The Value of Diploid Peanut Relatives for Breeding and Genomics

Wild species of Arachis are native to a large region of South America in tropical and subtropical areas. Eighty species have been named (Krapovickas and Gregory, 1994; Valls and Simpson, 2005) and others will likely be described from uncollected regions of South America. Wild peanut species have been important to man since before the cultivated peanut evolved, and they are still used as forages, for their esthetic value, and as sources of germplasm for crop improvement. For example, A. glabrata and A. pintoi are utilized for grazing (Mathews et al., 2000; Hernandez-Garay et al., 2004) and A. repens is used as a ground cover in residential areas and roadsides in tropical regions. Several species have been consumed for their seeds, but only A. hypogaea is economically important today as a human food source. The primary interest in wild species of Arachis for the past 50–60 years has been as sources of disease and insect resistances for crop improvement. Several species, including ones that will hybridize with the cultivated peanut, have extremely high levels of disease resistance and a few (e.g., A. diogoi) have virus resistance genes that are not present in the cultivated gene pool.

Arachis hypogaea is an allotetraploid (2n  =  4x  =  40) with a very large and complex genome. Chromosome behavior and morphology were reported by Husted (1936) and Stalker and Dalmacio (1986). Gregory (1946) reported the first chromosome number of a wild species (A. glabrata) as 2n  =  4x  =  40 and a year later Mendes (1947) reported diploid species (2n  =  2x  =  20). Not until 2005 were species with 18 chromosomes discovered (Penaloza and Valls, 2005). Most species in the genus are diploid, but tetraploids exist in sections Arachis and Rhizomatosae; and several species in sections Arachis and Erectoides are aneuploid (2n  =  2x  =  18) (Table 1). Polyploidy evolved independently in sections Arachis and Rhizomatosae (Smartt and Stalker, 1982); and Nelson et al. (2006) concluded that polyploidy evolved multiple times within section Rhizomatosae. Tallury et al. (2005) reported molecular evidence that the diploid section Rhizomatosae species (only one known) did not give rise to the tetraploids. Because A. glabrata will hybridize with species of both sections Erectoides and Arachis, Smartt and Stalker (1982) concluded that two diploids from sections Erectoides and Arachis likely hybridized and spontaneously doubled in chromosome number, evolving into the tetraploid species of section Rhizomatosae.

Krapovickas and Gregory (1994) concluded that Erectoides, Extranervosae, Heteranthae, Trierectoides and Triseminatae are “older” sections, while Arachis, Caulorrhizae, Procumbentes, and Rhizomatosae are more “recent” in origin. The largest subgeneric group is section Arachis, which includes the cultivated species, one other tetraploid (A. monticola), 26 diploid (2n  =  2x  =  20), and three aneuploid (2n  =  2x  =  18) species. These species are highly variable, and especially the annual species are continuing to differentiate. For example, A. duranensis has northern and southern groups which can be distinguished morphologically and with molecular markers (Stalker et al., 1995).

The first published attempt at interspecific hybridization in the genus was between the two tetraploids A. hypogaea (section Arachis) and A. glabrata (section Rhizomatosae) (Hull and Carver, 1938), but no hybrids were obtained. Krapovickas and Rigoni (1951) later hybridized A. hypogaea with A. villosa var. correntina and the F₁s were vigorous, but sterile. The cultivated peanut has since been hybridized with most species in section Arachis. Similar to other genera which have a polyploid series, crosses are usually more successful when the species at the higher ploidy level (in this case A. hypogaea) is used as the female parent. Triploid interspecific hybrids usually have 10 bivalents and10 univalents, but trivalents are also observed in pollen mother cells which indicates that some chromosome homology exists between the A and B genomes (Stalker, 1985).

Early cytological research identified one pair of significantly smaller chromosomes (termed ‘A’ chromosome) in species of section Arachis and a unique chromosome pair that had a large secondary constriction (termed ‘B’ chromosome) in the species A. batizocoi (Husted, 1936). Hybridization between diploid species was first reported between A. duranensis and A. villosa var. correntina (Raman and Kesavan, 1962) and meiosis was regular. Later studies indicated that hybrids between species having the small chromosome pair were partially to fully fertile and most will produce F₂ seeds; however, hybrids between the species with the small chromosome and A. batizocoi are sterile (Stalker and Simpson, 1995). Thus, the terminology ‘A’ and ‘B’ genome has been used in peanut to describe the two cytological groups. Because the cultivated peanut has one pair of smaller chromosomes and one pair of chromosomes with a large secondary constriction, it was described as an allotetraploid with AABB genomes. Stalker et al. (1991) crossed a series of species designated as having the A genome with A. batizocoi and found that F₁s had many univalents, and bivalents were loosely associated. Hybrids between either A or B genome species with A. glandulifera (D genome) also have many univalents and are sterile (Stalker et al., 1991). Thus, there is a considerable amount of cytological differentiation between the three genomes.

