Size, gene content and gene order of the olive chloroplast genome

The complete plastome of olive, cv. Frantoio has a total length of 155,889 bp (GenBank Accession Number GU931818), with the typical structure found in the unrearranged chloroplast genomes of Angiosperms. It includes an 86,614-bp Large Single Copy (LSC) and a 17,791-bp Small Single Copy (SSC) region separated by a pair of Inverted Repeats (IR), each 25,742 bp long (Figure 1). Coding DNA (92,095 bp) accounts for 59.08% of the genome and includes protein coding genes (80,252 bp), tRNAs (2,793) and rRNAs (9,050), while noncoding DNA (63,794 bp) accounts for the remaining 40.92% and includes introns (20,130 bp) and intergenic spacers (43,664 bp). The olive plastome contains 114 unique genes (80 CDS, 30 tRNA and 4 rRNA), with 19 of these genes (8 CDS, 7 tRNA and all 4 rRNA) duplicated in the IR for a total of 133 genes. In addition, the duplicated region includes a partial CDS for ycf1, as in other species like Typha [39]. There are 18 intron-containing genes, 15 of which contain one intron and 3 (ycf3, clpP and rps12) with two introns. The rps12 gene is trans-spliced, with the 5' end located in the LSC and the 3' end duplicated in the IR regions. The nucleotide composition of the olive chloroplast genome comprises 37.81% GC and 62.19% AT.

The repeat analysis also identified 14 interspersed repetitive sequences longer than 30 bp, each having 2-6 repetitions and a sequence identity higher than 85% (Table 2, Figure 2). These long interspersed repetitive sequences included two tandem repeats in the ycf2 gene and five palindromic sequences (two in the LSC, one in the SSC and two in the IR regions). Three of the four repeats found within the ycf2 exon were tandem repeats, as previously observed in V. vinifera [31]. There were only two inverted repeats, all the others were direct repeats. Five repeats were located within CDS, two repeats were found in the introns of the ycf3 and ndhA genes and all others were in the intergenic spacers (Table 2). Interspersed repeats did not cause any uncertainty during the sequencing process because they were quite short (< 61 bp), with a low number of repetitions and primers were never constructed on the repeats.

Repeat	Number of repeats	Size	Start(1)	Type	% Identity	Region	Gene position	Sequence
1	3	30	9,345	D	86.67	LSC	trnS-GCU-exon	[CA][AC]GGA[GA]AGAGAGGGATTCGAACCCTCG[AG]TA
			37,281	D		LSC	trnS-UGA-exon
			47,117	I		LSC	trnS-GGA-exon
2	2	31	10,849	D	90.32	LSC	trnG-UCC-exon 2	[AT][AG]A[CA]GATGCGGGTTCGATTCCCGCTA[CT]CCGC
			38,241	D		LSC	trnG-GCC-exon
3	1	30	14,401	P	93.33	LSC	atpF - atpH	AAATATGAAAAATA[TC][GA]TATTTTTCATATTT
4	2	45	40,451	D	88.89	LSC	psaB-exon	[AT]TGCAATAGCTA[AG]ATGATG[AG]TG[TA]GCAATATCGGTCAGCCATA[AG]AC
			42,675	D		LSC	psaA-exon
5	3	41	45,474	D	92.68	LSC	ycf3-intron	T[CA]CAGAACCGTAC[GA]TGAGATTTTCA[TC]CTCATACGGCTCCTC
			100,797	D		IR	rps12 - trnV-GAC
			122,052	D		SSC	ndhA-intron
6	2	31	56,736	D	90.32	LSC	atpB - rbcL	T[AT]CTTATTCATCCACTTGAAATTTTCAA[AG][AT]T
			56,777	I		LSC	atpB - rbcL
7	1	44	76,926	P	95.45	LSC	psbT - psbN	TTGAAGTAATGAGCCTCCC[CA]ATAT[TG]GGGAGGCTCATTACTTCAA
8	2	30	83,181	D	90	LSC	rps8 - rpl14	AATCTA[CG]T[AT][AC]TTAATCTAGTTCTTAATCTA
			83,193	D		LSC	rps8 - rpl14
9	2	30	91,385	TR	90	IR	ycf2-exon	TTTCTTTTTGTC[CT]AA[GC]TCACTTC[TC]TTTTTT
			91,427			IR	ycf2-exon
10	2	36	93,791	TR	94.83	IR	ycf2-exon	[AG]ATATTGATG[AC]TAGTGAC[AG]ATATTGATG[AC]TAGTGAC
			93,827			IR	ycf2-exon
11	1	48	96,252	P	91.67	IR	ycf15 - trnL-CAA	AGAGCTCGGATCGAATCGGTAT[TA][TG][AC][TA]ATACCGATTCGATCCGAGCTCT
12	2	30	109,623	D	93.33	IR	rrn 4.5 - rrn 5	CATTGTTCAA[AC]TCTTTGACAACA[CT]GAAAAA
			109,654	D		IR	rrn 4.5 - rrn 5
13	1	61	110,599	P	95.08	IR	trnR-ACG - trnN-GCU	AGAATTCTCAGATGTACTAGCACTGCATC[AT][AT][AT]GATGCAGTGCTAGTACATCTGAGAATTCT
14	1	36	118944	P	100	SSC	ndhD - psaC	AAAACCCGTGCTCCAAATATTTGGAGCACGGGTTTT

