Repetitive genomic elements in Campomanesia xanthocarpa: prospection, characterization and cross amplification of molecular markers

Abstract

Repetitive genomic elements were prospected in Campomanesia xanthocarpa, aiming to characterize these elements in a non-model plant species and to develop species-specific microsatellite markers. Approximately 4.12% of the partial genome of C. xanthocarpa is composed of repetitive elements, being retrotransposons the most widely represented. A total of nine polymorphic microsatellite markers were obtained: four nuclear-neutral, two nuclear EST, two plastidial and one mitochondrial. Levels of population genetic diversity of four natural populations of C. xanthocarpa were characterized using these markers. In addition, the cross-species amplification of the microsatellite markers was tested in seven species of tribe Myrteae (Myrtaceae). The characterized microsatellite markers revealed low to moderate levels of genetic diversity (expected heterozygosity range: 0.33–0.57; observed heterozygosity: 0.26–0.74 and number of alleles: 2.25–4.25). Cross-species amplification was successful for all loci, except Cxant76. These nine markers will contribute for studies on genetic diversity, gene flow, plant selection and breeding of this species, towards the conservation of natural populations, as well as its commercial use.

Electronic supplementary material

The online version of this article (10.1007/s13205-019-1953-8) contains supplementary material, which is available to authorized users.

Introduction

Repetitive genomic elements are ubiquitous DNA sequences comprising up to 90% of the plant genomes (Mehrotra and Goyal 2014, Shcherban 2015) and are an important component of the evolutionary forces. Plant speciation is frequently accompanied by rapid changes of the repetitive DNA sequence portion (Shcherban 2015). For instance, the distinct accumulation/exclusion of repetitive DNA was considered the most important factor driving the genome size variation across 23 species of Fabaceae (Macas et al. 2015). The wide occurrence of repetitive genomic elements across the genomes has also encouraged the use of such sequences as molecular markers in plants. Most of studies about these elements are concerned with the mobile genetic elements (retrotransposons and DNA transposons) and the satellite DNAs (mini- and microsatellites). Most recently, microRNAs have also received special attention given their importance in controlling gene expression (e.g., Guzman et al. 2012).

Retrotransposons and DNA transposons are genomic elements which are able to “jump” across the genome, moving their position. While DNA transposons use a mechanism of “cut-and-paste”, just moving within the genome, retrotransposons use the “copy-and-paste” strategy, thus enlarging the number of their own copies in the genome (Stefenon et al. 2019). The presence of conserved domains in some retrotransposon families (the LTR retrotransposons), enabled the development of retrotransposon-based DNA fingerprinting techniques, based on the variation of the length between two consecutives retroelements (IRAP markers; Kalendar et al. 1999) or between a retrotransposon and a microsatellite (REMAP markers; Kalendar et al. 1999).

Microsatellite molecular markers (or simple sequence repeats—SSRs) are locus-specific, co-dominant and high polymorphic markers, considered as efficient tools for generating information to advance the genetic knowledge of non-model and non-cultivated species (Lemos et al. 2018). These loci can be found in nuclear and organellar genomes and are suitable for studying population genetics, conservation, phylogeography, gene flow, breeding and management of species (Vieira et al. 2016).

Microsatellite markers found in expressed regions of the nuclear genome (expressed sequence tags, also known as ESTs) are known as EST–microsatellites and are quite useful in studies towards genetic breeding, since they may be linked to genes of relevance, like disease-resistance genes. On the other hand, microsatellite loci located in non-expressed regions of the nuclear genome are considered neutral markers, being useful for genetic population studies, measuring neutral DNA variation. Organellar microsatellite markers, in turn, can be found in the mitochondrial or the plastid genomes and are very useful for phylogenetic, phylogeographic and population genetic studies, mainly because of the predominantly uniparental heritability and the absence of recombination (Lemos et al. 2018).

The advent of the next-generation sequencing (NGS) platforms in the last decades enabled quick and cost-effective analyses of nuclear and organellar genomes, allowing in deep exploration of genomic resources including the characterization of gene families and repetitive genomic elements (Stefenon et al. 2019), the prospection of microRNAs (Guzman et al. 2012), the characterization of proteomic profiles (Guzman et al. 2014) and the discovery of microsatellite markers (Staton et al. 2015).

The impact of repetitive DNA content on plant genomes has been studied in model and cultivated plants (Macas et al. 2007), but little is known about this issue in wild tree species. Campomanesia xanthocarpa (Mart.) O. Berg (Fig. 1) is a fruit tree species native to Brazil, Argentina, Uruguay and Paraguay (Legrand and Klein 1977). The fruits of C. xanthocarpa are an important feed source for the local fauna, while their major economic uses are domestic consumption and small-scale marketing by small farmers (Lisbôa et al. 2011). In addition, the fruits of this species are indicated as a functional food and have important medicinal properties (Viecili et al. 2014). Currently, no studies on genomics of this species have been performed, limiting our knowledge about this issue.

