Synthesis and Characterization of a Cationic Poly (β-hydroxyalkanoate)

Jul 23, 2008 - Poly (PHAs) are biodegradable polyesters produced by a wide range of bacteria. The structures of these polymers may be tuned by control...
0 downloads 0 Views 540KB Size
Download PDF
August 2008

Published by the American Chemical Society

Volume 9, Number 8

 Copyright 2008 by the American Chemical Society

Communications Synthesis and Characterization of a Cationic Poly(β-hydroxyalkanoate) Jeff Sparks and Carmen Scholz* University of Alabama in Huntsville, Department of Chemistry, 301 Sparkman Drive, Huntsville, Alabama 35899 Received May 21, 2008; Revised Manuscript Received July 7, 2008

Poly(β-hydroxyalkanoates) (PHAs) are biodegradable polyesters produced by a wide range of bacteria. The structures of these polymers may be tuned by controlling the carbon source composition in the feed stock, but the range of functional groups accessible in this manner is limited to those that the organism is able to metabolize. Much effort has been made to chemically modify the side chains of these polymers to achieve new materials. Here, we report the synthesis of the first cationic PHA, poly(β-hydroxy-octanoate)-co-(β-hydroxy-11-(bis(2hydroxyethyl)-amino)-10-hydroxyundecanoate) (PHON). Pseudomonas putida Gpo1 was used to produce poly(βhydroxy-octanoate)-co-(β-hydroxy-10-undecenoate) (PHOU), whose vinyl-terminated side chains were first converted to terminal epoxides and then modified with diethanolamine. The modification of PHOU was examined using 1H, COSY, and HSQC NMR and GPC and resulted in a loss of molecular weight due to aminolysis and also in the introduction of side chains terminated with tertiary amine groups, which are protonated at physiological pH. The polycationic PHA is soluble in polar solvents such as DMSO, DMF, and water. The new biodegradable cationic polymers are envisioned as nucleic acid delivery systems.

Introduction Bacterial polyesters known as poly(β-hydroxyalkanoates) (PHA) are produced by a wide range of organisms.1 As the microorganisms sense environmental stress, usually through nitrogen deprivation, they produce PHA as a form of carbon storage. PHAs are sequestered intracellularly within vesicles and are inherently hydrophobic. Such polymers derived from renewable resources and produced from whole-cell catalysis are receiving consideration for potential application in the packaging and biomedical fields.2 Because they are of biological origin, they possess several desirable properties including biodegradability, biocompatibility, and recyclability. However, to be used in medically relevant applications requiring water solubility, PHAs must be chemically modified. Depending on the carbon supply in the local environment of the microorganism, the PHA side chains may contain a variety of terminal chemical structures, including phenyl, bromine, arylalkyl, ester, branched alkyl, and vinyl groups.3 To achieve further chemical diversity, * To whom correspondence should be addressed. Tel.: +1-256-824-6188. E-mail: [email protected].

others have modified these polymers through the side chains by conversion to other functional groups inaccessible through the microorganism, as well as by the attachment of other polymers.4 To produce a water-soluble PHA, a series of reaction steps were employed. P. putida GPo1 was used to produce medium-chain-length PHAs containing side chain-terminal vinyl groups. The cultures were fed equimolar concentrations of sodium octanoate (SO) and undecylenic acid (UA) to give the random copolymer, poly(β-hydroxyoctanoate-co-β-hydroxyundecenoate) (PHOU). The technique described by Bear et al.5 was then used to convert the vinyl groups to epoxides to give poly(β-hydroxyoctanoate-co-β-hydroxy-10-epoxyundecenoate) (PHOE). Finally, diethanolamine was used to modify the epoxide groups to give side chains terminated with the corresponding tertiary amine, resulting in the polymer poly(βhydroxy-octanoate)-co-(β-hydroxy-11-(bis(2-hydroxyethyl)amino)-10-hydroxyundecanoate) (PHON). This polymer was water soluble at a pH below its pKa due to protonation of the nitrogen atom. To the best of our knowledge, this represents the first report of a PHA containing side chain terminal amines, as well as the first report of cationic PHA.

10.1021/bm8005616 CCC: $40.75  2008 American Chemical Society Published on Web 07/23/2008

2092

Biomacromolecules, Vol. 9, No. 8, 2008

Communications

Figure 1. 1H NMR spectrum of PHOU. Table 1. Molecular Weight and Polydispersity Index (PDI) of PHOU, PHOE, and PHONa

PHOU PHOE PHON

Mn

Mw

PDI

100000 104000 20000

271000 287000 49000

2.71 2.76 2.45

a Data were obtained using a Waters Styragel HR 5E column with a mobile phase of 0.01 M LiBr DMF at a flow rate of 1.0 mL/min at 30°C. The system was calibrated with monodisperse polystyrene standards.

