Introduction

The cultivated peanut (Arachis hypogaea L.) originated from hybridization between diploid female species A. duranensis with the A genome and A. ipaensis with the B genome (Kochert et al. 1996). A single allotetraploidization event involved either chromosome doubling in diploid F1 (Favero et al. 2006) or unreduced gametes in F1 (Harlan and deWet 1975). Occurrence of diploid parental and wild allotetraploid species (A. monticola) indicates that the eastern foothills of the Andes in the region of northern Argentina and southern Bolivia could be the area of origin of A. hypogaea (Krapovickas and Rigoni 1957; Kochert et al. 1996; Ravi et al. 2010). Allotetraploid Arachis species were then put under selection for domestication (Kochert et al. 1996), and the domesticated peanut dispersed to South and Central America where both natural and artificial selection produced many landraces of domesticated peanut.

On the basis of several morphological differences, two subspecies fastigiata and hypogaea have been recognized in the cultivated peanut. Subspecies hypogaea is characterized by alternate branching, absence of flowers on the main axis, long duration (120–160 days), and presence of seed dormancy, whereas fastigiata is recognized by sequential branching, presence of flowers on the main axis, short duration (85–130 days), and lack of seed dormancy. On the basis of morphological differences, the two subspecies are subdivided into botanical varieties. Subspecies hypogaea is divided into var. hypogaea (Virginia bunch/runner) and var. hirsuta (Peruvian runner), whereas spp. fastigiata is classified into var. fastigiata (Valencia), var. vulgaris (Spanish bunch), var. aequatoriana, and var. peruviana (Krapovickas 1969; Krapovickas 1973). Subspecies hypogaea is thought to have originated by a mutation from within subspecies fastigiata (Singh 1988).

The morphological, physiological, and agronomic traits of cultivated peanuts are extremely variable. However, limited variation is found at the DNA level (Singh et al. 1998; Varshney et al. 2009; Khedikar et al. 2010) owing to the occurrence of a single polyploidization event followed by reproductive isolation from related diploid species (Young et al. 1996), self-pollination (Halward et al. 1991), and use of limited exotic germplasm in breeding programs (Knauft and Gorbet 1989). Hence it seems that most of the DNA-level variation among the cultivated types arose from various mutations (Perry 1968; Mouli et al. 1979; Prasad et al. 1984; Prasad 1989) rather than recombination. In a previous study, we hypothesized the important role of mutations possibly involving the activation of cryptic transposable elements (TEs) in the origin and evolutionary differentiation of the peanut (Gowda et al. 1996).

Miniature inverted repeat transposable elements (MITEs) are the predominant TEs among plant genomes (Wessler et al. 1995; Shan et al. 2005; Naito et al. 2006). Transposition preference for low copy genic regions emphasises the role of MITEs in modulating gene expression (Wessler 1998; Zhang et al. 2000; Wessler 2001) and aiding crop evolution (Shan et al. 2005; Naito et al. 2006). Previously, we reported an active peanut MITE (AhMITE1), the transposition of which was associated with the high-frequency origin of late leaf spot (LLS)-resistant mutants (Gowda et al. 2010). Here we have made an effort to study stress-induced activation of AhMITE1, and the relevance of its transposition with the mutational and evolutionary origin of botanical types.

Materials and methods

Plant material

Genotypes tested for AhMITE1 transposition included parents and mutants generated by various stresses (Table 1), and thirty-three South American landraces representing different botanical types (Table 2) and four accessions of wild allotetraploid (A. monticola) with the AB genome combination (Table 3). Parental wild types and their mutants developed via mutagenesis with EMS and gamma irradiation, and those recovered from spontaneous mutation were selected from our mutant collection. These stabilized mutants were phenotypically studied and characterized for morphological changes (Motagi et al. 1996). TAG 24 and its mutant (TAG 24 (M)) recovered through tissue culture as somaclonal variant were obtained from Bhabha Atomic Research Centre (BARC), Mumbai, India. South American landraces and accessions of A. monticola were obtained from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Hyderabad, India.

