Interplay of Photochemistry and Beer:
How Lightstruck Flavor Is Formed and How It Can Be Prevented
By Denis De Keukeleire, Ghent University, Belgium

Beer goes "skunky" when exposed to light! What an ill-fated scenario: having a beer or two with friends, preferably in a sun-drenched spot, and photochemistry ruining your happy hour by dispersing a disagreeable smell and taste. Yet, beer consumers all over the world are faced to this susceptibility of beer to light. From a scientific viewpoint, it is remarkable how little we know about the phenomenon of the so-called "lightstruck flavor" of beer. This account aims at reporting on the current state-of-the-art,1 thereby highlighting seminal contributions from our laboratory.

Obviously, beer is a popular drink, as ca. 60 ml is consumed daily per world inhabitant, about as much as milk and fivefold more than wine. There are good reasons for this liking. Beer is an alcoholic beverage, but it is very diluted (90%+ water). It is a great socializing aid and it confers a feel of relaxation. Some prefer beer when they are thirsty, others appreciate the diversified palette of so many beer flavors and tastes. There is also growing evidence supporting the nutritional and health benefits of moderate beer consumption as part of a healthy lifestyle.2

However, as the composition of beer is quite complex, chemical changes induced by the presence of oxygen, by temperature and, in particular, by light may affect beer flavor. Since all beers are colored—from very pale-yellow to almost black—a number of absorbers of visible and UV-A light are present and photochemistry must ensue on light exposure. The nature of most photochemical reactions is unknown, but it appears that they do not lead to readily observable organoleptic changes with one notable exception. The problem of a particular off-flavor in beer exposed to light was recognized as early as in 1875,3 and simple tests on the protective power of glass indicated that brown bottles were most effective.4 Gray et al. were first to show that thiols were involved in the development of an offending off-flavor.5 In the early sixties, Kuroiwa et al. used model systems to establish that a photochemical reaction in the wavelength range of 350-500 nm, involving a flavin such as riboflavin, beer bitter agents (isohumulones), and sulfur-containing compounds, led to the so-called "lightstruck flavor".6 Other drinks including champagne, wine, and milk are also sensitive to light, however, none produces the unique "skunky" odor and taste of lightstruck beer.

Role of Hops

The hop plant (Humulus lupulus L.) is an essential ingredient for beer brewing, together with malted barley, water, and yeast. Hops are special to brewers, because they distinguish beer from all other alcoholic drinks. Hop can be considered as a beer herb, since only 1 or 2 grams are needed to brew 1 liter of beer, whereas malted barley is used in much larger quantities, typically 200-300 grams, thus serving as the "body" of beer and providing most of the proteins, polyphenols, dextrins and other substances. In the brewing process, hops are boiled for about 1.5 hours with wort, a sweet tasting solution resulting from enzymic degradation of starch and proteins contained in malted barley. During the boiling in the brewing kettle, hop acids, called humulones (a mixture of 2 isomers and 1 homolog) are isomerized to cis- and trans-isohumulones in a ratio of approximately 7:3 (Scheme 1).

It should be mentioned that the cis-trans notation refers to the tertiary hydroxyl group and the prenyl side chain at vicinal carbon atoms of the five-membered ring in isohumulones. These compounds are extremely bitter-tasting with a threshold value of ca. 5 milligrams per liter (comparable to quinine). In this article, the term "isohumulones" refers to the group of 6 isomers and homologs including cis-trans epimers (the isohumulones with a 3-methylbutanoyl side chain constitute > 80% of the mixture, the isoadhumulones with a 2- methylbutanoyl side chain are very minor constituents). The total concentration of isohumulones in beer varies between 10 and 100 milligrams per liter. Subsequent to the boiling step, the bittersweet hopped wort is transferred to fermenting vessels and yeast converts sugars to ethanol and carbon dioxide within about one week. Young beer needs a maturation period at low temperature during several weeks prior to packaging.

