Mikrochim. Acta [Wien] 1989, I, 65--73
y by Springer-Verlag 1989
A Solid-Phase Enzyme-Linked Assay for Vitamin B12
Catherine D. Tsalta*, Sara A. Rosario, Geun Sig Cha,
Leonidas G. Bachas**, and Mark E. Meyerhoff***
Department of Chemistry, The University of Michigan, Ann Arbor, MI 48109-1055, USA
Abstract. A new solid-phase enzyme-linked competitive binding assay
for vitamin B12 (cyanocobalamin) is described. The assay is based on the
competition between analyte B12 molecules and a glucose-6-phosphate
dehydrogenase-vitamin B12 conjugate for a limited number of R-protein
binding sites immobilized on sepharose particles. After appropriate
incubation and washing steps, the enzyme activity bound to the solid-
phase is inversely related to the concentration of B12 in the sample.
Under optimized conditions, the method can detect B12 in the range of
3 x 10-1~ x 10 -8 M (using 100 #1 sample) with high selectivity over
other biological molecules.
Key words: vitamin B12 (cyanocobalamin), competitive binding assays,
enzyme-vitamin conjugates, vitamin tablet analyses.
Enzyme-immunoassays (EIA) are attractive alternatives to classical radio-
immunoassay procedures for the detection of biomolecules at trace levels.
Modern heterogeneous EIA methods are often based on the competition
between the analyte and enzyme-labeled analyte molecules for antibody
binding sites immobilized on a solid phase [1, 2]. After appropriate sepa-
ration and washing steps, the measured enzyme catalytic activity bound to
the solid support is inversely proportional to the concentration of analyte in
the sample. While selective antibodies have been the reagents most often
used to devise such methods, we recently demonstrated the advantages of
using immobilized natural binders [3--5] in place of antibodies, particularly
in cases where the preparation of antibodies toward the analyte is difficult.
We now report an extension of this new concept by describing a very simple
solid-phase enzyme-linked method suitable for the direct determination of
vitamin B12 (cyanocobalamin, CNCbl).
* Current address: Kallestad Lab., Inc., 200 Lake Hazeline, Chaska, MN 55318, USA
** Current address: Department of Chemistry, University of Kentucky, Lexington, KY
*** To whom all correspondence should be addressed
66 C.D. Tsalta et al.
The two bioanalytical methods most often used to detect B12 include
microbiological assays [6, 7] and radioassay competitive binding methods
[7--10]. The microbiological method is very slow (1--2 days) and yields
only semiquantitative values. The competitive binding technique involves
the use of a 57Co-cobalamin in conjunction with various vitamin B12
selective binders (R-protein, Intrinsic Factor, and TranscobalaminII)
insolubilized on solid supports [7--13]. While extremely sensitive, the
radioassay method is plagued by the need to use and dispose of radio-
Recently , we introduced a new homogeneous enzyme-linked assay
for vitamin B~2 based on the inhibition of glucose-6-phosphate dehy-
drogenase-vitamin B12 (G6PDH-Ba2) conjugates by soluble R-protein. This
homogeneous method was quite rapid and selective; however, even under
optimized conditions, the proposed assay could only detect B12 at levels
>10 nM. In the assay described here, we report results obtained when
using an immobilized form of R-protein in conjunction with similar
G6PDH-Bu conjugates. By employing a heterogeneous assay protocol, the
detection capabilities of the enzyme-linked competitive binding method are
improved significantly (down to 0.3 nM). The final assay is selective for B~2
and useful for the direct determination of B~2 in infant formula and vitamin
Enzyme activities were measured with a Gilford (Oberlin, OH) Stasar III Spectrophotometer
equipped with a vacuum-operated sampling system and temperature-controlled cuvette. The
cuvette chamber was maintained at 30 ~ C. The spectrophotometer was interfaced with a Syva
CP-5000 EMIT Clinical Processor.
Porcine R-protein (non-intrinsic factor), vitamin B12 , glucose-6-phosphate dehydrogenase
(G6PDH) (from Leuconostoc mesenteroides), as well as all other biochemicals were obtained
from Sigma Chemical Co. (St. Louis, MO).
