Comparative genomics revealed new insights into the plastome evolution of Ludwigia (Onagraceae, Myrtales)

Abstract

The primrose-willow (Ludwigia L.), a well-defined genus of the Onagraceae family, comprises 87 species widely distributed worldwide. In this study, we sequenced and characterized the complete chloroplast (cp) genomes of three species in the genus, including Ludwigia adscendens, Ludwigia hyssopifolia, and Ludwigia prostrata. Three Ludwigia cp genomes ranged from 158,354 to 159,592 bp in size, and each contained 113 genes, including 79 unique protein-coding genes (PCGs), four rRNA genes, and 30 tRNA genes. A comparison of the Ludwigia cp genomes revealed that they were highly conserved in gene composition, gene orientation, and GC content. Moreover, we compared the structure of cp genomes and reconstructed phylogenetic relationships with related species in the Onagraceae family. Regarding contraction/expansion of inverted repeat (IR) region, two kinds of expansion IR region structures were found in Oenothera, Chamaenerion, and Epilobium genera, with primitive IR structures in Ludwigia and Circeae genera. The regions clpP, ycf2, and ycf1 genes possessed highly divergent nucleotides among all available cp genomes of the Onagraceae family. The phylogenetic reconstruction using 79 PCGs from 39 Onagraceae cp genomes inferred that Ludwigia (including L. adscendens, L. hyssopifolia, L. prostrata, and Ludwigia octovalvis) clade was monophyletic and well-supported by the bootstrap and posterior probability values. This study provides the reference cp genomes of three Ludwigia species, which can be used for species identification and phylogenetic reconstruction of Ludwigia and Onagraceae taxa.

Introduction

The Ludwigia (Onagraceae family) genus, commonly called primrose-willow, comprises 87 herbaceous and shrubby plants distributed in tropical and subtropical regions. ¹ According to a revised classification by Wagner et al., the Onagraceae family was divided into 22 genera based on morphological and molecular data. ² The family is categorized into two subfamilies, of which subfam. Ludwigioideae consists of a single genus, Ludwigia. ² The species within this genus display specific characteristics of adaptation to aquatic habitats, such as morphological changes that allow them to thrive in floating or permanently flooded conditions.^2–4 Ludwigia has received in-depth investigations into cytology, ⁵ morphology,^6–8 reproductive traits, ⁹ wood and leaf anatomy, ¹⁰ environment adaptation, ¹¹ chemical isolation,^12,13 biological activities of antitumor and antibacterial properties,^14–16 environmental water improvement, ¹⁷ and molecular data.^18–26

Chloroplast (cp) is a vital organelle in plant cells, responsible for carrying out photosynthesis. This process harnesses the sunlight energy and converts it into high-energy molecules, which are then used to synthesize macro-biomolecules in living cells. ²⁷ The organelle possesses its own genome which typically has a double-stranded, circular structure with highly conserved gene order and content, especially in land plants. The cp genomes range from 120 to 217 kb in size, exhibit a modest evolutionary rate, haploid, uniparental inheritance (generally maternal) and rarely undergo recombination processes.^28–30 The recent introduction of next-generation sequencing technologies, such as Illumina, Ion Torrent, MGI, PacBio, and Nanopore, has made it more feasible to obtain a huge amount of cp genome data. The whole cp genome has been suggested to serve as a single superbarcode in replacement of discrete cp regions (rbcL and matK) for plant species identification.^31–34 Hence, the cp genome data has become a powerful tool for studying the phylogenetic relationships within a diverse range of plant taxa.^{28,29,35–37}

In addition to changes in nucleotide composition, cp genomes exhibit many structural variations, including gene loss, contraction and expansion of inverted repeat (IR), inversion, and translocation. ³⁸ These variations may provide potentially useful information insights into the cp genome evolution.^38,39 Previous studies identified several structural variations within some genera of the Onagraceae family. These include two types of IR region expansion in Epilobium, Chamaenerion, and Oenothera, a presence of a 56 kb rbcL-trnQ-UUG region in Oenothera subsect. Oenothera, and an intron loss in the clpP gene in Oenothera.^24,40 These variant structures appeared in Onagraceae as a results of inheritance of the cp genome that happens in each genus (maternal and biparental inheritance).^41–43

