Mild and efficient Lewis acid-promoted detritylation in the synthesis of N-hydroxy amides: A concise synthesis of (-)-cobactin T

Article abstract of DOI:10.1021/jo701411d

(Chemical Equation Presented) An efficient, high-yielding Lewis acid promoted deprotection of O-trityl hydroxylamine derivatives is described. A range of acid-labile protecting groups, such as N-Boc and O-TBS, were tolerated under these mild conditions. The present method is applicable to the synthesis of a broad range of hydroxylamine derivatives, including N-hydroxy amides (hydroxamic acids), N-hydroxy sulfonamides, and N-hydroxy ureas, which often exhibit significant biological activities. An application of this methodology for a concise synthesis of (-)-Cobactin T (18) is also demonstrated.

Full text of DOI:10.1021/jo701411d

SCHEME 1
Mild and Efficient Lewis Acid-Promoted
Detritylation in the Synthesis of N-Hydroxy
Amides: A Concise Synthesis of (-)-Cobactin T
Shyh-Ming Yang,* Bharat Lagu,† and Lawrence J. Wilson
TABLE 1. Deprotection of O-Trityl Hydroxamic Acid 1aa
Johnson & Johnson Pharmaceutical Research and DeVelopment,
L.L.C. 8 Clarke DriVe, Cranbury, New Jersey 08512
syang9@prdus.jnj.com
ReceiVed June 28, 2007
9a (yield, %)/1a (conversion, %)b
entry
reagent (equiv)
10 min
30 min
1 h
1
2
3
4
5
6
7
ZnBr2 (10)
ZnBr2 (5)
ZnCl2 (5)
Zn(OTf)2 (5)
MgBr2 (10)
MgBr2 (5)
BF3·OEt2 (2)
BF3·OEt2 (2)
Amberlyst 15
79/82
60/68
35/39
0/0
75/84
72/82
68/83
88/89
<10/12
60/87
72/82
38/45
2/3
91/94
80/93
67/84
94/98
14/21
55/89
77/85
39/45
3/5
84/96
83/96
66/85
-
8c
9d,e
25/30
a All reactions were performed on a 0.05 mmol scale in CH2Cl2 (1 mL)
at room temperature unless otherwise specified. b Percentages of 9a formed
and 1a that remained were determined by HPLC analysis relative to an
internal standard, 4-bromobiphenyl. c MeOH/CH2Cl2 (0.5 mL/0.5 mL) as
the solvent was employed. d Amberlyst 15 ion-exchange resin (60 mg) and
MeOH/CH2Cl2 (1 mL/1 mL) as the solvent were employed. e 9a (yield)/1a
(conversion) was 57/58, 80/81, and 85/87 at 3, 6, and 8 h, respectively.
An efficient, high-yielding Lewis acid promoted deprotection
of O-trityl hydroxylamine derivatives is described. A range
of acid-labile protecting groups, such as N-Boc and O-TBS,
were tolerated under these mild conditions. The present
method is applicable to the synthesis of a broad range of
hydroxylamine derivatives, including N-hydroxy amides
(hydroxamic acids), N-hydroxy sulfonamides, and N-hydroxy
ureas, which often exhibit significant biological activities.
An application of this methodology for a concise synthesis
of (-)-Cobactin T (18) is also demonstrated.
could potentially facilitate the removal of the trityl group through
the coordinated intermediate A (Scheme 1). Herein, we report
a mild and efficient Lewis acid-mediated deprotection of trityl
group in the synthesis of N-hydroxylamine derivatives.5
To test the feasibility of the detritylation process, compound
1a was initially selected and examined under various Lewis acid
conditions (Table 1). Attention was mainly focused on the use
of mild Lewis acid conditions. The yield of 9 and the percent
conversion of 1a were determined by HPLC analysis using
4-bromobiphenyl as an internal standard. Upon treatment of 1a
with 10 equiv of ZnBr2 in CH2Cl2, the detritylation took place
rapidly to reach over 80% conversion within 10 min at ambient
temperature.6 However, the yield of the desired product 9a
decreased to 55% after 1 h stirring, presumably due to
decomposition in the presence of excess Lewis acid (entry 1).
