Katy, bar the door! HIV entry inhibitors.

HIV replication requires the virus to enter human cells and become integrated into cellular DNA. Similarly, currently approved drugs must also enter those cells to provide a therapeutic effect. Perhaps because this battle between virus and drugs rages at a cellular level, patients and doctors have long anticipated the emergence of' long-term side effects. Side effects such as peripheral neuropathy, pancreatitis and anemia, etc. have been present all along, but these seemed isolated and drug-specific. Now as patients live longer, the broader and more fundamental toxicities are emerging with disturbing frequency.

Finding a way to keep HIV from entering and infecting human cells is an attractive strategy. Its conceptual simplicity is beguiling. If HIV cannot enter the cell, it cannot replicate--end of story. The idea of barring HIV from entry has many appeals, not the least of which is avoiding inadvertent drug/cellular interactions. These interactions are not thoroughly understood and man), scientists suspect they are responsible for the increased incidence of metabolic abnormalities and side effects such as lactic acidosis. In any case, putting more chemicals into cells hardly seems the best strategy to follow at this point. Desirable as barring HIV from cells may seem, the truth of the matter is that until recently no one found an effective way to do this.

Early attempts at entry inhibition. Knowledge of how HIV enters cells has grown tremendously in the last few years, but in the late 1980s the concept of viral entry was simplistic. Investigators knew that HIV attaches itself to a molecule called CD4 on certain specialized immune cells. The presence of this type of receptor on the surface of these cells identifies the cells and thus the cells themselves are known as CD4 T cells or more specifically, CD4+ T lymphocytes. Thus, all researchers knew was that the first step in the infection of a cell is the attachment of HIV to the CD4 molecule. Past that initial attachment there was little knowledge of how the virus enters the cell.

Armed with this limited knowledge, groups of scientists came up with the idea of flooding the body with "decoy" CD4 molecules to bind up the virus and keep it from attaching to the CD4 on the target immune cells. This seemingly ingenious method held promise. Around 1988 in vitro data suggested that recombinant (man-made) soluble CD4 could compete well enough with the natural CD4 of target cells to prevent infection. Further, soluble CD4 was synergistic in vitro with zidovudine (Retrovir), the only approved therapy at that time.

Despite the success of the test tube experiments, soluble CD4 was a complete failure when given to patients. It had no antiviral effect whatsoever. Subsequent experience confirmed a phenomenon that has since been a problem for several therapeutic approaches: laboratory-adapted strains of HIV do not always respond as naturally occurring strains in a patient's body. Also, soluble CD4 was metabolized rapidly in the body and effective therapeutic quantities could not be sustained even when soluble CD4 was infused intravenously.

Despite the disappointment of soluble CD4, other attempts at interfering with entry were tried (e.g. dextran sulphate) but none were successful. Some of these efforts proved to be detrimental in the end. Since these early days of disappointment, much has been learned about how HIV gains entry into cells.

The many steps of viral entry. As is often the case in the science of HIV, the order of discoveries makes for a confusing story. Identifying the steps of viral entry is no exception. Between 1995 and 1997 scientists isolated certain chemical substances known as beta chemokines that appeared to have an inhibitory effect on viral replication.

These chemokines, called RANTES, MIP-1alpha and MIP-1beta are involved in the body's inflammatory process, while another protein called stromal-derived factor (SDF-1) is involved in signaling lymphocytes to move to inflamed tissue. At the time, it was unclear why these chemokines have inhibitory effects on HIV replication. Only later was it determined that these chemokines bind naturally to 2 molecules on the CD4 T cell. The first set of chemokines (RANTES, MIP-1alpha and MIP-1beta) were found to bind to a molecule on the CD4 T cell known as CCR5. SDF-1 was found to bind naturally to another molecule on the CD4 T cell known as CXGR4. This was the first hint that these chemical substances might be interfering with viral entry.

In short, when these chemokines are present, HIV replication is inhibited. Only later did scientists realize that the reason these chemokines have an inhibitory effect on viral replication is that HIV not only has to engage the CD4 molecule to enter the cell, but also it has to dock with one of the 2 other receptors, namely CCR5 or CXCR4, to enter the target cell. If these chemokines are present in sufficient quantities, they engage all the CCR5 and CXCR4 receptors and HIV cannot complete its entry.

Attachment to the CD4 molecule then is merely a preliminary step in the whole entry procedure. A subsequent attachment either to CCR5 or CXCR4 is required as a second step in the process of entry. Further, scientists determined that CCR5 and CXCR4 bind different strains of HIV. CCR5 was noted to bind "M-tropic"or "nonsyncytium inducing" virus, most associated with initial infection and early disease; whereas CXCR4 was found to bind "T-tropic" or "syncytium inducing" viral strains associated with accelerated and advanced disease. Having identified 2 steps of viral entry--binding to CD4 and binding to either CCR5 or CXCR4--laboratories throughout the world began to work on methods of interfering with coreceptor attachment.

Help from Mother Nature. Uncertainties still underlie the blocking or manipulation of coreceptors. Alter all, these coreceptors have natural functions besides mediating HIV entry into the cell. Does interfering with these receptors block some essential function of cells? Coincidently, one reassuring observation emerged that encouraged scientists to pursue blocking CCR5. In the course of studying individuals who had multiple exposures to HIV but never became productively infected, researchers noted that these subjects had natural mutations that blocked the full expression of CCR5 receptors. Individuals with one mutation in their CCR5 receptors had slower disease progression than individuals without this mutation. Individuals with a double mutation in their CCR5 receptors appeared to have a great degree of protection from HIV infection. (The exact degree of this protection is still undetermined since it may be possible that HIV employs other receptors to complete entry under some circumstances.) Additionally, these coreceptor mutations appeared to have no ill effects for these individuals. Two favorable observations emerged from the study of these individuals: first, blocking or downregulating CCR5 may have no deleterious effect on cellular function and second, CCR5 may indeed play a major role in the viral infection of cells.

The issue of blocking CXCR4 is more complex. Scientists have genetically engineered mice not to express the CXCR4 coreceptor. These mice did not survive long enough to be born and the embryos showed serious abnormalities to their immune systems, hearts and central nervous systems. Researchers in Japan have shown that eliminating the gene for SDF-1, the chemokine that binds to CXCR4, also causes severe developmental abnormalities in mice. Others who have studied CXCR4 believe that the coreceptor may be essential during fetal development but unnecessary after birth (Science, 280, p. 825, 1998). Thus, research into finding ways to block CXCR4 continues.

More questions. While it is clear that blocking CCR5 might play a role in inhibiting infection of new cells, might it also result in a shift towards the use of CXCR4--a receptor associated with more advanced disease and more virulent strains of HIV? And could blocking both receptors lead to a further evolution in HIV that might allow it to use still other coreceptors? At the recent 7th Conference on Retroviruses and Opportunistic Infections, investigators presented preliminary data that indicate blocking CCR5 does not promote a shift towards the use of CXCR4 by HIV (7th CROI, Abstracts S17 & S18, oral presentations, San Francisco, 2000) and that a shift from CXCR4 to CCR5 is more likely. The possibility that HIV may evolve to use other receptors besides CCR5 and CXCR4 still exists, although scientists believe this evolution is unlikely (Science, 280, p. 825, 1998).

While the effects of interfering with CCR5 or CXCR4 are still not certain, several groups have started investigating compounds that either inhibit HIV binding with one of the coreceptors through competition or manipulate the expression of chemokine receptors. Table 1 lists agents under investigation.

TABLE 1

Agent                           Mechanism of Action

T-20 (pentafuside)              Blocks fusion by attaching
Requires subcutaneous           to gp41 during the "hair-
injection,                      pin" retraction. (See
                                RITA!, 5:2, p. 11, 1999).

T-1249                          Similar to T-20. Works
Requires subcutaneous           against T-20 resistant
injection,                      virus.

C-34                            Binds to the "pre-hairpin"
Low molecular weight,           pocket formed transiently
indicating it may be            by gp41 during fusion.
a candidate for the
development of orally
available drug.

AMD3100                         Antagonist of CXCR4, but
Low molecular weight,           not CCR5. In vitro syner-
currently administered          gy with T-20.
intravenously.

Met-SDF-1beta &                 Antagonists of chemokine
AOP-RANTES                      receptors CCR5 and
                                CXCR4.

SCH-C                           Antagonist of CCR5.
Oral absorption in
animal model.

PRO 542                         Competitive binding
Molecule designed               through attachment to
to treat and                    gp120; may strip viral
prevent infection,              particle of gp 120.

PRO 140                         Anti-CCR5 monoclonal
Inhibits CCR5 without           antibody,
affecting chemokine
activity.

PRO 367                         Binds to gp120 and
Radio-immunoconjugate           releases a lethal dose of
designed to recognize and       radiation.
destroy HIV-infected
cells.

TAK-779                         Molecular inhibitor of
                                CCR5 but not CXCR4.

Agent                           Status

T-20 (pentafuside)              Beginning Phase III
Requires subcutaneous
injection.

T-1249                          Phase I/II
Requires subcutaneous
injection.

C-34                            Preclinical
Low molecular weight,
indicating it may be
a candidate for the
development of orally
available drug.

AMD3100                         Phase II
Low molecular weight,
currently administered
intravenously.

Met-SDF-1beta &                 Preclinical
AOP-RANTES

SCH-C                           Preclinical
Oral absorption in
animal model.

PRO 542                         Phase II
Molecule designed
to treat and
prevent infection,

PRO 140                         Phase I is planned to start
Inhibits CCR5 without           by the end of 2000.
affecting chemokine
activity.

PRO 367                         Phase I/II are planned.
Radio-immunoconjugate
designed to recognize and
destroy HIV-infected
cells.

TAK-779                         Preclinical

Agent                           Investigator

T-20 (pentafuside)              Trimeris in conjunction
Requires subcutaneous           with Hoffmann-LaRoche
injection.

T-1249                          Trimeris in conjunction
Requires subcutaneous           with Hoffmann-LaRoche
injection.

C-34                            P. Kim, Howard Hughes
Low molecular weight,           Medical Institute &
indicating it may be            Massachusetts Institute of
a candidate for the             Technology
development of orally
available drug.

AMD3100                         AnorMED
Low molecular weight,
currently administered
intravenously.

Met-SDF-1beta &                 Harvard Medical School
AOP-RANTES                      & Universita' di Milano

SCH-C                           Schering-Plough
Oral absorption in
animal model.

PRO 542                         Progenics
Molecule designed               Pharmaceuticals
to treat and
prevent infection,

PRO 140                         Progenics
Inhibits CCR5 without           Pharmaceuticals
affecting chemokine
activity.

PRO 367                         Progenics
Radio-immunoconjugate           Pharmaceuticals
designed to recognize and
destroy HIV-infected
cells.

TAK-779                         Albert Einstein College of
                                Medicine, Aaron
                                Diamond Research
                                Center, The Rockefeller
                                University, Howard
                                Hughes Medical Institute,
                                and State University of
                                New York

A different approach. While some investigators have concentrated their efforts on the manipulation of coreceptors, others have concentrated on jamming the actual mechanism by which HIV penetrates the cell membrane. A previous issue of this journal discusses the mechanism of membrane fusion (see RITA!, 5:2, p. 11, 1999).

Agents like pentafuside (T-20) and its likely successor, T-1249, act by interfering with the so-called "hairpin" mechanism of one of HIV's proteins, which unfolds in a hinged fashion, extends itself to harpoon the cell membrane and then coils back onto itself to bring the virus close to the cellular membrane. The action of inhibitors like T-20 is akin to throwing in a monkey wrench at the precise moment that the mechanism is unfolded, preventing it from coiling back onto itself.

