Genomics 85 (2005) 143 – 151 www.elsevier.com/locate/ygeno

Mapping long-range chromatin organization within the chicken a-globin gene domain using oligonucleotide DNA arrays Elena Ioudinkovaa,1, Andrei Petrovb,c,1, Sergey V. Razina, Yegor S. Vassetzkyb,* a Institute of Gene Biology, Russian Academy of Sciences, Moscow, Russia UMR-8126, Institut Gustave Roussy, 39, Rue Camille-Desmoulins, 94805 Villejuif, France c Department of Chemistry, Moscow State University, Moscow, Russia

b

Received 10 February 2004; accepted 21 September 2004

Abstract We have analyzed the organization of the chicken a-globin gene domain using DNA miniarrays and have found two novel chromatin loop attachment regions. We have found a 40-kb loop domain that includes all the a-globin genes in cells of erythroid origin. One of the domain borders colocalizes almost exactly with a strong MAR element and with a block of enhancer-blocking elements found earlier at the upstream end of the a-globin gene domain. The domain structure was found to be different in a lymphoid cell line DT40. We propose to use the technique of DNA arrays to map the nuclear matrix attachment sites that define the borders of chromosome loop domains. The technique of DNA arrays permits a large number of DNA sequences to be immobilized on a glass or nylon matrix. This may prove useful for mapping chromatin loop positions within the human genome by using a pool of chromatin loop attachment regions as a probe in a hybridization with a DNA chip containing a specific DNA region. D 2004 Elsevier Inc. All rights reserved. Keywords: Chicken; a-globin gene domain; Nuclear matrix; Oligonucleotide array

Introduction A number of genomes have been sequenced recently. Despite this progress, the mechanisms that control gene expression, and the mechanisms of developmental control in particular, remain largely unknown [1]. Establishment of developmental programs and maintenance of differentiated cell lineages may require developmentally regulated specific chromatin organization, e.g., cell line-specific organization of nuclear territories and chromatin domains. Indeed, dynamic changes in large-scale chromatin organization have been detected during development [2]. Several levels of DNA compaction exist in the eukaryotic nucleus. The DNA is packed into nucleosomes, and the resulting chromatin is further compacted into 30-nm fibers * Corresponding author. Fax: +33(0)1 42 11 54 94. E-mail address: [email protected] (Y.S. Vassetzky). 1 These authors contributed equally to this article. 0888-7543/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygeno.2004.09.008

and chromatin loop domains. The loop domains can be visualized after extraction of histones from isolated nuclei or metaphase chromosomes where they are anchored to a proteinaceous nucleoskeleton, called also nuclear matrix or scaffold [3,4]. Chromatin loop domain size varies from 20 to 200 kb with many genes and clusters of functionally related genes being organized into distinct loops. Similarly, the organization of replicons within the genome also seems to be associated with loops, as shown by colocalization of DNA loop anchorage sites with replication origins [5]. Chromatin loop size also correlates with the size of the replicons [6,7]. In many species, the size of the chromatin loop domains increases from approximately 50 kb during early embryogenesis to 200 kb in the cells of the adult organism [6]. Chromatin loops are attached to the nuclear matrix via the chromatin loop attachment regions (LARs). These LARs are usually 500–3000 bp long (reviewed in [8]), but may be much longer in AT-rich isochores [9]. They may include

