Structure and Antigenic Activity of Rubella E l Glycoprotein Synthetic Peptides P. NERI,' M. CORTI,' 1. LOZZI,' and P.
E. VALENSIN'
'Dipartimento di Chimica and 'lstituto di Microbiologia, Universita di Siena, Siena, Italy
SYNOPSIS
Minimal sequences of rubella E l glycoprotein epitopes were previously identified as the tripeptide n50PER252 for the EP2 epitope, the tetrapeptide 260ADDP263 for the EP3 epitope, for the EP, epitope. and the tripeptide 273EVW275 plus the octapeptide 278PVIGSQAR285 In order to establish for each epitope the shortest sequence that was able to give the maximum binding with human antirubella immunoglobulins, synthetic peptides with increasing number of residues flanking these essential parts of rubella E l glycoprotein epitopes were synthesized and examined for their antigenic activity. Usually higher activity was observed with progressively longer homologues, whereas the additions of Pro-271, Pro-278 to 272GEVWVT277 peptide, and additions of Ala-248 to 249TPERP253 and 249TPERPR254, led to an abrupt decrease in binding. Taken together, our results indicated that the antigenic activity of the whole antigen could be dissected and reproduced using synthetic peptides of appropriate structure for each epitope.
I NTRODUCTlO N Synthetic peptides are enjoying increasing popularity in immunology as a tool both to understand the basic mechanism of the immune response, and to prepare new and better vaccines and diagnostic reagents as well immunoadsorbents for affinity chromatography.',' When dealing with the problem of using synthetic peptides as antigens, these must have or must be able to assume the proper three-dimensional structure to fit the paratope, in order to bind antibodies with the highest a f f i n i t ~ . ~ . ~ Since it is not easy to foresee which is the shortest sequence of a given B-cell epitope for obtaining optimal antigenic activity, multiple solid phase synthesis5 of large numbers of peptides related to a given epitope provides the systematic assessment of the effects of progressive amino acid elongation on recognition between epitopes and the corresponding antibodies.6
Hiopolymers, Vol 31, 631-635 (1991) (C 1991 .John Wdey & Sons, Inc
CCC 0006-3525/91/060631-05$04.00
By this method we have established the sequence of synthetic peptides able to reproduce the antigenic activity of E l Rubella virus glycoprotein, 7*8 in order to use them in place of the whole antigen in serological assay. Rubella virus ( R V ) consists of two envelope glycoprotein, E l and E2, and one nonglycosylated capsid protein, C.'-" Previously Terry e t al., '',I3 using El-specific monoclonal antibodies and peptides generated by proteolytic cleavage, showed that both the hemagglutination inhibition ( H I ) and the neutralization activities were associated with the E l protein and a n antigenic domain of 40 amino acids was l ~ c a l i z e d on ' ~ the rubella E l glycoprotein between residues 245-285. These authors identified a t least three epitopic regions ( EP1, E P 2 , E P 3 ) involved in the neutralization process of virus infection. In a previous paper we described the synthesis of overlapping octapeptides covering the region between residues 243 and 286 of E l RV glycoprotein, and in this way we were able to identify the essential part of the three epitopes: the tripeptide 250PER25', shared by all six octapeptides fairly active in the EP2 epitopic region; the tetrapeptide 260ADDP263, 631
632
NERI ET AL.
which is in common with five octapeptides found to be reactive in the EP3epitopic region; the tripeptide 273EVW275, which is shared by six fairly active octapeptides, plus the octapeptide 27sPVIGSQAR285, for the EP, epitopic region.15 On the basis of these results we can presume that each of these small peptides, corresponding or nearly corresponding to local maxima in acrophilicity plot,I6 could form the essential core of the corresponding epitope in a accessible region on the protein surface (Figure 1). In order to confirm that these sequences are the essential epitopic structures and to identify the minimum-length for maximum antibody binding, we have synthesized the predetermined sequences alone and also peptides of increasing length in the NH2 a n d / o r COOH terminal position, representing different-length peptide homologues of the regions identified a s critical for antigenic activities (Figure 2 ) . It must be noted that the binding of peptides with pooled human high-titer antirubella immunoglobulins from subjects hyperimmunized reflects the mean affinity of the reacting antibodies for the peptides.
