Louisiana State University LSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1991 The Alkaloids of Erythrina: Clonal Evaluation and Metabolic Fate. Lori Denise Payne Louisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

Recommended Citation Payne, Lori Denise, "The Alkaloids of Erythrina: Clonal Evaluation and Metabolic Fate." (1991). LSU Historical Dissertations and Theses. 5268. https://digitalcommons.lsu.edu/gradschool_disstheses/5268

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely afreet reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Ben & Howell Information Company 300 North Zeeb Road Ann Arbor Ml 48106-1346 USA 313 761-4700 800 521-0600 Order Number 0210668

The alkaloids ofErythrtnai Clonal evaluation and metabolic fate

Payne, Lori Denise, Ph.D. The Louisiana State University and Agricultural and Mechanical Col., 1091

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 THE ALKALOIDS OF ERYTHRINA: CLONAL EVALUATION AND METABOLIC FATE

A Dissertation Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Chemis&y

by Lon Denise Payne B.S., University of California at Davis, 1982 December 1991 ACKNOWLEDGEMENTS

There are many people I would like to take this opportunity to thank for contributing to the success of this project and for making my time at LSU enjoyable. First I would like to thank my research advisor, Dr. Joe P. Foley, for allowing me to do this work and for his flexibility and support during the research. I appreciate the time my fellow group members, Skip Little, Jeff Crow, Mark Jeansonne, Rene Arenas and Eric Ahuja, took in discussing research and practical problems of the laboratory, Jeff Corkern and Roberto Wong were instrumental in the GC/MS and LC/MS work respectively. Financial support in the form of a teaching assistantship from the Department of Chemistry is acknowledged. I thank Ravi Laxman for his prodding and los muchachos at CAT1E, especially Frank Lopez and Gerardo Rodriqu6z, for informative discussions and sample handling. I also acknowledge the help of the Interlibrary Loan Service in attaining several important scientific papers. I am grateful to the members of my dissertation committee, especially Dr. Nikolaus Fischer and Dr. Barbara Shane for their constructive comments. Gifts of B-erythroidine from Dr. V. Boekelheide of the University of Oregon, erysodine and erysovine from Dr. A. H. Jackson of the University of Cardiff, and erythraline and erythratine from Dr. M. Jui-ichi of the Mukogawa Women’s University in Japan are gratefully acknowledged. And finally I would like to thank my family, particularly my husband Alejandro Dittel and my parents, Walt and Sue Payne, for their emotional and financial support and immense patience.

ii TABLE OF CONTENTS

C&gfi ACKNOWLEDGMENTS ...... u LIST OF TABLES ...... vi LIST OF FIGURES ...... vii LIST OF ABBREVIATIONS ...... ix ABSTRACT ...... xii CHAPTER 1 INTRODUCTION AND OBJECTIVES ...... 1 1.1 Introduction ...... 2 1.2 The Genus Erythrina and its Traditional Uses ...... 5 1.3 Erythrina as Forage ...... 6 1.4 Antiqualitative Agents in Erythrina ...... 7 1.5 Objectives and Overview ...... 7 2 REVIEW OF LITERATURE ...... 9 2.1 Erythrina Alkaloids: Occurrence, Isolation and Identification ...... 10 2.1.1 Folkers, Unna, Majors and Koniuszy ...... 10 2.1.2 Deulofeu ...... 14 2.1.3 Boekelheide ...... 15 2.1.4 Other Studies on Structure Elucidation ...... 15 2.1.5 Occurrence of Erythrina Alkaloids ...... 16 2.1.6 Jackson, Chawla and Coworkers ...... 20 2.2 Erythrina Alkaloids: Biosynthesis ...... 21 2.2.1 Early Studies ...... 21 2.2.2 Barton and Coworkers ...... 22 2.3 Erythrina Alkaloids: Synthesis ...... 27 2.4 Erythrina Alkaloids: Pharmacology and Toxicology .. 28 2.4.1 Research at Merck ...... 29 2.4.2 Clinical Studies ...... 31 2.4.3 Other Pharmacological Studies ...... 32 2.5 Summary ...... 34

iii 3 SAMPLE COLLECTION, STATISTICAL ANALYSIS AND LIQUID CHROMATOGRAPHIC TECHNIQUES .... 35 3.1 Introduction ...... 36 3.2 Experimental ...... 37 3.2.1 Clonal Selection ...... 37 3.2.2 Sample Collection and Preparation ...... 39 3.2.3 Extraction of Fresh and Frozen Material ...... 39 3.2.4 Total Nitrogen Determination ...... 40 3.2.5 Dry Weight Determinations ...... 41 3.2.6 Blade to Leaf Ratios ...... 41 3.2.7 Statistical Analysis ...... 41 3.2.8 Thin Layer Chromatography ...... 41 3.2.9 Reversed Phase HPLC ...... 42 3.2.10 HPLC/MS Studies ...... 42 3.3 Results ...... 43 3.3.1 Statistical Analysis ...... 43 3.3.2 Reversed Phase HPLC ...... 44 3.3.3 HPLC/MS ...... 45 3.4 Disscussion and Conclusions ...... 48

4 CLONAL EVALUATION OF ERYTHRINA ALKALOIDS BY GAS CHROMATOGRAPHY AND MASS SPECTROMETRY ...... 53 4.1 Introduction ...... 54 4.2 Experimental ...... 55 4.2.1 Clonal Selection and Sample Preparation ...... 55 4.2.2 Extraction of Dry and Lyophilized Plant Material 56 4.2.3 Statistical Analysis ...... 56 4.2.4 GC/MS Analysis ...... 56 4.2.5 Identification of Erythrina Alkaloids ...... 58 4.3 Results and Discussion ...... 60 4.4 Conclusions ...... 79 5 POSSIBLE METABOLIC FATE OF ERYTHRINA ALKALOIDS ...... 80 5.1 Introduction ...... 81 5.2 Experimental ...... 86 5.2.1 Feeding Study ...... 86 5.2.2 Milk Analysis ...... 86 5.2.3 Rumen Incubations ...... 87 5.2.4 GC/MS Analysis ...... 88 5.2.5 Synthetic Hydrogenations ...... 88 5.2.6 Statistical Analysis ...... 88 5.3 Results ...... 89 5.4 Discussion ...... 100 5.5 Conclusion ...... 104

iv 6 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 105 6.1 Summary ...... 106 6.2 Conclusions ...... 107 6.3 Recommendations ...... 109

REFERENCES ...... 110 APPENDICES A SELECTED SPECTROSCOPIC DATA ON ERYTHRINA AND RELATED ALKALOIDS FROM LITERATURE SOURCES ...... 127 B GAS CHROMATOGRAPHIC AND MASS SPECTRAL DATA OF SELECTED ERYTHRINA ALKALOIDS DESCRIBED IN THIS STUDY ...... 154 VITA ...... 161 LIST OF TABLES

B&££ Table 2.1: Summary of Toxicological Data on Erythrina Alkaloids ...... 33 Table 3.1: Selected Erythrina Clones and Important Characteristics ...... 38 Table 3.2: Percent Dry Matter and Total Nitrogen Content of Erythrina Clones ...... 46 Table 4.1: Alkaloid Profiles of Selected Erythrina Clones ...... 70 Table 4.2: Reports of Erythrina Alkaloids in the Species Studied by Other Investigators ...... 76 Table 4.3: 8-Erythroidine and Nitrogen Content of Thirteen Erythrina Clones ...... 78 Table 5.1: Tablulated Mass Spectral Data as Percent of Base Peak of Biologically Hydrogenated Dihydroerythroidines (DHEI, DHEU and DHEIII), Catalytically Hydrogenated Dihydroerythroidines (DHEIV and DHEV), Tetrahydro- Erythroidine (THE) and 8-Erythroidine (E) ...... 91 LIST OF FIGURES page Figure 2.1: Some Erythrina Alkaloids ...... 11 Figure 2.2: The Dienoid Type of Erythrina Alkaloids ...... 12 Figure 2.3: The Alkenoid Type of Erythrina Alkaloids ...... 13 Figure 2.4: Eryme lan thine from E. meiancantha ...... 20 Figure 2.5: Biogenetic Scheme for Erytfirina Alkaloids Biosynthesis: Tyrosine to Erysodienone ...... 24 Figure 2.6: Erythrina Alkaloid Biosynthesis; Erysoenone to the Aromatic Alkaloids ...... 25 Figure 2.7: Erythrina Alkaloid Biosynthesis: Conversion of the Aromatic Erythrina Alkaloids to the Lactonic Alkaloids, a- and B- Erythroidine by (a) Extradiol or (b) Intradiol Cleavage . 26 Figure 2.8: The Structures of Dihydro- and B-Tetrahydro-B-erythroidine . 30 Figure 3.1: HLPC Trace of B-Erythroidine on Merck Lichrospher Column 47 Figure 3.2: HLPC Trace of B-Erythroidine on Merck Lichrospher Column with 1.4 mM Trielhylamine ...... 47 Figure 3.3: HLPC Trace of B-Erythroidine on Polymeric Phenomenex PLRP-S Column ...... 47 Figure 3.4: Total Ion Chromatogram from the HPLC/MS Analysis of a 500 PPM Solution of Oxo-B-erythroidine and B-Erythroidine 49 Figure 3.5: Mass Spectra of B-Erythroidine by HPLC/MS Analysis ..... 50 Figure 3.6: Total Ion Chromatogram from the HPLC/MS Analysis of a 7000 PPM Solution of an Extract from Clone 2674 .... 51 Figure 4.1: Extraction Procedure for a Representative Plant Material Sample ...... 57 Figure 4.2a: GC/MS Calibration Curve for B-Erythroidine from 0 to 20,000 PPM ...... 59 Figure 4.2b: GC/MS Calibration Curve for B-Erythroidine from 0 to 220 PPM ...... 59 Figure 4.3: Total Ion Chromatogram (TIC) of Erysodine and Erysovine with Trimethylsilanized (TMS) Erysodine and Erysovine 62 Figure 4.4; Total Ion Chromatogram (TIC) of Selected Concentrations of B-Erythroidine Used in Calibration ...... 63 Figure 4.5: Total Ion Chromatogram (TIC) of Extracts of Plant Material Lyophilized and Oven-Dried at 60°C ...... 64 Figure 4.6: Total Ion Chromatograms (TIC) of E. berteroana Clones ... 67 Figure 4.7: Total Ion Chromatograms (TIC) of Reclassified E. berteroana and E. costarricensis Clones ...... 68 Figure 4.8: Total Ion Chromatograms (TIC) of E. poeppigiana Clones .. 69 Figure 4.9: Total Ion Chromatogram (TIC) and Mass Spectra of Three Unidentified Compounds ...... 75 Figure 5.1: B-Erythroidine ...... 84 Figure 5.2: Schematic of the Milk Transfer of Toxins in Goats ...... 85 Figure 5.3: Mass Spectra of Dihydroerythroidine I (DHEI) from the Milk and Rumen Samples ...... 92

Figure 5.4: Mass Spectra of Dihydroerythroidine II (DHEII) from the Milk and Rumen Samples ...... 93

Figure 5.5: Mass Spectra of Tetrahydroerythroidine (THE) from the Milk and Rumen Samples ...... 94

Figure 5.6: Mass Spectra of the Principal Biological Dihydroerythroidine Isomer, DHEI (a), and the Principal Synthetic Dihydro- Erythroidine Isomer, DHEIV (b) ...... 95 Figure 5.7: Amount of Dihydroerythroidines (DHEs) Present in the Milk of Goats which Consumed E. poeppigiana and E. berteroana Foliage per Goat ...... 96

Figure 5.8: Average Amount of Dihydroerythroidines (DHEs) Present in the Milk of Goats which Consumed E. poeppigiana and E. berteroana Foliage by Species ...... 97 Figure 5.9: The Amount of Milk Produced by Each Treatment Group ... 98 Figure 5.10: The Amount of Feed Consumed by Each Treatment Group .. 98 Figure 5.11: Body Weight of Goats (kg) by Each Treatment Group 99 UST OF ABBREVIATIONS a asymmetry a. a. amino acid abs absolute ACN acetonitrile amu atomic mass unit br. broad c concentration in moles/liter CDCI3 deuterated chloroform conjug. conjugated dd doublet of doublets deriv. derivative DCM die h lorome thane DHE dihydroerythroidine e molar absorpdvity ED effective dose EtOH ethanol F flow rate fr. fresh gc gas chromatography gc/ms gas chromatography and mass spectrometry ha hectare HPLC h i gh pre s sure liqu id chromatography HRMS high resolution mass spectrometry i.d. internal diameter IL intralympathic infl inflection IP intra peritoneal IR infrared IV intraveneous JACS Journal of the American Chemical Society JCS Journal of the Chemical Society k' capacity factor kg kilogram LD100 lethal dose required to kill 100% of test animals LD50 lethal dose required to kill 50% of test animals log e logarithm of molar absorptivity m multiple! M+ or M parent ion rnM millimolar m.p. melting point m/z mass to charge ratio max maximum Me methyl MeO methoxy MeOH methanol mg milligram MHz megahertz min minimum ms mass spectrometry MW molecular weight N number of theroetical plates NMR nuclear magnetic resonance O oral OH hydroxy ppm parts per million q quartet s singlet sh shoulder spp. species TD threshold dose THE tetrahydroerythroidine TIC total ion chromatogram TLC thin layer chromatography TMS trimethylsilyl UV ultraviolet V0 column void volume wt. w eight [a]d optical rotation ABSTRACT

The foliage of thirteen genetic clones of the nitrogen-fixing trees Erythrina berteroana, E. poeppigiana and E. cos tarricens is growing under similar environmental conditions in the Latin American Nitrogen Fixing Tree Arboretum, CATEE, Turrialba, Costa Rica, was evaluated by gas chromatography/mass spectrometry for toxic alkaloid content. The major alkaloid identified was B-erythroidine, a naturally derived drug used in the 1950s and 1960s as a neuromuscular blocking agent in surgery and electroshock treatments. Other alkaloids identified were a-erythroidine, erybidine, erythraline, erysodine, oxo-B-erythroidine and several previously undescribed compounds. There was a difference of more than two orders of magnitude in the B-erythroidine concentration between the clones with the highest and the lowest content. Although all clones tested contained some amount of B-erythroidine, at least one clone of each species was notably low in this drug. The results indicate that the expression of Erythrina alkaloids is at least partially under genetic control. A feeding study was performed utilizing goats on three experimental diets: control, 40% E. berteroana foliage and 40% E. poeppigiana foliage. The B-erythroidine present in the foliage of Erythrina species was hydrogenated in the rumen and B-erythroidine, dihydroerythroidines and tetrahydroerythroidines were detected it. the milk of goats in the Erythrina foliage treatment groups, however, these compounds were not detected in the milk the control group. Synthetic experiments demonstrated that the dihydroerythroidine isomers present in the milk and rumen samples were not the same isomers produced by catalytic hydrogenation. During catalytic hydrogenation, the conjugated diene in B-erythroidine is reduced to a single double bond at the 1-6 position. Mass spectral data indicate that the conjugated diene is conserved during hydrogenation of B-erythroidine in the rumen and that the double bond in the lactone ring may be hydrogenated. Synthetically derived dihydroerythroidine is ten times more potent than the parent drug, B-erythroidine, therefore there is a possibility that the presence of biologically hydrogenated dihydroerythroidines in the milk of goats which consume Erythrina foliage may pose a health hazard to individuals consuming the contaminated milk. Further studies designed to investigate the toxicology of biologically hydrogenated dihydroerythroidines are recommended. CHAPTER ONE: INTRODUCTION AND OBJECTIVES

1 2

1.1 Introduction Costa Rica is a small country in Central America about the size of West Virginia. In spite of its small size, Costa Rica has received world wide scientific attention due to the immense biological diversity encountered in twelve distinct biological life zones (Standley, 1937). However, like many Latin American countries and tropical areas in general, Costa Rica is suffering from an accelerated rate of deforestation and land degradation. In part, the forests are being cleared for their valuable tropical wood, but the many of the reasons behind the deforestation are political and social. The rate of deforestation exceeds 60, 000 acres per year. The main purpose of deforestation is the clearing of land for grazing cattle. Once the forest is cleared the soil fertility decreases rapidly until the land reaches very low levels of productivity. To complicate matters, much of Central America is mountainous and the pressure on land with over a 40% slope for cattle grazing is threatening land and water resources for the future. In Costa Rica the majority of the farms are ten hectares or less in size and are usually farmed by a family (Ministerio de Agricultura, San Jos6, Costa Rica, personal communication). Although the majority of the farms are small, there are large landowners that possess vast areas of the jest land. Therefore, many small farmers are forced to farm marginal lands of poor soils and steep slopes. In general, the government gives title to farmers if they can show that they have improved and developed the land. Improvement means clearing the land and keeping it clear. The easiest way to accomplish this is to clear cut, bum the plant residue and introduce cattle. Ecologically, this is disastrous. Due to the fragility of the soils and the high precipitation, all the topsoil is eroded within a few years and the farmer and his family are obliged to move on and continue the process somewhere else. Already in Costa Rica these degradated areas are infamous and are particularly visible in the high population centers such as the 3

Central Valley and Puriscal, Unfortunately, alternatives to these practices, although available, have not been effectively transferred to the farmers and they continue to ruin the land in a new area where they become established. Tropical ecosystems and temperate ecosystems are very different. The majority of the available nutrients and organic matter in tropical ecosystems are found in live organisms and not in the soil. Nutrient cycles are very rapid. Due to the high humidity and warm temperatures, organisms decompose quickly and the nutrients that are liberated are taken up and used again almost immediately. Therefore, systems which maintain a permanent soil cover are the only systems that offer a reasonable solution to the problem of erosion. Sustainable systems which include partial clearing of the land and the incorporation of permanent crops or trees will have a better chance of success in the long term than clearcutting. For this reason, many farmers as well as researchers are turning to agroforestry systems to provide alternatives to the current clearcutting tradition. Agroforestry is a relatively new concept which involves the cultivation of trees in association with crops, pasture and/or animals in a sustainable agricultural production system. Nitrogen fixing species of trees offer the most flexibility and therefore the greatest number of potential uses within these agroforestry systems. As well as producing wood and maintaining cover over the soil, they also add nitrogen to the soil and are, in general, easy to propagate. Nitrogen fixing trees usually have a high crude protein content and can be used as a high protein supplement for many ruminants (National Science Foundation, 1979). The Centro Agrondmico Tropical para Investigacfon y Ensefianza (CATIE), an international agricultural research institution located in Turrialba, Costa Rica is actively involved with the promotion of agroforestry. Agroforestry systems which combine pasture with trees and which integrate animals have been developed at CATIE. The best results in terms of forage and biomass production have come from those systems which 4 utilize nitrogen fixing trees in association with pasture such as the king grass (Pennisetum purpureum x typhoides) and Por6 {Erythrina spp.) system (Benavides, 1983). The development of a high protein forage is necessary for the tropics where nitrogen fixing grasses are unable to compete with the natural vegetation. This need is especially felt in the dryer areas where pasture is essentially unavailable or of very poor quality during the dry season. Again, nitrogen fixing trees offer a possible solution because they can be managed to provide green forage all year round. The high crude protein content of these arboreal species can serve as a concentrated source of protein in the diet of the ruminant and increase its production potential. Several models of agroforestry systems for forage production have been developed and their incorporation into traditional agricultural practices may be relatively easy. Use of the foliage bom nitrogen fixing trees already established on farms as live fences or shade foT permanent crops would be an alternative technology easier to transfer to the farmer than other more unfamiliar technologies. Nevertheless, the development of a protein bank where nitrogen fixing trees are cultivated in pure stands or in association with pasture for cut and carry systems, non-traditional systems in Central America, would be ideal from a purely technical standpoint. In the past two decades, world wide tropical research has concentrated on the small fanner. Due to restrictions such as slope, extremely high rainfall and lack of capital, the majority of the tropical farmers use low input techniques on their farms. The typical small farm is a subsistence, risk minimizing, low investment, low production and low return operation. Emphasis is placed on spreading the risk and investment of land and labor over a variety of crops. Small ruminants integrated into these systems offer the opportunity to convert crop residues, weeds and otherwise unusable materials, such as nitrogen fixing tree foliage, into useful products for family user and cash income. For these reasons it has been postulated that integrating goats into agroforestry systems offers a more practical solution than using cattle as a source of protein (meat and milk) for the small farmers. This is very important is rural areas where the scarcity of animal protein is severe. Goats are smaller and, when kept in confinement, cause no destruction to fragile tropical soils. They are easier to handle and therefore are more adaptable to family managed production systems, where female or older members of the family are frequently in charge of the small animal production. In addition, it is known that goats can be fed for short periods of time up to 100% of their diet in the form of nitrogen fixing tree foliage and maintain adequate milk production and weight gain with no ill effects (Benavides, 1983; Samur, 1984). The main objective of the Nitrogen Fixing Tree Project at CATIE is to introduce technology which provides a source of animal protein in the form of milk, meat and cheese to the small farmer and his family.

1.2 The Genus Erythrina and its Traditional Uses The genus Erythrina of the family Papiionaceae contains about 110 species and is widely distributed in the tropical areas of both hemispheres. Several Erythrina species, locally known as por6, are naturalized in Costa Rica. One species, E. poeppigiana, was introduced into the Central Valley of Costa Rica approximately 80 years ago. Since that time it has spread to most of Costa Rica. It has played a very important role in the development of the coffee industry where it is used as a nurse or shade crop species along with E. fusca. Farmers have adopted E. berteroana and E. costarricensis as live fence posts and guides for climbing crops. Normally the trees are pruned once or twice a year and the foliage is left on the ground to decompose thus adding nitrogen and organic matter to the soil. At a planting rate of 230 trees/ha in a coffee plantation, 140 kg/ha of nitrogen can be added to the system per year, or the equivalent of 700 kg/ha of ammonium sulfate per year (Lessard, 1982). Experiments completed at CATIE comparing the utilization of Erythrina as shade with a similar non-nitrogen fixing species 6 have shown that the coffee yields increased when grown under the por6 when compared with other tree species when no commercial fertilizers were used (Lessard, 1982).

1.3 Erythrina as Forage Erythrina is usually vegetatively propagated by large cuttings of stakes approximately two meters in height and four to six inches in diameter. Because of its versatility, wide distribution and ease of propagation, alternative uses of of the genus have been investigated. Of special interest is the use of the foliage of Erythrina as a high protein forage for ruminant animals. E. poeppigiana contains 27-34% crude protein on a dry weight basis. The total digestible nitrogen is 70% and its potential overall digestibility is 74% (Nitrogen Fixing Tree Project, 1986). One tree of E. poeppigiana can produce 10 kilograms of dry foliage per year. The digestibility of the biomass can be affected by varying the pruning frequency. A higher pruning frequency produces more tender stems and leaves which have a higher crude protein content and digestibility than older counterparts (Benavides, 1986). Studies done by the Animal Production Department at CATIE have shown that E. berteroana is more effective as a goat feed than E. poeppigiana, or at least that it is more attractive to goats when offered a choice. In an experiment where goat kids were offered the foliage of the two Erythrina spp. and that of banana leaves, E. berteroana was significantly superior to the other treatments in terms of weight gain (Benavides, 1986). Benavides (1983) also demonstrated that the performance of goats on a Erythrina spp. diet could be improved when an adequate source of energy was provided, such as mature or green bananas. It seems that the animals need a starch source to efficiently utilize the protein available in the Erythrina leaves. 7

1.4 Antiqualitative Agents in Erythrina Given all of the above advantages of Erythrina, it is unfortunate that the seeds and foliage of the genus are known to contain toxic alkaloids. If the alternative use of Erythrina as a forage for animals is to be developed, the alkaloid profile of each species and specific genetic clones selected for development must be determined and the fate of the alkaloids investigated. If low alkaloid content clones can be identified, then these clones could be promoted over those clones with higher concentrations of alkaloids and eventually, the alkaloid content of selected low alkaloid content clones may be reduced further by genetic engineering and plant breeding.

1.5 Objectives and Overview The following pages describe the preliminary investigations undertaken to answer two major questions. 1) Are there differences in the alkaloid content of genetic clones and can low alkaloid containing clones be identified? 2) If toxic alkaloids are present in the foliage of Erythrina species, are these alkaloids or possibly toxic metabolites present in the milk of dairy goats fed Erythrina foliage? Following this introductory Chapter, Chapter two summarizes what is known about Erythrina alkaloids in the form of a review of the scientific literature. Chapter three describes theErythrina plant material, clonal selection and sample collection process in Costa Rica and presents a statistical analysis of the biomass data of each clone and the results of liquid chromatographic techniques applied to Erythrina alkaloid separation. Chapter four details the alkaloid profile of each clone analyzed by gas chromatography and mass spectrometry and compares the profiles of the thirteen gen^.ic clones. Finally, Chapter five discusses a feeding study performed at CATTE using goats and identifies the possible metabolic fate of the major Erythrina alkaloid present in the foliage of the selected Erythrina clones. A data base of parameters necessary for the the identification and characterization of a wide 8 variety of Erythrina alkaloids compiled from the scientific literature is presented as Appendix A. The gas chromatographic and mass spectral data of selected Erythrina alkaloids that were generated from the investigations described in the following chapters are presented in condensed form as Appendix B. Lists of tables, figures and abbreviations follow the table of contents for more convenient cross reference. CHAPTER TWO: REVIEW OF LITERATURE

9 10

This chapter consists of a review of the scientific literature on the discovery, use and structure eluciadation of Erythrina alkaloids,

2.1 Erythrina Alkaloids: Occurrence, Isolation and Identification Figures 2.1, 2.2 and 2.3 display the chemical structures of many of the Erythrina alkaloids to be discussed subsequently. Figure 2.1 shows the structures of B- (1) and a-erythroidine (3) along with 8-oxo-erythroidine (2), erysodienone (4) and crybidine (5). Figure 2.2 depicts the aromatic diene Erythrina alkaloids (6-21) whereas Figure 2.3 depicts the alkenoid type of Erythrina alkaloid (22-30).

2.1.1 Folkers, Unna, Majors and Koniuszy As early as the 1880s, it was observed that Erythrina seed extracts possessed curare­ like properties (Altamirano, 1888). Curare selectively paralyzes neurotransmission of the motor nerve endings of striated muscles. In 1937, Folkers and Majors of Merck and Co, Inc. isolated a pure crystaline material from the seeds of E. americana that they named erythroidine (1), the same name Altamirano used for his crude extracts. The compound was active when given orally whereas true curare is active only when injected intraveneously. Folkers and Unna (1938 , 1939) began surveying the genus and they found that extracts of at least half of the species were biologically active. Folkers and Koniuszy (1939a) isolated erythramine (22) from E. sandwicensis and E. subumbrans and further studies were done to identify the structure of the alkaloid (1939b). In the following decade additional Erythrina alkaloids were isolated from a variety of species and structure determination studies ensued (Folkers and Koniuszy, 1940a, 1940b, 1940c; Folkers, Shavel and Koniuszy, 1941; Folkers and Shavel, 1942a; Koniuszy, Wiley and Folkers, 1949). 11

o O

(1) X=H2 B-eryihroidine (3) a-erythroidine (2) X=0 8-oxo-B-erylhroidine

OH

HO

N— CH

O

<4) erysodienonc OCH

(5) erybidine

Figure 2.1 Some Erythrina Alkaloids. 12

NAME Rl R2 R3 R4 (6) erysotrine CH3 CH3 CH3 H (7) erythaline —CH2— CH3 H (8) 11-Methoxyerythaline —CH2— CH3 OCH3 (9) eryihristtmine CH3 CH3 CH3 OCH3 (10) erythrascine CH3 CH3 CH3 OOOCH3 (11) erysodine H CH3 CH3 H (12) erysovine CH3 H CH3 H (13) erythravine CH3 CH3 H H (14) erythrinine —CH2— c h 3 OH (15) erysopinc H H CH3 H (16) erysonine H CH3 H H (17) erysoline CH3 H H H (18) 11-oxoerysodine H CH3 CH3 =0 (19) 11 -hydroxy erysodine H CH3 CH3 OH (20) 11-tnethoxy erysodine H CH3 CH3 OCH3 (21) 11-oxoerythraline —CH2— CH3 ==0

Figure 2.2 The Dienoid Type Erythrina Alkaloids. 13

NAME R l R 2 R3 (22) erythramine —CH2— H (23) crythratidinc CH3 CH3 OH (24) crythratinc —CH2— OH (25) erysotine H CH3 OH (26) crysosalvine CH3 H OH (27) erysopitine H H OH (28) erythratidinone CH3 CH3 =0 (29) erythratinone —CH2— **0 (30) erysotinone H CH3 =0

Figure 2.3 The Alkenoid Type of Erythrina Alkaloids. 14

Folkers and co-workers divided the Erythrina alkloids into three classes. The "free" alkaloids were ex tractable by immiscible solvents from clean, fat-free plant extracts which had been rendered alkaline. They were given the prefix ERYTHR- and some examples are erythraline (7), erythramine (22) and erythratine (24). The "liberated" alkaloids were only extractable after acid hydrolysis. They have the prefix ERYSO- and erysodine (11), erysopine (15) and erysovine (12) are examples. The "combined" alkaloids were not extractable by ordinary methods, contained sulfur and were sources of the "liberated" alkaloids. Later, it was determined that the "combined” alkaloids were sulfoacetic acid derivatives of the "liberated" alkaloids. They were given the prefix ERYSOTHIO-. The first basic structure for the aromatic Erythrina alkaloids based on UV data and degradation studies was suggested by Folkers, Koniuszy and Shavel (1942). A structure was suggested for B-erythroidine (1) by Koniuszy and Folkers in 1950. In 1944, Folkers, Koniuszy and Shavel began investigating and isolating some of the combined alkaloids which were determined to be sulfoacetic acid esters of the corresponding liberated alkaloids. Colorimetric assay of B-erythroidine was devised in 1946 (Dietz and Folkers, 1946; Folkers and Shavel, 1946). Several U.S. patents were issued in conjunction with Folkers and coworkers' work with Erythrina alkaloids (Folkers and Major 1945a, 1945b; Folkers and Koniuszy, 1946; Major and Folkers, 1947).

2.1.2 Deulofeu At nearly the same time that Folkers and coworkers at Merck were investigating Erythrina species for alkaloids, Deulofeu was looking at Latin American species in Argentina (Deulofeu, Hug and Mazzocco, 1939). Following Folkers work, Deulofeu et al. isolated erythratine (24) and erysodine (II) from E. crista-galli (1947). Deulofeu 15 proposed an alternative structure for the aromatic Erythrina alkaloids (Labriola, Deulofeu and Berinzaghi, 1951; Deulofeu, 1952).

2.1.3 Boekelheide Boekelheide while at the University of Rochester began a series of studies aimed at elucidating the structure of B-erythroidine and identifying the structural features required for biological activity (Sauvage and Boekelheide, 1950). Merck and Co., Inc. provided the B-erythroidine required for the studies. Several B-erythroidine derivatives and decomposition products where found to be biologically active (Boekelheide and Agnello, 1951). In 1953, Boekelheide et al. published a structure of B-erythroidine that is now accepted as the true structure of B-erythroidine, based on UV, IR, degradation studies and comparison studies of the apo-derivative with lysergic acid. Studies on the degradation and structure of apo-erythroidine (Grundon and Boekelheide, 1953; Grundon, Sauvage and Boekelheide, 1953), the hydrogenation of desmethoxy-B- erythroidine (Boekelheide, Anderson and Sauvage, 1953) and the structure a- erythroidine (3) were also completed (Boekelheide and Grundon, 1953; Godfrey, Tarbell and Boekelheide, 1955). Boekelheide summarized his results and those of others in a review written with Prclog in 1955 and in Manske in 1960. After the structures of a and B-erythroidine were elucidated, experiments focused on determining the absolute configuration of a and B-erythroidine (Wenzinger and Boekelheide, 1963) and the other Erythrina alkaloids (Boekelheide and Wenzinger, 1964).

2.1.4 Other Studies on Structure Elucidation. Lapifre and Robinson (1951) were also involved in determining the structure of B- erythroidine. Lapi6re developed a specific test for the determination of B-erythroidine 16 upon degradation based on a reaction with ferric chloride. Hill and Schearer (1962) proposed the absolute configuraion of a-erythroidine at C-12. And finally in 1963, Hanson determined the crystal structure of dihydro-B-erythroidine hydro bromide and the absolute configuration at the spiro atom became known. Barton et al. (1970) published the crystal structure of a new Erythrina alkaloid named erythristemine (9) isolated from E. lysistemon. Following the publication of the crystal structure, other modem spectroscopic analyses were performed on the Erythrina alkaloids. Mass spectrometry was first performed on Erythrina alkaloids by Boar and Widdowson in 1970. The aromatic diene alkaloids exhibited a novel fragmentation and rearrangement reaction with a loss of 58 amu. Peaks attributible to a retro-Diels-Alker reaction in ring D were seen in the mass spectra of a and B-erythroidine. Mass spectral studies of synthesized erythrinane derivatives were done by Migron and Bergmann in 1977. In all cases the molecular ion peaks were relatively intense when electron impact ionization was used. Nuclear magnetic resonance (NMR) studies performed on Erythrina alkaloids were reviewed by Crabb in 1975. Decoupling and INDOR techniques proved helpful in the assignment of the NMR peaks (Barton et al., 1970; Ju-ichi et al., 1977 and 1978).

2.1.5. Occurance of Erythrina Alkaloids While some researchers concentrated on the elucidation of the chemical structure of Erythrina alkaloids, others were working on the isolation and characterization of new alkaloids or the investigation of unstudied Erythrina species. A new alkaloid, erysotrine (6) was isolated from E. suberosa by Singh and Chawla (1969). The discovery of erysotrine was important because it had been synthesized in the laboratory from other Erythrina alkaloids but had never before been found in nature. Ito et al. (1970) isolated erythrinine (14) from E. indica, which was novel in that ring C was 17 oxygenated at C-l 1 and this structural variation had not been seen before. This paper was also important in that the alkaloid was isolated from the leaves. All previous studies had focused on the seeds of Erythrina species although it was known that other plant parts possessed biological activity (Folkers and Unna, 1938). Ghosal, Ghosh and Dutta (1970) also began investigating other plant parts for Erythrina alkaloids in India. Their work followed reports of medicinal uses of various Erythrina plant pans. The highest alkaloidal content was found in the bark of E. variegata. In the bark and leaves, erysotrine was the major alkaloid whereas in the seed, hypaphorine was the major alkaloid. Erysodine, erythraline (7) and erysopine (15) where also identified. A similar pattern was seen when E. lithosperma was investigated (Ghosh and Majumdar, 1972). Ghosal, Chakraborti and Srivastava (1972) identified an O-acetylated alkaloid at C-l 1 called erythrascine (10) from the seeds of E. arborescens. In this species the highest alkaloid content was found in the seeds. Ghosal and Srivastava (1974) also isolated a quaternary indole alkaloid derivative of ersyovine from the seeds of £. arborescens. Singh and Chawla (1969) were also working in India on the identification of various Erythrina alkaloids in Indian species of Erythrina. They isolated erysodine and erysotrine from E. suberosa and erythraline and erysovine from £. variegata (Singh and Chawla, 1970a, 1970b and 1970c). They noted that different collections of £. suberosa seeds varied in their alkaloidal constitution. They did not isolate any Erythrina alkaloids from the leaves of £. suberosa, however they did isolate several Erythrina alkaloids from the bark of £. variegata. Miana et al. (1972) isolated erysotrine from the leaves of £. suberosa. A new structural variation was discovered by Letcher (1971). Methylation of the C- ring at carbon-11 was discovered in the leaves of £. lysistemson and the new alkaloid named 11-methoxyerythraline was identified. By this time the use of B-erythroidine and 18 the more potent dihydro-B-erythroidine in clinical practice had trailed off due to replacement by synthetic drugs, but interest in Erythrina alkaloids continued. Hargreaves et al. (1974) reviewed the occurrence of Erythrina alkaloids in twenty- two American species of Erythrina. They reported that the largest alkaloidal fraction in the seeds of the American species was the liberated alkaloid fraction (the eryso- alkaloids) which constituted about 0.1% of the dry seed weight. Only in exceptional cases did the percent alkaloids exceed 1% of the seed dry weight. They identified six new Erythrina alkaloids: erythravine, erysoline (17), erysosalvine, erysosalvinonc and erysoflorinone. Games et al. (1974) reviewed the occurrence of Erythrina alkaloids in the seeds of eighteen African, Asian, Polynesian and Australian species. The analysis was done by field desorption GC/MS. Games reported that erysodine and erysovine were the most abundant alkaloids found in the eastern species of Erythrina and they occured in every sample analyzed. The Hargreaves et al. (1974) and Games et al. (1974) papers were part of a special issue of Lloydia concerning the taxonomy of the genus Erythrina. In that same issue, Romeo and Bell (1974) measured the amino acid concentration and alkaloids in 78 species of Erythrina with representative from each of the 26 taxonomical sections. Amino acid concentration was usually low when alkaloid content was high and only common amino acids were found in the genus. No Erythrina alkaloids were detected in E. edules, an Erythrina species eaten by natives in Colombia. They found that a and B-erythroidine only occured in the sections of Micropteryx, Erythrina and Chirocalyx and their presence may be taxonomically important. When a and B-erythroidine occured they were often the major alkaloids. Romeo and Bell did not evaluate the alkaloidal variation between clones of Erythrina. Ito and co-workers continued to investigate Erythrina species found in Japan, New alkaloids, erythrinine and erybidine (5) were isolated and characterized from E. bidwillii . (Ito, Furukawa and Tanaka, 1973) and E. crysta-galli (Ito et al.,1973) as well 19 as other previously identified alkaloids. They isolated the new alkaloids 11- hydroxy cry so trine, erythrabine and cry so tra mi dine from E. arborescens (Ito, Furukawa and Haruna, 1974a and 1974b) and crystamidine (10,11 dehydro-8-oxo-erythraline) from £. crysta-galli (Ito et al., 1976). In the late 1970s, Erythrina type alkaloids were isolated from genera other than Erythrina. Dihydroerysovine was isolated from Cocculus trilobus (Ju-ichi et al., 1977) and was completely characterized by NOE and INDOR NMR and MS techniques. Ju- ichi used these techniques to determine the substitution pattern on the aromatic Erythrina alkaloids synthesized in that laboratory (Ju-ichi et al, 1978) and those isolated from Cocculus (Ju-ichi, Fujitani and Furukawa, 1985). Erythrina alkaloids have also been isolated from the flowers of Erythrina species. The flowers are sometimes used to make tea for medicinal purposes. El-Olemy et al. (1978) investigated the Erythrina alkaloids in the flowers and seeds of E. variegata in EygpL Erysotrine was the major alkaloid in the flowers while erysodine was the major alkaloid in the seeds. Aguilar, Giral and Espejo (1981) found that a and B-erythroidine comprised 0.11 % by weight of the dried flowers of E. americana. Dagne and Steglich (1984) discovered 8-oxoerythrinine in the flowers o ff. brucei, a tree found only in the highlands of Ethiopia and used for medicinal purposes. Other previously identified alkaloids were also isolated. A unique Erythrina alkaloid was isolated from the seed of E. melancantha (Dagne and Steglich, 1983). Erymelanthine (31) (Figure 2.4) contains a second nitrogen atom in the tetracyclic Erythrina skeleton. 20

Figure 2.4 Erymelanthine from E. melancantha.

2.1.6 Jackson, Chawla and Coworkers. Most recently another group at the University of Cardiff in Great Britain has begun to isolate and characterize the Erythrina alkaloids. Some of their work has been in collaboration with Chawla in India. Initial studies focused on the seeds of Erythrina species (Barakat et al., 1977, Abdullah et al., 1979). Further studies examined the leaves of E. poeppigiana, E. berteroana, E. macrophylla, and E. salviiflora (Jackson and Chawla, 1982) using the field ionizadon GC/MS method developed by Games et al. (1973). The alkaloids a and B-erythroidine were isolated from the leaves of the former species, but not from E. macrophylla. 8-oxoerythroidine (2) was detected in E. berteroana but not in E. poeppigiana. The seeds of various collections of other Erythrina species were examined for Erythrina alkaloids by GC/MS (Redha, 1983) and quantitative differences were found between samples which they attributed to different growing conditions (Jackson et al., 1982). Further studies were done on the synthetic interconversion of 11-oxygenated Erythrina alkaloids (Abdullah et al., 1982), on the NMR of Erythrina alkaloids (Chawla, Chunchatprasert and Jackson, 1983; Chunchatprasert, 1982) and on semi-preparative normal phase HPLC of Erythrina 21 alkaloids (Chawla, Jackson and Ludgate, 1982; Chunchatprasert, 1982; Chawla, Redha and Jackson, 1985). In the normal phase HPLC studies, a and B-erythroidine constituted more than 90% of the alkaloids in the seeds and more than 95% of the alkaloids in the leaves of E. berteroana. Mantle, Laws and Widdowson (1984) developed a reversed-phase HPLC method for the purification of 8-oxo-erythraline and crystamidine from the leaves of E. crista-galli using methanol and water as the mobile phase. Studies continued with Indian Erythrina species using GC/MS. E. blakei bark was investigated (Singh et al., 1981). Several common Erythrina alkaloids and a few new Erythrina alkaloids, which were N-oxides, were identified in E. crista-galli flowers (Chawla, Gupta and Jackson, 1987). A previous study had shown that the N-oxides were not artifacts but were real natural products (Sarragiotto, Filho and Marsaioli, 1985). The bark of E. variegata also contained mostly free alkaloids and few liberated alkaloids (Chawla et al., 1988). Much of the work by Chawla and Jackson is summarized in the reviews on Erythrina alkaloids in Natural Product Reports aim the Chemistry and Biology of Isoquinoline Alkaloids (Chawla and Jackson, 1984, Jackson, 1985, Chawla and Jackson, 1986). Reviews on the use of alkaloids in the chemotaxonomy of Erythrina and other leguminosae can be found in Smolenski et al. (1972), Kinghom and Smolenski (1978), and Mears and Mabry (1971).

2.2 Erythrina Alkaloids: Biosynthesis.

2.2.1 Early Studies Leete and Ahmad (1966) isolated radioactive a and B-erythroidine from E. berteroana plants which were fed radiolabled ^C-tyrosine (32), The radioactivity was equally 22 divided between the C-8 and C-10 positions suggesting that a symmetrical biosynthetic intermediate was involved in the biosynthetic pathway, Radiolabed phenylalanine was not incorporated. Leete suggested that dopamine (33) and 3,4 dihydroxy- phenylacetaldehyde (34), both derived from tyrosine, condensed to yield a Schiff base which is reduced to a symmetrical amine. Several workers had been successful in synthesizing the aromatic Erythrina ring system from the in vitro oxidation of the amine through the erysodienone (40) (Mondon and Ehrhardt, 1966; Gervag et al., 1966).

2.2.2 Barton and Coworkers. Barton et al. (1966) worked on the later stages of the biogenesis and suggested that erythratine (24) was the biogenically favored intermediate after ersyodienone. Later this group isolated erythratine as a new natural product from E. crysta-galli. Barton et al. (1968a and 1968b) also refined the early steps in Erythrina alkaloids biogenesis by demonstrating that the previously proposed symmetric amine precursor, synthesized in their lab in the radioactive and inactive forms, was not incorporated into the Erythrina alkaloids. N-norprotosinomenine (35) was incorporated, specifically and exclusively (Barton et al. 1974) and it was postulated as the new biogenetic pecursor making Erythrina alkaloids in reality benzylisoquinoline alkaloids. Later a few benzylisoquinoline alkaloids were isolated along with Erythrina alkaloids from several Erythrina species (Ghosal et al., 1971; Barton et al., 1973). Even though a symmetrical amine was thought to be a precursor of norprotosinomenine for a few years (Barton et al., 1970a and 1970b), it was ultimately discarded as a precursor of Erythrina alkaloids and a new symmetrical intermediate (38) formed from the benzylisoquinoline structure was suggested. A new biogenetic scheme was postulated and is presented in a modified form as Figures 2.5 and 2.6. Two moles of tyrosine condense to form the shiffs base as earlier postulated but the symmetrical amine intermediate is not formed, instead a dibenzazonine symmetrical intermediate (38) is formed from norprotosinomenine by para-para phenolic oxidative coupling and ring opening followed by reduction of the inline (37) to the dibenzazonine (Barton et al., 1974a). Subsequent attack by the lone pair of electrons on the nitrogen on the conjugated double bond of the diquinone (39) leads to erysodienone (40). The dienone is converted to erysoenone (41) which is reduced to th cErythrina alkaloid erysotrine (6). Erysotrine, an alkenoid type of Erythrina alkaloid, is converted through loss of water to the common alkaloid erysodine (11), a diene type alkaloid, or to erythratine (24) through the formation of the bisoxymethylene bridge. The order of bisoxymethylene formation is not important. Methylation of the hydroxy groups or demethylation of the methoxy groups yields the other common aromatic Erythrina alkaloids such as erysopine and erysovine. The lactonic Erythrina alkaloids, a and B-erythroidine, are thought to arise from the degradation of an aromatic precursor as shown in Figure 2.7 (Barton et al., 1974b; Grethe, 1981). It is still unknown if the biosynthetic sequence to the lactonic alkaloids occurs via intradiol cleavage (Figure 2.7a) or extradiol (Figure 2.7b) cleaveage. Extradiol cleavage of an aromatic ring in a natural product has been demonstrated in the betalains (Impellizzeri and Piattelli, 1972). In either case, cyclization would produce a- erythroidine and then B-erythroidine after isomerization. A plausible biogenetic route has been proposed for the formation of the double nitrogen containing erymelanthine (31) through extradiol cleavage of the aromatic ring of erysovine (Dagne and Steglich, 1983) providing additional evidence that the aromatic ring cleavage of Erythrina alkaloids may occur through extradiol cleavage. Good reviews of the biosynthesis and biogenesis of Erythrina alkaloids can be found in Mondon (1970) and Dyke and Quessy (1981). A recent review on the biosynthesis of benzylisoquinoline alkaloids in cell culture, including certain Erythrina alkaloids, can be found in Zenk (1989). 24

COOH COOH •CHO n h 2 NH

HO HO OH HO OH (33) DOPA (34) dlfcjdi

HO N-H N-H MeO MeO OH coupling MeO OMe

OH (35) N- benzyllaoqulnollne derivative

OH OH MeO MeO

N-H

(35) dlbtasaionlnc MeO (37) Imlae MeO OH OH ♦ HO MeO

MeO N-H MeO O MeO (40) « r ;H d l» * M

Figure 2.5. Biogenetic Scheme for Erythrina Alkaloid Biosynthesis: Tyrosine to Erysodienone. 25

HO HO

MeO MeO

MeO MeO O JOH (41) erysoenone (6) erysotrine I HO

MeO

MeO MeO OH (11) erysodine (24) erythraline I HO

HO

MeO MeO (15) erysopine (7) erythraline

MeO

HO

MeO' (12) erysovine

Figure 2.6 Erythrina Alkaloid Biosynthesis: Erysoenone to the Aromatic Alkaloids. 26

a% HO OHC HO,C a HO HO

MeO (15) erysopine

OH OHC

HOOC O

MeO MeO I I

isomerization O O

MeO’

(3) oc-erythroidine (I) B-erythroidine

Figure 2.7 Erythrina Alkaloids Biosynthesis. Conversion of the Aromatic Erythrina Alkaloids to the Lactonic Alkaloids, a and B-Erythroidine by (a) Extradiol or (b) Intradiol Cleavage. 27

2.3 Erythrina Alkaloids: Synthesis Much of the work on the synthesis of Erythrina alkaloids which occurred in the 1950 and 1960s by Boekelheide and Mondon focused on discerning probable biosysthetic routes. Belleau (1953) first published a synthesis of the erythrinane skeleton in a 12% overall yield from 2-carbethoxycyclohexanone in five steps. Unfortunately, no natural products resulted from this research for many years (Boekelheide, 1955). However, interconversions of various natural products were performed, for example hydroxylations or methylations or the isomerization of a and 8-erythroidines (Boekelheide and Morrison, 1958). Mondon and Ehrhardt published the total synthesis of dihydroerysodines in 1966, even though the natural product erysodine was not synthesized, however the paper was one of the first to present NMR spectal data on Erythrina derivatives. Erysodienone was produced during the synthesis. This compound was a proposed natural precusor in the biosynthetic process of Erythrina alkaloid formation. Erysodienone was isolated from E. lithosperma as a natural product in 1971 (Ghosal et al., 1971). Erysotrine was synthesized by Mondon and Nestler (1964) from the erythrinan skeleton, however this compound was not isolated as a natural product until later. There was some work done on the synthesis of Erythrina alkaloids or their biosynthetic intermediates from benzylisoquinoline derivatives (Franck andTeetz, 1971; Kupchan, Kim and Lynn, 1976). A B-erythroidine derivative was synthesized from the erythrinane skeleton in 1974 (Kitahara and Matsui, 1974). A new approach to erythrinane skeleton using an diphenyl amide was suggested by Ito et al. (1976). Ito's group also prepared erybidine (5) photochemically (Ito and Tanaka, 1974). Further work focused on preparing Erythrina alkaloids from the basic erythrinane skeleton (Haruna and Ito, 1976; Mondon and Nestler, 1964). Most recently, new methods of erythrinane skeleton formation have appeared (Ahmed- Scholfield and Mariano, 1985 and 1987; Westling, Smith and Livinghousc, 1986). Variations on the above methods have also appeared and are too numerous to list here, Chawla and Jackson (1989) have recently reviewed Erythrina alkaloid synthesis and readers interested in Erythrina alkaloid synthesis are referred to their paper

2.4 Erythrina Alkaloids: Pharmacology and Toxicology As mentioned in the introduction, interest in Erythrina began because of its use by South American Indians as an arrow poison. Various plant parts have been used in folk medicine for purgative, diuretic, sudorific or hypnotic effects (Unna et al.t 1944a and 1944b). In the 1880s, extracts of E. americana were shown to have curare-like activity in frogs (Altamirano and Dominquez, 1888). Lehman (1936) followed up on this work and reported preliminary pharmacological data using a crude ethanolic extract of E. americana. The crude drug was fatal to mice, rats, rabbits and pigeons in a range of 5 to 600 mg/kg depending on the route of administration and species. Intravenous injection was the most effective route of administration followed by intramuscular injection and then intraperitoneal injection. Lehman observed complete sciatic nerve paralysis at sublethal doses that lasted anywhere from 20 minutes to eight hours depending on route of administration and species. He found that phrenic nerve paralysis was peripheral and if artificial respiration was maintained, nanral respiration resumed even if 10 to 15 times the effective dose was given. Intravenous administration of the crude drug completely inhibited the typical convulsions which occur after strychnine poisoning. Lehman recommended pursuing the pharmacological activity of this drug as a possible anticonvulsant. A year later, Folkers and Major (1937) from Merck and Co., Inc. isolated B-erythroidine in its crystal form and members of that organization continued to evaluate compounds from Erythrina as possible commercial drugs. 2.4.1 Research at Merck Folkers and Major (1937) determined the LD50 of erythroidine hydrochloride in white mice to be 120 mg/kg when given perorally and 13 mg/kg when given subcutaneously (SC). 0.10-0.15 mg per frog caused complete motor paralysis when injected into the lymph sac (intralymphatic injection, IL). In order to ascertain whether the curare like activity was present in all the Erythrina species, a collaborative effort with the New York Botanical Gardens was undertaken (Folkers and Unna, 1938, 1939) All fifty-one species surveyed demonstrated curare like action in frogs to varying degrees. The frogs injected with a threshold dose recovered within three hours. Those injected with ten times the threshold dose recovered the next day. No deaths were reported. When the first aromatic Erythrina alkaloid, erythramine, was isolated, it was determined to have a threshold dose of 7 mg/kg as the hydrobromide in the frog (Folkers and Koniuszy, 1939). The threshold dose for erythraline hydrobromide and erythraline hydrobromide was 8 mg/kg and 75 mg/kg frog respectively by intralymphatic injection (Folkers and Koniuszy, 1940). The threshold dose for erysonine was 100 mg/kg frog (IL) (Folkers, Shavel and Koniuszy, 1941). The sulfoacetic acid derivatives proved to be more potent than the liberated alkaloids. A more detailed pharmacological study of B-erythroidine and its derivatives was performed by Unna, Kniazuk and Greslin (1944a and 1944b). Dihydro-B-erythroidine (43) (Figure 2.8.) was found to be the most potent drug with an LD50 *n cats of 2-3 mg/kg when given orally or subcutaneously. Cats were the most sensitive species tested, whereas rats were the most resistant. The alkaloids were rapidly absorbed from the gastrointestinal tract and elicited a response with the same time frame independent of route of administration. The animals died of nsphyxia. In chronic studies, rats were fed 40% of the LD50 of B-erythroidine for two months. No acute toxicological symptoms were seen, however, after two months the animals began to lose weight and died. Cats 30 fed 50% of the LD50 for three months displayed no toxic effects. Dihydro-C- erythroidine was also the most effective and longest lasting alkaloid in paralysis studies completed with frogs with an effective dose (ED) of 0.5 mg/kg. Administration of these alkaloids depressed blood pressure at doses below those that caused paralysis. Skeletal muscle was affected before the diaphragm. The margin of safety was small, but larger than that of curare itself. The major mode of action of Erythrina alkaloids was determined to be myo-neural transmission blockage of nerve impulses to the skeletal muscles. Smooth muscle and cholinesterase activity was not affected by B-erythroidine. This was confirmed by other researchers (Pichard and Luco, 1944a and 1944b). Prostigmine was an erythroidine antagonist. In a similar study of the the aromatic Erythrina alkaloids, the liberated alkaloids were less effective than the combined alkaloids and dehydrogenation diminished the curarized effect of the alkaloids. (Unna and Greslin, 1944a and 1944b).

O O

(42) dihydro-B-erythroidine (43) B-tetrahydro-B-erythroidine

Figure 2.8 The Structures of Dihydro- and B-Tetrahydro-B- erythroidine. 31

In later studies, it was determined that B-erythroidine had less of an effect on the central nervous system than curare and other curare-like compounds (Pick and Unna, 1945) and that B-erythroidine was synergistic with narcotics such as pentobarbital (Pick and Richards, 1947). Both drugs together could be administered in smaller amounts than either drug separately to obtain the same effects.

2.4.2 Clinical Studies. The most frequent use of B-erythroidine was in controlling seizures and convulsions in medical treatment for schizophrenia and other mental diseases (Lehman, 1936). Cottington (1941) describes the use of B-erythroidine in mctrazol treatment of children with schizophrenia to reduce the severity and duration of the induced seizures and to alleviate the dread felt by the patients associated with the treatment. Previously the treatment almost always resulted in compression fractures of the thoracic vertebrae. The administration of 100 mg of B-erythroidine intraveneously resulted in complete muscle relaxation and subsequent administration of metrozol produced mild or no seizures. This treatment was given three times a week. B-Erythroidine was used in the same capacity in electroshock treatments (Robbins and Lundy, 1947a). Curare and curare like compounds were used in surgery to produce a degree of anesthesia impossible to obtain with nitrous oxide or ethylene anesthesia (Robbins and Lundy, 1947b). B-Erythroidine was atractive for this purpose because it could be administered orally as well as intraveneously whereas curare had to be administered intraveneously. It also had a greater margin of safety than did curare (Berger and Schwartz, 1949), nevertheless, B-erythroidine’s margin of safety was small by today's standards. The LD50 was twice the paralyzing dose in mice (Berger and Schwartz, 1949; Savage, Berger and Boekelheide, 1949). In the 1940s and 1950s, curare was widely used in American operating rooms. The use of B-erythroidine and investigation 32 of its mode of action were inhibited by the lack of knowledge of the structure of fi­ ery throidine at that time and possibly by the induction of hypotension and bradycardia in some patients. Megirian, Leary and Slater (1955) surveyed a wide variety of B-erythroidine derivatives pharmacologically. Dihydro- (42) and fi-tetrahydro (43) derivatives remained the most potent drugs tested. The desmethoxy and apo derivatives had greatly reduced potency. The toxicological data available in the literature on Erythrina alkaloids is summarized in Table 2.1.

2.4.3 Other Pharmacological Studies. Dhar et al. (1968) discovered that Erythrina extracts had anticancer activity and although homoerythrina alkaloids isolated from Cephalotaxius harringtonia were pursued for this activity (Achenbach, 1977), the alkaloids isolated from Erythrina species never passed athe initial screening. The biological activity of Erythrina extracts has been frequently mentioned in folk medicine literature (Ghosal et al., 1970a and 1970b; Bartos and Bhattacharya, 1970; Bhattacharya et al., 1971; Gayum et al., 1971; Ghosal et al., 1972) One conclusion drawn from the review of Mondon's synthetic studies (1964 and 1966) was that thecisttrans configuration at the double bonds in the A and B rings was not important for pharmacological activity as all four optical isomers showed similar activities in biological assays (Snieckus, 1973). However, the position of the double bond in the D ring may be important. Unfortunately, there is no toxicological data available on the relative potencies of a and B-erythroidine or other Erythrina alkaloid derivatives that vary in the position of the double bond in the lactonic ring. Table 2.1 Summary of Toxicological Data on Erythrina Alkaloids.a

SPECIES

Chemical Frogs^ MiceE Ratsc Catsc H um ans Ref. B-erythroidine (O) 120 d 1-Dioo(0) 140 (SC) 45 LDioo(SC) 50 TD3 (0)75 510 30-40 e (SC) 48 1260 20-25 ED 1.5 24 ED 15-17 f ED50 14 ED 25 (IP) 30 ED50 0.4 methiodide 200 DihydrO'B-erythroidine 0.5 (O) 7.5 320 2-3 e (SC) 9.3 230 2-3 (IP) 4.5 ED30 08 f B-ietrahydro -8 0.5 (SC) 9.5 e ery throidine (IP) 10.6 ED50 .12 f erythramine 7 (SC) 104 e HBr 10 methiodide 40 dihydroerythamine 300 erythraline 8 (SC) 72 (0)80 hydrobromide 10 methiodide 50 erythratine 75 methiodide 300 dihydioerythratine 100 erysonine HQ 100 erysopine HQ 4 (SC) 15 (0)18 dihydroerysopine HBr 40 erysovine HC1 3 tetrahydroerysovine HBr 1 erysodine HQ 10 (SC) 100 (O) 155 tetrahydnwysodine HBr >300 aAll values in mg/kg. ^Threshold value for paralysis by intralymphatic injection. cED5o (dose required to kill 50% of test animals) unless otherwise noted. ^Folkers and Major, 1937. eUnna et al., 1944. ^Cited in Robbins and Lundy, 1947b. 2.5 Summary It has been known for many decades that Erythrina alkaloids were present in the seeds of a number of Erythrina species. Even though other plant parts were known to be biologically active, the leaves and bark of Erythrina species were not investigated for alkaloid content until the late 1970s and 1980s and only then in a restricted number of species. No studies were performed on the differences in alkaloid content between genetic clones of Erythrina. In studies on the taxonomic sections of Erythrina of the eastern hemisphere, the aromatic Erythrina alkaloids erysodine and erysovine seemed to be present in the highest concentrations. On the other hand, the lactonic Erythrina alkaloids, a and B-erythroidine seemed to dominate in the Erythrina species of the western hemisphere. Biosynthetically, Erythrina alkaloids are derived from two molecules of tyrosine which condense to form a benzylisoquinoline intermediate, A symmetrical intermediate is involved in a later biosynthetic step. The lactonic alkaloids are derived from aromatic ring cleavage of the aromatic Erythrina alkaloids. The principal biological action of Erythrina alkaloids is muscle paralysis. The lethality of each alkaloid depends on its structure, route of administration and the animal species. All Erythrina alkaloids possess some degree of biological activity. The dihydroderivatives of the lactonic alkaloids are more potent than the parent compounds, whereas the dihydrodcrivatives of the aromatic Erythrina alkaloids are usually less potent that the parent compounds. Pharmacokinetic and metabolic fate studies have never been performed on these drugs. CHAPTER THREE: SAMPLE COLLECTION, STATISTICAL ANALYSIS AND LIQUID CHROMATOGRAPHIC TECHNIQUES

35 36

3.1 Introduction The use of nitrogen fixing trees in agroforestry systems in general, and as feed for animals in particular, is attractive because the incorporation of nitrogen fixing trees to an agricultural system will help control erosion and add nitrogen to the system. Cassia , Leucaena and Gliricidia are a few species in addition to Erythrina that have been suggested for use in agroforestry systems. Leucaena is problably the best studied genus because of its ease of propagation, tolerance to a range of soil conditions and its palatibility to animals. It can contain up to 30% by weight crude protein and twice the carotene content of alfalfa. Unfortunately, it is also toxic to some animals due to its high mimosine content. Toxic symptoms include haemorrhagic cystitis and death in some animals. It was discovered, however, that certain rumen bacteria in some animals in specific geographical locations could degrade mimosine to non-toxic compounds (Molyneux et al., 1979). Therefore, Leucaena serves as a case study in forage assessment of nitrogen fixing trees. Crude protein content, biomass production and digestibility are important forage assessment parameters, but the presence of antiqualitadve compounds such as mimosine, tannins or alkaloids is also important. The presence of Erythrina alkaloids in the seeds of various Erythrina species was known in the 1930s, however, their presence in other Erythrina plant parts was not investigated until 1974 (Barton et al., 1974b). Most of the isolations of Erythrina alkaloids were performed using large amounts of seeds which were high in alkaloid content, approximately 1% by weight. The concentration in the leaves is known to be smaller, approximately 0.1% or less. Early work utilized traditional column chromatography over silica or alumina and preparative TLC. Gas chromatographic techniques using packed columns was first used in the 1970s by Boar and Widdowson (1971), who in addition to describing GC conditions used to separate several alkaloids, developed a theory of fragmentation of the diene and alkenoid type alkaloids, which has 37 been very important the the classification of newly discovered Erythrina alkaloids. More information on the mass spectra of Erythrina alkaloids will be discussed in Chapter 4 and 5. In addition to the use of OV 17 type packed columns for gas chromatographic Erythrina alkaloid analysis, several researchers began using normal phase HPLC to separate and isolate Erythrina alkaloids because several Erythrina alkaloids could be analyzed using the GC method. Reversed phase liquid chromatographic techniques up until the present have only been used for sample purification after column separation, not for Erythrina alkaloid analysis. This chapter describes the clonal selection of thirteen Erythrina clones from the Nitrogen Fixing Tree Arboretum at CAT1E, the subsequent plant collection and sample preparation and the development of TLC and reversed-phase HPLC techniques for the separation of Erythrina alkaloids. Ultimately, liquid chromatography was not used in the analysis of the Erythrina alkaloids. Gas chromatography/mass spectrometry proved to be the more sensitive method. However, the reversed phase HPLC technique could be used when GC/MS is not available.

3.2 Experimental

3.2.1 Clonal Selection Erythrina clones were selected on the basis of genetic quality, nutritional quality, and biomass availability after consultation with animal nutritionists, forest geneticists and other scientists at CAUL, Turn alba, Costa Rica. Clones were originally included in the arboretum on the basis of phenology, biomass production and suggestions from farmers. A summary of this data for each of the thirteen clones selected is presented in Table 3.1. 38

Table 3.1. Selected Erythrina Clones and Important Characteristics.

CLONE# PROVIDENCE NUTRITIVE PHENOLOGY^.* BIOMASS SPECIE 1 (ORIGIN) VALUE2-3 PRODUCTION^

E. berteroanaf* 2674 Sarapiqui med/low superior very high 2652 Naranjo high average high 2670 Iroquois high superior high 2667 La Suiza high superior high 2689 Santa Marfa high superior med/low 2668 Paso Canoas high average med/low 2691 Quebrada low superior med/low 2703 Palo Verde high average low

E. poeppigiana 2661 Tejar high superior med/high 2708 Alajuela high superior med/high 2687 StMarfa Dota high superior med/high 2700 San Pablo high average medium

E. costarricensis6

2750 Paso Canoas high average n/a?

1 Latin American Seed Bank Number, CATIE. 2perez, 1990. 3 Determined in consultation with Marfa Kass, Animal Nutritionist, Animal Production Dept., CATIE. 4D eterm ined in consultation with Edgar Viqudz, Forestry Geneticist, NFTP, CATIE, and tree register information. 5Nitrogen Fixing Tree Project (NFTP) evaluation, ^Clone Numbers 2668, 2691 and 2703 were originally classified as E. costarr teens is. 7Not available. 39

Five E. berteroanas, four E. costaricensis and four E. poeppigianas were selected in order to have a variety of clones from eacr. species. The thirteen clones selected were of priority interest to the Nitrogen Fixing Tree (NFT) Project at CATIE. They were growing under similar environmental and soil conditions in the Latin American Nitrogen Fixing Tree Orchard, San Juan Sur, CATIE, Turriabla, Costa Rica. Three of the E. costarricensis were reclassified as E. berteroana by the NFT project in June, 1991.

3.2.2 Sample Collection and Preparation Twelve individual trees represented each Erythrina clone. The twelve trees were separated into three groups of four at random providing three replicates for each clone. Leaves where detached from the second or third branch of the tree, collected and weighed. Immediately upon arrival at the laboratory, the leaf blades were clipped from the petiole and weighed. Two 100 g samples from each replicate were immediately frozen; one fresh frozen sample was extracted as fresh material and the other was lyophilized (Eyela Tokyo Tatakikai Freeze Dryer LFD1) for two to three days until dry. The remaining leaf blades were placed in aerated paper bags and dried at 60°C for 72 hours and then milled with a Thomas Wiley Model 4 Lab mill to a powder. All samples were collected in June or July of 1989 and transported to LSU after preparation.

