The use of antibiotics has revolutionized modern medicine, increasing our life expectancy and allowing for more complex medical procedures. Antibiotic resistance is becoming a pressing issue, exacerbated by the lack of antibiotic development, with only 5 new molecules in clinical Phases II or III, none of them representing a new class of compound. A second limitation with the use of traditional antibiotics is that they have a broad spectrum of action: many different bacterial species will be killed, including those that are beneficial. The disappearance or reduction of these species can leave free ecological niches that are sometimes colonized by disease-causing, resistant bacterial variants. There is therefore an urgent need for targeted therapies that enable a precise control of complex microbial ecosystems at will, including the reduction of antibiotic-resistant bacteria. Towards this goal, we developed the first programmable sequence-specific antimicrobials. Our technology relies on a biological vector derived from bacteriophages, called phagemid, which can deliver a therapeutic genetic circuit only to the specific bacteria of choice. We programmed a phagemid particle to deliver a genetic circuit containing CRISPR/Cas9, which encode very specific DNA-sequence targeting and cleavage elements. This approach allows for two control points to be implemented, as opposed to traditional antibiotic molecules: first, we take advantage of the natural capacity of the bacteriophages to recognize only a small subset of bacterial species; second, within those species, only those that contain the targeted DNA sequence will be killed, leaving the rest of the population intact. In this project, we will focus on one of the emergent and more serious threats caused by antibiotic resistant bacteria: we will generate a targeted therapy that will allow for the destruction of Extended-Spectrum Beta-Lactamase E. coli (ESBL-E.coli). To do this, we will first screen for new natural bacteriophages that can be used as delivery vehicles to target pathogenic bacteria. We are implementing high-throughput robotic technologies to accelerate the discovery process that will allow us to screen for phage variants with different characteristics, such as host range and injection capacities. State-of-the-art sequencing techniques will also help us shed light on the determination of viral DNA packaging elements, which we will then implement into the pipeline of pure, engineered phagemid production. In parallel, we are applying novel bioengineering techniques to modify both previously described phages and newly discovered variants to alter their host cell tropism, synthesize and pack our targeted genetic circuits and include mechanisms to bypass the natural bacterial defenses against phages. Finally, we will establish an ESBL-E.coli mouse colonization model to demonstrate the efficacy of our targeted antimicrobials. This will enable to optimize and quantify the efficiency of delivery and targeted killing of these pathogenic bacteria in a variety of conditions. In sum, the knowledge and tools which will be developed in this project will open the field of “bacterial gene-therapy”, enabling precise microbiome manipulation and targeted killing within complex microbial ecosystems, and could therefore lead to next-generation personalized therapies. Such a strategy could have impactful applications in human health, notably to control and treat several infectious diseases.

Global change and anthropogenic activities have caused a worldwide increase in reports of vibrio-associated diseases with ecosystem-wide impacts on humans and marine animals. This is illustrated by the recent outbreaks of Crassostrea gigas diseases in France associated with vibrios. Understanding of the ecological and evolutionary dynamics of these infectious agents is important for diagnosing, predicting and preventing diseases in farmed and wild species. We recently demonstrated that during oyster mortality events, particular ecological populations of Vibrio are more abundant in oyster tissues than in surrounding seawater. Here we define an ecological population as the genetic unit representing a cohesive ecology, gene flow and social attributes. We also showed that some of these ecological populations (such as V. crassostreae) contain a high proportion of pathogenic strains. The pathogenicity of V. crassostreae relies on the presence of a large mobilizable plasmid. Interestingly a preliminary study demonstrated that this plasmid is not detected in V. crassostreae strains isolated from oysters sampled in Northern Europe where mortalities are scarce. Our current hypothesis is that V. crassostreae is an oyster commensal that turned into a pathogen after a plasmid acquisition. While V. crassostreae is abundant in all diseased oysters, this population always co-occurs with diverse populations in each individual. Experimental infections have suggested that interactions between strains may play a role in the disease. Hence, populations are the unit of pathogenesis in juvenile oysters in a polymicrobial context. Therefore, an approach aimed at specifically eliminating a target population is needed to understand its specific role in the oyster disease process. One such strategy is to utilize naturally occurring predators of bacteria as Vibriophages that are extremely abundant in nature, present high host specificity and can thus be considered as an “eco friendly” method. A separate study suggested that co-evolution drives association of oysters and vibrios. Larvae from two genetically and geographically distinct sources (Texel and Sylt, North Europe) were exposed to diverse pathogenic vibrios isolated from both locations. For several vibrio species, sympatric combinations were associated with a lower mortality rate. The genome comparison of several strains highlighted two genes that clustered locally matching the virulence phenotype while showing incongruent patterns relative to the evolutionary history of the strains. The selective pressures leading to these geographic rather than phylogenetic matches are unknown, but it seems likely that they result from adaptations to living associated with oyster hosts. The overall aim of this project is to investigate the evolution of vibrio virulence in relation to the oyster as a host and phages as predators. This will allow to determine whether i) pathogenic genotypes of Vibrio emerge in areas experiencing mass mortalities; ii) Vibrio and oysters co-evolve; iii) Vibrio-phage infection networks are more highly connected in co-occurring phages and hosts; iv) elimination of a pathogenic population (V. crassostreae) by a phage cocktail improves oyster survival. To this end, we will use a combination of population modeling, comparative and functional genomic analysis in addition to experimental pathology. • In Task 1, we will explore the population structure of Vibrio in oysters and seawater in locations affected by a gradient of disease severity (Brest, Texel and Sylt) and investigate V. crassostreae diversity. • In T2, we will perform cross infection experiments using lab-bred specific pathogen free (SPF) oysters produced from genitors collected at Brest, Texel and Sylt and Vibrios isolated at the 3 different locations. • In T3, we will explore the differences in the phage-V. crassostreae infection network over time and space and test the effect of a phage cocktail on oyster disease.

SPELITEC is the first treatment for Shiga toxin-producing E. coli (STEC) infections, which affects primary young children and can be fatal in 1-5% of the cases. Currently there is no validated safe treatment: hospitalized patients receive only symptomatic treatments consisting of rehydration coupled with transfusions and dialysis for haemolytic uremic complications. The SPELITEC program is designed to finalise preclinical and initiate clinical studies for elimination of STEC bacteria (Shiga-toxin producing E. coli) in paediatric populations with a dedicated product: EB003. EB003 selectively eliminates STEC bacteria early in the disease progression; it is not expected to exert any toxicity in humans or animals; it is a multivalent mix of different capsid variants allows to target the whole spectrum of clinically relevant strains and pre-empts the acquisition of resistance through mutation of bacterial receptors; it is administered orally by a liquid/gel which is patient-friendly and adapted to young children and it represents a cost-effective alternative to antibodies against Shiga toxins. Our technology is the only one allowing the specific elimination of STECs without toxin release and without disrupting microbiota and inducing antibiotic resistance. Eligo will complete early drug development, which should ensure market authorization given the orphan and unmet medical need nature of the indication, and plans to out-license the compound to an established pharma company that will launch the product on the market. The management team of Eligo has solid business and technical management experience in multinational companies and a very strong drug development background. During the Phase 2 project we will validate this new individualised therapeutic approach and mechanism of action by conducting the validation of the pharmacological model, the production of a clinical batch and the safety tests in healthy animals.