The “PHB Depolymerase Inhibitor” of Paucimonas lemoignei Is a PHB

Biomacromolecules 2002, 3, 823-827

823

The “PHB Depolymerase Inhibitor” of Paucimonas lemoignei Is a PHB Depolymerase Simone Reinhardt, Rene´ Handrick, and Dieter Jendrossek* Institut fu¨r Mikrobiologie, Universita¨t Stuttgart, 70550 Stuttgart, Germany Received February 7, 2002; Revised Manuscript Received April 10, 2002

A ≈35 kDa protein that has been described to be secreted by Paucimonas lemoignei during growth on succinate and to inhibit hydrolysis of denatured (crystalline) poly(3-hydroxybutyrate) (dPHB) by extracellular PHB depolymerases of P. lemoignei (PHB depolymerase inhibitor (PDI)) was purified and characterized. Purified PDI (Mr, 36 199 ( 45 Da) inhibited hydrolysis of dPHB by two selected purified PHB depolymerases (PhaZ2 and PhaZ5) but did not inhibit the hydrolysis of water-soluble substrates such as p-nitrophenylbutyrate by PhaZ5 and PhaZ2. PDI revealed a high binding affinity to dPHB although it was not able to hydrolyze the crystalline polymer. However, purified PDI had a high hydrolytic activity if native (amorphous) PHB (nPHB) was used as a substrate. N-terminal sequencing of PDI revealed that it was identical to recently described extracellular PHB depolymerase PhaZ7 which is specific for nPHB and which cannot hydrolyze dPHB. To confirm that the inhibition of hydrolysis of dPHB by PhaZ7 is an indirect surface competition effect at high depolymerase concentration, the activity of PHB depolymerases PhaZ2 and PhaZ5 in the presence of different amounts of protein mixtures was determined. The components of NB or LB medium inhibited hydrolysis of the polymer in a concentration-dependent manner but had no effect on the hydrolysis of p-nitrophenylbutyrate by PHB depolymerases. In combination with PHB depolymerases PhaZ2 and PhaZ5 the protein PhaZ7 (“PDI”) enables the bacteria to hydrolyze dPHB and nPHB simultaneously. Introduction The ability to degrade extracellular poly(3-hydroxybutyrate) (PHB) and related polyhydroxyalkanoates (PHA) is widely distributed among bacteria and depends on the secretion of specific polyester depolymerases. PHB depolymerases are carboxyesterases (EC 3.1.1.75, 3.1.1.76) and hydrolyze the water-insoluble polymer to water-soluble monomers or oligomers. Paucimonas (formerly Pseudomonas) lemoignei1 is unique among PHA-degrading bacteria because it is able to synthesize at least seven different extracellular PHA depolymerases (PhaZ1-PhaZ7).2,3 Other PHB-degrading bacteria such as Alcaligenes faecalis T14 synthesize only one PHB depolymerase. Six of the 7 known PHB depolymerases of P. lemoignei (PhaZ1 to PhaZ6) are able to hydrolyze denatured PHA (dPHA) such as dPHB, denatured poly(3-hydroxyvalerate), dPHV, or denatured copolymers of 3-hydroxybutyrate and 3-hydroxyvalerate. Only the recently described extracellular depolymerase PhaZ7 is specific for the native (amorphous) form of PHB (nPHB) and is not able to hydrolyze crystalline dPHB.5 In contrast to most PHB-degrading bacteria, which repress PHB depolymerase synthesis in the presence of soluble carbon sources such as glucose or fatty acids, P. lemoignei synthesizes high PHB depolymerases activity during growth on succinate. The relationship between expression of PHB depolymerase activity and growth on succinate has been studied earlier.6 However, knowledge of the regulation of * To whom correspondence may be addressed. Tel.: +49-711-685-5483. Fax: +49-711-685-5725. E-mail: [email protected].

