Physical and Genetic Characterization Reveals a Pseudogene, an Evolutionary Junction, and Unstable Loci in Distal Xq28

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

Article

Physical and Genetic Characterization Reveals a Pseudogene, an Evolutionary Junction, and Unstable Loci in Distal Xq28 Swaroop Aradhya,1 Hayley Woffendin,2 Penelope Bonnen,1 Nina S. Heiss,3 Takanori Yamagata1, Teresa Esposito,4 Tiziana Bardaro,4 Annemarie Poustka,3 Michele D’Urso,4 Sue Kenwrick,2 and David L. Nelson1,* 1

Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza 902E, Houston, Texas 77030, USA 2 Department of Medicine, University of Cambridge, Cambridge CB2 2XY, UK 3 Department of Molecular Genome Analysis, Deutsches Krebsforschungszentrum (DKFZ), Im Neuenheimerfeld 290, Heidelberg, Germany 4 International Institute of Genetics and Biophysics (IIGB), Via G. Marconi 10, 80125 Naples, Italy *To whom correspondence and reprint requests should be addressed. Fax: (713) 798-5386. E-mail: [email protected].

A large portion of human Xq28 has been completely characterized but the interval between G6PD and Xqter has remained poorly understood. Because of a lack of stable, high-density clone coverage in this region, we constructed a 1.6-Mb bacterial and P1 artificial chromosome (BAC and PAC, respectively) contig to expedite mapping, structural and evolutionary analysis, and sequencing. The contig helped to reposition previously mismapped genes and to characterize the XAP135 pseudogene near the int22h-2 repeat. BAC clones containing the distal int22h repeats also demonstrated spontaneous rearrangements and sparse coverage, which suggested that they were unstable. Because the int22h repeats are involved in genetic diseases, we examined them in great apes to see if they have always been unstable. Differences in copy number among the apes, due to duplications and deletions, indicated that they have been unstable throughout their evolution. Taking another approach toward understanding the genomic nature of distal Xq28, we examined the homologous mouse region and found an evolutionary junction near the distal int22h loci that separated the human distal Xq28 region into two segments on the mouse X chromosome. Finally, haplotype analysis showed that a segment within Xq28 has resisted excessive interchromosomal exchange through great ape evolution, potentially accounting for the linkage disequilibrium recently reported in this region. Collectively, these data highlight some interesting features of the genomic sequence in Xq28 and will be useful for positional cloning efforts, mouse mutagenesis studies, and further evolutionary analyses. Key Words: int22h, BAC and PAC contig, mouse synteny, great ape evolution, XAP135 pseudogene, phylogenetic, single-nucleotide polymorphism

INTRODUCTION The first physical maps of human Xq28 spanned an interval of approximately 7.5 Mb between IDS and Xqter [1,2]. Transcriptional mapping and microdissection in this region led to identification of at least 50 genes [3–5]. The interval between SLC6A8 and G6PD has the highest gene density and complete sequencing [6] has helped identify multiple disease genes in this interval. The 400-kb Xq pseudoautosomal region (PAR) was also recently sequenced [7], leaving the 1.6-Mb section between G6PD and the XqPAR to be examined. Notably, almost none of the 20 genes in distal Xq28 is associated with a disease or function, except for F8C [8], DKC1 [9],

and NEMO [10]. However, several diseases have been mapped to distal Xq28, including incontinentia pigmenti (IP) [11–13], Waisman syndrome [14], familial skewed X inactivation and spontaneous abortions [15], and several nonspecific mental retardation syndromes [16–25]. In view of this, many genes within distal Xq28 lack functional information, which makes it difficult to evaluate them as candidate genes for diseases. Moreover, despite previous transcriptional mapping efforts, some genes could still be unidentified, as exemplified by the discovery of NEMO just distal to G6PD [10,26]. The previously isolated yeast artificial chromosome (YAC) and cosmid clones between G6PD and Xqter are poor tools for genomic analyses because of their size, thin coverage, and

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved. 0888-7543/01 $35.00

31

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

FIG. 1. BAC/PAC contig between G6PD and Xqter. A larger segment of Xq28 is shown on the top and the region of interest in this report is expanded below. The sequenced region is indicated in red and Xq PAR is marked in yellow. The int22h repeats are boxed in blue and the XAP135/FLI10975C is in green. A few PACs are included in the contig and are marked by blue text; the RP11 BACs are orange and the GS BACs are in black text. Note sparse coverage upstream of int22h-2. The suffix on BBOX2 indicates exon number.

instability. Therefore, to assist positional cloning efforts and other genomic studies, we have constructed a high-density bacterial and P1 artificial chromosome (BAC and PAC, respectively) contig to facilitate refined mapping, structural analysis, and sequencing. Our goal in this work was to characterize the G6PD to Xqter genomic interval (hereafter termed distal Xq28) in various respects to understand its nature and establish the groundwork for identifying linked human disease genes, particularly for IP (MIM 308300). In addition to mapping and sequencing, we paid attention to the evolution of distal Xq28 with respect to specific loci within the region and in relation to other species. This approach was intended to gain insight into how the region was formed and how its current state might relate to human diseases linked to this region. Distal Xq28 contains various repeat sequences, including the three int22h repeats that are particularly interesting because of their sequence, copy number, location, and orientation. Repeat sequences can predispose a genomic region to rearrangements depending on their sequence and structure [27,28]. The int22h repeats share > 99% sequence identity along a length of 9.5 kb [29], and inversions between int22h1 (within intron 22 of F8C) and either of the two repeats distal to F8C (int22h-2 and int22h-3) are the predominant mutations that cause the bleeding disorder hemophilia A [8]. Rearrangements between the int22h-2 and int22h-3 repeats have never been reported, although they lie in the same orientation. It was previously thought that a disease linked to this region, possibly IP, could result from deletions or dupli-

