Plant Physiol. (1992) 100, 902-9070032-0889/92/100/0902/06/$01 .00/0

Received for publication February 11, 1992Accepted February 21, 1992

Flavonols Stimulate Development, Germination, and TubeGrowth of Tobacco Pollen'

Bauke Ylstra2, Alisher Touraev, Rosa Maria Benito Moreno, Eva Stoger, Arjen J. van Tunen, Oscar Vicente,Joseph N. M. Mol, and Erwin Heberle-Bors*

Institute of Microbiology and Genetics, University of Vienna, Vienna Biocenter, Dr. Bohrgasse 9, 1030Vienna, Austria (B.Y., A.T., R.M.B.M., E.S., O.V., E.H.-B.); and Department of Genetics, Free University Amsterdam,

de Boelelaan 1087, 1081 HV Amsterdam, The Netherlands (A.J.v.T., J.N.M.M.)

ABSTRACT

The effect of anther-derived substances on pollen function wasstudied using pollen produced by in vitro culture of immaturepollen of tobacco (Nicotiana tabacum L.) and petunia (Petuniahybrida). Addition of conditioned medium consisting of diffusatesfrom in situ matured pollen strongly increased pollen germinationfrequency and pollen tube growth, as well as seed set after in situpollination. Thin-layer chromatography and depletion of phenolicsubstances by Dowex treatment indicated that flavonols are presentin the diffusate and may be the active compounds. When added tothe germination medium, flavonols (quercetin, kaempferol, myri-cetin) but not other flavonoids strongly promoted pollen germina-tion frequency and pollen tube growth in vitro. The best resultswere obtained at very low concentrations of the flavonols (0.15-1.5 Mm), indicating a signaling function. The same compounds werealso effective when added during pollen development in vitro.

Male gametophyte and gamete formation in plants occursby a close interaction with the surrounding sporophytic tis-sues, particularly the tapetum (1, 14, 28). A variety of factorshave been suggested to play a role in this interaction. How-ever, their function is still largely unknown.

Flavonoids are secondary plant products that include pig-ments (chalcones, anthocyanins) and colorless compounds(flavanones, flavones, and flavonols) that are involved inpollination, seed dispersal, UV light protection, and plant/pathogen interaction (9, 11, 23). Flavonoids are present inpollen of many species of angiosperms and gymnosperms, aswell as in spores of ferns and mosses (29). Flavonoid biosyn-thesis is initiated by chalcone synthase, followed by thesynthesis of flavanones by chalcone isomerase. Recently, itwas shown that the chs and chi genes are coordinately

l This work was supported by the Austrian 'Fonds zur Forderungder wissenschaftlichen Forschung', the 'Forschungsforderungsfondsfur die gewerbliche Wirtschaft, and 'btF-biotechnologische For-schungsges. m.b.H.,' Linz, Austria. B.Y. was partly financed by aStimuleringsprogramma voor Intemaltionalisering van het HogerOnderurijs TIR grant of the Dutch Ministry of Education and Science.

2 Present address: Center for Plant Breeding and ReproductionResearch (CPRO-DLO), P.O. Box 16, 6700 AA Wageningen, TheNetherlands.

expressed in the anther, with a peak of expression in themid-binucleate stage of pollen development (8, 23-25). Asimilar expression pattern was also observed in transgenictobacco plants (24; B. Ylstra, E. Stoger, unpublished obser-vation). Petunia plants lacking chs gene expression are malesterile (19, 21), indicating that anther-derived flavonoids mayplay a role in male gametogenesis.

In vitro culture of isolated microspores has shown that animportant function of the anther for the development of themicrospores and pollen grains is the stepwise provision oflow mol wt nutrients and anabolic precursors (2, 4, 18, 20).Germination frequency and seed set of this pollen are, how-ever, lower compared with mature pollen taken directly fromthe plant (2, 20). Obviously, in vitro pollen lacks factors thatare provided in vivo by the sporophyte. The in vitro pollencan thus be considered as 'minimal pollen' that fulfills theminimal requirements for pollination and fertilization butlacks factors for optimal reproductive success (28). It should,therefore, react very sensitively to added compounds that invivo are provided by the anther wall. Here, we show thatflavonols but not other flavonoids produced by the antherare present in diffusates of mature tobacco (Nicotiana tabacumL.) pollen and have a strong stimulatory effect on in vitropollen development, pollen germination, and pollen tubegrowth.

