Proc. Nadl. Acad. Sci.

USA
Vol. 88, pp. 9051-9055, October 1991
Genetics

Origin of human chromosome 2: An ancestral

telomere-telomere fusion
J. W. IJDO*t, A. BALDINIt§, D. C. WARDt, S. T. REEDERS**, AND R. A. WELLS*¶
*Howard Hughes Medical Institute and tDepartment of Genetics, Yale University School of Medicine, New Haven, CT 06510
Communicated by Alan Garen, July 8, 1991

ABSTRACT We have identified two allelic genomic a human genomic 300-base-pair (bp) Alu I fragment present
cosmids from human chromosome 2, c8.1 and c29B, each at the subtelomeric regions offive different chromosomes, as
containing two inverted arrays of the vertebrate telomeric well as at a single nonterminal locus at 2q13, described
repeat in a head-to-head arrangement, 5'(TTAGGG),,- elsewhere (8). Sixty positives were screened subsequently
(CCCTAA),,3'. Sequences fln g this telomeric repeat are with a (TTAGGG)" probe (n varies between 10 and 1000).
characteristic of present-day human pretelomeres. BAL-31 Fifteen cosmids hybridized to both pSC4 and (TTAGGG),.
nuclease experiments with yeast artificial chromosome clones Cosmid Mapping to Chromosome 2. Probe pSC4 detects
of human telomeres and fluorescence in situ hybridization five different size fragments on a Pst I digest of total human
reveal that sequences flanking these inverted repeats hybridize DNA representing the different chromosomal loci. Of these,
both to band 2q13 and to different, but overlapping, subsets of a 2.5-kb Pst I fragment was assigned to chromosome 2 by
human chromosome ends. We conclude that the locus cloned in means of hybridization of pSC4 to 34 somatic cell hybrid lines
cosmids c8.1 and c29B is the relic of an ancient telomere- (obtained, in part, from BIOS, New Haven, CT). We iden-
telomere fusion and marks the point at which two ancestral ape tified two cosmids, c8.1 and c29B, containing both
chromosomes fused to give rise to human chromosome 2. (TTAGGG), and the same 2.5-kb fragment detected by pSC4
that maps consistently to chromosome 2 only and, therefore,
Similarities in chromosome banding patterns and hybridiza- these two cosmids must originate from chromosome 2 (data
tion homologies between ape and human chromosomes sug- not shown).
gest that human chromosome 2 arose out of the fusion of two Restriction Mapping of Genomic Cosmids. Restriction maps
ancestral ape chromosomes (1-3). Molecular data show ev- of genomic cosmids c8.1 and c29B were constructed by using
idence that this event must have occurred only a few million partial digests ofcosmids hybridized to kinase-labeled T3 and
years ago (refs. 4 and 5 and the references therein). Although T7 primers, as well as cosmid double digests probed with
the precise nature of this putative fusion is unknown, cyto- subclone inserts.
genetic data point to either a centromeric or telomeric fusion DNA Sequencing. DNA sequencing was done by using the
in the vicinity of region 2ql (1, 2, and 6). The observation that dideoxynucleotide chain-reaction procedure. Subclones with
telomeric DNA is present in chromosomal band q13 suggests suitable insert sizes were generated from clone c8.1. Se-
that telomeres, the extreme ends of chromosomes, may have quences were determined either by sequencing both DNA
been involved in this fusion (7, 8). Normally, telomeres form strands or by sequencing the same DNA strand twice.
a dynamic buffer against loss of internal sequence and Fluorescence In Situ Hybridization. Standard metaphase
prevent chromosomes from fusing (for review, see ref. 9). By spreads were prepared from cultured phytohemagglutinin-
contrast, nontelomeric DNA ends are subject to degradation stimulated peripheral blood lymphocytes. Six unrelated in-
by nucleases and to fusion by ligation (10, 11). dividuals were studied. Chromosome preparations were hy-
The termini of human chromosomes consist of head-to-tail bridized in situ with probes biotinylated by nick translation,
tandem arrays of TTAGGG, running 5'--3' toward the end of under suppression conditions, essentially as described by
the chromosome, with average lengths of 5-10 kilobases (kb) Lichter et al. (18). The hybridization was done at 37°C in 2 X
in somatic cells (7, 12, 13). The proximal ends of these arrays standard saline citrate (SSC)/50o (vol/vol) formamide/10o
contain degenerate forms of this repeat, such as (TTGGGG), (wt/vol) dextran sulfate/DNase I (1.5 mg/ml)-cut human
and (TGAGGG)J, (14). Sequences adjacent to these simple genomic DNA (average size, 300-600 bp)/biotinylated probe
repeats have been characterized in a number of human at 3 ,ug/ml/sonicated salmon sperm DNA at 1 mg/ml. The
chromosomes and shown to consist of repetitive elements, probe was denatured in the hybridization mixture at 75°C for
each shared by a subset of all chromosomes (13, 15-17). In 10 min and annealed at 37°C for 20 min. Posthybridization
addition, stretches of telomeric repeats are present at inter- washing was at 42°C in 2x SSC/50%o formamide followed by
stitial sites, usually in subtelomeric regions but also at a three washes in 0.5x SSC at 60°C. Chromosome identifica-
distinct internal site within band 2q13 (8). We describe here tion was based on in situ hybridization banding produced by
the architecture of the sequence at this internal locus at adding heat-denatured digoxigenin-11-dUTP (Boehringer
2q13, which represents a relic of the fusion of two ancestral Mannheim)-labeled Alu-PCR products in the hybridization
ape chromosomes in the evolution ofhuman chromosome 2. 11 mixture (=2 ,ug/ml after the reannealing step). This technique
produces an R-banding pattern suitable for gene mapping
studies (19). Biotin-labeled DNA was detected with fluores-
MATERIALS AND METHODS cein isothiocyanate-conjugated avidin DCS (5 pkg/ml) (Vec-
Library Screening. Approximately 1.4 x 106 colonies from tor Laboratories). Digoxigenin-labeled DNA was detected by
a human genomic cosmid library containing Mbo I partial
digestion fragments of 35-41 kb in vector pWE15, propa- tTo whom reprint requests should be addressed.
gated in host NM554 (Stratagene), were screened with pSC4, §Present address: Imperial Cancer Research Fund, London, United
Kingdom.
Present address: Department of Internal Medicine, The Toronto
The publication costs of this article were defrayed in part by page charge Hospital, Toronto, Ontario, Canada.
payment. This article must therefore be hereby marked "advertisement" IThe DNA sequence reported in this paper has been deposited in the
in accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession no. M73018).
9051
9052 Genetics: IJdo et al. Proc. Natl. Acad. Sci. USA 88 (1991)
using a rhodamine-conjugated anti-digoxigenin antibody ing the telomere-like repeats from band 2q13. A human
(Boehringer Mannheim). Ten metaphase spreads were ana- genomic cosmid library was screened with a (TTAGGG)"
lyzed per experiment. probe and also with pSC4 (8). Cosmids containing both
(TTAGGG)" and pSC4 were assigned to chromosomes by
RESULTS means of hybridization to 34 somatic cell hybrid lines (data
not shown). We identified two cosmids, c8.1 and c29B, that
Identiflication and Characterization of Genomic Cosmids map consistently to chromosome 2, according to the specific
Mapping to Band 2q13. To investigate the architecture of this length of the pSC4-hybridizing fragment they contain, which
putative fusion point, we isolated genomic cosmids contain- is only shared with chromosome 2-containing human-rodent

