Aspartic protease inhibitors via C1-homologation of peptidic aldehydes and studies on reduced amide isosteres

Article abstract of DOI:10.1016/j.tetlet.2007.09.047

(R)-Configured isophthalic hydroxyethylamines play an important role in the inhibition of β-secretase (BACE1). We present the synthesis of a number of (S)-configured hydroxyethylamine derivatives via 2-iodoethanol intermediates and the comparison with the

Full text of DOI:10.1016/j.tetlet.2007.09.047

Tetrahedron Letters 48 (2007) 7990–7993
Aspartic protease inhibitors via C1-homologation of
peptidic aldehydes and studies on reduced amide isosteres
Hannes A. Braun,a Andrea Zall,a Manfred Brockhaus,b Marco Schutz,a
¨
Reinhard Meusingera and Boris Schmidta,*
aClemens Scho¨pf Institute for Organic Chemistry and Biochemistry, Darmstadt University of Technology,
Petersenstr. 22, 64287 Darmstadt, Germany
bDepartment of PRBD-N Neuroscience, F. Hoffmann-La Roche, 4070 Basel, Switzerland
Received 29 June 2007; revised 29 August 2007; accepted 7 September 2007
Available online 12 September 2007
Abstract—(R)-Configured isophthalic hydroxyethylamines play an important role in the inhibition of b-secretase (BACE1). We
present the synthesis of a number of (S)-configured hydroxyethylamine derivatives via 2-iodoethanol intermediates and the compar-
ison with the (R)-analogues. An N-substituted indole was investigated as a substitute for the isophthalamide moiety.
Ó 2007 Elsevier Ltd. All rights reserved.
Extracellular amyloid-b plaques are suspected to cause
Alzheimer’s disease.1,2 These plaques consist mainly of
amyloid-b peptides (Ab) which are generated from the
membrane bound APP (amyloid precursor protein) via
the consecutive cleavages by b-secretase (BACE1, b-site
amyloid precursor protein cleaving enzyme) and c-secre-
tase.3 Thus, the inhibition of BACE1 constitutes a
promising approach to block the onset of plaque forma-
tion by decreasing the Ab levels in the brain.
A number of highly potent, but peptidic BACE1 inhib-
itors have been available for several years.4,5 Inhibitors
1–3 have derived from the progressive elimination of
peptidic features from the original heptapeptidic
hydroxyethylenes (Fig. 1). The introduction of an iso-
Figure 1. BACE1 inhibitors by Elan pharmaceuticals and MSD.
phthalic moiety in 1 was regarded as a milestone in
the development of small BACE1 inhibitors.6 The
peptidic character of compound 1 was further reduced
by the replacement of the hydroxyethylene with a
hydroxyethylamine. Remarkably, the hydroxyethyl-
amine in 2 is the only transition state isostere in BACE1
inhibitors that displays an (R)-configured secondary
alcohol.7 Compound 3 comprises a reduced amide
transition state isostere with an additional sulfonamide
substituent in the P2 position and a modified P3 amide
residue.8 Compounds 1–3 are active at nanomolar con-
centrations in cell free assays and yet display different
cellular activities. They are good substrates for the
p-glycoprotein transporter, this was partially assigned
to the isophthalic amide moiety.9
Here, we present an approach to isophthalamide based
hydroxyethylamines and potential replacements thereof.
We compared a linear to a convergent access to this
class of compounds. The synthesis led to the (S)-config-
ured secondary alcohols which were compared to the
known (R)-analogue 2. Furthermore, we investigated
Keywords: b-Secretase; Transition state isostere; Hydroxyethylamine;
Reduced amide; Grignard reagent; Indole synthesis.
*
Corresponding author. Tel.: +49 6151 163075; fax: +49 6151
163278; e-mail: schmidt_boris@t-online.de
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.09.047

