For most people, HIV infection, if not treated, causes a long, slow decline in immune capacity to a point when they become susceptible to dangerous opportunistic infections. In some, this decline can happen within a few years; for a few, it hasn't happened yet after 20 years. In the pre-treatment era, the average time for progression to death was about 11 years. Nowadays, this can be delayed by successful suppressive antiretroviral drug therapy, although, again, some people have terrific, trouble-free results while others never manage to get the full benefits of treatment.

This variability in the course of HIV disease and in the success of treatment can be due to a number of complex, intertwining factors. An individual's genetics may come into play: a small number of people lack crucial receptors that HIV uses to infect new cells and there are surely many more host factors involved that we don't know enough about. Different viral genetics can make a huge difference: a flawed HIV protein called Nef makes for a much less virulent virus, while infection with drug-resistant HIV bodes poorly for the success of therapy. The effectiveness of treatment can obviously guide the impact that HIV has and treatment efficacy is influenced by both host and viral genetics as well as personal and cultural factors.

But the virus one starts out with may not be the virus that causes trouble down the line. Early in the epidemic scientists recognized that HIV has two faces: one attacks a limited set of immune cells slowly and steadily, wearing down defenses over time. This form of the virus gives AIDS its reputation as a slow but relentless killer. But another form of the virus, one that eventually develops in about half of those with HIV, shifts the disease into high gear as it begins taking out T cells aggressively, causing rapid immune cell loss that can quickly plunge a person into a dangerous state of AIDS.

The differences between the two forms of HIV have been traced to a few simple mutations on one of the virus' outer envelope proteins responsible for latching on to new target cells. Most of the mutations occur in a region called the V3 loop of the gp120 protein. When HIV first attaches to a potential target cell, the gp120 protein hooks up with the cell's CD4 molecule, which is the main cellular receptor for infection. But attachment is not enough; the gp120 also needs to plug into a co-receptor molecule on the cell that helps pull the virus into contact with the cell's surface so the two can merge. In the slow form of the virus, the V3 loop of gp120 is able to connect with a cellular co-receptor called R5 (CCR5). This R5 co-receptor is found on mature T cells that have already been primed to fight infection as well as on immune system sentinels, the macrophages. The R5-using HIV may cause a slower-paced infection because it has only a limited set of target cells to infect and because it seems to replicate at only a moderate rate.

But if the V3 loop changes its chemical properties slightly, the virus starts to be able to use a different co-receptor called X4 (CXCR4). The X4 co-receptor is found primarily on immature T cells that are still being formed in the immune system's incubator, the thymus, and on newly activated T cells that have recently met their antigen. The shift to an X4-using virus speeds up T cell destruction dramatically. One study saw the rate of T cell loss increase by three. A classic sign that X4-using virus is on the loose is when infected T cells clump together to form giant cells called syncytium; X4 HIV used to be called SI type, for syncytium-inducing. But X4-using HIV is also able to kill T cells in subtle and coldly effective ways that are just being discovered.

T-cell depletion is the central problem -- and central mystery -- of HIV disease, and there are several contending theories to explain what's going on. The earliest explanation held that HIV infected and killed T cells directly, end of story. Then it was noted that although most T cells never become infected, they were still being removed. From the beginning, other scientists held that excess immune activation was responsible for running the T-cell supply into the ground, possibly through exhaustion or self-attack. Some now think that apoptosis (programmed cell death) triggered by errant immune signaling, toxic viral byproducts or over-revved regulatory systems, is the main actor. Others argue that the T cells are simply retreating from the blood (where they are typically counted) back into hiding in lymph tissue. Most likely, several of these proposed mechanisms are overlapping and may operate at different stages of infection or under different conditions. One thing that surely heralds a change in the pace of T-cell depletion is the shift from R5 to X4-using virus.