3.2.3 Extraction of Fresh and Frozen Material Typical alkaloid extraction procedures were used. One hundred grams of the fresh- frozen material was extracted with 400 mL 50% aqueous ethanol macerated in a Servall Omni Mixer (Sorvall Inc., Norwalk, Conn.). The macerate was filtered through cheese cloth and rinsed with 50% aqueous ethanol. The filtrate was left overnight with 100 mL of hexane to extract the lipids. The hexane was removed and the filtrate adjusted to pH 10 with 25% NH4OH and extracted with 6 x 100 mL ethyl acetate. The ethyl acetate 40 extract was extracted with 5 x 100 mL 2% aqueous H2SO4 which was cleaned up by a 5 x 100 mL extraction with ethyl acetate. The aqueous mixture was adjusted to pH 10 with NaHC03 and 25% NH4OH and reextracted with 5 x 100 mL ethyl acetate which was concentrated to a small volume and dissolved in dichloromethane (DCM) for TLC experiments. The DCM solutions were extracted with 4 x 1 mL of 1 % HC1 for the HPLC experiments. The extraction was monitored using Mayer's reagent. Mayer’s reagent was made by dissolving 1.36 g HgCl2 in 60 mL of water and 5 g KI in 10 mL of water. The solutions were combined and diluted to 100 mL with water.

3.2.4 Total Nitrogen Determination Total nitrogen was determined using a modified microKjedahl method (Association of Official Agricultural Chemists, 1960). A sample of dried leaves (0.10-0.12 g) was weighed using copying paper and placed in a kjeldahl flask. Two mL of H2SO4 was added with the catalytic mixture (a few drops of a solution of 25 g K2SO4, 25 mL CUSO4 saturated water mixed with 2.5 g sodium selenate in 1 liter of concentrated H2SO4). After 10 minutes the flask was placed in the support apparatus and digested for 1 hour or until clear. The flask was cooled and then rinsed down the sides with water until the kjeldahl tube was 1/2 full. Boric acid (25 mL) which contained an indicator was added. The boric acid solution consisted of 7 mL of a solution of 0.450 g methyl red and 0.250 g methyl blue in 250 mL ethanol which was added to 1 L of H2O that contained 20 g of boric acid. The flask was placed in the distillation aparatus and 4 mL of 50% NaOH was added The solution was distilled until the distillate collected was 30 mL and green in color. The distillate was titrated with H2SO4 (0.0863N) until a grey blue. The percent nitrogen and cru3>; protein was calculated as follows; % total 41 nitrogen = [(mL H2S04-bIank)(0.0863 x 0.014 x 100)]/(weight of sample); % crude protein » [(mL H2SC>4-blank)(0.0863 x 0.014 x 6.25 x 100)]//(weight of sample).

3.2.5 Dry Weight Determinations Dry weights were determined at 60° and 100°. After the leaf blades were dried in the oven, the milled material was weighed. This value was taken as the dry weight. Three gram samples were taken from the 60° dried samples and further dried in crucibles at 100° for 24 hours and reweighed.

3.2.6 Blade to Leaf Ratios The total leaf weight was measured in the field and the leaf blade weight was determined after separation from the petiole in the laboratory. This ratio was on a fresh weight basis.

3.2.7 Statistical Analysis A NOVA and correlation statistical analysis was accomplished using STATVIEW 512+ software (Abacus Concepts).

3.2.8 Thin Layer Chromatography Aliquots of 10 to 25 ^L were spotted using a Hamilton syringe on Keisgelguhr 254 TLC glass plates (Merck). These were developed in 95% chloroform/5% methanol and sprayed with Draggendorfs reagent or Ehrlich's reagent. Draggendorfs reagent was made by dissolving 8.0 g Bi(N 03)3-5H 20 in 20 mL HNO3 (density 1.18 or concentration 30%) and dissolving 27.2 g KI in 50 mL of H2O. The two solutions were mixed and left for 24 hours. The solution was decanted and diluted to 100 mL with H2O. This reagent was used with acidic solutions. 42

3.2.9 Reversed Phase HPLC Two buffers were used in the reversed phase HPLC development: 33 mM phosphate buffer at pH 7.4 and 30 mM carbonate buffer at pH 10.4. The carbonate buffer was made by dissolving 1,045 g NaHC03 and 1.32 g Na2C0 3 in 450 mL of water, mixing and then diluting to 500 mL in a volumetric flask. No pH adjustment was attempted. All buffers as well as solvents were filtered and degassed before use. Two different types of HPLC columns were evaluated: 1) a 15 cm Lichrospher C-18 column (Merck, dp = 5 pm, Vm = 1.18 mL) and 2) a 15 cm polystyrene divinyl benzene column (Phenomenex PLRP-S, dp = 5 (im, Vm = 1.85 mL). The void volumes were measured using acetone in 100% acetonitrile. The instrument system consisted of three Rannin Rabbit HPLC pumps with mixer and a ISCO variable wavelength detector set at 254 nm. The flow rate was 1.0 mL/min. The system was controlled and the data acquired using a Dynamax software package.

3.2.10 HPLC/MS Studies The HPLC/MS studies were performed at the Institute for Environmental Studies, LSU. The liquid chromatograph consisted of two Waters model 510 pumps, a Waters model U6K Universal Liquid Chromatograph Injector, a Waters Lambda-Max model 481 LC spectrophometer and a Hewlett Packard 3390A integrator. Only filtered and degassed methanol and water were used ai solvents. The column was a 25 cm polymer coated polybutadiene column (Alcobond Unisphere, pore size = 60 A, dp = 10 pm, Vm = 3.16 mL), a 15 cm Lichrospher C-18 column (Merck, dp = 5 pm, Vm = 1.23 mL) or a 15 cm polystyrene divinyl benzene column (Phenomenex PLRP-S, dp = 5 pm, Vm = 2.29 mL). The void volumes were measured using acetone in 100% methanol. An Extrel ThermaBcam interface was interfaced to an Extrel ELQ400 triple quadrupole mass spectrometer. Optimization was accomplished by spiking caffeine or B-erythroidine in a 43 small amount of mobile phase to produce 100 ppm or 532 ppm solutions, respectively, and varying the liquid chromatographic and mass spectral parameters until a combination of adequate separation on the liquid chromatograph and maximum ion intensity on the mass spec was obtained. Optimum interface conditions were a flow rate of 0.8 mL/min with a capillary i.d. of 0.10 mm or a flow rate of 0.4 mlVmin with a 50 pm i.d. capillary. The vaporizer control temperature was 176° and the helium nebulization gas was set at 36 units. The stage one pressure of the interface was 6.8 torr and the second stage pressure was 0.3 tore. The mass spectrometer was tuned by Dr. Roberto Wong of the Institute of Environmental Studies. Electon impact at 70 eV was used as the ionization method. The methanol concentration in the mobile phase varied from 60% to

100%.

3.3 Results

3.3.1 Statistical Analysis Table 3.2 list the results of the statistical analysis of the Erythrina clones on a dry matter and total nitrogen basis. The blade to leaf ratio proved to be too variable to be useful in distinquishing clones, but there was significant differences between the species in relation to blade to leaf ratio. £. berteroana had a significantly higher proportion of its weight as leaf blade than did E. poeppigiana. There were statistically significant differences between clones of the same species and between species in dry matter and total nitrogen content. For example, E. berteroana clones 2652 and 2670 were statistically different from each other in both parameters with clone 2652 being lower in dry matter and nitrogen content than clone 2670. £. berteroana as a species was lower in nitrogen content than £. peoppigiana, but £. berteroana was generally higher in dry matter content. The reclassified clones 2668, 2691 and 2703 along with clone 2689 44 were significantly higher in dry matter content than the remaining E. berteroana clones. They were also generally lower and very similar in total nitrogen content There was no correlation between dry matter content and nitrogen content

3.3.2 Reversed Phase HPLC The major alkaloid in clone 2652 was B-erythroidinc (see Chapter 4). B-Erythroidine was isolated in 98% purity from the oven dried material of clone 2652 and this was used in the development of reversed phase HPLC. Boekclheide (1952) determined the pKa of B-erythroidine to be 6.8. Since it is generally desirable to separate compounds in reversed phase HPLC in their neutral form, the mobile phase must be buffered at a pH well above the pKa. Unfortunately, silica based reversed phase columns are limited to a pH range of 2 to 8, therefore a buffer of pH 7.4 was used when traditional silica based columns, such as the C-18 column, were used. At an optimized acetonitrile content of 25%, B-erythroidine was satisfactorily separated from early eluting compounds (Figure 3.1) with a k‘ of 1.54, however the asymmetry factor (a) of 3.7 and the resulting plate count (N) of 902 were unsatisfactory. The addition of 1.4 mM triethylamine changed the pH slightly to pH 7.6 and did improve the symmetry of the B-erythroidine peak (a « 3.3) and the plate count (N=1127), but the asymmetry of the peak was still unsatisfactory (Figure 3.2). Since it was desirable to run the HPLC separation well above the pKa of the B- erythroidine, a polymeric column which could be used up to a pH of 12 was investigated (Phenomenex PLRP-S). This column utilizing a carbonate buffer at pH 10 was much more efficient than the Merck column at the same organic modifier concentration (25% acetonitrile) (Figure 3.3). The plate count increased to 1535 and the asymmetry factor decreased to 2.3. B-Erythroidine was also retained longer with a k' of 45

1.98. However, even though the efficiency and asymmetry were better than on the silica based column, they were still poor. Using the latter system, B-erythroidine could be analyzed in 5 minutes by HPLC. It may be possible to use a scale up of this method to collect and isolate Erythrina alkaloids by HPLC. The drawback to this method, however, it that standards are required to identify the alkaloids because UV absorption at one wavelength, although appropriate for quantification is insufficient for identification. At this point only B-erythroidine as isolated by the author from clone 2652 was available as a standard and an alternative technique that would be able to give more information about the compounds of interest for identification purposes was desired.

3.3.3 HPLC/MS HPLC/MS studies were successful in separating and identifying two Erythrina alkaloids, but only at high concentrations. The best results were obtained using the polymeric column due to its high retentive ness. Ionization efficiency at the ThermaBeam interface dropped off precipitously when the methanol concentration dropped below 75%. At water concentrations above 25%, it was difficult to produce an adequate spray in the expansion chamber. Mobile phases with high water concentrations also tended to condense in the interface requiring a time consuming cleanup between runs. Using a high percentage of methanol in the mobile phase to avoid the condensation problem meant that the column had to retain the compounds adequately since the solvent strength was increased. 46

Tabic 3.2 Percent Dry Matter and Total Nitrogen Content of Erythrina Clones

CLONE______DRY MATTER______TOTAL NITROGEN (%) (%) E. berteroana 2652 14.1 al 4.3 a 2667 17.6 b 4.9 abdef 2670 17.8 bed 5.0 b 2674 16.9 d 4.8 d 2689 20.4 c 4.4 ac 2668* 20.7 c 4.4 ac 2691* 21.9 c 4,4 ac 2703* 17.9 b 4.5 c X 18.4 4.6 E. costarricensis 2750 16.4 d 4.5 c E. poeppigiana 2661 16.3 de 5.4 e 2608 14.4 a 5.9 g 2700 16.7 de 5.1 f 2708 15.3 e 4.7 d X 15.7 5.3 l Differing letters between clones indicates significant difference between the average values at die 95% confidence level using Fisher ANOVA analysis.

Optimized HPLC conditions as descibed in the previous secdon were 75% (vol/vol) carbonate buffer at pH 10.4 and 25% acetonitrile. These conditions could not be used in the HPLC experiments because no ions were detected under these conditions. In addition, a slower flow rate was required in order to completely vaporize the mobile phase at the interface. Depending on the i.d. of the capillary in the interface, a flow rate of 0.4 mL/min to 0.8 mL/min could be used. 47

3i

0 5.0 Figure 3,1 HLPC Trace of B-Erythroidine on Merck Lichrospher Column, Conditions: 25% ACN/75% 33 mM phosphate buffer at pH 7.4. Vm=1.18, F=1.0 mL/min.

5 0 0

Figure 3.2 HLPC Trace of B-Erythroidine on Merck Lichrospher Column with 1.4 mM trie thy lamine. Conditions: 25% ACN/75% 33 mM phosphate buffer at pH 7.6. Vm= 1.18, F= 1.0 mL/min.

3 t

0 7.0

Figure 3.3 HLPC Trace of B-Erythroidine on Polymeric Phenomenex PLRP-S Column. Conditions: 25% ACN/75% 50 mM carbonate buffer at pH 10.4. Vm=1.85, F=1.0 mL/min. 48

Despite the aforementioned problems, mass spectra were obtained. The total ion chromatogram of one standard sample is presented in Figure 3.4. Ox o-B-crythroidinc was detected at 2.88 minutes and B-erythroidine was detected at 3.11 minutes. The run was 7 minutes and 631 scans were collected. The mass spectra obtained of B- erythroidine is presented in Figure 3.5. The total ion chromatogram of an extract of a clone (2674) is depicted in Figure 3.6. Only mass spectra of B-erythroidine and oxo-B- erythroidine were obtained from this run. The other peaks could not be identified from their mass spectra due to the large amount of noise.

3.4 Discussion and Conclusions Statistical analysis of total nitrogen content and dry weight of the clones revealed that there are differences between clones of the same species with respect to these two parameters. As a species, E. poeppigiana in general was higher than either E. cosiarricensis and E. berteroana in nitrogen content, whereas E. berteroana was higher albeit more variable in dry matter content than E. poeppigiana. These differences suggest that Erythrina clones may vary in certain parameter due to genetic differences. Several investigators have used normal phase chromatography for Erythrina alkaloids analysis usually on a semi-preparative scale. Chawla et al. (1982) used silica columns with a 96/4/0.1 chloroform/ethanol/concentrated ammonia solution as the eluant in the alkaloid analysis of E. berteroana seeds and leaves. Redha (1983) used silica as the stationary phase and chloroform as a me bile phase in his analysis. Although normal phase chromatography offers the advantage of easy sample isolation because the mobile phases are volatile, it is much more difficult to obtain reproducible separations. In normal phase LC, the water content of the mobile phases must be scrupulously controlled. In order to develop a useful analytical method, reversed phase chromatography was employed because of its good reproducibility. 49

Figure 3.4 Total Ion Chromatogram from the HPLC/MS Analysis of a 500 PPM Solution of Oxo-B-erythroidine and B-Erythroidine. Conditions: Phenomenex column, 100% MeOH, F= 0.8 mL/min, inj. volume = 60 )iL, analysis time = 7 minutes. 50

PAP

91

65 77

! I' 1 30 156

103

:i4 £58 It*

( i I i i r t 60 70 SO 90 MO ! '0 130 1^0 ISO £10 ' 3^0 £^0 ?)o

Figure 3.5 Mass Spectra of B-Erythroidine by HPLC/MS Analysis. Conditions same as in Figure 3.4. Retention time at scan 280 is 3.11 minutes. 51

£ 1 ‘-'s1

Figure 3.6 Total Ion Chromatogram from the HPLC/MS Analysis of a 7000 PPM Solution of an Extract from Clone 2674. Conditions: Phenomenex column, 100% MeOH, F= 0.8 mLVmin, inj. volume = 20 jiL, analysis time = 25 minutes. 52

Mantle et al. (1984) did use reversed phase HPLC as a means of purification of Erythrina alkaloids from the leaves of E. crista-galli. After extraction and preparative TLC, one fraction from the TLC plate was further resolved on a semipreparative octyldecylsilyl column with a methanol/water mobile phase (4/1). They were able to separate crystamidine and oxoerythraline. Cryst ami dine is similar to oxoerythaline except for an additional double bond at the 10-11 position. Our studies have also shown that reversed phase HPLC can be useful in the analysis of Erythrina alkaloids. However, if trace amounts of alkaloids are present and need to be identified, there are other methods which are more sensitive (Chapter 4). 0- Erythroidine eluted within 5 minutes in our system; peak asymmetry was minimized by using a polymeric column and a mobile phase of relatively high pH. If the compounds of interest are known and standards are available, then reversed phase HPLC can be very useful. Unfortunately, this is frequently not the case. In an effort to develop a technique in which the alkaloids could be identified without the use of standards, LC/MS was investigated. Mass spectra of oxo-B-erythroidine and 3-erythroidine were obtained but only when large injection volumes and high concentrations of alkaloids were used. Optimal HPLC conditions could not be used because of the drop in efficiency at the ThermaBeam interface. Buffers could not be used in this system because of the high background which would have increased with the use of buffer Highly aqueous mobile phases developed during the optimization stage of this research could not be used because of condensation at the interface and the high pressures generated at relatively low flow rates (0.8 mL/min). Nevertheless, if samples are highly concentrated this method can be used to give preliminary identification via mass spectra of the compounds of interest. This is the first time LC/MS has been used to investigate Erythrina alkaloids. CHAPTER FOUR: CLONAL EVALUATION OF ERYTHRINA ALKALOIDS BY GAS CHROMATOGRAPHY AND MASS SPECTROMETRY

53 54

4.1 Introduction Erythrina is an important tree species distributed worldwide at low to mid elevations. It is important in Latin America, Africa, Australia and Asia. Erythrina is one of many leguminous trees which are used in agroforestry systems, especially in Central and South America, as living fences, shade for cash crops such as coffee or cocoa, windbreaks, support for climbing crops, and more. The widely distributed Erythrina genus has been used in agricultural systems since first described by taxonomists in Central America. Its high nitrogen content is exploited by fanners in coffee and cocoa plantations where the shade species is pruned and the foliage left on the ground to provide nitrogen to the cash crop. An additional use of the nitrogen fixing capabilities of Erythrina has been proposed by researchers. The proposed new use of Erythrina is as a diet supplement for ruminants just as legumes are used as supplements in the temperate zones (Benavides, 1983). Through microbial fermentation, ruminants have the capability of utilizing various types of organic matter as food. Microorganisms in the rumen of bovines, for example, obtain energy and protein from forage unusable by non-ruminants. Microorganisms in turn produce microbial protein, B vitamins and other valuable nutrients that can be utilized by the bovine (Hungate, 1966). Frequently, the limitation in the growth and quality of the animals is the amount and the availability of digestible nitrogen present in the feed. Many farmers supplement the pasture diet of ruminants with a high energy/high protein concentrate. However, for the small subsistence farmer, the cost of purchasing concentrate is prohibitive. As an alternative for the small farmer, other materials high in nitrogen, such as Erythrina, are readily available and may serve just as well as a concentrate in supplying high quality protein to ruminant animals. Erythrina contains 24.0 to 38.4% crude protein in the dry leaves, 70% of which is digestible by ruminants (Espinoza, 1984). When present in a coffee agroforestry 55 system the Erythrina tree may produce up to 90 kg of dry leguminous foliage per year per hectare (Russo, 1984). Other crops now in use as feeds for animals, such as alfalfa and canary grass, also contain undesirable compounds (Marten, 1981). With time and careful breeding strategies, these compounds have been essentially eliminated from domestic species greatly improving the forage quality of these types of pastures species and forestalling the possibility of milk or meat contamination. At the Centro Agrondmico Tropical de Investigacidn y Ensefianza (CATIE), Turrialba, Costa Rica, an arboretum of nitrogen fixing trees has been established (Viqu6z, Payne and Sanchez, 1985). The Latin American Nitrogen Fixing Tree Arboretum contains plots of individual clones of Erythrina species collected throughout Central America. Thirteen of these clones were selected as described in Chapter 3 for this preliminary study of the content and composition of Erythrina alkaloids in the foliage of three different Erythrina species, E. berteroana, E. poeppigiana and E. costarricensis, to determine if the expression of these alkaloids varies between genetic clones and to identify low alkaloid content clones of Erythrina to be used in the development of feeding programs for small family fanners of Central America. This investigation represents the first attempt ever to investigate Erythrina alkaloid content variation on a clonal basis. It may be the first study of clonal evaluation of any nitrogen fixing tree species and the first study to examine E, costarricensis leaves.

4.2 Experimental

4.2.1 Clonal Selection and Sample Preparation The process of clonal selection and sample preparation prior to extraction was described in Chapter 3. 56

4.2.2 Extraction of Dry and Lyophilized Plant Material The lyophilized and dried samples were extracted using 95% ethanol in a Soxhlet apparatus. Typical alkaloid extraction procedures were used. Figure 4.1 illustrates the flow chart of an extraction of a representative sample.

4.2.5 Statistical Analysis ANOVA and correlation tadstical analyses were performed using Statview 512+ (Abacus Concepts) software.

4.2.4 GC/MS Analysis One microliter of the extract dissolved in chloroform was injected into a Hewlett Packard 5890 gas chromatograph and analyzed with an Hewlett Packard 5971 quadrupole mass spectrometer. Gas chromatographic conditions were as follows: 40°C for 3 min., ramp of 20°C/min. to 250°C and held at 250°C for 25 minutes. A DB-5 (J&W Scientific) 30 m capillary column was used (0.25 mm i.d., 0.25 |im film thickness). The flow rate of helium was 30 cm/sec. The scan rate was 0.5 scans/sec and the mass range scanned was 15-550 mass units. Electron impact at 70 eV was used. The data was acquired and analyzed using DOS based ChemStation (Hewlett Packard) software. Quantitative analysis of six of the Erythrina alkaloids was accomplished using single ion areas of duplicate samples and calibration curves of authentic standards under the GC conditions described above except that the scan range was narrowed to 50-450 amu. 57

Lyophilized or Oven Dried Plant Material (20 g) Soxlet Extraction with Hexane (400 mL)

24 hours

Soxhlet Extraction with Ethanol (400 mL)

24 hours Discard Plant Material

Rotoevaporate Solvent, Dissolve in 2% H2S04 (100 mL)

Extract with 6 x 50 mL Chloroform

Discard Chloroform

Adjust to pH 10 with NaHC03 and 25% NH4OH

Extract with 5 x 50 mL Chloroform

aqueous organic

Acidify to < pH 2 with 25% HC1 Dry over Na2S04, Filter Heat to 60C for 1 hour Remove Solvent

Leave overnight Dissolve in Chloroform to 0.2-5 mg/mL

GC/MS Analysis

Figure 4.1 Extraction Procedure for a Representative Plant Material Sample 58

Ten data points were used for B-erythroidine calibration: 4.134, 20.69, 41.38, 103.4, 206.9, 413.8, 827.5, 2069, 8300, and 16,600 ppm. The calibration curve was linear over this range with an R2 value of 0.982 (Figure 4.2). Five data points between 40 and 850 ppm were used for oxo-B-erythroidine calibration with an R2 of 1.000. Three data points were used for erysodine, erysopine, erythratine and erythraline over the range of 0 to 1000 ppm resulting in a R2s of 0.996, 0.925, 0.978 and 0.949 respectively. The molecular ion (M+) was used as the quantification ion and the M+-15 and M+- 31 ions were used as the qualifier ions for all samples with a plus or minus 20% intensity tolerance. The duplicate samples were run over a three day period along with the standards using the same mass spec tune. If the qualifiers were satisfied on at least one of the duplicates then the alkaloid concentration was determined, however, if the molecular ion was detected in one or both duplicates but the qualifiers were not met for both runs then the compound was considered not to be present.

4.2.5 Identification of Erythrina Alkaloids Alkaloid identification was accomplished by mass spectral interpretation and comparison to authentic standards in the case of B-erythroidine, oxo-erythroidine, erysodine, erysovine, erysopine, erythraline and erythratine, by mass spectral interpretation, literature references and computer library searches in the case of erysotrine and mass spectral interpretation and comparison to literature values in the case of erybidine and a compound of m/z = 254. Spectral data of Erythrina alkaloids taken from the literature are listed in Appendix A. The mass spectra and tabulated spectra of selected alkaloids detected in this study are listed in Appendix B. Figure 4.2a GC/MS Calibration Curve for B-Erythroidinefor Curve PPM. 20,000 to Calibration 0 from GC/MS 4.2a Figure Figure 4.2b. GC/MS Calibration Curve for B-Erythroidinefor Curve PPM.Calibration 250 to 0 from GC/MS 4.2b. Figure

RESPONSE 1.00e+806 W Vi eu o Z Vi bi 2.5e+6 5.0e+5 1.0e+6 1.5e+6 0.00e+0 0 0 .35+ + .14+x FT2 1.0184e+4x + 2.7335e+4 205l+ 125l+x RA2=« 1.285 + 2.0250.982 le+4x le+5 y- 50 5000 100 PPM 10000 PPM 5 250 150 15000 200 20000 59 60

4.3 Results and Discussion Other researchers who have used GC/MS to investigate Erythrina alkaloids have utilized the trimethylsilyl (TMS) derivatization technique with OV-17 packed columns (Hargreaves et al., 1974, Barakat et al., 1979, Redha, 1983). We found, at least for the alkaloids examined in this study, that TMS derivatization was unnecessary. Alkaloids with or without hydroxy groups were well resolved on a DB-5 column under the chromatographic conditions described in the experimental section. Figure 4.3 demonstrates this point. Erysodine, erysovine and their TMS derivatives are present in the same total ion chromatogram (TIC). The underivatized as well as the TMS derivatized compounds are easily detected using this technique. Other Erythrina alkaloids with free hydroxyl groups such as erytratine were also detected with adequate sensitivity without TMS derivatization. The DB-5 cappillary column was easily overloaded with B-erythroidine, however quantitation was linear to 16,000 ppm. The calibration curves showed good linear correlation both at the high and the low ends (Figures 4.2a and 4.2b). However, the overloading problem can be seen in Figure 4.4 which shows the total ion chromatograms of several calibration points up to 8300 ppm. At 2069 ppm the fronting of the peaks can be discerned, therefore all samples for quantitation were run at or below 2069 ppm and a calibration curve of 0 to 2069 ppm was used for calibration. Alkaloid artefacts that oven-drying the leaf samples might introduce were also of concern. Bhukani et al. (1976) reported that some of the oxo Erythrina alkaloids may have been produced in the drying process. For this reason, all the samples in our study were divided into two groups, one which was lyophilized and one which was oven dried. However, there was no qualitative difference in the alkaloid profile between the oven dried or lyophilized samples (Figure 4.5). 8-Oxo-erythroidine (retention time =■ 29.8 minutes) was present in samples processed by both methods. There was a 61 difference, however, in the extraction efficiency between the two methods of sample preparation. Extraction of the oven dried material yielded slightly more alkaloid extract by weight per gram sample than did the extraction of the lyophilized samples. This may have occurred because the oven dried samples were milled whereas the lyophilized samples were crushed resulting in finer particles in the oven dried case. Recovery of the alkaloids from the lyophilized plant samples was estimated by performing a series calculation based on two subsequent extractions of the same sample for each sample of one replicate. The amount extracted in the second extraction was taken to be a constant percentage that remained and would remain if extractions were performed indefinitely. This method worked well as the majority of the alkaloids were extracted during the first 24 hour extraction period and recoveries varied between 80 and 100%. Most of the sample recoveries were above 90%. There has been some controversy as to whether the expression of secondary plant products such as alkaloids is under genetic or environmental control. It is commonly observed that the alkaloid composition of a plant may vary with environmental conditions such as soil nutrient level, soil type, soil pH, elevation, drainage, sunlight and stress, but no conclusions have been demonstrated using these parameters (Bick et al., 1953a and 1953b; Keeler, 1971; Blua and Hanscom, 1986). On the other hand a correlation has been found between alkaloid content and the developmental stage of the plant (Krukoff and Bameby, 1974). 62

Kbundanc TIC: LERYSODl.D

9000000-

8900000 - arysodina 8000000-

7500000

7000000■

6900000 -

6000000■

5500000 THS erysodlna

5000000

4500000

4000000 ERYS0D1NE STD FROH JACKSON

3500000

3000000

2500000

2000000

1500000 THS iryiovln*

ioooooo aryaovim 500000 -

0 1 ' ' I ' 1 ' 1 M ’ ' ' I ■ ' ,"'~l ' ' ■ ’ l *-*■» ' I ■ ■ ■ • i ■ 1 ' • i 1 1 • 1 I ' ' ' ri»a -> 20-00 21.00 22.00 23.00 24.00 25.00 26.00 27,00 28.00

Figure 4.3 Total Ion Chromatogram (TIC) of Erysodine and Erysovine with Trimethylsilanized (TMS) Eysodine and Erysovine. 63

414 ppi

4000000

2000000 -

6000000

Figure 4,4 Total Ion Chromatogram (TIC) of Selected Concentrations of B- Erythroidine Used in Calibration. 2500000

Lyophl1izad

1500000

500000■

> 20.00 26. 00 28.00 30 . OO

Dried 2000000 -

1500000

1000000

500000

rise ->20.00 22 .00 24 . 00 28.00

Figure 4.5 Total Ion Chromatogram (TIC) of Extracts of Plant Material Lyophilized and Oven Dried at 60°C. 65

In an effort to eliminate environmental variation in alkaloid composition in the present study, the clones utilized were grown under the same environmental conditions in the Nitrogen Fixing Tree Arboretum of Latin American in San Juan Sur, CATIE, Turrialba, Costa Rica. They were originally obtained from diverse regions of Costa Rica and therefore are genetically distinct. Any difference between the alkaloid content of the clones in this study should be primarily attributable to genetic differences and not environmental differences. Taking a larger genetic view, there have been attempts to use the presence of a particular alkaloid to taxonomically classify ambiguous plant specimens or species. In some genera, such as Nicotiana, the alkaloid composition was an unreliable parameter of phylogenetic position (Cramner and Turner, 1967). On the other hand, Hargreaves et al. (1974) were able to distinguish between old and new world Erythrina species by their alkaloid profiles. Our studies revealed that E. poeppigiana, E. berteroana and £. costarricensis, which are all members of the new world Erythrina group, are more similar than distinct with respect to the specific Erythrina alkaloids encountered and it may be difficult to classify these species by their alkaloid profiles. All three of the above species tested in our laboratory belong to the American species of Erythrina. E. poeppigiana belongs to the most modem Erythrina taxonomic section, Micropteryx, while E. berteroana and E. costarricensis belong to the section Erythrina, which is still evolving (Krukoff and Bameby, 1974). As a matter of fact, it is difficult to distinguish between E. berteroana and E. costarricensis and it is believed that these two species may interbreed (Folkers et al., 1941). As a case in point, three clones originally classified as E. costarricensis (2668, 2691 and 2703) were reclassified as E. berteroana during the course of this study by the Nitrogen Fixing Tree Project at CATIE where the arboretum is located. 66

The total ion chromatograms of a representative lypophilized sample of each clone are presented in Figures 4.6, 4.7 and 4,8. The Erythrina alkaloids eluted between 20 and 33 minutes, the range included in the figures. Figure 4.6 shows the five original E. berteroana clones, Figure 4.7 shows the three reclassified E. berteroana clones and the one E. costaricensis clone and Figure 4.8 shows the TICs of the E. poeppigiana clones. The most obvious peak in all the chromatograms is B-erythroidine at 21.6 minutes. Erysodinc eluted first at 20.8 minutes, and then erythaline at 21.2 minutes close to the B- erythroidine peak. a-Erythroidine elutes after B-erythroidine at 22.3 minutes. Erythratine elutes at 24.8 minutes; erybidine at 26.8 minutes and 8-oxo-B-erythroidine at 29.5 minutes. The predominant Erythrina alkaloids detected in the clones of Erythrina berteroana, £. costarricensis and E. poeppigiana are presented in Table 4.1. The variation in alklaloid content of the clones will be discussed later in the chapter. All the clones contained B-erythroidine (ERTH) as the major alklaloidal constituent. The alkaloid 8- oxo-erythroidine (OXOE) was found in all but four clones with representatives lacking this alkaloid from each species. In other words, oxoerythroidine was not species specific in this study. Erysodine (EYSD) was detected principally in the E. berteroana clones at low levels, while erythraline (ERTL) was detected in three E. berteroana clones in relatively high levels. Erythraline was detected in only one E. poeppigiana clone at about 1 ppm. The one E. costarricensis clone did not contain oxoerythroidine, erysodine or erythraline, Erybidine (ERBD) was almost ubiquitous in this study. Abundance Ion range 15.00 to 550.00 2652

4000000

2000000

rime ->2o.oo 2 2 . 00 24 .00 26.00 28 . 00 30. 00 Vbundance Ion range 15.00 to 550.00 2674

2000000 -

rime ->20.00 22 .00 24 . 00 26.00 28. 00 30. 00 \bundance Ion range 15.00 to 550.00 2670

4000000

2000000

Time ->20.00 22 . 00 24 . 00 26. 00 28 .00 30. 00 Sundance Ion range 15.00 to 550.00: 2667

4000000 -

2000000 •

rime ->2Q.00 22 .00 24 .00 26 . 00 28 ■ 00 30. 00 \bundance Ion range 15.00 to 550.00 2689

4000000-

2000000 -

Time ->20.00 22. 00 24 . 00 26. 00 28 . 00 30.00

Figure 4.6 Total Ion Chromatograms (TIC) of E. berteroana Clones. 68

Abundance Ion range 15.00 to 550.00 2668 4000000 -

3000000

2000000

1000000 yx—/V, rime ->20.00 22 . 00 24 .00 2 6 . 00 28 . 00 30 . 00 Ion r a n g e 15.00 to 550.00:

3000000

2000000

1000000

Time ->20.00 22 .00 24 .00 26. 00 28 . 00 30 . 00 Ion range 15.00 to 550.00: 2703

3000000

2000000

1000000 yv

Time ->20.00 22 . 00 24 .00 26.00 28 . 00 30. 00 Abundance Ion range 15.00 to 550.00 2750 4000000 -

3000000 -

2000000 - 1000000 - TV rime ->20.00 22.00 24 . 00 26. 00 28. 00 30. 00

Figure 4.7 Total Ion Chromatograms (TIC) of Reclassified E. berteroana and E. costarricensis Clones. Ion range 15.00 to 550.00 2661

4000000

2000000

Tine ->20.00 22.00 24 .00 26.00 28 . 00 30. 00 Ion range 15.00 to 550.00 2700

4000000

2000000

rime ->20.00 22.00 24 .00 26. 00 28 . 00 30.00 Ion range 15.00 to 550.00: 2708

4000000

2000000

rime ->20.00 22.00 24 . 00 26 . 00 2B.00 30.00 Ion range 15.00 to 550.00: 2687

4000000 -

2000000 -

rime ->20.00 22 . 00 24 .00 26 . 00 28.00 30.00

Figure 4.8 Total Ion Chromatograms (TIC) of E. poeppigiana Clones. Tabic 4.1 Alkaloid Profiles of Selected Erythrina Clones.