Although less than 100 milligrams of hop-derived compounds remain in 1 liter of beer, they exhibit various important functions. Hop acts as a natural preservative in beer, partly due to its bacteriostatic activity, partly because of its high content in antioxidative flavonoids. Isohumulones are essential to stabilize bubbles in beer foam and they account for most of the bitterness, particularly in pale beers including lager or pilsner-type beers. Unfortunately, isohumulones decompose rapidly either by direct exposure to UV-A light, or by indirect processes involving degradation of a "photosensitizer" by visible light

From Excited Isohumulones to the Lightstruck Flavor


formal mechanism for formation of beer lightstruck flavor in model systems, composed of isohumulones, riboflavin, and cysteine, on exposure to visible light, has been suggested by Kuroiwa et al. already in 1963. 6 Photoexcited riboflavin induces cleavage of isohumulones to a 4-methylpent-3-enoyl radical, which undergoes decarbonylation to a 3-methylbut-2-enyl radical. Trapping of this stabilized allyl radical by a thiol radical derived from cysteine leads to formation of 3-methylbut-2-ene-1-thiol (MBT) (Scheme 2). This mechanism has been frequently referred to in articles dealing with light-induced off flavors in beer and, in fact, all relevant data collected to date do not contradict the Kuroiwa-premise.

In particular, MBT has garnered a status as "skunky thiol". MBT, together with other sulfur-containing constituents, has been identified in malodorous secretions of the anal glands of skunks (Mustela vison L.). Its formation on exposure of beer to light has been confirmed, 7 while a range of techniques have been reported to quantify MBT.8 MBT is one of the most powerful taste- and flavor-active compounds known, while concentrations around 1 nanogram per liter (ca. 9.8 pM) can make (pale) beer unpalatable. Therefore, even very small photochemical conversion rates of isohumulones (0.28-2.8 mM in beer) can produce this effect. Recent work has pointed out that the "skunky flavor" is composed of a rather intricate mixture of sulfur-containing compounds.9,10 Nevertheless, MBT remains largely responsible for the offending flavor and odor (Figure 1).

Although the key components were identified by Kuroiwa, details of the mechanism have not been clarified. Some time ago, we showed that transisohumuolone was photolyzed to dehydrohumulinic acid on irradiation at 300 nm in methanol as indirect proof of the involvement of a Norrish Type 1a cleavage in the formation of the lightstruck flavor. (Scheme 3).11 We examined - in collaboration with Prof. Dr. Malcolm Forbes, Department of Chemistry, University of North Carolina, Chapel Hill, NC, USA – the primary photochemical events using time-resolved (continuous wave) electron paramagnetic resonance (TREPR) spectroscopy and computer simulations of the spectra to identify the free radicals produced.12 The TREPR experiment has superior time response (ca. 60 ns) over steady-state EPR methods, yet retains high structural resolution needed to measure hyperfine interactions used for assignment of the signal carriers. Due to the presence of chemically-induced electron spin polarization (CIDEP) in the TREPR signals, we have also obtained important new insight into the photophysis of the excited states leading to the observed radicals.

Figure 2 shows the X-band (9.5 GHz) TREPR spectrum obtained at a delay time of 0.3 µs after photolysis at 308 nm (excimer laser) of a toluene/methylcyclohexane solution (1:1 ratio) of a mixture of trans-ishumulones (trans-isohumulone, trans-isocohumulone, and trans-isoadhumulone), while a computer simulation using appropriate magnetic parameters is displayed also. The strong TREPR signal proves that free radicals are indeed produced, when UV-light strikes isohumulones. The high intensity in the center of the spectrum and weak signals at the edges indicate the presence

of at least two different radical species with significant electronnuclear hyperfine interactions. We have strong arguments to conclude that the five-membered ring radical, derived from trans-isohumulones, accounts for the intense signal (18 closely spaced lines), while the 3-methylbut-2-enyl radical has its intensity distributed over a very large number of hyperfines (128) that cover more than 100 Gauss. The net emission is generated via the triplet mechanism of CIDEP.

The TREPR data allowed us to propose a detailed mechanism for formation of the lightstruck flavor (Scheme 4). UV-light is absorbed by the b-tricarbonyl chromophore of the isohumulones and, after intersystem crossing, triplet energy transfer occurs on the submicrosecond time scale to the a-hydroxyketone, resulting from very strong throughbond coupling of both chromophores and relatively long electron spin-lattice relaxation times of the triplets involved in the energy transfer process. Subsequent photochemistry furnishes free radicals via a Norrish Type I a-cleavage. The events can be monitored by TREPR, because of strongly emissive triplet-mechanism CIDEP, which is produced initially and then propagated throughout the various photophysical and photochemical pathways. Simulation of the TREPR spectra provides unambiguous assignment of the signal carriers to the radicals proposed in

the mechanism. Future work will include attempts to obtain kinetic studies on the intersystem crossing, spin relaxation, and decarbonylation processes involved in this photochemistry. The primary photophysical and photochemical processes result in formation of a 3-methylbut-2-enyl radical, which, subsequently, is trapped by a thiol radical derived from an as yet unidentified sulfur source to give MBT and, consequently, lightstruck flavor. We have probed this important reaction by a preliminary TREPR study on trapping of radicals, derived from transisohumulones, using n-butanethiol. Very remarkably, we observed a time-dependent alteration of the reference TREPR pattern, which clearly indicates interaction in the time domain below 1 µs. Further studies are necessary to gain more insight into the role of the sulfur source with regard to formation of the lightstruck flavor.

