JIM-12286; No of Pages 8

Journal of Immunological Methods xxx (2017) xxxxxx

Contents lists available at ScienceDirect

Journal of Immunological Methods

journal homepage: www.elsevier.com/locate/jim

Research paper

Immunoassays for riboavin and avin mononucleotide using antibodies specic to

D-ribitol and D-ribitol-5-phosphate

G. Ravi, Yeldur P. Venkatesh

Department of Biochemistry and Nutrition, CSIRCentral Food Technological Research Institute (CFTRI), Mysuru 570020, Karnataka State, India

a r t i c l e i n f o a b s t r a c t

Article history: Riboavin (vitamin B2), a water-soluble vitamin, plays a key role in maintaining human health. Though, numer-
Received 2 January 2017 ous methods have been reported for the determination of total riboavin (TRF) content in foods and biological
Received in revised form 28 February 2017 samples, very few methods are reported for quantifying riboavin and its coenzymes [avin mononucleotide
Accepted 13 March 2017
(FMN); avin adenine dinucleotide (FAD)] individually. Recently, we have demonstrated that antibodies specic
Available online xxxx
to D-ribitol and D-ribitol-5-phosphate also recognize riboavin and FMN, respectively, and not vice-versa. In this
Keywords:
study, we have evaluated these two antibodies for the analysis of riboavin and FMN by indirect competitive
Flavin mononucleotide ELISA (icELISA) in selected foods and pharmaceuticals. Under the optimal assay conditions, 50% inhibition con-
Hapten-specic antibodies centration (IC50) and limit of detection (LOD, IC10) were 3.41 ng/mL and 0.02 ng/mL for riboavin, and
Indirect competitive ELISA 7.84 ng/mL and 0.24 ng/mL for FMN, respectively, with detectable concentration range between 0.1 and
Ribitol 100 ng of analytes and b 0.1% cross-reactivity with other water-soluble vitamins. The amounts of TRF in food sam-
Ribitol-5-phosphate ples, as analyzed by icELISA using ribitol antibody, were 9095% of the reported values in the literature or label
Riboavin values. Quantication of individual avins (riboavin and FMN) from the same food samples showed variation
in their values compared to TRF, and were in good agreement with values obtained from HPLC and AOAC
methods. Further, spiking and recovery analysis of food samples and pharmaceuticals showed no signicant ma-
trix effects. The immunoassays were validated in terms of accuracy and precision using inter- and intra-assays.
The immunoassays developed in this study are sensitive and appears feasible for screening a large number of
samples in the quantication of riboavin and FMN in various biological samples, pharmaceuticals and natu-
ral/processed foods.
2017 Elsevier B.V. All rights reserved.

1. Introduction
Riboavin (C17H20N4O6, 7,8-dimethyl-10-ribityl-isoalloxazine) is a
water-soluble vitamin (vitamin B2) present in a wide variety of foods
and has a key role in maintaining human health. Flavins exist naturally
as riboavin and its biologically active cofactor forms, avin mononu-
cleotide (FMN or riboavin-5-phosphate) and avin adenine dinucleo-
tide (FAD), which participate in a range of redox reactions that are
critical for metabolism and energy production (Massey, 2000; Pinto
Abbreviations: CV, coefcient of variance; FAD, avin adenine dinucleotide; FMN,
avin mononucleotide; icELISA, indirect competitive enzyme-linked immunosorbent
assay; KLH, keyhole limpet hemocyanin; LOD, limit of detection; Rbt, ribitol; Rbt-5-P,
ribitol-5-phosphate; TRF, total riboavin.
This work was presented at the International Conference on Healthcare and Technical
Research (ICHTR) held at Manipal University, Manipal, Karnataka State, India during 22
24 December 2015.
Corresponding author at: Department of Biochemistry and Nutrition, CSIRCentral
Food Technological Research Institute (CFTRI), Chaluvamba Vilas, KRS Road, Mysuru
570020, Karnataka State, India.
E-mail address: venkatyp@yahoo.com (Y.P. Venkatesh).
and Rivlin, 2013). Good sources of riboavin include milk, eggs, malted
barley, liver, kidney, heart and leafy vegetables. Riboavin is used to for-
tify several foods and also as a food-coloring agent [INS 101 (i)]
(Eitenmiller and Landen, 2010; Golbach et al., 2014). The recommended
daily allowance (RDA) for vitamin B2 is 0.4 mg/day for infants, 0.6
1.2 mg/day for children, and 1.8 mg/day for male adults and pregnant/
lactating females. The reference daily intake used for nutritional label
declaration is 1.7 mg/day (Eitenmiller and Landen, 2010; Pinto and
Rivlin, 2013; Rivlin, 2007). The structures of important compounds
and analytes related to this study are shown in Fig. 1.
Riboavin content in common foods varies based on types of food
sources: high (meat, egg, milk and milk products), moderate (spices, ce-
reals and pulses), and low (fruits and vegetables) as documented in the
literature and USDA National Nutrient Database (Gopalan et al., 1989;
USDA, 2015). Riboavin and its coenzymes are either present as free
form or covalently/non-covalently bound to macromolecules like pro-
teins and carbohydrates in foods and biological materials (Eitenmiller
and Landen, 2010; Golbach et al., 2014; Pinto and Rivlin, 2013; Rivlin,
2007); for this reason, acid hydrolysis, enzymatic hydrolysis, or both,
has been applied to the samples to obtain total riboavin (TRF) content

