Speaker: Christine Jacobs-Wagner
Department: Microbial Sciences Institute, Yale West Campus
Subject: How to achieve cellular replication without fail: lessons from bacterial cells.
Location: TU Delft (BN Seminar)
Author: Nemo Andrea
The topic of today’s talk was cellular replication, which, in Christine’s opinion, is the ability that separates the living from the non-living. In order to study this process, they study bacterial replication, as bacteria both divide rapidly and do so with high accuracy. The speaker stressed how, if one stops to think about it, achieving such short (20 minutes) division times in a varying environment with in the inherently highly stochastic environment of a living cell, is a truly remarkable feat. Thus, the robustness of bacterial replication and the relative simplicity of prokaryotic systems as compared to eukaryotic model systems make bacteria the candidate of choice for her research.
As there are many topics that can be explored within cellular replication, the speaker decided to focus on a single system that her group had recently done work on. Many bacteria have plasmids, which can contain vital functions for a bacterium, thus requiring efficient segregation of plasmids upon division. While plasmids diffuse throughout the cell and are thus, on average, equally divided among the two halves of a bacterium, suggesting that in the case of cell division no problem should arise, this is not the case. If a plasmid is present in high copy number, the chance of one daughter cell having significantly fewer copies of the plasmid is very small, but if a plasmid has low copy number the probability of one daughter cell ending up with no copies due to low number noise becomes significant. It is thus that bacteria have developed methods to effectively separate the plasmids that are present in low numbers. One could argue that it is not the bacterium driving the evolution of such a system but rather the plasmid itself, as plasmids that effectively do this ensure their survival, but that is really more a matter of perspective than anything else, and beside the point argued in this talk.
The plasmid that the group focused on displayed very curious behavior. The plasmids (in elongated bacteria) are distributed equidistantly along the long axis. The plasmids do diffuse, but stay in roughly the same area over time. It should be apparent that if such a distribution is maintained, the plasmids will be divided equally among daughter cells. As it turns out, this is achieved through a particular variant of the Par system. This variant uses just two proteins: PAR A/B, which are encoded on the plasmid itself. ParB binds to the plasmid, and does this by recognizing specific sequences on the plasmid, ensuring selective binding. ParA, on the other hand, binds to ParB, after which ParB will stimulate the ATPase activity of ParA, which will then unbind from the DNA. ParA also unspecifically binds to the DNA (of the bacterium). They observed (in case of a single plasmid) an oscillation of ParA from one side of the cell to the other, with the plasmid (with ParB) following the ParA signal. They then produced a first model, to test if simple Brownian dynamics of a diffusing plasmid could reproduce such behavior. They found that this was insufficient and concluded from this that a translocating force must be present to create this behavior.
The key insight that was missing from the first simulation was the fact that DNA is not static, but moves around in the cell. They determined that the movement of a locus in the DNA can essentially be seen as an elastic harmonic potential, with a force of around 0.04pN. This is a really small force, even by cellular standards. They redid the simulations with this idea incorporated and found that they were able to reproduce the behavior observed in bacteria. The conceptual change is now as follows: The plasmid with ParB is pulled around through its multiple connections with multiple harmonic potentials through ParA. It will then unbind from the ParA in that area and move in a direction (slightly). In the new situation, the plasmid will ‘see’ more ParA in the direction of movement than behind it (which is the previous location, where it made ParA unbind) essentially creating a gradient of forces in the direction of movement. The speed of the plasmid seemed to correspond well with the force of the harmonic potentials. The initial symmetry breaking is caused by the stochastic nature of ATP hydrolysis.
The speaker mentioned another recent model that was developed that may explain similar phenomena, but pointed out that certain parts of this specific system are most likely not reconcilable with that model. The model presented today was also extremely elegant requiring just two components; one that binds to chromosomes, and another that modulates that first component’s lifetime. I quite enjoyed the talk, as Christine was an excellent presenter. While I am generally not easily convinced by simulations, this model seems to make sense both mechanistically and matches real world experiments. We also had a brief session afterwards to ask some questions, which was cut short due to unforeseen circumstances. I do appreciate her taking the time to sit down with us and answer some questions – even the ones not related to the talk.