Blackwell Science, Ltd Secondary structure and conserved motifs of

Insect Molecular Biology (2000)

9

(6), 565–580

© 2000 Blackwell Science Ltd

565

Blackwell Science, Ltd

Secondary structure and conserved motifs of the frequently sequenced domains IV and V of the insect mitochondrial large subunit rRNA gene

T. R. Buckley,

1,2

C. Simon,

1,3

P. K. Flook

4

and B. Misof

5

1

Institute for Molecular Systematics, School of Biological Sciences, Victoria University of Wellington, Wellington, New Zealand,

2

Zoology Department, Duke University, Durham, NC, USA,

3

Department of Ecology and Evolutionary Biology, University of Connecticut, CT, USA,

4

Zoologisches Institut, Universität Basel, Switzerland, and

5

Institut fuer Evolutionsbiologie und Oekologie, Universitaet Bonn, Germany

Abstract

We have analysed over 400 partial insect mitochondriallarge subunit (mit LSU) sequences in order to identifyconserved motifs and secondary structures for domainsIV and V of this gene. Most of the secondary structureelements described by R. R. Gutell

et al

. (unpublished)for the LSU were identified. However, we present struc-tures for helices 84 and 91 that are not recognized inprevious universal models. The portion of the 16S genecontaining domains IV and V is frequently sequencedin insect molecular systematic studies so we havemany more sequences than previous studies whichfocused on the complete mitochondrial LSU molecule.In addition, we have the advantage of investigatingseveral sets of closely related taxa. Aligned sequencesfrom thirteen insect orders and nine secondary struc-ture diagrams are presented. These conserved sequencemotifs and their associated secondary structureelements can now be used to facilitate the alignmentof other insect mit LSU sequences.

Keywords: RNA secondary structure, alignment,molecular systematics, mitochondrial LSU rRNA, insectmitochondrial DNA.

Introduction

The identification of homologous characters is of funda-

mental importance in systematic studies. An alignmentof DNA sequences is a statement explicitly hypothesizinghomology among nucleotide character states (Mindell &Honeycutt, 1990). All methods of phylogenetic reconstructionassume that aligned nucleotides are homologous (Swofford

et al.

, 1996), and thus may produce incorrect phylogenetichypotheses if the aligned nucleotide sequences do not cor-rectly reflect homology (Kjer, 1995; Titus & Frost, 1996). In theanalysis of protein-coding genes the presence of conservedamino acids aids considerably in the alignment of nucleotidesequences (Doolittle, 1986). In addition to this, the codon-based nature of the genetic code makes the placement ofindels relatively straightforward in many cases. However, rRNAgenes are evolving under different constraints, and the align-ment of length variable sequences can be more problematic.Of particular concern is the need for the correct placementof ‘gaps’ or indels in the alignment, especially when theoverall degree of primary sequence conservation is low.Recent studies have shown that reference to a secondarystructure model can aid in this process (Kjer, 1995; Hickson

et al.

, 1996; Titus & Frost, 1996; Morrison & Ellis, 1997).The secondary structure of the large subunit (LSU) rRNA

gene has been the subject of intensive study by two researchgroups (de Rijk

et al.

, 1997; R. R. Gutell

et al

. unpublished).Although nuclear magnetic resonance and thermodynamicfolding studies have aided in the elucidation of rRNAsecondary structure, the covariation approach has beenparticularly useful in determining the size and relativeposition of secondary structure motifs (Gutell

et al.

, 1994).Covariation analysis operates on the premise that homo-logous rRNA genes with different nucleotide sequencesshould be able to be folded into similar secondary struc-tures. This is achieved by searching for base pairs thatcovary in tandem along a sequence when many differentsequences are compared. Evidence for the presence ofsecondary structural elements is obtained by searching forcompensatory substitutions on either side of a putativehelix (Gutell, 1996; Gutell & Damberger, 1996).

In insect molecular systematic, population genetic andconservation studies a portion of the 3

′

half of the mito-chondrial LSU rRNA (mit LSU) gene has been frequentlysequenced (Simon

et al.

, 1994; Kambhampati, 1995;

Received 4 February 2000; accepted 26 June 2000. Correspondence: ThomasBuckley, Zoology Department, Duke University, Durham, NC 27708–0325,USA. Tel.: 919–660–7431; fax: 919 6846168; e-mail: [email protected]

IMB220.fm Page 565 Tuesday, November 14, 2000 7:10 PM

566

T. R. B

uckley

et al.

© 2000 B

lackwell S

cience Ltd,

Insect Molecular B

iology

,

9

, 565–

580

Table 1.

Total number, taxonomic identity and source of sequences used in this study

Insect order Families included No. seqs Representative taxonAccession number Reference

Odonata Petaluridae, Gomphidae, Aeshnidae, Cordulegasteridae, Corduliidae,Libellulidae

23 Aeshna cyanea (dragonfly) AF268609 Misof

et al

. (2000)

Plecoptera Perlidae, Taeniopterygidae 2

Tacionemay

sp. (stonefly) P. K. Flook, unpublished

P. K. Flook (unpublished)

Dermaptera Spongiphoridae, Forficulidae 1

Circolabia pilicornis

(earwig) P. K. Flook, unpublished

P. K. Flook (unpublished)

Grylloblatodea Grylloblattidae 1 Grylloblata rothi (ice-crawler) Z93321 Flook & Rowell (1997a)

Phasmida Phyllidae, Pseudophasmatidae 2

Phyllium

bioculatum

(leaf insect) Z93319 Flook & Rowell (1997a)

Orthoptera Acrididae, Lentulidae, Pamphagidae, Pyrgomorphidae, Pneumoridae, Tanaoceridae, Xyronotidae, Trigonopterygidae, Proscopiidae, Euschmidtiidae, Thericleidae, Chorotypidae, Eruciidae, Eumastacidae, Tetrigidae, Batrachideidae, Tridactylidae, Rhipipterygidae, Tridactylidae, Cylindrachetidae, Stenopelmatidae, Mimnermidae, Rhaphidophoridae, Haglidae, Gryllidae

38

Ateliacris

annulicornis

(grasshopper) Z93291 Uhlenbusch

et al.

(1987); Flook

et al.

(1995); Flook

et al.

, 1999; Flook & Rowell (1997a, 1997b); P. K. Flook (unpublished)

Hemiptera Cicadidae, Cicadellidae 62

Unoka

sp. (leafhopper) L13015 Fang

et al

. (1993); T. R. Buckley

et al

. (unpublished); J. Tang

et al

. (unpublished)

Dictyoptera Panchloridae, Kalotermitidae, Termitidae, Rhinotermitidae, Termopsidae, Mastotermitidae, Blaberidae, Blattelidae, Blattidae, Cryptocercidae, Mantidae

48

Gromphadorhina

portentosa

(cockroach) Z97626 P. K. Flook, unpublished; Kambhampati (1995); Kambhampati

et al

. (1996)

Coleoptera Cicindelidae, Chrysomelidae 44

Cicindela dorsalis

(tiger beetle) L42911 Funk

et al

. (1995); Funk (1999); Vogler & Pearson (1996)

Diptera Simuliidae, Drosophilidae, Tephritidae, Otitidae, Platystomatidae, Culicidae

107

Delphinia picta

(fly) U39364 HsuChen

et al.

(1984); Clary & Wolstenholme (1985); DeSalle (1992); Xiong & Kocher (1991, 1993a, 1993b); Tang

et al

. (1995); Han & McPheron (1997); McPheron & Han (1997)

Mecoptera Panorpidae, Panorpodidae 32 – – Misof

et al

. (in press)

Lepidoptera Noctuidae, Lymantriidae, Nymphalidae, Satyridae, Papilionidae 2

Spodoptera

frugiperda

(moth) M76713 Pashley & Ke (1992); Davis

et al

. (1994)

Hymenoptera Pergidae, Cephidae, Orussidae, Evaniidae, Gasteruptiidae, Trigonalyidae, Megalyridae, Ceraphronidae, Megaspilidae, Pelecinidae, Proctotrupidae, Vanhorniidae, Roproniidae, Heloridae, Diapriidae, Scelionidae, Figitidae, Ibaliidae, Aphelinidae, Pteromalidae, Apidae, Ichneumonidae, Braconidae, Formicidae

126

Bombus

avinoviellus

(bumble bee) L22897 Vlasak

et al

. (1987); Cameron (1993); Crozier & Crozier (1993); Dowton & Austin (1994); Gimeno

et al.

(1997); Whitfield (1997); Dowton

et al.

(1998)

IMB

220.fm P

age 566 Tuesday, N

ovember 14, 2000 7:10 P

M

Insect mit LSU secondary structure

567

© 2000 Blackwell Science Ltd,

Insect Molecular Biology

,

9

, 565–580

Kambhampati

et al.

, 1996; Vogler & Pearson, 1996). Thisregion lies between the two highly conserved polymerasechain reaction (PCR) primers LR-N-13398 and LR-J-12887(Xiong & Kocher, 1991; Simon

et al.

, 1994), and correspondsto most of domains IV and V (Noller

et al.

, 1981) of the mitLSU gene. These primers, or modifications thereof, havealso been used to amplify a homologous portion of the mitLSU gene in other invertebrates (e.g. crabs: Cunningham

et al.

, 1992; spiders: Huber

et al.

, 1993; horseshoe crabs:Avise

et al.

, 1994; ticks: Black & Piesman, 1994), and awide variety of vertebrates (e.g. Orti

et al.

, 1996).Clary & Wolstenholme (1985) proposed a secondary

structure model for the complete mit LSU rRNA gene of

Drosophila yakuba

based on comparisons with the mousemit LSU secondary structure model (Glotz

et al.

, 1981).Although superseded by better models, the Clary &Wolstenholme (1985) model has been used by severalauthors to draw insect secondary structure diagrams and/orto align mit LSU rRNA sequences (e.g. Uhlenbusch

et al.

,1987; Fang

et al.

, 1993; Dowton & Austin, 1994; Flook &Rowell, 1997a). However, the considerably different modelsdescribed by R. R. Gutell

et al

. (unpublished, referred tohereafter as the Gutell model) and de Rijk

et al

. (1997;referred to hereafter as the DVD97 model) are based on amuch greater number of sequences and experimental andcovariation studies, and thus are likely to be more accurate.The de Rijk

et al

. (1997) structure is based on the modeldescribed by Noller

et al

. (1981) and Gutell & Woese (1990)and so is very similar to that of R. R. Gutell

et al

. (unpublished).Davis

et al

. (1994) published the complete mit LSUsecondary structure of the gypsy moth

Lymantria dispar

based on the

D. yakuba

and

Aedes albopticus

secondarystructure models from Gutell & Fox (1988). The Gutell &Fox (1988)

Drosophila melanogaster

mit LSU model hassince been updated by R. R. Gutell

et al

. (unpublished)and thus the Davis

et al

. (1994)

Lymantria

mit LSU modelrequires re-examination.

In this paper we report our analysis of over 400 partialinsect mit LSU rRNA sequences ranging from conspecificto interordinal taxonomic levels (Table 1). The alignment andsecondary structures presented are based on the analysisof sequences from a basal winged order (Odonata), sevenhemimetabolous orders (Orthoptera, Hemiptera, Dictyoptera,Dermaptera, Plecoptera, Gryloblattodea, Phasmatodea)and five more-derived holometabolous orders (Coleoptera,Lepidoptera, Mecoptera, Diptera and Hymenoptera). Inaddition, as in no other studies, we have focused onseveral data sets of closely related species. These include:(i) flies from the family Tephritidae (McPheron & Han, 1997);(ii) a large number of phylogenetically more diverse ortho-pterans (Flook & Rowell, 1997a, b; Flook

et al.

, 1999;unpublished data); and (iii)

Deltocephalus

-like leafhoppers(Fang

et al.

, 1993). Comparisons of sequences derived fromclosely related taxa (for example, the Tephritid sequences)

allow refinement of secondary structure elements in the morevariable regions and identification of variable sites withinconserved regions. Analysis of data from more divergenttaxa (e.g. many of the orthopterans) allows investigationof conserved nucleotide sites that are only free to vary overrelatively long periods of evolutionary time. Our analysis thusprovides new insight because the extensive LSU studiesof R. R. Gutell

et al

. (unpublished) and de Rijk

et al

. (1997)sample relatively few mitochondrial sequences and evenfewer of these are from insects despite the fact that insectsmake up the majority of animal biodiversity. Insects are poorlyrepresented in these other studies because few taxa havetheir mit LSU genes completely sequenced.

In this paper we present conserved sequence motifs andsecondary structures to aid in the alignment of domainsIV and V of the insect mit LSU gene. Hickson

et al

. (2000)demonstrate that the use of conserved structural motifs inalignment of ribosomal RNA (rRNA) molecules is crucialto the success of automated alignment programs becauseoptimal values for gap weight and gap length costs areunknowable

a priori

. The presentation of secondary struc-ture model diagrams from a phylogenetically diverse arrayof insects will facilitate the drawing of similar structures inother, as yet unsampled taxa.

Results and discussion

Secondary structure and alignment

The drawing of secondary structure diagrams is a tediousprocess. The existence of previous secondary models helpsbut the knowledge of conserved motifs speeds the processtremendously. The vast majority of papers on secondarystructure do not discuss conserved motifs other thanconserved G·U pairs (exceptions are Uhlenbusch

et al.

,1987; Hickson

et al.

