Although drugs such as protease and reverse-transcriptase inhibitors have had a huge effect on treating AIDS in the developed world, they are not without their problems: they do not eradicate HIV-1 from infected people1; long-term side effects have been observed2; and drug-resistant variants of HIV-1 are emerging3. There is a pressing need to identify new classes of anti-HIV-1 drugs, and, reporting in Cell, Eckert et al.4 signal a way forward. They have discovered that a cavity formed in the HIV-1 gp41 transmembrane glycoprotein can accommodate small, circular d-peptides — peptides that could prevent HIV-1 from fusing with, and entering, human cells.

Central to the immunodeficiencies of AIDS is a block to production of T cells bearing the cell-surface protein called CD4. When drugs reduce viral replication, the immune system gains time to repair this block5. The gp41 trimer mediates the initial fusion of HIV-1, and its potential as a drug target was revealed when large peptides derived from gp41 were found6,7 to inhibit HIV-1 infection in vitro. These peptides insert a metaphorical spanner into the works — they inhibit conformational changes within gp41, substituting for one or more components of the trimer and preventing the necessary intermolecular interactions that drive membrane fusion8,9.

One version of these peptides, known as T-20, has been shown in clinical trials to reduce viral load, an important proof of concept for HIV-1 fusion inhibitors10. But large peptides make imperfect drugs because they cannot be administered orally and they are rapidly metabolized. Indeed, most successful drugs have a relative molecular mass of less than 1,000. So do any inhibitors of gp41-mediated fusion not have the drawbacks of conventional peptides? Eckert and colleagues' results4 indicate that some, to a certain extent at least, do.

The same group previously described11,12 a cavity in gp41 that seemed to be of a suitable size and location to act as the binding site for a fusion inhibitor. This cavity is not identical to the binding site for T-20 and related peptides, but it is a part of it. Eckert et al. have now engineered a simplified version of the cavity, designated IQN17, by fusing a soluble, trimeric helical peptide (GCN4–pIQI) to 17 residues of the relevant gp41 region. This, in effect, produces a stable gp41 intermediate4. The authors then used IQN17 as a bait in a sophisticated screening procedure, based on phage-display library technology, to fish out the d-peptides that bind from a random pool. They showed that these peptides bind within the pocket of authentic gp41, successfully inhibiting membrane fusion (Fig. 1).

Figure 1: Plugging the gp41 pocket.
figure 1

Ribbon representation of a circular D-peptide inhibitor (purple) bound to an α-helical stretch of gp41 (green). Of the 16 residues in the peptide, only six interact directly with the gp41 pocket.

Full size image

Why did the authors choose d-peptides? These peptides contain d-amino acids, whereas all natural peptides contain the mirror-image l-form. But, because d-amino acids are unnatural, the peptide bonds between them resist digestion by cellular enzymes, increasing the stability of the peptide in vivo. Indeed, Eckert et al. note that cyclosporine A, a widely used drug that can be given orally to counter transplant rejection, contains d-amino acids and is not much smaller than the best of their inhibitory d-peptides.

The current d-peptide inhibitors are not very potent, with IC50 values (the concentration of inhibitor that is needed to block gp41-mediated fusion by half) in the 10–100-μM range. However, the skilled hands of medicinal chemists can work wonders with starting compounds. Furthermore, Eckert and colleagues outline a plausible strategy in which IQN17 and the currently available d-peptides could be used in high-throughput screening assays to identify small-molecule inhibitors from conventional chemical libraries. There are many steps along the path from the bench to the pharmacy, but the first is often the most difficult. Together with inhibitors aimed at other components of the HIV-1 fusion process13, derivatives of the newly identified d-peptides may eventually be valuable weapons with which to fight HIV-1 infection.