Gregory and Gregory (1979) conducted an extensive hybridization program using 91 Arachis collections and reported cross-compatibility relationships among species of the genus. Their results indicated that hybridization between species within the same section is more successful than crosses between sections; F₁s of intersectional crosses were highly sterile. To overcome crossing barriers, complex hybrids have been attempted (Gregory and Gregory, 1979; Stalker, 1981), but fertility has not been restored. Thus, introgression from wild Arachis species to A. hypogaea by conventional hybridization is believed to be restricted to members of section Arachis. Even within section Arachis there are many difficulties encountered to obtain hybrids due to genomic and/or ploidy differences.

Based on cross-compatibility data, Smartt and Stalker (1982) and Stalker (1991) concluded that genomic groups have evolved in the genus which mostly follow sectional designations (Am – Ambinervosae, T – Triseminatae, C – Caulorrhizae, EX – Extranervosae, and E – Erectoides, R – Rhizomatosae, and A, B and D – Arachis). The B genome was recently divided into B, F, and K genomes by Seijo et al. (2004) and Robledo and Seijo (2010). Based on rDNA loci and chromosomes with centromeric heterochromatin, Robledo et al. (2009) described three karyotypic subgroups within the A genome and grouped the cultivated peanut with A. duranensis, A. villosa, A. schininii, and A. correntina. Other studies support placing A. hypogaea closely with A. duranensis (Moretzsohn et al., 2004; Milla et al., 2005a; Bravo et al., 2006; Koppolu et al., 2010; Calbrix et al., 2012). The chromosomes of B genome species are karyologically more diverse than those with an A genome (Fernandez and Krapovickas, 1994; Seijo et al., 2004). The B genome species (as opposed to the F and K genomes) does not have centromeric heterochromatin and includes A. ipaensis (the B component of A. hypogaea), A. magna, A. gregoryi, A. valida, and A. williamsii (Seijo et al., 2004; Robledo and Seijo, 2010). The D genome species A. glandulifera is more distantly related to A. hypogaea than other species of section Arachis. Also, molecular analysis has indicated that the aneuploids in section Arachis are more closely related to the B (now classified as the B, F, and K genomes) and D genome species than to A-genome species (Tallury et al., 2005). Evolution is apparently continuing in section Arachis at a rapid pace and multiple translocations have been observed in diploid accessions of A. duranensis (Stalker et al., 1995) and A. batizocoi (Stalker et al., 1991; Guo et al., 2012). At least five different secondary constriction types have been observed in A. hypogaea, which were most likely from translocation events (Stalker and Dalmacio, 1986), and this species is also evolving cytologically.

Analyses of species outside section Arachis have been infrequent. Stalker (1985) reported that the two diploid section Erectoides species A. rigonii × A. paraguariensis hybrids had many univalents and Krapovickas and Gregory (1994) later placed these species in different sections. Intersectional hybrids also were reported by Mallikarjuna (2005) who used in vitro techniques to obtain F₁s.

In addition to morphological and cross-compatibility studies, molecular investigations have been used to better clarify the phylogenetic relationships among peanut species. Most of these investigations have involved species in section Arachis because of their close association with A. hypogaea. Many molecular systems have been utilized, including isozymes (Lu and Pickersgill, 1993; Stalker et al., 1994), seed storage proteins (Singh et al., 1991; Bianchi-Hall et al., 1993; Liang et al., 2006), Restriction Fragment Length Polymorphisms (RFLPs) (Kochert et al., 1991; Paik-Ro et al., 1992), Amplified Fragment Length Polymorphisms (AFLPs) (Milla-Lewis et al., 2005b); Simple Sequence Repeats (SSRs) (Hopkins et al., 1999; He et al., 2005; Hong et al., 2010; Guo et al., 2012; Nagy et al., 2012). Randomly Amplified Polymorphic DNA (RAPDS) (Halward et al., 1992; Lanham et al., 1992; Hilu and Stalker, 1995), and in situ hybridization (Raina and Mukai, 1999; Seijo et al., 2004). All of the studies have indicated that the cultivated peanut has significantly less molecular variation than diploid species, which supports the hypothesis that A. hypogaea originated from a single hybridization event. Additionally, there has been little or no apparent introgression from diploid species to A. hypogaea since its inception (Kochert et al., 1996).