Reducing the amount of ZnBr2 (5 equiv) gave a comparable
result without diminution of 9a (entry 2). As previously reported,
both ZnCl2 and ZnBr2 exhibited equal efficiency in the
detritylation of trityl-protected hydroxyl compounds.6a By using
ZnCl2 in this O-trityl hydroxamate example, however, the
detritylation proceeded to a significantly lesser extent (entry 3).
Zn(OTf)2 was found to be totally ineffective (entry 4). Ad-
ditionally, MgBr2 has been employed to remove the trityl group
Hydroxamic acid is one of the reoccurring motifs in the
structure of metalloproteinase inhibitors, including matrix met-
alloproteinase (MMP),1 neutral endopeptidase (NEP),2 and
human histone deacetylase (HDAC). Although the trityl group
(Tr) has been widely used as a protecting group for the hydroxyl
group,3 few reports have appeared wherein a trityl group was
used as a protecting group in the synthesis of hydroxamic acids.4
In those cases, trifluoroacetic acid (TFA) was commonly used
as the deprotecting reagent. By taking advantage of potential
bidentate chelation of the hydroxamate unit, a Lewis acid (LA)
† Present address. Novartis Institute for Biomedical Research, Cambridge,
MA.
(1) Representative reviews, see: (a) Whittaker, M.; Floyd, C. D.; Brown,
P.; Gearing, A. J. H. Chem. ReV. 1999, 99, 2735. (b) Matter, H.; Schudok,
M. Curr. Opin. Drug DiscoVery DeV. 2004, 7, 513, and the references cited
therein.
(2) Walz, A. J.; Miller, M. J. Org. Lett. 2002, 4, 2047, and the references
cited therein.
(3) Green, T. W.; Wuts, P. G. M. Protecting Group in Organic Synthesis;
John Wiley & Sons: New York, 1999; pp 102-104.
(4) (a) Hanessian, S.; Moitessier, N.; Wilmouth, S. Tetrahedron 2000,
56, 7643. (b) Hanessian, S.; Moitessier, N.; Gauchet, C.; Vian, M. J. Med.
Chem. 2001, 44, 3066. (c) Wittich, S.; Scherf, H.; Xie, C.; Brosch, G.;
Loidl, P.; Gerha¨user, C.; Jung, M. J. Med. Chem. 2002, 45, 3296. (d)
Remiszewski, S. W.; Sambucetti, L. C.; Bair, K. W.; Bontempo, J.; Cesarz,
D. J. Med. Chem. 2003, 46, 4609.
(5) To the best of our knowledge, except TFA, other conditions for
detritylation in the synthesis of hydroxamates, N-hydroxyl sulfonamide, and
N-hydroxyamine derivatives have not been investigated yet.
(6) (a) Kohli, V.; Blo¨cker, H.; Ko¨ster, H. Tetrahedron Lett. 1980, 21,
2683. (b) Lampe, T. F.; Hofmann, H. M. R. Tetrahedron Lett. 1996, 37,
7695.
10.1021/jo701411d CCC: $37.00 © 2007 American Chemical Society
Published on Web 09/19/2007
J. Org. Chem. 2007, 72, 8123-8126
8123

TABLE 2. Lewis Acid Promoted Detritylation of O-Trityl Hydroxlamine Derivativesa
a All reactions were performed on a 1.0-2.0 mmol scale with MgBr2 (5.0 equiv) in CH2Cl2, or BF3‚OEt2 (2.0 equiv) in MeOH/CH2Cl2 (1/1) at room
temperature unless otherwise specified. b Isolated yield. c Amberlyst 15 ion-exchange resin (1.0 g/mmol) in MeOH/CH2Cl2 (1/1) was employed. d MeOH/
CH2Cl2 (1/4) was employed. e 5% TFA/CH2Cl2 was employed. f 1b (5%) or 2 (8%) was recovered. g MeOH/CHCl3 (1/4) was employed. h ZnBr2 (5.0 equiv)
in MeOH/CH2Cl2 (1/1) was employed. i ZnBr2 (10.0 equiv) in CH2Cl2 was employed.