A 99H HH H

No8. 113b pSC4 612 M7 -Ap613

41--
V4.8

FRAGMENTA FRAGMENT S

c29B 1**j*
Nos.i 11 p6SC4 1RV
_JvB1f LRiI
79BVj H
W7 612 B19 617 Ap.613 HI
V4.3 R - kb

-. SEQUENCED

I4NVERTED REPEAT
B
1 GTGCCCCGGC GCCACGAGGG CGCTGGCGAC CACTGTAAGC AAGAGAGCC: -'TriCrCCTCh7R T (C!(CTIrcfi rrnCrCIZGCX
81 GGCCGGCCGC CTTTGCGATG GCGGAGTTCC GTTCTCCTCA GCAAGACCC GGAGAGCACC GCAGGGCGAC CTGCGTTG2TC
161 TCTCCACAA TGTT AC¶C.CCAM C TTCAC CM M CC Cx&TCAA
241 CGCGAGGCGA GCTGQQTTCT GCTCAGCACA GACCTGGGGG TCACCGTAAA GATGGAGCAG CATTCCCCTA AGCACAGAGG
321 TTGGGGCCAC TGCCTGGCTT TGTGACAACT CGGGGCGCAT CAACGGTGAA TAAAATCT T CCCGGTTGCA GCCGTGAATA
401 ATCAAGGTCA GAGACCAGTT AGAGCGGTTC AGTGCGGAAA ACGGGAAAGA AAAAGCCCCT CTGAATCCTG GGCAGCGAGA
481 TTATCCCAAA GCAAGGCGAG GGGCTGCATT GCAGGG

517 TGAGGG TGAGGG TGAGGG

TGAGGG TTAGGG TTTGGG TTGGGG
TTGGGG TTGGGG TTGGGG
577 TAGGG TTGGGG TTGGGG
TITGGG TTAGGG TTAGGGG TAGGG
TAGGGG TCAGGG TCAGGG
636 TCAGGG TTAGGG TTTTAGGG TTAGGG TTAGGG TTAAGG TrTGGGG TTGGGG TTGGGG TTGGGG
699 TTAGGGG TTAGGGG TTAGGGG TTAGGG TTGGGG TTGGGGG TTGGGG TTGGGG TTAGGGG TAGGGG
764 TAGGGG TAGGG TTAGGG TTAGGG TTAGGG TAAGGG TTAAGGG TTGGGG TTGGGG TTGGGG
824 TTAGGG TTAGGGG TTAGGG TTAG CTAA CCCTAA CCCTAA CCCCTAA CCCCTAA CCCCAA
883 CCCAAA CCCCAA CCCCAA CCCCAA CCCTA CCCCTA CCCCTAA CCCCAA CCCTTAA CCCTTAA
945 CCCTTAA CCCTTA CCCTAA CCCTAA CCCAAA CCCTAA CCCTAA CCCTA CCCTAA CCCAA
1004 CCCTAA CCCTAA CCCTA CCCTAA GCCTAAAA CCCTAAAA CCGTGA CCCTGA CCTTGA CCCTGA
1067 CCCTTAA CCCTTAA CCCTTAA CCCTAA CCCTAA CCATAA CCCTAAA CCCTAA CCCTAAA CCCTAA
1132 CCCTA CCCTAA CCCCAA CCCCTAA CCCTAA CCCCTATA CCCTAA CCCTAA CCCTA CCCCTA
1193 CCCCTAA CCCCAA CCCCAG CCCCAA CCCCAA CCCTTA CCCTAA CCCTA CCTAA CCCTTAA
1253 CCCTAA CCCCTAA CCCTAA CCCCTAA CCCTA CCCCAA CCCCAAA CCCAA CCCTAA CCCAA
1313 CCCTAA CCCAA CCCTAA CCCCTA CCCTAA CCCCTAA CCCTAA CCCCTA CCCTAA CCCCTAA
1374 CCCTAA CCCCTA CCCTAA CCCCTAA CCCTAG CCCTAG CCCTAA CCCTAA CCCTCA CCCTAA
1435 CCCTCA CCCTAA CCCTCA CCCTCA CCCTCA CCCTCA CCCTAA CCCAA

1482 CGTCTGTGC TGAGAAGAAT GCTCGTCCGC CTTTAAGGTG CCCCGTCTGTGCTCAA QCAGAGCAC GTCCGCCGTC

1561 CATCCCT rACCCCCCT CTCACTC Ara rATrCTr CT(!!rcCCTTc CCAATACCC Cr-AAGTCTCT QCCAC;AfCA-QA
1641 ACGCAGCTCC GCCCTCGCGA TGCTCTITCGG CTGTGTGCTA AAGAGAACGC AACTCCGCCC TCGCAAAGGC GGCGCCGCCG
1721 CCCAGCCGG AG CGCCGCCCGCCCGAG fACQGQA2AfG fGZGfaC2CfaC faGQACCCiG AGAG2C2CfaG QfiCfaC2GjG
1801 raCQia fG fGfiG~fCQG fAQ(XCfaG GAGraCfaCraG G.CfC2GAG rC G WXGQGCrCG C

