A chemical approach to searching for bioactive ingredients in cigarette smoke

Article abstract of DOI:10.1248/cpb.c12-00539

Cigarette smoke, a collection of many toxic chemicals, contributes to the pathogenesis of smoking-related diseases such as chronic obstructive pulmonary disease and cancer. Much work has been done on the chemical analysis of ingredients in cigarette smoke, but there are few reports on the active ingredients that can modify biomolecules. We used a sensitive liquid chromatography-mass spectrometry (LC/MS) and LC/ MS/MS method to show that L-tyrosine (Tyr), an amino acid with a highly reactive hydroxyl group, readily reacts with cigarette smoke extract (CSE) at body temperature (37°C) to form various Tyr derivatives. Among these derivatives were N -(3-oxobutyl)-Tyr and two acetylated compounds, N -acetyl-Tyr and O-acetyl-Tyr, which were synthesized by reaction of Tyr with methyl vinyl ketone and acetic anhydride, respectively, at 37°C. The presence of methyl vinyl ketone and acetic anhydride in CSE was confirmed by gas chromatography-mass spectrometry (GC/MS). These results indicate that Tyr can easily react with active ingredients in CSE. The present analytical methods should aid the search for active ingredients in cigarette smoke.

Full text of DOI:10.1248/cpb.c12-00539

Note
January 2013
Chem. Pharm. Bull. 61(1) 85–89 (2013)
85
A Chemical Approach to Searching for Bioactive Ingredients in
Cigarette Smoke
,a
Yuta Takahashi,a Shizuyo Horiyama,* Chie Honda,a Kiyoko Suwa,a
Kazuki Nakamura,a Masaru Kunitomo,a Shuichi Shimma,b Michisato Toyoda,b Hirofumi Sato,c
Motohiro Shizuma,c and Mitsuo Takayamad
a School of Pharmacy and Pharmaceutical Sciences, Mukogawa Women’s University; 11–68 Koshien-Kyuban-cho,
b
Nishinomiya, Hyogo 663–8179, Japan: Center for Advanced Science and Innovation, Osaka University; 1–1
c
Machikaneyama, Toyonaka, Osaka 560–0043, Japan: Osaka Municipal Technical Research Institute; 1–6–50
d
Morinomiya, Joto-ku, Osaka 536–8553, Japan: and International Graduate School of Arts and Sciences, Yokohama
City University; 22–2 Seto, Kanazawa-ku, Yokohama 236–0027, Japan.
Received June 18, 2012; accepted October 17, 2012
Cigarette smoke, a collection of many toxic chemicals, contributes to the pathogenesis of smoking-
related diseases such as chronic obstructive pulmonary disease and cancer. Much work has been done on the
chemical analysis of ingredients in cigarette smoke, but there are few reports on the active ingredients that
can modify biomolecules. We used a sensitive liquid chromatography-mass spectrometry (LC/MS) and LC/
MS/MS method to show that l-tyrosine (Tyr), an amino acid with a highly reactive hydroxyl group, read-
ily reacts with cigarette smoke extract (CSE) at body temperature (37°C) to form various Tyr derivatives.
Among these derivatives were N-(3-oxobutyl)-Tyr and two acetylated compounds, N-acetyl-Tyr and O-acetyl-
Tyr, which were synthesized by reaction of Tyr with methyl vinyl ketone and acetic anhydride, respectively,
at 37°C. The presence of methyl vinyl ketone and acetic anhydride in CSE was confirmed by gas chromatog-
raphy-mass spectrometry (GC/MS). These results indicate that Tyr can easily react with active ingredients in
CSE. The present analytical methods should aid the search for active ingredients in cigarette smoke.
