Documente Academic
Documente Profesional
Documente Cultură
11768-11775,1988
0 1988 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S. A.
Antigen-Antibody Interaction
SYNTHETICPEPTIDES DEFINE LINEAR ANTIGENIC DETERMINANTS RECOGNIZED BY MONOCLONAL
ANTIBODIES DIRECTED TO THE CYTOPLASMIC CARBOXYL TERMINUS OF RHODOPSIN*
The specificities of four monoclonal antibodies rho nikov et al., 1982; Hargrave et al., 1983), a model of rhodopsin
1D4, lCS,3A6, and 3D6 prepared by immunization of structure has been proposed in which the polypeptide chain
rod outer segments containing rhodopsin have been spans the membrane lipid bilayer seven times. Labeling and
defined using synthetic peptides. All of these antibodies proteolysis studies (Molday and MacKenzie, 1983; Clark and
interact within the 18 residues at theCOOH terminus Molday, 1979) have indicated that the COOH terminus is
of rhodopsin and recognize linear antigenic determi- exposed on the cytoplasmic side of the disc membrane whereas
nants of 4-11 residues. Twenty-seven synthetic pep-
tide analogs of varying lengths of native sequence or the NH2 terminus is exposed on the lumen or interdisc surface.
containing single amino acid substitutions at each po- Four monoclonal antibodies to bovine rhodopsin have been
sition of the COOH-terminal18 residues have provided shown to bind to the COOH-terminal region (MacKenzie et
some insight into the mechanism of antigen-antibody al., 1984). These monoclonal antibodies would be ideal can-
binding. Our results clearly demonstratethat antibod- didates to study the interaction of antibody and antigensince
ies can be highlyspecific at key positions as shown by the known antigenic region is small (18 residues or less) and
the loss of binding on single amino acid substitutions synthetic peptides are able to compete with native protein for
in the bindingsite. In contrast singleamino acid sub- binding to anti-protein antibodies raised to rhodopsin-con-
stitutions at other positions in the binding site only taining disc segments. Although it is generally thought that
affect affinity for some antibodies. Ionic interactions monoclonal antibodies recognize discontinuous epitopes (van
candominate immunogenic determinants. Immuno- Regenmortel, 1986; Benjamin et al., 1984), we report that
genic determinants are not restricted to highly charged monoclonal antibodies to the COOH terminus of intact rho-
hydrophilic regionson the surface of a protein andmay dopsin-containing discs recognize residues confined to small
be dominated by hydrophobic interactions. Although linear epitopes ranging from 4 to 11 residues. In a previous
certain side chains can dominate the interactionof the study of antigen-antibody interaction, we reported that a
antigen with antibody, our results are in agreement
with the interpretation that the free energies of all the NH2-terminal acetylated residue was critical for the binding
contact points are additive and a certain free energy of antibodies to EDP208 pilin protein, and this binding was
must be present to achieve binding. Antibodies with restricted to a NH2-terminal pentapeptide (Worobec et al.,
different specificities directed to the same region of 1985). Similarly the NH,-terminal acetylated amino acid was
the protein antigen can be produced in an immune found to be essential for the binding of synthetic peptides to
response. Peptide antigens representing regions of a anti-cytochrome c antibodies (Paterson, 1985). In this report
protein antigen bind best to the anti-protein antibody we describe results of antigen-antibody interactions at the
when the sequence is shortened to contain only those COOH terminus of rhodopsin, the importance of the free
residues binding to thespecificity site in theantibody. COOH-terminal a-carboxyl group, the additive binding effect
Cross-reactivity between protein antigens can be ex- of hydrophobic and hydrophilic residues, and thatmonoclonal
plained by conservation of the critical residues in the antibodies with a variety of specificities, both ionic and hy-
combining site. drophobic, can recognize small linear determinants.
MATERIALS AND METHODS
Rod Outer Segment Membranes-Bovine rod outer segments were
Rhodopsin is the major membrane glycoprotein in rod outer purified by sucrose-density centrifugation under dim red light as
segment disc membranes of vertebrate retinalrod photorecep- previously described (Wong and Molday, 1986). Rod outer segment
tor cells. On the basis of protein sequence analysis (Ovchin- disc membranes used as a source of rhodopsin were obtained by
hypotonic lysis of rod outer segments followed by flotation on 5%
* This work was supported by the Medical Research Council of Ficoll according to themethod of Smith et al. (1975). Protein concen-
Canada (R. S. H. and R. S. M.), equipment grants from the Alberta tration was determined by the Lowry method using bovine serum
Heritage Foundation for Medical Research (to R. S. H.), and National albumin as a standard (Lowry et al., 1951).
Institutes of Health Grant EY-02422 (to R. S. M.). The costs of Antibodies-Antibovine rhodopsin monoclonal antibodies (rho
Publication of this article were defrayed in part by the payment of 1D4, 3.46, 1C5 and rho 3D6) were derived from culture fluids of
page charges. This article must therefore be hereby marked adver- hybridoma cell lines (MacKenzie et al., 1984; Hicks and Molday,
tisement in accordance with 18U.S.C. Section 1734 solelyto indicate 1986).
this fact. Goat antimouse immunoglobulin antibody used in indirect solid-
$ To whom correspondence should be addressed. phase radioimmune assays was iodinated with NalZ5Iby the chlor-
11768
Antigen-Antibody Interaction 11769
amine-T method (Hunter and Greenwood, 1962) and had a specific
activity of 1-2 X 10 dpm/pg.
Solid-phase Radioimmune Assays-For radioimmune assays, 96-
well flexvinyl microtiter plates were coated with rhodopsin by drying
down 25 plof Triton X-100-solubilizedrod outer segment membranes.
The plates were rinsed in phosphate-buffered saline (0.01 M sodium
phosphate, 0.15 M NaC1) containing 1%bovine serum albumin prior
to use. Competition assays were carried out by incubating 25 p1 of
serial dilutions of the peptide antigen with 25 plof hybridoma culture
fluid a t a dilution which gave 80-90% antibody saturation as meas-
ured in standard solid-phase radioimmune assays (MacKenzie and
Molday, 1982). The concentration of all peptides were determined by
amino acid analysis of stock solutions, which has an error of +-5%,or
by weight which when compared to amino acid analysis had an
accuracy of &lo%. After 30-60 min at 25 C, 25 p1 from each well
was transferred to therhodopsin-coated microtiter plates. The micro-
titer plates were incubated for 30 min, washed in phosphate-buffered
saline, and treated with 25 11 of 261-labeledgoat antimouse immu-
noglobulin antibody (10 pg/ml; 1-2 X 10 dpm/pg). After 30-60 min,
the plates were again washed in phosphate-buffered saline, cut, and
counted in a Beckman 8000 gamma counter. Variations in the 160
values (peptide concentration required to obtain half-maximum in-
hibition of antibody binding to rhodopsin) were observed for different
stock hybridoma cell culture fluids and rod outer segment prepara-
tions. In repetitive assays, however, 160 values for a given peptide
agreed within -C50%. Variations in hybridoma supernatant are due to
the quantity of monoclonal antibody secreted per ml of fluid and the
dilution used in the competition assay. Differences in 160values were
used to get around this difference. When different hybridoma cell
culture fluids and rod outer segment preparations were used the I I I I
native peptide sequence was always used as a control to ensure the 10 20
relative accuracy of the peptide analog results. A variation in 160
values between different peptide analogs of 2-fold or greater is con-
sidered significant.