CLONE# ERTR OXOEEYSD ERTL ERBD 247A 247B 245 E. berteroana 2652 3548 172 * * 2667 1039 21 0.2 * * * 2670 804 * * * 2674 266 6 * * * * 2689 1456 2 303 * * • * 2668x1 166 2 * * • * 2691x 745 3 4 234 * * * * 2703x 657 1 192 * * * * E. costarricensis 2750 8 * * E. poeppigiana 2661 477 10 * * 2687 196 4 * * 2700 131 1 1 * 2708 836 8 1 * * *

ERTR=B-erythroidine, OXOE = 8-oxo-B-erythroidine, ERSD - erysodine, ERTL = erythraline, ERBD = erybidine, 247A, 247B, 245B, and 254 are unidentified compounds. All values arc averages of duplicate samples in |ig/g dry weight. See Table 4.3. for more details on quantitation. * indicates presence only. A space indicates that the alkaloid was not detected in either replicate. 1 These are clones that were orginally classified as E. costarricensis, but are now classified as E. berteroana. 71

Alkaloids not detected in these clones for which standards were available were erysovine, erysopine and erythrine. Erythratine and erythratidine may have been present in a few clones but in very small amounts. No standard was available for comparison purposes in the case of erythratidine and the qualifiers were not satisified in the case of erythratine. The mass spectra of these compounds are depicted graphically and in tabular form in Appendix B. Five compounds were detected in the clones that could not be identified. Several lower molecular weight compounds, two isomers with a molecular weight (MW) of 247 and one compound with a molecular weight of 245, eluted before the other Erythrina alkaloids with retention times between 15.5 and 16.5 minutes (Figure 4.9). The unknown of MW of 245 was almost ubiquitous in the samples analyzed. The 247 isomers were concentrated in E. berteroana and primarily accompanied by the unknown 245. The 247 isomers may be dihydroderivatives of the 245 compound. From their mass spectra it is apparent that they contain a methoxy group (M-15 and M-31). Trimethylsilation revealed one hydroxy group in each of the unknowns. The fragmentation pattern and the presence of a peak at m/z of 130 suggests that these unknowns may be dienoid type Erythrina alkaloids as yet undescribed. Additional work is required in order to further elucidate the structures of these unknowns. A compound of molecular weight 254 matches the description of a new type of Erythrina alkaloid reported by Redha (1983) which contains two nitrogens. The mass spectra are identical within experimental error, but an authentic standard of the alkaloid was unavailable for retention time matching via standard addition. The unknown of molecular weight 289 may be a dihydroderivative of a known Erythrina alkaloid. Examination of the alkaloid profile reveals that two of the reclassified clones (2691 and 2703) along with E. berteroana clone 2689 have similar alkaloidal characteristics. Clones 2689 and 2703 lack 8-oxo-erythroidine as does the one remaining E. 72 costarricensis clone. However, clones 2691, 2703 and 2689 also contain the unknowns 247A and 247B which are not present in E. costarricensis but are common in E, berteroana. Alternatively, only E. berteroana clones 2689, 2691, and 2703 contain erythraline, a dioxymethylene Erythrina alkaloid. It is possible that one or more of these three clones represents a hybrid between E. berteroana and E. costarricensis. They were certainly the most interesting in terms of alkaloid composition of all the clones studied. To fully assess the utility of alkaloid profiles for taxonomic classification of Erythrina species, additional genetic clones have to be analyzed for alkaloid content, especially of clones E. costarricensis which is poorly represented in this study. By far the alkaloid present in the greatest quantity in all three species was fi­ ery throidine. B-Erythroidine represented 75% or more by weight of the predominant alkaloids detected. B-Erythroidine had been detected in the seed of E. poeppigiana, E. berteroana and E. costarricensis, and had been reported as a major alkaloid in the leaves of E. berteroana and E. poeppigiana (Table 4.2; Romeo and Bell, 1974; Jackson et al., 1982). Usually when B-erythroidine occurs it is frequently the major alkaloid. This may be explained because B-erythroidine is the final biosynthetic product in Erythrina alkaloid biosynthesis. This subject is addressed later in this chapter. The alkaloid content of the leaves of E. costarricensis may have never been studied before this present study (Table 4.2), although erythroidine was reported in the seeds by Folkers in 1945. Jackson et al. (1982) detected 8-oxo-erythroidine in their samples of E. berteroana, but not in E. poeppigiana. In our study, 8-oxo-erythroidine was present in all E. poeppigiana clones. Erysodine was present in five of the clones, four E. berteroanas and one E. peoppigiana, however it wasn't as prominent as in other studies (Table 4.2). Erysovine and erysopine were not detected. Erythraline was reported by Barton et al. (1974b) in whole E. berteroana plants, but not in E. berteroana by Jackson et al. (1982). 73

Erythraline was reported by Ito et al. (1975) in the leaves of E. poeppigiana, but it was not detected in E. poeppigiana leaves by Jackson et al. (1982). In our studies, the larger quantities of erythraline were found in E. berteroana, however, the presence of erythraline was highly variable between clones and this may explain the varied results among research labs concerning the presence of erythraline in these two species. Erybidine had been reported in E. poeppigiana by Ito and Jackson (Table 4.2), but has never been reported before in E. berteroana. Of the unknowns, only MW 254 matches the description of an unknown compound in another work (Redha, 1983). This is the First time that a dinitrogen compound has been reported in any of these three species. The compounds with molecular weights of 245 and 247 have not been mentioned by other groups. It seems that the MW 247 compounds may be more species specific than the other Erythrina alkaloids detected. It would be interesting to pursue the presence of these compounds in other clones and other species of Erythrina to verify this theory. Romeo et al. (1974) reported that there was no correlation between amino acid content and alkaloid content in the species he tested. This was corroborated in our studies where there was no correlation (r2»-0.014) between total nitrogen content and fl- ery throidine content (Table 4.3). The non-correlation of alkaloid content with nitrogen or protein content has been observed in other nitrogen fixing tree species such as Leucaena (Gonzales, 1967). This is fortunate from a genetic engineering standpoint because it suggests that alkaloid content end nitrogen content are not genetically linked. Moreover, there is great variability between clones of the same species in fi-erythroidine content (Table 4.3). There are low alkaloid content clones of each species which might be preferred for use in the development of alternative animal feeds. This variability also suggests a means of breeding low alkaloid, high protein Erythrina varieties for animal feed. Importantly, variation between individuals of the same genetic clone was very 74 small both in total nitrogen (analyzed in triplicate) and B-erythroidine content (analyzed in duplicate). Once it was determined that B-erythroidine was the major alkaloid in these three Erythrina species, it was not surprising to find very little variation in alkaloid composition between the species. B-Erythroidine is formed biosynthetically by diol type cleavage of the aromatic Erythrina alkaloids (Chapter 2). The cleavage is probably ex era-diol because of the occurrence of Erythrina alkaloids with a nitrogen at the C-16 position. An example of a dinitrogen Erythrina alkaloid is melanacanthine reported by Redha (1983). Therefore, if B-erythroidine is the final product of the Erythrina alkaloid biosynthetic pathway in the Erythrina species examined, then it follows that all possible biosynthetic precursors, namely the aromatic types of Erythrina alkaloids, are shunted into this pathway and are convened principally to B-erythroidine, which is the major alkaloid detected. The aromatic Erythrina alkaloid precursors may not be present or detected in significant amounts because they may be rapidly converted to B-erythroidine. 2000000

1500000

500000-

Scan 120 (15.675 sin ) I LL26743N.D (-)

100000 -

194 214 1 4 1j ° 144 159 X70 I 232 M ’ i tk i i i‘M| i f*r? [T i - > 20 40 40 SO 100 120 140 190 190 200 220 240 Scan 124 (15.922 nin)i LL26T43H.D {-)

100000 -

194 144154 170

- > 20 40 60 90 100 120 140 160 190 200 220 240 ^cari- 334 (TeTlTt^mtnTi LlTSTvthTd- ?"^ 00000 92 T

77 91 29 51 144 16069 1#4 19#

1/Z -> 20 4060 90 100 120 140 160 19 0 200 2 20___240

Figure 4.9 Total Ion Chromatogram (TIC) and Mass Spectra of Three Unknowns (A) 247A, (B) 247B, and (C) 245 . 76

Tabic 4.2 Reports of Erythrina Alkaloids in the Species Studied by Other Investigators

Reference Plant Pan Alkaloids Identified Erythrina berteroana Folkers, 1940c seeds erysopine erysovine Folkers, 1942a seeds erysodine cry so nine Folkers, 1945 seeds erythroidine Barton, 1974b whole plants a, B-erythroidine erythraline Hargreaves, 1974 seeds erysovine erysodine erysopine erysonine erysoline a, B-erythroidine Romeo, 1974 seeds a, B-erythroidine Jackson, 1982 leaves a, B-erythroidine (major) 8 -oxo-ery throidine (1 unknown) Jackson, 1982 leaves/seeds erysodine erysovine erysopine erysoline erysonine a, B-eiythroidine (major) 8-oxo-ery throidine Erythrina costarricensis Folkers, 1941 seeds erysodine erysovine erysonine etysopine Folkers, 1945 seeds erythroidine 77

Table 4.2 Reports of Erythrina Alkaloids in the Species Studied by Other Investigators (continued)

Reference Plant Part Alkaloids Identified Erythrina poeppigiana Folkers, 1940c seed erysodine erysovine Folkers, 1944 seed erysothiovine Folkers, 1945 seed erythroidine Barton, 1973 leaves a, B-erythroidine Romeo, 1974 seed a, B-erythroidine Hargreaves, 1974 seed erysotrine erysodine erysovine erysopine erythratidine a, B-erythroidine Ito, 1975 leaves erysotrine erysodine 1 l-hydroxyerysotrine erythraline erythrinine erybidine Jackson, 1982 leaves erysodine erysopine erysotrine erythratidine (major) erythratidinone 11-hydroxyerythratidine 11-hydroxyepierythratidine erybidine (major) a, B-erythroidine (major) (3 unknowns) 78

Table 4.3 B-Erythroidine and Nitrogen Content of Thirteen Erythrina Clones

mg/lOOg fresh wt. M-g/g dry wt. % dry w Very High 2652 berteroana 61.58 al 3548 a 4.28

High 2689 berteroana 35.26 b 1456 b Mid Range 2667 berteroana 22.48 c 1039 be 4.92 2708 poeppigiana 18.92 cd 969 c 4.68 2691 berteroana 18.01 cd 745 cd 4.39 2670 berteroana 16.52 cde 804 cd 5.05 2703 berteroana 14.92 cde 657 cde 4.50 Moderately Low 2661 poeppigiana 8.44 def 477 df 5.43 2674 berteroana 6.33 ef 266 df 4.81 Lower 2668 berteroana 3.71 f 196 f 4.39 2687 poeppigiana 3.39 f 168 ef 5.86 2700 poeppigiana 2.77 f 131 f 5.08

Very Low 2750 costarricensis 0.17 f 8 f 4.48

1 Differing letters between clones indicates significant difference between the average values at the 90% confidence level using Fisher ANOVA analysis. For statistical analysis of total nitrogen content please see Chapter 3. 79

4.4 Conclusions GC/MS using a 5% phenyl/95% methyl column is a valuable technique for the analysis of Erythrina alkaloids. Clonal evaluation of E. poeppigiana, E. berteroana and E. costarricensis revealed that B-erythroidine, a drug previously used in surgery and electroshock treatments, was the major alkaloid in the leaves of these species. There was little statistical variation in the parameters measured between individual trees of the same genetic clone. There was wide variation in the concentration of B-erythroidine between Erythrina clones of the same species and between different Erythrina species, however there was little difference in the types of Erythrina alkaloids detected between clones or between species. Low alkaloid content clones of all three species were identified. Alkaloid content and nitrogen content were not correlated, which suggests that these important parameters are not genetically linked. Low alkaloid containing clones may be genetically engineered or bred to contain even lower amounts of alkaloids without affecting or even improving the total protein content. For the present, low alkaloids clones should be developed over high alkaloid content clones for alternative feeds. CHAPTER FIVE: POSSIBLE METABOLIC FATE OF ERYTHRINA ALKALOIDS

80 81

5.1 Introduction The use of nitrogen fixing crop species, such as alfalfa and clover, as a high protein animal feed is an established practice in U.S. agriculture. However, in many tropical countries nitrogen fixing pasture species are not commonly used. Temperate legumes do not grow well due to a combination of climate, competition and lack of traditional commercial inputs. On the other hand, nitrogen fixing tree species are abundant in the tropics. Many have been introduced or naturalized as a result of the coffee and cocoa industry where crops are managed below a prunable canopy of trees. Innovative uses of these trees have been developed by several international development organizations such as the Centro Agrondmico Tropical de Investigacidn y Ensefianza (CATIE) in Costa Rica. Initial studies have demonstrated that Erythrina is a very versatile and easy to use nitrogen fixing tree (Russo, 1984). Erythrina has been used successfully in several agroforestry systems, for example as live fences, as shade for coffee plantations, as guides for climbing crops and more recently as a feed supplement for animals. Several studies by the Nitrogen Fixing Tree Projoect at CATIE have indicated that the foliage Erythrina poeppigiana and E. berteroana can substitute for all or a portion of the protein required by a ruminant with no detectable adverse effects (Samur Rivero, 1984; Vargas Fournier, 1983; Benavides, 1986). On the other hand, Erythrina is not a suitable feed for monogastric animals. Toxic symptoms developed during feeding studies with Erythrina in several non-ruminant species. Rats fed a diet consisting of 51% (w/w) E. berteroana lost weight over a 28 day experiment while the controls more than tripled their weight (De Leon, 1984) and pigs fed a diet containing less than 3% E. poeppigiana failed to gain weight and were removed from a study due to their poor physical condition (Vargas Fournier, 1983). Sterility and death resulted in a study involving rabbits fed Erythrina (Martin, 1984). Additional evidence that Erythrina may be toxic to 82

non-ruminants animals lies in its use as a medicinal herb to combat insomnia in humans and its use by South American Indians as a fish poison (Dyke and Quessy, 1981). One explanation of the toxicity differences between ruminant and monogastric animals with respect to Erythrina species is that some toxic compound(s) in the foliage of the Erythrina species may be detoxified by the microorganisms in the rumen and may not be detoxified by the gastrointestinal tract of the monogastric animals. Alternatively, some compound(s) may be metabolically activated in the monogastric animals and not in the rumen of the ruminants. The reductive neutral atmosphere of the rumen may generate very different metabolites than the acidic stomach of the monogastric animal. As reviewed in Chapter 2, the genus Erythrina is known to contain curare like alkaloids (Dyke and Quessy, 1981). Clinically active alkaloids were isolated from the seeds of various Erythrina species in the 1930s and 1940s by Folkers and coworkers at Merck Pharmaceutical Company (Folkers and Major, 1937; Unna et al. 1944). These alkaloids were of interest because they retained neuromuscular transmission blocking activity even when administered orally. Further studies resulted in the isolation of several other "Erythrina-\ike" alkaloids from the seeds of various Erythrina species. The majority of new Erythrina alkaloids were less or as potent as the original drug, B- erythroidine (Figure 5.1). In the 1980s, B-erythroidine was isolated from the leaves of E. poeppigiana and E. berteroana (Jackson and Chawla, 1982). Work described in Chapter 4 confirmed that B-erythroidine was not only present in E. berteroana and E. poeppigiana, but was the major alkaloid present in these species (Payne and Foley, 1992). There is some circumstantial evidence that alkaloids of alkaloid containing plants may be transported to the milk of animals fed these plants. In the pioneering days of the United States, a disease known as milk sickness caused many fatalities. The disease was traced to the consumption of milk from cows that had eaten white snakeroot 83

(Eupatorium rugasum) (Chccke and Shull, 1985), In a more recent case, anagyrine present in the milk of diary animals which consumed Lupinus latifolius, may have caused birth defects in California (Kilgore et al. 1981). There is additional evidence that other types of alkaloids, namely ergot, indolizidine and quinolizidine alkaloids also may contaminate the milk of dairy animals (Cheeke and Shull, 1985). Milk transfer of toxins can occur with mycotoxins and ketones as well as the alkaloids mentioned (Cheeke and Shull, 1985). Much of the recent work demonstrating the milk transfer of naturally occurring toxins has centered on the species tanzy ragwort (Senecio jacobaea), a poisonous plant found in the Pacific Northwest known to contain hepatoxic pyirolizidine alkaloids. In one study (Johnson, 1976), histopathologic changes were not detected in the calves consuming milk of the cows eating 5. jacobaea, but clinical chemical tests showed hepatic biochemical lesions in calves. Rats fed goat's milk which contained 7.5 ng of pyrrolizidine alkaloids developed swollen hepatocytes of centrilobular distribution and biliary hyperplasia (Goeger et al. 1982). In another study in rats there was a high rate of transfer of the toxic substances through the milk as measured by lesions in the young. The dose given to the mother was not lethal but was lethal to the suckling young (Goeger et al., 1982). Additional studies by other investigators also indicate that Senecio alkaloids can be secreted into the milk of animals that consume Senecio spp. (Dickinson et al. 1976). In many instances, these naturally occurring toxins in the milk of dairy animals pose no threat to the general public in countries with a developed dairy industry because the toxins are highly diluted with uncontaminated milk at processing centers before distribution. However, naturally occurring toxins are of concern to the small family farmer in developing countries who may use only one or two animals to provide milk for the entire family over a long period of time. In this case, any toxins consumed by 84 the dairy animals may concentrate in the milk and pose a health threat to the family. Since Erythrina foliage is known to contain toxic alkaloids which cause muscle paralysis, it is imperative that the milk of the animals consuming Erythrina foliage be examined for alkaloid content and for the presence of toxic metabolites (Figure 5.2).

17 U

8 O 7

Figure 5.1 8-Erythroidine Figure 5.2 Schematic of the Milk Transfer of Toxins in Goats. 86

5.2 Experimental

5.2.1 Feeding Study Twenty dairy goats were preselected from the herd at the Animal Production Department, Centro Agrondmico Tropical de Investigacidn y Ensefianza (CATIE) in Costa Rica according to due date. The goats were crosses of Nubian, Totten berg, Saanen and Alpine and ranged in age from two to eight years. One month before expected parturition, the selected animals were fed a diet of 1 kg concentrate with grass and banana ad lib. This diet served as the control diet during the study. After parturition, the animals were separated into treatment groups. Treatment I was the control diet which was the same as the pre-experimental diet Treatment II consisted of a diet of 40% by weight E. poeppigiana foliage and 60% grass ad lib. with no concentrate. Treatment III consisted of a diet of 40% by weight E. berteroana foliage and 60% grass ad lib. The animals were conditioned on these diets for one month before milk samples were taken. Data collection related to the feeding experiment occurred between October, 1989 and January 1990; feed consumption was recorded daily, body weights were recorded monthly, milk production was recorded weekly and milk samples were taken weekly for four weeks and then monthly for two months. For the milk analysis, a representative sample of 100 mL of milk was taken from the total daily milk production of each goat, immediately frozen and then lyophilized for storage until analyzed.

5.2.2 Milk Analysis The lyophilized sample representing 100 mL of milk was placed in a Soxhlet apparatus and extracted with 400 mL of hexane for 24 hours. The solvent was evaporated and the residue dissolved in 20 mL of 2% H2SO4. The acid solution was 87 adjusted to pH 10 with NaHCC>3 and 25% NH4OH and extracted with 3x10 mL chloroform and dried over Na2S0 4 . After solvent evaporation, the residue was dissolved in enough chloroform to provide a 1-5 mg/mL solution for GC/MS analysis. The material remaining in the Soxhlet device after the hexane extraction was extracted with 400 mL ethanol. After rotoevaporation, the residue was dissolved in 50 mL H2SO4 and extracted with 4x10 mL chloroform which was discarded. The acidic solution was adjusted to pH 10 with NaHCC>3 and 25% NH4OH and extracted with 3x10 mL chloroform and dried over Na2SC>4. After solvent evaporation, the residue was dissolved in enough chloroform to provide a 1-5 mg/mL solution for GC/MS analysis. The basic solution remaining after chloroform extraction was readjusted to pH<2 with 25% HC1, heated to 60°C over a water bath for 1 hour and left overnight to encourage hydrolysis. This fraction was worked up for GC/MS analysis as described in the preceding paragraph.

5.2.3 Rumen Incubations Dried plant material (0.5 g) was incubated with 50 mL of a 20% rumen liquor/80% artificial saliva solution in sealed tubes flushed with CC>2 in an incubator at 40°C and under an atmosphere of CO2 for 48 hours. The rumen liquor was obtained by squeezing the rumen contents of a fistulated goat though cheese cloth. The artificial saliva was made from 9.8 g NaHC03, 7.0 g Na3P04, 5.7 g KC1,4.7 g NaCl and 1.2 g MgSC>4 in 10 L of distilled water. 1 mL of 5 3 g CaCl2 in 100 mL of water was added for each liter of mixed rumen liquor solution. The incubations were terminated by the addition of 5 drops of 25% HCI. The contents of the tubes were rinsed into lyophilization flasks, immediately frozen and then lyophilized for further analysis. The 88 lyophilized rumen samples were extracted and analyzed according to the procedure described in the milk analysis section.

5.2.4 GC/MS Analysis One microliter of the extract dissolved in chloroform was run on a Hewlett Packard 5890 gas chromatograph and analyzed with an Hewlett Packard 5971 quadrupole mass spectrometer. Gas chromatographic conditions were as follows: 40°C for 3 min., ramp of 20°C/min. to 250°C and held at 250°C for 25 minutes. The column was a DB-5 (J&W Scientific) 30 m capillary column with a film thickness of 0.25 pm and an i.d. of 0.25 mm. The flow rate was 30 cm/sec of helium. The data was acquired and analyzed using DOS based ChemStation (Hewlett Packard) software at a scan rate of 0.5 scans/sec from 15-550 amu. Ionization was electron impact at 70 eV. The mass spec was tuned according to the EPA semi-volatile method. The threshold was set at 1500.

5.2.5 Synthetic Hydrogenation B-Erythroidine (15 to 30 mg) was dissolved in 1 mL of water to which 62 mg of solid NaOH and 1 g of raney nickel (Aldrich) was added. The solutions were either left stirring overnight or placed under 3 atm. of H2 for 90 min. The mixture was filtered through 0.45 pm nylon filter, rinsed, acidified and left overnight. The acidic solution was adjusted to pH 10 with 25% NH4OH and extracted with 5x5 mL chloroform, which was then evaporated and the resulting residue dissolved in enough chloroform to make a solution of 1-5 mg/mL for GC/MS analysis.

5.2.6 Statistical Analysis Statistical analysis was performed using Statview 512+ (Abacus Concepts). 89

5.3 Results Through the use of mass spectral interpretation and comparison to the standard fi­ ery throidine, dihydroerythroidine isomers and tetrahydroerythroidine were identified. The dihydro- and tetrahydroerythroidines were detected in the milk of the experimental animals in the two treatment groups consuming E. poeppigiana and E. berteroana foliage and in the rumen incubation solutions. The tabulated mass spectra of these compounds and that of B-erythroidine are presented in Table 5.1. The various reduced compounds were well resolved on the DB-5 column and represented a very small portion of the total residue from the milk samples. The majority of the compounds in this fraction were long chain aliphatic compounds which were also well resolved. Three dihydroerythroidines (DHEI, DHEII and DHEIII) and one tetrahydroerythroidine (THE) were detected in the milk and rumen samples. The mass spectra of DHEI, DHEII and THE are depicted graphically in Figures 5.3, 5.4 and 5.5 respectively. DHEI constituted the major portion of the DHEs in the milk and rumen samples, DHE II was present in smaller amounts and DHEIII was present in trace quantities, if at all. Some unreduced fi-erythroidine was detected in two of the milk samples and in two of the rumen samples. No erythroidine, dihydroerythroidine or tetrahydroerythroidine compounds were detected in the milk collected from the goats in the control group consuming a diet free of Erythrina foliage. Two dihydroerythroidine (DHEIV and DHEV) isomers were the only products of the synthetic hydrogenations over rancy nickel. DHEIV was the major isomer. No parent B-erythroidine or tetrahydro- compounds were detected. The mass spectra of major DHE isomer from the biological hydrogenation (DHEI) and the major isomer from the synthetic hydrogenation (DHEIV) are presented in Figure 5.6. The synthetic dihydroerythroidines were used to estimate the amount of dihydroerythroidines in the milk samples using the areas in the total ion chromatogram and background subtraction. 90

Figure 5.7 and 5.8 display the results of the milk analysis over time on a per goat (Figure 5.7) basis and on a treatment averaged basis (Figure 5.8). Statistical analysis (t-tests) revealed no difference between the two Erythrina treatment groups with respect to the concentration of DHEs even though the concentration of DHEs in the E. poeppigiana treatment group was approximately twice that of theE. berteroana group. The Erythrina treatment groups were significantly different from the control groups in the concentration of DHE in most instances, but not in all due to variation in the data. There was no correlation between DHE concentration in the milk and fat content of the milk (correlation coefficient = -0.192) or the amount of milk produced (correlation coefficient = -0.186). There was no significant difference at the 95% confidence level (c.l.) with respect to milk production between the treatment groups and the control group until the later part of the study (Figure 5.9). There was a significant difference (95% c.l.) between all groups with respect to the amount of feed consumed (Figure 5.10). The diet containing £. poeppigiana was consumed n the greatest quantity, followed closely by the E. berteroana group. The control group ate significantly less foliage than either treatment group. The contribution of the concentrate was not included in the calculations of the total amount of foliage consumed for the control group. Figure 5.11 presents the data on the body weights of the goats throughout the feeding experiment. Although the goats in the Erythrina treatment groups began to lose weight when compared to the control after three months on the treatment diets, the weight difference was not statistically significant. 91

Table 5.1. Tabulated Mass Spectral Data as Percent of Base Peak of Biologically Hydrogenated Dihydroerythroidines (DHEI, DHEII and DHEIII), Catalytically Hydrogenated Dihdyroerythroidincs (DHEIV and DHEV), Tetrahydroerythroidine (THE) and B-Erythroidine (E).

Percent of Base Peak

m/z DHEI DHEII DHEIII DHEIY DHEV 275 20 42 46 16 37 260 58 91 94 1 10 245 38 31 20 7 16 244 44 76 66 23 81 230 100 93 95 3 2 217 5 49 59 100 49 216 19 74 74 8 21 204 17 20 6 2 2 200 15 20 24 7 4 191 6 62 31 <1 1 186 21 38 25 2 4 178 13 24 50 <1 2 172 19 100 100 24 100 158 16 56 50 5 34 146 15 20 17 2 7 132 15 54 39 9 40 130 14 33 20 9 9 77 28 45 50 4 6

THE m/z 2 l E m/z 2t 277 9 273 55 262 18 258 26 247 14 242 100 232 17 240 27 218 6 214 14 206 8 198 17 178 29 184 14 164 10 182 21 148 100 170 21 146 16 156 15 132 18 130 34 130 5 91 20 77 14 77 25 92

\bundanc«

14000

12000

10000 D ihydroerythroidine I from rumen

B000

6000

4000

2000 3 9 53

4/2 ~

120000

100000

Dihydroerythroidine 1 from milk BOOOO

60000

40000 77

42 65 275 20000 117 27

354 50 100 150 200 300250 350

Figure 5.3 Mass Spectra of Dihydroerythroidine I (DHEI) from the Milk and Rumen Samples 93

Vbundanc

6000-

5000 D ihydroerythroidine 11 from rumen

4000

3000

2000

1000

» k l B l 4/Z -> 20 40 60 80 100 120 140 160 IBP 200 220 240 260

260

12000 244 Dihydroerythroidine II from milk 216 10000 191 77 8000 158 132 91

6000 65 275

4000

2000

- > 20 40 60 80

Figure 5.4 Mass Spectra of Dihydroerythroidine II (DHEII) from the Milk and Rumen Samples 94

Tetrahydroerythroidine 5000

4000 from rumen

3000 178

2000 232

132 277 41 1000 164 204 91 193 29

60 80

Abundance

5000

Tetrahydroerythroidine 4000

from milk

3000 246

2000 178

262 1000 232

188 206 I 217 JHUUJ niM ...ii.iU.ijul MZ -> 20 40 60 100 120 140 160 180 200 220 240 260

Figure 5.5 Mass Spectra of Tetrahydroerythroidine (THE) from the Milk and Rumen Samples \bundanc 3500000 H

3000000

2500000

Dlhydroerythroidina IV 2000000

1500000

1000000 172 244 275 500000 - M 2 160 la 202

At ffiV h J* | ,-U‘Hi*!LL* vpl.li t t*tVt |*^ 1 L 1 2“I l*T)I tt/Z 20 40 60 SO 100 120 140 160 180 200 220 240 260 280

90000

80000

70000 Dihydroerythroidina I 60000 260 50000 244 40000

77 91 186 20000 42 216 275 65 132146 204 117 17 10000

40 60 80 100 120 140 160 180 200 220 240 260 280

Figure 5.6 Mass Spectra of the Principal Biological Dihydroerythroidine Isomer, DHEI, and the Prinicpal Synthetic Dihydroerythroidine Isomer, DHEIV. WEEK

Figure 5.7 Amount of Dihydroerythroidines (DHEs) Present in the Milk of Goats which Consumed E. poeppigiana and E. berteroana Foliage per Goat. Treatment goats only. Goat identification numbers shown in legend. All control goats where devoid of dihydroerythroidines. 97

80

60 o o

0 20 =L

4 6 6 10 12 14 WEEK

1200

« 1000 TJ (b) •n. *1 800 10 o \o* 600 M HI Q 400

OzL 200 0 6 8 10 12 14 WEEK

Figure 5.8 Average Amount of Dihydroerythroidines (DHEs) Present in the Milk of Goats which Consumed E. poeppigiana (•) and E. berteroana (A) and Foliage by Species. Key: (a) pg per 100 mL of milk; (b) total pg DHE produced per goat per day. Vertical bars indicate plus or minus one standard deviation. 4 Q M U 3 aD sb 2 K •4 1 (9 M 0 0 2 4 6 8 10 12 14 WEEK

Figure 5.9 The Amount of Milk Produced (kg) by Each Treatment Group, (x ) Control; (*) E. poeppigiana; (A) E. berteroana.. Vertical ban indicate plus or minus one standard deviation.

