Plant Foods Hum Nutr (2008) 63:105–109 DOI 10.1007/s11130-008-0078-8

ORIGINAL PAPER

Genetic Diversity for Seed Mineral Composition in the Wild Teramnus labialis

Michael A. Grusak

Published online: 19 June 2008 # Springer Science + Business Media, LLC 2008

Abstract Teramnus labialis (L.) Spreng. is a wild, tropical protein, energy (in the form of carbohydrates or lipids), legume whose seeds are collected and used as a food source several vitamins, and many human essential minerals. by tribal populations. In order to assess the potential of this Various legume species have been harnessed for large-scale legume to provide dietary minerals for humans, fourteen cultivation in diverse climactic zones; however, several diverse accessions were grown under controlled, nutrient- wild legume species are also collected and consumed on a replete conditions and seeds were harvested for mineral smaller scale by rural or tribal populations [3–6]. The analysis. The germplasm originated from Indonesia, , potential conversion of these wild into cultivated the Caribbean, and South America. Seed concentrations of crops has received considerable attention in recent years, phosphorus (P), potassium (K), sodium (Na), iron (Fe), especially as efforts have expanded to increase the mineral copper (Cu), manganese (Mn), and zinc (Zn) were found to concentration of seed foods for humans [7]. Populations fall within the range of published values for several that are at-risk for micronutrient mineral deficiencies (e.g., cultivated grain legumes, while calcium (Ca) and magne- Fe, Zn, I) or macronutrient mineral deficiencies (e.g., Ca) sium (Mg) were higher in T. labialis seeds. Mineral could be helped with nutrient-dense, biofortified legume concentrations across the diverse accessions showed ranges and cereal grains [8]. Recent studies have assessed the of 1.3- to 2.3-fold for the macronutrient minerals (Ca, Mg, genetic diversity of seed mineral concentrations in several P, K) and 1.8- to 15.9-fold for the micronutrient minerals cultivated crops, in order to use selected lines for breeding (Fe, Cu, Mn, Zn, and Na). The existing genetic diversity in or investigative research purposes [9]. Germplasm of wild, this wild legume, especially for the essential minerals Ca edible species should also be studied [10], thereby and Mg, could be exploited to develop T. labialis as a new providing opportunities for developing these legumes either cultivated legume for tropical regions of the world. as new food crops themselves, or as biological models for the discovery of genes relevant to seed mineral accretion. Keywords Calcium . Germplasm . Magnesium . One such wild legume that has received only limited Seed minerals . Wild legume attention as a grain crop is Teramnus labialis (L.) Spreng. This legume grows in tropical areas of Asia, Africa, and the Americas [11], and has recently been reported to be Introduction consumed as a seed food by tribal sects in South [12]. As with several other legumes, T. labialis is used as a Legume seeds are consumed worldwide as an important forage crop in certain parts of the world [13, 14]; thus, its source of dietary nutrients for humans [1, 2]. They provide current utilization by rural populations as forage makes it a good candidate for expansion into a cultivated, human food M. A. Grusak (*) source. Interestingly, the limited nutritional data on T. USDA/ARS Children’s Nutrition Research Center, labialis suggests that it is a very good source of essential Department of Pediatrics, Baylor College of Medicine, minerals, including Ca, Mg, and K. In particular, the 1100 Bates Street, Houston, TX 77030, USA reported seed concentration for Mg (812 mg/100 g dry e-mail: [email protected] weight) [12] is one of the highest seed Mg values we have 106 Foods Hum Nutr (2008) 63:105–109 found in the literature, and is much higher than that of solution to each pot, three times a day. Sufficient solution existing, cultivated legume (or cereal) crops [15]. was delivered to ensure soil saturation; excess solution was In order to gain a better understanding of the seed allowed to drain from the pots. The nutrient solution mineral composition of this wild legume, fourteen diverse contained the following concentrations of mineral salts:

T. labialis germplasm accessions were grown for seed 1.0 mM KNO3, 0.4 mM Ca(NO3)2, 0.1 mM MgSO4, harvest and analysis. were maintained under con- 0.15 mM KH2PO4,25μM CaCl2,25µMH3BO3,2µm trolled environmental conditions, with continuous mineral MnSO4, 2 µM ZnSO4, 0.5 µM CuSO4, 0.5 µM H2MoO4, fertilization used to prevent mineral inadequacies. The data 0.1 µM NiSO4, and 1 µM Fe(Ш)-N, N′-ethylenebis[2-(2- from the diverse accessions are presented and compared hydroxyphenyl)-glycine] (Sprint 138; Becker-Underwood, with seed values for various cultivated and wild legumes. Inc., Ames, Iowa, USA). The potential for developing T. labialis into a cultivated The environmental conditions within the greenhouse legume is also discussed. were a temperature regime of 22±3 °C day and 20±3 °C night, with relative humidity ranging from 45% to 65% throughout the day/night cycle. Sunlight was supplemented Materials and Methods with metal halide lamps set to a 15-h day, 9-h night photoperiod (lights on at 700 h), and a minimum intensity Plant Material and Growth Conditions (i.e., lamps only) of photosynthetically active radiation of 200 μmol photons m−2 s−1 at the top of the plants. Fourteen accessions of T. labialis (L.) Spreng. () were used in this study (Table 1). These were obtained from Seed Harvest and Tissue Analysis the USDA germplasm collection (USDA/ARS Plant Ge- netic Resources Conservation Unit, Griffin, Georgia, USA; Plants were grown for up to 15 months, which allowed all accessions labeled with PI numbers) or from the perennial accessions time to reach reproductive age and to enable Glycine germplasm collection (University of Illinois, adequate pod production for seed collection. Seeds were Champaign-Urbana, Illinois, USA) (both the CIAT identi- harvested as pods matured and were combined from all six fiers and the Champaign-Urbana [CU] identifiers are plants per accession. Seeds were removed from pods, provided). extraneous material was removed, and seeds were dried Plants were grown in 5-l black plastic pots filled with a for a minimum of 3 days in a 60 °C, forced-air drying oven. 2:1 (v/v) mixture of synthetic soil (Metro-Mix 360; Scotts- Average seed weights were calculated by averaging the Sierra Horticultural Products Co., Marysville, Ohio, USA) weight of three sets of 100 random seeds for each and vermiculite (Strong-Lite Medium Vermiculite, Sun Gro accession. Horticulture Co, Seneca, Illinois, USA). Plants of each For seed mineral analysis, a minimum of 3 g of dried accession were grown in two pots, with plants thinned to seeds for each accession was homogenized. Three aliquots three seedlings per pot, 4 d after emergence. Pots were (0.25 g each) of each accession were digested at 125 °C randomly assigned positions in a greenhouse. An automat- (3 h) and 200 °C (digestate taken to dryness) using trace- ed drip irrigation system was used to deliver nutrient metal grade hydrogen peroxide and nitric acid. Digestates

Table 1 Source country and mean seed weights of Accession identifier Country of origin Mean seed weight (mg/100 seeds±standard error) Teramnus labialis germplasm used in this study CIAT 926, CU 409 Antigua and Barbuda 650±1 CIAT 4986, CU 411 Indonesia 615±2 CIAT 7442, CU 412 Cuba 745±5 CIAT 20072, CU 416 Colombia 267±0 PI 200233 Kenya 668±7 PI 277511 Virgin Islands 758±4 PI 365055 South Africa 841±5 PI 365056 South Africa 824±5 PI 365057 South Africa 796±2 PI 406170 Kenya 698±3 PI 490301 Kenya 763±4 PI 517204 Ethiopia 1,250±4 PI 538317 Virgin Islands 706±0 PI 538318 Virgin Islands 953±7 Plant Foods Hum Nutr (2008) 63:105–109 107 were resuspended in 2% ultra-pure nitric acid and analyzed from 1.74 to 11.99 mg/100 g dry weight; Zn varied 1.9- for Ca, Mg, K, P, Fe, Zn, Mn, Cu, and Na concentrations. fold, ranging from 4.48 to 8.53 mg/100 g dry weight; and Elemental analysis was performed using inductively cou- Na varied 6.1-fold, ranging from 0.48 to 2.95 mg/100 g dry pled plasma–optical emission spectrometry (CIROS ICP weight. Model FCE12; Spectro, Kleve, Germany). Certified rice flour standards (SRM 1568A; National Institute of Stand- ards and Technology, Gaithersburg, Maryland, USA) were Discussion digested and analyzed along with samples to verify the reliability of the procedures and analytical measurements. Although grown in a greenhouse with somewhat atypical conditions (i.e., non-tropical temperatures or humidity), all fourteen T. labialis accessions grew well and produced Results many pods in this study. Seed weights (Table 1) were also indicative of this species, suggesting that the plants were Fourteen T. labialis accessions obtained for this study healthy and producing normal seeds. Mineral nutrients were represented germplasm collected from Asia, Africa, the provided to the plants on a daily basis with the intent of Caribbean, and South America (Table 1). Inspection of leaf ensuring complete, non-limiting nutrient conditions. This characters, growth habit, pod phenotype (data not shown), was done to assess the full genetic potential of each accession and seed weight (Table 1) suggested that all the material with respect to the transport of minerals to developing seeds. was T. labialis germplasm, as described by Verdcourt [11]. All plants appeared healthy throughout the study with no Pods of all plants exhibited the elongated hook that is evidence of leaf chlorosis or necrotic lesions. characteristic of Teramnus species [11]). Seed weights, Seed mineral concentrations were highly variable across which ranged from 267 to 1,250 mg/100 seeds, also were the accessions (Table 2), reflecting the presumed diversity typical for this species. of this germplasm that originated from several regions of Mineral concentrations of the harvested, mature seeds the world (Table 1). The concentration ranges, being greater showed a high degree of variation across the accessions for the micronutrients than the macronutrients, are consis- (Table 2). Among the macronutrient minerals, Ca varied tent with concentration ranges reported in earlier germ- 2.3-fold, ranging from 250 to 575 mg/100 g dry weight; Mg plasm studies for Phaseolus vulgaris [16] or soybean [17]. varied 1.9-fold, ranging from 263 to 489 mg/100 g dry A comparison of the range of mineral values obtained in weight; K varied 1.3-fold, ranging from 1,038 to 1,316 mg/ this study, with mineral values of the previous report for 100 g dry weight; and P varied 1.7-fold, ranging from 339 seeds of T. labialis collected from a forest habitat [12], to 584 mg/100 g dry weight. Among the micronutrient show similarities for some minerals (Ca, P, and Cu), higher minerals, Fe varied 1.8-fold, ranging from 5.25 to 9.21 mg/ concentrations for a few minerals (Fe, Mn, and Zn), and 100 g dry weight; Cu varied 15.9-fold, ranging from 0.15 to lower to much lower levels for the others (Mg, K, and Na). 2.39 mg/100 g dry weight; Mn varied 6.9-fold, ranging With respect to the micronutrients Fe, Mn, and Zn that were

