Olefin cross-metathesis on proteins: Investigation of allylic chalcogen effects and guiding principles in metathesis partner selection

Article abstract of DOI:10.1021/ja104994d

Olefin metathesis has recently emerged as a viable reaction for chemical protein modification. The scope and limitations of olefin metathesis in bioconjugation, however, remain unclear. Herein we report an assessment of various factors that contribute to productive cross-metathesis on protein substrates. Sterics, substrate scope, and linker selection are all considered. It was discovered during this investigation that allyl chalcogenides generally enhance the rate of alkene metathesis reactions. Allyl selenides were found to be exceptionally reactive olefin metathesis substrates, enabling a broad range of protein modifications not previously possible. The principles considered in this report are important not only for expanding the repertoire of bioconjugation but also for the application of olefin metathesis in general synthetic endeavors.

Full text of DOI:10.1021/ja104994d

Published on Web 11/04/2010
Olefin Cross-Metathesis on Proteins: Investigation of Allylic
Chalcogen Effects and Guiding Principles in Metathesis
Partner Selection
Yuya A. Lin, Justin M. Chalker, and Benjamin G. Davis*
Department of Chemistry, UniVersity of Oxford, Chemistry Research Laboratory,
12 Mansfield Road, Oxford OX1 3TA, U.K.
Received June 8, 2010; E-mail: ben.davis@chem.ox.ac.uk
Abstract: Olefin metathesis has recently emerged as a viable reaction for chemical protein modification.
The scope and limitations of olefin metathesis in bioconjugation, however, remain unclear. Herein we report
an assessment of various factors that contribute to productive cross-metathesis on protein substrates.
Sterics, substrate scope, and linker selection are all considered. It was discovered during this investigation
that allyl chalcogenides generally enhance the rate of alkene metathesis reactions. Allyl selenides were
found to be exceptionally reactive olefin metathesis substrates, enabling a broad range of protein
modifications not previously possible. The principles considered in this report are important not only for
expanding the repertoire of bioconjugation but also for the application of olefin metathesis in general synthetic
endeavors.
Scheme 1. Sulfur-Assisted Cross-Metathesis
Introduction
Precise modification of proteins allows the dissection and
analysis of many biological systems.1,2 While many advances
in bioorthogonal ligation have been reported, there is still only
a limited set of transformations that are inert to the biological
milieu.2 This privileged set of transformations must be chemose-
lective and efficient in aqueous media for general utility in
protein modification. Olefin metathesis has emerged among a
list of popular transition metal-mediated transformations as a
potential candidate for selective carbon-carbon bond formation
on proteins.3 Olefin metathesis enables the installation of a
carbon-carbon bond which is largely inert to a range of
biological processes. Olefin cross-metathesis (CM) is also an
attractive chemical challenge since aqueous olefin metathesis,
while advancing,4-6 is still in its infancy.
In our exploratory work in aqueous cross-metathesis,7 the
amino acid S-allylcysteine (Sac) was found to be a reactive
substrate in olefin metathesis. When compared to its all-carbon
analogue homoallylglycine (Hag), S-butenyl- and S-pentenyl-
cysteine, Sac was the most reactive and was the only residue
to afford a synthetically useful amount of CM product.7 Allyl
sulfide reactivity in CM, much less aqueous CM, was at first
counterintuitive since sulfur is known to confound many metal-
mediated reactions.8 The unique reactivity of Sac was explained
with a sulfur-assisted mechanism whereby sulfur precoordination
of the allyl sulfide to the ruthenium center increases the effective
concentration of the alkylidene and alkene metathesis partner
without detrimental chelation (Scheme 1). While allyl sulfides
(1) (a) Carrico, I. S. Chem. Soc. ReV. 2008, 37, 1423–1431. (b) Gamblin,
D. P.; van Kasteren, S. I.; Chalker, J. M.; Davis, B. G. FEBS J. 2008,
275, 1949–1959. (c) Hermanson, G. T. Bioconjugation Techniques,
2nd ed.; Academic Press: New York, 2008. (d) Basle´, E.; Joubert, N.;
Pucheault, M. Chem. Biol. 2010, 17, 213–227.
