Common bean (Phaseolus vulgaris L.) is the most important food legume in terms of providing directly consumed nutrients and dietary protein in developing countries of Latin America, Africa as well as in traditional diets of the Middle East and the Mediterranean with over 23 M tons grown around the tropics, sub-tropics and temperate zones for anywhere from on-farm to local market, and within-country consumption or exports
. Common bean is usually grown in areas with sufficient rainfall but has also extended to regions where drought is endemic and supplemental irrigation is scarce such as in northeastern Brazil, coastal Peru, the central and northern highlands of Mexico, and in lower elevations of East Africa
 as well as the western plains of the United States and Canada
. Therefore, increasing drought tolerance through common bean breeding has become a common goal of national and international breeding programs.
Relatively few sources of drought tolerance have been identified in common bean compared to other legumes. Most studies have concentrated on advanced lines from a few commercial classes
 and not from the wild and landrace collections which are considerable reservoirs of naturally-adapted genotypes for drought-stress environments. Therefore, searching wild and cultivated collections of common bean is another goal of plant breeding programs and has recently been assigned a high level of funding within the context of the food security programs of the International Agricultural Research Centers. Initial testing has shown high variability for drought tolerance traits and certainly common bean contains a lot of allele richness
Drought tolerance is a genetically, physiological and mechanistic complex trait. In terms of genetics, the multiple individual traits that make up drought tolerance are usually inherited quantitatively with very few major genes for drought tolerance mechanisms known, although Blair et al. did find some quantitative trait loci for drought tolerance. Epigenetic and environmental components of drought stress exist, as well. One transcription factor that is often involved in signaling of drought stress is abscisic acid whose levels are often correlated with plant parts and whole plants that are suffering from drought stress
. Some of the mechanisms of drought tolerance are controlled through an ABA responsive pathway
, while others are independent of ABA
[8, 10]. In particular, the transcription factors of the Asr (abscisic acid, stress, ripening induced) family of genes are plant-specific and stress-regulated components of the ABA- dependent pathway, with further proof of their role found in their interaction with ABRE elements
[11, 12]. Sucrose synthase genes are thought to be downstream Asr genes
. The number of Asr genes found in plant genome databases varies from one in Vitis vitifera, four in Brachypodium distachyon, six in Oryza sativa, and up to seven in Sorghum bicolor. Expression analysis have demonstrated their explicit role in conferring increased drought and salt tolerance in tomato, rice and lilies
[14–16] but to date no analysis of their role in the legumes has been put forward.
In this regards, diversity analysis of the ASR family has been illustrative of adaptive selection of crop plants to help deal with environmental conditions. For example, studies of the extent of nucleotide diversity in Asr genes in Solanaceae species and in wild and cultivated rice provided some evidence of non-neutral evolution, adaptive, or demographic events in dry areas
[14, 17, 18]. Moreover, genetic mapping of Asr1 co-localized this gene with QTLs for xylem sap ABA content, for anthesis–silking interval responsive to mild water deprivation and for leaf senescence in maize
[19, 20]. Finally, Maskin et al. showed a DNA-binding activity, Konrad & Bar-Zvi
 revealed that the unstructured form of tomato ASR1 proteins presented a chaperone-like activity that stabilized other proteins against denaturation caused by heat and freeze–thaw cycles, and Cakir et al. described an association with a grape hexose transporter promoter.
Common bean is a good model to study drought related candidate genes and especially the less complex Asr family because of its rich evolutionary history in the wild across two continents (South and Central America) and multiple domestication process (in the Andean and Mesoamerican regions of the Americas). Wild Phaseolus vulgaris beans are diverse in the western hemisphere ranging from temperate Argentina to dry land parts of Mexico. However, wild beans are thought to have evolved from an original gene pool in Ecuador and northern Peru, after which radiation to various regions north and south of there, gave rise to an Andean, a Colombian, and a Mesoamerican gene pool
. The Andean and Mesoamerican wild beans were then subjected to domestication in each region giving rise to cultivars of both gene pools
Mesoamerican beans were domesticated in the region of Jalisco
, although this does not preclude more than one domestication event in another part of Mesoamerica giving rise to a diversity of chloroplast haplotypes
. For the Andean gene pool, southern Bolivia may have been the center of domestication
 with introgression from the wild occurring in the extension of cultivated types northwards towards the equator. Both domestications occurred 5,000–8,000 years ago
. Therefore in common bean, populations structure is divided into gene pools (Andean and Mesoamerican for cultivated beans and four or more groups in wild beans); while additional structure within each of these gene pools is then found. Within the cultivated Andean gene pool the races Nueva Granada, Peru and Chile are identifiable
[5, 32–35]. Within the cultivated Mesoamerican gene pool the race Mesoamerica, the complex Durango-Jalisco and the race Guatemala are observable
As mentioned above, the cultivated gene pool structure contrasts with the structure obtained for wild common bean in which four main clusters are seen: the Colombo-Mesoamerican, the Mexican, the Andean and the Peruvian-Ecuadorian
. This simplistic model has been further challenged by results from Kwak et al. where wild common beans from Central America were divisible into different groups including those from Guatemala and Costa Rica.
Introgression between gene pools, between the races and between cultivated and wild genotypes has been a historical, long-term and re-iterative process
[34, 37, 38]. However, the structure of the wild populations is maintained by geographic barriers along the length of the Andean to Mesoamerica arc of mountains and varying terrain. In terms of habitat ecology, race Durango-Jalisco is the only one of the groups of races that has significant drought tolerance, with part of this group distributed in semi-arid areas of Mexico
; race Chile has adaptation to relative drier areas as well, but is only found in the southern Andes
[35, 40]. Races Mesoamerica and Guatemala, or Nueva Granada and Peru occupy low to mid altitude or highland regions of Latin America, respectively
[5, 41, 42]. Some highland and mid-elevation sites are also drought susceptible especially in the equatorial regions where bimodal rainfall events are often short. Although cultivars from the Durango-Jalisco complex have the highest level of drought tolerance, one may expect to find still higher levels in certain wild germplasm
Nucleotide diversity surveys are powerful tools for the study of reference collections of cultivated and wild genotypes that allow population genetic tests to be made for departure from the neutral equilibrium models and to identify the diverse selective modes that shaped the evolution of specific genes
. For instance, sequence information for specific genes can show their genealogy and suggest how different groups of accessions evolved from the wild and how this influenced the genotypes involved in domestication events. Therefore, the particular role of duplication, lineage sorting, sub-functionalization, sub-speciation and ecological constrains on gene evolution can be inferred. Furthermore, population structure, SNP diversity and phenotype association are other activities which can we undertaken
The specific goals of this research were to evaluate the allele diversity of the Asr1 and Asr2 genes in wild and cultivated common bean and to determine (1) the extent of haplotype diversity, (2) the allele distribution in relation with the gene pool origins and probable drought tolerance based on geographic origin, (3) the differences at these candidate genes between wild and cultivated common beans, and (4) the patterns of nucleotide variation as related to local adaptation to ecological environments.