Immunological characterization of riboflavin carrier proteins using monoclonal antibodies

Molecular Immunology, Vol. 24, No. 9, pp. 969-974, Printed in Great Brhin.

Ol61-589Oj87 $3.00 + 0.00 Pergamon Journals Ltd

1987

IMMUNOLOGICAL CHARACTERIZATION RIBOFLAVIN CARRIER PROTEINS USING MONOCLONAL ANTIBODIES

OF

SANDHYA S. VISWESWARIAH, ANJALI A. KARANDE and P. RADHAKANTHA ADIGA* Department of Biochemistry, Indian Institute of Science, Bangalore012, India (Received

7 November

1986; accepted

in revised form

19 January

1987)

Abstract--Monoclonal antibodies to chicken riboflavin carrier protein have been produced by fusing immunized mouse spleen cells with myeloma SP,/O-Ag 14. The three different monoclonal antibodies with respect to specifically bound ‘251-labelled chicken riboflavin carrier protein and were characterized their affinities to bind the antigen, subclass and isotype. These three monoclonal antibodies had similar affinities for holo-, apo- and SDS-denatured riboflavin carrier protein but were unable to recognize the reduced and carboxymethylated protein indicating that they were directed to specific conformational epitopes on the native avian protein. Succinylation of the vitamin carrier protein while still retaining flavin binding characteristics totally abolished the cross-reactivity with all the three monoclonal antibodies indicating that lysine residues were involved at the antigenic sites of the protein. This shows that the antigenic loci may be distinct from the flavin binding sites in the protein. All three antibodies were able to recognize riboflavin carrier protein present in the sera of pregnant rats, monkeys and humans indicating that the epitopes to which they are directed are conserved throughout evolution. These antibodies can

therefore be effectively used for radioimmunoassays protein in higher mammals.

and further studies on the functional aspects of this

latory concns of riboflavin carrier proteins (RCPs) in both the rat and the bonnet monkey (M. radiata) are modulated by oestrogen in a fashion analogous to the chicken protein (Adiga and Murty, 1983). These mammalian RCPs appear to be essential for transplacental transport of riboflavin during gestation, since selective neutralization of rat or monkey RCP by antibodies to cRCP causes drastic curtailment of flavin uptake by the foetoplacental unit (Murty and Adiga, 1981; Krishnamurthy et al., 1984) and culminates in pregnancy termination due to foetal wastage (Murty and Adiga, 1982; Adiga et al., 1986). The above investigation employed polyclonal antiserum to the cRCP for radioimmunoassays (RIAs) and immunoneutralization of mammalian RCPs. It was therefore reasonable to hope that monoclonal antibodies (MAbs) produced by the hybridoma technique of Kahler and Milstein (1975), would provide a specific tool for epitope analysis of the evolutionarily conserved protein in different mammals, and help to define the regions of the cRCP molecule essential for flavin binding and for recognition of the putative oocyte receptor involved in protein endocytosis. In this communication, we report the generation and characterization of MAbs to cRCP. In view of the fact that these MAbs do not differentiate between holo-, apo- and partially denatured RCP, they have the potential for effecive use for RIAs for RCP quantification in biological materials. Moreover, these MAbs are able to recognize RCPs isolated from mammals including humans, showing that the epitopes recognized by the monoclonals are remarkably conserved throughout vertebrate evolution.

INTRODUCTION Chicken riboflavin carrier protein (cRCP) is a phosphoglycoprotein (M, 37,000) present in the egg white, yolk, and serum of laying hens (Rhodes et al., 1959; Ostrowski et al., 1962; Winter et al., 1967). This vitamin carrier is an oestrogen inducible gene product (Murthy and Adiga, 1978; DurgaKumari and Adiga, 1986) and functions as the obligatory mediator of circulatory transport and oocyte deposition of riboflavin for use by the developing avian embryo (Farrell et al., 1969). We have been interested in studying the evolutionary conservation of this protein, in terms of its structure and functions, in view of the finding that, by using specific polyclonal antiserum to the egg white cRCP, a protein crossreacting with the avian vitamin carrier can be detected in the pregnancy sera of several higher animals including primates (Adiga and Murty, 1983). These mammalian RCPs exhibit physicochemical characteristics such as p1 (<4), M, (- 37,000) and hormonal modulation very similar to cRCP, and show a preferential affinity towards riboflavin compared to FMN and FAD (Visweswariah, 1986). Furthermore, circu-

*Author to whom correspondence should be addressed. Abbreviations: BSA, bovine serum albumin; cRCP, chicken riboflavin carrier nrotein: DTT. dithiothreitol: EDTA. ethylene diamine tetraacetic acid; FAD, flavin adenine diphosphate; FMN, flavin mononucleotide; IgG, immunoglobulin G; i.p., intra-peritoneal; MAb, monoclonal antibody; M,, Mol. wt; PEG, polyethylene glycol; RCP, riboflavin carrier protein, RIA, radioimmunoassay; SDS, sodium, dodecyl sulphate. 969

SANDHYA

970

S.

