T. Rundlof et al., Potential nitrite scavengers as inhibitors of the formation of N-nitrosamines in solution and tobacco matrix systems, J AGR FOOD, 48(9), 2000, pp. 4381-4388
The ability of 20 compounds, all but one tobacco constituents, to inhibit t
he formation of tobacco specific N-nitrosamines (TSNA) was investigated in
buffer and detergent solution and in tobacco midrib and lamina systems. In
solution at pH 5.5, ascorbic acid and the phenolic acids caffeic and feruli
c acid were the most potent inhibitors of the reaction between nornicotine
and nitrite, with nearly complete inhibition at molar ratios test compound/
nitrite > 1:I. Also, cysteine > dihydrocaffeic acid > protocatechuic acid a
pproximate to catechin acted as strong inhibitors with >90% inhibition at a
ratio of 3:1. Lower inhibitions were observed with chlorogenic acid > p-co
umaric acid > sclareol > serine. Rutin showed an inhibition of 34% at a rat
io of 0.1:1. Sclareol, alanine, proline, and serine did not significantly a
ffect the N-nitrosonornicotine (NNN) formation. alpha-Tocopherol:and glutat
hione enhanced NNN formation at pH 5.5 but were inhibitors at pH 3. Cinnami
c acid, vanillic acid, eugenol, and esculin enhanced NNN formation. Increas
ed NNN formation was also observed for dihydrocaffeic acid, chlorogenic aci
d, protocatechuic acid, and catechin at a less-than-equimolar ratio of test
compound to nitrite. The tobacco matrix experiments were performed with ai
r-cured, ground tobacco midrib and lamina. Caffeic acid, ferulic acid, dihy
drocaffeic acid and catechin were potent inhibitors of the formation of TSN
A in the midrib as well as in the lamina. Also protocatechuic acid, glutath
ione, ascorbic acid, p-coumaric acid, chlorogenic acid and cysteine were in
hibitors, while alpha-tocopherol and rutin inhibited the reaction in the mi
drib but not in the lamina. Cinnamic acid, vanillic acid, eugenol, alanine,
proline and serine showed small effects only. The molar ratio of secondary
alkaloid(s)/nitrite in the test systems were 0.1:1 (solution), similar to
0.25:1 (midrib), and similar to 1:1 (lamina) and is most likely the major c
ontributor to the observed order of inhibition potency (solution > midrib >
lamina) of the test compounds. The vicinal phenolic hydroxyl groups of pol
yphenols and the simultaneous presence of a phenol group and an olefinic bo
nd in hydroxycinnamic acids were the most characteristic structural element
s of the potent inhibitors.