HYBRIDOMA Volume 10, Number 4, 1991 Mary Ann Lieben, Inc., Publishers

Monoclonal Antibodies to Sweet Taste Proteins. I. Analysis of Antigenic Epitopes

Thaumatin

on

by Competitive Inhibition Assays

CHITRA MANUAL, FRANCINE SHIRLEY, JERRY M. ANCHIN, CHHABINATH MANUAL, and D. SCOTT LINTHICUM Department of Veterinary Pathobiology, College of Veterinary Medicine, College Station, Texas 77843-4467

Texas A &M

University,

ABSTRACT

Using a competitive inhibition binding immunoassay, we have examined some of the antigenic epitopes present on thaumatin, an intense sweet tasting protein from the African fruit katemfe. We have developed a library of monoclonal antibodies which Some of these monoclonal react with different surface antigenic epitopes on thaumatin. antibodies also cross-react with monellin, another unrelated sweet tasting protein. The competitive binding immunoassay examines the immunoreactivity of both solid-phase and liquid-phase monoclonal antibodies. At least six major antigenic epitopes on thaumatin were identified by our library of monoclonal antibodies. This type of competitive binding analysis may prove useful in the discovery of "sweet taste determinants" on plant proteins and in the development of tandem immunoassays for quantitation of sweet tasting proteins in plant extracts. INTRODUCTION Thaumatin (from the African fruit katemfe, Thaumatococcus danielli), monellin (from the serendipity berries, Dioscoreophyllum cumminsii) and pentadin (from Pentadiplandra brazzeana) are among the sweetest natural molecules known to man.

These plant proteins have a very high specificity for the sweet taste receptor and thaumatin is 100,000 times sweeter than sugar on a molar basis (1). The native conformations of these proteins are required for their sweet taste and cooking denatures the sweet taste. There is little amino acid sequence homology among these molecules with the exception of several tripeptides which are shared between thaumatin and monellin; thaumatin is a single chain of 207 residues, whereas monellin is made up of Some of the polyclonal antibodies that have two chains of 44 and 50 amino acids (2,3). been made to thaumatin and monellin cross-react with each other. These antisera have been reported to bind other sweet tasting compounds (e.g., aspartame or saccharin) in a rank order related to their sweet taste intensity (4). The three dimensional structures for thaumatin and monellin have been determined by x-ray crystallographic techniques, but the structural features responsible for their intense sweet taste are not Furthermore, the presence of structural features that might be shared apparent (5). between numerous small synthetic sweet taste compounds and the large plant proteins remains obscure. One objective of our research, and the work presented herein, is to use monoclonal antibodies which react to thaumatin and monellin to localize the "sweet taste epitopes" on these proteins. We have prepared a library of monoclonal antibodies to thaumatin and monellin. These antibodies have been used to characterize the antigenic 459

epitopes which are either unique or common to both proteins. In this study we report a competitive inhibition immunoassay that identified the immunoreactive sites recognized by the library of monoclonal antibodies. MATERIALS AND METHODS Production of Monoclonal Antibodies (MAbs) Thaumatin and monellin used in this study were gifts from Professor Goran Hellekant (University of Wisconsin, Madison). Female BALB/c mice were immunized with 100 ng of protein emulsified in Freunds complete adjuvant. Booster immunizations were given thirty days later using incomplete adjuvant. Prior to the fusion procedure, the mice were given an intrasplenic booster injection of antigen (50 ng) in saline (5). Four days later, the spleen cells were harvested and prepared for the hybridoma fusion with myeloma line SP2/0. From five separate hybridoma fusion preparations, fifteen MAb secreting hybridoma cell lines were isolated, cloned, and subjected to further study. The enzyme-immunoassays for detecting antibody activity, Ig isotyping, isoelectric focusing, and immunoblotting procedures were also used for the initial characterization of the different MAbs (6). All of the MAbs in this study were isotyped as IgGl, G2a or G3 and all had kappa light chains as shown in Table 1. MAb clones #A-M were obtained from mice immunized with thaumatin; clones A thru E and G thru J are specific for thaumatin, whereas clones F, K thru M, cross-react with monellin (Table 1). MAb clones N and O were obtained from mice immunized with monellin and these MAb react with both monellin and thaumatin. TABLE 1

MAb Clone A B C D E F G H I J K L M N

Ig Isotvpe

G3,k Gl.k G2a,k G2a,k Gl,k Gl,k Gl,k G2a,k G2a,k Gl,k G2a,k G2a,k

Specificity* T T T T T

T.M T T T T

G2a,k

T,M T,M T,M M,T

Gl.k

M.T

nd

Indicates if the MAb reacts with thaumatin (T) or monellin (M) or both; first antigen indicated was initial antigen used in the immunization. Immunoreactivity was determined using a solid phase EIA (5).

Radio-iodination of Thaumatin

Thaumatin

was

radio-iodinated

using

IL)) (oxidative capacity, 0.55 mM/bead) and

Sephadex

G25 column.

Competition

Assay for

The

specific activity

Iodo-beads (Pierce Chemical Co. Rockford, separated from free radio-iodine using a of the radiolabeled protein was 15 u.Ci/u.g. was

Epitope Specificity

Hybridoma MAb tissue culture supernatants or hybridoma induced MAb ascites immobilized by absorption to the surface of flexible polyvinyl 96 well microtiter plates. Each well was coated for 2 hrs at room temperature with 50 \i\ of the MAb solution diluted in borate buffered saline (BBS), pH 8.6; an appropriate dilution for each MAb solution was previously determined by a direct binding assay with radiolabeled were

460

thaumatin. Following the absorption of the MAb, the wells were "post-coated" with 100 After u.1 of bovine serum albumin (1% (w/v)) to block non-specific protein absorption. washing with BBS containing Tween20 (0.05%), each well received a mixture of 25 ul of 125I-labeled thaumatin and 25 u.1 of MAb hybridoma tissue culture supernatants from related or unrelated MAb clones. This mixture was allowed to incubate overnight at 4°C After the incubation period, the plates were thoroughly or 6 hrs at room temperature. The amount of washed with BBS-Tween20 and the individual wells were cut off. The radiolabel antigen bound to the solid phase was determined by gammaspectroscopy. control wells which contained no solution phase MAb in the antibody-antigen incubation mixture showed the highest binding and these values were used as the maximum values of radiolabel bound (Cq); the wells containing MAb from the same clone as in the solid phase showed the least binding (Cp) (maximal inhibition or competition). The individual inhibition values (Ca) were obtained from wells which contained a The extent of inhibition (I) solution phase MAb different than that in the solid phase. for each MAb was calculated using following equation :

(c0 CA)

I

-

x

=

100%

(Co CF)

-

The extent of inhibition for all MAb competitors are finally expressed as the percent of inhibition by the culture supernatant containing same antibody in the solution phase as in the solid phase. For a particular MAb coated to the solid-phase surface, all the antibodies of the panel (15 in all) were used as inhibitors. This experiment was repeated for the entire panel of solid phase MAbs (n=15). A computer algorithm (in GWBASIC) has been developed (7) to identify groups of The epitope specificity can MAbs based on these competitive binding inhibition results. be quantitated as the distance between the antibody in the solid phase and a competitor in the solution phase by subtracting the extent of inhibition (I) of the antibody from 100 (the value assigned to the solid phase antibody in the solution phase). This epitope specificity distance serves as an index of "closeness" in the ability to compete for the same epitope. The distances for the solid phase MAb from all other MAb in solution were graphically displayed by plotting the clones (labeled as "a" to "o" in lower case) around These the centrally placed solid phase MAb clone (shown as an upper case letter). individual plots represent the observed relative inhibition, but they are insufficient in correlating mutual relations between MAbs when considered only in pairs. In order to derive a unique mutual relationship between any two of the 15 MAb clones (total number of pairs= 210), the epitope specificity distances were further analyzed using a modified algorithm (6). This algorithm analyzes the mutual relationship between any two MAb clones and defines six categories with increasing epitope specificity distances. A limiting value of the specificity scale was used in deriving these relations. Relationship tables were constructed using a number of limiting values. This approach has been successfully used to analyze antigenic epitopes on alpha-fetoprotein (7). -

RESULTS The extent of thaumatin binding inhibition for each MAb in the solid phase upper corner) and all other MAb clones in the solution phase (lower case letters on x-axis) for all the clones studied is shown in Fig. 1. These inhibition data were used for the grouping of the various MAbs as described below. The relative inhibition activity of clones C, E, K and L are summarized in Fig. 2. In Fig. 2 the relative positions of epitopes identified by the MAb clones in terms of their epitope specificity distances from that recognized by the solid phase MAb clone are shown. The concentric rings represent 15% inhibition levels between the solid phase Panel 1 MAb (Capital letter, center) and the solution phase MAbs (lower case letters). and 2 show the relative positions of all the MAbs in the solution phase from the solid phase clone C and E, respectively. It may be noted that e (in solution phase) is the closest (within 15%) clone in the graph of the clone C (in solid phase) in panel 1 and c (in solution phase) is the closest clone (within 15%) to the clone E ( in solid phase) as shown in panel 2. These results indicate that the binding of the antibody of clone c to thaumatin is fully competitive to that of clone e, and vice versa, with the 15% limit. In panel 3 the solution phase clone k is the closest (within 15%) to the solid phase clone L, but clone 1 (in solution phase) is >45% distant from the solid phase clone K as shown in panel 4.

