© 2012 Macmillan Publishers Limited. All rights reserved.

Combinatorial evolution of site- andenantioselective catalysts for polyene epoxidationPhillip A. Lichtor and Scott J. Miller*

Selectivity in the catalytic functionalization of complex molecules is a major challenge in chemical synthesis. The problemis magnified when there are several possible stereochemical outcomes and when similar functional groups occurrepeatedly within the same molecule. Selective polyene oxidation provides an archetypical example of this challenge.Historically, enzymatic catalysis has provided the only precedents. Although non-enzymatic catalysts that meet some ofthese challenges became known, a comprehensive solution has remained elusive. Here, we describe low molecular weightpeptide-based catalysts, discovered through a combinatorial synthesis and screening protocol, that exhibit site- andenantioselective oxidation of certain positions of various isoprenols. This diversity-based approach, which exhibits featuresreminiscent of the directed evolution of enzymes, delivers catalysts that compare favourably to the state-of-the-art for theasymmetric oxidation of these compounds. Moreover, the approach culminated in catalysts that exhibit alternative-siteselectivity in comparison to oxidation catalysts previously described.

Analysis of biosynthetic pathways reveals that functional groupselectivity in enzymatic reactions is a persistent issue. Forexample, olefins may undergo oxidation in favour of C–H

bonds, or instead of alcohols, and different enzymes are engagedto reverse the order of events. Similarly, the derivation of strategiesfor the chemical synthesis of complex structures requires constantconsideration of selectivity. The complexity of the situation is ampli-fied when identical functional groups occur repeatedly within thesame molecule. Now, for example, catalysts may be called on tooxidize a singular olefin in the presence of others, and different cat-alysts may be needed to target alternative sites. Even so, selectivityfor one of the multiple sites represents but one aspect of the chemi-cal challenge, because many functional groups may also react inways that yield new isomers, which adds stereochemical diversityto the positional selectivity outcomes. Thus, the selective oxidationof polyenes presents a profound challenge.

For many years, the selective functionalization of similarly reac-tive olefins within polyenes was observed primarily through thestudy of enzymes1. Today, small-molecule catalysts are known thatalso meet some of these challenges, yet catalyst types that deliversubstantially different site selectivity are not well established, andneither are generalizable protocols for their discovery. Solutions tothese problems provided by venerable catalysts enable allylic olefinepoxidation2–5, oxidation of subterminal olefins of polyenes6,7 andselective oxidation of conjugated systems8,9. Some strategiesprovide access to internal epoxides of polyenes using stoichiometricauxiliary-based ‘template-directed’ strategies10–12. In this article, wedescribe catalysts based on low molecular weight peptides13,14, dis-covered through a combinatorial synthesis and screening protocol,that exhibit site- and enantioselective oxidation of various isopre-nols. The diversity-based approach15,16 delivers catalysts thatcompare favourably to the state-of-the-art for the asymmetric oxi-dation of certain isoprenols, such as geraniol, farnesol and geranyl-geraniol. Moreover, the approach leads to catalysts that offeralternative site selectivities in ways that are not known for oxidationcatalysts described until now.

Previously, we developed a catalytic cycle that employed an aspar-tic acid-based carboxylic acid and applied this methodology to the

asymmetric oxidation of allylic urethanes17 and indoles18. Duringthe course of the catalytic cycle, the carboxylic acid side chain ofaspartic acid is activated with a carbodiimide, which generates a tran-sient peracid on reaction with hydrogen peroxide (Fig. 1a). Catalyticturnover of this cycle is enhanced through the addition of commonnucleophilic co-catalysts such as 1-hydroxybenzotriazole (HOBt)and N,N-dimethylaminopyridine (DMAP), an approach distinctfrom enzymatic ‘epoxidases’19, which often exploit cofactors. Whenapplied to farnesol (1), this catalytic cycle, which employs propionicacid (2) as the catalyst, furnishes the same product selectivity as thetextbook-level reagent meta-chloroperbenzoic acid (m-CPBA),which is used as a stoichiometric reagent (Fig. 1b).

The work began with an assessment of peptide-embeddedaspartates to realize the site-selective oxidation of farnesol (1).Could catalysts based on aspartic acid be found that wouldoxidize farnesol with high levels of site selectivity and ultimatelyenantiocontrol? Could catalysts be found that would functionalizethe allylic alkene20 or other olefins of 1? Oxidation of 1 with (þ)-or (2)-DIPT/Ti(O-i-Pr)4 (DIPT¼ diisopropyl tartrate) under theconditions of Sharpless delivers 3 with excellent selectivity(82–95% enantiomeric excess (e.e.), Fig. 1c)20–26. However, welacked a clear structure-based hypothesis for how peptide-basedcatalysts might achieve our goals. Therefore, we applied anon-bead screening protocol to discover peptide oxidation catalysts27

for the selective oxidation of 1 (Fig. 1d).

Results and discussionOur approach employs the one-bead-one-compound (OBOC)approach to library design28, which capitalizes on split-pool combi-natorial synthesis29,30. This method also allows for an evolutionarydevelopment of catalyst optimization31, in analogy to the strategyemployed in the directed evolution of enzymes31–34.

For our initial screening, we made a number of strategicdecisions. For example, we elected to run reactions to a low conver-sion so as to allow a preliminary assessment of catalyst krel values. Wedefine krel as the relative rate of epoxidation at each olefinic site of 1(2,3-position to give 3; 6,7-position, 4; 10,11-position, 5). It is alsoappreciated that reaction conversion (that is, the extent of overoxidation)

Department of Chemistry, Yale University, PO Box 208107, New Haven, Connecticut 06520-8107, USA. *e-mail: scott.miller@yale.edu

ARTICLESPUBLISHED ONLINE: 14 OCTOBER 2012 | DOI: 10.1038/NCHEM.1469

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry 1

© 2012 Macmillan Publishers Limited. All rights reserved.

can influence product distributions and muddle the extraction ofmeaningful krel comparisons among catalysts35. Thus, we plannedto validate the best ‘hit’ catalysts at a later stage under syntheticallymeaningful conditions, with sufficient stoichiometry of the terminaloxidant to achieve good yields of products (see below).

We began with the evaluation of a small library of resin-boundpeptide catalysts we had synthesized in a parallel fashion, as wellas a diverse split-and-pool library for the oxidation of 1. (Thedata from the initial screen are given in the SupplementaryInformation.) Both the parallel synthesis library and manymembers of the unbiased split-pool library displayed some selectiv-ity, mostly towards the 2,3-position of 1 (Table 1, selected results).These observations were auspicious, given that many catalystsexhibited significant differences from the product distributionobserved with catalytic propionic acid (5:4:3¼ 2.6:2.2:1.0, Fig. 1b).From the initial screening, catalyst 6 displayed the mostsignificant deviation from the distribution afforded by propionicacid (5:4:3¼ 1.0:1.0:3.9, Table 1, entry 1).

To enhance the selectivity for 3, a small OBOC combinatoriallibrary was synthesized and biased towards peptide 6, with four vari-able positions adjacent to the catalytic aspartic acid (Fig. 2a). To biasthe library, at each variable position approximately 50% of the func-tionalized polystyrene beads were coupled to the parent residue(which corresponds to catalyst 6) at each position, and the remain-ing 50% of the beads were distributed evenly for coupling to eitherseven different residues (positions iþ 1 and iþ 2) or six differentresidues (positions iþ 3 and iþ 4, Fig. 2). The resulting library pos-sesses a theoretical size of about 3,000 unique peptide sequences

(n¼ 7 × 7 × 8 × 8, see the Supplementary Information). On com-pletion, the theoretical distribution of peptides in this libraryfollows the histogram shown in Fig. 2b: 6% of beads contain theparent sequence (6), 25% contain a single substitution, 38%contain two substitutions and so on.

Approximately 228 beads from this library were tested, includingcatalyst 7, which favoured epoxide 3 with an approximate doublingof selectivity (5:4:3¼ 1.3:1.0:8.2, Table 1, entry 2). This result rep-resented a further departure of the performance of a peptide-based catalyst from that of the control catalyst 2, data that stimulatedthe evaluation of a second biased library.

For the new OBOC library, now biased towards 7, we fixed theiþ 1 residue as Pro and iþ 2 position as Asn(Trt), where Trtindicates trityl, with the hypothesis that each was important forselectivity. Then the identities of three adjacent residues (iþ 3,iþ 4 and iþ 5) were varied. Biased library 2 (Table 1, entries 3and 4) produced catalysts 8a and 9a, which exhibited an evengreater site selectivity, again nearly double the preference for 3.Catalyst 8a demonstrated a 5:4:3 ratio of 1.2:1.0:18.0. Catalyst 9aproduced a ratio of 1.2:1.0:14.6. In contrast, the truncated analogue10 exhibited a reduced selectivity of 1.3:1.0:5.0 (entry 5).

Next we assessed the performance of the catalysts on resynth-esis and evaluation in solution under synthetically meaningfulconditions (for example, 1.0 theoretical equiv. oxidant, controlledby the stoichiometry of N,N ′-diisopropylcarbodiimide (DIC)).As discussed below, we observed that catalysts provide betterresults when they are dissolved in solution, as opposed to whenthey are immobilized on beads. Perhaps of greater importance

R3HNNHR4

O

OH

O

R3HNNHR4

O

O

O

O

H

R3HNNHR4

O

O

O

NHR5

NR5

R2R1

O

H2O2

R5N

C

NR5

a b

2†

1*

†*

Relativeconversion‡

65%

67%

2,3-

1.0

1.0

Entry10,11-

2.7

2.6

2.2

2.2

6,7-Oxidant/acid

0.1 equiv.

