Persoonia Sm. (Proteaceae) is an endemic Australian genus of woody perennial plants containing 100 species, 43 of which are found in Western Australia (Weston, 1995). Persoonia elliptica R. Br. and P. longifolia R. Br. are widespread, small trees within the southwestern Australian jarrah forest and are priority species for mine site restoration practitioners. These two sympatric congeners are key components of the jarrah forest and share pollinators and seed dispersers, but contrast markedly in their population densities. There are no species-specific molecular markers available for these two diploid species. Here we report the isolation and characterization of microsatellite markers for both species using (1) cloning of microsatellite-enriched libraries (P. elliptica) and (2) 454 GS-FLX shotgun sequencing (P. longifolia). This will enable the examination of genetic diversity, range-wide genetic differentiation, and mating system parameters in the two species.
METHODS AND RESULTS
Fresh leaf material was collected from P. elliptica plants in the southwest of Western Australia (population code: Pe-AE, 31.572°S 116.181°E, collection number: JS102; population code: Pe-AW, 31.613°S 116.151°E, collection number: JS102). Herbarium material is deposited at the Kings Park and Botanic Garden Herbarium (KPBG), Perth, Western Australia, Australia. Fresh leaf material was also collected from P. longifolia plants in southwestern Western Australia (population code: Pl-MD, 32.176°S 116.245°E, voucher: R. Davis 3717 [PERTH 0481230]; and population code: Pl-CE, 33.356°S 116.345°E, voucher: A. Gundry 486b [PERTH 04801881]). The fresh leaf material was stored at−80°C and DNA extracted from the frozen material using the procedure described in Jobes et al. (1995) with the following modifications: 4 µL of proteinase-K was added prior to placing samples in the water bath and DNA was precipitated in the final stages by added 95% ethanol rather than isopropanol.
Genetic Identification Services (La Cañada Flintridge, California, USA; http://www.genetic-id-services.com/) was employed to develop microsatellite-enriched libraries for P. elliptica for four different repeat motifs (CAn, GAn, ACCn, and ATGn). Briefly, genomic DNA was restricted with seven blunt-end cutting enzymes (Rsa1, HaeIII, BsrB1, Pvull, StuI, ScaI, EcoRV). Fragments in the size range of 300–750 bp were linker adapted with oligonucleotides that contained a HindIII site and then subjected to magnetic bead capture (CpG Methy1Quest DNA Isolation Kit; EMD Millipore, Billerica, Massachusetts, USA). Molecules were restricted with HindIII and ligated into the HindIII site of the pUC19 plasmid. Ligation products were introduced into E. coli strain DH5α (ElectroMax; Invitrogen, Carlsbad, California, USA) by electroporation. Blue-white selection was used to identify recombinant clones for sequencing on an ABI PRISM 377 DNA sequencer (Applied Biosystems, Carlsbad, California, USA) using Amersham's DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Biosciences, Little Chalfont, Buckinghamshire, United Kingdom).
One hundred and six clones were sequenced, including 22 from the CA library, 22 from the GA library, 20 from the AAC library, and 22 from the ATG library. Eighty-two different microsatellite-containing clones were identified from the four libraries, and primers were designed for 72 sequences using DesignerPCR version 1.03 (Research Genetics Inc., www.lifetechnologies.com) and synthesized for 24. These primer pairs were initially tested to verify amplification, determine the optimum annealing temperature, and to establish size ranges for later PCR multiplexing, using DNA from eight individuals of P. elliptica. PCR was carried out in a total volume of 10 µL, containing approximately 10 ng genomic DNA template, 1× PCR Polymerization Buffer containing dNTPs (Fisher Biotech, Wembley, Western Australia, Australia), 0.2 µM each of unlabeled forward and reverse primer (GeneWorks, Hindmarsh, South Australia, Australia), 0.5 units Taq DNA polymerase (Fisher Biotech), and 2 mM of MgCl2 (Fisher Biotech). PCR was carried out in a Veriti 96-Well Thermal Cycler (Applied Biosystems) using the following reaction conditions: an initial activation step at 95°C for 15 min; followed by 35 cycles of 95°C for 30 s, annealing at 59°C for 90 s, extension at 72°C for 90 s; followed by a final extension at 72°C for 15 min. PCR products were separated on a 2% agarose gel stained with SYBR Safe DNA Gel Stain (Invitrogen), and fragment sizes determined by comparison to a Low DNA Mass Ladder (Invitrogen).
The nine loci that amplified successfully were then screened on eight individuals of P. elliptica to test for polymorphism and trial multiplexing groups. Multiplexing was performed on four groups of primers (Table 1) using the QIAGEN Multiplex Kit (QIAGEN, Valencia, California, USA) in 12.5-µL reaction volumes containing 5–30 ng DNA template, 6.25 µL QIAGEN Multiplex PCR Master Mix, 1.25 µL Q solution, 0.1 µM of each forward primer (labeled; unique to primer), 0.1 µM of each reverse primer (unlabeled), and sterile H2O to 12.5 µL. The multiplex PCR was conducted in a Veriti 96-Well Thermocycler (Applied Biosystems) with the following conditions: an initial activation step at 95°C for 15 min; followed by 35 cycles of 95°C for 30 s, annealing at 59°C for 90 s, extension at 72°C for 90 s; followed by a final extension at 72°C for 15 min. Following PCR, samples were diluted 1 : 30 in sample loading solution (Beckman Coulter, Brea, California, USA) with the addition of 0.4 µL fluorescently labeled 400-bp size standard per sample (Beckman Coulter) for capillary electrophoresis on a CEQ 8800 Genetic Analysis System (Beckman Coulter). Fragment peaks were visualized using CEQ Genetic Analysis System software (Beckman Coulter), and fragment (allele) sizes were scored manually. Eight individuals of P. longifolia were also genotyped using the above protocol, to test for cross-species amplification for these nine loci.
Genomic DNA of P. longifolia was sent to the Australian Genomic Research Facility (AGRF), Adelaide, Australia, for shotgun sequencing on a Titanium GS-FLX (454 Life Sciences, a Roche Company, Branford, Connecticut, USA), following Gardner et al. (2011). The sample occupied 6.25% of a plate and produced 108,806 individual sequences, with an average read length of 367 bp. The average GC content of these data was 37.82%. The program QDD version 1 (Meglécz et al., 2010) was used to screen the raw sequences for eight or greater di-, tri-, tetra-, or pentabase repeats, remove redundant sequences, and design primers using Primer3 (Rozen and Skaletsky, 2000). Software running parameters were set to default values, except PCR product lengths, which were set to 80–180 bp. We identified 14.2% of all reads containing microsatellite loci. Dinucleotide motifs were the most frequent (79,604), followed by tri-, tetra-, and then pentanucleotide motifs (14,550, 1072, and 445, respectively) (Meglécz et al., 2012). Primer pairs were designed for 108 loci and, from these, we excluded all loci that contained imperfect repeats, had a >2°C difference between the forward and reverse primer annealing temperature, short repeat motifs within the flanking region or primer sequence, or had poly A/T runs of >7 bp as there is an association between a high degree of poly A/T and instability (Li et al., 2002).
Characteristics of microsatellite primers developed in Persoonia elliptica and P. longifolia.
From the remainder, we arbitrarily selected 28 loci and then followed guidelines from Gardner et al. (2011) for further development; the loci were initially trialed for amplification. Initial PCR amplification was carried using the same method as for cloned microsatellites, mentioned above.
The 12 loci that amplified successfully were then screened on eight individuals of P. longifolia to test for polymorphism and trial multiplexing groups. Multiplexing was performed on four groups of primers (Table 1) using the QIAGEN Multiplex Kit in 12.5-µL reaction volumes containing 5–30 ng DNA template, 6.25 µL QIAGEN Multiplex PCR Master Mix, 1.25 µL Q solution, 0.1 µM of each forward primer (labeled; unique to primer), 0.1 µM of each reverse primer (unlabeled), and sterile H2O to 12.5 µL. The multiplex PCR was conducted in a Veriti 96-Well Thermocycler (Applied Biosystems) with an initial activation step at 95°C for 5 min; followed by nine cycles of 95°C for 30 s, a 1°C touchdown starting at 65°C for 180 s, and 72°C for 15 s; followed by 25 cycles at 95°C for 30 s, 56°C for 180 s, and 72°C for 15 s; and then a final extension at 60°C for 30 min. Capillary electrophoresis and fragment scoring was carried out using the same method as for cloned microsatellites, mentioned above. Eight individuals of P. elliptica were also genotyped using the above protocol, to test for cross-species amplification for these 12 loci.
One of the nine loci isolated from microsatellite-enriched libraries for P. elliptica did not amplify reliably in the target species (Table 1). Six out of 12 loci isolated from 454 pyrosequencing for P. longifolia were monomorphic, and three did not amplify reliably (Table 1). Cross-species amplification from P. elliptica to P. longifolia was successful for three loci (Table 1). Cross-species amplification from P. longifolia to P. elliptica was successful at one locus (Table 1).
More than 15 individuals from the two populations of both P. elliptica and P. longifolia were then genotyped using nine and six loci, respectively; all loci were polymorphic and amplified reliably from the previous screening. Genetic diversity parameters and deviation from Hardy–Weinberg equilibrium (HWE) were calculated using GenAlEx version 6.4 (Peakall and Smouse, 2006) (Table 2). We used MICRO-CHECKER 2.2.3 (van Oosterhout et al., 2004) to check each locus for evidence of null alleles, scoring error due to stuttering, and large allele dropout.
Results of primer screening in two populations each of Persoonia elliptica and P. longifolia. a
In P. elliptica, between three and 14 alleles per locus were found. Observed (Ho) and expected (He) heterozygosities ranged from 0.46 to 0.93 and 0.42 to 0.88, respectively (Table 2). For P. longifolia, between two and 13 alleles per locus were found. Ho and He ranged from 0.04 to 0.88 and 0.04 to 0.84, respectively (Table 2). Significant departure from HWE was detected in four of the nine loci for P. elliptica and three of the six loci for P. longifolia (Table 2). No loci showed significant null allele frequencies, large allele dropout, or evidence of scoring error due to stuttering.
The microsatellite markers developed here will enable an examination of population genetic patterns and realized dispersal of pollen and seed in P. elliptica and P. longifolia. The effort required in developing microsatellite markers has long been an important consideration, especially for low-budget studies on nonmodel organisms (Squirrell et al., 2003). Next-generation sequencing techniques such as 454 pyrosequencing have now been widely used for the development of microsatellite markers more rapidly and cheaply than traditional cloning approaches (Gardner et al., 2011; Malausa et al., 2011). Our experience suggests that, although there were efficiencies to be gained from employing next-generation sequencing for microsatellite marker development, particularly in the identification of microsatellite loci and primer sequences, it is not a panacea. A significant investment of time and effort in the screening, as well as optimization of polymorphic microsatellite markers, is still required with this relatively new approach to marker development.