HIV Shell Structure Cracked With Help Of Supercomputer

A new study that features on the cover of Nature this week describes how researchers in the US have for the first time cracked the chemical structure of the capsid or protein shell of the human immunodeficiency virus (HIV). The breakthrough, which likely opens the way to powerful new drugs against the virus that causes AIDS, was made possible with the help of a new "petascale" supercomputer.
Scientists have been trying for some time to crack the precise chemical structure of HIV's cone-shaped capsid, a protein shell that protects the virus's genetic material. The capsid is thought to be the key to virulence of HIV and has become an attractive target for new antiretroviral drug development.

As senior author of this new Nature study, Peijun Zhang, an associate professor of structural biology at the University of Pittsburgh School of Medicine, says in a statement:

"The capsid is critically important for HIV replication, so knowing its structure in detail could lead us to new drugs that can treat or prevent the infection."

"This approach has the potential to be a powerful alternative to our current HIV therapies, which work by targeting certain enzymes, but drug resistance is an enormous challenge due to the virus' high mutation rate."

Previous studies have described attempts to chip away at the capsid structure bit by bit. To try and see the atomic-level detail of the shell, made of over 1,300 identical proteins, researchers have used a range of sophisticated lab tools, from nuclear magnetic resonance spectroscopy and X-ray crystallography, to cryo-electron microscopy and cryo-EM tomography.

But it was only when they added the processing power of the new petascale Blue Waters supercomputer at the National Center for Supercomputing Applications at the University of Illinois, to the already impressive array of tools, that Zhang and colleagues were able to fathom the chemical structure of the entire capsid.

A petascale computer has a number-crunching rate measured in "petaflops", or petas (quadrillions, 10¹⁵) of floating point instructions per second. To put this into context, a petascale computer can perform in one second the same number of instructions as it would take everyone on Earth doing one calculation per second for 1.5 days.

The simulations that added the missing pieces to the HIV capsid puzzle were conducted during testing of Blue Waters by co- authors Klaus Schulten, a physics professor, and Juan R. Perilla, a post-doc researcher, both at the University of Illinois.

Commenting on the HIV capsid challenge, Schulten says:

"This is a big structure, one of the biggest structures ever solved."

"It was very clear that it would require a huge amount of simulation - the largest simulation ever published - involving 64 million atoms," he adds.


From previous studies that had found the HIV capsid contains a number of identical proteins, the researchers already knew these proteins are arranged as pentagons and hexagons, and they had a hunch that the pentagons formed the tight round corners of the cone-shaped capsid they could see under an electron microscope.

But exactly how many of these proteins it takes to make the capsid, or how the pentagons and hexagons fit together, remained a mystery.

Zhang and the structural biology team at Pittsburgh found that when exposed to high concentrations of salt, the protein building blocks assemble into tubes made only of hexagons.

From further experiments they found that certain regions of the proteins interact with one another in a way that is "critical for capsid assembly and stability, and for viral infectivity," they note.

They then managed to get a rough idea of the overall shape of the capsid by taking cryo-electron tomographs of it sliced into sections.

From these results, and their own simulations of how the hexamers and pentamers might interact, Schulten and Perilla carried out a series of large-scale computer simulations.

Schulten says that they could only match the 64-million-atom capsid structure to the "diverse" experimental data using a unique approach they developed themselves that they call "molecular dynamic flexible fitting".

"You basically simulate the physical characteristics and behavior of large biological molecules but you also incorporate the data into the simulation so that the model actually drives itself toward agreement with the data," he explains.

With these techniques the researchers found that the HIV protein shell comprises 216 hexagons and 12 pentagons arranged in the way the experimental data suggested.

The proteins in the hexagons and pentagons were identical but the angles through which they attached to each other were different among different regions of the structure.

Schulten says this is what puzzled them: such a protein would have to be inherently flexible to form such a varied structure.

By having pentagons as well as hexagons, the capsid can form a closed structure, explain the researchers, describing the property the pentagons bring as "induced acute surface curvature". (A quick look at the structure of fullerenes, or even soccer balls for that matter, and you get an idea of what they are talking about).

