Reprinted from the Journal of the American Chemical Society, 1990,IIZ. Copyright @ 1990by the American Chemical Society and reprinted by permission of the copyright owner.

Variationsin the Heterogeneityof the Decayof the Fluorescencein Six ProcyanidinDimers

DongbwanCbo,t Ru~iangTian,t Lawrence J. Porter,' Richard W. Hemingway,.l and Wayne L. Mattice.'

Contribution from the Institute of Polymer Science. The University of Akron. Akron, Ohio 44325- 3909, Department of Chemistry, Bowling Green State University. Bowling Green. Ohio 43403. Chemistry Division. Department of Scientific and Industrial Research.Petone, New Zealand, and Southern Forest Experiment Station. Pineville. Louisiana 71360. ReceivedOctober 10, 1989

Abstract: The decay of the fluorescence has been measured in 1.4-dioxane for six dimen of (lR.3R)-(-)-epicatechin and (2R.3S)-(+)-catechin. hereafter denoted simply epicatechin and catechin. The dimen are epicatechin-(4,8-+8)-catechin. epicatechin-(4.8-8)-epicatechin. catechin-( 4a--8)-catechin. catechin-( 4a--8 )-epicatechin. epicatechin-( 4,8-6)-epicatechin. and epicatechin-(4,8-+8;2,8-O-7)-epicatechin. The monomers and the bridged dimer have a fluorescence that decays as a single e~ponential. The remaining five dimen have a heterogeneousdecay that can be described by the sum of two exponentials. The heterogeneity is most apparent in the two dimen with 4a--8 interflavan bonds. In view of the molecular origin of the heterogeneousdecay in the presenceof two rotational isomen at the interflavan bond, polymeric JX"CXYanidinswith predominantly a stereochemistryfor the interflavan bond at C(4) should be more disordered and more compact than those with predominantly .8 stereochemistry.