H. A. Braun et al. / Tetrahedron Letters 48 (2007) 7990–7993
7991
Figure 2. Synthesis of epoxides 10a and 10b and ester 12. Reagents and conditions: (a) KOH, MeOH, rt; (b) SOCl2, CHCl3, reflux, (overnight); (c) n-
Pr2NH, CH2Cl2, 0 °C to rt; (d) L-phenylalaninol, CH2Cl2, 0 °C to rt; (e) IBX, DMSO, rt, 15 h; (f) i-PrMgCl, CH2I2, THF, À78 °C (15 min), 0 °C
(2 h); (g) K2CO3, MeCN, rt (4 h); (h) 4-bromobenzylamine, MeCN, reflux (4 h); (i) water, THF, trifluoroacetic acid, 80 °C (1 h), rt (15 h).
an indole as a replacement for the isophthalamide por-
tion, combined it with a reduced amide isostere and
compared it to the isophthalamide analogue.
ative (e.g., 15) or its respective epoxide, which does not
undergo cyclization. This difference in carbamate versus
amide reactivity was applied to the synthesis of the
desired hydroxyethylamines. Boc-phenylalaninol 13
was oxidized to 14 with IBX and converted to b-iodo-
ethanols 15a and 15b (de >99%, crude ds = 3:2, Fig. 3).
A reaction sequence starting from diester 4a was
employed for the generation of 5a, 6a and finally the
isophthalamide 7a comprising two different amide sub-
stituents on either carboxyl functionality (Fig. 2).10
Alcohol 7a was oxidized by IBX (2-iodooxybenzoic
acid) furnishing aldehyde 8a in a high yield and free of
racemisation. The aldehyde was converted to a 2-iodo-
ethanol using a moderately stereoselective C1-homo-
logation.11 The reaction occurs upon treatment with
i-PrMgCl/CH2I2 forming a mild Grignard reagent that
undergoes addition to peptidic aldehydes in a high che-
moselectivity leaving the a-chiral position unchanged.
Product 9a (crude ds = 3:1) was obtained in 75% yield
as a diastereomeric mixture and further purification
delivered pure 9a. Treatment of 9a with K2CO3 yielded
epoxide 10a. Surprisingly, 9a was converted to oxazine
11a upon treatment with any benzylamine in MeCN
under reflux. 11a could be opened to ester 12 by treat-
ment with trifluoroacetic acid in a water/THF mixture.
A small amount of the diastereomer of 11a was gener-
ated likewise, and the cyclic structures of both diastereo-
mers of 11a served to assign the configuration of
compounds 9–12. The reactions to the analogous com-
pounds 5b–11b starting from dimethyl 2,6-pyridinedi-
carboxylate 4b were conducted in a similar manner.
The higher yield of 9b (de = 100%) compared to 9a is
explained by the higher diastereoselectivity of the
C1-homologation (crude ds = 9:1). Even though the
generation of 9a or 9b seemed to be a promising
approach for a variety of benzylic hydroxyethylamines,
the iodo compounds reacted under a variety of condi-
tions to the undesired oxazines. This undesired reaction
was circumvented in previous synthetic approaches:12
(a) by the protection of the amide nitrogen or (b) by
the conversion of an N-Boc-protected amino acid deriv-
Either diastereomer was obtained in high purity from
the crude diastereomeric mixture (53%). Substitution
of the iodide by various benzylamines provided
hydroxyethylamines 16. Deprotection and reaction with
chloride 6a led to the desired (S)-configured alcohols
17–20 and (R)-alcohol 2. The configuration was assigned
by single crystal X-ray analysis of 15a and subsequent
conversions of 15b to known compounds and compari-
son of the NMR data.11
Figure 3. Reagents and conditions: (a) IBX, DMSO, rt (15 h); (b)
i-PrMgCl, CH2I2, THF, À78 °C (15 min), 0 °C (2 h); (c) amine,
MeCN, reflux (4 h); (d) trifluoroacetic acid, CH2Cl2; (e) 6a, HOBt,
CH2Cl2 (overnight).

