process. It has therefore been suggested that the ex-
traction conditions should be optimized for each type
of food sample in order to achieve reproducible and
good recoveries of the complete analysis.
Microbiological Assay
0020 Over the years, the growth response of various micro-
organisms has been used to analyze cobalamins in
foods and other biological material. The micro-
organisms that have been used are Lactobacillus
delbrueckii (previously, leichmanni), Escherichia
coli, Euglena gracilis,andOchromonas malhamen-
sis. Below, only the AOAC (Association of Official
Analytical Chemists) Official method (reference
method) based on Lactobacillus delbrueckii is pre-
sented. The other microorganisms have different dis-
advantages compared with Lactobacillus delbrueckii,
i.e., less sensitive and/or specific also towards less
active biological analogs. In addition, Euglena graci-
lis and Ochromonas malhamensis grow more slowly
and are thus more painstaking.
0021 The AOAC Official Method 952.20: Cobalamin
(Vitamin B
12
-activity) in Vitamin preparations use a
microbiological assay with Lactobacillus delbrueckii
(ATCC 7830) (45.2.02). This method was originally
collaborated for use on vitamin preparations, but the
AOAC Task Force on Methods for Nutrition Label-
ing recommended the procedure for use on all food
matrices. The Lactobacillus delbrueckii growth re-
sponse is sufficiently sensitive to quantify cyanoco-
balamin at concentrations approaching 1.0 pg per
milliliter of assay growth medium (corresponding to
less than 0.5 mg per 100 g). It has a variable response
to various cobalamins. A similar growth response has
been reported for cyanocobalamin, hydroxocobala-
min, sulfitocobalamin, nitritocobalamin, and dicya-
nocobalamin. However, adenosylcobalamin produces
a greater response, and methylcobalamin a lower
growth response. This will generally not cause a prob-
lem, since excess of cyanide or sulfite is used during
the extraction procedure prior to analysis, thereby
converting all different native forms of cobalamins
except methylcobalamin into one stable form.
0022 An important aspect in all methods of analysis is
their specificity. Lactobacillus delbrueckii can utilize
vitamin B
12
analogs, e.g., deoxyribonucleotides and
deoxyribonucleosides, in addition to biologically
active cobalamins. Older literature suggests that dilu-
tion of deoxyriboside levels (e.g., thymidine) to less
than 1 mg per milliliter of the assay medium will elim-
inate the effect.
Radioprotein-Binding Assays
0023 Radioprotein-binding assays have been used rou-
tinely for blood and tissue analysis since early work
by Rothenberg and several others during the 1960s.
The principle of this assay is based on the competition
of added isotope-labeled cyanocobalamin (
57
Co)
versus cyanocobalamin from the unknown sample
for a limited number of binding sites on cobalamin-
binding proteins (often intrinsic factor from hog). The
amount of labeled cobalamin bound to the binding
protein is quantified in a gamma counter, and the
concentration of cobalamin in the unknown sample
is inversely proportional to the concentration of
labeled cobalamin, and calculated from a calibration
curve.
0024Early types of radioprotein-binding assays for de-
termination of vitamin B
12
in clinical specimens and
foods were often inaccurate because the binding pro-
tein could bind to active forms of vitamin B
12
as well
as biologically inactive analogs. The specificity of
such assays has been greatly improved through
the use of a vitamin B
12
-binding protein (generally
porcine intrinsic factor, IF) that is specific for the
biologically active forms of the vitamin. Radioligand
binding methods based on monospecific antisera that
eliminate any cross-reactions with other cobalamins
are now available and should be used routinely in all
clinical laboratories.
0025The relative binding affinities of porcine IF for
various cobalamins have also been studied. An
equivalent affinity has been found for cyano-,
dicyano, nitrito-, and methylcobalamins, whereas
significantly different binding affinities have been
found for hydroxo-, sulfito, and adenosylcobala-
mins. Therefore, with cyanocobalamin as the calibra-
tion standard, it is necessary to add cyanide to the
samples prior to quantification during the extraction
procedure to convert all native forms to cyanocobal-
amins.
0026Studies by several research groups indicate that
radioprotein-binding assays with IF can be used for
food analysis. When optimized extraction procedures
have been used, radioprotein binding assays and
microbiological procedures show agreement for
most foods; however, differences have been noted
often enough to conclude that the two methods
are not universally interchangeable for the assay of
all foods. Thus, recent work has shown that the
radioprotein binding assay could be reliable for food
analysis, providing careful method validation includ-
ing control of selectivity. One should bear in mind
that not only the food matrix itself but also sub-
stances used during the extraction procedure, e.g.,
buffer strength, pH, cyanide concentration, and
added enzymes needed to release vitamin B
12
bound
to the food matrix, might cause nonspecific binding
and therefore must be accounted for in order to
achieve reliable results.
1424 COBALAMINS/Properties and Determination