- Email: [email protected]
Isolation and characterization of a family of sequences dispersed on the human X chromosome
3,32-38
GENOMICS
Isolation
(1988)
and Characterization of a Family of Sequences on the Human X Chromosome B. BARDONI,*
Dispersed
5. GUIOLI,* E. RAIMONDI,* R. HEILIG,t J. L. MArum,t S. OTTOLENGHI,* AND G. CAMERINO*~’
*Dipartimento di Genetica e Microbiologia, Universitd di Pavia, Pavia, Italy; VJ. 184 (INSfRM) and LGMf (CNRS), Universit6 Louis Pasteur, Strasbourg, France; and SDipartimento di Genetica e di Biologia dei Microorganismi, Universita’ di Milano, Milan, Italy ReceivedJanuary21,
1988;revised
Press,
16, 1988
duplications of DNA segments [for instance, globins (Efstratiadis et aZ., 1980)], duplication of whole chromosomes [for instance, chromosomes 11 and 12 (Craig et al, 1986)] or even polyploidization, and dispersal of DNA copies of mRNAs [retrotranscripts pseudogenes (Heilig et ai., 1984)]. Thus sequence families can appear either as genes clustered in one or a small number of loci (globin, interferon, fibrinogen, growth hormone, etc.) or as dispersed gene families (actin). Pseudogene families in most cases are dispersed on several chromosomes, although clusters of retrotranscripts can be found [for instance, a cluster of ornithine aminotransferase pseudogenes on the X chromosome (Ramesh et al., 1987)]. Thus in almost all cases, families of related sequences either are found within a relatively short genomic region or are dispersed onto many chromosomes. Until now, families of dispersed chromosome-specific sequences have been identified only on the human (Vergnaud et aZ., 1986) and mouse (Lamar and Palmer, 19% Bishop et aZ., 1985) Y chromosomes. These arose probably as a result of duplications that occurred in a chromosome that appears to evolve rapidly and to have a very low coding capacity. We report here the characterization of a family of at least seven related sequences that are present on the X chromosome at a minimum of four different locations in the distal portions of the chromosome (~22 and q28). This observation is of interest given the conservation of the genetic content of the X chromosome in mammals (Ohno, 1969) and in light of the current hypothesis of the mechanisms of X inactivation that postulate a role for X-specific sequences in the spread of inactivation (Gartler and Riggs, 1983).
During a systematic search for X-specific sequences we isolated a DNA fragment (called 61.3) that hybridizes to six further homologous X-specific genomic fragments that map to at least four different regions of the human X chromosome. Genomic segments of 1 l-30 kb (called 61.3 a, b, c, d, and e or DNF22Sl to DNF22SS) have been subsequently cloned for five of the seven repetitions and characterized by restriction mapping. Single-copy sequences have been used to analyze homology between cloned repetitions, to confirm X specificity, and to regionally localize the repetitions. Sequence homology between members of this family seemsto be very high (SO-SOW) and to extend over at least 5 to 12 kb. In situ hybridization and Southern blotting experiments with a panel of human-rodent hybrid cell lines demonstrated that four of the cloned sequences map to three different regions within Xp21.2-pter and the fifth one (G1.3~) maps to Xq28. The family is present with the same complexity and X specificity in macaques (20-30 X 10’ years divergence with man), whereas no related sequences were detected in the mouse. To our knowledge small families of dispersed chromosome-specific sequences have been described only for the human Y chromosome. The possible functional or evolutionary significance of this family is discussed. Q 1988 Academic
May
Inc.
INTRODUCTION
Small families of related sequences detectable by hybridization or by nucleotide sequence comparisons are found in increasing numbers in genomes from higher eukaryotes, including man. Their presence reflects mechanisms that operate in evolution: tandem
MATERIALS * To whom correspondence should dress: Biologia Generale e Genetica Pavia, Italy. OMa-7543/88$3.00 Copyright 0 1988 by Academic All
rights
of reproduction
be addressed at present adMedica, C.P. 217-1, 27100
in any form
Inc. reserved.
METHODS
Cell lines used in gene-dosage experiments or for subchromosomal localization have been described 32
Press,
AND
HUMAN
X-SPECIFIC
(Oberl8 et al., 1986). Cell lines A2.4 (Ray et al., 1985) and PeCH.B (Couturier et al., 1979) were obtained from Drs. Worton and Hors-Cayla, respectively. Total genomic DNA was extracted from human or cercopithecoid leukocytes or cultured cells, digested to completion with restriction endonucleases, fractionated by electrophoresis on 0.9% agarose gels as described (Oberlb et al., 1986), and blotted onto diazobenzyloxymethyl (DBM) paper (Alwine et aZ., 1979; Bellard et al., 1980). In situ hybridization was performed essentially as described by Mattei et al. (1985) with modified stringency conditions: hybridization to chromosome spreads was performed in the presence of 2X SSC, 50% formamide at 37°C with 100 rig/ml of 3H nick-translated G1.3 probe (sp act, 4.9 X 10’ cpm/pg); slides were then washed in 2X SSC, 37% formamide at 39°C and exposed for 7-30 days. The two genomic libraries used in this study were obtained by BglII complete digestion of total genomic 46,xX DNA cloned into the BanHI site of XEMBL4 (S. Ottolenghi, unpublished) and by partial Mb01 digestion of total genomic DNA obtained from a human 49,XXXXY cell line cloned into the BamHI site of X2001 (R. Heilig, unpublished). RESULTS
A Probe That Detects a Set of X-Linked Related Sequences A 2.6-kb EcoRI fragment (laboratory acronym G1.3) devoid of repetitive sequences was isolated from a DNA library specific for the human X chromosome (Davies et aZ., 1981) by using a systematic screening procedure (Oberle et al, 1986). X linkage was assessed by a gene-dosage experiment with genomic DNAs obtained from normal males (46,XY), normal females (46,xX), and human cell lines with 48,XXXX and 49,XXXXY karyotypes. Probe G1.3 detected seven EcoRI fragments in human DNA. The fragments differed in size and in intensity of the hybridization signal and included one fragment with the size (2.6 kb) of the cloned probe (Fig. 1). Five of these fragments (designated a to e) are clearly X-chromosome-specific as shown by their dosage, which paralleled the concentration of X-chromosome sequences in the various DNAs. The hybridization of the same blot with autosomal probes showed that approximately equal amounts of DNA were present in each line (results not shown). Five different genomic sequences containing the EcoRI fragments a to e have been subsequently cloned and shown to be X linked (see below). For the two largest fragments (f and g), which hybridize more faintly, it was not possible to determine dosage unambiguously. Analysis of various somatic hybrid lines strongly suggests that these two se-
33
SEQUENCES
9 $1 A&
12.0 -9.6 -7.7
-4.2 b-2.6
a-
FIG. 1. Gene-dosage analysis of fragments hybridizing to orobe Gl.3 in human DNA dieested with EcoRI. After hvbridization &h probe G1.3 (in 40% formamide at 42’C), the blot was washed in 1X SSC containing 0.1% NaDodSO, at 60°C (medium stringency). Origin of DNAs:XY, unrelated males; XX, unrelated females; 4X and 4XY, lymphoblastoid cell lines with 48,XXXX and 49,XXXXY karyotypea, respectively. The sizes of fragments detected by probe G1.3 are indicated in kb. The additional band detected in lane 4X is probably due to a digestion artifact.
quences also belong to the X chromosome (see below). In addition, all fragments detected by probe G1.3 in human DNA digested with HindIII and TaqI showed X dosage (results not shown). Thus this sequence family appears completely X specific. However, under reduced stringency conditions (hybridization in 40% formamide at 37”C, washing in 2X SSC at 50°C) faint bands which might be autosomal are detected. Regional Localization In order to establish the regional localization of the DNA fragments detected by G1.3 on the X chromosome, we used a panel of rodent-human cell lines derived from human parental cells carrying X-autosome translocations with various breakpoints on the X chromosome. This panel has been used extensively to localize a large number of X-specific probes, and is thus well characterized (Oberlh et al., 1986). The DNAs from the various lines were digested with EcoRI, PstI, or TaqI and blotted onto DBM paper (Fig. 2a and results not shown). Hybridization with G1.3 showed that the fragments detected map to at least four different regions of the X chromosome. In EcoRI digests, fragment a, which has the size of the cloned probe, is present only in hybrids containing the distal portion of the short arm of the X chromosome (GM194, Cer.S, 58.6, PeCH.B; see lanes 4,8,12). Furthermore, it is present in one line (PI.7-2; lane 13) but absent in two others (HRBCB, 34.X, lanes 3,14) that are defined as containing the p22.3-qter portion of the human X chromosome (Oberle et al., 1986). We conclude that fragment a is localized in the p22.3-pter
34
BARDONI 1 2
3
4
5 6 7 8
ET AL.
