Pii: s0014-5793(98)01258-7

The role and source of 5P-deoxyadenosyl radical in a carbon skeleton rearrangement catalyzed by a plant enzyme Sandrine Ollagnier, Eric Kervio, Jaènos Reètey* Lehrstuhl fuër Biochemie, Institut fuër Organische Chemie, Universitaët Karlsruhe, Kaiserstr. 12, D-76128 Karlsruhe, Germany Received 15 August 1998; received in revised form 24 September 1998 Abstract The last step in the biosynthesis of tropane alkaloids the major source of 5P-deoxyadenosyl radical [13]. More re- is the carbon skeleton rearrangement of littorine to hyoscyamine.
cently, S-adenosylmethionine (SAM) was also identi¢ed to The reaction is catalyzed by a cell-free extract prepared from play a similar role [14]. Therefore, SAM has been called the cultured hairy roots of Datura stramonium. Adenosylmethionine stimulated the rearrangement 10^20-fold and showed saturation Here we report on the involvement of SAM in the rear- kinetics with an apparent Km of 25 WM. It is proposed that rangement of littorine to hyoscyamine.
S-adenosylmethionine is the source of a 5P-deoxyadenosyl radical which initiates the rearrangement in a similar manner as it does in analogous rearrangements catalyzed by coenzyme B dependent enzymes. Possible roles of S-adenosylmethionine as a radical source in higher plants are discussed.
z 1998 Federation of European Biochemical Societies.
They were obtained from commercial sources and were the highest purity normally available. SAM and [2,8,5P-QH]ATP (65% QH in posi- tion 5P) were obtained from Amersham, Aylesbury, UK, dithiothreitol Key words: Tropane alkaloid; Carbon skeleton (DTT) from Promega, ethylene diamine tetraacetic acid (EDTA) from rearrangement; S-Adenosylmethionine as 5P-deoxyadenosyl Serva, B5 medium from Bioproduct (Boehringer Ingelheim).
2.2. Preparation of the phenyl-lactoyl tropine (littorine) According to a modi¢ed procedure [16] L-(3)-phenyllactic acid (Fluka) (116 mg, 0.7 mmol) dried over phosphorus pentoxide was added to tropine (Acros) (0.5 mmol) previously dried over sodium hydroxide pellets. The mixture was heated to 145³C on an oil bath and dry hydrogen chloride was bubbled through the solution for 3.5 h.
The last and most intriguing step in the biosynthesis of the (Note that esteri¢cation attempts with para-toluene sulfonic acid or tropane alkaloids hyoscyamine and scopolamine is the carbon under montmorrillonite KSF as alternative procedures gave no satis- skeleton rearrangement depicted in Scheme 1 (for a review see factory results.) The cooled mixture was dissolved in a 10 ml HPSOR solution (50 mM), ¢ltered and made alkaline with a 10% aqueous [1]). Due to its similarity to the rearrangement of methylma- ammonium hydroxide solution, then extracted into chloroform lonyl CoA to succinyl-CoA it was originally speculated that (6U10 ml). The solvent was removed under reduced pressure and the residue puri¢ed by partition chromatography on Celite (Merck) IP is involved in the reaction [2^4]. Since, in spite of some claims [5^7], no trace of vitamin B containing 0.5 M phosphate bu¡er (pH 6.8) eluted with chloroform.
GC-MS analysis of the combined fractions (97 mg, yield: 48%) in- found in plants this idea was abandoned [8,9]. More recently dicated almost pure (98%) littorine.
it has been surmised that cytochrome P450 is the agent ab- Retention time (1,4-dioxane) on a Optima 1-02 capillary column stracting the benzylic hydrogen atom and redonating a hy- (25 mU0.2 mm), ramp 100^300³C at 20³C/min): 12.2 min. MS m/z droxyl group after the rearrangement [10,11]. Accordingly, (rel. int.): 361(4), 193(3), 140(4), 125(12), 124(100), 123(1), 96(9), clotrimazole, a known P-450 inhibitor, has been said to inhibit 95(5), 94(12), 91(8), 83(15), 73(19) [17].
