key: cord-0333339-uotvipv8
authors: Rider, Paul J.F.; Dulin, Harrison; Uche, Ifeanyi K.; McGee, Michael C.; Breitenstein, Blake; Tan, Gene S.; Huang, Weishan; Kousoulas, Konstantin G.; Hai, Rong
title: A Herpes Simplex Virus Type-1-derived influenza vaccine induces balanced adaptive immune responses and protects mice from lethal influenza virus challenge
date: 2021-08-05
journal: bioRxiv
DOI: 10.1101/2021.08.05.455241
sha: 31f963063c9c1701c67e8f9ca82297b26dc91648
doc_id: 333339
cord_uid: uotvipv8

Influenza virus is a major respiratory viral pathogen responsible for the deaths of hundreds of thousands worldwide each year. Current vaccines provide protection primarily by inducing strain-specific antibody responses with the requirement of a match between vaccine strains and circulating strains. It has been suggested that anti-influenza T-cell responses, in addition to antibody responses may provide the broadest protection against different flu strains. Therefore, to address this urgent need, it is desirable to develop a vaccine candidate with an ability to induce balanced adaptive immunity including cell mediated immune responses. A live viral vector technology should exhibit safety, immunogenicity, effectiveness in the presence of pre-existing immunity, and the ability to induce mucosal immune responses. Here, we used VC2, an established Herpes Simplex Virus type 1 vaccine vector, to express the influenza HA protein. We show that this virus is capable of generating potent and specific anti-influenza humoral and cell-mediated immune responses. We further show that a single vaccination with the VC2-derived influenza vaccine protects mice from lethal challenge with influenza virus. Our data support the continued development of VC2-derived influenza vaccines for protection of human populations from both seasonal and pandemic strains of influenza. Finally, our results support the potential of VC2-derived vaccines as a platform for the rapid development of vaccines against emerging and established pathogens, particularly respiratory pathogens.

Current approaches to influenza vaccines have limited efficacy and leave human populations 71 susceptible to emerging influenza strains (1, 2). Indeed, since the development of the first 72 influenza vaccine there have been three pandemics responsible for more than 2 million deaths 73 worldwide and seasonal influenza infection is currently responsible for the deaths of 250,000-74 500,000 individuals per year worldwide (3) (4) (5) . In order to produce vaccine in time for the influenza 75 A season it is necessary to predict which influenza A strains will be circulating in the future (6). As 76 might be expected this process is imperfect and predicted strains do not always match the 77 circulating strains (6, 7). In addition to mismatch problems seasonal influenza vaccines exhibit 10-78 60% efficacy and this leaves human populations vulnerable to more severe illness and death 79 from influenza infection (8-10). Finally, current influenza vaccines leave human populations 80 susceptible to emerging influenza strains for which there is no vaccine, as in pandemic scenarios 81 (11). Therefore, a major focus of influenza vaccine research is to generate a vaccine that will 82 protect human populations from circulating as well as potentially emerging pandemic influenza 83 strains, a so-called "universal" influenza vaccine (12, 13) . 84

The incomplete protection provided by current influenza vaccines is due to the unique biology 86 of influenza virus as a segmented RNA virus: 1) RNA virus replication is error-prone and 87 mutations can occur that lead to the emergence of novel influenza A strains, and 2) influenza A 88 viruses can swap segments, which leads to the emergence of novel subtypes of virus (5, 14) . 89

Current influenza vaccination strategies induce strain specific neutralizing antibodies directed 90 against the influenza hemagglutinin (HA) protein (15, 16) . These antibodies inhibit the receptor 91 binding activity of HA and effectively neutralize virus entry preventing infection (17) (18) (19) (20) (21) (22) . As 92 mentioned above the virus can: 1) mutate the receptor binding site (drift) and 2) reassort to 93 include a novel HA segment (shift). This inherent ability to shift and drift results in the emergence 94 of influenza strains for which there is no protection in the human population and for which 95 vaccines do not exist (6). As such, current influenza vaccines require constantly updated vaccine 96 components to match viruses which are predicted to be circulating in a given season. 97 98 A major focus of influenza vaccine research is the generation of universal vaccines that can 99 protect against all possible strains of influenza. Critical to achieving the goal of a universal 100 vaccine is technology that is capable of eliciting anti-influenza T-cell responses. The goal of such 101 a vaccine is to generate antibodies and/or T-cell responses that are capable of preventing 102 infection by emergent strains of influenza virus regardless of shift or drift mutations (11, 23) . 103

Unlike antibody responses, which typically target HA with neutralizing antibodies that are strain 104 specific, T-cells are capable of targeting infected cells using epitopes that are essential, 105 conserved between different strains of a virus and not subject to shift or drift mutations. It has 106 been suggested that development of vaccines capable of inducing anti-influenza T-cells, 107 particularly at mucosal surfaces, is an important goal of efforts to generate more effective 108 influenza virus vaccines (9, 24 large genome allows the insertion of a number of antigenic transgenes, 2) molecular virology is 120 relatively well understood which informs rational attenuation strategies, 3) relative safety, and 4) 121 the ease of genetic manipulation. Importantly, in clinical and pre-clinical studies, herpesvirus 122 vectors, including HSV-1, have been shown to be unaffected by pre-existing immunity (28) (29) (30) (31) (32) . 123

For the current study we have used the herpesvirus vaccine candidate VC2 as the vector to 124 generate an influenza virus vaccine. VC2 contains mutations in the HSV-1 envelope proteins gK 125 and UL20 (33), which render the virus unable to enter into neurons via axonal termini (34-36). As 126 neurons are the site of HSV-1 latency, the mutation in VC2 preclude the establishment of a latent 127 infection. As the majority of clinical symptoms from HSV-1 infection occur after reactivation from 128 latency the inability of VC2 to establish latency greatly enhances its safety profile. In multiple 129 studies using a viariety of animal models, VC2 has been shown to be a safe, effective vaccine 130 against genital herpes in murine, guinea pig, and non-human primate models (33, (37) (38) (39) (40) (41) . 131

Notably, these studies described the development of significant, durable HSV-1 mucosal 132 immunity in mice that protected against ocular HSV-1 challenge (39). These studies also 133 described the development of anti-HSV-1 mucosal immunity in guinea pig and non-human 134 primate models (40, 41) . 135

In this study, we introduced influenza PR8 HA gene into VC2 and established an optimized 137 design to achieve the maximal expression of HA protein. We further characterized its 138 immunogenicity in mice, finding a robust, specific induction of anti-HA antibody and T-cell 139 responses. Importantly, T-cell responses exhibited both effector and memory phenotypes. We 140 report here that vaccination of mice with a single dose of our VC2-derived influenza vaccine 141 protects mice from lethal challenge with influenza virus. Our data supports the utility of VC2 as a 142 live-attenuated vaccine vector, which has the potential to be applied to the development of a 143 universal influenza vaccine as well as vaccines against significant emerging and re-emerging 144 pathogens, such as SARS-CoV-2. This BAC plasmid was used to construct VC2-HA. Briefly, the new VC2-HA plasmid were 160 constructed in Escherichia coli SW105 cells, using the two-step bacteriophage lambda Red-161 mediated recombination system, as described previously (15, 46) . The pCAGGS-PR8-HA 162 sequence of was amplified by PCR using primers P1 and P2. To construct PR8HA-Kan r , the 163 kanamycin resistance (Kanr) gene adjoining the I-SceI site was amplified by PCR from plasmid 164 pEPkan-S using primers P3 and P4, fused with PR8HA through fusion PCR. The PR8HA-Kanr 165 gene was amplified by PCR using primers P5 and P6 and then cloned into VC2 to replace gC. 166

The kanamycin resistance cassette was cleaved after expression of I-SceI from plasmid pBAD-I-167

