Altered metabolism is a nearly universal characteristic of cancer. Similarly, normal lymphocytes require metabolic adaptations to satisfy various proliferative and effector function demands. However, the nutrient regulation, preferences and requirements that allow different cancer cells to survive or different immune cells to properly function remain poorly understood. Therefore, the significant interest to exploit cancer metabolism or to modulate immunometabolism for patient benefit will require an improved understanding of cellular metabolism and the influence of metabolite availability on other cellular processes in these diverse cell types.
Cell culture certainly offers an experimental system of unmatched scope, flexibility, and accessibility both to investigate cell physiology and to evaluate drug efficacy and toxicity. And implicit to the use of cultured cells for these aims is an expectation that in vitro phenotypes reasonably reflect in vivo cell behavior. However, while it has become better appreciated that environmental factors influence metabolism and other intertwined cellular processes, our understanding of cancer (and immune cell) physiology and drug toxicity is largely based on cells cultured in media that poorly resemble the composition of human blood. The overarching hypothesis of our group is that conventional model systems used to examine human cells have masked otherwise critical biological and pharmacological insights that have greater physiologic relevance for cancer intervention.
- Tool Development
Our lab previously developed human plasma-like medium (HPLM), a new culture medium that contains polar metabolites and salts at concentrations comparable to those of normal adult human plasma (Cantor et al., Cell, 2017). In published work, we showed that relative to reference conventional media, HPLM had widespread effects on the metabolism of cultured cells, including alterations to the metabolome and glucose carbon utilization. Through the use of HPLM, we also discovered an example of metabolic regulation mediated by a metabolite (uric acid) that, among basal culture media, was uniquely defined in HPLM, and whose plasma concentrations differ by up to 10-fold between humans and mice. We went on to demonstrate that, as a consequence of this unforeseen regulation, HPLM could decrease the relative cytotoxicity of 5-fluorouracil (5-FU), a classic chemotherapeutic that remains in wide use.
We now have interest in exploring the development of additional HPLM derivatives based on metabolic conditions that are reflective of different diets, (patho)physiologic states, or even other sites within the human body.
Interested in trying HPLM?
While we continue progress toward making HPLM commercially available, please contact Jason to inquire about receiving a sample of the homemade formulation.
We are also interested in utilizing in vitro systems that complement conventional cell culture approaches. In particular, drawing from concepts in bioreactor design typically associated with chemical engineering, we have adapted mammalian cell bioreactors to operate as chemostats. In contrast to typical (batch) cell culture, chemostats are continuous flow systems that enable cell growth under tightly controlled and constant conditions, i.e. operation at steady state. A primary advantage of steady state cell culture is the potential to study phenotypic consequences of altering a single environmental parameter while keeping all others constant, thus limiting secondary effects.
Blood cancer cell line collection
Cancer is a remarkably complex collection of diseases, which makes it unsurprising that different cancer (sub)types are marked by heterogeneous metabolic preferences and dependencies, therapeutic regimens, and treatment outcomes (Cantor and Sabatini, Cancer Discov., 2012). Given our interest in studying hematological cancers, we previously constructed a collection of over 50 human blood cancer cell lines that together sample various subtypes of leukemia, lymphoma, and myeloma. Across this collection, we have also engineered individual cell lines to harbor unique DNA barcodes, which in turn, permits several cell lines to be cultured in a pooled fashion and tracked via deep sequencing. Such an approach was described previously by others (Birsoy et al., Nature, 2014), and an iteration of our barcoded blood cancer cell line collection was used analogously in a recent study (Kanarek et al., Nature, 2018).
- Research Areas
Among a panel of diverse blood cancer cell lines, we previously showed that metabolic alterations induced by culture in HPLM could be categorized as either shared or cell line-specific, underscoring how cell-intrinsic and cell-extrinsic factors together can contribute to cell physiology. Using biochemistry and various approaches in metabolomics, we aim to further explore and characterize the physiologic functions of exogenous metabolites and to understand how the metabolic activities across diverse human blood cancers are affected by environmental factors that better recapitulate more physiologically relevant contexts.
