Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase II sites

A region immediately adjacent to the origin of replication of bovine papilloma virus type 1 interacts in vitro with the nuclear matrix

We have investigated the interaction of the 69% transforming fragment of the Bovine Papilloma Virus type 1 (BPV1) with the nuclear matrix from 1361.5 cells (NIH-3T3 cells transformed by a BPV chimeric construct). In vitro studies performed with end-labelled DNA fragments and nuclear matrices prepared using a high-salt extraction procedure demonstrate the binding of a 672 bp fragment adjacent to the viral origin of replication and containing the plasmid maintenance sequence (PMS-1). This fragment can be cleaved into two pieces (393 and 279 bp), both interacting equally well with the nuclear matrix. This indicates that a least two regions of the 672 bp DNA fragment are involved in the interaction.

Silencing of the expression of the immunoglobulin kappa gene in non-B cells

Although the activating factor NF-kappa B can be present in the nucleus of many cell types, transcription and rearrangement of the immunoglobulin kappa chain gene is restricted to cells of the B lineage. Part of this specificity is determined by sequences within the major intron of the kappa gene that specifically silence gene expression in non-B cells (T cells and HeLa cells). These sequences are found in a 232-bp fragment located 5' of the NF-kappa B binding sequence of the enhancer. When this fragment is added back upstream of an active NF-kappa B site, it specifically decreases the expression of a linked gene by more than 10-fold in activated T cells but it has no effect on expression in B cells. The kappa silencer region acts in an orientation- and distance-independent manner and appears to be composed of multiple negative elements. The kappa silencer may act to restrict transcription and rearrangement of the C kappa locus to cells of the B lineage.

A position-effect assay for boundaries of higher order chromosomal domains

Eukaryotic chromosomes are thought to be organized into a series of discrete higher order chromatin domains. This organization is believed to be important not only in the compaction of the chromatin fiber, but also in the utilization of genetic information. Each domain would define an independent unit of gene activity, insulated from the regulatory influences of adjacent domains. Critical to this model of chromosome organization and function are the domain boundaries: the special nucleoprotein structures that delimit each higher order domain and segregate the chromosome into units of independent gene activity. In the work reported here we have tested whether two putative domain boundaries, scs and scs', from the Drosophila 87A7 heat shock locus can establish a domain of independent gene activity in vivo and insulate against chromosomal position effects.

Gene deletions causing human genetic disease: mechanisms of mutagenesis and the role of the local DNA sequence environment

Reports describing short (less than 20 bp) gene deletions causing human genetic disease were collated in order to study underlying causative mechanisms. Deletion breakpoint junction regions were found to be non-random both at the nucleotide and dinucleotide sequence levels, an observation consistent with an endogenous sequence-directed mechanism of mutagenesis. Direct repeats of between 2bp and 8bp were found in the immediate vicinity of all but one of the 60 deletions analysed. Direct repeats are a feature of a number of recombination, replication or repair-based models of deletion mutagenesis and the possible contribution of each to the spectrum of mutations examined was assessed. The influence of parameters such as repeat length and length of DNA between repeats was studied in relation to the frequency, location and extent of these deletions. Findings were broadly consistent with a slipped mispairing model but the predicted deletion of one whole repeat copy was found only rarely. A modified version of the slipped mispairing hypothesis was therefore proposed and was shown to possess considerable explanatory value for approximately 25% of deletions examined. Whereas the frequency of inverted repeats in the vicinity of gene deletions was not significantly elevated, these elements may nevertheless promote instability by facilitating the formation of secondary structure intermediates. A significant excess of symmetrical sequence elements was however found at sites of single base deletions. A new model to explain the involvement of symmetric elements in frameshift mutagenesis was devised, which successfully accounted for a majority of the single base deletions examined. In general, the loss of one or a few base pairs of DNA was found to be more compatible with a replication-based model of mutagenesis than with a recombination or repair hypothesis. Seven hitherto unrecognized hotspots for deletion were noted in five genes (AT3, F8, HBA, HBB and HPRT). Considerable sequence homology was found between these different sites, and a consensus sequence (TGA/GA/GG/TA/C) was drawn up. Sequences fitting this consensus (i) were noted in the immediate vicinity of 41% of the other (sporadic) gene deletions, (ii) were found frequently at sites of spontaneous deletion in the hamster APRT gene, (iii) were found to be associated with many larger human gene deletions/translocations, (iv) act as arrest sites for human polymerase alpha during DNA replication and (v) have been shown by in vitro studies of human polymerase alpha to be especially prone to frameshift mutation.(ABSTRACT TRUNCATED AT 400 WORDS)

Silencing of the expression of the immunoglobulin kappa gene in non-B cells

Although the activating factor NF-kappa B can be present in the nucleus of many cell types, transcription and rearrangement of the immunoglobulin kappa chain gene is restricted to cells of the B lineage. Part of this specificity is determined by sequences within the major intron of the kappa gene that specifically silence gene expression in non-B cells (T cells and HeLa cells). These sequences are found in a 232-bp fragment located 5' of the NF-kappa B binding sequence of the enhancer. When this fragment is added back upstream of an active NF-kappa B site, it specifically decreases the expression of a linked gene by more than 10-fold in activated T cells but it has no effect on expression in B cells. The kappa silencer region acts in an orientation- and distance-independent manner and appears to be composed of multiple negative elements. The kappa silencer may act to restrict transcription and rearrangement of the C kappa locus to cells of the B lineage.

Scaffold-attached regions from the human interferon beta domain can be used to enhance the stable expression of genes under the control of various promoters

We have transfected DNA corresponding to the complete chromatin domain of human interferon beta (huIFN-beta) gene into mouse L cells. In this construct, which is flanked by scaffold-attached regions (SARs), the gene's transcription was enhanced 20-30-fold with respect to DNAs containing only the immediate regulatory elements. To elucidate the role of SAR elements in the transcriptional enhancement, their position was varied relative to several artificial promoter-gene combinations. It was found that SARs enhance general promoter functions in an orientation- and partially distance-independent manner; their effect is restricted to the integrated state of transfected templates. During the phase of transient expression, SAR elements were generally found to have an antagonizing effect.

