Possibilities for controlling a PHB accumulation process using various analytical methods

Journal of Biotechnology 117 (2005) 119–129

Possibilities for controlling a PHB accumulation process using various analytical methods K.-D. Wendlandt ∗ , W. Geyer, G. Mirschel, F. Al-Haj Hemidi 1 Departments of Environmental Biotechnology, Analytical Chemistry, UFZ Centre for Environmental Research, Leipzig-Halle, P.O. Box 2, D-04301 Leipzig, Germany Received 9 August 2004; received in revised form 23 December 2004; accepted 7 January 2005

Abstract Poly-␤-hydroxybutyrate (PHB) and other polyesters can be produced by various species of bacteria. Of the possible carbon sources, methane could prove to be one of the most suitable substrates for the manufacture of PHB. The methanotrophic strain Methylocystis sp. GB 25 DSM 7674 was applied in order to accumulate PHB in a rapid, non-sterile process. Cultivation was performed in two stages: a continuous growth phase (dilution rate 0.17 h−1 ) and a PHB accumulation phase under deficiency conditions of an essential nutrient (e.g. phosphorus) in batch culture. The PHB content of the biomass was as high as 51%; efficiency was the highest during the first 5 h of the product formation process. The PHB produced is of very high quality, having a high molecular mass of up to 2.5 × 106 Da. In order to monitor and control the process, a rapid analysis method based upon turbidimetry in the visible range (438 nm) was applied. Moreover, the PHB content of the biomass was determined using an FTIR-spectroscopic method with ATR sampling and multivariate calibration. We achieved a value of 1.4% as the best standard error of cross validation. The nitrogen content of the PHB final product (a product quality parameter) was estimated by spectroscopic method in the visible range. © 2005 Elsevier B.V. All rights reserved. Keywords: Poly-␤-hydroxybutyrate; Methanotrophic bacteria; FT-IR; ATR; Product quality

1. Introduction

Abbreviations: DCU, digital control unit; SDDC, safe direct digital control; FT-IR, Fourier transform infra red (spectroscopy); ATR, attenuated total reflection; PLS, partial least squares (regression) ∗ Corresponding author. Tel.: +49 341 235 2281; fax: +49 341 235 2402. E-mail address: [email protected] (K.-D. Wendlandt). 1 Formerly UFZ. 0168-1656/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jbiotec.2005.01.007

Poly-␤-hydroxybutyrate (PHB) and other polyesters can be produced by various species of bacteria (Lafferty et al., 1988; Anderson and Dawes, 1990; Steinb¨uchel, 1995; Lee, 1966). Due to its biodegradability and other extraordinary properties, PHB could be ideal for use in medicine (e.g. for hard and soft tissue implants, dressing and sewing material,

120

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

tissue growth materials), pharmacy (controlled-release drugs), food packaging and agriculture. Of the possible carbon sources, methane could prove to be one of the most suitable substrates for the production of PHB. Methane is very cheap and plentifully available not only as natural gas but also as biogas. Methanotrophic bacteria can be divided into two types depending on their membrane structure and metabolic route. Type II bacteria, which perform metabolization using the serine route, are more effective at producing PHB than type I bacteria. Methylocystis sp. GB 25 is an example of a type II bacterium that can accumulate PHB intercellularly to a level of up to 51% (by weight). The changing PHB concentration during the accumulation process is an important parameter for assessing the success of a given experiment. The PHB concentration can be determined by means of various different analysis methods such as flow cytometry, gas chromatography, NMR spectroscopy and IR spectroscopy as well as by means of turbidimetry in the visible range after using a hypochlorite solution to destroy all cell components of a sample which do not contain PHB. The determination of the PHB concentration by means of gas chromatography (Riis and Mai, 1988) requires a calibration process to be performed. A more serious disadvantage of this method in the current application is that it necessitates a time-intensive sample preparation involving a hydrochloric acid propanolysis of the dry biomass, thus making the method inappropriate for continuous monitoring of the PHB accumulation process. Using a sodium hypochlorite solution to destroy all cell components of a sample, which do not contain PHB and then performing turbidimetry in the visible range (Williamson and Wilkinson, 1958) allows a rapid (40 min) assessment of the current PHB concentration per unit volume during the accumulation process. It is therefore possible to apply this method for process monitoring and control. Flow cytometry, as applied by various authors (Degelau et al., 1995; M¨uller et al., 1999), is a relative method, which requires the cells to be coloured with Nile red. A direct relationship between the concentration of the coloured cells and the fluorescence only exists up to cell counts of the order of 108 ; a dilution is necessary in the case of higher cell densities. The