SceI. The inserted PR8HA was verified by capillary DNA sequencing by using primers P7 to P12. well of a six-well dish of 80% confluent Vero cells was infected (MOI of 2) with the indicated 176 recombinant HSV viruses or mock infected with phosphate-buffered saline (PBS) for 1 h at 33°C. 177

At 24 h postinfection (hpi), cells were lysed in 1X protein loading buffer as described previously. 178

The reduced cell lysates were subjected to Western blot analysis by using monoclonal antibody performed on diluted serum samples as described earlier (16). Briefly, serum was obtained from 197 mice right before viral challenge and stored at -80°C. We coated 96-well ELISA plates (Immulon4; 198

Dynex, Chantilly, VA) with 50 µl (10 µg/ml) of the purified influenza A PR8 viruses. After being 199 washed with PBS, coated wells were blocked with PBS containing 1% BSA and then incubated 200 with diluted serum. After 1 h of incubation at room temperature, wells were rinsed with PBS and 201 incubated with a secondary anti-mouse IgG conjugated to peroxidase (Invitrogen, Carlsbad, CA). 202

Rinsed wells were incubated with 100 µl of SigmaFast OPD (ophenylenediamine dihydrochloride) 203 substrate (Sigma-Aldrich) for 30 min and stopped with 50 µl of 3M hydrochloric acid. The plates 204

were then read with a plate reader that measured the optical density at 490 nm (OD405; DTX880 205 multimode detector; Beckman Coulter). influenza virus (Fig. 1a) . To enhance the expression of HA protein, we replaced the entire gC 272 gene with PR8 HA and a pCAGGS promoter fused at its N terminus since pCAGGS promoter is a 273 strong mammalian Pol II promoter. This virus was designated VC2-HA. We predicted that this 274 promoter will drive higher HA expression compared to the original HSV gC promoter. The HA 275 insertion was verified by PCR-assisted sequencing of the insertion site. VC2-HA was able to 276 express PR8 HA in Vero cells, as indicated by the detection of HA with HA-specific antibody, 277 PY102 (Fig. 1b) . To ensure that insertion of HA did not affect virus replication we performed a 278 multi-step growth curve to compare replication of VC2-HA to the parental VC2 virus in Vero cells 279 at an MOI 0.01. Both VC2-HA and parental VC2 viruses exhibited a similar growth pattern (Fig.  280   1c) . 281 282 Immunization with VC2-pCA-HA protects mice from lethal challenge with PR8 virus 283

To evaluate the protection efficacy of recombinant VC2 expressing PR8 HA, we performed a 284 lethal challenge mouse study using the VC2-HA vaccine. Specifically, 6-8-week-old Black 6/J 285 mice (n=5) were vaccinated intramuscularly with VC2-HA at 1X10 5 , or 1X10 6 PFU or VC2 286 parental virus at 1X10 6 PFU per animal. Previous studies have shown that intramuscular route of 287 inoculation with VC2 is safe in murine, guinea pig and non-human primate models and elicits 288 immune responses to targeted antigens expressed in VC2 viruses (33, (40) (41) (42) . For these 289 experiments we used a single vaccination regimen. During the vaccination period there were no 290 significant weight differences among the groups (data not shown). Three weeks after 291 vaccination, mice were challenged intranasally with 1X10 3 PFU of influenza A PR8 H1N1 virus. 292

The two groups of mice vaccinated with different doses VC2-HA did not exhibit any weight loss or 293 clinical signs, such as fur ruffling and heavy breathing during the two weeks post challenge 294 To evaluate the immunogenicity of recombinant VC2-HA, we first investigated the humoral 302 responses induced by vaccination of mice with VC2-HA virus. 6-8-week Black/6J mice were 303 vaccinated with VC2-HA at 1X10 5 PFU, or VC2 parental virus at 1X10 6 PFU per mouse 304 intramuscularly since the intramuscular vaccination of VC2. Sera were collected from mice before 305 prime, pre-boost (after prime) and three weeks after the boost for in vitro serological assays (Fig.  306   3a) . Serum PR8 HA specific IgG titers were measured by ELISA using plates coated with purified 307 PR8 viruses. The endpoint titers of serum IgG were used as the readout (Fig. 3a) (Fig. 3b) . Collectively, the vaccination with VC2-316 HA elicited PR8 HA specific antibodies with robust neutralizing activity against PR8 virus. 317 318

There is increasing evidence that the induction of T cells is important for the broadly 320 protective properties of influenza vaccines (43). T cell responses to either HA or inactivated 321 influenza virus were measured using splenocytes isolated 9 days post boost from either VC2 or 322 VC2-HA vaccinated mice. Splenocytes were activated with either HSV-1 specific peptide as a 323 control, or inactivated PR8 virus particles. gB498-505 is an immunodominant HSV-1 epitope for 324 C57BL/6 mice (44). CD8+ T-cells from both VC2 and VC2-HA vaccinated mice were activated 325 with gB498-505 (Fig. 4a) . In contrast, only mice vaccinated with VC2-HA responded with IFN-g and 326

TNF-a expression after stimulation with heat-inactivated PR8 (HI PR8) virus (Fig. 4a) and 327 expanded upon re-stimulation by HA protein (Fig. 4b) . Furthermore, we observed an increase in 328 CD4+ and CD8+ T cells with an effector/memory phenotype (CD44high CD62L-) in VC2-HA 329 vaccinated mice, as compared to VC2 vaccinated controls (Fig. 4b) . The utility of viruses as vectors for vaccines requires that they are safe, immunogenic, and 341 capable of inducing protection in the presence of pre-existing immunity. Our novel vector, VC2 342 possesses mutations that abrogate infection of neurons via neuronal axons (34). This is a 343 rational design that disrupts the ability of the virus to establish latency. The inability to establish 344 latency is expected to preclude the majority of clinical symptoms associated with HSV-1 infection. 345 VC2 has been shown to be safe and immunogenic in murine (including SCID mice), guinea pig 346 and non-human primate models (33, (38) (39) (40) (41) . Our data here show that VC2-HA promotes strong 347 humoral and cell mediated immune responses and protects mice from lethal influenza virus 348 challenge. Consistent with this, we have previously shown that VC2 can be used to create a 349 safe, immunogenic vaccine against equine herpesvirus 1 (42). 350

All viral vector-mediated therapeutic or intervention strategies can be potentially affected by 352 pre-existing immunity in the target population. This is a concern for HSV-1-derived vectors due to 353 their high prevalence in human populations. However, it is known that individuals are capable of 354 being re-infected by differing strains of HSV throughout their lifetime (48). Further, a hallmark of 355 herpesvirus infection is their ability to reactivate and spread in the presence of sizable host anti-356 herpesvirus immune responses (49). These characteristics of the natural history of herpesvirus 357 infection, as well as a number of studies demonstrating that pre-existing immunity had no 358 substantial effect on vector efficacy (38-42) inform the usage of herpesvirus-derived vectors for 359 use in anti-cancer and vaccine applications. Specifically, clinical trial data from T-Vec™, the FDA 360 approved HSV-1-derived oncolytic virotherapy, found no difference in treatment efficacy between 361 seropositive and seronegative individuals. Consistent with this, we and others have found no 362 effect of HSV-1 seropositivity on the efficacy of oncolytic virotherapy in mice (29, 49, 50) . While 363 the current study was not performed on HSV-1 seropositive mice our data shows we are able to 364 get increased anti-HA immune responses after a prime and boost compared to a single 365 vaccination. As we were unable to detect HA on virion particles of VC2-HA (data not shown), this 366 suggests VC2-HA boost virus was able to infect and express HA in seropositive mice. 367

There are a number of groups currently working on HSV-vectored vaccines against a number 369 of human and animal pathogens. Multiple approaches ranging from HSV based amplicons, 370 replication defective, attenuated or conditionally defective vectors have been considered (45) (46) (47) (48) . 371

Our approach to altering the pathogenic potential of HSV-1 is unique in that it allows the virus to 372 replicate essentially as a wild type virus without the potential for causing disease as seen in other, 373 even attenuated viral vectors. This wild type like replication is important as it likely allows for the 374 maximum potential immunogenicity of the vectored antigen. Furthermore, we have shown that 375 VC2 possesses the gK31-68 mutation that forces the virus to enter only via endocytosis instead 376 of fusion of the viral envelope with cellular plasma membranes (51). This altered mode of entry 377 prevents the virus from entering into neuronal axons as well as potentially altering innate immune 378 responses and downstream adoptive immune responses as evidenced by the fact the the VC2 379 intramuscular immunization was more effective in conferring against lethal ocular challenge with 380 the human clinical strain HSV-1 (McKrae) (39). In this regard, VC2 appears to possess an 381 "adjuvant" effect that can facilitate robust immune responses against heterologous antigens 382 expressed via the VC2 vector. 383 384 Current influenza vaccines induce neutralizing antibodies as their primary mechanism of 385 protection. These vaccines induce strain-specific immunity and as a result leave human 386 populations at risk for infection with other influenza strains. It has been suggested that a T-cell 387 response may be necessary to induce broad protection against multiple strains of influenza (9). 388

Our data demonstrate that our vaccine is capable of producing both cellular and humoral anti-389 influenza immune responses. With our approach, in addition to neutralizing antibodies, we 390 achieved significant induction of anti-influenza CD4+ and CD8+ T-cells. In this study, we saw 391 greater induction of anti-influenza CD4+ T-cells than CD8+ T-cells. We believe that this is likely 392 due to the use of inactivated virus and recombinant HA protein to stimulate splenocytes in our 393 experiments. It is possible that HA peptide libraries would be more effective at identifying HA 394 specific CD8+ immune responses. Regardless, our data demonstrate that VC2-HA is capable of 395 inducing significant influenza specific T-cell responses. In future experiments it will be critical to 396 test the ability of our vaccine to induce broad protection against multiple strains of influenza. 

The authors declare that an intellectual property application for VC2-HA has been filed on which 421 R.H., P.J.F.R. and K.G.K are the named inventors. K.G.K. has intellectual property rights to the 422 Generates Anti-EHV-1 Immune Responses in Mice. J Virol 91. 542 intramuscularly with either 1X10 5 , or 1X10 6 PFU VC2-HA. As a negative control, mice were 590 vaccinated intramuscularly with 1X10 6 PFU of VC2. For these experiments we used a single 591 vaccination. Three weeks post vaccination mice were challenged intranasally with 1X10 3 PFU of 592 influenza A PR8 H1N1 virus. Mice were observed for weight loss (b.) and survival (c.). p < *0.05, 593 **0.01, ***0.001, NS= not significant, by unpaired student's t test. 594 595 Figure 3 . VC2-HA vaccination induces HA specific antibody responses. 6-8-week C57BL/6 596 J mice were vaccinated intramuscularly with VC2-HA at either 1X10 5 PFU or 1X10 6 PFU, or VC2 597 vector alone at 1X10 6 PFU. For these experiments we used a prime and a boost immunization 598 after three weeks. Sera were collected from mice before prime, pre-boost (after prime) and three 599 weeks after the boost. a. HA-specific IgG titers in sera were measured by ELISA using plates 600 coated with purified PR8 viruses (Fig. 2a) 

(-) gB PR8 (-) gB PR8

Efficacy and effectiveness of influenza vaccination

The efficacy and cost 429 effectiveness of vaccination against influenza among elderly persons living in the community

Estimates 435 of global seasonal influenza-associated respiratory mortality: a modelling study

The annual impact of seasonal influenza in the US: measuring disease 439 burden and costs

Orthomyxoviridae: the viruses and their replication

Overcoming Barriers in the Path to a Universal Influenza Virus 443 Vaccine

Current and future influenza vaccines

Mismatch between 446 the 1997/1998 influenza vaccine and the major epidemic A(H3N2) virus strain as the cause of 447 an inadequate vaccine-induced antibody response to this strain in the elderly

A Universal Influenza Vaccine: The Strategic Plan for the National Institute of 451 Allergy and Infectious Diseases

Comparing influenza vaccine efficacy against mismatched and matched strains: a 454 systematic review and meta-analysis

Development of a Universal Influenza Vaccine

Chasing Seasonal Influenza -The Need for a 458 Universal Influenza Vaccine

460 Influenza viruses expressing chimeric hemagglutinins: globular head and stalk domains 461 derived from different subtypes

Breadth of Antibody Responses during Influenza Virus Infection 465 and Vaccination

Influenza B 467 virus NS1-truncated mutants: live-attenuated vaccine approach

Is It Possible to Develop a "Universal" Influenza Virus Vaccine? Potential for 469 a Universal Influenza Vaccine

The structure and receptor binding properties of the 1918 472 influenza hemagglutinin

Complete structure of the hemagglutinin gene from the human 475 influenza A/Victoria/3/75 (H3N2) strain as determined from cloned DNA

Structure and 477 receptor specificity of the hemagglutinin from an H5N1 influenza virus

Structure of the 479 uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus

Structural basis for receptor specificity of influenza B virus 482 hemagglutinin

Front Immunol 11:513. 485 24. Iwasaki A, Omer SB. 2020. Why and How Vaccines Work

Sci Immunol 2. 488 26. van Riel D, de Wit E. 2020. Next-generation vaccine platforms for COVID-19

Talimogene laherparepvec: First in class 491 oncolytic virotherapy

Effect of preexisting anti-herpes immunity on the efficacy of herpes simplex viral 494 therapy in a murine intraperitoneal tumor model

ICP34.5 deleted herpes simplex virus with enhanced 497 oncolytic, immune stimulating, and anti-tumour properties

Effect of prior exposure to herpes 499 simplex virus 1 on viral vector-mediated tumor therapy in immunocompetent mice

Herpes simplex virus vectors elicit durable immune 502 responses in the presence of preexisting host immunity

Programming Multifaceted Pulmonary T Cell Immunity 544 by Combination Adjuvants

Defining the herpes simplex virus-546 specific CD8+ T cell repertoire in C57BL/6 mice

Humanized Mice Redirects T Cell Responses to Gag and Reduces Acute HIV-1 Viremia

Controlled Herpesvirus Vectors

Evaluation of the immunogenicity of a recombinant 555 HSV-1 vector expressing human group C rotavirus VP6 protein

An attenuated herpes simplex virus type 1 558 (HSV1) encoding the HIV-1 Tat protein protects mice from a deadly mucosal HSV1 challenge

Novel Oncolytic Herpes Simplex Virus 1 VC2 Promotes Long-Lasting

Systemic Anti-melanoma Tumor Immune Responses and Increased Survival in an 563 Immunocompetent B16F10-Derived Mouse Melanoma Model

CD8(+) T-565 cell Immune Evasion Enables Oncolytic Virus Immunotherapy

The Amino 567 Terminus of Herpes Simplex Virus 1 Glycoprotein K (gK) Is Required for gB Binding to Akt, 568 Release of Intracellular Calcium, and Fusion of the Viral Envelope with Plasma Membranes

Figure 4. VC2-HA vaccination induces both CD4+ and CD8+ effector/memory T-cells

Splenocytes from vaccinated mice were analyzed 9 days post boost

peptide 606 or heat-inactivated PR8 virus for 5 hours in vitro. b. Representative FACS plots and summary of 607 percentages of naïve and effector/memory CD4+ and CD8+ T cells. Summary of percentages of 608 expanded CD4+ and CD8+ T cells post re-stimulation