Inhibition of HSV-1 proliferation by decoy phosphodiester oligonucleotides containing ICP4 recognition sequences

Transcriptional control in eukaryotes results from the interplay between DNA sequences in promoters, enhancers, or silencers and transcription factors. Selective control of gene expression can thus be achieved by inhibiting specific transcription factor/DNA interactions. Transcriptional activity of DNA binding transcription factors can be inhibited by competition with double-stranded oligonucleotides (decoys) that contain their specific recognition sequences. The immediate early protein ICP4 of herpes simplex virus type 1 (HSV-1) is a sequence-specific DNA binding protein that is essential for viral replication. We synthesized double-stranded hairpin phosphodiester oligonucleotides carrying ICP4 sites and demonstrated their ability to specifically titrate ICP4. Upon addition to Vero cells, ICP4 hairpin decoys significantly reduced HSV-1 titers (IC50 = 0.3 microM), whereas a control hairpin oligonucleotide had no activity. Antiviral activity of ICP4 hairpin decoys was correlated to their relative binding affinities. These results show that phosphodiester oligonucleotides can compete for binding of specific transcription factors within cells, thus providing a potential therapeutic tool to control disease-causing genes.

Genetic analysis of case/control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer's disease

There is growing debate over the utility of multiple locus association analyses in the identification of genomic regions harboring sequence variants that influence common complex traits such as hypertension and diabetes. Much of this debate concerns the manner in which one can use the genotypic information from individuals gathered in simple sampling frameworks, such as the case/control designs, to actually assess the association between alleles in a particular genomic region and a trait. In this paper we describe methods for testing associations between estimated haplotype frequencies derived from multilocus genotype data and disease endpoints assuming a simple case/control sampling design. These proposed methods overcome the lack of phase information usually associated with samples of unrelated individuals and provide a comprehensive way of assessing the relationship between sequence or multiple-site variation and traits and diseases within populations. We applied the proposed methods in a study of the relationship between polymorphisms within the APOE gene region and Alzheimer's disease. Cases and controls for this study were collected from the United States and France. Our results confirm the known association between the APOE locus and Alzheimer's disease, even when the epsilon 4 polymorphism is not contained in the tested haplotypes. This suggests that, in certain situations, haplotype information and linkage disequilibrium-induced associations between polymorphic loci that neighbor loci harboring functional sequence variants can be exploited to identify disease-predisposing alleles in large, freely mixing populations via estimated haplotype frequency methods.

Efficient synthesis of DNA dumbbells using template-induced chemical ligation in double-stranded polynucleotides closed by minihairpin fragments

The chemical ligation of 17 50-54-membered nicked DNA dumbbells with different closing fragments, nick positions, and nucleotides facing the nick were investigated. T4, T5, GTA4C, GCGA2GC, and GCGA3GC sequences were chosen as the closing fragments. The nicks were placed in the center of the duplex stem or were adjacent to the closing fragments. N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide and cyanogen bromide were used as the condensing agents. We showed that the ligation efficiency is 10%-90% depending on the sequence of the closing fragments, nick position, and nucleotides facing the nick. Coupling yields of 80%-90% were observed when the nick was situated in the middle of the molecule between two T residues or was adjacent to GCGA2GC or GCGA3GC minihairpins. In the last case, the reacting 3'-phosphate and 5'-hydroxy groups were brought close together by only two base pair minihairpins. The coupling yields did not depend on the nature of the condensing agent. On the basis of the results obtained, we believe a rational design of nicked DNA dumbbells has been developed for efficient chemical synthesis of closed dumbbells.