MATERIALS AND METHODS

In Vitro Culture of Immature Pollen

Microspores or young binucleate pollen grains were iso-lated from tobacco (Nicotiana tabacum L. cv Petit HavanaSR1) and petunia (Petunia hybrida cv Wi 15) flowers. Micros-pores were cultured in a MR26 medium consisting of theMurashige-Skoog (16) minerals, 0.5 M sucrose, 3 mm gluta-mine, 100 mg. L` of inositol, 2% coconut water, and 1 g* L`of lactalbumin hydrolysate (pH 7) in wells (330 ,L of pollensuspension at a density of 105 grains/mL) of tissue clusters24 (Costar, Cambridge, MA) at 280C in darkness. Youngbinucleate pollen grains were cultured in a AMGLU mediumconsisting of Murashige-Skoog minerals, 0.5 M sucrose, and3 mm glutamine (pH 7) to maturity following the procedureof Benito Moreno et al. (2). Maturity was judged by micro-scopical analysis (33 ,m in diameter, fully packed with starch

902

FLAVONOLS AND POLLEN

grains) and by germination ability in a germination medium(Brewbaker-Kwack [3] minerals with a double amount ofboric acid, 10% sucrose, GK medium). In situ pollination withthe in vitro matured pollen was done as described by BenitoMoreno et al. (2).To obtain a homogeneous population of mid-binucleate

tobacco pollen (no starch grains, nucleolus in vegetativenucleus present, generative cell detached from intine), pollenwas isolated from flower buds of 18 to 19 mm in length intoa solution of 0.4 M mannitol. After two washings, the pollensuspension was layered on top of a 60% Percoll in 0.4 M

mannitol solution and was centrifuged at 300g for 15 min.The remaining pollen on top of the Percoll solution was

removed, washed three times, and cultured in AMGLU me-

dium with or without flavonoids.

In Vitro Germination Assays

The in vitro pollen was washed twice and transferred toGK medium. After several hours of incubation, the pollengrains had formed tubes and were analyzed. Each experimentwas repeated at least three times. The quantitative data are

from one representative experiment. The pollen with tubeswere counted, and pollen tube length was measured on

projected slides. To determine germination frequency, usually500 pollen grains were counted. Tube length of 50 to 100pollen grains was measured. Only tubes longer than half thesize of a pollen grain were judged as germinated.

Preparation of Conditioned Medium

Pollen was collected from open anthers and incubatedovernight in GK medium at a density of 105 pollen grains/mL. The pollen was removed by centrifugation, and theresulting medium was filter sterilized. Mature pollen was alsoisolated from closed tobacco flowers just before anthesis bygently crushing the anthers in GK medium under asepticconditions to obtain a suspension of a density of 4 x 1 05/mL. This suspension was stirred by vortexing for 5 min. Thepollen was removed by centrifugation, and the medium was

filter sterilized.

Treatments of the Conditioned Medium

A strongly basic anion exchange resin (Dowex, Sigma, 50-100 mesh, stock No. lx2-100) was added to the conditioned

medium and was then gently shaken for 10 min. The resinwas removed by centrifugation, and the medium was filtersterilized. The conditioned medium was dialyzed (SpectraPOR 1, mol wt cutoff 6000-8000) twice for 8 h against 100volumes of fresh GK medium at 0°C. For heat inactivation,the samples were heated for 10 min at 800C.

Flavonoids

Naringenin chalcone (4,2',4',6'-tetrahydroxychalcone[21]), naringin, and naringenin were isolated from petuniaflowers (8). Quercetin (3,3',4',5,7-pentahydroxyflavone dih-ydrate), kaempferol [3,5,7-trihydroxy(4-hydroxyphenyl)-4H1-benzopyran-4-one] and myricetin (3,3',4',5,5',7-hexahy-droxyflavone) were purchased from Sigma. All flavonoidswere dissolved in DMSO immediately before addition to themedia. In control experiments, no effect of DMSO (<0.3%)on pollen germination was observed.

Detection of Flavonoids

TLC (8) was used to detect flavonoids in the conditionedmedium. The sucrose in the GK medium was omitted in theseexperiments. After chromatography, the TLC plates were

placed on a UV (312 nm) transilluminator and photographed.

