Modular, programmable proteins enable tracking and control of gene expression
Researchers at MIT have developed a new class of engineered proteins that can be programmed to bind virtually any RNA sequence. These customizable RNA-binding proteins make it possible to visualize RNA inside living cells, monitor how often a specific mRNA is translated, and even modulate RNA activity without altering the cell’s DNA.
The approach adapts naturally occurring human RNA-binding domains involved in embryonic development and retools them into easily targeted, modular components. By converting these domains into interchangeable parts, the team created a straightforward way to design proteins that recognize chosen RNA sequences.
“You can use these proteins to measure RNA production or to monitor translation from RNA to protein,” says Edward Boyden, associate professor of biological engineering and of brain and cognitive sciences at the MIT Media Lab. “This capability could have broad utility across biology and bioengineering.”
Unlike earlier protein-based methods for targeting RNA, the MIT system emphasizes modularity. The researchers argue that modular, repeatable parts simplify design and accelerate development of diverse RNA tools.
Boyden is the senior author of the study, published in the Proceedings of the National Academy of Sciences. The paper’s lead authors are postdoctoral researcher Katarzyna Adamala and graduate student Daniel Martin-Alarcon.
Modular code
Cells contain many different types of RNA, each with distinct roles. Messenger RNA (mRNA) carries the protein-coding instructions copied from DNA to ribosomes, where the instructions are translated into proteins. Observing mRNA dynamics offers insight into which genes are active in a cell, and manipulating translation of target mRNAs provides a means to alter gene expression without editing the genome.
To create programmable RNA binders, the team focused on Pumilio homology domains—natural RNA-binding proteins that recognize specific nucleotide “letters” (adenine, uracil, and guanine). Previous efforts to redesign these proteins required case-by-case optimization. The MIT researchers systematically tested many amino-acid combinations and established a reliable code linking particular amino-acid motifs to each RNA base at any position within a target sequence.
They named the platform Pumby (Pumilio-based assembly). Using this modular code, the researchers built proteins that can target sequences ranging from six to eighteen bases in length with predictable specificity.
“This is a significant advance,” says Robert Singer, professor of anatomy and structural biology, cell biology, and neuroscience at Albert Einstein College of Medicine, who was not involved in the study. “Previous approaches often required modifying the RNA to include a binding sequence; with this technique you design the protein to match the native RNA, so you can target any RNA in any cell without altering the RNA itself.”
RNA manipulation
In cultured human cells, the team demonstrated several practical applications. To image specific mRNAs, they designed two Pumby proteins that bind adjacent sequences on the same transcript. Each Pumby was fused to a complementary fragment of green fluorescent protein (GFP). When both proteins bind their respective sites, the GFP fragments assemble and become fluorescent, signaling the presence of the target mRNA.
The researchers also found that GFP fluorescence is dynamic during translation: the fluorescent tag can be displaced when a ribosome translates the mRNA and then rebinds after translation, enabling measurement of translation frequency for individual transcripts. This gives a quantitative readout of how often a given mRNA is being read by the protein-synthesis machinery.
Beyond imaging and measurement, Pumby proteins can be used to control translation. By fusing a translation-initiator protein to a Pumby binder, the team significantly increased translation of an mRNA that is normally translated only rarely, effectively turning up protein production from that message without genomic modification.
“We can increase translation of arbitrary genes in the cell without changing the genome at all,” says Daniel Martin-Alarcon.
The researchers are expanding the method to label and manipulate different mRNAs within neurons, testing the idea that distinct mRNAs are stored in separate neuronal compartments to support functions such as memory formation. Until now, observing and controlling those localized mRNAs has been difficult; the new platform offers a direct way to do both.
These programmable RNA binders could also serve as molecular scaffolds to assemble enzymes in a defined order, creating nanoscale assembly lines for biosynthesis of drugs or other valuable molecules.
Source: Anne Trafton – MIT
Image credit: Christine Daniloff/MIT.
Original research: The study appears in Proceedings of the National Academy of Sciences (PNAS), week of April 25, 2016.