Analysis of the distribution of binding sites for a tissue-specific transcription factor in the vertebrate genome

Hepatocyte nuclear factor 1 (HNF1) is a dimeric homeoprotein expressed in hepatocytes and in a few other epithelial cells where it helps regulate the expression of a specific subset of genes. In an attempt to identify novel target genes for HNF1 and to assess the distribution of its target sites within the vertebrate genome, we performed a computer-assisted search within the available databases using a weighted matrix. Several hundred potential target sequences were identified within the GenBank and EMBL data banks. DNA binding assays demonstrated that more than 95%, of the new sites tested (52 sites among 54) bound HNF1. Surprisingly many HNF1 target sites were found in genes that are transcribed in cell types that do not contain the protein. On the other hand these sites are 2.5 to five times more frequent in hepatic genes than expected. It seems that the presence of HNF1 sites in liver-specific genes was favoured, but that no counter-selection occurred within the rest of the genome. HNF1 binding sites in liver genes are more often associated in clusters with sites for other transcription factors and the enrichment is more pronounced in promoter regions. We identified more than 100 liver specific genes that are potentially regulated by HNF1.

Crosslinking of double-stranded oligonucleotides containing O-methyl-substituted pyrophosphate groups to the HNF1 transcription factor in nuclear cell extract

Probing of the HNF1 (hepatocyte nuclear factor I) DNA-binding region using a set of DNA duplexes containing pyrophosphate or O-methyl-substituted pyrophosphate internucleotide groups at different positions of the HNF1 recognition sequence was performed. The histidine-tagged HNF1/1-281 DNA binding domain and nuclear extract from rat liver were used. We showed that HNF1 from these species specifically binds to modified DNA duplexes. A correlation in binding affinity of both types of duplexes was detected. Crosslinking of the HNF1 DNA-binding domain and HNF1 in nuclear liver extract to DNA duplexes carrying O-methyl-substituted pyrophosphate groups was observed. The crosslinking efficiency of HNF1 in liver extract to substituted pyrophosphate-modified DNA duplex, containing a reactive internucleotide group between nucleotides G and T of the GT dinucleotide immediately 5' to the TAAT recognition sequence, amounts to 40% of the efficiency of non-covalent association. Nonspecific crosslinking of the reactive DNA duplexes to other components of nuclear extract was not observed. These results indicate that DNA duplexes carrying substituted pyrophosphate internucleotide groups can specifically bind and crosslink with DNA-binding proteins, especially transcription factors in crude preparations and could constitute a potential tool to control the expression of disease-causing genes.

Mechanisms of inhibition of in vitro dimerization of HIV type I RNA by sense and antisense oligonucleotides

Retroviruses display a strong selective pressure to maintain the dimeric nature of their genomic RNAs, suggesting that dimerization is essential for viral replication. Recently, we identified the cis-element required for initiation of human immunodeficiency virus type I (HIV-I) RNA dimerization in vitro. The dimerization initiation site (DIS) is a hairpin structure containing a self-complementary sequence in the loop. We proposed that dimerization is initiated by a loop-loop kissing interaction involving the self-complementary sequence present in each monomer. We tested the ability of sense and antisense oligonucleotides targeted against the DIS to interfere with a preformed viral RNA dimer. Self-dimerization and inhibition properties of the tested oligonucleotides are dictated by the nature of the loop. An RNA loop is absolutely required in the case of sense oligonucleotides, whereas the nature and the sequence of the stem is not important. They form reversible loop-loop interactions and act as competitive inhibitors. Antisense oligonucleotides are less efficient in self-dimerization and are more potent inhibitors than sense oligonucleotides. They are less sensitive to the nature of the loop than the antisense oligonucleotides. Antisense hairpins with either RNA or DNA stems are able to form highly stable and irreversible complexes with viral RNA, resulting from complete extension of base pairing initiated by loop-loop interaction.

Hairpin, dumbbell, and single-stranded phosphodiester oligonucleotides exhibit identical uptake in T lymphocyte cell lines

The cellular binding, uptake, and intracellular distribution of structured double-stranded phosphodiester oligonucleotides (decoys) have been examined in T lymphocytes using fluorescein-labeled molecules. Intracellular localization of hairpin and dumbbell decoys was similar to that of single-stranded oligonucleotides. At short incubation times, oligonucleotides were localized only in cytoplasmic vesicles, whereas at longer times, they were also found in the nucleus. Cellular uptake was dependent on temperature, time, and extracellular concentration. Oligonucleotide efflux was similar for all types of molecules and was very rapid (t1/2 = 10-15 minutes). These results imply that phosphodiester double-stranded oligonucleotides can enter cellular compartments of interest for their potential biologic function and, thus, provide a potential tool to control gene transcription.

