3.2. The Design and Medicinal Chemistry of Ligands based on Molecular Recognition

pp. 28-29 in Bioethics and the Impact of Human Genome Research in the 21st Century

Author: Yoshiaki Kiso (Kyoto Pharmaceutical University)

Editors: Norio Fujiki, Masakatu Sudo, and Darryl R. J. Macer
Eubios Ethics Institute

Copyright 2001, Eubios Ethics Institute All commercial rights reserved. This publication may be reproduced for limited educational or academic use, however please enquire with the author.
The analysis of gene structures has progressed, and state of the art modeling and docking techniques that are based on structural scienc e have become essential to medicinal chemistry. By way of introduction I will review the HIV protease. The structure of the HIV virion consists of glycoproteins, structural proteins, and the main enzymes (protease, reverse transcriptase, integrase). The cell membrane lipid bilayer is that of a human cell membrane. The HIV protease consists of 99 amino acids, with the active site containing the Asp-Thr-Gly sequence characteristic of aspartic proteases. The enzyme functions as a homo-dimeric form. This is the active site.

This protease cleaves gag and gag-pol polyproteins to produce structural proteins and key enzymes, the protease itself, reverse transcriptase and integrase. The HIV protease can recognize Phe-Pro and Tyr-Pro sequences as the retrovirus specific cleavage site, but mammalian proteases do not exhibit this particular specificity.

In order to obtain selective inhibitors we focused attention on the retro-virus specific Phe-Pro scissile site, and introduced the hydroxymethylcarbonyl isostere as a transition-state mimic. The stereochemistry of the OH group is important for inhibitory activity. Against expectations, in the case of HIV protease inhibitors the syn-form was more potent. We have named this unnatural syn-form amino acid allophenylnorstatine (Apns).

We designed some substrate-based inhibitors. This heptapeptide, KNI-93, was a potent inhibitor. We next found a small-sized, tripeptide HIV protease inhibitor, KNI-102. After lead-optimization, we discovered a highly selective and superpotent HIV protease inhibitor KNI-272. This hydroxymethylcarbonyl isostere interacts with the enzyme active site aspartic acids in the same mode as does the transition state.

The stereochemistry of the hydroxyl group is important. The anti-form, KNI-529, was one thousand times less potent than KNI-272. The hydroxyl group of KNI-272 interacts well with both aspartic acids in the active site. The interaction of the hydroxyl group of KNI-529 with these active-site amino acids is weak. The solution, crystalline and complexed structures are similar except for the P3 residue. This led us to think that a small-sized dipeptide, lacking the P3 part, would be advantageous, so we started looking at these. The P2-P2' moiety may represent a core structure for achieving enzyme inhibition. P was removed, and P2 was replaced by several Val-mimicking alkyl and hydrophilic carboxyl groups.

Of these compounds, one called KNI-413, containing a dimethyl group, exhibited HIV protease inhibitory activities, but only weak antiviral activities. In order to enhance this activity we introduced a dimethylphenoxyacetyl moiety, which gave the potent inhibitor KNI-727. Furthermore, we cyclized the P2 moiety of KNI-549 to make a rigid structure, and thus succeeded in obtaining the potent inhibitor KNI-577. This hydroxymethylbenzoyl-Apns-Dmt-NHBut showed remarkably high anti-HIV activity, having an EC50=0.02mM. This anti-HIV activity was higher than that of tripeptide KNI-272, while the compound's cytotoxicity was very low, and its bioavailability was 48%.

We also pursued design based on symmetry considerations, and introduced a 2-methyl benzylamine in P2'. This compound, KNI-764, exhibited highly potent enzyme inhibition, and its anti-HIV activity was surprisingly good, while no cytotoxicity was observed, and its bioavailability was 41%, and plasma half-life 92 minutes. It had good pharmacokinetics.

In peptide studies cell membrane permeability is very important. Recent papers describe the peptide transporters. PepT1 transports di- and tri-peptides. PepT1 is a 707 amino acid protein with twelve trans membrane domain channels. PepT1 recognizes the amino-terminus by ionic interaction, this peptide bond by hydrogen bonding, the other peptide bond through the imidazole of histidine, and this side chain by hydrophobic interaction.

A membrane nucleoside transporter is also known. Also, a nucleoside receptor is known. Thus nucleosides have affinity for the cell membrane. Taking the above into consideration we designed a novel hybrid-type anti-HIV agent. As mentioned already, the carboxylic acid-containing HIV protease inhibitor exhibited very low anti-viral activity. We therefore esterified these by coupling them with a nucleoside RT inhibitor, azidothymidine, using DCC/DMAP. An AZT transporter protein has also been reported.

KNI-413 was esterified by coupling with the nucleoside RT inhibitor, azidothymidine, using DCC/DMAP. We obtained a potent hybrid-type anti-HIV agent, KNI-684. The antiviral activity of KNI-684 was increased 6.6 times over that of AZT. As already mentioned, a 2-methylbenzylamide P2' moiety was better than a t-butylamide , so we prepared conjugates of AZT with an HIV protease inhibitor containing 2-methylbenzamide at P2'. The resulting KNI-694 exhibited highly potent antiviral activity. This antiviral activity of KNI-694 was 46 times more potent than that of AZT.

These data suggest a mechanism for the prodrug. The nucleoside-containing protease inhibitor can penetrate the cell membrane, and then the prodrug is cleaved to produce two different types of inhibitor.

We also synthesized another hybrid-type anti-HIV agent. The HIV protease inhibitor KNI-727 was conjugated with AZT through a linker. At first we used a succinic acid linker, but this ester was too stable for KNI-727 to be regenerated. Therefore we used succinyl-glycine and glutaryl-glycine linkers in order to release KNI-727 by spontaneous imide formation. This imide formation was faster for the former than the latter, as expected for the energetically favorable five-membered ring formation. In practice, this spontaneous imide formation was too fast for prodrug use.

This hybrid-type compound KNI-1039, containing the glutaryl-glycyl linker exhibited highly potent anti-HIV activity, 62 times more potent than AZT. The conjugate using the succinylglycine linker was not so effective. These results suggest that KNI-1039 is relatively stable outside the cell, but undergoes degradation gradually on entering the cell.

These examples give some idea of the design of new drugs.

Please send comments to Email < asianbioethics@yahoo.co.nz >.

To contents page
To Eubios book list
To Eubios Ethics Institute home page