€
A.J. Ruger et al. / Tetrahedron 68 (2012) 8823e8829
8826
Again the amide functionality showed a high nucleophilicity, be-
cause intramolecular cyclization was faster than addition of mod-
erately nucleophilic alcohol 27 and yielded the bicyclic system 33 in
good yield. Changing the base to potassium carbonate also led to
the intramolecular addition albeit in lower yield (33%). The use of
the stronger base sodium hydride led to a coupled product 34 in
a small amount and in addition caused defluorination at the
trifluoroethyl-unit. Due to low yield only one isomer (cis or trans) of
compound 34 could be isolated and identified, though a mixture of
isomers was employed for alcohol 27.
As it is shown in Scheme 9 further steps for the synthesis of both
retroamides proceeded in good to excellent yields. We were able to
get an X-ray of the cis isomer of target structure 4, which helped us
once more to determine the configuration of the cyclohexyl-
substituents. The solid state structure of this bisamide is surpris-
ing since both carbonyl groups point in a similar direction (Fig. 7).
Importantly, there are no intramolecular hydrogen bonds between
the amide groups, thus we assume the structure to be less rigid
than is the case for the lead structure 3, which we hoped would
enhance water solubility.
The second bromide 32 could easily be employed for the syn-
thesis of amine 35 and ether 36 without degradation or other side
reactions (Scheme 7). To facilitate substitution with phenol 25 we
added tetrabutylammonium iodide to the reaction mixture.
2.5. Amide-mimicseurea derivatives
Besides retroamides urea derivatives were accessible via Curtius
rearrangement too.16 Again starting from carboxylic acids 12 and 13
the significant moiety was introduced within one step. Whereas
urea formation at position 3 of the pyrazole occurred at room
temperature we observed amide synthesis at position 4 using the
same reaction conditions. Therefore we heated the reaction mix-
ture to 90 ꢀC before adding the amine to facilitate rearrangement of
the azide to the intermediate isocyanate (Scheme 10).
Saponification of ester 46 led directly to the bicyclic pyr-
imidinedione 49. The free acid 48 was only obtained as a side
product. Deprotection in acidic media or under nucleophilic con-
ditions using lithium iodide did not lead to free acid 48 either
(Scheme 11). Therefore we decided to convert ester 46 directly into
an amide using a protocol described by Weinreb.17 Trimethylalu-
minium allows the one step conversion of esters into amides
through activation of the corresponding amine in situ. Thus we
were able to synthesize urea 50 in low yield, but unreacted ester 46
could be recovered after column chromatography.
A standard protocol using EDC was used for the synthesis of the
other urea derivative 5. Employing saponification of ester 47 fol-
lowed by amide-synthesis we were able to synthesize urea 5 in
medium yield over two steps (Scheme 12).
Scheme 7. Synthesis of amine 35 and ether 36. Reagents and conditions: (a) K2CO3,
DMF, 90 ꢀC, 5 he17 h, up to 46%; (b) K2CO3, Bu4NI, DMF, 90 ꢀC, 17 h to days, up to 77%.
2.4. Amide-mimicseretroamides
For the synthesis of retroamides 4 and 43 nitrogen substituted
pyrazoles were needed. Those could easily be obtained from car-
boxylic acids 12 and 13 through Curtius rearrangement.16 After
saponification both Boc-protected amino carboxylic acids 38 and 40
were synthesized in good yields and readily crystallized (see
Scheme 8, X-ray crystal structures shown in Fig. 6).
2.6. Other bioisosteres
In order to improve the water solubility of the lead structure, we
introduced
a pyridine-moiety as a replacement for the tri-
fluoromethyl group. Since this position is easily accessible we
derivatized the functional group on the central pyrazole core. As
was the case for the synthesis of trifluoroethyl substituted pyrazole
18 we obtained only one regioisomer 53 in the condensation of
b
-enaminoketone 16 with hydrazine 52 (Scheme 13).12 Further
deprotection and coupling steps provided pyrazolebiscarboxamide
9 in good to excellent yields (48% overall, 5 steps).
Another approach to enhance water solubility is the in-
troduction of further heteroatoms. Thus we regarded acetals 10 and
11 as interesting targets, whose syntheses are based on diol 58,
which can be obtained in good yield from acid 21 (Scheme 14).
There is very little literature available about the synthesis of acetals
derived from trifluoroacetaldehyde and we initially followed
a route via activation of the diol as described by Smithers.18 With
tosylchloride we did not achieve monotosylation of the diol, and
instead the bicyclic product 59 was isolated (Scheme 14).
Scheme 8. Synthesis of amino carboxylic acids 38 and 40.
Since the linear route did not succeed, we examined a building
block approach with protected serinol 60 (Scheme 15).18,19 It was
possible to cyclize activated diol 61 with in situ generated fluoral
63, but the products obtained were supposed to be oxazolidines 65,
not acetals 64, as determined by NMR spectroscopy. It appears that
the reaction is not reproducible and not suitable for the synthesis of
those oxazolidines or acetals, such as 64.
Amide formation with carboxylic acid 38 led to amide 41, which
could also be crystallized (Fig. 6). The X-ray crystal structure of
amide 41 (pure trans isomer crystallized) proofed the configuration
of the cyclohexane moiety to be correct, which was determined by
comparison of 1H NMR spectra with the data for the lead com-
pound 3 before. Thus we continued to elucidate cis or trans con-
figurations via NMR analyses.
Other approaches for the synthesis of acetal 64 employing the
ethyl hemiacetal of trifluoroacetaldehyde, which is easier to handle,