Complete nucleotide sequence of the Drosophila transposable element copia: homology between copia and retroviral proteins

Retrotransposons and the evolution of mammalian gene expression

Transposable elements, and retroviral-like elements in particular, are a rich potential source of genetic variation within a host's genome. Many mutations of endogenous genes in phylogenetically diverse organisms are due to insertion of elements that affect gene expression by altering the normal pattern of regulation. While few such associations are known to have been maintained over time, two recently elucidated examples suggest transposable elements may have a significant impact in evolution of gene expression. The first example, concerning the mouse sex-limited protein (Slp), clearly establishes that ancient retroviral enhancer sequences now confer hormonal dependence on the adjacent gene. The second example shows that within the human amylase gene family, salivary specific expression has arisen due to inserted sequences, deriving perhaps from a conjunction of two retrotransposable elements.

The linear plasmid pMC3-2 from Morchella conica is structurally related to adenoviruses

pMC3-2, one of two linear plasmids localised in the mitochondria of the ascomycete Morchella conica, was completely sequenced. It is 6044 bp in size, contains terminal inverted repeats of 713 and 710 bp length and two open reading frames, ORF1 and ORF2, spanning 2706 bp and 918 bp, respectively. ORF1 probably encodes a viral B-type DNA-polymerase. Concerning ORF2, no homology to any other published protein- or DNA-sequence could be detected. According to the structure of DNA-polymerases, linear plasmids can be grouped into two classes reflecting their localisation either in the cytoplasm or within the mitochondria. In general, the structure of plasmid pMC3-2, as well as of other linear plasmids from filamentous fungi, indicates a close relationship of these genetic elements to adenoviruses.

Preferential Y chromosomal location of TRIM, a novel transposable element of Drosophila miranda, obscura group

We have isolated a novel transposable element from the Y chromosome of Drosophila miranda (obscura group) which shows an organization intermediate between that of typical retroviruses and the I factor of Drosophila melanogaster. The site of integration of this element, TRIM, is an inverted repeat. On the 5' side the central region of TRIM is bordered by a sequence homologous to the tRNA primer binding site (PBS) and on the 3' side by a sequence homologous to the polypurine tract (PPT). The 5' LTR (long terminal repeat) contains transcription signals, such as a TATA box and a polyadenylation sequence. Three long open reading frames (ORFs) are found within the central region of the transposon. ORF1 and ORF2 show amino acid homology to the I factor of D. melanogaster and to conserved amino acid residues of retroviral reverse transcriptases. When used as a probe for in situ hybridization the TRIM element labels about 25 euchromatic sites and the chromocenter in polytene nuclei of D. miranda females. In the chromosome complement of males the polytenized part of the Y chromosome shows about the same number of additional TRIM homologous elements. Thus, as a result of this enrichment of TRIM elements in the Y chromosome, the genomic size of the TRIM population in D. miranda males is approximately doubled.

Characterization of the genome of rice tungro bacilliform virus: comparison with Commelina yellow mottle virus and caulimoviruses

Rice tungro disease is caused by an infection of two different viruses, rice tungro spherical virus (a (+) sense RNA virus) and rice tungro bacilliform virus (RTBV) with a genome of circular double-stranded DNA. The genome of an RTBV isolate from the Philippines was cloned, sequenced, and found to be 8000 bp in length. It contains four open reading frames (ORFs) on a single strand, with ORF 1 having an internal termination codon (TAA). The 5' and 3' ends of a polyadenylated viral RNA transcript, of genome length, were mapped by primer extension and cDNA sequence analysis, respectively. The transcript is terminally redundant by 265-268 nucleotides. Purified virus particles contain two major proteins with molecular masses of 37 and 33 kDa, although only the 37-kDa protein was detected in the infected rice tissues. The N-terminal amino acid sequence of the 33-kDa protein was determined and its coding region was identified on the RTBV genome. The identity of the coat protein gene was further confirmed by expressing a region of the genome in Escherichia coli, the products of which reacted with anti-RTBV antibody. The unusually long ORF 3 of RTBV is predicted to encode a polyprotein of 194.1 kDa that includes: the coat protein(s), viral proteinase, reverse transcriptase, and ribonuclease H. The sections of the polyprotein show varying degrees of similarity to the counterparts of Commelina yellow mottle virus (a member of the proposed badnavirus group) and caulimoviruses. The functions of the other three ORFs are unknown.

Proteolytic processing of pol-TYB proteins from the yeast retrotransposon Ty1

Using antibodies directed against the TYB1 protein of the transpositionally competent retrotransposon Ty1-H3, we have identified three mature proteins of 23, 60, and 90 kDa and processing intermediates of 140 and 160 kDa that are derived from the 190-kDa TYA1-TYB1 polyprotein. Mature proteins and variable amounts of the precursors cofractionate with Ty viruslike particles. The map locations and precursor-product relationships of mature TYB1 polypeptides suggest that p23 is Ty1 protease, p90 is integrase, and p60 contains reverse transcriptase and RNase H. Immunoprecipitation and immunoblot analyses of Ty1 proteins show that p190 is cleaved to form p160. The p160 intermediate is cleaved to form p23 and p140, and p140 is cleaved to form p90 and p60. Processing of TYB1 proteins is dependent on Ty1 protease. Immunoblot analysis of TYB proteins from different Ty1 isolates reveal that correct processing of TYB1 proteins is a characteristic of functional Ty1 elements, whereas aberrant processing is a common defect found in transposition-incompetent elements.

How do viral reverse transcriptases recognize their RNA genome?

Reverse transcription is not solely a retroviral mechanism. Hepadnaviruses and caulimoviruses have RNA intermediates that are reverse transcribed into DNA. Moreover non-viral retroelements, retrotransposons, use reverse transcription in their transposition. All these retroelements encode reverse transcriptase but each group developed their own expression modes capable of assuring a specific and efficient replication of their genomes.

Identification of an unusual structure in the Drosophila melanogaster transposable element copia: evidence for copia transposition through an RNA intermediate

The Drosophila melanogaster transposable element copia is usually 5 kb long with long terminal repeats (LTRs), and its major transcripts are a full-length 5-kb RNA and a 2-kb RNA. We have previously shown that the 2-kb RNA is generated through splicing. Here, we have cloned a genomic intronless copia using an oligodeoxyribonucleotide probe which is specific for the junction of the two exons. The unusual copia is bounded by two LTRs and lacks precisely the intron of the 2-kb copia RNA. Identification of genomic intronless copia strongly suggests that copia transposes through an RNA intermediate. Moreover, we have found that copia virus-like particles (VLPs), in which reverse transcription of copia RNA seems likely to occur, packages the spliced copia RNA much less efficiently than the full-length copia RNA. This result leads to the suggestion that much lower copy number of genomic intronless copia, as compared with that of 'normal' copia, may be responsible for the inefficient packaging of the spliced copia RNA into the VLP.

Autoprocessing of Drosophila copia gag precursor to generate a unique laminate structure in Escherichia coli

Drosophila copia protease is likely to be encoded in the gag gene. We have expressed copia gag polyprotein precursor in E. coli. The gag precursor was correctly processed to generate a unique laminate structure in E. coli. The processing was almost completely blocked by a mutation at the putative active site of copia protease, and resulted in accumulation of the precursor. Furthermore, the laminate structure was not found in E. coli expressing the mutant precursor. These results indicate that the protease is involved in cleaving the gag precursor itself. Also, the assembly of copia gag protein should correlate to the autoprocessing of copia gag polyprotein precursor.

A family of retrotransposons and associated genomic variation in wheat

A family of related retroelements was characterized in the genomes of some Graminease species. The structure of these retroelements indicates that they are retrotransposons containing reading frames with sequence similarity to the polyproteins of copia and Ty. This family of retroelements (termed WIS-2) occurs in the genomes of barley, wheat, rye, oats, and Aegilops species. Ongoing genomic variation both within individual plants of a wheat variety and within and between varieties of wheat is associated with some members of the WIS-2 family.

Processed pseudogenes in Drosophila

Two species of Drosophila, D. yakuba and D. teissieri, possess pseudogenes of Adh. These pseudogenes lack introns and map to chromosome arm 3R, rather than to chromosome arm 2L, wherein are located the functional Adh genes. Their structure suggests that the pseudogenes arose from reverse transcripts. Because the pseudogenes map to homologous sites in both species, they presumably arose before these species diverged. Remarkably, the pattern of base substitution in the pseudogenes differs between sites that correspond to degenerate and non-degenerate codon positions in their functional paralogs.

Characterization of products of TY1-mediated reverse transcription in Saccharomyces cerevisiae

Transposition of the yeast transposable element, Ty, has been shown to require a reverse transcription process. By analysing the extrachromosomal Ty-specific nucleic acid molecules associated with overproduced Ty virus-like particles (Ty-VLPs), we identified several reverse transcribed cDNA strands. Most of them resemble the characteristic intermediates of the reverse transcription process described for authentic retroviruses: a (-) strong-stop DNA strand covalently bound to an RNA primer, two elongated (-) strands with one or two long terminal repeat (LTR) sequences and a (+) strong-stop DNA. Surprisingly, complete (+) strands and full-length linear duplex Ty DNA could not be detected. The structural features of two additional (+) strands may indicate some differences between the mechanisms of (+) strand synthesis in Ty and other retrotransposons or retroviruses.

Zinc finger motif for single-stranded nucleic acids? Investigations by nuclear magnetic resonance

Nuclear magnetic resonance (NMR) methods have been used to address issues regarding the relevance and feasibility of zinc binding to "zinc finger-like" sequences of the type C-X2-C-X4-H-X4-C [referred to as CCHC or retroviral-type (RT) zinc finger sequences]. One-dimensional (1D) NMR experiments with an 18-residue synthetic peptide containing the amino acid sequence of an HIV-1 RT-zinc finger domain (HIV1-F1) indicate that the sequences are capable of binding zinc tightly and stoichiometrically. 1H-113Cd spin echo difference NMR data confirm that the Cys and His amino acids are coordinated to metal in the 113Cd adduct. The 3D structure of the zinc adduct [Zn(HIV1-F1)] was determined to high atomic resolution by a new NMR-based approach that utilizes 2D-NOESY back-calculations as a measure of the consistency between the structures and the experimental data. Several interesting structural features were observed, including (1) the presence of extensive internal hydrogen bonding, and (2) the similarity of the folding of the first six residues to the folding observed by X-ray crystallography for related residues in the iron domain of rubredoxin. Structural constraints associated with conservatively substituted glycines provide further rationale for the physiological relevance of the zinc adduct. Similar NMR and structural results have been obtained for the second HIV-1 RT-zinc finger peptide, Zn(HIV1-F2). NMR studies of the zinc adduct with the NCP isolated directly from HIV-1 particles provide solid evidence that zinc finger domains are formed that are conformationally similar (if not identical) to the peptide structures.(ABSTRACT TRUNCATED AT 250 WORDS)

