of 12 /12
Insect Molecular Biology (1999) 8 (4), 469 – 480 © 1999 Blackwell Science Ltd 469 Blackwell Science, Ltd Phylogenetic relationships of seven palearctic members of the maculipennis complex inferred from ITS2 sequence analysis M. Marinucci, R. Romi, P. Mancini, M. Di Luca and C. Severini Laboratory of Parasitology, Istituto Superiore di Sanità, Rome, Italy Abstract The sequences of the second internal transcribed spacer (ITS2) of ribosomal DNA (rDNA) were determined from seven palearctic mosquitoes species belonging to the Anopheles maculipennis species complex, namely An. atroparvus , An. labranchiae , An. maculipennis , An. messeae , An. melanoon , An. sacharovi and An. martinius . The length of the ITS2 ranged from 280 to 300 bp, with a GC content of 49.4–54.1%. With the excep- tion of An. messeae , negligible levels of intraspecific polymorphism and no intrapopulation variation were observed. The phylogenetic relationships among the members of the maculipennis complex were inferred by maximum-parsimony analysis of the PAUP program and the neighbour-joining and maximun-likelihood analysis of the PHYLIP program. All of the trees obtained were almost identical in topology, although the relation- ships among three species, i.e. An. maculipennis , An. messeae and An. melanoon , remained unresolved. The phylogenies were in good agreement with the previous gene–enzyme and polytene chromosome banding pattern studies. Keywords: rDNA, ITS2, philogeny, Anopheles maculi- pennis complex, malaria vectors. Introduction In the early decades of this century, ecological studies on the most important species of malaria vector in Europe, Anopheles maculipennis Meigen (Diptera: Culicidae), led to the finding that it was actually a complex of species with overlapping morphological characters, providing the explanation for the ‘Anophelism without malaria’ (Hackett & Missiroli, 1935). The group of anopheline mosquitoes usually referred to as An. maculipennis complex (White, 1976, 1978) includes palearctic as well as nearctic cryptic species. The identification of the palearctic members of this complex, which includes both malaria vector and non- vector species (the number of species depending on taxo- nomic treatment), is largely dependent on egg morphology (Falleroni, 1926; Corradetti, 1934), the adults being difficult or impossible to distinguish by morphological characters. Polytene chromosome banding patterns (Kitzmiller et al ., 1967) and electrophoretic analysis (Bullini & Coluzzi, 1982) proved to be a useful approach to study systematics in the palearctic species. However, an unequivocal phylogeny has not emerged because of the close relationships within of this group of mosquitoes and the lack of DNA data. In insects, ribosomal DNA (rDNA) consists of tandemly repeated transcriptional units with each unit containing the genes for the 18S, 5.8S and 28S ribosomal RNA (rRNA) (reviewed in Gerbi, 1985; Hillis & Dixon, 1991). The internal transcribed spacers (ITSs) ITS1 and ITS2 flanking 5.8S gene separates it from 18S and 28S, and an external transcribed spacer (ETS) on 5 end of the 18S completes the transcriptional unit. The basic rDNA repeat (rDNA array) thus consists of the transcriptional unit plus an adjacent intergenic spacer (IGS) which is not tran- scribed. The rDNA arrays vary in the copy number from approximately 100 in Sciara coprophila , to more than 1000 in Calliphora erythrocephala and in several Drosophila species (Beckingham, 1982). These rDNA repeated units do not evolve independently, but at a relatively homo- geneous evolutionary rate within individuals and within the species. Different mechanisms were proposed for the process of concerted evolution (Dover & Coen, 1981; Arnheim, 1983; Gerbi, 1985; Tautz et al ., 1987), including gene conversion and unequal crossing over (Dover, 1982a, b; Seperack et al ., 1988). While the coding regions of the rDNA tend to be highly conserved in evolution, the spacer regions appear relatively free to diverge, even in closely related organisms. Thus, the ITS sequences proved useful for resolving the evolutionary affiliations at different taxonomic levels, including recently diverged Received 6 July 1998; accepted 15 March 1999. Correspondence: Dr Marino Marinucci, Laboratorio di Parassitologia, Istituto Superiore di Sanità, viale Regina Elena, 299, 00161 Rome, Italy. E-mail: [email protected]

Phylogenetic relationships of seven palearctic members of the maculipennis complex inferred from ITS2 sequence analysis

Embed Size (px)

Citation preview

Insect Molecular Biology (1999)

8

(4), 469–480

© 1999 Blackwell Science Ltd

469

Blackwell Science, Ltd

Phylogenetic relationships of seven palearctic members of the

maculipennis

complex inferred from ITS2 sequence analysis

M.

Marinucci, R.

Romi, P.

Mancini, M. Di

Luca and C.

Severini

Laboratory of Parasitology, Istituto Superiore di Sanità, Rome, Italy

Abstract

The sequences of the second internal transcribedspacer (ITS2) of ribosomal DNA (rDNA) were determinedfrom seven palearctic mosquitoes species belonging tothe

Anopheles maculipennis

species complex, namely

An. atroparvus

,

An. labranchiae

,

An. maculipennis

,

An. messeae

,

An. melanoon

,

An. sacharovi

and

An.martinius

. The length of the ITS2 ranged from 280 to300 bp, with a GC content of 49.4–54.1%. With the excep-tion of

An. messeae

, negligible levels of intraspecificpolymorphism and no intrapopulation variation wereobserved. The phylogenetic relationships among themembers of the

maculipennis

complex were inferred bymaximum-parsimony analysis of the

PAUP

program andthe neighbour-joining and maximun-likelihood analysisof the

PHYLIP

program. All of the trees obtained werealmost identical in topology, although the relation-ships among three species, i.e.

An. maculipennis

,

An.messeae

and

An. melanoon

, remained unresolved. Thephylogenies were in good agreement with the previousgene–enzyme and polytene chromosome bandingpattern studies.

Keywords: rDNA, ITS2, philogeny,

Anopheles maculi-pennis

complex, malaria vectors.

Introduction

In the early decades of this century, ecological studies onthe most important species of malaria vector in Europe,

Anopheles maculipennis

Meigen (Diptera: Culicidae), ledto the finding that it was actually a complex of specieswith overlapping morphological characters, providing the

explanation for the ‘Anophelism without malaria’ (Hackett& Missiroli, 1935). The group of anopheline mosquitoesusually referred to as

An. maculipennis

complex (White,1976, 1978) includes palearctic as well as nearctic crypticspecies. The identification of the palearctic members ofthis complex, which includes both malaria vector and non-vector species (the number of species depending on taxo-nomic treatment), is largely dependent on egg morphology(Falleroni, 1926; Corradetti, 1934), the adults being difficultor impossible to distinguish by morphological characters.Polytene chromosome banding patterns (Kitzmiller

et al

.,1967) and electrophoretic analysis (Bullini & Coluzzi, 1982)proved to be a useful approach to study systematics in thepalearctic species. However, an unequivocal phylogenyhas not emerged because of the close relationships withinof this group of mosquitoes and the lack of DNA data.

In insects, ribosomal DNA (rDNA) consists of tandemlyrepeated transcriptional units with each unit containingthe genes for the 18S, 5.8S and 28S ribosomal RNA(rRNA) (reviewed in Gerbi, 1985; Hillis & Dixon, 1991).The internal transcribed spacers (ITSs) ITS1 and ITS2flanking 5.8S gene separates it from 18S and 28S, and anexternal transcribed spacer (ETS) on 5

end of the 18Scompletes the transcriptional unit. The basic rDNA repeat(rDNA array) thus consists of the transcriptional unit plusan adjacent intergenic spacer (IGS) which is not tran-scribed. The rDNA arrays vary in the copy number fromapproximately 100 in

Sciara coprophila

, to more than 1000in

Calliphora erythrocephala

and in several

Drosophila

species (Beckingham, 1982). These rDNA repeated unitsdo not evolve independently, but at a relatively homo-geneous evolutionary rate within individuals and withinthe species. Different mechanisms were proposed forthe process of concerted evolution (Dover & Coen, 1981;Arnheim, 1983; Gerbi, 1985; Tautz

et al

., 1987), includinggene conversion and unequal crossing over (Dover,1982a, b; Seperack

et al

., 1988). While the coding regionsof the rDNA tend to be highly conserved in evolution, thespacer regions appear relatively free to diverge, evenin closely related organisms. Thus, the ITS sequencesproved useful for resolving the evolutionary affiliationsat different taxonomic levels, including recently diverged

Received 6 July 1998; accepted 15 March 1999. Correspondence:Dr Marino Marinucci, Laboratorio di Parassitologia, Istituto Superiore di Sanità,viale Regina Elena, 299, 00161 Rome, Italy. E-mail: [email protected]

IMB140.fm Page 469 Monday, October 18, 1999 9:56 AM

470

M. Marinucci

et al.

© 1999 Blackwell Science Ltd,

Insect Molecular Biology

,

8

, 469–480

taxa, such as sibling species of mosquitoes (Collins &Paskewitz, 1996; Xu & Qu, 1997). Both ITSs are flanked byhighly conserved regions which facilitate their examinationthrough construction of primers for use in polymerasechain reaction (PCR). In particular, ITS2 is generally moreconserved among species than the ITS1, providing anefficient phylogenetical marker for closely related species.

