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

J. Mol. Biol. (1988) 203, 689-697

Retrotransposon-like Sequences are Expressed in Y Chromosomal Lampbrush Loops of Drosophila hydei Peter Huijsert, Christiane KirchhoQ Dirk-Henner Lankenau and Wolfgang Hennig Department

of Molecular and Developmental Genetics, Katholieke Toernooiveld, 6525 ED Nijmegen, The Netherlands (Received

7 March

Universiteit

1988)

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 lo3 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.

1. Introduction The Y chromosomal lampbrush loops that are formed in primary spermatocytes of Drosophila hydei are composed of middle repetitive DNA sequences (Vogt & Hennig, 1983; Hennig et al., 1983; and see Hennig et aZ., 1987). Members of some of these families of repeated DNA sequences not only occur within the Y chromosome but are widespread throughout the entire genome (Vogt & Hennig, 1983, 19863; Hennig et al., 1983). This type of DNA was classified as “Y-associated” DNA (Vogt & Hennig, 1983). Y-associated DNA sequences are transcribed as parts of the Y chromosomal fertility genes and occur interspersed between another class of repeated Y chromosomal DNA sequences, the “Y-specific” type, which in its genomic location is restricted to the Y chromosome (see Vogt & Hennig, 1986a,b). Our studies of these different classes of Y chromosomal DNA sequences were extended to different strains and to other

7 Present address: Max-Planck-Institut fiir Ziichtungsforschung, Egelspfad 3, D-5000 Kiiln 30, F.R.G. f. Present address: Institut fiir Hormon- und Fortpflanzungsforschung, Grandweg 64, 2000 Hamburg 54. F.R.G. 0022-2836/88/ 190689-09 $03.00/O

Drosophila species. The results suggested that the widespread genomic distribution of the members of DNA families of the Y-associated class is the result of transpositions (Hennig et al., 1983; Vogt & Hennig, 1986b; Vogt et al., 1986; Huijser & Hennig, 1987; Hennig et al., 1987; see also Hareven et al., 1986). The Y-specific class may be derived from originally belonged to the sequences that Y-associated class, or from any other genomic sequence that has become inserted into the Y chromosome (Huijser & Hennig, 1987; see also Hareven et al., 1986). However, until now our data did not permit unequivocal demonstration of the transposable nature of sequence elements of the Y-associated class. In this paper we present the detailed sequence analysis of members of one of the Y-associated DNA sequence families of D. hydei. We discovered a high degree of similarity between a putative protein product encoded by this DNA sequence family and the reverse transcriptase of retroviruses. Also other sequence elements such as the “finger” domain and a protease-like region typically present in retroviruses and retrotransposons were found. These and other properties show that these sequences form a class of mobile genetic elements, in RNA-mediated principle capable of transposition.

689

0 1988 Academic Press Limited

690

I? Huvjser et al.

2. Materials and Methods (a) Drosophila The laboratory

wild-type strain collection.

(d) In-situ hybridization Transcript in-situ hybridization on testes syuashes was carried out as described by Vogt et al. (1982). As a probe, 3H-labelled cRNA was synthesized. Cytoplasmic in-situ hybridization was carried out. as described (Huijser &. Hennig, 1987).

strains

of D. hydei

was from

our

(b) Isolation of nucleic acids E’ chromosomal DNA fragments obtained by microdissection of the Y chromosomal lampbrush Ioop threads or pseudonucleolus were cloned into the EcoRI site of the i vector 641 by the method of Scalenghe et al. (1981). Previously we described the 2 clones dhMiF2 and dhMiF8 (Hennig et al., 1983), both of which are derived from the lampbrush loop threads. Both clones have internal EcoRI sites (see Fig. 1). The generated EcoRI fragments were subcloned in pBR328 and designated as dhMiF2A (3.1 kbt), dhMiF2B (1.8 kb), dhMiF8A (2.4 kb) and dhMiF8B (2.0 kb). A 3rd EeoRI fragment of 0.6 kb of the original 1641 clone dhMiF8 is not further considered. since its nucleotide sequence suggests that it is present as the result of an additional ligation during the microcloning procedure. Recombinant DNA was isolated according to standard protocols (Maniatis et al., 1982). RNA was isolated according to the method of Chirgwin et al. (1979) after homogenization of the tissue in 7*5 w guanidinium . HCI. (c) Northern blotting Denatured RNA (20 pg/slot) was separated by length. blotted on to nitrocellulose. and hybridized as described by Thomas (1980).

t Abbreviations used: kb, 103 bases or base-pairs; bp, base-pairs; ORF. open reading frame: LTR, long terminal direct repeat.

I

*B

E

SS

1

I

(0)

(b)

E

R Pv

I

(

The homology matrix program of Pustell & Kafatos (1982) was used to compare the sequences of dhMiF2. dhMiF8 and the copia-like element 17.6.

3. Results (a) General ch.aracterization

Hy microdissection of t’he Y chromosomal lampbrush loop threads from primary spermatocytes of II. hydei and subsequent cloning a,t a microscale. several recombinant, DNA clones were obtained, which were described in a prior publication (Hennig et al., 1983). The inserts of all these clones hybridize in irk-xitu hybridization rxperiments with neuroblast metaphase chromosomes to the dist,a,l part’ of the long arm of t.h(> J7 chromosome. i.e. to i he @sition of thr \ chromosome expected from t,he position of t.hr

BA

1 H

Ps Xh

*SE

b

tiiF2

Selected as well as random restriction fragments were subcloned in the appropriate Ml3 vectors constructed by Messing & Vieira (1982) and Yanisch-Perron et al. (1985), and single-stranded templates were isolated according to the protocols published hy Amersham (1984). The nucleotide sequences were determined by the dideox) chain termination method of Sanger et al. (1977).

*

b

Pv A K

H KAXbE l(

\I( I

----I AXbA

XbB

YI

v

E I

R

PsA AXhC

1

I r

E uw UY

VHII sbp

I

Figure 1. Restriction maps of (a) dhMiF8 and (b) dhMiF2. Thick filled arrows mark tlw putative LT11 regions in both csloned fragments. Vertical arrowheads mark the position of the conserved poly(A) signal. Horizontal arrowheads mark the 19 bp imperfect repeats. Thick vertical lines indicate the positions of the tandemlp repeat,ed sequences (24 hp. 35 bp). Stop codons in the 6 reading frames of dhMiF2 are indicated by vertical lines below t,he restri&ion map (c). The putative reverse transcriptase coding region is indicated by dots (RT); the DNA-binding finger motif is marked (F). The protease-like region (P) is also marked. For further details, see the t,rxt. Restriction enzymes are: :I. ilccI: B. Hn:mHI: C. PEaI: E. EcoRI; H, HindIIT: K, I(@; Ps, P&I; Pv, PuuIT; Sp. &&I: Qs. S&I: Xh. XbtzT: Xh. X&)1.

Retrotransposon-like

Sequences in Lampbrush Loops

microdissected loops. On polytene chromosomes, however, the different cloned DNA fragments hybridize to different X chromosomal and autosomal sites. The only clones that share a number of sites in polytene chromosomes are dhMiF2 and dhMiF8. From cross-hybridization experiments we recognized that the inserts belong to the same family of Y-associated DNA sequences. Our obser-

691

than 80% (Figs 1 to 3). The interruption in the regions of sequence similarity suggests that in the course of the evolution either an insertion in the dhMiF8 sequence or a deletion in the dhMiF2 sequence has occurred, or that both events have taken place. The most likely event is, however, an insertion of about 1 kb into the dhMiF8 fragment,

of repeated DNA sequences of Drosophila melanogaster led to a detailed analysis of both cloned DNA fragments. The comparison by a matrix analysis of the DNA sequences of dhMiF2 and dhMiF8 displayed two separate regions in each of both clones with an

as a comparison with homologous DNA fragments shows (Lankenau et isolated from D. melanogaster al., unpublished results). This possibility is also supported by the presence of an (imperfect) duplication of 19 nucleotides close to the borders of the putative insertion of dhMiF8 (positions 876 to 895 and 1984 to 2002) (Fig. 3). Since this duplication is not found in dhMiF2, it is

almost

in

vation that a high exists with a family

entirely

degree

colinear

of sequence

sequence

similarity

homology

1 61 121 181 241 301 361 421 481 541 601 661 721 781

gaettccte*tttcstcetsctaaaaaatattgcettg~~t~~g~~g~teett~~~~~~t~ett~ gtcctttgcattttgtcgctttaagaagttccgttatagc=g~~~g~gtt~te~t~ tggttcggtgcgaggcagagaaacgaaaaggcggtgagttcaactttgaatcgag~agcc tgcgcatcgaeactcgctgcgctattegagccatggaaaat~~te~~ge~g~t~~~e~~t cagaagtgggatcagcaccacctcctgcattctcega)gg~t~~gtgg~g=et~etttt~~ aaetgcegaaccagastctgctggagcttgcaagaaccatgcaag~gt~~~~~~~gg~g~ aacagcagactcgtggcgtcactttgccaaaatttttctctgacactgcaggtgcagatg cctcaaagtggtccatcacagctgacattattaacggaacag=~~=t~g~~gg~*~~~ gactcatcatggcactaagcgactgatggtgggcagtacctcacagtggctaactcaaac tgcatatcccggcattaagtggattgaatttaaggaattattcct~~g~~g~tac~~a~~ tagtgaaactgcgtcggcaatgtttctaaatctcacteaacsgtcgcccaactaacggcg aatgtcttgcggtttatgctagccgtttggtcaccacttt~~~tt~~~~ttgg~~~~~~~ aaaacatcgaagaaatctcgctgtttcaattgtgctcgcg~ettt~g=~~~tgtg~~t~~ ccgcttggaacggettgtgttcacgtcaaaagttcaaacteg~~gtge~ttg=~~t~~~~

841

~.‘AlaHisSerProLevLysLysHisArgAsnAlaAspArgAspAspAspSerGlyP~‘o aCteggCgCettCtCCttte8888BBCstCgtBBtgCtgB~~

of more

all

probability

part

of

the

inserted

2161

ThrValCluLeuGlyLeuCluArgLeuLysIleValLeuAspIleLe~ThrAspAla~~~~ cBc*gtcgaattaggattggeaagBttgaaaatagtactagacattctcacagatgctag

2281

GluValArgAleClyCluIleArgProAsnSerArgLysIleValAlaLeuThrSerLeu tgaggtaagagcaggggeaattcgaccaaactcgcgaaag*ttgttg~gtt~~~~t=gtt

2341

HisAlaThrAspClyPheTrpCysThrAlaValTyrTrpSerCysIleLeuPheSerCln gCaCgCCBCagacggtttttggtgtacggc8gtttattggtcttgcatcctattttcgca

2401

IleArgThrArgIleLeuSerSerHisLysAlaValValPheProAsnIleArgLysGly BsttcgtaCccggattctCtc8agtcataeaBccgttgtatt=~~t~e=~t~~gg~eggg LysPheCysValAspTyr