Fig. 1.

Campomanesia xanthocarpa. a Adult tree in its natural range of occurrence. b Voucher of the species containing ripe fruits

In this study, we explored the content of repetitive genomic elements from the draft genome of Campomanesia xanthocarpa and compared it to data from Eugenia uniflora, two tree species of the family Myrtaceae (tribe Myrtae). In addition, we characterized a set of microsatellite markers for C. xanthocarpa, including four neutral nuclear, two EST, and three organellar polymorphic microsatellite markers, and tested the cross amplification of these loci to seven other fruit species of the tribe Myrteae (Acca sellowiana, Campomanesia aurea, Eugenia pungens, Eugenia uniflora, Psidium austral, Psidium cattleianum and Psidium grandiflorum), all native to the Atlantic Forest.

Materials and methods

Total genome sequencing and assembling

Samples of young leaves of Campomanesia xanthocarpa (Myrataceae) were obtained from an adult individual collected in the Department of Botany, Federal University of Santa Catarina (27°36.094′′S, 48°31.310′′W). Total genomic DNA was isolated from fresh leaves using the CTAB method (Doyle and Doyle 1990).

The total DNA was sequenced using an NGS Illumina MiSeq platform, to generate a partial sequencing of the genome of C. xanthocarpa. The sequencing libraries were prepared using 1.0 ng of DNA with the Nextera XT DNA Sample Prep Kit (Illumina Inc., San Diego, CA), according to the manufacturer’s instructions. Libraries were sequenced using MiSeq Reagent Kit v3 (600 cycles) on the Illumina MiSeq Sequencer (Illumina Inc., San Diego, California, USA). After sequencing, CLC Genomics Workbench 8.0.1 software (Qiagen Inc.) was employed to filter the reads, to remove low quality sequences (Phred score < 20) and adapters. The same software was employed to perform a de novo assembling of the reads (mean length of 203 bases), yielding a whole genome set of assembled sequences (WGSAS). This Whole Genome Shotgun project was deposited at DDBJ/ENA/GenBank under the Bioproject PRJNA534508, BioSample SAMN11491736.

Characterization of genomic repetitive elements

Genomic repetitive elements within the C. xanthocarpa WGSAS were identified using the Smith–Waterman alignment algorithm, as implemented in the RepeatMasker web server (http://www.repeatmasker.org/cgi-bin/WEBRepeatMasker). RepeatMasker identifies repetitive elements within the evaluated genome by aligning each of the query sequence with each of the repeat consensus sequences in the repeat library file. Arabidopsis thaliana was employed as reference repeat database (RepeatMasker combined database: Dfam_Consensus-20181026, RepBase update 20181026).

Prospection of microsatellite loci

Aiming to develop molecular markers for genetic studies of C. Xanthocarpa, we further focused in the prospection and characterization of microsatellite loci using two independent sets of DNA sequences. The first set is the WGSAS, used for the characterization of genomic repetitive elements and contains sequences of both nuclear and organellar genomes. The second set of sequences (hereafter called “plastid genome”) was originated from the complete plastid sequence of C. xanthocarpa (GenBank Accession Number KY392760).

Microsatellite loci were prospected from both sets of sequences using the MISA tool (Thiel et al. 2003), considering a minimum number of 10, 5, 4, 3, 3 replicates for mono-, di-, tri-, tetra-, and pentanucleotide motifs, respectively. Primers for the microsatellite loci were designed with the software PRIMER3 (Rozen and Skaletsky 2000) by setting products size ranges from 100 to 500 bp. The possible occurrence of secondary structures as hairpin loops, dimers, bulge loops, and internal loops were verified with Gene Runner (www.generunner.com).

Characterization of microsatellite loci

While the microsatellite loci prospected in the plastid genome set are certainly from plastid origin, loci from the total genome set may have different genomic origins (mitochondrial, plastid, nuclear genomic or nuclear EST). Therefore, the genomic location of each microsatellite locus prospected in the total genome set was determined by performing a BLASTn search into the GenBank database of nucleotide sequences. The complete contig in which the microsatellite locus was identified was used as the query sequence for the BLAST. To identify the exact location of the microsatellite loci prospected in the plastid genome set, a search was performed directly within the sequence of the complete plastid genome (GenBank Accession Number KY392760).