The ~t commonnaturally occurring polymeric procyanidins higher polymers. Upon excitation near 280 nm in dilute solution are polymersof two 2.3-flavan-3-o1s.(2R.3R)-(-)-epicatechin and in 1,4-dioxane, both monomers show a single emission band at (2R.3S)-(+ )-catechin.hereafter referred to simply as epicatechin 320-324 nm.II.12 The decay of the intensity of the nuorescence, and catechin. Thesepolymers are found in the leaves.fruits. and /(t). can be described by a single exponential. barks of many woodyand herbaceousplants.1 Interest in these /(t) = a exp(-tIT) (J) polymersis increasingbecause of their potential as a renewable source of useful chemicals.1their probable use by plants as a The nuorescence lifetimes, T, for the two monomers are indis- defensemechanism.2 and their formation of complexeswith a tinguishable.11 Higher oligomers exhibit an emission maximum variety of naturally occurring and synthetic polymers.3-1 in the same spectral region, but with a lower nuorescencequantum An interflavan bond from C( 4) of one monomer unit to C(8) yield, Q, than that obtained with the monomers.II.12 Although of its neighOOris the ~t commonlinkage between monomer units a ronstrained dimer, procyanidin A J, exhibits a nuorescence decay in the naturally occurring polymers.' Interflavan bonds from that can be described by ~ I, two unconstrained dimers, pro- C(4) to C(6) alsooccur. Figure I depictsthe covalentstructures cyanidins 8 J and 87, exhibit a heterogeneous decay that requires of six dimers of catechin and/or epicatechin. The trivial no- the sum of two exponentials for an adequate description.11 menclature of the six unbridged dimers is procyanidin BI. B2. /(t) = al exp( -t ITI) + a2 exp(-t I T2) (2) B3. 84. B5. and B7. The full nomenclaturedescribed by Hem- ingway et al.9 is epicatechin-(4p--8)-catechin. epicatechin- (4.8-8 )-epicatechin.catechin-( 4a-+8)-a techin. catechin-( 4a-+- (I) Hemingway. R. W., Karcbely. J. J.. Eds. Chemistry and SlglliftCGN:e o/ConJellSu TannillS; Plenum PIal: New York, 1989. 8)-epicatechin. epicatechin-(4p-6)-epicatechin. and epi- (2) Haslam. E. BlocMm. J. 1974. 139. 285. catechin-(4p-6)-atechin. respectively. Thesenames are based (3) Oh, H. I.; Hoff. J. E.; ArmstfODJ, G. S.; Hoff. L. A. J. Agrlc. F()Od on the constituent monomer units and the location and stereo- Chern. 1980, 18. 394. (4) Hagerman, A. E.; Butler. L G. J. BIoi. Chern. 1981. 