IQC NMR and IH NMR spectra were recorded on a Bruker DRX the formation of hyoscyamine [1], and the 3-H‚e atom ab- 500 spectrometer. Deuteriochloroform was used as the solvent with stracted from the benzylic position was lost during the bio- TMS as an internal reference for IH. The abbreviations used are as synthetic experiment [12]. When the OH group of littorine was follows: N, shift; s, singlet; d, doublet; t, triplet; m, multiplet; b, labelled with the isotope IVO, 25^29% of the label was lost in the product [11]. These results were interpreted as the involve- Q) N : 1.65 (1H, HT or HU, d, JQrr = 15.8 Hz) ; 1.85 (1H, HU or HT, d, JQrr = 15.8 Hz); 1.94^2.20 (2H, HT or HU, m); 2.00 ment of two enzymes, a mutase and a dehydrogenase, in the (N-CHQ, s); 2.54^2.72 (4H, HR and HP, m); 2.78 (1H, HS or HI, t, conversion of littorine into hyoscyamine, the geminal diol JQrr = 15.8 Hz); 2.80 (1H, HI or HS, t, JQrr = 15.8 Hz); 2.98 (2H, being an intermediate in the reaction. The dehydratation HQH, d, JQrr = 6.5 Hz); 3.57 (b, OH); 4.37 (1H, HPH, t, JQrr = 6.5 Hz); thereof to the aldehyde could be partially stereospeci¢c ex- 4.99 (1H, HQ, t, JQrr = 4.5 Hz); 7.11^7.22 (5H—rom, m) [18]. IQC NMR (CDClQ) N: 25.1, 25.3 (CU and CT); 35.1, 35.2 (CR and CP); 37.3 (CQH); plaining the less than 50% loss of IVO in the overall reaction.
40.7 (CHQ-N); 60.0 (CI and CS); 71.6 (CQ); 74.3 (CPH); 126.9 (CUH); Only a few agents are known in biochemistry that are able to 128.5, 129.3 (CSHYTHYVHYWH); 136.4 (CRH); 173.5 (CIH).
abstract hydrogen atoms from non-activated positions. Beside cytochrome P450, nature uses the 5P-deoxyadenosyl radical SAM synthetase was puri¢ed by the procedure of Markham et al.
for this purpose. In animals and bacteria coenzyme BIP is [19], with some minor modi¢cations: growth in the presence of oxy- tetracycline (30 Wg/ml) and disruption of the cells by sonication (Ban- delin Nonopuls HD 2200). The speci¢c activity of the enzyme was *Corresponding author. Fax: (49) (721) 608-4823.
E-mail: [email protected] 2.4. Preparation of the S-[2,8,5P-QH]adenosylmethionine S-[2,8,5P-QH]Adenosylmethionine was enzymatically synthesized 0014-5793/98/$19.00 ß 1998 Federation of European Biochemical Societies. All rights reserved.
PII: S 0 0 1 4 - 5 7 9 3 ( 9 8 ) 0 1 2 5 8 - 7 S. Ollagnier et al./FEBS Letters 437 (1998) 309^312 from [2,8,5P-QH]ATP (speci¢c activity: 41 mCi/mmol) and unlabelled mercially prepared reagent (bicinchoninic acid) and bovine serum al- methionine, using a SAM synthetase preparation.