Occurrence of a copia-like transposable element in one of the introns of the potato starch phosphorylase gene

The gene coding for starch phosphorylase (EC 2.4.1.1) was isolated from a potato genomic library constructed in lambda EMBL3. It is an unusually long plant gene (16.4 kb) which encodes a preprotein of 966 amino acids. The phosphorylase coding sequence is interrupted by 14 introns whose positions do not match those of the introns in the human glycogen phosphorylase gene. A 78 amino acid central peptide unique to plant plastidial phosphorylases is hypothesized to have arisen through the mis-splicing of an intron-exon junction site in an ancestral gene. The fifth intron of the phosphorylase is very large (approximately 7 kb) and contains a copia-like transposable element inserted in the opposite orientation to that of the phosphorylase gene. This element has been named Tst1; it is bordered on the 5' and 3' sides by long terminal repeats of 285 and 283 bp respectively, which define an internal domain of 4492 bp. Tst1 contains 4 open reading frames (ORFs) that encode protein domains for a reverse transcriptase, an integrase, an RNA-binding site and a protease. Transcription of the phosphorylase gene appears to proceed unimpaired through the copia element.

Single copy sequences in galago DNA resemble a repetitive human retrotransposon-like family

Galago DNA contains a few single copy sequences that are homologous to the human THE 1 family of repeats. Two of these galago loci have been isolated as genomic clones and their structures are compared to the THE 1 consensus sequence. Whereas the human sequence resembles a proretroviral transposon, the galago sequences provide no evidence for a proretroviral sequence organization. The two galago clones share a common repeat sequence, which is homologous to the U5 region of the THE 1 long terminal repeat. Immediately 3' to this repeat, each galago clone contains sequences that are homologous to mutually exclusive regions of the internal THE 1 sequence. Thus, the human THE 1 sequence can be represented as a mosaic of the two ancestrally related galago loci. The galago loci are transcribed in vivo, so that their conservation in the primate genome could be selected. Human THE 1 repeats apparently resulted by recruiting preexisting cellular sequences via a retrovirally mediated process.

Interactions among modifiers of retrotransposon-induced alleles of the white locus of Drosophila melanogaster

Mutations in five loci that modify the phenotype of whiteapricot (wa), caused by the retrotransposon, copia, were examined in two-way combinations to determine whether their effects were additive or epistatic. All two-way combinations of mutations in these five loci, mottler of white (mw), suppressor of forked (su(f], suppressor of white apricot (su(wa], Enhancer of whiteapricot, (E(wa] and Darkener of apricot (Doa), are additive in their effects on wa, implying that each second-site modifier locus affects a different process. Three other copia-induced mutations, HwUa, whd81b25 and ctns were also examined for responsiveness to mutations in these modifier loci. None clearly responded. Mutations associated with B104 insertions, including Gl, vgni, ctn and wric were also examined for responsiveness to mw mutations, which have specificity for this element as well. Both vgni and wric respond to mutations in mw. The former interaction demonstrates that mw is capable of interacting with B104 elements in loci other than white. The significance of the results with respect to the nature of second-site modifier loci is discussed.

Identification of a second retrotransposon-related element in the genome of Physarum polycephalum

The repetitive fraction of the genome of the eukaryotic slime mould Physarum polycephalum is dominated by Tp1, a family of retrotransposon-like sequences. Tp1 elements are arranged in scrambled clusters probably arising from integration of the element into copies of its own sequence. The present report describes a second sequence family, Tp2, which has been identified within cloned DNA segments of scrambled Tp1 sequences. Like Tp1, the Tp2 element is structurally related to retrotransposons, having long terminal direct repeats and being flanked by an apparent target site duplication, but its relatively short length (1.68 kb) indicates that it is probably incapable of encoding all the functions necessary for its own mobilisation. Analysis of the coding potential of the Tp2 element supports this view, although a striking homology to a nucleic acid binding domain common to many retrotransposons was identified. As with Tp1, putative regulatory signals can be identified in the LTRs of Tp2. Identical arrangements of Tp2 with respect to Tp1 in more than one independently derived clone indicate that non-functional copies of Tp2 may be mobilised as part of a Tp1 transcriptional unit.

Autonomous expression of RTVL-H endogenous retroviruslike elements in human cells

Northern (RNA) blot analysis of RNA from various human cell lines and tissues has demonstrated that elements belonging to the RTVL-H family of human endogenous retroviruslike sequences are expressed in several cell types. The highest levels of RTVL-H-related RNAs were observed in teratocarcinoma cell line NTera2D1, HeLa cells, two bladder carcinoma cell lines, and normal amniotic tissue. Expression was also observed in normal chorion and in some other cell lines. The RTVL-H transcription pattern varied among the different cell types, but several expressed a unit-length 5.6-kilobase transcript. Characterization of cDNA clones corresponding to transcripts present in NTera2D1 cells indicates that the complex transcription pattern observed in these cells is generated by the following: (i) transcription of both full-length and deleted genomic elements, which is initiated within the 5' long terminal repeat (LTR) and, in all but one case, polyadenylated in the 3' LTR; (ii) the splicing of both unit-length transcripts and transcripts from a deleted element; (iii) transcription involving solo LTR sequences; and (iv) transcription which, in one case, reads through the 3' LTR into flanking cellular sequences. Sequence data obtained from 25 cDNA clones revealed that at least 13 RTVL-H elements are expressed in NTera2D1 cells. The positions of several termination codons within the pol region are the same among nine different elements, indicating that an ancestral RTVL-H element bearing these mutations dispersed within the genome. We also found that RTVL-H expression varied among samples of amnion and chorion tissue isolated from different individuals. These findings demonstrate that regulated autonomous expression of RTVL-H sequences occurs in human cells.

Type I (R1) and type II (R2) ribosomal DNA insertions of Drosophila melanogaster are retrotransposable elements closely related to those of Bombyx mori

Approximately 50% of the ribosomal DNA (rDNA) units of Drosophila melanogaster are inactivated by two different 28 S RNA ribosomal gene insertions (type I and type II). We present here the nucleotide sequence of complete type I and type II elements. Conceptual translation of these sequences revealed open reading frames (ORFs) encoding amino acid residues conserved in all retrotransposable elements. Full-length type I elements are 5.35 x 10(3) base-pairs in length and contain two overlapping ORFs. The smaller ORF (471 amino acid residues) has similarity to gag genes, while the larger ORF (1021 residues) has similarity to pol genes. Full-length type II elements are 3.6 x 10(3) base-pairs and contain one large ORF (1056 residues) that appears to represent a gag-pol fusion. Type I and type II elements are similar in structure, in the proteins they encode, and in insertion specificity to the R1Bm and R2Bm retrotransposable elements of Bombyx mori. We suggest that the D. melanogaster elements be called R1Dm and R2Dm, to reflect their structure as retrotransposons. Comparison of the R1 and R2 elements from these two widely different species revealed regions of the ORF that are likely to play an important role in the propagation of the elements. Four distinct regions of sequence conservation separated by regions of little or no sequence similarity were detected for both the R1 and R2 elements: (1) cysteine motifs of the gag gene, with three such motifs for R1 and one motif for R2; (2) a reverse transcriptase domain; (3) an integrase domain located carboxyl terminal to the reverse transcriptase region; and (4) a small region amino terminal to the reverse transcriptase domain, whose function is not known. The level of identity of the amino acid residues for these segments is 28 to 34% between the R1 elements, and 34 to 39% for the R2 elements. Finally, it may be predicted that the mechanism of unequal crossover might eventually eliminate R1 and R2 from the rDNA locus. The long history of selection at the protein level exhibited by these elements indicates that it is their active transposition that maintains them in the locus. The high level of sequence homogeneity between copies of each element within the same species is consistent with the high turnover rate expected to result from these processes.

S71 is a phylogenetically distinct human endogenous retroviral element with structural and sequence homology to simian sarcoma virus (SSV)

Human endogenous retroviral element S71 had previously been shown to contain gag- and pol-related regions and a 3' LTR-like sequence. The nucleotide sequence of S71 was determined and compared with the corresponding regions of SSV and its helper virus SSAV. The 1.48-kb S71 gag region consists of matrix protein p15 (MA)-, capsid protein p30 (CA)-, and nucleocapsid protein p10 (NC)-related sections and the 1.82-kb pol region of tether, RNase H (RH), and endonuclease/integrase (IN) sections. The S71 nucleotide sequence contains a 167 amino acid open reading frame encompassing MA. The boundaries of the S71 element are delimited by direct repeats and the entire element is 5.4 kb long. Similarity between S71 and the v-sis-bearing, defective SSV provirus also covers overall structural organization, including the presence of presumably nonretroviral sequences. Both the gag and the pol regions of S71 contain sequences highly conserved in numerous retroviruses. Phylogenetic analysis with conserved CA, RH, and IN sequences showed that of all other (C-type) human retroviral elements available for comparison, S71 is most closely related to infectious primate and murine retroviruses. This suggests that S71 represents a phylogenetic subgroup of its own. In addition we identified short ranges of conserved amino acid sequences within C-type retroviral gag and pol genes sufficient for phylogenetic analysis. Use of these may facilitate large-scale phylogenetic evaluation of C-type retroviral elements and allow rapid classification of new elements.

Structural and functional characterization of the unusually short long terminal repeats and their adjacent regions of a novel endogenous avian retrovirus

We have cloned the long terminal repeats and their flanking regions from four different proviruses belonging to a large, highly conserved, novel family of avian endogenous retroviruses. This family, termed the endogenous avian retrovirus (EAV) family, is distinct from the previously characterized avian endogenous and exogenous retroviruses. We have analyzed the sequences of the long terminal repeats and their adjacent noncoding viral sequences, including the gag leader region and the 3' noncoding region, of several different members of the EAV family and have found that the regulatory region of these novel viruses contains several unique features. The LTRs of the EAV proviruses are extremely short (243 bp long) but contain all of the essential regulatory features of longer avian retrovirus LTRs. The gag leader region and the 3' noncoding region of the novel EAVs are only weakly related to those of other avian retroviruses. Northern blot hybridization analysis of RNA from Line-0 chicken embryos reveals several transcripts derived from the EAV proviruses. Primer extension analysis indicates that all transcripts initiated from 5' proviral LTRs are initiated at the predicted +1 position within the EAV LTRs. The relative shortness, sequence divergence from other known LTRs, and the retention of the transcriptional integrity of the EAV LTRs make these LTRs an interesting model system for LTR function and for study of the potential involvement of such highly conserved retroviral elements in development.

Plus-strand priming by Moloney murine leukemia virus. The sequence features important for cleavage by RNase H

The reverse transcriptase-associated RNase H activity is responsible for producing the plus-strand RNA primer during reverse transcription. The major plus-strand initiation site is located within a highly conserved polypurine tract (PPT), and initiation of DNA replication at this site is necessary for proper formation of the viral long terminal repeats (LTRs). We present here a compilation of PPT sequences from an evolutionarily diverse group of retroviruses and retrotransposons, which reveals that there is a high degree of sequence conservation at this site. Furthermore, we found previously that secondary plus-strand origins, identified in vitro, also show strong similarity to the PPT. Taken together, these data suggest that RNase H recognizes a specific sequence at the PPT as a signal to cleave the RNA at a precise location, producing a primer for the initiation of plus-DNA strands. We have analyzed the RNase H recognition sequence by producing a large number of single and double mutations within the PPT. Our findings suggest that no single residue in the +5 to -6 region (where the cleavage occurs between -1 and +1) is essential; mutations at these positions introduced heterogeneity at the cleavage site, but cleavage is still predominantly at the correct location. Furthermore, base-pairing is not required at the +1 position of the RNase H cleavage site, but a mismatched base-pair at the -1 position causes imprecision in the cleavage reaction. Interestingly, the A residue at position -7 seems to be critical in positioning the RNase H enzyme for correct cleavage. The preference of the enzyme for cleaving between G and A residues may play a minor role in determining the specificity.

The intron of a plastid gene from a green alga contains an open reading frame for a reverse transcriptase-like enzyme

Plastid (pt) and mitochondrial (mt) DNA were isolated from the unicellular green alga KS3/2, which is presumed to be a species of the genera Ankistrodesmus or Monoraphidium. The DNA species are characterized by their different densities (pt, 1.685 g/ml; mt, 1.695 g/ml), individual restriction patterns, and their respective sizes of 130 and 47 kb. Using an intronic sequence from fungal mitochondria as a hybridization probe, intron sequences from mtDNA and from ptDNA were identified. DNA sequencing of the cloned ptDNA intron revealed that the plastid gene for subunit IV of the cytochrome b6/f petD complex is interrupted by a group II intron of unusual length (3533 bp). This was confirmed by transcript analysis. The intron encodes an open reading frame (ORF) which shows significant homology with reverse transcriptase genes from various genetic elements. This discovery is unique for plastomes and indicates that introns, retrotransposons, insertion elements and retroviruses may have a common evolutionary origin.

Composite transposable elements in the Xenopus laevis genome

Members of two related families of transposable elements, Tx1 and Tx2, were isolated from the genome of Xenopus laevis and characterized. In both families, two versions of the elements were found. The smaller version in each family (Tx1d and Tx2d) consisted largely of two types of 400-base-pair tandem internal repeats. These elements had discrete ends and short inverted terminal repeats characteristic of mobile DNAs that are presumed to move via DNA intermediates, e.g., Drosophila P and maize Ac elements. The longer versions (Tx1c and Tx2c) differed from Tx1d and Tx2d by the presence of a 6.9-kilobase-pair internal segment that included two long open reading frames (ORFs). ORF1 had one cysteine-plus-histidine-rich sequence of the type found in retroviral gag proteins. ORF2 showed more substantial homology to retroviral pol genes and particularly to the analogs of pol found in a subclass of mobile DNAs that are supposed retrotransposons, such as mammalian long interspersed repetitive sequences, Drosophila I factors, silkworm R1 elements, and trypanosome Ingi elements. Thus, the Tx1 elements present a paradox by exhibiting features of two classes of mobile DNAs that are thought to have very different modes of transposition. Two possible resolutions are considered: (i) the composite versions are actually made up of two independent elements, one of the retrotransposon class, which has a high degree of specificity for insertion into a target within the other, P-like element; and (ii) the composite elements are intact, autonomous mobile DNAs, in which the pol-like gene product collaborates with the terminal inverted repeats to cause transposition of the entire unit.

