Addressing mechanistic issues in the coupling of isonitriles and carboxylic acids: Potential routes to peptidic constructs

Article abstract of DOI:10.1021/ja804709s

Mechanistic issues surrounding the two component coupling (2CC) reaction of carboxylic acids with isonitriles have been investigated. Experimental details suggest the formimidate carboxylate mixed anhydride intermediate exists in both interdictable and noninterdictable form. Furthermore, the 2CC reaction has been applied to the synthesis of a tripeptide featuring two formyl functional groups. Copyright

Full text of DOI:10.1021/ja804709s

Published on Web 09/11/2008
Addressing Mechanistic Issues in the Coupling of Isonitriles and Carboxylic
Acids: Potential Routes to Peptidic Constructs
Xuechen Li,† Yu Yuan,† Cindy Kan,† and Samuel J. Danishefsky*,†,‡
Laboratory for Bioorganic Chemistry, Sloan-Kettering Institute for Cancer Research, 1275 York AVenue,
New York, New York 10065, and Department of Chemistry, Columbia UniVersity, HaVemeyer Hall, 3000 Broadway,
New York, New York 10027
Received June 27, 2008; E-mail: s-danishefsky@ski.mskcc.org
In previous papers, we have described some of the chemistry
which follows from the reaction of carboxylic acids (1) and
isonitriles (2) under suitable thermolytic means (Figure 1).1 To
differentiate these studies from earlier aspects of isonitrile chem-
istry,2 we suggest the term “two-component coupling” (2CC) for
our program. In the absence of external acyl acceptors, quite
encouraging yields (70-90%) of the synthetically versatile N-
formylamides of the type 4 are obtained. As was shown, such novel
amidic (or, more precisely, imidic) structures provide new op-
portunities for the synthesis of modified peptide-like structures. In
the presence of nucleophilic trapping agents, the rearrangement of
3f4 competes with the formation of interdiction products. For
instance, ester-type 5 is generated from intervention by an external
alcohol-type nucleophile.1b In the presence of amine-based acyl
acceptors, simple amides can be produced (see 6).1a The formation
of these products can be mechanistically accommodated via the
intermediacy of the formimidate carboxylate mixed anhydride
(FCMA, cf. 3). We have shown that, notwithstanding several claims
to the contrary,3,4 the isolation of such a FCMA under usual reaction
conditions has not been accomplished.5
Figure 2
view, the distribution between 4 vs 5 (or 6) really reflects, to a
great extent, the distribution in the formation of FCMA stereoiso-
mers 3a anti (a) and 3s syn (s). Were equilibrium between 3a and
3s to be more rapid than either 1,3-OfN acyl transfer or interdiction
by nucleophiles, then, the two pathways would, in effect, still be
competing for control of product distribution.
The 2CC reaction of acids and isonitriles carries with it significant
potentialities for the synthesis of biologically active amidic or
peptidic products, including patterns not readily accessible by
current methods.8 With new insights into the finer workings of the
underlying chemistry of the 2CC method, could well come new
Figure 1
opportunities to enlarge upon its growing menu of applications.
The experiments delineated herein particularly focus on gaining
better insight into the feasibility of 2CC processes in the presence
of potential acyl acceptors. The findings described below chart a
course for future discovery.
Two formulations present themselves to reach products 4, 5, and
6. In one view, a high energy FCMA is produced, which can
undergo the 1,3-OfN acyl transfer leading to 4 or nucleophilic
trapping leading to 5 or 6. Alternatively, the apparent competition
We first interrogated a 2CC system by evaluating an alternative
between interdiction and N-formylamide formation may really
reflect the consequence of concurrent formation of 3a and 3s.6,7
pathway, which need not actually involve 1,3-OfN acyl rear-
rangement within 3 for producing 4. In this proposal, interdiction
One of the stereoisomers, 3a, is perhaps noninterdictable. It is, in
(perhaps solely on the nonrearranging 3s form) occurs by nucleo-
effect, “committed” to undergoing very rapid thermolytically
induced 1,3-OfN acyl transfer leading to 4. By contrast, production
philic attack of acid 1 at the carboxyl carbonyl group of FCMA 3
(Figure 2). This step would coproduce formamide 7 and symmetric
of the nonrearranging FCMA (3s) in the presence of a nucleophilic
anhydride 8.9 N-acylation of 7 with 8 would afford, directly,
acceptor would lead to 5 or 6 (as well as formamide 7). Of course,
in the absence of acyl acceptor XH, the noncompetent acyl transfer
FCMA form, 3s, would revert to 3a, which goes on to 4. In this
N-formylamide product 4. Of course, O-acylation of 7 with 8 would
be an “invisible option” in that it would simply result in the
reappearance of 3 and, thence, 4.
