Einstein Quart. J. Biol. and Med. (2001) 18(3):113-122

Chromatin Higher Order Structure and Chromosomal Positioning in the Nucleus

Matthew Schmerer (Nelson, 1986). The nuclear matrix also seems to be a site Department of Developmental and Molecular Biology for chromatin attachment, which may provide a founda- Albert Einstein College of Medicine tion for the higher order structure of the chromatin Bronx, New York 10461 (Nelson, 1986; van Driel, 1995). This concept of nuclear compartmentalization alluded to the hypothesis that Abstract whole chromosomes or parts of chromosomes may occu- py certain regions of the nucleus (Manuelidis, 1985, While it has been known for some time that chromatin 1990). is highly packaged, it has only recently been discovered that chromatin is attached to the nuclear matrix. It has The existence of chromatin loops and chromosome also been recently shown that interphase chromosomes domains lend support to the contention that the nucle- remain separate from one another in so called, “chro- us is a highly structured organelle. The first level of chro- mosome domains.” This author suggests a functional matin organization is chromatin loops. The chromatin of relationship between chromatin attachment to the uncondensed chromosomes in interphase nuclei is nuclear matrix and chromosomal domains in the inter- attached to the nuclear matrix at regular intervals mak- phase nucleus. Herein, the evidence for the presence of ing loops of 5 to 100 kb in length (Cook and Brazell, matrix attachment sites and chromosomal domains is 1976; Paulson and Leammli, 1977). The second level of presented, and it is proposed that chromatin attachment chromatin organization is the location of the uncon- to the nuclear matrix provides the physical foundation densed chromosomes in the nucleus. Chromosomes for interphase chromosomes to be localized to chromo- occupy unique domains in the nucleus such that individ- some domains. ual chromosomes do not overlap (Manuelidis, 1985, 1990). This author would like to suggest that these two Introduction levels of organization are linked together, such that chromatin loops, which attach to the nuclear matrix pro- It has long been known, based on electron microscopy vide a foundation for the localization of interphase studies, that chromatin is packaged into nucleosomes chromosomes to unique domains. (Olins and Olins, 1974), then into 10 nm fibers (Cameron et al., 1979), and then into 30 nm solenoids (Finch and Chromatin Loops and SARs/MARs Klug, 1976; Felsenfeld and McGhee, 1986). This knowl- edge raised, among other issues, the question of how The chromatin of interphase nuclei is attached to the chromatin gets packaged into metaphase chromosomes. nuclear matrix at specific sites called matrix association Beyond the issue of chromatin packaging is the organi- regions (MARs) or scaffold attachment regions (SARs). zation of the interphase nucleus and where chromatin, These names refer to the same sequences that are asso- as linear uncondensed chromosomes, fits in. This ques- ciated with the nuclear matrix after high salt treatment tion was first addressed by David Comings in 1968 when or ionic detergent treatment (Razin and Gromova, 1995) he described, “The Rationale for an Ordered and will be hereto referred to as MARs only (see Fig. 1). Arrangement of Chromatin in the Interphase Nucleus” The purpose of the high salt treatment is to remove the (Comings, 1968). In this paper, Comings described the histones from chromatin, so that the DNA will be more evidence at that time for the non-random localization of accessible to nuclease treatment. Unfortunately, this chromatin in the nucleus. This includes some of the first preparation method leaves behind actively transcribing evidence that DNA replication occurs, to a greater and replicating DNA as well as nuclear matrix-associated extent, at specific points on the nuclear membrane and DNA sequences (Razin and Gromova, 1995). This makes it along the edge of the nucleolus, as well as some sugges- difficult to map specific sequences, but is sufficient for tive observations that chromatin is ordered in a specific the initial visualization of the chromatin loops them- way in the nucleus. Since then we have learned a great selves (Paulson and Laemmli, 1977; Hancock and Hughes, deal about the internal structure of the nucleus and how 1982). it is compartmentalized. We now have evidence for rRNA processing in the nucleolus (Sheer and Benavente, The presence of chromatin loops was also suggested 1990) and for RNA Polymerase II transcription at certain based on ultra-centrifugation of high salt-extracted locations in the nucleus, but not the nucleolus (van Driel, nuclei (Cook and Brazell, 1976). Based on those initial 1995). It is also known that the nucleus has an internal studies, a model was proposed where the chromatin skeleton, called the nuclear matrix, where the machinery loops are composed of the 30 nm solenoid, which is con- for DNA replication and transcription is anchored stricted between MARs (see Figure 2) (Marsden and 114 The Einstein Quarterly Journal of Biology and Medicine

