BrooRGANrc cHEMTSTRY 17, 4 l -_52 (1989)

Enzymatic Baeyer-Vi l l iger Type Oxidat ions of KetonesCatalyzed by Cyclohexanone Oxygenase

OssroraNR AnnrL, Cenol C. RyEnsoN, CsRrsropHER Wnr-sH.eNo GEoncE, M. WHrrEsrogsl

Department of Chemistry, Haruard Uniuersity, Camhridge, Massac'husetts 02138, und Departmentof Chemistry, Massachusetts Institute of Technologt,, Cumbridge, Mussachusett.s 02138

Rec'e iued Fehruury 9, 1988; accepted August 31, 1988

Cyclohexanone oxygenase (EC l l4. l3. - ) produced by Acinetobat ' ter NCIB 9871 is af lav in-contain ing NADPH-dependent monooxygenase that ut i l izes dioxygen to convert cy-c l ic ketones into lactones. A var iety of ketones were examined in order to determine thesubstrate speci f ic i ty . regioselect iv i ty , and enant ioselect iv i ty of cyclohexanone oxygenase.Lactones were synthesized using immobi l ized enzymes. The nicot inamide cotactor requiredby cyclohexanone oxygenase was regenerated in . r i l r r r . r , i th g lucose 6-phosphate and glucose-6-phosphate dehydrogenase f rom Laucono.\ to( ' mesenleroi t le.s. . i9g9 Acactemic press. rnc.

INTRODUCTION

Microorganisms are capable of effecting a wide range of oxidative transforma-tions of organic compounds (1). Among these transformations, the biologicalequivalent of the Baeyer-Vi l l iger react ionl (2)- the insert ion of an oxygen atominto organic ketones with the product ion of lactones-has been impl icated inmic rob ia l degradat ions o f a lkanes . s te ro ids . and cyc l i c ke tones (3 -15 ) . The ob-ject ive of th is work was to explore the ut i l i ty of one enzyme-cyclohexanonemonooxygenase (EC 1 .14 .13 . - ) -capab le o f ca ta lyz ing th is t ran format ion fo rut i l i ty in organic synthesis.

The microorganisms that are capable of removing side chains from steroids arealso general ly able to carry out an addi t ional Baeyer-Vi l l iger oxidat ion on the C-17ketones to give lactones. The microbial double Baeyer-Vil l iger oxidation of pro-gesterone by Penicil l ium c'hrysogenum, for example, gives testololactone in 70Vayield (9). The microbiological Baeyer-Vil l iger reactions of steroids are not l imitedto C-17 side chains and D-rings. Eburicoic acid undergoes cleavage of the A-ringby Glomerella fusarioides and yields 4-hydroxy-3,4-seco-eburica-8,24(28)-diene-3,2l-dioic acid in l0% vield (10).

I To whom correspondence should be addressed.2 A Baeyer-Vi l l iger rearrangement is def ined as an oxygen insert ion react ion resul t ing f rom the

treatment of a ketone with a peracid or other peroxy compound.

(x)4-s-2068/u9 1;3.00[ ' o p l , r i g h t t t 1 9 1 { 9 h r A c a d e m i c P r e s s . l n c ,

Al l r ights of reproduct ion in anr lbrm rcservet l .

4 l

42 ABRIL ET AL.

Analogous microbiological Baeyer-Vil l iger oxidations of simple cyclic ketonesto the corresponding lactones have also been reported (l 1- 1 5). The oxidation of 2-heptylcyclopentanone or of 2-pentylcyclopentanone by Pseudomonas oleouoransgives the corresponding lactones in low yield, with some suggestion that stereo-chemical selectivity may be involved (11). Microbial oxidation of fenchone withCorynebacterium species afford a mixture of 1,2-fencholide and 2,3-fencholide in42% yield (12). In addition, cell-free enzyme preparations from pseudomonadspecies catalyze the lactonization of camphor and 2,5-diketo camphane (14, l5).

Key enzymes in these microbiological oxidations are the dioxygenases andmonooxygenases that catalyze the introduction of an oxygen functionality andinit iate the carbon-carbon bond cleavage step. Although the most extensivelyinvestigated monooxygenases are the aromatic hydroxylases, the lactone-formingenzymes are now well established as important catalysts in the degradation of arange of organic compounds. These enzymes are flavin dependent and carry out anet four-electron reduction of dioxygen: one atom of oxygen is incorporated intothe organic substrate; the second appears as water.