from 1-O-Benzyl-3-O-tritylglycerol to give 1-O-benzylglycerol.7
In that report, the deprotection required 24 h in refluxing
benzene to reach completion or 6 days with 90% conversion at
room temperature. Upon treatment of 1a with MgBr2 (10 equiv),
surprisingly, the cleavage of trityl group went to completion
within 30 min at ambient temperature (entry 5). In fact, 5 equiv
of MgBr2 was sufficient to achieve high conversion with
excellent yield within 1 h (entry 6). These results suggest that
the bi-dentate coordination of the hydroxamate moiety with
Mg2+ accelerate the cleavage of the trityl group. The use of
BF3‚OEt2 (2 equiv) in dichloromethane was next examined.8
Under this condition, only a moderate yield of 9a was detected
(entry 7). By using methanol as the cosolvent, the detritylation
proceeded cleanly within 10 min and a higher yield was
observed (entry 8). Finally, the heterogeneous removal of the
trityl group using Amberlyst 15 ion-exchange resin proceeded
slowly, but smoothly, to reach completion at approximately 10
h (entry 9).9
With these practical detritylation conditions in hand, we
turned our attention to expanding the scope of this process. A
variety of chemotypes with different functionalities were studied
under optimized conditions, namely MgBr2 (5 equiv) in CH2-
Cl2 or BF3‚OEt2 (2.0 equiv) in CH2Cl2/MeOH at room tem-
perature (Table 2). The deprotection of 1a with MgBr2 or BF3‚
OEt2 proceeded as expected, completing within 40 min to
provide 9a in 86 and 89% isolated yield, respectively (entries
1-2). The use of Amberlyst 15 ion-exchange resin provided a
viable alternative to 9a (82%, entry 3). A range of functional
groups such as O-TBS, N-Cbz, and especially N-Boc were well
tolerated under MgBr2 and BF3‚OEt2 conditions with good
(7) Haraldsson, G. G.; Stefansson, T.; Snorrason, H. Acta Chim. Scand.
1998, 52, 824.
(8) Cabaret, D.; Wakselman, M. Can. J. Chem. 1990, 68, 2253.
(9) Malanga, C. Chem. Ind. (London) 1987, 856.
8124 J. Org. Chem., Vol. 72, No. 21, 2007

CHART 1. Representative Nature Products Containing
Hydroxamic Acid Unit
SCHEME 2. Synthesis of (-)-Cobactin T (18)a
a Reagents and Conditions: (a) EDC, HOBt, N-methylmorpholine,
TrONH2, CH2Cl2, rt, 2 h, 89%; (b) Allyl methyl carbonate, Pd(PPh3)4 (0.02
equiv), MeCN, 45 min, 86%; (c) Grubbs II catalyst (0.05 equiv), CH2Cl2,
40 °C, 3 h, 85%; (d) H2 (15 psi), 5 wt % Pt/C (0.05 equiv of Pt), EtOAc,
rt, 1.5 h; (e) Et3N, MeCN, rt, 12 h, 68% (two steps); (f) (R)-3-hydroxybutyric
acid, EDC, HOAt, N-methylmorpholine, CHCl3, rt, 1 h, 70%; (g) BF3‚OEt2,
CH2Cl2/MeOH (4/1, 0.1 M), rt, 5 min, 45%; or (h) Amberlyst 15
ion-exchange resin, CH2Cl2/MeOH (1/4), rt, 2 h, 80%.
isolated yields (66-81%, entries 4-5 and 7-9).10 In a
comparison between substrates 1b and 2, both cyclic and acyclic
trityl-protected hydroxamates exhibited similar reactivity. The
success of selective cleavage of the trityl group in the presence
of N-Boc demonstrates the efficiency of this coordinative
detritylation process. Not surprisingly, N-Boc did not survive
under trifluoroacetic acid (TFA) condition. Even under 5% TFA/
CH2Cl2 solution, both 1b and 2 gave desired product in low
yields (ca. 28-30%, entries 6 and 10) with low recovery of
starting material. These results may be attributed to the
simultaneous loss of the Boc group during detritylation process.