Experimental Section Materials. Pseudomonas putida GPo1 was purchased from ATCC (ATCC 29347) and maintained according to standard protocol on agar plates composed of modified E* medium (Per 1 L: 1.1 g (NH4)2HPO4, 5.8 g K2HPO4, 3.7 g KH2PO4, 2.5 g sodium octanoate, 10 mL of a 100 mM MgSO4 solution, 1 mL of a µ-element solution: per 1 L of 1 M HCl solution, 2.78 g FeSO4 · 7H2O, 1.98 g MnCl2 · 4H2O, 2.40 g CoSO4 · 6H2O, 1.67 g CaCl2 · 2H2O, 0.17 g CuCl2 · 2H2O, 0.29 g ZnSO4 · 7H2O) at 4 °C. All salts, m-chloroperbenzoic acid (mCPBA), and diethanolamine were obtained from Sigma-Aldrich, and used without further purification. All solvents were obtained from Fisher and used as received, unless otherwise noted. Polymer Synthesis. PHOU. A 200 mL dented shake flask containing E* medium and 15 mM sodium octanoate as the carbon source was inoculated with P. putida GPo1 cells taken from an agar plate. The culture was grown in an incubator/shaker (30 °C, 180 rpm) for 24 h after which the entire contents of the flask were transferred into 1.8 L containing 7.5 mM sodium octanoate and 7.5 mM undecylenic acid as the carbon sources. This 2 L culture was grown under identical conditions for 48 h. A pellet of wet cells was obtained after centrifuging at 4 °C for 20 min at 8000 × g. The cells were frozen and freeze-dried for 24 h. The PHA was extracted by refluxing under methylene chloride

for 8 h, filtered to remove cell debris, and precipitated in cold methanol. The PHA was collected, dried under vacuum, and stored at -20 °C until further use. PHOE. A total of 500 mg PHOU (5 µmol), containing 1.7 mmol vinyl groups, was dissolved in 10 mL of dry methylene chloride. A 2-fold molar excess (compared to vinyl groups) of mCPBA was dissolved in 10 mL dry methylene chloride and added to the PHOU solution under argon. The solution was stirred under argon at room temperature for 12 h, after which a substantial amount of white precipitate formed. Pure PHOE was obtained by precipitating the solution in cold methanol and drying. PHOE underwent slight crosslinking upon complete drying, which resulted in a swelled, insoluble material. To prevent cross-linking, PHOE was collected from methanol and transferred directly into a Schlenk flask for the next reaction. PHON. A total of 200 mg (2 µmol) wet PHOE, containing 0.7 mmol epoxide groups, was dissolved in 4 mL freshly distilled THF. To the PHOE solution, 1.3 mL (13.4 mmol) diethanolamine was added under argon. The solution was heated to 50 °C and stirred under argon for 24 h. The reaction was terminated by the addition of 7.6 mL of a 1.8 M HCl solution, which was added in order to prevent further hydrolysis of the polyester backbone by forming HCl salts of all amines. The slightly cloudy solution was immediately placed in a dialysis membrane (Spectra/Por 7 MWCO 1 kDa) and dialyzed for 24 h (reservoir changed at 4 h intervals) to remove unreacted diethanolamine HCl salts. The transparent solution was freeze-dried and dry PHON was stored at -20 °C until further use. Polymer Characterization. GPC measurements were performed with a Waters 1515 HPLC system equipped with a Waters 2414 refractive index detector. A Waters Styragel HR 5E column was used with a mobile phase of DMF (0.01 M LiBr) at a flow rate of 1.0 mL/ min at 30 °C. The system was calibrated with monodisperse polystyrene standards. 1H, COSY, and HSQC experiments were performed with a Varian Unity Inova 500 MHz spectrometer equipped with an an indirect

Communications

Biomacromolecules, Vol. 9, No. 8, 2008

2093

Scheme 1. Conversion of PHOU (a) to PHOE (b) to PHON (c)

pulsed field gradient with deuterated chloroform (CDCl3) or deuterated DMSO (DMSO-d6) containing TMS as an internal standard. All experiments were performed at 25 °C.

Results and Discussion A hydrophobic bacterial polyester, PHOU, was effectively converted to a cationic water soluble polyester, PHON, through a series of reactions at the polymer side chains. PHAs containing vinyl-terminated side chains were synthesized using P. putida GPo1 through batch culture techniques. The PHOU consisted of a random bacterial polyester backbone with vinyl and methyl terminated side chains in the ratio of 42 to 58%, respectively. The polymer yield achieved was 15% based on cellular dry weight. The structure of PHOU was confirmed in the 1H NMR spectrum (Figure 1), while the molecular weight and polydispersity of the polymer were determined by GPC (Table 1). The two-step synthesis of PHON from PHOU is outlined in Scheme 1. The first reaction involved the transformation of the vinyl-terminated side chains of PHOU to epoxide groups using the method employed by Bear et al.,5 resulting in PHOE. A 2-fold molar excess of mCPBA was sufficient to yield 100% conversion of PHOU to PHOE, as seen in the 1H NMR analysis (Figure 2). The vinyl proton peaks at 5.8 and 5.0 ppm disappeared, while epoxide protons appeared at 2.5, 2.8, and 2.9 ppm. A 4% molecular weight increase (Mn) was observed upon epoxidation, however, this slight increase may be accounted for by the substitution of an oxygen atom in the epoxide for two hydrogen atoms of a carbon-carbon double bond in the vinyl group (Table 1). Purification of PHOE required special

measures due to polymer cross-linking. Upon drying and attempting to redissolve the polymer in chloroform or THF, it was observed that substantial amounts of polymer swelled, but did not redissolve. This presumably was the result of a small number of epoxide rings opened by methoxide ions present under these conditions in methanol, which, upon drying, initiated cross-linking reactions. To prevent this cross-linking, PHOE was precipitated in methanol and collected as a wet, swelled material. THF was then added immediately, thus preventing the polymer from drying. The small amount of transferred methanol did not interfere with subsequent reactions. To generate a PHA with cationic side chains, the introduction of a protonable amine at functionalized side chains was necessary. It is with this goal in mind that vinyl conversion to epoxide was employed. Traditional epoxide resins are crosslinked via an amine-epoxide reaction, so it was envisioned that this general reaction might serve as the basis for our requirements. However, this reaction normally requires high temperatures, which would likely result in backbone cleavage of PHA. Shechter, et al.6 have suggested a more facile approach by using hydroxyl groups in the solvent to stabilize the opening of the epoxide ring. PHAs are not soluble in polar, hydroxyl-containing solvents, however, we hypothesized that if the amine used in the reaction were to contain hydroxyl groups, and the amine were used at sufficiently high concentration, conditions might be favorable for the reaction to occur. In addition, aminecontaining polyesters would have the potential for self-aminolysis, so a secondary amine (which would convert to a much less reactive tertiary amine) was required. The smallest hy-

2094

Biomacromolecules, Vol. 9, No. 8, 2008

Figure 2. 1H NMR spectrum of PHOE.

Figure 3. HSQC spectrum of PHOE.

Communications

Communications

Figure 4. HSQC spectrum of PHON.

Figure 5. COSY spectrum of PHON.

Biomacromolecules, Vol. 9, No. 8, 2008

2095

2096

Biomacromolecules, Vol. 9, No. 8, 2008

Communications

Table 2. Solubility of PHOU, PHOE, and PHON

methylene chloride chloroform THF DMF DMSO water

PHOU

PHOE

PHON

+ + + + -

+ + + + + -

+ + +

droxyl-containing commercially available secondary amine, diethanolamine, was used to convert PHOE to PHON (Scheme 1). PHOE was dissolved in THF and diethanolamine was added at a 20-fold molar excess. The solution was heated to 50 °C an allowed to proceed for 24 h. Due to peak overlap in the 2.3-2.7 ppm region of the 1H spectrum (data not shown), twodimensional COSY and HSQC were used to monitor the reaction progress. HSQC experiments allow the determination of correlations between protons and carbons that are directly attached to each other, while COSY experiments allow the determination of correlations between protons that are attached to adjacent carbons. Comparison of the HSQC and COSY spectra for PHOE (Figure 3) and PHON (Figures 4 and 5) reveal correlations that indicate successful transformation of the side chain terminal epoxides by diethanolamine to create PHON [Note: Single and double tick marks are used here to differentiate PHOE assignments (′) from PHON assignments (′′) for the same atoms.]. In the HSQC spectrum, the primary correlation which indicates epoxide-opening by diethanolamine is that of the shift of the methine epoxide carbon and proton from g′ (Figure 3) to g′′ (Figure 4). Also, correlations corresponding to the methylene epoxide carbon and its protons shift from h′ and i′ (Figure 3) to h′′ and i′′ (Figure 4). In addition, the original correlations for the methine and methylene carbons and protons of the epoxide (g′, h′, and i′, Figure 3) are completely absent in Figure 4, indicating complete conversion to the amine-terminated side chains. Due to the slight (∼15%) impurity of unreacted diethanolamine in the PHON product, one can see the shift upon attachment of diethanolamine to PHOE within the PHON HSQC spectrum (Figure 4). Although the -CH2-OH corrleations (k′′) do not shift, R-N-(CH2)2- correlations shift from l′′ to j″. The unreacted diethanolamine impurity is likely the result of a small amount of diethanolamine, which remained entrapped through noncovalent interactions within the polymer chains, but to prevent unnecessary polyester hydrolysis or other degradation from extended dialysis time, dialysis purifications were limited to 24 h. Also, differential solvent washing with toluene and THF was attempted but ultimately unsuccessful. As its chemical signature could easily be separated from that of bound diethanolamine, PHON was used at this level of purity. In the COSY spectrum, the primary correlation which indicates epoxide-opening by diethanolamine is the correlation of the methine proton with both its adjacent methylene groups (g′′ f h′′, i′′ and g′′ f f′′) (Figure 5). In addition, the -CH2CH2-OH correlation is observed which indicates protons k′′ have populations that couple to protons j′′ (PHON, major contributor) and to protons l′′ (free diethanolamine, minor contributor; Figure 5).

The GPC chromatogram revealed an extensive decrease in molecular weight (Table 1). The degradation is the result of aminolysis of the polymer backbone esters due to the large excess of amine, specifically on average 6 hydrolysis reactions per polymer chain, or 1% of the total number of ester bonds in the polymer chain. However, as the resulting amides (aminolysis of ester bond) would merely terminate a single end of the polymer chains, their presence could not be detected in any of the NMR experiments due to their comparatively small concentrations. Since the application of polycationic PHAs are envisioned in the biomedical field, specifically in plasmid DNA delivery, the decrease in molecular weight could potentially prove beneficial due to the general toxicity of higher molecular weight polymers. The successful side chain conversion was further substantiated by the change in solubility when converting PHOU to PHOE to PHON (Table 2). As the functionalized side chains became more polar, the polymer became soluble in more polar solvents. Ultimately, PHON was completely soluble in pure water.

Conclusions Microorganisms do not produce PHAs with terminal amino groups, presumably because cationic polyelectrolytes are toxic and are, therefore, not produced. As a result, covalent modification of precursor PHAs is required. Several attempts by large, experienced PHA research groups failed in producing this type of polymer. Hence, two novel contributions were made to the field of PHA research: (1) production of a PHA with pendant amine side groups and (2) production of cationic PHA. Acknowledgment. This work was supported by the Alabama Space Grant Consortium and the Partnership for Biotechnology Research. J.S. would like to thank Drs. Willy Vayaboury and Bernhard Vogler for assistance.

References and Notes (1) (a) Findley, R. Appl. EnViron. Microbiol. 1983, 45, 71–78. (b) Steinbuchel, A. Polyhydroxyalkanoic Acids; Macmillan Publishers: New York, NY, 1992. (2) Lenz, R.; Marchessault, R. Biomacromolecules 2005, 6, 1–8. (3) (a) Fritzsche, K.; Lenz, R.; Fuller, R. Macromol. Chem. 1990, 191, 1957–1965. (b) Kim, Y.; Lenz, R.; Fuller, R. Macromolecules 1992, 25, 1852–1857. (c) Hazer, B.; Lenz, R.; Fuller, R. Polymer 1996, 37, 5951–5957. (d) Scholz, C.; Fuller, R.; Lenz, R. Macromol.Chem. Phys. 1994, 195, 1405–1421. (e) Shah, D.; Tran, M.; Berger, P.; Aggarwal, P.; Asrar, J.; Madden, L.; Anderson, A. Macromolecules 2000, 33, 2875–2880. (f) Hazer, B.; Lenz, R.; Fuller, R. Macromolecules 1994, 27, 45–49. (g) Fritzsche, K.; Lenz, R.; Fuller, R. Int. J. Biol. Macromol. 1990, 12, 92–101. (h) Fritzsche, K.; Lenz, R.; Fuller, R. Int. J. Biol. Macromol. 1990, 12, 85–91. (4) (a) Lee, M.; Park, W.; Lenz, R. Polymer 2000, 41, 1703–1709. (b) Renard, E.; Ternat, C.; Langlois, V.; Guerin, P. Macromol. Biosci. 2003, 3, 248–252. (5) Bear, M.; Leboucher-Durand, M.; Langlois, V.; Lenz, R.; Goodwin, S.; Guerin, P. React. Funct. Polym. 1997, 34, 65–77. (6) Shechter, L.; Wynstra, J.; Kurkjy, R. J. Ind. Eng. Chem. 1956, 48, 94–97.

BM8005616