Resources> Research

SUMMARIES OF MAJOR RESEARCH INTERESTS (1992 to January 2006)

1.0 NEW OPTIONS FOR MILD AND SELECTIVE POLYMERIZATIONS USING LIPASES

Abstract.

New and versatile biocatalytic methods were developed that offer mild and efficient options for polymer synthesis. A wide range of aliphatic polyesters, mixed linkage copolymers, and poly(ester-polyols) were developed. Lipase B from Candida antartica (CALB) physically immobilized on hydrophobic macroporous resins was found to be a remarkable catalyst for these non-natural substrates. Polymerizations were conducted by both ring-opening and step-condensation reactions. A new family of aliphatic polyesters was discovered using polyols as building blocks. The step-condensation polymerizations proceed without solvent under mild conditions (e.g. 50 to 90oC). Lipase regioselectivity allows the direct copolymerization of polyols with a range of diols and diacids to give non-crosslinked products. Polycondensations with natural polyols such as sorbitol give functional polyesters with weight average molecular weights up to 200 000. The mild reaction conditions allow the incorporation of chemically and/or thermally sensitive co-monomers. Studies of fundamental aspects of the reactions have lead to a better understanding of the polymerization mechanism and remarkable improvements in polymerization efficiency. These studies led to discoveries that: i) CALB catalyzes rapid transesterification reactions between high molecular weight substrates in the melt, ii) carbohydrates can be substituted for water as initiators to obtain polyesters with multifunctional end-groups, iii) polyols can be polymerized without protection/deprotection steps giving chains that are highly linear compared to their chemical counterparts, iv) synthesis of silicone-containing polyesters and polyamides, v) developing a new family of ω-hydroxyfatty acids.


Enzyme technology offers environmental benefits:

Industrial competitiveness increasingly depends on our ability to adapt and incorporate new technological advances. Furthermore, technological advances to reduce cost will often be associated with needs to reduce potentially toxic metals and reduce the energy requirements of synthetic pathways. Increased reaction efficiencies based on advances in catalysis will continue to be critical to industry. There is a clear need to pave new pathways that do more than provide incremental improvements in existing conditions. This was well stated in a quote by Professor Barry Trost “The issues cannot simply be addressed by minor tinkering with current processes to improve their performance although, undoubtedly, that is one of many of the strategies that will be helpful. Fundamental new science derived from basic research that creates new paradigms for synthesis is also required as well as everything in between” (Green Chemistry, 1998). Our laboratory has become increasingly engaged in developing enzyme-based routes to monomer, oligomer and polymer synthesis as well as polymer modification.1 The guiding principles of this work are as follows:

By using the expanding arsenal of enzymes in non-traditional ways in polymer synthesis and modification, important environmental, economic and product performance benefits will result.2 This will be a direct outcome of milder reaction conditions, increased reaction efficiencies, processes that require less discrete steps, the avoidance of heavy metals, and the ability to develop well defined highly functionalized polymeric products.

We selected synthetic challenges that are of practical importance and, therefore, if successful could be translated into commercial successes. In addition, we have been developing fundamental knowledge that will help build a strong base for the further development of enzyme-catalysis as an alternative synthetic pathway to build and modify polymers by green chemistry. Our work has established new methods that can be generally applied to a broad based segment of chemical manufacturers. New polyol-polyesters3,4 prepared by lipase-catalysis may be used as reactive components in polyurethane coatings. They also are under-evaluation for incorporation into cosmetic products. Polyesters prepared by lipase-catalysis from long chain hydroxylfattyacids are strong-tough plastics that offer properties that are intermediate between poly(ε-caprolactone) and polyethylene.5,6 Polyesters prepared with carbohydrate terminal groups represent a new family of functional macromers.7-9 The discovery that lipases catalyze transesterification reactions can be applied to a wide array of hydroxyl or carboxyl terminated chains (e.g. polybutadiene, silicones) to create new and previously unavailable family of multiblock copolymers.10-15 Furthermore, the polyol-polyesters from glycerol or sorbitol are currently under-study as bioresorbable and biocompatible implant materials.

This program has also found new ways to improve the usage of natural agro-based resources such as polyols (e.g. glycerol, sorbitol) and functional lipids (e.g. ω-hydroxyfattyacids). Furthermore, the methods developed offer new opportunities for the usage of these and other renewable feedstocks.

Project Description: Enzymes in Polymer Synthesis - New materials by in-vitro enzymatic processes

1.1 Solventless lipase-catalyzed polycondensations of polyols:

A practical new enzymatic process was developed for making polyol-containing polyesters and other novel compositions of matter. 3,4 Briefly, the process is a one-step enzyme-catalyzed polymerization of various acid, hydroxyl, and/or polyol building blocks conducted in the absence of solvent, with high regioselectivity, without activation of the diacid.

click here for full size image

Scheme 1 : Glycerol substitution pattern for lipase-catalyzed step- condensation polymerizations

By using various mixtures of natural polyols with other building blocks, the polar polyols are partially or completely solubilized resulting in highly reactive condensation polymerizations. By this method, organic solvents and activation of acids are not needed. The polymerizations are performed at temperatures between 60 and 95°C, in-vacuo (to remove water), and give products of high molecular weight (M w up to 200 000) with narrow polydispersities (as low as 1.3). 3 Furthermore, the condensation reactions with natural polyol building blocks proceed with high regioselectivity. Thus, of the ≥3 hydroxyl groups of the polyols, only two of the hydroxyl groups are highly reactive in the polymerization. The method offers simplicity, mild reaction conditions, and the ability to incorporate a wide range of renewable polyols into polyesters without protection-deprotection steps.

