Oxidation of isohumulones induces the formation of carboxylic acids by hydrolytic cleavage

Article abstract of DOI:10.1021/jf501826h

The degradation of isohumulones in mechanistic experiments was investigated. Incubation of trans-isohumulone in the presence of l-proline led to the formation of carboxylic acids and their corresponding proline amides. In the context of isohumulones unknown amides were verified first in model incubations and then in beer for the first time by comparison with authentic reference standards via LC-MS analyses. Carboxylic acids and amides were formed preferably under oxidative conditions and increasing pH. Stable isotope experiments excluded the incorporation of molecular oxygen into carboxylic acids, strongly indicating a hydrolytic mechanism via β-dicarbonyl cleavage. The proposed mechanism includes oxidation and thereby incorporation of molecular oxygen to the isohumulone ring structure followed by hydrolytic cleavage leading to acids and amides.

Full text of DOI:10.1021/jf501826h

Article
pubs.acs.org/JAFC
Oxidation of Isohumulones Induces the Formation of Carboxylic
Acids by Hydrolytic Cleavage
Stefan Rakete, Robert Berger, Steffi Bohme, and Marcus A. Glomb*
̈
Institute of Chemistry, Food Chemistry, Martin-Luther-University Halle-Wittenberg, Kurt-Mothes-Straße 2, 06120 Halle/Saale,
Germany
ABSTRACT: The degradation of isohumulones in mechanistic experiments was investigated. Incubation of trans-isohumulone
in the presence of L-proline led to the formation of carboxylic acids and their corresponding proline amides. In the context of
isohumulones unknown amides were verified first in model incubations and then in beer for the first time by comparison with
authentic reference standards via LC-MS analyses. Carboxylic acids and amides were formed preferably under oxidative
conditions and increasing pH. Stable isotope experiments excluded the incorporation of molecular oxygen into carboxylic acids,
strongly indicating a hydrolytic mechanism via β-dicarbonyl cleavage. The proposed mechanism includes oxidation and thereby
incorporation of molecular oxygen to the isohumulone ring structure followed by hydrolytic cleavage leading to acids and amides.
KEYWORDS: hop bitter acids, isohumulones, carboxylic acids, carboxylic acid amides, β-dicarbonyl cleavage,
amino acid modifications, beer
not supported by mechanistic investigations and seems to play
only a minor role in beer aging.11
INTRODUCTION
■
Beer is one of the most popular beverages worldwide due to its
unique flavor. The particular flavor is the result of a tremendous
number of volatile and nonvolatile substances. The nature and
reactivity of these substances are important subjects to the
scientific community. Both raw materials and chemical
alterations are responsible for the final flavor composition.
Although chemical reactions occurring in bottled beer generally
are of no risk to human health, they are often related to the
generation of negatively perceived flavors. Vanderhaegen et al.
published a review on aging-related reactions in beer.1
Especially hop-based flavor compounds, in particular bitter
acids, are of great interest to researchers. During wort boiling
the α-acids (humulones) are isomerized to iso-α-acids
(isohumulones), the main bitter compounds in beer, resulting
in a trans/cis mixture for each α-acid. The trans/cis ratio is
typically around 0.4 due to the higher thermodynamic stability
of the cis isomer and depends on various parameters, such as
pH and temperature.2 Upon beer aging the trans isomers are
depleted significantly more quickly.3,4 De Cooman et al.
assumed a higher electron density leading to an increased
susceptibility toward oxidative degradation.3 In contrast,
Intelmann et al. showed the formation of polycyclic compounds
under nonoxidative conditions exclusively from trans-iso-α-
acids and related them to a harsh, lingering bitter taste.4,5
Furthermore, various oxidation products of the intact iso-α-acid
structure were identified.6,7 The formation of volatile, aroma
active substances from iso-α-acids is of major concern. In the
case of the so-called light-struck flavor, a photo-oxidative
mechanism was presented.8 Also, the hop-originated precursor
3-methyl-2-buten-1-ol is known to generate onion-like off-
flavors.9 Hashimoto and Eshima found evidence for the
formation of staling aldehydes from the alkanoyl side chain
by oxidative degradation of iso-α-acids.10 However, the
formation of these aldehydes by iso-α-acid degradation was
The formation of carboxylic acids from iso-α-acids was
reported under oxidative conditions, apparently also from the
alkanoic and alkenoic side chain.12−14 No mechanism for their
formation is given in the literature. As a matter of fact, iso-α-
acids can also be recognized as β-dicarbonyl structures and may
show the same reactions as other β-dicarbonyls mentioned in
the literature. In this context Davidek et al. showed that acetic
acid is the main cleavage product from 1-deoxyglucosone
formed by hydrolytic β-dicarbonyl cleavage in Maillard reaction
systems.15 Today, hydrolytic β-dicarbonyl cleavage is consid-
ered to be one of the major degradation pathways for
carbohydrates.16,17
The scope of the present work was to elucidate the molecular
mechanisms leading to flavor active carboxylic acids from
degradation of iso-α-acids. The formation of these acids was
studied under defined conditions. As a strong proof for
hydrolytic β-dicarbonyl cleavage the corresponding carboxylic
acid amides from the reaction with amines were found in such
incubations.18 This included the syntheses of authentic
reference compounds to unambiguously verify these amides.
Thus, the interaction of iso-α-acids and amines was established
for the first time in a mechanistic experiment. In particular,
trans-isohumulone, the most relevant trans-iso-α-acid in beer,
was incubated in the presence of L-proline, which represents
approximately 50% of the total free amino acids in beer.19
MATERIAL AND METHODS
■
Chemicals. Chemicals of the highest available grade were obtained
from Sigma-Aldrich (Munich, Germany) unless otherwise indicated.
Received: April 16, 2014
Revised: July 3, 2014
Accepted: July 5, 2014
Published: July 15, 2014
© 2014 American Chemical Society
7541
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
Acetonitrile and methanol (HPLC gradient grade) were purchased
from VWR (Darmstadt, Germany). Silica gel for column chromatog-
raphy (0.06−0.2 mm) was obtained from Roth (Karlsruhe, Germany).