Product analysis of the reaction mixture obtained after irradiation of trans-isohumulone at 254 nm in methanol confirms the proposed mechanism, as typical photoreaction products are derived from recombination of the 3-methylbut-2-enyl radical with the incipient five membered ring radical following Norrish Type I a-cleavage (Scheme 4). The major photodecomposition product of trans-isohumulone is decarbonylated dehydrated isohumulone (Figure 3, peak 9)


and the immediate precursors, decarbonylated cis-isohumulone and decarbonylated trans-isohumulone (peaks 5 and 7), are present as well. Other peaks are due to interfering photochemical reactions including retro-oxa-di-p-methane rearrangement to humulone13 and photoenolization of the b,g-enone chromophore in the side chain of trans-isohumulone sequenced by further dark reactions. It can readily be understood that, in the presence of a suitable sulfur source, the 3-methylbut-2-enyl radical should be converted to MBT rather than recombine with other isohumulone-derived radicals, thus leading to the lightstruck flavor.

From Excited Riboflavin and Isohumulones to the Lightstruck Flavor

The lightstruck flavor is most likely formed on exposure of beer to visible light, the most damaging wavelengths being in the range of 350-500 nm. Since the isohumulones as key constituents do not absorb in this wavelength region, the intervention of a suitable "photosensitizer" is necessary. Riboflavin (RF) or vitamin B2 is present in beer in concentrations around 1 milligram per liter and other flavins have been detected as well.14 In model systems, as applied by Kuroiwa,6 RF absorbs visible light and interaction with isohumulones leads to degradation products and radicals that eventually give rise to formation of MBT.

We envisaged to exploit TREPR to get insight into the mechanism for formation of the lightstruck flavor using RF or flavin mononucleotide (FMN) as a "photosensitizer" under visible light conditions. Since RF exhibits two broad absorption bands with maxima at 350 nm and 450 nm, the third harmonic of a pulsed YAG-laser (355 nm) was a most suitable excitation source. When RF and isohumulones were irradiated, a broad emissive/absorptive EPR signal was detected with a maximum intensity at a delay time of ca. 750 ns. Similarly, a very strong predominantly emissive signal was observed on excitation of FMN in the presence of isohumulones, together with a superimposed emissive signal, which we were able to attribute to a 3-methylbut-2-enyl radical.

Additional experiments and comparison with EDTA as an efficient electron donor led us to establish a reasonable mechanism for the initial photochemical events leading to the lightstruck flavor (Scheme 5).First of all, it was demonstrated by Hastings et al. that the values of the triplet energies of RF (about 210 kJ/mol) and isohumulones (about 300 kJ/mol) exclude direct energy transfer. 15 Hence, RF is incapable of generating excitedstate isohumulones and, thus, RF-initiated photosensitization is not operative. Triplet-excited RF is a very strong electron acceptor16 and electron transfer from isohumulones affords an oxidized species. The electron may be released from the b-tricarbonyl unit or from either oxygen of the a-hydroxyketone. No details are known yet, but for each of the possible radical cations, stabilization pathways can be envisaged that lead, via a formal Norrish Type I a-cleavage of the a-hydroxyketone, to a 3-methylbut-2-enyl radical en route to the lightstruck flavor. One clue regarding the nature of oxidized species, derived from isohumulones, has been the identification of oxydehydrohumulinic acid (Scheme 5) as a major photoproduct from the visible-light irradiation of trans-isohumulone in the presence of RF. The mechanism for formation of this oxidized derivative of trans-isohumulone remains elusive pending further research.

It should be clearly pointed out that RF does not function as a photosensitizer according to definition, since it is itself reduced in the process. Rather, RF or flavins are the photoreactive entities in model systems–and very likely also in beer–which induce decomposition of isohumulones. It is, furthermore, very remarkable that both direct and “photosensitized” excitation of isohumulones channel the breakdown to the pivotal 3-methylbut-2-enyl radical, although accesses are entirely different.