An unusually compact retrotransposon in maize

Unlike any other known plant transposon, the maize transposable element Bs1 is similar to the retrotransposons previously described in yeast and Drosophila. Bs1 is bounded by 302 bp identical long terminal repeats (LTRs), and it contains open reading frames with apparent amino acid sequence similarity to reverse transcriptase and other retroviral pol gene enzymes. Bs1 is 3203 bp long, very short for a retrotransposon, and the apparent organization of its genetic information is significantly different from any previously described element. Although transcription of Bs1 has not been observed, it is probably an active transposon, since it was observed to transpose in a maize line that contains only two sequences hybridizing to Bs1 probes. Both of these sequences share 35 restriction sites with the cloned Bs1 element, and thus must be very similar or identical with it.

Characterization of Moloney murine leukemia virus mutants with single-amino-acid substitutions in the Cys-His box of the nucleocapsid protein

To study the function of the retroviral nucleocapsid protein (NC), we have constructed point mutations in the gag gene of Moloney murine leukemia virus (MuLV) that affect a conserved cysteine-histidine motif of NC. The mutants were characterized biologically and biochemically. Cell lines producing the mutant virions were constructed in NIH 3T3 and rat2 cells, and the viral particles released by these cells were characterized for protein and RNA content. The results indicated that most mutations block replication and specifically inhibit the packaging of the MuLV genomic RNA. In some of the mutants, the packaging of the endogenous rat VL30 RNA was not affected as profoundly as was MuLV RNA. NC also seems to have another function distinct from dimer formation and packaging: one mutation reduced viral RNA packaging by only fivefold but completely abolished viral cDNA synthesis, suggesting a defect in reverse transcription.

Characterization of a nondeleterious L1 insertion in an intron of the human factor VIII gene and further evidence of open reading frames in functional L1 elements

We have characterized an insertional event in IVS-10 of the factor VIII gene in a pedigree containing a hemophilia A patient (JH-25). The inserted DNA is a 5' truncated L1 element that is 681 bp long followed by a 3'66-bp poly(A) tract. The L1 element is inserted 154 bp 5' to the start of exon 11 and is flanked by a 13- to 17-bp target site duplication. The L1 insertion is present in four generations of the patient's family. The maternal grandfather who carries the insertion does not have hemophilia A, indicating that the insertion is not the cause of hemophilia A in the patient. We have sequenced this insertion and two previously reported de novo L1 insertions in the factor VIII gene in patients JH-27 (3785 bp) and JH-28 (2132 bp). The three nucleotide sequences differ by 0.2-0.8%. All three of these L1 insertions have open reading frames (ORFs) (1192, 642, and 157 aa) and the three derived amino acid sequences are 98-99% identical. The previously reported sequence similarity between L1 3' ORFs and the polymerase domain of reverse transcriptases is maintained in the ORFs of the JH-27 and JH-28 L1 insertions. The presence of open reading frames and the close sequence similarity of these recently inserted L1 elements provide indirect evidence for the existence of a set of functional L1 elements that encode one or more proteins necessary for their retrotransposition.

Retrotransposon transposition intermediates are encapsidated into virus-like particles

Virus-like particles (VLPs) possessing reverse transcriptase activity are persistently present in Drosophila melanogaster cultured cells and are formed in yeast induced for transposition. Different retrotransposon transposition intermediates consistent with those expected from the model of reverse transcription pathway of retrotransposon transposition have been detected during the analysis of nucleic acids isolated from VLPs. These data indicate that the act of reverse transcription takes place in VLPs which may be considered as functional intermediates of transposition.

Synergistic activation and repression of transcription by Drosophila homeobox proteins

We have used a transient expression assay employing Drosophila tissue culture cells to study the potential of several Drosophila homeobox proteins to function as transcriptional regulators. A 96 bp fragment from the promoter region of the segment polarity gene engrailed, previously shown to contain five copies of a 10 bp consensus binding site for these proteins, enhanced transcription in the presence, but not the absence, of several different homeobox protein expression vectors. It is interesting that cotransfection with combinations of expression vectors encoding the homeobox proteins fushi tarazu, paired, and/or zen resulted in substantial synergistic increases in expression. In contrast, the products of the even-skipped and engrailed genes were found to repress, or quench, the activation induced by the other proteins. We discuss the implications of these results with respect to the role of homeobox genes in the control of embryonic development, and propose a "multi-switch" model whereby the activity of a target gene depends on the interactions of different homeobox proteins with multiple copies of a common binding site.

Tnt1, a mobile retroviral-like transposable element of tobacco isolated by plant cell genetics

Transposable elements can be identified by their ability to induce mutant alleles at new loci. The retrotransposon family is thought to transpose through an RNA intermediate and has many similarities to vertebrate proretroviruses. In plants, retrotransposons have been described in maize, Arabidopsis and wheat, and non-viral retroposons in maize. Most of these elements, however, have been found as non-mobile integrated units. Here, we report the isolation of the first tobacco (Nicotiana tabacum) transposable element, Tnt1, which seems to be the most complete mobile retrotransposon characterized in higher plants. Tnt1 has been isolated after its transposition into the nitrate reductase (NR) structural gene of tobacco, and transposition events have been detected through in vitro selection of spontaneous NR-deficient (NR-) mutant lines in cell cultures derived from tobacco mesophyll protoplasts. Tnt1 is 5,334 nucleotides long, contains two 610-base-pair-long terminal repeats and a single open reading frame of 3,984 nucleotides. Comparison of the Tnt1 open reading frame coding potential with those of the Drosophila melanogaster copia retrotransposon, yeast Ty retrotransposon, and vertebrate proretroviruses revealed that Tnt1 is closely related to copia and carries all the functions known to be required for autonomous transposition by reverse transcription.