Figure 1: Mapping of relative positions of unique DNA sequences within the DNA loops by progressive detachment of DNA from the nuclear matrix. Three extended DNA loops (after histone removal) are shown, one of them bearing three sequence elements (a, b and c) whose relative positions in the loop are to be determined. After mild (A) or more intensive (B) treatment of the sample with a nuclease, the matrix-bound and cleaved-off DNA are separated by differential centrifugation. The distantly located sequence element a and b, while element c will still be recov- ered in a matrix-bound DNA fraction. Knowing the average loop length and the percentage of DNA in the nuclear matrix preparations possess- ing different subsets of probed sequence elements, one can estimate the approximate positions of these elements within the loop. (Adapted from Razin and Gromova, 1995).

Laemmli, 1979). Firm evidence for the existence of these chicken β-globin domain (see Figure 4 for a schematic of loops came from the identification of the MAR the procedure). These data showing MAR sequences in sequences. These sequences were identified using different genes of one organism and in genes from dif- Exonuclease III (Exo III) sensitivity assays followed by ferent organisms provide strong evidence for the evolu- restriction mapping to identify the location of fragments tionary conservation of chromatin loops (see Cockerill that are resistant to Exo III digestion. Blasquez et al. and Garrard, 1986b for further discussion). (1989) used this technique to identify MAR sequences in the mouse light chain immunoglobulin gene (see Figure What is the significance of these loops? Do they have 3). The numbered bands (1-11) in the matrix lanes indi- any function other than as a means of packaging DNA in cate sequences that were protected from Exo III diges- the nucleus? The observation that the nuclear matrix tion, because they were bound to the nuclear matrix. It and replication origins are bound to each other was first is important to note the periodicity of the attachment made almost thirty years ago by David Comings (1968) sites. It is this periodicity that leads to the loop struc- and was demonstrated more recently by Razin et al. tures. Other groups used similar approaches to deter- (1986). These chromatin loops and their attachment to mine the MAR sites within the Drosophila histone gene the nuclear matrix through MARs may be important for cluster (Gasser and Laemmli, 1986) and the Drosophila initiating replication or for facilitating the process of hsp70 gene (Mirkovitch et al., 1984). A fortuitous obser- replication by bringing the DNA to the replication vation was that MARs often contain topoisomerase II machinery. They may also be important for localizing binding sites (Cockerill and Garrard, 1986a). Razin et al. certain sequences to specific compartments in the nucle- (1993) took advantage of the presence of topoisomerase us in order to silence them or bring them near the tran- II binding sites as a way to map MAR sequences in the scriptional machinery, but this is only speculation. Chromatin Higher Order Structure 115

Figure 2: DNA domains in nuclei. Elements that organize chromosomal loops in nuclei, termed “matrix association regions” (MARs), are depict- ed as solid rectangles (DNA sequences) and circles (proteins) (Adapted from Blasquez et al., 1989).

Evidence for DNA localization to certain domains for the tion in the nucleus, it is clear that interphase chromo- purpose of silencing or activating stretches of the DNA is somes remain separate entities and the domains they present more at the level of chromosome domains. form remain exclusive of the nucleolus.

