Mitogenome and phylogenetic analysis of typhlocybine leafhoppers (Hemiptera: Cicadellidae)

Mitogenomes have been widely used to estimate phylogenetic relationships among insects and provide data useful for augmenting traditional morphological characters in delimiting species. Here, complete mitogenome sequences of two closely related typhlocybine leafhoppers, Cassianeura cassiae (Ahmed, 1970) and C. bimaculata Dworakowska, 1984, were obtained and found to be 15,423 bp and 14,597 bp in length, respectively. The gene order was found to be similar to other published leafhopper mitogenomes, but the control region of C. bimaculata is the shortest among known leafhoppers and lacks tandem repeats. Phylogenetic analysis of 13 protein-coding genes (PCGs), the first and second codons of 13 PCGs, 13 PCGs and two rRNAs formed three well-supported tree topologies. The topologies of phylogenetic trees inferred from three datasets were almost identical, which was consistent with previous molecular phylogenies of this group. Comparative morphological study of the ovipositors revealed several characters potentially useful for diagnosing genera and resolving their phylogenetic relationships. Phylogenetic analysis of these and other morphological characters yielded a tree that is mostly consistent with the tree obtained from analysis of mitogenome sequences. In both molecular and morphological phylogenenies, Typhlocybini and Zyginellini clustered into one clade, but neither was recovered as monophyletic.

www.nature.com/scientificreports/ of Zyginellini, and four species of Empoascini. These data are therefore far from sufficient to facilitate detailed comparative and phylogenetic research on Typhlocybinae. In order to supplement of available mitogenome data of Typhlocybinae and exploration of its internal phylogenetic relationship, two species of the genus Cassianeura of the Erythroneurini were sequenced and annotated in this study ( Table 1). The structure and composition of mitochondrial genomes of two of the three species of Cassianeura were analyzed and compared. Specimens of the two species were recently collected by us from the Cassia tree in Pattaya, Chunburi Province, Thailand 14 .

Materials and methods
Sample collection and DNA extraction. The specimen information for leafhoppers collected in this study are shown in Table 1, and the data downloaded from GenBank are shown in Table 2. The specimens were immersed in absolute ethanol and stored in a refrigerator at − 20 °C. After morphological identification, the prepared voucher specimens with male genitalia and female genitalia were deposited in the insect specimen room of Guizhou Normal University.
Genome sequencing, assembly, and annotation. Mitogenomes were obtained by PCR amplification and sequencing. The PCR reaction was performed with LA-Taq polymerase. The cycling conditions comprised a predenaturation step for 2 min at 94 °C, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, elongation at 72 °C for 1 min/kb, and then the final extension for 10 min at 72 °C. PCR products were then directly sequenced. After quality-proofing of the obtained DNA fragments, DNAStar 15 was used to manually assemble the complete mt genome sequence, and homology search was performed through the Blast function in NCBI to verify whether the amplified sequence matched the target gene regions 16,17 . The positions and secondary structures of 22 tRNA genes were determined using tRNAscan SE version 1.21 18 and ARWEN version 1.2 19 . The base composition of each gene, relative synonymous codon usage (RSCU), and A + T content values were analyzed with MEGA 6.06 20 . Tandem repeat sequences of the control area were identified by the online search tool Tandem Repeats Finder 21   www.nature.com/scientificreports/ overcome of base compositional heterogeneity and explore the effect of method choice on phylogenetic reconstruction, three datasets were assembled: (1) a concatenated nucleotide sequence alignment of the first and second codons of 13 PCGs (PCG12); (2) a concatenated nucleotide sequence alignment of 13 PCGs (PCG123); (3) a concatenated nucleotide sequence alignment of the first and second codons of 13 PCGs and two rRNAs (PCR12R). The Gblocks Server online platform was used to eliminate poorly aligned positions and divergent areas of the DNA protein alignment and check and correct all the alignments in MEGA 6.06 40 . Phylogenetic trees were estimated using the Bayesian Inference (BI) method and the Maximum Likelihood (ML) method. The BI analysis was performed using MrBayes 3.2.7 41 . BI selected GTR + I + G as the optimal model, running 10 million generations twice, sampling once every 1000 generations, with a burn-in of 25%, and the remaining trees used to generate a consensus tree and calculate the probability of each branch of the posterior probability (PP). The ML analysis was conducted using IQ-TREE 42 under an ML + rapid bootstrap (BS) algorithm with 10,000 replicates used to calculate bootstrap scores for each node (BP).
Photographing and illustration of female genitalia. An Olympus SZX16 dissecting microscope was used for specimen study and an Olympus BX53 stereoscopic microscopes for drawing of the dissected female genitalia. Morphological terminology used in this study follows Dietrich & Rakitov 43 . All specimens examined were deposited in the collection of the School of Karst Science, Guizhou Normal University, China (GZNU). Photographs were taken by using a KEYENCE VHX-5000 digital microscope.
Morphology-based phylogenetic analysis. 63 morphological features of 14 species of Typhlocybinae and 1 species of Dikraneura (outgroup) were selected to establish a morphological phylogenetic tree, of which 8 new characteristics from the female valvulae I, II and III of the ovipositor were choosen for the differences of the female genitalia of leafhoppers (Table S3). The remaining 55 are from Dietrich & Dmitriev 44 , representing characteristics that show obvious differences among Typhlocybinae. The character states of the female genitalia are unordered and assigned state numbers "0,1,2,3,…". For the remaining characters, "0" represents the hypothesized ancestral state, and "1" represents the derived state. For multi-state characters, states are arranged in order from ancestral to derived. A data matrix compiled based on the above codes was then analyzed using the traditional search method in TNT.

Results and discussion
Mitochondrial genomic structure and composition. Preivously known mitogenomes of Typhlocybinae range from 14,803 bp to 16,945 bp 27,28 ( Table 2). The complete mitogenomes of C. cassiae and C. bimaculata are 15,423 bp and 14,597 bp in size, respectively ( Fig. 1, Table 1). The smaller size of the complete mitogenome of C. bimaculata is mainly due to the difference in the length of the A + T-rich (control) region. The mitogenomes of C. cassiae and C. bimaculata contain the usual 13PCGs, 22 tRNA genes, two rRNA genes and non-coding regions found in other insects, no gene rearrangements are present. Twenty-three genes encode in the minority strand (H-strand) while the others encode in the majority strand (J-strand) ( Table S1).
The mitogenome of C. cassiae has 12 intergenic spacers, 47 bp in total, ranging in length from 1 to 10 bp. The longest intergenic spacer is between trnA and trnR; 14 gene overlap regions are present, ranging in length from 1 to 8 bp, with the longest between trnW and trnC (Table S1). The mitogenome of C. bimaculata has 10 intergenic spacers of 41 bp in total, and the length also varies from 1 to 10 bp. The longest intergenic spacer is between trnQ and trnM; 14 gene overlap regions are also present in this species, ranging in length from 1 to 8 bp, with the longest gene in the same position as in C. cassiae (Table S1). Such gene structure is common among leafhoppers 27,45 . www.nature.com/scientificreports/ The A + T content of C. cassiae and C. bimaculata (Table 3) is 78.4-93.6% and 73.3-87.8%, respectively. The highest A + T content of both appears in the control region (CR), the lowest value of C. bimaculata appears in the 1st codon position while in C. cassiae it occurs in the N-strand. All genes of the two species mostly showed positive AT-skew and negative GC-skew. The nucleotide composition of the J-strand of C. cassiae and C. bimaculata show relatively high A + T content, 86.2% and 79.3%, respectively, consistent with most other metazoan species 46 . Such mitogenome structure is common in Arthropoda 47,48 . However, the N-strand of C. cassiae and C. bimaculata show negative AT-skew and negative GC-skew.
Protein-coding genes and codon usage. As in most leafhoppers, only four genes (nd5, nd4, nd4L, and nd1) of the 13 PCGs of C. cassiae and C. bimaculata are encoded on the N-strand, while other genes are encoded on the J-strand (Table S1). Except for atp8 and nad5 that use TTG as the start codon, the remaining 13 PCGs start codons all follow the ATN rule. Among the 13 PCGs, standard stop codons (TAA and TAG) and an incomplete stop codon (T) are used. The cox2 of both species are terminated by T, the nad5 gene of C. bimaculata is also terminated by T, but the nad5 gene of C. cassiae is terminated by TAA. Most of the start and stop codons of the two species are the same; only the four genes nad2, atp8, nad3, and nad5 are different. Such subtle differences are common among Cicadellidae 45,49 .
The relative synonymous codon usage (RSCU) for the 13 PCGs is summarized in Table S2 and Fig. 2. Both C. cassiae and C. bimaculata genes encode a total of 3,655 amino acids. In the two mitochondrial genomes, the frequency of the codon UUA (Leu2) is much higher than other codons, and the RSCU for the two species is 3.93 and 3.12, respectively. The codon GCG (Ala) has not been found in C. cassiae, and absence of this codon has been observed in previous studies. For example, Ricania speculum (Fulgoroidea: Ricaniidae) lacks Thr (ACG), and Aphaena discolor nigrotibiata (Fulgoridae) lacks CCG (Pro), GCG (Ala) and ACG (Thr) 50,51 . In addition, both species showed greater codon bias.
Transfer and ribosomal RNA genes. The total tRNA lengths of C. cassiae and C. bimaculata are 1453 bp and 1433 bp, respectively, of which 14 tRNAs are on the N-strand, and the rest are on the J-strand. The 22 tRNAs of the two species are between 62 and 72 bp in length. The trnK are longest of the two species, and the shortest tRNA of C. cassiae is trnC while C. bimaculata has many, including trnC, trnY, trnD, trnG, trnA, and trnR (Table S1). The secondary structure of 22 tRNAs of the two species are shown in Fig. S1.
Secondary structures of tRNAs of C. cassiae and C. bimaculata resemble those of other leafhoppers, except that the dihydrouridine (DHU) arm of tRNA-S1 forms a simple loop, the remaining 21 tRNAs exhibit a typical clover-leaf secondary structure 45,52,53 . 17 and 16 weak G-U base pairs are found in the tRNAs of C. cassiae and C. bimaculata, respectively. In addition, there are differences in the number of base pairs on the Anticodon arm of Arg (R), Asp (D), and Met (M) of the two species. This difference is not common in two species of the same genus 45,49 .
The two highly conserved rRNA genes of C. cassiae and C. bimaculata are encoded on the short-chain (J-strand). 16S are located between trnL2 and trnV, with lengths of 1184 bp and 1168 pb, respectively, and 12S are located after trnV, with lengths of 733 bp and 731 bp, respectively (Table S1). Both 16S and 12S have positive AT-skew and negative GC-skew (Table 3).
Control region. The control region, also known as the A + T region, is the initiation region of mitochondrial DNA replication and the largest non-coding region in the metazoan mitogenome; due to variation in the length of tandem repeat units (TRs) and the difference in copy number, the control region has much greater size variation than other regions of the mitotic genome 48,54,55 . The lengths of the control regions in the mitotic genomes of C. cassiae and C. bimaculata are 1078 bp and 302 bp, respectively, and the A + T% are 93.6% and 87.8%, respectively ( Table 3). The control region of C. cassiae has TRs with 746 bp and 289 bp repeating units (Fig. 3). The length of the control region of C. bimaculata is the smallest known so far in the Typhlocybinae and  (Figs. 4, S2, S3). The BI and ML analysis produced a consistent tree topology comprising ((Zyginellini + Typhlocybini) + Erythroneurini) + Empoascini (Fig. 4), with the two Cassianeura species forming a sister group with Mitjaevia protuberanta. However, some species of Zyginellini were  This distinction has been accepted for many years, e.g., Zhang 58 but more systematic analyses are clearly needed before conclusions can be made on the validity of the tribe.

Female genitalia.
Previous taxonomic studies have shown that different leafhopper species may vary in the shape of female abdominal sternite VII 43,[59][60][61] . Female sternite VII of nine species in subfamily Typhlocybinae were compared for this study (Fig. 5). All have a convex posterior margin narrower than the base, and a transverse or slightly convex base; except for E. sipra, which has the base distinctly emarginate. Sternite VII of different species differs in coloration, length, and in the shape of the posterior margin. M. protuberanta has a relatively long female sternite VII, which covers about half of the ovipositor, and isstrongly produced posteriorly; the female sternite VII of E. sipra and C. cassiae conceal the basal third of the ovipositor, with a shorter, rounded medial lobe; the female sternite VII of E. dwalata has the posterior margin slightly concave medially; the posterior margins of C. bimaculata, E. minuscula and E. gracilivramus are less produced than in other species, being broadly and shallowly rounded posteriorly; the female sternite VII of L. lingchuanensis is transparent; the female sternite VII of Z. minuta covers most of the ovipositor, with two longitudinal dark oval areas in the middle. The female valvulae I of the nine species are shown in Fig. 6. The paired valvulae I are connected basally by membranes. Each valvula I is wide at the base and tapers toward the apex, with imbricate sculpture dorsally and ventroapically. Besides the valvula I of Z. minuta, which is relatively wide, other species are elongated. In previous studies, most typhlocybine leafhoppers have been shown to have long and slim valvulae I 59,60 , with wide valvulae I uncommon. However, Most Deltocephalinae and Cicadellinae leafhoppers have wider female valvulae I 43,61,62 . In addition, the end of female valvula I of E. dwalata is slightly blunt, while the remaining species are acute.
The valvulae II are divided into two branches and in Typhlocybinae the two branches are slightly different in shape, a unique characteristic of this subfamily 43 . It is unique trait that appears the asymmetry of the right and left valvulae for Typhlocybinae in Cicadellidae. The female valvulae II of the 9 species of leafhoppers examined here are shown in Fig. 7. In lateral view, valvulae II are narrow at the base, distinctly expanded in the middle and then gradually narrowed towards the apex, which is blunt. Valvulae II of M. protuberanta and L. lingchuanensis have coarse serrations on the dorsal margin apically, the area between each larger tooth has many small serrations, and the ventral preapical tooth is distinct. Z. minuta also has some stout serrations on the dorsal margin and the part between each larger tooth also has numerous small serrations. However, compared with M. protuberanta and L. lingchuanensis, Z. minuta has more blunt serrations on the dorsal side, and the ventral apical teeth are absent. The valvulae II of E. sipra and E. dwalata have serrations of varying sizes on the dorsal side and the ventral apical tooth is present. C. cassiae and C. bimaculata have almost uniform serrations on the dorsal margin apically and ventral margin smooth. The valvulae II of E. minuscula and E. gracilivramus have slanted large serrations at the apex dorsally, but E. gracilivramus has sharper serrations on the dorsal side. The valvulae II of most leafhoppers has an obvious saw-like structure, which is used to cut the leaf epidermis during oviposition 63 . www.nature.com/scientificreports/ The female valvula III of 9 species of leafhoppers is shown in Fig. 8. Valvulae III form a sheath partially enclosing valvulae I and II. Valvula III has its base connected with valvula I and is shaped like a small boat, with some irregularly arranged small setae on the ventral surface. M. protuberanta, E. sipra, E. dwalata, E. gracilivramus, and Z. minuta have many slender and dense setae on surface of valvula III, while the other studied leafhoppers have relatively few setae. The color of valvula III of different leafhopper species is quite different, mostly similar to the color of sternite VII 43,61 .
In our comparative study of the ovipositors of 9 existing species, we found that the female valvulae I, II, and III of species in the same genus and tribe are similar to each other. Combining the characteristics of body morphology and male genitalia, a morphological phylogenetic tree was obtained (Tables S3, S4, Fig. 9). The internal relationship of the morphological phylogenetic tree is: (Zyginellini + Typhlocybini) + Erythroneurini. Comparing with the molecular phylogenetic tree, the structures of the two are basically the same, all showing the two tribes of Typhlocybini and Zyginellini clustered into one clade. However, relationships within Erythroneurini, the position of the clade comprising M. protuberanta and Illinigina sp. differs. This may be attributable to the differences in taxon sample between the two analyses, i.e., the absence of Empoascini in the morphology-based phylogeny and the different outgroups used to root the trees. www.nature.com/scientificreports/

Conclusions
The Oriental leafhopper genus Cassianeura (Cicadellidae: Typhlocybinae: Erythroneurini) contains three species which feed and reproduce on Cassia trees. In this study, we sequenced and annotated the complete mitogenomes of two Cassianeura species (C. cassiae and C. bimaculata) and compared them with other Typhlocybinae leafhoppers. Comparative mitogenome analysis showed that the gene content and gene arrangement of the two new mitogenomes are both conservative. Phylogenetic analysis of the nucleotide sequences of 13 PCGs resulted in a well-supported topology, with most branches receiving strong support and most relationships consistent Comparative morphological study of the ovipositors of 9 typhlocybine leafhoppers showed that species in the same genus have similar structure, but there are obvious morphological differences among genera and the shape of female sternite VII of different species is quite different. Morphology based phylogenetic analysis of Typhocybinae yields a tree similar to that obtained by analyzing mitogenome sequences. Both suggest that Typhlocybini and Zyginellini are closely related but neither support the monophyly of either tribe. The results of this study www.nature.com/scientificreports/ provide further comparative data on the mitogenomes of Typhlocybinae and suggest that morphological and mitogenome data provide consistent phylogenetic signal useful for elucidating relationships among genera and tribes of this diverse group of plant-feeding insects.