, 1996).Secondary structure diagrams for representatives of eight

insect orders are shown in Figs 1 and 2. Representativesfrom a selection of insect orders are presented. An alignmentof representative insect taxa derived from these structuresis presented in Fig. 3, in addition to the sequence of

D. melanogaster

, with helices marked following the Gutellmodel. Most of the secondary structure motifs proposedin the Gutell model were identifiable in the insect sequences.However, the structures presented here differ from thatof the Gutell model for helices 84 and 91. We have illus-trated the contrasting structures for these two helices inFig. 4. Helix numbering follows that of Larsen (1992) andKambhampati

et al

. (1996).Conserved motifs present in the 400+ sequences are

described in Figs 1 and 3. These motifs can be used asanchors to construct preliminary sequence alignments.Putative helices can then be located by searching foruninterrupted base pairing in locations similar to thoseshown in Figs 1 and 3. Compensatory substitutions help


568

T. R. Buckley

et al.



,

9

, 565–580

Figure 1.

Secondary structure drawing of the 3

′

half of the

Spodoptera frugiperda

mit LSU gene. Boxed bases represent conserved motifs observed in at least 95% of all sequences we examined. Helices 65, 64, 61, 73, 90 and 93 have been drawn following the model of Gutell

et al

. (unpublished). Dominant G·U interactions are shaded. Canonical Watson–Crick interactions are represented by a dash. Noncanonical guanine–uracil interactions are represented by a dot and guanine–adenine interactions and uracil–uracil interactions by a hollow circle. The shaded U–A interaction in helix 71 is G·U in most other insect taxa. We have also indicated the proposed lone pair interaction between

E. coli

positions 1935 and 1962, between helices 69

′

and 67

′

, by a dashed line (see text for further details).



569



,

9

, 565–580

Figure 2.

Secondary structure models for eight representative insect mit LSU sequences. Helical regions where only one strand is available are drawn following the secondary structure model of R. R. Gutell

et al

. (unpublished).


570

T. R. Buckley

et al.



,

9

, 565–580

Figure 2.

(continued)



571



,

9

, 565–580

Figure 3.

Alignment of insect mit LSU rRNA sequences derived from secondary structure comparisons. Helices are identified by bold horizontal lines and numbered. Nucleotides involved in base pairing are highlighted. Vertical lines delimit adjacent helices. Spaces indicate missing data and indels are represented by dashes. With the exception of the first 3 bp, the remaining nucleotides of helix 75 are poorly conserved making homology assessment difficult. Additional areas of doubtful homology are the terminal loops of helices 66, 68, 81, 84 and 88, the bulge in helix 61 and the unpaired regions between helix 72

′

and 73, 81

′

and 84 and 84

′

and 88. These problematic areas have been indicated by an asterisk in the alignment.


572

T. R. Buckley

et al.



,

9

, 565–580

Figure 3.

(continued)



573



,

9

, 565–580

to verify paired regions. The alignment helps refine thesecondary structure and vice versa. Thus, alignment andsecondary structure prediction becomes an iterative pro-cess. Kjer (1995) describes this procedure more fully.

We have constructed tables that list helical substitutionsin the

Deltocephalus

-like leafhoppers, orthopterans, tephritidflies, and flies of the genus

Rhagoletis

. These tables areavailable from the Correspondent upon request.

Helix 66

The initial 5 bp of helix 66 form hydrogen bonds in mosttaxa. Although the terminal half may pair in many taxa,these interactions are less conserved. For example, in thedipteran

Delphinia picta

, it is possible to form a 10 bphelix (Fig. 2). However, hydrogen bonds in the terminal halfof the helix are broken in many related dipteran species.This helix is especially difficult to construct for some termitespecies. Accordingly, Kambhampati

et al

. (1996) leave itunpaired in some species. However, in the preying mantis,

Mantis religiosa

, it can be drawn as a 12 bp helix terminated

by a 4 bp loop (data not shown). However, we note that Kjer(who constructed the model described in Kambhampati

et al

. (1996) ) intentionally truncates helices that may formwhen there are taxa in the data set that contradict anotherwise possible extended helix, in favour of a moreuniversal minimum conserved structure (pers. comm.). Thefirst nucleotide pair in this helix often forms a U·U interaction,as has been observed in some helices of the insect SSUrRNA gene (Hickson

et al.

, 1996). Compensatory sub-stitutions among the leafhopper sequences, orthopteransand tephritid flies provide evidence for the existence ofthis helix.

Helix 67

Kambhampati

et al

. (1996) draw this helix as containingonly 3 bp. However, in almost all other insect taxa, helix67 is 4 bp long. In all of the termites, except

Mastotermesdarwiniensis

and

Incistitermes snyderi

, helix 67 can bedrawn as a 4 bp structure. Davis

et al

. (1994) includes thepreceding nucleotides from the region between 66

′

and

Figure 4. Alternative secondary structure models for helices 84 and 91. The fly sequence is that of Dephinia picta and the grasshopper is Ateliacris annulicornis.


574

T. R. Buckley

et al.



,

9

, 565–580

67, and 67

′

and 64′ as part of helix 67. However, evidencefor this interaction is lacking when other insect taxa areexamined. A conserved C–G bond in most taxa termin-ates helix 67. Compensatory substitutions are observedin the leafhoppers, and orthopterans.

Helix 68

This helix is typically initiated by three highly conservedcouplets (usually composed of Gs and Cs) followed by anasymmetric bulge (usually C·A). The sixth base pair is aG·U interaction that is conserved among all insects, withthe exception of two Simulium (Diptera) species from Tanget al. (1995) and several orthopterans from Flook & Rowell(1997a). The terminal couplets of helix 68 are not con-served, but can form canonical hydrogen bonds in mosttaxa. Kambhampati et al. (1996) only identify the first 3 bpin this helix. However, the G·U interaction at the sixthbase pair in the helix is conserved among all termite taxa.Other hydrogen bonds adjacent to this interaction are bro-ken in some of the termites; however, compensatory sub-stitutions among leafhoppers and Rhagoletis flies providesupport for their existence.

Davis et al. (1994) pair part of the asymmetric bulge inLymantria. This arrangement is not supported by com-pensatory substitutions in other insect taxa. The Lymantriasequence can be redrawn to fit the identical Gutell andDVD97 models. The loop terminating this helix is highlyvariable in nucleotide sequence and length, and difficult toalign, even among relatively closely related taxa.

Helix 69

This helix is highly conserved in length (6 bp) and primarysequence, and is terminated by a hairpin loop that is 6 bplong, except in some members of the Hymenoptera andOdonata. The second couplet involves a dominant G·Ubond. Compensatory substitutions among the orthopteransprovide evidence for its location. This model is drawnidentically in the Gutell and DVD97 models.

Helix 70

Helix 70 as drawn in the DVD97 model is a 1–2 bp longstructure in insects. However, we find no support in theform of compensatory substitutions for its existence, andbecause the proposed interactions are broken in some ofthe Hymenoptera. Gutell draws this region differently tothe DVD97 model. In the Gutell model this helix is formedfrom a single base pair interaction between the first nucle-otide downstream of helix 69′ and another nucleotide fourbases downstream. Because it is so highly conserved thisinteraction is difficult to evaluate with the data we havehere; however, we favour the Gutell structure for this regionbecause we have no data that contradict the Gutell model,and the interaction can form in almost all of the insectsequences we have analysed.