9 10 11 12 13 1415
b)
A (4)
lml
B C 03) tn
D (14)
E
F
E H I (l.U (10) (9)
J (7)
K (6)
1 (5)
l
N (8)
0
l
l
l
+
l
+
l
l
l
l
GlJA
N (12)
10628.0169
l
+
l
+
l
Gl.k
l
l
l
l
l
l
+
&m6
+t*+
+
l
l
l
Gl.k,r
l
l
+
l
l
l
-
l
*****
l
l
.
+
l
+
+
l
l
ml
+
-
.++*++t
F9
l
at*++++++
U.k
+
+
D(sy.FB
l
*+++***+.
l
+
l
t
l
+
+
l +
+
+
t
l
t
+
+
+
+ +
+
+
+
FIG. 2. Regional localization on the X chromosome of the fragments detected by the Gl.3 probe. (a) The DNAs from the various rodent-human hybrid cell lines and from control rodent or human cells were digested and analyzed as in Fig. 1. Origin of DNA8 in lane 1, normal female; 2, mouse; 3, (C); 4, (A); 6, (L); 6, (K); 7, (J); 8, (N); 9, (I); 10, (H); 11, (G); 12, (M); 13, (B); 14. (D); 15, Chinesehamster. Letters in parentheses correspond to hybrid cell lines described in b. The band in lane 7 that migrates slightly slower than fragment b is due to plamnid contamination of the hybrid DNA. The band in lane 4 that migrates faster than fragment b is probably the result of incomplete d&&ion (see the reduced intensity of fragment a compared to the intensity in other lanes). (b) Results of hybridization experiments: +, presence of human X-specific fragment; *, not tested. The human-rodent somatic hybrids and the portion of human X chromosome they retain are A, GM194 (pter-q28); B, PI.7-2 (p22.3-qter); C, HRBCB (p22.3-qter); D, 34.X (p22.3-qter); E, A2-4 (p21.2-qter); F, GO.4 (pll-qter); G, Cer.H (qll-qter); H, 58.2b (q13-qter); I, 56.0 (q21-qter); J, GM69 (q23-qter); K, 63.R (q26-qt.er); L, GM97.613 (q26-qter); M, Cer.S (pter-qll); N, 58.6 (pter-q21); 0, PeCH.B (pter-q27). Numbers in parentheses on top indicate the corresponding lanes in a.
region, distal to RC8 (DXSS) in Xp22.2 but proximal to MIA (DXS31) (which is absent from hybrid PI) (Fig. 2b). Fragments b, d, and e are all located distal to the Duchenne locus defined by the X-21 translocation present in hybrid A2-4 (lane E, Fig. 2b). They are differentiated only by the Cer.H hybrid (fragment b is present only in CerS, while fragments d and e are present in both Cer.S and Cer.H; lanes 11 and 12). Cer.H was originally described as containing only the qll-qter portion of the X chromosome, while the Cer.S hybrid, derived originally from the same translocation, contains the complementary qll-qter region. However, of about 50 probes tested, we have found 4 that are present in both hybrids, although with a reduced concentration in Cer.H. This can probably be explained by the presence of an extra X-chromosome fragment in hybrid Cer.H. Three of these probes, including DXSl, previously localized to qll-q12, are also present in the GO.4 hybrid (which contains pll-qter; lane F, Fig. 2b). The fourth probe, pXUT23 (DXSlG), previously localized to ~22.2 is absent from GO.4 and shows exactly the same pattern with respect to the hybrid panel as that of fragments d and e (see Fig. 2b). Thus fragments b, d, and e are present in two different regions within Xp21.2-~22.3, with fragments d and e probably being in the ~22.2 region close to pXUT23 (DXSlS). Fragment c is located in the distal end of the long arm of the X chromosome (q27-qter), as indicated by its presence in the GM97 line (q26-qter; lane 5, Fig. 2b) and its absence from the PeCH.B line (pter-q27; lane 0, Fig. 2b).
Faint bands corresponding to fragments f and g were present in hybrids GM194, 58.6, CerS, and PeCH.B. Although these data are not sufficient for a precise localization of these fragments, they strongly suggest that these fragments belong to the short arm of the X chromosome. In Situ Hybridization To provide an independent localization of the sequences homologous to probe G1.3, we hybridized it to human metaphase chromosomes (Fig. 3). Of the 350 grains observed, 36% were on the X chromosome; 42% of X-linked grains were present in the terminal portion of the X short arm (p22.3-pter), an additional 21% were distributed within bands p21 and ~22.1, and 8% in band ~21. This scatter does not fit with a unique hybridization region and suggests that in addition to the ~22.3 localization (which corresponds to G1.3a itself; see above), other cross-reacting sequences have a more proximal localization. Significant hybridization (13%) was observed in the q28 region, in agreement with the localization of fragment G1.3c (see above). However, 16% of the grains appeared localized in a region around the centromere. We do not know at present whether this corresponds to additional members of the sequence family (not detected in the blot hybridization experiments) or to nonspecific hybridization to centromeric sequences. Cloning and Mapping
of the Homologous
Regions
In order to characterize this sequence family further, we screened two human genomic libraries with
HUMAN
CHROMOSOME
FIG. 3. distribution
X-SPECIFIC
_ _ _ _ _ _
NUMBER
In situ hybridization with probe G1.3. (A) Distribution on the X chromosome.
probe G1.3. A first library was derived from a BglII complete digest cloned into bacteriophage EMBL4, and allowed us to obtain several identical recombinant phages containing a 4.2-kb EcoRI fragment hybridizing to the G1.3 probe (similar in size to the b fragment seen in EcoRI digests of genomic DNA). A single different clone contained the 7.7-kb EcoRI fragment characteristic of fragment c. To obtain clones corresponding to the other repeats, we screened a complete library constructed by cloning a Sau3A partial digest of human 49,XXXXY DNA in the X2001 vector; 24 positive phages were further characterized by restriction mapping with EcoRI, HindIII, and Sac1 and by localization of human repetitive sequences. This allowed us to identify groups of overlapping clones (Fig. 4). We have thus isolated five different genome regions of 11 to 30 kb. Each region contains a unique EcoRI fragment homologous to the G1.3 probe; the sizes of these fragments correspond to the five prominent fragments (a to e) detected in genomic blots. This was further proven by subcloning unique sequences from each phage and using them as probes (under high-stringency conditions) on dosage and localization blots as shown in Figs. 1 and 2 for the initial probe. The patterns obtained confirmed that all five cloned regions are X linked and localized as expected (not shown). The human gene-mapping nomenclature for the five cloned regions is DNF22Sl to DNF22S5. In order to study the extent of homology among the cloned members of the family, we hybridized six unique subcloned fragments (belonging to different phages) and probe Gl.3 to blots containing single and double digests of phage DNAs (with EcoRI, HindIII, and SacI). In addition, each probe was tested on blots containing human genomic DNAs digested with different restriction enzymes. This analysis gives a rough estimate of the size of the homologous fragments. Three of the probes used (probes 1,2, and 3)
35
SEQUENCES
... .. .. ... ..
of 350 grains on 36 mitoses (the thick bar represents 5
x
grains); (B)
recognize homologous fragments in all five members of the family. The total size of fragments hybridizing varies from about 4.5 kb in c to about 9.5 kb in a, with fragments hybridizing to probe 1 adjacent to those hybridizing to probe 3 in c, d, and e and separated by about 2 and 4 kb in b and a, respectively. Probe 3 recognizes seven to eight X-linked fragments in various restriction digests of human DNA and a Y-chromosome-specific fragment (but no autosomal fragments) (Fig. 5, lanes 1 and 2). Probe 4 hybridizes to
SHSHE I 111
H I
E I
E -
E
E I
ESHE
G1.3 EESHHS III,,
Ilk-
SS I,
G1.3a Xp223-plw
-
SES HS I, I ,I
EH I
HSm I,
E 1
H I
Gl.3b
I-
Xp21.2-~22.3
ESiEE 1 \If .___ -
E I
ESESSHEHH 1 I ‘Iif lr
HE
q
E S E I III _______.__-_-
EHH H I
S I
___-__.-_. . . . .._-
H 1
H SSH I II
HE
f
S HHH Illlll 888se-e
E
SHS
G1.3~ xq20
EH Gl.3d Xp21.2-~22.3
E I
H I
H I
H I G1.3e Xp21Pp22.3
FIG. 4. Restriction map of the five genomic regions cloned and representation of the homologies among them. Restriction enzyme sites: E, EcoRI; H, HindIII, S, SacI. The following symbols indicate the regions of homology to the probes used (the letters in parentheses indicate from which cloned member of the G1.3 family each probe originates): 0 - 0 - 0, probe 1 (a); -, probe 2 (e); X - X - X, probe 3 (c); - - -, probe 4 (c); I, probe 5 (d); 0 - 0 - 0, probe 6 (d); 0 - 0 - 0, probe 7 (d).