measurement value acquired is a relative fluorescence intensity. Because of their preparation complexity, NMRspectroscopic methods are usually only applied after extraction for the characterization of the polymer together with its copolymers and blends (Avella et al., 1998; Schu´e et al., 2000). The IR-spectroscopic method described in this paper can usually be performed quickly and simply. Moreover, in the case of IR absorption, a large amount of information is revealed. For example, changes in the constituents of bacteria cells can be identified from FTIR spectra during the growth of the microorganisms. In addition, multivariate mathematical methods such as PLS (partial least squares) regression enable complex changes to be identified from complete spectra and from spectral regions (Kansiz et al., 2000). This IR-spectroscopic method for determining the PHB concentration of biomass uses attenuated total reflection (ATR) (Harrick, 1960) as its sampling technique. The application of a Golden Gate singlereflection diamond ATR unit allows a rapid determination of the PHB concentration with a low sample volume requirement and with good reproducibility. For calibration purposes, the IR spectra were correlated with the data obtained via gas chromatography. In addition, the purity of the PHB produced (in terms of its nitrogen content) was characterized with a UV/vis-spectroscopic method. The methodology described enables rapid process analysis.

2. Material and methods 2.1. Microorganisms and culture medium The cultivation of the methanotrophic mixed culture used mainly consisted of the strain Methylocystis sp. GB 25 DSM 7674 (≥90% of biomass) as previously described (Wendlandt et al., 1983). The experiments were carried out in a mineral salt medium with the following composition per litre dist. water: 0.028 ml H3 PO4 (80%), 35 mg KH2 PO4 , 25 mg MgSO4 ·7H2 O, 0.785 mg CuSO4 ·5H2 O, 1.389 mg MnSO4 ·H2 O, 1.678 mg FeSO4 ·7H2 O, 0.322 mg ZnCl2 , 0.036 mg Co SO4 ·7H2 O, 0.186 mg Al2 (SO4 )·18H2 O, 0.883 mg Ca(NO3 )2 ·4H2 O, 0.041 mg Na2 MoO4 ·2H2 O, 1.286 mg H3 BO3 , 0.077 mg CrCl3 ·6H2 O, 0.109 mg NiSO4 ·

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

7H2 O. This nutrient solution is calculated for the growth of 1 g/l dry wt. 2.2. Bioreactor The experiments were carried out using a 70 l pressure bioreactor (p ≤ 5 bar, UD 50, Bio Braun Biotech International, Germany) consisting of a culture vessel with a stirrer system and individual supply systems for gases (methane, air, nitrogen) and dosage (nutrient, water, NH4 OH, NaOH, H2 SO4 , antifoam). The measurement and control of the pH, pressure, temperature, weight, dissolved oxygen, agitation speed, and gas flow were performed by a digital measuring system and two control systems (DCU and SDDC) connected to a control computer (IRIS-NT) which was used to log and store process data as well as to control and optimize the fermentation process. The process was then scaled-up to a test scale with a working volume of 400 l using a 600 l pressure bioreactor (p ≤ 5 bar, TF 600, Infors AG, Basle, Switzerland). The dissolved oxygen concentration (pO2 ) in the fermentation medium was kept constant by cascade control (agitator, air quantity and pressure for the UD 50; air quantity and pressure for the TF 600). The input of methane was controlled in both bioreactors via the dissolved methane concentration (pCH4 ), which was determined with a measuring head produced by Biotechnologie Kempe GmbH and kept constant during the process by an IRIS-NT control sequence. In addition, the ammonium–nitrogen and phosphate concentrations in the 600 l fermenter were measured on-line during the process by means of a sensor using a twochannel FIA system (ASIA, Ismatec SA, Glattbrugg, Switzerland). 2.3. Fermentation conditions All fermentation was performed in a non-sterile manner in two stages: (a) a growth phase (temperature 38 ◦ C, pH 5.7, dilution rate 0.17 h−1 , gas flow 50–100 l/min, agitation speed 800–1000 rpm, pressure ≤ 2.5 bar, pO2 = 15% and pCH4 = 15% (UD 50) and temperature 38 ◦ C, pH 5.7, dilution rate 0.17 h−1 , gas flow 46–58 m3 /h, agitation speed 600 rpm, pressure ≤ 4 bar, pO2 = 15% and pCH4 = 30% (TF 600)) with the pH being controlled by NH4 OH; (b) the PHB accumulation phase under phosphorus deficiency. The