PHB depolymerase synthesis on the molecular level is still low. A few years ago Doi and co-workers described a 35 kDa protein that was secreted from P. lemoignei during growth on succinate and that inhibited the activity of dPHB depolymerases in vitro (PHB depolymerase inhibitor, PDI).7 Since production of dPHB depolymerase activity during growth on succinate in the absence of the substrate dPHB is useless for the cells and since PDI apparently was not synthesized during growth on dPHB, the authors assumed that the physiological function of PDI might be the inhibition of the dPHB depolymerase active site as long as the substrate is not present.7 In this contribution we purified PDI and two dPHB depolymerases and analyzed the function of PDI by in vitro experiments. Materials and Methods Bacteria, Media, and Growth Conditions. Paucimonas [formerly Pseudomonas] lemoignei (DSMZ7445)1,8 was grown at 30 °C in Stinson and Merrick’s mineral salts medium9 with 20-50 mM sodium succinate or carbon sources as indicated. Other media were nutrient broth (NB; 8 g/L), Luria-Bertani medium (LB; 5 g/L yeast estract, 10 g/L tryptone, 10 g/L sodium chloride), and Super Rich medium (SR, 25 g/L Bacto tryptone, 20 g/L yeast extract, 3 g/L K2HPO4). dPHB Granules. Semicrystalline dPHB was isolated from gluconate-grown cells of Ralstonia eutropha H16 (DSMZ428) by sodium hypochlorite digestion and subsequent solvent extraction with acetone/diethyl ether as described elsewhere.10

10.1021/bm025519x CCC: $22.00 © 2002 American Chemical Society Published on Web 05/10/2002

824

Biomacromolecules, Vol. 3, No. 4, 2002

nPHB Granules. nPHB granules with intact surface layer were prepared from crude extracts (French press) of accumulated cells by two subsequent glycerol density gradient centrifugations as described recently.11 Purification of PDI, PHB Depolymerase A (PhaZ5), and PHB Depolymerase B (PhaZ2). Cell-free culture fluid of a 10 L culture on succinate was concentrated by ultrafiltration and ammonium sulfate precipitation (30-85%). The dissolved (20 mM Tris-HCl, pH 8) proteins were diafiltrated against 10 mM succinate-NaOH, pH 4.7 containing 1 mM CaCl2 and 5% (v/v) glycerol and were applied to a CM-Sepharose CL-6B column (diameter, 26 mm; bedvolume, 100 mL). After the column was washed with dialysis buffer, proteins were eluted by a linear NaCl gradient (0-200 mM in equilibration buffer, 750 mL). Supernatant of succinate-grown cells contained high levels of PHB depolymerase A (PhaZ5), PHB depolymerase B (PhaZ2), and PDI which eluted at 15-20, 50-55, and 55-65 mM NaCl, respectively. PHB depolymerase A (PhaZ5) turned out to be homogeneous and was stored at - 20 °C. Peaks of PHB depolymerase B (PhaZ2) and of PDI partially overlapped. Separation of PHB depolymerase B from PDI was performed by chromatofocusing on a Mono-P column (diameter, 5 mm; bedvolume, 4 mL; flow rate, 0.5 mL/min). Proteins were eluted with 1:60 diluted Pharmacia Pharmalyte polybuffer, pH range 8-10, pH (HCl) 8.0. PDI appeared between pH 9.3 and 8.8 with a maximum at pH 9.2. PhaZ2 was completely separated from PDI and appeared between pH 9.6 and 9.4. Fractions containing high amounts of PDI or PhaZ2 were pooled separately and stored on ice or frozen at -20 °C. Protein concentration was determined according to the Bradford method.12 Enzyme Assays. Activity of p-nitrophenylbutyrate esterase activity was performed photospectroscopically (405 nm) in 1 cm cuvettes (1 mL) containing 100 mM Tris-HCl, pH 7.9, 1 mM CaCl2, and 10 µL of a 30 mM solution of pnitrophenylbutyrate in ethanol at 37 °C. Activity of dPHB depolymerases was assayed photospectroscopically at 650 nm and at 37 °C. The assay mixture contained 100 mM Tris-HCl, pH 8.0, 1 mM CaCl2, and 180500 µg/mL of sodium hypochlorite-purified dPHB granules. One unit of activity is defined as the hydrolysis of 1 µg of PHB in 1 min. For assay of nPHB depolymerase PhaZ7 activity the assay mixture contained 100 mM Tris-HCl, pH 9.0, 1 mM CaCl2, and 500 µg/mL of nPHB granules purified from R. eutropha as described earlier.5 PHB-Binding Assay. One hundred micrograms of purified PDI was added to 300 µL of a suspension of 15 mg of sodium hypochlorite and aceton/ether extracted dPHB granules and mixed by vortexing. After 5 min of incubation at 8 °C the mixture was centrifuged (10 000 rpm) and the amount of PDI in the supernatant and in the pellet was determined. PDI bound to the pellet could be liberated by vortexing the pellet with 300 µL of 50% 2-propanol and incubation for 5 min. After centrifugation the supernatant was evaporated, and liberated PDI was rehydrated in water or buffer. The same experiment without dPHB granules served as a control.

Reinhardt et al.

Figure 1. Reducing SDS-PAGE analysis of PDI, PhaZ5, and PhaZ2 at various stages of purification. The proteins were separated by SDS-PAGE and stained with silver. Molecular mass standards (lane 1), concentrated cell-free culture fluid (lane 2), PDI after ammonium sulfate precipitation and dialysis (lane 3), PDI after CM-Sepharose (lane 4), purified PhaZ5 after Mono P (lane 5), purified PhaZ2 after Mono P (lane 6), purified PDI after Mono P (lane 7). Arrows indicate positions of the respective proteins.

Results Purification of PHB Depolymerase Inhibitor (PDI). P. lemoignei was grown in mineral salts solution with succinate as a carbon source. PHB depolymerase A (PhaZ5), PHB depolymerase B (PhaZ2), and a 36 kDa protein (putative PHB depolymerase inhibitor [PDI]) were purified from the cell-free culture fluid by two chromatography steps as described in materials methods. Fractions containing PHB depolymerase activity and/or PDI were analyzed by SDSPAGE and stained with silver before fractions were pooled, respectively, to ensure the highest possible purification. All three proteins were homogeneous and did not contain any contamination of one of the other proteins (Figure 1). The faint band below purified PhaZ5 at ≈43 kDa is a thermal degradation product of PhaZ5 appearing during heat denaturation as shown by immunological reaction with antibodies against PhaZ5 (data not shown). Test for PHB depolymerase activity revealed that two of the purified proteins, namely, PHB depolymerase A (PhaZ5) and PHB depolymerase B (PhaZ2) had high activity with dPHB granules (1.6 × 104 and 1.5 × 104 U/mg, respectively). N-terminal amino acid sequencing confirmed that both proteins with depolymerase activity were identical to previously sequenced PhaZ5 and PhaZ2.13,14 The third protein (putative PDI) was inactive with denatured PHB regardless whether the turbidometric assay or a titristat method was used. An apparent molecular weight of 36 ( 3 kDa and of 36 199 ( 45 Da was calculated by reducing SDS-PAGE and matrix-assisted laser desorption ionization time-of-flight (MALDI-ToF) mass spectrometry analysis, respectively, and was almost the same as described for PDI (35 kDa).7 Inspection of the protein pattern of cellfree culture fluid revealed that the 36 kDa protein represents the strongest protein band in the area of 35-36 kDa. We assume that the purified 36 kDa protein is identical to PDI. Inhibition of PHB Depolymerase Activity of PhaZ5 and PhaZ2 by PDI. The PHB depolymerase activities of purified PhaZ5 and purified PhaZ2 in the absence or presence of different amounts of PDI were determined. As shown in plots

PHB Depolymerase Inhibitor

Biomacromolecules, Vol. 3, No. 4, 2002 825

Figure 2. Inhibition of PHB depolymerase PhaZ2 (A and C) and PHB depolymerase PhaZ5 (B and D) by PDI (A and B) and by NB, LB, and SR medium (C and D). Experiments without depolymerase (black circles in A and B) or without inhibitor (diamonds in C and D) served as controls. The activity of PHB depolymerase was followed photospectroscopically at 650 nm as described in material and methods.

A and B of Figure 2, PDI strongly inhibited hydrolysis of dPHB granules in a concentration-dependent manner. The concentration of PDI necessary for a 50% reduction of depolymerase activity amounted 1.5 and 3.7 µg PDI/ml for PhaZ5 and PhaZ2, respectively. The inhibition of PHB depolymerases by PDI could be based on two reasons: (i) PDI might be a specific inhibitor of the depolymerase by a yet unknown interaction of PDI with the active site or with the PHB-binding domain of the depolymerase as suggested by Mukai et al.;7 (ii) alternatively, PDI might have a high affinity to the surface of the substrate (dPHB) and thus could inhibit dPHB depolymerase indirectly by prevention of the dPHB depolymerase to bind to the surface of the polymer. To investigate which of the two assumptions could be true, binding of PDI to dPHB granules was investigated as described in materials and methods. Almost all PDI (g 98%) was removed from the supernatant by the addition of solid dPHB granules and was detected in the pellet fraction (Figure 3). PDI could be solubilized from the dPHB pellet by addition of 50 vol % 2-propanol or by cooking in SDS denaturing solution. Inhibition of PHB Depolymerase PhaZ5 and of PhaZ2 by Nutrient Broth, Luria-Bertani Broth, and Super Rich Media and by Bovine Serum Albumin. We conclude from the experiments described above that PDI has a strong affinity to the surface of dPHB granules. If this is true, other proteins also might be able to bind to dPHB granules resulting in reduction of PHB depolymerase activity. The effect of different amounts of oligopeptide mixtures on dPHB hydrolysis by PHB depolymerases PhaZ5 and PhaZ2 was analyzed. Nutrient broth (NB), Luria-Bertani broth (LB), and Super Rich medium (SR) were added to the PHB depolymerase assay mixture 5 min before the reaction was started by the addition of the respective depolymerase. As shown in plots C and D of Figure 2, all three media inhibited

Figure 3. Binding of PDI to dPHB granules. Purified PDI was added to a suspension of dPHB granules and mixed by vortexing. After centrifugation the amount of PDI in the supernatant and in the pellet was determined. PDI bound to the pellet was vortexed with 50% 2-propanol. After centrifugation the supernatant was evaporated, and liberated PDI was rehydrated in water. Purified PDI (lane 1); protein bound to dPHB, liberated by 50% 2-propanol and resuspended in buffer (lane 2); supernatant of PDI after addition of dPHB and centrifugation (lane 3); molecular weight standard (lane 4).

the reaction in a concentration-dependent manner with the highest and lowest inhibition for SR and NB media, respectively. This corresponded with the absolute amount of oligopeptides present in NB, LB, and SR medium (38, 45, and 170 µg/mL, respectively). In control experiments, in which only buffer instead of medium had been added, no reduction of depolymerase activity was found. We conclude that complex media such as NB, LB, and SR contain compounds that interfere with the depolymerase reaction. Similar results were obtained with bovine serum albumin (BSA, data not shown). We assume that the inhibition is caused by an unspecific reaction of medium compounds such as oligopeptides which are able to bind to the surface of

826

Biomacromolecules, Vol. 3, No. 4, 2002

Reinhardt et al.

Figure 4. Growth (black circles) and production of PDI (columns) during growth of P. lemoignei on acetate (A), valerate (B), 3-hydroxybutyrate (C), and dPHV (D). Growth of the bacteria was measured turbidometrically in a Klett-photometer, PDI was assayed by its activity with nPHB granules and by SDS-PAGE.

dPHB granules and prevent dPHB depolymerases from binding to the substrate. Inhibition of the p-Nitrophenylbutyrate Esterase Activity of PhaZ5 and PhaZ2 by PDI, NB, LB, and SR Media. The results shown above suggested that inhibition of PHB depolymerase activity by PDI probably is caused by a binding of PDI to the PHB granules rather than by an interaction between the depolymerase and PDI. If this assumption is true, the presence of PDI should not have a significant effect on the hydrolysis of soluble esters such as p-nitrophenylbutyrate. All dPHB depolymerases analyzed so far have low but significant activity with soluble substrates such as p-nitrophenylbutyrate. The inhibition of p-nitrophenylbutyrate esterase activity of PhaZ5 and PhaZ2 by PDI and by components of complex media (NB, LB, and SR medium) was determined. Addition of either 6.3 µg/mL of purified PDI or of 10 vol % NB or LB medium did not change the esterase activity of PhaZ5 or PhaZ2, respectively (data not shown). We conclude that neither PDI nor components of NB or LB could inhibit the active sites of PhaZ5 or PhaZ2. SR medium itself enhanced spontaneous hydrolysis of p-nitrophenylbutyrate in the absence of PHB depolymerase and could not be used. Identification of the True Function of the 35 kDa Protein (PDI). From the results described above it is unlikely that the physiological function of PDI is the inhibition of PHB depolymerase activity during growth on succinate, i.e., in the absence of dPHB. In that case the depolymerase and the esterase activity should be inhibited by PDI. However, that was not the case. P. lemoignei produces high amounts of PHB depolymerase isoenzymes during growth on succinate in batch culture.6 Recently, we had isolated a new PHB depolymerase, PhaZ7, from succinate-grown culture fluid.5 It came out that the purified new depolymerase had an