156,4494. chemistryof the interflavanbond betweenmonomer units. Figure (5) Bergmann. W. R.; Mattice, W. L ACS Symp. SQ. 1987, No. 358, 162- 2 depictsthe covalentstructure of two bridgeddimers. procyanidin (6) Tilstra. L. F.; Maeda. H.; Mattice, W. L. J. CAem.Soc.. Perk/" TrGlU. A I and A2. Their longer names are epicatechin-(4p-8;2P- 11988, 1613. o-7)-catechin and epicatechin-(4p-8;2p-O-7)-epicatechin. (7) olstra, L F.; ClIo. D.; Bergmann, W. R.; Mattice, W. L.ln Chemistry respectively. The bridges from C(2) to C(7) in procyanidin Al olld Slglliftco/lt'e 0/ CoIIIlellSed TollllillS; Hemingway, R. W., Karcbesy, J. J., Eds.; Plenum Press: New York, 1989; P 335. and A2 prevent internal rotation about the interflavan bond. (8) Thompson. R. S.; Jaajues, D.; Hulam. E.; Tanner, R. J. N. J. CArm. Sinceit wasfirst reporteda few yearsago.lo,ll the fluorescence Soc., Perk/" TrollS. 1 1972, 1387. of the oligomeric procyanidins has played an important role in (9) Hemingway, R. W.; Foo, t. Y.; Porter, L J. J. Chem. Soc.. Perki" the developmentof an understandingof the conformationsof the TrollS. I 1982, 1209. (10) Bergmann, W. R.;Barkley. M. D.; Mattice. W. L. PoI,m. Prep'. (Am. Chem. Soc., DID. PoI,m. Chrm.) 1986, 17 (2), 320. (II) Bergmann. W. R.; Barkley, M. D.; Hemingway, R. W.; Mattice. W. 'The University or Akron. I Bowling Green State University. L. J. Am. CAem. Soc. 1987, 109,6614. . Depanment or Scientific and Industrial Research. (12) Cho, D.; Porter, L J.; Hemingway. R. W.; Mattice, W. L. PoI...lMr J. Southern Forest Experiment Station. 1989,30. 19S5. 4274 J. Am. C"~m.Soc.. Vol. 111.No. 11.1990 Cho~, 0/.

two rotational isomers deduced from the analysis of the 81 ("pH 82 fluorescence decay curves for procyanidins BI and B7,Ieads to O HO .0:: the conclusion that the polymers are disordered and more compact HOr("r'° ~ than polystyrene chains of the same molecular weight. ~. ) ~...VOH.04 ,("tON (It ,o~"r'..~ HO 'OH0 0:OH A more extensi\'C examination of thed~y of the nuoresceJU I ," in tbis class of compounds is reported here. Reported for the first time are tbe decays of the nuorescences of procyanidins B2, B3, ~ " c.t OH ""OH 84, 85, and A2. While confirming the earlier conclusions aOOut the origin of the heterogeneous nuorescence d«ay from rotational isomerism, the data for the more extensi\'C set of' dimers show that there is a mucb broader range for the relative contributions of tbe two terms in cq 2 than was apparent from the study of pr<>- cyanidins Bl and 87. When combined with rotational isomeric HO~: state theory, this result bas profound implications for the con- 831GQ::x0:'. formational versatility of the polymeric procyanidins. It implia tbat plants can manipulate the stiffness and the extension of their OH OH polymeric procyanidins over a broad range by manipulation of the location and stereochemistry of the interfiavan bond.

