EVALUATION OF ALFALFA PLANTS FOR SAPONiNS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MERLYN JONES 1969 .HL-AJLU m. LIBRARY W /////I///I///l/////I/// /I//////I/I//////l/////I//I ‘ M University This is to certify that the thesis entitled EVALUATION OF ALFALFA PLANTS FOR SAPONINS presented by Marlyn L. Jones has been accepted towards fulfillment of the requirements for Ph.D. degree in_9_1:9.p_§s:1ence QW‘L 6’, CE W Major professor Date W9.— 0-169 BIN‘DING BY nuns & snu3' mg mm mc. \ nanFRQ ABSTRACT EVALUATION OF ALFALFA PLANTS FOR SAPONINS by Marlyn Jones Three bio-assays utilizing the hemolytic property and one based on the ichthyotoxic nature of saponin were deve10ped to rapidly and effectively evaluate saponin levels in alfalfa. Hemolytic and minnow deter- minations were significantly correlated with chemical determinations. The three erythrocyte assays formed the basis of alfalfa saponin investigations. Saponin was highly heritable in the two pepula- tions of Medicago sativa ;. studied. Sufficient variation existed in all alfalfa pepulations studied to facilitate selection for high and low saponin lines. Selection of high and low saponin lines by Erythrocyte Assay I was shown to be effective in four alfalfa populations. Alfalfa saponin was stable both within the plant and in water extracts. Extracted saponin reacted with cholesterol and lost nearly all its hemolytic Merlyn Jones activity. Heat did not affect the hemolytic activity of the same extracts. Extracts from plants low and high in saponin did not exhibit abnormal hemolytic activities when analyzed in mixtures. Saponin levels in'!. falcata changed very little over a two month period of first growth. 'Hemolytic ' values" of leaves were nearly three times as high as those for stems and the tap one-third of tepgrowth was higher in hemolytic activity than the bottom third. The organ of saponin synthesis was studied through the use of grafting techniques. Both leaves and roots appear to be capable of synthesizing ' saponins. EVALUATION OF ALFALFA PLANTS FOR SAPONINS By L/ . ,- MerlyniJones A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop Science 1969 PLEASE NOTE: . Several pages contain colored illustrations. Filmed in the best possible way. UNIVERSITY NICROFILHS The union u: guid :epint Apr irrisor lid to 1 hit a: Elem . 5n u. .n. “first; 'lfe', ‘34 in "Mia ACKNOWLEDGMENTS The author wishes to express his sincere appre- ciation to DB. Fred 0. Elliott for his encouragement and guidance during the course of this study and in preparation of the manuscript. Appreciation is also expressed to Dr. Carter M. Harrison for his critical review of the manuscript and to the American Red Cross, Lansing, Michigan for their assistance in supplying blood for the devel- opment of assays in this study. Special gratitude is due to my wife, Shirley Ann, and daughters, Kim and Kay, for their unselfish understanding through three years of study. My wife's assistance with field and laboratory work and in typing and reviewing the manuscript is especially appreciated. ii m TABLE OF CONTENTS Page ACKNOWLEDGMENTS . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . vii INTRODUCTION . . . . . . . . . . . . 1 REVIEW or LITERATURE . . . . . . . . . h Alf‘lf‘ saponin! e e e e e e e e 9 Genetic studies . . . . . . . . 12 Erythrocyte assay . . . . . . . 13 Fish assay . . . . . . . . . . I“ ChOmic‘l ‘fi..y e e e e e e e e 15 MATERIALS AND METHODS . . . . . . . . . 16 Evaluation of Alfalfa Saponin . . . 17 As EFYthrOOYt. ".‘y I e e e e e 17 B. Erythrocyte assay 11 . . . . . 20 c. Erythrocyte assay III . . . . . 21 De Fish ‘.8" e e e e e e e e 22 Be Chonic‘l ‘IC‘y e e e e e e e 22 I. 0th0r O O O O O O O O O O 23 RESULTS AND DISCUSSION . . . . . . . . . 2b Erythrocyte ‘..‘y I e e e e e e 0 2“ Erythrocyte assay 11 . . . . . . 26 Erythrocyte assay III . . . . . 28 Comparison of erythrocyte assays . . 30 Filh assay e e e e e e e e 31 Comparison of erythrocyte assay I, fish, ‘nd Omnicu any. e e e e e 32 Distribution and heritability of s.p0n1n 1n ‘lf‘lf‘ e e e e e e 35 iii Some properties of alfalfa saponins. Seasonal trend and areas of top- growth concentration of alfalfa .‘ponin. e e e e e e e e e Organ of saponin synthesis. . . . CONCLUSIONS 0 0 O O O O O O O O O 0 LITERATURE CITED . . . e . . . . . . iv Page “9 5h 60 66 68 ' Table l. 2. 7. 8. 9. 10. LIST OF TABLES Erythrocyte, minnow, and chemical assay values of leaf samples from twenty alfalfa plants. . . . . . Correlation coefficients for three assays of crude saponin extracts of alfalfa leaves . . . . . . . 'Hemolytic value” ranges and means of three Medicago species and three go I‘tiV‘ ks V‘riotio. e e e e e Analysis of variance and estimates of saponin heritability of two 5. sativa L, populations. . . . . . Mean progeny "hemolytic values” of Vernal alfalfa parentage. . . . . Mean progeny 'hemolytic values" of Culver alfalfa parentage. . . . . Analysis of variance and estimates of saponin heritability in leaf, stem, and whole samples of a diallel cross between seven Vernal clones. . . . Mean progeny leaf, stem, and whole plant 'hemolytic values" of Vernal alfalfa parentage . . . . . . . Correlation coefficients of leaf and stem 'hemolytic values" and 1.‘f/.t°m r‘tios e e e e e e e Saponin levels of progeny from selected low or high saponin parents, two unselected varieties, and one partially selected alfalfa increase . Page 33 3h 36 38 39 he #1 #2 Nb “7 Table 11. 12. 13. 1h. 15. ‘Effect of cholesterol and heat treatment on the hemolytic activity of crude saponin extracts of Vernal alfalfa stem and leaf samples. . . . . . Expected and actual hemolytic activity of crude saponin extracts of alfalfa whole plant, stem, or leaf samples. Leaf and stem ”hemolytic values“ from four asexually propagated‘g. falcata plants harvested over a two-month ”riod O O O O O O O O O 'Hemolytic values“ of stem and leaf samples from three different positions of four M. sativa plants . . . “Erythrocyte Assay II hemolytic values“ of leaves from forty-seven grafts and of the stock and scion parents.. vi Page Figure l. 2. 3. 5. 7. LIST OF FIGURES Ten tube dilution series of three plant extracts evaluated for hem- olytic activity by Erythrocyte Assay I. ”Set A” - low saponins, “Set B” - average saponins, and ”80': C" - high S‘pOflil'IIe e e e e e Micrograph of 0% hemolysis at an erythrocyte preparationxplant extract ratio of 1.00:0.85. . . . . Micrograph of 50% hemolysis at an erythrocyte preparationxplant extract ratio of 1.00:1.00. . . . . Micrograph.ef 100$.hemolysis at an erythrocyte preparationxplant extract ratio of 1.00:1.15 . . . . Hemolysis time sequence of ten alfalfa leaf samples analyzed by Erythrocyte Assay III. Horizontal series-Sets A, B, and C - are hemolysis after i, 2, and b hours of ten plant samples arranged in vertical rows. Distribution of progeny from plants selected for low and high saponin . . Hbmolytic trend of M, falcata leaves and stems over a two—month first growth period and trend of percent leaves for the same population. . . . . . . vii Page 25 27 27 27 29 N6 58 INTRODUCTION Identification and evaluation of natural plant constituents affecting nutrition should be as important as evaluation of the many factors affecting plant yields. Until recently, applied plant research was directed predominantly toward the improvement of yield, with the exception of cases where very severe animal growth depression or death resulted from certain plant diets. Alfalfa has been improved in yield and persistency of stand through both breeding and managerial efforts. Improved alfalfa quality, on the other hand, has been achieved primarily through managerial efforts, even though a potential for quality improvement exists through plant breeding. Alfalfa possesses stimulative as well as inhibitive growth properties to many biolog- ical systems. Some of the compounds responsible for these properties may be controlled by relatively simple genetic systems. Consequently, rapid shifting in the levels of certain compounds may be realized. 2 One of the poorly understood and ill-defined chemical components of alfalfa is a group of steroidal and triterpenoid glycosides (saponins) which exhibit a broad spectrum of biological activity. They were identified in alfalfa fifty years ago, but have not received sufficient attention to adequately under- stand their growth depressing effects on microor- ganisms, monogastric or ruminant systems, or to appreciate their role in plant metabolism. Much of the research concerned with biological activity and mode of action of compounds such as saponin is necessarily of an empirical nature. In order to aid this research, alfalfa populations of high and low saponin content should be available. Also, a better understanding of alfalfa saponin distribution within Medica 0, distribution within individual plants, and a better knowledge of its properties would be most helpful. This study was initiated for the purpose of support- ing more comprehensive research on alfalfa saponins. Its objectives were to (1) develop rapid, sensitive, and reliable assays for the evaluation of alfalfa saponin, (2) obtain estimates of alfalfa saponin heritability and establish alfalfa base populations of very high and very low saponin content to supply sufficient material for 3 evaluations in higher organisms, and (3) establish certain properties of alfalfa saponin, their dis- tribution within plants, and the organ or organs where they are synthesized. cons suga cert numb 0f 5. fr0m Nma. PM Glad] “pox rmu dist; REVIEW OF LITERATURE Most saponins are nitrogen-free glycosides, each consisting of an aglycone (sapogenin) and one or more sugar units. Many of these diverse compounds possess certain common biological preperties. Review articles by Hiller‘gt‘gl. (1966), Heftman (1967a), and Basu ggmgl. (1967)describe many aspects of steroidal and triterpene saponin structure, isolation, purification, chemistry, pharmacology, toxicology, and distribution. Both of the major saponin groups occur throughout many plant families. Often a plant saponin is actually a mixture of two or more saponins. In a 12 year survey of over 6000 plants representing 208 families and 1397 genera, steroidal saponins were ascertained in a number of monocot and dicot families (Hall gt 5;. 1961a). Of 5&2 plant species of Malaya, belonging to 295 genera from 89 families, lhfi gave a positive saponin reaction (Amarasinghamngt‘gl. 196%). In Turkmenia 51 of 236 plant species (13 of #8 families) contained saponin Gladkikhtggigl. 1965). According to Lindner (19h6) saponins.occur in various genera of the Leguminosae , family. It is difficult to discern a pattern of saponin distribution from these studies. 