Our focus on Salmonella infection of macrophages is carefully chosen to both have specific, immediate impact on a global biomedical challenge and to generally enhance our understanding of complex diseases. Macrophages are critical mediators of the innate immune system, and their response dictates the outcome of many infectious diseases. Salmonella is a pathogen of global importance that causes systemic diseases resulting in upwards of 27 million cases and
over 200,000 deaths per year worldwide. Salmonella is also a disease on the rise: although the number of human infections caused by another foodborne pathogen (Escherichia coli O157) dropped by almost 50% in the U.S. between 1997‐2010, Salmonella cases increased by 10% (Frieden, CDC 2013). Moreover, while antibiotics are used to treat many Salmonella cases, antibiotic‐resistant strains have emerged, threatening the possibility of future large‐scale pandemics (Kariuki et al., J Clin Microbiol 2010).
These Salmonella strains are part of a still larger problem ‐ the emergence and rise of infections that are largely uncharacterized and recalcitrant to antibiotic treatment. In 2013, it was estimated that over 2 million patients in the U.S. alone were infected by resistant bacteria, nearly 28,000 of whom lost their lives (Frieden, CDC 2013). The economic cost was equally staggering, an estimated 55 billion dollars due to increased healthcare costs and lost labor.
Importantly, our medical system depends on the routine ability to control these infections ‐ otherwise, every medical procedure with a significant infection risk (from organ transplants to routine operations) would be deemed too risky to perform. Finally, globalization has put the developed world into contact with previously unencountered microbes, exposing naive populations to brand new diseases. A natural consequence of this fact ‐ powerfully demonstrated by the recent Ebola epidemic ‐ is that diseases that were thought to only affect the Third World now directly affect the developed world as well.
The vision of our center is to open an unprecedented window into infections of macrophages by the global pathogen Salmonella, integrating cutting‐edge modeling, computation and experimental measurements of both host and pathogen into a combined systems approach. Our proposal therefore stands at the nexus of modeling, measurement and immunology ‐ all of which were recently identified by the Allen Family Foundation as critical interest areas ‐ and lays the foundation for predictive systems medicine based on whole‐cell and multiscale mechanistic modeling.
Over the last fifteen years, the creation of a “whole‐cell” computer model that explicitly takes all of the known information about each gene and molecule into account to predict cell behaviors was called the “ultimate goal” of systems biology, and “a grand challenge for the 21st century” (Meyer, Cell 2011; Tomita, TiBTech 2001). Significant progress has been made in both whole‐cell and multiscale modeling since that time ‐ including completion of the first
whole‐cell model in 2012, an achievement that was recently highlighted in the journal Cell as one of the most important advances reported by that journal in 40 years (Karr et al., Cell 2012).
The integration of multiscale and whole‐cell computational modeling has the potential to be a major transformative force in medicine, in particular with regard to complex disease, because it can explain complex pathologies in terms of cellular components and their interactions. Multiscale models integrate data at different levels of biological scale – from gene to protein to cell to collections of cells within a tissue, and ultimately to tissues and organs – and are,
therefore, capable of predicting how perturbations at one level of scale (e.g. gene expression) affect important outcomes at other levels of scale (e.g. phenotype and function). Multiscale modeling research efforts thus far have focused on bridging cell‐level characteristics (e.g. cardiomyocyte electrophysiology) to organ‐level outputs (e.g. heart beat rhythms). While those types of multiscale models have already proven useful in the design of surgical strategies to
treat atrial fibrillation, for example, they are fundamentally unable to reveal the underpinning molecular mechanisms of disease. Indeed, a requirement for understanding the molecular basis of a disease – which is often necessary for developing cures – is the creation and deployment of multiscale models that comprehensively represent whole cells (at the levels of genetic and epigenetic regulation, signaling, metabolism, physical structure, etc.), as well as their dynamic
environment and interactions with one another. Our team has unique strengths in whole‐cell, multi‐cell and biophysical modeling, and through the proposed project we will be the first to couple our approaches to generate a new breed of computational model that has unprecedented detail and predictive power.
Our multiscale modeling will integrate heterogeneous data sets of widely varying features into a unified model that spans from individual genes to tens of thousands of cells, so that the world’s combined discoveries and insight can be brought to bear simultaneously on understanding host‐pathogen interactions in the context of Salmonella infections. Our model will also suggest experiments with the highest likelihood of generating new knowledge and thereby shorten
the path to new breakthroughs (Carrera and Covert, TiCB 2015). Most importantly, our model will be able to predict or diagnose complex, multi‐network phenotypes ‐ within individual cells and as a result of cell‐to‐cell interactions and cell‐to‐cell heterogeneity. Notwithstanding recent developments, such approaches are only in their infancy, and extensive work remains to help this technology achieve its potential. The needs of the biomedicine community have given us an
obligation and urgency to expedite the maturation of this field, motivating our proposal for this Allen Discovery Center.
We have chosen Salmonella typhimurium‐macrophage interactions as our model system to integrate multiscale and whole‐cell computational modeling because the outcome of systemic Salmonella infections (e.g. typhoid fever) is dictated by the outcome of this interaction. Salmonella replication is heterogeneous between macrophages (Figure 1; see also Helaine et al., Science 2014). In some macrophages, Salmonella grows aggressively, while in neighboring cells infections result in little or no growth. The subpopulations of non‐replicating or slowly replicating drug‐tolerant bacteria (termed “persisters”) are hypothesized to cause relapsing infections after antibiotic treatment. Moreover, a survey of the recent literature reveals that the putative mechanisms underpinning bacterial heterogeneity are multifaceted and multiscale, invoking nearly all aspects of cell biology, including cell division, cell metabolism, cell signaling, and genetic and epigenetic regulation. For example, heterogeneity in division rates of single Salmonella have been found to correlate with antibiotic efficacy (Claudi et al., Cell 2014). Moreover, formation of Salmonella persisters has been shown to depend on macrophage metabolic factors (Eisele et al., Cell Host Microbe 2013), and heterogeneity in Salmonellalipopolysaccharide modifications has been found to drive the variable interferon response of infected macrophages (Avraham et al., Cell 2015). These recent studies clearly demonstrate the important role of reciprocal feedback between Salmonella and the host macrophages in producing and maintaining cell heterogeneity. Further, they motivate why it is necessary to evaluate these multifaceted mechanisms inside whole cells, as well as across populations of many cells if we hope to understand how cell‐to‐cell variations contribute to infection outcomes and the evolution of drug resistance. Why and how these drug‐tolerant sub‐populations emerge over time are wide open questions that have broad relevance, not only to infectious disease but also to other medical problems, such as cancer metastasis, where aninsufficient understanding of the mechanisms that drive cell heterogeneity has similarly confounded therapy design.