When the Shift Comes Down

The shift from R5-using to X4-using happens in stages and there are a few in-between forms of R5/X4 HIV that are able to use either co-receptor. As virulent as the X4 virus is, one of the mysteries of HIV is why newly infected people almost always carry the R5 virus exclusively. This may be because only the R5 virus is able to infect immune cells that patrol the mucous membrane frontiers where sexually transmitted HIV first takes hold. Yet even in people infected with blood-borne X4 HIV, there is a nearly immediate shift to the R5 variant in the new host. It's not clear why the R5 co-receptor is preferred at first; some have proposed an inhibitory factor, others think that macrophages are a preferred target, or it may be that R5-bearing mature T cells replicate faster and have not yet been exhausted in a healthy host. Although X4 T cells outnumber R5s, they may be outposted to tissues and less available to infection. Typically, the R5-using stage of infection can carry on for five or more years.

The reason for the shift is not clear. Some believe that the virus is trying to escape antibody attack directed at its R5-using site; others think that the virus simply starts to look for different co-receptors once the supply of mature R5-bearing T cells becomes too scarce. Once the shift begins, however, the use of X4 becomes more and more common until in a few people the body's predominant strain of HIV is using X4 exclusively. Clinically, this is bad news, for although antiretrovirals are able to suppress X4 HIV as well as its R5-using ancestor, T-cell destruction now proceeds at an alarming pace. If the R5-using virus is like a sniper picking off selected target cells, the X4 virus is a weapon of mass destruction in the thymic maternity ward.

This may be one of the strongest reasons not to delay starting antiretroviral therapy too long: once the shift to X4 virus begins, it may be very difficult to recover lost immune capacity. One study found the shift often occurring in the 400-500 CD4 cell range. Another showed a greater rate of switching below 500 than above, with those in the 250-500 range switching at a rate similar to those in the 0-250 range. One day it could be common to test for co-receptor usage in the clinic; Virologic has developed a phenotypic test that classifies a virus as X4-using, R5-using, or dual type that could be available by 2004.

What Is X4-Using HIV Doing in There?

A recent paper published in the Journal of Virology may shed some light on why the X4 strain is so destructive. Most of what we know about HIV comes from experiments conducted under laboratory conditions using cell systems and special viral strains that have been adapted to live and reproduce under artificial conditions. There is a limit to what these systems can say about a disease process that affects the complex interactions of immunity in living beings. This is why Andreas Jekle and colleagues from the Gladstone Institute of Virology and Immunology in San Francisco decided to use a model of infection that preserves much of the ecology of the T cell's environment, including a mix of cells at all stages of maturity carrying either the R5 or X4 receptors, or both.

They began with lymphoid tissue harvested from children's tonsil operations then infected the tissue with various strains of X4-, R5- and mixed X4/R5-using HIV. They were particularly interested in looking for evidence of cell destruction caused by apoptosis, a natural mechanism of cell death than can be triggered by a number of internal or external factors. It had been recognized in the mid 1990s that apoptosis was a contributing factor in T-cell depletion. Furthermore, it was noted that not only infected cells but also uninfected "bystander" cells were somehow receiving signals to activate their self-destruct mechanisms.

Jekle and colleagues began by looking for a few characteristic markers that appear whenever apoptosis has been activated. One of the first things they noticed was that signs of apoptosis were far more common among cells that were dosed with the X4- and dual X4/R5-using strains than among cells infected with R5-using HIV. Soon after apoptosis markers began to appear in the X4-infected system they noted that a large number of CD4 T cells were being depleted. Meanwhile, the R5-infected batch of cells only became slightly depleted. Although some studies have shown that CD8 cells were depleted in the presence of X4 HIV, in this study CD8 T cells were not depleted by either type of virus.

