Importance of grass pea (Lathyrus sativus L.)

History of the crop

Grass pea (Lathyrus sativus L.) is one of the oldest cultivated crops with a long history of domestication. In archaeological excavations in Turkey and Iraq, seeds of Lathyrus species had been found as collected or cultivated items. Similarly, seeds from 2500 bc were identified in the oldest excavations in India (Kislev 1989) and already in the Balkan in 8000 bc. According to the legend, grass pea was brought to Ethiopia (Abyssinia) by the queen of Sheba after visiting King Salomon in the tenth century bc (Budge 1928).

The origin of the name Lathyrus is from the Greek word “la thyros” that hints at something exciting, referring to the aphrodisiacal qualities ascribed to grass pea (Loudon et al. 1855). Although the term “lathyrism” used to designate a disease caused by overconsumption of grass pea was coined by Cantani of Naples only in 1873, the history of lathyrism dates back to ancient times (Barrow et al. 1974). In the funeral offerings found in the Egyptian pyramids, various legume seeds were present, including grass pea. Apparently, grass pea was considered a special food to be offered to kings, in contrast to the present-day reputation of being the survival food of the poorest of the poor. Ancient civilizations in the Middle East and the Americas included grain legumes and cereals in diets with well-balanced composition in essential amino acids.

Foods prepared from grass pea seeds are very tasty. They have been and are still popular in many European countries (e.g., Spain, France, Portugal, Italy, and Poland), in Africa (for instance, Ethiopia), and in South Asia (e.g., India, Bangladesh, and Nepal). In Ethiopia, grass pea dishes include boiled whole seeds (nifro), roasted whole seeds (kollo), the traditional sauce (shiro wott and kik wott), the local drink (areke), and the green unripe seeds eaten as snack by young boys on their way from school to home (eshet) (Fikre et al. 2011a). Moreover, many European languages have also their own local names for Lathyrus (Gry et al. 1998). For example, a gastronomic festival of chícharo (grass pea) in Alvaiázere, a small village in the central region of Portugal, is organised yearly for one whole week since 2001. The purpose of the festival is to promote the local knowledge on this forgotten crop and its associated traditions. Moreover, it provides opportunities for local farmers to sell their products outside the conventional commercial systems. With the support of many local restaurants, a great diversity of grass pea menus is offered during the festival days (Vaz Patto 2009). In different parts of Spain, titos, almortas, muelas, or guijas are some of the names used for grass pea. During the Spanish Civil War and World War II in the Greek maquis, this presently neglected species was an important crop (Lambein 2015). It was used for different purposes, from animal feed, as supplement to cereals, to human consumption. In these regions, one of the most “healthy” and amazing custom for almost all village children until the 1960s was “corer los titos” (“running for grass peas”). The aim of the game was to slip into the neighbours’ grass pea fields and to eat the immature green seeds with the hope not to be caught by the owners. Today, in these same countries, grass pea is an almost forgotten crop, just sown in a few small plots to be consumed mainly during some religious festivities (Caminero Saldaña and Grajal Martín 2009).

Grass pea, a geographically successful crop

Grass pea is an important annual legume crop cultivated in several drought-prone areas of the world, including Eurasia, North America, temperate South America, and East Africa (Vaz Patto et al. 2006a; Yan et al. 2006; Dixit et al. 2016). Whereas in India, it is cultivated up to 1300 m above sea level (Duke 1981), in some parts of Ethiopia, it can be grown at a height of 2500–3000 m with annual rainfall averaging 1000 mm (Fikre et al. 2011b). Grass pea is farmed as a summer crop in the high altitudes of Kashmir and Nepal and as a winter crop at low altitudes, for instance, in Bangladesh (Girma and Korbu 2012). Unlike most orphan crops often adapted to local environments and with a limited geographical distribution, grass pea has spread over three continents and has adapted well to various harsh environments, from the highland volcanic soils of Ethiopia to heavy clay in the paddy rice (Oryza sativa) fields of Bangladesh (Fig. 1).

Fig. 1
figure 1

Grass pea cultivation in Ethiopia: a grass pea in the greenhouse, b grass pea in the field, c grass pea on Ethiopian volcanic soil

Full size image

The total acreage of grass pea is estimated at 1.50 million ha with an annual production of 1.20 million tons, mainly in South Asia and in East Africa (Sammour 2014). In Ethiopia, the production of the crop has increased both in area coverage and production volume, from 185,490 tons in 2005 to 287,674 tons in 2015. The productivity per hectare has also risen from 1260 kg to 1808 kg in the same period (Central Statistical Authority 1990–2009). In Bangladesh, grass pea (Khesari) is cultivated on 239,343 ha with a mean yield of 728 kg/ha both in dry areas and in flooded rice fields and it is often used as intercrop in wheat (Triticum sp.) or rice fields (Malek et al. 2000).

The genus Lathyrus, which belongs to Fabaceae (Leguminosae), is represented by more than 160 species worldwide (Allkin et al. 1986), some of which are economically important as food (pulse), fodder, forage, ornamental crops, and even in traditional medicine (mainly L. sativus, L. cicera, and L. odoratus) (Vaz Patto and Rubiales 2014). In Europe, the interest in grass pea was revived because of the urgent need to recover marginal land and to provide an efficient alternative to land overexploited by cereal cultivation (Granati et al. 2003). A list of the species known to be historically or currently cultivated for agriculture or horticulture is presented (Table 1).

Table 1 Historic or currently cultivated Lathyrus species.
Full size table

Grass pea is a typical orphan legume crop, such as cowpea (Vigna unguiculata), groundnut (Arachis hypogaea), and bambara groundnut (Vigna subterranea) (Cullis and Kunert 2017). Orphan or underutilized crops, including orphan legume crops, play a significant role in many developing countries, because they provide food and nutrition security to consumers, as well as an income to resource-poor farmers. They are so-called orphan crops, because they have largely been neglected and their cultivation and use are poorly documented by the international scientific community, the research networks, and the industry, when compared to cash crops, such as rice, corn (Zea mays), and wheat (Foyer et al. 2016). This is mainly due to their limited economic importance in the global market and the weak or lack of formal seed supply systems (Cullis and Kunert 2017).

Orphan legumes are primarily grown by resource-poor farmers, predominantly women, who use their own seeds on small landholdings in specific agro-ecological niches and on marginal and sub-marginal lands to provide families with high nutritional value food (Diane 2016). In particular, the vital importance of grass pea in many low-input farming systems (Vaz Patto et al. 2006a) originates from its tolerance to abiotic stresses, such as drought, salinity, flood, and waterlogged conditions (Kumar et al. 2011; Jiang et al. 2013; Piwowarczyk et al. 2016; Zhou et al. 2016). In years of severe flooding or drought, grass pea is the only crop that survives in the field. Interestingly, in Eastern Poland, cultivation of grass pea (Podlaska lentil) was stimulated after several crop failures due to drought in the 1990s (Milczak et al. 1997). Grass pea has a hardy and penetrating root system suited to a wide range of soil types, from low fertility soils to heavy clays (Girma and Korbu 2012). Grass pea also responds positively to a favorable environment, giving yields as high as 5 tons/ha under good agro-climatic and management conditions (Briggs et al. 1983).

The robustness of grass pea along with its intrinsic capacity to fix biologically atmospheric nitrogen through its rhizobia-containing roots, a feature shared with other legumes, makes it an attractive crop to grow under adverse agricultural conditions (Campbell et al. 1993). As an efficient nitrogen fixer, grass pea is a good green manure that improves soil fertility by supplementing approximately 67 kg/ha of nitrogen in a single season. This specific ability confers yield and protein benefits for the subsequent non-legume crop (Kumar et al. 2013). In India, most of the grass pea acreage (~ 70%) lies in a rice-growing region where additional irrigation is available only for rice and, consequently, not for subsequent winter crops. Hence, grass pea is the only alternative crop that can be cultivated after rice (Yadav and Mehta 1995), because it efficiently uses moisture retained in the soil and has beneficial effects on soil fertility and soil health. Similarly, in the Eastern Indo-Gangetic plains, a sizeable area is mono-cropped with rice. The lack of irrigation water and the delay in harvesting the rice fields do not normally permit double cropping. Under such conditions, relay cropping of Lathyrus could transform these mono-cropped areas into double-cropping areas, increasing the productivity and sustainability of the agricultural system (Dixit et al. 2016).

Grass pea, boon or bane?

Nutritional, nutraceutical, and pharmaceutical properties of grass pea

Grass pea can serve a variety of purposes, such as animal feed and fodder, but also as human food, thanks to 18–34% and 17% of protein content in seeds and mature leaves, respectively (Rizvi et al. 2016). These values are higher than those of field pea (Pisum sativum subsp. arvense (L.) Asch) (23%), faba bean (Vicia faba) (24%), or lupine (Lupinus albus) (32%) (Petterson et al. 1997), but lower than those of soybean (Glycine max) (42%) (Ravindran and Blair 1992). The L. sativus seed protein is composed of albumins (14%), globulins (66%), glutelins (15%), and prolamins (5%) (Chandna and Matta 1994). Whereas animal feed from grass pea usually consists of ground or split grain or flour and is used primarily to feed lactating cattle or other draft animals, human diets include Lathyrus as grains that are boiled and then either consumed whole or processed for split dal (Enneking 2011). Moreover, grass pea is highly suitable for human consumption because 58% of the fatty acids are polyunsaturated (Grela et al. 2010). As a nutrient-dense food/feed crop and its high drought tolerance with minimal input requirements for its cultivation, grass pea provides food and nutrition security to many low-income communities.

Grass pea, like other orphan legumes, is still an untouched treasure for compounds that can contribute to human health. For instance, it is the only known dietary source of l-homoarginine. Therefore, as nutraceutical, grass pea is an excellent example of a potential “functional food” (Singh and Rao 2013; Llorent-Martínez et al. 2017). The amino acid l-homoarginine provides benefits in cardiovascular disease treatments (Rao 2011; Singh and Rao 2013; van Wyk et al. 2016) and in overcoming the consequences of hypoxia, i.e., the inadequate oxygen supply at the tissue level, associated with cancer tumor development (Ke and Costa 2006; Jammulamadaka et al. 2011). Thus, a daily dietary intake of l-homoarginine through small quantities of grass pea may be valuable for human health and deserves to be studied further (Rao 2011).

Furthermore, as most grain legumes, grass pea is deficient in the essential sulfur-containing amino acids, methionine and cysteine, but is rich in lysine that is low in cereals (Gatel 1994; Ravindran and Blair 1992; Mahler-Slasky and Kislev 2010). The amino acid profiles of L. sativus are similar to those reported for many grain legumes (Hanbury et al. 2000). The deficiency of essential sulfur-containing amino acids, such as methionine that plays a vital role in the central nervous system (Amara et al. 1995), may be overcome with a balanced diet containing cereals (Lambein and Kuo 2004).

However, grass pea suffers from a reputation of being toxic, because under certain circumstances its overconsumption has caused neurolathyrism, a neurodegenerative disease in humans and domestic animals (Lambein and Kuo 2009), due to its content of the neuroexcitatory β-N-oxalyl-l-α,β-diaminopropionic acid (β-ODAP) (Vaz Patto and Rubiales 2014). This crippling, but not lethal, disease is more pronounced when grass pea is the dominant component of the diet and accounts for at least 30% of the caloric intake for a period of at least 3–4 months (Kumar 1998). Despite this stigma, β-ODAP in grass pea has also been described as a multifunctional plant metabolite. One of the physiological functions with possible therapeutic potentials is the activation of protein kinase C, which adds a new dimension to explore its potential in the treatment of Alzheimer’s disease, hypoxia, and long-term potentiation of neurons essential for memory (Singh and Rao 2013). Recently, the biological properties (antioxidant, enzyme inhibitory, and cytotoxic effects) and phytochemical profiles of selected Lathyrus species (L. czeczottianus and L. nissolia) (Llorent-Martínez et al. 2017) have been studied for the first time. Whereas L. czeczottianus exhibited remarkable antioxidant abilities with the highest concentration of phenolics, L. nissolia had promising enzymatic inhibitory effects against cholinesterase, amylase, and glucosidase. Thus, these species could be utilized to design new phytopharmaceutical and nutraceutical formulations.

To further challenge the bad reputation of grass pea and its metabolite β-ODAP, a Chinese group has obtained a patent for the use of β-ODAP as a hemostatic agent during surgery (Lan et al. 2013). β-ODAP is also present in the longevity-promoting Panax ginseng root (Kuo et al. 2003) that under the name Dencichin is known for its hemorrhage-stopping property and thrombopoiesis treatment (Ding et al. 2018). In China, it is even added in some toothpaste brands and used as an herbal medicine to avoid bleeding. Recent investigations have revealed that β-ODAP induces wound healing and can be considered as a natural wound-curative agent (Sharma et al. 2018). An increasing number of therapeutic applications derived from Lathyrus may be developed in the coming years.

