Fine chemicals producers have flocked to international symposia, such as ChemSpec Europe 96 held recently in Basel, Switzerland, as part of a marketing strategy to present their latest technological advances.
"Specialist fine chemicals producers - sunrise or sunset?" This is the challenge that analyst René Willhalm of the consulting firm SRI International, Zurich, posed to the Conference on Pharmaceutical Ingredients (CPhI) in Frankfurt last November. "Over the last few years," he explained, "fine chemicals producers have come under increased competitive pressure, which is having a considerable effect on the way this traditionally profitable industry conducts its business."
At the ChemSpec Europe 96 symposium and exposition last month in Basel, Switzerland, Wolfgang Bernhagen of Hoechst delivered the industry's response to Willhalm's challenge: technological innovation. "Only the firm that can reduce an n-step synthesis to n-minus-one or n-minus-two will stay in the business," added Bernhagen, who manages new business development and European marketing of pharmaceutical and agrochemicals in Frankfurt.
The technological innovation achieved over the past year and demonstrated so strikingly in Basel covers a range of intermediates and processes from possible future high-tonnage workhorse compounds to intriguing research specialties in gram lots. And exhibits of chiral drug intermediates in near-100% enantiomeric excesses stood side by side with offerings of simple, symmetrical, yet versatile molecules.
Bernhagen presented examples of Hoechst's cultivation of three organic reaction types: the Heck reaction of olefins to form arylolefins, replacement of aryl nitro groups by chlorine atoms, and air oxidation. The company's aim in all this is to cut steps and substitute cleaner chemistry to provide intermediates and make contract manufacturing available to customers.
"Pharmaceuticals like naproxen are just begging for new synthetic routes," Bernhagen said, and he showed three such routes. "Quite astoundingly, all of them start from the same intermediate 6-methoxy-2-vinylnaphthalene, which is accessible via a Heck reaction." Once a chemist reaches the vinyl compound by reaction of bromonaphthalene with ethylene, Bernhagen showed, the possibilities next are asymmetric hydroformylation or hydrocyanation to the enantiomeric aldehyde or nitrile, respectively, or carboxylation and resolution.
For what he calls denitrating chlorination, Bernhagen described the deceptively simple synthesis of o-chlorofluorobenzene, which is useful to make antibacterial drugs and pesticides. His preferred route is reaction of commercially available o-fluoronitrobenzene with chlorine. Among the rejected routes, reduction of the nitro compound to aniline, followed by diazotization and a Sandmeyer reaction, has more steps, which makes it more costly. Also, conversion of o-chloroaniline to a diazonium fluoborate and strong heating of that uses more expensive materials. And chlorination of fluorobenzene gives the ortho product in only 10% yield.
Reemphasizing his original theme of the need for innovation in fine chemicals, Bernhagen said: "In search of more competitiveness, environmentally cleaner reactions, and more intelligent solutions to problems, the fine chemicals industry will continue to test new chemical reactions. The future will show which ones will be added to the existing arsenal of well-established chemical reactions."
Intermediates debut
Among new offerings in fine chemicals
is &agr;,&agr;-dichlorodimethyl ether in commercial quantities from chemical, defense, and aerospace
giant SNPE of France, which
the firm trade names Chloromyl. This compound has a
wide range of reactions, including conversion of organic acids to acid chlorides, addition of carbonyl groups to
molecules, and generation of
methoxycarbene.
Chloromyl has been limited to existence as a laboratory curiosity for 40 years, though, because the preparation was a messy chlorination of methyl formate with phosphorus pentachloride. SNPE has devised a cleaner catalytic reaction of methyl formate with phosgene, with carbon dioxide as the only coproduct.
Using Chloromyl, phenylacetic and adipic acids are converted to their acid chlorides with volatilization of hydrogen chloride and methyl formate (bp 32 °C). Titanium(IV), tin(IV), or aluminum chlorides serve as catalysts in formylation of mesitylene to 2,4,6-trimethylbenzaldehyde. The coproducts are hydrogen chloride and methanol. Treatment with methyllithium converts the chloroether to methoxycarbene, which in turn reacts with tetramethylethylene to form the corresponding methoxycyclopropane.
