Microbiology:
Findings Recommend Herpes Vaccine for Human Trials
Immunology:
Genetic Battleground Marked Between HIV and Host
Biological Chemistry:
Disparate Proteins Structurally Identical
Neuroscience:
Brain Region Senses Setbacks, Redirects Action
Faculty Development:
First African-American Pediatrics Professor at HMS Asserts Value of Mentoring
Anti-angiogenesis Drug Improves Response To Radiation Therapy
Brain Wave Abnormalities May Explain Schizophrenic Hallucinations
Extended Shifts Raise Interns’ Risk of Having A Car Wreck
Cancer Institute Announces Awards to Speed Cancer Biomarker Discovery
MD-PhD Program Splits into Two Tracks
Faculty Applications Sought for Fellowship in Medical Education
Dale Named Dean for Research Integrity
News Brief
Honors and Advances
Baghdad to Boston and Back
Disparate Proteins Structurally Identical
What do mammalian transcription factor SCAN domains have in common with the C-terminal domains of capsid proteins from retroviruses such as HIV? Not much, it would appear. These peptides have no sequence homology and perform vastly different functions. Yet they turn out to be structurally identical. So says a study in the Jan. 7 Molecular Cell by Gerhard Wagner, the Elkan Blout professor of biological chemistry and molecular pharmacology at HMS, and his colleagues.
Gerhard Wagner (right) has solved the structure of SCAN domains using NMR spectroscopy. Unexpectedly, these motifs, first identified by Tucker Collins (left) as essential components of some zinc-finger transcription factors, adopt almost exactly the same three-dimensional shape as certain retroviral capsid proteins. (Photo by Steve Gilbert)
The structure of the SCAN domains provides an unexpected insight into retroviral particle assembly,” said co-author Tucker Collins, the S. Burt Wolbach professor of pathology at HMS and Children’s Hospital Boston. As such, it could lead to ways of preventing assembly of retroviruses and, hence, their replication.
Family Features
SCAN domains were first identified by Collins in 1995 as a major feature of one of the largest families of regulatory proteins, the zinc-finger transcription factors. (The acronym derives from the first letter in the names of the four proteins initially found to contain this domain.) About two years ago, Collins’s group showed that these domains form stable dimers in solution. But when the researchers had no luck growing crystals that would yield high-resolution structural information—a prerequisite to fully understand the workings of the dimers—they turned to Wagner for help. Wagner, who uses nuclear magnetic resonance (NMR) spectroscopy to determine the three-dimensional structure of proteins in solution, has pioneered techniques specifically suited for solving dimer structures.
Dmitri Ivanov, then a postdoctoral fellow in Wagner’s lab, and James Stone, then at Children’s and now an HMS assistant professor of pathology at Massachusetts General Hospital, worked closely to obtain NMR spectra of the SCAN domain from ZNF174, a zinc-finger transcription factor from humans. Early on, the project was dogged by poor-quality spectra that indicated there were several different “species” of dimer in the NMR tube.
“We reasoned that this might be due to spontaneous isomerization around a proline residue,” explained Wagner. And indeed, this turned out to be the case. When Stone mutated the protein, replacing proline 111 with leucine, the spectral quality improved immensely. Wagner suggests that this isomerization-induced heterogeneity probably prevented crystals of the peptide from forming.
With the mutant ZNF174 in hand, Ivanov was able to use an asymmetric labeling method, developed by Wagner, to obtain high-res NMR spectra. The method involves mixing monomers made with nitrogen-15 and deuterium (which is NMR silent) with monomers made with the more common nitrogen-14 (also NMR silent) and hydrogen. Magnetic resonance (or the nuclear Overhauser effect, to be precise) between nitrogen-15 and hydrogen atoms then reflects spatial proximity between atoms on different monomers. The technique allows the dimer interface to be accurately mapped.
By the time the spectroscopy data was gathered, Ivanov had accepted an industrial research position in Switzerland, where he continued to analyze the spectra. These revealed that the ZNF174 peptide exists as a dimer in which components are “swapped” between each monomer. Such domain-swapped dimers have been previously described, notably in plant viral capsid proteins. In the SCAN domain, this swapping manifests itself as the installation of one whole helical hairpin (about one fifth of the total SCAN domain) of one monomer into the body of the other monomer and vice versa. “Domain swapping offers a much tighter dimer than simple head-to-head dimerization,” said Wagner.
A Missing Link
This is an important finding in and of itself because it explains how a large family of transcription factors dimerizes (there are 64 unique SCAN domain proteins in the human genome). But what came as an even greater surprise were the results of a structural similarity search. When Ivanov reviewed the Protein Data Bank for structural homologs, viral capsid proteins were the only hits. The C-terminal domain of the HIV-1 capsid protein, for example, though it has absolutely no primary sequence homology with SCAN domains, exhibits virtually the same structure. The implications are that HIV capsid proteins may also form domain-swapped dimers, despite X-ray crystallographic evidence to the contrary.
Fair trade. One half of the mammalian SCAN dimer (left), showing domain swapping between monomers (gray and light red) is virtually identical to the known crystal structure of the C-terminal domain (CTD) of the HIV capsid protein (right). Structural similarity suggests that the capsid protein can also form a domain-swapped dimer. (Image courtesy of Gerhard Wagner)
Capsids, which form a structural envelope around retroviral cores, play a crucial role in the life cycle of HIV. The finding that they may form domain-swapped dimers resolves some outstanding questions. The major homology region of the HIV-1 capsid, for example, a stretch of about 20 amino acids that is highly conserved among all retroviral capsids, has had no obvious functional role to play, until now. “Domain swapping puts that region right in the dimer interface,” explained Collins.
“Mutations of HIV-1 capsid at positions 184 and 185 that form the dominant contacts in the proposed head-to-head in-terface do prevent the weak dimerization seen in solution, but these mutations have little effect on viral assembly,” noted Wagner. This observation supports the domain-swapping model, which predicts that amino acids 184 and 185 would constitute only a small part of the dimer interface.
Indeed, searching for capsid dimers in solution may have been a wild goose chase to begin with. Capsid proteins from viruses closely related to HIV, such as equine infectious anemia virus, do not dimerize in solution. This suggests that the head-to-head dimer of the HIV-1 capsid that does form in solution may not be physiologically relevant, again supporting the domain-swapped model for dimerization. In fact, though structural biologists have traditionally studied only the capsid itself, or a fragment of it, the peptide is part of a much larger polyprotein called gag, which is cleaved into smaller fragments only after the virus has been assembled and released from the infected cell. “The dimerization process may, therefore, be a much more complicated affair requiring the support of the complete gag protein,” said Wagner.
The discovery of domain-swapping as the glue that holds HIV capsid proteins together also suggests a new approach to keeping the virus in check. Small molecules that would interfere with domain-swapping, for example, could prevent viral replication by blocking assembly of new viral particles. The groups are now so interested in pursuing this and other research questions raised by SCAN–capsid structural similarity that Ivanov has returned to work at the labs of Wagner and Collins.
— Tom Fagan