4
H.-X. Wei et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
O
OH
N
N
42
39
41% at 50 µM, Abeta 40
82% at 100 µM, Abeta 40
Gamma-secretase inhibition
Gamma-secretase inhibition
O
O
O
N
S
Cellular assays:
N
IC50 = 0.12 µM, Abeta 40
IC50 = 0.13 µM, Abeta 42
IC50 = 10 µM, Notch
Cl
Et
44
Et
CO2H
43 58% at 100 µM, Abeta 40
Gamma-secretase inhibition
Figure 2. Evolution of naphthyl amino alcohols.
hydroxyl group. As shown in Table 1, shortening (2) or increasing
(3) the linker by one methylene group significantly decreased the
the b-amino group. These compounds (Table 5) were synthesized
via the Mannich reaction, using our new one-pot reaction, coupling
acrolein (37% in water) and isopropyl benzyl amine in ethanol at
90–92 °C for 40 h to form an amino-aldehyde intermediate. After
evaporation of all solvents and drying (oil pump), the amino-
aldehyde residue was dissolved in anhydrous THF at 0 °C, and
the corresponding Grignard reagents were added to give the desire
products.
inhibitory effects on
c-secretase-mediated Ab40 production and
altered the Notch-sparing feature seen in compound 1. In contrast,
methylation of the hydroxyl group (4) did not change the Notch-
sparing profile whereas inhibition on Ab40 production was some-
what reduced.
Analogs in which the naphthyl ring of compound 1 replaced
with various aryl groups (R1) are shown in Table 2. The results indi-
cate that the presence of two aromatic rings in the aryl moiety (R)
appears to be critical for preserving the Notch-sparing property.
Although substituted mono-aryl analogs (e.g., 12–17) did show a
range of low to moderate potency, the Notch-sparing effects were
abolished.
The data on these aryl substituted b-amino alcohols illustrated in
Table 2 indicated that the original naphthyl analog 1 was still the
most potent analog with the strongest inhibition on c-secretase-
mediated Ab40 production. We investigated the activity of each
enantiomer of analog 1 by reducing ZM39923 (Fig. 1) with (R)-(+)-
2-methyl-CBS oxazaborolidine and also with (S)-(+)-2-methyl-CBS
oxazaborolidine to give the respective (R)- and (S)-enantiomers of
Entry 1. Neither of these stereoisomers showed an improvement
in Abeta 40 activity. Hence, our attention was then focused on intro-
It was found that introducing two geminal alkyl groups (i.e.,
methyl and ethyl) at the b-position of compound 1 did not abolish
the inhibition on
c-secretase-mediated Ab40 production, while
Notch-sparing effects were preserved (e.g., 38–41). The most inter-
esting dialkyl substituted analog 39 showed very good inhibition in
cells against Abeta 40 and 42 (80% and 100% @ 100
tively) as well as very good selectivity for the Notch receptor with
an IC50 >75 M.
Oxidation of the hydroxyl group of 39 gave the keto-analog 42
which showed modest inhibition (40% @ 50 M). Studies ongoing
lM, respec-
l
l
in our laboratories at this time were investigating aryl amides
(e.g., 42) and various size linkers.
These studies converged and led to more potent and selective
sulfonamide analogs like 44 (Fig. 2).
In summary, based on the structure of JAK3 inhibitor, ZM39923,
ducing various
c-amino groups (Tables 3 and 4).
an unstable b-amino ketone, we synthesized a series of c-amino
The results shown in Table 3 suggested that alkyl substitution
on the amino group significantly influences inhibitory effects on
c
groups (e.g., isobutyl, 21) abolished the activity on inhibition of
Ab40 production. The isopropyl group (R2) is by far the best sub-
stituent within the simple alkyl groups investigated.
alcohol analogs (I). Within this class of compounds, the presence
of two aromatic rings, such as naphthalene-2-yl and phenoxy-
phenyl, in the immediate proximity of the alcohol functionality,
consistently provided Notch-sparing effects. The benzyl group
seemed to be another important feature that affected the inhibi-
-secretase. Notably, replacing the isopropyl of R2 with larger alkyl
tion of c-secretase-mediated Ab40 production and Notch-sparing
Interestingly, replacing the benzyl group with a phenyl ring (24)
properties. Additionally, substitution of the two b-protons did
maintained similar potency on inhibition of
c-secretase-mediated
not alter the overall profile of Notch-sparing inhibition of
Ab40 production, but the Notch-sparing property was lost. Further,
replacing the entire amino group of (1) with phenyl substituted
piperizine groups (25–29) seemed tolerated, but the potency var-
ied and highly relied on the substitution of the phenyl ring. The
aforementioned structural modifications indicated the importance
of several fragments, including the naphthalene-2-yl, iso-propyl
and benzyl functionalities.
c-secretase, but should add chemical stability to the framework
in that the amino functionality cannot be eliminated by retro-
Michael addition. Compounds of this type provided a valuable
template that led to sulfonamide analogs for animal studies and
potential differentiation from failed clinical candidates.
Acknowledgments
We further introduced some substituents on the benzyl group
of the amino moiety (Table 4). The results suggested that fluoro-
and chloro- groups are well-tolerated. Although no significant
improvement in potency was observed, halogen substitutions
could potentially help brain penetration and effect metabolic
stability.
We acknowledge Jian Chen and Katherine M. Brogan for their
assistance with the biological testing. We acknowledge the Alzhei-
mer Drug Discovery Foundation, the Harvard NeuroDiscovery Cen-
ter and members of the Center for Neurologic Diseases for funding
support.
As previously mentioned, conversion of the ketone group of
ZM39923 into a hydroxyl group in compound 1 (Scheme 2), we
reduced the chemical instability of this b-aminoketone skeleton.
However, it is also possible that the hydroxyl group of I may not
be very stable in vivo. Hence, we intentionally introduced two alkyl
Supplementary data
Supplementary data associated with this article can be found, in
groups (R7) at the
a-position to prevent the possible elimination of