CRISPR-based loss-of-function (LOF) screens offer a promising approach for characterizing gene function and identifying cancer cell vulnerabilities. In recent years, such screens have been used to catalog essential genes in different cancer cell lines and, along with similarly motivated RNAi-based screens, to identify genetic perturbations that impair cancer cell proliferation in the context of specific cell-intrinsic characteristics, such as genetic background or lineage. However, while environmental factors influence metabolic and intertwined processes of cancer cell physiology, large-scale pooled CRISPR (and RNAi) screening efforts have been largely limited to cells cultured in media that poorly mimic physiologic conditions. Therefore, we hypothesize that paired LOF CRISPR-based screens, performed in either a conventional medium or HPLM, will reveal non-cell-autonomous context-dependent genetic vulnerabilities that have greater physiologic relevance toward understanding protein functional roles and identifying opportunities for cancer therapy. We will integrate such screens with methods in traditional biochemistry and metabolomics to better understand gene-nutrient interactions, and ultimately hope that these findings will reveal unforeseen therapeutic opportunities.
Across diverse human cancers, the use of cell culture remains essential to virtually all efforts in drug discovery and development. However, clinical success has often failed to follow from the identification of promising preclinical drug candidates, as largely attributed to the limited modeling capacity of cell culture systems. And indeed, while the influence of environmental factors on cancer physiology and drug responses has become better appreciated, high-throughput compound screens and ensuing validation efforts have been restricted to cells cultured in media that poorly mimic physiologic conditions. Guided by our published results describing the relative antagonism of 5-FU cytotoxicity in HPLM, we hypothesize that paired small molecule screens, performed in either a conventional medium or HPLM, will reveal additional unforeseen context-dependent liabilities and metabolite-drug interactions that have greater physiologic relevance for cancer therapy. Guided by such screens, we will again use various approaches to better understand mechanism(s) that dictate medium-dependent drug phenotypes.
Blood cancer nutrient dependencies
Several cancer types are auxotrophic for one or more amino acids and therefore rely on import of a particular amino acid(s) from the serum to survive. This type of liability can be exploited through systemic depletion of the circulating “cancer-essential” amino acid (Cantor et al., Methods Enzymol., 2012). The most notable example of this strategy is the clinical success of asparaginase in treating acute lymphoblastic leukemia (ALL). Whereas normal cells can produce the non-essential amino acid asparagine (L-Asn) de novo via asparagine synthetase (ASNS), certain ALL cells express very low levels of ASNS and require uptake of serum L-Asn to survive. Asparaginase depletes serum L-Asn, resulting in selective induction of apoptosis in the auxotrophic ALL cells. However, ASNS levels do not provide an absolute predictive indication of susceptibility to asparaginase treatment, strongly indicating that unbiased approaches could more effectively identify the heterogeneous dependencies of blood cancers to different exogenous biomolecules. We will first utilize our unique set of tools to identify such liabilities, and then combining approaches ranging from expression analyses to metabolomics and CRISPR-based screens, we will aim to determine why certain cells are sensitive to the absence of a particular nutrient, whereby such phenotypes may be indeed dependent upon culture in a more physiologically relevant medium.
There has been a recent surge of interest in understanding how the metabolic phenotypes of different immune cells are directly linked to differentiation states and effector functions. Within this space, there has been a growing appreciation that environmental factors can have dramatic effects on immune cell function, with a particular focus on (altered) availability or accumulation across a small number of exogenous nutrients. Further, like our understanding of cancer cell metabolism, that of normal lymphocytes and other immune cells is largely based on in vitro experimentation using cell lines or cells isolated from tumors or blood. Combining the tools that we have developed with methods in both metabolomics and immunobiology, we are interested in examining how the metabolic phenotypes of normal lymphocytes are affected by environmental factors that more closely reflect physiological contexts. We aim to better characterize and understand the functional relationship between exogenous metabolites and lymphocyte metabolism, and ultimately, we hope to translate these findings into new approaches in modulating immune cell function.