RESULTS

Pollen Diffusates Stimulate in Vitro Pollen Germinationand in Situ Seed Set

In initial experiments, a conditioned germination mediumproduced by incubating mature in vivo pollen overnight inthe germination medium and added to in vitro maturedtobacco pollen promoted pollen germination in vitro and seedset after in situ pollination (Table I). The same effect was

achieved by a conditioned medium obtained by 5 min ofvigorous shaking of pollen isolated from nearly mature an-

thers. Tobacco pollen tubes grew 2.5 times as fast in theconditioned medium (Fig. 1, middle) as they did in thenonconditioned germination medium (Fig. 1, top).

Flavonols Stimulate in Vitro Pollen Germination

To identify the germination-promoting compounds presentin mature pollen, the conditioned medium was subjected toseveral treatments. No effect of heat treatments was ob-

Table 1. Effect of a Conditioned Germination Mediuma (GKc) on Germination Frequency in Vitro andon Seed Set after in Situ Pollination of Minimal Pollen Produced by in Vitro Culture of IsolatedTobacco Microspores

No. of No. of Seed Pods No. ofGermination . No. of per Seeds perMedium Pollen Pollina-. SedprGrains Frequency tions Seed Pods Pollina- Pod

tions

%b %

GK 507 19.3 25 4 16 100-200GKc 390 54.8 26 13 50 200-500

a Made from pollen from open anthers incubated overnight. b Similar data were obtained fromthree repetitions.

903

Plant Physiol. Vol. 100, 1992

A fc

4li

V* . is.. H qi *,;' o i' (>7 ,$

*qq* F

bear.:. '., ..;t' Is

F COT- g , §^\r

be at .is' "hi. +

i .' ;. ., .d ..

To identify more precisely the type of compounds involvedin growth stimulation of tobacco pollen tubes, a TLC analysiswas performed. Flavonoids are known to be present in pollen(29), and in fact, TLC revealed the presence in the conditionedmedium of compounds with the mobility and color of theflavonols quercetin and kaempferol used as standards (Fig.

:*^ 3). To show directly the effect of flavonols on pollen tubegrowth, various flavonols and other flavonoids were addedto minimal pollen produced by in vitro culture of isolatedmicrospores. All of the flavonols tested (quercetin, kaemp-ferol, myricetin) had a strong stimulatory effect on germina-tion frequency and pollen tube length (Figs. 1, top andbottom, and 4) at concentrations ranging from 0.15 to 1.5A-M. Maximum restoration of the effect of the conditionedmedium was 80%. The other flavonoids tested (naringeninchalcone, naringin, naringenin) had no effect (data notshown).

Addition of the tobacco pollen diffusate and of flavonolsto in vitro matured petunia pollen also had a stimulatoryeffect on pollen germination (data not shown). Petunia pol-len, however, usually gave a less efficient response. The poorsynchronism of in vivo pollen development in this speciesdid not allow homogeneous starting populations. Also, thein vitro culture conditions established for tobacco pollen may

at, not have been adequate for petunia pollen.

C

'K>."

A..*O

Figure 1. Minimal pollen germinating overnight in control GK me-

dium (top), in conditioned GK medium made from pollen isolatedfrom closed anthers by 5-min incubation (middle), and in GKmedium containing 1.5 AM quercetin (bottom).

served, indicating that heat-labile substances (including mostproteins) were not directly involved in growth control oftobacco pollen tubes (Fig. 2A). Dialysis produced a reductionof the effect of the conditioned medium, suggesting that lowmol wt substances are involved in growth control. The effectof dialysis was, however, small. Depletion of compoundsfrom the conditioned medium by incubation in the presenceof a Dowex resin that can bind phenolic substances reducedpollen germination frequency and pollen tube growth to thelevels of the nonconditioned germination medium serving as

a control (Fig. 2B).

