A new and practical boronic acid catalyzed intramolecular cyclodehydration of ureas for the synthesis of LFA-1 antagonists

Article abstract of DOI:10.1016/j.tetlet.2011.03.007

A streamlined and practical approach to 1H-imidazo[1,2-α]imidazol-2- one LFA-1 antagonist (1) that features as the key step a novel intramolecular boronic acid catalyzed cyclocondensation of a urea and an amide is reported.

Full text of DOI:10.1016/j.tetlet.2011.03.007

Tetrahedron Letters 52 (2011) 2465–2467
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
A new and practical boronic acid catalyzed intramolecular cyclodehydration
of ureas for the synthesis of LFA-1 antagonists
⇑
Rogelio P. Frutos , Thomas Tampone, Jason A. Mulder, Yibo Xu, Diana Reeves, Xiao-Jun Wang, Li Zhang,
Dhileepkumar Krishnamurthy, Chris H. Senanayake
Chemical Development US, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Rd./PO Box 368, Ridgefield, CT 06877-0368, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A streamlined and practical approach to 1H-imidazo[1,2-a]imidazol-2-one LFA-1 antagonist (1) that fea-
tures as the key step a novel intramolecular boronic acid catalyzed cyclocondensation of a urea and an
amide is reported.
Received 13 January 2011
Revised 15 February 2011
Accepted 1 March 2011
Available online 6 March 2011
Ó 2011 Elsevier Ltd. All rights reserved.
of triphenylphosphine oxide. In order to facilitate the large scale
synthesis of LFA-1 antagonists like 1, a more streamlined approach
to 2 that would be more suitable for scale-up and would eliminate
the formation of large quantities of copper salts or triphenylphos-
phine oxide was needed. Herein we describe a new approach to
intermediate 2 that features a novel intramolecular boronic acid
catalyzed cyclocondensation that eliminates the formation of
undesired by-products. It was previously observed that the cyc-
lodehydration of 44 (Scheme 1) or 115 (Eq. 5) to give 5 or 12,
respectively, fails with well known dehydration reagents such as
Burgess reagent, p-TsCl/Et3N or Ph3PꢀBr2/Et3N. The cyclodehydra-
tion was successful only when using CCl4ꢀPPh3/Et3N or Ph3PꢀCl2
following a modification of the protocol reported by Appel and
co-workers.6 Due to the environmental, health and safety concerns
associated with the use of CCl4 and the difficulty in removing the
triphenylphosphine oxide by-product, we embarked upon a study
aimed at developing alternatives to the above reaction conditions.
Our results are described herein.
R1
R1
O
O
Cl
Cl
O O
R2
S
N
N
N
N
ð1Þ
N
R3
N
N
Cl
2
Cl
1
R1 = Br, CN, OCF3, aryl or heterocycle
R2 = H or alkyl
R3 = H or alkyl
The synthesis of 1H-imidazo[1,2-a]imidazol-2-one LFA-1 inhibi-
tors such as 1 and their evaluation as potential clinical candidates
for the treatment of inflammatory and immune disorders was re-
cently reported.1–5 Synthetic approaches toward 1 make use of
intermediates 2, for which several synthetic routes have been
developed (Eq. 1). The earliest approach to 2 featured the conden-
sation of an aza-phosphonium ylide with a thiohydantoin as the
key step,1 requiring the handling of a potentially unstable azide re-
agent. In addition, the use of the aza-phosphonium ylide resulted
in the quantitative formation of triphenylphosphine oxide, which
was difficult to remove. Another approach to 2 utilized a Cu-
promoted intramolecular cyclocondensation of a thiourea and an
amide.3 Although this approach could be scaled-up to make
multi-kilogram quantities of 2, it generated large quantities of cop-
per salts that had to be removed by laborious filtration. Similarly,
another approach to 2 (Scheme 1) utilized CCl4ꢀPPh3/Et3N or
Ph3PꢀCl2 to promote the intramolecular cyclocondensation of a
urea and an amide,2,4 also resulting in the quantitative formation
R1
R1
References 4 and 9
O
O
ð2Þ
Cl
Cl
H2N
NH
N
N
O
CH3O
CH3O
H
H
NH
4
3
Cl
Cl
We had hypothesized that the formation of 2 upon treatment of 4
(Scheme 1) with CCl4/PPh3 may occur by the formation of a transient
carbodiimide intermediate.5 Accordingly, literature reports on the
synthesis of carbodiimides from ureas using reagents normally
employed for amide and peptide synthesis, such as Mukaiyama’s
reagent (2-chloro-1-methylpyridinium iodide),7 prompted us to
investigate the use of these reagents. We investigated the reaction
⇑
Corresponding author.