Secondary structural analysis of retrovirus integrase: characterization by circular dichroism and empirical prediction methods

The retrovirus integrase (IN) protein is essential for integration of viral DNA into host DNA. The secondary structure of the purified IN protein from avian myeloblastosis virus was investigated by both circular dichroism (CD) spectroscopy and five empirical prediction methods. The secondary structures determined from the resolving of CD spectra through a least-squares curve fitting procedure were compared with those predicted from four statistical methods, e.g., the Chou-Fasman, Garnier-Osguthorpe-Robson, Nishikawa-Ooi, and a JOINT scheme which combined all three of these methods, plus a pure a priori one, the Ptitsyn-Finkelstein method. Among all of the methods used, the Nishikawa-Ooi prediction gave the closest match in the composition of secondary structure to the CD result, although the other methods each correctly predicted one or more secondary structural group. Most of the alpha-helix and beta-sheet states predicted by the Ptitsyn-Finkelstein method were in accord with the Nishikawa-Ooi method. Secondary structural predictions by the Nishikawa-Ooi method were extended further to include IN proteins from four phylogenetic distinct retroviruses. The structural relationships between the four most conserved amino acid blocks of these IN proteins were compared using sequence homology and secondary structure predictions.

A copia-like transposable element family in Arabidopsis thaliana

The fast generation time, small genome size and extensive genetic map of the crucifer Arabidopsis thaliana have made it the subject of an increasing number of studies in plant molecular genetics. As transposable elements have greatly facilitated genetic analysis in a variety of species, we have attempted to identify an endogenous A. thaliana transposable element. We report here the discovery of a family of such elements, which we refer to as Ta1 elements. Sequence analysis of one such element shows that it is closely related to retrotransposons and integrated retroviral proviruses, being bound by a direct sequence repeat and having an open reading frame with clear sequence similarity to the polyprotein of the Drosophila melanogaster retrotransposon copia. The sequence of an empty target site of a Ta1 element shows that insertion is accompanied by a five-base-pair target-site duplication and that Ta1 has transposed in the period of time since divergence of two races of A. thaliana.

Functional expression of a sequence-specific endonuclease encoded by the retrotransposon R2Bm

A fraction of the 28S ribosomal genes in certain insect species is interrupted by the insertion elements R1 and R2. These two elements from the silkworm Bombyx mori (R1Bm and R2Bm) are retrotransposons capable of transposing in a highly sequence-specific manner. We report here the functional expression in E. coli of the entire single open reading frame of R2Bm and show that it encodes a double-stranded endo-nuclease (integrase) that can specifically cleave the 28S gene at the R2 insertion site. The resulting cleavage is a 4 bp staggered 5' overhang. Deletion analysis of the 28S gene revealed that the DNA sequence required for specific cleavage is asymmetric with respect to the actual insertion (cleavage) site, with fewer than 10 bp required at one side and at least 24 bp at the other side of the site. A model is proposed based on these and previous data to account for the sequence-specific integration of the R2 retrotransposon.

Retrotransposon-like sequences are expressed in Y chromosomal lampbrush loops of Drosophila hydei

The DNA sequence family micropia consists of repeated DNA sequences that occur dispersed in the genome of Drosophila hydei. Members of this DNA sequence family were recovered from two recombinant DNA clone banks obtained by microdissection of the two Y chromosomal lampbrush loop threads and pseudonucleolus from primary spermatocyte nuclei. Nucleotide sequence analysis of two of the recombinant DNA clones revealed homology to the DNA region coding for a reverse transcriptase-like protein in retroviruses and retrotransposons. Homologous tissue-specific transcripts of a size of 1.2 x 10(3) base-pairs were found in testes. Transcript in-situ hybridization shows that at least parts of these transcripts are synthesized in these Y chromosomal lampbrush loops, which were originally used for microdissection. Also the cytoplasm of primary spermatocytes contains homologous RNA species. These observations are discussed in the context of lampbrush loop function and evolution.

Chromosomal distribution of the RTVL-H family of human endogenous retrovirus-like sequences

We have analyzed the chromosomal distribution of a large family of human endogenous retrovirus-like sequences termed RTVL-H. In situ hybridizations suggest that these sequences are found on all human chromosomes. These results also indicate that clusters or concentrations of RTVL-H elements may exist on chromosomes 1p and 7q. Southern blotting experiments using somatic cell hybrids containing either the human chromosome 3 or the X chromosome confirm the presence of multiple dispersed RTVL-H sequences on these two chromosomes. These experiments also demonstrate that distinct RTVL-H banding patterns can be detected for each chromosome. Thus, RTVL-H probes may be useful in genome mapping studies.