As opposed to the cultivated species, large amounts of molecular variation have been documented among wild species of the genus. Although there have been differences observed among marker systems regarding species relationships, and there remain questions about species positions in sectional groupings (Friend et al., 2010), the molecular data generally fits the sectional relationship model proposed by Krapovickas and Gregory (1994). For example, Hoshino et al. (2006) used microsatellites to evaluate species in the nine peanut sections, and while most species grouped as expected, several species in the Procumbentes grouped with species from section Erectoides, and others clustered into sections Trierectoides and Heteranthae. Galgaro et al. (1998) also indicated that species in section Heteranthae did not group together. Friend et al. (2010) conducted a more comprehensive investigation of Arachis species and found that sections Extranervosae, Triseminatae, and Caulorrhizae each separated into distinct groups based on trnT-trnF sequences; but species in sections Erectoides, Heteranthae, Procumbentes, Rhizomatosae, and Trierectoides formed a major lineage. Species in section Arachis grouped into two major clades, with (i) the B (renamed the B, F, and K genomes), the D genome species, and 18-chromsome aneuploids being in one group and (ii) the A genome species being in the second group.

As compared to lines of A. hypogaea, extremely high levels of resistance have been identified in Arachis species for many important peanut pathogens (Stalker and Moss, 1987; Dwivedi et al., 2007) (Table 2). Mehan et al. (1986) identified four Arachis species that are resistant to aflatoxin production and Xue et al. (2004) found pre-harvest aflatoxin resistance in A. duranensis. Subrahmanyam et al. (2001) found 12 Arachis species accessions to be immune to groundnut rosette virus as opposed to none in the cultivated species. Arachis diogoi was the only species identified with no infection to peanut bud necrosis virus (Subrahmanyam et al., 1985b); this species is also the only one with immunity to tomato spotted wilt virus (Lyerly et al., 2002). None of 7,000 cultivated lines screened for peanut clump virus (PCV) had useful resistance whereas four Arachis accessions of A. kuhlmannii, A. duranensis, and A. ipaensis were immune (Dwivedi et al., 2007). ICRISAT scientists also have evaluated Arachis species for late and early leaf spots and they identified highly resistant materials (Dwivedi et al., 2007). Many Arachis species also have been evaluated for insect pests and extremely high levels of resistance observed as compared to the cultivated peanut (Table 3).

Because the domesticated peanut is an allotetraploid with two genomes and the species being utilized for introgression are diploids, sterility barriers result from ploidy differences and genomic incompatibilities between the species. Traits of interest from Arachis species have been difficult to follow in progenies of interspecific hybrids and disease and insect resistances in 40-chromosome progenies have been lost because of low population sizes and inadequate methods for selection in single plants. Utilizing molecular markers associated with traits of interest may help overcome many of these problems, but unfortunately, few molecular markers have been available to enhance selection efficiency. Molecular research to date indicates that introgression from Arachis species to A. hypogaea appears to be in large blocks (Garcia et al., 1995; Nagy et al., 2010) rather than as single genes or small chromosome segments. Thus, linkage drag of undesirable traits can restrict the use of genetic resources, and molecular markers also have great potential for selection against these characters.

The first peanut cultivars released from interspecific hybridization were from a cross between A. hypogaea and the second tetraploid species in section Arachis (A. monticola Krapov. & Rigoni). Biologically, A. monticola could be considered a weedy subspecies of A. hypogaea. Spancross was released by Hammons (1970) and Tamnut 74 was later released by Simpson and Smith (1975). Neither of these cultivars had phenotypic traits that could be identified as being derived from the wild species, which is not surprising because A. monticola has most of the same disease and insect problems as found in A. hypogaea.

Several methods have been utilized to create populations of fertile A. hypogaea interspecific hybrids and to restore plants to the tetraploid level. First, crosses can be made by hybridizing A. hypogaea with diploids to produce triploid (3x  =  30) F₁s, after which cuttings can be colchicine-treated to restore fertility at the hexaploid (6x  =  60) level. Many triploids will also produce a few seeds through fusion of unreduced gametes, especially if they are left in the field or grown in the greenhouse for long periods of time (Singsit and Ozias-Akins, 1992). Backcrossing the hexaploids with A. hypogaea results in pentaploids (5x  =  50) that are usually vigorous, but only partially fertile. Additionally, they produce few flowers and are difficult to use in crossing programs; but they sometimes yield a few seeds when selfed and the ploidy level stabilizes in progenies at the tetraploid level. A major problem with this scheme has been the few seeds produced at the hexaploid and pentaploid levels, and the lack of selection methods for traits of interest during the semi-sterile generations has resulted in many hundreds of tetraploid lines without traits of interest for crop improvement. To date, no useful germplasm has resulted from backcrossing hexaploids with A. hypogaea. Although backcrossing hexaploids with diploids will theoretically drop the chromosome number to the tetraploid level in one generation, these 6x × 2x crosses (or reciprocals) have not produced viable progenies.

An alternative method to backcrossing hexaploids with the cultivated species is to allow 6x plants to self-pollinate and, by selecting fertile progenies, a few plants may spontaneously lose chromosomes and stabilize at the 40-chromosome level. The loss of chromosomes appears to be infrequent and random, but the advantage of this procedure is associating chromosomes in different species at a high ploidy level which can increase the frequency of recombination. For example, A. hypogaea × A. cardenasii hexaploids were selfed for five generations after which they produced 40-chromosome progenies that were highly variable for seed size, color and other morphological traits (Company et al., 1982). Garcia et al. (1995) analyzed introgression from A. cardenasii to A. hypogaea with RFLPs and found wild species-specific markers on 10 of 11 linkage groups on the diploid RFLP map developed by Halward et al. (1993). Most of the introgression (88%) was apparently in the A genome of A. hypogaea, with the remaining 12% in the B genome. Germplasm lines have been released from this cross with resistance to early leaf spot, nematodes, and several insect pests (Stalker et al., 2002a, b; Stalker and Lynch, 2002; Isleib et al., 2006). The cultivar Bailey was released after utilizing these lines as sources of multiple disease resistances (Isleib et al., 2010).

A second method to introgress germplasm from diploid species to A. hypogaea is to first double the chromosome number of the diploid species to the tetraploid level. This method has the advantage of avoiding several generations of mostly sterile hybrids and recovering tetraploids and is much faster than by going through the triploid – hexaploid procedure; but autotetraploids generally have low vigor, and when annual species are used as parents, they are short lived. Ideally, A and B genome species would be hybridized at the diploid level and then the chromosomes doubled to produce AABB genome allopolyploids to be crossed with the cultivated species. However, chromosome doubling of the sterile AB genome diploids can be highly problematic. Examples of success with this methodology are TXAg-6 and TX Ag-7 (Simpson et al., 1993) which originated from the complex hybrid [A. batizocoi (B genome) × (A. cardenasii (A genome) × A. diogoi (A genome))]^4x. TxAG-6 had very good nematode resistance, but also significant linkage drag which resulted in low yields and poor seed and pod quality. RFLP markers linked to the nematode gene conferring resistance were used to select favorable genotypes (Church et al., 2000). The nematode-resistant cultivars COAN (Simpson and Starr, 2001) and NemaTAM (Simpson et al., 2003) were released after introgressing genes from TXAg-6 to A. hypogaea. By using SSR markers, Nagy et al. (2010) showed that recombination was greatly reduced in the chromosome region where the nematode-resistance gene is located because a large introgressed segment from the wild species comprised a third to half of a chromosome. This germplasm was used in the development of the nematode-resistant cultivar Tifguard (Holbrook et al., 2008), which also has good resistance to tomato spotted wilt virus.

Tracking introgression from Arachis species to cultivated peanut would greatly facilitate selection of desirable progenies in advanced generations of interspecific hybrids. Molecular markers offer a method for following introgression from Arachis species to A. hypogaea. This is important because recombination between the cultivated genomes and those of other species is rare, thus restricting selection for desired traits in interspecific hybrid derivatives (Holbrook and Stalker, 2003). Guimarães et al. (2010, 2011) identified eight genes in A. stenosperma roots that correspond with resistance to M. arenaria, and QTLs for resistance to late leaf spot also have been identified (Leal-Bertioli et al., 2009) in section Arachis species.

For tracking introgressed segments from wild accessions into cultivated germplasm, markers that allow more agile assays would be ideal. A different type of marker, transposon-based, is being evaluated for their use for introgressed segments. For example, Shirasawa et al. (2012b) developed 535 markers derived from transposon-enriched genomic libraries. These markers show great potential and detect higher polymorphism levels than genomic microsatellite markers (Shirasawa et al., 2012b). Also, Single Nucleotide Polymorphism (SNP) markers constitute the most abundant molecular markers in the genome and can be carried out with high throughput genotyping methods. SNP markers have been widely used in many plant species. However, they have had limited use in peanut due to the difficulties of their implementation in polyploid plants, and in Arachis it will require separation of A and B-genome sequences. A SNP-based map of diploid Arachis was developed by Nagy et al. (2012) wherein a high-density genetic map of the A genome was developed from an intra-species cross within A. duranensis, and 598 SSRs, 37 single-stranded DNA conformation polymorphism (SSCP) markers, and 1,054 SNPs were mapped. SNP-based markers have not yet been extensively used on the tetraploid Arachis, but they are expected to greatly accelerate genetic mapping and marker-assisted selection when genotyping-by-sequencing (Elshire et al., 2011) can be routinely implemented.

The narrow genetic base of peanut makes the development of molecular markers a difficult task. The nature and the possibility of using molecular markers have been evolving as have all aspects of the agronomical sciences. The first molecular marker studies used for peanut were based on isozymes and proteins (Grieshammer and Wynne, 1990; Krishna and Mitra, 1988; Lu and Pickersgill, 1993), followed by Restriction Fragment Length Polymorphism - RFLPs (Kochert et al., 1991; Kochert et al., 1996; Paik-Ro et al., 1992), Random Amplified Polymorphic DNA - RAPDs (Halward et al., 1991; Halward et al., 1992; Hilu and Stalker, 1995; Subramanian et al., 2000; Dwivedi et al., 2001; Raina et al., 2001) and Amplified Fragment Length Polymorphism – AFLPs (He and Prakash, 1997; Gimenes et al., 2000; Gimenes et al., 2002; Herselman, 2003; Milla et al., 2005a, 2005b; Tallury et al., 2005).

In recent years, microsatellite or Simple Sequence Repeat (SSR) markers have become the assay of choice for genetic studies in Arachis since they are multiallelic, co-dominant, transferable among related species, PCR-based markers, and usable in tetraploid genomes. Efforts by several research groups to develop microsatellite markers for peanut have resulted in more than 5,000 SSRs (see Pandey et al., 2011, 2012 for reviews). This large effort to produce and characterize SSRs has enabled the phylogenetic evaluation of the genus Arachis (Krishna et al., 2004; Barkley et al., 2007; Tang et al., 2007; Varshney et al., 2009b, Moretzsohn et al., 2013) and production of moderately dense genetic maps for cultivated peanut, which will be described below.

To date, the number of genes associated with molecular markers in peanut is relatively small, but the large number of molecular markers becoming available has great potential for utilization in crop improvement programs. Bertioli et al. (2003) described numerous linkages of resistance genes in peanut. Pandey et al. (2012) listed QTLs for some of the important traits found in the cultivated peanut. Chu et al. (2011) outlined a breeding scheme to utilize marker-assisted selection to pyramid nematode resistance and the high oleic acid trait in peanut cultivars, and the system has greatly increased efficiency for developing breeding lines.

By the use of molecular markers to map genetically structured populations, a number of Quantitative Trait Loci (QTLs) have been identified and some are already used for tracking wild segments introgressed into A. hypogaea.

The first markers for an agronomically useful, wild species-derived trait in peanut were for resistance to root-knot nematode (M. arenaria) from A. cardenasii. Two closely-linked sequence characterized amplified region (SCAR) markers were identified for genes for reduced galling and egg number (Garcia et al., 1996). Simultaneously, three RAPD markers were associated with nematode resistance in several backcross breeding populations derived from the interspecific hybrid TxAG-6 [A. batizocoi × (A. cardenasii × A. diogoi)]^4x (Burow et al., 1996). Marker-assisted selection then was used to develop NemaTAM, a high yielding, nematode resistant cultivar (Simpson et al., 2003). In this case, it was demonstrated that use of markers was more efficient than phenotypic selection because plants selected with markers for the homozygous resistance gene have fewer escapes compared to plants from phenotypic selection. Marker-assisted selection and an accelerated backcross breeding program were also used in development of a high-oleic variety with nematode and tomato spotted wilt virus resistances, called high O/L Tifguard (Chu et al. 2011). The effectiveness of selection of nematode resistance has been the most successful use of MAS in peanut to date. However, in spite of the success of this work, it is now thought that the use of a single gene trait that confers near immunity may be subject to breakdown of resistance under high selection pressure, and is cause for concern. Therefore, new sources of resistance for nematodes, such as amphidiploids derived from A. stenosperma, which is highly resistant to fungi and nematodes (Leal-Bertioli et al., 2010; Proite et al., 2008; Santos et al., 2011; Singsit et al., 1995), would be useful for peanut breeding.

Resistance to late leaf spot (LLS) (Cercosporidium personatum) has multiple components, including percent defoliation, incubation period, latency period, lesion number and diameter, sporulation, and pod yield (Green and Wynne, 1986; Chiteka et al., 1988a, 1988b; Anderson et al., 1993; Waliyar et al., 1993, 1995). High levels of resistance have also been associated with low yield suggesting linkage or pleotropic effects (Iroume and Knauft, 1987), thus breeding for high yielding cultivars with resistance requires this linkage to be broken.

Stalker and Mozingo (2001) identified three RAPD markers associated with early leaf spot lesion diameter in a peanut population derived from a cross between an A. hypogaea × A. cardenasii introgression line with NC 7. Two breeding lines developed from this material have been placed into advanced line trials.

Mapping of RFLP markers on BC₃F₁ lines in greenhouse studies identified five markers for leaf spot resistance (Burow et al., 2008), including three QTLs for incubation period, and one each for latency period, lesion number and diameter. Those QTLs for latency period and lesion number were overlapping, suggesting either linkage between the two or a QTL with pleiotropic effects.

Leal-Bertioli et al. (2009) reported the mapping of 34 resistance gene analogues (RGAs) and five QTLs for LLS resistance (% diseased leaf area) on detached leaves of the F₂ plants of the A-genome mapping population derived from A. duranensis × A. stenosperma, and suggested additive or partial dominance gene action. One QTL explained almost half of the phenotypic variance observed and some QTLs mapped near RGA markers. In another QTL study based on cultivated genotypes, with GPBD-4 as one parent, Khedikar et al. (2010) reported 11 QTLs for LLS; each QTL explained 2 to 7% of phenotypic variation in three environments, suggesting that the genes controlling LLS resistance in this cross are relatively minor. In maps from two other populations, again using GPBD-4 and a larger number (188 and 181) of markers and six trials, a major QTL for LLS was reported, which explained from 10% to 62% of phenotypic variance, depending on the environment (Sujay et al., 2011). These are being used for marker assisted selection breeding at ICRISAT in India (Varshney, 2012, personal communication).

QTL analysis using a partial genetic map of a mapping population with 67 marker loci derived from the cross TAG 24 × GPBD-4 and multiple season phenotyping data on both rust and LLS detected 12 QTLs explaining between 1.70 and 55.20% of the phenotypic variation for each disease, respectively (Khedikar et al., 2010). The SSR marker tightly linked to the major QTL (IPAHM103; QTLrust01), was then validated among a diverse set of genotypes as well as another mapping population (Sarvamangala et al., 2011) derived from the cross TG 26 × GPBD-4. The SSR marker (IPAHM 103) was deployed to introgress the rust resistance QTL into three elite groundnut varieties (ICGV 91114, JL 24 and TAG 24) using the donor GPBD-4 through marker-assisted backcrossing. GPBD-4 was a common parent in all of these crosses, and since it was derived from the wild species A. cardenasii, resistance incorporated into cultivars was likely derived from wild species germplasm.

The aphid-transmitted groundnut rosette virus is an important pathogen of peanut in Africa and Asia, causing severe stunting and loss of yield. Herselman et al. (2004) tested 308 AFLP primer combinations, and they were able to devise five linkage groups consisting of 12 markers with one marker linked to aphid resistance.

Tomato spotted wilt virus is transmitted by tobacco thrips (Frankliniella spp.) and causes serious losses in the U.S.A. segregating population of F₂ plants of the A-genome diploid cross A. kuhlmannii × A. diogoi was screened for resistance to TSWV, and five linked AFLP markers on one chromosome were associated with resistance at a high statistical threshold (Milla et al., 2004; Milla, 2003). In Brazil, interspecific populations and wild species have also been found as promising for introgression of resistance to the thrips, Enne-othrips flavens (Janini et al., 2010).

Mapping of RFLP markers on BC₃F₂ lines of the cross A. hypogaea × A. cardenasii (Burow et al., 2011) identified 29 markers for the domestication-related traits of main stem length, number of lateral branches, pod size, and seed size. Foncéka et al. (2012) produced populations derived from crosses of cultivar Fleur 11 and an amphidiploid (A. ipaensis × A. duranensis)^4x (Fávero et al., 2006) to investigate introgression of wild segments. A population composed of a mixture of BC₃F₁ and BC₂F₂ individuals, each self-pollinated to produce BC₃F₂ and BC₂F₃ families, was used for phenotyping and QTL detection. Domestication-related trait QTLs were found, including ones associated with days to flowering, plant architecture, pod and seed morphology, and yield components. Perhaps surprisingly, it was shown that wild alleles contributed positive variation to several agronomic traits such as flowering precocity, seed and pod number per plant, and length, size and maturity of pods. Moreover, the comparison of QTLs obtained under well-watered and water limited conditions revealed that QTLs for stress tolerance indices for pod and seed numbers with favorable alleles could be attributed to the wild parents. These could be involved in reproductive trade-offs between producing large seeds versus producing more, but smaller, seeds under water stress.

One of the main uses of molecular markers in peanut has been the construction of genetic linkage maps, which are used to study the genome structure and organization, identify regions of particular interest (e.g., resistance to diseases) and for marker-assisted selection in breeding programs. Due to the very low genetic variation in cultivated peanut, wild species were initially used for the construction of linkage maps in Arachis because of their simple genetic structure and higher polymorphism levels. The first map used RFLPs to analyze variation between the diploid species A. stenosperma and A. cardenasii where a total of 117 RFLP markers were mapped into 11 linkage groups (Halward et al., 1993). An AFLP-based A-genome map was generated from an F₂ population developed from the cross A. kuhlmannii × A. diogoi (Milla, 2003), and a RAPD-based map of A. stenosperma × A. cardenasii also was developed by Garcia et al. (2005).

It was only with the large abundance of SSR markers that moderately saturated maps were created. The first SSR-based map was constructed for an F₂ population derived from a cross of two A-genome diploid species (A. duranensis and A. stenosperma) where 170 loci mapped into 11 linkage groups covering 1231 cM of total map distance (Moretzsohn et al., 2005). Additional markers were subsequently included in the diploid A genome map, resulting in 369 markers (188 SSRs, 80 anchor markers, and 35 resistance gene analogues, among others), mapped into 10 linkage groups, which correspond to the 10 chromosomes of the haploid genome (Leal-Bertioli et al., 2009). Nagy et al. (2012) published a more saturated map having 1724 SNP, SSR, and single stranded DNA conformation polymorphism (SSCP) makers from a cross between two A. duranensis accessions.

A diploid F₂ population derived from the cross A. ipaënsis × A. magna was used to develop a map for the B genome of Arachis, with 149 codominant markers (mostly microsatellites) mapped into 10 linkage groups covering of 1294.4 cM (Moretzsohn et al., 2009). Fifty-one common markers presented evidence of the high synteny of the B and the A genomes. Guo et al. (2012) created a more dense map by crossing two accessions of A. batizocoi. They compared high density A and B genome maps and observed a large amount of synteny between the A and B genomes, but also several inversions and translocations.

The first tetraploid map also was created with RFLPs by using progenies of a cross between the cultivar Florunner × the synthetic amphidiploid TxAG-6 [A. batizocoi × (A. cardenasii × A. diogoi)]^4x where 383 markers were mapped (Burow et al., 2001). A genetic map was created based on a population of 88 BC₁F₁ individuals of a cross of a synthetic amphidiploid (A. ipaënsis × A. duranensis) by Fávero et al. (2006) with A. hypogaea cv. Fleur11. This SSR-based linkage map for the tetraploid genome was composed of 298 markers and 21 linkage groups, spanning 1843.7 cM (Fonceka et al., 2009). A comparative analysis of this map with the A genome map of Moretzsohn et al. (2005) suggested the occurrence of a chromosomal translocation event prior to the peanut's tetraploidization (Fonceka et al., 2009).

The partial first linkage map from a cross between accessions of A. hypogaea was constructed using an F₂ population (Herselman et al., 2004), in which five linkage groups with 12 markers spanned 139 cM of the genome. More complete maps were later reported, with the first being in 2008 by using 142 individuals of a recombinant inbred line (RIL) population derived from a cross between one accession of A. hypogaea subsp. hypogaea and one accession of the fastigiata subspecies (Hong et al., 2008). New markers were added to this map and two additional maps were constructed based on RIL populations having accessions of the subspecies fastigiata as parents (Hong et al., 2010). A reference map was developed, with 175 loci and 22 linkage groups, covering a total distance of 885.4 cM. The marker order was in general collinear to the A genome map (Moretzsohn et al., 2005). Another intraspecific map for peanut was developed using a RIL population composed of 318 F₈/F₉ plants (Varshney et al., 2009a); and of the 1145 microsatellite markers screened, only 135 mapped in 22 linkage groups spanning 1270.5 cM. More recently, Wang et al. (2012) published a SSR map based on an F₂ population of 94 individuals derived from an A. hypogaea subsp. hypogaea × A. hypogaea subsp. fastigiata hybrid. This map consisted of 385 polymorphic SSRs covering 318 loci and 21 linkage groups that spanned 1674.4 cM. To date, there are 23 maps of peanut published using RFLPs, AFLPs, SSRs, SNP, SCAR and CAPS markers which are summarized in Pandey et al. (2012). A recent consensus map of A. hypogaea map comprised of 897 marker loci was constructed by Gautami et al. (2012).

These maps are very useful for breeding because they incorporate QTLs for agronomically important traits, such as disease resistance and drought related traits. They also have been used to develop markers closely associated with a nematode resistance gene (Nagy et al., 2010). Further, as the cultivated genome is sequenced, the diploid maps will be of great importance for sorting out the chromosome and genetic organization of the tetraploid genome.

In general, it is possible to map only limited numbers of molecular markers in a given mapping population due to polymorphism constraints, and this is especially true for peanuts. In order to improve the knowledge on the genetic and genomic structure, assignment of marker and QTL positions, genetic maps developed for different mapping populations can be used for creating consensus maps, which can be also be anchored to other species maps. Consensus maps allow investigators to (i) map a large number of marker loci onto a single map, (ii) determine relative position of common markers across different mapping populations, (iii) determine stability of marker locus position across the genomes, and (iv) provide evidence for chromosomal rearrangements, gene duplication and assists in the assignment of linkage groups to chromosomes (Gautami et al., 2012).

With the present availability of markers, the international Arachis community has been striving towards developing a consensus genetic map. By integration of the eleven linkage maps of the artificial amphidiploids and cultivated tetraploid peanuts, a reference consensus map consisting of 897 marker loci has been constructed and given the consensus nomenclatures for the linkage groups (a01 to a10 and b01 to b10) (Gautami et al., 2012).

Subsequently, Shirasawa et al. (2012a) created the most comprehensive consensus map to date. They mapped a large number of DNA markers from two cultivated peanut maps (Shirasawa et al., 2012b) onto the previously reported linkage maps of the A and B genomes (Moretzsohn et al., 2005) and the artificial amphidiploid map (Fonceka et al., 2009), and integrated these with nine maps of cultivated peanut (Gautami et al., 2012; Hong et al., 2010; Khedikar et al., 2010; Sarvamangala et al., 2011; Qin et al., 2012; Ravi et al., 2011; Sujay et al., 2011; Varshney et al., 2009b; Shirasawa, 2012a, 2012b). This consensus map was then subjected to comparative analysis with four legume genomes (Cajanus cajan, Glycine max, Lotus japonicus, and Medicago truncatula) to clarify the feature of genome structure in the genus Arachis. The comparison of diploid and tetraploid maps revealed probable genome rearrangements within the Arachis genomes. Bertioli et al. (2009) used sequence characterized markers, and a high proportion of low or single copy gene markers to anchor the AA-diploid Arachis map onto the fully sequenced genomes of L. japonicus and M. truncatula (Sato et al., 2008); www.medicago.org). This alignment was represented as “genome plots” (Bertioli et al., 2009). Inspection of these plots shows surprising degrees of synteny considering the time of species divergence (estimated 55 million years). Although there are some regions of double affinities between Arachis and these model legumes, most synteny blocks have a single main affinity and not multiple affinities. Genome evolution (e.g., chromosomal translocations and inversions) progressively breaks down syntenic relationships between species over evolutionary time.

The comparative genomics of the legumes revealed that the Arachis genome structure was moderately divergent from those of the compared legumes. This information will be useful for selecting highly informative and uniformly distributed markers for developing new genetic maps, background selection and diversity analysis, aligning new genetic and physical maps, performing QTL analysis in a multi-populations design, evaluating the genetic background effect on QTL expression, and serving other genetic and molecular breeding activities in peanut.

The chromosomes of A. hypogaea are highly diplodized which indicates that there may be little recombination between the A and B genomes except when the infrequent quadrivalent is formed. To support this, genomic comparison of A. duranensis and A. ipaensis indicates that there is significant divergence in repetitive DNA; however, the repetitve DNA in the tetraploid species has not significantly recombined since A. hypogaea evolved (Araujo et al., 2012). Thus, genomic investigations of the progenitor diploid species should have direct application to the cultivated peanut. For example, sequencing the diploid genome of either A. duranensis or A. ipaensis could serve as a template for sequencing the genomes of the tetraploid species. This is important because developing a high quality diploid genetic sequence will be significantly less expensive than developing an entire sequence for the tetraploid genomes and trying to sort out the genetic duplications between the A and B genomes. Using a diploid progenitor species also will lead to fewer assembly errors. In addition, utilizing A. duranensis or A. ipaensis for sequencing templates will solve problems associated with chimeric contigs in de-novo assemblies of the tetraploid genomes.

The incorporation of wild alleles into crops is a proven strategy for the development of improved varieties with pest and disease resistance, tolerance to abiotic stress, improved yield and quality, and even male fertility and fertility restoration. However, the extent of utilization of the useful allele reservoir in wild species and its impact on peanut breeding has been limited, mainly due to the restrictions of the plant itself in terms of crossability, multiplication rate, and, until recently, to the lack of appropriate molecular tools to analyze and follow traits in hybrids. Arachis cardenasii has been one of the most useful sources of genes from wild species to date, but crosses involving other species also have been used. The recent use of the two most probable ancestors of peanut A. duranensis and A. ipaënsis in a systematic introgression program opens the way for extensive and detailed characterization of the peanut genome and wild allele interactions for a wide range of traits. As new materials are being created and genotyping strategies are becoming more advanced, variability from the wild species is being harnessed to the benefit of world agriculture.