5 McNair (1932) reported that saponins resemble alkaloids in diminution of toxicity from temperate to tropical climates. Some saponin-containing plants (spinach, beetroot, and asparagus) are components of the human diet while others are consummed as animal feed. Extracted saponins have found use in such products as soft drinks, soaps, fire extinguishers, confectionaries, and pharmaceuticals. The many demonstrated in-vitro activities of saponin and its detrimental effect on growth of many organisms have prompted legal action to control its use in human foods (George 1965). In spite of numerous investigations the activities and mode of action of saponin in in-vivo monogastric and ruminant animals remains poorly understood. Many properties of saponins, known to exist at the turn of the century, were described by Sollmann (19b2). These include affinity for cholesterol, hemolytic activity, fish and snail toxicant, surface active, mucous membrane irritant, and general solubility in water. The early therapeutic uses of plants containing saponin dating back to the very early 16th century, were based upon effects on absorption, circulation, and hemolysis. Many important biological activities of saponin have been studied only recently. The antitumoral acti- vity of saponin in rats and mice (Horvath 23 5;, 1967; 6 and Kupchan 22.£l° 1967) has been investigated only recently. Saponin induced enzyme inhibition has been reported for at least four enzymes (Mircevova gt 3;. 1968; Ishaaya gt_=l,19653 Ishikawa gt 5;, 1960; and Birk gt_;l. 1963). Cardioactive properties in the frog and guinea pig (Von Blumencron 1901; and Roy 32,2l. 1963), effect on the electrical activity of the rabbit brain (Sokolov 1965), ability to influence atherosclerosis in rabbits and mice (Gladkikh 1965a; and Yeti-ova‘gg‘gl. 1966), and abortifacient activity in rabbit, goat and cow (Dollahite gghgl. 1962) are among the many reported activities of saponin. The antimicrobial activity of saponin is well established (Walters 1968; Kahl 22.5;, 1966; and Tschesche aglgl. 1965) and a bioassay for alfalfa saponins was developed on the basis of this activity (Zimmer ggigi, 1967). Interestingly, however, micro- organism cultures capable of degrading saponin have been observed in silage of Agave (Sanchez-Marroquin g§.g;, 1967) and in the rumen of cattle (Gutierrez _£.gl, 1959). The role saponins play in plant metabolism and growth, if any, is not known, nevertheless, they have ,been shown to influence plant deve10pment in many ways. Aaapplication of saponin solutions to leaves stimulated the.development of shoots and roots in Begonia, induced tumors in Hedera helix, influenced chlorophyll 7 synthesis in Euonymus Japonicus, and inhibited root growth in other plants (Bonner and Varner 1966). This review also reports a sUppression of root hair development in water cross and an acceleration of germination of pea, maize and tomato seeds. Germination is generally inhibited at high saponin concentrations and promoted at low concentrations. Plant growth-regulating activity of some saponins has also been demonstrated (Vendrig 196k). Distribution of saponins in plant parts is not uniform and maximum levels may occur in different organs of various plant genera at different stages of maturity. Dormant seeds, as well as the tubers, roots, stems, and leaves of Dioscorea contained saponins (Baker'gg‘;l. 1966). The greatest accumulation occurred in the tubers when the spring shoots were in full bloom. Saponins were detected only in the roots of two Gypsophila species and percent content was highest (as high as 19$»of root dry matter) when the first leaves appeared (Kolodzieski‘gg‘2l. 1965). Lowest saponin values of the subterranean organs occurred at flowering in four other plant genera (Drozdz 196“). Differences between the nature of saponins in various plant parts and even in the same part during its development are not uncommon. 8 Present knowledge of steroidal saponin biosynthesis indicates that cholesterol may be a precursor of certain sapogenins. Cholesterol has been shown to exist in many plant tissues and has been converted by plants to sapogenins and alkaloids (Heftman‘gt al. 1967a). The same review disclosed that the incorporation of radio- active mevalonic acid into sapogenins was higher and qualitatively different in the Dioscorea gpiculiflora shoot system than the tubers. Kessar et al. (1968) formed other adducts which readily led to steroidal sapogenins. Many extraction procedures and evaluation tech- niques of plant saponins have been utilized. Cholesterol precipitation (Walter‘s: 2;. 195b), activated carbon- solvent systems (Van Atta 23.31, 1961b), paper chromatography and electrophoresis (Coulson 1957a), and ion exchange resins and column chromatography (Jackson ;£,;;. 1959) have been utilized to isolate saponin fractions. Reaction of extracted saponins with a Lieberman-Burchard reagent provides a chemical assay for saponin (Gestetner 33,51. 1966; Van Atta g£_;l. 1958a). Quantitative and qualitative evaluation of plant saponins has been based primarily on their biological activities. Many alfalfa and soybean saponin evaluations have employed biological assays similar to those developed for saponins of other sources. Special attention is currently being paid to the biological 9 properties and physiological activities of compounds found in these two saponin containing plants in order to improve their present and future utilization. Alfalfa Saponins Evaluations of the energy, protein, fiber, mineral, fat, and vitamin content of alfalfa have been useful criteria for quality, but in many cases could not be closely correlated with biological responses. The search for other factors affecting alfalfa nutritive utilization has revealed the presence of estrogens (Bickoff gtflgl. 1957a), purine derivatives (Bickoff 33.5l. 1968b), and saponins (Lourens 23.21. 1951). Although an extract from alfalfa possessing the preperties of saponins was obtained early in the century (Jacobson 1919), little research was directed toward understanding the biological and chemical properties of alfalfa saponins until recently. As many as seven to ten alfalfa saponins have been isolated using paper chromatographic techniques (Lourens 1961; and Coulson 1962b) and several biological activities have been attributed to these water-soluble compounds. Biological activities, extraction, purification, and characterization of alfalfa saponins are reviewed by Scardavi g£_;l. (1967). Articles on saponin inhibition of seed germination (Guenzi g£_gl. 196“; Pedersen 1965; and Megie Zfinfil' 1967), control of the causative fungus 10 of Avocado root rot (Zentmyer at ll; 1967), association with ruminant bloat (28), and the development of an assay based on a saponin sensitive fungus (91) concur with observations cited in the review by Scardavi 23.;l. (70). Other investigations as to the cause of ruminant bloat, however, indicate that alfalfa saponins are probably not solely responsible for this condition (Lindahl _£,;;. 1957; Stifel gt_=l. 1968; and McArthur 196h). The many activities of alfalfa saponins are thought to be primarily the action of triterpenoid saponins. Extraction and purification methods have been based primarily on water or alcohol extracts and the property of saponins to be adsorbed on charcoal, or combined with cholesterol. Detection and characterization of alfalfa saponins have been based on reactions with Liebermann-Burchard reagent, inhibition of the respira- tion of rat diaphragm, chick and fungus growth in- hibition, hemolytic activity, melting point, specific rotation and spectographic information. Pedersen 23'31. (1967b) was able to confirm the report of Hanson 23.51. (1963) which demonstrated quantitative differences between varieties. He also reported higher saponin concentrations in the leaf than the stem and quantitative changes within varieties sampled throughout one growing season. This is not surprising as quantitative changes within seasons have 11 been observed in other saponin containing plants (3, #5, 15). In the same report Pedersen g£_gl. (62) demonstrated the existence of qualitative alfalfa saponin differences in varieties by the use of chicks, fungus, and lettuce seed germination assays. ‘M. sativa and‘M. falcata are reportedly two of the highest saponin containing species of the Medicago genus (Jaretzky 19h0). Saponin biosynthetic activity of the Dioscorea apiculiflora shoot system was greater than the roots (Bennett 25‘;;.1965) and leader shoots were given as possible sites of saponin synthesis in two other Dioscorea species (3). Nevertheless, saponins occurred throughout the plants. Saponins occur in all alfalfa organs; roots, leaves, stems, and flowers (Morris 1965; 1961; and Pedersen 1967b), but may be qualitatively different. One site of synthesis is unlikely if saponins cannot diffuse through living cells or across grafts as was the case in Calendula officinalis (Fischer et al. 1951). Beat processing of extracted soybean saponins decreased hemolytic activity, but had no influence on saponin within heat treated seed (8). Heating saponins of other sources did not change inflammatory or hemolytic properties, however, treatment with cholesterol decreased hemolytic, but not inflammatory properties (Richou ggflgl. 1965). Cholesterol had no 12 influence on hemolytic activity of a group of acid saponins or a saponin extracted from Chelidonium mains L. (Ruyssen g£_=l. 19h6; and Kwasniewske 1958). Genetic Studies The general and/or specific combining ability (gca/sca)- as determined by Griffing's (26) analysis of dialled systems - of a number of quantitative char- acters in alfalfa have been studied recently. Significant effects for gca existed for alfalfa forage yield (Porceddu 1968; and Daday 1967). Likewise, gca of the followinglfl. sativa plant characteristics was highly significant; estrogen content, leaf/stem ratio, protein content, in-vitro dry matter digest- ibility, cell wall constituents, acid detergent lignin, and acid detergent fiber (G11 3373;. 19671 and Stuthman 1967). Buker (1963) reported significant gca for percent leaves. Breeding for improved meth- ionine levels may also be feasible (Singleton gt‘gl. 1952). Two reports have directed attention to possible progress in lowering alfalfa saponin levels through plant breeding (29, 62). An estimate of saponin heritability exists for one crop, Dioscorea floribunda. This plant is cultivated for yields of steroidal saponins used in the synthesis of steroidal drugs (Martin 33,5l. 1967). Clones were compared for fresh 13 weight, dry weight, precent sapogenin, and total sapo- genin. Percent sapogenin was the only trait with an estimate of heritability which would be useful. Erythrocyte Assay Saponins, as determined by hemolytic activity, have been reported in many plant families (22). Alfalfa saponins possess hemolytic activity (#8), how- ever, this activity has not been used as a major screening criterion in alfalfa saponin studies. The hemolytic activity of saponins has been attrib- uted, by some researchers, to a chemical reaction of sap- onins with cholesterol of the cell membrane (Joos 21 a_l_. 