Helix 71

In almost all insect taxa this helix is 6 bp in length and isterminated with a dominant G·U bond. Compensatorysubstitutions occur in this helix among the leafhoppersand orthopterans. Kambhampati et al. (1996) have drawnthis helix 4 bp long, leaving the terminal 2 bp unpaired,because the fifth base pair cannot form in most of thetermites. Because this helix can form a 6 bp structure in mostother insect taxa and is well supported by evidence fromcompensatory substitutions, we have drawn it as paired.The fifth base pair is also well supported by compensatorysubstitutions in the orthopterans, and the leafhoppers.

Helix 64 and 61

Only the 61′ and 64′ halves of these two helices lie in theregion considered in this study. We have aligned this regionand constructed secondary structure drawings accordingto the Gutell model. Two conserved bulges described inthe Gutell model in helices 61′ and 64′ were detected inour alignment and we have indicated these in our secondarystructure diagrams. The complete helices are shown forthe moth Spodoptera frugiperda in Fig. 1.

Helix 72

This helix is 5 bp in all insect taxa. It is usually terminatedby a G–C bond and an 8-bp loop. In the DVD97 modelthis helix is drawn as 8 bp long, forming a bond betweenthe preceding two bases. This proposed interaction is notsupported by compensatory substitutions and cannot formin many of our insect mit LSUs. Thus, as in the Gutellmodel, we have drawn this helix as 5 bp.

Helix 73

The downstream half of helix 73 lies outside the regionsequenced in most insect systematic studies. However,among the taxa for which it has been sequenced, its sizeand position is highly conserved among insects and otherorganisms (R. R. Gutell et al. unpublished). As describedin the Gutell model, helix 73 contains two highly conservedsingle base pair bulges that we have indicated in ourstructures. This helix ends with five conserved bases in itsupstream half. A highly conserved unpaired region con-nects this helix to the next.

Helix 74

This helix is highly conserved in both sequence andsecondary structure. Two C–G bonds, conserved in mostother organisms (R. R. Gutell et al. unpublished) initiatethis helix. The terminal nucleotides of helix 74 are lessconserved and compensatory substitutions in this regionoccur in the orthopterans and leafhoppers.

Helix 75

In terms of secondary structure and alignment, helix 75 is


Insect mit LSU secondary structure 575

© 2000 Blackwell Science Ltd, Insect Molecular Biology, 9, 565–580

the most problematic region under consideration here. Itis highly variable in both length and secondary structure,with no apparent conserved motifs to aid in align-ment. Compensatory substitutions among closely relatedRhagoletis species suggest that base pairing is involved,but the lack of conservation of both sequence and helicallength makes alignment of even closely related taxadifficult. The high level of AT richness in insects makesalternative secondary structures possible, reflecting the easeof base pairing sequences composed largely of two com-patible nucleotides. In both the Gutell and DVD97 modelsthe first half of this region is drawn as paired, and the terminalhalf as a loop. The first three bonds are conserved in mosttaxa. A conserved A–U bond usually initiates the helix.

Due to difficulties in aligning sequences for this region,we have excluded it from the analyses of compensatorysubstitutions. The difficulty in aligning this region meansthat it is often excluded from further phylogenetic analysis(e.g. Flook & Rowell, 1997a). This region is also highlyvariable in other animal taxa (e.g. Orti et al., 1996).

Helix 80

Helix 80 is a conserved 4 bp structure, terminated by ahighly conserved 5 bp loop, composed of largely guaninesin almost all taxa. In most insects the final base pair formsa dominant G·U bond. Compensatory substitutions in thehelix are observed in the Orthoptera and the Lepidoptera.

Helix 81

Helix 81 is a 6 bp structure in most insects. It is terminatedby a length variable loop. A cytosine usually precedes 81′,but thymine is also observed, especially in the Hymenoptera.Compensatory substitutions occur in the Orthoptera andleafhoppers.

Helix 84

The single major difference between the Gutell, DVD97models and structures presented here lies in the regionbetween helix 81′ and 88. Like Kambhampati et al. (1996),we have drawn this region as a 3 bp helix. Accordingly, wehave numbered it helix 84 as in Kambhampati et al. (1996).In Fig. 4 we present the contrasting secondary structuremodels for this region for our structures, and the Gutell andDVD97 models. This helix is preceded by a length- andsequence-variable stretch of bases. We have been unableto derive a secondary structure for the region betweenhelices 81 and 84, thus have left it unpaired. In our struc-tures, helix 84 is terminated by a sequence- and length-variable loop. In some taxa the terminal couplet is broken;however, compensatory substitutions are observed in theorthopterans, hymenopterans and cicadas. The first basepair in helix 84 is highly conserved; however, compensatorysubstitutions are observed in the termites (Kambhampatiet al., 1996; Fig. 3), hymenopterans and cicadas.

Helix 88

Helix 88 can be constructed in most taxa. However, in thetermites, leafhoppers and many of the Hymenoptera, forexample, it is difficult to draw. For this reason Kambhampatiet al. (1996) leave the region between 84′ and 74′ unpaired.Compensatory substitutions among Tephritidae fliesand the orthopterans provide evidence for its presence.A conserved C–G bond initiates this helix, althoughcompensatory substitutions can occur at this position,especially in the Hymenoptera.

Helix 89

In the DVD97 model an interaction between the fourthnucleotide of 89 and the fifth nucleotide of 89′ is proposed.Compensatory substitutions do not support this interaction,and as in the Gutell model we leave the fourth nucleotideunpaired. The extreme rarity of indels in this helix makesalignment straightforward despite the lack of sequenceconservation across all insects.

Helix 90

Most of the 90′ side of this helix is excluded from the regionconsidered here, as it contains the PCR primer bindingsite for LR-J-12887 (Simon et al., 1994). Helix 90 is highlyconserved in length and primary sequence among insecttaxa with few indels observed. In our structures and theassociated alignment, we have drawn it as in the Gutellmodel.

Helix 91

The initial nucleotides of this helix are highly variablemaking a stable secondary structure difficult to derive. Inmost insect taxa this helix is formed by ten nucleotidecouplets. Indels can occur in the bulge near the start ofthe helix. Support from compensatory substitutions for thesecond nucleotide couplet is moderate. In the DVD97model this area is constructed differently in various insecttaxa. For example, a bulge is placed after the fifth basepair in the dipteran (Fig. 4), hymenopteran and lepidop-teran structure. Consistent with the Gutell model we donot recognize this bulge. In the Gutell model this helix hasa 1–2 bp bulge near the start of the helix, whereas in ourstructure this bulge is shorter. The structure presentedhere is conserved across a wide range of insect taxa,although some interactions are broken, as in the leafinsect (see Fig. 3). Compensatory substitutions among theOrthoptera and leafhoppers support this structure. Thepenultimate couplet often involves a G·A interaction asdescribed in the Gutell model. This interaction is notrecognized in the DVD97 model despite the fact that itforms a canonical bond in many insect taxa.

