Characterization of Simple Sequence Repeat (SSR) Markers and Genetic Relationships within Cultivated Peanut (Arachis hypogaea L.)

Peanut (Arachis hypogaea L.) probably originated in southern Bolivia or northern Argentina in South America (Gregory et al., 1980; Kochert et al., 1996). The genus Arachis contains approximately 70 species ( Krapovickas and Gregory, 1994), and almost all of them are diploid with either an A genome or a B genome. In contrast, the cultivated peanut species is a tetraploid (2n  =  4x  =  40 chromosomes) composed of both the A and B genomes.

In the past, morphological similarity was widely used for estimating variation within species (Ayana and Bekele, 1999), and variations in morphological, physiological, and agronomic traits have also been reported in peanuts. Recently, development of DNA molecular markers, such as RAPD (random amplified polymorphic DNA), AFLP (amplified fragment length polymorphism), and SSR (simple sequence repeat), have been used to assess genetic variability and conduct evolutionary studies in different crops (Burr and Burr, 1991; Akkaya et al., 1992; Rongwen et al., 1995; Cho et al., 2000; Temnykh et al., 2000; Gethi et al., 2002; Robertson-Hoyt et al., 2006). There are many advantages of using these molecular markers including high repeatability, no genotype × environment (G × E) interaction, and no off season limitation. In addition to assessing genetic variability, these markers also can be very useful for characterization of individuals and breeding lines for choosing parental genotypes in breeding programs (Ribaut and Hoisington, 1998). Previous research has indicated a low level of genetic variability in cultivated peanut (Halward et al., 1991, 1992; Kochert et al., 1991; Stalker et al., 1994; He and Prakash, 1997; Gupta and Varshney, 2000). Compared to other kinds of markers such as RFLPs and RAPDs, SSR markers are co-dominant and multi-allelic in inheritance, have higher level of DNA polymorphism in cultivated peanut, and are easier to amplify with less DNA quantity and low cost (Hopkins et al., 1999; Tang et al., 2003; He et al., 2003). The identification and characterization of molecular markers in cultivated peanut has lagged behind other economically important crops such as soybean, maize, and rice because of the limited genetic variability in cultivated peanut, and the added complexity of having two genomes.

In this research, the majority of SSR polymorphic markers detected by different research groups were collected and screened with sixteen cultivated peanut genotypes having a range of field reactions to Tomato spotted wilt virus (TSWV), Cercospora arachidicola (early leaf spot) and Cercosporidium personatum (late leaf spot). The objectives of this study were: 1) to characterize the SSR markers based on allele-length ranges, average PIC (polymorphism information content) score and heterozygosity among 16 genotypes of cultivated peanut; 2) to detect the genetic relationships among these 16 genotypes; and 3) to explore potential genotypes for use as parents for recombinant inbred line (RIL) population development for genetic mapping.

A diverse array of 16 genotypes tested in this study was obtained from USDA-ARS and the University of Georgia peanut programs at the University of Georgia Coastal Plain Experiment Station in Tifton, GA in 2006 (Table 1). These genotypes included commercial cultivars as well as breeding lines and germplasm accessions from the U.S. and China. Some of the lines have been previously evaluated for resistance to TSWV and leaf spot pathogens, and wide ranges of variation for resistance to TSWV, C. arachidicola and C. personatum have been observed. ‘Georgia Green’ (Branch, 1996) was a predominant cultivar planted in Georgia with a moderate level of field resistance to TSWV and has been used as a standard moderate resistance cultivar to TSWV ( Culbreath et al., 1999, 2008). ‘Georganic’ (tested previously as C11-2-39) (Holbrook and Culbreath, 2008) and ‘Georgia-01R’ (Branch, 2002) are cultivars with resistance to TSWV and moderate levels of resistance to the early and late leaf spot pathogens ( Holbrook et al., 2008a, 2008b). ‘Tifguard’ (tested previously as C724-19-15) and a near-isogenic sister line ‘C724-19-25’ also have field resistance to TSWV but differ in susceptibility to the peanut root knot nematode ( Holbrook et al., 2008b). ‘Tifrunner’ ( Holbrook and Culbreath, 2007) was released in 2007 as a new cultivar with field resistance to TSWV and moderate resistance to early and late leaf spots. ‘SunOleic 97R’ (Gorbet and Knauft, 2000) was a cultivar developed by the University of Florida with good agronomic traits including high oleic acid oil composition (>80%), but it is susceptible to TSWV (Culbreath et al., 2005) and leaf spot pathogens. The breeding line ‘F NC94022-1-2-1-1-b3-B’ (hence forth referred to as ‘NC94022’) has been reported to have a high level of field resistance to TSWV(Culbreath et al., 2005).

