Promise, Disappointment Mark AIDS Vaccine Quest : Science: Years of work by biotech firms produce two strong candidates. But results fall short and U.S. halts tests.


In the depths of a converted warehouse, stuck in the corner of a vast refrigerated room, sits a padlocked steel cage that rarely is opened. Locked inside are nine bottles of a crystal clear liquid that, scientists here believe, has the power to cripple the spread of AIDS.

This is the home of Genentech Inc., the father of biotechnology. The liquid is a protein called gp120--the key ingredient in a vaccine that, company researchers say, could prevent infection with the AIDS virus in as many as 60% of people who take it.

For the past two years, the protein has sat, untouched, in cold storage. There is enough here to make 998,013 doses of the vaccine.

But the vaccine will not be made.


After preliminary tests in about 200 people, a panel of independent scientists has concluded that the vaccine does not work as well as Genentech researchers say. In the laboratory, it has flunked a crucial test--it kills only viruses grown in a petri dish, not strains that come from infected people. Now, outside experts and AIDS activists want to wait for something better to come along.

And so in June, the federal government shelved further testing of its two most promising vaccines: the Genentech invention and a similar one manufactured by the Biocine Co. of Emeryville, Calif. The decision was a crushing acknowledgment that a vaccine to prevent AIDS--considered crucial to stemming the worldwide epidemic--may be at least another decade away.

And it is driving Phil Berman and Don Francis crazy.

Francis is a veteran AIDS warrior and a bit of a renegade--an impish, blunt-talking virologist who, while working for the U.S. Centers for Disease Control and Prevention, helped in the worldwide effort to eradicate smallpox with a vaccine. Now, he runs Genentech’s AIDS vaccine testing program--a program that, with the government’s recent decision, is going nowhere fast.


Berman is a self-effacing, 44-year-old molecular biologist whom Francis describes as “a wizard of modern technology.” He has spent the past decade toiling in his Genentech laboratory, trying to work his technological magic into an effective vaccine.

The effort has been marked by disappointment, triumph and disappointment again.

There were experiments that worked in guinea pigs but not on chimpanzees--a failure that prompted the company to pull the plug on the project in 1986, just a year after it began. So Berman and his partner, Tim Gregory, forged ahead on their own time, spending two agonizing years coming up with a better formula.

Then there was Christmas, 1988, when this scientific duo sweet-talked Genentech vice presidents into letting them experiment with chimps that had been purchased for another project. Genentech had paid $500,000 for the rights to use the chimps, but the animals were in Texas and the rights were to expire New Year’s Day.


In a mad dash, Berman and Gregory shipped the vaccine to Texas, and the chimps got vaccinated just in the nick of time. When the results came in, they rejoiced--the animals were protected against HIV, and Genentech resurrected the vaccine.

At every low point, Berman and his crew went back to the lab. At every high, they treated themselves to a bottle of champagne. “Veni, vidi, vaccini, " their company T-shirts proudly proclaim. “I came, I saw, I vaccinated.”

But there are no champagne toasts at Genentech these days.

Instead, there is frustration and anger--at other scientists, activists and officials of the National Institute of Allergy and Infectious Diseases who, Berman and Francis say, lack the courage and political will to run a field trial that would test their invention on thousands of people, thus answering once and for all the question of how well it works--or whether it works at all.


“They totally missed the boat,” Berman says of the AIDS Research Advisory Council, the independent panel that on June 17 recommended that the National Institute of Allergy and Infectious Diseases abandon the tests. “Fifty thousand people are becoming infected each year. If we could cut that down, even to 35,000, I think that would be a wonderful thing.

“If we’re ever going to have a vaccine, we are going to have to test this product sooner or later. We’ve built this airplane. Now we have to see how it flies.”

But the trial would be expensive--as much as $30 million or more, depending on its length and size--and might require as many as 10,000 volunteers. If the vaccine flopped, it would be a colossal embarrassment for the government, and that could make it more difficult to sign up people for future tests--an argument Francis dismisses as ridiculous.

An epidemic is raging, he insists. Every day without an AIDS vaccine is another day added to the human and economic costs of this devastating disease. In this country alone, every new infection will cost an estimated $100,000 to treat. We cannot, he says, afford to wait.