Partial revertants of the transposable element-associated suppressible allele white-apricot in Drosophila melanogaster: structures and responsiveness to genetic modifiers

The eye color phenotype of white-apricot (wa), a mutant allele of the white locus caused by the insertion of the transposable element copia into a small intron, is suppressed by the extragenic suppressor suppressor-of-white-apricot (su(wa] and enhanced by the extragenic enhancers suppressor-of-forked su(f] and Enhancer-of-white-apricot (E(wa]. Derivatives of wa have been analyzed molecularly and genetically in order to correlate the structure of these derivatives with their response to modifiers. Derivatives in which the copia element is replaced precisely by a solo long terminal repeat (sLTR) were generated in vitro and returned to the germline by P-element mediated transformation; flies carrying this allele within a P transposon show a nearly wild-type phenotype and no response to either su(f) or su(wa). In addition, eleven partial phenotypic revertants of wa were analyzed. Of these, one appears to be a duplication of a large region which includes wa, three are new alleles of su(wa), two are sLTR derivatives whose properties confirm results obtained using transformation, and five are secondary insertions into the copia element within wa. One of these, waR84h, differs from wa by the insertion of the most 3' 83 nucleotides of the I factor. The five insertion derivatives show a variety of phenotypes and modes of interaction with su[f) and su(wa). The eye pigmentation of waR84h is affected by su(f) and E(wa), but not su(wa). These results demonstrate that copia (as opposed to the interruption of white sequences) is essential for the wa phenotype and its response to genetic modifiers, and that there are multiple mechanisms for the alteration of the wa phenotype by modifiers.

Processing of TY1 proteins and formation of Ty1 virus-like particles in Saccharomyces cerevisiae

We have analysed functional properties of putative proteins encoded by the yeast transposable element, Ty1, by overexpression of TY genes. High-level expression was achieved by appropriate fusion of a Ty sequence, TY9C, to the yeast ADH1 promoter and transformation of yeast cells with this construction. As shown recently by others (Garfinkel et al. 1985; Mellor et al. 1985c) TY overexpression leads to an increase in particle-bound reverse transcriptase activity and to an intracellular accumulation of virus-like particles (Ty-VLPs). We have used a number of deletions in the second open reading frame (TYB) to identify functional domains required for processing and assembly of Ty proteins. Deletions in the TYB region with homology to acid proteases result in overproduction of an unprocessed form of the TYA protein (pro-TYA) which represents the major protein of Ty-VLPs. One particular mutant construction, TY9C-delta 36, led to the accumulation of a particle-bound, 160 kDa protein which cross-reacted with a mouse antiserum raised against purified pro-TYA protein. This supports the hypothesis that TYB is expressed as a TYA/TYB fusion protein which is processed by a TYB-encoded protease activity. Ty-VLPs are formed in the absence of protein processing and even when the TYB gene is not expressed. Thus, we assume that the assembly of Ty particles occurs prior to processing of Ty proteins.

Production of virus-like particles by the transposable genetic element, copia, of Drosophila melanogaster

Using the calcium-phosphate procedure the plasmid cDm2055, which carries the copia transposable element of Drosophila melanogaster, was co-transfected with the Neor-carrying plasmid pSVCneo-1 into the D. hydei cell line KUN-DH-33 which was free of copia. The Neor transfectants stably carried both plasmids as tandem oligomers integrated in the chromosome and virus-like particles (VLP) were produced specifically in the transfectants that received the copia plasmid. The particles were quite similar in various aspects such as size, morphology, density, RNA content and molecular weight of the major protein component, to retrovirus-like particles (RVLP) that spontaneously appear in cultured cells of D. melanogaster: the reverse transcriptase activity however seemed to be low compared to that of the D. melanogaster RVLP. This finding demonstrates that the retrovirus-like particles (RVLP) in Drosophila cultured cells are produced by the transposable genetic element copia resident in the host chromosomes.

Transposable elements controlling I-R hybrid dysgenesis in D. melanogaster are similar to mammalian LINEs

I-R hybrid dysgenesis in D. melanogaster is controlled by transposable elements known as I factors. We have determined the base sequences of one complete I factor and the ends of six others. The ends of these elements are highly conserved and are flanked by target site duplications varying in length from 10-14 bp. There are no terminal repeats, and the 3' end of one strand is A-rich, having 4-7 tandem repeats of the sequence TAA. This sequence organization is similar to that of mammalian LINEs, or L1 elements. The complete I factor sequence contains two long open reading frames, ORF1 and ORF2, of 1278 and 3258 bp. ORF1 encodes a possible nucleic acid-binding protein, and part of the amino acid sequence of ORF2 is similar to that of viral reverse transcriptases and polypeptides encoded by L1 elements. These results suggest that I factors transpose by reverse transcription of a full-length RNA.

Conserved and non-conserved features among the yeast Ty elements

We have isolated and characterized a Ty element from a yeast cosmid library which exhibits several unusual features: it is flanked by non-homologous delta elements and directly associated with a singular delta element. A tRNA(Glu3) gene and tRNA(Cys) gene are found in conjunction with this element, located in opposite orientation on either end of it. The sequence information now available for several Ty elements has been used in a detailed comparative analysis to determine conserved features among the Ty elements, preferably between class I elements and a class II element. Highly conserved sequence motifs appear to be located at the borders of particular segments that correspond to the putative protein domains of the Tys. Furthermore, we include a comparison of the best-conserved amino acid homologies for these putative proteins of Ty elements, transposable elements from other organisms and several retroviral proviruses to confirm their close structural resemblance.