Research interests

Community evolution

How can communities evolve? Can we design an experiment where a constiutent organism is forced to evolve in a way that is beneficial to the community, but not necessarily to itself? I am conducting experiments at the Max Planck Institute for Evolutionary Biology to address these questions with a yeast-bacterial community consisting of Saccharomyces cerevisiae and Lactococcus lactis.

Bacterial colonization of plants

It is possible to use microbial communities to make traits of the host plant more desirable. This is usually done by coating the seeds of the plants with the desired microbial community. However, it is very hard to know how many bacteria actually end up on the plants. This is a glaring gap in knowledge that could be vital to improving plant traits. At the Max Planck Institute for Evolutionary Biology, I am using barcoded bacteria to quantify what fraction and diversity of bacteria on the seeds end up on the leaves and how these change as the plants grow.

Effect of biotic and abiotic factors on community assembly

Is it possible to distinguish the impact of biotic and abiotic factors in the assembly of microbial communities? To answer this question, we grew a complex bacterial community derived from soil with and without a lab strain of alga, Chlamydomonas reinhardtii. These communities were grown for different growth periods and serially diluted into fresh medium for 10 cycles. We found that at the shorter growth periods, the algae has no influece on community assembly, whereas at the larger dilutions, the algal impact is strong - communities with the algae converge taxonomically. This convergence occurs due to loss of certain taxa, rather than the recruitment. We also showed that the ability of bacteria to consume algal secretions can determine their dynamics over transfer in the communites where the impact of algae is strong. This project was done in Univeristy of Chicago and University of Illinois at Urbana-Champaign.

Carbon cycling in closed ecosystems

We developed a novel technique to measure the amount of carbon cycled in closed ecosystems over day-night cycles. This technique uses low cost pressure sensors (purchasable through Amazon) embedded in the lid of glass vials. Because Henry’s constant is higher for carbon dioxide, the pressure measured is the partial pressure of oxygen, which, given the temperature and volume can be converted to moles of oxygen consumed/produced. During the day phase, due to predominant photosynthesis, the pressure of oxygen increases and during the night phase due to predominant respiration, pressure of oxygen decreases. Assuming glucose respiration, the number of moles of oxygen produced and consumed can be converted to moles of carbon. Using this technique on communities consisting of soil derived bacteria and a lab strain alga Chlamydomonas reinhardtii grown under day-night cycles, we found that the communities are functionally similar, but taxonomically divergent, indicating functional redundancy. This project was done in Univeristy of Chicago and University of Illinois at Urbana-Champaign.

Pattern formation in Dictyostelium discoudeum indicates developmental state

Dictyostelium discoudeum are fascinating amoebae. When starved they undergo a transition to a multi-cellular organism via a complex signalling mechanism. The start of the signalling mechanism indicates that the amoebae have transitioned into a development phase. When a large population of such amoebae are starved on a pertridish and observed by a darkfield microsope, the response of the cells to the signalling chemical, cyclic Adenosine Mono Phosphate (cAMP), appears as large scale (on the order of centimeters) patterns of spiral waves and target patterns. By forcing the cells to begin patter formation at different stages of their development phase, I was able to correlate the characteristic patterns with the developmental state. Further, by mixing populations at different development phases, I correlated patterns to inhomogenous development phases. Further, by modyifying an existsing model to reproduce the experimental results, I was able to predict how important parameters such as rate of release of cAMP, rate of degradation of cAMP andthe excitability of the system change with time, something that is difficult to measure experimentally. The modified model was also independently able to re-capture experiments performed by other labs. This was part of PhD, and was performed at the Max Planck Institute for Dynamical Systems and Cornell University.

Starvation memory in Dictyostelium discoideum

Entering and leaving the development phase is a costly process for Dictyostelium discoideum. In this project, I wanted to investigate whether cells retain some memory of starvation, which would make the transition easier. From the results of the project described above, it is possible to correlate patterns with the devlopmental state. With this knowledge I performed a series of experiments where the populations of cells were starved for a given amount of time, then fed for a fixed amount of time and starved again. By varying the initial starvation duration and the feeding duration, I found that the cells retain the memory of starvation for an amount of time dependent on the initial starvation duration. After this time, the cells “forget” that they were starved before, and leave the development phase completely. A simple feedback model is able to capture these dynamics. This was part of PhD, and was performed at the Max Planck Institute for Dynamical Systems and Cornell University. I thank Prof. Stan Leibler for the conversation that led to this project.

Cooperation of the haves and the have-nots

The production of the signalling chemical, cAMP is important for pattern formation and the subsequent transition to the multicellular phase. What happens when some cells “cheat” and do not produce cAMP? During my weekends, I puzzled over this question and decided on a series of expeirments to investigate. First, what happens when you reduce the cell density? I found, at low enough cell densities the populations fail to form the multicellular aggregates. So, this meant that something is not produced sufficiently at these low densities, which prevents the proper transition to the multi-cellular state. If as I initially suspected, this was cAMP, adding mutants that are deficient in production of cAMP (have-nots) should not restore the transition. However, to my surpise, adding such mutants was sufficient to restore the transition! So what was going on? With the help of Dr. Albert Bae, I performed a series of experiments that led us to conclude that the deficit chemical was not cAMP, but rather Phosphodiesterase (PDE), the chemical that degrades cAMP. The cAMP deifiencet mutants, the have-nots, provide enough PDE so that together with the cAMP produced by the wild type cells, the haves, the whole population can make the transition to the multi-cellular phase.