Molecular relationships of Campomanesia xanthocarpa within Myrtaceae based on the complete plastome sequence and on the plastid ycf2 gene

Abstract

Plastomes are very informative structures for comparative phylogenetic and evolutionary analyses. We sequenced and analyzed the complete plastome of Campomanesia xanthocarpa and compared its gene order, structure, and evolutionary characteristics within Myrtaceae. Analyzing 48 species of Myrtaceae, we identified six genes representing ‘hotspots’ of variability within the plastomes (ycf2, atpA, rpoC2, pcbE, ndhH and rps16), and performed phylogenetic analyses based on: (i) the ycf2 gene, (ii) all the six genes identified as ‘hotspots’ of variability, and (iii) the genes identified as ‘hotspots’ of variability, except the ycf2 gene. The structure, gene order, and gene content of the C. xanthocarpa plastome are similar to other Myrtaceae species. Phylogenetic analyses revealed the ycf2 gene as a promissing region for barcoding within this family, having also a robust phylogenetic signal. The synonymous and nonsynonymous substitution rates and the Ka/Ks ratio revealed low values for the ycf2 gene among C. xanthocarpa and the other 47 analyzed species of Myrtaceae, with moderate purifying selection acting on this gene. The average nucleotide identity (ANI) analysis of the whole plastomes produced phylogenetic trees supporting the monophyly of three Myrtaceae tribes. The findings of this study provide support for planning conservation, breeding, and biotechnological programs for this species.

Introduction

Plastidial genomes (plastomes) are useful tools to perform comparative analyses associated with phylogenetic and evolutionary studies. The relatively small size, mostly uniparental inheritance, high gene synteny, and elevated copy number in green plant cells are the main characteristics that make plastids useful for such studies. In seed plants, plastome sizes range from 70 to 218 Kbp (Xiao-Ming et al., 2017) and typically present a quadripartite structure with two inverted repeat regions (IRs) divided between the large (LSC) and the small (SSC) single-copy regions (Bock, 2007).

The plastome contains essential genes in conserved open reading frames (ORFs). However, some plastidial ORFs have unknown function and are called hypothetical chloroplast open reading frame (ycf). The largest plastome coding sequence (ORF2280 or ycf2) encodes a plastidial protein (Glick and Sears, 1993) whose function has been hypothesized to exhibit similarities with fstH, such as ATPase-related activities, chaperone function, and activity associated with cell division (Wolfe, 1994).

The elevated substitution rates of the ycf2 gene led to a pseudogenization process (Downie et al., 1994; Oliver et al., 2010; Wolf et al., 2011). This gene is absent in some plastomes (Downie et al., 1994; Millen et al., 2001), especially in monocot grasses, such as maize, rice, and sugarcane (Maier et al., 1995; Matsuoka et al., 2002; Asano et al., 2004). Thus, the high variability of ycf2 makes it a potential candidate for species-level DNA barcoding (Kumar et al., 2009).

Investigating the sequence variation and phylogenetic characteristics of the ycf2 gene in angiosperms, Huang et al. (2010) showed that it provides generally well-supported phylogenies, consistent with those inferred from the most comprehensive multigene data. Within angiosperms, Myrtaceae is a large family of shrubs and trees with well-known ecological and economic importance in tropical and subtropical regions of the globe. According to The Plant list (2013), this family is composed by 144 genera and 5,970 species distributed across the world, with predominantly Neotropical and Southern Hemisphere distribution. Its main center of diversity is the wet tropics, predominantly in South America, Australia, and tropical Asia (Grattapaglia et al., 2012). Evolutionary and phylogenetic trends within this family have been studied using single and combined plastidial genes, as well as by complete plastome sequencing (Steane, 2005; Bayly et al., 2013, 2016; Jo et al., 2016; Eguiluz et al., 2017; Machado et al., 2017).

Campomanesia xanthocarpa Berg. is a fruit tree species of the family Myrtaceae, native to South America, occurring in Brazil, Argentina, Uruguay, and Paraguay (Lorenzi, 1992). In addition to feeding several mammal and bird species, its fruits are appreciated by local people, being consumed fresh and as component of cakes, ice cream, and beverages. As a heliophyte species, C. xanthocarpa is indicated for recovery of degraded areas and as an ornamental plant (Souza and Lorenzi, 2005). Controlled studies have shown that extracts of C. xanthocarpa leaves induced reduction in weight gain and glycemia in rats (Biavatti et al., 2004) and were efficient in avoiding gastric ulceration (Markman et al., 2004). In hypercholesterolemic human patients, encapsulated dried leaves of C. xanthocarpa significantly decreased total cholesterol and LDL levels (Klafke et al., 2010). This plant also demonstrated anti-inflammatory (Klafke et al., 2016), antimicrobial (Capeletto et al., 2016) and anti-oxidant (Viecili et al., 2014) activities, and may also have therapeutic applications during pregnancy, reducing reabsorption sites, increasing placenta weight and the number of live fetuses (Auharek et al., 2013).

