Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Heterocyclic Aromatic Amines in Cooked MeatProducts: Causes, Formation, Occurrence, andRisk AssessmentMonika Gibis
Abstract: Meat products are sources of protein with high biological value and an essential source of other nutrients,such as vitamins and minerals. Heating processes cause food to become more appetizing with changes in texture,appearance, flavor, and chemical properties by the altering of protein structure and other ingredients. During heattreatment, heterocyclic aromatic amines (HAAs), potent mutagens/carcinogens, are formed due to the Maillard reaction.The HAAs are classified in at least 2 groups: thermic HAAs (100 to 300 °C) and pyrolytic HAAs (>300 °C). This reviewfocuses on the parameters and precursors which affect the formation of HAAs: preparation, such as the marinating ofmeat, and cooking methods, including temperature, duration, and heat transfer, as well as levels of precursors. Additionally,factors are described subject to pH, and the type of meat and ingredients, such as added antioxidants, types of carbohydratesand amino acids, ions, fat, and other substances inhibiting or enhancing the formation of HAAs. An overview of thedifferent analytical methods available is shown to determine the HAAs, including their preparation to clean up thesample prior to extraction. Epidemiological results and human daily intake of HAAs obtained from questionnaires showa relationship between the preference for very well-done meat products with increased HAA levels and an enhancedrisk of the incidence of cancer, besides other carcinogens in the diet. The metabolic pathway of HAAs is governed bythe activity of several enzymes leading to the formation of DNA adducts or HAA excretion and genetic sensitivity ofindividuals to the impact of HAAs on human cancer risk.
Keywords: antioxidants, β-carbolines, heterocyclic aromatic amines, meat, precursors
IntroductionIndividuals have been exposed to a range of toxic substances
throughout human history. These can be naturally occurring mu-tagens, which are mainly found in plant substances, or process-induced mutagens, which can arise during manufacture, such asfrom heating. Nitrosamines, polycyclic aromatic hydrocarbons,and heterocyclic aromatic amines are typical heat-induced com-pounds (Ferguson 2010; Oostindjer and others 2014). Widmark(1939) reported for the first time that extracts of roasted horse meatinduced cancer in the mammary glands of mice when multiple-swabbed on the back. The heterocyclic aromatic amines (HAAs)that are the focus of this article belong to the process-inducedmutagens that cannot be detected in unheated products (Skog andothers 1998b). In general, HAAs are formed from heated prod-ucts which contain sources of nitrogenous compounds, mainlyheated foods of animal origin, such as proteins and creatine (Skogand others 1998b). The formation of HAAs is principally de-pendent on the temperature, heat transfer, and heating conditions
MS 20151495 Submitted 3/9/2015, Accepted 1/12/2015. Author Gibis is withDept. of Food Physics and Meat Science, Inst. of Food Science and Biotechnology, Univ.of Hohenheim, Garbenstrasse 21/25, 70599 Stuttgart, Germany. Direct inquiries toauthor Gibis (E-mail: [email protected]).
used (Murkovic 2004a; Sugimura and others 2004). The struc-tures of HAAs which are formed at temperatures between 100 and300 °C are called thermic HAAs, IQ-type HAAs (imidazoquino-line or imidazoquinoxaline) or aminoimidazoazaarenes (Jägerstadand others 1998). Above 300 °C, pyrolysis of proteins and individ-ual amino acids occurs and HAAs are formed which are knownas pyrolytic HAAs or non-IQ-type HAAs (Jägerstad and others1998). Another classification of HAAs subdivides them into non-polar and polar HAAs (IQ-type compounds) due to their chem-ical properties. The names, abbreviations, properties, mutagenic-ity (Ames test), and molecular structures of HAAs are shown inTable 1.
Occurrence of Heterocyclic AminesBesides the important formation factors of cooking temperature
and duration of the heat treatment, thermic HAAs occur in almostall heated foods of animal origin, such as meat and fish, becausecreatine, the precursor needed for their formation, is containedin these products (Jägerstad and others 1998; Murkovic 2000).2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) hasbeen detected not only in products of animal origin, but alsoin wine, beer (Manabe and others 1993), and smoked cheese(Skog and others 1994; Naccari and others 2009). The sourceis possibly environmental, since PhIP can occur in burning
C© 2016 Institute of Food Technologists®doi: 10.1111/1541-4337.12186 Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 269
Heterocyclic amines in cooked meat products . . .
Tabl
e1–
Chem
ical
stru
ctur
e,na
me,
and
mol
ecul
arw
eigh
tof
30H
AA
s;so
me
ofth
eir
prop
erti
esan
din
dica
tion
ofth
eir
mut
agen
icit
y(A
mes
assa
yin
the
pres
ence
ofS-
9m
ixus
ing
Salm
onel
laTy
phim
uriu
mst
rain
sTA
98,T
A10
0an
dTA
1538
.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
2-A
min
o-im
idaz
oqui
nolin
esIQ 2-
Am
ino-
3-m
ethy
limid
azo[
4,5-
f]qu
inol
ine
7618
0-96
-6
192.
2po
lar
pKa
5.86
±0.
40c
433
(TA
98)7
(TA
100)
300
(TA
1538
)
Iso-
IQ2-
Am
ino-
1-m
ethy
limid
azo[
4,5-
f]qu
inol
ine
1024
08-2
5-3
192.
2po
lar
pKa
5.99
±0.
40c
na
MeI
Q2-
Am
ino-
3,4-
dim
ethy
limid
azo[
4,5-ƒ]
quin
olin
e77
094-
11-2
212.
3po
lar
pKa
6.22
±0.
40c
661
(TA
98)3
0(T
A10
0)85
9(T
A15
38)
2-A
min
o-im
idaz
oqui
noxa
lines
IQx
2-A
min
o-3-
met
hylim
idaz
o[4,
5-ƒ]
quin
oxal
ine
1083
54-4
7-8
199.
3po
lar
pKa
1.96
±0.
50c
75.4
(TA
98)1
.5(T
A10
0)10
3(T
A15
38)
8-M
eIQ
x2-
Am
ino-
3,8-
dim
ethy
limid
azo[
4,5-ƒ]
quin
oxal
ine
7750
0-04
-0
213.
3po
lar
pKa
2.20
±0.
50c
145
(TA
98)1
4(T
A10
0)84
.7(T
A15
38)
(Con
tinu
ed)
270 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .Ta
ble
1–Co
ntin
ued.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
4-M
eIQ
x2-
Am
ino-
3,4-
dim
ethy
limid
azo[
4,5-ƒ]
quin
oxal
ine
1083
54-4
8-9
213.
3po
lar
pKa
2.32
±0.
50c
1162
(TA
98)5
1(T
A10
0)10
42(T
A15
38)
4,8-
DiM
eIQ
x2-
Am
ino-
3,4,
8-tr
imet
hylim
idaz
o[4,
5-ƒ]
quin
oxal
ine
9589
6-78
-9
227.
3po
lar
pKa
2.56
±0.
50c
183
(TA
98)8
(TA
100)
225
(TA
1538
)
7,8-
DiM
eIQ
x2-
Am
ino-
3,7,
8-tr
imet
hylim
idaz
o[4,
5-ƒ]
quin
oxal
ine
9218
0-79
-5
227.
3po
lar
pKa
2.49
±0.
50c
163
(TA
98)9
.9(T
A10
0)18
9(T
A15
38)
TriM
eIQ
x2-
Am
ino-
3,4,
7,8-
tetr
aim
ethy
limid
azo[
4,5-ƒ]
quin
oxal
ine
1328
98-0
7-8
241.
3po
lar
pKa
2.85
±0.
50c
na
4-CH
2O
H-8
-MeI
Qx
2-A
min
o-4-
hydr
oxym
ethy
l-3,8
-dim
ethy
limid
azo[
4,5-ƒ]
quin
oxal
ine
1539
54-2
9-1
243.
3po
lar
pKa
2.17
±0.
50c
13.5
2±
0.50
d
99(T
A98
)2.6
(TA
100)
(Con
tinu
ed)
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 271
Heterocyclic amines in cooked meat products . . .
Tabl
e1–
Cont
inue
d.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
7-M
eIgQ
x2-
Am
ino-
1,7-
dim
ethy
l-1H
-imid
azo[
4,5-
g]qu
inox
alin
e93
4333
-16-
1
213.
2po
lar
pKa
3.95
±0.
50c
na
7,9-
MeI
gQx
2-A
min
o1,7
,9-t
ridim
ethy
l-1H
-imid
azo[
4,5-
g]qu
inox
alin
e15
6243
-39-
9
227.
3po
lar
pKa
4.65
±0.
50c
0.67
(TA
98)
2-A
min
o-im
idaz
opyr
idin
esPh
IP2-
Am
ino-
1-m
ethy
l-6-p
heny
l-im
idaz
o[4,
5-b]
pyrid
ine
1056
50-2
3-5
224.
3po
lar
pKa
7.72
±0.
30c
1.8
(TA
98)0
.12
(TA
100)
2.9
(TA
1538
)
4’O
H-P
hIP
2-A
min
o-1-
met
hyl-6
-(4’h
ydro
xyph
enyl
)-im
idaz
o[4,
5-b]
pyrid
ine
1268
61-7
2-1
240.
6po
lar
pKa
7.79
±0.
30c
8.27
±0.
15d
0.00
2(T
A98
)
DM
IP2-
Am
ino-
1,6-
dim
ethy
limid
azo[
4,5-
b]py
ridin
e13
2898
-04-
5
162.
2po
lar
pKa
8.16
±0.
