DNA loop domains in mammalian spermatozoa

The highly condensed and tightly packaged DNA of hamster spermatozoa was found to be organized into topologically constrained DNA loop domains attached at their bases to a nuclear matrix. The loop domains of the sperm nuclei differed from somatic cell loop domains from the same animal in two aspects. Sperm loop domains were 60% smaller than somatic cell loop domains, with an average DNA length of 46±7 kb in sperm as compared with 16±11 kb in brain. Secondly, unlike virtually all somatic cell DNA known which is negatively supercoiled, sperm DNA was devoid of detectable supercoiling. The presence of the loop domain structure in the highly condensed DNA of mammalian spermatozoa suggests that this motif is a fundamental aspect of eukaryotic DNA organization.     

DNA packaging and organization in mammalian spermatozoa: comparison with somatic cells

Mammalian sperm DNA is the most tightly compacted eukaryotic DNA, being at least sixfold more highly condensed than the DNA in mitotic chromosomes. To achieve this high degree of packaging, sperm DNA interacts with protamines to form linear, side-by-side arrays of chromatin. This differs markedly from the bulkier DNA packaging of somatic cell nuclei and mitotic chromosomes, in which the DNA is coiled around histone octamers to form nucleosomes. The overall organization of mammalian sperm DNA, however, resembles that of somatic cells in that both the linear arrays of sperm chromatin and the 30-nm solenoid filaments of somatic cell chromatin are organized into loop domains attached at their bases to a nuclear matrix. In addition to the sperm nuclear matrix, sperm nuclei contain a unique structure termed the sperm nuclear annulus to which the entire complement of DNA appears to be anchored when the nuclear matrix is disrupted during decondensation. In somatic cells, proper function of DNA is dependent upon the structural organization of the DNA by the nuclear matrix, and the structural organization of sperm DNA is likely to be just as vital to the proper functioning of the spermatozoa.     

Ribosomal DNA sequences attached to the nuclear matrix

The organization of rat liver ribosomal DNA (rDNA) as matrix-attached DNA loops was examined using a protocol which fractionates chromatin from discrete regions of DNA loops. Southern blot analysis of matrix-attached and solubilized chromatin DNA fragments demonstrated that rDNA is associated with the matrix via its 5' and 3' nontranscribed spacer sequences (NTS). Although the 45 S rRNA coding sequences were approximately threefold enriched in matrix preparations, the recovery of this DNA (unlike the NTS) was dependent on the extent of nuclease digest and proportional to the length of the matrix-attached DNA fragments. The data suggest that rDNA is organized as matrix-attached DNA loops and only the NTS are directly involved in matrix binding. Further, we demonstrated that while the kinetics and extent of nuclease digestion were similar in all regions of the DNA loops, the nuclease digestion pattern of bulk nuclear and matrix DNA showed a typical nucleosome organization, but the rDNA fragments retained with the nuclear matrix did not.     

Mapping long-range chromatin organization within the chicken alpha-globin gene domain using oligonucleotide DNA arrays

We have analyzed the organization of the chicken alpha-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 alpha-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 alpha-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.     

Molecular dissection of a specific nuclear domain: the chromatin region of the ribosomal gene cluster in Xenopus laevis.

Molecular dissection of the nuclear domain corresponding to the ribosomal chromatin cluster was investigated. The experimental scheme was based on the ability of restriction enzymes to digest the whole genome without affecting this region (several megabases in length). Such a strategy involved the judicious choice of restriction enzymes, which is possible in Xenopus laevis, where the rDNA sequence is known and the repeated units are organized into one unique cluster. SalI, XhoI, and EcoRV digestion produced frequent cutting of the genome leaving the ribosomal cluster intact. Isolation of the rDNA cluster was confirmed by separation of the digested DNA by pulsed-field electrophoresis. When applied to purified nuclei, this approach allowed the isolation of the ribosomal chromatin cluster under very mild conditions: no cleavages (either enzymatic or mechanical) were detectable. Since the purification scheme depends only on the DNA sequence outside of the rDNA cluster, it permits the obtention of this domain in different functional states. Electron microscopic analysis demonstrated that the domain organization is substantially preserved and maintains its looped organization (the size and the full number of loops were preserved). This purification scheme provides a powerful tool for studying the structure-function relationships within the ribosomal nuclear domain.     

Spatial Organization of the Chicken α-Globin Gene Domain in Cells of Different Origins

Hybridization with an oligonucleotide array was used to map the regions of DNA anchorage to the nuclear matrix. Matrix-associated DNA served as a hybridization probe. To obtain the oligonucleotide array, 60-mer oligonucleotides regularly distributed throughout the genome region of interest at 2-kb intervals were immobilized on a nylon filter. The organization of DNA into loop domains was studied in a 100-kb region of chicken chromosome 16, including the α -globin gene cluster. A 40-kb DNA loop, which was fixed to the nuclear matrix and harbored all α-globin genes, was observed in erythroid cells. One of its anchorage regions colocalized with matrix associated region (MAR) and an insulator found previously in the 5′ region of the chicken α-globin gene domain. The spatial (domain-loop) organization of the α-globin gene cluster in lymphoid cells proved to be strikingly different from that in erythroid cells.     

Structural differences in the DNA binding domains of human p53 and its C. elegans ortholog Cep-1

The DNA binding domains of human p53 and Cep-1, its C. elegans ortholog, recognize essentially identical DNA sequences despite poor sequence similarity. We solved the three-dimensional structure of the Cep-1 DNA binding domain in the absence of DNA and compared it to that of human p53. The two domains have similar overall folds. However, three loops, involved in DNA and Zn binding in human p53, contain small alpha helices in Cep-1. The alpha helix in loop L3 of Cep-1 orients the side chains of two conserved arginines toward DNA; in human p53, both arginines are mutation hotspots, but only one contacts DNA. The alpha helix in loop L1 of Cep-1 repositions the entire loop, making it unlikely for residues of this loop to contact bases in the major groove of DNA, as occurs in human p53. Thus, during evolution there have been considerable changes in the structure of the p53 DNA binding domain.     

Tertiary structure and computer modeling of plant 5S ribosomal RNA

A new model of secondary and tertiary structure of higher plant 5S rRNA is proposed. It consists of three domains. Domain alpha includes stem I and loop A; domain beta contains stems II and III and loops B and C; domain gamma consists of stems IV and V and loops D and E. We propose that the domains beta and gamma adopt RNA-A like structure due to irregularities caused by the different in size internal loops B and E and the bulges occurring in the model. A suggested bending of RNA could bring single stranded fragments of domains beta and gamma close enough to each other to allow tertiary interactions. The new model of plant 5S rRNA differs from those suggested previously for eukaryotic 5S rRNA, by arrangement of the domains beta and gamma and the base pairing scheme of domain gamma. The model is based on our results of partial digestion obtained with single and double strand specific nucleases. The experimental results were confirmed by computer aided secondary structure prediction analysis of all higher plant 5S rRNAs and computer modeling using energy minimalization approach. Further support of our model have been provided by experiments including alpha sarcin, ribonuclease H and chemical modifications.     

Higher plant 5S rRNAs share common secondary and tertiary structure. A new three domains model

A new model of secondary and tertiary structure of higher plant 5S RNA is proposed. It consists of three helical domains: domain alpha includes stem I; domain beta contains stems II and III and loops B and C; domain gamma consists of stems IV and V and loops D and E. Except for, presumably, a canonical RNA-A like domain alpha, the two remaining domains apparently adopt a perturbed RNA-A structure due to irregularities within internal loops B and E and three bulges occurring in the model. Bending of RNA could bring loops B and E and/or C and D closer making tertiary interactions likely. The model differs from that suggested for eukaryotic 5S rRNA, by organization of domain gamma. Our model is based on the results of partial digestion obtained with single- and double-strand RNA specific nucleases. The proposed secondary structure is strongly supported by the observation that crude plant 5S rRNA contains abundant RNA, identified as domain gamma of 5S rRNA. Presumably it is excised from the 5S rRNA molecule by a specific nuclease present in lupin seeds. Experimental results were confirmed by computer-aided secondary structure prediction analysis of all higher plant 5S rRNAs. Differences observed between earlier proposed models and our proposition are discussed.     

DNA Topoisomerase II Is Not Detectable on Lampbrush Chromosomes but Enriched in the Amplified Nucleoli of Xenopus Oocytes

In somatic cells DNA topoisomerase II (topo II) is thought to be involved in the domain organization of the genome by anchoring the basis of chromatin loops to a chromosomal scaffold. Lampbrush chromosomes of amphibian oocytes directly display this radial loop organization in cytological preparations. In order to find out whether topo II may play a role in the organization of these meiotic chromosomes, we performed immunofluorescence studies using antibodies against Xenopus topo II. Our results indicate that topo II is apparently absent from lampbrush chromosomes and is hence unlikely to act as a "fastener" of the numerous lateral chromosomal loops. Topo II was, however, enriched in the amplified nucleoli of Xenopus oocytes.     

Nuclear matrix support of DNA replication

In higher eukaryotic cells, DNA is tandemly arranged into 10(4) replicons that are replicated once per cell cycle during the S phase. To achieve this, DNA is organized into loops attached to the nuclear matrix. Each loop represents one individual replicon with the origin of replication localized within the loop and the ends of the replicon attached to the nuclear matrix at the bases of the loop. During late G1 phase, the replication origins are associated with the nuclear matrix and dissociated after initiation of replication in S phase. Clusters of several replicons are operated together by replication factories, assembled at the nuclear matrix. During replication, DNA of each replicon is spooled through these factories, and after completion of DNA synthesis of any cluster of replicons, the respective replication factories are dismantled and assembled at the next cluster to be replicated. Upon completion of replication of any replicon cluster, the resulting entangled loops of the newly synthesized DNA are resolved by topoisomerases present in the nuclear matrix at the sites of attachment of the loops. Thus, the nuclear matrix plays a dual role in the process of DNA replication: on one hand, it represents structural support for the replication machinery and on the other, provides key protein factors for initiation, elongation, and termination of the replication of eukaryotic DNA.     

Remodeling of chromatin loops does not account for specification of replication origins during Xenopus development

We have investigated the possible relationship between replicons and chromatin loops during Xenopus development. In early embryos, replication of the ribosomal RNA genes (rDNA) can initiate at apparently any sequence. Nevertheless, the need for a regular spacing of replication origins suggests that some periodic chromatin folding might dictate which sites are actually used for initiation. After the midblastula transition, replication initiation is restricted to the rDNA intergenic spacers. A remodeling of chromatin folding could account for this change in origin usage. Here, it is reported that nuclear matrix anchorage of the Xenopus rDNA occurs at multiple, apparently random sequences, throughout embryonic development as well as in adult cells. In vitro matrix rebinding assays confirmed the lack of specific anchoring sequences in the rDNA, before as well as after specific replication origins are established. Thus, no change in loop attachment sites could explain the change in origin usage at this locus. Nonspecific loop anchorage was a special feature of the rDNA locus, since the same nuclear matrices were able selectively to bind the scaffold attachment region (SAR) of the Drosophila histone gene cluster in vitro. Blastula and gastrula nuclear matrices bound a higher amount of SAR sequences than matrices from later stages or adult cells. This developmental change in SAR binding might explain the increase in size of the bulk of genomic DNA loops that occurs after the gastrula stage. However, no change in chromatin loop organization that could explain the midblastula stage transition from small to large replicons was observed. Received: 15 January 1998; in revised form: 4 March 1998 / Accepted: 9 March 1998     

Chromosome length and DNA loop size during early embryonic development of Xenopus laevis

The looped organization of the eukaryotic genome mediated by a skeletal framework of non-histone proteins is conserved throughout the cell cycle. The radial loop/scaffold model envisages that the higher order architecture of metaphase chromosomes relies on an axial structure around which looped DNA domains are radially arranged through stable attachment sites. In this light we investigated the relationship between the looped organization and overall morphology of chromosomes. In developing Xenopus laevis embryos at gastrulation, the bulk of the loops associated with histone-depleted nuclei exhibit a significant size increase, as visualized by fluorescence microscopy of the fully extended DNA halo surrounding high salt treated, ethidium bromide stained nuclei. This implies a reduction in the number of looped domains anchored to the supporting nucleoskeletal structure. The cytological analysis of metaphase plates from acetic acid fixed whole embryos, carried out in the absence of drugs inducing chromosome condensation, reveals a progressive thickening and shortening of metaphase chromosomes during development. We interpret these findings as a strong indication that the size and number of DNA loops influence the thickness and length of the chromosomes, respectively. The quantitative analysis of chromosome length distributions at different developmental stages suggests that the shortening is timed differently in different embryonic cells.