Leaves of 16 genotypes were collected from greenhouse grown plants at Tifton, GA for total DNA extraction. The fresh leaf tissues were frozen at −80°C after harvesting, and then ground in liquid nitrogen with a mortar and pestle. The total genomic DNA was extracted according to a modified CTAB method (Tang et al., 2002) from fresh leaves. DNA concentration and quality were determined by use of a spectrophotometer at 260 nm wavelength and a ratio of 260nm/280nm. Afterwards, DNA was diluted in sterile water to 10 ng/ul for PCR reaction.

A total of 709 pairs of SSR markers were collected and used for screening polymorphisms in this study (Table 2). Among them, 97 pairs of primers were newly developed from genomic sequences in Dr. Steve Knapp’s Laboratory at the University of Georgia (Table 3). An additional 612 pairs of primers were chosen from 6 different research groups’ published data. Initially, SSR markers were screened on 1.5% agarose gels for utility, functionality, and length estimation using the bulk DNA of 4 randomly chosen cultivated peanut samples. From the initial screen, 153 SSR markers were deleted with no amplification. Therefore, only 556 SSR markers passed the initial screen and were used to assess polymorphisms in the 16 genotypes. These SSR primers were synthesized by MWG Biotech (High Point, N.C., USA). Forward primers were modified by adding different fluorescent phosphoramidite (6FAM, HEX, or TAMRA) to the 5′ ends. Two hundred thirty-seven primers were labeled with 6-FAM (6-carboxyfluorescein), 236 primers were labeled with HEX (hexachloro-carboxyfluorescein), and the remaining 236 primers were labeled with TAMRA (tetramethyl rhodamine) (Tang et al., 2002, 2003).

‘Touchdown’ PCR (Don et al. 1991) was used to minimize spurious amplification. An 11.5 ul reaction volume in 384-well plates was used for PCR reaction, containing 1 ul of forward and reverse primers, 2 ul of DNA template, 1.15 ul of 10 × PCR buffer, 0.25 ul of dNTP (2.5 mM) , and 0.1 ul of Taq enzyme. The ‘touchdown’ thermal cycle of the PCR reaction was decided by the thermal temperature (Tm) for different kinds of primers separately and was applied with either 52°C or 56°C ‘touchdown’ cycle reaction. The ‘touchdown’ amplification program is as follows: 94°C for 1 min to allow samples to denature, followed by 6 cycles of 94°C for 30s, 62°C for 30s, and 72°C for 30s, the annealing temperatures were decreased 1°C per cycle in subsequent cycles until the temperature reached 52°C or 56°C for the different kinds of ‘touch down’ programs. Products were subsequently amplified for 36 cycles at 94°C for 20s, 56°C for 20 s, and 72°C for 30 s with a final extension for 20 min.

PCR-multiplexes were used based on the multiplex PCR principles described by Edwards and Gibbs (1994) and Henegariu et al. (1997), and performed as described by Tang and Knapp (2003) using post-PCR multiplexing of six amplicons (samples were diluted 60- to 100-fold). The criteria used to select SSR markers for the PCR-multiplexes were primer compatibility, genotyping performance when amplified by multiplex PCR, and allele length range (estimated from the SSR polymorphism initial screen). The SSR markers were sorted by allele-length range and combined so as to minimize the comigration of identically labelled non-allelic bands.