36
BARDONI 12345676
ET AL.
being intermediate. We estimate that there is at least 80% sequence homology between members of the G1.3 family since the cross hybridization is detected in genomic blots under conditions of medium stringency: this conclusion is derived from sequence analysis of other fragments that show cross hybridization under similar conditions (Heilig et al, 1980; Mandel et al., 1986) and is confirmed by preliminary sequence analysis of fragments from the G1.3 a and b regions. The Family Is Present in Cercopithecoids but Not in Rodents
FIG. 5. Hybridization of probes from different regions of the G1.3 family to human genomic DNA digested with HindIII. Lanes 1,3,5, and 7, DNAs from normal females; lanes 2,4, 6, and 8, DNAs from normal males. Probes used: lanes 1 and 2, probe 3; lanes 3 and 4, probe 4; lanes 5 and 6, probe 5; lanes 7 and 8, probe 6 (origin of probes as in Fig. 4).
only three members of the family (c, d, and e). However, it detects about six to seven X-linked sequences on genomic blots (and no autosomal sequences). For instance, in HindI digests, probe 4 detects seven fragments, five of which hybridize also to probe 3 (Fig. 5, lanes 3 and 4). This suggests that sequences homologous to those in probe 4 are present in most or all members of the family, although they might be located more distantly in regions a and b. Probe 5, originating from region d, hybridizes only to region e, at a comparable position and intensity. In genomic blots this probe detects a complex pattern at medium-low stringency, as for a middle repetitive sequence (Fig. 5, lanes 5 and 6). However, it detects at least three Xlinked EcoRI fragments, two of which correspond to regions d and e. It must thus be present in at least one other member of the family. Probes 6 and 7 (originating from region d) detect only the cognate fragments both on phage (Fig. 4) and on genomic blots (Fig. 5, lanes 7 and 8) and are thus located outside of the region of homology. A comparison of the relative intensities of hybridization signals in the different regions suggests that regions d and e have the highest homology. This is further supported by the similarity of their restriction maps and the similarity of their organization with respect to five of the seven probes used. Since these two regions share the same localization, determined by means of the somatic hybrid panel, it is possible that they are clustered as a result of a recent duplication event. Regions a and b seem to be the most distantly related to d and e, with region c
We have hybridized probe G1.3 to DNAs from male and female cercopithecoids, lemurs, and rodents. In several macaque species and in African green monkey (Cercopitheques), a family of strongly hybridizing fragments with a complexity similar to that observed in humans (Fig. 6) was found. Comparison of signals in males with those in females strongly suggests that all fragments are X linked (the presence of polymorphisms between individuals does not allow a comparison of dosage for all bands). The family is also present in ateles, a new world monkey of more ancient divergence with the human lineage; however, we have not yet checked for X linkage in this species. In contrast, no clear pattern was observed in lemur DNA with probe G1.3, and virtually no hybridization was present with probe 3 under our standard reaction HS -MT--MN--MticMM-+CVMMMFFMMFFMFFMMFFMMFF
FIG. 6. Pattern of hybridization of G1.3 in cercopithecoids. DNAs were digested with ToqI. Source of DNAs: M, males, F, females; HS, Homo sapiens; MT, Mocaca tonkeana; MN, Macaca nemestrina; MA, Macaca arctoides; MM, Macaca mulatta; CV, Cercopithecue
ueruett.
HUMAN
X-SPECIFIC
conditions. No signal was ever observed in mouse or hamster DNA, using all the six probes described above. DISCUSSION
We have partially characterized a family of six to seven X-chromosome-specific sequences and have cloned 11-30 kb of genomic DNA corresponding to five of them. The sequences appear distributed in at least four different locations on the X chromosome: three in the distal region (p21.2-pter) of the short arm, and one in the distal portion of the long arm (q28). The presence of cross-reacting sequences around the centromere was suggested by in situ hybridization, but could not be confirmed using a panel of somatic cell hybrids. This discordance may be due to differences in the hybridization conditions and/or the signal-to-noise ratio between the two types of techniques. Sequence homology between members of the family appears to extend over at least 5 to 12 kb. The fragments detected correspond to single-copy sequences, as indicated by the Mendelian inheritance of RFLPs detected in three members of the family (unpublished results). To our knowledge small families of dispersed chromosome-specific sequences have been described only for the Y chromosome (Vergnaud et al., 1986; Lamar and Palmer, 1984; Bishop et al., 1985). This chromosome is peculiar in its apparently very low coding capacity and in its rapid evolution. Many sequences present on the Y chromosome are recent borrowings from the X and perhaps other chromosomes (Page et al., 1984; Koenig et al., 1985; Bickmore and Cooke, 1987). Since the Y is not involved in meiotic exchange (apart from the pseudoautosomal region), chromosomal rearrangements that may lead to duplication and dispersion of a given sequence or even of whole chromosome regions may be less deleterious for this than for other chromosomes. Other chromosome-specific sequence families appear clustered: this is the case for small gene families (globins, interferons, etc.), even the ornithine transaminase pseudogene family (also on the X) (Ramesh et al., 1987), and for larger ones such as the chromosome-specific alphoid sequences present in centromeric regions (Willard et al., 1986) and the X-specific family clustered in Xq28 described by Miiller et al. (1986). None of the mechanisms proposed for the duplication or amplification of a sequence (retrotranscription, unequal crossingover, etc.) might account for the chromosome specificity and dispersion of this family. Does the G1.3 family have a functional significance that would account both for its conservation (the cross hybridization observed suggests at least 80-90s homology between members) and for its dispersion?
37
SEQUENCES
One attractive hypothesis would be an involvement in the mechanisms of X-chromosome inactivation. In order to explain the phenomenon of incomplete spreading of inactivation onto autosomal sequences in X-autosome translocations, it has been hypothesized that in addition to a main controlling elementthe inactivation center-specific sequences on the X would be required for the transmission and maintenance of inactivation (the way station model; Gartler and Riggs, 1983; Mohandas et al., 1987). The fact that we could not detect sequences homologous to the G1.3 family in rodents (and lemurs) does not a priori favor the hypothesis that the strong homology between the family members is related to a role as important controlling elements. Another problem is that no member of the family has yet been unequivocally identified in the major part of the X chromosome, from ~21.2 to q27. Experiments are underway to try to analyze the DNA methylation and chromatin structure of these sequences on the active and inactive X chromosome, in search of functionally significant correlations. It is also possible that the high homology among members of the family in man reflects a recent evolutionary origin. We have shown that the family has the same complexity and X specificity in macaques (20-30 X lo6 years divergence with man), while we have not yet been able to trace its presence in lemurs (SO-80 X lo6 years divergence). Given the apparently slower rate of evolution in higher primates (Li and Tanimura, 1987; Savatier et al., 1987), sequences that diverged 30 X lo6 years ago would still be able to show cross hybridization comparable to that observed. One hypothesis that would account for the dispersion of the family on the X chromosome is that rearrangements that occurred during evolution of the X chromosome led to a dispersal of a previously clustered family. This would imply that these sequences were preferential sites for such rearrangements. In this respect it is interesting to note that genes that map in the q28 region (GGPD, Hemophilia A) and in ~21.2 (Duchenne) in man, appear closely linked in mouse, while genes that are in the p21.2-pter region in man (DMD and hypophosphatemia or steroid sulfatase) are separated in the mouse by genes belonging to the human long arm (Avner et al., 1987; Brockdorff et al., 1987; Heilig et al., 1987). In situ hybridization analysis of the location of this sequence family and of defined X-linked genes in primates should allow a test of this hypothesis. ACKNOWLEDGMENTS This work was supported by grants from the CNRS (Action Incitatives “EUROPE” 920193/011) to J. L. Mandel and by the Progetto Finalizzato “Ingegneria Genetica e Basi Molecolari delle Malattie Ereditarie” of CNR to G. Camerino.
38
BARDONI
1.
ALWINE, J. C., KEMP, D. J., PARKER, B. A., REISER, J., RENART, J., STARK, G. R., AND WAHL, G. M. (1979). Detection of specific RNAs or specific fragments of DNA by fractionation in gels and transfer to diazobenzyloxymethyl paper. In “Methods in Enzymology” (R. Wu, Ed.), Vol. 68, pp. 220-242, Academic Press, New York. AVNER, P., AMAR, L., ARNAUD, D., HANAUER, A., AND CAMBROU, J. (1987). Detailed ordering of markers localizing to the Xq26-Xqter region of the human X chromosome by the use of an interspecific Mus spretus mouse cross. Proc. Natl. Acad. Sci. USA 84: 1629-1633. BELLARD, M., Kuo, M. T., DRETZEN, G., AND CHAMBON, P. (1980). Differential nuclease sensitivity of the ovalbumin and @globin chromatin regions in erythrocytes and oviduct cells of laying hen. Nucleic Acids Res. 8: 2737-2750. BICKMORE, W. A., AND COOKE, J. H. (1987). Evolution of homologous sequences on the human X and Y chromosomes, outside of the meiotic pairing segment. Nucleic Acids Res. 16: 6261-6271. BISHOP, C. E., BOURSOT, P., BARON, B., BONHOMME, F., AND HATAT, D. (1985). Most classical Mus musculus domesticus laboratory mouse strains carry a Mus musculus musculus Y chromosome. Nature (London) 316: 70-72. BROCKDORFF, N., CROSS, G. S., CAVANNA, J. S., FISHER, E. M. C., LYON, M. F., DAVIES, K. E., AND BROWN, S. D. M. (1987). The mapping of a cDNA from the human X-linked Duchenne muscular dystrophy gene to the mouse X chromosome. Nature (London) 228: 166-168. COUTURIER, J., DUTRILLAUX, B., GARBER, P., RAOUL, O., CROQUETTE, M.-F., FOURLINNIE, J. C., AND MAILLARD, E. (1979). Evidence for a correlation between late replication and autosomal gene inactivation in a familial translocation t(X21). Hum. Genet. 49: 319-326. CRAIG, S. P., BUCKLE, V. J., LAMOUROUX, A., MALLET, J., AND CRAIG, I. (1986). Localization of the human tyrosine hydroxylase gene to 11~15: Gene duplication and evolution of metabolic pathways. Cytogenet. Cell Genet. 42: 29-32. DAVIES, K. E., YOUNG, B. D., ELLES, R. G., HILL, M. E., AND WILLIAMSON, R. (1981). Cloning of a representative genomic library of the human X chromosome after sorting by flow cytometry. Nature (London) 293: 374-376. EFSTRATIADIS, A., POSAKONY, J. W., MANIATIS, T., LAWN, R. M., O’CONNELL, C., SPRITZ, R. A., DERIEL, J. K., FORGET, B. G., WEISSMAN, S. W., SLIGHTOM, J. L., BLECHL, A. E., SMITHIES, O., BARALLE, F. E., SHOULDERS, C. C., AND PROUDFOOT, N. J. (1980). The structure and evolution of the human @globin gene family. Cell 2 1: 653-668. GARTLER, S. M., AND RIGGS, A. D. (1983). Mammalian Xchromosome inactivation. Annu. Reu. Genet. 1'7: 155-190. HEILIG, R., PERRIN, F., GANNON, F., MANDEL, J. L., AND CHAMBON, P. (1980). The ovalbumin gene family: Structure of the X gene and evolution of duplicated split genes. Cell 20: 625-637. HEILIG, R., HANAUER, A., GRZESCHIK, K. H., HORS-CAYLA, M. C., AND MANDEL, J. L. (1984). A&n-like sequences are present on the X and Y chromosomes. EMBO J. 3: 1803-1807. HEILIG, R., LEMAIRE, C., MANDEL, J. L., DANDOLO, L., AMAR, L., AND AVNER, P. (1987). Locahsation of the region homologous to the Duchenne muscular dystrophy locus on the mouse X chromosome. Nature @md.on) 328: 168-170. KOENIG, M., MOISAN, J. P., AND MANDEL, J. L. (1985). HOmologies between X and Y chromosomes detected by DNA probes: Localisation and evolution. Nucleic Acids Res. 13: 5485-5501.
REFERENCES
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13. 14.
15.
ET
AL.
16. LAMAR, E. E., AND PALMER, E. (1984). Y-encoded, speciesspecific DNA in mice: Evidence that the Y chromosome exists in two polymorphic forms in inbred strains. CeU 37: 171-177. 17. LI, W.-H., AND TANIMURA, M. (1987). The molecular clock runs more slowly in man than in apes and monkeys. Nature (London) 326: 93-96. 18. MANDEL, J. L., ARVEILER, B., CAMERINO, G., HANAUER, A., HEILIG, R., KOENIG, M., AND OBERLE, I. (1986). Genetic mapping of the human X chromosome: Linkage analysis of the q26-q28 region that includes the fragile X locus and isolation of expressed sequences. CoM Spring Harbor Symp. Quant. Biol. 51: 195-203. 19. MATTEI, M. G., PHILIP, N., PASSAGE, E., MOISAN, J. P., MANDEL, J. L., AND MATTEI, J. F. (1986). DNA probe localization at 18~113 band by in situ hybridization: An identification of a small supernumerary chromosome. Hum. Genet. 69: 268-271. 20. MOHANDAS, T., GELLER, R. L., YEN, P. H., ROSENDORFF, J., BERNSTEIN, R., YOSHIDA, A., AND SHAPIRO, L. J. (1987). Cytogenetic and molecular studies on a recombinant human X chromosome: Implications for the spreading of X chromosome inactivation. Proc. NatL Acad. Sci. USA 04: 4954-4968. 21. MWLLER, U., TANTRAVAHI, U., MONACO, A., STROH, H., KUNKEL, L. M., AND LATT, S. A. (1986). Repeated DNA sequences in the distal long arm of the human X chromosome. Hum. Genet. 74: 24-29. 22. OBERLI?, I., CAMERINO, G., KLOEPFER, C., MOISAN, J. P., GRZESCHIK, K. H., HELLKUHL, B., HORS-CAYLA, M. C., VAN CONG, N., WEIL, D., AND MANDEL, J. L. (1986). Characterizetion of a set of X-linked sequences end of a panel of eomatic cell hybrids useful for the regional mapping of the human X chromosome. Hum. Genet. 72: 43-49. 23. OHNO, S. (1969). Evolution of sex chromosomes in mammals. Annu. Rev. Genet. 2: 496. 24. PAGE, D. C., HARPER, M. E., LOVE, J., AND BOTSTEIN, D. (1984). Occurrence of a transposition from the X chromosome long arm to the Y chromosome short arm during evolution. Nature (London) 311: 119-123. 26. RAMESH, V., EDDY, R., BRUNS, G. A., SHIH, V. E., SHOWS, T. B., AND GUSELLA, J. F. (1987). Localization of the omithine aminotransferase gene and related sequences on two human chromosomes. Hum. Genet. 76: 121-126. 26. RAY, P. N., BELFALL, B., DUFF, C., LOGAN, C., KEAN, V., THOMPSON, M. W., SYLVESTER, J. E., GORSKI, J. L., SCHMICKEL, R. D., AND WORTON, R. G. (1985). Cloning of the breakpoint of an X;21 translocation associated with Duchenne muscular dystrophy. Nature (London) 318: 672-675. 27. SAVATIER, P., TRABUCHET, G., CHEBLOUNE, Y., FAURE, C., VERDIER, G., AND NIGON, V. M. (1987). Nucleotide sequence of the delta-beta-globin intergenic segment in the macaque: Structure and evolutionary rates in higher primates. J. Mol. Evol. 24: 297-308. 28. VERGNAUD, G., PAGE, D. C., SIMMLER, M. C., BROWN, L., ROUYER, F., NOEL, B., BOTSTEIN, D., DE LA CHAPELLE, A., AND WEISSENBACH, J. (1986). A deletion map of the human Y chromosome based on DNA hybridization. Amer. J. Hum. Genet. 38: 109-124. 29. WILLARD, H. F., WAYE, J. S., SKOLNICK, M. H., SCHWARTZ, C. E., POWERS, V. E., AND ENGLAND, S. B. (1986). Detection of restriction fragment length polymorphisms at the centromeres of human chromosomes by using chromosome-specific (Ysatellite DNA probes: Implications for development of centromere-based genetic linkage maps. Proc. Natl. Acad. Sci. USA 83: 5611-5615.