144

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

topoisomerase II binding sites [10] as well as other sequence motifs. Considering that the organization of chromatin loops is quite varied and complex, it is unsurprising that neither a consensus DNA motif nor a specific protein has been shown to be responsible for the attachment of DNA loops to the nuclear matrix (for review see [11]). We have recently shown that there is a developmental change from apparently random to specific attachments of the rDNA and a-globin domains to the nuclear matrix during development in Xenopus laevis [2]. The specification of nuclear matrix attachment regions (MARs) observed in this system may be involved in the establishment of stable programs of transcription during development and may contribute to the determination of stable cell lineages in the embryo. Rearrangement of DNA loops was also found in transformed cells in which the average loop size was found to decrease [12]. Additionally, the loop size in several human cancer cell lines was found to be smaller than in normal cells [13]. This may reflect a reversal in the differentiated state of the normal cells. Hence, a study of the chromatin loop domain organization may provide important data on the large-scale mechanisms involved in development and oncogenic transformation. The DNA loop domains are attached to the nucleoskeleton (matrix), so one way of characterizing the domains is to study the attachment of DNA to the nuclear matrix at the anchorage points of the domains. The conventional technique consists of the treatment of isolated nuclei with nonionic detergents and/or high salt to remove histones, followed by DNase I treatment [14– 16]. The loops are digested, while DNA fragments associated with the nuclear matrix remain protected (for a review see [17]). The fraction of the nuclear matrixassociated DNA obtained in this way may be used as a probe in Southern blotting to test the cloned areas of a genome (Fig. 1A). The major drawback of this method is caused by the abundance of repetitive sequences in the genomes of higher eukaryotes. Indeed, the presence of a repetitive sequence in the nuclear matrix fraction and its subsequent hybridization with a fragment of cloned DNA does not determine definitely if the repeat is actually associated with the nuclear matrix within the area of interest (for example see [18]) (Fig. 1B). Another inconvenience of this method is the necessity of manipulating large fragments of cloned DNA due to the average size of the domains in human chromatin being approximately 80 kDa [19]. Recent advances in sequencing various genomes have made available a vast pool of sequence data, allowing an improvement in the reliability of the technique of chromatin domain mapping. This is achieved by using a pool of chromatin loop attachment regions as a probe in a hybridization with a DNA chip containing a specific DNA region or even a chromosome.

A DNA microarray consists of an orderly arrangement of DNA fragments representing the genes or DNA of an organism. Oligonucleotide arrays use small synthetic DNA fragments, while PCR arrays are based on the spotting of PCR products, generally 200–600 nucleotides long. Oligonucleotide arrays have higher hybridization specificity, due to probe design. In the present study, we have studied the organization with respect to the nuclear matrix of the chicken a-globin gene domain and surrounding regions within a 100-kb sequenced DNA fragment (AY016020 [20]) in cells of different lineages using oligonucleotide DNA arrays.

Results Large-scale chromatin organization in the domain of chicken a-globin genes revealed by DNA array technique Chicken a-globin genes have been intensively studied over the past 20 years (reviewed in [21]). The most prominent features of the globin gene domain are summarized in Fig. 2A. Isolation of the nuclear matrix has allowed identification and mapping of the DNA regions that structurally define the bases of the chromatin loop domains, the scaffold/matrix attachment sites [22,23]. Extensive treatment of the isolated nuclei with DNase I, under conditions similar to those used for in vivo footprinting, leads to the digestion of nonprotein-associated DNA. Subsequent extraction of the nuclei in a high-salt buffer removes histones and other highly soluble proteins with associated DNA. The remaining nucleoskeleton contains chromatin LARs. This DNA fraction was purified, radiolabeled, and used as a probe to examine the chromatin loop organization of the chicken aglobin gene domain. To determine the organization of the a-globin gene domain, we have employed the DNA array technique: the sequence of the domain having been analyzed for the presence of DNA repeats and the oligonucleotides designed to avoid the presence of repetitive sequences. The oligonucleotides were quantified by PAGE and slot-blotted onto a nylon membrane. The chicken genome has not been sequenced completely yet, therefore we have checked whether the chosen oligonucleotides in the array contained repetitive sequences by hybridization with total chicken DNA. Fig. 2B shows that total chicken DNA hybridizes almost equally to the globin DNA array, with the exception of the oligonucleotides positioned at 28, 60, and 84 kb, which may be represented by several copies in the genome or contain some short simple sequence motifs. The hybridization pattern observed with the nuclear matrix DNA is quite different: two LARs at positions 34 and 78 kb were found. In addition, one of the multicopy DNA elements we have found (positioned at 60 kb) was

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

145

Fig. 1. Limitations in the conventional nuclear matrix attachment site mapping technique. (A) The principle of the chromatin loop domain mapping technique. (B) Chromatin domain mapping and DNA repeats.