MATERIALS A N D METHODS Synthesis was performed by the solid-phase procedure on acrylic acid-coated polyethylene rods, assembled in a 96-well microtiter plate pattern, using fluorenylmethyloxycarbonyl / t-butyl amino acid derivatives and pentafluorophenyl esters a s activated carboxyl groups.
1
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
PER TPER PERP TPERP ATPERP TPERPR ATPERPR GATPERPR ATPERPRL GATPERPRL VGATPERPRL GATPERPRLR ADDP DADDP ADDPL DADDPL VDADDPL DADDPLL VDADDPLL LVDADDPLL VDADDPLLR LVDADDPLLR EVW GEVW EVWV GEVWV PGEVWV GEVWVT PGEVWVT GPGEVWVT PGEVWVTP GPGEVWVTP PGPGEVWVTP GPGEVWVTPV VIGSQA PVIGSQA VIGSQAR PVIGSQAR TPV IGSQAR PVIGSQARK TPVIGSQARK
Figure 2. Sequences of peptides representing differentlength homologues of the rubella E l glycoprotein regions previously identified as critical for antibody binding.
After synthesis, peptides were deprotected using a mixture of trifluoroacetic acid/phenol/ethanedithiol(95/2.5/2.5) and, still bound to the rods, tested by ELISA for antibody-binding. Control peptides allowed to check the quality of peptide synthesis by amino acid analysis. T h e tips of the rods were coated for 1 h a t 25°C with 1% ovalbumin/ 1% bovine serum albumin/ -2 0.1% Tween 20 in phosphate-buffered saline p H 7.2 o~ block nonspecific adsorption, incubated -4 I D C S R L V G A T P ~ R P R ~ R ~ V ~ ~ D D P L L R T A P G ~ ~ E V ~ W V ~ V ~ G(~PQB AS R K) ~tG 240 250 260 270 280 290 overnight a 4°C with pooled human high-titer anSEQUENCE tirubella immunoglobulins from subjects hyperimmunized ( Istituto Sierovaccinogeno Italiano, S. AnFigure 1. Acrophilicity profile of 240-290 of rubella E l timo, Naples, Italy; H I activity 1/ 1024) and finally glycoprotein. Positions are numbered according to the washed in PBS/Tween 20. amino acid sequence. 4 1
I
633
STRUCTURE AND ACTIVITY OF RUBELLA E l GLYCOPROTEIN
4
T
0.4
- 100
-
0.3
75
r75f
1
0,3
t-
W
Ya
I! Z
w
m
2m
L
-
0.2
50
P t
g
Z
a m
a
T
a
W
I I-
. 25
0.1
c
15 W
3
W
50
0.2
9 Z
W
P Iz
U
2m
a
a
25
0.1
a:
w
I t
a
-I
W
LI
1
2
3 4
5
6
1 3 14 r'5
7 8 9101112
Ik
1'7
r
- 100
19 20 21 2 2
r
- 100
8
8
>
0 O.3 .
-
~
75
I t-
% I
> !I
!I
-
0 O.3 e41
75
a I!
I!
l -
Z
z
W
- 50
s?I-
W
- 50
z a
-1
t-
I
2
n
32 33 34
z
W
4W
a:
19 30
t-
I W
23 24 25 26 27
P a
W T
L
I-
0
a:
35 313 37 38
Figure 3. Results in ELISA of binding experiments of pooled human antirubella immunoglobulins by synthetic peptides shown in Figure 2. Results are expressed as the average of triplicate samples -t SD. Antibody-binding activity for each peptide is shown as a vertical line proportional to the absorbance value obtained after subtracting background absorbance value. Relative antigenic activity was expressed as percent ratio between the readings obtained using peptides and that using the whole antigen.
634
NERI ET AL.
Then, a solution of horseradish peroxidase/conjugate mouse antihuman immunoglobulins ( Sigma, St. Louis, MO) was added. The bound antibodies were detected by addition of the substrate 2,2'-azino-di-3-ethyl-benzthiazodin sulphonate (ABTS) and the colored resulting solution was measured in a Bio-Rad model 2550 EIA plate reader. Moreover, a positive control was obtained by linking rubella virus proteins ( Behring S.p.A., L'Aquila, Italy) to the rods by glutaraldehyde. Optical density (OD ) values were the mean of three determinations obtained after subtracting background (OD = 0.1) and the relative antigenic activities (raa) were expressed as a percentage of the value obtained using the whole antigen.
RESULTS AND DISCUSSION The results in Figure 3a indicated that the minimum-length region of the EP2 epitope capable of significant binding was the 250PERP253 tetrapeptide, even if its antigenic activity was lower than that of progressively longer homologues. In fact, whereas additions of residue Ala-248 to peptides 249TPERP 253 and 249TPERPR254 resulted in an abrupt decrease in the related OD values, additional residues to the C- and N-terminal positions led to an increase of binding. Remarkably, the addition of residue Val-246 to the peptide 247GATPERPRL255 resulted in a decapeptide that showed an antigenic activity similar to that of the whole antigen and about threefold higher than that of 250PERP253. Similarly, higher binding was observed when flanking residues were added to the essential region of EP3 epitope, the antigenically silent tetrapeptide 260ADDp263 However, in this case the best result was obtained by the sequence 257LVDADDPLLR266 with a raa of 65% (Figure 3 b ) . Finally, the two epitopes in the EP, region, the tripeptide 273EVW2 7 5 and the octapeptide 27sPVIGSQAR285, displayed a raa of 23 and 42%, respectively. By elongating these sequences, maximum raa of 84% was obtained with the 272GEVWVT277 peptide, whereas the additions of Pro-271 to 272GEVWVT277 peptide, of Pro-278 to 270GPGEVWVT277 peptide, and of Pro-269 to 270GPGEVWVTP278 peptide led to a substantial lower binding (Figure 3c). In the second case, the strong increase in activity
with the addition of Pro-278 and Arg-285 to 279VIGSQA284 peptide indicated that both these residues were essential to obtain an active structure. Further additions in C- and N-terminal positions increased the ability to bind specific antibodies, and the best result was obtained by the 277TPVIGSQARKZs6peptide (80% raa) (Figure 3d). That additional residues resulted in an increase of activity could be consistent with both the hypothesis that these amino acids directly participate to the epitope structure or only stabilize the conformation present in the parent protein. In conclusion, as far as the rubella E l glycoprotein, it appears that critical chain lengths and residues must be established for each epitope in order to obtain the best antigenic activity. Preliminary studies using peptide analogues of the 246VGATPERPRL255 containing single residue substitutions allowed more information on the critical amino acids: binding of pooled human antirubella immunoglobulins was drastically reduced by Pro-250, Glu-251, and Arg-252 replacements, confirming the tripeptide 250PER252 as the essential part of the EP2 epitope. Antigen-antibody interaction was not affected, carrying substitution of Arg-252 to a Lys residue. This suggested that a positive charge would play an important role, e.g., permitting the correct orientation of the peptide in the antibody paratope, without residue Arg-252 being directly involved in binding. Alternatively, this peptide could recognize, in the pool of antibodies used in the test, those directed toward an hypothetical mutant of E l glycoprotein with Lys in place of Arg-252. We are currently investigating the usefulness of such optimized epitopic structures as antigens in rubella serological analysis. Determination of their related antibodies is of relevant diagnostic interest and the use of synthetic peptides in place of the whole antigen would give a better test reliability. The authors wish to thank Miss Silvia Scali for her skillful technical assistance. This work has been supported by a grant of Consiglio Nazionale delle Ricerche, Progetto Finalizzato Chimica Fine.
REFERENCES 1. Arnon, R., Shapira M. & Jacob, C. 0. (1983) J. Immunol. Method 61, 261-273. 2. Gazin, C., Rigolet, M., Briand, J. P., Van Regenmortel,
STRUCTURE AND ACTIVITY OF RUBELLA E l GLYCOPROTEIN
3.