The characterization of the microsatellite loci concerning PCR amplification and polymorphism was achieved in samples obtained from plants of C. xanthocarpa sampled in four populations, located in Santa Catarina State, southern Brazil: Coronel Freitas (26°48′33′’S, 52°42′22′’W), Iomerê (27°00′26′’S, 51°14′28′’W), São Joaquim (28°28′38′’S, 50°04′04′’W) and São José do Cerrito (27°43′30′’S, 50°36′21′’W). Total DNA was isolated from fresh leaf samples collected in 25 plants of C. xanthocarpa in each population. The voucher specimens were deposited in the herbaria FLOR (vouchers FLOR 63645 and FLOR 63646 consigned populations Coronel Freitas and Iomerê, respectively), upheld by the Department of Botany of the Federal University of Santa Catarina (Florianópolis, SC, Brazil), and LUSC (vouchers LUSC 8707 and LUSC 7870 consigned populations São Joaquim and São José do Cerrito, respectively), upheld by the University of the State of Santa Catarina (Lages, SC, Brazil). Total DNA was extracted from leaves dried in silica gel, using the CTAB method (Doyle and Doyle 1990).

A total of 89 microsatellite loci (77 from the total genome set and 12 from the plastid genome set) were selected from the initial prospection (see Results section) and tested for polymorphism using DNA samples from eight plants, two of each sampled population. PCR amplification was performed in 20 μL reactions containing 20 ng of DNA, 0.2 mM of each dNTP (Invitrogen, Carlsbad, CA, USA), 1 U of Taq DNA polymerase (QuatroG, Porto Alegre, RS, Brazil), 1X buffer, 1.5 mM of MgCl₂ and 0.2 μM of each primer. Amplifications were performed in a Veriti™ Thermal Cycler (Applied Biosystems, Foster City, CA, USA) with an initial denaturing step at 95 °C for 3 min, followed by 30 cycles of 94 °C for 30 s, annealing temperature (see Table 1) for 30 s and extension at 72 °C for 1 min, with a final extension step of 72 °C for 30 min. PCR products were denatured and resolved through electrophoresis in 8% denaturing polyacrylamide gels and stained with silver nitrate.

Table 1.

Overall characteristics and PCR conditions of six nuclear microsatellite loci (Cxant26–Cxant76) and three organellar microsatellite loci (Cxant22, Guabi05 and Guabi11) developed for Campomanesia xanthocarpa

Locus	GenBank ID	Primer sequences (5′-3′)	Repeat motif	Allele size range (bp)	PCR conditions
					Ta (°C)	Primers [μM]	MgCl2 [mM]	dNTPs [mM]
Neutral nuclear markers
Cxant26A	MG557629	F: ATGCAAAATCCCTACGTGCT R: ATGACACATTTCGGCTGTGA	(ATCG)3	325–345	57	0.05	Master MixB	Master MixB
Cxant50	MG557634	F: CGCACAACCAGCACAAAAC R: CTATCACCGAGGGAGGCAAG	(CTTT)3	475–495	66	0.30	1.50	0.30
Cxant59C	MG557635	F: GAGGGACTTTCAGTTTGTGTGTC R: GACCGTTTCCAACATTTCCA	(GA)10	197–225	55	0.15	Master MixB	Master MixB
Cxant66	MG557636	F: GCGAGACCATAAGCCACTAC R: TGAGAAGGAGACACACACAAAT	(AGA)4	204–208	57	0.12	1.00	0.20
EST nuclear markers
Cxant69	MG557637	F: CCCAACACTCTCCACAATCC R: TCCTTCCCTCTTCTCTCCATC	(GA)7	289–309	60	0.10	1.00	0.20
Cxant76	MG557638	F: ATGTTTTTGTGCGTTCTGG R: TTGACCTTTGTTCCTCTTCCT	(AAG)6	342–355	60	0.12	1.00	0.20
Mitochondrial markers
Cxant22A	MG557627	F: GCTTGGTGGTGCCTCTCTC R: GCTCTTCCCTTTGCCTCTCT	(TCCA)3	209–213	57	0.05	Master MixB	Master MixB
Plastid markers
Guabi05C	MG973216	F: TTCTCGTGATTTGTATCCAAGG R: TGCTTCAATCTTTCCTATCGAA	(T)10	199–205	55	0.10	Master MixB	Master MixB
Guabi11A	MG973217	F: TTGATTCAGGGAACAAATTCAA R: TGGCTAGTGTGGTTCATTCAG	(ATTA)4	222–250	57	0.40	Master MixB	Master MixB

^ALoci amplified as a triplex set; ^BMicrosatellite loci amplified using the Qiagen Multiplex PCR Master Mix; ^CLoci amplified as a biplex set

Out of the 89 evaluated microsatellite loci, nine revealed polymorphism and were employed for further characterization. Four of these microsatellite loci (Cxant50, Cxant66, Cxant69, and Cxant76) were individually amplified using the formerly described PCR conditions. The other five loci were combined in biplex (Cxant59/Gauabi05) or triplex (Cxant22/Cxant26/Gauabi11) sets for PCR amplification. To perform multiplex amplification, it was used a PCR reaction volume of 6.2 μL reaction mix, containing 12 ng of DNA, 2.1 μL of QIAGEN Multiplex PCR Master Mix (Qiagen Inc., Valencia, CA, USA) and 0.10-0.40 μM of each primer (Table 1). Amplifications were carried out in a Veriti™ Thermal Cycler with an initial denaturing step at 95 °C for 15 min, followed by 25 cycles of 94 °C for 30 s, annealing temperature (55 °C or 57 °C; Table 1) for 30 s and extension at 72 °C for 1 min, with a final extension step of 60 °C for 30 min. To all nine loci, the forward primers were labeled with fluorescent dyes (PET, 6-FAM, NED or VIC).

