Single Cell Analysis
The cell is a complex dynamic system in which components change in time and space, with outputs in functional outcomes as simple as response to changed substrate availability, as complex as growth and differentiation. Genomics and the ability to measure parameters globally, such as transcripts and proteins, are setting the stage for a truly systems-level understanding of how the cell works. Network analysis of biological systems, whether control systems, metabolic systems, or ecosystems, suggests a framework of nodes, interlinked by a relatively small number of connections. This modular view of biological systems is supported by evolutionary studies, suggesting that functional units, or modules, have evolved within the larger system as subsystems. Such a framework is also amenable to predictive modeling, in which the complexity of the entire system can be reduced to a smaller number of subsystems and effectively modeled by inputs and outputs.
The Achilles heel of systems biology is the reliance on population-based analysis. All global analysis currently depends on using large numbers of cells and averaging measured parameters. This approach has the advantage of averaging out natural fluctuation, but has the disadvantage of covering up the presence of functionally-important subpopulations. Emerging results suggest that such subpopulations are a fundamental property of prokaryotic populations. A basic assumption exists in microbial biology, that exponentially-growing populations of isogenic bacteria are roughly the same physiologically, given some Poisson-like distribution. However, a growing body of evidence suggests that this assumption is not correct. In fact, it appears that bacterial populations continually generate substantial physiological diversity at the individual cell level in fundamental characteristics such as translation rate and growth rate (1-19). At the ends of the spectrum of variability, the cells in these physiological states are significantly different from the population average. Current data suggest this diversity is due to stochastic, or random mechanisms. In addition, it appears that such cells are poised to respond differently to sudden change, such as stress or nutrient addition. In some cases, this difference can be dramatic, such that a minor subpopulation can dominate population behavior (Fig. 1).
The Lidstrom group has gathered data on Methylobacterium extroquens AM1 (a facultative methylotroph) at the single-cell level, and we have shown that in this organism not only do both gene expression and growth rate vary by a factor of 2-3 in exponentially-growing populations, the two parameters do not correlate with each other. Therefore, whatever the mechanisms are for generating this heterogeneity, they are distinct.
The implications of this growing body of data are significant for our fundamental mechanistic understanding of the cell as a complex dynamic system, as well as understanding behavior of natural populations of bacteria. In addition, analysis of single-cells is the first step to understanding cell-cell interactions in populations. Separating responses of cells into those that are cell-specific (not involving cell-cell communication) and those that are population-specific (involving cell-cell communication) requires single-cell analysis.
Although many elegant studies have been published recently focused on stochasticity or noise in gene expression (e.g. 3-11,15,16), the grand challenge in this nascent field of single cell heterogeneity is relating stochasticity (cell-to-cell variation) to phenotype. Since the cell functions as a complex, dynamic, and interconnected system, it is important to understand phenotype at the whole organism level, in order to make the link between cell-to-cell variation and subsequent cellular and population response. The major barrier inhibiting progress on understanding how cell-to-cell physiological variation impacts cellular outcome, and how that in turn impacts population response, is technological. It is essential to measure both gene expression and phenotypic parameters in individual living cells in real time, in high throughput, over multiple growth cycles, and until recently, that capability did not exist.
The NHGRI-funded Center of Excellence in Genomic Sciences, the Microscale Life Sciences Center at the University of Washington, has developed core technology for single-cell analysis of mammalian cells (see Fig. 2) based on video microscopy that incorporates multi-color fluorescence detection, a 99-well glass chip platform that allows isolation and observation of single cells, a sensor system for measuring respiration rates, and off-chip analyses such as single-cell proteomics by 2D capillary electrophoresis and qRT-PCR of single cells. For the first time, this technology has provided the means to measure both gene expression and multiple phenotypic parameters in single mammalian cells. This platform has the potential to be adapted to the study of prokaryotic cells with further technology development. By judicious selection of gene fusions and dyes used, targeting key events and physiological outcomes, important components of the cell system can be mapped out in 1000s of individual cells before and during response to change.
Fig. 2. Schematic of the MLSC well-based platform for single cell analysis. The movable lid allows intermittent closure to detect O2 consumption via a sensor in the wells, then opening for flow-through incubation.
In addition, this type of technology can also be applied to cell-cell interaction studies. This technology will provide the ability to spatially locate individual cells at specific distances and within specific fluid volumes with regard to other cells of either the same or different types, and obtain gene expression and physiological data on either individual cells or the aggregate group of cells.
Research. The nascent field of single-cell analysis has the potential to revolutionize our understanding of prokaryotic cells and populations. For the first time, it offers the possibility to study complex events in prokaryotes at the biologically-relevant level, the individual cell, and link genotype to phenotype. In addition, it opens up an entirely new field of study, physiological heterogeneity in populations. Although at this time, technology is restricted to measuring a few parameters, advances in single molecule analysis, bar-coded nanobeads, and polony sequencing suggest the future will allow genome-level analysis of individual cells at the transcript, protein, and DNA level. The key is to adapt an automated platform such as that being developed in the MLSC to these new measurement techniques. The need to address important fundamental questions at the single-cell level will drive such technology development.
In the Lidstrom Lab, we are using Methylobacterium extorquens AM1 as a model system to study the link between heterogeneity in gene expression and heterogeneity in physiological outcomes, such as growth rate, respiration rate, and excretion of metabolic end products. We are using single cell analysis technology to assess the role of physiologically-distinct subpopulations in response to changing conditions, which in this case are focused on response to formaldehyde stress.
1. Maloney, P. C. & Rotman, B. Distribution of suboptimally induces beta-D-galactosidase in Escherichia coli. The enzyme content of individual cells. J. Mol. Biol. 73, 77–91 (1973).
2. Siegele, D.A., & Hu, J.C. Gene expression from plasmids containing the araBAD promoter at subsaturating inducer concentrations represents mixed populations. Proc Natl Acad Sci U S A 94, 8168-72 (1997).
3. Fraser HB, Hirsh AE, Giaever G, Kumm J, Eisen MB.2004. Noise minimization in eukaryotic gene expression. PLoS Biol. 4 Jun;2(6):e137.
4. Rosenfeld N, Young JW, Alon U, Swain PS, Elowitz MB. 2005. Gene regulation at the single-cell level. Science. 2005 Mar 25;307(5717):1962-5.
5. Pedraza, J. M. & van Oudenaarden, A. 2005. Noise propagation in gene networks. Science 307:1965–1969.
6. Colman-Lerner A, Gordon A, Serra E, Chin T, Resnekov O, Endy D, Pesce CG, Brent R. 2005. Regulated cell-to-cell variation in a cell-fate decision system. Nature 437(7059):699-706.
7. Kaern M, Elston TC, Blake WJ, Collins JJ. 2005. Stochasticity in gene expression: from theories to phenotypes. Nat Rev Genet. 6(6):451-64.
8. Le, T.T., Harlepp, S, Guet CC, Dittmar K, Emonet T, Pan T, Cluzel P. 2005. Real-time RNA profiling within a single bacterium. Proc Natl Acad Sci U S A. 28;102(26):9160-4.
9. Golding I, Paulsson J, Zawilski SM, Cox EC. 2005. Real-time kinetics of gene activity in individual bacteria. Cell. 123(6):1025-36.
10. Becskei A, Kaufmann BB, van Oudenaarden A. 2005. Contributions of low molecule number and chromosomal positioning to stochastic gene expression. Nat Genet. 37(9):937-44.
11. Kuang Y, Biran I, Walt DR. 2004. Simultaneously monitoring gene expression kinetics and genetic noise in single cells by optical well arrays. Anal Chem. 76(21):6282-6.
12. Kelly, C.D. & O. Rahn. 1932. The growth rate of individual bacterial cells. J. Bacteriol. 23:147-153.
13. Umehara S, Wakamoto Y, Inoue I, Yasuda K. 2003. On-chip single-cell microcultivation assay for monitoring environmental effects on isolated cells. Biochem Biophys Res Commun. 2003 Jun 6;305(3):534-40.
14. Booth IR. 2002. Stress and the single cell: intrapopulation diversity is a mechanism to ensure survival upon exposure to stress. Int J Food Microbiol. 78(1-2):19-30.
15. Thattai M, van Oudenaarden A. 2004. Stochastic gene expression in fluctuating environments. Genetics. 167(1):523-30.
16. Aertsen A, Michiels CW. 2005. Diversify or die: generation of diversity in response to stress. Crit Rev Microbiol. 2005;31(2):69-78.
17. Balaban NQ, Merrin J, Chait R, Kowalik L, Leibler S. Bacterial persistence as a phenotypic switch. Science. 2004 Sep 10;305(5690):1622-5. Epub 2004 Aug 12.
18. Kussell E, Kishony R, Balaban NQ, Leibler S. Bacterial persistence: a model of survival in changing environments. Genetics. 2005 Apr;169(4):1807-14. Epub 2005 Jan 31.
19. Kussel, E., and S. Leibler. 2005. Phenotypic diversity, population growth, and information in fluctuating environments. Science 309:2075-8.
20. Lidstrom, M. and D. Meldrum. Life on a Chip. 2003. Nat Rev Microbiol. 1(2):158-64.