Helix 92

As this helix lies close to the priming site of LR-J-12887, it


576 T. R. Buckley et al.


is often excluded from published sequences. This 5 bphelix is terminated with a highly conserved UCUUG loop.Compensatory substitutions occur in the Orthoptera andleafhoppers. Deletions occasionally occur in the proximaland distal arms of this helix (e.g. Delphinia picta, Fig. 2).

Lone pairs

Lone-pair or tertiary interactions are base pairs that are notimmediately adjacent to other base pairs (Larsen, 1992;Gutell & Damberger, 1996). The Gutell model describes anumber of such interactions, supported by covariation ana-lysis and in some cases by experimental evidence (e.g.Mitchell et al., 1990). In the Gutell model, three lone-pairinteractions can be evaluated using the data presentedhere. Two occur between helices 81′ and 73′, and theother between 68′ and 67′. The first interaction is betweennucleotides 2282 and 2427 of the Escherichia coli sequence.We find no support in the form of covarying substitutionsin insect mtDNA for this proposed interaction despite thefact that substitutions are common in this region and theinteraction is supported by crosslinking experiments (Mitchellet al., 1990). This interaction is unable to form in many of thetiger beetles (Vogler & Pearson, 1996), odonates (unpub-lished data) and cockroaches (Kambhampati, 1995) forexample.

In the Gutell model the third base of helix 84 is includedin a lone-pair interaction with another base between helices88′ and 74′. We find no support for this lone pair amonginsect taxa. In addition to this, the inclusion of this base inhelix 84 precludes its inclusion in a lone pair.

The Gutell model describes a lone-pair interaction betweenEscherichia coli positions 1935 and 1962. Among insectsthese two positions are highly conserved, making thisinteraction difficult to evaluate. However, an A–U → G·Usubstitution has occurred in several of the cicindelidbeetles from Vogler & Pearson (1996) and the OrthopteranPneumora inanis from Flook & Rowell (1997a). We havemarked this interaction on Fig. 1.

Dominant G·U bonds and other non-canonical interactions

Gautheret et al. (1995) emphasized the functional import-ance of conserved G·U pairs in all rRNA sequences. Thesemost frequent of non-Watson–Crick base pairs changehelix stacking properties and are often found in tandem orwith the U located 5′ of single base bulges. They tend tobe found in specific locations and can be either unvaried(found in 100% of taxa) or dominant (found in greater than55% of taxa). In the insect mit LSU gene we have detectedseven dominant G·U bonds in helices 68, 69, 71, 80 and89. No unvaried G·U pairs were found. We have markedthe dominant interactions on the Spodoptera structure inFig. 1. The highly conserved nature of such bonds impliessome of these G·U interactions may have a functional

significance and not represent intermediate characterstates between conventional Watson–Crick bonds.

Hickson et al. (1996) observed that C·A and A·C bondswere relatively common in some helices of the avian mitSSU. However, they also observed that such bonds wereless common in the odonate mit SSU sequences. Sim-ilarly, we have observed that C·A and A·C interactions arerare in the insect mit LSU sequences. We only detectedone G·A bond, terminating helix 91 (see above).

Comparison of models

Over 15 years of study of the LSU rRNA gene (beginningwith Noller et al., 1981) has facilitated the development ofa stable and robust secondary structure model. Thus, mostof the secondary structure elements proposed by the Gutelland DVD97 models have been proven to exist in the insectmit LSU rRNA gene. At the same time, several aspects ofthe evolution of these rRNA secondary structures, such asdeviations from common patterns in specific evolutionarylineages, remain poorly understood. Understanding struc-tural and functional shifts that may occur in moleculesemphasizes the need for new phylogenetic models thattake into account such covariotide shifts (Lockhart et al.,1998; Philippé & Laurent, 1998). Our detailed studies ofspecific evolutionary lineages such as insect mtDNA sug-gest that there is a continuing need to refine the model ofthe mit LSU. This is clearly demonstrated by the insect mitLSU sequences, which differ from sequences from manyother taxonomic groups examined in at least two respects.The first of these concerns the relationship between thefrequency of different canonical pairings and their thermo-dynamic stability, which has previously been reported (onthe basis of comparison of a large sample of LSU rRNAsequences) to be positively correlated, i.e. G–C > C–G >U–A > A–U > G·U > U·G (Gutell & Damberger, 1996). Inthe insect data, this pattern is not observed. Instead, A–Uand U–A pairs are twice as common as G–C and C–Gpairings. The trend towards A–U/U–A interactions probablyreflects the selective pressure to maintain AT richness ininsect mitochondrial DNA (Simon et al., 1994).

A second difference between the mit LSU secondarystructure of insects and other groups concerns the occur-rence of tetraloops (i.e. hairpin loops of length 4 bp). Ther-modynamic studies have indicated that these structuresmay be particularly stable. In the E. coli LSU structure ofR. R. Gutell et al. (unpublished), for the region under con-sideration here, the ratio of tetraloops to the total numberof hairpin loops is 5 : 17 (29%). The same ratio for thenematode Caenorhabditis elegans mit LSU is 1 : 10 (10%),Homo sapiens 3 : 13 (23%), the cow Bos taurus 3 : 13 (23%),the frog Xenopus laevis 2 : 15 (13%). The Spodopterafrugiperda structure presented here has a ratio of 1 : 13(8%). In the partial mitochondrial SSU model of Hicksonet al. (1996) there are no tetraloops conserved across the




metazoa. In the prokaryote LSU, two examples of conservedtetraloops are found terminating helices 66 and 68. In theinsect mit LSU, neither of these two tetraloops are conserved.Both these helices are truncated relative to the prokaryoteLSU and appear to have lost their terminal tetraloops.

Fields & Gutell (1996) observed that of the 256 possiblecombinations of four nucleotides, the motifs GHGA, GVAAand UWCG were the most abundant in the LSU rRNAgene. By contrast, in the structures presented here, wedetected only one conserved tetraloop, not conforming toany of the motifs described by Fields & Gutell (1996), ter-minating helix 91 (also recognized in the Gutell model).This loss of tetraloops in the insect mit LSU gene mayactually reflect different functional constraints relative toother genomes. Fields & Gutell (1996) have previously notedthat the mitochondrial genome has the lowest proportionof tetraloops conforming to the above motifs relative to theArcheal, Eubacterial, Chloroplast or Eucaryal genomes.Konnings & Gutell (1995) made a similar observation forthe SSU rRNA gene.