1.2 Rapid lactone polymerizations: When we began this research program(1993), lipase-catalyzed polymerizations were slow and gave low molecular weights. For example, polymerizations using 20% lipase (w/w rel. to monomer) would require 4-days to reach 80% monomer conversions giving polymers with M n about 2 000. Since then, our laboratory developed methods that transformed this work from an academic curiosity to a technology that now is under active development by our industrial sponsors. It is remarkable that the polymerizations can be conducted at temperatures from 20 to 108 oC giving monomer conversions of >60 mol% within 2 h (see Figure 1). 16 An important step was the identification of physically immobilized Lipase B from Candida Antarctica (immobilized CALB or Novozym 435, Novozymes Inc.). 17,18

click here for full size image

Figure 1: Monomer conversion as a function of time for e-caprolactone polymerizations. The water content was kept constant (0.6 % by wt).

click here for full size image

Scheme 3:Lipase catalyzed polymerization of w -pentadecalactone (a macrolactone) .  

By using CALB (1% w/w relative to monomer), running reactions at high monomer concentrations, in low polarity medium, and limiting the number of initiation events by using low water concentration (e.g. 0.5% by wt), remarkable improvements were realized in monomer conversion efficiencies and product molecular weights. For example, e -caprolactone polymerizations at 70 oC that previously required 4-days to reach M n 2 000 (see above) now needed only 4 hours to give polycaprolactone (M n 44 800 g/mol) in nearly quantitative yield. 18 Polymerizations of macrolactones that proceed slowly by chemical methods were efficiently conducted by Novozyme-435 catalysis. Within 20 min. at 55 oC polymer was formed in >90% yield with M n 86 K. 12

The lack of sensitivity of the polymerizations to oxygen, and tolerance to water (e.g. 0.6% w/w in reactions is typical), allows these reactions to be performed without the need for specialized reactors. Furthermore, we have shown unequivocally that for systems where monomer/polymer mixtures remain fluid to 90 oC, polymerizations can be conducted without the addition of solvent. 10,17 Poly( w -pentadecalactone) forms films with impressive properties. The films are best described as hard-tough with T m 97 oC, T g -27 oC, %-crystallinity ~60, and %-elongation 100 to 200. 5 If or when long chain w -hydroxyfatty acids become inexpensive commodity chemicals, their corresponding homo- and copolymers will have considerable interest as biodegradable polyethylene analogs. 2

1.3 Lipase-catalyzed transesterification reactions between polymers of high molecular weight:

Our group discovered that lipases are highly active for the catalysis of transesterification reactions between polyesters. 10,11 For polymers that have melting points below 100 oC the reactions can be conducted in-bulk. Transacylation reactions between polyesters involve intrachain cleavage by the lipase to form an enzyme-activated-chain segment, followed by reaction of this activated segment with the terminal hydroxyl unit of another chain. Immobilized Candida antartica lipase B (Novozyme-435) catalyzed transacylation reactions between pre-formed polyester chains with M n > 40 K. These reactions, conducted under mild conditions, were used to regulate block lengths along copolymer chains. In fact, transesterification reactions can occur to such a high extent that copolymers with random repeat unit sequence distributions are formed. 10 The rate of these lipase-catalyzed transacylation reactions is a function of the polyester chain length and the availability of chain-end hydroxyl groups. As the chain length increases, or the hydroxyl chain-end concentration decreases, the rate of transacylation reactions decreases. 10 Examples of suitable polyesters for transacylation reactions are poly( e -caprolactone), PCL, and poly( w -pentadecalactone), PPDL.

1.4 Use of carbohydrates as initiators for lactone ring-opening polymerization: Our laboratory is exploring how enzyme-catalysis can be used to prepare well-defined molecular architectures that would be difficult if not impossible to synthesize relying exclusively on traditional chemical catalysts. The enantio- and regioselectivity of enzymes is envisioned as a powerful tool that can be used to circumvent tedious protection-deprotection steps. As an example, a,b -ethylglucopyranoside

(EGP) was used as a multifunctional monosaccharide initiator for PPL-catalyzed e -CL ring-opening polymerizations. 8 For the e -CL/EGP ratio of 7:1, in the absence of solvent, a product resulted that had an M n and M w/M n of 2200 g/mol and 1.3, respectively. Structural analysis by 2-D NMR techniques showed that the reaction was highly regiospecific. In other words, oligo( e -CL) chains formed were attached by an ester group to the primary hydroxyl moiety of EGP.

Scheme 4: Ethyl glucopyranoside (EGP) initiated ring-opening polymerization of e-caprolactone

Subsequently, the lipase PS-30 catalyzed the selective acetylation of the w -oligo(CL) hydroxyl terminus. The acetylated product was used as a macroinitiator for lactide ring-opening polymerization. This gave a multi-arm heteroblock copolymer with spatially organized oligo(CL) and oligo(lactide) chain segments. Our plan in the future is to combine classical chemical and enzymatic methods to develop efficient routes to important molecular architectures that would either be difficult or currently not possible to prepare by conventional chemical methods. 9

