A Blog by

Herding HIV into an evolutionary dead end – study finds the virus’s weak spots

“The discovery of HIV enables us to develop a vaccine to prevent AIDS in the future… We hope to have such a vaccine ready for testing in approximately two years.” – Margaret Heckler, Secretary of Health and Human Services, April 23, 1984.

Twenty-seven years have passed since Heckler’s comment and no vaccine exists. There is a simple reason for this: HIV out-evolves us. HIV can produce around 100 billion new virus particles every day, and it does so with unusual imprecision. When most genetic material is copied with great fidelity, HIV goes for a sloppier approach. It duplicates itself with errors galore, creating a swarm of genetically variable viruses. It leaves a host looking very different to when it entered.

In the face of this rapid shape-shifting, any drug or vaccine soon becomes obsolete. Fighting HIV is like fighting a hydra – there are several heads and every time you lop one off, two more grow in its place.

To tackle this continually changing adversary, some scientists are looking for parts of HIV’s proteins that tend to stay the same. These “conserved sites” tend to be important in some way; mutating them would compromise the virus’s ability to reproduce. By training the immune system to attack these sites, we would give the virus an unenviable choice – do nothing and die, or change and become weaker.

But there is a third option. The virus could alter a different part of its protein so the conserved site is no longer important. That way, it can change the site and escape the immune system’s attention, without doing itself any harm. HIV gets away again.

Vincent Dahirel and Karthik Shekhar from MIT think we need a different approach. In a study led by Arup Chakraborty and Bruce Walker, they have identified large groups of conserved sites on HIV proteins, which they call “HIV sectors”. These aren’t just sites that stay the same; they stay the same en masse while the rest of the virus warps around them. Dahirel and Shekhar’s idea is to train the immune system to attack all of the sites in a sector. To escape these “multiple points of immune pressure”, HIV would have to develop many different mutations that, together, would almost certainly cripple it.

This approach is a new one. “It came from bringing new blood into the HIV field,” says Walker, who has been studying HIV almost since it was discovered. “When we described some of the challenges in the field, Arup Chakraborty thought we were taking too narrow a view of evolution and came up with this approach.”

To identify the sectors, Dahirel and Shekhar used a technique called “random matrix theory”, which detects groups of objects that change in a concerted way. It has its origins in high-energy physics, but economists have also used it to study fluctuating stock prices. The team used it to analyse HIV proteins, using sequences taken from a massive database of patients. “It allows people to assess how one stock might be linked to the performance of other stocks,” says Walker. “In the same way we have asked if evolution at one site is at all linked to evolution at other sites. And it is.”

They focused on Gag, the protein that makes up HIV’s outer coat. If any protein is an inviting target for the immune system, it’s this outwardly facing one. Dahirel and Shekhar identified five sectors within Gag, each consisting of sites that evolve together, independently of each other and the rest of the protein. One of these – Sector 3 – was particularly constrained. At these sites, groups of mutations are very rare indeed.

The team soon found out why. The Sector 3 sites are important for binding different Gag proteins together. Without them, the virus wouldn’t be able to assemble its outer shell. If one of these sites changes, the virus can probably cope. If several of them change, the result is a naked, hobbled virus.

Dahirel and Shekhar reasoned that if the immune system targets several sites in Sector 3, HIV will find it difficult to evolve its way out of trouble. It’s an attempt to beat a virus that evolves too quickly by herding it into an evolutionary dead end. And, as it happens, that’s exactly what some people have already done.

Around one in every three hundred people infected with HIV never go on to develop AIDS. These “elite controllers” can live with the virus throughout their long, healthy lives, without the need for medication. Dahirel and Shekhar found that their immune cells disproportionately target parts of the Gag protein that belong to Sector 3. This helps to explain what makes the controllers special. To survive their immune assault, their viruses have been forced to mutate in ways that make them weak and unfit, making them easier for their hosts to control.

The study is a wonderful collision of theoretical number-crunching and real-world experiments. “What we show in this paper is theoretical,” says Walker, “but it goes one step further in that it has been tested against human data in the elite controllers. This makes us more confident.”

That’s fine for the elite controllers, but the team think that it should also be possible to get other people to mount a similar defence. You’d need to expose the immune system to fragments of proteins that mimic parts of Gag, including as many of the Sector 3 sites as possible. They system should develop a memory for these shapes and during a natural infection, they would “hit HIV where it hurts early”. The team are now working with other HIV researchers to turn these ideas into reality. Walker says, “We need now to make immunogens [substances that trigger specific immune responses – Ed] with this new technique and see if this gets over the major hurdle we face in an HIV vaccine – that of viral diversity and viral evolution to escape immune responses.”

This is one of a growing number of studies where scientists are trying to control viruses by studying how they evolve, and by looking at combinations of mutations rather than individual ones.

For example, Joshua Plotkin’s group at the University of Pennsylvania is looking at ways of predicting the evolution of flu viruses by looking for mutations that occur in tandem. They could compensate for one another, or boost each other’s effects; either way, the whole is greater than the sum of the parts. Plotkin is trying to use these interactions to identify seemingly innocuous mutations that herald the arrival of more serious ones – to develop a crystal ball for the future of flu.

Plotkin is impressed with Dahirel’s approach. He says, “This paper, along with several others works in recent years, exemplifies the practical value of deciphering the rules that govern evolution, especially the evolution of microbes and viruses.”

Reference: Dahirel, Shekhar, Pereyra, Miura, Artyomov, Talsania, Allen, Altfeld, Carrington, Irvine, Walker & Chakraborty. 2011. Coordinate linkage of HIV evolution reveals regions of immunological vulnerability. PNAS http://dx.doi.org/10.1073/pnas.1105315108

Image by AJC1

More on HIV and protein evolution: