Biology Reference
In-Depth Information
5 Hit Confirmation Using Assays with Human PDE5
From 1,242 retested compounds, 67 were selected for hit confirmation using
hPDE5. The selection was based on chemical properties (e.g., elimination of
reactive compounds), biological properties (e.g., elimination of compounds that
are active on multiple targets), and physicochemical properties. Considering that
the affinity of cGMP for the regulatory GAF domain is lower than for the catalytic
domain (Corbin et al.
2000
; Rybalkin et al.
2003b
), an allosteric inhibitor of the
GAF domain is expected to inhibit hPDE5 only under conditions of increased
intracellular cGMP levels, while basal PDE5 activity at low cGMP concentrations
should remain unaffected. On the other hand, a classical PDE5 inhibitor that
competes with cGMP at the catalytic center is expected to inhibit more effectively
at lower cGMP concentrations.
To compare the activity of nonactivated and activated PDE5, a two-step assay
procedure has been described. Protein is preincubated with up to 100
m
M of cGMP
on ice to activate PDE5, rapidly diluted, and, after addition of a
3
H-cGMP spike,
enzyme in 5 min at 37
C activity is determined. Under these conditions, a four- to
sixfold activation by cGMP is observed (Rybalkin et al.
2003a
). Using a similar, but
nonradioactive assay setup, we were unable to validate this assay with in-house
identified inhibitors of the tandem GAF domain in PDE5-GAF-CyaB1. Instead, we
used an alternative approach: Hydrolysis of fluorescein-tagged cGMP (Molecular
Devices) by native hPDE5 was detected by capillary electrophoresis using the
Labchip 3000 reader (Caliper Life Sciences). The enzymatic reaction was carried
out in 384 assay plates. Afterward, fluorescein-tagged cGMP and 5
0
-GMP were
separated in a microfluidic device, the 12 sipper chip, and detected fluorimetrically
after excitation with a 488 nm laser by a CCD camera combined with a 510-550 nm
band pass filter (Fig.
6a
). Peak heights of fluorescein-tagged cGMP and 5
0
-GMP are
used to calculate the percent of cGMP hydrolysis. Nonfluorescent constituents are
not detected and do not interfere. Fluorescein-tagged cGMP (200 nM) was used as
substrate either alone or in combination with untagged cGMP in order to obtain a
total cGMP concentration of either 5 or 20
m
M. Control experiments demonstrated
that hydrolysis of fluorescein cGMP is representative of total cGMP hydrolysis.
Thus, this assay is suitable for a broad range of cGMP concentrations. According to
preliminary experiments (e.g., Fig.
6b
), suitable enzyme amounts for the three total
cGMP concentrations were determined in order to obtain a turnover of 35% within
90 min at room temperature and a linear dependence of the turnover on enzyme
amount and reaction time.
A lysate of human platelets served as source for native hPDE5 (WO 2008/095835):
Platelet-rich plasma was prepared by centrifugation of citrate-treated blood (200
g
,
10min), and 1/10 volume of 85 mMNa
3
citrate/111 mM
D
-glucose/71 mMcitric acid/
pH 4.4 was added. After centrifugation (1,400
g
,10min),thepelletwasfirst
suspended in 140 mM NaCl/3.8 mM KCl/1 mM EGTA/1 mM MgCl
2
/20 mM
Tris-HCl/1 mM beta-mercaptoethanol/pH 8.2 plus a protease-inhibitor cocktail,
then sonicated, and centrifuged (15 min, 10,000
g
).
