Metabolic pathways are essential for maintaining cellular homeostasis, and in proliferative cells the demand to accumulate biomass requires metabolic pathway alteration. Indeed, altered bioenergetics is a hallmark of cancer. In contrast to normally proliferating cells, those within a tumor are frequently starved for nutrients due to a combination of increased nutrient consumption and unreliable tumor vasculature. Fluctuations in nutrient availability affect cancer cell metabolism, and may be part of a selective process that occurs during tumorigenesis. Thus, understanding both the altered metabolism of cancer cells and their response to nutrient deprivation will enable identification of metabolic liabilities that can be exploited for cancer therapy.

The Possemato Lab seeks to understand the pathways that are altered in cancer, the processes that these altered pathways support, and the environmental, genetic, and epigenetic contexts in which these pathways are important.

To this end, we focus on three key areas: iron-cluster biosynthesis, fluctuations in metabolism that are cell cycle dependent, and serine biosynthesis and the key enzyme PHGDH.

Iron-Cluster Biosynthesis

We found that the requirement for iron-cluster biosynthesis is dependent on environmental oxygen concentration. We are asking whether activation of this pathway is an important step in lung tumorigenesis and can be exploited by increasing the sensitivity of cells to ferroptosis, a type of oxidative cell death.

Fluctuations in Metabolism that Are Cell Cycle Dependent

We utilize traditional cell cycle synchronization methods and a murine liver model of tissue regeneration to identify key metabolic regulators of proliferative metabolism.

Serine Biosynthesis and the Key Enzyme PHGDH

We identified PHGDH, now a member of the PAM50 gene set, as being highly expressed and required for the viability of estrogen receptor (ER)-negative breast cancer cells. We are investigating how this pathway supports tumor metabolism and the requirement for individual pathway components across cancer types.

Key Findings From Major Publications

In our paper Allosteric regulation of CAD modulates de novo pyrimidine synthesis during the cell cycle, published in Nature Metabolism in 2023, key findings included the following:

• CAD activity fluctuates during the cell cycle, peaking in S phase.
• UTP levels change during the cell cycle, opposite to CAD activity.
• Changes in CAD activity are explained by allosteric inhibition of its activity by UTP, not by phosphorylation or oligomerization downstream of growth factor signaling.
• Phosphorylation of CAD at S1859 (S6K site) does not affect its oligomerization.
• Two flexible loops in the CAD allosteric domain evolved in animals to mediate regulation by UTP.
• UTP binding regulates the formation of a substrate channel that permits catalytic activity, mediated by a transmission loop.
• Mutations in these CAD Loops can eliminate allosteric regulation, rendering the enzyme constitutively active.

In our paper Iron-sulfur cluster deficiency can be sensed by IRP2 and regulates iron homeostasis and sensitivity to ferroptosis independent of IRP1 and FBXL5, published in Science Advances in 2021, key findings included the following:

• Inhibiting iron-sulfur cluster (ISC) synthesis triggers sensitivity to ferroptosis in cancer cells that upregulate the transferrin receptor.
• Transferrin receptor upregulation is necessary to sensitize cells to ferroptosis upon ISC inhibition.
• IRP1 is dispensable for altered iron metabolism upon ISC inhibition.
• IRP2 stability is altered upon ISC synthesis inhibition only under atmospheric oxygen conditions.
• IRP2 binding to mRNAs involved in iron metabolism is stimulated by ISC synthesis inhibition, independent of FBXL5 or changes in IRP2 stability at tissue-relevant oxygen concentrations.
• Deletion of both IRP1 and IRP2 renders cells dependent upon iron supplementation or forced expression of the transferrin receptor.
• Cells lacking both IRP1 and IRP2 are resistant to ferroptosis, even when ISC synthesis is inhibited.