++‘ValGlyAsnAspSerC1nGluAsnCysAspTyrSerT~~~

2461 aaagttttgcgtggactactgegttggaaacgattcgcaagaaasttgtgactattctt~ 2521

LysAcgThrSerSeffilyAsnIleArgProAlaIleProTyrArgIleHisThrAspAla ceeecgaaccagttctggteatattcgacccgcaataccctetegeet~~e=e~~ge=~~

2581

SerSerIleClyPheClyAlaMetLeuLe~H~ArgIleAsnAsnLysLeuHisV~lV~l tagctcsatcggatttggtgccatgcttttacatcgastte~=~~t~~g~tt~~=gttgt

2701

ThrLeuAlaValPheAlaSerValLysHisPheArgHisTyrLeuLeu~lyAr~luPhe ~~CtCtggCsgtetttgC*tCCgttaaaCatttCCgtCaCte~ttg~tt~g=eg~ge~tt

2761

ValVelTyrThrAspCysAsnSerLeuLysAlaSerArgThrLysIleAspLeuThrPro tgtggtttatecagattgtaactcgttaaaggcatcgcgaaccaaaatsgacttaactcc

PheValProGluProThrGlyVelMetMeLGlnLe~GlyGluThrTy~ProIleS~~~Phe gctttgtsccag*gcCBactggagtgatgatgcagctcggtg~g*~tt~=~~~~t~t~tt

2821

ArgValHisArgTrpTrpSerTyrLeuClnSerPheLysPheAspIleGlnTyrArgGlu tagsgttcatcgctggtggtcttattttacaatctttcaagtttgaca~ac~gtacagaga

1141

2881

GlyLysArgMetAlaHisValAspPheLeuSerArgAsnProPheThrGlnLysGlnLeu aggaaaecgtatggcacatgtcgacttcttatctagaeacccatccactcaga~~caatt

1201

AspSerGlyAlaGluCysSerLe”IleLysCluLysTyrValValSero** l * ArgCluIleCysSerLysLeuAlaGlyLys ttgattctggcgcagagtgttcactcattaaagagaaatatgtegt~eg~teg~tggt~~

2941

IleLeuCluLysIleProC1uLysArgIleAsnIleAlaGluIleSerAsnAsnTrpLeu aattttggaaaaeattcccgageagcgaatasacategcatagcagaaatctcaaataattggtt

1261

ArgPheAsnThrValIleMetLeuLysAspIleGlyAspAlaGlyIleTyrSerThrLeu gcgttttaacacagttatcatgctaaaagatataggcgatg~tgg~etetet~g=e~~tt

3001

PheAlaCluClnGlnLeuAspProAspIleAleAlaLeuValSerLysLeuAsnSerAsn gtttgCCgaaCBgCBettggaCCCggatattgCtgCCCttgt=~gt~~e=tg~~tt~~~~

1321

GlnIleLeuSerClnValThrIleAsnGluAsnAlaLeuClnIleLeuPheHisValVa~ acaaatattaagccaagtcaccataaatgagaacgcactt~~~*t~ttgtt=~~tgt~gt

3061

AspLeuLeuAspAspGluLeuLysHisMetIle*** tgacctactagacgatgaactaaeacatsrgatgatttgagaaaagggatgttgtttcgtaaa

1381

GlnAspAsnHisLeuLysSerAspIleValIleGlyArgCluIleLeuLysGlnGlyPhe gcBBgataafcatttgesaagcgaCSttgttattggccgCg~~~tt=tt~=~~~=gg~tt

1441

AsnVa1lleIleSerClnAsnLysPheAsnValValLysAlaLysAlaVelAsnlleCys taatgtcattatatcacaaaacaagtttaecgtagtaaaag~~~~~g~~gtee~t~t~tg

1501

SerIleAspArgCluLeuValAspLeuThrAsnLeuAspThrGluLeuArgGluAl~Asp ctcaatsgatagggaactagttgaccttecaaatctcgat~~~g~~ttg=g~g~~g~tg~

1561

LysAspLyeLeuIleLysMetLeuCiuAsnTyrSerThrAlaLeuV~lLysGlyAl~~~~~ taaagacaaactaataeeeatgcttgagaattactccactg~e=ttgt~~=~gg~g~~~~

1621

SerThrArgValThrThrGlyCluMetLysIleArgLeuIleAspProThrLysTl~rVal atctacacgagtcacgacaggggaaetgaaaattcgcttgattgetccaeccaaaactgt

1681

GlnArgArg~~~TyrArgLeuSerProGluCl”ArgAspIleValAr~l”LysLc”Sc~ acaacgccggccctatagacttagccctgaagaaagagscattgtccgagaaaagtt~ag

1741

GluLeuLeuLysCysAsnIleIleArgProSerCysSerProPheAlaSerPro~etLe~~ tgagttacttaaatgcaacatcattcgacccagctgctcc~~ttttg==~g~~~~~t~~t

Le”ValLysLysLysAsnClyThrAspArgLeuCysValAspTyrAr~luLeuAsnSe~ 1801 gcttgtcaagaagaaaaatggtscagaccgactttgtgtcgactacagagagttaa~ctc 1861

AsnThrIleAlaAspAr~y~P~~LeuProLeuIleSerAspGl”IleAlaArgLeu~~~~ tseCacaatcgcsgacaggtaccctttgccgctcartagcge==~g~t~g=~~gg~~~cg

1921

GlyAlaAsnTyrPheThrCysLeuAspMetAlaSerGlyPheHisGlnlleProIle~is tggCgCBBBttaCttcaCgtgt~ttgetatggcgagrggtggttt=~~==ee~tt~=t~~~~~

ProAspSerIleGluTyrThrAlaPheValThrProAspClyClnTyrCluPheLeuTh~ 1981 CcCtgattCCatsgaataCactgcatttgtgaccccggac~ 2041

MetProPheGlyLeuLysAsnAlaProSerVslPheGlnAr~AlaValMetA~g.~l~L~~l cstgccgttcggactcaaasatgcgcCatccgttttCCa8CgggCegtC~tgCgt~CC~~

2101

GlyAspLeuAlaTyrSerTyrValIleVslTyrMetAsPAspII~MetIleVul~~~~'~~~ gggtgaCttggCCtattCstacgtgattgtttscatggatgatatcatg~tcgt~~c~~~~

Figure 2. Complete putative LTR re@on. polvadenylation slgnal regions are given with with L-shaped arrows. marking the ORFs.

DNA

3121 3181 3241 3301 3361 3421 3481 3541 3601 3661 3721 3781 384 1 3901 3961 4021 4081 4141 4201 4261 4321 4381 4441 4501 4561 4621 4681 4741 4801

atccaaaaaaatggtaagecacgatgcttgccagttgtgc=~=g~e~ttt=~getgg~~~ gtagttsacca~tacatgegtcaatcatgcaccttggttggcaaaaaacacttgacaaa gtttecceettttectggtttgagaaaatgaaceaatatgt~~gtseetet~ttg~~~st tgtataacctgcaggacagcaaagtcacattccggcaaagttcaagcagaattacatcct ataccaaaggttegcatcccttggcscaccgtccatatggacattactggcaaattaagt ggtaaaagtgatttaaaggaatatetcattgttcaaatrg*~~g~tt=~~e~eett=gt~ CaCCtgtacCaCaCtgtatCagggccataaagtcttcaacttccttgtttggtgcaccaa

nucleotide sequence of dhMiF2. The thick arrow below the sequence indicates the position of a Thin tandemly arranged arrows above the sequence mark tandem repeat clusters. A conserved is boxed. ORFs including the putative reverse transcriptase and the DNA-binding finger coding their amino acid residues. Termini of the reverse transcriptase amino acid sequence are marked The finger motif is marked by a dotted line above the amino acid sequence. ***. Stop codons

692

P. Iluijser

1 61 121 181 241 301 361 421 481

gaattccgactatgccattcggactaaaaatgcgccatctgattttcaactggcagtcat gtgagcataggcgatataataatcgagttacccacagtcgagttaggattggaaaggtta aaaacagtactagacattcttacagaggctgtattcattataaatgtcagcaaatgtaac tttttBBB*aCagtCCttCBgtattttggt~t~tgaggtae~~~ ccgaaaaattgctgcgtcctttecacaccgttacaaatagtttttggtgt~cgacagtttat tggtcttgcstcctaagttttgttaagaatacaattcgca~g~~~~ttttg*=g~tt=tt aCaaaCgeacCagttctggtgatattcgacccgcaecacc~t*t*g~~tt~=~~~~tg~~ gCCBgCtCBstCggatttggtgCCCstBCttttatatCBgtg~~=tt=~=gt tgtgaaatactteagcaaacaacttcgcce4gccgagtctaggtactact~ttatg~gct

661 721 781 841 901 961 1021 1081 1141 1201 1261 1321 1381 1441 1501 1551 1621 lh81 1741 1801 1861 1921 2041 1981

cctagagttcatcgttagrggtcttctttacaatccttcacttttgatatacaagttaac acBtaggBatgcccatgaatactttgtataagaatcttgc~~~~t~~t~~t*ttt*~~~~ gttactccaaccatattccgtgacgcatgagttcagcgcc~=g~~=*~~*tggt==g~~t cagtaaactgatacttgactgagtcggcttgccggctgtagttttgtagcttatBcgccc aacttttaggtttagcattattccaattttgtaataagtacagtcgtacgagcgtcaacg ggtacacatatcggtgtttaaagaataaacccgcacaactggc4cacaccgccgtaetca BattgttatatttCtgBtcatcgctgagccaceaggccagggtg~~ gtcBaaccctccaacatastcatcataagaagattataatt~~~~g~t=*~=gttgg~~~ tgccaccgacgcctagttcatcgctggaccagceacttactt~~tttgt~t~~~tg~~g~ tttgggCBgtta*CBattgCgccgttaacaacgcttcgat~tg=~~~~gtggttt~~~~t gacacctgcgcgcgtgatctacaagctggeggcacggcacgetgccacactcagcccsgc CatttgaaaCctattacagttgtggetgetgatggcattgtggtggtgaatgaaagcccagcg cgaatcagcaccgatgacagccccgagatcgcagtcaacgt~~~ttt~g~~=g~~~~g=~ ttcataaacggaagtaagtacgtaaccggcggcgtgagatcctaaacagaacccctgggcat agCagCCtCaCCsttgCtCaacatcattagacacg0gtcagttcta0gcatgccgttcct acataggatgaacaaccacaaccttcacatcatcaacaaacttcaagaagacgtagtatc agccggatcgcctaseacttggtccattgctggagttgaa~t=sg~~t~gtttt~~g~gg ttccatcctcctctgccgagtctgaggagaaagagagcct=~~g~*~~~~~~~g*~g~t~ gtcaacBgctataagatgaacgaglcacggtcacagttctgaggggggagtagttaacaa tgCCCCCtgCBCCCagttgCttagactaaaacaacagcBct~gt~~~~g=tg~t=ttg=~ gegctcngttcatcgcccgastcatcatacggtcagcgtcggtaacctgatactttgcatgta tcgcaggagcagcgttatcagttatgctggctagccgaccgaattcttgtagctaaagaa teactgtagtttgtagctttt~attccattttatttttactttaagt~cagtcctgtac~ acagttctaccaataaacacatattgctgagccacaaggc~t~t~tt~g 2101 ttatatccc8tcataatctgcctaicaatttgatgtaact~t~ 2161 cccctaatgtggaacctgaacgaacagaatccataatccaaattgatctacattgtactc

et al.

2221 2281 2341 2401 2461 2521 2581 2641 2701 2761 2821 2881 2941 3001 3061 ;:i: 3241 :g: 3421 3481 3541 ‘,2:: ::;: 3841 ;;“6: 4021 4081 4141 4201 4261 4321 4381

Figure 3. Complete nucleotide sequence of dhMiF8. The thick arrow below the sequence indicates the position of’ t,he putative LTR copy. Thin tandemly arranged arrows above the sequence mark tandem repeat clusters. Thin arrows below the sequence mark the imperfect duplication discussed in the text. A conserved polyadenylation signal is boxed. Uncertainties in the sequence (1, 2. 3, 4) are probably C, T. A and G, respectively.

sequence. However, we cannot define precisely the exact site of an insertion. Another remarkable sequence element occurs in both clones (in positions 4164 to 4319 in dhMiF2 and positions 2924 to 2971 in dhMiF8). These regions contain short tandemly repeated sequences of a basic unit length of 24 nucleotides (Figs 1 to 3). A functional relevance of the 24 bp repeat cluster within the sequence contexts is suggested by its revolutionary conservation between D. melanogaster and D. hydei (Lankenau et al.; unpublished results). Both sequenced DNA fragments contain several open reading frames (ORFs). The longest of the various ORFs in dhMiF2 (see Fig. 1 (c)) is 1248 nucleotides in the subfragment dhMiF2A (positions 1232 to 2479) (Figs 1 and 2). This sequence segment is only in part present in dhF8 (cf. Fig. I).

(b) Comparison of dhMiF2 with DNA sequences of other organisms We screened the EMBL genebank with the sequence of dhMiF2 for similarities with other DNA sequences and found a prominent degree of to DNA of retroviruses or sequence similarity retrovirus-like sequences like TVMV (tobacco vein mottling virus), TMV (tobacco mosaic virus), CaMV (cauliflower mosaic virus), HPV (human papilloma virus), the transposable element Ty of yeast and the copia-like element 17.6 of Drosophila. All these genetic elements are considered to be related to retroviruses because of their potential to undergo

transposit.ions with the aid of a reverse t,ranscrip tase that is encoded wit.hin their genome. These sequence properties indicate a retrotransposon character of the dhMiF2 family of repeated DNA sequences. Because of its sequence relationship to the putative reverse transcriptase coding domain in copia-like elements and its experimental origin we called this DNA sequence family the micropia family. Of the sequences with homology to dhMiF2, the copia-like element 17.6 from D. melanogaster is the evolutionary most closely related sequence. Therrfore we compared the 17.6 DNA sequence in detail with dhMiF2. The strongest homology is located in the region of the longest ORF of dhMiF2 (positions 1232 to 2479). In 17.6 it represents the DNA section coding for a reverse transcriptase-like protein (Saigo et al., 1984). If the amino acid sequences of the putative reverse transcriptase coding regions of dhMiF2 and 17.6 are compared, approximately half of the amino acid residues are found to be identical or chemically equivalent (Fig. 4). This also holds t,rue for the relationship with other (putative) retroviral trans~ criptases (Toh et aZ., 1983, 1985; Saigo et al., 1984: Boeke et al., 1985; Mount & Rubin, 1985; Fawcett ef al., 1986; Loeb et al., 1986; Fanning & Singer, 1987: Kimmel et al.. 1987; Schwarz-Sommer et a,Z.. 1987). The homologous regions of the dhMiF2 reverse transcriptase involve those regions of the enzyme identified as highly conserved throughout evolution (Toh et al., 1983, 1985; Patarca & Haseltine, 1984). Besides the homology of the reverse transcriptase-like protein coding region, another sequence

Retrotransposon-like

Sequences in Lampbrush Loops

--

693

Figure 4. Comparison of the amino acid residues of the putative reverse transcriptase encoded by the copia-like element 17.6 and by dhMiF2. L-shaped arrowheads mark the reverse transcriptase (cf. Saigo et a,l., 1984). Identical and rhemically similar amino acid residues (cf. Fawcett et al., 1986) in both sequencesare boxed. Gaps (indicated by dots) are introduced bo maximize homology. *. a stop codon.

element typical of a variety of retrotransposons has been discovered in dhMiF2. Upstream from the reverse transcriptase-like protein-coding region, a 405 bp ORF (positions 846 to 1250) (Fig. 2) is located. Part of its derived amino acid sequence displays homology to the consensus sequence C-X2-C-X3-GH-X4-C (Fig. 5). This sequence resembles the finger domain of putative DNAbinding proteins (Miller et al., 1985). In the retrovirus genome this motif is encoded by the much longer gag region upstream from the pol gene for the reverse transcriptase complex. However, this short consensus is cleaved off as part of a small peptide from the Gag polypeptide in retroviruses and may therefore be coded as a separate entity (Covey, 1986). A finger motif has also been discovered in non-viral retrotransposons lacking long t,erminal direct repeats (LTRs) where they occur upstream and downstream from the reverse transcriptase (see, for example, Fawcett et al., 1986; cf. Schwarz-Sommer et al., 1987). Additional CC-HC- finger motifs downstream from the reverse transcriptase-like protein-coding region were not found in dhMiF2.

in dhMiF2 and positions 3450 to 3670 in dhMiF8) (Fig. 6) with a degree of homology higher than 70%. These regions may represent a left and a right LTR of micropia elements. LTRs are often characterized by several sequence elements such as promoters of transcription and polyadenylation signals. We cannot recognize such sequence elements within the 200 nucleotides that to represent opposite LTRs. are considered However, it cannot be excluded that t#he actual length of the LTRs is larger. Polyadenylation signals found upstream from the 200 nucleotide segment in dhMiF8 in positions 3215 to 3220 and in positions 4492 bo 4497 of dhMiF2 (Figs 2 and 3) support this idea. These polyadenylation signals have the same distance to a 24 bp repeat that is present in both clones (Fig. 1). The 24 bp repeat regions occur in position 4164 to 4319 in dhMiF2 and in position 2924 to 2971 in dhMiF8 (Figs 1 to 3) and they are also found in D. melanogaster (Lankenau et al., unpublished results). From their position upstream from the putative 3’ LTR of dhMiF8 (see below and Fig. I), one would expect that the 3’ LTR should be present in the 3’ region of dhMiF2. This is not the case. but, as is evident from Figures 1 and 2, a second repeat sequence related to the 24 bp repeat cluster is found

(c) Other cowerved structural sequence elements

We expected other structural components in dhMiF2 typical of transposable elements. Since the copia-like element 17.6 is characterized by the presence of LTRs we searched for such sequence elements in dhMiF2 and dhMiF8. Within dhMiF2 a LTR-like repeat is not present. However, a comparison with 17.6 shows that dhMiF2 may be too short to accommodate both LTRs. The matrix comparison with 17.6 shows that dhMiF2 may be extends further toward a 3’ direction if the regions of homology with dhMiF2 are matched (Fig. 1). Comparing the 5’ end of dhMiF2 with the 3’ end of dhMiF8 reveals regions of homology extending over more than 200 nucleotides (positions 1 to 218

RSV

6 L s V T*?6 S P 6’ji V Q A 0% P K ERCOLCN6NSHNAKECRK

I-factor

LRCKKCLRF6HPTPICKS

Fr-elerent V U C T N C R E V 6 H T R S V C T L dhlliF2

IV~HFCSKPStjKT6ECRB

Figure 5. Comparison of the amino acid sequences of the DSA-binding finger domains from various organisms. RSV: ROW sarcoma virus (Schwartz et al.. 1983); T-factor (Fawcett et al., 1986): Fw-element (DiNocera R: Casari, 1987); dhMiF2: this work. Cysteine and histicline residues are boxed. Note the conserved glycine residue in front of the histidine residue.

P. Huijser et al.

694

dhlif2 dhMf6

hc.~,~,~~l,~,~,~,~,~~,~~~~~~~ tutt ut u tttutI 111t ttuu I I I “~f@MM~~~hlI66. .,~~~,,~,~,,~~,r~~~~6,~~,~,6~~~~,”,6,~~-~~~,,

Figure 6. Comparison of the nucleotide sequences of t,he putative LTR region of dhMiB2 and dhMiP8. Gaps (indicated by dots) are introduced to increase positional identity. Numbering is according to the sequences shown in Figs 2 and 3.

in dhMiF2. The insertion of this second repeat cluster may have caused a shift of the 3’ LTR. which is consequently expected in a position downstream from the 3’ EcoRI site of dhMiF2 and therefore absent from this cloned insert. However, other types of transposable genetic elements do not carry LTRs (Finnegan, 1985). Their nature as retrotransposons is deduced from their sequence information, since they also code for reverse transcriptase-like proteins. Such genetic elements like the mammalian LINES (Fanning & Singer, 1987), the Drosophila I- and F-factors (Fawcett et aE., 1986; DiNocera & Casari: 1987), the Ingi-3 element of Trypanosoma (Kimmel et al., 1987) and the Cin-4 element of maize (SchwarzSommer et al.? 1987) lack terminal repeats but. in signals, addition to putative polyadenylation possess an oligo(dA) track or at least dA-rich regions at their 3’ ends. In dhMiF2 and dhMiF8 we cannot find similar A$T-rich regions downstream from the polyadenylation signals discussed above. An unequivocal definition of the termini of the dhMiF2 sequence family members must therefore await additional sequence information in the regions flanking our cloned DNA segments.

cones and pseudonucleolus (Hennig, 1985). Transcripts cross-hybridizing to dhMiF2A were also found in t’he cytoplasm of primary spermatocytes (Fig. 7(c) and (d)). We also hybridized the subcloned EcoRl fragments dhMiF2A and dhMiF2B separately to blots of RNA from testis, female and male carcasses (Fig. 8). With both probes we found a prominent reaction with testis RNA that, however, differs for both subclones. Fragment dhMiF2A hybridizes t,o an RNA species of 1.2 kb (Fig. 8(a)). Remarkably, the size of 1.2 kb agrees closely to the putative reverse transcriptase coding region in this restriction fragment, which has a 1248 nucleotide ORF (see Fig. 1). The fragment dhMiF2B hybridizes to RNA species of heterogeneous sizes (Fig. 8(b)). It carries a (CA/GT),-rich region at, its 3’ end. Such repeats have been demonstrated to be highly expressed in

(d) Testis-spec@c transcription of the dhMiF2 sequence family Since the dhMiF2 family DNA sequence was recovered by microdissection of a t,ranscriptionally active chromosome region in primary spermatocytes, it is expected that it reacts with testis RNA. This can be verified by transcript in-situ hybridization. We used the subloned dhMiF2A fragment as a probe. Homologous transcripts were identified in the lampbrush loop threads and pseudonucleolus (Fig. 7(a) and (b)). This agrees with the origin of the dhMiF2 clone (we dissected the threads) and with the fact that the dhMiF2-homologous sequences were discovered in the 2641 clone banks established by microcloning the pseudonucleolus (Huijser & Hennig, unpublished results). Both lampbrush loop pairs thus contain DNA sequences homologous to the micropia sequence family. A correlation to this situation may be the complex genetic interrelationship between the fertility gene loci A to C corresponding to the lampbrush loop pair threads.

Figure 7. Transcript in-situ hybridization of dhMiF2A to primary spermatocytes. (a) Phase contrast, (b) bright.field picture of the same spermatocyte nucleus. (c) and (d) Bright,-field photographs of cytoplasmic in-situ hybritlization to I pm sections of testes. Bars represent, IO pm.

Retrotransposon-like

Sequences in Lampbrush Loops

23*79*5-

4.3-

z*o-

0.6-

Figure

8. Hybridization

of labelled

(a) dhMiF2A

and

(h) dhMiFZB to total RNA isolated from female flies (p), testes (t) and carcasses (c) (i.e. males after dissection of the testes). HindIII-digested and denatured 1 DNA was used as a marker (on the left). The arrowhead marks the 1.2 kb RNA species hybridizing with dhMiF2A.

testes of Drosophila and their transcripts display a length heterogeneity comparable to that of dhMiF2B-homologous RNA (Huijser et al., 1987; Huijser et aE., unpublished results). We suppose that the heterogeneous size pattern of the RNA seen with dhMiF2B as probe is mainly caused by the (CA/GT), repeats. 4. Discussion (a) Identijcation of a new family of retrotransposons From our previous studies we had circumstantial evidence that the Y-associated sequence type belongs to the transposable class of DNA sequences (see Hennig et al., 1987). The DNA sequences of dhMiF2 and dhMiF8 now provide additional and direct evidence that Y chromosomal DNA sequences from an active fertility gene belong to a class of repetitive DNA elements that in their sequence content and structure closely resemble integrated forms of known retroviruses or retrovirus-like elements. retrotransposons of the class In Drosophila, carrying direct LTRs have been described as copialike transposable elements (for a review, see Finnegan, 1985). DNA with sequence similarity to

695

copia is not represented in the D. hydei genome (Martin et al., 1983; P. Vogt, unpublished results). However, the comparison of the DNA sequences of dhMiF2 and dhMiF8 with the DNA sequence of the copia-like element 17.6 revealed similarities. The most obvious correlationship is the potential to code for a reverse transcriptase-like protein. Internal mutations, nevertheless, make it unlikely that transcripts of dhMiF2 and dhMiF8 in D. hydei ‘can be translated into a functional protein. However, amino acid residues that have been identified as strongly conserved during evolution in a widely divergent range of organisms are also conserved within the putative D. hydei protein. Since the haploid genome of D. hydei carries approximately 30 copies of this repetitive sequence family (Hennig et al., 1983), we do not know whether other copies carry a DNA sequence capable of being converted into a functional protein. The DNA sequences of dhMiF2 and dhMiF8 are only partially overlapping, and it is likely that both together represent the length of a full size member of the repetitive DNA family. This is strongly supported by the homologous DNA fragments recovered from the D. melanogaster genome, which will be reported in a separate publication (Lankenau et al., unpublished results). From the combined sequence information we can derive that this family of retrotransposons has LTRs of at least 200 bp (Fig. 7). Since the putative 5’ LTR of dhMiF2 starts with the EcoRI restriction site used for cloning, it cannot be excluded that the actual size of the suspected terminal regions of the transposable elements extends upstream from this 200 bp. (b) Expression of the dhMiF2 family DNA sequence8

of repeated

Transcript in-situ hybridization with the DNA fragment including the reverse transcriptase-like protein-coding sequence to primary spermatocyte nuclei confirmed that this DNA sequence is represented in transcripts of the lampbrush loop threads. This lampbrush loop pair was microdissected and yielded the DNA fragments dhMiF2 and dhMiF8 after microcloning (Hennig et al., 1983). Transcripts of the lampbrush loop pair pseudonucleolus also were labelled by in-.&u hybridization. This observation confirms the results of our screening microclones from pseudonuclei with the dhMiF2A probe. We recovered several cloned DNA fragments with homology to dhMiF2 (Huijser et al., unpublished results; see also Huijser, 1987). These data are consistent with other results indicating that the two lampbrush loop pair threads and pseudonucleolus share members of the same families of repeated DNA sequences (Huijser & Hennig, 1987). This may be correlated to the complex genetic interrelationship between the fertility gene loci A to C corresponding to the lampbrush loop (see pair threads, cones and pseudonucleolus Hennig, 1985; Hackstein et al., unpublished results).

696

t’. Huijser

In Miller spreading experiments it was noticed that, between the giant transcripts of the pseudonucleolus, small transcripts are interspersed, which has been considered as an indication of secondary initiation sites within this lampbrush loop (Loos et al., 1984). It is tempting to speculate that these smaller transcripts are initiated at the promoter regions. Transcripts with homology to t’he dhMiF2 family of repeated DNA sequences are present as a prevalent RNA fraction in testis. With the fragment dhMiF2A encoding the putative reverse transcriptase-like protein, an RNA species of a size equivalent to the length of this polypeptide was identified. The question arises as to whether this RNA species is derived from the Y chromosomal copies of the micropia family or from copies in other genomic positions. A definite answer to this question cannot be given. However, we assume that part of this RNA may be of Y chromosomal origin, since the Y chromosomal copies are expressd with a high transcriptional activity, as in-situ hybridization has shown. This, of course, does not prove that the Y chromosomal t’ranscripts are indeed spliced and transferred into the cytoplasm. Moreover. we cannot decide whether any functional coding sequence for the reverse transcriptase-like protein exists in the D. hydei genome. Arguments in favour of the presence of functional copies of the reverse transcriptase-like protein are derived from our studies of the location of members of the copia family in relat’ed species, such as Drosophila neohydei and D. eohydei (Hennig et al.. 1983, and unpublished results). In these species the polytene chromosome banding patterns are highly conserved (Hennig, 1978). However, the micropia sequences outside t,he Y chromosome are in genomic sit,es entirely different from those of D. hydri. This can be explained only by transpositions of members of this family of repeated DNA sequences during evolution. The intense testis-specific expression of the micropia family merits particular attention. For Y chromosomal DNA copies of t,his sequence family the expression is not unexpected, since it’ is well established that transcriptional activity of large amounts of Y chromosomal DNA of Drosophila is required during spermatogenesis. Whether other genomic copies are transcriptionally active in testes is unknown. In the context, of our knowledge on other transposable elements this seems likely. In several organisms, transposable elements have been demonstrated to cause mutations in cases where they are transcriptionally active in the germ line and are induced to undergo transpositions (Finnegan, 1985; Rubin et al., 1985). Several authors proposed that activity in transposition is an important evolutionary role of transposable elements, since they are able to generate variability in protein-coding regions on the basis of t’heir excision mechanism (see Schwarz-Sommer, 1987). In addition, transposable elements may become involved in the regulation of transcriptional

et al.

activity of genes helonging t,o a common dt~velol~& 8aedler. mental pathway (Schwarz-Sommer 1987). Such arguments fit. well t,o our hypothesis sequences are important that transposable constit,uents of the Y chromosomal fert,ility genes of Drosophila (Vogt et al.. 1986; Hennig d al.. 19X7). The Y chromosomal t,ranscripts of the micropin family and ot’her DNA elements of similar nature may be important for interactions with their family members outside the T chromosome. The identification of dhMiF2 as a memhcxr of a family of retrotransposons confirmed and sub stantiated our earlier conclusions on the rnolecula,l struct,ure and evolut’ion of the Y chromosomal WV lampbrush loops of Drosophila ( for surnmariw Hennig et al., 1987: Hennig, 1987). Future stjudirs will have to explain the biological role of’ i he transposable element’s within the \’ chromosomal fertility genes. IVr thank Dr %z. Schwarz-Snmmrr for hrr support IIL sc~rrrn t’hr EMIT, grnrbank and R. T,ugtighrid f’ol carrying out the c~toplasmic in-sift/ hybridizat,iorl experiments. To our wIleagues Rein Brand. .lohannrs Hackstein. Ron Hochstenbach. Hannie Kremrr and Koos Miedema, we are grateful for discussions and for (.ritiwI reading of the manuscript.

References Amersham Handhook.

(1984). .WZ3 Amersham

C’lonirry trrrd ‘+ycrrr,circg [ntrrnat~ional. A-Imc~rsham.

l’.K. T3oekr. .I. I)., (:artinkel, I). .I.. St?ltAs. (‘. .I. & Fink ( :. f:. (19%). (‘Pll. 40. 491-500. Cirgwin, ,I. M.. T’rz~byla. A. E.. Jla~l~onaltl. I 659661. Sanger, F., Nicklen, S. & Co&on, A. R. (1977). Proc. Nat. A cad. Sci., U.S. A. 74. 5463-5467.

697

Scalenghe, F., Turco, E., Edstrom, J.-E.. Pirrotta, V. & Melli, M. (1981). Chromosoma, 82. 20%216. Schwarz-Sommer, Zs. (1987). In Structure and Function of Eukaryotic Chromosomes (Hennig, W.. ed.), vol. 14, pp. 213-221, Springer-Verlag, Berlin and Heidelberg. Schwarz-Sommer, Zs. & Saedler, H. (1987). &foZ. Gen. Genet. 209, 207-209. Schwartz-Sommer, Zs., Leclerq, L., Giibel. E. & Saedler. H. (1987). EMBO J. 6, 3873-3880. Thomas, P. S. (1980). Proc. Nat. Acad. Sri., I'.S.A. 77, 5201-5205.

Toh. H.. Hayashida, H. & Miyata. T. (1983). Nature (London), 305, 827-829. Toh. H.. Kikuno, R., Hayashida. T., Kugimiya, W., Inouye, S., Yuki, S. & Saigo, K. (1985). EMBO J. 4, 1267-1272. Vogt, P. & Hennig, W. (1983). J. Mol. Biol. 167, 37-56. Vogt,, P. & Hennig, W. (1986a). Chromosoma, 94, 44% 458. Vogt, P. & Hennig, W. (19863). Chromnsoma, 94. 458p 467. Vogt, P., Hennig, W. & Siegmund, I. (1982). Proc. Nat. Acad. Sci., U.S.A. 79, 5132-5136. Vogt, P.. Hennig, W., Hacken, D. ten & Verbost,, P. (1986). Chromosoma, 94, 367-376. Yanisch-Perron. C., Vieira, J. & Messing. .J. (1985). Gene, 33. 103-l 19.

Edited by R. A. Laskey