Influenza antibody profiling by protein microarray
Marion Koopmans, Professor of public health virology Head Viroscience department
Influenza serology applications
1. Evaluation of vaccine response (functional antibodies)
Hemagglutination inhibition, microneutralisation
2. Evaluation of influenza infection at population level (any antibodies)
HI when specific viruses are known
More difficult when studying human animal interface and unknown exposures
3. Emerging disease preparedness and ability to obtain serological data at BSL2 level (both)
Level of immunity?
Rate of asymptomatic infections Spread of infection in households
Effect of control measures on spread of infection
Which antibodies?
Many antibodies are produced upon
exposure
HA, NA, M, NP
Functional antibodies:
Binding blocking (HI, MNT) Targeting surface proteins
Most specific, and considered dominant in seasonal influenza
Insights from pandemic H1:
Broadly reacting neutralising stem Ab Broadly reactive antibodies HA1 (ADCC?)
Challenges for sero-epidemiology studies
Current reference assays are difficult to standardise Require knowledge of virus persons are exposed to (which may be lacking in surveillance at human animal interface)
Single mutations can affect ability to detect HI antibodies
Need for testing to panels of viruses > costly, labor intensive and requiring a lot of serum
> development of a more standardized method to assess exposure for population-based studies
H2.57 A/Canada/720/05 H3.03 A/Wyoming/3/03 H3.07 A/Brisbane/10/2007 H5.04 A/Vietnam/1194/2004 H7.03 A/Chicken/Netherlands/1/03 H9.99 A/Guinea fowl/Hong Kong/WF10/99 A/South Caroline/1/18 H1.18 A/WS/33 H1.33 A/Puerto Rico/8/34 H1.34 A/New Caledonia/20/99 H1.99 A/Brisbane/59/2007 H1.07 A/California/6/2009 H1.09
HA1 proteins expressed in HEK293 cells, HIS tag purified 16 or 64 wells / slide
Quality control serum for slide-slide comparison and normalisation
Quantitative signal (fluorescence)
Calculation of titre from dilution curves
117b Verdunning (log-scale) si gna l 1: 10 1: 20 1: 40 1: 80 1: 160 1: 320 1: 640 1: 1280 1: 2560 1: 5120 0 10000 20000 30000 40000 50000 60000 70000 H1.18 H1.33 H1.34 H1.99 H1.07 H1.09i H2.05 H3.03 H3.07 H5.04 H7.03 H9.99
Rabbit serum 1918 H1 1999 H1 1957 H2 2003 H3 2004 H5 2003 H7 1999 H9 anti-H1 1999 62 231 20 20 20 20 20 anti-H2 1957 20 20 186 20 39 20 20 anti-H3 2005 20 20 20 245 20 20 20 anti-H5 2005 20 20 20 20 176 20 20 anti-H7 2003 20 20 20 20 20 1327 175 anti-H9 2003 20 20 20 20 20 20 1603 anti-H1 2009* 552 139 20 20 20 20 20
rabbit immunized with HA1 recombinant proteins
Validation for use in humans
Problem: “ true negative” human sera do not exist
Proxy is supernatants from differentiated plasma cells, and serum samples from children with primary infection
PBMC culture to stimulate plasma-cell differentiation from individual specific memory B cells
Testing of individual wells (human monoclonals) by proteinarray
Age stratified Ab profile prepandemic population
sample
(n = 122, 2008)HI and PA profiles pandemic cases (n=18)
HA1 A(H1N1)pdm09
responses in neonatal
screening samples
Rationale
• Convenience sample
• Standardized sampling material (blood spots)
• Profile of antibody responses based on women of childbearing age (maternal Ab)
Validation
• Bloodspots collected before April 2009 used to set cutoff (30375 fluorescence)
Standardization
• Single batch of antigens for total study
• Corrected for test-to-test variation to the mean value of positive control serum (09/194, NIBSC), for each slide separately
13 countries 5 continents
Other seroprevalence studies P er c en ta g e po s it iv e/ w ee k ( x 10 0)
A(H1N1)pdm09 epidemic period
Start vaccination
Phase 5 alert
Phase 6 alert
Smoothed blood spot results with 95% CI intervals; Smoothing by generalized additive model (GAM)
Woods et al. 2003, 2004, 2006, 2011
Europe, Lebanon, South Africa
P er c en ta g e po s it iv e/ w ee k ( x 10 0)What other information can be provided from
the profiles of antibodies?
Is it possible to distinguish Ab profiles within subtype? What is the effect of
vaccination?
Are the Ab measured for H9 Low level cross reactive antibodies or true
Patterns of (broad/cross reactivity) are very
diverse, likely related to an individual’s exposure
history
Patterns of (broad/cross reactivity) are very
diverse, likely related to an individual’s exposure
history
Patterns of (broad/cross reactivity) are very
diverse, likely related to an individual’s exposure
history
Patterns of (broad/cross reactivity) are very
diverse, likely related to an individual’s exposure
history
Huijskens et al., 2013
Potential applications: comparative analysis of vaccination and infection response
Potential applications: avian influenza surveillance Guangdong
H9 H7
Since 2012: enhanced surveillance
Market environmental sampling and testing for AI by PCR
Periodic serological sampling of poultry workers and blood donors > Analysis of seroresponses in relation to virus exposure
Summary
HA1 based profiling of immune response has potential for population studies of virus exposure in humans and animals
Comparative evaluation against biological assays needed to fully leverage the potential (some studies planned in Flucop project, IMI)
Acknowledgements
ErasmusMC Erwin de Bruin
Gudrun Freidl
Guus Rimmelzwaan
Hendrik Jan van Ham
Guandong CDC Ke Changwen Xin Zhang OUCRU Vietnam Maciej Boni Stacey Todd
Nhat Nguyen Thi Duy