Amplified alleles were resolved through capillary electrophoresis using an ABI 3500 XL DNA automatic sequencer (Applied Biosystems™), with the POP-7™ polymer (Applied Biosystems, Framingham, MA, USA). The reactions were adjusted to a final volume of 11 μL, including 1 μL of PCR product, 0.3 μL of GeneScan 600 Liz genotyping standard (Applied Biosystems, Foster City, CA, USA) and 9.7 μL Hi-Di™ formamide (Applied Biosystems, Woolston, WA, USA). Allele sizing was carried out using the software GeneMapper^® version 4.1 (Applied Biosystems, Foster City, CA, USA).

To test the usefulness of these markers in other species of tribe Myrteae (Myrtaceae), transferability and polymorphism of the nine microsatellite loci were tested in each of five samples of the species A. sellowiana, C. aurea, E. pungens, E. uniflora, P. austral, P. cattleianum, and P. grandiflorum. PCR and genotyping conditions were the same described above for the individually amplified microsatellite loci.

Microsatellite polymorphism analysis

The presence of null alleles and genotyping errors of nuclear microsatellite loci were verified with the program Micro-Checker version 2.2.4 (van Oosterhout et al. 2004). The nonrandom association of alleles at different pairs of loci was evaluated through the analysis of linkage disequilibrium (LD). Although physical linkage and LD are distinctly different, they are related, and high levels of LD may be evidence of physical linkage (Flint-Garcia et al. 2003). A pair of loci was considered physically linked if it presents significant LD in each of the studied populations as well as across all populations. LD was tested with GenePop (Raymond and Rousset 1995, Rousset 2008) on web version 4.6 (http://genepop.curtin.edu.au) using 1000 dememorization steps, 100 batches and 1000 iterations per batch as parameters for the Markov chain. The significance (p < 0.05) of deviation from Hardy–Weinberg equilibrium (HWE) at each locus for each population was tested using FSTAT version 2.9.3.2 (Goudet 2001). The number of alleles (A), expected heterozygosity (He), observed heterozygosity (Ho) and polymorphic information content (PIC) were estimated for the nuclear markers using Cervus version 3.0.3 (Kalinowski et al. 2007).

For the plastid and mitochondrial markers, the number of alleles per locus (A), haploid diversity (h) and Shannon’s index (I) were estimated using the software GenAlEx version 6.5 (Peakall and Smouse 2012).

Results

Sequencing output

A total of 735,621,315 bases were generated as output of the Illumina sequencing. Considering that the haploid genome size of C. xanthocarpa estimated through flow cytometry is 205.38 Mbp (1Cx = 0.21 pg; our unpublished data), this output corresponds to a coverage of 3.6 × , on average, of the whole genome of C. xanthocarpa.

The de novo assembling of the reads (mean size of 203.43 bases) generated a total of 5551 contigs (average length of 1538 bases; N25 = 1888; N50 = 1453; N75 = 1239), totaling 8,539,780 bases (Table 2). Including scaffolded regions, the shortest contig had 981 bases, while the longest had 38,536 bases in length (Table 2).

Table 2.

Data summary of the genome sequencing and assembling, and the repetitive elements characterized in the C. xanthocarpa total genome set of sequences

Sequencing and assembling
Number of sequenced bases	735,621,315
Estimated coverage	3.6×
Average quality per read (Phred score)	32
Number of contigs	5551
N75 contig length (bases)	1239
N50 contig length (bases)	1453
N25 contig length (bases)	1888
Longest contig (bases)	38,536
Average contig size (bases)	1538
Total size of contigs (bases)	8,539,386
GC content	42%

Repetitive elements characterization	Number of elements	Total length (bases)	% of the WGSAS
			C. xanthocarpa	E. unifloraa
Retroelements	373	242,262	0.028	0.010
LINE elements
L1-CIN4	45	13,848	0.002	7.43 × 10−5
LTR retrotransposons
Ty1-Copia	262	193,438	0.023	0.007
Gypsy-DIRS1	52	37,089	0.004	0.002
DNA transposons
hobo-Activator	8	893	1.00 × 10−4	2.17 × 10−5
Tourist/Harbinger	3	229	2.68 × 10−5	–
Unclassified elements	1	68	7.96 × 10−6	1.49 × 10−5
Small RNA	20	7939	0.001	8.45 × 10−5
Satellites	1	199	2.33 × 10−5	–
Simple repeats	1730	64,379	0.008	0.002
Low complexity	467	22,515	0.003	4.00 × 10−4
Total interspersed repeats	–	256,841	0.030	0.010