121

PHB production process was performed in batch culture. The dissolved oxygen was kept at a constant level of 15% saturation by varying the pressure and agitation speed and adjusting the gas flow. The dissolved methane was also kept at a constant level of 15% (UD 50) and 30% (TF 600) by an IRIS-NT controlling sequence. 2.4. Analyses 2.4.1. Gas analysis The composition of inlet and exhaust gas was determined in two different ways: (a) using a gas analysis system based on paramagnetic gas analysis (Magnos 6G, Hartmann and Braun, Germany) for oxygen, and infrared absorption gas analysis (Uras 10P, Hartmann and Braun, Germany) for methane and carbon dioxide (UD 50); (b) applying a Micro gas chromatograph (M/P200 Agilent Technologies Deutschland GmbH, Germany) for oxygen, methane and carbon dioxide (TF 600). 2.4.2. Biomass concentration 1. Absorbance was measured at 600 nm (Spekol 1100, analytikjena, Germany; blank: aqua dist.); the biomass concentration was evaluated using a calibration curve. 2. Dry weight: 50 ml of cell suspension was centrifuged at 14,000 rpm for 20 min, and the sediment was dried at 105 ◦ C for 24–36 h (depending on the PHB concentration) until the weight remained constant. 2.4.3. Ammonium-nitrogen concentration The analysis was based on the condensation reaction of formaldehyde and ammonium to tetramethylene tetramin. After the neutralisation of 5 ml supernatant and 10 ml distilled water with 0.5N KOH (indicator phenolphthalein), 5 ml formaldehyde was added. The hydrogen ions formed were estimated by titration with 0.5N KOH (blank: 15 ml aqua dist.). 2.4.4. Phosphate-phosphorus concentration The concentration of phosphate-phosphorus was determined by measuring the absorbance of the coloured solution formed by the reaction of phosphate with vanadate–molybdate reagent at 435 nm (Spekol 1100,

122

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

analytikjena, Germany). The phosphate-phosphorus concentration was evaluated using a calibration curve. 2.4.5. Poly-β-hydroxybutyrate extraction After centrifugation, the biomass was freeze-dried. Lipids and coloured substances were then removed by extraction with methanol/water (80%, v/v, 1 h, 50 ◦ C). In the second step, PHB was extracted from the biomass with 1,2-dichloroethane (1 h, 80 ◦ C), the non-PHB cell matter was removed by filtration, and the dissolved PHB was precipitated with methanol (80%, v/v). The PHB was washed twice with methanol, separated by filtration and dried at 60 ◦ C for 2 h. 2.4.6. Poly-β-hydroxybutyrate analysis 2.4.6.1. Hypochlorite-method (as rapid method for determining the PHB concentration). In order to monitor the PHB accumulation by the method of destroying all cell components not containing PHB, 0.2 ml biomass suspension were added to 4.8 ml hypochlorite solution and incubated for 40 min at 38 ◦ C. The PHB concentration was determined by turbidimetry at 436 nm (blank: aqua dist. with hypochlorite solution) (Spekol 1100, analytikjena, Germany) and evaluated using a calibration curve in correlation to gas-chromatographic data. 2.4.6.2. Gas-chromatographic method for determining the PHB content of the biomass. The PHB content was determined by a gas-chromatographic method published by Riis and Mai (1988). About 40 mg dried biomass powder was suspended with 2 ml 1,2-dichloroethane, 2 ml propanol containing hydrochloric acid (1:4, v/v hydrochloric acid (conc.) and propanol) and 200 ␮l standard solution (4%, w/v benzoic acid in propanol), and incubated at 100 ◦ C for 2 h. After cooling to room temperature, 4 ml of distilled water was added and the samples were shaken for 30 s. After 24 h, the organic phase was directly analysed using gas chromatography (HP 6890 system, G1530A chromatograph, G 1512AX GC autosampler, FID detector, capillary column: HP 19091J-413 [HP 5 containing 5% phenylmethylsiloxane], 30 m × 320 ␮m × 0.25 ␮m). Pure poly-3hydroxybutyrate (ICN Biomedicals, Germany) was used as a reference. 2.4.6.3. Determination of the PHB content of the biomass using IR spectroscopy. Biomass containing

