Rhizopogon Fr. (Rhizopogonaceae) is a fungal genus of species that form ectomycorrhizal symbioses with conifers and are pivotal components of temperate forest ecosystems (Molina et al., 1999). Four species (R. kretzerae Grubisha, Dowie & Mill., R. salebrosus A. H. Sm., R. ellenae A. H. Sm., and R. arctostaphyli A. H. Sm.) in Rhizopogon subgenus Amylopogon (A. H. Sm.) Grubisha & Trappe are obligately ectomycorrhizal with several pine species and are the symbiotic mycobiont host to the mycoheterotrophic plant, Pterospora andromedea Nutt. (Ericaceae; Bidartondo and Bruns, 2002; Dowie et al., 2011, 2012; Hazard et al., 2012; Grubisha et al., 2014), a North American endemic divided into eastern and western populations (Bakshi, 1959). Pterospora andromedea primarily associates with two closely related Rhizopogon species: western P. andromedea are symbiotic primarily with R. salebrosus while eastern populations are found with R. kretzerae (Grubisha et al., 2014). Rhizopogon kretzerae is currently only known from collections of P. andromedea roots and only found in association with Pinus strobus L. (Hazard et al., 2012; Grubisha et al., 2014). Eastern P. andromedea populations have undergone recent decreases in population size and possible local extinction from a variety of anthropogenic factors (Schori, 2002).
Copopulation genetic studies of the fungal host and mycoheterotrophic plant will (1) provide information about levels of genetic variation and population structure at different hierarchical levels in plant and fungal populations, and (2) be important in developing conservation management plans for the rare to endangered eastern populations of both species. While microsatellite loci have been previously developed for distantly related species in three of the five Rhizopogon subgenera (Kretzer et al., 2000, 2004; Grubisha et al., 2005), the microsatellite loci reported here, which were developed from R. kretzerae and R. salebrosus, are the first for any species in Rhizopogon subgenus Amylopogon.
METHODS AND RESULTS
Genomic DNA (gDNA) was isolated from 60 mg of silica gel-dried pure cultures of R. kretzerae isolate Miller207 (Michigan; Table 1) and R. salebrosus isolate Miller322 (Washington; 46.20545°N, 117.7704°W) separately using the DNeasy Plant Mini Kit (QIAGEN, Valencia, California, USA). These two cultures were obtained from P. andromedea monotropoid roots following Grubisha et al. (2014). Each culture was ground in liquid nitrogen to a fine powder. Twice the recommended volume of buffer AP1, RNaseA, and buffer AP2 were used. Library construction and 454 shotgun pyrosequencing were performed by the Advanced Genetic Technologies Centre (AGTC) at the University of Kentucky (Lexington, Kentucky, USA). Libraries for each species were prepared using the GS FLX+/XL+ Titanium Rapid Library Preparation Kit (454 Life Sciences, a Roche Company, Branford, Connecticut, USA). Instead of nebulizing, gDNA was sheared using a Bioruptor NGS (UCD-600TS; Diagenode, Denville, New Jersey, USA) by sonicating at 4°C with six cycles of 5 s on and 90 s off. Shotgun 454 pyrosequencing was performed on 25% of a plate on a Titanium GS-FLX (454 Life Sciences, a Roche Company) yielding a total of 152,007 reads with an average length of 414 bp for R. kretzerae and a total of 92,252 reads with an average length of 517 bp for R. salebrosus. At the AGTC, raw sequence reads were initially processed using prinseq-lite.pl ( http://prinseq.sourceforge.net/) and involved removing all (1) reads with >2% length with uncalled bases (N), (2) duplicate reads, (3) reads with Q < 17, (4) reads too short in length [average length − 2(standard deviation)], and (5) reads too long in length [average length + 2(standard deviation)]. Processed reads were assembled de novo using Newbler version 2.8 (454 Life Sciences, a Roche Company) at the AGTC, which produced 10,190 contigs for R. kretzerae (average contig size = 1502 bp, N50 = 1701 bp) and 5178 contigs for R. salebrosus (average contig size = 1302 bp, N50 = 1428 bp).
Genetic properties of 17 microsatellite loci developed from Rhizopogon kretzerae and one locus from R. salebrosus , and potential applicability in R. arctostaphyli and R. ellenae
Perfect microsatellite repeats were identified from the assembled contigs using MSATCOMMANDER version 1.0.8 (Faircloth, 2008) for R. kretzerae and R. salebrosus, yielding 43 and 21 dinucleotides with at least six repeat units, 137 and 48 trinucleotides with at least five repeat units, 34 and 12 tetranucleotides with at least four repeat units, five and 11 pentanucleotides with at least four repeat units, and five and 11 hexanucleotides with at least four repeat units, respectively. Within MSATCOMMANDER, Primer3 (Rozen and Skaletsky, 2000) was used to generate primers for all loci using default parameters except that difference in acceptable primer melting temperature was 2°C with an optimum of 60°C; maximum poly X was set to three; GC clamp = yes; and a target PCR product size of 90–210 bp. Selection of loci for initial screening was based on Primer3 output tabulated in a spreadsheet by MSATCOMMANDER, and those that had no lowercase letters in the primer sequences that would reflect nucleotide mismatches across reads in assembled contigs. Forty-eight loci (38 identified from R. kretzerae and 10 from R. salebrosus) were initially screened across four samples (two R. kretzerae, two R. salebrosus; Table 1) for positive PCR products using agarose gel electrophoresis. Loci were PCR amplified using 0.1 µM of each forward and reverse primer in a 10-µL reaction volume that included 1× CoralLoad PCR Buffer with 1.5 mM MgCL2 (QIAGEN), 200 µM each dNTP, 2.5 units Taq DNA Polymerase (QIAGEN), and 1.0 µL of 1:10 diluted genomic DNA. Thermocycler parameters were 94°C for 3 min, 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s, followed by a final extension for 10 min at 72°C. PCR products were visualized in a 2% agarose gel stained with GelRed (Phenix Research Products, Candler, North Carolina, USA) on a UV transilluminator. Of the 48 loci assessed, 27 loci (23 from R. kretzerae and four from R. salebrosus) yielded no more than one or two bright PCR product bands within the approximate expected size range (target size +40 bp, −10 bp) and no PCR products outside the expected size range. The lower threshold was set to target size −10 bp due to the relatively low number of repeats.
