Pyrazine formation from serine and threonine

Article abstract of DOI:10.1021/jf9813687

The formation of pyrazines from L-serine and L-threonine has been studied. L-Serine and L-threonine, either alone or combined, were heated at 120 °C as low temperature for 4 h or at 300 °C as high temperature for 7 min. The pyrazines formed from each reaction were identified by GC/MS, and the yields (to the amino acid used, as parts per million) were determined by GC/FID. It was found that pyrazine, methylpyrazine, ethylpyrazine, 2-ethyl-6- methylpyrazine, and 2,6-diethylpyrazine were formed from serine, whereas 2,5- dimethylpyrazine, 2,6-dimethylpyrazine, trimethylpyrazine, 2-ethyl-3,6- dimethylpyrazine, and 2-ethyl-3,5-dimethylpyrazine were formed from threonine. Mechanistically, it is proposed that the thermal degradation of serine or threonine is composed of various complex reactions. Among these reactions, decarbonylation followed by dehydration is the main pathway to generate the α-aminocarbonyl intermediates leading to the formation of the main product, such as pyrazine from serine or 2,5-dimethylpyrazine from threonine. Also, deamination after decarbonylation generates more reactive intermediates, α-hydroxycarbonyls. Furthermore, aldol condensation of these reactive intermediates provides α-dicarbonyls. Subsequently, these α- dicarbonyls react with the remaining serine or threonine by Strecker degradation to form additional α-aminocarbonyl intermediates, which then form additional pyrazines. In addition, decarboxylation and retroaldol reaction may also involve the generation of the intermediates.

Full text of DOI:10.1021/jf9813687

4332
J. Agric. Food Chem. 1999, 47, 4332−4335
P yr a zin e F or m a tion fr om Ser in e a n d Th r eon in e
Chi-Kuen Shu†
Bowman Gray Technical Center, R. J . Reynolds Tobacco Company, Winston-Salem, North Carolina 27105
The formation of pyrazines from L-serine and L-threonine has been studied. L-Serine and L-threonine,
either alone or combined, were heated at 120 °C as low temperature for 4 h or at 300 °C as high
temperature for 7 min. The pyrazines formed from each reaction were identified by GC/MS, and
the yields (to the amino acid used, as parts per million) were determined by GC/FID. It was found
that pyrazine, methylpyrazine, ethylpyrazine, 2-ethyl-6-methylpyrazine, and 2,6-diethylpyrazine
were formed from serine, whereas 2,5-dimethylpyrazine, 2,6-dimethylpyrazine, trimethylpyrazine,
2-ethyl-3,6-dimethylpyrazine, and 2-ethyl-3,5-dimethylpyrazine were formed from threonine.
Mechanistically, it is proposed that the thermal degradation of serine or threonine is composed of
various complex reactions. Among these reactions, decarbonylation followed by dehydration is the
main pathway to generate the R-aminocarbonyl intermediates leading to the formation of the main
product, such as pyrazine from serine or 2,5-dimethylpyrazine from threonine. Also, deamination
after decarbonylation generates more reactive intermediates, R-hydroxycarbonyls. Furthermore, aldol
condensation of these reactive intermediates provides R-dicarbonyls. Subsequently, these R-dicar-
bonyls react with the remaining serine or threonine by Strecker degradation to form additional
R-aminocarbonyl intermediates, which then form additional pyrazines. In addition, decarboxylation
and retroaldol reaction may also involve the generation of the intermediates.
Keyw or d s: Serine; threonine; pyrazine formation; decarbonylation; decarboxylation; deamination;
R-aminocarbonyls; Strecker degradation; aldol condensation
INTRODUCTION
The objective of the present study was to determine
the major alkylpyrazines formed qualitatively and
quantitatively from serine and threonine as well as from
a combination of serine and threonine under different
temperatures. On the basis of the results obtained, a
mechanism for the formation of alkylpyrazines from
serine and threonine is proposed.