1.5 Recent developments in polyol-polyester polymerizations:

Progression of Products Formed during Copolymerizations. 19 Scheme 5 summarizes the general characteristics of CAL-B catalyzed bulk polycondensations of glycerol copolyesters and its products formed at selected reaction times. Results from experiments for polycondensations with monomer feed ratio adipic acid:1,8-octane diol: glycerol (A:O:G) 1.0:0.8:0.2 (by mol) were used to illustrate distinct stages and product types found in reactions at 5 min, 45 min to 2 h, 6 to 18 h, and 42 h. Proton NMR of the copolymer at 5 min showed it consisted exclusively of octanediol-adipate units. Also, SEC revealed this product was a mixture of unreacted monomer and an oligomer (P[OA]) with M n and M w/M n of 1300 and 1.4, respectively (see Scheme 5). Thus, within the first 5 min of the polymerization, 1,8-octanediol reacts faster than glycerol and oligomers of narrow distribution are formed. Products at 45 min and 2 h had little or no unreacted monomers, M n of 2250 and 2700, respectively, and M w/M n of 1.2 and 1.6. The relative percent linear, terminal and dendritic groups for the 2 h product revealed that chains have (i) a 3:1 ratio of L 1,3:L 1,2 glycerol repeat units, (ii) a 1:1:1.7 ratio of T G to T O to T A end groups, and (iii) no dendritic glycerol units (see Table S-2). From this information, generalized structures of products formed between 45 min and 2 h was constructed and is displayed in Scheme 2. The extension of the polycondensation from 2 h to 6 and 18 h resulted in: (i) substantial increases in M n, (ii) broadening of the molecular weight distribution, (iii) decrease in T G units, and (iv) increase in L 1,3 units. A key finding was that CALB regioselectivity circumvented branching for the products formed to 18 h. Extension of the reaction time to 42 h results in remarkable changes in the product structure. Most notable is the formation of hyperbranched chains with dendritic glycerol units. In addition, M w/M n increased from 3.2 to 5.6 with little change in M n.

click here for full size image

click here for full size image

Scheme 5. Hypothetical Depiction of the Progression of Molecular Species That Evolve as a Function of Reaction Time for A:O:G of 1.0:0.8:0.2 Polymerization of 1,3-bishydroxybutyric acid20:

click here for full size image

Scheme 6. Novozyme 435 Catalyzed Copolymerization of Bis(hydroxymethyl)butyric Acid (BHB) with Adipic Acid and 1,8-Octanediol To Form Linear Aliphatic Polyesters Containing Pendant Carboxylic Acid Groups.
A simple, environment friendly, one-pot biocatalytic route to functional aliphatic polyesters bearing pendant carboxylic acid groups is described. By using the immobilized lipase CAL-B (Novozyme 435) as the biocatalyst, bis(hydroxymethylbutyric acid) was copolymerized with adipic acid and 1,8-octanediol without the need to first protect its carboxylic acid group. A series of aliphatic polyesters containing up to 22 mol % bis(hydroxymethylbutyric acid) units and molecular weights up to 22 000 were synthesized. The selectivity of Novozyme-435 resulted in the exclusive esterification of the bis(hydroxymethylbutyric acid) hydroxymethyl groups while leaving the carboxyl groups unreacted. Thus, instead of branched copolymers that would result if a chemical polymerization catalyst had been used, linear copolymers where each bis(hydroxymethylbutyric acid) repeat unit along the chain provides a carboxylic acid groups were obtained. The carboxylic groups of the functional copolyesters are highly versatile since they can be converted by simple chemical transformations to many other functional entities. Alternatively, the pendant carboxyl groups can be used to directly conjugate bioactive, photo-cross-linkable, or optically interesting molecules. Thermal analysis of the polymers showed that they have high thermal stability and are low melting. WAXS measurements confirm that the degree of crystallinity decreases upon copolymerization with BHB. The concurrent decrease in the degree of crystallinity with the increasing content of the BHB units in the polyester chains indicates that these copolyesters will have tunable bioresorption rates.

1.6 Poly(ω-hydroxyfatty acids): Potential new biopolyesters that mimic polyethylene structure and properties:

Our group discovered that, instead of using macrolactones that are expensive to prepare, linear ω-hydroxyfattyacids can be directly polymerized using CAL-B as the catalyst (Scheme II)21 The relative reactivity for step-condensation polymerizations of 10-, 12- and 16-carbon ω-hydroxyfattyacids was almost identical. Typically, the polymers formed had DPavg = 120, Mn about 30 000, and Mw/Mn <=1.5.The reactions were complete within 8 hours when the rate of water removal was controlled. That is, the rate of water removal is an important parameter that controls the kinetics of polymer synthesis. The successful synthesis of high molecular poly(ω-hydroxypentanoic acid) and related polymers in substantial quantities paved the way for physical property studies that validated the usefulness of these polymers as materials.

The solid-state properties of poly(ω-hydroxypentadecanoate) were investigated by thermogravimetric analysis coupled with mass spectrometry (TGA-MS), differential scanning calorimetry (DSC), stress-strain measurements, wide angle X-ray diffraction, dynamic mechanical and dielectric spectroscopies.5 Poly(ω-hydroxypentanoate) is a highly crystalline polyester that melts at about 100oC and, due to its long methylene sequence, shows structural similarities to polyethylene. The glass transition of the polyester is -27°C. Furthermore, it has good thermal stability, with a mainTGA weight loss centered at 425°C. Moreover, poly(ω-hydroxypentadecanoate) has a stress-strain curve that reflects its hard-tough behavior with elongation up to about 100 to 200%. The elastic modulus and yield parameters are comparable with those of low density polyethylene.