http://dx.doi.org/10.1016/j.jim.2017.03.010
0022-1759/ 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
2 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx
Fig. 1. Structures of haptens [D-ribitol and D-ribitol-5-phosphate] and analytes [riboavin and avin mononucleotide (riboavin-5-phosphate or FMN)] related to the immunoassays.
of foods (Eitenmiller and Landen, 2010; Golbach et al., 2014; Pinto and
Rivlin, 2013; Rivlin, 2007).
Numerous methods like uorometric (Russell and Vanderslice,
1992a), microbiological (Snell and Strong, 1939), HPLC (Eitenmiller and
Landen, 2010; Russell, 2012), chemiluminescence (Prez-Ruiz et al.,
1994) and capillary electrophoresis (Cataldi et al., 2003) have been re-
ported for the determination of riboavin content in foods and biological
samples. The standard method for riboavin determination in foods is the
AOAC method (970.65) using uorometry (LOD: 3 ng/mL) (AOAC, 2012).
Numerous HPLC methods have been reported for the quantication of
riboavin alone or along with other water-soluble vitamins with varying
degree of detectability (LOD: 0.130 ng/mL) (Eitenmiller and Landen,
2010; Golbach et al., 2014; Russell, 2012; Valls et al., 1999; Zand et al.,
2012). These methods involve tedious steps in the pre-treatment of sam-
ples, pre-derivatization with critical derivatizing agents, often have the
disadvantages of being less sensitive, and with a limited application to
some food samples, due to the presence of several interfering compounds,
and varied food matrices; moreover, most of these methods measure only
the TRF content of foods.
To quantify individual avins in free as well as in non-covalently
bound forms of riboavin, FMN, and FAD, a limited number of HPLC
methods have been developed for mostly clinical applications using TCA
extraction and methanol/methylene chloride extraction (Lopez-Anaya
and Mayersohn, 1987; Petteys and Frank, 2011). Very few methods
have been described for individually quantifying riboavin and its two
coenzymes in foods (Gliszczyska-Swigo and Kozioowa, 2000; Lumley
and Wiggins, 1981; Russell and Vanderslice, 1992b; Vias et al., 2004).
Immunoassays with their applicability for wide range of analytes,
high sample throughput, relatively low cost, their specicity for the an-
alyte of interest and high sensitivity are widely used in clinical laborato-
ries, food and pharmaceutical industries (Wild, 2013). Various
immunoassays have been developed for the quantitation of several vita-
mins including folic acid (Wild, 2013; Zhang et al., 2012) and vitamin
B12 (Selva Kumar and Thakur, 2011; Wild, 2013). Enzyme-linked ligand
sorbent assays (ELLSA) using competition between riboavin-binding
protein and 3-carboxymethylriboavin-enzyme conjugates or biotin/
avidin-riboavin conjugates have been developed for riboavin analysis
(Cha and Meyerhoff, 1988; Kim et al., 1995; Kozik, 1996); drawbacks of
Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
these assays include the preparation of riboavin-binding protein and
riboavin-enzyme conjugates.
It has been shown previously that antibodies specic to ribitol (Rbt)
and ribitol-5-phosphate (Rbt-5-P) specically recognize riboavin and
FMN, respectively; however, the cross-reactivity of ribitol- and ribitol-
5-phosphate-specic antibodies (hereafter referred to as Rbt and Rbt-
5-P antibodies) is low for Rbt-5-P and Rbt, respectively (Ravi and
Venkatesh, 2014a, 2014b). Therefore, it appeared interesting to investi-
gate the feasibility of using Rbt antibodies for the quantication of ribo-
avin, and similarly, Rbt-5-P antibodies for the quantication of FMN in
foods. Since riboavin and FMN are small molecules (molecular mass
b 500 Da), an indirect competitive enzyme-linked immunoassay
(icELISA) format was utilized for their quantication. The present
study describes the development of an icELISA to quantify riboavin
and FMN in selected foods and pharmaceuticals by using hapten
afnity-puried Rbt and Rbt-5-P antibodies, respectively.
2. Materials and methods
Riboavin (USP grade), FMN, FAD, lumichrome (7,8-dimethylalloxazine),
thiamine, pantothenic acid, pyridoxine, biotin, folic acid, cyanocobalamin,
ascorbic acid, niacin and sodium dithionite (reducing agent; Na2S2O4)
were obtained from SigmaAldrich Chemical Co., St. Louis, MO. Flat
bottomed 96well ELISA microtiter plates were purchased from Greiner
Bio-One, Frickenhausen, Germany. Goat anti-rabbit IgG-alkaline phospha-
tase (ALP) conjugate and p-nitrophenyl phosphate (pNPP) were products
of Bangalore Genei, Bengaluru, India. Whatman 41 (ashless) lter paper
was purchased from Wipro GE Healthcare Pvt. Ltd., Chennai, India. All
other chemicals and reagents used were of analytical grade.
Wheat germ, wheat our, wheat bran, skimmed milk powder (Kar-
nataka Co-operative Milk Producers' Federation Ltd., Bengaluru, India)
and Kellogg's cornakes (Kellogg India Pvt. Ltd., Raigad, Maharashtra,
India) were purchased from the local market. Zevit (vitamin B-
complex capsule; contains riboavin, zinc sulfate monohydrate, thia-
mine mononitrate, nicotinamide, pyridoxine, cyanocobalamin, calcium
pantothenate, folic acid, biotin and vitamin C as active ingredients;
Glaxo SmithKline Pharmaceuticals Ltd., Bengaluru, India) and Polybion
SF (vitamin B-complex syrup; contains riboavin sodium phosphate,
 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx
thiamine hydrochloride, pyridoxine hydrochloride, nicotinamide, D-
panthenol, cyanocobalamin as active ingredients, and propylene glycol,
glycerine and sorbitol as solution base; Merck Ltd., Aurangabad, India)
were purchased from the local pharmacy.
2.1. Immunogens and antibodies for the immunoassays
Ribitol-KLH conjugate, ribitol-5-phosphate-KLH conjugate, and
hapten-specic (afnity-puried) rabbit Rbt and Rbt-5-P antibodies
were prepared as described previously by Ravi and Venkatesh (2014a,
2014b).
2.2. Preparation of food samples
2.2.1. Acid hydrolysis for TRF quantication (AOAC, 2012)
To 1 or 2 g of food sample placed in a 50 mL conical ask, 20 mL of
0.1 N HCl were added and autoclaved at 121 C and 15 psi for 30 min
to obtain TRF; the acid hydrolysis step converts FMN and FAD to free ri-
boavin. Proteins were removed by adjusting the pH sequentially
(twice or thrice) to 4.5 and 6.8 using 1 N HCl and 1 N NaOH, respectively,
