Introduction to Computational Vaccinology and iVAX by EpiVax
-
Upload
annie-de-groot -
Category
Technology
-
view
513 -
download
1
description
Transcript of Introduction to Computational Vaccinology and iVAX by EpiVax
Using Computa.onal Vaccinology to Design Genome-‐Derived Vaccines for Infec.ous Diseases,
Cancer, Allergy and Autoimmune Disease
1
Anne S. De Groot, Lenny Moise, Leslie Cousens, Frances Terry, William Mar<n Ins<tute for Immunology and Informa<cs, University of Rhode Island and EpiVax, Inc. www.epvax.com www.immunome.org
22 January 2014
1
Your Speaker – Annie De Groot MD
2
The Company: EpiVax
3 hOp://bit.ly/EpiPubs
EpiVax Collaborates with the Ins*tute for Immunology and Informa*cs @ URI
Collabora<ve Research on Immunome-‐Derived Accelerated Vaccine Design and Development Funded by the NIH CCHI U19, COBRE, and P01 awards. www.immunome.org
4 hOp://bit.ly/EpiPubs
Addi.onal Collaborators
Bill Mar<n Lenny Moise Frances Terry Leslie Cousens Ryan Tassone Howie La<mer Mindy Cote Lauren Levitz Chris<ne Boyle
Alan Rothman Carey Medin Andres Gui<errez Danielle Aguirre Joe Desrosiers Thomas Mather Wendy Coy Loren Fast
Don Drake, Brian Schanen
Sharon Frey Mark Buller Jill Schreiwer
Hardy Kornfeld Jinhee Lee Liisa Selin
Connie Schmaljohn Lesley C. Dupuy
Ted Ross
Mark Poznansky Tim Brauns Pierre LeBlanc
AI058326, AI058376, AI078800, AI082642
5 hOp://bit.ly/EpiPubs
• Why Computational Immunology • Tools to Produce IDVs
– Antigen selection – Vaccine design – New concepts
• Case Studies
6
Outline
Predic<ng the future is something that weather experts do with the assistance of informa<cs models. These forecasts enable us to make decisions on a daily basis, and they are accurate enough to mobilize millions if and when severe storms are predicted. Why then, are we so slow to use informa<cs in vaccine and protein therapeu<cs design?
In todays talk, I will discuss the use of immunoinforma<cs tools for vaccine design, mechanism of ac<on studies, and efficacy evalua<ons. I believe that the <me is ripe for vaccine developers to ac<vely apply, evaluate and improve vaccines through the use of computa<onal immunogenicity predic<on tools.
“Old Style” Vaccines
Grow . . . and use whole pathogen
Whole (live/killed) vaccines
Subunit vaccines (Flu, Hepa<<s B, HPV vaccines, for
example)
Genome-‐Derived, Epitope Driven (GD-‐ED)
Vaccines
BeOer understanding of vaccine MOA
Improve vaccine safety and efficacy
Accelerate Vaccine Design
The focus of our work Can we make vaccines beJer/faster
10 hOp://bit.ly/EpiPubs
iVAX Vaccine Design Toolkit
• For Example: – HIV – HCV – Malaria – Universal Influenza Vaccine – Vaccines against Cancer – Vaccines for immunotherapy of AI – Vaccines for diseases affecting food animals
Why? New Vaccines Needed
• For Example: Pandemic influenza 2009 – Traditional flu vaccine production methods
require large lead time – 20 weeks to first vaccine dose – “Pandemic” influenza had already peaked by
the time the first shots were being delivered. – Vaccine manufacturing failed the test. – Is H7N9 the next pandemic? If so, we are
worried. . .
Why? Unacceptable Delays
Emergent H7N9 disease in China
14 hOp://bit.ly/EpiPubs
Spread to Beijing on 4/13/13 . . . Spread to Hong Kong on 12/6/ 13
15
Markedly Increased ac.vity in late 2013 and early 2014!
16 hOp://bit.ly/EpiPubs
Con.nuing Expansion of H7N9 First confirmed cases occurred in Shanghai (3/30/13) but case ac<vity rapidly increased in Zheijang and Jiangsu provinces shortly aier. Now, we have a problem!
Image credit to VDU and Dr. Ian M Mackay hOp://www.uq.edu.au/vdu/VDUInfluenza_H7N9.htm 17 hOp://bit.ly/EpiPubs
Ci.es that are one stop from H7N9
An es<mated 70% of the world popula<on resides within two hours’ travel <me of des<na<on airports (calculated using gridded popula<on-‐density maps and a data set of global travel <mes, map supplied by A. J. Tatem, Z. Huang and S. I. Hay (2013).
