摘要:
1972 919Chemistry of a Colour Test for Abscisic AcidBy R. Mallaby and G. Ryback,. Shell Research Ltd., Milstead Laboratory, Sittingbourne, KentAbscisic acid and its methyl ester are converted by acid-catalysed dehydration into neutral products, one of whichhas been identified as an unsaturated y-lactone. It gives an intense violet-red colour with alkalis, which fades asthe lactone is hydrolysed. This reaction is a useful qualitative test for abscisic acid.WHEN abscisic acid (I) is heated with a mixture of formicand concentrated hydrochloric acids, the major product isthe lactone (11), which in the presence of alkalis gives anintense, characteristic, violet-red colour lasting someseconds or minutes (depending on the concentrations oflactone and alkali). The mass spectrum of (11) re-sembles that of abscisic acid below m/e 246, except thatthe peak a t m/a 111 representing the side-chain of thelatter is absent; and the i.r., u.v., and lH n.1n.r.spectrashow the expected similarities with those of a lactone(111) reported by 0hkuma.l The U.V. spectrum and thechemical shift of the methine H signal, 6 3.41 p.p.m.,exclude structure (IV), that of the likely first product ofdehydration; the isolated lactone (11) can arise from thispresumed intermediate by an allylic rearrangement.The lactone (11) is optically inactive when made from(+)-abscisic acid. (We thank Dr. D. R. Robinson forthis result .)The mixed formic-hydrochloric acid gives a good yieldof lactone (11). With other strong acids, the reaction ismore complex and at present little understood, althoughlactone (11) is always one of the products.Thus, hotformic acid alone converted abscisic acid into a mixtureof neutral products, separated by t.1.c. into three majorcomponents, A ( R p 0.25), B (0.4), and C ( O G ) , withpronounced streaking between them. A, B, C, and thearea between them were coloured violet-red when thechromatogram was sprayed with aqueous sodiumhydroxide. Component B was the lactone (11), but Aand C were unstable and were largely converted into (11)on attempted isolation. A single peak, correspondingto lactone (11), was observed when the mixture wasanalysed by g.1.c. The structures of A and C are notyet known; they appear to be lactones, C at leastan isomer of (11).A similar pattern of products, oftenwith additional components, was seen when otherreagents were used : hot aqueous-alcoholic mineral acids,1 K. Ohkuma, Agvic. a d Biol. Chem. (Jnpait), 1965, 29, 262;1966, 30, 434.boron trifluoride-ether (which reacted only if dilutedwith, e.g. benzene), sulphonic acid ion-exchange resinsin hot dichloromethane, molten oxalic acid, etc ; weakeracids such as hot acetic acid effected no dehydration.Methyl abscisate behaved similarly, being converted ingood yield into the lactone (11) if heated with a formic-hydrochloric acid mixture, and in rather lower yield andaccompanied by other products if formic acid alone wasused. The 2-trans-isomer of abscisic acid is also readilydecomposed by hot formic acid-not, however, intoneutral products but into a t least five as yet unidentifiedacids, one of which gives a strong violet-red colour withalkalis.The colour of alcoholic solutions of lactone (11) whenaqueous sodium hydroxide is added is due to theappearance of an absorption band at 577 nm, which weattribute to the formation of the extended delocalisedanion (V) by removal of a proton from C-1'.Forcomparison, the open-chain, unsubstituted anion-O*[CH=CH],*CHO absorbs at 547.5 nm in dimethyl-formamide.2 An analogous reaction (pink colour,A,, 507 nm, 1hfl 570 nm) is given by a mixture ofmethyl 1 '-deoxyabscisate (VI) and its 2-lrans-isomer.If dilute (104-10-3~) solutions of lactone (11) are used,the colour develops slowly, then fades as the lactone ringis hydrolysed (cj.below). For reasons which are notquite clear, a strong colour is obtained only in thepresence of water; sodium ethoxide in dry ethanol givesbut a faint colour, which intensifies when water is added.It was found that a high absorbance reading at 577 nmcould be obtained by adding 0.12 ml of 2N-aqUeOUSsodium hydroxide to a few micrograms of lactone (11)in 2.5 ml ethanol; the colour reached its maximumintensity (E 11,800, based on weight of lactone) in ca. 10min, before fading slowly. The by-products A and Cmentioned above give similarly coloured solutions, withhm, -577 nm, and it is probable that they give rise tothe same anion as does lactone (11).An interesting incidental observation was the appear-ance, during the first few minutes of heating, of a rela-tively weak but distinct violet-red colour when methylabscisate was heated with formic acid.This colour(maximal absorption at 560 nm in formic acid) wassimilar to that given by lactone (11) in the presence ofstrong bases, suggesting a chromophore of the same typeas that discussed above. If (VII) is invoked as an inter-mediate in the demethanolation, removal of a protonfrom C-4 by formate acting as a base, in competition with2 S. S. Malhotra and >I. C. Whiting, J . Chew. Soc., 1960,3812920 J.C.S. Perkin I1removal of the methyl group, could give small concentra-tions of the zwitterion (WIT) which would account forthe transient colour.We hoped to devise a simple spectrophotometi-ic assayfor abscisic acid based on this colour reaction.Althoughhydrolysis and fading could not be avoided (in lion-hydroxylic solvents, e.g. using sodium hydride in tetra-hydrofuran or dimethyl sulphoxide, a violet precipitateand calourless supernatant were obtained), microgramamounts of the pure lactone (11) could be estimated bystandardising the measurements and using solutions ofknown concentration for calibration. However, thetotal dehydration products from abscisic acid behaveddifferently when alkali was added (even if a formic-hydrochloric acid mixture was used for the dehydration) ;the colour (AmL still at 577 nm) appeared instantaneously,the initial rate of fading was very rapid and no meaning-ful absorbance value could be measured.This is due tothe presence in the mixture of minor by-products,among them the component A mentioned above, whichreact more rapidly with the base and appear to give arelatively more intense colour than does lactone (XI). Anassay in which lactone (11) is isolated by t.1.c. and esti-mated spectrophotometrically or by g.1.c. would bepracticable but time-consuming, dependent on a constantyield of the lactone and therefore unsatisfactorj-. As asensitive qualitative test for abscisic acid or esters, thereaction is useful. The sample to be tested is heated withformic acid, which is then removed in vacuo; the residueis transferred to a depression in a white tile with ether orethanol; and a drop of aqueous-ethanolic sodiumhydroxide is added.A test on 0.1 pg of pure abscisicacid gives a distinct, transient violet-red colour, andsomewhat larger amounts of the hormone can be readilydetected in this way in fractions of plant extracts thatare not themselves too strongly coloured. The colourobtained with 2-trans-abscisic acid is similar but lessintense. We have not yet met any other compound inthe acidic fraction of plant extracts which gives apositive reaction.* The numbering shown in formulae (11) and (XI) , and used forn.m.r. results, is based on that of abscisic acid.A spectrofluorimetric assay for abscisic acid isbased on the fact that it gives green-fluorescing spots ont.1.c. plates that have been sprayed with dilute sulphuricacid and heated. Similarly, during our dehydrationexperiments, the reaction mixtures usually acquired anintense green fluorescence.The origin of this is notclear, since lactone (11) does not fluoresce (even when re-heated with acids), and only bhe-fluorescing minor by-products have been detected when the products wereanalysed by t.1.c.The fading of alkaline solutions of lactone (11) isaccompanied by hydrolysis of the lactone ring, andnothing can be extracted from the faded solutions withether. After acidification, however, a new lactonewhich is isomeric with abscisic acid can be isolated. Itsspectroscopic properties indicate structure (XI) ; thei.r. spectrum shows that hydroxy-groups are absentand there are no n.m.r. signals ascribable to hydrogensattached to oxygen-bearing carbon atoms.Presumablylactone (11) is hydrolysed to a keto-acid salt (IX andtautomers) which cyclises, via a lactol (X), when thesolution is acidified and worked up. The formation ofphaseic acid provides an analogy for the step (X) +(XI). When heated with a mixture of formic and con-centrated hydrochloric