Based on these comparisons we therefore agree withGutell et al. (1994) that the LSU rRNA core secondarystructure is now highly refined and will probably need fewfurther revisions. The slight differences between the Gutellmodel and the structures presented here are not in them-selves sufficient basis to challenge the universality of theGutell model, and may in part reflect general differencesin evolutionary patterns between insect mtDNA and DNAfrom other taxonomic groups and genomes (see Simon et al.,1994). However, they do emphasize the lack of understand-ing concerning the precise nature of constraints imposedon sequence evolution by secondary structure. In addition,the small number of differences between the structures pre-sented here and the Gutell model may reflect mitochondrialspecific or even insect mitochondrial specific changes inrRNA structure. Support from compensatory substitutions inother taxonomic groups for the refined structures we presenthere should be assessed.

Secondary structure and alignment

The alignment of rRNA sequences using secondary struc-ture can be a time-consuming task. The molecular system-atist may well question the value of such an approach giventhe availability and widespread use of multiple sequencealignment programs (e.g. Higgins et al., 1992; Wheeler &Gladstein, 1992). During the analysis of the sequencesutilized in this investigation, it was noted that many datasets aligned using multiple sequence alignment programshad incorrectly placed gaps in length conserved helices,according to the model described here. In an analysis ofOplurid lizard mit rRNA sequences, Titus & Frost (1996)observed that constraining homologous helices to alignwith each other led to an increase in the number ofalignments that recovered a well-supported tree based on

morphological characters. It has been shown that all multiplesequence alignment programs failed to correctly align theconserved motifs described in the Hickson et al. (1996)secondary structure model for domain III of the mit SSUrRNA gene (Hickson et al., 2000). Even among closelyrelated taxa, secondary structure can be essential for align-ment of hypervariable sites in length variable regions. Forexample, within the alignment produced by McPheron & Han(1997) for their Rhagoletis sequences, gaps were insertedin the terminal half of helix 67 that disrupted the highlyconserved G·U interaction described above. The Rhagoletissequences can be realigned to accommodate this bond.

The process for obtaining an alignment derived froma secondary structure model using conserved motifs asanchors has been well described by Kjer (1995). However,the use of a secondary structure in aligning rRNA sequencesdoes not guarantee obtaining the correct alignment forevery base. This is especially true in length and sequencevariable loops.

It is essential that an up-to-date secondary structure modelbe used to derive an alignment. The most recent LSUsecondary structure models can be obtained from the RNAWWW server (de Rijk et al., 1997; http://rrna.uia.ac.be), theRNA Secondary Structure WWW home page (R. R. Gutellet al., unpublished; http://www.rna.icmb.utexas.edu/) andthe Ribosome Database Project WWW home page (Maidaket al., 1997; http://www.cme.msu.edu/RDP). The abovestudies involve large numbers of sequences from aphylogenetically wider array of taxa and genomes thandescribed here. However, studies involving relatively closelyrelated taxa such as in Hickson et al. (1996) and the currentinvestigation are useful in refining the more variable regionswithin genomes.

Experimental procedures

The insect mit LSU sequences used in this study are given inTable 1. This compilation includes sequences obtained fromGenBank, EMBL and unpublished sources. Initially secondarystructure diagrams were constructed by the comparative method,using conserved motifs as anchors, for one representative ofeach available insect order following the D. melanogaster draw-ing of R. R. Gutell et al. (unpublished). Within each order, substi-tutions that indicate paired interactions were identified using theprogram SMAPP (written by PKF) and then examined on the second-ary structure model. SMAPP searches a given alignment of rRNAsequences for Watson–Crick interactions and non-Watson–Crick(non-canonical) hydrogen bonds, including C·A, G·A and G·Uinteractions. We define a compensatory change as two substitu-tions occurring sequentially that maintain base pairing in a givenposition in a helix. The observation that two or more Watson–Crick(or G·U interactions) at the same location in a putative helixindicates that there is selection to maintain base pairing andthus support for the helical model. We used SMAPP to analysethirty-eight orthopteran (Flook & Rowell, 1997a; 1997b; Flooket al., 1999; unpublished data), thirty-seven Tephritid fly (Han




& McPheron, 1997; McPheron & Han, 1997) and thirty-twoDeltocephalus-like leafhopper (Fang et al., 1993) sequences. TheC·A interactions were included in our search because they havebeen shown to be common in the mitochondrial small subunit(SSU) rRNA of some animal taxa (Hickson et al., 1996). We alsosearched for G·U interactions, as they are also common in rRNAsequences (Gutell, 1996). Hickson et al. (1996) have shown thatseveral ‘dominant’ (Gutell et al., 1994) G·U interactions, conservedin greater than 55% of known SSU sequences are also found inthe animal mitochondrial SSU gene. Thus, we searched for suchdominant G·U interactions in the insect mit LSU gene. We alsosearched the entire data set for bases that were unvaried in atleast 95% of all sequences to identify conserved primarysequence motifs. Helix numbering follows that of Larsen (1992)and Kambhampati et al. (1996). The different domains of the LSUgene are numbered following Noller et al. (1981). The alignmentof representative sequences in Fig. 3 was constructed by aligninglength-conserved helices and conserved motifs in helices, loopsand unpaired regions.

Acknowledgements

We thank Mary Jane Spring for preparing the secondarystructure drawings. Bruce McPheron, Mark Dowton, JimWhitfield and Hervé Philippé provided us with sequences,GenBank accession numbers and/or unpublished manu-scripts. Robin Gutell kindly supplied us with unpublishedarthropod secondary structure drawings and commentson the manuscript. T.R.B. thanks Geoffrey Chambers forlaboratory accommodation, general support and commentson the manuscript. Lin Field, Karl Kjer, and an anonymousreviewer provided comments on a previous version of themanuscript. Liz MacAvoy and Lesley Milicich also aidedT.R.B. with laboratory work and other logistics. B.M. thanksK.P. Sauer (Institute of Evolutionary Biology and Ecology,University of Bonn) for access to laboratory facilities.Funding for this study was provided in part by the Sci-ence Faculty Leave and Grants Committee and the Schoolof Biological Sciences, Victoria University of Wellington(T.R.B.), The National Geographic Society (T.R.B. andC.S.), The University of Connecticut (C.S.), The NationalScience Foundation (C.S.), The Fulbright Foundation (C.S.),The Swiss National Science Foundation (P.K.F.) andThe Austrian Science Foundation (B.M.). B.M. was alsosupported by a Schroedinger Post Doctoral Fellowship.

References

Avise, J.C., Nelson, W.S. and Sugita, H. (1994) A speciationalhistory of ‘living fossils’: molecular evolutionary patterns inhorseshoe crabs. Evolution 48: 1986–2001.

Black, W.C. and Piesman, J. (1994) Phylogeny of hard- and soft-tick taxa (Acari: Ixodida) based on mitochondrial 16S rDNAsequences. Proc Natl Acad Sci USA 91: 10 034–10 038.

Cameron, S.A. (1993) Multiple origins of advanced eusociality inbees inferred from mitochondrial DNA sequences. Proc NatlAcad Sci USA 90: 8687–8691.