Mapping of the duplicated rabbit immunoglobulin kappa light chain locus

The rabbit has two isotypic forms of the immunoglobulin kappa light chain, K1 and K2, which probably arose by duplication. In the normal rabbit, only traces of K2 light chains are produced. However, K2 levels are elevated in allotype-suppressed rabbits and in the Basilea strain which does not produce K1 because of a K1 mRNA splice site mutation. Previous cloning and sequencing showed that each isotype has its own set of J kappa genes but it was not known whether the two isotypes utilize shared or separate sets of V kappa genes. In addition, although genetic linkage of allotypes associated with the K1 and K2 genes has been demonstrated, physical linkage had not been previously demonstrated by overlapping cosmid or phage clones. We used pulsed field and transverse alternating field electrophoresis to obtain megabase maps and to estimate the size of the duplication of the rabbit kappa light chain locus. We found that the two C kappa genes are about 1 megabase apart. One explanation for the poor expression of K2, could be great physical distance from V kappa genes. However, we found that there are V kappa, J kappa and C kappa 2 genes within a approximately 105-kb fragment. Thus, physical distance of V kappa from C kappa 2 may not be the basis for poor K2 expression.

Association of topoisomerase II with the hepatoma cell nuclear matrix: the role of intermolecular disulfide bond formation

Previous studies have resulted in conflicting data regarding the recovery of the nuclear enzymes topoisomerase (topo) II and topo I in the nuclear matrix fraction. In the present study we have assessed the effect of systematically altering a single extraction procedure on the distribution of these enzymes during the subfractionation of nuclei from HTC hepatoma tissue culture cells. When nuclear monolayers (prepared by treating attached cells in situ with the neutral detergent Nonidet-P40 at 4 degrees C) were isolated in the presence of the irreversible sulfhydryl blocking reagent iodoacetamide, subsequent treatment with DNase I and RNase A followed by 1.6 M NaCl resulted in structures which were extensively depleted of intranuclear components as assessed by phase contrast microscopy and conventional transmission electron microscopy. These structures contained 12 +/- 4% of the total protein present in the original nuclear monolayers. The lamins and polypeptides with molecular weights comparable to those of actin and vimentin were the predominant polypeptides present on SDS-polyacrylamide gels. Western blotting revealed that less than 5% of the total nuclear topo II molecules were present in these structures. In contrast, when the sulfhydryl cross-linking reagent sodium tetrathionate (NaTT) was substituted for iodoacetamide, the same extraction procedure yielded structures containing components of the nucleolus and an extensive intranuclear network. These structures contained a wide variety of nonlamin, nonhistone nuclear polypeptides including 23 +/- 4% of the total nuclear topo II. SDS-polyacrylamide gel electrophoresis performed under nonreducing conditions revealed that topo II in these nuclear matrices was present as part of a large disulfide cross-linked complex. Treatment of these structures with reducing agents in 1.6 M NaCl released the topo II. In contrast, topo I did not form disulfide cross-linked oligomers and was not detectable in any of these nuclease- and salt-resistant structures prepared at 4 degrees C. To assess the effect of in vitro heat treatment on the distribution of the topoisomerases, nuclear monolayers (isolated in the absence of iodoacetamide and NaTT) were heated to 37 degrees C for 1 h prior to treatment with nucleases and 1.6 M NaCl. The resulting structures (which retained 26 +/- 5% of the total nuclear protein) were morphologically similar to the NaTT-stabilized nuclear matrices and contained 15 +/- 4% of the total nuclear topo II. High-molecular-weight disulfide cross-linked oligomers of topo II were again demonstrated. Attempts to demonstrate these disulfide cross-linked oligomers in intact cells were unsuccessful.

Identification of a T-cell-specific transcriptional enhancer located 3' of C gamma 1 in the murine T-cell receptor gamma locus

A transcriptional enhancer element has been localized 3 kilobases 3' of the murine T-cell receptor C gamma 1 locus using a chloramphenicol acetyltransferase reporter gene construct. As a monomer the enhancer functions only in PEER gamma delta cells and Jurkat alpha beta cells of the T-cell lines tested. However, a tetramer of the enhancer functions in virtually all T-cell lines tested, including alpha beta T-cell lines, but not in other cell types. These results suggest that elements other than the enhancer are responsible for the failure of rearranged C gamma 1 genes to be expressed in alpha beta T cells. The enhancer has been localized to a 200-base-pair Rsa I restriction fragment, which contains sequence motifs similar to those found in the other T-cell receptor enhancers but not in the immunoglobulin enhancers.

Chromosome assembly in vitro: topoisomerase II is required for condensation

The role of topoisomerase II (topo II) in chromosome condensation was studied in a mitotic extract derived from Xenopus eggs by specific immunodepletion. HeLa nuclei, which have a high complement of endogenous topo II, are converted to mitotic chromosomes in the topo II-depleted extract equally well as in the control. Chicken erythrocyte nuclei, however, which have a very low content of topo II, do not convert to condensed chromosomes in the depleted extract, although their condensation is normal upon addition of purified topo II. Dosage experiments support the possible notion of a structural involvement of topo II in chromosome condensation. In the topo II-depleted extract the erythrocyte nuclei progress to precondensation chromosomes, which lack the nuclear membrane-lamina complex and consist of a cluster of swollen chromatids.

A matrix/scaffold attachment region binding protein: identification, purification, and mode of binding

Matrix/scaffold attachment regions (MARs/SARs) partition chromatin into functional loop domains. Here we have identified a chicken protein that selectively binds to MARs from the chicken lysozyme locus and to MARs from Drosophila, mouse, and human genes. This protein, named ARBP (for attachment region binding protein), was purified to homogeneity and shown to bind to MARs in a cooperative fashion. ARBP is an abundant nuclear protein and a component of the internal nuclear network. Deletion mutants indicate that multiple AT-rich sequences, if contained in a minimal approximately 350 bp MAR fragment, can lead to efficient binding of ARBP. Furthermore, dimerization mutants show that, to bind ARBP efficiently, MAR sequences can act synergistically over large distances, apparently with the intervening DNA looping out. The binding characteristics of ARBP to MARs reproduce those of unfractionated matrix preparations, suggesting that ARBP is an important nuclear element for the generation of functional chromatin loops.

Structure-function relationships in eukaryotic nuclei

It may be that eukaryotic nuclei contain a collection of operationally independent units (genes), each controlled through its interactions with soluble protein factors which diffuse at random throughout the nucleoplasmic space. Alternatively, nuclei might be organized in such a sophisticated fashion that specific genes occupy distinct sites and that spatially ordered RNA synthesis, processing and transport delivers mature RNAs to predestined sites in the cytoplasm. Different fields of research support each of these extreme views. Molecular biologists inspecting the precise details of specific interactions, usually in vitro, inevitably favour the former, while cell biologists working with far more complicated systems generally assume that more elaborate arrangements exist. In considering the importance of nuclear architecture, I have attempted to relate a collection of experiments each of which intimates some close relationship between structural aspects of chromatin organization and the precise mechanisms underlying nuclear function. I will argue that higher-order structures are crucial for achieving the observed efficiency and coordination of many nuclear processes.