7992
H. A. Braun et al. / Tetrahedron Letters 48 (2007) 7990–7993
Epoxides 10a, 10b, oxazine 11a, ester 12 and hydroxy-
ethylamines 2, 17–20 were tested in BACE1-assays (fluo-
rescence resonance energy transfer (FRET) assay,13
RLBA (radioactive ligand binding assay)14 in the pres-
ence of detergent (Tween 20) or serum albumin (BSA))
and a FRET Cathepsin D assay13 (Table 1). The epox-
ides proved to be inhibitors equally of both BACE1
and Cathepsin D. In further studies using the RLBA as-
say compound 10a showed an irreversible, time-depen-
dent inhibition. The apparent IC50 was 39 lM after
10 min and 3 lM after 60 min incubation. The same
compound was not active in a cellular system measuring
Ab production by HEK293 cells. Compound 11a was
investigated because of the structural similarity to an
oxazole reported by Rajapakse et al.15 While the oxazole
was inactive, its hydrolyzed linear form was a highly
potent inhibitor. Similar to the oxazole, oxazine 11a
displayed no inhibition of BACE1 and its ring-opened
derivative 12 was a moderate inhibitor of BACE1.
Interestingly, the (S)-configured hydroxyethylamines
displayed no inhibition of BACE1. This proves the
importance of the opposite (R)-configuration on the
secondary alcohol in hydroxyethylamines compared to
other transition state isosteres.
Figure 4. Reagents and conditions: (a) (i) SOCl2, CHCl3, reflux
(overnight) (ii) MeOH, Et3N, CH2Cl2, 0 °C to rt; (b)
HC(NMe2)(OMe)2, CuI, DMF, microwave (180 °C, 10 min); (c)
Pd/C, H2, MeOH, 3 d; (d) MeSO2Cl, THF, rt (4 h); (e) NaH, MeI,
THF, rt (1 h); (f) K2CO3, i-BuI, DMF, 80 °C (2 d); (g) KOH, MeOH,
reflux (2 d); (h) (i) EDAC, HOBt, CH2Cl2, (ii) 29b, Et3N.
In search of an isophthalamide replacement we and
others12,16 hypothesized that the indole nitrogen in a
(4-amino-6-carboxyl)-indole (e.g., 23) could be attached
to an alkyl substituent mimicking the P3 residue, while
the amine functionality can be converted into a sulfon-
amide and the carboxylate serves as a linker to a transi-
tion state isostere. Such indoles display H-bonds and
thus may exhibit a reduced affinity for the p-glyco-
protein transporter9 and the enhanced rigidity might
increase their activity. The synthesis started from the
C2-symmetrical acid 21 (Fig. 4). After esterification, an
enamine was generated that underwent cyclization to
the indole in a subsequent Pd/C catalyzed reduction.17
The resulting amine was converted into a methyl sulfon-
amide and the amide nitrogen was methylated after
selective deprotonation. After a second deprotonation,
the indole nitrogen was alkylated with i-BuI and the es-
ter was saponified to furnish building block 25.
A reduced amide transition state isostere was chosen for
elongation. Coburn et al.8 reported on a variety of
reduced amides similar to 3 and two of the most active
isosteres were generated via reductive amination based
on a Z-strategy. Z-L-alanine 27 was coupled with
i-BuNH2 or cyclopropylamine (c-PrNH2) to the respec-
tive amides (Fig. 5). The resulting amides were deprotec-
ted and converted to 28a and 28b by treatment with
N-Z-L-phenylalaninal and Na(AcO)3BH. The primary
amines were deprotected to provide the transition state
isosteres 29a and 29b. Compound 29b was coupled to
indole 25 in a final step leading to inhibitor 26. The
secondary amine 29a was coupled with 6a to provide
compound 30 in order to compare the activity of indole
26 with other P2–P3 mimetics. Indole 26 is active at
micromolar concentrations and exhibits a similar activ-
Table 1. IC50 of compounds 2, 10a, 10b, 11a, 12, 17–20, 26 and 30
Compounds
BACE1
CatD
FRET RLBA (tween) RLBA (BSA) FRET
IC50
IC50 (lM)
IC50 (lM)
IC50
(lM)
(lM)
2
1.2
106
95.7
0.14
10.4
15.4
—
37.9
34.2
—
200
—
25
0.19
11.0
16.2
—
36.4
37.9
—
127
—
20
0.1
63.0
194.2
197.6
39.0
19.3
63.3
12.3
69.2
75
10a
10b
11a
12
17
18
19
20
26
30
—
91.1
76.5
>200
200
—
Figure 5. Reagents and conditions: (a) (i) EDAC, HOBt, CH2Cl2, (ii)
amine, Et3N; (b) Pd/C, H2, MeOH, 3 d; (c) N-Z-L-phenylalaninal,
Na(AcO)3BH, Et3N, MgSO4, CH2Cl2, rt (overnight); (d) 6a, HOBt,
CH2Cl2, rt (15 h).
35
17.0
4.6
8.6
84.9
No inhibition: —.