Elsewhere, Austria-based DSM Chemie Linz has expanded use of guanidine carbonate to synthesis of pyrimidine building blocks for weed killers. The company buys guanidine carbonate from its former sister company, Agrolinz Melamin, which in turn gets it as a coproduct during production of melamine monomer.
Guanidine carbonate reacts with such dicarbonyl compounds as acetylacetone to yield 2-amino-4,6-dimethylpyrimidine, a raw material for DuPont's herbicide sulfometuron methyl. Reaction with dimethyl malonate gives the 4,6-dihydroxy derivative. Phosphorus oxychloride replaces the hydroxy groups with chlorine atoms, and treatment of the 4,6-dichloro compound with sodium methoxide forms the 4,6-dimethoxy intermediate, which is a raw material for several herbicides, such as Monsanto's halosulfuron methyl.
Research group leader Gerhard C. Stucky of Basel-based Lonza described an alternative route to the 2-amino-4,6-dimethoxypyrimidine herbicide intermediate in Basel. This route uses the company's malononitrile and installs the methoxy groups early on by a sequence that avoids chlorination with phosphorus oxychloride and the resulting problem of treating phosphate waste. Stucky's malononitrile route might thus become the more economical way to make the dimethoxy compound. However, someone might rehabilitate the guanidine carbonate route with a cleaner chlorination step based on phosgene.
Also in Basel, research group leader Jim Schwindeman of FMC Corp.'s lithium division, based in Gastonia, N.C., described a series of functionalized organolithium reagents newly available from the company as polymerization initiators or synthetic intermediates. The new products are &ohgr;-tert-butyldimethylsilyloxyalkyllithiums in cyclohexane solutions.
The reagents are more stable in hydrocarbon than in ether solvents, and have the advantage of being nonpyrophoric. The tert-butyldimethylsilyl group is easily removed, and the resulting silanol is relatively harmless. The original technology of producing the organolithium reagents in such solutions was invented by the Defense Evaluation & Research Agency of the U.K., and FMC licensed that patent. Alkyl groups available include n-propyl, 2,2-dimethylpropyl, n-hexyl, and n-octyl.
In another development, Germany-based Degussa is test-sampling the ethylene acetal of acrolein (2-vinyl-1,3-dioxolane) as a means of shipping and using that toxic, lachrymatory aldehyde in protected form. Customers interested in operations on the carbon-carbon double bond might carry those out on the acetal before removing the ethylene glycol at a later step.
In perhaps the most exotic recent fine chemicals offering, BNFL Fluorochemicals of the U.K. and Oakwood Products of West Columbia, S.C., are betting that the - SF5 functional group may be as valuable as trifluoromethyl has been to impart biological activity to compounds. The two firms are offering research quantities of 4-nitrophenylsulfur pentafluoride (p-O2NC6H4SF5) at $185 per g and the 3-nitrophenyl compound at $275 per g. Roy D. Bowden, who heads up new product and process development at BNFL, said the firm would welcome inquiries for kilogram amounts. BNFL makes both compounds by direct fluorination of the corresponding nitrothiophenols. Oakwood sells the compounds in the U.S.
German chemical company Bayer is test-sampling &agr;-chloroacrylonitrile, a multifunctional intermediate that undergoes about two dozen different reaction types. The company chlorinates acrylonitrile to &agr;,&bgr;-dichloropropionitrile, then cracks hydrogen chloride out of that to get the unsaturated nitrile.
New chiral products
Given the increasing importance of chirality in drug and agrochemical
development, it is not surprising that such heavy hitters as Air Products
& Chemicals, BASF, and Zeneca should weigh in with entries for this
market. Air Products, Allentown, Pa., has introduced several dozen
enantiomeric secondary amines made by reactions of ketones with amines,
followed by asymmetric hydrogenations of the resulting imine Schiff bases.
The company introduced the new amines in May at a Chiral USA 96 exposition in Boston, but with no information about enantiomeric excesses. Even so, if enantiomeric excesses are halfway decent, the achievement is impressive, because Schiff bases are labile and capable of syn-anti geometrical isomerism. This makes their enantioselective hydrogenation difficult.
The primary amine starting materials for the Air Products chiral secondary amines are methyl-, ethyl-, and n-propylamine, methyl esters of alanine and valine, and 1-methoxy-2-aminopropane. The ketone starting materials are methyl ethyl and methyl isopropyl ketones, acetophenone and propiophenone, and 1-indanone.
The chiral amines offered by Germany-based BASF are all primary amines, made by enzyme-catalyzed enantioselective N-acylation of one isomer in a racemate. The company incubates a racemic amine such as &agr;-phenethylamine with ethyl methoxyacetate and a lipase in methyl tert-butyl ether solvent. The particular lipase used can acylate amines as well as hydrolyze lipids, so the products are R-methoxyacetamide and the unreacted S-amine. In this way, the company resolves 16 primary amines for sale in enantiomeric excesses greater than 95%.
Zeneca LifeScience Molecules clearly intends to participate in workhorse chiral intermediates but is being circumspect in talking about it for now. The recent foundation of the LifeScience Molecules business unit and name are suggestive. The company may be about to introduce enantiomeric secondary amines analogous to those of Air Products by asymmetric alkylation.
Of all this, business technology manager Geoff E. Evans at the company's Huddersfield, England, works, tells C&EN that Zeneca is undertaking a three-year program of process research into large-scale asymmetric synthesis. "This," he says, "will augment our process expertise in producing chiral molecules by biotransformation and kinetic resolution where we already have operating processes at scales up to thousands of [metric tons] per year."
Evans says three areas are encompassed by the program: chiral reduction, chiral alkylation, and chiral oxidation. "Many catalyst systems have been developed for homogeneous chiral hydrogenation," he explains, "but the results are very dependent on substrate structure, and enantiomeric selectivity often trails off on scale-up."
The majority of homochiral intermediates for the pharmaceutical and agrochemical industries can theoretically be produced via an enantioselective reduction, according to Evans. "However, this approach is not always successful, particularly where the prochiral substrate does not have a high degree of asymmetry. We have technology for asymmetric alkylation or arylation, which we have shown to be successful in cases where an approach via reduction fails to give good selectivity. Our aim is to get a better handle on the complementarity of approaches and the factors [that] enable good selectivity at large scale."
Among other workhorse intermediates, there are new developments in three- and four-carbon enantiomeric intermediates. Such simple building blocks are amenable to construction of a wide variety of structures.
The most obvious recent approaches to such compounds have been the asymmetric epoxidations, aminohydroxylations, and dihydroxylations of olefins of organic chemistry professor K. Barry Sharpless of Scripps Research Institute, La Jolla, Calif., and the asymmetric epoxidation reaction of organic chemistry professor Eric N. Jacobsen of Harvard University. But these methods are patented and licensed variously to ChiRex, Wellesley, Mass., and Sipsy, based in France.
Indeed, ChiRex managers went right from their booth at the Basel exposition to their plant site in Dudley, England, where they formally opened a $13.5 million research and development center-cum-pilot plant operating under current good manufacturing practices (cGMP). cGMP is the U.S. Food & Drug Administration's code of manufacturing, which continues to evolve with new additions and interpretations, and which fine chemicals producers worldwide must observe if they hope to participate in drug markets.
ChiRex itself is newly born of a merger between SepraChem, the fine chemicals producing arm of U.S.-based Sepracor, and Sterling Organics, which is the onetime Sterling Winthrop drug subsidiary of Eastman Kodak. The merger combines the proprietary enantioselective technology panoply of SepraChem with the large cGMP production capacity of Sterling. And the new technical center/pilot plant will bridge research to production by process chemistry.
Alternative routes that help customers avoid the patents licensed by ChiRex and Sipsy were presented to CPhI 95 in Frankfurt by Keishiro Nagao, marketing manager for chiral chemicals at Daiso Co. in Osaka, Japan, and to the Basel symposium by chiral chemistry manager Oreste Piccolo of Chemi in Cinisello Balsamo near Milan, Italy.
One Daiso approach is through microbial resolutions of 2,3-dichloro-1-propanol and 3-chloropropane-1,2-diol. Nagao pointed out that the enantiomeric diol is a synthetic equivalent to single-isomer glycidol. The diol has the advantages that it is stable on storing six months in the dark, is available in greater than 98% enantiomeric excess, and is easily convertible to glycidol in reaction mixtures, in which case there is no need to isolate the glycidol itself.
Daiso uses enantiomeric chlorinated propanols and propanediols to make and sell glycidyl derivatives. Nagao said he cautions customers, however, that enantiomeric glycidyl p-toluenesulfonates and m-nitrobenzenesulfonates are still covered in the U.S. and Canada by composition-of-matter patents licensed by Sipsy.
Chemi attacks the problem of the patents licensed by ChiRex and Sipsy from the pathway of such enantiomeric glycerol derivatives as acetonides (4-hydroxymethyl-2,2-dimethyl-1,3-dioxo-lanes). Piccolo pointed out that the five-membered ring acetonides are more stable than glycidyl epoxide rings. Also, the acetonides are less toxic than epoxides. And there are larger reactivity differences among the functional groups attached to the three carbons of glycerol in acetonides.
Piccolo and his coworkers have published two commercial routes to chiral acetonides. In one, they resolve diastereoisomeric salts of chiral amines with hydrogen phthalate esters of the acetonides. In the other, they selectively acetylate one glycerol-2,3 cyclic carbonate isomer by enzyme catalysis.
Chemi's approach from a glycerol starting point means that the company can also make such enantiomeric phospholipids as phosphatidylcholine, -cholamine, and -serine. In particular, Piccolo said, U.S. fans of nutritional supplements value phosphatidylserine as "a remarkable brain cell nutrient." The compound can be isolated from soybeans and eggs, but the synthetic technology is preferred for making large amounts.
NSC Technologies of Mount Prospect, Ill., has parlayed the enantioselective expertise gained through production of the synthetic dipeptide sweetener aspartame into syntheses of amino acids and other chiral intermediates. NSC Technologies is a subsidiary of Monsanto and a sister business unit within Monsanto of NutraSweet Kelco, which markets aspartame under the NutraSweet brand.
NSC Technologies had already branched out into amino alcohols, cyclohexyl amino acids and derivatives, and tetrahydroisoquinolines based on d- and l-phenylalanine and aspartic acid, the raw materials for aspartame. Now the company offers such nonnatural amino acids as l-&agr;-aminobutyric acid, "l-tert-leucine" (trimethylalanine), d-phenylalanine, and d-isoleucine, made by actions of d- or l-transaminase enzymes on the corresponding &agr;-keto acids. A wide range of &agr;-keto acids comes from reaction of organolithium compounds with diethyl oxalate, followed by hydrolysis. Also available from NSC are such chiral auxiliaries and ligands as amino alcohols, oxazolidinones, and bis(oxazolines) made from a broad spectrum of amino acids, as well as &agr;-hydroxy acids made by biocatalytic reductions and asymmetric hydrogenations of &agr;-keto acids.
NutraSweet Kelco has begun to supply NSC with sugars such as glucuronic acid and l-rhamnose, made by fermentation of polysaccharide gums. In addition, NSC will sell d-&bgr;-hydroxybutyric acid and enantiomeric compounds made from that as a result of Monsanto's purchase of Zeneca's polyhydroxybutyrate biodegradable resin business.
Another event taking place as the Monsanto-NutraSweet-NSC saga has played out has been the $50 million purchase of NutraSweet Kelco's production plant in University Park, Ill., by Takasago International Corp. The move makes Takasago an even bigger force in commercial chiral chemistry. As Takashi Miura, manager of the fine chemicals research laboratory in Kanagawa, Japan, told Chiral USA 96 in Boston, the company has produced 1,000 metric tons per year of synthetic menthol by asymmetric hydrogenation since 1984. The company will use the University Park plant to make aroma chemicals.
Also in Boston, Miura described an example of Takasago's ingenuity in asymmetric transformations. The object is to make an azetidone linked at one corner to the &agr;-carbon of a propionic acid molecule. The azetidone is to be the &bgr;-lactam of a semisynthetic penicillin or cephalosporin. The carboxyl group of the propionic acid serves to cobble a five- or six-membered ring alongside. The methyl tail of the propionic acid becomes a therapeutically efficacious methyl substituent on that second ring.
Miura and his coworkers tack on the propionic acid group by base-catalyzed addition of diallyl methylmalonate to the azetidone. They subject this diallyl methylmalonate substituent to hydrogenolysis, which converts the diallyl ester to a methylmalonic acid, which, in turn, decarboxylates to the propionic acid.
The overall sequence is doubly elegant. First, hydrogenolysis of a diallyl ester is more atom-economical than hydrogenolysis of a dibenzyl ester would be. Second, the Takasago team does not use high-pressure hydrogenation but catalyzed hydrogen transfer from ammonium formate. Hydrogen transfer does not need the high-pressure equipment that hydrogenation does and is well suited to such small-batch operations.
Just as at NSC Technologies, single-enantiomer natural and nonnatural amino acids are also very much Degussa's business, which the firm carries on through its subsidiary Rexim in Courbevoie, France. Karlheinz Drauz from Degussa's technical center in Hanau, Germany, outlined some of these activities in Boston. Drauz combines the directorship of research for pharmaceuticals and intermediates at Degussa with a professorship in organic chemistry at the University of Würzburg, Germany.
The Degussa approach to tert-leucine is from trimethylpyruvic acid, which the company gets from pinacolone. The enzyme leucine dehydrogenase uses ammonium ion to replace the keto group of trimethylpyruvic acid with an amino group reductively to form d-tert-leucine. The enzyme cofactor is regenerated continuously by oxidation of formate ion.
Alternatively, the company reacts pivalaldehyde with hydrogen cyanide and ammonium carbonate in a Strecker synthesis of racemic tert-butylhydantoin. A d-hydantoinase hydrolyzes that to d-tert-leucine. In addition to serving as a building block for several drug companies' peptidelike drugs, tert-leucine is part of a chiral auxiliary available from Degussa. The company sells both enantiomers of N-&bgr;-naphthoyl-tert-leucine for this use.
Degussa has begun production of l-trimethyllactic acid by reduction of trimethylpyruvic acid by hydroxyisocaproic acid dehydrogenase. The company derives a variety of chiral diols and amino alcohols from that intermediate.
In other developments, several chiral building blocks are newly available from Synthon of Lansing, Mich., a start-up company founded by organic chemistry professor Rawle I. Hollingsworth of Michigan State University in East Lansing. The company commercializes his process to make (S)-&bgr;-hydroxy-&ggr;-butyrolactone from cheap carbohydrate raw materials.
For example, Hollingsworth buys lactose, which is a by-product of cheese making, from a Land O'Lakes plant locally for 8 to 14 cents per lb. A sequence of alkali- and hydrogen-peroxide-mediated reactions cleaves and converts the glucose portion of the disaccharide to the enantiomeric butyrolactone. In addition to multikilogram amounts of the lactone, produced so far for him in a pilot plant at Zeeland Chemicals, Zeeland, Mich., Hollingsworth has worked out conversion of that lactone to three dozen or so other chiral intermediates, which Synthon offers in gram amounts.
Newest offerings from Salford Ultrafine Chemicals & Research, based in Manchester, England, are enantiomeric bicyclic lactams originally developed by organic chemistry professor Albert I. Meyers of Colorado State University in Fort Collins. Available from Salford Ultrafine in 1- to 25-g lots, these intermediates are useful to make asymmetric cyclopentenones, cyclopropanes, pyrrolidines, and compounds with quaternary asymmetric atoms.
For example, levulinic acid reacts with enantiomeric phenylglycinol to form single-isomer 5-methyl-2-phenyl-4-oxa-1-azabicyclo[3.3.0]octan-8-one. A further step introduces a C6 - C7 double bond. Aldrich Chemical, Milwaukee, has also offered these enantiomeric lactones for a few years.
Interestingly, neither Meyers nor Colorado State ever patented the lactams. "I'm not a patenting person," smiled Meyers. "I don't believe academic research should be used in that way." He added that the National Institutes of Health funded exploration of the lactam chemistry over almost a decade.
U.K.-based Oxford Asymmetry has also deliberately chosen to leave its new BINAP process unpatented. Said Managing Director Edwin Moses, "It's so hard for a small firm to police patents, so our patent lawyers advised us to keep it as 'know-how.' " BINAP is the acronym for 2,2'-bis(diphenylphosphino)-1,1'-binaphthyl, a chiral ligand available as either enantiomer useful to make asymmetric hydrogenation, hydroformylation, and hydrocyanation catalysts.
In another development, Spanish company Deretil has introduced d-(-)-&agr;-(o-hydroxyphenyl)glycine as a side chain for semisynthetic penicillin and cephalosporin antibiotics. Manuel Esteban, director of technical and business development for Deretil, suggested that the o-hydroxyl group will contribute greater water solubility than with p-hydroxyphenyl side chains, leading to greater bioavailability of such antibiotics.
Synthetic organic technology
In addition to developing intermediates for sale, fine chemicals companies
are devising technologies that they can
carry out to special order. Sometimes
the reaction is hazardous, and a customer would rather leave it to experts.
Or perhaps the technology is ideal for
the customer's target compound but
proprietary to the inventor firm.
One example of touchy chemistry is the custom sulfopropylation offered by Germany-based Raschig. The company uses 1,3-propane sultone, which is the cyclic internal ester of 3-hydroxypropanesulfonic acid, for sulfopropylations of alcohols, amines, and carboxylic acids. The process introduces a sulfonic acid group without the severity of a sulfonation reaction.
Thus, methacrylic acid reacts to form a sulfopropyl ester, which may be useful as a water-soluble monomer. Mor-pholine yields the N-sulfopropyl compound called MOPS, which is a buffer compound to maintain physiological pH in biological experiments. And ethoxylated p-nonylphenol affords a sulfopropyl anionic surfactant.
The 1,3-propane sultone may come from reaction of propylene with sulfuric acid or sodium bisulfite. The sultone is a possible carcinogen, so chemists may prefer to leave the sulfopropylation with Raschig.
Meanwhile, Catalytica of Mountain View, Calif., has found a way to make indolones by catalytic oxidation of indoles. Another of the firm's processes yields unsymmetrically substituted biaryls.
The virtue of the indolone process is that it avoids a route through isatin as well as use of highly reactive chlorosulfonyl isocyanate. Catalytica originally worked out a synthesis of 5-chloroindol-2-one as an intermediate for Pfizer's anti-inflammatory drug tenidap. Pfizer's original process presented a certain impurity profile based on isatin as an intermediate. Catalytica's process eliminates the impurities of the isatin pathway without introducing any new ones. Catalytica now has a portfolio of proven methods to make a number of chloro-, bromo-, and methoxyindolones for customers.
The unsymmetrically substituted biaryls may be intermediates for anti-inflammatory drugs that act by inhibition of phospholipase A2. The Catalytica route to these compounds specifically avoids Suzuki coupling of aryl halides with aryl boron derivatives. Suzuki coupling cannot make chlorobiaryls.
Yet another bit of new technology is a method of making unsymmetrical hydrazines without having to use carcinogenic nitrosamine intermediates. A number of promising oral antidiabetic agents, diuretics for high blood pressure, and antiprotozoal and antituberculosis drugs feature various substitution patterns about the N - N bond.
Vittorio Rossetti of Laporte Organics Francis in Caronno Pertusella near Milan reported to the Basel meeting on his company's method to make such compounds. Rossetti, who is manager of research and development and commercial development, gave the example of a new approach to Upjohn's oral antidiabetic drug tolazamide.
Instead of nitrosating hexamethyleneimine, which is one end of an intended unsymmetrical hydrazine structure, Francis' chemists treat it with sodium cyanate, which produces N,N-hexamethyleneurea. This unsymmetrical urea is only a Beckmann rearrangement away from N,N-hexamethylenehydrazine. Reaction of that hydrazine with p-toluenesulfonylurea forms tolazamide.
The overall yield in the plant is only 30 to 50%. But Rossetti explained that the method avoids nitrosamines and results as well in little waste to treat. So he said the plant is overall economically viable.
Biotechnology in new processes
In addition to processes based on synthetic organic chemistry, yet other
recent innovations owe their invention to biotechnology. One firm that is
deepening its commitment to biotechnology is Lonza.
In 1992, the company acquired a fermentation plant in the Czech Republic. Interestingly, the plant had been used to make food products. That means that the five 4,000-gal fermentors and the five 13,000-gal ones are larger than those often found in drug-producing plants, yet not as massive as those in breweries - in short, just right for fine chemicals.
Last month, Lonza announced acquisition of Celltech Biologics for $64 million from Celltech Group of the U.K. (C&EN, June 24, page 14). The acquired firm specializes in producing monoclonal antibodies and recombinant proteins in mammalian cell culture. In addition to a 45,000-sq-ft producing plant in Slough, near London, and a new 76,000-sq-ft plant in Pease, N.H., Lonza is getting the company's glutamine synthase (GS) system of gene expression, which stimulates product protein yields of more than 1 g per L.
The GS approach uses some mammalian cells that do not produce enough glutamine to survive in a glutamine-free medium. By inserting gene sets containing GS genes in addition to genes for the desired product protein, biotechnologists can culture all cells in glutamine-free medium to select for those that have the inserted gene set.
In the case of mammalian cells that normally have enough GS to survive, biotechnologists add methionine sulfoximine, which is a GS inhibitor, to the medium. As a result, only the cells that have extra inserted GS genes survive.
Other biotechnological developments involve enzymes and other proteins for specialized chemical reactions and molecular recognitions. For example, Altus Biologics of Cambridge, Mass., has introduced products based on cross-linked enzyme crystals.
The company crystallizes enzymes in very pure form, then treats the crystals with a reagent like glutaraldehyde, which cross-links peptide molecules through pendant amino groups on lysine residues. The resulting crystals retain enzymatic activity and function stably in organic solvents and at high temperatures.
Altus got a big boost last month as a result of an agreement with Ciba-Geigy of Switzerland to develop cross-linked enzymes for liquid laundry detergents (C&EN, June 3, page 11). The big soap producers routinely add enzymes to solid detergents, but there have been storage stability problems in the highly alkaline climate of liquids. First efforts will be to apply the protease subtilisin for removal of proteinaceous stains on clothing. Beyond the immediate money to be made, this project will multiply the company's production manyfold and deepen its expertise.
The company introduced a kit of enzymes at the CPhI Frankfurt exposition for chiral resolution. The kit, trade named ChiroKit TE (the TE stands for transesterification) contains 25 biocatalysts for acylation of racemic alcohols and amines. As Michael D. Grim explained at the Basel symposium, the kit is supposed to be used in conjunction with the earlier introduced ChiroKit EH (ester hydrolysis) to see whether an acylation or hydrolysis is best for kinetic resolution of a test compound. Grim, who is a technical support chemist, said that after the customer has narrowed the field down to, say, three enzymes, Altus will ship those three in larger amounts for optimization of the process.
Also in Basel, Altus introduced a penicillin acylase derived from Escherichia coli for phenylacetylation of amines or hydrolysis of phenylacetamides. In addition to kinetic resolution of amines, the company suggested uses to add phenylacetyl protecting groups or removal of protecting groups.
Recombinant Biocatalysis of Sharon Hill, Pa., uses another approach to secure enzymes for use at high temperatures and other extreme conditions. The company isolates enzymes from bacteria that thrive in deep-sea vents, geothermal pools, the Arctic and Antarctic, and rain forests. These isolated aminotransferases, esterases, lipases, glycosidases, and phosphatases are adapted to function in wide ranges of temperatures, pressures, solvent conditions, and pH.
The company has cloned genes for 150 such enzymes and produces them in familiar organisms like E. coli. Beyond hunting for wild enzymes capable of functioning in extreme environments, the company offers to modify these still further to adapt them very specifically to a customer's application.
During the past year, Boehringer Mannheim of Germany received worldwide marketing rights to two lipases and two proteases from Novo Nordisk. These additions swell the already massive array of Chirazyme chiral enzymes for industrial use.
Elsewhere, use of the phthalimido group as an amino protecting group may get a closer look as a result of development of a phthalimidase enzyme at Eli Lilly, Indianapolis. Reactions of amines with phthalic anhydride to protect them as phthalimides date back to 1887. But modern synthetic chemists have avoided this protective group because of harsh alkaline hydrolyses or hydrazinolyses needed to remove it. An enzymatic deprotection would be gentler.
Research scientist Milton Zmijewski told an audience at Chiral USA 96 in Boston of finding the enzyme in the soil bacteria Xanthobacter agilis. The enzyme dissociates the phthalimide linkage whether the rest of the molecule is chiral or not. Thus it can be a general means of protecting group removal. As to chiral preferences in particular, the phthalimidase lyses the phthalimide of d-phenylglycine rather than the L-isomer, Zmijewski said, but little else is known of its enantioselectivity.
In addition to enzymes, biotechnology serves to discover protein ligands that will bind to a target molecule to isolate a target compound from a complex mixture that may contain similar substances, resolve a racemate, or purify a preparation of the target compound. These proteins can be immobilized on column packings for use in this way.
At Dyax Corp., Cambridge, Mass., the tactic is to start with a simple, known protein, make a library of 10 million or 20 million variations on the amino acid sequence of that protein, and screen the library for affinity to the target compound. As John Maclennan, director of new business development, said in Basel, Dyax biotechnologists begin with a microprotein of low molecular weight, about 800 to 7,000, with two or three cystine disulfide bonds.
An example is aprotinin, the usual biological activity of which is inhibition of the protease trypsin. To Dyax engineers, its value is that its known sequence of 58 amino acids can be varied by as much as 60% by DNA mutation techniques to generate several million variations of different charges, hydrophobicity or hydrophilicity, and three-dimensional shape, and therefore different binding capabilities.
Maclennan estimates that it takes three to six months' work and $100,000 to develop a single microprotein with optimum binding to the client's target compound. Thus, he said, it is cheaper overall than column packings based on monomolecular chiral reagents. And the small, highly bridged structures are more specifically binding and robust than monoclonal antibodies, surviving a month in 1 M phosphoric acid. As Maclennan put it, "They're bulletproof."
To make and carry each peptide in the intended library, Dyax needs a "genetic package." The genetic package is a bacteria-infecting phage virus called M13. The varied DNA for the microprotein is spliced to what is called gene III of the phage. Protein expressed by gene III is displayed on the phage surface, jutting from one end. Thus, all operations are carried out by processing and replicating batches of phage and not of unbound microprotein.
Dyax screens the library by passing it through a column of immobilized target compound. Microprotein-bearing phage that binds to the column is recovered by elution with organic solvent. Phage that does not bind is discarded. If resolution of a racemate is desired, then Dyax passes phage over a column of the "wrong" enantiomer. This step selects out microprotein that binds to that unwanted isomer.
In addition to resolving racemates, immobilized microprotein columns separate compounds from a complex mixture. An example is separation of tissue plasminogen activator from the hundreds of other proteins in blood plasma. Or such columns can complete purification of a desired compound.
Fine chemicals flourish
The upshot of the Basel show is that speakers and exhibitors showed off a
broad and vigorous array of technology that attendees sampled with
enthusiasm. These ranged from Hoechst's proprietary replacement of aromatic
nitro groups by chlorine atoms to the carcinogen-free route to unsymmetrical
hydrazines pioneered by Laporte Organics Francis.
Such new intermediates appeared as the chlorinated ether from France's SNPE, and the chiral lactone from the start-up company of Michigan State's Hollingsworth. Air Products and BASF entered strong bids in the single-enantiomer market. Also, biotechnology furnishes several methods to make and isolate fine chemicals.
And just as Bernhagen of Hoechst and Willhalm of SRI International carried on a symbolic conversation, moving from the conference in Frankfurt to the symposium in Basel, participants in fine chemicals shows continue their give and take from one international venue to the next.
ACS Home Page