GENOMICS
Isolation
(1988)
and Characterization of a Family of Sequences on the Human X Chromosome B. BARDONI,*
Dispersed
5. GUIOLI,* E. RAIMONDI,* R. HEILIG,t J. L. MArum,t S. OTTOLENGHI,* AND G. CAMERINO*~’
*Dipartimento di Genetica e Microbiologia, Universitd di Pavia, Pavia, Italy; VJ. 184 (INSfRM) and LGMf (CNRS), Universit6 Louis Pasteur, Strasbourg, France; and SDipartimento di Genetica e di Biologia dei Microorganismi, Universita’ di Milano, Milan, Italy ReceivedJanuary21,
1988;revised
Press,
16, 1988
duplications of DNA segments [for instance, globins (Efstratiadis et aZ., 1980)], duplication of whole chromosomes [for instance, chromosomes 11 and 12 (Craig et al, 1986)] or even polyploidization, and dispersal of DNA copies of mRNAs [retrotranscripts pseudogenes (Heilig et ai., 1984)]. Thus sequence families can appear either as genes clustered in one or a small number of loci (globin, interferon, fibrinogen, growth hormone, etc.) or as dispersed gene families (actin). Pseudogene families in most cases are dispersed on several chromosomes, although clusters of retrotranscripts can be found [for instance, a cluster of ornithine aminotransferase pseudogenes on the X chromosome (Ramesh et al., 1987)]. Thus in almost all cases, families of related sequences either are found within a relatively short genomic region or are dispersed onto many chromosomes. Until now, families of dispersed chromosome-specific sequences have been identified only on the human (Vergnaud et aZ., 1986) and mouse (Lamar and Palmer, 19% Bishop et aZ., 1985) Y chromosomes. These arose probably as a result of duplications that occurred in a chromosome that appears to evolve rapidly and to have a very low coding capacity. We report here the characterization of a family of at least seven related sequences that are present on the X chromosome at a minimum of four different locations in the distal portions of the chromosome (~22 and q28). This observation is of interest given the conservation of the genetic content of the X chromosome in mammals (Ohno, 1969) and in light of the current hypothesis of the mechanisms of X inactivation that postulate a role for X-specific sequences in the spread of inactivation (Gartler and Riggs, 1983).
During a systematic search for X-specific sequences we isolated a DNA fragment (called 61.3) that hybridizes to six further homologous X-specific genomic fragments that map to at least four different regions of the human X chromosome. Genomic segments of 1 l-30 kb (called 61.3 a, b, c, d, and e or DNF22Sl to DNF22SS) have been subsequently cloned for five of the seven repetitions and characterized by restriction mapping. Single-copy sequences have been used to analyze homology between cloned repetitions, to confirm X specificity, and to regionally localize the repetitions. Sequence homology between members of this family seemsto be very high (SO-SOW) and to extend over at least 5 to 12 kb. In situ hybridization and Southern blotting experiments with a panel of human-rodent hybrid cell lines demonstrated that four of the cloned sequences map to three different regions within Xp21.2-pter and the fifth one (G1.3~) maps to Xq28. The family is present with the same complexity and X specificity in macaques (20-30 X 10’ years divergence with man), whereas no related sequences were detected in the mouse. To our knowledge small families of dispersed chromosome-specific sequences have been described only for the human Y chromosome. The possible functional or evolutionary significance of this family is discussed. Q 1988 Academic
May
Inc.
INTRODUCTION
Small families of related sequences detectable by hybridization or by nucleotide sequence comparisons are found in increasing numbers in genomes from higher eukaryotes, including man. Their presence reflects mechanisms that operate in evolution: tandem
MATERIALS * To whom correspondence should dress: Biologia Generale e Genetica Pavia, Italy. OMa-7543/88$3.00 Copyright 0 1988 by Academic All
rights
of reproduction
be addressed at present adMedica, C.P. 217-1, 27100
in any form
Inc. reserved.
METHODS
Cell lines used in gene-dosage experiments or for subchromosomal localization have been described 32
Press,
AND
HUMAN
X-SPECIFIC
(Oberl8 et al., 1986). Cell lines A2.4 (Ray et al., 1985) and PeCH.B (Couturier et al., 1979) were obtained from Drs. Worton and Hors-Cayla, respectively. Total genomic DNA was extracted from human or cercopithecoid leukocytes or cultured cells, digested to completion with restriction endonucleases, fractionated by electrophoresis on 0.9% agarose gels as described (Oberlb et al., 1986), and blotted onto diazobenzyloxymethyl (DBM) paper (Alwine et aZ., 1979; Bellard et al., 1980). In situ hybridization was performed essentially as described by Mattei et al. (1985) with modified stringency conditions: hybridization to chromosome spreads was performed in the presence of 2X SSC, 50% formamide at 37°C with 100 rig/ml of 3H nick-translated G1.3 probe (sp act, 4.9 X 10’ cpm/pg); slides were then washed in 2X SSC, 37% formamide at 39°C and exposed for 7-30 days. The two genomic libraries used in this study were obtained by BglII complete digestion of total genomic 46,xX DNA cloned into the BanHI site of XEMBL4 (S. Ottolenghi, unpublished) and by partial Mb01 digestion of total genomic DNA obtained from a human 49,XXXXY cell line cloned into the BamHI site of X2001 (R. Heilig, unpublished). RESULTS
A Probe That Detects a Set of X-Linked Related Sequences A 2.6-kb EcoRI fragment (laboratory acronym G1.3) devoid of repetitive sequences was isolated from a DNA library specific for the human X chromosome (Davies et aZ., 1981) by using a systematic screening procedure (Oberle et al, 1986). X linkage was assessed by a gene-dosage experiment with genomic DNAs obtained from normal males (46,XY), normal females (46,xX), and human cell lines with 48,XXXX and 49,XXXXY karyotypes. Probe G1.3 detected seven EcoRI fragments in human DNA. The fragments differed in size and in intensity of the hybridization signal and included one fragment with the size (2.6 kb) of the cloned probe (Fig. 1). Five of these fragments (designated a to e) are clearly X-chromosome-specific as shown by their dosage, which paralleled the concentration of X-chromosome sequences in the various DNAs. The hybridization of the same blot with autosomal probes showed that approximately equal amounts of DNA were present in each line (results not shown). Five different genomic sequences containing the EcoRI fragments a to e have been subsequently cloned and shown to be X linked (see below). For the two largest fragments (f and g), which hybridize more faintly, it was not possible to determine dosage unambiguously. Analysis of various somatic hybrid lines strongly suggests that these two se-
33
SEQUENCES
9 $1 A&
12.0 -9.6 -7.7
-4.2 b-2.6
a-
FIG. 1. Gene-dosage analysis of fragments hybridizing to orobe Gl.3 in human DNA dieested with EcoRI. After hvbridization &h probe G1.3 (in 40% formamide at 42’C), the blot was washed in 1X SSC containing 0.1% NaDodSO, at 60°C (medium stringency). Origin of DNAs:XY, unrelated males; XX, unrelated females; 4X and 4XY, lymphoblastoid cell lines with 48,XXXX and 49,XXXXY karyotypea, respectively. The sizes of fragments detected by probe G1.3 are indicated in kb. The additional band detected in lane 4X is probably due to a digestion artifact.
quences also belong to the X chromosome (see below). In addition, all fragments detected by probe G1.3 in human DNA digested with HindIII and TaqI showed X dosage (results not shown). Thus this sequence family appears completely X specific. However, under reduced stringency conditions (hybridization in 40% formamide at 37”C, washing in 2X SSC at 50°C) faint bands which might be autosomal are detected. Regional Localization In order to establish the regional localization of the DNA fragments detected by G1.3 on the X chromosome, we used a panel of rodent-human cell lines derived from human parental cells carrying X-autosome translocations with various breakpoints on the X chromosome. This panel has been used extensively to localize a large number of X-specific probes, and is thus well characterized (Oberlh et al., 1986). The DNAs from the various lines were digested with EcoRI, PstI, or TaqI and blotted onto DBM paper (Fig. 2a and results not shown). Hybridization with G1.3 showed that the fragments detected map to at least four different regions of the X chromosome. In EcoRI digests, fragment a, which has the size of the cloned probe, is present only in hybrids containing the distal portion of the short arm of the X chromosome (GM194, Cer.S, 58.6, PeCH.B; see lanes 4,8,12). Furthermore, it is present in one line (PI.7-2; lane 13) but absent in two others (HRBCB, 34.X, lanes 3,14) that are defined as containing the p22.3-qter portion of the human X chromosome (Oberle et al., 1986). We conclude that fragment a is localized in the p22.3-pter
34
BARDONI 1 2
3
4
5 6 7 8
ET AL.
