Biomacromolecules 2003, 4, 1698-1702
1698
Physical Properties of Microbial Polythioesters: Characterization of Poly(3-mercaptoalkanoates) Synthesized by Engineered Escherichia coli Jumpei Kawada,† Tina Lu¨tke-Eversloh,‡ Alexander Steinbu¨chel,‡ and Robert H. Marchessault*,† Department of Chemistry, McGill University, 3420 University St., Montreal, QC, H3A 2A7, Canada, and Institut fu¨r Molekulare Mikrobiologie und Biotechnologie, Westfa¨lische Wilhelms-Universita¨t Mu¨nster, Corrensstrasse 3, D-48149 Mu¨nster, Germany Received May 1, 2003; Revised Manuscript Received July 15, 2003
Physical properties of chiral poly(thioesters), PTEs, prepared by engineered Escherichia coli, were examined by GPC, 13C CP/MAS solid-state NMR, X-ray diffraction, and thermal analysis. Microbial homopolymers of PTEs, poly(3-mercaptopropionate), PMP, and poly(3-mercaptovalerate), PMV, showed different solubility characteristics compared to poly(hydroxyalkanoates), PHAs. Generally, PTEs required higher temperatures for dissolution. Poly(3-mercaptobutyrate), PMB, and PMV dissolve in chloroform, and the molecular weight values were revealed by GPC as 176 000 and 165 000, respectively. The density values for PMP and PMB were 1.42 and 1.27 g/cm3, respectively. These values are similar to those for oxygen analogues. The NMR spectra for PTEs showed that carbonyl carbons are greatly shifted downfield by the sulfur atoms in the chain backbone compared to the PHA family. X-ray powder diffraction data indicated that PTEs are crystalline materials, but they do not crystallize as well as in the PHA family. The melting point, Tm, for PMP was 170 °C, which is about 100 °C higher than the equivalent oxygen analogue, poly(3-hydroxypropionate), PHP, and almost the same as that of bacterial poly(3-hydroxybutyrate), PHB. According to thermal analysis, only the PMP sample had enhanced heat stability, e.g., the decomposition temperature for PMP was 277 °C at 5% weight loss, whereas the values for PHP and PHB were 233 and 260 °C at the same weight loss, respectively. Introduction Poly(thioesters), PTEs, have recently been synthesized in a “biological factory” by engineered Escherichia coli.1 This production of new bacterial polyesters through a recombinant strain of E. coli is a developing polymer bioproducts synthesis revolution. Marvel and Kotch made PTEs2 based on polycondensation in 1951 by using dibasic acid chlorides and dithiols which mimicked the diacid and diol condensation method pioneered by Carothers. Synthesis of poly(thioglycolide) was first described in 1960 by Scho¨berl,3 eight years later Overberger and Weise synthesiszed poly(-thiocaprolactone) using ring-opening polymerization, and they reported a 45 °C higher melting point, Tm, than for the oxygen analogue i.e. poly(-caprolactone).4 Afterward, polymer chemists made various PTEs5-8 such as poly(thiolactide)9 and poly(3-mercaptopropionate).10 Nowadays, PTEs are mainly made by ring-opening polymerization,11,12 and melting points for the mercapto versions were higher than for their oxygen analogues. Recently, polyhydroxyalkanoate (PHA) accumulating bacteria have been shown to synthesize copolymers containing * To whom correspondence should be addressed. Phone: 1-514-3986276. Fax: 1-514-398-7249. E-mail:
[email protected]. † McGill University. ‡ Westfa ¨ lische Wilhelms-Universita¨t Mu¨nster.