30c
0.00
8(T
A15
38)
(Con
tinu
ed)
272 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Tabl
e1–
Cont
inue
d.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
1,5,
6TM
IP2-
Am
ino-
1,5,
6-tr
imet
hylim
idaz
o[4,
5-b]
pyrid
ine
1610
91-5
5-0
176.
2po
lar
pKa
8.66
±0.
30c
100
(TA
1538
)
3,5,
6TM
IP2-
Am
ino-
3,5,
6-tr
imet
hylim
idaz
o[4,
5-b]
pyrid
ine
5766
7-51
-3
176.
2po
lar
pKa
8.36
±0.
30c
na
IFP
2-A
min
o-1,
6-di
met
hyl-f
uro[
3,2-
e]im
idaz
o[4,
5-b]
pyrid
ine
3573
83-2
7-8
202.
3po
lar
pKa
7.52
±0.
40c
̴10(T
A15
38)
α-C
arbo
lines
Aα
C2-
Am
ino-
9H-p
yrid
o[2,
3-b]
indo
l26
148-
68-5
183.
2no
npol
arpK
a6.
79±
0.30
c
14.8
7±
0.40
d
0.3
(TA
98)0
.02
(TA
100)
MeA
αC
2-A
min
o-3-
met
hyl-9
H-p
yrid
o[2,
3-b]
indo
l68
006-
83-7
197.
2no
npol
arpK
a7.
08±
0.30
c
15.2
5±
0.40
d
0.2
(TA
98)0
.12
(TA
100)
(Con
tinu
ed)
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 273
Heterocyclic amines in cooked meat products . . .
Tabl
e1–
Cont
inue
d.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
β-C
arbo
lines
Har
man
1-m
ethy
l-9H
-pyr
ido[
4,3-
b]in
dole
486-
84-0
182.
2no
npol
arpK
a8.
62±
0.30
c
15.8
7±
0.40
d
co-m
utag
enic
Nor
harm
an9H
-pyr
ido[
4,3-
b]in
dole
244-
63-3
168.
2no
npol
arpK
a7.
85±
0.10
c
15.4
4±
0.30
d
co-m
utag
enic
γ-C
arbo
lines
Trp-
P-1
3-A
min
o-1,
4-di
met
hyl-5
H-p
yrid
o[4,
3-b]
indo
le62
450-
06-0
211.
3no
npol
arpK
a10
.88
±0.
10c
16.0
2±
0.40
d
39(T
A98
)1.7
(TA
100)
Trp-
P-2
3-A
min
o-1-
met
hyl-5
H-p
yrid
o[4,
3-b]
indo
le62
450-
07-1
197.
2no
npol
arpK
a10
.59
±0.
30c
15.5
9±
0.40
d
104.
2(T
A98
)1.8
(TA
100)
δ-C
arbo
lines
Glu
-P-1
2-A
min
o-6-
met
hyld
ipyr
ido[
1,2-
a:3’
2’-d
]imid
azol
e67
730-
11-4
198.
3no
npol
arpK
a6.
33±
0.30
c
49(T
A98
)3.2
(TA
100) (C
onti
nued
)
274 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .Ta
ble
1–Co
ntin
ued.
Stru
ctur
eA
bbre
viat
ion/
Chem
ical
nam
eCA
SN
oM
olec
ular
wei
ght
(g/
mol
)Pr
oper
ties
aM
utag
enic
ity
(×10³r
ev/μ
g)b
Glu
-P-2
2-A
min
o-di
pyrid
o[1,
2-a:
3’2’
-d]im
idaz
ole
6773
0-10
-3
184.
3no
npol
arpK
a5.
80±
0.30
c
1.9
(TA
98)1
.2(T
A10
0)
Oth
erpy
roly
ticH
AA
s
Phe-
P-1
2-A
min
o-5-
phen
ylpy
ridin
e33
421-
40-8
170.
2no
npol
arpK
a6.
32±
0.13
c
na
Orn
-P-1
4-A
min
o-6-
met
hyl-1
H-2
,5,1
0,10
b-te
traa
zaflu
oran
then
e78
859-
36-6
237.
3no
npol
arpK
a9.
52±
0.20
c
na
Cre-
P-1
4-A
min
o.1,
6-di
met
hyl-2
-met
hyla
min
o-1H
,6H
-pyr
rolo
-[3,4
-f]be
nzim
idaz
ole-
5,7-
dion
e13
3883
-91-
7
259.
3no
npol
arpK
a4.
83±
0.20
c
19(T
A98
)0.4
(TA
100)
Lys-
P-1
1,2,
3,8-
Tetr
ahyd
ro-c
yclo
pent
a[c]
pyrid
o[3
,2-a
]car
bazo
le69
477-
66-3
258.
3no
npol
arpK
a6.
87±
0.20
c
15.8
3±
0.20
d
na
aCa
lcul
ated
usin
gA
dvan
ced
Chem
istr
yD
evel
opm
ent(
ACD
/La
bs)S
oftw
are
V11
.02
(C©19
94–2
015
ACD
/La
bs).
bRe
fere
nces
:Sug
imur
aan
dot
hers
2004
;Wys
sand
Kadd
urah
-Dao
uk20
00;n
a,no
tana
lyze
d.c M
ostb
asic
(tem
pera
ture
25°C
).d
Mos
taci
dic
(tem
pera
ture
25°C
).
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 275
Heterocyclic amines in cooked meat products . . .
processes. Heterocyclic aromatic amines are similarly detected inrain water and cigarette smoke condensate (Xu and others 2010;Liu and others 2013). Individual HAAs were also detected in parti-cles from diesel exhaust fumes (Kataoka 1997). Over 25 mutagenicHAAs have been isolated and identified since 1977 (Alaejos andAfonso 2011) (Table 1). However, the concentrations detected infood vary widely, as will be discussed later.
The HAAs occurring most often in meat are PhIP, MeIQx,4,8-DiMeIQx, IQ, MeIQ, and AαC (Skog and others 1998a, b).The PhIP concentrations in most studies are between 1 and 70ng/g meat (Gross and Grueter 1992; Alaejos and Afonso 2011).Concentrations of MeIQx up to 23 ng/g can be detected (Skogand others 1995). As a rule, concentrations of up to 6 ng/g meat arefound (Skog and others 1998b). 4,8-DiMeIQx is mostly detectedonly in the lower concentration range of around 1 ng/g meat. Insome studies, IQ could not be detected at all (Skog and others1998b). Table 2 to 4 show an overview of the occurrence ofHAAs in beef patties, chicken breasts, and pork using differentpreparation methods.
While meat and meat products are frequently studied, fewerstudies were found about processed meat (Table 5). However, IQlevels of 3.8 to 10.5 ng/g were recorded in fried bacon (Johans-son and Jägerstad 1994). IQ and MeIQ could also be detected ingrilled sausages (Gerbl and others 2004). The HAA levels of ba-con prepared in different ways were investigated in a study (Sinhaand others 1998). Levels of 0.4 to 4.3 ng MeIQx/g and 0 to4.8 ng PhIP/g could be determined with pan-frying. MeIQxlevels of 0 to 4 ng/g with ovenroasting differed only a littlefrom frying. However, significantly higher PhIP levels of 1.4 to30.3 ng/g were detected for oven roasting (Sinha and others 1998).Further studies show that high HAA concentrations can be foundin bacon (Table 5). The use of various preparation methods affectsthe formation of HAAs differently for different meat products. Thegrill sausages investigated were pan-fried or grilled (Abdulkarimand Smith 1998). Along with different heating times, the heatingtemperature was also varied. With both cooking methods, alongwith norharman and harman, MeIQx and PhIP could also bedetected (Table 5). In offal products (beef liver, lamb kidney, andbeef tongue), which were thermally processed, HAAs were foundin concentrations only near the detection limit, except for norhar-man and harman. Both β-carbolines were found in concentrationsbelow 2 ng/g; only DMIP, MeIQx, and 4,8-DiMeIQx were de-tectable in cooked kidney and tongue up to 0.25 ng/g (Khan andothers 2009). The reason for the low contents of IQ-type HAAsmay be the lack of the precursor creatine (Harris and others 1997).The wide-ranging variation of precursors observed in studies withmeat from various animal species (Zimmerli and others 2001; Skogand Solyakov 2002; Sun and others 2010; Puangsombat and oth-ers 2012; Zaidi and others 2012), clearly affected the formation ofHAAs (Liao and others 2010, 2011b; Gibis and Weiss 2015).
Also the pH-value is known to influence the Maillard reactions(Cremer and Eichner 2000) and thereby also the formation ofHAAs. This was shown that the content of HAAs in pork increasedabout 22% on average (MeIQx – 33%; 4,8-DiMeIQx – 17%; andharman – 9%) in the PSE (pale, soft, and exudative) meat sampleswith lower pH values when grilled than the normal muscle at acore temperature of 95 °C (Polak and others 2009b). At the lowercore temperature of 70 °C, no remarkable difference in HAAformation could be observed between PSE meat and normal meat(Table 4). PSE is associated with pale color, low pH, and a highdrip loss causing by preslaughter stress and genetics of pigs witha fast post mortem glycolysis (Polak and others 2009b). In a study
investigating normal muscle meat of different animal species, themeasured pH values of the uncooked meat of the different animals(range of pH: 5.56 to 6.12) showed no significant linear correlationto the HAA levels, except for PhIP with a very weak correlation(r = 0.27, p < 0.05) (Gibis and Weiss 2015). Additionally, theduration of aging influenced the HAA concentrations for bothPSE and normal pork. Significantly higher levels were observedafter longer aging of pork (Polak and others 2009b) and beef (Polakand others 2009a).
Meat from horses showed threefold higher glucose levels thanbeef (Rossier 2003), which reduced the content of HAAs (Gibisand Weiss 2015). Similar results were observed in pork containinghigh and low levels of glucose (Olsson and others 2002). Contrar-ily, chicken had very low glucose levels, but a similar content ofcreatine, which increased, in particular, the concentration of PhIPby about a factor of 10 (Gibis and Weiss 2015). The same decreasein concentrations of PhIP, and less in levels of other HAAs, wasobserved in a model system by the addition of saccharides (Skogand Jägerstad 1990).
Formation of Heterocyclic Aromatic AminesFormation of imidazoquinolines and imidazoquinoxalines
The pyridines and pyrazines formed from hexoses and aminoacids, respectively, in the Maillard reaction via the Strecker degra-dation serve as building blocks for the IQ compounds. The reac-tion is depicted in Figure 1A. The creatine cyclizes to creatinineduring heating and reacts in an aldol reaction with the pyridineor pyrazine derivatives, respectively, to generate imidazoquinolineand imidazoquinoxaline (Skog and Jägerstad 1993). The aldehy-des arising, together with creatinine, also play an important rolein the formation of the imidazole rings of the polar HAAs. The2 parts can be joined to each other via a Strecker aldehyde to aSchiff base. The mechanism was confirmed for IQx, MeIQx, and4,8-DiMeIQx by using 14C-labeled glucose (Skog and Jägerstad2005).
Formation of PhIPPhenylalanine, reducing sugars and creatinine could be detected
in PhIP formation. It could be shown in a model trial with radioac-tively labeled carbon in the phenylalanine molecule that the phenylring was completely built into the PhIP molecule (Zöchling andMurkovic 2002). Further trials showed that creatinine forms apart of the imidazole ring. The authors were able to identifythe following reaction steps in PhIP formation in a model trial:First, phenylacetaldehyde is formed from phenylalanine via theStrecker degradation. The phenylacetaldehyde formed reacts inan aldol reaction with creatinine to form an intermediate prod-uct. In the subsequent condensation reaction, PhIP arose from thissubstance (Zöchling and Murkovic 2002). The mechanism of thereaction is shown in Figure 1B. The formation of formaldehydefrom phenylacetaldehyde and phenylalanine, and the combinationof both formaldehyde and ammonia in the generation of PhIPfrom phenylacetaldehyde and creatinine were reported in the re-action pathways that produce PhIP (Zamora and others 2014). Inthe presence of oxidized lipid, other amino acids competed withphenylalanine for the lipid, and amino acid degradation productswere formed, among which α-keto acids seemed to play a role inthese reactions (Zamora and others 2013b). However, unoxidizedlipids did not contribute to PhIP formation (Zamora and others2012).
276 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Tabl
e2–
Occ
urre
nce
ofH
AA
sin
grou
ndbe
efpa
ttie
sus
ing
diff
eren
the
atin
gco
ndit
ions
.
Cook
ing
Cook
ing
tim
eTe
mpe
ratu
rePh
IPM
eIQ
x4,
8-D
i-MeI
Qx
Nor
harm
anH
arm
anO
ther
spr
oced
ure
(min
)(°
C)(n
g/g)
b(n
g/g)
b(n
g/g)
b(n
g/g)
c(n
g/g)
c(n
g/g)
bcRe
fere
nce
Frie
d4–
2015
0nd
–1.8
nd–0
.6nd
1.1
na(K
nize
and
othe
rs19
94)
190
nd–9
.80.
1–1.
30.
40.
15na
230
1.3–
320.
4–7.
3nd
1.6
naFr
ied
1219
84.
94.
31.
3na
naAα
C(2
1)(T
hiéb
aud
and
othe
rs19
95)
277
6816
4.5
Frie
d8
165
0.08
0.2
ndna
na(J
ohan
sson
and
othe
rs19
95b)
200
1.5
1.6
0.4
Frie
d5–
715
00.
01nd
ndna
na(S
kog
and
othe
rs19
95)
225
1.1
2.2
0.8
Frie
d–
–67
.516
.44.
5na
na(K
nize
and
othe
rs19
97a)
Gril
led
––
502
ndBa
rbec
ued
6+
620
02.
350.
27nd
1.87
0.61
Trp-
P-2
(1.7
)(A
bdul
karim
and
Smith
1998
)(1
5%fa
t)3.
5+
3.5
240
0.10
0.80
nd2.
170.
88Tr
p-P-
2(1
.5)
Frie
d3
+3
150
0.43
nd0.
310.
960.
31(1
5%fa
t)5
+5
190
0.57
0.84
2.0
0.83
7.5
+7.
523
00.
251.
005.
651.
7Fr
ied
12–2
017
50.
9–6.
20.
5–0.
80.
8–0.
9na
naIQ
(0.7
-1.3
),M
eIQ
(0.1
-0.3
)(B
alog
han
dot
hers
2000
)20
04.
0–25
.41.
5–4.
20.
9–4.
5IQ
(1.7
-4.4
),M
eIQ
(0.5
-2.1
)22
513
.3–1
.43.
5–5.
83.
0–4.
8IQ
(2.8
-5.3
),M
eIQ
(2-3
.5)
Frie
d–
–nd
–1.5
nd–1
.3nd
nana
(Zim
mer
lian
dot
hers
2001
)Fr
ied
323
00.
21.
0nd
2.1
1.5
(Jau
tzan
dot
hers
2008
)4.
523
00.
92.
00.
45.
03.
56
230
3.8
4.8
3.0
10.4
8.9
Frie
d/10
180
0.7
1.0
0.2
nana
dIQ
(0.1
/0.
05),
IFP
(0.1
/nd
)(N
iand
othe
rs20
08)
Ove
n-br
oile
d10
186
ndnd
ndIQ
[4,5
-b](
0.3/
nd),
IgQ
x(0
.5/
nd),
7-M
eIgQ
x(2
.4/
0.3)
,6,7
-DiM
eIgQ
x(0
.2/
0.05
),7,
9-D
iMeI
gQx
(0.7
/nd
)Fr
ied/
1518
92.
73.
00.
6na
nad
IQ(0
.1/
0.06
),IF
P(0
.6/
nd)
IQ[4
,5-b
](0.
3/0.
4),I
gQx
Ove
n-br
oile
d15
189
0.06
0.02
0.02
(1.5
/0.
03),
7-M
eIgQ
x(9
.5/
0.1)
,6,7
-DiM
eIgQ
x(0
.3/
nd),
7,9-
DiM
eIgQ
x(2
.2/
0.02
)Fr
ied/
2019
12.
93.
70.
7na
nad
IQ(0
.2/
0.1)
,IFP
(0.7
/0.
14)
IQ[4
,5-b
](0.
3/0.
15),
IgQ
xO
ven-
broi
led
2019
11.
230.
380.
1(1
.8/
0.3)
,7-M
eIgQ
x(1
1.7/
0.3)
,6,7
-DiM
eIgQ
x(0
.4/
nd),
7,9-
DiM
eIgQ
x(3
.0/
0.12
)Fr
ieda
4.5
230
2.6
4.9
1.8
13.5
21.4
(Gib
is20
09)
Frie
da2-
3̴19
00.
1–0.
20.
2–1.
7nd
–0.2
0.2–
0.9
0.7–
1.7
(Gib
isan
dW
eiss
2010
)
aFr
ied/
grill
edon
ado
uble
-pla
tegr
ill;b
ndno
tdet
ecte
d;c n
ano
tana
lyze
d;d
HA
Ang
/g
(frie
d/ov
en-b
roile
d).
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 277
Heterocyclic amines in cooked meat products . . .
Tabl
e3–
Occ
urre
nce
ofH
AA
sin
chic
ken
brea
st.
Cook
ing
proc
edur
eCo
okin
gti
me
(min
)
Cook
ing
tem
pera
ture
(°C)
PhIP
(ng/
g)c
MeI
Qx
(ng/
g)bc
4,8-
DiM
eIQ
x(n
g/g)
bcN
orha
rman
(ng/
g)b
Har
man
(ng/
g)c
Oth
ers
(ng/
g)Re
fere
nce
Pan-
frie
d14
–36
197–
221
12–7
01–
31–
4na
na(S
inha
and
othe
rs19
95)
Barb
ecue
d10
–43
177–
260
27–4
80nd
–9nd
–2na
naBo
iled
9–17
79–8
66–
150
nd-3
ndna
naPa
n-fr
ied
1617
50.
7na
nana
na(P
erss
onan
dot
hers
2002
)18
200
10.5
nana
nana
1222
529
.7na
nana
naPa
n-fr
ied
12–3
414
0–22
5<
0.1–
38.2
0.1–
1.8
0.1–
0.6
0.5–
6.9
0.3–
7.5
(Sol
yako
van
dSk
og20
02)
Roas
ted
24–4
017
5–24
0nd
ndnd
<0.
1–3.
3<
0.1–
1.7
Dee
p-fa
tfrie
d11
160
<0.
1a<
0.1a
<0.
1a0.
50.
3Bo
iled
3820
0nd
nd0.
1a<
0.1a
<0.
1aPa
n-fr
ied
14–3
619
7–21
18.
8–48
.50.
6–2.
30.
8–3.
6na
naIQ
(0.1
-0.2
),IQ
x(0.
1-0.
2),
IFP(
2.7-
15.9
),IQ
[4,5
-b](
a-0
.2),
IgQ
x(0.
3-0.
9),
(Nia
ndot
hers
2008
)
7-M
eIgQ
x(1.
5-8.
7),
6.7-
DiM
eIgQ
x(a
-0.1
),7.
9-D
iMeI
gQx(
0.2-
0.9)
,Aα
C(a
-0.1
),M
eAα
CaO
ven-
broi
led
9–17
179–
186
5.6–
72.0
0.1–
2.8
0.1–
2.0
nana
IQ(0
.2-0
.3),
IQxa
,IF
P(0.
4-11
.3),
IQ[4
,5-b
](a
-0.3
),Ig
Qx(
0.2-
0.4)
,7-
MeI
gQx(
1.0-
11.1
),6.
7-D
iMeI
gQx(
a-0
.2),
7.9-
DiM
eIgQ
x(0.
1-1.
8),
Aα
C(0.
2-9,
4),
MeA
αC(
a-1
.3)
Frie
dd5
220
2.4
2.1
0.11
0.07
0.09
(Gaš
perli
nan
dot
hers
2009
)G
rille
de18
220
0.3
0.2
0.13
0.11
0.05
Frie
dd4.
523
03.
80.
2nd
0.8
0.9
(Gib
is20
09)
Pan-
frie
d5
+5
180
18.3
1.8
1.1
1.4
2.8
IQ(1
.8),
Aα
C(0.
2),
MeA
αCa
(Lia
oan
dot
hers
2010
)
Dee
p-fa
tfrie
d10
180
2.2
0.8
0.4
5.4
12.3
Aα
C(0.
3),M
eAα
CaCh
arco
al-g
rille
d10
+10
200
31.1
1.2
3.6
32.2
31.7
Trp-
P-2(
3.6)
,Trp
-P-1
(1.5
),Aα
C(5.
6),M
eAα
C(1.
6)Ro
aste
d20
200
0.04
ndnd
3.1
0.7
Aα
Ca,M
eAα
Ca
aTr
aces
(�0.
05ng
/g)
.b
na,n
otan
alyz
ed.
c nd,
notd
etec
ted.
dFr
ied
ona
doub
le-p
late
grill
.e
Gril
led
onan
infr
ared
grill
.
278 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Tabl
e4–
Occ
urre
nce
ofH
AA
sin
pork
.
Mat
eria
lCo
okin
gpr
oced
ure
Cook
ing
tim
e(m
in)
Cook
ing
tem
pera
ture
(°C)
PhIP
(ng/
g)a
MeI
Qx
(ng/
g)a
4,8-
DiM
eIQ
x(n
g/g)
aN
orha
rman
(ng/
g)b
Har
man
(ng/
g)a
bO
ther
s(n
g/g)
bRe
fere
nce
Pork
chop
Frie
d8–
9.5
150–
225
nd–4
.8nd
–2.6
nd–1
.1na
na(S
kog
and
othe
rs19
95)
Pork
chop
Frie
d5–
1517
5nd
nd–3
.8nd
nana
(Sin
haan
dot
hers
1998
)Po
rkpa
ttie
s(70
g)RN
− alle
leFr
ied
3+
320
01.
91.
90.
4na
na(O
lsso
nan
dot
hers
2002
)N
onRN
− alle
leFr
ied
0.2
1.5
0.2
nana
Pork
chop
sRN
− alle
lepH
5.32
Pan-
frie
d3
+3
160–
200
0.05
–0.1
0.1–
0.2
nd0.
6–1.
71.
1–0.
7IQ
x(0
.1–0
.2)
(Ols
son
and
othe
rs20
05)
Non
RN− a
llele
pH5.
58Pa
n-fr
ied
0.1–
3.3
0.1–
0.8
ndnd
ndIQ
x(n
d)Po
rkpa
ttie
sBo
iled
8–16
100
nd0.
4–1.
0nd
nana
(Shi
n20
05)
Broi
led
12–1
917
7–22
5nd
–2.7
1.2–
1.6
0.3–
0.7
nana
Pan-
Frie
d9–
2117
7–22
50.
3—10
–50.
6–5.
00.
3–1.
7na
naPo
rkst
eakc
Pan-
frie
d22
00.
8–3.
10.
6–4.
60.
2–1.
00.
2–0.
50.
9–0.
2(P
olak
and
othe
rs20
09b)
pH5.
48co
rete
mp
pH5.
6070
–95
0.8–
2.7
0.9–
3.1
0.2–
0.9
0.2–
0.5
0.9–
0.2
Pork
top
loin
Pan–
frie
d8
+8
204
1.8
1.1
1.2
nana
IQx
(nd)
(Pua
ngso
mba
tand
othe
rs20
12)
Bake
d70
2.2
0.2
0.9
nana
IQx
(nd)
Pork
patt
yFr
ied
(50
g)5
180
18.4
3.2
0.7
nana
(Zha
ngan
dot
hers
2013
)Po
rkm
eat-
ball
5.3
1.0
0.3
nana
Pork
strip
5.4
1.0
0.3
nana
Pork
loin
Pan-
frie
d20
4co
rete
mp
7713
.17.
61.
6na
na(V
angn
aian
dot
hers
2014
)
Pork
patt
ies(
80g)
pH5.
56G
rille
dd2.
722
02.
31.
10.
51.
10.
5(G
ibis
and
Wei
ss20
15)
ana
nota
naly
sed,
bnd
notd
etec
ted.
c M.l
ongi
ssim
usdo
rsi
dfr
ied
ona
doub
le-p
late
grill
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 279
Heterocyclic amines in cooked meat products . . .
Tabl
e5–
Com
pari
son
ofth
eco
nten
tof
the
mos
tco
mm
onH
AA
sin
cook
edba
con
and
proc
esse
dm
eat.
Hea
ting
cond
itio
nH
AA
a
Prep
arat
ion
met
hod
Tim
e(m
in)
Tem
p.(°
C)M
eIQ
x(n
g/g)
4,8-
DiM
eIQ
x(n
g/g)
PhIP
(ng/
g)N
orha
rman
(ng/
g)H
arm
an(n
g/g)
Refe
renc
e
Baco
nfr
ied
12–1
617
00.
9–27
0.5–
2.4
nd–5
2nd
–22
nd–3
0(G
ross
and
othe
rs19
93)
Baco
nm
icro
-wav
ed(6
00W
)3
0.1
ndnd
3.3
ndBa
con
pan-
frie
d5
150
2.8
3.4
0.2
nana
(Joh
anss
onan
dJä
gers
tad
1994
)Ba
con
pan-
frie
d4–
1617
6–17
70.
4–4.
3<
0.2
<0.
2–4.
8na
na(S
inha
and
othe
rs19
98)
Baco
nov
en-b
roile
d4–
717
5–18
50.
2–4
<0.
21.
4–30
.3na
naBa
con
mic
ro-w
aved
1.8–
3.3
<0.
2–1.
5<
0.2
<0.
2–3.
1na
naBa
con
pan-
frie
d1.
523
08.
14.
528
.4na
na(G
uyan
dot
hers
2000
)Ba
con
grill
ed1.
523
01.
60.
95.
0na
naBa
con
pan-
frie
d16
.117
63.
00.
74.
9na
na(N
iand
othe
rs20
08)
Baco
nov
en-b
roile
d7.
517
52.
65.
215
.9na
naBa
con
pan-
frie
d3
204
4.0
3.6
6.9
nana
(Pua
ngso
mba
tand
othe
rs20
12)
Baco
npa
n-fr
ied
3–6
160
1.5–
4.9
nd0.
1–1.
15–
14.1
0.4–
1.7
(Gib
isan
dot
hers
2015
)2–
321
02.
4–5.
6nd
1.3–
2.6
13.7
–19.
91.
3–1.
6Fa
lun
saus
age
frie
d5
160
0.6
ndnd
nana
(Joh
anss
onan
dJä
gers
tad
1994
)Sa
usag
efr
ied
616
00.
70.
20.
1na
naH
ampa
n-fr
ied
5–19
175
<0.
2–1.
8<
0.2
<0.
2–0.
3na
na(K
nize
and
othe
rs19
97b)
Hot
dogs
pan-
frie
d4–
1817
5–17
7nd
ndnd
nana
(Sin
haan
dot
hers
1998
)H
otdo
gsov
en-b
roile
d3–
1018
0–18
5nd
ndnd
nana
Hot
dogs
grill
/ba
rbec
ued
5–15
232–
252
ndnd
ndna
naPo
rksa
usag
efr
ied
6–15
150–
230
nd–0
.7nd
nd–1
.1nd
–0.8
nd–3
.1(A
bdul
karim
and
Smith
1998
)Po
rksa
usag
eba
rbec
ued
12/
720
0/24
00.
35/
0.8
nd0.
1/1.
36.
1/1.
160.
57/
4.2
Saus
age
frie
d9
175–
200
<0.
2<
0.5
<0.
10.
30.
3(B
usqu
etsa
ndot
hers
2004
)Po
rksa
usag
epa
ttie
sfrie
d21
179
5.1
0.7
0.2
<0.
3<
0.3
(Nia
ndot
hers
2008
)Pi
zza
topp
ing
sala
mio
ven-
bake
d12
–20
230
nd–2
.6nd
nd–0
.410
7.4–
186.
111
.4–2
4.7
(Gib
isan
dW
eiss
2013
)10
–12
250
nd–0
.2nd
0.1–
0.3
143.
2–14
6.1
15.3
–21.
0Pi
zza
topp
ing
ham
oven
-bak
ed12
–20
230
0.2–
3.1
0.5–
2.1
0.2–
0.8
4.5–
10.3
2.5–
4.8
10–1
225
00.
1–0.
20.
5–0.
60.
3–0.
55.
6–7.
04.
3–4.
9a
na,n
otan
alyz
ed;n
d,no
tdet
ecte
d.
280 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Tabl
e6–
Effe
ctof
diff
eren
tan
tiox
idan
tin
gred
ient
son
the
form
atio
nof
HA
As
infr
ied
orgr
illed
beef
prod
ucts
,inh
ibit
ion
ofH
AA
form
atio
n(%
)and
incr
ease
ofH
AA
leve
lsco
mpa
red
toth
eco
ntro
lsw
itho
utan
tiox
idan
tco
mpo
nent
sas
indi
cate
dby
+;na
indi
cate
sno
tan
alyz
ed;n
din
dica
tes
not
dete
cted
.
Prod
uct
Ingr
edie
ntCo
ncen
trat
ion
Cook
ing
tim
e(m
in)
Cook
ing
tem
p.(°
C)Ph
IP(%
)M
eIQ
x(%
)
4,8-
DiM
eIQ
x(%
)N
orha
rman
(%)
Har
man
(%)
Oth
ers
(%)
Refe
renc
e
Beef
patt
yCh
erry
tissu
e11
.5%
8+
817
093
6281
nana
IQ(7
2),M
eIQ
(50)
(Brit
tand
othe
rs19
98)
Beef
stea
kRo
sem
ary
Spre
ada
2018
075
3839
nana
IQ(7
2),M
eIQ
(64)
(Mur
kovi
can
dot
hers
1998
)G
arlic
Spre
ada
5471
78na
naIQ
(32)
,MeI
Q(4
0)Sa
geSp
read
a10
040
100
nana
IQ(1
00),
MeI
Q(7
7)Th
yme
Spre
ada
7561
100
nana
IQ(7
4),M
eIQ
(61)
Beef
patt
yO
leor
esin
rose
mar
y1%
10+
1022
544
3077
nana
IQ(7
2),M
eIQ
(87)
(Bal
ogh
and
othe
rs20
00)
10%
4512
68na
naIQ
(72)
,MeI
Q(7
2)V
itam
inE
1%69
4879
nana
IQ(8
6),M
eIQ
(79)
10%
7226
71na
naIQ
(88)
,MeI
Q(6
4)Be
efpa
tty
(100
g;15
.4%
fat)
Min
ced
garli
c4.
5%10
+10
225
3413
24na
na(S
hin
and
othe
rs20
02a)
18.2
%,
7162
62na
naSu
lfurb
Com
p.0.
17m
M51
2846
nana
0.67
mM
8266
81na
na1.
01m
M90
7081
nana
Beef
patt
yV
irgin
oliv
eoi
lfre
sh40
g5
+5
200
605
nd50
40(P
erss
onan
dot
hers
2003
a)(9
0g)
Stor
edfo
r1ye
ar75
65nd
800
Beef
patt
yG
rape
-see
d0.
5%10
+10
210
126
100
48+3
493
IQ(1
00),
MeI
Q(1
7),
Aα
AC(
29)
(Ahn
and
Gru
en,2
005b
)
1%25
6410
064
+694
9IQ
(100
),M
eIQ
(34)
,Aα
AC(
43)
Pine
bark
0.5%
3747
100
5528
IQ(1
00),
MeI
Q(1
00),
Aα
AC(
+11)
1%36
6210
061
32IQ
(100
),M
eIQ
(100
),Aα
AC(
29)
Ole
ores
inro
sem
ary
0.5%
396
2449
23IQ
(100
),M
eIQ
(26)
,Aα
AC(
100)
1%58
2310
010
055
IQ(1
00),
MeI
Q(1
00),
Aα
AC(
100
BHA
/BH
T0.
02%
12+3
1552
27IQ
(100
),M
eIQ
(4),
Aα
AC(
17)
Beef
patt
yG
rape
-see
d0.
1%6
+6
210
7267
66na
na(C
heng
and
othe
rs20
07)
App
le0.
1%69
5963
nana
Elde
rber
ry0.
1%45
619
nana
Pine
appl
e0.
1%13
2718
nana
Beef
patt
yCa
rvac
rol
1%20
0co
rete
mp
7078
72nd
nana
MeI
Q(5
8)(F
riedm
anan
dot
hers
2009
)
Beef
patt
y(1
00g)
Min
ced
garli
c4.
8%5
+5
220
1651
nd6
6Tr
p-P-
1(15
),Tr
p-P-
2,G
lu-P
-1,&
Glu
-P-2
(100
),Aα
AC
(+39
)
(Jun
gan
dot
hers
2010
)
13%
8310
0nd
9688
Trp-
P-1(
100)
,Trp
-P-2
,G
lu-P
-1,&
Glu
-P-2
(100
),Aα
AC
(100
)Be
efpa
tty
(70
g)H
ibis
cusm
arin
adec
0.2%
2.7
230
1836
nd+1
+2(G
ibis
and
Wei
ss20
10)
0.8%
6456
nd+6
6+3
3
(Con
tinu
ed)
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 281
Heterocyclic amines in cooked meat products . . .
Tabl
e6–
Cont
inue
d.
Prod
uct
Ingr
edie
ntCo
ncen
trat
ion
Cook
ing
tim
e(m
in)
Cook
ing
tem
p.(°
C)Ph
IP(%
)M
eIQ
x(%
)
4,8-
DiM
eIQ
x(%
)N
orha
rman
(%)
Har
man
(%)
Oth
ers
(%)
Refe
renc
e
Beef
patt
yG
alan
gal
0.2%
5+
520
419
18na
nana
(Pua
ngso
mba
tand
othe
rs20
11)
(100
g)Fi
nger
root
0.2%
3731
nana
naTu
rmer
ic0.
2%38
41na
nana
Cum
in0.
2%6
1na
nana
Coria
nder
seed
s0.
2%6
3na
nana
Rose
mar
y0.
2%36
50na
nana
Beef
patt
y(7
0g,
froz
en)
Gra
pe-s
eedd
0.2%
2.7
230
5012
nd+3
+6(G
ibis
and
Wei
ss20
12)
0.8%
7193
nd+2
5+5
5Ro
sem
ary
extr
acte
0.12
%35
16nd
+34
+55
1.5%
9154
nd+5
03
Beef
patt
y(3
0g,
6.2×
1.2
cm)
Asc
orbi
cac
id0.
02m
M3
+3
200
1917
14na
na(W
ong
and
othe
rs20
12)
Nia
cin
1919
15na
naPy
rido
xam
ine
4342
38na
na
Beef
stea
k(1
.4cm
thic
k)
Beer
50%
3+
318
050
>99
>98
nana
IQ(>
97)
(Vie
gasa
ndot
hers
2012
)Be
er+
spic
esf
91>
7664
nana
IQ(>
97)
Win
e+1
>76
70na
naIQ
(>97
)W
ine
+sp
ices
f32
3364
nana
IQ(7
2)D
Ag
Beer
+sp
ices
f61
52>
98na
naIQ
(>97
)D
Ag
Win
e+
spic
esf
7751
>98
nana
IQ(>
97)
Beef
patt
y(6
2.5
g,10
mm
thic
k)
Soy
leci
thin
h1%
L2.
722
0+2
047
nd+2
+21
(Nat
ale
and
othe
rs20
14)
5%L
+90
26nd
+8+1
6So
yle
cith
inh+
GSE
i1%
L+
0.1%
GSE
5047
nd1
+24
5%L+
0.1%
GSE
7037
nd55
471%
L+
0.2%
GSE
1042
nd2
+21
5%L+
0.2%
GSE
+40
26nd
+20
+24
Pure
GSE
i0.
1%50
37nd
313
0.2%
2032
nd+2
3+5
aSp
ices
spre
adon
surf
ace.
bO
rgan
osul
furc
ompo
unds
:dia
llyld
isul
fide,
dipr
opyl
disu
lfide
,dia
llyls
ulfit
e,al
lylm
ethy
lsul
fide,
ally
lmer
capt
an,c
yste
ine,
cyst
eine
.c P
lant
extr
acts
,.5
gof
mar
inad
eus
edon
each
side
ofth
epa
tty.
dG
rape
-see
dor
Hib
iscu
sext
ract
inw
ater
-insu
nflow
eroi
lem
ulsi
on(m
arin
ade)
(em
ulsi
fierE
472c
:citr
icac
ides
ters
ofm
ono
and
digl
ycer
ides
);1.
5g
ofm
arin
ade
used
onea
chsi
de.
eRo
sem
ary
extr
acti
nsu
nflow
eroi
lmar
inad
e;1.
5g
ofm
arin
ade
used
onea
chsi
de.
f Mar
inad
eco
ntai
ning
ging
er(2
.8%
),ga
rlic
(2.9
%),
rose
mar
y(0
.4%
),th
yme
(0.2
5%
),an
dre
dch
ilipe
pper
(0.1
%,w
/v)
.g
DA
,dea
lcoh
oliz
edw
ine
orbe
er.
hSo
yle
cith
inco
ntai
ning
0.18
%D
L-α
-toc
ophe
rols
hom
ogen
ized
at22
500
psit
olip
osom
es(L
)use
das
mar
inad
e(1
0g,
mar
inat
ing
time
3h
at8°C
)com
pare
dto
cont
rol(
rape
seed
oilr
efine
d).
i Pur
eor
inlip
osom
esin
corp
orat
edG
SE(g
rape
-see
dex
trac
t).
282 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Figure 1–Formation of heterocyclic amines: (A) mechanism of thermophilic HAAs with creatine, sugar, and amino acids (modified) (Jägerstad andothers 1983), (B) mechanism of formation of PhIP without sugar (modified) (Zöchling and Murkovic 2002), and (C) formation of norharman asβ-carboline (modified) (Roenner and others 2000).
Formation of β-carbolineThe reaction of the formation of β-carboline norharman is
shown in Figure 1C (Roenner and others 2000). A similarreaction is postulated for both β-carbolines (norharman and har-man). Both substances have no free amino group, which causes nomutagenicity in the Ames test (Pfau and Skog 2004). In compari-son to other carbolines (pyrolytic HAAs), the reaction of norhar-man and harman could be formed at lower temperatures and theyoccurred in cooked meat, fish, and meat extract (Ziegenhagen
and others 1999) as well as in processed meat, particularly suchpizza toppings as salami and cooked ham (Gibis and Weiss 2013).Tryptophan, in particular, was found as a precursor and glucoseenhanced the formation (Pfau and Skog 2004). Under the samecondition with an excess of tryptophan, harman and norharmanwere found at levels 70 and 20 times higher, respectively (Skog andothers 2000). The other carbolines are normally formed as typicalpyrolysis products of amino acids above a temperature of 300 °C(Jägerstad and others 1998).
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 283
Heterocyclic amines in cooked meat products . . .
Precursors of Heterocyclic Aromatic AminesCreatine and creatinine
Creatine, as a nonproteinogenic amino acid, occurs in the formof creatine phosphate in the muscle of vertebrates and serves as anenergy store. Heating transforms it into the physiologically inertcyclization product creatinine (Murkovic and others 1999). Theresulting creatinine is needed to form the imidazole ring. If nocreatinine is available, no imidazoquinoline and imidazoquinox-aline are formed. By contrast, the formation of nonpolar HAAsis independent of the presence of creatinine (Murkovic 2004b).The longer the heating time, the more creatine is converted tocreatinine. The higher the addition of creatine/creatinine duringthe heat treatment of meat and fish, the higher is the mutagenicityof the products (Felton and Knize 1991). A promoting indicationthat creatine is significantly involved in the formation of thermicHAAs is that there are only small amounts of creatine in protein-rich foods, such as cheese, beans, liver, and shrimp. Only very smallamounts of mutagenic activity could be detected in these productsafter preparation (Cheng and others 2006; Chen and others 1990).The molar ratios of total creatine/glucose was found from 0.89 upto 9.84 for beef, pork, mutton, chicken, duck, and goose (Liao andothers 2011b) and 1.2 up to 12 for veal, beef, pork, lamb, horse,turkey, and ostrich, respectively, except for vension and chickenwith ratios up to 60.7 (Gibis and Weiss 2015) in meat from dif-ferent animals. Trials with a model system showed that creatinineforms a part of the PhIP molecule (Zöchling and Murkovic 2002).It was shown in a further trial that beef with a high creatine level(1.5 mg/g) demonstrated a higher mutagenic activity than beefwith a low creatine level (Jackson and others 1994).
Free amino acidsFree amino acids are, similar to creatine, necessary for the
formation of HAAs. However, several different amino acids canform the same mutagens, for example, threonine, glycine, lysine,alanine, and serine all lead to MeIQx formation (Jägerstad andSkog 1991). Several amino acids were heated in the presence ofcreatinine and glucose in model trials and all the amino acidsshowed mutagenic activity (Johansson and others 1995a). Otherauthors showed that some amino acids had less mutagenic activitywith dry heating. According to this study, the amino acids serineand threonine had the highest mutagenic activity (Overvik andothers 1989). Which mutagens are formed depends on the aminoacids present. PhIP is formed from the amino acid phenylalanine(Manabe and others 1992, 1993). Other authors determined,using 13C NMR spectrometry, that a 13C-labeled phenylalaninemolecule is built into the structure of the PhIP molecule(Murkovic and others 1999). In model studies, phenylalanine, as aknown precursor for PhIP, is highest in the chicken model system,and the compound is also formed from tyrosine and isoleucine,which are similarly highest in chicken (Skog and others 1998b).However, no correlation between the amount of phenylalanineand the levels of PhIP was observed in pork (Olsson and others2002). MeIQx, by contrast, arose from the amino acids glycine,threonine, alanine and lysine (Johansson and Jägerstad 1996;Johansson and others 1995a). In the model, the amino acidsthreonine, alanine, and lysine took part in the formation of4,8-DiMeIQx (Johansson and others 1995a). The amino acidsglycine, alanine, and lysine were responsible for the formationof 7,8-DiMeIQx. With the exception of L-cysteine, whichreduced the mutagenicity in the Ames test, all amino acidsincreased the mutagenic activity (Lee and others 1994). When4-oxo-2-nonenal, a compound of oxidization of fat, was present,only the addition of methionine, glycine, and serine significantly
increased the amount of PhIP formed (p < 0.05), whereascysteine, lysine, tryptophan, histidine, tyrosine, and alaninereduced (p < 0.05) PhIP significantly (Zamora and others 2013b).The generation of PhIP is also associated with the capacityof 4-oxo-2-nonenal to constitute the Strecker degradation ofphenylalanine to phenylacetaldehyde (Zamora and others 2013a).Amino acids also formed HAA amino acid adducts, for example,the PhIP adduct with glycine was formed easily within 5 min byheating at 200 °C, which is probably based on the dehydrationcondensation of the amino group of PhIP and carboxyl groupof glycine (Kataoka and others 2012). The content of free aminoacids as the sum of all encoded and nonencoded amino acids is nothighly correlated with the formation of HAAs in grilled beef; butderivatives of amino acid, such as creatinine, and the amino acidsalanine, phenylalanine, and lysine are related to the formation ofaminoimidazoarenes and PhIP, while the other amino acids do notparticipate in the formation of HAAs (Szterk and Waszkiewicz-Robak 2014). Similar findings were observed in pork, beef, andchicken, only lysine, tyrosine, phenylalanine, proline, isoleucine,and aspartic acid showed significantly positive correlations to levelsof PhIP, while no significant correlation between any amino acidand contents of other HAAs was found (Gibis and Weiss 2015).
CarbohydratesThe presence of reducing sugars plays an important role in the
formation of HAAs. Along with creatine, creatinine, and aminoacids, reducing sugars such as glucose, fructose, ribose, erythrose,and lactose are also necessary for the formation of these muta-gens (Skog and Jägerstad 1993). Glucose was necessary for form-ing mutagenic activity in the Ames test both in aqueous modelsystems and with the heating of meat samples (Skog and others1992). The sugars occurring naturally in meat are present in ap-proximately half the molar quantity of creatine and amino acids.If the sugar concentration is increased beyond the naturally oc-curring ratio, a reduction is observed in HAA formation (Skogand Jägerstad 1990). The same was observed for glucose, fructose,sucrose, and lactose as for sugar, however, the monosaccharidesshowed the most distinct inhibitory effects (Skog and Jägerstad1990). Some authors found higher ratios of total creatine/glucosewith values from 0.89 up to 9.84 (Liao and others 2011b) andup to 60.7, respectively, (Gibis and Weiss 2015) for animals ofdifferent species. These ratios of the precursors to each other alsoplay a key role in the formation of HAAs. Higher levels werefound for MeIQx by adding lactose in model systems (molecularratio lactose/amino acids/creatinine – 2:1:1) with 5% water con-tent (Dennis and others 2015). The inhibition of the formationof the mutagenic compounds by an excess of sugars is proposedto be an effect of Maillard reaction products, which may blockthe formation of imidazoquinoxalines by attacking creatine (Skogand Jägerstad 1990). Therefore, the reason is due to the formationof Maillard products which changes the reaction path and resultsin a reduction in HAA concentration (Shin and others 2002c). Asignificant reduction in mutagenicity and in the content of MeIQ,PhIP, DiMeIQx, IQ, IQx, and norharman was attained in grilledchicken that was marinated with honey containing glucose andfructose in the same ratio and a small amount of sucrose (Hasnoland others 2014; Shin and Ustunol 2004). It was also observed in astudy with pork, from pigs as carriers of the RN allele, containingincreased concentrations of residual glycogen that this fact causedabout 90% lower levels of PhIP and 50% lower of total mutagenicHAAs in cooked meat compared with cooked meat from normalpigs (Olsson and others 2002).
284 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
Different oligosaccharides, such as fructooligosaccharide, galac-tooligosaccharide, isomaltooligosaccharide, or inulin, reduced theHAA formation in cooked meat products (Shin and others 2003,2004; Persson and others 2004) . The addition of oligosaccharidesaffected the mass transfer of the precursors to the surface, whichresults in higher water-holding capacity as well as thereby a lowerweight loss of cooked products and reduction of mass transfers ofprecursors (Persson and others 2004).
Physical FactorsHeating time and heating temperature
The formation of HAAs depends on the heating time and theheating temperature depending on the used cooking method. Theheating temperature has the largest effect on HAA formation.The formation of mutagenic substances begins at around 125 °C(Jägerstad and others 1998; Skog and others 1998b). Howeveraccording to Ahn and Gruen (2005a), few or no HAAs were de-tected at temperatures (160 °C) and a heating time of 15 min inground beef as a model systems with glass test tubes. Only afterincreasing the heating temperature and time, the HAA concen-tration rose. At 180, 200, and 220 °C, significant concentrationsof polar HAAs (IQ, MeIQ, MeIQx, 4,8-DiMeIQx, and PhIP)and nonpolar HAAs (harman, norharman, and AαC) were deter-mined after 20 min. By contrast, neither MeIQ nor 4,8-DiMeIQxcould be detected at 180 °C and a cooking time of 10 min. TheMeIQ concentrations detected were under 2 ng/g at all 3 tem-peratures (Ahn and Gruen 2005a). The HAA concentrations inthe trials performed increased only after about 5 min, but thendisproportionally. This observation could be explained by the sur-face temperature being under 100 °C in the first few minutes ofthe heating process. The temperature range that is required forthe formation of the mutagenic substances was only reached afterabout 5 min (Ahn and Gruen 2005a).
It could be shown in further experiments (Bordas and others2004) that the concentration of HAAs was increased with an in-crease in the temperature and an extension of the cooking time. A4,8-DiMeIQx concentration of 1.2 ng/g was detected in an aque-ous model system with meat flavor extract (bouillon, extract/water1:2) after heating for 1 h at 175 °C in an oven. With the same tem-perature and a doubling of the heating time, a threefold quantityof 4,8-DiMeIQx was detected. In this meat flavor model system indry conditions, changing the heating times and temperatures from100 °C for 1 h to 2 h, 150 °C for 30 min to 1 h, and 200 °C for10 min up to 30 min, respectively, made the effect of temperatureand time on HAA formation clear. While a MeIQx concentrationof 3.5 ng/g was found at 100 °C for both times, at 150 °C, therewas a concentration of around 4 ng/g for both heating times. Af-ter increasing to 200 °C, the concentration increased from 3.7 to10.7 ng/g for the longer heating time (Bordas and others 2004).A significant increase of PhIP concentration was already observedat 100 °C after 2 h compared to a heating time of 1 h (Bordas andothers 2004).
The results showed that HAA formation depends on temper-ature, time, and cooking methods. Further studies are shown inTable 2 to 4 concerning cooked meat samples (beef patties, chickenbreasts, and pork) under different cooking methods and at variousheating conditions.
Heat and mass transfer in cooked meat productsThe objectives of cooking meat products are to warrant mi-
crobial safety and to produce flavor in meat. Some chemical and
physical alterations take place during the cooking of meat, such asprotein denaturation, melting of fat, water evaporation, changesin texture and shape, water transport of solutes, and formationof crust on the surface (Kondjoyan and others 2014). The de-naturation of meat proteins during heating results in shrinkage,hardening, and the release of juice. First, the denaturation of themyosin takes place at approximately 53 to 58 °C, second, colla-gen shrinks and denatures at approximately 70 to 74 °C similaras sarcoplasmic proteins, and third actin denaturation at about 80to 82 °C results in a decrease in water-holding capacity with therelease of serum (Bertram and others 2006). Free amino acids andsugars in the outer parts of the meat react via the Maillard reac-tion during frying and form a variety of reaction products, whichare important for the color and flavor of cooked meat; aminoacids containing sulfur are known as precursors for the final flavor(Shahidi and others 2014).
The effect of temperature and time on HAA formation wasdescribed in the previous section. However, the cooking methodused also has a big effect on the formation of mutagenic activ-ity. Authors found in different studies that no amounts of HAAsor mutagenicity occurred with gentle cooking methods, such asstewing, steaming, and boiling as temperatures are below 120°C (Persson and others 2002; Joshi and others 2015) (Table 2and 3). With dry-heating methods, such as frying in a pan, sig-nificantly higher mutagen levels occur than with oven-roasting(Table 2 to 4). In addition to the temperatures and heating time,the methods differed between the type of heat transfer, such asconvection, conduction, or radiation, and the surrounding me-dia, such as metal, water, fat, or air, which result in different heattransfer coefficients (Houšová and Topinka 1985; Pan and Singh2002; Zorrilla and Singh 2003; Kondjoyan and others 2013). Itcould be shown in model trials that the preparation of hamburg-ers in a pan led to a much higher mutagenic activity and HAAconcentrations than oven-roasting. A comparison of the resultsof preparation in a deep-fat fryer, a convection oven, and on acontact grill showed clear differences (Persson and others 2002)(Table 2 and 3). MeIQx levels after preparation on a contact grillare around 10 times higher than with the other methods. Withpreparation in a deep-fat fryer or a convection oven, lower lev-els of HAAs were formed. The differences are explained by thesignificantly better heat transfer in methods with direct contact,such as grilling or frying in a pan (heat transfer coefficient of pan150, air 40 Wm2/K, and up to 10000 Wm2/K for boiling wa-ter), than in methods without direct contact (air convection 30to 40 Wm2/K) (Pan and Singh 2002; Sprague and Colvin 2011).There is an indirect heat application in ovenroasting as the heathas to be transferred by convection. There is also an indirect heattransfer using microwave methods, as the energy here is trans-ferred in the form of radiation (Haskaraca and others 2014; Singhand Heldman 2014). That is the reason why preparations of meatproducts using a microwave pretreatment result in very low HAAconcentrations compared to only charcoal-grilled or deep-friedchicken or beef samples (Felton and others 1994; Jinap and oth-ers 2013). However, Barrington and others (1990) showed a highmutagenic activity in 1 beef steak sample which were microwave-cooked for 5 to 7 min per side in contrast to lower times. Thistreatment most probably resulted in high water loss. The high-est HAA levels occurred when frying on a contact grill (Panand others 2000). The reason for this is the direct contact withthe heating medium, which results in a higher conductive heattransfer.
C© 2016 Institute of Food Technologists® Vol. 15, 2016 � Comprehensive Reviews in Food Science and Food Safety 285
Heterocyclic amines in cooked meat products . . .
Kinetics of HAA formationThe kinetics of HAA formation was studied by using model
systems prepared with precursors, such as creatinine, glucose,dipeptides, and free amino acids, analogously to levels in beef(Arvidsson and others 1997) and beef juices as a model system,which were obtained from roasted beef (Arvidsson and others1999). The HAAs are stable at ambient temperature, but they aredisposed to degradation at higher temperatures (Jackson and Har-graves 1995; Arvidsson and others 1997). Degradation occurred at100 °C in solutions of standard HAAs and increased significantlywith temperatures at 200 to 225 °C. In this connection, PhIP wasfound to be the most disposed to degradation, followed by MeIQx,4,8-DiMeIQx, and IQx (Chiu and Chen 2000). However, degra-dation in meat juice systems or meat was different because theformation can balance more due to various parameters, such as heattransfer, mass transfer, vaporization of water, and crust formation,which complicate the kinetics calculations. Besides the generationreactions, the kinetics of HAAs also included a subsequent degra-dation of HAA compounds. The formation of HAAs ([HAA]levels in ng/g, t for time) is supposed to follow a first-order reac-tion equation with a rate constant given by the Arrhenius equation(Arvidsson and others 1997, 1999; Tran and others 2002):
∂ [HAA]
∂ t= A · e (− EaRT ) (1)
k = kb Th
· e ( �SR ) · e (− �HRT ), (2)
where Ea is the activation energy, R is the gas constant (8.3145J/(mol·K)), and A is the unknown exponential prefactor. Forthe reaction mechanism, the modulation of Eq. 1 was used inthe Eyring Eq. 2 for the determination of the temperature-dependence (T in K) of rate constant of the formation wherek is the rate constant, kb is the Boltzmann constant (1.381 × 10−23J/K), h is the Planck constant (6.626 × 10−34 J s), and �H isthe activation enthalpy (Arvidsson and others 1997, 1999). Theactivation entropy �S calculated showed that the rate-limitingstep for the formation of PhIP follows rather a monomolecularreaction, whereas for all the IQx-type compounds the formationprobably follows a bimolecular reaction of pseudo-first order (1 ofthe 2 reactants being in large excess) for MeIQx, 4,8-DiMeIQx,and IQx (Jägerstad and others 1998). The limiting step of the ratecould be the reaction between creatinine, aldehyde, and pyrazine(Arvidsson and others 1997, 1999). Using numerical methodswith different mathematical and computational modeling of pan-frying can show the predictive values of the transport of waterand the temperature distribution in patties, as well as the asso-ciated formation of HAAs, for turning once and multi-turning(Sprague and Colvin 2011). The modeling of the formation ofHAAs in slices of beef (musculus longissimus thoracis and semimem-branosus) subjected to jets of hot air depicted the influence ofextreme dehydration (low water activity) obtained, which slowedthe formation of IQx, MeIQx, and, particularly, 4,8-DiMeIQxcompared with superheated steam treatments. By contrast, a re-verse effect was found for PhIP levels, which increased 1.4- to5.5-fold (Kondjoyan and others 2010a; 2010b). In this connec-tion, the knowledge is important that many reactions favoringat various environmental conditions can simultaneously proceed.In model systems with high temperature around 225 °C or longcooking times, formation and degradation of the HAA in partic-ular PhIP were reported (Chiu and Chen 2000; Arvidsson and
others 1997). Additionally there is known from model studies thatdry conditions (freeze dried meat juice) resulted in an increasedformation of PhIP, DMIP, TMIP, and IFP, but in the wet system(meat juice), MeIQx and 4,8 DiMeIQx was favored (Borgen andothers 2001).
Chemical FactorsFat content and lipid oxidation products
No clear statement could be made on the role of fat in thedevelopment of mutagenic activity. The fat content of the productsaffected the formation of mutagenic substances, but it is not clearwhether this was due to physical or chemical influences (Johanssonand Jägerstad 1993; Hwang and Ngadi 2002).
Moreover, the precursors being necessary for HAA formation,such as creatine/creatinine, free amino acids, and carbohydrates,are present almost exclusively in lean meat, and a higher fat con-tent results in a dilution of the precursors (Knize and others 1985)leading to a reduction of the mutagenic activity in the Ames test(Chen and others 1990). However, it has also been shown thata higher fat content in meat results in a shorter time needed toreach a fixed meat surface temperature due to a more effectiveheat transfer and thereby an increased formation (Abdulkarim andSmith 1998). Furthermore, if the fat is oxidized the HAA forma-tion is enhanced that was shown in different studies (Johansson andJägerstad 1993; Zamora and others 2012). Nonetheless, many sci-entists agree that there is an optimal fat content for the formationof the highest possible HAA concentration in a product.
Regarding beef patties, HAAs such as MeIQx, PhIP, norhar-man, harman, and Trp-P-2 were found at higher levels with about5% fat than those with 15% fat after frying at 150, 170, and190 °C (Abdulkarim and Smith 1998). In beef patties with 15%and 30% fat content, which were cooked to a core temperature of100 °C on a propane grill using a cooking time from 10 up to26 min (weight loss from 46% to 55%), the low-fat patties had thehigher levels of PhIP, but lower levels of AαC than the high-fatpatties (Knize and others 1997c).
Lipids had increased the yield of pyrazines and Strecker aldehy-des in model trials (Arnoldi and others 1990). A possible reason forthis is fat oxidation. Oxidized fats led to radicals which supportedthe formation of HAAs by this oxidation (Johansson and Jägerstad1993). The content of HAAs in meat, including pan residue fry-ing with different frying fats, or oils (butter, margarine, margarinefat phase, liquid margarine, rapeseed oil, and sunflower seed oil),was significantly lower after frying in sunflower seed oil or mar-garine than after frying with the other fats (Johansson and others1995b). The variations in generation of MeIQx and DiMeIQxcould be stated with regard to the concentrations of antioxidants,such as vitamin A, vitamin E, and tocopherols/tocotrienols, andstatus of oxidation (peroxide and anisidine value) (Johansson andothers 1995b) as antioxidants can reduce the formation while theoxidized fat can increase it. Other researchers reported that fatsalso have an enhancing influence on the yield of HAAs in modelsystems after the addition of iron ions (Fe2+ or Fe3+), probably byfree radicals formed during thermally induced lipid oxidation. Inthis model system, the level of MeIQx formed was nearly dou-bled, probably due to iron-catalyzed lipid peroxidation and, thus,formation of free radicals (Felton and others 2000).
Most authors assume that the heat transfer is improved with anincrease in the fat content to an optimal fat–water ratio (Hwangand Ngadi 2002) leading to a shortening of the cooking time. Inthis study using high fat meat emulsions, the authors observed thatthe activation energy of HAA formation was reduced at high-fat
286 Comprehensive Reviews in Food Science and Food Safety � Vol. 15, 2016 C© 2016 Institute of Food Technologists®
Heterocyclic amines in cooked meat products . . .
levels due to a more thermodynamically favored reaction (Hwangand Ngadi 2002). An increase in the fat content over the optimalratio results in a decreased HAA concentration (Persson and others2008; Abdulkarim and Smith 1998).
Moisture content and aw-valueWater serves as a reaction and transport medium during the
heating of meat and meat products. With water, the precursor sub-stances creatine/creatinine, amino acids, glucose, and so on, canbe transported to the surface of the product (Persson and others2002). At the surface, these substances are exposed to higher tem-peratures and so contribute to the formation of HAAs (Overvikand others 1989). The formation of mutagenic substances couldbe significantly reduced by reducing water vaporization during thecooking process (Persson and others 2004). The mutagenic activityis reduced with a reduction of the water content and an increasein the fat content; the capacity of the transport medium declinesand the precursors become diluted (Knize and others 1985).
An increase in water-holding capacity could, thus, be achievedby the addition of water-binding substances into minced meat(Persson and others 2003b, 2004; Wang and others 1982). Theauthors showed that with the addition of a tripolyphosphate plussodium chloride mixture (Persson and others 2003b), soy protein(Wang and others 1982) or carbohydrates (Shin and others 2003,2004; Persson and others 2004) to minced meat products, waterwas bound, leading to a reduction in HAA formation. Wrap-ping hamburgers with carrageenan had a similar effect (Schoch2003). The extreme dehydration obtained with the hot-air jetsresulted in a low water activity and slowed down the formationof IQx, MeIQx and, particularly, 4,8-DiMeIQx compared withsuperheated steam treatments (Kondjoyan and others 2010a,b).But the reverse effect was detected for PhIP concentrations whichincreased (Kondjoyan and others 2010b). In pan-fried bacon, theheat treatment of bacon with water activity around 0.93 caused ahigh content of MeIQx (Gibis and others 2015).
Antioxidants and reducing agentsNext to physical influences (temperature, time, and prepara-
tion method), added substances, such as nitrite, antioxidants, orspices, also showed an inhibiting effect on HAA formation assummarized for beef products as an example in (Table 6). Bothmain components such as lipids and proteins may be oxidized in aseries of radical reactions that include steps of initiation, propaga-tion, and termination with simultaneous formation of free radicals(Weiss and others 2010). The antioxidants did not only inhibitfat oxidation. It is known that free radicals may be involved inthe mechanism of HAA generation and Maillard reaction. Theantioxidants showed that they would scavenge the free radicalsand reduce HAA formation, and this could inhibit the radicalreactions in HAA formation, which was shown in an electronparamagnetic resonance experiment (Kikugawa 1999). Depend-ing on the substances added, HAA formation could be stimulatedor inhibited. When nitrite was added to meat products, severalHAAs, such as Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2, and AαC,were transformed under acidic conditions (pH 2) into their hy-droxy derivatives in model systems and so lost their mutagenicactivity (Furihata and Matsushima 1986). Nitrite is known as anadditive with properties as a reducing agent, at low pH values, thatcan reduce the concentration of non-IQ-type HAAs in modelsystems (Tsuda and others 1985; Shin 2005). Nitrite can also reactwith the amino acid cysteine causing the formation of an antiox-idant, and free radicals can, thus, be blocked (Shin 2005). The
generation of IQ-compounds is reduced by the addition of to-copherol (vitamin E) (Balogh and others 2000; Lan and others2004). Tocopherol is a naturally occurring antioxidant. The inhi-bition is due to the direct blocking of free radicals (Shin 2005). Inaddition, the breakdown products of tocopherol react with pre-cursors, which are then no longer available for the formation ofmutagens. Ascorbic acid or sodium ascorbate as a naturally occur-ring antioxidant and reductone, respectively, could also similarlybring about an HAA reduction (Kato and others 2000; Kiku-gawa and others 2000; Dundar and others 2012; Wong and others2012). A moderate inhibitory effect (approximately 20%) on theformation of PhIP, 4,8-DiMeIQx, and MeIQx was found forwater-soluble vitamins, such as niacin and ascorbic acid; whereaspyridoxamine reduced the concentrations of all 3 HAAs by ap-proximately 40% (Wong and others 2012). This study showedthat pyridoxamine reduced the level of PhIP significantly by trap-ping the phenylacetaldehyde and reacted with the latter to forma pyridoxamine-phenylacetaldehyde adduct which was confirmedby using LC–ESI–MS/MS and NMR spectroscopy (Wong andothers 2012). Rosemary extract reduced the content of HAAs infried beef patties (Puangsombat and Smith 2010; Damašius andothers 2011; Gibis and Weiss 2012). An inhibiting effect of thetomato carotenoid fraction on the formation of imidazoquino-lines (IQx, MeIQx, and DiMeIQx) was reported in model sys-tems containing freeze-dried bovine meat juice (Vitaglione andothers 2002). Using carotenoid extract at a concentration of 1000mg/kg, inhibitions of 13% of MeIQx and of 5% of 4,8-DiMeIQxin the meat juice model system were observed. The effect of themain tomato flavonoid, quercetin, gave an inhibition of MeIQxformation between 9% and 57% with a maximum effect of 67%at 10 mg/kg (Vitaglione and others 2002).
Butylated hydroxyanisole (BHA) is a synthetic antioxidant. Itwas seen in a study that the HAA level could be lowered by 40%with BHA (Weisburger 2005). A few years later, this inhibitoryeffect was confirmed and it was discovered that other phenolicantioxidants, such as epigallocatechin gallate and sesamol (Oguriand others 1998), had a positive effect on the reduction of HAAformation (Weisburger and others 1994).
Extracts of vegetables and fruit could also lead to a reduction inmutagenic activity (Britt and others 1998). Next to pure antioxi-dants, extracts with polyphenols present in plants were particularlyeffective (Ahn and Gruen 2005b; Cheng and others 2007; Gibisand Weiss 2012; Liao and others 2011a). The levels of differentHAAs, such as MeIQx or PhIP, can be lowered by heating modelsystems with polyphenols, such as quercetin, rutin, catechin, cat-echin gallate, and n-propyl gallate (Arimoto-Kobayashi and others2003; Ahn and Gruen 2005b). The results indicated that phe-nols having 2 hydroxy groups at meta positions of the aromaticring were the most efficient inhibitors. The presence of alkyl orcarboxylic groups as additional substituents in the aromatic ring re-duced the inhibitory effect slightly (Arimoto-Kobayashi and others2003). On the other hand, the introduction of additional hydroxyand amino groups mostly cancelled the inhibitory effect, whichwas also mostly absent in ortho and para dihydroxy derivatives;in complex phenols, the presence of several rings with oppositeeffects produced a reduced inhibitory effect (Arimoto-Kobayashiand others 2003). An isotope-labeling study showed that all frag-ments contained had phenylalanine as the origin. The reactionemployed phenylacetaldehyde and epigallocatechin gallate, whichfurther confirmed the ability