Hiv Protease Inhibitors

When in 1984 it was discovered that the human immunovirus (HIV) caused AIDS it was the start of an intensive drug-hunting process. The DNA in the virus encodes for a number of enzymes, e.g., a reverse transcriptase, an integrase, and a protease. Each of these represented a potential drug target, and drugs have been developed subsequently for each of the enzymes. Presently, cocktails of inhibitors against at least two of these enzymes are used therapeutically. In the following text, we will concentrate on the HIV-1 protease and how biostructure-based design has been applied extensively to this target.

The first 3D structure of an HIV-1 protease in complex with an inhibitor, MVT-101, was reported shortly after it had been shown that inhibition of the HIV-1 protease prevented the virus in producing new virions. MVT-101 binds to the enzyme in an extended conformation and forms a network of hydrogen bonds between the ligand and enzyme (Figure 2.7). Hydrophobic substituents on the inhibitor occupy hydrophobic pockets in the enzyme, and a water molecule mediates contact between the inhibitor and two residues in two flexible beta-sheets, normally referred to as the flaps. Today, close to 300 structures of complexes between HIV-1 protease and ligands have been determined, which makes HIV-1 protease one of the structurally most extensively studied proteins.

The first HIV-1 protease inhibitors like, e.g., indinavir (Crixivan), nelfinavir (Viracept), and saquinavir (Invirase, Fortovase) (Figure 2.8) were derived from the polypeptide sequences cleaved by the protease in the HIV. Accordingly, they were very peptide-like and had poor bioavailability.

Unfortunately, the HIV rapidly developed resistance against these first-generation inhibitors. The shapes of the hydrophobic pockets are sensitive to mutations in the enzyme and accordingly, the virus could easily prevent an inhibitor from binding by mutation of residues forming the hydro-phobic pockets.

H2N^NH

N ynvVtnh2

H2N O

FIGURE 2.7 3D structure of the HIV-1 protease with bound MVT-101 (pdb-code 4HVP). (A) Side view showing that the active form of the HIV-1 protease is a homodimer (colored red and green, respectively) and that the inhibitor MVT-101 binds between the two monomers. (B) Top view showing the extended form of the inhibitor and that the inhibitor via a structural water molecule (cyan) binds to the flaps. (C) The structural water molecule makes four hydrogen bonds to the inhibitor and to Ile50 in the flaps. (D) Chemical structure of MVT-101.

o-Shh O

H2N^NH

N ynvVtnh2

H2N O

FIGURE 2.7 3D structure of the HIV-1 protease with bound MVT-101 (pdb-code 4HVP). (A) Side view showing that the active form of the HIV-1 protease is a homodimer (colored red and green, respectively) and that the inhibitor MVT-101 binds between the two monomers. (B) Top view showing the extended form of the inhibitor and that the inhibitor via a structural water molecule (cyan) binds to the flaps. (C) The structural water molecule makes four hydrogen bonds to the inhibitor and to Ile50 in the flaps. (D) Chemical structure of MVT-101.

Indinavir Amprenavir
Nelfinavir KNI-764

Saquinavir

Tipranavir

FIGURE 2.8 Examples on HIV-1 protease inhibitors. Indinavir, nelfinavir, and saquinavir are examples of first generation, amprenavir and KNI-764 of second generation, and tipranavir of third-generation HIV-1 protease inhibitors.

Saquinavir

Tipranavir

FIGURE 2.8 Examples on HIV-1 protease inhibitors. Indinavir, nelfinavir, and saquinavir are examples of first generation, amprenavir and KNI-764 of second generation, and tipranavir of third-generation HIV-1 protease inhibitors.

In order to circumvent this problem, inhibitors binding not only primarily due to hydrophobic effects but also by hydrogen bonding to the enzyme backbone atoms (amide NH group and amide carbonyl oxygen atom) were designed. These hydrogen-bonding contacts are not sensitive to mutations, since they do not involve the side chains but only the backbone atoms. These second-generation inhibitors like, e.g., amprenavir (Agenerase) and KNI-764 (Figure 2.8) showed significantly different binding characteristics.

Thermodynamic determination of the enthalpy and entropy components to the free energy of binding (AG = AH - TAS) can be determined by isothermal titration calorimetry (ITC), and this method has led to a much more detailed understanding of the energetics associated with the process of binding a ligand to a protein. By using ITC to guide the design of new inhibitors it has been possible to optimize the binding characteristics of these new inhibitors.

For the first-generation HIV-1 protease inhibitors, the majority of the free energy of binding is due to an entropy gain associated with filling hydrophobic pockets in the HIV-1 protease with hydrophobic substituents on the inhibitor. The second-generation inhibitors were characterized by both enthalpy and entropy now contributing to the free energy of binding, making the enzyme less likely to develop resistance.

Recently, third-generation HIV-1 protease inhibitors have been developed based on the careful optimization of the structural and energetic contributions to binding. Tipranavir (Aptivus) (Figure 2.8) is an example of an HIV-1 protease inhibitor with unique binding characteristics. It is a highly potent inhibitor (Ki = 19 pM), which primarily binds to the wild-type HIV-1 protease due to entropy effects. The unusually high binding entropy is most likely caused by release of buried water molecules from the active site of the HIV-1 protease. When binding to the multidrug-resistant HIV mutants tipranavir only looses little in potency, because the reduction in binding entropy is compensated by a gain in binding enthalpy.

In 1994, researchers at the DuPont Merck Pharmaceutical Company reported an important observation. They had realized that in most of the complexes between HIV-1 protease and the peptide-like inhibitors a structural water molecule bridged the ligand and enzyme. They concluded that by

"Structural water" MeO

"Structural water" MeO

HO OMe H-bond donor/acceptor

HO OMe H-bond donor/acceptor

"Structural water" O

"Structural water" O

HO OH H-bond donor/acceptor

HO OH

FIGURE 2.9 The DuPont Merck design process leading to the seven-membered urea HIV-1 protease inhibitors. (A) Symmetric dihydroxyethylene inhibitor used to define a pharmacophore. (B) Initial hit from 3D database search. (C) Initial synthetic scaffold. (D) Scaffold modified to accommodate two hydrogen-bond donors/ acceptors. (E) Final scaffold optimized for synthetic feasibility and improved hydrogen bonding.

designing an inhibitor that would displace this water molecule, a favorable entropy contribution to the binding energy would be obtained. In addition, selectivity should be gained since this structural water molecule was unique for the viral proteases.

The DuPont Merck scientists defined a pharmacophore (see Chapter 3) from known dihydroxyethylene inhibitors (Figure 2.9A), used this pharmacophore for searching a database and obtained a hit (Figure 2.9B), which subsequently gave the idea to the six-membered cyclic ketone as lead structure (Figure 2.9C). Ring expansion to the seven-membered cyclic ketone (Figure 2.9D) enabled incorporation of two hydrogen-bonding donor/acceptor groups and synthetic reasons led to the

FIGURE 2.10 Schematic representations of the binding of the seven-membered urea HIV-1 protease inhibitor XK-263 to HIV-1 protease (pdb-code 1HVR). In (A), Bn and Np refer to benzyl and 2-naphthyl, respectively. In (B), the two Asp25 and Ile51 residues making hydrogen bonds to the inhibitor are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively. The surface is colored accordingly.

FIGURE 2.10 Schematic representations of the binding of the seven-membered urea HIV-1 protease inhibitor XK-263 to HIV-1 protease (pdb-code 1HVR). In (A), Bn and Np refer to benzyl and 2-naphthyl, respectively. In (B), the two Asp25 and Ile51 residues making hydrogen bonds to the inhibitor are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively. The surface is colored accordingly.

series of seven-membered cyclic urea HIV-1 protease inhibitors (Figure 2.9E). Using the structural information available from the many peptide-like inhibitors, the nature and stereochemistry of the substituents on the cyclic urea could be designed so they were preorganized for binding to the enzyme (Figure 2.10). By preorganization of a ligand for binding, the conformational energy penalty, often associated with ligand binding, is reduced and the ligand may be more potent.

Although the DuPont Merck cyclic urea inhibitors were based on a brilliant structural idea and potent inhibitors were designed, the inhibitors generally had low bioavailability and the HIV quickly developed resistance against the compounds. Thus, none of these inhibitors are among the drugs used today. Other companies have subsequently adopted the same idea in their design processes. One example is the previously mentioned HIV-1 protease inhibitor tipranavir (Aptivus) where the oxygen atom in the carbonyl group replaces the structural water molecule.

Was this article helpful?

0 0

Post a comment