Porter & Collins (1991) carried out a phylogenetic ana-lysis of some members of the

An. maculipennis

complexusing this molecular target. They studied the nucleotidesequences of the rDNA ITS2 region of three nearcticspecies of the complex (namely the two cryptic species

An. hermsi

Barr & Guptavanij and

An. freeborni

Aitken, and

An. occidentalis

Dyar & Knab) as a base of comparison.In this paper, we present the results of the analysis of

the ITS2 sequences of seven palearctic taxa of the same

An. maculipennis

complex, namely

An. atroparvus

Van Thiel,

An. labranchiae

Falleroni,

An. maculipennis

Meigen,

An.messeae

Falleroni,

An. melanoon

Hacket,

An. sacharovi

Favre and

An. martinius

Shingarev.Our ITS2 sequences and those of the three nearctic

species of the same complex available in EMBL gene bankwere examined for phylogenetic relationships to predictthe most probable evolutionary three.

Results

Internal transcribed spacer 2 sequences

The rDNA ITS2 regions were amplified by PCR fromgenomic DNA of individual mosquitoes. No specimens of

An. subalpinus

and

An. becklemishevi

, the other palearcticmembers of the

maculipennis

complex, were available. OneITS2 clone from five to seven different individual mosquitoesof each of the field-collected samples listed in Table 1 wassequenced and the levels of spacer variability withineach sample were found to be negligible. To assess any

intraindividual ITS2 variability (of which none was detected),three to five ITS2 clones from a single specimen ofeach species and population were sequenced.

The PCR-amplified sequences varied from 421 bpin

An. maculipennis

to 442 bp in

An. sacharovi

. Thesequence alignments of the seven

Anopheles

species areshown in Fig. 1. Presumptive boundaries of 5.8S and 28Sgenes were deduced from comparisons of alignmentswith the sequences of other mosquitoes (Porter & Collins,1991) and considerations based on the ITS2 secondarystructure of

An. maculipennis

(data not shown). The initialninety-three bases were 5.8S DNA and the last fifty baseswere 28S DNA. All the Nematocera taxa examined so far,with the exception of the Culicidae, had a divided 5.8SrRNA, consisting of a mature 5.8S rRNA and a 2S rRNAseparated by a transcribed spacer (Shimada, 1992; Miller

et al

., 1997). This indicated that the presence of continuous5.8S rRNA gene is a synapomorphy defining the Culicidae(Miller

et al

., 1997).The 5.8S and 28S coding regions flanking the ITS2

sequence were well conserved and almost identical tothose reported for nearctic taxa of the

An. maculipennis

complex (Porter & Collins, 1991). In the 3

end of the 5.8Ssequence, there was either a thymine or a cytosine at posi-tion 89, and

An. sacharovi

departed from the consensusby a thymine instead of an adenine at position 92. Aslightly higher propensity for variants (indels and basechanges) occurred in the 5

end of the 28S region (Fig. 1).Although there was some divergence among the spacers,

including some insertion/deletion differences, the alignmentsamong the species were unequivocal because of the pre-sence of the highly conserved regions (Fig. 1). The lengthof the ITS2 ranged from 280 bp in

An. maculipennis

to300 bp in

An. sacharovi

. The overall base compositionwas (range in parentheses) A, 24.9% (23.3–28.7%); C,27.6% (25.7–28.6%); G, 24.9% (23.7–25.7%); T, 22.6%

Species and name abbreviations Geographical origin Source

Anopheles atroparvusATS Toscana, Italy Istituto Superiore di Sanità (ISS)

Anopheles labranchiaeLAG Toscana, Italy ISSLAP Puglia, Italy ISSLAS Sardegna, Italy ISS

Anopheles maculipennisMAP Veneto, Italy ISSMAS Toscana, Italy ISS

Anopheles martinius Samarcanda, Uzbekistan ISSAnopheles melanoon Lazio, Italy ISSAnopheles messeae

MEP Piemonte, Italy ISSMER Lazio, Italy ISSMET Tomsk, Russia Tomsk State University, RussiaMEK Pavlodar, Kazakhstan Tomsk State University, Russia

Anopheles sacharoviSAT Adana, Turkey ISS

Table 1. Mosquito populations used in the study.

IMB140.fm Page 470 Monday, October 18, 1999 9:56 AM

ITS2 sequence of palearctic

Anopheles maculipennis 471

© 1999 Blackwell Science Ltd,

Insect Molecular Biology

,

8

, 469–480

Figure 1. Sequence alignment of the ITS2 and flanking 5.8S and 28S coding regions of the rDNA of the palearctic An. maculipennis taxa. Start and end of the presumptive ITS2 region are marked by asterisks above the consensus sequence. The consensus sequences underlined represent primers used in PCR amplification. Insertions or deletions (indels) in alignment are denoted by a dash (–). Taxa are abbreviated as in Table 2.

IMB140.fm Page 471 Monday, October 18, 1999 9:56 AM

472

M. Marinucci

et al.

© 1999 Blackwell Science Ltd,

Insect Molecular Biology

,

8

, 469–480

(22.0–24.2%). The GC content varied from 49.4% in

An.sacharovi

to 54.1% in

An. atroparvus

.