Cyclohexanone oxygenase produced by Acinetobacter NCIB 9871, belongs tothe class of bacterial f lavoprotein monooxygenases that catalyze lhe conversion ofcycl ic ketones into the corresponding lactones (16). Speci f ical ly. cyclohexanoneoxygenase catalyzes the oxygen insertion and ring expansion of cyclohexanone toform e-caprolactone,

A\,

CyclohexononeOxygenose

+ NADPH + O^ ---_----------->'

E n z . F A DNADP ' + HZO ( t )

ol l

to/ \ +

Mechanistic studies on cyclohexanone oxygenase support the formation of a4a-hydroperoxyf lavin intermediate (17-19). This react ive peroxide acts as a nu-cleophile toward cyclohexanone to form a mixed peroxide, which then decom-poses by a Baeyer-Vil l iger rearrangement to 4a-hydroxyflavin and e-caprolactone( 17).

We were interested in the potent ia l synthet ic ut i l i ty of cyclohexanone oxy-genase for several reasons. First , cyclohexanone oxygenase displays a broadsubstrate speci f ic i ty for cycl ic ketones (201 . Second. work by Schwab (2 1) demon-strates that the lactonization of 2-methylcyclohexanone catalyzed by cyclohexa-none oxygenase exhibits regioselectivity comparable with that found for thechemical Baeyer-Vil l iger oxidation; moreover, there is evidence that the enzyme-catalyzed reaction proceeds with some enantioselectivity (21).

Thus, we examined a variety of readily available ketones in order to determinethe substrate speci f ic i ty, regioselect iv i ty, and enant ioselect iv i ty of cyclohexanoneoxygenase immobilized in a polyacrylamide gel and to ascertain whether thismethod offered advantages over conventional Baeyer-Vil l iger procedures. Thereduced nicotinamide cofactor NADPH required by cyclohexanone oxygenasewas regenerated in situ with glucose 6-phosphate and glucose-6-phosphate dehy-drogenase from Leuconostoc' mesenteroides (Scheme I).

ENZYMATIC OXIDATIONS OF CATALYZED KETONES 43

o

*A* ' + o"Cyclohexonone O

Oxygenose llnAo + Hzo

/ \ l/ \ R '/ i

NADPH NADP+

H O HOH O

f-"' 'OPHga_Lox

HO\\_'\-c02.

HO

ScHEue I . General scheme for the enzymat ic synthesis of lactones catal l ,zed by cyclohexanone

oxygenase wi th regenerat ion of NADPH using glucose 6-phosphate and glLrcose-6-phosphate deh-v-

drogenase.

RESULTS AND DISCUSSION

Determination of Substrate Specfficity of Cyclohexanone Oxygenose andMeasurement of Relatiue Rates

Table I l ists the relative activity of cyclohexanone oxygenase toward severalketones. These relat ive rates were measured by two methods: by fo l lowing therate of consumption of NADPH spectrophotometrically and by following the rateof oxygen uptake with an oxygen electrode. These measurements should be inter-preted cautiously since all compounds were employed at the same nominal con-centration with no regard to differences in solubil ity or value of K,,,. These assayswere intentionally run under conditions that should saturate the enzyme andtherefore approximate Vru^.

Cyclohexanone oxygenase was found to catalyze the oxidation of NADPH atthe same rate as the reduction of 02, in the presence of the various ketonesexamined. Addition of catalase to the assay system had no effect on the rate ofoxygen uptake. Substrates and substrate analogs often act as effectors withNADPH-l inked monooxygenases (22).Effectors combine with the oxidized en-zyme (F,nz' FAD) and accelerate the rate of f lavin reduct ion by NADPH. In thepresence of d ioxygen, th is process would resul t in an increased rate of consump-tion of dioxygen, with some of the dioxygen being reduced to hydrogen peroxide.That is, oxygen insertion into the ketone moiety would be uncoupled from oxida-

OP

o o

44 ABRIL ET AL.

TABLE I

Substrate Specif ici ty of Cyclohexanone Oxygenaseand Measurement of Relat ive Rates"

CompoundRe la t i ve

rale ( ,'/r I

Cyclc lhexanone2-Norbornanoneo-Fenchonel -Fenchone2-Adamantanone( + )-Dihydrocarvone.r ln-7- B e nzyI ox y meth y I -2-norbone n--5 -one

2-Ace ty l cyc lohexanone2-Cyc lohexene- l -one( + )-ClamphorProgesterone1 .4 -Cyc lohexaned ioneI .3 -Cyc loheraned ione4- Pheny lc yc lohe xanone2-Pheny lcyc lohexanone

4-t t , r l -BuIv lcvclohe xanone

" Rclat i l 'c n l tes were measrrrecl spectrophotometr i -cal ly at -140 nm and'n, ' - d iorvgen uptake at 2-5 'C. Mea-su rements were pe r fb rn ted in I m l o f g l y , c ine -NaOHbuffer ( [ i0 my. pH f l .0) contain ing ] -5 pr .mol of the indi-ca ted ke tone ( i n i t i i r l l l d i sso lved in - s0 p l o f MeOH) .N A D P H ( 0 . 1 6 m r r ) . c y c l o h e x a n o n e o x y g e n a s e ( 0 . 6