Substrate solutions (glucose-6-phosphate, fl-NAD +) were prepared in 0.050M
Tris(hydroxymethyl)aminomethane-hydrochloric acid (Tris-HCl) buffer, pH 7.8, containing
0.10 M NaC1 and 0.01% (w/v) NaN3 (assay buffer). Conjugates, standards, and binding
protein solutions were prepared in the same buffer also containing 0.1% (w/v) gelatin
(Tris-gel buffer). Gelatin was added to reduce non-specific adsorption of the enzyme-B12
conjugate or analyte B12 molecules onto the walls of the test tubes and the solid-phase R-
protein beads. The vitamin preparations and the infant formula analyzed were commercially
available and their manufacturers and compositions are listed in Table 1.
A standard stock solution of vitamin B12 was prepared by dissolving a given amount of
cyanocobalamin in assay buffer. The concentration of this stock solution was determined
spectrophotometrically using a molar extinction coefficient for CNCbl at 361nm,
E361 = 28060 1 mo1-1 cm -1. Standards solutions of B12 in the range of 10-11--10 6 Mwere
prepared by diluting this stock solution with Tris-gel buffer.
A Solid-Phase Enzyme-Linked Assay for Vitamin B12
Table 1. Composition of analyzed samples
Content B complex a
B complex b Folic acid c
(rag/5 fl oz)
Vitamin B 2
Vitamin B 6
Para-amino benzoic acid
a Bio-genics, Woodland Hills, CA
b The Kroger Co., Cincinnati, OH
~ Makers of KAL, Inc., Woodland Hills, CA
Similac, Ross Laboratories, Columbus, OH
* Contains numerous inorganic salts
Preparation of G6PDH-B12 Conjugate
A G6PDH-B12 conjugate containing approximately 6.4 B12 molecules per
enzyme was synthesized from monocarboxylic acid derivatives of B12 using
the procedure described in an earlier paper . This conjugate possessed
83% of its original enzymatic activity following the conjugation reaction.
Preparation of R-Protein Solid-Phase Reagent
R-protein solid-phase beads were prepared by covalently attaching
0.485 mg of R-protein to 900 mg of commercial CNBr activated Sepharose
4B particles according to the method suggested by the manufacturer .
Briefly, the activated gel was swollen and washed in 1 mM HC1 solution (in
order to remove dextran and lactose, which are added to the activated gel to
preserve its activity under freeze-drying). The protein to be coupled
was dissolved in coupling buffer (0.1 M NaHCO3, pH 8.3, 0.5 M NaC1)
and then added to the gel suspension and incubated for 2 h. Excess
protein was washed away and the remaining active groups in the
beads were blocked using glycine. After the coupling was completed
and the remaining active groups blocked, the excess blocking reagent
and adsorbed protein were washed away by alternatively washing the
beads with high and low pH buffer solutions. The solid-phase was
68 C.D. Tsalta et al.
finally resuspended in a 1 : 5 suspension (ratio of beads to total volume of
Tris-gel buffer) and stored at 4~ This suspension was further diluted
(1:66) with Tris-gel buffer for use as a reagent in the heterogeneous
competitive binding assay.
The solid phase binding beads could be stored for at least three months
without significant changes in their ability to bind G6PDH-B12 conjugates.
However, attempts to regenerate (by washing with an alkaline glycine
buffer, pH 12.9 ) the solid-phase after use in the B12 assays resulted in
beads that retained only 30% of their initial conjugate binding capacity.
Thus, fresh solid-phase reagent was always used for each B12 assay and to
obtain calibration data.
Heterogeneous Assays for Vitamin B12
The assays were carried out in a single incubation (equilibrium) mode. This
was accomplished by mixing 100#1 of the standard or sample solution with
100 yl of a conjugate solution and 200 #1 of an R-protein suspension with
400 r Tris-gel buffer. After an incubation period of 2.5 h, the beads were
centrifuged and washed three times with assay buffer. One hundred r of
each enzyme substrate, G6P (0.10 M) and fl-NAD + (0.063 M), and 800 j.tl of
assay buffer were then added and the tubes were incubated again for 2 h.
After this time, the beads were centrifuged and the absorbance of the super-
natant was measured at 340 nm.
Sample Preparations and Vitamin B12 Determinations
Infant formula. Samples of the liquid infant formula were used as is or
diluted 1 : 2 with Tris-gel buffer.
Vitamin tablets. For the folic acid with B12 vitamin preparation, 12 tablets
were ground and the amount equivalent to 3 tablets was weighed and mixed
with deionized H20 in a 50 ml centrifuge tube. The solution was shaken,
end to end, for 1 h at 4 ~ C. The suspension was then centrifuged four times
at 2400 rpm for 10 rain intervals. The supernatant was transferred, after
each centrifugation step, to a 100 ml volumetric flask. The volume was then
completed up to 100 ml with deionized H20. Several dilutions of the super-
natant were made with Tris-gel buffer. For the B-complex preparations
(with and without Vitamin C), 10 tablets were ground together and the
amount equivalent to one tablet was weighed and added to a 1000 ml volu-
metric flask. The volume was completed to mark with deionized H20 and
this solution was stirred for 24 h at 4 ~ C. From this sample solution, further
dilutions in Tris-gel buffer were prepared.
The resulting samples were stored at 4~
These samples were analyzed according to the heterogeneous assay
protocol described above, along with the standards. The unknown concen-
trations were estimated graphically from the calibration curve. Only the
steep portion of the dose response curve was used for analytical purposes.
and protected from light.
A Solid-Phase Enzyme-Linked Assay for Vitamin B12 69
For the recovery study, a concentrated infant formula solution was
spiked with 100 ~1, 500 r or 1500/.tl of a 3.5 • 10-7MB12 standard solution.
These spiked solutions were then diluted 1 : 5 with Tris-gel buffer. The rest
of the procedure was the same as for the analysis of the unspiked infant
Results and Discussion
Since the binding interaction of G6PDH-B12 and R-protein in solution
previously had been found to induce inhibition of enzymatic activity ,
we initially wanted to determine whether the immobilized form of R-protein
would yield an analogous effect. To investigate this, varying amounts of the
solid-phase R-protein reagent were incubated for 30 min with a given
amount of G6PDH-B12 conjugate (1.25 x 10 -9 M), in the presence of the
substrates G6P and fl-NAD +. After centrifugation, the absorbance of the
supernatant was measured and compared to that obtained for the same
experiment in the absence of solid-phase R-protein (full activity). While
nearly 30% inhibition was observed in the presence of excess (i. e., 30 r
absolute volume) solid-phase R-protein (compared to 65% for the same
conjugate when soluble R-protein is employed ) insignificant effects
(2--4% inhibition) were found when the amount of solid-phase R-protein
equalled that normally employed in the solid-phase enzyme-linked compet-
itive assay protocol (2--4/.tl; limiting reagent). Thus, we concluded that the
inhibition induced by the immobilized R-protein would not cause problems
when attempting to utilize G6PDH-Bn conjugates in devising a heteroge-
neous competitive binding assay for Ba2.
I I I
' ; ' s ; ,, ,;',;' b'
abs. voL of beods (pL)
Fig. 1. G6PDH-B12 conjugate (9.4x 10-1~ M) binding as a function of increasing immo-
bilized R-protein solid-phase (volume refers to settled volume of R-protein beads). Data
points are average of two determinations
70 C.D. Tsalta et al.
In order to determine the capacity of the solid-phase R-protein for
binding the G6PDH-B~2 conjugate (not the actual binding constant), varying
amounts of the R-protein reagent were incubated with a fixed level of
conjugate. After 3 h, the activity of conjugate bound to the solid phase was
determined. Fig. 1 illustrates the results of such an experiment. At higher
levels of R-protein beads, the amount of bound conjugate activity levels off
to about 40% of that initially added to the reaction mixture. Given that at
high levels of R-protein beads there is also some inhibition of the conjugate
(see above discussion), we estimate that about 60% of the starting G6PDH-
B12 conjugate can be bound by the immobilized R-protein. The remainder of
the conjugate molecules probably cannot bind to the solid-phase due to
steric factors (i. e., their covalently bound B~2 molecules are in regions inac-
cessible for binding to the solid phase). Generally, for the heterogeneous
competitive binding assays, we chose an amount of solid phase R-protein
which fell on the steep portion of the binding curve (e.g. 1--3/,tl absolute
bead volume) (see Fig. 1).
The kinetics of the binding of G6PDH-B12 to immobilized R-protein was
also examined (Fig. 2). We observed that binding of the conjugate to the
solid phase continues to increase somewhat, even after a 3 h equilibration
period. However, the rate of binding is significantly slowed after the first
2 h and thus, in subsequent competitive binding assays, we chose a 2.5 h
time period for all incubations between the conjugate, immobilized binder,
and standards or samples. Interestingly, in related work , we observed
that the overall kinetics of conjugate binding are enhanced when the solid-
phase has lower levels of protein coverage. Therefore, in the present effort,
f I I I I I I I I I I L
gO 80 120 160 200 2gO
Fig. 2. G6PDH-B12 conjugate (1.88 x 10-9/140 binding to R-protein beads (4.5 #l absolute
bead volume) as a function of increasing incubation time. Data points are average of two
A Solid-Phase Enzyme-Linked Assay for Vitamin B12 71
Fig. 3. Typical competitive binding dose-response curve toward B12 (M) using 1.9 x 10 -9 M
G6PDH-B12 conjugate and 1.5 #1 absolute bead volume of solid-phase R-protein. Data
points are average of two trials
we purposely prepared our solid-phase reagent using high mass ratios of
Sepharose beads to R-protein.
Fig. 3 shows a typical B12 dose-response curve obtained using the
various reagents in a competitive binding assay protocol. The detection
ranges, steepness of the curves, and detection limits vary depending on the
concentrations of reagents used. To obtain optimized detection limits
(Fig. 3; 3 x 10 -1~ M B12 in sample) low levels of conjugate (1.90 x 10 -9 M)
and binder (1.5 yl absolute volume) were employed. Although bound
enzyme activities could be monitored using shorter incubation times, we
chose a 2 h time period to improve the precision of the method (i.e., longer
reaction times allow for slight differences in the time between reagent addi-
Table 2. Results for the determination of B12 in
S ample Found a Claimed
B complex with C
Folic acid with B12
Infant formula 4.2_+ 0.4 nM 2.5 nM
a Average of four determinations (+ std. dev.)
for vitamin preparations and six determina-
tions (+ std. dev.) for infant formula
72 C.D. Tsalta et al.
tions, a situation which occurs when working with a large number of assay
The precision, selectivity, and ultimate analytical utility of the proposed
solid-phase assay was evaluated by determining the B12 content of various
commercial vitamin and infant formula preparations (see Tables 1 and 2).
Results obtained for the vitamin tablets were in good agreement with the
manufacturer's claims. This agreement, even when B12 is the minor
component, illustrates that the method offers high selectivity over other
common vitamins. Results for the infant formula were substantially higher
than the labeled values. However, according to the manufacturer, the
difference between minimum (labeled) and actual determined values can be
quite high to ensure an adequate shelf-life forthe product. Indeed, subse-
quent analytical recovery studies using the protein-based infant formula
preparation (Table 3) yielded satisfactory accuracy, suggesting that the
method can be used for direct B12 measurements in rather complex samples.