In an effort to elucidate the evolutionary history of South American tree species, the plastome of different native species has been sequenced (Vieira et al., 2014b, 2016a,b; Lopes et al., 2018a,b), including species from the Myrtaceae family (Machado et al., 2017). Aiming to contribute to this attempt and to generate useful information for future efforts towards biotechnology, breeding, and genetic conservation of C. xanthocarpa, we used a next-generation sequencing technology to sequence the complete plastome of this species and describe here its genome structure and gene content. Considering previous studies on plastidial genomes, suggesting strong signatures of positive selection between close-related species of tribe Myrteae, this study aimed to answer three main questions: (i) Does the C. xanthocarpa plastome resemble the plastidial structure of the species from the tribe Myrteae and family Myrtaceae? (ii) Does the C. xanthocarpa ycf2 gene present signatures of positive selection within the tribe Myrteae and family Myrtaceae? (iii) Does the ycf2 gene have a strong taxonomic/phylogenetic signal for resolving relationships within the tribe Myrteae and family Myrtaceae compatible to classic plastidial phylogenies?

Material and Methods

Plant material and plastidial DNA isolation

For plastidial DNA isolation, fresh leaves were collected from a single individual of C. xanthocarpa in the Department of Botany, Federal University of Santa Catarina (UFSC), Brazil (27º36.094” S, 48º31.310” W). The plastidial DNA was obtained according to Vieira et al. (2014a), with the modification that plastidial lysis was achieved by incubating the chloroplast pellet with 8 mL of DNA isolation buffer [1.5 mL 20% SDS, 450 μL 2-mercaptoethanol, and 50 μL proteinase K (10 mg/mL)] in a centrifuge tube at 55 °C for 4 h or overnight.

Plastome assembly and annotation

The sequencing libraries were prepared using 1 ng of plastidial DNA with the Nextera XT DNA Sample Prep Kit (Illumina Inc., San Diego, CA), according to the manufacturer's instructions. Libraries were sequenced using the MiSeq Reagent Kit v3 (600 cycles) on an Illumina MiSeq Sequencer (Illumina Inc., San Diego, California, USA). The obtained paired-end reads (2 x 300 bp) were used for de novo assembly performed with CLC Genomics Workbench v8.0.1. The same software was used to estimate plastome coverage. Initial annotation of the C. xanthocarpa plastome was performed using DOGMA - Dual Organellar GenoMe Annotator (Wyman et al., 2004). From this initial annotation, putative start, stop, and intron positions were determined based on comparisons to homologous genes in other plastomes. The tRNA genes were further verified using tRNAscan-SE (Schattner et al., 2005). A circular plastome map was drawn using OGDRAW - OrganellarGenomeDRAW (Lohse et al., 2007). REPuter (Kurtz and Schleiermacher, 1999) was used to identify and locate the IRs in the C. xanthocarpa plastome by forward versus reverse complement (palindromic) alignment, with minimal repeat size was set to 30 bp, and identity of repeats ≥ 90%. REPuter was also used to identify and locate LSC/IRb/SSC/IRa sizes and boundaries in 12 other previously published chloroplast genomes.

The complete C. xanthocarpa plastome sequence reported in this study was deposited in the GenBank database under accession number KY392760.

Taxonomical relationships within Myrtaceae based on the whole plastome

Aiming to investigate the taxonomical relationships within Myrtaceae based on the whole plastome sequences, we employed the average nucleotide identity (ANI) analysis. This analysis is a measure of nucleotide-level genomic similarity between two genomes, where the averages reflect the degree of divergence between coding regions of the compared genomes and, consequently, evolutionary distances between these genomes. It consists in calculating the percentage nucleotide identity of the matching regions of two genomes, as an average for all matching regions.

Sequences from plastomes of 47 Myrtaceae species including the C. xanthocarpa plastome reported in the present study and 46 Myrtaceae species, which have whole plastome sequences deposited in the GenBank database, were used in this analysis, including 30 species of genus Eucalyptus, six of the genus Corymbia, two of the genus Angophora, and one each of the Allosyncarpia, Eugenia, Stockwellia, Syzygium, Acca, Pimenta, Plinia and Psidium genera (Table 1). The plastome sequence of Lagerstroemia fauriei (Myrtales: Lythraceae) was used as outgroup.

Table 1. Comparison of chloroplast genomes of Myrtaceae species and outgroup analyzed in this study.

Species	Accession	Size	LSC b	SSC c	IR d
Campomanesia xanthocarpa a	KY392760	158,131	87,596	18,595	25,970
Acca sellowiana	KX289887	159,370	88,028	18,598	26,372
Allosyncarpia ternate	KC180806	159,593	88,218	18,571	26,402
Angophora costata	KC180805	160,326	88,769	18,773	26,392
Angophora floribunda	KC180804	160,245	88,715	18,746	26,392
Corymbia eximia	KC180802	160,012	88,522	18,672	26,409
Corymbia gummifera	KC180800	160,713	88,310	17,197	27,603
Corymbia henryi	KP015032	160,095	88,589	18,688	26,409
Corymbia maculate	KC180801	160,045	88,557	18,670	26,409
Corymbia tessellaris	KC180803	160,127	88,617	18,692	26,409
Corymbia torelliana	KP015033	159,994	88,494	18,682	26,409
Eucalyptus aromaphloia	KC180789	160,149	88,925	18,468	26,378
Eucalyptus baxteri	KC180773	160,032	88,926	18,368	26,369
Eucalyptus camaldulensis	KC180791	160,164	88,874	18,492	26,399
Eucalyptus cladocalyx	KC180786	160,213	89,045	18,376	26,396
Eucalyptus cloeziana	KC180779	160,015	88,867	18,446	26,351
Eucalyptus curtisii	KC180782	160,038	88,828	18,448	26,381
Eucalyptus deglupta	KC180792	160,177	88,936	18,425	26,408
Eucalyptus delegatensis	KC180771	159,724	88,490	18,498	26,368
Eucalyptus diversicolor	KC180795	160,214	88,994	18,416	26,402
Eucalyptus diversifolia	KC180774	159,954	88,901	18,315	26,369
Eucalyptus elata	KC180776	159,899	88,762	18,401	26,368
Eucalyptus erythrocorys	KC180799	159,742	88,691	18,287	26,382
Eucalyptus globulus	AY780259	160,286	89,012	18,488	26,393
Eucalyptus grandis	HM347959	160,137	88,872	18,475	26,395
Eucalyptus guilfoylei	KC180798	160,520	89,054	18,096	26,685
Eucalyptus marginata	KC180781	160,076	88,828	18,476	26,386
Eucalyptus melliodora	KC180784	160,386	89,073	18,557	26,378
Eucalyptus microcorys	KC180797	160,225	89,051	18,410	26,382
Eucalyptus nitens	KC180788	160,271	89,005	18,468	26,399
Eucalyptus obliqua	KC180769	159,527	88,293	18,498	26,368
Eucalyptus patens	KC180780	160,187	88,902	18,543	26,371
Eucalyptus polybractea	KC180785	160,268	88,944	18,530	26,397
Eucalyptus radiate	KC180770	159,529	88,295	18,498	26,368
Eucalyptus regnans	KC180777	160,031	88,860	18,447	26,362
Eucalyptus saligna	KC180790	160,015	89,041	18,426	26,274
Eucalyptus salmonophloia	KC180796	160,413	89,173	18,466	26,387
Eucalyptus sieberi	KC180775	159,985	88,848	18,401	26,368
Eucalyptus spathulata	KC180793	161,071	88,729	17,116	27,613
Eucalyptus torquata	KC180794	160,223	89,018	18,439	26,383
Eucalyptus umbra	KC180778	159,576	88,864	18,658	26,027
Eucalyptus verrucata	KC180772	160,109	88,890	18,481	26,369
Eugenia uniflora	KR867678	158,445	87,459	18,318	26,334
Pimenta dioica	KY085891	158,984	87,572	18,586	26,413
Plinia trunciflora	KU318111	159,512	88,097	18,587	26,414
Psidium guajava	KX364403	158,841	87,675	18,464	26,351
Stockwellia quadrifida	KC180807	159,561	88,247	18,544	26,385
Syzygium cumini	GQ870669	160,373	89,081	18,508	26,392
Lagerstroemia fauriei (Lythraceae)e	KT358807	152,440	83,923	16,933	25,792

^aSpecies with plastid genomes sequenced in this study

^bLarge Single Copy Region

^cSmall Single Copy Region

^dInverted Repeat Region

^eOutgroup

ANI was calculated for the whole plastomes using the Pyani script (Python module) for average nucleotide identity analyses; (https://github.com/widdowquinn/pyani), aligning the sequences with the MUMmer algorithm (Goris et al., 2007).

Evolutionary and phylogenetic patterns within Myrtaceae based on the ycf2 gene

In order to evaluate the phylogenetic signal of the ycf2 gene within Myrtaceae, we analyzed evolutionary and phylogenetic patterns of this gene among representatives of this family using sequences obtained from the same species employed in the ANI analysis (Table 1).

The evolutionary patterns of the plastidial ycf2 gene in family Myrtaceae were evaluated by estimating the Ka/Ks ratio for the ycf2 protein-coding gene. The evolutionary characteristics, nonsynonymous (Ka) and synonymous substitution rates (Ks), as well as Ka/Ks ratio, were calculated using Model Averaging in the KaKs_Calculator program (Zhang et al., 2006). The genes were pairwise aligned using the MUltiple Sequence Comparison by Log-Expectation (MUSCLE) algorithm (Edgar, 2004) to identify synonymous and nonsynonymous substitution.