Peptide Synthesis-Unless otherwise stated,all chemicals used ELUTION TIME ( m i d
were reagent-grade. Diisopropylethylamine, dichloromethane FIG. 1. Representative reversed-phase chromatographic
(DCM), trifluoroacetic acid, and dimethylformamide were obtained elution profiles on analytical columns of the crude synthetic
from General Intermediates of Canada. Diisopropylethylamine, peptides. Panel A, peptide 5, Fig. 1 [Ac(Ala2)R(1-12)-OH]; Col-
DCM, and trifluoroacetic acid were redistilled before use. HPLC- umn, SynChropak RP-8, 4.1 mm inner diameter X 150 mm, 6.5 pm,
grade acetonitrile was obtained from either Fischer Scientific, or 300 A, (2-8. Panel B , peptide 1, Fig. 1 [AcR(1-18)-OH]; Column,
Caledon Laboratories. Double-distilled water was purified by passage Pharmacia PEP RPC, 5 mm inner diameter x 50 mm, 5 pm, 100 A,
througha Milli-Q water purification system. t-Butyloxycarbonyl- C-18. The solvent system used for both columns was a linear AB
amino acids were obtained from Institut Armand Frappier (Laval, gradient where A = 0.05% trifluoroacetic acid/HzO and B = 0.05%
Quebec), Protein Research Foundation (Osaka, Japan), andBachem trifluoroacetic acid/acetonitrile. Gradient rate was 1%B/min with
Fine Chemicals. Co-poly (styrene, 1%divinylbenzene) chloromethyl flow rates of 1 ml/min. The crude peptides were purified as outlined
resin (0.9 mmol Cl/g resin) was obtained from Pierce Chemical Co. under Materials and Methods.
Co-poly (styrene, 1%divinylbenzene) benzhydrylamine-HC1 resin
(0.9 mmolNHz/g resin)was obtained from Institut Armand Frappier.
Twenty-seven analogs were synthesized using the general proce- described previously and acetylated in asolution of toluene, pyridine,
dure for solid-phase peptide synthesis described by Merrifield (Stew- and acetic anhydride (3:3:1) for 1 h. The programs used for attach-
art and Young,1984; Erickson and Merrifield, 1976) on either a ment of each aminoacid using the Beckman 990 synthesizer were the
Beckman 990 Peptide Synthesizer, or an Applied Biosystems 430A same as previously described (Worobec et al., 1985; Parker and
Peptide Synthesizer. All peptides with a free carboxyl terminus were Hodges, 1985). Coupling procedures for the Applied Biosystems syn-
initiated by esterification of the cesium salt of the COOH-terminal thesizer have been described (Kent andLewis, 1985).
amino acid to co-poly (styrene, 1%divinylbenzene) chloromethyl Peptides werecleaved from the resin with HF (20 ml/g resin)
resin (Stewart and Young, 1984). Substitutions of approximately 0.9 containing anisole (2 ml/g resin), and ethanedithiol(20drops/g resin)
mmol of amino acid/g resin were obtained. For peptides with an a t 4 Cfor 45 min. The solvents were removed under reduced pressure
amide COOH terminus, co-poly (styrene, 1%divinylbenzene) benz- a t 4 Cfor at least 3h. The resin was washed with ether, and extracted
hydrylamine-HC1was neutralized with 5% diisopropylethylamine for with 3 X 10 ml trifluoroacetic acid. The trifluoroacetic acid was
1 h and theCOOH-terminal aminoacid was coupled to theresin with evaporated and thepeptide redissolved in water and lyophilized.
dicyclohexylcarbodiimide for 1 h (1:l equivalent, protected amino Purification by HPLC-The crude peptides were purified using
acid/dicyclohexylcarbodiimide). reversed-phase chromatography on either a SynChropak RP-P C-18,
Side-chain protecting groups for t-butyloxycarbonyl-aminoacids 300 A, 6.5pm column (250 X IO mm inner diameter) or a Rainin
used were: Lys(2-chlorobenzyloxycarbonyl), Asp(0-benzyl), Glu(0- Dynamax Macro C-18, 300 A, 1 2 pm column (250 X 21.4 mm inner
benzyl), Ser(benzyl), and Thr(benzy1). Removal of the t-butyloxycar- diameter) with a linear AB gradient where A = 0.05% trifluoroacetic
bony1 protecting group at each cycle was done in 50% trifluoroacetic acid/HzO and B = 0.05% trifluoroacetic acid/acetonitrile. The gra-
acid/DCM for 20 min. This was followed by neutralization in 5% dientrate was 1%B/min with flow rates of 2 ml/min for the
diisopropylethylamine. All amino acids were coupled using DCC- SynChropak column and 10 ml/min for the Rainin column. Repre-
activated couplings (1:l equivalent, protected amino acid/DCC) in sentative HPLC elution profiles of the 12-residue and 18-residue
DCM, except Thr which was coupled as a symmetric anhydride (2:l crude synthetic peptides before purification are shown in Fig. 1.