We studied the possible formamide (cf. 7)-symmetrical anhy-
dride (cf. 8) pathway, in the context of the high yielding reaction
† Sloan-Kettering Institute for Cancer Research.
‡ Columbia University.
9
10.1021/ja804709s CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 13225–13227 13225

C O M M U N I C A T I O N S
a
Table 1
entry
MeOH
22a
24
25
1
2
1.0 equiv
5.0 equiv
30%
14%
4%
11%
38%
41%
a The distribution of 22, 24, 25 was determined by 1H NMR of the
crude sample.
Figure 3. The distribution of 22, 23 was determined by LC-MS of the
crude sample.
committed to a particular program (vide supra, see pathways in
Figure 1). To get at this problem, we studied the consequences of
administering 5 equiv of methanol (rather than 1) in the 2CC
reaction of acid 15 and isonitrile 16. Remarkably, the yield of
methyl ester did not increase significantly, even following a sharp
upgrade in the concentration of the trapping agent (methanol). We
take this experiment to suggest that already at the level of a single
equivalent of methanol, there exists an intermediate whose thermally
induced methanolysis reaction is sufficiently fast as to provide for
highly efficient trapping. Thus, administration of additional equiva-
lents of methanol does not materially change the amount of methyl
ester which is generated. What does change (see Table 1) is the
ratio of N-formyl compound 22 to amide 24. This result reflects
the role of excess methanol in increasing the level of methanolytic
transformation of 16 to the corresponding amine 26. The latter goes
on to produce amide 24, as previously shown. However, the
important message is that a large increase of the trapping agent
has not affected to any significant extent the level of interdiction.
It is interesting to note that our studies1 have covered a varied
range of acyl acceptors such as formamide 17 (or 19), methanol,
and amines. Yet, globally, there are produced roughly comparable
levels of nucleophilic trapping (ca. 30-50%), relative to 1,3-OfN
acyl migration. These admittedly preliminary findings tend to
suggest the idea that under the microwave thermolytic conditions,
the stereochemistry of the FCMA which is produced (3a vs 3s)
determines to a great extent whether it undergoes 1,3-OfN acyl
transfer or interdiction.
Finally, we report on some early yet encouraging results pointing
to the applicability of the 2CC reaction to the synthesis of novel
peptides and peptidic constructs. We presumed that central to the
2CC method is the intermediacy of a FCMA of the general type
3a. As shown above, products can arise from either trapping by an
external acyl acceptor or 1,3-OfN acyl transfer. In contemplating
application to peptide and modified peptide synthesis, we noted
that the potential acyl acceptor is intramolecular, that is, the amide
bond flanking the site of proposed elongation of the C-terminus
through the 2CC reaction.
Our feasibility-exploring foray started with the phthalimidoyl
leucine derivative (27, Scheme 1). The free acid was coupled to
glycine isonitrile ester (28). Obviously, the selection of glycine
served to bypass issues of racemization at the isonitrile terminus.
Happily, reaction of 27 with 28 gave rise to 29 in 70% yield. We
next attempted to couple compound 31 with phenylalanine isonitrile
ester 32. This compound was prepared in optically pure form by a
known method.13 Given the strong acyl donating character of the
presumed FCMA (derived from the reaction of the carboxyl group
of 31 with the isonitrile function of 32), we were in effect asking
whether it would be intercepted by the neighboring amide bond of
31. In the event, reaction of 31 and 32 gave at best traces of
tripeptide 35. Presumably, reaction of 31 and 32 produced FCMA
of acid 9 with benzylisonitrile 10 under microwave pyrolysis.
Compound 12 was obtained in 85% yield. To evaluate the viability
of the alternative (symmetrical anhydride) route, we independently
examined the reaction of benzylformamide 13 and separately
synthesized symmetrical anhydride 14,10 under simulated micro-
wave thermolysis conditions. In this experiment, we were clearly
providing a far more favorable setting for such a pathway than could
be available via low levels of the same intermediates generated in
the course of the 2CC reaction. In the event, reaction of 13 and 14
under 2CC conditions afforded a maximum 16% yield of 12. While
we cannot assign a limiting maximal number for intervention of
the symmetrical anhydride pathway in the actual 2CC processes,
we conclude that the symmetrical anhydride pathway (cf. 14) is, at
best, a very minor contributor. Moreover as noted above, even in
the minor pathway in which 14 may be involved, reaction with 13
may actually occur by O-acylation (the invisible reaction), taking
one back to FCMA 11, which advances to 12 via 1,3-OfN acyl
transfer.