PHB was separated from the fermentation liquids by drying at 105 ◦ C. It was then examined by IR spectroscopy over a range of wave number between 650 and 4000 cm−1 (resolution 8 cm−1 , 256 accumulations) using the ATR technique in simple reflection with a Golden Gate ATR unit (Specac). The Golden Gate ATR system contains a 45◦ ATR diamond crystal as the upper plate and ZnSe (zinc selenide) lenses to focus the IR beam. The application of pressure ensures intensive contact with the solid sample. The empty ATR unit (cleaned with methanol) was used as the reference. 2.4.6.4. Data analysis and calibration. To calibrate the IR-spectroscopic determination of the PHB content, dried fermentation samples whose PHB contents had been determined by gas chromatography were used. The spectroscopic data were quantified on the basis of individual bands and larger spectral ranges. The evaluation techniques employed were band integration and multivariate data analysis using PLS (partial least squares) regression. PLS is a powerful tool that is increasingly being used for the quantification of spectroscopic signals. The spectral data (including multivariate data analysis) were processed using the GRAMS/32 software with the PLSplus add-on (Galactic Industries Corporation). The gas-chromatographic PHB analyses were used as reference data. In connection with the PLS method, calibration models were calculated by correlation with the spectroscopic data (Geyer et al., 1998). This model enabled the PHB content of biomass dry substance to be calculated from the IR spectra. 2.4.6.5. Molecular mass. The molecular mass and its distribution were determined with a gel permeation chromatography (GPC) system (KNAUER, Germany) using RI-detector relative to polystyrene standards (solvent 1,2,4-trichlorobenze, concentration 0.22–0.4 mg/ml). 2.4.6.6. Spectroscopic determination of the nitrogen content of PHB products. Before investigating the absorption behaviour of the PHB product in the UV/vis spectral range, the polymer fibres were pressed into foils. A press device (Specac) with heatable and coolable plates was used, along with a tool to produce polymer foils (LOT) and a hydraulic press (Specac).

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

Foils with a thickness of 5–18 ␮m were produced at a temperature of 160 ◦ C and a pressing force of 3 ton. The foils were used to record absorption in the range of 300–700 nm in a special foil-holder in the CARY 3 UV/vis spectrophotometer (VARIAN; spectral width of slit 1 nm, 2 accumulations). The data from the spectra obtained were correlated with the nitrogen values determined in conformance with the DIN 38409 H27 standard by means of a CHNS-1000 Microanalyser (LECO).

3. Results and discussion The ability of the methanotrophic strain Methylocystis sp. GB 25 DSM 7674 to accumulate PHB in bioreactors under non-sterile conditions has already been reported (Wendlandt et al., 1997, 1998, 2001; Helm et al., 2002).