(Capital letter,

461

abcdefghijklmi

abcdefghijklm

SOLUTION CLONES

SOLUTION CLONES

abcdefghijklmno

abcdefghijklm SOLUTION CLONES

SOLUTION CLONES

abcdefghijklmno

abcdefghijklmno

SOLUTION CLONES

SOLUTION CLONES

abcdefghijklmno

abcdefghijklmno

SOLUTION CLONES

SOLUTION CLONES

When all the clones are compared in pairs, their relative positions to each other, such as that displayed by clones C and E, are usually reciprocal. However, in some cases other MAb pairs, such as clones K and L, do not show this reciprocal nature. The relative positions of all the clones are presented in Table 2 where the solution phase clones are arranged in order of increasing distance with respect to a solid phase clone. In order to get a unique mutual relationship between any two clones, a primary grouping was done The first with a given value of limiting distance dividing the clones in three groups. group contains clones having distances from 0 to 15%, the second group containing clones above 15% and below 85% and the third group containing clones from 85 to 100%. Using the data derived from Table 2, the analysis algorithm generates the relationships between any pair of clones in six categories (Table 3). The essential feature of this final grouping is that any two clones bear the same mutual relationship to each other. The pair of clones (e.g., A and C) are "closest" to each other in terms of This means that they compete with their epitope specificity distance scale (within 15%). On the other each other with equal efficiency whether in the solid or solution phase. hand, a group of clones (e.g., N and O) are "most distant" in their relation to the epitope recognized by clone H (more than 85% difference in the epitope specificity distance A pair of clones having a "closer" relationship compete with fairly high scale). efficiency. A pair of clones showing a "more distant" relationship compete poorly with of each other. If the distance is within the range of 15-85% with either of the clones in the solid phase, the relationship is called "intermediate." A pair is said to have an "anomalous" relationship if the inhibitory strengths drastically differ by interchanging the partners from solution to solid phase and vice versa. None of the pairs presented in Table 3 showed an anomalous relationship.

462

abcdefghijk

abcdefghijklmno SOLUTION CLONES

SOLUTION CLONES

abcdefghijklmno

abcdefghijkirr

SOLUTION CLONES

SOLUTION CLONES

abcdefghijklmno

abcdefghijklmn

SOLUTION CLONES

SOLUTION CLONES

Fig. 1 The binding by

inhibition of thaumatin different solution phase clones is shown for each solid phase MAb examined.

abcdefghijklmno SOLUTION CLONES

TABLE 2 Inhibition Levels Between Solid Phase MAb Clones (SPC) and Solution Phase MAb Clones

£EC_

Within 15%

Between

A B

cge

C

ea

D E F G H I J K L M N O

gji

bjmoifkhdnl aimlgohjdknf hbdijgm

15-85^

dfeab

gdj dg

bfmchlok clnkom

k

cfghba

ce

kfoln

aenmohblckf

hkjmgbildonf gkibcadomehln idjkonmlfhacbe gckjlim

ca

Above 85%

jiafgnlohebd jmibdaohgf jlindko olkagfmieb

lngkambefi_

463

J no ane

efhab c m

nee e

jhed hied

2 Diagrammatic representation of the competitive inhibition levels for several different MAbs antibodies (center ring, capital letter); rings represent levels of 15% difference.

Fig.

TABLE 3

SEC A B C D E F G H

I J K L M N O

Relationships between pairs of clones which exhibit mutual competition _Closest_Closer_Intermediate_Distant_Most Distant bofkdnl gmh J1 m h ailgodknf jce m ea bf koln hdijg n o aemblckf gh ji h jmbil kgdonf m h c gkibadoeln konlfhcbe idjma dfeabm gckli j ane bfmchlok dj id

clkm

1 k

cfghba

noehab

jiafgnohebd jmibdaohgf jlindo okagfmeb lngkambefi

464

f

c

m

c

ne

ke

jd

lijd

hc hc

Table 4 presents the number of epitopes (total of six) clearly identified by different groups of MAbs. These groups were selected from the MAbs which have either "closet" or "closer" relationships as shown in Table 3. These groups of MAbs represent the minimal number of distinct epitopes or closely related epitopes as indicated by the competitive binding inhibition assay. TABLE 4

Epitopes and Specific MAb Clones Epitope Number_Specific Clones_ Number of

I II III IV V

A,CE D,I,J K,L NO M, H

V]_B_ DISCUSSION

study we have examined a panel of fifteen different MAbs and used them to their epitope specificities for thaumatin using a competitive inhibition binding assay combined with computer-aided grouping. The results, expressed as the epitope specificity distances, have identified antibodies with distinct antigen epitope The members of the "closest" MAb pairs react with the same or specificities. topographically similar antigen epitopes. Such MAb binding competitions may arise due to recognition of the same antigen epitope or due to steric (in close proximity) or allosteric (action from a distance) effects of one MAb on the binding of other MAb. On the other hand, it may be confidently stated that the members of the "most distant" pairs of MAbs recognize two different antigen epitopes which are far enough from each other to rule out any steric interactions since the binding of one MAb has no affect on the binding of the other MAb to any significant degree. For example, MAb clone C binds to an epitope on thaumatin which is clearly different from the epitopes recognized by clones K, L, N and O (Table 3). The antibodies made against monellin which cross-react with thaumatin (clones N and O) appear to react with a distinctive thaumatin epitope (IV). In contrast, the antibodies made to thaumatin which were found to cross-react with monellin appear to react with a different epitope (Group III). The reactivity of clone F was not distinct enough to assign its reactivity to a specific group. We have reported here a panel of monoclonal antibodies made against the sweet taste proteins thaumatin and monellin. The relative epitope specificities found on thaumatin recognized by these antibodies have been analyzed so that antibodies which This type of have distinct non-interactive epitope specificities can be identified. analysis can be very useful in analyzing the differential expression of cross-reactive epitopes in these two unrelated sweet taste proteins. Further analysis using peptide digestion fragments or synthetic peptides may help identify the precise amino acid residues which comprise the antigenic epitopes identified in this study. The identification of antibodies which recognize non-interactive epitopes for thaumatin may prove useful in the design of tandem immunoassays for the quantitation of protein in plant extracts, gene expression systems and site-directed mutagenesis experiments. In this

establish

ACKNOWLEDGMENTS This research was supported by grants from the American Heart Association (891066) Drs. Chitra Mandai and Chhabinath Mandai are visiting scientists and the NutraSweet Co. from the Indian Institute of Chemical Biology, Calcutta. REFERENCES Van de Wei, H. and Loeve, K. Isolation and characteristics of thaumatin I and II, the tasting protein from Thaumatococcus danielli. Benth. Eur. J. Biochem. 31:221-225, 1972. 1.

sweet

465

2.

Frank, G. and Zuber, H.

sweet

protein

monellin.

The complete amino acid sequences of both subunits Hoppe-Seyler's Z. Physiol. Chcm. 357:585-592, 1976.

of the

3. Iyengar, R. B., Smits, P., Van der Ouderaa, F., and Van der Wei, H., Van Brouwershaven, J., Ravenstein, P., Richters, G. and Van Wassennaar, P. D. The complete amino acid sequence of the sweet protein thaumatin I. Eur. J. Biochem. 96:193-204, 1979. 4.

Hough,

taste

C. A. M. and Edwardson, J. A. receptor. Nature 271:381-383, 1978.

5. Kim, S.H., de Vos, A. and Ogata, C. Trends in Biochem. Sei. 13:13-15, 1988.

Antibodies to thaumatin

Crystal

structures of the

as a

model of the

intensely

sweet

sweet

proteins.

6. Kussie, P.H., Albright, G. and Linthicum, D. S. Production and characterization of monoclonal idiotypes and anti-idiotypes for small ligands. Meth. Enzymol. 178:49-63, 1989.

Chakraborty, M., Mandai, C. and Mandai, C. Epitope analysis of the oncofetal antigen alpha-fetoprotein using monoclonal antibodies. Mol. Immunol, (in press) 1991. 7.

Direct reprint requests to: Dr. D. S. Linthicum Dept. Vet. Pathobiol. College of Vet. Med. Texas A&M University College Station, TX 77843-4467

466