FIG. 1. (A) Restriction maps of cosmids c8.1 and c29B indicate that they are allelic. Thick bars indicate positions of subclones that were
used in hybridization studies to show that the two cosmids contain homologous sequences in the same order. B, BamHI; Bg, Bgi H; H, HindIII;
R, EcoRI; and V, EcoRV. (B) DNA sequence of region indicated in A shows degenerate head-to-head arrays of a tandem repeat having the
consensus TTAGGG. The underlined flanking sequences show 80% identity over 269 bp (when inverted) and were 95% (left of telomere repeat)
and 90%o (right of telomere repeat) identical to sequences found adjacent to the telomeric tandem repeats in clones pTH14 and TelSau2.0,
respectively (13, 17). The dashed line indicates sequence consisting of a 25-bp repeat unit with an 88% G + C content.
 Genetics: Udo et al. Proc. Natl. Acad. Sci. USA 88 (1991) 9053
A B The inverted arrangement of the 1TAGGG array and the
adjacent sequences, which are similar to sequences found at
0 60 120 180 240 0 30 60 120
S 80 240 present-day human telomeres, is precisely that predicted for
a head-to-head telomeric fusion of two chromosomes. Alter-
natively, a small duplication and inversion, which could have
arisen by chance, might account for this structure. To dis-
tinguish between these possibilities we isolated subclones 817
and Apa813, which flank the telomeric repeat in clone c8.1
An.~~~~~~~~ but lie outside the inverted-repeat region demonstrated by
sequencing (Fig. 1B). These flanking subelones detect BAL-
31-sensitive bands, respectively, in HTY243 and HTY275,
two independently isolated yeast artificial chromosome
clones that contain different human telomeres (Fig. 2) (20).
These data provide strong evidence that the inverted repeats
in c8.1 arose from the head-to-head fusion of ancestral
telomeres.
C tion f the Subteomeric Origin of the Squences
Flanking the Teloineric Array. Confirmation that sequence
blocks from either side of the 2q13 telomere repeat are similar
to sequences at human subtelomeres was obtained by fluo-
rescence in situ hybridization experiments. Fragments A and
B (Fig. 1A) from either side of the telomere repeat in c8.1,
used as probes in fluorescence in situ hybridizations of
metaphase chromosomes, showed a different, but overlap-
FIG. 2. Sensitivity of the sequences hybridizing to probe 817 in ping, chromosomal distribution (Fig. 3 A and B). Both
HTY243 (A) and to probe Apa813 in HTY275 (B) to digestion with fragments hybridized to most telomeric bands of several
BAL-31 nuclease. High-molecular-weight DNA isolated from chromosomes, as well as to the interstitial band 2q13 (Fig. 4
HTY243 and HTY275 was treated with BAL-31 nuclease for the A and B). An additional interstitial hybridization signal was
indicated time in min at 30"C. The DNA was then cleaved with Pst observed with fragment B at 3p14 in 5 ofthe 20 chromosomes
I (HTY243) or EcoRI (HTY275) and, after transfer to membrane, analyzed. We also observed a wide range of signal intensity
hybridized to probes 817 and Apa813, respectively. We have attrib- between different chromosomes and between homologous
uted the focusing of the Apa813-hybridizing fragments observed at chromosomes, especially with fragment B. In some cases
120 min to a decreased BAL-31 digestion rate when the enzyme only one homologue was labeled in each of the 10 metaphases
reaches the C + G-rich area (21), marked with a dashed line in Fig.
1B. studied (i.e., the telomeric region on chromosome 6p with
fragment A). The pattern of signal was consistent from cell to
hybrids. Fluorescence in situ hybridization experiments lo- cell within an individual but not from individual to individual
calized these cosmids to band 2q13. Restriction analysis and (e.g., in an unrelated individual, fragment A hybridized to
studies of eight subclones of these cosmids revealed that they only one copy of chromosome 20q) (data not shown). These
are allelic (Fig. 1A). Both cosmids contain arrays of a
data, along with the observations of others (17, 22), suggest
that the terminal regions ofhuman chromosomes are dynamic
telomere-like repeat. By sequence analysis, the telomere- structures, from which stretches of sequence are gained and
hybridizing region in clone c8.1 was found to consist of two lost at a relatively high frequency.
degenerate arrays of TTAGGG in an inverted (head-to-head)
arrangement (Fig. 1B). The flanking sequences are also
arranged in an inverted fashion and show 80% identity over DISCUSSION
269 bp and are also 95% and 90%o identical with sequences We have isolated two allelic genomic cosmids that were
found adjacent to cloned "true" telomeres, pTH14 and localized to chromosome 2, each containing two arrays of
TelSau2.0, respectively (13, 17). telomeric repeat TTAGGG in an inverted arrangement.
A 20-

