enzymes in otganic synthesis - harvard university · pressures of environmental constraints, and...

of 8 /8
./ Enzymes in otganic synthesis Alan Akiyama Mark Bednarski Mahn-Joo Kim EthanS. Simon HerbertWaldmann GeorgeM. Whitesides Illaturing of the petroleum-based chemical industry, pressures of environmentalconstraints, and the explosive developmentof biotechnology have increased interest in enzymology. Enzymes are now an attractiveproposition as catalysts for new classes of reagents and products, especially sugars, chiral synthons,metabolites, and food components. Enzymesare catalytic proteinsthat control the ratesof most biological reactions; they have been used ex vivo in small-scale applications for both the analysis (l) and the synthesis of research biochemicals (2, S) and in a small but significant number of large-scale chemicalprocesses (4-6), Their usein the latter was limited because most enzvmes were expensive and unstable, and they were not the best catalysts for the reactions of greatest interest in the chemical process industry or the pharmaceutical industry (Table 1). Several developments in industry and biology have changed the value of enzymes as catalysts. The r"ublution in molecular genetics has made enzymes available at dramaticallylowered costs (5, T),At the same time, the targetsof central interest in the pharmaceutical industry have shifted away from low molecular weight hydrocarbons and heterocycles to complex biological substances-for example, tissue plasminogen activator (TPA), lymphokines, and cvclosporin A-or substances derived from them. The creation of chiral centers has become a major strategicconcern in chemical synthesis, especially in drug synthesis, Moreover, public concern with environmentalissues hasincreased the attractiveness of processes that operate with high selectivities and thus minimize the problems of waste and byproduct disposal in environmentallyacceptable solvents. Enzymesshould not be considered as replacements for existing catalysts; the idea of replacing the chemical process for the conversion of propyleneto propyleneoxide b19d on organic peroxides and transition--"t"1 catalysts (B)bV an oxidativeenzymaticprocess (9) was never a good one. Rather, they should be considered as a new clais of catalysts for which new uses and new processes must be developed. The processes that currently use enzymes have been extensively reviewed(5, -10). Already the use of enzymes as Table 1. Perceived characterislicsof enzymes Advantagee Olsadvaniage* Highly selective for substrates Highly enantioselective Environmentally acceptable Applicable to products relevant in pharmaceuticals, food, biotechnology, agriculture Large numbers of catalytic activities available Costcan be lowered and properties improved using biotechnology Expensive Unstabte Restricted to predominantly aqueous environments Inapplicable to substrates not occurring in nature Inapplicable to many important types of reactions Difficult to manipulate components in detergent formulations (l/) common in Europe and the United States has proved economical, Enzymesare a small part of detergent formulations,but they remove proteinaceousstains sufficiently well to justify their production on a large scale, The conversion of starchto glucose, and glucose to high-fructose corn syrup, is the largest scale chemical transformation effected by enzymes (5, 12). The vearly quantity of high-fructose corn syrup production in the United States is now more than a billion pounds, The Novo process for the production of semisynthetic insulin (lS) provides the first example of what will probably be a widespread use of enzymes, that is, to modify proteins and other biological macromolecules derived from recombinant DNA technology. The efficiency and economicsof this processand the high quality of the product establish the value of enzymatic transformations. The range of amino acids now derived from enzymatictransformations illustrates the importance of enzymology asa technology for resolving and producing chiral centers (2, 4, 14).For example,biologicalprocesses for synthesizing r--dopa(2-amino-3[3,4-di-hydroxyphenyl] propanoicacid) are now displacingtechnologies based on asymmetric hydrogenationusing rhodium catalysts, Before discussing reactions and processes now under development, let us summarizesomeissues related to the use of enzymes as catalysts, Enzyme production. Enzymes used in chemical processes are mainly derived by extraction from natural sources, especially microorganisms. In synthetic applications, it is usually satisfactory to work with

Author: others

Post on 27-Sep-2020




0 download

Embed Size (px)


  • ./

    Enzymes inotganic synthesisAlan AkiyamaMark BednarskiMahn-Joo KimEthan S. SimonHerbert WaldmannGeorge M. Whitesides

    I l laturing of the petroleum-based chemical industry,pressures of environmental constraints, and the explosivedevelopment of biotechnology have increased interest inenzymology. Enzymes are now an attractive proposition ascatalysts for new classes of reagents and products,especially sugars, chiral synthons, metabolites, and foodcomponents.

    Enzymes are catalytic proteins that control the rates ofmost biological reactions; they have been used ex vivo insmall-scale applications for both the analysis (l) and thesynthesis of research biochemicals (2, S) and in a small butsignificant number of large-scale chemical processes (4-6),Their use in the latter was limited because most enzvmeswere expensive and unstable, and they were not the bestcatalysts for the reactions of greatest interest in thechemical process industry or the pharmaceutical industry(Table 1).