Flavonols Stimulate in Vitro Pollen Development

Flavonols were also tested for their effect on tobacco pollendevelopment in vitro. A highly homogeneous population ofmid-binucleate pollen produced by density centrifugation inPercoll was cultured in AMGLU medium, and aliquots ofpollen suspension were transferred to germination mediumat different time intervals. Germination frequency was deter-mined 10 h after transfer to germination medium. After 48 hof development, no pollen tubes were seen, indicating thatthe pollen was not yet mature (Table II). After 60 h of pollendevelopment in control medium, some pollen (6.2%) germi-nated, whereas pollen developed in quercetin-containing me-dium germinated at a frequency of 40.9%. Pollen tubes ofquercetin-treated pollen were longer than in the control (Fig.5). After 72 h, 25.5% of the pollen developed in controlmedium germinated, whereas 59.9% germinated when pollendevelopment took place in quercetin-containing medium.Kaempferol and myricetin also stimulated pollen develop-ment but to a lower extent. These results indicate that mid-binucleate pollen reach maturity (maximum germination fre-quency) within 72 h of in vitro culture and that flavonolsreduce the time required to reach maturity.

DISCUSSION

The data presented in this article strongly suggest that theeffect of pollen diffusates on pollen germination and tubegrowth is largely due to the presence of flavonols in maturepollen. Other flavonoids such as chalcones, flavanones, andflavones had no such effect. We have shown that the lowmol wt flavonoid compounds stimulate pollen germinationand tube growth. On the other hand, removal of smallmolecules by dialysis reduced the germination-promoting

904 YLSTRA ET AL.

--. u,,E, 4

.... .!, ", -.g?NO

..s

.;. A-lw-41'

1.O.,

t'.:. I '. ..

FLAVONOLS AND POLLEN

gk preconditioned dialysis heat (80C)

0)(0C0

40

-20 .ac

-3

C

0

10

c0CT

EL0

10)7mc

BE

r-

.-0c

0)

D

4,

m

4,Q0

gk (control) preco. + dowex preconditionedTreatm ent Treatment

Figure 2. A, Effect of dialyzed or heat-treated conditioned media (made from pollen from closed anthers by 5-min incubation) on germinationof minimal pollen. Germination frequency, average; pollen tube length, average ± SD. B, Effect of Dowex-treated conditioned medium (madefrom pollen from closed anthers by 5-min incubation) on germination of minimal pollen. Germination frequency, average; pollen tube length,average ± SD. preco., Preconditioned medium.

S

* ,0,*.*w

I

3 41 2

I start

OHNoI toH myricetin

OH 0OH

HO_1 OH quercetin

OH 0

HO- 0oH kaempferol

OH 0

effect of the diffusate only a little. This may indicate thathigh mol wt compounds, i.e. lipids or proteins (6), may beadditional factors. Possibly, their effect is by binding theflavonols.An interesting finding was that the effective concentrations

were very low, i.e. in the micromolar range. This fact, togetherwith the chemical specificity, suggests that the flavonols actas signal molecules, similar to plant hormones. Even the thirdclassical definition of a hormone, apart from effectiveness atlow concentration and chemical specificity, applies to theflavonols: synthesis at one location and action at another.The genes for flavonoid biosynthesis in petunia and tobacco

-

E

-E

r-

.-i0

0

front

samples

Figure 3. TLC of pure flavonols (sample 1, kaempferol; sample 2,quercetin; sample 3, mixture of kaempferol, quercetin, and myri-cetin) and of tobacco pollen diffusates (sample 4). In addition to RFvalue, the specific yellow fluorescence in UV light was diagnosticfor flavonols. The two lower spots in the pollen diffusate samplewere brown in visible light and did not fluoresce. They seem to besugars.

gk (control) kaempferol quercetin myricetin

Addition

Figure 4. Effects of flavonols on germination of minimal pollenwhen added to the germination medium. Germination frequency,average; pollen tube length, average ± SD.

AE 800

C

0)c 6000

0

cx0 400-

200-

0'

905

0Im

-15 c

-0v

L-

L.100W

CT0

'5 cEL-

-1'40

c

EL-

cn

Plant Physiol. Vol. 100, 1992

4M4k 7

4~.

Figure 5. Germination of minimal pollen in GK medium (after 6 h) that had developed for 60 h in AMGLU medium (left) and for 60 h inAMGLU medium plus 10 AM quercetin (right).

anthers are expressed in the sporophyte, i.e. in the antherwall tissues (22; B. Ylstra, E. Stbger, unpublished observa-tion), and the flavonoids synthesized by the tapetal flavonoidenzymes are incorporated in the exine cavities of the pollen(30). Flavonoid action, however, is in the pollen grains afterthey have landed on a stigma, as suggested by their stimu-latory effect on pollen germination. For these reasons, wesuggest that during the development of the male gameto-phyte, flavonols act as plant growth regulators or planthormones.

