Chemistry Reference
In-Depth Information
inhibitors in vivo, there are scant reports in the literature regarding the ecacy
of chronic DPP4 inhibition in rodent models. These observations highlight the
limitations of these pre-clinical disease-pertinent models in fully or accurately
mirroring the pathophysiology of human disease.
1.11 Peptidase Selectivity of Saxagliptin
At program initiation, the precise requirements for DPP4 selectivity versus
inhibition of related proteases were not entirely clear. SAR progression and
compound optimization outpaced the full implementation of our protease
selectivity panel, and it was not until identification of saxagliptin as a devel-
opment candidate that all elements of the selectivity panel were in place. Saxa-
gliptin (similar to other analogs within the series) was found to be highly
selective (44000-fold) for inhibition of DPP4 as compared to a variety of other
proteases, including FAP, PEP, and DPP2. Against the most closely related
enzymes DPP8 and DPP9, selectivity was attenuated but still high (400- and 75-
fold, respectively).
45,57
Contributing to saxagliptin's enhanced selectivity is the
observation that, unlike its slow tight-binding kinetics for the inhibition of
DPP4, saxagliptin did not display slow tight-binding when inhibiting DPP8 or
DPP9. Additionally, since DPP8 and DPP9 are intracellular enzymes, the poor
cellular permeability of saxagliptin attenuates access to these off-target pro-
teases. As a result, saxagliptin behaves as a highly selective inhibitor of DPP4
with little or no propensity for off-target protease inhibition.
In 2005, scientists at Merck published results of two-week rat (10, 30, and
100mg kg
1
) and single-dose dog (10mg kg
1
) toxicity studies with a cohort of
compounds which included a selective DPP4 inhibitor closely related to sita-
gliptin, a selective inhibitor of DPP8 and DPP9 (
L
-allo-isoleucyl isoindolide),
and a somewhat weak and non-selective inhibitor of all three enzymes (
L
-allo-
isoleucyl thiazolidide), as well as several other controls.
58
Both the DPP8/9
selective compound and the weak non-selective compound (which share a
common structural element in the
L
-allo-isoleucine moiety) caused alopecia,
thrombocytopenia, anemia, reticulocytopenia, splenomegaly, and mortality in
rats, and bloody diarrhea in dogs. None of these effects were observed in
animals dosed with the DPP4 selective compound. Based on these empirical
findings, the Merck team concluded that the observed toxicities were a direct
result of inhibition of DPP8 and/or DPP9. In formal toxicology studies, neither
saxagliptin nor vildagliptin produced the spectrum of pathology changes
ascribed by Merck to DPP8/9 inhibition, even at doses predicted to give plasma
levels much higher than the K
i
's required for inhibition of DPP8/9. Thus, the
findings suggested both saxagliptin and vildagliptin are highly selective DPP4
inhibitors in vivo, and/or the theory regarding the consequences of DPP8/9
inhibition proposed by the Merck group may be incorrect. An alternate
hypothesis, that the observed toxicology profiles of the Merck tool compounds
were structure (compound) related, rather than based on their DPP8/9 inhi-
bition profiles, was not initially embraced as a possible cause of the observed
Search WWH ::
Custom Search