Fate of xanthohumol and related prenylflavonoids from hops to beer

Article abstract of DOI:10.1021/jf990101k

The fate of three prenylated flavonoids of the chalcone type, xanthohumol, desmethylxanthohumol, and 3'-geranylchalconaringenin, was monitored with LC/MS-MS from hops (Humulus lupulus L.) to beer in two brewing trials. The three prenylchalcones were largely converted into their isomeric flavanones, isoxanthohumol, prenylnaringenins, and geranylnaringenins, respectively, in the boiling wort. Losses of prenylflavonoids were due to incomplete extraction from the hops into the wort (13-25%), adsorption to insoluble malt proteins (18-26%), and adsorption to yeast cells (11-32%) during fermentation. The overall yield of xanthohumol, after lagering of the beer and largely in the form of isoxanthohumol, amounted to 22-30% of the hops' xanthohumol. About 10% of the hops' desmethylxanthohumol, completely converted into prenylnaringenins, remained in the beers. 3'- Geranylchalconaringenin behaved similarly to desmethylxanthohumol. Solubility experiments indicated that (1) malt carbohydrates form soluble complexes with xanthohumol and isoxanthohumol and (2) solubility does not dictate the isoxanthohumol levels of finished beers.

Full text of DOI:10.1021/jf990101k

J. Agric. Food Chem. 1999, 47, 2421−2428
2421
F a te of Xa n th oh u m ol a n d Rela ted P r en ylfla von oid s fr om Hop s to
Beer
J an F. Stevens,*,† Alan W. Taylor,† J eff E. Clawson,‡ and Max L. Deinzer†
Departments of Chemistry and Food Science and Technology, Oregon State University,
Corvallis, Oregon 97331
The fate of three prenylated flavonoids of the chalcone type, xanthohumol, desmethylxanthohumol,
and 3′-geranylchalconaringenin, was monitored with LC/MS-MS from hops (Humulus lupulus L.)
to beer in two brewing trials. The three prenylchalcones were largely converted into their isomeric
flavanones, isoxanthohumol, prenylnaringenins, and geranylnaringenins, respectively, in the boiling
wort. Losses of prenylflavonoids were due to incomplete extraction from the hops into the wort
(13-25%), adsorption to insoluble malt proteins (18-26%), and adsorption to yeast cells (11-32%)
during fermentation. The overall yield of xanthohumol, after lagering of the beer and largely in the
form of isoxanthohumol, amounted to 22-30% of the hops’ xanthohumol. About 10% of the hops’
desmethylxanthohumol, completely converted into prenylnaringenins, remained in the beers. 3′-
Geranylchalconaringenin behaved similarly to desmethylxanthohumol. Solubility experiments
indicated that (1) malt carbohydrates form soluble complexes with xanthohumol and isoxanthohumol
and (2) solubility does not dictate the isoxanthohumol levels of finished beers.
Keyw or d s: Beer; hops; Humulus lupulus; prenylated flavonoids; xanthohumol
INTRODUCTION
investigators have studied the removal of these pheno-
lics from beer by adsorption to nylon and polyvinylpoly-
pyrrolidone [see, e.g., McMurrough et al. (1995)]. The
prenylflavonoids of hops have not received nearly as
much attention for two main reasons. First, the value
of these polyphenols to the brewer is not well under-
stood. However, recent studies revealing various biologi-
cal effects of hop prenylflavonoids in vitro (Miranda et
al., 1999; Tabata et al., 1997; Tobe et al., 1997) have
fueled a renewed interest in these phenolics as dietary
flavonoids. Second, the levels of prenylflavonoids in
typical American beers (0.6 ppm; Stevens et al., 1999)
are low in comparison with the levels of iso-R-acids (13-
20 ppm; Peacock, 1998) and proanthocyanidins [1-40
ppm, depending on the method used; see, e.g., McGuin-
ness et al. (1975) and Delcour (1988)]. This makes HPLC
with UV detection less suitable for analysis of beer
prenylflavonoids without sample pretreatment. We have
recently developed a sensitive and selective method for
direct quantitation of prenylflavonoids in hops and beer
by HPLC/tandem mass spectrometry (MS) (Stevens et
al., 1999). In the present study, we used this method in
two brewing trials to investigate the fate of xanthohu-
mol and related prenylflavonoids from hops to beer.
Hops, the female inflorescences of the hop plant
(Humulus lupulus L.), are used in the brewing industry
to add flavor and bitterness to beer. The bitter principles
of beer, the iso-R-acids, are formed from the hops’
R-acids by thermal isomerization in the brew kettle.
Because the bitter acids occur together with prenylated
flavonoids of the chalcone type in the resin fraction of
hops (Ha¨nsel and Schulz, 1988; Stevens et al., 1997),
conversion of prenylchalcones into their isomeric fla-
vanones (Bohm, 1989) is expected to take place in the
brewing process. In fact, beer contains higher levels of
prenylflavanones than prenylchalcones (Stevens et al.,
1999). Beer prenylflavanones are largely derived from
the hop chalcones because hops are very poor in pre-
nylflavanones, and other beer ingredients such as
barley, wheat, and rice lack prenylated flavonoids (see
Figure 1 for structures and nomenclature).
Earlier studies on the fate of phenolic compounds
during the brewing process have been summarized by
Charalambous (1981). Most of these studies, however,
were focused on total phenolics or on proanthocyanidins
as a group of polyphenols. More recently, Moll et al.
(1984) investigated changes in the levels of individual
catechins from barley to finished beer by high-perfor-
mance liquid chromatography (HPLC) with ultraviolet
(UV) detection. They attributed the loss of procyanidin
dimer B3 (∼75%) to incomplete extraction from barley
and to chemical degradation (depolymerization) during
the later brew stages. Because proanthocyanidins have
been associated with the formation of beer haze, many
EXPERIMENTAL PROCEDURES
Ch em ica ls. Xanthohumol, isoxanthohumol, desmethyl-
xanthohumol, and 3′-geranylchalconaringenin were isolated
from hops as before (Stevens et al., 1997). Prenylnaringenins
were prepared by prenylation of naringenin (Indofine, Som-
erville, NJ ) according to the method of J ain et al. (1978). A
similar procedure was used for the preparation of 6-geranyl-
naringenin and is described below. The internal standard, 2′,4-
dihydroxychalcone, was purchased from Indofine.
