Polyploidy means that two or more complete sets of chromosomes of the same (autopolyploid) or different (allopolyploid) genomes are present in the same nucleus. It is a prominent and significant process in plant evolution [1
]. Polyploidy has been considered important in conferring adaptive value to some cultivated species by increasing the allelic diversity, maintaining genome-wide heterozygosity and allowing the emergence of novel phenotypic variation [3
]. The stages of polyploid formation usually include reproductive isolation from the progenitors, resulting in severe genetic bottlenecks. However, as most polyploid species have been formed recurrently from their wild progenitors [7
], a moderate level of polymorphism has been kept in polyploid plants. Peanut (Arachis hypogaea
L.) is an allotetraploid (2n = 4x = 40) native from South America with an AB genome. In contrast to the recurrent formation of several polyploid species, the allopolyploid structure of cultivated peanut is likely derived from a single hybridization between two wild diploid species followed by chromosome doubling [8
]. Consequently, its monophyletic origin and domestication effects have greatly narrowed the genetic basis of the cultigen.
The peanut primary gene pool comprises elite breeding lines and landraces of the cultivated species A. hypogaea
and A. monticola
, a closely related wild tetraploid species of cultivated peanut [9
]. Although a large amount of phenotypic variation is conspicuous in this gene pool, only a limited level of DNA polymorphism between genotypes has been observed [10
]. With the increase of the number of molecular markers, efforts have been invested for developing genetic maps [15
]. QTLs for physiological parameters and yield component traits linked with drought tolerance have recently been reported [17
]. However, there has still been little progress on the integration of molecular markers in intraspecific peanut breeding programmes despite the challenge to obtain new varieties with resistance to diseases and tolerance to abiotic stresses.
The secondary gene pool of cultivated peanut mainly comprises wild diploid species (2n = 2x = 20) and represents an important source of novel alleles that can be used to improve the cultigen. Extensive work has been done to characterize genetic relationships between species of this gene pool and cultivated peanut using molecular markers [18
] and cytogenetics [24
]. Several wild diploid species have been hypothesized as the possible ancestors of the cultivated species. Recent studies have proposed A. duranensis
(A genome) and A. ipaensis
(B genome) as the most probable wild progenitors [24
]. Favero et al.
] produced a synthetic amphidiploid resulting from a cross between A. ipaensis
and A. duranensis
and doubling of the chromosome number. This amphidiploid has produced fertile hybrids when crossed with each of the botanical varieties of A. hypogaea
. Furthermore, resistances to several diseases have been identified in wild species [30
] and QTLs for disease resistance were recently mapped in a cross involving wild diploid species [34
]. Introgression of disease resistance genes from the wild diploid species A. cardenasii
into an elite peanut variety has also been reported [35
]. However, the effective transfer of genes from peanut wild species to cultivated species was reported to be labor intensive [37
] and the introgression of genes involved in the variation of complex traits such as yield has to our knowledge never been reported. Hence, genetic variation existing in wild species remains largely underexploited.
Wild relatives represent an important source of genes that has been successfully tapped to improve productivity and adaptation in various crops [38
]. Tanksley et al.
] have proposed an efficient advanced backcross-QTL (AB-QTL) approach to detect and map valuable QTLs and to simultaneously transfer them from wild to cultivated species. This approach has been widely adopted for mapping and introgressing QTLs involved in complex traits in several species [42
QTL mapping in crosses between crops and their wild progenitors is also a powerful means for identifying genomic regions involved in morphological and physiological changes that distinguish crops from their wild relatives [43
]. These morphological and physiological differences that have resulted from plant evolution under anthropogenic influences have been included in a generic term known as the "domestication syndrome" [44
]. Features of the domestication syndrome have been shared in almost all agronomically important domesticated species. The targets of domestication include the loss of mechanisms for seed dispersal and dormancy, changes in plant growth habits and increases in the size of harvested plant parts [46
]. Although pod dehiscence is absent in peanut, the long peg and isthmus observed solely in wild species have been identified as a potential mechanism for seed dispersal [49
]. In cultivated peanut pods, the isthmus is virtually nonexistent and has given way to a more or less deep pod constriction that may represent a vestige of the isthmus. Cultivated peanut also displays a more compact growth habit compared to wild species. However a large range of variation still exists in cultivated species. Varieties belonging to subspecies fastigiata
are characterized by an erect growth habit accompanied by traits such as loss of pod constriction and of seed dormancy. Prostrate growth habits are generally accompanied by small fruits with marked constriction and seeds demonstrating dormancy. These characters, which could be considered primitive, are mainly found in varieties belonging to the hypogaea
Notwithstanding the identification of the two most probable wild progenitors of cultivated peanut, little is known about peanut evolution under domestication, and the genomic regions associated with domestication have never been reported. We recently published an SSR-based genetic map constructed using a BC1
population derived from a cross between the amphidiploid (A. ipaensis
× A. duranensis
and a cultivated peanut variety, and an analysis of the genome-wide introgression of wild DNA fragments in the BC2
]. As a follow-up to this study, we have produced an AB-QTL population that represents a great opportunity to map QTLs involved in peanut domestication and to explore the reservoir of agronomically interesting alleles remaining in the wild species. As peanut is mainly grown under rainfed conditions in the arid and semi-arid tropics, it often faces moderate to severe drought conditions [52
]. An important breeding objective is thus to develop varieties that can produce suitable yields under water-limited conditions.
In this article, we present a detailed QTL analysis of several traits involved in peanut productivity and adaptation under two water regimes as well as in the domestication syndrome. Based on these results, we report the identification of wild alleles that contribute positive variations to complex traits, we outline several regions of the peanut genome involved in the domestication process and we compare the distribution of QTLs in the subgenomes.