equivalent, protected amino acid/DCC) in DCM, and Gln and Asn,
which were coupled in dimethylformamide as N-hydroxybenzotriazole RESULTS
active esters. Double couplings of 1 h each were performed at each Four antibovine rhodopsin monoclonal antibodies, rho1D4,
step of the synthesis. The completed peptides were deprotected as 3A6, 1C5, and 3D6, were previously shownto bind to bovine
The abbreviations used are: DCM, dichloromethane; DCC, dicy- rhodopsin in the COOH-terminal 18 residues, 1-18 where
clohexylcarbodiimide; HPLC, high performance liquid chromatogra- 1 is the COOH-terminalresidue(MacKenzie et al., 1984;
phy. Hicks and Molday, 1986). To localize further the antibody-
11770 Antigen-Antibody Interaction
binding sites and examine the importance of each residue in rhodopsin (Hicksand Molday,1986). Therefore, synthetic
this 18-residue peptide to antibody binding, the 27 peptides analogs of lengths l ' - l Z ' and 1'-18' seemed appropriate to
shown in Fig. 2 were synthesized, purified, and examined in test antigen-antibody binding of these four monoclonal anti-
a solid-phase competitive inhibition assay. To demonstrate bodies. Based upon the results with these analogs, the mini-
the effect of each amino acid side chain on antibody binding, mum lengths for maximum antibody binding would then be
the following single amino acid substitutions of each side synthesized.
chain were carried out; when the amino acid side chain in the All peptides representing internal regions of the rhodopsin
native sequence was alanine, aglycine residue was substituted; sequence were synthesized as the Ne-acetylated and COOH-
all larger side chains were substituted by alanine except for terminal amides to prevent the introduction of charged groups
the acidic residues aspartic and glutamic acid (positions 8 ', near the antibody-binding site. This is of particular impor-
17', and 18', Fig. 2) which were substituted by the uncharged tance, as the minimum peptide chain length is synthesized
isosteric residues, asparagine and glutamine, respectively which maximizes antibody binding. Talbot andHodges (1981)
(peptides 11, 20, and 21, Fig. 2). These Asn and Gln substi- clearly demonstrated the importance of not introducing N"
tutions would examine the importance of ionic interactions or C" charged groups into a binding site peptide which origi-
in antigen-antibody binding while leaving all other interac- nates from an internal sequence of protein molecule. These
tions with that residue intact (van der Waals, hydrophobic or workers determined the minimal inhibitory peptide of the
hydrogen bonding). The two peptide chain-lengths of 1'-12' troponin I inhibitory sequence is 12 residues and that the N"-
or 1'-18' were selected based upon previous results where amino groups or C"-carboxyl group significantly affected the
competitive inhibition studies indicated that antibody rho ability of this peptide to inhibit the actomyosin ATPase.
3A6 required a length 1'-12' and larger whereas antibody rho This precaution of N"-acetylation and C"-amides should
1C5 required peptide 1'-18' (MacKenzie et al., 1984). Anti- always becarried out since these groups are isosteric with the
body rho 1D4 binding was not inhibitedby peptides 2'-13' or peptide bonds at either end of the peptide. Although, in this
3'-18', indicating that theCOOH-terminal alanine residue of study the 150 values for Ne-acetylated 1'-12' or 1'-18' were
rhodopsin was required. Competition studies using COOH- indistinguishable from the nonacetylated 1'-12' or 1'-18'
terminal peptides l"4' to 1'-18' were equally effective inhib- peptides when tested with rho 1D4 and rho 3A6, this precau-
itors of antibody rho 3D6 binding to immobilizedbovine tion was observed for all shorterpeptides.