also detected in the fraction of loop anchorage regions. The pattern of weaker hybridization signals observed with nuclear matrix DNA also was distinct from that observed with total DNA. While total DNA gave almost equal hybridization signals with all slots (with the exception of the three slots mentioned above), the matrix DNA probe demonstrated preferential hybridization with oligonucleotides positioned at 34–90 kb and gave virtually no signal with oligonucleotides positioned at 0–33 kb. A similar pattern was observed with the fraction of LARs from erythroblasts isolated from anemic chicken (data not shown). Hybridization of the soluble DNA fraction with the array revealed a pattern similar to that of the total DNA. This is not surprising since the non-matrix-associated DNA constitutes 95–98% of the genome. The organization of the chromatin domains varies depending on the cell lineage and functional state of the genes. To determine whether the observed organization of the a-globin gene domain is specific for cells of erythroid lineage, we isolated the nuclear matrix from cultured cells of lymphoid lineage, DT40, and used the nuclear matrix-associated DNA as a probe in hybridization with the globin DNA array. This probe gave preferential hybridization with oligonucleotide positioned at 84 kb.

Large-scale chromatin organization in the chicken a-globin gene domain monitored by FISH The DNA loops fixed at the nuclear matrix or scaffold of metaphase chromosome can be visualized in histonedepleted nuclei (metaphase chromosome) as a crown of DNA surrounding the proteinaceous core element (chromosomal scaffold or nuclear matrix) [3,24]. Using FISH on high-salt-extracted nuclei (nuclear halos) or chromosomes, one can detect individual loops and determine whether short individual sequences are located on the nuclear matrix or in DNA loops [25–28]. We also used fluorescence in situ hybridization (FISH) to validate our data. Nuclear halos were prepared by salt extraction of the nuclei immobilized on glass slides. One can see a halo of DNA loops (light blue, Fig. 3B) originating from the nuclear matrix (bright blue, Fig. 3B). The nuclear matrix can also be visualized in these preparations using staining with lamins. It coincides with the bright 4,6diamindo-2-phenylindole (DAPI) staining (data not shown). Nuclear matrices were prepared by in situ digestion of the nuclear halos with DNase I. Loop DNA was almost completely digested, while the nuclear matrix-associated DNA remained protected by the nuclear matrix proteins (Fig. 3C). These preparations as well as intact HD3 nuclei were used for the FISH.

146 E. Ioudinkova et al. / Genomics 85 (2005) 143–151 Fig. 2. Mapping the nuclear matrix attachment site in the chicken a-globin gene domain by DNA array technique. (A) Map of the chicken a-globin domain showing the positions of repeated DNA and the nuclear matrix attachment sites. The positions of the transcribed a-globin genes are shown. (B) Nuclei from HD3 and DT40 cells were treated as described for Fig. 1A. Total DNA and DNA remaining attached to the nuclear matrix was radiolabeled and hybridized with the DNA array to probe for specific regions in the a-globin genomic domain. (C) Quantification of the array data. The data represent the ratios of hybridization of nuclear matrix vs. total DNA. The averages of three independent experiments (two hybridizations per experiment) are presented. The signal of the total DNA hybridization was subtracted from the MAR hybridization signal.

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

The DNA fragment corresponding to one of the detected LARs positioned at 78 kb was isolated by PCR from genomic DNA, purified, labeled, and used as a probe in FISH with the nuclear halos or matrices from HD3 cells. Almost exclusive hybridization of this probe with the nuclear matrix was observed (Figs. 3B and 3C, green color). In 100 cells inspected 97% of signals given by this LAR probe were detected on the nuclear matrix. The signals from two homologous chromosomes were located far from each other, in agreement with the previously published observations that usually homologous chromosomal territories do not occupy adjacent positions [29], thus validating our technique of S/MAR analysis. Similar results were obtained when a LAR positioned at 34 kb was tested (data not shown). In contrast, the probe derived from the region with coordinates 4–8 kb gave signals preferentially in loop DNA (Fig. 3B, red color), again as one would expect based on results of array analysis. In 100 cells inspected 78% of the signal given by the probe positioned at 4–8 kb was detected in loop DNA. The presence of 22% of the signals on the nuclear matrix represents, most probably, a result of distortion of the three-dimensional loop halos in the course of immobilization on the microscope slide. And on the other hand, the signal derived from the given probe, in contrast to the LAR probe, is absent in at least 95% of nuclear matrix preparations analyzed (Fig. 3C).