Phylogenetic analysis

The presumptive nucleotide sequences of the ITS2 regionof the rRNA transcript of seven palearctic (this study) and

three nearctic taxa (Porter & Collins, 1991) of the

maculi-pennis

species complex and the

An. quadrimaculatus

species A (Cornel

et al

., 1996) were aligned and examinedfor phylogenetic analysis. The alignments are presentedin Annex 1. The alignments of the ITS2 sequences resultedin a total of 382 nucleotide sites, of which 151 were

Figure 1. (continued).

Table 2. Pair-wise Kimura two-parameter distances for rDNA ITS2 sequences. All insertions/deletions were removed from the data set. Taxa are abbreviated as first two letters of genus and first three letters of species names (e.g. An. maculipennis = An. mac). An. mac = MAS in Table 1; An. mes = MER in Table 1; An. lab = LAS in Table 1.

1 2 3 4 5 6 7 8 9 10 11

1 An. mac 00.00 0.0450 0.0685 0.0806 0.0790 0.1154 0.2766 0.3704 0.3287 0.3573 0.23752 An. mes 0.0000 0.0790 0.0875 0.1133 0.1450 0.2778 0.3673 0.3277 0.3631 0.26193 An. mel 0.0000 0.0895 0.1190 0.1551 0.3098 0.3807 0.3455 0.3573 0.25534 An. mar 0.0000 0.1108 0.1400 0.2653 0.3315 0.2988 0.3191 0.22435 An. lab 0.0000 0.0994 0.3353 0.3806 0.3536 0.3774 0.24876 An. atr 0.0000 0.3686 0.4161 0.3739 0.3799 0.33177 An. sac 0.0000 0.3714 0.3567 0.3973 0.35018 An. her 0.0000 0.0274 0.0939 0.15499 An. fre 0.0000 0.0824 0.1660

10 An. occ 0.0000 0.196811 An. qua 0.0000

IMB140.fm Page 472 Monday, October 18, 1999 9:56 AM

ITS2 sequence of palearctic

Anopheles maculipennis 473

© 1999 Blackwell Science Ltd,

Insect Molecular Biology

,

8

, 469–480

variable (38.5%) and eighty-seven (22.8%) were informativeunder the conditions of parsimony.

The average transition/transvertion ratio was 0.89 (SE0.14). The lowest and the highest ratios from taxa wereobserved between

An. hermsi

and

An. freeborni

(0.33)and between

An. maculipennis

and

An. melanoon

(1.57).The overall nucleotide composition (A, T, C and G) of theeleven taxa considered here was 24.4%, 22.2%, 28.0%and 25.4%, respectively.

Estimates of Kimura’s two-parameter distances are shownin Table 2 for all pairs of the eleven ITS2 sequences. Abootstrap consensus neighbour-joining (NJ) tree producedfrom these distance values is shown in Fig. 2. Most of

the phylogenetically critical nodes were significantly sup-ported by bootstrap values (> 90%), but the phylogenet-ical relationships were unresolved for

An. maculipennis

,

An. messeae

and

An. melanoon

. The topology of the NJtree was nearly identical to the bootstrap consensusmaximum-likelihood (ML) tree, presented in Fig. 3. Parsimonyanalysis using the exhaustive search procedure in

PAUP

resulted in two equally parsimonious trees of 234 steps,differing in grouping

An. maculipennis

,

An. messeae

and

An. melanoon

, with consistency index (CI) of 0.842,homoplasy index (HI) of 0.158, CI excluding uninformat-ive characters of 0.761, HI excluding uninformative char-acters of 0.239, retention index (RI) of 0.800 and rescaled

Figure 2. Phylogenetic tree based on rDNA ITS2 sequence data. The tree was constructed using neighbour joining with Kimura two-parameter distances (scale bar) and rooted with An. quadrimaculatus species A. Numbers are bootstrap percentages (1000 replications) for clades supported above the 50% level. Taxa are abbreviated as in Table 2.

IMB140.fm Page 473 Monday, October 18, 1999 9:56 AM

474

M. Marinucci

et al.

© 1999 Blackwell Science Ltd,

Insect Molecular Biology

,

8

, 469–480

consistency index (RC) of 0.764. The bootstrap consen-sus maximum-parsimony (MP) tree is presented in Fig. 4.

Discussion

Internal transcribed spacer 2 sequences

The ITS2 sequences of the seven palearctic species of the

An. maculipennis

complex varied in length from 280 bpin

An. maculipennis

to 300 bp in

An. sacharovi

. The ITS2length of the species examined here was similar to thatfound in the other

Anopheles

species:

An. nuneztovari

(363–369 bp; Fritz

et al

., 1994), the

An. quadrimaculatus

complex (287–329 bp; Cornel

et al

., 1996) and the North

American species of the

An. maculipennis

complex(305–310 bp; Porter & Collins, 1991). Considerably longerITS2 sequences were found in other

Anopheles

speciescomplexes (Xu & Qu, 1997; Paskewitz

et al

., 1993;Beebe & Saul, 1995).

The GC content of the ITS2 region for the seven

Anopheles

taxa (52.5%) fell within the range of 50–60%of the anopheline and culicine species. Relatively higherGC contents (69%) were found in the

An. dirus

species Aand D (Xu & Qu, 1997).

Spacer pair-wise sequence differences among the

An.maculipennis

palearctic species (here and in the follow-ing estimates, all insertions/deletions were removed from

Figure 3. A maximum-likelihood consensus bootstrap tree based on rDNA ITS2 sequence data. The tree was rooted with An. quadrimaculatus species A. Numbers are bootstrap percentages (100 replications) for clades supported above the 50% level. Taxa are abbreviated as in Table 2.

IMB140.fm Page 474 Monday, October 18, 1999 9:56 AM

ITS2 sequence of palearctic

Anopheles maculipennis 475

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

alignments) ranged from 5.0% between An. maculipennisand An. messeae to 25.2% between An. atroparvus andAn. sacharovi. Lower spacer variation (2.7%) was foundbetween two sibling nearctic species of the An. maculi-pennis complex, An. hermsi and An. freeborni (Porter &Collins, 1991), which compares to 3.0% spacer variationbetween An. dirus species A and D (Xu & Qu, 1997).Between An. hermsi and the more distantly related An.occidentalis, ITS2 sequence variation amounted to 7.7%(Porter & Collins 1991). In other Anopheles species com-plexes, the interspecies differences ranged from 0.5 to1.6% between recently diverged members of the An.gambiae complex (Paskewitz et al., 1993) to 4.6–14.5%between members of the An. quadrimaculatus complex(Cornel et al., 1996).

Although the intraspecies ITS2 variability within thepalearctic members of the An. maculipennis complex is thesubject of further investigation, preliminary results indic-ated low levels of intraspecific ITS2 sequences variation(up to 0.35%) in An. maculipennis and An. labranchiae.This compares to intraspecies spacer variabilities reportedfor sibling species An. freeborni and An. hermsi (Porter& Collins, 1991), An. quadrimaculatus species A and D(Cornel et al., 1996), An. nuneztovari (Fritz et al., 1994), andsibling species An. gambiae and An. arabiensis (Paskewitzet al., 1993). Slightly higher levels (up to 1.36%) of spacervariation were detected in the An. messeae populationslisted in Table 1. There were differences at seven posi-tions, including one single-nucleotide mismatch and fiveindels (data not shown). Intrapopulation variability was not

Figure 4. Phylogenetic tree based on rDNA ITS2 sequence data and generated by maximum-parsimony method. The tree was rooted with An. quadrimaculatus species A. The phylogram has a branch length proportional to the number of inferred changes. The tree is the 50% majority rule consensus of 1000 bootstrap replicates with the branch-and-bound search option. Numbers are percentage bootstrap support for individual nodes. Tree length = 234 steps, with consistency index (CI) excluding uninformative characters = 0.761, homoplasy index (HI) excluding uninformative characters = 0.239, retention index (RI) = 0.800 and rescaled consistency index (RC) = 0.764. Taxa are abbreviated as in Table 2.

IMB140.fm Page 475 Monday, October 18, 1999 9:56 AM

476 M. Marinucci et al.

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

observed in An. maculipennis, An. labranchiae and An.messeae.

Internal transcribed spacer 2 inferred phylogeny

NJ analysis (Fig. 2) separated An. maculipennis nearcticand palearctic species as two distinct lineages, havingbootstrap supports of 100% and 99%, respectively. Thisarrangement is in concordance with a number of infer-ences (reviewed, for example, by Kitzmiller et al., 1967)indicating the genetic divergence of the two major line-ages within the An. maculipennis complex. Other stronglysupported nodes (> 90%) included individual cladeswithin both the nearctic and palearctic groups. In agree-ment with the whole existing information, An. occidentaliswas placed in a sister group relation to the strongly sup-ported (94%) An. hermsi–An. freeborni clade, the siblingspecies An. freeborni and An. hermsi being indistinguish-able by polytene chromosome banding patterns, enzymeloci, mitochondrial restriction fragment length polymorph-ism and morphology (Fritz et al., 1991). Based upon Sharp& Li (1989) substitution rate, Porter & Collins (1991)estimated divergence of An. hermsi and An. freeborni0.8 million years ago.

Evidence of phylogenetic affiliations within the nearcticclade concurs with prior morphologically based inferences.The nearctic taxa, with the exception of An. hermsi, haveevolved slight but consistent morphological differencesat all the life stages. In contrast, morphologically based,well-corroborated relationships among the palearctictaxa examined here are unavailable for comparisonpurposes: adults of sibling species An. atroparvus, An.labranchiae, An. maculipennis, An. melanoon and An.messeae are impossible to distinguish by morphologicalcharacters, and systematic work is chiefly based on eggmorphology.