H. A. Braun et al. / Tetrahedron Letters 48 (2007) 7990–7993
7993
3. Weidemann, A.; Eggert, S.; Reinhard Friedrich, B. M.;
Vogel, M.; Paliga, K.; Baier, G.; Masters Colin, L.;
Beyreuther, K.; Evin, G. Biochemistry 2002, 41, 2825–
2835.
ity like isophthalamide 30 (Table 1). Surprisingly, this
results in an approximately 1000-fold decreased activity
compared to amide 3 (IC50 = 24 nM).
4. Guo, T.; Hobbs, D. W. Curr. Med. Chem. 2006, 13, 1811–
1829.
In conclusion, we confirmed the (S)-configured hydroxy-
ethylamines to be inactive and thus to constitute an
exception to all other transition state isosteres employed
in BACE1 inhibition. However, (S)-configured hydro-
xyethylamines are known as isosteres in inhibitors
targeting other aspartyl proteases. The depicted C1-
homologation resulted in useful intermediates for such
protease inhibitors. The direct nucleophilic opening of
acyl N-protected a-aminoepoxides is an attractive syn-
thesis to bioactive molecules, but turned out to be inac-
cessible. However, the initially undesired cyclization
allowed us to unambiguously establish the stereochemis-
try due to cyclic constraints. The combination of an
indole bearing an i-butyl P3 residue and a methyl sulfon-
amide P2 residue proved equally potent as the N,N-
dipropylisophthalamide analogue, but neither of the
compounds displayed activities comparable to the most
potent BACE1 inhibitors.
5. Schmidt, B.; Baumann, S.; Narlawar, R.; Braun, H. A.;
Larbig, G. Neurodegenerative Dis. 2006, 3, 290–297.
6. Hom Roy, K.; Gailunas Andrea, F.; Mamo, S.; Fang
Larry, Y.; Tung Jay, S.; Walker Donald, E.; Davis, D.;
Thorsett Eugene, D.; Jewett Nancy, E.; Moon Joseph, B.;
John, V. J. Med. Chem. 2004, 47, 158–164.
7. Maillaird, M.; Hom, C.; Gailunas, A.; Jagodzinska, B.;
Fang, L. Y.; John, V.; Freskos, J. N.; Pulley, S. R.; Beck,
J. P.; Tenbrink, R. E., WO2002002512, 2002.
8. Coburn, C. A.; Stachel, S. J.; Jones, K. G.; Steele, T. G.;
Rush, D. M.; DiMuzio, J.; Pietrak, B. L.; Lai, M. T.;
Huang, Q.; Lineberger, J.; Jin, L.; Munshi, S.; Katharine
Holloway, M.; Espeseth, A.; Simon, A.; Hazuda, D.;
Graham, S. L.; Vacca, J. P. Bioorg. Med. Chem. Lett.
2006, 16, 3635–3638.
9. Schinkel, A. H. Adv. Drug Delivery Rev. 1999, 36, 179–
194.
10. Umbreen, S.; Brockhaus, M.; Ehrenberg, H.; Schmidt, B.
Eur. J. Org. Chem. 2006, 4585–4595.
11. Braun, H. A.; Meusinger, R.; Schmidt, B. Tetrahedron
Lett. 2005, 46, 2551–2554.
Acknowledgements
12. Kleinman, E. F.; Murray, J. C., WO2006032999, 2006.
13. Gruninger-Leitch, F.; Schlatter, D.; Kung, E.; Nelbock,
P.; Dobeli, H. J. Biol. Chem. 2002, 277, 4687–4693.
14. Brockhaus, M.; Doebeli, H.; Grueninger, F.; Huguenin,
This work was supported by the Deutsche Forschungs-
gemeinschaft (SPP1085 SCHM1012-3) and the EU Con-
tract LSHM-CT-2003-503330 (APOPIS).
P.;
Kitas,
E.
A.;
Nelboeck-Hochstetter,
P.,
US2003125257, 2003.
Supplementary data
15. Rajapakse, H. A.; Nantermet, P. G.; Selnick, H. G.;
Munshi, S.; McGaughey, G. B.; Lindsley, S. R.; Young,
M. B.; Lai, M. T.; Espeseth, A. S.; Shi, X. P.; Colussi, D.;
Pietrak, B.; Crouthamel, M. C.; Tugusheva, K.; Huang,
Q.; Xu, M.; Simon, A. J.; Kuo, L.; Hazuda, D. J.;
Graham, S.; Vacca, J. P. J. Med. Chem. 2006, 49, 7270–
7273.
16. Redshaw, S.; Demont, E. H.; Walter, D. S.,
WO2005058915, 2005.
17. Siu, J.; Baxendale, I. R.; Ley, S. V. Org. Biomol. Chem.
2004, 2, 160–167.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.2007.
09.047.
References and notes
1. Selkoe, D. J. Physiol. Rev. 2001, 81, 741–766.
2. Adalbert, R.; Gilley, J.; Coleman, M. P. Trends Mol. Med.
2007, 13, 135–142.