To explore variations in the physical properties of poly(ω-hydroxypentadecanoate), copolymers were prepared using either trimethylene carbonate or ε-caprolactone. Interestingly, the copolymers with equimolar comonomer content and close-to-random distribution are highly crystalline.6 This shows that neighboring ω-hydroxypentadecanoic acid (PDL)/ ε-caprolactone (CL) and ω-hydroxypentadecanoic acid/trimethylene carbonate (TMC) units co-crystallize. PDL-CL copolymers showed a single crystal phase whose melting temperature changed with composition from that of PPDL to that of PCL. In PDL-TMC copolymers the result was a mixture of higher and lower melting crystal phases. The higher melting transition corresponds to poly(ω-hydroxypentadecanoic acid) crystals while the lower corresponds to the crystallization of alternate PDL-TMC units. This low-melting

click here for full size image

Figure 3. Melting temperature of linear polyhydroxy acids and lactones as a function of the methylene-to-ester group ratio: a) polyglycolic acid (Brandrup et al. 1999), b) poly(3-hydroxypropionate) (Shimamura et al. 1994), c) poly(4-hydroxybutyrate) (Nakamura et al. 1992), d) poly(ε-caprolactone) (Brandrup et al. 1999), e) poly(10-hydroxycapric acid) (Brandrup et al. 1999), f) poly(PDL), this work. Dotted line: melting temperature of
polyethylene (Brandrup et al. 1999).

phase showed by X-ray diffraction a fiber axis periodicity larger than that of PPDL, that corresponds to the long PDL-TMC crystallizing units.

Another important method to change the physical properties of poly(ω-hydroxyfatty acids) is by varying the methylene-to-ester group ratio, i.e change the repeat unit length5 Figure 3 shows the melting temperature of linear poly(ω-hydroxyfatty acids) (or their equivalent polymers from lactone polymerization) as a function of the methylene-to-ester group ratio. According to a well- known behavior of aliphatic polyesters, after an initial steep drop caused by reduction of intermolecular polar interactions with increasing repeating unit length, Tm slowly increases reflecting the predominant effect of ‘diluting’ the flexible ester group along the polymer chain when n(CH2)/m(COO) > 3. Figure 3 shows that the Tm of poly(ω-hydroxypentanoate) fits the increasing trend of the melting temperature towards the limiting value of polyethylene (dotted line).

1.7 Sweet Silicones22:

Pure organosilicon-sugar conjugates were prepared in a one-step reaction, without protection-deprotection steps. This simplification of an otherwise tedious reaction was a result of the inherent regioselectivity of the lipase catalyst. The lipase-catalyzed

click here for full size image

Scheme 7. Lipase-catalyzed esterification between ethylglucoside and diacid-terminated siloxanes (PDMS Diacids) where x = 0, 7, and 65 reactions did not require activation of the acid groups. In comparison to organic materials, the hydrophobic organosilicones were acceptable substrates. Given the proficiency of lipases to perform a selective reaction and maintain the integrity of the siloxane bonds with lipase, the ability to synthesize structurally defined organosilicon carbohydrates with a diversified set of functional groups may be used to create new materials such as fibers, films, coatings, gels, and surfactants with novel properties.

1.8 References and Patents (most closely related to this work):

1. R. A. Gross, A. Kumar, B. Kalra, Chemical Reviews, 101(7), 2097-2124 (2002).
2. R. A. Gross and B. Kalra, Science, Vol 297, 803-806 (2002).
3. R. A. Gross et al., “Enzyme-Catalyzed Polycondensations” United States Patent Application 20040019178, issued January 29 2004.
4. A. Kumar, A.S. Kulshrestha, W. Gao, R.A. Gross;
   Macromolecules; 36(22); 8219-8221 (2003); A. Mahapatro, A. Kumar, B. Kalra, B. Macromolecules 37(1); 35-40 (2004).
5. M.L. Focarete, M. Scandola, A. Kumar, R. A. Gross, J. Polym. Sci., Part B: Polym. Phys. 39(15), 1721-1729 (2001).
6. M.L. Focarete, M. Gazzano, M. Scandola, A. Kumar, R.A. Gross, Macromolecules, 35(21); 8066-8071 (2002).
7. R.A.Gross, K. Bisht, D. Kaplan, G. Swift, F. Dang US Patent 5,981,743 November 9, 1999.
8. K. S. Bisht, F. Deng, R. A. Gross, D. L. Kaplan and G. Swift, J. Am. Chem. Soc., 120, 1363-1367 (1998).
9. R. Kumar, R.A. Gross, J. Am. Chem. Soc. 124(9), 1850-1851 (2002).
10. A Kumar, R.A Gross, J. Am. Chem. Soc. 122, 11767-11770 (2000).
11. R. A. Gross and A. Kumar, U.S. Pat. Appl., Filed January 21 2000.
12. A Kumar, B Kalra, A Dekhterman, R. A. Gross, Macromolecules, 33, 6303-6309 (2000).
13. A Kumar, K Garg, R. A. Gross, Macromolecules, 34, 3527-3533 (2001).
14. A. Kumar, R.A. Gross, Y. Wang, M.A. Hillmyer, Macromolecules 35(20); 7606-7611 (2002).
15. Bankova, M.; Kumar, A.; Impallomeni, G.; Ballistreri, A.; Gross, R. A.; Macromolecules, 35(18); 6858-6866 (2002).
16. Y. Mei, A. Kumar, R.A. Gross, Macromolecules, 35, 5444-5448 (2002).
17. F. Deng, R. A. Gross, Int. J. Biol. Macrom.. 25, 153-159 (1999).
18. A Kumar and R.A. Gross, Biomacromolecules, 1, 133-138 (2000).
19. A. S. Kulshrestha, W. Gao, R.A. Gross Macromolecules; 38(8); 3193-3204 (2005).
20. A. S. Kulshrestha, B. Sahoo, W. Gao, H. Fu, R.A. Gross Macromolecules; 38(8); 3205-3213 (2005).
    21. Mahapatro, A.; Kumar, A.; Gross, R. A.; Biomacromolecules; 5(1); 62-68 (2004).
21. Sahoo, B.; Brandstadt, K. F.; Lane, T. H.; Gross, R. A.; Org. Lett.; 7(18); 3857-3860
    (2005)