There was strong support (100%) for the divergenceof An. sacharovi pre-dating the clade of the remainder ofthe palearctic species. Our phylogenetic analysis herestrongly supported the inference based on polytene chro-mosome banding patterns and other data (Stegnii, 1982)of the genetic divergence of An. martinius, placed in a sistergroup relation (100%) to the remainder of the palearcticgroup, and An. sacharovi. The strongly supported (99%)An. labranchiae–An. atroparvus clade was the sistergroup (91%) of a weakly supported (< 50%) clade includingAn. melanoon, An. maculipennis and An. messeae. Theinferred phylogeny of An. labranchiae and An. atroparvusis congruent with that based on polytene chromosomebanding patterns, hybridization studies and other data(reviewed by Kitzmiller et al., 1967), but the neighbour-joining analysis did not unambiguously resolve the genea-logy of the three An. melanoon, An. maculipennis and An.messeae individual lineages. An An. maculipennis–An.messeae clade was placed in a sister group relation to

An. melanoon by < 50% of bootstrap trees. Tree topolo-gies derived from maximum likelihood and maximumparsimony analyses (Figs 3 and 4) were identical to thatinferred from NJ analysis. Compared with NJ analysis,there was an attenuation of phylogenetic informationpertaining some genealogies, as indicated by a reduction(more consistent in ML analysis) of the bootstrap supportfor some nodes. In particular, the affiliations of An. maculi-pennis, An. melanoon and An. messeae were depictedas emanating from a trichotomus node in the bootstrapconsensus MP tree (Fig. 4), while a weakly supported(< 50%) An. maculipennis–An. melanoon clade was foundin the bootstrap consensus ML tree. Thus, ITS2-basedphylogenetic inferences within palearctic taxa are in goodagreement with the genealogies inferred by allozymeanalysis by Bullini & Coluzzi (1982) and do not conflictwith the relationships suggested by polytene chromo-some patterns and all other data. Whereas the close rela-tionship between An. atroparvus and An. labranchiae wasalways strongly supported by all the phylogenetic ana-lyses presented here, weakly supported or incongruentphylogenies of An. maculipennis, An. melanoon and An.messeae are inferred from ITS2 sequence analyses. Thispossibly reflects the differentiation of these species fromneighbouring taxa within a brief evolutionary time-framethat dispensed insufficient differences to support theseindividual lineages. Nevertheless, we cannot exclude thatadditional information from other nucleotide sequences(e.g. mitochondrial sequences and/or coding regions ofnuclear genes) could better resolve the relationshipsbetween the three species

Experimental procedures

Mosquito strains

The species, name abbreviation and geographical origin of themosquito field populations used in the present study and theInstitutions which provided them are listed in Table 1. Adultanophelines were collected by oral or battery-powered aspiratorsfrom animal shelters and brought to the laboratory. Blood-fedor gravid females were used to obtain eggs for species identifica-tion, using the keys of Angelucci (1955). The identified specimenswere frozen at −80 °C or dry stored.

Polymerase chain reaction and sequencing

Genomic DNA was extracted from individual adult mosquitoesaccording to the procedure of Coen et al. (1982). The rDNA ITS2regions were amplified by the PCR using primers based on con-served sequences of the 5.8S and 28S coding regions (Paskewitz& Collins, 1990; Collins & Paskewitz, 1996). The primer sequenceswere: 5.8S primer 5′-TGTGAACTGCAGGACACATGAAC-3′ and28S primer 5′-ATGCTTAAATTTAGGGGGTAGTC-3′. Amplificationwas performed with AmpliTaq polymerase in 50 ml total reactionvolume following GeneAmp kit recommendations (Perkin ElmerCetus). The PCR cycling programme was 5 min at 94 °C followed

IMB140.fm Page 476 Monday, October 18, 1999 9:56 AM

ITS2 sequence of palearctic Anopheles maculipennis 477

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

by thirty cycles of 0.5 min at 94 °C, 0.5 min at 50 °C and 1 minat 72 °C with 7 min at 72 °C after the last cycle. Cloning of thePCR products was performed using the vector and the competentcells supplied in TA Cloning Kit (Invitrogen) following the protocolsin the instruction manual.

Clones were sequenced in both forward and reverse directionby the dideoxy method using the material and protocols suppliedwith the T7 Sequencing Kit (Pharmacia Biotech) and [a35 S]dATP(Amersham). Compression zones were resolved with Deaza G/AT7 Sequencing mixes (Pharmacia Biotech).