Key words cigarette smoke extract; l-tyrosine; LC/MS; LC/MS/MS; GC/MS
Cigarette smoking is the major risk factor for cardiovascular
disease, atherosclerosis, lung disease and cancer.1) Cigarette
Experimental
Chemicals Frontier Lights (cigarette brand name) was
smoke is an extremely complex mixture of over 4800 chemi- purchased from Japan Tobacco Inc. (Tokyo, Japan), Cam-
cals.2) Its gas phase components are considered to penetrate bridge filters from Borgwaldt (Germany). Commercially
the small airways and alveoli, and play a significant role in purchased Tyr was purified by HPLC (Shimadzu Co., Kyoto,
the development of pulmonary parenchymal and systemic Japan). N-Acetyl-Tyr was purchased from Sigma-Aldrich Inc.
diseases.3,4) Cigarette smoke extract (CSE) is known to in- (Tokyo, Japan), O-acetyl-Tyr from Santa Cruz Biotechnol-
duce cytotoxicity in various cells.5–8) However, what is not ogy, Inc. (Santa Cruz, CA, U.S.A.), methyl vinyl ketone and
well understood is whether or not the ingredients in cigarette crotonaldehyde from Tokyo Chemical Industry Co., Ltd.,
smoke can act on biomolecules and lead to the development of (Tokyo, Japan). LC/MS grade H2O, CH3OH and guaranteed
smoking-related diseases.
grade acetic acid were obtained from Wako Pure Chemical
The chemical analysis of ingredients in cigarette smoke is Industries, Ltd. (Osaka, Japan), and LC/MS grade formic acid
one of the most challenging tasks for analysts, and extensive and guaranteed grade acetic anhydride from Nacalai Tesque,
work has been done using many analytical techniques. GC Inc. (Kyoto, Japan). An ODS column (Cosmosil 5C18-AR-II
and GC/MS are usually used for the analysis of volatile com- 4.6mm×150mm) was purchased from Nacalai Tesque, Inc.
pounds such as reactive carbonyl compounds.9,10) Despite such
Preparation of CSE CSE was prepared by a modifica-
work, there have been no reports regarding the analysis of tion of the technique described in a previous report.11) Briefly,
reaction products from the compounds in cigarette smoke and CSE was prepared by bubbling into phosphate-buffered saline
functional biomolecules.
(PBS) (1mL per three cigarettes) the mainstream of smoke
To find a clue as to whether biomolecules can be modified (gas phase) from which the particulate phase, including tars
by active ingredients in cigarette smoke, we tried to identify and nicotine, had been almost completely removed by passage
the products formed by reaction of nicotine- and tar-removed through a Cambridge filter using an aspiration pump. The
CSE, i.e., the gas-phase extract of cigarette smoke, with pump flow rate was kept constant (1L/min). Smoke was bub-
chemically reactive amino acids that make up biomolecules. bled only for 1min after lighting the cigarettes. The CSE so-
In this study, the amino acid chosen for study was l-tyrosine lution was immediately filtered through a 0.22-µm filter. The
(Tyr) because it has a hydroxyl group and its reaction prod- resulting solution was designated the 100% CSE, and stored at
ucts with CSE are easy to analyze. The reaction temperature −80°C. CSE solutions were reacted with 2mm Tyr solution at
was set at 37°C (average normal human body temperature). A 37°C for 24h, except for the experiment examining the reac-
highly sensitive LC/MS/MS system was utilized to detect and tive time course. The products in the reaction solutions were
identify trace amounts of the reaction products. We also tried analyzed by LC/MS and LC/MS/MS.
to identify the active ingredients in CSE.
Triple-Quadrupole Mass Spectrometer and HPLC Con-
ditions A Quattro Premier triple-quadrupole LC/MS (Mi-
cromass, Manchester, U.K.) with an electrospray ionization
The authors declare no conflict of interest.
© 2013 The Pharmaceutical Society of Japan
*To whom correspondence should be addressed. e-mail: horiyama@mukogawa-u.ac.jp

86
Vol. 61, No. 1
Fig. 1. Mass Chromatogram of the CSE Solutions before (a) and after (b) Reaction with the Tyr Solution for 24h at 37°C
For LC condition, a binary mobile phase consisting of 0.05% HCOOH (solvent A) and CH3OH (solvent B) was used in the following program: 2min, 5% B; 12–15min,
40% B; 16min 5% B.
(ESI) source was used for the positive and negative ion mode was set at 27°C and the injection volume was 5µL.