E-mail address: rfrutos@rdg.boehringer-ingelheim.com (R.P. Frutos).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.03.007

2466
R. P. Frutos et al. / Tetrahedron Letters 52 (2011) 2465–2467
R1
CN
F3C
CF3
6
CN
R1
references
B(OH)2
CCl4/PPh3
1-5
2
or
O
O
O
10-20 mol %
X
Ph3PCl2
O
Cl
ð3Þ
ð4Þ
Cl
Cl
Cl
NH
N
H
O
N
N
NH
N
O
N
N
(MeO)3CH,
H
NH
Toluene, reflux
12-24 h
NH
HN
MeO
NH
OMe
Cl
Cl
Cl
Cl
7
8
OMe
MeO
OMe
MeO
CN
F3C
6
CN
4
5
B(OH)2
Scheme 1. Synthesis of 5 from 4.
CF3
O
O
10-20 mol %
Cl
Cl
O
X
N
N
N
H
N
N
of 4 with Mukaiyama’s reagent, EDC, trichloroisocyanuric acid and
propanephosphonic acid cyclic anhydride, but use of these reagents
gave little or no cyclic 5 or 2, resulting in the recovery of unreacted
starting material or unidentifiable decomposition products. We
continued our screening with less common reagents and were
able to obtain intermediate 2 directly from 4, without detecting 5,
by using an unprecedented boronic acid mediated cyclodehydration
procedure similar to that described by Yamamoto and co-workers
for the synthesis of amides.8 An assortment of reaction conditions
were screened in order to optimize the yield of 2 and reduce the
amount of undesired 9. Hence, different solvents such as toluene,
xylenes, isobutyl acetate and butyronitrile were evaluated for the
reaction as well as several boronic acids such as boronic acid,
3,5-difluorophenylboronic acid, 3-nitrophenylboronic acid, 3-trif-
luoromethylboronic acid and 3,5-bis(trifluoromethyl)boronic acid
(6). Moreover we also investigated the effect of additives such as
triethylamine, K2CO3, molecular sieves, methanol, MgSO4, and trim-
ethylorthoformate on the reaction, but only the addition of trimeth-
ylorthoformate was beneficial. Accordingly, best results were found
to be with 6 as the catalyst in 10–20 mol % relative to the starting
urea 2, toluene as the solvent and trimethylorthoformate in an
amount that did not exceed the amount of the boronic acid catalyst.
The above-mentioned transformation was applied to a number of
analogs 4a–d prepared according to the previously described proto-
cols9 from known amino-amides (Eq. 2). Accordingly, submission of
4a–d4,10 to the reaction conditions,11 which consisted of treatment
with a catalytic amount of the electron deficient arylboronic acid 6
with the simultaneous removal of water by azeotropic distillation,
provided 2a–d4,12 in good yields (Table 1).
(MeO)3CH,
Toluene, reflux
12-24 h
N
Cl
Cl
2a
9
The main side products detected from the above transformations
were cyclic ureas like 9 (Eq. 4) which formed in approximately
5–10%. The addition of a catalytic amount of trimethyl orthofor-
mate, presumably acting as a water scavenger, was crucial to min-
imize the formation of 9; without the orthoformate, up to 30% of 9
could be observed in some cases. Nevertheless, urea 9 could be
easily removed by silica gel chromatography or by crystallization
of 2. To further improve the reaction, we evaluated the acetal
group itself and found that dimethylacetal gave better results than
diethylacetal or dioxolane. This was somewhat surprising since we
expected that an acetal more stable to acid-promoted hydrolysis
would be less prone to the formation of 9 and give a higher yield
of 2.