P r ep a r a tion of Ger a n yln a r in gen in s. 6-Geranylnarin-
genin was prepared by treatment of naringenin with (()-
linalool in the presence of BF3-etherate as follows. Boron
trifluoride-etherate (0.54 mL) was added dropwise to a stirred
* Author to whom correspondence should be addressed
[telephone +1 (541) 737 1776; fax +1 (541) 737 0497; e-mail
stevensf@ucs.orst.edu].
† Department of Chemistry.
‡ Department of Food Science and Technology.
10.1021/jf990101k CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/30/1999

2422 J. Agric. Food Chem., Vol. 47, No. 6, 1999
Stevens et al.
F igu r e 1. Major prenylated flavonoids of hops (left) and their conversion products formed during the brewing process (right).
solution of naringenin (0.55 g) in dioxane (10 mL) under
nitrogen. Linalool (0.8 mL) was then added dropwise to the
reaction mixture. The solution was stirred for 2.5 h at room
temperature and then diluted with ethyl ether (150 mL). The
ether layer was washed with water (3 × 100 mL), dried over
anhydrous Na2SO4, and evaporated to dryness on a rotary
evaporator. The residue was taken up in MeOH and submitted
to preparative HPLC. The reaction products were separated
on a 10 µm Econosil RP-18 column (22 × 250 mm) using a
linear gradient starting from 40 to 100% acetonitrile in 1%
aqueous formic acid over 45 min at a flow rate of 11.2 mL/
min. The UV trace was recorded at 290 nm. Peak fractions
were collected manually and dried by rotary evaporation and
lyophilization. 6-Geranylnaringenin was obtained as a white
powder (15 mg): UV (MeOH) λmax (log ꢀ) 294 (4.24) and 336
(3.55) nm; APCI-MS, m/z (rel intensity) 409 [MH]+ (100); MS-
F igu r e 2. LC/MS chromatogram of a methanolic extract of
the cold trub, recovered from the single-hopped wort. For key
to peaks, compare Figure 1. Peaks 2 and 3 are included to
indicate their relative positions in the chromatogram.
+
MS (collision energy, 11 V) m/z 409 (71), 285 [MH - C9H16
]
(100), 165 [1,3AH-C9H16
]
+ (25); 1H NMR (DMSO-d6, 600 MHz)
δ 12.43 (s, OH-5), 10.72 (s, OH-7), 9.57 (s, OH-4′), 7.30 (2H, d,
J ) 8.5 Hz, H-2′/H-6′), 6.78 (2H, d, J ) 8.5 Hz, H-3′/H-5′), 5.96
(1H, s, H-8), 5.40 (1H, dd, J ) 2.8 and 12.8 Hz, H-2), 5.12 (1H,
t, J ) 7.1 Hz, H-2′′), 5.04 (1H, t, J ) 7.1 Hz, H-7′′), 3.23 (1H,
dd, J ) 12.8 and 17.0 Hz, H-3ax), 3.12 (2H, d, J ) 7.0 Hz, H-1′′),
2.67 (1H, dd, J ) 2.9 and 17.0 Hz, H-3eq), 1.99 (2H, t, J ) 7.4
Hz, H-6′′), 1.90 (2H, t, J ) 7.6 Hz, H-5′′), 1.69 (3H, s, H-4′′),
1.60 (3H, s, H-10′′), 1.53 (3H, s, H-9′′); 13C NMR (DMSO-d6,
150 MHz) δ 196.5 (C-4), 164.2 (C-7), 160.6 (C-5), 160.5 (C-9),
157.7 (C-4′), 133.8 (C-3′′), 130.6 (C-8′′), 129.0 (C-1′), 128.3 (C-
2′/C-6′), 124.1 (C-7′′), 122.4 (C-2′′), 115.1 (C-3′/C-5′), 107.5 (C-
6), 101.6 (C-10), 94.3 (C-8), 78.3 (C-2), 42.1 (C-3), 39.0 (C-5′′),
26.2 (C-6′′), 25.5 (C-10′′), 20.6 (C-1′′), 17.5 (C-9′′), 15.9 (C-4′′).
8-Geranylnaringenin was prepared by isomerization of
6-geranylnaringenin. The latter compound (12 mg) was dis-
solved in 20 mL of 5% NaOH in MeOH. The mixture was
refluxed under nitrogen for 1 h and then poured into 200 mL
of 2 N HCl. The aqueous mixture was extracted with EtOAc
(1 × 150 mL, 2 × 75 mL). The combined EtOAc layers were
washed with water (3 × 100 mL), dried over Na2SO4, and
evaporated. The major product, 8-geranylnaringenin, was
isolated from the residue by preparative HPLC along with 3′-
geranylchalconaringenin and 6-geranylnaringenin. 8-Geranyl-
naringenin was obtained as a white powder (3 mg): UV
(MeOH) λmax (log ꢀ) 294 (4.23) and 338 (3.58) nm; APCI-MS,
H-6′′), 1.87 (2H, t, J ) 7.6 Hz, H-5′′), 1.59 (3H, s, H-10′′), 1.53
(3H, s, H-4′′), 1.52 (3H, s, H-9′′); 13C NMR (DMSO-d6, 150 MHz)
δ 196.7 (C-4), 164.3 (C-7), 161.1 (C-5), 159.7 (C-9), 157.5 (C-
4′), 133.7 (C-3′′), 130.6 (C-8′′), 129.2 (C-1′), 128.0 (C-2′/C-6′),
124.1 (C-7′′), 122.4 (C-2′′), 115.1 (C-3′/C-5′), 106.9 (C-8), 101.8
(C-10), 95.2 (C-6), 78.2 (C-2), 41.9 (C-3), 38.2 (C-5′′), 26.1 (C-
6′′), 25.5 (C-10′′), 21.1 (C-1′′), 17.5 (C-9′′), 15.8 (C-4′′).