in order to completely precipitate the proteins. It is important to keep
the pH below 7.2 during pH adjustment as riboavin gets destroyed
above pH 7.2. Finally, sample solutions were ltered using Whatman
41 lter paper. All extractions were carried out in the dark; the extracts
were stored at 4 C for b 24 h prior to analysis.
2.2.2. Non-degradative extraction method for avins (riboavin, FMN and
FAD)
This was essentially performed by the described method (Russell
and Vanderslice, 1992b). Aliquots of methanol (9.0 mL) and of methy-
lene chloride (10.0 mL) were added to the extraction tube containing
1 or 2 g of food samples and, homogenized at 25 C for 12 min. Then,
10 mL of 10 mM potassium phosphate solution (pH 5) were added to
the extract and homogenized for another minute. The extracts were
centrifuged at 5000 rpm for 10 min. The aqueous phase was decanted,
ltered using Whatman 41 lter paper. All extractions were carried
out in the dark to prevent photodegradation of the avins; the extracts
were stored at 4 C for b 24 h prior to analysis.
2.3. Generic indirect competitive enzyme-linked immunosorbent assay
(icELISA)
The icELISA format used in this study is based on similar formats de-
veloped in our laboratory for the quantication of erythritol and xylitol
in foods using rabbit hapten-specic erythritol and xylitol antibodies,
respectively (Sreenath and Venkatesh, 2008, 2010). Flat bottom poly-
styrene microtiter plate wells were coated with antigens or hapten-
protein conjugates at 100 ng/well in 0.1 M carbonate-bicarbonate buff-
er, pH 9.6 and incubated at 4 C overnight. The wells were washed thrice
between steps using 250 L of PBS containing 0.05% Tween-20 (PBS-T).
Blocking was done using 200 L of 1% gelatin in PBS-T at 37 C for
30 min. Afnity-puried antibodies (25 ng in 50 L PBS), and different
dilutions of competitors or food extracts in PBS (nal volume: 100 L)
were pre-incubated at 37 C for 1 h and added to the microtiter plate
wells and incubated at 37 C for 1 h. Goat anti-rabbit IgG-ALP secondary
antibody (1:5000 dilution in blocking buffer; 100 L/well) was added
and incubated at 37 C for 1 h. Color development was done using
pNPP (1 mg/mL; 100 L/well) in 1% diethanolamine buffer, pH 9.8, at
37 C for 30 min. The reaction was stopped by adding 3 M NaOH (50
L/well) and the absorbance was read at 405 nm in a microtiter plate
reader (model 680; Bio-Rad Laboratories, Hercules, CA). ELISA results
were expressed as B/B0, where B0 is the absorbance value of blank sam-
ple and B is the absorbance value of the sample with different concen-
trations of riboavin or FMN. The competition curves were obtained
by plotting normalized B/B0 versus riboavin or FMN concentration
with the 4-parameter logistic non-linear regression curve t mode in
Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
3
GraphPad Prism 5 software (GraphPad, San Diego, CA). The IC50 and
IC10 values were calculated using the 4-parameter logistic nonlinear re-
gression curve t mode in GraphPad Prism 5 software (GraphPad, San
Diego, CA). The 50% inhibition concentration (IC50) and the limit of de-
tection (LOD) dened as 10% inhibition concentration (IC10) were calcu-
lated from these calibration curves.
2.4. Competitive inhibition curve and calibration curve for riboavin and
FMN immunoassays
A dose-dependent competitive inhibition curve was prepared for
riboavin analysis by icELISA with afnity-puried Rbt antibodies and
different known amounts (0.00110,000 ng) of standard riboavin in
competition with the ribitol epitopes of ribitol-KLH antigen coated on
the microtiter plate wells. A standard calibration curve was also pre-
pared using 1100 ng of riboavin.
Similarly, a dose-dependent competitive inhibition curve was pre-
pared for FMN analysis by icELISA with afnity-puried Rbt-5-P anti-
bodies and different known amounts (0.00110,000 ng) of standard
FMN in competition for binding to the ribitol-5-phosphate epitopes of
ribitol-5-phosphate-KLH antigen coated on the microtiter plate wells.
A standard calibration curve was prepared using 1100 ng of FMN.
2.5. Specicity and cross-reactivity of the antibodies used in immunoassays
Specicity and cross-reactivity of the afnity-puried Rbt and Rbt-5-
P antibodies towards riboavin, FMN, FAD, photolytic product of riboa-
vin (lumichrome) and water-soluble vitamins (thiamine, pantothenic
acid, pyridoxine, biotin, folic acid, cyanocobalamin, ascorbic acid and ni-
acin) were studied by icELISA as described above. IC50 value of each
compound (inhibitor) was calculated using the 4-parameter logistic
nonlinear regression curve t mode in GraphPad Prism 5 software
(GraphPad, San Diego, CA). The percent cross-reactivity for various in-
hibitors was calculated using the formula [IC50 for ribitol (or ribitol-5-
phosphate) IC50 for test compound] 100.
2.6. icELISA for application in selected foods and pharmaceuticals
Selected food samples like wheat our, wheat germ, wheat bran,
Kellogg's corn akes and skimmed milk powder that had been subjected
to sample preparation were used as the source of competing analyte (ri-
boavin or FMN) in the immunoassay. The content of one capsule of
Zevit amounting to 5 mg riboavin was dissolved in distilled water;
for FMN quantication, 1 mL of Polybion SF syrup containing 0.5 mg
FMN was diluted in distilled water. All samples were diluted in distilled
water to obtain a nal analyte concentration of 1020 ng/mL and were
used in the immunoassay.
2.7. Recovery analysis
For the spiking and recovery analysis, riboavin or FMN was added
to skimmed milk powder at a nal concentration of 5, 10 or 20 ng/mL
during extraction, and in the case of Zevit capsule or Polybion SF
syrup, riboavin or FMN was added directly to diluted solutions at
nal concentration of 5, 10 or 20 ng/mL. From the optimized ELISA con-
ditions, icELISA was carried out in 3 sets within a microplate in a day
(intra-assay variability) and on three different days (inter-assay vari-
ability) in order to evaluate the precision and accuracy. The percent re-
covery was calculated based on the difference in the mean
concentration of the individual analyte in a spiked sample and an
unspiked sample using the following formulae:
%Recovery spiked sample valueunspiked sample value
amount of spike added  100
4 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx

2.8. HPLC analysis of riboavin and FMN in food extracts and
pharmaceuticals
Reversed-phase HPLC was performed on an Agilent LC system, the
separation being accomplished with a C18 analytical column (4.6
250 mm, 5.0 m particle size) by isocratic elution at a ow rate of
1 mL/min using mobile phase of 10 mM potassium phosphate
buffer:methanol (70:30, v/v). The sample injection volume was 20 L.
An excitation wavelength of 450 nm and an emission wavelength of
520 nm were used to detect both riboavin and FMN. Peak area of
analytes in question was measured and used for quantication.
did not show reactivity with other water-soluble vitamins, and its
cross-reactivity with other vitamins of the B-complex and vitamin C
was found to be b0.1%.
Considering 100% cross-reactivity for ribitol-5-P, Rbt-5-P antibodies
showed ~230% cross-reactivity towards FMN, ~10% cross-reactivity to-
wards riboavin and b 4% reactivity towards ribitol, FAD and
lumichrome (Table 1). The Rbt-5-P antibody did not recognize other
water-soluble vitamins, and its cross-reactivity with these molecules
was found to be b0.1%.
3.2. Immunoassay for quantication of riboavin and FMN

2.9. Spectrouorimetric analysis of riboavin in food extracts and
pharmaceuticals
The uorometric analysis was carried out as described in the AOAC
ofcial method (970.65) (AOAC, 2012), using an excitation wavelength
of 440 nm and an emission wavelength of 565 nm. Briey, to 10 mL of
food sample extract solution or riboavin standard solution
(100 ng/mL), 1 mL of glacial acetic acid was added and mixed well to
measure the uorescence intensity. Twenty milligrams of powdered so-
dium dithionite were added to the above solutions with mixing and the
uorescence intensity was measured within 5 s. On the basis of aliquots
taken, the riboavin concentration (in mg/mL nal sample solution)
was calculated using the formula


IQ = I0 Q 0  0:1  0:001
where I and I refer to uorescence intensities of test sample and riboa-
vin standard, respectively, and Q and Q refer to uorescence intensities
of test sample and riboavin standard, respectively, following sodium
dithionite addition.
To quantify riboavin, icELISA was carried out using afnity-puried
hapten-specic Rbt antibodies involving competition between ribitol
epitopes of the coating antigen (ribitol-KLH conjugate) on the microti-
ter plate wells with various amounts of riboavin; a typical inhibition
curve was obtained which is depicted in Fig. 2, panel a. The riboavin
concentration showing 50% decrease in maximal binding (50% B/B0 or
IC50) is calculated to be 3.41 ng/mL; the limit of detection (LOD),
taken as the concentration giving a 10% inhibition of maximal absor-
bance (10% B/B0 or IC10), is 0.02 ng/mL. Riboavin concentration be-
tween 1 and 100 ng/mL showed a linear relationship (R2 = 0.965, n
= 4) with obtained B/B0 values (Fig. 2, panel b).
Similarly, a dose-dependent inhibition curve was obtained for FMN
analysis by icELISA using afnity-puried Rbt-5-P antibodies involving
competition for binding to the ribitol-5-phosphate epitopes of ribitol-
5-phosphate-KLH antigen or FMN; this is shown in Fig. 3, panel a. IC50
and LOD (IC10) values were calculated to be 7.84 ng/mL and

3. Results

3.1. Specicity and cross-reactivity of the antibodies used in immunoassays

The percent crossreactivity of Rbt and Rbt-5-P antibodies towards

other related compounds and water-soluble vitamins is shown in
Table 1. Considering 100% cross-reactivity for ribitol, Rbt antibodies
showed ~ 800% cross-reactivity towards riboavin, b 3% cross-
reactivity towards Rbt-5-P, FMN, FAD and lumichrome. Rbt antibody

Table 1
Specicity and cross-reactivity of ribitol (Rbt) and ribitol-5-phosphate (Rbt-5-P) antibod-
ies towards related compounds and water-soluble vitamins.

Inhibitors Antibodies to Antibodies to

ribitol ribitol-5-phosphate

IC50 CRa IC50 CRa (%)

(ng/mL) (%) (ng/mL)

Ribitol 27.0 100.0 522.1 3.4

Ribitol-5-phosphate 926.4 2.9 17.9 100.0
Riboavin 3.4 792.1 182.5 9.8
FMN 1086.1 2.5 7.8 227.8
FAD 2246.3 1.2 2612.8 0.7
Lumichrome 6595.4 0.4 7686.4 0.2
Thiamine n.r.b b0.1 n.r.b b0.1
Niacin n.r. b0.1 n.r. b0.1
Pyridoxine n.r. b0.1 n.r. b0.1
Folic acid n.r. b0.1 n.r. b0.1
Biotin n.r. b0.1 n.r. b0.1
Pantothenic acid n.r. b0.1 n.r. b0.1
Cyanocobalamin n.r. b0.1 n.r. b0.1
Fig. 2. Inhibition and calibration curve for icELISA using riboavin (n = 4). (a) Inhibition
Ascorbic acid n.r. b0.1 n.r. b0.1
curve; (b) calibration curve. Coating antigen: ribitol-KLH conjugate (100 ng); primary
The signicance of bold refers to the cross-reactivity values of the haptens and analytes antibody: afnity-puried anti-ribitol antibodies (25 ng/well); competitor: standard
used in this study. riboavin at different concentrations; secondary antibody: goat anti-rabbit IgG-ALP
a
CR, cross-reactivity. conjugate (1:5000 dilution). In B/B0, B0 is the absorbance value of blank sample and B is
b
n.r., no response. the absorbance value of the sample at different concentrations of riboavin.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx 5

Table 2
Analysis of TRF, riboavin and FMN in selected foods and pharmaceuticals using
icELISA (n = 4).

Samples Reported or as icELISA

per label*
TRF Riboavin FMN
(mg/100 g)
(mg/100 g) (g/100 g) (g/100 g)

Skimmed milk powder 1.70* 1.61 0.06 1261.3 81.1 119.2 7.8
Kellogg's corn akes 1.30* 1.18 0.07 1018.0 56.6 14.7 3.5
Wheat bran 0.58 0.53 0.04 149.4 15.2 28.4 0.9
Wheat germ 0.50 0.47 0.08 128.7 06.8 32.5 1.2
Wheat our 0.19 0.17 0.03 46.4 04.2 13.2 0.1
Zevit capsule 10.00* n.a. 10.1 00.2a n.a.
Polybion SF 50.00* n.a. n.a. 48.4 2.1b

n.a. = not applicable.

a
mg/capsule.
b
mg/100 mL
*
The signicance of asterisks refers to the values shown on the label of the respective
product.

3.4. Recovery analysis

In order to evaluate the precision and accuracy of immunoassays,

Fig. 3. Inhibition curve and calibration curve for icELISA using riboavin-5-phosphate (n
= 4). (a) Inhibition curve; (b) calibration curve. Coating antigen: ribitol-5-phosphate-
KLH conjugate (100 ng); primary antibody: afnity-puried anti-ribitol-5-phosphate
antibodies (25 ng/well); competitor: standard riboavin-5-phosphate at different
concentrations; secondary antibody: goat anti-rabbit IgG-ALP conjugate (1:5000
dilution). In B/B0, B0 is the absorbance value of blank sample and B is the absorbance
value of the sample at different concentrations of FMN.
0.24 ng/mL, respectively. A linear calibration curve (R2 = 0.976, n = 4)
was obtained for B/B0 values plotted against FMN concentration in the
range of 1 to 100 ng/mL (Fig. 3, panel b).
3.3. Immunoassay for riboavin and FMN quantication in selected foods
and pharmaceuticals
Food samples were subjected to acid hydrolysis to obtain TRF, or ex-
tracted with methanol/methylene chloride to obtain non-degraded in-
dividual avins; these food extracts were used as the sources of
competing analyte in the icELISAs. TRF was detected using Rbt antibody
whereas FMN was detected using Rbt-5-P antibody. The amounts of
TRF, riboavin and FMN present in 100 g food samples like wheat
our, wheat germ, wheat bran, Kellogg's corn akes and skimmed
milk powder, and pharmaceuticals like Zevit capsule (containing ribo-
avin) and Polybion SF syrup (containing FMN) were calculated using
calibration curves prepared for ELISAs, and the results are presented in
Table 2. The amounts of TRF (per 100 g food samples) in wheat germ
(0.47 mg), wheat our (0.17 mg), wheat bran (0.53 mg), skimmed
milk powder (1.61 mg) and Kellogg's corn akes (1.18 mg) were
found to be 9095% of the reported values in the literature (USDA,
2015) or as declared on the label, and correlated well with the values
from HPLC and AOAC spectrouorometric methods. Further, the riboa-
vin content of Zevit capsule agreed perfectly with the label value;
similarly, the FMN content of Polybion SF syrup by immunoassay
was found to be in agreement with the label value.
skimmed milk powder, Zevit capsule and Polybion SF syrup were
used for spiking and recovery studies, with inter-assay and intra-assay
variability along with analysis of recovery and coefcient of variance;
the results of the analyses are presented in Tables 3 and 4. On the aver-
age, riboavin icELISA showed intra-assay recovery of 94105%, with CV
ranging from 1.23 to 3.53%; for inter-assay, the recovery was found to be
90108%, with CV ranging from 1.51 to 5.8%. In the case of icELISA for
FMN, intra-assay recovery was 95102%, with CV ranging from 1.51 to
6.24%; for inter-assay, the recovery was 94102%, with CV ranging
from 1.93 to 7.37%.
3.5. HPLC and spectrouorimetric analyses of riboavin and/or FMN in food
and pharmaceutical extracts
HPLC analysis of riboavin showed retention time of 6.27 min (Fig. 4,
panel a) with linear calibration curve (R2 = 0.999) for riboavin con-
centration in the range of 5 ng/mL to 10 g/mL as shown in Fig. 4,
panel b. Similarly, HPLC analysis of FMN showed retention time of
2.37 min (Fig. 5, panel a), with linear calibration curve (R2 = 0.998)
for FMN concentration in the range of 5 ng/mL to 10 g/mL as shown
in Fig. 5 (panel b).
Food samples analyzed by HPLC method for TRF, riboavin and FMN
(Table 5) showed values which are in good agreement with those ob-
tained by icELISAs for these two analytes. Spectrouorimetric analysis
of food samples using AOAC method for TRF (Table 5) resulted in values
comparable to those obtained with icELISA for TRF.
Quantication of individual avins (riboavin and FMN) from the
same food samples showed variation in their values compared to TRF;
in the case of wheat germ, wheat our and wheat bran, riboavin and
FMN amounted to 2030% and 510%, respectively, of TRF, and the bal-
ance of TRF might be deduced as FAD, suggesting that FAD is the most
abundant avin present in tissues representing 6070% of TRF
(Eitenmiller and Landen, 2010; Golbach et al., 2014; Rivlin, 2007). In
Kellogg's corn akes, riboavin amounted to 70% of TRF; this appears
to be due to the fortication of the cereal with riboavin. In skimmed
milk powder, riboavin accounts for 7580%, FMN 510% and FAD
1520% of TRF, and the values are comparable to the literature values
(Gliszczyska-Swigo and Kozioowa, 2000; Kanno et al., 1991; Lumley
and Wiggins, 1981; Vias et al., 2004). The results obtained with
Zevit capsule containing riboavin, and Polybion SF syrup contain-
ing FMN are also in good agreement with values reported on the respec-
tive pharmaceutical labels; other active ingredients and excipients do
not interfere in the immunoassay of riboavin and FMN.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
6 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx

Table 3
Recovery and CV analyses of riboavin quantication in a selected food and pharmaceutical by icELISA in intra-assay and inter-assay.

Sample Spiked (ng/mL) Intra-assay Inter-assay

Detected (ng/mL) Recovery (%) CV (%) Detected (ng/mL) Recovery (%) CV (%)

Skimmed milk powder 0 13.15 0.19 n.a. 1.51 13.63 0.61 n.a 4.54
5 17.86 0.22 94.20 1.23 18.15 0.27 90.40 1.51
10 22.99 0.65 98.40 2.83 22.98 0.95 92.70 4.13
20 32.56 0.99 97.05 3.04 32.45 0.62 94.10 1.93
Zevit capsule 0 10.04 0.33 n.a. 3.29 9.84 0.57 n.a. 5.80
5 15.17 0.55 102.60 3.53 14.80 0.32 99.20 2.14
10 20.58 0.26 104.60 1.28 20.60 0.89 107.60 4.33
20 30.19 0.43 104.30 1.45 30.15 0.73 101.55 2.44

CV = coefcient of variance; n.a. = not applicable.

4. Discussion pharmaceuticals; the cross-reactivity of riboavin antibodies towards

In the present study, Rbt and Rbt-5-P antibodies were used to develop
icELISAs for the quantication of riboavin (LOD: 0.02 ng/mL) and FMN
(LOD: 0.24 ng/mL), respectively. The cross-reactivity results displayed
high specicity of the Rbt and Rbt-5-P antibodies towards riboavin and
FMN, respectively. The high cross-reactivity of Rbt antibodies towards ri-
boavin may likely be attributed to the presence of hydrophobic residues
in the proximity of the binding sites of Rbt antibody interacting with the
isoalloxazone ring, as reported earlier (Ravi and Venkatesh, 2014a). The
remarkable specicity towards riboavin indicated the feasibility of utiliz-
ing the Rbt antibodies for riboavin quantication by icELISA. Similarly,
the recognition of the Rbt-5-P moiety of FMN by Rbt-5-P antibodies
(Ravi and Venkatesh, 2014b) strongly suggested the feasibility of utilizing
Rbt-5-P antibodies for FMN quantication by icELISA.
In both the immunoassays, the LODs (taken as the IC10 values in this
study) indicated excellent detectability of the immunoassay for riboa-
vin and FMN comparable to the limit of detection with other methods
like HPLC, uorometry (AOAC method) and immunoassay for riboavin
using antibodies to whole riboavin molecule (Eitenmiller and Landen,
2010; Russell, 2012; Wang et al., 2013). Thus, spiking and recovery anal-
ysis of food samples showed no signicant matrix effects with excellent
recovery in the range of 90108%, and the mean coefcient of variance
for inter-assay and intra-assay was found to be 2.3% and 4.4%, respec-
tively; these results signify the accuracy and precision of both immuno-
assays indicating that the proposed methods should be of use for the
routine quantication of riboavin and FMN in large number of food
samples in a single analysis.
The results of icELISA were compared with those of HPLC and AOAC
methods and the values are in close agreement with one another, thus
demonstrating the usefulness of the developed ELISA method for the
quantication of both riboavin and FMN in various food samples. It is
of interest to note that, recently, an indirect competitive immunoassay
(LOD: 1.07 ng/mL) has been developed by Wang et al. (2013) using rab-
bit polyclonal antibodies generated against riboavin conjugated to
cationized BSA with carbodiimide, which measures TRF in foods and
FMN has not been reported in their study (Wang et al., 2013).
Several commercial ELISAs are available for the quantication of ri-
boavin, mostly as TRF; in most of these, the antibodies are raised
against whole riboavin molecule using carrier protein conjugates of ri-
boavin derivatives many of which are not commercially available. The
advantages of the developed ELISA is the use of a simple sugar (D-ri-
bose) for the preparation of ribitol-protein conjugates by reductive
amination (Ravi and Venkatesh, 2014a), and Rbt antibodies can be
used for the quantication of either TRF or free riboavin in a sample.
Moreover, free FMN in a sample can be quantitated by utilizing Rbt-5-
P antibodies prepared by the reductive amination of proteins with the
phosphorylated derivative of D-ribose, viz., D-ribose-5-P (Ravi and
Venkatesh, 2014b).
Ribitol occurs naturally in very few plants and fungi in small
amounts with the exception of sweet vernal (Adonis vernalis), where
its concentration is 40,000 ppm (Duke, 2000; Pfyffer and Rast, 1980);
hence, ribitol is unlikely to interfere in the quantication of riboavin
in various foods in the present icELISA. Though numerous methods
have been reported for the determination of TRF content in foods and
other biological samples (Eitenmiller and Landen, 2010; Pinto and
Rivlin, 2013; Rivlin, 2007; Russell, 2012), this appears to be the rst
study in using Rbt and Rbt-5-P antibodies to develop icELISA for the in-
dividual detection of riboavin and FMN, respectively, in foods and
pharmaceuticals. In conclusion, the icELISAs developed in this study is
sensitive and seems to be of utility for the routine analysis of riboavin
and FMN independently in various biological samples, pharmaceuticals
and natural/processed foods.
Funding sources
This work was supported by the Council of Scientic and Industrial
Research (CSIR), New Delhi, India by an institutional project, MLP-
0094 to YPV. GR received a research fellowship from the University
Grants Commission (UGC), New Delhi.

Table 4
Recovery and CV analysis of FMN quantication in a selected food and pharmaceutical by icELISA in intra-assay and inter-assay.

Sample Spiked (ng/mL) Intra-assay Inter-assay

Detected (ng/mL) Recovery (%) CV (%) Detected (ng/mL) Recovery (%) CV (%)

Skimmed milk powder 0 1.34 0.08 n.a 6.24 1.24 0.09 n.a. 7.37
5 6.15 0.19 96.20 3.01 6.09 0.29 97.00 4.71
10 10.89 0.32 95.50 3.03 10.84 0.71 96.00 6.52
20 20.91 0.42 97.85 2.02 21.01 0.79 98.85 3.76
Polybion SF syrup 0 9.52 0.21 n.a. 2.19 9.47 0.37 n.a. 3.94
5 14.45 0.32 98.60 2.22 14.18 0.27 94.20 1.93
10 19.69 0.34 101.70 1.73 19.62 0.95 101.50 4.86
20 29.25 0.44 98.65 1.51 28.58 0.64 95.55 2.26

CV = coefcient of variance; n.a. = not applicable.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx 7

Table 5
Analysis of TRF, riboavin and FMN in selected foods and pharmaceuticals by HPLC and
AOAC methods (n = 3).

Sample HPLC AOAC

method

TRF Riboavin FMN TRF

(mg/100 g) (g/100 g) (g/100 g) (mg/100 g)

Skimmed milk powder 1.65 0.12 1345.9 49.8 113.5 5.4 1.71 0.13
Kellogg's corn akes 1.26 0.02 1063.6 46.9 11.8 2.8 1.31 0.09
Wheat bran 0.57 0.03 152.4 17.9 25.8 3.1 0.56 0.01
Wheat germ 0.51 0.01 123.5 05.3 31.2 2.3 0.50 0.06
Wheat our 0.19 0.01 49.2 05.1 13.6 0.6 0.19 0.04
Zevit capsule n.a. 10.2 00.4a n.a. n.a.
Polybion SF n.a. n.a. 49.06 1.2b n.a.

n.a. = not applicable.

a
mg/capsule
b
mg/100 mL.

Role of the funding source

The funding agency (CSIR, New Delhi) did not have any role in the
preparation of the article, in the study design, in the collection, analysis
and interpretation of data, in the writing of the report, and in the deci-
sion to submit the article for publication.

Acknowledgments

We thank Dr. Ram Rajasekharan, Director, CSIRCFTRI, Mysuru for

Fig. 4. Quantitation of riboavin by HPLC. (a) Representative HPLC chromatogram of
riboavin (RT = 6.266 min), and (b) calibration curve for riboavin. LU = luminescence
units.
Declaration of interest
The authors declare no conict of interest.
Fig. 5. Quantitation of FMN by HPLC. (a) Representative HPLC chromatogram of FMN (RT
= 2.373 min), and (b) calibration curve for FMN. LU = luminescence units.
his keen interest and encouragement.
References
AOAC, 2012. Ofcial Methods of Analysis of AOAC International. 19th ed. Gaithersburg,
Maryland, AOAC International, p. 2012.
Cataldi, T.R.I., Nardiello, D., Carrara, V., Ciriello, R., De Benedetto, G.E., 2003. Assessment of
riboavin and avin content in common food samples by capillary electrophoresis
with laser-induced uorescence detection. Food Chem. 82, 309314.
Cha, G.S., Meyerhoff, M.E., 1988. Solid phase enzyme-linked competitive binding assay for
riboavin. Anal. Biochem. 168, 216227.
Duke, J.A., 2000. Handbook of Phytochemical Constituents of GRAS Herbs and Other Eco-
nomic Plants: Herbal Reference Library. CRC Press, Boca Raton, Florida, p. 2000.
Eitenmiller, R.R., Landen Jr., W.O., 2010. Vitamin Analysis for the Health and Food Sci-
ences. CRC Press, Boca Raton, Florida, p. 2010.
Gliszczyska-Swigo, A., Kozioowa, A., 2000. Chromatographic determination of riboa-
vin and its derivatives in food. J. Chromatogr. A 881, 285297.
Golbach, J.L., Ricke, S.C., O'Bryan, C.A., Crandall, P.G., 2014. Riboavin in nutrition, food
processing, and analysis a review. J. Food Res. 3, 2335.
Gopalan, C., Sastri, B.V.R., Balasubramanian, S.C., 1989. Nutritive Value of Indian Foods.
Indian Council of Medical ResearchNational Institute of Nutrition, Hyderabad,
India, p. 1989.
Kanno, C., Shirahuji, K., Hoshi, T., 1991. Simple method for separate determination of
three avins in bovine milk by high performance liquid chromatography. J. Food
Sci. 56, 678681.
Kim, M.J., Kim, H.J., Kim, J.M., Kim, B., Han, S.H., Cha, G.S., 1995. Homogeneous assays for
riboavin mediated by the interaction between enzyme-biotin and avidin-riboavin
conjugates. Anal. Biochem. 231, 400406.
Kozik, A., 1996. Microtitre-plate enzyme-linked ligand-sorbent assay of riboavin (vita-
min B2) in human plasma and urine. Analyst 121, 333337.
Lopez-Anaya, A., Mayersohn, M., 1987. Quantication of riboavin, riboavin 5-
phosphate and avin adenine dinucleotide in plasma and urine by high-
performance liquid chromatography. J. Chromatogr. 423, 105113.
Lumley, I.D., Wiggins, R.A., 1981. Determination of riboavin and avin mononucleotide
in foodstuffs using high-performance liquid chromatography and a column-
enrichment technique. Analyst 106, 11031108.
Massey, V., 2000. The chemical and biological versatility of riboavin. Biochem. Soc. Trans.
28, 283296.
Prez-Ruiz, T., Martnez-Lozano, C., Sanz, A., Toms, V., 1994. Photokinetic determination
of riboavin and riboavin 5-phosphate using ow injection analysis and chemilu-
minescence detection. Analyst 119, 18251828.
Petteys, B.J., Frank, E.L., 2011. Rapid determination of vitamin B (riboavin) in plasma by
HPLC. Clin. Chim. Acta 412, 3843.
Pfyffer, G.E., Rast, D.M., 1980. Systematic distribution of ribitol in the fungi. New Phytol.
85, 163168.
Pinto, J.R., Rivlin, R.S., 2013. Riboavin (vitamin B2). In: Zempleni, J., Suttie, J.W. (Eds.),
Handbook of Vitamins. 2013. CRC Press, Boca Raton, Florida, pp. 191266.
Ravi, G., Venkatesh, Y.P., 2014a. Recognition of riboavin and the capsular polysaccharide
of Haemophilus inuenzae type b by antibodies generated to the haptenic epitope D-
ribitol. Glycoconj. J. 31, 247258.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010
8 G. Ravi, Y.P. Venkatesh / Journal of Immunological Methods xxx (2017) xxxxxx

Ravi, G., Venkatesh, Y.P., 2014b. Recognition of avin mononucleotide, Haemophilus
inuenzae type b and its capsular polysaccharide vaccines by antibodies specic to
D-ribitol-5-phosphate. Glycoconj. J. 31, 573585.
Rivlin, R.S., 2007. Riboavin (vitamin B2). In: Rucker, R.B., Zempleni, J., Suttie, J.W.,
McCormick, D.B. (Eds.), The Handbook of Vitamins. 2007. Marcel Dekker, New
York, New York, pp. 255273.
Russell, L.F., 2012. In: Nollet, L.M.L., Toldra, F. (Eds.), In Food Analysis by HPLC, third ed.
2012. Marcel Dekker, New York, New York, pp. 325442.
Russell, L.F., Vanderslice, J.T., 1992a. Comments on the standard uorometric determina-
tion of riboavin in foods and biological tissues. Food Chem. 43, 7982.
Russell, L.F., Vanderslice, J.T., 1992b. Non-degradative extraction and simultaneous quan-
titation of riboavin, avin mononucleotide, and avin adenine dinucleotide in foods
by HPLC. Food Chem. 43, 151162.
Selva Kumar, L.S., Thakur, M.S., 2011. Competitive immunoassay for analysis of vitamin
B(12). Anal. Biochem. 418, 238246.
Snell, E.E., Strong, F.M., 1939. A microbiological assay for riboavin. Ind. Eng. Chem. Anal.
Ed. 11, 346350.
Sreenath, K., Venkatesh, Y.P., 2008. Analysis of erythritol in foods by polyclonal antibody-
based indirect competitive ELISA. Anal. Bioanal. Chem. 391, 609615.
Sreenath, K., Venkatesh, Y.P., 2010. Quantication of xylitol in foods by an indirect com-
petitive immunoassay. J. Agric. Food Chem. 58, 12401246.
USDA, 2015. USDA Agricultural Research Service, Nutrient Data Laboratory. USDA Nation-
al nutrient database for standard reference, release 27 (revised). Version current:
May 2015. http://www.ars.usda.gov/ba/bhnrc/ndl (24 August 2016).
Valls, F., Sancho, M.T., Fernndez-Muio, M.A., Checa, M.A., 1999. Determination of total
riboavin in cooked sausages. J. Agric. Food Chem. 47, 10671070.
Vias, P., Balsalobre, N., Lpez-Erroz, C., Hernndez-Crdoba, M., 2004. Liquid chromato-
graphic analysis of riboavin vitamers in foods using uorescence detection. J. Agric.
Food Chem. 52, 17891794.
Wang, P., Yin, Y., Eremin, S.A., Rybakov, V.B., Zhang, T., Xu, Z., Ren, L., He, X., Meng, M., Xi,
R., 2013. Indirect competitive immunoassay for detection of vitamin B2 in foods and
pharmaceuticals. J. Agric. Food Chem. 61, 70487054.
Wild, D., 2013. The Immunoassay Handbook: Theory and Applications of Ligand Binding,
ELISA and Related Techniques. fourth ed. Elsevier Science, Amsterdam, The
Netherlands, p. 2013.
Zand, N., Chowdhry, B.Z., Pullen, F.S., Snowden, M.J., Tetteh, J., 2012. Simultaneous deter-
mination of riboavin and pyridoxine by UHPLC/LC-MS in UK commercial infant
meal food products. Food Chem. 135, 27432749.
Zhang, T., Xue, H., Zhang, B., Zhang, Y., Song, P., Tian, X., Xing, Y., Wang, P., Meng, M., Xi, R.,
2012. Determination of folic acid in milk, milk powder and energy drink by an indi-
rect immunoassay. J. Sci. Food Agric. 92, 22972304.

Please cite this article as: Ravi, G., Venkatesh, Y.P., Immunoassays for riboavin and avin mononucleotide using antibodies specic to d-ribitol
and d-ribitol-5-phosphate, J. Immunol. Methods (2017), http://dx.doi.org/10.1016/j.jim.2017.03.010