December 2001
Vol. 31, No. 12, pp 33–39.
Developing Technology

Table of Contents

Rob Schoevaart
Tom Kieboom

Combined catalytic reactions—Nature’s way

Going from traditional step-by-step methods to one-pot coupled conversion saves raw materials and energy and reduces waste.

figureOrganic synthesis—chemistry as it is done in the laboratory or manufacturing plant—traditionally uses a step-by-step approach. In a typical sequence, a starting material A is converted into a final product D, and intermediate products B and C have to be isolated and purified in each conversion step.

Such multistep organic syntheses are still quite common in today’s fine chemical industry, and they suffer from several disadvantages. They are often carried out noncatalytically using relatively large amounts of reagents that produce many kilograms of waste per kilo of final product. The separation and purification steps needed after each conversion step produce waste heat as energy is consumed. They require extra energy to overcome the thermodynamic hurdles to produce and isolate intermediates B and C if they lie in high-energy states.

figureOn the other hand, biosynthesis—chemistry as Nature performs it in the cells of organisms—goes through a multistep cascade to convert starting material A to final product D without separation of intermediates B and C.

Such multistep combined syntheses are common in everyday life. They are carried out in a fully catalytic way by using enzymes with relatively limited amounts of reagents (cofactors) and thus produce much less waste. The mutual compatibility and high selectivity of the enzymatic conversions make it possible to proceed without intermediate recovery steps. They save energy by avoiding the separation and isolation of intermediates B and C.

The potential power of combined catalytic conversions to overcome thermodynamic hurdles in multistep syntheses is demonstrated here for the well-known glycolysis pathway.

figureThe equilibria between fructose 1,6-diphosphate (A), dihydroxyacetone phosphate (B), and glyceraldehyde 3-phosphate (C) are quite unfavorable: [A]/[B]/[C] > 1000:20:1. Nevertheless, complete transformation of A to lactic acid (D) occurs through a coupled multienzymatic conversion of the more highly energetic C intermediate. In addition, the success of this overall conversion is brought about by the very fast aldolase- and isomerase-catalyzed equilibria, A left right arrow B and B left right arrow C, respectively.

Challenges in synthesis
For catalysis of the next generation of organic synthesis, the challenge is to

  • combine the power of chemical, enzymatic, and microbial conversions;
  • overcome problems associated with the disparities in the sizes of catalytic species (see the box, “Biological and chemical diversity”);
  • search for multistep conversions that do not require recovery steps—one-pot multicatalytic procedures—such as Nature uses; and
  • fine-tune reaction conditions and catalytic systems to permit the desired concerted reactions without intermediate isolation or purification steps.

The result of these efforts should drastically diminish the costs of energy consumption and waste treatment characteristic of current multistep processes in the fine chemical industry.

For the next generation of biosynthesis (microbial conversions), the challenge is to

  • improve productivity in terms of carbon source efficiency of microorganisms, toward a much higher product/ biomass ratio;
  • broaden the scope of products, from natural products toward the modified products often required for drugs; and
  • reengineer metabolic pathways as the most efficient way to reach these two goals.

These steps should free fermentation processes from excessively using renewables that end up in undesired, low-value biomass byproducts and minimize additional chemical modification steps to obtain the final product.

An industrial example—Biotechnology
The challenges we have presented for organic synthesis and biosynthesis are in no way wishful thinking. This is demonstrated by the evolution during the past 30 years of the industrial process route for the synthesis of the antibiotic cephalexin by DSM (1).

In the 1970s, cephalexin synthesis consisted of one fermentation reaction (step 3) together with seven organic synthesis steps (2):

1. Benzaldehyde → D-phenylglycine

2. D-Phenylglycine → D-phenylglycine chloride

3. Sugar + phenylacetic acid → penicillin G

4. Penicillin G → penicillin G sulfoxide

5. Penicillin G sulfoxide → penicillin G sulfoxide trimethylsilyl ester

6. Ring enlargement of penicillin G sulfoxide trimethylsilyl ester to fully protected 7-aminodeacetoxypenicillanic acid (7-ADCA)

7. Hydrolysis of fully protected 7-ADCA to 7-ADCA

8. D-Phenylglycine chloride + 7-ADCA → cephalexin

This multistep sequence used large amounts of energy (e.g., in low-temperature conversions in steps 5–7 and many recovery and solvent recycle steps), reagents and organic solvents (steps 2 and 4–8), and phenylacetic acid (no recycling was possible after the chemical hydrolysis step 7).

Figure 1. Chemistry, enzymatic conversion, and fermentation work together to make cephalexin.
Figure 1. Chemistry, enzymatic conversion, and fermentation work together to make cephalexin.
The integration of chemistry, enzymatic conversion, and fermentation (Figure 1) is the basis of an efficient, economic, environmentally friendly production method. By 2000, cephalexin synthesis consisted of an organic synthesis step (1), a fermentation step (2), and two enzymatic steps (3 and 4):

1. Benzaldehyde → D-phenylglycine amide

2. Sugar + adipic acid → 7-ADCA adipic acid amide

3. 7-ADCA adipic acid amide → 7-ADCA + adipic acid (recycled for use in step 2)

4. D-Phenylglycine amide + 7-ADCA → cephalexin

This sequence avoids low-temperature conversions, most of the previously used reagents and organic solvents, and consumption of adipic acid. The key improvements were

  • enzymatic coupling of D-phenylglycine amide with 7-ADCA in water (avoiding the synthesis of the acid chloride and the use of reagents and organic solvents);
  • a metabolic pathway–engineered microorganism that ferments sugar directly to the desired 7-ADCA moiety (avoiding three chemical conversion steps) (1); and
  • aqueous enzymatic hydrolysis of the adipyl side chain that allows easy recycling of the adipic acid for reuse in the fermentation step.

The following reagents (often used in more than stoichiometric amounts) and solvents (that could not be recycled completely) are no longer needed in the new cephalexin synthesis (2):


Peracetic acid
Phosphorus pentachloride
Trimethylsilyl chloride Dichloromethane
Bis(trimethylsilyl)urea n-Butanol
Pyridine hydrobromide Phenylacetic acid
Trimethylamine hydrochloride D-Phenylglycine acid chloride hydrochloride

The two-step process reduced solvents by ~90%, auxiliary reagents by 100%, energy consumption by ~40%, and waste by ~15%. Highlights of the process are in the box, “DSM’s process for 7-ADCA”.

DSM's process for 7-ADCA
  • The process is based on the high productivity of the Penicillium strain.
  • The Penicillium strain required the implanting expandase enzyme genes from a Cephalosporium strain.
  • The engineered strain is able to convert sugar directly to the 7-ADCA moiety.
  • The more efficient process makes an existing chemical plant starting from Penicillium unnecessary.
  • The new process requires a U.S. $25 million production unit scheduled to start up in late 2001 at DSM in Delft, The Netherlands.
  • This is the first large-scale example to demonstrate the power of metabolic pathway engineering.
Combinations of enzymes and chemical catalysts
In addition to the use of metabolic pathway–engineered microorganisms on a multiton scale, several promising multistep one-pot catalytic conversions have been described on the laboratory and industrial scale, using either a combination of enzymes or a combination of an enzyme and a chemical catalyst (see Table 1).

We can sum up the current state of this field as follows: Most of the multistep one-pot conversions have been reported in carbohydrate synthesis, using combinations of enzymatic conversions. Combinations of metal-catalyzed and biocatalytic conversions are not yet common. After some early examples in the 1980s, more than a decade of inactivity ensued, followed by an increasing interest during the past few years for combined catalytic conversions.

The very first example of the combined action of an enzyme and a metal catalyst was the direct one-pot conversion of D-glucose to D-mannitol. In this process, the enzyme glucose isomerase converts glucose to a ~1:1 glucose–fructose mixture and ensures that this mixture remains in equilibrium. At the same time, the copper catalyst preferentially hydrogenates the fructose to mannitol (Figure 2).

Figure 3 Complete conversion of a racemate into an optically pure compound by the cooperation of two catalysts in one pot. Conditions: 70 ºC, cyclohexane solvent.
Figure 2. An early example of combined enzyme and metal catalysts: The one-pot conversion of D-glucose to D-mannitol. Conditions: 70 atm H2, 70 °C, aqueous solvent, pH 7.5 (3 ,4).
At first glance, this combination approach appears to be quite simple, but in practice several fine-tuning measures were taken to achieve a balanced “cooperation” of the two simultaneous catalytic conversion steps:

  • The enzyme had to be immobilized on silica to prevent poisoning of the copper metal by protein sulfur groups.
  • The enzyme was protected by a chelating agent (EDTA) to avoid inhibition by traces of copper ion from the catalyst.
  • The correct compromise of hydrogen pressure and temperature had to be used to meet the stability and activity requirements for both catalyst systems.
  • The system had to be maintained at a slightly basic pH to avoid the mutarotation of D-glucose. (The interconversion of the α- and β-glucopyranose forms becomes rate-limiting because the enzyme only converts the α-form.)

The kinetic picture is quite complicated (Table 2). The three types of kinetics are expressed in TONs (turnover numbers). In the enzymatic isomerization (Michaelis–Menten kinetics), only two of the six sugar forms are substrates for the enzyme. In the heterogeneous hydrogenation on copper, the adsorption constants vary from 3 to 10 L/mol. Only ~25% of the copper surface is covered with D-fructose, but that sugar reacts much faster than the adsorbed glucose. In the acid–base homogeneous catalysis of mutarotation, the rate for D-glucose is 50 times slower than that for D-fructose.

This is a good example of the selective conversion of one of two enantiomers in equilibrium. The fine tuning of simultaneous catalytic conversion is important for the synthesis of enantiomerically pure compounds in 100% yield from racemic starting materials.

figure

Some recent examples on the laboratory scale have been reported (11–15) for the concomitant action of a chemical and a biocatalyst for the required racemization and enantioselective conversions, respectively.

figure

Figure 3 Complete conversion of a racemate into an optically pure compound by the cooperation of two catalysts in one pot. Conditions: 70 ºC, cyclohexane solvent.
Figure 3. Complete conversion of a racemate into an optically pure compound by the cooperation of two catalysts in one pot. Conditions: 70 ºC, cyclohexane solvent.
In this way, racemic 1-arylalcohols can be converted to esters with high enantiomeric excess and in high yield using transition-metal catalysts with lipases in organic solvents (see Figure 3). There is no doubt that this kind of synthesis from relatively inexpensive racemates will find its way into the fine chemical industry in the near future— as already demonstrated by the industrial hydantoinase process for amino acids, in which spontaneous racemization occurs.

A multienzyme one-pot example
Another elegant multistep one-pot approach recently developed (5, 6) is the four-enzyme–catalyzed, four-step, one-pot conversion of glycerol into a D-heptose sugar, in which a pH switch method is applied to temporarily turn off the phytase enzyme during the second and third steps of the concerted synthesis (Figure 4).

The four consecutive enzymatic conversion steps in one reactor without any separation of intermediates consist of

  • Phosphorylation. Glycerol is phosphorylated with pyrophosphate by phytase at pH 4.0 at 37 ºC. Racemic glycerol 3-phosphate is obtained in 100% yield (based on pyrophosphate) in 95% glycerol after 24 h.
  • Oxidation. By raising the pH to 7.5, phytase activity is “switched off”, and hydrolysis is prevented. Oxidation of l-glycerol 3-phosphate to dihydroxyacetone phosphate (DHAP) by glycerol-L-phosphate oxidase (GPO) at 55 vol% glycerol is quantitative. Catalase is added to suppress the build-up of hydrogen peroxide. The D-isomer is converted back to glycerol and phosphate in the last step.
  • Aldol reaction. More than 20 aldehydes are known to be substrates for the aldolases from Staphylococcus carnosus and S. aureus. Stereoselectivity of the aldolases must be looked at for each acceptor substrate, because isomers are formed in different proportions. The oxidation and aldol reaction can be carried out simultaneously.
  • Dephosphorylation. Lowering the pH back to 4 “switches on” phytase’s activity, and hydrolysis of the aldol adduct is initiated.

Combined with the broad substrate specificity of DHAP aldolases, this sequence constitutes a simple procedure for the synthesis of a wide variety of carbohydrates from readily available glycerol and pyrophosphate (5, 6).

Investigations of such multistep synthesis methods without isolation of intermediate products require appropriate in situ analytical methods so that the investigator knows what is happening during the consecutive conversions. The combination of NMR spectroscopy with starting materials selectively enriched in isotopes such as 13C, 15N, and 17O provides a powerful window (18, 19). In this way, a sequence of conversions can be well characterized, even in complicated matrices of catalysts, reagents, and mixed solvent systems.

This approach is illustrated in Figure 5 for the cross-linking of in situ–generated galactohexodialdose from lactose protein mixtures, as studied by using selective 13C labeling (20). The undesired background from 13C NMR signals of the other components of the reaction mixture is suppressed, allowing straightforward detection of the reaction species of interest.

A synthesis renaissance
In conclusion, the concepts and various examples given show that we may foresee a renaissance in synthesis methods by integration of bio- and organic syntheses for fine chemicals by one-pot multistep catalytic procedures and by metabolically engineering the microbes that perform the catalysis (21). Such a merger of chemistry and biotechnology through a cell-factory catalysis approach will include

  • multistep catalytic conversions, including acids, bases, metals, enzymes, and microorganisms;
  • combinations of these catalysis methods, including integration of conversion and separation techniques in continuous processing; and
  • use of isotope labeling and NMR as the analytical tools to study such complex transformations in situ.

In this respect, future clean synthesis methods should be inspired by the achievements in the field of modern detergent formulations that have up to six different enzymes in them (22): an advanced multicatalytic one-pot conversion of “dirty laundry” to “clean laundry” plus “dirt”, with the washing machine as the in-house catalytic reactor that simultaneously separates the product (clean laundry) from waste (dirt).

Economic and environmental goals will cause chemical and biotechnological conversion methods to merge into integrated biology–chemistry routes, based on optimum feedstocks. Conservation of matter and energy from starting material to end product is required to achieve sustainable conversion processes. Neither chemistry nor biotechnology, neither fossil fuels nor renewable feedstocks will be the ultimate winner—only the combination of these disciplines and resources.

References

  1. DSM Magazine 1998, 147, 18–21.
  2. Verweij, J.; de Vroom, E. Recl. Trav. Chim. Pays-Bas 1993, 112, 66–81.
  3. Makkee, M; Kieboom, A.P.G.; van Bekkum, H.; Roels, J. A. J. Chem. Soc., Chem. Commun. 1980, 930–931.
  4. Makkee, M.; Kieboom, A.P.G.; van Bekkum, H. Carbohydr. Res. 1985, 138, 237–245.
  5. Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. Chem. Comm. 1999, 2465–2466.
  6. Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. Tetrahedron: Asymmetry 1999, 10, 705–711.
  7. Inagaki, M.; Hiratake, J.; Nishioka, T.; Oda, J. J. Am. Chem. Soc. 1991, 113, 9360–9361.
  8. Bommarius, A. S.; Drautz, K.; Hummel, W.; Kula, M. R.; Wandrey, C. Biocatalysis 1994, 10, 37–47.
  9. Galkin, A.; Kulakova, L.; Yamamoto, H.; Tanizawa, K.; Tanaka, H.; Esaki, N.; Soda, K. J. Ferment. Bioeng. 1997, 83, 299–300.
  10. Orru, R.V.A.; Mayer, S. F.; Kroutil, W.; Faber, K. Tetrahedron 1998, 54, 859–874.
  11. Reetz, M. T.; Schimossek, K. Chimia 1996, 50, 668–669.
  12. Larsson, A.L.E.; Persson, B. A.; Bäckvall, J. E. Angew. Chem., Int. Ed. 1997, 36, 1211–1212.
  13. Allen, J. V.; Williams, J.M.J. Tetrahedron Lett. 1996, 37, 1859–1862.
  14. Lee, D.; Huh, E. A.; Kim, M.-J.; Jung, H. M.; Koh, J. H.; Park, J. Org. Lett. 2000, 2, 2377–2379.
  15. Dinh, P. M.; Howarth, J. A.; Hudnott, A. R.; Williams, J.M.J.; Harris, W. Tetrahedron Lett. 1996, 37, 7623–7626.
  16. Wong, C.-H.; Haynie, S. L.; Whitesides, G. M. J. Org. Chem. 1982, 47, 5416–5418.
  17. Koeller; K. M.; Wong, C.-H. Chem. Rev. 2000, 100, 4465–4493, and references cited therein.
  18. Lugtenburg, J.; de Groot, H.J.M. Photosynth. Res. 1998, 55, 241–245.
  19. Verdegem, P.J.E.; Lugtenburg, J.; de Groot, H.J.M. In Stable Isotopes in Pharmaceutical Research; Browne, T. R., Ed.; Pharmacochemistry Library, Elsevier: Amsterdam, 1997; Vol. 26, and references cited therein.
  20. Schoevaart. R.; Kieboom, T. Carbohydr. Res. 2001, 334, 1–6.
  21. Heijnen, J. J.; Haasnoot, C.A.G.; Bruggink, A.; Bovenberg, R.A.L.; Meijer, E. W.; Feringa, B. L.; Driessen, A. Integration of Biosynthesis and Organic Synthesis, A New Future for Synthesis; Programme Proposal of the National Scientific Foundation: The Hague, Oct 9, 2000.
  22. McCoy, M. Chem. Eng. News 2001, 79 (8), 23–25.


Rob Schoevaart is a postdoctoral fellow at the Leiden Institute of Chemistry (Department of Industrial Fermentative Chemistry, Leiden University, Einsteinweg 55, PO Box 9502, Leiden, 2300 RA, The Netherlands; w.schoevaart@chem.leidenuniv.nl). He received his M.Sc. in organic chemistry and microbiology from Nij megen University, The Netherlands, and his Ph.D. in organic chemistry from the Delft University of Technology, The Netherlands. At Leiden, he works on combined biological and chemical catalytic conversion procedures. His experience includes the application of multienzyme systems in organic synthesis and with in situ NMR analysis, and he has published six scientific papers.

Tom Kieboom is general science manager at DSM Food Specialties R&D (Postbus 1, Delft, 2600 MA, The Netherlands; tom.kieboom@dsm.com) and part-time professor of organic chemistry in the Department of Industrial Fermentative Chemistry, Leiden Institute of Chemistry (a.kieboom@chem.leidenuniv.nl). He received his M.Sc. and Ph.D. in organic chemistry from Delft University of Technology, where he later was appointed as full professor in bioorganic chemistry. His main research interests and expertise are in the application of chemical and biocatalysis in organic chemistry, with special emphasis toward carbohydrate conversions in combination with in situ NMR. He has published more than 160 scientific papers.

Return to Top || Table of Contents