BCR-ABL antisense oligodeoxyribonucleotides suppress the growth of leukemic and normal hematopoietic cells by a sequence-specific but nonantisense mechanism

We have examined the effect of BCR/ABL junctional antisense phosphodiester oligodeoxyribonucleotides (ODNs) on BV173 and other chronic myeloid leukemia (CML) cell lines. Various control ODNs were used to understand the mechanism of the observed antiproliferative effect. Not only the antisense ODNs but also several control ODNs inhibit the proliferation of the leukemic cell lines. All the ODNs that inhibit the cell proliferation share a TAT consensus sequence at their 3' end. A 1-base mismatch within this consensus sequence abolishes the antiproliferative effect. Mismatches of several bases at any other position within the sequence of the active ODNs do not suppress the observed effect. Similar experiments on normal or CML CD34+ cell fraction led to the same observations. We conclude that the antiproliferative effect of the phosphodiester BCR/ABL antisense ODNs cannot be attributed to an antisense mechanism but rather to a nonelucidated effect of a 3' terminal TAT sequence. This effect is not CML specific.

[Synthesis of modified oligonucleotides and use of a duplex from them with covalently bound chains]

A method for the synthesis of DNA duplex with covalently linked strands was elaborated, and the thermal and hydrolytic stability of the duplex was studied. The strands were connected via an amide bond between carboxyl and aliphatic amino groups in the presence of water-soluble carbodiimide. For this purpose, a series of modified 5- to 26-mer oligonucleotides with primary amino or carboxyl group were prepared, and their properties were investigated.

Image-guided automated core biopsies of the breast, chest, abdomen, and pelvis

PURPOSE

A prospective study was undertaken to measure success at obtaining definitive histologic diagnoses with an automated biopsy gun technique.

MATERIALS AND METHODS

Image-directed biopsies (n = 300) were performed between November 1989 and June 1992 in 288 patients. Biopsy specimens of breast lesions were obtained with ultrasound (US) or stereotaxic mammographic guidance. Biopsy samples of lesions from other areas of the body were obtained with computed tomographic or US guidance. Clinical follow-up and surgical biopsy were used to measure diagnostic accuracy.

RESULTS

Overall, a definitive diagnosis was reached with core specimen biopsy techniques in 95.3% of cases (286 of 300), and all 286 definitive findings were accurate.

CONCLUSION

Because of the success of this approach, the adoption of an automated biopsy gun technique that includes histologic examination of a core specimen should be considered by all radiologists who perform image-guided biopsies.

[Synthesis of cyclic oligodeoxyribonucleotides with non-nucleotide inserts]

An effective method for the oligonucleotide cyclization using BrCN-induced chemical ligation was developed. The novel idea to incorporate non-nucleotide inserts makes the process of cyclizations independent of the inner secondary structure of the linear precursor. An set of assays were developed to confirm the cyclic structure of the compounds obtained.

Improved anti-herpes simplex virus type 1 activity of a phosphodiester antisense oligonucleotide containing a 3'-terminal hairpin-like structure

We synthesized a series of 20-mer antisense phosphodiester oligonucleotides constituting of a 5'-dodecameric sequence, complementary to the acceptor splice junction of herpes simplex virus type 1 (HSV-1) pre-mRNAs IE4 and IE5, flanked in 3' by octameric sequences adopting hairpin-like structures of different stabilities. The presence of the minihairpins in 3' protected the 20-mer phosphodiester oligonucleotides against serum nuclease degradation, this protection being well correlated to the reported melting temperatures of the minihairpins, and to the gel mobilities of the 20-mer oligonucleotides. While no protection was observed using a linear 8-mer, the addition in 3' of the most stable minihairpin--H8--increased more than eightfold the nuclease resistance of the linear antisense dodecamer. We analyzed the effect of such a protection on the anti-HSV-1 antisense activities of the oligonucleotides. When bearing H8 in 3', the antisense dodecamer was 10 times more active than in the absence of 3'-flanking sequence, while a linear 20-mer control containing the antisense sequence was only 3 times more active. This work provides the basis for a further rational design of phosphodiester antisense oligonucleotides, taking advantage of the specific properties conferred by their conformations.

Oligonucleotide circularization by template-directed chemical ligation

An efficient method for producing the covalent closure of oligonucleotides on complementary templates by the action of BrCN was developed. A rational design of linear precursor oligonucleotides was studied, and the effect of factors such as oligonucleotide concentration and oligomer-template length ratio was evaluated. The efficiency of circularization was shown to correlate well with the secondary structure of the precursor oligomer (as predicted by a simple computer analysis), hairpin-like structures bearing free termini clearly favouring the circularization reaction. A novel idea, consisting of the incorporation of non-nucleotide insertions in the precursor oligomer (namely, 1,2-dideoxy-D-ribofuranose residues), may render this method universal and highly effective. An original set of assays was developed to confirm the circular structure of the covalently closed oligonucleotides.

Ex vivo regulation of specific gene expression by nanomolar concentration of double-stranded dumbbell oligonucleotides

Inhibition of specific transcriptional regulatory proteins is a new approach to control gene expression. Transcriptional activity of DNA-binding proteins can be inhibited by the use of double-stranded (ds) oligodeoxynucleotides that compete for the binding to their specific target sequences in promoters and enhancers. As a model, we used phosphodiester dumbbell oligonucleotides containing a binding site for the liver-enriched transcription factor HNF-1 (Hepatocyte Nuclear Factor 1). Binding affinity of HNF-1 to dumbbell oligonucleotides was the same as that to ds oligonucleotides, as determined by gel retardation assays. HNF-1 dumbbells specifically inhibited in vitro transcription driven by the albumin promoter by more than 90%. HNF-1-dependent activation of a CAT reporter plasmid was specifically inhibited when the HNF-1 dumbbell oligonucleotide was added at nM concentration to transiently transfected C33 cells. On the contrary, HNF-1 ds oligonucleotides, which displayed the same activity as the dumbbell oligonucleotides in the in vitro assays, were no more effective in the ex vivo experiments. These results might reflect the increased stability of the circular dumbbell oligonucleotides towards cellular nuclease degradation, as shown in vitro with nucleolytic enzymes. Dumbbell oligonucleotides containing unmodified phosphodiester bonds may efficiently compete for binding of specific transcription factors within cells, then providing a potential therapeutic tool to control disease-causing genes.

Self-cleavage of a 71 nucleotide-long ribozyme derived from hepatitis delta virus genomic RNA

Self-cleavage efficiency of ribozymes derived from hepatitis delta virus (HDV) has been shown to depend on the RNA structure, which in turn may be determined by the length of the considered sequences. Here we describe the construction and functional analysis of a 71 nucleotide-long RNA genomic fragment, Rz71, which carries an 18 nucleotide deletion in a very stable GC-rich stem-loop (stem IV), predicted to be present in several computer-derived secondary structure models. Rz71 is able to undergo self-cleavage under non-denaturing conditions (the t1/2 of the reaction at 37 degrees C is 3 min). The deletion, however, is not neutral, since under the same conditions the non-deleted ribozyme cleaves to 50% in less than 15 sec. Therefore, stem-loop IV seems to play a structural role, not being directly involved in the catalytic reaction, but contributing to the correct positioning of the catalytic core of the HDV ribozyme. Rz71 is the smallest self-cleaving sub-fragment of HDV genomic RNA reported so far.