Diethylenetriaminepentaacetic acid (DTPA) was obtained from Merck
(Darmstadt, Germany). Methanol-d4 (CD3OD) and trichlorome-
further purified by preparative HPLC as described below (water/
methanol (25:75, v/v)) to give pure humulone. To obtain trans-
isohumulone, humulone was isomerized by using a modified literature
protocol.22 One gram of pure humulone was dissolved in 250 mL of
methanol and flushed with argon before the isomerization was begun.
The solution was radiated with a mercury-vapor lamp until the educt
was completely vanished (monitored by HPLC). After removal of the
solvent under reduced pressure, the crude product was purified by
preparative HPLC as described below (water/methanol (50:50, v/v))
to yield trans-isohumulone with a purity of 98% (HPLC-DAD).
Structure was confirmed by NMR experiments (1H, 13C; CD3OD) in
accordance with literature data.23
Preparation of trans-Humulinic Acid. Pure humulone was
converted into trans- and cis-humulinic acid by using a modified
literature protocol.24 Humulone was stirred in 2 M aqueous sodium
hydroxide at 95 °C for 2 h to give a mixture of trans- and cis-humulinic
acid (89:11). After acidification with hydrochloric acid, the solution
was extracted with diethyl ether and concentrated under reduced
pressure. The individual isomers were isolated by preparative HPLC as
described below (water/methanol (65:35, v/v)) to yield pure trans-
humulinic acid. NMR spectra (1H, 13C; CD3OD) were in line with the
literature.24,25
Preparation of Hydroxy-trans-alloisohumulone. Pure trans-
isohumulone was incubated in 0.1 M phosphate buffer at pH 7 after
the addition of 10% vol ethanol for 1 week at 50 °C. After acidification
with hydrochloric acid, the incubation was extracted with ethyl acetate,
dried over sodium sulfate, and concentrated in vacuo. The crude
extract was further processed by preparative HPLC (water/methanol
(80:20, v/v)) to give hydroxy-trans-alloisohumulone with a purity of
95% (HPLC-DAD). NMR spectra (1H, 13C; CD3OD) were in line
with the literature.6
MLCCC. The MLCCC system (Ito, multilayer separator-extractor
model, P. C. Inc., Potomac, MD, USA) was equipped with a constant-
flow pump (Waters 510, Waters Corp., Milford, MA, USA) and a Jasco
UV-2075 detector (Jasco, Gross-Umstadt, Germany) operating at 250
nm. Eluted liquids were collected in fractions of 16 mL with a fraction
collector (LKB Ultrorac 7000). Chromatograms were recorded on a
plotter (Servogor 200). The multilayer coil was prepared from a 1.6
mm inner diameter polytetrafluoroethylene (PTFE) tubing. The total
capacity was 270 mL. The MLCCC was run at a revolution speed of
790 rpm and a flow rate of the mobile phase of 2 mL/min in head-to-
tail modus. Separation was performed by using binary solvent systems
(upper phase as stationary phase and lower phase as mobile phase).
The sample was dissolved in a 1:1 mixture of upper and lower phase
(10 mL) and injected into the system. The collected fractions were
dried for further processing.
thane-d (CDCl ) were purchased from Armar Chemicals (Dottingen,
̈
3
Switzerland). 4-Methyl-2-pentenoic acid and isovaleryl chloride were
obtained from Alfa Aesar (Karlsruhe, Germany). The hop extract was
kindly provided by the Hasseroder brewery (Wernigerode, Germany).
̈
4-Methyl-3-pentenoic Acid. 4-Methyl-3-pentenoic acid was
prepared according to the literature with some modifications.20 One
milliliter of 4-methyl-2-pentenoic acid (8.4 mmol) was dissolved in 20
mL of 6 M potassium hydroxide solution and heated to 105 °C for 48
h. After cooling to room temperature, the reaction mixture was
acidified to pH 1 with 2 M hydrochloric acid and extracted three times
with dichloromethane (30 mL). The combined organic layers were
dried with sodium sulfate, and the solvent was removed under reduced
pressure to afford a mixture of the target compound and the educt as a
colorless liquid (90% 4-methyl-3-pentenoic acid, determined by gas
1
chromatography as its trimethylsilyl derivative and NMR). H NMR
(400 MHz, CDCl3), δ 1.65 (s, 3H), 1.79 (d, 3H, 4J = 0.9 Hz), 3.08 (d,
2H, 3J = 7.2 Hz), 5.30 (m, 1H); 13C NMR (400 MHz, CDCl3), δ 18.1,
25.8, 33.6, 115.2, 136.5, 178.3.
General Method for the Preparation of N-Acyl-L-proline
Derivatives. The preparation was based on a method described in the
literature with some modifications.21 One equivalent of L-proline was
suspended with 1.1 equiv of triethylamine in 10 mL of dry
dichloromethane. The mixture was flushed with argon and cooled to
0 °C. Then 1 equiv of carboxylic acid chloride dissolved in 5 mL of dry
dichloromethane was added dropwise. The reaction mixture was then
stirred at room temperature for 15 h. After completion, the reaction
mixture was washed two times with 15 mL of 1 M hydrochloric acid.
The aqueous phase was re-extracted with 25 mL of dichloromethane.
The combined organic layers were dried with sodium sulfate, and the
solvent was removed under reduced pressure. The crude product was
purified using flash chromatography (silica gel 60, n-hexane/acetone/
trifluoroacetic acid 66:33:1). Fractions containing the desired product
(TLC (silica gel 60, n-hexane/acetone/trifluoroacetic acid 66:33:1);
detection, mixture of ninhydrine and p-toluenesulfonic acid (1%, 3%)
in isopropanol) were collected and concentrated in vacuo.
N-Isovaleryl-L-proline (3-Methylbutyric Acid Proline Amide).
