Effects of sulphur nutrition during potato cultivation on the formation of acrylamide and aroma compounds during cooking

Article abstract of DOI:10.1016/j.foodchem.2010.03.049

Lack of sulphur nutrition during potato cultivation has been shown to have profound effects on tuber composition, affecting in particular the concentrations of free asparagine, other amino acids and sugars. This is important because free asparagine and sugars react at high temperatures to form acrylamide, a suspect carcinogen. Free amino acids and sugars also form a variety of other compounds associated with colour and flavour. In this study the volatile aroma compounds formed in potato flour heated at 180 °C for 20 min were compared for three varieties of potato grown, with and without sulphur fertiliser. Approximately 50 compounds were quantified in the headspace extracts of the heated flour, of which over 40 were affected by sulphur fertilisation and/or variety. Many of the 41 compounds found at higher concentrations in the sulphur-deficient flour were Strecker aldehydes and compounds formed from their condensation, whereas only one compound, benzaldehyde, behaved in the same way as did acrylamide and was found at higher concentrations in the sulphur-sufficient flour. The reasons for these effects are discussed.

Full text of DOI:10.1016/j.foodchem.2010.03.049

Food Chemistry 122 (2010) 753–760
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Effects of sulphur nutrition during potato cultivation on the formation
of acrylamide and aroma compounds during cooking
a
b
b
b
a
J.S. Elmore a, , A.T. Dodson , N. Muttucumaru , N.G. Halford , M.A.J. Parry , D.S. Mottram
*
a Department of Food Biosciences, University of Reading, Whiteknights, Reading RG6 6AP, UK
b Plant Science Department, Rothamsted Research, Harpenden, Hertfordshire AL5 2JQ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 15 July 2009
Received in revised form 5 January 2010
Accepted 10 March 2010
Lack of sulphur nutrition during potato cultivation has been shown to have profound effects on tuber com-
position, affecting in particular the concentrations of free asparagine, other amino acids and sugars. This is
important because free asparagine and sugars react at high temperatures to form acrylamide, a suspect
carcinogen. Free amino acids and sugars also form a variety of other compounds associated with colour
and flavour. In this study the volatile aroma compounds formed in potato flour heated at 180 °C for
20 min were compared for three varieties of potato grown, with and without sulphur fertiliser. Approxi-
mately 50 compounds were quantified in the headspace extracts of the heated flour, of which over 40 were
affected by sulphur fertilisation and/or variety. Many of the 41 compounds found at higher concentrations
in the sulphur-deficient flour were Strecker aldehydes and compounds formed from their condensation,
whereas only one compound, benzaldehyde, behaved in the same way as did acrylamide and was found
at higher concentrations in the sulphur-sufficient flour. The reasons for these effects are discussed.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Potato
Acrylamide
Sulphate fertiliser
Amino acids
Sugars
Volatile compounds
1. Introduction
phur deprivation and to examine the relationship between acryl-
amide formation and aroma formation. Heated flour was studied
Several recent papers have highlighted the effects of sulphur
deficiency, during cultivation, on the chemical composition of
wheat grain and potato tubers, particularly in relation to the for-
mation of the suspect carcinogen acrylamide from its precursors,
free asparagine and sugars (Curtis et al., 2009; Elmore et al.,
2007; Granvogl, Wieser, Koehler, Von Tucher, & Schieberle, 2007;
Muttucumaru et al., 2006; Weber et al., 2008). The changes in free
amino acid and sugar levels in wheat grain caused by sulphur
deprivation during cultivation also resulted in large changes in
the volatile aroma compound composition of cooked flour (Elmore,
Parker, Halford, Muttucumaru, & Mottram, 2008). Sulphur depriva-
tion altered the ratio of sugars to free amino acids; in sulphur-de-
prived (Sꢀ) wheat grain, free amino acids were present in excess
compared with sugars whereas, in grain from wheat grown under
normal conditions (S+), sugars were in excess. As a result, aroma
compounds that form from sugar degradation, such as furfural
and 2,3-pentanedione, were present at higher levels in flour from
S+ wheat, while compounds derived from amino acids, such as
Strecker aldehydes (for example, 3-methylbutanal) and their con-
densation products (for example, 2-isopropyl-(E)-2-butenal) were
at higher concentrations in flour from Sꢀ wheat.
because sample sizes were small, precluding the cooking of French
fries, for example, and because the use of flour removed any vari-
ation which may have been caused by reductions in tuber size
brought about by sulphur deprivation. The three varieties of potato
that were studied were King Edward and Maris Piper, two popular
varieties available all year round in the United Kingdom, and Prai-
rie, an experimental variety.
The potato material used in this experiment had been studied
previously, with regard to the effect of sulphur deprivation on
acrylamide formation in heated flour (Elmore et al., 2007). In that
study, the sugar and free amino acid compositions of the potatoes
were measured, as well as the acrylamide concentrations in flour
cooked at 160 °C. There was a general rise in the levels of free
amino acids in potato tubers in response to sulphate deprivation
during cultivation, and there was a clear difference between the
varieties with respect to the predominant amino acid that accumu-
lated. Sulphur-deprived tubers of the King Edward and Maris Piper
varieties were particularly high in glutamine, while asparagine lev-
els fell. In contrast, while the tubers of the Prairie variety also accu-
mulated high levels of glutamine in response to sulphate
deprivation, they also showed a large increase in asparagine levels.
With regard to the formation of acrylamide, the effect of sulphur
deprivation on asparagine as a proportion of the free amino acid
pool (as opposed to asparagine concentration per se) was the key
parameter, and sulphur deprivation led to less acrylamide forma-
tion in heated tuber flour from all three varieties.
In this paper we have heated potato flour at 180 °C for 20 min in
order to determine how its volatile composition is affected by sul-
* Corresponding author. Fax: +44 118 310 0080.
E-mail address: j.s.elmore@reading.ac.uk (J.S. Elmore).
0308-8146/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2010.03.049

754
J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
All analyses were performed on a Perkin–Elmer Clarus 500 GC–
2. Materials and methods
MS system (Perkin–Elmer Life and Analytical Sciences Inc., Wal-
tham, MA) running Turbomass Software (Version 4.5; Perkin–El-
mer). A Perkin–Elmer Turbomatrix automated thermal desorber
(ATD) was used to transfer the volatiles from the Tenax trap onto
2.1. Chemicals
1,2-Dichlorobenzene (99%), 2-methylpropanal (>99%), acetalde-
hyde (>99.5%), sulphuryl chloride (97%), ethyl acetoacetate (99%),
methanethiol sodium salt (sodium thiomethoxide, 95%), 3-bro-
mo-2-butanone (97%), tetrabutylammonium hydrogensulphate
(97%), anhydrous sodium acetate (99.5%), 2-methylpropionamide
(99%), 2-methylbutyryl chloride (97%), 1-hydroxy-2-butanone
(Rare Chemical Library Substance), formaldehyde (37% in water),
the front of
a J&W DB-5 fused silica capillary column
(60 m ꢁ 0.32 mm i.d., 1 m film thickness; Agilent Technologies,
Santa Clara, CA), via a cold trap. In the ATD, the Tenax trap was
heated at 300 °C for 10 min; volatiles were desorbed onto the cold
trap, which was held at ꢀ30 °C. After desorption, the cold trap was
then heated to 300 °C at 40 °C per second to release the volatile
material onto the GC column. After desorption, the temperature
of the oven was held at 40 °C for a further 2 min, before being
raised at 4 °C/min to 250 °C. Helium, at 145 kPa, was used as the
carrier gas. A series of n-alkanes (C5–C25) in diethyl ether was ana-
lysed, under the same conditions, to obtain linear retention index
(LRI) values for the cooked potato aroma components.