Q1 scan and MS/MS analysis coupled to the Alliance HT
High-Resolution (HR) MS Conditions
A HR MS
2795 Separations Module (Waters Co., Milford, MA, U.S.A.). matrix-assisted laser desorption/ionization (MALDI)-Spiral
The mass spectrometer was operated at low resolution for both time-of-flight (TOF)/TOF (JEOL Ltd., Tokyo, Japan) was used
Q1 and Q3 in the selected reaction monitoring (SRM) mode. for elemental analysis. N-Acetyl-Tyr and its Na and K adduct
The optimized conditions were as follows: source temperature, ions as well as platelet-activating factors were used as internal
120°C; desolvation temperature, 350°C; flow rate of cone ni- standards for mass calibration.
trogen, 50L/h and flow rate of desolvation nitrogen, 800L/h.
Gas Chromatography (GC)-MS Conditions A mass
Capillary and cone voltages were 3.0kV and 15–25V, respec- spectrometer (Automass SUN, JEOL Ltd., Tokyo, Japan)
tively. The flow rate of argon collision gas for fragmentation equipped with a GC (6890N, Agilent Technology Inc., Santa
in the SRM mode was 0.3mL/min (3.37–3.39×10−3 mbar) by Clara, CA, U.S.A.) was used for analysis of active compounds
which the collisional energy was optimized for the fragment that can react with Tyr in CSE. GC/MS conditions were as
ions of Tyr and other analytes (10–13eV). All chromatograph- follows: ionization energy, 70eV; current, 300µA; PM volt-
ic separations were performed using a Cosmosil 5C18-AR-II age, 500V; source temperature, 250°C; interface temperature,
column (4.6mm×150mm). The mobile phase was composed 250°C; inlet temperature, 250°C; He gas, 1.0mL/min (at con-
of water containing 0.05% formic acid as solvent A and meth- stant flow); splitless mode. Total ion current (TIC) chromatog-
anol as solvent B, and the flow rate was set at 0.4mL/min. A raphy and selected ion monitoring (SIM) chromatography GC/
linear gradient analysis was used for LC conditions in the sep- MS were used to identify and quantify the search compounds.
aration. Main gradient analysis was follows: the initial elute Chromatographic separations were performed using a Zebron
solvent consisted of 5% solvent B until t=2min, and using a capillary GC column ZB-WAX (Phenomenex Inc., Torrance,
linear gradient, it was raised to 40% solvent B at t=12min, CA, U.S.A.), 30m long×0.25mm i.d.×1.00 µm film thickness,
and then held for 3min until t=15min, and using a linear gra- phase: 100% polyethylene glycol. GC conditions to separate
dient, it was raised further to 95% solvent B at t=18min and active compounds in CSE were as follows: initial oven tem-
then held at 95% solvent B for 5min until t=23min. Finally, perature, 40°C; fold time, 2min; rate temperature 4°C/min
the gradient solvent was lowered to 5% solvent B by 24min rise to 100°C; fold time, 2min; rate temperature 15°C/min rise
and held for 2min. Other conditions of gradient analysis are to 220°C, hold time, 3min; rate temperature 30°C/min down
given in the figure caption (Fig. 1). The LC oven temperature to 40°C, hold time, 1min.

January 2013
87
Table 1. Comparison of Retention Time tR and SRM Analyte Peak Area Ratio between Tyr Derivatives (m/z 224, tR 15.2 and 16.8min) Produced by
CSE and the Respective Authentic O-Acetyl-Tyr and N-Acetyl-Tyr
Peak area ratio of SRM chromatogram
Compound
tR (min)
222>180/224>178
222>205/224>178
Mr 223 (Tyr+42)
O-Acetyl-Tyr
15.2
15.2
0.006
0.0061
0.0023
0.0025
224>178/224>182
222>180/224>182
Mr 223 (Tyr+42)
N-Acetyl-Tyr
16.8
16.8
1.431
1.485
0.095
0.093
HPLC conditions: flow rate, 0.4mL/min; column oven temperature, 27°C; linear gradient analysis system. Mass transition pattern: negative ion mode, m/z 222>180,
222>205; positive ion mode, m/z 224>178, 224>182.
Chemical Reactivity of CSE with Tyr To determine op-
timal reaction time, the fresh CSE solution was gently mixed
with 2mm Tyr solution for 0.017 (1min), 1, 3, 6, 12 and 24h
at 37°C. After the reaction, the solutions were analyzed in the
SRM mode. The peak area in the SRM chromatogram of reac-
tion products at each reaction time was compared with that at
24h.
Synthesis of N-(3-Oxobutyl)-Tyr N-(3-Oxobutyl)-Tyr,
whose IUPAC name is 3-(4′-hydroxyphenyl)-2-(3″-oxobutyl-
amino)propanoic acid, was synthesized by mixing 5.5eq of
Tyr, methyl vinyl ketone and NaHCO3 in 100mL water at
80°C for 4h. The reaction mixture was condensed and then
the title compound was purified and collected by HPLC, under
the following conditions: mobile phase, 5% CH3CN and 95%
water containing 0.1% trifluoroacetic acid; flow rate, 1.4mL/
min; Cosmosil 5C18-AR-II packed column (3.0mm×150 mm).
1H-, 13C-, hetero-nuclear multiple quantum coherence (HMQC)
and hetero-nuclear multiple-bond connectivity (HMBC) NMR
spectra were obtained on an ECP-500 (JEOL) at 500MHz in
Fig. 2. Structures of O-Acetyl Tyrosine (a), N-Acetyl Tyrosine (b) and
CD3OD solution. High-resolution FAB mass spectra were ob-
tained by a JMS-700 (JEOL Ltd., Tokyo, Japan).
N-(3-Oxobutyl)-tyrosine
propanoic Acid] (c)
[3-(4′-Hydroxyphenyl)-2-(3″-oxobutylamino)-
Results and Discussion
Analysis of Compounds Produced by Reaction of Tyr MS chromatograms (Table 1). Ultimately, the compound of tR
with CSE To analyze the compounds produced by reaction 15.2min was identified as O-acetyl-Tyr and that of tR 16.8min
of Tyr with CSE, LC/MS analysis, the MS/MS method (prod- as N-acetyl-Tyr (Figs. 2a,b).
uct ion scan, neutral loss scan) and highly sensitive SRM were
Subsequently, the elemental composition of Mr 251 (Tyr+70)
used. First, we optimized the MS parameters and obtained the was determined using a high-resolution mass spectrometer.
protonated and deprotonated molecular ions of Tyr, [Tyr+H]+, The accurately measured mass was 252.1233 0.00048. The
m/z 182 and [Tyr−H]−, m/z 180 under positive and negative difference between the accurately measured mass and the
ESI conditions. Next, we compared differences in the mass calculated exact mass of C13H18NO4 was −0.00017u, and the
spectra of the CSE solutions before and after reaction with the mass error was 0.67ppm. This indicates that the elementary
Tyr solution (Fig. 1). As a result, three compounds displaying composition (molecular formula) of the compound Tyr+70 is
a constant neutral loss scan (m/z 46u) on mass spectra were C13H18NO4.
detected. The loss of m/z 46u, that is, the decrement from
Time Courses of the Amounts of Compounds Produced
[Tyr+H]+ m/z 182 to m/z 136, is characteristic of Tyr in MS/ by Reaction of Tyr with CSE The time courses of the
MS spectra. We next tried to determine their structural for- product yields of three Tyr derivatives, N-acetyl-Tyr, O-
mulas. The molecular weight (Mr) of the three peaks was pre- acetyl-Tyr and Tyr+70, which were produced by reaction of
sumed to be 223 for retention times (tR) of 15.1 and 16.8min, Tyr with CSE, were examined to determine their respective
and 251 for tR 11.2min from the mass value of the positive optimal reaction times by using LC/MS and SRM methods.
and the negative ion mode mass spectra. The compounds of The mass transition pattern of each product was monitored
Mr 223 (Tyr+42) were estimated to be acetylated forms of by the product ion mass spectra (Table 2). The peak area in
Tyr, because Mr 223 has m/z 42u higher mass values than Tyr the SRM chromatogram of each compound was determined
(Mr 181). N-Acetyl-Tyr and O-acetyl-Tyr were identified by at 0.017 (1min), 1, 3, 6, 12 and 24h after reaction of Tyr with
using their respective authentic standards on the basis of the CSE. The yield of each compound was expressed as the ratio
retention time and product ion spectral pattern of their LC/ against the value at 24h. The peak area of O-acetyl-Tyr of tR

88
Vol. 61, No. 1
Table 2. Mass Transition Patterns of Three Tyr Derivatives Obtained by
Reaction of Tyr with CSE
Tyr+70. Analysis data for compound Tyr+70 synthesized by
reaction of methyl vinyl ketone with Tyr agreed with those
of compound Tyr+70 obtained by reaction of CSE with Tyr.
