Bivalent vaccines derived from measles vaccine
Major childhood diseases, such as polio or measles, have been virtually eliminated since the 60s thanks to the use of live-attenuated vaccines. Today, millions of children worldwide are still exposed to terrible infectious diseases for which no vaccine is available: AIDS, malaria, dengue fever, infants bronchiolitis. Our laboratory is working on a combined strategy to develop pediatric prophylactic vaccines. Our strategy is derived from measles live-attenuated vaccine, a mostly safe and effective human vaccine, engineered to express additional antigens. We are devoting our efforts to develop this strategy from the conception of antigens and vaccine vectors to clinical trials. The measles vaccine has been administered to hundreds of millions of children over the last 40 years and proved to afford life-long immunity after a single administration. Easily produced in many countries this vaccine is distributed through the Expanded Program on Immunization of WHO. These characteristics led us to consider its use as a pediatric vector designed to protect children against measles and simultaneously to immunize them against other infections. Such affordable recombinant vaccines are attractive for the developing world. We realize that all these vaccines will not be used simultaneously. However, depending on the region of the world where diverse epidemics threaten children, some combinations may still be used successfully.
With this aim, we demonstrated the strong and stable expression from this vector of genes encoding proteins from HIV, WNV, DENV, CHIKV, SARScoV, H5N1, P. falciparum or mTB. These vectors are immunogenic in mice transgenic for MV receptor and in non-human primates, inducing long-term neutralizing antibodies and cellular immunity. In the case of HIV and chikungunya, we succeeded, in collaboration with industrial partners, to introduce two recombinant vaccine candidates in phase I clinical trials in adults. GMP batches of recombinant MV virus expressing HIV clade B Gag-Pol-Nef or CHIKV structural proteins as virus-like-particles (VLP) (see Figure 1A) were generated and shown to be immunogenic in preclinical models. Their toxicity was assessed in macaques, showing a biodistribution and shedding similar to that of standard measles vaccine, both vaccines replicating preferentially in secondary lymphoid organs and epithelium-rich tissues. These vaccines were well tolerated and did not induce local or systemic toxicity. Next generation of MV-HIV vectors, either promoting the budding of assembled HIV1 Gag/Env VLPs, or chimeric virus with HIV1 Env on the viral surface, have been produced and proved to be immunogenic in preclinical models. These programs are still ongoing.
We are also developing a measles-dengue preventive vaccine. To avoid the interference problems of using a vaccine mixture containing the four serotypes of DENV, we generated a single MV vector expressing a tetravalent combined dengue antigen derived from the four serotypes of DENV. This recombinant virus elicited durable specific neutralizing antibodies to the four serotypes of dengue virus in mice and partially protected macaques from DENV infectious challenge. This vaccine is currently improved by adding newly identified DENV antigens before clinical evaluation.
Because measles vaccine provides efficient protection against a respiratory acquired disease, this strategy appears suitable to other respiratory viral diseases. To test this possibility we generated recombinant MV expressing the spike protein of SARS coronavirus or the hemaglutinin and neuraminidase genes of H5N1 influenza virus. Both recombinant vaccines were highly immunogenic in mice and protected all immunized mice from lethal challenge. This program is currently continued with the aim of generating a universal measles-flu vaccine for pediatric use. Similarly, we are working on several strategies to generate MV-RSV vaccine candidates to protect infants from bronchiolitis.
Since several years, we have developed in collaboration with our colleagues of Inserm-892 (Nantes) the use of modified MV vectors for oncolytic treatment of aggressive tumors. We have demonstrated in a series of original articles the capacity of these vectors to specifically kill tumor cells from human mesothelioma (asbestos pleural cancer), melanoma, lung, and colon adenocarcinoma (in vitro and in vivo in mouse engrafted models). We also demonstrated the capacity of these vectors to elicit antitumoral CD8 T cell immunity through cross presentation in MoDC and pDC. This work is now progressing to clinical trials in cancer patients.
A new vaccine platform derived from viral nanoparticles
We have recently developed a new vaccine platform based on measles virus nucleoprotein (N). This protein, which composes the helical viral nucleocapsid, has the property to auto-assemble around RNA molecules (viral genome or any RNA) to form highly stable multimeric ribonucleoprotein rods (RNP) with a diameter of 20 nm and variable length from 100 to 500 nm (see Figure 1B). These biological nanoparticles can be produced in mammalian cells or in yeast. We are expressing the N protein in fusion with heterologous antigens in Pichia pastoris yeast. As a proof-of-concept, we demonstrated the immunogenicity and protective capacity of whole heat-inactivated yeast expressing a well-known malaria antigen, the circumsporozoite protein (CSP) multimerized on RNP, without adjuvant. We are currently extending this technology to other antigens.