Six hundred and thirty-three milligrams of L-proline (5.5 mmol), 837
μL of triethylamine (6 mmol), and 671 μL of isobutyryl chloride (5.5
mmol) were used according to the general procedure. The product
(TLC, Rf 0.31) was yielded as a yellowish oil (98%) from which the
trans derivative was precipitated as colorless crystals at room
1
Preparative HPLC-UV. A Besta HD 2-200 pump (Wilhelmsfeld,
Germany) was used at a flow rate of 20 mL/min. Elution of material
was monitored by a UV detector (Jasco UV-2075). Detection
wavelength for α-acids, iso-α-acids, and trans-humulinic acid was set
at 270 and 250 nm for hydroxy-trans-alloisohumulone. Chromato-
graphic separations were performed on a stainless steel column
(KNAUER, Berlin, Germany; Eurospher-100-10 C18, 250 × 20 mm,
10 μm). All target substances were separated isocratically with
mixtures of water and methanol containing 0.2% formic acid. The
fractions containing target material were combined and dried under
reduced pressure.
temperature. H NMR (500 MHz, CDCl3, trans derivative), δ 0.97−
3
0.99 (2d, 6H, J = 6.4 Hz), 1.85−2.35 (m, 7H), 3.45−3.68 (m, 2H),
4.54−4.59 (m, 1H), 11.00 (s br, 1H); 13C NMR (125 MHz, CDCl3,
trans derivative), δ 22.6. 22.7, 24.8, 25.7, 43.2, 48.0, 59.6, 173.9, 174.4;
HR-MS, m/z 198.1135 (found), m/z 198.1135 (calcd for C10H17NO3
[M − H]−).
N-4-Methyl-3-pentenoyl-L-proline (4-Methyl-3-pentenoic
Acid Proline Amide). One milliliter of 4-methyl-3-pentenoic acid
(7.6 mmol) was dissolved in 1 mL of thionyl chloride (13.8 mmol)
and refluxed for 2 h. Excess thionyl chloride was removed under
reduced pressure. The product, 967 mg of L-proline (8.4 mmol), and
1.25 mL of triethylamine (9 mmol) were used according to the general
procedure. The product (TLC, Rf 0.31) was yielded as a brown oil
(74%). 1H NMR (500 MHz, CDCl3, trans derivative), δ 1.63 (s, 3H),
1.73 (s, 3H), 1.88−2.17 (m, 2H), 2.18−2.35 (m, 2H), 3.10 (d, 2H, 3J
trans-Isohumulone Incubation Setup. Individual iso-α-acids
were incubated in 0.1 M ethanolic phosphate buffer (10% vol. ethanol)
at two different pH values (5 and 7) with 25 mM L-proline for up to
21 days at 50 °C. The final iso-α-acids concentration was 5 mM. The
incubations were performed under aerated and deaerated conditions.
Deaerated conditions were achieved by adding diethylenetriamine-
pentaacetic acid (1 mM), degassing the buffer with helium, and
flushing the residual headspace with argon. All samples were prepared
in triplicate.
3
= 6.5 Hz), 3.38−3.68 (m, 2H), 4.48−4.57 (m, 1H), 5.23 (t, 1H, J =
6.3 Hz), 10.30 (s br, 1H); 13C NMR (125 MHz, CDCl3, trans
derivative), δ 18.2, 24.8, 25.7, 28.4, 34.5, 47.8, 59.7, 115.2, 136.3,
173.7, 174.2; HR-MS, m/z 210.1135 (found), m/z 210.1136 (calcd for
C11H17NO3 [M − H]−).
Preparation of n-Humulone and trans-Isohumulone. A
mixture of α-acids was obtained by multilayer coil countercurrent
chromatography (MLCCC) using n-hexane, methanol, and water as
tertiary system (10:8.2, v/v/v). The crude MLCCC product was
Incubation of trans-Isohumulone with H218O. The buffer (300
μL) described above (aerated, pH 7) was lyophilized and redissolved
with 270 μL of H218O (97 atom % 18O). Thirty microliters of an
ethanolic trans-isohumulone solution (50 mM) was added to meet the
7542
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
conditions described above. Samples were incubated for 21 days at 50
°C.
Table 1. Mass Spectrometer Parameters for Quantitation
(MRM Mode)
Analytical HPLC-UV for Determination of Residual Iso-α-
acid Levels. Analyses were performed on a Jasco HPLC system (PU-
2089plus, AS-2055plus, UV-2075plus; Gross-Umstadt, Germany).
Chromatographic separations were performed on a stainless steel
column (KNAUER; Eurospher 100-5 C18, 250 × 4.0 mm, 5 μm)
using a flow rate of 1 mL/min at 20 °C. The mobile phase used
consisted of eluent A (water, 20 mg/L DTPA, 0.2% formic acid) and
eluent B (acetonitrile/water (90:10, v/v), 0.2% formic acid). The
incubation solutions were diluted with methanol and injected at 50%
B. After 20 min, B was raised to 100% within 5 min (held for 5 min)
and returned to starting conditions within 2 min (held for 8 min). The
detection wavelength was set at 270 nm.
Extraction and Derivatization of Carboxylic Acids from Iso-
α-acid Incubations. First, all of the samples (200 μL) were spiked
with 100 μL of aqueous butyric acid solution (0.1 mM) as internal
standard. After acidification with concentrated hydrochloric acid (25
μL), the precipitate was removed by centrifugation. An aliquot of the
supernatant (200 μL) was saturated with sodium chloride and
extracted with 300 μL of toluene. Two hundred microliters of the
organic layer was dried with sodium sulfate. An aliquot was then
derivatized with an equal volume of hexamethyldisilazane at 50 °C for
3 h prior to GC-MS analysis. External calibration samples were
prepared in the same manner.
