Immunochemical studies on riboflavin carrier proteins.

Abstract

Investigations detailed in this thesis constitute a part of a continuing programme of research undertaken in this laboratory on riboflavin carrier proteins (RCPs), with particular reference to the production and characterization of monoclonal antibodies (MAbs) to chicken RCP, immunochemistry of its antigenic epitopes, their evolutionary conservation, and functional significance during embryonic development as probed with MAbs. RCPs having extensive immunological and physicochemical similarities to chicken RCP have been identified earlier in this laboratory in various mammalian species including primates (and humans). The vital role of this protein in transplacental flavin transport during gestation in rats and monkeys was established by experiments in which immunoneutralization of RCP in female rats and monkeys, using antibodies to chicken RCP, led to fetal wastage and early termination of pregnancy. These results prompted the investigations detailed in the present molecularimmunological studies, namely: (1) the production and detailed characterization of a panel of MAbs to chicken eggwhite RCP; (2) the use of these immunological probes to analyse the evolutionary conservation of antigenic epitopes across species from chicken to humans; (3) delineation of a crucial epitope and its peptide sequence involved in transplacental transport of RCP, and the effect of specific immunoblockade of this epitope on early pregnancy in mice; and (4) restoration of assembled and segmental epitopes during complete unfolding and refolding of chicken RCP under optimal in vitro conditions.

Chapter I Chapter I describes the production and characterization of a battery of MAbs to chicken eggwhite apoRCP. Quantitative immunoprecipitation studies with rabbit polyclonal antisera indicated that 6-7 IgG molecules can bind simultaneously to RCP, implying that at least six epitopes are exposed on the protein’s surface. Earlier, only three MAbs (5B1D3, 6H10F7, 5G1E11) were available, and these were insufficient to map the complete immunotopography. Using hybridoma technology, 18 stable, highaffinity IgGsecreting clones were produced (including the three earlier MAbs). Competitiveinhibition RIA classified these clones into seven subgroups, each recognizing a distinct epitope. One representative clone from each group (6B2C12, 5A2E6, 6H10F7, 5B1D3, 5C4C6, 5G1E11, and another) was selected for further analysis. Among these, 5A2E6 and 5B1D3 recognized partially overlapping epitopes. Five clones belonged to the 6B2C12 subgroup, suggesting that this epitope is immunodominant. Monoclonality was confirmed using: (1) analytical gel filtration of ^125Ilabelled RCP-antibody complexes by FPLC, and (2) heavy and lightchain isotyping. All MAbs had type light chains; heavy chains were IgG2a (3 MAbs), IgG1 (3 MAbs), or IgG2b. Affinity constants were in the range of 10¹¹-10¹² M¹ (liquidphase RIA). To distinguish assembled vs. segmental epitopes, RCP was denatured using SDS, SDS + mercaptoethanol, or completely unfolded by reduction and carboxymethylation. Six MAbs recognized assembled epitopes, susceptible to denaturation; only 6B2C12 recognized a segmental epitope. Neither dephosphorylation nor deglycosylation altered MAb reactivity, indicating epitope determinants reside mainly in the unmodified aminoacid backbone. The flavinbinding site was nonantigenic. Crossreactivity studies showed conservation of most epitopes across rat, cow, and human RCPs, except for the 5G1E11 epitope, absent in bovine RCP.

Chapter II This chapter focuses on identifying the crucial antigenic epitope whose immunoblockade leads to embryonic resorption. Chicken eggwhite and yolk RCP differ in two respects: (1) composition of oligosaccharides, and (2) yolk RCP lacks the Cterminal 11-13 amino acids due to selective proteolysis during endocytosis. All assembledepitopespecific MAbs recognized both forms; however, 6B2C12 did not recognize yolk RCP, suggesting its epitope lies in the Cterminal tail. A synthetic 17aminoacid peptide corresponding to this region displaced ^125IRCP from 6B2C12 in RIA, confirming location of the epitope. However, a high molar excess was needed, suggesting that correct local conformation is required for effective binding. The 6B2C12 epitope was conserved across chicken, rat, and human RCPs. Passive immunization of pregnant mice with 6B2C12 ascites caused 100% early embryonic resorption, while controls (Sp2/0Ag14 ascites or 6H10F7 ascites) produced normal litters. This strongly implicates the 6B2C12 epitope in transplacental riboflavin transport, likely by interfering with receptor recognition or ligand uptake. Partial mapping of assembled epitopes was achieved by isolating a 133aminoacid Cterminal fragment, generated by selective AspPro cleavage. This fragment reacted with three MAbs (6B2C12, 5B1D3, 5A2E6). An 181-204 phosphopeptide bound only 5B1D3.

Protein Refolding Studies Chicken RCP (36 kDa) is highly stabilized by nine intramolecular disulfide bonds, explaining why most epitopes are conformationdependent. Conditions for optimal refolding of completely unfolded RCP were standardized: 2 mg/ml, pH 8.6, 20°C. Refolding was monitored by:

free-SH content, fluorescence spectra, ^14Cflavin binding, gel filtration (Superose12), mobility in native PAGE.

Unfolded RCP reacted only with 6B2C12, indicating retention of the Cterminal segmental epitope, consistent with predictions that this region forms a stable helix. Upon refolding, all assembled epitopes were restored, demonstrating that MAbs recognize determinants dependent on tertiary structure rather than local secondary structure.