Moreover, Francis argues, private companies such as Genentech will not foot the bill for research if the government won’t put up the money to test their products. The AIDS research council meeting made his blood boil. When it broke up, he collared a committee member.

“Do you have any idea what you did?” he blurted out. “You have essentially killed the vaccine initiative in the United States.”


Such are the passions and pitfalls of developing a vaccine to protect people from infection with one of the trickiest viruses known--the human immunodeficiency virus, HIV.


“People say this is the end of a vaccine for AIDS,” says Dr. Anthony Fauci, who as director of the National Institute of Allergy and Infectious Diseases made the final decision not to conduct the expanded clinical trials. “It’s not. (But) we have a long way to go before we can feel scientifically very comfortable about going ahead with a vaccine.”

The point of a vaccine is to fake out the immune system, which functions by recognizing and destroying foreign invaders. In essence, vaccines are impostors; they contain just enough virus to trick the immune system into thinking it has seen the enemy. When they work, the body teaches itself how to kill the virus. And when the real enemy comes, it is ready.

The standard formula for most vaccines--smallpox, polio, measles--is straightforward, employing a weakened, or attenuated, form of the live virus. Such “live attenuated” vaccines can cause a minor infection--enough to teach the immune system its lesson, but not enough to cause disease.

After studies in monkeys last year at Harvard University, some experts think the live-attenuated approach is the most promising for preventing AIDS. But there is a problem: Because HIV mutates so easily, any vaccine that used the whole virus--even in a weakened form--could be dangerous. What if it mutated back into a deadly strain, infecting the person it was supposed to protect?


For this reason, scientists and government officials discounted the live-attenuated formula in their early vaccine efforts. Instead, after the discovery of HIV was announced in 1984, vaccine development fell to biotechnology firms such as Genentech.

Through modern gene-splicing techniques, scientists such as Berman can snip off crucial parts of HIV--the protein gp120, for instance--to fashion vaccines. These “subunit,” or “recombinant DNA,” vaccines cannot cause AIDS because they do not contain the disease-causing genetic material of the virus.

But it may take a lot of tinkering to get them to work.

“If only we were dealing with a virus that wasn’t so mutable,” sighs Dr. Steven Wolinsky, an AIDS researcher at Northwestern University. “It’s much easier to shoot at a target that remains in place.”


Whatever vaccine is developed, experts say, it will likely evolve over time. Unlike the polio vaccine, which worked right off the bat and is effective after a single dose, an AIDS vaccine may take years to reach a point where it works for everyone. Berman likens it to automobile safety; first there were lap belts, then shoulder belts and now air bags. Each step improved a driver’s chances of surviving a crash. AIDS vaccines, he says, are still in the lap belt stage.

Moreover, there is tremendous disagreement among scientists about just what it will take to make a vaccine work. Still unknown are the “correlates of immunity"--the precise immune system responses that a vaccine must induce to keep HIV in check.

Some think the answer lies with antibodies--custom-made proteins that attach themselves to the virus, rendering it harmless. The Genentech vaccine primarily produces antibodies, although studies indicate it also prompts a mild form of a different kind of immunity, called “cellular immunity.”

In cellular immunity, killer immune cells recognize and destroy other cells infected with the virus. Some experts think this is the key, and several vaccines now being tested rely on this theory; they work by sending HIV proteins into the body on the backs of viruses such as canarypox or vaccinia.


Still others think protection rests with “mucosal immunity,” in which the virus is attacked when it passes the mucous membranes that line the entrances to the body.

And some think any vaccine must produce all three.

“The whole (scientific) community seems to be fragmented right now as to what they think a vaccine is supposed to be,” says Dani Bolognesi, a highly regarded AIDS vaccine expert at Duke University who serves on the AIDS research council. “None of us know what it is going to take to do this.”

Some AIDS researchers think more laboratory studies are necessary to answer these questions before designing a vaccine. On the other hand, Bolognesi notes, there are the old-time vaccinologists--veterans of the polio epidemic--who say all the lab work in the world won’t take the place of testing a vaccine in people.


And who cares precisely how it works? The important thing is that it works.

Says Bolognesi: “The vaccinologists would say, ‘You could work until the cows come home on these issues. You’ll never solve them. You’re going to have to put your toe in the water and find out.’ ”

This is precisely Francis’ point. Above his desk, he keeps a quote from Maurice Hilleman, a prominent 74-year-old virologist who sits on a panel that evaluated the Genentech vaccine in April.

“This vaccine,” Hilleman declared then, “is not bad enough not to test.”



Phil Berman thinks his vaccine is a lot better than not bad.

Berman came to Genentech in 1982, when the company was just 6 years old and a stock market darling. Genentech’s story is well-known in the world of science: In the 1970s, a venture capitalist named Robert Swanson got the idea that the new recombinant DNA technology could be used to make pharmaceuticals.

Swanson hooked up with biochemist Herbert Boyer of UC San Francisco, and over beers at a downtown bar, they agreed to became partners, each putting up $500 to found Genentech in 1976. The company became so successful so fast that in 1980, when it went public, the stock price more than doubled in less than an hour.


Although it has since fallen on harder times, Genentech remains one of the world’s biggest biotech companies, with revenues last year of $649 million and 2,500 employees who occupy a sprawling industrial complex on the wind-swept edge of San Francisco Bay.

The freewheeling atmosphere upon which Genentech was founded still exists. Employees wind down on Friday nights with company-hosted parties. Each year, there is a “ho-ho,” a Hawaiian theme event. Swanson, now chairman of the board, routinely turns up in drag--flowered bikini top, grass skirt and blond wig.

Berman didn’t come for the party. He came to do good science.

Having worked in Boyer’s lab at UC San Francisco, he was well-schooled in genetic engineering. But Genentech was trying something novel. While most recombinant DNA technology employed yeast or bacteria to manufacture proteins, Genentech hoped to use cells from mammals. These mammalian cells, from the ovaries of Chinese hamsters, are more similar to human cells, and thus more effective at producing the proteins. But nobody thought it would work.


“The thing that knocked my socks off was the possibility of making recombinant proteins in mammalian cells,” Berman says. “It was an outrageous idea at the time.”

But it did work, and it has been critical to the making of Genentech’s AIDS vaccine.

The main component of the vaccine, gp120, is especially difficult to reproduce. It is a chain of 510 amino acids, 26 of them linked to complex carbohydrates, or sugars, and it has nine cross-links forming loops.

The protein is like an origami bird that must be folded just right for its wings to move. Its sugars and cross-links must be in just the right place for it to “fold” in the unique way that binds it to HIV. The trouble is, nobody knows what it really looks like folded up.


But what Berman does know is this: Inside the Chinese hamster ovary cells, gp120 folds like a dream.

By 1985, just a year after the announcement of HIV, Genentech had made enough of the protein to immunize guinea pigs and rabbits with the first version of its current vaccine, called rgp120 (the r stands for recombinant.) Early tests showed that blood taken from the vaccinated animals killed the virus.

That marked the start of the champagne toasts. “It was,” Berman recalls, “a great day.”

Since then, there have been several different incarnations of the vaccine and a host of experiments--on guinea pigs, chimpanzees and, ultimately, healthy uninfected people. The first of the human trials began two years ago, a small test with just 57 volunteers--39 women and 18 men.


The volunteers received four doses of the vaccine--an initial shot and three booster shots, given over 48 weeks. The results--to be reported this week at the 10th International AIDS Conference in Yokohama, Japan--were extremely encouraging, according to Dr. Robert Belshe of St. Louis University’s School of Medicine, where the study was conducted.

Researchers found that the vaccine was safe and that it prompted an immune response against the virus. Lab tests showed that 46 volunteers developed antibodies to HIV after three injections. After the fourth injection, more than one-third of the volunteers also showed cellular immunity.

The early findings were so promising that last year the government launched “Phase II” clinical trials of the Genentech and Biocine vaccines in 296 patients. But these studies, along with other research on these vaccines, have created some doubts.

There are two main sticking points--and Berman and Francis have answers to them both.


First, a total of six people who were vaccinated with either the Genentech or Biocine products have become infected with the AIDS virus. No one is suggesting that the vaccines caused the infection, but many scientists see it as a signal that the shots do not work.