Clary, D.O. and Wolstenholme, D.R. (1985) The ribosomal RNAgenes of Drosophila DNA. Nucleic Acids Res 13: 4029–4045.

Crozier, R.H. and Crozier, Y.C. (1993) The mitochondrial genomeof the honeybee Apis mellifera: complete sequence and genomeorganisation. Genetics 133: 97–117.

Cunningham, C.W., Blackstone, N.W. and Buss, L.W. (1992)Evolution of king crabs from hermit crab ancestors. Nature355: 539–542.

Davis, R.E., Kelly, T.J., Masler, E.P., Thyagaraja, B.S., Rote, C.A.and Imberski, R.B. (1994) Complete base sequence for themitochondrial large subunit ribosomal RNA of the gypsy mothLymantria dispar (L.). Insect Mol Biol 3: 219–228.

de Rijk, P., Van De Peer, Y. and de Wachter, R. (1997) Databaseon the structure of the large ribosomal RNA. Nucleic AcidsRes 25: 117–123.

DeSalle, R. (1992) The phylogenetic relationships of flies in thefamily Drosophilidae deduced from mtDNA sequences. MolPhylogenet Evol 1: 31–40.

Doolittle, R.F. (1986) ORF’s and URF’s University ScienceBooks, Mill Valley, CA.

Dowton, M. and Austin, A.D. (1994) Molecular phylogeny of theinsect order Hymenoptera: apocritan relationships. Proc NatlAcad Sci USA 91: 9911–9915.

Dowton, M., Austin, A.D. and Antolin, M.F. (1998) Evolutionaryrelationships among the Braconidae (Hymenoptera: Ichneu-monoidea) inferred from partial 16S rDNA gene sequences.Insect Mol Biol 7: 129–150.

Fang, Q., Black, I.V.W.C., Blocker, H.D. and Whitcomb, R.F. (1993)A phylogeny of New World Deltocephalus-like leafhopper generabased on mitochondrial 16S ribosomal DNA sequences. MolPhylogenet Evol 2: 119–131.

Fields, D.S. and Gutell, R.R. (1996) An analysis of large rRNAsequences folded by a thermodynamic method. Folding Design1: 419–430.

Flook, P.K., Klee, S. and Rowell, C.H.F. (1999) A combinedmolecular phylogenetic analysis of the Orthoptera (Arthropoda:Insecta) and its implications for their higher systematics. SystBiol 48: 233–253.

Flook, P.K. and Rowell, C.H.F. (1997a) The effectiveness ofmitochondrial rRNA gene sequences for the reconstructionof the phylogeny of an insect order (Orthoptera). MolecPhylogenet Evol 8: 177–192.

Flook, P.K. and Rowell, C.H.F. (1997b) The phylogeny of theCaelifera (Insecta, Orthoptera) as deduced from mtrRNA genesequences. Mol Phylogenet Evol 8: 89–103.

Flook, P.K., Rowell, C.H.F. and Gellissen, G. (1995) The sequence,organization and evolution of the Locusta migratoria genome.J Mol Evol 41: 928–941.

Funk, D. (1999) Molecular systematics of cytochrome oxidase Iand 16S from Neochlamisus leaf bettles and the importanceof sampling. Mol Biol Evol 16: 67–82.

Funk, D.J., Futuyma, D.J., Ortí, G. and Meyer, A. (1995) Mito-chondrial DNA sequences and multiple data sets. a phylogeneticstudy of phytophagous beetles (Chrysomelidae: Ophraella).Mol Biol Evol 12: 627–640.

Gautheret, D., Konings, D. and Gutell, R.R. (1995) G·U basepairing motifs in ribosomal RNA. RNA. 1, 807–814.

Gimeno, C., Belshaw, R. and Quicke, D.L.J. (1997) Phylogeneticrelationships of the Alysiinae/Opiinae (Hymenoptera: Bra-conidae) and the utility of Cytochrome b mit LSU and 28S,D2 rRNA. Insect Mol Biol 6: 273–284.




Glotz, C., Zweib, C. and Brimacombe, R. (1981) Secondary struc-ture of the large subunit ribosomal RNA from Escherichia coli,Zea mays and human and mouse mitochondrial ribosomes.Nucleic Acids Res 9: 3287–3306.

Gutell, R.R. (1996) Comparative sequence analysis and thestructure of mit, LSU, 23S rRNA. Ribosomal RNA: Structure,Evolution, Processing, and Function in Protein Biosynthesis(Zimmermann, R.A. & Dahlberg, A.E., eds), pp. 111–128CRC Press, Boca Raton, FL.

Gutell, R.R. and Damberger, S.H. (1996) Comparative sequenceanalysis of experiments performed during evolution. RibosomalRNA and Group I Introns (Green, R. and Schroeder, R., eds),pp. 15–33. R. G. Landes Company Austin, Texas.

Gutell, R.R. & Fox, G.E. (1988) A compilation of large subunitRNA sequences presented in a structural format. NucleicAcids Res 16: 175–269.

Gutell, R.R. and Woese, C.R. (1990) Higher order structuralelements in ribosomal RNAs: pseudo-knots and the use ofnon-canonical pairs. Proc Natl Acad Sci USA 87: 663–667.

Gutell, R.R., Larsen, N. and Woese, C.R. (1994) Lessons froman evolving rRNA mit LSU and 23S structures from a com-parative perspective. Microbiol Rev 58: 10–26.

Gutell, R.R., Subashchandran, S., Schnare, M., Du, Y., Lin, N.,Madabusi, L., Muller, K., Pande, N., Yu, N., Shang, Z., Date, S.,Konings, D., Schweiker, V., Weiser, B. and Cannone, J.J.(unpublished) Comparative sequence anal prediction RNAstructure and the web. <http://www.rna.icmb-utexas.edu/>.

Han, Y. and McPheron, B.A. (1997) Molecular phylogenetic studyof Tephritidae (Insecta: Diptera) using partial sequences ofthe mitochondrial mit LSU ribosomal DNA. Mol PhylogenetEvol 7: 17–32.

Hickson, R.E., Simon, C., Cooper, A., Spicer, G.S., Sullivan, J.and Penny, D. (1996) Conserved sequence motifs, alignment,and secondary structure for the third domain of animal 12SrRNA. Mol Biol Evol 13: 150–169.

Hickson, R.E., Simon, C. and Perrey, S.W. (2000) The perform-ance of several multiple-alignment programs in relation tosecondary-structure features for an rRNA sequence. Mol BiolEvol 17: 530–539.

Higgins, D.G., Bleasby, A.J. and Fuchs, R. (1992) CLUSTAL V:improved software for multiple sequence alignment. ComputAppl Biosci 8: 189–191.

HsuChen, C.C., Kotin, R.R. and Dubin, D.T. (1984) Sequences forthe coding and flanking region of the large ribosomal subunitRNA of mosquito mitochondria. Nucleic Acids Res 12: 7771–7785.