Sequence analysis

Sequences are given in GenBank accession numbers Z50102-Z50105, Z83198, AJ224329, AJ224330. The homologoussequences of An. hermsi, An. freeborni, An. occidentalis (accessionnumbers M64482–M64484) and An. quadrimaculatus species A(accession number U32503) were published previously by Porter& Collins (1991) and Cornel et al. (1996), respectively.

Presumptive ITS2 sequences were aligned with CLUSTAL Wsoftware (version 1.7; Thompson et al., 1994) and manuallyadjusted if necessary. NJ and ML analyses were performedusing the programs NEIGHBOUR and DNAML of the PHYLIP pack-age (version 3.5c; Felsenstein, 1981), with bootstrap analyses(Felsenstein, 1985) performed using the SEQBOOT (1000 replicatesin NJ analysis, 100 replicates in ML analysis) and CONSENSE

programs in the same package. Estimates of Kimura’s two-parameter distances (Kimura, 1980) were determined with theDNADIST program of the PHYLIP package. MP analysis was carriedout with PAUP 3.1.1 (Swofford, 1993) using the exhaustive searchoption. A bootstrap consensus tree was determined using thebranch-and-bound search option (1000 replicates). Gaps wereexcluded from the analysis and characters were unweighted.NJ, ML and MP trees were rooted using the sequence of An.quadrimaculatus species A as the outgroup.

Acknowledgements

We thank Vladimir N. Stegnii, Tomsk State University,Russia, Adriana Sabatini and Guido Sabatinelli, IstitutoSuperiore di Sanità, for providing us with mosquito popu-lations and for information about the ecology and distribu-tion of these species.

References

Angelucci, A. (1955) Tavole sinottiche sugli anofelini italiani.ACIS. Monografia degli Annali di Sanità Pubblica 1.

Arnheim, N. (1983) Concerted evolution of multigene families,pp. 38–61. Evolution of Genes and Proteins (Nei, M. andKoehn, R.K., eds) Sinauer, Sunderland, MA.

Beckingham, K. (1982) Insect rDNA. The Cell Nucleus (Busch, H.and Rothblum, L., eds), pp. 205–269. Academic Press, NewYork.

Beebe, N.W. and Saul, A. (1995) Discrimination of all membersof the Anopheles punctulatus complex by polymerase chainreaction restriction fragment length polymorphism analysis.Am J Trop Med Hygiene 53: 478–481.

Bullini, L.A. and Coluzzi, M. (1982) Evolutionary and taxonomic

inference of electrophoretic studies in mosquitoes. RecentDevelopments in the Genetic of Insect Disease Vectors(Steiner, W.M., Tabachnick, W.J. Rai, K.S. and Narang, S.,eds), pp. 465–481. Stipes Publishers, Champaign, IL.

Coen, E.S., Thoday, J.M. and Dover, G.A. (1982) The rate ofturnover of structural variants in the ribosomal gene family ofDrosophila melanogaster. Nature 295: 564–568.

Collins, F.H. and Paskewitz, S.M. (1996) A review of the useof ribosomal DNA (rDNA) to differentiate among crypticAnopheles species. Insect Mol Biol 5: 1–9.

Cornel, A.J., Porter, C.H. and Collins, F.H. (1996) Polymerasechain reaction species diagnostic assay for Anophelesquadrimaculatus cryptic species (Diptera: Culicidae) basedon ribosomal DNA ITS2 sequences. J Med Entomol 33: 109–116.

Corradetti, A. (1934) Ricerche sugli incroci tra le varietà diAnopheles maculipennis. Riv Malariol 13: 707–720.

Dover, G. (1982a) A molecular drive through evolution. Bio-science 32: 526–533.

Dover, G. (1982b) Molecular drive: a cohesive mode of speciesevolution. Nature 299: 111–117.

Dover, G. and Coen, E. (1981) Springcleaning ribosomal DNA:a model for multigene evolution? Nature 290: 731–732.

Falleroni, D. (1926) Fauna anofelica italiana e suo ‘habitat’(paludi, risaie, canali). Metodi di lotta contro la malaria. RivMalariol 5: 553–593.

Felsenstein, J. (1981) Phylogeny Inference Package (PHYLIP),Version 3.5. University of Washington, Seattle.

Felsenstein, J. (1985) Confidence limits on phylogenies: anapproach using the bootstrap. Evolution 39: 783–791.

Fritz, G.N., Conn, J., Cockburn, A. and Seawright, J. (1994)Sequence analysis of the ribosomal DNA internal transcribedspacer 2 from populations of Anopheles nuneztovari (Diptera,Culicidae). Mol Biol Evol 11: 406–416.

Fritz, G.N., Narang, S.K., Kline, J.A., Seawright, J.A.,Washino, R.K., Porter, C.H. and Collins, F.H. (1991) Diagnosticcharacterization of Anopheles freeborni and An. hermsi byhybrid crosses frequencies of polytene X chromosomes andrDNA restriction enzyme fragments. J Am Mosquito ControlAssoc 7: 198–206.

