Aqueous-phase deactivation and intramolecular [2 + 2 + 2] cycloaddition of oxanorbornadiene esters

Article abstract of DOI:10.1021/ol103153f

Both inter- and intramolecular degradation pathways were identified for the aqueous phase deactivation of oxanorbornadiene (OND) electrophiles, and propargylic OND esters were found to undergo facile intramolecular [2 + 2 + 2] homo-Diels-Alder cycloaddition in polar media.

Full text of DOI:10.1021/ol103153f

ORGANIC
LETTERS
2011
Vol. 13, No. 7
1832–1835
Aqueous-Phase Deactivation and
Intramolecular [2 þ 2 þ 2] Cycloaddition
of Oxanorbornadiene Esters
Alexander A. Kislukhin, Cody J. Higginson, and M. G. Finn*
Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines
Road, La Jolla, California 92037, United States
mgfinn@scripps.edu
Received February 7, 2011
ABSTRACT
Both inter- and intramolecular degradation pathways were identified for the aqueous phase deactivation of oxanorbornadiene (OND) electrophiles,
and propargylic OND esters were found to undergo facile intramolecular [2 þ 2 þ 2] homo-Diels-Alder cycloaddition in polar media.
7-Oxabicyclo[2.2.1]hept-2,5-diene-2,3-dicarboxylates
(oxanorbornadienedicarboxylates, OND) have recently
attracted attention due to their potential as bioconjugation
reagents. The unique electrophilicity of a maleate moiety
embedded in a homoconjugated bicycle1 is manifested by
several processes more typical for electron-deficient al-
kynes than alkenes. These include facile Diels-Alder addi-
tion of a second furan,2 room temperature reactivity with
simplicity of preparation make OND reagents viable
competitors to traditional thiol-labeling reagents.
We previously reported the reactivity and stability of OND
reagents to be inversely, but not linearly, correlated.7 For
instance, dimethyl ester 1 (eq 1) was slightly more reactive
toward glutathione (k2 = 104 M-1 s-1) and less stable (half-
life in aqueous buffer = t1/2 = 9.3 days) than diethyl ester 2
(k2 = 63.4 M-1 s-1, t1/2 = 20.6 days).7 The slightly more
electron-deficient dipropargyl ester 3 was about twice as
reactive as 1 (k2 = 197 M-1 s-1), yet dramatically less stable
(t1/2 = 1.8 days). Compound 4, containing a bridgehead
methyl group, was 5-fold less reactive than 3 (k2 = 39.9 M-1
s-1), but the stability was improved only 2-fold (t1/2 = 3.2
days).7 In order to make the best use of OND electrophiles, we
investigated the pathways responsible for their deactivation,
with the goal of finding effective compromises between
reactivity and aqueous stability.
organic azides (k2 ≈ 10-3 M-1 s-1 3-6
)
and aliphatic thiols7
(k2 ≈ 102 M-1 s-1), and the ability to quench a pendant
dansyl fluorophore via photoinduced electron transfer.7,8
Despite enhanced electrophilicity and formidable strain,
OND-dicarboxylates are soft electrophiles that are re-
markably stable under ambient conditions. High aqueous
stability, exceptional reactivity toward thiols, and the
(1) Cristina, D.; De Amici, M.; De Micheli, C.; Gandolfi, R. Tetra-
hedron 1981, 37, 1349–1357.
(2) Lautens, M.; Fillion, E. J. Org. Chem. 1997, 62, 4418–4427.
(3) Berkel, S. S. v.; Dirks, A. J.; Marjoke, F. D.; Floris, L. v. D.;
Jeroen, J. L. M. C.; Roeland, J. M. N.; Floris, P. J. T. R. ChemBioChem
2007, 8, 1504–1508.
(4) Van Berkel, S. S.; Dirks, A. J.; Silvie, A. M.; Dennis, L. L. P.;
Otto, C. B.; Peter, L.; Floris, L. v. D.; Jeroen, J. L. M. C.; Floris, P. J. T.
R. ChemBioChem 2008, 9, 1805–1815.
(5) Laverman, P.; Meeuwissen, S. A.; van Berkel, S. S.; Oyen,
W. J. G.; van Delft, F. L.; Rutjes, F. P. J. T.; Boerman, O. C. Nucl.
Med. Biol. 2009, 36, 749–757.
(6) Canalle, L. A.; Berkel, S. S. v.; Haan, L. T. d.; Hest, J. C. M. v.
Adv. Funct. Mater. 2009, 19, 3464–3470.
(7) Hong, V.; Kislukhin, A. A.; Finn, M. G. J. Am. Chem. Soc. 2009,
131, 9986–9994.
(8) Guy, J.; Caron, K.; Dufresne, S.; Michnick, S. W.; Skene, W. G.;
Keillor, J. W. J. Am. Chem. Soc. 2007, 129, 11969–11977.
OND diesters such as 1-4 have two electrophilic sites.
While soft nucleophiles such as thiolates attack the double
bond, hard oxygen-centered nucleophiles such as hydroxide
r
10.1021/ol103153f
Published on Web 03/04/2011
2011 American Chemical Society