Neurolathyrism

The symptoms of both neurolathyrism and konzo are sudden onset of non-progressive symmetric spastic paraparesis. Intriguingly, these two diseases with identical clinical symptoms are linked to the consumption of totally different plant foodstuffs. The occurrence of konzo (meaning tied legs in the Yaka language) is associated with ingestion of the carbohydrate-rich roots of Manihot esculenta (cassava) that contain residual cyanogens, whereas neurolathyrism (guaya beshita in Amharic, meaning broken legs) is correlated with the consumption of the protein-rich seeds of L. sativus that contain a neuro-excitatory amino acid. The distinction between konzo and neurolathyrism can only be made by the historical consumption of cassava roots and grass pea seeds, respectively (World Health Organization 1996). The etiology of both diseases involves oxidative stress that has many contributing factors, such as imbalance of micronutrients, depletion of reducing power and of glutathione by the lack of sulfur-containing amino acids and antioxidants in the diet.

The epidemiology and the socio-economic background of the two diseases that occur in different geographical areas are also very similar, because both diseases happen mainly among very poor rural people, living as subsistence farmers on drought-prone and marginal lands. Epidemics of konzo in Mozambique and Tanzania (1980s and 1990s), epidemics of neurolathyrism in Bangladesh (1980s), India (1980s), and Ethiopia (1980s and 1990s) had all spread during or after drought periods (Bradbury and Lambein 2011). The question arises whether konzo and neurolathyrism are caused by intoxication, by an essential nutrient deficiency, or a combination thereof. Very plausibly, a nutrient deficiency together with toxic insults could cause a common syndrome (Bradbury and Lambein 2011). Both conditions can be considered socio-economic famine diseases.

From epidemiological surveys in Ethiopia and Bangladesh, neurolathyrism occurs seemingly only when three conditions are present: extreme poverty of the grass pea farmers, illiteracy of the consumers, and availability of grass pea as the cheapest food during drought-triggered famines, because, as mentioned, it is the only crop that can survive in the field, with prolonged overconsumption as a consequence. In Bangladesh, during the last neurolathyrism epidemic, the price of grass pea was the lowest of all foodstuffs. Even, although the grass pea production has increased in Bangladesh, no new cases of neurolathyrism have been reported. The price of grass pea is now equal or higher than that of rice and, hence, overconsumption as the cheapest food is no longer a risk for the poor (Haque et al. 1997).

Recently, stress has been identified as a risk factor for neurolathyrism. Experiments with new-born rats that received daily injections of β-ODAP revealed a 4.5-fold higher incidence of neurological symptoms when put under stress (Kusama-Eguchi et al. 2010). The impact of stress may be the reason for the extremely high incidence of neurolathyrism among the inmates in a World War II labor camp in Vapniarca (Ukraine) (Lambein et al. 2001). Stress may also have played a role during the civil war in Spain or in the maquis in Greece during World War II, when cases of neurolathyrism were reported. It may also affect the integrity of the blood/brain barrier that in a healthy person protects the central nervous system from toxic metabolites in the bloodstream, including β-ODAP after grass pea consumption (Kusama-Eguchi et al. 2010).

The incidence of neurolathyrism can be reduced by simple dietary measures. The occurrence of the disease can be prevented by addition of sufficient cereals to improve the amino acid score of the diet and of antioxidant-containing nutrients (fruits and vegetables) (Getahun et al. 2003). Moreover, seeds can be partly detoxified by various food processing methods (such as soaking in water or boiling) and the grains can be roasted at 150 °C (Kumar et al. 2011). Furthermore, fermentation has been found to reduce β-ODAP by 80–90% (Kuo et al. 1995), although an attempt to use fermentation with Bacillus species might not become useful due to practical difficulties. Chemical compounds, such as lime water, sodium chloride, sodium bicarbonate, ascorbic acid, etc., can easily be used in households to reduce the β-ODAP content in grass pea seeds without any effect on the quality parameters, namely content of protein, carbohydrates, and sugar (Geda et al. 1995).

Importantly, both konzo and neurolathyrism are restricted to rural areas, in contrast to urban areas, where the availability of other foodstuffs is higher. Contrary to many infectious diseases, konzo and neurolathyrism are not reportable diseases. However, these socio-economic diseases can be considered as alarming signals of extreme poverty and food insufficiency. Regarding toxicity, it is good to go back to Paracelsus (1493–1541) who mentioned that “everything is toxic and nothing is free of toxicity, only the dose determines toxicity”. Indeed, for instance, 500 g of kitchen salt or sodium chloride in a minimal volume of water is lethal, but kitchen salt is never labeled with toxicity symbols. After β-ODAP had been discovered as a neuroactive non-protein amino acid in grass pea seeds in 1964 (Rao et al. 1964), it was believed to be the cause of neurolathyrism. Neurotoxicologists considered neurolathyrism as a special case of neurodegeneration for which the origin was known, in contrast to other neurodegenerations, such as Alzheimer’s disease and amyotrophic lateral sclerosis (ALS). Although numerous studies were published on the “toxicity” of this unusual amino acid and the plant producing it, the presence of β-ODAP could not explain all the results. Neither the threshold toxicity level, nor the 50% effect level of the neuroexcitant could ever be established. In contrast to the unprocessed roots of bitter cassava and high amounts of kitchen salt, ingestion of grass pea is not fatal. An intake of at least 50 kg during a limited, uninterrupted period and without mixing with cereals or antioxidant-rich condiments would be necessary to obtain a 5% chance of irreversible crippling, without impact on longevity or cognitive functions.

In India and Nepal, a ban on the sale and storage of grass pea that had been imposed in 1961 after an epidemic of neurolathyrism has been the main reason behind a drastic reduction in grass pea cultivation (Kumar et al. 2011). The epidemic was in part triggered by the local tradition to pay poor labourers in kind with cheap food, sometimes exclusively with grass pea (Rutter and Percy 1984), a practice that was later prohibited. Nevertheless, the production area decreased from 1.30 million ha to approximately 850,000 ha in recent decades. Proposals and discussions to lift this ban are ongoing in the Indian media, even though a research panel headed by the Indian Council of Medical Research (ICMR) has cleared the consumption of khesari dal—considered the poor man’s dal due to its low cost (Utkarsh 2016). Similarly, before the 1960s, grass pea was widely planted in North-West China (in Gansu province, more than 20,000 ha), but in the 1970s, a severe drought caused a terrible famine in this area and a severe neurolathyrism epidemic. Grass pea was then forbidden and the remaining seeds were confiscated (Yang and Zhang 2005).

The biochemistry behind neurolathyrism

The etiology of neurolathyrism is complex. As mentioned, the cause has been suggested to be the neuroexcitatory amino acid β-N-oxalyl-l-α,β-diaminopropionic acid (β-ODAP), present in the Lathyrus species (Bell 1964). This small molecule has many physiological and biochemical activities and its level in the seeds is affected by genetic and environmental factors (Lambein et al. 2007). β-ODAP is biosynthesized from a heterocyclic amino acid β-isoxazolin-5-on-2-yl-alanine (BIA) that is only abundantly present at the seedling stage of grass pea. BIA is also found in seedlings of the garden pea (Pisum sativum) and lentil (Lens culinaris), where, in contrast with grass pea, it is not further metabolized into β-ODAP (Kuo et al. 1998). Drought, zinc depletion, and excess iron or cadmium in the soil can stimulate the plant to increase its production of β-ODAP. Recently, the biosynthesis pathway of β-ODAP has been studied in grass pea by means of metabolomics and has clearly been linked with the primary metabolism (Liu et al. 2017). The β-ODAP biosynthesis was found to be co-regulated with the nitrogen and sulfur metabolisms, more specifically with serine and cysteine that were inversely proportional to the β-ODAP accumulation, and β-cyanoalanine synthase was identified to be the key enzyme (Liu et al. 2017). However, the crucial metabolic step from BIA to β-ODAP remained unexplored. This is the fundamental difference with lentil and garden pea, in which the similarly abundant BIA is not further metabolized into β-ODAP. Nevertheless, the enzymatic step leading from BIA to β-ODAP had been studied previously (Ikegami et al. 1999). More recently, a novel cysteine synthase (LsCSase) gene from grass pea has been identified and characterized through bioinformatics approaches (Chakraborty et al. 2018). The LsCSase gene was up-regulated in young seedling tissues as well as in young seeds with an increased expression level under zinc-iron stress conditions and polyethylene glycol-induced osmotic stress. Hence, this preliminary investigation implies a correlation between the LsCSase gene and the β-ODAP biosynthesis pathway (Chakraborty et al. 2018). Further studies are needed, for example through the development of genetically engineered grass pea lines in terms of the CSase gene to explore its exact activity during the β-ODAP biosynthesis.

The best known and most studied property of β-ODAP is the excitation of a subgroup of glutamate receptors, the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors (Pearson and Nunn 1981) that trigger a cascade of neurophysiological reactions, including Ca2+ transport and NO production (Van Moorhem et al. 2010). β-ODAP inhibits not only the transport of the natural neuroexcitant glutamate, increasing and extending the excitation (Kusama et al. 2000), but also the enzymes involved in amino acid metabolism (Shasi Vardhan et al. 1997) and the mitochondrial activity (Ravindranath 2002). Besides the excitation of glutamate receptors, β-ODAP has many additional physiological and biochemical effects (Lambein et al. 2007). Recent sophisticated experiments, in which a human glioma cell line was treated with β-ODAP (Tan et al. 2017), confirmed the decrease in mitochondrial membrane potential, leading to the Ca2+ release from the mitochondria into the cellular matrix. The increased Ca2+ in the cellular matrix activated a new pathway and brought about the overexpression of specific proteins on the cell membrane surface. This new pathway led to cytoskeleton polymerization inhibition, cell microfilaments distortion, and, ultimately, cytoskeleton collapse. New-born chicks with an immature blood/brain barrier are susceptible to β-ODAP that generates transient neurological symptoms different from those of neurolathyrism (Anil Kumar et al. 2018). Proteomic analyses of chick brains revealed that several proteins involved in cytoskeletal structure, signalling, cellular metabolism, free radical scavenging, oxidative stress, and neurodegenerative disorders were initially up-regulated 2 h after treatment and reverted to normal levels by 4 h (Anil Kumar et al. 2018). Research on the effect of low concentrations of β-ODAP on human neuroblastoma cell lines indicated that β-ODAP can be explored as a stimulator of various reactive oxygen species (ROS)-mediated cell signalling pathways that are not detrimental to cells (Anil Kumar et al. 2018). Unravelling the fundamental steps and regulation of the β-ODAP biosynthesis in grass pea will be important for the development of improved varieties of this orphan legume.

Research areas and achievements in grass pea improvement

Grass pea, a climate-resilient crop

Grass pea is cultivated worldwide and is one of the cheapest sources of dietary protein in the developing world (Enneking 2011). It is an excellent example of an orphan crop adapted to drought and salinity and with a high water use efficiency (Kumar et al. 2011; Jiang et al. 2013; Piwowarczyk et al. 2016; Zhou et al. 2016). Under the prevailing climate change and the serious concerns about sustainability of agricultural production and food security worldwide, investigation of the basis of this high tolerance to drought, heat, and salinity is of great interest taking full advantage of grass pea as an elite germplasm resource (Boukecha et al. 2017). Grass pea is among the priority crops in the Kew’s Millennium Seed Bank and the Global Crop Diversity Trust project “Adapting Agriculture to Climate Change”. A primary objective of the Kew’s project is to collect and protect the genetic diversity of plants harboring traits that could be used for adaptation of the world’s most important food crops to new climatic conditions. Such adaptation is considered a key component in securing the world’s future food production (Dempewolf et al. 2014). Grass pea is largely considered as an “insurance crop” in areas that are prone to abiotic stresses, because it can produce reliable yields when all other crops fail due to drought (Yan et al. 2006; Vaz Patto and Rubiales 2014).

Water shortages negatively influence plant growth and development, consequently heavily limiting plant productivity. Cell turgor is lost by the reduced water availability and, subsequently, cell elongation and plant growth are hampered (Taiz and Zeiger 2010). Although the exact physiological and molecular mechanisms behind grass pea tolerance to abiotic stresses are still unknown, reports hint at a high ability to adjust the plant’s osmotic potential (Jiang et al. 2013; Piwowarczyk et al. 2014) and at the accumulation of polyamines and β-ODAP (Xing et al. 2001; Xiong et al. 2006). Furthermore, the mechanism of salinity tolerance in grass pea was found to result probably from the elevated activity of antioxidant enzymes in the root cells, manifested by the increased accumulation of phenolic compounds and peroxidase activity (Jiang et al. 2013; Piwowarczyk et al. 2016) that might provide a better adaptation to emerging stresses (Hura et al. 2009). Additionally, the defense responses of grass pea were stronger than those of garden pea when subjected to drought stress, including the up-regulated expression of genes encoding antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX), and glutathione peroxidase (GPX), an enhanced antioxidant defense system, and a high content of two osmoprotectants, proline and soluble sugars (Jiang et al. 2013).

Compared to other legumes, grass pea has developed several morphological drought tolerance traits as adaptive mechanisms, including narrow leaves, winged stems, and a deep and extensive root system. Therefore, under drought stress, the grass pea performance is closely related to the root system development and root system distribution (Bultynck et al. 2004; Franks 2013; Koevoets et al. 2016). Indeed, the greater the root depth and root biomass, the more efficiently any available soil moisture is extracted (Blum 2011; Fenta et al. 2014). Moreover, to avoid drought, the plant matures early, a characteristic that can prevent any terminal desiccation. To forestall dehydration, both the green leaf area can be reduced, keeping a higher water status for photosynthesis maintenance, and the stomatal openings can be diminished to minimize the transpiration rate (Cullis and Kunert 2017). Under stress, grass pea also decreases its flower and pod production and concentrates its resources in a small number of surviving pods (Gusmao et al. 2012). Noteworthy, transpiration and other physiological processes in drying soil do not begin to decrease immediately after water is withheld, but only after a threshold soil water content has been reached (Kong et al. 2015). Grass pea, unlike garden pea, curls or rolls its leaves to diminish water loss through transpiration, which is an evolutionary adaptation (Jiang et al. 2013) identified as one of the most common responses to water deficit stress in some plant species (Turgut and Kadioglu 1998). Moderate leaf curling or rolling helps to improve photosynthetic efficiency, accelerates dry-matter accumulation, increases yield, reduces the solar radiation on leaves, and decreases leaf transpiration under drought stress (Lang et al. 2004).

Variety amelioration research

Grass pea has a tremendous potential as a source of multiple stress-tolerant genes for general crop improvement with suitable breeding strategies (Hao et al. 2017; Wang et al. 2015). For example, donor germplasm with the desirable phenotypes can be used to form new breeding materials and available genomic tools can be applied to speed up the breeding process. However, when compared to other legume crops, the genomic resources for grass pea are still scarce because of its big complex 8.2-Gb genome (Bennett and Leitch 2012). In addition, biotechnological investments remain limited. Hence, to date, very limited research efforts have been devoted to genetic improvement of the crop. Grass pea is a diploid with a chromosome number of 2n = 14. Moreover, it has a very slow breeding process (Hao et al. 2017), possibly due to a very narrow range of genetic variations, resulting from self-pollination and interspecific incompatibility (Nerkar 1976). Increasing the genetic erosion presents a challenge to sustain its genetic diversity, which provides the basis for all plant improvements.

Therefore, there is still a pressing need to collect, conserve, and characterize the genetic variation of grass pea present across diverse geographical regions by means of both ex situ (for instance, seed storage in gene banks) and in situ (i.e., within the natural habitats) techniques (Vaz Patto and Rubiales 2014). For instance, the International Center for Agricultural Research in Dry Areas (ICARDA) in Syria has collected and conserved nearly 3600 accessions, representing 45 wild Lathyrus species from more than 40 countries (Dixit et al. 2016). Since 2016, 2720 Lathyrus accessions are conserved in collections of the National Bureau of Plant Genetic Resources (NBPGR) in New Delhi (Dixit et al. 2016). For details on other important ex situ Lathyrus collections, the reader is referred to Vaz Patto and Rubiales (2014). In addition to ex situ collections, in situ preservation is recommended, because this method safeguards both the population and the evolutionary processes that enable the population to adapt to changing environments by management of the organisms in their natural state or within their normal range. The potential for a high-level improvement exists within these materials, because high variability found in the primary gene pool within L. sativus accessions (Vaz Patto and Rubiales 2014; Abate et al. 2018; Khillari et al. 2018) allows the enhancement of yield, quality, and adaptability. In the different collections, the Lathyrus accessions have been identified for agronomically/economically important traits, including earliness (< 100 days), number of pods (> 50 per plant), grain yield (> 10 g/plant), seed size (> 12 g/100 seed), and β-ODAP content (< 0.2%) (Dixit et al. 2016). In particular, biotic and abiotic stress tolerances and nutritional and technological qualities are also important target traits to be characterized in the Lathyrus germplasm (Vaz Patto and Rubiales 2014). Broadly, the grass pea germplasms can be divided into two groups: the Asian and Mediterranean groups with small and large seeds and on average a high and low β-ODAP content, respectively (Fig. 2).

Fig. 2
figure 2

Diversity of Lathyrus sativus seed varieties. Broadly, grass pea germplasms can be divided into two groups: the Asian and Mediterranean groups with small and large seeds and on average a high and low β-ODAP content, respectively, a Bangladesh, b China, c Ethiopia, d Canada, e India, f Nepal, g Portugal, h Poland, i China (scale: 5 mm)

Full size image

Moreover, the results of interspecific hybridization in grass pea offer many opportunities for improvement through the identification and transfer of desirable traits from exotic and wild germplasm (Yunus and Jackson 1991). Taxonomic relationship studies have shown that only two species, L. amphicarpos and L. cicera, produce viable hybrids when crossed with L. sativus (Yunus and Jackson 1991). Yet, to overcome strong reproductive barriers among different species, biotechnological tools, including tissue culture, somaclonal variation, and protoplast fusion techniques may also be utilized (Ochatt et al. 2001; Piwowarczyk and Pindel 2015; Tripathy et al. 2016).

Historically, the crop improvement of grass pea followed three phases. A first phase lasted for almost 20 years (1940–1960) and focused on yield improvement (Gautam et al. 1998) with the development of several cultivars recommended for cultivation. A second phase (1974–1990s) involved the development of improved varieties with low β-ODAP content. Evaluation of grass pea collections worldwide has revealed a considerable variation in β-ODAP content of the seeds, ranging from 0.02 to 2.59% (Kumar et al. 2011), but no β-ODAP-free plants have ever been identified in either germplasm or wild species (Emmrich 2017; Xu et al. 2017). Whereas the highest β-ODAP concentrations are measured in some of the Indian germplasm (Nagarajan and Gopalan 1968), the lowest level of β-ODAP (0.02%) has been reported in Australian germplasm from ICARDA (Hanbury et al. 1999). Several improved grass pea cultivars, all with a β-ODAP content < 0.1%, have been released as the result of various national and international breeding initiatives (Ali-Bar in Kazakhstan; Ceora and Chalus in Australia; Gurbuz 1 in Turkey; Wasie in Ethiopia; Prateek, Mahateora, Ratan, and Pusa 24 in India; Bari Khesari 1, Bari Khesari 2, and Bina Khesari 1 in Bangladesh) (Abd El Moneim et al. 2001; Kumar et al. 2011). In a third phase (1990s onwards), several varieties and lines were developed that combined low β-ODAP (< 0.1%) with high-yield potential (up to 1.5 tons/ha) and resistance to a variety of biotic and abiotic stresses (Kumar et al. 2013).