These 27 loci were screened in 65 isolates from two populations each for R. kretzerae (n = 14 and n = 11; Table 1) and R. salebrosus (n = 20 both; Table 1), and for cross-amplification in two distantly related Rhizopogon subgenus Amylopogon species that are also hosts for P. andromedea: R. arctostaphyli (n = 2 for both; Table 1) and R. ellenae (n = 2 for both; Table 1). Pterospora andromedea ectomycorrhizal roots were collected and fungal DNA was isolated as previously described (Dowie et al., 2012; Grubisha et al., 2014). Due to the extremely small sample of rootball collected, voucher specimens in herbaria were not made except for samples from Michigan (DAOM 242738) and Quebec (DAOM 242739) that were previously deposited at the Canadian National Mycological Herbarium (Grubisha et al., 2014). The forward primer was 5′ end-labeled with one of four dyes in the AB I DS-33 dye set (Table 2). Loci were PCR amplified in multiplex reactions using the QIAGEN Multiplex PCR Kit in a 10-µL volume with 1× QIAGEN Multiplex PCR Master Mix, 50 nM each primer, 0.25–0.5x Q-solution, and 1.0 µL of 1:10 diluted genomic DNA. Touchdown thermocycler conditions were: 95°C for 15 min; 10 cycles of 94°C for 30 s, 67°C for 90 s, decreasing 1°C each cycle, and 72°C for 30 s; 25 cycles of 94°C for 30 s, 57°C for 90 s, 72°C for 30 s; with a final extension of 45 min at 60°C. Primers with a higher or lower touchdown annealing profile are listed in Table 2. Fragment analysis was conducted using the GeneScan 500 LIZ Size Standard (Applied Biosystems, Foster City, California, USA) on an ABI 3730 DNA Analyzer (Applied Biosystems) by the Biotechnology Resource Center (BRC) at Cornell University. Allele sizes were called manually using the Microsatellite Plugin in Geneious version R6.1.5 (Drummond et al., 2011). Loci selected for the final set were PCR amplified independently to verify allele size ranges.
Of the 23 loci developed in R. kretzerae and screened with fluorescently labeled primers in R. kretzerae populations, 12 loci were polymorphic (Table 1), seven loci were monomorphic, and four loci had excessive stutter. In R. salebrosus populations, 21 of the 23 loci developed in R. kretzerae successfully PCR amplified and 11 loci were polymorphic, four of which were monomorphic in R. kretzerae populations (Table 1). The four loci isolated in R. salebrosus did not PCR amplify in R. kretzerae populations screened. In R. salebrosus populations, only one locus (Rsa12) was polymorphic (Table 1). A few loci cross-amplified in R. arctostaphyli and R. ellenae isolates and may prove to be useful when screened at a larger scale (Table 1). Locus Rkre 19 had stutter problems in both R. kretzerae and R. salebrosus but was polymorphic in R. arctostaphyli and R. ellenae (Table 1).
Characteristics of 12 polymorphic and five monomorphic microsatellite loci developed in Rhizopogon kretzerae and one polymorphic locus developed in R. salebrosus.
Number and size range of alleles, and observed (Ho) and expected heterozygosities (He) were calculated in GenAlEx version 6.5 (Peakall and Smouse, 2006, 2012; Table 1). When both populations for each species were considered (R. kretzerae, n = 25; R. salebrosus, n = 40), for the loci isolated in R. kretzerae (Rkre3–Rkre38; Table 1), there were one to eight alleles, and He ranged from 0.00 to 0.64 in R. kretzerae populations (Table 1); and there were one to nine alleles and He values of 0.00 to 0.76 in R. salebrosus populations (Table 1). For locus Rsal2 there were one to three alleles, and He ranged from 0.00 to 0.10. Scoring errors due to stuttering, large allele dropout, and null alleles were checked using MICROCHECKER 2.2.3 (van Oosterhout et al., 2004). For R. kretzerae populations, two loci showed signs of possible null alleles (Rkre5 in Quebec and Rkre37 in Michigan). In the combined R. salebrosus populations, possible null alleles and scoring errors due to stuttering were detected at four loci (Rkre3, Rkre 17, Rkre27, Rkre34).
These are the first microsatellite loci developed for any species in Rhizopogon subgenus Amylopogon. Presently we are using these loci to assess the conservation genetic status of R. kretzerae, which will be important for developing conservation management plans for eastern R andromedea. We are also conducting copopulation genetic studies to determine if this fungal host and mycoheterotrophic plant share similar patterns of population genetic structure and variation at different hierarchical levels. Cross-amplification of loci isolated in R. kretzerae with R. salebrosus and two distantly related Rhizopogon subgenus Amylopogon species suggests they may be useful in studies of other species in this subgenus.