Figure 1. Degradation pathways of OND electrophiles.
and alkoxides were expected to react with the ester
groups.3,7,9,10 Indeed, exposure of dialkyl OND diesters
1-4 to an aqueous hydroxide ion led only to sequential
saponifications (Figure 1, path A), selectively giving fluor-
escent monoacids 5 when 1 equiv of base was used.7 To study
the stability of 1-4 in neutral aqueous buffers, we originally
incubated 0.1 mM 1-4 in 1:9 DMSO/aqueous phosphate
(0.1 M, pH 7.0). Interestingly, upon consumption of 1 and 2
the samples did not become fluorescent, but they gradually
lost their turn-on capacity (i.e., addition of a thiol did not
result in the appearance of fluorescence). Mass spectrometry
revealed a single set of peaks with the same m/z values as the
starting OND, suggesting an isomerization process.
Incubation of dimethyl ester 1 on a preparative scale in
4:1 acetonitrile/phosphate buffer (pH 7) over several days
did not result in its degradation, presumably due to a lower
buffer strength (20 mM) or a smaller aqueous component
in the reaction medium. This suggests that the degradation
pathway(s) of 1 involves a general acid or base or has a
negative volume of activation which responds well to
the high cohesive energy density of water. At elevated
temperatures (50 °C, 5 days) partial saponification to
monoacid 5 was observed (Figure 1, path A). Interestingly,
when excess acetylacetone (HAcAc) was added to the
reaction mixture, ester 1 gave the nonfluorescent isomeric
(E)-olefin 7 in excellent yield. It is likely that the acetyla-
cetone anion acts as a base and deprotonates the sulfo-
namide11 of 1, which undergoes intramolecular 4-exotrig
conjugate addition to give 6. This highly strained azetidine
should then be subject to fast retro-Diels-Alder (RDA)
cycloreversion to form (E)-7 (path B1). Alternative me-
chanisms involving intermolecular conjugate addition-
RDA fragmentation sequences could not be ruled out.
Plausible intermediates 8 and 9 formed by conjugate
addition of water or methanol to 1 could not be detected.
Blocking the sulfonamide group (methylated compound
10) diverted the pathway to the simple conjugate addition
of methanol (path B2), giving 12 in 75% yield. The
addition of water (to give 11) was not detected under these
conditions or when acetonitrile was used in place of
methanol as a cosolvent. Methanol adduct 12 was found
to be much less susceptible to RDA fragmentation than
thiol adducts, although furan 13 and methoxymaleate 14
(9) Baran, P. S.; Li, K.; O’Malley, D. P.; Mitsos, C. Angew. Chem.,
Int. Ed. 2006, 45, 249–252.
(10) Michieletto, I.; Fabris, F.; De Lucchi, O. Tetrahedron: Asym-
(11) Koike, T.; Watanabe, T.; Aoki, S.; Kimura, E.; Shiro, M. J. Am.
Chem. Soc. 1996, 118, 12696–12703.
metry 2000, 11, 2835–2841.
Org. Lett., Vol. 13, No. 7, 2011
1833

