J. Biochem. 83, 1109-1116 (1978)

Accumulation of in Yeast Grown

in the Presence of Ethionine

Nakao ARIGA,* Hiroshi HATANAKA ,' Jun NAGAI,2 and Hirohiko KATSUKI

Department of Chemistry, Faculty of Science , Kyoto University, Sakyo-ku, Kyoto 606, and *Department of Chemistry , Faculty of General Education, Gifu University, Nagara, Gifu 502

Received for publication, September 17, 1977

In order to identify the methyl acceptor for the methylation of side-chains in , Saccharolnyces cerevisiae (wild type) was grown in the presence and absence of ethionine which was expected to be an inhibitor of the methylation. Gas-liquid chromato graphic analyses of the in the cells grown in the absence of ethionine showed that ergosterol was the most abundant sterol. On the other hand, a sterol, named sterol Z, accounted for more than 50% of the total sterols in the cells grown in the presence of ethionine. As a result of experiments to raise the yield of sterol Z, the best concentration of DL-ethionine for the production was found to be 1.0 mm. The use of the methionine-less mutant was less effective for the production of sterol Z. Sterol Z was isolated by repeated TLC and was identified as zymosterol from its melting point, GLC and mass spectrometry. The role of zymosterol and other sterols as the methyl-acceptor sterol in ergosterol bio synthesis is also discussed.

Although it is well known that ergosterol is syn a methylated sterol intermediate in yeast, 4a thesized via in yeast as in the case of methyl-5ƒ¿-ergosta-8,24-dien-3 ƒÀ-ol which has a biosynthesis in animals, much remains methyl group at position 4. Akhtar et al. (2) unsolved on the synthesis of ergosterol from synthesized 3H-labeled 4,4,14ƒ¿-trimethyl-5 ƒ¿-ergo lanosterol, especially on the methylation reaction sta-8,24(28)-dien-3 ƒÀ-ol and demonstrated that the of the sterol side-chain. Two theories have been compound was converted into ergosterol in yeast. proposed concerning the problem whether the They claimed that the methylation occurs before reaction occurs before or after complete elimination complete elimination of the three methyl groups. of the three methyl groups attached to the nucleus On the other hand, Katsuki and Bloch (3) examined of the lanosterol molecule. Barton et al. (1) found the methylation reaction with cell-free extracts of

yeast. According to their studies, most of the methylated sterols had a carbon skeleton of ergo 1 Present address: Mitsubishi-Kasei Institute of Life Sciences, Tokyo. stane, suggesting that Cg,-sterol is the methyl 2 Present address: Department of Biochemistry, Fa acceptor in the methylation reaction. Moore and culty of Medicine, Mie University, Tsu. Gaylor (4, 5) investigated the substrate specificity

Vol. 83, No. 4, 1978 1109 1110 N. ARIGA, H. HATANAKA, J . NAGAI, and H. KATSUKI

of S-adenosylmethionine: Ģ21 sterol methyltrans flask was, after sterilization, inoculated with 0.5 ml

ferase with a partially purified enzyme preparation . of an overnight culture of the yeast. The cultiva

They claimed that the methylation occurs at the tion was carried out with shaking at 30•Ž for the

stage of C27-sterol, since zymosterol showed the indicated periods. Aliquots of the culture were

highest activity as methyl acceptor so far examined withdrawn aseptically and diluted with an appro-. and 4ƒ¿-methyl-5ƒ¿-cholesta-8 ,24-lien-3ƒÀ-o1 and priate volume of saline. The absorbance at 660 4,4-dimethyl-5a-cholesta-8,24-dien-3ƒÀ-ol scarcely nm was measured in a Hitachi Model 101 Spectro showed any activity. Recently, we studied the photometer. The dry weight of the cells was methylation reaction with cell-free extracts of yeast estimated from the absorbance on the basis of the and reported that homocysteine inhibited the predetermined standard curve. methyltransfer reaction from S-adenosylmethionine Respiratory adaptation of yeast was carried

to sterols, giving rise to an accumulation of 5 out in the following way: 800 ml of the growth ƒ¿ cholesta-7,24-dien-3ƒÀ-ol (6). Parks et al. (7) also medium in a 1 liter Erlenmeyer flask was, after

reported the occurrence of this sterol in yeast. In sterilization, inoculated with 50 ml of an overnight