3-mercaptopropionate (3MP) or 3-mercaptobutyrate (3MB), in addition to 3-hydroxybutyrate (3HB), representing the first examples of PTEs as a novel class of biopolymers.13,14 Afterward, a recombinant strain of E. coli expressing a nonnatural PHA biosynthesis pathway, was used as a biotechnological tool to produce novel homopolythioesters.1 These homopolymers of PTEs are poly(mercaptopropionate), PMP, poly(mercaptobutyrate), PMB, and poly(mercaptovalerate), PMV. The chemical formulas for PTEs are identical to those for PHAs, except for the sulfur atoms in the chain backbone (cf. Table 2). All microbial PTEs are 100% isotactic polymers i.e. there is a singular chilarity of the asymmetric carbon in the chain backbone.1 Relatively little is known about the biodegradability of PTEs. Recently, a novel bacterium, Schlegelella thermodepolymerans, was isolated which used poly(3HB-co-3MP) as the sole carbon source for growth.15 However, it is not generally known whether PTEs are degraded by the wellstudied PHA depolymerases.16 The biodegradability of PTEs is currently under investigation.15 Our incentive for examining physical properties of PTEs is based on the marked difference in melting point, Tm, between polyoxyethers and polythioethers as shown in Figure 1. The Tm for polythioethers is always higher than that for polyoxyethers. Because PTEs are relatively unknown, this
10.1021/bm0341327 CCC: $25.00 © 2003 American Chemical Society Published on Web 08/23/2003
E. coli Engineered Poly(3-mercaptoalkanoates)
Biomacromolecules, Vol. 4, No. 6, 2003 1699
Table 1. Solubility Results of PMP, PMV, and PHVa
a
The temperature required for dissolution is noted.
Figure 1. Melting points of polyethers with the repeat unit -[(CH2)nX]-.
study focuses on their physical properties and comparing these new microbial entities with the well-known PHAs. Materials and Methods All PTE samples (PMP, PMB and PMV) were synthesized and purified as described previously1. Poly(3-hydroxypropionate), PHP, was provided by Dr. Nishida of Kinki University, Japan. Poly(3-hydroxybutyrate), PHB, was ob-
tained from Imperial Chemicals Ltd., ICI. Poly(3-hydroxyvalerate), PHV, was obtained from cells of Chromobacterium Violaceum grown on valeric acid as described.17 Solubility Test. The samples (PMP, PMV, and PHV, the latter was used as reference) were tested for solubility in the following solvents: chloroform, dimethyl formamide, 2-methoxyethyl ether, 1-methyl-2-pyrrolidone, nitrobenzene, dimethyl sulfoxide, ethyl-(S)-lactate, and triacetin. Gel Permeation Chromatography (GPC). GPC analyses were performed at room temperature with chloroform as eluent, at a flow rate of 1 mL/min. Two Waters Styragel columns HR3 and HR4 connected in series were used, and the detector was a Hewlett-Packard refractive index HP 1047 RI. Polystyrene standards were used for calibration. Density Measurement. Density measurements of PTEs were carried out using two solvents, which neither dissolve nor swell the sample, such as xylene and carbon tetrachloride. A specimen, cut into small pieces, was put into a mixture of xylene and carbon tetrachloride (CCl4) in a ground glass stoppered measuring cylinder which was kept in a thermostat bath at 25 ( 1.0 °C. Xylene or CCl4, as necessary, were added to the suspension until the sample floated, indicating that the solution and the sample are of the same density.
1700
Biomacromolecules, Vol. 4, No. 6, 2003
Kawada et al.
Table 2. Physical Properties of PHAs and PTEs
* From Polymer handbook, 4th ed.; Brandrup, J., Immergut, E. H., Grulke, E. A., Eds.; John Wiley & Sons, Inc.: New York, 1999.
The density of the solution was calculated by the weight and the volume of the components. 13C CP/MAS Solid State NMR. Cross-polarization/magic angle spinning (CP/MAS) NMR spectra were recorded by a Chemagnetics CMX-300 instrument operating at 75.4 MHz for 13C. Samples were packed in 7.5 mm PENCIL rotors and spun at 4000 Hz. X-ray Diffraction Analysis. X-ray diffraction data were collected by using a Rigaku D/Max 2400 diffractometer operated at 40 kV and 160 mA at room atmosphere. A copper X-ray anode was used with graphite monochromator to provide CuKa radiation. Transmission Electron Microscopic Analysis. Single crystals of PMV samples were grown by a method modified from that of Hocking et al. 18 PMV was dissolved in chloroform, to which was added warmed ethanol (ca. 60 °C) to give an ethanol/chloroform ratio of 2:1, at a polymer concentration of 0.02%. The turbid solution was brought to 70 °C and held for 30 min until disappearance of the turbidity, and afterward it was