A new mathematical model of HIV-fighting drugs reveals the biology beneath the varying success of such treatments. Infection with HIV was a death sentence until the introduction of multidrug cocktails, yet the differential effectiveness of the combinations has remained a puzzle. The research, published July 13 in Science Translational Medicine, could help refine therapies for HIV and other viruses such as hepatitis C.
Slightly boosting the dose of some HIV drugs has a profound effect if those drugs are attacking multiple targets, the new model reveals. Finding that more bullets can kill more targets may seem obvious, says AIDS researcher and Howard Hughes Medical Institute investigator Robert Siliciano of Johns Hopkins University. But the realization required a shift in thinking about a very old idea: the relationship between a drug’s dose and its effect.
For centuries, drug effectiveness has been visualized with what’s called the dose-response curve. This relationship often takes on a stretched-out “S” shape when graphed.
But in 2008, Siliciano’s then graduate student Lin Shen realized that the steepness of the incline of the “S” — its slope — varied with different classes of HIV drugs. A gradual climb meant that increases in drug concentration gradually improved the response. But a very steep slope meant that tiny increases in a drug’s concentration could wipe out significantly more target molecules.
That the steepness of this slope mattered for HIV drugs was puzzling, says Siliciano. “The differences were huge — orders of magnitude,” he says.
For example, increasing the dose of the most effective protease inhibitors, drugs that block an HIV protein that snips up virus parts for assembly, can make them billions of times more powerful against the virus, he says. But increasing the amount of the drug AZT, which attacks virus machinery that translates genetic material, might yield an effect only 10 times greater than the lesser dose.
Incremental increases in dose that yield a vast improvement in response is a phenomenon that usually happens with drugs that attack a target molecule at multiple sites, an effect known as cooperative binding. Yet HIV has only one site that drugs can latch onto, so more of a drug shouldn’t necessarily be more effective.
But the researchers realized that at certain times in the HIV life cycle, there is so much virus or viral machinery to attack that the drugs are cooperatively binding, but to many, many targets rather than many sites on one target.
“It wasn’t obvious,” says Siliciano. “Looking at what drug concentration gives you 50 percent inhibition is very linear thinking. But the virus replicates exponentially. Each infected cell releases enough virus to infect 10 more cells. So we had to think in those kinds of terms.”
The new model takes into account that some drugs enter the battle during parts of HIV’s life cycle when infection is halted only if a critical number of targets are killed.
The team tested the model by creating viruses that would offer a different number of targets than usual, such as viruses that didn’t crank out their usual number of protease enzymes. When the team infected kidney cells with these altered viruses and calculated the dose-response curves, the slopes were different than those for normal HIV. If the altered virus had fewer working enzymes for the drug to disable, the virus was inhibited with a lower dose, the team found.
The concepts outlined in the new model don’t just shed light on fighting HIV but could also inform efforts to attack other viruses, such as hepatitis C, says HIV researcher and clinician Steven Deeks of the AIDS Research Institute at the University of California, San Francisco, who coauthored a commentary on the new work.
“HIV is a replicating machine,” says Deeks. “It mutates constantly and the immune system is ineffective at controlling it. We’ve never understood, given how effective the virus is at doing what it does, why these drug combinations have worked for so long.”
Now that math has revealed the secret of these drugs’ success, the word needs to get out, says Deeks. “The math is so dense,” he says. “I finally understand it.”
