Please click here for my list of publications.
I'm currently a PhD graduate student in the Muller lab at UCSD. My research is mainly focused on origins of life, with some minor forays into broader RNA biochemistry.
The scientific consensus on when life originated on Earth is about 3.4 billion years ago. The current leading model for how the first organisms existed is called the RNA world hypothesis. This is a theory that posits that RNA served both genetic (currently DNA) and catalytic (currently proteins/enzymes) roles for the earliest organisms. Providing evidence towards supporting this theory is a little complicated. Because we don't have a time machine, we cannot provide direct evidence for the RNA world hypothesis. What we mean by this is no more "RNA fossils" exist (if they even did in the first place) because all the RNAs would have degraded by the present day, billions of years after being generated. The only way to provide evidence is to imagine the essential functions for an RNA world organism and then creating novel RNAs that can catalyze/perform those functions in a laboratory setting.
Catalytic RNAs, also known as ribozymes, can catalyze functions just like proteinaceous enzymes. Creating novel ribozymes in the lab to catalyze RNA world reactions can be done using a method called in vitro selection.
A general layout for an in vitro selection: Starting from step A, a DNA pool containing trillions of randomized DNA sequences is transcribed into RNA. Depending on their respective sequences, certain RNAs will fold into a shape that can be “functional” for a specific reaction with a given substrate. The entire pool is incubated with the substrate (B). Only the functional RNAs survive via their success with catalyzing the substrate for a specific reaction (C). The reactive RNAs are then reverse transcribed back into DNA to be amplified for the next round of selection (D, E).
The specific RNA world reaction we want catalyzed is the generation of chemically activated molecules. These are essential for any organism in order to overcome energetically unfavorable reactions. Triphoshorylated molecules can be used as an energy currency by an organism to drive aforementioned reactions. Triphosphorylation, therefore, can be used to generate a chemically activated molecule. Specific to RNA, a prebiotically plausible synthesis would be triphosphorylating the 5' hydroxyl group using trimetaphoshphate (shown below).
A triphosphorylation reaction: The ribozyme on the left manages to triphosphorylate itself under the following conditions: 100 mM Mg2+, 50 mM trimetaphosphate (Tmp), and 50 mM Tris/HCl pH 8.3. This specific ribozyme came out of an in vitro selection in our lab. This selection, and the subsequent ribozymes that were generated, demonstrated two very important points: triphosphorylation can be catalyzed by a ribozyme, and triphosphorylation can be done using trimetaphosphate as a substrate. Using trimetaphosphate is particularly interesting because it has prebiotic relevance (it was likely available at the time).
With a triphosphorylation ribozyme (TPR1) finally available, we focused on characterizing and optimizing it using a doped selection. A doped selection is an in vitro selection that has an initial pool stemming from a single construct as opposed to starting out with unique randomized ones. The result is the selection searches the local sequence space of the parent ribozyme to find variants of it that might react better. This doped selection resulted in TPR1e, a triphosphorylation ribozyme that's 24-fold faster than its parent ribozyme while containing only 7 mutations.
TPR1e: A doped selection on TPR1 resulted in the generation of TPR1e, a much faster triphosphorylation ribozyme. TPR1e acquired 7 mutations that differentiated it from its parent TPR1 (shown in blue). All the major secondary structures remained the same between the two ribozymes.
Analyzing the beneficial mutations of TPR1e made us speculate that the ribozyme contained a new secondary structure motif. We performed SHAPE analysis with base covariation (wikipedia has a great article on SHAPE) in order to confirm our speculations.
TPR1e revised structure: The ribozyme on the left is TPR1 in its old structure. The ribozyme on the right is the optimized TPR1e with its revised structure. The blue nucleotides were indicative of being double stranded during SHAPE analysis. The red nucleotides were indicative of being single stranded. This overlay shows the agreement between experimental data and our predictive secondary structure.
In pursuit of a TPR1e, it was discovered that the P4 helix (see above) would tolerate mutations without greatly affecting the function of the ribozyme. This was a sign for us to investigate further. Removing unnecessary regions of the ribozyme would make it more RNA world plausible (the shorter the sequences the better), might make it faster, and might provide biochemical insight as to how this specific ribozyme works. Our first step was to split the entire ribozyme into two fragments (fragmented along the middle of the loop in the P4 helix) and see if the fragments could come together and still function (see 2016 paper). After confirming it still worked, I successively truncated the P4 helix (the respective 3' end of the 5' fragment and the 5' end of the 3' fragment) while measuring activity. Surprisingly, the entire P4 helix could be removed while still retaining function. The new construct was named TPR1f.
What's amazing about TPR1f isn't just the fact that it consists of three individual RNAs coming together to function, at lower temperatures TPR1f is faster than TPR1e. This finding was large enough to be published by the science news magazine "Chemistry World" (link here).
The figure above is a temperature kinetic analysis of TPR1e and tPR1f. TPR1e is in blue and TPR1f is in green. The x-axis is the reaction temperature the experiment took place at and the y-axis is the Kobserved (the "speed") of the reaction. The higher the Kobserved the faster the reaction goes to completion. The interesting part is that TPR1f is faster than TPR1e below 20 degrees Celsius. This may have implications if the early Earth was cold. There is no general consensus about what the temperature of early Earth was like, but in the even it was cold shorter fragmented ribozymes might have been favored assuming this trend carries over into other ribozymes.