32

cations between int22h-2 and int22h-3. It was also not known when the int22h loci originated or if they have always been unstable. We examined these repeats in various great apes that represent different lineages and evolutionary periods to address that question. Syntenic homology mapping across species is also often useful for describing the evolution of specific genomic regions. Comparison of the human and mouse X chromosomes reveals four major segments of similarity, where segment IV in the top-middle section of the mouse X chromosome corresponds to human Xq28, from HPRT to Xqter [30–33]. Until recently, the mouse F8c (homolog of human F8C) was the most distally mapped locus in segment IV. The locus after F8c was Dmd in segment III; in humans, F8C maps to distal Xq28 and DMD maps to Xp11. Thus, the question was how far mouse syntenic homology extended after F8c and whether the break in synteny corresponded to an interesting position within distal Xq28. In particular, this information would be essential for large-scale genomic deletion projects in the mouse aimed at identifying linked human disease genes, such as IP, through phenotypic modeling. Because Xq28 contains numerous repeat sequences, including int22h, we had entertained the idea that they could affect recombination in this region. Indeed, linkage analysis in a 3-Mb interval between DXS52 and Xqter had failed to identify recombinants among a large number of families with IP [11–13] (S.A., unpublished data). This led us to speculate that the frequency of crossovers was reduced in this interval. A recent report clearly demonstrated extensive linkage

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

FIG. 2. Mapping and structural analysis of XAP135/FLI10975 and its Xq28 pseudogene copy. (A) FISH analysis shows that biotin-labeled BAC GS-476P10, containing the XAP135/FLI10975C, maps to Xq28. (Right) FISH hybridization of biotin-labeled BAC GS-34F24, containing the XAP135/FLI10975 gene, to 6qter and 18p (GS-45J7 showed the same result). Both metaphase spreads were also hybridized with digoxigenin-labeled a-control probes as indicated (Oncor). (B) Structure of the XAP135/FLI10975C (above) is compared with the fulllength copy of XAP135/FLI10975. Exons and introns are color-coded to highlight differences in structure. Various groups of ESTs that match XAP135/FLI10975 are aligned to show the overlap, positions of introns, and presence of inaccurate, intron-containing EST sets A, B, F, and G. (C) PCR analysis on BAC clones with primers designed from the EST consensus sequence yielded larger-thanexpected products, which is indicative of introns (gray boxes). Primer locations are shown on the left. GenBank accession numbers: EST set A, AF055030; EST set B, W90464, W67631, W94664, W94808, R54772; EST set C, AA35340, AA633574, IMAGE1505626, IMAGE1593952, IMAGE1350711; EST set D, AA508475, T64415, X97515; EST set E, AA992270, AI1004847; EST set F, AI003809; EST set G, AA626472; EST set H, AA324890.

A

C

B

disequilibrium (LD) between ALD and NEMO, although the reasons for the LD were not apparent [34]. We have used single-nucleotide polymorphisms (SNPs) in an overlapping but smaller region in an attempt to validate this reported LD and consider it from an evolutionary standpoint. Together, these different approaches have yielded a comprehensive view of the nature and evolution of distal Xq28. The data presented here provide the groundwork for various types of studies, including positional cloning, mouse mutagenesis, genomic evolution, and LD. As this report describes, distal Xq28 has an interesting history through its evolution from lower mammals and great apes, and specific loci within it are characterized by an instability that has predisposed them to cause certain diseases. Individually, the reported findings highlight some familiar themes associated with other genomic locations, thus underscoring the dynamic nature of genomes.

453H20, 126B15, 376G16, 368J9, 461C6, 140O10, 123F5, 517K15, 570J12, and 441B11; Fig. 2). The lengths of all BACs were in the expected range of 80–200 kb. There was relatively sparse coverage between VBP1 and XAP135, with only four clones containing both loci. A few clones also appeared to be prone to rearrangements. Specifically, some preparations of GS-482C22 and GS-433O22 demonstrated partial deletions, suggested by PCR failure for markers XAP136, XAP140, and VBP1 (data not shown). Similarly, some DNA samples of GS461C6 were negative for XAP135 and BBOX2. Lastly, GS491E10 was positive for X10 and markers nearly 300 kb upstream, which is not typical of BAC clone sizes. In addition, we noted unexpected FISH signals for GS-453H20 and GS126B15 in normal human female lymphoblasts. Clone GS453H20 hybridized to chromosomes 2p, 9q, and Xq28, whereas GS-126B15 showed signals on chromosomes 6p, 18q, and Xq28 (data not shown). Finally, another group had

RESULTS

TABLE 1: Exon-intron boundaries for XAP135 (FLI10975) gene on 6qter/18p

Construction of the BAC/PAC Contig, Positioning of Mismapped Loci, and Unstable Clones Four library screens yielded 211 positive BAC and PAC clones, of which 137 (65%) mapped between G6PD and Xqter (Fig. 1). Fingerprinting analysis with EcoRI and BamHI supported the map order and clone overlaps (data not shown). Fluorescence in situ hybridization (FISH) confirmed the Xq28 localization of 16 clones (RP11 clones 211L10, 107C18, 402H20, 476P10, and 95M2; Genome Systems (GS) clones 379A24,

Location

5’ end sequence

.....

3’ end sequence

Exon 1

ATACTATATG

.....

GAGAATAGTG

Intron 1

GTAAGTATTG

.....

TTTGATTTAGa

Exon 2

GCCATCCTTC

.....

ATTCCATCAG

Intron 2

GTAAAGATTA

.....

TTTTTCCTAG

Exon 3

GTCGCTGGAT

.....

TTTTTTTTGC

a

Splice sites are underlined and italicized.

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

33

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

A

FIG. 3. Investigation of int22h repeats in great apes. Southern blot analysis of great ape DNA samples with the F8A probe hybridized to a BclI digest showed that the two chimpanzee and gorilla samples had three copies, whereas the pygmy chimpanzee and orangutan lanes carried only two copies. The three int22h copies in the human samples were as expected. PTR (common chimpanzee, P. troglodytes), PPA (pygmy chimpanzee, P. paniscus), GGO (gorilla, G. gorilla), PPY (orangutan, P. pygmaeus), HSA (human, H. sapiens).