Hepatic nuclear factor 1 (HNF1) shows a wider distribution than products of its known target genes in developing mouse

Hepatic nuclear factor 1 (HNF1) is a highly diverged homeoprotein that is crucial for transcription of many liver-specific genes including albumin. In particular, a minimal promoter, consisting of an HNF1-binding-site and a TATA box, is highly active only in hepatoma cell lines. The expression of the HNF1 and albumin genes has been examined in mouse embryos by in situ hybridization. At 10.5 days of gestation, the HNF1 mRNA was detected in both the hepatic primordia and visceral endoderm of the yolk sac whereas the albumin transcript was present only in the nascent liver. At later stages of development, HNF1 was detected in liver, in the epithelial cells of most of the digestive tract and in the cortex of the kidney, whereas albumin was again found only in the liver. The presence of HNF1 protein in adult kidney was demonstrated by immunodetection in gel-retardation assays and western blot analysis. These experiments show that, even though the HNF1 homeo-protein is essential for expression of many liver-specific genes, it cannot, by itself, force high expression levels of these genes, in non-hepatic tissues.

A distal dimerization domain is essential for DNA-binding by the atypical HNF1 homeodomain

Hepatic Nuclear Factor 1 (HNF1, also referred to as LFB1, HP1 or APF) is a liver-specific transcription factor required for the expression of many hepatocyte specific genes. We report here the purification of this rat liver nuclear protein and the cloning of its cDNA using a PCR-derived approach. Seven independent clones reveal 3 alternative polyadenylation sites and a unique open reading frame. Both a motif homologous to the homeodomain and a distal dimerization domain are required for specific DNA binding. Sequence comparisons reveal several atypical features at key positions in the segment corresponding to helices III and IV of the Antaennapedia homeodomain as well as a potential 24 amino acid loop in place of the universal turn between helices II and III. Together with its property to dimerize in the presence or absence of DNA, these features place HNF1 as the prototype of a novel subclass of transcription factors distantly related to homeoproteins.

Measurement of cervical length in pregnancy: comparison between vaginal ultrasonography and digital examination

Evaluation of the gravid cervix uteri is an important part of prenatal care, especially in the patient at risk for preterm birth. Seeking a method of cervical length measurement that could be used easily regardless of patient habitus, location of the cervix, and gestational age, we used a vaginal probe with a 240 degrees scanning angle in gravidas at various gestational ages to test the theoretical advantages of the wide scanning angle. Among the first 201 examinations, cervical length was measured successfully in 99.5% of cases. This success rate compares favorably with those of abdominal sonography and vaginal sonography using the standard 90 degrees scanning angle sector probes. We also compared this method with digital examination in a double-blind fashion. Only a fair degree of association between sonographic cervical measurements and measurements obtained by digital examination was found, reflected in a correlation coefficient of 0.49.

Interactions between structure transitions in a torsionally constrained DNA

We used S1 nuclease cleavage in conjunction with gel electrophoresis to evaluate torsion-induced cruciform extrusion at two inverted repeat sequences, IRS-B and IRS-C of plasmid pUC12. These structure transitions affect each other through competition for the available torsional free energy according to their relative energies of activation and the magnitude of DNA duplex unwinding associated with each transition. They can be modulated by the level of DNA negative torsion. Interplays between transition sequences occur over long distances and are independent of relative orientation of transition sites. DNA binding factors that enhance or repress structural transitions of specific sequences may, thus, regulate the structural and functional properties of torsionally coupled, distal sequences.

A liver-specific factor essential for albumin transcription differs between differentiated and dedifferentiated rat hepatoma cells

We have identified and characterized two mutually exclusive nuclear proteins that interact with a single crucial element of the albumin promoter. One, albumin proximal factor (APF), is found only in liver or differentiated hepatoma cells and is probably identical to the liver-specific factor named HNF1, alpha 1TFB, or HP1-binding protein. The other, variant albumin proximal factor (vAPF), is present in dedifferentiated hepatoma cells as well as in somatic cell hybrids that show extinction of the expression of liver-specific proteins, including albumin. Reversion to the hepatic phenotype of either a dedifferentiated variant or an extinguished somatic hybrid clone is accompanied by loss of vAPF and reappearance of APF. These two proteins differ in their thermostability and in their molecular weight, while displaying identical sequence specificities. Both proteins interact with a homologous motif present in promoter regions of several other liver-specific genes. In vitro transcription assays, using a rat liver nuclear extract, indicate that the binding of APF to its target sequence is required for albumin transcription. These results suggest that a modification in the primary structure of a transcription factor is correlated with the differentiated state of the hepatic cell.