Mosaicism of tyrosinase-locus transcription and chromatin structure in dark vs. light melanocyte clones of homozygous chinchilla-mottled mice

The chinchilla-mottled (cm) mutation at the mouse tyrosinase-encoding locus leads to a transversely striped pattern of dark- and light-grey coat colors in homozygotes. The same basic pattern occurs in various other genotypes and has previously been found to represent the clonal developmental history of melanocytes. In a homozygote such as cm/cm, cis-acting mechanisms would be expected to account for the color differences. To search for these mechanisms, the genomic structure of the mutation was examined and compared with the wild-type, and its function was compared in cultured melanocyte clones of the respective colors. Evidence from restriction mapping indicated that the coding region of the mutant gene resembles that of the fully and uniformly pigmented wild-type. However, the upstream sequences are rearranged in the mutation. The rearrangement begins 5 kb 5' of the transcription initiation site and is estimated to encompass at least 30 kb of distal upstream sequence. At least two stable functional states of the cm gene were detectable: Light-cell clones have low levels of tyrosinase-specific transcription, reduced DNAase I sensitivity of tyrosinase chromatin, and no detectable hypersensitive sites near the gene; dark-cell clones have higher (but subnormal) levels of transcription, greater sensitivity of chromatin to DNAase I, and a hypersensitive site in the promoter region. The changed relation between the structural gene and its upstream region may separate it from cis-acting control elements, resulting in reduced and variable ability to achieve the appropriate chromatin configuration near the time of melanocyte determination; differences in expression among clonal initiator cells are then mitotically perpetuated.

Chromatin structure of hormono-dependent promoters

Transient transfections of mutated MMTV LTRs, driving the luciferase reporter gene, have shown the presence of at least one cis-acting element cooperating with the GREs. Studies of the chromatin structure of two glucocorticoid-regulated promoters, the mouse mammary tumor virus (MMTV) long terminal repeat (LTR), a retroviral promoter, and the rat tyrosine aminotransferase (TAT) promoter, demonstrate that both DNAs are organized into precisely positioned nucleosomes. Hormonal activation of transcription is accompanied by structural changes of one (MMTV LTR) or two (TAT promoter) nucleosomes associated with the hormone-response elements (HREs). These changes can be visualized by the appearance of DNasel hypersensitive sites. Association of the hormone-receptor complex with the nucleus is necessary to induce the DNasel hypersensitive site and to maintain transcription, but is not necessary to maintain DNasel hypersensitivity. Anti-hormones, even when able to promote a strong binding of the receptor to the nucleus, are unable to induce the chromatin structural change. Using cell lines containing approx. 200 copies of a MMTV LTR/Hv-ras chimeric construct, we have demonstrated a strong, hormono-independent nuclear matrix interaction of sequences located just upstream and downstream of the ras coding sequences.

A very large spontaneous deletion at aprt locus in CHO cells: sequence similarities with small aprt deletions

Spontaneous deletion mutants can be isolated from CHO cell lines heterozygous at the adenine phosphoribosyltransferase locus at frequencies up to 10(-4), i.e., about 100-fold higher than spontaneous point mutations. Indirect evidence has suggested that most deletions were in the megabase range. We have fully characterized one such mutant, Del 155, and shown that it resulted from an illegitimate recombination that took place between overlapping tetranucleotides. Comparisons with sequences of other deletions at various loci revealed a number of similarities, most striking of which was a CHI-like motif found within 6 bp of the upstream breakpoint of Del 155 and breakpoints of 8/21 previously described short deletions at the CHO aprt locus. Homology also existed between the downstream breakpoint of Del 155 and breakpoints of retinoblastoma gene deletions (3/6 cases) and also a 20-bp stretch of an Alu sequence in which breakpoints at the low-density lipoprotein receptor locus have been shown to cluster. The magnitude of the deletion event in Del 155 was assessed by pulsed field (PF) gel analysis and found to be at least 2100 kb long. PF analysis also indicated that the downstream breakpoint was near a region of structural differences between the two chromosomes carrying apart. These structural differences were probably not implicated in the mechanism of the high frequency event, since no indication of breakpoint clustering among a large collection of mutants was found either by conventional or PF electrophoretic analyses.

Paranemic structures of DNA and their role in DNA unwinding

A DNA structure is defined as paranemic if the participating strands can be separated without mutual rotation of the opposite strands. The experimental methods employed to detect paranemic, unwound, DNA regions is described, including probing by single-strand specific nucleases (SNN), conformation-specific chemical probes, topoisomer analysis, NMR, and other physical methods. The available evidence for the following paranemic structures is surveyed: single-stranded DNA, slippage structures, cruciforms, alternating B-Z regions, triplexes (H-DNA), paranemic duplexes and RNA, protein-stabilized paranemic DNA. The problem of DNA unwinding during gene copying processes is analyzed; the possibility that extended paranemic DNA regions are transiently formed during replication, transcription, and recombination is considered, and the evidence supporting the participation of paranemic DNA forms in genes committed to or undergoing copying processes is summarized.

Telomere and centromere DNA are associated with the cores of meiotic prophase chromosomes

Mouse (Mus musculus) whole-mount, surface-spread, meiotic prophase chromosomes have an axial core structure, the synaptonemal complex, SC, from which extend chromatin loops. This arrangement permits a novel approach to the analysis of chromosome structure. Using in situ hybridization, the types of DNA sequences preferentially associated with the SC and the types located primarily in the chromatin loops can be determined. With biotinylated probes, detected by avidin conjugated to FITC, we present evidence for differential chromatin-SC interaction. The telomere sequence (TTAGGG)n is associated exclusively with the two ends of each autosomal SC rather than with the chromatin loops. The minor satellite DNA sequences are predominantly localized to the centromeric region of the SC, as defined by CREST serum anti-centromere antibodies. In contrast, the major satellite DNA probe hybridizes to the chromatin loops of the centromeric heterochromatin, and a probe containing a LINE sequence hybridizes to chromatin loops in general with no obvious preference for the SC. These observations demonstrate that, depending on the type of DNA sequence, the chromatin has different properties in regard to its association with the SC.

Boundaries of somatic mutation in rearranged immunoglobulin genes: 5' boundary is near the promoter, and 3' boundary is approximately 1 kb from V(D)J gene

To investigate why somatic mutations are spatially restricted to a region around the rearranged V(D)J immunoglobulin gene, we compared the distribution of mutations flanking murine V gene segments that had rearranged next to either proximal or distal J gene segments. 124 nucleotide substitutions, nine deletions, and two insertions were identified in 32,481 bp of DNA flanking the coding regions from 17 heavy and kappa light chain genes. Most of the mutations occurred within a 2-kb region centered around the V(D)J gene, regardless of which J gene segment was used, suggesting that the structural information for mutation is located in sequences around and within the V(D)J gene, and not in sequences downstream of the J gene segments. The majority of mutations were found within 300 bp of DNA flanking the 5' side of the V(D)J gene and 850 bp flanking the 3' side at a frequency of 0.8%, which was similar to the frequency in the coding region. The frequency of flanking mutations decreased as a function of distance from the gene. There was no evidence for hot spots in that every mutation was unique and occurred at a different position. No mutations were found upstream of the promoter region, suggesting that the promoter delimits a 5' boundary, which provides strong evidence that transcription is necessary to generate mutation. The 3' boundary was approximately 1 kb from the V(D)J gene and was not associated with a DNA sequence motif. Occasional mutations were located in the nuclear matrix association and enhancer regions. The pattern of substitutions suggests that there is discrimination between the two DNA strands during mutation, in that the four bases were mutated with different frequencies on each strand. The high frequency of mutations in the 3' flanking region and the uniqueness of each mutation argues against templated gene conversion as a mechanism for generating somatic diversity in murine V(D)J genes. Rather, the data support a model for random point mutations where the mechanism is linked to the transcriptional state of the gene.

Specific organization of genes in relation to the sperm nuclear matrix

We have recently reported that mammalian sperm DNA, the most highly condensed and functionally inert eukaryotic DNA, is organized into DNA loop domains attached at their bases to a sperm nuclear matrix (Chromosoma, 98: 153, 1989). We report here the specific arrangement of genes within the sperm loop domains by measuring the proximity of six hamster genes to the sperm nuclear matrix. After restriction endonuclease treatment of sperm nuclear matrices, five genes were clearly not associated with the sperm nuclear matrix and only one, alpha A-crystallin, was within 2 Kb of the matrix. This suggests that sperm DNA is organized in a specific manner in relation to the sperm matrix.

Drosophila scaffold-attached regions bind nuclear scaffolds and can function as ARS elements in both budding and fission yeasts

Histone-depleted nuclei maintain sequence-specific interactions with genomic DNA at sites known as scaffold attachment regions (SARs) or matrix attachment regions. We have previously shown that in Saccharomyces cerevisiae, autonomously replicating sequence elements bind the nuclear scaffold. Here, we extend these observations to the fission yeast Schizosaccharomyces pombe. In addition, we show that four SARs previously mapped in the genomic DNA of Drosophila melanogaster bind in vitro to nuclear scaffolds from both yeast species. In view of these results, we have assayed the ability of the Drosophila SARs to promote autonomous replication of plasmids in the two yeast species. Two of the Drosophila SARs have autonomously replicating sequence activity in budding yeast, and three function in fission yeast, while four flanking non-SAR sequences are totally inactive in both.

DNA-binding protein activated by gamma radiation in human cells

DNA damage-inducible responses in mammalian cells tend to lack specificity and can be activated by any one of a number of damaging agents. Although a number of different induced proteins have been described, their involvement in DNA processing and transcriptional control remains unresolved. We describe the appearance of a previously unreported, specific DNA-binding protein in nuclei from human cells exposed to ionizing radiation, which was not detected in nuclear extracts from unperturbed cells. The distal part of the simian virus 40 enhancer (without the AP-1 site) and oligonucleotide sequences derived from that sequence were used in binding studies. The appearance of this activity was dose dependent and transient, reaching a maximum at 1 h postirradiation and disappearing from nuclei by 9 h. This protein was induced in cells by a mechanism not requiring de novo protein synthesis, and the response was specific for ionizing radiation and radiomimetic agents; neither UV nor heat shock invoked a response. The DNA-binding protein was present in the cytoplasm of untreated cells, apparently being translocated to the nucleus only after radiation exposure. Southwestern (DNA-protein) analysis demonstrated that the nuclear and cytoplasmic proteins were approximately the same size, 43,000 daltons. The protected DNA-binding motif, using the distal fragment of the simian virus 40 enhancer as the substrate, was shown by DNase I footprint analysis to be pTGTCAGTTAGGGTACAGTCAATCCCAp. This was confirmed by dimethyl sulfate footprinting.

Studies of an 800-kilobase DNA stretch of the Drosophila X chromosome: comapping of a subclass of scaffold-attached regions with sequences able to replicate autonomously in Saccharomyces cerevisiae

We have previously mapped scaffold-attached regions (SARs) on an 800-kilobase DNA walk from the Drosophila X chromosome. We have also previously shown that the strength of binding, i.e., the ability of SARs to bind to all nuclear scaffolds or only to a fraction of them varied from one SAR to another one. In the present study, 71 of the 85 subfragments that bind scaffolds and 38 fragments that do not bind scaffolds were tested for their ability to promote autonomous replicating sequence (ARS) activity in Saccharomyces cerevisiae. Sixteen SAR-containing fragments from the chromosome walk were also examined for association to yeast nuclear scaffolds in vitro. All identified ARSs (a total of 27) were present on SAR-containing fragments, except two, which were adjacent to SARs. There is thus a correlation between ARS and SAR activities, and this correlation defines a SAR subclass. Moreover, the presence of an ARS on a DNA fragment appeared to be highly correlated with the strength of binding. The binding activity was highly conserved from Drosophila melanogaster to yeast. These data suggest that Drosophila DNA sequences responsible for binding to components of the nuclear scaffold from either D. melanogaster or yeast may be involved in the process of heterologous extrachromosomal replication in yeasts.