The reaction mixture contained 200 Wl of 0.1 M diluted solution of [2,8,5P-QH]ATP (speci¢c activity: 424 700 cpm/nmol) in 3 ml Tris-HCl 2.9. Extraction and analysis of alkaloids bu¡er (pH 7.4), adjusted to pH 7,6 by addition of a 2 M KOH Harvested roots were freeze-dried, then powdered and soaked over- solution, 1.5 eq. MgClP, 5 eq. KCl, 1.2 eq. L-methionine added as night in an EtOH-28% NHROH (19:1) mixture. This macerated ma- solids, 100 Wl reduced glutathione (25 mg/ml) and 11 mg of SAM terial was centrifuged for 5 min at 1300Ug. Extraction with the basic synthetase were incubated for 45 min at 35³C. The reaction was ter- alcohol was repeated twice and the combined alcohol fractions were minated by addition of 40 Wl of trichloroacetic acid (6 M). The result- evaporated to dryness at 35³C. The dry residue was dissolved in 2 ml ing suspension was centrifuged at 4³C and the extracts pooled and of 0.1 N HCl and the acidic aqueous solution was ¢ltered through stored at 4³C. After lyophilization, the residue was dissolved in ¢lter No. 2. One milliliter of the ¢ltrate was made alkaline with 2 ml doubly distilled water and applied to a carboxymethyl sepharose of 1 M NaPCOQ-NaHCOQ bu¡er (pH 10) and 1 ml of this alkaline high £ow column (0.8U20 cm FPLC column, Pharmacia Biotech, solution was loaded onto a Extrelut-1 column. After 10 min, 6 ml of Freiburg, Germany) previously equilibrated with water. The column CHClQ was passed through the column and the chloroform fractions was washed with 2 volumes of water and then S-[2,8,5P- were evaporated to dryness at 35³C. The dry residue was dissolved in QH]adenosylmethionine was eluted with 1 M NaCl gradient (4 ml/ a mixture of 1,4 dioxane/N,O-bis-(trimethylsilyl)-acetamide (v/v 4:1).
min). Fractions containing S-[2,8,5P-QH]adenosylmethionine (speci¢c Alkaloids were measured with a gas chromatograph coupled with a activity: 123 000 cpm/nmol) were mixed, lyophilized and stored at mass spectrometer model Hewlett Packard 5890 Serie II using a capil- lary column Optima 1-02 Wm (25 mU0.2 mm). The column temper- ature was 350³C, the carrier gas was HP at a £ow rate of 20 ml/min.
Roots from Datura stramonium D15/5 were a gift from Dr. N.J.
The two alkaloids identi¢ed for this study were littorine and hyos- Walton (Institute of Food Research, Norwich, UK). Roots cultures were initiated by wounding surface sterilized explants of leaves on the Littorine RT (retention time): 12.2 min. MS m/z (rel. int.): 361(4), midrib with an overnight suspension of Agrobacterium rhizogenes.
193(3), 140(4), 125(12), 124(100), 123(1), 96(9), 95(5), 94(12), 91(8), Tissue bearing emergent roots were excised and placed into 8 ml of Gamborg's B5 medium supplemented by 500 Wg/ml ampicillin. Roots Hyoscyamine RT: 12.1 min. MS m/z (rel. int.): 361(5), 193(0.5), for enzyme extraction and puri¢cation were grown at 25³C, under 140(7), 125(10), 124(100), 123(6), 104(6), 96(10), 95(6), 94(13), stirring at 90 rpm. Rapidly growing roots were subcultured every 2 weeks into 50 ml of the same medium in a 250 ml Erlenmeyer £ask.
After eight subcultures it should be possible to omit ampicillin from the medium. After this time, roots were harvested by ¢ltration in An Incapharm 100 RP-18 TS (250U4.6 mm) column was equili- vacuo, washed twice with water and frozen immediately at 370³C.
brated at room temperature with 0.1% tri£uoroacetic acid in water solution. 5 min after sample injection (20 Wl in same bu¡er), a gra- dient of 0^60% solvent B (0.1% tri£uoroacetic acid in acetonitrile) in All procedures were done at 4³C. Roots (6 g) were crushed in a 40 min was developed by a Hewlett-Packard 1050 T system at a £ow mortar with sea sand, mixed with 15 ml of extraction bu¡er (100 mM rate of 1.2 ml/min. The eluted substances were monitored at 215 and KPHPOR pH 8, 3 mM DTT, 5 mM NaP EDTA and 250 mM sucrose) 254 nm. Labelled products were analyzed by liquid scintillation count- and then ground during 30 min until a homogeneous suspension was ing (Tri-carb 2100TR, Packard, a Canberra company, Meriben, USA) obtained. The homogenate was clari¢ed by passage through Mira- after adding 15 ml of scintillation liquid (Lumasafe plus^Lumac LSC) cloth (Calbiochem, Novabiochem, Frankfurt a/M, Germany) and cen- trifugation (30 000Ug, 4³C, 30 min). Low molecular weight com- pounds were removed using a PD-10 desalting column (Pharmacia 2.11. Determination of the isotope exchange Biotech AB, Uppsala, Sweden) equilibrated in a one tenth strength Tritium exchange with water was measured by liquid scintilla- extraction bu¡er. The proteins were eluted with 3.5 ml of the same tion counting (Tri-carb 2100 TR) after loading 10 ml of scintillation bu¡er. The yellow solution obtained was concentrated to 2^4 mg/ml liquid to 10 Wl of water resulting from bulb-to-bulb distillations. The with centricon 5000 Da (Pharmacia Biotech AB, Uppsala, Sweden).
QH]adenosylmethionine solution (speci¢c activity: 123 000 cpm/nmol) and the resulting distilled water and between aqueous phase (0.1% The activity of the enzyme was determined by measuring the con- tri£uoroacetic acid in doubly distilled water) before HPLC analysis version of littorine into hyoscyamine by the GC-MS method. The standard reaction mixture, in a volume of 200 Wl, contained 200 WM SAM (unlabelled or labelled as prepared above), 100 Wl of the protein solution (200^400 Wg), 2 mM of littorine, adjusted to 200 Wl with the extraction bu¡er. After 40 min incubation at 33³C, the reaction was terminated by addition of 100 Wl 30% NHROH. Control experiments 3.1. Dependence of the rearrangement littorine/hyoscyamine on lacking substrate were routinely included. Reaction mixtures were loaded on an Extrelut-1 column (Merck Art 1.13076, Darmstadt, Germany). After 15 min the alkaloids were eluted with 13 ml of The experiments were carried out as described in Section CHClQ. The chloroform fractions were evaporated to dryness at 2.7 and are listed in Table 1. The silylated products were 35³C, dissolved in 200 Wl of a fresh solution of 1,4 dioxane/N,O-bis- detected by GLC coupled with MS. Omission of SAM from (trimethylsilyl)-acetamide (4:1 v/v), analyzed by GC-MS (experiments the assay mixture resulted in only 3.5% hyoscyamine while 96.5% remained littorine. Addition of SAM (200 WM) in- creased the conversion leading to 61% hyoscyamine. When Protein was assayed by the method of Smith et al. [20] using com- the hairy root extract was kept frozen at 320³C for a week Dependence of the rearrangement of tropane alkaloids on SAMHairy roots S. Ollagnier et al./FEBS Letters 437 (1998) 309^312 The 10^20-fold enhancement of the carbon skeleton rear- rangement, littorine to hyoscyamine (Scheme 1) by SAM in cell-free extracts of a root culture of Datura stramonium sug- gests a new role for this cofactor in plant biochemistry. The well-known function of SAM as methyl donor is operative in both prokaryotic and eukaryotic organisms and requires a heterolytic cleavage of a carbon sulfur bond. More recently, Scheme 1. Carbon skeleton rearrangement of tropane alkaloids.
Frey and coworkers discovered and documented [14,15] a case in which SAM is the source of a 5P-deoxyadenosyl radical, the homolytic cleavage of the carbon sulfur bond being promoted the conversion under the same conditions without and with reductively by an FeRSR cluster-containing enzyme.
SAM decreased to 2.3 and 28% hyoscyamine, respectively.
We surmise a similar role for SAM in the rearrangement of After 4 weeks of freezing little if any mutase activity of the littorine to hyoscyamine. Indeed, addition of SAM to the assay mixture not only increased the conversion of littorine to hyoscyamine, but showed saturation kinetics with a Km of 3.2. Determination of the Michaelis constant for SAM As the enzyme source we used a hairy root extract, thawed The low activity of the hairy root extracts without addition after 1 week in the deep-freeze. The kinetic experiments were of SAM may be due to trace amounts of endogenous SAM carried out under conditions of the enzyme assay (see Section that remained in the extract after the PD-10 desalting column.