In str u m en ta tion a n d An a lysis. Prenylflavonoids were
analyzed in hops, wort, and beer by LC/MS as described
previously (Stevens et al., 1999). In short, hop extracts and
brew products were separated on a 5 µm C18 column (250 × 4
mm) using a linear solvent gradient from 40 to 100% B
(acetonitrile) in solvent A (1% aqueous formic acid) over 15
min, followed by 100% solvent B for 5 min. The flow rate was
0.8 mL/min. The HPLC instrument was connected to a PE
Sciex API III Plus triple-quadrupole mass spectrometer
equipped with an atmospheric pressure chemical ionization
(APCI) source. The APCI source was operated in the positive
ion mode. The HPLC effluent was introduced into the mass
spectrometer via a heated nebulizer interface at 500 °C.
Ionization of the analyte vapor mixture was initiated by a
corona discharge needle at ∼8 kV and a discharge current of
∼6 µA. Six prenylflavonoids were selectively detected by
multiple reaction monitoring (MRM) in the tandem MS mode.
Argon-nitrogen (9:1) was used as target gas (∼1.8 × 1014
atoms/cm2). The collision energy was 30 V. For MRM, specific
daughter ions produced by collision-induced dissociation of
[MH]+ ions of selected prenylflavonoids were detected (for a
typical chromatogram, see Figure 2). Quantitation was carried
out using 4,2′-dihydroxychalcone as the internal standard.
Calibration curves were recorded for each of the six prenylfla-
m/z (rel intensity) 409 [MH]+ (100); MS-MS (collision energy,
+
11 V) m/z 409 (100), 285 [MH - C9H16
]
(58), 165 [1,3AH-
+
1
C9H16
] (39); H NMR (DMSO-d6, 600 MHz) δ 12.10 (s, OH-
5), 10.74 (s, OH-7), 9.55 (s, OH-4′), 7.30 (2H, d, J ) 8.5 Hz,
H-2′/H-6′), 6.78 (2H, d, J ) 8.5 Hz, H-3′/H-5′), 5.97 (1H, s, H-6),
5.41 (1H, dd, J ) 2.9 and 12.5 Hz, H-2), 5.09 (1H, t, J ) 7.1
Hz, H-2′′), 5.03 (1H, t, J ) 7.1 Hz, H-7′′), 3.19 (1H, dd, J )
12.6 and 17.1 Hz, H-3ax), 3.08 (2H, d, J ) 7.1 Hz, H-1′′), 2.72
(1H, dd, J ) 3.1 and 17.1 Hz, H-3eq), 1.97 (2H, t, J ) 7.4 Hz,

Fate of Xanthohumol from Hops to Beer
J. Agric. Food Chem., Vol. 47, No. 6, 1999 2423
vonoids on the days the samples were analyzed; samples and
standards were run at least twice.
(7) Spent Hops. After whirlpool and transfer of the wort to
the fermenter, solids remaining in the kettle were allowed to
settle for ∼5 min, and the spent hops were quantitatively
recovered using a coarse sieve. The weight of the wet spent
hops was determined. A weighed portion was lyophilized. The
lyophilized material was weighed, ground, and further treated
as hops.
(8) Hot Trub. Fine solids were allowed to resettle after
recovery of the spent hops. The deposit, consisting mainly of
coagulated protein (trub) and finer hop particles, was drained
as a slurry from the bottom of the brew kettle. The slurry was
weighed, and an aliquot was filtered through Whatman No. 1
paper. The weight of the hot trub was determined after
lyophilization. A 0.400 g sample of the lyophilized material
was extracted with 100 mL of MeOH by sonication for 10 min.
The extract was centrifuged (1300g, 15 min), and 25.0 mL of
the supernatant was spiked with 50 µL of internal standard
solution and made up to 100 mL with MeOH/HCOOH (99:1,
v/v).
(9) Preyeast Pitch Wort. After the cold trub was removed
from the fermenter, a sample of the supernatant was taken
prior to inoculation with the yeast and further treated as
unhopped wort.
(10) Cold Trub. During the overnight chilling of the wort in
the fermenter, another portion of coagulated protein and other
solids settled in the coned-shaped bottom of the fermenter,
which was drained as a slurry and further treated the same
as hot trub.
(11) End of Primary Fermentation. A sample was drawn
from the fermenter and treated the same as the unhopped
wort.
(12) Seven Days of Secondary Fermentation. A sample was
drawn after 7 days of secondary fermentation and treated as
the unhopped wort.
(13) Yeast Slug. The yeast settled out during the secondary
fermentation and was drawn off as a slurry from the cone-
shaped bottom of the fermenter. The yeast slurry was further
treated the same as hot trub.
(14) Raw Beer. Raw beer was treated the same as the
unhopped wort.
(15) Beer after Lagering. This was treated the same as the
unhopped wort.
Solu bility of Xa n th oh u m ol a n d Isoxa n th oh u m ol in
Differ en t Med ia . The following media were studied: water,
5% ethanol in water (v/v), beer (a major U.S. brand containing
no detectable amounts of prenylflavonoids), low-gravity wort
(8.5 °P = 8.4 g of extract solids/100 g of wort), and high-gravity
wort (13 °P = 12.9% w/w). Before the start of the experiments,
the worts were refrigerated and then centrifuged (1300g, 15
min) to remove any cloudiness. To 1.0 mg portions of xantho-
humol and isoxanthohumol was added 25.0 mL of the above
media. The mixtures were shaken from time to time and kept
at 80 °C for 30 min, resulting in complete dissolving of the
solids. The mixtures were cooled and kept at room temperature
for at least 4 h and then centrifuged (1300g, 15 min). Aliquots
of the supernatants (100 µL) and 50 µL of the internal
standard solution (0.10 mg/mL DMSO) were diluted to 10.0
mL with MeOH/H2O/HCOOH (50:49:1 by vol). The centrifuged
mixtures were then resuspended by shaking and kept at 8 °C
for at least 4 h. After centrifugation (1300g, 15 min), 100 µL
aliquots were diluted as described for the room temperature
supernatants. Additional 1.0 mg portions of isoxanthohumol
were mixed with 5.0 mL of the beer and worts and further
treated and analyzed as described above.