Huber, K., Haider, T.S., Müller, M.W., Huber, B.A., Schweyen, R.J.and Barth, F.G. (1993) DNA sequence data indicates polyphylyof the family Ctenidae (Areneae). J Arachnol 21: 194–201.

Kambhampati, S. (1995) A phylogeny of cockroaches and relatedinsects based on DNA sequence of mitochondrial ribosomalRNA genes. Proc Natl Acad Sci 92: 2017–2020.

Kambhampati, S., Kjer, K.M. and Thorne, B.L. (1996) Phylogeneticrelationships among termite families based on DNA sequenceof mitochondrial mit LSU ribosomal RNA gene. Insect MolBiol 5: 229–238.

Kjer, K.M. (1995) Use of rRNA secondary structure in phylo-genetic studies to identify homologous positions: an example ofalignment and data presentation from the frogs. Mol PhylogenetEvol 4: 314–330.

Konnings, D.A.M. and Gutell, R.R. (1995) A comparison of ther-modynamic foldings with comparatively derived structures of16S and 16S-like rRNAs. RNA 1: 559–574.

Larsen, N. (1992) Higher level interactions in 23S rRNA. ProcNatl Acad Sci USA 89: 5044–5048.

Lockhart, P.J., Steel, M.A., Barbrook, A.C., Huson, D.H.,Charleston, M.A. and Howe, C.J. (1998) A covariotide modelexplains apparent phylogenetic structure of oxygenic photo-synthetic lineages. Mol Biol Evol 15: 1183–1188.

Maidak, B.L., Olsen, G.J., Larsen, N., Overbeek, R., Caughley, M.J.and Woese, C.R. (1997) The RDP (Ribosomal DatabaseProject). Nucl Acids Res 25: 109–111.

Mindell, D.P. and Honeycutt, R.L. (1990) Ribosomal RNA invertebrates: evolution and phylogenetic applications. AnnuRev Ecol Syst 21: 541–566.

Mitchell, P., Osswald, M., Schueler, D. and Brimacombe, R.(1990) Selective isolation and detailed analysis of intra-RNAcross-links induced in the large ribosomal subunit of E. coli: amodel for the tertiary structure of the tRNA binding domain in23S RNA. Nucleic Acids Res 18: 4325–4333.

McPheron, B.A. and Han, H. (1997) Phylogenetic analysis of NorthAmerican Rhogoletis (Diptera: Tephritidae) and related generausing mitochondrial DNA sequence data. Mol PhylogenetEvol 7: 1–16.

Misof, B., Erpenbeck, D. and Sauer, K.P. (in press) Mitochondrialgene fragments suggest paraphyly of the genus Panorpa(Mecoptera, Panorpidae). Mol Phyl Evol in press:.

Misof, B., A.M. Rickert, T.R., Buckley, G., Fleck, K.P. Sauer. (in press)Phylogenetic signal, its decay in mitochondrial SSU and LSUrRNA gene fragments of dragonflies. Molecular Biology andEvolution in press.

Morrison, D.A. and Ellis, J.T. (1997) Effects of nucleotide sequencealignment on phylogeny estimation: a case study of 18S rDNAsof Apicomplexa. Mol Biol Evol 14: 428–441.

Noller, H.F., Kop, J., Wheaton, V., Brosius, J., Gutell, R.R.,Kopylov, A.M., Dohme, F., Herr, W., Stahl, D.A., Gupta, R.and Woese, C.R. (1981) Secondary structure model for 23Sribosomal RNA. Nucl Acids Res 9: 6167–6189.

Orti, G., Petry, P., Porto, J.I.R., Jegu, M. and Myer, A. (1996)Patterns of nucleotide change in mitochondrial ribosomalRNA genes and the phylogeny of piranhas. J Mol Evol 42:169–182.

Pashley, D.P. and Ke, L.D. (1992) Sequence evolution in mitochon-drial ribosomal and ND-1 genes in Lepidoptera: Implicationsfor phylogenetic analyses. Mol Biol Evol 9: 1061–1075.

Philippé, H. and Laurent, J. (1998) How good are deep phylogenetictrees? Curr Opinion Genet Dev 8: 616–623.

Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. andFlook, P. (1994) Evolution, weighting, and phylogenetic utilityof mitochondrial gene sequences and a compilation ofconserved polymerase chain reaction primers. Ann EntomolSoc Am 87: 651–701.

Swofford, D.L., Olsen, G.J., Waddell, P.J. and Hillis, D.M. (1996)Phylogenetic inference. Molecular Systematics (Hillis, D.M.,Moritz, C. and Mabee, P., eds), pp. 407–514. Sinauer,Sunderland, MA.

Tang, J., Toe, L., Back, C., Zimmerman, P.A., Pruess, K. andUnnasch, T.R. (1995) The Simulium damnosum speciescomplex: phylogenetic analysis molecular identification basedupon mitochondrially encoded gene sequences. Insect MolBiol 4: 79–88.

Titus, T.A. and Frost, D.R. (1996) Molecular homology assessmentand phylogeny in the lizard family Opluridae (Squamata: Iguania).Mol Phylogenet Evol 6: 49–62.

Uhlenbusch, I., McCracken, A. and Gellissen, G. (1987) The gene




for the large (mit LSU) ribosomal RNA from the Locusta. CurrGenet 11: 631–638.

Vlasak, I., Burgschwaiger, S. and Kreil, G. (1987) Nucleotidesequence of the large ribosomal RNA of honeybee mitochon-dria. Nucleic Acids Res 15: 2388.

Vogler, A.P. and Pearson, D.L. (1996) A molecular phylogeny ofthe tiger beetles (Cicindelidae): congruence of mitochondrialand nuclear rDNA data sets. Mol Phylogenet Evol 6: 321–338.

Wheeler, W.C. and Gladstein, D.L. (1992) A heuristic multiplesequence alignment procedure based on parsimony. Computerprogram. American Museum of Natural History, New York.

Whitfield, J.B. (1997) Molecular and morphological data suggest

a single origin of the polydnaviruses among braconid wasps.Naturwissenschaften 84: 502–507.

Xiong, B. and Kocher, T.D. (1991) Comparison of mitochondrialDNA sequences of seven morphospecies of black flies(Diptera: Simullidae). Genome 34: 306–311.

Xiong, B. and Kocher, T.D. (1993a) Intraspecific variation in siblingspecies of Simulium venustrum and Simulium verecundumcomplexes (Diptera: Simuliidae) revealed by the sequence of themitochondrial mit LSU rRNA gene. Can J Zool 71: 1202–1206.

Xiong, B. and Kocher, T.D. (1993b) Phylogeny of sibling speciesof Simulium venustrum and S. verecundum (Diptera: Simuliidae)based on sequences of the mitochondrial mit LSU rRNA gene.Mol Phylogenet Evol 2: 293–303.


Documents

Blackwell Science, Ltd Secondary structure and conserved motifs of