Drosophila melanogaster H1 histone is phosphorylated stably

Phosphorylation of histone H1 is developmentally regulated in Drosophila spp. It cannot be detected in preblastoderm embryos or polytene salivary gland cells, but in cellular blastoderm, postblastoderm embryo, and amitotic adult head nuclei, it occurs with a frequency of roughly 4 x 10(5) molecules per nucleus. We used pulse-labeling to study the relationship between H1 synthesis and modification in cultured cells. These results reveal that the H1-associated phosphate is stable and suggest that Drosophila H1 is synthesized, translocated to the nucleus, associated with chromatin, and then phosphorylated. Partial tryptic digestion of Drosophila H1 revealed that the phosphorylation site is located within the globular, central domain of the protein. Thus, the developmentally regulated phosphorylation of Drosophila H1 presents two contrasts with previously studied H1 phosphorylation. It is not correlated with DNA replication, and it is located in the central domain of the protein.

Drosophila histone H2A.2 is associated with the interbands of polytene chromosomes

Drosophila chromatin contains two antigenically distinct H2A histones, H2A.1 and H2A.2. Indirect immunofluorescence analyses revealed that anti-H2A.1 binding was distributed throughout polytene chromosomes, whereas anti-H2A.2 binding was interband-specific. Thus, H2A.2 probably contributes to the less compacted structure of interbands. Since each band-interband region is thought to contain a single gene, our results suggest that the distribution of H2A.2 echoes the functional organization of the Drosophila genome. Similar H2A histones occur in eukaryotes ranging from protozoa to mammals. Their placement might be an important determinant of chromatin structure.

Drosophila virilis histone gene clusters lacking H1 coding segments

Approximately 30-40% of Drosophila virilis DNA complementary to cloned Drosophila histone genes is reduced to 3.4-kilobase-pair (kbp) segments by Bgl I or Bgl II digestion. The core histone genes of a 3.4-kbp Bgl II segment cloned in the plasmid pDv3/3.4 have the same order as the D. melanogaster core histone genes in the plasmid cDm500: H2B H3 H4 H2A. Nonetheless, pDv3/3.4 and cDm500 have different histone gene configurations: In pDv3/3.4, the region between the H2B and H3 genes contains 0.35 kbp and cannot encode histone H1; in cDm500, the region contains 2.0 kbp and encodes histone H1. The lack of an H1 gene between the H2B and H3 genes in 30-40% of D. virilis histone gene clusters suggests that changes in histone gene arrays have occurred during the evolution of Drosophila. The ancestors of modern Drosophila may have possessed multiple varieties of histone gene clusters, which were subsequently lost differentially in the virilis and melanogaster lineages. Alternatively, they may have possessed a single variety, which was rearranged during evolution. The H1 genes of D. virilis and D. melanogaster did not cross-hybridize in vitro under conditions that maintain stable duplexes between DNAs that are 75% homologous. Consequently, D. virilis H1 genes could not be visualized by hybridization to an H1-specific probe and thus remain unidentified. Our observations suggest that the coding segments in the H1 genes of D. virilis and D. melanogaster are greater than 25% divergent.(ABSTRACT TRUNCATED AT 250 WORDS)