Isom er iza tion of Xa n th oh u m ol in a Mod el System . The
reaction medium was prepared by dissolving 10 g of sucrose
in a mixture of 10 mL of ammonium acetate buffer (100 mM,
adjusted with acetic acid to pH 5.5) and 50 mL of water, finally
diluted with water to 100 mL. This solution was brought to a
boil, and xanthohumol (200 µg, dissolved in 100 µL of DMSO)
was added. The reaction mixture was left boiling (101 °C)
under reflux conditions, and 4 mL samples were drawn from
the reaction mixture at 5, 15, 30, and 60 min after addition of
Nuclear magnetic resonance spectra were recorded on a
multinuclear Bruker DRX 600 instrument at 600 MHz for 1H
and at 150.9 MHz for 13C. Samples were dissolved in DMSO-
d6 and analyzed at room temperature. DMSO resonances (δH
2.50 and δC 39.51) were used as internal chemical shift
references. 1H-13C HMQC and HMBC experiments were
performed using standard pulse sequences. One-dimensional
NOE experiments were recorded using the double-pulsed field
gradient spin-echo (DPFGSE) sequence as described by Stott
et al. (1995).
Ultraviolet spectra were recorded in MeOH on a Cecil 3021
spectrophotometer.
Beer P r od u ction . Two beers were produced in 265 L
batches according to normal brewing procedures. Both beers
were of the lager type, and they were brewed using 68% two-
rowed malted barley (Great Western Malting) and 32% rice
syrup solids. Before boiling, the wort volume was 303 L (pH
5.6). The volume after wort boiling was 265 L. Lactic acid was
added to the wort during the boil, resulting in pH 5.2 at the
end of the boil.
The hopping rate for each beer was intended to result in
the same number of bittering units (BU), that is, 12 BU. For
the first brew, whole hops (172 g, cv. Willamette harvested in
1997) were added in one portion 15 min after the wort was
brought to boiling. Boiling was continued for 75 min. For the
second brew, whole hops from the same bale were added in
two portions. The first portion (128 g, accounting for 75% of
the BU) was added 15 min after the wort was brought to
boiling; the second portion (137 g, accounting for 25% of the
BU) was added 5 min before the end of the 90 min-boiling
period.
After wort boiling, a whirlpool was formed in the kettle
using a paddle, and the solids were allowed to settle for 15
min. The wort was then pumped to the fermentation tank via
a heat exchanger, which cooled the wort to 13 °C. During the
transfer, the wort [specific gravity ) 8.5 °Plato (°P)] was
oxygenated with sterile oxygen. The finished wort was cooled
to 4 °C and held overnight to allow further settling. The wort
was then warmed to 13 °C and inoculated with a lager yeast
(Wyeast Laboratories, strain 2035). Primary fermentation was
carried out at 13 °C for 7 days. The temperature was raised
to 15.5 °C for 48 h (diacetyl rest) and then dropped to 3 °C for
secondary fermentation, which lasted 15 days. The raw beer
was transferred into 19 L kegs and stored at 2 °C for 7 days;
it was then aged at 1 °C for another 2-3 weeks (lagering
period).
Br ew Sa m p les a n d Sa m p le P r ep a r a tion . Samples were
collected at various stages during the brewing process and
treated as follows prior to LC/MS analysis:
(1) Hops. A 5 g sample of hop cones was ground with a Wiley
mill to pass a 20 mesh sieve. An aliquot (0.400 g) was extracted
in 100.0 mL of MeOH by sonication for 10 min. A portion of
the extract was centrifuged (1300g, 15 min). An aliquot of the
supernatant (2.5 mL) and 50 µL of internal standard solution
(2′,4-dihydroxychalcone, 1 mg/mL DMSO) were placed in a 100
mL volumetric flask, and the contents were made up to volume
with MeOH/HCOOH (99:1 v/v). The rest of the ground hops
was lyophilized to determine the moisture content (9.3 and
9.7% for hop cones used in the first and second brews,
respectively).
(2) Unhopped Wort. A 0.5 L sample of the wort was taken
before the hops were added. A portion of the liquid was
centrifuged (1300g, 15 min), and 100.0 mL of supernatant was
spiked with 50 µL of internal standard solution.
(3) Wort Sampled 15 min after Addition of Hops. This was
treated in the same manner as the unhopped wort.
(4) Wort Sampled 30 min after Addition of Hops. This was
treated the same as the unhopped wort.
(5) Wort Sampled 70 min after Addition of Hops. This was
treated the same as the unhopped wort.
(6) Wort Sampled after Whirlpool. This was treated the same
as the unhopped wort.

2424 J. Agric. Food Chem., Vol. 47, No. 6, 1999
Stevens et al.
F igu r e 3. Prenylflavonoids measured in single-hopped brew. Flavonoids: ISO, isoxanthohumol; XAN, xanthohumol; 6-PN,
6-prenylnaringenin; 8-PN, 8-prenylnaringenin; DMX, desmethylxanthohumol. Brewing stages: for more details, see Brew Samples
and Sample Preparation under Experimental Procedures.
xanthohumol. After the samples had cooled to room temper-
ature, 2.5 mL aliquots were spiked with 5 µL of internal
standard solution (2′,4-dihydroxychalcone, 1.0 mg/mL DMSO)
and made up to 10 mL with 10 mM ammonium acetate buffer.