Liu et al. classified Ludwigia species into 23 different sections based on molecular data of two nuclear loci (ITS and waxy) and five cp loci (rps16, rpl16, trnL-trnF, and trnL-trnG). ¹⁸ Additionally, several molecular phylogenetic analyses have been conducted to elucidate the evolutionary relationships of Ludwigia and other genera within the Onagraceae family.^{19–22,25,26} The complete cp genome of Ludwigia octovalvis was reported by Liu et al. in 2016, making it the first cp genome available in the Ludwigia genus. ⁴⁴ Recently, Zhang et al. and Luo et al. integrated the newly sequenced cp genomes and reconstructed evolution trees to explore the relationship of species within the Onagraceae family.^23,24 In the present study, we sequenced and characterized the complete cp genomes of three Ludwigia species including Ludwigia adscendens, Ludwigia hyssopifolia, and Ludwigia prostrata collected in Vietnam. The three newly sequenced cp genomes were used to perform comparative genomic analysis with other cp genomes of Onagraceae species, and to reconstruct the phylogenetic tree of Onagraceae family to determine the phylogenetic position of Ludwigia.

Materials and methods

Sample collection, DNA extraction, and sequencing

The leaves of L. adscendens, L. hyssopifolia, and L. prostrata were collected from July to September 2019 from Kien Giang Province, Vietnam. The fresh leaves were washed and dried with tissue paper before being stored at 4°C from 3 to 7 days before further processing. The whole plants were collected and stored in the herbarium at NTT Hi-tech Institute, Nguyen Tat Thanh University (number of specimens: NTT_2019.09_01–03, contact person: dhdkhoa@ntt.edu.vn). The identities of the three species were identified based on their morphological characteristics (Supplemental Figures S1, S2, and S3) according to the previous descriptions. ⁴⁵

The chloroplast fraction was enriched from fresh leaves using the sucrose gradient (20%/45%) method. ⁴⁶ The cetyltrimethylammonium bromide method was employed to extract DNA from cp-enriched fragment. ⁴⁷ The extracted DNA samples were purified by using the Monarch Genomic DNA Purification Kit (#T3010, New England Biolabs) according to the manufacturer's instructions. Following the assessment of quality using the Nanodrop OneC spectrophotometer (Thermo Fisher Scientific), the DNA samples were subjected to sequencing on the MiniSeq system (Illumina), which generated 150 bp paired-end reads at KTest Science Co. Ltd. in Ho Chi Minh City, Vietnam (www.ktest.com.vn).

Chloroplast genome assembly and annotation

The raw read data was used for the de novo assembly of cp genomes using NOVOPlasty version 4.3.1 ⁴⁸ and further assessed by Geneious Prime version 2023.01.01 to obtain complete cp genomes and coverage depths. The assembled cp genomes were first annotated using GeSeq tool. ⁴⁹ Then, the annotation of transfer RNA (tRNA) genes was independently verified using tRNAscan-SE2 with “Mixed (general tRNA model)” option for the sequence source, and “Default” option for the search mode. ⁵⁰ Annotations of protein-coding genes (PCGs) were manually checked and reconfirmed using homologous genes identified through the BLAST tool available on NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The complete annotated cp genomes of the three species were submitted to the GenBank database with the accession numbers of NC_081012 for L. adscendens, NC_081013 for L. hyssopifolia, and NC_081014 for L. prostrata. The OrganellarGenomeDRAW (OGDRAW) version 1.3.1 was used to illustrate the cp genome maps. ⁵¹

Characterization of cp Genome and IR expansion/contraction

The cp genomes of three Ludwigia species were compared with seven available cp genomes of Onagraceae species, namely L. octovalvis, Circaea cordata, Epilobium cylindricum, Chamaenerion angustifolium, Oenothera argillicola, Oenothera lindheimeri, and Oenothera villaricae. The length and GC content of the entire cp genomes, single copy (SC) regions, and IR regions were compared among the 10 selected species of Onagraceae. Furthermore, the cp genome content, including exons, introns, and intergenic spacers (IGSs) across these Onagraceae species was compared using the mVISTA web server. ⁵² Gene organization among species was determined using the MAUVE alignment algorithm, integrated into the Geneious Prime software. The adjacent genes to the IR/SC junction were visualized using IRscope. ⁵³

Analysis of SSR, long repeat, and relative synonymous Codon usage

The amino acid frequency and relative synonymous codon usage (RSCU) values of the PCGs in cp genomes were determined using CodonW (https://codonw.sourceforge.net/). The RSCU values were shown in a stack column chart using ggplot2 in the R environment. ⁵⁴ The Phobos package integrated into Geneious Prime software was employed to find simple sequence repeats (SSRs), with the minimum length of SSRs being as follows: 10 for mononucleotide, 12 for dinucleotide, 16 for trinucleotide, 20 for tetranucleotide, 25 for pentanucleotide, and 24 for hexanucleotide. The online tool REPuter was utilized to identify four types of oligonucleotide repeats, namely forward (F), complementary (C), reverse (R), and palindromic (P), with the following parameters: minimum repeat size ≥ 24 and hamming distance = 3. ⁵⁵

Analysis of synonymous/nonsynonymous substitution and nucleotide diversity

The synonymous (K_s) and nonsynonymous (K_a) substitutions for 79 shared PCGs were computed using DnaSP version 6.12.03 ⁵⁶ for the cp genomes of three Ludwigia species, with Epilobium cyclindricum and Oenothera curtiflora as references. For predicting gene selection, K_a/K_s ratios were determined with values below 1 indicating purifying selection, a ratio of 1 representing neutral selection, and values over 1 suggesting positive selection.

Nucleotide diversity levels of Onagraceae family and Ludwigia genus were calculated and compared. In total, the 36 cp genomes of Onagraceae species were aligned using MAUVE-alignment; a list of 36 Onagraceae species was presented in Supplemental Table S1. Regarding nucleotide divergence of Ludwigia genus, cp genomes of four Ludwigia (L. adscendens, L. hyssopifolia, L. octovalvis, and L. prostrata) were aligned. The nucleotide diversity level was measured by Pi values calculated using DnaSP6 software, with the parameters of window length—800 bp and step size—200 bp.

Phylogenetic analysis among Onagraceae species

For reconstructing phylogeny, we retrieved the available cp genomes of 37 species including 33 Onagraceae species, the three newly sequenced Ludwigia species, and a Heimia apetala from the family Lythraceae as an outgroup from the GenBank database. The sequences of 79 PCGs from the 37 cp genomes were extracted and aligned using the MAUVE-alignment algorithm in Geneious Prime.

The maximum likelihood (ML) tree was reconstructed by IQ-tree ⁵⁷ with a General time-reversible model with gamma distribution as the best fit model (Akaike information), which was predicted by jModelTest. ⁵⁸ The Bayesian inference (BI) phylogenetic tree was reconstructed by using MrBayes version 3.2.7a following the same evolutionary model. ⁵⁹ The ML and BI phylogenetic trees were reconstructed with the same parameters as described previously. ⁶⁰ The Figtree version 1.4.4 was used to illustrate the topology and bootstrap/ probability properties for each node in the phylogenetic trees. ⁶¹

Results

The features of Ludwigia cp genomes

Using the Illumina platform, we obtained the sequences of whole cp genomes of L. adscendens, L. hyssopifolia, and L. prostrata with an average coverage depth of 327×, 336×, and 270×, respectively. The sizes of L. adscendens, L. hyssopifolia, and L. prostrata cp genomes were 159,592 bp, 158,354 bp, and 158,660 bp, respectively (Figure 1). The cp genomes had a typical four-segment structure, which consisted of a large SC region (LSC (89,323–90,292 bp)) and a small SC region (SSC (19,535–19,776 bp)), separated by a couple of IR regions (24,748–24,776 bp). The overall GC content in the cp genomes of three Ludwigia species was approximately 37.3%, which was similar to that of other species in the Onagraceae family.

Figure 1.

The cp genome maps of three Ludwigia species. The largest circle depicts the gene location and orientation, outer genes are transcribed clockwise, and insider genes are transcribed counterclockwise. Different functional group genes are color-coded, with the legend below the gene map. In a small circle, the darker gray color and the lighter gray color correspond to the GC content and the AT content of the cp genomes, respectively. SSC: small single copy; LSC: large single copy; IRa, IRb: inverted repeat regions.

The three newly sequenced cp genomes were highly conserved in gene content and orientation. The cp genome of Ludwigia species contains 113 individual genes, which consisted of 79 PCGs, 30 tRNA genes, and four rRNA genes (Table 1). Among these, seventeen genes were duplicated due to their presence in the IR regions. These duplicated genes included seven PCGs (rpl2, rpl23, rps7, rps12, ndhB, ycf15, and ycf2), four rRNA genes (rrn4.5, rrn5, rrn16, and rrn23), and seven tRNA genes (trnA-UGC, trnI-GAU, trnI-CAU, trnL-CAA, trnN-GUU, trnR-ACG, and trnV-GAC). Nine PCGs (atpF, ndhA, ndhB, rpl2, rpl16, petD, rps16, rpoC1, and petB) contain a single intron each, while two genes (clpP and ycf3) are characterized by the presence of two introns.