    Several developments in industry and biology havechanged the value of enzymes as catalysts. The r"ublutionin molecular genetics has made enzymes available atdramatically lowered costs (5, T),At the same time, thetargets of central interest in the pharmaceutical industryhave shifted away from low molecular weighthydrocarbons and heterocycles to complex biologicalsubstances-for example, tissue plasminogen activator(TPA), lymphokines, and cvclosporin A-or substancesderived from them. The creation of chiral centers hasbecome a major strategic concern in chemical synthesis,especially in drug synthesis, Moreover, public concernwith environmental issues has increased the attractivenessof processes that operate with high selectivities and thusminimize the problems of waste and byproduct disposal inenvironmentally acceptable solvents.

    Enzymes should not be considered as replacements forexisting catalysts; the idea of replacing the chemicalprocess for the conversion of propylene to propylene oxideb19d on organic peroxides and transition--"t"1 catalysts(B)bV an oxidative enzymatic process (9) was never a goodone. Rather, they should be considered as a new clais ofcatalysts for which new uses and new processes must bedeveloped.

    The processes that currently use enzymes have beenextensively reviewed (5, -10). Already the use of enzymes as

    Table 1. Perceived characterislics of enzymesAdvantagee Olsadvaniage*

    Highly selective forsubstrates

    Highly enantioselectiveEnvironmentally

    acceptableApplicable to products

    relevant inpharmaceuticals, food,biotechnology,agriculture

    Large numbers ofcatalytic activitiesavailable

    Cost can be lowered andproperties improvedusing biotechnology


    UnstabteRestricted to

    predominantly aqueousenvironments

    Inapplicable to substratesnot occurring in nature

    Inapplicable to manyimportant types ofreactions

    Difficult to manipulate

    components in detergent formulations (l/) common inEurope and the United States has proved economical,Enzymes are a small part of detergent formulations, butthey remove proteinaceous stains sufficiently well tojustify their production on a large scale, The conversion ofstarch to glucose, and glucose to high-fructose corn syrup,is the largest scale chemical transformation effected byenzymes (5, 12). The vearly quantity of high-fructose cornsyrup production in the United States is now more than abil l ion pounds,

    The Novo process for the production of semisyntheticinsulin (lS) provides the first example of what wil lprobably be a widespread use of enzymes, that is, tomodify proteins and other biological macromoleculesderived from recombinant DNA technology. Theefficiency and economics of this process and the highquality of the product establish the value of enzymatictransformations. The range of amino acids now derivedfrom enzymatic transformations illustrates the importanceof enzymology as a technology for resolving and producingchiral centers (2, 4, 14). For example, biological processesfor synthesizing r--dopa (2-amino-3[3,4-di-hydroxyphenyl]propanoic acid) are now displacing technologies based onasymmetric hydrogenation using rhodium catalysts,

    Before discussing reactions and processes now underdevelopment, let us summarize some issues related to theuse of enzymes as catalysts,

    Enzyme production. Enzymes used in chemicalprocesses are mainly derived by extraction from naturalsources, especially microorganisms. In syntheticapplications, it is usually satisfactory to work with

  • relatively crude preparations, and immobilized whole cellsare often the most convenient and economical form inwhich to use an enzyme. Immobilization facilitatesrecovery and improves the stability of the enzyme' A rangeof techniques for immobilizing enzymes is available (4,

    14-17). The method most commonly used in large-scalepreparations is based on crosslinking reactions withgluiaraldehyde (4,75,18). This procedure works well withrelatively stable enzymes but often fails with the moredelicate enzymes required for complex synthesis. For thelatter group, reactive organic polymers, such aspolyacrylamide (PAN) (19), are more effective.-

    Cofactors. Most of the enzymatic reactions used in

    synthesis require cofactors. Essentially all of the common

    cofactors-ATP, NAD, NADH, and CoASH-are too

    expensive to use in stoichiometric quantities' We have

    developed in situ cofactor regeneration schemes that allowus to easily regenerate ATP from ADP or AMP by using

    either aceiyl phosphate (20) or phosphoenol pyruvate (2I)

    as the phosphate donor. The best method for regeneratingNADH from NAD is based on formate as the hydride

    donor (17, 22). Regeneration of NAD remains slightly

    complicated (23, 24),but the best regeneration scheme isprobably one based on a-ketoglutarate as the }ydride

    acceptoi (23). Using stoichiometric amounts of a-ketoglu-tarate is expensive and limits this type of regeneration.

    Enzyme purification. In certain cases, especially thoseinvolving cofactors or delicate substrates, it is important to

    consider highly purified enzymes. For example, it can beparticularly advantageous to exclude proteases andATP"s"t because proteases can shorten the lifetime of the

    enzyme of interest by degrading it proteolytically, andATPases can decrease the efficiency of ATP usage in a

    cofactor recycling scheme. In addition, certaincontaminating

    "iry^e activities are especially

    troublesome in chiral transformations where the required

    substrate may be accepted by more than one enzyme in a

    crude enzyme preparation (25). ff these competingenzymes have different enantioselectivities, the result can

    be unacceptably low enantiomeric excess (ee) in theproduct. Establishing that a crude preparation of enzymeshas the purity and reproducibility required for a stable anduseful process is an important step in designing an enzyme-based reaction sequence.

    Specific activity. The specific activity of an enzyme is ameasure of its catalytic activity, usually given inmicromoles of product generated per minute per

    milligram of enzyme. It is important to establish this valueearly in any consideration of an enzyme-based process

    because it determines the upper limit to the productivity ofan enzymatic reactor. It is currently not practical toincrease the specific activity of an enzyme significantlybeyond its intrinsic value. If the specific activity is too lowto form the basis for an acceptable process, other enzymesshould be examined for higher activity.

    Kinetic considerations. Enzymatic reactions take placein fluid phases and are thus slower than many of the vapor-phase ptocett"s widely used in large-scale chemicaliynthesis. A number of enzymatic processes-for example,those involving lipases (enzymes that require a water-hydrocarbon interface for activity) or other enzymes in

    two-phase water-hydrocarbon systems-introduceproblems in interfacial mass transport that have not yet

    been thoroughly analyzed. In addition, enzymes

    sometimes suffer from intrinsic kinetic limitations (26)-

    for example, product inhibition (2a)-.that are part of theenzyme regulation system in vivo but are a nuisance in

    synthesis.We have explored a number of strategies to avoid

    product (and sometimes reactant) inhibition (24), but

    convenient general strategies are not available.Reactor design. Relatively little work has been done on

    the development of reactors engineered specifically to

    take advantage of the catalytic characteristics of enzymes.Many existing processes use column reactors adapted more

    or less directly from those used with heterogeneouscatalysts in classical chemical synthesis' New reactorsbased on membranes (27) and hollow fibers (28) are now

    beginning to attract interest.thr following sections illustrate some applications of

    enzymes to current problems in organic syntlesis. First

    *"'il look at the generation of a low molecular weightchiral synthon.

    Ghiral epoxy alcoholsThese versatile synthetic intermediates, which are

    widely useful in pharmaceutical synthesis, have been

    studied extensively by Sharpless and co-workers (29' 30).

    Sharpless's chemistry uses asymmetric epoxidizingsystems, usually based on an organic peroxide in

    combination with a complex between a titanium salt and a

    chiral ligand such as tartaric acid. Thus we have an

    opportunity to compare the characteristics of enzymatic

    "nd nott"nzymatic processes for the synthesis of these

    enantiomerically enriched compounds'Most of our work has focused on the kinetic resolution of

    epoxy alcohols by enantioselective hydrolysis of thederived esters using lipases, Chiral synthons with an ee

  • cH3coNH


    E T















    }--co2H/E I

    E t . '


    /\, r /

    E r


    >cozH (e5)


    ococaHT Cholesterol esterase ococaHT

    L-Lactate dehydrogenase

    l t \o r,rnox ulf\\/--corH

    E I

    H O)t

    ;'-co2{E I

    H O\),-cozH

    E I

    H Oo .

    \---lE T

    Glycerol dehydrooenase \ ./r / I E r 'O H (>98)

    l n [ ( I - c ) ( L - e e , ) l

    l n [ ( t - c ) ( I * e e , ) ]

    l n [ l - c ( l * a e r ) ]

    l n [ l - c ( I - e e o ) )

    where c is extent of conversion, ees is enantiomericexcess of remaining -starting material, and oep isenantiomeric excess of product.

    E values are useful parameters but are sensitive to themeasured extent of conversion, one must look carefully atthe accuracy of measurements of both enantiomeric excessand conversion in order to realistically assess whetherchanges in methods of manipulating the enzymes andreaction conditions significantly improve theenantioselectivity of an enzymatic reaction.

    Are enzymatic methods based on enantioselectivehydrolysis better than the sharpless reaction for synthesisof chiral epoxy alcohols? We can't give a general answer.For certain compounds, enzymatic methods work well.The.1 ?Ie simple and easy but are not universallyapplicable, Because they are kinetic resolutions, however,


    ligure 1. Four synthetic routes to optically enriched butene oxide using enzymes. Reaction conditions z a,2 N HCl, 100 "C,2h;bt 6 N HCl, NaNOz; c, BH3, THF,0 oC,20 h; d, KoH; e, C5H11Ox, i rxrr6H, room temperature; f ,307o HBr.AcoH,- 1 5 o C - 2 5 o C , 3 h

    o Low enantiomeric selectivity in a kinetic hydrolysisis sometimes intrinsic to the enzyme being used, andsometimes reflects competing activities by othere_nzymes present as impurities. We have improvedthe enantioselectivity of crude enzyme m^ixtureswithout carrying out elaborate columnchromatographies. Useful strategies involve partialdenaturation of a crude mixture of enry-es bv heattreatment, by shifting pH, or by adding denaturants.With luck, the undesired impuritier *ill prove lessstable to some of these conditions than will the desiredmajority enzyme, and a relatively simple process canthen be developed that selectively destroyi unwantedimpurity activities while retaining a large part of thedesired activity.

    o Analysis. One issue that is often slighted in itsimportance in the development of processes based onenantioselective hydrolysis is analysis ofenantioselectivity. The procedure commonly used toanalyze enantioselectivity is based on a conversion-independent measure of enantioselectivity proposedby Sih and co-workers (83), the so-called E value.

    [ - -

    p =

    CHFIITtrCI{ r)(]T.1Rtra loaa aro

  • o

    *A* + Ho - ll * oPor2' Aldolase-.......---"""""....,.,..'........4




    R=Me, iPr, CHrOC6H5, etc.

    o+ HO-A-.oPo32-

    Aldolase - HO

    Figurc 2, Ftuclose-1,6-dlphosphale aldolase calalyzos a stereospecllic aldol condensation belween dihydroxyacetonephosphate (DHAP) and a varlely ol aldehydes. The method is especlally usolul in lhe preparation ot rare carbohydrales asexemplltled above by the synthesls ol D-xylulose


    the ultimate yield is intrinsically limited. The Sharplessreaction is a chiral synthesis and thus, in principle, capableof giving product in higher yield. Workup of reactionmixlures, though, can be a problem. Certain classes ofcompounds are accepted well by enzymes and performpoorly in the Sharpless reaction; the reverse is also true.Enzymatic methods are not limited to epoxy alcohols,Thus evaluation of the relative applicability of enzvmesand organometallic catalysts for production of chiralepoxy alcohols and derivatives must also be conducted on acase-by-case basis,

    Chiral butene oxideWe also have examined a number of strategies for the

    preparation of chiral epoxides that are not epoxy alcohols.Both the Sharpless procedure and lipase-catalyzed kinetichydrolyses are restricted to epoxy alcohols and relatedspecies. Chiral epoxides are, however, generally usefulchiral synthons, and we wished to evaluate routes to thembased on enzymatic procedures. We have chosen buteneoxide as a representative structure. Figure 1 summarizesfour synthetic strategies for its preparation. We emphasizethe distinction between strategies based on kineticresolutions, in which the yield of one enantiomer from aracemic mixture is in principle l imited to 50%-but inwhich both enantiomers may be produced-from thosebased on asymmetric syntheses (in which a singleenantiomer can, in principle, be generated in I00% yield).

    Lipase. The enantioselectivity of lipase-catalyzedhydrolysis of bromohydrin and chlorohydrin estersdepends markedly on the structures involved. Pancreaticlipase is only moderately enantioselective in catalyzing thehydrolysis of the bromohydrin ester derived from butene,although it works considerably more selectively with otherbromohydrin and chlorohydrin esters. To produce buteneoxide with high enantioselectivity following this route, onemust carry the extent of conversion of the racemic startingmixture relatively far. It would be desirable to find moreenantioselective lipases to catalyze the hydrolysis, or toimprove the enantioselectivity of the porcine lipase bychanging reaction conditions.

    Acylase I. The kinetic resolution of amino acids byacylase I is broadly applicable and highly enantioselective.However, the route to butene oxide based on acylase I hasfour reaction steps, with one exotic reaction: conversion ofthe a-amino acid to an a-chloroacid by nitrosation. It hasthe advantage that it generates other useful intermediates,especially the chiral a-chloroacid.




    2)H* I uzOOH


    Lactate dehydrogenase. Routes to butene oxide basedon reduction of prochiral a-ketobutyric acid are chiralsyntheses. In the particular case of lactate dehydrogenase,both enantiomers of a-hydroxybutyrate are available withhigh enantiomeric excess, because both o- and l-lactatedehydrogenase are available and inexpensive. The t--lactate dehydrogenase has, however, much broadersubstrate specificity than does the n enzyme. For largera-ketoacids, the availability of both enantiomers is morerestricted, Lactate dehydrogenase requires NADH as acofactor, and its use is more complex and expensive thanthat of simple hydrolases.

    Glycerol dehydrogenase. Reduction of l-hydroxy-2-butanone with glycerol dehvdrogenase is anotherasymmetric synthesis that requires a cofactor. It is anefficient reaction but it requires low molecular weightsubstances because of the limited substrate specificity ofglycerol dehydrogenase (24).

    Unnatural sugarsWe have explored the usefulness of aldolases as catalysts

    for the preparation of sugars. Many different enzymeshave aldolase activity (2). The most readily available,rabbit muscle aldolase, catalyzes the condensation ofdihydroxyacetone phosphate with a variety of aldehydes(34). Past work has seen the application of this system tothe synthesis of naturally occurring monosaccharides'Recent efforts have demonstrated that this system is usefulfor transferring the chirality present in hexoses andpentoses into C6 and Ce sugars (35). These compounds areinteresting as precursors for and analogues of sialic acids(amino sugars containing nine or more carbon atoms) andrelated species (Figures 2 and 3),

    Another useful class of enzvmes include those thatcondense pyruvate and pvruvate derivatives withaldehydes. For example, the use of ketodeoxyoctanoic acid(KDO) synthetase to synthesize KDO-B-phosphate fromarabinose-S-phosphate and phosphoenol pyruvate (PEP)

    il lustrates one such enzymatic activity (36)' We are nowexamining the ability of this enzyme to accept analogues ofarabinose phosphate and PEP and are developing apractical large-scale synthesis of KDO.

    Metabolic intermediatesEnzymatic synthesis is ideal for preparing many classes

    of metabolic intermediates and their analogues, such asglycerol-3-phosphate (37) and phosphoribosylpyrophosphate (38). The former compound is optically

  • ronroL-oH


    Glycerokinar" \

    f ot osHe


    OpO,H^ Acide I ohosoha-

    o tase

    r{r,,, ,/|-ot'Lor osHa

    active and has the correct stereochemistry to be a precursorfor lipids for possible use as surfactants in liposomes andrelated drug delivery systems; the latter is valuable as abranch point intermediate in the synthesis of manynucleosides and nucleotides, For both compounds classicalchemical synthesis cannot compete with enzymaticmethods for ease and specificity.

    Oligosaccharides and polysaccharidesOligosaccharides are important as cell surface markers

    and are potentially useful in drug delivery systems. Theyare difficult or impossible to make in large quantities byusing classical synthetic techniques, although they havebeen the object of elegant and highly successfulsmall-scalesyntheses (39). Polysaccharides are important for theirinfluence on rheology of aqueous solutions and for certainspecific biological activities. Enzymatic methods areapplicable to both. The biosynthesis of these classes ofsubstances follows two biosynthetic pathways. One, theLeloir pathway, which uses sugar moieties activated asnucleoside di- or monophosphate sugars, has beendemonstrated by preparations of di-, tri-, andtetrasaccharides (40). These syntheses, which can be usedto prepare f 0-g quantities of oligosaccharides, arerelatively complex because they require regeneration ofthe nucleoside phosphates. The principal limitation ofLeloir pathway syntheses is the availabiliiy of the requiredglycosyl transferases. The second pathway uses sugarphosphates per se. It is not broadly applicable but isespecially good for the preparation of deitrans (47) andlaminarins (a2) (Figure 4).

    Multienzyme systemsOne of the features that distinguishes enzymatic from

    nonenzymatic catalysts is the broad mutual compatibility


    \ Hg4 Fructose-l,6-(pOsH2)2

    l- oHHO

    ,//r/ r/

    ,/nraoaset// triose phosphate/ tsomerase

    CHO OH

    )e,s )z,go OHoPo3H2 oPo3H2

    Figure 3. The elaboration of simple C5 and C5 fionosdccharides into octuloses and nanuloses using fructose-1,6-diphosphate aldolase as a catatyst



    of enzymatic catalvsts, Most enzymes operate satisfactorilyin aqueous solution at pH 7 and room temperature;relatively few broadly used nonenzymatic catalystsoperate under conditions that are compatible with oneanother. It is thus possible almost uniquely with enzymes toconsider the construction of complicated serial andparallel catalytic sequences involving the combination ofdifferent enzymes to provide complementary catalyticactivities. This area of catalyst development-thedevelopment of multienzyme systems-is still in theresearch stage. Many of the examples cited above haveinvolved groups of three to five cooperating enzymes.Systems of this level of complexity are easily controlledand assembled using commonly available enzymes. Amore complex example is that illustrated in Figure 5. Thecomplexities of working with this system are such that,although it provides a useful system with which to studythe mechanisms of metabolic regulation, it seems unlikelyto be synthetically useful. Our experience to date suggeststhat the limit in complexity that can be tolerated in mostsynthetic applications is between 5 and l0 cooperatingenzymes. Numbers larger than 10 become too difficult tobe synthetically practical; numbers smaller than this canbe handled with more or less facilitv.


    ^ Enzymes are clearly attractive and practical as catalystsfor a number of types of synthesis. Where will they be usedin practice? What are the problems currently limitingtheir use?o Chiral sEnthesis. The now-established capability of

    enzymes to produce chiral synthons will be widelyused in the pharmaceutical industry for the synthesisof complex drugs. Among the major problems to beovercome in this area are the identification of classes of


  • F/

    / UDPGE

    -/oHHot\-,.,,-o-\..-- "\

    Ho lO-UDP



    Figure 4. Enzymatic synthesis of lactosamine. GT, galactosyltransferase; PK, pyruvate kinase; PGM, phosphoglucomutase;

    UOpGp, ggp-glucosepyrophosphorylase; UDPGE, UDP-glucose epimerase; PPase, inorganic pyrophosphatase;

    PP;, inorganic pyrophosphate; UTP, uridine 5'-triphosphate



    tlt l

    Ee l lt lll



    OH- z -


  • enzymes with useful enantioselectivitv. thedevelopment of conditions for their use, and theidentification of chiral products sufficiently valuableto warrant process optimization,Metabolic intermediates and analogues-aminoacids, sugars, oligosaccharides, ond polgpeptides.Again, the value of enzymes in manipulating theseclasses of substances is clear. A numbei of substantialquestions remain in these areas, however, particularlyquestions that concern the usefulness of enzvmes ascatalytic entities relative to other types of catalysts.Enzymes can, for example, be used as dehydratingagents to catalyze the formation of peptide bonds (21,43). Classical methods of peptide bond formation arehighly developed, and it remains to be seen howbroadly enzymes will be used. In principle, they havethe advantages of high enantioselectivity, and tLey donot need activating reagents, On the other hand, theyare more limited in their applicability.Process aids in biotechnolog,l. Recombinant DNAtechnology has made available a broad range ofproteins. Enzymes could be used to manipulate theseproteins, for example, to attach sugar moieties, to doselective proteolytic clipping or stitching ofpolypeptide fragments, or to destrol' unwantedimpurities, Until recently attention of the emergingrecombinant industry has been focused on theproblems of producing the proteins-genetics,microbiology, and cell culture. These problems are byno means completely solved, but the attention is norvshifting over to the problems of isolation,modification, and purification.Large-scale sEnthesis. It is presently unclear howuseful enzymes will be in large-scale synthesis. Theproduction of high-fructose corn syrup represents aspecial case in that it is a one-step conversion of asugar, the type of process in which enzymes havespecial utility, Hopes of replacing or improvingcommodity hydrocarbon transformations-especiallvthose based on oxidative functionalization-do notseem realistic. The most plausible applications forenzymes are hydrations or dehydrations. Thusformation of esters or amides is a plausible target forenzyrnology, as are reactions such as the hydrolysis ofnitriles to amides or acids (44). After initial andunrealistic expectations for the application of enzymesin large-scale process chemistry, proposals in this areafell into a certain disrepute. As the characteristics ofenzymes have become clearer, they are now beingreexamined in a more realistic light,Waste treatment, Enzymes work well as catalysts inwater, and two of the most pressing problems facingthe chemical process industry are waste disposal andwater purification at plant sites. One of the majorstrategies for waste-water treatment-dieestion ofimpurities in bioponds-is already un

    "nry*-e-basedprocedure in the sense that the microorganismsresponsible for water purification are based onenzymatic activities, It is plausible to speculate thatisolation and concentration of desirable enzvmaticactivities might be useful for the treatment of certaintypes of waste streams: for example, removal ofcyanide (45) or phenol (40) from dilute aqueoussolution. These types of applications are now beingactively explored.

    . Food processing. A number of problems inimprovement of f lavor and controlof rheology dependon chemical transformations whose nature either is orwill soon be understood at the molecular level.Enzymes are attractive in that they can be used ascatalysts in situations where many classical organicreagents are not permitted by regulatory agencies.

    FutureSeveral areas are now receiving active attention. First is

    the development of improved processes. The use ofenzymes as catalysts in two-phase organic aqueous systemsis an attractive method of circumventing some of theproblems in solubilities that have limited the applicabilityto water-soluble substrates. Supercritical fluids mayprovide a way of improving mass transport in enzymaticreactors and of continuously extracting products (47),Continuous extraction and related schemes providestrategies for circumventing the problems ofnonc^ompetitive product inhibition and facilitatingpurif ication.

    - Second, exploratory work in enzymology continues-that is, examining a range of enzymes to identify those thatcombine the activitr ' , stabil ity, and breadth ofapplicabil ity required to be useful as catalvsts in organicsynthesis.

    Third, genetic engineering is increasingly being appliedto improve the catalvtic characteristics of enzymes. Theinitial efforts in this area have been focused on relativelysmallmodifications in the amino acid sequences of existingenzymes, with the intent of altering kinetics or changingstabil it ies (aB). These init ial efforts have been rema.kiblvsuccessful and suggest that site-specific mutagenesis rvii lbe useful in the rational improvement of existingenzymatic activit ies in the relatively immediate future.The broader problem of large changes in amino acidsequence to produce major changes in catalytic properties,or of de nouo design and construction of new .ut"lyti"activit ies, require substantialadvances in basic science andwill not be solved soon.

    The most important issue is, however, a basic change inphilosophy in large-scale synthesis. In the past, enzymeshave been (probablv correctly) regarded less useful ascatalysts than the more commonly used inorganic andorganic substances for the classes of problems important inpractical-scale chemistry. The maturing of the petroleum-based chemical process industrv, the pr"rr.rr", ofenvironmental constraints, and the explosive developmentof biotechnology have now shifted the focus of problems incatalysis such that enzymology has become much moreattractive. The init ial considerations of enzymes aspotential catalysts for the chemical process industry werebased on the hypothesis that enzymes might be useful inproviding more economic processes for existingpetrochemically derived products, This hypothesis may becorrect in certain specialized cases-especially thoseinvolving reactions that add or remove water fromreactants or products-but probably is not generally valid.The use of enzymes as catalysts for new classes of reagentsand products-especially sugars, water-soluble mateiials,chiral synthons, biopolymers, metabolites, foodcomponents, and related substances-seems entirelyrealistic, and to the extent that chemistry wil lbe concernedwith these types of problems, enzymes will play animportant role.


  • AcknowledgmentsThis work was supported by National Institutes of Health grant GM30367 and the Naval Rir Syslems Command MDA 903-86-M-0505. Wegratefully acknowledge feliowships to M. Bednarski (American CancerSocietv postdoctoral* fellow tgSS-tggg; grant PF-2762) and H'Waldmann (Deutsche Fortschungs-gemeinschaft postdoctoral fellow1985-r986).

    References(I) Bersmeyer, H. U., Ed. Methods of Enzgmatic Analgsis, Srd ed;

    VerIag-Chemie: Weinheim. I 986.(2) White"s ides, G. M.; Wong, C.H. Angeu. Chem.lnt . Ed. Engl .1985,

    25, 617.(3) Iones, f . B. In Asgmmetric S7nthesis; Morrison, J. D., Ed. Academic:

    New York, 1985; Vol . 5. P. 309.(4) Chibata, I., Ed. lmmobilized Enzqmes-Research and

    Deoelopment; Halsted: New York, 1978.(5) "Commercial Biotechnology: An International Analysis,

    'Office of'

    Technology Assessment, U.S. Congress: Washington' D.C., 1984;oTA-BA-218.

    (6) Sonnet, P. CHEMTECH, February 1988, p. 94.(Z) fe.sht, A. R.; Winter, G. P. ln Enzymes in Organic Sgnthesis, CIBA

    Foundation Sgmposium III; Porter, R.; Clark S., Eds.; Pitman:London, 1985; p. 204.

    (8) Mimoun, H., ef al. J. Am. Chem. Soc. 1986, l0B, 37ll(9) Neidleman, S. L. ; Amon, W. F. , Jr . ;Geigert , J . U.S. Patent4 '284'723,

    I98l ; Cetus Corp.(10) Scott, D. In Kirk-Othmer Concise Encgclopedia of Chemical'

    Technology,Srded.; Gravson, M., Ed.; Wilev: New York, 1980; Vol.I , p. 173.

    (lI) Siarce, C.; Barford, H. C. In Concise l.ncgclopedia ot' ChemicalTechnologg, Srd ed.; Kirk-Othmer, Ed. Wilev: New York, 1980;Vol. 9, p. 138.

    (12) Venkatasubramanian, K, ln Enzgmes, the Interface betueenTechnologg and Economics; Danehy, J. P.; Wolnak, B', Eds;Marcel Dekker: New York, 1980; P. 35.

    (13) Markussen, J.; Schaumburg , K. ln- Peptides; Plaha, K.; Malon, P ,Eds.; de Gruyter: Berlin, 1983; P 387

    (I4) Izumi, Y, ;Chibata, T. ; I toh, T. Angeu. Chem.Int . Ed. Engl .1978,17. 176.

    (15) Mosbach, K., Ed. Methods Enzgmol. L976,14.(16) Mathiasson, G. lmmobtltzed Cells and Organelles; CRC Press:

    Cleveland, Ohio, 1982.(17) Wandrey, C.; Bruckmann, A. F.; Kula, M. R. Biotechnol. BioenS

    1981 ,23 , 1789 .(18) Klibanov, A. M. Science 1983,219,722.(19) Pol lak, A. , et a l . J .Am Chem. Soc 1985, 102,6324.(20) Kazlauskas, R. J. ; -Whitesides, G. M. J Ore Chem.1985,50, 1069(2I) Hirschbaum, B. L.; Mazenod, F. P.; Whitesides, G. M / Org. Chem

    1982.47.3765.Shaked, Z. ; Whi tesides, G. M. J.Am.Chem. Soc. 1980, 102,7104.Lee , L . G . ;Wh i tes ides , G . M. J .Am,Chem. Soc . 1985 , 107 ,6999 .Lee, L. G.; Whi tesides, G. M. J.Ore, Chem,1986,5l , 25.Sih, C. J.; Chen, C. S. Angeu. Chem. lnt. Ed. Engl. 1984,23,570.Engel, P. C. Enzgme Kinetics; Chapman and Hall: New York, 1981.Wichmann, R., et al. Biotechnol. Bioeng, 1981, 23, 2789.Kotano, H.; Ise, N. Trends Biotechnol. 1984,2,5.Behrens, C. H.; Sharpless, K. B. Aldrichimica Acta L983,16,67.Sharpless, K. B. CHEMTECH, November 1985, p. 692.Ladner, W. E.; Whitesides, G. M / Am. Chem. Soc. 1984, 106,7250.

    (32) Klibanov, A. CHEMTECH, June 1986, p. 354.(33) Chen, C. S., et al. J. Am. Chem. Soc. 1982, 104,7294.(34) Wong, C. H. ; Whi tesides, G M / Org. Chem.1983, 48, 3199.(ss) n"a;;ski, M. D.;waldmann, H. j.;wh-itesides, G. M. Tetrahedron

    Lett . 1986,27,5807.(36 ) Ray ,P H / Bac te r io l .1980 ,193 ,635 .(37) Rids-Mercadillo, V. M.; Whitesides, G M / Am. Chem. Soc. 1979,

    101.5828.(38) Gross, A. , et a l . J . Am.Chem. Soc. 1983, 105,7428.(39) Paulsen, H. Chem. Soc. Reu. 1984, I3, 15.(40) wong, C. H.; Haynie, S. C.; Whitesides, G. M, J. Org. Chem.1982,

    4 7 , 5 4 1 6 ,(4I) Sanford, P. ; Baired, J. ln The Polgsacchar ides; Aspinal , G. O., Ed. ;

    Academic: New York, 1983; Vol . 2, p. 4 l I .(42) Waldmann, H. et al. Carbohgdr. Res. 1986, 157, C4.(43) Jakubke, H. D. ; Kuhland, P. ; i (oenecke, A, Angew. Chem. Int . Ed.

    Engl . 1985,24,85.(44) Asano, Y. , et a l . Agr ic. Bio l . Chem. 1982,46, 1183.(45) Vennesland, B., et i l., Eds. Cganide in Biologg; Academic: London,

    1 9 8 l .(46) Klibanov, A. M. ln Biotechnol. Mar. Sci., Proc. Annu' MIT Sea

    Crant Lect , Semin. ls t 1982; Colwel l , R. R. ; Sinskey, A. i . ; Par iser,E. R. , Eds. ; Wi ley: New York, 1984; p. 259.

    (47) Hoyer, G. G. CHEMTECH, July 1985, p.440.(48) Winter , G., et a l . l la ture 1982,299,756.

    Alan Akiyama is a doctoral candidate inorganic chemistry with George Whitesidesat Harvard University. He is studying thepractical production of optically pure

    compounds using enzyme-catalvzedreactions. He received his A.n. from theUniversity of Pennsylvania.

    Mahn Joo Kim studied under GeorgeWhitesides and received his Ph.D. from

    Harvard in 1987. He is currently apostdoctoral research associate with Chi-Huey Wong at Texas A&M UniversitY(College Station, TX 77343; 409-845-9163).He grew up in Chuncheon, Korea, andreceived his B.S. at Seoul NationalUniversity.

    Herbert Waldmann is an AssistantProfessor of Chemistry at JohannesGutenberg Universitdt, Mainz Institut frir

    Organische Chemie (D-6500 Mainz,Federal Republic of Germany). His Ph.D.research with Horst Kunz centered on

    developing methods of synthesizingglycopeptides for immunological studies.He was a postdoctoral fellow with GeorgeWhitesides.

    9 9 )

    23)24)25)26)9 1 \

    28)t o \

    30)3 l )


    W1f r

    George M. Whitesides is MallinckrodtProfessor of Chemistry and DepartmentChairman at Harvard University (12

    Oxford St . , Cambridge, MA 02138; 617-195-9430). He received an A.B. fromHarvard and a Ph.D. from Caltech (with J.D. Roberts). His present research interestsinclude reaction mechanisms, organo-metall ic chemistry, applied biochemistry,surface chemistry, catalysis, and materialsscience.

    Mark D. Bednarski is an Assistant Professorof Chemistry at the University of

    California, Berkeley (Berkeley, CA 94720;415-642-6372), He received his Ph.D.(1986) under the direction of SamuelDanishefsky and did postdoctoral workwith George Whitesides. His researchinterests include the synthesis of cell-surfacemimics and biocompatible materials, andthe design of proteins for syntheticchemistry.

    Ethan S. Simon is a doctoral candidate inorganic chemistry in George Whiteside'sgroup at Harvard University (1986-87

    DuPont Fellow). He holds an A.B. inbiochemistry from Harvard (1985). Hisresearch focuses on the development oftechniques for the combined chemical andenzymatic synthesis of sugars, nucleotides,and glycolipids.