Table 1 Activation of AhMITE1 in peanut mutants
Full size table
Table 2 Activity of AhMITE1 at FST1-linked site among South American peanut landraces
Full size table
Table 3 Activity of AhMITE1 at FST1-linked site among the primitive allotetraploids of peanut
Full size table

Detection of AhMITE1 insertion/excision

Genomic location of the AhMITE1 was determined (Bhat et al. 2008) by recovering its 5′-flanking sequence tag from TMV 2, a Spanish bunch cultivar, using TAIL-PCR with a set of degenerate primers and nested primers designed based on the reported MITE sequence (Patel et al. 2004). Presence or absence of AhMITE1 at this pre-determined site (FST1-linked site) in various genotypes/mutants was checked by developing an AhMITE1-specifc PCR (Gowda et al. 2010). The forward primer (5′ GGGAGAAGAAAAGGATGAGA 3′) was designed on the basis of the AhMITE1 flanking sequence tag (FST1) recovered from the TMV 2 genotype of peanut (Bhat et al. 2008) whereas the reverse primer (5′ TCTCATGAAGATGCTTTGGT 3′) was specific to AhMITE1 (Patel et al. 2004). Amplification of a 242-basepair (bp) product would indicate the presence of AhMITE1 at the FST1-linked site in the genome.

For AhMITE1-specifc PCR, genomic DNA of various genotypes and mutants was isolated from young leaves collected at the 3–4 leaf stage of the seedling using the GenEluteTM plant genomic DNA miniprep kit (Sigma–Aldrich). A final volume of 20 μl containing 100 ng genomic DNA, 1× PCR buffer, 0.5 mM dNTPs, 10 pmol of each primer, and one unit of Taq DNA polymerase (Genei, Bangalore, India) was used for PCR. The amplification reaction was carried out in a Mastercycler® (Eppendorf, Germany) by setting the conditions for one cycle of pre-denaturation (94°C for 5 min), 35 cycles of denaturation (94°C for 1 min), annealing (60°C for 1 min), and extension (72°C for 1 min). One cycle of final extension (72°C for 1 min) was included before the PCR product was stored at 4°C until further use. Presence of the PCR product was checked on 1.2% agarose gel by electrophoresis.

Results

EMS and gamma mutagens and tissue culture could activate the transposition of AhMITE1. The parent Dharwad Early Runner (DER) did not amplify the 242-bp product (Fig. 1), hence was found to be free of AhMITE1 insertion at the FST1-linked site. Mutants generated by gamma irradiation (SB 2, SB 4, SB 5, and SB 6) and EMS treatment (SB 10 and SB 11) amplified the 242-bp product with AhMITE1-specific PCR, indicating the insertion of AhMITE1 into the FST1-linked site. These mutants resembled Spanish bunch type in their erect growth habit, lack of seed dormancy, and increased pod size compared with the DER type (Table 1). Another EMS mutagenesis effort with VL 1, a Valencia mutant from DER, could generate three Spanish bunch mutants (28-2, 45, and 110). Compared with VL 1, these mutants had more primary and secondary branches, short main stem and primary branches, small leaves, high pod yield and test weight, and resistance to late leaf spot. They showed insertion of AhMITE1 into the FST1-linked site. Contrastingly, EMS mutagenesis could also activate excision of AhMITE1 from the FST1-linked site as observed with the narrow leaf mutant (NLM) from TMV 2. Origin of NLM, a Virginia bunch type mutant (with alternate branching and main stem flowering) belonging to spp. hypogaea from TMV 2, a Spanish bunch cultivar, involved excision of AhMITE1 from the pre-determined site, indicating a gross morphological change.