D: direct, I: inverted, P: palindrome, TR: tandem repeat (imperfect).

⁽¹⁾ The start base position of the interspersed repeats and palindromic sequences refers to the cv. Frantoio sequence.

Bold nucleotides refer to the indel P32.

The actual size of the olive plastome is larger than the size estimated on the basis of RFLP analysis, which predicted a range from 132 to 134 Kb [16].

Olive chloroplast genome organisation

The sequence of the olive chloroplast genome represents one of the first contributions to deciphering the genetic background of this important tree crop species and was used to verify that rearrangements observed in the plastomes of other genera of Oleaceae, such as Jasminum and Menodora, were not represented in O. europaea. In fact, in contrast to what observed in the Jasminum and Menodora plastomes [2], the olive chloroplast maintains a size range, organisation and gene order typical of most land plants, such as members of the Vitis, Populus, Citrus, Eucalyptus, Coffea and Arabidopsis genera. Based on the phylogeny of Oleaceae inferred from the ndhF and rbcL genes [2], Jasminum and Menodora were already known to be unusual genera within the family, and all other tribes, including Oleae, to which the Olea genus belongs, do not share their combination of multiple mutational events. The highly conserved plastome organization of the olive allowed universal primers and genome walking with consensus primers to be used to amplify most of the LSC region.

Identification of new plastid markers to discriminate between olive cultivars

To detect intervarietal polymorphisms, a preliminary screening of the intergenic spacer trnS-GCU - trnG-UCC, previously demonstrated to be polymorphic among olive varieties [40], was performed on a set of 30 cultivars having different geographical distributions and representing a wide range of morphological and agronomical phenotypes (data not shown). A sub-set of eight highly variable cultivars (Table 4) was further examined for 100 potentially polymorphic regions.

The comparison of the Frantoio chloroplast sequence with ESTs deriving from fruits of cvs. Coratina and Tendellone showed some sequence mismatches, but they were not confirmed by resequencing the corresponding genomic regions in Coratina and Tendellone cultivars.

Overall, the analysis of cpDNA sequences from the eight cultivars resulted in the identification of 40 polymorphic sites, 30 of which represent new and never-described plastid variants (Table 3, Table 4, Figure 2). Sixteen polymorphic sites were mono-nucleotide SSRs: eight poly-A, including one with an irregular motif; seven poly-T and one poly-C. The remaining polymorphisms included 20 SNPs and 4 indels. Thirty-three polymorphic sites (P1-P33) were located within the LSC region, one (P34) within the IR and six (P35-P40) within the SSC (Figure 2). The indel P32 was identified within the repeat of the rps8 - rpl14 spacer, but none of the other repetitive regions was polymorphic between cultivars.

The chloroplast sequence of cv. Frantoio was also compared with all previously sequenced regions of the olive chloroplast, particularly with the plastome sequence of cv. Bianchera, which has been recently deposited in the Genbank database (NC_013707.1). More than 200 mismatches were detected between the Bianchera and Frantoio sequences. Surprisingly, not one of these polymorphisms fell within the previously identified cultivar-specific polymorphic regions. To verify if these mismatches might represent real sequence differences between the two varieties, most of the ambiguous regions were reamplified and resequenced in both cultivars (Bianchera sample was provided by the CRA-OLI of Spoleto, Perugia, Italy). These analyses confirmed the sequence of cv. Frantoio and showed an absolute sequence identity with that obtained from the cv. Bianchera in all of these regions, including the exons of the rpoC1 and ndhF genes, carrying 27- and 9-bp indels, respectively. The differences detected between the two olive plastome sequences can not derive from an incorrect identification of the Bianchera genotype because, in that case, mutations should have been found in the polymorphic sites and not randomly along the chloroplast genome. More likely, divergences may be attributed to sequence uncertainties in the Bianchera plastome sequence deposited in GenBank.

The new markers identified in this study can distinguish six haplotypes among eight cultivars. Therefore, these new markers hold great promise for the identification of new cultivar haplotypes and for use in DNA barcoding systems to distinguish between different cultivars.

