Partial mitochondrial genome sequences of Ostrinia nubilalis and Ostrinia furnicalis.

Contiguous 14,535 and 14,536 nt near complete mitochondrial genome sequences respectively were obtained for Ostrinia nubilalis and Ostrinia furnicalis. Mitochondrial gene order was identical to that observed from Bombyx. Sequences comparatively showed 186 substitutions (1.3% sequence divergence), 170 CDS substitutions (131 at 3(rd) codon positions), and an excess of transition mutation likely resulting by purifying selection (d(N)/d(S) = omega congruent with 0.15). Overall substitution rates were significantly higher at 4-fold (5.2%) compared to 2-fold degenerate codons (2.6%). These are the 3(rd) and 4(th) lepidopteran mitochondrial genome reference sequences in GenBank and useful for comparative mitochondrial studies.

1. Introduction

Mitochondrial genomes of 16 insect species are completely sequenced and published with a majority from the order Diptera; D. yakuba 1, Ades gambiae 2, Anopheles quadrimaculatus 3, D. melanogaster 4, Ceratitis capitata 5, Cochliomiyia hominivorax 6, D. simulins 7, and Bactrocera oleae 8. Complete sequences also have been published from a hymenopteran, Apis mellifera 9, an orthopteran, Locusta migratoria 10, a phthirapteran H. macropus 11, thysanuran, T. inaginis 12, a hemipteran Triatoma dimidiata 13, coleopteran Crioceris duodecipunctata 14, and lepidopterans Bombyx mori and B. mandarina 15.

Larvae from corn borer species Ostrinia nubilalis and Ostrinia furnicalis (Lepidoptera: Crambidae) are pests of agricultural crop plants and cause major crop production losses 16, 17. Ostrinia nubilalis and O. furnicalis are sister species 18, 19, with difference residing in female O. nubilalis and O. furnicalis emission of E- and Z- stereoisomers of Δ11- 20, and Δ12-tetradecenyl acetates 21, respectively. The pheromone binding protein gene sequences showed little nucleotide variance between O. nubilalis and O. furnicalis 22, and 7 allozyme markers indicated a high similarity between Chinese populations of O. nubilalis and O. furnicalis suggesting recent speciation 18. Similarly, mitochondrial cytochrome c oxidase subunit II (coxI) gene alignment estimated 1.63% interspecies divergence 19. The present study compares GenBank annotated mitochondrial genomes from O. nubilalis (accession AF442957) and O. furnicalis (AF467260).

2. Materials and methods

2.1 Samples and amplification

A single bivoltine female Z-pheromone race O. nubilalis adult was collected from the Iowa State University Uthe Farm, Ames, Iowa, USA. One adult multivoltine O. furnicalis female of indeterminate pheromone composition collected from Hengshui, Hebei Province, China was contributed by Dr. Wang Zhen-ying, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, China. DNA extractions used Qiagen DNeasy kits (Qiagen, Valencia, CA).

Primers combinations TY-J-1460 with TK-N-3785, J-11545 with N-12854, and N1-J-12585 with SR-N-14588 23 were used to PCR amplify fragments 2, 9, and 10 (Fig. 1). Bombyx mori (GenBank:AF149768 and AY048187) and D. yakuba (GenBank: MIDYRRN) 1 mitochondrial genomes were aligned using AlignX software (Informax, San Francisco, CA) to identify regions of sequence similarity, from which regions PCR primers were designed to amplify remaining fragments using Primer3 24. All PCR reactions were performed in a 50 μl volume with 1.7 U of Tli polymerase (Promega Corp., Madison, WI), 100 ng of DNA, 5 μl 10X thermal polymerase buffer (Promega), 2.5 mM MgCl2, 200 μM dNTPs, and 20 pmol of each primer. Fragments 1, and 3 to 8 were amplified on a PTC-100 thermocycler (MJ Research, Watertown, MA) with denaturation at 95ºC for 2 m, followed by 40 cycles at 94ºC for 30 s, 50 to 54 ºC for 40 s, a 2.5 ºC/s ramp for +15 ºC, and 70 ºC for 1.5 to 3 m depending on fragment length. Fragments 2, 9, and 10 were amplified by denaturing template at 95ºC for 2 m, followed by 40 cycles at 94ºC for 30 s, 44 ºC for 1 m, a 2.0 ºC/s ramp for +23 ºC, and 70 ºC for 3 m.

Figure 1

Ostrinia mitochondrial genome map of sequenced regions. Protein coding genes represented by arrows indicating direction with left-facing arrows on major strand. The tRNA genes are labeled by single letter codes and * indicating coding sequence on minor strand. Underscores indicate positions of ten overlapping PCR amplified genome fragments.

