Biology Microscale Life Sciences Center
Dissect the mechanism involved in live/die decisions at the individual cell level
In this goal we use two model systems: proinflammatory cell death (pyroptosis), a condition underlying heart disease and stroke; and neoplastic progression of Barrett’s esophagus (BE), a precancerous condition with a 0.5-1% progression rate.
- Measure physiological, genomic, proteomic and transcriptomic parameters in single cells.
- Assess response to stimuli involved in live/die decisions via real-time multiparameter measurements.
We have now compiled an expanding database of thousands of single-cell respiration rate measurements as a baseline for comparison to cells of different type, neoplastic progression stage and exposure to varying stimuli. We report our investigations into the response of Barrett’s Esophagus (BE) single cells containing different genotypes (with and without the tumor suppressor gene CDKN2A/p16 and combined CDKN2A/p16,TP53 abnormalities) representative of clonal populations in BE patients in vivo, to factors known to increase risk for neoplastic progression of premalignant BE cells (e.g., acid, bile, gastrin) or known to protect from neoplastic progression (e.g., nonsteroidal anti-inflammatory drugs, selenium), by serially measuring physiological states (respiration, cell death, cell division and DNA content). We have also focused on defining physiological heterogeneity in resting macrophage populations. We are correlating our measurements of cell respiration rates to resistance or susceptibility to proinflammatory cell death, and quantifying the temporal ordering of physiological parameters during pyroptosis. We are using this data to identify common pathways. We are examining how heterogeneity at the single cell level affects development of resistance to hypoxia, a common selective pressure that occurs during neoplastic progression. We have generated hypoxia resistant BE cell lines from patients having different somatic genetic backgrounds. We are currently characterizing the heterogeneity at the genetic and metabolic levels in these resistant lines to better understand the relationship between population heterogeneity and risk of cancer progression.
- There are significant variations in oxygen consumption rates (OCR) among populations of CP-A (metaplastic BE) and CP-C (dysplastic BE) cells, differing by as much as a factor of six among individual cells.
- Measurements using CP-A cells in G1 phase of the cell cycle revealed that the observed heterogeneity in respiration rates cannot be explained by differences in cell cycle phase.
- We observed a relatively small subpopulation of fast respiring CP-A cells, which implies that at least two subpopulations of CP-A cells exist with different levels of metabolic activity.
- OCR heterogeneity is less pronounced in slow respiring cells compared to fast respiring cells.
Single-cell gene transcription level profiling
- We observe marked variability in the transcription levels of hypoxia-related genes in response to acute (30 minutes) hypoxia in metaplastic (CP-A) and dysplastic (CP-C) BE epithelial cells.
- Only a fraction of studied individual CP-A and CP-C cells exhibited increased levels of hypoxia response genes in response to acute hypoxia, whereas other cells did not show upregulated levels of the corresponding genes.
- CP-A cells have lower mtDNA copy numbers and show significantly less heterogeneity than CP-C cells.
- Mitochondrial membrane potential (MMP) levels of CP-A cells are lower than CP-C cells under both normal and hypoxic conditions, indicating more loss of membrane integrity and leakage of apoptotic factors under hypoxic stress in CP-A cells.
- 16s rRNA and coxI genes were significantly upregulated in hypoxic CP-A cells, which may reflect more active mitochondria to produce ROS, especially more H2O2 leaking into the cytosol via electron transfer complex IV.
- More hypoxia response genes were upregulated in hypoxic CP-A compared to CP-C cells, which may result from more ROS produced by mitochondria under hypoxia. In contrast, CP-C cells were much more resistant to hypoxia since less ROS was produced by mitochondria under hypoxia.
- Caspase-1-dependent pore formation executes osmotic lysis during terminal stages of pyroptosis; temporal events in the pathway were discovered and resolved by studying single cells.
- Endonuclease activation during pyroptosis results in DNA cleavage and is independent of the canonical caspase activated DNase involved in apoptosis.
- Pathological ion fluxes are inhibited by physiological concentrations of glycine; permitted observation of intact cells undergoing terminal events in pathway.
- Pore formation during pyroptosis has molecular dimensions of 1.1-2.4 nm and results in Ca++ fluxes.
- Diverse upstream signaling pathways lead to a conserved downstream mechanism of pyroptosis.
- Caspase-1-dependent Ca++ fluxes result in lysosomal exocytosis; independently caspase-1 is responsible for maturation and release of inflammatory cytokines IL-1b and IL-18.
- Coordinated secretory events during pyroptosis include the release of PAMPs that we demonstrated can activate cells and render resistant cells susceptible to pyroptosis; partially degraded antigens that can prime immune responses; antimicrobial compounds, release of lysosomal contents co-temporally with IL-1b and IL-18 by an independent mechanism; lysis of the cell is also often host beneficial by eliminating intracellular niches of propagating microbes.
Increase the information content of single-cell datasets and develop a deeper understanding of disease pathways
Toward this goal we developed novel technology to interface with the Living Cell Array analysis platform to increase the richness of single-cell datasets to gain more detailed insight into disease states.
- Develop and apply new intracellular fluorescent sensors that will simultaneously measure intracellular pH, ions, and biomolecules and extracellular optical sensors for oxygen consumption rates, cytokine production, and ion and protein concentration and fluxes.
- Develop loss of heterozygosity (LOH) and transcriptomic measurement capabilities.
- Develop module interconnects to enable real-time (physiological) and post-fixation (genomic, transcriptomic and proteomic) measurements on the same cell.
We have significantly transitioned from a technology development mode to a data gathering and biological inquiry and understanding mode. The refinement of our primary instrumentation, the Living Cell Array (LCA) analysis platform, was substantial. Three continuously operating LCA platforms were brought on line to enable physiological observations of single cells and groups of cells. Development of LCA-driven technology, sensors and single-cell genomic and transcriptomic methods continues to advance and improve the richness and quality of our biological observations by the introduction of multiple new sensors (intracellular and extracellular), detection methods (optical and electrical), cell handling instruments (array loading and harvesting), and stimulus control and delivery mechanisms.
Apply technologies to study cells in isolation, cells interacting with other cells, and cells tissues
Toward this goal we apply the new capabilities to the same two biological models (esophageal cancer and inflammation) to study cells in isolation (Phase I), in groups (Phase II), and in tissue (Phase III), to broaden our understanding of underlying mechanisms.
We have built a physiological response data set (respiration rates) of thousands of single cells and of numerous cell types. Thanks to an ARRA supplement award we were able to add resources to our CEGS team specifically to study the effects of cell-cell interactions on these observations. While analysis of cells in tissues (Phase III) remain a future goal, we have developed methods and sensor delivery approaches this year that enable the extension of the LCA to much higher throughput (order of magnitude increase) on larger groups of cells, and potentially to tissue preparations and real time in vivo devices.
We have quantified a non-linear increase in oxygen consumption rates of individual cells as a function of the number of interacting cells. Three interacting CP-A cells showed most significant (4.4 times) increase in OCR compared with single, non-interacting cells of the same type. However, bulk cell measurements with as many as 50,000 CP-A cells revealed average (per-cell) OCR rates similar to those of single cells.