The comparison of percentage of each element in relation to the total WGSAS is giving in comparison to Eugenia uniflora

^aData from Stefenon et al. (2019)

Genomic repetitive elements

A total of 351,785 bases were characterized as repetitive elements, representing 4.12% of the C. xanthocarpa WGSAS. Retroelements were the most represented ones, with a total of 373 elements and 242,262 bases in length (Table 2). Individually, LTR retrotransposons of the Ty1 family were the most largely represented, with 262 elements, totalizing 193,438 bases (2.26% of the total) (Table 2). The second most important element corresponds to the simple repeats, with 1730 elements, reaching 64,379 bases in length, followed by Gypsy retrotransposons (52 elements and 37,089 bases) and low complexity sequences (467 elements and 22,515 bases in length). Retroelements of the L1-CIN4 family, DNA transposons (hobo-Activator and Tourist/Harbiger families), small RNAs, one satellite and one unclassified element were also identified (Table 2).

In comparison to the data reported for a draft genome of E. uniflora (Stefenon et al. 2019), C. xanthocarpa has a threefold larger amount of all classes of repetitive genomic elements including Tourist/Harbiger and satellite elements, that were not identified in E. uniflora (Table 2). A higher proportion of Copia in relation to Gypsy retrotransposons was revealed by both species.

Nuclear and organellar microsatellite loci

Based on the determined prospection parameters, a total of 364 microsatellite loci were identified in the WGSAS set and 84 loci in the plastid genome set. Out of the 364 loci of the WGSAS set, primer pairs were designed and used to amplified 77 of them. Six of them failed amplifying, 64 were monomorphic, and seven loci revealed polymorphism. Among these polymorphic microsatellites, six were located in the nuclear and one in the mitochondrial genomes. Out of the 84 microsatellite loci prospected in the plastid genome set, 12 were tested and two were polymorphic. Sequences of all characterized microsatellite loci were deposited in the GenBank (Table 1).

Polymorphism of the nuclear loci

The BLAST characterization of the nuclear loci revealed four neutral (i.e., not matching any known genomic region in the BLAST analysis), while two matched sequences of Eucalyptus grandis deposited in the GenBank database. The contig containing locus Cxant69 (1244 bases in length) revealed 91% of similarity to the alcohol dehydrogenase class-3 gene, containing 3 exons and two introns of this gene (Supplementary File 1). The contig containing locus Cxant76 (1112 bases in length) revealed 86% similarity to the 541 bases of the end-region of the disease-resistance protein RGA2 of Eucalyptus grandis (Supplementary File 1).

Micro-Checker identified significant evidence of null alleles in loci Cxant50 and Cxant76. Significant linkage disequilibrium (LD), as evidence of putative physical linkage between pair of loci, was verified in four pairs of loci at the population level and in three pairs of loci across all populations, but no pair of loci showed LD in all populations (Table 3). A total of 24 alleles were identified with the nuclear markers (mean ranging from 2.25 for loci Cxant26, Cxant59 and Cxant66 to 4.25 for locus Cxant76; Table 4). Significant deviation from HWE (p < 0.05) was detected for loci Cxant50 and Cxant76 in two and three populations, respectively, while loci Cxant26, Cxant59, Cxant69, and Cxant66 did not deviate from HWE in any population (Table 4).

Table 3.

Estimates of linkage disequilibrium between pair of loci

Pairs of loci	Population	p value
Cxant59–Cxant66	Coronel Freitas	0.101
Cxant59–Cxant69	Coronel Freitas	0.007
Cxant66–Cxant76	São Joaquim	0.059
Cxant69–Cxant76	São Joaquim	0.003
Cxant66–Cxant76	São José do Cerrito	0.028
Cxant59–Cxant69	Across all Populations	0.052
Cxant66–Cxant76	Across all Populations	0.014
Cxant69–Cxant76	Across all Populations	0.053

Table 4.