GC-LCI-MS Determination of Carboxylic Acid Trimethylsilyl
Derivatives. One microliter of the samples described above was
injected to a Thermo Finnigan Trace GC Ultra coupled to a Thermo
Trace DSQ (Thermo Fisher Scientific, Bremen, Germany). Separation
was carried out on a DB-5MS GC column (30 m × 0.25 mm × 0.25
μm; Agilent Technologies, Palo Alto, CA, USA). Helium 5.0 was used
as carrier gas in constant flow mode (linear velocity = 28 cm/s, flow =
1 mL/min). The split flow was set at 20 mL/min. Injector and transfer
lines were heated to 220 and 250 °C, respectively. The oven program
was as follows: 40 °C (1 min), 4 °C/min to 100 °C (0 min), 20 °C/
min to 270 °C (10 min). LCI mass spectra were obtained at 70 eV
(source, 190 °C; emission current, 50 μA) in SIM mode using
methanol as reagent gas. The retention times and SIM parameters for
the silylated acid derivatives were as follows: butyric acid at 7.7 min
(m/z 161), isovaleric acid at 9.2 min (m/z 175), and 4-methyl-3-
pentenoic acid at 14.3 min (m/z 187).
LC-MS/MS Analyses of Humulinic Acid and Hydroxy-trans-
alloisohumulone. Analyses were performed on a Jasco HPLC
system (AS2057plus and PU-2080plus). Chromatographic separations
were performed on a stainless steel column (KNAUER; Eurospher
100-5 C18, 250 × 4.6 mm) using a flow rate of 1 mL/min at 25 °C.
The mobile phase used consisted of eluent A (30 mM ammonium
formate buffer, pH 5.5) and eluent B (methanol). For analyses of
incubations, samples were diluted with methanol (1:1). The respective
samples (20 μL) were injected to the LC system at 25% B. After 10
min, B was raised to 30% within 12 min and then to 70% within 3 min
(held for 5 min) and returned to starting conditions within 1 min
(held 5 min). The LC system was connected to the probe of the mass
spectrometer. The mass analyses were performed using an Applied
Biosystems API 4000 quadrupole instrument (Applied Biosystems,
Foster City, CA, USA) equipped with an API source using an
electrospray ionization (ESI) interface. Nitrogen was used as sheath
and auxiliary gas. For electrospray ionization in negative mode the
following specifications were used: sprayer capillary voltage, −4.5 kV;
nebulizing gas flow, 50 mL/min; heating gas, 60 mL/min at 500 °C;
and curtain gas, 40 mL/min. For multiple reaction monitoring
(MRM), declustering potential (DP), collision energy (CE), and cell
exit potential (CXP) were optimized by syringe injection of reference
compounds. Quantitation was based on matrix-assisted calibration.
The LC-MS/MS parameters are given in Table 1A.
mass (amu)
tR
DP
CE
CXP
(V)
(min)
Q1
Q3
(V)
(eV)
A
trans-humulinic acid
21.2
23.5
265.1
377.1
154.9
195.2
−75
−90
−30
−41
−24
−16
hydroxy-trans-
alloisohumulone
B
3-methylbutyric acid
proline amide
22.0
24.7
200.2
116.1
116.2
50
56
20
25
6
4-methyl-3-pentenoic
acid proline amide
212.1
10
eluent A (water, 1.2 mL/L heptafluorobutyric acid) and eluent B
(methanol/water (70:30, v/v), 1.2 mL/L heptafluorobutyric acid). For
analyses the samples were diluted with methanol (1:1) and injected to
the LC system at 20% B. After 10 min, B was raised to 70% within 20
min and then to 100% within 2 min (held for 13 min) and returned to
starting conditions within 5 min for equilibration (10 min). The LC
system was connected to the probe of the mass spectrometer. The
mass analyses were performed using an Applied Biosystems API 4000
quadrupole instrument equipped with an API source using an
electrospray ionization (ESI) interface. Nitrogen was used as sheath
and auxiliary gas. For electrospray ionization in positive mode the
following specifications were used: sprayer capillary voltage, 4 kV;
nebulizing gas flow, 70 mL/min; heating gas, 70 mL/min at 500 °C;
and curtain gas, 50 mL/min. For MRM, declustering potential (DP),
collision energy (CE), and cell exit potential (CXP) were optimized by
syringe injection of reference compounds. Quantitation was based on
matrix-assisted calibration. The LC-MS/MS parameters are given in
Table 1B. Proline amides were extracted from Pilsener type beer after
acidification to pH 2 with hydrochloric acid by means of solid phase
extraction (Bakerbond C18 Polar Plus; Avantor, Deventer, The
Netherlands). After elution with methanol, the extract was
concentrated and redissolved in methanol/water (7:3, v/v). The
crude extract was filtered and fractionated by analytical HPLC runs
with the setup described above to isolate 3-methylbutyric acid proline
amide. Fractions were collected by an automated fraction collector (1
min/fraction; CHF122SB, Advantec, Japan). Fractions containing the
desired amide were dried and redissolved in a mixture of methanol and
water (1:1) for LC-MS analysis. MS2 spectra were generated by using
the parameters described above (collision energy = 19 eV). The same
experiment was performed with the authentic reference compound.
Nuclear Magnetic Resonance Spectroscopy (NMR). NMR
spectra were recorded on a Varian VXR 400 spectrometer operating at
400 MHz for 1H and 100 MHz for 13C or on a Varian Unity Inova 500
instrument operating at 500 MHz for 1H and 125 MHz for 13C,
respectively. Chemical shifts are given relative to external SiMe4.
Accurate Mass Determination (HR-MS). The high-resolution
positive ion ESI mass spectra (HR-MS) were obtained from a Bruker
Apex III Fourier transform ion cyclotron resonance (FT-ICR) mass
spectrometer (Bruker Daltonics, Billerica, MA, USA) equipped with an
Infinity cell, a 7.0 T superconducting magnet (Bruker, Karlsruhe,
Germany), a radio frequency (RF)-only hexapole ion guide, and an
external electrospray ion source (Apollo; Agilent, off-axis spray).
