86
S. B. Christensen et al.
10.1 Substituent at O-8
The pharmacophore model emphasizes that a flexible lipophilic side chain at O-8 is
needed for a low K D value. Unexpectedly, even very long and bulky side chains did
not significantly decrease the affinity although the binding cavity was too small for
these side chains [138]. X-ray analysis revealed that the side chains penetrate the
openings between the M3 and M5 transmembrane cavities, allowing the voluminous
molecule to create a space to enable binding [151]. The advantage of this discovery
was taken to construct prodrugs of thapsigargin (1) [94, 152]. As expected, removal of
the O-8 side chain substantially reduces the affinity from the sub-nanomolar K D value
for 1 to 10 nM for 8-O-debutanoyl-1 [79]. Replacement of the butanoate groups with
non-flexible acyl groups also decreases the affinity as compared to 1, and as would
be expected since these long side chains cannot penetrate the cavities between the
M3 and M5 transmembrane segments [152]. Nevertheless, thapsigargin analogs with
such replacements and which maintain a K D value below 10 nM, have been generated
[138], thus demonstrating the feasibility of generating prodrugs that can unmask
highly cytotoxic thapsigargin analogs with high affinity for SERCA. Inversion of
the stereochemistry at O-8 afforded an analog with a negligible affinity for SERCA
[153].
10.2 Substituent at O-10
Systematic replacement of the substituent at O-10 of thapsigargin revealed that this
side chain could only be varied to a very limited extent without severely impairing
SERCA-binding. Replacement with a butanoate group reduced the affinity about 100
times, and the introduction of a p-phenyl benzoate substituent reduced the affinity to
be outside the windows that were measured [18].
10.3 Substitution at C-2
A lipophilic substituent in this position is not important for SERCA binding, as
evidenced by the affinity of trilobolide (17), in which C-2 is a methylene group.
The affinity constant K D of trilobolide to SERCA is significantly higher than that of
thapsigargin, but is still in the nanomolar range [79]. Even a sidechain anchored to
a bead shows so large an affinity that affinity chromatography for isolating 1 from a
cell lysate failed because it was not possible to remove the ligand from the stationary
phase (Section 11) [75].
S. B. Christensen et al.
10.1 Substituent at O-8
The pharmacophore model emphasizes that a flexible lipophilic side chain at O-8 is
needed for a low K D value. Unexpectedly, even very long and bulky side chains did
not significantly decrease the affinity although the binding cavity was too small for
these side chains [138]. X-ray analysis revealed that the side chains penetrate the
openings between the M3 and M5 transmembrane cavities, allowing the voluminous
molecule to create a space to enable binding [151]. The advantage of this discovery
was taken to construct prodrugs of thapsigargin (1) [94, 152]. As expected, removal of
the O-8 side chain substantially reduces the affinity from the sub-nanomolar K D value
for 1 to 10 nM for 8-O-debutanoyl-1 [79]. Replacement of the butanoate groups with
non-flexible acyl groups also decreases the affinity as compared to 1, and as would
be expected since these long side chains cannot penetrate the cavities between the
M3 and M5 transmembrane segments [152]. Nevertheless, thapsigargin analogs with
such replacements and which maintain a K D value below 10 nM, have been generated
[138], thus demonstrating the feasibility of generating prodrugs that can unmask
highly cytotoxic thapsigargin analogs with high affinity for SERCA. Inversion of
the stereochemistry at O-8 afforded an analog with a negligible affinity for SERCA
[153].
10.2 Substituent at O-10
Systematic replacement of the substituent at O-10 of thapsigargin revealed that this
side chain could only be varied to a very limited extent without severely impairing
SERCA-binding. Replacement with a butanoate group reduced the affinity about 100
times, and the introduction of a p-phenyl benzoate substituent reduced the affinity to
be outside the windows that were measured [18].
10.3 Substitution at C-2
A lipophilic substituent in this position is not important for SERCA binding, as
evidenced by the affinity of trilobolide (17), in which C-2 is a methylene group.
The affinity constant K D of trilobolide to SERCA is significantly higher than that of
thapsigargin, but is still in the nanomolar range [79]. Even a sidechain anchored to
a bead shows so large an affinity that affinity chromatography for isolating 1 from a
cell lysate failed because it was not possible to remove the ligand from the stationary
phase (Section 11) [75].