Mapping of structural and transcription-related matrix attachment sites in the alpha-globin gene domain of avian erythroblasts and erythrocytes

The positions of preferential DNA interaction with the nuclear matrix were mapped within the domain of the chicken alpha-globin genes in transcriptionally active erythroblast nuclei and inactive nuclei of mature erythrocytes. In the latter, only two major distinct attachment sites were observed, close to the A + T-rich sequences previously found at the boundaries of the domain. Sequencing of these structural matrix attachment points revealed several known DNA motifs; some of them were present on both sides of the domain. In actively transcribing erythroblast nuclei of adult animals, a large fraction of the transcribed area was represented in nuclear matrix DNA, including upstream and downstream elements. In particular, adult alpha A- and alpha D-globin genes were found in matrix DNA, while the transcribed but translationally unexpressed embryonic pi gene was underrepresented. The data are discussed in terms of the existence of stable or structural and expression-related matrix attachment sites; correlations to the origin of replication and the units of transcription of the domain are shown.

Drosophila scaffold-attached regions bind nuclear scaffolds and can function as ARS elements in both budding and fission yeasts

Histone-depleted nuclei maintain sequence-specific interactions with genomic DNA at sites known as scaffold attachment regions (SARs) or matrix attachment regions. We have previously shown that in Saccharomyces cerevisiae, autonomously replicating sequence elements bind the nuclear scaffold. Here, we extend these observations to the fission yeast Schizosaccharomyces pombe. In addition, we show that four SARs previously mapped in the genomic DNA of Drosophila melanogaster bind in vitro to nuclear scaffolds from both yeast species. In view of these results, we have assayed the ability of the Drosophila SARs to promote autonomous replication of plasmids in the two yeast species. Two of the Drosophila SARs have autonomously replicating sequence activity in budding yeast, and three function in fission yeast, while four flanking non-SAR sequences are totally inactive in both.

Studies of an 800-kilobase DNA stretch of the Drosophila X chromosome: comapping of a subclass of scaffold-attached regions with sequences able to replicate autonomously in Saccharomyces cerevisiae

We have previously mapped scaffold-attached regions (SARs) on an 800-kilobase DNA walk from the Drosophila X chromosome. We have also previously shown that the strength of binding, i.e., the ability of SARs to bind to all nuclear scaffolds or only to a fraction of them varied from one SAR to another one. In the present study, 71 of the 85 subfragments that bind scaffolds and 38 fragments that do not bind scaffolds were tested for their ability to promote autonomous replicating sequence (ARS) activity in Saccharomyces cerevisiae. Sixteen SAR-containing fragments from the chromosome walk were also examined for association to yeast nuclear scaffolds in vitro. All identified ARSs (a total of 27) were present on SAR-containing fragments, except two, which were adjacent to SARs. There is thus a correlation between ARS and SAR activities, and this correlation defines a SAR subclass. Moreover, the presence of an ARS on a DNA fragment appeared to be highly correlated with the strength of binding. The binding activity was highly conserved from Drosophila melanogaster to yeast. These data suggest that Drosophila DNA sequences responsible for binding to components of the nuclear scaffold from either D. melanogaster or yeast may be involved in the process of heterologous extrachromosomal replication in yeasts.

Mapping of structural and transcription-related matrix attachment sites in the alpha-globin gene domain of avian erythroblasts and erythrocytes

The positions of preferential DNA interaction with the nuclear matrix were mapped within the domain of the chicken alpha-globin genes in transcriptionally active erythroblast nuclei and inactive nuclei of mature erythrocytes. In the latter, only two major distinct attachment sites were observed, close to the A + T-rich sequences previously found at the boundaries of the domain. Sequencing of these structural matrix attachment points revealed several known DNA motifs; some of them were present on both sides of the domain. In actively transcribing erythroblast nuclei of adult animals, a large fraction of the transcribed area was represented in nuclear matrix DNA, including upstream and downstream elements. In particular, adult alpha A- and alpha D-globin genes were found in matrix DNA, while the transcribed but translationally unexpressed embryonic pi gene was underrepresented. The data are discussed in terms of the existence of stable or structural and expression-related matrix attachment sites; correlations to the origin of replication and the units of transcription of the domain are shown.

DNA-binding protein activated by gamma radiation in human cells

DNA damage-inducible responses in mammalian cells tend to lack specificity and can be activated by any one of a number of damaging agents. Although a number of different induced proteins have been described, their involvement in DNA processing and transcriptional control remains unresolved. We describe the appearance of a previously unreported, specific DNA-binding protein in nuclei from human cells exposed to ionizing radiation, which was not detected in nuclear extracts from unperturbed cells. The distal part of the simian virus 40 enhancer (without the AP-1 site) and oligonucleotide sequences derived from that sequence were used in binding studies. The appearance of this activity was dose dependent and transient, reaching a maximum at 1 h postirradiation and disappearing from nuclei by 9 h. This protein was induced in cells by a mechanism not requiring de novo protein synthesis, and the response was specific for ionizing radiation and radiomimetic agents; neither UV nor heat shock invoked a response. The DNA-binding protein was present in the cytoplasm of untreated cells, apparently being translocated to the nucleus only after radiation exposure. Southwestern (DNA-protein) analysis demonstrated that the nuclear and cytoplasmic proteins were approximately the same size, 43,000 daltons. The protected DNA-binding motif, using the distal fragment of the simian virus 40 enhancer as the substrate, was shown by DNase I footprint analysis to be pTGTCAGTTAGGGTACAGTCAATCCCAp. This was confirmed by dimethyl sulfate footprinting.

Transvection and long-distance gene regulation

Numerous genes contain regulatory elements located many tens of kilobases away from the promoter they control. Specific mechanisms must be required to ensure that such distant elements can find and interact with their proper targets but not with extraneous genes. This review explores the connections between transvection phenomena, the activation of domains of homeotic gene expression, position effect variegation and silencers. These various examples of long-distance effects suggest that, in all cases, related forms of chromatin packaging may be involved.

Nucleosome positioning as a critical determinant for the DNA cleavage sites of mammalian DNA topoisomerase II in reconstituted simian virus 40 chromatin

We have assessed the ability of nucleosomes to influence the formation of mammalian topoisomerase II-DNA complexes by mapping the sites of cleavage induced by four unrelated topoisomerase II inhibitors in naked versus nucleosome-reconstituted SV40 DNA. DNA fragments were reconstituted with histone octamers from HeLa cells by the histone exchange method. Nucleosome positions were determined by comparing micrococcal nuclease cleavage patterns of nucleosome-reconstituted and naked DNA. Three types of DNA regions were defined: 1) regions with fixed nucleosome positioning; 2) regions lacking regular nucleosome phasing; and 3) a region around the replication origin (from position 5100 to 600) with no detectable nucleosomes. Topoisomerase II cleavage sites were suppressed in nucleosomes and persisted or were enhanced in linker DNA and in the nucleosome-free region around the replication origin. Incubation of reconstituted chromatin with topoisomerase II protected nucleosome-free regions from micrococcal nuclease cleavage without changing the overall micrococcal nuclease cleavage pattern. Thus, the present results indicate that topoisomerase II binds preferentially to nucleosome-free DNA and that the presence of nucleosomes at preferred DNA sequences influences drug-induced DNA breaks by topoisomerase II inhibitors.

Attachment of DNA to the nucleoskeleton of HeLa cells examined using physiological conditions

Although it is widely believed that eukaryotic DNA is looped by attachment to a nucleoskeleton, there is controversy about its composition and which sequences are attached to it. As most nuclear derivatives are isolated using unphysiological conditions, the criticism that attachments seen in vitro are generated artifactually has been difficult to rebut. Therefore we have re-investigated attachments of chromatin to the skeleton using physiological conditions. HeLa cells are encapsulated in agarose microbeads and lysed using Triton in a 'physiological' buffer. Then, most chromatin can be electroeluted after treatment with a restriction enzyme to leave some at the base of the loops still attached. Analysis of the size and amounts of these residual fragments indicates that the loops are 80-90kbp long. The residual fragments are stably attached, with about 1kbp of each fragment protected from nuclease attack. This is very much longer than a typical protein-binding site of 10-20bp.

A non-curved chicken lysozyme 5' matrix attachment site is 3' followed by a strongly curved DNA sequence

Matrix attachment regions (MARs) partition the genome into functional and structural loop-domains. Here, we determined the relative matrix affinity of cloned fragments of the chicken lysozyme 5' MAR. We show that this region contains a non-curved high-affinity binding site, which is 3' followed by a strongly curved DNA sequence that exhibits weak matrix binding. DNA curvature is not a physical property required for strong matrix binding. Possible biological functions of this sequence arrangement, particularly of the strongly curved DNA, are discussed.

Supragenic loop organization: mapping in Drosophila embryos, of scaffold-associated regions on a 800 kilobase DNA continuum cloned from the 14B-15B first chromosome region

The supragenic loop organization of the Drosophila genome was investigated on a 800 kilobase (kb) DNA continuum from the 14B-15B first chromosome region. Nuclear scaffolds from 0-18 hr embryos were prepared with Laemmli's low-salt, detergent procedure and digested with restriction enzymes. Scaffold-associated regions (SARs) were mapped by probing Southern transfers of total, scaffold-associated and free DNA with a set of 70 recombinant phages overlapping the investigated genomic region. In all, 85 restriction fragments showed association to scaffolds. 12 of them were present in the majority of scaffolds. They bore strong SARs organizing the DNA molecule as consecutive loops with sizes ranging from 15 to 115 kb. 44 were present in only a fraction of scaffolds. They contained weak SARs subdividing the basic loops into smaller ones. 29 additional restriction fragments were present in a very small fraction of scaffolds. The position of SARs with respect to transcribed regions was investigated. Strong SARs appeared to be located on untranscribed DNA and to frame transcription units. In contrast, at least some weak SARs were shown to comap with transcribed regions or to reside within characterized transcription units. Statistical analyses established that strong and weak SARs were periodically positioned on the DNA continuum and that there was a potential contact point between scaffolds and the DNA continuum every 11 kb, or multiples thereof. Implications for SAR role(s) are discussed.

Adenovirus terminal protein mediates both nuclear matrix association and efficient transcription of adenovirus DNA

Adenovirus DNA is tightly bound to the nuclear matrix throughout the course of infection. Analysis of adenovirus DNA from infected HeLa cell nuclei after extraction with lithium diiodosalicylate and digestion with restriction enzymes demonstrated that the sites of tightest attachment occur in the terminal fragments of the linear viral chromosome. Analysis of viruses mutated in the precursor terminal protein coding sequence demonstrated that the terminal protein, which is covalently attached to the 5' end of each DNA strand, mediates the tight binding. Virions containing chromosomes with mutant terminal proteins were unpackaged and viral DNA accumulated in the nucleus at a normal rate and competed for the limiting component during transcription complex formation, but their early genes were transcribed at reduced efficiency by both RNA polymerases II and III. The transcriptional defects were not complemented by coinfection with a wild-type virus. We propose that the adenovirus chromosome may exist as a single chromatin domain during infection and that binding of DNA to the nuclear matrix may play a critical role in adenovirus transcription.

Analysis of somatic mutations in kappa transgenes

We have examined the nature and localization of somatic mutations in three kappa transgenes cloned from IgG-secreting hybridomas. All of the mutations identified were single base substitutions. Mutations were localized to the variable (V) region and its flanking sequences. In every case, the nuclear matrix association region, kappa enhancer, and C gene were spared. These data indicate that the rearranged kappa gene contains the necessary sequences for targeting of the mutation process, and suggest that the observed localization of mutations to the V region reflects the inherent specificity of this mutation process.

Inhibition of topoisomerase II does not inhibit transcription of RNA polymerase I and II genes

Injection of VM-26 (teniposide) into Xenopus oocytes inhibits the activity of topoisomerase II but does not inhibit transcription by RNA polymerases I and II. A specific assay for topoisomerase II, resolution of catenated DNA molecules into product rings, was used to quantitate VM-26 inhibition in vivo. When catenanes were injected without VM-26, about 60% of them were separated into product rings in the first 5 min after injection, and decatenation of the remainder was complete within 15 min. When VM-26 was coinjected, 60% of the catenanes were separated into product rings in the first 5 min after injection, but the remaining 40% were stable over the next 40 min. At 1 h after injection catenanes were no longer detected in the gel analysis, but the increasing numbers of linear product rings indicated that topoisomerase II continued to be inhibited by VM-26. These results suggest that a short lag of approximately 5 min is required for VM-26 to inhibit topoisomerase II and that after this initial period topoisomerase II is inhibited by more than 90%. There was no detectable decrease in transcription of injected rRNA and thymidine kinase (TK) genes or in the activity of the rRNA enhancer when these transcription templates were coinjected with VM-26. The time required for assembly of injected DNA into chromatin doubled in the presence of VM-26.

A nuclear matrix protein binds very tightly to DNA in the avian beta-globin gene enhancer

Current evidence suggests that DNA is covalently attached to proteins in the nuclear matrix of eukaryotic cells and that specific DNA sequences are tightly associated with the nuclear matrix. However, it has not been documented that specific DNA sequences can become covalently attached to nuclear matrix protein. We have examined the binding of cloned DNA sequences that contain the avian beta-globin gene enhancer, a region previously shown to be matrix associated in erythroid cells in vivo, with nuclear matrices from several avian tissue sources to determine if covalent DNA-protein bonds are formed. Our results indicate that sequence-specific DNA-protein complexes that are resistant to denaturation by SDS, boiling, and phenol and disulfide reduction are formed. Excess protein, capable of forming very tight bonds with DNA that contains the beta-globin gene enhancer, is present in cells in which matrix attachment of this DNA sequence is not detected in vivo. Evidence is presented that suggests that the protein to which DNA forms very tight bonds is not topoisomerase II. These results are discussed in relation to current models of the nuclear matrix and the utility of in vitro assays of matrix attachment regions using cloned DNA.

Inhibition of topoisomerase II does not inhibit transcription of RNA polymerase I and II genes

Injection of VM-26 (teniposide) into Xenopus oocytes inhibits the activity of topoisomerase II but does not inhibit transcription by RNA polymerases I and II. A specific assay for topoisomerase II, resolution of catenated DNA molecules into product rings, was used to quantitate VM-26 inhibition in vivo. When catenanes were injected without VM-26, about 60% of them were separated into product rings in the first 5 min after injection, and decatenation of the remainder was complete within 15 min. When VM-26 was coinjected, 60% of the catenanes were separated into product rings in the first 5 min after injection, but the remaining 40% were stable over the next 40 min. At 1 h after injection catenanes were no longer detected in the gel analysis, but the increasing numbers of linear product rings indicated that topoisomerase II continued to be inhibited by VM-26. These results suggest that a short lag of approximately 5 min is required for VM-26 to inhibit topoisomerase II and that after this initial period topoisomerase II is inhibited by more than 90%. There was no detectable decrease in transcription of injected rRNA and thymidine kinase (TK) genes or in the activity of the rRNA enhancer when these transcription templates were coinjected with VM-26. The time required for assembly of injected DNA into chromatin doubled in the presence of VM-26.

Lack of somatic mutation in a kappa light chain transgene

Analysis of mice transgenic for immunoglobulin genes should allow definition of the cis-acting DNA sequences required to target somatic mutation to antibody V genes. We have looked for mutations in a chimeric kappa transgene encoding a V region specific for the hapten 2-phenyloxazolone (phOx) linked to a rat C kappa gene. Two independent lines of transgenic mice were hyperimmunized with phOx and splenic hybridomas established. In B cells that had been selected by antigen and which used mouse anti-phOx genes, the endogenous sequences were found to be mutated whereas the transgene remained unchanged. These results suggest either that (a) if the transgene is a "passenger" gene expressed at a low level, transgene mutation is a rare event, or that (b) sequences far from the kappa coding region are necessary to direct somatic mutation.

Chromosome reproduction: units of DNA for segregation

Evidence is summarized which indicates that the DNA loop anchoring proteins in chromosomes are effectively heterodimers that stack and are fastened into a bilaterally symmetrical array along the chromonemal axis. The evidence consists primarily of the observations made twenty five to thirty years ago on the pattern of sister chromatid exchanges and the way the DNA chains are sorted in the formation of diplochromosomes in cells that have undergone endoreduplication. The evidence indicates that each chain of DNA in the single duplex, which is assumed to run the length of a chromosome, is anchored to a bilaterally symmetrical axis of heterodimers that sort the two original chains among the four derived chromatids of each diplochromosome in a very precise way. These observations are considered in the context of investigations on the nature of scaffold proteins and the loop anchorage sequences, as well as the advances being made on the nature of DNA binding proteins and the roles of topoisomerase II.

Somatic diversification of chicken immunoglobulin light chains by point mutations

The light-chain locus of chicken has 1 functional V lambda 1 gene, 1 J gene, and 25 pseudo-V lambda-genes (where V = variable and J = joining). A major problem is which somatic mechanisms expand this extremely limited germ-line information to generate many different antibodies. Weill's group [Reynaud, C. A., Anquez, V., Grimal, H. & Weill, J. C. (1987) Cell 48, 379-388] has shown that the pseudo-V lambda-genes diversify the rearranged V lambda 1 by gene conversion. Here we demonstrate that chicken light chains are further diversified by somatic point mutations and by V lambda 1-J flexible joining. Somatic point mutations were identified in the J and 3' noncoding DNA of rearranged light-chain genes of chicken. These regions were analyzed because point mutations in V lambda 1 are obscured by gene conversion; the J and 3' noncoding DNA are presented in one copy per haploid genome and are not subject to gene conversion. In rodents point mutations occur as frequently in the V-J coding regions as in the adjacent flanking DNA. Therefore, we conclude that somatic point mutations diversify the V lambda 1 of chicken. The frequency (0-1%) and distribution of the mutations (decreasing in number with increased distance from the V lambda 1 segment) in chicken were as observed in rodents. Sequence variability at the V lambda 1-J junctions could be attributed to imprecise joining of the V lambda 1 and J genes. The modification by gene conversion of rearranged V lambda 1 genes in the bursa was similar in chicken aged 3 months (9.5%) or 3 weeks (9.1%)--i.e., gene conversion that generates the preimmune repertoire in the bursa seems to level off around 3 weeks of age. This preimmune repertoire can be further diversified by somatic point mutations that presumably lead to the formation of antibodies with increased affinity. A segment with structural features of a matrix association region [(A + T)-rich and four topoisomerase II binding sites] was identified in the middle of the J-C lambda intron (where C = constant).

The lengths of DNA fragments light-induced in the presence of a photosensitizer localized at the nuclear membrane of human cells

DNA damage was apparently introduced selectively in the parts of DNA localized close to the nuclear membrane of human NHIK 3025 cells. This was obtained by illumination of the cells in the presence of a hydrophobic photosensitizer, Photofrin II, which was located in membrane structures but not in the nucleus. Photofrin II sensitizes DNA to light mainly via singlet oxygen, which diffuses about 0.1 microns in its intracellular lifetime. By measuring DNA unwinding in alkali after illumination or X-irradiation of cells, the distribution of the number of DNA bases between two adjacent, photodamaged DNA sites was estimated. The average length of these DNA fragments was found to be 155 kilobases (kb).