Table 1.

List of genes in the chloroplast genomes of three Ludwigia species.

Gene categories	Gene groups	Gene names	No. of genes
Self-replication	Ribosomal RNAs	rrn4.5 (2x), rrn5 (2x), rrn16 (2x), rrn23 (2x)	8
	Transfer RNAs	trnA-UGC* (2x), trnC-GCA, trnD-GUC, trnE-UUC, trnF-GAA, trnG_UCC*, trnG-GCC, trnH-GUG, trnI-CAU, trnI-GAU* (2x), trnK-UUU*, trnL-UAA*, trnL-UAG, trnL-CAA (2x), trnfM-CAU, trnM-CAU (2x), trnN-GUU, trnP-UGG, trnQ-UUG, trnR-UCU, trnR-ACG (2x), trnS-GCU, trnS-UGA, trnS-GGA, trnT-GGU, trnT-UGU, trnV-UAC*, trnV-GAC (2x), trnW-CCA, trnY-GUA	37
	Large subunit ribosome	rpl2* (2x), rpl14, rpl16*, rpl20, rpl22, rpl23 (2x), rpl32, rpl33, rpl36	11
	Small subunit ribosome	rps2, rps3, rps4, rps7 (2x), rps8, rps11, rps12* (2x), rps14, rps15, rps16*, rps18, rps19	14
	DNA dependent RNA polymerase	rpoA, rpoB, rpoC1*, rpoC2	4
Photosynthesis	ATP synthase subunit	atpA, atpB, atpE, atpF*, atpH, atpI	6
	Photosystem I subunit	psaA, psaB, psaC, psaI, psaJ, ycf3*, ycf4	7
	Photosystem II subunit	psbA, psbB, psbC, psbD, psbE, psbF, psbH, psbI, psbJ, psbK, psbL, psbM, psbN, psbT, psbZ	15
	Cytochrome b/f complex subunit	petA, petB*, petD*, petG, petL, petN	6
	NADH dehydrogenase	ndhA*, ndhB* (2x), ndhC, ndhD, ndhE, ndhF, ndhG, ndhH, ndhI, ndhJ, ndhK	12
	Rubisco subunit	rbcL	1
Other genes	ATP-dependent protease subunit P	clpP*	1
	Envelope membrane protein	cemA	1
	Acetyl-CoA-carboxylase subunit	accD	1
	Maturase	matK	1
	C-type cytochrome synthesis gene	ccsA	1
	Translational initiation factor	infA	1
Unknown function	Conserved open reading frame	ycf1 (2x), ycf2 (2x)	4

Note: The asterisk “*”—gene containing introns; (2x)—duplicated gene.

Contraction and expansion of SC/IR junctions

We found that the size of IR regions varied in the Onagraceae cp genomes; and adjacent genes in the IR/SC junctions of these cp genomes were also different (Figure 2). These variations occurred due to the expansion of IR regions in the cp genomes of Epilobium, Chamaenerion, and Oenothera species, otherwise, the LSC/IR junctions were similar in other species of the Onagraceae family. Eight genes (consisting of rps19, rpl2, trnN-GUU, ndhF, rpl32, ccsA, ndhD. and ycf1) were found in the LSC/IR and the SSC/IR borders of the candidate species. Based on genes related to the SC/IR junctions, the structure of IR regions in Onagraceae was divided into three main types. In Ludwigia and Circeae species, 17 genes were found in the IR regions, ranging from 24,748 bp to 24,776 bp. In Chamaenerion, Epilobium, and most Oenothera species, an additional gene (ndhF) was found in the IR regions, bringing the total number of genes to 18. As a result, the length of these IR regions was much longer than those in other cp genomes of Ludwigia and Circeae species, ranging from 27,400 bp (C. angustifolium) to 28,773 bp (O. argillicola). The cp genomes of Oenothera subsect. Munzia species (O. villaricae and O. picensis subsp. picensis) exhibited the largest cp genome sizes in the Onagraceae family due to the expansion of the IR region into three genes (including rpl32, trnL-CAA, and ccsA). That resulted in the number of genes and the length of IR regions being 21 and 30,884 bp, respectively, in the cp genomes of Oenothera subsect. Munzia species.

Figure 2.

Comparison of SC/IR junctions of Onagraceae cp genomes. The cp genome structure of Onagraceae species consists of four regions denoted as LSC (blue), SSC (green), IRa and IRb (orange), with lengths shown for each region. Adjacent genes are illustrated with different colored codes; genes are located above or below the cp genome, depending on the transcriptional direction of the genes. The arrows indicate the distance between genes and junctions. JLB: junction of LSC/IRb; JSB: junction of IRb/SSC; JSA: junction of SSC/IRa; JLA: junction of IRa/LSC.