Prevention of Formation of the Lightstruck Flavor

Apart from storing beer in light–proof containers, such as dark glass, kegs or cans, or immediate consumption, the photosensitivity can be circumvented by quenching of the excited triplet state of the isohumulones and/or RF (not readily achievable) or by the use of chemically modified isohumulones, whereby formation of the lightstruck flavor is prohibited. This is currently achieved on an industrial scale in several hop processing plants around the world reflecting the state-of-the-art of hop technology aimed at controlling not only the light-stability of beer, but also bitterness levels and desired foam features. About half of the hops produced worldwide is extracted with liquid or supercritical carbon dioxide (hops ranks first in the application of this technology thereby preceding decaffeination of coffee).17,18 After removal of the hop essential oil, humulones are isomerized to isohumulones according to a very efficient procedure (more than 90% yield) developed some time ago in our laboratory,19 whereas conversion of humulones in the brewery is very low (less than 30%) when hop cones are used. The isohumulones, commercially available as standardized aqueous solutions, can be used to dose exactly the bitterness of beer. In a further manipulation, isohumulones are (quantitatively) reduced to


dihydroisohumulones and tetrahydroisohumulones (Scheme 6; both are mixtures of isomers, homologs, and stereomers) by sodium borohydride reduction and catalytic hydrogenation, respectively. These so-called advanced hop products are now widely used in the brewing practice. Introduction of this advanced hop technology during the last decade has led to a much improved overall hop utilization. Dihydroisohumulones, in which the photosensitive a-hydroxyketone is reduced to a photoinactive 1,2-diol entity, are fully light-resistant, since a Norrish Type I a-cleavage can not occur. Consequently, beers, bittered with dihydroisohumulones, can be kept in clear glass bottles without any harm being done by light. In contrast, tetrahydroisohumulones are as photoreactive as isohumulones, however, MBT cannot be formed subsequent to Norrish Type I a-cleavage, since the double bond in the photosensitive side chain is lacking. In principle, it may be feasible that some sort of a lightstruck flavor is developed from tetrahydroisohumulones, but until now a distinct flavor change in beers, bittered with tetrahydroisohumulones, has not been observed. In this respect, many consider tetrahydroisohumulones to be light-proof, which is obviously not correct. It may be added that tetrahydroisohumulones have gained fame, because they are twice as bitter as isohumulones and they have a strong impact on the stability of the head on a glass of beer and on the foam cling to the glass.

The light-associated characteristics of dihydroisohumulones and tetrahydroisohumulones have been nicely confirmed by our TREPR experiments. Photolysis of trans-tetrahydroisohumulones (trans-tetrahydroisohumulone, transtetrahydroisocohumulone, and trans-tetrahydroisoadhumulone) at 308 nm yielded the TREPR signal, displayed


in Figure 4, which, again, has a simulation shown immediately below it. The shape is altered significantly with respect to that of trans-isohumulones (see Figure 2). It is noteworthy that the signal of the trans-tetrahydroisohumulones has narrowed considerably, which is to be expected if, after Norrish Type I a-cleavage, decarbonylation to a stabilized allyl radical is no longer taking place on the µs-time scale and, consequently, the hyperfine structure due to a 3-methylbut-2-enyl radical disappears. There are two additional transitions seen on the perimeter of the spectrum, one emissive line on the low-field side and a weaker absorptive line on the high-field side. These transitions, marked with an asterisk (*), are due to an unknown radical, which could not be simulated. The signal may arise from a competing photochemical reaction or from secondary photochemical processes. The fact that this signal is more strongly spin-polarized by the radical pair mechanism of CIDEP indicates that it does not result from a primary process. It is more likely to have been generated at a later delay time, after the strong net emissive polarization from the parent ketone triplet has relaxed.

Photolysis of a solution of dihydroisohumulones at 308 nm does not lead to observable TREPR signals. This is the first direct spectroscopic evidence for the resistance of dihydroisohumulones to photolysis and it is the most conclusive proof to date that the photochemistry leading to the lightstruck flavor in beer requires activation of the a-hydroxyketone moiety.

To provide additional support for the mechanism, proposed in Scheme 4, the triplet spectrum of frozen dihydroisohumulones in methylcyclohexane solution was generated and observed by TREPR (Figure 5A). This experiment should characterize the b–tricarbonyl triplet in the absence of energy transfer, as observed for isohumulones. The emissive half– field transitions, indicated by an asterisk, strongly suggest that the net emissive polarization, shown in Figure 2, originates from the a–tricarbonyl chromophore. A rough estimate of the dipolar interaction D in the triplet can be made by measurement of the separation (in Gauss) between the outermost dm = 1 transitions, indicated in dashed vertical lines. This gives D = 1100 ± 100 Gauss or approximately 0.1 cm–1, which is very consistent with a delocalized triplet such as this b-tricarbonyl unit.

We further irradiated dihydroisohumulones in the presence of 2,2,6,6–tetramethylpiperidine–1–oxyl (TEMPO), a stable free radical. In the TREPR spectrum an emissively polarized three–line spectrum of TEMPO is observed (Figure 5B). No TREPR spectrum is obtained with either dihydroisohumulones alone or TEMPO alone. The emissive polarization of the TEMPO free radical is due to the radical triplet–pair mechanism of CIDEP,20 which results from the diffusive encounter and magnetic interaction between an excited triplet molecule, in this case that of the dihydroisohumulones, and a doublet-state free radical. The net emission must arise from the polarized triplet state of the dihydroisohumulones, as TEMPO does not absorb the light, nor does it have any spin polarization it can acquire on its own. Both spectra in Figure 5 allow us to conclude that the emissive polarization, observed in the TREPR spectra in Figure 2, is originally generated in the intersystem crossing process from S1 to T1 of the b–tricarbonyl moiety, the primary light–absorbing chromophore in the isohumulones.

Conclusions

Photochemistry undoubtedly deteriorates the quality of beer and protection against light is absolutely necessary. Isohumulones, the main beer bitter compounds derived from hops, undergo light-induced decomposition either on direct illumination with UV-A light, or via a photoredox reaction involving excitation of a visible-light absorber such as riboflavin. Both routes lead to formation of a 3-methylbut-2-enyl radical, which is trapped by a thiol radical originating from a suitable sulfur source thereby affording 3-methylbut-2-ene-1-thiol, known as “skunky thiol”. This lightstruck flavor, observable at a concentration of the thiol of around 1 nanogram per liter, is the most offending off flavor known to occur in beer. Our studies of this phenomenon by means of time-resolved electron paramagnetic resonance, applied to the photochemistry of individual beer bitter compounds and model systems, in conjunction with identification of photoreaction products, have permitted to establish reliable reaction mechanisms for formation of the lightstruck flavor. The results obtained so far should form a fundamental basis for further complicated studies on beer itself.

Physical prevention against the lightstruck flavor calls on the use of dark glass bottles (preferably brown-colored) or any opaque package, but modern hop technology enables application of high-tech and efficient chemical protection. Dihydroisohumulones are perfectly light-stable and beers bittered with these advanced hop products can safely be packaged in clear glass bottles. Tetrahydroisohumulones are sensitive to light, but, since 3-methylbut-2-ene-1-thiol can not be formed, a possible flavor change does not compare with the lightstruck flavor. In the brewing practice, a hopping procedure using a combination of dihydroisohumulones (between 15 and 20 milligrams per liter) and tetrahydroisohumulones (few milligrams per liter) is a very attractive option as an alternative for hop cones. Indeed, dihydroisohumulones account for light-stability, while tetrahydroisohumulones accentuate bitterness and greatly improve the stability and cling of beer foam.

Acknowledgments

I would like to thank my co-workers for their contributions in the field of hop and beer research. Financial support by the Interbrew-Baillet Latour Foundation (Leuven, Belgium), Interbrew/Cobrew (Leuven, Belgium), and the Labatt Brewing Company (London, Ontario, Canada) is gratefully acknowledged.

