My First Paper: tRNAs as Genetic Switches

It thrills me beyond words to announce that I am now a published scientist, and that in this post I can cite Sawyer et al. (2012)! The first figure of the paper depicts synthetic biology research as a winding path between tackling engineering questions and generating new knowledge about basic science. But, as you'll see, the publication of this paper is itself another turn in my own journey of scientific discovery.

A Stretch in the Winding Path of Synthetic Biology Research

Synthetic biology encompasses a huge array of ideas and techniques, but my paper¹ focuses on parts-based synthetic biology in E. coli. Genes and regulatory DNA can, to an approximation, be treated as interchangeable parts like resistors, lights, and switches in a circuit. Individual projects contribute to synthetic biology by building new parts or combining existing ones in new ways to generate new functions. Most projects, including this one, do a combination of both.

We were trying to develop a new category of tRNA parts to act like switches. tRNAs are the adapter molecule that, in the process of translation, convert the genetic code of the nucleic acids DNA and RNA to the sequence of amino acids that makes a protein. Experiments in the 20th century revealed that three bases of DNA or RNA correspond to one amino acid in a protein. Each triplet of bases is called a codon, because the collection of bases codes for one amino acid. During translation, the ribosome exposes a codon in the mRNA strand, and the complementary tRNA attaches its amino acid to the growing protein chain.

Previous experiments have shown that that is not always the case. Incorporating additional bases into the anticodon loop of a tRNA (the part that "reads" the codon) can increase the length of its recognition codon. This has important implications because of how the genetic code works.

The English language is in a sense also a code. We combine letters into words that encode a particular meaning. There is nothing special, or sheep-like, about the collection of letters in the word "sheep." Nor is there anything special in "ovis," whose collection of letters is used by a different language (Latin) to mean the same thing.

Like in RNA, we can construct a sentence using three-letter words ("codons"): THE-CAT-CAN-RUN-FAR. But the message changes when the same ordering of letters is rearranged: THEC-ATC-ANR-UNF-AR. Likewise, if a tRNA reads four bases as a single codon, the DNA message is changed.

Sometimes genes are mutated by the addition or removal of a nucleotide, called a frameshift mutation because the triplet "reading frame" of translation is changed. The original message is lost when re-grouped into threes as follows: THE-CAT-CAN-RUN-FAR becomes, e.g., when mutated THE-XCA-TCA-NRU-NFA-R. However, if a tRNA that can read THEX as a single codon word is present, the correct reading frame can be recovered as THEX-CAT-CAN-RUN-FAR. This slight alteration to the original sentence is much less devastating than the frameshift. These tRNAs do in fact exist, and they are called frameshift suppressor tRNAs because they suppress the effect of a frameshift mutation (Fig. 2B, a model of a frameshift suppressor tRNA I designed in high school).

My group set out to develop frameshift suppressor tRNAs as a molecular switch for gene expression. A gene containing a disabling frameshift mutation is OFF unless a corresponding frameshift suppressor tRNA is expressed in the cell (Figs. 2C and 2D: functional GFP is produced only when the frameshift is suppressed by a frameshift suppressor tRNA).

We envisioned suppressor tRNAs as a method for encoding logic expressions into living cells. The figure at the top of this post is an example control system that could, in theory, be implemented using these tRNAs. Imagine, for example, a population of E. coli in a fermenting vat refining a drug. The cells are fed two precursors which they combine and modify using a series of enzymatic reactions. Four suppressor tRNAs control the circuit. tRNA A is produced under DNA damage, tRNA B under oxidative stress from free radicals, tRNA C when precursor X runs out, and tRNA D when precursor Y runs out. If the cell is experiencing DNA damage or oxidative stress, then stress response proteins are produced to keep the cell alive and producing. If substrate X or substrate Y accumulates, the cell produces more enzyme to pick up the slack of the reaction. And finally, if the cell is experiencing DNA damage or oxidative stress, and one or both of the substrates is in excess (things are in total disarray!), the cell breaks itself open to conserve resources at the population level and release the drugs it has produced so far.