9 10 11 12 13 1415
b)
A (4)
lml
B C 03) tn
D (14)
E
F
E H I (l.U (10) (9)
J (7)
K (6)
1 (5)
l
N (8)
0
l
l
l
+
l
+
l
l
l
l
GlJA
N (12)
10628.0169
l
+
l
+
l
Gl.k
l
l
l
l
l
l
+
&m6
+t*+
+
l
l
l
Gl.k,r
l
l
+
l
l
l
-
l
*****
l
l
.
+
l
+
+
l
l
ml
+
-
.++*++t
F9
l
at*++++++
U.k
+
+
D(sy.FB
l
*+++***+.
l
+
l
t
l
+
+
l +
+
+
t
l
t
+
+
+
+ +
+
+
+
FIG. 2. Regional localization on the X chromosome of the fragments detected by the Gl.3 probe. (a) The DNAs from the various rodent-human hybrid cell lines and from control rodent or human cells were digested and analyzed as in Fig. 1. Origin of DNA8 in lane 1, normal female; 2, mouse; 3, (C); 4, (A); 6, (L); 6, (K); 7, (J); 8, (N); 9, (I); 10, (H); 11, (G); 12, (M); 13, (B); 14. (D); 15, Chinesehamster. Letters in parentheses correspond to hybrid cell lines described in b. The band in lane 7 that migrates slightly slower than fragment b is due to plamnid contamination of the hybrid DNA. The band in lane 4 that migrates faster than fragment b is probably the result of incomplete d&&ion (see the reduced intensity of fragment a compared to the intensity in other lanes). (b) Results of hybridization experiments: +, presence of human X-specific fragment; *, not tested. The human-rodent somatic hybrids and the portion of human X chromosome they retain are A, GM194 (pter-q28); B, PI.7-2 (p22.3-qter); C, HRBCB (p22.3-qter); D, 34.X (p22.3-qter); E, A2-4 (p21.2-qter); F, GO.4 (pll-qter); G, Cer.H (qll-qter); H, 58.2b (q13-qter); I, 56.0 (q21-qter); J, GM69 (q23-qter); K, 63.R (q26-qt.er); L, GM97.613 (q26-qter); M, Cer.S (pter-qll); N, 58.6 (pter-q21); 0, PeCH.B (pter-q27). Numbers in parentheses on top indicate the corresponding lanes in a.
region, distal to RC8 (DXSS) in Xp22.2 but proximal to MIA (DXS31) (which is absent from hybrid PI) (Fig. 2b). Fragments b, d, and e are all located distal to the Duchenne locus defined by the X-21 translocation present in hybrid A2-4 (lane E, Fig. 2b). They are differentiated only by the Cer.H hybrid (fragment b is present only in CerS, while fragments d and e are present in both Cer.S and Cer.H; lanes 11 and 12). Cer.H was originally described as containing only the qll-qter portion of the X chromosome, while the Cer.S hybrid, derived originally from the same translocation, contains the complementary qll-qter region. However, of about 50 probes tested, we have found 4 that are present in both hybrids, although with a reduced concentration in Cer.H. This can probably be explained by the presence of an extra X-chromosome fragment in hybrid Cer.H. Three of these probes, including DXSl, previously localized to qll-q12, are also present in the GO.4 hybrid (which contains pll-qter; lane F, Fig. 2b). The fourth probe, pXUT23 (DXSlG), previously localized to ~22.2 is absent from GO.4 and shows exactly the same pattern with respect to the hybrid panel as that of fragments d and e (see Fig. 2b). Thus fragments b, d, and e are present in two different regions within Xp21.2-~22.3, with fragments d and e probably being in the ~22.2 region close to pXUT23 (DXSlS). Fragment c is located in the distal end of the long arm of the X chromosome (q27-qter), as indicated by its presence in the GM97 line (q26-qter; lane 5, Fig. 2b) and its absence from the PeCH.B line (pter-q27; lane 0, Fig. 2b).
Faint bands corresponding to fragments f and g were present in hybrids GM194, 58.6, CerS, and PeCH.B. Although these data are not sufficient for a precise localization of these fragments, they strongly suggest that these fragments belong to the short arm of the X chromosome. In Situ Hybridization To provide an independent localization of the sequences homologous to probe G1.3, we hybridized it to human metaphase chromosomes (Fig. 3). Of the 350 grains observed, 36% were on the X chromosome; 42% of X-linked grains were present in the terminal portion of the X short arm (p22.3-pter), an additional 21% were distributed within bands p21 and ~22.1, and 8% in band ~21. This scatter does not fit with a unique hybridization region and suggests that in addition to the ~22.3 localization (which corresponds to G1.3a itself; see above), other cross-reacting sequences have a more proximal localization. Significant hybridization (13%) was observed in the q28 region, in agreement with the localization of fragment G1.3c (see above). However, 16% of the grains appeared localized in a region around the centromere. We do not know at present whether this corresponds to additional members of the sequence family (not detected in the blot hybridization experiments) or to nonspecific hybridization to centromeric sequences. Cloning and Mapping
of the Homologous
Regions
In order to characterize this sequence family further, we screened two human genomic libraries with
HUMAN
CHROMOSOME
FIG. 3. distribution
X-SPECIFIC
_ _ _ _ _ _
NUMBER
In situ hybridization with probe G1.3. (A) Distribution on the X chromosome.
probe G1.3. A first library was derived from a BglII complete digest cloned into bacteriophage EMBL4, and allowed us to obtain several identical recombinant phages containing a 4.2-kb EcoRI fragment hybridizing to the G1.3 probe (similar in size to the b fragment seen in EcoRI digests of genomic DNA). A single different clone contained the 7.7-kb EcoRI fragment characteristic of fragment c. To obtain clones corresponding to the other repeats, we screened a complete library constructed by cloning a Sau3A partial digest of human 49,XXXXY DNA in the X2001 vector; 24 positive phages were further characterized by restriction mapping with EcoRI, HindIII, and Sac1 and by localization of human repetitive sequences. This allowed us to identify groups of overlapping clones (Fig. 4). We have thus isolated five different genome regions of 11 to 30 kb. Each region contains a unique EcoRI fragment homologous to the G1.3 probe; the sizes of these fragments correspond to the five prominent fragments (a to e) detected in genomic blots. This was further proven by subcloning unique sequences from each phage and using them as probes (under high-stringency conditions) on dosage and localization blots as shown in Figs. 1 and 2 for the initial probe. The patterns obtained confirmed that all five cloned regions are X linked and localized as expected (not shown). The human gene-mapping nomenclature for the five cloned regions is DNF22Sl to DNF22S5. In order to study the extent of homology among the cloned members of the family, we hybridized six unique subcloned fragments (belonging to different phages) and probe Gl.3 to blots containing single and double digests of phage DNAs (with EcoRI, HindIII, and SacI). In addition, each probe was tested on blots containing human genomic DNAs digested with different restriction enzymes. This analysis gives a rough estimate of the size of the homologous fragments. Three of the probes used (probes 1,2, and 3)
35
SEQUENCES
... .. .. ... ..
of 350 grains on 36 mitoses (the thick bar represents 5
x
grains); (B)
recognize homologous fragments in all five members of the family. The total size of fragments hybridizing varies from about 4.5 kb in c to about 9.5 kb in a, with fragments hybridizing to probe 1 adjacent to those hybridizing to probe 3 in c, d, and e and separated by about 2 and 4 kb in b and a, respectively. Probe 3 recognizes seven to eight X-linked fragments in various restriction digests of human DNA and a Y-chromosome-specific fragment (but no autosomal fragments) (Fig. 5, lanes 1 and 2). Probe 4 hybridizes to
SHSHE I 111
H I
E I
E -
E
E I
ESHE
G1.3 EESHHS III,,
Ilk-
SS I,
G1.3a Xp223-plw
-
SES HS I, I ,I
EH I
HSm I,
E 1
H I
Gl.3b
I-
Xp21.2-~22.3
ESiEE 1 \If .___ -
E I
ESESSHEHH 1 I ‘Iif lr
HE
q
E S E I III _______.__-_-
EHH H I
S I
___-__.-_. . . . .._-
H 1
H SSH I II
HE
f
S HHH Illlll 888se-e
E
SHS
G1.3~ xq20
EH Gl.3d Xp21.2-~22.3
E I
H I
H I
H I G1.3e Xp21Pp22.3
FIG. 4. Restriction map of the five genomic regions cloned and representation of the homologies among them. Restriction enzyme sites: E, EcoRI; H, HindIII, S, SacI. The following symbols indicate the regions of homology to the probes used (the letters in parentheses indicate from which cloned member of the G1.3 family each probe originates): 0 - 0 - 0, probe 1 (a); -, probe 2 (e); X - X - X, probe 3 (c); - - -, probe 4 (c); I, probe 5 (d); 0 - 0 - 0, probe 6 (d); 0 - 0 - 0, probe 7 (d).
