Page 1 of 12
Molecular Ecology Resources
Development and characterization of gene-derived markers in Brassica oleracea L.
Lorenzo Raggi, Simona Ciancaleoni and Valeria Negri
Dipartimento di Biologia Applicata, Università degli Studi di Perugia, Borgo XX Giugno 74, 06121 Perugia, Italy
Brassica oleracea, broccoli, gene-derived SSR, candidate genes, flowering, PIC Lorenzo Raggi, Borgo XX Giugno 74, +39 075 5856224,
[email protected]
Development of gene-derived SSR markers
Abstract Brassica oleracea L. species include extremely important vegetables for human and animal consumption and represent an essential part of many national diets. Unfortunately few molecular tools like gene-derived SSR, related to phenological specific traits, are available for this species. This study developed a set of markers for studying flowering control and cold stress response genes diversity that is useful to explore among and within population B. oleracea diversity. A total of 9 gene-derived markers were developed and 8 of them are based on SSR polymorphism. The markers have been tested in 64 genotypes of 4 Italian broccoli landraces with a total of 38 alleles produced and PIC value that ranged from 0.30 to 0.62. The developed markers, easy to score and highly polymorphic, represent a useful tool for the increase of knowledge on the genetic composition of wild populations, landraces and varieties of B. oleracea. and all the species that belong to its primary gene pool.
1
Molecular Ecology Resources
Page 2 of 12
Microsatellites or Simple Sequence Repeat (SSR) loci are valuable targets for the development of molecular markers because of their trend to evolve into different alleles based on contraction and expansion (Hancock 1999). They are user friendly polymorphic markers suitable for several diversity analysis, like differentiation of cultivars (Cipriani et al 2010), identification of duplicates in collections (Lund et al. 2003) and selection of appropriate parents in breeding programs (Zhou et al. 2006). The development of SSR markers have been limited because of time consuming and labor-intensive requirement to construct, enrich and sequence genomic libraries (Edwards et al. 1996). Thanks to the increase of available DNA sequences in several species, the production of gene-derived markers is easier today than before (Thiel et al. 2003), and many of them are available for model species. Gene-derived SSRs can be presently identified within existing public sequence databases also for Brassica oleracea L., although a few markers of this type are available (Love et al. 2004). The aim of this study was to identify, develop and test SSR markers located in candidate genes which are involved in flowering and in cold stress control in B. oleracea var. italica Plenck (broccoli). Initially we considered 14 gene involved in flowering control, namely FLOWERING LOCUS A (FCA), FLOWERING LOCUS C1 (FLC1), FLOWERING LOCUS C3 (FLC3), FLOWERING LOCUS C4 (FLC4), CONSTANS HOMOLOG (BN1CON19), CONSTANS LIKE (COL), LEAFY (LFY), FLOWERING LOCUS T (FT), APETALA 1 (AP1), APETALA 3 (AP3), CRYPTOCHROME 1 (CRY1), BABY BOOM (BBM), FRIGIDA (FRI), GIGANTEA (GI) (Chen et al. 2010; Babula 2007, Engelmann and Purugganan 2006; Robert at al. 1998; Okazaki et al. 2007; Pankin et al. 2008; Vorobiev et al. 2005, Passarinho et al. 2008; Lagercrantz et al. 1996; Kramer et al. 1998; Chatterjee et al. 2006; Shindo et al. 2005; Irwin et al. 2012; Razi et al. 2008; Lin et al. 2005, Wang et al. 2009; Carr and Irish 1997) and 1 gene related to cold stress response, namely COLDREGULATED PROTEIN (COR25) (Chen et al. 2011). The sequences of above mentioned genes were retrieved from the National Centre for Biotechnology (NCBI) database. SSR prediction, primer development and validation were performed as described for barley by Raggi and Negri (Molecular Ecology Resources Primer Development Consortium, 2012). A total of 20 SSRs located in promoter sequences, untranslated regions (UTRs), exons and introns of the 14 above mentioned genes were identified. SSRs in traslated as well as in untraslated regions can affect gene activation, gene regulation or gene transcription (Lawson and Zhang 2006). In addition, the presence of SSRs in the 5′-UTRs is required for the 2
Page 3 of 12
Molecular Ecology Resources
expression of some genes and their expression can change according to SSR length (Kalia et al.2010). Regarding FRIGIDA gene, since no repeated motif were identified, primers were designed on conserved sequences between Brassica and Arabidopsis with the aim to amplify a fragment including an intron. Introns flanking sequences are generally highly conserved between orthologous genes and therefore useful for primer development (Rogozin et al. 2005), while introns variability can be exploited to detect polymorphisms between individuals. For the research, 50 mg of fresh leaf tissue were collected from 24 genotypes selected from 3 different Sicilian landraces (popF, popI and popQ) and 40 genotypes from a landrace of Cental Italy (popSYN). In experimental trials these genotypes showed high morphological and phenological variation (Chiarenza 2008; Ciancaleoni et al. 2012). By using the DNeasy 96 Plant Kit (Quiagen), high quality genomic DNA was isolated from all the 64 genotypes. DNA quality and concentration were checked through spectrophotometry using NanoDrop 2000 (Thermo Scientific) and 2% (w/v) agarose gel separation. The markers developed in the above mentioned genes were initially tested on 8 randomly chosen genotypes. The Polymerase Chain Reaction (PCR) was performed in a 2720 Thermal Cycler (Applied Biosystems) in a volume of 12.5 µl containing the following: 20 ng of DNA, 1× PCR reaction Buffer with MgCl2, 0.5 mM dNTPs, 0.6 µM of each primer and 1U of Taq polymerase. The reaction mixture was initially denaturated at 94°C for 4’, followed by 10 cycles of amplification at 94°C for 30’’, 61°C for 30’’ and 72°C for 30’’. Then 28 cycles of amplification at 94°C for 10’’, 59°C for 15’’ and 72°C for 20’’ and 20’ at 72°C for final extension were performed. One or two amplification bands were visualized in 20 out of the 21 developed primers combinations after separating 5 µl of PCR products in a 3% (w/v) agarose gel. Only the primer pair developed for COL gene resulted in no amplification. According to polymorphism and amplification quality, 11 out of the 21 tested SSR were chosen for further analysis. The forward primer was 5’-end-labeled with one of the fluorescent tags 6-FAM, HEX, PET and VIC (Applied Biosystems). In order to test amplification quality and polymorphism, 1 µl of 1:20 dilution of the fluorescent amplicons, obtained from the 8 randomly chosen genotypes, was analyzed on AB3130xl sequencer (Applied Biosystems). Amplicons were then sized according to the internal size standard GeneScanTM 500 LIZTM (Applied Biosystems), visualized and scored using the GENEMAPPER software (Applied Biosystems).
3
Molecular Ecology Resources
Page 4 of 12
Peak morphology analysis allowed to finally select 8 SSRs and 1 gene derived marker that were considered to be the most useful for an accurate genotyping. Locus name, gene name, accession number, primers sequence, repeat motifs, fluorescence dye, expected and observed product size and position within the gene (i.e. intron or exon) of developed markers are reported in Table 1. The 9 developed markers were then amplified in the 64 genotypes; number of analyzed genotypes (N), observed (Na) and effective (Ne) alleles, observed (Ho) and expected (He) heterozygosity, null alleles frequencies [F(null)], Fixation index (F) and the Polymorphic Information Content (PIC) were worked out using GENALEX (Peakall and Smouse, 2006), PowerMarker (Liu and Muse, 2005) and CERVUS (Kalinowski et al., 2007) software. All the calculated parameters are reported in Table 2. Test for deviations from Hardy-Weinberg equilibrium (HWE) and genotypic linkage disequilibrium (LD) were performed using Arlequin 3.5.1.2 (Excoffier and Lischer, 2010). Tests for LD and HWE were performed for each population separately. The 9 developed markers resulted in 38 different alleles. Na ranged from 3 (Bo_FT-SSR9) to 5 (Bo_FLC1SSR1, Bo_CRY1-SSR17 and Bo_FRI-genederived19) with an average value of 4.22 (± 0.22 SE). Ne ranged from 2.98 (Bo_FRI-genederived19) to 1.44 (Bo_GI-SSR20) with an average value of 2.20 (± 0.16 SE). The Ho and He values ranged from 0.13 to 0.86 (Bo_COR25-SSR21 and Bo_FLC1-SSR1) and from 0.30 to 0.66 (Bo_GI-SSR20 and Bo_FRI-genederived19), respectively (average values of Ho=0.42 ± 0.08 SE and He=0.52 ± 0.04 SE). 90% of analyzed genotypes were scored as homozygous for Bo_COR25-SSR21 and 14% for Bo_FLC1-SSR1. F values ranged from -0.6 to 0.73 with an average value of 0.22 ± 0.13 SE. In four cases F values were positive and showed a trend to inbreeding. This result is not surprising since landraces are populations evolved under human and natural selection across generations and could have alleles close to fixation for specific characteristics (Polegri and Negri, 2010). In addition, positive F values are common in both natural populations as well as in landraces (Pressoir and Berthaud, 2003; Khaldi et al., 2012 ). F values calculated for the other loci underlined a trend to Hardy-Weinberg equilibrium or to outbreeding (Bo_FLC1SSR1). Null alleles with frequencies > 0.2 were present in 4 out of the 9 examined loci namely: Bo_AP3SSR13, Bo_CRY1-SSR17, Bo_GI-SSR20, Bo_COR25-SSR21. However CERVUS software estimate null alleles frequencies with the implicit assumption that genotype frequencies are in approximate HardyWeinberg equilibrium but in the studied populations this assumption is not always respected and null alleles frequencies can be overestimated because of the low values of Ho and He. The very low percentage of 4
Page 5 of 12
Molecular Ecology Resources
unsuccessfully amplifications (0.05%), which might result from the presence of null alleles in homozygous condition, allows to assume that the real percentage of null alleles is lower than estimated and therefore not negatively conditioning the usefulness of developed markers. PIC values calculated for developed markers ranged from 0.30 (Bo_GI-SSR20) to 0.55 (Bo_FLC3-SSR3) in trinucleotide SSRs and from 0.42 (Bo_COR25-SSR21) to 0.53 (Bo_CRY1-SSR17) in dinucleotide SSRs. In (Bo_FRI-genederived19) it was equal to 0.62. Significant deviation from HWE was detected for 2 loci: Bo_FLC1-SSR1 (P