Experimental Methods and Details Catechin and epicatKhin were purchased from Sigma Chemical Co. The details of the preparation and purirlCation of the dimen and char- aCterization via 'JC nuclear maanetic resonance, spccirH: rotation. and FAB mass spectrometry.12 have been described elsewhere. ILlS-I? 1.4- Dioxane (spectrophotometric grade) purchased from Aldrich Chemical Co. was uted without funber purification. FrabJy prepared loIutions were IISecJtbnxaah tbis study due to the seasiiMty of loIutions of the sampia to ezpllUn to lilbt f« lona periods 01 ti~ OoH 6c.. -. Ultraviolet absorption measurements were performed at ambient temperature with a Perkin-Elmer Lambda Array 3840 UV /vis spectro- photometer cquiwcd with a photodilxle array and a deuterium lamp and OH OH interfaced with Perkin-Elmer Computerized Spectroscopic Software. FII- I. ~leat structuresof six un~ dimers.PrOC)'anMSim: 8 Absorbences at 29S nm for the monomen and 290 nm for the dimen. I. which were the wavelengths used for excitation in the nuoresccnce life- epicatechin-(4.8-8)-cat~hin: 82. epicatechin-(4.8-8)-epicat~hin; 83. time measurements. were in the range of 0.01-0.11. catechin-(4a-8)-cat~hjn: 84. catechin-(4a-8)-epicatechin; 8S. epi- Time-ra()!ved nu«accncc measurements achieved by time-correlated catechin-(4.8-6)-epic.t~hin: 87. epicatechin-(48-6)-catechin. single-photon counting"'" were used to determine the decay curves for the nu~ of the monomeric and dimeric procyanidins in solution. The excitation source was a Model 102 ultrafast dye laser (Coherence Co.). synchrtWlOUslypwnpeci by a frequency doubled. mIxic.locked Model Alltara 16-5 Nd:Y AG laser (Cobercncc Co.). A 1220 cavity dumper (Coherence Co.) was IISecJin the dye laser systems to provide selectable ootIKJt IKJlserepetition rates. A Model S-14A autCK:orTClator(Inred Co.) was used to measure the temlMJral width of the IKJlscemitted from a hip repetition rate mcxIc-1ockcd laser. The fwhm (full width. half maximum) or the dye laser pulse wu about 4 S-. The fwhm of the instrumental FcsIX'fISCfunction was about S80 s-. A repetition rate of S.4 MHz wu used with the time-to-amplitude amverler. An R928 photomultiplier. 01 manufactured by Hamamatsu Corp., was used. The filter used was a 30S-nm cutoff filter (WG-30S. Rolyn Arcadia. CA). A counting rate OH of aboot 18K counts/s was used for detecting the decay of the fluora- ccIIcc. The collection of data was amtinued until the decay curves had above 2OK counts in the peak. The d«8y curves were stored in the ranac of G-441 channels. Time calibration was 0.03623 ns/channel. The data ()~ aatuisition time required was usuaRy in the range of 2()-40 s. Electronics manufactured by EG &;. G OnK Co. wen used for photon timina. !to HO('",O~ M~-.n.le conditions were employed. The solvent blank for 1,4-di- ~ ~OH """"~O ozane was subtracted. Measurements were performed at ambient tem- HO~,~O. perature. The standard fluoropilore f« Qlibratioo wu aIIthra~ ill cydo- A2 HO,A OH beune. The lifeti~ 01 the aIItbra~ in cydoIIczallC wu about 4.29 lIS, with a decay described by a single cxJX)ne"tial. This value agreed with the literature valuel' of 4.10 ns. Coff« creamer was used as a scatterer. F1&we 1. Covalent structures or two bridseci dimers. Procyanidins: A I. All samples were maintained at ambient temperature. Data a«luisition epicatechin-(~8:28-O-- 7)-catechin: A2. epicatechin-(4.8-8:28- was controlled by 1xJth multichannel analyzer (MCA) and d~voIution 0-+7)-epicatechin. softwan produced by Edinburgh Instruments, Ltd. (Scotland). A n0n- linear least-iquara method was used to analyze the time-resolved decay The heterogeneousdecay in the unoonstraineddimers arisesfnxn the presenceof two rotational isomersat the interflavan bond, interconversionof which is slow on the time scalefor fluorescence. (14) Bcra-nn. W. R.; Viswanadhao,V. N.: Mattice. W. L J. CAlm. Soc., P,rkilt TraIlS. .' 1988. 45. The relative populationsof the two rotational isomers can be (IS) Lea. A. G. H. J. Sci. Food Alric. 1978.19.471. extracted from the contribution to the fluorescenced~y curves (16) CZOt'banska,Z.; Foo, L. Y.; Newman, R. H.; Porter, L J. J. CAlM. of the two terms in ~ 2.11 Rotational isomeric state analysisof Soc~ P"kilt TraM. I I_, 2278. the higher polymen,l .14utilizing the relative populationsof the (17) Dcrddiocb.. G.; Jcrumanis, J. J. CArtNllGfDl'. 1914. 115, 231. (18) Dcmas, J. N. E~df~ Sfaf~ U/ninw MH.nImrwltts; Academic ~ New York. 1983: Chapter 9. ~ (19) O'CmJI«, D. V.: PbiI1i~ D. Tinw-COI"nlaud Si.," PltoIOIt COUll- (.13).y~w..~n.' V. ~.; Bcrlmann. W. R.; Matti«. W. L. M«rorrto/. liltt:~~~I!em~~: .Nc--:."!o.rk.] 984. , , -. .~- ~ .- LHcay of Ih~ Fluor~sc~nc~in Procyanidin Din'~rs J. Am. Chem.Soc.. Vol. 111,No. II, 1m 4275

T.ba. I. Doublc-ExJK>ncnti.1F1u~ DecaysfM FiveDimcrs in

1.4-Dioxanc- - ! lifetime (ns) . .'

B. ~ D88er.The n~ decaycurve or prtx:yanidin A2 is described by a single ex~nential, eq I, with a T of 1.07:t: 0.01 ns. This value of T is virtually identical with the result (1.06 :t: 0.10 ns) obtained previously for another bridged dimer, pro- cyanidin A 1.11 The only difference in the covalent structure between procyanidin AI and A2 is in the configuration at C(3) of the bottom monomer unit. A single-ex~nential fit for the procyanidin A2 decay curve is depicted in Figure 3, representing the random fluctuation or the residuals around zero with the value of Xl being 1.41. When the double-ex~nential function in eq 2 was assumed to fit the decay curve, neither the XZ value nor the residuals were improved at all. Furthennore, the oontribution by the longer lifetime com~nent was negligible (less than 1%). These results for procyanidin A2 confirm the prior oonclusionll that JX'eYentionor rotation about the interflavan 00cxI in a lxidged dimer results in a single-ex~nential decay for the fluorescence. In oontrast, a heterogeneous decay is seen in most dimers that are not bridged, as described in the next section. C. DiDlers without. Brid&i81 Ri... The fluorescence decay curvu for five unbridged dimers, procyanidin BI, 82, 83, 84, and 85, were measured in 1,4-dioxane. The results are summarized in Table I. The decay curve for epicatechin-(4p-8)-catcchin (procyanidin 8 I) was adequately fitted by a double-ex~nential function, with random fluctuation of the residuals. The major com~nent of the decay has the longer lifetime (1.52 ns), with oflhe nuorcsccnce. The goodnessof fit is determined by the value of the fraction of the intensity from this oom~nent beingabout 94%. x2, which is calculated by Ihe following equation}1 The minor com~nent is the one with the shorter lifetime (0.86 ns). When the present d~y data were fitted to a singie-expo- x1 ~ LW,(R(t) - R.(t)P (3) nential function, a slightly JKK>rerfit was obtained, as indicated Here III, is a statistical wcightin, factor (II R(t» and (R(t). - R(t)] is by the r value of 1.77. The small contribution to the overall do:ay Ihe residual. The residuals must be randomly distributed around zero by the process with the sooner lifetime has a detectable influence. for a good fit. Tbe fraction of Ihe nuorcscenccintensity conlnouled by Fitting this decay curve to the sum of three ex~nentials ~'as an individual componentis determined by impossible because it led to a negative preex~nential factor for /, - o""/L(o,,,,) (4) one of the com~nents. An earlier experiment also found that the decay of the fluorescence of procyanidin 8 I is described by The °1 and "j are varied durin, calculations until the best fit is achicvcd. a sum of two ex~nentials, but the TI and Ji were somewhat 0 neDtwith the nm. The maximum for the bridged dimer was at 277 nm. shorter lifetime and to no signifK:ant imp-ovement in the r (frOOt F1uor~ Ufetime Measurements. A. Monomen. The 1.55 to 1.58) ujX)n changing from a single- to a double-expxiential decaysof the nuorcscenceof the monomers were measuredin fit. Fitting to three ex~nential functions was impossible for the 1,4-dioxane. The decay curves were adequately fitted to a sin- same reason as described ab)ye in the case of epicatechin-(4p- gle-~xponentiald~cay function, with this d~cay function being 8)-<:atechin. ind~pcndentof emissionwavelength in the rang~ of 310-350 nm. The decay curves for catechin-(4a-8)-<:atechin (procyanidin The values of the x1 were 1.55-1.67, the residuals fluctuated 83) and catechin-(4a-8)-epicatechin (procyanidin 84) in 1,4- randomly around l~ro. and the quality of the fit did not improve dioxane were best described by eq 2. When these data were fitted when the decay function was assumedto be a sum of two expo- to singie-exJX}nentialfunctions. the values of the X2 were very large nentials. The valuesof Tare 2.06.i: 0.01 ns for catechin and 2.05 (3.06 for catechin-(4a-8)-<:atechin and 12.17 for catechin- .i: 0.02 ns for epicatechin. PreviouslyBergmann et al,ll reJX)rted (4a-8)-epicatechin). Fitting to the sum of two ex~nentials T'Sof 2.00 .i: 0.05 and 2.04.i: 0.05 ns for catechin and epicatechin. resulted in a large improvement of the Xl values, to 1.37 and 1.46, respectively,in 1,4-dioxane. respectively, and random fluctuation of the residuals around zero. There was no further improvcment in the Xl when the ~ys we~ 4276 J. Am. Chern.Soc., Vol. //2, No. II. /990 Clio et al.

