D i o s g e n i nP r o d u c t i o n i nNorth America

A Brief History

22

Ray F. Dawson1

Additional index words. steroid hormones, cortisone, fertility regulants, synthesis intermedi-ates, progesterone, Mexican yam, Dioscorea cultivation, low-tension soil water, stigmasterol,sitosterol

Summary. Diosgenin is a steroidal aglycone occurring in certain species of Dioscorea nativeprincipally to eastern Mexico. In the 1940s, diosgenin became a much-sought-after intermedi-ate for the chemical synthesis of certain corticosteroids and structurally related fertilityregulants. Various difficulties of access to native sources led to attempts at plantation produc-tion. One of these, supported by the Upjohn Company between 1962 and 1980, was locatedon the Pacific coast of Guatemala and is described herein from the standpoint of technologydevelopment. The Dioscorea plant produces a long, coarse vine that requires support. Thedeep-growing, fleshy rhizome contains the diosgenin and, at harvest, must be removed fromsoil depths up to 1 m. Dry rhizome yield depends on supply of readily available (low-tension)soil water. Sites located over abundant water reserves give satisfactory rhizome yields, butdiosgenin concentrations fall to uneconomically low levels under such circumstances. By 1980,diosgenin had been displaced by two products of soya oil processing, stigmasterol and sitos-terol, which became available as a result of advances in microbial fermentation technology.Consequently, the cultivation of Dioscorea was abandoned.

The recent history of diosgenin pro-duction and use is marked by anunusual convergence of interestsranging from industrial botany

through medicine, pharmaceutical chemia-try, business, and international politics. It is atale of extremes. Need for diosgenin aroseprecipitously in the 1940s, when no com-mercial source existed. It remained a favored

1Consultant in tropical agriculture (retired). 40 Palmer Avenue,Winter Park, FL 32789.

synthesis intermediate in the pharmaceuticalindustry for almost 40 years. During thoseyears, great quantities were extracted fromthe rhizomes (root or tuber in trade parlance)of Dioscorea spp. native to and collected fromthe wild in eastern Mexico. Various pressuresled to attempted cultivation of some of thesespecies on plantations. Perhaps the mostcomprehensive and long-lived effort was thatof the Upjohn Company in Guatemala. Un-der intensive cultivation, allowances had tobe made for support of the long (up to 30 m)woody vines and for removal of rhizomes

American Society for Horticultural Science

a

medi-terol,

ltive,ermedi-r~roduc-locatedlologyChelfrom:ension)butay 1980,sitos-

logy.

laceuticaling those:ted fromparlance)ctedfrompressuresIeof thesethe most1:wasthatmala.Un-:eshad to)to 30 m)rhizomes

:ural Science

from soil depths up to 1 m at harvest. Themost exacting site requirement was for acontinuous and abundant source of low-tension soil water. Eventually, diosgenin lostits preferred position, so far as the UpjohnCompany was concerned when stigmasteroland sitosterol became accessible microbiallyforthe preparation of avariety of commerciallyuseful steroid intermediates. As so oftenhappens, displacement of a reigning botani-cal resulted from advances in industrial mi-crobiology and in organic chemistry. But,while demand remained high, it would bedifficult to name a more-challenging projectin the field of industrial botany than was theattempted cultivation in the American tropicsof the Mexican yam, Dioscorea compositaHems!' (Matuda, 1953) and its variants.

Diosgenin is found in the st<?rageorgans(rhizomes) of certain species of Dioscorea thatoccur in Mexico and adjacent Guatemala.Chemically,diosgenin isthe steroidal aglycone(Fig. 1), released upon acid hydrolysis, of asaponaceous glycoside, dioscin. Dioscin-containing plant parts, when macerated inwater, impart a surfactant property that theindigens of Mexico and Central Americaemployed for cleaning purposes. The starchyrhizomes were eaten, albeit without relish, inseasons of maize crop failure (de Landa,1566). Dioscin is also hemolytic (Coursey,1967). Pre-Columbian inhabitants of the re-gion dispersed pulped rhizomes in slowlymoving streams, whereupon fish suffocated,floated to the surface, and could be retrievedeasilyby hand. The archeological museum atthe Tikal ruins in Guatemala contains anincised bone depiction of a Mayan fishingparty removing very limp fish from foam-covered waters. More recently, North Ameri-canpatent medicines of the late 19th and early20th centuries made use of dioscin-contain-ing extracts in the concoction of remedies for"female troubles," but uses still to comecould not have been foreseen by either stone-age Maya or our forebearers.

Diosgenin was isolated first from therhizomes of anAsiaticspecies,Dioscorea tokoroMakino (Tsukamotoetal., 1936). Fouryearslater, R.E. Marker (1940a, 1940b, 1943), ofthe Pennsylvania State College, publishedthe molecular structure and botanical occur-rences of this aglycone. Marker's interestwent far beyond the academic to the realm ofprescience. At that time, knowledge of thechemistry and physiology of mammalian hor-mones was expanding rapidly. Researchquantities were needed of those constituentsof the animal body that, up to that time, hadbeen known only in trace amounts. Shouldtherapeutic uses emerge, Marker reasoned,the pharmaceutical industry would be whollyunprepared to produce the bulk quantities

required. The small amounts then availablewere said to be under monopolistic controlby three European manufacturers (see SenateHearings, 1956). Only 3 years ahead of thecrisis of supply, Marker et al. (1947) deviseda synthesis of progesterone , the corpus luteumhormone, from diosgenin. On a hunch(Marker, 1987), he explored the Dioscoreaceaeof eastern Mexico, where he found substan-tial quantities of diosgenin-containing rawmaterial. His successes in both chemical andplant exploratory facets of this field have wonfor him far greater attention in Mexico(Lehmann et al., 1973)than they have in hisnative land. He ex-tracted diosgenin fromthe woody rhizomes ofthe "cabeza de negro,"Dioscorea mexicanaGuillemin, and pre-pared from it 3 kg of

.progesterone (Marker,1987), by far the larg-est quantity ever beforeseen. Marker had not long to wait. In 1949,two developments in medicine set off theexplosive demand that he had foreseen. Fromthe Mayo Clinic there came an announce-ment of the dramatic effect of administeredcortisone upon symptoms of rheumatoid ar-thritis. And, from the Worcester Inst. forDevelopmental Biology, came the first prac-tical fertility regulants. Although the earliestintermediates for cortisone synthesis were thesteroidal bile acids, both cortisone and thefertility regulants could be made, at that time,much more easily from diosgenin. Conse-quently, a world -wide search began for abun-dant and inexpensive supplies of intermedi-ates, including diosgenin, that could be usedadvantageously for the manufacture of desiredsteroid compounds (for two interesting ex-amples, see Landrum, 1986).

Total synthesis of steroids from thesimplest organic compounds was still years inthe future-and that route would be pro-hibitively expensive in any case. In the wordsof George Pucher, at one time of the Con-necticut Agricultural Experiment Station,New Haven, the plant world came to beregarded as a storehouse of rare chemicals,and it was ransacked from top to bottom insearch of ever more useful intermediates.Marker scored a third time by discovering,also in eastern Mexico, enormous populationsof Dioscorea composita Hems!., known locallyas "barbasco," a far more tractable source ofdiosgenin than was D. mexicana. Mexicanresources began to be "mined" by rhizomecollectors. Mexican authorities estimated thatthe states of Vera Cruz and Oaxaca alonecontained up to 25 t of dry root equivalent!

Fig. l.Diol8enin.

HortTechnology.Oct/Dec. 1991 23

km2, and that ≈ 5500 t of air-dried mate-rial was then being removed annually (SenateHearings, 1956). Giral (1958) stated that thediosgenin concentration in this materialranged between 3% and 4% on a dry basis,despite many more optimistic reports. As aconsequence of Marker’s work, and of con-tinuing improvements in chemical technol-ogy, the price of progesterone to manufac-turers of hormonal products fell from about$200/g in 1940, to $80 in 1943, and to 30¢in 1955 (Senate Hearings, 1956). Owing toits fortunate position with respect to thegeographical distribution of diosgenin-con-taining species, Mexico became the worldcapital of the steroid intermediates industry,a position that it held for more than 20 years.A parallel, although much smaller, industrysprang up in neighboring Guatemala that wasbased on the occurrence there of D. flori-

24 American Society for Horticultural Science

bunda Mart et Gal., a less vigorous speciespossessing a higher average diosgenin con-tent (≈ 5%) and recognized by the yellow fleshof its rhizomes (those of D. composita arewhite). Substantial quantities were processedby the Compania Agricola IndustrialGuatemalteca (CAIGSA).

Ultimately, Mexico lost its preferred po-sition as a consequence of several ongoingfactors. First, continued exploitation of natu-ral stands of Dioscorea led to ever highertransportation costs as distances betweenjungle and processing centers increased andas second-growth stands suffered in terms ofyield per unit area of terrain and of assay.Second, other intermediates were broughtalong, including stigmasterol and sitosterolfrom soya oils and hecogenin from sisal wastes.The third factor was Mexico’s attempt toprotect its national industry in a classic paral-lel to the earlier attempts of Peru to preserveCinchona bark production in the Andes andof Brazil to prevent loss of Hevea rubber toforeign planters. Political repercussions ofMexico’s efforts in this direction were airedbefore a Senate Subcommittee on the ‘Won-der Drugs” (1956) chaired by Sen. JosephO’Mahoney. Testimony before this groupaffords a broad insight concerning the technolo-gies, the botany, and chemistry, as well as thebusiness and international trade aspects of anew and turbulent industry.

Inevitably, commercial-scale cultivationof suitable Dioscorea spp. was contemplatedby those firms unable to gain unrestrictedaccess to Mexican raw materials and by Mexi-can firms concerned about the uncertainty offuture supplies. Edible yams of the samegenus had been cultivated for a very long timein Africa and the Indies, and considerabletechnology had accumulated (Coursey,1967). The earliest attempts to cultivate genin-containing yams in the western hemisphereare believed to have taken place in Guatemala(Bruhn et al., 1972), in Costa Rica and inPuerto Rico. One of the more systematic andcomprehensive efforts was that of the U. S.Dept. of Agriculture in Mayaguez and an-other at the Puerto Rican Experiment Stationin Rio Piedras. The first Guatemalan andCosta Rican attempts were terminated pre-maturely. Yield extrapolations from field plotsto commercial-scale plantings were publishedby Puerto Rican workers (Martin et al., 1966),with recommendations for cultivation. It isunderstood that Irving Sollins, one of themore knowledgeable witnesses before theO’Mahoney subcommittee, formed a com-pany to make use of Puerto Rican technolo-gies. However, the clay soils and inclementweather combined to make harvesting bothdifficult and costly. An earlier journal paper

(Kennard and Morris, 1956) had alreadyencouraged those who would attempt culti-vation under more favorable conditions forharvesting activities. In 1957, Percy L. Julian,well-known organic chemist and my one-time mentor in that subject at DePauw Univ.,came to New York to discuss possibilities forgrowing Dioscorea successfully in Guatemala,where I had some experience with the intro-duction of Cinchona. On the evidence avail-able, a rough estimate of 60% in favor ofsuccess was enough to set Julian on a courseofaction. For the purpose, he formed EmpresaAgro-Quimica Guatemalteca, S.A., where-upon I began a 20-year-long schedule ofcommuting between New York and Guate-mala.

Empresa’s first plantings were made latein 1958, employing material purchased fromB.A. Krukoff of Ensayos Agricolas, S.A.(Landrum, 1986). First harvests were disap-pointing. The project survived only becauseone field or block on the farm yielded at or evenabove our best expectation. We took this ex-ception as a measure of species potential and sodefined our task in terms of learning how toachieve the latter routinely. In 1962, the FineChemicals Division of the Upjohn Companyacquired the project, moved it to an area of low-bulk-density soils that were derived from recentvolcanic ash falls (see Dawson et al., 1980), andproceeded patiently to develop technology.This is designated herein as the central farm.Production objectives were set minimally at 18t·ha -l of dry rhizome material assaying 4%diosgenin, all within a 3-year crop cycle. Therewas reason to believe that these figures mightbe improved to as much as 28 t of dry materialwith a content of 5.5% to 6.0% diosgenin.Harvested crops on the new site gave on aver-age 4% of diosgenin, but dry rhizome yields fellto 50% of the minimum expected or less.Consequently, a three-tiered experimental ap-proach was devised to identify the limitingfactor(s) and means for rectification. Field plottests employing the usual management vari-ables failed to affect either growth rates ordiosgenin concentrations. Off-farm tests scat-tered from Florida to Costa Rica gave clearindication that this heretofore wild plant re-quired unusually great quantities of soil waterand that this water must be within reach of thelong coarse roots but not the rhizome meris-tems. A mini-scale greenhouse experimentalsystem that measured relative growth rates ofvery small tubers formed by young seedlingsnot only confirmed interpretation of the sitetesting outcome, but also gave an explanation.That was, D. composita rhizome growth ratesdepend linearly upon the supply of readilyavailable (low-tension) soil water (Dawson etal., 1980, 1983).

Three site categories well-suited to rapid

growth of Dioscorea rhizomes in terms of drymatter accumulation were identified by theabove means. The first occupied benches ofalluvium above the flood plains of rivers suchas the Motagua in Guatemala and the Aguanin Honduras. The second lay over relativelyshallow permanent water tables such as theblack sands below Tiquesate, Guatemala, orthe peats of the Florida Everglades. In bothplaces, water tables could be controlled bydry-season irrigation or by poldering andpumping. The third was characterized by afriable topsoil over a barrier to percolation.Old marine sediments (Bucul clay with 14%organic matter in the topsoil) near Retalhuleu,Guatemala, and an old lava flow from volcanoTurrialba in Costa Rica, were representative.The principal point is that subsoil profile withrespect to the storage and facile delivery ofwater to yam roots is much more importantto the growth of this crop than is topsoilclassification. Gross yields on each of thesesites came close to the initial objective. Forexample, 32 ha of cotton land below Tiquesategave 15 t of dry rhizome/ha in 3 years. In allsuch cases, however, diosgenin concentra-tion (assay) in the dry material fell to 3% orlower. Because of this totally unexpecteddevelopment, estimated diosgenin yieldequivalents at Tiquesate, 400 kg·ha-1 in 3 years,fell slightly below those (450-kg equivalents)on the central farm. This negative relationshipbetween vigorous rhizome growth and high

HortTechnology · Oct./Dec. 1991 25

Sapogenins. 69. Isolation andstructure ofthirteennew steroidal sapogenins. New sources for knownsapogenins. Amer. Chem. Soc. J. 65:1199-1209.

Marker, R.E., R.B. Wagner, P.R.. Ulshafer, E.L.Wittbecker, D.P.J. Goldsmith, and C.H. Ruof. 1947Sterols CLX. Sapogenins. 72. Steroidal sapogenins.Amer. Chem. Soc. J. 69:2167-2230. [See alsoUS. patents no. 2,352,852,4 July 1944, and no.2,420,489,13 May 1947, issued to RE. Marker.]

Marker, R.E. 1987. The early production of ste-roidal hormones. CHOC News of the Center forHistory of Chemistry 4(2):3-6. [A somewhatdifferent view occurs in what is purported to be aninterview with Marker by members of a Germantelevision crew that was printedwithout the author’sname in LA SEMANA, a weekly newspaper oncepublished in Guatemala City. A translation of thisarticle from the Spanish has been depositedwith theArchives of The Pennsylvania State Univ.]

Martin, F. W., E. Cabanillas, and M.H. Gaskins.1966. Economics of the sapogenin- bearing yam asa crop plant in Puerto Rico. Agr. Res. J., Univ.Puerto Rico (50(1):53-64.

Matuda, E. 1953. Las Dioscoreas de Mexico. An.Inst. Biol. México 24(2):279-390.

Senate subcommittee hearings of the Committee ofthe Judiciary, 5-6 July 1956.1957. “The WonderDrugs”. Senate Res. no. 167. U.S. Govt. PrintingOffice, Washington, D.C.

Tsukamoto, T. and Y. Ueno. 1936. Isolation ofdioscin and diosgenin from Dioscorea tokoroMakino. Pharm. Soc. Jpn. J. 56:802. [German, p.135-140. Chem. Abstr. 32:7470/1938.]

AcknowledgementsThe manuscript was read by Alexander W.Schneider, vice-president (retired), FineChemicals Division, the Upjohn Company,who sponsored the project; by Jeffery E.Shrum, Jr., (retired) formerly projectmanager, and by Lloyd W. Nystrom,formerly assistant manager and presentlyDirector of International Purchasing,Mallinckrodt Specialty Chemical Company.

HortTechnology · Oct./Dec. 1991 27