147

oligonucleotide array and by using the relatively long fragments of nuclear matrix DNA as a hybridization probe; we ensure that all the existing LARs will be detected, including those located in the vicinity and within the DNA repeats. We have clearly identified two types of DNA interactions with the nuclear matrix within the 100-kb-long DNA fragment studied in the present work. These can be characterized by strong (positions at 34 and 78 kb) and weak hybridization signals (positions 34–90 kb). The strong signals are likely to identify the loop anchorage regions. Permanent positioning at the nuclear matrix of the regions located at 34 and 78 kb was supported by the results of FISH on nuclear halos and matrices. Interestingly, one of these regions (position 34) colocalizes almost exactly with a strong MAR element [34] and with a block of enhancerblocking elements found at the upstream end of the a-globin gene domain [35]. Furthermore, this loop anchorage region is located in proximity to the LCR of the a-globin domain [20] and to the upstream border of the chromatin domain, characterized by a particular type of histone acetylation in erythroid cells [36]. This transcribed domain (including the cluster of a-globin genes) is characterized by preferential (although temporal and distributive) interaction with the nuclear matrix resulting in the weak positive signals observed with nuclear matrix DNA (positions 34 and 90), compared to total DNA. This result may reflect a temporal interaction of transcribed DNA sequences with the nuclear matrix [30,37,38].

Discussion Oligonucleotide DNA arrays provide a new method for S/MAR analysis Conventional methods for identifying S/MARs are based either on hybridization of nuclear matrix DNA with cloned probes [22,30] or on fishing up cloned DNA fragments possessing an ability to interact with the nuclear matrix in vitro [31] (for a review see [17,32]). Both approaches have their internal problems. The first involves manipulation of large fragments of cloned DNA, often containing repeats, thereby limiting its use over large genomic regions. The second approach ensures the isolation of a subset of DNA sequences that do bind nuclear matrix in vitro but may or may not be involved into the organization of DNA loop anchorage sites in vivo [33]. We propose to use oligonucleotide DNA arrays instead of the digests of cloned DNA used in conventional methods of mapping DNA regions bound to the nuclear matrix. This procedure allows the use of Cot-1 DNA, which is commercially available only for human and murine genomes, to be avoided and excludes the influence of repeated DNA on the results obtained. Indeed, prehybridization with Cot-1 DNA would saturate the repeated DNA, but at the same time exclude DNA repeats from the analysis. We have chosen a different strategy by excluding the repeats at the level of the

Cell lineage-specific organization of the chicken a-globin gene domain It is interesting that clear-cut differences between the patterns of interaction with the nuclear matrix of the region under study were observed when nuclear matrix DNA probes from erythroid and nonerythroid cells were compared. Even the permanent interactions observed in erythroid cells were missing in lymphoid cells. This result supports the previous observations of the Lichter group [27] and shows that topological organization of DNA in the interphase nucleus may be cell-lineage dependent. It seems reasonable to expect the changes in this organization may also occur in connection with malignant transformation. If true, application of DNA array technology to studies of DNA spatial organization may prove to be a promising diagnostic tool. To this end it is important to say that the sensitivity of our oligonucleotide array allows the detection of LARs using the nuclear matrix DNA fraction isolated from 105 cells. Speaking more generally, one may say that a study of the chromatin loop domain organization may provide important data on the large-scale mechanisms involved in development and in oncogenic transformation. The simple and reliable microarray technique may constitute an important tool for these studies.

148

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

Fig. 3. FISH of the nuclei, nuclear halos, and matrices. Hybridization of PCR fragments corresponding to the 4-kb regions positioned at 4–8 and 76–80 kb within the a-globin gene domain to the (A) nuclei, (B) nuclear halos, and (C) nuclear matrices. The positions of the FISH signal are indicated by arrows (76- to 80-kb PCR fragment) or arrowheads (4- to 8-kb PCR fragment); nuclei and nuclear halos/matrices were counterstained with DAPI (blue). From left to right, DNA staining, FISH signals from 76- to 80-kb PCR fragment, FISH signals from 4- to 8-kb PCR fragment, and combined images are presented. (D) The association of the 32- to 36-kb region with the nuclear matrix, and (E) the association of the 54- to 57-kb region with the loop DNA are shown. The PCR fragments were labeled by random priming with biotin–14-dCTP and FISH signals were detected using corresponding antibodies conjugated with TAMRA (red). The positions of the FISH signal are indicated by arrows (32- to 36-kb PCR fragment) or arrowheads (54- to 57-kb PCR fragment); nuclei and nuclear halos/matrices were counterstained with DAPI (blue). From left to right, DNA staining, FISH signals, and combined images are presented. (F) A summary of our findings, showing the organization of the DNA loop domain in the cells of erythroid origin, as based on DNA array and FISH data. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Materials and methods