The PCR products were first checked with 1.5% agarose gel to ensure the successful amplifications, and then were diluted 60- to 100- fold before analysis using an ABI3730. Those six different diluted amplicons with different fluorescent labels and amplification length were mixed into one well by 1 uL with 9 ul of formamide containing a GeneScan 500 internal lane standard labeled with ROX. GeneScan Filter Set D and the ROX 500 internal-standard were used for analyses of amplicons labeled with FAM, HEX, and TAMARA. Gene Mapper 4.0 was used for allele scoring for the 556 functional SSR markers based on the multiplex mixture with different fluorescent labels and amplification lengths.

The screening results for the 556 markers were recorded as 1 for presence of the amplification band, 0 for absence of the amplification band. According to the amplification quality and reliability across 16 genotypes, the markers were classified as 1–5, where 1 =  Excellent, indicated by perfect amplification with unambiguous product within all 16 genotypes; 2  =  Good, indicated by clear peak with some magnification; 3  =  Fair, indicated by clear peak with high magnification and 1–2 genotypes with null amplification results; 4  =  Poor, indicated by bad amplification results and 2–4 genotypes with null amplification results; 5  =  Not good, indicated that more than half genotypes had null amplification results. Failed PCR amplifications were scored as missing data for genotyping.

The PIC (polymorphism information content) value, defined by Botstein et al. (1980) as a closely related diversity measure is a measure of the polymorphism of a marker (SSR marker) for linkage. The formula for this estimation of PIC score is:

Where l  =  index for marker ‘l’;  =  proportion of marker ‘l’ alleles which are of allele type ‘u’;  =  proportion of marker ‘l’ alleles which are of allele type ‘v’; k  =  number of alleles types present for marker ‘l’ (Shete et al., 2000). Another estimator, heterozygosities (H) were estimated for each SSR marker as described by Ott (1991) according to the formula:

Where Pi is the frequency of i^th allele in the genotype population (Shete et al., 2000). The estimation of PIC, amplified allele number, allelic frequency and observed heterozygosity were obtained by PowerMarker (Liu and Muse, 2005).

Estimators of genetic diversity within these 16 genotypes were based on the shared alleles distance (Dps) in pairwise comparisons. The estimates of genetic distance were calculated in MicroSat as the parameter: [1- Dps]. The cladogram trees were obtained by the PHYLIP program with Neighbor-Joining method based on the genetic distance matrix obtained as described above. Trees were drawn using Tree View.

Among 556 amplifiable markers, 80 pairs of primers were suitable for 52°C ‘touchdown’ PCR reaction, and the others worked with 56°C ‘touchdown’. The number of primers in the five classified rating groups was 281, 95, 85, 28, and 67 from excellent to not-good, respectively. Out of the 556 functional markers in the original screen, the 281 (50.5%) markers classified in the excellent group gave reliable unambiguous results and were used for the analysis of genetic diversity among the 16 genotypes. A total of 235 (42.3%) markers showed polymorphisms. The percentage of polymorphism detected in this study was relatively high when compared to other DNA markers such as RFLPs and RAPDs (He et al., 2005). However, only 83 of the polymorphic markers (29.5%) were classified among the 281 excellent markers, which was lower than expected. There were 42 (44.2%) and 49 (57.6%) polymorphic markers from good and fair groups, respectively. The primers that generated ambiguous bands and no amplification within one or more genotypes produced higher estimates of polymorphism. Therefore, we used the data generated by the 461 markers from excellent, good, and fair groups with reasonable amplification quality and repeatability detected by GeneMapper. The average heterozygosity estimated from 556 SSR markers was 0.225 with a range of 0 to 0.992 for individual SSR markers and the average of PIC was 0.209.

