but not in SINEs, whereas the distal 700 kb is SINE rich and LINE and LTR poor. Both regions, however, share similar fractions of total interspersed repeats. A preponderance of paralogous segments in the
proximal 400 kb is reminiscent of the organization of other pericentromeres (Eichler et al. 1996, 1997; Régnier et al. 1997; Ritchie et al. 1998; Jackson et al. 1999; Horvath et al. 2000). A more
detailed analysis of the sequences in the 22q pericentromere is
described below. These differences suggest a boundary exists just
proximal to IL-17R delineating the gene-poor and
duplication-rich pericentromeric region from the rest of the q arm of
chromosome 22 (Fig. 2). This boundary also corresponds to the
divergence of conserved synteny on mouse chromosome 6, which is
disrupted at mouse Il-17r, adjacent to the gene
Rbbp2, whose human ortholog maps to the most telomeric known
locus on 12p13.3
(http://www.ncbi.nlm.nih.gov/genemap99/map.cgi?BIN=406&MAP=G3). Bid and Atp6e have been genetically mapped proximal
to a region of conserved synteny with human 12p13.3 (Puech et al. 1997;
Footz et al. 1998). This suggests that the CESCR-homologous interval is
oriented with Il-17r towards the telomeric end of mouse
chromosome 6. The mouse is not expected to possess genes orthologous to
CECR7 and CECR8 because of their high degree of
similarity to other regions of the human genome, suggesting these
duplication-divergence events occurred recently (see below).
|
Further Delineation of the Distal Boundary of the CESCR
Previously, we demonstrated that BID maps telomeric to the
CESCR (Footz et al. 1998) and that at least part of ATP6E is
within the CESCR region (Mears et al. 1995). To refine the boundary of the CESCR, we performed dosage analysis with probes corresponding to
candidate genes in the region. The intensity of fragments on autoradiographs was compared using DNA from patient 25105, whose dicentric r(22) chromosome defines the CESCR; a CES patient (CES01) with a typical supernumerary chromosome that should contain four copies
of the region; and a normal control. A probe from MIL1 exon 9 (Fig. 4) showed the presence of two copies
in patient 25105 and the control, as well as four copies in CES01,
indicating that the 3'-most coding region maps outside the CESCR (data
not shown). This does not eliminate the possibility of the shorter
MIL1 transcript (exons 1-8) being present on the ring
chromosome. However, as the predicted 5' ends of MIL1 and
ATP6E are close (142 bp), this result also suggested that the
CESCR distal boundary may be within ATP6E. Unexpectedly, both
a cDNA probe to ATP6E exons 1-8 and a probe within the first
intron showed an apparent two copies in all three individuals. Two
copies of the cDNA probe were confirmed in a second CES patient known
to contain four copies of the region distal to ATP6E (McDermid
et al. 1996). These results suggested the presence of additional
unprocessed copies of ATP6E in the genome, similar enough in
sequence to produce identical bands on an autoradiograph. This would
interfere with dosage analysis by increasing the overall copy number of
the band in the normal genome, making an increase of one or two copies
in a patient difficult to detect. BLASTN searches of the
ATP6E cDNA against the high-throughput genomic sequence
database (htgs) revealed the presence of apparently processed
pseudogenes of ATP6E, but no unprocessed copies. We
hypothesized that there are unprocessed copies of the region around
ATP6E in areas of the genome not yet sequenced, possibly in
pericentromeric regions of other chromosomes. Therefore, primers from
the first intron of ATP6E were used to amplify products from a
monochromosomal human/rodent hybrid panel. Products of 328 bp were seen
in hybrid cell lines containing human chromosomes 12, 14, and 22 (results not shown). These products were directly sequenced, revealing
a 1-bp difference between the copies: At position 65 the chromosome 22 copy has a C, whereas the chromosome 12 and 14 copies have a G. This
supports our hypothesis that at least part of the region between the 3'
ends of ATP6E and MIL1 has undergone recent
duplication and transposition to elsewhere in the genome. Because these
additional copies cross-hybridize with the gene on chromosome 22, it is
not possible to determine by dosage analysis the exact location of the
25105 breakpoint, which defines the CESCR.
|
Characterization of Genes in the Distal 700 kb
For each gene, GenBank accession numbers and structural information
are given in Table 2, and gene structure is
diagrammed in Figure 4. Expression data is given in Table
3 and illustrated in Figure
5. Primers are given in Table 4 below.
|
|
|
Known Genes
IL-17R
A novel cytokine receptor, IL-17R, was previously mapped to the 22q11.2 region using a radiation hybrid panel (Yao et al. 1997),
but the published mRNA sequence did not define the 3' end of the gene.
Our Northern blot analysis using a probe from the 3' end of the coding
region showed eight different transcripts, suggesting alternative
splicing or polyadenylation. Two clusters of ESTs distal to the coding
sequence suggested alternative 3' ends for IL-17R. This was
confirmed by Northern blot analysis using cDNA probes from each of the
EST clusters (Table 3). Also, RT-PCR using primers
designed from the 3' end of the coding region and the distal EST
cluster (primers IL17F22 and R5) produced a product linking the two, as
expected.
|
). The PipMaker prediction of the genomic organization
of Il-17r demonstrated conservation of exon size and
identified extensive sequence similarities within introns 1, 2, 3, and
11 and in the 3' UTR (Fig. 3).
ATP6E
The
subunit of the vacuolar H+-ATPase originally was
cloned from selected hnRNA from a chromosome 22 somatic cell hybrid (Baud et al. 1994). PipMaker analysis revealed murine sequence conservation restricted to the exons and small regions of
flanking introns, as well as two small regions in the middle of intron
1 (Fig. 3).
Genes Identified Through Exon Amplification and EST Analysis
SLC25A18
SLC25A18 (solute carrier family 25, mitochondrial carrier, member 18) was identified originally by an EST cluster. Northern blot analysis suggested tissue- and development-specific alternative splicing (Table 3; Fig. 5). The predicted ORF shows 43% identity and 53%-54% similarity to mitochondrial transport proteins citrin (SLC25A13, accession no. AF118838) and Aralar 1 and 2 (NP_003696; CAB62206.1). Although the sequence identifies this gene as a member of the mitochondrial carrier protein superfamily preserving the six membrane-spanning domains and the internal tripartite structure (Nelson et al. 1993), the carried solute is unknown. PipMaker analysis shows sequence conservation restricted to the coding exons (4-11) and around exons 1 and 2 in the 5' UTR
(Fig. 3).
CECR1
Details of the structure and expression of CECR1 were published earlier (Riazi et al. 2000). Comparison of human and mouse sequence in this region using PipMaker revealed an absence of sequence conservation in the corresponding position of the mouse
contig (Fig. 3). Similarity was seen around the first exon of
CECR1 (44% identity, 488 bp of human sequence to 737 bp of mouse sequence), but the ORF was not conserved. There was also 63%
identity (904 bp of human sequence vs. 859 bp of mouse sequence) in the
region of the CECR1 third intron. Fragments can be faintly visualized on an autoradiograph from a mouse genomic Southern blot
hybridized with the human exon 1 as probe at reduced stringency (60°C
hybridization and wash), but whether this represents hybridization to
the exon 1 remnant or to Cecr1 in another region of the mouse genome is not clear. It is possible that this gene may not exist in mouse.
CECR2
CECR2, which encodes a bromodomain, was identified by a combination of GENSCAN predictions, EST-identified cDNA sequencing, and 5'/3' RACE. An in-depth analysis of this gene is underway. PipMaker analysis demonstrates sequence conservation of the CECR2 exons of the human and mouse orthologs. It also reveals many regions of sequence conservation between the human and mouse genomes within intron 1, which extends ~106 kb (between 284 and 390k in Fig. 3). It is unlikely that this represents another gene within the intron, as computer analysis of the intron alone shows no ESTs or predicted genes in either orientation. These conserved regions likely represent either alternate 5' ends to the gene or regulatory sequences.CECR4
CECR4 was recognized both through exon amplification and cDNAs identified by analysis of genomic sequence. Sequencing of two cDNAs (321686 and 462605) indicated possible alternative splicing (Fig. 4). The cDNA 321686 encodes part of a possible incomplete ORF of 108 amino acids that shows no similarity to genes identified previously. However, cloning the remainder of this putative gene by 5' RACE and RT-PCR has been unsuccessful, possibly because of the presence of a CpG island covering the most 5'-identified exon. At least 99 bp of overlap exists between CECR4 and the proximally adjacent gene, CECR5, which is transcribed in the opposite direction (Fig. 2). The shared region contains an alternative start codon of CECR5 but is located within the 3' UTR of CECR4. We have considered the possibility that CECR4 is an unprocessed pseudogene, however we have been unable to find evidence of a paralogous second location in the human genome. Southern blot analysis of genomic DNA digested with BglII, BsrGI, EcoRV, HindIII, PvuII, SspI, or SstI showed only one band of the size expected from the sequence of the 22q11.2 region (data not shown), suggesting a single location in the genome or a very recent duplication. Digestion with PstI and TaqI identified RFLPs. Comparison of human and mouse sequence in this region using BLAST2 sequences and the PipMaker program revealed sequence conservation of CECR5, but only exon 1 of CECR4 (Fig. 3).CECR5
