Ejm. Konings, A validated liquid chromatographic method for determining folates in vegetables, milk powder, liver, and flour, J AOAC INT, 82(1), 1999, pp. 119-127
A liquid chromatographic (LC) method was elaborated for determining folates
in foods. Folates were extracted by homogenizing in buffer and heat treatm
ent. A portion was incubated with an enzyme preparation containing conjugas
e, amylase, and protease. After purification by affinity chromatography, fo
late monoglutamates were determined by reversed-phase LC with fluorescence
and diode array detection. Gradient elution with phosphate buffer and aceto
nitrile was used to separate vitamers. The most abundant folate forms natur
ally present in foods were detected, including tetrahydrofolic acid, 5-meth
yltetrahydrofolic acid, and 5-formyltetrahydrofolic acid. 10-Formylfolic ac
id could be detected by applying a second fluorescence detector. Folic acid
, used for fortification, might also be quantitated with this system. The d
ifference between folate concentrations in sample extracts, with and withou
t treatment of conjugase, is a measure of the quantity of polyglutamates in
the food matrixes. An additional treatment with conjugase, amylase, and pr
otease reflects the amount of matrix-bound folates. The LC system gave a li
near response over the range 0-100 ng/mL. Detection limit for these compoun
ds were 7 pg/mL for tetrahydrofolic acid and 5-methyltetrahydrofolic acid a
nd 59 pg/mL for 10-formylfolic acid (signal-to-noise ratio greater than or
equal to 3) when 100 mu L was injected. Detection limits for 5-formyltetrah
ydrofolic acid and folic acid were 1 ng/mL. Repeatability relative standard
deviation values for separate folates in 3 candidate Certified Reference M
aterials (CRMs)-mixed vegetables (CRM 485), pig liver (CRM 487), and wholem
eal flour (CRM 121)-and a Certified Reference Material milk powder (CRM 421
) varied from 3.3 to 21.0% for the concentration range 1.8-1440 mu g/100 g.
Recoveries ranged from 73 to 109%. Use of amylase and protease was advanta
geous. Use of a commercially available folate-binding protein for cleanup s
aved time and money and was effective. Results for 5-methyltetrahydrofolic
acid were in good agreement with results obtained with other LC methods. Re
sults for total folates were lower than results obtained with microbiologic
al methods.