The structure of compound Tyr+70 was identified as N-
(3-oxobutyl)-Tyr[3-(4′-hydroxyphenyl)-2-(3″-oxobutylamino)-
propanoic acid] (Fig. 2c) by 1H-, 13C- and 2D- (HMQC,
HMBC) NMR spectra. This compound was newly formed
Mass transition pat-
tR (min)
Estimated Mr
tern from [M+H]+ and
[M−H]−
11.2
15.2
16.8
N-(3-Oxobutyl)-Tyr
m/z 252>194, 148
m/z 250>180
m/z 224>178, 180
m/z 222>205
m/z 224>182, 178
m/z 222>180
251
O-Acetyl-Tyr
223
N-Acetyl-Tyr
223
1
in the mixture. The analysis data of H-, 13C-NMR and high-
1
resolution FAB-MS were as follows: H-NMR (CD3OD) δ:
2.17 (3H, s), 2.91 (2H, dd, J=6.4Hz), 3.19 (2H, d, J=6.4Hz),
3.26 (2H, d, J=6.4Hz), 4.18 (H, dd, J=6.4Hz), 6.77 (2H,
d, J=8.7Hz), 7.11 (2H, d, J=8.7Hz). 13C-NMR (CD3OD) δ:
29.6, 35.9, 39.6, 43.1, 63.0, 116.9, 125.6, 131.6, 158.5, 170.9,
and 207.6. HR-FAB-MS m/z ([M+H]+); Calcd for C13H18NO4,
252.1235. Measured: 252.1238.
HPLC conditions: flow rate, 0.4mL/min; column oven temperature, 27°C; linear
gradient analysis system.
15.2min (Mr 223) reached maximum at around 1min after the
These results show that the synthesized N-(3-oxobutyl)-Tyr
reaction and thereafter the value was sustained for at least is obtained by the Michael reaction, not the Schiff base
24h. The production rate of O-acetyl-Tyr of tR 15.2min (Mr reaction of carbonyl compounds with primary amines. Re-
223) was the fastest among the three compounds and that of cently, a similar result has been reported in the literature;
N-acetyl-Tyr of tR 16.7min (Mr 223) was second. Its yield was α,β-unsaturated carbonyl compounds, methyl vinyl ketone
about 50% after 1h and almost 100% after 24h. Tyr+70 (tR and acrolein, react with lysine residues in proteins via their
11.2min, Mr 251) showed the slowest production rate and its unsaturated bond to form Michael addition adducts, and this
yield was about 50% after 3h and almost 100% after 24h. The reaction is more efficient than Schiff base adduction via their
peaks of those compounds were not detected in the CSE solu- carbonyl group.12) On the other hand, it has been shown fluo-
tion itself.
rometrically that crotonaldehyde and acrolein in CSE can form
GC/MS Analysis of Active Ingredients in CSE To carbonyl adducts at lysine, cysteine, and histidine residues in
identify the active ingredients in CSE that can produce the human serum albumin.13) The future challenge is to evaluate
acetylated compounds Tyr+42 and the compound Tyr+70 by whether CSE reacts with such amino acids to produce car-
reacting with Tyr, GC/MS analysis was used. With regard bonyl adducts.
to compounds Tyr+42, it was confirmed that the acetic acid
contained in CSE does not react with Tyr at all, but the acetic
Conclusion
anhydride easily reacts with Tyr in PBS solution to produce
The present study demonstrated that CSE can easily react
the acetylated Tyr derivatives, N- and O-acetyl-Tyr. This result with Tyr at 37°C to produce N-acetyl-Tyr, O-acetyl-Tyr and
indicates that acetic anhydride is present in the fresh CSE so- N-(3-oxobutyl)-Tyr. It is noteworthy that the O-acetylation re-
lution and gradually hydrolyzes to acetic acid with time. Thus, action of Tyr proceeded extremely quickly. We also elucidated
analytical data for the authentic standard of acetic acid were the presence of active ingredients in CSE, i.e., methyl vinyl
compared with those of acetic anhydride by GC/MS. Peaks of ketone and acetic anhydride, although the latter compound
acetic acid and acetic anhydride were observed at tR 23min could not be definitely determined because the spectral data
on the TIC chromatograms and also appeared with the same of acetic acid and acetic anhydride are very similar. It is cer-
base peak of m/z 43 in both mass spectra. From these results, tain, however, that acetic anhydride is present in CSE. We
we can surmise that acetic anhydride is easily decomposed also clarified that Michael addition adducts are formed by the
to acetic acid by hydrolysis during passage through the GC reaction of Tyr with methyl vinyl ketone in CSE. Our find-
column.
ings, although preliminary, offer a clue to further elucidating
Next, we tried to identify the active ingredients in CSE that the health risks of smoking. However, it is not certain whether
can produce compound Tyr+70. Two compounds were detect- CSE can react with Tyr residues in peptides and proteins as
ed at tR 6.3min and 9.5min from the mass chromatogram of with Tyr itself. Further studies are needed to assess what
m/z 70. From library search of their mass spectra, we identi- reaction products of CSE with peptides containing Tyr are
fied the peak at tR 9.5min as crotonaldehyde and the peak at tR produced.
6.3min as methyl vinyl ketone. These retention times were the
same as those of their respective authentic samples. From the
Acknowledgments We would like to thank Ms. S. Shirota
results of quantification in the SIM mode, the concentrations and Mr. H. Fukushima (JEOL Ltd.) for their useful sugges-
in the CSE solution of crotonaldehyde and methyl vinyl ketone tions and technical support for the GC-MS analysis, and Dr.
were approximately 450µm and 550µm, respectively.
A. Yamamoto (Osaka City Institute of Public Health and
Identification of Compound Tyr+70 To clarify the Environmental Science) for his useful advice and analytical
structure of compound Tyr+70, crotonaldehyde and methyl support in identifying the new compound. This study was
vinyl ketone, components assumed to be present in CSE, were supported in part by a Grant from the Smoking Research
reacted with Tyr at 37°C. The peak of compound Tyr+70 ap- Foundation, Japan.
peared on the SRM spectra of the mixed solution of Tyr and
methyl vinyl ketone, but was hardly detected on that of the
mixed solution of Tyr and crotonaldehyde. This shows that
methyl vinyl ketone reacts with Tyr to produce compound
References
1) Shah P. K., Helfant R. H., Chest, 94, 449–452 (1988).
2) Hoffmann D., Hoffmann I., El-Bayoumy K., Chem. Res. Toxicol.,

January 2013
14, 767–790 (2001).
89
(2006).
3) Gould N. S., Min E., Gauthier S., Martin R. J., Day B. J., Respir.
Res., 12, 133 (2011).
8) Guzik K., Skret J., Smagur J., Bzowska M., Gajkowska B., Scott D.
A., Potempa J. S., Cell Death Dis., 2, e131 (2011).
4) Kunitomo M., Yamaguchi Y., Kagota S., Yoshikawa N., Nakamura
K., Shinozuka K., J. Pharmacol. Sci., 110, 354–361 (2009).
5) Hoshino S., Yoshida M., Inoue K., Yano Y., Yanagita M., Mawatari
H., Yamane H., Kijima T., Kumagai T., Osaki T., Tachiba I., Kawa-
se I., Biochem. Biophys. Res. Commun., 329, 58–63 (2005).
6) Baglole C. J., Bushinsky S. M., Garcia T. M., Kode A., Rahman I.,
9) Miyake T., Shibamoto T., J. Chromatogr. A, 693, 376–381 (1995).
10) Fujioka K., Shibamoto T., Environ. Toxicol., 21, 47–54 (2006).
11) Yamaguchi Y., Matsuno S., Kagota S., Haginaka J., Kunitomo M.,
Atherosclerosis, 172, 259–265 (2004).
12) Kaminskas L. M., Pyke S. M., Burcham P. C., Chem. Res. Toxicol.,
18, 1627–1633 (2005).
Sime P. J., Phipps R. P., Am. J. Physiol. Lung Cell. Mol. Physiol., 13) Colombo G., Aldini G., Orioli M., Giustarini D., Gornati R., Rossi
291, L19–L29 (2006).
7) Ramage L., Jones A. C., Whelan C. J., J. Inflamm. (Lond.), 3, 3
R., Colombo R., Carini M., Milzani A., Dalle-Donne I., Antioxid.
Redox Signal., 12, 349–364 (2010).