Proteomics of virus-host interactions
Virulence, pathogenesis and emergence of viruses are directly linked to the characteristics of their molecular interactions with host-cell components. These interactions both modulate the host immune response and allow viruses to hijack the cellular machinery. To identify virus-host interactions at a proteome level, we have developed a large-scale mapping project based on high-throughput technologies, including recombination-based cloning of viral coding sequences, expression in human cells and state-of-the-art yeast two-hybrid (Y2H). Using this tool, we have identified hundreds of virus-host protein-protein interactions, taking advantage of the unique collection of viruses available at Institut Pasteur. Besides collaborative programs developed inside or outside of this institution, our efforts mostly focus on two viral families of higher interest for our research group: Paramyxoviridae and arboviruses from flavivirus and alphavirus genera such as chikungunya virus. As part of this research program, we have recently studied Tioman virus (TioV), a member of Paramyxoviridae family related to mumps virus (MuV). TioV was first isolated from giant fruit bats in Southeast Asia together with Nipah virus (NiV), a related Paramyxoviridae that is extremely pathogenic in human. Despite serological evidence of close contacts between TioV and human populations, whether TioV is associated to some human pathology remains undetermined. Using our screening pipeline to characterize virus-host protein interactions, we have shown that in contrast to V proteins of MuV or NiV, the V protein of TioV hardly interacts with human STAT2 and cannot block IFN-alpha/beta signaling in human cells. Altogether, these observations question the capacity of TioV to control type I interferon signaling in both human and giant fruit bats, which suggest some limited virulence as opposed to closely related Paramyxoviridae. Additionally, we developed a new method based on recombinant viruses expressing tagged viral proteins to capture both direct and indirect physical binding partners during infection. Using this technology and mass spectrometry, we identified a prosperous list of cellular partners of MV-V protein.
Identification of new antiviral compounds
A large fraction of virus-host interactions that we identified are involved in the inhibition of innate immune response, in particular cellular response to type I interferons. These antiviral cytokines determine the establishment of a resistance state through the induction of a large panel of antiviral cellular factors. Chemical compounds inducing expression of the antiviral gene cluster in virus-infected cells would be of a major clinical interest in the treatment of viral infections but have not been identified yet. To identify such immunostimulatory molecules, we have designed an in vivo functional assay, and used this system to screen 44.000 small molecules in collaboration with the Organic Chemistry Unit (H. Munier-Lehmann). Three families of compounds identified with this approach exhibit a broad-spectrum antiviral activity (see Figure 2), blocking the replication of numerous RNA viruses and boosting the expression of antiviral cellular genes in human cells. While searching for their mode of action, these compounds were characterized as inhibitors of pyrimidine biosynthesis pathway, a key enzymatic cascade that is essential to nucleotide synthesis. This established a yet unsuspected link between pyrimidine biosynthesis and the innate antiviral response. Furthermore, the antiviral activity of these compounds was found strictly dependent on cellular gene transcription and required Interferon Regulatory Factor 1 (IRF1). Altogether, our results better explain the antiviral property of pyrimidine biosynthesis inhibitors and unravel a novel pathway that induces cell resistance to RNA virus infections. Our objective is now to establish the therapeutic interest of these molecules in the treatment of viral infections using in vivo models.
Virus-host ribonucleoprotein complexes and innate immunity
The RIG-I-like receptors (RLRs) play a major role in sensing intracytoplasmic RNA virus infection to initiate and modulate antiviral immunity. These RNA helicases interact with particular signatures of viral RNA but also with cellular protein and RNA partners, most of them being still unknown. To decipher the network of interactions between RLRs and proteins or RNA molecules in a viral context, we developed cutting edge high-throughput approaches based on tagged protein affinity purification, next-generation sequencing of RNA molecules and mass-spectrometry analysis of protein complexes. We obtained extensive lists of specific virus-host interactions between RLRs and protein/RNA molecules in the context of measles and chikungunya infections. Investigating these complexes into action within infected cells will help to understand the molecular features underlying innate immune response to these viruses. A similar project is developed in the context of HIV infection. Understanding the characteristics of viral RNA ligands recognized by RLRs and the proteins that control RLR signaling should guide RLR-targeted therapeutics for antiviral and immune-modulation applications.