Protein Diffusion from a DNA Point Source

by Mitchell Wendt & Alexander Engel

Introduction

The Noireaux laboratory has developed a method for in vitro (in glass) transcription and translation. We wanted to use this method to explore the diffusion of proteins when they are being constantly produced, as this protein diffusion is important in biological systems. To simulate a point source, we attached plasmid DNA that codes for GFP (Green Fluorescent Protein) to 10μm diameter microbeads. GFP is a protein originally part of jellyfish that emits green light (λ=509nm) when exposed to 488nm excitation light. Using fluorescence microscopy and time-lapse photography, we could then track the production of GFP and measure its diffusion from the beads.

Above: DNA attached to a microbead [Source: 3]

Theory

Diffusion is the stochastic movement of particles in solution due to collisions with solvent molecules. Mathematically, it can be described with Fick’s second law, which relates the time derivative of a particle’s concentration Ï• to its spatial derivative with the diffusion constant D, which can be calculated with the Stokes-Einstein relation. This gives the value for spherical particles in terms of the Boltzmann constant k_B, absolute temperature T, radius of the particle r, and viscosity of the solution η. For GFP with a radius of 2.82nm, D_GFP = 87μm²/s.

Methods

The Noireaux laboratory provided us with an in vitro reaction solution that we combined with DNA-tagged beads and observed with an Olympus IX-81 fluorescence microscope. The figure below shows the biological processes that occur in the experiment.

Fluorescence microscopy uses an excitation laser that causes fluorescent molecules to emit light. The emitted light is then transmitted to a camera. Our experiment involved combining the in vitro reaction solution with microbeads coated with a plasmid coding for GFP on a slide and measuring the brightness observed when GFP was produced. We allowed oxygen to reach the solution by using PDMS to cover the solution. This also ensured that the solution was in a very thin layer, which allowed us to ignore the effects of diffusion in the vertical direction. To prevent photobleaching (the loss of fluorescence due to overexposure to excitation light), we narrowed our excitation light to a thin slit and positioned a bead within it. Data was collected at the rate of one photo per minute for 40 minutes.

Above: depiction of how transcription and translation occur to create proteins [Source: 2]

Above: a couple images of microbeads taken with the microscope. The one on the left has just plain beads, which still have some visibility despite the fluorescence filters. The one on the right has beads which have DNA attached and are in a reaction solution.

Analysis and Results

We had difficulty obtaining consistent success with the reaction itself (only 1 reaction produced a measurable signal). We used !ImageJ and wrote a Java program to measure the pixel intensities of the photos along the full excitation slit and compared them to calibrations based on known concentrations of GFP. We then applied a discrete approximation of Fick’s Second Law using the intensity changes at each position within the slit for each time data was taken. To reduce the effect of random fluctuations between pixels, we binned the pixel data. Using a weighted average of diffusion coefficients calculated for each bin, we calculated a D_GFP = 49±10μm²/s, a 4σ deviation from the expected value; various systematic errors were likely present. Also, using the calibrated brightness of known GFP concentrations, we calculated the approximate quantity of GFP at each position and time. With this, we determined that the in vitro reaction was producing GFP at a rate of 137 ± 27 proteins/sec (but we were not able to determine how many DNA strands were attached to the microbead).

Above: two images of fluorescence viewed through the thin slit. The top one is of a solution of 5 μM GFP while the bottom one is of the main reaction in which the bead with DNA attached, at the left, produces GFP which then diffuses away. The features that are due to the slit being non-uniform are the same in each image, and thus can be removed by comparing the two images on a pixel-by-pixel basis.

Above: graph of some of the data obtained; the concentration of GFP with respect to the distance from the source (bead) in a reaction run. Three different times are plotted, with time zero being the beginning of data acquisition.

Above: an image of the java program that was written to efficiently analyze the data. !ImageJ was also used, but the full analysis of the set of reaction images involved specifics that cannot be simply done in !ImageJ (and would be cumbersome to perform in Excel).

Conclusion

Though the in vitro reaction did not occur or produced an undetectable (too small) quantity of GFP in the majority of our trials, we were able to clearly observe its efficacy in a few trials. Furthermore, the image analysis process was highly variable and difficult due to artifacts present in our images and a high degree of background noise. However, we were able to extract a GFP diffusion coefficient from the image analysis.

Future research could use Fluorescence Fluctuation Spectroscopy to more accurately measure the presence of GFP to find a more accurate diffusion coefficient. The errors in our results would also be greatly reduced if the GFP production rate could be increased.

References and Acknowledgements

We would like to acknowledge our advisers Dr. Vincent Noireaux and Dr. Elias Puchner for their assistance and support throughout the experiment. We would also like to thank Jon Garamella for his help in preparing the biological samples we used.

References (first one is the main paper for the DNA reaction in vitro)

• 1. Noireaux, V., Bar-Ziv, R., Libchaber, A. Principles of cell-free genetic circuit assembly. Proc. Nat. Acad. Sci. USA 100, 12672-12677 (2003). Retrieved February 22, 2015, fromhttp://www.noireauxlab.org/html pages/docs website/publications/Noireaux, Bar-ziv, Libchaber 2003.pdf

  2. "Ghost in Your Genes." PBS. Nova Online, n.d. Web. 08 Mar. 2015. Retrieved March 8, 2015, from http://www.pbs.org/wgbh/nova/education/activities/3413_genes.html

  3. Darios et al., PNAS, 2010, vol. 107, 18197–18201. Retrieved March 8, 2015, from http://pure.rhul.ac.uk/portal/en/persons/mikhail-soloviev(2b06f036-2cc0-4375-87e1-57f662534d0d).html