123

PHB formation can be initiated by various deficiency conditions; the interruption of the nutrient supply changes continuous growth (D = 0.17 h−1 ) into a batch process. Phosphate deficiency was chosen as the initiating factor in our experiments. The pressure reactors used allowed a biomass concentration of 25–65 g/l to be achieved. In contrast to PHB synthesis from sugars with accumulation times of 72 h (e.g. Madden et al., 1999), the PHB accumulation with Methylocystis sp. GB 25 takes place in a ‘short time’ process, i.e. the accumulation is almost complete after 24 h. During the course of the PHB formation process, as shown in Fig. 1 for a series of experiments, the PHB content rises to about 40% of the biomass within the first 12 h. This already represents 80–85% of the PHB content after 24 h. The PHB formation and the specific PHB formation rate (related to PHB-free residual biomass) show that the efficiency of the accumulation is the highest during the first 5 h of product formation; an example is given

Fig. 1. Course of the PHB accumulation over 24 h in selected experiments.

124

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

Fig. 2. Specific formation rate qPHB based on the PHB-free residual biomass and PHB concentration during PHB accumulation.

in Fig. 2. The maximum yield is 0.54 g PHB/g CH4 . The PHB yield achieved closely coincides with theoretical yield calculations published by Asenjo and Suk (1986) (0.67 g PHB/g CH4 ) and Yamane (1993) (0.54 g PHB/g CH4 ). The rapid method of using hypochlorite to determine the PHB concentration proved to be suitable for monitoring and controlling the course of the process.

Fig. 3 shows a correlation of the values obtained for five fermentations with concentrations determined by means of gas chromatography. Higher precision can be achieved with FT-IR-spectroscopic methods. The IR spectra of biomass dry substances containing PHB, as obtained using the simple reflection ATR technique, are excellent; they are, in contrast to transmission methods (Kansiz et al., 2000) independent of the

Fig. 3. Correlation of the PHB concentrations obtained via the hypochlorite and GC methods.

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

125

Fig. 4. FT-IR spectra of dried biomass using the Golden Gate single-reflection diamond ATR unit; the absorbance in the frequency range from 1800 to 900 cm−1 and the absorption bands with maxima about 980, 1180, 1280 and 1740 cm−1 reflect the PHB content (see also Table 1).

probe layer thickness and do not show any interference caused by water or any atmospheric constituents. Fig. 4 shows the spectral range affected by the PHB content. As the PHB content rises, the IR spectra contain some areas in which the intensity of the bands increases and others in which their intensity decreases. This indicates that the cell constituents are broken down in favour of PHB. To produce a data set for calibration, the IR spectra of biomass dry substances from two fermentation experiments and the corresponding PHB contents determined with gas chromatography were selected. Calibration using PLS resulted in the best correlation results when the entire spectral range of wave numbers between 1850 and 860 cm−1 was measured using SECV (standard error of cross validation) as a relative value. Table 1 summarizes the results of the PLS regression calculations on the basis of various spectral ranges. This model was used to calculate the PHB content of ‘unknown’ samples from the FT-IR spectra and for comparison with the PHB contents determined by gas chromatography. These samples were:

Table 2 Comparison of PHB content in dried biomass of final samples from various fermentation experiments (numbered) as determined using the GC and FT-IR methods No.

Value GC

Value IR

1 2 3 4 5 6 7 8 9 10

45.2 47.3 45.9 41.5 34.2 51.6 39.5 42.3 42.0 46.5

46.2 49.1 48.9 43.1 34.2 50.2 41.5 43.9 40.9 46.5

• the final samples of six different fermentations (Table 2 shows the results of the values determined with the various methods); • 68 samples taken during the course of 5 fermentation experiments (2 examples are given in Figs. 5 and 6).

Table 1 Result of the PLS correlation using different frequency ranges for FT-IR difference spectra of the dried biomass samples Peak maximum/frequency range (cm−1 )

Factors

R2

S.E.C.V.

S.E.C.V. rel. (%)

980 1180 1280 1740 1800–1480 1480–860 1850–860

8 5 7 7 7 7 8

0.986 0.989 0.995 0.995 0.998 0.998 0.999

1.623 1.459 0.971 0.989 0.499 0.523 0.339

6.7 6.1 4.0 4.1 2.1 2.2 1.4

R2 —squared correlation coefficient; S.E.C.V.—standard error of cross-validation; S.E.C.V. rel.—SECV/average PHB content of the sample set.

126

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

Fig. 5. Content of PHB in dried biomass of fermentation a experiment using the GC and FT-IR methods; regression analysis between the values from GC and FT-IR.

ATR on solid samples is suitable as a rapid method for the IR spectroscopic determination of the PHB content of biomass both during the course of fermentation and in final samples: it provided the best congruence with GC values, and only limited sample preparation is required. The PHB formed by Methylocystis sp. GB 25 and obtained by solvent extraction is a homopolymer characterized by high purity (99.9%) and high molecular masses ≥ 2.5 × 106 Da. Comparison with other microorganisms and other substrates for PHB synthesis indicates that such high (and even higher) molecular weights could previously only be produced by using genetically modified strains (Daniel et al., 1992; Valentin and Dennis, 1997). The high molecular weight is ad-

vantageous because the molar masses may be reduced during thermoplastic processing of the polymers. Another indicator of the degree of purity is the total nitrogen content of the PHB ≤ 0.02%. The PHB’s biocompatibility and these outstanding quality parameters make it suitable for medical purposes. The assumption that the purity of the PHB final product in terms of its nitrogen content is related to its colour was proven using information from the absorption spectra in the visible spectral range of the PHB samples pressed into foils. Using derivative spectroscopy (second derivation), bands with absorption minima of about 405 nm could be distinguished from the extinction curves monotonously rising according to low wavelength. Following

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

127

Fig. 6. Content of PHB in dried biomass of fermentation b experiment using the GC and FT-IR methods; regression analysis between the values from GC and FT-IR.

Fig. 7. Linear regression of the visible absorption of PHB products (second derivation, area of bands with minimum of about 405 nm) as a function of the nitrogen content of the samples.

128

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129

the integration of these bands and the standardization of the data to a uniform foil thickness (10 ␮m), a correlation was observed with the nitrogen content of the PHB products as determined in accordance with DIN 38409 H27. This provides a way of rapidly assessing the quality of PHB products. Because the investigation method used is non-destructive, there is no inherent limit to the number of repeated measurements, which can be made in order to reduce the measurement uncertainty. Fig. 7 shows the measurement error in terms of the calculation of the confidence area for 95% probability together with the correlation coefficient (R2 = 0.872). Thus, the described method provides a good basis for a routine application in quality control of the PHB. 4. Conclusion The methanotrophic strain Methylocystis sp. GB 25 (DSM 7674) is able to accumulate PHB with a high molecular mass in a rapid two stage processes. Various methods for estimation of the PHB content of biomass were discussed. FT-IR-ATR on dried biomasses was established as a method for investigating the accumulation process. The congruence between the PHB levels determined from biomass-dried substances using gas chromatography and this method is very high. Moreover, sample preparation for this method entails less material consumption and is less time-consuming than gas chromatography. The next step should be to define a method, which uses ATR spectra of the aqueous fermentation samples. The UV/vis-spectroscopy provides a way of rapidly assessing the quality of PHB products as regards to the nitrogen content. References Anderson, A.J., Dawes, E.A., 1990. Occurrence, metabolism, metabolic role and industrial uses of bacterial polyhydroxyalkanoates. Microbiol. Rev. 54, 450–472. Asenjo, J.A., Suk, J.S., 1986. Microbial conversation of methane into poly-␤ hydroxybutyrate (PHB): growth and intracellular product accumulation in a type ii methanotroph. J. Ferment. Technol. 64, 271–278. Avella, M., Errico, M.E., Immirzi, B., Malinconico, M., Falgicno, L., Paollilo, L., 1998. Preparation of poly(␤-hydroxybutyrate)/poly(methyl methacrylate) blends by reactive blending and their characterisation. Macromol. Chem. Phys. 199, 1901–1907.