Repeat analysis and RSCU

A total of 39, 30, 49, and 28 SSRs were identified in the cp genomes of L. adscendens, L. hyssopifolia, L. prostrata, and L. octovalvis, respectively (Table 2). We found that the majority of mononucleotide SSRs consisted of either poly A or poly T repeats, whereas all dinucleotide SSRs were either AT or TG repeats. Tetranucleotide was only detected in L. adscendens and L. hyssopifolia but not in other species. The majority of identified SSRs were found in IGS regions, including 28 (71%), 22 (73%), 30 (61%), and 18 (64%) SSRs in the cp genomes of L. adscendens, L. hyssopifolia, L. prostrata, and L. octovalvis, respectively.

Table 2.

Quantities of SSRs and long repeats identified in the chloroplast genomes of Ludwigia species.

	Repeat unit	Ludwigia adscendens	Ludwigia hyssopifolia	Ludwigia prostrata	Ludwigia octovalvis
SSR type
Mono-	A/T	35	25	48	27
	G/C	1	2	-	-
Di-	AT	2	2	-	1
	TG	-	-	1	-
Tetra-	CAAC	1	-	-	-
	TTAT	-	1	-	-
Total		39	30	49	28
Location of SSR
Exon		7	6	7	3
Intron		4	2	12	7
IGS		28	22	30	18
Long repeat type
Forward		10	8	7	9
Palindromic		1	2	2	2
Total		11	10	9	11
Location of long repeat
Exon		5	3	3	5
Intron		1	4	3	2
IGS		5	3	3	4

The dash “-” means “not found.”

All four Ludwigia species possessed the similar number of long repeats in their cp genomes. The cp genomes of L. adscendens, L. hyssopifolia, L. prostrata, and L. octovalvis had a total of 11, 10, 9, and 11 long repeats, respectively. These long repeats were categorized into two types: forward repeats and palindromic repeats. Forward repeats were the most common in the cp genomes of Ludwigia species (10 in L. adscendens cp genome, eight in L. hyssopifolia cp genome, seven in L. prostrata cp genome, and nine in L. octovalvis cp genome). For palindromic repeats, we found one in L. adscendens cp genomes and two in three other Ludwigia cp genomes. These long repeats were evenly distributed in three regions, namely exon, intron, and IGS, in L. hyssopifolia, L. prostrata, and L. octovalvis cp genomes. In the L. adscendens cp genome, an intron region had a single long repeat, while the exon and IGS regions each contained five long repeats.

Codon usage bias refers to the uneven use of synonymous codons during mRNA translation. This is a common phenomenon observed in most organisms, and it plays an important role in determining the efficiency and accuracy of protein synthesis.^62,63 The RSCU in the genomes of many species is not random, ⁶⁴ but rather influenced by mutation, natural selection, and genetic drift.^65,66 Codon usage plays an important role in the evolution of cp genomes. ⁶⁷ Therefore, we analyzed the RSCU of all PCGs and determined the frequency of each amino acid present in the cp genomes of Ludwigia species. The total number of codons was 22,790 in L. adscendens, 22,740 in L. hyssopifolia, and 22,735 in L. prostrata (Figure 3). Of these, Leucin codon was most frequently used in the three species (L. adscendens—2422; 10.63%, L. hyssopifolia—2419; 10.64%, and L. prostrata—2408; 10.60%). Meanwhile, cysteine was the least commonly used amino acid (L. adscendens—253; 1.11%, L. hyssopifolia—256; 1.13%, and L. prostrata—252; 1.11%). The two amino acids methionine and tryptophan showed no biased usage (RSCU = 1), since they were each decoded by a single codon. Besides, the RSCU ratios of all amino acids (except Met and Trp) were greater than one, indicating a deviation in codon usage. Most preferred codons (28 in total) end with either A or U.

Figure 3.

The RSCU values of 79 PCGs in the cp genomes of three Ludwigia species. The codons and their RSCU values are signed by color groups, with the highest RSCU value signed by blue, followed by red, brown, yellow, purple, and dark blue. The RSCU values of different codons encoding the same amino acid are illustrated by a stacked column. The three columns are arranged from left to right as follows: L. adscendens, L. hyssopifolia, and L. prostrata.