inaccurately mapped three ESTs (XAP127, XAP129, and XAP135), possibly because of unstable physical reagents [4]. The depth of our BAC/PAC contig indicated that XAP127 and XAP129 were located between int22h-2 and int22h-3, where XAP123 also mapped (Fig. 1). XAP135 appeared to be proximal to int22h-2. The locations of all other genes concurred with previously reported data [4]. Identification of the XAP135 Pseudogene The XAP135 locus was first considered a candidate gene for IP. However, initial mapping data indicated that it was a truncated pseudogene. Five GS BACs (34F24, 34L7, 45J7, 77K3, and 99D13) were positive by PCR with primers XAP135–F/R at the 39 end of XAP135 (GenBank acc. no. X97515; partial) but not for Xq28 markers. Homology searching indicated that XAP135 matched a full-length cDNA called FLI10975 (GenBank acc. no. NM_018288). Primers XAP135-3F/4R (Table 5, see supplementary data) from the 59 end of FLI10975 amplified products from a different set of 10 clones, which were also positive for Xq28 markers (Fig. 1). All 5’-positive clones localized to Xq28 by FISH, whereas the two 39-positive clones we tested (GS-34F24 and GS-45J7) showed signals on 6qter and 18p (Fig. 2A). These results indicated that at least two copies of XAP135/FLI10975 were present in the genome and that Xq28 lacked the 39 part of this gene (Fig. 2B). Anchored Alu-PCR with primer XAP135-4R on Xq28+ hybrids as templates yielded a sequence containing multiple missense and nonsense mutations and an Alu repeat toward the 59 end of XAP135/FLI10975 (Fig. 2B). The National Center for Biotechnology Information unfinished high-throughput genomic sequence database had a BAC sequence (GenBank acc. no. AC027048) containing XAP135/FLI10975c, which confirmed our data and established that the Xq28 copy was a partial pseudogene (GenBank acc. no. AF338734) containing only the 59 end of XAP135/FLI10975 with an internal Alu repeat insertion in exon 1. Because PCR data suggested that the 6qter/18p clones contained the entire XAP135/FLI10975 gene, we proceeded to elucidate its gene structure. A BLAST

34

B

FIG. 4. Analysis of homology between distal Xq28 and mouse segment IV. (A) Comparison of human and mouse X chromosomes. Loci marked in red were mapped in the mouse. Expanded regions represent Xq28 and the corresponding mouse interval. Note that VBP1 marked the distal end of the homology between human distal Xq28 and mouse segment IV. Dmd was adjacent to Vbp1 in the mouse but mapped to Xp11 in humans. In addition, Bbox2 was located near the centromere in the mouse but was adjacent to VBP1 in humans. (B) Mapping panel results for Mtcp1 and Bbox2. (Bottom) Mapping the Vbp1 pseudogene to chromosome 3.

search with the XAP135 cDNA identified several expressed sequence tags (ESTs) that overlapped in eight distinct sets with two major gaps (Fig. 2B). Primers across the consensus sequence (Fig. 2B; and Table 4, see supplementary data) yielded various PCR products (Fig. 2C) that validated the EST

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

TABLE 2: SNPs identified in distal Xq28 and ASO probes for haplotype analysis Gene

Accession no.

Base changea

Exon

Amino acid change

NSHDL

U47105

A219G

2

Met9Valb

IRAK

L76191

C587T→

5

Ser196Phe

exon 6+15C‡G

intron 6

none: intronic

FLN1

DNL1L

SEX

GDX

X53416

L40823

L44140

J03589

SNPs for haplotyping

ASO probe (5’ →3’)

T1674C

12

Leu532Ser

IRAK12B

CCG GGC ATT T/C GGA GGC CGC

C835T

4

Pro221Pro

FLN1-835

CAG CAA GCC C/T GTT ACC AAT

A1054G

7

Thr294Thr

C1458T

9

Thr429Metc

FLN1-1458

AGA AGG GCA C/T GGT AGA GCC

FLN1-int47

TGG GCA ACC C/T GGG CCC CCG

SEX-3462

CTA AGG GCG C/A CGT CGG CCG

SEX-int23

TAC CTG CCT C/T GCT CTA TCC

NEMO-165

GTT TCT ACT C/G CTC CCT CCT

d

exon 47+11C→T

intron 47

none: intronic

exon 6–5T→C

intron 5

none: intronic

T1418C

9

none: 3’ UTR

G1421C

9

none: 3’ UTR

C2289T

12

Phe763Phe

G2529A

13

Thr843Thr

G2589C

14/15

Glu863Aspc

G3462T

20

Ala1154Ala

G3723C

21

Ala1241Ala

exon 23+10C→T

intron 23

none intronicd

G563A

4

none: 3’ UTR

G1460A

4

none: 3’ UTR

NEMO

AF091453

C21G

1C

none: 5’ UTR

CLIC2

AJ000217

exon 2–36G→T

intron 1

none: intronic

BBOX2

T.E., A.C.e

G1266T

7

Ile379Metc

a

These base alterations were found by conformation sensitive gel electrophoresis and the numbers refer to position from start of mRNA unless intronic. bPreviously reported in [36]. Not conserved sites. dPreviously reported in [37]. eT.E. and A.C., unpublished data.

c

alignments and revealed consensus splice sites at the exon–intron boundaries (Table 1). Some ESTs also contained partial sections of the full-length cDNA, possibly representing alternatively spliced forms (Fig 2B). Thus, the intact copy of the XAP135/FLI10975 gene was ~ 4.1 kb with three exons (GenBank acc. no. AF338735). Conservation of int22h in Great Apes Southern blot hybridization of a BclI blot with the F8A probe showed the expected three bands in human DNA, where the 21.6-kb band represented int22h-1, located within intron 22 of F8C (Fig. 3). The 16-kb band corresponded to int22h-3 and the 14-kb fragment indicated the presence of int22h-2. The bands in the other great ape DNAs varied in size and number compared with those of the human sample. The common and pygmy chimpanzee samples showed fragments of the same size, except the pygmy chimpanzee DNA lacked the intermediate of three bands. The gorilla sample also exhibited three bands, whereas the orangutan had only the two upper fragments, matching in size those observed in gorilla DNA.