Chromosome domains The most popular techniques for determining chromoso- mal domains are in situ hybridization with biotin labeled Uncondensed chromosomes in the interphase nucleus do probes followed by electron microscopy, fluorescent in not unravel and then randomly associate with each situ hybridization (FISH), pulse labeling of cells with other, but remain segregated from each other in their BrdU (or similar analog) followed by confocal microscopy own chromosomal domains (Manuelidis, 1990). The (see Figure 5), or injection of a fluorescent nucleotide domain organization of interphase chromosomes has analog (Cy3-AP3-dUTP) followed by video microscopy or been observed in Drosophila (Mathog and Sedat, 1989), confocal microscopy (Leitch et al., 1990; Sadoni et al., plants (Leitch et al., 1990), and mammals (Zink et al., 1999; Zink et al., 1998). The in situ hybridization tech- 1998; Manuelidis and Borden, 1988). This demonstrates niques are very powerful techniques for localizing DNA that these domains have been conserved throughout in the nucleus, but they require fixation of the cells at evolution in eukaryotic organisms. During metaphase, some point during the procedure, which could introduce the condensed chromosomes adopt what has become artifacts. These artifacts could distort the actual chromo- known as the Rabl orientation, where the centromeres somal positioning in vivo. The technique developed by are collected at one side of the cell and the telomeres Zink et al. (1998) provides a much better approach, are at the other side (Rabl, 1885; Manuelidis and Borden, because labeled chromatin can be observed in live cells, 1988). It has been suggested that during interphase, the which effectively removes the artifacts introduced from uncondensed chromosomes continue to adopt this ori- fixation. This technique provides the added benefit of entation leaving a gap in the middle of the nucleus for video recording labeled chromatin in live nuclei to deter- the nucleolus (Avivi and Feldman, 1980). Manuelidis and mine, in real time, if chromosome domains move Borden (1988) proposed that this may not be a ubiqui- through the nucleus or remain fixed in place. In fact, tous feature. Their studies of human central nervous sys- their results demonstrate that there is some shifting of tem cells using biotin labeled probes, in situ hybridiza- chromatin within domains, but the domains themselves tion, and three-dimensional reconstruction of nuclei remain fairly stationary (see Figure 6). In contrast to this, indicate that some interphase chromosomes do not Brown et al. (1999) and Bridger et al. (2000) demonstrate adopt this Rabl orientation. Regardless of their orienta- that when cells change from a proliferative state to a 116 The Einstein Quarterly Journal of Biology and Medicine

Figure 3: Association of the X gene MAR with the nuclear matrix occurs through multiple specific binding sites. A 341-bp fragment encom- passing the X gene MAR was 5' end labeled, in either the upper or lower strand, opposite to a 3' overhand. Exonuclease III digestion was per- formed after specific binding to plasmacytoma matrices (prepared using micrococcal nuclease instead of Dnase I digestion), and matrix-bound DNA was purified for electrophoretic separation on a sequencing gel. Naked DNA was similarly digested as a control. Barriers toward the 5' side (1-5) and the 3' side (6-11) are indicated. Four pairwise combinations of these barriers are shown, each of which leads to approximately 76 bp segments (Adapted from Blasquez et al., 1989). quiescent state, either certain sets of genes or whole and that this movement has a functional relationship to chromosomes change their position in the nucleus. the proliferitive state of the cell. Brown et al. (1999) showed that a set of transcriptional- ly inactive genes associate with a centromeric hete- Comings (1968) suggested that chromosomes may be rochromatin region in the nucleus in dividing lympho- localized to certain locations in the nucleus. There seems cytes, but not in non-dividing lymphocytes. They also to be some evidence that heterochromatic regions of demonstrated that when non-dividing cells are stimulat- chromosomes are localized to the nuclear membrane or ed to divide, there is a shift of this set of genes back to the outer periphery of the nucleolus while euchromatic the centromeric heterochromatin. This demonstration of regions are localized internally (Avivi and Feldman, a shift in nuclear chromosome positioning is at the gene 1980). The best example of a whole chromosome being level; Bridger et al. (2000) demonstrated a similar phe- localized to a reproducible location in the nucleus is the nomenon at the level of whole chromosomes. They inactivated X chromosome in mammals. Dyer et al. looked at the positions of chromosomes 18 and 19 in (1989) did in situ hybridization studies of human female proliferative and quiescent human fibroblasts. In prolif- cell nuclei and showed that the inactive X chromosome erative fibroblasts, chromosome 18 is associated with the almost always associates with the heterochromatic sex nuclear periphery, and chromosome 19 is associated with chromosome body (SCB). They also showed that in the nucleolus. As cells exit the cell cycle and enter a qui- hybrids between human and mouse nuclei, this close escent state after serum starvation, chromosomes 18 and association of the X chromosome with the SCB is lost. 19 lose their associations with the nuclear periphery and This seems to suggest a breakdown of the heterochro- the nucleolus, respectively. Surprisingly, when serum is matic region in these hybrids. Sanchez et al. (1997) did added back followed by several passages, chromosomes localization studies of certain chromosomes, including 18 and 19 reestablish their associations with the nuclear the X and Y chromosomes, in human neutrophil nuclei. periphery and nucleolus, respectively. This suggests that They determined that the X and Y chromosomes seemed chromosomes may actively move around the nucleus, to localize to protrusions of the neutrophil nucleus (see Chromatin Higher Order Structure 117