4. 5.
6. 7. 8. 9. 10.
M. H. V. & Galibert, F. (1986) EMBO J . 5, 22412250. Dyson, H. J., Cross, K. J., Houghten, R. A., Wilson, I. A., Wright, P. E. & Lerner, R. A. (1985) Nature 3 1 8 , 480-483. Sela, M. (1969) Science 166,1365-1374. Geysen, H. M., Rodda, S. J., Mason, T. J., Tribbick, G. & Schoofs, P. G. (1987) J . Immunol. Method 102, 259-274. Matthews, R. E. F. (1982) Intervirology 17, 1-199. Oker-Blom, C. (1984) J. Virol. 51, 354-358. Oker-Blom, C., Kalkkinen, N., Kaariainen, L. & Petterson R. F. (1983) J . Virol. 46,964-973. Waxham, M. N. & Wolinsky J. S. (1983) Virology 196, 194-203. Ho-Terry, L., Cohen, A. & Tedder, R. S. (1984) J . Med. Microbiol. 17, 105-109.
635
11. Ho-Terry, L. & Cohen, A. (1980) Arch. Virol. 65, 113. 12. Umino, Y., Sato, T. A., Katow, S., Matsuno, T. & Sugiura, A. (1985) Arch. Virol. 83, 33-427. 13. Terry, G. M., Ho-Terry, L., Loudesbourgh, P. & Rees, R. K. (1988) Arch. Virol. 98, 189-197. 14. Ho-Terry, L., Terry, G . M., Cohen, A. & Loudesbourgh, P. (1986) Arch. Virol. 90, 145-152. 15. Lozzi, L., Rustici, M., Corti, M., Cusi, M. G., Valensin,
P. E., Bracci, L., Santucci, A., Soldani, P., Spreafico, A. & Neri, P. (1990) Arch. Virol. 110, 271-276. 16. Hopp, T. P. (1986) J. Immunol. Method 88, 1-18.
Received July 28, 1990 Accepted September 7, 1990
E. VALENSIN'
'Dipartimento di Chimica and 'lstituto di Microbiologia, Universita di Siena, Siena, Italy
SYNOPSIS
Minimal sequences of rubella E l glycoprotein epitopes were previously identified as the tripeptide n50PER252 for the EP2 epitope, the tetrapeptide 260ADDP263 for the EP3 epitope, for the EP, epitope. and the tripeptide 273EVW275 plus the octapeptide 278PVIGSQAR285 In order to establish for each epitope the shortest sequence that was able to give the maximum binding with human antirubella immunoglobulins, synthetic peptides with increasing number of residues flanking these essential parts of rubella E l glycoprotein epitopes were synthesized and examined for their antigenic activity. Usually higher activity was observed with progressively longer homologues, whereas the additions of Pro-271, Pro-278 to 272GEVWVT277 peptide, and additions of Ala-248 to 249TPERP253 and 249TPERPR254, led to an abrupt decrease in binding. Taken together, our results indicated that the antigenic activity of the whole antigen could be dissected and reproduced using synthetic peptides of appropriate structure for each epitope.