The 3' end of CECR5 was identified by a cluster of ~30 ESTs. The transcript was extended by sequencing cDNA 1953625, which contains exons 3-9 and part of exon 2. RT-PCR between the originally predicted first exon (exon 2; primer PSL-F6) and exon 3 (primer PSL-R1) confirmed the presence of a start codon with a partial Kozak consensus sequence (Kozak 1991, 1995), preceded by an in-frame stop codon. RT-PCR
between exon 3 and an exon located 15 kb upstream (identified through
exon amplification; primer EX86-3) revealed an alternate 5' end with
an alternate start codon. Each 5' end is located within or near a CpG
island. The two transcripts would not differ significantly in size. The
predicted proteins of 393 and 423 amino acids show 30% identity and
51% similarity to Schizosaccharomyces pombe CDP-alcohol
phosphatidyl transferase (accession no. Z99295), as well as 40%
identity and 61% similarity to a phosphatidyl synthase-like gene in
Caenorhabditis elegans (accession no. AF125443). Comparison of
human and mouse sequence in this region using BLAST 2 sequences and the PipMaker program revealed sequence conservation around all coding exons except exon 1, as well as interspersed segments within introns 2-5 (Fig. 3).
CECR6
CECR6 was first identified by a cluster of ~30 ESTs. Sequencing of selected cDNAs and RT-PCR products revealed two overlapping predicted ORFs in different frames (209 and 578 amino acids), neither of which shows similarity to any sequences available in public databases. The longer ORF contains a leucine zipper motif. Comparison of human CECR6 sequence with mouse sequence revealed 86% identity and 89% similarity for the predicted proteins. Only the larger, leucine zipper-containing predicted ORF was conserved in the mouse. PipMaker analysis revealed conservation of sequence around the entire CECR6/Cecr6 gene including the coding sequence and 5' and 3' UTRs (Fig. 3).MIL1
MIL1 was identified by numerous EST clusters indicating the presence of at least four alternatively spliced transcripts. Five transcripts were identified by Northern blot analysis using a cDNA (1541822) from the most proximal cluster, which most likely represents the smallest, 1.3-kb transcript. The cDNA 1854295 from the most distal EST cluster hybridizes to the 2.4- and 10-kb transcripts. The 2.4-kb transcript identified by this double-stranded DNA probe represents an alternative transcript of BID, the distally adjacent gene (Footz et al. 1998) (Fig. 2). The transcripts from each gene overlap by
1030 bp, both within the 3' UTRs on opposite strands (data not shown).
No overlapping ESTs have been found for murine Mil1 and
Bid. The human cDNA 663843 is derived from alternative splicing within exon 8 (Fig. 4), producing an extended ORF into exon 9, presumed to be common to all the larger transcripts of MIL1.
The cDNA 663843 represents the 4.0-kb message and cDNA 34798 represents
the 5.0-kb message. Additional EST analysis predicts an additional
2.4-kb transcript (Fig. 4) that uses an alternative 5' end (exon 3) and
an alternative-length terminal exon 8 (data not shown).
Comparison of human and mouse genomic sequence revealed extensive
conservation of the MIL1 locus. PipMaker showed conservation of the coding exons (4-9), as well as regions of homology
in the 3' UTR and in every intron except intron 5 (Fig. 3). The amount
and extent of conservation in noncoding regions of
MIL1/Mil1 is unique compared with the adjacent CESCR
genes, however sequence conservation at the amino acid level is
somewhat lower for this gene. Analysis of the predicted MIL1
amino acid sequence did not reveal an obvious function for this gene.
However, while this study was in progress, an mRNA sequence for this
locus was deposited in GenBank (accession no. AF146568) naming the gene
MIL1 and indicating that it encodes a protein, localized to
the mitochondria, that promotes cell survival. Thus, MIL1
would have the opposite role of the overlapping gene, BID.