The methodology can be further expanded to incorporate other
N-hydroxylamine derivatives, such as N-hydroxy carbamate
(12), N-hydroxy ureas (13-14), and N-hydroxy sulfonamide
(15).11 The efficiency of detritylation seems to depend on the
nature of hydroxamate unit and Lewis acid employed. For
example, the reaction of O-trityl N-hydroxy carbamate (4)
reacted more rapidly with BF3‚OEt2 than with MgBr2, although
both afforded 12 in excellent yields (entries 12-13). In the case
of an urea-sulfonamide-hybrid hydroxamate (5), the removal
of trityl group occurred with equal efficiencies and yields under
both Lewis acid conditions (entries 14-15). The detritylation
of 611a and 7 took place at a relatively slower rate than
compounds 1-5 as described above (entry 16-18). In addition,
MgBr2 did not cleave the trityl group from O-trityl N-hydroxy
sulfonamide 7. The trityl group of 7 could be successfully
removed, but to a lesser extent under ZnBr2 (5.0 equiv)
conditions using methanol as a cosolvent (entry 19). Finally,
an oxime-containing hydroxylamine derivative 8 was examined
under several conditions described above. This oxime moiety
was extremely sensitive to acidic conditions such as TFA, HCl,
and Amberlyst 15 ion-exchange resin. It was also unstable under
a diverse set of Lewis acids including AlCl3, MgBr2, and BF3‚
OEt2, presumably due to rapid hydrolysis of the oxime unit
under these conditions. Gratifyingly, the desired product 16 was
obtained in 82% isolated yield (entry 20) by using 10 equiv of
ZnBr2 in CH2Cl2 within 10 min. In most cases, the desired
product in an analytically pure form could be easily obtained
by recrystallization or precipitation from the crude mixture.
To further demonstrate the versatility and effectiveness of
this process, we applied the detritylation methodology to a novel
synthesis of (-)-Cobactin T (18). Mycobactin T (17) and
Cobactin T are the siderphore growth promoters isolated from
mycobacteria. Both macromolecules contain a novel cyclic
hydroxamate unit (Chart 1).12 The importance of mycobactin
family in the aspect of drug resistance in strains of tuberculosis
has stimulated a significant level of effort in synthesis of diverse
analogs relative to natural mycobactins. Among them, Cobactin
T, one of the major fragments of Mycobactin T, had been
previously synthesized.12,13 These reports feature intramolecular
lactam formation or Mitsunobu type cyclization to construct the
novel N-hydroxy lactam ring. In our retrosynthetic analysis, we
envisioned that the ring closing metathesis (RCM) would
provide the key lactam ring (B) in a highly efficient way. For
further installation of the amide side chain, the Fmoc was chosen
as the protecting group for the R-amino moiety since it could
be easily removed under basic conditions without affecting the
acid-labile trityl group.
The synthesis of 18 began with the commercially available
N-Fmoc L-allylglycine 19 (Scheme 2). The coupling of 19 with
TrONH2 using 1-ethyl-3-[3-(dimethylamino)propyl]carbodiim-
ide hydrochloride (EDC) and 1-hydroxybenzotriazole (HOBt)
as the coupling reagents and N-methylmorpholine as the base
gave 20 in 89% yield. Under this condition, only a trace amount
of de-Fmoc product was observed by HPLC analysis instead
of ca. 10% of de-Fmoc product when 4-dimethylaminopyridine
(10) N-Boc can be removed under ZnBr2 conditions, see: (a) Kaul, R.;
Brouillette, Y.; Sajjadi, Z.; Hansford, K. A.; Lubell, W. D. J. Org. Chem.
2004, 69, 6131. (b) Nigam, S. C.; Mann, A.; Taddei, M.; Wermuth, C.-G.
Synth. Commun. 1989, 19, 3139. One Boc group of N(Boc)2 can be removed
using MgBr2, see: (c) Burkhart, F.; Hoffmann, M.; Kessler, H. Angew.
Chem., Int. Ed. 1997, 36, 1191. Deprotection of N-Boc using BF3‚OEt2
has been reported, see: (d) Evans, E. F.; Lewis, N. J.; Kapfer, I.; Macdonald,
G.; Taylor, R. J. K. Synth. Commun. 1997, 27, 1819.
(11) Compound 6 has been reported as a potent acyl-CoA cholesterol
acyltransferase (ACAT) inhibitor, see (a) Trivedi, B. K.; Holmes, A.;
Purchase, T. S.; Essenburg, A. D.; Hamelehle, K. L.; Krause, B. R.; Hes,
M. S.; Stanfield, R. L. Bioorg. Med. Chem. Lett. 1995, 5, 2229. N-Hydroxy
sulfonamides, such as 15, are potent carbonic anhydrase (CA) inhibitors,
see: (b) Scozzafava, A.; Supuran, C. T. J. Med. Chem. 2000, 43, 3677. (c)
Mincione, F.; Menabuoni, L.; Briganti, F.; Mincione, G.; Scozzafava, A.;
Supuran, C. T. J. Enzyme Inhib. 1998, 13, 267.