Nitrogen was used as drying gas at 150 °C. The samples were
dissolved in methanol, and the solutions were introduced continuously
via a syringe pump at a flow rate of 120 μL/h. The data were acquired
with 256K data points and zero filled to 1024K by averaging 32 scans.
RESULTS
■
LC-MS/MS Analyses of Amides. Analyses were performed on a
Jasco HPLC system (AS2057plus and PU-2080plus). Chromato-
graphic separations were performed on a stainless steel column (Vydac
218TP54, RP18, 5 μm 250 × 4.6 mm; Hesperia, CA, USA) using a
flow rate of 1 mL/min at 25 °C. The mobile phase used consisted of
Preparation of N-Acyl-L-proline Derivatives. For the
analyses of the amino acid amides authentic reference material
was necessary. In the case of proline the synthesis led to cis- and
trans-configured amides due to the partial double character of
7543
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
the amide moiety. In this context the trans isomer is considered
to be the thermodynamically more stable species, which was
confirmed by the performed syntheses.26,27 The proline-
originated α-methine proton was used as a probe for the cis/
trans ratio (1:5.7−6.5, Figure 1). The crystals precipitated from
Figure 1. 3-Methylbutyric acid proline amide 1H NMR spectra
(excerpt) showing the α-methine proton of the trans (4.57 ppm) and
cis (4.41 ppm) isomer.
3-methylbutyric acid proline amide were identified as the trans
derivative by crystal structure analysis (data not shown).
Interestingly, redissolution of trans crystals in deuterated
chloroform, followed by NMR analysis, resulted in the
reconstitution of the isomeric ratio stated above. Furthermore,
the isomers could not be distinguished by LC-MS analyses.
Determination of Carboxylic Acids in Incubations. For
quantitation the carboxylic acids had to be separated from the
matrix. Residual trans-isohumulone led to artifacts during
derivatization with hexamethyldisilazane and thus had to be
removed by precipitation through acidification with hydro-
chloric acid. The carboxylic acids were extracted from the
isohumulone-free supernatant (verified by HPLC-UV) with
toluene. Butyric acid was used as internal standard because it
cannot be formed from trans-isohumulone. The optimized
workup provided a minimization of matrix effects and led to
reproducible results.
Formation of trans-Isohumulone Degradation Prod-
ucts. trans-Isohumulone was incubated from 3 to 21 days at
two different pH values under aerated and deaerated
conditions. An overview of the substances analyzed in
incubations is given in Figure 2.
Figure 3 (1) shows the degradation of trans-isohumulone. At
acidic pH values it was almost completely vanished after 3
weeks (up to 97% degradation), although slightly more slowly
at deaerated conditions (95%). On the contrary, relatively high
amounts of unaltered trans-isohumulone were remained at pH
7, at which 72% was degraded in deaerated samples and 85%
under oxidative conditions.
Figure 2. Overview of the structures analyzed in trans-isohumulone
(1) incubations: 3-methylbutyric acid (2a, R = OH), 3-methylbutyric
acid proline amide (2b, R = L-proline), 4-methyl-3-pentenoic acid (3a,
R = OH), 4-methyl-3-pentenoic acid proline amide (3b, R = L-
proline), trans-humulinic acid (4), hydroxy-trans-alloisohumulone (5).
third of the level determined at pH 7. A pH of 5 and deaeration
led to concentrations of 24 mmol/mol after 3 weeks. The
results for the formation of 4-methyl-3-pentenoic acid (MPA)
(Figure 3, 3a) revealed an almost identical pattern as seen for
MBA. However, MPA levels were about 4−6 times lower
compared to MBA concentrations. Again, the maximum was
found at pH 7 and oxidative conditions (20 mmol/mol),
whereas under deaeration half as much MPA was formed (10
mmol/mol). At pH 5 the levels were about 3 times lower after
3 weeks (4.1 mmol/mol for deaerated vs 6.6 mmol/mol for
aerated samples). The carboxylic acids in the stable isotope
experiment with H218O were analyzed analogously to the
regular incubations. The incorporation of 18O was monitored
by analyzing additionally the [M + 2] molecular ion. For pH 7
and oxidative conditions about 90% of both carboxylic acids
incorporated 18O. As an example, the SIM chromatograms for
the unlabeled (m/z 175) and labeled (m/z 177) MBA are given
in Figure 4. Additional incubations were also performed
without proline. However, there were no significant differences
in the levels of carboxylic acids (data not shown).
On the basis of isolated authentic reference compounds,
trans-humulinic acid and hydroxy-trans-alloisohumulone were
also quantitated in the incubations. trans-Humulinic acid was
preferentially formed at a pH of 7 under deaerated conditions
(Figure 3, 4; 11.6 mmol/mol after 3 weeks). Oxidative
conditions led to levels up to 9.4 mmol/mol at the same pH.
In contrast, in acidic incubations trans-humulinic levels were
significantly higher at oxidative conditions (4.9 vs 3.0 mmol/
mol in deaerated incubations). Hydroxy-trans-alloisohumulone
had its maximum for pH 5 and oxidative conditions after 3 days
Along with the degradation of trans-isohumulone, the levels
of 3-methylbutyric acid (MBA) increased over the incubation
time in all incubation setups, preferably at neutral and aerated
conditions with concentrations up to 86 mmol/mol trans-
isohumulone (Figure 3, 2a). The concentrations found in
deaerated samples (35 mmol/mol) were approximately 3 times
lower. At pH 5 the highest concentrations were also observed
under oxidative conditions (34 mmol/mol), that is, about one-
7544
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
Figure 3. Degradation of trans-isohumulone (1) and formation of 3-methylbutyric acid (2a), 4-methyl-3-pentenoic acid (3a), trans-humulinic acid
○
;
△
▲
(4), hydroxy-trans-alloisohumulone (5), and 3-methylbutyric acid proline amide (2b) at pH 7 (aerated, ; deaerated, ) and pH 5 (aerated,
deaerated, ).