Genetic characterization of the ten newly developed polymorphic microsatellite loci of Campomanesia xanthocarpa at population level and overall populations (mean value)

Nuclear	Coronel Freitas			Iomerê			São Joaquim			São José do Cerrito			Overall mean
	A	Ho	HAe	A	Ho	HAe	A	Ho	HAe	A	Ho	HAe	PIC	A	Ho	He
Cxant26	3	0.375	0.377ns	2	0.500	0.383ns	2	0.280	0.246ns	2	0.400	0.327ns	0.281	2.25	0.39	0.33
Cxant50+	4	0.040	0.257*	2	0.286	0.418ns	4	0.333	0.669*	4	0.524	0.502ns	0.458	3.50	0.30	0.46
Cxant59	2	0.520	0.510ns	2	0.400	0.444ns	2	0.200	0.184ns	3	0.520	0.554ns	0.395	2.25	0.41	0.42
Cxant66	2	0.417	0.337ns	2	0.240	0.372ns	3	0.560	0.598ns	2	0.440	0.458ns	0.399	2.25	0.41	0.44
Cxant69$	3	0.522	0.679ns	3	0.739	0.571ns	2	0.840	0.497ns	3	0.875	0.520ns	0.504	2.75	0.74	0.57
Cxant76+$	7	0.261	0.744*	5	0.240	0.612*	2	0.150	0.296ns	3	0.391	0.505*	0.569	4.25	0.26	0.54
Mean	3.5	0.356	0.484	2.7	0.401	0.467	2.5	0.394	0.415	2.8	0.525	0.478	–	–	–	–

Organellar	A	h	I	A	h	I	A	h	I	A	h	I		A	h	I
Guabi05	3	0.627	0.998	2	0.080	0.168	4	0.577	0.994	3	0.293	0.551	–	3.00	0.39	0.68
Guabi11	2	0.518	0.690	1	0.000	0.000	1	0.000	0.000	2	0.280	0.440	–	1.50	0.20	0.28
Cxant22	2	0.518	0.690	1	0.000	0.000	1	0.000	0.000	1	0.000	0.000	–	1.25	0.13	0.17
Mean	2.3	0.554	0.792	1.3	0.027	0.056	2	0.192	0.331	1.92	0.241	0.377	–	–	–	–

A = number of alleles; Ho = observed heterozygosity (for nuclear loci only); He = expected heterozygosity (for nuclear loci only); PIC = polymorphism information content; h = haploid diversity (for organellar loci only); I = Shannon’s index (for organellar loci only)

^ASignificance of deviation from Hardy–Weinberg equilibrium: *P < 0.05; ns = not statistically significant

⁺Significant evidence of null alleles detected (P < 0.05)

^$EST nuclear loci

At population level (Table 4), the observed heterozygosity of nuclear loci revealed from low (Ho = 0.040; locus Cxant50 in population Coronel Freitas) to high diversity (Ho = 0.875; locus Cxant69 in population São Joaquim). The expected heterozygosity ranged from He = 0.184 (locus Cxant59 in population São Joaquim) to He = 0.744 (locus Cxant76 in population Coronel Freitas). PIC values ranged from 0.281 (locus Cxant26) to 0.569 (locus Cxant76) (Table 4).

Polymorphism of the organellar loci

The BLAST analysis revealed the presence of one polymorphic mitochondrial locus within the total genome set. Locus Cxant22 (contig containing 32,215 bases in length) showed 99% of similarity with a region of 16,583 bases of the mitochondrial genome of E. grandis deposited in the GenBank (Supplementary File 1). This microsatellite locus is located between tRNA-Asn and tRNA-Cys genes.

The two polymorphic plastid loci prospected are located between the trnS-GCU and the trnG-UCC regions (locus Guab05) and between the trnY-GUA and the trnT-GGU regions, including the complete trnE-UUC gene (locus Guab11).

Organellar loci are evaluated as haploid markers, since both mitochondrion and plastid present a circular haploid genome. Organellar loci revealed a total of eight alleles (mean ranging from 1.25 for locus Cxant22 to 3 for locus Guabi5; Table 4).

The organellar locus with lowest genetic diversity was Cxant22 (mean h = 0.17 and mean I = 0.13, Table 2), while locus Guabi05 revealed the highest mean values (h = 0.39 and I = 0.68) as well as the highest diversity estimates at the population level (Table 4).

Transferability of the microsatellite loci

The cross amplification of the microsatellite loci to seven Myrteae species (Table 5) revealed a percentage of transferability of 77.8% in A. sellowiana (7/9), 77.8% in C. aurea (7/9), 44.4% in E. pungens (4/9), 66.7% in E. uniflora (6/9), 88.9% in P. australe (8/9), 88.9% in P. cattleianum (8/9) and 90% in P. grandiflorum (8/9). Loci Cxant59, Cxant22, Guabi05, and Guabi11 successfully amplified products in the seven analyzed species. Locus Cxant76 did not amplify in any of the tested DNA species.

Table 5.