1967b, and Schmidt-Theme _e_t_:_ a_l_. 1950). Not all hemolytic saponins, however, form cholesterides (Joos 1966a; Kwasniewski 1958; and Ruyssen gig. 19h6). Complete hemolysis occurred at concentrations equal to or lower than would be needed to form a single monolayer on the cell surface (Carter 33 Sl° 1931; and Granick 19159). In one study (Kesten 23 LL. 1928) hemolysis was inversely proportional to cell concen- tration and directly preportional, through the greater portion of the reaction, to the square of saponin con- centration. With a given saponin, hemolytic velocity varies with the source of the erythrocyte, being low for man and high for sheep (71). Rate of saponin hemolysis of different normal human blood samples Van lb was essentially the same under constant conditions (Elbel g_t_:_ 2;. 19315; and Kesten 1928). Erythrocyte resistance to saponin hemolysis is about the same, irrespective of source, at equal saponin and erythro- cyte concentrations (Jung 2: _a_l_. 1950). Richou _e_t_ _a_l_. (66) found no relation between hemolytic activity, lethal dose in mice, and inflam- matory effect of different lots of saponin. Lindahl g}; 3;. (#8) also reported no definite correlation between blood hemolysis and other physiological actions of alfalfa saponins, however, Awe _e_t_: _a_l_. (1950) found that the surface tension and hemolytic indices of extracts of certain roots and fruit had parallel variations. Vacek et al. (1962) showed that correlations existed between foam number, superficial tension, hemolytic activity, and toxicity after subcutaneous administration of four saponins. Fish Assay Fish were used early in the 20th century to evaluate toxic effects of certain plant extracts (Priess 1911). An ”equation of toxicity” was developed for surface-tension lowering poisons and successfully tested on minnows 90 - 100 m in length of Alburus lucidus in 19h2 (Macovski 23 a_l_. 19152). A fish assay was proposed as a simple method for verifying the presence of saponin in water extracts of 15 iohthyotoxic plants (Meyer 19h2) and a fish index, based upon Lebistes reticulatus, was later used to evaluate saponin extracts from a Brazilian shrub (Uasicky'gg‘gl. 19b9). Saponins were found to control predeceous fish in shrimp ponds in Taiwan (Tang 1961). gaggical.Assay A procedure described by Hall gg‘gl. (1952b) for extraction, isolation, and identification of steroidal sapogenins has been employed in the isolation of alfalfa saponins (50, 91). A gravimetric determination of total alfalfa saponin was developed by Van Atta 2£_;l. (81). The latter method of extraction has been used to study alfalfa saponins as related to variety, cutting, and location variables (29), and to compare chemical and biochemical assays (62). MATERIALS AND METHODS Plant material was obtained primarily from eight alfalfa populations of three Medicago species (3. sativa, g. falcata, and M, glutinosa). Only one of the populations had been previously selected for saponin content. Populations of three diallel crosses (two of Vernal and one of Culver parentage), two USDA releases (MBA and MSB), DuPuits, g. falcata (Russian source 22506), and M. glutinosa (Russian source 29003) were transplanted on 0.91h m centers in the summer of 1965 at East Lansing, Michigan on Conover loam with a pH of 6.5. Phosphorous levels were adequate for good alfalfa growth, however, potassium levels were low. A selected Culver x Vernal population was established in the same area in 1966. Harvests of individual clones were made at first bloom on June 10, 1967 and 1968. Following air drying, leaf-stem separations were made on much of the 1968 harvested material. Field and greenhouse materials were dried at #3 - #7 C, ground to pass through a 1 mm screen, and stored at 0 - 6 C. 16 17 Greenhouse materials were of two types; 1. established field clones selected and transferred as potted plants to the greenhouse for further study and 2. seedlings grown from selected crosses. Evaluation of Alfalfa Saponin Hemolytic activity and ichthyotoxicity, were the basis of the four assays in this study for evaluating individual plants. A. Erythrocyte Agggy'I Crude saponin extracts of alfalfa were prepared by soaking 2 g of ground leaf, stem, or whole plant samples in 23 ml of physiological saline solution for 2 hr. (The saline extracts were not different in hemolytic activity from distilled water extracts made isotonic with salt after extraction.) The mixture was then filtered through two layers of Kimwipes tissue and the filtrate collected and refrigerated at 2 C. Specific amounts of the extracts were added to a 2% human erythrocyte solution. The erythrocyte solution was prepared by centrifuging citrated human blood at 5,900 g for five min, washing the collected red blood cells twice, and adding 20ml to 980 ml of physiological saline solution. Ten tubes, each containing 1 m1 of the erythrocyte preparation, were used in a series to evaluate each plant 18 extract. One ml of the plant extract was added to tube #1, mixed, and 1 ml transferred with a 1 ml syringe to tube #3. The contents of tube #3 were mixed and transfer- red, as above, through tubes }5, 7, and 9. One ml of a diluted plant extract (1.3 ml plant extract + 0.7 ml saline solution was placed in tube #2, mixed, and transferred in the same manner through tubes in, 6, 8, and 10. The series was placed in a water bath (3? C) for 30 min, removed from the bath and allowed to remain at room temperature for 15 - 20 min before reading. A series containing no saponin was included periodically to aid percent hemolysis estimates. This preparattion differed from a standard dilution series. Plant extract concentrations in the reaction mixture were obtained by transferring one-half the initial plant extract-erythrocyte mixture through a series of tubes containing only erythrocyte preparation. In this series the plant extract concentration decreased and, unlike the standard series, the concentration of erythrocytes increased. This allowed a rapid method for obtaining concentration gradients and the results did not differ significantly from those obtained in a standard series. The lowest concentration of plant extract causing complete hemolysis was termed "titer”. Complete hemolysis was defined as absence of erythrocyte sedimentation at the bottom of the test tube. Titer values were established 19 by recording the last tube containing completely hemolyzed erythrocytes and estimating the percent hemolysis occurring in the adjacent tube. The ”hemolytic value" was calculated as volume of 2% egythrocyte preparation volume of undiluted extract at titer concentration . These proportions for tubes #1 - 10 respectively were 1.oo/l.oo, 1.oo/o.65, 1.50/o.50, 1.50/o.33, 1.75/0.25. 1.75/o.16, 1.88/0.13, 1.9b/o.06, 1.9h/o.ou, and the resulting "hemolytic valueS” were 1.0, 1.5, 3.0, “.6, 7.0, 10.8, 15.0, 23.1, 31.0, and “7.7 respectively. In a few instances the plant extracts did not have sufficient hemolytic activity to cause lysis in tube #1. These samples were freeze-dried and reconstituted to increase the extract concentration as much as fivefold. Two hemolytic determinations were made for each sample. Modifications in preparing plant samples for the hemolytic assay included the use of fresh and frozen plant samples. Uater extract of fresh plant material was obtained by chopping 7 g of alfalfa leaves with 50 ml saline solution for 1.5 min at high speed in a Haring blender and filtering the resulting mixture through two layers of Kimwipes. Plant juices were also collected by freezing a small fresh sample and after thawing, placing it under 2,270 kg pressure in a Carver press. Extracts from samples thus prepared were analyzed in the same manner as the extracts of dried material. 20 B. Erythrocyte Agggy ;; Sample size was limiting in one of the studies. As little as 30 - “0 mg of dried alfalfa leaves was available. In order to accurately evaluate the hemolytic activity of this size sample, microscOpic observations were made of the erythrocyte-extract solutions. Saline extracts of alfalfa leaves (1 part saline solution 8 0.017 parts alfalfa leaf) were obtained using a 2 hr soaking period. This extract was filtered through one layer of Uhatman #1 filter paper and added to an 0.08% human erythrocyte preparation in a micro concave slide. Different volume combinations of plant extract and erythrocyte preparation were used to obtain the desired erythrocyte preparationxplant extract ratios. Pipettes of 10 - loo/(capacity allowed proper volume combinations. The reaction mixture was stirred twice during the f hr reaction period and covered at other times to avoid evaporation losses. A dilution series contained ten erythrocyte- extract mixtures and included erythrocyte preparations plant extract ratios from 2.0:1.0 to 1.030.07. In most cases this series of ten mixtures included the desired hemolytic range (0 - 100% hemolysis). If the hemolytic range was not included, ratios adequate to include this range were prepared. After a i hr incubation period the erythrocytes were observed at a magnification of 125 X to determine the degree of lysing. Titer values were 21 established by recording the last preparation containing completely hemolysed erythrocytes and estimating the percent hemolysis occurring in the adjacent preparation. Erythrocyte preparationsplant extract ratios as small as 1.0:6.0 and as great as 1.030.02 were prepared. Two hemolytic determinations were made for each plant extract. C. Erythrogyte Agggy II; An abbreviated form of Assay I was develOped to screen large numbers of alfalfa plants for hemolytic activity. Sample size ranged from 2 - 7 leaflets (approximately 23 - 28 mg) depending upon leaflet size. One sample was collected from each plant in an area of young, fully expanded leaves. Leaflets were placed in 3 ml test tubes, dried, crushed with a glass rod, and 2 ml of 1% human erythrocyte preparation added. Each tube was shaken twice in the following 20 min and visually rated for degree lysing after i and “ hr. Samples causing complete hemolysis in the first i hr were rated as having high hemolytic activity and those causing no hemolysis after “ hr were rated very low in activity. Only plants evaluated as being very low in saponin were re-evaluated. 22 no usages: Locally available fathead (Pimephales promelas) minnows were purchased. The minnows were sized, “.0 - “.5 cm minnows saved, and held at 20 C for no more than two days. Alfalfa crude saponin extracts for the minnow assay were prepared by soaking 2 g ground alfalfa leaves in 23 ml distilled water for 120 min. The mixture was then filtered through two layers of Kimwipes tissue and the filtrate collected and refrigerated at 2 C. One minnow was placed in a 360 g wax paper cup containing 90 ml distilled water (20 C) for each plant assay. If the minnow appeared calm and healthy 5 min after its transfer, 10 ml of the alfalfa extract was added. The minnows were closely observed for the duration of the assay (180 min maximum). The time at which the minnow became immobilized was recorded. At least two determinations were made for each plant extract. E. Chemical Agggy Alfalfa leaves were chemically analyzed for total saponin by the method of Van Atta 23.;l. (81). The leaf samples were finely ground so 3 g of plant material were used rather than the suggested 15 g and proportion- ally smaller volumes of liquids were used to prepare the extract solution. One determination was made for each leaf sample. 23 The above assays are the basis for all saponin evaluations to be described. Lesa: Griffing (26) Model 11, Method 2 was used to partition mean squares for crosses into general and specific combining ability. Individual plants from within crosses were used to estimate the within family variance. Estimates of heritability were calculated according to the formula 2? Tips of alfalfa stems (1.0 - 1.5 cm) selected for high or low hemolytic activity were grafted onto stems of selected plants. The grafting procedure was the same as reported by Gorz 23.;l. (2“) in which Melilotus 2l22‘2225; scions were grafted onto stems of the same Species. Thirteen M, sativa plants from four different populations were selected as source material. Scions of high and low hemolytic activity were grafted onto stocks of both high and low activity. A limited number of scions were grafted beck unto the plant from which they were removed. As many as six grafts were made of each graft combination with a total of more than 250 grafts. Scions were harvested after attaining maximum growth and the leaves analyzed by Erythrocyte Assay III. RESULTS AND DISCUSSION The erythrocyte assays provided a means of better studying the nature of saponins in alfalfa. Each erythrocyte assay was a rapid, sensitive, and repro- ducible method of evaluating hemolytic, water soluble alfalfa saponins. The degree of lysing was visually observed in each of the assays with no difficulty because distinct breaking points existed between lysed and non-lysed preparations, even at similar plant extract concentrations. Erythrocyte Aggy I A typical range of hemolytic activity for Assay I is presented in Figure 1. Each ”Set" was the dilution series of one plant extract. Tubes were numbered from left to right. The titer of “Set A” was between tubes #1 and 2. Erythrocytes were completely lysed only at the highest leaf extract concentration (tube #1). Saponins were not present in sufficient quantity in tubes #2 - 10 and intact erythrocytes settled. The titer of "Set B"was between tubes #5 and 6. Very high hemolytic activity existed in the extract of ”Set C." The titer was between tubes #9 and 10. Between “0 - 70% hemolysis occurred in tube #10. 2“ 25 Figure 1. Ten tube dilution series of three plant extracts evaluated for hemolytic activity by Erythrocyte Assay I. I'Set A“ - low saponins, "Set B" - average saponins, and "Set 0' - high saponins. The ten tube hemolysis series had a "hemolytic value" range of 1 - “8. 'Hemolytic values" as low as 0.2 have been recorded after concentrating plant extracts of low activity by freeze-drying. Saponin preparations possessing "hemolytic values" greater than “8 could easily be evaluated by continuing the dilution series beyond tube #10. Values ranging from 0.“ to “5.0 are reported in Table 1; over a hundredfold differ- ence in hemolytic activity. 26 Larger populations of alfalfa plants could be evaluated by Assay I by modifying the harvesting and sampling procedures. Hater extracts of plants freshly harvested and homogenized in a Haring blender exhibited the same order of hemolytic activity as dried, ground, and water extracted plants. Similar activity ‘180 resulted from plant juices collected from fresh frozen samples. Erythrocyte Assay III would be adequate for screening most plant populations for saponin but Assay I could also be effectively employed by using only five tubes in the hemolysis series (tubes #1,3, 597’ ‘nd 9)- Erythrocyte Agggy_I; Low concentrations of erythrocytes were used for this assay (0.08% by volume) and plant extracts were filtered through filter paper to facilitate observation of erythrocytes under the microscOpe. The break between complete and no hemolysis was nearly as distinct as in Assay I. A gradation of hemolysis usually occurred (from 0 - 100%) within three preparations of a series. Figures 2, 3, and “ illustrate this gradation (0, 50, and 100% hemolysis respectively). Figure 2. Micrograph of 0% hemolysis ht‘an erythrocyte , preparationsplant extract ratio of 1.00:0.85. Figure 3. Micrograph of 50% hemolysis at :A erythrocyte preparationtplant extract ratio of 1.00:1.00. Figure “. Micrograph of 100% hemolysis at an erythrocyte preparationsplant extract ratio of 1.00:1.15. 28 In this case gradation of hemolysis occurred at erythro- cyte preparations: plant extract ratios of 1.00:0.85, 1.00:1.00, and 1.00:1.15 respectively. Ratios as low as 1.00:5.00 werereqUired to obtain complete hemolysis in some samples of low saponins and as high as 1.00:0.027 to cause complete hemolysis in samples with very high saponins. Over eighty alfalfa samples were evaluated by this assay and a representative portion of them re-evaluated by Erythrocyte Assay I and III to compare results. Erythrocyte £2.21 LI Over 1500 greenhouse seedlings and established field clones were evaluated by Erythrocyte Assay III in the spring of 1969. The analysis of individual plants is illustrated in Figure 5. A vertical series of three tubes formed a time sequence for one plant. The hori- zontal series “Set A' is all ten plant extract-erythrocyte preparation mixtures after i hr. "Set B" is the same set of mixtures after 2 hr and "Set 0" after “ hr. Plant extracts 2,3,5,6,7,9, and 10, from the left of the horizontal series, exhibited no hemolytic activity after i hr. Complete hemolysis had already occurred in plant extracts 1,“, and 8 after the Q hr reaction period. Only plant extracts 3,6, and 9 exhibited no hemolysis after the 2 hr reaction period and these same extracts demonstrated no hemolytic activity after the 29 “ hr period ("Set C”). The plants of this series were classified as follows; plants 1,“, and 8 --high saponins; 3,6, and 9--very low saponins; and 2,5,7, and 10--moder- ate saponins. Figure 5. Hemolysis time sequence of ten alfalfa leaf samples analyzed by Erythrocyte Assay III. Horizontal series-Sets A, B, and C- are hemolysis after *, 2, and “ hours of ten plant samples arranged in vertical rows. Plant sample preaparaticn was minimal for this assay, however, it was sufficient to allow the saponins of the alfalfa leaf to express themselves as hemolytic agents. Hemolytic activity of whole uncrushed leaflets was very low, but crushing the dried leaflets slightly, even with the fingers, was sufficient to allow the same 30 hemolytic activity as if the sample was dried and ground through a mill. Comparison 22 Erythrocyte Assays Information obtained through Assay I forms the basis of the studies reported. Assays II and III were developed to allow evaluations of very small and numerous samples. Although the assays have the use of hemolytic activity in common, each has its unique characteristics. Thirty samples were prepared and analyzed per day with Assays I and II. Over 200 plants could be evaluated in one day by one individual with Assay III. The variance of determinations was very low. Standard deviation on a single determination basis of one preparation of 1“ samples was 0.3 and 0.5 hemolytic units respectively for material analyzed by Assay I and II. Larger variations would be expected in Assay III because the sample size was not strictly controlled. Uhen values from Assay II were adjusted to the common basis of "hemolytic value" a lower range of values was obtained than with Assay I. The highest "hemolytic value” from Assay II would have been 7.0 and the lowest would have been less than the lowest recorded in Assay I. This reduced activity is due in part to filtering of the extract, but is thought to be caused primarily by the increased surface area provided by the micro concave 31 slide. As reaction mixtures are spread thinly over a large surface the hemolytic activity of surface-active saponin will decrease considerably (Ponder 19“8). Assay III values compared well with those of Assay I. If hemolysis did not occur in Assay III after “ hrs it would have a ”hemolytic value” less than 3. "Hemolytic values” less than 3 would also be considered low in Assay I. In spite of differences in actual ”hemolytic values” the three assays, rated plants in the same relative order. Fish Assay The biological activity of saponins was readily expressed by using the simple materials and methods of the fish assay. Defining the time of assay termination was important for consistent and meaningful results. The assay may be terminated when the fish becomes immobilized or dies. Immobilization was reported in the study because it varied less between determinations and was more easily observed than death. Time from immobilization until death may range from 1 - 15 min depending upon toxicity of the sample. Standard deviation on a single determination basis of 20 samples evaluated for minnow toxicity was 3.9 minutes. The period from assay initiation to minnow immobilization ranged from 10 min to termination of the assay (180 min)(Table 1). I? M’YKV-J-V-m-s-nr's. .-s- -—~ A-_ 32 A 0.0135% solution of commercial saponin (Nutritional Biochemicals Corporation) immobilized the minnows in 8 - 9 min. Local California shiner minnows and female Gambusia fish have also been successfully used to detect alfalfa saponins. Campgrison g£_Erythrocyte Assay I, Fish and Chemical Assays Values of each assay for alfalfa leaf saponins (Table 1) were correlated with the other two assay values (Table.2). Correlation coefficients for trans- formed fish and erythrocyte and non-transformed chemical values ( —0.91, 0.90, and -0.81) as well as non- transformed fish and erythrocyte values indicated that the two saponin assay procedures accurately described alfalfa populations for crude saponin levels. The square root transformation of erythrocyte values improved only slightly the correlation between erythrocyte and chemical assays. High correlation coefficients have also been obtained by evaluating.alfalfa stem and whole plant samples with the three saponin assays. The fish assay was not as highly correlated with the chemical deter- mination as was Erythrocyte Assay I and it lacked other assay qualities possessed by the erythrocyte assay. 33 However, the availability of human blood or a uniform source of minnows may be the major criterion for selecting between the two assays. Table 1. Erythrocyte, minnow, and chemical assay values of leaf samples from twenty alfalfa plants. 'Hemolytic Plant extract Chemical value" toxicity to minnows t rm at n min required for mg crude saponin; immobilization 3 g sample X Y z “5.0 11 78 31.0 22 69 28.0 1“ 62 26.0 , 10 72 26.0 12 70 21.0 1“ 56 19.0 30 “8 16.0 20 “6 16.0 3“ “8 16.0 39 31 7.0 78 ' 38 7.0 “o “3 7.0 58 “6 7.0 105 “5 7.0 117 ““ 3.0 180 35 2.0 180 39 1.9 180 33 0.6 180 26 o.“ 180 20 3“ Table 2. Correlation coefficients for three assays of crude saponin extracts of alfalfa leaves. v? z W -o.91 0.90 x -0.82 0.89 ‘VY -0.81 Y -o.76 The erythrocyte assay may be affected by presence of a substance(s) or condition(s) altering the nature or rate of hemolysis. Accelerators or inhibitors of hemolysis could produce reversible or irreversible effects on the cells of the hemolytic system or affect the properties of the hemolysin. Sugars, proteins, electrolytes, pH, and temperature are some of the factors which may affect the rate or degree of hemolysis. Variations in these or other factors, however, were not major deterrents to the reliability of the erythrocyte assay. The fish.assay would be affected less likely by moderate fluctuations of the above variables. Both the erythrocyte and fish assays undoubtedly reflected qualitative as well as quantitative saponin differences, however, for those materials studied, the quantitative 35 aspect had the greatest influence on assay values. Distribution and Heritability of Saponin in Alfalfa Studies evaluating the presence, distribution, and inheritance of saponins in diverse populations of the plant kingdom have most often been directed toward identifying specific species able to yield saponin of high pharmaceutical value. Consequently, plant species containing high levels of steroidal saponin have been studied more completely than those containing a predominance.of the triterpene types. Saponin contain- ing plants which may have negative nutritional effects on animals consuming them, such as alfalfa, have not received adequate attention. Past saponin evaluations of alfalfa have relied primarily upon M. sativa for source material. Many characteristics of saponin from this source are similar to those of other plant saponins. Plant saponin surveys based upon these common characteristics indicated that its presence often occurs in many species of a genus containing.saponin. Therefore, its distribution in Medicago species would also be expected to be rather broad. Individual.whole plants of three Medicago species and three _M. sativa varieties were sampled in 1967 and 1968 and analyzed for crude saponin by Hemolytic Assay I (Table 3). Differences of saponin content within species 36 was much greater than differences between. '5. sativa exhibited the greatest range of "hemolytic values" of the three species, however, it was also the most thoroughly sampled. The mean "hemolytic value" of g. falcata was double that of the other two species. Of the varieties, Vernal, Culver, and DuPuits plants possessed "hemolytic values".of 103, 16, and 3 fold differences, reapectively. .DuPuits exhibited the highest and Culver the lowest mean values. Table 3. 'Hemolytic value" ranges and means of three Medicago species and three M. sativa L. varieties. Species or Number of "Hemolytic Value" variety plants range mean ,3. glutinosa “0 3.0-15 7 ‘5. falcata “0 3.0-35 16 g. sativa 250 0.2-31 6 DuPuits 30 5.0-15 9 Culver 50 1.0-16 2 Vernal 80 0.3-31 jg. dzawkhetica as well as other plant introductions and varieties have also been sampled and shown to contain saponin. The level of saponin was not always as high ~ as that of the three species sampled in detail. Moapa, a variety of‘g. sativa developed from plant introductions 37 from Africa, was shown to be very low in saponin on the basis of the fish and Erythrocyte assays. . The only available information regarding the inheritance of saponin, based upon a study conducted in Puerto Rice on Dioscorea floribunda (50), revealed that the heritability of percent sapogenin was ”moderate.” The heritability of saponin in alfalfa was investigated in the present study by evaluating the saponin content of individual plants from three‘g. sativa populations. -_.c....- 36-...-.- “mar .»..' A... Estimates of saponin heritability were found to be “moderate - high.” One-half of the diallel crosses within Vernal and Culver parentage were harvested in 1967; two years after establishment. The parents were part of a comprehensive alfalfa quality study but had not been selected pre- viously for saponin content. Six Vernal parent and five Culver parent clones formed the basis for each of the diallel crosses. At least ten plants were established (1965) for each of the resulting 15 Vernal and 10 Culver crosses. Ten randomly chosen Vernal crosses were harvested as bulks and up to twelve plants from each of six crosses were harvested and analyzed in- dividually for an estimate of within family variance. Six Culver crosses were randomly chosen and harvested as bulks.and the within family variance estimated from six to ten plants from each of five crosses harvested and analyzed individually. 38 The "hemolytic values” of six Vernal parents ranged from 1.3 - 25 and those of 5 Culver parents from 1.“ - 7. The gca was highly significant in both populations (Table “). The sea was also significant in both populations (5% level in Culver and 1% level in Vernal). Table “. Analysis of variance and estimates of saponin heritability of two g. sativa L. populations. df 88 «MB Source Vernal Culver Vernal Culver Vernal Culver gca 5 u 320.15 30.01 6“.03**-7.50** sea 15 10 139.7“ 5.10 9.32"”!I 0.51* error 51 39 0.“! 0.20 Vernal Culver Estimate of heritability 0.60 0.79 * significant at 5% level ** significant at 1%tlevel 39 A major portion of the highly significant sea in Vernal was due to the transgressive effect exhibited by progeny of two combinations; V 51 X V 21“ and V 37 X V 21“ (Table 5). The progeny of these combinations as well as V 37 X V 132 had mean "hemolytic values“ lower than either parent. A trend toward plants of lower saponin content than would be expected from parental levels existed in this population. Over “5% of all individual plant values from this Vernal parentage were lower than either parent; less than “% were higher. Table 5. Mean progeny "hemolytic values" of Vernal alfalfa parentage. v__z v 21“ v 122 v 51 y_h_6_ L21 1.“ 7.7 5e5 300 A.“ 309 V 7 v 21“ 2“.6 8.0 6.3 “.7 7.9 v 132 6.3 6.3 3.6 5.5 v 51 7.0 3.9 7.0 V “6 1.3 3.6 V 37 8.3 ' ‘. “#1, ,“4- -."-l- .4-' . “0 This trend did not exist in the Culver population (Table 6). Only one progeny mean value fell outside the parental values, C 58 X C 13, and it was higher than expected. Only 7% of the plant values were lower than either parent, but 28% were higher. Table 6. Mean progeny ”hemolytic values“ of Culver alfalfa parentage. c M. 2.2 c1; c 268 c 58 c an 2.2 3.1 3.1 I.“ 3.0 0'5 7.0 5.9 3.0 “.7 c 13 3.0 2.3 5.0 c 268 t.“ 3.0 C 58 3.9 Progeny of a diallel cross between Vernal plants previously selected for growth characteristics under different fertility conditions formed the basis of a more exhaustive study of saponin inheritance in alfalfa. \Seven parent clones formed the basis of the diallel crosses. VTen.plants of every combination of one-half of this diallel were harvested on June 10, 1968 and leaf-stem separations made. Samples were analysed by Erythrocyte Assay I. The gca of leaves, stems, and whole “1 plant were highly significant (Table 7). In no case was the sea significant. The estimates of heritability were high; loaf 0.903, stem 0.916, and whole plant 0.902. This indicated that the genetic system control- ling saponin levels in alfalfa was expressed equally well in the leaf and stem. Mean progeny leaf, stem, and whole plant "hemolytic values" are reported in (T‘bl. 8) e Table 7. Analysis of variance and estimates of saponin heritability in leaf, stem, and whole samples of a diallel cross between seven Vernal clones. Source df SS MS leaf stem -whole leaf stem whole gca 6 765.10 105.07 37“.37 127.52.. 17.51** 62.“0** sea 21 52.15 7.33 30.87 2.“8 0.35 1.“? error 189 3.00 0.27 1.31 leaf stem whole Estimate of heritability 0.903 0.916 0.902 ** significant at 1% level “2 Table 8. Mean progeny leaf, stem, and whole plant _ ”hemolytic values“ of Vernal alfalfa parentage. Mean progeny leaf "hemolytic values". _ELJLlLil-Lii H1 15.0 H3 1“.6 21.0 H“ 16.2 19.“ 17.0 L1 9.2 1“.8 16.6 8.9 L3 10.2 12.7 11.6 6.5 “.6 L“ 6.0 11.8 7.5 3.3 3.8 1.1 H5 11.9 10.5 10.6 “.“ 3.7 3.5 3.0 Mean progeny stem "hemolytic values". H1 H3 H“ L1 L3 35 H1 5.3 6.9 5.2. 2.7 11.0 2.6 11.11 32.. 7.9 5.“ “.1 3.6 3.8 “.0 H“ '5.8 3.6 3.0 1.7 2.7 L1 3.0 2.0 1.0 1.“ 0.2 0.8 1.0 “3 Table 8 (cont'd.) Mean progeny whole plant ”hemolytic values". .23.. .89.. .83. £1. £2_. LE. 52. H1 10.“ n3 10.0 15.5 H“ 10.7 13.3 11.6 L1 5.9 9.8 10.2 6.8 L3 7.0 8.5 7.7 “.“ 2.8 L“ 3.5 7.9 “.8 2.1 2.“ 0.7 as 8.5 6.9 5.6 2.8 2.“ 2.1 2.0 In this population one-third of the leaves and whole plants, and one-fourth of the stems were higher or lower in saponin content than either of the parents. Unlike the previous Vernal population, leaf, stem, and whole plant values falling outside parental values were equally divided between those which were higher and those which were.lower. Saponin levels of leaf and.stem were significantly correlated (t.01), however, the correlation (0.7“) was less than might be expected (Table 9). “Leaf saponin/ stem saponin“ values ranged from 0.8 - “5. Stem saponins averaged one-third that of the leaves. In spite of the fact that both stem and leaf saponin were significantly .~s_h- .' pl ““ correlated (t.01) with the "leaf/stem ratio”, 0.21 and 0.20 respectively, the correlations were not sufficiently high to hamper selection for leafy plants of low saponin content. Table 9. Correlation coefficients of leaf and stem ”hemolytic values” and leaf/stem ratios. lé‘f Beponin leaf/stem ratio i stem saponin 0.7“** 0.20** 1 leaf saponin 0,21ee ** significant at 1% level Sufficient variation in saponin content existed in all alfalfa populations studied to allow selection for high and low saponin lines. Progeny from plants selected for low or high saponins should be lower or higher in saponins.than the average non-selected population. Fixation of this character at a low level should be rapid if the estimates of heritability were accurate. Likewise, high saponin levels should be realized in progeny of plants selected for high saponin content. These observations were found to be correct. Over 1500 plants of 2nd. cycle saponin selection or unselected populations were analyzed by Erythrocyte Assay III in “5 1969. Most of the plants were greenhouse seedlings resulting from 1968-1969 crosses. Plants were rated as low, medium, or high in saponins on the basis of hemolytic activity exhibited by leaflets from young seedling plants or the 2 - “ leaflets from established plants required to obtain a leaf sample of 23 - 27 mg. Parental combinations and saponin levels of their progeny are presented in Table 10. Unselected Vernal, Culver, and M88 populations had normal saponin distributions when individual plants were classified as low, medium, or high. The DuPuits population was skewed toward high saponins. The distribution of all progeny from selected low and high combinations, however, was very skewed (Figure 6). Nearly 79% of the progeny from selected low plants were classified "low" and over 86% of the progeny from selected high plants were classified ”high". The distribution of saponins in unselected Saranac (Table 10) was skewed to the high side. This is not surprising as Saranac is closely related to DuPuits. The other variety, Team, was developed for weevil resistance. One parent of the California increase had previously been selected for low saponins. This selection was apparently very effective for lowering the level of saponins. 285 270 255 2“0 225 number of plants sues-sens... pee-e" ONUMO‘OWs-e \nOMOUuOVtO 90 75 60 “5 30 15 Figure 6. Progeny of low saponin plants. “6 Progeny of high saponin plants. low medium high :— low medium high a u: C: \n o u. c: \n c u: number of plants WW OVIQUI OUIOUIGM Saponin Content Distribution of progeny from plants selected for low and high saponin. “7 Table 10. Saponin levels of progeny from selected low or high saponin parents, two unselected varieties, and one partially selected alfalfa increase. Source Low Medium High Low saponin--2nd. cycle 01 07v 1 22 2 5 c 2 07v '2' 23 u u c 2 07V 3 20 5 3 v 1 v 2 2“ 3 1 “-1’; v 1“ 11 2 V—é v 20 1 0 v 6 V 1 2“ 3 3 L2 V 10 “ 0 c v 2 c v 25 0 0 c v cév 6 13 1 0 B 1 B 2 22 2 1 B 1 B 3 19 6 1 1% B 19 6 1 §_2 B 2 22 0 1 I._ 4 ._-. _.......1. J...“ . ..- ms “8 Table 10 (cont'd.) High saponin—-2nd. cycle 1.9.! £22.12! ELSE 9.11 B 5 0 2 25 B B E 0 “ 21 V 8 D 1 0 7 22 D 1 C 3 0 2 27 Varieties previously unselected for saponin Saranac 8 23 “6 Team 21 28 53 Alfalfa increase partially selected for saponin L“909 F2 increase 266 56 39 N 529 V - Vernal D - DuPuits C - Culver B - MSB “9 Some Properties of Alfalfa Saponins Due to the diversity of compounds classified as saponins, different biological activities and chemical properties would be expected of saponins from various sources. Cholesterol influences the hemolytic activity of some saponins, but not all. Heat is also variable in its influence on hemolytic activity of different saponins. It is possible that the effect of these two factors may even vary from plant to plant in a hetero- zygous polyploid plant such as alfalfa. Crude saponin extracts from nineteen Vernal leaf and stem samples (1st. crop 1968) were analyzed by Erythrocyte Assay I for heat lability and the forming of cholesterides (Table 11). The samples were drawn from six crosses designated A - F. Individuals within these crosses were recorded as subscripts. Cholesterol, when added as 2% of the crude saponin extract and heated momentarily to 80 C, effectively decreased the hemolytic activity to one-tenth its original value. Heating the crude saponin extracts at 120 C for 20 min did not decrease hemolytic activity. Hemolytic activity of low saponin samples actually increased under these conditions. Although hemolytic activity of all alfalfa saponin extracts studied was affected only slightly by heat, other biological activities of saponins may be affected to a greater extent. This was demonstrated in an 50 Table 11. Effect of cholesterol and heat treatment on the hemolytic activity of crude saponin extracts of Vernal alfalfa stem and leaf samples. S2252; "Hemolytic value" extract extract + extract + heat cholesterol A L 29 29 1.5 81L 29 27 1.5 01L 2? 23 1.5 ch«l/ 27 23 1.5 325 1“ 17 1.5 31L 9 9 1.5 D;L 8 8 0.8 DzL 8 11 0.8 033 7 7 1.5 C45 7 7 0.8 B38 7 7 0.6 22L 3 “ 0.“ F L 2.8 3 0.5 D38 1.6 2 0.3 has 1.6 2 0.3 E3L 1.1 3 0.3 EuL 1.2 3 0.0 '25s 0.3 0.7 0.0 253 0.3 0.5 0.0 51 Table 11 (cont'd.) L - leaf 8 - stem 1] Source materials identified with the same letter are different plants resulting from the same cross. experiment in which Trichoderma gp. grew normally in the presence of some crude saponin extracts treated with heat but grew only poorly in other heat treated extracts.- Alfalfa saponins are reasonably stable; both in the plant and in water extracts. Plants harvested and stored under a variety of conditions retained their original hemolytic activity. Drying temperature; speed of drying; and storage temperature, light, and period did not cause major changes in hemolytic activity of saponin extracts. Fresh plant samples were stored at 5 C for over one week with no deleterious effect on the contained saponins. Uater extracts of crude saponin were stored for five days in waxed paper cups at 5 C with no change in hemolytic activity. The activity of extracted saponins is apparently not altered over a period of time by other factors in the extract. The effect that extracts from different plants might have upon extracted saponins was studied in two experiments (Table 12). Crude extracts from different plants were mixed in the proportions 3:1, 111, or 113 and hemolytic activity of the mixture, based on 52 Expected and actual hemolytic activity of crude saponin extracts of alfalfa whole plant, stem, or leaf samples. T.b1° 12 e Experiment 1 Expected Plant ”Hemolytic value” "Hemolytic value” A 18.0 B 1.3 c 13.9 1/“A+3/“B “.3 5.“ 1/2A+1/28 8.9 9.6 3/“A+1/“B 11.0 13.8 1/ZB+1/20 6.“ 7.6 D 1.3 E 17.0 r 39.0 1/“D+3/“E 11.0 13.1 1/2D+1/2E 7.0 9.2 3/“D+1/“E 3.8 5.3 1/“D+3/“r 30.0 30.9 1/2D+1/2F 19.0 20.“ 3/“D+1/“F 9.8 10.9 1/“s+3/“r 29.1 33.9 1/2s+1/2r 27.1 28.2 3/“s+1/“r 19.0 22.6 Experiment 2--source material was the same as reported in Table 12. 1/2A L+1/2C1 L 29.0 28.0 1/2A L+1/ZE5 s 1“.0 1“.6 1/2A L+1/2E3 L 1“.0 15.0 1/2A L+1/2D3 S 1“.o 15.3 1/2c; L+1/2C3 s 15.0 17.0 1/283 S+1/2B1 L 15.0 18.0 1/222 L+1/202 L 15.0 15.0 1/233 S+1/232 s 10.0 10.5 1/282 S+1/2E5 s 7.0 7.1 1/232 S+1/2Dg s 7.0 7.8 1/221 L+1/2Eu L 3.8 5.1 1/221 L+1/2E5 s “.2 “.6 1/233 s+1/2r L 3.“ “.9 1/225 s+1/2D, L “.6 “.1 l/ZDZ 1.1-1,233 L 3.0 ‘+05 53 Erythrocyte Assay I, was compared with the expected activity. Expected activity was calculated only on the basis of dilution. The mixture of two extracts possess- ing ”hemolytic values" 15 and 5, if mixed in equal portions, should be 10. In the first experiment ”hemolytic values“ of the mixtures were slightly, but significantly (t.01) lower than expected values. This difference could have i been due to factor(s) in the extract of lower activity which decreased the activity of the higher. It could also have been due to the loss of some saponins during the mixing of extracts. The surface active nature of saponins have created problems of this type in the past. Greater care was taken to protect against transfer losses of saponins in the second experiment. The hemolytic activity of the mixtures remained consistently and significantly (t.05) lower than expected. These differences, however, were smaller than those of exper- iment 1 and were not of an important magnitude. The discrepancy between actual and expected ”hemolytic values“ may be explained as the loss of saponins due to lack of perfect transfer procedures. ' Mixtures of crude saponin leaf and stem extracts, when properly reconstituted to represent the whole plant, exhibited the same hemolytic activity as the whole plant. Mixtures of leaf extracts from closely or more distantly related plants had little or no effect on their 5“ combined expected hemolytic activity. The same lack of hemolytic stimulation or inhibition of mixtures was observed for stem X stem and stem X leaf mixtures. Seasonal Trend andgéreas of Topgrowth Concentration of Alfalfa Saponins Selected alfalfa plants grown in the field or greenhouse under a variety of environmental conditions were consistently placed in the same order on the basis of saponin content. This was the case irrespective of the type of sample preparation. In order to study seasonal trend of saponin levels in different alfalfa genotypes several clones of four '5. falcata plants, previously selected only for other quality factors, were analyzed for saponins over a two-month period of 1968 first growth. As many as three plants were harvested on the first sampling date and as few as one on the last date. Leaf-stem separa- tions were made and the dried and ground samples analyzed for saponins by Erythrocyte Assay I. The four selected plants contained "medium to medium-high” levels of saponins and although P 21 and P “9 generally contained less saponins than P “8 and P 11 (Table 13) the differences were not of an important magnitude. Leaf and stem samples from the four selections exhibited somewhat different trend patterns of saponins over the two-month period. "Hemolytic values" which 55 Table 13. Leaf and stem "hemolytic values“ from four asexually propagated‘g. falcata plants harvested over a two-month period. Source Harvest date 1968 5-“ 5-1“—' 5-257 6-3 6-13 6-23 3"'7S“ P 11 leaf 12.5 7.0 16.0 1“.5 13.7 13.9 9.8 stem “.2 2.6 “.6 3.0 3.0 2.6 3.0 P 21 leaf 7.0 7.0 12.5 7.5 11.7 13.9 13.9 stem “.2 3.“ 3.0 2.6 1.6 1.5 3.0 P “8 10.11. 11.3 9.8 12.5 lass 13.7 lSeO 8e9 stem 6.“ “.6 3.8 3.6 3.1 “.2 3.0 2.22 leaf 7.0 9.8 10.8 9.3 11.3 9.8 7.0 stem 1.3 2.3 1.5 1.5 1.5 1.5 1.5 Stage of maturity vegetative 5-“, early bud 5-2“ mid bud 6-3 1st. flower 6-13 full flower 6-23 early seed pod7-“ 5-1“ — 3.3—“ ‘5‘; 'm‘- I 1 56 differed by as much as 1.6 were significantly different (Tukey's w-procedure), however, differences of this magnitude were not considered meaningful or important. There was no significant difference in Fhemolytic values” between clones of the same selection on any given harvest date. An ”over-all trend" in alfalfa saponin content from very early growth through full flower was apparent when the average leaf and stem "hemolytic values“ were taken for each date (Figure 7). Ten days after first harvest the average "hemolytic value“ decreased from 9.3 to 8.“. From May 2“ - June 23 saponin content remained relatively high and finally decreased on the last harvest date, July “. Average stem saponin levels remained rather constant throughout the sampling period, however, there was a trend toward decreasing levels from first to last harvest. The leafsstem nhemolytic value" ratio increased from first harvest through June 13 and then decreased at the last harvest. Ratio values from first through last harvest were 2.3, 2.6, “.1, “.3, 5.5, 5.3, and 3.8 respectively. Unlike other factors contributing to forage quality, saponin levels remained rather constant throughout first growth. During the same period percent leaves decreased from 72 to “2 and whole plant cell wall constituents in- creased from 2“ to 51 %; lignin increased from 3.7 to 9.0%, and cellulose increased from 15% to 32%. 57 "Hemolytic Values" 3 G .n .. no a-.. fiancee V. smu.: mum: mum mu.u mumu us: moaowwnuo «Hose 0% 3. anyone» Hon e.n m > a n.m s..m.m c.m n > c.n : > o o.m m.o. c.~ m > o.o~ n > m o.m a.m.s.n. o.~ u > c.o~ N o : o.e. a..a..m..a..n. 6.: . o o.n~ _ > n o.e~ -.a~.m~.au.mm.su o.n~ . > o.: u o N o.~n on.mn.mn..n c.n~ . > c.5— oV> u codascdnaeo emsnoss one canvas easscu>avna cached sends» edahaoaoms senses acsflm nodcscdnaoo .auconsa codes one Scene on» he one semen» coboalhanou Scam sobeoa mo seesasb edahuoaon HH heae< cahoonnshnms .m— canes 63 ml.-. lltfillllvulvlttti . . -1. (a i! m o.m. o..eu o.ou e. > o.» .. > :. c.n. ou....o e.c~ o. > 6.. w > n. e.o m.o.~.o.n.e.o.. n.. o > . o.:. N > u. m.°. n..m.m.o. e.a. om» o.n a > .. c.e. e..n..m. o.a. om> c.eu e > c. a.n. n..n..n. 6.: . o e.a. oV> a c.:. w..~. o.a. om> e... a o m EOdvdfldn—BOO oneness one cased) unencubdvnd unseen sends» odamnolems senses ocean coaaacanaoo ..e.»=.o. n. oases 6“ An exception to this specific.stock effect is demonstrated in combinations 8, 10, and 11 in which 'V/C' is the common stock. In one case the average leaf ”hemolytic value" of the scion is lower than either parent, however, in the other two combinations the scion values are intermediate to the parents. Uhen “VIC” was the scion source (combination 9) "hemolytic values“ were again intermediate. The only other apparent specific stock effect was that of ”V 8' in combination 12. In those combinations in which scion leaves were intermediate in "hemolytic values" to either parent neither the stock nor scion exhibited a controlling force. Scion leaf “hemolytic values” were, in most cases, near the average of the parents. The question of which plant organ, root or leaf, might be the more logical area of alfalfa saponin synthesis was not resolved. Both the root and leaf probably produce saponins. If only one of these organs produced saponin the distribution of these compounds in the leaves of high and low scions would have been much different. In this study leaves of scions were as much as “.5 times as active as leaves from the stock parent and as much as 13 times as active as leaves from the scion parent, but still intermediate in activity to the parents. - . .. .. . .w ix‘.::x‘-__.. I ‘7“"- "1‘ r. *' M” .w 65 Also, leaves of scions were less than one-half as active as leaves from the stock parent in certain instances and nearly one-half as active as the leaves from the scion parent in others, yet intermediate in activity to the parents. The graft itself may have increased hemolytic activity somewhat. Scions gflafted directly back ?w to their parent produced leaves with hemolytic % fi« activity somewhat greater than the parent. This % increase was as great as 10 "hemolytic value units,” however, it never.changed.a.low value into a medium or a medium to a high saponin value. This increase and other small increases of leaf hemolytic activity over parent plants may be due.to a younger average age of scion leaves than parent leaves. Another variable which may have influenced the hemolytic activity of scion leaves was the traversing of a graft by saponins. This variable, however, could not eXplain all cases of the diverse results. On the bases of the data presented and the above discussion, this.study of organ of synthesis should serve primarily as added direction for future inves- tigations. l. 2. 3. 5. 7. CONCLUSIONS Hemolytic activity of saponins in alfalfa may be used as the basis of assays of hemolytic saponin content of individual alfalfa plants Erythrocyte Assay I, an adaptation of other hem- olytic assays, is effective for rapidly and accurate- ly evaluating the saponin content of dried, fresh, or frozen alfalfa samples. Erythrocyte Assays II and III were further adapta- tions of Erythrocyte Assay I and allowed evaluations of small samples and large numbers of samples. Chemical.and biological (fish assay) evaluations agree with.erythrocyte assay results. Accelerators or inhibitors of hemolysis, if present in the reaction mixture, do not alter the realia- bility of the three erythrocyte assays. Alfalfa.saponins change very little in hemolytic activity under various storage conditions. Differences of saponin content within alfalfa species is much greater than between species. General combining ability for saponins is highly significant in g. sativa populations. 66 9. 10. ll. 12. 13. In. 15. ‘6. l7. 67 Variation in alfalfa saponin content within populations and saponin heritability are suffic- iently large to facilitate selection for high and low saponin lines. Alfalfa leaves contain about three times as much saponins as stems. Estimates of heritability of g. sativa leaf and stem saponins are nearly identical. Saponins of alfalfa form cholesterides resulting in a major decrease in.hemolytic activity. Based on their hemolytic activity, alfalfa saponins are not heat labile. The hemolytic activity of mixed alfalfa plant extracts is not greatly different from what would be expected on a strict dilution basis. Leaf and stem‘!. falcata saponin levels do not change greatly over a twodmenth first growth period. There is a trend toward higher saponin content in younger.alfalfa leaves and stems than in older.’ Both alfalfa leaves and roots are likely areas of saponin synthesis. 1. 7. 9. LITERATURE CITED Amarasingham, R. D., N. G. Bisset, A. H. Millard, and M. 0. Woods. 196“. A phytochemical survey of Malaya. Part III. Alkaloids and saponins. Economic Bot. 18:270-278. Awe, V., and H. Hausermann. 1950. .Vertbestimmung von saponinhaltigen pflansenaussugen durch messung der oberflichenspannung im vorgleich mit der ermittlung des index haemolyticus initialis. Arch. Pharm. U. Ber. Deutsch. Pharm. Ges. 283(55):7-25. Biological Abstracts 25:8h339071. B‘kCr’ Es As. Je To mrtin, ‘nd Aphrt Po Hillone 1966. The distribution of diosgenin in Dioscorea 322. Ann. Appl. Biol. 58:203-211. Basu, N., and R. P. Rastogi. 1967. Triterpenoid saponins and sapogenins. Phytochemistry 63 Bennett, Raymond D., and Erich Heftmann. 1965. Biosynthesis of Dioscorea sapogenins from cholesterol. Phytochemistry “3577-586. Bickoff, E. M., A. N. Booth, 3. L. Lyman, A. L. Livingston, C. R. Thompson, and F. DeEds. 1957. Coumesterol, a new estrogen isolated from forage crops. Science 126:969-970. , R. R. Spencer, S. C. Witt, B. E. Knuckles, and J. B. Stark. 1968. Purine derivatives in alfalfa as growth stimulants for Bacillus subtilis. J. Agr. Food Chem. 16(253253-251. Birk, Yehudith, A. Bondi, B. Gestetner, and I. Ishaaya. 1963. A thermostable haemolytic factor in soybeans. Nature 19721089-1090. Bonner, James, and J. E. Varner. 1966. Plant biochemistry. Academic press; New York. 68 10. 11. 12. 13. I“. 15. 16. 17. 18. 19. 20. 59 Buker, Robert Joe. 1963. General and specific combining ability in alfalfa. Ph. D Thesis, Purdue University. Coulson, C. B. 1957. Fractionation of isolated lucerne and other triterpenoid saponins. Nature 180:1297-1298. , and T. Davis. 1962. Saponins II. Fractionation and pharmacological properties of lucerne saponins. Possible relation of these to bloat. J. Sci. Food Agr. 13:53-57. Daday, H. 1967. Heritability and genotypic and environmental correlations of creeping root and persistency in Medicago sativa 5. Aust. J. Agric. Res. 19:27-§FT_— Dollahite, James V., Ted Shaver, and Bonnie J. Camp. 1962. Injected saponins as abort- ifacients. Amer. Jour. Vet. Res. 23:1261-1263. Drosds, B. 196“. Variations in the saponin content and in the hemolytic activity in Saponaria officinalis, Primula officinalis, Polemonium caeruleum, and Glzcyrrhisa glabra. Pharmasie 191553533-5h0. Chemical abstracts 61:12319F. Elbel, G., and F. J. Holser. 193E. Bestehen gruppenspesifische unterschiede im verhalten der blutkgrperchen gegen hlmolytika. Zeitschr. Immunitfitsforsch. 82(3):175-178. Chemical Abstracts 28:5522(h). Fischer, Robert, and maria Vybiral. 1951. Zur diffusion von saponin aus der pflansenselle. Phyton (Ann. Roi. Bot.) 3(3/h):232-2b1. George, A. J. 1965. Legal status and toxicity of saponins. Ed. Cosmet. Toxicol. 3:85-91. Gestetner, B., Yehudith Birk, A. Bendi, and Y. Tencer. 1966. Soya bean saponins VII. A method for the determination of sapogenin and saponin contents in soya beans. Phytochemistry 58803-806. Gil, Hernan Chaverra, R. L. Davis, and R. F. Barnes. 1967. Inheritance of in-vitro digestibility and associated characteristics in medicago sativa‘g. Crop Sci. 7:19-21. 21. 22. 23. 2h. 25. 26. 27. 28. 29. 30. 7O Gladkikh, A. S. 1965. Vliyanie saponina is tsiklamena grusinskogo na rasvitic eksperimental' nogo aterosklerosa u krolikov. Farmakol Toksikol 28(2):1h7-152. Biological Abstracts #732315327681. , I. A. Gubanov, and A. A. Meshcheryakov. 1965. Saponin content in plants found in Turkmenia. Isv. akad. nauk. Turkm. SSR Sor. Biol. Nauk. 1:22-35. Chemical abstracts 66: 10h853112925z. Gorter, E., F. Grendel, and H. A. Seder. 1931. Saponin hemolysis. Proc. Acad. Sci. Amsterdam 3hxh71-h73. Chemical Abstracts 25:“290. Gorz, H. J., and F. A. Haskins. 1962. Trans- location of coumarin across a graft union in sweetclover. Crop Sci. 23255-257. Granick, S. 19h9. The chemistry and functioning of the mammalian erythrocyte. Blood “snob-hbl. Griffing, Bruce. 1956. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. of Biol. Sci. 9xh63-b93. Guensi, V. D., U. R. Kehr, and T. M. McCalla. 196k. Water-soluble phytotoxic substances in alfalfa forage. Variation with variety, cutting, year, and stage of growth. Agron. Jo 56 81399-500. Gutierrez, Jose, and R. E. Davis. 1959. Character- istics of saponin-utilising bacteria from rumen of cattle. Appl. Microbiol. 7(5):30h-3oa. Hansen, 0. H., G. O. Kohler, J. V. Dudley, E. L. Sorensen, G. R. Van Atta, I. H. Taylor, M. V. Pedersen, H. L. Carnahan, C. P. Vilsio, V. R. Iehr, C. C. Lowe, E. H. Stanford, and J. A. Yungen. 1963. Saponin content of alfalfa as related to location, cutting, variety, and other variables. USDA Res. Report, ABS 3h-hh. Heftmann, E. 1967. Biochemistry of steroidal saponins and glycoalkaloids. Lloydia 30(3): 209-230. among-é h: 31. 32. 33. 3h. 35. 36. 37. 33. 39. ho. 71 , Re Re LiebOr, find Re De mnn’tte 1937. Biosynthesis of tomatidine from cholesterol in L co rsicon pimpinellifolium. Phytochemistry 3225-229. Killer, K., M. Keipert, and B. Linzer. 1966. Triterpensaponine. Pharmazie 21(12):?13- 751. Rorvath, Antonio, Francisco Alvardo, Joseph Szocs, Zoila Ney do Alvarado, and Gilberto Padilla. 1967. Metabolic effects of calagual- ine, an antitumoral saponin of Pol odium leucotomas. Nature 21h31256-1258. Ishaaya, Isaac, and Yehudith Birk. 1965. Soybean saponins IV. Effect of proteins on the inhibitory activity of soybean saponins on certain enzymes. J. Food Sci. 30(1):118-120. Ishikawa, Yoshinori, and Yataro Obata. 1960. Inhibition pattern of beet-saponin on amylose formation by potato phosphorylase. Bull. Agr. Chem. Soc. Japan 2&3699-702. Biological Abstracts 3683153338620. Jackson, Horace D., and Ralph A. Shaw. 1959. Chemical and biological properties of a respiratory inhibitor from alfalfa saponins. Arch. Biochem. Biophys. 8h(2):h11-hl6. Jacobson, C. A. 1919. Alfalfa saponin: Alfalfa investigation VII. J. Amer. Chem. Soc. ' hlx6bo-6h8. Jaretzky, Robert. 1900. Uber saponinvorkommen bei arten der gattatg Medicago. Angewandte Bet. 22(2):1b7-156. Biological Abstracts 153227032h337. Joos, P. 1966. Ideal behavior and interaction of adsorbed surfactants at the interface. Medad. Kon. Vlaam. Acad. Netonsch. Belg., x1. Vetensch. 28(3). , and R. Ruyssen. 1967. Spread mixed monolayer of senegin and seneginin with cholesterol at the air-water interface. Bulls SOCe Chill. 80180.. 76(5-6)3308-3I5e Chemical Abstracts 6736637370619r. 72 “1. Jung, F., and Lotteliese Nirth. 1950. Saponin hemolysis II. Arch. Exptl. Path. Pharmakol. 210:328-335. Chemical Abstracts “5:526“h. “2. Kahl, Uladyslaw, Krystyna Huszlik-Jochym, and Jan Starzyk. 1966. Microbiological investigations of the components of horse-chestnut Aesculus hippocastanum 5. seeds; I. Mycostatic action of the raw extract from the horse- chestnut seeds and of the saponin glycoside aescin. Acta. Biol. Cracov. Ser. Bot. 9(2):115-12“. Biological Abstracts “9: 37“;“087. “3e KQIQ‘r, So Veg Ye Po Gupt‘, Re Ks M‘h‘J‘n’ Ge Se Joshi, and A. L. Rampal. 1968. Synthetic studies in steroidal sapogenins and alkaloids V. Synthesis of kryptogenin, diosgenin and yanogenin. Tetrahedron 2“(2):899-90“. ““. Kesten, H. D., and T. F. Zucker. 1928. Saponin hemolysis of normal human blood with some observations on anemia blood. Am. J. Physiol. 87:27“-279. ‘ “5. Kolodzieske, Jozef, and Leokadia Stecka. 1965. Occurrence of saponins in various parts of Gypsophila paniculata and‘g. altissima. Farmac. polska 213751-75“. Chemical Abstracts 6“:20196e. “6. Kupchan, Morris 8., Richard J. Hemingway, John R. Knox, Stanley J. Barboutis, Dieter Verner, and Maureen A. Barboutis. 1967. Tumor inhibitors XXI. Active principles of Acer no undo and C clamen pgrsicum. J. Pharm. Sci. 56(5):603-60. ~ “7. Kwasniewski, Victor. 1958. Discovery of a saponin in celandine, Chelidonium maJus L. Arch. Pharm. 291(“):209-211. Chemical Abstracts 5381630h. “8. Lindahl., Ivan L., R. E. Davis, R. T. Tertell, G. E. Uhitmore, R. U. Dougherty, U. T. Shalkep, C. R. Thompson, G. R. Van Atta, E. M. Bickoff, E. D. Walter, A. G. Livingston, Jack Guggolz, Robert H. Wilson, Martin B. Sideman, and Floyd DeEds. 1957. Alfalfa saponins. Studies on their chemical, , pharmacological, and physiological preperties in relation to ruminant bloat. USDA Tech. Bull. 1161,1-83. --‘ l‘l“W“"_.’-l‘+m .- .. . I I .'~ 'm 73 “9. Lindner, N. 19“6. The occurrence of saponins in Leguminosae. Pharmazie 1(“)3177-182. Biological Abstracts 2531105311887. 50. Lourens, V. A., and M. B. O'Donovan. 1961. Studies on the lucerne saponins. I. The isolation of saponins. South African Journal of Agricultural Sci. “(2)3151-156. 51. Macovski, Eugen, and Adela Stancu. l9“2. Untersuchungen uber die permeabilitat lebender membranen. X. Prufung der vergiftungsleichung (C-Co)(f- O-m)=X an der saponinpenetration in das kiemenepithelium von Alburnus lucidus Heck. Biochem. 2. 31332502 265. Biological Abstracts 19313“5312“22. 52e mrtin, I. We, ‘nd Ee C‘b‘nill‘ae 1967s Herit- ability of yields in Dioscorea floribunda. Trop. Agric. Trin. “5(15855-51- 53. McArthur, J. M., J. E. Miltimore, and M. J. Pratt. 196“. Bloat investigations. The foam stabilizing protein of alfalfa. Can. J. Animal Sci. ““3200-206. ‘ 5“. McNair, James F. 1932. Some properties of plant substances in relation to climate of habitat- volatilo oils, saponins, cyanogenetic glucosides, and carbohydrates. Amer. Jour. Bate 19(2)8178-193e 55. Megie, Christian A., R. U. Pearson, and A. E. Hiltbold. 1967. Toxicity of decomposing crop residues to cotton germination and seedling growth. Agron. J. 593197-199. 56. Meyer, J. R. 19“2. Processo simples para se verificar a presenca de saponina nos extratos aquosos de plantas ichthiotoxicas. Biologico 8(12)3283-286. Biological Abstracts 1732“2“32509“. 57. Mircevova, L., A. Simonoya, C. Michalec, and z. Holman. 1968. The influence of hemolysis on the AtPase activity and the composition of the erythrocyte membrane (human). Enzymol. Biol. Clin. 9(1)331-“0. Biological Abstracts “9399033109606. d“ .~ 3...; ”I“ e L I'ln‘l num‘._r1‘u\“'e' , I 58. 59- 60. 61. 62. 63. 6“. 65. 66. 67. 7“ Morris, R. J., and E. N. Hussey. 1965. A natural glycoside of medicagenic acid. An alfalfa blossom saponin. J. Org. Chem. 30(1)8166-168o , N. B. Dye, and P. S. Gisler. 1961. Isolation, purification, and structural identity of an alfalfa root saponin. J. Org. Chem. 26312“1-12“3. Nielsen, K. F., T. F. Cuddy, and U. B. Woods. 1959. The influence of the extract of some crops and soil residues on germination and growth. Canadian Jour. of Plant Sci. xi “03188-197. é: A Pedersen, M. N. 1965. Effect of alfalfa saponin on cotton seed germination. Agron. J. 57‘516-517e i F x , De Eo Zimmer, De Re MCAlliltOr’ Jo 0e :.{ Anderson, M. D. Wilding, G. A. Taylor, and g} C. F. McGuire. 1967. Comparative studies of saponin of several alfalfa varieties using chemical and biochemical assays. Crap. Sci. 7’3u9-352o Ponder, Eric. 19“8. Hemolysis and related phenomena. Grune and Stratton Inc., New York, New York. Porceddu, E. 1968. Stima degli effetti dell' attitudine generals e specifica alla combin- azione in incroci tra cloni di erba medica. Genetica Agraria 22389-102. Priess, Hans. 1911. Plant lactones as fish poisons Ber. Pharm. Geo. 213267-270. Chemical Abstracts 53332“. Richou, R., P. Lallouette, R. Jensen, and Cl. Bolin. 1965. Research on saponin3 an adjuvant substance and stimulant of immunity. Rev. Immunol. 29(“-5)3205-219. Chemical Abstracts 6“311692c. Roy, A., and M. L. Chatterjee. 1963. Pharma- cological study of saponins. Bull. Calcutta Sch. Trop. Med. 11(“)3156. Biological Abstracts “53625337811“. 68. 69. 70. 71. 72. 73. 7“. 75- 76. 77- 75 Ruyssen, R., and J. Huble. 19“6. Hemolysis by saponins. Bull. Soc. Chim. Biol. 283 532-5“O. Chemical Abstracts “133518h. Sanchez-Marroquin, Carlos Larios, and Pilar Fernandez. 1967. Microorganisms aislados de ensilados do maguey Agave atrovirens Xarev. que utilizan saponinas. Ciencia '25(5)3173-176. Biological abstracts “9352602. Scardavi, Anna, and Fred C. Elliott. 1967. A review of saponins in alfalfa and their bioassay utilizing Trichoderma s . Mich. Agr. Expt. Sta. Quart. Bul. 50 2)3163-177. Schmidt-Theme, Josef, and Friedrich Prediger. 1950. Hemolysis with saponins. z. Physiol. Chem. 2863127-138. Chemical Abstracts “533“““f. Singleton, V. L., E. T. Mertz, and R. L. Davis. 1952. The hydrolysis and amino acid assay of alfalfa and the methionine range in 100 selections. Agron. J. ““33“6-3“8. Sokolov, S. Ya. 1965. Vliyanie saponinov aralii man'chzhurskoi na elektricheskuyu aktivnost'- golovnogo mozga. Byul. Eksp. 8161. Med. 60(8)373-77. Biological Abstracts “73 ““88352653. Sollmann, Torald. 19“2. A manual of pharmacology and its applications to therapeutics and toxicology. N. B. Saunders Company. Philadelphia. pp. 556-559. Stifel, Fe Bo, Bo Le Vottor, Be So All‘n, ‘nd Ho To Horner Jr. 1968. Chemical and ultrastructural relationships between alfalfa leaf chloroplasts and bloat. Phytochemistry 73355-36“. Stuthman, n. D., a. L. Davis, and Martin Stob. 1967. Combining ability of six Medicago sativa L. clones for uterotrophic activity. Crop Science 73119-121. Tang, Yun-An. 1961. The use of saponin to control predaceous fishes in shrimp ponds. Progressive Fish Culturist 233“3-“5. Biological Abstracts 363216“326296. ‘. n.muvu.a’.pu. ‘ -1 LL.‘ «file—J, 73. 79. 80. 81. 82. 83. 8“. 85. 86. 87. 76 Tschesche, R., G. Uulff. 1965. Uber die anti- mikrobielle wirksamkeit von saponinen. Z Naturforsch 206(6)35“3-5“6. Biological Abstracts “736308373782. Vacek, Lubor, and Bretislav Sedl‘k. 1962. Comparison.of physical, chemical, and bio- logical properties of saponins. Gunma Jour. Med. Sci. 11(1)31-6. Biological Abstracts “3310“3677. Van Atta, G. R, and Jack Guggolz. 1958. Detection of saponins and sapogenins on paper chromat- ograms by Liebermann-Burchard reagent. Agricultural and Food Chemistry 6(11)3 8“9-850. , and C. Ray Thompson. 1961. Determination of saponins in alfalfa. J. Agr. Food Chemistry 9377- 79. Vendrig, J. C. 196“. Growth-regulating activity of some saponins. Nature 20331301-1302. Von Hlumencron,.Uilhelm. 19“1. Vorgleichende untersuchung der wirkung verschiedener saponins auf das froschhorz. 19“1. Arch. Exptl. Path. Pharmakol. 197358“-589. Wall, Monroe E., J. N. Garvin, J. J. Uillaman, Ghentin Jones, and Bernice G. Schubert. 1961. Survey of plants for steroidal sapogenins and other constituents. J. Pharmaceutical Sci. 5031001-103“. , Merle M. Krider, Edward S. Rothman, and C. Roland Eddy. 1952. Steroidal sapo- genins. I Extraction, isolation, and ident- 1f1°.t1°no Jo 8101 Chemo 1983533-5u3e V‘lt.r’ Re D., Go Re V‘nAtt‘, Co Re Thomp.°n’ and N. D. Maclay. 195“. Alfalfa saponin. J. Am. Chem. Soc. 7632271-2273. Uasicky, Richard, and Clara Ferroira. 19“9. As saponinas da raiz de Brodemoyera flori- bunda Nilld, droga da medicine popular brasil- eira. Anals Fac. Farm. E Odontl. Univ. Sao nfiuic 7:3»1-350. Biological Abstracts 25319513215“O. 88. 89. 90- 91. 77 Uolters, Bruno. 1968. Saponins als pflanzliche pilzabwehrstoffe. Zur antibiotischen wirkung von saponinen III. Planta 79(1)377-83. Yefimova, T. H., H. P. Pivnenko, V. A. Nutsevych, and N. Ya. Zykova. 1966. Vplyv saponiniv oplodnya myl'noho dereva na krov'ya nyitysk i kholesterynemiyu tvaryn. Farm 2h. 21(6)3 “5-“9. Biological Abstracts “931690319003. Zentmeyer, G. A. and C. R. Thompson. 1967. The effect of saponins from alfalfa on Phyto- thora cinnamomi in relation to control of root rot of avocado. Phytopathology 57(11)31278-1279. Zimmer, D. E., M. U. Pedersen, and C. F. McGuire. 1957. A bioassay for alfalfa saponins using the fungus, Trichoderma viride Pers. ex. Fr. Crop Sci. 73223-22“. ___-...__.__ anal-_— . ..—-.........—.. . . L, 1 ‘ .J HICHIGRN STnTE UNIV. LIBRQRIES (I?)IIIWIWIIWIWWWWWWIVW’HIIHI 31293104579242