apparent molecular mass of 36 kDa. The depolymerase PhaZ7 was inactive with dPHB but was highly active with nPHB granules. Could it be that PDI is identical to PhaZ7? We tested purified PDI for activity with nPHB granules and found a high specific activity of 20.5 × 104 U/mg, similar to that of previously purified PhaZ7.5 Moreover, purified PDI had almost no activity with p-nitrophenybutyrate and low activity with p-nitrophenyloctanoate as had been determined for PhaZ7. The N-terminal amino acid sequence of purified PDI was determined by Edman degradation (NH2-LTXGTNSGFV) and was identical to that of PhaZ7.5 The expression of PDI during growth of P. lemoignei on other substrates such as acetate valerate, 3-hydroxybutyrate, and dPHA was determined. PDI was found at the end of the exponential growth phase on all four substrates (Figure 4). However, in the case of dPHA as a substrate, PDI was bound to the polymer granules and could be detected in the cell-free culture fluid only after almost complete hydrolysis of the polymer. In all cases when PDI was found, PHB depolymerase PhaZ7 was also detected. From all these results it is obvious that PhaZ7 and PDI are identical. Discussion The expression of PHB depolymerase in PHB-degrading bacteria is regulated. As far as PHB-degrading bacteria have been analyzed, the majority of them repress the synthesis of PHB depolymerase activity in the presence of soluble carbon sources such as glucose, fatty acids, high concentrations of 3-hydroxybutyrate, or other substrates that support growth of the bacteria. Only during growth on PHB or during starvation for carbon is the synthesis of most extracellular dPHB depolymerase derepressed.1,15 P. lemoignei seemed to be an exception because it produces the highest amounts of

PHB Depolymerase Inhibitor

dPHB depolymerases during growth on succinate in batch culture.8,9 It turned out that expression of dPHB depolymerase during growth on succinate was caused by starvation for carbon due to an unusual succinate uptake carrier that transports succinate efficiently only at pH below 6.5 if the substrate is present in its protonated (H-succinate-) form. Since the pH in succinate batch culture of P. lemoignei increases contineously, the bacteriasafter a short time of unrestricted growth at pH above 6.5 at the beginningscannot take up succinate anymore and starve for carbon. This is a signal to derepress PHB depolymerase synthesis similar to that seen in other PHB-degrading bacteria (for details see refs 6 and 15). From the view of P. lemoignei cells it makes sense to secrete active PHB depolymerases in succinate batch culture: PHB depolymerases could improve the supply with carbon from polymers such as PHB as the succinate carrier fails to take up succinate when the pH increases and the bacteria begin to starve for carbon. It is unlikely that a bacterium is able to measure the absence or presence of an utilizable extracellular insoluble polymeric carbon source such as PHB directly unless it has tried to degrade it to soluble products, e.g., by PHB depolymerase. From these arguments the secretion of a protein (PDI) with the function of a specific inhibition of PHB depolymerase during growth on succinate has no adventage for the bacteria. Moreover, “PDI” was also found during growth of P. lemoignei on dPHB and dPHV. This is in contrast to results reported by Mukai et al.7 who found no “PDI” during growth on dPHB. We assume that “PDI” was also synthesized in these experiments but could not be detected in cell-free culture fluid because it was still bound to residual dPHB. The aim of this study was to determine the true function of “PDI”. The “PDI” protein was purified from succinategrown culture fluid. The apparent molecular weight (36 kDa) was very close to the value that had been reported for PDI (35 kDa).7 The inhibitory effect of “PDI” on dPHB depolymerase activity could be reproduced for PhaZ5 and PhaZ2 in this study. The apparent ki values of purified “PDI” for PhaZ5 and PhaZ2 depolymerase inhibition amounted to 1.5 and 3.7 µg/mL and were 2.2- and 3.8-fold lower than had been determined by Mukai et al.7 The differences probably are due to the different protein assay methods (Bradford versus Lowry method). Protein determination of purified PHB depolymerases usually results in 2-3-fold lower values for the Bradford method (unpublished observation). The inhibitory effect of “PDI” on PHB depolymerase activity could be replaced by complex media such as NB, LB, and SR medium or even by BSA. This suggested that the inhibition of PHB depolymerase might not be very specific. To find out whether the inhibition was caused by an inhibition of the depolymerase protein or by covering of the insoluble substrate (PHB), the effect of “PDI” and of medium components on the soluble esterase activity of PHB depolymerases with soluble p-nitrophenylbutyrate was tested. Since