(2) (a) Hackenberger, C. P. R.; Schwarzer, D. Angew. Chem., Int. Ed.
2008, 47, 10030–10074. (b) Sletten, E.; Bertozzi, C. Angew. Chem.,
Int. Ed. 2009, 48, 6974–6998. (c) Lim, R. K. V.; Lin, Q. Chem.
Commun. 2010, 1589–1600.
(3) (a) Antos, J. M.; Francis, M. B. Curr. Opin. Chem. Biol. 2006, 10,
253–262. (b) Binder, J. B.; Raines, R. T. Curr. Opin. Chem. Biol.
2008, 12, 767–773. (c) Chalker, J. M.; Wood, C. S. C.; Davis, B. G.
J. Am. Chem. Soc. 2009, 131, 16346–16347. (d) Lin, Y. A.; Chalker,
J. M.; Davis, B. G. ChemBioChem 2009, 10, 959–969.
(4) (a) Hong, S. H.; Grubbs, R. H. J. Am. Chem. Soc. 2006, 128, 3508–
3509. (b) Jordan, J. P.; Grubbs, R. H. Angew. Chem., Int. Ed. 2007,
46, 5152–5155.
(7) Lin, Y. A.; Chalker, J. M.; Floyd, N.; Bernardes, G. J. L.; Davis, B. G.
J. Am. Chem. Soc. 2008, 130, 9642–9643.
(5) (a) Binder, J. B.; Blank, J. J.; Raines, R. T. Org. Lett. 2007, 9, 4885–
4888. (b) Gułajski, Ł.; Michrowska, A.; Naroznik, J.; Kaczmarska,
Z.; Rupnicki, L.; Grela, K. ChemSusChem 2008, 1, 103–109.
(6) (a) Lipshutz, B. H.; Aguinaldo, G. T.; Ghorai, S.; Voigtritter, K. Org.
Lett. 2008, 10, 1325–1328. (b) Burtscher, D.; Grela, K. Angew. Chem.,
Int. Ed. 2009, 48, 442–454.
(8) (a) Armstrong, S. K.; Christie, B. A. Tetrahedron Lett. 1996, 37, 9373–
9376. (b) Mascaren˜as, J. L.; Rumbo, A.; Castedo, L. J. Org. Chem.
1997, 62, 8620–8621. (c) Shon, Y.-S.; Lee, T. R. Tetrahedron Lett.
1997, 38, 1283–1286. (d) Fu¨rstner, A.; Seidel, G.; Kindler, N.
Tetrahedron 1999, 55, 8215–8230.
9
10.1021/ja104994d  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 16805–16811 16805

A R T I C L E S
Lin et al.
have been used in olefin metathesis previously,9 the enhanced
reactivity relative to other alkenes went unnoticed and was not
exploited in synthesis. Our findings showed that allyl sulfides
are not simply tolerated; they can enhance the rate of olefin
metathesis. This enhanced rate is critical in aqueous systems
where catalyst decomposition pathways may compete.10 The
reactivity of allyl sulfides in metathesis motivated the incorpora-
tion of Sac into proteins. Indeed, ready chemical access to Sac
on protein surfaces enabled the first cross-metathesis on protein
substrates.7,11
Scheme 2. Alkene Substrates Used in This Investigation
These preliminary reports demonstrated the enhanced reactiv-
ity of Sac in olefin metathesis and its use in covalent protein
modification. However, the CM substrates tested were largely
limited to simple allyl ethers, and the full scope of CM as a
method for bioconjugation remains unclear. In fact, only a single
example in these reports proceeded to completion: the cross-
metathesis between the protein substrate and allyl alcohol. For
general use of olefin metathesis as a bioconjugation technique,
understanding the scope and limitations of metathesis substrates
is essential. Indeed, Grubbs has established some guiding
principles in substrate selection for CM in organic solvent,12
yet additional factors must be considered for successful CM on
protein substrates. Most notably, the reaction must proceed
rapidly in water at or near room temperature. The metathesis
partners and any intervening linker must therefore be selected
with these stringent requirements in mind. With a clear
understanding of the scope and limitations of metathesis
partners, olefin metathesis may be deployed more routinely in
bioconjugation. Moreover, progress in the genetic incorporation
of alkene-containing unnatural amino acids further motivates
the development of olefin metathesis as a method for protein
modification.13,14 Finally, an increased understanding of chemi-
cal behavior in aqueous olefin metathesis is useful in general
synthetic endeavors.15 These considerations motivate our in-
vestigation of aqueous CM on protein substrates. Herein, we
report our investigation into allylic chalcogen activation effects
in olefin metathesis and the scope of CM partners useful in
protein cross-metathesis.