I Rc-Rrp-Glu-Rla-Ser-Tr-Thr-Val-Ser-Lyr-Thr-Glu-Thr-Ser-Gln-UaI-Rla-Pro-Rla-OH
5
W
6 RC"
7 Rc-Ual
20
21 Rla-OH
22 ~c-Rsp-6lu-Rla-Ser-~r-Thr-Ual-Ser-Lyr-Tr-6lu-amide
23 Rc-Thr-Thr-Val-Ser-Lys-Thr-Glu-amide
24 Ac-Val-Ser-Lyr-Rr-Glu-amide
25 Ac-Thr-Glu-Thr-Ser-GIn-Ual-Ala-Pro-RIa-OH
26 Rc-Thr-Ser-Gln-Val-Ala-Pro-Ala-OH
27 Ac-Ual-Ala-Pro-Rla-OH
Antigen-Antibody Interaction 11771
Importance of the COOH-terminal a-Carboxyl Group to An- with glutamine (Iw values of2000, 63, 125, 10, and 6.3 pM,
tibody Binding-Synthetic peptides 2 and 3, AcR(1'-12')-OH respectively, when compared to peptide 2 of0.9 pM repre-
and AcR(1'-12')-amide, respectively, were compared in the senting a7-2220-fold decrease). The most important residues
solid-phase competitive inhibition assay. The results shown interacting with antibody rho 1D4 werealanine 3', glutamine
in Table I indicate that theCOOH-terminal a-carboxyl group 5', and threonine7' where no detectable binding was observed
is essential for antibody binding to both rho 1D4 and rho when these positions were substituted by glycine 3', alanine
3D6. The Iso value increased from 0.9 to 158 p~ when com- 5', and 7', respectively. The results of the competitive inhi-
paring peptide 2 and 3, respectively, for antibody rho 1D4 bition assays for monoclonal rho 1D4 antibody binding are
binding and peptide 3 showed no binding to antibody rho 3D6 summarized graphically in Fig. 3 where the solid bars indicate
(Table I).In contrast antibody rho 3A6 bound equally well to the importance of each residue to antibody binding.
peptides 2 and 3 indicating that the a-carboxyl group of the The results shown in Fig. 3 suggested that the minimum
COOH-terminal alanine residue is not involved in binding to peptide sequence required for maximum antibody rho 1D4
this antibody. binding was residues 1'-9'. This conclusion was verified by
Importance of Each Amino Acid Side C h i n to Antibody the data shown in Table I1 where binding increased approxi-
Binding-Systematic replacement of each residue in the mately 6-fold whencomparing peptide 25 (1'-9') with peptide
COOH-terminal 18-residue sequence was made in either the 2 (1'-12'). It appears that the 1'-9' peptide reproducibly
12- or 18-residue peptide. The substitutions for each of the serves as a more effective antigen than the 1'-12' peptide.
first 12 positions of the COOH-terminal sequence were made Interestingly, peptide 26 (1'-7') which contained the residues
in analogs of the 1'-12' sequence and are shown in Fig. 2, demonstrated to be critical for binding (Fig, 3) showed no
peptides 4-15. The substitutions for each of the positions 13- detectable binding. It appears as if the loss of binding energy
18were made in analogs of the 1'-18' sequence and areshown from residues 8' and 9', although small, was substantial
in Fig. 2, peptides 16-21. All four antibodies used in this study enough to prevent binding from a peptide containing residues
bound to the 18-residue peptide containing the COOH-ter- 1'-7'. MacKenzie et al. (1984) showed that peptide 1'4'
minal residue of rhodopsin. In addition, three antibodies, rho bound to antibody rho 1D4. Another possibility is the loss of
1D4,3A6, and 3D6 bound to the12-residue peptide containing hydrogen bonds to thepeptide bonds on removing residues 8'
the COOH-terminal residue of rhodopsin. and 9'.