I NTRODUCTlO N Synthetic peptides are enjoying increasing popularity in immunology as a tool both to understand the basic mechanism of the immune response, and to prepare new and better vaccines and diagnostic reagents as well immunoadsorbents for affinity chromatography.',' When dealing with the problem of using synthetic peptides as antigens, these must have or must be able to assume the proper three-dimensional structure to fit the paratope, in order to bind antibodies with the highest a f f i n i t ~ . ~ . ~ Since it is not easy to foresee which is the shortest sequence of a given B-cell epitope for obtaining optimal antigenic activity, multiple solid phase synthesis5 of large numbers of peptides related to a given epitope provides the systematic assessment of the effects of progressive amino acid elongation on recognition between epitopes and the corresponding antibodies.6
Hiopolymers, Vol 31, 631-635 (1991) (C 1991 .John Wdey & Sons, Inc
CCC 0006-3525/91/060631-05$04.00
By this method we have established the sequence of synthetic peptides able to reproduce the antigenic activity of E l Rubella virus glycoprotein, 7*8 in order to use them in place of the whole antigen in serological assay. Rubella virus ( R V ) consists of two envelope glycoprotein, E l and E2, and one nonglycosylated capsid protein, C.'-" Previously Terry e t al., '',I3 using El-specific monoclonal antibodies and peptides generated by proteolytic cleavage, showed that both the hemagglutination inhibition ( H I ) and the neutralization activities were associated with the E l protein and a n antigenic domain of 40 amino acids was l ~ c a l i z e d on ' ~ the rubella E l glycoprotein between residues 245-285. These authors identified a t least three epitopic regions ( EP1, E P 2 , E P 3 ) involved in the neutralization process of virus infection. In a previous paper we described the synthesis of overlapping octapeptides covering the region between residues 243 and 286 of E l RV glycoprotein, and in this way we were able to identify the essential part of the three epitopes: the tripeptide 250PER25', shared by all six octapeptides fairly active in the EP2 epitopic region; the tetrapeptide 260ADDP263, 631
632
NERI ET AL.
which is in common with five octapeptides found to be reactive in the EP3epitopic region; the tripeptide 273EVW275, which is shared by six fairly active octapeptides, plus the octapeptide 27sPVIGSQAR285, for the EP, epitopic region.15 On the basis of these results we can presume that each of these small peptides, corresponding or nearly corresponding to local maxima in acrophilicity plot,I6 could form the essential core of the corresponding epitope in a accessible region on the protein surface (Figure 1). In order to confirm that these sequences are the essential epitopic structures and to identify the minimum-length for maximum antibody binding, we have synthesized the predetermined sequences alone and also peptides of increasing length in the NH2 a n d / o r COOH terminal position, representing different-length peptide homologues of the regions identified a s critical for antigenic activities (Figure 2 ) . It must be noted that the binding of peptides with pooled human high-titer antirubella immunoglobulins from subjects hyperimmunized reflects the mean affinity of the reacting antibodies for the peptides.
MATERIALS A N D METHODS Synthesis was performed by the solid-phase procedure on acrylic acid-coated polyethylene rods, assembled in a 96-well microtiter plate pattern, using fluorenylmethyloxycarbonyl / t-butyl amino acid derivatives and pentafluorophenyl esters a s activated carboxyl groups.
1
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
PER TPER PERP TPERP ATPERP TPERPR ATPERPR GATPERPR ATPERPRL GATPERPRL VGATPERPRL GATPERPRLR ADDP DADDP ADDPL DADDPL VDADDPL DADDPLL VDADDPLL LVDADDPLL VDADDPLLR LVDADDPLLR EVW GEVW EVWV GEVWV PGEVWV GEVWVT PGEVWVT GPGEVWVT PGEVWVTP GPGEVWVTP PGPGEVWVTP GPGEVWVTPV VIGSQA PVIGSQA VIGSQAR PVIGSQAR TPV IGSQAR PVIGSQARK TPVIGSQARK
Figure 2. Sequences of peptides representing differentlength homologues of the rubella E l glycoprotein regions previously identified as critical for antibody binding.