(12) (a) Hu, J.; Miller, M. J. J. Am. Chem. Soc. 1997, 119, 3462. (b)
Xu, Y.; Miller, M. J. J. Org. Chem. 1998, 63, 4314.
(13) (a) Maurer, P. J.; Miller, M. J. J. Am. Chem. Soc. 1983, 105, 240.
(b) Maurer, P. J.; Miller, M. J. J. Org. Chem. 1981, 46, 2835. (c) Hu, J.;
Miller, M. J. Tetrahedron Lett. 1995, 36, 6379.
(14) The N-allylation of O-benzyl hydroxamate moiety was predominated
when the reactive allylating reagent, such as allyl bromide, was employed,
see: Johnson, J. E.; Springfield, J. R.; Hwang, J. S.; Hayes, L. J.;
Cunningham, W. C.; McClaugherty, D. L. J. Org. Chem. 1971, 36, 284.
J. Org. Chem, Vol. 72, No. 21, 2007 8125

(DMAP) or triethylamine was used as the base. Attempts made
to introduce the allyl group onto the nitrogen atom selectively
under basic allylation conditions proved to be problematic.
Several combinations of bases (K2CO3 or Cs2CO3) and solvents
(THF, MeCN, or acetone) were examined with allyl bromide
or allyl iodide as the alkylating partner.14 With all conditions
examined, the O-allylated product was the predominant product.
Exclusive N-allylation of 20 was finally achieved under the
palladium-catalyzed allylic substitution condition. By employing
2 mol % of Pd(PPh3)4 as the catalyst and 2 equiv of allyl methyl
carbonate as the allylating reagent, the reaction went to
completion within 1 h with excellent isolated yield (86%).15
The ring closing metathesis of 21 using Grubbs II ruthenium
complex (5 mol %) in refluxing CH2Cl2 gave the key lactam
intermediate 22 in 85% yield (ca. 66% yield from 19).16,17
Initial attempt to selectively hydrogenate the double bond
without cleavage of trityl group by using [Ir(cod)py(PCy3)]PF6
complex (Crabtree catalyst, 5 mol %) with H2 (1atm) was
unsuccessful.18 The homogeneous iridium complex seemed to
serve as a Lewis acid, which caused ca. 30% detritylation based
on LC-MS analysis. To avoid the potential coordination, 5 wt
% Pt/C (5 mol % of Pt loading) was used as the catalyst under
1 atm of H2 to afford 23 cleanly with no detectable amount of
detritylation product.19 The deprotection of N-Fmoc group under
basic conditions, triethylamine in acetonitrile at rt, afforded 68%
of 24 from 22. The amide formation of 24 with (R)-3-
hydroxybutyric acid using EDC and HOAt (1-hydroxy-7-
azabenzotriazole) as the coupling reagents gave O-trityl cobactin
T (25) in 70% isolated yield. Unfortunately, MgBr2 was
ineffective for the removal of the trityl group in 25. Although
BF3‚OEt2 did provide clean cleavage of trityl group with the
same efficiency, relatively low isolated yield (45%) was obtained
presumably due to the loss of product during extraction. To
overcome the potential issue of aqueous solubility of the product,
Amberlyst 15 ion-exchange resin in CH2Cl2/MeOH was em-
ployed. After stirring for 2 h, the product was filtered and then
recrystallized from ethyl acetate and hexane to furnish (-)-
In conclusion, the Lewis acid mediated detritylation in the
synthesis of N-hydroxyl amine derivatives, especially the
hydroxamates, has been investigated. Although the coordination
nature of O-trityl hydroxamate unit plays an important role in
reaction efficiency, MgBr2, ZnBr2, and BF3‚OEt2 were found
to be superior for the removal of trityl group. In addition, the
Amberlyst 15 ion-exchange resin also provided a viable
alternative in terms of the ease for purification. These mild and
highly efficient detritylative conditions clearly demonstrate the
potential utility in the synthesis of metalloproteinase inhibitors
containing hydroxamate moiety. The synthetic application using
trityl as the protecting group in the synthesis of (-)-Cobactin
T was achieved in seven steps with 25% overall yield. In
addition, the N-hydroxy lactam ring was rapidly assembled by
ring closing metathesis which provides the opportunity for the
synthesis of dehydro-mycobactin analogs that maybe useful in
studies of drug resistance in strains of mycobacteria.