Alkylpyrazines are generally considered as trace
important flavor components in foods (Maga, 1992). The
formation of alkylpyrazines has been widely investi-
gated and reviewed (Vernin and Parkanyi, 1982; Heath
and Reineccius, 1986; Ohloff et al., 1985; Vernin and
Metzger, 1981). Categorically, there are two widely
accepted mechanisms for the formation of pyrazines:
the Strecker degradation and the ammonia/acyloin
reaction. The Strecker degradation (Schonberg and
Moubacher, 1952; Rizzi, 1972) is involved with R-amino
acids and the reductones (R-dicarbonyls), which are
derived either from the Maillard reaction or from
caramelization of carbohydrates (Hodge, 1967). During
the Strecker degradation, the R-dicarbonyls are con-
verted into R-aminocarbonyls, which, in turn, condense
to form alkylpyrazines. The ammonia/acyloin reaction
is involved with ammonia and R-hydroxycarbonyls,
which are also derived from caramelization of carbohy-
drates (Hodge, 1967). During this reaction, R-hydroxy-
carbonyls are converted into R-aminocarbonyls and then
to alkylpyrazines. This reaction takes place very readily
even at room temperature (Rizzi, 1988; Shu and
Lawrence, 1995). Basically, these two mechanisms for
alkylpyrazine formation require a carbohydrate source
to provide the carbohydrate degradation products, either
as R-dicarbonyls or as acyloins. However, without a
carbohydrate source, alkylpyrazines were also found
from the pyrolysis of hydroxyamino acids (Kato et al.,
1970; Wang and Odell, 1973). These authors concluded
that R-dicarbonyls are not required to form such alkyl-
pyrazines, but they did not explain how such alkyl-
pyrazines were formed.
EXPERIMENTAL PROCEDURES
Ma ter ia ls. L-Serine and L-threonine were purchased from
Ajinomoto Co. (Tokyo, J apan), and n-hexadecane was pur-
chased from Aldrich Chemical Co. (Milwaukee, WI).
P r ep a r a tion of th e Rea ction Mixtu r es. In an enclosed
reaction vessel (Parr Instrument Co., Moline, IL), 300 mg of
reactant and 36 µL of water (as 12% moisture level) were
heated in an oven at 120 °C for 4 h or at 300 °C for 7 min. The
reactants included L-serine, L-threonine, and a combination
of 50% L-serine and 50% L-threonine (w/w). Each reaction
mixture obtained was cooled to room temperature and ex-
tracted with methylene chloride (5 mL × 4). The methylene
chloride extracts, which were combined from the reaction
performed at 300 °C for 7 min, were directly analyzed by GC/
MS, while the methylene chloride extracts, which were com-
bined from the reaction performed at 120 °C for 4 h, were
further concentrated under a stream of nitrogen to 0.5 mL
prior to GC/MS analysis.
GC/MS An a lysis. Each extract as prepared above was
analyzed by GC/MS on a DB-Wax fused silica column (60 m
× 0.32 mm, 0.15 µm film thickness) with a mass selective
detector (EI; 70 eV). The oven temperature was programmed
from 50 to 200 °C at 6 °C/min.
Qu a n tita tion of th e Alk ylp yr a zin es F or m ed in th e
Rea ction Mixtu r e. Each extract prepared above was also
analyzed under the same chromatographic conditions as
described above except that a flame ionization detector (FID)
was used. n-Hexadecane was added to each extract as an
internal standard for the quantitation of each alkylpyazine;
† Fax (336) 741-6343.
10.1021/jf9813687 CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/04/1999

Pyrazines from Serine and Threonine
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4333
Ta ble 1. P yr a zin es Id en tified a n d th e Qu a n tita tive Da ta ,
a t P a r ts p er Million
120 °C/4 h
300 °C/7 min
from from from from from from
Ser Thr Ser/Thr Ser Thr Ser/Thr
pyrazine identified
pyrazine
methyl-
97.6
6.1
32.3 1477
354
880
898
501
348
1077
281
705
247
32.0
9.9
245
2,5-dimethyl-
2,6-dimethyl-
ethyl-
2-ethyl-6-methyl-
2-ethyl-5-methyl-
trimethyl-
2,6-diethyl-
2-ethyl-3,6-dimethyl-
2-ethyl-3,5-dimethyl-
11.4
1100
390
3.1
8.2 1025
21.3
12.6
131
10.0
0.4
7.4
271
391
11.4
632 2291
83 458
total
125.0 21.8 116.9 3269 2476 8040
it was assumed that the FID response factor of n-hexadecane
was the same as those of the alkylpyrazines. The yield of each
alkylpyrazine was reported as parts per million of serine or/
and threonine used (ppm).
RESULTS AND DISCUSSION
Table 1 shows the pyrazines identified from all of the
reactions along with the quantitative data. In general,
heating at high temperature (300 °C/7 min) generated
greater amounts of alkylpyrazines than heating at low
temperature (120 °C/4 h). Also, under each condition,
it is obvious that L-serine generates a higher yield of
pyrazines than L-threonine, probably due to the melting
point of L-serine (222 °C), which is lower than the
melting point of L-threonine (256 °C). This melting point
effect on the yield appears to be more significant for
heating at low temperature than at high temperature.
When L-serine is heated to 120 °C, pyrazine, eth-
ylpyrazine (major), and methylpyrazine (minor) were
found, whereas when it is heated to 300 °C, pyrazine,
ethylpyrazine, 2,6-diethylpyrazine (major), methylpyra-
zine, and 2-ethyl-6-methylpyrazine (minor) were found.
Taking into account these results and relevant litera-
ture information, it is suggested that during thermal
degradation of serine, alkylpyrazines are formed as
shown in Figure 1.
Figure 1A shows the pathway to form the parent
pyrazine and the reactive intermediates. Decarbonyla-
tion of serine followed by dehydration generates a
precursor, R-aminoacetaldehyde (1), which is dimerized
to form pyrazine. Meanwhile, decarbonylation followed
by deamination forms R-hydroxyacetaldehyde (glycolal-
dehyde), which can also be converted to 1. In addition,
decarboxylation followed by deamination can lead to the
formation of acetaldehyde. These reactive intermediates
participate further in the reactions as described later.
Decarbonylation of serine is essential to describe this
mechanism. Basically, it is analogous to the acid-
catalyzed decarbonylation of R-hydroxy acid, in which
the primary products include CO and water. Carbon
monoxide and other low molecular weight compounds
were already found from the pyrolysis of amino acids
(Lien and Nawar, 1974).
Figure 1B shows how the other major alkylpyrazines,
ethylpyrazine and 2,6-diethylpyrazine, may be formed.
Aldol condensation of the reactive intermediates, gly-
colaldehyde and acetaldehyde, followed by dehydration,
generates 1,2-butanedione, which can participate in the
Strecker degradation with the remaining serine to form
F igu r e 1. Mechanism proposed for the formation of pyrazines
from serine: (A) formation of parent pyrazine; (B) formation
of ethylpyrazine and 2,6-diethylpyrazine; (C) formation of
2-ethyl-6-methylpyrazine and methylpyrazine.

4334 J. Agric. Food Chem., Vol. 47, No. 10, 1999
Shu
precursors 1-amino-2-butanone (2) and 2-aminobutanal
(3). As a result, the combination of 1 and 2 or 1 and 3
can form ethylpyrazine, whereas the combination of 2
and 3 can form 2,6-diethylpyrazine. Because the amount
of parent pyrazine found is greater than that of other
pyrazines (Table 1), it appears that decarbonylation
followed by dehydration is the predominant pathway,
compared with all other pathways. Similarly, Figure 1C
shows the possible formation of precursors 1-amino-2-
propanone (4) and 2-aminopropanal (5) from glycoalde-
hyde via aldol condensation and retroaldol reaction.
Therefore, the combination of 2 and 5 or 3 and 4 can
generate 2-ethyl-6-methylpyrazine, whereas the com-
bination of 1 and 4 or 1 and 5 can generate methylpyra-
zine.
When threonine is heated to 120 °C, 2,5-dimeth-
ylpyrazine, trimethylpyrazine (major), and 2-ethyl-3,6-
dimethylpyrazine (minor) were found, whereas when it
is heated to 300 °C, 2,5-dimethylpyrazine, 2,6-dimeth-
ylpyrazine, trimethylpyrazine, 2-ethyl-3,6-dimethylpyra-
zine (major), and 2-ethyl-3,5-dimethylpyrazine (minor)
were found. On the basis of these results and the
literature information, a formation mechanism of the
alkylpyrazines is proposed in Figure 2. In general, these
formation pathways from threonine are similar to those
described earlier for serine.
Figure 2A shows the main pathway through decar-
bonylation and dehydration to form precursor 4, which
is dimerized to yield 2,5-dimethylpyrazine. Also, decar-
bonylation followed by deamination generates 1-hy-
droxy-2-propanone and 2-hydroxyypropanal, the reac-
tive intermediates, which may react back with ammonia
to form precursors 5 and 4. Furthermore, self-combina-
tion of 4 or 5 can form 2,5-dimethylpyrazine, whereas
cross-combination of 4 and 5 can form 2,6-dimeth-
ylpyrazine. Because decarbonylation/dehydration is the
main pathway, 2,5-dimethylpyrazine is preferentially
formed over 2,6-dimethylpyrazine.
Figure 2B shows that the aldol condensation of
1-hydroxy-2-propanone followed by retroaldol reaction
and Strecker degradation generates precursors 3-amino-
2-pentanone (6) and 2-amino-3-pentanone (7). As a
result, the combination of 6 and 4 can form 2-ethyl-3,6-
dimethylpyrazine, whereas the combination of 7 and 4
can form 2-ethyl-3,5-dimethylpyrazine. Similarly, Fig-
ure 2C shows the pathway of the reaction of 1-hydroxy-
2-propanone and formaldehyde to generate the precur-
sor 3-amino-2-butanone (8). The combination of 8 and
4 can form trimethylpyrazine.
It should be noted that in addition to the pyrazines
listed in Table 1, there are several other pyrazines
present in trace amounts. For instance, trace amounts
of 2,5-dimethylpyrazine and 2,6-dimethylpyrazine were
also found from the thermal degradation of serine,
probably due to reaction of 4 and 5 (Figure 1C). Self-
combination of 4 or 5 forms 2,5-dimethylpyrazine,
whereas cross-combination of 4 and 5 forms 2,6-di-
methylpyrazine as in the case of threonine (Figure 2A).
However, during this study, the detailed information
of the formation of trace level pyrazines was not
considered.
It is interesting to examine the pyrazines formed from
the reactions of a mixture of 50% serine and 50%
threonine (Table 1). It has been found that all of the
pyrazines formed from the reaction of serine or threo-
nine are formed from the reaction of serine/threonine,
but the major pyrazines formed from the reaction of
F igu r e 2. Mechanism proposed for the formation of pyrazines
from threonine: (A) formation of 2,5-dimethylpyrazine and 2,6-
dimethylpyrazine; (B) formation of 2-ethyl-3,6-dimethylpyra-
zine and 2-ethyl-3,5dimethylpyrazine; (C) formation of tri-
methylpyrazine.

Pyrazines from Serine and Threonine
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4335
LITERATURE CITED
serine/threonine are different from those formed by the
reaction of serine or threonine alone. For example,
heating at low temperature, methylpyrazine is one of
the two most abundant pyrazines formed from the
reaction of serine/threonine. In contrast, methylpyrazine
is not the major pyrazine formed from the reaction of
either serine or threonine. The difference may involve
the precursors of methylpyrazine, R-aminoacetaldehyde
(1) and 1-amino-2-propanone (4), which are derived from
serine and threonine, respectively It is also observed
that with heating at high temperature, 2-ethyl-3,6-
dimethylpyrazine and 2-ethyl-6-methylpyrazine are the
most abundant pyrazines formed from the reaction of
serine/threonine, but they neither is the most abundant
pyrazine from the reaction of serine or threonine. This
may also involve their precursors, R-aminocarbonyls.
Heath, H. B.; Reineccius, G. Flavor Chemistry and Technology;
AVI Publishing: Westport, CT, 1986; pp 71-111.
Hodge, J . E. Origin of flavor in foods, nonenzymatic browning
reactions. In Chemistry and Physiology of Flavors; Schultz,
H. W., Day, E. A., Libbey, L. M., Eds.; AVI Publishing:
Westport, CT, 1967; pp 465-491.
Kato, S.; Kurata, T.; Ishitsuka, R.; Fujimaki, M. Pyrolysis of
â-hydroxy amino acids, especially L-serine. Agric. Biol.
Chem. 1970, 34, 1826-1832.
Lien, Y. C.; Nawar, W. W. Thermal decomposition of some
amino acids. J . Food Sci. 1974, 39, 911-913.
Maga, J . A. Pyrazine update. Food Rev. Int. 1992, 8 (4), 479-
558.
Ohloff, G.; Flament, I.; Pickenhagen, W. Flavor Chemistry. In
Food Reviews International; Teranishi, R., Hornstein, I.,
Eds.; Dekker: New York, 1985; Vol. 1, pp 99-148.
Rizzi, G. P. A mechanistic study of alkylpyrazine formation
in model systems. J . Agric. Food Chem. 1972, 20, 1081-
1085.
SUMMARY OF PROPOSED MECHANISM
The results obtained from this study indicate that the
thermal degradation of serine or threonine, either at low
temperature (120 °C) or at higher temperature (300°),
is composed of various complex reactions including
decarbonylation, dehydration, deamination, amination,
aldol condensation, retroaldol reaction, decarboxylation,
and Strecker degradation. Of these reactions, decarbo-
nylation followed by dehydration is the main pathway
to generate the R-aminocarbonyl intermediates leading
to the formation of the main product, such as pyrazine
from serine or 2,5-dimethylpyrazine from threonine. In
addition, decarbonylation and decarboxylation followed
by deamination generate reactive intermediates such
as R-hydroxycarbonyls and aldehydes. Furthermore,
aldol condensation and retroaldol reaction of these
reactive intermediates lead to the formation of R-dicar-
bonyls. Subsequently, these R-dicarbonyls react with the
remaining serine or threonine by Strecker degradation
to form additional R-aminocarbonyl intermediates and
additional pyrazines. When serine and threonine are
combined in a mixture for the thermal reaction, the
pyrazine profile obtained quantitatively depends on the
R-aminocarbonyl intermediates formed.
Rizzi, G. P. Formation of pyrazines from acyloin precursors
under mild conditions. J . Agric. Food Chem. 1988, 36, 349-
352.
Schonberg, A.; Moubacher, R. The Strecker degradation of
R-amino acids. Chem. Rev. 1952, 50, 261-277.
Shu, C.-K.; Lawrence, B. M. Formation of 2-(1-hydroxyalkyl)-
3-oxazolines from the reaction of acyloins and ammonia
precursors under mild conditions. J . Agric. Food Chem.
1995, 43, 2922-2924.
Vernin, G.; Metzger, J . La Chimie des Aromes: Les Hetero-
cycles. Bull. Soc. Chim. Belg. 1981, 90, 553-587.
Vernin, G.; Parkanyi, C. Occurrence and formation of hetero-
cyclic compounds. In The Chemistry of Heterocyclic Flavour-
ing and Aroma Compounds; Vernin, G., Ed.; Ellis Hor-
wood: New York, 1982; pp 192-198.
Wang, P. S.; Odell, G. V. Formation of pyrazines from thermal
treatment of some amino-hydroxy compounds. J . Agric. Food
Chem. 1973, 21, 868-870.
Received for review December 18, 1998. Revised manuscript
received J uly 16, 1999. Accepted J uly 16, 1999.
J F9813687