same course as 10. In other words, despite having an
additional decomposition route available (pathways B1
and B2), the NH-sulfonamide 1 was more stable than
methylated 10 (only pathway B2).7 This suggested the
presence of a stabilizing hydrogen-bonding interaction
available to 1 but not 10.
Table 1. Homo-Diels-Alder Reaction of OND Alkynesa
Degradation of various OND derivatives in unbuffered
aqueous solutions required more forcing conditions
(80-100 °C). For example, benzamide 15 gave a mixture
of furan 16 and 1:1 furan-OND adduct2 17; the latter was
the major product when the reaction was run at high
concentration. This suggests that intermolecular conjugate
addition of water does occur and is followed by rapid RDA
fragmentation at high temperatures (path B2). A significant
amount of polar compounds is formed under these condi-
tions; mono- and diacids 18 could be detected in the reaction
mixtures by ESI-MS. It is unlikely that retro-Diels-Alder
cleavage occurs in 15 or any other OND derivatives de-
scribed here since such fragmentations typically require very
harsh conditions and favor the formation of acetylene and
substituted furan12 19, which was not detected.
Instead of the above hydrolysis, Michael, and retro-
Diels-Alder pathways, the dipropargyl esters 3 and 4 were
each found to give two fluorescent products, 20a/b and
21a/b, derived from cycloadditions on opposite sides of the
oxanorbornadiene core. This reaction occurred to ap-
proximately 50% completion over six months of storage
(room temperature, capped flask) and in only a few days in
aqueous buffer at the same temperature. The structures
were assigned on the basis of 1H, 13C, COSY, and 1H-13C
HMBC NMR spectroscopy and by X-ray crystallographic
analysis of one of the products (21a) obtained from 4
(Supporting Information).
The intramolecular [2 þ 2 þ 2] transformations of
propargylic OND substrates proceeded under mild condi-
tions and were accelerated in aqueous acetonitrile, presum-
ably by virtue of the standard hydrophobic effect common
for cycloaddition reactions (Table 1).13-16 Heating at reflux
was found to be best for preparative-scale reactions (entry 4).
OND diester 26 gave unsubstituted pentacycle 27 in situ
under the conditions of the Diels-Alder reaction (entry 5),
or after isolation and heating (entry 6). As expected, internal
propargylic alkynes underwent the same reaction, com-
pound 28 providing pentacycles 29a,bin good yield (entry 7).
The cyclization rate was quite sensitive to linker length.
Homopropargyl ester 30 was much more stable than the
analogous propargyl esters 3 and 4, and the expected δ-
lactone product could not be detected (entry 8). The use of
copper(I) to catalyze the reaction of 30 was not effective.
Similarly, dihomopropargyl ester 31 and bis(pent-4-yn-1-
yl) ester 33 gave only trace amounts of the corresponding
a Prg = propargyl. b 0.7 mmol scale.
(12) Rickborn, B. The Retro-Diels-Alder Reaction Part I. C-C
Dienophiles. Organic Reactions; John Wiley & Sons: 1998.
(13) Rideout, D. C.; Breslow, R. J. Am. Chem. Soc. 1980, 102, 7817–
7818.
(14) Breslow, R. Acc. Chem. Res. 1991, 24, 159–64.
(15) Hunt, I.; Johnson, C. D. J. Chem. Soc., Perkin Trans. 2 1991,
1051–1056.
could be detected in the reaction mixture by TLC and
ESI-MS. Since the alternative degradation reaction of 10
took place at lower temperature than the decomposition of
(16) Repasky, M. P.; Jorgensen, W. L. Faraday Trans. 1998, 110,
379–389.
1, it is significant that the reaction of 1 did not follow the
1834
Org. Lett., Vol. 13, No. 7, 2011