Allele size range of ten microsatellite loci developed for Campomanesia xanthocarpa and tested for cross amplification in five samples of seven species of the tribe Myrteae, family Myrtaceae

Locus	Acca sellowiana	Campomanesia aurea	Eugenia pungens	Eugenia uniflora	Psidium australe	Psidium cattleianum	Psidium grandiflorum
Cxant26	–	–	–	–	325	325	325
Cxant50	496	477	–	470–471	461–489	459–474	461
Cxant66	202	204	–	202	205–209	205–209	205
Cxant69	282	294–302	–	–	294	298	292–294
Cxant76	–	–	–	–	–	–	–
Cxant59	198	199	197	198–199	199	199	199
Cxant22	213	209	209	209	213	213	213
Guabi05	199–200	199–200	198	199–200	200	199–200	200
Guabi11	245	219	228	245	231	231	231
Total of loci	7	7	4	6	8	8	8

– Unsuccessful amplification

Discussion

The present study reports the characterization of repetitive elements within the WGSAS of C. xanthocarpa, with focus in the prospection and characterization of species-specific nuclear and organellar microsatellite markers, using the next-generation sequencing (NGS) method. The NGS technology has increasingly been used for the partial genome sequencing of different species aiming the overall characterization of the genomes (Stefenon et al. 2019) and prospection and characterization of microsatellite markers in tree species (e.g., Owusu et al. 2013, Staton et al. 2015, Wu et al. 2017, Lemos et al. 2018).

Lineage-specific accumulation/exclusion of transposons is a critical element in genome evolution (Macas et al. 2015), driving the genome size variation in flowering plants. Among these repetitive genomic elements, the two major superfamilies of LTR retrotransposons, Gypsy and Copia, are recognized to be differentially distributed across many species. Frequently, Gypsy superfamily is more highly amplified than Copia in plant species (Barghini et al. 2014). An opposite result was obtained with WGSAS of C. xanthocarpa (present work), as well as for E. uniflora (Stefenon et al. 2019). Stefenon et al. (2019) suggested that such trend could be due to a particular dynamic of repetitive elements distribution in E. uniflora. Likewise, an analysis of transcriptionally active LTR retrotransposons in Eucalyptus genus identified a higher number of Copia elements than Gypsy ones (Marcon et al. 2015). A higher frequency of Copia retrotransposons was also reported in other tree species: Nyssa sylvatica, Liquidambar styraciflua, Quercus alba, Persea borbonia, Juglans nigra, Fraxinus americana, F. pennsylvanica and Acer saccharum (Staton et al. 2015). Take into account the similar results in different tree species, we can raise the hypothesis, to be further tested, that a higher number of Copia elements than Gypsy ones is characteristic to wood plant species.

The present WGSAS of C. xanthocarpa is composed by about 4.12% of repetitive elements, a proportion lower than observed in E. uniflora (7.13%; Stefenon et al. 2019). As expected, DNA transposons has a lower frequency than retroelements, much likely due to the different mode of transposition of these elements. While DNA transposons just move across the genome (the cut-and-paste mode of transposition), retroelements copy itself for transposition (copy-and-paste mode of transposition).

Repetitive sequences are conjectured to effect cytoplasmic, cellular and developmental processes not only by changing genome size, but also affecting chromosomal recombination. Simple sequence repeats represent recombination “hotspots” of genome reorganization (Mehrotra and Goyal 2014) and has been largely used as genetic markers of non-model species (e.g., Staton et al. 2015, Lemos et al. 2018). Simple sequences regions (also known as microsatellite regions) were the second most common class of repetitive elements observed in the C. xanthocarpa WGSAS, spanning a total of 64,379 bases, which made up about 0.008% of the WGSAS, distributed into 1730 loci. In Pisum sativum, a legume species with an estimated haploid genome of 4300 Mb, microsatellite sequences corresponded to about 1.5% of the genome (Macas et al. 2007). Using low-coverage sequencing of the whole genome strategy for characterization of microsatellite markers in 10 hard-wood species, Staton et al. (2015) identified from 891 to 18,167 potentially amplifiable loci/species, without correlation between number of identified loci and estimated genome size.

In comparison to other studies reporting the development and characterization of microsatellite markers for species of tribe Myrteae, our markers revealed lower polymorphism.