The mass spectrometer operated in electron impact mode with
an electron energy of 70 eV. The mass spectrometer scanned from
m/z 29 to m/z 350 at 2.5 scans/s. Compounds were identified by
comparing their mass spectra with those contained in the NIST/
EPA/NIH Mass Spectral Database or in previously published litera-
ture. In all cases, identities were confirmed by comparison of mass
spectra and linear retention index (LRI) values with those of
authentic standards, either purchased or synthesised. Approximate
quantities of the volatiles in the headspace were estimated by
comparison of their peak areas with that of the 1,2-dichloroben-
zene internal standard, obtained from the total ion chromato-
grams, using a response factor of 1.
a-ionone (90%), 1-hydroxy-2-propanone (90%), ethylenediamine
(>99%), manganese (IV) oxide (85%), methyl 2-pyrrolecarboxylate
(97%), gallium (III) chloride (anhydrous, beads, ꢀ10 mesh, 99.99%
trace metals basis), triethylene glycol (99%) and N-bromosuccini-
mide (99%) were all purchased from Sigma–Aldrich Corporation
(St Louis, MO). 3-Methylbutanal (98%), isobutyllithium (1.6 M
solution in heptane), N,N-diethylaniline (99%), anhydrous sodium
sulphate (99%) and dibenzoyl peroxide (75%) were all purchased
from Acros Organics (Loughborough, UK). 2-Methylbutanal (95%),
2-chloropropane (99%) and trimethylpyrazine (99%) were pur-
chased from Alfa Aesar (Heysham, UK). Ammonium acetate
(97%), sodium thiosulphate (99%), anhydrous disodium hydrogen
orthophosphate (>99%), sodium dihydrogen orthophosphate
(>98%), potassium hydroxide pellets (>85%), iodine (>99.5%), meth-
anol, heptane, dichloromethane, diethyl ether, tetrahydrofuran and
petroleum ether (all AnalaR grade) were purchased from Fisher
Scientific (Loughborough, UK). Ammonia gas was supplied by
S.I.P Analytical, Ltd. (Sandwich, UK).
An aroma extract of heated Sꢀ Maris Piper potato flour was also
analysed, under the same GC–MS conditions, using a VF-WAXms
2.2. Growing of potato plants
column (60 m ꢁ 0.32 mm ꢁ 1.00
lm; Varian, Middleburg, The
Netherlands) instead of a DB-5 column, in order to provide addi-
The procedures for growing the potato plants and preparing po-
tato flour were first described by Elmore et al. (2007). Briefly, three
varieties of potato, King Edward, Prairie and Maris Piper, were
grown in a glasshouse. Ten plants were grown of each variety; five
were watered with medium containing sulphate ions (1 mM
MgSO4; these were designated S+), while five were watered with
the same medium lacking the sulphate (these were designated
Sꢀ). Within 24 h of harvesting, the tubers were peeled, sliced
and freeze-dried. The dried samples were powdered, vacuum-
packed and stored at ꢀ18 °C prior to analysis.
tional retention time data for compound identification.
2.5. Analysis of acrylamide
Acrylamide was extracted from these samples with 25% aque-
ous methanol and converted to the dibromo- derivative, prior to
analysis by GC–MS, using the method of Castle, Campos, and Gil-
bert (1991), with the modifications described by Elmore, Koutsidis,
Dodson, Mottram, and Wedzicha (2005). 13C3-Acrylamide was
used as the internal standard.
The brominated extracts (2 ll) were injected onto an Agilent
2.3. Preparation of cooked samples for flavour and acrylamide analysis
5975 GC–MS system in pulsed splitless mode at 250 °C, the splitter
opening after 0.5 min. The helium carrier gas pressure was 145 kPa
in pulsed mode, falling to 66.2 kPa for the rest of the run. A DB-17
Tuber flour samples (0.50 g) in unsealed glass ampoules (1 ml
capacity) were heated for 20 min, at 180 °C. Copper wire was tied
around the necks of the filled ampoules, and the ampoules were
heated by suspending by the wire from the ceiling of a GC oven.
Six replicates were prepared for each flour sample; three were ana-
lysed for their acrylamide content and three for their volatile
composition.
MS capillary column was used (30 m ꢁ 0.25 mm i.d., 0.15
lm film
thickness; Agilent). The oven temperature was 85 °C for 1 min, ris-
ing at 8 °C/min to 200 °C, then 30 °C/min to 280 °C for 10 min. The
transfer line was held at 280 °C and the ion source at 180 °C. The
mass spectrometer was operated in electron impact mode with se-
lected ion monitoring. Two ions were used to monitor brominated
13C3-acrylamide (m/z 153 and 155) and another two ions were
used for brominated acrylamide (m/z 150 and 152). The ion m/z
155 was used to quantify brominated 13C3-acrylamide and the
ion m/z 150 was used to quantify brominated acrylamide. Each
sample was prepared and analysed in triplicate.
2.4. Analysis of volatiles
Aroma isolates were collected on Tenax TA. The cooked flour
was transferred to a 250 ml conical flask with a screw-thread neck
and 100 ml of HPLC-grade water were added. Nitrogen, at
40 ml minꢀ1, flowed over the cooked flour sample, which was held
2.6. Preparation of compounds to confirm identities
at 35 °C or 45 min, sweeping volatiles onto
a glass trap
(3.5 in. ꢁ 0.25 in.) containing 85 mg Tenax TA (Supelco, Bellefonte,
PA). A standard (100 ng of 1,2-dichlorobenzene in 1 l of methanol)
was added to the trap at the end of the collection, and excess sol-
vent and any water retained on the trap were removed by purging
the trap with nitrogen at 100 ml/min for 10 min.
2-Methyl-2-(methylthio)propanal and 3-methyl-2-(methyl-
thio)butanal were prepared from 2-methylpropanal and 3-methyl-
butanal, respectively. 2-Chloroaldehydes were prepared by
reaction of sulphuryl chloride with the appropriate aldehyde
(Dejaegher & De Kimpe, 2004). These were then heated with

J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
755
methanethiol sodium salt in water (Mathew & Ulmer, 1976), to
give the 2-(methylthio)aldehydes.
propylpyrrole by heating it with potassium hydroxide in
triethylene glycol at 230 °C (Groves, Anderson, & Nagy, 1971).
1,2-Dihydro-1,1,6-trimethylnaphthalene was prepared by con-
6-Methyl-3-hepten-2-one was prepared from 3-methylbutanal
and ethyl acetoacetate, using the method of Kourouli, Kefalas,
Ragoussis, and Ragoussis (2002). Base-catalysed aldol condensa-
tion was used to prepare (E)-4-methyl-2-hexenal from 2-methyl-
butanal and acetaldehyde (Elmore et al., 2008).
4,5-Dimethyl-2-isopropyloxazole and 4,5-dimethyl-2-(1-meth-
ylpropyl)oxazole were prepared by heating 3-bromo-2-butanone
with 2-methylpropionamide and 2-methylbutyramide, respec-
tively (Theilig, 1953). 2-Methylbutyramide was made from 2-
methylbutyryl chloride and ammonia, using the method of Art-
man, Balandrin, Moe, Van Wagenen, and Pesyan (2006).
2,5(6)-Dimethyl-3-(2-methylpropyl)pyrazine and 2,5(6)-di-
methyl-3-(3-methylbutyl)pyrazine were prepared by heating
ammonium acetate and sodium acetate in water, and then adding
1-hydroxy-2-propanone and a Strecker aldehyde dropwise to the
refluxing mixture. 2-Methylpropanal was used to prepare 2,5(6)-
dimethyl-3-(2-methylpropyl)pyrazine, and 3-methylbutanal was
used to prepare 2,5(6)-dimethyl-3-(3-methylbutyl)pyrazine (Chen,
1994). The same method was used to prepare 3,5-diethyl-2-meth-
ylpyrazine, using 1-hydroxy-2-butanone and formaldehyde.
2-(2-Methylbutyl)-3-methylpyrazine and 2-(3-methylbutyl)-
3-methylpyrazine were prepared from their respective 5,6-dihy-
dropyrazines by oxidation with manganese dioxide in alkaline
ethanol at room temperature (Akiyama, Enomoto, & Shibamoto,
1978). 2-(2-Methylbutyl)-3-methyl-5,6-dihydropyrazine and 2-
(3-methylbutyl)-3-methyl-5,6-dihydropyrazine were prepared
by reacting ethylenediamine with 5-methyl-2,3-heptanedione
and 6-methyl-2,3-heptanedione, respectively. 5-Methyl-2,3-hep-
tanedione and 6-methyl-2,3-heptanedione were prepared from
the reactions of 1-hydroxy-2-propanone with 2-methylbutanal
and 3-methylbutanal, respectively (Winter, Gautschi, Flament,
& Stoll, 1976).
verting
a-ionone to ionene, using iodine, and then methylating
the ionene using N-bromosuccinimide, benzoyl peroxide and
N,N-diethylaniline (Miginiac, 1990). Isobutyltrimethylpyrazine
was prepared by reacting isobutyllithium with trimethylpyrazine
under nitrogen at 0 °C (Rizzi, 1968; warning! – isobutyllithium is
both extremely moisture-sensitive and extremely air-sensitive).
2.7. Physicochemical data from synthesised compounds
Sixteen compounds were synthesised to confirm their presence
or absence in the heated flour extracts. In all cases, products were
extracted in diethyl ether, dried with anhydrous sodium sulphate
and analysed by GC–MS (HP 5972 MSD) in split mode (20:1), using
a similar temperature programme to that used for the analysis of
the aroma extracts, with the mass spectrometer scanning from
m/z 10 to m/z 300. The capillary columns that were used were of
the same phase types as those used for the aroma extract analysis
but of different dimensions:
(60 m ꢁ 0.25 mm ꢁ 0.25 m; Varian) and a polar HP-Innowax col-
m; Agilent). The mass spectra and
a non-polar VF5-ms column
l
umn (60 m ꢁ 0.25 mm ꢁ 0.25
l
LRIs for the synthesised compounds are listed in Table 1, when
appropriate.
2.8. Statistical analysis
XLStat 2006 (Addinsoft, Paris, France) was used to perform two-
way analysis of variance on the data.
3. Results and discussion
3.1. Effect of sulphur deprivation during potato cultivation on the
precursors of acrylamide and flavour in tuber flour
Methyl 4-isopropylcarboxylate was prepared from 2-chloropro-
pane and methyl 2-pyrrolecarboxylate, using gallium chloride as a
catalyst (warning! – gallium chloride is extremely moisture-sensi-
tive). Methyl 4-isopropylcarboxylate was then converted to 3-iso-
Tables 2–4 have been created from the data presented in the pa-
per of Elmore et al. (2007), in order to understand how sulphur
Table 1
Mass spectra and linear retention indices of synthesised compounds.
a
b
Compound
LRIVF5
855
LRIWax
Mass spectral data, m/z (relative intensity)c
2-Methyl-2-(methylthio)propanal
1238
1291
89, 41(59), 49(37), 39(35), 45(19), 47(15), 29(14), 118(11), 61(11),
59(11), 43(8), 42(7), 46(6), 91(5), 90(5), 75(5)
55, 83(71), 41(61), 39(50), 97(28), 56(23), 53(22), 69(18), 43(16),
(E)-4-Methyl-2-hexenal
920
79(12), 51(11), 84(9), 67(8), 50(8), 42(8), 112(4)
4,5-Dimethyl-2-isopropyloxazole
3-Methyl-2-(methylthio)butanal
975
978
1280
1386
NISTd
55, 103(83), 39(33), 29(27), 41(26), 45(25), 61(24), 132(19),
47(13), 90(11), 53(8), 69(7), 49(7), 46(7), 43(7)
6-Methyl-3-hepten-2-one
995
1008
1055
1368
1695
1326
43, 69(84), 55(58), 41(47), 39(36), 84(25), 111(23), 71(13), 108(9),
39(9), 56(8), 29(8), 53(7), 83(5), 126(4), 93(4)
94, 109(33), 93(16), 67(13), 39(11), 95(7), 65(7), 41(7), 92(5),
68(5), 28(5), 80(4), 77(4), 53(4)
3-Isopropylpyrrole
4,5-Dimethyl-2-(1-methylpropyl)oxazole
124, 125(47), 138(20), 43(20), 153(19), 41(13), 55(12), 70(10),
42(9), 39(9), 27(8), 54(7), 29(6), 111(5), 110(5), 69(5)
3,5(or 6)-Diethyl-2-methylpyrazine
1171
1204
1218
1248
1507
1561
1590
1608
NIST
NIST
NIST
2,5(or 6)-Dimethyl-3-(2-methylpropyl)pyrazine
2,5(or 6)-Dimethyl-3-(2-methylpropyl)pyrazine
2-(2-Methylbutyl)-3-methylpyrazine
108, 41(10), 39(9), 109(7), 149(6), 135(6), 42(6), 67(5), 52(4),
121(3), 80(3), 66(3), 53(3), 40(3), 164(0)
2-(3-Methylbutyl)-3-methylpyrazine
1255
1625
108, 121(15), 41(10), 39(10), 149(9), 109(8), 42(8), 67(6), 53(5),
52(4), 40(4), 93(3), 80(3), 66(3), 43(3), 29(3), 164(0)
2-(2-Methylpropyl)-3,5,6-trimethylpyrazine
2,5(or 6)-Dimethyl-3-(3-methylbutyl)pyrazine
2,5(or 6)-Dimethyl-3-(3-methylbutyl)pyrazine
1,2-Dihydro-1,1,6-trimethylnaphthalene
1282
1318
1332
1369
1585
1694
1724
1826
NIST
NIST
NIST
NIST
a
b
c
Linear retention index on VF5-ms column (Varian).
Linear retention index on an HP-Innowax column (Agilent).
First number is the base peak; molecular ion in bold type.
Spectrum is present in the NIST/NIH/EPA database (2005).
d

756
J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
Table 2
ma formation, those amino acids which decompose during the
Maillard reaction to give Strecker aldehydes, should be examined.
Of particular importance in this respect are glycine, alanine, valine,
leucine, isoleucine and phenylalanine, which were all present at
levels at least 2.5 times higher in the sulphate-deprived samples
compared with their levels in the sulphate-fed samples (Table 3).
In fact, levels of leucine, isoleucine and phenylalanine were at least
seven times higher in the Sꢀ potatoes of all three varieties, and this
would be expected to have an effect on aroma characteristics.
Although statistically significant, the effect of sulphate deprivation
on levels of sugars was much less marked (Table 4). However, a
4.5-fold increase in the glucose levels of Sꢀ Maris Piper potatoes
was observed.
Free asparagine, total free amino acids (excluding arginine) and total sugars (glucose,
fructose, and sucrose) in flour from three varieties of potato, fertilised with (S+) or
without (Sꢀ) sulphur (adapted from Elmore et al. (2007)).
