Gene, 121 (1992) 347-352 0 1992 Elsevier Science Publishers

GENE

B.V. All rights reserved.

341

0378-l 119/92/$05.00

06736

The 28s ribosomal RNA-encoding gene sequences in the retrotransposon-rich regions (Parasitoid

wasp; bees; mobile elements;

Yves Bigot, Franqoise

Lutcher,

retrotransposon;

Marie-H&be

male-haploid

Hamelin

of

Hymenoptera:

inserted

system)

and Georges PCriquet

Institut de Bioctkotique Expkimentale des Agrosyskmes, Fact& des Sciences. Pare Grandmont, 37200 Tours. France, Tel. (33-47)366967 Received

by H.M.

Krisch:

13 December

1991; Revised/Accepted:

1 May/l3

May 1992; Received

at publishers:

28 July 1992

SUMMARY

The genomes of two parasitoid wasps, Diadromuspulchellus and Eupelmus vuilleti, and the honey bee, Apis melllifra, contain few interspersed repeated sequences corresponding to transposons (Tn). This suggests that the genomic organisation of Hymenoptera could be due to the elimination of deleterious Tn in haploid males. We have used restriction-fragment length polymorphism analysis to show that nondeleterious Tn are present in the DNA (rDNA) encoding ribosomal RNA of twelve species of Hymenoptera. Sequence analysis of the 28s rDNA type-1 and type-II insertion-rich regions of 80 species showed that this region is very highly conserved (95.8%). A consensus sequence and restriction map of the rDNA region were established. These sequence data were used to develop a strategy for detecting inserted elements in the rDNA fragments containing type-1 or type-II insertion sites, and this strategy was used to screen twelve hymenopteran species and four non-Hymenoptera control species. The rDNA fragments from the Hymenoptera and control species contained inserted sequences in the area where type-1 and type-II elements are inserted in the 28s rDNA retrotransposon-rich region of Diptera and Lepidoptera. The hymenopteran genomes therefore appear to contain repeated elements, the mobility and nature of which remain to be determined.

INTRODUCTION

One ofthe most fascinating evolutionary questions raised by the haplo-diploid system of Hymenoptera, in which the males come from parthenogenetic eggs with n chromosomes and the females from fertilized eggs with 2 n chromosomes, is the possible relationship between this genomic organi-

Correspondence to: Dr. Bigot Y., IBEAS, Grandmont,

37200 Tours,

Tel. (33-47)36.69.76; Abbreviations:

Fact&e

des Sciences,

Part

France.

Fax (33-47)36.70.40.

B., Bombyx; bp, base pair(s); kb, kilobase

or 1000 bp;

D., Drosophila; Di, Diadromus; Din., Dinarmus: E., Eupelmus; r-, ribosomal; rDNA, DNA encoding rRNA; RFLP, restriction-fragment length polymorphism; RTn, retrotransposon(s); TAE, 40 mM Tris,OH/20 mM acetic acid/2 mM EDTA pH 8.1; TBE, 89 mM Tris,OH/89 mM boric acid/2.5

mM EDTA pH 8.3; Tn, transposon(s).

sation and the low level of genetic variability. The adaptative capability of this system has apparently enabled the Hymenoptera to diversify as efficiently as many other insect orders. They have developed a highly effective ecological and ethological strategy of parasitism (Bouletreau, 1986) a very low level of enzymatic polymorphism (Grauer, 1985), and a male haploidy which may cause the loss of deleterious mutations (Woods and Guttman, 1987). This low level of genetic variability raises questions concerning the relative contributions of Tn, recombination and base mutation, the evolution and fluidity of the hymenopteran genome. The genomic organisation and the interspersion of repeated sequences in the two parasitoid wasps D. pulchellus and E. vuilleti (Bigot et al., 1991), and in Apis mellifra (Cram et al., 1976) differ from those of D. melanogaster. These studies suggest that hymenopteran genomes have

.a

.A-.

a.

.--.

.c. Y.

.a..

AL_...

.

.--.

cl.. Y..

qtc.. . . . . . . ..c.

.d._yl.

.

.

.---a

.

.

.LL_UY.

. t. . .a..qc.

........ ... a............................c......t ............. .... ..t.......~~.qc................q .a.......,.~...........~_. .............. 7 q .....