The chicken lysozyme 5' matrix attachment region increases transcription from a heterologous promoter in heterologous cells and dampens position effects on the expression of transfected genes

Matrix attachment regions (MARs) are DNA elements that dissect the genome into topologically separated domains by binding to a chromosomal skeleton. This study explored the putative influence of the MAR located 5' of the chicken lysozyme gene on expression of heterologous genes in heterologous cell systems. Expression of a construct with the chloramphenicol acetyltransferase (CAT) indicator gene controlled by the herpes simplex virus thymidine kinase promoter (TC) and a construct in which the same transcriptional unit is flanked by chicken lysozyme 5' MARs (MTCM) was assayed after stable transfection into rat fibroblasts. Median CAT activity per copy number in MTCM transfectants was elevated approximately 10-fold relative to that in TC transfectants. Total variation in normalized CAT activity decreased from more than 100-fold among TC transfectants to nearly 6-fold among MTCM transfectants. The steady-state level of transcripts and the relative rate of transcription were increased in MTCM transfectants, as shown by S1 nuclease and run-on transcription assays, respectively. The chicken lysozyme 5' MAR thus can confer elevated, less position-dependent expression on a heterologous promoter in cells of a different species by increasing the density of transcribing RNA polymerase molecules. MAR-mediated transcriptional enhancement suggests that MARs are important for gene expression and not just for DNA packaging.

Nuclear matrix attachment occurs in several regions of the IgH locus

The genome is thought to be divided into domains by DNA elements which mediate anchorage of chromosomal DNA to the nuclear matrix or chromosome scaffold. The positions of nuclear matrix anchorage regions (MARs) have been mapped within the 200 kb mouse immunoglobulin heavy chain constant region locus, thereby allowing an estimate of the size of DNA domains within a segment of the genome. MARs were identified in four regions, which appear to divide the locus into looped DNA domains of 30, 20, 30 and greater than 70 kb in length. These DNA domain sizes fall within the range of DNA loop sizes observed in histone-extracted nuclei and chromosomes. In two regions, large clusters of MARs were identified, and many of these MARs lie on DNA fragments that include repetitive DNA elements, perhaps indicating that repetitive DNA integrates into the genome close to MARs, or that some classes of repeats could themselves act as MARs.

The chicken lysozyme 5' matrix attachment region increases transcription from a heterologous promoter in heterologous cells and dampens position effects on the expression of transfected genes

Matrix attachment regions (MARs) are DNA elements that dissect the genome into topologically separated domains by binding to a chromosomal skeleton. This study explored the putative influence of the MAR located 5' of the chicken lysozyme gene on expression of heterologous genes in heterologous cell systems. Expression of a construct with the chloramphenicol acetyltransferase (CAT) indicator gene controlled by the herpes simplex virus thymidine kinase promoter (TC) and a construct in which the same transcriptional unit is flanked by chicken lysozyme 5' MARs (MTCM) was assayed after stable transfection into rat fibroblasts. Median CAT activity per copy number in MTCM transfectants was elevated approximately 10-fold relative to that in TC transfectants. Total variation in normalized CAT activity decreased from more than 100-fold among TC transfectants to nearly 6-fold among MTCM transfectants. The steady-state level of transcripts and the relative rate of transcription were increased in MTCM transfectants, as shown by S1 nuclease and run-on transcription assays, respectively. The chicken lysozyme 5' MAR thus can confer elevated, less position-dependent expression on a heterologous promoter in cells of a different species by increasing the density of transcribing RNA polymerase molecules. MAR-mediated transcriptional enhancement suggests that MARs are important for gene expression and not just for DNA packaging.

cis-acting enhancement of RNA polymerase III gene expression in vitro

The Xenopus laevis S-150 cell-free extract catalyzes in vitro transcription of several RNA polymerase III genes. Among these are the Xenopus 5S RNA gene (somatic type) and the Xenopus methionine tRNA gene. In this report we present an analysis of the transcriptional activity of these two genes either in trans-competition experiments or when the genes are co-localized in the same circular plasmid. In the "cis" arrangement, elevated levels of 5S and tRNA gene expression are observed, which are dependent on the relative orientation of the two genes (convergent or in tandem) and the distance between them. The results of these analyses reveal important parameters affecting the expression of juxtaposed genes.

Replication forks are associated with the nuclear matrix

It has been proposed that DNA in eukaryotic cells is synthesized via replication complexes that are fixed to a proteinaceous nuclear matrix. This model has not been universally accepted because the matrix and its associated DNA are usually prepared under hypertonic conditions that could facilitate non-specific aggregation of macromolecules. We therefore investigated whether different ionic conditions can significantly affect the association of nascent DNA with the nuclear matrix in cultured mammalian cells. Matrices were prepared either by a high salt method or by hypotonic or isotonic LIS extraction. Chromosomal DNA was subsequently removed by digestion with either DNAse I or EcoRI. With all methods of preparation, we found that newly synthesized DNA preferentially partitioned with the nuclear matrix. Furthermore, when the matrix-attached DNA fraction was analyzed by two-dimensional gel electrophoresis, we found that it was markedly enriched for replication forks. We therefore conclude that attachment of DNA to the matrix in the vicinity of replication forks is not induced by conditions of high ionic strength, and that replication may, indeed, occur on or near the skeletal framework provided by the nuclear matrix. From a practical standpoint, our findings suggest a strategy for greatly increasing the sensitivity of two important new gel electrophoretic methods for the direct mapping of replication fork movement through defined chromosomal domains in mammalian cells.

Structural features of transposed human VK genes and implications for the mechanism of their transpositions

The genes encoding the variable, joining and constant regions of human immunoglobulin light chains have been localized to the short arm of chromosome 2. However, several VK genes lie outside of the locus: a single copy cluster of five VK genes is located on chromosome 22; an isolated but amplified VkI gene is found on chromosome 1; and several isolated VkI genes are on as-yet-unidentified chromosomes other than chromosome 2. Vk genes not contained within the kappa locus are termed orphons. We have attempted to gain insight into the mechanism of transposition of both the chromosome 22 cluster and the several amplified VkI genes by searching in the kappa locus for a parent copy of the former, and by analyzing the junctions between transposed VKI-containing segments and adjacent non-amplified regions. The chromosome 22 orphon cluster must have been non-duplicatively transposed. Sequence features at the junctions of this and other orphon regions are direct and inverted repeats, and, in one case, an Alu repeat. These unusual features may have predisposed the orphon regions to transposition by serving as target sites for enzymes involved in recombination.