Francis, however, says such “breakthrough infections” are to be expected. They occur because the volunteers have not received the full course of shots and have engaged in risky behavior.

Second, and far more troubling to most experts, is the issue of the “field isolates.”

Although scientists can deliberately inject the deadly AIDS virus into animals, they cannot do so in humans. So they must rely on lab tests in which a vaccinated person’s blood is “challenged"--mixed with the virus in a test tube.


These tests show that the blood of people who have gotten the gp120 vaccines can kill strains of virus grown in a laboratory. But their blood is not able to kill “field isolates"--strains of wild virus taken from infected people.

Berman argues that those lab tests are not reliable. He says the isolates, although drawn directly from people, are then grown in unnatural laboratory conditions that use potent chemical activators, which give the virus advantages it does not have inside the body. Moreover, he points to studies of vaccinated chimps, which show that the animals can fight off the wild virus.

Others are not so convinced.

“From what we know about these vaccines, you would not predict that they would have a high level of efficacy,” says Dr. Ashley T. Haase, a University of Minnesota researcher who chairs the AIDS research council. “They simply don’t neutralize the strains that are out there.”


Haase says the best course is to wait until other vaccines get further along in the testing process, and then test those, possibly in combination with the Genentech or Biocine vaccines.

Top officials at Genentech, meanwhile, are not optimistic.

The company has invested $150 million in AIDS research and has not been able to bring a single product to market, according to G. Kirk Raab, Genentech’s president and chief executive officer. Raab says that unless Berman and Francis come up with something that looks much more promising than rgp120, the company is not likely to continue to invest in AIDS vaccine research.

“Other scientists don’t think it will work, there’s no funding for it and I’m not sure that a vaccine that is effective 60% of the time is a good idea medically,” Raab says. Of Francis and Berman, he adds: “I appreciate and admire their enthusiasm, but I think the situation is more complicated and less black and white than they see it.”


Berman, however, seems undeterred. He is busy in his lab, trying to add other strains to the vaccine formula to make it work better. He shrugs when asked if he feels gloomy.

“There have been a lot of setbacks,” he says philosophically. “The thing about AIDS research is everything can turn on a dime.”

On his desk, meanwhile, sits an unopened bottle of champagne.

The Making of rgp120


Like a number of the AIDS vaccines in various stages of human experimentation, the Genentech rgp120 vaccine is formulated through sophisticated genetic engineering, or “recombinant DNA,” techniques. Here is how it is made:

STEP 1: Laboratory scientists snip the gene that encodes the crucial protein--gp120--from the human immunodeficiency virus. The gene is inserted into a plasmid--a small, circular piece of DNA--that enables scientists to alter it to grow in mass quantities.

STEP 2: The plasmid goes into a culture of Chinese hamster ovary--or CHO--cells, which incorporate the gene, in essence becoming microscopic factories to make gp120. A cell especially good at making the protein is plucked out.

STEP 3: The cell--called an “overproducer"--is put into a mixture of nutrients and grown in a petri dish, where it makes copies of itself. The copies--up to 1 million cells per milliliter of nutrients--are transferred into tiny containers called ampuls, where they are frozen in liquid nitrogen.


STEP 4: Manufacturing begins. Cells from an ampul are thawed out and “seeded” into a larger vessel to grow in another nutrient mixture. The cells grow for a month and are transferred to Genentech’s giant “fermentation hall"--a vast room of stainless steel fermenters connected by miles of pipe.

STEP 5: Fermenters gently agitate the fragile cells, which must be blended with just enough oxygen to grow. All the while, they are making the crucial protein. Sterile conditions are a must; a single bacterium could destroy the process in a day.

STEP 6: After fermenting for many days--the exact time is a trade secret, Genentech says--the protein is “recovered.” Other proteins secreted by the CHO cells are removed, leaving the pure gp120.

STEP 7: The pure protein--a crystal clear liquid--is ready to make into rgp120 vaccine. (The “r” stands for recombinant.) The protein is shipped to a Genentech manufacturing plant, diluted with a buffer solution and packaged in vials.