•« Q Ed 2 3 »» Z o C5

10 -

020 60 80 100 DAY

Figure 5.10 The Amount of Feed Consumed (kg) by Each Treatment Group. (— ) Control. ( ------) E.poeppigiana ; (...... ) E. berteroana.. 60

50

40

50

MONTH

Figure 5.11 Body Weight of Goats (kg) by Each Treatment Group. (x ) Control; (•) E. poeppigiana; (A) E. berteroana. Vertical bars indicate plus or minus one standard deviation. 100

5.4 Discussion The mass spectra of the dihydroerythroidines and tetrahydroerythroidine detected in the rumen samples exactly match the dihydro- and tetrahydroerythroidines detected in the milk samples of the treatment groups of each species. These compounds were not detected in the milk of the control group. The B-erythroidine present in the foliage of the E. poeppigiana and E. berteroana fed animals is evidently hydrogenated in the rumen of the dairy goats and transported to the milk of these animals. In the rumen environment at a pH of 5.5 to 7.0, the absence of oxygen and the presence of microorganisms promotes reductive reactions (Dickinson et al. 1976). Hydrogenation is one of many reductive chemical reactions that takes place in the rumen. Most reports of hydrogenation focus on the rapid hydrogenation of unsaturated fatty acids by rumen bacteria for the construction of bacterial cells (Ward, 1964). During this type of hydrogenation, double bond migrations are known to occur (Wilde and Dawson, 1966). The mass spectral data of the biologically formed DHEs and the synthetic DHEs indicate that the two hydrogenation processes produce different DHE isomers. The biologically hydrogenated DHEs have major molecular ion peaks as well as major peaks corresponding to (M-CH3)+, and (M-OCH3)+. The base peak of DHEI (m/z - 230) corresponds to a loss of an additional loss of CH2 as well as a loss of OCH3. These results are consistent with the fragmentation pattern of dienoid type of Erythrina alkaloids proposed by Boar and Widdowson (1970). If the conjugated diene is intact after hydrogenation, then the double bond present in the lactone ring of B-erythroidine (Figure 1) is hydrogenated in the rumen which may result in the formation of two diastereomers. Inspection of the mass spectra of the major synthetic isomer, DHEIV, reveals that the molecular ion and (M-CH3)+ peaks are weak when compared to the mass spectra of the 101 biological DHEs and that the base peak of m/z =217 represents a loss of 58 mass units. This loss is typical of a alkenoid type Erythrina alkaloids (Boar and Widdowson, 1970). Hydrogenation of the double bond at the 2-3 position, probably the most facile position to synthetically hydrogenate, and subsequent double bond migration would result in a double bond at the 1-6 position. A retro-Diels Alder reaction would result in a loss of 58 amu and the second mechanism of fragmentation proposed by Boar and Widdowson would also result in a loss of 58 units. These results support the hypothesis that although catalytic and bacterial hydrogenations are similar in many instances, conjugated double bonds are resistant to microbial hydrogenation (Garton, 1964). It should be noted that the pharmacological studies done by Unna (Unna et al. 1944) and others utilized DHE formed upon catalytic hydrogenation of 0-crythroidine. Therefore it is likely that the synthetic dihydroerythroidine investigated in the pharmacological and toxicological drug studies of the 1940s also contained a double bond at the 1-6 position and the 12-13 position. The wide variability in the concentration of dihydroerythroidines in the milk of the treated animals was expected. Neither the experimental animals nor the foliage used as feed in the feeding study was genetically homogeneous. This variability may be environmental, genetic or biochemical in nature. The alkaloid content of the leaves may vary with environmental conditions thus causing variability in the amount of 0- erythroidine the animals were consuming. It is interesting that both Erythrina species varied similarly. The in concert variation of the DHE content of the milk in both treatment groups may be indicative of an environmental influence on the goats or on the trees which is superimposed on the other variables. The bacteria in the rumen of each animal may be slightly different resulting in different rates of hydrogenation of the 0- erythroidine and/or competing biochemical reactions in the rumen. Alternatively, each animal may metabolize the dihydroerythroidine produced differently or excrete them at 102 different rates. Other contributions to the variability can be hypothesized. These results emphasize the importance of utilizing genetically homogeneous material in scientific investigations, particularly in feeding studies where biological variation may be great. Only a small fraction of the B-erythroidine consumed may reach the milk in an unaltered or hydrogenated form. Studies described in Chapter 4 with Erythrina clones have shown thatErythrina foliage is approximately 0.1% by weight of B-erythroidine (Payne and Foley, 1992). If the goats eat on the average 2.2 kg of Erythrina foliage per day, then approximately 440 mg of B-erythroidine are consumed per day per goat assuming a dry weight of 20%. Estimates of the total amount of dihydroerythroidines in the milk are between 1 fj.g and 1 mg per day per goat. So the DHE found in the milk represents less than 0.2% of the total amount consumed if the above assumptions are used. It is possible that the foliage used in the feeding study was exceptionally low in alkaloid content, however there may be other explanations for the possibly low amount of DHEs detected in the milk of the experimental animals, such as environmental variation and differences in bacterial composition in the rumen of the experimental animals as previously discussed. The results of the rumen incubations suggest that almost all the B-erythroidine found in the foliage is metabolized. Very little parent compound is detected in the rumen solutions after 48 hours, so it is unlikel) hat the pareant compound B-erythroidine is eliminated intact through the feces. Since the dihydro- compounds are probably transported via the circulatory system at a pH at which they are soluble and because they have no particular affinity for lipids, it is possible that DHE are eliminated through the urine as well as the milk. However, experiments with frog urine indicated that no pharmacologically active compounds were eliminated via the urine in frogs (Unna et al., 1944). The parent compound, B-erythroidine, is known to be soluble in blood serum since intravenous injection was used in clinical administration of the drug. There have 103 been no studies done on the degradative metabolism of B-erythroidine in humans or animals, however clinical use of the drug suggests that B-erythroidine is rapidly degraded or eliminated in human subjects (Unna, 1939). There was no correlation between the milk fat content and the amount of DUE detected suggesting that the DHEs are not preferentially sequestered in the fat of the animal. No Erythrina alkaloids were detected in the hexane fraction during extraction. In addition, most of the treatment animals maintained or lost weight, so fat deposition is an unlikely method of elimination. The B-erythroidine may be degraded by rumen bacteria into compounds not present in the alkaloid fraction analyzed by GC/MS. Although several types of heterocyclic alkaloids, for example pyrrolizidine and indole alkaloids (Carlson and Breeze, 1984), are not substantially degraded by ring fission in the rumen, there are examples of heterocyclic fissions of other types of compounds such as coumarins where the lactone moiety is attacked (Gutenmann et al. 1972) or the example of diquat which is converted to a methyl urea derivative and CO2 (Williams, 1972). Thus it is possible that much of the drug is degraded and excreted before it reaches the milk. It is possible that the hydrogenation of B-erythroidine or some other bacterial reactions is a detoxification mechanism. Even though adult goats do not develope toxic symptoms when fed Erythrina, several young animals that were given milk from goats on a diet containing Erythrina developed toxic symptoms (NFT project, CATIE, personal communication). It was unclear whether these symptoms resulted from a disease or a response to a toxin in the milk. It could be that the kids had not yet developed the bacterial population necessary to detoxify the possible toxins and therefore were susceptible to B-erythroidine or its metabolites. It is difficult to assess the human health risk associated with the results presented in this paper. Dihydroerythroidine is one of the most potent forms of B-erythroidine with 104 an effective dose for complete paralysis in mice of 80 |ig/kg (Robbins and Lundy, 1947a and 1947b). This amount was exceeded in many milk samples on a total amount of DHEs produced per goat per day basis (Figure 5.8). However, mass spectral data presented here suggest that the DHEs formed in the rumen are different from the synthetic DHE tested in toxicological studies in the 1940s and 1950s (Boekelheide and Prelog, 1955) and therefore, the pharmacological data accumulated concerning the toxicity of DHE may not be pertinent to the isomers detected in the milk and rumen of the dairy goats in this study. From a review of the synthetic studies done on synthetically hydrogenated erythroidine derivatives, Snieckus concluded that in the case of B-erythroidine, the pharmacological properties of the drug were not dependent on the double bond configuration of the diene in the C and D rings (Snieckus, 1973). Several different optical isomers showed similar activities in biological assays. Therefore the dihydroerythroidines formed biologically may indeed possess biological activity. Furthermore, long term chronic exposure to these alkaloids, as well as other types alkaloids (Peterson, 1985), which may occur when a family consumes the contaminated milk regularly, is unknown.

5.5 Conclusion B-Ery throidine, a paralyzing drug which is present in the foliage of Erythrina berteroana. and E. poeppigiana tree species, is rapidly hydrogenated in the rumen of dairy goats. Ery throidine, dihydroerythroidines and tetrahydroerythroidine were detected in the milk of the two treatment groups fed E. poeppigiana and E. berteroana foliage and were detected in rumen incubations which included dried plant material. The DHEs detected in the milk were not the same isomers formed upon catalytic hydrogenation of B-erythroidine and therefore the human health hazard associated with the consumption of milk which contains biologically hydrogenated DHEs is uncertain. CHAPTER 6 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

105 106

6.1 Summary After introducing the concept of agroforestry and describing the integration of animals in agricultural systems, the objectives of this research were presented. The objectives encompassed two main areas: the clonal evaluation of Erythrina clones and the identification of the metabolic fate of the major Erythrina alkaloids identified by gas chromatography and mass spectrometry analysis. Chapter 2 reviewed the scientific literature that pertained to Erythrina alkaloids. It was known that all Erythrina alkaloids possessed biological activity, however the exact mode of action has never been elucidated. Although the structure and stereochemistry of Erythrina alkaloids were determined, the exact structural requirements for biological activity were never revealed. Pharmacological studies focused on the duration of the drugs' effect and possible side effects. Chapter 3 presented the basic experimental methodology for the plant collection, preparation and extraction. Additionally, liquid chromatographic techniques that could be used for Erythrina alkaloid analysis were investigated and discussed. These techniques included reversed phase liquid chromatography and LC/MS analysis. Chapter 4 detailed the GC/MS studies of the clone replicates. The potential importance of the alkaloid profiles in taxonomic classification was discussed as well as the implications of the large clonal variation in B-erythroidine content Clone 2652 was identified as the highest alkaloid containing clone, whereas, clone 2750 was very low in alkaloid content. There were low alkaloid content varieties of all three species evaluated. The feeding study was described and discussed in Chapter 5. The milk of goats from three treatment groups were studied. Dihydroerythroidines and tetrahydroerythroidines were discovered in the milk of treated animals. Rumen incubations revealed that B-erythroidine was hydrogenated in the rumen. A discussion of the structural differences between synthetically hydrogenated and biologically 107 hydrogenated dihydroerythroidines was included.

6.2 Conclusions Statistical analysis of total nitrogen content and dry weight of the clones revealed that there are differences between clones of the same species with respect to these two parameters. As a species, E. poeppigiana in general was higher than either E. costarricensis and E. berteroana in nitrogen content, whereas E. berteroana was higher albeit more variable in dry matter content than E. poeppigiana. These differences suggest that Erythrina clones may vary in certain parameter due to genetic differences. Studies presented here reveal that reversed phase HPLC can be useful in the analysis of Erythrina alkaloids. However, if trace amounts of alkaloids are present and need to be identified, gas chromatography /mass spectrometry is more sensitive. 8 - Ery throidine eluted within 5 minutes in the reversed phase HPLC system developed. Peak asymmetry was minimized by using a polymeric column and a mobile phase of relatively high pH. If the compounds of interest are known and standards are available, then reversed phase HPLC can be very useful. In an effort to develop a technique in which the alkaloids could be identified without the use of standards, LC/MS was investigated. An Extrel mass spectrometer, Waters HPLC system and a Thermabeam interface was utilized. Mass spectra of oxo-B- erythroidine and B-erythroidine were obtained but only when large injection volumes and high concentrations of alkaloids were used. Optimal HPLC conditions could not be used because of the drop in efficiency at the Thermabeam interface. Buffers could not be used in this system because of the high background which would have increased with the use of buffer. Highly aqueous mobile phases developed during the optimization stage of this research could not be used because of condensation at the interface and the high pressures generated at relatively low flow rates. Nevertheless, if samples are 108 highly concentrated this method can be used to gri'e preliminary identification via mass spectra of the compounds of interest. This is the first time LC/MS has been used to investigate Erythrina alkaloids. GC/MS using a 5% phenyl/95% methyl column is a valuable technique for the analysis of Erythrina alkaloids. Clonal evaluation of E. poeppigiana,. berteroanaE and E. costarricensis revealed that B-erythroidine, a drug previously used in surgery and electroshock treatments, was the major alkaloid in the leaves of these species. There was little statistical variation in the parameters measured between individual trees of the same genetic clone. There was wide variation in the concentration of B-erythroidine between Erythrina clones of the same species and between different Erythrina species, however there was little difference in the types of Erythrina alkaloids detected between clones or between species. Low alkaloid content clones of all three species were identified. Alkaloid content and nitrogen content were not correlated, which suggests that these important parameters are not genetically linked. Low alkaloid containing clones may be genetically engineered or bred to contain even lower amounts of alkaloids without affecting or even improving the total protein content. For the present, low alkaloids clones should be developed over high alkaloid content clones for alternative feeds. B-Ery throidine is rapidly hydrogenated in the rumen of dairy goats. Ery throidine, dihydroerythroidines and tetrahydroerythroidine were detected in the milk of the two treatment groups fed E. poeppigiana and E. berteroana foliage and were detected in rumen incubations which included dried plant material. The DHEs detected in the milk were not the same isomers formed upon catalytic hydrogenation of B-erythroidine. Since pharmacological studies have only included synthetically hydrogenated dihydroerythroidines, the human health hazard associated with the consumption of milk which contains biologically hydrogenated Ul lEs is unclear. 109

5.6 Recommendations Further studies should be performed in order to assess the toxicity of biologically formed dihydroerythroidines. As an example, goats could be dosed with a high level of B-erythroidine using a high alkaloid content clone identified in this study or by introducing seeds of high B-erythroidine content into the rumen of a fistulated goat. The milk collected from these animals could be freeze-dried and incorporated into the feed of mice in a controlled feeding study to determine toxic effects in mice. The results of this study could then be compared with the toxicity data accumulated using synthetically derived dihydroerythroidines. Until the toxicity of the metabolites of B-erythroidine can be assessed, every effort should be made to utilize low alkaloid containing clones in feeding studies or in projects whereErythrina foliage will be used as a feed for animals. REFERENCES

110 Ill

Abdullah, M.I.; Barakat, I.E.; Games, D.E.; Ludgate, P.; Mavraganis, V.G.; Ratnayake, V. U.; Jackson, A. H. 1979. Ann. Missouri Bot. Gard. 66:533- 540. Abdullah, M.I.; Chawla, A.S.; Jackson, A.H. 1982. J. Chem. Soc., Chem. Commun. 1982:904-905. Achenbach, H. 1977. ia Wagner, H.; WolfT, P. eds. New Natural Products and Plant Drugs with Pharmacological. Biological or Therapeutical, Activity Springer-Verlag, New York; pp 115-116. Aguilar, M.I.; Giral F,; Espejo, O. 1981. Phytochem. 20(8):2061-2062. Ahmed-Schofield, R.; Mariano, P.S. 1985. J. Org. Chem. 50:5667-5677. Ahmed-Schofield, R.; Mariano, P.S. 1987. J. Org. Chem. 52:1478-1482. AUcroft, R.; Camaghan, R.B.A. 1962. Vet. Rec. 74(31):863-864. Allcroft, R.; Roberts, B. A. 1968. Vet. Rec. 1968:116-118, Alston, R. E.; Rosier, H.; Naifeh, K.; Mabry, T. J. 1965. Proc. National Acad. Set. 54:1458-1465. Alston, R.E. 1962. Science 137:1048-1050. Altamirano A.; Dominguez 1888. Gaceta Medica de Mexico 23:369. Arnold, G. W.; Hill, J. L. 1972. ia Harbome, J.B. ed„ Phytochemical Ecology Academic Press, New York p. 72. Association of Official Agricultural ChemisL 1960. Official Methods of Analysis, 9th ed., Association of Official Agricultural ChemistsWashington, D.C.; p 832. Ayensu, E.S. 1977. Lloydia 40:436. Balsevich, J.; Hogge, L.R.; Berry, A.J.; Games, D.E.; Mylchreest, I.C. 1988. J. Natural Prod. 51(6): 1173-1177. Bansal, R.D.; Seaney, R.R. 1970. Indian J. Anim. Sci. 40(4):413. Barakat, I.; Jackson, A.H.; Abdullah, M.I. 1977. Lloydia 40(5):471-475. Baietta-Kuipers, T. 1982. Allertonia 3(1 ):53. Banos, G.S.G.; Matos, F.J.A.; Vieir., J.E.V.; Sousa, M.P.; Medeiros, M.C. 1970. J. Pharm. Pharmac. 22:116-122. Barton, D.H.R.; Gunstilaka, A.A.L.; Letcher, R.M.; Lobo, A.M.F.T.; Widdowson, D.A. 1973. /. Chem. Soc., Perkin Trans. 1 1973:874-880. 112

Barton, D.H.R.; Boar, R.B.; Widdowson, D.A. 1970a. J. Chem. Soc. (Q 1970:1208-1213. Barton, D.H.R.; Boar, R.B.; Widdowson, D.A. 1970b. J. Chem. Soc. (C) 1970:1213-1218. Barton, D.H.R.; Brae ho, R.D.; Potter, C.J.; Widdowson, D.A. 1974a. /. Chem. Soc. Perkin 7:2278-2283. Barton, D.H.R.; James, R.; Kirby, G.W.; Turner, D.W.; Widdowson, D.A. 1966. Chem. Comm. 10:294-295. Barton, D.H.R.; James, R.; Kirby, G.W.; Turner, D.W.; Widdowson, D.A. 1968. J. Chem. Soc. (C) 1968:1529-1537. Barton, D.H.R.; Jenkins, P.N.; Letcher, R.; Widdowson, D.A.; Hough, E.; Rogers, D. 1970c. Chem. Commun. 1970:391-392. Barton, D.H.R.; Potter, C.H.; Widdowson, D.A. 1974. J. Chem. Soc., Perkin Trans 1 1974:346-348. Belleau, B. 1953. J. Am. Chem. Soc. 75:5765-5766 Benavides, J.E. 1983. Utilizacidn de follaje de forrajes de origen arbdreo en la alimentacidn de mmiantes men ores, Curso Corto Agroforestal, 11-21 January, 1983, CATIE, Turrialba, Costa Rica. Benavides, J.E. 1986. Utilizacidn de follaje de por 6 (Erythrina poeppigiana) para alimentar cabras bajo condiciones de tropico humedo, 2do Congreso de la Asociacidn Mexicana de Zootecnistas y Tdcnicos en Caprinocultura, March 19- 22, 1986, Mazatl^n, Mexico. Berger, F. M.; Schwartz, R.P. 1948. J. Pharmacol. Exper. Therap. 93:362-367. Berry, A.J.; Games, D.E.; Perkins, J.R. 1986. J. Chromatogr. 363:147-158. Bhakuni, D.S.; Jain, S. 1980. Tetrahedron 36:3107-3114. Bhakuni, D.S.; Khanna, N.M. 1959, J. Sci. Industr. Res. 18B;494-495. Bhakuni, D.S.; Uprety, H.; Widdowson, D. A. 1976. Phytochem. 15:739-741. Bhat, S.V.; Domauer, H.; De Souza, N.J. 1980. J. Natural Prod. 43(5):588-591. Bhattacharya, S.K.; Debnath, P.K.; Sanyal, A.K.; Ghosal, S. 1971. J. Res. Ind. Med. 6:235-241. Bhutani, K.K. et. al., 1984. Phytochem., 23(8):1765-1769. Bick, I.R.C.; Doebel, K.; Taylor, W.I.; Todd, A.R. 1953a. J. Chem. Soc. 1953:692-695. 113

Bick. I.R.C.; Taylor, W.I.; Todd, A.R. 1953b. J. Chem. Soc. 1953:695-700. Blua, M. J.; Hanscom, Z. 1986. J. Chem. Ecol. 12(6): 1449-1458. Boar, R.B.; Widdowson, D.H., 1970. J. Chem. Soc. (B) 1970:1590. Bochefontaine, A.; Rey, P. 1881. Gaz. Med. Pahs (6) 3:196. Boekelheide, V. 1955. Record of Chem. Prog. 16{4):226. Boekelheide, V. 1960. ia Manske, cd. The Alkaloids. Vol. VII., Academic Press, New York, Chapter 11. Boekelheide, V.; Agnello, E. 1951. J. Am. Chem. Soc. 73:2286-2289. Boekelheide, V.; Anderson, A.E.; Sauvage, G.L. 1953. J. Am. Chem Soc. 75:2558-2563. Boekelheide, V.; Grundon, M.R. 1953. J. Am. Chem. Soc. 75:2563-2566. Boekelheide, V.; Morrison, G. 1958. J. Am. Chem. Soc. 80:3905-3908. Boekelheide, V.; Prelog, V. 1955. Prog, in Organic Chem. 3:242. Boekelheide, V.; Weinstock, J.; Grundon, M.F.; Sauvage, G.L.; Agnello, E.J. 1953. J. Am. Chem. Soc. 75:2550-2558, Boekelheide, V.; Wenzinger, G.R. 1964. J. Org. Chem. 29:1307-1310. Bremner, J.B. 1990. Studies of Natural Products Chemistry. Vol. 6 Part D 467- 502. Browne, C.A.; Sim, F.R.; Rae, I.D. Keeler, R.F. 1984. J. Chromatogr. 336:211-220. Buckmaster, G.W. et. al 1976. J. Animal Sci. 43(2):464. Bull, L.B.; Dick, A.T. 1959. J. Pathol. Bact. 78:483-502. Carlson, J.R.; Breeze, R.G. 1984. J. Animal Sci. 58(4): 1040-1049. Carmack, M.; McKusick, B.C.; Prelog, V. 1951. Helv. Chim. Acta 34:1601- 1609. Chan, W.W.-C.; Maitland, P. 1966. J. Chem. Soc. (C) 1966:753-757. Chawla, A.S.; Chunchatprasert, S.; Jackson, A.H. 1983. Organic Magnetic Resonance 21(1):39-41. Chawla, A.S.; Gupta, M.P.; Jackson, A.H, 1987. J. Natural Prod. 50(6): 1146- 1148. 114

Chawla, A.S.; Jackson, A.H. 1984. Natural Product Reports 1:371-373. Chawla, A.S.; Jackson, A.H. 1986. Natural Product Reports 3:555-564. Chawla, A.S.; Jackson, A.H. 1989. Natural Product Reports 6:55-66. Chawla, A.S.; Jackson, A.H.; Ludgate, P. 1982. J. Chem. Soc. Perkin Trans. 1:2903-2907. Chawla, A.S.; Krishnana, T.R.; Jackson, A.H.; Scalabrin, D.A. 1988. Planta Medica 54(6):526-528. Chawla, A.S.; Redha, F.M.J.; Jackson, A.H. 1985. Phytochem. 24(8): 1821- 1823. Cheeke, P.R.; Shull, L.R. 1985. Natural Toxicants in Feeds and Poisonous Plants. Avi Publications. Chopra, R.N. et al. 1956. Glossary of Indian Medicinal Plants. Council of Sci. and Ind. Res., New Dehli. Chunchatprasert, S. 1982. Chemical and Spectroscopic Studies of Erythrina Alkaloids, Ph.D. Thesis, University of Wales, Cardiff, Great Britain. Chung, C-T.A.; Staba, B.J. 1984. J. Chromatogr. 295:276-281. Clezy, P.S.; Gellert, E., Lau, D.Y.K.; Nichol, A.W. 1966. Aust. J. Chem. 19:135-142. Cook, J.W. 1957. Ag. Food Chem. 5(11):859. Cordell, G.A. 1981. Introduction to Alkaloids: a biogenetic approach. John Wiley, New York, 450-462. Cottington, F. 1941. Am. J. Psychiatry 98:397-400. Crabb, T.A. 1975. in Mooney E.F, ed., Annual Reports on NMR Spectroscopy. Vol. 6A, Academic Press, New York, pp 249-387. Craig, A.M.; Blythe, L.L.; Lassen, E.D. 1985. in Seawright, A.A. et al. eds. Plant Toxicology, Animal Research Institute, Yeerongpilly, Australia. 200- 208. Craig, L.E. 1955. The Alkaloids. Vol.5: Pharmacology, Academic Press, New York. Crane, T.D. 1973. Vet. Bull. 43:231-249 Crane, T.D. 1973. Vet. Bull. 43:165-177 Cramner, M.F.; Mabry, T.J. 1966. Phytochem. 5:1133-1138. 115

Cranmcr, M.F.; Turner, B.L. 1967. Evolution 21:508-517. Critchett, C.D.; Dharmaratne, H.R.W.; Sotheeswaran, S.; Galal, A.M.; Schiff Jr., P.L.; Bick, R.C. 1989. Aust. J. Chem. 42:2043-2046. Daccarett, M.; Blydcnstcin, J, 1968. Turrialba 18(4):405-408. Dagne, E.; Stcglich, W. 1983. Tetrahedron Lett. 24(4):5067-5070. Dagnc, E.; Steglich, W. 1984. Phytochem. 23(2):449-451. Dalton, D.R. 1979, The Alkaloids, Marcel Dekker, Inc., New York; pp 365-370. Dean, J.A. ed. Lange's Handbook of Chemistry. 13th ed., McGraw Hill Book Co., New York; p 7-698. De Silva, S.O.; Snieckus, V.A. 1978 . in Grundon, M.F. ed. The Alkaloids. Burlington House Chemical Society, London; Vol. 8 , p 144. De Leon, R.M, 1984. Evaluacidn Nutricional de la Harina del Follaje de Arbol de Pito (Erythrina berteroana Urban, Symb), en Animates de Laboratorio, Master's Thesis, Universidad de San Carlos, Guatemala. Deulofeu, V. 1952. Chem. Ber. 85:620-623 Deulofeu, V.; Hug, E.; Mazzocco, P. 1939. J. Chem. Soc. 1939:1841-1842. Deulofeu, V; Labriola, R; Hug, E., Fondovila, M.; Kauffmann, A, 1947. J. Org. Chem. 12:486-489. Dhar, M.L.; Dhar, M.M.; Dhawan, B.N.; Mehrotra, B.N.; Ray, C. 1968. Indian J. Exp. Biol. (6):232-247. Dickinson, J.O.; Cooke, M.P.; King, R.R.; Mohamed, P.A. 1976. J. Amer. Vet. Med. Assoc. 169(11):1192-1196. Dickinson, J.O.; King, R.R. 1978. in Keeler, R.F. et al. eds. Effects of Poisonous Plants on Livestockx Academic Press, New York; pp 201-208. Dietz, E.; Folkers, K. 1946. J. Am. Pharm. Assoc. 35:48-49. Dyke, S.F.; Quessy, S.N. 1981. The Alkaloids, Academic Press, New York; Vol. XVIII, p 1. El Olemy, M.M.; Ali, A.A.; El Mottaleb, M.A. 1978. Lloydia 41(4):342-347. Espinoza Bran, J.E. 1984. Characterizacidn Nutritiva de la Frace id n Nitrogenada del Forraje de Madera Negro Gliricidia sepium y Pord Erythrina poeppigiana, Master's Thesis, CAT1E, Turrialba, Costa Rica. Fernandes, J.; Van de Kamer, J.H.; Weijers, H.A. 1962. J. Clinical Investigation. 41(3):488-494. 116

Folkers, K.; Koniuszy, F. 1939a. J. Am. Chem. Soc. 61:1232-1235. Folkers, K.; Koniuszy, F. 1939b. J. Am. Chem. Soc. 61:3053-3055. Folkers, K.; Koniuszy, F, 1940a. J. Am. Chem. Soc. 62:436-441. Folkers, K.; Koniuszy, F. 1940b. J. Am. Chem. Soc. 62:1673-1676. Folkers, K.; Koniuszy, F. 1940c. J. Am. Chem. Soc. 62:1677-1683. Folkers, K.; Koniuszy, F. 1946, Chem. Abstr. 40:1283. Folkers, K.; Koniuszy, F. 1950. J. Chem. Soc. 72:1832-1833. Folkers, K.; Koniuszy, F.; Shavel jr., J. 1942. J. Am. Chem. Soc. 64:2146- 2151. Folkers, K.; Koniuszy, F.; Shavel jr., J. 1944. J. Am. Chem. Soc. 66:1083- 1087. Folkers, K.; Major, R.T. 1937. J. Am. Chem. Soc. 59:1580-1581. Folkers, K.; Major, R.T. 1945a. Chem. Abstr. 37:5408. Folkers, K,; Major, R.T. 1945b. Chem. Abstr. 39:3549. Folkers, K.; Shavel jr., J. 1942. J. Am. Chem. Soc. 64:1892-1896. Folkers, K.; Shavel jr., J. 1946. Chem. Abstr. 40:1284. Folkers, K.; Shavel, J. jr.; Koniuszy, F. 1941. J. Am. Chem. Soc. 63:1544- 1549. Folkers, K.; Shavel, J. jr.; Koniuszy, F. 1944. J. Am. Chem. Soc. 66:1083- 1087. Folkers, K.; Unna, K, 1938. J. Pharm. Sci. 27:693-699. Folkers, K.; Unna, K. 1939. J. Pharm. Sci. 28:1019-1028. Franck, B.; Teetz, V. 1971. Angew. Chem. internal. Edit. 10(6):411-412. Furton, K.G.; Rein, J. 1990. Anal. Chim. Acta 236:99-114. Games, D.E.; Jackson, A.H.; Khan, N.A.; Millington, D.S. 1974. Lloydia 37(4):581-588. Games, D.E.; Jackson, A.H.; Millington, D.S. 1973. Tetrahedron Lett. 32:3036- 3066. 117

Garton, G.A. 1964. in R.M.C. Dawson and D.N. Rhodes, eds. Metabolism and Physiological Significance of Lipids, Wiley, New York. Gentile, R.A.; Labriola, R. 1942. J, Org. Chem. 7:136-139. Gervay, J.E.; McCapra, F,; Money, T.; Sharma, G.M. 1966. Chem Comm. 1966:142-143. Ghosal, S.; Baneijee, P.K. 1969. Aust. J. Chem. 22:2029. Ghosal, S.; Baneijee, P.K.; Baneijee, S.K. 1970a. Phytochem. 9:429-433. Ghosal, S.; Chakraborti, A.; Srivastava, R.S. 1972. Phytochem. 11:2101-2103. Ghosal, S.; Dutta S.K. 1971. Phytochem. 10:195. Ghosal, S.; Dutta, S.K.; Bhattacharya, S.K. 1972. J. Pharm. Sci. 61(8): 1274- 1277. Ghosal, S.; Ghosh, D.K.; Dutta, S.K. 1970b. Phytochem. 9:2397-2398. Ghosal, S.; Majumdar, S.K.; Chakraborti, A. 1971. Aust. J. Chem. 24:2733- 2735. Ghosal, S.; Srivastava, R.S. 1974. Phytochem. 13:2603. Ghosh, D.K.; Majumdar, D.N. 1972. Current Sci. 41(15):578. Gibson, J.B. 1957. British Medical Journal 2:159. Glasby, J.S. 1975. Encyclopedia of the Alkaloids, Plenum Press, New York; pp 540-549. Godfrey, J.C.; Tarbell, D.S.; Boekelheide, V. 1955. J. Am. Chem. Soc. 77:3342- 3348. Goeger, D.E.; Cheeke, P.R.; Schmitx, J.A.; Buhler, D.R. 1972. Am. J. Vet. Res. 43(9): 1631. Grethc, G. ed. 1981. Isoquinolines: The chemistry of heterocyclic compounds* Wiley, New York.; 38(l):325-329. Griffith, H.R. 1951. Ann. New York Acad. Sci. 54:493-497. Grundon, M.F.; Boekelheide, V. 1953. J. Am. Chem. Soc. 75:2537-2541. Grundon, M.F.; Savage, G.L.; Boekelheide, V. 1953. J. Am. Chem. Soc. 75:2541-2543. Gunn, U.R.; Bamers, D.E. 1977. Lloydia 40(5):454. 118

Gutenmann, W.H.; Serum; J.W.; Lisk, D.J, 1972. J. Agr. Food Chem. 20(5):991-993. Hanson, A.W. 1963. Proc. Chem. Soc. 1963:52. Harbome, J.B. ed. 1972. Phytochemical Ecology, Academic Press, New York. Hargreaves, R.T.; Johnson, R.D.; Millington, D.S.; Mondal, M.H.; Beavers, W.; Becker, L.; Young, C.; Rinehart Jr., K.L. 1974. Lloydia 37(4):569-580. Haruna, M.; Ito, K. 1976. J. Chem. Soc. Chem. Comm. 1976:345-346. Harvey, A. M.; Masland, R. L. 1941. . Pharmacol. Exper. Therap 73:304-311. Hill, R.K. 1967. in Manske, R.H.F., ed., The Alkaloids, Academic Press, New York; Volume IX. Hill, R.K.; Schearer, W.R. 1962. J. Org. Chem. 27:921-924. Hungate, R.E. 1966. The Rumen and Its Microbs, Academic Press, New York. Impellizzeri, G.; Piattelli, M. 1972. Phytochem. 11:2499-2502. Ito, K.; Furukawa, H.; Hanura, M. 1974a. Chem. Abstr. 80:68387p. Ito, K.; Furukawa, H.; Haruna, M. 1974b. Chem. Abstr. 80:48212z. Ito, K.; Furukawa, H.; Haruna, M.; ho, M. 1974. Chem. Abstr. 80:68391k. Ito, K.; Furukawa, H.; Haruna, M.; Lu, S-T. 1974. Chem. Abstr. 80:68390j. Ito, K.; Furukawa, H.; Tanaka, H. 1970. Chem. Comm. 1970:1076-1077. Ito, K.; Furukawa, H.; Tanaka, H. 1971. Chem. Pharm. Bull. 19(7): 1509-1511. Ito, K.; Furukawa, H.; Tanaka, H. 1973. Yakugaku Zasshi 93(8): 1211-1214. Ito, K.; Furukawa, H.; Tanaka, H. 1973. Chem. Abstr. 79:146715q. Ito, K.; Furukawa, H.; Tanaka, H. 1973. Chem. Abstr. 79:146713n. Ito, K.; Furukawa, H.; Tanaka, H,; Runshu, R. 1973, Chem. Abstr. 79:146714P. Ito, K.; Haruna, M.: Furukawa, H. 1975a. Chem. Abstr. 82:167515b. Ito, K.; Haruna, M.; Furukawa, H. 1975b. J. Chem. Soc., Chem. Commun. 1975:681. Ito, K.; Haruna, M.; Jinno, Y.; Furukawa, H. 1976. Chem. Pharm. Bull. 24:52- 55. 119

Ito, K.; Tanaka, H. 1974, Chem. Pharm. Bull. 22:2108. I vie, G.W. 1974. J. Agri. Food Chem 22(4):632. Jackson, A.H. 1985. in Philipson, J.D. et al. cds.The Chemistry and Biology of Isoquinoline Alkaloids, Springer Verlag, New York. Jackson, A.H.; Chawla, A.S. 1982. Allertonia 3(1):39. Jackson, A.H.; Ludgate, P.; Mavraganis, V.; Redha, F. 1982. Allertonia 3(1):47. Jain, S.: Madhusudana, K.P.; Bhakuni, D.S. 1987. Indian J. Chem. 26B:308- 311. Johnson, A. E. 1976. Am. J. Vet. Res. 37(1): 107. Johnson, A.E.; Molyneux, R.J. 1985. in Seawright, A.A. et al. eds. Plant Toxicology. Animal Research Institute, Yeerongpilly, Australia. 209-218. Jones, R.J. 1979. World Animal Review 31:13-23. Jones, R.J.; Megarrity, R.G. 1983. Aust. J. Agric. Res.. 34:781-790. Ju-ichi, M.; Ando, Y.; Satoh, A.; Kunitomo.J.; Shingu, T.; Furukawa, H. 1978. Chem. Pharm. Bull. 26(2):563-570. Ju-ichi, M.; Ando, Y.; Yoshida, Y.; Kunitomo, J.; Shingu, T.; Furukawa, H. 1977. Chem. Pharm. Bull. 25(3):533-534. Ju-ichi, M.; Fujitani, Y. 1981. Heterocycles 16(4):555-558. Ju-ichi, M; Fujitani, Y.; Furukawa, H. 1985. Chem. Abstr. 102:21 195. Kametani, T.; Fukumoto, K. 1972. Synthesis 1972:657-674. Keeler, R.F. 1975. Lloydia 38(1):56. Keeler, R.F. 1973. Teratology 7:23-30. Keeler, R.F.; Binns, W. 1971. Phytochem. 10:1765-1769. Kessler, C.D. J. 1981. The International Tree Crops Journal (1981):245-272. Kharat, S.T. 1980. Ind. J. Animal Sci. 50(8):638 Kilgore, W.W.; Croshy, D.G.; Craigmill, A.L.; Poppen, N.K. 1981. California Agriculture 35:6. King, R.R.; McQueen, R.E. 1981. 29:1101-1103. Kinghom, A.D.; Smolenski, S.J. 1978. Advances in Legume Systematics, Royal Botanic Gardens, England; Part 2, pp 585-598. 120

Kingsbury, J. M. 1980. in Doull, J.; Klassen, C.D.; Amdur, M.O. eds. Toxicology, Macmillan Pub, Co., Inc., New York; pp 578-590, Kitahara, T.; Matsui, M. 1974. Agr. Biol. Chem. 38(1): 171-174. Koniuszy, F.; Folkers, K. 1950. J. Am. Chem. Soc. 72:5579-5581. Koniuszy, F.; Wiley, P.F.; Folkers, K. 1949. J. Am. Chem. Soc. 71:875-878. Krukoff, B.A. 1979. Ann, Missouri Bot. Garden 66:422-445. Krukoff, B.A.; Bameby, R.C. 1974. Lloydia 37(3):332-459. Kupchan, S.M.; Kim, C-K.; Lynn, J.T. 1976. J. Chem. Soc.t Chem. Commun. 1976:86. Labriola, R.A.; Deulofeu, V.; Berinzaghi, B. 1951. J. Org. Chem. 16:90-93. Lapiere, C.; Coppee, G. 1948. Experientia 4(10):387-388. Lapiere, C.; Robinson, R. 1951. Chem. Ind. 1951:650-651. Lcboeuf, M. et al. 1982. J. Nat. Products 45(5):617. Leete, E.; Ahmad, A., 1966. J. Am. Chem. Soc. 88(20):4722-4725. Lehman, A. J.; Chase, H. F.; Yonkman, F.F. 1942. . Pharmacol. Exper. Therap 75:270-276. Lehman, A.J. 1936. Proc. Soc. Exptl. Therap. 33:500-503. Lessard, L. G. 1982. Project Summary: Erythrina Project, International Development Research Center, Toronto, Canada; File 3-P-82-0015, pp 1-17. Letcher, R.M. 1971. J. Chem. Soc. (C) 1971:652. Macek, K. 1967. in Heftmann, E. ed., Chromatography, Reinhold Pub. Corp., New York; pp 606-628. Major, R.T.; Folkers, K. 1947. Chem. Abstr. 41:781. Mantel, P.G.; Laws, I.; Widdowson, D.A. 1984. Phytochem. 23(6): 1336-1338. Marafion, J.; Santos, J.K. 1932. Chem. Abstr. 26:5609. Maraflon, J.; Santos, J.K. 1932. Philippine Journal of Science 48:563-577. Marion, L. 1952. in_Manske, R.H.R.; Holmes H.L. eds. The Alkaloids: Chemistry and Physiology, Academic Press, New York; Vol. n. Marten, G.C.; Jordan, R.M.; Hovin, A.W. 1976. Agron. J. 68:909-914. 121

Marten, G.C.; Jordan, R.M.; Hovin, A.W. 1981, Crop Science 21:295-298. Martin, F. W. 1984. Nitorgen Fixing Tree Research Reports 2:57-58. Marum, P.; Hovin, A.W.; Marten, G.C. 1979. Crop Science 19:539-544. Mears, J.A.; Mabry, T.J. 1971. in Harbome, J.B.; Boulter, D.; Turner, B.L. eds. Chemotaxonomy of the Leguminosae. Academic Press, New York; pp 73-178. Megarrity, R.G.; Jones, R.J. 1983. Aust. J. Agric. Res.. 34:791-798, Megirian, D.; Leary, D.E.; Slater, I.H. 1955. . Pharmacol. Exper. Therap 113:212-277. Merck Index, 8 th ed. 1968. Merck, Rahway, New Jersey. Miana, G.A. et. al., 1972. Lloydia 35(1 ):92. Migon, Y.; Bergmann, E.D. 1977. Org. Mass Spectrometry 12(8):500-507. Millbum, P. 1978. in Harbome, J.B. ed. biochemical Aspects of Plant and Animal Coevolution, Academic Press, New York; pp 35-73. Millington, D.S.; Steinman, D. H.; Rinehart Jr., K.L. 1974. J. Am. Chem. Soc. 96(6): 1909-1917. Molyneux, R.J.; Johnson, A.E.; Roitman, J.N.; Benson, M.E. 1979. J. Agric. Food Chem. 21(3): 494-499. Mondon, A. 1970. in Pelletier, S.W. ed. Chemistry of the Alkaloids. Van Nostrand Reinhold Co., New York; pp 173-197. Mondon, A.; Ehrhardt, M. 1966. Tetrahedron Lett. 23:2557-2568. Mondon,A.; Nestler, H.J, 1964. Angew. Chem. internal. Edit. 3(8):588. Mondon, A.; Seidel, P.R. 1970. Chem. Ber. 103:1298. National Science Foundation, 1979. Tropical Legumes: Resources for the Future. National Academy of Science, Washington, D.C.; pp 5-9. Nitrogen Fixing Tree Project, 1986. Final Technical Report: Erythrina Project, Phase /. International Development Research Center, CATIE, Turrialba, Costa Rica; english Translation by L.D. Payne. Oaks, A.J.; Skov, O. 1962. Tropical Ag. 39(4):281-287. Paton W.D.M. 1949. J. Pharm. Pharacol. l(5):273-286. Patterson, H. D. 1951. J. Roy. Stat. Soc. Series B. 13:256. 122

Payne, L.D.; Foley, J.P. 1992a in Ahuja, S. ed. Chromatography and Pharmaceutical Analysis, American Chemical Society Symposium Series, American Chemical Society, Washington, D.C. (accepted) Payne, L.D.; Foley, J.P. 1992b in Agarwal, V.K. ed. Antibiotic/Drug Residues in Food Products of Animal Origin, Plenum Press, New York, (accepted) Perez Castellon, E.E. 1990 . Evaluacion del Ensayo Clonal de Erythrina spp. en San Juan Sur, Turrialba, Costa Rica, Master's tTiesis, CAT 1E, Turrialba, Costa Rica. Peterson, J.E. 1985. in Seawright, A.A. et al. eds. Plant Toxicology. Queensland Poisonous Plants Committee, Yecrongpilly. 191-218. Pichard, R.; Luco J.V. 1944. . Pharmacol. Exper. Therap 80:62-69. Pichard, R.; Luco J.V. 1944. Chem. Abstr. 38:1563. Pick, E.P.; Richards, G.V. 1947, J. Pharm. Exp. Therap. 91:1-13 Pick, E.P.; Unna, K. 1945. J. Pharm. Exp. Therap. 83:59-70 Pineda, O. 1986. Utilization of Pord (Erythrina poeppigiana) Foliage as Feed for Dairy Calves. Master's Thesis, UCR-CATIE, Turrialba, Costa Rica. Polan, C.E. 1974. J. Agri. Food Chem 22(4):635. Polhill, R. M.; Raven, P.H. eds. 1978. Advances in Legume Systematics, Royal Botanic Gardens, England; Part 2. Prakash, O.; Roy, R.; Jain, S.; Bhakuni, D.S, 1988. J. Nat. Prod. 51(3):603- 605. Prelog, V. 1957. Angew. Chem. 69(112):33. Prelog, V.; Wiesner, K. Khorana, H.G.; Kenner, G.W, 1949. Helv. Chim. Acta 32(59):453-461. Purchase, I.F.H.; Steryn, M. 1967. J. Assoc. Off. Anal. Chem. 50(2):363-366. Qayum, A. et al. 1972. Chem. Abstr. 77:148526m. Rahario, Y.C.; Cheeke, P.R. 1986. Nitrogen Fixing Tree Research Reports 4:31- 32. Randall, L.O. 1951. Annals New York Academy of Sciences 54:460-479. Rao, J.V. 1945. Current Sci. 8:198. Raven, P.H. 1979. Ann. Miss. Bot. Garden 66:417-421. Raven, P.H. 1974. Lloydia 37(7):32l-331. 123

Redha, F.J.M. 1983. Chromatographic and Spectroscopic Studies of Erythrina Alkaloids, Ph.D. Dissertation, University of Wales, Cardiff, Great Britain. Rinehart K.L. et al. 1990. Pure and Appl. Chem. 62(7): 1277-1280. Ripperger, H.; Preiss, A.; Diaz, M. 1983. Phytochem. 22(11):2603-2605. Robbins, B.H.; Lundy, J.S. 1947a. Anesthesiology 8:252-265. Robbins, B.H.; Lundy, J.S. 1947b. Anesthesiology 8:348-357. Robinson, T. 1968. Biochemistry of Alkaloids, Springer-Verlag, New York. Rohaijo, Y.C.; Cheeke, P.R. 1986. Nitrogen Fixing Tree Research Reports 4:31- 32. Roldan Perez, G. 1981. Degradacidn Ruminal de Algunos Forrajes Proteicos en funcidn del Consumo de Banano Berde Suplementario, Master's Thesis, CATIE, Turrialba, Costa Rica. Romeo, J.T.; Bell, E.A. 1974. Lloydia 37(4):543-568. Ruiz, M.E. 1983. Avarices en la Investigacion in Sistemas Silvo-pastoriles. Curso Corto Agroforestal, CATIE, 11-21 enero 1983 Russo, R.O. 1984. Erythrina: un gencro versatil en sistemas agroforestales del tropico humedo, INFORAT, CATIE, Turrialba, Costa Rica. Saenz R., J.A. 1964. Rev. Biol. Trop. 12(l):67-74. Saenz R., J.A. 1965. Rev. Biol. Trop. 13(2):207-212. Samur, C. 1984. Produccidn de Leche de Cobras Alimentadas con King Grass (Pennisetum purpureum) y Pord (Erythrina poeppigiana), Suplementadas con Fruto de Banano (Musa sp. 'Cavendish'),cv. Master's Thesis, CATIE, Turrialba, Costa Rica. Sangster, A.W.; Stuart, K.L. 1965. Chem. Rev. 65:69-130. Sarragiotto, M.H.; Filho, H.L.; Marsaioli, A.J. 1981. Can. J. Chem. 59:2771- 2775, Sauvage, G.R.; Berger, F.M.; Boekelheide, V. 1949. Science 109:627 Sauvage. G.L.; Boekelheide, V. 1950. J. Am. Chem. Soc. 72:2062-2064. Scheline, R.R. 1968. J. Pharm. Sci. 57(12):2021-2037. Schoental, R. 1959. J. Pathol. Bact. 77:485-495. Schoental, R. 1968. Can. Res. 28:2237-2246 124

Schoental, R ; Magee, P.N. 1959. J. Pathol. Bad. 78:471-482. Shamma, M. 1960. Experientia 16:484-485. Shamma, M.; Yao, S.Y.; Pai, B.R.; Charubala, R. 1971. /. Org. Chem. 36(21):3253-3254. Shull, L.E. 1980. in Church, D.C. ed., Digestive Physiology and Nutrition of Ruminants, O and B Book, Inc, Corvallis, Oregon; pp 337-359. Shull, L.R.; Cheeke, P R. 1983. J. Animal Sci. 57(suppl. 2):330, Singh, A.N.; Bhakuni, D. S. 1977. Indian J. Chem. 15B:388-389. Singh, A.N.; Pande, H. Bhakuni, D.S. 1977. Lloydia 40(4):322-325. Singh, A.N.; Pande, H.; Bhakuni, D.S. 1976. Experientia 32:1368-1369. Singh, H. Chawla, A.S. 1970. Planta Medica 19:378-379. Singh, H.; Chawla, A.S. 1969. Experientia 25:1. Singh, H.; Chawla, A.S. 1970. Planta Medica 19:71-74. Singh, H.; Chawla, A.S. 1970. J. Ph~rm. Sci. 59(8): 1179-1182, Singh, H.; Chawla, A.S.; Jindal, A.K. 1975. Lloydia 38(2):97-100. Smolenski, S.J.; Silins, H.; Farnsworth, N.R. 1972. Lloydia 35(1): 1-34. Snieckus, V.A. 1971. in Grundon, M.F. ed. The Alkaloids. Burlington House Chemical Society, London, Vol. 1:145. Snieckus, V.A. 1972. in Grundon, M.F. ed. The Alkaloids. Burlington House Chemical Society, London, Vol. 2:199. Snieckus, V.A. 1973. in Grundon, M.F. ed. The Alkaloids. Burlington House Chemical Society, London, Vol. 3:180. Snieckus, V.A. 1974. in Grundon, M.F. ed. The Alkaloids, Burlington House Chemical Society, London, Vol. 4:273. Snieckus, V.A. 1975. in Grundon, M.F. f d. The Alkaloids. Burlington House Chemical Society, London, Vol. ^.176. Snieckus, V.A. 1977. in Grundon, M.F. ed. The Alkaloids. Burlington House Chemical Society, London, Vol. 7:176. Somerville, C.; Browse, J. 1991. Science 252:80-87. Srivastava, V.K.; Moheshwari, M.L. 1985. J. Assoc. Off. Anal, Chem. 68(4):801-803. Standley R.C. 1937. Flora of Costa Rica. Field Museum of Natural History: Botandcal Scriers, Chicago; pp 1-63. Still, W.C.; Kahn, M.; Mitra, A. 1978. /. Org. Chem. 43(14):2923-2925. Strombon, J.; Brahn, J.G. 1978. J. Chromatogr. 147:513 Sudhajain, K.P.; Madhusudana, K.P.; Bhakuni, D. S. 1987. Indian J. Chem. 26B:308-311. Svoboda, G.H. 1964. Lloydia 27(4):299-301. Swick, R.A.; Cheeke, P.R.; Ramsdell, H.S.; Buhler, D.R. 1983. J. Anim. Sci. 56(3):645-647. Unna, K. 1939, American Physiological Society Proceedings 126.P644-P645. Unna, K.; Greslin, J.G. 1944a. . Pharmacol. Exper. Therap 80:54-61. Unna, K.; Greslin, J.G. 1944b, Chem. Abstr. 38:1562-1563. Unna, K.; Kniazuk, M.; Greslin, J.G. 1944a. . Pharmacol. Exper. Therap 80:39- 51. Unna, K.; Kniazuk, M.; Greslin, J.G. 1944b. Chem. Abstr. 38:1562. Van der Maeden F.P.B. 1977. /. Chromatogr. 142:715-723 Vargas Fournier, A. 1983. Estudio Preliminar: Sso de la Hanna de Pescado y Hojas de Pord (Erythrina poeppigiana) en el Desarrollo y Engorde de Cerdos Alimentados con Banano (Musa sp. cv, "cavendish'), CATIE, Turrialba, Costa Rica. Vargas Fournier, A. 1987. Evaluacidn del Forraje de Pord (Erythrina cocleata) como Suplemento Proteico para Toretes en Pastoreo, Master's Thesis, CATIE, Turrialba, Costa Rica. Verpoorte, R. Baerheim Svendsen, A. 1984. Chromatography o f Alkaloids. Journal of Chromatography Library, Elsevier Amsterdam; Vol. 23B. Vfquez, E.; Payne, L.D.; Sanchlz, G.A. 1986. El Huerto Latinoamericano de Arboles Fijadores Nitrdgeno, First Nadonal Forestry Congress, November 11- 14, 1986, San Jos 6, Costa Rica. Wada, K.; Maramuo, S.; Munakata, K. 1966 . Tetrahedron Lett. 42:5179-5184. Wada, K.; Munakata, K. 1967. Ag. Biol. Chem. 31:336-339. Watson, W.C. 1965. Clin. Chim. Acta 12:340-342. 126

Weiss, U., Ziffer, H. 1963. Experientia 19:108-112. Weiss, U., Ziffer, H. 1963. Experientia 19:660. Wenkert, E. 1953. Chem. Ind. 72:1088-1089. Wenzinger, G.R.; Boekelheide, V. 1963. Proc. Chem. Soc. 1963:53. Westermarch, H. 1959. Acta Vet. Scand. 1:67-73 Westling, M.; Smith, R.; Livinghouse, T. 1986. J. Org. Chem. 51 (8 ): 1159-1165. Westwood, S.; Games, D.E.; Sheen, L. 1981. J. Chromatogr. 204:103-107. White, R.D. et al. 1984. J. Animal Sci. 58(5): 1245. Wilde, P.F.; Dawson, R.M.C. 1966. Biochem. J. 98:469-475. Willaman, J. J. 1962. USDA Tech. Bull. 1234. Williams, A.H. 1955. Nature 175:213. Williams, P. 1971. J. Agri. Food Chem 19(6):1198. Williams, R.T. 1972. Tox. Appl. Pharmac. 23:769-781. Wilson, A.D. 1977. Aust. J. Agric. Res.. 28:501. Woodward, R.B. 1942. J. Chem. Soc. 64:72-75. Zarga, M.H.A.; Shamma, M. 1982. J. Nat. Products 45(4):471. Zenk, M.H. 1989. in Poulton, J.E.; Romeo, J.T.; Conn, E.E. eds. Recent Advances in PhytochemA Vol. 23: 429-450 APPENDIX A

SELECTED SPECTROSCOPIC DATA ON ERYTHRINA AND RELATED ALKALOIDS FROM LITERATURE SOURCES

127 128

The units for the data listed below are as follows: optical rotation (degrees), UV (nanometers), IR (reciprocal centimeters), HPLC (mobile phase %volume), MS (m/z, % base peak), and NMR ( 6). BOLDINE Shamma, 1960 i [aid (EtOH) +73 UV Oog e) 303 (4.2), 284 (4.2), 220 (4.6) CRYSTAMIDINE Chunchatprasert, 1982 MS m/z 309 (M+, 100), 294 (-15), 276 (-33) HPLC Uchroprcp EtOH/CHCl3 10/90 NMR 400 MHz 6.32 (1), 6.90 (2), 3.67 (3a), 2.69 (4e), 1.40 (4a), 6.06 (7), 6.88 (14), 6.72 (17), 3.26 (OCH 3), 5.92 and 5.96 (O- CH2-O) Ito, 1976 pale yellow oil [aid +840 (c=0.5, CHCI 3) UV (log e, EtOH) 235 (4.13), 267 (4.15), 357 (3.31) IR 1695 (C=0) MS 309 (M+), 294, 278, 276 (100) NMR (CDCI3) 6.73 (s), 6.70 (s), 6,93 (dd), 6.32 (br. d), 6.07 (m), 6.91 (d), 6.11 (d), 5.95 (q), 3.29 (s), 3.73 (m), 2.72 (dd), 1.40 (t) Mantle, 1984 MS 309 (M+, 85), 294 (-15, 28), 278 (-31. 60), 277 (65), 276 (100), 266 (35), 250 (20) HRMS 309.0997 NMR (CDCI3) 6.30 (dd, 1), 6.92 (dd, 2), 3.65 (dd, 3), 1.6 (4a), 2.70 (ddd, 2.70), 6.05 (s, 7), 6.75 (d, lOe), 6.1 (d, lie), 6.71 (s, 14), 6.88 (s, 17), 3.28 (s, OCH 3), 5.92 and 5.97 (OCH 2O) DEMETHYLERYTHRALINE C 17 H 17 NO3 Redha, 1983 MS 283 (M+), 268 (-15), 266 (-17, 100), 254 (-29), 266 (-57), 213 (- 70) NMR 90 MHz 5.94 (1), 6.42 (2), 4.34 (3a). 1.80 (4a), 2.40 (4), 5.70 (7), 3.82 and 3.30 ( 8 ), 3.30 (lOe), 2.70 (10a), 2.55 (lie), 2.72 (11a), 6.74 (14), 6,60 (17), 5.85 (O-CH 2-O) DIHYDROERYSODINE C 18 H23NO3 Chunchatprasert, 1982 MS m/z (TMS) HPLC lichroprep EtOH/ CHCI3 (10/90) NMR 400 MHz 5.58 (1), 2.51 (2), 2.70 (2), 3.80 (3a), 1.56 (4a), 2.30 (4e), 2.23 and 2.43 (7). 270 and 296 ( 8 ), 3.13 (10a), 3.50 (lOe), 2.51 (lie), 2.91 (11a), 6.64 (14), 6.55 (17), 3.26 (OCH 3). 3.80 (aryl OCH 3) Millington, 1974 (seeds) MW. 301.1692 (HRMS) NMR (100 MHz) 6.73 and 6.65 (s, aromatic); 5.67 (, =); 3.60 (m, next to OCH 3); 3.89 and 3.34 (s, OCH 3); 3.30-1.50 (c, -CH2-) Prelog, 1949 [aid +239 UV Cog e, acid) 285 (3.7) 235 (4.0) DIHYDROERYSOVINE C 18 H 23NO3 129

Ju-ichi, 1977 ( Cocculus) [a]d +223 oil UV (log e, EtOH) 232 (3.83), 299 (3.55) IR 3500 (OH) MS 301 (M+), 243 (-58, 100), 242 NMR (C6D6) 6.99 and 6.30 (m, aryl), 5.37 (m, olefinic); 5.11(br, OH), 3.89 (m, CH 3OCH), 3.27 and 2.98 (s, OCH 3), 2.49 (q), 1.93 (t) DIHYDROXYDIMETHOXYTETRAHYDROBENZYLISOQUINOLINE C 19H 23N O 4 Ghosal ctal. 1972 brown oil UV (log e, EtOH) 220-222 (infl, 4.24) 283-284 (3.81) min 255-257 (2.95) MS 329 (M+), 192 (100), 177 (48), 137 (52), 134 (26) DIMETHYLTRYPTOPHAN METHYL ESTER (N,N) Ghosal et al., 1972 brown gum, m.p. methiodide 208-209 crystallized from EtOH as prisms UV (log e) 222-224 (4.42), 282 (3.79), 292-295 (3.76) [aid -59 (c 0.71, EtOH) IR 3312 (NH), 2855 (NCH), 1722 (CO 2CH3) MS 246 (M+, 78), 183 (-CO 2CH3), 130 (100, -CH(C0 2 CH3)- N(CH3)2), 116 (-CH2CH(C02CH3)-N(CH3)2) NMR (60 MHz, CDCI3) 2.9 (s, N(CH3)2), 3.25 (dd), 3.32 (s, OCH3), 3.54 (dd), 4.45 (s), 7.0-8.2 (complex, 5H) EPIERYTHRATIDINE Chunchatprasert, 1982 MS m/z (TMS) 403, 388, 345, 257 (100) NMR 400 MHz 5.61 (1), 4.35 (2), 3.65 (3a), 2.30 (4e), 1.68 (4a), 3.04 and 2.74 ( 8 ), 2.43 amd 2.23 (7), 3.47 (lOe), 3.16 (10a), 2.57 (lie), 3.04 (11a), 6.78 (14), 6.58 (17), 3.82 and 3.83 (aryl OCH 3), 3.27 (OCH 3) ERYBIDINE C20H25NO4 Chunchatprasert, 1982 MS m/z (TMS) 413, 400, 372, 358 (100), 343 ...... 73 Crabb, 1975 NMR in CDCI3: 3.92, 3.87, and 3.87 (CH 3O); 2.32 (NCH3) Ito, 1971, 1973 and 1976 m.p. 178-180 [aid (EtOH) +0 UV (log e, EtOH) 216 (4.50), 284 (3.92) IR (CHCI3) 3500 (OH) MS 343 (M+), 328, 300, 286, 285, 256, 255 NMR 60 MHz (CDCI3) 2.32 (s, NCH3), 2.42-2.78 ( 8 H), 3.87 and 3.92 (OCH3), 6.78, 6.70, 6.68 , and 6.67 (s, aromatic) Jackson, Allertonia, 1982 (leaves of poeppigiana) Redha, 1983 (leaves of poeppigiana) 130

ERYMELANTH1NE C 18 H20N2O4 Redha, 1983 MS 312, M-15, M-31 (100%) IR 1730 cm-1 (ester carbonyl) NMR 3.96 (methoxy carbonyl) 360 MHz spectra shown ERYSODIENE Barton, 1968 m.p. 226-230 MW 313.131868 IR vrnax in CHCI 3 3545, 3300, 1675, 1655, 1625 cm-1 UV (e, EtOH,) 285 (3900) and 243 (20,700) MS 313 (M+), 282 (100), 270,254 NMR CDCI3 at 100MHz 6.67 (s, 17); 3.37 (s, 14); 6.28 (t, 1); 5.99 (s, 4); 3.70 (s, OCH 3); 3.61 (s, OCH3); 3.65-2,10 (complex 8 methylenes) Ghosal, 1971 m.p. 222-224 stTaw colored needles IR vrnax 3525, 3305, 1672, 1648, 1620 UV (log e) 238-242 (4.18) and 282-285 (3.05) MS 313 (M+, 62); 298 (17), 282 (100) ERYSODIENONE C 18 H 19NO4 Barton, 1968 m.p. 226-230 (EtOH) UV (c, EtOH) 285 (3900), 243 (20,700) IR (CHCI3) 3545, 3300, 1675, 1655, 1625 MS 313, 282(100), 270, 254 NMR 6.64, (s, 17); 6.37 (s, 14); 6.28 (t, 1); 5.99 (s, 4); 3.7 (s, OCH3), 3.61 (s, OCH3), 3.65-2.1 (m, 8 f methylene H) Barton, 1970 m.p.226-229 UV (e) 243 (20,700) and 285 (3900) IR 3545, 3300, 1675, 1655, and 1625 NMR 6.65 (s), 6.37 (s) 6.28 (t), 6.00 (s), 3.71 and 3.61 (s, OCH 3), 3.65-2.1 (complex 8 H) Gervay, 1966 UV 242, 283 IR 3500, 1692, 1665, 1630, 1595 NMR (CDCL3) 6.64 and 6.38 (s, aryl), 6.01 (s, olefinic), 3.28 (t, olefinic), 6.21 (OH), 3.72 and 3.62 (OCH 3), 3.5-2.2 (br CH2 muldplets) Ghosal, 1971 straw coloured needles from EtOH m.p. 222-224 UV flog e) 238-242 (4.18), 282-285 (3.50) IR 3525, 3305, 1672, 1648, 1620 MS 313 (M+, 62), 298 (17), 282 (100) Ghosal, 1972 m.p. 222-225 UV (log e) 240-242 (4.32), 282 (3.55) IR 3533, 3286, 1672, 1655, 1614 MS 298 (35), 282 (100), 270 (18), 254 (22) NMR (60 MHz, CDCI3) 6.66 (s), 6.28 (t), 5.98 (s), 3.72 (s, OCH 3), 3.63 (s, OCH3), 2.2-3.7 (complex, 8 H) 131

Mondon, 1966 UV (log e) 241 (4.12), 283 (3.47) IR 1681, 1655, 1618 (conjugated diene) MS 313 (M+), 282 (100) ERYSODINE C 18 H21NO3 Boar, JCS, 1970 MS M+ (58%), M-CH 3, M-CH3O (100%), M-CH50, M-C3H80, M- C4H70, M+ C5H80 Boekelheide, 1960 pKa 6.29 UV (log e) 285 (3.5); 235 (4.3) Chawla, 1982 (seeds and leaves of berteroana) MS of TMS deriv. 371 (M+, 58); 356 (M-CH 3, 49); 340 (M-OCH3), 100); 337 (15); 313 (10); 310(10); 287 (10); 73 (26) Chawla, 1988 UV (log e, CH3OH) 228 (4.3); 283 (3.7); MS m/z (%) 299 (M+, 72), 284 (47), 268 (100) NMR (CDCI3) 6.78 (s, 14), 6.68 (s, 17), 6.57 (dd, 2), 5.99 (d, 1), 5.72 (br. s, 7), 3.74 (s, aryl OCH 3), 3.28 (s, CH-OCH 3), 2.52 (dd, 4e), 1.76 (t, 4a) Chunchatprasert, 1982 MS m/z (TMS) 371, 356, 340 (100), ....73 HPLC Lichroprep EtOH/CHCl3 10/90 NMR 360 MHz 6.00 (1), 6.57 (2), 4.05 (3), 1.83 (4a), 2.52 (4e), 5.7 (7), 3.75 ( 8 ), 3.55 (8 ), 3.49 (lOe). 2.92 (10a), 2.92 (11a), 2.61 (lie), 3.77 (14), 6.68 (17), 3.78 (aryl OCH 3), 3.34 (OCH3) Crabb, 1975 NMR in C6D6: 6.65 and 6.76 (aromatic H) El Olemy, 1972 m.p. 203-205, picrate 210-212, methiodide 229-30 IR 3400 (OH) Fokers, JACS, 71, 1949 crystallizes from EtOH containing excess HC1 Folkers, JACS, 62, 1940 m.p. 204-205 yellowish, granules [aid +248 c~ 0.311 in EtOH Folkers, JACS, 63, 1941 (seeds of costarricensis) m.p. 200-201 [aid +248 EtOH Folkers, JACS, 64, 1942 weak yellow-green fluorescence on alumina column where deposited w/ chloroform and eluted with EtOH Gentile, 1942 m.p. 204-205 [aid +250 (EtOH, c 0.3212, dm) Ghosh, 1972 m.p. 202 MS 299 (M+), 284, 268, 266 Glasby, 1975 m.p. 204-205 colorless needles from EtOH [aid +248 132

Hargreaves, 1974 (seeds of poeppigiana, costarricensis and berteroana, not in leaves) Ito, 1973 and 1976 m.p. 202-204 [aid +231 (c 0.7, CHCI 3) UV (log e, EtOH) 228 (4.25), 284 (3.60) IR (CHCI3 ) 3520 (OH) MS 299 (M+), 284, 268 (100), 266, 241, 228, 215 NMR 60 MHz (CDCI3) 1.82 (t, 4), 3.33 and 3.78 (s, OCH 3), 5.75 (br. s, 7), 6.02 (br.d, 1 \ 6.60 (br. dd, 2), 6.72 and 6.83 (s, arom.) Jackson, Allertonia, 1982 (leaves of poeppigiana) Marion, 1952 (seeds of berteroana, costarricensis and poeppigiana) m.p. 204-205 [aid +248 Millington, 1974 m.p. 205-205.5 Prelog, 1949 [aid +267 UV (log e, acidic) 287 (3.5); 238 (4.4) Redha, 1983 (leaves of poeppigiana) Sangster, 1965 UV (log e, EtOH/O.OlN H+) 235 (4.32), 285 (3.57); (EtOH/0.01 N OH-) 230-240 (sh, 4.28) and 297 (3.66) Singh, 1969 m.p. 208-210 HC 1 206-208 ERYSOLINE C17H19NO3 Chawla, 1982 (seeds of berteroana) MS of TMS deriv. 429 (M+, 100); 414 (M-CH 3, 16); 398 (m-OCH 3, 3); 340 (51); 338 (42); 326 (12); 313 (21); 308 (12); 287 (10); 73 (60) Chunchatprasert, 1982 MS m/z (TMS) 429, 414, 340 (100), 338 ...... 73 NMR 220 MHz 5.96 (1), 6.53 (2), 4.48 (3a), 1.83 (4a), 2.54 (4e), 5.74 (7). 3.68 ( 8 ), 2.94 (10a), 2.94 (11a), 2.54 (lie), 3.50 (lOe), 6.87 (14), 6.65 (17), 3.87 (aryl OCH 3) Hargreaves, 1974 (seeds of berteroana) Millington, 1974 (seeds) MW 285.1362 (HRMS) MS 285 (M+, 100) 268 M-OH, 89); 266 (M-OH, 21); 254 (M-CH 3O, 22) NMR (60 Mhz) 6.69 and 6.46 (aromatic s); 6.34, 5.78 and 5.56 (m, -); 4.20 (m, next to OCH 3); 3.79 (s, OCH 3 ); 3.70-1.55 (complex-CH 2 ‘) UV (c, EtOH) 225 (11,900) 283 (2950) Redha, 1983 M S (T M S ) 429, 414, 398, 328 (100) .... 73 ERYSONINE C 17 H 19NO3 Boekelheide, 1960 pKa 6.50 Chawla, 1982 (seeds of berteroana) MS of TMS deriv. 429 (M+. 100); 340 (58); 338 (38); 326 (12); 313 (24); 308 (10); 73 (58) 133

Chunchatprasert, 1982 MS m/z (TMS) 429, 414, 340 (100), 338, ,...73 Crabb, 1975 NMR in CftDfi: 6.65 and 6.76 (aromatic H) Folkers, JACS, 63, 1941 (seeds of costarricensis) chromatographed with chloroform and methanol on alumina, sparingly soluble in most solvents incl alcohols. Soluble in 5CV50 methanol and chloroform or in morpholine. m.p. 236-237 and 238-239, difficult to obtain clear m.p. [aid +285-288 (0.5%HC1); +272 (morpholine) Glasby, 1975 m.p. 236-237 colorless needles from EtOH [aid +289 Hargreaves, 1974 (seeds of costarricensis and berteroana) Marion, 1952 (seeds of costarricensis) no characteristic green color, phenolic m.p. 238-239 [aid +285-288 Millington, 1974 (seeds) MS 285 (M+, 100) 268 M-OH, 89); 266 (M-OH, 21); 254 (M-CH 3O, 22) ERYSOPINE C17H19NO3 Boar, JCS, 1970 MS M+ (83%), M-CH 3, M-CH3O (100%), M-CH50, M-C3H8 O, M- C4H7 O, M+ C5H8 O Boekelheide, 1960 pKa 6.60 UV (log e) 292 (3.6); 240 (4.3) Chawla, 1982 (seeds of berteroana) MS of TMS deriv. 429 (M+, 62); 414 (M-CH 3, 24); 398 (M-OCH 3, 100), 340 (12); 73 (45) Chunchatprasert, 1982 MS m/z (TMS) 429, 414, 398 (100)...... 73 HPLC Lichroprep EtOH/ CHCI3 10/90 Folkers, JACS, 62, 1940 (seeds of poeppigiana) m.p. 240-241 white (green color with FeCl 3 solution w/ 1 drop HC1) [aid +263.4 c= 0.291 in 60% absolute EtOH and 40% glycerol Folkers, JACS, 63, 1941 (seeds of costarricensis) m.p. 242-243 [aid +263 (40% glycerol and 60% EtOH); +276 (c=18.413 mg/2.023 mL of 0.5% HC1); +225 (c=l3.860 mG/2.023mL morpholine) Gentile, 1942 m.p. 241-242, white prisms, with ferric chloride gives green color [aid +265 (EtOH-glycerol, c 0.0846, 2 dm) Ghosh, 1972 m.p. 241-242 MS 285 (M+), 270, 254, 252 Glasby, 1975 m.p. 241-242 colorless crystals from EtOH [aid +265.2 sparingly soluble in CHCI 3, H2O and protic solvents 134

Hargreaves, 1974 (seeds of poeppigiana, costarricensis and berteroana, not in leaves of poeppigiana) Jackson, Allertonia, 1982 (leaves of poeppigiana) Marion, 1952 (seeds of berteroana and poeppigiana) less soluble in water, chloroform and alcohols, forms characteristic green color with ferric chloride in acid solution, base HC1 prisms m.p. 241-242 [aid +265.2 Millington, 1974 m.p. 230-235 Redha, 1983 MS (TMS) 429, 414, 398, 328 (100)....73 ERYSOPITINE C 17 H 21NO4 Ghosal, 1972 m.p. 168-171 [aid +148 (c 0.52, EtOH) UV (log c) 285-287 (loge 3 4.31 MS 303 (M+, 92), 288 (18), 271 (100), 245 (41, -CH2=CHOCH3) ERYSOSALVINE C 18 H23NO4 Glasby, 1975 Millington, 1974 (seeds) MS 317 (M, 8); 257 (a, 75); 258 (a, 23); 243 (b, 100); 242 (b, 26) NMR (60 MHz) 6.60 and 6.55 (s, aromatic); 4.45 and 5.87 (m, =); 3.60 (m, next to OCH 3); 3.86 and 3.36 (s, OCH 3); 3.30 and 1.75 (c, -CH2 -) ERYSOSALVINONE Chunchatprasert, 1982 MS m/z (TMS) 461, 446, 430, 403 (100), 315...... 73 Millington, 1974 (seeds) MS 317 M, 8 ); 257 (a, 75); 258 (a, 23); 243 (b, 100); 242 (b, 26) ERYSOTHIOPINE C 19H2 INO 7 S Folkers, JACS, 66, 1944 (seeds of poeppigiana) m.p. 168-169 [aid +193 (c=0.183, EtOH) Glasby 1975 m.p. 168-169 [aid +194 Marion, 1952 (seeds of berteroana) m.p. 168-9 [aid 194 CH3rck Index, 1968 MW 407.44 m.p. 168-169 [aid +194 (c=0.103 in EtOH) ERYSOTHIOVINE C20H23NO7 S Folkers, JACS, 66, 1944 (seeds of poeppigiana) m.p. 187 short white crystals [aid +208 (c=0.359 in EtOH) Glasby, 1975 m.p. 187 [aid +208 Marion, 1952 (seeds of berteroana and poeppigiana) 135

m.p. 187 [aid 208.5 CH3rck Index, 1968 MW 421.48 m.p. 187 [aid +208 (c=0.359, EtOH) ERYSOTINE Chunchatprasert, 1982 MS m/z (TMS) 461, 446, 430, 703 (100), 315 ...... 73 Millington, 1974 (seeds) MW 317.1930 (HRMS) m.p. 225-227 colorless crystals MS 317 (M, 8 ); 257 (a, 75); 258 (a, 23); 243 (b, 100); 242 (b, 26) NMR (60 MHz) 6.68 and 6.48 (s, aromatic); 4.45 and 5.87 (m, =); 3.60 (m, next to OCH 3); 3.8 and 3.36 (s, OCH 3); 3.30 and 1.80 (c, - CH2 ) ERYSOTINONE CI8 H21NO4 Boar, JCS, 1970 MS m/z 315 (M+, 20%), 272, 257 (100%), 229, 228, 201, 200 Glasby, 1975 Millington, 1974 (seeds) MW 315.1467 (HRMS) m.p. 177-179 colorless needles [aid +342 (c 0.28 abs EtOH) UV (e) 226 (17,600) IR vrnax (CHCI3) 1675 cm-1 (C=0) MS 315 (M+, 3); 257 (75); 256 (11); 229 (70); 228 (100); 242 (15); 272 (25) NMR (100MHz) 6.80 and 6.58 (s, aromatic) 6.18 (m, =); 4.10 (q, next to OCH 3); 3.84 and 3.56 (s, OCH 3); 3.5-2.08 (c, -CH 2-) Mondon, 1970 m.p. 196-197 (racemic) UV (log e) 228 (4.5) IR (KBr) 1680 cm -1 (C=0) NMR 6.77 and 6.55 (s, aromatic); 6.12 (m, =); 4.07 (q, next to OCH3); , 3.80 and 3.51 (s, OCH 3) ERYSOTRAMIDINE C19H21NO4 Chunchatprasert, 1982 MS m/z (TMS) 327 (M+, 100), 312 (-15), 294 (-33) IR CHCL3 1665 CM-! HPLC Lichroprep EtOH/CHCl3 10/90 NMR 400 MHz 6.32 (1), 6.89 (2), 3.86 (3a), 1.70 (4a), 2.80 (4e), 6.02 (7), 3.99 (10c), 3.60 (10a), 3.08 (11a), 2.98 (lie), 6.78 (14), 6.70 (17), 3.34 (OCH 3), 3.76 (aryl OCH 3), 3.86 (aiyl OCH 3) Glasby, 1975 ERYSOTRINE C 19H 23NO3 Boar, JCS, 1970 MS M+ (40%), M-CH3, M-CH3O (100%), M-CH5O, M-C3H8 O, M- C4H7 O, M+ C5H8 O 136

Chawla, 1988 UV Gog e, CH 3OH) 230 (4.3); 280 (3.8); MS m/z (%) 313 (M+, 84), 298 (81), 282 (100) NMR (CDCI3) 6.71 (s, 14), 6.48 (s, 17), 6.38 (dd, 2), 5.85 (d, 1), 5.58 (br. s, 7), 3.95 (m, 3), 3.74 (s, aiyl OCH3), 3.64 (s, aryl OCH3), 3.22 (s, OCH3), 2.48 (dd, 4e), 1.76 (t, 4a) Chunchatprasert, 1982 MS m/z (TMS) 313, 298, 282 (100), HPLC Lichroprep EtOH/CHCl3 10/90 NMR 220 NHz 6.05 (1), 6.69 (2), 4.10 (3a), 1.85 (4a), 2.55 (4c), 5.77 (7), 3.71 (8), 3.63 (8), 2.67 (lie), 3.55 (lOe), 2.90 (10a), 2.90 (11a), 6.87 (14), 6.68 (17), 3.36 (OCH3), 3.78 (aryl OCH3), 3.88 (aryl OCH3) El Olemy, 1972 HC1 pale yellow needles m.p. 205-207 pricrate 160-162 UV (log e, EtOH) 280(3.5), IR 2250 (amine salt) 1615 MS 313 (M+), 298 (-15), 282 (-31), 281 (100), 280, 266 (-47) NMR (CDOD) 3.85 and 3.75 (s, aryl OCH3), 3.35 (s, alkyl OCH3), 6.9 and 6.7 (s, aryl), 6.55 (dd), 2.50 (ddO, 1.80 (t) Ghosal, 1970 pale brown syrupy liquid, yellow picrate m.p. 161, picrolonate, 189- 190 [aid (EtOH) +142 IR 1640, 1622, 1258, 1020, 805 UV (e) 235 (21,880) 282-285 (3262) MS 313 (M+), 298 (-CH3), 282 (-OCH3), 179 Ghosh, 1972 m.p. picrate 160 MS 313 (M+), 298, 282, 280 Glasby, 1975 m.p. 95-97 colorless crystals from light pet UV 228, 280 Hargreaves, 1974 (seeds of poeppigiana) Ito, 1976 oil, picrate 161-162 [aid +140 (c 0.40, EtOH) UV (EtOH) 229 and 280 MS 313 (M+), 298, 282, 280 NMR in CDCI3 3.82 (s), 3.75 (s), 3.32 (s), 6.56 (dd), 5.98 (br. d/0, 5.70 (br. s), 6.83 (s), 6.63 (s), 4.05 (m), 2.53 (dd), 1.83 (t) Jackson, Allertonia, 1982 (leaves of poeppigiana.) Letcher, 1971 m.p. 95-96 from light pet. picrate 162-163 [aid+142 (c 0.40, EtOH) UV (log e , EtOH) 228 (4.17) and 280 (3.52) MS 313, 298, 282(100) NMR in C6D6 100MHz 5.93, 6.37, and 5.43 («, 1,2,7), 4.07 (next to OCH3), 2.11 (-CH2-.4), 6.97 and 6.46 (aromatic H), 3.46, 3.44 and 3.03 (OCH3) 137

Miana, 1972 m.p. 94-98 colorless crystals picrate 164 HQ 205-206 UV 232,284 NMR CDQ3 3.33, 3.78, 3.85 (s, OCH3) MS 313 (M+), 298 (M-CH3) and 282 (M-OCH3, 100%) Redha, 1983 (leaves of poeppigiana) Sarragiotto, 1981 UV (c, EtOH) 230 (12100), 280 (3300), IR 1610 MS 313 (M+, 65); 298 (41); 282 (100); 280 (12) NMR (100 MHz, CDCI3) 6.86 (s, 14); 6.66 (s. 17); 6.56 (d, 2); 6.04 (br d. 1); 5.76(br s, 7); 3.86 and 3.34 (s, OCH3); 1.89 (t. 4a) Singh, 1969 (seeds) m.p. HQ 206-208 UV (e) 232 (19,800) 284 (3,490) MS 313 (M+, 33), 298 (M-CH3, 35), 282 (M-CH3O, 100) ERYSOTRINE-N-OXIDE C19H23NO4+ Sarragiotto, 1981 [aid+ 78.3 (ETOH, c 1) UV (log e, EtOH) 231(14,000), 278 (2900 IR 1610 MS 329 (M+, 5); 313 (10), 311 (10), 298 (12), 282 (20), 296 (90), 254 (36), 239 (100), 224 (55), 165 (62), 152 (52) NMR (100 MHz, CDQ3) 6.83 (s, 1), 6.76 (s, 14), 6.86 (d, 2), 4.06- 4.12 (m, 4H), 3.95, 3.84 and 3.43 (s, OCH3), 3.36-3.08 (m, 3H), 2.14 (dd, 4e) ERYSOVINE CI8H21NO3 Boar, JCS, 1970 MS M+ (46%), M-CH3, M-CH3O (100%), M-CH5O. M-C3H8O, M- C4H7O, M+ C5H8O Boekelheide, 1960 pKa 6.45 UV (log e) 285 (3.5); 239 (3.8) Chawla, 1982 (seeds and leaves of berteroana) MS of TMS deriv. 371 (M+, 58); 356 (M-CH3, 49); 340 (M-OCH3, 100); 338 (14); 313 (8); 310 (18); 73 (28) Chawla, 1988 UV Gog e, CH3OH) 228 (4.3); 283 (3.6); MS m/z (%) 299 (M+, 39), 284 (41), 268 (100) NMR (CDCI3) 6.83 (s, 14), 6.59 (s, 17), 6.52 (2), 5.96 (1), 5.66 (br. s, 7), 3.84 (s, aryl OCH3), 3.28 (s, OCH3), 2.50 (dd, 4e), 1.78 (t, 4a) Chunchatprasert, 1982 MS m/z (TMS) 461, 446, 430, 403 (100), 315...... 73 HPLC Lichroprep EtOH/CHCl3 10/90 NMR 400 MHz 5.98 (1), 6.51 (2), 4.03 (3a), 1.82 (4a), 2.50 (4e), 5.68 (7), 3.68 (8), 3.52 (8), 3.47 (lOe), 2,96 (10a), 2,90 (11a), 2.67 (lie), 6.85 (14), 6.60 (17), 3.86 (aryl OCH3), 3.32 (OCH3) Crabb, 1975 NMR in C6D6: 6.65 and 6.76 (aromatic H) Folkers, JACS, 62, 1940 (seeds of berteroana and poeppigiana) m.p. 178-179.5 large prisms 138

[a](j +2232-234 Gentile, 1942 m.p. 175-176 Glasby, 1975 m.p. 178-179 colorless prisms from Et20 [aid +252 readily soluble in CHCI3 and moderately so from EqO and EtOH Hargreaves, 1974 (seeds of poeppigiana and berteroana, not in costarircensis) Jackson, 1985 NMR 360 MHz spectra shown Marion, 1952 (seeds of berteroana, costarricensis and poeppigiana) m.p. 178-9.5 baseHCl needles taJd +252 Millington, 1974 m.p. 178-179 Singh, 1970 m.p. 181-182 [aid +188 (c=0.47 ub /CHCI3) MS m/z 299 (M+, 44%), 284 (-CH3, 47), 268 (-CH3O, 100) UV (log e) 285 (3.6) Singh, 1975 mp 169-170 TLC n-Bu0 H-Ac0 H-H20 (4:1:2) UV (log e, EtOH) 284 (3.89) MS m/z 299 (M+, 53), 284 (-CH3, 54), 268 (-OCH3, 100) NMR 682(14-H), 6.58 (17-H) aromatic, 1, 2,and 7 H 5.95, 6.43, 5.64 (diffuse); 3.28, 3.82 (methoxy) ERYSTEMINE C20H25NO4 Glasby, 1975 m.p. 127-129 pale yellow prisms from light pet. pricrate m.p. 145-150 [aid +189 ERYTHAMIDE C19H24NO3 Ju-ichi, 1981 (Cocculus) colorless cubes m.p. 87-89 [aid +262 (EtOH) IR (CHCI3) 3500, 3350, 1660 MS 328 (M+), 270 (-58, 100), NMR (acetone, d6) 4.87 and 3.99 (OCH3), 5.58 (m, olefinic H), 6.93 and 7.91 (s, aromatic) ERYTHLAURINE C20H25NO5 Ju-ichi, 1981 (Cocculus) oil [aid +232 (EtOH) IR (CHCI3) 3500, 1710 MS 273 (-58, 100) NMR (CDCI3) 3.30 (s, OCH3), 3.91 and 3.93 (x. OCH3 and COOCH3), 5.63 (m, olefinic H), 7.18 (s, aromatic) ERYTHRALINE C18H19NO3 Boar, JCS, 1970 MS M+ (72%), M-CH3, M-CH3O (100%), M-CH5O, M-C3H80 , M- C4H7O, M+ C5H8O Boekelheide, 1960 139

pKa 5.97 UV (log c) 292 (3.6); 238 (4.3) Chunchatprasert, 1982 MS m/z 297, 282, 266 (100), HPLC Lichroprep EtOH/CHCl3 10/90 NMR 220 MHz 5.98 (1), 6.55 (2), 3.96 (3), 1.84 (4a), 2.48 (4e), 5.72 (7), 3.75 (8), 3.48 (3), 3.48 (lOe), 2.9 (10a), 2.85 (11a), 2.7 (lie), 6.78 (14), 6.52 (17), 5.9 (O-CH2-O), 3.32 (OCH3) Dalton, 1979 m.p. 106-107 [aid +212 (EtOH) Deulofeu, 1947 HI m.p. 252-253 yellow needles [aid +176 (water, c=0.229) Folkers, JACS, 1940 UV (e) 291 (14.5) off scale after 240 Folkers, JACS, 1940 m.p. 106-107 [aid +211.8 (c=0.944, abs. EtOH) Ghosh, 1972 m.p. HBr 256 297 (M+), 282, 266, 264 Glasby, 1975 m.p. 106-107 [aid +211 Hargreaves, 1974 (seeds of poeppigiana and berteroana, leaves of poeppigiana, not in seeds of costarricensis) Ito, 1973 and 1976 oil HBr m.p. 245-248 [aid +223 (c 0.6, H2O) UV 232,290 MS 297 (M+), 288, 266 (100), 264, 226, 213 NMR 60MHz (CHCI3) 1 .S3 (t, 4); 3.33 (OCH3); 5.73 (br. s, 7); 5.89 (s, OCH2O); 5.90 (br. d, 1); 6.62 (br. dd, 2); 6.64 and 6.78 (s, aromatic) Marion, 1952 m.p. 106-7 [aid +211.8 Prelog, 1949 [aid 228 UV (log e, acid) 290 (3.7); 238 (4.3) base 290 (3.6) 235 (4.2) Sangster, 1965 UV Gog e, EtOH/O.lN H+) 238 (4.29), 290 (3.66); (EtOH/O.OlN OH- ) 225 (4.16), 273 (sh, 3.23), 290 and (3.59) Singh, 1970 m.p. HBr 250-251 UV (log e) HBr 291 (3.6) 140

ERYTHRAMINE C18H21NO3 (DIHYDROERYTHRALINE) Barton, 1968 m.p. 120-121 [a]d+168 (0.33) MS 241 (-58, 100) NMR in CDCI3 at 100MHz 5.55 (1); 2.34 (2,2) 3.73 (3); 1.67 (4a); 2.78 (4c); 6.60 (14); 6.60 (17); 5.85 (OCH2O); 3.26 (OCH3) Boar, JCS, 1970 MS m/z 299 (M+, 20%), M-15, 268 (-31), 241 (-58,74%), 240 (100%) Crabb, 1974 NMR in CD Cl 3; 5.85 (bridged methylene); 6.6 and 6.6 (aromatic H), 3.26 (CH3O); 3.73 (H next to CH3O); 2.34 (aliphatic next to =); 5.55 (=); 1.57 and 2.78 (aliphatic next to spiro) Folkers, JACS, 1939 m.p. HC1 249 free base unstable m.p. 104-105 [a]d HC1 +220 Folkers, JACS, 1940 UV (log e) 291 (4.2) and 238 Ghosh, 1972 m.p. 120-121 MS 399 (M+) 268, 241,240 Glasby, 1975 mp. 103-104 m.p. HI 249 orange-yellow needles form EtOH Marion, 1952 m.p. 103-4 small crystals Prelog, 1949 [aid +223 UV (log c, acid and base) 292 (3.65); 241 (3.65) Sangster, 1965 UV (log e, EtOH/O.lN H+) 241 (3.65), 291 (3.65); (EtOH/O.OlN OH-) 241 (3.65) and 291 (3.67) ERYTHRARTINE C19H23NO4 El Olemy, 1972 yellow needles, m.p. 166-168 UV (log e, EtOH) 230 (8.2), 280 (3.3) IR 3200, 1620 MS 329 (M+), 314 (-15), 311 (-18), 298 (-31)297,(100), 296 (-18, 15), 280 (-18, 31), 279 (-18, 32) NMR (CDCI3) 3.80 and 3.90 (s, aryl OCH3), 3.4 (alkyl OCH3), 6.95 and 7.05 (s, aryl), 6.6 (dd), 6.0 (br. d), 5.8 (br s), 4.70 (t), 4.15 (m) Sarragiotto, 1981 [aid + 256 (CHCI 3 , c 1.4) UV (e), EtOH) 227 (15,135), 275 (3630) IR 1610 MS 329 (M+, 75); 314 (26), 298 (100), 296 (28), 280 (12) NMR (100 MHz, CDCI3) 7.08 (s, 17), 6.92 (s, 14), 6,68 (dd, 2), 6.10 (br d, 1), 5.82 (br s, 7), 4.80 (t, lie), 3.95, 3.83 and 3.37 (s, OCH3), 3.67 (dd, 10a), 3.14 (dd, lOe) 141

ERYTHRARTINE-N-OXIDE C19H23NO5 Sarragiotto, 1981 [a]d + 88.57 (ETOH, c 1.4) UV (e, EtOH) 231(14,000), 278 (2900) IR 1610 MS 345 (M+, 4) 329 (26); 326 (29), 298 (70), 239 (100), 224 (45), 165 (47), 152 (39) NMR (100 MHz, CDCI3) 7.08 (s, 1), 6.68 (s, 14), 6.74 (d, 2), 6.18 (d, 1), 5.82 (br s, 7). 5.28 (br d, 11a), 4.4-4.0 (m, 4H), 3.92, 3.78 and 3.38 (s, OCH3), 3.16 (t, 4a), 2.09 (dd, 4c) ERYTHRASCINE Ghosal et al., 1972 cream colored needles m.p. 138-140 [aid +152 (c 0.51, CHCI3) UV (log c, EtOH) 210-212 (4.88), 233-235 (4.36), 284-288 (3.34) MS 371 (<+, 42), 356 (22), 340 (100), 339 (27), 329 (31), 313 (24), 311(17) ERYTHRATIDINE C19H25NO4 Barton, JCS, 1973 m.p. 120-120.5 [aid +258 IR vmax 3509 and 3387 cm-1 UV (e) 219 (22,500), 232infl (5700) and 284 (2600) MS m/z 331 (M+), 300, 273, 257 (100%), 244 NMR 6.58 (l,s), 5.81 (l,m), 4.43 (l,m), 3.6 (l,m), 3.74 (3,s), 3.3 (3,s), 3.8-1.68 (11, complex aliphatics) Chawla, 1988 UV (log e, CH3OH) 234 (3.75); 283 (3.49); NMR (CDCI3) 6.64 (s, 14), 6.50 (s, 17), 5.93 (1), 4.50 (2), 3.87 (s, aryl OCH3), 3.82 (s, aryl OCH3), 3.68 (m, 3) Chunchatprasert, 1982 MS m/z (TMS) 403, 388, 345 (100), 257 ..... 73 NMR 400 MHz 5.84 (10, 4.46 (2), 3.66 (3a), 1.83 (4a), 2.04 (4e), 2.46 (7), 2.22 (7), 2.65 and 2.96 (8), 3.13 (10a), 3.49 (lOe), 3.00 (1 la), 2.58 (1 le), 6.47 (14), 6.60 (17), 3.84 and 3.78 (aryl OCH3), 3.34 (OCH3) Glasby, 1975 m.p. 120-121 colorless needles from EtOH as hemihydrate [a]d +222.5 Jackson, Allertonia, 1982 (leaves of poeppigiana and berteroana) Millington, 1974 (seeds) MW 331.1770 (HRMS) C19H25N04 m.p. 119-120 [aid +273 (c 0.3 abs EtOH) UV (e, abs EtOH) 232 (7820) 287 (3550) MS 331 (M+, 100), 300 (M-CH3O, 12) 273 (81), 272 (18), 257 (100), 256 (20) NMR (IOOMHz) 6.70 and 6.59 (s, aromatics); 4.56 and 5.93 (m, =); 3.70 (m, next to OCH3); 3.94, 3.89 and 3.44 (s, OCH3); 3.8-1.80 (c, -CH2-) 142

Redha, 1983 (leaves of poeppigiana) HPLC EtOH/ CHCI3 10/90 ERYTHRATIDINONE C19H23NO4 Banon, JCS, 1973 m.p. 119-120 [aid +358 (c = 1.121 in CHCI3) IR vmax 1675 cm-1 UV (e, EtOH) 231 (19,600). 284 (3950) MS m/z 329 (M+), 301, 298, 286, 272, 271 (100%), 243, 242, 228, 215, 214,and 197 NMR 6.68br (l,s), 3.57 (l,s), 6.13 (l,m), 4.05 (l,m), 3.88 (3,s). 3.78 (3,s), 3.5 (3,s) 3.32-2.28 (10, complex aliphatics) Chunchatprasen, 1982 NMR 400 MHz 6.02 (1), 3.96 (3a), 2.11 (4a), 2.52 (4e), 2.44 and 2.61 (7), 2.71 and 2.94 (8), 3.12 (10a), 2.95 (11a), 2.58 (lie), 6.67 (17), 6.48 (14), 5.86 and 5.86 (O-CH2-O), 3.46 (OCH3) Glasby, 1975 m.p. 119-120 Jackson, Allertonia, 1982 (leaves of poeppigiana) Redha, 1983 (leaves of poeppigiana) ERYTHRATINE C18H21NO4/2H2O Banon, 1968 IR 3605 (intramolecular H-bond) NMR in CDCI3 at 100MHz 5.56 (1); 4.28 (2); 3.60 (3); 1.63 (4a); 2.30 (4e); 6.73 (14);6.59 (17); 5.85 and 3.28 (OCH3) Boar, JCS, 1970 MS m/z 315 (M+, 14%), 384, 257, 256, 241, 240 (100%), 228, Crabb, 1975 NMR in CDCI3: 5.85 (bridged methylene); 6.59 and 6.73 (aromatic H); 3.28 (CH3O); 3.60 (H next to CH3O); 4.28 (H next to OH); 5.56 (double bond H) 1.63 and 2.30 (aliphatic next to spiro) Deulofeu, 1947 m.p. 171-172 HI m.p. 240 HC1 (long prisms) m.p. 250 [aid +138.6 (EtOH, c=0.39) HI +110 (water, c=0.209) HC1 +165.10 water, c=0.333) Folkers, JACS, 1940 m.p. 170 [aid +145.5 c« 0.371, absolute EtOH) Folkers, JACS, 1942 UV (e) 293 (4.3-4,8) increases past 240 Glasby, 1975 m.p. 170-175 crystalizes as hemihydrate from Et20-pet ether Marion, 1952 m.p. 170-170.5 [aid 145.5 143

ERYTHRATINONE C18H19N04 Barton, 1968 m.p. 136-137 MW 313.130931 [ald +409 (c 0.35) IR vmax (CHCI3) 1675 UV (e) 297 (4800) and 230 (15,500) MS 313 (M+), 270, 255 (100), 227, 226, 199, 198 NMR CDCI3 at 100MHz 6.65 (s,17); 6.56 (s, 14); 5.9 (s, OCH2O); 3.98 (dd, 3); 3.51 (s. OCH3); 3.4-2.9 (complex 10 CH2s) Boar, JCS, 1970 MS m/z 313 (M+, 20%), 270, 255 (100%), 227, 226, 199, 198 Singh, 1972 m.p. 135-136 MS 313 (M+), 255, 227,226 ERYTHRAVINE C18H21NO3 Chunchatprasen, 1982 MS m/z (TMS) 371, 356. 282 (100) ...... 73 HPLC lichroprep EtOH/ CHCI3 40:60 Glasby, 1975 Millington, 1974 (seeds) MW 299.1520 (HRMS) yeUow oil UV (e) 225 (14,000) and 282 (3070) MS 299 (M+, 94), 282 (M-OH, 100); 280 (M-OH, 17); 266 (M-CH3O- H2 23 NMR (100 MHz) 6.88 and 6.68 (aromatic s); 6.57 (q =); 6.05 (d, =) 5.77 (s, =); 4.53 (m next to OCH3); 3.90 and 3.80 (s, OCH3, 1.75- 3.75 (complex -CH2- 8, 10, 11) Redha, 1983 NMR 220 MHz 5.98 (1), 6.57 (2), 4.51 (3a), 1.83 (4a), 2.54 (4e), 5.75 (7), 3.63 (8), 3.70 (7), 2.66 (10a), 3.55 (10c), 2.97 (11a), 2.98 (lie), 6.86 (14), 6.64 (17), 3.77 and 3.87 (aryl OCH3) ERYTHRINADIENONE Crabb, 1975 NMR in CDCI3: 3.85 (aromatic CH3O); 6,63 and 6.36 (aromatic H); 6.00 (H-OCH3) ERYTHRININE C18H19NO4 Crabb, 1975 NMR 5.94 (bridged methylene); 7.00 and 6.82 (aromatic H); 3.62 and 2.95 (ring C aliphatic), 3.32 (methoxy); 4.73 (H next to OH) Ghosh, 1972 m.p. 197-198 MS 313 (M+), 298, 283, 282 Glasby, 1975 m.p. 197-200 [aid +204 UV 209, 230, and 289 144

Ito, Chem Comm, 1970 m.p. 197-200 [a]d +204 (CHCI3) MS m/z 313 (M+), 298, 283, 282, 280 UV (log e) 209 (4.35), 230 (4.26) and 289(3.70) IR vmax 3560 cm-1 (OH) NMR 3.32 (3H, s, OCH3), 5.94 (2H,m, OCH2O), 6.82, 7.00 (2H, ss, para aromatic protons), 4.73 (1H, t, HOCH), 3.62, 2,95 (2H, pair of quartets), 6.00 (1H, broad doublet) and 6.60 (1H, broad double doublet) Ito, 1973 and 1976 m.p. 197-200 [aid +204 (c 1.0, CHCI3) IR (CHCI3) 3500 (OH) UV (log e) 209 (4.35) 230(4.26), 289 (3.70) NMR 60 MHz 3.32 (OCH3), 3.62 (q); 4.73 (t); 5.75 (br. s); 5.94 (m, OCH3O); 6.00 (br.d); 6.6 (br. dd); 6.82 and 7.00 (s, aromatics), ERYTHRISTEMINE C20H25NO4 Barton, 1970 m.p. 127-129 [aid+189 (c 0.4 in CHCI3) MS MW 343.1789 m/z 343 (M+), 328, 312, 311 (100%), 310, 296, 280 IR v max 1610 (no hydroxy or carbonyl) UV (e, EtOH> 283 (3100) and 235 (20,000) NMR in C6D0 7,09 and 6.93 (aromatic); 3.43, 3.43, 3.21, 3.02 (s, OCH3); 4.06 ( H at OCH3); 2.49 (q,4a); 2.08 (t, 4c); 6.32 (2); 5.88 ( 1) Barton, JCS, 1973 m.p. 127-129, pale yellow prisms [aid +189 (0.4, CHCI3) IR vmax 1610 cm-1 UV (log e, EtOH) 235 nm (4.3) and 283 (3.5) MS m/z 343 (M+), 328, 312, 311 (100), 310, 296, 280 NMR 7.09 (1H, s), 6.93 (1H, s), 6.32 (1H, dd). 5.88 (lH,d),5.38 (lH,s), 4.06 (1H, X of ABMX), 3.94 (1H, dd), 3.97 (1H, A), 4.67 (1H, B), 3.43 (6H,s)3.21 (3H, s), 3.02 (3H,s), 3.S-2.8 (2H, dd), 2.49 (lH.dd). 2.08 (lH.dd) Chunchatprasert, 1982 MS m/z (TMS) HPLC lichroprep EtOH/ CHCI3 10/90 NMR 400 MHz 6.01 (1), 6.58 (2), 4.09 (3), 1.79 (4a), 2.42 (4e), 5.71 (7), 3.96 and 3.78 (8), 3.43 (lOe), 3.28 (lOe), 4.12 (lie), 6.84 (14), 6.89 (17), 3.32 (OCH3), 3.56 (II-OCH3), 3.76 and 3.89 (aiyl OCH3) Crabb, 1975 NMR in CfeDfi; 7.09 and 6.93 (aromatoc H); 6.32, 5.88, and 5.38 (double bond Hs); 3.94 and 4.06 (H geminal with CH3OS); 2.08 and 2.49 (ring A aliphatic); 3.88 and 4.05 (aliphatic next to N); 3.43 (CH3OS) Glasby, 1975 m.p. 127-129 pricrate 145-150 [aid +189 UV 235 and 282 Jackson, 1985 NMR spectra included in paper (w/o values) Letcher, 1971 MS M+ (31), M-15 (20), M-30 (31), M-31 (96), M-32 (100), M-33 (78)k, M-47 (31) NMR 100MHz in C6D6 5.88, 6.32 and 5.38 (=, 1,2,7) 2.08 and 2.49 (-CH2-, 4), 4.06 and 3.94 (next to OCH3), 6.93 and 7.08 (aromatic H); 3.43, 3.43,3.21 and 3.02 (OCH3) ERYTHROCARINE Jackson, in Philipson, 1985 MS 283, 282 (M-H), 266 (M-OH), 264 (M-H2O-H) NMR 5.85, 6.59 and 6.72 (aromatic H) Jackson, Phytochem, 1985 MW 283.121 UV 235 MS 283 (M+), 282 (M-H), 266 (M-CH3), 264 (M-OH) NMR 6.72 and 6.59 (aromatic) 2.99, 6.48, and 5.71 (diffuse, =, 1,2, 7); 5.85 (methylene dioxy); 3.24 (m, 3); 1.78 (t, 4a); 2.46 (q, 4e) a-ERYTHROIDINE CI6H19NO3 (berteroana, costarricensis, poeppigiana) mp 236 [aid +89 (water) UV (log e) 233 (4.12); 254 (3.65) A233/254 = 2.95 pKa (80%MCS) 5.42 IR 770, 810, 845, 860, 890, 902, 935, 943, 980, 990, 1010, 1022, 1032, 1054. 1072. 1109. 1140, 1163. 1193. 1246, 1260, 1272, 1295, 1321, 1342, 1640, 1715 Aguilar, 1981 (flowers, americana) UV (log e, EtOH) 226 (4.18) [aid+123.85 (c 2.39, H2O) IR film v max cm-1 2840, 1730, 1090, 890, 650, NMR 60 MHz in CDCI3 6.43 (q, =, 2); 5.91 (d, =, 1); 5.80 (s, =, 14); 5.76 (m, =, 7); 4.46 and 4.03 (q, -CH2-, 17)3.5-3.96 (m, next to OCH3, 3); 3.65 (s, -CH2-, 8); 3.4 (s, OCH3); 2.33-3.23 (m, -CH2-, 10, 11, 12); 1.66-1.93 (m, -CH2-,4) Barton, JCS, 1973 (leaves of poeppigiana) m.p. 220-221 Boar, JCS, 1970 MS M+ (83%), M-CH3, M-CH3O (100%), M-CH5O, M-C3H8O, M- C4H7O, M+ C5H8O Boekelheide, 1953 m.p. 58-60 beautiful white crystals HC1226-228 told +136 (c 0.5% in water) IR HC1 5.85 p (unsaturated lactone ring) and approx. 4.1, 7.0, 8.35, 9.1, 9.5, 11.4, 11.6, 13.0 (spectrum shown) UV (log e, EtOH) 224 (4.5) 146

Boekelheide, JCS, 1953 (alpha HCL) (seeds of berteroana and poeppigiana) m.p. 58-60 HC1 m.p. 226-228 IR approx v max 5.85, others 3.0, 4.1 9.1, 11.2, 11.9 UV (log e) 224 nm (4.5) (aid HC1 +118 (0.5% in water) base +136 Chawla, 1982 (seeds and leaves of berteroana) glass IR vmax 1720 (lactonic CO) UV (log e) 227 (4.18) MS 273 (M+, 72); 258 (M-CH3, 34); 242 (M-OCH3), 100) NMR in CDCI3 5.88 (d. 1); 6.38 (dd. 2); 3.76 (m, 3); 1.69 (q. 4a); 2.57 (q, 4c); 5.75 (br s. 7); 3.60 (s, 8); 2.83 (m, lOe); 1.75 (q, 11a); 3.25 (m, 12); 5.75 (s, 14); 4.03 (q, 17a); 4.42 (q, 17e); 3.35 (OCH3) Glasby, 1975 m.p. 94-95 Jackson, Allertonia, 1982 (leaves of poeppigiana and berteroana) CH3rck Index, 1968 m.p. 58-60 [aid +136 (c=0.5 in water) Millington, 1974 (seeds of berteroana) Mondon, 1970 m.p. 58-60, no color with sulfuric acid or ferric chloride, sensitive to air, stable as salt [a]d +136 (water) UV (log e) 224 (4.5) IR 5.78 (1730 cm-1) Redha, 1983 (leaves of poeppigiana) Sangster, 1965 UV lmax HC1 (EtOH) 229 (log e 4.54) and 232 (sh, 4.46) Smith Kline and French Laboratories m.p. 236 [aid +89 (H20) IR vmax 1715 cm-1 others none 3200-3600 and 1193, 1109, 1054, 980, 860, 770 UV (log e) 233 (4.12) pK(80%mcs) 5.42 fl-ERYTHROIDINE C16H19NO3 Aguilar, 1981 (flowers, americana) m.p. 100 methiodide 219-220 MS 273.1359 (M+) IR vmax (KBr) in cm-1 2810, 1720, 1090, 810, 645 IR vmax (KBr) in cm-1 2820, 1740, 1110, 840, 810 NMR 60 MHz ir CDCI3 6.43 and 5.85 (q, «, 2, 1); 5.71 (m, =, 7); 4.63 (s, -CH2- next to ether); 4.1 (m, next to OCH3); 3.58 (m, - CH2- next to N); 3.38 (s, -OCH3); 2.35-3.23 (m, 14, 10, 11) Barton, JCS, 1973 (leaves of poeppigiana) m.p. 222.5 Boar, JCS, 1970 MS M+ (84%), M-CH3, M-CH3O (100%), M-CH5O, M-C3H8O, M- C4H7O, M+ C5H8O 147

Boekelheide, JACS, 1953 (beta) (seeds of berteroana and poeppigiana) m.p. HC1 228-231 base 99-100 fold +85 HC1+107 IR vmax 5.78 p. (carbonyl of d-lactone) and approx 3.3, 4.15, 9.1, 9.4, 10.2, 12.2, 14.5 (spectrum shown) UV (log e) 238 (4.4) Chawla, 1982 (seeds and leaves of berteroana) glass IR vmax 1720 (lactonic CO) UV (log e) 230 (4.16) MS 273 (M+, 63); 258 (M-CH3, 32); 242 (M-OCH3, 100) MS of TMS deriv. 345 (M+, 40); 330 (M-CH3, 17); 314 (M-OCH3, 40); 224 (65); 196 (75), 130 (100); 73 (64) NMR 5.90 (d, 1); 6.44 (dd, 2); 4.11 (m, 3); 1.79 (q, 4e); 2.55 (q, 4e); 5.71 (s, 7); 3.63 (d, 8); 3,17 (q, 10a); 3,22 (q, lOe); 1.68 (m, 11a); 2.40 (m, lie); 2.98 (d, 14a); 3.07 (d, 14e); 4.57 (d, 17a); 4.77 (d, 17e); 3.38 (OCH3) Folkers, CA 37:5408 (seeds of berteroana and costarricensis) m.p. 228 Folkers, JACS, 1937 ( americana seeds) m.p. HC1 228-229 base 94-96 [a]d HC1 +109.7 (c=0.501 in H2O) MW 273 Glasby, 1975 m.p. 98.5-99.5 [aid +88.8 Jackson, 1985 (seeds of berteroana) Jackson, AUertonia, 1982 (leaves of poeppigiana and berteroana) Marion, 1952 (seeds of berteroana and costarricensis) m.p. 99.5-100 [aid +88.8 CH3rck Index, 1968 m.p. 99.5-100 [a]d +88-8 soluble in water, benzene, chloroform, methanol, EtOH Millington, 1974 (seeds of berteroana) Mondon, 1970 m.p. 99-100 characteristic color with sulfuric acid or ferric chloride [a)d +89 (water) UV (loge) 238 (4.4) IR 5.78(1730 cm-1) Redha, 1983 (leaves of poeppigiana) Sangster, 1965 UV (log e, HCl-EtOH) 232 (4.4). 240 (4.4), 245 (sh, 4.27) and 264 (sh, 3.28) 148

HYDROXYERYSODINE (Cll) C18H21NO4 Chunchatprasen, 1982 IR 3600 UV (log c, EtOH) 280 (378) 230 (280) for beta OH 280 (3424) 230 (11640) for alpha OH MS m/z (TMS) 459, 444, 428 (100), 369, 338 ..... 73, 315 (M+), 300 (-15)k 284 (-31, 100), 266 (-49) HPLC Lichroprep EtOH/CHCl3 10/90 NMR 400 MHz (beta) 6.0 (KK 6.58 (20< 4.03 (3a), 1.79 (4a), 2.40 (4e), 5.74 (7), 3.97 and 3.85 (8), 3.56 (lOe), 3.10 (10a), 4.66 (lie), 6.82 (14), 7.0 (17), 3.31 (OCH3), 3.78 (aryl OCH3) NMR 400 MHz (alpha) 6.02 (1), 6.55 (2), 4.22 (3a), 1.85 (4a), 2.70 (4e), 5.70 (7), 3.66 and 3.70 (8), 3.45 (lOe), 3.18 (10a), 4.93 (1 la), 6.85 (14), 7.09 (17), 3.35 (OCH3), 3.78 (aryl OCH3) HYPAPHORINE C14HI8N2O2 Deulofeu, 1947 m.p. 230-235 Folkers, 1939, JACS m.p. 231-232 free base 236-237 [a]d +89.6 c=0.502 in H2O base +113.1 c=0.516, H2O Folkers, JACS, 63, 1943 (seeds of costarricensis) found in every species so far examined Langes Handbook large monoclinic cyrstals/aqueous, very soluble in water and alcohol m.p. 255 (anhydrous) Marion, 1952 (seeds of costarricensis) Sarragiotto, 1981 m.p.247.8-248.8 HC1 233.1-234.4 IR 3200, 1640 NMR (100 MHz, CF3COOH) 7.77-7.30 (6 aromatic protons), 4.53 (t, 1), 3.65 (d, 2), 3.47 (s, N(CH3)3) ISOBOLDINE C19H21NO4 Barton, JCS, 1973 m.p. HC1 118-123 Chan, 1966 amorphous white powder, m.p. 123-135 UV (log e, EtOH) 280, 302 (4.03, 4.05) IR (Nujol) 3300-3500, 1608, 1585, 1515s, 1415, 1335, 1315, 1280s. 1250s, 1110, 1080 NMR 8.05, 6.85, 6.60 (s, aromatics) Clezy, 1966 m.p. as pinkish needles or colourless plates 120-124 [aid 41.3 (c, 0.7, EtOH) UV (log e) 280 (4.11), 305 (4.15) NMR 2.50 (NCH3), 3.85 (2xOCH3), 2.6-3.3 (m, complex aliphatics), 6.49 (3), 6.77 (8), 7.98 (11) Jackson, Allertonia, 1982 (leaves of poeppigiana) 149

ISOCOCCULIDINE Praskash 1988 NMR 400 MHz 6.03 (m, 1), 5.91 (m, 2), 3.53 (m, 3), 1.75 (m, 4a) 1.98 (m, 4e), 2.76 and 1.85 and 2.17 (m, 7), 2.79 and 3.20 (m, 8), 2.64 (m, 10a), 3.38 (m, lOe), 6.76 (m, 14), 6.73 (m, 16), 7.09 (m, 17), 3.26 and 3.76 (OCH3) MELANACANTH1NE C18H20NO3 Redha, 1983 m.p. 156-157 small needles crystalized from a mix of equal parts of EtOH/H20 IR 1720 cm-1 (ester carbonyl) UV (log e, EtOH) 230 (1870), 265 (infl, 562) HPLC EtOH/CHCl3 5:95 MS (TMS) 312 (M+), 297 (-15), 282, 281 (-31, 100), 269, 221... NMR 400 MHz 6.08 (1), 6.59 (2), 3.97 (3a), 1.90 (4a). 5.75 (7), 3.54 a d 3.74 (8), 3.54 (lOe), (3.07 (10a), 2.76 (lie), 3.07 (11a), 8.00 (14), 8.51 (17), 3.33 (OCH3), 3.96 (OCOCH3) METHOSYERYTHRATINE Chawla, 1987 MS (TMS) 417 (M+, 100); 329 (-88, 40); 328 (-89, 52) and 73 (94) NMR 360 MHz (CDCI3) 5.75 (d, 1); 4.31 (m, 2); 3.59 (m, 3); 1.77 (t, 4e); 2.19 (dd, 4e); 2.95 and 2.6 (m, 8), 4.53 (t, 11), 6.47 and 6.64 (s, 14/17); 3.40 and 3.36 (OCH3) METHOXYERYSODINE (Cll) C19H23NO4 Chunchatpraseit, 1982 UV (log £, EtOH) 230(1711), 280(658) IR 3560 cm-1 (OH) MS m/z (TMS) 401, 386, 370 (100),354, 339 .... 73, 329 (M+), 314 (-15), 298 (-31), 297 (-32, 100). 282 (-47), 266 (-63) HPLC Lichroprep EtOH/CHCl3 10/90 NMR 400 MHZ 6.0 (1), 6.58 (2), 4.05 (3a), 1,78 (4a), 2.43 (4e), 5.71 (7), 3.96 and 3.77 (8), 3.3 (lOe), 4.08 (lie), 6.81 (14), 6.96 (17), 3.32 (OCH3), 3.53 (11 OCH3), 3.77 (aryl OCH3) Jackson, 1985 MS 329 (M+), M-15, M-31, M-32 (-CH3OH), M-47 (-CH3 and CH3OH), M-63 (M-OCH3 and CH3OH) NMR spectra shown w/o values METHOXYERYSOPINE (Cll) Chunchatprasert, 1982 MS m/z (TMS) 459, 444, 428 (100), 427 ..... 73 Redha, 1983 MS m/z (TMS) 459 (M+) , 428 (-31), 427, 396, 329, ....73 (100) METHOXYERYTHR ALINE Chunchatprasert, 1982 MS m/z (TMS) 327, 312, 296 (100), 294, 264 HPLC Lichroprep EtOH/CHCl3 10/90 or PartisU CHCI3 NMR 400 MHz 6.0 (1), 6.54 (2), 4.02 (3), 1.86 (4a), 5.72 (7), 3.81 and 3.81 (8), 3.18 (10a), 3.46 (lOe), 4.15 (lie), 6.78 (14), 6.91 (17), 3.35 (OCH3), 3.56 (II-OCH3). 5.89 and 5,95 (O-CH2-O) 150

Letcher, 1971 MW 327.1461 (HRMS) pale yellow gum ta]d +199 IR vmax (film) 1616 cm-1 UV (log e) 227 (4.15) and 286 (3.5) MS M+ (24), M-15 (16), M-30 (22), M-31 (90), M-32 (100), M-33 (55), M-47 (32) MS 327, 312, 297, 296, 295 (100), 294, 280 NMR in C6D6 100MHz 5.80, 6.32, and 5.37 (« 1,2,7); 4.95 and 3.94 (next to methoxy, 3, 11); 2.1 amd 2.48 (CH2, 3); 3.99 and 7.04 (aromatic); 3.17 and 3.01 (OCH3); 5.33 (O-CH2-O) NMR in CDCI3 well resolved signals 4.08 (t) METHOXYERYTHRALINE-N-OXIDE Chawla, 1987 MS 343 (M+, 2), 327 (-16, 44), 312 (-31. 21), 296 (-47, 100), 295 (- 48. 94), 294 (-49, 68), 280 (46) NMR 360 MHz (CDCI3) 6.18 (d, 1); 6.65 (dd, 2); 4.19 (m, 3); 3.09 (t, 4a); 2.08 (dd, 4e); 5.79 (br. s, 7); 5.09 (d, 8); 4.71 (m, 8); 4.71 (m, 10a); 3.91 (dd, lOe); 4.50 (d, 11), 6.55 and 6.86 (s, 14/17), 6.03 and 5.97 (OCH2O); 3.63 and 3.41 (s, OCH3) NORORIENTALINE C18H21NO4 OR Cj8H22N04Br Barton, JCS, 1973 (leaves of poeppigiana isolated as HBr salt) m.p. 257 from CH30H-Et20H (HBr salt) free base m.p. 145-147 [aid +42 (c 0.18 in chloroform) UV (e) 285 (8800) and 231 (16,500) in NaOH-EtOH 301 (10,500) and 247 (19,500) v max (nujol) 3300, 3200, 3100, 1605, and 1525 cm-1 NMR (TFA) 7.4 (2, NH2+), 7.87-7.63 (5, arotnatics), 4.72 (l.m), 3.89 (3, s), 34.34-3.12 (6, complex aliphatics) MS m/z 315 (M+ -HBr), <78 (100%), 163, and 137 Ito, 1973 and 1976 m.p. 104-106 HCL 158-160 [aid -41.3 (c 0.46, CHCI3) IR (CHCI3) 3550 (OH) UV (log e, EtOH) 210 (4.60), 228 (4.12 ), 286 (3.83) MS 315 (M+), 179 (100), 178, 177, 163 NMR 60MHz (CDClv) 3.82 and 3.83 (OCH3), 6.93-6.63 (4 aromatic H), 6.57 (s, aromatic) NORPROTOSINOMENINE C18H21N04HCI Banon, 1968 m.p. HC1 241-242 colorless needles MW base 315.147048 IR vmax 3350, 3100, 2720, 2660, 2620, 2555, 1605 cm-1 UV (e, EtOH) 286 (7100) and 222 (infl 13,800) MS 315 (M+), 178 (100), 163, 137 NMR CDCI3 at 100 MHz 9.99 (br s. NH/OH) 9.85 (s, NH/OH); 6.79 (br s, 2, 4, 5); 6.63 (s, 8); 6.35 (s, 5) 4.49 (br m, 1); 3.77 (s, OCH3); 3.57 (s, OCH3); 3.92-2.30 (complex methylene +20H/NH) 151

Ghosal, 1971 m.p. 242-244 colorless needle:: [aid +28 (c 0.88 in water) UV (log e) 222-225 (hump 4.13) and 284-285 (3.84) IR v max (base) 3342, 3304, 1611 MS 315 (M+,<1%), 178 (100), 163 (32), 137 (16), 122 (9), 94 (6) NMR 8.98 (broad s, NH); 8.87 (s, OH); 6.76 (broad, s, 3H), 6.62 (s, 1H), 6.36 (s, 1H); 4.50 (broad m, 1H); 3.72 (s, OCH3). 3.69 (s, OCH3), 2.5-4.1 (comeplex, 8H) ORIENT ALINE C20H25NO4 Barton, JCS, 1973 [a]d -35.5 (c=0.2 in CH32CO UV (c, NaOH-EtOH) 310 (e 6900) and 252 (e 14,800) NM R in D-acetone 6.83br (2, s), 6.61 (1H, d), 6.48 (1, dd), 6.02 (1, s), 5.1 (1, dd)) 3.78 (3, s), 3,72 (3, s), 3.07br (3,s) 3.33br (3,s) and 3.9-3.1 (6, complex aliphatics) MS m/z 343 (M+), 312, 206, 192 (100%), 177, 149, and 137. MW 343.1775 Ito, 1973 NMR 60 MHz 2.48 (s, NCH3), 3.79 and 3.85 (s, OCH3), 6.43 (s, aromatic) OXOERYSOVINE at C8 C18H19NO4 Chunchatprasert, 1982 UV (e, EtOH) 230 (2504) 284 (1052) IR 1670 cm -1 or 1675 MS m/z 313 (M+, 100), 298 (-15), 282 (-31), 280 (-33) HPLC Lichroprep EtOH/ CHCI3 10/90 NMR 400 MHz 6.29 (1), 6.84 (2), 3.82 (3), 1.63 (4a), 2.74 (4e), 5.97 (7), 3.94 (lOe), 3.57 (10a), 3.04 (11a), 2.93 (lie), 6.80 (14), 6.69 (17), 3.29 (OCH3), 3.86 (aryl OCH3) OXOERYTHRALINE at C8 C18H17NO4 Chunchatprasert, 1982 IR (CHCI3) 1670 cm-1 MS m/z 311 (M+, 100), 296 (-CH3,41), 280 (-OCH3, 58),294 (-47) HPLC Partis il CHCI3 and Lichroprep EtOH/CHCl3 10/90 NMR 220 MHz 6.34 (1), 6.91 (2), 3.80 (3), 1.70 (4a), 2.81 (4e), 6.06 (7). 3.65 (10a), 3.91 (lOe), 3.02 (lie), 3.12 (11a), 6.88 (14), 6.85 (17), 3.37 (OCH3), 5.92 and 5.96 (O-CH2-O) Ito, 1976 IR (CHCI3) 1665 cm-1 (C=0) MS m/z 311 (M+) NMR (CHCI3) 6.91 (dd), 6.33 (br d), 6.78 (s), 6.75 (s), 5.93 (dd), 3.35 (s) Jackson, 1985 MS 311 (M+), 295 (M-15), 280 (M-31), 278 (100%), 266, 258, 225, 213, 189, 167, 149, 126, 105, 91, 71, 58 152

Mantle, 1984 MS 311 (M+, 100), 296 (-15,50), 280 (-31, 65), 279 (33), 278 (76), 268 (15), 250 (30) HRMS 311.1163 NMR (CDCI3) 6.30 (dd, 1), 6.87 (ss, 2), 3.78 (Odd, 3), 1.67 (ddd,4a), 2.80 (ddd. 4e), 6.0 (s, 7), 3.87 (ddd, 10a), 3.62 (ddd, lOe), 3.10 (ddd, 11a), 2.96 (ddd, lie), 6.74 (s, 14), 6.72 (s, 17), 3.33 (s, OCH3), 5.91 and 5.92 (OCH2O) OXOERVTHRATINONE at C8 CI8H17NO5 C18H17NOS Chunchatprasert, 1982 UV (e, EtOH) 230 (2780) 285 (1250) MS m/z 327 (M+), 241 (-15, 100), 212 (-115) HPLC Lichroprep EtOH/ CHCI3 10/90 NMR 400 MHz 7.30 (1), 3.49 (3a), 1.68 (4a), 2.59 (4e), 4.12 and 3.63 (7), 3.58 (10a), 3.72 (lOe), 3.09 (11a), 2.92 (lie), 6.48 (14), 6.66 (17), 3.40 (OCH3), 5.93 and 5.90 (O-CH2-O) OXOERYTHRININE C18H17NO5 Dagne, 1984 m.p. crystals 200-201 MW 327.1111 (HRMS) [aid +100 (CHCI 3 c=0.35) IR 3460 (OH), 3040, 2960, 1680 (C=0), 1505, 1485, 1250, 1100, 1040, 930, 870 UV (e, EtOH) 357 (1022), 272 (6740), 238 (12,600) MS 327 (M+, 93); 312 (72); 296 (85); 295 (68), 294 (100), 282 (49), 238 (64) NMR (400 MHz, CDCI3) 1.67 (dd, 7a); 2.67 (br. dd, 4e); 3.66 (dd, lOe); 4.22 (dd, 10a); 4.84 (br t, 11); 5.92 and 5.96 (OCH2O); 6.03 (s, 7); 6.34 (br d, 1); 6.83 (s, 14); 6.85 (br dd, 2); 7.02 (s, 17) OXO-a-ERYTHROIDINE Chawla, 1982 (seeds and leaves of berteroana) m.p. 183 crystals from acetone [aid+137.9 (c 0.116, EtOH) UV (log e) 222 (4.19) and 253 (4.16) IR vmax 1700 cm-1 (CO); 1745 cm-1 (5 membered lactam CO) MS 297 (M+, 100); 272 (M-CH3, 17); 256 (M-OCH3, 47) NMR 6.25 (d, 1); 6.74 (dd, 2); 3.75 (m, 3); 1.62 (q, 4a); 2.57 (q, 4e); 6.02 (s, 7); 3.11 (m, 10a); 4.35 (dq, lOe); 1.45 (dq, 11a); 1.93 (m, lie); 3.03 (m, 12); 5.83 (d, 14); 3.97 (q, 17); 4.46 (q, 17e); 3.41 (OCH3) OXO-B-ERYTHROIDINE Chawla, 1982 (seeds and leaves of berteroana) glass MW 287.1146 (HRMS) [aid +32 (c 3.156, EtOH) UV (log e) 254 (3.83) IR vmax 1700 cm-1 (CO); 1745 cm-1 (5 membered lactam CO) MS 287 (M+, 100), 272 (M-CH3, 27); 256 (M-OCH3, 46) MS of TMS deriv. 359 (M+, 10); 344 (M-CH3, 16); 242 (62); 227 (78); 73 (100) NMR 6.25 (d, 1); 6.77 (dd, 2); 4.08 (m, 3); 1.72 (q, 4a); 2.82 (q, 4e); 5.96 (s. 7); 3.22 (dq, 10a); 4.34 (dq, lOe); 2.02 (br dd, 11a); 2.40 (m, 1 le); 3.10 (d, 14 a and e); 4.59 (d, 17); 4.73 (d, 17e); 3.42 (OCH3) Jackson, 1985 (seeds and leaves of berteroana) OXOMELANACANTHINE at C8 C18H18NO4 Redha, 1983 IR 1720 cm-1 (ester carbonyl), 1677 (carbonyl) UV (e, EtOH) 232 (1478), 256 (1174) MS 326 (M+), 311 (-CH3), 295 (-OCH3), 294 (-OCH3, H2), 283 (- C2H3O), 269 .... NMR 400 MHz 6.42 (1), 6.93 (2), 3.77 (3a), 1.79 (4a), 2.75 (4e), 6.04 (7), 4.17 (lOe), 3.61 (10a), 3.15 (lie,a), 7.95 (14), 8.62 (17), 3.33 (OCH3). 3.96 (OCOCH3) OXOMETHOXYERYTHRALINE

GAS CHROMATOGRAPHIC AND MASS SPECTRAL DATA OF SELECTED ERYTHRINA ALKALOIDS DESCRIBED IN THIS STUDY

154 254 m/z 3 l (17.649 Pin) 254 35 239 27 227 22 ■ 000 223 100 221 29 7 000 198 27 6000 196 37 5000 183 20 130 182 30 4000 170 21 169 19 3 000 167 16 130 43 63 18 iooo -] 83.5 20 354 63 11 51 14 50 100 150 200 2 50 300

ERYB1DINE m/z Jfc 344 21 343 100 Vlbundanca (26.7 16 Bin) 328 31 9000 312 5 300 10 ■ 000 287 18 286 70 7 000 286 285 64 271 17 6000 269 13 5000 255 257 12 4 000 - 256 26 5B 255 45 • 42 IIS 254 17 3000 225 12 2000 - 211 11 3 00 197 9 IOOO - 165 8 o - jllijL lll, 153 8 H/Z -> ______SO______100 150 200 250 300 139 6 115 9 71 11 58 35 156

ERYSODINE m/z _9fe Abundance (20.754 Bln) 300 9 299 47 4 50000 285 7 400000 284 38 269 26 350000 268 100 300000

266 15 350000 254 8 253 10 200000 241 9 150000 228 10 215 19 100000 215 130 19 50000 130 , 91 9 hMtW A, Jk . | 77 8 4/Z -> 20 40 60 00 100 120 140 160 150 300 220 240 360 350 300

ERYSODINE-TMS

votz 372 11 \bundam (21.167 Bin) 371 38 500000 357 10 356 36 700000 341 33 600000 340 100 338 10 500000 310 11 308 7 400000 300 7 371 287 12 300000

257 5 200000 223 3 147.5 4 100000 3??310 146.5 6 1920023 130 4 0 127.5 5 73 17 ERYSOVINE

Ibundanco (22.594 Bin) mlz Jfe 300 9 299 48 9000 - 285 7 0000 - 284 34 269 24 1 0 0 0 ■ 268 100 266 14 6000 - 254 7 5000 ■ 253 8 241 8 4000 228 8 215 11 3000 ■ 130 6 2000 - 91 7 215 77 8 1000 - 77 103 2739 65 I I J 151167 ' 126341 0 ■ lu VjujU jjiymiii J mn 1/Z -> 50 100 150______200

ERYSOVINE-TMS m/z _2b 372 11 371 40 (Eunoanci (21.314 357 10 356 32 50000 341 32 340 100 40000 338 9 310 10 308 7 10000 - 300 6

287 8 20000 - 257 5 223 3 147.5 3 10000 146.5 7 91103 130 4 U j L 127.5 5 73 20 158

ERYTHRALINE m/z 3k 298 9 Mt>u nilanc* (21.229 297 44 1 6 282 35 14 0000 - 267 24 266 100 120000 264 21 252 7 100000 237 11 226 14 80000 213 22 297 180 6 60000- 154 4 40000• 152 4 212 131.5 4 130 4 20000 - 77 102 120*52 . , 102 4 254 446 'f I ~r 91 4 30 100 ISO 200 250 2 00 250 4 00 77 4

ERYTHRATIDINE m/z 3k 331 5 (24.210 Bin) 300 12 4bund»nce 284 3 273 71 5000 271 272 19 258 23 257 100 4000 256 37 244 56 3 0 0 0 - 242 14 230 10 228 7 2000 214 8 28 180 7 1000 300 12Q 165 7 i 677 7 *5 311 148 5 l l 120 8 A/Z -> 90 100 1< 190 200 250 300 91 7 77 8 159

ERYTHRATINE

m/z Jfe 315 13 298 2 pibundanca {34.947 ■ITT 284 12 1000000 258 9 257 59 900000 256 3 900000 242 16 239 241 100 700000 240 71 600000 228 63 500000 214 10 213 10 400000 198 5 300000 180 12 300000 149 11 190 I 148 10 100000 ; A‘I ^ I i'J| t Ji | if' i1 115 8 100'130 140 160 190 200 220 340 360 280 300 91 11 77 13

ERYTHRATINE-TMS

m/z Jfe 387 3 372 5 \bundance (33.533 ■in) 356 5 331 4 600000 330 14 329 56 500000■ 328 18 382 3 400000 266 5

252 8 300000 - 242 16 241 100 300000 240 58 228 3 227 1 100000 - 226 7

213 8 1/Z -> 89 9 73 40 160

B-ERYTHROIDINE m/z _2ft 274 10 273 55 258 243 242 100 350000 241 240 300000 228 214 198 273 196 200000 184 150000 182 130 170 77 100000 168 65 170 19B 21 * 156 50000 156 154 103 144 IB 130 20 40 60 BO BO 200 220 240 260 128 117 14 103 8 91 20 77 25 OXO-B-ERYTHROIDINE pVz 3 l 287 100 286 23 hbundance (29.377 272 33 256 54 9000 - 255 78 254 33 8000 - 244 17 228 33 7000 - 227 31 6000 226 30 214 22 5000 - 197 26 184 50 4000 - 183 55 3000 ■ 182 56 170 30 2000 169 28 128 23 1000 - 117 28 o■ 91 34 1/Z -> 40 60 BO 100 170 140 160 IBP 200 220 240 240 2B0 77 58 VITA

Lori Denise Payne was bora in Stockton, California to Walter Payne and Susan Rood Payne. She grew up in the Southwest and attended Rio Americano High School in Sacramento California. At Rio, Dr. Payne was involved in sports, especially volleyball and badminton. She won the Best-All-Around Student award during her senior year in high school. She attended the University of California at Davis and graduated from there with a B.S. degree in Environmental Biochemistry. After graduation from UCD, Dr. Payne began work at Stauffer Chemical Company in Richmond, California where she was a member of a pilot group called the Biochemical Design Group. In 1985, Dr. Payne joined the Peace Corps and was assigned to Costa Rica. She worked at the Centro Agrondmico Tropical para Investigacidn y Enseflanza (CATIE), Turrialba, Costa Rica, where she performed dual roles as a Forestry Extensionist and Research Assistant in the Nitrogen Fixing Tree Project. During her Peace Corps assignment, Dr. Payne met and married Mr. Alejandro Dittel of Turrialba, Costa Rica. They both returned to the United States in 1987 so that Dr. Payne could pusue a doctoral degree at Louisiana State University. As a graduate student in chemistry, Dr. Payne worked on the technical part of a project that was developed while she was a CATIE. In January of 1989, Serra Raquel Dittel joined the Payne-Dittel family. Dr. Payne graduated from LSU with a Ph.D. degree in chemistry with a specialization in analytical chemistry and a minor in environmental studies. She has joined Merck Sharp and Dohme Research Laboratories in Rahway, New Jersey as a Senior Research Chemist

161 DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidatei Lori Denise Payne

Major Fieldt Chemistry

Title Dissertation! -phe Alkaloids of Erythrina: Clonal Evaluation and Metabolic Fate

Approvedi

Q w S '. i * h u - - Major Professor u n Cbai

Dean of the Graduate School

EXAMINING COMMITTEE:

^ A . ' ______

.^ z r £U-

Date of Examination;

< ! / / / ? /