Grubbs second generation catalyst (1),16 the substrate scope of
the reaction was not fully assessed. Examples to date include
simple allylic alcohols and ethers.7,11 Since we are interested
in biorelevant protein modifications, allyl ethers containing
carbohydrates, oligo(ethylene glycols), and charged groups were
among the metathesis substrates synthesized for this study.
Compounds containing an allyl sulfide, N-allyl amines, or longer
alkene tethers are also metathesis partners of interest for the
assessment of substrate scope (Scheme 2).
Reactions were monitored for up to 2 h, the reaction time
necessary for complete conversion with allyl alcohol as the
metathesis partner (Table 1, entry 1). Reaction conversions were
determined by ESI-MS.17 When protein 2 was tested with each
of the substrates in CM, the best results were obtained with
allylic alcohols, ethers, and hexenyl glucoside 10 (Table 1,
entries 1-7). The reaction worked moderately well with allyl
glycosides 6-8 and oligo(ethylene glycol) derivative 9, with
conversions ranging from 30 to 65%. CM with hexenyl
glucoside 10 importantly revealed the sensitivity to linker length,
with full conversion to the modified protein after only 1 h at
room temperature (Table 1, entry 7). This result compares
favorably to the allyl glycosides in entries 3-5 and was the
first carbohydrate-bearing substrate to proceed with full
conversion.
As an additional guide to mechanism, CM with self-
metathesis product of allyl alcohol (5) was carried out to test
whether and under what conditions it is a reactive substrate.
The self-metathesis product of allyl alcohol used in Table 1
was largely the E isomer isolated from a model cross-metathesis
in water (see ESI). Cross-metathesis of 5 with protein 2 only
reached 28% conversion under the same reaction conditions as
allyl alcohol (Table 1, entry 2). This result suggests that the
CM in entry 1 is mainly with allyl alcohol and not with 5. The
difference in conversion is likely due to a higher rate of
metathesis of 4 than 5 since the latter is a more substituted
alkene and therefore generally slower in olefin metathesis.18
However, when heated to 37 °C, the reaction with 5 proceeded
Results and Discussion
Assessing the Substrate Scope of Protein Cross-Metathesis
at S-Allylcysteine. The model protein used for our studies was
a single cysteine mutant of subtilisin from Bacillus lentus (SBL-
S156C). Sac was installed on SBL-S156C by direct allylation
with allyl chloride, in accordance to our previous report.11 While
we have previously disclosed several examples of CM on the
single Sac mutant of SBL (2, SBL-156Sac) using Hoveyda-
(9) (a) Spagnol, G.; Heck, M.-P.; Nolan, S. P.; Mioskowski, C. Org. Lett.
2002, 4, 1767–1770. (b) Toste, F. D.; Chatterjee, A. K.; Grubbs, R. H.
Pure Appl. Chem. 2002, 74, 7–10.
(10) (a) Dinger, M. B.; Mol, J. C. Organometallics 2003, 22, 1089–1095.
(b) Dinger, M. B.; Mol, J. C. Eur. J. Inorg. Chem. 2003, 2827–2833.
(11) Chalker, J. M.; Lin, Y. A.; Boutureira, O.; Davis, B. G. Chem.
Commun. 2009, 3714–3716.
(12) Chatterjee, A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360–11370.
(13) (a) van Hest, J. C. M.; Tirrell, D. A. FEBS Lett. 1998, 428, 68–70.
(b) Zhang, Z.; Wang, L.; Brock, A.; Schultz, P. G. Angew. Chem.,
Int. Ed. 2002, 41, 2840–2842.