RPMI 1640 medium supplemented with 10% fetal bovine serum, 2% chicken serum, and 0.05 mM h-mercaptoethanol.

Cell cultures Purification of nuclei and nuclear matrices The chicken AEV-transformed cell line HD3 (clone A6 of line LSCC [39]) was grown in suspension in DulbeccoTs modified EagleTs medium supplemented with 8% fetal bovine serum and 2% chicken serum. The chicken DT40 cell line of lymphoid origin was grown in suspension in the

Nuclei were purified from chicken cells as described elsewhere [40]. Nuclear matrices were prepared by treatment of the isolated nuclei with DNase I followed by extraction with 2 M NaCl essentially as described in [41].

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

The size distribution of the loop domain attachment sites (the matrix-bound DNA fragments) was in the range of 200–1000 bp. The fraction of the loop domain attachment sites constituted 2–5% of the total DNA. In the pilot experiments, we used the loop domain attachment sites with a size range of 200–500, 200–1000, 500–2000, and 500– 5000 bp. The largest fraction gave a much higher background, and one LAR (located at the 78 kb) was missing when we used the smallest fraction. We chose the 200–1000 size range as it gave reproducible results with a relatively small background. Nuclear matrices were digested with proteinase K and extracted with phenol–chloroform. Isolated nuclear matrix DNA was treated with RNase A and, after being labeled, used as a probe in a Southern/slot blot with the DNA array covering the 100 kb of the chicken aglobin gene domain. Nuclear matrices obtained from three independent experiments were used for hybridizations. Each labeling/hybridization experiment was carried out in duplicate. Similar amounts of DNA were taken for the labeling reaction and the signal level detected after the hybridization was not significantly different. DNA array The DNA array consisted of 50 60-mer oligonucleotides spaced 2 kb apart (see Supplementary Data). The oligonucleotides had a similar T m. Prior to hybridization, the oligonucleotides were analyzed in silico to avoid repetitive DNA sequences. The oligonucleotides were slot-blotted onto Hybond N+ filters in 10 SSC and fixed by baking at 808C for 2 h. The hybridization was carried out at 658C in modified Church buffer (0.5 M phosphate buffer, pH 7.2, 7% SDS, 10 mM EDTA) overnight. The blot was washed subsequently in 2 SSC, 0.1% SDS two times for 5 min, then in 1 SSC, 0.1% SDS two times for 10 min, and finally in 0.1 SSC, 0.1% SDS three times for 10 min. The blots were exposed to PhosphoImager Fuji FLA-3000 for 3–12 h. Nuclear halo/matrices preparation for FISH Chicken nuclei were prepared as described elsewhere [40]. To obtain the nuclear halos nuclei were pelleted at 200g for 10 min onto glass slides. Slides were then incubated in a buffer containing 10 mM Pipes, pH 6.8, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 0.5% Triton X-100, 0.1 mM CuSO4, 1 mM PMSF for 10 min on ice, followed by treatment in buffer containing 10 mM Pipes, pH 6.8, 2 M NaCl, 10 mM EDTA, 0.1% digitonin, 0.05 mM spermine, 0.125 mM spermidine for 4 min. Slides were then passed through 10, 5, 2, and 1 PBS followed by 10, 30, 50, 70, and 95% ethanol solutions, air-dried, and finally fixed at 708C for 2 h. Preparation of nuclear matrices on the microscope slides was carried out essentially as described elsewhere [42].

149

FISH analysis DNA fragments corresponding to the 4- to 8-kb and 76- to 80-kb regions of the a-globin domain were amplified by PCR using the Expand Long Template PCR System (Roche). Total genomic DNA of HD3 was applied as a template. Primers Glo6-f (5V-CAAAAGCTTTCCCACTGCTGTTTCACC-3V) and Glo6-r (5V-GATTTCAGATGTTAGCAGAACAATATG3V) were used for amplification of the fragment located at 4–8 kb, Glo34-f (5V-TTCCATATAGAACCATGTAAATTCTTTCC-3V) and Glo34-r (5V-CCATGTTCCCATTTCAGAGACCAATGTCC-3V) were used for the amplification of the 32–36 fragment containing a LAR identified by oligonucleotide array screening, Glo56-f (5V-CTTCACCACAACGTCAGTTTTGATGGAGG-3V) and Glo56-r (5V-CCCCTCAGCGCGGCCGCTCGGGGCTC-3V) were used for amplification of the 54–57 fragment, and Glo78-f (5V-ACGCCAAGCTCACCATCACCGTCACCATG-3V) and Glo78-r (5V-CCTCCTTCCTTAGATGGCAAACCCTGCAG-3V) were applied for amplification of the 76- to 80-kb fragment containing a LAR. The former 4-kb PCR product was labeled by random priming with DIG–11-dUTP (Roche), the latter PCR fragment was labeled in the same manner with biotin–14-dCTP (Invitrogen BioPrime DNA Labeling System). Hybridization to the slides containing nuclei, nuclear halos, or matrices was performed as described earlier [42]. For detection we applied anti-biotin mouse antibodies conjugated with AlexaFluor 488 (Molecular Probes) or anti-DIG sheep antibodies conjugated with 5-carboxy-tetramethyl-rhodamin-N-hydroxy-succinimide ester (TAMRA; Roche). In the case of hybridization with nuclear matrices the signal amplification kit for mouse antibodies (Molecular Probes) was applied. To amplify the signal derived from DIG-labeled DNA anti-sheep donkey antibodies conjugated with AlexaFluor 546 (Molecular Probes) were used. The nuclei, nuclear halos, and matrices were counterstained by DAPI (0.5 Ag/ml) in Vectashield antifade mounting medium (Vector Laboratories). Slides were examined on an Olympus Provis fluorescence microscope with a 100 oil immersion objective and the appropriate filters. Images were captured with a cooled charge-coupled device video camera (SenSys KAF-1400; Photometrics), using RSImage software (Scanalytics).

Acknowledgments We thank Dr. Alan Hair for critically reading the manuscript. This research was supported by grants from the ARC, the FRM, the Ligue contre le Cancer (Comite´ de Charante), and the Science Support Foundation to Y.S.V. and the Russian Academy of Sciences Program for Support of Physico-Chemical Biology to E.I. and S.V.R. E.I. acknowledges the ICRETT fellowship from the IUCC. A.P. was supported by CNRS and a FEBS postdoctoral fellowship.

150

E. Ioudinkova et al. / Genomics 85 (2005) 143–151

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ygeno.2004. 09.008.

[19] [20]

References [21] [1] S.V. Razin, A.P. Petrov, A. Hair, Y.S. Vassetzky, Chromatin domains and territories: flexibly rigid, Crit. Rev. Eukaryotic Gene Expression (2003) (in press). [2] Y. Vassetzky, A. Hair, M. Mechali, Rearrangement of chromatin domains during development in Xenopus, Genes Dev. 14 (2000) 1541 – 1552. [3] J.R. Paulson, U.K. Laemmli, The structure of histone-depleted metaphase chromosomes, Cell 12 (1977) 817 – 825. [4] R. Hancock, M.E. Hughes, Organization of DNA in the eukaryotic nucleus, Biol. Cell 44 (1982) 201 – 212. [5] S.V. Razin, M.G. Kekelidze, E.M. Lukanidin, K. Scherrer, G.P. Georgiev, Replication origins are attached to the nuclear skeleton, Nucleic Acids Res. 14 (1986) 8189 – 8207. [6] M. Buongiorno-Nardelli, G. Micheli, M.T. Carri, M. Marilley, A relationship between replicon size and supercoiled loop domains in the eukaryotic genome, Nature 298 (1982) 100 – 102. [7] M. Marilley, G. Gassend-Bonnet, Supercoiled loop organization of genomic DNA: a close relationship between loop domains, expression units, and replicon organization in rDNA from Xenopus laevis, Exp. Cell Res. 180 (1989) 475 – 489. [8] Y.S. Vassetzky, A. Hair, S.V. Razin, Rearrangement of chromatin domains in cancer and development, J. Cell Biochem. S35 (2000) 54 – 60. [9] E.Y. Svetlova, S.V. Razin, M. Debatisse, Mammalian recombination hot spot in a DNA loop anchorage region: a model for the study of common fragile sites, J. Cell Biochem. 81 (2001) 170 – 178. [10] S.V. Razin, Y.S. Vassetzky, R. Hancock, Nuclear matrix attachment regions and topoisomerase II binding and reaction sites in the vicinity of a chicken DNA replication origin, Biochem. Biophys. Res. Commun. 177 (1991) 265 – 270. [11] Y. Vassetzky, J.M. Lemaitre, M. Mechali, Specification of chromatin domains and regulation of replication and transcription during development, Crit. Rev. Eukaryotic Gene Expression 10 (2000) 31 – 38. [12] M.H. Linskens, A. Eijsermans, P.A. Dijkwel, Comparative analysis of DNA loop length in nontransformed and transformed hamster cells, Mutat. Res. 178 (1987) 245 – 256. [13] F. Oberhammer, J.W. Wilson, C. Dive, I.D. Morris, J.A. Hickman, A.E. Wakeling, P.R. Walker, M. Sikorska, Apoptotic death in epithelial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation, EMBO J. 12 (1993) 3679 – 3684. [14] S.V. Razin, V.L. Mantieva, G.P. Georgiev, The similarity of DNA sequences remaining bound to scaffold upon nuclease treatment of interphase nuclei and metaphase chromosomes, Nucleic Acids Res. 7 (1979) 1713 – 1735. [15] J. Basler, N.D. Hastie, D. Pietras, S.I. Matsui, A.A. Sandberg, R. Berezney, Hybridization of nuclear matrix attached deoxyribonucleic acid fragments, Biochemistry 20 (1981) 6921 – 6929. [16] G. Farache, S.V. Razin, F.R. Targa, K. Scherrer, Organization of the 3Vboundary of the chicken alpha globin gene domain and characterization of a CR 1-specific protein binding site, Nucleic Acids Res. 18 (1990) 401 – 409. [17] S.V. Razin, I.I. Gromova, The channels model of nuclear matrix structure, Bioessays 17 (1995) 443 – 450. [18] G. Farache, S.V. Razin, W.J. Rzeszowska, J. Moreau, F.R. Targa, K. Scherrer, Mapping of structural and transcription-related matrix

[22]

[23]

[24] [25]

[26]

[27]

[28] [29]

[30] [31]

[32] [33]

[34]

[35]

[36]

[37] [38]

[39]