......................... ..... t...........1.....t ................................................

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . ..__--_..........g......................c........................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..c......................c........c................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..~..........y......................c........c................... .__._..____..._...._.._.......___........ .q......................c........c............. .._.........._...............~L..............__. t. ................................................ t. ............ ..‘~.........t . . ............. .g..........:

CGG~~C-TTGC-CI,c.~.:;TCl~GCGl~G

FhAC,~~r,ACCCTG'I'TG~~~~T~ACTCTAc-CT

... ..aaq.gc.t.......................t.........c ..... c...ca.3 ....................... a.at.......................,.,., ............

... .._~.._ ............ ........

._L ...... .......... a.t...t ...................... a.t...t..................t...__........a .......... a.t...t..................t..._.,........dtc .......... ..

Y.. .g.

...... .

.

.Y. .gt .g.

. . .c.

t

.. .. ... . . .. .

. .. .. .. . . .. . . . . .. . ... .

. ._. . .LL.L.

..LLY.

..L_.

t

.-..

.

._a-*.

.

J-L_.

.......................

pigmentosa

Dictyostelium

..t ..a

.......................

Crithidia fasciculata discordeum

Coenorhabdilis elegans Ascaris lumbricordes

...........

......... ..ya.~a.......................,__

. .

SEQUENCE

Tetrahymona

..........

... a

.

CONSENSUS

I

...... ..a.a.~..........................,_~,

... .

.. .. t

.. . . . . ... . . . .

LA>._.

g. .

Calliphra erythrocephala Drosophila melanogaster Drosophila virilis Sciara coprophila Bombyx mori Xenopus laevis Mus musculus Rattus norvegicus Homo sapiens Saccharomyces

cerevisiae

Physarum polycephalum Citrus limon Oryza sa tiva

Fig. 1. Sequences of the RTn-rich regions of the 28s rRNA genes of 18 eukaryotic species: protozoans (T. pigmentosa, C. fasciculatu, D. dkcoideum, annelids (C. elegans, A. lumbricoides), insects (C. erythrocephala, D. melanogaster, D. virilis,S. coprophila, B. man), vertebrates (M. musculus,R. norvegicus, H. sapiens),fungi (S. cerevisiae, P. pofycephalum) and plants (C. limon, 0. sativa). The underlined, large-type sequences correspond to insertion sites of the type-1 and type-II elements present in Diptera and B. mori. x indicates are given in the figure.

indefinite nt position.

Empty positions

indicate unavailable

data. Full species names

349 fewer Tn than other eukaryotic all. The

eukaryotic

species, or perhaps

rRNA-encoding

multigene

type-l site

none at

I

type-l site

family

(rDNA) is generally composed of a few hundred to a few thousand copies of repeated units arranged in tandem arrays on one or more chromosomes. One of the common causes of variation in rDNA units is the length of the intergenic spacer. A second cause of variation in rDNA unit length is the interruption of the 28s ribosomal genes by two types of insertions, which have been studied in detail in D. melanogaster and Bombyx mori (Jackubczak et al., 1990; 1991 for review). Type-I (Rl) elements are inserted at exactly the same position in the 28s gene and

pDmr628

fragment

generate identical target site duplications. The type-II (R2) insertions have been reported to be less widespread and interrupt the 28s gene at the same position in both species, 75 bp upstream from type-1 elements, but without duplication of the insertion site. However, the type-1 and type-II elements of these species show no sequence or size similarities. It is not clear why this highly conserved stretch of about 200 bp within the 28s coding region is so rich in inserted elements. All the available data indicate that both types of elements are transposable (Xiong et al., 1988) and correspond to RTn elements. These inserted elements are frequent in the r-genes of several insect taxa, suggesting that examination of the r-region would be a good starting point for a study of mobile sequences in hymenopteran genomes. We have therefore analysed the rDNA sequence in hymenopteran species and established a consensus restriction map of the region where type-1 and type-II elements are inserted in Diptera and Lepidoptera. These sequence data were used to develop a strategy for detecting inserted elements in the type-1 and type-II RTn-rich region. This strategy was employed to screen twelve hymenopteran species and four non-hymenopteran control species. All the hymenopteran and control species tested contained inserted sequences in the regions where type-1 and type-II elements are inserted in the 28s rDNA RTn-rich region of dipteran and lepidopteran species.