Suppressor tRNAs lend themselves well to producing OR and AND (Fig. 3). A gene is shifted out of frame with, e.g., the sequence shown. If either suppressor tRNA A or suppressor tRNA B is expressed, the original reading frame is restored and a functional protein is produced. Here, the tRNAs are shown reading five mRNA nucleotides as a single codon, which is precisely what we used in our design.

Things Take an Unexpected Turn

Everything I have said so far is all well and good, but research often takes unexpected turns. We built tons of constructs shifted out of frame in a way that could be restored with different combinations of six tRNAs. But something funny was happening. We designed our constructs so that the E. coli would glow red if the combination of tRNAs they were provided satisfied the logic constraint. But the pattern of glowing E. coli did not match the pattern predicted by the logic statements. Instead what we found is that the tRNAs were actually reading four nucleotides as a codon, instead of five (Fig. 4 shows the mechanism)! That was surprising, since it went against what was already published in the literature.

However, in our experiments the tRNAs were operating under very different conditions from the previous literature. Our experiments did not exert selective pressure on the cells; we just looked to see if they glowed red. The experiment from the literature demanded that the cells suppress a five-base codon or they would be killed by an antibiotic. Thus, four-base codon events were not detected in their assay, only five (which we believe to be rarer; otherwise we would have observed them in our assay). And so, because of this snag, the paper reports this unexpected suppressor tRNA behavior. Had we started knowing that these tRNAs read four-base codons rather than five, maybe we could have produced something like the schematic at the top of the page.

Although we were not successful in designing a circuit like the one above, labs around the world can pick up where we left off by requesting our tRNAs from the Registry of Standard Biological Parts. There has been a lot of exciting work recently in RNA synthetic biology, from designer RNA aptamers that regulate gene expression by binding to small molecules, to riboswitches, and more. It was a thrill to make a small contribution.

My Own Winding Path Through Science: The Story Behind the Paper

It's hard to say when my interest in science first appeared. I would say that every child is intrinsically interested in science, requiring concerted effort to extinguish it. But even in elementary school I took notice to the world of science and medicine. I even took a summer school class all about the subject! At the time I could not have known that, on an ordinary day in my sixth grade summer, I met the biologist with whom I would coauthor a paper a decade later.

Years later, in high school, I was fortunate to be given the opportunity to do original research in an elective science class. In search of guidance and mentorship, I contacted the biologist—Todd Eckdahl—who advised me on a couple projects. Then, in the spring of my junior year in high school, he offered me a summer research position in his lab learning molecular and synthetic biology techniques in addition to modeling molecular structures on the computer. This experience would change the course of my life, from the respectable medical profession to the noble scientific one. I was sold, then and there. Research was my passion. It wasn't easy, but the personal rewards were worth it.

I got to meet and travel to work alongside the lab's collaborators from Davidson College. It's no coincidence that I now write this from Davidson College. Without my exposure to research in high school, I would not have found the college which, for better or worse, I now owe so much. But before coming here, I got to travel to MIT with the whole team to present our work at iGEM, the International Genetically Engineered Machine competition. I had never seen anything like it. Undergraduates and their advisors came from all over the world to share their synthetic biology projects all in one breakneck weekend. It was a great time, with great people.

Before coming to Davidson, Dr. Eckdahl took me for a second summer. I was much more independent, holding my own in designing constructs and performing experiments. When I got to Davidson, I took a slightly-less-fulfilling year off from research as I struck off graduation requirements and settled into college life. I then spent my first summer at Davidson, as well as the subsequent fall semester, working to finish the project I had started in high school. Out of the collective ideas and painstaking labor of the paper's many authors, we discovered something unexpected and new.

I don't know what field of research I will eventually come to claim as my own, but no matter how my future unfolds I will always remember this fateful project that carried me into the world of science, setting me on my own winding path of wonder and discovery. I am forever indebted.

Image credits: Figure 4 is from ref. 1. The rest were made by me, specifically for this post. Figures from the paper, as well as my figures in this post, can be reproduced with attribution.

Reference:

1. Sawyer, E. M. et al. Bacterial Logic Devices Reveal Unexpected Behavior of Frameshift Suppressor tRNAs. Interdisciplinary Bio Central 4, 10 (2012).

Acknowledgments:

Thanks to all the student and faculty coauthors on the paper, and for support and facilities from Missouri Western State University and Davidson College.