They then looked at how many cells had actually become infected with HIV. With the R5-using virus, the number of infected cells was low, and apoptosis levels, as was seen before, were also low. With the X4-using virus, the number of infected cells was similar to that of the R5 virus, but, as seen earlier, the number of cells with apoptosis markers was very high. It seems that the X4-using virus was able to stimulate cell death without directly infecting the cells. This phenomenon is called bystander apoptosis and in a number of other experiments it was shown that X4-associated apoptosis did not depend on establishing a productive infection, and that X4, but not R5, viral strains could induce widespread apoptosis in bystander CD4 T cells. Furthermore, the R5-using virus infected only CD4 T cells that carried R5 and produced a low level of apoptosis in these cells but caused no apoptosis in cells that lacked R5. In contrast, X4-using virus caused extensive apoptosis, predominantly in uninfected bystander cells, including some that also carried R5.

Since the only difference between the X4-using virus and the R5-using virus was a few changes in the V3 loop of the envelope protein, the investigators theorized that the interaction of the viral envelope with the cellular co-receptor was likely responsible for setting off apoptosis. They tested this by adding drugs that block X4-using virus from binding to the X4 receptor on T cells. They found that blocking X4 effectively protected the cells from bystander killing by X4 viruses. Treatment with an R5 blocker did not protect the cells. The authors concluded that binding of the gp120 viral envelope protein of an X4-using virus to the cellular X4 co-receptor was the trigger for bystander apoptosis in their tissue culture system. But is this true in living bodies as well?

In this experiment, a small number of X4 virus particles were able to deplete a large proportion of CD4 T cells -- even when pre-treated by the reverse transcriptase inhibitor AZT. While AZT could prevent cells from becoming infected, it could not prevent apoptosis triggered by exposure to gp120. If this finding is also true in people, then the rapid drop in T-cells seen after the switch to an X4-using virus may be primarily due to the killing of bystander cells that never actually become infected. This may explain why some studies found that viral load does not soar when CD4 counts drop soon after the switch.

Finally, since the immature T cells that are depleted by bystander apoptosis are the precursors to the mature cells, attacking X4-bearing cells may be shutting off the supply of T cells at its source. This could be another reason why only modest rates of T-cell decline are seen during chronic infection with R5 virus, and why T-cell depletion speeds up so much once the shift from R5 to X4 occurs. However, the authors caution, while the gp120 interaction with X4 seems to be necessary to induce apoptosis, there may still be other unknown factors that contribute to this effect. It is also not yet known if the gp120 must be bound to an intact virus for it to trigger apoptosis or if freely floating particles have the same effect.

"Pop" Stoppers

If blocking the binding of gp120 to a cell's X4 co-receptor can stop bystander apoptosis and the resulting rapid CD4 T-cell depletion that occurs for some during the dangerous later stages of HIV disease, then X4 binding inhibitors could be a valuable form of salvage therapy for tens of thousands of people with AIDS. So what do we know about drugs that block X4?

A number of R5 binding inhibitors are being developed as HIV therapy. The R5 receptor is an attractive target because R5-using HIV is the predominant strain in life and has such a long, slow course of infection. Also, because some people are born without R5 receptors (a genetic anomaly that occurs in about 1% of Caucasians) and because rats modified to lack R5 suffer no overt ill effects, it's hoped that blocking R5 won't have toxic consequences. Effective blocking of R5-mediated infection, some believe, could preclude the need for having to ever deal with an X4-using strain. Yet there's been a great deal of concern that blocking R5 binding would push the virus to start using the X4 receptor, although that has not been borne out in laboratory studies. Still, that possibility makes the need for an X4 blocker even more important.

One of the drugs that Jekle and colleagues used in their experiment to show that blocking X4 stopped bystander apoptosis is called AMD3100. The drug had been recognized as an HIV entry inhibitor even before the R5 and X4 co-receptors were discovered in 1996. By 2000, several studies had shown that AMD3100 could not only block infection by X4-using HIV but could also arrest HIV-associated apoptosis. AMD3100 was explored in phase I clinical trials with HIV-infected people where it apparently showed limited activity against HIV. Unfortunately, heart rhythm abnormalities were detected in several patients and development of AMD3100 for HIV therapy was halted in 2001. The drug's sponsor, Anormed, of British Columbia, Canada, is now developing a new, orally available compound called AMD070 that is active in the laboratory against X4-using HIV. Human testing of AMD070 should begin this year.

One of the potential problems with blocking X4 is that, unlike R5, it may perform some essential jobs in the body that shouldn't be messed with. It is also found on a greater variety of cell types than just immune cells. While mice born without R5 do okay, mice with the X4 gene deleted can't survive. And although AMD3100 has been scrapped for treating HIV, it has other kinds of biologic activity and is still under investigation for inhibiting cancers and accelerating recovery after heart attack -- all of which suggests that indiscriminately blocking X4 may have unintended consequences. Finding highly specific medicinal molecules that block HIV co-receptor function without stepping on the toes of any of the receptor's natural tasks is the goal.

New Ideas, New Tools

The other type of X4 blocker used in the Jekle study was a monoclonal antibody that attached to the receptor and blocked access by gp120. One speculative idea is that perhaps a vaccine could induce the body to make its own anti-X4 antibodies, thus using the immune system to supply the therapeutic molecules.

A recent experiment has reported that "gene silencing" through RNA interference was able to suppress expression of the cell's X4 gene and block infection by X4 strains. This new technique gives a powerful tool for understanding the role these receptors play in the immune system and they may one day offer an approach to therapy. Another study observed that the viral protein Tat caused cells to display more CXCR4 receptors on their surfaces, possibly making them more susceptible to X4 virus. If this is significant, then perhaps a Tat inhibitor could be synergistic with an X4 blocker by reducing the number of X4 targets on a cell's surface. Another therapeutic avenue might be to stimulate the molecules that bind to R5 and X4 receptors in nature. These chemical messengers, called chemokines, send signals that also decrease the expression of the receptors on cells.

Resistance Is Always With Us

One issue with co-receptor blockers -- as with every other HIV drug -- concerns the near certainty that viral resistance will develop after a while -- and resistance to AMD3100 has already been shown to develop in laboratory experiments. When the drug was used against an exclusively X4-using virus, gp120 accumulated mutations that allowed it to use X4 in spite of the drug. But there are also suggestions that the mutations that allow escape from the drug also make the virus less fit and less pathogenic. In one experiment, all of the X4 isolates that evolved resistance to AMD3100 after serial passage in cell culture exhibited reduced fitness compared to wild type. In a clinical trial of AMD3100 in patients with dual X4/R5 HIV the virus simply switched over to using R5. Since all previous strains of HIV are likely to persist in viral reservoirs, blocking an evolved X4-using virus would probably tend to cause an earlier R5-using strain to eventually re-emerge. While not a perfect solution, having an active R5 strain may be better than the alternative.

Receptor blocking is still in its infancy, although several R5 blockers are moving forward in clinical trials. While resistance may be a problem, it may also provide an opportunity. One therapeutic strategy that is likely to come into greater use may be called "guided resistance," which seeks to back HIV into a corner of diminished fitness and destructive potential. Indeed this is already the only strategy left for many people with multi-drug resistant virus who find that a failing regimen, if tolerable, is far better than no regimen at all. Since the virus appears impossible to eradicate, maybe shutting off one of its more destructive aspects, such as the shift to X4 type, can help to keep expanding the possibilities for living with HIV.

Until then, entry inhibitors will continue to have an important role to play in helping scientists understand the basic science of HIV pathogenesis and T-cell depletion. It may be a race to see which avenue first benefits the greatest number of people: another new drug to suppress HIV or a new understanding that unlocks the secret to something much better.

Jekle A, et al., In Vivo Evolution of Human Immunodeficiency Virus Type 1 toward Increased Pathogenicity through CXCR4-Mediated Killing of Uninfected CD4 T Cells. J Virol_, May 2003, p. 5846._