2.0 Bioengineering of Emulsans and Sophorolipids - New Families of Glycolipid Bioemulsifiers, Vaccine Adjuvants, and Pharmaceuticals

Emulsan, an extracellular lipoheteropolysaccharide generated by the bacterium Acinetobacter calcoaceticus that forms oil-in water emulsions, was studied.

click here for full size image

Schematic: Emulsan general structure

The key concept that was demonstrated in this research was the extraordinary ability to which the metabolic flexibility of the biosynthetic pathway could be exploited to incorporate non-native structures into the polymer. Side chain lipids that were highly fluorinated, hydroxylated in unusual positions, and had high levels of unsaturated were incorporated as side-chains of the polysaccharide backbone. Importantly, these alterations in structural features resulted in large changes in emulsification behavior. For example, it was shown that, as the degree of fatty acid substitution of the polysaccharide backbone increased to about one fatty acid per seven sugar residues, the emulsification activity was optimized. Since these polymers have analogous structures to endotoxins, work was begun in collaboration with Professors David Kaplan and Juliet Fuhrman (Tufts University) to explore whether these lipoheteropolysaccharides could function for immunoregulation. Studies thus far, by macrophage screening and a mouse study, showed that when properly designed, these 'tailorable' lipoheteropolysaccharides have comparable adjuvant activity to an industry standard. In a similar fashion, a family of sophorolipids with interesting surface activity has been produced by fermentation of Candida bombicola in our laboratory. Methods have been developed to produce pure compounds and carry out lipase-catalyzed site-selective modifications. Studies using these compounds as anti-cancer agents have thus far proved promising. Other biomedical applications of site-selectively modified glycolipids, and their incorporation as side chains in water-soluble polymeric carriers have been conducted.

• Leading References -

• Jinwen Zhang, Alexander Gorkovenko, Richard A. Gross, Alfred L. Allen and David L. Kaplan "Incorporation of 2-Hydroxyl Fatty Acids by Acinetobacter calcoaceticus RAG-1 to Tailor Emulsan Structure, Intern. J. Biol. Macromol., 20, 9-21 (1997).
  
• Alexander Gorkovenko, Jinwen Zhang, Richard A. Gross, Alfred L. Allen and David L. Kaplan, "Bioengineering of Emulsifier Structure: Emulsan Analogs", Can. J. Microbiol, 43, 384-390 (1997).

3.0 Sophorolipids: a unique class of microbial membrane glycolipids

Sophorolipids and related structures are a diverse family of microbial glycolipids, which can be isolated from natural sources, and easily chemo-enzymatically modified. They naturally occur as disaccharide sophoroses linked glycosidically to the hydroxyl group at the penultimate carbon of primarily C18 chain-length fatty acids (Figure 1).

Figure 1 Structure of lactonic and acidic forms of Sophorolipid mixture produced by Candida bombicola.
They are fermentatively produced by yeasts such as Candida bombicola, Yarrowia lipolytica, Candida apicola, and Candida bogoriensis. First described in 1961, sophorolipids occur as a mixture of macrolactone and free acid structures that are acetylated to various extents at the primary hydroxyl position of the sophorose ring (Fig. 1). Analytical studies revealed that at least eight structurally different sophorolipids are produced during fermentation. Our laboratory has developed fed-batch fermentation methods that have 350 g/L sophorolipids in shake-flask reactors1.

These types of bioengineered sophorolipid compounds have widespread implications in many areas of clinical medicine, including pharmaceuticals, drug delivery and components of biomaterials. We have recently demonstrated that sophorolipids, which is a unique member of the glycolipid family, has shown that it has immense potential as therapeutic agents. Sophorolipids can act as an immunomodulator and display antibacterial, antifungal, antiviral (HIV-1) antiseptic, anti-spermicidal and anticancer activities2-4. Part of its mechanism of action appears in its ability to perturb membrane integrity and is cell specific in that it disrupts algae and other microbes but has no impact on macrophage viability.

Our laboratory has pioneered chemo-enzymatic methods to synthesize a range of pure sophorolipid analogs from the microbially produced natural mixture.5,6 In a recent study, novel enzyme-mediated synthetic routes were developed to provide a new family of

click here for full size image

Scheme 1: a a (i) vinyl acetate, Novozym 435, dry THF, 40 C, 2.5 h; (ii) vinyl ester (vinyl acetate for 3 and vinyl methacrylate for 4), Lipase PS-C, dry THF, 40 C, 72 h; (iii) primary amine (tyramine for 5, phenethylamine for 6, (p-tolyl)amine for 7, p-methoxyphenethylamine for 8, p-fluorophenethylamine for 9), Novozym 435, dry THF, 50 C, 24 h. sophorolipid derivatives and glycolopid-based amphiphilic monomers. As discussed above, these compounds are of great interest for their potential use in immunoregulation, as well as for other biological properties. An efficient lipase-catalyzed conversion of sophorolipid ethyl ester to (a) the 6'-monoacylated derivatives using Novozym 435, (b) 6' '-monoacylated derivatives using Lipase PS-C, (c) secondary amide derivatives using Novozym 435, and (d) 6',6' '-diacylated amide derivatives using Novozym 435 in an one-pot reaction and (e) the regioselective monoacylation of an amide derivative at the 6'- and 6' '-positions using Novozym 435 and Lipase PS-C, respectively, were reported. The reaction conditions and products synthesized are shown in Scheme 1, above.

