Abstract
Ph.D.
Despite nearly three decades of intensive research, the HIV/AIDS pandemic remains a major challenge to
modern medicine. The discovery and development of antiretroviral agents acting against various essential
viral processes and enzymatic targets have greatly enhanced the quality of life for infected individuals, but
no cure or preventative vaccine is available as yet and HIV infection is currently considered irreversible.
Furthermore, the emergence of viral resistance to every class and type of antiretroviral treatment agent
necessitates the continued discovery of antiretroviral agents with novel mechanisms of action. The first
antiretroviral agent targeting the retroviral integrase enzyme (InsentressTM, Raltegravir) received regulatory
approval from the United States Food and Drug Administration during 2007, validating HIV-1 integrase as a
therapeutic target and providing a much-needed second- or third-line treatment option for treatment
experienced patients. This enzyme was selected as a target for the current work.
As limited data were available on the primary and secondary structure of the biologically relevant HIV-1
integrase enzyme, a first step in the present work was the construction of monomeric, dimeric and
tetrameric models of the enzyme with biologically relevant catalytic centres incorporating both viral and
host co-factors and DNA. The models were constructed to identify potential inhibitors of the strandtransfer
reaction of HIV-1 integrase and were based on observations and interactions reported in the
literature and on crystal structure data of HIV-1 integrase sub-domains and related structures available in
the Protein Data Bank. The monomeric model was used as the macromolecular target in docking studies
with “drug-like” compound databases, identifying the pyrrolidinone compound class as an in silico hit
candidate for further development.
Initial activity screening of a number of commercially available pyrrolidinone analogues against
recombinant HIV-1 subtype B integrase in direct enzyme assays confirmed the predicted potential for
strand transfer inhibition of the compound class, and provided initial support in the further development of
this compound class as inhibitors of HIV-1 integrase that target the strand-transfer step. Retrosynthetic
analysis of the pyrrolidinone hit candidates provided a facile one-pot, three-component synthetic pathway
from readily available starting materials, which generally gave the proposed products cleanly and in
acceptable yields. A range of closely related analogues were designed and synthesised. The analogues
making up this series generally differed by only one functional group, in order to enable initial structureactivity
relationship investigations during later stages of the project.
Foreword Page XVI
The synthesised pyrrolidinone analogues were screened through a range of direct and cell-based in vitro
assays to determine the toxicity and strand-transfer activity of each. In general, the pyrrolidinone
compounds proved well-tolerated in PM1 cell culture, with clear potential to further develop the strandtransfer
inhibition of the compound family in second- and further-generation optimisation stages.
Furthermore, the aqueous solubility and membrane permeability of each compound were determined in
vitro, providing initial biological profiles of each test compound. As no in vivo testing was performed with
any of the compounds during this first round of drug discovery and optimisation, several computational
models were employed to extrapolate the in vitro and structural data to possible in vivo scenarios.
Two pyrrolidinone analogues (11.6 and 15.2) were identified as low micro-molar strand-transfer inhibitors
of wildtype-equivalent HIV-1 integrase, with low toxicity in cell culture and favourable solubility and
permeability profiles. Resistance screening of these two compounds against four mutant HIV-1 integrases
(Q148H; Q148H/G140S; N155H and N155H/E92Q) has shown some promise, with compound 15.2 retaining
a measure of activity against the Raltegravir-resistant N155-mutants. These hit candidates will form the
basis of structure-activity relationship optimisations in second- and further generation design stages.