U) . and an equ i l i b r i um concen t ra t i c ln o f i i tmospher i cd iooxygen (0 .24 mn) . See F ,xper imen ta l f i r r de ta i l s .

t ion of n icot inamide. In an assay system measur ing oxygen uptakc. the fbrmat ionof hydrogen peroxide would appear as act iv i ty. Since catalase fai led to al ter therate of d ioxygen consumption. we therefore conclude that those ketones thatsigni f icant ly st imulate oxygen uptake are t rue substrates fbr cyclc lhexanone oxy-genase. This observat ion is also consistent wi th the conclusion of Ryerson et a l .(17): d i rect react ion between H:O: and substrate is an unl ikely mechanism for theenzyme-catalyzed conversion of ketones to lactones.

The fol lowing conclusions concerning the structure-react iv i ty relat ionship canbe drawn from the resul ts presented in Table l . First , cyclohexanone oxygenasedispfays a broad substrate speci f ic i ty toward al icyc: l ic 'ketones. Second. the fo l -lowing general types of ketones are nol substrates: d,B-unsaturated ketones and1,3-diketones (acycl ic and al icycl ic) . These observat ions are consistent wi th thef indings of Donoghue er ol . (16). Walsh and Chen (20) have also demonstratedthat the enzyme shows even broader speci f ic i ty. working on some acycl ic ketonesand ary l ke tones .

The racemic ketone .ryn-7-benzvloxvmethvl-2-norbornen--5-one (Corev's ke-

I 00l t t il060

( l

')-l l 0

0U

46tl

900( ,

l 9l lt +

ENZYMATIC OXIDATIONS OF CATALYZED KETONES

tone), an intermediate in the synthesis of prostaglandins (23), was a substrate witha relative rate of 110% (relative to cyclohexanone). A total consumption assay runwith a l imiting amount of the racemic ketone did not exhibit biphasic kineticbehavior, as would have been expected were the K,,, values for each enantiomersignificantly different. Product analysis to conclusively determine the enantiose-lectivity, if any, of the enzyme was not, however, carried out.

Two steroids were investigated as substrates for cyclohexanone oxygenase:androstanolone, which was a substrate with a relative rate of ca.47o. and proges-terone, which was not a substrate.

En4tmatic Syntheses of Lactones

Lactones were synthesized on ca. 30-80 mmol scale from 2-norbornanone (1),r-fenchone (2), o-fenchone (3), (+)-camphor (4), and (+)-dihydrocarvone (5)(Table 2); the react ion sequence is out l ined in Scheme I . The procedure ut i l izescyclohexanone oxygenase and a regeneration system for NADPH that is based onglucose 6-phosphate (G-6-P),3 and glucose-6-phosphate dehydrogenase (G-6-PDH) from L. mesenteroides.In a typical synthesis, the react ion mixture con-t a i n e d k e t o n e ( 0 . 1 m o l ) , G - 6 - P . N a : ( 0 . 1 2 m o l ) , N A D P - ( l m m o l ) . M g c l : ( 6mmol) , EDTA (1 mmol) , and the enzymes immobi l ized in PAN: cyclohexanoneoxygenase and G-6-PDH. Although the pH optimum for cyclohexanone oxy-genase is 9.0 (16), the syntheses of lactones were carr ied out at pH 8.0 in order tominimize the extent of epimerization of the ketone and the lactone. At this valueof pH the enzyme has ca. 707o of its maximum activity.

The react ions were al lowed to run for 5-10 days: GLC analysis showed com-plete conversion of ketone to product. After removal of the polyacrylamide gelscontaining the enzymes, the product was isolated by extraction.

The enzyme-catalyzed oxidation of 2-norbornanone (0.1 mol) afforded, after 5days, 2-oxabicyclo[3.2.1]octan-3-one (6) in 8l% yield rH NMR and GC/MS daraof the enzymatic product were indistinguishable from authentic lactone obtainedby oxidation of 2-norbornanone with m-chloroperbenzoic acid (24). Enzymaticoxidation of l-fenchone (0.1 mol) afforded a mixture of 1,2- and 2,3-fencholides in76%yield The rat io of 1,2- and2,3- lactones was 88:12.Enzymatic oxidat ion ofo-fenchone (0.1 mol) afforded the mixture of fencholides in essentially the sameproportions, in 79% yield: rH NMR spectra of the product mixtures obtained wereindist inguishable. The spectra exhibi ted the two signals at 6 l . l9 and A 1.24 due tothe gem-dimethyl and a single signal at 6 1.47 due to the shielded angular methylgroup of the 1,2-fenchol ide. Two weak signals at 6 1.35 and 6 1.38 were alsopresent corresponding to the gem-dimethyl group of the minor component, 2,3-fencholide. The conversion of (+)-camphor to its corresponding lactone pro-ceeded in 89% yield. 6-Isopropenyl-3-methyl-2-oxaheptanone (10) was obtainedin75% yield from (+)-dihydrocarvone. These products were identif ied by compar-ison with authentic materials obtained by oxidation of the keton es with 40Voperacetic acid (25).

I Abbreviat ions used: G-6-P, g lucose 6-phosphatel G-6-PDH. glucose-6-phosphate t lehydrogenase;PAN, poly(acry lamide-N-acry loxysuccin imide ) .

45

46 ABRIL ET AL.

TABLE 2

Summary of the Large-Scale Enzymatic Syntheses of Lactones

Reactant (mmol) Products

React iont ime

(days )

Yie ld

mmol %

8 rt i l

o2-Norbornanone ( l )

l -Fenchone (2)

( t00)

2-Oxab icyc lo [3 .2 . I ]oc tan-3-one (6 )

1 .2 -Fencho l ide (7 )

1 .2 -Fencho l idc (7 ) l . . l -Fencho l ide (8 )o-Fenchone (3 )

4-5

o2-Oxab icyc lo [3 . 2 . I I I .7 .7 - t r imethy loc tan- - ] -one (9 )

( + ) -D ihydrocarvone (5 ) 6 - lsopropeny l -3 -meth-v l -2 -oxac l c loheptanone ( l0 )

Regeneration of IVADPH

The procedure for regenerat ion of the reduced nicot inamide is based on thcdehydrogenation of G-6-P catalyzed by G-6-PDH from L. mesenteroitle.r (26).

This method has several advantages. First , G-6-PDH is commercial ly avai lable,inexpensive, and easy to manipulate and immobi l ize. Second. because G-6-PDHfrom L. mesenteroides has no essent ia l th io l groups. th is enzyme is relat ivelystable under aerobic condi t ions and can be used for resenerat ion of reduced

7 sl0

o2 . 1 - F e n c h o l i d e ( 8 )

(+)-Camphor (4)

ENZYMATIC OXIDATIONS OF CATALYZED KETONES

nicotinamide cofactors in dioxygen-containing environments; dioxygen is a sub-strate of cyclohexanone oxygenase. Third, the equilibrium constants for reducednicotinamide formation are high, because the reaction is rendered essentiallyirreversible by hydrolysis of the initially formed 6-phosphogluconolactone to 6-phosphogluconate. The disadvantage of this system is that both G-6-P and 6-PGare general acid catalysts for the hydration of the reduced nicotinamide cofactorsQ7) .4

CONCLUSIONS

This work describes studies concerning the substrate specificity and regioselec-tivity of cyclohexanone oxygenase; this work was carried out in order to deter-mine whether this enzymatic oxidation of ketones offers advantages over conven-tional Baeyer-Villiger procedures.

We successfully synthesized several lactones on a 30-80 mmol scale, in highyields, using immobilized enzymes and in situ cofactor regeneration. These syn-theses should be easily amenable to further scale-up.

Although the data obtained from the substrate specificity experiments show thatcyclohexanone oxygenase accepts a broad range of alicyclic ketones as sub-strates, it does not accept a,B-unsaturated or 1,3-diketones.

The results obtained in the large-scale oxidations of 2-norbornanone, (+)-cam-phor, and (+)-dihydrocarvone parallel those of the chemical Baeyer-Vil l iger oxi-dation: the migratory aptitude of the alkyl groups appears to follow the sequenceof tert-alkyl > sec-alkyl ) prim-alkyl. It is of interest that the enzymatic oxidationof fenchone gives both possible lactones. The proportions of the 7,2- and 2,3-fencholides obtained (88: 12) contrast with the chemical oxidation of fenchonewith peracetic acid: the 2,3-fencholide is the major product in a 40: 60 mixture.Although the similarity in the spatial arrangements of the methyl groups at thetertiary C-1 and C-3 positions would suggest a low degree of regioselectivity, a1,2-attack is favored.

Regarding enantioselectivity, it is clear from the results obtained in the largescale oxidations of racemic 2-norbornanone and o- and l-fenchone that cyclohex-anone oxygenase does not exhibit a useful degree of enantioselectivity: bothenantiomers are turned over by the enzyme. However, recently, Taschner et al.(28) have found that cyclohexanone oxygenase is capable of enantioselectivesynthesis of lactones from mesomeric cyclohexanones.

With respect to regioselectivity, cyclohexanone oxygenase does not seem tooffer major advantages over conventional Baeyer-Vil l iger reaction using peracids.Although, when confronted with a substrate containing a ketone and a sulfidefunctional group in the same molecule, e.g., 3-thia- and 4-thiacyclohexanones, theenzyme catalyzes the Baeyer-Vil l iger oxygenative ring expansion exclusively(29). This is in contrast to the greater ease of sulfur oxidation in nonenzymatic

a The phosphate-catalyzed decomposition of NADPH by G-6-P could probably be avoided by substi-tution of glucose 6-sulfate.

47

48 ABRIL ET AL.

oxygen-transfer chemistry and is an example of functional group differentiation byan enzyme that is different from conventional oxidation reagents.

We conclude that this enzyme has the combination of availability, specificactivity, and stability required to make it a candidate for practical organic synthe-sis on scales up to 0.1-l mol. Its enantioselectivity remains to be establishedcarefully. Its structural specificity will be different than that of peracids and maybe useful in special circumstances.

EXPERIMENTAL

General

Spectrophotometric measurements were performed at 25"C using a Perkin-Elmer Model 552 spectrophotometer equipped with a constant temperature cell.tH NMR spectra were recorded at 250 MHz on a Bruker Model WH-250 instru-ment, using CDCI3 as solvent (internal lock) (kindly performed by Marifaith Hack-ett). Chemical shifts were measured relative to internal tetramethylsilane. Infra-red spectra were recorded on a Perkin-E,lmer Model 598 instrument. GC/MS datawere obtained on a Hewlett-Packard Model 59904 instrument at an ionizingvoltage of 70 eV, using a UC-W98 column (6 ft x 0.125 in.) (kindly performed byRobert DiCosimo). GLC analyses were performed on a Perkin-Elmer Model39208 instrument equipped with a flame ionization detector. using l07c Carbowax(10 f t x 0.125 in.) and l0% UC-W98 (6 f t x 0.12-5 in.) columns. Measurements foroxygen concentrat ion were carr ied out using a Yel low Spr ings oxygen monitor(Yel low Spr ings Instrument Co., Yel low Spr ings. OH).

Materials

Cyclohexanone oxygenase was produced from a culture of Ac'inetobacter NCIB9871 maintained from a culture donated by Dr. P. W. Trudgil l, University Collegeof Wales (Aberystwyrth, UK). Glucose-6-phosphate dehydrogenase from L.mesenteroides and NADP* were purchased from Sigma. Glucose 6-phosphatewas prepared enzymatical ly according to publ ished procedure (J0). t - -Fenchoneand o-fenchone were obtained from K and K Laborator ies. Inc. (Plainview. NY).Al l other organic chemicals were obtained from Aldr ich. The organic compoundswere used without further purif ication.

Assay Methods5

G-6-P, G-6-PDH, and NADP+ were assayed according to the procedure ofBergmeyer (J1) .

Assay of cyclohexanone oxygenase (32). The determination of cyclohexanoneoxygenase activity is based on the decrease in optical density at 340 nm whichresults from the conversion of NADPH to NADP. The assay mixture (final vol-

5 Assays were carr ied out at 25'C. One uni t of enzymat ic act iv i ty is def ined as that amount ofenzyme which catalyzes the format ion of I pmol of product per minute at 2-5 'C.

ENZYMATIC OXIDATIONS OF CATALYZED KETONES

ume,3 ml) contained glycine-NaOH buffer (80 mu, pH 8.0), NADPH (0.16 mM),cyclohexanone (48.5 pM), cyclohexanone oxygenase (0.01-0.1 prra; concentrationbased on flavin , t440 : 11 .3 mw-r cm l), and oxygen at ambient concentrations(0.24 mM). For the determination of enzymatic activity of immobilized cyclohexa-none oxygenase, aliquots (50-100 g.l) of the enzyme-containing polyacrylamidegel were used. Units of activity were calculated according to the equation

U n i t s . m l - 1 - A r + o - Y 9 l u . . ' u v .

Growth of Acinetobacter MIB 9871 Cells and Isolation and Purif ication ofCy c lohe xano ne Oxy g e nas e

A large-scale (25-liter) fermentation of Acinetobacter NCIB 9871 and the isola-tion and purif ication of cyclohexanone oxygenase from 80- 100 g of cells, obtainedfrom the large-scale fermentation, were carried out following procedures de-scr ibed elsewhere (17, 32).

Since the completion of this work an improved protocol for growth of thebacteria and purif ication of the cyclohexanone oxygenase has been developed.The bacterial growth can be performed on a laboratory bench top with essentiallyno specialized equipment (JJ).

Immobilization of Cyc'lohexanone Oxy genase

PAN 1000 (2 g) (34) was dissolved in 0.3 rra Hepes buffer (8 ml, pH 7.6) contain-ing NADPH (0.2 mu) and cyclohexanone (70 p.w). The enzyme solution (15 mg ofcyclohexanone oxygenase in 1.5 ml of 2 l mv potassium phosphate buffer, pH 7.1,containing 1.2 mrra dithiothreitol and 20% glycerol) was added, followed by L-5 rratriethylenetetraamine (0.5 ml). The resulting gel was ground and washed as de-scr ibed elsewhere (34). The immobi l izat ion y ie ld was 30Vc.

Determination of Substrate Specific' it1' und Measurements of Relatiue Rutes(Tab le I )

( i ) Spec' t rophotometr ic method. Cyclohexanone oxygenase act iv i ty was mea-sured using the standard assay procedure described above. The rate of decrease ins3a6 wirs measured , at 25"C, following the addition of 25 pmol of ketone dissolvedin 50 p" l of MeOH to I ml of a solut ion containing glycine-NaOH buffer (80 mu,pH 8.0), NADPH (0.16 mM), soluble cyclohexanone oxygenase (0.6 U), and atmo-spheric oxygen (0.24 mM).

(i i) Oxygen-uptake method. Assays were carried out in a stirred oxygen-elec-trode cell at 25'C. The decrease in oxygen concentration was measured by theaddition of 10 pr,l of cyclohexanone oxygenase (0.6 U)to I ml of a solution contain-ing glycine-NaOH buffer (80 mr ' l , pH 8.0), NADPH (0.16 mM), and 25,umol ofketone previously dissolved in 50 pl of MeOH. Catalase (ca. 100 U) was addedafter 3 min and the reaction allowed to continue.

Subsequent to the completion of this work, an improved protocol for kinetic

49

50

assays was developedgradual loss of enzyme

ABRIL ET AL.

(-l-t). Assays are now conducted at 15"C because of aactivity at higher temperatures.

Enzymatic Syntheses of Lactones

2-Oxabicyclo[3.2.1]octan-3-one (6). To a l - l i ter solut ion containing G-6-P '

Naz (34 .6 9 ,0 .12 mol ) , NADP+ (0 .85 g , I mmol ) , MgCl2 0 .2 g ,6 mmol ) , EDTA(330 mg, I mmol) , and 2-norbornanone (11 .4 g,0.1 mol) was added PAN-immobi-Iized G-6-PDH (100 U) and cyclohexanone oxygenase (50 U). The reaction wasconducted at pH 8.0 by the addition of 2 N NaOH using a pH controller, and thecourse of the reaction was followed by GLC for norbornanone derivative and byenzymatic assays for NADP(H) and G-6-P. After 3 days the total concentration ofNADP+ and NADPH was 55% of the original value. Additional NADP* (0.85 g, Immol) was introduced into the system and the reaction continued. After an addi-tional 2 days GLC analysis showed complete conversion of 2-norbornanone toproduct. The polyacrylamide gels containing the enzymes were separated by cen-trifugation and the supernatant was continuously extracted with dichloromethanefor I h. The CH2CI2 extract was dried (MgSOa) and concentrated to afford an oilwhich crystall ized upon cooling (11.4 g). Gas chromatography on l07c Carbowaxor l0% UC-W98 gave only a single peak. The product was further purif ied bymolecular dist i l lat ion which af forded white crystals (10.2 g. 8 l mmol ,8170 yield);

mp: 58-60'C ( l i t . (J i ) mp 56-57'C). The rH NMR spectrum of the product ob-tained was identical to that of authentic 2-oxabicyclo[3.2.1]-octan-3-one obtainedby oxidation of 2-norbornanone with rr-chloroperbenzoic acid (24). GC/MS dataobtained for the enzymatic product were similar to those of the synthetic product.

The residual activit ies of the recovered immobilized enzymes were cyclohexa-none oxygenase, 777o; G-6-PDH , 807a.

Cyclohexanone oxygenase-c'atalyzed oxidation of r-fenchone (2). The sameprotocol described for 2-norbornanone was followed for the enzymatic oxidationof l - fenchone (0.1 mol, 15.6 g). Af ter 8 days of react ion GLC analysis showedcomplete conversion of l-fenchone to product. The reaction mixture was workedup to yield off-white crystals (12.7 9,76 mmol, 76% yield). mp: 72-76"C (l it. (Ji)

mp: 61-63"C). Gas chromatography showed a s ingle broad peak. Analysis of therH NMR spectrum indicated that both the 1,2-fencholide (7) and the 2,3-fencho-lide (S) had formed, in a ratio of 88: 12. The rH NMR exhibited the two signals at 61.19 and 61.24 of the gem-dimethyl and the signal at 6 I .47 of the shielded angularmethyl for the 1,2-fenchol ide, and the two signals at 6 1.35 and 6 1.38 correspond-ing to the gem-dimethyl of the 2,3-fencholide.

Cyclohexanone oxygenase-catalyzed oxidation of o-fenchone (3). The sameprotocol described for 2-norbornanone was followed for the oxidation of o-fen-chone (0.1 mol, 15.8 g). Af ter 10 days of react ion GLC analysis showed completeconversion of o-fenchone to product. The reaction mixture was worked up asdescr ibed previously to af ford a mixture of 1,2- and 2,3-fenchol ides (13.2 9,79mmol ,79Vo yield) which was similarby GC and rH NMR to the product formed inthe enzymatic oxidation of l-fenchone. The 'H NMR spectrum indicated that therat io of lactones formed was ca. 88:12 with the l ,2- fenchol ide as the majorcomponent.

ENZYMATIC OXIDATIONS OF CATALYZED KETONES

2-Oxabicyctol3 .2 . l l t ,7 ,7- t r imethyl oc ' tan-3-one (9). The enzymatic oxidat ion

of (+)-camphor (50 mmol, 7.6 g was carried out following the procedure de-

scr ibed for 2-norbornanone. The mixture obtained af ter l0 days of react ion was

worked up ro afford 2-oxabicyclo[3 ,2.111,7,7-trimethyl octan-3-one (7.5 e. 44.5

mmol, 89% yield). The product was identif ied by comparison. by GC and NMR,

with the synthetic compound obtained from the oxidation of (+)-camphor with

40Vo peracetic acid (25).

6-lsopropenyl-3-methyl-2-oxct-heptanone (10). (+)-Dihydrocarvone (40 mmol,

6.2 g) was converted to the corresponding lactone following the protocol de-

scr ibed. After l0 days, the react ion was complete and 6-methyl-3- isopropenyl-6-

hexanolactone was isolated in 75% yield (50 g. 30 mmol). The product was identi-

f ied by comparison with synthetic material obtained from the chemical oxidation

of ( + )-dihydrocarvone with 40% peracetic acid (25 ).

ACKNOWLEDGMENTS

We thank ou r co l l eagues R . D iCos imo fo r GC/MS measurements . M . Hacke t t f o r l - s0 -MHz mea-

surements. ani i B. P. Branchaud for h is advice concerning potent ia l substrates for cy 'c lohexanone

oxygenase. This work was supported by NIH Grants GM 26-s43 and 30367 to GMW and Gr:rnt GM

2 1 6 4 3 t o C W .

REFERENCES

l . F o N x E N . G . S . . A N D J o H N s o N . R . A

New York.

( l g lD Chemica l Ox ida t ions w i th N{ i c roo rgan ism ' Dekker '

2. Mnncn. J. (1968) Advanced Organic Chemistr-"" :Reac t ions . Mechan ism. i t nc l S t r t t c tu re . 2nd ed ' .

pp . 618 , 822 . McGraw-H i l l . Ner ' r 'Yo rk .

3 . V tscHEn, A . , nNo WEr rs rE tN . A ' (19 -53 ) E ' rpe r ien t ia 9 ,371-312 '

4 . F o n N p y . F . W . . A N D M n n x o v r - r z . A . J . ( 1 9 7 0 ) J . B u c t e r i o l . l 0 2 , 2 t l l - l t t l

- 5 . C e p e r . A . . H n N c , O . . M e c e x . K . . - I n e o q . M ' . . A . N D R I E o t ' - T u v o v e " I ' - ( 1 9 5 6 ) ' \ t t l t t r r | t ' s ' r c n -

.s< hu.ft c n 43, 47 1 .

6 . P n , r t n l n . R . L . . A N D T e l - n t - e v . P . ( 1 9 6 3 ) B i t t t h e m i s u r ' 2 , 2 0 3 - 2 0 8 '

7 . R A H I v . M . A . . A N D S r H . C . J . ( 1 9 6 6 ) J . B i r t l ' C h e m ' 2 4 1 , 3 6 1 - s - 3 6 2 3 '

8 . PEr r ,nsoN. G ' E . . f uove . R . W ' . PERt - lu lRN ' D" qNo FRIeo ' J ' (19 -57 ) J ' f J t t t t t r i t t l T4 ' 6 t i 4 -686 '

9 . F n r r n . J . . ' l ' H O n a n . R . W . . A N D K I - l N c S s r - R . A . ( 1 9 - 5 1 ) J , A n e r . c h e m . . l o r ' . 7 5 . r l 6 l t - - s 7 6 9 '

1 0 . L , q s r t x . , , \ . I . . l l R e e o w l c ' n . P . . M p . t e R s . C . o e L . . r N o F n l n o ' J ' ( 1 9 6 4 ) J ' ' \ l c d ' ( ' h e n t ' 7 " 1 0 6 -

'109.

l l . SHnw. R . (1966) , \ i t r t t r r t ' ( L t tn t k tn ) 209 . 1369 '

1 2 . C H n p r r r n x . p . J . . M E r n r r , \ N . C ' . . l N o ( l u x s , . \ t t , s . I . C . ( 1 9 6 5 \ B i o t ' h e n t , B i o p h t s R t s ' ( - t t n t t n r t n '

20 . 104- lO t i .

1 3 . B n e o s n a w . W . H . . C o r u n a o . H . E . . C o n r , v . t : . J . . ( i t r N s q ' t u s ' I ' C " e N o l ' p n r t t r ' n ' D ( 1 9 ( 9 ) J '

Amer . Chem. 5n , . . 91 , -5 -507 .

1 4 . C o N n n o . H . E . . H e o E R c , q . r n o . J . . G u N s e . t . u s . t . t l ' . C l t l n p - r ' 1 1 ' J " q N o L . l o r ' i " l r l 9 6 s l ' l e t r t t l t c -

dron Let t . , -56 l - -56-5.

15 . l ' nuoc r r - r - . P . W. . DuBus . R . . ,qNo GuNse t -us . L C . (1966) J . B i t i l - Ch t ' r r t ' 24 l " l l l t t t - ' 1190 '

f 6 . D o N o c H u E . N . A . . N o n n r s . D . U . . A N D T R u o c r l t . P . W . ( 1 9 T 6 l E u r . ' 1 . B i r t t h c t t ' 6 3 . l 7 s - 1 9 2 '

17 . RyEnsoN, c . c . . Be lLou . D . P . . ANo wn l -sF l . c - . (1982) B io t ' he rn i . s t r . r ' 21 .26 '1 '1 -16 -s -s '

18 . We l -sH , C . . Jecossor ' r . F ' , AND RvpnsoN ' C ' C ' ( l 9 t t0 ) A t lu ' Chen ' r " \ c r ' l 9 l ' I l 9 - l 3 t l '

1 9 . B n u r c E . T . C . . N o n n , J . t s . . B n r r . . s . s . . . c N o V F N K A T A R A I \ { . u . v ' ( l 9 t t ' l } J . ' A n r e r ' c h e t n ' S t t t '

105. 2.1-52-2463.

_51

52 ABRIL ET AL.

20 . Wnss , C . , eNro CHEN, Y . -C . J . (1988) Angev , . Chem. In t . Ed . En91 .27 ,333-343 .

2 1 . S c u w n e , J . M . , L I , W . , A N D T H o M R S , L . P . ( 1 9 8 3 ) / . A m e r . C h e m . S o r : . 1 0 5 , 4 8 0 0 - 4 8 0 8 .22 . F lessNEn, M. S . , nNo Messey , V . (1974) i n Mo lecu la r Mechan ism o f Oxygen Ac t i va t i on

(Haya ish i . O . , Ed . ) , pp .245-283 , Academic P ress . New York .

23 . Conev , E . J . , R IvTNoRaNATHAN, T . , nNo Tenesnrue , S . (1971) J . Amer . Chem. Sor ' . 93 , 4326-

4327.24 . Mensue l l , J . L . , Bnooxs , J . P . , eNo HaTzENBUEHLER, G . W. I l l ( 1969) J . Org . Chem.34 ,4193-

4195.

25 . Sa ,uens , R . R . , AND AnrRRN, G. P . (1961) J . Amer . Chem. Sor ' . 83 ,2759-2762 .26. CneNnur-r , H. K. , nNo Wurresroes, G. M. (1987) Appl . Biochem. Biotec 'hnol . 14, 147-197.

27 . WoNc, C . -H . , GoRDoN, J . , CooNev , C . L . , AND Wsr res roes , G . M. (1981) J . Org . Chem.46 ,

4676-4679.

28. TnscnNER, M. J. , AND BLACK, D. J. (1987) l94th Nat ional Meet ing of the American Chemical

Society, New Orleans, LA, Amer. Chem. Soc. , Washington, DC, ORGN 109.

29 . L r rue ,u , J . A . , AND WALSH, C . (1987) J . Amer . Chem. Soc ' . L09 ,3421-3427 .

30. Porrnx, A. , BeucHr-r , R. L. , aNo WHITESTDES, G. M. (1977) J. Amer. Chent. Sor ' . 99, 2366-2367.31 . Bnnc r r , rEyER, H . U . (19 '14 ) Me thods o f Enzymat i c Ana lys i s . Ve r lag Chemie . Ncu Yurk .

32 . RvnnsoN, C . C . (1980) Ph .D . thes is , M IT , Cambr idge .3 3 . B n e N c H A U D , B . P . , e N o W n l s H , C . ( 1 9 8 5 ) J . A m e r . C h e m . S n r . 1 0 7 , 2 l - 5 3 - 2 1 6 1 .

34 . Por - lex , A . , BLUMENFELD, H . , Wex , M. , Bnucn , R . L . , AND WHt rgs to r -s . G . M. (1980) " / .Amer. Che m. Soc. 102, 6324-6336.

35 . MprNweLD, J . . AND FnaupNcr -ess , E . (1960) J . Amer . Chem. .5o r ' . 82 ,5235-5239 .