References

1. Guilmanov V., Ballistreri, A. Impallomeni, G. Gross, R. A. Biotechnol. Bioeng., 77, 489 (2002).
   
2. Bluth MH, Kandil E, Mueller CM, Shah V, Lin YY, Zhang H, Dresner L, Lempert L, Nowakowski M, Gross R, Schulze R, Zenilman ME. Sophorolipids block lethal effects of septic shock in rats using a cecal ligation and puncture model of experimental sepsis. Crit Care Med Vol. 34, No. 1 (available online ASAP) (2006).
   
3. Shah V, Doncel GF, Seyoum T, Eaton KM, Zalenskaya I, Hagver R, Azim A, Gross R. Sophorolipids, microbial glycolipids with anti-human immunodeficiency virus and sperm-immobilizing activities. Antimicrob Agents Chemother 49:1-8 (2005)
   
4. Scholz C, Mehta S, Bisht K, Guilmanov V, Kaplan D, Nicolosi R, Gross R. Bioactivity of extracellular glycolipids – investigation of potential anti-cancer activity of sophorolipids and sophorolipid-derivatives. Am. Chem. Soc. Polymer Preprints (1998).
   
5. Bisht KS, Gross RA, Kaplan DL. Enzyme-Mediated Regioselective Acylations of Sophorolipids. J Org Chem. 64: 780-789 (1999)
   
6. Singh, S. K.; Felse, A. P.; Nunez, A.; Foglia, T. A.; Gross, R. A.; Regioselective Enzyme-Catalyzed Synthesis of Sophorolipid Esters, Amides, and Multifunctional Monomers, J. Org. Chem.; 68(14); 5466-5477 (2003).

4.0 Biosynthesis of Novel Polysaccharide Copolymers - New Families of Polysaccharide-Based Biomaterials

There have been intensive studies of the biosynthesis of modified or nonnative proteins and polyesters in bacteria. However, there has been little study of this phenomenon in the bacterial synthesis of polysaccharides. Polysaccharides perform an incredibly diverse set of critical functions in the biosphere including structural, adhesion and barrier functions. Polysaccharides are more problematic for studies of structure-function relationships because, unlike proteins, the direct link between genetic blueprints and polymer structure is absent. Thus, more complex control pathways to regulate synthesis and processing are involved. The manipulation of the biosynthetic pathway flexibility of exopolysaccharide synthesis can lead to new insights into the rational synthesis of polysaccharides with controlled compositions and chemistries. This methodology will in turn provide the requisite family of structurally defined polymers with which to interrogate assembly into macromolecular materials. The first goal has been to establish the synthetic methods by co-opting the biosynthetic machinery. To this end, we have studied bioengineering of polysaccharides, including changes in the composition of glucose and galactose in zoogloea gum from Zoogloea ramigera and changes in the content of glucose and mannose in pullulan from the fungus, Aureobasidium pullulans. As an alternative approach to modify pendant groups on polysaccharide main chains, an enzymatic route was used to regioselectively form fatty acid esters on the backbone of amylose and cellulose. Curdlan, an unbranched homo-(1,3)-glucan produced by Agrobacterium sp., was modified in vivo by the direct incorporation of the carbon source 3-O-methyl-D-glucose. Up to 12 mol% of this modified polyglucan was 3-O-methyl-D-glucose. Studies recently completed on the biosynthesis of cellulose and modified cellulose by Acetobacter xylinum support the concept that the non-glucose analogs 2-amino-2-deoxy-glucose and 2-acetimido-2-deoxy-glucose can be directly incorporated into microbial cellulose formed by A. xylinum. This strategy resulted in the formation of novel copolymers, glucose/2-amino-2-deoxy-glucose (cellulose-chitosan) and glucose/2-acetimido-2-deoxy-glucose (cellulose-chitin). These types of cellulose-chitin and cellulose-chitosan copolymers have not been previously reported and represent a new group of polysaccharides based on solubility, functional properties and structure. These types of 'tailorable' polysaccharides may lead to novel biomaterials.

References -

• Jin W. Lee, Water G. Yeomans, Alfred L. Allen, David L. Kaplan, Frank Deng, and Richard A. Gross. Exopolymers from curdlan production-incorporation of glucose-related sugars by Agrobacerium sp. ATCC 31749. Can. J. Microbiol. 43, 149-156 (1997).
  
• Jin W. Lee, Water G. Yeomans, Alfred L. Allen, David L. Kaplan, and Richard A. Gross. "Production of zoogloea gum by Zoogloea ramigera with glucose analogs" Biotechnol. Lett. 19(8), 799-802 (1997).

