Owing to the life-prolonging success of three-drug anti-HIV therapy - a cocktail of antiretrovirals and protease inhibitors - complacency may be creeping onto the AIDS-prevention scene. It's what the virus has been waiting for.

Molecular oncologist Steven Dowdy, Howard Hughes Medical Institute investigator at Washington University, in St. Louis, sounds a warning, and offers a remedy that he calls large-protein therapy. "The triple-drug regimen works very efficiently," he told BioWorld Today, "but the virus has the ability to mutate itself around the protease inhibitors, the same way it mutated itself around reverse transcriptase inhibitors. In the clinics now, we're starting to see the emergence of viral strains resistant to the current protease inhibitors. Although the infection rate is obviously going down in the industrialized world, thanks to the protease inhibitor cocktail, new infections will propagate viruses that have mutated the ability to bind to these inhibitors.

"Not that these protease inhibitors aren't effective," Dowdy hastened to add, "and aren't going to continue to be effective, but the virus mutates itself very rapidly. So, the pharmaceutical industry is now moving on to the next protease-inhibiting protein, integrase, which appears to be a very good target for small molecules.

"Picture those protease inhibitor molecules, which are currently so successful, as the size of an eraser on the end of your pencil," Dowdy said. "Your fingers are their target - the HIV protease enzymes. So, the inhibitors hang out in the structural hinge of the protease, and stop the virus from replicating. A major problem with the present protease inhibitors," he pointed out, "is their small size - about 650 Daltons, 800 maximum. This amounts to from three or four amino acids, up to six or eight at most. The inhibitor is so small because it must be able to permeate the infected cell. If it can't cross the cell membrane, it won't be able to carry out its functions.

"So, when you get above 800 Daltons," he explained, "you hit a wall, through which you can't get larger molecules into cells efficiently. We addressed this problem, and we can now put in proteins that go from that eraser size to the size of your fist. Our technology allows us to put in full-length proteins that are 30 or 65 or 110 kiloDaltons in size. They go in very rapidly, and into 100 percent of target cells - at least experimentally."

Dowdy is senior author of a paper in the January 1999 issue of Nature Medicine, titled "Killing HIV-infected cells by transduction with an HIV-protease-activated caspase3 protein." Its co-senior author is HIV virologist Lee Ratner.

Their article reports the first application, albeit in vitro, of what Dowdy called "the brave new world of protein therapy." Riding his bike to work one day, he recalled, "I thought, 'Why not use the viral protease to kill the infected cell instead of using a drug to inactivate the protease?'"

From Daydream To Killer Complex

That notion led him and his co-authors to construct a 290-amino-acid protein complex designed to make HIV-infected T cells commit suicide.

"That's a lot easier than trying to stop the protease," he said. "So, we took a naturally occurring human protein, caspase3, which is a pro-apoptosis factor with three molecular domains. It's normally cleaved by a cellular protease that clips it into three monomeric pieces. Two of these come together - dimerize - and force the cell to commit suicide. Once caspase becomes activated, it autoactivates other, inactive, caspase3 molecules to carry that domino-cascade forward."

To prime the caspase protein for its starring assisted-suicide role, the co-authors genetically engineered it into a transducible protein by replacing the front third of it with a protein transduction domain. Then, they replaced the infected target cell's endogenous proteolytic cleavage sites with that of an uninfected protease site, together with HIV recognition cleavage sites.

"Then we transduced this Trojan-horse protein into all cell populations, where it sat dormant until the HIV protease clipped it into thirds," Dowdy said. "At that point, two of the three caspase3 domains dimerized. So, now we'd got it into T cells. In a normal cell, nothing happened to it. But in an infected cell, the protease recognized the HIV cleavage sites and clipped them. This then forced the caspase to become active, and induce apoptosis in that HIV-infected cell. So, it's a very simple strategy."

Then came the caveat: Could the HIV get around this strategy? Not likely, Dowdy hypothesized.

"The virus has eight to 20 cleavage sites in its genome," he pointed out, "plus the specificity of the protease to recognize those sites. Now it's virtually impossible - physically improbable - that it could ever mutate all eight or 10 cleavage sites, plus the specificity of their protease, at the exact same time.

"But it can mutate itself against protease inhibitors, because they require only one or two point mutations," Dowdy said. "So, if there are HIV strains out there that, in fact, have divergent protease specificity, or divergent protease cleavage sites, then we can adapt our killing molecule to those strains by merely sequencing those cleavage sites, and placing them into our protein.

"Therefore," he continued, "we can adapt the system to kill other infectious diseases that require proteases for their life cycle, as does HIV. Protease is absolutely required for hepatitis C virus and malaria parasites. In theory, we could kill malarial infected cells, and treat now-untreatable hepatitis C infections.

Pending: Patents, Animal Trials, Drugs, Clinical Trials

The university, with five patent applications pending on the co-authors' protein-therapy technology, has licensed it to Idun Pharmaceuticals Inc., of La Jolla, Calif., and Life Technologies Inc. (LTI), in Rockville, Md.

"We carved it up into two exclusive fields," Dowdy said, "which are animals and above, for Idun; research and development - which is anything below animals - for LTI. So, if pharmaceutical or biotech companies would like to do this in-house, or develop their own therapies in in vitro models, they get a sublicense from LTI. If you're an academic and you want the protein's plasmid precursor, I've been giving them out free now for over a year."

He concluded: "If you move from target validation into an actual therapy, as a biotech or pharmaceutical [company], then you go to Idun for a sublicense to practice it on animals, in anticipation of clinical trials." *