spite of the studies reported, it has not been culture of the yeast. After bubbling nitrogen gas established yet whether the methylation of the through the growth medium under sterile conditions

side-chain occurs before or after the demethylation for 1 h, the outlet of the glass tubing was sealed

of lanosterol in intact yeast cells and, even if the with a mercury trap. The yeast was grown under

latter possibility is valid, whether one or both of anaerobic conditions for 38 h at 28•Ž with stirring. the two C27-sterols is the methyl acceptor. In The harvested cells were washed twice with 0.1 mt connection with this, Fryberg et al. (8) proposed potassium phosphate buffer, pH 7.0, resuspended multiple pathways for ergosterol biosynthesis and in the same buffer containing 3 % glucose (0.4 mg

they considered that the methylation reaction of dry weight of cells per ml). The cells were occurs at the stage of zymosterol or a 4-methyl incubated with shaking for 4 h at 28•Ž for respira derivative of it and that is an important tory adaptation. The absorbance at 660 nm of intermediate in the biosynthesis. We have found the culture increased by 25 % during the incubation . that ethionine is an effective inhibitor of the Analysis of Sterols-The cells harvested by methylation of sterol side-chains in intact yeast centrifugation, were saponified at 80•Ž for 2 .5 h cells. This paper describes the accumulation of with 10% KOH in 50% methanol in a nitrogen

zymosterol in yeast grown in the presence of atmosphere. Non-saponifiable lipids were ex

ethionine under various conditions. The role of tracted by the usual method and analyzed with a zymosterol in the methylation reaction of the Shimadzu GC-5A or Hitachi K-53 gas-liquid

side-chain is also discussed. chromatograph equipped with a flame ionization

detector. A glass column (4 mm •~ 1 .5 m) was

MATERIALS AND METHODS used. The liquid phase was SE-30 (1 .5•“) ad sorbed on acid-washed Chromosorb W (Shimadzu Growth of Yeast-The yeast strains used were Co., Ltd.). Column temperature was 230•Ž and

Saccharomyces cerevisiae ATCC 12341 (wild strain) flow-rate of nitrogen carrier gas was 50 ml per and a methionine-less mutant (strain No . 16) minute, unless otherwise indicated . For the iden obtained by UV-irradiation of the haploid-type of tification of sterol Z, GLC with the following three

Saccharomyces cerevisiae (mating type a, galactose columns was carried out , using cholestane as an fermentable, pantothenate-non-requiring). The internal standard: 5 % DEGS , 1.5% OV-17 basal growth medium for the wild strain was the (Shimadzu Co., Ltd.) and Diasolid ZS (Nihon same as described previously (3), except for the Chromato Works, Ltd.). The determination of addition of DL-ethionine, and that for the mutant non-saponifiable lipids was carried out as follows:

strain was the same as described by Naiki and , sterol Z and ergosterol were determined

Iwata (9), except for the addition of the indicated by measuring the height of each peak . Total concentrations of DL-ethionine and 100 PM DL sterols were determined by weighing tracings of all methionine. For most of the growth experiments , areas of sterol peaks, using ƒÀ-cholestanol as a 50 ml of the growth medium in a 250-m1 Erlenmeyer standard. The amount of other sterols including

J. Biochem. ACCUMULATION OF ZYMOSTEROL IN YEAST 1111

lanosterol, , and episterol, etc. was ob yeast cells, the effects of various compounds on the tained by subtracting the amounts of sterol Z and incorporation of radioactivity from L-[methyl-14C-]

ergosterol from that of total sterols. methionine into sterols were examined. Among Mass Spectrometry-For the analysis of sterols the compounds tested, only ethionine inhibited the by mass spectrometry, a Hitachi RMU-6D was methylation and other compounds such as betaine, used. The electron energy and temperature of sarcosine, and norleucine were ineffective (data the sample heater were 70 ell and 120•Ž, respec not shown). tively. Injection of the sample heater was carried The yeasts were grown in the presence and out by the direct injection method. absence of ethionine and were harvested at the Thin-Layer Chromatography-For the isolation late logarithmic growth phase, and the non

of sterol Z, TLC was carried out. Kieselguhr G saponifiable lipids extracted from the cells were

(Merck, Art. 8129) was used as the adsorbent. analyzed by GLC as described in " MATERIALS The solvent system used was cyclohexane-ethyl AND METHODS. " The chromatograms in Fig. 1

acetate (99.5 : 0.5, v/v). A part of the chromato show a typical example of gas-liquid chromato

plate, after development, was sprayed with 20 grams of non-saponifiable lipids obtained from the SbCl5 solution in chloroform. wild strain of yeast grown under both conditions.

Cleavage of Sterol Side-Chain-The cleavage of Three major peaks representing squalene, unidenti

the sterol side-chain was carried out as described fied sterol Z and ergosterol, and three minor peaks

previously (3) with some modifications. Thesterol representing lanosterol, fecosterol, and episterol are sample was allowed to react with osmium tetroxide seen in the yeast grown in the absence of ethionine in ether. The osmium complex with the sterol was (Fig. IA). In the yeast grown in the presence of

degraded with sodium sulfite and the resulting ethionine, in contrast, sterol Z is the most abundant

compound was extracted with ether. It sterol, and ergosterol, fecosterol, and episterol

was oxidized with periodic acid and the remaining

periodic acid was decomposed with an excess amount of tartaric acid. The carbonyl compound

formed from the side-chain was converted into its

2,4-dinitrophenylhydrazone derivative. The hy drazone was extracted with benzene and then

separated from the hydrazone of glyoxylic acid

formed from tartrate by shaking the benzene

solution with a bicarbonate solution. The hydra zone in the benzene solution was concentrated and

subjected to TLC on Silica Gel G (Wako Co.,

Ltd.), using isopropyl ether-ligroin (1 : 1, v/v) as

a solvent system.

Materials-L-[Methyl-14C-]methionine (5.6 Ci/ mol) was purchased from Daiichi Pure Chemicals

Co. The sample of authentic zymosterol was a Fig. 1. Gas-liquid chromatograms of non-saponifiable gift from Professor C. Djerassi, Stanford Univer lipids from the cells grown in the presence and absence sity. L-Methionine, DL-ethionine, and other chemi of ethionine. The wild strain of yeast was grown cals of reagent grade were obtained from Nakarai aerobically for 40 h in the absence (A) and presence Chemicals Co., Ltd. (B) of 1 MM DL-ethionine. The cells were saponified and non-saponifiable lipids were extracted. They were RESULTS analyzed by GLC. A column of 1.5•“ SE-30 adsorbed on Chromosorb W was used. The column temperature Gas-Liquid Chromatograms of Sterols from for analysis was 230•Ž. Sq, Erg, Lano, Feco, and Epi

the Wild Strain in the Presence and Absence of represent squalene, ergosterol, lanosterol, fecosterol, and

Ethionine-In order to identify inhibitors of the episterol, respectively. Z is the sterol under investiga tion. methylation reaction of the side-chain with intact

Vol. 83, No. 4, 1978 1112 N. ARIGA, H. HATANAKA, J. NAGAI, and H. KATSUKI decreased (Fig. 1B). This led us to the supposi Effects of Methionine and Ethionine on Sterol tion that sterol Z might be the acceptor sterol of Composition of the Methionine-Less Mutant-The the methyl group in the reaction or a sterol having effects of ethionine on the sterol composition of a structure close to the acceptor. cells were investigated with the methionine-less Effect of Ethionine on Sterol Composition of mutant of yeast. Since the yeast required meth the Wild Strain-In order to find out the conditions ionine for growth, it was grown at varying concen for accumulation of sterol Z, the effects of increas trations of ethionine in the presence of 0.1 mm ing concentrations (0.5-10 mm) of DL-ethionine on L-methionine. The growth of cells was suppressed the sterol Z content of the cells were examined. by much lower concentrations of ethionine than Table I shows the results of analysis of non those for the wild strain. Table ‡U shows the saponifiable lipids of the cells grown for 40 h results of analysis of non-saponifiable lipids of the under these conditions. By increasing the concen mutant. Although the addition of ethionine in tration of ethionine to 2.0 mm, growth of the the presence of 0.1 mm methionine, a limiting con wild-strain was inhibited by 40%, whereas the centration for the cell growth, did not affect the non-saponifiable lipids content was not changed. non-saponifiable lipids content so much, it steadily Squalene and sterol Z increased, concomitant with increased the proportion of sterol Z in it. When the decrease of ergosterol. When the concentra 0.2 MM DL-ethionine was added, the proportion of tion of ethionine was more than 5 mm, the non sterol Z increased 2.5 times of that observed in saponifiable lipids content and sterols decreased. the absence of ethionine. Although the proportion The results show that the optimal concentration of sterol Z in the non-saponifiable lipids of the of ethionine for the accumulation of sterol Z is mutant was almost comparable with that of the 1.0 mm. wild strain under optimal conditions, the yield of