acids, the spiro-lactone (XI)regenerates the dehydrated lactone (11), detected by thecharacteristic violet-red colour with alkalis (the spiro-lactone gives a bright yellow colour), by the reappearanceof an absorption band at 277 nm, and by t.1.c. analysis.EXPERIMEN TXLM.p.s are corrected. Thin-layer chromatography was onKieselgel-GF,,, (Merck) developed in 3 : 2 (v/v) lightpetroleum-EtOAc. Abscisic acid and derivatives wereracexnic.A sample of methyl 1'-deoxyabscisate (mixtureof 2 4 s - and 2-fvans-isomers) was kindly provided by Dr.M. Anderson, Shell Research Ltd. We thank Mrs. J. Tuckerfor mass spectra (A.E.I. MS-9) and Mr. D. M. Barnett forp.m.r. spectra (Varian HA-100).2,5-Dihydro-4-methyl-5- (2,6,6-trimethy1-4-oxocyclohex-2-enylmethylene) furan-2-one (11) *.-(a) Abscisic acid (200 mg),formic acid (2 nil), and concentrated hydrochloric acid(0-3 ml) were heated a t 95" for 30 min, and then evap-orated under reduced pressure a t 95". The residue wastaken up in ether and washed with aqueous NaHCO,(acidification and re-extraction of which yielded negligibleamounts of material). Removal of the ether, decolouris-3 R. Antoszewski and R. Rudnicki, Analyt. Biochem., 1969,32, 233.B.V. Milborrow, Chem. Comm., 1969, 9661972ation with charcoal in ethanol, and recrystallisation fromether-methanol gave the lactone (11) (107 mg) as colourlessneedles, moderately soluble in ether, m.p. 130--131*5O(Found: C, 72.7; H, 7.1. C,,H,,O, requires C, 73.1; H,7.4%); Amax. (EtOH) 277 nm (E 22,400), shoulders a t 238(E 10,000) and 233 ( E 9500) nm; vmsL (CS,) 1783vs, 17SOinfl,and 1666vs cm-l; 8 (CDCI,)* 1.01 and 1.10 (each 3 H, s, 6’,6’-di-CH3), 1-92 (3H, d, J 1-5 Hz, 2‘-CH,), 2-21 (3H, d, J1.5 Hz, 3-CH3), 2.18 and 2.41 (AB, each lH, d, J 17 Hz, 5’,-5-H), 5-92 (lH, broad, 2-H), and 6.05 p.p.m. (lH, broad,3’-H); m/e 246 (rel. int. 8%; M+), 231 (1-5), 190 (loo),162 (3.6), 147 (2*6), 134 (12*6), 119 (2-3), 106 (3), 91 (6*6),metastables a t 146.7 (246 --t 190), 138.1 (190 ---t 162),110.9 (162 --w 134), 105.7 (134- 119), 94-9, 83-8 (134---t 106), 78.1 (106 + 91), 76.2.The 2,kdinitro-pkenylhydrazone had m.p. 226-228”, Lx. (EtOH) 381 and270 nm, vmX. (Nujol) 1770 cm-1.( b ) Abscisic acid (653 pg), formic acid (0.1 ml), and con-centrated hydrochloric acid (0.01 ml) were heated a t 100”for 30 min, and then evaporated in vacuo. The residuewas separated by t.1.c. and the major u.v.-absorbing band,coinciding with a marker of lactone (11), was scraped offand eluted with ethanol; the yield was estimated spectro-photometrically a t 81 yo.(G) Methyl abscisate (425 pg) was treated exactly as in( b ) , above. The yield of lactone (11) was 86% (spectro-photometric); it gave the correct U.V.and mass spectral5’-H,), 3-41 (lH, d, J 11 Hz, 1’-H), 5-18 (lH, d, J 11 Hz,921characteristics and gave a violet-red colour with aqueous-alcoholic sodium hydroxide.benzo[b] furan-2-spiro-2’-fu~an-5‘,6-dione (XI) * .-The lactone(11) (42 mg) in ethanol (2 ml) and N-aqueous NaOH (1 ml)was left a t room temperature until the initial intense violetcolour had faded to a pale red (-1 h) . Partial evaporation,dilution with water, and ether-extraction yielded < 0-5 mgof material. Acidification of the aqueous layer and furtherextraction gave an oil (44 mg) from which the laclone (XI)(30 mg) was obtained as colourless needles, m.p. 146-148O, by crystallisation from ether (Found: C, 68.2; H,7.9. C,,H,,O, requires C, 68.2; H, 7.6%); the U.V.spectrum (EtOH) showed strong end-absorption only ;vmX. (CCl,) 1779vs, 1721vs cm-l; 6 (CC1,)* 1.07 and 1-20(each 3H, s, 6’,6’-di-CH3), 1.60 (3H, s, 2’-CH,), 2.00 (3H, d,J 1.5 Hz, 3-CH3), 2.1-2.7 (7H, complex), and 5.85 p.p.m.(lH, J 1.5 Hz, 2-H); m/e 264 (rel. int. 37%; M+), 249 (21),231(2), 222(2), 220(3), 208(14), 207(100), 190(4), 189(4),166(12), 165(37) (metastable for 207 + 165 at 131.6),164(15), 154(21), 152(12), 151(18), 124(18), 122(18), 111(18),110(8), 99(12), 96(26), 95(15), 83(24), 68(24), 55(12), 43(44).2,2‘, 3,3a,4,5,6’, 6,7,7a-Decahydro-3‘,4,4, 7a-tetramethyl-\Ye are grateful to Professor J. W. Cornforth for valuable[1/2400 Received, 14th Decenrbev, 19711discussions, and to a referee for pointing out reference 2.* See footnote on page 920
ISSN:1472-779X
DOI:10.1039/P29720000919
出版商:RSC
年代:1972
数据来源: RSC