5.0 Functional Polylactides for Biomedical Applications

Although there has been considerable progress towards the development of bioresorbable biomaterials, additional work is urgently needed to address current demands for polymer therapeutic materials. We have developed new monomers that help fill gaps in existing biomaterials. As an example, the synthesis of 1,2-O-isopropylidene-D-xylofuranose-3,5-cyclic carbonate (IPXTC) from xylofuranose was carried out in high yield after two simple steps. This work builds on our experience that substituted cyclic carbonate monomers can be copolymerized with medically important monomers such as L-lactide, trimethylene carbonate and e-caprolactone. An emphasis has been placed on the development of monomers from natural sugars where ketal-protected diols can incorporated along chains and subsequently be deprotected to generate hydroxyl functionalized biomaterials. Work is now underway to evaluate these new bioresorbable copolymers for a range of medical needs.

Figure 1 - Copolymerization of L-LA with IPXTC

click here for full size image

Figure 2 - Deprotection of IPXTC ketal groups to form vicinol diol pendant groups.

click here for full size image

References -

• X. Chen and R. Gross, "Versatile Copolymers from [L]-Lactide and [D]-Xylofuranose", Macromolecules, 32, 308-314 (1999).
  
• "Polycarbonates from Sugars: Ring-Opening Polymerization of 1,2-O-Isopropylidene-D-Xylofuranose-3,5-Cyclic Carbonate (IPXTC)", Macromolecules, 32, 2799-2802 (1999).
  
• "Aliphatic Polycarbonates with Controlled Quantities of D-Xylofuranose in the main Chain", Macromolecules, 32, 3891-3897 (1999).

6.0 The In-Vivo Formation of Natural-Synthetic Diblock Copolymers

Our laboratory pioneered what is a new research area that we termed 'polyethylene oxide (PEG) modulated fermentation'. It was discovered that the addition of PEG to culture media during microbial polyester formation resulted in the following: i) PEG acted as an in-vivo chain terminating agent that allowed the 'tailoring' of microbial polyester molecular weight, and ii) regulation of the repeat unit compositions and sequence distribution along chains. For the first time block copolymers that consisted of a synthetic segment (PEG) and a natural polymer was formed in vivo. This research has identified a route by which synthetic-natural block copolymers can be synthesized in-vivo by the identification of synthetic oligomers or polymers that act as terminators of microbial polymer synthesis.

References -

• R. D. Ashby, F.-Y. Shi and R.A. Gross, "A Tunable Switch to Regulate the Synthesis of Low and High Molecular Weight Microbial Polyesters" Biotechnology and Bioengineering, 62 (1), (1999).
  
• Richard Ashby, Feng-Ying Shi, and Richard A. Gross, "Use of poly(ethylene glycol) to control the end group structure and molecular weight of poly(3-hydroxybutyrate) formed by Alcaligenes latus DSM 1122," Tetrahedron, 53(45), 15209-15223 (1997).

7.0 Microbial Synthesis of γ-poly(glutamic acid)
γ-PGA is a water soluble, polyanionic, biodegradable, extracellular polymer produced by the bacterium, B. licheniformis 9945a . The polymer can be biosynthesized from many different carbon sources including starch, glucose, amino acids, glycerol, citric acid and gluten.

Previous Studies on γ-PGA Formation
The bacterial capsular γ-PGAs are distinct in structure when compared with conventional ⟩-linked poly(amino acids) and would be best classified as pseudo-poly(amino acids). ⟩-PGA is formed mainly by specific species of the genus Bacillus and often diffuses freely into the fermentation medium. This polymer consists of glutamic acid repeating units that are linked between the Αlpha-amino and γ-carboxylic acid functionalities (Figure 1). Reviews on γ-PGA have been published by Housewright (1962), Nitecki and Goodman (1971), Troy (1982), and Gross (1997).

Figure 1. Structure of γ-PGA (free acid form).

Beginning in 1969, several strains of bacillus that produced γ-PGA were isolated from sewage, soil, and grain (Murao et al., 1969; 1971). B. subtilis 5E emerged as the best overall γ-PGA producing strain at approximately 18 g/L of crude γ-PGA when grown on a glucose/urea medium (Sawa et al., 1971). It is also significant that the mucin from "Natto" contains ?-PGA as well as polysaccharide components. Natto is a popular food in Japan as well as some parts of China and Thailand. The fact that "Natto" is a traditional fermented food suggests that this biopolymer may be safely consumed (Hara et al., 1982). This will lead to potential applications as food and pharmaceutical formulations.

γ-PGA is produced at a higher rate when a citrate and glycerol medium is augmented with glutamic acid. Early work in the field had shown the value of glutamic acid as a component of the growth medium for γ-PGA synthesis (Ward et al., 1963). 13C labeling studies (Cromwick and Gross, 1995a) showed that added glutamate is incorporated directly into the polymer undiluted by other primary metabolites. There are many examples where various Bacillus species have used glucose, sucrose, amino acids, citrate, glycerol and gluten as carbon sources for ?-PGA formation in high yields (Thorne et al., 1954; Leonard et al., 1958; Ward et al., 1963; Murao et al., 1969; Sawa et al. , 1971; Murao et al., 1971; Sawa et al., 1973a; 1973b; Troy, 1973a; 1973b; Cheng et al., 1989; Goto and Kuniota, 1992). These examples demonstrate the potential for γ-PGA synthesis by the biological conversion of renewable resources in an efficient manner. It is important to note that B. licheniformis does not require any specific vitamins for growth and γ-PGA formation. This permits selection of bulk renewable substrates such as starch without regard to their specific growth factor content.