Figure 4: Excision of genomic DNA loops by topoisomerase II-mediated DNA cleavage at matrix attachment sites. (A) In living cells, the 30 nm chromatin fibril is arranged into loops by periodic attachment to the nuclear scaffold. The latter contains topoisomerase II, which can cut DNA at loop basements. In addition, soluble topoisomerase II can interact with DNA at a number of accessible sites, creating an apparently random cleavage pattern (as far as the low-resolution analysis by PFGE is considered). The resulting cleavage pattern (panel I) will represent the superim- posing of this random cleavage pattern over the regular pattern of cleavages at loop basements. (B) With the soluble topoisomerase II removed by high-salt extraction, the cleavages can occur only within matrix attachment sites. Varying the concentration of VM-26, it is possible to cut DNA at all loop basements (panel 3) or less frequently so that multimers of loops are also released (panel 2) (Adapted from Razin and Gromova, 1995).

Figure 7), while somatic chromosomes seem to localize in the centromere, non-randomly associate with certain a more random pattern. This argues against a general portions of the nucleus. Manuelidis (1984) demonstrated theory of specific localization of chromosomes in the this using biotin-labeled, centromere-specific satellite nucleus. An interesting observation made by Borden and DNA that was hybridized in situ to fixed tissue from Manuelidis (1988) was that in cells derived from human mouse cerebellum. She showed that individual neuronal females with epilepsy, the inactive X chromosome loses cell types display reproducible patterns of centromere its association with the heterochromatic region. Somatic localization to the nuclear membrane or to areas adja- chromosomes and the Y chromosome were also tested, cent to the nucleolus. However, no consistent pattern of and their heterochromatic regions remain associated centromere localization was seen in all of the cell types with the periphery of the nucleolus or the nuclear mem- that were studied. brane. This seems to suggest a connection between the disease state and the localization of certain chromo- Ferreira et al. (1997) and Sadoni et al. (1999) did studies somes in the nucleus. of chromosomal replication patterns in multiple cell types. They employed the use of BrdU replication-label- Certain domains within chromosomes seem to localize in ing alone, Cy3-dUTP replication-labeling alone, or a a non-random way within the nucleus. Even if the whole unique double labeling procedure where they pulse chromosome is not specifically positioned, certain por- labeled during S phase with Cy3-dUTP and then labeled tions of it may be localized. Satellite sequences, which in the next S phase with BrdU. The double labeling pro- cluster in certain portions of the chromosome, including cedure allowed them to distinguish between DNA 118 The Einstein Quarterly Journal of Biology and Medicine

Figure 5: Metaphase spread with two differently sized (non-homologous), replication-labeled chromosomes. Note that only one of the two sis- ter chromatids contains the label. Chromosomes were counterstained with DAPI (blue). Interphase nuclei in B counterstained with DAPI from the same slide as the metaphase shown in A depict replication-labeled and segregated chromatid territories. Small patches of replication-labeled chro- matin (arrow) most likely result from sister chromatid exchanges (Adapted from Zink et al., 1998).

