Adaptive Immunity:Understanding cross-reactive responses

Lorena E. BrownThe University of Melbourne

Australia

Is vaccination of everyone from birth onwards with current inactivated vaccines desirable?

Would a population devoid of infection-induced cross-protective immunity be worse off in situations of:

• Vaccine mismatch/shortage• Emergence of a new subtype• Elderly experiencing OAS or waning B cell immunity• Lack of vaccine efficacy in >30% population

Modelling of the evolution of influenza virus suggests short-lived crossreactive immunity is essential to explain the linear nature of antigenic drift.

• Without this would we start to see more strain diversity and corresponding difficulty in infection control?

What adaptive immune mechanisms are cross-subtype reactive?

• CD8+ cytotoxic T cells (CTL)– kill cells infected with all type A viruses– can recognise peptides from internal proteins– thought to speed recovery – if present in high numbers at site of infection as

memory cells may lead to subclinical outcome

• Non-neutralising antibodies that bind to viral antigens on infected cells– Lyse infected cells by Ab +C’ or ADCC– Best studied is M2 ectodomain

What adaptive immune mechanisms are cross subtype reactive?

Stem domain:– Mabs from human B cells (Throsby et

al. PLoS ONE 3(12): e3942; Ekiert et al. Science. 2009 324:246; Corti et al. J. Clin

Invest 2010 120:1663)– Induced by DNA/ split vaccine

prime boost (Wei et al., Science 2010 329, 1060)

Site B epitope– Yoshida et al. PLoS Pathog. 2009 5(3):

e1000350

10.1126/science.1195116

• Antibodies to conserved regions of HA

Site B

Investigating cross-protective mechanisms against H5N1 in the ferret model

Days

3-5 mth old male/femaleseronegative

Challenge 106 EID50 wt H5N1

A/Vietnam/1203/2004

VaccinergA/Vietnam/1194/2004 or

other split virus formulations i.m.

• serology (HI, micro virus neut.)• weight• temperature (rectal, transponder)• clinical symptoms • activity score• post mortem tissues for histology

• virus isolation- rectal swab- nasal wash- oral swab- organs post mortem

Days 3, 5 and 7 and 14 or at

humane endpoint

(systemic and pulmonary disease)

0 3 wk 7 wk

H5N1 inactivated split vaccines are fully protective when formulated with adjuvant

Alert and playfulAlert, play only when inducedAlert but not playfulNeither alert nor playfulCulled at humane endpoint

*IMX = ISCOMATRIXTM

Middleton et al. J. Virol 2009 83:7770

Activity scoreD

ay

po

st

ch

all

en

ge

165 166 167 168 177 178 179 180161 162 163 164

1

3

567

14

4

2

PBSFluvaxTM + AlPO4

FluvaxTM + IMX*

FluvaxTM

157 158 159 160

* IMX = ISCOMATRIXTM adjuvant

Seasonal influenza vaccine can protect against H5N1

173 174 175 176 169 170 171

H3N2 + IMX

H1N1 + IMX

The H1N1 component is inducing the cross-protection

Despite any crossreactive immunity measured by HI or VN assays

Vaccine Pre-challenge HIGMT fraction respond

Pre-challenge VNGMT fraction respond

H5N1 + AlPO4 45 4/4 45 4/4

H5N1 + IMX 45 4/4 128 4/4

Fluvax + AlPO4 <4 0/4 <4 0/4

Fluvax + IMX <4 0/4 <4 0/4

* IMX = ISCOMATRIXTM adjuvant

Crossreactive responses potentially against N1 BUTreassortant H3N1 vaccine shows no protection against H5N1

0

0.5

1

1.5

2

2.5

1 2 4 8 16 32 64 128

Dilution

NA

acti

vit

y

H3N1H1N1H3N2rNA

However, to maintain the HA/NA activity balance the H3N1 virus: • has 8-fold less enzyme activity (fetuin cleavage assay)• contains 8-fold less NA (Western blot)

Da

y p

os

t c

ha

lle

ng

e

1

3

567

14

4

255 256 257 258

H1N1 + IMX

251 252 253 254

H3N1 + IMX

259 260 261 262

H3N2+ IMX

PBS

271 272 273 274

Activity score

Ferret number

There is an imperfect correlation between neuraminidase inhibition activity of ferret serum and protection

PBS H1N1 H3N1 H3N2 rNA rNP

Vaccines formulated with ISCOMATRIXTM adjuvant

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

= survivors

NI

acti

vity

of

pre

-ch

alle

ng

e fe

rret

ser

a a

gai

nst

H5N

1

Conclusions from ferret data:

• Seasonal influenza vaccine can provide some protection against H5N1, particularly when formulated with an adjuvant.

• The H1N1 component provides the cross protective immunity

• This is not reflected by HI or VN assays but imperfectly by NI assays, so the N1s share epitopes for Ab and we should not discount their role in crossprotection.

• An H3N1 virus was non-protective, possibly due to low N1 levels, nor was H3N2 virus so these crossreactive responses are NOT HETEROSUBTYPIC

• To induce true heterosubtypic responses we can prime CD8+ T cells (CTL)

• These are not induced efficiently by inactivated split virus vaccines

• A different type of vaccine that delivers antigen to the cytoplasm of dendritic cells is required

• Can’t adequately study these in ferrets

Testing benefits of adding CTL-inducing component to split virus vaccine for

seasonal influenza

Suboptimal dose of seasonal split vaccine• mimics people responding poorly to vaccine• mimics situation of vaccine mismatch

CD4+ T cell epitope

CD8+ T cell epitopeK

Pam2Cys

Suboptimal dose of lipopeptide• will induce influenza-specific CD4+ and CD8+ T cells• will not prevent infection but aid clearance

TLR2

immature DC mature DC

lipopeptide

costimulatory molecule

• lipopeptide binds to TLR2 on DC surface via Pam2Cys

• lipopeptide enters cells by TLR2-mediated endocytosis so Ag can enter the class II processing pathway

• Ag can also escape the endosome and so can enter the class I processing pathway

• DC maturation is induced by TLR signalling; costim. and MHC II upregulated

Help for:•Antibody production•priming of long lived memory CD8+ T cells

T

MHC molecules

CD8+T cells

The lipopeptide used here do not have any B cell epitopes

Lipopeptides

CD4+T cells

T

Kill virus-infected cells

10ug HA intranasally

0

1

2

3

4

5

0

1

2

3

4

5

6

10ug HA subcutaneously

0

1

2

3

4

5

0

1

2

3

4

5

6

Mixtures of suboptimal split vaccine and lipopeptide induce improved viral clearance

Mice given a single dose of vaccine

3 weeks later mice are bled then challenged with A/Memphis/1.71 virus

5 days later lung virus titres are determined

10ug HA intranasally

0

1

2

3

4

5

0

1

2

3

4

5

6

10ug HA subcutaneously

0

1

2

3

4

5

0

1

2

3

4

5

6

Mixtures of suboptimal split vaccine and lipopeptide induce improved viral clearance

Slight enhancement of Ab levels; sub cut combination vac. is approaching live virus levels

0.3ug HA intranasally

0

1

2

3

4

5

0

1

2

3

4

5

6

0.3ug HA subcutaneously

0

1

2

3

4

5

0

1

2

3

4

5

6

With even lower doses of split vaccine the improvement is best seen when vaccine is given by the i.n. route

Though slightly enhanced, Ab levels are still low – other mechanisms may need to come into play

0

5000

10000

15000

20000intranasal subcutaneous

Large numbers of activated CD8+ T cells are present in the lungs of intranasally-primed mice 5 days after

challenge with influenza

influenza-specific IFN-producing CD8+ T cells in the lungs measured by intracellular cytokine staining 5 days after infection

CTLs expanded by infected cells as not enough antibody to remove virusthese are as numerous as when induced by prior virus infection

note that 5 days is before any T cells are present as a result of the challenge virus

Conclusions from lipopeptide/split virus data:• Suboptimal doses (10µg and 0.3µg) of split virus leads to poor antibody

production and viral clearance

• Viral clearance can be improved by the addition of a suboptimal dose of lipopeptide that induces T cell responses

• Extra “help” from CD4+ T cells plus increased DC activation during priming may provide a greater quantity or quality of Ab to enhance the 10 µg dose of split virus.

• When antibody is insufficient to control infection rapidly, as with the 0.3 µg dose of split virus, CTL-memory cells in the lungs can be activated and expanded to aid clearance

• Triggering of CTL memory cells already present in the lung (i.n. delivery) may be more efficient than relying on trafficking of memory cells from other organs (sub. cut. delivery)

Different arms of the adaptive immune response are acting in synergy in these responses to the combination vaccines

Conclusion• Co-induction of T cell and antibody responses by influenza

vaccines may provide better protection against disease when seroconvertion has not been adequate due to: – vaccine mismatch, – waning B cell responsiveness, – original antigenic sin, – individuals not recently vaccinated – for emergence of a new subtype.

• Only through a greater understanding of the mechanisms of cross-protection and their induction can we hope to create vaccines that provide us with the much needed “safety net” of heterosubtypic immunity.

University of Melbourne• David Jackson• Joanna Cobbin • Weiguang Zeng

WHO Collaborating Centre for Influenza Reference and Research

• Ian Barr

Main contributors

CSL Limited• Steve Rockman• Martin Pearse

CSIRO Australian Animal Health Laboratory

• Deborah Middleton

Urgent Research into a Potential Avian Influenza-Induced Pandemic Grant Scheme