1 0-

o -rII
I
1
,1
2
, i rrl
3 4
irT-
5
Ir ,
6
---
7
-rT-ir
8
i
9
r-it
10
, 11 I
12
lmvI
rr14
13
I
lrr-
15 16 17
rTIrTil^rlflT-
18 19 20
21 22
X
, ITII
Y

B
10L
20 -

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X Y

FIG. 3. Hybridization signals obtained with fragments A and B from cosmid c8.1. Ten banded metaphase spreads (20 chromosomes) were
examined with each probe. The number of chromosomes hybridizing were summated for each point at which signal was observed. Both fragments
hybridize to the most telomeric bands of several chromosomes (at 400-band resolution) as well as region 2q13 (confirmed at 800-band resolution)
and show a different but overlapping chromosomal distribution. An additional interstitial hybridization was seen with fragment B at region 3p14
in 5 of 20 chromosomes. Note that, in some cases, only one telomere was labeled in each metaphase spread. For example, only one 6p terminus
hybridized with fragment A in this individual; hybridization of both 6p telomeres was never observed (sensitivity of the technique is such that
there was failure to detect hybridization above background in 3 of 10 spreads). Likewise, one homologue was labeled on lq, 3q, 8p, 10p, 12p,
13q, and 16p with fragment B.
9054 Genetics: Udo et aL Proc. Natl. Acad. Sci. USA 88 (1991)