Mouse Syntenic Homology Mapping for Genes in Distal Xq28 Three genes—Vbp1, Mtcp1, and Bbox2—were chosen for mapping in the mouse because they spanned the distal Xq28 interval that lacked evolutionary homology information (Fig. 4). Vbp1 appeared to be near the mouse chromosome 3 centromere at 1 centimorgan (cM), along with Csrp and Lbp1 (Fig. 4B). However, the PCR product we used for mapping contained several base changes compared with the Vbp1 cDNA sequence (data not shown); it became evident that the primers were in different exons and had amplified from a processed, intronless pseudogene on chromosome 3. The functional copy on the X chromosome presumably was not amplified because of the large intron between primer sites. Another group concurrently reported that Vbp1 was located in the expected interval, distal to F8c [35]. We mapped Mtcp1 to the expected location at 30.48 cM between Flna and Dmd, where Mpp1 and Ctsh-rs1 exist. Bbox2 was located close to the centromere at 0.5 cM, near Sybl1, Fsc1, and DXMit26.

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

35

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

FIG. 5. Cladogram demonstrating relationships between Xq28 haplotypes. Two ancestral haployptes (3 and 4) were found in all ethnic groups and have the highest frequencies. Haplotype 4 was also found in other great apes. Six haplotypes were found in multiple ethnic groups and another six were found exclusively in a single group.

with P = 0.20 and haplotype 4 with P = 0.32) had a difference of only one SNP allele between them. Finally, phylogenetic analysis indicated that all haplotypes could be related to each other through mutation without evidence for recombination (Table 3).

DISCUSSION

Haplotype Analysis We used SNP genotype data to construct haplotypes in the ~ 250-kb region between IRAK and NEMO. We identified 22 SNPs in an initial testing group of 10 individuals [10,36,37] (Table 2). Among these SNPs, seven were chosen for haplotyping because of their location and relatively high minorallele frequency in the testing group (Fig. 6 and Table 2). A panel of 94 male DNA samples that included five ethnic groups and four great apes were genotyped by allele-specific oligonucleotide (ASO) hybridization. We found 12 different haplotypes, and the 2 with the highest overall frequencies (3 and 4) were present in all ethnic groups (Table 3). The common chimpanzee and gorilla samples were also positive for haplotype 4. The bonobo chimpanzee haplotype (haplotype 13) was not found in humans, but it differed from haplotype 4 by one allele. Six haplotypes (haplotypes 2–6 and 8) were found in multiple ethnic groups, whereas the remaining six (haplotypes 1, 7, and 9–12) were unique to a specific group. The frequencies of some SNPs (FLN1-4, FLN1-9, and NEMO2) were too low for conclusive evaluation of LD and recombination and therefore are not reported here. Instead, to understand the evolutionary relationships between haplotypes, we constructed a phylogenetic tree using the DNA parsimony method DRAWTREE [38]. The haplotype frequency and sharing information were added to the tree to create Fig. 5. Three major clades were present, with each clade consisting of one or two haplotypes with a frequency of > 0.09. In addition, there were several private haplotypes in some ethnic groups. The two most frequent haplotypes (haplotype 3

36

Our multifaceted approach to understanding the genomic nature of distal Xq28 has provided several interesting insights. We constructed a highdensity BAC/PAC contig in distal Xq28, which helped characterize a pseudogene, reposition incorrectly mapped genes, and detect instability in the distal int22h repeats. Mouse homology mapping revealed an evolutionary junction near the distal int22h repeats, and analyses of great apes suggested that these repeats are unstable and have existed since a common ancestor. Finally, polymorphism analysis showed that an ancestral Xq28 haplotype has been preserved throughout higher primate evolution, offering a potential basis for the previously reported LD in this region. Together, these data provide the foundation for various types of genomic studies in distal Xq28, particularly with respect to human diseases. XAP135/FLI1095c adds to a growing list of pseudogenes in Xq28 [39–41]. This Xq28 copy of XAP135 contains missense and nonsense mutations, lacks introns, and has an Alu insertion within exon 1. The functional copy contains three exons, with two potential alternatively spliced forms that likely account for some of the bands on an initial northern blot [4]. The BAC clones containing the functional XAP135/FLI10975 showed FISH signals on both 6qter and 18p. However, XAP135/FLI10975 may exist in only one of those locations, because the FISH signals could be due to a separate sequence common to both 6qter and 18p. The XAP135/FLI10975 gene contains zinc fingers, which typically are found in proteins involved in transcriptional regulation. Thus, our findings should be useful for functional studies of this gene, particularly if it is associated with a human disease. It is interesting that XAP135/FLI10975c is located near int22h-2, because the instability of this repeat locus may have allowed the pseudogene insertion. The BAC/PAC contig has also provided evidence for the instability of the int22h

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

Article FIG. 6. Summary of findings in distal Xq28. Low-copy repeats are indicated as triangles. Gray arrowheads mark locations of SNPs used in this work. Loci mapped in the mouse are shown in green; gray box and green arrowheads represent location of the evolutionary junction. Blue arrowheads at the bottom delineate mutated regions in hemophilia A and familial skewed X inactivation and male lethality. Location of the NEMO/LAGE2 duplication is also shown. Red segment indicates sequenced interval; yellow segment represents Xq PAR.

repeats. First, clone coverage is very thin near int22h-2, despite multiple library screenings. Second, multiple clones containing the distal int22h repeats have demonstrated rearrangements. Finally, the repeats are involved in hemophilia A [8] and have caused instability in cosmid and YAC clones [4,42]. In view of these findings, one might expect other types of rearrangements between these repeats, especially between int22h-2 and int22h-3, which are aligned in the same direction. One reason such deletions have not been detected could be because they are extremely rare or lethal even in the heterozygous state. Genes located between these repeats may have indispensable functions, and thus it is important that we have accurately mapped them. Interestingly, an X-linked dominant and male-lethal disorder arises from deletions between int22h-1 and int22h-3 [15]. Analysis of IP patients by FISH with BAC clones failed to detect rearrangements between any of the int22h loci (S.A., unpublished data). The int22h repeats seem to have been unstable at least since the advent of the great ape lineage. The mouse has a partial, single copy of int22h [29,32]. It is not known whether Old World monkeys have three copies of int22h, but our analysis has confirmed that orangutans, one of the first great ape lineages, have only two copies. Hence, at least two copies likely existed in the common ancestor to great apes, around 20 million years ago (mya). Because the gorilla has three copies, duplication likely gave rise to the third copy after divergence of and in the lineage leading to present-day gorillas, chimpanzees, and humans. Furthermore, in the later pygmy chimpanzee lineage (around 5 mya), one copy was probably lost, while human and common chimpanzee lineages maintained all three copies. Thus, the great apes have different numbers of int22h copies, possibly due to inherent instability throughout their evolution. This situation is reminiscent of Charcot–Marie–Tooth disease (CMT1A), which is caused by rearrangements between unstable repeats that were duplicated during speciation of the chimpanzee–human lineage [43]. We also recently reported a large duplication involving the NEMO and LAGE2 genes that originated after the gorilla–chimpanzee–human lineage diverged from that of the orangutan [44].

It is interesting that an evolutionary junction exists near the distal int22h copies, although it is not likely related to the instability of the int22h loci. Our earlier question was where does the mouse syntenic homology end in comparison to distal Xq28? The mapping of Bbox2 near the centromere, close to Sybl1 [45] and Spry3 (T.E. and T.B., personal communication), has now established that a break in syntenic homology exists between VBP1 and BBOX2 in humans (Figs. 4 and 5). In short, an evolutionary junction splits human distal Xq28 into two homologous segments on the mouse X chromosome. The precise location of the junction will be found only by mapping mouse homologs of genes between the distal int22h repeats, although none have been identified yet. In this respect, it is important that we have remapped the human genes in this location. Human BBOX2, SYBL1, and HSPRY3 are distal to VBP1, and the latter two are within the Xq PAR. The location of Sybl1 near the mouse X-chromosome centromere previously led to the idea that the original PAR was on the eutherian X chromosome and later transposed to Yqter in primates [45,46]. Our mapping of Bbox2 now suggests that the entire interval between human int22h-3 and SYBL1 originated from the mouse X-chromosome centromeric region and became part of Xq28 as a single unit with a portion later duplicating to Yqter [47]. In this regard, it is interesting that BBOX2 did not become part of the XqPAR in humans, although that is probably because it was not initially transposed to Yqter and consequently remained subject to X inactivation. Distal Xq28 and further upstream regions contain several types of repeats and pseudogenes that could cause instability. Regions rich in repeats or unstable sequences have been associated with elevated LD [48–50] and repeats have been known to influence recombination by providing added homology [51]. These observations may help explain the reported LD in Xq28 [34] (Fig. 5). Despite a small population sampling, our phylogenetic analysis in an overlapping region has revealed a distinct ancestral haplotype (haplotype 4) that appears to have survived stable transmission since the common ancestor of the great apes 15–20 mya. Moreover, the second most frequent haplotype (haplotype 3) appears to have originated from haplotype 4 and propagated at least since the advent of the human lineage. The fact that an

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

37

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

age studies aimed at identifying disease genes in Xq28. Albeit interesting, none of the findings JAPA S-IND reported here was finally relevant to IP, n = 18 n = 18 except for the clones themselves. IP proved to be due to mutations in a gene called – – NEMO [10,53], located at the 59 end of our – 0.06 BAC/PAC contig. Using BAC RP11-211L10 0.22 0.06 from the contig, we quickly elucidated the 0.67 0.39 25-kb structure of NEMO (GenBank acc. no. AJ271718) to screen for mutations. Most – 0.06 genome researchers will agree that the – 0.39 importance of a thoroughly characterized – – physical map cannot be underestimated. 0.06 0.06 Thus, in addition to assisting with the analysis of NEMO, the BAC and PAC clones have – – been indispensable in identifying, mapping, – – and characterizing a second copy – – (DNEMO), which is located about 25 kb downstream and adds to a growing list of 0.06 – interesting loci in distal Xq28 [44]. Sequence generated from these clones will contribute to the final version of the human X-chromosome data to be submitted by the Human Genome Project. These sequence data will also define other interesting aspects of distal Xq28 and greatly accelerate positional cloning, expression analyses, comparative mouse studies, and evolutionary and LD analyses in this dynamic region.

TABLE 3: SNP haplotype analysis in distal Xq28 between IRAK and NEMO Frequencies All samples AFAM n = 90 n = 18

Haplotypes

CAUC n = 18

HISP n = 18

1

C C C C C T G

0.01

–

0.06

–

2

T C C T G C C

0.13

0.11

0.33

0.17

3

T C C C G C C

0.20

0.06

0.33

0.33

4

C C C C G C C

0.32

0.06

0.11

0.39

5

C C C T G C C

0.02

–

0.06

–

6

C C C C C T C

0.14

0.28

0.06

–

7

T C T C G C C

0.01

–

0.06

–

8

T C C C C T C

0.09

0.22

–

0.11

9

C T C C C T C

0.01

0.06

–

–

10

T C C C C C C

0.02

0.11

–

–

11

C C C C C C C

0.02

0.11

–

–

12

T T C C G C C

0.01

–

–

–

Lower primate haplotypes

All samples n=4

PTR n=2

PPA n=1

GGO n=1

4

C C C C G C C

0.75

1.00

13

C T C C G C C

0.25

1.00 1.00

ancestral chromosome is detectable at such a high frequency in all the human ethnic groups and in other great apes indicates that this section of Xq28 has witnessed little recombination, an observation that might account for one cause of the LD observed in Xq28. Several factors together could contribute to a lowered rate of recombination. In view of the high repeat content in Xq28, nondeleterious genomic rearrangements between repeats may cause structural interference so that a normal chromosome fails to pair effectively with its rearranged partner and consequently avoids recombination. For example, the benign FLN1/EMD inversion in Xq28 arises from rearrangement between flanking repeats [52] (Fig. 5). We also recently discovered the NEMO/LAGE2 duplication, distal to the FLN1 and EMD genes, which predisposes to several types of rearrangements [44]. In addition, lethal genomic rearrangements also arise from recombination between misaligned repeats, as exemplified by two disorders linked to distal Xq28 [10,15,53]. These mutations may not only prevent other normal recombination events nearby through negative interference but may fail to survive themselves because of the associated lethality. Finally, Xchromosome hemizygosity in males naturally reduces the recombination rate and there may be additional undefined pressures against recombination, such as the nature of the genomic sequence itself in distal Xq28. At present, these findings help explain why a 3-Mb segment between DXS52 and Xqter has failed to reveal crossovers in IP pedigrees [11–13] (S.A., unpublished data), and they will similarly be relevant to other link-

38

Supplementary data for this article are available on IDEAL (http://www.idealibrary.com).

