Download Development of a predictive model for VM for sertotype 0, M.Mahapatra

Development of a predictive model for
vaccine matching for serotype O FMDV
from serology and capsid sequence
D. Borley, S. Upadhyaya, D. Paton, R. Reeve
and Mana Mahapatra
Pirbright Laboratory
United Kingdom
FMDV Genome Organization
Lb
Cleavage sites
AUG
Lab
L
3C
2A
AUG
Unknown
3B 1-3
VPg
L
VP0
VP3
VP1 2A
P1-2A
2B
2C
3A
3C
P2
3D
AAA n
P3
3B 1-3
VP0
VP3
VP1 2A
2B
2C
3A
3C
3D
3B1-3
VP0
VP4
VP3
VP1
2A
VP2
2B
2C
2B
2C
3A
3C
3D
3A
3A
3C
3C
3D
3D
Non-Structural Proteins
Structural
Proteins
Capsid
RNA Synthesis/Replication, Helicase /polymerase/protease
Antigenicity, receptor binding
activity, virulence
Main Objectives
a)
- Develop a predictive model for
vaccine matching for serotype O
FMDV
- Genetic and/or structural
determinants of antibodymediated protection
5x
b)
To develop/improve FMDV
vaccine strain selection, and
design novel vaccine
3x
2x
VP1- blue
VP2- red
VP3- green
3x
Selection of suitable vaccine strains
Vaccination using killed viral capsid antigens is very important
for disease control
Seven serotypes and multiple subtypes of FMDV
Antigenic mis-match is one important cause for vaccine failure
Vaccine strains must be selected from the available pool and
the need for new vaccines must be identified– eg. muiltiple
outbreak in S. Korea
SAT3
Current vaccine selection - VNT or ELISA tests
SAT2
to measure the cross-reaction of a bovine
vaccinal sera with the field strain in question
Few studies have been carried out to measure
cross-protection directly
A
SAT1
O
C
Asia1
Problems with serological methods
● Time needed to grow up field viruses and not all grow
equally well
● Need for panels of vaccine strains and antisera
● Antisera are inherently variable
● Difficult to standardise tests or to have full confidence in
results without many repetition
● Vaccine selection based on serological methods always
not give same results as in vivo
Alternative approaches
I.
Matching ELISA using type specific monoclonal
antibodies (mAbs)
-Need panel of well defined mAbs able to recognise differences between
vaccine and field viruses
-Need to know which antigenic sites are the most important
(Mahapatra et al., 2008)
II. Antigenic Cartography
III. Sequencing of viral capsid and correlation of amino acid
changes to antigenic matching
-Still need to know which sites are the most significant
Viruses used -serotype O
EURO-SA
Serotype O viruses (n = 80)
Cathay
Asia
Africa
Bovine serum (n = 5)
Heatmap of viruses/antisera
III- Correlating capsid sequence to serology
5 sera X~80 viruses
- Capsid sequence determined
- Surface accessibility of each capsid residue determined
* 48 discrete regions (1-40 aa)
* Optimised serology result
* LME prediction model developed
VNT
ELISA
O BFS Multimer: VP1- Red, VP2 – Blue, VP3 - Green
(Borley et al. - manuscript in prep.)
Prediction model using VNT and Capsid sequence
For the first time serotype O vaccine strain can be
selected without recourse to serology
Predicted regions using model
VNT
Blue: VP2 70-79
Green: VP3 84-85
Red: VP3 219-220
ELISA
Blue: VP1 194-204
Green: VP3 84-85
Red: VP1 132-162
Individual aa tested
VNT
ELISA
VP2 191
VP1 198
VP1 138
VP3 219
VP2 191 Site 2
VP1 138
VP1 174 Site 1
VP2 194
O1 BFS
Neu. Ab Titre
500
Ab. repertoire of bovine sera
Epitope prediction (structure)
400
300
200
100
O BFS
O1K
Red. O BFS Red. O1K
ASN 190
ASN 190
ASN 190
ASN 190
SER 190 ASN 190 ASP 190
THR 191
THR 191
THR 191
THR 191
ASN 191 THR 191 GLN 191
GLU 192
GLU 192
GLU 192
GLU 192
ALA 192 THR 192
ASP 69
ASP 69
ASP 69
ASP 69
SER 70
0
Site1
Site 2
Site3
mAb escape mutants
(Mahapatra et al., 2012)
Site 4
Site 5
ASP 71
ASP 71
ASP 71
ARG 218
ARG 218
ALA 219
GLU 220
ASP 71
A
C
SAT-1
ASP 69
SER 69
ASP 70
GLY 70
THR 71
SER 71
GLU 221 GLN 219 ARG 220
GLN 220 GLN 221
(Borley et al., submitted)
VP2
VP3
Reverse genetics to test the residues
Type O cDNA clone
Mutations introduced – VP2 191 and VP3 219
Recombinant virus - recovered, characterised
Antigenic properties of the virus - serology
Serum antibody titre
100
80
60
40
20
0
Parent VP3 219 VP2 191
Conclusion
• Serological methods- time-consuming, still useful
• mAb-based assay- vaccine-sp, not feasible (limited resources)
• Sequence vs serology model: -promising, needs further work
- better rep. sera/viruses
- testing and validation
- may change in future
• VNT and ELISA – different results (aa)
• New epitope – VP2 191, VP3 219
What we would like to do next
• Refine the prediction model (type O):
-Include additional sera against antigenically distant vaccine
-Develop further model – to confirm the residues predicted in
the current model
• Test and validate the type O model:
- Introduce in to use by FMD Ref. labs
- Cross-protection studies
• Test the predicted aa residues using a cDNA clone
• Extend the work to serotype A
Acknowledgements
Pirbright
Daryl Borley
Sasmita Upadhyaya
Amin Asfor
Fufa Bari
David Paton
Jef Hammond
Univ. of Glasgow
Richard Reeve
Dan Haydon
Nico Visser
Danny Gooverts
Oxford
Elizabeth Fry
Dave Stuart
IZSLER, Brescia
Emi Brocchi
Santina Grazioli