In our paper Hyperactive CDK2 activity in basal-like breast cancer imposes a genome integrity liability that can be exploited by targeting DNA polymerase ε, published in Molecular Cell in 2020, key findings included the following:

• Basal-like breast cancer (BLBC) cells are especially sensitive to suppression of iron-sulfur cluster (ISC) biosynthesis, compared to other breast cancer subtypes or non-transformed mammary epithelial cells.
• The catalytic subunit of DNA polymerase epsilon (POLE1) requires ISC cofactor binding to support replication and cell viability. BLBC is selectively susceptible to partial suppression of POLE1, but not the other replicative polymerases, indicating that POLE1 is the ISC protein that underlies the sensitivity of this breast cancer subtype to suppression of ISC synthesis.
• Suppression of POLE1 results in DNA damage, expression of inflammatory cytokines, and a senescence-like state or cell death, selectively in BLBC cells.
• Non-BLBC cells and non-transformed mammary epithelial cells depend on properly regulated activity of the cyclin dependent kinase CDK2 to maintain viability when POLE1 or ISC synthesis is suppressed.
• Human BLBC tumors have increased phosphorylation of CDK2 target genes compared to other breast cancer subtypes, consistent with genetic alterations that promote CDK2 activity in BLBC.
• Hyperactive CDK2 may underlie the sensitivity of BLBC cells to ISC or POLE1 suppression, and may represent a liability for BLBC cells that can be targeted for anti-cancer therapy.

In our paper NFS1 undergoes positive selection in lung tumours and protects cells from ferroptosis, published in Nature in 2017, key findings included the following:

• Oxygen level is a key factor in understanding the response to metabolic perturbation. We describe how cancer cells in culture are more sensitive to suppression of enzymes utilizing oxygen as a substrate compared to these same cells in a tumor. These data suggest that growing cells in culture at tissue level oxygen would be one simple and important way to mimic in vivo conditions.
• Lung tumors select for a metabolic pathway that protects against damage from a nutrient in the tumor microenvironment. We demonstrate that NFS1, a critical enzyme in the biosynthesis of iron-sulfur cluster cofactors, is upregulated in lung adenocarcinoma and when cells are placed in high oxygen environments. Iron-sulfur clusters are present in numerous cell-essential proteins and are sensitive to oxidative damage.
• Physiologic changes in oxygen level affect iron-sulfur cluster proteins more than exogenous oxidants do. We describe how iron-sulfur cluster containing proteins lose activity in high oxygen, but not upon moderate stress with hydrogen peroxides. Similarly, suppressing iron-sulfur cluster cofactor production blocks proliferation, which can be rescued by low oxygen culture, but not by anti-oxidants.
• Suppressing iron-sulfur cluster biosynthesis activates the iron-starvation response and increases sensitivity to oxidative cell death, also known as ferroptosis. We show that the classic iron-starvation response (increase in iron uptake and intracellular iron release) is dramatically upregulated upon iron-sulfur cluster biosynthesis suppression. Treatment with exogenous oxidants results in a type of cell death called ferroptosis, likely because the increase in free intracellular iron potentiates oxidative damage by producing superoxide radicals. These results lead us to the hypothesis that activating the iron-starvation response can trick cancer cells into increasing iron uptake, leaving them vulnerable to death by ferroptosis.

In our paper Serine catabolism by SHMT2 is required for proper mitochondrial translation initiation and maintenance of formylmethionyl-tRNAs, published in Molecular Cell in 2018, key findings included the following:

• A mitochondrial pathway that catabolizes the amino acid serine is required to support mitochondrial function. Mitochondrial one-carbon metabolism harvests carbon units from serine for use in several metabolic reactions requiring one carbon donation, particularly in the de novo synthesis of nucleotide bases. We show that complete loss of this pathway also impairs mitochondrial oxidative phosphorylation because of a loss of mitochondrially encoded proteins. A parallel pathway of one carbon metabolism in the cytoplasm can largely make up for loss of the mitochondrial pathway by supporting nucleotide metabolism, but cannot provide sufficient one carbon units to the mitochondria to maintain its function. These findings indicate that cells maintain two parallel pathways in the mitochondrial and cytoplasm to support this critical mitochondrial function.
• Carbon units harvested from serine are used to modify mitochondrial tRNAs, supporting translation in the mitochondria. The mitochondrial genome encodes 13 proteins involved in oxidative phosphorylation, which are transcribed and translated in the mitochondria. Prokaryotes and mitochondrial ribosomes, but not eukaryotic cytoplasmic ribosomes, prefer to start translation with a modified initiator tRNA in which the methionine is formylated. We show that this formyl group originates from the amino acid serine via mitochondrial one carbon metabolism. Complete blockade of mitochondrial one carbon metabolism impairs production of formyl-methionine tRNAs, which results in the loss of translation of several mitochondrial proteins.

In our paper Metabolic determinants of cancer cell sensitivity to glucose limitation and biguanides, published in Nature in 2014, key findings included the following:

• Continuous flow cell culture can be used to maintain cells in limiting nutrient conditions for extended periods, permitting genetic screening in these conditions. We describe the development of an apparatus that we call the Nutrostat, in which cells are maintained in 1 mM glucose for weeks. These conditions permit exponential growth approaching nutrient replete levels and high-throughput screening.
• Cell-barcoding can be used to perform competition assays between cell lines in nutrient limiting conditions. We develop a technique to compare the sensitivity of cell lines to changes in the nutrient levels of the environment. Dozens of non-adherent cell lines are tagged by stable barcodes transduced lentivirally. We show that these cell lines can be grown together in large cultures to simultaneously measure the differential response of these lines to glucose limitation.
• Cancer cells grown under low glucose conditions require robust oxidative phosphorylation. Using loss-of-function genetic screening in Nutrostats, we show that genes encoding core components of oxidative phosphorylation complexes are differentially required in low glucose culture.
• Cancer cell lines sensitive to low glucose culture harbor mutations in mitochondrially-encoded proteins or exhibit low expression of genes required for glucose utilization. Using the cell line barcoding technique described above, we show that cell lines with an inability to activate oxidative phosphorylation in response to glucose limitation exhibit exquisite sensitivity to low glucose culture. These cell lines either harbor loss-of-function mutations in mitochondrially-encoded proteins, which encode core subunits of oxidative phosphorylation, or express low levels of glycolytic genes and glucose transporters.
• Cancer cell lines sensitive to low glucose culture are sensitive to biguanide treatment. Biguanides such as metformin and phenformin can inhibit mitochondrial oxidative phosphorylation Complex I. Indeed, we show that the effects of phenformin on oxygen consumption, and its toxicity on cells in culture and on xenograft tumors can be rescued by the expression of a yeast protein, Ndi1, that can bypass Complex I. Cell lines harboring mutations in mitochondrial proteins or expressing low levels of glycolytic genes and glucose transporters are particularly sensitive to phenformin treatment. These data suggest that in tumors that harbor such mutations in the mitochondrial genome, biguanide treatment may be efficacious.

In our paper Functional genomics reveal that the serine synthesis pathway is essential in breast cancer, published in Nature in 2011, key findings included the following:

• Serine synthesis is upregulated in basal-like breast cancer. We show that basal-like breast tumors exhibits high expression of the enzymes of serine biosynthesis, especially phosphoglycerate dehydrogenase (PHGDH). Cell lines derived from these tumors convert metabolites of glucose to serine, and can also import serine from the environment. Serine is consumed for several important downstream biosynthetic pathways including nucleotide metabolism, membrane lipid synthesis, and synthesis of the amino acids glycine and cysteine. PHGDH is 1 of 50 genes whose expression is used in a common classifier of breast cancer into subtypes, the PAM50, because it is highly specific for basal-like disease. Subsequent work has shown that expression of serine biosynthetic enzymes is required in a variety of tumor types to support downstream biosynthetic pathways.
• Activation of serine synthesis drives flux into branching metabolic pathways. We demonstrate that by increasing serine biosynthesis, basal-like breast cancer cells also increase flux of glutamine into the TCA cycle, as transamination by PSAT1 also results in the conversion of glutamate to alpha-ketoglutarate. In fact, in serine synthesis high cell lines, the PSAT1 reaction is responsible for about half of the flux from glutamine into the TCA cycle.
• Serine synthesis is silenced in a subset of breast tumors, particularly luminal subtypes, and cancer cell lines exhibiting silencing of these genes are dependent on exogenous serine. We show that some cell lines with low expression of PHGDH, PSAT1, or PSPH are serine auxotrophs, which are dependent upon serine uptake.