Comparative chloroplast genome analysis of Ludwigia and related species

The cp genome comparison demonstrated that the coding regions were more conserved than noncoding regions among the nine Onagraceae species (Figure 4). The highest divergence was found among the IGSs, including trnH-psbA, matK-rps16, rps16-psbK, psbI-atpA, rpoB-psbD, psaA-rps4, rps4-ndhJ, ndhC-atpE, rbcL-accD, and psbE-psaJ in LSC; and ndhF-rpl32, rpl32-ccsA, ccsA-ndhD, and ndhG-ndhI in SSC. In addition, a high level of nucleotide divergence was also found in the coding regions of clpP, accD, rps3-rps19, ycf2, ndhF, ndhD, and ycf1. Among the highly diverse cp PGCs in examined species in the genera Ludwigia, Circaea, Epilobium, and Chamaenerion, the clpP gene exhibited the highest variation (Figure 5). Nevertheless, the clpP gene structure remains stable, consisting of three exons and two introns across the four genera. The notable structural variation of the clpP gene was observed in the Oenothera genus, in which the gene lost its second intron in the Oenothera sect. Gaura (O. lindhermeri). Meanwhile, the clpP gene in Oenothera sect. Oenothera (O. villaricae and O. argillicola) lost both introns and was left with a single coding sequence constituted by three exons. Besides, the mVISTA alignment result revealed a large gap between trnQ-UUG and rbcL region (56 kb) in the cp genome of O. argillicola (subsect. Oenothera). This gap in the alignment resulted from the inversion of the trnQ-UUG-rbcL region in the cp genomes of O. argillicola (Supplemental Figure S4).

Figure 4.

Alignment of ten Onagraceae cp genomes using mVISTA program. The top of the graph shows the name and orientation of genes (depending on transcriptional direction, shown by arrows). A cut-off of 70% identity was used for the plots, and the Y-scale indicates the percent identity between 50% and 100%. The X-scale denotes location in cp genomes. The cp genome regions are color coded as exon, intron, and conserved noncoding sequences (CNS).

Figure 5.

Structural comparative analysis of clpP gene among 10 Onagraceae species. The numbers next to the species names indicate the length of the gene. The alignment of the clpP gene consists of gray box (homologous sequences), black box—(nonhomologous sequences), and line (gaps). The yellow boxes below the sequences indicate the coding sequence of the clpP gene.

Nucleotide diversity among Onagraceae species

The results of nucleotide divergence analysis showed a high level of divergence in nucleotides among Ludwigia species (Figure 6A) as well as Onangraeae species (Figure 6B). In Ludwigia genus, several hotspots were discovered, including psbA-matK (Pi = 0.03), accD (Pi = 0.032), clpP (Pi = 0.056), ycf2 (Pi = 0.028), ndhF (Pi = 0.022), ccsA (Pi = 0.025) and ycf1 (Pi = 0.054). Whereas in the Onagracaea, hotspot regions were also detected in accD (Pi = 0.175), clpP (Pi = 0.14), ycf2 (Pi = 0.11), ndhF - rpl32 (Pi = 0.09), and ycf1 (Pi = 0.24). The data clearly showed that, some similarity in divergence hotspot regions were found in accD, clpP (in the LSC region), ycf2 (in the IR region), ndhF, and ycf1 (in the SSC region) at both family and genus levels. This reflected that the cp genome evolution process of the Ludwigia genus is conserved in the Onagraceae family. Thus, these cp regions could be evaluated to develop potential molecular markers for plant identification and resolving complex relationships of Ludwigia genus, it could also be expanded to apply for species delimitation of the Onagraceae family.

Figure 6.

Nucleotide divergence in cp genomes of Onagraceae. (A) within four Ludwigia species. (B) within 37 species of the Onagraceae family. The horizontal axis represents the nucleotide position on the cp genome, the vertical axis indicates the nucleotide diversity (Pi value). The black line represents the Pi value in each region. Divergent hotspots are signed by the name of gene at each peak.

Codon selective pressure analyses

The results showed that the K_a/K_s values are similar when comparing both E. cylindricum and O. curtiflora (Figure 7A, B, Supplemental Table S2). It revealed that most of the genes in the three Ludwigia cp genomes had a K_a/K_s ratio of less than 0.5, meaning that nucleotide substitutions lead to amino acid changes in these genes, which were eliminated in evolution. In contrast, three out of 79 PCGs genes (atpE, petA, and rps12) had positive selection, with the K_a/K_s ratios being greater than 1, suggesting that mutations that replace amino acids encoding in the codons of these genes are being rapidly evolved due to beneficial changes and are retained by natural selection.

Figure 7.

The K_a/K_s ratio of 79 shared PCGs of cp genomes. (A) Three Ludwigia species compared with Epilobium cylindricum. (B) Three Ludwigia species compared with Oenothera curtiflora. The genes on the cp genome are denoted by the numbers, including (1)—accD, (2)—atpA, (3)—atpB, (4)—atpE, (5)—atpF, (6)—atpH, (7)—atpI, (8)—ccsA, (9)—cemA, (10)—clpP, (11)—infA, (12)—matK, (13)—ndhA, (14)—ndhB, (15)—ndhC, (16)—ndhD, (17)—ndhE, (18)—ndhF, (19)—ndhG, (20)—ndhH, (21)—ndhI, (22)—ndhJ, (23)—ndhK, (24)—petA, (25)—petB, (26)—petD, (27)—petG, (28)—petL, (29)—petN, (30)—psaA, (31)—psaB, (32)—psaC, (33)—psaI, (34)—psaJ, (35)—psbA, (36)—psbB, (37)—psbC, (38)—psbD, (39)—psbE, (40)—psbF, (41)—psbH, (42)—psbI, (43)—psbJ, (44)—psbK, (45)—psbL, (46)—psbM, (47)—psbN, (48)—psbT, (49)—psbZ, (50)—rbcL, (51)—rpl14, (52)—rpl16, (53)—rpl2, (54)—rpl20, (55)—rpl22, (56)—rpl23, (57)—rpl32, (58)—rpl33, (59)—rpl36, (60)—rpoA, (61)—rpoB, (62)—rpoC1, (63)—rpoC2, (64)—rps11, (65)—rps12, (66)—rps14, (67)—rps15, (68)—rps16, (69)—rps18, (70)—rps19, (71)—rps2, (72)—rps3, (73)—rps4, (74)—rps7, (75)—rps8, (76)—ycf1, (77)—ycf2, (78)—ycf3, (79)—ycf4. Genes harbor K_a/K_s values greater than 1 which was in bold.

Phylogenetic relationships between Ludwigia and related taxa

The sequences of cp genomes are widely employed for reconstructing phylogenetic trees and studying population genetics, due to their simplicity and rich evolutionary information. ⁶⁸ The results of phylogenetic analysis revealed that all Onagraceae species constituted a monophyletic group with robustly supported values (bootstrap and posterior probability), which was consistent with the present taxonomic data. The topology of phylogenetic trees also agreed with other phylogenetic studies based on the cp genomes of Onagraceae species.^24–26

A total of 36 Onagraceae species were classified into two clades which correspond to two subfamilies, Ludwigoideae and Onagroideae (Figure 8). Ludwigia is a unique genus in the Jussiaceae tribe, Ludwigoideae subfamily. In contrast, the tribes Onagraceae, Epilobeae, and Ciraeeae had close relationships and were grouped into the Onagroideae subfamily. In the genus Oenothera, there are two sections, including O. sect. Oenothera and O. sect. Gaura. Within section O. sect. Oenothera, there are two subsections, namely O. subsect. Oenothera and O. subsect. Munzia. These genera/sections/subsections were also divided according to evolutionary events that occurred in the cp genomes of these species in a previous study. ²⁴

Figure 8.

Phylogenetic trees inferred from the sequence of 79 PCGs in the cp genomes of 37 Onagraceae species using ML and BI methods. The species marked in red are assembled cp genome species in this study. The bootstrap and posterior probability values are represented beside each node. The scale bar represents the number of nucleotide substitutions per site.

The four species in the genus Ludwigia formed a well-supported clade in Jussiaeeae tribe, Ludwigoideae subfamily that considered early diverged genera in Onagraceae. ²⁴ In which, two species (L. octovalvis and L. adscendens) formed a group and were closely related. Our findings additionally demonstrated that using cp genomes can offer a reliable approach for identification and a better understanding of plant species relationships.