Genes Identified by Comparing Human and Mouse Sequence
CECR3
CECR3 originally was identified by exon amplification of a 123-bp clone from PAC 238M15. This putative exon identified a single cDNA from a fetal brain cDNA library. The sequence of this 526-bp cDNA and the surrounding genomic region showed no similarities to known genes or ESTs, nor could a gene be identified by gene or exon prediction programs. We therefore compared the human and mouse sequence of this region (Fig. 3) to predict possible conserved segments. This analysis revealed 12 segments, 32 to 259 bp in size, with sequence identities between 77% and 93%. These conserved regions extended from the putative polyadenylation signal in the 3' UTR of the cDNA to 29,066 bp upstream. RT-PCR between these conserved regions using cDNA reverse transcribed from human lung RNA has confirmed three exons, representing a transcript of ~1.9 kb. No ORF has been identified, suggesting that the reason this gene was not discovered by gene prediction programs is that it codes for a functional RNA product.CECR9
An exon of 199 bp amplified from PAC 238M15 is located between CECR3 and CECR2 (Fig. 2). BLASTN searches of the exon sequence and the genomic sequence surrounding it show no similarity to known genes or ESTs. The exon is 75% identical to the mouse genomic sequence (Fig. 3). The BLAST2 sequences and PipMaker programs identify other regions of mouse/human similarity in this region, suggestive of exons for CECR9 or distant regulatory elements of the adjacent CECR2. GENSCAN predicts the structure of a gene, with moderate scores, which includes this exon (data not shown). The structure of this putative gene is being investigated.Annotation of the Proximal 400 kb
Pericentromeric Repeat Pattern
Figure 6 depicts the duplicated segments found in the proximal 400 kb of the 1.1-Mb contig of 22q11.2. Similarities between chromosome 22 sequences and other chromosomes were discovered through BLASTN queries of the nonredundant (nr) and htgs databases with repeat-masked sequence from the human contig. Throughout the region, similarity to segments from all other chromosomes was found. In many cases similarity is shared between more than two chromosomal locations. For example, a large portion of the IgK pseudogene cluster in the CESCR is shared with 2p11, whereas smaller portions are present on various other chromosomes (3, 5, 6, 7, 8, 10, 11, 12, 15, 16, 17, 18, and 21). Analysis of another region shared with chromosomes 1, 3, 8, 19, Xp22, and Y indicates that it is present in at least three copies on the Y chromosome.
|
) and the IgK pseudogene cluster
(Lotscher et al. 1986, 1988). When the sequence of this region was used
to search dbEST, numerous cDNAs were identified that were similar, but
not identical, to chromosome 22. These presumably represent genes on
other chromosomes that have homology to pericentromeric 22q11.2.
However, two EST clusters showed identity to chromosome 22 rather than
the homologous chromosomal regions and, therefore, represent putative
chromosome 22-specific genes. A preliminary description is given below.
Putative Genes
CECR7
CECR7 was identified originally by two exons (exons 2 and 3, Fig. 4) trapped from a cosmid pool (containing cosmids c1A3, c1H9, and c88G3) mapping to PAC 109L3. cDNA clones 23249 and 1367959 and two additional cDNAs isolated from a CaCo cDNA library were used to elucidate gene structure, although the 5' end of the gene remains to be cloned. Exons 2 and 3 show 86% and 84% identity to exons of the human homolog of a rat kidney-specific gene (accession no. AAC23497). The genomic sequence between exons 2 and 3 shows 86% identity to genomic sequence from chromosome 16 and is similar to sequence on various other chromosomes as well (Fig. 6). Exon 1 is 99% identical to sequence on chromosome 13 and 21. The paralogous region of chromosome 21 is part of the 3' UTR of a GENSCAN-predicted gene. Three alternative 3' ends of CECR7 are associated with three different exons. Exon 6, the 3' end associated with the cDNA 1367959 and one of the CaCo cDNAs, is 84% identical to a region of chromosome 12. The cDNA 1461704 maps to this region of chromosome 12 with 100% identity and also shows 87% identity to the chromosome 22 exon 6. Exons 4 and 5 are rich in repeat sequences. This data indicates that the CECR7 transcripts contain exons originating from several other chromosomes.CECR8
CECR8 was identified by 3' ESTs of two cDNAs (1840041 and 1840206), which showed 98%-100% identity to chromosome 22 when sequenced. Both cDNAs originate from testis cDNA libraries. Two exons have been identified (Fig. 4) that are 83% and 89% identical to sequence on chromosome Y (accession no. AC006040). The cDNA 1840206 may represent an intronless alternative 3' end of the gene. Northern analysis showed a band only in the testis (Fig. 5), which likely represents expression from both the chromosome 22 locus and at least one other locus on the Y chromosome. Sequencing of a second group of dbEST cDNAs (accession nos. AA435549, AI469188, AA394010, and AI798750) showed 98%-100% identity to the chromosome Y locus and only 86%-89% identity to chromosome 22. Additional ESTs similar to CECR8 match unmapped genomic clones. The CECR8-related sequence on the Y chromosome is part of a GENSCAN-predicted gene showing 47% amino acid identity to the LW-1 gene (accession no. AAD45879, L. Wang and S.N. Thibodeau, unpubl.) on the X chromosome.| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
The overexpression of genes, associated with trisomy or partial chromosome duplication, is a common cause of congenital malformations and mental retardation. Gene overexpression is, however, difficult to study at the molecular level because of the large region of the genome usually involved. CES provides an ideal model system for studying the effects of the overexpression of genes during human development, because of the small region of the genome duplicated and the multiple organ systems affected. Fourteen putative genes in and adjacent to the CESCR have now been identified through a combination of exon amplification, EST analysis, gene prediction, and the comparison of human and mouse genomic sequence. Of these 14 genes, only 3 previously were known to be present in this region (IL17R, ATP6E, and BID).