(14) Ai, H.-w.; Shen, W.; Brustad, E.; Schultz, P. G. Angew. Chem., Int.
Ed. 2010, 49, 935–937.
(16) Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2000, 122, 8168–8179.
(15) (a) Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3012–3043. (b)
Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed.
2005, 44, 4490–4527. (c) Hoveyda, A. H.; Zhugralin, A. R. Nature
2007, 450, 243–251.
(17) For a discussion and validation of ESI-MS as a tool for monitoring
reaction conversion on protein substrates, see Supporting Information.
(18) Kirkland, T. A.; Grubbs, R. H. J. Org. Chem. 1997, 62, 7310–7318.
9
16806 J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010

Olefin Cross-Metathesis on Proteins
A R T I C L E S
Table 1. Substrate Scope of Cross-Metathesis with SBL-156Sac
Scheme 3. Comparison in Rate of Self-Metathesis between 12a
and 13a
3) reveals that the acetamide at C-2 in 12a must contribute some
adverse steric, electronic, or chelating influence that impedes
metathesis. Allyl sulfide 13a was synthesized with the intention
of overcoming these obstacles, since this linker might be
expected to be more reactive. However, only self-metathesis
was observed (see ESI). To further investigate this observation,
a model reaction comparing the rate of self-metathesis between
12a and 13a was carried out (Scheme 3). The formation of the
self-metathesis product of 13a was observed within the first 30
min of reaction. After 2.5 h, the self-metathesis product 13b
was isolated in a yield of 59%. In contrast, the self-metathesis
of 12a only resulted in >95% recovery of starting material under
the same reaction conditions. When 13b was tested in CM with
protein 2, no reaction was observed (Table 1, entry 11).
Apparently, the self-metathesis product 13b, unlike 5, cannot
re-enter the metathesis cycle. We attribute this low reactivity
of 13b to its hindered structure. The propensity for self-
metathesis and the resulting reactivity of the self-metathesis
product are therefore important considerations in substrate
selection. Reaction with vinyl C-glucoside 14 was unfruitful,
likely because of the steric congestion at the alkene. Allyl
acetamide 15 performed poorly in CM with protein 2, possibly
due to the formation of a stable six-member ring chelate via
carbonyl oxygen coordination to ruthenium, poisoning the
catalyst.
From these initial results in Table 1, it seems that in order
for CM on Sac-containing proteins to work efficiently, the
metathesis partner needs to be slightly less reactive than the
allyl sulfide. If the metathesis partner is highly reactive (e.g., 4
or 13a), the product of self-metathesis must be able to re-enter
the catalytic cycle or no protein modification is observed. Of
the substrates tested, allylic alcohols and ethers, and hexenyl
glucoside 10 stood out as the most productive metathesis
partners for Sac-containing protein 2. The remaining substrates
(Table 1, entries 8-14) provide a benchmark of challenging
transformations that can perhaps be achieved by altering the
protein metathesis partner or linker. Accordingly, we turned next
to an assessment of the accessibility of the Sac residue on the
protein and its effect on cross-metathesis.
Making the Reactive Site More Accessible: Linker-Ex-
tended S-Allylcysteine on Protein Substrates. CM failed to reach
full conversion with metathesis partners such as allyl glycosides
and oligo(ethylene glycols). In other cases we saw no productive
CM, particularly for sterically demanding and electron-poor
substrates. According to the proposed sulfur-assisted metathesis
mechanism, coordination of the Sac residue to the ruthenium
is critical. We therefore investigated the influence of sterics at
the protein surface. Accordingly, acrylamide 17 was used to
install a Sac residue at the same protein site, but extended from
the protein surface to create a less hindered protein olefin.