Table 2 Mineral concentrationsa in mature seeds of diverse Teramnus labialis germplasm

Accession Ca Mg K P Fe Cu Mn Zn Na

CIAT 926, CU 409 474±7 486±5 1,096±5 406±2 7.01±0.16 0.38±0.01 2.49±0.07 6.01±0.02 1.12±0.14 CIAT 4986, CU 411 515±3 402±3 1,122±3 526±2 8.83±0.02 0.70±0.01 8.19±0.12 8.13±0.12 1.26±0.21 CIAT 7442, CU 412 441±10 314±5 1,260±8 339±4 5.67±0.05 0.45±0.01 4.56±0.08 5.29±0.11 0.94±0.07 CIAT 20072, CU 416 287±2 343±3 1,038±2 422±1 8.43±0.14 0.45±0.01 2.96±0.10 5.84±0.10 0.94±0.15 PI 200233 508±8 383±2 1,316±2 464±2 6.92±0.11 0.18±0.00 1.96±0.01 4.78±0.03 2.95±0.07 PI 277511 517±4 489±4 1,075±12 418±3 7.57±0.18 0.50±0.01 2.23±0.03 6.86±0.07 1.61±0.28 PI 365055 384±7 293±3 1,255±9 490±4 5.25±0.11 0.23±0.01 1.90±0.04 5.66±0.05 1.06±0.10 PI 365056 386±7 275±2 1,241±13 497±4 5.88±0.09 0.15±0.00 2.01±0.01 5.36±0.04 0.98±0.07 PI 365057 374±16 263±1 1,254±23 479±15 5.45±0.04 0.17±0.00 1.74±0.09 5.21±0.19 1.17±0.24 PI 406170 448±11 343±4 1,171±7 416±1 6.69±0.13 0.52±0.01 6.16±0.27 5.99±0.07 0.48±0.04 PI 490301 575±7 384±3 1,191±10 526±2 8.87±0.14 1.35±0.03 4.86±0.11 6.57±0.01 1.78±0.24 PI 517204 250±2 288±4 1,121±15 584±7 7.58±0.11 0.62±0.02 7.72±0.29 8.53±0.15 1.20±0.32 PI 538317 502±4 444±4 1,205±9 479±2 9.21±0.17 2.39±0.00 11.99±0.64 6.69±0.07 1.20±0.17 PI 538318 437±12 346±4 1,303±4 445±5 6.09±0.09 1.14±0.02 3.14±0.11 4.48±0.05 1.02±0.12 a Mean values (mg/100 g dry weight±standard error) determined from three samples 108 Plant Foods Hum Nutr (2008) 63:105–109