Ten putative nuclear microsatellite markers were developed for C. xanthocarpa by Góes et al. (2019), and revealed a mean expected heterozygosity ranging from He = 0.303 to He = 0.796 (mean He = 0.579), observed heterozygosity ranging from Ho = 0.00 to Ho = 0.958 (mean Ho = 0.275), and number of alleles ranging from A = 2 to A = 8 (mean A = 3.95). Using NGS technology (IonTorrent PGM™ platform), 12 microsatellite markers were characterized for Eugenia uniflora (Sarzi et al. 2019) and moderate to high level of diversity was reported for the number of alleles (mean A = 6.1, ranging from A = 3 to A = 12), expected heterozygosity (mean He = 0.70, ranging from 0.57 to 0.91) and observed heterozygosity (mean Ho = 0.23, ranging from 0.00 to 0.57). Twenty-three microsatellite markers developed for Acca sellowiana using SSR-enriched library strategy (Santos et al. 2008, Klabunde et al. 2014) revealed higher levels of diversity ranging from A = 2 to A = 15 (mean A = 8.0), from He = 0.458 to He = 0.949 (mean He = 0.768) and from Ho = 0.200 to Ho = 1.000 (mean Ho = 0.672). The lower polymorphism and diversity of the Illumina-sequencing-derived microsatellite loci in comparison to SSR-enriched library-based markers was also observed in Accer saccharum (Khodwekar et al. 2015). Such lower polymorphism of nuclear microsatellite markers generated using this methodology has been attributed to a biased-sequencing towards short and AT-rich regions occurring in Illumina-sequencing runs (Khodwekar et al. 2015). Our total genome set of sequences (8.53 Mbp) was formed mainly by reads of 300-304 bases in length and revealed an AT content of about 60%. Out of the 364 microsatellite loci prospected, 154 presented the A/T repeat motif, 15 the AT/TA motif and five the AAT/TTA motif, reaching 47.8% of the total prospected loci (data not shown). On the other hand, only 9.09% of the microsatellite loci characterized (seven out of 77) contained repeat motifs AT/TA or AAT/TTA and no loci with the motif A/T was evaluated. Therefore, this relatively low polymorphism may be characteristic of the species and the microsatellite markers described in this study seems sufficiently variable for genetic studies of C. xanthocarpa populations.

The characterization of both, nuclear and organellar markers allows elaborating studies concerning biparental vs uniparental gene flow using the estimations of genetic differentiation (F_ST) obtained individually from each marker category (Petit et al. 2005a, b; Lemos et al. 2018). It is known that the mode of inheritance has a key effect on subdividing genetic diversity. Studies based on maternally inherited markers have significantly higher estimations than those based on biparentally inherited markers (Petit et al. 2005a, b), corroborating the much higher estimation of F_ST (AMOVA) for organellar markers (fourfold higher than nuclear markers) observed for C. xanthocarpa in this study (data not shown).

Another important insight from our study is the characterization of nuclear-EST markers. Locus Cxant69, linked to a disease-resistance protein and locus Cxant76, linked to the ADH-3 gene (alcohol dehydrogenases are related to plant growth, development, adaptation, fruit ripening, and aroma production), may be quite useful in programs of selection and breeding of C. xanthocarpa.

The absence of linkage between pairs of nuclear microsatellite loci is an important feature of the developed markers, meaning a widest and independent coverage of the C. xanthocarpa genome when using these loci in genetic studies. Since none of the pairs of loci showed significant LD in all populations, we can consider that they are not physically linked. The significant LD values may be attributed to high levels of inbreeding (Flint-Garcia et al. 2003), as revealed for locus Cxant76, which revealed significant deviation from HWE (Table 3) and evidence of the occurrence of null alleles.

The reported deviation from HWE is not surprising in natural populations of tree species, mainly due to the occurrence of inbreeding and limitation of inter-populational gene flow. However, this pattern does not occur for all loci, supporting the evidence for the presence of null alleles detected in our analysis. This is an important aspect that may be considered in studies of natural populations of C. xanthocarpa using the developed markers.

As expected, the organellar microsatellite markers were successfully transferred to all species of the tribe Myrteae tested in the cross-amplification analysis. This fact can be due to the highly conserved organellar genomes across species of this tribe (Machado 2017). On the other hand, one out of four nuclear markers (Cxant59) successfully amplified DNA from all seven tested species, while the other eight markers amplified at least DNA from three species. EST–microsatellite markers would be expected to show a high degree of transferability, since they are found inside conserved expressed regions. Intriguingly, the EST locus Cxant76 did not transfer to any species. Further tests trying to optimize the PCR conditions for the amplification may improve the transferability of this locus to other species.

Here, we report the first study related to the characterization of the repetitive elements in C. xanthocarpa, as well as a set of microsatellite markers specifically developed for this species. The microsatellite markers tested in the present study seem to be, on average, slightly less variable than markers reported for other Myrtaceae species from Brazil. Anyway, this set of markers revealed enough diversity for future populational genetic studies. Since C. xanthocarpa is largely distributed in the Atlantic Forest biome, which is a highly fragmented environment, genetic characterization studies in populations of this species are needed. This set of species-specific markers will largely improve the current knowledge about C. xanthocarpa populations, since microsatellite markers are useful for assessing genetic diversity, inter- and intra-populational genetic structure and gene flow, as well as the fitness of natural populations (e.g., Nagel et al. 2015, Stefenon et al. 2016). Improving genetic knowledge will help to better design conservation and sustainable use strategies, which is of utmost importance for the future of this promising fruit tree from the Atlantic Forest.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Ethical approval

This article does not include any studies with human participants or animals performed by any of the authors.

Informed consent

This article does not involve any informed consent.

Conflict of interest

All authors hereby declare that there is no conflict of interest.