Monoclonal Antibody rho 104 Binding Studies-As shown Monoclonal Antibody rho 3A6 Binding Studies-As shown
in Fig. 3 and Table I, the region 1'-9' is important for peptide in Fig. 3 and TableI, the side chains in the region 8'-12' are
binding to monoclonal antibody rho 1D4. These results sug- important for peptide binding to monoclonal antibody rho
gest that side chains of residues l', 10'-18' make only weak 3A6. These resultssuggest that theside chains of residues 1'-
or insignificant contributions to antibody rho 1D4 binding. 7' and 13'"' make only weak or insignificant contributions
Substantial decreases in antibody binding, as reflected by the to antibody rho 3A6 binding. Substantial decreases in anti-
150 values in Table I, were seen with peptides 5, 7, 9, 11, and body binding, as reflected by the IsOvalues in Table I were
12, where proline 2', valine 4', serine 6', and threonine 9' seen for peptides 12 and 15 where threonine 9' and valine 12'
were replaced with alanine and glutamic acid 8' was replaced were replaced with alanine (I50values of 1000 and 350 p ~ ,
TABLE I
Effect of single amino acid substitutions on peptidebinding to anti-rhodopsinmonoclonal antibodies
Monoclonal antibody
Peptide Peptide Group altered 1D4
1C5 3A6 3D6
no. region from -+ to ~ -
IW" LogIMl-LogI"50b 154 LogI~O-LogI"50
2 l"12' None 0.9 0 28 0
3 l"12' COOH + amide 158 2.3 28 0
4 l"12' 1'Ala + Gly 2.5 0.5 25 -0.05
5 l"12' 2'Pro + Ala 2000 3.4 35 0.09
6 l"12' 3'Ala + Gly - - 35 0.09
7 l"12' 4'Val+ Ala 63 1.8 56 0.30
8 l"12' 5'Gln + Ala - - 24 -0.07
9 1'-12' 6'Ser + Ala 125 2.2 35 0.09
10 1'-12' 7'Thr + Ala - - 31 0.04
11 l"12' 8'Glu + Gln 6.3 0.9 - -
12 l"12' 9'Thr + Ala 10.0 1.1 1000 1.55
13 l"12' 1O'Lys + Ala 1.0 0.1 - -
14 l"12' 11'Ser + Ala 0.5 -0.3 - -
15 l"12' 12'Val- Ala 0.3 -0.5 350 1.09
0
Ov)
U
FIG. 3. Panel A , effects of single
amino acid substitutions inthe synthetic 0"
-1
2.4
104
"""" - ""
1C5 I I 306
I
TABLE I1 lysine lo', and serine 11' where no detectable binding was
Effects of chain length on peptide binding to anti-rhodopsin observed, when these positions were substituted by glutamine
monoclonal antibodies 8', alanine lo', and ll',respectively. The importance of serine
Monoclonal Peptide Peptide region 11' andthreonine 9' also has been suggested in previous
no. or ROS" I,*
antibody studies indicating that kinase-catalyzed phosphorylation of
PM rhodopsin inhibits the binding of antibody rho 3A6 to rho-
l"12' 1D4 2 6.3 dopsin (Molday and MacKenzie, 1985). The results of the
25 1'-9' 1.3
26 l"7' -
competitive inhibition assays for monoclonal rho 3A6 anti-
24 8"12' - body binding are summarized graphically in Fig. 3 where the
ROS 0.02 hatched bars indicate the importance of each residue to anti-
3A6 2 l"12' 59' body binding.
8"18' 22 1.38 The results shown in Fig. 3 suggestedthat minimum peptide
23 8"14' 3.58
8"12' 24 19.60
sequence required for maximum antibody rho 3A6 binding is
ROS 0.08 residues 8'-12'. This conclusion wasverified by the data
1C5 1 l"18' 0.40 shown in Table I1 and Fig. 4. Fig. 4 is representative of the
2 l"12' - data obtained by solid-phase radioimmune competitive inhi-
22 8'-18' 0.21 bition assays displayed in Tables I and 11. Binding increased
23 8"14' 263
24 8"12' >loo0
approximately 3-fold when peptide 24 (8'-12') is compared
ROS 0.03 with peptide 2 (1'-12'). However, as the8'-12' sequence was
l"12'3D6 2 140 extended to 8'-14' (peptide 23) and 8"'' (peptide 22) bind-
l"4' 27 140 ing increased by 5- and 14-fold, respectively, when compared
ROS 0.06 to peptide 24 (8'-12') (Table 11). Interestingly, peptide 1 (1'-
ROS denotes rhodopsin outer segment disc membranes. 18') binds 7 times better than peptide 2 (1'-12') (Table I)
The dash (-) indicates no measurable binding to themonoclonal while peptide 22 (8'"') binds 43 times better than peptide
antibody. 2 (1'-12') (Table 11). These results suggest that, although
The IW values determined vary slightly with the preparation of
ROS, and themonoclonal antibody containing hybridoma cell culture extension of the 1'-12' peptide to 1'"' increases the binding
fluid. For example the I50 value for peptide 1'-12' was 28 / r (Table
~ affinity to monoclonal rho 3A6, the removal of unnecessary
I) and 59 ~ I M(Table 11). For this reason all peptides are compared to residues in the 1'-7' region results in a further substantial
the native sequences 1'-12' or 1'-18' in each assay. increase in binding affinity.
Monoclonal Antibody rho1C5 Binding Studies-In a similar
respectively, when compared to peptide 2 of 28 pM represent- fashion to the monoclonal binding studies with rho 3A6, the
ing a 36- and 13-fold decrease). The most important residues binding region for antibody rho 1C5 was localized to the 8'-
in interacting with antibody rho 3A6 were glutamic acid 8', 18' region (Fig. 3B, Table 11) and the importance of residues
Interaction
Antigen-Antibody 11773
respectively, in peptide 2 (1-12). The contributions of all
other side chains in the COOH-terminal sequence were insig-
nificant to antibody rho 3D6 binding. These results suggest
8 that the minimum sequence required for maximal antibody
binding is 1-4. This conclusion is in agreement with the
3 results of Hicks and Molday (1986) who showedthat peptides
0 6 of lengths 1-4 to 1-18 were equally effective inhibitors of
X
0 rho 3D6 antibody binding to immobilized bovine rhodopsin.
a
a 4
No inhibition was observed with the 1-2 peptide or with
peptide 2-13. Our results clearly show that removal of either
2n the ionized carboxyl group by substitution with an isosteric
amide group or removal of the @-carbon of alanine 1 by
2 replacement with glycine completely removed peptide binding
to antibody rho 3D6.
A very interesting observation can explain the cross-reac-
0 tivity between rho 3D6 and opsin from cone cells (Hicks and
-3 -2 -1 0 1 2
Molday, 1986). The sequence of the COOH terminus of rod
Log Antigon Concentration ( I (Hargrave and Fong, 1977) and green and red cone protein
FIG.4. Competitive inhibition of monoclonal antibody rho (Nathans et al., 1986) is shown in Fig. 5. Absolute identity in
3 A 6 binding to rhodopsin by synthetic peptide analogs of the sequence is observed for the residues suggested in this study
COOH-terminal region of rhodopsin. The closed circles denote to be important for rho 3D6 binding, that is, the side chains
peptides 2, AcR(1-12)-OH; the open triangles, peptide 24, AcR(8- of 4, 2, and 1 and the a-COOH groups of 1.The sequence
12)amide; the open squares,peptide 23, AcR(8-14) amide,and the
closed triangles, peptide 22, AcR(8-18) amide.
differences occur at positions 3 and 5 which were shown to
be unimportant for peptide binding to the antibody rho 3D6.
Obviously increase in the size of the side chain at position 3
1
from alanine in rhodopsin to serine in opsin is not able to
Rod
0 0 0
Thr - Ser - G l n - Val - Ala - Pro - Ala - COOH
interfere with antibody binding, suggesting that this side
chain may not be in the antibody-binding site.