Posts Tagged ‘positional information’

Only accessible information is useful: insights from gradient-mediated patterning

Mikhail Tikhonov, Shawn C. Little, and Thomas Gregor. Roy. Soc. Open Sci. 2: 150486 (2015). (more…)

Positional information, positional error, and read-out precision in morphogenesis: a mathematical framework

Gasper Tkačik, Julien O. Dubuis, Mariela D. Petkova, and Thomas Gregor. Genetics 199 (1): 39–59 (2015).  (more…)

Positional information, in bits

Julien O. Dubuis, Gasper Tkacik, Eric F. Wieschaus, Thomas Gregor and William Bialek, PNAS 110, 16301-16308 (2013).

Current projects

The following paragraphs summarize the projects that are currently progressing in the lab to contribute to the execution of the program outlined above:

  1. Originally instigated in the summer of 2009 by an undergraduate project, we developed a novel mRNA quantification method in fly embryos, the core concept of which was published last year as a proof of principle. We essentially tag individual mRNA particles with fluorescent probes in fixed tissue and use diffraction-limited confocal microscopy to image and count. The real challenge of this method, however, was to convince ourselves that we are indeed looking at individual molecules (and not spots or aggregates of molecules) and that we can indeed count all of them. A further challenge was to go beyond simple spot identification, and to quantify the actual intensities that we obtain per spot. Intensities at the DNA location of active native transcription reveal the actual transcriptional state of that site at the moment of specimen fixation. Using multiple colors on the same site we will be able to even extract dynamic parameters such as binding rates and lifetimes from our data. We have resolved most of these challenges over the last year in a very talented team composed of a postdoc, a graduate student and an undergraduate. Using this powerful method and some of the other tools currently developed in the laboratory I am convinced that we will be able to reveal many of the molecular underpinnings of transcriptional regulation in the early fly embryo over the next few years.
  2. The wealth of approaches developed in molecular biology over the past decades gives us the luxury in this field to do the same experiment twice, but with a completely different set of systematic errors. Consequently, over the past two years we have developed an independent way to measure the mean number of mRNA molecules in single embryos. Our method exploits a combination of polymerase-chain-reactions (PCR) with techniques typically used by biochemists, and by making sure at each step we understand what we do to our absolute numbers of initial mRNA molecules. The key is then to compare the numbers we get from single embryos with a precisely quantified molecular standard. The means of the two methods match within less than either of the two standard deviations.
  3. Again in the spirit of an alternative approach to the same problem, we are also developing a direct handle on transcriptional dynamics by directly labeling mRNA strands with fluorescent proteins. This technique has been pioneered in bacteria to trace individual mRNA molecules and in fly embryos to trace mRNA aggregates. We are extending this application to reveal the dynamics of nascent mRNA molecules as they are assembled at the DNA transcription site in living fly embryos. By simultaneously measuring fluorescently labeled input transcription factors (as we have done in the past) and mRNA output dynamics we will be able to extract dynamic input-output relations of transcription in a natural, which has never been done in any biological system. We thus will have a fine tool at hand to put models of transcriptional regulation to a real quantitative test.
  4. We are collaborating with Bill Bialek (Princeton/Physics) and Eric Wieschaus (Princeton/Molecular Biology) to measure the amount of information transmitted through the gene network, testing the idea whether optimization principles govern and apply during early fly patterning, i.e. is there an optimal matching between the distribution of transcription factor concentrations and the noise levels in the control mechanism? What would be the structure of of genetic regulatory networks that maximize information transmission, if they optimize their regulatory power and precision while using a limited number of molecules? We are currently measuring information that is transmitted through the gene network that controls the early patterning events in the fly embryo, and initial results are about to be disclosed in full.
  5. Perturbation experiments have a long tradition in both the physical and biological sciences to understand complex processes and networks. In biology, however, it is customary to perform such perturbations in terms of a mutation analyses which result in the partial or complete disruption of the function of a particular protein. In reality, such a mutation is much more than a simple perturbation, as it leads to a transformation of the original network to a new and completely different network that has one less node. To avoid such a disruptive perturbation, my laboratory has developed a system in which we have a dial knob for the network??s input concentration. We genetically generated a set of fly lines that express the input protein Bicoid at dosages that differ from its wild type expression level in increments of 10-20% up to factors of 2.5-3 in either direction. We are currently in the process to carefully measure the response of the network to these subtle absolute input concentration changes. Our preliminary results indicate that the traditional picture of a local concentration readout may no longer be in agreement with our observations.

The role of input noise in transcriptional regulation.

G. Tkačik, T. Gregor, W. Bialek, PLoS One 3, e2774 (2008).


Probing the limits to positional information.

T. Gregor, E. F. Wieschaus, D. W. Tank, W. Bialek, Cell 130, 153-164 (2007).