 |
Structure and Function of Nucleic Acid Regulatory Complexes
Summary: Leemor Joshua-Tor studies the protein components of the RNAi machinery in cells, which can selectively silence genes, as well as complexes involved in DNA replication.
My research program reflects my long-standing interest in nucleic acid regulatory processes. Trained as a chemist, I was initially drawn to nucleic acid chemistry and structure. But soon my focus broadened to the regulation of nucleic acid function and, in particular, the protein–nucleic acid complexes at the center of many basic life processes, including replication and gene expression. I have been especially attracted to areas that present unique mechanistic challenges, though often of clinical relevance. Deciphering the mechanisms that underlie double-stranded RNA gene silencing is a major effort in my laboratory. Another is to obtain a molecular view of the initiation of DNA replication and how the proteins involved manipulate DNA.
RNA Interference
The emergence of a new pathway for gene regulation, namely RNA interference (RNAi), sparked my interest. I was struck by how genetics, biochemistry, molecular biology, and bioinformatics were being used to understand this new paradigm in gene regulation. I wanted to add structural biology to the mix, moving the mechanistic descriptions beyond the "blobology" that depicted the various factors involved as blobs devoid of the molecular, chemical, and mechanistic features that might explain how they work in the various RNAi-related pathways.
RNAi is triggered by double-stranded (ds) RNA to elicit gene silencing. Similar components in all RNAi-related pathways appear to suppress gene expression through mRNA degradation, translation, or chromatin structure. The best-characterized pathway, the one predominantly used in gene-knockdown technology, is post-transcriptional gene silencing (PTGS) or "slicing." Here, the RNA-induced silencing complex (RISC) produces an endonucleolytic cut in the mRNA target, preventing gene expression from proceeding.
We originally targeted Argonaute, the signature component of RISC, because it is involved in all RNAi-related processes. It was also the most mysterious component; nothing was known, except that it contained two conserved but uncharacterized domains: PAZ and PIWI. We initially solved the structure of the PAZ domain. With this and the results of biochemical studies, we proposed that the PAZ domain has a role in siRNA 3'-end recognition. The structural studies of PAZ-RNA complexes by Dinshaw Patel (Memorial Sloan-Kettering Cancer Center) and Michael Sattler (European Molecular Biology Laboratory) subsequently confirmed this.
But it was our structure of a full-length Argonaute protein (Figure 1) that instantly solved a long-standing puzzle in the RNAi field. We saw at once that Argonaute was the long sought Slicer, which performs the critical slicing event—cleavage of the mRNA. This revelation arose from the fact that the PIWI domain looks very much like an RNase H enzyme, consistent with the biochemistry of the slicing reaction. We demonstrated that a large, positively charged groove along the protein could accommodate siRNA and substrate (mRNA), with the scissile phosphate at the active site. We then turned to Gregory Hannon (HHMI, Cold Spring Harbor Laboratory), who had just shown that only immunoprecipitations of human Argonaute-2 (hAgo2), but not the other human Argonautes, contain slicer activity. Together we mutated the key aspartates identified at the active site to alanine; this abolished slicing, showing hAgo2 to be the catalytic engine of RNAi. This is a clear case of structural biology identifying the critical role of a key protein in an important biological pathway.
To prove unequivocally that Argonaute is indeed Slicer, we made recombinant hAgo2 in Escherichia coli, an organism that lacks RNAi. Together with the Hannon laboratory, we showed that all that is needed to carry out slicing is an active Argonaute protein and an siRNA. We could thus make a "minimal" recombinant RISC.
We have identified a third clade of the Argonaute family, the group III Argonautes, or WAGOs, which appear to be specific to worms. These consist predominantly of nonslicing Argonautes: they do not have an intact catalytic motif. I thought this was rather curious and approached Craig Mello (HHMI, University of Massachusetts Medical School). We realized that all the deletions in a multiple Argonaute mutant in the worm Caenorhabditis elegans that he had been characterizing belong to this third clade. Further analysis revealed that they act in a separate downstream silencing step. These secondary Argonautes (SAGOs) act following an initial trigger step, carried out by the slicing Argonaute RDE-1, that produces secondary siRNAs. How SAGOs specifically recognize secondary siRNAs that carry a unique 5'-end modification and how the nonslicing SAGOs bring about silencing are questions we are starting to address.
RNAi-mediated transcriptional gene silencing (TGS) directed toward heterchromatin establishment and maintenance is another area of intense effort in my lab. We set out to understand the function of complexes involved in TGS, the role of their components, and the mechanism underlying heterochromatization. The fission yeast Schizosaccharomyces pombe is the best-characterized system to study TGS because the protein complexes involved are well defined. It is also easy to translate structural, biochemical, and biophysical findings to in vivo studies. Silencing is dependent on several multiprotein complexes. RITS (RNA-induced transcriptional silencing complex), a specialized RISC complex, consists of the Argonaute protein, Ago1; the chromodomain-containing protein Chp1; the novel protein Tas3; and siRNA. We asked whether slicing activity is important for TGS, given that Ago1 is the sole Argonaute protein in S. pombe. Together with my colleague Robert Martienssen (Cold Spring Harbor Laboratory), who made some of the initial seminal discoveries on the link between heterochromatin and RNAi, we showed that Ago1 is an active Slicer in vitro and that slicing is required for TGS.
The Argonautes, central to all RNAi pathways, are stretching their tentacles into diverse forms of gene silencing. Our Ago-centric approach (funded in part by a grant from the National Institutes of Health) is enabling us to reveal some of the biochemical mechanisms underlying these pathways. Since we are addressing these mechanisms across broad biological systems, we have the opportunity to understand the evolution of this mechanism, particularly the intrinsic catalytic activity and its involvement in diverse silencing pathways.
Replication Initiation
Replication initiation, a basic biological process, is important in controlling developmental processes and is altered in diseases such as cancer. It begins with binding of an origin recognition complex, such as DnaA in E. coli and ORC in yeast and higher eukaryotes, which binds to dsDNA. It progresses by subsequent DNA melting and assembly of a helicase and recruitment of many factors that enable replication. In E. coli, three proteins are required just to establish the replicative helicase, DnaB, at the origin. In eukaryotes, many more proteins are required. But small DNA tumor viruses, such as papillomavirus and SV40, require a single initiator protein for origin recognition, helicase loading, and helicase unwinding. Intriguingly, different oligomeric forms of these initiator proteins control the various activities. Papillomavirus E1 forms a dimer with sequence-specific dsDNA recognition. Through a series of discrete intermediate oligomeric states that progressively distort and subsequently melt the DNA, it ultimately forms two hexameric helicases assembled on the origin. This fascinating phenomenon—where various oligomeric forms of the same protein lead to different functional roles—was a strong incentive for pursuing this system from a structural and biophysical perspective.
Several years ago we determined crystal structures of the DNA-binding domain (DBD) of E1, both unbound and in two stages of assembly at the origin as complexes (dimer and tetramer) with dsDNA. This led us to propose a model in which the transition from the dimer to the ultimate double hexamer results in strand separation. The loading and assembly of this protein separates the two strands of the DNA, such that each hexameric helicase in the final complex encircles one strand of DNA. Once assembly is completed, the helicase uses its ATP-driven motor to translocate on the DNA or "pump" the single-stranded (ss) DNA through the hexameric ring. Several competing mechanisms for helicase unwinding were proposed over the years for these types of helicases, many based on crystal structures of hexameric helicases and other AAA+ motor proteins that were determined in the absence of DNA.
We determined a crystal structure of hexameric E1 with ssDNA discretely bound in the central channel and nucleotides at the subunit interfaces (Figure 2). This structure demonstrates that, unlike some previous models, only one strand of DNA passes through the hexamer channel and that the DNA-binding hairpins of each subunit form a spiral staircase that sequentially tracks the oligonucleotide backbone. Consecutively grouped ATP, ADP, and apo configurations correlate with the height of the hairpin in the channel, suggesting a straightforward DNA translocation mechanism, which we called a "coordinated escort mechanism" (Figure 3). Each subunit sequentially progresses through ATP, ADP, and apo states while the associated DNA-binding hairpin travels from the top staircase position to the bottom, escorting one nucleotide of ssDNA through the channel. These events permute sequentially around the ring from one subunit to the next, resembling six hands that grab the DNA to pull it through the channel (see movies). (This work has been funded by a grant from the National Institutes of Health.)
As their name suggests, the AAA+ (ATPases associated with a variety of cellular activities) family proteins are involved in many important biological processes and pathways, from a chaperone function in protein degradation, vesicular fusion, to DNA manipulation events, such as the ones we are studying. All use a multisubunit motor function. The motor function in our mechanism for DNA translocation, for example, is already being discussed in these other contexts. We continue to be interested in the evolution of these motors for performing these diverse tasks in biology.
Our aim in these studies is to dissect, decipher, and study the evolution of the intricate mechanisms underlying these fundamental biological processes.
Last updated July 07, 2008
|
|
|
|