FIG. 4. Fluorescence in situ hybridization of fragment A (a) and fragment B (b) to metaphase chromosomes. Metaphase spreads were
prepared from cultured lymphocytes and simultaneously hybridized with a biotinylated A or B probe (shown in red) and digoxigenin-labeled
Alu-PCR products (shown in green) that generate an R-banding pattern, according to published protocols (19).
Flanking sequences are characteristic for the preterminal repeat, as described here, might be expected to facilitate the
regions of human chromosomes. The data we present here formation of secondary, cruciform structures. The ability of
demonstrate that a telomere-to-telomere fusion of ancestral inverted-repeat sequences to form cruciform structures has
chromosomes occurred, leaving a pathognomonic relic at been demonstrated in vitro (30, 31). Because of the formal
band 2q13. This fusion accounts for the reduction of 24 pairs analogy between cruciform structures and Holliday junc-
of chromosomes in the great apes (chimpanzee, orangutan, tions, both are subject to site-specific cleavage and, hence,
and gorilla) to 23 in modern human and must, therefore, have resolution by single-strand-specific nucleases (32, 33). We
been a relatively recent event. Comparative cytogenetic suggest such a phenomenon could have resulted in progres-
studies in mammalian species indicate that Robertsonian sive shortening of this inverted sequence until relative sta-
changes have played a major role in karyotype evolution (23, bility was reached.
24). This study demonstrates that telomere-telomere fusion, The cosmid clones described here will allow testing of the
rather than translocation after chromosome breakage, is hypothesized association between the telomere-like se-
responsible for the evolution of human chromosome 2 from quence at region 2q13 and the rare folate-sensitive fragile site
ancestral ape chromosomes.
Fusion of telomeres is a rare occurrence in normal lym- (FRA2B), which also maps to band 2q13 (7, 8, 34). Although
phoblasts and fibroblasts, although it has been observed in the mechanism underlying chromosome fragility has not been
20-30% of the cells of certain tumors, where it appears to be determined, it is clear that rare fragile sites, including
nonclonal (25-29). The telomere-telomere fusion at region FRA2B, segregate as codominant traits, so that a heritable
2q13 must have been accompanied or followed by inactiva- cis-acting difference must exist in the fragile chromosome.
tion or elimination of one of the ancestral centromeres, as Hastie and Allshire (35) have cited several features of a
well as by events that stabilize the fusion point. Hybridiza- putative telomere-telomere fusion that make it an attractive
tion studies suggest that there is a remnant of an ancestral candidate for the fragile site in this band. In the sole example
centromere at band 2q21, which is consistent with the telo- of a constitutional telomere-telomere fusion in human, a high
meric fusion proposed here (A.B., unpublished data). rate of chromosome gaps and breaks was reported at the
The frequency with which telomere-telomere fusion has fusion point between chromosomes 6 and 19, although the
participated in chromosome evolution cannot readily be presence of residual telomere repeats at the fusion point was
assessed. More ancient fusions than the one described here not confirmed (36).
may not easily be detected because of subsequent mutation The sequences cloned in c8.1 and c29B promise to be
of the telomere-like repeats and their flanking regions. The extremely useful reagents for further study of chromosome
observation of an additional weak interstitial hybridization evolution. Comparison of nucleotide sequence of regions
signal at band 3p14 in 5 of 20 chromosomes with fragment flanking the inverted telomere repeats at band 2q13 with
B could be explained by the presence of a more degener- homologous sequences at human and ape telomeres should
ate subtelomeric remnant of another telomere-telomere fu- cast some light on the nature of telomere evolution, as the
sion. Telomere-related sequences have also been found interstitial location of c8.1 will have sheltered the sequences
in subtelomeric regions of many human chromosomes (8). therein from genomic turnover mechanisms peculiar to sub-
It has not, however, been determined whether these have
arisen by telomere-telomere fusion or by another mechanism telomeric sequences.
of illegitimate recombination. In the single example that
has been studied in detail, an inverted telomere array was We are grateful to M. Rocchi (University of Bari, Italy) for
not present. providing the somatic cell hybrid line RJ38791-CT8, U. Francke
The likeliest explanation for the relatively short stretch of (Howard Hughes Medical Institute, Stanford University Medical
telomere-like repeat in cosmid clone c8.1, compared with the Center, Stanford, CA) for use of a hybrid panel and H. C. Riethman
average length of human telomeres, is the instability of a long (Wistar Institute, Philadelphia) for the half-yeast artificial chromo-
inverted tandem repeat sequence. An extensive inverted somes HTY243 and HTY275.
 Genetics: Udo et al. Proc. Natl. Acad. Sci. USA 88 (1991) 9055