2.2 DNA sequence and analysis

PCR reaction products for fragments 1 to 10 were purified using Qiaquick PCR purification columns (Qiagen), and diluted to 2.5ng/μl/100 bp of product length. Sequencing was performed in duplicate at the DNA sequencing core facility at Iowa State University, Ames, IA. Overlapping fragments were assembled into a single contiguous sequence using Contig Express software (Informax). Ostrinia nubilalis and O. furnicalis mitochondrial genome sequences were aligned with B. mori (GenBank AF149768, and GenBank AY048187) using AlignX software (Informax), and gene features were annotated using Vector NTI 7.0 (Informax). Contiguous mitochondrial DNA sequence of 14535 and 14536 nt were respectively submitted to GenBank for O. nubilalis (AF442957) and O. furnicalis (AF467260).

Substitution rate and transition/transversion ratio for Ostrinia mitochondrial DNA sequences were calculated with MacClade 4.03 25. Twenty one tRNA gene structures were predicted with M-fold 3.1 26, and viewed using RNAviz 2.0 27. Codon usage was evaluated by the Countcodon program version 4 (http://www.kazusa.or.jp/codon/ countcodon.html). Average per site rates of synonymous (dS) and nonsynonymous nucleotide substitution (dN) were calculated according to 28 using MEGA 29.

3. Results and discussion

3.1 Ostrinia mitochondrial genomes

Contiguous O. nubilalis (GenBank accession: AF442957) and O. furnicalis (AF467260) mitochondrial genomes were assembled from overlapping PCR product sequence (Fig. 1). Each GenBank record includes 13 open reading frames (ORFs), a large ribosomal RNA (rrnL) gene, 21 tRNAs, and part of trnM and small ribosomal RNA (rrnS) genes (Fig. 1). Gene order and orientation were identical to Drosophila 1, 4, except for translocation of trnM to a position preceding trnI as was observed in Bombyx 15. Major strand of O. nubilalis (41.3% A, 38.8% T, 8.0% G, and 11.8% C; 80.2% AT) and O. furnicalis (41.5% A, 38.9% T, 7.9% G, and 11.7% C; 80.4% AT) showed a bias toward A and T nucleotides that is typical of insect mitochondrial genomes 30.

The O. nubilalis and O. furnicalis mitochondrial genomes have 3731 codons; 3718 amino acid encoding and 13 termination codons (Table 1). Codons had a prevalence of A and T in 3rd positions and bias may reflect selection for optimal tRNA use 31, speed of genome replication, genome bias, or DNA repair efficacy (Table 2). The O. nubilalis and O. furnicalis mitochondrial peptides comparatively showed 24 predicted amino acid changes (24 of 3718; 0.646%; peptide similarity ≅ 99.22%, and identity ≅ 99.78%) 32. All ORFs were initiated by ATA or ATT codons, except cox1. Initiation of coxI translation is ambiguous, but may occur by a TATTAG sequence in O. nubilalis and O. furnicalis, that is similar to TTTTAG in the B. mori. Hexanucleotides, initiation signals TATCTA from Penaeus monodon 33, or ATTTAA from A. gambiae 2, A. quadrimaculatus 3 and C. capitata 5 have been proposed. Alternatively, an ATAA tetranucleotide sequence was predicted to initiate cox1 translation in Drosophila, L. migratoria 10, and Daphnia pulex 34. Termination codons were either TAA or TAG in O. nubilalis and O. furnicalis, except for cox2 and atp6 that have incomplete stop codons T and TA, respectively. Incomplete stop codons may become function after polycistronic transcript cleavage and polyadenylation mechanisms 35, 36.

Table 1

Codon usage for 3718 amino acid residues and 13 nonsense codons among protein coding regions from each O. nubilalis (On) and O. furnicalis (Of) using the invertebrate mitochondrial genetic code.

Codon	On	Of	Codon	On	Of	Codon	On	Of	Codon	On	Of
UUU-Phe	347	354	UCU-Ser	93	95	UAU-Tyr	175	170	UGU-Cys	29	30
UUC-Phe	31	28	UCC-Ser	12	10	UAC-Tyr	12	16	UGC-Cys	2	1
UUA-Leu	459	454	UCA-Ser	94	94	UAA-Ter	*12	*12	UGA-Trp	89	89
UUG-Leu	15	17	UCG-Ser	5	4	UAG-Ter	1	1	UGG-Trp	5	5
CUU-Leu	20	22	CCU-Pro	60	60	CAU-His	55	58	CGU-Arg	14	14
CUC-Leu	0	1	CCC-Pro	8	7	CAC-His	7	4	CGC-Arg	1	1
CUA-Leu	32	32	CCA-Pro	55	52	CAA-Gln	61	61	CGA-Arg	33	33
CUG-Leu	0	0	CCG-Pro	1	2	CAG-Gln	3	3	CGG-Arg	3	3
AUU-Ile	449	455	ACU-Thr	71	72	AAU-Asn	232	234	AGU-Ser	23	25
AUC-Ile	28	28	ACC-Thr	14	10	AAC-Asn	24	22	AGC-Ser	2	2
AUA-Met	262	265	ACA-Thr	71	75	AAA-Lys	92	94	AGA-Ser	90	89
AUG-Met	26	20	ACG-Thr	1	0	AAG-Lys	9	8	AGG-Ser	0	1
GUU-Val	72	70	GCU-Ala	70	71	GAU-Asp	59	59	GGU-Gly	56	56
GUC-Val	1	2	GCC-Ala	9	8	GAC-Asp	4	4	GGC-Gly	1	5
GUA-Val	58	62	GCA-Ala	46	47	GAA-Glu	65	62	GGA-Gly	117	115
GUG-Val	6	1	GCG-Ala	2	1	GAG-Glu	8	10	GGG-Gly	29	25

* Includes stop codons from cox2 (T) and atp6 (TA), completed by adenylation.