The time ) 0 min sample (pH 5.4) was prepared by adding
xanthohumol (200 µg, dissolved in 100 µL of DMSO) to 100
mL of reaction medium at room temperature; a 2.5 mL aliquot
was diluted to 10 mL with 10 mM ammonium actetate buffer
after the addition of 5 µL of internal standard solution. These
solutions were analyzed by LC/MS as described above. At the
end of the experiment and after cooling to room temperature,
the reaction mixture was at pH 5.4.
waste materials were determined and expressed as
percent of the prenylflavonoid amounts in the hops
(Figure 3). Because 13% of the hops’ xanthohumol was
found in the spent hops, at least 87% must have been
extracted. However, the amount of xanthohumol mea-
sured in the post-whirlpool wort was equal to 69%. The
difference (18%) could largely be accounted for by
adsorption to the coagulated proteins (hot trub, 12%),
which settled out of the hot wort at the end of the boil.
Later in the brew, an additional 17% was lost due to
adsorption to the cold-precipated proteins (cold trub, 6%)
and yeast (11%). The overall yield of xanthohumol
dropped to 30% in the beer after lagering; at this stage,
98% of the amount of xanthohumol was present as
isoxanthohumol. A substantial amount of xanthohumol
(27%) could not be accounted for.
Similar results were obtained with desmethylxantho-
humol (Figure 3). This prenylchalcone went more quickly
into solution than xanthohumol, most likely due to its
higher hydrophilicity, and showed a much faster con-
version into its isomeric prenylflavanones. In fact,
desmethylxanthohumol had completely isomerized by
the time the first wort sample was drawn from the brew
kettle, 15 min after addition of the hops. The overall
yield of desmethylxanthohumol (in the form of 6- and
8-prenylnaringenin) amounted to 11% in the beer after
lagering. The combined losses (spent hops, trub, and
yeast) totaled 60%, leaving 29% of the amount of
desmethylxanthohumol initially present in the hops
unaccounted for.
Effect of La te Hop p in g on th e Over a ll Yield . It
is common practice in the brewing industry to add a
second portion of hops late in the wort boil to retain a
greater portion of the hops’ essential oils in the beer,
resulting in a stronger aroma, usually referred to as
“late hop aroma”. The effect of late hopping on the fate
of xanthohumol and desmethylxanthohumol was also
investigated (Figure 4). To obtain a beer with ∼12 BU,
the total amount of hops added was increased to 265 g,
of which 128 g was added as “kettle” hops and 137 g as
“finish hops”. The kettle hops remained in the boiling
wort for 75 min, and the finish hops were added during
the last 5 min of the wort boil. Shortly after removal of
the (combined) spent hops, the xanthohumol amount in
the wort was 53% of the amount in the total of hops
added (as compared to 69% in the first brew), reflecting
RESULTS AND DISCUSSION
The prenylated flavonoids of hops are predominantly
of the chalcone type (Figure 1). No less than 95% of the
prenylflavonoid fraction of hops consists of xanthohu-
mol, desmethylxanthohumol, and 3′-geranylchalconar-
ingenin (3′-geranyl-2′,4′,6′,4-tetrahydroxychalcone). Hops
contain minor quantities of prenylated flavanones,
which are most likely formed during drying and storage
of commercial hops (Stevens et al., 1997). By contrast,
beer contains much higher levels of prenylflavanones
than prenylchalcones (Stevens et al., 1999). Obviously,
beer prenylflavanones are formed from their isomeric
chalcones during the brewing process. The conversion
of chalcones into flavanones requires the presence of a
2′- or 6′-hydroxyl group. Because xanthohumol has one
hydroxyl available for ring closure, isoxanthohumol is
the only possible cyclization product. Desmethylxan-
thohumol and 3′-geranylchalconaringenin, on the other
hand, have two hydroxyl functions that can participate
in cyclization, yielding two flavanones for each chalcone
(Figure 1). Other ingredients of beer, barley (Hordeum
vulgare), rice (Oryza sativa), and corn (Zea mays), do
not contain prenylflavonoids. The flavonoids of barley
are of the proanthocyanidin type (McMurrough, 1981)
which do not interfere with the prenylflavonoids of hops
in our LC/MS method.
Effect of Ea r ly Hop p in g on th e Over a ll Yield . In
the first brewing experiment, a lager-type beer was
brewed with a single dose of hops (172 g), calculated to
give a beer with ∼12 BU. The hops were added 15 min
after the wort was brought to a boil and remained in
contact with the boiling wort for 75 min. Flavonoid
amounts present in the brew at various stages or in the

Fate of Xanthohumol from Hops to Beer
J. Agric. Food Chem., Vol. 47, No. 6, 1999 2425
F igu r e 4. Prenylflavonoids measured in double-hopped brew. See Figure 3 for keys to flavonoids and brewing stages.
Ta ble 1. Ap p a r en t Solu bility of Xa n th oh u m ol a n d
the brief contact between the finish hops and the boiling
Isoxa n th oh u m ol
wort. At this stage, 40% of total added xanthohumol was
found in the spent hops and the hot trub. Further losses
concna (mg/L)
medium
water
temp (°C)
xanthohumol
isoxanthohumol
were due to adsorption to the cold-precipitated proteins
(cold trub, 9%) and to the yeast slug (14%). About 22%
of the hops’ xanthohumol was found in the beer after
lagering, largely in the form of isoxanthohumol. The
losses totaled 64%, leaving 14% unaccounted for.
With desmethylxanthohumol, a significant increase
in the levels of 6- and 8-prenylnaringenin was seen 20
min after addition of the finish hops (this period
included 5 min of wort boiling and 15 min of whirlpool
settling) (Figure 4). Moreover, this increase almost
equals the amount of desmethylxanthohumol present
in the finish hops, indicating a virtually quantitative
extraction of desmethylxanthohumol. By contrast, a
mere 45% was found in the wort 15 min after addition
of the kettle hops, and the total amount of prenylnar-
ingenins dropped to 36% of the hops’ desmethylxantho-
humol just before addition of the finish hops. Appar-
ently, extraction and isomerization of desmethylxantho-
humol are complete within a few minutes, after which
time a gradual loss of prenylnaringenins from solution
sets in as a result of adsorption to the spent hops (21%)
and insoluble proteins (hot trub, 13%). Although a
further 14% precipitated with the cold trub, most of the
prenylnaringenins were lost during fermentation as a
result of adsorption to the yeast cells (32%). About 10%
of the hops’ desmethylxanthomumol remained, in fla-
vanone form, in the lager beer, leaving 11% unaccounted
for.