A peculiar characteristic of γ-PGA formed by B. licheniformis 9945a as well as some other γ-PGA producing strains is that the polymer stereochemistry can be controlled by the composition of the medium used in polymer production (Cromwick and Gross, 1995a). Specifically, the PI's collaborator recently has published a detailed study on the effects of Mn (II) on various aspects of B. licheniformis physiology and polymer formation. These data showed that γ-PGA having percent [L] compositions of 59% to 10% were obtained in media containing 0 to 615 mM MnSO4, respectively. Work by others has demonstrated that other divalent metals such as Zn and Co may also be used to modulate γ-PGA stereochemical composition (Leonard et al., 1958). Also, evidence has been presented by Thorne and Leonard (1958) which suggests that the γ-PGA formed by B. licheniformis 9945a is a mixture of polymer chains which contain a high predominance of either L- or D- stereoisomers (approximate stereochemically pure chains).

Hikichi et al. (1990) have used NMR to investigate copper and manganese binding with γ-PGA and McLean et al. (1990) studied the binding affinity of γ-PGA to a wide range of metal salts. Flocculation could be induced by the addition of Cu2+, Al3+, Cr3+, and Fe3+ with formation of the corresponding γ-PGA-metal complexes (McLean et al., 1990). This demonstrates potential applications of γ-PGA in waste water treatment, mining, and in other products which utilize its metal chelation characteristics.

Work in our Laboratory on γ-PGA Biosynthesis

We have studied the effects of physiological variables on γ-PGA stereochemistry, molecular weight and product yield using B. licheniformis 9945a (Giannos et al., 1990; 1991; Birrer et al., 1994; Cromwick et al., 1994; Cromwick and Gross, 1995a; 1995b). In addition, analytical methods to determine molecular weight, yield and stereochemical composition have been established in the laboratory.

Physiological Effects on γ-PGA Molecular Weight, Yield, and Stereochemistry: Significant changes in γ-PGA molecular weight, ranging from 100,000 to 2 million g/mole (number average molecular weight, Mn) have been obtained by varying culture conditions. In general, γ-PGA molecular weight decreases during the fermentation period, presumably due to the presence of depolymerase enzyme(s). However, we have previously identified certain media conditions, such as decreased media sulfate concentration and increased oxygenation, that lead to more rapid depolymerization of the polymer. Furthermore, increased medium sodium chloride concentration apparently decreases the rate of polymer breakdown.

Initial studies have also shown that [L]-glutamate is not required for polymer formation and that a decreased sulfate media concentration leads to significant increase in polymer yield using modifications of medium E (medium E in g/L consists of: [L]-glutamic acid, 20.0; citric acid, 12.0; glycerol, 80.0; NH4Cl, 7.0; K2HPO4, 0.5; MgSO4 .7H2O, 0.5; FeCl3.6H2O, 0.04; CaCl2 .2H2O, 0.15; MnSO4.H2O, 6.15 x 10-4M). Our prior results indicate that augmentation of medium E with increased citrate results in a significant enhancement in the γ-PGA yield. Controlled pH experiments show increased polymer formation at pH 6.5 relative to 5.5 and 7.4. The rate and total volumetric production of γ-PGA increased dramatically with higher oxygenation conditions, 250 to 800 rpm and 0.5 to 2.0 liter air/liter, respectively. Finally, variation of the media manganese sulfate concentration from 6.15 x 10-4 M to 0 M resulted in γ-PGA % [D]-glutamate compositions of approximately 90% and 50%, respectively. This is consistent with reports by Thorne et al. (Leonard et al., 1958; Thorne and Leonard, 1958) and contrasts with claims made by Troy (1973a; 1973b; 1982).

Analytical Methods For γ-PGA Analysis: We have developed convenient analytical methods to quantify γ-PGA yield, molecular weight and stereochemistry. Gel permeation chromatography (GPC) with Shodex KB800 series columns (two KB80M, and one of KB802.5) and a mobile phase containing 0.3M Na2SO4 brought to a pH of 4.0 allows elution of the polymer from the column to separate various molecular weight components of the product. Other GPC resins investigated for this purpose retained the polymer on the column. This GPC method is now used routinely in our laboratory to monitor the molecular weight and yield of the product by direct injection into the GPC of a small filtered culture aliquot. Quantitation of the yield from this analysis is obtained by measuring peak area against a calibration constructed from purified γ-PGA standard samples. The repeat unit stereochemistry of the γ-PGA is measured by reverse phase HPLC. Samples are prepared by filtering an aliquot of the crude culture through a 0.45 mm cellulose acetate membrane to remove cells. Salts and other low molecular weight impurities are removed by dialysis and the solution is concentrated by filtration using Amicon 30,000 molecular weight cutoff Centricon Semi-PrepTM filters. The retentate containing the polymer is hydrolyzed in acid, derivatized using Marfey's reagent (Marfey, 1984) and the resulting diasteriomers analyzed by HPLC with UV detection (for addition details see Cromwick and Gross, 1995b).

8.0 Biodegradable Polymers - Probing Structure-Biodegradability Relationships

8.1 Synthetic analogs: Syndiotactic Poly(3-hydroxybutyrate).

Our laboratory discovered that tin-based catalyst systems could be used to synthesize predominantly syndiotactic poly(3-hydroxybutyrate) which is a stereochemical isomer of the natural isotactic enantiopure polyester. These syndiotactic analogs of bacterial poly(3-hydroxybutyrate) was found to degrade at unexpectedly rapid rates in the presence of some depolymerase enzymes. These findings stimulated a number of other laboratories to begin research on this new polymer type that had very different properties then it's natural analog.

Leading References -