'Identical' cells? Not so much

Genetically identical cells may be far more different than previously believed. Published this week in linkurl:Science,; researchers find striking variation in levels of gene expression among individual, genetically identical E. coli, seemingly the result of simple chance. "The paper is quite rich," said linkurl:Sanjay Tyagi,; a molecular biologist at New Jersey Medical School who was not involved in the research. "People thi

By | July 29, 2010

Genetically identical cells may be far more different than previously believed. Published this week in linkurl:Science,; researchers find striking variation in levels of gene expression among individual, genetically identical E. coli, seemingly the result of simple chance. "The paper is quite rich," said linkurl:Sanjay Tyagi,; a molecular biologist at New Jersey Medical School who was not involved in the research. "People think that if an organism has a particular genotype, it determines its phenotype -- that there's a one-to-one relationship," said Tyagi. "But as it turns out, [differences in gene expression] can arise just from chance."
Microfluidic device allows multiplex
imaging of library strains.

Image courtesy of Yuichi Taniguchi,
Paul Choi, Gene-wei Li, and Huiyi Chen,
Harvard University
In traditional gene expression studies, researchers grind up a population of cells, then identify overall amounts of gene products from the resulting mixture. Researchers at Harvard University instead studied cells one by one, still calculating averages but also capturing variation in the population with single molecule sensitivity -- and found cells expressing genes at wildly different levels. "It's single molecules meet systems biology," said linkurl:Sunney Xie,; senior author on the paper and a chemical biologist at Harvard University. Xie's team, along with collaborators at the University of Toronto in Canada, tagged 1018 genes -- about one-fourth of the E. coli genome -- with fluorescent labels, then counted protein and mRNA copies in individual cells using a high-throughput system. They found that mRNA and protein copy numbers vary greatly from cell to cell, what researchers call "noise." Genes were being expressed at different levels, including some not at all: At any given moment, a fraction of cells didn't have a single molecule of mRNA or protein from a given gene, and a surprising subset of genes -- more than 20 percent of those analyzed -- expressed one or fewer copies of protein per cell. The ability to measure with this kind of single-molecule sensitivity is valuable for single-cell studies, said Xie. "We are able to characterize the protein distribution at every expression level," from single molecules to thousands of proteins, he added. So what causes the variation among genetically identical cells? Chance, researchers said. A single E. coli cell typically contains only a few copies of transcribing complexes at a given moment, resulting in a random synthesis of mRNA and proteins. So whether or not a given gene gets transcribed is a simple numbers game: "It's a mathematical outcome, like throwing a coin," said Tyagi. The researchers also found that, at any moment in time, there is no correlation between a single bacterial cell's protein and mRNA copy numbers, defying the long-held expectation that mRNA levels should correlate with the protein levels in the same cell. Others have showed a similar lack of correlation in populations of yeast. Several factors may be at play, Tyagi writes in an accompanying perspective in Science. First, mRNA have a short lifespan, often only minutes, while proteins are long-lived, sometimes for hours. In addition, because of the short lifespan of an E. coli, dividing every 30 minutes, long-lasting proteins are randomly passed along to daughter cells, so unequal inheritance of proteins may likely lead to some of the noise in bacterial cells. Because mRNA and protein copy numbers don't correlate, scientists need to be careful when attempting to infer protein number from mRNA quantities in a cell. "Single-cell transcriptomes don't tell you what's going on," said Xie. And counting mRNA in a single cell only offers a snapshot, a view of a cell at a moment in time, he added. Scientists can better describe gene expression by examining cells over time. "This is a cautionary note for those who want to do single-cell analysis," said Xie. Y. Taniguchi. "Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells," Science, 329:533-8, 2010.
**__Related stories:__***linkurl:Surpassing the Law of Averages;
[1st September 2009]*linkurl:Lac on, lac off;
[1st June 2009]*linkurl:Seeing single protein production;
[16th March 2006]


Avatar of: eve barak

eve barak

Posts: 85

July 29, 2010

Not surprising to me.\n\nIf you think this is amazing, try looking at "identical" cells from the same tissue of a eukaryotic organism. And then try adding the additional dimension of time! \n\nSome results are obvious -- but the experiments are very much worth doing nonetheless. We need to understand the mechanisms that are at play.\n\nFodder for the "systems analysis" folks!\n\n
Avatar of: anonymous poster

anonymous poster

Posts: 2

July 29, 2010

The title of the original article is fundamentally misleading because of semantic mismatch. Escherichia coli is a species of bacterium, but each cell is also a separate organism, with its own history. Therefore, E. coli cells may be no more closely related than the cells of any two organisms of other species (or even different genera). Moreover, the recent history of each colony may drastically alter gene expression. A zebu cow and a Holstein bull might be equally relevant as comparisons to different E. coli cells/organism for testing for identical gene expression. Or humans vs. spider monkeys. \nDeep Eddy
Avatar of: Vinod Nikhra

Vinod Nikhra

Posts: 48

July 29, 2010

It appears that ubiquitin proteasome system (UPS)works as an effective biological brake system. It is as necessary to control and stream-line cell genesis as is to stimulate cell division and growth. \nFurther, no two cells even from the same tissue may be similar in every aspect. The may differ in their age and cell-age-markers like telomere length etc. Thus, biologically different they are, they carry their difference in their physiological properties. \nWe have lot to understand in the arena of inhibitory controls in the biological systems.\n\nVinod Nikhra, M.D.\

July 30, 2010

These results highlight the need to think of a 'field-like' approach to biological regulation. The main player of biological regulation is the tissue and the cell population not the single cell, it is like physical phenomena like magnetism where an ensemble of basically random agents (the single dipoles) acquire a structuring at the global population level. This 'global scale' organization is very robust (the entire genome-wide mRNA profile of two samples of the same tissue reaches correlation levels near to unity spanning thirty thousands genes) and largely independent of details like expected in the case of collective phenomena and ends up into a remarkable low number of different profiles (the observed different tissues/ gene expression profiles are very few)so pointing to very strict constraints acting on possible 'stable states' of tissues that can be considered as dynamical systems wandering in a rugged landscape.\nThis tissue-dynamics perspective, envisaged by Waddington in the fifties and now confirmed in a lot of experimental works, asks for a complete change in perspective when studying biological systems.
Avatar of: Steven S. Clark

Steven S. Clark

Posts: 3

July 30, 2010

So, why is it I can sign on to post a comment, but am unable to read the comments already posted?
Avatar of: anonymous poster

anonymous poster

Posts: 2

August 3, 2010

Were the E. coli bacterial strains used grown as synchronous cultures? (not mentioned)...\nEven if this were the case, the total time for the analysis was 40 minutes for the 96 strains tested.\nCould not the expected variation of gene expression during various phases of a 30-minute-long bacterial cell cycle degrade the scientific worth of individual cell analyses and later summations? Have I missed something here?..\n\njust curious\n\n
Avatar of: Shi Liu

Shi Liu

Posts: 32

August 6, 2010

I do not know why my comments were repeatedly deleted. Thus I am posting a shortened comment the third time.\nI think this new study suffered the same methodological mistake as I have pointed out before. Thus it generated confusing results that compromised the conclusions. Some of the conclusions are actually wrong. Because I am writing a detailed Technical Comment on this so I will not disclose further information here.\nBut I have saved copies of my comments and the web pages of this news each time I post a comment and each time my comment was deleted, just in case someone want to see them in the future.\n

Follow The Scientist

icon-facebook icon-linkedin icon-twitter icon-vimeo icon-youtube

Stay Connected with The Scientist

  • icon-facebook The Scientist Magazine
  • icon-facebook The Scientist Careers
  • icon-facebook Neuroscience Research Techniques
  • icon-facebook Genetic Research Techniques
  • icon-facebook Cell Culture Techniques
  • icon-facebook Microbiology and Immunology
  • icon-facebook Cancer Research and Technology
  • icon-facebook Stem Cell and Regenerative Science
Life Technologies