Ad sor p tion of P r en ylfla von oid s to Tr u b P r o-
tein s a n d Yea st. Several processes determine the fate
of xanthohumol from hops to beer: extraction, isomer-
ization, and adsorption. Whereas extraction and isomer-
ization have already been dealt with in some detail, the
process of adsorption is more complex and needs further
discussion. Many studies on the interaction between
polyphenols and proteins [see, e.g., McMurrough et al.
(1995) and Siebert and Lynn (1998)] have led to the
general belief that hydrophobic forces and hydrogen
bonding play important roles in the interaction between
phenolic groups and proteins (Outtrup, 1992; Haslam,
1996). Most of these studies were focused on catechin,
epicatechin, and their oligomers (condensed tannins),
which are much more soluble in water than are the hop
prenylflavonoids.
23
8
23
8
23
8
23
8
1.3
1.1
3.5
1.8
5.3
4.0
8.7
7.0
9.5
8.9
5.0
3.9
5.5
5% ethanol
beer
4.4
27.2
23.8
39.6
37.6
43.3
37.3
low SG wortb
high SG wortc
23
8
a
Mean of two measurements by LC/tandem MS or LC/UV.
Specific gravity ) 1.034 (8.5 °P). c Specific gravity ) 1.052 (13
°P).
b
As with other polyphenols, adsorption of prenylfla-
vonoids is assumed to be an equilibrium process in
which the distribution of bound and unbound molecules
depends on the nature of the complexing materials and
the solubility of the prenylflavonoid in the medium. The
solubilities of xanthohumol and isoxanthohumol were
determined in five different aqueous media (Table 1).
Isoxanthohumol is more soluble in the true solvents,
water and 5% ethanol, than xanthohumol. It was also
more soluble in the beer and in the worts. The beer and
the worts contain carbohydrates (1, 8.4, and 13% w/w,
respectively), which keep a higher amount of both
prenylflavonoids in solution, presumably as soluble
complexes (referred to as “apparent solubility” in Table
1). In support of the complexation hypothesis, an
increase in the apparent solubility was observed with
the wort having a higher specific gravity, that is, a
higher concentration of carbohydrates.
Taking the results of the brew trials and the solubility
experiments together, it seems reasonable to assume
that the amounts of freely dissolved xanthohumol and
isoxanthohumol are small compared to the amounts
bound to trub proteins and wort carbohydrates. After
whirlpool settling, there is much more isoxanthohumol
than xanthohumol in the brew as a result of isomeriza-
tion, yet more xanthohumol was recovered from the hot
trubs in both brews. Isoxanthohumol shows a higher
affinity for the wort carbohydrates than xanthohumol
(Table 1), and this interaction may reduce the pool of
free isoxanthohumol available for binding to trub pro-

2426 J. Agric. Food Chem., Vol. 47, No. 6, 1999
Stevens et al.
teins. Differences in the relative affinity of xanthohumol
and isoxanthohumol for the trub and yeast surfaces
could also play a role; however, this was not investigated
in this study. It is hypothesized that these differences
in prenylflavonoid-macromolecule affinities result in
the greater proportion of xanthohumol measured in the
protein residues of the trub materials and the yeast.
Yeast proteins accessible to prenylflavonoids are located
in the outer layer of the cell walls (Stewart and Russell,
1981). In the solubility experiments, the contribution
of trub proteins to prenylflavonoid-macromolecule com-
plexation was assumed to be minimal because any
cloudiness formed on refrigeration of the worts was
removed by centrifugation prior to addition of the
flavonoids.
There is little doubt that adsorption caused a gradual
decrease of the prenylnaringenin levels during wort
boiling in both brewing trials. Wort boiling induces
denaturation of proteins, thereby enhancing the surface
available for interaction with prenylflavonoids (Haslam,
1996). Denaturated proteins may form aggregates,
which become visible as a trub-like precipitate on
cooling of the wort samples taken during the boil. These
precipitates were removed by centrifugation prior to LC/
MS analysis, hence the apparent decrease of the pre-
nylnaringenin levels in the supernatants. Whirlpool
settling resulted in precipitation of trub particles while
the wort was still hot. Under these circumstances, it is
more favorable for the prenylnaringenins to remain in
solution, even after cooling. This may explain why lower
prenylnaringenin levels were measured in the pre-
whirlpool worts. Although less prominent, the same
probably underlies the sudden jump in the xanthohumol
and isoxanthohumol levels after whirlpool settling in
the first brew. Unlike the prenylnaringenins, the com-
bined xanthohumol/isoxanthohumol concentration shows
a net increase, caused by a positive balance between
continuing release of xanthohumol from the hops and
loss due to adsorption.
The solubility experiments also demonstrate that
solubility hardly dictates the isoxanthohumol levels of
commercial beers, which vary greatly, from absence in
beers bittered with hop extract preparations to 3.44
mg/L in highly hopped beers (Stevens et al., 1999). From
Table 1 it can be seen that a saturated solution of
isoxanthohumol in beer contains ∼8 times more isox-
anthohumol than a beer richest in prenylflavonoids
(3.44 mg/L). The brew trials suggest rather that adsorp-
tion of prenylflavonoids is more important than the hop
dosage for the final prenylflavonoid concentration in the
beer. Despite the extra xanthohumol added in the
second brew, both brews yielded beers with about the
same isoxanthohumol levels (0.62 and 0.65 mg/L, re-
spectively).