Conjugate addition of the cysteinyl residue on SBL-S156C to
a Determined by LC-MS. b First 2 h at RT.
with over 95% conversion after 30 min. This result indicates
that while slower than allyl alcohol, the self-metathesis partner
can successfully re-enter the metathesis cycle (Table 1, entry
2). This observation is consistent with other reports on the
reversible nature of olefin CM.19
A number of CM partners failed to react with Sac-containing
proteins. No product formation was observed in the CM between
protein 2 and substrates 11-16 (Table 1, entries 8-14). In all
of these attempts only starting material was detected on LC-
MS. These results revealed structural features of metathesis
partners that may adversely affect the rate of reaction. For
instance, alkenes 11 and 16 contain electron-withdrawing
ammonium groups known to retard olefin metathesis.4 N-
Acetylglucosamine-derived substrates 12a and 13a also gave
no detectable CM product. Comparison of GlcNAc derivative
12a (Table 1, entry 9) with glucose derivative 6 (Table 1, entry
(19) (a) Fu¨rstner, A.; Thiel, O. R.; Ackermann, L. Org. Lett. 2001, 3, 449–
451. (b) Smith, A. B., III; Adams, C. M.; Kozmin, S. A. J. Am. Chem.
Soc. 2001, 123, 990–991.
9
J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010 16807

A R T I C L E S
Lin et al.
Scheme 4. Installing Extended S-Allylcysteine on Protein Surface
proceeded with full conversion. Again, shorter reaction time
was required with the protein containing the extended linker.
The influence of steric effects observed on our model system
should be considered when a modification is desired at an active
site or hindered helix. Yet steric hindrance is only one dominant
factor. Substrates 13a-16 remained challenging, even for the
Sac-extended protein 18. Substrate 13a again only resulted in
unproductive self-metathesis (see ESI), whereas compounds
14-16 are either too electron-deficient or sterically demanding
to participate in CM.
Table 2. Cross-Metathesis on SBL 18-Extended Sac
Se-Allylselenocysteine: A Metathesis Substrate Superior to
S-Allylcysteine in Aqueous Cross-Metathesis. From our initial
report on allyl sulfides7 and the results above, it was increasingly
apparent that allylic heteroatoms modulate the rate of olefin
metathesis. When considering these observations alongside
reports of the positive influence of allylic alcohols20 and ethers14
in olefin metathesis, it is tempting to consider if this enhanced
reactivity was general for allylic chalcogenides. Pursuing this
hypothesis, we next examined the CM reactivity of allyl
selenides. Accordingly, Se-allylselenocysteine (Seac) derivative
21a was synthesized and then tested along with Sac derivative
20a in model aqueous CM with allyl alcohol under identical
reaction conditions. Indeed, the reaction with Seac 21a was
higher yielding than the Sac case, with respective yields of 72
and 56% (Scheme 5a). The difference in reactivity may be
attributed to the softness of selenium which makes the coordina-
tion to ruthenium even more favorable than the sulfur in Sac.
Remarkably, there are few examples in the literature describing
olefin metathesis with selenium-containing compounds. In one
instance, Koketsu and co-workers used RCM of an allyl selenide
derivative as a key step for the synthesis of selenium-containing
bicyclic ꢀ-lactams.21 However, the scope of olefinic selenoethers
in olefin metathesis was not manifested in these reports, and
the enhanced reactivity of allyl selenides was not noted.
Next, a more complex and biochemically important carbo-
hydrate metathesis partner 22 was used in place of allyl alcohol.
The CM reaction of sugar 22 with Sac 20a and Seac 21a gave
moderate yields of 45 and 53%, respectively. Initial inspection
of the yields of CM products 20d and 21d suggests no difference
in reactivity between Sac and Seac. However, self-metathesis
product 21c was also isolated from the reaction of Seac, whereas
no self-metathesis product of Sac (20c) was observed (Scheme
5b). Total CM yields (CM and self-metathesis) are therefore
73% for Seac and 45% for Sac, a clear indication that not only
are allyl selenides reactive in cross-metathesis but also they are
more reactive than allyl sulfides. We sought to take advantage
of this reactivty of allyl selenides in protein conjugation. In
particular, the more reactive allyl selenide was tested for its
ability to promote challenging CM with substrates that were
sluggish or unreactive with Sac.
a Determined by LC-MS.
acrylamide 17 led to full conversion of the alkylated protein
18 after incubation at 37 °C for 1 h (Scheme 4). The reaction
at cysteine was verified with Ellman’s assay (see ESI).
With a protein containing a linker-extended Sac in hand, CM
with allyl alcohol (4) and ether substrates 6 to 12a were carried
out. Notably, all reactions proceeded to full conversions with
the exception of ethanolamine 11 and GlcNAc 12a, which gave
29 and 53% conversion, respectively (Table 2). Protein modi-
fication via CM with compound 11 and 12a, though only
achieved with low conversions, was promising because these
were the only two allyl ether substrates that had failed to work
with protein 2. The results summarized in Table 2 demonstrated
that steric effects are indeed important for CM on the surface
of SBL. CM with allyl alcohol reached full conversion after
just 30 min of reaction time at room temperature, while the
same reaction with protein 2 required 2 h to proceed to
completion. Protein CM with hexenyl glucoside 10 also
Chemical Access to Se-Allylselenocysteine on a Protein
Surface. To directly compare the CM reactivity with protein
18, Seac-containing protein 24 was synthesized in a similar
manner using Seac acrylamide 23 (Scheme 6). Again, reaction
at cysteine was verified with Ellman’s assay (see ESI).
The remarkable reactivity of allyl selenide-containing protein
24 was clear after the first test for CM activity. CM between
allyl alcohol and protein 24 required only 15 min at room
(20) (a) Hoye, T. R.; Zhao, H. Org. Lett. 1999, 1, 1123–1125. (b) Hoveyda,
A. H.; Lombardi, P. J.; O’Brien, R. V.; Zhugralin, A. R. J. Am. Chem.
Soc. 2009, 131, 8378–8379.
(21) Garud, D. R.; Garud, D. D.; Koketsu, M. Org. Biomol. Chem. 2009,
7, 2591–2598.
9
16808 J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010

Olefin Cross-Metathesis on Proteins
A R T I C L E S
Scheme 5. Comparison of Reactivity between S-Allylcysteine and Se-Allylselenocysteine in Aqueous Cross-Metathesis
Scheme 6. Installing Extended Se-Allylselenocysteine on Protein
Surface
Table 3. Cross-Metathesis on SBL 24-Extended Seac
temperature to reach completion (Table 3, entry 1). The LC-
MS data of this reaction is shown in Scheme 7. The total ion
chromatogram (TIC) was typically analyzed between 13 and
16 min, the time of elution for all protein material, both
unmodified and modified. Ethers also reached full conversion
under mild reaction conditions (Table 3, entries 2-8). Notably,
these substrates included the more challenging ethanolamine
11 and GlcNAc 12a, which resulted in poor conversions in
previous attempts with protein 18. CM with the reactive hexenyl
glucoside 10 also gave full conversion (Table 3, entry 6).
Moreover, allyl acetamide 15, a substrate that was unreactive
in all previous CM reactions, also gave productive CM with
protein 24 (Table 3, entry 11). Either the unhindered Seac on
protein 24 was able to initiate rapid CM with 15 before the
catalyst was sequestered by the acetamide or the allyl selenide
is simply a better ligand for ruthenium than the chelating
acetamide. Among the CM reactions carried out on protein 24,
the modifications with GlcNAc 12a and acetamide 15 are
particularly biologically relevant modifications. The GlcNAc
moiety on the glycosylated protein 25h is an anchor for many
modifications and bioprocesses such as celluar signaling.22
Moreover, acetylation such as that found in 25i is an important
protein posttranslational modification (PTM) and often occurs
either at the N-terminus or at lysine residues of proteins.23 The
attachment of an N-acetyl group on proteins by CM might
suffice as a mimic for the natural PTM of lysine residues in
proteins.24 The results in Table 3 are highly promising, and it
a Determined by LC-MS.
is clear that unhindered allyl selenides allow unprecedented
reactivity in bioconjugation by olefin metathesis. Nonetheless,
some limitations for CM using 24 remain. Substrates 13a, 14,
and 16 did not participate in productive CM. Likely C-vinyl
glucoside 14 is too hindered and 16 is both too hindered and
(22) Hart, G. W.; Akimoto, Y. In Essentials of Glycobiology; 2nd ed.; Cold
Spring Harbor Laboratory Press: New York, 2009; pp 263-279.
(23) Walsh, C. T.; Garneau-Tsodikova, S.; Gatto, G. J., Jr. Angew. Chem.,
Int. Ed. 2005, 44, 7342–7372.