36
BARDONI 12345676
ET AL.
being intermediate. We estimate that there is at least 80% sequence homology between members of the G1.3 family since the cross hybridization is detected in genomic blots under conditions of medium stringency: this conclusion is derived from sequence analysis of other fragments that show cross hybridization under similar conditions (Heilig et al, 1980; Mandel et al., 1986) and is confirmed by preliminary sequence analysis of fragments from the G1.3 a and b regions. The Family Is Present in Cercopithecoids but Not in Rodents
FIG. 5. Hybridization of probes from different regions of the G1.3 family to human genomic DNA digested with HindIII. Lanes 1,3,5, and 7, DNAs from normal females; lanes 2,4, 6, and 8, DNAs from normal males. Probes used: lanes 1 and 2, probe 3; lanes 3 and 4, probe 4; lanes 5 and 6, probe 5; lanes 7 and 8, probe 6 (origin of probes as in Fig. 4).
only three members of the family (c, d, and e). However, it detects about six to seven X-linked sequences on genomic blots (and no autosomal sequences). For instance, in HindI digests, probe 4 detects seven fragments, five of which hybridize also to probe 3 (Fig. 5, lanes 3 and 4). This suggests that sequences homologous to those in probe 4 are present in most or all members of the family, although they might be located more distantly in regions a and b. Probe 5, originating from region d, hybridizes only to region e, at a comparable position and intensity. In genomic blots this probe detects a complex pattern at medium-low stringency, as for a middle repetitive sequence (Fig. 5, lanes 5 and 6). However, it detects at least three Xlinked EcoRI fragments, two of which correspond to regions d and e. It must thus be present in at least one other member of the family. Probes 6 and 7 (originating from region d) detect only the cognate fragments both on phage (Fig. 4) and on genomic blots (Fig. 5, lanes 7 and 8) and are thus located outside of the region of homology. A comparison of the relative intensities of hybridization signals in the different regions suggests that regions d and e have the highest homology. This is further supported by the similarity of their restriction maps and the similarity of their organization with respect to five of the seven probes used. Since these two regions share the same localization, determined by means of the somatic hybrid panel, it is possible that they are clustered as a result of a recent duplication event. Regions a and b seem to be the most distantly related to d and e, with region c
We have hybridized probe G1.3 to DNAs from male and female cercopithecoids, lemurs, and rodents. In several macaque species and in African green monkey (Cercopitheques), a family of strongly hybridizing fragments with a complexity similar to that observed in humans (Fig. 6) was found. Comparison of signals in males with those in females strongly suggests that all fragments are X linked (the presence of polymorphisms between individuals does not allow a comparison of dosage for all bands). The family is also present in ateles, a new world monkey of more ancient divergence with the human lineage; however, we have not yet checked for X linkage in this species. In contrast, no clear pattern was observed in lemur DNA with probe G1.3, and virtually no hybridization was present with probe 3 under our standard reaction HS -MT--MN--MticMM-+CVMMMFFMMFFMFFMMFFMMFF
FIG. 6. Pattern of hybridization of G1.3 in cercopithecoids. DNAs were digested with ToqI. Source of DNAs: M, males, F, females; HS, Homo sapiens; MT, Mocaca tonkeana; MN, Macaca nemestrina; MA, Macaca arctoides; MM, Macaca mulatta; CV, Cercopithecue
ueruett.
HUMAN
X-SPECIFIC
conditions. No signal was ever observed in mouse or hamster DNA, using all the six probes described above. DISCUSSION
We have partially characterized a family of six to seven X-chromosome-specific sequences and have cloned 11-30 kb of genomic DNA corresponding to five of them. The sequences appear distributed in at least four different locations on the X chromosome: three in the distal region (p21.2-pter) of the short arm, and one in the distal portion of the long arm (q28). The presence of cross-reacting sequences around the centromere was suggested by in situ hybridization, but could not be confirmed using a panel of somatic cell hybrids. This discordance may be due to differences in the hybridization conditions and/or the signal-to-noise ratio between the two types of techniques. Sequence homology between members of the family appears to extend over at least 5 to 12 kb. The fragments detected correspond to single-copy sequences, as indicated by the Mendelian inheritance of RFLPs detected in three members of the family (unpublished results). To our knowledge small families of dispersed chromosome-specific sequences have been described only for the Y chromosome (Vergnaud et al., 1986; Lamar and Palmer, 1984; Bishop et al., 1985). This chromosome is peculiar in its apparently very low coding capacity and in its rapid evolution. Many sequences present on the Y chromosome are recent borrowings from the X and perhaps other chromosomes (Page et al., 1984; Koenig et al., 1985; Bickmore and Cooke, 1987). Since the Y is not involved in meiotic exchange (apart from the pseudoautosomal region), chromosomal rearrangements that may lead to duplication and dispersion of a given sequence or even of whole chromosome regions may be less deleterious for this than for other chromosomes. Other chromosome-specific sequence families appear clustered: this is the case for small gene families (globins, interferons, etc.), even the ornithine transaminase pseudogene family (also on the X) (Ramesh et al., 1987), and for larger ones such as the chromosome-specific alphoid sequences present in centromeric regions (Willard et al., 1986) and the X-specific family clustered in Xq28 described by Miiller et al. (1986). None of the mechanisms proposed for the duplication or amplification of a sequence (retrotranscription, unequal crossingover, etc.) might account for the chromosome specificity and dispersion of this family. Does the G1.3 family have a functional significance that would account both for its conservation (the cross hybridization observed suggests at least 80-90s homology between members) and for its dispersion?
37
SEQUENCES
One attractive hypothesis would be an involvement in the mechanisms of X-chromosome inactivation. In order to explain the phenomenon of incomplete spreading of inactivation onto autosomal sequences in X-autosome translocations, it has been hypothesized that in addition to a main controlling elementthe inactivation center-specific sequences on the X would be required for the transmission and maintenance of inactivation (the way station model; Gartler and Riggs, 1983; Mohandas et al., 1987). The fact that we could not detect sequences homologous to the G1.3 family in rodents (and lemurs) does not a priori favor the hypothesis that the strong homology between the family members is related to a role as important controlling elements. Another problem is that no member of the family has yet been unequivocally identified in the major part of the X chromosome, from ~21.2 to q27. Experiments are underway to try to analyze the DNA methylation and chromatin structure of these sequences on the active and inactive X chromosome, in search of functionally significant correlations. It is also possible that the high homology among members of the family in man reflects a recent evolutionary origin. We have shown that the family has the same complexity and X specificity in macaques (20-30 X lo6 years divergence with man), while we have not yet been able to trace its presence in lemurs (SO-80 X lo6 years divergence). Given the apparently slower rate of evolution in higher primates (Li and Tanimura, 1987; Savatier et al., 1987), sequences that diverged 30 X lo6 years ago would still be able to show cross hybridization comparable to that observed. One hypothesis that would account for the dispersion of the family on the X chromosome is that rearrangements that occurred during evolution of the X chromosome led to a dispersal of a previously clustered family. This would imply that these sequences were preferential sites for such rearrangements. In this respect it is interesting to note that genes that map in the q28 region (GGPD, Hemophilia A) and in ~21.2 (Duchenne) in man, appear closely linked in mouse, while genes that are in the p21.2-pter region in man (DMD and hypophosphatemia or steroid sulfatase) are separated in the mouse by genes belonging to the human long arm (Avner et al., 1987; Brockdorff et al., 1987; Heilig et al., 1987). In situ hybridization analysis of the location of this sequence family and of defined X-linked genes in primates should allow a test of this hypothesis. ACKNOWLEDGMENTS This work was supported by grants from the CNRS (Action Incitatives “EUROPE” 920193/011) to J. L. Mandel and by the Progetto Finalizzato “Ingegneria Genetica e Basi Molecolari delle Malattie Ereditarie” of CNR to G. Camerino.
38
BARDONI
1.