VKWESWARIAH et al.

MATERIALS AND METHODS

Purl$cution of cRCP. The purification of RCP from the chicken egg white was as detailed earlier (Murthy and Adiga, 1977), and the preparation was homogeneous by standard electrophoretic and immunological criteria (Weber and Osborn, 1969). Protein estimation was performed by the dye-binding assay (Bradford, 1976). Purified apo-cRCP was incubated with an excess of riboflavin to saturate the protein with the vitamin and then subsequently used for RIA. The radioiodination of cRCP was performed by the iodogen method (Fraker and Speck, 1978) and the ‘2SI-labelled protein had a sp. act. of 3.2-7 mCi/mg RCP. Immunization and fusion.Swiss mice (3-4 months old, 25 g body wt) of either sex were immunized with a subcutaneous priming dose of holo-cRCP (50 pg) emulsified with Freund’s complete adjuvant (Difco). Two booster injections of the same amount of the antigen were given in incomplete adjuvant at 3-weekly intervals following which the mice were rested for 2-3 months. The spleen was removed and splenocytes recovered for cell fusion on the fourth day following a single in LGXIinjection (i.p.) of 200 pg cRCP in saline. Blood was collected from the mice just before splenectomy to monitor circulating antibody titres by RIA. This particular immunization protocol was found to be optimal in obtaining the maximum number of antibody secreting hybrids subsequently. If, on the other hand, multiple small doses of cRCP were administered for 7 days prior to splenocyte isolation for fusion, very few antibody secreting hybrids were obtained, even though circulating antibody titres were relatively higher at this time (data not shown). Cell fusion was performed with the myeloma line SP,/O-Ag 14 (Shulman et al., 1978) according to the method of Westerwoudt ef al. (1984) and fused splenocytes plated at 0.2 x l@/microtitre well. Wells containing hybrid clones were screened for the presence of antibodies between 12 and 15 days following fusion. Hybrid cells in the antibody positive wells were subcloned by limiting dilution on feeder cells, and single cluster wells tested for antibody production. Positive clones were expanded and culture supernatants stored at -20°C until used. Screening procedures. Solid phase immunoassays were performed by incubating culture supernatants with cRCP (0.1-I pg/well) coated to polystyrene plates (Nunc). Antibodies bound to the wells were detected either by addition of anti-mouse IgG conjugated to alkaline phosphatase (Sigma) or by assaying to the binding of “51-labelled protein A subsequent incubation with rabbit anti-mouse IgG antiserum (Brown et al., 1983). In the second assay, after incubation of immobilized cRCP with culture supernatants, an appropriate dilution of rabbit anti-mouse IgG serum was added to each well and left for 60 min at 37°C. This was followed by incubation with radio-

iodinated protein A. The inclusion of the anti-mouse IgG serum enhanced the ‘251-protein A binding 20-fold, when compared to binding of “‘I-protein A to mouse IgG alone (Brown et al., 1983). Liquid phase RIAs were performed by incubating the culture supernatants with ‘251-labelled cRCP and the antibody-bound ratioactivity quantitated after double antibody precipitation. Briefly, ‘2sI-labelled cRCP (lo- ‘OM) was incubated with varying vols of supernatants from confluent cultures of different clones for 6 hr at 37°C in a total vol of 300~1 of phosphate buffered saline (containing 0.2% BSA and 0.02% sodium azide). One hundred microlitres of mouse serum (1: 100 diluted) and 200 k 1 of 1: 10 diluted goat antiserum to mouse IgG were added and incubation continued for 15 min. PEG 6000 was added at a final concn of 2.5% and incubated on ice for 15 min prior to centrifugation and quantification of radioactivity in the pellet measured in a gamma counter (LKB, Model 1275 minigamma). For determination of affinity constants, aliquots of culture supernatants capable of binding 50% of input radioactivity in the absence of inhibitor were incubated at 37°C for 8-16 hr with f251-labelled RCP and varying known concns of unlabelled holo-cRCP. Specifically bound radioactivity was monitored by the double antibody PEG-precipitation method. The affinity constants were calculated (Miiller, 1983) after determining the maximum amount of “‘1-labelled cRCP immunoprecipitable with each antibody precipitation (Murthy and Friesen, 1985). Subclass and isotype determination c.$ MAhs. The subclass and isotype of the various MAbs were determined by solid phase assay employing ‘2SI-labelled protein A (Brown et al., 1983). Rabbit antisera specific to the mouse light chains and subclasses (Miles) were added to the wells with the MAbs bound to the cRCP coated-polystyrene plates. Detection and quantification of the bound antibodies was performed by adding “SI-labeiled protein A to the wells and measuring the bound radioactivity. SDS-denaturation oj” cRCP. For the SDS denaturation of cRCP, a solution of RCP in water (10 mg/ml) was treated with SDS (1% w/v final concn) at 100°C for 10 min. It was found unnecessary to remove SDS from the protein preparation since the final concn of SDS (0.01%) in the protein solution after appropriate dilution prior to RIA, did not interfere with the binding of the antibody to radiolabelled cRCP. Succinyiation of cRCP. The holo-cRCP was succinylated at pH 8.0 by the method of Klapper and Klotz (1972). Briefly, succinic anhydride (50 mol,‘mol of peptidyl lysine) was added in small portions over a 1 hr period to a stirred solution maintained at pH 8.0. Unreacted anhydride was removed by gel filtration on a prepacked PD-10 column (G-25, Pharmacia), and protein appropriately diluted prior to RIA. Reduction and carhoxymethylation. This was per-

971

Monoclonal antibodies to riboflavin carrier protein formed according to the method of Konigsberg (1972). A 2% (w/v) solution of cRCP was prepared in 0.5 M Tri-HCl buffer, pH 8.1, containing 6M guanidine hydrochloride and 0.002 M EDTA. After flushing with nitrogen, the vial was kept at 50°C for 30 min, following which DTT (50 mol/mol of S-S in protein) was added. Reduction was allowed to proceed for 4 hr at 50°C and the solution was then warmed to 25°C. Iodoacetamide (100 mol/mol of disulphide) was added and the solution kept for 20min in the dark, after which it was desalted as described earlier. Cross-reactivi(ty of antibodies with ape, holo- and modljied cRCP. Three MAbs were selected for investigating their cross-reactivities with the native and chemically modified forms of cRCP. Antibodies at dilutions capable of binding 50% of input radioactivity (cpm) were incubated with varying known concns of apo-, holo- and SDS-treated cRCP along with ‘251-labelled cRCP for 16 hr at 37°C. Specifically bound radioacttvity was determined by the double antibody PEG-precipitation method. The affinities of apo-, holo- and SDS-treated protein to the different MAbs were determined by a Scatchard analysis as described earlier. Cross-reactivity of antibodies with mammalian RCPs. We have earlier reported that mammalian RCPs have very similar physicochemical properties to cRCP and these include pl and M, (Visweswariah, 1986). Consequently, pregnant rat, monkey and human sera were dialysed individually against 0.1 M sodium acetate buffer, pH 4.5 and subjected to ionexchange chromatography on DEAE-cellulose at pH 4.5. Proteins bound to the anion exchanger were eluted with I M NaCl in acetate buffer, dialysed and concentrated. These DEAE-protein fractions were enriched in the RCPs as shown earlier (Visweswariah, 1986) and were subsequently used to determine their cross-reactivity with the three MAbs. The three antibodies at dilutions capable of binding 2&40% of input radioactivity (cpm) were incubated with varying known concns of either rat RCP, monkey RCP or human RCF’ for 6 hr. “51-labelled cRCP was then added and incubation continued for another 6 hr. Specifically bound radioactivity was determined by double antibody precipitation, RESULTS

The immunization protocol described above elicited circulating antibodies to cRCP in all the mice immunized. However, antibody titres as defined by the dilution of antibody capable of binding 20% of input ‘251-~RCP varied from 1:400 to 1: 10,000. More than 50% of the 200 wells which were plated with the hybrid cells contained specific antibodies to cRCP. All the hybrid cell lines were stable in terms of antibody secretion for more than a month in culture. Subcloning by limiting dilution was performed only after the cell lines were well-established to culture

1 Volume

10 of Supernatant

(~1)

Fig. 1. Binding of ‘251-labelled cRCP to hybridoma supernatants. Doubling dilutions of supernatants (0.1-50 ~1) from confluent cultures of antibody secreting clones were incubated with ‘251-labelled cRCP (ca 50,000 cpm) for 6 hr at 37°C in a total vol of 0.3 ml. Specific radioactivity was monitored by the PEG-precipitation method as detailed in Materials and Methods.

conditions (Westerwoudt et al., 1984). We have been able to adapt and grow the hybridomas in 2% foetal calf serum and 5% newborn calf serum satisfactorily. Three hybrid cell lines were chosen at random for further characterization and subcloning. All the MAbs secreted by these three hybrid cell lines bound ‘251-labelled cRCP effectively as measured by liquid phase RIA (Fig. 1). Fifty percent of maximum binding was obtained with as little as 0.4 ~1 of supernatant from 5Bl hybrids whereas 6 ~1 of supernatant was required in the case of 6HlO clone and 8 ~1 of 5GlD6 clone. This is probably a reflection of the relative affinities of the respective MAbs for the holo-cRCP (Heyningen et al., 1983) and was confirmed by determining the affinities of the various MAbs by Scatchard analysis. Figure 2 shows that increasing concns of unlabelled holo-cRCP were able to inhibit the binding of 12SI-labelled cRCP from each of these antibody preparations, in a linear dose-dependent manner. Calculation of the affinities of the antibodies for holo-cRCP according to the method of Miiller (1983) showed that all the plots were linear (Fig. 2, insets) indicating the presence of a single class of binders in each of these culture supernatants. The binding affinities calculated for the various monoclonal antibodies are summarized in Table 1, which also indicates the subclass and light chain type of the three MAbs; all of these have light chains of K type. It was ascertained that 6HlO MAb is of IgGl subclass while those of 5Bl and 5Gl D6 belong to the IgG2a subclass. It was of interest to determine whether the MAbs show an appreciable difference in terms of their affinities to apo- and holo-cRCP and whether they would recognize the SDS-treated antigen. Apo- and holo-cRCPs were prepared as described in Materials and Methods and used in the homologous RIA. The results shown in Table 2 clearly reveal that, despite significant conformational changes that accompany

912

SANDHYA

S. VISWETSWARIAH et al. Table

60

Chicken

RCP (M)

Fig. 2. Displacement of ‘251-i~belled cRCP from monoclonal antibodies. Varying concns of cRCP (lo-“-IO-‘M) were incubated with ‘251-labelled cRCP (IO-” M) along with a concn of MAb sufficient to bind 50% of input radioactivity, Immunoprecipitation was performed by the double anti-

body method. Inset shows the Scatchard analysis of the binding of MAbs to cRCP.

the interaction of riboflavin with cRCP (Nishikimi and Kyogoku, 1973), there is no appreciable alteration in affinities of the protein to the three MAbs, as a consequence of ligand binding to the apo-cRCP. To find out if the MAbs are directed solely to conformation-specific antigenic determinants on the protein, cRCP was treated with 1% SDS to perturb its conformational characteristics and then tested in RIA. Table 2 shows that this treatment did not lead to significant differences between SDS-treated cRCP and the holo-protein in terms of their relative affinities to the selected MAbs. It has previously been noted, however, that cRCP treated with SDS has a significantly reduced affinity for riboflavin (Murthy et al., 1976). It would appear therefore that the flavin binding site is distinct from the major antigenic determinants recognized by the three MAbs. However, total denaturation of RCP by reduction and

Table 2. Cross-reactwity

1. Characteristics

of monoclonal

Clone

Light chain type

IgG subclass

5GlD6 5Bl 6H10

K K K

IgG2a IgG2a IgGl

5GlD6 5Bl 6HlO

The above results represent the first report on the production and characterization of MAbs to cRCP. SDS gel electrophoresis of ‘251-labelled holo-RCP associated with the MAbs indicates that the antibodies recognize intact native protein (data not shown). The fact that the MAbs are able to recognize apo- and holo-RCP to comparable extents suggests that the high-affinity antibodies may be useful for homologous RIAs in place of the polyclonal antisera used to date. The specificities of the antibodies as

of monoclonal

antibodies

with modified

n.i. = no inhibition. Values in parentheses

(5.0 x IO*) (1.3 x IO? (9.0 x IO*)’

represent

RCP

Apo-RCP

Holo-RCP

80 1.5

5 x 108 1.3 x 10’0 9 x IO’