Fig. 1
figure 1

Mutagen-induced transpositional activity of AhMITE1 in some of the mutants and their parents. Amplification of the 242-bp product upon AhMITE1-specific PCR indicates the presence of the element at the FST1-linked site (M: 100 bp DNA ladder)

Full size image

Two spontaneous mutants (VR 1 and VR 2) resembling Virginia runner were isolated from the Spanish bunch variety JL 24. These mutations represented the morphological changes from spp. fastigiata to spp. hypogaea, and accompanied the excision of AhMITE1. Likewise, 28-2(s), 45(s), and 110(s) obtained spontaneously from 28-2, 45, and 110, respectively, were also associated with excision of the transposable element. Significant morphological changes, for example susceptibility to late leaf spot and pod shape and size, were observed in these mutants. Tissue culture induced somaclone, TAG 24 (M), from TAG 24 indicated excision of the element. This excision resulted in discernible changes, for example increase in the number of branches, decrease in seed size, and prostrate habit. Thus, the origin of Spanish bunch mutants belonging to spp. fastigiata were generally associated with insertion whereas spp. hypogaea mutants were associated with excision of the TE from the FST1 site.

This association was further validated by checking South American landraces comprising native subspecies and botanical types for the presence of AhMITE1. Of the nine genotypes belonging to spp. hypogaea, only one (ICG 10479) had AhMITE1 at the FST1-linked site. In the spp. fastigiata, four of the eleven Valencia, seven of nine in the Spanish bunch, and one of two peruviana (ICG 11088) had the AhMITE1 insertion, and it was absent from the two aequatoriana genotypes tested. Proportions of the genotypes within the two subspecies carrying AhMITE1 at the FST1-linked site were compared statistically using the z test (standard normal deviate test for proportion) (Nageswara Rao 2007). Because the calculated z value (2.06) was significantly higher than the critical value of 1.96 at the 5% level of significance, it was concluded that most of the landraces belonging to spp. fastigiata types had AhMITE1 at the FST1-linked site, but spp. hypogaea types lacked it.

Four accessions of the allotetraploid A. monticola (AABB) and a cultivar (TMV 2) were screened for AhMITE1 at the FST1-linked site to trace the possible time of transposition during the course of evolution and peanut domestication. It was found that all the accessions of A. monticola were devoid of the element at the site, indicating that insertion occurred late in the evolutionary differentiation of A. hypogaea into subspecies and botanical types.

Discussion

Among plants, allopolyploidy is common (Stebbins 1951; Masterson 1994; Leitch and Bennett 1997), and regarded as “dynamic” (Soltis and Soltis 1995) and a major force in evolution compared with diploidy (Liu and Wendel 2003). It is known to be of major importance in genome diversification and adaptation (Cronn and Wendel 2004). However, the underlying genetic and molecular basis of allopolyploidy-mediated evolution remains obscure. Recent studies indicate that epigenetic mechanisms, for example DNA methylation, chromatin remodelling, and RNA-silencing regulate reprogramming of the gene expression and developmental patterns in allopolyploids (Chen 2007). Methylation of cytosine and histone modifications are of major importance in regulating the activity of transposable elements (TE) belonging to both long terminal repeats (LTR) and MITEs (Wessler et al. 1995; Madlung et al. 2005) that are known to induce transposon-mediated variations fuelling adaptation and evolution (McClintock 1984).

Numerous reports record activation of a rice MITE, mPing, by diverse stresses such as gamma rays (Nakazaki et al. 2003), tissue culture (Jiang et al. 2003; Kikuchi et al. 2003), environmental conditions (Jiang et al. 2003), and genomic shock, because of interspecific hybridization (Shan et al. 2005). Enhanced transpositional activity was observed for En-Spm-like transposons of Arabidopsis thaliana after remodelling of CG methylation on allopolyploidization (Madlung et al. 2005). Hence, it seems that the response of a plant to such stresses may involve epigenetic changes leading to transpositional activity (Chen 2007).

In peanut, of the several copies of MITE identified using Southern hybridization, a particular MITE was induced by diethyl sulfonate (DES) to cause high oleate mutation (Patel et al. 2004). We named this AhMITE1, and determined its site of integration in TMV 2, a Spanish cultivar by recovering a 151 bp 5′-flanking sequence tag (FST1) (Bhat et al. 2008). Homology search indicated that a small region (23/24 bp) of the FST1 corresponded to a genomic sequence of A. batizocoi (GenBank acc. no. DX508954) of the methylation filtered library (LibID703) (Bhat et al. 2008). So the genomic location of AhMITE1 as identified by the FST1 was referred to as the FST1-linked site.

Analysis of several spontaneous and induced mutants originating from mutagenesis (EMS and gamma irradiation) and tissue culture revealed AhMITE1 transposition. EMS-induced NLM from TMV 2 involved excision of AhMITE1 from the FST1-linked site. Most interestingly, this mutation involved morphological changes from subspecies fastigiata (TMV 2) type to those resembling hypogaea (NLM), for example sequential branching to alternate branching and main stem flowering to main stem non-flowering. Likewise, many mutations characterized previously (Motagi et al. 1996), were associated with drastic morphological changes resembling alternate subspecies or botanical types, and the presence of AhMITE1 at the FST1-linked site was predominant in spp. fastigiata. As in the previous studies (Prasad et al. 1984; Gowda et al. 1996), the origin of mutants belonging to spp. hypogaea from spp. fastigiata types was more common than the other way round. This could be because of higher chances of excisions of AhMITE1 at the FST1-linked site compared with fresh insertions at that site.

Among the cultivated species of peanut, spp. fastigiata (both Spanish bunch and Valencia) were found to carry an AhMITE1 insertion at the FST1-linked site, whereas genotypes classified under spp. hypogaea rarely contained AhMITE1, indicating a strong association between AhMITE1 transposition and intraspecific differentiation. Such preferential activity of MITEs has been reported in japonica versus indica types of rice (Jiang et al. 2003). The exceptions, for example VL 1, a Valencia type mutant from DER, and some genotypes belonging to different botanical types in spp. fastigiata, which were devoid of the element, indicated a probable role of additional mechanisms. Genotypes belonging to A. hypogaea spp. fastigiata var aequatoriana missed AhMITE1 at the site, similar to many spp. hypogaea types. Previous efforts with AFLP markers also indicated its closer resemblance to spp. hypogaea than to spp. fastigiata (He and Prakash 2001).

Lack of AhMITE1 at the FST1-linked site in four accessions of A. monticola but its presence in some forms of cultivated species indicates de-repression of MITE, possibly because of “genomic shock” triggered by hybridization followed by allopolyploidization (Shan et al. 2005; Naito et al. 2006). As far as we are aware, this is the first report indicating the role of an MITE in peanut genome differentiation that took place after allopolyploidization. This unique association supports the view that the spp. hypogaea is much closer to the wild allotetraploid (Paik-Ro et al. 1992; Singh et al. 1993; He and Prakash 2001), and that AhMITE1 transposition at an FST1-linked site could have been of major importance in the origin of spp. fastigiata.

This study reveals a possible reason for extensive morphological diversity, despite low DNA polymorphism in the cultivated species. MITEs, by virtue of their inherent property, are preferentially inserted into low copy genes and their regulators (Singh et al. 1998), and the transpositions are known to modulate gene expression because of methylation, gene disruption, and frame-shift mutations (Nakazaki et al. 2003; Shan et al. 2005). Such events could also be important precursors generating morphologically distinct subspecies and botanical types. Genomic characterization of the FST1-linked site for candidate gene and future elucidation of other MITEs/LTRs using the transposon display technique (Van den Broeck et al. 1998) in mutant collection, landraces, and wild types will lead to better understanding of peanut genome differentiation and domestication.