From HTS to Market: the Discovery and Development of Maraviroc, a CCR5 Antagonist for the Treatment of HIV - Accounts in Drug Discovery Case Studies in Medicinal Chemistry - page 187

LE ¼ 0.21). 27 These are shown in Figure 9.2, 27-32 alongside a number of other
CCR5 antagonist leads found by others through screening campaigns. Com-
parison of these structures show that they possess similar molecular weights,
lipophilicities and, in most cases, a central basic centre.
In addition to their lipophilicity and relatively high molecular weight, these
hits showed only weak anity for the CCR5 receptor and no antiviral activity.
While clearly desirable, the initial lack of antiviral activity was not critical to
progression, since we had little understanding of the dynamic range of the
antiviral assay. However, if antiviral activity had continued to be lacking with
more potent CCR5 binders, then a re-evaluation of this approach would have
become essential.

9.5 Defining Objectives and an Appropriate Screening
Sequence
The programme was initiated with the objectives of identifying a safe, orally
effective agent that would be administered twice (but preferably once) daily. To
be safe, a compound would need to be a selective CCR5 antagonist that
avoided drug-drug interactions when dosed in conjunction with other agents
used to treat HIV (see HAART above). This strategic objective was translated
into a lab profile as follows:

antiviral (AV) IC 90 o 10 nM
4100-fold selective over other targets (specifically other GPCRs)
low turnover in human liver microsomes (HLM), t 1/2 4 30min
apical to basal flux in caco-2 cells 45 (believed predictive of oral
absorption)
no cytochrome P450 inhibition (to minimise potential drug-drug
interactions)

While antiviral assays were available which measured the ability of com-
pounds to block HIV Bal from infecting PM-1 or PBL cells, these were low
throughput and somewhat capricious, certainly not able to deliver sucient
data to drive structure-activity relationships (SAR). However, without
recourse to regular antiviral screening, we ran the risk of optimising binding to
a region of the CCR5 receptor which had no influence on the binding of viral
gp120. This drove the parallel development of additional higher throughput
assays that would give better translation to antiviral activity than our initial
MIP-1b binding assay. The solution was a cell fusion assay run in a 384-well
plate format. It consisted of two cell lines: (1) HeLa P4 expressing CCR5, CD4
and HIV-1-LTR-b-Gal and (2) a CHO cell line expressing viral gp160 and
TAT. If cell fusion takes place, then the TAT in the CHO cell would trans-
activate the HIV-1LTR present in HeLa cell, leading to the measurable
expression of the b-galactosidase enzyme. 33 A retrospective comparison of the
MIP-1b binding, cell

fusion and antiviral assays (Figure 9.3) shows an