Potato variety
Free asparagine
Total free amino acids
Total sugars
(mmol kgꢀ1
)
(mmol kgꢀ1
)
(mmol kgꢀ1
)
King Edward Sꢀ
King Edward S+
Maris Piper Sꢀ
Maris Piper S+
Prairie Sꢀ
95.4
177
51.8
91.6
126
62.5
663
364
338
178
310
112
34.1
32.3
47.3
30.6
37.1
30.6
Prairie S+
Table 3
Sulphur deprivation caused similar changes in the relative
amounts of sugars and amino acids in wheat flour (Muttucumaru
et al., 2006). When the wheat flour was cooked, there were large
increases in volatile compounds formed from Strecker aldehydes,
and decreases in compounds that could be formed from sugar
decomposition, such as furfural (Elmore et al., 2008). Wheat flour
usually contains substantially greater amounts of sugar than free
amino acid, while the converse is true for potato flour if it is pre-
pared immediately after harvesting, as in this experiment. Sulphur
deprivation, however, increased the total free amino acids in wheat
grain to a concentration greater than that of total sugars, making
the free amino acid: sugar ratio more similar to that seen in potato.
Ratios of free amino acids ([amino acid]S+ = 1) in potato flour grown without sulphur
(Sꢀ), compared with potato flour grown with sulphur (S+) (adapted from Elmore et al.
(2007)).
Amino acid
Potato variety
King Edward
Maris Piper
Prairie
Asparagine
Glutamine
Tryptophan
Alanine
Glycine
Valine
0.54
3.63
9.89
4.22
3.80
3.44
12.7
8.90
3.80
1.37
6.22
1.61
1.01
3.25
1.24
7.24
5.22
12.4
2.75
24.0
0.57
4.89
10.4
4.52
15.7
5.61
12.6
9.12
5.31
0.94
6.43
1.38
0.80
2.65
0.84
10.3
15.6
8.39
3.93
66.5
2.02
7.50
15.5
2.54
3.05
6.34
12.4
10.7
2.88
1.48
4.20
3.28
1.00
2.37
1.51
15.7
11.6
11.3
4.38
1.67
Leucine
Isoleucine
Threonine
GABA
Serine
Proline
Aspartic acid
Methionine
Glutamic acid
Phenylalanine
Ornithine
Lysine
3.2. Effect of sulphur deprivation on acrylamide formation in cooked
potato
Elmore et al. (2007) found that acrylamide formation in potato
flour during heating at 160 °C for 20 min. was always lower in flour
from Sꢀ plants than in flour from S+ plants. When the levels of
acrylamide formed on heating were plotted against the concentra-
tion of asparagine, expressed as a percentage of the total free ami-
no acid pool, a close correlation was revealed (r2 = 0.835). This
result suggested that changes in asparagine concentration per se
were outweighed by the large increases in the size of the total ami-
no acid pool resulting, in the main, from the increase in glutamine
concentration. Acrylamide formation was reduced in heated Sꢀ
flour because of competition between asparagine and other amino
acids for participation in the Maillard reaction, where sugar levels
were limiting.
Histidine
Tyrosine
Total free amino acids
1.82
1.90
2.77
Table 4
Ratios of sugars ([sugars]S+ = 1) in potato flour grown without sulphur (Sꢀ), compared
with potato flour grown with sulphur (S+) (adapted from Elmore et al. (2007)).
In the present study, acrylamide formation was measured after
heating at 180 °C (Table 5), to allow comparison with the data for
aroma compounds, which were acquired at the higher tempera-
ture, in order to produce sufficient peaks for analysis. The amounts
of acrylamide formed at 180 °C were 3–4 times greater than those
formed at 160 °C, for each treatment. However, the ratios: [acryl-
amide]S+: [acrylamide]Sꢀ remained relatively constant for all varie-
ties (2.85 and 2.87 for King Edward, 2.68 and 2.33 for Maris Piper,
and 1.60 and 1.49 for Prairie, at 180 °C and 160 °C, respectively),
showing that the hypotheses made on data acquired at 160 °C by
Elmore et al. (2007) were also valid at 180 °C. In fact, the correla-
tion between free asparagine as a proportion of the total free ami-
no acid pool (calculated from Table 2) and acrylamide formed was
higher at 180 °C, with r2 > 0.90, even when the trend-line was
forced through the origin (y = 0.237x; r2 = 0.905). At both 160 °C
and 180 °C, there were no correlations between acrylamide and
absolute asparagine concentration, or acrylamide and sugar
concentration.
Sugar
Variety
King Edward
Maris Piper
Prairie
Glucose
Fructose
Sucrose
1.50
1.84
1.00
4.50
1.45
1.36
1.23
2.20
1.20
Total sugars
1.06
1.55
1.22
deprivation may impact on aroma formation in cooked potatoes,
by examining its effect on free amino acid and sugar concentra-
tions in the potato flour. Table 2 compares concentrations of aspar-
agine, total free amino acids and total sugars across the six
samples. In order to highlight which amino acids were most af-
fected by sulphur deprivation, Table 3 displays the relative concen-
tration of each free amino acid and the relative concentration of
total free amino acids in the Sꢀ samples, compared with the S+
samples, for each variety, while Table 4 does the same for each
sugar.
The effect of sulphur deprivation on asparagine, expressed as a
proportion of the free amino acid pool, was the key parameter,
with regard to the formation of acrylamide; sulphur deprivation
led to less acrylamide formation in heated tuber flour from all
three varieties (Elmore et al., 2007). However, with regard to aro-
3.3. Effect of sulphur deprivation on the volatile compounds in cooked
potato flour
Forty-nine compounds were present in at least one of the
headspace extracts of the heated flour with an average total

J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
757
Table 5
Effect of applying sulphur fertiliser during the growing of potatoes on the headspace volatile compoundsa and the acrylamide contentb of flour cooked for 20 min at 180 °C.
LRIc
Compound
King Edward
Maris Piper
Prairie
SE
Effect of sulphurd
Effect of varietye
Sꢀ
269
15.4
7.62
346
S+
Sꢀ
673
S+
Sꢀ
225
15.4
8.60
416
S+
**
**
554
2-Methylpropanal
148
248
189
48.4
1.44
4.41
47.7
152
***
597
2-Butanone
7.80
5.42
68.3
171
23.9
9.77
611
2030
34.4
26.9
637
10.3
26.8
146
12.2
23.4
128
NS
599
Hexane
NS
NS
***
**
655
3-Methylbutanal
***
***
***
*
664
2-Methylbutanal
873
459
943
337
723
2-Vinylfuran
39.0
10.6
448
9.19
17.6
38.0
73.7
9.41
33.5
12.3
2.62
3.32
14.8
2.67
8.84
13.5
2.11
2.60
13.9
9.88
28.3
4.77
0.81
13.4
18.6
1.73
16.6
7.94
4.73
11.5
8.40
1.18
3.32
3.81
13.6
6.40
31.9
2.96
6.54
2.28
45.5
0.94
1.03
1.27
0.77
0.22
7.74
12.4
7.11
24.0
129
25.6
20.2
149
6.84
22.7
80.2
85.0
9.46
48.5
14.9
2.93
4.80
30.4
2.78
20.6
15.2
5.82
5.52
17.9
16.0
27.2
3.97
1.19
27.2
23.1
2.04
19.4
9.08
7.07
15.9
7.84
1.71
6.15
7.53
18.8
10.6
36.8
3.16
10.3
2.17
48.6
1.46
1.37
1.60
0.80
0.33
10.2
12.6
3.32
1.63
53.6
3.08
5.32
2.19
1.93
3.04
2.61
2.31
1.92
1.27
0.41
1.74
2.27
1.93
4.36
1.01
2.35
2.69
5.60
0.48
3.37
3.64
1.29
3.60
1.83
4.24
2.75
4.41
3.94
2.38
1.96
1.97
5.02
1.77
3.03
2.66
3.68
5.53
2.44
3.48
1.98
2.16
0.889
**
738
1-Methylpyrrole
NS
***
***
746
Dimethyl disulphide
748
Pyrrole
76.8
47.0
32.2
24.1
25.0
18.4
18.6
16.1
10.3
22.3
17.1
17.0
30.0
26.3
22.9
14.0
22.7
80.3
10.6
23.5
52.1
20.9
32.9
16.4
30.1
13.0
25.9
22.6
32.6
27.7
36.4
42.4
19.3
20.5
52.2
31.0
52.5
14.8
29.0
20.0
20.7
4.35
75.7
64.4
28.5
32.3
35.6
34.1
5.33
23.3
9.73
40.0
22.2
25.1
31.4
63.5
28.5
32.3
28.9
54.8
13.9
39.4
34.3
16.0
43.7
32.3
53.3
33.9
54.5
49.6
36.4
19.9
42.1
51.4
24.9
34.6
61.4
38.4
48.6
28.6
33.8
15.6
26.5
4.81
84.1
11.3
32.8
13.3
4.41
6.45
8.65
3.20
9.79
15.2
5.33
6.39
16.0
15.7
34.8
7.35
2.55
32.1
12.7
3.18
17.5
8.75
6.41
17.4
12.0
4.13
7.71
5.96
16.0
8.29
36.1
3.00
8.88
2.18
54.5
3.85
1.32
2.02
1.40
0.53
26.5
12.9
92.3
40.7
35.6
10.9
12.6
9.45
10.5
12.6
9.27
16.0
11.6
10.1
11.2
25.6
26.5
12.5
9.36
21.1
21.7
14.6
9.36
6.03
18.4
12.9
39.3
11.4
21.3
22.8
11.1
5.92
29.4
5.14
5.45
4.86
41.0
6.74
6.10
4.23
5.47
1.99
7.43
7.87
NS
NS
***
**
769
Toluene
801
Hexanal
NS
NS
***
***
828
Methylpyrazine
***
***
*
***
***
**
856
875
890
2-Methyl-2-(methylthio)propanal
2-Isopropyl-(E)-2-butenal
2-Heptanone
***
***
897
Styrene
903
Heptanal
NS
NS
***
***
916
917
919
920
2,5(and 6)-Dimethylpyrazine
(E)-5-Methyl-2-hexenal
(E)-4-Methyl-2-hexenal
Ethylpyrazine
***
***
*
***
***
NS
***
**
***
951
1-Isobutylpyrrole
**
971
Benzaldehyde
***
***
***
***
***
***
*
974
974
982
4,5-Dimethyl-2-isopropyloxazole
3-Methyl-2-(methylthio)butanal
Dimethyl trisulphide
992
2-Pentylfuran
NS
***
***
***
***
***
***
***
***
***
***
***
***
993
(E or Z)-6-Methyl-3-hepten-2-one
2-Ethyl-5(or 6)-methylpyrazine
2-Ethyl-3-methylpyrazine + trimethylpyrazine
3-Isopropylpyrrole
2-Ethyl-5(or 6)-methylpyrazine
Phenylacetaldehyde
4,5-Dimethyl-2-(1-methylpropyl)oxazole
1-(2-Methylbutyl)pyrrole
1-(3-Methylbutyl)pyrrole
3-Ethyl-2,5(or 6)-dimethylpyrazine
2,6-Diethylpyrazine
***
***
***
***
***
***
***
***
***
***
1001
1005
1006
1007
1052
1053
1059
1062
1081
1082
1106
1141
1159
1203
1208
1248
1256
1280
1318
1332
1376
Nonanal
NS
NS
***
***
2-Methyl-3-isobutylpyrazine
3,5(or 6)-Diethyl-2-methylpyrazine
2,5(or 6)-Dimethyl-3-(2-methylpropyl)pyrazine
Decanal
2-(2-Methylbutyl)-3-methylpyrazine
2-(3-Methylbutyl)-3-methylpyrazine
2-(2-Methylpropyl)-3,5,6-trimethylpyrazine
2,5(or 6)-Dimethyl-3-(3-methylbutyl)pyrazine
2,5(or 6)-Dimethyl-3-(3-methylbutyl)pyrazine
1,2-Dihydro-1,1,6-trimethylnaphthalene
Acrylamide
***
***
**
**
NS
NS
***
***
***
***
***
***
***
***
**
***
***
***
NS
***
a
b
c
Peak area values relative to that of 100 ng of 1,2-dichlorobenzene (peak area = 100). Each value is the mean of three replicates.
Values are in mg per kg, based on weight of uncooked sample. Each value is the mean of three replicates.
Linear retention index on J&W DB-5 column.
d,e
Probability that there is a difference between means for (d) sulphur feeding or (e) variety; NS, no significant difference between means (p > 0.05).
Significant at the 5% level.
*
**
Significant at the 1% level.
Significant at the 0.1% level.
***
ion chromatogram peak area greater than or equal to 20% of that
of 100 ng of internal standard (Table 5). Forty-one compounds
were affected by sulphur treatment and 42 compounds were af-
fected by variety. Benzaldehyde was the only compound that
was present at increased levels in cooked S+ potato flour,
although its formation did not correlate with that of acrylamide.
Several authors have reported the formation of benzaldehyde
from phenylalanine (Chu & Yaylayan, 2008), and its formation
is generally thought to be associated with that of phenylacetal-
dehyde, the Strecker aldehyde of phenylalanine. Phenylacetalde-
hyde was 3–5 times higher in the aroma volatiles of the Sꢀ
potato flour, mirroring phenylalanine, which was 7–15 times
higher; therefore, it seems unlikely, in this case, that this is
the only route to benzaldehyde.
Table 5 shows that nearly all of the compounds that were
formed at higher levels in cooked Sꢀ potato flour compared with
cooked S+ flour were derived from the Maillard reaction. These
compounds included Strecker aldehydes, alkylpyrazines, and
a,b-
unsaturated aldehydes that are formed from aldol condensation
of two Strecker aldehydes (Elmore et al., 2008). Some pyrazines,
particularly those containing a branched side-chain, indicative of
the presence of 2-methylpropanal (from valine), 2-methylbutanal
(from isoleucine) or 3-methylbutanal (from leucine), were over
ten times higher in heated Sꢀ potato flour.

758
J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
Aldol condensation products reported in Table 5 are (E)-5-
ine in the plot, a straight line was also observed (y = 1.35x,
r2 = 0.979). When the same analysis was performed for the other
amino acids, no correlation was observed. It is assumed that the
2-vinyl moiety is obtained from an amino acid, in the same way
as for acrylamide and styrene. Intriguingly, styrene and 2-vinylfu-
ran were two of the compounds which were most negatively cor-
related with acrylamide (r2 = 0.897 and 0.934, respectively).
It is clear from Table 5 that sulphur deprivation has little effect
on the fatty acid composition of potato flour, because only one sig-
nificant change was observed in the volatile compounds typically
associated with lipid degradation (Grosch, 1982), such as hexanal,
heptanal, nonanal, decanal, 2-heptanone and 2-pentylfuran. 2-
Heptanone showed an inconsistent effect of sulphur deprivation.
methyl-2-hexenal, 2-isopropyl-(E)-2-butenal, (E)-4-methyl-2-hex-
enal and (E or Z)-6-methyl-3-hepten-2-one. Both (E)-5-methyl-2-
hexenal and 2-isopropyl-(E)-2-butenal have been identified in
heated wheat flour (Elmore et al., 2008), and derive from the con-
densation of acetaldehyde (from alanine) with 3-methylbutanal.
(E)-4-Methyl-2-hexenal has been identified in potato crisps (But-
tery, 1973) and could be formed from the condensation of 2-meth-
ylbutanal with acetaldehyde. 6-Methyl-3-hepten-2-one could be
formed from an aldol condensation of 3-methylbutanal with ace-
tone. It has been reported in grape brandy (Schreier, Drawert, &
Winkler, 1979) but no other foods to date.
An increase in methionine led to an increase in methanethiol,
which, as well as forming dimethyl disulphide and dimethyl trisul-
phide (Belitz, Grosch, & Schieberle, 2004), also reacted with valine
and isoleucine to give 2-methyl-2-(methylthio)propanal and 3-
methyl-2-(methylthio)butanal, respectively. Neither of these com-
pounds has previously been identified in food.
4,5-Dimethyl-2-isopropyloxazole and 4,5-dimethyl-2-(1-meth-
ylpropyl)oxazole are likely to be formed from an aminocarbonyl
reacting with valine and isoleucine, respectively. 4,5-Dimethyl-2-
isopropyloxazole has been reported in several foods, including
French fries (Carlin, Jin, Huang, Ho, & Chang, 1986), while 4,5-di-
methyl-2-(1-methylpropyl)oxazole does not appear to have been
reported previously.
Four 1-alkylpyrroles are reported in Table 5, of which 1-(2-
methylpropyl)pyrrole, 1-(2-methylbutyl)pyrrole and 1-(3-meth-
ylbutyl)pyrrole all increased in cooked Sꢀ potato flour. It is clear
that the 1-substitution in these compounds is due to 2-methyl-
propanal, 2-methylbutanal and 3-methylbutanal, respectively.
What is less clear is how the rest of the pyrrole ring is formed,
although Yaylayan and Keyhani (2001) suggest that serine can
decompose to form unsubstituted pyrrole. 1-(2-Methylpropyl)pyr-
role, 1-(2-methylbutyl)pyrrole and 1-(3-methylbutyl)pyrrole were
all present in heated wheat flour but, in that food system, they
were not affected by sulphur deprivation (Elmore et al., 2008). 3-
Isopropylpyrrole has not previously been found in food. We pro-
pose that 3-methylbutanal from leucine forms the isopropyl side-
chain, as well as C2 and C3 of the pyrrole ring.
3.4. Effect of variety on the volatile compounds in cooked potato flour
In general, the aroma compositions of the S+ samples were sim-
ilar for the three varieties. However, sulphur deprivation affected
the aroma composition of Prairie to a far lesser degree than it did
the other two varieties; total aroma compounds increased in Maris
Piper and King Edward by 3.33 and 3.29 times, respectively, but in-
creased in Prairie by only 1.70 times. The largest effects for com-
pounds derived from leucine and isoleucine were found in King
Edward; sulphur deprivation led to a 30-fold increase in four pyr-
azines with methylbutyl groups in their side-chains. Compounds
derived from valine, such as 2-methylbutanal, 1-isobutylpyrrole
and 4,5-dimethyl-2-isopropyloxazole, and compounds derived
from phenylalanine, such as phenylacetaldehyde, styrene and tol-
uene, showed a relatively large effect of sulphur deprivation in
Maris Piper potatoes, compared with the other varieties.
Although Sꢀ Maris Piper potatoes contained 4.5 times more
glucose than did S+ Maris Piper potatoes, it is not clear if any of
the differences observed in the aroma compounds of the cooked
potato flour could be attributed to this large increase.
3.5. Correlations between amino acid precursors and aroma
compounds
Many of the aroma compounds formed in cooked potato flour
derive from the thermally-induced breakdown of amino acids,
and in many cases it is clear which amino acid is the precursor
of a particular aroma compound, e.g., phenylacetaldehyde from
phenylalanine. We have already proposed that the formation of
acrylamide from asparagine may be dependent on the amount of
asparagine present, relative to the other free amino acids. With this
in mind, we examined the formation of certain aroma compounds
that had been significantly affected by sulphur deprivation. It
seemed likely that those aroma compounds that were highly corre-
lated with the relative amount of a precursor amino acid needed a
sugar fragment in their formation, as in the case of acrylamide,
whereas those compounds that were highly correlated with the
absolute amount of a precursor amino acid were less dependent
on sugar fragments for their formation. Hence, this exercise might
provide information about the formation pathways for particular
groups of aroma compounds.
Styrene (vinylbenzene) is of interest because it is an analogue of
acrylamide. Acrylamide is formed from a carbonyl moiety reacting
with asparagine, while styrene is formed from a carbonyl reacting
with phenylalanine (Stadler et al., 2004). It is assumed that toluene
is formed in a similar way to styrene. When the mean concentra-
tion of styrene was plotted against the mean concentration of tol-
uene for the six potato flour samples studied (data not shown), a
linear relationship was found in which:
½tolueneꢂ ¼ 2:90 ꢁ ½styreneꢂ:
When Baltes and Mevissen (1988) heated phenylalanine with
glucose, styrene was reported as a major product, while toluene
was present in smaller amounts. Seck and Crouzet (1981) also re-
ported toluene formation when phenylalanine was heated with
glucose, fructose or ascorbic acid.
2-Vinylfuran was present in heated Sꢀ wheat flour at concen-
trations up to 25 times higher than in S+ wheat, although its source
was uncertain (Elmore et al., 2008). In heated potato flour, a simi-
lar, although less extreme effect, was noted. Although serine in-
creased in Sꢀ potato flour, its increase in Sꢀ wheat flour was
much greater. When the relative concentrations of 2-vinylfuran
([2-vinylfuran]Sꢀ/[2-vinylfuran]S+) were plotted against the rela-
tive concentrations of serine ([serine]Sꢀ/[serine]S+) in heated wheat
Strecker aldehydes require an
a-dicarbonyl or related com-
pound, usually derived from a sugar, for their formation to proceed
(Belitz et al., 2004). In Sꢀ potatoes, absolute concentrations of free
amino acids are relatively high (Table 2), and hence increased com-
petition between free amino acids for sugar fragments results in a
non-linear increase in Strecker aldehydes in absolute terms. How-
ever, when the four major Strecker aldehydes present in heated po-
tato flour (2-methylpropanal, 2-methylbutanal, 3-methylbutanal
and phenylacetaldehyde) were plotted against their precursor
amino acids (valine, isoleucine, leucine and phenylalanine,
respectively), with the precursor amino acids expressed relative
and potato flour,
a straight line was observed (y = 0.653x,
r2 = 0.987), suggesting that serine may participate in the formation
of 2-vinylfuran. However, when threonine was replaced with ser-

J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
759
Table 6
sugar reacts with the amine group of a free amino acid (Belitz
et al., 2004), subsequently generating aroma and colour. Hence,
in theory, if reducing sugars are limiting, addition of any free ami-
no acid to a food will cause changes in all the aroma compounds
formed via the Maillard reaction, due to increased competition
for a limited number of sugar fragments. Likewise, addition of a
reducing sugar to a food where free amino acids are limiting will
have the same effect. However, the amino acids threonine and ser-
ine can behave like sugars with regard to their participation in the
Maillard reaction (Shu, 1999) and their behaviour would also have
to be considered when predicting the effects of precursor addition.
The effects of competition could also explain why the aroma
compounds of Prairie potatoes were affected the least by sulphur
deprivation. Table 7, which is again adapted from data reported
by Elmore et al. (2007), shows how the amounts of the free amino
acids in the potato flours, relative to the total free amino acid con-
tent, were affected by sulphur deprivation. Two amino acids
important in aroma formation, alanine and glycine, increased in
relative, as well as absolute terms, in Maris Piper and King Edward,
whereas they showed no relative increase in Prairie. In fact, alanine
decreased slightly in Prairie in relative terms and may be the rea-
son why some pyrazines are actually lower in Sꢀ Prairie than in S+
Prairie, e.g., ethylpyrazine, 2-ethyl-5-methylpyrazine and 2-ethyl-
6-methylpyrazine (compounds where acetaldehyde from alanine
breakdown may provide the ethyl side-chain (Low, Parker, & Mot-
tram, 2007)). Threonine also showed no relative increase in Sꢀ
Prairie potatoes and the relative increase in serine was small com-
pared with that observed in Maris Piper and King Edward.
Correlations (r2) between some aroma compounds formed in heated potato flour and
their precursor amino acids, amino acid concentrations expressed in absolute terms
([aa]abs) or relative to total free amino acids ([aa]rel).
Aroma compound
Precursor
amino acid
Correlation
coefficient (r2)
[aa]abs
[aa]rel
2-Methylpropanal
2-Methylbutanal
3-Methylbutanal
Phenylacetaldehyde
Dimethyl disulphide
4,5-Dimethyl-2-isopropyloxazole
4,5-Dimethyl-2-(1-methylpropyl)oxazole Isoleucine
1-(2-Methylpropyl)pyrrole
1-(2-Methylbutyl)pyrrole
1-(3-Methylbutyl)pyrrole
2-Methyl-3-isobutylpyrazine
2-Methyl-3-(2-methylbutyl)pyrazine
2-Methyl-3-(3-methylbutyl)pyrazine
Valine
Isoleucine
Leucine
Phenylalanine 0.812
Methionine
Valine
0.399
0.693
0.742
0.683
0.936
0.943
0.932
0.360
0.944
0.929
0.906
0.880
0.906
0.576
0.588
0.655
0.949
0.723
0.772
0.624
0.792
0.666
0.864
0.886
0.892
Valine
Isoleucine
Leucine
Valine
Isoleucine
Leucine
Table 7
Free amino acids in potato flour grown with and without sulphur, expressed as a
percentage of total free amino acids (adapted from Elmore et al. (2007)).
Amino acid
King Edward
Maris Piper
Prairie
Sꢀ
S+
Sꢀ
S+
Sꢀ
S+
Asparagine
Glutamine
Tryptophan
Alanine
Glycine
Valine
14.4
44.2
0.13
48.6
22.1
0.02
15.3
39.6
0.31
51.5
15.5
0.06
40.7
30.0
0.10
1.00
0.39
3.90
0.80
1.38
0.04
0.54
1.37
1.22
4.32
2.82
0.87
6.55
2.74
0.34
0.73
0.18
0.02
55.8
11.1
0.02
4.03
0.95
3.11
0.71
1.06
0.06
1.40
1.03
10.57
6.00
1.72
0.75
6.03
1.57
0.07
0.39
0.11
1.69
1.74
0.46
0.02
1.65
0.10
0.22
0.15
0.67
1.37
3.10
6.75
3.10
0.42
8.84
0.40
0.02
0.06
0.07
2.14
3.52
6.68
1.30
2.38
0.19
1.66
1.41
6.53
2.64
2.20
1.78
5.32
3.16
0.60
1.12
0.17
1.96
0.90
0.43
2.26
0.20
0.49
0.12
0.60
2.86
1.93
3.64
5.25
1.27
1.09
0.36
1.71
0.18
0.36
0.13
0.52
2.56
0.80
3.65
7.81
1.02
4. Conclusion
Leucine
We have previously shown (Elmore et al., 2007) that acrylamide
formation was reduced in flour from Sꢀ tubers heated at 160 °C,
because the relative amount of free asparagine compared with to-
tal free amino acids was lower in Sꢀ tubers than in S+ tubers. Sugar
levels were similar in both Sꢀ and S+ tubers and much lower than
levels of free amino acids; competition between amino acids for
sugar fragments, as part of the Maillard reaction, meant that less
acrylamide was produced in heated Sꢀ flour. In the current study,
we have shown that this effect is also observed at 180 °C, with
acrylamide levels 3–4 times higher in all heated flours than at
160 °C.
This competition between amino acids also had a significant ef-
fect on the volatile aroma compounds produced through the Mail-
lard reaction. Amino acids with alkyl and aryl side-chains, as well
as serine and threonine, were found in relatively high amounts in
Sꢀ tubers compared with S+ tubers, resulting in increased levels of
many compounds, in particular Strecker aldehydes, aldehydes and
ketones formed from aldol condensation, alkylpyrazines and N-
substituted alkylpyrroles, in heated Sꢀ flour. Only benzaldehyde
was found at higher concentrations in heated S+ flour.
Isoleucine
b-Alanine
Threonine
GABA
Serine
Proline
Aspartic acid
Methionine
Glutamic acid
Phenylalanine
Ornithine
Lysine
12.0
12.0
0.58
0.07
0.25
0.08
0.06
0.48
0.08
0.18
0.12
0.03
Histidine
Tyrosine
to total free amino acids, correlations were much improved
(Table 6).
These Strecker aldehydes can participate in further reactions,
providing the alkyl group in the following compounds. 2-Alkyl-
4,5-dimethyloxazoles and 1-alkylpyrroles also seem to be highly
correlated with their precursor amino acids in relative terms, while
dimethyl disulphide composition appears to be unaffected by su-
gar content, as does 2-methyl-3-alkylpyrazine formation. 2-
Methyl-3-alkylpyrazine formation may result from the interaction
of a Strecker aldehyde with serine and threonine breakdown prod-
ucts, obviating the need for a reducing sugar (Shu, 1999).
Acrylamide mitigation strategies that involve the addition of an
amino acid, such as glycine, have been proposed in potatoes and
their products (Bråthen, Kita, Knutsen, & Wicklund, 2005; Low
et al., 2006). Low et al. (2006) showed that addition of glycine to
a potato model system resulted in a fall in the Strecker aldehydes
2-methylpropanal, 2-methylbutanal, 3-methylbutanal and phenyl-
acetaldehyde, and proposed that this effect was due to glycine
competing with the other amino acids for sugar breakdown prod-
ucts. In the Maillard reaction, the carbonyl function of a reducing
We have shown that, for freshly-harvested potatoes, sulphur
deprivation during cultivation resulted in reduced acrylamide for-
mation in cooked tuber flour and an overall increase in aroma vol-
atiles. However, sulphur deprivation led to small tubers and a large
reduction in yields in all three varieties. The total yields for each
treatment, across the five plants studied, are shown in Table 8. In
Table 8
Tuber yields (from totals of S plants) from potato plants studied.
Yield (g)
Sꢀ
199
619
155
S+
608
1110
1260
King Edward
Maris Piper
Prairie

760
J.S. Elmore et al. / Food Chemistry 122 (2010) 753–760
Elmore, J. S., Koutsidis, G., Dodson, A. T., Mottram, D. S., & Wedzicha, B. L. (2005).
Measurement of acrylamide and its precursors in potato, wheat, and rye model
systems. Journal of Agricultural and Food Chemistry, 53(4), 1286–1293.
Elmore, J. S., Mottram, D. S., Muttucumaru, N., Dodson, A. T., Parry, M. A. J., &
Halford, N. G. (2007). Changes in free amino acids and sugars in potatoes due to
sulfate fertilization and the effect on acrylamide formation. Journal of
Agricultural and Food Chemistry, 55(13), 5363–5366.
Elmore, J. S., Parker, J. K., Halford, N. G., Muttucumaru, N., & Mottram, D. S. (2008).
Effects of plant sulfur nutrition on acrylamide and aroma compounds in cooked
wheat. Journal of Agricultural and Food Chemistry, 56(15), 6173–6179.
particular, yields of S+ Prairie were over eight times greater than
those of Sꢀ Prairie, while yields of S+ Maris Piper were less than
twice those of Sꢀ Maris Piper. Thus, potato plants that are severely
deprived in sulphur will not produce commercially viable potatoes.