ALWINE, J. C., KEMP, D. J., PARKER, B. A., REISER, J., RENART, J., STARK, G. R., AND WAHL, G. M. (1979). Detection of specific RNAs or specific fragments of DNA by fractionation in gels and transfer to diazobenzyloxymethyl paper. In “Methods in Enzymology” (R. Wu, Ed.), Vol. 68, pp. 220-242, Academic Press, New York. AVNER, P., AMAR, L., ARNAUD, D., HANAUER, A., AND CAMBROU, J. (1987). Detailed ordering of markers localizing to the Xq26-Xqter region of the human X chromosome by the use of an interspecific Mus spretus mouse cross. Proc. Natl. Acad. Sci. USA 84: 1629-1633. BELLARD, M., Kuo, M. T., DRETZEN, G., AND CHAMBON, P. (1980). Differential nuclease sensitivity of the ovalbumin and @globin chromatin regions in erythrocytes and oviduct cells of laying hen. Nucleic Acids Res. 8: 2737-2750. BICKMORE, W. A., AND COOKE, J. H. (1987). Evolution of homologous sequences on the human X and Y chromosomes, outside of the meiotic pairing segment. Nucleic Acids Res. 16: 6261-6271. BISHOP, C. E., BOURSOT, P., BARON, B., BONHOMME, F., AND HATAT, D. (1985). Most classical Mus musculus domesticus laboratory mouse strains carry a Mus musculus musculus Y chromosome. Nature (London) 316: 70-72. BROCKDORFF, N., CROSS, G. S., CAVANNA, J. S., FISHER, E. M. C., LYON, M. F., DAVIES, K. E., AND BROWN, S. D. M. (1987). The mapping of a cDNA from the human X-linked Duchenne muscular dystrophy gene to the mouse X chromosome. Nature (London) 228: 166-168. COUTURIER, J., DUTRILLAUX, B., GARBER, P., RAOUL, O., CROQUETTE, M.-F., FOURLINNIE, J. C., AND MAILLARD, E. (1979). Evidence for a correlation between late replication and autosomal gene inactivation in a familial translocation t(X21). Hum. Genet. 49: 319-326. CRAIG, S. P., BUCKLE, V. J., LAMOUROUX, A., MALLET, J., AND CRAIG, I. (1986). Localization of the human tyrosine hydroxylase gene to 11~15: Gene duplication and evolution of metabolic pathways. Cytogenet. Cell Genet. 42: 29-32. DAVIES, K. E., YOUNG, B. D., ELLES, R. G., HILL, M. E., AND WILLIAMSON, R. (1981). Cloning of a representative genomic library of the human X chromosome after sorting by flow cytometry. Nature (London) 293: 374-376. EFSTRATIADIS, A., POSAKONY, J. W., MANIATIS, T., LAWN, R. M., O’CONNELL, C., SPRITZ, R. A., DERIEL, J. K., FORGET, B. G., WEISSMAN, S. W., SLIGHTOM, J. L., BLECHL, A. E., SMITHIES, O., BARALLE, F. E., SHOULDERS, C. C., AND PROUDFOOT, N. J. (1980). The structure and evolution of the human @globin gene family. Cell 2 1: 653-668. GARTLER, S. M., AND RIGGS, A. D. (1983). Mammalian Xchromosome inactivation. Annu. Reu. Genet. 1'7: 155-190. HEILIG, R., PERRIN, F., GANNON, F., MANDEL, J. L., AND CHAMBON, P. (1980). The ovalbumin gene family: Structure of the X gene and evolution of duplicated split genes. Cell 20: 625-637. HEILIG, R., HANAUER, A., GRZESCHIK, K. H., HORS-CAYLA, M. C., AND MANDEL, J. L. (1984). A&n-like sequences are present on the X and Y chromosomes. EMBO J. 3: 1803-1807. HEILIG, R., LEMAIRE, C., MANDEL, J. L., DANDOLO, L., AMAR, L., AND AVNER, P. (1987). Locahsation of the region homologous to the Duchenne muscular dystrophy locus on the mouse X chromosome. Nature @md.on) 328: 168-170. KOENIG, M., MOISAN, J. P., AND MANDEL, J. L. (1985). HOmologies between X and Y chromosomes detected by DNA probes: Localisation and evolution. Nucleic Acids Res. 13: 5485-5501.
REFERENCES
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13. 14.
15.
ET
AL.
16. LAMAR, E. E., AND PALMER, E. (1984). Y-encoded, speciesspecific DNA in mice: Evidence that the Y chromosome exists in two polymorphic forms in inbred strains. CeU 37: 171-177. 17. LI, W.-H., AND TANIMURA, M. (1987). The molecular clock runs more slowly in man than in apes and monkeys. Nature (London) 326: 93-96. 18. MANDEL, J. L., ARVEILER, B., CAMERINO, G., HANAUER, A., HEILIG, R., KOENIG, M., AND OBERLE, I. (1986). Genetic mapping of the human X chromosome: Linkage analysis of the q26-q28 region that includes the fragile X locus and isolation of expressed sequences. CoM Spring Harbor Symp. Quant. Biol. 51: 195-203. 19. MATTEI, M. G., PHILIP, N., PASSAGE, E., MOISAN, J. P., MANDEL, J. L., AND MATTEI, J. F. (1986). DNA probe localization at 18~113 band by in situ hybridization: An identification of a small supernumerary chromosome. Hum. Genet. 69: 268-271. 20. MOHANDAS, T., GELLER, R. L., YEN, P. H., ROSENDORFF, J., BERNSTEIN, R., YOSHIDA, A., AND SHAPIRO, L. J. (1987). Cytogenetic and molecular studies on a recombinant human X chromosome: Implications for the spreading of X chromosome inactivation. Proc. NatL Acad. Sci. USA 04: 4954-4968. 21. MWLLER, U., TANTRAVAHI, U., MONACO, A., STROH, H., KUNKEL, L. M., AND LATT, S. A. (1986). Repeated DNA sequences in the distal long arm of the human X chromosome. Hum. Genet. 74: 24-29. 22. OBERLI?, I., CAMERINO, G., KLOEPFER, C., MOISAN, J. P., GRZESCHIK, K. H., HELLKUHL, B., HORS-CAYLA, M. C., VAN CONG, N., WEIL, D., AND MANDEL, J. L. (1986). Characterizetion of a set of X-linked sequences end of a panel of eomatic cell hybrids useful for the regional mapping of the human X chromosome. Hum. Genet. 72: 43-49. 23. OHNO, S. (1969). Evolution of sex chromosomes in mammals. Annu. Rev. Genet. 2: 496. 24. PAGE, D. C., HARPER, M. E., LOVE, J., AND BOTSTEIN, D. (1984). Occurrence of a transposition from the X chromosome long arm to the Y chromosome short arm during evolution. Nature (London) 311: 119-123. 26. RAMESH, V., EDDY, R., BRUNS, G. A., SHIH, V. E., SHOWS, T. B., AND GUSELLA, J. F. (1987). Localization of the omithine aminotransferase gene and related sequences on two human chromosomes. Hum. Genet. 76: 121-126. 26. RAY, P. N., BELFALL, B., DUFF, C., LOGAN, C., KEAN, V., THOMPSON, M. W., SYLVESTER, J. E., GORSKI, J. L., SCHMICKEL, R. D., AND WORTON, R. G. (1985). Cloning of the breakpoint of an X;21 translocation associated with Duchenne muscular dystrophy. Nature (London) 318: 672-675. 27. SAVATIER, P., TRABUCHET, G., CHEBLOUNE, Y., FAURE, C., VERDIER, G., AND NIGON, V. M. (1987). Nucleotide sequence of the delta-beta-globin intergenic segment in the macaque: Structure and evolutionary rates in higher primates. J. Mol. Evol. 24: 297-308. 28. VERGNAUD, G., PAGE, D. C., SIMMLER, M. C., BROWN, L., ROUYER, F., NOEL, B., BOTSTEIN, D., DE LA CHAPELLE, A., AND WEISSENBACH, J. (1986). A deletion map of the human Y chromosome based on DNA hybridization. Amer. J. Hum. Genet. 38: 109-124. 29. WILLARD, H. F., WAYE, J. S., SKOLNICK, M. H., SCHWARTZ, C. E., POWERS, V. E., AND ENGLAND, S. B. (1986). Detection of restriction fragment length polymorphisms at the centromeres of human chromosomes by using chromosome-specific (Ysatellite DNA probes: Implications for development of centromere-based genetic linkage maps. Proc. Natl. Acad. Sci. USA 83: 5611-5615.