Although a number of these genes were identified initially by exon
amplification, gene prediction and EST analysis were the main methods
used for identification. However, it is striking that the primary means
to establish the presence of CECR3 and CECR9, which
were not represented by any ESTs in the database, was through homology
to mouse sequence. Northern analysis of CECR3 shows it is
strongly expressed in lung, but it is unclear why there are no database
ESTs for the transcript. It is particularly interesting that
CECR3 and CECR9 would have been overlooked by standard computer-based techniques if mouse sequence had not been available, underscoring the utility of such comparative sequence analyses. Comparison of human and mouse genomic DNA has proved to be a
very sensitive method for detecting genes in other regions (Hood et al.
1993; Ansari-Lari et al. 1998; Jang et al. 1999; Lund et al. 2000).
Conservation of noncoding regions can also indicate the presence of
sequences involved in gene regulation (Hardison et al. 1997a,b; Oeltjen
et al. 1997). Numerous stretches of homology in noncoding regions
throughout the CESCR indicate the possible locations of regulatory
regions that may prove important in the future study of these genes.
One goal of this study was to identify genes that would be reasonable
candidates for involvement, either singly or in combination, in the
production of the CES phenotypic features. Although at this point it is
difficult to completely rule out any as candidate genes, some genes are
unlikely to be involved. SLC25A18 is similar to
citrin (OMIM 603859 ), which causes the autosomal recessive disorder citrullinemia when mutated. Thus, based on similarity to
citrin, SLC25A18 is unlikely to be sensitive to
increased gene dose. The similarity of CECR5 to yeast
phosphatidyl synthases suggests that it is an enzyme involved in fatty
acid metabolism, possibly in the production/processing of membrane
phospholipids (Antonsson 1997; Yamashita and Nikawa 1997) and,
therefore, is unlikely to be sensitive to dosage changes. The
testis-restricted expression of CECR8 makes it an unlikely
candidate for CES. Finally, IL-17R, although expressed
ubiquitously (Yao et al. 1997), functions in the immune response, which
is not reported to be disrupted in CES patients.
Our understanding of several of the CESCR genes is limited by technical difficulties, which currently preclude prediction of whether they are reasonable CES candidate genes. It is not yet clear whether CECR4 is a pseudogene with a near-identical copy elsewhere, a spurious transcript from a bidirectional promoter of CECR5, or a gene that overlaps CECR5 in the opposite direction and its exons are resistant to prediction and cloning. The possible candidacy of ATP6E and the shortest transcript of MIL1 is confounded by uncertainty over whether each complete gene is present in the CESCR, as defined by dosage analysis with the r(22) 25105 CES patient. Resolving this uncertainty will require an alternate approach, such as fluorescence in situ hybridization of patient 25105 cells.
We consider CECR1 to be the most suitable CES candidate gene
identified to date (Riazi et al. 2000). This gene is homologous to a
novel family of growth factors first characterized in invertebrates (Sossin et al. 1989; Homma et al. 1996). Sites of expression in a
35-day human embryo include the outflow tract and atrium of the forming
heart, as well as the VII/VIII cranial nerve ganglion, suggesting
potential involvement in the heart and facial defects of CES (Riazi et
al. 2000). The C terminus of the CECR1 protein shows homology
to adenosine deaminase, suggesting that CECR1 may function by
regulating the concentration of extracellular adenosine.
There are several other promising CES candidate genes. CECR2,
with a leucine zipper and a bromodomain (G. Banting and H.E. McDermid,
unpubl.), may be involved in chromatin remodeling (Jeanmougin et al.
1997; Collingwood et al. 1999; Winston and Allis 1999). CECR6
is a single-exon gene encoding a leucine zipper motif, suggesting potential protein-protein interactions exhibited by many transcription factors and gene regulatory proteins (Busch and Sassone-Corsi 1990;
Hagerman 1996). Genes involved in such interactions have the potential
to be sensitive to dosage effects.