Schulten says that knowing more about the detailed structure of the HIV capsid will help researchers understand how it functions, and this helps drug developers work out how to disrupt those functions.

He explains how the HIV capsid has to perform two opposing functions. It has to remain intact to protect its genetic material, but it also has to be able to release it in a timely manner once inside the host cell so it can replicate.

"That has to happen with really good timing - too quick is not good, too slow is not good. And this is a moment when you can throw a wrench into the system," says Schulten.

"The timing of the opening of the capsid is essential for the degree of virulence of the virus. This is where we could perhaps best interfere with HIV infection," he adds.

Funds for the study came from the National Institute of General Medical Sciences at the National Institutes of Health and the National Science Foundation, which also funds the Blue Waters supercomputer.

Earlier this year, scientists in the UK developed a vaccine against foot and mouth disease that uses a synthetic virus capsid to provoke an immune response.

To determine the structure of that virus shell, and identify mutations that would improve it, they used Diamond Light Source, the UK's national synchrotron facility.

Mosquitoes With Altered Smell Gene Lose Preference For Humans

By changing one gene, scientists have bred a mosquito that does not seek out the smell of humans in preference to other animals. The team behind one of the first successful attempts to genetically engineer mosquitoes believes their work not only shows what can be done with the latest genetic techniques, but also helps us better understand the insect's attraction to humans and therefore how to block it.
Lead researcher Leslie Vosshall, a Howard Hughes Medical Institute (HHMI) investigator at The Rockefeller University in New York, says in a statement:

"The time has come now to do genetics in these important disease-vector insects. I think our new work is a great example that you can do it."

"By disrupting a single gene, we can fundamentally confuse the mosquito from its task of seeking humans," she adds.

Vosshall and colleagues write about their work in a paper published online in Nature on 29 May.

Their report follows another study published recently in PLOS ONE, where researchers from the London School of Hygiene & Tropical Medicine in the UK describe how malaria-carrying mosquitoes are more strongly attracted to the smell of humans.

Starting Point Was a Gene in Flies

After scientists in 2007 announced they had sequenced the complete genome of Aedes aegypti, the mosquito that carries dengue and yellow fever, Vosshall switched her lab's focus from Drosophila flies to mosquitoes and set about trying to alter their genes.

From working with genetically engineered flies, she and her team already knew of a gene called orco that was important for the fly's sense of smell. So, as Vosshall explains, they started working on this gene in mosquitoes:

" ... we had some hints that mosquitoes interact with smells in their environment, so it was a good bet that something would interact with orco in mosquitoes."

Genetic Engineering Tools

To mutate the orco gene in Aedes aegypti, the team used "zinc-finger nucleases" (ZFNs), powerful tools that can be designed to target and cleave specific sequences of genomic DNA.

First, they injected ZFNs into mosquito embryos and when these matured, they sought out mutant individuals and used them to generate mutant strains so they could study the behavior of the orco gene in mosquitoes.

They discovered that brain cells linked to sensing odors were not as active in the genetically engineered mosquitoes. But they also found some other interesting changes.

Less Preference for Human Odor

Normally, when presented with a choice between humans and other animals, non-mutant Aedes aegypti mosquitoes fly toward humans, attracted by their smell.

But when Vosshall and colleagues gave their mutant mosquitoes a choice between human scent and that of guinea pigs, they did not show a preference for humans. This was the case even in the presence of carbon dioxide, which is supposed to enhance the attraction of mosquito to humans.

It appears that changing a single gene, the orco gene, disrupts the mosquito's ability to seek human prey.

However, this experiment did not establish precisely how the mutated mosquito lost the preference for human smell.

For example, did the mutated insect lose its ability detect that the guinea pig smell is not a preferred one, or did it lose the ability to discriminate that the human smell is the one to go for? Or did the altered gene cause both these changes?

Response to DEET

In a second part of their study, Vosshall and colleagues found that the mosquitoes with orco mutations were attracted to human skin even when it was protected by the common insect repellant DEET.

They exposed them to two human arms: one slathered in a solution of 10% DEET, and the other untreated. The insects flew equally to both arms, showing therefore that they could not smell the DEET.

However, once the mutant mosquitoes landed on the arms, they quickly flew away from the one slathered in DEET solution.

Two Different Odor-Sensing Mechanisms Identified

The team concluded that their experiments with DEET on human arms showed the mosquitoes are using two separate mechanisms to sense the DEET.

"One is what's happening in the air, and the other only comes into action when the mosquito is touching the skin," Vosshall explains.

There has been talk of a dual mechanism, but this is the first experiment to show it.

Vosshall's team now wants to explore how the orco protein interacts with the mosquito's smell receptors to shape its sense of smell.

"We want to know what it is about these mosquitoes that makes them so specialized for humans," she says.

"And if we can also provide insights into how existing repellants are working, then we can start having some ideas about what a next-generation repellant would look like," she adds.

In another recently published study, US researchers suggest it may be possible to use a bacterium that stops malaria parasites developing in mosquitoes.

Rice University Scientists Turn Soy Mash Into Succinic Acid

The humble soybean could become an inexpensive new source of a widely used chemical for plastics, textiles, drugs, solvents and as a food additive.

Succinic acid, traditionally drawn from petroleum, is one focus of research by Rice chemists George Bennett and Ka-Yiu San. In 2004, the Department of Energy named succinic acid one of 12 "platform" chemicals that could be produced from sugars by biological means and turned into high-value materials.

Several years ago, Rice patented a process by Bennett and San for the bio-based production of succinic acid that employed genetically modified E. coli bacteria to convert glucose into succinic acid in a way that would be competitive with petroleum-based production.

The new succinate process developed by Bennett, San and Chandresh Thakker and reported recently in Bioresource Technology promises to make even better use of a cheap and plentiful feedstock, primarily the indigestible parts of the soybean.

"We are trying to find a cheaper, renewable raw material to start with so the end product will be more profitable," said Thakker, a research scientist in the Bennett lab at Rice's BioScience Research Collaborative and lead author of the study. "The challenge has been to make this biomass process cost-competitive with the petrochemical methods people have been using for many years."

Bennett feels they have done that with soybean-derived feedstock as an inexpensive source of the carbon that microorganisms digest to produce the desired chemical via fermentation. "A lot of people use plant oils for cooking - corn or soybean or canola - instead of lard, as they did in the old days," he said. "The oils are among the main products of these seeds. Another product is protein, which is used as a high-quality food.

"What's left over is indigestible fiber and small carbohydrates," said Bennett, Rice's E. Dell Butcher Professor of Biochemistry and Cell Biology. "It's used in small amounts in certain animal feeds, but overall it's a very low-value material."

The Rice researchers are changing that with the help of E. coli bacteria engineered to process soy meal that generally gets discarded. Certain microbes naturally produce succinic acid from such feedstock, but manipulating E. coli's metabolic pathways (by eliminating pathways that produce other chemicals like ethanol, for instance) can make it far more efficient.

Expanding on their success in producing succinic acid from glucose, the new microbes are engineered to metabolize a variety of sugars found in soybean meal. The theoretical ideal is a 1:1 ratio of feedstock (the extracted sugars) to product, which they feel is achievable by industry. In the lab, under less controlled conditions, they still found the process highly efficient. "We're demonstrating a very high yield," Thakker said. "We're achieving in a flask a non-optimized formation of succinate that is close to the theoretical goal."

Bennett said his lab has been looking at soybeans for nearly three years. "We're always interested in low-cost feedstock," he said. "We were able to get a connection with a soybean group that is very interested in technologies to make better and more profitable use of their crop.

"There's a fair amount of oilseed residuals available, including cottonseed carbohydrates, that are not used for any high-value product, and we're in the space of microbial engineering to enable these sorts of materials to be used in a good way," he said.