For the phylogenetic relationships of Myrtaceae based on the ycf2 gene, sequences were aligned by Multiple Alignment using Fast Fourier Transform -MAFFT (Katoh and Standley, 2013). The substitution model was selected by means of the Akaike information criterion using jModelTest (Darriba et al., 2012) with seven substitution schemes, as this set covers all the possible models present in MrBayes software. Bayesian inference was conducted using MrBayes v3.2.6 at CIPRES Science Gateway V. 3.3, with the general time reversible (GTR) model of substitution incorporating invariant sites (GTR + I), as suggested by the model test selection. Markov Chain Monte Carlo (MCMC) simulations were run for 6,000,000 generations (average standard deviation of split frequencies = 0.003027), discarding the first 25% of trees as burn-in. The remaining trees were represented and edited using FigTree v1.4.1.

In addition, we performed a sliding window analysis of the total plastid genome of all analyzed species using the software DNAsp v.5 (Librado and Rozas, 2009). The window length and the step size were set as 200 and 50 bp, respectively. The genes representing hotspots of sequence divergence were identified and used for a new phylogenetic analysis, both including and excluding the ycf2 region, in order to evaluate the individual contribution of this gene to the phylogenetic patterns in comparison to other plastid genes highly variable within Myrtaceae.

Results

Sequencing results

The Illumina MiSeq sequencing resulted in a high plastome coverage (∼370x) with a total of 3,240,548 raw reads, an average read length of 147.9 bp, and a total number of 479,391,832 base pairs. After trimming (quality score limit of 0.05) a total of 3,228,689 reads were mapped in aligned pairs with mean length of 147.52 bp, generating a total of 58,815,423 bp, which were used for the de novo assembly.

General features of the Campomanesia xanthocarpa plastome

The C. xanthocarpa plastome has 158,131 bp in length, with a GC content of 36.98% and the general quadripartite structure, consisting of a pair of IRs (25,970 bp) separated by the LSC (87,596 bp) and SSC (18,595 bp) regions (Figure 1; Table 2). It is the smallest plastidial genome size within the reported plastomes of Myrtaceae, 2,582 bp shorter than the plastome of Corymbia gummifera, the longest plastidial genome reported for this family. The IR region of the C. xanthocarpa plastome is the shortest among the 48 Myrtaceae species evaluated herein (Table 2).

Figure 1. Gene map of the Campomanesia xanthocarpa chloroplast genome. Genes drawn inside the circle are transcribed clockwise, and genes drawn outside are transcribed counterclockwise. Genes belonging to different functional groups are color-coded. The darker gray in the inner circle corresponds to GC content, and the lighter gray corresponds to AT content.

Table 2. Summary of Campomanesia xanthocarpa chloroplast genome characteristics.

Characteristics of plastome	C. xanthocarpa
Plastome Size (bp)	158,131
LSC size in bp (%)	87,596 (55.39)
SSC size in bp (%)	18,595 (11.76)
IR length in bp	25,970
Different genes	115
Different PCG	77
Different tRNA genes	30
Different rRNA genes	4
Different pseudogenes	4
Different genes duplicated by IR	20
Different genes with introns	18
Overall % GC content	36.98
% GC content in LSC	34.8
% GC content in SSC	30.6
% GC content in IR	42.9

PCG = Protein-coding gene

The plastome contains 112 genes and four pseudogenes, with the same gene order and gene clusters as other Myrtaceae. The presence of pseudogenes is a known feature of Myrtaceae plastomes. Out of the 112 genes, 90 were single copy and 19 were duplicated (Figure 1; Table S1 ^{(92.1KB, pdf)} ). In addition, 18 were intron-containing genes (Table 3), including nine protein-coding genes with a single intron, two protein-coding genes with a double intron, six tRNA genes with a single intron and one trans-splicing gene (rps12). Among intron-containing genes, 12 are located in the LSC region, one in the SSC region, and four in the IR region.

Table 3. Genes with introns in Campomanesia xanthocarpa chloroplast genome and length of exons and introns.

Region	Gene	Exon I (bp)	Intron I (bp)	Exon II (bp)	Intron II (bp)	Exon III (bp)
LSC	rps16	206	889	38
LSC	rpoC1	1616	730	452
LSC	atpF	410	747	146
LSC	petB	5	772	647
LSC	petD	8	753	473
LSC	rpl16	398	1005	8
LSC	ycf3	152	725	227	760	125
LSC	clpP	227	620	290	871	68
LSC	trnK-UUU	34	2530	36
LSC	trnG-UCC	22	750	48
LSC	trnV-UAC	36	595	38
LSC	trnL-UAA	36	504	49
SSC	ndhA	539	1061	551
LSC/IRs	rps12 *	113	–	209	–	26
IR	rpl2	434	662	392
IR	ndhB	755	695	776
IR	trnI-GAU	36	956	33
IR	trnA-UGC	37	804	34

^*rps12 is trans-spliced with the 5’-end located in the LSC region and the duplicated 3’-end in the IR regions