DISCUSSION

cRCP (ng) Reduced carboxy-

SDS-treated

75 1.4 60

to cRCP

K, (litres/mol)

carboxymethylation eliminates recognition of the modified protein by the MAbs (Table 2). This shows that none of these MAbs recognize a linear sequence of amino acids, per se and that at least a partial native conformation of cRCP is essential for interaction with the antibodies. It is intriguing that succinylated cRCP is also not able to inhibit the binding of “‘I-native cRCP to these MAbs even at a IOO-fold excess concn, indicating that lysine residues are involved in the recognition of the protein by these antibodies. The results with polyclonal antiserum also show a similar pattern (Ramanathan et al., 1980). It is interesting to note that succinylated RCP is still able to bind riboflavin significantly (Ramanathan et al., 1980) indicating that lysine residues are not critically involved at, or near, the ligand-binding site. The results in Fig. 3 clearly show that the epitopes to which the three MAbs are directed on cRCP are conserved in RCPs from the three mammalian species. Each antibody could recognize the mammalian RCPs as indicated by the inhibition of binding of ‘251-labelled cRCP to the respective MAb. This remarkable cross-reactivity observed with MAbs is confirmatory to the earlier reports using polyclonal antisera (Adiga and Murthy, 1983) and further emphasises the high degree of evolutionary conservation amongst RCPs.

Amount of modified cRCP required for 50% inhibition Monoclonal antibody

antibodies

_~

methylated ~ -

Succinylated

(4.8 x IO”)

125

(3.2 x IO”)

“.I.

(1.16 x IO”)

2

(0.9 x 10’0)

n.i.

n.i. n.i.

(9.0 x IO”)

60

(9.0 x 10X)

n.i.

“.I.

60

the K,, of each antIbody

for the modified

protein

in litres/mol.

Monoclonal

antibodies

to riboflavin

carrier

protein

913

gies amongst the proteins in the various species, as an approach to understanding the role such RCPs may have during embryonic development in mammals. Acknowledgements-The financial assistance from the Indian Council of Medical Research, New Delhi is gratefully acknowledged. We wish to thank Miss Vani R. for competent technical assistance and help in preparation of the manuscript.

” 10

1

m

100

REFERENCES

o+*c,ml

““”

IO ’ I

’

100 c RCP (“g/tube)

r/m/h

“““’

’

1000

1W fr':ctmn ifJg/tube)

RCP

Fig. 3. Displacement of ‘Z51-labelled cRCP from monoclonal antibodies. A vol of culture supernatant of confluent clones was taken such that 25% of the input radioactivity was bound and incubated with unlabelled cRCP or fractions obtained from ion-exchange chromatography of pregnant rat (r), human (h) or monkey (m) serum for 12 hr at 37°C. ‘251-labelled cRCP (ca 50,OOOcpm) was added and incubation continued for a further 6 hr at 37°C. Immunoprecipitation was performed by the double antibody method.

indicated either

by the results totally

ymethylation) antibodies

or

MAbs

by

most to

succinylation

cRCP

by suggest

carboxthat

the

exclu-

conformations

with cRCP.

significant cRCP

by modifying

followed

are able to recognize

sively associated The

obtained

(reduction

are

finding able

to

is the recognize

fact

that the

the

RCPs

purified from higher mammals, thereby underscoring the extensive evolutionary conservation of this protein. This implies that the epitopes on cRCP to which these MAbs are directed are present in mammalian RCPs, implying therefore, that the primary sequences of all the proteins are retained to near identity. MAbs are able to distinguish even a few amino acid changes in the sequence within an epitope (Harris, 1983) and the results presented above indicate that few major changes in the conformational characteristic have occurred in the RCPs through evolution. It is clear therefore that these MAbs can profitably be used to probe deeper into the regions of structural homolo-