Discussion

This study reported the complete cp genomes of L. adscendens, L. hyssopifolia, and L. prostrata, ranging from 158,354 bp to 159,592 bp in length. Three cp genomes exhibited a typical quadripartite conserved structure (LSC/IRa/SSC/IRb), with 131 genes annotated, including 86 PCGs, 37 tRNA genes, and eight rRNA genes. The annotated genes in these genomes are similar to previously reported angiosperm cp genomes.^28–30 Our findings show that the four cp genomes of this genus have no structural rearrangements. The cp genomes in angiosperms are conserved in terms of the number and order of genes across different lineages. ⁶⁹ In general, gene arrangement and variation in the structure of cp genomes are uncommon in most angiosperms. ⁶⁹ The rearrangements with large regions have rarely been detected, but they infrequently occurred in a few lineages, namely Asteraceae, ⁷⁰ Campanulaceae, ⁷¹ Fabaceae, ⁷² Geraniaceae, ⁷³ Plantaginaceae, ⁷⁴ and Poaceae. ⁷⁵ In the Onagraceae family, cp genome conservation was observed, albeit with some minor changes (including IR expansion, inversion, and loss of genes), which provided essential information for further studies into the evolution of cp genomes. In Oenothera subsect. Oenothera, the cp genomes contained a large inversion from rbcL-trnQ-UUG (approximately 56 kb) in the LSC region, which was previously reported by Greiner et al. (2008). ⁴⁰

The dynamics in IR regions play an important role in evolutionary process.^76,77 The IR expansions have been found in various lineages, including as Geraniaceae, ⁷⁸ Euphorbiaceae, ⁷⁹ Solanaceae, ⁸⁰ and Bignoniaceae. ⁸¹ Compared to a typical angiosperm genome, in Ludwigia cp genomes, the IR region was similar in size and gene content, which extended from the rps19 gene to trnN-GUU with a total of 17 genes. In the Onagraceae family, the length of cp genomes displayed variation, ranging from 158.3 kb to 165.8 kb, while the number of genes in cp genomes was also diverse, ranging from 131 to 141 genes. The main reason for these variations was the expansion of the IR regions into the SSC region.

The phenomenon of gene loss or reduction in the cp genome is a common occurrence in many parasitic plant species, as mentioned by a previous study. ⁸² In the case of the nonparasitic Onagraceae family, we conducted whole cp genome alignments and identified instances of gene reduction within the clpP. The clpP gene encodes a component of an ATP-dependent protease that is essential for cp biogenesis. ⁸³ In previous research, the degradation of introns in clpP gene was identified in Geraniaceae, Hypericaceae, mimosoid legumes, and tobacco.^83–86 These studies found that the clpP gene was lost in various Oenothera species. This event has not been observed in any other genera within the Onagraceae family. Thus, more research is required to comprehend the molecular mechanisms of the loss of clpP gene in Onagraceae family, especially the Oenothera genus.

In previous studies, some loci present in the cp genomes were found to be effective to infer phylogenetic relationships in Onagraceae family, such as ITS, phyC, waxy, matK, trnH-psbA, rbcL, rps16, rpl16, rpl32-trnL, trnL-trnF, and trnL-trnG.^11,18,21 Based on a comparison of 36 Onagraceae cp genomes, we identified several specific regions (accD, clpP, ycf2, ndhF, and ycf1) that exhibited the highest level of nucleotide divergence among the compared species. These hotspot regions could serve as candidate barcoding regions in phylogenetic analyses, population genetics, and species identification. Further studies were necessary to determine if these highly divergent regions could be applied. On the other hand, a part of IR (from ycf15 to trnN-GUU) was more stable than ones in LSC and SSC regions.^23,87 However, the gene ycf2 displayed high nucleotide divergence, which was similar to the levels of divergence observed in the LSC and SSC regions.

With the advancement of sequencing technology, it is now possible to fully unveil the entire sequence and features of every cp genome, providing the groundwork for genetic classification and a thorough understanding of plant evolution.^28,29,39 However, the systematic relationships among Ludwigia taxa have been unclear, resulting from the lack of genomic data. For example, the evolutionary position of Ludwigia in the Onagraceae family was previously poorly resolved due to the use of limited loci in cp and nuclear genomes. ¹⁸ In the current study, we employed cp sequences to rebuild the phylogenetic relationships within the Onagraceae. Although not all species of Onagraceae were included, the new results of phylogenetic analysis provided useful information for elucidating the evolutionary history of Ludwigia species and Onagraceae taxa.

Conclusions

In this study, the cp genomes of three Ludwigia species were assembled and characterized, of which the total length ranged from 158,354 bp to 159,592 bp. The chloroplast genomes of Ludwigia were highly conserved regarding gene content and genome structure. Moreover, the base composition, codon usage, SSRs, long repeats, structure of IR regions, and Ks/Ks values in the Ludwigia chloroplast genomes were fully characterized. These results provided valuable genomic resources for further studies examining the evolutionary history of these plants and related species in Onagraceae. Additionally, the phylogenetic results revealed that four Ludwigia species had a close relationship and formed a well-supported clade, which provides a better understanding of relationships among Ludwigia species.