After synthesis, peptides were deprotected using a mixture of trifluoroacetic acid/phenol/ethanedithiol(95/2.5/2.5) and, still bound to the rods, tested by ELISA for antibody-binding. Control peptides allowed to check the quality of peptide synthesis by amino acid analysis. T h e tips of the rods were coated for 1 h a t 25°C with 1% ovalbumin/ 1% bovine serum albumin/ -2 0.1% Tween 20 in phosphate-buffered saline p H 7.2 o~ block nonspecific adsorption, incubated -4 I D C S R L V G A T P ~ R P R ~ R ~ V ~ ~ D D P L L R T A P G ~ ~ E V ~ W V ~ V ~ G(~PQB AS R K) ~tG 240 250 260 270 280 290 overnight a 4°C with pooled human high-titer anSEQUENCE tirubella immunoglobulins from subjects hyperimmunized ( Istituto Sierovaccinogeno Italiano, S. AnFigure 1. Acrophilicity profile of 240-290 of rubella E l timo, Naples, Italy; H I activity 1/ 1024) and finally glycoprotein. Positions are numbered according to the washed in PBS/Tween 20. amino acid sequence. 4 1
I
633
STRUCTURE AND ACTIVITY OF RUBELLA E l GLYCOPROTEIN
4
T
0.4
- 100
-
0.3
75
r75f
1
0,3
t-
W
Ya
I! Z
w
m
2m
L
-
0.2
50
P t
g
Z
a m
a
T
a
W
I I-
. 25
0.1
c
15 W
3
W
50
0.2
9 Z
W
P Iz
U
2m
a
a
25
0.1
a:
w
I t
a
-I
W
LI
1
2
3 4
5
6
1 3 14 r'5
7 8 9101112
Ik
1'7
r
- 100
19 20 21 2 2
r
- 100
8
8
>
0 O.3 .
-
~
75
I t-
% I
> !I
!I
-
0 O.3 e41
75
a I!
I!
l -
Z
z
W
- 50
s?I-
W
- 50
z a
-1
t-
I
2
n
32 33 34
z
W
4W
a:
19 30
t-
I W
23 24 25 26 27
P a
W T
L
I-
0
a:
35 313 37 38
Figure 3. Results in ELISA of binding experiments of pooled human antirubella immunoglobulins by synthetic peptides shown in Figure 2. Results are expressed as the average of triplicate samples -t SD. Antibody-binding activity for each peptide is shown as a vertical line proportional to the absorbance value obtained after subtracting background absorbance value. Relative antigenic activity was expressed as percent ratio between the readings obtained using peptides and that using the whole antigen.
634
NERI ET AL.
Then, a solution of horseradish peroxidase/conjugate mouse antihuman immunoglobulins ( Sigma, St. Louis, MO) was added. The bound antibodies were detected by addition of the substrate 2,2'-azino-di-3-ethyl-benzthiazodin sulphonate (ABTS) and the colored resulting solution was measured in a Bio-Rad model 2550 EIA plate reader. Moreover, a positive control was obtained by linking rubella virus proteins ( Behring S.p.A., L'Aquila, Italy) to the rods by glutaraldehyde. Optical density (OD ) values were the mean of three determinations obtained after subtracting background (OD = 0.1) and the relative antigenic activities (raa) were expressed as a percentage of the value obtained using the whole antigen.
RESULTS AND DISCUSSION The results in Figure 3a indicated that the minimum-length region of the EP2 epitope capable of significant binding was the 250PERP253 tetrapeptide, even if its antigenic activity was lower than that of progressively longer homologues. In fact, whereas additions of residue Ala-248 to peptides 249TPERP 253 and 249TPERPR254 resulted in an abrupt decrease in the related OD values, additional residues to the C- and N-terminal positions led to an increase of binding. Remarkably, the addition of residue Val-246 to the peptide 247GATPERPRL255 resulted in a decapeptide that showed an antigenic activity similar to that of the whole antigen and about threefold higher than that of 250PERP253. Similarly, higher binding was observed when flanking residues were added to the essential region of EP3 epitope, the antigenically silent tetrapeptide 260ADDp263 However, in this case the best result was obtained by the sequence 257LVDADDPLLR266 with a raa of 65% (Figure 3 b ) . Finally, the two epitopes in the EP, region, the tripeptide 273EVW2 7 5 and the octapeptide 27sPVIGSQAR285, displayed a raa of 23 and 42%, respectively. By elongating these sequences, maximum raa of 84% was obtained with the 272GEVWVT277 peptide, whereas the additions of Pro-271 to 272GEVWVT277 peptide, of Pro-278 to 270GPGEVWVT277 peptide, and of Pro-269 to 270GPGEVWVTP278 peptide led to a substantial lower binding (Figure 3c). In the second case, the strong increase in activity