BID and the most telomeric coding exon of MIL1 are
located outside the CESCR as defined by the r(22) 25105 CES patient,
however they are present in four copies in cases of CES with a typical CES chromosome. Because the patient that defines the CESCR (25105) died
at 17 days (Mears et al. 1995), further physical and mental development
of this patient could not be determined. It is therefore possible that
overexpression of MIL1 and BID may be involved in physical and mental development of CES patients. Interestingly, the
overlap of the 3' UTR of these two genes suggests the possibility of
antisense regulation (Spencer et al. 1986).
We have established that the mouse genome carries a single linkage
group containing genes orthologous to the CESCR. However, the ortholog
of CECR1, the most promising candidate gene for involvement in
CES features, is absent from our sequenced contig, although it may map
elsewhere in the mouse genome. This precludes straightforward modeling
of CES in mice by manipulation of murine ES cells to engineer a
duplication of the homologous linkage group using the Cre/loxP system
(Ramirez-Solis et al. 1995). Additionally, if CECR7 and
CECR8 are involved in producing the CES phenotype, it would
likely be impossible to model their effect in a mouse, as their
evolution from apparently recent duplications argues against the
existence of rodent orthologs (see below). CECR8 is expressed only in testis by Northern blot analysis and, therefore, is unlikely to
be involved in CES. The expression pattern of CECR7 is
currently unknown. Genes distal to CECR7, however, should be
amenable to testing the effects of overexpression in a mouse model. We
are therefore producing transgenic mice using the BAC and PAC clones of
the human contig to determine whether overexpression leads to
abnormalities similar to those seen in CES.
A unique junction was discovered within the distal half of the CESCR,
between the gene-poor pericentromeric portion and the remaining
gene-rich q arm of chromosome 22. Although a small amount of chromosome
22-specific sequence may be present in the proximal 400 kb of the
distal CESCR, the majority of this region appears to represent
duplications of syntenic and nonsyntenic loci in the genome.
Preliminary analysis indicates this pattern of duplications extends
over an additional 1 Mb to the centromere. Analysis of the
pericentromeric regions of chromosome 10, 16, and others (Eichler et
al. 1996, 1997; Régnier et al. 1997; Ritchie et al. 1998; Jackson et
al. 1999; Horvath et al. 2000) has led to the hypothesis that genomic
duplications are recruited to pericentromeres, possibly serving as a
buffer between heterochromatin and euchromatin (Eichler et al. 1999;
Horvath et al. 2000). Because many of these fragments contain portions
of genes, the pericentromere appears to be a "junkyard" of gene
fragments. However, our discovery of transcribed elements within the
duplication segments also suggests that this region may serve as a
nursery for the evolution of new genes (Eichler et al. 1997; Jackson et
al. 1999; Horvath et al. 2000). Because these duplications appear to be
evolutionarily recent, the creation of new genes in this region would
likely be specific to human and higher primates (Horvath et al. 2000).
The mechanism by which gene duplication recruitment occurs has recently
been speculated to be driven by recombinogenic structures composed of
an array of CAGGG elements and various subtelomeric-like repeats
(Eichler et al. 1999). This is not apparent in the duplication region
examined in this study. Although a cluster of CAGGG repeats was
discovered in the proximal 400 kb of the distal CESCR, its organization
is not reminiscent of the previously described arrays (Eichler et al.
1999) as it does not contain the repeat exclusively on one strand.
Instead, the 2100-bp interval at positions 161205-163304 bp (Fig. 6)
contains 13 CAGGG elements intermixed with 16 CCCTG repeats (i.e., the
reverse complement), and the 2100-bp CAGGG-rich cluster differs from
the "consensus" clusters (Eichler et al. 1999) in that no known
subtelomeric repeats were detected in the vicinity. The 22q duplication
flanked by CAGGG repeats studied by Eichler et al. (1999) is situated
~6-7 Mb from the centromere, within an Ig
cluster (accession no.
D87003/018). Also within the Ig cluster, ~200 kb centromeric to
this duplication, is a stretch of 
-satellite repeats (Kawasaki et
al. 1997). This region may therefore represent an isolated group of
centromere-like repeats, distinct from the duplications directly
adjacent to the centromere. The mechanism for recruiting gene
duplications to these separate regions of 22q may turn out to be distinct.