Daniel, M., Choi, J.H., Kim, H.J., Lebeault, J.M., 1992. Effect of nutrient deficiency on accumulation and relative molecular weight of poly-␤-hydroxybutyric acid by methylotrophic bacterium Pseudomonas 135. Appl. Microbiol. Biotechnol. 37, 702–706. Degelau, A., Scheper, T., Bailey, J.E., Guske, C., 1995. Fluorometric measurement of poly-␤-hydroxybutyrate in Alcaligenes eutrophus by flow cytometry and spectrofluorometry. Appl. Microbiol. Biotechnol. 42, 653–657. Geyer, W., Br¨uggemann, L., Hanschmann, G., 1998. Prediction of properties of soil humic substances from FT-IR spectra using partial least squares regression. Int. J. Environ. Anal. Chem. 71, 181–193. Harrick, N.J., 1960. Surface chemistry from spectral analysis of totally internally reflected radiation. J. Phys. Chem. 64, 1110–1114. Helm, J., Wendlandt, K.-D., Jechorek, M., Rogge, G., Kappelmeyer, U., Stottmeister, U., 2002. Natural gas—an attractive carbon source for the production of PHB. In: Proceedings of the International Symposium on Biological Polyesters, M¨unster, Germany, p. 110. Kansiz, M., Billman-Jacobe, H., McNaughton, D., 2000. Quantitative determination of the biodegradable polymer poly(␤hydroxybutyrate) in a recombinant Escherichia coli strain by use of mid-infrared spectroscopy and multivariative statistics. Appl. Environ. Microbiol. 66, 3415–3420. Lafferty, R.M., Korsatko, B., Korsatko, W., 1988. Microbial production of poly-␤-hydroxybutyric acid. In: Rehm, H.-J., Gees, G. (Eds.), Biotechnology, vol. 6b. VCH-Verlagsgesellschaft, Weinheim, pp. 135–176. Lee, S.Y., 1966. Bacterial polyhydroxyalkanoates. Biotechnol. Bioeng. 49, 1–14. Madden, L.A., Anderson, A.J., Shah, D.T., Asrar, J., 1999. Chain termination in polyhydroxyalkanoate synthesis: involvement of exogenous hydroxy-compounds as chain transfer agents. Biol. Macromol. 25, 43–53. M¨uller, S., Bley, T., Babel, W., 1999. Adaptive responses of Ralstonia eutropha to feast and famine conditions analysed by flow cytometry. J. Biotechnol. 75, 81–97. Riis, V., Mai, W., 1988. Gas chromatographic determination of poly␤-hydroxybutyric acid in microbial biomass after hydrochloric acid propanolysis. J. Chrom. 445, 285–289. Schu´e, F., Jaimes, C., Dobreva-Schu´e, R., Giani-Beaune, O., Amass, W., Amass, A., 2000. Synthesis and degradation of polyesters. Polym. Int. 49, 965–974. Steinb¨uchel, A., 1995. Diversity of bacterial polyhydroxyalkanoic acids. FEMS Microbiol. Lett. 128, 219–228. Valentin, E.H., Dennis, D., 1997. Production of poly(3hydroxybutyrate-co-4-hydroxybutyrate) in recombinant Escherichia coli grown on glucose. J. Biotechnol. 58, 33–38. Wendlandt, K.-D., Wandt, A., Br¨uhl, E., Karbaum, K., Schurig, K.H., Wagler, D., 1983. Verfahren zur Kultivierung von Mikroorganismen, 7, DD 283738. Wendlandt, K.D., Jechorek, M., Helm, J., Stottmeister, U., 1997. Production of PHB with a high molecular mass from methane. In: Steinb¨uchel, A. (Ed.), Proceedings of the Conference on Biodegradable Polymers and Macromolecules, Strasbourg, France, pp. 1991–1994.

K.-D. Wendlandt et al. / Journal of Biotechnology 117 (2005) 119–129 Wendlandt, K.-D., Jechorek, M., Helm, J., 1998. Production of PHB with a high molecular mass from methane. Polym. Degrad. Stability 59, 191–194. Wendlandt, K.-D., Jechorek, M., Helm, J., Stottmeister, U., 2001. Production of poly-3-hydroxybutyrate with a high molecular mass from methane. J. Biotechnol. 86, 127–133.

129

Williamson, D.H., Wilkinson, J.F., 1958. The isolation and estimation of the poly-␤-hydroxybutyrate inclusions of Bacillus species. J. Gen. Microbiol. 19, 198–209. Yamane, T., 1993. Yield of poly-d(−)-3-hydroxybutyrate from various sources: a theoretical study. Biotechnol. Bioeng. 41, 165– 170.