2.7), except that the concentration of SAM was varied be- Indeed, in previous studies no SAM was added to the cell-free tween 50 and 500 WM. The Km and Vm—x values were deter- system, but much more extract and longer incubation times mined using Lineweaver-Burk plot (Fig. 1).
were used. Moreover, most of the incorporation experiments The following values were found: Km = 25 þ 5 WM, Vm—x = were carried out with growing root cultures and the incuba- tion times were 10^28 days. In these cases endogenous SAM must have been provided by the in vivo system.
3.3. Conducting the rearrangement in the presence of Preliminary experiments using [2,8,5P-QH]SAM showed no incorporation of tritium into littorine or hyoscyamine but When the enzymic rearrangement of littorine to hyoscy- some loss of the radioactivity to water. This is in agreement amine was carried out in the presence of [2,8,5P-QH]SAM with the observed loss of the benzylic pro-R H-atom of littor- (see Section 2.4 for preparation), no incorporation of tritium ine during the rearrangement [12]. Since all work has been into the alkaloids was observed. However, about 2% of the done with root cultures or crude cell-free extracts it is possible tritium (calculated for the 5P position) was washed out into that the migrated H-atom is washed out by redox enzymes.
the medium. Only a trace of tritium was found in the solvent This could also explain the 25^29% loss of IVO of littorine in a control experiment without hairy root extract. HPLC during the rearrangement [11] since the intermediate aldehyde showed no appreciable amounts of 5P-deoxyadenosine after could be in equilibrium with its hydrate.
the reaction. If the latter was an intermediate, then only in The steric course of the migration at the relevant centers is enzyme-bound catalytic amounts. If the putative 5P-deoxyade- also of importance. After contradictory results [2,3,21] it has nosyl radical abstracted the benzylic pro-R H-atom, then this been established that the substitution occurs with inversion at process must have been irreversible. Possible reasons for both migration centers [12]. The same steric course has been the loss of tritium from the 5P position of SAM are given determined for the lysine 2,3-aminomutase reaction [22]. In the latter, however, the migrating hydrogen atom is not lost Fig. 1. Lineweaver-Burk determination of Km and Vm—x. X = % GC-MS hyoscyamine.
S. Ollagnier et al./FEBS Letters 437 (1998) 309^312 to the medium. This discrepancy can only be clari¢ed when [8] Freer, I., Pedrochi-Fantoni, G., Picken, D.J. and Overton, K.H.
pure littorine mutase is available. Another plant enzyme that (1981) J. Chem. Soc. Chem. Commun. 80^82.
catalyzes the rearrangement of K-phenylalanine to L-phenyl- [9] Leete, E. (1990) Planta Med. 56, 339^352.
[10] Robins, R.J., Chesters, N.C.J.E., O'Hagan, D., Parr, A.J., Wal- alanine is the phenylalanine aminomutase. This reaction is ton, N.J. and Woolley, J.G. (1995) J. Chem. Soc. Perkin Trans.
part of taxol biosynthesis. Recently it has been established that the substitution occurs with retention of con¢guration [11] Wong, C.W., Hamilton, J.T.G., O'Hagan, D. and Robins, R.J.
at both migration centers [23]. This fact does not rule out (1998) J. Chem. Soc. Chem. Commun. 1045^1046.
[12] Chesters, N.C.H.E., Walker, K., O'Hagan, D. and Floss, H.G.
the possibility that the same cofactor, namely SAM, is oper- (1996) J. Am. Chem. Soc. 118, 925^926.
ative also in the phenylalanine aminomutase reaction. Prece- [13] Frey, P.A., Essenberg, M.K. and Abeles, R.H. (1969) J. Biol.
dence for opposite steric courses is found in the coenzyme [14] Chang, C.H., Ballinger, M.D., Reed, G.H. and Frey, P.A. (1996) IP-dependent carbon skeleton rearrangements. Methylma- lonyl-CoA mutase operates with retention [24,25], methylene Biochemistry 35, 11081^11084 and previous papers cited therein.