TABLE I. Effect of ethionine on the composition of non-saponifiable lipids of the wild strain. The cells were grown aerobically for 40 h at varying concentrations of DL-ethionine in 250-m1 Erlenmeyer flasks with shaking at 28•Ž. "Other sterols" are indicated in the text. NSL indicates non-saponifiable lipids.

TABLE ‡U. Effect of ethionine on the composition of non-saponifiable lipids of the methionine-less mutant. The cells were grown aerobically for 69 h in the presence of 0.1 mm L-methionine and varying concentrations of DL ethionine.

J. Biochem. ACCUMULATION OF ZYMOSTEROL IN YEAST 1113

sterol Z per liter of the mutant culture was only were separated by TLC on Kieselguhr G as de

one-tenth of that of the wild strain, since both the scribed in " MATERIALS AND METHODS."

cell yield and the non-saponifiable lipids content Sterol Z was located just in the upper portion of in the mutant cells were lowered. the ergosterol fraction. It was clearly distin Effects of Ethionine and Cycloheximide on the guishable from ergosterol since it showed a brown

Composition of Non-Saponifiable Lipids of the color upon spraying with 5 % SbCl5 in chloroform

Wild Strain during the Respiratory Adaptation-It on part of the chromatogram, whereas ergosterol

is well known that when anaerobically-grown yeast gave a pink color which then turned to purple. cells are aerated in a phosphate buffer containing Figure 2 shows a typical thin-layer chromatogram

glucose, the sterol content increases accompanied with the appearance of respiratory activity. With

these aerated yeast cells, the effects of ethionine

on sterol Z content were investigated. As shown in Table ‡V, the non-saponifiable lipids content

increased 3-fold upon aeration of the anaerobically

grown cells for 4 h. Analysis of the lipids indicated an increase of sterols by more than 10 times, since

squalene did not increase during the aeration. The addition of cycloheximide resulted in the inhibi

tion of the synthesis of non-saponifiable lipids and

in the conversion of squalene into sterols, especially into sterols Z, episterol, and lanosterol. However, Fig. 2. Thin-layer chromatogram of sterols obtained when the aeration was carried out in the presence from non-saponifiable lipids of yeast grown in the of ethionine, formation of considerable amounts presence of ethionine. The sterol sample was applied of non-saponifiable lipids was observed and the to a Kieselguhr G plate and it was developed using accumulation of sterol Z was marked. cyclohexane-ethyl acetate (99.5: 0.5, v/v). A part of Identification of Sterol Z-For the identifica the chromatogram obtained was visualized by spraying tion of sterol Z, the wild strain of yeast was grown 5% SbCl5 solution in chloroform. (A) and (B) show for 48 h in the presence of 1 MM DL-ethionine in a the spots of authentic samples of ergosterol and lano large scale under the conditions described. The sterol, respectively, (C) shows a chromatogram of

harvested cells were hydrolyzed with alkali by the sterols obtained from non-saponifiable lipids of the

usual method and non-saponifiable lipids extracted yeast.

TABLE ‡V. Effects of ethionine and cycloheximide on the composition of sterols during respiratory adaptation. Wild strain yeast cells grown anaerobically for 38 h were harvested, washed and resuspended in phosphate buffer containing 3% glucose. They were aerobically shaken at 28°C for 4 h as described in " MATERIALS AND METHODS."

a Before aerobic adaptation. b The cells were aerated for 4 h in the presence of cycloheximide c The ceps were aerated for 1 h in the absence of cycloheximide and for a further 3 h in the presence of it. d The cells were aerated for 4 h without the addition.