Gerbi, S.A. (1985) Evolution of ribosomal DNA. Molecular Evolu-tionary Genetics (MacIntyre, R.J., ed.), pp. 419–517. Plenum,New York.

Hackett, L.W. and Missiroli, A. (1935) The varieties of Anophelesmaculipennis and their relation to the distribution of malaria inEurope. Riv Malariol 14: 45–109.

Hillis, D.M. and Dixon, M.T. (1991) Ribosomal DNA: molecularevolution and phylogenetic inference. Q Rev Biol 66: 411–453.

Kimura, M. (1980) A simple method for estimating evolutionaryrates of base substitution through comparative studies ofnucleotide sequences. J Mol Evol 16: 111–120.

Kitzmiller, J.B., Frizzi, G. and Baker, R. (1967) Evolutionand speciation within the maculipennis complex of thegenus Anopheles. Genetics of Insect Vectors of Disease(Wright, J.W. and Pal, R., eds), pp. 151–210. Elsevier Publishing,London.

Miller, B.R., Crabtree, M.B. and Savage, H.M. (1997) Phylo-genetic relationships of the Culicomorpha inferred from 18S and5.8S ribosomal DNA sequences (Diptera: Nematocera).Insect Mol Biol 6: 105–114.

IMB140.fm Page 477 Monday, October 18, 1999 9:56 AM

478 M. Marinucci et al.

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

Paskewitz, S.M. and Collins, F.H. (1990) Use of polymerasechain reaction to identify mosquito species of the Anophelesgambiae complex. Med Vet Entomol 4: 367–373.

Paskewitz, S.M., Wesson, D.M. and Collins, F.H. (1993) Theinternal transcribed spacers of ribosomal DNA in fivemembers of the An. gambiae species complex. Insect MolBiol 2: 247–257.

Porter, C.H. and Collins, F.H. (1991) Species-diagnostic differ-ences in a ribosomal DNA internal transcribed spacer fromthe sibling species Anopheles freeborni and Anopheleshermsi (Diptera, Culicidae). Am J Trop Med Hyg 45: 271–279.

Seperack, P., Slatkin, M. and Arnheim, N. (1988) Linkage dis-equilibrium in human ribosomal genes: implications formultigene family evolution. Genetics 119: 943–949.

Sharp, P.M. and Li, W.H. (1989) On the rate of DNA sequenceevolution in Drosophila. J Mol Evol 28: 398–402.

Shimada, T. (1992) Distribution of split 5.8S ribosomal RNA inDiptera. Insect Mol Biol 1: 45–48.

Stegnii, V.N. (1982) Genetic adaptation and speciation in siblingspecies of the Eurasian maculipennis complex. Recent Develop-ments in the Genetic of Insect Disease Vectors (Steiner, W.M.,

Tabachnick, W.J., Rai, K.S. and Narang, S., eds), pp. 454–464.Stipes Publishers, Champaign, IL.

Swofford, D.L. (1993) PAUP: phylogenetic analysis using parsimony,Version 3.11. Illinois Natural History Survey, Champaign, IL.

Tautz, D., Tautz, C., Webb, D. and Dover, G.A. (1987) Evolution-ary divergence of promoters and spacers in the rDNA familyof four Drosophila species: implications for molecular coevolu-tion in multigene families. J Mol Biol 195: 525–542.

Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL

W: improving the sensitivity of progressive multiple sequencealignment through sequence weighting, positions-specific gappenalties and weight matrix choice. Nucleic Acids Res 22:4673–4680.

White, G.B. (1976) The place of morphological studies in theinvestigation of the Anopheles species complexes. WorldHealth Organization, VBC/SC/76.14.

White, G.B. (1978) Systematic reappraisal of the Anophelesmaculipennis complex. Mosquito Syst 10: 13–44.

Xu, J.N. and Qu, F.Y. (1997) Ribosomal difference between spe-cies A and D of the Anopheles dirus complex of mosquitoesfrom China. Med Vet Entomol 11: 134–138.

IMB140.fm Page 478 Monday, October 18, 1999 9:56 AM

ITS2 sequence of palearctic Anopheles maculipennis 479

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

Annex 1. Alignment of rDNA ITS2 sequences. Insertions or deletions (indels) in alignment are denoted by a dash (–). Taxa are abbreviated as in Table 2. ITS2 sequences of An. hermsi, An. freeborni and An. occidentalis from Porter & Collins (1991); ITS2 sequence of An. quadrimaculatus species A from Cornel et al. (1996).

IMB140.fm Page 479 Monday, October 18, 1999 9:56 AM

480 M. Marinucci et al.

© 1999 Blackwell Science Ltd, Insect Molecular Biology, 8, 469–480

Annex 1. (continued).

IMB140.fm Page 480 Monday, October 18, 1999 9:56 AM