Adiga P. R. and Murty C. V. R. (1983) Vitamin carrier proteins during embryonic development in birds and mammals. In Molecular Biology of Egg Maturation, Ciba Foundation Symposium 98, pp. 111.Pitman, London. Adiga P. R., Seshagiri P. B., Malathy P. V. and Sandhya S. V. (1986) Reproduction specific vitamin carrier proteins involved in transplacental vitamin transport in mammals including primates. In Pregnancy Proteins in Animals (Edited by Hau, J.), pp. 317-329. Walter de Gruyter, Berlin. Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilising the principle of protein-dye binding. Analyi. Biochem. 72, 248. Brown J. P., Hellstrom K. E. and Hellstrom I. (1983) Indirect ‘251-labelled protein A assay for monoclonal antibodies to cell surface antigens. Meih. Enzym. 92, 160. DurnaKumari B. and Adiaa P. R. (1986) Hormonal induction of riboflavin carrier protein in the chicken oviduct and liver: a comparison of kinetics and modulation. Molec. cell. Endocr. 44, 285. Farrell H. M., Mallette M. F., Buss E. G. and Clagett C. 0. (1969) The nature of the biochemical lesion in avian riboflavinuria-III. The isolation and characterisation of the riboflavin binding protein from egg albumen. Biochim. biophys. Acta 154, 433. -Fraker P. J. and Soeck J. C. (1978) Protein and cell membrane iodinat;ons with a ‘sparingly soluble chloramide, 1,3,4,6-tetrachloro-3a,6a-diphenylglycouril. Biothem. Biophys. Res. Commun. 80, 849. Harris H. (1983) Application of monoclonal antibodies in enzyme genetics. A. Reu. Genet. 17, 219. Heyningen V. V., Brock D. J. H. and Heyningen S. V. (1983) A simple method for ranking the affinities of monoclonal antibodies. J. Immun. Meth. 62, 147. Klapper M. H. and Klotz I. M. (1972) Acylation using dicarboxylic anhydrides. Meth. Enzym. 25, 531. Kiihler G. and Milstein C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495. Konigsberg W. (1972) Reduction of disulphide bonds in proteins with dithiothreitol. Meth. Enzym. 25, 185. Krishnamurthy K., Surolia N. and Adiga P. R. (1984) Mechanism of foetal wastage following immunoneutralisation of riboflavin carrier protein in the pregnant rat: disturbances in flavin co-enzyme levels. FEBS Left. 178, 87. Mtiller R. (1983) Determination of affinity and specificity of antibodies by competitive radioanti-hapten immunoassay. Meth. Enzym. 92, 589. Murthy G. S. and Friesen H. G. (1985) Practical considerations in analyzing radioreceptor assays by Scatchard analysis and capacity determination in irreversible hormone receptor interactions. J. Biosci. 7, 135. Murthy U. S. and Adiga P. R. (1977) Riboflavin binding protein of hen’s egg: purification and radioimmunoassay. Indian J. Biochem. Biophys. 14, 118. Murthy U. S. and Adiga P. R. (1978) Oestrogen induction of riboflavin binding protein in immature chicks. Biothem. J. 170, 331.

974

SANDHYA S. VISWE~WARIAH et al.

Murthy U. S., Podder S. K. and Adiga P. R. (1976) The interaction of riboflavin with a protein isolated from hen’s egg white: a spectrofluorometric study. Biochim. biophys. Acta 434, 69. Murty C. V. R. and Adiga P. R. (1981) Mechanism of foetal wastage following immunoneutralisation of riboflavin carrier protein in the pregnant rat. FEBS L&r. 135, 281. Murty C. V. R. and Adiga P. R. (1982) Pregnancy suppression by active immunisation against gestation-specific riboflavin carrier protein. Science 216, 191. Nishikimi M. and Kyogoku Y. (1973) Flavin-protein interaction in egg white flavoprotein. J. Biochem., Tokyo 73, 1233. Ostrowski W., Skahzynski B. and Zak Z. (1962) Isolation and properties of flavoprotein from the egg yolk. Biochim. biophys. Acta 59, 515. Ramanathan L., Guyer R. B., Karakawa W. W., Buss E. G. and Clagett C. 0. (1980) Chemical modifications of riboflavin binding protein: effects on function and antigenicity. Molec. Immun. 17, 267.

Rhodes M. B., Bunnet N. and Feeney R. E. (1959) The flavoprotein-apoprotein system of egg white. J. biol. Chem. 234, 2054. Shulman M., Wilde C. D. and Kohler G. (1978). A better cell line for making hybridomas secreting specific antibodies Nature 276, 269. Visweswariah S. S. (1986) Studies on riboflavin carrier proteins: physicochemical, biosynthetic and immunological aspects. Ph.D. thesis, Indian Institute of Science, Bangalore. Weber K. and Osborn M. (1969) The reliability of molecular weight determinations by dodecyl-sulphate polyacrylamide gel electrophoresis. J. biol. Chem. 244, 4406. Westerwoudt R. J., Naipal A. M. and Harrison C. M. H. (1984) Improved fusion technique II. Stability and purity of hybrid clones. J. Immun. Meth. 68, 89. Winter W. P., Buss E. G., Clagett C. 0. and Boucher R. V. (1967) The nature of the biochemical lesion in avian renal riboflavinuria-I. Effect of genotype on renal riboflavin metabolism. Comp. Biochem. Physiol. 22, 889.