Broadly Reactive H2 Hemagglutinin Vaccines Elicit Cross-Reactive Antibodies in Ferrets Pre-Immune to Seasonal Influenza A Viruses

Influenza vaccines have traditionally been tested in naïve mice and ferrets. However, humans are first exposed to influenza viruses within the first few years of their lives. Therefore, there is a pressing need to test influenza virus vaccines in animal models that have been previously exposed to influenza viruses before being vaccinated. In this manuscript, previously described H2 computationally optimized broadly reactive antigen (COBRA) HA vaccines (Z1, Z5) were tested in influenza virus ‘pre-immune’ ferret models. Ferrets were infected with historical, seasonal influenza viruses to establish pre-immunity. These pre-immune ferrets were then vaccinated with either COBRA H2 HA recombinant proteins or WT H2 HA recombinant proteins in a prime-boost regimen. A set of naïve pre-immune or non pre-immune ferrets were also vaccinated to control of the effects of the multiple different pre-immunities. All of the ferrets were then challenged with a swine H2N3 influenza virus. Ferrets with pre-existing immune responses influenced recombinant H2 HA elicited antibodies following vaccination as measured by HAI and classical neutralization assays. Having both H3N2 and H1N1 immunological memory regardless of the order of exposure significantly decreased viral nasal wash titers and completely protected all ferrets from both morbidity and mortality, including the mock vaccinated ferrets in the group. While the vast majority of the pre-immune ferrets were protected from both morbidity and mortality across all of the different pre-immunities, the Z1 COBRA HA vaccinated ferrets had significantly higher antibody titers and recognized the highest number H2 influenza viruses in a classical neutralization assay compared to the other H2 HA vaccines. Importance H1N1 and H3N2 influenza viruses have co-circulated in the human population since 1977. Nearly every human alive today has antibodies and memory B and T cells against these two subtypes of influenza viruses. H2N2 influenza viruses caused the 1957 global pandemic and people born after 1968 have never been exposed to H2 influenza viruses. It is quite likely that a future H2 influenza virus could transmit within the human population and start a new global pandemic, since the majority of people alive today are immunologically naïve to viruses of this subtype. Therefore, an effective vaccine for H2 influenza viruses should be tested in an animal model with previous exposure to influenza viruses that have circulated in humans. Ferrets were infected with historical influenza A viruses to more accurately mimic the immune responses in people who have pre-existing immune responses to seasonal influenza viruses. In this study, pre-immune ferrets were vaccinated with WT and COBRA H2 recombinant HA proteins in order to examine the effects of pre-existing immunity to seasonal human influenza viruses have on the elicitation of broadly cross-reactive antibodies from heterologous vaccination.


276
Waltham, MA, USA). The MDCK cells were then washed with distilled water to remove the 277 crystal violet. Any well having an HA activity of 1:2 was defined as positive for the analysis.

278
HA activity was confirmed by >10% of CPE in wells that was positive for HA activity.

303
Vaccination of ferrets with pre-existing influenza virus immunity

304
Fitch ferrets (n=20) were made pre-immune with one of three influenza virus subtypes.

305
The

336
The non pre-immune ferrets did not have any influenza infection prior to vaccination.

372
In

383
None of the ferrets in the H3N2-H1N1 pre-immune or the H1N1-H3N2 pre-immune groups had 384 any detectable viral titers in their nasal washes on any day (Fig. S7). Of all the time points and 385 pre-immunity background there was no significant difference between the vaccinated groups 386 and the mock groups, except for non pre-immune D1 NW, where Z1 was significantly lower than 387 the mock group with p-value of 0.0078, with a one-way ANOVA + Tukey.

388
A three-way ANOVA looking at the main effects of vaccine received, the ferret pre-389 immunity, and the day of the nasal wash, indicated that overall when adjusting for pre-immunity 390 and day post infection that the mean viral nasal wash titers of the Z1 COBRA was significantly 391 lower than that of the mock vaccinated group by 0.322 log 10 viral titer (p adjusted < 0.001).

392
Furthermore, the Z1 COBRA also had a titer 0.219 log 10 lower after adjustment compared to 393 the Z5 COBRA (p adjusted = 0.039) (Fig S2). Only the non pre-immune ferrets were 394 significantly different from the other pre-immunities after controlling for vaccine received and day 395 post infection ( Fig S2). All other pre-immunities have nonsignificant mean viral titers. When 396 comparing the day post infection, Day 1 and day 3 were not significantly different, but day 5 had The HAI titers varied greatly between the pre-immune groups. The H2N2 pre-immune 401 ferrets were the only pre-immune group to have HAI titers to VLPs in the H2 panel on the day of 402 prime vaccination (Fig. 4). Ferrets in all five of the vaccination groups had average HAI titers 403 above 1:40 to multiple VLPs in the H2 panel, but none of the vaccination groups had detectable 404 titers to all 12 of the H2 HA VLPs (Fig. 4).

405
After the first vaccination, the H2N2 pre-immune ferrets had a geometric mean HAI titer 406 of 1:40 to all twelve of the VLPs in the panel excluding the mock vaccination group (Fig. 5A-E).

407
The H3N2 pre-immune ferrets had low HAI titers to the H2 VLPs after the first vaccination ( Fig.   408 5F-J). Only the Z1 vaccinated ferrets had a geometric mean HAI titer of 1:40 to more than one 409 H2 VLP (Fig. 5H). The mock vaccinated ferrets in H1N1 pre-immune group did not have

418
After the second vaccination, the HAI titers for the ferrets in the H2N2 pre-immune group 419 did not drastically change (Fig. 6A-E). In the H3N2 pre-immune group, HAI titers in the

426
In the H1N1 pre-immune ferrets, the HAI titers of all of the vaccination groups besides 427 the mock vaccination group increased after the second vaccination ( Fig. 6K-O

434
In the non pre-immune group, the HAI titers in the Mal/NL/01, Mal/WI/08, Z1 and Z5 435 vaccination groups all increased after the second vaccination (Fig. 6P-T)    but the titers were 2-4 fold lower on average than Group 1 imprinted ferrets. The Group 2 523 imprinted ferrets likely have B-cells that are highly specific to epitopes on the H3 HA (35).

524
These epitopes are far less similar to the epitopes on H2 HA than the epitopes on H1 HA.

525
Without any similar immune memory, the H3N2 influenza virus imprinted ferrets would likely

564
The Z1 COBRA HA also outperformed the Z5 COBRA and the two wild-type vaccines.

585
The author TMR has a patent on the COBRA methodology.  Table 1: Ferret Survival Percentage

593
The survival rate among ferrets in each of the vaccination groups across all of the pre-594 immunities. The columns correspond to the pre-immunity, the vaccine and the survival 595 percentage. Each row is a vaccination group with its pre-immunity followed by the survival 596 percentage following viral challenge. Ferret groups either contained n=3 or n=4 ferrets.

599
Neutralization titers were obtained from pooled sera. Titers were obtained by taking the 600 geometric mean titer of the replicates for each of the vaccine groups. Each column is a virus 601 that was used in the neutralization assay. Each row is the sera from a vaccination group. The

711
Significance between groups are analyzed in Supplemental Table 1.