δ- (32) and ε-lactones (34) (entries 9, 10). Dihomopropargyl
esters 30 and 31 were recovered largely unchanged, while
diester 33 was completely degraded after heating at 100 °C,
presumably by hydrolysis and Michael addition at the higher
temperature. N-Propargylamide 35 (entry 11) also did not
cyclize, presumably because a strong intramolecular hydro-
gen bond locks the amide moiety in a planar conformation.17
The attempted formation of saturated oxadeltacyclanes
from diallyl ester 36 was also unsuccessful (entry 12).
The oxadeltacyclene products are formed by intra-
molecular homo-Diels-Alder (HDA) reaction, or [2 þ
2 þ 2] cycloaddition. This process is well precedented
with norbornadienes and oxanorbornadienes but pro-
ceeds only under harsh conditions (>90 °C, several
days) with poor chemoselectivity, unless catalyzed by
cobalt18 or ruthenium19 complexes. Intramolecular,
intermolecular, and asymmetric versions are known.20
Deltacyclenes prepared from norbornadienes are use-
ful building blocks for the synthesis of polycyclic
natural products,19 but reports of HDA adducts of
7-oxanorbornadienes are relatively rare. The Diels-
Alder adduct between 2,5-dimethylfuran and ethyl
propiolate has been reported to undergo an HDA
reaction with itself, producing oxadeltacyclenes as
well as many other byproducts;21 bis-furan adducts
of electron-deficient dialkynylcarbinols undergo this
formal dimerization intramolecularly.22
In summary, analysis of the aqueous deactivation of
OND electrophiles revealed base-catalyzed isomerization
of dansylamide-bearing bicycles and a facile intramolecu-
lar homo-Diels-Alder process. Since sulfonamidomaleate
7, the degradation product of the simplest dimethyl ester 1,
is not fluorescent, partial decomposition should not be
problematic for the detection and fluorogenic labeling of
thiols. The somewhat limited stability of diesters 1 and 2 is
only a minor drawback, given that they are the easiest to
prepare among any fluorogenic thiol-reactive reagents.
Furthermore, the cyclization-resistant (and therefore
stable) acetylenic esters are highly useful as “clickable”
connectors, since they retain their high reactivity toward
thiols. For example, the dihomopropargyl ester 30 reacted
with glutathione with a second-order rate constant of 57.6
( 2.7 M-1 s-1, comparable to the case of diethyl ester 2,
and no degradation over extended exposure (several
months at room temperature) to aqueous buffers was
observed. N-Propargylamide 35 was successfully used as
a linker for attaching BSA (single thiol group) to an azide-
functionalized Qβ protein nanoparticle.25 Further investi-
gationsofthe performanceand utilityof OND reagentsare
in progress.
Unlike related and more strained oxaquadricyclanes,23
the pentacyclic oxahemiquadricyclane lactones produced
here proved to be quite stable. Compounds 21a and 27
were unchanged after heating at 120 °C for several days.
Attempted acid-catalyzed rearrangement of 27 performed
under conditions that cleave three out of four cycles in
oxaquadricyclanes24 (5% methanolic H2SO4) resulted
only in sequential lactone opening (37) and transesterifica-
tion (38) (eq 2).
Acknowledgment. Support from the NIH (RR021866)
and The SkaggsInstituteforChemicalBiologyis gratefully
acknowledged. We thank Dr. Michal Sabat (University of
Virginia) for the X-ray crystal structure of 21a.
Supporting Information Available. Experimental pro-
cedures and characterization data of new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(17) Yi, L.; Li, H. Y.; Sun, L.; Liu, L. L.; Zhang, C. H.; Xi, Z. Angew.
Chem., Int. Ed. 2009, 48, 4034–4037.
(18) Lautens, M.; Tam, W.; Lautens, J. C.; Edwards, L. G.; Crudden,
C. M.; Smith, A. C. J. Am. Chem. Soc. 1995, 117, 6863–6879.
(19) Tenaglia, A.; Gaillard, S. Org. Lett. 2007, 9, 3607–3610.
(20) Lautens, M.; Lautens, J. C.; Smith, A. C. J. Am. Chem. Soc.
1990, 112, 5627–5628.
(23) Tochtermann, W.; Olsson, G. Chem. Rev. 1989, 89, 1203–1214.
(24) O’Malley, D. P.; Li, K.; Maue, M.; Zografos, A. L.; Baran, P. S.
J. Am. Chem. Soc. 2007, 129, 4762–4775.
(21) McCulloch, A. W.; Smith, D. G.; McInnes, A. G. Can. J. Chem.
1974, 52, 1013–1018.
(22) Barlow, M. G.; Tajammal, S.; Tipping, A. E. J. Chem. Soc.,
Perkin Trans. 1 1992, 2485–2494.
(25) Hong, V.; Presolski, S. I.; Ma, C.; Finn, M. G. Angew. Chem.,
Int. Ed. 2009, 48, 9879–9883.
Org. Lett., Vol. 13, No. 7, 2011
1835