Satellite DNA-correlated nucleosomal proteins in Drosophila virilis

Three major satellite DNAs comprise 40-45% of the genome of Drosophila virilis. Since these satellites are not substrates for most restriction enzymes, we were able to digest D. virilis nuclei with HaeIII and micrococcal nuclease and isolate chromatin fractions containing variable levels of satellite DNA. Electrophoretic analysis of these chromatin fractions revealed that the level of the acid-soluble chromosomal protein, cp17.3, was directly related to the percentage of satellite DNA in chromatin. The correlation between cp17.3 and satellite DNA abundance suggests that cp17.3 is involved in the heterochromatic condensation of satellite DNAs. cp17.3 occurs at a frequency of one molecule per 10-20 nucleosomes. It is detected in an electrophoretically distinguishable class of mononucleosomes, provisionally identified as MN1uH2A, which contains ubiquitinated histone H2A (uH2a) but lacks histone H1. It is not detected in MN1, a second class of mononucleosomes, which lacks uH2A and H1. Since cp17.3 is correlated with satellite DNAs and present in nucleosome cores, it might be a histone variant specifically associated with satellite DNAs.

Preparation of histone variants and high-mobility group proteins by reversed-phase high-performance liquid chromatography

Methods have been developed for the preparation of histone variants and high-mobility group (HMG) proteins by high-performance liquid chromatography (HPLC). The individual HPLC fractions were recovered as a dry powder in 95% yield by direct lyophilization from the column effluent. Perchloric acid-soluble H1 variants and HMG proteins from Chinese hamster cells (line CHO) were separated on a mu Bondapak CN column using a 0-50% linear acetonitrile gradient in water containing 0.2% trifluoroacetic acid (TFA). The proteins were eluted in the following order: HMG-E/G (an HMG-14/17 class proteins from CHO cells), HlO, Hl, HMG-2, and HMG-l. HMG-E/G, Hl, and an unidentified protein were recovered electrophoretically pure. HlO contained contaminants which could be removed by subsequent chromatography on a mu Bondapak C18 Radial-Pak column, but HMG-l and HMG-2 could not be completely resolved. Nucleosomal core histones were fractionated on a mu Bondapak C18 Radial-Pak column using a 30-55% linear acetonitrile gradient containing 0.2-0.3% TFA. They were eluted in the following order: H2B, (LHP)H2A, (MHP)H2A, H4, LHP(H3), and (MHP)H3, (where LHP and MHP refer to less-hydrophobic and more-hydrophobic variants). If the gradient containing 0.3% TFA was interrupted with an isocratic elution at 43% acetonitrile, the H2B, (LHP)H2A, (MHP)H2A, and H4 proteins were completely resolved, thus providing a good preparative method for these proteins. The H2A class of Drosophila histones was also fractionated on a mu Bondapak C18 Radial-Pak column using a 30-35% linear acetonitrile gradient containing 0.2% TFA. Drosophila melanogaster H2A, obtained as a single fraction by chromatography on Biol-Gel P-100, was eluted from the C18 column as three proteins. The order of elution was identified by electrophoresis to be: H2Aox (an oxidized form of H2A), D2 (a Drosophila-specific subtype), and H2A.

Drosophila nucleosomes contain an unusual histone-like protein

Mononucleosomes prepared from Drosophila melanogaster nuclei contain the four core histones H2A, H2B, H3, and H4 plus an additional histone-like, acid-soluble, chromosomal protein. It is probably the protein designated D2 by Alfageme et al. [Alfageme, C.R., Zweidler, A., Mahowald, A. & Cohen, L.H. (1974) J. Biol. Chem. 249, 3729-3736]. D2 elutes with histone H2A from a Bio-Gel P-100 column, but can be distinguished electrophoretically from H2A and from the other standard Drosophila core histones. The amino acid composition of D2 resembles the compositions of H2A and H2B. However, peptide mapping reveals that D2 is not a simple sequence variant of either H2A or H2B. D2 is present in nuclei from embryos and adult heads, and thus is not restricted to a narrowly defined developmental period. It is present in D. melanogaster and D. virilis, and thus appears to be conserved during the evolution of Drosophila. D2 is present in D. melanogaster chromatin with an approximate frequency of one molecule per five nucleosomes, and must therefore be associated with a subset of nucleosomes. The function of this protein is not known. Its presence in nucleosomes, evolutionary conservation, and comparatively large abundance all suggest that it is an important nucleosomal element. It will be interesting to learn whether this histone-like protein is encoded in a subset of the Drosophila histone gene cluster or is encoded separately.