MATERIALS AND METHODS Genomic library screening and clone characterization. Three sets of high-density filters—Genome Systems (GS) human female BAC library, RP11 human male BAC library, and RP6 human female PAC library—were prehybridized in NaHPO4 buffer. Overgo probes were designed from GenBank-derived EST sequences and used in the hybridization along with end-labeled oligonucleotides (Table 4, see supplementary data). Briefly, overgo oligonucleotide probes are 22mers with an 8-bp overlap so that annealing and extension with Klenow, in the presence of [a-32P]dATP and [a-32P]dCTP isotopes, produces a 36-mer radioactive probe [54]. Labeling and hybridization were done as described [54]. The NEMO cDNA probe was also used but only on the RP11 library. All filters were hybridized overnight and washed to a stringency of 23 SSC/0.1% SDS at 588C. Exposure times ranged from 1 to 4 days. Positive clones were identified with the provided grids. Isolated clones were characterized by PCR with our EST-derived markers or previously published STS/STR markers between G6PD and Xqter (Table 4, see supplementary data). Fingerprinting was carried out by digesting 8–10 mg of miniprep DNA overnight and electrophoresing on a 1% agarose gel at 40 V for 14 hours. The gel was then stained with SYBR Gold dye (Molecular Probes) and visualized on the Molecular Dynamics Fluorimager that contains software to detect bands and size them relative to a 1-kb extended ladder (Gibco). FISH. BAC clones were labeled with biotin or digoxigenin (Oncor). If biotin was used to label the BACs, the control probes were labeled with digoxigenin. Five hundred nanograms of BAC probe were labeled for each FISH experiment, performed as described [55].

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

Mouse syntenic homology mapping. All mouse loci were mapped on The Jackson Laboratory BSS panel ((C57BL/6JEi 3 SPRET/Ei)F1 3 SPRET/Ei). Primers F5 and B3 designed from the mouse Vbp1 cDNA produced a 220-bp product that revealed an SspI restriction site in C57BL/6J only (Table 6, see supplementary data). The PCRs on DNA samples from the BSS panel were digested with SspI to detect the 120- and 100-bp products. For Mtcp1, C57BL/6J showed a BanI restriction site that was absent in Mus spretus in a 349-bp PCR product amplified with primers Mtcp1-6F and -6R (Table 6, see supplementary data). Digestion with BanI produced 200- and 149-bp fragments in C57BL/6J. For Bbox2, primers 1F and 1R were designed across mouse cDNA in exon 1 (Table 6, see supplementary data). The 155-bp PCR product contained a SNP (C in C57BL/6J and T in M. spretus) at position 113 from the 59 end of the product. This polymorphism was tracked in the BSS panel by ASO hybridization (oligonucleotide probes in Table 6, see supplementary data), performed as described [56]. The membranes were washed to a final stringency of 23 SSC at 458C for 30 minutes. Great ape int22h analysis. DNA samples (5 mg) from gorilla (Gorilla gorilla, GGO), orangutan (Pongo pygmaeus, PPY), common chimpanzee (Pan troglodytes, PTR), bonobo chimpanzee (Pan paniscus, PPA), and humans (Homo sapiens, HSA) were digested with 5 ml of BclI (10,000 units/ml) overnight and electrophoresed at 90 V for 6 hours on a 0.8% agarose gel. The digested DNA was transferred onto a nylon membrane and hybridized with the F8A probe (ATCC, catalog number 57203) overnight. The filters were washed to a stringency of 23 SSC/0.1% SDS at 658C and autoradiographed for 2 days. SNP identification and ASO hybridization. SNPs were identified from positional cloning efforts aimed at identifying the gene for IP. Some have been previously reported [10,36,37,53]. The SNPs were identified by conformation sensitive gel electrophoresis (CSGE) of PCR products amplified from 10 IP patient DNAs (see below) with primers designed to amplify exons of listed genes (Table 7, see supplementary data). CSGE was done as described [57]. PCR products that caused CSGE bandshifts were sequenced to identify the causative base alteration or SNP (Table 2). For ASO hybridizations, PCR products were generated from a panel of 94 unrelated male individuals comprising five ethnic groups and three great apes (see below). A unique female IP patient DNA sample was used for each SNP as a positive heterozygote control because the SNP was originally identified in this person. PCR and ASO hybridization were done as described [56] with oligonucleotide probes containing the SNP (Table 2). Human and other great ape DNA samples. Blood samples for IP patients were obtained through an IRB approved protocol and DNA was extracted with conventional salt precipitation protocols. For SNP genotyping, males from five ethnic groups were sampled: African American (n = 18), Caucasian (n = 18), Hispanic (n = 18), Japanese (n = 18), and south Indian (n = 18). The African American, Caucasian, and Hispanic ethnic samples were part of a collection of 941 DNAs purified from anonymous blood donors in community-based blood drives in southeast and central Texas. Japanese DNA samples were obtained from Takanori Yamagata from the Department of Pediatrics, Jichi Medical School, Tochigi, Japan. South Indian DNA samples were obtained from James Fielding Hejtmancik of the National Eye Institute, National Institutes of Health, Bethesda, Maryland. Genomic DNA samples from the great apes genotyped in this study were isolated from lymphoblast cell lines maintained by D.L.N. PCR and sequencing. All PCRs were performed on the Perkin Elmer 9700 thermal cycler in 30-ml reaction mixtures. To place BACs into a contig, either 0.5 ml of culture or 400 ng of miniprep DNA was used as template with 20 mM primers. Template DNA of 100 ng was used for somatic cell hybrids, YACs, or normal human controls to analyze the XAP135 pseudogene. For sequencing, PCR mixtures were purified with a PCR purification kit (Qiagen) and the products were eluted in 20-ml reaction mixtures for dye-terminator fluorescent sequencing by SeqWright, Inc. (Houston, TX). Evolutionary tree analysis. Phylogenetic trees were constructed by using the DNA parsimony algorithm (v. 3.5) and DRAWTREE (v. 3.5) by Joseph Felsenstein, University of Washington.