Quick numbers... • Total confirmed human cases of
influenza A virus H7N9: > 200
• Total deaths aOributed to infec<on with influenza A virus H7N9: > 50
• Case Fatality Rate (CFR): 29% (current)
• Average <me from illness onset to first confirma<on of H7N9 (days): <10
• Median age of the H7N9-‐confirmed cases (including deaths; years): 63
• Males: 71% of cases, 74% of deaths
• Younger pa<ents are recovering . . .
hOp://pandemicinforma<onnews.blogspot.com hOp://www.uq.edu.au/vdu/VDUInfluenza_H7N9.htm
H7N9 Morbidity and Mortality
19
Virus Transmission Mechanism – source is s.ll at large • Human to human transmission has not been proved (or disproved) many cases show uninfected family members
• Poultry iden<fied as poten<al natural host and H7N9 samples were found in poultry market environment in Shanghai. However not many poultry vendors infected and many cases have no indica<on of poultry exposure Image credit to VDU and Dr. Ian M Mackay hOp://
www.uq.edu.au/vdu/VDUInfluenza_H7N9.htm 20 hOp://bit.ly/EpiPubs
Distribu.on of Cases
This picture shows the
geographically wide distribu<on of flu cases -‐ sugges<ng widespread
distribu<on of the virus rather than a point outbreak.
21 hOp://bit.ly/EpiPubs
Why are immunoinforma.cs tools important in this sedng?
• Immunoinforma<cs predicted low immunogenicity of ‘cri<cal an<gen’ H7 HA
• hOp://bit.ly/H7N9_2013
(reminder) Flu Vaccine – HA protein
Ian Mackey hOp://www.uq.edu.au/vduVDUInfluenza_H7N9.htm 23
hOp://bit.ly/EpiPubs
What Can We Learn About H7N9?
HA (hemagglu<nin) is the ‘Cri<cal An<gen’ used for Flu vaccines, especially recombinant vaccines – – which are currently in produc*on.
24 hOp://bit.ly/EpiPubs
H7N9 is a unique virus
• Low conserva<on of HA, NA surface proteins is not surprising
• Internal proteins are more conserved
25 hOp://bit.ly/EpiPubs
gB-2 (EPX Score: -24.56)
- 80 -
- 70 -
- 60 -
- 50 -
- 40 -
- 30 -
- 20 -
- 10 -
- 00 -
- -10 -
- -20 -
- -30 -
- -40 -
- -50 -
- -60 -
- -70 -
- -80 -
Thrombopoietin
Human EPO
Tetanus Toxin
Influenza-HA
Albumin
IgG FC Region
EBV-BKRF3
Fibrinogen-Alpha
Follitropin-Beta
HA A/California/07/2009 (H1N1)
HA A/Victoria/361/2011 (H3N2)
HA A/Texas/50/2012 (H3N2)
HA A/Shanghai/1/2013 (H7N9) . . . . . . . .. . . . . . . . -‐8.11 HA A/mallard/Netherlands/09/2005 (H7N7) . . . . . . -‐8.63
Random Expecta.on
HA A/mallard/Netherlands/12/2000 (H7N3) .. . . . . .-‐9.91
HA A/chicken/Italy/13474/1999 (H7N1) . . . . . . . . . -‐6.23
H7 HA Immunogenic Poten.al
New H7N9 Flu is Predicted to be POORLY IMMUNOGENIC
hOp://bit.ly/EpiPubs
Why are immunoinforma.cs tools important in this sedng?
• Immunoinforma<cs predicted low immunogenicity of ‘cri<cal an<gen’ H7 HA
• Vaccine was developed but is low immunogenicity as predicted.
hOp://bit.ly/H7N9_NovaVax
Unadjuvanted Influenza Vaccine Effectiveness
Why are immunoinforma.cs tools important in this sedng?
• Immunoinforma<cs predicted low immunogenicity of ‘cri<cal an<gen’ H7 HA
• Vaccine was developed but is low immunogenicity as predicted
• Sero-‐conversion is delayed, diminished in pa<ents infected with H7N9.
hOp://bit.ly/H7N9_Serology
. . . Low and Slow . . .
Why are immunoinforma.cs tools important in this sedng?
• Immunoinforma<cs predicted low immunogenicity of ‘cri<cal an<gen’ H7 HA
• Vaccine was developed but is low immunogenicity as predicted
• Sero-‐conversion is delayed, diminished in pa<ents infected with H7N9.
• New vaccine approaches are needed. • . . . Now that you are convinced, let’s talk about computa<onal vaccine design
• Why Computational Immunology • Tools to Produce IDVs
– Antigen selection – Vaccine design – New concepts
• Case Studies
31
Outline
Computational Vaccinology: Genomes-to-Vaccines
• Lots of Genomes now Published! • On line tools for Pathogen Gene finding
(GLIMMER, ORPHEUS, GeneMark) • Tools for selecting subsets of protein –
such as subcellular localization of hypothetical proteins (PSORTb, CELLO, Proteome Analyst)
Selection of vaccine antigens is key
Strain 1
Strain 3
Strain 2
core genome dispensable genes
strain-‐specific genes pangenome
Comparative Genomics Impacts Vaccine Immunogen Selection
. . . Need “informa*on” = T cell and B cell epitopes
. . . And the correct “milieu”
= delivery vehicle, adjuvants/TLR ligands
“Fine tune” the immune response?
. . And there is ample evidence that this approach to vaccine design produces protective immunity
Immunome-Derived Vaccines . . .
Payload
Adjuvant
Delivery Vehicle
Vaccine
HLA (Human MHC), are comprised of peptide specific pockets
EpiMatrix predicts how well a peptide sequence will bind to a specific pocket.
Binding is the prerequisite for immunogenicity
8 class II HLA supertypes which taken together incorporate 95% of human
populations (and pockets) worldwide.
Each 9-mer/10-mer is analyzed for binding potential to each of those 8
allele matrices. .
Payload: Predic.ng Epitopes that Drive Immune Response is our Exper.se
Mature APC
Protein MHC II Pocket
Southwood et al. J. Immunology 1998 Sturniolo et al. Nature Biotechnology, 1999
The EpiMatrix Score describes the binding affinity of the pep<de sequence to the HLA complex
Peptide Epitope
37 hOp://bit.ly/EpiPubs
epitope
Vaccine an<gen
1 + 1 + 1 = Response
epitope epitope
Immune response to a vaccine an<gen can be predicted by measuring the number of T cell epitopes contained in the an<gen with immunoinforma<cs tools.
How do we measure Immunogenicity?
hOp://bit.ly/EpiPubs
Non Immunogenic
proteins
Immunogenic proteins
“Immunogenicity Scale”
41 hOp://bit.ly/EpiPubs
Easy easy to deliver as pep<des
42
DRB1*0101
DRB1*0301
DRB1*0401
DRB1*0701
DRB1*0801
DRB1*1101
DRB1*1301
DRB1*1501
ClustiMer: Screen for Epitope Clusters
43
Conservatrix: Overcome the Challenge of Variability
HIV HCV Influenza
44
Identifying the most conserved 9-mers allows for protection against more strains with fewer epitopes
Conservatrix Finds Conserved 9-mers
Conserved epitope
CTRPNNTRK
CTRPNNTRK CTRPNNTRK
CTRPNNTRK CTRPNNTRK
CTRPNNTRK
CTRPNNTRK
45
BlastiMer: Epitope Exclusion
Confidential
In all of our vaccines we eliminate cross-‐reac<ve epitopes
Self Foreign
Human
Pathogen
Human
Microbiome
Protec.ve epitopes
Poten.ally detrimental cross-‐reac.ve epitopes
Poten.ally detrimental cross-‐reac.ve epitopes
Epitope Cross-‐Reac<vity Impacts Vaccine Immunogen Selec<on
46 hOp://bit.ly/EpiPubs
Each MHC ligand has two faces, The MHC-binding face (aggretope), and the TCR-interacting face (epitope)
JanusMatrix
TCR
MHC
The JanusMatrix algorithm searches for putative MHC ligands which are identical at the contact residues but may vary at the MHC-binding residues.
MHC/HLA
TCR
• Identical T cell-facing residues • Same HLA allele and minimally
different MHC-facing residues
Find predicted 9-mer ligands with:
http://bit.ly/JanusMatrix
48
HCV T Effector Epitopes
HCV_G1_1605
HCV_G1_DEXDC_1246
HCV_G1_NS5A_1988
HCV_G1_NS4B_1725
HCV_G1_2898
HCV_G1_2913
HCV_G1_ENV_359
HCV_G1_2941
HCV_G1_NS4B_1910
HCV_G1_ENV_255
HCV_G1_2440
HCV_G1_NS2_732
HCV_G1_NS2_748
HCV_G1_2840
HCV_G1_1941
HCV_G1_NS4B_1769
HCV_G1_NS2_909
HCV_G1_2485
HCV_G1_NS4b_1798 HCV_G1_NS4B_1790
HCV_G1_NS4B_1876
HCV_G1_2879
Treg-‐like-‐Epitope: HCV
HCV_G1_NS2_794
• Why Computational Immunology • Tools to Produce IDVs
– Antigen selection – Vaccine design
• Case Studies
51
Outline
Epi-Assembler
Immunogenic consensus
CTRPNNTRK CTRPNNTRK
CTRPNNTRK CTRPNNTRK
CTRPNNTRK
CTRPNNTRK
EpiAssembler Constructs Immunogenic Consensus Sequences
STRAIN 01 Q X S W P K V E Q F W A K H X W N X I S X I Q Y LSTRAIN 02 Q A S W P K V E X F W A K H M W N F I S G I Q Y LSTRAIN 03 Q X S W P K X E Q F W A K H M W N F I S G I Q Y XSTRAIN 04 Q A S W X K V E Q F W A K H M W N F X S X I Q Y LSTRAIN 05 Q X S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 06 Q A S W P K X E Q F W A X H M W N F I S G I Q Y XSTRAIN 07 Q X S W P K V E Q F W A K H M X N F I S G I Q Y LSTRAIN 08 Q A S W X K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 09 Q X S W P K X E Q F W A K H M W N F X S X I X Y XSTRAIN 10 Q A S W P R V E Q F W A K H M W N F I X G I Q Y LSTRAIN 11 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 12 Q A S W X K V E Q F W A X H M W N F I S G I Q Y XSTRAIN 13 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 14 Q A S W X K X E Q F W A K H M W N F I S X I Q Y LSTRAIN 15 Q A S W P K V E X F W X K H M W N F I S G I Q Y LSTRAIN 16 Q X S W P K V E Q F W A K H M W N F I X G I Q Y LSTRAIN 17 X A S W X K V E Q F W A K H M W N F I S G I Q Y XSTRAIN 18 Q X S W P K X E Q F W A K H M W N X I S G I Q Y LSTRAIN 19 Q A S W X K V E Q F W A K H M W N F I S X I Q Y LSTRAIN 20 Q A S W P K V E Q F W A X H M W N F I S G I Q Y L
x
F W A K H M W N F
EpiAssembler: Core Epitope
STRAIN 01 Q X S W P K V E Q F W A K H X W N X I S X I Q Y LSTRAIN 02 Q A S W P K V E X F W A K H M W N F I S G I Q Y LSTRAIN 03 Q X S W P K X E Q F W A K H M W N F I S G I Q Y XSTRAIN 04 Q A S W X K V E Q F W A K H M W N F X S X I Q Y LSTRAIN 05 Q X S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 06 Q A S W P K X E Q F W A X H M W N F I S G I Q Y XSTRAIN 07 Q X S W P K V E Q F W A K H M X N F I S G I Q Y LSTRAIN 08 Q A S W X K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 09 Q X S W P K X E Q F W A K H M W N F X S X I X Y XSTRAIN 10 Q A S W P R V E Q F W A K H M W N F I X G I Q Y LSTRAIN 11 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 12 Q A S W X K V E Q F W A X H M W N F I S G I Q Y XSTRAIN 13 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 14 Q A S W X K X E Q F W A K H M W N F I S X I Q Y LSTRAIN 15 Q A S W P K V E X F W X K H M W N F I S G I Q Y LSTRAIN 16 Q X S W P K V E Q F W A K H M W N F I X G I Q Y LSTRAIN 17 X A S W X K V E Q F W A K H M W N F I S G I Q Y XSTRAIN 18 Q X S W P K X E Q F W A K H M W N X I S G I Q Y LSTRAIN 19 Q A S W X K V E Q F W A K H M W N F I S X I Q Y LSTRAIN 20 Q A S W P K V E Q F W A X H M W N F I S G I Q Y L
x
F W A K H M W N FW P K V E Q F W A
Q A S W P K V E Q N F I S G I Q Y LM W N F I S G I Q
EpiAssembler: Flanking Epitopes
STRAIN 01 Q X S W P K V E Q F W A K H X W N X I S X I Q Y LSTRAIN 02 Q A S W P K V E X F W A K H M W N F I S G I Q Y LSTRAIN 03 Q X S W P K X E Q F W A K H M W N F I S G I Q Y XSTRAIN 04 Q A S W X K V E Q F W A K H M W N F X S X I Q Y LSTRAIN 05 Q X S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 06 Q A S W P K X E Q F W A X H M W N F I S G I Q Y XSTRAIN 07 Q X S W P K V E Q F W A K H M X N F I S G I Q Y LSTRAIN 08 Q A S W X K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 09 Q X S W P K X E Q F W A K H M W N F X S X I X Y XSTRAIN 10 Q A S W P R V E Q F W A K H M W N F I X G I Q Y LSTRAIN 11 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 12 Q A S W X K V E Q F W A X H M W N F I S G I Q Y XSTRAIN 13 Q A S W P K V E Q F W A K H M W N F I S G I Q Y LSTRAIN 14 Q A S W X K X E Q F W A K H M W N F I S X I Q Y LSTRAIN 15 Q A S W P K V E X F W X K H M W N F I S G I Q Y LSTRAIN 16 Q X S W P K V E Q F W A K H M W N F I X G I Q Y LSTRAIN 17 X A S W X K V E Q F W A K H M W N F I S G I Q Y XSTRAIN 18 Q X S W P K X E Q F W A K H M W N X I S G I Q Y LSTRAIN 19 Q A S W X K V E Q F W A K H M W N F I S X I Q Y LSTRAIN 20 Q A S W P K V E Q F W A X H M W N F I S G I Q Y L
x
F W A K H M W N FW P K V E Q F W A
Q A S W P K V E Q N F I S G I Q Y LM W N F I S G I Q
Q A S W P K V E Q F W A K H M W N F I S G I Q Y L
EpiAssembler: Final Immunogenic Consensus Sequence
VaxCAD Identifies and Eliminates Junctional Epitopes
VaxCAD will identify junctional epitopes and rearrange chosen epitopes to reduce junctional epitope formation
57
-10
0
10
20
30
40
50
HP
4117
H
P41
79
HP
4007
H
P41
11
HP
4018
H
P40
70
HP
4034
H
P41
93
HP
4065
H
P41
81
HP
4157
H
P40
60
HP
4068
H
P41
64
HP
4160
H
P41
75
HP
4127
H
P41
20
HP
4126
H
P41
54
HP
4168
H
P41
19
HP
4100
H
P40
01
HP
4061
EpiM
atrix
Clu
ster
Sco
re
Peptides in Default order in construct HP_IIB
Epitope Cluster Score Junctional Cluster Score
-10
0
10
20
30
40
50
HP
4117
H
P40
61
HP
4181
H
P41
11
HP
4018
H
P40
70
HP
4060
H
P41
57
HP
4065
H
P40
01
HP
4193
H
P40
34
HP
4068
H
P41
68
HP
4160
H
P41
75
HP
4127
H
P41
26
HP
4007
H
P41
54
HP
4164
H
P41
19
HP
4100
H
P41
20
HP
4179
EpiM
atrix
Clu
ster
Sco
re
Peptides in Optimized order in construct HP_IIB
Epitope Cluster Score Junctional Cluster Score
VaxCAD Example
58
DNA Vector
DNA insert
Intended Protein Product: Many epitopes strung together in a “String-of-Beads”
Protein product (folded)
Multi-Epitope Gene Design
DNA – chain of epitopes, or pep<de in liposomes ICS-‐op<mized proteins in VLP ICS-‐op<mized whole proteins
Immunogenic Consensus Sequence Formulations
HLA A2
HLA DR3
HLA B7
HLA DR2
HLA A2/DR1
HLA DR4
In Vivo Model for Validation: HLA Transgenic Mice
• Why Computational Immunology • Tools to Produce IDVs • Case Studies
– Tularemia – Smallpox – H. pylori – VEEV (multi-pathogen vaccine) – Influenza
61
Outline
Burk/Tuly/MP
Current Vaccine Design Pipeline
Epitope Discovery
Epitope Validation
Construct Design
Immuno-genicity
HIV/TB Epitope Discovery
Epitope Validation
Construct Design
Immuno-genicity
Tularemia Epitope Discovery
Epitope Validation
Construct Design
Immuno-genicity
Animal Model Validation
Smallpox Epitope Discovery
Epitope Validation
Construct Design
Animal Model Validation
VEEV
Epitope Discovery
Epitope Validation
Construct Design
Animal Model Validation H. pylori
Epitope Discovery
Epitope Validation
Construct Design
Animal Model Validation
Animal Model Validation
Animal Model Validation
Immuno-genicity
Immuno-genicity
Immuno-genicity
62
Epitope Discovery
Epitope Validation
Construct Design
Animal Model Validation
Immuno-genicity Influenza
GDV Approach Applied to F. tularensis
63
McMurry JA, Gregory SH, Moise L, Rivera DS, Buus S, and De Groot AS. Diversity of Francisella tularensis Schu4 an<gens recognized by T lymphocytes aier natural infec<ons in humans: Iden<fica<on of candidate epitopes for inclusion in a ra<onally designed tularemia vaccine. Vaccine 2007 Apr 20;25(16):3179-‐91.
In 24 months:
• Took one genome
• Mapped class I + Class II
• Selected 165 epitopes
• Confirmed in human
• Cloned into vaccine
• Performed Challenge studies. . .
High Responder Frequency to Class II Epitopes in Pa.ents with Prior Exposure
64
Percent of subjects responding by IFN gamma ELISpot Significant Spot Forming Cells averaged across subjects
22/25 pep<des; Average response to the pool was over 1,000 gamma producing cells per million above background.
TulyVax: 6 epitope in LVS Challenge Strain
0
50
100
150
200
250
30030
04
3005
3017
3018
F102
F176
3001
3003
3015
3019
3007
3023
3024
3025
Schu4 peptides with perfect LVS match
Schu4 peptides with partial LVS match
Schu4 peptides without LVS match
IFN
-g S
FC
/10^
6 sp
leno
cyte
s ov
er b
ackg
roun
d
Placebo-immunizedFT_II_v1-immunized
950 -
900 -
TulyVax Immunogenicity in HLA Tg Epitope-‐specific IFNγ Response
Nearly identical immunogenicity profile observed in HLA DR3 mouse immunizations performed in collaboration with Dr. Terry Wu (UNM), illustrating broad reactivity of immunoinformatic predicted epitopes.
57%
0%0%
20%
40%
60%
80%
100%
0 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21Days after lethal bacterial challenge
Perc
ent S
urvi
val
TuliVax Immunized MicePlacebo Recipient Mice
14 epitopes: T cell-‐epitope-‐immunized mice were protected against live challenge with tularemia. Placebo-‐recipient mice died within 10 days.
Rapidity: from genome to candidate vaccine in 24 months . . . Efficacy: 14 epitope vaccine protects against live challenge
TulyVax Efficacy
McMurry et al. Vaccine 2007;25:3179-91 and Gregory et al. Vaccine 2009 27:5299-306
Vaccine
Immunogenic Epitopes
Shared Immunogenic Epitopes
smallpox
vaccinia
Immunome-Derived Smallpox Vaccine: VennVax
88% of predicted T cell epitopes confirmed in vitro using hu PBMC
20
VennVax Class II Epitopes are Antigenic in Dryvax Vaccinees
Moise et al. Vaccine. 2009 27:6471-9
Immunogenicity Day 56
1. epitope DNA vaccine prime (IM) 2. epitope peptide boost (IN)
Immunizations Days 0, 14, 28, 42
Challenge Day 65
VennVax Immunization in HLA DR3 Transgenic Mice
Moise L et al. Vaccine. 2011;29:501-11
Survival of VennVax-‐Vaccinated Mice Aqer Aerosol Challenge
73
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Perc
ent S
urvi
val
Day Post Immunization
Placebo
Vaccinated
DNA DNA boost boost Challenge 17%
0 20 40 60 80 100
100% survival of Vaccinated mice vs. 17% of placebo
Moise et al. Vaccine. 2011; 29:501-11
0
0.5
1
1.5
2
2.5
3
100 200 400 800 1600 3200 6400 12800
OD
490
1/Dilution Factor
Pre-challenge Placebo
Pre-challenge Vaccine
Post-challenge Placebo
Post-challenge Vaccine
Post-challenge
Pre-challenge
Protection Without Vaccine-Induced Antibodies
Therapeutic H. pylori Vaccination
Week 0 Week 6 Week 12-19 Week 51
IFN-gamma and IL-4 ELISpot
Histology
1. epitope DNA vaccine prime IM 2. epitope peptide boost IN
H. pylori SS1
H. pylori SS1
H. pylori SS1
H. pylori SS1 lysate IN
1. epitope DNA vaccine prime IN 2. epitope peptide boost IN
1. control DNA prime IN 2. control peptide boost IN
H. pylori SS1
IFN-gamma Secretion in Response to Splenocyte Restimulation following immunization
0
100
200
300
400
500
600
700
HP
4009
HP
4029
HP
4032
HP
4040
HP
4054
HP
4055
HP
4067
HP
4071
HP
4077
HP
4152
HP
4153
HP
4156
HP
4165
HP
4174
HP
4189
HP
4197
HP
4199
HP
POO
L 1
HP
POO
L 2
HP
POO
L 3
HP
4018
HP
4060
HP
4068
HP
4070
HP
4111
HP
4117
HP
4119
HP
4120
HP
4127
HP
4154
HP
4157
HP
4160
HP
4164
HP
4175
HP
4179
HP
POO
L 4
HP
POO
L 5
HP
POO
L 6
Con
A
SFC/
10̂6
over
bac
kgro
und
Average Helico-Vax
Average SS1
SS1 (whole lysate-immunized mice) recognized few epitopes (white bars); HelicoVax-immunized mice recognized 45 of 50 (dark bars). 45/50 were immunogenic.
HelicoVax: Broad Epitope Recognition
Lysate pVAX DNA IM DNA IN
0
20
40
60
80
100
120
140
160
180
600
800
H. p
ylor
i qPC
R(S
SA/G
APDH
)
*** P<0.001
** P<0.01
*** P<0.001
HelicoVax Eradicates H. pylori Infection
This result accomplished in just over 24 months . . .
Moss et al, Vaccine 2011;29:2085-91
Two Whole Gene Constructs – Ebola Zaire GP – VEEV 26S* – subcloned into pWRG-7077
*Dupuy LC, Richards MJ, Ellefsen B, Chau L, Luxembourg A, Hannaman D, Livingston BD, Schmaljohn CS. A DNA Vaccine for Venezuelan Equine Encephalitis Virus Delivered by Intramuscular Electro-poration Elicits High Levels of Neutralizing Antibodies in Multiple Animal Models and Provides Protective Immunity to Mice and Nonhuman Primates. Clin Vaccine Immunol. 2011 Mar 30.
One Multi-Epitope Construct – Ebola Zaire/Sudan GP epitopes – VEEV 26S epitopes – subcloned into pWRG-7077
VS.
VEEV IDV Development: Comparison with Whole Antigen Vaccine
IFNγ ELISpot responses to VEEV peptide pools
VEEV E1 VEEV E2
USAMRIID DR3 Mouse StudyVEEV Challenge Group ELISA
Day 56 Serum Samples
Neg Con Arm Pos Con Arm Vaccine Arm0
1
2
3
4
5
Log 1
0 Ti
ter
VEEV IDV Elicits Antibody Response
Negative Control
Negative Control
Whole Antigen Vaccine
Whole Antigen Vaccine
Epitope-Driven Vaccine
Epitope-Driven Vaccine
VEEV IDV Protects Against Lethal Challenge
USAMRIID DR3 Mouse StudyVEEV Challenge Survival
0 5 100
102030405060708090
100Neg Con ArmPos Con ArmVaccine Arm
Days postchallenge
Perc
ent s
urvi
val
USAMRIID DR3 Mouse StudyVEEV Challenge Weights
0 1 2 3 4 5 6 7 8 9 10 11 12 1350
60
70
80
90
100 Neg Con ArmPos Con ArmVaccine Arm
Days Postchallenge
% M
ean
Star
ting
Wei
ght
Whole Antigen Negative Control
Epitope-Driven
Vaccine
Negative Control
Whole Antigen Vaccine
Epitope-Driven Vaccine
T helper Epitopes B cell epitopes
Other? CTL? Th2?
Subset of Th epitopes stimulate IFNγ secretion""Combination of immunogenic Th epitopes that overlap B cell epitopes???" "Contribution from other Th epitopes (stimulate other cytokines) that overlap with B-cell epitopes""""Th epitopes that stimulate different subpopulations""""What is clear: that whole Ag is not necessary for protection"
What Drives Protection?
T cells = Immune System Body Armor
T cell response cannot prevent Infec<on but . . .
T cell response can arm against Disease
The "New" Flu (H1N1 2009 California)
84 hOp://bit.ly/EpiPubs
2009 Worry: CDC – No Cross-‐reac.ve Ab
• Preliminary studies of individuals showed that an<bodies induced by seasonal influenza vaccina<on were not cross-‐reac<ve with novel H1N1.
• What if the T cell epitopes were cross-‐reac<ve? Would that help?
• (Note that the situa<on is very similar for H7N9 – no cross-‐reac<ve an<body).
Centers for Disease Control and Preven<on. Serum an<body response to a novel influenza A (H1N1) virus aier vaccina<on with seasonal influenza vaccine. MMWR Morb Mortal Wkly Rep 2009;58(19):521–4.
85 hOp://bit.ly/EpiPubs
2009 H1N1 contains conserved epitope Sequences – Predicted Cross Protec.on
TIV
2008-‐2009 HA and NA
Novel H1N1
HA and NA
Conserved T-‐Cell Epitopes
Immunogenic T cell
epitopes
TIV
2008-‐2009 HA and NA
Novel H1N1
HA and NA
Conserved T-‐Cell Epitopes
Immunogenic T cell
epitopes
TIV
2008-‐2009 HA and NA
Novel H1N1
HA and NA
Conserved T-‐Cell Epitopes
Immunogenic T cell
epitopes
TIV
2008-‐2009 HA and NA
Novel H1N1
HA and NA
Conserved T-‐Cell Epitopes
Immunogenic T cell
epitopes
Conserved T-Cell
Epitopes
Immunogenic T cell
epitopes
De Groot et al. Vaccine 2009;27:5740-7
Enough Cross-‐protec<ve Epitopes that Seasonal Flu vaccina<on or
exposure may protect
86 hOp://bit.ly/EpiPubs
hOp://www.ncbi.nlm.nih.gov/pubmed/19660593
EpiVax Predicted Cross-‐Protec.on
87 hOp://bit.ly/EpiPubs
1.00E+06
1.00E+07
1.00E+08
Placebo FluVax 2009
Placebo FluVax 2009
PFU/m
l
2 Days 4 Days
*P= 0.002
Immuniza.on with FluVax cross-‐conserved T cell epitopes decreases lung viral load
108
107
106
Post-‐Infec.on
A handful of conserved epitopes protected against disease
hOp://bit.ly/Moise_Universal_Flu hOp://bit.ly/H1N1_DR3_2013
90 hOp://bit.ly/EpiPubs
H1N1 Conclusions
• This work recapitulates other projects already completed: Complete protection using ONLY T cell epitopes (H. pylori, Tularemia, VennVax)
• Results of our published studies demonstrate that conserved T cell epitope sequences, important to viral fitness, also may be immunologically significant contributors to protection against newly emerging influenza strains.
• The conserved epitope approach promises to answer the need for prompt preparedness and delivery of a safe, efficacious vaccine without requiring a new vaccine for every emergent influenza strain.
hOp://bit.ly/Moise_Universal_Flu hOp://bit.ly/H1N1_DR3_2013
91 hOp://bit.ly/EpiPubs
What about H7N9?
92 hOp://bit.ly/EpiPubs
What Can We Learn About H7N9? Epitopes Novel or Conserved?
H7N9 Circula<ng Flu
As it turns out -‐ -‐ -‐ Very Poor Cross-‐Conserva<on – Only within Internal Proteins
93 hOp://bit.ly/EpiPubs
gB-2 (EPX Score: -24.56)
- 80 -
- 70 -
- 60 -
- 50 -
- 40 -
- 30 -
- 20 -
- 10 -
- 00 -
- -10 -
- -20 -
- -30 -
- -40 -
- -50 -
- -60 -
- -70 -
- -80 -
Thrombopoietin
Human EPO
Tetanus Toxin
Influenza-HA
Albumin
IgG FC Region
EBV-BKRF3
Fibrinogen-Alpha
Follitropin-Beta
HA A/California/07/2009 (H1N1)
HA A/Victoria/361/2011 (H3N2)
HA A/Texas/50/2012 (H3N2)
HA A/Shanghai/1/2013 (H7N9) . . . . . . . .. . . . . . . . -‐8.11 HA A/mallard/Netherlands/09/2005 (H7N7) . . . . . . -‐8.63
Random Expecta.on
HA A/mallard/Netherlands/12/2000 (H7N3) .. . . . . .-‐9.91
HA A/chicken/Italy/13474/1999 (H7N1) . . . . . . . . . -‐6.23
H7 HA Immunogenic Poten.al
New H7N9 Flu is Predicted to be POORLY IMMUNOGENIC
hOp://bit.ly/H7N9_HVandI
This is a unique virus
• Low conserva<on of HA, NA surface proteins is not surprising
• Internal proteins are more conserved • And – HA is has unusually low immunogenicity • Could that explain why infec<on is widespread?
• Difficult to make an<bodies to the HA
96 hOp://bit.ly/EpiPubs
Differen<al Cross-‐reac<vity with the human genome-‐ significance?
H1N1 H7N9
97 hOp://bit.ly/EpiPubs
New and unpublished: The “Classic Epitope” Is much more cross-‐conserve with the human genome in the case of H7N9.
This is a unique virus
• Unusually low immunogenicity • Cross-‐reac<vity with human genome • How do we overcome this problem?
98 hOp://bit.ly/EpiPubs
99 hOp://bit.ly/EpiPubs
• EpiMatrix – maps T cell epitopes • ClustiMer - Promiscuous / Supertype Epitopes • BlastiMer - Avoiding “self” - autoimmunity • Conservatrix – Identifies Conserved Segments • EpiAssembler - Immunogenic Consensus Sequences • Aggregatrix – Optimizing the coverage of vaccines • VaxCAD - Processing and Assembly
Immunoinforma.cs Toolkit
Seamless Vaccine Design
Integrated toolkit is
unique to iVax
100 hOp://bit.ly/EpiPubs
FastVax: Vaccines on demand
• High throughput computing
• Immunoinformatics
• Vaccine design algorithms
• Vaccine Production
• Delivery device
• Animal safety/tox/immunogenicity/validation
• Deployment by established distribution systems
Prebuilt
Rapid deployment when genome
sequence is in hand
Pilot program Funded by DARPA
101 hOp://bit.ly/EpiPubs
20 hours -‐ April 05 – April 06 2013 Extremely Rapid H7N9 Vaccine Design
April 05, 2013: Obtain H7N9 Sequences (4 human-‐sourced; GISAID)
EpiMatrix Analysis: Iden<fica<on of H7N9 Class I and Class II Epitopes
101 H7N9 ICS* Class II Epitopes + 586 Class I Epitopes
April 06, 2013: H7N9 Vaccine: Two Constructs, Class I and Class II
Eliminate Epitopes highly conserved with Human Design vaccine: 12 hours (Logged).
Compare with previous epitopes (IEDB) And other H7N9 strains; create final list 20 hours (Logged).
Obtain all available H7N9 sequences
102 hOp://bit.ly/EpiPubs
Regulatory Agency approval
As Currently Proposed with Genome-‐derived Epitope-‐driven Influenza Vaccines (R21 / NIAID / NIH)
Gedng FastVax into the clinic: 4 Steps
1. In silico Design
2. Produc<on and Packaging
3. Clinical Trial
(correlates of immunity)
4. Deployment
Emergency use authoriza<on
104 hOp://bit.ly/EpiPubs
H7N9 at EpiVax
• String-‐of-‐epitopes DNA vaccine (Doug Lowrie) • String-‐of-‐epitopes Phage vaccine (Ft. Detrick) • Op<mized HA (fix epitopes) recombinant (TBD?)
• Op<mized HA + epitope string VLP (Ted Ross) • Collabora<on with NIID/Japan – in progress
EpiVax Contacts: Anthony Marcello, BDA, [email protected] Anne S. De Groot CEO/CSO [email protected]
105
DNA – chain of epitopes, or pep<de in liposomes ICS-‐op<mized proteins in VLP ICS-‐op<mized whole proteins
H7N9 Delivery vehicles
And . . . Cancer, Allergy and Autoimmune Disease?
107 107
• Cancer Vax = Epitopes + Adjuvant + ?
• Tregitope = Novel “adjuvant” that induces tolerance
• Allergy Vax = Epitopes +Tregitope+Delivery vehicle
• Autoimmunity Vax= AutoAg+Tregitope+Del. vehicle
• Payload+Adjuvant+ Delivery vehicle = Vaccine
• Why Computational Immunology • Tools to Produce IDVs
– Antigen selection – Vaccine design – New concepts
• Case Studies • . . . Questions?
108
Outline