Our
Research
The Bailis lab aims to understand how metabolism impacts the immune system and human health. Our long-term goal is to translate fundamental mechanistic insights into how cells navigate their own biochemistry and the diverse biochemical environments they find themselves in to metabolically target disease, enhance immunotherapies, and develop dietary treatments that can lower the cost of care while broadening access.
Our work is currently focused in three major areas:
How do single-cell differences in metabolism pattern cell fate?
The body tightly controls when cells are allowed to grow, divide, and differentiate. This is traditionally understood through the action of signaling and gene regulation, with pro-growth signals licensing cells to enter the cell cycle and the mix of genes turned off or on instructing cells to behave differently. While this perspective has provided great insight into why different cells behave differently (“cellular heterogeneity”) and helped inform therapeutic development, we at the Bailis Lab think this overlooks a big ingredient: cell-to-cell variation in metabolism.
Imagine the cells in our body each are a different bowl of soup. Each has it’s own mix of ingredients – like noodles, meat, and vegetables – and is bathed in its own broth. Research tends to focus on the ingredients (i.e. the RNA’s and proteins) and say, “As long as we know the number of noodles, pieces of chicken, and which vegetables there are, I can tell you what type of soup this is.” So we go about doing things like single-cell sequencing and conclude that a cell is a certain type of cell because it has a certain mix of RNA ingredients. At the Bailis Lab, we like to ask, “Would that soup be the same soup if all those noodles, meats, and vegetables were bathed in plain old water? Is the soup really just the ingredients, or does the broth matter too?”
Put another way, you can imagine having single-cell information on every molecule of RNA and protein, every epigenetic mark, every post-translational modification on every protein in every cell in your body. You could then computationally cluster all those cells and feel confident you could make a true “human cell atlas.” Now imagine, you had the ability to keep all those molecules and cell membranes in place, but could magically turn everything inside and outside the cell into deionized water. You could re-sequence and re-run proteomics and get the exact same results, but would the cells still be programmed the same? NO! First, all the RNA and protein lose structure, but even if you could somehow keep them folded, enzymes don’t do magic. Enzymes run naturally occurring reactions at physiologic conditions, which means they need to have substrates and co-factors to work at all. The broth matters!
Taking this soupy thought experiment back to earth, you can imagine that any given transcriptome/proteome will behave differently when placed along a gradient of biochemical states. By tuning the levels of substrates, cofactors, and ions, the same set of RNA’s and proteins will act differently (like ordering your ramen in shoyu or tonkatsu broth, same ingredients, different meal!). So following this through, one might hypothesize that single-cell differences in metabolism could drive cellular heterogeneity alongside signaling and gene regulation.
We have begun to test this hypothesis and found that it holds true in helping explain how T and B cells behave. We recently discovered that NAD biosynthesis acts as a metabolic hub that converts information on how well T and B cells recognize their targets into the cell biology we know as “clonal selection” (i.e. the process in which the body grows out T and B cells that detect pathogens, vaccines, and cancer cells the best). We found that even among cells with the same antigen receptor receiving comparable stimulation, there was broad cell-to-cell variation in NADH and this can explain large facets of heterogeneity in T and B cell responses. Just sorting cells by how much NADH they have less than 24 hours after activation –before they ever divide – can predict the rate of cell cycle, proliferation, and the acquisition of effector function. Much to our surprise, when we transferred cells sorted by early single-cell differences in NADH, it could predict the rate these cells contracted after an infection and persisted into the memory phase weeks later!
We believe this work illustrates how initial differences in metabolism can pattern how cells behave, with profound implications for how we design vaccine strategies, immunotherapies, and treat autoinflammation. Future work in the lab is focused on translating this discovery into those therapeutic areas and uncovering the basic mechanisms explaining how this all works.
How is the “eukaryotic operon” organized?
The immune system is a liquid organ, with immune cells routinely circulating from our blood into our tissues. When an infection occurs or a tumor forms, the metabolic environment within the affected tissues can undergo considerable change, such as depletion of oxygen and nutrients. While we know a fair amount about how those depleted environments can impact immune cells that reside there long-term, there is a moment in time when they pass from nutrient rich blood into the depleted tissue. When that happens, those cells experience a rapid change in their metabolic environment. How immune cells adapt during this transition and how this impacts their biology long-term is poorly understood, with major implications for cancer immunotherapy, infectious disease, and autoinflammatory disorders.
We know a great deal about how bacteria manage this biological problem: using operons, suites of co-regulated genes under the same promoter. In the context of cellular metabolism, operons can contain multiple enzymes in the same network, kept transcriptionally silent by repressors sensitive to key metabolites within the operon-regulated pathway. Thus, when a metabolite reaches a high enough concentration in the environment, there is simultaneous transcription and translation of all genes in a metabolic network that can respond to the changes in the environment. In contrast to prokaryotes, cells in our bodies encode genes for metabolic pathways across multiple chromosomes. This spatial separation of genetic information means that locus control, transcription, and translation all occur independently for each enzyme in a metabolic process. Although we know that eukaryotes also have metabolic sensors that can coordinate gene expression, how eukaryotes maintain coordination of translation and protein abundance of enzymatic networks thereafter and the extent to which transcriptional regulation is sufficient to drive this remains poorly understood. Asked another way, what does the “eukaryotic operon” look like?
We have been working to systematically dissect this question using innate and adaptive immune cells as a biomedically relevant model. Combining RNA-seq, polysome profiling, and whole cell proteomics with CRISPR/Cas9 editing of primary immune cells we have started to map the molecular mechanisms explaining how metabolic adaptation occurs in primary mammalian cells. We have made several major findings that are now the focus of our work: 1) rapid and selective changes in mRNA translation, rather than transcription or protein stability, are the major hallmark of acute-phase responses in dynamic nutrient environments; 2) long-term adaptation and recovery of immune cell function results from a large number of immunologically important transcription factors being upregulated and translated early during nutrient stress responses; 3) even though metabolites are primarily sense through sensors that lead to stereotyped signaling (e.g. mTORC1, GCN2, AMPK), cells “decode” their metabolic environments and have unique responses to specific metabolites.
These observations have led us to now ask: How do cells differentially sense and respond to specific metabolites? What are the cellular processes and RNA binding proteins that lead to selective or altered translation in immune cell responses? How does experiencing metabolic stress and the subsequent remodeling of the transcription factor landscape alter the trajectory of cells fate? We are actively exploring these questions in the context of cancer immunotherapy, T cell differentiation, and programmed cell death in order to therapeutically target and manipulate these pathways to develop new, adaptation-targeted therapeutic strategies.
How does chronic malnutrition lead to immune dysfunction?
Nearly 1 billion people live with restricted food access and experience undernutrition. As a result, the leading cause of childhood mortality in the world is infection in the context of malnutrition and the leading cause of acquired immunodeficiency is nutritionally acquired immunodeficiency syndrome (NAIDS). Despite these being well documented phenomena for decades, we still have extremely limited insight into why undernourished people are so vulnerable to infection, respond poorly to vaccines, and exhibit other broad features of immunodeficiency. Much of the work that has been done has been focused on understanding how caloric and dietary restriction can be used as a therapy in developed countries to enhance immune responses, rather than the goal of treating NAIDS.
While stopping world hunger is beyond the scope of what our small group can hope to do, we are focused on elucidating the molecular and cellular mechanisms explaining NAIDS with the goal of defining dietary components or supplements that could be used to enrich diets and help revitalize the immune system in these individuals. Moreover, we believe the insights we gain in studying the immune systems in these extreme settings will reveal previously underappreciated processes that can be used to modulate inflammation and applied more broadly in therapies. To this end we have been working with a chronic model of malnutrition and found that even after refeeding protocols that recover weight and immune cell production, these animals still remain vulnerable to infection weeks after recovery. We believe this suggests that poverty and limited food access are critical immunologic variables in one’s health history that have largely been overlooked. We are actively investigating how the immune system “remembers” prior malnutrition and exploring the different ways the undernutrition can dysregulate immune responses.