●
(3.9 mmol/mol) and then started slowly to decrease to a final
level of 1.1 mmol/mol (Figure 3, 5). The curve for deaerated
incubation was similar, although concentrations were slightly
lower (maximum, 3.0 mmol/mol; 0.7 mmol/mol after 3
weeks). At pH 7 levels of hydroxy-trans-alloisohumulone were
considerably higher. The maximum at aerated conditions was
reached after 3 days at 87 mmol/mol. Thereafter, the level
decreased to 36 mmol/mol within 3 weeks. At nonoxidative
conditions the maximum was reached after 2 weeks at 80
mmol/mol and decreased to 46 mmol/mol afterward.
In parallel to the carboxylic acids, the analogous L-proline
amides were found in the incubations. The results for 3-
methylbutyric acid amide (MBA amide) are given in Figure 3
(2b). Whereas for pH 7 relatively high concentrations of MBA
amide were found (up to 1800 μmol/mol), about 50 times
lower compared to MBA itself, only minor amounts (up to 54
μmol/mol) were found at pH 5. The ratio of the MBA amide
levels between aeration and deaeration at pH 7 were almost
identical to the ratio found for the analogous acid (2.4 to 2.1).
In contrast, 4-methyl-3-pentenoic acid proline amide (MPA
amide) reached a plateau after 3 days regardless of the presence
7545
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
evidence for a hydrolytic cleavage was the formation of
carboxylic acid amides. The nucleophilic L-proline nitrogen can
react analogously to water or hydroxide, respectively. Obviously
because of its weaker nucleophilicity compared to water, the
amounts of the amides remain low. However, from a
mechanistic point of view their formation is very important.
As MBA amide was preferentially formed at oxidative
conditions and the ratio between aerated and deaerated
conditions was almost equal to that of the free acids, an
identical mechanism must be assumed. To strengthen the
postulated hydrolytic mechanism a stable isotope experiment
with H218O was performed, revealing the incorporation of
approximately 90% 18O into the carboxylic acids, which
supported the proposed hydrolytic cleavage. Incorporation of
18O into free carboxylic acids was excluded by incubation of
MBA and MPA with H218O, resulting in no enrichment of the
heavy isotope into the acids at these conditions. The remaining
10% must be explained by incorporation of molecular oxygen
by yet unknown pathways, presumably including reactive
oxygen species as intermediates. Thus, the difference in the
formation of MBA under aeration versus deaeration is mainly
not caused by the incorporation of molecular oxygen to the
acid. Obviously there is an upstream oxidation step promoting
the hydrolytic cleavage of trans-isohumulone.
trans-Isohumulone is a β-dicarbonyl structure and also a
carboxylic acid with a pKa of 3.23 and is almost completely
deprotonated at pH 5 and entirely at pH 7.28 Because of the
β,β′-tricarbonyl system a negative charge after deprotonation is
distributed over seven atoms. Thus, the anticipated carbonyl
activity at the alkanoic side chain is limited in deprotonated
isohumulones and does not support the preferred generation of
carboxylic acid by hydrolytic β-dicarbonyl cleavage at neutral
pH. Considering the given hydrolytic cleavage at oxidative
conditions, we therefore postulate the mechanism shown in
Figure 6.
Figure 4. GC-LCI-MS chromatogram (SIM) of unlabeled (m/z 175;
[M + H]+) and labeled (m/z 177; [M + H]+) 3-methylbutyric acid
from trans-isohumulone incubation with H218O. The ratio of 1:9
(unlabeled vs labeled) was calculated on the basis of the peak areas.
of oxygen and remained constant at levels around 170 μmol/
mol at pH 7. An additional experiment showed that MPA
amide was stable at these conditions. Analogous to MBA amide,
MPA amide levels at pH 5 were extremely low (1.8 μmol/mol).
MBA amide was also verified in beer by LC-MS/MS
experiments (Figure 5). Although the spectrum of the isolate
After deprotonation, the isohumulate ion is oxidized via a
one-electron oxidation, according to Huavere et al., giving a
carbon-centered trans-isohumulone radical.29 Reaction with
molecular oxygen leads to the formation of the corresponding
hydroperoxy radical. After abstraction of hydrogen, the
hydroperoxide undergoes a Fenton-like redox reaction to
form the alkoxy radical and subsequently the alcohol after a
second abstraction of hydrogen. The now oxidized isohumu-
lone has lost its ability to enolize, which comes along with a
significantly increased carbonyl activity. Nucleophilic attack of
water or L-proline at the carbonyl function at the alkanoic or
alkenoic side chain then gives the free carboxylic acid or the
carboxylic acid amide based on β-dicarbonyl cleavage,
respectively. The difference between the concentrations of
MBA and MPA is likely caused by lower reactivity of the
alkenoic side chain compared to the β,β′-tricarbonyl system, a
potential steric hindrance due to the prenyl side chain at the C5
of the isohumulone and the coinciding oxidation to hydroxy-
trans-alloisohumulone. Hydrolytic cleavage of the latter would
give 4-hydroxy-4-methyl-2-pentenoic acid.
The identification of the remaining cleavage products with
intact ring structure would have been unambiguous proof for
the present mechanism. Unfortunately, it was not possible to
isolate these compounds. In contrast to the carboxylic acids and
their amides, which represent stable reaction endproducts, the
remaining isohumulone structures must be evaluated as highly
reactive and thus can be further degraded. For instance, the
cleavage of MBA gives a reductone structure, which is highly
Figure 5. MS2 of 3-methylbutyric acid proline amide (2b) isolated
from beer (A) and synthesized reference standard (B) (product ion
scan of m/z 200.2, collision energy of 19 eV).
(A) was much less intense, the fragmentation pattern was
virtually identical to that of the authentic standard (B). For
MPA amide, only traces were found in beer.
DISCUSSION
■
For the first time the degradation of trans-isohumulone in the
presence of L-proline was thoroughly investigated in a model
experiment. The clear impact of pH and oxidative conditions
on the formation of carboxylic acids provided insights into the
molecular mechanisms. In general, oxidative and hydrolytic
reactions increased with higher pH. The question was whether
the formation of carboxylic acids is based on hydrolytic or
oxidative mechanisms or even a combination of both. Strong
7546
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
Figure 6. Proposed mechanism leading to the formation of 3-methylbutyric acid (2a) and 4-methyl-3-pentenoic acid (3a) from trans-isohumulone
(1) by hydrolytic β-dicarbonyl cleavage after oxidation of the isohumulone ring structure.
susceptible to oxidation. This would also entail a loss or change
of chromophoric properties to withdraw the products from the
given detection systems. Therefore, in the present work, only
trans-humulinic acid and hydroxy-trans-alloisohumulone were
monitored as degradation products of trans-isohumulone.
Humulinic acids are known to form under harsh alkaline
treatment. Here, we were able to detect trans-humulinic acid
under mild conditions, preferably at pH 7. This per se
nonoxidized structure must be formed by direct hydrolysis of
the alkenoic side chain, which is not part of the tricarbonyl
system and maintains its carbonyl activity. trans-Humulinic acid
appeared to be stable under oxidative conditions and was not
degraded significantly in an independent incubation (data not
shown). Interestingly, higher amounts of trans-humulinic acid
were generated at pH 7 under deaeration than in aerated
incubations. This is reasonable considering the oxidation of
trans-isohumulone according to above proposed mechanism
leading to MBA and MPA. However, at pH 5 the picture is
reversed and may be the result of a suppression of the initial
oxidative step. Nevertheless, from our experiments the
hydrolytic cleavage is evident at least for the alkenoic chain
without previous oxidation. This is further supported by the
fact that the levels of trans-humulinic acid and MPA were
almost equal at pH 5, whereas MPA levels were significantly
higher than trans-humulinic concentrations at pH 7 on the
contrary. Obviously the oxidation of trans-isohumulone
decreases the yield of trans-humulinic acid.
As a major degradation product under oxidative conditions at
pH 7, 4-hydroxy-trans-alloisohumulone was isolated and
analyzed in incubations. Together with its precursor, 4-
hydroperoxy-trans-alloisohumulone, it has already been de-
scribed by Intelmann et al.6 The structure was relatively
unstable and degraded after reaching a maximum more quickly
under aerated conditions. In support, independent incubations
of hydroxy-trans-alloisohumulone indeed revealed the forma-
tion of significant amounts of MBA, most likely according to
our proposed mechanism (data not shown). We therefore must
assume that MBA is not generated by only one precursor
structure. As an alternative mechanism for the degradation of
7547
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
(6) Intelmann, D.; Hofmann, T. On the autoxidation of bitter-tasting
iso-α-acids in beer. J. Agric. Food Chem. 2010, 58, 5059−5067.
(7) Taniguchi, Y.; Matsukura, Y.; Ozaki, H.; Nishimura, K.; Shindo,
K. Identification and quantification of the oxidation products derived
from α-acids and β-acids during storage of hops (Humulus lupulus L.).
J. Agric. Food Chem. 2013, 61, 3121−3130.
(8) Burns, C. S.; Heyerick, A.; de Keukeleire, D.; Forbes, M. D. E.
Mechanism for formation of the lightstruck flavor in beer revealed by
time-resolved electron paramagnetic resonance. Chem.−Eur. J. 2001, 7,
4553−4561.
(9) Vermeulen, C.; Lejeune, I.; Tran, T. T. H.; Collin, S. Occurrence
of polyfunctional thiols in fresh lager beers. J. Agric. Food Chem. 2006,
54, 5061−5068.
(10) Hashimoto, N.; Eshima, T. Oxidative-degradation of iso-
humulones in relation to flavor stability of beer. J. Inst. Brew. 1979, 85,
136−140.
(11) de Clippeleer, J.; de Rouck, G.; de Cooman, L.; Aerts, G.
Influence of the hopping technology on the storage-induced
appearance of staling aldehydes in beer. J. Inst. Brew. 2010, 116,
381−398.
(12) Green, C. P. The volatile water-soluble fraction of hop oil. I.
Formation by the oxidation of soft resins. J. Inst. Brew. 1970, 76, 36−
40.
(13) Clarke, B. J.; Hildebrand, R. P. The isomerization of humulone
II. Preparation of reference compounds. J. Inst. Brew. 1967, 73, 60−70.
(14) Williams, R. S.; Wagner, H. P. Contribution of hop bitter
substances to beer staling mechanisms. J. Am. Soc. Brew. Chem. 1979,
37, 13−19.
(15) Davidek, T.; Devaud, S.; Robert, F.; Blank, I. Sugar
fragmentation in the Maillard reaction cascade: isotope labeling
studies on the formation of acetic acid by a hydrolytic β-dicarbonyl
cleavage mechanism. J. Agric. Food Chem. 2006, 54, 6667−6676.
(16) Voigt, M.; Glomb, M. A. Reactivity of 1-deoxy-D-erythro-hexo-
2,3-diulose: a key intermediate in the Maillard chemistry of hexoses. J.
Agric. Food Chem. 2009, 57, 4765−4770.
(17) Smuda, M.; Glomb, M. A. Novel insights into the Maillard
catalyzed degradation of maltose. J. Agric. Food Chem. 2011, 59,
13254−13264.
hydroxy-trans-alloisohumulone, Intelmann et al. assumed the
formation of polycyclic compounds.4
In conclusion, we present a novel degradation mechanism
that comprehensively explains the so far unknown formation of
carboxylic acids from isohumulones. Stable isotope experiments
with H218O as well as the generation of carboxylic acid amides
unambiguously revealed a dominant hydrolytic mechanism. We
showed that this reaction is relevant to brewing by the
verification of MBA amide in beer. The lacking fragmentation
counterparts and other reactive intermediates are subjects of
current research. The actual influence of carboxylic acids
derived from hops on the staling flavor of beer still has to be
clarified. Nevertheless, our results are in good agreement with
Williams and Wagner, who found increased levels of fatty acids
during beer aging caused by the addition of hops.14
Alternatively, MBA can also be formed by Strecker degradation
of leucine, and molecular mechanisms may not be distinguish-
able in the aging of beer. It is very likely that the hydrolytic
cleavage of isohumulones is also occurring during wort boiling
after the addition of hops. High temperatures and exposure to
oxygen would support the hydrolytic cleavage of isohumulones
based on the proposed mechanism. Consequently, the
formation of carboxylic acids from isohumulones at this stage
would also promote the so-called staling esters during
fermentation or beer storage. As we proved for L-proline,
other nucleophiles besides water are able to react with
isohumulones. Thus, also sulfur-containing substances may
lead to potent odor-active compounds, such as S-methylthioi-
sovalerate.30
AUTHOR INFORMATION
Corresponding Author
*(M.A.G.) Phone: ++049-345-5525784. E-mail: marcus.
glomb@chemie.uni-halle.de.
■
Notes
(18) Smuda, M.; Voigt, M.; Glomb, M. A. Degradation of 1-deoxy-D-
erythro-hexo-2,3-diulose in the presence of lysine leads to formation of
carboxylic acid amides. J. Agric. Food Chem. 2010, 58, 6458−6464.
(19) Vanderhaegen, B.; Neven, H.; Verstrepen, K. J.; Delvaux, F. R.;
Verachtert, H.; Derdelinckx, G. Influence of the brewing process on
furfuryl ethyl ether formation during beer aging. J. Agric. Food Chem.
2004, 52, 6755−6764.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. D. Strohl from the Institute of Organic
̈
Chemistry, Halle, Germany, for recording NMR spectra and
Dr. J. Schmidt from the Leibniz Institute of Plant Biochemistry,
Halle, Germany, for performing accurate mass determination.
(20) Caspi, E.; Varma, K. R. Synthesis of (−)-(3S)- and (+)-(3R)-4-
methyl-3D1-pentan-1-ols and (−)-(3S)- and (+)-(3R)-3D1-isocaproic
acids. J. Org. Chem. 1968, 33, 2181−2188.
REFERENCES
■
(21) Kim, B. H.; Lee, H. B.; Hwang, J. K.; Kim, Y. G. Asymmetric
induction in the conjugate addition of thioacetic acid to
methacrylamides with chiral auxiliaries. Tetrahedron Asymm. 2005,
16, 1215−1220.
(22) Sharpe, F. R.; Ormrod, I. H. L. Fast isomerization of humulone
by photo-reaction − preparation of an HPLC standard. J. Inst. Brew.
1991, 97, 33−37.
(23) Intelmann, D.; Haseleu, G.; Hofmann, T. LC-MS/MS
quantitation of hop-derived bitter compounds in beer using the
ECHO technique. J. Agric. Food Chem. 2009, 57, 1172−1182.
(24) Tynan, P. J. T.; Murrough, I. M.; Byrne, J. Preparation,
purification and separation by high-performance liquid-chromatog-
raphy of humulinic acids, dehydrohumulinic acids, and hulupones. J.
Inst. Brew. 1990, 96, 137−141.
(25) Borremans, F.; Depotter, M.; Dekeukeleire, D. C-13 NMR-
spectroscopy of hop bitter substances. Magn. Reson. Chem. 1975, 7,
415−417.
(1) Vanderhaegen, B.; Neven, H.; Verachtert, H.; Derdelinckx, G.
The chemistry of beer aging − a critical review. Food Chem. 2006, 95,
357−381.
(2) Jaskula, B.; Kafarski, P.; Aerts, G.; de Cooman, L. A kinetic study
on the isomerization of hop α-acids. J. Agric. Food Chem. 2008, 56,
6408−6415.
(3) de Cooman, L.; Aerts, G.; Overmeire, H.; de Keukeleire, D.
Alterations of the profiles of iso-α-acids during beer ageing, marked
instability of trans-iso-α-acids and implications for beer bitterness
consistency in relation to tetrahydroiso-α-acids. J. I. Brewing. 2000,
106, 169−178.
(4) Intelmann, D.; Haseleu, G.; Dunkel, A.; Lagemann, A.; Stephan,
A.; Hofmann, T. Comprehensive sensomics analysis of hop-derived
bitter compounds during storage of beer. J. Agric. Food Chem. 2011,
59, 1939−1953.
(5) Intelmann, D.; Demmer, O.; Desmer, N.; Hofmann, T. O-18
stable isotope labeling, quantitative model experiments, and molecular
dynamics simulation studies on the trans-specific degradation of the
bitter tasting iso-α-acids of beer. J. Agric. Food Chem. 2009, 57, 11014−
11023.
(26) Nishihara, H.; Nishihara, K.; Uefuji, T.; Sakota, N.
Conformation and circular dichroism of several N-acyl-L-prolines.
Bull. Chem. Soc. Jpn. 1975, 48, 553−555.
7548
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549

Journal of Agricultural and Food Chemistry
Article
(27) Pauling, L.; Corey, R. B. Configurations of polypeptide chains
with favored orientations around single bonds − 2 new pleated sheets.
Proc. Natl. Acad. Sci. U.S.A. 1951, 37, 729−740.
(28) Simpson, W. J. Ionization behavior of hop compounds and hop-
derived compounds. J. Inst. Brew. 1993, 99, 317−326.
(29) Huvaere, K.; Andersen, M. L.; Olsen, K.; Skibsted, L. H.;
Heyerick, A.; de Keukeleire, D. Radicaloid-type oxidative decom-
position of beer bittering agents revealed. Chem.−Eur. J. 2003, 9,
4693−4699.
(30) Lermusieau, G.; Bulens, M.; Collin, S. Use of GC-olfactometry
to identify the hop aromatic compounds in beer. J. Agric. Food Chem.
2001, 49, 3867−3874.
7549
dx.doi.org/10.1021/jf501826h | J. Agric. Food Chem. 2014, 62, 7541−7549