Furthermore, a lack of sensory data means that we cannot be sure
how the increases in aroma compounds will affect sensory quality.
Clearly, it is important that an optimum level of sulphur availabil-
ity is established that supports healthy growth and tuber yield,
while neither exacerbating the risk of acrylamide formation during
processing nor compromising palatability.
Granvogl, M., Wieser, H., Koehler, P., Von Tucher, S.,
& Schieberle, P. (2007).
Influence of sulfur fertilization on the amounts of free amino acids in wheat.
Correlation with baking properties as well as with 3-aminopropionamide and
acrylamide generation during baking. Journal of Agricultural and Food Chemistry,
55(10), 4271–4277.
Grosch, W. (1982). Lipid degradation products and flavours. In I. D. Morton & A. J.
MacLeod (Eds.), Food flavours (pp. 325–398). Amsterdam: Elsevier.
Groves, J. K., Anderson, H. J., & Nagy, H. (1971). Pyrrole chemistry. Part XIII. New
syntheses of 3-alkylpyrroles. Canadian Journal of Chemistry, 49(14), 2427–2432.
Kourouli, T., Kefalas, P., Ragoussis, N., & Ragoussis, V. (2002). A new protocol for a
Acknowledgements
This study was financially supported by the Biotechnology and
Biological Sciences Research Council (BBSRC) of the United King-
dom in conjunction with the United Kingdom’s Food Standards
Agency (Grants BB/C508634/1 and BB/C508669/1). Rothamsted
Research receives grant-aided support from the BBSRC.
regioselective aldol condensation as an alternative convenient synthesis of
ketols and
a-
a,b-unsaturated ketones. Journal of Organic Chemistry, 67(13),
4615–4618.
Low, M. Y., Koutsidis, G., Parker, J. K., Elmore, J. S., Dodson, A. T., & Mottram, D. S.
(2006). Effect of citric acid and glycine addition on acrylamide and flavor in a
potato model system. Journal of Agricultural and Food Chemistry, 54(16),
5976–5983.
Low, M. Y., Parker, J. K., & Mottram, D. S. (2007). Mechanisms of alkylpyrazine
References
formation in
Agricultural and Food Chemistry, 55(10), 4087–4094.
Mathew, C. T., Ulmer, H. E. (1976). alpha-Formyl sulfides and 2-
a potato model system containing added glycine. Journal of
Akiyama, T., Enomoto, Y., & Shibamoto, T. (1978). A new method of pyrazine
synthesis for flavour use. Journal of Agricultural and Food Chemistry, 26(5),
1176–1179.
Artman, L. D., Balandrin, M. F., Moe, S. T., Van Wagenen, B. C., & Pesyan, A. (2006).
Preparation of analogs of isovaleramide for treating central nervous system
conditions or diseases. World patent no. WO2006012603.
Baltes, W., & Mevissen, L. (1988). Model reactions on roast aroma formation. 6.
Volatile reaction products from the reaction of phenylalanine with glucose
during cooking and roasting. Zeitschrift für Lebensmittel-Untersuchung und
Forschung, 187(3), 209–214.
Belitz, H.-D., Grosch, W., & Schieberle, P. (2004). Food chemistry (3rd ed.). Berlin:
Springer-Verlag.
Bråthen, E., Kita, A., Knutsen, S. H., & Wicklund, T. (2005). Addition of glycine
reduces the content of acrylamide in cereal and potato products. Journal of
Agricultural and Food Chemistry, 53(8), 3259–3264.
Buttery, R. G. (1973). Unusual volatile carbonyl components of potato chips. Journal
of Agricultural and Food Chemistry, 21(1), 31–33.
Carlin, J. T., Jin, Q. Z., Huang, T.-C., Ho, C.-T., & Chang, S. S. (1986). Identification of
alkyloxazoles in the volatile compounds from french-fried potatoes. Journal of
Agricultural and Food Chemistry, 34(4), 621–623.
Castle, L., Campos, M.-J., & Gilbert, J. (1991). Determination of acrylamide monomer
in hydroponically grown tomato fruits by capillary gas chromatography–mass
spectrometry. Journal of the Science of Food and Agriculture, 54(4), 549–555.
Chen, T.-K. (1994). Process for the preparation of 2,3,5(6)-trialkylpyrazines. European
patent no. EP590296.
&
hydrocarbylthioaldoximes. US patent no. US3931331.
Miginiac, P. (1990). A facile synthesis of 1,1,6-trimethyl-1,2-dihydronaphthalene.
Synthetic Communications, 20(12), 1853–1856.
Muttucumaru, N., Halford, N. G., Elmore, J. S., Dodson, A. T., Parry, M., Shewry, P. R.,
et al. (2006). Formation of high levels of acrylamide during the processing of
flour derived from sulfate-deprived wheat. Journal of Agricultural and Food
Chemistry, 54(23), 8951–8955.
Rizzi, G. (1968). Some reactions of methylpyrazines with organolithium reagents.
Journal of Organic Chemistry, 33(4), 1333–1337.
Schreier, P., Drawert, F., & Winkler, F. (1979). Composition of neutral volatile
constituents in grape brandies. Journal of Agricultural and Food Chemistry, 27(2),
365–372.
Seck, S.,
& Crouzet, J. (1981). Formation of volatile compounds in sugar-
phenylalanine and ascorbic-acid phenylalanine model systems during heat-
treatment. Journal of Food Science, 46(3), 790–793.
Shu, C.-K. (1999). Pyrazine formation from serine and threonine. Journal of
Agricultural and Food Chemistry, 47(10), 4332–4335.
Stadler, R. H., Robert, F., Riediker, S., Varga, N., Davidek, T., Devaud, S., et al. (2004).
In-depth mechanistic study on the formation of acrylamide and other
vinylogous compounds by the Maillard reaction. Journal of Agricultural and
Food Chemistry, 52(17), 5550–5558.
Theilig, G. (1953). Untersuchungen in der Oxazolreihe und Umwandlungen von
Oxazolen in Imidazole mittels Formamids (Formamid-Reaktionen, II. Mitteil.).
Chemische Berichte, 86(1), 96–109.
Weber, E. A., Koller, W. D., Graeff, S., Hermann, W., Merkt, N., & Claupein, W. (2008).
Impact of different nitrogen fertilizers and an additional sulfur supply on grain
yield, quality, and the potential of acrylamide formation in winter wheat.
Journal of Plant Nutrition and Soil Science, 171(4), 643–655.
Winter, M., Gautschi, F., Flament, I., & Stoll, M. (1976). 5-Methylfurfuryl furfuryl ether.
US patent no. US3952026.
Chu, F. L., & Yaylayan, V. A. (2008). Model studies on the oxygen-induced formation
of benzaldehyde from phenylacetaldehyde using pyrolysis GC–MS and FTIR.
Journal of Agricultural and Food Chemistry, 56(22), 10697–10704.
Curtis, T. Y., Muttucumaru, N., Shewry, P. R., Parry, M. A. J., Powers, S. J., Elmore, J. S.,
et al. (2009). Effects of genotype and environment on free amino acid levels in
wheat grain: Implications for acrylamide formation during processing. Journal
of Agricultural and Food Chemistry, 57(3), 1013–1021.
Yaylayan, V. A., & Keyhani, A. (2001). Elucidation of the mechanism of pyrrole
Dejaegher, Y., & De Kimpe, N. (2004). Rearrangement of 4-(1-haloalkyl)- and 4-(2-
formation during thermal degradation of 13C-labeled
L-serines. Food Chemistry,
74(1), 1–9.
haloalkyl)-2-azetidinones into methyl
aziridines and azetidines. Journal of Organic Chemistry, 69(18), 5974–5985.
a-alkylaminopentenoates via transient