any alteration in the interpretation of the molecularorigin of the decay parameters. The interpretation is that each of the two rotational isomersat the internavan bond is responsiblefor one of the two terms in eq 2. Of course,restriction of the populations to a single rotational isomer, by the introduction of a bridging 1 ring, must produce the simpler decay described by eq I, as 0 measuredinitially for procyanidin A II' and as confirmed here g for procyanidinA2. The populationof the two rotationalisomers I in procyanidin B I was extracted from the two preexponential N factors in eq2. The value of a,(a. + a2)-" 0.75 % 0.15,rejX}rted T earlier I is not oonvincinglydifferent from the value,0.90 % 0.01, E found in the present work. N S a,(a, + a~-' =/,1"2[1", - /,(1", - 1"~]-' (5) I T The resultsreported here fr. prprocyanidin B2. Discussion This range has profound implications for the values of (r}o Three of the compounds(epicatechin, catechin, procyanidin obtained for polymersof high molecular weight becausethe ro- 81) studiedhere with a laseras the sourceof excitation were also tational isomericstate analysis showsthat the meansquare un- studied earlierll with a systemthat employeda deuterium flash perturbeddimensions increase dramatically when one rotational lamp. Comparison can also be made betweenprocyanidins AI isomerhas a strongdominance over the other. In the limit where and A2, becausethese two constraineddimers differ only in the only a single rotational isomer is populated, the chains become stereochemistryof the attachmentof the hydroxyl group at C(3) rodlike with a local helical character. When the experimental of the lower monomeric unit, as depicted in Figure 2. valuesfor al(a. + a2)-1reported here are interpreted in light of The two studies yield experimentally indistinguishable decay the rotational isomeric state analysis,13.14 it is apparentthat the curvesfor epicatechinand catechinin 1,4-dioxane, and the decay conformationsof the high polymers can range all the way from curve for procyanidin A2 in this solventis indistinguishablefrom mol~les that are disorderedand quite compactto moleculesthat the one measuredearlier for procyanidin A I. All of thesedecay are highly extendedwith a local helical structure. Hence.plants curves can be describedby-eq I with a value of T that is about can control both the overall extension and the helical nature of 2.05 ns (for epicatechinand catechin)or 1.07ns (for procyanidins the polymeric procyanidinsthat they produceby selec:tionof the A I and A2). Hence,both systemsyield the sameresults for t~ location and stereochemistryof the interflavan bond. samplesthat have a fluorescencedecay describable by a single The valuesofl. for the five dimef$ in Table I separatenaturally exponential with T > 1 ns. into two grou~. ProcyanidinsBl, B2, and B5 havea,(al + a~-' The decaysobtained for procyanidin 81 in 1,4-dioxane,with > 0.9, and procyanidinsB3 and 84 haveal(al + a~-1 < 0.6. The either system,cannot be successfullydescribed by eq I, but they three procyanidinsin the first group have .8stereochemistry for can be describedby ~ 2. Syntheticdecay curves calculated from the attachment of the internavan bond at C(4), and the two the two setsof decayparameters are comparedin Figure 4. The procyanidinsin the latter group have a stereochemistryfor this differences betweenthe two curves are very small over the first attachment. The results suggestthat high polymers with inter- decade. The value of TI in the presentwork, 1.523 *' 0.004 ns, navan bondswith predominantlya stereochemistrymay be more is very close to the weightedaverage of TI (1.25 *' 0.08 ns) and disorderedthan thosewith predominantly .8stereochemistry. It T2(2.0 *' 0.4 ns) from the previouswork. The presentresults are is difficult to draw any conclusions about the influence of the likely to be the more accurate becausethey were obtained with location of the interflavan bond, i.e., C(4)-C(6) vs C(4)-C(8), instrumentation with a faster instrument responsefunction and becauseonly one dimer, procyanidin B5, has the former type of under conditions wheredata collection required a much shorter interflavan bond. time. Control of the extension and local helical structure may be The minor alteration in the descriptionof the heterogeneityof related to the biophysical role of the polymeric procyanidinsas the decayof the nuorescenceof procyanidin 81 docs not require a plant defensemechanism if it provides a basis for specificity 4277 in the f«mation of a)lnplexcsbet~n the polymericprocyanidins gratefully acknowledge the use of ~uipnent of Prof~ Michad and other types of biopolymers. It may also becomeimportant Rodgers' laboratory at Bowling Green State University. in potential commercial applications of these polymers. Re&istry No. Procyanidin BI, 20315-25-7; procyanidin B2, 29106- Acknowtedlment.This researchwas supported by a grant from 49-8; procyanidin B3. 23567-23-9; procyanidin 84. 29106-51-2; pro- the U.S. Department of Agriculture. 87-FSTY -9-o2S6. We cyanidin B5. 12798-57-1;procyanidin AI. 12798-56-0.