By November 2003, 40 million people worldwide – 5 million more than the year before – were infected with HIV. In 2003, three million died of AIDS, bringing the total number lost to the epidemic to nearly 32 million people, the size of the population of Canada.
This insidious disease continues to prove itself. When this virus turns on, modern medicine can attack and kill, but it cannot cure. HIV hides. It slips inside other cells and waits. It can wait in reservoirs for years, probably longer than a person would live after contracting HIV. So scientists are trying to lure this killer out of hiding, turn it on, and destroy it. So far, some latent virus persists no matter what physicians throw at it, because today's drugs cannot detect the hidden virus. Moreover, HIV does not just infect cells: It integrates into the host cell's chromosome and DNA.
"I think that HIV hides in every place that it can," says Dean Hamer of the National Cancer Institute, "and it's pretty clever at doing so." He points to evidence from France Pietri-Rouxel of the Cochin Institute in Paris that HIV might even hide in fat cells.1 Mulling over all the reservoir possibilities, Hamer concludes: "I would say any cell that's living and breathing and has CD4 is a target for HIV."
Scientists have found the virus in macrophages, monocytes, dendritic cells, and of course, CD4+ T cells. Finding every place that HIV hides, however, depends on how adept scientists become at uncovering it. Today's methods can find almost any amount of HIV; the problem is learning how to eradicate it (see related story p. 20). It may be that HIV cannot be wiped out. "It's like most of internal medicine," says Roger Pomerantz of Thomas Jefferson University in Philadelphia. "We don't cure much of anything." Diabetes, heart disease, and hepatitis, like HIV, all remain uncured. Some scientists think that defeating HIV will demand a new approach.
A SAD TALE
What they do know is that HIV binds to CD4 receptors on the surface of T lymphocytes. The infected T cells work as virus factories for HIV, which soon kills the cells. With enough T cells destroyed, a person cannot mount an immune response, so almost anything foreign, such as a bacterium, fungus, or another virus, poses a serious threat. The person infected with HIV usually dies from such opportunistic infections. Ongoing work, however, reveals many more players, making elimination of this virus an immense challenge.2
To keep track of the HIV infection, scientists measure the amount of viral mRNA in patients' plasma. Modern clinical techniques can detect as few as 50 copies of mRNA per milliliter; smaller amounts are considered undetectable. Moreover, the well-known drug regimen HAART (highly active antiretroviral therapy) can reduce HIV to 50 copies or less in many patients. When asked if HIV could be maintained at low levels and provide good outcomes for patients, the University of Barcelona's José Gatell replies, "This has been clearly demonstrated, at least for limited periods of time, six years." He adds, "Any potent regimen of HAART can achieve this goal, if the tolerance is reasonably good and the compliance to prescribed medication is above 90% to 95%."
If a patient goes off HAART though, the virus resurfaces, sometimes rebounding in just one week to predrug levels or higher. Some HIV probably remains in the blood; so if just one copy lingers, it rushes into reproduction as soon as HAART stops. And then there are the reservoirs, the places where HIV replicates little or not at all, going unnoticed by HAART. Those reservoirs have been found to cover more ground than was expected.
STRATEGIES TO REDUCE LATENT RESERVOIRS OF HIV
The first step is to induce the expression of the quiescent HIV genome in latently infected resting CD(4)T cells using agents that activate transcription from the viral promoter. This is done in the presence of anti-retroviral drugs (HAART) to prevent spreading infection by newly synthesized virus. The second step is to destroy the HIV+ cells, either by action of the virus itself, through the immune system, or by chimeric toxins that recognize the viral envelope glycoprotein Env present on the surface of infected cells.
TAKING THE PULSE
Scientists have turned most often to T cells as the reservoir source, not only because HIV attacks T cells, but also because the T cells' normal function makes them possible reservoirs. Says Mario Stevenson of the University of Massachusetts Medical School, "There's a T cell that is essentially in a dormant state. It's called a quiescent T cell." Those cells remember past infections and mobilize rapidly if the same infection returns. Otherwise, this T cell sits and waits for years, even decades, mostly in tissues and lymph nodes.