Figure 6: Time-lapse recording of one nucleus from the neuroblastoma cell line. Cy3-AP3-dUTP-microinjected cells were grown for several cell cycles. Each panel shows a projection of optical sections. 3D stacks were recorded every 20 minutes for 5 hours. The nucleus performed rota- tional and translational movements during the entire period. The framed segregated territory is clearly separated from the rest of the labeled ter- ritories. After a major initial rearrangement (indicated by large arrowheads), this territory maintained its overall shape. The inset (upper right) within the first six panels shows an enlargement (factor 2.3) of subdomains (arrow). The subdomains display changing patterns of foldings and extensions (some examples are indicated by arrowheads) (Adapted from Zink et al., 1998). Chromatin Higher Order Structure 119

Figure 7: Deconvolved volumetric views of neutrophil nuclei from healthy females (top two panel pairs) and males (bottom two panel pairs). The leftmost panel of each corresponds to a volumetric view of a nucleus stained for total DNA with DAPI (negative image). The rightmost panel of each pair corresponds to a volumetric view of a nucleus hybridized for the X chromosome and the Y chromosome centromere sequences. In these panels, the outline of the nuclei is indicated by a white dotted line and the asterisks mark the position of the sex-specific appendages. Notice the presence of an X chromosome in the appendage of the female nucleus, a Y chromosome in the appendage of the male nucleus, and the absence of chromosome pairing when the sex chromosomes reside in the same lobe. Bar represents 5 µm (Adapted from Sanchez et al., 1997). labeled at a first S phase and DNA labeled at the next S sion patterns of different genes. It is noteworthy that phase. Using these labeling techniques, Ferreira et al. housekeeping genes are mostly localized in R bands, (1997) demonstrated that early replicating portions of whereas tissue specific genes are almost exclusively locat- chromosomes (R bands that are identified by a light ed in G/Q bands (Craig and Bickmore, 1993, 1994). Giemsa staining on chromosomes) preferentially localize to interior regions of the nucleus, while late replicating Conclusion portions (G/Q bands that are identified as dark Giemsa staining on chromosomes) preferentially localize to Two levels of chromatin structure have been presented, peripheral regions. Sadoni et al. (1999) verified these that of chromatin loops and chromosome domains. results (see Figure 8) and further showed, using their Marsden and Laemmli (1979) proposed that the 30 nm double labeling technique, that S phase labeling patterns solenoid chromatin filaments were attached at regular are maintained from division to division. So, not only are intervals to the nuclear matrix forming chromatin loops. regions of chromosomes that replicate at different times The sequences that mediate the attachment of the DNA in S phase localized to different portions of the nucleus, to the nuclear matrix (MARs) have been extensively stud- but that pattern is maintained in a cell type-specific man- ied (Blasquez et al., 1989; Gasser and Laemmli, 1986; ner with every cell division. These results provide a poten- Mirkovitch et al., 1984; Razin et al., 1993). The attach- tial structural explanation for the tissue-specific expres- ment of chromatin to the nuclear matrix has also been 120 The Einstein Quarterly Journal of Biology and Medicine

Figure 8: S phase replication labeling patterns are preserved. Cells ([a and b] Hv, human diploid fibroblasts; [c and d] SH-EP N14 human neu- roblastoma cells; [e and f] HeLa cells; [g and h] CHO cells; and [i and j] C2C12 mouse myoblasts) were replication-labeled with BrdU (a-d, i, and j; 30-min pulses) or Cy3-dUTP (e-h). Cells were fixed immediately after BrdU labeling or 30 minutes after microinjection of Cy3-dUTP to obtain labeled S phase ells (left, a,c,e,g, and i). S phase cells display the typical replication labeling patterns (indicated on the left): (a) type I, (c) type II, (e) type III, (g) type IV, and (i) type V. Similarly labeled cells were grown for one to five hours. after labeling and fixed after this growth period (right, b,d,f,h, and j). The presence of similar patterns (types indicated on the right) several days after S phase labeling suggests that DNA repli- cating at a particular time during S phase always occupies similar nuclear positions. Arrowheads mark the nucleoli that are always unlabeled. Arrowheads in c-e mark perinucleolar label. Arrows in b and d indicate unlabeled chromosome territories. The presence of unlabeled territories demonstrates that cells do not display similar patterns because they stopped cycling, but rather that they went through at least two mitoses after labeling. Note also the two very close nuclei in f displaying similar type III patterns, suggesting that similar patterns were conserved in sister nuclei after cell division. All pictures display single light optical sections through midnuclear planes except i and j, which show epifluorescence images. Bar in a is similar for all images (Adapted from Sadoni et al., 1999). Chromatin Higher Order Structure 121

implicated in the process of replication initiation (Razin under which the two levels of chromatin organization et al., 1986). The suggestion that interphase chromo- can be understood. Undoubtedly, more work needs to somes are specifically positioned in the nucleus was first be done in order to fully understand the functional made by Comings in 1968. It has since been shown that organization of the nucleus and the functional relation- chromosomes occupy specific domains in the nucleus and ships between chromatin loops, chromosome territories, do not overlap in the interphase nucleus (Manuelidis, and the replication and transcription machinery. 1990; Leitch et al., 1990). There is also a great deal of evi- dence to suggest that certain portions of chromosomes Acknowledgments are non-randomly arranged in the nucleus (Manuelidis, 1984; Ferreira et al., 1997; Sadoni et al., 1999). The author would like to thank Rajarshi Ghosh, Nicole Rempel, and Dr. U. Thomas Meier for their help and con- How do these two levels of chromatin organization con- structive criticism. nect with one another? What is their functional connec- tion, if any? This author believes that the chromatin References attachments to the nuclear matrix in the form of chro- matin loops may provide the “foundation” for the local- Avivi, L. and Feldman, M. (1980) Arrangement of chromosomes in the ization of, at least, specific domains of chromosomes if interphase nucleus of plants. Hum. Genet. 55:281-295. not whole chromosomes. Some of the most striking evi- Blasquez, V.C., Sperry, A.O., Cockerill, P.N., and Garrard, W.T. (1989) dence for a structural basis to this model comes from the Protein:DNA interactions at chromosomal loop attachment sites. work of Ma et al. (1999). When they did a classical Genome 31:503- 509. extraction for nuclear matrix in the absence of DNaseI followed by FISH, they showed that chromosome Borden, J. and Manuelidis, L. (1988) Movement of the X chromosome in domains remain intact. In contrast to this, when they did epilepsy. Science 242:1687-1691. a complete matrix extraction by doing RNase A treat- Bridger, J.M., Boyle, S., Kill, I.R., and Bickmore, W.A. (2000) Remodelling ment, to remove interactions with RNAs or RNPs, fol- of nuclear architecture in quiescent and senescent human fibroblasts. lowed by 2.0 M NaCl, the chromosome domains are dis- Curr. Biol. 10:149-152. rupted. This suggests that chromosome domains are attached to the nuclear matrix most likely through the Brown, K.E., Baxter, J., Graf, D., Merkenschlager, M., and Fisher, A.G. attachment of chromatin loops to the matrix. This model (1999) Dynamic repositioning of genes in the nucleus of lymphocytes is also functionally supported by the observation that preparing for cell division. Mol. Cell 3:207-217. chromatin is attached to the nuclear envelope at the site Cameron I.L., Pavlat, W.A., and Jeter, J.R. (1979) Chromatin substructure: of replication origins (Razin, 1986), which would bring an electron microscopic study of thin-sectioned chromatin subjected to these regions of chromosomes near the nuclear pore sequential protein extraction and water swelling procedures. Anat. Rec. complexes. The proximity of the DNA and replication 194:547-562. complexes near the nuclear pore complexes means that these complexes have ready access to nucleotide precur- Cockerill, P.N. and Garrard, W.T. (1986a) Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in region sors and other cytoplasmic factors. Razin and Gromova’s containing topoisomerase II sites. Cell 44:273-282. (1995) “channels model” of nuclear matrix structure incorporates some of this idea. Their model postulates Cockerill, P.N. and Garrard, W.T. (1986b) Chromosomal loop anchorage that the nuclear matrix forms channels near the nuclear sites appear to be evolutionarily conserved. FEBS Lett. 204:5-7. pore complexes. These channels have large lateral exten- sions coming off of the main channel. The chromatin Comings, D.E. (1968) The rationale for an ordered arrangement of chro- matin in the interphase nucleus. Am. J. Hum. Genet. 20:440-460. loops are then attached to the matrix along these chan- nels. The model provides a good explanation for the dif- Cook, P.R and Brazell, I.A. (1976) Conformational constraints in nuclear ferential sensitivity of loop attachment sites to nuclease DNA. J. Cell Sci. 22:287-302. treatment after the different extraction methods. It also supports the observation that newly synthesized mRNAs Craig, J.M. and Bickmore, W.A. (1993) Chromosome bands–flavours to seem to track toward the nuclear pore complex and pre- savour. BioEssays 15:349-354. sumably out of the nucleus (Lawrence et al., 1989; Craig, J.M. and Bickmore, W.A. (1994) The distribution of CpG islands in Kramer et al., 1994). These nuclear matrix channels may mammalian chromosomes. Nat. Genet. 7:376-382. be the means by which mRNAs exit the nucleus; this remains to be seen. These channels may exist within Cremer, T., Kurz, A., Zirbel, R., Dietzel, S., Rinke, B., Schrock, E., Speicher, chromosome territories or, as Cremer et al. (1993) sug- M.R., Mathieu, U., Jauch, A., and Emmerich, P. (1993) Role of chromo- gest, they may exist between chromosome territories. some territories in the functional compartmentalization of the cell nucle- They suggested that the interchromosomal spaces may us. Cold Spring Harb. Symp. Quant. Biol. 58:777-792. provide channels for newly synthesized mRNAs and Dyer, K.A., Canfield, T.K., and Gartler, S.M. (1989) Molecular cytological other precursor molecules to move in and out of the differentiation of active from inactive X domains in interphase: implica- nucleus. Both of these models provide a framework tions for X chromosome inactivation. Cytogenet. Cell Genet. 50:116-120. 122 The Einstein Quarterly Journal of Biology and Medicine

Felsenfeld, G. and McGhee, J.D. (1986) Structure of the 30 nm chromatin Mirkovitch, J., Mirault, M.E., and Leammli, U.K. (1984) Organization of fiber. Cell 44:375-377. the higher-order chromatin loop: Specific DNA attachment sites on nuclear scaffold. Cell 39:223-232. Ferreira, J., Paolella, G., Ramos, C., and Lamond, A.I. (1997) Spatial organ- ization of large-scale chromatin domains in the nucleus: A magnified Nelson, W.G., Peinta, K.J., Barrack, E.R., and Coffey, D.S. (1986) The role view of single chromosome territories. J. Cell Biol. 139:1597-1610. of the nuclear matrix in the organization and function of DNA. Ann. Rev. Biophys. Biophys. Chem. 15:457-475. Finch, J.T. and Klug, A. (1976) Solenoidal model for superstructure in chromatin. Proc. Natl. Acad. Sci. U.S.A. 73:1897-1901. Olins, A.L. and Olins, D.E. (1974) Spheroid chromatin units (v bodies). Science 183:330-332. Gasser, S.M. and Laemmli, U.K. (1986) The organization of chromatin loops: characterization of a scaffold attachment site. EMBO J. 5:511-518. Paulson, J.R. and Laemmli, U.K. (1977) The structure of histone-depleted metaphase chromosomes. Cell 12:817-828. Hancock, R. and Hughes, M.E. (1982) Organization of DNA in the eukary- otic nucleus. Biol. Cell. 44:201-212. Rabl C. (1885) Über Zellteilung. Morphol. Jahrb. 10:214-330.

Kramer, J., Lachar, L., and Bingham, P.M. (1994) Nuclear pre-mRNA Razin, S.V. and Gromova, I.I. (1995) The channels model of nuclear matrix metabolism: channels and tracks. Trends Cell. Biol. 4:35-37. structure. BioEssays 17:443-450.

Lawrence, J.B., Singer, R.H., and Marselle, L.M. (1989) Highly localized Razin, S.V., Hancock, R., Iarovaia, O., Westergaard, O., Gromova, I., and tracks of specific transcripts within interphase nuclei visualized by in situ Georgiev, G.P. (1993) Structural-functional organization of chromosomal hybridization. Cell 57:493-502. DNA domains. Cold Spring Harb. Symp. Quant. Biol. 63:25-35.

Leitch, A.R., Mosgöller, W., Schwarzacher, T., Bennett, M.D., and Heslop- Razin, S.V., Kekelildze, M.G., Lukanidin, E.M., Sherrer, K., and Georgiev, Harrison, J.S. (1990) Genomic in situ hybridization to sectioned nuclei G.P. (1986) Replication origins are attached to the nuclear skeleton. shows chromosome domains in grass hybrids. J. Cell Sci. 95:335-341. Nucleic Acids Res. 14:8189-8207.

Ma, H., Siegel, A.J., and Berezney, R. (1999) Association of chromosome territories with the nuclear matrix: Disruption of human chromosome Sadoni, N., Langer, S., Fauth, C., Bernardi, G., Cremer, T., Turner, B.M., and territories correlates with the release of a subset of nuclear matrix pro- Zink, D. (1999) Nuclear organization of mammalian genomes: Polar chro- teins. J. Cell. Biol. 146:531-542. mo-some territories build up functionally distinct higher order compart- ments. J. Cell Biol. 146:1211-1226. Manuelidis, L. (1984) Different central nervous system cell types display distinct and nonrandom arrangements of satellite DNA sequences. Proc. Sanchez, J.A., Karni, R.J., and Wangh, L.J. (1997) Fluorescent in situ Natl. Acad. Sci. U.S.A. 81:3123-3127. hybridization (FISH) analysis of the relationship between chromosome location and nuclear morphology in human neutrophils. Chromosoma Manuelidis, L. (1985) Individual interphase chromosome domains 106:168-177. revealed by in situ hybridization. Hum. Genet. 71:288-293. Scheer, U. and Benavente, R. (1990) Functional and dynamic aspects of Manuelidis, L. (1990) A view of interphase chromosomes. Science the mammalian nucleolus. BioEssays 12:14-21. 250:1533-1540. Strouboulis, J. and Wolffe, A.P. (1996) Functional compartmentalization Manuelidis, L. and Borden, J. (1988) Reproducible compartmentalization of of the nucleus. J. Cell Sci. 109:1991-2000. individual chromosome domains in human CNS cells revealed by in situ hybridization and three-dimensional reconstruction. Chromosoma 96:397-410. van Driel, R., Wansink, D.G., van Steensel, B., Grande, M.A., Schul, W., and de Jong, L. (1995) Nuclear domains and the nuclear matrix. Int. Rev. Marsden, M.P. and Leammli, U.K. (1979) Metaphase chromosome struc- Cytol. 162A:151-189. ture: evidence for a radial loop model. Cell 17:849-858. Zink, D., Cremer, T., Saffrich, R., Fischer, R., Trendelenburg, M.F., Mathog, D. and Sedat, J.W. (1989) The three-dimensional organization of Ansorge, W., and Stelzer, E.H.K. (1998) Structure and dynamics of polytene nuclei in male Drosophila melanogaster with compound XY or human interphase chromosome territories in vivo. Hum. Genet. ring X chromosomes. Genetics 121:293-311. 102:241-251.