Table 3 Mean seed mineral concentrationsa in Teramnus labialis compared to several cultivated grain legumes

Mineral Teramnus labialisb Teramnus labialisc Lentild Chickpeae Black beanf Soybeang (Lens culinaris) (Cicer arietinum) (Phaseolus vulgaris) (Glycine max)

Ca 436±24 521 50±1 93±3 109±10 253±5 Mg 361±20 812 109±2 102±4 152±2 256±8 K 1,189±23 2,167 856±16 774±26 1,320±0 1,644±26 P 464±17 336 404±8 324±11 313±12 644±10 Fe 7.1±0.4 3.4 6.8±0.2 5.5±0.1 4.5±0.5 14.4±0.7 Cu 0.7±0.2 0.9 0.5±0.0 0.8±0.0 0.8±0.0 1.5±0.0 Mn 4.4±0.8 1.2 1.2±0.1 2.0±0.1 0.9±0.0 2.3±0.1 Zn 6.1±0.3 1.3 4.3±0.1 3.0±0.1 3.3±0.3 4.5±0.1 Na 1.3±0.2 68.4 5.4±1.3 21.2±2.9 4.5±0.6 1.8±1.0 a Presented values are mg/100 g dry weight±standard error b Values calculated from the 14 accessions reported in Table 2; N=14 c Values derived from a single wild collection as reported in [12] d Values are for mature, whole seeds (USDA/ARS, 2007; NDB No. 16069); N=4–14 e Values are for mature, whole seeds (USDA/ARS, 2007; NDB No. 16056); N=25–47 f Values are for mature, whole seeds (USDA/ARS, 2007; NDB No. 16014); N=4–28 g Values are for mature, whole seeds (USDA/ARS, 2007; NDB No. 16108); N=5–78 higher in the 14 accessions, relative to the wild collected For Ca, although the values from the Food Nutrient material (Tables 2, 3), it should be noted that the complete Database [15] for these legumes appear representative of mineral nutrition employed in this study may have allowed several published studies, there also are reports of higher these lines to express their full transport potential for these values. These include concentrations as high as 480 mg/ minerals [9]. Whether the wild material from the forest 100 g dry weight for Glycine soja germplasm [17]andas habitat was limited in these metals, or truly had a lower high as 363 mg/100 g dry weight for germplasm transport potential, is difficult to assess without growing it [18], demonstrating that these cultivated legumes can under similar conditions to those used in this study. achieve seed Ca concentrations that approach those seen Similarly, the levels of Na, which were lower in the 14 in T. labialis. accessions, relative to the wild material, may be due to low It appears that T. labialis and a few other wild legumes Na availability in the current study. Sodium salts were not (e.g., Cassia sps. and Abrus precatorius)[4, 19]havethe added to the drip line solution and the only source of Na potential to mobilize Mg or K more effectively and to would have been from the synthetic soil, the vermiculite, or accumulate these minerals at higher concentrations in seeds, as contamination in the other mineral salts. relative to many common grain legumes. In combination Seed values for the macronutrients Mg and K were with a good protein concentration [12], these results suggest significantly lower in the 14 accessions, relative to the that it would be beneficial to promote T. labialis as a human previous report [12]. Values (mg/100 g dry weight) ranged food, especially where it is currently being grown as a forage from 263 to 489 (Table 2) versus 812 [12] for Mg and crop. Additionally, this species could serve as a model for ranged from 1,038 to 1,316 (Table 2) versus 2,167 [12] for understanding the mechanisms controlling Ca, Mg or K K. These differences appear to be true genotypic variations transport from vegetative tissues to developing seeds [9, 20]. because the higher values are not without precedent in other Interestingly, T. labialis is closely aligned with soybean, as species. A seed Mg value of 813 mg/100 g dry weight has both are members of the Glycineae subtribe, and thus the been reported for Cassia obtusifolia [4] and K values close molecular resources developed for soybean [21] are likely to 2,000 mg/100 g dry weight have been reported for transferable to and could be used to study mineral transport several legumes [4, 6]. Unfortunately, we were unable to and partitioning in this wild legume. obtain Indian germplasm of T. labialis for the present study, which under our optimal growth conditions may have Acknowledgements This work was funded in part by funds from demonstrated even higher seed concentrations of Mg or K. USDA-ARS under Agreement No. 58-6250-6-001. The contents of Relative to commonly grown field-cultivated cultivars of this publication do not necessarily reflect the views or policies of the US other legumes (lentil, chickpea, black bean, and soybean), Department of Agriculture, nor does mention of trade names, commercial T. labialis exhibits seed mineral concentrations that are products, or organizations imply endorsement by the US Government. The author wishes to thank Brad Morris (USDA-ARS, Griffin, Georgia generally comparable to these species (Table 3). The USA) and Theodore Hymowitz (University of Illinois, Urbana, Illinois exceptions are Ca and Mg, which are higher in T. labialis. USA) for providing the germplasm used in this study. Plant Foods Hum Nutr (2008) 63:105–109 109

References 12. Viswanathan MB, Thangadurai D, Vendan KT, Ramesh N (1999) Chemical analysis and nutritional assessment of Teramnus labialis (L.) Spreng. (Fabaceae). Plant Foods Hum Nutr 54:345–352 1. Wang TL, Domoney C, Hedley CL, Casey R, Grusak MA (2003) 13. Funes F, Perez C (1976) Agronomical studies on perennial Can we improve the nutritional quality of legume seeds? Plant soybean. I. Comparison between Glycine wightii and Teramnus Physiol 131:886–891 labialis under cutting and grazing conditions. Cuban J Agric Sci 2. Grusak MA (2005) Legumes: types and nutritional value. In: 10:199–209 Allen L, Prentice A (eds) Encyclopedia of human nutrition, 2nd 14. Rika IK, Mendra IK, Gusti Oka M, Oka Nurjaya MG (1991) edn. Elsevier, New York, pp 120–125 Forage species for coconut plantations in Bali. In: Shelton HM, 3. Rodrigues BF, Torne SG (1991) A chemical study of seeds in Stür WW (eds) Forages for plantation crops. ACIAR Proceedings three species. Trop Sci 31:101–103 No.32. Canberra: ACIAR, pp 41–44 4. Mohan VR, Janardhanan K (1995) Chemical determination of 15. U.S. Department of Agriculture, Agricultural Research Service nutritional and antinutritional properties in tribal pulses. J Food (2007) USDA National Nutrient Database for Standard Reference, Science Technol 32:465–469 Release 20. Nutrient Data Laboratory Home Page, http://www.ars. 5. Seena S, Sridhar KR, Jung K (2005) Nutritional and antinutri- usda.gov/nutrientdata tional evaluation of raw and processed seeds of a wild legume, 16. Beebe S, Gonzalez AV, Rengifo J (2000) Research on trace Canavalia cathartica of coastal sand dunes of India. Food Chem minerals in the common bean. Food Nutr Bull 21:387–391 92:465–472 17. Raboy V, Dickinson DB, Below FE (1984) Variation in seed total 6. Vadivel V, Janardhanan K (2005) Nutritional and antinutritional phosphorus, phytic acid, zinc, calcium, magnesium, and protein characteristics of seven South Indian wild legumes. Plant Foods among lines of Glycine max and G. soja. Crop Sci 24:431–434 Hum Nutr 60:69–75 18. Abbo S, Grusak MA, Tzuk T, Reifen R (2000) Genetic control of 7. Bouis HE (2005) Micronutrient fortification of plants through seed weight and calcium concentration in chickpea seed. Plant plant breeding: can it improve nutrition in man at low cost? Proc Breeding 119:427–431 Nutr Soc 62:403–411 19. Siddhuraju P, Vijayakumari K, Janardhanan K (1995) Nutritional 8. Pfeiffer WH, McClafferty B (2007) HarvestPlus: breeding crops and antinutritional properties of the underexploited legumes for better nutrition. Crop Sci 47:S88–S105 Cassia laevigata Willd. and Tamarindus indica L. J Food Comp 9. Grusak MA, Cakmak I (2005) Methods to improve the crop-delivery Anal 8:351–362 of minerals to humans and livestock. In: Broadley MR, White PJ 20. Grusak MA, DellaPenna D (1999) Improving the nutrient (eds) Plant nutritional genomics. Blackwell, Oxford, pp 265–286 composition of plants to enhance human nutrition and health. 10. Tanksley SD, McCouch SR (1997) Seed banks and molecular Ann Rev Plant Physiol Plant Mol Biol 50:133–161 maps: unlocking genetic potential from the wild. Science 21. Gepts P, Beavis WD, Brummer EC, Shoemaker RC, Stalker HT, 277:1063–1066 Weeden NF, Young ND (2005) Legumes as a model plant family. 11. Verdcourt B (1970) Studies in the Leguminosae-Papilionoïdeae Genomics for food and feed report of the cross-legume advances for the ‘Flora of Tropical East Africa’: II. Kew Bull 24:235–307 through genomics conference. Plant Physiol 137:1228–1235