References

  1. The present article updates a previous paper by the same author on Photochemistry of Beer, which was published in The Spectrum 1991, 4(2), 1-7.
  2. Denke, M. A. Am. J. Med. Sci. 2000, 320, 320-326.
  3. Lintner, C. Lehrbuch der Bierbrauerei; Verlag Vieweg und Sohn: Braunschweig, 1875.
  4. Brand, J. Zeitschr. ges. Brauwesen 1908, 31, 333.
  5. Gray, P.; Stone, I.; Rothschild, H. Wallerstein Lab. Commun. 1947, 10, 5.
  6. Kuroiwa, Y.; Hashimoto, N.; Hashimoto, H.; Kokubo, E.; Nakagawa, K. Proc. Am. Soc. Brew. Chem. 1963, 181-193.
  7. Gunst, F.; Verzele, M.; J. Inst. Brew. 1978, 84, 291.
  8. Templar, J.; Arrigan, K.; Simpson, W. J. Brewers Digest 1995 (May), 18-25.
  9. Hill, P. G.; Smith, R. M. J. Chromatogr. A. 2000, 872, 203-213.
  10. Hughes, P. Cerevisia 1999, 24(2), 21-25.
  11. Blondeel, G. M. A.; De Keukeleire, D.; Verzele, M. J. Chem. Soc. Perkin trans. 1 1987, 2715-2717.
  12. Burns, C. S.; Heyerick, A.; De Keukeleire, D.; Forbes, M. D. Chem.-Eur. J. 2001, in press.
  13. Blondeel, G. M. A.; De Keukeleire, D. Tetrahedron Lett. 1979, 1343-1346.
  14. Moll, M.; Louis, C. J. Inst. Brew. 1977, 83, 132.
  15. Hastings, D. J.; McGarrity, M. J.; Bordeleau, L.; Thompson, D. J. In Abstracts of the Brewing Congress of the Americas, 20-24 September 1992, St. Louis, MO, p 97.
  16. Lu, C. Y.; Wang, F. W.; Lin, W . Z.; Han, Z. H.; Yao, S. D.; Lin, N. Y. J. Photochem. Photobiol. B: Biol. 1999, 52, 111-116.
  17. Benitez, J. L.; Forster, A.; De Keukeleire, D.; Moir, M.; Sharpe, F. R.; Verhagen, L. C.; Westwood, K. T. Hops and Hop Products; Verlag Hans Carl: Nürnberg, 1997.
  18. Spent hops, after extraction by liquid or supercritical carbon dioxide, contains most of the hop flavonoids that are associated to medicinal properties. We have recently isolated and identified 8-prenylnaringenin as a very potent phytoestrogen present in hops. 8-Prenylnaringenin is more active than any established phytoestrogen including genistein from soy and coumestrol from clover. See: Milligan, S. R.; Kalita, J.; Heyerick, A.; Rong, H.; De Cooman, L.; De Keukeleire, D. J. Clin. Endocrinol. Metab. 1999, 83, 2249-2252, and Milligan, S. R.; Kalita, J. C.; Pocock, V.; Van de Kauter, V.; Rong, H.; De Keukeleire, D.; Stevens, J. F.; Deinzer, M. J. Clin. Endocrinol. Metab. 2000, 85, 4916-4920. The hop phytoestrogen, 8-prenylnaringenin is present in most beers. See: Rong; H.; Zhao, Y., Lazou, K.; De Keukeleir e, D.; Milligan, S. R.; Sandra, P . Chromatographia 2000, 51, 545-552.
  19. Verzele, M.; De Keukeleire, D. Chemistry and Analysis of Hop and Beer Bitter Acids; Elsevier: Amsterdam, 1991.
  20. Blättler, C.; Jent, F.; Paul, H. Chem. Phys. Lett. 1990, 166, 375.

About the Author

Dr. Denis De Keukeleire received his Ph.D. in chemistry at the Ghent University, Belgium, in 1971. He worked as a NATO-postdoctoral fellow on the photochemistry of trichromophores with George Hammond at the California Institute of Technology, Pasadena, California, during 1971-1972. From 1972 until 1991 he was a senior research associate of the Belgian Fund for Scientific Research with a position as photochemist and phytochemist at the Laboratory of Organic Chemistry, Ghent University. In 1991, he joined the Laboratory of Plant Biochemistry and became full professor in 1992 at the Faculty of Pharmaceutical Sciences, Ghent University. His research interests are in photochemical synthesis and natural products, in particular hops. The expertise on beer bitter compounds derived from hops and on hop aroma is currently being complemented by focusing on health-beneficial and medicinal properties of hops, mainly on phytoestrogenicity. His address is Ghent University, Faculty of Pharmaceutical Sciences, Laboratory of Pharmacognosy and Phytochemistry, Harelbekestraat 72, B-9000 Ghent, Belgium (+3292648055); e-mail: Denis.DeKeukeleire@rug.ac.be; http://allserv.rug.ac.be/~ddkeukel.