[Agr. Biol. Chem., Vol. 28, No. 4, p. 230•`233, 1964]

Two of Derivatives in Sweet Potato Roots Infected by Ceratocystis Fimbriata

By Takao MINAMIKAWA,TAKASHI AKAZAWA and IKUZO URITANI Laboratoryof BiochemistryNagoya University, Anjo, Aichi ReceivedJanuary 13, 1964

Two glucosides of coumarin derivatives were separated from sweet potato roots with black rot, by the combination of silica gel-coated thin layer chromatography and paper chromatography, and identified with and . The magnitude of synthesis of both bound was lower than that of the corresponding free coumarins, whereas they were detectable in neither cut nor fresh root tissues.

sues7). During the course of this work, we INTRODUCTION also tried to find out whether or not the Several workers have reported the produc fungus infection would induce the produc tion of coumarins in plant tissues infected by tion of the bound forms of the respective pathogenic microorganisms1•`4) Post-infec coumarins in the root tissues. Hughes and tion synthesis of (7-hydroxy- Swain4) reported that about 25-fold increase coumarin) and (6-methoxy-7-hy of scopolin (scopoletin-7-ƒÀ-) was in droxycoumarin) in sweet potato roots attack duced by the infection of Phytophthora in ed by the black rot fungus, Ceratocystis festans in potato tubers, whereas the chloro fimbriata, was reported by Uritani and Hoshi genic acid increase was only 2- to 5-fold in - ya5), and recently the third coumarin com the infected region. pound, esculetin (6,7-dihydroxycoumarin), This paper describes the chromatographic has been added to the list6). We have also identification of skimmin (umbelliferone-7-ƒÀ- carried out an analytical study on the synthe sis of the former two coumarins in the in glucoside) and scopolin in the sweet potato tissue with black rot. Silica gel-coated thin fected root tissues in the hope that this might layer chromatography (TLC) was found to reveal their possible role in disease resistance, be useful for the separation and identifica and found that their synthesis was most tion of these bound coumarins. marked in the fungus-penetrated root tis RESULTS AND DISCUSSION * Part 38 of Phytopathological Chemistry of Sweet Potato with Black Rot. Methods of the fungus inoculation on the 1) R. J. Best. Aunt, J. Exptl. Biol. Med. Sci., 22, 251 (1944), 26, sweet potato roots were described in the pre 223 (1948). 2) W. A. Andreae, Can. J. Res., 26C, 31 (1948). ceding papers7,8), The starting material of 3) W. H. Fuchs, Angew, Bot., 30, 141 (1956). 4) J. C. Hughes and T. Swain, Phytopath., 50, 398 (1960). the present experiment was the aqueous re- 5) J. Uritani and 1. Hoshiya, J. Agr. Chem. Soc. Japan, 27, 161 (1953). 7) T. Minamikawa, T. Akazawa and 1. Uritani, Plant Physiol., 38, 6) T. Minamikawa, T. Akazawa and 1. Uritani, Nature, 195, 726 493 (1963). (1962). 8) T. Akazawa and K. Wada, ibid., 36, 139 (1961). Two Glucosides of Coumarin Derivatives in Sweet Potato Roots Infected by Ceratocystisfimbriata 231 sidue of the coumarin extraction as reported to a filter paper (Toyo No. 51A), and de veloped by the ascending method using iso previously6). A hot 80% ethanol extract of the diseased sweet potato roots (about 1kg) amyl alcohol/acetic acid/water (4:1:2, by was concentrated, dissolved in 5% sodium vol.) as solvent. Their RF values were ex

carbonate solution, shaked vigorously with actly identical with those of authentic skim ethyl ether, and finally extracted continuously min and scopolin and no other fluorescent with ethyl ether at pH 6.5 to 6.8, leaving the compounds were detected. Each band was aqueous fractions. These fractions contain- extracted with ethanol and was subjected to ing water-soluble were pooled and (a) chromatographic identification (paper extracted three times more with ethyl acetate. and TLC) of skimmin and scopolin respec The total extract was concentrated in vacuo, tively using several solvent systems, and (b)

and the residue was finally dissolved in a ultraviolet spectral analysis by Cary Model 14 small volume of water. The aqueous mix Automatic Recording Spectrophotometer. ture was again extracted with ethyl ether to As will be seen in Table I (a), the R,, remove ether-soluble material. The final re values of I and II were perfectly in agreement sidue was applied to a silica gel TLC (15 with those of authentic skimmin and scopo •~25cm, 1mm in thickness), prepared as de lin, and their ultraviolet absorption spectra scribed in the preceding paper9). The chro were also identical with those of the respec matography was carried out in a closed cham tive authentic samples (Fig. 1). To identify ber, with 10% ethanol in ethyl acetate (v/v) the sugar and aglycone moieties of bands I as a solvent; four fluorescent bands were and II, acid hydrolysis was carried out as detected under ultraviolet lamp. Among follows: 0.7ml of 2% hydrochloric acid was them, the locations of the two fast-moving added to a small aliquot of each glucoside bands (I: RF 0.29, faint blue, and II: RF 0.19, and the mixture refluxed in a boiling water blue) were identical with those of the authen bath for 1 hour. The hydrolyzates were ex tic bound coumarins, skimmin and scopolin tracted several times with ethyl ether, and respectively, which were added as markers. each of the ethyl ether-soluble fraction and We have not investigated further into the residue was subjected to paper chromato nature of the other two compounds, although graphy and silica gel TLC for the identifica one of the later which remained almost at tion of the sugar and aglycone component. the starting line appeared to be chlorogenic In the latter case, the pH of the sample was acid from its fluorescent property and other adjusted after the hydrolysis. The spraying color reactions. To obtain a better separa reagent for the sugar was benzidine trichoro- tion of bands I and II, the plate was de acetic acid10) and the location of aglycone veloped two times more by the same solvent; was detected by its fluorescence under ultra- each zone was scratched from the plate with violet lamp. As shown in Table I (b) the a spatula, extracted with hot ethanol, con sugar liberated from bands I and II was centrated to a small volume in vacuo, and proved to be glucose. Also, the results show again applied to a silica gel TLC. For the that a multiple development could well dis final purification of the two compounds, the tinguish glucose from galactose. Table I (c) ethanol extracts of bands I and II from the shows the identities of the aglycones of I and second plate were applied as a streak (10cm) II as umbelliferone and scopoletin, respective-

9) T. Akazawa and K. Wada, This Journal, 25, 30 (1960). 10) J. S. D. Bacon and J. Edelman, Biochem. J., 48, 114 (1951). 232 Takao MINAMIKAWA, Takashi AKAZAWA and Ikuzo URITANI

TABLE I. CHROMATOGRAPHIC IDENTIFICATION OF SKIMMIN AND SCOPOLIN AND THEIR ACID HYDROLYZATES

* Rsklmmin by multiple development (three times) . ** Rglucose by multiple development (three times).

ly. The free forms of the coumarins fluores tectable in both cut and fresh root tissues, cence more strongly than the bound forms. the magnitude of their synthesis was lower The ultraviolet spectra of the aglycones than that of the free coumarins, e.g., umbel eluted from the paper chromatogram (sol liferone and scopoletin. As was discussed vent: 5% acetic acid) were identical with previously, the amount of coumarins synthe those of umbelliferone and scopoletin, respec sized in the diseased tissues was not high tively. enough to account for their role in the de The experimental data obtained above fense mechanism7,11,12) . Hence, the even clearly demonstrate the presence of two smaller amounts of their glucosides , now con- bound coumarins, skimmin and scopolin , in firmed to be present in the diseased tissues, sweet potato roots with black rot. Unfor may not have a significant role in the host- tunately, their concentrations in the diseased parasite relationship. However, identifica root tissues were not high enough to make tion of these coumarin glycosides in diseased possible their crystallization. Although the tissues merits further consideration in their production of bound coumarins in the di 11) I. Uritani, in Symposium on "Biochemistry of Plant Phenolic seased root tissues was induced specifically by Substances" (ed. G. Johnson and T. A. Geissman), Colorado State Univ., 1962, p. 98. the fungus infection and they were not de 12) I. Uritani, in Symposium on "Natural Biochemical Resistance to Disease in Plants", The Connecticut Agricultural Expt . Station, 1963. Two Glucosides of Coumarin Derivatives in Sweet Potato Roots Infected by Ceratocystisfimbriata 233

scopoletin) from the corresponding glycosides (skimmin and scopolin) by enzymatic hydro lysis can now be excluded in view of the experimental results in hand. The former compounds are assumed to be synthesized by a mechanism similar to that of polyphenol formation via dehydroquinic acid11). It has been claimed13) that latent plant glycosides are enzymatically hydrolysed to produce phe nolic substances and other toxic compounds in response to pathogenic infection; however, our present observations do not seem to cor roborate this claim. Acknowledgement.The authors wish to ex press their most sincere gratitude to Dr. J. B. Harborne of the John Inness Institute, Bay- fordbury, Hertford, England for sending us

FIG. 1. Ultraviolet Absorption Spectra of Band I, the authentic samples of skimmin and scopo Band II and the Authentic Compounds. lin. They also thank Dr. H. Imaseki in this laboratory for his many valuable advices dur possible relation to the mechanism of bio- ing the experiment. synthesis of coumarins. The possible forma tion of free coumarins (umbelliferone and 13) J. B. Pridham, "Phenolics in Plants in Health and Disease" (ed. J. B. Pridham), Pergamon Press, London. 1960, p. 9.