slowly cooled to room temperature. The turbid solution was reheated and held at 65 °C for 10 h to permit self-seeding crystal growth and was then slowly cooled to room temperature. Drops of PMV single crystals were deposited on carbon coated grids and allowed to dry. The grids were examined with a JEOL 2000FX transmission electron microscope (TEM) operating at 80 kV for imaging. Electron micrographs were recorded electronically. Thermal Analysis (DTA and DSC). Differential thermal analysis, DTA, was performed using a SEIKO TG/DTA 220 instrument over the temperature range from room temperature to 400 °C by ramping at 10 °C/min. For differential scanning calorimetry (DSC), a DSC Q1000 calorimeter from TA
Instrument was used. Pretreatment of samples was as follows: first, sample endotherms were recorded from room temperature to 190 °C by 10 °C/min ramping; second, they were cooled from +190 to -50 °C by 40 °C/min ramping and kept at that temperature for 10 min. Afterward, the second DSC run, with ramping at 10 °C/min, provided the melting temperature (Tm) from the observed endotherm. The endotherms were examined for glass transition temperatures (Tg) over the temperature range studied. Polarization Microscopy. A PMP sample was heated until it melted on a microscope slide. The melted sample was then cooled to room temperature, and micrographs were recorded using a polarization microscope, Nikon MICROPHOT-FXA. Results and Discussion In reviewing the features of thiopolymers, one notes that increased thermal stability is a general characteristic. The following compilation is far from comprehensive, but given the 100% isotactic structure1 of PMB and PMV using this engineered E. coli biosynthesis, it is important to record some of their basic physical properties. Solution Properties. In our previous paper19 on the physical characterization of PTE copolymers, all samples were soluble in chloroform at room temperature. However, the solubility characteristics (Table 1) of both homopolythioesters (PMP and PMV) were different from the copolymer of polythioesters. Thus, PMV was soluble in CHCl3 at room temperature like PHV, whereas PMP was insoluble at any temperature up to the boiling point. In all other solvents, the temperature had to be raised above room temperature to bring about solubility except for triacetin which was a nonsolvent. Of the three samples, PMP always required the
E. coli Engineered Poly(3-mercaptoalkanoates)
Figure 2.
Figure 3.
13C
13C
Biomacromolecules, Vol. 4, No. 6, 2003 1701
Figure 4. X-ray powder diffraction traces of PTEs.
CP/MAS solid-state NMR spectra of PTEs.
CP/MAS solid-state NMR spectra of PHAs.
Table 3. 13C CP/MAS NMR Resonance Comparison between PTEs and PHAs 13C
chemicals shift [ppm]
carbon
PMP
PMB
PMV
PHP
PHB
PHV
C1 C2 C3 C4 C5
197 46 24
197 49 37 22
197 46 42 27 10,12
173 31 62
170 43 69 21
170 40 71 29 10
highest temperature which is tentatively attributed to higher cohesive energy and higher degree of crystallinity. The temperature required to dissolve in most solvents may relate to the melting points which are in the order PMV < PHV < PMP (cf. the Thermal Analysis section). Because PMB and PMV dissolve in CHCl3 at room temperature, molecular weights were measured by GPC, yielding the values for PMB and PMV: 176 000 and 165 000, respectively. The molecular weight for these homopolythioester was lower than reported for copolymer polythioesters which range from 100 000 to 1 000 000.19 The polydispersity was 1.9 for PMB and 2.0 for PMV. Solid State Properties. The densities for PTEs as revealed by a flotation method had values for PMP and PMB of 1.42 and 1.27 g/cm3, respectively. These values are similar to those calculated for the equivalent oxygen analogues from unit cell data (100% crystalline) as shown in Table 2. It is noteworthy that PMV dissolved in CCl4; hence, its density could not be determined. The 13C CP/MAS solid-state NMR spectra of PTEs and PHAs are shown in Figures 2 and 3, respectively. Table 3 shows all of the resonance assignments from which the effect of sulfur on the chemical shifts can be compared. The carbonyl carbons (C1) for PTEs are shifted downfield relative