Table 2

Nucleotide frequencies partitioned among O. nubilalis (On) and O. furnicalis (Of) mitochondrial genome regions. IGS = non-coding intergenic spacer regions.

	Protein Coding Sequence
	1st position		2nd position		3rd position		rrnL		rrnS		tRNAs		IGS
% nt	On	Of	On	Of	On	Of	On	Of	On	Of	On	Of	On	Of
% AT	74.4	74.5	70.4	70.5	92.8	93.3	84.9	85.0	82.3	82.8	82.2	82.1	91.9	93.4
% GC	25.6	25.5	29.6	29.5	7.2	6.7	15.1	15.0	17.7	17.2	17.8	17.9	8.1	7.6

Complete nucleotide sequence was obtained for 21 O. nubilalis and O. furnicalis mitochondrial tRNAs. Seven substitutions were observed, and 0.49% sequence divergence was estimated from 1429-shared sites. Insertion-deletion (indel) mutation occurred in loop structures of trnA, trnD, trnG, and trnT, and, except for trnR, did not affect predicted two-dimensional tRNA structures (Fig. 2). Variable mitochondrial tRNA loops in Bombyx were assumed not to affect biological function 15. The complete rrnL gene sequence was 1339 nt for O. nubilalis and O. furnicalis, and alignment comparatively showed a single C to T transition. A partial rrnS sequence was obtained from O. nubilalis (434 nt) and O. furnicalis (435 nt), and comparatively showed a single nucleotide deletion.

Figure 2

Predicted secondary structures for A) O. nubilalis and B) O. furnicalis trnR.

3.2 Nucleotide substitution pattern

A 14543 nt consensus mitochondrial genome alignment identified 186 substitutions between O. nubilalis and O. furnicalis: 138 transition (ts) and 48 transversion (tv) mutations (ts:tv = κ ≅ 2.88). This ratio deviated significantly from neutral expectation (1:2; χ2 = 141.447, d.f. = 1, P < 0.001), indicating evolutionary pressures are acting upon O. nubilalis and O. furnicalis mitochondrial genomes. Excess transition mutation also was reported between D. melanogaster subgroup members (κ= 761/180 ≅ 4.23) and attributed to non-neutral evolutionary forces or population effects 7.

Additionally, mitochondrial protein coding sequences (CDS) comparatively showed 170 substitutions between O. nubilalis and O. furnicalis; 131 at 3rd codon positions. The ratio of the rate of nonsynonymous changes at nonsynonymous sites (dN) to synonymous changes at synonymous sites (dS) in Ostrinia ORFs indicated a 7-fold excess of silent mutation (dN/dS = ω ≅ 0.15) 28. High peptide similarity (≅ 99.22%) may reflect regency O. nubilalis and O. furnicalis speciation, but effects of purifying selection can be inferred since synonymous substitutions are very prevalent. Alternatively, similar environmental selection after speciation could lead to peptide conservation co-occurring with a background of random genetic drift at neutral nucleotide positions. The observed mutation rate at Ostrinia 4-fold degenerate codons (μ4-fold = 5.22%) was significantly higher than at 2-fold degenerate codons (μ2-fold = 2.60%; χ2 = 35.157, d.f. = 1, P < 0.001). Results suggest a greater susceptibility of 4-fold degenerate codons to synonymous substitution.

3.3 Divergence time estimates

The divergence time between O. nubilalis and O. furnicalis mitochondrial was estimated by assuming a linear rate of substitution in short-term evolution (molecular clock) 37 of 2% per million years 38. Nucleotides in rRNA and tRNA may lack independence due to structural dependence, and purifying selection may act at 1st and 2nd codon positions. The intergenic sequence (IGS) and 3rd codon positions only sites that are nearly neutral. IGS region and 3rd codon positions showed 3.54% nucleotide difference between O. nubilalis and O. furnicalis, indicating that speciation occurred 1.8 mya 38. Alternatively, 3rd position and IGS region data give a pairwise genetic distance of 0.3284 ± 0.0348 using the Kimura-2-parameter model 39. Estimates of 0.1 distance unit (D) per 1.0 myr 40 suggest divergence at 3.3 mya. These molecular-based divergence time estimates are supported by highly similar morphology of O. nubilalis and O. furnicalis 41.