attachment sites in the alpha-globin gene domain of avian erythroblasts and erythrocytes, Mol. Cell Biol. 10 (1990) 5349 – 5358. D.A. Jackson, P. Dickinson, P.R. Cook, The size of chromatin loops in HeLa cells, EMBO J. 9 (1990) 567 – 571. J. Flint, C. Tufarelli, J. Peden, K. Clark, R.J. Daniels, R. Hardison, W. Miller, S. Philipsen, K.C. Tan-Un, T. McMorrow, et al., Comparative genome analysis delimits a chromosomal domain and identifies key regulatory elements in the alpha globin cluster, Hum. Mol. Genet. 10 (2001) 371 – 382. F. Recillas-Targa, S.V. Razin, Chromatin domains and regulation of gene expression: familiar and enigmatic clusters of chicken globin genes, Crit. Rev. Eukaryotic Gene Expression 11 (2001) 227 – 242. J. Mirkovitch, M.E. Mirault, U.K. Laemmli, Organization of the higher-order chromatin loop: specific DNA attachment sites on nuclear scaffold, Cell 39 (1984) 223 – 232. S.M. Gasser, U.K. Laemmli, The organization of chromatin loops: characterization of a scaffold attachment site, EMBO J. 5 (1986) 511 – 518. B. Vogelstein, D.M. Pardoll, D.S. Coffey, Supercoiled loops and eucaryotic DNA replication, Cell 22 (1980) 79 – 85. W.A. Bickmore, K. Oghene, Visualizing the spatial relationships between defined DNA sequences and the axial region of extracted metaphase chromosomes, Cell 84 (1996) 95 – 104. N. Kikyo, P.A. Wade, D. Guschin, H. Ge, A.P. Wolffe, Active remodeling of somatic nuclei in egg cytoplasm by the nucleosomal ATPase ISWI, Science 289 (2000) 2360 – 2363. A. Ratsch, S. Joos, P. Kioschis, P. Lichter, Topological organization of the MYC/IGK locus in BurkittTs lymphoma cells assessed by nuclear halo preparations, Exp. Cell Res. 273 (2002) 12 – 20. K. Byrd, G. Corces, Visualisation of chromatin domains created by the gypsy insulator of Drosophila, J. Cell Biol. 162 (2003) 565 – 574. T. Cremer, G. Kreth, H. Koester, R.H. Fink, R. Heintzmann, M. Cremer, I. Solovei, D. Zink, C. Cremer, Chromosome territories, interchromatin domain compartment, and nuclear matrix: an integrated view of the functional nuclear architecture, Crit. Rev. Eukaryotic Gene Expression 10 (2000) 179 – 212. E.M. Ciejek, M.J. Tsai, B.W. O’Malley, Actively transcribed genes are associated with the nuclear matrix, Nature 306 (1983) 607 – 609. P.N. Cockerill, W.T. Garrard, Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase II sites, Cell 44 (1986) 273 – 282. S.V. Razin, Functional architecture of chromosomal DNA domains, Crit. Rev. Eukaryotic Gene Expression 6 (1996) 247 – 269. O. Iarovaia, R. Hancock, M. Lagarkova, R. Miassod, S.V. Razin, Mapping of genomic DNA loop organization in a 500-kilobase region of the Drosophila X chromosome by the topoisomerase II-mediated DNA loop excision protocol, Mol. Cell Biol. 16 (1996) 302 – 308. S.V. Razin, K. Shen, E. Ioudinkova, K. Scherrer, Functional analysis of DNA sequences located within a cluster of DNase I hypersensitive sites colocalizing with a MAR element at the upstream border of the chicken alpha-globin gene domain, J. Cell Biochem. 74 (1999) 38 – 49. V. Valadez-Graham, S.V. Razin, F. Recillas-Targa, CTCF-dependent enhancer blockers at the upstream region of the chicken alpha-globin gene domain, Nucleic Acids Res. (2004) (in press). E. Anguita, C.A. Johnson, W.G. Wood, B.M. Turner, D.R. Higgs, Identification of a conserved erythroid specific domain of histone acetylation across the alpha-globin gene cluster, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 12114 – 12119. D.A. Jackson, P.R. Cook, A general method for preparing chromatin containing intact DNA, EMBO J. 4 (1985) 913 – 925. S.V. Razin, W.J. Rzeszowska, J. Moreau, K. Scherrer, Localization of regions of DNA attachment to the nuclear matrix within the chicken alpha-globin genes in functionally active and functionally inactive nuclei, Mol. Biol. 19 (1985) 456 – 466. H. Beug, A. von Kirchbach, G. Doderlein, J.F. Conscience, T. Graf, Chicken hematopoietic cells transformed by seven strains of defective

E. Ioudinkova et al. / Genomics 85 (2005) 143–151 avian leukemia viruses display three distinct phenotypes of differentiation, Cell 18 (1979) 375 – 390. [40] J. Rzeszowska-Wolny, S. Razin, E. Puvion, J. Moreau, K. Scherrer, Isolation and characterization of stable nuclear matrix preparations and associated DNA from avian erythroblasts, Biol. Cell 64 (1988) 13 – 22.

151

[41] S.M. Gasser, Y.S. Vassetzky, in: H. Gould (Ed.), Chromatin: A Practical Approach, Oxford Univ. Press, Oxford, UK, 1998, pp. 111 – 124. [42] S. Cai, T. Kohwi-Shigematsu, Intranuclear relocalization of matrix binding sites during T cell activation detected by amplified fluorescence in situ hybridization, Methods 19 (1999) 394 – 402.