Although a number of cultivars with a low β-ODAP have been released, the long-term results of these efforts are often questioned. The β-ODAP content is highly influenced by climatic and edaphic conditions and has a strong genotype × environment effect (Tadesse 2003; Fikre et al. 2011b; Jiao et al. 2011; Girma and Korbu 2012). For instance, the β-ODAP content of the Ethiopian variety Wasie has increased from < 0.08 to > 0.2% from the laboratory to field tests (Fikre et al. 2011b). Various factors that affect β-ODAP accumulation in grass pea include the plant growth stage, nutrients, and abiotic stresses, such as drought, salinity, water, and heavy metals (Jiao et al. 2011). Indeed, water stress can double the β-ODAP content in the plant (Hanbury et al. 1999; Xing et al. 2001) and application of zinc sulphate to the soils has been suggested to reduce the β-ODAP content by 10–40% (Lambein et al. 1994; Kumar et al. 2011), but the mechanism is still unknown. Grass pea tends to synthesize more of this neuroexcitatory amino acid when stressed than when it grows under optimal conditions (Abd El Moneim et al. 2001).

Pest and disease resistance

One of the advantages of grass pea is that it does not suffer much from pests and diseases (Campbell 1997). In fact, the Lathyrus gene pool offers a source of resistance to important legume diseases caused by fungi, such as Ascochyta blight (Mycosphaerella or Didymella pinodes) (Robertson and Abd El Moneim 1996; Gurung et al. 2002), downy mildew (Peronospora lathyri-palustris), and powdery mildew (Erysiphe spp.) (Vaz Patto et al. 2006b). Only in South Asia is downy mildew an important disease of grass pea and in parts of Ethiopia are rust diseases (Uromyces spp.) problematic (Campbell 1997). There are many more reports on the biotic than on the abiotic stress resistance evaluation of Lathyrus germplasm collections. Studies under natural infection conditions identified promising lines without visible symptoms of downy mildew (Narsinghani and Kumar 1979; Asthana and Dixit 1998; Pandey et al. 1998) and lines with a moderate resistance to powdery mildew (Lal et al. 1985; Gautam et al. 1998; Vaz Patto et al. 2006b). Lines with an exceptional resistance or immunity to leaf spot Cercospora pisi sativae were also characterized (Mishra et al. 1986). Screenings of a set of 50 Lathyrus accessions for rust disease established the presence of high resistance in all accessions against Uromyces viciae-fabae and Uromyces ciceris arietini, and a partial resistance to Uromyces pisi (Vaz Patto and Rubiales 2014). Besides, Sclerotinia sclerotiorum, a necrotrophic fungal pathogen, instigates white mould on more than 400 plant species, including important crops, such as cotton (Gossypium hirsutum), tomato (Solanum esculentum), sunflower (Helianthus annuus), soybean, and grass pea (Heller and Witt-Geiges 2013). Oxalic acid (OA) is considered as a pathogenesis factor for this devastating phytopathogenic fungus that synthesizes and secretes OA during host colonization (Malenčić et al. 2010; Kabbage et al. 2015; Uloth et al. 2015).

Wild Lathyrus species, including L. ochrus, L. clymenum (Sillero et al. 2005), and L. cicera (Fernández-Aparicio et al. 2009; Fernández-Aparicio and Rubiales 2010) are resistant to broomrape (Orobanche crenata), a common parasitic weed to the faba bean, a trait that is absent in the cultivated gene pool. In addition to its reduced β-ODAP content (varying between 0.07 and 0.51% across 142 accessions), L. cicera may be a promising source of other important agronomic traits, including earliness and cold tolerance (Robertson and Abd El Moneim 1996).

Moreover, the root knot nematode (Meloidogyne artiella) and cyst nematode (Heterodera ciceri) can attack grass pea, but resistance to both has been identified at ICARDA (Campbell 1997). Thrips (Caliothrips indicus) are a pest of high economic concern for grass pea cultivation, in particular in India (Dixit et al. 2016). Screening of 56 accessions against thrips at Raipur revealed two lines with moderate resistance (Asthana and Dixit 1998; Pandey et al. 1998). Resistant Lathyrus genotypes may serve as a source of new and useful genetic traits in the breeding of related major legume crops, such as peas, lentils, and vetches.

Molecular toolbox for variety amelioration

The genetic potential of grass pea could be exploited to enhance food and nutrition security, to provide feed for livestock, and to improve soil fertility as green manure in harsh drought-prone environments. Regardless of the availability of low toxin lines, listing grass pea as a toxic plant and banning seed sales in some countries have considerably limited funding for genetic improvement of grass pea through modern breeding techniques as well as genetics and omics tools. As a consequence, sparse genomic resources exist today for this orphan legume, hampering its potential exploitation in crop breeding (Hao et al. 2017).

Mutagenesis

Improvement strategies focusing on mutation breeding have been explored as a valuable addition to conventional breeding. As a mutation is defined as a change in the heritable constitution of an organism, mutation breeding increases the genetic variability, thus widening the scope on the isolation and characterization of novel genetic variants (Girma and Korbu 2012). Mutation breeding in grass pea has generated several viable diploid mutants with altered plant characters, such as growth habit, branching, stem and internodes, leaflets, stipules, flower colour, pod features, and seed traits (Talukdar 2009). For instance, mutants were generated with erect, determinate, and semi-dwarf characteristics, with also an increased number of primary and secondary branches, along with a high grain yield and low seed β-ODAP content (Talukdar and Biswas 2006). These mutants have been used to identify and map the desirable morphological markers on specific grass pea chromosomes (Talukdar 2009). Through induced mutagenesis, a robust cytological stock was developed as well, comprising trisomics, tetrasomics, double trisomics, various reciprocal translocations, and autotetraploid lines (Talukdar 2012a). Similarly to morphological changes, marked biochemical changes have been introduced into grass pea by mutagenic treatments. Most mutations have been reported to contribute directly or indirectly to the plant antioxidant defense systems. A glutathione (GSH)-deficient mutant (gshl-1) revealed a greater sensitivity to cadmium than its mother plants (Talukdar 2012a). By contrast, grass pea mutants were developed with an enhanced tolerance to salt and arsenic toxicity (Talukdar 2011, 2013). In grass pea, arsenic exposure significantly reduces seed germination, inhibits photosynthetic activities and plant growth, but significantly enhances the level of the seed neuroexcitatory β-ODAP (Talukdar 2012b, 2013).

Molecular markers development

Breeder-friendly genomic tools become more and more available to improve the efficiency of breeding protocols in grass pea (Bohra et al. 2014). Unlike morphological descriptors, tools, such as molecular markers, detect diversity and variation among and within species directly at the DNA level and independently of environmental factors. They can be used to determine number, position, and individual effects of genes/quantitative trait loci (QTLs) that control interesting traits, such as disease or pest resistance, reduced β-ODAP concentration, increased protein content, and other characteristics of agronomic importance, via, for instance, genetic mapping and QTL analysis (Campbell et al. 1993).

Genetic diversity in the Lathyrus genus can be detected by means of diverse molecular markers, including, among others, restriction fragment length polymorphism (RFLP), random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP), and sequence-related amplified polymorphism (SRAP) (Chtourou-Ghorbel et al. 2001; Marghali et al. 2016; Nosrati et al. 2012; Tavoletti and Iommarini 2007). Another widely used class of molecular markers are simple sequence repeat (SSR) or microsatellite markers that can be obtained by searching in the expressed sequence tags (ESTs) available at the National Center for Biotechnology Information (NCBI) GenBank for Lathyrus (http://www.ncbi.nlm.nih.gov/genbank/dbest/dbest_summary/) (Lioi et al. 2011; Ponnaiah et al. 2011; Shiferaw et al. 2012; Sun et al. 2012; Lioi and Galasso 2013; Soren et al. 2015; Gupta et al. 2018). Today, whereas the number of ESTs is limited for L. sativus and L. cicera (178 and 126, respectively) at the NCBI GenBank, many more ESTs (8702) are available from L. odoratus. This EST-SSR marker system generally has a high degree of conservation and can be transferred among species. Thus, cross-species and cross-genus amplification of molecular markers facilitate comparative genomic mapping, providing an alternative for the development of new molecular markers for orphan species (Gutierrez et al. 2005). Hence, they have allowed the use of Lathyrus as a source of interesting traits for other related species and, vice versa, the increase in the number of molecular markers available for Lathyrus species, in particular for L. cicera and L. sativus (Chandra 2011; Lioi et al. 2011; Shiferaw et al. 2012; Almeida et al. 2014a). For instance, a large number of molecular markers from Medicago truncatula (barrel medic), garden pea, lentil, lupine, and faba bean were shown to be transferable to L. cicera and L. sativus for their future applications in mapping and diversity studies (Almeida et al. 2014a).

Thus far, two linkage maps have been developed for L. sativus by means of molecular markers (Chowdhury and Slinkard 1999; Skiba et al. 2004). One of these maps was used to locate two QTLs, that explained the relatively small percentage of observed variation in the resistance against Ascochyta blight (Skiba et al. 2004), but no candidate genes were identified at that time for these resistance QTLs, hampering their use in precision breeding. Moreover, these two linkage maps were not adequately saturated with markers and presented many gaps and short linkage groups, so that they could not be aligned and compared with linkage maps of other legume species (Almeida et al. 2015).

Recently, the application of next-generation sequencing (NGS) to grass pea has provided a means to exploit the repertoire of genomic resources for the orphan legume Lathyrus. NGS platforms were used to generate large-scale SSR-enriched sequence data and develop SSR markers that will facilitate the construction of high-resolution maps for positional cloning and QTL mapping (Yang et al. 2014; Wang et al. 2015). Thanks to the sensitive and quantitative high-throughput transcriptome analysis, referred to as RNA-sequencing technology, new genetic markers could be identified for some orphan legumes, such as Bituminaria bituminosa (Arabian pea or pitch trefoil) (Pazos-Navarro et al. 2011). The availability of new tools for both the de novo assembly of transcriptomes and their alignment to newly assembled genomes may help identify differential gene expression in response to stress, for instance. In 2014, by the RNA-sequencing technology, the first high-throughput transcriptome assemblies were generated from control and with rust (Uromyces pisi)-inoculated leaves of susceptible and rust-resistant grass pea genotypes (Almeida et al. 2014b). The study generated a large number of new gene-based molecular tools, e.g., ESTs, EST-SSRs, and single nucleotide polymorphism (SNP)-based markers. These markers will be instrumental for future work on high-throughput mapping of the genetic basis of disease resistance in L. sativus and, eventually, comparative mapping with other legume species. Meanwhile, this transcriptome analysis provided a comprehensive insight into the molecular mechanisms underlying rust resistance in L. sativus. The differences in rust resistance between the two (resistant and susceptible) genotypes seemed to be mainly due to the activation of several pathogenesis-related genes, including the “mildew resistance locus O (MLO)-like protein” involved in signalling in response to biotic stresses. The MLO-encoding gene, first identified in barley (Hordeum vulgare), conferred resistance to powdery mildew (Jørgensen 1992). Later, by coupling high-throughput sequencing (Illumina) technology with serial analysis of gene expression (SAGE) analyses, a set of differentially expressed genes was identified in the leaves of L. sativus in response to Ascochyta lathyri inoculation (Almeida et al. 2015). Functional categorization associated these genes with processes involved in biotic/abiotic stress, cell wall metabolism, and hormone signalling (Almeida et al. 2015). Very recently, the L. cicera transcriptome was investigated in response to rust (Uromyces pisi) and different types of molecular markers were developed to construct the first L. cicera linkage map with potential for genetic mapping of resistance in this robust species and its most closely related species L. sativus (Santos et al. 2018).

Regeneration and genetic transformation

To exploit the potential of grass pea as a source of multiple stress-tolerant genes for general crop improvement through modern approaches, such as genetic transformation, good in vitro culture and regeneration protocols are a first prerequisite. Legumes, and grass pea in particular, are very challenging to regenerate and transform, because of the problematic somatic embryogenesis or organogenesis (Iantcheva et al. 2013). After many attempts, a regeneration protocol of fertile plants from meristematic tissues was established for L. sativus in 2002 (Zambre et al. 2002). Using this regeneration protocol, Ethiopian grass pea could be transiently and genetically transformed for the first time using Agrobacterium-mediated transformation (Girma 2010). Here, in an effort to boost the seed protein quality, the genetic transformation was conducted with a common bean (Phaseolus vulgaris L.) gene coding for additional amounts of methionine, but without further analyses (Girma 2010). In contrast, a genetic transformation procedure was designed with epicotyl segments of Indian L. sativus varieties co-cultivated with Agrobacterium and expressing both a reporter gene (β-glucuronidase) and a selectable marker gene (neomycin phosphotransferase II) (Barik et al. 2005). Unfortunately, only one event could be stably transformed in the progenies. Similarly, currently, only one broadly applicable virus-induced gene silencing (VIGS) reverse genetics approach has been reported for L. odoratus that resulted in a bleaching phenotype, hinting at a down-regulation of the phytoene desaturase (PDS) gene expression (Grønlund et al. 2008).