F3C
CN
CN
6
B(OH)2
CF3
O
O
ð5Þ
10-20 mol %
Cl
Cl
N
N
X
NH
N
O
H
(MeO)3CH,
Toluene, reflux
12-24 h
MeO2C
NH
12
HN
Cl
Cl
CO2Me
11
We considered possible that the cyclization proceeded in a step-
wise manner through intermediates like 5 (Scheme 1) or perhaps
even 9 (Eq. 4). However, 7 failed to give 8 (Eq. 3) and urea 9 failed
to give 2a (Eq. 4) when submitted to the boronic acid promoted
cyclization conditions. Similarly, urea 11 bearing an ester group
instead of the dimethyl acetal failed to provide 12 (Eq. 5) when
submitted to the reaction conditions. Interestingly, 5, 8 and 12
could be obtained upon treatment of their corresponding urea
precursors, 4, 7 and 11 with CCl4/PPh3. Furthermore, when urea
4 was treated with protic acids, the monocyclic urea 9 was the
only observable product, suggesting the boronic acid promoted
reaction is unlikely to be a simple acid catalyzed transformation.
Thus, it appears that neither 5 nor 9 are intermediates for the
reaction, and that the choice of acetal group and boronic acid
are both crucial for the desired reaction to proceed. Although
no rigorous mechanistic studies have been carried out and no
reaction intermediates have been observed, it is possible that
the reaction may proceed through transient intermediates in
which the acetal group and the boronic acid strongly interact
through a pathway resembling the one depicted in Scheme 2. This
is consistent with the diminished yields we observed when we
switched from the dimethyl acetal to the diethyl acetal or dioxo-
lane. The beneficial role of the orthoformate is likely to come
Table 1
Synthesis of intermediates 2a–d
R1
R1
F3C
CF3
6
B(OH)2
O
O
Cl 10-20 mol %
Cl
NH
N
O
N
N
(MeO)3CH,
H
toluene, reflux
12-24 h
(75-90%)
HN
MeO
N
Cl
Cl
OMe
4a-d
2 a-d
Entry
Urea
Product
R1
6
Yield (%)
1
2
3
4
4a
4b
4c
4d
2a
2b
2c
2d
CN
OCF3
Br
0.2 equiv
0.1 equiv
0.2 equiv
0.2 equiv
84
80
95
75
5-pyrimidine

R. P. Frutos et al. / Tetrahedron Letters 52 (2011) 2465–2467
2467
ArIIB(OH)2
+
Acknowledgment
4
The authors wish to acknowledge Professor Peter Wipf for valu-
able suggestions and insightful comments.
MeOH
ArI
N
ArI
N
ArI
O
O
O
References and notes
N
Ar
H
N
N
Ar
NH
H
N
Ar
H
H
H
N
O
O
1. Wu, J.-P.; Emeigh, J.; Gao, A. D.; Goldberg, D. R.; Kuzmich, D.; Miao, C.; Potocki, I.;
Qian, K. C.; Sorcek, R. J.; Jeanfavre, D. D.; Kishimoto, K.; Mainolfi, E. A.; Nabozny,
G., Jr.; Peng, C.; Reilly, P.; Rothlein, R.; Sellati, R. H.; Woska, J. R., Jr.; Chen, S.;
Gunn, J. A.; O’Brien, D.; Norris, S. H.; Kelly, T. A. J. Med. Chem. 2004, 47, 5356.
2. Frutos, R. P.; Eriksson, M.; Wang, X.-J.; Byrne, D.; Varsolona, R.; Johnson, M.;
Nummy, L.; Krishnamurthy, D.; Senanayake, C. H. Org. Process Res. Dev. 2005, 9,
137.
3. Wang, X-j.; Zhang, L.; Xu, Y.; Krishnamurthy, D.; Varsolona, R.; Nummy, L.;
Shen, S.; Frutos, R. P.; Byrne, D.; Chung, J. C.; Farina, V.; Senanayake, C. H.
Tetrahedron Lett. 2005, 46, 273.
4. X.-J. Wang, T. Wirth, T. Nicola, L. Zhang, R. P. Frutos, Y. Xu, D. Krishnamurihy, L.
J. Nummy, R. J. Varsolona, J. Kroeber, C. H. Senanayake. Patent number WO
2006/014828 A1, 2006.
5. Frutos, R. P.; Johnson, M. Tetrahedron Lett. 2003, 44, 6509.
6. Appel, R.; Kleinstuck, R.; Ziehn, K. Chem. Ber. 1971, 104, 1335.
7. Siu, J.; Baxendale, I. R.; Lewthwaite, R. A.; Ley, S. V. Org. Biomol. Chem. 2005, 3,
3140.
N
O
H
MeOH
O
II
O
O
MeO
B
Ar
13
ArI
B
O
ArII
15
H
O
H
14
B
O
II
Ar
ArI
O
O
2
ArIIB(OH)2
N
N
Ar
N
N Ar
H
O
O
NH
HO
B
H
O
H2O
N
ArII
B
O
II
Ar
16
17
8. (a) Ishihara, K.; Kondo, S.; Yamamoto, H. Synlett 2001, 1371; (b) Ishihara, K.;
Ohara, S.; Yamamoto, H. J. Org. Chem. 1996, 61, 4196.
Scheme 2. Possible mechanism for the cyclodehydration.
9. Frutos, R. P.; Stehle, S.; Nummy, L.; Yee, N. K. Tetrahedron: Asymmetry 2001, 12,
101.
10. Compound 4a: 1H NMR (400 MHz, CDCl3) d 1.27 (s, 3H); 3.00–3.60 (m, 4H), 3.34
(s, 6H), 4.39 (dd, J = 5.6, 5.6 Hz, 1H), 6.12 (dd, J = 6.0, 6.0 Hz, 1H), 6.23 (s, 1H),
7.28 (dd, J = 2.0, 2.0 Hz, 1H), 7.33–7.37 (m, 2H), 7.77–7.81 (m, 2H), 7.81 (d,
J = 2.0, 2.0 Hz, 2H) 9.98 (s, 1H).
11. Typical experimental procedure: Urea 4 (0.96 mmol) was charged to a 50 mL 1-
neck round bottom equipped with a Dean–Stark trap, N2 line, and reflux
from trapping water, preventing the reaction of water with a
transient intermediate like 14, and the formation of urea 9, which
itself was ruled out as a possible intermediate for the transforma-
tion (Eq. 4). In conclusion, a new, safe and efficient synthesis for
the core of 1H-imidazo[1,2-a]imidazol-2-one LFA-1 inhibitors
condenser. Boronic acid
6
(0.19 mmol), toluene (15 mL), and
was developed by means of an unprecedented boronic acid cyc-
lodehydration. The new protocol provides the desired products
cleanly without the isolation issues caused by reagents used in
earlier approaches and it was implemented on a >1 kg scale.
The exact sequence of events involved in this transformation is
not well defined, but control experiments shed light on possible
mechanistic pathways.
trimethylorthoformate (0.19 mmol) were charged and the mixture was
placed in a preheated oil bath (135 °C) and refluxed for 16 h. Afterward, the
solution was cooled and washed with dilute NaHCO3 solution. The organic
layer was collected and concentrated under reduced pressure (rotorvap) to
afford the desired crude product as an amber oil in 80–95% assay yield.
12. Compound 2a: 1H NMR (400 MHz, CDCl3) d 1.80 (s, 3H), 3.26 (d, J = 13.6 Hz, 1H),
3.39 (d, J = 13.6 Hz), 6.94 (d, J = 1.2 Hz, 1H), 6.97 (d, J = 1.2 Hz, 1H), 7.01 (d,
J = 8.4 Hz, 2H), 7.26 (t, J = 1.8 Hz, 1H), 7.46 (d, J = 8.4 Hz, 2H), 7.75 (d, J = 1.8 Hz,
2H).