EXPERIMENTAL

AND DISCUSSION

(a) Sequence analysis of the RTn-rich region of the 28s rRNA-encoding genes Eighteen sequences were collected from the EMBL and GeneBank databases. Multiple alignment was carried out both manually and automatically with the CLUSTAL Program of CIT12 (Fig. 1). A consensus sequence was deduced for the highly conserved 251 bp containing the two insertion sites described in Diptera and B. mori. This RTnrich region showed an internal sequence conservation of

Fig. 2. Restriction

maps of the pYDmrl2

of D. melanogaster)

and pDmr628

rRNA genes of D. melanoguster). type-1 and type-II insertion B, BumHI;

C, @I;

AspHI fragment; fragment.

fragment

fragments

(rDNA repeated

(RTn-rich

The arrows indicate the locations

sites. NTS, nontranscribed

E, EcoRI;

H, HindIII;

probe 2, 71-bp @I-CfiI

Open boxes correspond

spacer:

S, AspHI; fragment;

with transcribed

unit

region of the 28s

probe

of the

A, A[uI; 1, 211-bp

probe 3, CfoI-AluI

spacers and blackened

boxes with 18S, 5.8s and 28s r-genes.

95.8% for the 18 sequences, from taxa as different as protozoa, invertebrates, mammals, fungi and plants. A search for all restriction sites present in the 25 1 bp of the consensus sequence revealed four highly conserved restriction sites (Fig. 1): two were in the 5’ region of the type-II insertion site (AspHI and @I), one was between the type-1 and type-II insertion sites (@I), and one was in the 3’ region of the type-1 insertion site (AluI). These data suggested that a study of the restriction fragment length polymorphism of 211-bp AspHI-AluI, 71-bp CfoI-C.‘oI and 107-bp @I-AluI sequences would detect insertions, limited only by the sensitivity of Southern blots and the control of hybridisation conditions. Experimentally, 211-bp AspHI-AZuI fragment or its subfragments: 71-bp C&I-CfiI and 107-bp C’I-AZuI were purified from the pDmr628, a subclone of pDmrY12 (Fig. 2 Terracol and Prud’homme, 1986), and used as probes. Fragments labelled by primer extension were chromatographed on Sephadex G50, and fragments shorter than 50 bp were eliminated. The hybridisation and washing temperatures were 62°C for probes 1 and 3, and 57°C for probe 2. (b) Verification of the methodology using Drosophila melanogaster The sensitivity and effectiveness of the protocol were tested by looking at length variants on the 211-bp AspHIAluI, 107-bp CfoI-AluI and 71-bp CfiI-CfoI rDNA frag-

350 a 1

2

3

4

5

6

7

8

ments of D. melanogaster in rDNA. Type-I elements interrupt about 50% of the rDNA repeats in this species and

9 10 11 12 13 14 15 16

three forms have been described: a complete form approx. 5 kb long and two 5’-truncated forms (0.5 and 1 kb long). Several sizes of type-II elements are present in about 15 % of the rDNA repeats, with the shorter variants again representing 5’ truncations of the longest 3.5-kb variant (see Jakubczak et al., 1990, for review). Male genomic DNA was used because there is no amplification of uninterrupted rDNA repeats in this sex, thus maximizing variant length fragments. The result obtained AspHI-AfuI fragment type-1 and type-II 16. Five variants