We then asked whether a formamide would be sufficiently
nucleophilic to intervene if it were present from the beginning as
an added component in the 2CC reaction. In the light of the
experiment described above, for a formamide such as 13 to interdict
in the context of the 2CC reaction, there would have to be generated
a much stronger acyl donor than is a symmetrical anhydride such
as 14. To approach this question, we studied the 2CC reaction of
aspartate derivative 15 with cyclohexylisonitrile 16 in the presence
of 4-methylcyclohexylformamide 17.11 As shown in Figure 3, there
are produced substantial amounts of crossover products (see 22 and
23). Essentially, the same level of crossover occurred in the reaction
of 4-methylisonitrile 18 in the presence of cyclohexylformamide
19 (see products 22 and 23). We take these experiments to teach
that during the course of the 2CC reactions, there must be produced
highly reactive acylating agents (presumably FCMA structures 20
and 21) capable of aspartylating even a relatively weak nucleophile
such as a formamide (cf. 17 or 19). Such a step accounts for the
substantial crossover observed. The simplest interpretation of the
events envisions the FCMA (20 or 21) reacting with the NH group
of cyclohexylformamide 17 or 19 to give directly, 22 or 23.12 Of
course, an alternative “invisible” possibility envisions O-acylation
of the 4-methylcyclohexyl formamide 17 leading to FCMA 21
which goes on to 23 by the postulated 1,3-OfN acyl transfer.
We next probed the question of intervention in the 2CC chemistry
by methanol (presumably via its reaction with FCMA 20) to produce
methyl ester 25. This reaction had been demonstrated earlier,1b but
was now interrogated at a more subtle level. We asked whether
trapping by methanol and formation of 1,3-OfN acyl transfer
product are truly competitive within the same molecular species,
or whether the apparent competition really stems from the formation
of two molecular entities 3a and 3s each of which is, in effect,
9
13226 J. AM. CHEM. SOC. VOL. 130, NO. 40, 2008

C O M M U N I C A T I O N S
Scheme 1a
In summary then, the message at this stage of the inquiry is that,
at least in the presence of N-formylamide, the intermediate peptidyl
FCMA is sustainable and can participate in the next 2CC reaction.
Future results describing applications of the 2CC reaction to novel
peptides and peptidic constructs will be forthcoming.
Acknowledgment. Support for this research was provided by
the National Institutes of Health (CA28824 to S.J.D.). We thank
Rebecca Wilson for editorial consultation and Dana Ryan for
assistance with the preparation of the manuscript. We also thank
Dr. George Sukenick, Ms. Hui Fang, and Ms. Sylvi Rusli, (NMR
Core Facility, Sloan-Kettering Institute) for mass spectral and NMR
spectroscopic analysis.
Supporting Information Available: Detailed experimental proce-
dures, copies of all spectral data, and compound characterization data.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) (a) Li, X.; Danishefsky, S. J. J. Am. Chem. Soc. 2008, 130, 5446. (b) Li,
X.; Yuan, Y.; Berkowitz, W. F.; Todaro, L. J.; Danishefsky, S. J. J. Am.
Chem. Soc. 2008, 130, 13222. (c) Jones, G. O.; Li, X.; Hayden, A. D.;
Houk, K. N.; Danishefsky, S. J. Org. Lett. 2008, 10, 4093.
(2) (a) Passerini, M. Gazz. Chim. Ital. 1921, 51, 181. (b) Ugi, I.; Meyr, R.;
Fetzer, U.; Steinbru¨ckner, C. Angew. Chem. 1959, 71, 386 For reviews,
see: (c) Banfi, L.; Riva, R. Org. React. 2005, 65, 1. (d) Do¨mling, A.; Ugi,
I. Angew. Chem., Int. Ed. 2000, 39, 3168. (e) Do¨mling, A. Chem. ReV.
2006, 106, 17.
(3) Shaabani, A.; Soleimani, E.; Rezayan, A. H. Tetrahedron Lett. 2007, 48, 6137.
(4) (a) Gloede, J.; Gross, H. Z. Chem. 1968, 8, 219. (b) Gloede, J. J. Prakt.