with the addition of Pro-278 and Arg-285 to 279VIGSQA284 peptide indicated that both these residues were essential to obtain an active structure. Further additions in C- and N-terminal positions increased the ability to bind specific antibodies, and the best result was obtained by the 277TPVIGSQARKZs6peptide (80% raa) (Figure 3d). That additional residues resulted in an increase of activity could be consistent with both the hypothesis that these amino acids directly participate to the epitope structure or only stabilize the conformation present in the parent protein. In conclusion, as far as the rubella E l glycoprotein, it appears that critical chain lengths and residues must be established for each epitope in order to obtain the best antigenic activity. Preliminary studies using peptide analogues of the 246VGATPERPRL255 containing single residue substitutions allowed more information on the critical amino acids: binding of pooled human antirubella immunoglobulins was drastically reduced by Pro-250, Glu-251, and Arg-252 replacements, confirming the tripeptide 250PER252 as the essential part of the EP2 epitope. Antigen-antibody interaction was not affected, carrying substitution of Arg-252 to a Lys residue. This suggested that a positive charge would play an important role, e.g., permitting the correct orientation of the peptide in the antibody paratope, without residue Arg-252 being directly involved in binding. Alternatively, this peptide could recognize, in the pool of antibodies used in the test, those directed toward an hypothetical mutant of E l glycoprotein with Lys in place of Arg-252. We are currently investigating the usefulness of such optimized epitopic structures as antigens in rubella serological analysis. Determination of their related antibodies is of relevant diagnostic interest and the use of synthetic peptides in place of the whole antigen would give a better test reliability. The authors wish to thank Miss Silvia Scali for her skillful technical assistance. This work has been supported by a grant of Consiglio Nazionale delle Ricerche, Progetto Finalizzato Chimica Fine.
REFERENCES 1. Arnon, R., Shapira M. & Jacob, C. 0. (1983) J. Immunol. Method 61, 261-273. 2. Gazin, C., Rigolet, M., Briand, J. P., Van Regenmortel,
STRUCTURE AND ACTIVITY OF RUBELLA E l GLYCOPROTEIN
3.
4. 5.
6. 7. 8. 9. 10.
M. H. V. & Galibert, F. (1986) EMBO J . 5, 22412250. Dyson, H. J., Cross, K. J., Houghten, R. A., Wilson, I. A., Wright, P. E. & Lerner, R. A. (1985) Nature 3 1 8 , 480-483. Sela, M. (1969) Science 166,1365-1374. Geysen, H. M., Rodda, S. J., Mason, T. J., Tribbick, G. & Schoofs, P. G. (1987) J . Immunol. Method 102, 259-274. Matthews, R. E. F. (1982) Intervirology 17, 1-199. Oker-Blom, C. (1984) J. Virol. 51, 354-358. Oker-Blom, C., Kalkkinen, N., Kaariainen, L. & Petterson R. F. (1983) J . Virol. 46,964-973. Waxham, M. N. & Wolinsky J. S. (1983) Virology 196, 194-203. Ho-Terry, L., Cohen, A. & Tedder, R. S. (1984) J . Med. Microbiol. 17, 105-109.
635
11. Ho-Terry, L. & Cohen, A. (1980) Arch. Virol. 65, 113. 12. Umino, Y., Sato, T. A., Katow, S., Matsuno, T. & Sugiura, A. (1985) Arch. Virol. 83, 33-427. 13. Terry, G. M., Ho-Terry, L., Loudesbourgh, P. & Rees, R. K. (1988) Arch. Virol. 98, 189-197. 14. Ho-Terry, L., Terry, G . M., Cohen, A. & Loudesbourgh, P. (1986) Arch. Virol. 90, 145-152. 15. Lozzi, L., Rustici, M., Corti, M., Cusi, M. G., Valensin,
P. E., Bracci, L., Santucci, A., Soldani, P., Spreafico, A. & Neri, P. (1990) Arch. Virol. 110, 271-276. 16. Hopp, T. P. (1986) J. Immunol. Method 88, 1-18.
Received July 28, 1990 Accepted September 7, 1990