b 1

2

3

4

5 6

7

8

910

11 1213

14 1516

I

4

4220 4190 0175

.I07

0

70

c 1 2

3

4

5

6

7 8

9

10 1112131415

16

bp -460

the detection

of

by RFLP analysis of the 211-bp (probe 1, Fig. 2), which contains

insertion sites, is shown in Fig. 3, lane (800 bp, 680 bp, 600 bp, 530 bp and

395 bp) of the major uninterrupted 2 11-bp rDNA fragment were detected, corresponding to the extremities of the major inserted elements (the complete form and the two 5’-truncated forms of type-1 elements and the complete form of type-II elements). The RFLP analysis of the 107-bp CfiI-&I fragment (probe 3, Fig. 2) which contains a type-1 insertion site, is shown in Fig. 3b, lane 16. Four variants (220, 190, 175 and 70 bp) of the major uninterrupted 107-bp rDNA fragment were detected. They corresponded to the extremities of type-1 inserted elements (the complete form and the two 5’ truncated forms). The RFLP analysis of the 71-bp @I-CfoI fragment (probe 2, Fig. 2) is shown in Fig. 3c, lane 16. The major uninterrupted 7 1-bp rDNA fragment contained 35-50-bp variants corresponding to type-II elements, which migrated as a broad band. Thus the strategy used will detect variant length fragment reflecting the presence of inserted elements in the insect 211-bp RTn-rich region. The data also show that accurate location of these insertions is possible in the type-l insertion site region (CfiI-A/u1 107-bp fragment; Fig. 3b) and the type-II insertion site region (71-bp CfiI-C’foI fragment; Fig. 3c) described in Diptera and B. mori. However, although this strategy detects insertions, it provides no proof of their absence as there may be detection problems if the

Ictopectis tunetana (5),

Triapsis luteides (6),

Dinarmus busalis (7),

Trichogramma caceociae (8), Apis melrifra (9), Bombus terrestris (lo), Xylocopa violacea (1 l), Myrmica ruginodis (12) and four non-hymenopteran species:

Locusta migratoria (13) Acrolepiosis assectella (14), Bruchidius

atrolineatus (15), Drosophila melanogaster

(16).Genomic

DNAs

were di-

gested with AspHI and AluI in panel a, C&I and AluI in panel b, and CfoI in panel c. Fragments b,c) NusieveGTG

Fig. 3. Southern

blot analysis

of the complete AspHI-ALI

210-bp (panel

were separated

on 3.5”/, (panel a) or 4% (panels

agarose gels (FMC product),

1 x TAE (panel

a) or TBE

(panels b,c), blotted for 2 h and hybridized with probe I (panel a), 2 (panel c) or 3 (panel b) (Fig. 1) from D. melanogaster. The data shown are the best autoradiograph intensities for each species. Small arrows between

a) RTn-rich region of the 28s rDNA, 107-bp type 1 (panel b) and type 2 (panel c) RTn-rich subregion in twelve Hymenoptera: Eupelmus vuilleti

lanes indicate

(I), Eupelmus orient& (2), Diadromuspulchellus (3), Diadromus collarik(4),

inserted

the presence

of weak autoradiograph

on the side of lane 16 indicate

variant

element in the 28s rDNA

fragments

bands.

Arrowheads

corresponding

of D. melanogaster.

to the

351 levels of the inserted number

of rRNA

elements,

or the ratio between

genes and the genome

the

size is too low.

(d) Conclusions (I) This study was carried out to determine whether interspersed repeated sequences corresponding to Tn are

(c) Analysis of the RTn-rich region of twelve hymenopteran species The three restriction length analyses were applied

present in the genomes of hymenopteran parasitoid wasps and bees. The published data on the type-1 and type-II insertions of D. melanogaster (Jackubczak et al., 1990 for

(Fig. 3a) to twelve parasitoid

reviews) permitted us to test and define the limits of the strategy used. The digests of D. melanogaster, AspHI-AluI

wasps (for full names of gen-

era, see Fig. 3 legend): E. vuilleti, E. orientalis, T. luteides, Din. basalis, Di. pulchellus, Di. collaris, I. tunetana, T. caceociae (lanes l-S), three Apinae: A. mellifra, B. terrestris, X. violacea (lanes 9, 10, ll), one Formicidae: M. ruginodis (lane 12) and four non-hymenopteran species (control), including one Orthoptera: L. migratoria; one Lepidoptera: A. assectella; one Coleoptera: B. atrolineatus, and one Diptera: D. melanogaster (lanes 13, 14, 15, 16). The genomic DNA of males or workers of the social hymenopteran species (A. melhyera, B. terrestris and A4. ruginodis) was used to maximise detection of variant length fragments. Mass sexing of Trichogramma is impossible, because the insects are too small. All studies were performed using DNA from a species generating only females by telytokous parthenogenesis (T. caceociae). Fig. 3a shows the analysis of 211-bp insect RTn-rich regions. The intact fragment in lanes 1 to 15 is shorter than that of D. melanogaster. The intact subfragments (107 and 71 bp) have the same length in all species (Fig. 3, b and c). There is therefore an extra AspHI or AluI site located between the AspHI and AluI sites of the consensus sequence (Fig. 1) in the 5’ region of the type-II insertion site in all non-dipteran species. All the hymenopteran species tested contained variant length fragments. The band intensity in each lane suggested that among the twelve hymenopteran species tested, only T. luteides had a small proportion of its rDNA repeats interrupted by insertions. Variant length fragments were detected in the four control species (lanes 13-16), in agreement with the literature (Schaeffer and Kunz, 1987). Those of L. migratoria probably correspond to the 7-kb and 0.7-kb elements inserted within its RTnrich region of the 28s rDNA. Fig. 3b shows the analysis of the 107-bp type-1 RTn-rich region. All the species tested had numerous variant length fragments, indicating that this subregion is rich in inserted elements. Fig. 3c shows the analysis of the 71-bp type-II RTn-rich region. Eight of the twelve hymenopteran species tested had variant length fragments, indicating the presence of inserted elements in this sub-region. The results for Di. collaris, T. luteides, Din. basalis and M. ruginodis suggest that there are probably no inserted elements in the 7 1-bp type-II Tn-rich region. However, the gel electrophoresis and Southern blots were at their limits of detection in analysing this region, and we may not have detected a low copy number of inserted elements.

and AluI-CfoI fragments hybridised with probe 1, 2 or 3 (Fig. l), were used to detect RFLP resulting from the presence of type-1 and type-II RTn. However, rare variant length fragments generated by 5’-truncated forms of the element might not be detected by our method. The strategy detects the presence of insertions, but cannot show the absence of insertions in a species. This limitation is aggravated by the fact that RTn different from types I and II, and inserts in other regions of the 28s rRNA-encoding genes have been found in Anopheles gambiae, Anopheles arabiensis (Paskewitz and Collins, 1989) and Sciara coprophila (Kerrebrock et al., 1989). In addition, the variant length fragments in AluI + CfoI digests hybridised by probe 2 or 3 (Fig. 1) do not imply that RTn are present in the same type-1 or type-II insertion sites as in Diptera and B. mori. They only indicate that inserted elements are present in the 107-bp @I-AluI type-1 or 71-bp @I-CfiI type-II insertion-rich fragments. This second limitation is exemplified by the fact that Ascaris lumbricoi’des contains a type-1 RTn-like element inserted 39 bp upstream of the type-1 insertion site of D. melanogaster (Back et al., 1984). However, from the twelve insertions sequenced in this region of the 28s genes from insect species, only one has been in neither the type-1 nor type-II site. It is therefore most probably the same in the hymenopteran 28s gene. (2) Like several gene families, rRNA-repeated units within individuals and populations undergo concerted evolution of nt sequences by molecular mechanisms such as unequal crossing over and biased gene conversion (Hillis et al., 1991). One of the consequences of buffering features of concerted evolution might be the ‘elimination’ of units containing inserted elements. In individuals and populations, the maintenance of a Tn family in any unit of a gene family probably requires that the elements retain a minimum mobility during evolution. Hence, the inserted elements most likely correspond to Tn of the type-1 or type-II RTn families in all twelve of the hymenopteran species examined. These results are in agreement with the recent paper of Jacubczak et al. (1991). These authors showed the presence of inserted elements in a fraction of the 28s region of three hymenopteran species, the sequences of which proved to be type-1 and type-II RTns. The method presented here is capable of ascertaining the presence or absence of inserted sequences in such sites. (3) Hymenopteran species show little genetic variability.

352 It has been suggested that this is because male haploidy directly eliminates all ‘disadvantageous’ and deleterious genes (Lester and Selander, 1979; Grauer, 1985). Studies of the genomic organisation of A. meZZz$zru (Crain et al., 1976), Di. pulchellus and E. vuilleti(Bigot et al., 1991) show

Bouletreau,

interactions

between D. (Eds.), of

DNA sequence arrangement in Apis melliferu (honey bee) and Musca domestica (housefly). Chromosoma 56 (1976) 309-326. Grauer,

with repeat patterns which vary from one species to another. However, these specifically inserted putative Tn may not have direct deleterious effects on insect viability, because of the repeated structure of the 28s rRNA genes and

Jakubczak,

G.: Gene diversity

in Hymenoptera.

Evolution

39 (1985) 190-

199. Hillis, D.M.,

Moritz, G., Porter,

ased gene conversion

C.A. and Baker, R.J.: Evidence

in concerted

evolution

of ribosomal

for bi-

DNA. Sci-

ence 251 (1991) 308-310. J.L., Xiong, Y. and Eickbush,

(RII) ribosomal

DNA

T.H.: Type-I (RI) and type 11

of Drosophila melunogaster are rc-

insertions

elements closely related to those of Bombyx mori. J.

trotransposable

Mol. Biol. 212 (1990) 37-52. Jakubczak, J.L., Burke, W.E. and Eickbush, elements Rl and R2 interrupt

T.H.:

Retrotransposable

the rRNA genes of most insects. Proc.

Natl. Acad. Sci. USA. 88 (1991) 3295-3299. Kerrebrock, A.W., Srivastava, R. and Gerbi, S.A.: Isolation acterisation

of ribosomal

DNA

variants

and char-

from Sciura coprophila. J.

Mol. Biol. 20 (1989) 1-13. Paskewitz,

S.M. and Collins, F.H.:

tion elements quence

ACKNOWLEDGEMENTS

and co-evolutionary

and their hosts. In: Waage, J. and Greathead,

Insect Parasitoids. Academic Press, London, 1986, pp. 169-200. Crain, W.R., Davidson, E.H. and Britten, R.J.: Contrasting patterns

that interspersed repeated sequences corresponding to Tn are rare or absent. It has also been suggested that this feature might be a consequence of male haploidy. Our study shows that putative Tn are present in 28s rRNA genes,

therefore would not be eliminated by male haploidy. According to this hypothesis, the insertions in hymenopteran genomes would occur only in nonessential regions or in rRNA-encoding genes and other tandem genes.

M.: The genetic

parasitoids

Site-specific

ribosomal

DNA inser-

in Anopheles gumbiue and A. arabiensis: nucleotide

of gene element

boundaries.

Nucleic

se-

Acids

Res. 20 (1989)

gene amplification

does not occur

8125-8133.

The authors thank R. Terracol for supplying the plasmid and the technical personnel working with G. Devauchelle, B. Paintureau and B. Leroux for supplying the insects. This work was supported by C.N.R.S. (URA 1298) and M.E.N.-DRED (Evolution 89-1616).

Schaefer,

M. and Kunz, W.: Ribosomal

in the oocytes Terracol,

of Locusta migratorkz. Dev. Biol. 120 (1987) 418-424.

R. and Prud’homme,

N.: Differential

elimination

of rRNA genes

in bobbed mutants ofDrosophila melanogaster. Mol. Cell. Biol. 6 (1986) 1023-1031. Woods,

P.E. and Guttman,

menoptera:

Symphyta,

S.I.: Genetic Diprionnidae),

levels of genetic diversity

variation

in Neodiprion

sawflies and a comment

within Hymenoptera.

(Hyon low

Ann. Entomol.

Sot.

Am. 80 (1987) 590-599. REFERENCES

Xiong, Y. and Eickbush, element

Back, E., Meir, E.V., Mueller, F., Schaller, D., Neuhaus, H., Aeby, P. and Tobler, H.: Intervening sequences in the ribosomal RNA genes of Ascaris lumbricoides: DNA sequences nization.

and genomic

orga-

EMBO J. 3 (1984) 2523-2529.

Bigot, Y., Hamelin, genomic

at junctions

M.H.

organization

and Periquet,

rotransposon.

analysis

of the

Diadromus pulchellus and

Eupelmus vuilleti.J. Evol. Biol. 4 (1991) 541-556.

T.H.: The site-specific

belongs

DNA insertion

ribosomal

DNA insertion

to a class of non-long-terminal-repeat

ret-

Mol. Cell. Biol. 8 (1988) 114-123.

Xiong, Y., Burke, D.W., Jakubczak, quence

G.: Molecular

of the Hymenoptera

RlBm

elements

specific manner

RlBm

J.L. and Eickbush,

H.: Ribosomal

and R2Bm can transpose

to locations

Acids Res. 16 (1988) 10561-10573.

outside

in a se-

the 28s gene. Nucleic