1. Dutrillaux, B., Rethord, M. 0. & Lejeune, J. (1975) Ann. 19. Baldini, A. & Ward, D. C. (1991) Genomics 9, 770-774.
Genet. 18, 153-161. 20. Riethman, H. C., Moyzis, R. K., Meyne, J., Burke, D. T. &
2. Yunis, J. J. & Prakash, 0. (1982) Science 215, 1525-1529. Olsen, M. V. (1989) Proc. Natl. Acad. Sci. USA 86, 6240-6244.
3. Wienberg, J., Jauch, A., Stanyon, R. & Cremer, T. (1990) 21. Legerski, R. J., Hodnett, J. L. & Gray, H. B. (1978) Nucleic
Genomics 8, 347-350. Acids Res. 5, 1445-1464.
4. Koop, B. F., Goodman, M., Xu, P., Chan, K. & Slightom, 22. Wilkie, A. 0. M., Higgs, D. R., Rack, K. A., Buckle, V. J.,
J. L. (1986) Nature (London) 319, 234-238. Spurr, N. K., Fischel-Ghodsain, N., Ceccherini, I., Brown,
5. Andrews, P. (1985) Nature (London) 314, 498-499. W. R. A. & Harris, P. C. (1991) Cell 64, 595-606.
6. De Grouchy, J., Turleau, C., Finaz, C. & Roubin, M. (1975) 23. Ward, 0. G., Wurster-Hill, D. H., Ratty, F. J. & Song, Y.
Chromosome Variation in Human Evolution, ed. Boyce, A. J. (1987) Cytogenet. Cell Genet. 45, 177-186.
(Taylor & Francis, London), pp. 17-38. 24. Wurster-Hill, D. H., Ward, 0. G., Davis, B. H., Park, J. P.,
7. Allshire, R. C., Gosden, J. R., Cross, S. H., Cranston, G., Moyzis, R. K. & Meyne, J. (1988) Cytogenet. Cell Genet. 49,
Rout, D., Sugawara, N., Szostak, J. W., Fantes, P. A. & 278-281.
Hastie, N. D. (1988) Nature (London) 332, 656-659. 25. Fitzgerald, P. H. & Morris, C. M. (1984) Hum. Genet. 67,
8. Wells, R. A., Germino, G. G., Krishna, S. K., Buckle, V. J. & 385-390.
Reeders, S. T. (1990) Genomics 8, 699-704. 26. Morgan, R., Jarzabek, V., Jaffe, J. P., Hecht, B. K., Hecht, F.
9. Zakian, V. A. (1989) Annu. Rev. Genet. 23, 579-604. & Sandberg, A. A. (1986) Hum. Genet. 73, 260-263.
10. McClintock, B. (1941) Genetics 26, 234-282. 27. Mandahl, N., Heim, S., Kristoffersson, U., Mitelman, F.,
11. Haber, J. E., Thornburn, P. C. & Rogers, D. (1984) Genetics Rooser, B., Rydholm, A. & Willen, H. (1985) Hum. Genet. 71,
106, 185-205. 321-324.
12. Moyzis, R. K., Buckingham, J. M., Cram, L. S., Dani, M.,
Deaven, L. L., Jones, M. D., Meyne, J., Ratliff, R. L. & Wu, 28. Mandahl, N., Heim, S., Arheden, K., Rydholm, A., Will6n, H.
J. R. (1988) Proc. Natl. Acad. Sci. USA 85, 6622-6626. & Mitelman, F. (1988) Cancer Genet. Cytogenet. 30, 23-33.
13. de Lange, T., Shiue, L., Myers, R. M., Cox, D. R., Naylor, 29. Kovacs, G., Muller-Brechlin, R. & Szics, S. (1987) Cancer
S. L., Killery, A. M. & Varmus, H. E. (1990) Mol. Cell. Biol. Genet. Cytogenet. 28, 363-365.
10, 518-527. 30. Grosschedl, R. & Hobom, G. (1979) Nature (London) 277,
14. Allshire, R. C., Dempster, M. & Hastie, N. D. (1989) Nucleic 621-627.
Acids Res. 17, 4611-4672. 31. Williams, W. L. & Muller, U. R. (1987) J. Mol. Biol. 196,
15. Cross, S., Lindsey, J., Fantes, J., McKay, S., McGill, N. & 743-755.
Cooke, H. (1990) Nucleic Acids Res. 18, 6649-6657. 32. Mizuuchi, K., Kemper, B., Hays, J. & Weisberg, R. A. (1982)
16. Cheng, J.-F., Smith, C. L. & Cantor, C. R. (1990) Nucleic J. Mol. Biol. 156, 229-243.
Acids Res. 19, 149-154. 33. Lilley, D. M. & Kemper, B. (1984) Cell 36, 413-422.
17. Brown, W. R. A., MacKinnon, P. J., Villasantd, A., Spurr, N., 34. Sutherland, G. R. & Mattei, J. F. (1987) Cytogenet. Cell Genet.
Buckle, V. J. & Dobson, M. J. (1990) Cell 63, 119-132. 46, 316-324.
18. Lichter, P., Tang, C. J. C., Call, K., Hermanson, G., Evans, 35. Hastie, H. & Allshire, R. C. (1989) Trends Genet. 5, 326-331.
G. A., Housman, D. & Ward, D. C. (1990) Science 247, 64-69. 36. Drets, M. E. & Therman, E. (1983) Chromosoma 88, 139-144.