That long life creates a problem when it involves HIV. If HIV finds its way into a quiescent T cell, then it should be able to live there for decades. (HIV kills cells, but not the quiescent T cells.) Instead, Stevenson says, "The popular theory is that a quiescent cell is in a state of dormancy. So when HIV gets in a cell that enters dormancy, HIV goes into dormancy along with it." Scientists can study HIV-infected, resting T cells and not find virus. But, if they stimulate those T cells so they become turned on, HIV starts replicating. "Basically, the virus has its finger on the pulse of the T cell," says Stevenson.
The T-cell reservoir could be very small, maybe as few as 100 cells. Nonetheless, experiments from Robert Siliciano's lab at the Johns Hopkins University School of Medicine put the half-life of those cells at about four years. It could take 60 years of HAART to eliminate all the virus hiding in the T-cell reservoir.3 Even worse, the reservoir is not limited to T cells.
OF MONOCYTES AND MACROPHAGES
Dormant cells are not the only places where researchers are having a tough time drawing out the HIV virus. Other white blood cells, the monocytes, also play a role in HIV. Suzanne Crowe of the Macfarlane Burnet Institute for Medical Research and Public Health in Melbourne, Australia, explains that monocytes in the blood can be considered as members of two populations.4 Two surface antigens, CD14 and CD16, distinguish the monocyte populations. One group expresses lots of CD14 proteins on its surface and very few CD16 proteins. These monocytes comprise most of the population, and tend to resist HIV infection. The other segment sports few CD14 proteins on its surface but many CD16s; this latter group expresses a receptor called CCR5, which binds HIV.
To explore this further, Crowe and her colleagues collected monocytes from HIV-infected patients. Then, the scientists divided the monocytes into two populations, one high in CD16 and one high in CD14. Crowe says, "Preliminary data show that the monocytes high in CD16 are more susceptible to HIV infection both in vivo, in blood from infected patients, as well as ... in vitro." Moreover, the high-CD16 monocytes may go places that the other population cannot easily reach, such as the brain. So, the high-CD16 monocytes could get infected with HIV and then hide where drugs might not go, contributing to the HIV reservoir problem.
As monocytes mature, they become macrophages, living for about two weeks in some tissues. But they may survive for months – maybe decades – in other places, including the brain. HIV-infected macrophages turn up in the brain, lung, lymph nodes, and spleen. Worse still, macrophages make lots of virus. Even if infected T cells outnumber the infected macrophages by as much as 100-fold, it does not ensure that the T cells produce most of the virus. Because the infected macrophages do not become dormant like a resting T cell, the macrophages keep producing virus.
Monocytes also make dendritic cells which help the immune system recognize foreign antigens. Dendritic cells show up at every possible entry point: the oral cavity, intestine, vagina, and rectum, where they pick up foreign particles and present them to T cells and B cells, eliciting an immune response.
HIV-infected dendritic cells pose real trouble. Thomas Hope of the University of Illinois at Chicago and his colleagues studied the interaction between HIV-infected dendritic cells and T cells.5 By using time-lapse microscopy, Hope watched dendritic cells come into contact with T cells. In one experiment, HIV was distributed evenly over a dendritic cell before contact. Then, as the dendritic cell spread out over a T cell, the HIV moved to the place where the two cells first touched. In addition, a series of HIV receptors, including CD4, CCR5, and CXCR4, on the T cell also moved to that site. So, when a dendritic cell contacts a T cell, the HIV virus and its receptors become correctly positioned for a transfer to occur.
No matter how much the HIV moves around, finding it depends first on very sensitive tests, and those already exist. Says the University of Minnesota's Ashley Haase: "If you had a single copy of the virus in a cell, you could certainly detect it. The problem is really sampling." For example, he points out that about 60% of the secondary lymph system is in the gut, and that covers a lot of area. "You might hit something, you might not." From extensive sampling, Haase concludes that T cells, monocytes, and macrophages hold the majority of the latent HIV. He agrees that dendritic cells hold some virus, too. Nonetheless, he adds, "Most of us think that most of the latently infected cells are resting CD4+ T cells."
MASTER OF THE GAME
HIV hides in many places. Once it crosses the body wall, HIV quickly invades quiescent T cells and these cells slip away to tissues and lymph nodes. HIV also infects monocytes, which may hide in various organs, including the brain. As monocytes mature they make macrophages, which also can be infected.