F igu r e 5. (A) Isomerization of xanthohumol in a model
system (boiling water, 10% sucrose, pH 5.4). (B) Plot of ln-
[xanthohumol] vs boiling time (half-life ) 30 min).
et al., 1999). However, the prenylflavonoids in beer are
largely in the flavanone form, which can be quantified
by the method with sufficient accuracy ((5% for iso-
xanthohumol and 6-prenylnaringenin, (8% for 8-pre-
nylnaringenin). Furthermore, the accuracy with regard
to xanthohumol was shown to vary within (5% when
dissolved in methanol, also used in this study for
extraction of hops, spent hops, trubs, and yeast. In
addition, the extraction of the trubs and yeast with
methanol proved to be satisfactory: a second extraction
of the remaining solids yielded only 0-3.7% of the
flavonoids found in the initial extracts. Therefore, some
cause(s) other than experimental error must account for
the bulk of the missing 27% of the initial xanthohumol.
Another explanation for the missing amounts could
be the participation of xanthohumol and desmethyl-
xanthohumol in chemical reactions besides isomeriza-
tion. Therefore, the conversion of xanthohumol into
isoxanthohumol was carried out in a model system to
eliminate loss of xanthohumol as a result of adsorption.
The decrease of xanthohumol was in very good balance
with the formation of isoxanthohumol in boiling water
(Figure 5). Assuming first-order kinetics, the half-life
of xanthohumol was calculated to be 30 min. These
results led to the conclusion that the contribution of
possible side reactions to the degradative loss of xan-
thohumol is negligible in boiling water. It is assumed
the same is true for boiling wort.
F a te of th e “Missin g” F la von oid Am ou n ts. Sub-
stantial amounts of the hops’ xanthohumol (27%) and
desmethylxanthohumol (29%) were missing from the
total of amounts measured in the final products (i.e.,
beer, yeast, trubs, and spent hops). Part of these
amounts could be attributed to less than full recovery
of “lost” fractions and measurement errors. Although
best efforts were made to recover the spent hops, trubs,
and yeast slug, some losses did occur; on the basis of
visual inspection, these are thought to be <5% in all
cases. Regarding measurement error, it can be argued
that the LC/MS method underestimates xanthohumol
in wort and beer, as shown in a previous study (Stevens
F u n ga l Meta bolism . It is known that fungi such as
Botrytis cinerea and Saccharomyces cerevisiae (brewers’
yeast) contain monooxygenases that are capable of
epoxidizing double bonds of isoprenoid moieties (Satoh
et al., 1993; Tanaka and Tahara, 1997). Xanthohumol
and related prenylflavonoids could be epoxidized in a

Fate of Xanthohumol from Hops to Beer
J. Agric. Food Chem., Vol. 47, No. 6, 1999 2427
similar fashion and then hydrolyzed or reduced to
alcohols. To test this hypothesis, a concentrated metha-
nolic extract of yeast, recovered from the raw beer, was
analyzed by LC/MS using single ion monitoring and
daughter ion scanning. Several compounds were de-
tected that had masses consistent with that of oxygen-
ated (iso)xanthohumol (Mr 370); daughter ion mass
spectra allowed these products to be characterized as
prenylated chalcones or flavanones. These compounds
were present only in trace amounts, and some occur
naturally in hops, which left doubt as to whether these
trace constituents are produced by the hops or by the
yeast. At any rate, these xanthohumol derivatives do
not nearly make up for the amount of missing xantho-
humol in the brew trials (14-27%).
Ger a n ylfla von oid s. The geranyl analogue of desm-
ethylxanthohumol, 3′-geranylchalconaringenin, is the
only known geranylchalcone from hops (Stevens et al.,
1997). The fate of this geranylflavonoid could not be
quantified because it and 8-geranylnaringenin were not
available in sufficient amounts at the time the brewing
trials were conducted. However, the chromatographic
profiles indicated that 3′-geranylchalconaringenin be-
haved very similarly to desmethylxanthohumol in both
brewing experiments. After addition of the hops to the
boiling wort, 3′-geranylchalconaringenin was rapidly
converted into two products, which were identified as
8- and 6-geranylnaringenin by LC/MS comparison with
authentic markers (Figure 2).
unaltered. This was taken as proof for H-2′′ being
located trans to the CH3-4′′ group, thus confirming the
presence of a geranyl, not neryl, substituent.
ACKNOWLEDGMENT
We are grateful to Monika Ivancic and Victor L. Hsu
for help with the NMR experiments and to Donald A.
Griffin for technical support.
LITERATURE CITED
Bohm, B. A. Chalcones and aurones. In Methods in Plant
Biochemistry, Vol. 1. Plant Phenolics; Harborne, J . B., Ed.;
Academic Press: London, U.K., 1989; pp 237-281.
Charalambous, G. Involatile constituents of beer. In Brewing
Science; Pollock, J . R. A., Ed.; Academic Press: London,
U.K., 1981; Vol. 2, pp 167-254.
Delcour, J . A. Malt and Hop Flavanoids in Pilsner Beer. In
Modern Methods of Plant Analysis, Vol. 7, Beer Analysis;
Linskens, H. F., J ackson, J . F., Eds.; Springer-Verlag:
Berlin, Germany, 1988; pp 225-240.
Ha¨nsel, R.; Schulz, J . Desmethylxanthohumol: Isolierung aus
Hopfen und Cyclisierung zu Flavanonen. Arch. Pharm.
(Weinheim) 1988, 321, 37-40.
Haslam, E. Natural polyphenols (vegetable tannins) as
drugs: possible modes of action. J . Nat. Prod. 1996, 59,
205-215.
J ain, A. C.; Gupta, R. C.; Sarpal, P. D. Synthesis of lupinifolin,
di-O-methylxanthohumol and isoxanthohumol and related
compounds. Tetrahedron 1978, 34, 3563-3567.
McGuinness, J . D.; Laws, D. R. J .; Eastmond, R.; Gardner, R.
J . The estimation and significance of catechin and dimeric
catechin in beer. J . Inst. Brew. 1975, 81, 237-241.
McMurrough, I. High-performance liquid chromatography of
flavonoids in barley and hops. J . Chromatogr. 1981, 218,
683-693.