Attempts to use the existing assay for measurements of serum B12 levels
were not undertaken since physiological concentrations of B12 fall at or
below the detection limits of the new enzyme-linked method . The
detection limits obtainable in the enzyme-linked method appear to be
restricted by the binding affinity of the immobilized R-protein toward the
enzyme-labeled conjugate and analyte B12 molecules. Although previous
workers have found that radioactive Ba2 binds to immobilized R-protein
with high affinity, as suggested by low detection limits on competitive ra-
dioassay [13, 17], using a Scatchard analysis method, we found that the
G6PDH-Ba2 conjugate employed in this work bound to our R-protein solid-
phase with an affinity constant of approximately 3.4 x 10 8 M -a . This
value is nearly 3 orders of magnitude lower than that reported (3 x 10 aa
M -1) for the binding of radioactive Ba2 with a similar R-protein solid-phase
. Thus, steric effects owing to the bulkiness of the Ba2 conjugate and/or
the specific spatial orientation of the immobilized R-protein may be
contributing to this low value. Improved binding of the enzyme conjugate
to the solid-phase, and consequently better competitive binding assay
detection capabilities may be achieved by replacing R-protein with a
different natural Ba2 binder. Indeed, preliminary studies using immobilizing
intrinsic factor are now in progress . However, since intrinsic factor is
more selective in its binding (i.e., it can differentiate among the mono-
Table 3. Results from recovery studies for the addition of
B12 to infant formula
Sample Vitamin B12 (nM)
Added Total found a
a Average of two determinations
A Solid-Phase Enzyme-Linked Assay for Vitamin B12
carboxylic acid isomers of B12 [19, 20]), synthesis of new conjugates
prepared with only the "e"-monocarboxylic acid derivative of B12 , rather
than a mixture of monocarboxylic acids, will be required.
Acknowledgements. The authors gratefully acknowledge support of this work by the National
Science Foundation (Grant # 8506695).
 D. Monroe, Anal Chem. 1984, 56, 920A.
 A. Voller, D.E. Bidwell, in: Alternative Immunoassay (W. P. Collins, ed.), Wiley,
NewYork, 1985, pp. 77--86.
 L. G. Bachas, P. Lewis, M. E. Meyerhoff, Anal Chem. 1984, 56, 17213.
 L.G. Bachas, C. D. Tsalta, M. E. Meyerhoff, Biotechniques 1986, 4, 42.
 G. S. Cha, M. E. Meyerhoff, Anal Biochem. 1988, 168, 216.
 D. L. Mollin, B. B. Anderson, J. E. Bauman, Clin. Haematol. 1976, 5' (3), 52h.
 I. Chen, M. I. Sperling, L. A. Heminger, in: Methods in Clinical Chemistry (A. J. Pesce,
L. A. Kaplan, eds.), Mosby, St. Louis, MO, 1987, pp. 569--573.
 K. S. Lau, C. Gottlieb, L. R. Wasserman, V. Herbert, Blood 1965, 26, 202.
 G. Tibbling, Clin. Chim. Aeta 1969, 23, 209.
 M. Caeska, U. Lundkuist, Clin. Chim. Acta 1971, 32, 339.
 E. Nexo, P. Gimsing, Seand. J. Chim. Lab. Invest. 1981, 41, 465.
 I.W. Chen, E. B. Silberstein, H. R. Maxon, M. Sperling, E. Barnes, J. Nucl. Med. 1981,
22 (5), 447.
 N. P. Kubasik, M. Ricotta, H. E. Sine, Clin. Chem. 1980, 26 (5), 598.
 C. D. Tsalta, M. E. Meyerhoff, Anal Chem. 1987, 59, 837.
 Affinity Chromatography, Principles and Methods, Pharmacia Fine Chemicals,
Piscataway, N J, 1979.
 R. Gr~isbeck, U. H. Stenman, L. Puutula, K. Visuri, Biochim. Biophys. Acta 1968, 158, 292.
 M. Muir, I. Chanarin, Br. J. Haematol. 1983, 53, 423.
 C. D. Tsalta, Ph. D. Dissertation, University of Michigan, 1987.
 E. Hippe, F. Haber, H. Olesen, Bioehim. Biophys. Aeta 1971, 243, 75.
 D. L. Anton, H. P. C. Hogenkamp, T. E. Walker, N.A. Matwigoff, J. Am. Chem. Soc.
1980, 102, 2215.
Received September I0, 1987. Revision December 7, 1987.