ACKNOWLEDGMENTS We thank Sally L. Haydel for mapping some of the BAC clones; Zhe Fang and Dimitra Trikka for assisting with ASO hybridizations; Catherine Kashork (Kleberg Cytogenetics Laboratory of Baylor College of Medicine) for help with FISH imaging; and Kerry L.

Article

Wright (Baylor College of Medicine) for editing the manuscript. This work was supported by NIH grants 5 R01 HD35617 and 2 P30 HD24064 to D.L.N. and by Telethon-Italy grant E0927 to M.D. RECEIVED FOR PUBLICATION APRIL 6; ACCEPTED OCTOBER 30, 2001.

REFERENCES 1. Palmieri, G., et al. (1994). YAC contig organization and CpG island analysis in Xq28. Genomics 24: 149–158. 2. Rogner, U. C., et al. (1994). A YAC clone map spanning 7.5 megabases of human chromosome band Xq28. Hum. Mol. Genet. 3: 2137–2146. 3. Bione, S., et al. (1993). Transcriptional organization of a 450-kb region of the human X chromosome in Xq28. Proc. Natl. Acad. Sci. USA 90: 10977–10981. 4. Rogner, U. C., et al. (1996). Transcriptional analysis of the candidate region for incontinentia pigmenti (IP2) in Xq28. Genome Res. 6: 922–934. 5. Yokoi, H., et al. (1994). Isolation of expressed sequences encoded by the human Xq terminal portion using microclone probes generated by laser microdissection. Genomics 20: 404–411. 6. Chen, E. Y., et al. (1996). Long-range sequence analysis in Xq28: thirteen known and six candidate genes in 219.4 kb of high GC DNA between the RCP/GCP and G6PD loci. Hum. Mol. Genet. 5: 659–668. 7. Ciccodicola, A., et al. (2000). Differentially regulated and evolved genes in the fully sequenced Xq/Yq pseudoautosomal region. Hum. Mol. Genet. 9: 395–401. 8. Lakich, D., Kazazian, H. H. Jr., Antonarakis, S. E., and Gitschier, J. (1993). Inversions disrupting the factor VIII gene are a common cause of severe haemophilia A. Nat. Genet. 5: 236–241. 9. Heiss, N. S., et al. (1998). X-linked dyskeratosis congenita is caused by mutations in a highly conserved gene with putative nucleolar functions. Nat. Genet. 19: 32–38. 10. International IP Consortium. (2000). Genomic rearrangement in NEMO impairs NF-kB activation and is a cause of incontinentia pigmenti. Nature 405: 466–472. 11. Sefiani, A., et al. (1989). The gene for incontinentia pigmenti is assigned to Xq28. Genomics 4: 427–429. 12. Sefiani, A., et al. (1991). Linkage relationship between incontinentia pigmenti (IP2) and nine terminal X long arm markers. Hum. Genet. 86: 297–299. 13. Jouet, M., et al. (1997). Linkage analysis in 16 families with incontinentia pigmenti. Eur. J. Hum. Genet. 5: 168–170. 14. Gregg, R. G., Metzenberg, A. B., Hogan, K., Sekhon, G., and Laxova, R. (1991). Waisman syndrome, a human X-linked recessive basal ganglia disorder with mental retardation: localization to Xq27.3-qter. Genomics 9: 701–706. 15. Pegoraro, E., et al. (1997). Familial skewed X inactivation: a molecular trait associated with high spontaneous-abortion rate maps to Xq28. Am. J. Hum. Genet. 61: 160–170. 16. Matsuo, M., et al. (2000). Mother and daughter with 45,X/46,X,r(X)(p22.3q28) and mental retardation: analysis of the X-inactivation patterns. Am. J. Med. Genet. 91: 267–272. 17. Lubs, H., et al. (1999). XLMR syndrome characterized by multiple respiratory infections, hypertelorism, severe CNS deterioration and early death localizes to distal Xq28. Am. J. Med. Genet. 85: 243–248. 18. Armfield, K., et al. (1999). X-linked mental retardation syndrome with short stature, small hands and feet, seizures, cleft palate, and glaucoma is linked to Xq28. Am. J. Med. Genet. 85: 236–242. 19. Gendrot, C., et al. (1999). X-linked nonspecific mental retardation (MRX16) mapping to distal Xq28: linkage study and neuropsychological data in a large family. Am. J. Med. Genet. 83: 411–418. 20. Wolff, D. J., Schwartz, S., Montgomery, T., and Zackowski, J. L. (1998). Random X inactivation in a girl with a balanced t(X;9) and an abnormal phenotype. Am. J. Med. Genet. 77: 401–404. 21. Pai, G. S., et al. (1997). A new X linked recessive syndrome of mental retardation and mild dysmorphism maps to Xq28. J. Med. Genet. 34: 529–534. 22. des Portes, V., et al. (1997). A gene for dominant nonspecific X-linked mental retardation is located in Xq28. Am. J. Hum. Genet. 60: 903–909. 23. Hamel, B. C., et al. (1996). A gene for nonspecific X-linked mental retardation (MRX41) is located in the distal segment of Xq28. Am. J. Med. Genet. 64: 131–133. 24. Lindsay, S., et al. (1996). PPM-X: a new X-linked mental retardation syndrome with psychosis, pyramidal signs, and macroorchidism maps to Xq28. Am. J. Hum. Genet. 58: 1120–1126. 25. Auricchio, A., et al. (1996). The locus for a novel syndromic form of neuronal intestinal pseudoobstruction maps to Xq28. Am. J. Hum. Genet. 58: 743–748. 26. Jin, D. Y., and Jeang, K. T. (1999). Isolation of full-length cDNA and chromosomal localization of human NF- kB modulator NEMO to Xq28. J. Biomed. Sci. 6: 115–120. 27. Lupski, J. R. (1998). Genomic disorders: structural features of the genome can lead to DNA rearrangements and human disease traits. Trends Genet. 14: 417–422. 28. Ji, Y., Eichler, E. E., Schwartz, S., and Nicholls, R. D. (2000). Structure of chromosomal duplicons and their role in mediating human genomic disorders. Genome Res. 10: 597–610. 29. Naylor, J. A., et al. (1995). Investigation of the factor VIII intron 22 repeated region (int22h) and the associated inversion junctions. Hum. Mol. Genet. 4: 1217–1224. 30. Davisson, M. T. (1987). X-linked genetic homologies between mouse and man. Genomics 1: 213–227.