Biomacromolecules, Vol. 3, No. 4, 2002 827

the esterase activities of dPHB depolymerases PhaZ2 and PhaZ5 were not significantly effected at all by PDI or by medium components it is only the polymeric substrate dPHB that is prevented from hydrolysis by binding of “PDI” or medium components to the surface of dPHB. This assumption could be confirmed by finding a more likely physiological function of “PDI”, namely, the hydrolysis of nPHB granules. Is is not surprising that a protein that hydrolyses nPHB granules is also able to bind efficiently to dPHB granules although it cannot hydrolyze dPHB. It is known that nPHB granules, which can be released by dying bacteria to the environment, sooner or later lose the crystallizationprotecting surface layer and become (partially) crystalline (dPHB). Therefore, nPHB and dPHB can be present in one ecosystem simultaneously. A bacterium that has the capacity to hydrolyze both forms of the polymer should have an advantage over other bacteria. It is unlikely that the concentration of PhaZ7 in the environment can become high enough to cause a significant inhibition of dPHB degradation. All results described above can be easily explained if “PDI” and PhaZ7 are identical. The identity of both proteins is supported by (i) almost identical Mr (36 versus 35 kDa), (ii) comparable degree of inhibition of hydrolysis of dPHB by PHB depolymerases PhaZ5 and PhaZ2, (iii) identical order of elution of PhaZ5, PhaZ2 and “PDI”/PhaZ7 during cation exchange chromatography, and (iv) identity of the N-terminal amino acid sequence of purified “PDI” with PhaZ7. Acknowledgment. This work was supported by a grant of the Deutsche Forschungsgemeinschaft. References and Notes (1) Jendrossek, D. Int. J. Syst. EVol. Microbiol. 2001, 51, 905-908 (2) Jendrossek, D. In Biopolymers, Part 3b, Polyesters; Steinbu¨chel, A., Doi, Y., Eds.; Wiley-VCH: Weinheim, 2001; pp 41-83,. (3) Jendrossek, D. AdV. Biochem. Eng. Biotechnol. 2001, 71, 293-325. (4) Tanio, T., Fukui, T., Shirakura, Y., Saito, T., Tomita, K., Kaiho, T. Eur. J. Biochem. 1982, 124, 71-77 (5) Handrick, R., Reinhardt, S., Focarete, M. L., Scandola, M., Adamus, G., Kowalczuk, M., Jendrossek, D. J. Biol. Chem. 2001, 276, 3621536224 (6) Terpe, K., Kerkhoff, K., Pluta, E. Jendrossek, D. Appl. EnVironm. Microbiol. 1999, 65, 1703-1709. (7) Mukai, K., Yamada, K., Doi, Y. Int. J. Biol. Macromol. 1992, 14, 235-239. (8) Delafield, F. P., Doudoroff, M., Palleroni, N. J., Lusty, C. J., Contopoulos, R. J. Bacteriol. 1965, 90, 1455-1466 (9) Stinson, M. W., Merrick, J. M. J. Bacteriol. 1974, 119, 152-161 (10) Jendrossek, D., Mu¨ller, B., Schlegel, H. G. Eur. J. Biochem. 1993, 218, 701-710. (11) Handrick, R., Reinhard, S., Jendrossek, D. J. Bacteriol. 2000, 182, 5916-5918 (12) Bradford, M. Anal. Biochem. 1976, 72, 248-254 (13) Briese, B. H., Schmidt, B., Jendrossek, D. J. EnViron. Polym. Degrad. 1994, 2, 75-87. (14) Jendrossek, D., Frisse, A., Behrendes, A., Andermann, M., Kratzin, H. D., Stanislawski, T., Schlegel, H. G. J. Bacteriol. 1995, 177, 596607 (15) Jendrossek, D., Knoke, I., Habibian, R. B., Steinbu¨chel, A., Schlegel, H. G. J. EnViron. Polym. Degrad. 1993, 1, 53-63.

BM025519X