Recently, in vitro direct multiple shoot induction was obtained from mature seed embryo with two cotyledons as explants. This protocol has been shown to be a time-saving approach for regenerating whole plants from this type of explant. It by-passes the callus induction phase and thus decreases somaclonal variations usually seen in plants produced by tissue culture (Barpete et al. 2017). It can be helpful in future genetic transformation and biotechnology-based breeding of grass pea. Meanwhile, soybean and grass pea genetically engineered by Agrobacterium-mediated transformation had an improved nutritional quality and an enhanced tolerance to fungal pathogens, without any adverse impact on plant growth, development, and seed protein quality (Kumar et al. 2016). The reduced level of the antinutritional metabolite oxalic acid (OA) in transgenic seeds of soybean (up to 73%) and grass pea (up to 75%) was observed by the constitutive and/or seed-specific expression of an oxalate-degrading enzyme, oxalate decarboxylase (FvOXDC) of the fungus Flammulina velutipes. In addition, the β-ODAP level of grass pea seeds had also decreased up to 73%. In transgenic grass pea lines, the reduced OA content was interrelated with the associated increase in seed micronutrients, such as calcium, iron and zinc, manganese, and magnesium. Moreover, the constitutive expression of FvOXDC led to an improved tolerance against the fungal pathogen Sclerotinia sclerotiorum that requires OA during host colonization (Kumar et al. 2016). Although the genotype used was not mentioned in the study, this recent genetic transformation protocol for grass pea, if successfully reproducible, may help future proof-of-function studies of putative Lathyrus genes via overexpression, deletion, or silencing.

Conclusions

Grass pea is a plant with a combination of exceptional agronomic properties, such as tolerance to drought, flooding, salinity, a high nitrogen fixation capacity that increases the yield of subsequent crops, easy cultivation with minimal inputs, and adaptability to different climates and soils. As a legume, it is also a highly nutritious food and fodder crop. Considered one of the most climate-resilient crops, grass pea is the only crop that survives in the field during severely dry years, providing no alternative to many low-income communities but to overconsume the seeds as the only dietary component. Uncounted thousands of people have survived famines thanks to grass pea, hence deemed a life-saver crop. However, overconsumption of grass pea has caused the crippling disease neurolathyrism in many European countries in historical times and more recently in Asia and Africa. The forced overuse of grass pea in diets during severe famines has given it its present undeserved reputation of being a toxic plant. Due to this stigma, investments to improve Lathyrus varieties are considerably limited and, consequently, sparse genomic resources exist, hampering its potential exploitation in crop breeding (Hao et al. 2017).

Although drought-caused famine has not happened in the last few years, climate change and, particularly, global warming are a serious source of concern for the sustainability of the agricultural production and the food security worldwide, thus, calling for the development of climate-smart crops. Grass pea holds a tremendous potential as a source of multiple stress-tolerant genes for general crop improvement (Hao et al. 2017). The basis of the high tolerance to drought and salinity should be further investigated and the crop could be used as an elite germplasm resource for traits to adapt the world’s most important crops to new climatic conditions (Boukecha et al. 2017). Moreover, the high variability found in the primary gene pool within the Lathyrus accessions allows the improvement of yield, nutritional quality, and adaptability of grass pea itself. Modern approaches and, especially the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) or CRISPR-associated protein 9 (CRISPR/Cas9) system will enable genome editing in grass pea, both as a research tool for the engineering of signalling pathways or regulatory mechanisms involved in the protection of grass pea against drought and as a mean to accelerate crop improvement (Emmrich 2017).

For more than 3 decades no new cases of neurolathyrism have been reported (Singh and Rao 2013). Importantly, when neurolathyrism occurred, it was restricted to rural areas among very poor people, living as subsistence farmers on marginal lands. This socio-economic disease can be considered as alarming signals of extreme poverty and food insufficiency. The low incidence of clinical symptoms caused by neurolathyrism among grass pea consumers and in experimental animals was first explained by the individual variability, without considering the total diet. Only afterwards, deficiency in the essential amino acids methionine and cysteine was taken into account. These sulfur-containing amino acids play important roles in the redox homeostasis of the cell and protect against oxidative stress. Addition of methionine to grass pea as chicken feed shielded against neurological symptoms (Fikre et al. 2010), whereas sufficient cereals supplemented to a grass pea diet is a protective factor in the neurolathyrism epidemiology (Getahun et al. 2003). An old experiment by Justus von Liebig (reported by Pasero 2017), in which a soldier ate solely garden peas for several months, resulted in neurological symptoms described as “erbsenwahn” or pea madness. Although this ancient experiment on a single subject needs confirmation, it implies that legumes considered wholesome should not be used as exclusive diet.

In view of the climatic change pressure, this plant could be made into a wonder crop for drought-prone and marginal lands. To this end, grass pea should be used properly in a well-balanced diet and research investments should be increased to improve the nutritional quality of seeds with focus on the essential amino acid balance. The past attempts to strive for zero or low β-ODAP content in grass pea caused researchers to look for other improvement options (Girma and Korbu 2012). Advances in genomics and genetic engineering could help boost the seed protein quality and enhance the content of essential amino acids. In the near future, exploitation of the non-neurotoxic potentials of β-ODAP as nutraceutical and development of economically important traits in grass pea, such as tolerance to biotic and abiotic stresses, for general crop improvement will presumably take the stage. Moreover, instead of focusing on the reduction of β-ODAP content, research efforts to develop grass pea varieties with increased methionine and cysteine contents may be more relevant to improve the nutritional value without loss of its hardy nature. Additionally, other aspects should be considered, such as the very high nitrogen-fixing capacity and the potential role played by β-ODAP. The presence of homoarginine in grass pea seeds is hardly mentioned in the literature, albeit its positive effect on cardiovascular health (Rao 2011). The impact of a regular and moderate consumption of grass pea on longevity and strong cardiovascularity should be a priority in epidemiological research.

Author contribution statement

All authors contributed to the design of the article. FL and ST mainly wrote the paper. All authors gave input and approved the final version of the manuscript.