OH

CH3CH3

H3C

CH3

10,11 6,7 2,3

Farnesol epoxides

Conditions

Monoepoxide ratio‡

Farnesol (1)

O

O

O

H

Cl

H3C

O

OH

OH

CH3CH3

H3C

CH3

OH

CH3CH3

H3C

CH3

OH

CH3CH3

H3C

CH3

O

O

O

2,3-epoxyfarnesol (3)

6,7-epoxyfarnesol (4)

10,11-epoxyfarnesol (5)

2

R1

R2

OH

CH3CH3

H3C

CH3Goal

Peptide A

Peptide B

Peptide C

This work found peptide A and peptide B

d

Ti-(O-i-Pr)44 Å molecular sieves

t-BuOOH–50 to –20 ºC

Six reports (1993–2010)82%–95% e.e.

(references 21–26)OH

CH3CH3

H3C

CH3

OH

CH3CH3

H3C

CH3

O

HO OH

CO2i-Pri-PrO2Cc

Figure 1 | Precedent and goal of catalytic oxidation of farnesol. a, Catalytic cycle of aspartic acid-mediated epoxidations. b, Site selectivity of m-CPBA and

epoxidation based on propionic acid. c, Precedent of the Sharpless epoxidation of 1. d, The goal of the present study. *Entry 1, 10 mol% acid, HOBt

(10 mol%), DMAP (10 mol%), 1.0 equiv. DIC, 2.0 equiv. H2O2. †Entry 2, 1.0 equiv. m-CPBA, Na2HPO4 (2.0 equiv.), DCM, H2O. ‡Determined by uncalibrated

GC integrations (see Supplementary Information for details).

ARTICLES NATURE CHEMISTRY DOI: 10.1038/NCHEM.1469

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry2

© 2012 Macmillan Publishers Limited. All rights reserved.

was our observation of significant enantioselectivity in theseexperiments. We had hypothesized that catalyst-dependent siteselectivity might also be coupled to the observation of enantio-selectivity for a given monoepoxide isomer. The notion thatthe same intrinsic transition-state organization associated with aspecific site selectivity might also translate into the organizationrequired for enantioselectivity finds analogy in our previous rate-based selections for catalysts that also turned out to be enantiose-lective29. With these considerations in mind, we synthesized 8aand 9a as their corresponding C-terminal methyl esters, 8b and9b. When examined under conditions designed to lead to full con-version to monoepoxide (1.0 theoretical equiv. oxidant (1.0 equiv.DIC, 2.0 equiv. H2O2)), methyl esters 8b and 9b both providedexcellent overall selectivity. Catalyst 8b delivered 3 with essentiallytotal site selectivity and in 82% e.e. (Table 2, entry 1a). Catalyst 9balso exhibited high site selectivity, and 3 showed a slightlyimproved 86% e.e. (entry 1b). Strikingly, these peptide-basedcatalysts exhibited site selectivities and enantioselectivities thatcompare favourably to the Sharpless catalyst for the oxidationof 1 (Fig. 1c).

Given the analogy, we also evaluated catalyst 9b for the asym-metric epoxidation of other terpenes to assess its generalitytowards allylic alcohols. Peptide 9b maintained excellent site selec-tivity for geraniol (.100:1 site selectivity, 87% e.e., Table 2, entry 2).Nerol showed somewhat better enantioselectivity (entry 3, 93% e.e.)and retained site selectivity for the 2,3-cis-allylic position. Prenolalso gave a favourable result, with the corresponding epoxide

formed with 92% e.e. (entry 4). Catalyst 9b showed no selectivityfor the 2,3-olefin of farnesyl methyl ether, which supports ahydroxyl-directed mechanism.

The initial evaluation of the libraries also revealed catalyst 11,which exhibited a surprising, albeit modest, selectivity for the 6,7-position of 1 to favour 4 (Table 1, entry 6; 5:4:3¼ 1.5:1.9:1.0).Catalyst 11, with D-Pro in the iþ 1 position, was unique in that pre-viously we had not observed even this modest level of alternativeregioselectivity. We therefore prepared a new library, biasedtowards the structure of 11, wherein the iþ 1 D-Pro was fixed andthe adjacent three residues were varied. Therein, we found catalyst12a, which displayed an improved selectivity for 4 (Table 1, entry7; 5:4:3¼ 1.5:2.9:1.0). When resynthesized as the C-terminalmethyl ester 12b and evaluated in CHCl3, this catalyst yielded aproduct distribution of up to 1.0:4.1:1.3 (5:4:3). Next, we evaluateda second-generation library to target 6,7-selective catalysts deliber-ately. Intriguingly, this study did not reveal catalysts that exhibiteddramatically improved selectivity for 4, although several comparablecatalysts were found (see Supplementary Table S1). The sequencesof the peptides unveiled by this library prompted us to prepare ana-logues of catalyst 12a in a hypothesis-driven manner. We projectedthat a C-terminal amide might better resemble the catalysts in theiron-bead screening form (wherein peptides are linked to the resin asthe amide). Catalysts 12c (with a C-terminal n-Bu amide) and 12d(with C-terminal Gly–OMe) were examined. Although both cata-lysts favoured production of epoxide 4, catalyst 12d exhibitedgreater selectivity (5:4:3¼ 1.2:8.0:1.0) under optimized conditions

Table 1 | Results from on-bead screening.

Resin-bound peptideHOBt, DMAP, DIC (0.3 equiv.)

H2O2 (1 equiv.)DCM

Run to low conversions

OH

CH3CH3

H3C

CH3

10,11 6,7 2,3

Farnesol (1) OH

CH3CH3

H3C

CH3

O

OH

CH3CH3

H3C

CH3

OH

CH3CH3

H3C

CH3

O

O

3

4

5

Entry i i 1 1 i 1 2 i 1 3 i 1 4 i 1 5 Catalyst Approximateconversion (%)

10,11-(5)

6,7-(4)

2,3-(3)

Standarddeviation*

Library

1 Boc-Asp Pro D-Thr(O-t-Bu)

Asn(Trt) Leu - 6 13 1.0 1.0 3.9 0.1 Initialscreening

2 Boc-Asp Pro Asn(Trt) D-Phe Thr(OBn) - 7 13 1.3 1.0 8.2 0.1 Directedlibrary No. 1

3 Boc-Asp Pro Asn(Trt) D-Phe Thr(OBn) Asn(Trt) 8a 15 1.2 1.0 18.0 0.5 Directedlibrary No. 24 Boc-Asp Pro Asn(Trt) D-Phe Pro Asn(Trt) 9a 12 1.2 1.0 14.6 0.7

5 Boc-Asp Pro Asn(Trt) D-Phe - - 10 14 1.3 1.0 5.0 0.06 Boc-Asp D-Pro Thr(OBn) Leu - - 11 14 1.5 1.9 1.0 0.0 Truncated

directedlibrary No. 2

7 Boc-Asp D-Pro Thr(OBn) Asn(Trt) Tyr(O-t-Bu)

- 12a 16 1.5 2.9 1.0 0.0 Directedlibrary No. 3

Variations of selective peptide catalysts

N

O

NH

BocHNO

O

OH

HN N

O

NHTrt

O

O

NH

O

R

OTrtHN

O

N

O

NH

BocHNO

O

OH

HN

O

NHTrt

O

O

NH

O

HN

O

R

OBnH3C

O

NHTrt

NH

OtBu

O

RHN

O

NHTrt

O

NH

H3C OBn

ON

O

BocHN

O

OH

O

8a, R = NH-resin8b, R = OCH3

9a, R = NH-resin9b, R = OCH3

12a, R = NH-resin12b, R = OCH312c, R = NH-n-Bu12d, R = NH-Gly-OCH3

*Reactions for 2,3-selective catalysts were run in triplicate from trials run at the same time with the same batch of reagents, and conversions and monoepoxide ratios were estimated by GC (see theSupplementary Information for details).

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1469 ARTICLES

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry 3

© 2012 Macmillan Publishers Limited. All rights reserved.

(Fig. 3a,b). Furthermore, this product ratio simplified an otherwisedifficult separation of products by chromatography as it allowed theisolation of 4 in 43% yield as a .94:6 mixture of 4:5 isomers.

The appreciable, catalyst-dependent selectivity for the 6,7-positionof farnesol is quite unusual. As in the case of the 2,3-selective cata-lysts, experiments with farnesyl methyl ether exhibited little to noselectivity (data not shown). Thus, the observed 6,7-selectivity maybe largely the result of a hydroxyl-directed mechanism, despite thedistance of the hydroxyl, which is six bonds away from the reactingolefin. Yet, unlike the results obtained with the best 2,3-selective cat-alysts, the production of 4 through the action of catalyst 12d occurredwith only a modest, but measurable, 10% e.e. The observation thatboth high site selectivity and enantioselectivity can track togetherqualitatively with 2,3-selective catalysts, but do not appear as tightlycoupled with 6,7-selective catalysts, is intriguing. This may resultfrom the higher number of rotatable bonds between the directinghydroxyl group and the preferred site of oxidation. In our screeningstudies (in corresponding non-optimized experiments) we observedcatalysts that gave a somewhat lower site selectivity (SupplementaryTable S2, 5:4:3, �1:3–4:1), but slightly higher enantioselectivity(�20–30% e.e.). As our data sets are not large, this concept requiresfurther study for a full assessment. Even so, to our knowledge theobserved 6,7-site selectivity exhibited by catalyst 12d greatly contrastswith the product distribution induced by m-CPBA and that of anyother reported catalyst (Fig. 3b).

That catalyst 12d shows a preference for the 6,7-position of 1stimulated us to assess it for the oxidation of a more complexpolyene, geranylgeraniol (13), which presents four competingolefins for functionalization. Figure 3 shows the raw data for the oxi-dation of 1 (Fig. 3b) and 13 (Fig. 3c) with m-CPBA and with thecatalysts described above. With geranylgeraniol (13, Fig. 3c), theresults are quite striking. With m-CPBA, the four monoxides wereformed essentially without control. Catalyst 9b, as with the oxidationof 1, exhibited a high site selectivity to give the 2,3-monoxideof geranylgeraniol (14) in 85% yield, 86% e.e. Catalyst 12d againexhibited a striking site selectivity for an internal alkene, whichwe assigned as the 6,7-monoxide 15. The monoepoxide 14 andthe other monoepoxides (that is, the 10,11- and 14,15-epoxides,see the Supplementary Information) were also produced in thereaction, but as minor products. To our knowledge, the catalyticsite-selective epoxidation of geranylgeraniol (or farnesol) at a

Table 2 | Selectivity of peptide 9b with different linear terpenoid alcohols.

Peptide (10 mol%)HOBt, DMAP, DIC (1 equiv.)

H2O2 (2 equiv.)DCM, 4 ºC, 7 h

OH

R1

R2

R1

R2 OHO

Site selectivity

Entry Substrate Catalyst 10,11- 6,7- 2,3- Yield (%) e.e. (%) Product*

1a

OH

CH3CH3

H3C

CH3

1

8b 1.0 1.0 .100 n.d. 82

OH

CH3CH3

H3C

CH3

O31b 9b 1.0 1.0 .100 81 86

2CH3CH3

H3 OC H

9b 1.0 .100 80 87CH3CH3

H3 OC HO

3CH3

H3C

CH3

OH

9b 1.0 .100 79 93CH3

H3C

CH3

OH

O

4CH3

H3C OH

9b 75† 92CH3

H3C OHO

*See Supplementary Information for details of absolute configuration assignment. †GC yield.

Number of residues different from parent

~50%

~50%

Initial hit: 6Split-and-poollibrary synthesis

Bias

Per

cent

of l

ibra

ry

a

b 40

30

20

10

00 1 2 3 4

Boc-Asp-Pro-D-Thr(O-t-Bu)-Asn(Trt)-Leu-

Boc-Asp-Xaa-Xaa-Xaa-Xaa-linker-

First generation library

Ile

Asn(Trt)

Tyr(O-t-Bu)

D-Phe

Thr(OBn)

D-Val

Pro

i + 2

D-Thr(O-t-Bu)

Pro

Tyr(O-t-Bu)

D-Phe

Thr(OBn)

D-Thr(O-t-Bu)

Val

i + 3

Asn(Trt)

Phe

Tyr(O-t-Bu)

Lys(Ac)

D-Ala

Thr(OBn)

Val

i + 4

Leu

i + 1

Pro

Hyp(O-t-Bu)

Asn(Trt)

Lys(Ac)

Tyr(O-t-Bu)

D-Phe

Thr(OBn)

D-Val

Figure 2 | Catalyst library design by split-and-pool synthesis via the OBOC

library method. a, Design of the first directed library. b, Histogram of

theoretical library composition.

ARTICLES NATURE CHEMISTRY DOI: 10.1038/NCHEM.1469

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry4

© 2012 Macmillan Publishers Limited. All rights reserved.

position other than the 2,3-olefin has not been reported previously.The challenge of the task is manifest in the nearly equivalent reac-tivity of the 14,15-, 10,11- and 6,7-olefins. Thus, the observationsshown in Fig. 3 may represent an important step forwards,enabled by the iterative peptide-based catalyst screening approach.

ConclusionsIn summary, we discovered different peptide catalysts capable offurnishing different isomers of epoxyfarnesol and related com-pounds that are minor products of generic peracid-based olefinepoxidation. Peptide 9b appears to operate via a hydroxyl-directingmechanism, analogous to the Sharpless asymmetric epoxidationand with comparable selectivity. Peptide 12d provides unprece-dented selectivity for a catalyst for the internal olefin of farnesol,with substantial site selectivity but modest enantioselectivity.Discovery of catalysts for complex selectivity situations, wherein amyriad of possible products may be formed as a function of a cata-lytic mechanism of bond formation, is a state-of-the-art challengefor asymmetric catalysis endeavours. The application of diversity-based approaches, such as that described herein, may prove fruitful,and may also offer some analogy to the directed evolution strategiesemployed by natural and engineered enzymatic systems.

MethodsGeneral procedure for the epoxidation of allylic alcohols with peptide 9b. To atest tube with a stir bar and Teflon-lined screw cap was added peptide 9b (0.1 equiv.),allylic alcohol (1.0 equiv.), a freshly prepared/sonicated solution that containedHOBt.H2O (0.1 equiv.) and DMAP (0.1 equiv.) in dichloromethane (DCM, to aconcentration of 0.2 M of substrate), and then 30% aqueous H2O2 (2.0 equiv.). Thetest tube was placed in ice (if running the reaction below room temperature) andallowed to chill before adding DIC (1.0 equiv.). The reaction tube was then sealedwith a screw cap under ambient conditions (without the exclusion of air) and

brought to a cold room (4 8C), where the reaction was stirred vigorously. Sevenhours after the addition of DIC, the reaction was quenched with a saturated aqueoussolution of Na2SO3 and stirred for a few moments, sitting in ice, before it was allowedto warm to room temperature. Saturated aqueous NaHCO3 and hexane or EtOAcwere added, the mixture was vortexed, allowed to settle and then the organic layerwas removed, followed by 2–3 additional hexane or EtOAc extracts, performedsimilarly. If the site selectivity was to be determined, an aliquot of the combinedorganics was removed, diluted with additional hexanes and analysed by gaschromatography (GC). The GC sample was returned to the organics, which werethen concentrated in vacuo or under a stream of N2 and purified by flash silica-gelcolumn chromatography.

General procedure for the epoxidation of 6,7-olefin in polyisoprenols withpeptide 12d. To a test tube with stir bar and Teflon-lined screw cap was addedpeptide 12d (0.1 equiv.), substrate (1.0 equiv.), about two-thirds of a freshlyprepared/sonicated solution that contained HOBt.H2O (0.1 equiv.) and DMAP (0.1equiv.) in CHCl3 (the CHCl3 was passed through basic alumina prior to mixing withDMAP/HOBt), and then 30% aqueous H2O2 (2.0 equiv.). The test tube was placedin ice and allowed to chill before DIC (1.0 equiv.) was added, followed by theremaining CHCl3 solution (to a concentration of 0.2 M of substrate) rinsed down thesides of the reaction tube. The reaction tube was then sealed with a screw cap underambient conditions (without the exclusion of air) and placed in an i-PrOH bathmaintained at 212 8C to 218 8C by a cryostat. The reaction was then stirredvigorously. About 25–48 hours after the addition of DIC, the reaction was quenchedwith a saturated aqueous solution of Na2SO3 and stirred for a few moments, still at alow temperature, before it was allowed to warm to room temperature. Saturatedaqueous NaHCO3 and EtOAc were added and the mixture was vortexed and allowedto settle, after which the organic layer was removed, along with three additionalEtOAc extracts performed similarly. An aliquot of the combined organics wasremoved, diluted with additional EtOAc and analysed by GC. The organics wereconcentrated in vacuo and then purified by flash column chromatography.

Additional experimental procedures, descriptions of compounds and analyticalmethods are given in the Supplementary Information.

Received 24 May 2012; accepted 28 August 2012;published online 14 October 2012

OH

CH3CH3CH3CH3

H3COH

CH3CH3

H3C

CH3

OH

CH3CH3CH3CH3

H3COH

CH3CH3

H3C

CH3

HOBt (10 mol%), DMAP (10 mol%)1.1 equiv. DIC, 1.5 equiv. H2O2,

CHCl3, –12 ºC to –18 ºC

OO

4, 43% yield 15, 42% yield

Farnesol (1) Geranylgeraniol (13)

OH

CH3CH3CH3CH3

H3CO

1485% yield86% e.e.

Peptide 9bConditions in Table 2

m-CPBA

12d

9b

Selectivity2,3-

1.0

>100

1.0

6,7-

2.2

1.0

48 50 52 54 min 50 52.5 55 57.5 min

8.2

10,11-

2.7

1.0

1.2

Catalyst oroxidant

Selectivity

1.0

>100

1.5*

2,3-

3.1

1.0

1.0

14,15-

3.1

1.0

1.8

10,11-

2.5

1.0

13.2

6,7-

13

1415

GC trace

m-CPBA

12d

9b

Catalyst oroxidant

Peptide 12d (10 mol%)

1

3

54

GC trace

a

b c

Figure 3 | Results and raw data for the site-selective oxidation of farnesol and geranylgeraniol with m-CPBA, 2,3-selective catalyst 9b and 6,7-selective

catalyst 12d. a, Optimized oxidation conditions with 12d. b,c, Comparison of product ratios and GC spectra from crude reaction mixtures with m-CPBA,

catalyst 9b and catalyst 12d, with the relevant area of GC spectra for farnesol (b) and geranylgeraniol (c). *The peak has a shoulder that is integrated

and so the area is overestimated. Conditions for m-CPBA reactions are given in Fig. 1b, for reactions with catalyst 9b in Table 2 and for reactions with

catalyst 12d in (a).

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1469 ARTICLES

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry 5

© 2012 Macmillan Publishers Limited. All rights reserved.

References1. van Tamelen, E. E. & Heys, J. R. Enzymic epoxidation of squalene variants.

J. Am. Chem. Soc. 97, 1252–1253 (1974).2. Katsuki, T. & Martin, V. S. Asymmetric epoxidation of allylic alcohols: the

Katsuki–Sharpless epoxidation reaction. Org. React. 48, 1–299 (1996).3. Zhang, W., Basak, A., Kosugi, Y., Hoshino, Y. & Yamamoto, H. Enantioselective

epoxidation of allylic alcohols by a chiral complex of vanadium: an effectivecontroller system and a rational mechanistic model. Angew. Chem. Int. Ed. 44,4389–4391 (2005).

4. Malkov, A. V., Czemerys, L. & Malyshev, D. A. Vanadium-catalyzed asymmetricepoxidation of allylic alcohols in water. J.Org. Chem. 74, 3350–3355 (2009).

5. Egami, H., Oguma, T. & Katsuki, T. Oxidation catalysis of Nb(salan) complexes:asymmetric epoxidation of allylic alcohols using aqueous hydrogen peroxide.J. Am. Chem. Soc. 132, 5886–5895 (2010).

6. Barlan, A. U., Basak, A. & Yamamoto, H. Enantioselective oxidation of olefinscatalyzed by a chiral bishydroxamic acid complex of molybdenum. Angew.Chem. Int. Ed. 45, 5849–5852 (2006).

7. Corey, E. J. & Zhang, J. Highly effective transition structure designed catalystfor the enantio- and position-selective dihydroxylation of polyisoprenoids.Org. Lett. 3, 3211–3214 (2001).

8. Chang, S., Lee, N. H. & Jacobsen, E. N. Regio- and enantioselective catalyticepoxidation of conjugated polyenes. Formal synthesis of LTA4 methyl ester.J.Org. Chem. 58, 6939–6941 (1993).

9. Burke, C. P. & Shi, Y. Regio- and enantioselective epoxidation of dienes by achiral dioxirane: synthesis of optically active vinyl cis-epoxides. Angew. Chem.Int. Ed. 45, 4475–4478 (2006).

10. Breslow, R. & Maresca, L. M. Template-directed epoxidation of farnesol andgeranylgeraniol as conformational probes. Tetrahedron Lett. 10, 887–890 (1978).

11. Saito, I., Mano, T., Nagata, R. & Matsuura, T. Inter- and intramolecularepoxidation utilizing silyl-protected peroxy esters and copper salt.Tetrahedron Lett. 28, 1909–1912 (1987).

12. Gnanadesikan, V. & Corey, E. J. A strategy for position-selective epoxidation ofpolyprenols. J. Am. Chem. Soc. 130, 8089–8093 (2008).

13. Colby Davie, E. A., Mennen, S. M., Xu, Y. & Miller, S. J. Asymmetric catalysismediated by synthetic peptides. Chem. Rev. 107, 5759–5812 (2007).

14. Wennemers, H. Asymmetric catalysis with peptides. Chem. Commun. 47,12036–12041 (2011).

15. Francis, M. B., Jamison, T. F. & Jacobsen, E. N. Combinatorial libraries oftransition-metal complexes, catalysts and materials. Curr. Opin. Chem. Biol. 2,422–428 (1998).

16. Kuntz, K. W., Snapper, M. L. & Hoveyda, A. H. Combinatorial catalyst discovery.Curr. Opin. Chem. Biol. 3, 313–319 (1999).

17. Peris, G., Jakobsche, C. E. & Miller, S. J. Aspartate-catalyzed asymmetricepoxidation reactions. J. Am. Chem. Soc. 129, 8710–8711 (2007).

18. Kolundzic, F., Noshi, M. N., Tjandra, M., Movassaghi, M. & Miller, S. J.Chemoselective and enantioselective oxidation of indoles employing aspartylpeptide catalysts. J. Am. Chem. Soc. 133, 9104–9111 (2011).

19. Thibodeaux, C. J., Chang, W-C. & Liu, H-W. Enzymatic chemistry ofcyclopropane, epoxide and aziridine biosynthesis. Chem. Rev. 112,1681–1709 (2012).

20. Sharpless, K. B. Searching for new reactivity (Nobel Lecture). Angew. Chem.Int. Ed. 41, 2024–2032 (2002).

21. Kotaki, T., Shinada, T. & Kaihara, K. Structure determination of a new juvenilehormone from a Heteropteran insect. Org. Lett. 11, 5234–5237 (2009).

22. Koohang, A. et al. Enantioselective inhibition of squalene synthase by aziridineanalogues of presqualene diphosphate. J. Org. Chem. 75, 4769–4777 (2010).

23. Tanuwidjaja, J., Ng, S-S. & Jamison, T. F. Total synthesis of ent-dioxepandehydrothyrsiferol via a bromonium-initiated epoxide-openingcascade. J. Am. Chem. Soc. 131, 12084–12085 (2009).

24. Uyanik, M., Ishibashi, H., Ishihara, K. & Yamamoto, H. Biomimetic synthesis ofacid-sensitive (–)-caparrapi oxide and (þ)-8-epicaparrapi oxide induced byartificial cyclases. Org. Lett. 7, 1601–1604 (2005).

25. Marshall, J. A. & Hann, R. K. A cascade cyclization route to adjacentbistetrahydrofurans from chiral triepoxyfarnesyl bromides. J. Org. Chem.73, 6753–6757 (2008).

26. Dittmer, D. C. et al. A tellurium transposition route to allylic alcohols:overcoming some limitations of the Sharpless–Katsuki asymmetric epoxidation.J. Org. Chem. 58, 718–731 (1993).

27. Lichtor, P. A. & Miller, S. J. One-bead-one-catalyst approach to aspartic acid-based oxidation catalyst discovery. ACS Comb. Sci. 13, 321–326 (2011).

28. Lam, K. S., Lebl, M. & Krchnak, V. The ‘one-bead-one-compound’combinatorial library method. Chem. Rev. 97, 411–448 (1997).

29. Furka, A., Sebestyen, F., Asgedom, M. & Dibo, G. General method for rapidsynthesis of multicomponent peptide mixtures. Int. J. Pept. Protein Res. 37,487–493 (1991).

30. Lam, K. S. et al. A new type of synthetic peptide library for identifying ligand-binding activity. Nature 354, 82–84 (1991).

31. Singh, J. et al. Application of genetic algorithms to combinatorial synthesis: acomputational approach to lead identification and lead optimization. J. Am.Chem. Soc. 118, 1669–1676 (1996).

32. Reetz, M. T. Laboratory evolution of stereoselective enzymes: a prolific source ofcatalysts for asymmetric reactions. Angew. Chem. Int. Ed. 50, 138–174 (2011).

33. Brustad, E. M. & Arnold, F. H. Optimizing non-natural protein function withdirected evolution. Curr. Opin. Chem. Biol. 15, 201–210 (2011).

34. Copeland, G. T. & Miller, S. J. Selection of enantioselective acyl transfer catalystsfrom a pooled peptide library through a fluorescence-based activity assay: anapproach to kinetic resolution of secondary alcohols of broad substrate scope.J. Am. Chem. Soc. 123, 6496–6502 (2001).

35. Schreiber, S. L., Schreiber, T. S. & Smith, D. B. Reactions that proceed with acombination of enantiotopic group and diastereotopic face selectivity can deliverproducts with very high enantiomeric excess: experimental support of amathematical model. J. Am. Chem. Soc. 109, 1525–1529 (1987).

AcknowledgementsThis work is supported by National Institutes of Health (NIH R01-GM096403) to S.J.M.,and P.A.L. was partially supported by NIH CBI-TG-GM-067543. P.A.L. thanks B. Fowlerfor collaborating to build the library that resulted in catalyst 6, and S. Alexander and theSchepartz laboratory for assistance.

Author contributionsP.A.L. designed and performed the experiments and S.J.M. oversaw the project. Bothauthors analysed data and co-wrote the manuscript.

Additional informationSupplementary information and chemical compound information are available in theonline version of the paper. Reprints and permission information is available online athttp://www.nature.com/reprints. Correspondence and requests for materials should beaddressed to S.J.M.

Competing financial interestsThe authors declare no competing financial interests.

ARTICLES NATURE CHEMISTRY DOI: 10.1038/NCHEM.1469

NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry6