This week we profile a recent publication in Nature Genetics from Pei-Hsuan Wu (pictured below) in the laboratory of Dr. Phillip Zamore (pictured above) at the Howard Hughes Medical Institute and RNA Therapeutics Institute at UMass Medical School.
Can you provide a brief overview of your lab’s current research focus?
Argonautes are the only known family of proteins that can be programmed with any RNA or DNA sequence to make sequence-specific regulators of transcription, mRNA stability, or translation. Our lab seeks to understand the biology and mechanism of paradigmatic examples of Argonaute proteins and pathways, and, ultimately, to use these insights to design and improve small RNA-guided therapies for human diseases. Indeed, studying how Argonautes work and how their small RNA guides are made has led to the development and FDA approval of small RNA drugs. Nevertheless, fundamental questions about the specificity and function of Argonaute protein-mediated pathways remain unanswered.
Despite >20 years of study, for example, we still cannot predict how Dicer enzymes will cleave a pre- miRNA based only on its sequence. We will use biochemical and structural approaches to identify the features that determine where Dicer cleaves a pre-miRNA and how Dicer partner proteins alter this process.
In animals, the PIWI subfamily of Argonaute proteins uses 23–30-nt pachytene PIWI-interacting RNA (piRNA) guides to silence transposons or regulate gene expression in germ cells. piRNAs are made from specific long, single-stranded precursor RNAs. Our research seeks to explain why some genomic regions and transcripts are destined to make piRNAs, while others are excluded. By studying piRNAs in flies, moths, and mice, we hope to identify both evolutionarily ancient and newly evolved strategies that animals use to designate piRNA precursors and to convert them into functional complexes with PIWI proteins. While experimental and computational studies have dramatically improved our ability to predict miRNA targets, similar advances have not yet been made for piRNAs. In the spermatocytes of placental mammals, pachytene piRNAs are nearly as abundant as ribosomes, but we still do not know what or how they regulate. Mutations in the proteins that make pachytene piRNAs cause male infertility, suggesting that pachytene piRNAs promote sperm development. We use biochemistry and mouse genetics to study the function and specificity of pachytene piRNAs.
Finally, many archaeal and bacterial genomes encode Argonautes, yet we rarely know what they do or how they acquire their guides. Some Argonautes, such as Thermus thermophilus Ago, even use DNA guides to target DNA. Understanding the function of bacterial Argonaute proteins will not only broaden our evolutionary understanding of this remarkable protein family, but may ultimately lead to the development of novel antibiotics.
What is the significance of the findings in this publication?
Fundamentally, they suggest that in mammals—including human beings—a convoluted network of selfish small RNAs is using spermatogenesis to ensure its own propagation. Such selfish genes are typically associated with reproductive isolation, so these testicular small RNAs, called piRNAs, are using mammals to maintain themselves, rather than mammals using the small RNAs to regulate genes. piRNAs first evolved in the ancestor of all animals in order to silence transposons. Along the way, some picked up specialized functions in regulating gene expression. Our work suggests that in the case of the mammalian “pachytene” piRNAs, an entire class of small RNAs has gone over to the dark side, using us to ensure their transmission to our offspring and perhaps competing with each other for survival. We think that these small RNAs act like a network of non-autonomous transposons, ensuring their own replication by providing a tiny advantage to the production of successful gametes while creating incompatibilities among members of the same species.
What are the next steps for this research?
We are developing a mouse model system to test our ideas. It will probably end up taking three to five years to get a result, so I hope it will be informative.
This work was funded by:
The work was made possible by HHMI and a (now ended) P01 grant from the NIHCD.