The mechanism of T-20 is well established in clinical trials and this agent is likely to be the first entry inhibitor to gain US Food & Drug Administration (FDA) approval. Phase III trials are imminent. As promising as T-20 appears to be, it is not without some problems. First, the agent is difficult to manufacture. Second, it is a rather large peptide that with the current formulation requires subcutaneous injections twice a day. The size of the peptide (36 amino acids) prevents it from being formulated for oral dosing. Finally, resistance to the compound has been well documented and, like all other current antiretroviral agents, patients will have to take it with other active agents to avoid resistance. On the positive side, T-20 might work with all strains of HIV since all strains use the same mechanism of action for membrane fusion. Further, preliminary data indicate that T-20 may be synergistic with other entry inhibitors under development (see Table 1 on p. 10).

The question of the availability of T-20 under an expanded access program prior to FDA approval has been at the top of the agenda of many activist organizations, including The Center for AIDS. At a meeting with the pharmaceutical developers of T-20, Trimeris and Roche Pharmaceuticals, company officials indicated that an expanded access program was under consideration. However, manufacturing and supply difficulties might constrain the timing and the size of the program. A reformulation of the agent is currently being tested that will cut down on the number of injections required. (Early trial participants were required to administer up to 4 subcutaneous :injections per day to maintain sufficient quantity of the agent in the body.) Also, the developers report that they have made advances that may affect manufacturing favorably.

Despite the advances in manufacturing and formulation, as of this writing company officials could not commit to any specific details of an expanded access program. This represents a tragic situation for persons with AIDS who are unable to construct a viable and effective drug regimen without T-20. Trimeris and Roche have proposed a Phase III trial to the FDA that would include patients who are highly experienced with current medications. The design calls for "best available" background therapy as determined by genotypic and phenotypic tests, current clinical status and drug history. To enter the trial participants will be required to have viral loads greater than or equal to 5000 copies/mL without any upper limit. There will be no restrictions on the number of CD4 T cells of participants under this proposed protocol. While this Phase III trial will cover many who might be in need of third and fourth line regimens, it is no substitute for expanded access. Even if the FDA approves this protocol as proposed, persons in parts of the country where there are no clinical trials of T-20 will be deprived of a valuable agent that in some cases may be life-saving. Company officials expect that negotiations with the FDA will occur in June and that details of the Phase III trial and expanded access program may be forthcoming in July 2000.

Beyond membrane fusion. Other investigators working on entry inhibitors have made new and exciting observations that may provide a glimpse into the future of entry inhibitors. Several groups are working on smaller molecules that will enable entry inhibitors to be orally bioavailable. At least one company is working on a compound that will not only attach to the virus but will infuse a lethal dose of radiation into infected cells (see Table 1).

An investigator who garnered much attention at the recent 7th CROI is Peter S. Kim, PhD, of the Whitehead Institute for Biomedical Research at the Massachusetts Institute of Technology and the Howard Hughes Medical Institute. Kim and his group were the first to identify what appears to be a vulnerable deep pocket in one of HIV's critical outer proteins, gp41.

This pocket is only transiently displayed during the process of viral fusion, but it appears that binding it effectively blocks viral entry. Also, the pocket can be blocked by a compound of low molecular weight, making it possible to create an orally available drug. This pocket is highly conserved, that is to say, it does not vary from virus to virus or strain to strain. Kim and his colleagues have used X-ray crystallography to decipher the structural architecture of gp41 and of this key pocket. The process of identifying molecules for this target has begun and may be some of the most promising research on the long-term horizon.

On the whole, entry inhibitors appear to be the next new weapon against HIV. Unfortunately, much of the research is young and with the exception of T-20, the approval of actual drugs is still some time away.

THE PROCESS OF VIRAL ENTRY

1. HIV approaches a human immune cell

2. HIV attaches to CD4 receptor

3. HIV attaches to coreceptor, either CCR5 or CXCR4

4. Membrane fusion leading to entry