The LSC/IRb/SSC/IRa boundary regions were examined to compare four junctions (JLA, JLB, JSA, and JSB) among 12 species of three tribes of Myrtaceae and an outgroup, from Lythraceae family. The IR lengths ranged from 25,792 bp to 26,414 bp, and the position of IRs boundaries varied for each species. The boundary between LSC and IR regions was located within the rps19 gene, resulting in the formation of an rps19 pseudogene in C. xanthocarpa, A. ternata, S. quadrifida, P. dioica, P. trunciflora, P. guajava and L. fauriei chloroplast genomes (Figure S1 ^{(373.4KB, pdf)} ). In the other six species, the LSC comprises an intact rps19 gene together with 2 bp (E. grandis), 3 bp (A. sellowiana and E. uniflora), 6 bp (S. cumini), and 8 bp (A. costata and C. maculata) of non-coding region beyond the LSC-IRb border. The IRa-LSC border in these six species is located in the intergenic spacer (IGS) between rpl2 and trnH. The trnH gene in C. xanthocarpa, A. ternata, S. quadrifida, P. dioica, P. trunciflora, P. guajava and L. fauriei extends to the IRa by 31 bp, 5 bp, 5 bp, 1 bp, 4 bp, 11 bp and 3 bp, respectively, whereas the same gene for A. sellowiana, E. uniflora, A. costata and C. maculata, E. grandis, and S. cumini is, respectively 53 bp, 44 bp, 9 bp, 9 bp, 2 bp, and 56 bp away from the IRa-LSC border. The boundary of the SSC-IRb junction in Myrtales plastomes was located within the ycf1 gene, also resulting in the formation of a ψycf1 pseudogene, which varied in length between 1,007 bp and 2,251 bp. In the A. ternata and L. fauriei chloroplast genomes, the ndhF gene at the IRb-SSC border extends 35 bp and 38 bp into the IRb region, respectively. This gene is located in the SSC region in the 11 other species, and is separated from the IRb-SSC border by five to 225 bp.

Taxonomic relationships within Myrtaceae based on the whole plastome

The average nucleotide identity (ANI) analysis represents a mean of identity values between homologous regions shared by two genomes, and was used to compare the complete plastomes of 47 species of Myrtaceae and Lagerstroemia fauriei as outgroup species. The plastomes sequence analysis (∼160 kb) indicated an ANI above the 95% threshold for all species within Myrtaceae. Despite the high overall ANI value observed, two major clades can be identified in the dendrogram and are closely apportioned in the heatmap (Figure 2). The Myrteae clade consists of six species [Campomanesia + [[Acca + Psidium] + [[Pimenta + Plinia] + Eugenia]]]. The second clade, which includes species from both tribes Syzygieae and Eucalypteae, is subdivided in four subgroups, one containing all Eucalyptus species (all pairwise comparisons with ANI values near 100%), a second with representatives of Stockwellia and Allosyncarpia, a third subgroup comprising Corymbia and Angophora, and a fourth subgroup with the unique Sysygium species, from tribe Syzygieae (Figure 2).

Figure 2. Heatmap of ANIm percentage identity for 48 Myrtaceae species and the outgroup species Lagerstroemia fauriei (Myrtales: Lythraceae; KT358807). Cells in the heatmap corresponding to 95% ANIm sequence identity are in red color; Cells corresponding to 75% ANIm sequence identity are in blue. Color intensity fades as the comparisons approach 95% ANIm sequence identity. Color bars above and to the left of the heatmap correspond to the species level that was analyzed.

Evolutionary and phylogenetic patterns within Myrtaceae based on the ycf2 gene

The translated product of ycf2 from the plastome of C. xanthocarpa contains 2,295 amino acids (6,885 bp). It is longer than the sequence reported from all other species of Myrtaceae included in this study (ranging from 2,288 to 2,289 amino acids), except for Eucalyptus spathulata. This difference is observed in the C-terminal portion of the translated product (Figure 3). The translated product of this gene in C. xanthocarpa has the same amino acid sequence as the other species up to position 2,283 (Figure 3). The amino acids from position 2,284 to 2,288 are conserved in all other species, with exception of E. spathulata. The amino acids of positions 2,289 and 2,290 in P. guajava and from positions 2,289 to 2,295 in E. spathulata are different from the amino acids of C. xanthocarpa (Figure 3). The analysis of the pairwise diversity between C. xanthocarpa and 46 species of Myrtaceae (Table 4) showed a Ka ranging from 0.0019 (Pimenta dioica) to 0.0054 (Eucalyptus grandis, E. deglupta and Corymbia gummifera) with a mean Ka = 0.0048. Estimations of Ks ranged from 0.0010 (Acca sellowiana and Plinia trunciflora) to 0.0138 (Eucalyptus erythrocorys), with a mean Ks = 0.0114. The average Ka/Ks ratio was 0.4620, ranging from 0.03207 (Stockwellia quadrifida) to 3.0 (Plinia trunciflora).

Figure 3. Comparison of the C-terminal region of the amino acid sequences of the ycf2 gene in 47 species of Myrtaceae.

Table 4. Comparison of polymorphisms (segregating sites), nonsynonymous (Ka) and synonymous (Ks) substitution rates and Ka/Ks ratio of the ycf2 gene among Campomanesia xanthocarpa and 46 species of Myrtaceae.

Species	C. xanthocarpa/Species
	Polymorphisms	Ka	Ks	Ka/Ks
Tribe Myrteae
Acca sellowiana	13	0.0021	0.0010	2.0114
Eugenia uniflora	15	0.0021	0.0028	0.7413
Pimenta dioica	13	0.0019	0.0020	0.9677
Plinia trunciflora	19	0.0032	0.0010	3.0000
Psidium guajava	26	0.0035	0.0049	0.7127
Tribe Eucalypteae
Corymbia eximia	40	0.0046	0.0118	0.3921
Corymbia gummifera	42	0.0054	0.0101	0.5314
Corymbia henryi	42	0.0051	0.0110	0.4677
Corymbia maculata	42	0.0051	0.0110	0.4677
Corymbia tessellaris	43	0.0052	0.0117	0.4420
Corymbia torelliana	44	0.0052	0.0127	0.4053
Eucalyptus aromaphloia	42	0.0050	0.0116	0.4309
Eucalyptus baxteri	40	0.0046	0.0118	0.3921
Eucalyptus camaldulensis	43	0.0052	0.0117	0.4433
Eucalyptus cladocalyx	44	0.0052	0.0127	0.4062
Eucalyptus cloeziana	40	0.0046	0.0118	0.3921
Eucalyptus curtisii	44	0.0053	0.0121	0.4403
Eucalyptus deglupta	44	0.0054	0.0115	0.4704
Eucalyptus delegatensis	41	0.0048	0.0119	0.4043
Eucalyptus diversicolor	44	0.0052	0.0124	0.4199
Eucalyptus diversifolia	41	0.0048	0.0116	0.4189
Eucalyptus elata	40	0.0046	0.0118	0.3921
Eucalyptus erythrocorys	46	0.0053	0.0138	0.3811
Eucalyptus globulus	43	0.0052	0.0114	0.4584
Eucalyptus grandis	45	0.0054	0.0125	0.4318
Eucalyptus guilfoylei	41	0.0048	0.0119	0.4038
Eucalyptus marginata	42	0.0050	0.0117	0.4302
Eucalyptus melliodora	45	0.0053	0.0128	0.4173
Eucalyptus microcorys	39	0.0044	0.0121	0.3655
Eucalyptus nitens	42	0.0050	0.0116	0.4309
Eucalyptus obliqua	41	0.0048	0.0119	0.4043
Eucalyptus patens	39	0.0046	0.0108	0.4304
Eucalyptus polybractea	44	0.0052	0.0127	0.4062
Eucalyptus radiata	41	0.0048	0.0119	0.4043
Eucalyptus regnans	41	0.0046	0.0128	0.3594
Eucalyptus saligna	42	0.0051	0.0119	0.4283
Eucalyptus salmonophloia	44	0.0050	0.0136	0.3641
Eucalyptus sieberi	41	0.0046	0.0128	0.3591
Eucalyptus spathulata	44	0.0053	0.0121	0.4362
Eucalyptus torquata	43	0.0052	0.0117	0.4429
Eucalyptus umbra	42	0.0050	0.0120	0.4162
Eucalyptus verrucata	41	0.0048	0.0116	0.4186
Stockwellia quadrifida	34	0.0037	0.0115	0.3207
Tribe Syzygieae
Syzygium cumini	35	0.0043	0.0093	0.4587

The matrix of the ycf2 gene used for the phylogenetic inference was composed of 6,969 nucleotide positions, and the Bayesian phylogenetic tree with highest log-likelihood (lnL = −11,575.58) is shown in Figure 4. Monophyly of the tribes of Myrteae and Eucalypteae was confirmed with posterior probability (PP) of 1.0 (Figure 4). The tribe Syzygieae is sister to Eucalypteae with PP = 1.0.

Figure 4. Bayesian phylogeny based on the cp ycf2 sequence of 48 Myrtaceae species and the outgroup species Lagerstroemia fauriei (Myrtales: Lythraceae; KT358807). Branch length is proportional to the inferred divergence level. The scale bar indicates the number of inferred nucleic acid substitutions per site.

The clades within the Myrteae tribe were highly supported, with posterior probability of 0.9788 for the clade Eugenia uniflora, Acca sellowiana, Campomanesia xanthocarpa, Pimenta dioica, Plinia trunciflora, and Psidium guajava, and PP = 1.0 for the node that ties A. sellowiana and C. xanthocarpa. Similarly, the Eucalypteae tribe segregated into two well-supported monophyletic clades: [Eucalyptus [Angophora + Corymbia]] (PP = 1.0) and [Allosyncarpia ternata + Stockwelia quadrifida] (PP = 1.0).

The sliding window analysis (Figure 5) revealed six genes as hotspots of sequence divergence (π 3 0.03): ycf2, atpA, rpoC2, pcbE, ndhH, and rps16. The phylogenetic tree constructed with all six genes (Figure S2 ^{(87.4KB, pdf)} ) revealed the same topology obtained in the tree built only with the ycf2 gene sequence (Figure 4). When the phylogenetic analysis was performed using the atpA, rpoC2, pcbE, ndhH, and rps16 genes (Figure S3 ^{(139KB, pdf)} ), i.e., excluding the ycf2 gene sequence, Syzygieae was positioned basal to Myrteae and Eucalypteae, differing from the trees built using all six genes (Figure S2 ^{(87.4KB, pdf)} ) and based only on the ycf2 sequence (Figure 4). However, the support for the node of Myrteae in this tree was low (PP = 0.55; Figure S3 ^{(139KB, pdf)} ).

Figure 5. Sliding window analyses of aligned whole plastomes for the family Myrtaceae, the tribe Myrteae and the tribe Eucalypteae. The regions with high nucleotide variability (Pi > 0.03) are indicated. Pi = nucleotide diversity within each window.

Discussion

Structural patterns of the C. xanthocarpa plastome in comparison to other Myrtaceae

Conservation of plastome structure has been reported for species of Myrtaceae concerning the total length, as well as the length of the SSC, LSC and IR regions, and the number and position of genes (Machado et al., 2017, Xiao-Ming et al., 2017). Fitting with these previous findings, the assembly and annotation of the C. xanthocarpa plastome revealed a general structure, including number and category of the genes, similar to that of other 47 species of Myrtaceae, suggesting overall stability. The reduction of the C. xanthocarpa plastome is not related to a loss of genes or pseudogenes. Large expansions and reductions in plastome sizes in higher plants have been attributed to the extension of inverted repeats into neighboring single copy regions (Ravi et al., 2008). The difference between the C. xanthocarpa plastome and the plastomes of the other species included in this study, which ranged from 314 bp (in comparison to E. uniflora) to 2,940 bp (in comparison to E. spathulata), was due to indels occurring in intergenic regions of both, IR and LSC regions of all species, except for P. dioica (Table 1).

While the structure of the C. xanthocarpa plastome presented such stability, the analysis of the ycf2 gene suggested the occurrence of different patterns of selection at family and at tribe levels. In this study, we found very low variability in this gene among published plastomes of Myrtaceae species, with 45 species presenting the same length and same amino acid sequences for the ycf2 gene. However, the ycf2 gene of C. xanthocarpa is 18 bp longer than those of these species, 15 bp longer than that of Psidium guajava and has the same size as that of Eucalyptus spathulata. Moreover, having the same amino acid sequences does not mean having the same nucleotide sequences, what enables the exploitation of the nucleotide sequences of the ycf2 gene as a barcode region in plants (Kumar et al., 2009). The ycf2 gene has an essential, but yet little known function in higher plants. Kikuchi et al. (2018) demonstrated the function of a protein encoded by this gene, associated to five related nuclear-encoded FtsH-like proteins, in the translocation of preproteins across the inner membrane of the chloroplast. Silencing or reduction in the mRNA synthesis of this gene has been shown to induce cell apoptosis (Drescher et al., 2000). It seems, however, that the elongation of this gene and the substitution of some amino acids in the C-terminal portion of the sequence has no deleterious effect in C. xanthocarpa, P. guajava and E. spathulata.

Modest signatures of positive selection for the C. xanthocarpa ycf2 gene within Myrtaceae

Increased substitution rates and elevated Ka/Ks ratios for similar sets of plastid genes have been reported for several plant species. However, it remains uncertain whether these patterns reflect positive selection, relaxed purifying selection, changes in underlying mutation rates, a breakdown in DNA repair mechanisms, such as gene conversion, or some combination of these (Barnard-Kubow et al., 2014).

Concerning the evolutionary patterns of the ycf2 gene of C. xanthocarpa within Myrtaceae, the results of the present study showed evidence of purifying selection acting over this region. The number of segregating sites, not surprisingly, revealed the higher polymorphism when comparing C. xanthocarpa to species of tribe Eucalypteae and Syzygieae. For the species of tribe Myrteae, the highest polymorphism was observed between C. xanthocarpa and P. guajava, supporting the phylogenetic relationship recovered by the ycf2 gene alone, and also using the six plastid genes with high diversity (Figure 4).

In relation to tribe Eucalypteae and Syzygieae (Ka/Ks ratios < 0.5314), the pattern of purifying selection is more evident, towards equilibrium in relation to other species from the Myrteae tribe (Ka/Ks ratio > 0.7127). Machado et al. (2017) observed a Ka/Ks ratio of 0.30 for this gene by comparing the plastomes of the close-related Myrtaceae species A. sellowiana and E. uniflora, evidencing purifying selection. Evidence of relaxed purifying selection over the ycf2 gene was also reported for Campanulastrum americanum (Campanulaceae) by Barnard-Kubow et al. (2014). However, the estimated Ka/Ks ratio is a mean over the full length of the gene and, considering its very large size, some regions of ycf2 are likely experiencing stronger selection, while other regions are more conserved. This hypothesis has to be closely tested, but the sliding window analysis of Tajima's D estimations (data not shown) revealed such a pattern of gene regions with significant (p < 0.05) negative values within ycf2, suggesting signatures of positive selection.

Taxonomic and phylogenetic patterns of Myrtaceae as revealed by whole plastome and the ycf2 gene

The taxonomic patterns within Myrtaceae revealed by the analysis of the whole plastome sequences clearly clustered species at the tribe level, with ANI > 95%. Studies involving prokaryotic organisms with ANI values > 95% indicate that they belong to the same species (Goris et al., 2007; Richter and Rossello-Mora, 2009). In our study, we applied the ANI to investigate the identity level of the whole plastome among Myrtaceae species. This is the first time such an analysis was performed for organellar genomes, and it seems to be a useful approach, since the obtained outcomes fit the results obtained using classical phylogenetic analyses based on plastidial genes.

Similarly, the Bayesian phylogenetic inference using the plastidial ycf2 gene confirmed the monophyly of the tribes Myrteae, Eucalypteae and Sysygieae, as already suggested through phylogenetic inferences based on plastidial and nuclear genes (Sytsma et al., 2004; Wilson et al., 2005; Biffin et al., 2010; Thornhill et al., 2012, 2015), as well as through combined analysis of plastidial DNA regions (matK and ndhF) (Biffin et al., 2010), 78 protein-coding and four rRNA genes (Jo et al., 2016), 57 plastidial protein-coding genes (Huang et al., 2013; Eguiluz et al., 2017), and complete plastidial sequences (Machado et al., 2017). Also, the topology of the tribe Eucalypteae is equivalent to that proposed based on plastidial genomes (Bayly et al., 2013, 2016) and on nuclear ribosomal ITS sequences (Parra-O et al., 2006). The ycf2 gene was one of the six genes with higher polymorphism at family and at tribe levels (Figure 5), corroborating the potential of this plastid region for species-level DNA barcoding, as proposed by Kumar et al. (2009).

The internal topology of the tribe Myrteae phylogeny obtained with the ycf2 gene needs to be evaluated with caution, because it is the most species-rich tribe within Myrtaceae (Thornhill et al., 2015), and we have the ycf2 sequence from just six species representing this taxon. At a higher level, the phylogeny presented by Lucas et al. (2007), based on ITS, ETS, psbA-trnH and matK sequences places the Pimenta group as sister of the Eugenia group, while the Plinia group is placed externally to these two groups. In the ycf2 phylogeny, the species from Pimenta (Acca sellowiana, Campomanesia xanthocarpa, Psidium guajava and Pimenta dioica) and Eugenia (Eugenia uniflora) groups were not separated, and Plinia trunciflora was placed externally, as in the phylogeny presented by Lucas et al. (2007).

The shortest plastome and longest gene within Myrtaceae

Addressing three main questions in this study concerning the structure and evolution of the plastome of C. xanthocarpa, we highlighted some relevant features. First, although the plastome of C. xanthocarpa conserves the same general structure as in other 47 studied species, regarding number and position of genes, it is the shortest recorded plastidial genome within the family.

Second, the ycf2 gene of C. xanthocarpa is the longest among the Myrtaceae species that had sequences of this gene deposited in GenBank at the time of this study. Signatures of moderate purifying selection were observed for the ycf2 gene of C. xanthocarpa within Myrtaceae, more apparent in relation to tribe Eucalypteae and tending to equilibrium relative to tribe Myrteae. In addition, the ycf2 gene revealed a robust phylogenetic signal at the family level, generating a Bayesian inference of phylogenetic relationships equivalent to the taxonomic classification presented using the whole plastome sequences and average nucleotide identity analysis.

Although in the starting steps, these findings have important implications for thinking about the genetics, evolution, conservation, breeding, and biotechnology of C. xanthocarpa, a fruit tree species with high biotechnological and agricultural potential that is still underexploited. Understanding the evolutionary and taxonomic/phylogenetic relationships of this species relative to other species from Myrtaceae enables the elaboration of conservation, breeding and biotechnology programs with a consistent scientific basis. Thus, enterprises towards safeguarding and managing the species’ genetic resources, as well as selecting and developing cultivars for agricultural or biotechnological uses will be greatly benefited by the results of this study. For instance, plastid SSR markers for C. xanthocarpa will be soon released by our group, with direct applicability for marker assisted selection.

Supplementary material

The following supplementary material is available for this article: