Neglected treasures in the wild – legume wild relatives in food security and human health

Abstract

The legume family (Fabaceae) is the third-largest flowering family with over 18 000 species worldwide that are rich in proteins, oils, and nutrients. However, the production potential of legume-derived food cannot meet increasing global demand. Wild legumes represent a large group of wild species adaptive to diverse habitats and harbor rich genetic diversity for the improvement of the agronomic, nutritional, and medicinal values of the domesticated legumes. Accumulating evidence suggests that the genetic variation retained in these under-exploited leguminous wild relatives can be used to improve crop yield, nutrient contents, and resistance/tolerance to environmental stresses via the integration of omics, genetics, and genome-editing technologies.

Introduction

It is predicted that the global crop productivity must be doubled by 2050 to meet growing global demand [1]. However, the current increase rate of crop production cannot meet the challenge of food security because of two main reasons: 1) Modern crop cultivars are unable to adapt to harsh environments or changing climates due to low genetic variation after the domestication bottleneck; 2) Agricultural land is continuously decreasing due to urbanization, industrialization, and as well as the increasing demand for animal production (land competition) due to human dietary changes [2]. There is an urgent need in developing diverse crop cultivars that can grow in harsh environments with broad-spectrum biotic stress resistance. On the other hand, crop wild relatives (CWRs), the ancestor of modern crops, thrive in diverse and challenging environments, are under-utilized. They harbor rich genetic diversity that can provide novel genes/alleles for the development of crop cultivars more resistant and resilient to harsh growing conditions. The CWRs are usually classified into G1-G4 groups based on its relationship to existing crops. G1 is the most closely related group, while G4 is the most distant one [3,4]. In some cases, when the immediate wild relatives are unknown or extinct, the relatively distant wild species could be used as the alternative gene pools to enrich the reduced genetic diversity of cultivated crops. Over the past decades, tremendous efforts have been made to explore and utilize the genetic diversity in CWRs for crop improvements, such as pest resistance and abiotic stress tolerance [3,5]. Thus, CWRs serve as a vast reservoir of agriculturally-important genes. However, the majority of CWRs are still under-exploited, under-utilized, and not preserved. Figure 1 shows the strategies that can be used for the exploration, utilization, and conservation of leguminous wild relatives.

Figure 1.

A flowchart depicting the exploration and use of environment-adaptive leguminous wild relatives in developing climate resilient varieties. In situ and ex situ conservation are two major methods to maintain biodiversity. The in-situ conservation retains the species in their adaptive natural habitats, while ex situ conservation preserves the seeds from diverse habitats in a seed bank offering excellent opportunities for research and breeding. Applying high-throughput genotyping and phenotyping (including phenotypic characterization, transcriptomics (T), proteomics (P), metabolomics (M)) technologies on these in situ conserved natural wild populations may advance the dissection of ‘genotype-phenotype’ associations using various bioinformatics approaches. The use of the environmental variables is useful in identifying adaptation-associated genomic regions. For breeding, successful crossing allows the transfer of a superior trait from wild species into an elite variety for new cultivar improvement. Vigorous and fertile seeds produced between species can facilitate linkage mapping development for identifying candidate genes/loci associated with traits of interest. The molecular mechanisms elucidated by association and linkage analyses would shed light on the efficient management of in situ conservation of the leguminous wild relatives. These strategies aided with high-throughput biotechnologies could facilitate the identification of closely-linked markers for marker-assisted selection (MAS) or causal genes that could be edited for favorite traits.

The legume family (Fabaceae) is the third-largest plant family, of which many species are cultivated worldwide. Leguminous plants also fix nitrogen by interacting with symbiotic nitrogen-fixing bacteria [6], and produce usable forms of nitrogen that can be used by other organisms. However, when grown under stressful environmental conditions, many legume yields are significantly decreased. One reason is that cultivated crops have lost a considerable amount of genetic diversity during domestication process [3,7–9]. Additionally, compared to cereals, relatively less funding available for legume community could be another major barrier constraining extensive studies for legume improvement. In turn, there is increasing evidence indicating that leguminous wild relatives retain considerable novel alleles useful for crop improvement [10], such as salt tolerance [11••] and yield-related traits [12,13].

Here, we first review recent progress in the utilization of genetic variation in wild legume species involving biotic stress resistance, abiotic stress tolerance, nutrition, and phytocompounds. We then discuss the challenges and opportunities with the surge of genomics and metabolomics data, as well as cutting-edge genome-editing technologies. We provide a few examples for each trait, which are summarized in Table 1.

Table 1.

Recent studies exploiting the exotic genetic resources of leguminous wild relatives for legume crop improvement using various biotechnological approaches

Trait classification	Specific traits of interest	Wild leguminous species	Research strategy	References
Abiotic stress tolerance	Salt tolerance	Wild Vigna species	Phenotype characterization, genetic mapping	[14–17]
		Glycine soja	Gene cloning, metabolomics	[11••,18–22]
	Cold tolerance	Wild Cicer species	Phenotype characterization	[81,82]
	Drought tolerance	Wild Cicer, Arachis, common bean, and Lentil species	Phenotype characterization, GWAS	[8,26••,28,81]
Biotic stress resistance	Nematodes (Heterodera glycines, Meloidogyne arenaria) resistance	G. soja, perennial Glycine species	Phenotype characterization, RNA-Seq, GWAS, KASP, genetic mapping	[32,33,75,83]
		Wild Arachis species	Phenotype characterization	[36,37]
	Pathogen, late leaf spot and rust resistance	Wild Arachis spieces	Gene cloning	[84••]
	Stemphylium, Ascochyta, anthracnose, and Stemphylium blight resistance	Wild Lentil species	Genetic mapping	[38–41]
	Aphid resistance	G. soja	Genetic mapping	[3,43–45]
	Butterfly and Plume moth resistance	Wild Cajanus species	Genetic mapping	[46,47]
Yield-related	100-seed weight	G. soja	Genetic mapping, RNA-seq, gene function analysis	[12,50,51,65]
Seed composition	Protein	G. soja, perennial Glycine species	1-DE, GWAS	[64,85]
	Oil	G. soja	RNA-seq, phenotype characterization	[64,65]
Domestication	Seed weight, pod dehiscence, flowering, plant height, seed coat color	Wild soybean, lupin, pigeonpea, cowpea species	NGS, microarray	[2,86,87,88•,89•,90]
Metabolomics	Flavonoids, phenolics, isoflavones	G. soja	Genetic mapping	[91,92]
	Saponin	G. soja	LC–MS	[66,67]
Genome sequencing	Whole genome, plastome	G. latifolia, G. soja, G. gracilis	NGS	[93,94,95•]
Ecology	Adaptive traits, drought, salt	G. soja, wild chickpea species	NGS, GWAS	[26••,73•,74]

Note: GWAS, genome-wide association study; RNA-Seq: RNA sequencing; KASP assay (Kompetitive Allele Specific PCR assay): is a fluorescence-based genotyping strategy to distinguish two different alleles; 1-DE: 1-dimensional gene electrophoresis; NGS: next-generation sequencing; LC–MS: liquid chromatography–mass spectrometry.

Abiotic stress tolerance

Salt tolerance

Most legume crops are glycophytes; therefore, salt stress can significantly limit legume yield productivity. However, some of their wild relatives have adapted to various coastal and saline areas. Chankaew et al. [14] identified genetic variation associated with salt tolerance in the halophyte Vigna marina. The quantification of Na⁺ and K⁺ concentrations with improved methods has indicated different salt tolerance mechanisms among wild Vigna species [15–17]. For example, a group of wild Vigna species prevents salt-caused damages by excluding salt from the whole plants, while another group accumulates excessive Na⁺ in vacuoles [16,17]. Iseki et al. [17] suggested that salt tolerance in the genus Vigna has evolved multiple times independently after diverging from its ancestor. These Vigna plants adaptive to saline coastal areas are useful materials for the dissection of the underlying salt-tolerance mechanisms.

Recent studies in Glycine soja, the wild ancestor of cultivated soybean (Glycine max), have identified a number of genotypes with enhanced salt tolerance [11••,18]. A novel salt tolerance gene, GmCHX1, has been recently identified and cloned in G. soja via whole-genome sequencing [11••]. In addition, many candidate genes [18–22] from a single salt-tolerant G. soja genotype, G. soja07256, demonstrated improved salt tolerance when expressed in model plant species, Arabidopsis and Medicago. Given the importance of this genotype, it is worth pinpointing the causal gene(s) conferring salt tolerance with the genomics-linkage mapping approach as previously described [11••]. It is also important to test how these genes behave in saline soils when transferred to soybean cultivars via breeding or gene engineering. Meanwhile, it remains unclear if these genes will result in tradeoffs when used to develop salt-tolerant cultivars. Recently, metabolomics comparisons have provided new insights into salt tolerance mechanisms. For example, when under salt stress conditions, higher accumulation of carbohydrates, fatty acids, phenylalanine, proline, and palmitic acid was observed in the salt-tolerant G. soja plant than that in salt-sensitive soybean cultivar [23,24].

Drought tolerance

Unlike salt tolerance with relatively well-understood mechanisms, drought tolerance is more complex and usually found associated with many loci and each with small effects. Wild legumes are excellent materials for studying drought tolerance because most of them grow in arid regions [25,26••]. For example, many drought-tolerant wild lentil species (such as Lens odemensis, Lens tomentosus) [8] and wild chickpeas [26••] adapt to drought-prone areas. Under moisture-controlled soil conditions, these wild lentil genotypes can increase their capacity to avoid or tolerate drought stress by reducing transpiration rates, and some due to its deep root system. Meanwhile, various phenotypic trade-offs have been observed, such as delayed flowering, reduced plant height and growth rate [8]. Other than deep rooting and increased lateral roots in grain legumes for drought tolerance [27], wild Arachis duranensis maintains relatively high levels of transpiration and photosynthesis rate under dehydration treatment [28]. Even though physical differences exist among these wild species, the drought tolerance mechanisms remain similar: for example, enhanced water uptake, and reduced water loss by growing smaller leaves to maintain water balance under water deficits [28].

Plant root systems are crucial in drought tolerance and they affect aboveground yields. To date, many QTLs associated with root-related traits have been identified in legumes using interspecific mapping populations [27]. More QTLs will likely be identified in the near future as more advanced biotechnologies, such as 2D/3D root imaging systems, as well as more affordable sequencing platforms, are applied to root-related traits studies. Further determination of their genetic relevance to drought response could be helpful for root trait-directed breeding for drought management. Alternatively, a global transcriptomics analysis of drought-stressed wild A. duranensis revealed osmotic regulation as a primary mechanism in drought tolerance [28]. From an ecological adaptation perspective, drought tolerance in wild relatives could also be dissected with landscape genomics, that is, association analysis of their environmental factors combined with genomic variation [25,29,30].

Biotic stress resistance

Nematode resistance

Biotic stresses, such as phytoparasitic nematodes, can cause significant yield loss in a wide range of leguminous hosts. For example, soybean cyst nematode (Heterodera glycines, SCN) is the most devastating soybean pest, which causes over one billion yield losses per year in US [31]. Interactions between SCN, as well as root-knot nematode (Meloidogyne spp., RKN), and their hosts have been intensively studied because of their agricultural importance. To identify new sources of SCN resistance to widen the existing limited resistance gene pool (mainly from soybean cultivar PI88788), Zhang et al. [32] evaluated over 200 wild soybean (G. soja) genotypes and suggested to prioritize the screening of Japanese G. soja accessions for increased chances of finding resistant ones. While most G. max and G. soja are SCN-susceptible, 97.3% of 223 perennial soybean relatives, subgenus Glycine, exhibited varying levels of resistance to three SCN types [33], which makes this subgenus a valuable resource for breeding new cultivars with durable resistance to SCN. Despite the challenges due to cross-incompatibility between Glycine perennials (GP-3) and soybean, recent inspiring studies have shown that it is feasible to produce fertile seeds between Glycine tomentella (polyploidy) and soybeans [34,35••], indicating that the superior traits retained in polyploidy perennial could be potentially used to improve soybeans after significant improvements in hybridization techniques.

The applications of nematode-resistant wild resources in other legume species lag behind but examples are emerging. Arachis stenosperma, a wild relative of peanut (Arachis hypogaea), was identified to be a new source of RKN resistance [36]. It was then used to cross with peanut to develop mapping populations for identifying RKN-resistance genes [36]. A recent study on an expansin superfamily from A. duranensis and Arachis ipaensis, the wild progenitors of cultivated peanut, found that expansin-like B gene (AraEXLB8) shows RKN resistance by overexpressing it in soybean hairy roots [37]. With further functional validation, this gene could be useful in developing RKN-resistant peanut cultivars by transferring resistant alleles to peanut.

Disease resistance

Wild legumes also represent a genetic reservoir by providing exotic genes/alleles conferring broad-spectrum resistance to many different diseases, such as pathogen and fungi. Many wild relatives are resistant to various diseases. For example, Lens orientalis is resistant to stemphylium blight (Stemphylium botryosum) [38], and Len ervoides and Len nigricans are resistant to ascochyta blight (Ascochyta lentis) [39–41]. Wild lentils were also found to be resistant to anthracnose (Colletotrichum truncatum), fusarium wilt (Fusarium oxysporum), powdery mildew (Erysiphe polygoni), and rust (Uromyces fabae) [42]. Further assessment of these resistance sources is needed to evaluate the breadth of the resistance spectrum and to discover the resistance loci for lentil crop improvement.

Sap-sucking pest resistance

Wild legumes are also rich exotic resources for sap-sucking pest resistance. The soybean aphid (Aphis glycines Matsumura) causes severe yield loss annually in legume crops. It colonizes on soybean and other legumes such as Medicago and Trifolium [43]. For effective aphid management, aphid resistant genotypes have been identified in wild leguminous species, such as annual G. soja and perennial Glycine species (Glycine falcate, Glycine clandestina) [43,44]. Novel loci have been identified in these wild soybean relatives by genetic mapping [3,45]. Likewise, the source of resistance to other sap-sucking pests was also identified in wild legume relatives such as Cajanus scarabaeoides (a pigeon pea wild relative) for Plume moth (Exelastic atomosa) [46], Cicer wild relatives (Cicer reticulatum) for Helicoverpa armigera [47]. Next generation sequencing of segregating populations as used in wild soybean G. soja [11••,45] may significantly accelerate the dissection of the underlying mechanisms in these under-exploited legumes.

Combination and broad-spectrum resistance and tolerance

In nature, plants usually experience a combination of biotic and abiotic stressors rather than an individual stressor. Therefore, enhancing the legume crops with tolerance/resistance to multiple stresses has become increasingly important. It is a primary goal of legume breeding in the light of future climate instability. Interestingly, wild leguminous relatives are currently identified either resistant to broad-spectrum biotic stresses or tolerant to multiple abiotic stresses. However, wild legumes that combine biotic stress resistance and abiotic stress tolerance are rarely studied. For example, wild pigeon pea genotypes, C. scarabaeoides, were reported to be resistant to both blue butterfly and plume moth [46], the wild lentil (L. ervoides) is resistant to both anthracnose and Stemphylium blight [48], and wild common bean is tolerant to both drought and subzero temperatures [49•]. Ideally, some wild relatives such as wild chickpeas, adaptive to local habitats and resistance to biotic stress, could be an ideal system for analyzing the combination of stress resistance [26••]. In addition, the transfer of multiple desired traits from a wild relative to its elite descendant is still challenging, unless they are controlled by a single or closely linked loci [46]. Nevertheless, interspecific introgression of multiple abiotic stresses into cultivated peanut cultivars has been successful [49•], making this strategy feasible and promising.

Yield-related traits

One of the main contributions of legume wild relatives in yield-related traits is that they are often used as one of the two parents when developing mapping populations for identifying candidate genes associated with seed size, weight, and number, which are essential yield-related traits [50,51,52•]. Interestingly, alleles from wild relatives benefiting yield improvement have also been reported [e. g. Ref. [53]]. For example, Li et al. [53] identified a QTL for seed yield in a backcross population between wild soybean G. soja and cultivated soybean G. max. The lines carrying G. soja alleles demonstrated a 6.3% yield increase than lines homozygous for the G. max allele. Considering that the yield traits are complex and mostly attributed to additive and epistatic effects of many loci [12], transcriptomics-based comparisons between wild and cultivated soybeans resulted in the identification of a WRKY gene correlated to seed size [54•]. Candidate genes, such as WRKY, phosphatase 2C-1 (PP2C-1), as well as the environmentally stable yield-enhancing QTLs, identified in the leguminous wild relatives [12,50,52•,55,56] merit further efforts for function validation and field testing. Their potential applications in legume yield improvements make pre-breeding practices worthwhile.

Nutrition and phytocompounds

Seed composition

A high proportion of proteins, fats, carbohydrates, dietary fibers, B-group vitamins, and minerals can be obtained from legumes [57]. A study on all-cause mortality found out that a daily intake of 20 g of legumes can reduce the rate of mortality [58]. According to the Dietary Guidelines for Americans, an intake of three cups of legumes weekly is recommended for people who consume about 2000kcal/day [59]. These nutrient-related traits were selected during domestication, for example, starch and fat in adzuki bean (Vigna angularis) [60], and protein and oil in soybeans [61]. Intensive studies on these traits have uncovered a number of novel loci within the same genomic regions [62,63•,64]. For example, in soybean, the environmentally stable QTLs on chromosome 20 for seed protein and oil have been repeatedly identified among many different interspecific populations [63•,65]; thus deserving further identification of useful haplotypes/alleles to facilitate seed improvement. Functional markers developed from these important loci, such as dCAPS [62], may significantly accelerate the improvement of seed composition.

In turn, human selection of agriculturally important traits resulted in unintended accumulation of secondary compounds in legume seeds, such as phenolics, flavonoids [66], and saponins [66,67]. High accumulation of these compounds in G. soja might confer resilience to environmental stresses. Meanwhile, they have also been demonstrated to be beneficial to human health, which holds great potential in preventing human disease and developing alternative medicine. Figure 2 demonstrates this ‘two birds, one stone’ scenario of phytocompounds.

Figure 2.

The ‘two birds, one stone’ scenario of phytocompounds: high accumulation of phytocompounds under biotic and abiotic stress conditions can help plants confer resistance to environmental stress, and they are also beneficial to human health.

Phytocompounds

Legumes are a crucial source of many phytochemicals that serve as human-health promoting ingredients in alternative medicines [68]. For example, isoflavonoid formononetin, common in wild legume species, has anti-hyperglycemic activity [69], and reduces insulin resistance and hyperglycemia [70], holding great potential for assisting in the treatment of diabetes. Phytoalexin glyceollins can be induced in soybean and wild soybean, G. soja, (Song lab, unpublished data). They showed anti-proliferative effects on breast cancer cells [71]. Von Wettberg et al. [26••] reported that chickpea wild relatives are rich in health-beneficial phytochemicals, such as polyphenolics, flavonoids, etc. Table 2 summarizes the recent studies on health-beneficial phytocompounds in wild legume species.

Table 2.

Recent studies exploiting the health-beneficial phytocompounds from leguminous wild relatives

Legume wild relatives	Phytocompounds	Nutritional values/alternative medicine	References
Glycine soja (wild soybean)	Soyasaponins, Soybean agglutinin Bioactive peptides, Lunasin, Genistein Formononetin	Anti-inflammatory; Natural antioxidant agent; Vasodilation; Anticancer activity	[96,97•,98]
Medicago sativa and Medicago lupulina	Flavonoids (Formononetin, Biochanin A, Daidzein, Genistein), Polyols, Coumestrol	Hypocholesterolemic effects and can prevent associated diseases; Antimicrobial activity	[99,100]
Medicago trunculata	Proanthocyanidins (PAs),	PAs reduce risks of cardiovascular disease and Alzheimer’s disease; Anticancer activity	[101]
Butea monospermea (Flame-of-the-forest and bastard teak)	Genistein (flavonoid)	Antioxidant; Anti-inflammatory; Vasodilation; Antimalarial activity	[97•,102]
Trifolium pretense (red clover) and Trifolium medium (zigzag clover)	Formononetin, Daidzein, Genistein, Biochanin A	Nutritional, mineral and bioactive values, antimicrobial activity	[99]
Glycyrrhiza uralensis (Chinese liquorice)	Flavonoids (flavanones, chalcones, and isoflavones), Triterpene saponin	Anti-inflammatory, antiviral, antimicrobial, antioxidative, antidiabetic, antiasthma, antiallergic and anticancer activities as well as immunomodulatory, gastroprotective, hepatoprotective, neuroprotective and cardioprotective effects	[103]
Clitoria ternatea (Butterfly pea)	Phenolic metabolites, Hydroalcoholic extracts, Steroids, Saponins, Flavonoids, Lectins, Tocopherols, Phytosterols	Antifungal; Anti-inflammatory; Antibacterial; Anti-hyperlipidemia; Antiasthmatic; Antidiabetic, nootropic, anxiolytic, anticonvulsant, sedative, antipyretic, and analgesic functions; Antioxidant; Decrease laryngeal cancer cell (HEp-2) viability	[98,104,105]
Pueraria thunbergiana (Kudzu bean)	Puerarin (isoflavone), Saponin	Effective in the treatment of osteoporosis (OP) and osteoarthritis (OA); Highly beneficial in treating liver damage; Inhibits HIV-1 initial attachment and inhibits HIV-1 replication; Inhibits HIV-2 and simian immunodeficiency virus; Anti-inflammatory; Antioxidant activity	[106–108]
Lupinus albus (Lupine)	Polyphenols (Quercetin, Caffeic acid, Ferulic acid, p-Coumaric acid); Tannins	Antioxidant activity	[68,109]

Emerging studies have shown that health-beneficial nutrients and phytocompounds in wild legumes could be used to develop alternative medicines for treating human disease. However, only a small fraction of the immensely diverse plant metabolites have been explored for the production of new medications or health supplements. Studies on dissecting the genetic basis of compound production/induction are even more scarce. The isolation of key enzymes for undefined steps in the modification of secondary metabolites, such as triterpenoid saponins in legumes, will shed light on the metabolic engineering of useful phytocompounds [72].

Uncover adaptive loci from an ecological perspective

Given that CWRs grow in diverse environments with broad ecological and geographic distributions, the environmental variations may drive population differentiation for local adaptation, as shown in G. soja [32,73•,74,75], wild Yigna [17], wild common beans [25,76], and wild chickpeas [26••]. During the past few years, landscape genomics has been an emerging integrative research field in identifying adaptation-related loci/SNPs by association studies using environmental variables as quantitative traits. Anderson et al. [73•] recently identified candidate SNPs in G. soja associated with environmental variables, such as temperature, soil, and precipitation. It is expected that more adaptation-associated loci will be identified in other wild legumes once the high-throughput genomic data and ecological variables are available. On the other hand, comparative transcriptome analyses between locally adaptive wild legume populations could lead to the identification of ecological adaptation-related transcripts or allelic variants. By linking the variants with its origin information, breeding scientists may prioritize the locally adapted relatives to advance the improvement of abiotic tolerance through marker-assisted selection.

Challenges and perspectives

Trait transfer

Traditional genetic crossing remains one of the primary strategies for transferring the target traits from a wild relative to an elite variety, despite the fact that the linkage drag may have negative effects on the elite genotype. The recent successful cross of polypoid G. tomentella [34,35••] and diploid soybean showed the feasibility of transferring genes/alleles from the gene pool of GP3 or GP4 to elites after efficient crossing methods are developed. Alternatively, the useful alleles identified in wild species could be ‘duplicated’ into cultivated varieties using the CRISPR/Cas9 system, which is the most powerful tool to date used to make genetic changes in a DNA fragment [77].

Trait dissection

The use of the genetic variation in wild relatives to benefit crop improvements relies largely on the accurate measurements of the trait variation. However, large-scale and accurate measurements of trait variation in legume wild relatives are still challenging because most phenotypically look like weeds and are difficult to handle. Along with the high-throughput phenotyping strategies [78], high-resolution SNP data are indispensable for the identification of the associated loci by genome-wide association studies (GWAS) and linkage mapping. Such analyses can be extended to use the expressed transcriptome, proteome, and metabolome to identify the regulatory/expressed QTLs, considering epistatic effects on complex traits variation. Given that either genomic or regulatory variation may cause trait variation, rather than focusing on a single wild genotype, a pan-genome approach is an effective strategy to capture the genomic variation associated with diverse traits within species [78,79••,80].

A long way to go

Although immense progress has been made, it is just the beginning of using these untapped genetic resources. The knowledge obtained from what we currently know on CWRs is likely the tip of the iceberg; thus, more systematic efforts are needed to increase our understanding and advance legume breeding. These efforts include efficient conservation of these resources in-situ or ex-situ, integration of multiple-dimensional data generated from different platforms, systematic data curation, data/germplasm sharing, advanced biotechnology development, high throughput genotyping and phenotyping, and more affordable technologies quantifying human-beneficial phytocompounds.