It is interesting that the flavonols also had an effect onpregermination pollen development, indicating a second con-trol point for flavonols. This coincides with the finding thatchs-antisense mutants (with a chimeric anther-specific pro-moter) are male sterile caused by an arrest in pollen devel-opment (21).

In chs-cosuppressor mutants (introduction of a chs trans-gene into wild-type petunia), seed set was prevented,whereas pollen development was apparently unaffected (19).Male fertility was restored when mutant pollen was used topollinate wild-type stigmas, probably by uptake of flavonolssynthesized by the stigma (conditional male fertility). Wepredict that restoration can also be achieved by using the so-called mentor effect (7), i.e. by mixing the mutant pollen with

Table II. Germination Frequencya of in Vitro Pollenb after ThreeTime Intervals in the Presence of Flavonols (all 10 lsM)The data are means ± SE of three experiments; six Petri dishes

were used for each treatment.Time AMGLU AMGLU + AMGLU + AMGLU +

Interval Control Quercetin Kaempferol Myricetinh % % % %48 0 0 0 060 6.2 ± 0.76 40.9 ± 2.5 26.5 ± 0.7 30.3 ± 1.672 25.5 ± 1.8 59.5 ± 1.7 26.6 ± 1.5 36.1 ± 2.4

a Ten hours after transfer to GK medium. b A highly homo-geneous population of mid-binucleate pollen produced by Percollcentrifugation.

wild-type pollen to pollinate mutant stigmas. Flavonols pro-duced by the wild-type pollen should promote pollen tubegrowth of the mutant pollen, resulting in fertilization andseed set.

Little is known about the mechanism of action of flavonolsas signal molecules (11). Recently, Rubery (17) proposed thatquercetin and other flavonoids bind as natural ligands to thephytotropin or N-1-naphthylphthalamic acid receptor (5,confirmed in ref. 26), which interacts with and controls theauxin efflux carrier in the plasma membrane involved inpolar auxin transport (10, 15). We imagine that in pollen ablock of the auxin efflux carrier by flavonols increases theintracellular concentration of auxin, which, in turn, promotesthe polar tube growth. This view is corroborated by Vesperand Kuss (27) who found in the presence of N-1-naphthyl-phthalamic acid a stronger stimulation by IAA of elongationgrowth of maize coleoptiles. Alternatively, pollen tubegrowth promotion may result from the inhibition by flavon-oids of oxidative destruction of IAA (12, 13), resulting againin an increased endogenous level of auxin. Further experi-ments are required to unravel the precise mode of action offlavonols in pollen development.

ACKNOWLEDGMENT

We are thankful to Bernd Zbell, Heidelberg, Germany, for valuablesuggestions.

LITERATURE CITED

1. Albertini L, Souvre A, Audran JC (1987) Le tapis de l'anthereet ses relations avec les microsporocytes et les grains de pollen.Rev Cytol Biol Veg Bot 10: 211-242

2. Benito Moreno RM, Macke F, Alwen A, Heberle-Bors E (1988)In situ seed production after pollination with in vitro maturedisolated pollen. Planta 176: 145-148

3. Brewbaker JL, Kwack BH (1963) The essential role of calciumion in pollen germination and pollen tube growth. Am J Bot50: 859-865

4. Heberle-Bors E (1989) Isolated pollen culture in tobacco: plantreproductive development in a nutshell. Sex Plant Reprod 2:1-10

906 YLSTRA ET AL.

FLAVONOLS AND POLLEN

5. Jacobs M, Rubery PH (1988) Naturally occurring auxin transportregulators. Science 241: 346-349

6. Kirby EG, Vasil IK (1979) Effect of pollen-protein diffusates ongermination of eluted pollen samples of Petunia hybrida invitro. Ann Bot 44: 361-367

7. Knox RB, Gaget M, Dumas C (1987) Mentor pollen techniques.Int Rev Cytol 107: 315-332

8. Koes RE, van Blokland R, Quattrochio F, van Tunen AJ, MolJNM (1990) Chalcone synthase promoters in Petunia hybridaare active in pigmented and unpigmented cell types. PlantCell 2: 379-392

9. Kuhn DN, Chapall J, Boudet A, Hahlbrock K (1984) Inductionof phenylalanine ammonia-lyase and 4-coumarate:CoA ligasemRNAs in cultured plant cells by UV induction or fungalelicitor. Proc Natl Acad Sci USA 81: 1102-1106

10. Lembi CA, Morre DJ, Thomson K-S, Hertel R (1971) N-1-naphthylphthalamic-acid-binding activity of a plasma mem-brane-rich fraction from maize coleoptiles. Planta 99: 37-45

11. Long S (1989) Rhizobium-legume nodulation: life together in theunderground. Cell 56: 203-214

12. Marigo G, Boudet AM (1977) Relations polyphenols-croissance:mise an evidence d'un effect inhibiteur des composes pheno-liques sur le transport polarise de l'auxine. Physiol Plant 41:197-202

13. Marigo G, Boudet AM (1979) Effects of an increase in levels ofphenolic compounds on the auxin content and growth ofLycopersicum esculentum. Z Pflanzenphysiol 92: 33-38

14. Mascarenhas JP (1990) Gene activity during pollen develop-ment. Annu Rev Plant Physiol Plant Mol Biol 41: 317-338

15. Michalke W, Schmieder B (1979) Fractionation of particulatematerial from maize coleoptile homogenates with polyethyl-ene glycol. Planta 145: 129-135

16. Murashige T, Skoog E (1962) A revised medium for rapidgrowth and bioassays with tobacco tissue cultures. PhysiolPlant 15: 473-497

17. Rubery PH (1990) Phytotropins: receptors and endogenous li-gands. Soc Exp Biol Symp 44: 119-146

18. Stauffer C, Benito Moreno RM, Heberle-Bors E (1991) Seedset after pollination with in-vitro-matured, isolated pollen ofTriticum aestivum. Theor Appl Genet 81: 576-580

19. Taylor LP, Jorgensen R (1992) Conditional male fertility inchalcone synthase-deficient petunia. J Hered 83: 11-17

20. Tupy J, Rihova L, Zarsky V (1991) Production of fertile tobaccopollen from microspores in suspension culture and its storagefor in situ pollination. Sex Plant Reprod 4: 284-287

21. van der Meer IM, Stam ME, van Tunen AJ, Mol JNM, StuitjeAR (1992) Inhibition of flavonoid biosynthesis in petuniaanthers by antisense approach results in male sterility. PlantCell 4: 253-262

22. van Tunen AJ, Mol JNM (1987) A novel purification procedurefor chalcone flavanone isomerase from Petunia hybrida andthe use of its antibodies to characterize the Po mutation. ArchBiochem Biophys 257: 85-91

23. van Tunen AJ, Mol JNM (1990) Control of flavonoid synthesisand manipulation of flower colour. In D Grierson, ed, PlantBiotechnology Series. Blacky and Son, Glasgow, Scotland, pp94-130

24. van Tunen AJ, Mur LA, Brouns GS, Rienstra J-D, Koes RE,Mol JNM (1990) Pollen and anther-specific chiA and B pro-moters from Petunia hybrida: tandem promoter regulation ofchiA gene expression. Plant Cell 2: 393-401

25. van Tunen AJ, Mur LA, Recourt K, Gerats AGM, Mol JNM(1991) Regulation and manipulation of flavonoid gene expres-sion in anthers of Petunia: the molecular basis of the Po-mutation. Plant Cell 3: 39-48

26. Venis MA (1991) Recent progress with auxin and phytotropinreceptors. In M Kutacek, MC Elliot, I Machackova, eds, Mo-lecular Aspects of Hormonal Regulation of Plant Develop-ment. SPB Academic Publishing bv, The Hague, The Nether-lands, pp 177-184

27. Vesper MJ, Kuss CL (1990) Physiological evidence that theprimary site of auxin action in maize coleoptiles is an intracel-lular site. Planta 182: 486-491

28. Vicente 0, Benito Moreno RM, Heberle-Bors E (1991) Pollencultures as a tool to study plant development. Cell Biol Rev25: 295-305

29. Wiermann R (1968) Untersuchungen zum Phenylpropanstoff-wechsel des Pollens. Ber Dtsch Bot Ges 81: 232-238

30. Wiermann R, Vieth K (1983) Outer pollen wall, an importantaccumulation site for flavonoids. Protoplasma 118: 230-233

907