Comparison of plastome variation between cultivars and with other Oleataxa

Based on previous chloroplast sequence analyses, olive cultivars belong to the cp-II lineage and have been classified into three sublineages (E1, E2 and E3) and four chlorotypes (1, 2, 9 and 13) [19, 40]. These chlorotypes were defined by evaluating length variations in the psbK-psbI, trnS-trnG, rps2-rpoC2, trnE-trnT and atpB-rbcL regions among more than 140 cultivars [17, 19, 40].

Several polymorphisms had been previously identified in the partial sequence of the trnK intron (AF359497-AF359504) by analysing the subspecies cuspidata, laperrinei, maroccana, cerasiformis, guanchica, europaea var. sylvestris (wild olive) and the Cornicabra cultivar, but none of these polymorphisms were found among the cultivars we have analysed. The psbK-psbI and trnS-GCU-trnG-UCC regions, spanning the polymorphic sites P8, P9, P10, P11, P12, P13 and P14, were analyzed by Besnard et al. [12] as fragment length variation on a set of different O. europaea taxa including cultivars. That analysis revealed intercultivar variability only at P11, P12 and P13 but was unable to keep the C/T and G/T SNPs in the P9 and P14 sites, respectively. We treated the A/T/- polymorphism, closely linked to P11, as a different polymorphism (P10) because the A/- indel is present in most varieties while the T is a rare mutation carried by few cultivars.

The spacer rps2-rpoC2, spanning the polymorphic sites P16, P17 and P18, generated five different chlorotypes among the eight varieties analysed, demonstrating a high level of rearrangement within cultivars. This region corresponds to the ccmp5 microsatellite [42, 43], but previous studies that analysed only length polymorphisms were unable to capture the complexity of this region. P28 includes ccmp7 [40, 42] and an additional SNP polymorphism (P27) captured in the flanking region.

Intrieri et al. [18] reported the identification of 5 SNPs and 4 indels in the trnD-trnT region of 13 cultivars. Analyzing a different set of cultivars, Besnard [19] did not detect these polymorphisms. Similarly, only two polymorphisms were confirmed in our cultivar set: the poly-A SSR (P20) and the C/T SNP (P21).

Other regions previously analysed in different Olea taxa, such as trnL-F and rps16 [4], trnL-trnF [13], and trnT-trnL [11] were not polymorphic among our cultivars.

No differences between the eight cultivars were found within the matK and psbA exons or the rps16 intron, regions used for species barcoding. In contrast, the psbK-psbI and trnH-psbA barcoding regions, both representing markers for plant species identification [44, 45], correspond to our P8, P2, P3 and P4 polymorphisms. This observation indicates that these markers may not accurately discriminate between some species, given their potential intra-specific genetic variations [46].

Plant material and DNA extraction

For the plastome sequence analysis, leaves of cv. Frantoio were collected from the accession present at the CRA-OLI olive cultivars collection (Collececco, Spoleto).

For the detection of intervarietal polymorphisms, a subset of eight cultivars was used, chosen among 30 cultivars pre-selected on the basis of their haplotypes for the intergenic trnS-GCU - trnG-UCC spacer (Table 4).

Total DNA was extracted by the DNeasy Plant Mini Kit (Qiagen, Hilden, Germany), following the manufacturer's instructions.

Sequencing strategy: primer design and PCR amplification

Sequencing of the olive plastome was performed by designing a series of PCR primer pairs that produced partially overlapping amplicons and spanned the entire chloroplast genome.

For the Large Single Copy (LSC) region, 38 primer pairs located within conserved regions and designed by Grivet et al. [34] were used, avoiding gaps between successive fragments along the cpDNA molecule. Five primer pairs (5-14-22-27-38) produced double bands, and two (16-28) did not produce any amplification. Thus, new primers for those regions were constructed, following the strategy used for the amplification of the IR and SSC regions. For primer sequences see Additional File 1, Table S2.

For the SSC and the IR regions, primers were constructed from conserved sequences derived by the alignment of the plant chloroplast genomes of Jasminum nudiflorum (DQ673255), Populus tricocharpa (EF489041), Vitis vinifera (DQ424856), Eucaliptus globulus (AY780259), Arabidopsis thaliana (AP000423), Gossypium hirsutum (DQ345959), Citrus sinensis (DQ864733), Cucumis sativus (AJ970307), Morus indica (DQ226511), Panax ginseng (AY582139), Solanum lycopersicum (AM087200) and Nicotiana tabacum (Z00044). These sequences were retrieved from GenBank and aligned using Muscle V. 3.7 [47], and the primers were designed using PerlPrimer v1.1.6 [48]. Because the average size of the amplified fragments was approximately 2,500 bp, internal primers to sequence the entire amplicons were also designed. The primer sequences and positions, along with their respective amplicon lengths, are given in Additional File 1, Table S1.

