5574
R. Guidetti et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5572–5575
and 23% in dog from these studies. Although these values were
lower than determined for compound 1,4 PK-PD modelling pre-
dicted that human PK parameters and doses would be acceptable
to support a QD regimen. A cerebrospinal fluid (CSF) and brain
PK profile was also obtained post 1 mg/kg oral dose. Plasma, brain
and CSF exposures correlated very closely over time (data not
shown) with a total brain/plasma ratio of 0.4 and a total CSF/
plasma ratio of 0.1. Correcting the brain/ plasma ratio for unbound
exposures18 gave a Kpu,u value of 0.2, comparable to compound 1.4
Compound 6 was not found to be a substrate for the human P-gp
transporter in a bi-directional permeability study using an MDCK
cell line19 although other potentially relevant brain efflux trans-
porters were not evaluated.18 Profiling of the compound indicated
the compound was neither a significant inducer nor inhibitor of the
clinically relevant CYP isoforms (3A4, 1A2, 2D6, 2C9) up to concen-
Figure 3. Dose dependent effect of CB2 agonist 6 in the MIA induced joint pain
assay. Mean SEM n = 5 animals per group. The CB2 agonist 6 at doses (and
unbound plasma exposures) of 0.1 (3.0 nM), 0.3 (10 nM), and 1 mg/kg (30 nM) and
trations of 10 l
M.12 Using recombinant human P450 enzymes, we
diclofenac at 5 mg/kg significantly reduced pain compared to vehicle (⁄/⁄⁄ ⁄⁄⁄p <0.05
/
by Student’s t-test). Each increasing dose of 6 was significantly different from the
preceding one. Day 12 post MIA injection, differential weight bearing measured 2 h
post compound dosing.
had shown that Cyp3A4 was only isoform involved in the oxidative
turnover of compound 1. Similar studies with compound 6 showed
involvement of CYPs 3A4, 2C9 and 1A2 in the turnover with no one
isoform contributing more than 80% to the metabolite formation.
In addition, analysis of rodent urine and bile after an oral dose of
6 indicated the presence of phase 2 glucuronide and sulfate metab-
olites. Taken together, this data indicates the potential for com-
pound 6 to be cleared via multiple mechanisms and to be at less
risk than compound 1 of drug–drug interactions in clinical use.
To establish efficacious doses and exposures in rat, compound 6
was tested in a monoiodoacetic acid (MIA) assay of joint pain20
(Fig. 2). The injection of MIA into the knee joint of rats produces
an acute inflammatory insult which over a period of 7–10 days
then develops into chronic degeneration of the tissues in the
injected joint. The pain resulting from the joint injury can be mea-
sured via differential weight bearing of the hind legs using an inca-
pacitance tester. Efficacy is measured by the ability of a test
compound to normalize weight distribution presumably by reduc-
ing joint pain. The MIA model has been extensively described in
the literature21 and has been used to demonstrate pain reversal
for a variety of mechanisms with compounds showing efficacy at
plasma exposures comparable to clinically effective human
exposures.22
in this study and the effect was completely blocked by the CB2
antagonist SR-144528 (the antagonist alone having no effect).
Compound 6 was then tested for a dose dependent response in
the MIA model (Fig. 3) alongside the NSAID diclofenac, a standard
of care for the treatment of inflammatory joint pain. Compound 6
demonstrated a clear dose related reversal of MIA induced joint
pain with an ED50 of 0.16 mg/kg po where it displayed comparable
efficacy to diclofenac dosed at 5.0 mg/kg po. This dose of diclofenac
was selected as it gave an exposure in this assay comparable to
that measured in humans at a clinically effective 50 mg dose. The
ED50 of compound 6 was comparable to that determined for com-
pound 1 under the same MIA assay conditions confirming the
equivalent efficacy of these molecules. To demonstrate that toler-
ance24 does not develop on repeat dosing in the MIA model, com-
pound 6 was dosed at 1.0 mg/kg bid for 5 days. On day 5, the
efficacy of compound 6 in the MIA assay was in fact statistically
superior to that measured on day 1 with exposures of the com-
pound remaining constant over this time (results not shown).
In support of a lack of central CB1 activity, oral administration
of 6 did not induce hypothermia25 or behavioral impairment in a
rotorod study26 in rats up to 30 mg/kg po (data not shown). In con-
trast, the nonselective cannabinoid agonist CP-55,940 (0.5 mg/kg
ip) tested as a positive control, caused a significant hypothermia
and impaired rotorod performance thereby confirming the high
selectivity of 6 for CB2 over CB1 receptors in vivo (data not shown).
The general synthetic routes for the preparation of the purines
described in this communication have been discussed in detail
elsewhere3,4 and are shown in Scheme 1 for the preparation of
compound 6.27 The synthesis proceeded in three steps using com-
mercially available materials in an overall yield of 55%. The dichlo-
ropyrimidine (12) was reacted sequentially with 2-chlorobenzoyl
chloride and 2(R)-aminopropanol to give the intermediate (13).
This material then underwent a second SnAr reaction with 1-meth-
ylpiperazine and under the forcing conditions of the reaction (high
temperature and pressure), a concomitant cyclization occurred to
give the desired purine (6). The absolute configuration was con-
firmed as (R) by single crystal X-ray crystallography.
In the experiment summarized in Figure 2, rats were unilater-
ally injected intra-articularly with 0.3 mg of MIA into the right
knee joint with saline injected into the left knee joint on day 0.
Twenty days later, the rats were randomized and dosed orally with
either vehicle, 1 mg/kg of compound 6, 3 mg/kg of a selective CB2
antagonist SR-14452823 or a combined dose where antagonist was
administered 30 min prior to compound 6. Efficacy was measured
using incapacitance testing 2 h post agonist dosing. Compound 6
significantly inhibited differential weight bearing versus vehicle
25
20
15
*
10
5
In summary, we have further explored the SAR of a series of
purine CB2 agonists culminating in the discovery of compound 6,
a brain penetrating CB2 selective agonist with excellent drug-like
properties (mol weight = 401, logP = 3.1, pKa = 7.5, PSA = 70, aque-
ous solubility = 1.8 mg/mL, pH 6.5). Issues relating to off target
cross reactivity and potential drug-drug interaction liabilities
observed with compound 1 were resolved by the introduction of
a hydroxyalkyl side chain at N9 of the purine scaffold. The attrac-
tive efficacy, PK and safety profile of compound 6 has led to its
0
Vehicle
1mg/kg
3mg/kg
combined dose
LSN3038404 HCl
Compound 6
LSN624185
SR-144528
Figure 2. Efficacy of a selective CB2 agonist 6 in the MIA induced joint pain model20
and reversal with
a
selective CB2 antagonist SR-144528.23 Mean SEM n = 6
animals per group. ⁄Indicates p <0.5 versus vehicle (Dunnett’s test). Vehicle is 1%
carboxy methyl cellulose (CMC) in water including 0.25% Tween 80.
Guidetti, Rossella
