Biomacromolecules 2003, 4, 1092-1097
1092
Fourier Transform Infrared Spectroscopy for Screening and Quantifying Production of PHAs by Pseudomonas Grown on Sodium Octanoate Solo Randriamahefa, Estelle Renard, Philippe Gue´ rin, and Vale´ rie Langlois* Laboratoire de Recherches sur les Polyme` res, UMR 7581, 2 a` 8 rue Henri Dunant, 94 320 Thiais, France Received April 4, 2003; Revised Manuscript Received May 7, 2003
Poly(hydroxyalkanoates) PHAs are synthesized by many bacteria as inclusion bodies and their biodegradability and structural diversity have been studied with a view to their potential application as biodegradable materials. A method based on FT-IR was developed to carry out rapid qualitative and quantitative analysis of PHAs in Pseudomonas, when they were grown on sodium octanoate. Using absorbance of the ester band of PHAs, a rapid method was reported to distinguish PHB and PHO and to determine polymer content in intact bacteria. Relative areas in which the CdO area was normalized to the area of the peak representing the amid group (1656 cm-1) characteristic of bacteria were calibrated to the polymer content which was determined after solvent extraction. Polymer contents vary from 0% to 53% and depend on the nature of the bacteria. Among 27 strains of Pseudomonas belonging to the rRNA homology group I, a very low amount of bacteria were able to produce PHB. The majority of strains were able to produce a copolymer, PHO, in which the major constituent unit is 3-hydroxyoctanoate. The FT-IR results were further confirmed by gas chromatography analysis after methanolysis of polymer, but FT-IR method requires less preparation of sample than gas chromatography and it is very useful for screening a large variety of Pseudomonas. Introduction A wide variety of prokaryotic microorganisms are used in the synthesis of poly(hydroxyalkanoates), PHAs, which are considered as good candidates for biodegradable and biocompatible polymers.1,2 These polymers serve as intracellular and energy reserve materials and can be produced in large quantities by a fermentation process. Poly(3-hydroxybutyrate), PHB, is the commonest member of the PHA family; it belongs to the short chain length PHAs, scl PHAs, with its monomer units containing 4 carbon atoms, whereas PHAs containing monomer units consisting of 6-16 carbon atoms have been termed medium chain length, mcl PHAs (Figure 1).3 It has been clearly demonstrated by several authors that FT-IR spectra from bacteria can be used for identification and differentiation. FT-IR can also be used to detect and identify particular cells constituents such as capsules or storage materials.4,5,6 It has been reported that PHB is observable in FT-IR spectra in Bacillus megaterium.7 In a recent study, Hong8 extended the observation that not only PHB but also mcl PHAs can be rapidly detected by FT-IR in intact cells of Pseudomonas. Recently Wu9 has developed an approach for rapid differentiation between scl PHAs and mcl PHAs by spectrofluorimetry. Several methods have already been used to determine polymer content in bacteria, but they require extensive and complicated sample preparation, like extraction, purification, or methanolysis.10 * To whom correspondence should be addressed. E-mail: langlois@ glvt-cnrs.fr. Fax: 33 1 49 78 12 08.
Figure 1. Chemical structures of some scl and mcl PHAs
In this paper, we used FT-IR analysis which is a rapid analysis involving minimal sample preparation, and we showed that FT-IR is particularly suitable for screening large amounts of Pseudomonas belonging to the 16S rRNA homology group I,11 for their abilities to produce PHAs. We developed a rapid method to distinguish PHB and PHO and to quantify the content of PHAs in cells when they are growing on sodium octanoate, without extraction of the polymer. When Pseudomonas were cultivated on sodium octanoate, most of them are able to synthesize in certain conditions a copolymer called poly(hydroxyoctanoate) in which the major repetition unit is the 3-hydroxyoctanoate.12 Experimental Section Strains belonging to the genus Pseudomonas were studied. All bacteria were cultivated as previously described12 at 30 °C in 500 mL Erlenmeyer flasks containing 50 mL of mineral salt medium and agitated to 200 rpm. Composition of the mineral salt medium was as follows: CaCl2, 2H2O: 15 mg; MgSO4, 7H2O: 123 mg; KH2PO4: 680 mg; K2HPO4, 3H2O: 2610 mg; NaCl: 7 g; (NH4)2SO4: 525 mg for 1 L
10.1021/bm034104o CCC: $25.00 © 2003 American Chemical Society Published on Web 06/06/2003
Production of PHAs by Pseudomonas
Biomacromolecules, Vol. 4, No. 4, 2003 1093
Figure 2. Comparison of FT-IR (transmission) spectra of PHO film (a) and PHB film (b).