PCR amplifications were performed in a final volume of 50 μL containing 1-20 ng of template DNA, 10× PCR buffer, 200 μM of each dNTP, 10 pmol of each primer and 2 U of EuroTaq polymerase (EuroClone). For those fragments that were longer than 5,000 bp, 1 unit of LA Taq polymerase (TaKaRa) was used instead. The amplifications were performed with the PCR System 9600 (Applied Biosystems, Foster City, CA) using the following cycling conditions: an initial denaturation step of 95°C for 5 min, followed by 35 cycles of 95°C for 30 sec, 60°C for 30 sec and 72°C for 25 sec and a final elongation step of 72°C for 30 min. For those amplifications including LA Taq polymerase in the PCR mix, the following cycling conditions were used instead: an initial denaturation step of 94°C for 1 min, followed by 30 cycles of 98°C for 60 s and 68°C for 10 min and a final extension step of 72°C for 10 min. Negative controls (no template DNA) were included in all experiments.

The PCR products were checked by electrophoresis on 2% agarose gels, then purified with the JetQuick PCR purification kit (Genomed) and directly sequenced in both directions using the ABI Prism BigDye Terminator V.3.1 Ready Reaction Cycle Sequencing Kit (Applied Biosystems) on an ABI 3130 Genetic Analyzer (Applied Biosystems-Hitachi). The sequences were assembled using BioEdit v7.0.9 software (Ibis Biosciences, Carlsbad, CA).

The DOGMA program [49] was used for the initial genome annotation, which was then manually refined using Artemis version 11 [50] and NCBI Blast searches. The annotation of tRNA genes was checked using tRNAscan version 1.21 [51]. The genome map was generated using OGDRAW software V. 1.0 [52].

Evaluation of repeat structures

Msatfinder v. 2.0.9 [53] was used to identify simple sequence repeats (SSR), with the following settings: a six-repeat threshold for mono-nucleotide SSRs, a five-repeat threshold for di- and tri-nucleotide SSRs and a three-repeat threshold for tetra-, penta- and esa-nucleotide SSRs. The SSR density in the different regions of the chloroplast genome was calculated by dividing the number of SSRs by the length of the given region. Interspersed repeats were identified with REPuter [54] by setting the minimum repeat size to 30 bp and the Hamming distance to 3. The presence and distribution of the repetitive element were verified manually using Artemis and computationally by performing an intragenomic Blast search. For this purpose, the sequence was interrogated using a local installation of NCBI Blast and a Blast database created with formatDB software http://www.ncbi.nlm.nih.gov/staff/tao/URLAPI/wwwblast/.

Identification of polymorphic regions among olive cultivars

To identify sequence polymorphisms, the following potentially variant domains were tested: i) regions containing mono-, di-, tetra- and penta-nucleotide microsatellites; ii) regions previously reported as polymorphic among Olea subspecies, iii) regions containing high sequence variations among 12 species (see materials and methods for chloroplast sequencing strategy); iv) barcoding regions previously identified for species discrimination that had never been tested in olive cultivars; and v) plastid ESTs derived from massive sequence analyses of fruit cDNAs [55].

Candidate SSRs were selected among those having the highest number of repeats (Table 1 and Figure 2). Although no mono-nucleotide SSRs with repeats shorter than 10 bp were considered, some were indirectly included in the analyses of other regions.

PCR amplifications were performed in a final volume of 25 μl containing 25 ng of template DNA, 2,5 μl of 10 × PCR buffer, 0.5 mM of each dNTP, 1 μM of each primer and 1.5 U/μl of PerfectTaq DNA Polymerase (5-PRIME). The amplifications were run on a thermal cycler Mastercycler Gradient (Eppendorf) using the same conditions as previously indicated for plastid sequencing.

After an initial evaluation by electrophoresis on a 2% agarose gel, amplicons were sequenced in both directions using the ABI Prism BigDye Terminator V.3.1 Ready Reaction Cycle Sequencing Kit and run on an ABI 3130 Genetic Analyzer (Applied Biosystems-Hitachi).

The sequences of each region were aligned to evaluate the presence of SNPs, indels or polymorphic microsatellites among the six cultivars. To use these polymorphisms as chloroplast markers able to distinguish olive cultivars from each other, specific primers localizing within conserved flanking regions were constructed (Additional File 1, Table S1). The resulting fragments ranged in size from 145 to 688 bp and could be amplified at an annealing temperature of 60°C. Some amplicons included from two to five polymorphisms. All 40 polymorphisms can be amplified by a set of 21 primer pairs.