IL-17R is situated at the boundary between the duplicated segments and unique genes in the CESCR. Close inspection reveals that exon 1 and part of the first intron are actually shared with the pericentromere of 12p11 (accession no. AC010198). As this exon is orthologous with the BAC contig on mouse chromosome 6, it is likely that this portion of 22q11.2 was duplicated and transposed to 12p11. The juxtaposition of Il-17r and Rbbp2 on mouse chromosome 6 suggests that either these genes were separated in the human lineage following the divergence of mouse and man (with IL-17R and RBBP2 being incorporated as the outermost unique genes on their respective human chromosomes, 22qcen and 12pter) or that in the mouse lineage there was a fusion of the linkage groups carrying Il-17r and Rbbp2.
Because the region of pericentromeric repeats on chromosome 22 is unlikely to produce many transcripts, most if not all of the genes associated with CES will be located between IL17R and ATP6E. Further characterization of these genes should shed light on the developmental processes that lead to the CES phenotype upon their overexpression.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Hybrid Cell Lines
Cell lines from the NIGMS human/rodent somatic cell hybrid monochromosomal panel no. 2 (version 2) were grown under recommended conditions and harvested for DNA.
Exon Amplification
Exon amplification was performed using the Gibco BRL Exon Trapping System according to the manufacturer's instructions. Exons were trapped from 2 BACs (95A8, 233A2), 3 PACs (109L3, 238M15, and 120N18), and 27 cosmid clones that cover a portion of the distal CESCR.
Sequencing
Human and mouse genomic clones were sequenced by a random shotgun strategy (protocol available from http://www.genome.ou.edu). PCR products, exons, and cDNAs were sequenced either manually using a Thermo Sequenase radiolabeled terminator cycle sequencing kit (Amersham Life Science) according to the manufacturer's instructions, or using ABI and LI-COR automated sequencers.
Hybridization
Hybridization of exon and cDNA probes to total human or bacterial
clone DNA was carried out as described previously (Mears et al. 1994)
with the following exceptions. DNA probes were purified in 0.7% or
1.0% low-melt agarose (SeaPlaque, FMC); transferred DNA was separated
on 0.7%-2.0% agarose; the first post-hybridization wash consisted of
1.5× SSC/0.2% SDS; and all Southern blot washes ranged from 5-20
min. Human/human or mouse/mouse hybridizations were performed at
65°C, whereas interspecific hybridizations were done at
50°C-55°C with washes no higher than 60°C.
Mouse BAC clones were obtained by screening a library for strain 129SV (Research Genetics). Hybridizations to eight high-density membranes (containing >200,000 double-spotted unique clones in total) were performed in 5× SSC/5× Denhardt's/0.5% SDS with washes as described for Southern analysis above. Commercial cDNA probes were used for the following genes: Il-17r (accession no. W12281), Cecr6 (accession no. AA030766), Mil1 (accession no. AA683716), and Bid (accession no. AA104077). PCR products were used for Cecr3 (1.6-kb mouse genomic fragment with primers M38F + M38R) and Cecr2 (human exons 1 and 2, see Fig. 3; G. Banting and H.E. McDermid, unpubl.).
Northern Hybridizations
Human multiple tissue Northern blots were purchased from Clontech Laboratories and Invitrogen. Each Northern blot contains ~2 µg of poly(A)+ RNA per lane from various different human adult or fetal tissues. Probes were labeled using the Strip-EZ DNA kit (Ambion).
Screening cDNA Libraries
A fetal brain cDNA library (Stratagene) and a cDNA library made from CaCo cell RNA (library kindly provided by Dr. Johanna Rommens, Hospital for Sick Children, Toronto) were screened to identify clones containing exons identified previously through the exon amplification procedure. Plaque lifts were performed using Hybond-N (Amersham Life Science) nylon membranes as per the manufacturer's instructions.
RT-PCR
RT-PCR was performed using the ThermoScript RT-PCR System (Life Technologies) on 2-5 µg of total RNA isolated from various human or mouse tissues. In cases where subsequent PCR was to be performed using primers not flanking intronic sequence, the RNA was pretreated with DNase following the manufacturer's instructions (either DNase I [10 U/µL, Boehringer Mannheim]; DNase I, amplification grade [1 U/µL, Life Technologies]; or DNA-free system [Ambion]). In these situations, a negative control reverse transcription was also performed as above but without the addition of the ThermoScript reverse transcriptase enzyme. Primers used for RT-PCR and PCR are given in Table 4.
PCR
PCR was performed using a PTC-100 programmable thermal controller (MJ Research). The PCR conditions typically consisted of a touchdown procedure: 95°C for 1-2 h (Tminitial, 0.5°C-1.0°C/cycle for 30 min; 72°C for 1 h per kb; 94°C for 30 min) 10 times; (Tmfinal, 30 min; 72°C for 1 h per kb; 95°C for 30 min) 20-30 times; Tmfinal, 30 min; 72°C for 10 h. When PCR was performed on CG-rich templates, the PCRx Enhancer System (Life Technologies) was used according to manufacturer's instructions or with 5% DMSO.
Rapid amplification of cDNA ends (Frohman et al. 1988) was performed
using Marathon cDNA kits (Clontech) and RNA from human testis, brain,
heart, and fetal heart. The kits used were prepared from
poly(A)+ RNA from either human adult heart, testis, or brain,
or fetal brain. A primary RACE was performed using an internal
gene-specific primer and the Marathon Adapter Primer (AP1), followed by
a nested PCR using a nested gene-specific primer and the nested
Marathon Primer (AP2).
Computer Manipulations
GeneTool v1.0 for Macintosh (BioTools) was employed for sequence chromatograph interpretation, in silico gene modeling, CpG island prediction (parameters used a window size of 500 bp and a Y-value > 0.6, %GC > 0.5) and the annotation of genomic sequence used in Figures 2, 5, and 6.
Human and rodent interspersed repeats were detected and masked using RepeatMasker (A.F.A. Smit and P. Green, unpubl. software) at the Web interface http://ftp.genome.washington.edu/cgi-bin/RepeatMasker. Simple sequence repeats and low complexity regions were not masked.
The BLAST algorithm (Altschul et al. 1990)
(http://www.ncbi.nlm.nih.gov/blast/) was used for DNA and protein
similarity searches against the nr, expressed sequence tag (dbEST), and
htgs databases using the default parameters (or without filtering). The
BLAST 2 sequences algorithm (Tatusova and Madden 1999) for
pairwise alignments used the default parameters (or without filtering).
All sequences were repeat-masked before querying a database. At least
one sequence was repeat-masked when compared against another by
BLAST 2 sequences.
Exon and gene structure predictions in repeat-masked sequence were made
using GENSCAN (Burge and Karlin 1997 ) at http://bioweb.pasteur.fr/seqanal/interfaces/GENSCAN.htm,
GRAIL2 at
http://compbio.ornl.gov/Grail-1.3-bin/OrgForm.DoPost, MZEF at http://sciclio.cshl.org/genefinder/, Genie at
http://www.fruitfly.org/seq_tools/genie.html, and
FGENES at
http://dot.imgen.bcm.tmc.edu:9331/seq-search/gene-search.html.
Mouse genomic DNA, corresponding to 453,257 bp of noncontinuous
sequence in four contigs covering the region of Il-17r to Mil1 (Fig. 2), was compared with repeat-masked human sequence (from IL-17R to MIL1, 651,660 bp of continuous
sequence) using the PipMaker program accessed at the web
site http://bio.cse.psu.edu/pipmaker/. Individual gap-free alignments,
ranging in size from 3 to 850 bp, exhibiting >50% identity are
depicted in the percent identity plot (PIP; Schwartz et al. 2000) in
Figure 3.
Genomic clones, paralogous to the pericentromeric portion of the CESCR, being sequenced by the Washington University Genome Sequencing Center, were searched at the web site http://genome.wustl.edu/cgi-bin/ace/ctc_choices/ctc.ace to identify their chromosomal location.
| |
ACKNOWLEDGMENTS |
|---|
We thank Angela Johnson, Dana Shkolny, Lisa Lane, and Heather Wright for technical support. Dr. Roseline Godbout and Paul Grundy provided DNA from some somatic cell hybrid lines. This research was supported by grants from the Medical Research Council of Canada (H.E.M.) and NIH- NHGRI grant no. HG00313 (B.A.R.).
The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.
| |
FOOTNOTES |
|---|
5 These authors contributed equally to this work.
6 Corresponding author.
E-MAIL hmcdermi{at}ualberta.ca; FAX (780) 492-9234.
Article and publication are at www.genome.org/cgi/doi/10.1101/gr.154901.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
-actin or GAPD is shown.
Numbers beside the Northern blots indicate sizes in kb. (He) Heart;
(Br) brain; (Pl) placenta; (Lu) lung; (Li) liver; (Sk) skeletal muscle;
(Ki) kidney; (Pa) pancreas; (Mu) muscle; (Ut) uterus; (Co) colon; (Sm)
small intestine; (Bl) bladder; (St) stomach; (Sp) spleen; (Th) thyroid;
(Pr) prostate; (Te) testis; (Ov) ovary; (Pe) peripheral blood
leukocytes.