to PHAs because of the nonconjugated state of the sulfur atom with respect to C1 for PTEs. By contrast, the C3 resonances in PTEs move upfield relative to PHAs because sulfur is less electronegative than oxygen. The C3 resonances of PMB and PMV are different from that for PMP because of the substituent effects. Similary, C3 resonances in PHB and PHV were also shifted by the same substituent effect compared to PHP. A doublet resonance at C5 in the PMV spectrum may tell us that C5 has two different conformations. The X-ray powder diffraction traces of PTEs are shown in Figure 4. The observed d spacings for PTEs are different from those reported for equivalent oxygen analogues. PMP has a sharp diffraction peak at 21.3°, 2θ, which was also detected for poly(3HB-co-3MP) having high 3MP content.19 The PMB diffractometer trace lacks resolution, which shows that PMB is of much lower crystallinity than the other two samples (PMP and PMV). PMV shows a long d spacing at 7.7°, 2θ, (1.1 nm) which the other two samples and PHV do not display. As we reported in our previous paper,1 the PHA family has ester dipole crystallization forces which are weaker in PTEs. In other words, when a sulfur atom is inserted in the chain backbone, replacing an oxygen atom, crystallization may be hindered because the electronegativity of sulfur (2.58) is much lower than that of oxygen (3.44). Sulfur’s value is the same as that for carbon (2.55). In the case of PHB, ester dipole interactions stabilize its helical conformation,20 sulfur atoms may not have this favorable Coulombic influence on helix stability. PMV single crystals were examined by TEM. PMV crystals are poorly defined; in other words, crystals lacked the edge sharpness and well-defined thickness characteristic of PHA single crystals. One notes two kinds of PMV single crystals; one is a small round type (“blob”), and the other is a large hexagon. Both compositions were confirmed to be PMV by energy dispersive spectroscopy (EDS). Thermal Analysis. All PTEs and PHAs were analyzed by DSC, and the Tg and Tm results are shown in Table 2. The PMP sample melted at 170 °C which is much higher than for PHP. This is in keeping with the observed trend in the melting points for aliphatic polyethers compared to the thio-equivalents as shown in Figure 1. As suggested above, a change in helix stabilization by intramolecular Coulombic forces when oxygen is replaced by sulfur may be responsible for a decrease in enthalpy. Interestingly, poly(-thiocaprolactone) has an enhanced Tm by ca. 45 °C compared to its oxygen analogue.4 Perhaps, thiopolymers have higher melting points when the chains are planar zigzag and do not have
1702
Biomacromolecules, Vol. 4, No. 6, 2003
Figure 5. Thermal decomposition thermograms of PMP and PHP.
side chains, like polythioethers as shown in Figure 1. The rule (higher Tm for thiopolymers) can be applied not only to polar thiopolymers but also to nonpolar. For example, Tm for poly(ethylenedithioladipate), poly(butylenedithioladipate), poly(hexamethylenedithioladipate), poly(ethylenedithiolterephthalate), and poly(butylenedithiol-terephthalate) are significantly higher than those for the corresponding oxygen analogues.2,21,22 However, the Tm for poly(thioglycolide)3,7 is not higher than that for poly(glycolide). All three PTE samples were examined in a polarizing microscope following a crystallization treatment. After heating past the melting point and cooling, only the PMP sample showed a spherulitic appearance between crossed polars. In other words, both PMB and PMV did not crystallize well after melting and cooling. This is the reason PMB and PMV displayed a readily detectable Tg on thermal analysis since crystalline polymers often have a Tg which is difficult to detect. Thermal Decomposition. Thermal gravimetric traces for PTEs and PHAs were examined, and only the data for PMP and PHP are shown in Figure 5. The heat stability for PMP was better than PHP. The residual weight value of PHP was only 11.8% at 280 °C on decomposition, whereas the value for PMP was 93.2% at the same temperature. The values at 5% decomposition weight loss, Td(5%), are also listed in Table 2 where PMP has the highest temperature value. The thermograms of PMP leveled off at a residual weight value of 16%, which persisted to 400 °C. Ongoing heating caused PHAs to decompose completely to H2O and CO2, whereas PMP left a black char. The thermograms of PMB and PMV are similar to that of PHB. It is difficult to explain why only PMP showed black char, but the char itself suggests formation of cross-links during the pyrolysis process. Conclusions Genetically engineered bacteria can become polymer factories of the future as shown in the PTEs which were biosynthesized by using a nonnatural pathway in bacteria. The microbial polymers produced are 100% isotactic which are difficult to synthesize, especially for PTEs. PTEs are interesting materials because of their different physical properties compared to PHAs. For example, PTEs
Kawada et al.
are more difficult to crystallize than PHAs, and the spherolytic texture in PTEs develops less readily than that for PHAs. The melting point of PMP, a nontactic homopolymer, is almost the same as for PHB, and PMP has greater heat stability than PHP or PHB. Also PMP requires higher temperature to dissolve in a range of solvents, which suggests that PMP has high cohesive energy. Because there are some solvents to dissolve PMP, film casting and fiber spinning are technologically feasible. PMP truly follows the unwritten rule that thiopolymers are usually more heat stable than the oxygen analogue. If PTEs have some desirable properties because of the sulfur content (e.g., antibacterial property), then PMP may be the most promising candidate to use in such a practical application. It responds to compression molding better than PMB and PMV. Further studies should provide missing information such as molecular weight, dynamic mechanical properties, biodegradability, orientability, etc., and because there is no asymmetric carbon in PMP, the chemical synthesis challenge is more feasible. Acknowledgment. We are grateful to Professor Nishida of Kinki University for supplying the PHP sample. We also thank Dr. Fred Morin of the McGill Chemistry Department for recording and discussing the solid state NMR spectra. Financial support from NSERC (Natural Sciences and Engineering Research Council of Canada) is gratefully acknowledged. References and Notes (1) Lu¨tke-Eversloh, T.; Fischer, A.; Remminghorst, U.; Kawada, J.; Marchessault, R. H.; Bo¨gershausen, A.; Kalwei, M.; Eckert, H.; Reichelt, R.; Liu, S.-J.; Steinbu¨chel, A. Nat. Mater. 2002, 1, 236-240. (2) Marvel, C. S.; Kotch, A. J. Am. Chem. Soc. 1951, 73, 1100-1102. (3) Scho¨berl, V. A. Makromol. Chem. 1960, 37, 64-70. (4) Overberger, C. G.; Weise, J. K. J. Am. Chem. Soc. 1968, 90, 35333537. (5) Bu¨hrer, H. G.; Elias, V. H.-G. AdV. Chem. Ser. 1973, 129, 105130. (6) Goethals, E. J. In Topics in sulfur chemistry; Senning A., Magee, P. S., Eds.; Gerog Thieme Publishers: Stuttgart, Germany, 1977; pp 3-61. (7) Elias, V. H.-G.; Bu¨hrer, H. G. Makromol. Chem. 1970, 140, 21-39. (8) Kricheldorf, H. R.; Bo¨singer, K. Makromol. Chem. 1973, 173, 6780. (9) Bu¨hrer, V. H. G.; Elias, H.-G. Makromol. Chem. 1970, 140, 41-54. (10) Kricheldorf, H. R. Makromol. Chem. 1973, 173, 81-89. (11) Kricheldorf, H. R.; Probst, N.; Schwarz, G.; Schulz, G.; Kru¨ger, R.P. J. Polym. Sci., Polym. Chem. Ed. 2000, 38, 3656-3664. (12) Sanda, F.; Jirakanjana, D.; Hitomi, M.; Endo, T. J. Polym. Sci., Polym. Chem. Ed. 2000, 38, 4057-4061. (13) Lu¨tke-Eversloh, T.; Bergander, K.; Luftmann, H.; Steinbu¨chel, A. Microbiology 2001, 147, 11-19. (14) Lu¨tke-Eversloh, T.; Bergander, K.; Luftmann, H.; Steinbu¨chel, A. Biomacromolecules 2001, 2, 1061-1065. (15) Elbanna, K.; Lu¨tke-Eversloh, T.; Van Trappen, S.; Mergaert, J.; Swings, J.; Steinbu¨chel, A. Int. J. Syst. EVol. Microbiol. 2003, 53, 1165-1168. (16) Jendrossek, D. In Biopolymers; Doi, Y., Steinbu¨chel, A., Eds.; WileyVCH: Weinheim, Germany, 2002; Vol. 3b, pp 41-77. (17) Marchessault, R. H.; Debzi, E. M.; Revol, J.-F.; Steinbu¨chel, A. Can. J. Microbiol. 1995, 41, 297-302. (18) Hocking, P. J.; Revol, J.-F.; Marchessault, R. H. Macromolecule 1996, 29, 2467-2471. (19) Lu¨tke-Eversloh, T.; Kawada, J.; Marchessault, R. H.; Steinbu¨chel, A. Biomacromolecules 2002, 3, 159-166. (20) Cornibert, J.; Marchessault, R. H. J. Mol. Biol. 1972, 71, 735-756. (21) Polymer handbook, 4th ed.; Brandrup, J., Immeergut, E. H., Grulke, E. A., Eds.; John Wiley & Sons: New York, 1999.
BM0341327