(24) Davis, B. G. Science 2004, 303, 480–482.
9
J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010 16809

A R T I C L E S
Lin et al.
Scheme 7. Cross-Metathesis between SBL 24 and Allyl Alcohol
Scheme 8. Summary of Substrate and Linker Selection from Cross-Metathesis on Model Protein SBL
electron deficient. Again, 13a underwent preferential self-
metathesis, and no CM product was detected.
that extending the site of the reaction from the surface of the
protein enhances reactivity simply through steric relief and
increased solvent accessibility. However, we cannot rule out
other subtle changes in the complex chemical environment of
the protein that may account for this difference in reactivity.
For Sac- and Seac-containing proteins, allyl ethers make good
metathesis partners because they undergo slow self-metathesis
compared to allyl sulfides, allowing sufficient amount of
unsubstituted alkene for productive CM. However, allyl ethers
are not the only effective metathesis partners as other olefins
such as hexenyl glucoside 10 can also be reactive in CM.
Importantly, the metathesis partner must not form a stable
chelate. If this occurs, the metathesis rate drops and little or no
Conclusions
In summary, by examining a range of sterically and electroni-
cally diverse olefin substrates in protein CM, we have gleaned
some guiding principles for successful CM on model protein
substrates (Scheme 8). In general, allyl sulfides or allyl selenides
extended from the protein surface at the site of modification
are desirable. The role of this olefin partner is distinct since
there is minimal risk of protein self-metathesis.25 We suspect
(25) No protein self-metathesis was observed by SDS-PAGE gel analysis
(see ESI).
9
16810 J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010

Olefin Cross-Metathesis on Proteins
A R T I C L E S
protein modification is observed. Ideally, for fast protein CM
the metathesis partner should also have an unhindered and
nonchelating alkene tether. These results are summarized in
Scheme 8.
were joined efficiently by virtue of the innate affinity of allyl
sulfides and allyl selenides for ruthenium, an affinity that
orchestrates rapid, productive metathesis of the alkene and
alkylidene.
Throughout the course of our investigation, we have dem-
onstrated that by relieving steric hindrance around the alkene
and protein surface, the rate of cross-metathesis is increased.
However, the steric-sensitive nature of olefin metathesis also
means that modification at more hindered protein sites is a
current, unmet challenge with conventional metathesis catalysts.
This limitation prompts the need for a new class of metathesis
catalysts for bioconjugation, where the ligand binding to the
metal should be both small and water-solublesa significant
challenge given that the sterically encumbered NHC ligands
impart the stability necessary for use in air and water. Addition-
ally, allyl selenides were discovered to be superior to allyl
sulfides in aqueous CM. For unhindered allyl selenide-containing
proteins, efficient CM was achieved with several substrates
including carbohydrates, oligo(ethylene glycols), allyl aceta-
mides, and even alkenes with electron-withdrawing ammonium
salts. It is also worth noting that, at the time of our first report,7
examples of homogeneous cross-metathesis in water were
largely limited to simple alkenols.4,5 In this report, a new
benchmark in substrate complexity is set for olefin cross-
metathesis. Complex macromolecules and metathesis partners
The promising results from CM of allyl selenides are driving
our effort in developing chemical and genetic strategies for
incorporation of Seac and other allyl selenide derivatives on
protein surfaces. We are also investigating further the directing
effect of chalcogens in olefin metathesis. We anticipate the
application of these concepts and techniques in bioconjugation,
and synthetic chemistry will allow a largely untapped potential
of allyl sulfides and allyl selenides to be realized.
Acknowledgment. We thank our sources of generous financial
support: EPSRC (Y.A.L.); Rhodes Trust and the National Science
Foundation (J.M.C). B.G.D. is a Royal SocietysWolfson Research
Merit Award recipient and is supported by an EPSRC LSI platform
grant.
Supporting Information Available: Full experimental proce-
1
dures, including H and 13C NMR spectra for all novel com-
pounds and ESI-MS for all protein samples. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA104994D
9
J. AM. CHEM. SOC. VOL. 132, NO. 47, 2010 16811