Novel Structure of Ty3 Reverse Transcriptase
Structure of the asymmetric Ty3 RT homodimer in the presence of its cognate polypurine-tract-containing RNA/DNA hybrid. In keeping with HIV-1 RT, subdomains of Subunit A and B are designated fingers (blue), palm (red), thumb (green), while the RNase H domains are represented in gold. In contrast to the HIV-1 enzyme, Ty3 RT lacks the “tether” or connection separating its thumb and RNase H domains. Lower, Structures of Ty3 RT subunits A and B. While the overall folding of each subdomain is essentially similar for the two subunits, their overall topology is quite different, with Subunit B more closely resembling the p51 subunit of HIV-1 RT. In contrast to lentiviral and gammaretroviral RTs, Ty3 RT dimerization is strictly nucleic acid-dependent.
Retrotransposons are mobile genetic elements that self amplify via a single-stranded RNA intermediate, which is converted to double-stranded DNA by an encoded reverse transcriptase (RT) with both DNA polymerase (pol) and ribonuclease H (RNase) activities. Categorized by whether they contain flanking long terminal repeat (LTR) sequences, retrotransposons play a critical role in the architecture of eukaryotic genomes and are the evolutionary origin of retroviruses, including human immunodeficiency virus (HIV).
Ty3, a yeast LTR retrotransposon, encodes an RT that has been studied extensively for its biochemical activity and substrate binding interactions. Unlike retroviral enzymes, Ty3 RT lacks a conserved tether domain between the DNA pol and RNase domains and has a smaller distance, as measured by substrate binding, between its active sites, which cannot be explained by known retroviral RT protein structures. To understand how Ty3 RT functions, Stuart Le Grice, Ph.D., of CCR’s HIV Drug Resistance Program Retroviral Replication Laboratory, along with colleagues from his lab and the International Institute of Molecular and Cell Biology in Warsaw, Poland, set out to determine the structure of Ty3 RT.
The researchers co-crystalized purified Ty3 RT in the presence of a 16-base pair RNA-DNA hybrid, the sequence of which corresponded to the Ty3 RT poly-purine tract (PPT) primer and RNase cleavage site, and refined the structure at 3.1 Angstroms resolution. In contrast to previous reports of Ty3 RT as a monomer in solution, their structure showed an asymmetric homodimer in the presence of the nucleic acid hybrid.
Labeling the dimer subunits A and B, the investigators found that subunit A had a similar structure to retroviral RTs. Its DNA pol domain resembled a right hand with palm, fingers, and thumb subdomains. The RNase domain, however, was in a position analogous to the tether domain of retroviral RTs. While the sequence of the A and B subunits is identical and the structures of their subdomains were generally similar, the scientists noted a significant difference in the positioning of the RNase and DNA pol thumb subdomains between the subunits. In subunit B the RNase domain was positioned between the DNA pol fingers and palm subdomains, blocking the DNA pol substrate binding region and displacing the thumb. The interface between the subunits was polar and involved two main contact points, namely insertion of the B subunit fingers between the palm and RNase domains of subunit A, and the interaction of both RNase domains.
Looking more closely at these domains, the researchers found two main differences between Ty3 and cellular or retroviral RNases. The first was a lack of 10 residues in the Ty3 enzyme, which shortened one strand of the central β-sheet. The second was a rearrangement of α-helices. Since the Ty3 RNase H active site resembles these other enzymes, it likely uses a similar mechanism, but the investigators were unable to identify important substrate association residues, particularly those in the phosphate-binding domain. In addition, neither subunits’ RNase domain interacted with the RNA, making it unclear which subunit actually contributes RNase activity. Using previous in vitro data, the scientists modeled conformational changes that would be necessary to bring either subunit’s RNase near the RNA backbone. Only the movement of subunit B was possible without disrupting dimerization, suggesting it is responsible for nuclease activity.
The researchers then focused on interactions between the RNA-DNA hybrid and Ty3 RT. They found that the hybrid adopted a confirmation between A- and B-form DNA and noted a fairly uniform structure along its length. The hybrid sat in a positively charged cleft of the dimer made up of domains of both subunits. The two main interacting regions involved contacts between subunit A DNA pol and nucleotides at the RNA 5’ end and DNA 3’ end and between RNases H of both subunits and the DNA 5’ end. The observed contacts between subunit A and the hybrid and the similar structures of subunit A and HIV-1 RT suggested that subunit A is responsible for DNA pol activity. Additional contacts observed and modeled onto the crystal structure by the investigators supported a number of previous biochemical studies.
To verify and extend the observations from their Ty3 RT crystal structure, the scientists generated a series of mutant proteins and tested their abilities to dimerize and to carry out enzyme activities. Substituting alanine for a number of residues within the putative dimerization and substrate binding domains reduced or eliminated dimer formation, prevented efficient DNA polymerization, and also severely affected RNase activity. The researchers then took advantage of the unique orientation of the Ty3 RT dimers to investigate their conclusion that the DNA pol and RNase activities are contributed by separate subunits. They combined two mutants, R140A/R203A and D426N, which are critical for the dimer domain of subunit A but not B and for RNAse H activity, respectively. These mutants should form two combinations of dimers: D426N homodimers, which are inactive, and a mixed dimer with a double alanine mutant in subunit B, which should retain RNase activity. Encouragingly, the investigators observed nuclease activity confirming their model.
In comparing this first crystal structure of a retrotransposon RT with that of HIV-1 RT, the scientists pointed out two important differences: the requirement for substrate binding for Ty3 dimer formation and the fact that both enzyme activities reside in the same subunit of HIV-1 RT. However, the overall topologies of the RTs are remarkably similar, suggesting that the Ty3 RT structure represents its physiological architecture. Future studies that focus on substrate induced Ty3 dimerization and whether other LTR retrotransposon RTs have a similar mechanism will enhance the understanding of these genetically important enzymes.Summary Posted: 04/2014
Nowak E, Miller JT, Bona MK, Studnicka J, Szczepanowski RH, Jurkowski J, Le Grice SFJ, and Nowotny M. Ty3 reverse transcriptase complexed with an RNA-DNA hybrid shows structural and functional asymmetry. Nature Structural & Molecular Biology. March 9, 2014 PubMed Link