McMurrough, I.; Madigan, D.; Smyth, M. R. Adsorption by
polyvinylpolypyrrolidone of catechins and proanthocyanidins
from beer. J . Agric. Food Chem. 1995, 43, 2687-2691.
Miranda, C. L.; Stevens, J . F.; Helmrich, A.; Henderson, M.
C.; Rodriguez, R. J .; Deinzer, M. L.; Barnes, D. W.; Buhler,
D. R. Antiproliferative and cytotoxic effects of prenylated
flavonoids from hops (Humulus lupulus) in human cancer
cell lines. Food Chem. Toxicol. 1999, in press.
A small amount of 6-geranylnaringenin was obtained
by alkali treatment of 3′-geranylchalconaringenin iso-
lated from hops. The position of its geranyl substituent
was determined by 2-D NMR: in the 1H-13C HMBC
spectrum, cross-peaks showing interaction between the
chelated 5-hydroxyl and C-6 and between H-1′′ of the
geranyl group and C-6 left no doubt that the geranyl
group was located ortho to the hydroxyl at C-5. Fur-
thermore, the only aromatic A-ring proton showed cross-
peaks with all A-ring carbons but C-5 and was therefore
assigned to H-8.
A larger amount of 6-geranylnaringenin was obtained
by geranylation of naringenin with (()-linalool. The
semisynthetic product was identical to the above 6-gera-
nylnaringenin by UV, LC/MS, and 2D-NMR. It was
treated with alkali to give 3′-geranylchalconaringenin
and an earlier eluting product, which was identified as
8-geranylnaringenin by MS and 2-D NMR spectroscopy.
In the HMBC spectrum of the latter compound, the
aromatic A-ring proton was assigned to H-6 because it
showed cross-peaks with all A-ring carbons but C-9.
Carbon-9 interacted with H-1′′ of the geranyl substitu-
ent at C-8. The identity of carbons 5, 6, and 10 was
confirmed by interactions between these carbons and
the hydrogen-bonded 5-hydroxyl proton. Full assign-
ments of proton and carbon resonances for both gera-
nylnaringenins are listed under Experimental Proce-
dures.
The condensation of naringenin with linalool results
in the formation of a double bond between C-2′′ and C-3′′
of the monoterpenyl substituent (Figure 1). Double-bond
formation may give rise to cis and trans isomers, that
is, neryl- and geranylnaringenins, and this problem was
investigated by means of 1-D NOE experiments. Ir-
radiation of the H-2′′ resonance of 8-geranylnaringenin
resulted in enhancement of the H-1′′ and H-5′′ signals,
whereas the intensity of the CH3-4′′ singlet remained
Moll, M.; Fonknechten, G.; Carnielo, M.; Flayeux, R. Changes
in polyphenols from raw materials to finished beer. Tech.
Q.sMaster Brew. Assoc. Am. 1984, 21 (2), 79-87.
Outtrup, H. Proanthocyanidins, the brewing process, and the
quality of beer. In Plant Polyphenols; Hemingway, R. W.,
Laks, P. E., Eds.; Plenum Press: New York, 1992; pp 849-
858.
Peacock, V. Fundamentals of hop chemistry. Tech. Q.sMaster
Brew. Assoc. Am. 1998, 35 (1), 4-8.
Satoh, T.; Horie, M.; Watanabe, H.; Tsuchiya, Y.; Kamei, T.
Enzymatic properties of squalene epoxidase from Saccha-
romyces cerevisiae. Biol. Pharm. Bull. 1993, 16, 349-352.
Siebert, K. J .; Lynn, P. Y. Comparison of polyphenol interac-
tions with polyvinylpolypyrrolidone and haze-active protein.
J . Am. Soc. Brew. Chem. 1998, 56, 24-31.
Stevens, J . F.; Ivancic, M.; Hsu, V. L.; Deinzer, M. L. Pre-
nylflavonoids from Humulus lupulus. Phytochemistry 1997,
44, 1575-1585.
Stevens, J . F.; Taylor, A. W.; Deinzer, M. L. Quantitative
analysis of xanthohumol and related prenylflavonoids in
hops and beer by liquid chromatography-tandem mass
spectrometry. J . Chromatogr. A 1999, 832, 97-107.
Stewart, G. G.; Russell, I. Yeast Flocculation. In Brewing
Science; Pollock, J . R. A., Ed.; Academic Press: London,
U.K., 1981; Vol. 2, pp 61-92.
Stott, K.; Stonehouse, J .; Keeler, J .; Hwang, T.-L.; Shaka, A.
J . Excitation sculpting in high-resolution nuclear magnetic
resonance spectroscopy: application to selective NOE ex-
periments. J . Am. Chem. Soc. 1995, 117, 4199-4200.

2428 J. Agric. Food Chem., Vol. 47, No. 6, 1999
Stevens et al.
Tabata, N.; Ito, M.; Tomoda, H.; Omura, S. Xanthohumols,
diacylglycerol acyltransferase inhibitors from Humulus
lupulus. Phytochemistry 1997, 46, 683-687.
Tanaka, M.; Tahara, S. FAD-Dependent epoxidase as a key
enzyme in fungal metabolism of prenylated flavonoids.
Phytochemistry 1997, 46, 433-439.
Tobe, H.; Muraki, Y.; Kitamura, K.; Komiyama, O.; Sato, Y.;
Sugioka, T.; Maruyama, H. B.; Matsuda, E.; Nagai, M. Bone
resorption inhibitors from hop extract. Biosci., Biotechnol.,
Biochem. 1997, 61, 158-159.
Received for review February 1, 1999. Revised manuscript
received March 23, 1999. Accepted March 28, 1999. The Sciex
API III Plus mass spectrometer was purchased in part through
grants from the National Science Foundation (BIR-9214371)
and from the Anheuser-Busch Companies. This work was
supported by the Anheuser-Busch Companies, the Miller
Brewing Co., and the Hop Research Council.
J F990101K