Satellite Ic: a possible link between the satellite DNAs of D. virilis and D. melanogaster

In this study, we isolated and characterized a previously undetected cryptic satellite DNA comprising 0.1% of the total nuclear genome of D. virilis. This satellite is hidden from detection in neutral CsCl by satellite I and is therefore designated cryptic satellite I or Ic. Sequence analysis reveals that Ic is the repeating heptanucleotide [poly d(AATATAG): d(CTATATT)]. It is more closely related to the three simple sequence satellite DNAs of D. melanogaster, a distantly related species, than it is to any of the major D. virilis satellite DNA sequences. Ic may therefore be a link between the simple sequence satellites of D. virilis and D. melanogaster. As an extension of this theory, we have constructed a "family tree" linking the satellites of D. virilis and D. melanogaster by a series of "simple" operations. Only one intermediate required by this evolutionary scheme has not yet been identified.

Anomalous electrophoretic mobility of Drosophila phosphorylated H1 histone: is it related to the compaction of satellite DNA into heterochromatin?

In embryonic nuclei of Drosophila virilis, 45% of the DNA is satellite, and congruent to 50% of the H1 histone is phosphorylated. In polytene salivary gland nuclei, less than 1% of the DNA is satellite, and less than 10tion. The phosphorylated H1's migrate 4% slower than the unphosphorylated H1's on SDS-acrylamide gels. The mobility difference may arise because the phosphorylated and unphosphorylated H1's have different conformations in SDS. This putative conformational difference could be essential to the compaction of satellite DNA into heterochromatin.

Correlation between phosphorylated H1 histones and satellite DNAs in Drosophila virilis

Drosophila virilis DNA contains satellites I, II, and III. D. novamexicana DNA contains satellite I. D. virilis H1 histone contains subfractions a, b, c, d, and e; D. novamexicana H1 contains subfractions a, b, and c. Therefore, satellites II and III might be correlated with H1d and H1e. To test the validity of this correlation, the H1 histones of polytene nuclei, which contain less than 1% satellite DNA, were analyzed. Polytene nuclei of D. virilis contain substantially decreased levels of H1c and H1e and marginally decreased levels of H1d. Polytene nuclei of D. novamexicana contain decreased levels of H1c.H1c is correlated with satellite I (common to D. virilis and D. novamexicana); H1e is correlated with satellites II and III; H1d is not correlated with any satellite DNA, because its level is virtually unchanged in polytene cells lacking detectable amounts of satellite DNA. Alkaline phosphatase digestion of the H1 histones reveals that H1c is the phosphorylated form of H1b and H1e is the phosphorylated form of H1d. Therefore, the under-replication of satellite DNAs is correlated with the decreased phosphorylation of H1 histones. In vitro, D. virilis H1 histones preferentially bind D. virilis DNAs in the progression III greater than II greater than I greater than main band, whereas D. virilis core histones do not preferentially bind any D. virilis DNA. As an extension of these results, we suggest that phosphorylated H1 histones bind D. virilis satellite DNAs in vivo and are involved in the compaction of heterochromatin.

A family of three related satellite DNAs in Drosophila virilis

Isolated single DNA strands of satellites II or III of Drosophila virilis form hybrid duplexes with the complementary single strands of satellite I. The hybrids denature at a higher temperature than controls, renature rapidly, and form bands of hybrid density in neutral CsCl. Isolated single strands of satellite II do not form clear-cut duplexes with the complementary strands of satellite III. Mixtures of satellites II and III denature at the same temperature as controls, and do not form bands in neutral CsCl. Therefore, satellite I-II and I-III complexes are extensively base paired, while satellite II-III complexes are minimally base paired. These experiments demonstrate partial homology among the three satellites, and suggest an evolutionary relationship among them. They also suggest that satellites and other repeated sequences from related species, which do not form hybrids in vitro, could have a common evolutionary origin, but have accumulated enough base substitutions to lose interspecific homology demonstrable by hybrid duplex formation.