Experimental Section
Representative procedures for detritylation using MgBr2 or BF3‚
OEt2 are described below.
Preparation of N-Hydroxy 2-(9-Fluorenylmethoxycarbonyl)-
amino-4-pentenoamide (9a). To a mixture of 1a (1.0 mmol, 594
mg, 1.0 equiv) and MgBr2 (5.0 mmol, 920 mg, 5.0 equiv) was added
CH2Cl2 (10 mL) at room temperature. The yellow suspension was
stirred at rt for 40 min and was then poured into EtOAc/H2O (50
mL/50 mL). The organic layer was washed with brine (50 mL),
dried (Na2SO4), and filtered. After removal of solvent, acetone (10
mL) and hexane (50 mL) were added sequentially to the crude
product. The resulting solid was filtered and was washed with Et2O/
hexane (1/1, 50 mL). The product was dried to give 303 mg of 9a
(86%) as a white solid.
Preparation of N-Hydroxy (9-Fluorenyl)methyl carbamate
(12). To a mixture of 4 (2.0 mmol, 994 mg, 1.0 equiv) in CHCl3/
MeOH (16 mL/4 mL) was added BF3‚OEt2 (4.0 mmol, 0.5 mL,
2.0 equiv) at room temperature. The mixture was stirred at rt for
45 min and was then poured into EtOAc/H2O (100 mL/100 mL).
The organic layer was washed with brine (100 mL), dried (Na2-
SO4), and filtered. After removal of solvent, CH2Cl2 (10 mL) and
hexane (30 mL) were added sequentially to the crude product. The
resulting solid was filtered and was washed with Et2O/hexane (2/
3, 20 mL). The product was dried to give 474 mg of 12 (93%) as
a white solid.
Cobactin T (18) in 80% yield. The spectroscopic data of the
20
synthetic sample were consistent in all respects (mp, [R]D
,
1H NMR, and 13C NMR) to those reported by Miller’s
group.12a,13a
(15) (a) Miyabe, H.; Yoshida, K.; Yamauchi, M.; Takemoto, Y. J. Org.
Chem. 2005, 70, 2148. (b) Miyabe, H.; Yoshida, K.; Matsumura, A.;
Yamauchi, M.; Takemoto, Y. Synlett 2003, 567.
(16) Benzylidene[1,3-bis-(2,4,6-trimethylphenyl)-2-imidazolidinylidene]-
dichloro(tricyclohexylphosphine)ruthenium (Grubbs II catalyst) was pur-
chased from Aldrich.
Acknowledgment. S.-M.Y. thanks Dr. Maud Urbanski for
helpful discussion during the preparation of this manuscript.
(17) For some recent reviews, see: (a) Grubbs, R. H.; Chang, S.
Tetrahedron 1998, 54, 4413. (b) Fu¨rstner, A., Ed. Alkene Metathesis in
Organic Synthesis. Springer: Berlin, 1998. (c) Trnka, T. M.; Grubbs, R.
H. Acc. Chem. Res. 2001, 34, 18.
(18) Schultz, A. G.; McCloskey, P. J. J. Org. Chem. 1985, 50, 5905.
(19) (a) Rishel, M. J.; Hecht, S. M. Org. Lett. 2001, 3, 2867. Trityl group
can be removed under a condition of H2 (2.5 atm) and 5% Pd/C (ca. 10
mol% Pd loading), see: (b) Mirrington, R. N.; Schmalzl, K. J. J. Org. Chem.
1972, 37, 2877.
Supporting Information Available: Detailed experimental
procedure and characterization data for all O-trityl protected
precursors (1-8), detritylation products (9-16), intermediates (20-
25), and (-)-Cobactin T (18) are provided. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO701411D
8126 J. Org. Chem., Vol. 72, No. 21, 2007