The emergence, evolution and spread of (drug resistant) infectious diseases are a multi-faceted global public health challenge.

The current treatment against many bacterial infections is a cocktail of broad-spectrum antibiotics, which is increasingly leading to treatment failure, infection recurrence and rising rates of antibiotic resistance (>2.8 million infections per year in the U.S. alone). Antibiotic resistance and treatment failure are thereby driving a growing problem, with millions of casualties and untreatable infections every year, which is further exacerbated by the stalling of traditional antibiotic development. Importantly, approaches such as antibodies engineered to target specific molecules on cancer cells have been successful as anti-cancer treatments, while interest in vaccines targeting bacterial and viral infections, and even cancer, have received an insurgence in attention at least partially due to their success against COVID-19. Consequently, antibody-based drugs and bacterial vaccines, hold enormous potential in the treatment of and protection against infectious diseases. However, this partially depends on the efficient identification of potent antibody and vaccine bacterial antigens. Moreover, drugs to treat acute, and chronic infections are also instrumental in limiting and preventing the further emergence and spread of resistant strains and species. With the advent of new experimental and computational tools, opportunities are arising to achieve a deeper, integrated understanding of how pathogens interact with drugs, their environment and host. We believe that this in turn can help enable data driven identification of antigens that enhance the immune system’s ability to efficiently prevent, inhibit or clear an infection. Moreover, these tools enable data driven characterization of drugs, optimization of their efficacy, and the design of new strategies to combine them with other drugs and approaches.

Fueling the development of new antimicrobial strategies.

New technology drives innovation and discovery, which is why experimental and computational tool development are foundational to the lab. These tools are geared towards achieving two main goals:  1. To map out bacterial immune evasion mechanisms, for data-driven antigen discovery, and the development of immune potentiating antimicrobial strategies. 2. To identify the paths leading to drug sensitivity and evasion, to develop data-driven drug potentiation strategies.

Experimental & Computational Tool Development

A lot of the work we do is focused on capturing how an entire pathogen responds to and interacts with its environment, such as the host and or drugs. Thereby, when we develop a new tool we always consider how to capture as much of the organism as possible. For instance, a simple example is to capture a pathogen’s entire transcriptome in response to a specific environment, or how all the genes in the genome contribute to a specific phenotype. A second consideration is how to make the tool high throughput, for instance, instead of measuring one response to one condition or drug, how can we measure how these responses change over time, and in response to hundreds of conditions? Consequently, and this is a key point, such approaches generate large amounts of data, which require appropriate tools to assist in processing and analysis. This creates opportunities to develop machine learning/AI-mediated approaches to uncover non-obvious patterns in the data that not only reveal new biological insights, but can be leveraged to design novel treatment strategies. Examples of experimental and computational tools are: 

Tn-Seq and Hii-TnSeq: Advanced techniques for genome-wide mapping of bacterial responses to environmental stresses and host immune factors, highlighting genetic determinants of drug sensitivity and immune evasion.

Droplet Tn-Seq: A microfluidic approach enabling the analysis of thousands of single bacterial cells simultaneously, unlocking insights into cell-to-cell communication and antibiotic interactions.

CRISPRi-TnSeq and dualCRISPRI: CRISPR interference, possibly combined with Tn-Seq to create detailed genetic interaction networks, expanding our understanding of a genome-wide bacterial architecture.

Machine Learning Models like PREDDIE: Predicting bacterial responses to various drug combinations, allowing for the identification of optimal therapeutic strategies in clinical settings.

Aerobio: Our cloud-based platform that automates the analysis of high-throughput genomic data, empowering researchers to derive actionable insights from large datasets.

Below you’ll find two ‘movies’ made about the lab back in 2019.

Data Driven Immune Potentiating Antigen Discovery

Many of the tools we have developed over the years enabled us to build ‘multidimensional profiles’ (MDPs) for a variety of bacterial pathogens. These profiles contain a wide variety of information for instance their transcriptional (e.g., (sc)RNA-Seq) and phenotypic (e.g., Tn-Seq, CRISPRi) responses to many (in vitro and in vivo) conditions. We showed that through network integration these data can be represented as a network, which can be computationally mined to obtain leads for new gene-function, filling gaps in pathways and identifying mechanisms underlying specific phenotypes. Since this approach has been so successful in uncovering new biology, we aimed to take this a step further. We hypothesized that by identifying the key genetic determinants that aid pathogens in immune evasion and integrating this information into a MDP, we could develop a data driven approach towards antigen discovery and support the development of immune potentiating strategies. Importantly we have now indeed developed several approaches that are able to trigger a potent immune response, and rescue up to 90% of host (e.g., mice) during an acute infection. Moreover, our MDPs correctly predict that specific immune potentiating targets also trigger synergy with certain antibiotics. This shows that by identifying key immune evasion nodes strategies can be developed that resensitize bacteria to the immune system, trigger a potent and infection clearing immune response, and can even simultaneously (re)sensitize bacteria to antibiotics.

Data Driven Drug Evasion & Potentiation Strategies

The rise in antibacterial treatment failure due to drug resistance is projected to significantly increase in the coming years, requiring the development of new antimicrobial strategies. This evolving complex problem is only solvable if besides developing new drugs we also learn to untangle the exact (genetic) processes that affect drug sensitivity and enable drug tolerance and resistance to emerge. While for many antibiotics we know which genomic changes can cause resistance, the evolutionary paths that lead to resistance have rarely been interrogated. For instance, it is often not clear if tolerance or lowered drug sensitivity precedes resistance. More vexing is the fact that clinical strains isolated from patients who have suffered antibiotic treatment failure (ATF) may lack known resistance markers and instead contain multiple changes that have no known role in resistance. Therefore, understanding which genes and biochemical pathways drive altered drug susceptibility contributes toward identifying genomic changes that both sensitize and desensitize bacteria to clinically effective drugs. Importantly, we hypothesize that this creates opportunities to computationally predict conditions that optimize drug efficacy, and drug-interactions that are most effective in clearing infections. Importantly, we have now developed powerful machine learning tools that exploit our multidimensional profiles and are delivering unique clinical application opportunities, by predicting conditions that optimize drug efficacy and drug interactions, which is a combinatorial space that is too large to experimentally test.