SOME EXPERIMENTAL THERAPIES
With HAART, clinicians can kill HIV in active cells, but the dormant cells cause the worst treatment problems. In one attempt to attack the HIV reservoir, Pomerantz and his colleagues studied three HIV-infected men who had been on HAART for more than a year. The men had no detectable virus in their blood samples, meaning fewer than 50 copies per milliliter of plasma.6 Then, Pomerantz gave these patients two drugs, didanosine and hydroxyurea, which knock down viral replication even in resting cells, and two other drugs, OKT3 (an anti-CD3 monoclonal antibody) and interleukin-2, to turn on resting T cells. This treatment cut the viral load to less than one copy per milliliter. Every indicator showed no virus in these patients, so Pomerantz offered them drugless trials. "The good point," Pomerantz says, "is that they lasted significantly longer than the usual seven to 10 days before they rebounded. One lasted six months. Two rebounded after six weeks, which is still much, much longer, but they all came back. So we got close, but we didn't get it to zero."
Hamer and Jerome Zack of the David Geffen School of Medicine at the University of California, Los Angeles, put together a research team to try another tactic, immunotoxins.7 The immuno component of their immunotoxin recognized a specific glycoprotein in the envelope of HIV, and the
In short, the HIV reservoir escapes virtually all available therapies. Anthony S. Fauci, director of the National Institute of Allergy and Infectious Diseases, says, "A number of studies throughout the years – from our lab, from Robert Siliciano's lab, from other labs – have indicated that, despite the suppression of virus replication for up to several years with antiretroviral therapies, when you look for the reservoir, you inevitably find it, virtually without exception. And when you discontinue the drug, in most everybody with very few exceptions, the virus bounces back." As a result, HIV currently demands a course of antiretroviral therapy that continues indefinitely, but that, too, might fail (see sidebar).
The question is whether the localization of HIV particles was altered when dendritic cells were observed shortly after contact with adherent, CD(4)-positive cells. The answer is yes. In the first frame, the virus was evenly distributed throughout both dendritic cells. Within six minutes, the cell in the top of the frame began to spread out on the target, and the majority of the HIV relocated to the initial site of contact. The researchers observed movement in the other cell at 18 minutes. In both cells, the majority of the particles moved within one, three-minute time frame.
Instead of looking only to drug treatments, some scientists look at the human immune system itself: Perhaps HAART could reduce the virus, and then the immune system could clean up what's left. For example, a nucleic-acid-editing enzyme called APOBEC3G is made inside HIV-infected cells. This enzyme deactivates viruses by changing the nucleic acids in the viral DNA. In HIV, APOBEC3G converts cytosine to uracil in the viral DNA that is generated during reverse transcription, thereby blocking further infection.
Nevertheless, "HIV almost never gets killed by APOBEC3G," emphasizes Didier Trono of the University of Geneva.8 "Instead, APOBEC3G edits the genome of the virus a little bit, which makes HIV even harder to fight." Trono's group analyzed one variant of APOBEC3G that apparently leads to a faster progression to AIDS. "We found this variant slightly more active," Trono says. "It might accelerate disease progression by inducing a faster genetic drift for HIV because of a stronger APOBEC3G." Instead of APOBEC3G helping in the fight against HIV, it might make it worse. "This antiviral factor and its interplay with HIV probably contributes to the reservoir's diversity and difficulty to track down," says Trono.
To get a better view of how HIV works, perhaps scientists need to work less in vitro and more in vivo. Several years ago, Haase and his colleagues discovered that resting T cells produce some virus, but this appears only in vivo, not in tissue culture.9 To explore the very beginning of HIV infection, he turned to the simian immunodeficiency virus found in monkeys. "When the virus first enters a host," Haase says, "it does not see lots of activated T cells. It sees many resting T cells. In contrast to a lot of people, I think the latent cells get infected very early on." Haase says he believes that the virus goes into a reservoir right away. Moreover, his work shows that even the reservoir replicates a small amount of virus. "It's diabolical," Haase says. The reservoir reproduces just enough to trigger an immune response that helps the HIV infect activated T cells, where the virus replicates most efficiently.
HIV continues to slip out of every trap that scientists create. "What we know," concluded Stevenson last year, "represents only a thin veneer on the surface of what needs to be known."2