Recombinant modified vaccinia virus Ankara efficiently restimulates human cytotoxic T lymphocytes in vitro

Abstract

The immunogenicity of recombinant modified vaccinia Ankara, a highly attenuated vaccinia virus, expressing influenza nucleoprotein (MVA-NP) and HIV-1 gag (MVA-gag) was investigated. Restimulation of peripheral blood lymphocytes of healthy subjects with MVA-NP led to expansion of CTL with specificity for known NP epitopes. These CTL efficiently lysed NP peptide-pulsed targets and released interferon-γ (IFN-γ) on contact with epitope peptide. MVA-NP-stimulated CTL specific for the HLA-B8 epitope, NP380-88, stained with a tetrameric complex of HLA-B8 refolded with the NP380-88 peptide and anti-CD8 antibody on flow cytometry. CTL were also elicited from two HIV-1 seropositive donors by restimulation with MVA-HIV-1 gag and showed specificity for immunodominant gag epitopes. These data indicate that restimulation of human CTL with recombinant MVA is effective and suggest that MVA will elicit CTL responses in humans in vivo.

Introduction

Recombinant vaccinia viruses are being considered as potential vaccines against a variety of infectious diseases and tumours. Following the success of vaccinia virus as a smallpox vaccine, recombinant vaccinia constructs expressing foreign proteins have been investigated for their immunogenicity and ability to protect against disease in several animal models [1]. The urgent need for a vaccine against HIV led to the evaluation of a non-host range-restricted vaccinia against HIV envelope glycoprotein as a vaccine candidate [2]. Although it was shown to be safe and immunogenic in humans [3], adverse effects including the increased risk of vaccinia replication in immunocompromised individuals are major limiting factors.

The attenuated vaccinia strain, modified vaccinia virus Ankara (MVA), is an attractive alternative to conventional strains. MVA was derived by serial passage in chick embryo fibroblasts, causing deletion of 15% of the parental genome. This has resulted in deletion or inactivation of many non-essential genes. In particular, it has lost host range functions with consequent limited replicative capacity in mammalian cells, and it has lost several functional cytokine receptors which may contribute to virulence and immune evasion [4]. It was used at the end of the smallpox vaccination campaign and was found to be safe in 120,000 healthy and high risk individuals, with no recorded dissemination or post-vaccinial encephalitis [5].

Recombinant MVA has been tested in several animal models of disease. In a murine influenza model, MVA expressing haemagglutinin (HA) and nucleoprotein (NP), when administered via a variety of routes induced cellular and humoral immunity and protected mice against a lethal respiratory challenge as effectively as a recombinant wild type vaccinia virus strain (WR) [6]. More recently, it was shown that mice given a series of vaccinations comprising a DNA prime-MVA boost expressing a Plasmodium circumsporozoite protein were completely protected against a sporozoite challenge [7]. Macaques which received intramuscular MVA expressing simian immunodeficiency virus (SIV) proteins gag, pol and env, had a reduced viral load following a homologous SIV challenge and disease-free survival was increased relative to animals vaccinated with replication competent vaccinia recombinants, although SIV replication was detected in all animals in subsequent reports [8], [9]. The DNA prime-MVA boost regimen was used to vaccinate macaques with a multi-epitope gene which contained sequences encoding SIV CTL epitopes [10]. Although two out of three monkeys were not protected against a pathogenic SIV challenge, this vaccine regimen was able to elicit strong SIV-specific CTL responses, which are likely to be an important component of protective immunity.

As yet, no recombinant MVA vaccines have been tested in humans though such vaccines are being developed for HIV-1 and malaria. We have therefore used MVA constructs expressing influenza virus and HIV proteins to determine whether they can elicit human antigen-specific CTL in vitro. Conventionally, influenza-specific CTL have been generated in vitro either by infection of peripheral blood mononuclear cells with influenza virus or by stimulation with synthetic peptides based on known epitopes. The former method is relatively inefficient and the latter requires prior knowledge of the epitopes of interest. Recombinant non-attenuated vaccinia viruses have also been used but protocols require inactivation of virus to avoid cytotoxicity.

Here, we show that MVA expressing influenza NP and HIV-1 gag can induce CTL specific for immunodominant CTL epitopes which are restricted through diverse HLA alleles. The protocol employed has advantages over other methods in that it is technically simple as there is no need for virus inactivation and it may be useful for the detection and expansion of CTL specific for novel CTL epitopes. Recombinant MVA technology is potentially applicable to the study of immune responses to a large number of proteins from diverse pathogens and tumours. The absence of receptors used by vaccinia virus to subvert the immune response, namely for interferon-α/β, IFN-γ, TNF and CC-chemokines [4], or possibly other uncharacterised changes in MVA, may even enhance its immunogenicity as well as reducing its cytopathic effects.