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

39

Article

doi:10.1006/geno.2001.6680, available online at http://www.idealibrary.com on IDEAL

31. Angel, T. A., et al. (1993). Genetic mapping of the X-linked dominant mutations striated (Str) and bare patches (Bpa) to a 600-kb region of the mouse X chromosome: implications for mapping human disorders in Xq28. Mamm. Genome 4: 171–176. 32. Levin, M. L., et al. (1996). A comparative transcription map of the murine bare patches (Bpa) and striated (Str) critical regions and human Xq28. Genome Res. 6: 465–477. 33. Mallon, A. M., et al. (2000). Comparative genome sequence analysis of the Bpa/Str region in mouse and man. Genome Res. 10: 758–775. 34. Taillon-Miller, P., et al. (2000). Juxtaposed regions of extensive and minimal linkage disequilibrium in human Xq25 and Xq28. Nat. Genet. 25: 324–328. 35. Hemberger, M., et al. (1999). Expression of the von Hippel-Lindau-binding protein-1 (Vbp1) in fetal and adult mouse tissues. Hum. Mol. Genet. 8: 229–236. 36. Aradhya, S., et al. (2000). Human homologue of the murine bare patches/striated gene is not mutated in incontinentia pigmenti type 2. Am. J. Med. Genet. 91: 241–244. 37. Aradhya, S., et al. (2000). Filamin (FLN1), plexin (SEX), major palmitoylated protein p55 (MPP1), and von Hippel-Lindau binding protein (VBP1) are not involved in incontinentia pigmenti type 2. Am. J. Med. Genet. 94: 79–84. 38. Golding, B., and Felsenstein, J. (1990). A maximum likelihood approach to the detection of selection from a phylogeny. J. Mol. Evol. 31: 511–523. 39. Das, S., and Gitschier, J. (1994). Identification of a TXREB pseudogene (TXREBP) located between the genes for p55 (MPP1) and G6PD on Xq28. Genomics 21: 275–278. 40. Maric, S. C., Crozat, A., Louhimo, J., Knuutila, S., and Janne, O. A. (1995). The human Sadenosylmethionine decarboxylase gene: nucleotide sequence of a pseudogene and chromosomal localization of the active gene (AMD1) and the pseudogene (AMD2). Cytogenet. Cell Genet. 70: 195–199. 41. Timms, K. M., et al. (1995). 130 kb of DNA sequence reveals two new genes and a regional duplication distal to the human iduronate-2-sulfate sulfatase locus. Genome Res. 5: 71–78. 42. Freije, D., and Schlessinger, D. (1992). A 1.6-Mb contig of yeast artificial chromosomes around the human factor VIII gene reveals three regions homologous to probes for the DXS115 locus and two for the DXYS64 locus. Am. J. Hum. Genet. 51: 66–80. 43. Boerkoel, C. F., Inoue, K., Reiter, L. T., Warner, L. E., and Lupski, J. R. (1999). Molecular mechanisms for CMT1A duplication and HNPP deletion. Ann. NY Acad. Sci. 883: 22–35. 44. Aradhya, S., et al. (2001). Multiple pathogenic and benign rearrangements occur at a 35kb duplication involving the NEMO and LAGE2 genes. Hum. Mol. Genet. 10: 2557–2567. 45. D’Esposito, M., et al. (1997). Differential expression pattern of XqPAR-linked genes SYBL1

40

46. 47.

48. 49. 50.

51. 52. 53.

54. 55.

56.

57.

and IL9R correlates with the structure and evolution of the region. Hum. Mol. Genet. 6: 1917–1923. Graves, J. A. M., Wakefield, M. J., and Toder, R. (1998). The origin and evolution of the pseudoautosomal regions of human sex chromosomes. Hum. Mol. Genet. 7: 1991–1996. Kvaloy, K., Galvagni, F., and Brown, W. R. (1994). The sequence organization of the long arm pseudoautosomal region of the human sex chromosomes. Hum. Mol. Genet. 3: 771–778. Chakravarti, A., et al. (1984). Nonuniform recombination within the human b-globin gene cluster. Am. J. Hum. Genet. 36: 1239–1258. Lichten, M., and Goldman, A. S. (1995). Meiotic recombination hotspots. Annu. Rev. Genet. 29: 423–444. Liu, X., and Barker, D. F. (1999). Evidence for effective suppression of recombination in the chromosome 17q21 segment spanning RNU2-BRCA1. Am. J. Hum. Genet. 64: 1427–1439. Shaffer, L. G., and Lupski, J. R. (2000). Molecular mechanisms for constitutional chromosomal rearrangements in humans. Annu. Rev. Genet. 34: 297–329. Small, K., Iber, J., and Warren, S. T. (1997). Emerin deletion reveals a common X-chromosome inversion mediated by inverted repeats. Nat. Genet. 16: 96–99. Aradhya, S., et al. (2001). A recurrent deletion in the ubiquitously expressed NEMO (IKKg) gene accounts for the vast majority of incontinentia pigmenti mutations. Hum. Mol. Genet. 10: 2171–2179. McPherson, J. D. (1999). Genome Analysis: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Page, S. L., and Shaffer, L. G. (1998). Chromosome stability is maintained by short intercentromeric distance in functionally dicentric human Robertsonian translocations. Chromosome Res. 6: 115–122. DeMarchi, J. M., Richards, C. S., Fenwick, R. G., Pace, R., and Beaudet, A. L. (1994). A robotics-assisted procedure for large scale cystic fibrosis mutation analysis. Hum. Mutat. 4: 281–290. Ganguly, A., Rock, M. J., and Prockop, D. J. (1993). Conformation-sensitive gel electrophoresis for rapid detection of single-base differences in double-stranded PCR products and DNA fragments: evidence for solvent-induced bends in DNA heteroduplexes. Proc. Natl. Acad. Sci. USA 90: 10325–10329.

GENOMICS Vol. 79, Number 1, January 2002 Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.