Vol. 83, No. 4, 1978 1114 N. ARIGA, H. HATANAKA, J. NAGAI, and H. KATSUKI

of the sterols. By repeating the TLC three times, According to Djerassi, zymosterol gave m/e 384 a pure preparation of sterol Z could be obtained. [M, 100], 369 [M-15, 601,351 [M-15-18, 141, 281

The compound showed a melting point of 107 [21], 271 [37], 255 [19], 231 [28], 229 [35], and 219

110•Ž and its benzoate derivative, mp 126•Ž. [49] as the main mass peaks having m/e values

They agreed with the melting points of authentic greater than 200 (11). These completely coincided samples of zymosterol (110•Ž) (4) and its benzoate with those of sterol Z and the authentic sample of derivative (124-126•Ž) (10), respectively. For zymosterol. further confirmation of sterol Z, it was subjected to GLC. Relative retention times of sterol Z on

the several columns and temperatures used are as follows: 2.20 at 230•Ž on 1.5% SE-30, 3.59 at

240•Ž on 1.5% OV-17, 4.77 at 220•Ž on 5

DEGS (as methyl ether), 2.46 at 210•Ž on 0.56

SE-52, and 4.10 at 240•Ž on Diasolid ZS. These values completely agreed with those of the authen

tic sample of zymosterol. For determination of the double bond in the side-chain of sterol Z, it

was subjected to the treatment for cleavage of the side-chain as described in " MATERIALS AND METHODS. The 2,4-dinitrophenylhydrazone of Fig. 3. Thin-layer chromatogram of the 2,4-dinitro the carbonyl compound obtained was identified as phenylhydrazones of carbonyl compounds obtained the derivative by TLC on Silica Gel G from the cleavage of the double bond in the side-chain

(Fig. 3). This result indicates the existence of a of sterol Z. The cleavage of the side-chain of sterol Z double bond at the 24-position in the side-chain. was carried out as described in " MATERIALS AND Identification of sterol Z was also made by METHODS." The same operation as above except for the omission of sterol Z was run as the control experi mass spectrometry. Figure 4 shows a mass spec ment. The finally obtained 2,4-dinitrophenylhydrazone trogram of sterol Z. The authentic sample of fractions were applied to a TLC plate and developed with zymosterol gave exactly the same spectrogram isopropyl ether-ligroin (1 : 1, v/v). (A) shows the (not shown in the figure). authentic sample of acetone-2,4-dinitrophenylhydrazone. It showed the molecular ion at m/e 384. (C) shows the hydrazone obtained from the cleavage of Intense peaks were observed at in/e 369 and 351. the sterol Z side-chain. (B) shows the hydrazone frac A weak peak was also observed at m/e 271, cor tion obtained from the control experiment. The spots, responding to the loss of the side-chain together d, e, and f were the decomposition products of the hy with two hydrogen atoms from the steroid nucleus. drazone and the hydrazine reagent.

Fig. 4. Mass spectrogram of sterol Z.

J. Biochem. ACCUMULATION OF ZYMOSTEROL IN YEAST 1115

sterol in yeast is not clear, since the sterol is formed

DISCUSSION through isomerization from zymosterol and is not clearly separable from ergosterol in GLC. Even

Methionine, after being converted to S-adenosyl if the methylation occurs with the three sterols,

methionine, acts as a methyl donor for various the extent of the contribution of each methylation methylated compounds such as choline and ergo activity must vary, depending on the growth condi

sterol. Ethionine inhibits the methylation reac tions, intracellular concentrations of the three

tions besides protein synthesis by competing with sterols and other factors.

methionine. In order to inhibit the methylation Furthermore, the observations that sterols in

of the sterol side-chain and accumulate the methyl yeast exist in free and esterified forms (14) and

acceptor, the ethionine addition experiment to the that the esterified sterols do not serve as substrate

yeast culture was carried out. The accumulated for the methylation (14) and isomerization (12), sterol was found to be zymosterol in the present complicate these problems. According to studies

investigation. by Parks et at. (7) and our group (15), zymosterol

Homocysteine which was effective for inhibi is mostly in the form of fatty acid ester in the

tion of the methylation in the cell-free system (6) aerobically-grown cells, showing the highest ratio was not effective in the intact-cell system. Simi of esterified to free forms among sterols. This

larly, ethionine was not effective in the cell-free phenomenon might act as a control mechanism of

system (6). Presumably this was due to high ergosterol biosynthesis by yeast.

activity of an other metabolism of homocysteine Although a considerable amount of zymosterol

other than its methylation to methionine in the is usually found in yeast cells, its preparation has intact-cell system, and to weak activity of activation not necessarily been easy, owing presumably to

of ethionine to S-adenosylethionine, a competitive the difficulty in separation from other sterols. The

inhibitor or of S-adenosylmethionine utilization, in predominant accumulation of the sterol under the the cell-free system. conditions reported made its isolation much easier.

The observation that zymosterol was not The method reported here must be useful for its

accumulated by the addition of cycloheximide preparation. suggests direct inhibition by ethionine of the methy We are grateful to Professor C. Djerassi, Stanford lation of the side-chain, not indirect through pro University, for the kind gift of zymosterol and a copy of tein synthesis. a mass-spectrogram of zymosterol, and to Professor N.

The results obtained, together with the studies Naiki, Gifu University for the kind gift of the methio by Moore and Gaylor (5), might suggest that the nine-less mutant of yeast. Thanks are also due to

acceptor sterol for methylation is zymosterol. Professor S. Okuda, Dr. A. Kawaguchi, and Mrs. N. Morisaki, Institute of Applied Microbiology of Tokyo However, the formation of 4ƒ¿-methyl-5ƒ¿-cholesta University, for the mass-spectrometry analysis. 8,24-dien-3ƒÀ-ol and its conversion to ergosterol in

the intact cells (8) indicate that it is also able to REFERENCES1 serve as the methyl acceptor. Moreover, 5

ƒ¿ cholesta-7,24-dien-3 ƒÀ-0l was also found to be; . Barton, D.H.R., Harrison, D.M., & Widdowson, D.A. (1968) Chem. Comrnun. 17-19 accumulated in the cell-free system upon the inhibi 2. Akhtar, M., Parvz, M.A., & Hunt, P.E. (1966) tion of methylation by homocysteine (6) and to act Biochem. J. 100, 38C-40C as the methyl acceptor (12). Although Fryberg 3. Katsuki, H. & Bloch, K. (1967) J. Biol. Chem. 242, et al. proposed a multiple pathway for the bio 222-227 synthesis of ergosterol through the methylation of 4. Moore, J.T., Jr. & Gaylor, J.L. (1970) J. Biol. Cheni. 4ƒ¿-methyl-5ƒ¿-cholesta-8,24-dien-3 ƒÀ-ol and zymo 244,6334-6340 sterol (8), our group recently suggested that the 5. Moore, J.T., Jr. & Gaylor, J.L. (1970) J. Biol. Chem. methylation mainly occurs at the stage of C27 245,4684-4688

sterol(s) and that 4ƒ¿-methyl-5ƒ¿-cholesta-8,24-dien 6. Hatanaka, H., Ariga, N., Nagai, J., & Katsuki, H.

3 ƒÀ-ol is a minor acceptor (13). However, how (1974) Biochem. Biophys. Res. Commun. 60, 787-793 7. Parks, L.W., Anding, C., & Ourisson, G. (1974) much the methylation through 5ƒ¿-cholesta-7,24 Eur. J. Biochem. 43, 451-458 dien-3ƒÀ-ol contributes to the biosynthesis of ergo

Vol. 83, No. 4, 1978 1116 N. ARIGA, H. HATANAKA, J. NAGAI, and H. KATSUKI

8. Fryberg, M., Oehlschlager, A.C., & Unrau, A.M. 13. Osumi, T., Taketani, S., & Katsuki, H. (1978) (1973) J. Amer. Chem. Soc. 95, 5747-5757 J. Biochem. 83, 681-691 9. Naiki, N. & Iwata, M. (1962) Sci. Rep . Fac. Lib. 14. Nagai, J., Kawamura, S., & Katsuki, H. (1977) Arts Educ., Gifu Univ. (in Japanese) 3, 70-74 J. Biochem. 81, 1665-1673 10. Alexander, J.G. & Schwenk, E. (1957) Arch. 15. Taketani, S., Osumi, T., Nagai, J., & Katsuki, H. Biochem. Biophys. Acta 66, 381-387 (1975) Proc. Japan Conf. Biochem. Lipids (in 11. Wylli, S.G. & Djerassi, C. (1968) J. Org. Chem. 33 , Japanese) 17, 85-88 305-313 12. Yabusaki, Y., Nishino, T., Katsuki, H., & Ariga, N. (1976) Seikagaku (in Japanese) 48, 465

J. Biochem.