The average number of SSR alleles among 709 pairs of primers tested in this study was 2.5 with a range of 1 to 13 among the 16 peanut genotypes. However, only one allele was found in 410 of these SSR markers in the 16 genotypes, indicating that diversity of cultivated peanuts is very limited. Among those 125 polymorphic SSR markers from the excellent and good quality groups, 26.4% of the markers were dinucleotide GA repeat motif SSR markers, followed by dinucleotide CT (10.4%), and trinucleotide TAA (9.6%). This is in agreement with the earlier report (Ferguson et al., 2004). This research validated that the dinucleotide and trinucleotide repeat motifs are the most abundant type of SSR, and dinucleotide GA repeat motif shows a higher polymorphism compared to other trinucleotide and polynucleotide repeat motifs (Moretzsohn et al., 2005).

The genetic distance matrix was estimated by shared allele distance in pairwise comparisons of 16 peanut genotypes based on 281 excellent quality markers with unambiguous bands (Table 4). The genetic distances among the 16 tested genotypes varied from 0.002 to 0.169. The smallest genetic distance was between ‘Tifton8’ and ‘Tifrunner’, where both were developed in Tifton, Georgia, but no apparent similarity in genetic background (Holbrook and Culbreath, 2007; Coffelt et al., 1985) and the largest genetic distance was between ‘Georganic’ (a typical runner-type) and ‘GTC9’ (a typical Spanish-type).

A horizontal rectangular cladogram of the 16 genotypes was constructed based on the genetic distance matrix with Neighbor-Joining method (Figure 1). The Neighbor-Joining method assigned the 16 genotypes into A, B, C, D, E clades or clusters and three outliers (‘Tifton 8’, ‘Tifrunner’, and ‘Georganic’) (Figure 1). Basically, the assigned clusters were in agreement with either pedigree or origin. The sister lines ‘Tifguard’ and ‘C 724-19-25’ were grouped as cluster A. Two Virginia type peanuts, ‘GPNCWS13’ and ‘NC6’, were grouped as cluster B. ‘Georgia Green’ (GaGreen) was grouped together with the released cultivar ‘SunOleic 97R’ in cluster C. Cluster D included two Chinese released cultivars (‘GTC20’ and ‘GTC9’) and one U.S. released cultivar ‘Spancross’ (Hammons, 1970), which are Spanish-type. Cluster E consisted of two Chinese breeding lines (‘PE1’ and ‘PE2’) and another U.S. released cultivar ‘Tennessee Red’ (TNRed), which are Spanish-type and Valencia-type, respectively.

A total of 709 SSR markers were collected from public databases which will be a valuable asset to the peanut research community; 556 SSRs passed an initial screen and were used to characterize 16 diverse peanut genotypes. PIC scores and heterozygosity indices were calculated to assess the genetic diversity of SSR markers, and genetic distances were also estimated from shared allele distances for construction of a cladogram to illustrate the genetic relationships among the tested genotypes. Although the majority of the SSR markers used in this study have been developed and screened in other Arachis genotypes, polymorphism information has been expanded using different genotypes. This is obvious because different genotypes carry different SSR alleles. The previous genotypes used for polymorphism screening were mainly from South America (He et al., 2003; Ferguson et al., 2004; Moretzsohn et al., 2004, 2005). In this study, 12 genotypes are from the U.S. and 4 genotypes are from China. The SSR markers developed by Ferguson et al. (2004) were not as polymorphic in this study in comparison to their report, in which 110 markers out of 192 showed polymorphism. The peanut genotypes used in their research were from different origins and different market types, so they may have had more diverse genetic backgrounds compared to the genotypes in this study.

