Resistance networks

Organisme : Wellcome Leap
Programme : Resistance networks
Financement : 50 M $ pour les 3 volets, budget variable par projet
Durée : 3 ans et plus

Description 

Le programme Wellcome Leap est une initiative internationale de financement de la recherche biomédicale créée par la fondation Wellcome afin d’accélérer des percées scientifiques majeures en santé humaine. Inspiré du modèle de la DARPA américaine, il soutient des projets ambitieux, à haut risque et à fort potentiel d’impact, en réunissant des chercheurs, cliniciens, ingénieurs et partenaires industriels du monde entier autour d’objectifs précis et mesurables à atteindre sur un horizon de 5 à 10 ans. Contrairement aux mécanismes traditionnels de financement, Wellcome Leap investit à grande échelle dans des programmes thématiques ciblant des défis complexes, tels que la santé des femmes, les maladies cardiovasculaires, la santé mentale, les maladies infectieuses, le développement de l’enfant ou les technologies biomédicales de rupture. Son approche repose sur des jalons de performance rigoureux, une collaboration internationale multidisciplinaire et une volonté de produire rapidement des innovations susceptibles de transformer les soins et les résultats de santé à l’échelle mondiale.

The goal of Resistance Networks is to build an epidemiological model that can predict whether a plasmid carrying antibiotic resistance genes (pARG) has epidemic-like potential — an R₀ greater than 1 — to spread within and between human bacterial networks during and after antibiotic use.

The program seeks 80% balanced predictive accuracy, sufficient to enable the design and testing of new individual and public health measures. Such a model would provide a leading indicator of resistance trajectories — a signal of rising pARG before resistance reaches the levels at which empiric therapies must change. Current approaches rely on resistance surveillance as a tool for population-level management; for example, changing empiric prescribing guidelines once E. coli resistance exceeds 20% in local isolates.50

If successful, model-based interventions could slow the emergence of resistance by as much as a factor of 2 or more — averting more than 1,300 deaths from antibiotic resistance per day by 2050, more than a new antibiotic class targeting priority pathogens alone.22,51,52 And if plasmids are a primary driver of resistance to new antibiotics, such approaches could extend the working life of the drugs that follow and change the market dynamics that have made antibiotic development so difficult.

Thrust 1: Determine, at 80% balanced predictive accuracy, whether antibiotic-amplified pARGs exhibit epidemic-like transmission (R₀ > 1) within and between gut bacterial networks in response to antibiotic use.

Thrust 1 counts new transmission events — between bacterial cells inside the antibiotic-treated gut, and between individuals in close-contact networks — across focal communities that differ in antibiotic exposure and microbiome composition, and identifies the levers for intervention at each scale. It separates transmission into two coupled processes: R₀-within (how efficiently a pARG spreads between bacterial cells inside a treated gut) and R₀-between (how reliably it transfers to close contacts afterward). The thrust pairs a multiscale modeling effort with the longitudinal cohort measurements that feed it — sampling the gut microbiome before, during, and after antibiotic exposure, alongside close contacts and household environments, at single-cell resolution that links resistance genes to their plasmids and bacterial hosts.45 Of interest are longitudinal cohort designs with linked household sampling, single-cell sequencing that preserves plasmid-to-host linkage, multiscale transmission modeling,53–55 and wastewater and isolate based surveillance of resistance able to independently test whether a pARG sits above the epidemic threshold.56

Thrust 2: Advance cost-effective tools that enable single-cell sequencing technology to scale measurements of plasmid transmission.

Thrust 2 develops scalable library preparations that read which resistance gene sits on which plasmid inside which bacterial host cell — as complete, circularized host–plasmid sequence pairs — across roughly 1,000 Enterobacterales cells per complex microbiome sample. The goal is to make the core Thrust 1 measurements affordable at epidemiological scale, where the dominant cost is long-read sequencing and per-cell library preparation. Promising approaches include encapsulation or barcoding strategies that pool thousands of single cells into one library without per-clone culture, hydrogel immobilization, and linked-long-read methods adapted for bacterial cells.57 Success is a validated workflow that demonstrates ≥95% linkage concordance with full long-read sequencing,13 processes ≥1,000 cells per sample with retained linkage at a 5- to 10-fold per-cell cost reduction, and biobanks cell pools for cross-program use.

Thrust 3: Test in silico, model-guided key parameters and interventions in controlled in vivo systems, targeting the dominant driver(s) identified by Thrust 1.

Thrust 3 validates the Thrust 1 transmission models in a controlled in vivo system: if the model is right, targeting the barriers it identifies should lower R₀. The aim is to show that acting on a dominant driver — e.g., within-gut amplification, conjugation rate, or between-individual transmission — can push R₀ below 1, or reduce the probability of crossing it. Because Thrusts 1 and 3 run in parallel, proposals should commit to a concrete starting intervention while building in flexibility to re-target as the R₀ decomposition sharpens. Interventions may act within the gut (microbiome restoration, competitive exclusion, conjugation inhibition, phage, or CRISPR-based approaches)58–61 or on between-animal transmission interventions (cohorting and environmental controls); swine systems are of particular interest, both as the closest proxy for human plasmid networks49,62,63 and as a test of interventions against zoonotic transmission. Success means a measurable reduction in at least one component of R₀ — for example, a ≥10-fold reduction in conjugation rate or shedding, or a ≥10-percentage-point increase in the fraction of gut bacteria protected from plasmid acquisition.

(tiré du site web du programme)