of distilled water. The medium was complemented with 10 mL of microelements (1 L of microelements stock solution contains: H3PO4 (85% v:v): 1.15 mL; FeSO4, 7H2O: 56 mg; ZnSO4, 7H2O: 29 mg, MnSO4, 4H2O: 22 mg; CuSO4, H2O: 2.5 mg; Co(NO3)2, H2O: 3 mg; H3BO3: 6 mg. Sodium octanoate was provided with a concentration of 20 mM. After reaching the early stationary growth phase, the cells were harvested by centrifugation and the biomass was frozen and lyophilized. The polymer produced were extracted by refluxing in chloroform, then precipitated twice into ethanol and dried under vacuum. PHAs content was determined by the ratio of weight of dry polymer to the lyophilized bacteria weight. The lyophilized bacteria were analyzed by FTIR (Golden Gate ATR). The spectra were recorded on a TENSOR 27 BRUKER apparatus (Digi Tect DLaTGS detector, 32 scans, 1 cm-1). PHAs films were cast from 1 to 2wt % solution in CHCl3 on a KBr support and vacuum-dried. Films were analyzed by transmission with a Tensor 27 Bruker apparatus. For the determination of the composition of PHAs, 5 mg of sample was mixed with 1 mL of chloroform in a small capped test tube. Then, 0.85 mL of 2% (v:v) sulfuric acid in methanol was added, and the solution was heated for 2 h at 100 °C. After cooling the sample, 1 mL of demineralized water was added. The chloroform layer was analyzed by GC. Tetradecane was used as an internal standard. The solutions of methanolyzed polymers were analyzed with a Varian 3300 chromatograph equipped with a 15 m × 0.53 mm Ohio Valley column OV 351. The column temperature was programmed from 50 to 170 °C at a rate of
5 °C/min, then from 170 to 230 °C at a rate of 20 °C/min. The FID detector temperature was set at 280 °C, and the injector at 230 °C. The melting temperature (Tm) and glass transition (Tg) were measured by using a Perkin-Elmer differential scanning calorimetry (DSC 7) at a rate of 10 °C/min. Molecular weights were determined by SEC with a Waters apparatus (510 pump). The mobile phase was tetrahydrofuran, with an eluent flow rate of 1 mL/min, using two PLgel mixed C type columns (30 cm × 7.5 mm internal diameter). A differential refractometer detector (Waters 410) was used for detection. Results and Discussion 1. PHAs Characteristic Bands in FT-IR Spectra. We have compared the FT-IR spectra of PHB extracted from B. cepacia LMG 1222 and PHO produced by P. sp. Go1 105816 (Figure 2). Polyesters were extracted from bacteria and twice precipitated in ethanol. After vacuum-drying, films of polyesters were prepared by dissolution in CH2Cl2 and solvent evaporation. This figure clearly indicates the differences between PHO and PHB. The methylene C-H vibration near 2900 cm-1 has the strongest band in spectra of PHO, the weakest one in PHB. The most prominent marker band for PHAs is the ester carbonyl band near 1720-1740 cm-1, and we can distinguish PHB and PHO. PHB shows its strongest band at 1724 cm-1, whereas PHO shows its ester band at 1742 cm-1 (Table 1). When the PHB film is obtained after dissolution in chloroform and solvent evaporation, Bloembergen et al.13
1094
Biomacromolecules, Vol. 4, No. 4, 2003
Randriamahefa et al.
Table 1. Characteristics of PHAs wavenumber (cm-1)c PHBa PHOb
Cd0
C-O
CIc
Tm (°C)d
∆Hf(J/g)d
1724 1742
1185 1165
0.91 0.41
175.2 50.2
86.0 17.3
a PHB extracted from B. cepacia LMG 1222. b PHO obtained from P. sp Gpo1 105816. c Wavenumbers (cm-1) and crystallinity index (CI) of PHAs samples determined by FT-IR. d Determined by DSC.
showed the effect of crystallinity on the CdO peak wavenumber. After 1 min of solvent evaporation, the band was at 1740 cm-1, whereas it was at 1725 cm-1 after 17 h of solvent evaporation. We can conclude that our PHB film crystallinity is maximal because the observed band is at 1724 cm-1. With this same film preparation, PHO shows its ester band at 1742 cm-1. FT-IR was also used by Xu14 to evaluate the crystallinity of PHB quantitatively during melting and crystallization. The bands at 1742 and 1725 cm-1 were attributed to the stretching vibration of the amorphous and crystalline carbonyl groups respectively.15 This shift of the ester band is due to the variation of the dipole moment.16 Other accompanying bands near 1185 cm-1 helped to identify the types of PHAs: PHB shows its band at 1185 cm-1, whereas PHO shows its band at 1165 cm-1. The absorption bands of hydroxyl groups are observed near 3500 cm-1, but in a very low amount. A relative measure of the degree of crystallinity was obtained by normalizing the absorbance at 1185 cm-1 in the case of PHB (and at 1165 cm-1 for PHO) to that of 1385 cm-1 band (δs), which is insensitive to the degree of crystallinity. The bands at 1185, 1228, and 1279 cm-1 are crystallinity-sensitive bands but the band at 1185 cm-1 is better resolved than the others. These bands are characteristic of νa (C-O-C). Hence, a crystallinity index, CI, was defined as the ratio of the bands intensities at 1385 and 1185 cm-1. This crystallinity index is not to be confused with an absolute degree of crystallinity but is useful as a comparison criterion.17 Table 1 shows the crystallinity index (CI) measured by FT-IR. Crystallinity has a tendency to decrease when the side chain length increases. These results were confirmed by DSC. The crystallinity index of PHB is equal to 0.91 and the PHB presents a melting temperature of 175.2 °C, whereas PHO has a lower CI (0.41) and a lower Tm (50.2 °C). 2. Intact Bacteria. Different Pseudomonas were cultivated on sodium octanoate, and FTIR analyses were done on lyophilized bacteria. The infrared spectra of bacteria are clearly dominated by the characteristic absorptions of the cellular proteins.18 A strong amide I band was detected within 1600 and 1700 cm-1, and additional bands at 1546 and 1454 cm-1 stand for polypeptides structures (amide II) and CH bending vibrations of methyl and methylene groups, respectively. These features indicate the existence of polypeptide. The most prominent marker band for polyester storage compound is the ester carbonyl band near 1730 cm-1 (Figure 3). It is more difficult to distinguish PHO and PHB contained in bacteria by studying the wavenumber of the CdO group
Figure 3. FT-IR (ATR) spectra of various cells containing PHA: (a) cells of P. fluorescens DSM 50415 containing PHO, (b) cells of P. pseudoalcaligenes LMG 1225 containing PHB.
as it was possible on extracted polymer. Intracellular PHB presents an ester band at 1724 cm-1 and intracellular PHO has an ester band at 1731 cm-1. The difference between these two wavenumbers is lower (6 cm-1) than the observed difference in PHO and PHB films (18 cm-1). The accompanying bands visible in extracted PHA were also observable in intact cells, and we showed that the bands at 1180 and 1163 cm-1 are respectively characteristic of PHB and PHO. We used this band to identify the nature of PHA (PHB or PHO). P. pseudoalcaligenes LMG 1225 containing PHB had a band at 1180 cm-1, whereas P. fluorescens DSM 50415 which produced PHO had its band at 1163 cm-1. This demonstrates further that the bands at 1185 and 1165 cm-1 are respectively characteristic of PHB and PHO. No band was observable near 1185 or 1165 cm-1 when P. sp GP01 accumulated no PHA, further demonstrating that the band near 1185 or 1165 cm-1 is characteristic of PHAs (Figure 4). 3. Determination of PHA Content by FT-IR. To estimate polymer content in intact cells after cultivation on sodium octanoate, we used the non destructive ATR analysis. By comparison of the absorbance at 1731 cm-1, Figure 4 shows that P. fluorescens DSM 50415 accumulates more intracellular PHO than P. fluorescens LMG 5940. The
Production of PHAs by Pseudomonas
Biomacromolecules, Vol. 4, No. 4, 2003 1095
Figure 4. FT-IR (ATR) spectra of lyophilized bacteria. (a) P. sp Gpo1 CIP 105816 not containing PHA. (b) P. fluorescens LMG 5940 containing PHO. (c) P. fluorescens DSM 50415 containing PHO.
Figure 5. Variation of the ratio of the IR absorbance of the CdO band at 1731 cm-1 to the absorbance of the amid band at 1656 cm-1 with polymer content determined after solvent extraction.
FT-IR spectrum of the lyophilized cells of P. sp Gpo1 CIP 10581 growing without sodium octanoate contains a band at 1731 cm-1, which has a very low intensity, probably due to the presence of protein content. To make quantitative analysis of polymer content, we calculated the ratio of the intensities of the band at 1731 cm-1, due to PHA, and the band at 1656 cm-1, due to strain. This ratio was correlated to the PHA content, which was determined after solvent extraction. Polymer content was determined by the proportion of dry extracted polymer from intact lyophilized cells. Figure 5 shows the calibration curve for PHA content determined by FT-IR. A correlation coefficient, r, of 0.991 was determined by linear regression analysis. By this fermentation process, strains are not able to accumulate more than 53% of polymer, that is the reason the calibration curve stopped at 53% of PHA content.
This very rapid and nondestructive method has been correlated to GC analysis, and the results shown in Table 2 were in good agreement. It can be concluded that FTIR is a rapid method to determine polymer content in bacteria. We have determined PHO content in different cells cultivated on sodium octanoate, with a view to test the ability of different Pseudomonas to produce biopolyesters. 4. Screening of PHAs Production by Different Pseudomonas. We investigated the PHA composition of 27 different Pseudomonas belonging to the rRNA homology group I, as defined by Yamamoto.11 Strains were cultivated on sodium octanoate. Intact cells containing PHAs were studied by FT-IR to identify and quantify the bacterial polyester produced by bacteria. The results were compared with GC analysis to confirm FT-IR results (Table 2). FT-IR spectroscopy has been demonstrated to be a powerful tool to identify and quantify the content of PHAs, but GC analysis were done, after methanolysis, to know more precisely the exact composition of the copolyester. The composition of the copolymer resulting from the fermentation process is varying with the microorganisms. In the case of sodium octanoate as substrate, 3-hydroxyoctanoate is the major repeating unit. The percentage of this 3-hydroxyacid repeating unit changes in a narrow scale. The PHOs were always composed of approximately 4-17 mol-% of 3-hydroxycaproate (HC), 78-92 mol-% of 3-hydroxyoctanoate (HO), and 2-11 mol-% of hydroxydecanoate (HD). On the contrary, P. oleoVorans LMG 2229, P. pseudoalcaligenes LMG 1225 produced PHB polyester when grown on sodium octanoate. Some strains as P. stutzeri and P. syringae seemed unable to synthesize PHA in our conditions. It was admitted that one characteristic of Pseudomonas is their inability to accumulate poly(3-hydroxybutyrate), but we have shown that in our conditions some strains are able to produce PHB when
Biomacromolecules, Vol. 4, No. 4, 2003
Randriamahefa et al.
Table 2. PHA Content Determined by GC and FT-IR
when they are growing on sodium octanoate, the PHOs are not quite different enough.
1096
PHAs content (%) strains
FT-IR
GC
PHAs compositiona (mol-%)
P. chlororaphis DSM 50083 P. putida bv A LMG 2257 P. alcaligenes LMG 6353 P. marginalis pv marginalis LMG 5177 P. taetrolens LMG 2336 P. alcaligenes LMG 6353 P. mendocina LMG 6396 P. fluorescens LMG 5940 P. monteilii CIP 104883 P. putida bv B LMG 1246 P. pseudoalcaligenes LMG 5517 P. denitrificans DSM 1650 P. resinovorans ATCC 14235 P. aeruginosa LMG 5031 P. sp. Gpo1 CIP 105816 P. plecoglossicida CIP 106493 P. cichorii CIP 106704 P. aureofaciens LMG 1245 P. marginalis pv alfafae LMG 2214 P. fluorescens DSM 50415 P. pseudoalcaligenes LMG 1225 P. oleovorans LMG 2229 P. sp. LMG 13970 P. syringae CIP 106698 P. stutzeri CIP 103022 P. viriflava CIP 106699 P. agarici CIP 106703
17.5 17 18.5 2.8
16 18 18 2.5
12HC,81HO,7HD 5HC,92HO,3HD 4HC,90HO,6HD 11HC,82HO,9HD
3 19 7 12.5 4 24 7.3
3.3 22 10 12 4.5 21 8
10HC,82HO,8HD 4HC,88HO,8HD 10HC,82HO,8HD 6HC,90HO,4HD 4HC,88HO,8HD 17HC,81HO,2HD 5HC,92HO,3HD
43 42 3 26 40 30 28 3
40 40 4 24.3 43 33 30 3.5
8HC,90HO,2HD 17HC,78HO,4HD 12HC,80HO,8HD 6HC,80HO,14HD 11HC,85HO,4HD 12HC,85HO,3HD 11HC,84HO,5HD 12HC,83HO,5HD
25 53
27 50
11HC,83HO,6HD 100HB
25 34 0 0 0 0
24.7 32.5 0 0 0 0
100HB 100HB
a Composition (mol-%) determined by GC analysis after methanolysis; HC: 3-methyl hydroxycaproate; HO: 3-methyl hydroxyoctanoate; HD: 3-methyl hydroxydecanoate, HB: 3-methylhydroxybutyrate.
Table 3. Characterization of PHOs Produced by Different Pseudomonas Belonging to RRNA Homology Group I P. aureofaciens P. chlororaphis P. fluorescens P. sp. Gpo1 Mn (g/mol) Ip Tg (°C) Tm (°C) ∆H (J/g)
LMG 1245
DSM 50083
DSM 50415
CIP 105816
97 000 1.5 -40 50 17.2
62 000 2.1 -39 52 18.6
85 000 1.6 -41 40 7.9
80 000 1.5 -39 50.2 17.3
they are grown on sodium octanoate. These results are particularly important and revealed the presence of a different metabolism. We have shown a heterogeneity among the Pseudomonas belonging to the rRNA homology group. The PHA synthesis could be a new criterion to classify these strains, and the results would be very interesting for polymer synthesis. Four strains were selected and PHOs have been produced in large quantities by these strains, to be characterized by SEC and DSC experiments, after solvent extraction and precipitation. Results were presented in Table 3. The PHOs produced are semicrystalline and the degree of crystallinity varies. No simple explanation can be proposed, but a key could be in the distribution of the different repeating units in the macromolecular chain. We can conclude that when strains are able to produce mcl PHO
Conclusion We have developed a FTIR method to determine poly(hydroxyalkanoate) content in intact Pseudomonas grown on sodium octanoate. By comparison of the absorbance of the ester band, characteristic of the polymer, and the absorbance of the amid group of bacteria, a calibration curve was used to determine polymer content. FT-IR does not require extensive sample preparation. This method was validated with 27 Pseudomonas strains by determination of polymer content by the conventional method based on gas chromatography after methanolysis of polymer. The results were in good agreement, and we used FT-IR, which is a more rapid and nondestructive method, for screening a large variety of Pseudomonas belonging to the rRNA homology group I. We can easily distinguish intracellular PHO and PHB by the value of the wavenumber of the C-O-C band (1165 cm-1 for PHO and 1185 cm-1 for PHB). FT-IR is at the same time a qualitative and a quantitative method to determine the type of PHA and polymer content in intact bacteria. The calibration curve is valid between 3% and 53% of polymer content. The upper value (53%) is due to the fact that in our conditions the bacteria were not able to accumulate more than 53% of polymer. Only very low polymer content (