Chem. 1982, 324, 667.
(5) Recently, Rebek and colleagues described the apparent intervention of a
FCMA in the context of molecular encapsulation: (a) Hou, J-L.; Ajami,
D.; Rebek, J., Jr J. Am. Chem. Soc. 2008, 130, 7810. (b) Restorp, P.; Rebek,
J., Jr. J. Am. Chem. Soc. 2008, 130, 11850.
a Key: (a) MV, 160 °C, 30 min, ClCH2CH2Cl, 70%; (b) MV, 150 °C,
30 min, CHCl3; (c) HCO2H, 4 h, room temp; (d) NaHCO3, MeOH, room
temp, 2 h, 60%.
(6) The tendency of anti and syn imidic-carboxylic mixed anhydrides to
rearrange to imides and the interconversion of these isomers have been
discussed; see ref 7.
33. In its anti form (cf. 3a in Figure 1) 33 should be a candidate
for rearrangement to 35 via the usual 1,3-OfN acyl migration to
complete the 2CC reaction. We postulated that the failure of the
reaction to occur perhaps reflects formation of the intermediate
oxazalone 34,14 wherein the neighboring peptidic leucine amidic
bond interdicts the FCMA function in 33. Apparently, 34 is not a
sufficiently competent acyl donor to go on to 35. Rather, 34 suffers
eventual hydrolysis back to 31.
Given this interpretation, a fascinating possibility presented itself.
The thought was that the intramolecular acyl acceptor ability of an
N-formylamide system 30 (which is in fact derived from the primary
product of the 2CC reaction between 27 and 28) would be much
reduced relative to the amide system within 31. Accordingly, the
2CC reaction between compounds 30 and 32 was studied. Happily,
the projected reaction occurred to produce the bis formylamide 36.
While the yield of this coupling step producing the novel diformyl
compound 36 is presently rather modest, we see this as a most
encouraging direction for the buildup of polypeptides through the
2CC method. In the case at hand, cleavage of both carbonyl groups
provided a 60% yield of tripeptide 37.
(7) (a) Brady, K.; Hegarty, A. F. J. Chem. Soc., Perkin Trans, 2 1980, 121.
(b) Darbeau, R. W.; White, E. H.; Nunez, N.; Coit, B.; Daigle, M. J. Org.
Chem. 2000, 65, 1115. (c) Curtin, D. Y.; Miller, L. L. J. Am. Chem. Soc.
1967, 89, 637. (d) Sheehan, J. C.; Corey, E. J. J. Am. Chem. Soc. 1952,
74, 4555. (e) Schwarz, J. S. P. J. Org. Chem. 1972, 37, 2906. (f) Quast,
H.; Aldenkortt, S. Chem.sEur. J. 1996, 2, 462. (g) Schulenberg, J. W.;
Archer, S. Org. Reactions 1965, 14, 31. (h) Hegarty, A. F.; Tynan, N. M.;
Fergus, S. J. Chem. Soc., Perkin Trans. 2 2002, 1328.
(8) N-methyl peptides possess interesting biological profiles: (a) Subtelny, A. O.;
Hartman, M. C. T.; Szostak, J. W. J. Am. Chem. Soc. 2008, 130, 6131. (b)
Sagan, S.; Karoyan, P.; Lequin, O.; Chassaing, G.; Lavielle, S. Curr. Med.
Chem. 2004, 11, 2799. (c) Aurelio, L.; Brownlee, R. T. C.; Hughes, A. B.
Chem. ReV. 2004, 104, 5823. With our methodology, such scaffolds can
be readily reached as shown in ref 1a.
(9) Gautier, A. Liebigs Ann. 1896, 151, 240.
(10) Dhimitruka, I.; SantaLucia, J., Jr Org. Lett. 2006, 8, 47.
(11) Formamides 17 and 19 were prepared from the corresponding amines and
formic acetic anhydride. Isonitrile 18 was prepared by dehydration of
compound 17 (POCl3, Et3N).
(12) Compounds 17 and 18 are a diastereomeric mixture of trans/cis isomers,
as is 23.
(13) Skorna, G.; Ugi, I. Angew. Chem., Int. Ed. 1977, 16, 259.
(14) (a) McGahren, W. J.; Goodman, M. Tetrahedron 1967, 23, 2017. (b)
McGahren, W. J.; Goodman, M. Tetrahedron 1967, 23, 2031.
JA804709S
9
J. AM. CHEM. SOC. VOL. 130, NO. 40, 2008 13227