[15] Frey, P.A., Reed, G.H., Moss, M.L., Petrovich, R.M., Ballinger, glutarate mutase [26] and glutamate mutases [27] with inver- M.D., Lieder, K.W., Wu, W., Chang, C.H., Bandarian, V., Ru- sion of con¢guration. Even the same enzyme, the coenzyme zicka, F.J., Lo Brutto, R. and Beinert, H. (1998) in: Vitamin BIP IP-dependent ethanolamine ammonia-lyase, operates with IP-Proteins (Kraëutler, B., Arigoni, D. and Golding, B.T., opposite steric courses, inversion and retention with (2R)- [16] Robins, R.J., Bachman, P. and Woolley, J.G. (1994) J. Chem.
and (2S)-propanolamines, respectively [28]. We thus propose that SAM may be the cofactor in the BIP-like rearrangements [17] Hashimoto, T., Yukimune, Y. and Yamada, Y. (1986) J. Plant occurring in plants and the reaction is initiated by the 5P- deoxyadenosyl radical derived from SAM.
[18] Cannon, J.R., Joshi, K.R., Meehan, G.V. and Williams, J.R.
That plants use SAM as the source of 5P-deoxyadenosyl [19] Markham, G.D., Hafner, E.W., Tabor, C.W. and Tabor, H.
radical has been established for the last step of biotin biosyn- (1983) Methods Enzymol. 94, 219^222.
thesis in Arabidopsis thaliana [29]. A similar mechanism oper- [20] Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gart- ates in Bacillus sphaericus and Escherichia coli [30,31]. We ner, F.H., Prorenzano, M.D., Fujimoto, E.K., Goeke, N.M., thus believe that while bacteria are able to use both coenzyme Olson, B.J. and Klenk, D.C. (1985) Anal. Biochem. 150, 76^85.
[21] Platt, R.V., Opie, C.T. and Haslam, E. (1984) Phytochemistry 23, BIP and SAM as sources of 5P-deoxyadenosyl radical, animals and humans inherited from them the former and higher plants [22] Aberhart, D.J., Gould, S.J., Lin, H.J., Thiruvengadam, T.K. and Weiller, B.H. (1983) J. Am. Chem. Soc. 105, 5461^5470.
[23] Walker, K.D. and Floss, H.G. (1998) J. Am. Chem. Soc. 120, Acknowledgements: The work was supported by the European Union and the Fonds der Chemischen Industrie. We thank Dr. Nick Walton, [24] Sprecher, M., Clark, M.J. and Sprinson, D.B. (1964) Biochem.
Institute of Food Research, Norwich, UK, and Professor Meinhart Zenk, University of Munich, Germany, for help with establishing the [25] Reètey, J. and Zagalak, B. (1973) Angew. Chem. Int. Ed. Engl. 12, [26] Hartrampf, G. and Buckel, W. (1986) Eur. J. Biochem. 156, 301^ [27] Sprecher, M., Switzer, R.L. and Sprinson, D.B. (1966) J. Biol.
[1] O'Hagan, D. and Robins, R.J. (1998) Chem. Soc. Rev. 27, 207^ [28] Diziol, P., Haas, H., Reètey, J., Graves, S.W. and Babior, B.M.
(1980) Eur. J. Biochem. 106, 211^224.
[2] Leete, E. (1984) J. Am. Chem. Soc. 106, 7271^7272.
[29] Baldet, P., Alban, C. and Douce, R. (1997) FEBS Lett. 419, 206^ [3] Leete, E. (1987) Can. J. Chem. 65, 226^228.
[4] Ansarin, M. and Woolley, J.G. (1995) J. Chem. Soc. Perkin [30] Guianvarc'h, D., Florentin, D., Tse Sum Bui, B., Nuzi, F. and Marquet, A. (1997) Biochem. Biophys. Res. Commun. 236, 402^ [5] Fries, L. (1962) Physiol. Plant. 15, 566.
[6] Poston, J.M. (1977) Science 195, 301^302.
[31] Shaw, N.M., Birch, O.M., Tinschert, A., Venetz, V., Dietrich, R.
[7] Poston, J.M. (1978) Phytochemistry 17, 401^402.
and Savoy, L.-A. (1998) Biochem. J. 330, 1079^1085.

Source: http://www.ioc.uni-karlsruhe.de/Professoren/Retey/febsletters1998437.pdf