Out of 56 designed markers developed from a microsatellite enriched library by He et al. (2003), 19 markers (34%) were polymorphic in their research, but 22 markers showed polymorphism in this study. Those 24 genotypes reported by He et al. (2005) consisted of different botanical varieties including hypogaea, hirsuta, fastigiata, peruviana, aequatoriana, and vulgaris with the origin of South America. Similar polymorphism results were also observed for the markers developed by Moretzsohn et al. (2004, 2005) who reported that 3 markers out of 67 designed markers were polymorphic within 5 cultivated peanut lines and 66 markers out of 271 designed markers were polymorphic within six A. hypogaea accessions mainly from Brazil. Although it is apparent that different sets of genotypes exhibit different levels of polymorphisms of SSR markers, the origin of DNA sequence information that was used for development of SSR markers also could be responsible for some of the observed differences. The markers developed from Arachis pintoi instead of Arachis hypogaea (Palmieri et al., 2002, 2005) did not show polymorphism among the 16 genotypes in this study. Among the 97 newly developed SSR markers (Table 3), 40 markers showed polymorphisms, which is a relatively high rate of polymorphism.

This research also corroborated that the dinucleotide and trinucleotide repeat motifs were the most abundant type of SSRs, and dinucleotide GA repeat motif showed a higher polymorphism compared to other trinucleotide and polynucleotide repeat motifs (Moretzsohn et al., 2005). The most frequent repeat family identified was dinucleotide GA, which was also reported by Ferguson et al. (2004), followed by dinucleotide CT (10.4%), and trinucleotide TAA (9.6%).

Most of the genetic relationships revealed from the cladogram by SSR alleles corresponded well to the pedigrees and origins of the genotype pairs. One of the closest genetic relationships was from ‘Tifguard’ and ‘C 724-19-25’. These are sister lines and have been reported as near-isogenic (Holbrook et al., 2008b). Although they both were developed by crossing of ‘C-99R’ and ‘COAN’, they differ greatly in reaction to the peanut root-knot nematode. ‘Tifguard’ is a released cultivar with a high level of resistance to peanut root-knot nematode, whereas, ‘C724-19-25’ is susceptible to the nematode (Holbrook et al., 2008b). The largest genetic distance was between ‘Georganic’ and ‘GTC9’. ‘GTC9’ belongs to the Spanish type and was developed in China, while ‘Georganic’ is a runner type peanut developed in Georgia. The genetic backgrounds of ‘Georganic’ and ‘GTC9’ are distinct. Two Chinese breeding lines ‘PE1’ and ‘PE2’ were grouped together in cluster E with U.S. released cultivar ‘TNRed’, which was unexpected. The two Chinese breeding lines belong to Virginia type, whereas TNRed is a Valencia type. The similarity of ‘NC 6’ and ‘GPNPWIS-13’, placed together in cluster B, was also not surprising since ‘NC 6’ was one of the parents used to develop ‘GP-NC WS 13’ (Stalker et al., 2002). However, there is no apparent close relationship between ‘Tifton8’ and ‘Tifrunner’, which had the smallest genetic distance. ‘Tifunner’ was developed from a cross of a component line of the cultivar ‘Florunner’ and PI 203396 (Holbrook and Culbreath, 2007), whereas ‘Tifton-8’ was developed as a Virginia-type peanut from a selection from a Spanish-type, PI 261976 (Coffelt et al., 1985).

The information gained in this study has been used in genotype selection for genetic mapping population development and the markers screened in this study are also used in genotyping of the developed RILs (recombinant inbred lines). Two RIL populations have been developed from crosses of Tifrunner × GTC20 and SunOleic 97R × NC94022 (Guo et al., unpublished data). Furthermore, this study will contribute to the construction of genetic linkage maps for cultivated peanut, which will be an important research goal to facilitate quantitative trait locus (QTL) analysis and gene tagging for use in marker-assisted breeding.

We thank Billy Wilson and Jake Fountain for technical assistance in the field and the members of Dr. Knapp’s Laboratory particularly Drs. Tang and Ma. This research was partially supported by USDA Specific Cooperative Agreement 58-6602-6-121 with the University of Georgia, and the graduate student assistantship was partially supported by funds provided by Georgia Peanut Commodity Commission and the National Peanut Board. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture.