Molecular & Biochemical Parasitology 221 (2018) 32–35 Contents lists available at ScienceDirect Molecular & Biochemical Parasitology journal homepage: www.elsevier.com/locate/molbiopara Heterologous expression of three antigenic proteins from Angiostrongylus cantonensis: ES-7, Lec-5, and 14-3-3 in mammalian cells T Bianca B. Cognatoa, Sukwan Handalib, Alessandra L. Morassuttia, , Alexandre Januário da Silvac, Carlos Graeff-Teixeiraa ⁎ a Laboratorio de Parasitologia Molecular, Instituto de Pesquisas Biomédicas and Laboratorio de Biologia Parasitaria, Faculdade de Biociências da Pontifícia Universidade Catolica do Rio Grande do Sul (PUCRS), Av Ipiranga 6690, 90690-900, Porto Alegre, RS, Brazil b Division of Parasitic Diseases and Malaria, Centers for Disease Control and Prevention, 1600 Clifton Road, MS D-64, Bldg 23, Room 9-440, Atlanta, GA, 30329, USA c U.S. Food and Drug Administration, Center for Foods Safety and Applied Nutrition, Office of Applied Research and Safety Assessment, Division of Food and Environmental Microbiology, USA A R T I C LE I N FO A B S T R A C T Keywords: Angiostrongylus cantonensis Recombinant proteins Chinese hamster ovary cells Human embryonic kidney cells Glycoprotein Angiostrongylus cantonensis is a parasitic nematode and the main causative agent of human cerebral eosinophilic meningoencephalitis (EoM). A definitive diagnosis of EoM usually requires serologic or molecular analysis of the patient’s clinical sample. Currently, a 31 kDa antigen is used in immunological tests for this purpose, however as a crude antigen preparation it may present cross-reactivity with other helminthic infections, especially echinococcosis. Heterologous expression studies using prokaryotic systems failed on producing antigenic proteins. The aim of this study was to express and purify three recombinant glycoproteins representing A. cantonensis antigens: ES-7, Lec-5, and 14-3-3, in Chinese hamster ovary (CHO) cells and ES-7 in human embryonic kidney (HEK) cells to develop a source of specific antigens to be used in the diagnosis of angiostrongyliasis. The potential diagnostic value of these three proteins was subsequently characterized in one- and two-dimensional electrophoresis and Western blot to dot blot analyses, with Angiostrongylus-positive sera, normal human sera (NHS), and a pool of Echinococcus-positive sera (included as a specificity control) used for detection. In addition, recognition of these three proteins following treatment with N-glycosidase F was examined. The ES-7 proteins that were expressed in HEK and CHO cells, and the Lec-5 protein that was expressed in CHO cells, were specifically recognized by A. cantonensis-positive sera in the 2D electrophoresis analysis. This recognition was shown to be dependent on the presence of glycidic portions, making mammalian cells a very promising source of heterologous expression antigenic proteins from Angiostrongylus. 1. Introduction Angiostrongylus cantonensis is the most common causative agent of EoM. Humans are an incidental host and they become infected following the ingestion of raw or undercooked snails or slugs. Paratenic hosts, such as prawns or contaminated vegetables, serve as hosts for infective larvae (L3). There have been many cases reported worldwide, especially in Asia and the Pacific Islands [1], but it has expanded worldwide [2–5]. A parasitological diagnosis of A. cantonensis is challenging since larvae are rarely detected in cerebrospinal fluid. Consequently, indirect methods are used for diagnosis. An antigen that has been used as a reliable immunodiagnostic for angiostrongyliasis corresponds to a 31 kDa protein band from whole crude antigen from female worms [6]. Previous work have identified and characterized several protein components present in this 31 kDa band using two-dimensional (2D) electrophoresis and demonstrated the that glycidic moieties in such proteins were immunogenic [7]. With the purpose of developing a more accurate source of antigen for diagnosis of human angiostrongyliasis, Morassutti and collaborators [8] expressed in prokaryotic cells the 14-3-3 protein, one of the components identified in the 31 kDa band, as well as the excretion and secretion products, Lec-5 and ES-7, also found as promising diagnostic targets in another study [9]. All three recombinant proteins were indiscriminately recognized by normal human sera (NHS). Based on these results, it was hypothesized that the incorporation of carbohydrates during protein expression may be crucial for protein antigenicity. It was also recognized that protein folding in a prokaryotic expression system ⁎ Corresponding author at: Laboratorio de Biologia Parasitaria da Pontifícia Universidade Catolica do Rio Grande do Sul, Avenida Ipiranga 6690, Predio 12, Bloc°C, Sala 241, 90690 900, Porto Alegre, RS, Brazil. E-mail address: almorassutti@gmail.com (A.L. Morassutti). https://doi.org/10.1016/j.molbiopara.2018.03.001 Received 14 December 2017; Received in revised form 13 March 2018; Accepted 15 March 2018 Available online 16 March 2018 0166-6851/ © 2018 Elsevier B.V. All rights reserved. Molecular & Biochemical Parasitology 221 (2018) 32–35 B.B. Cognato et al. 2.3. SDS-PAGE electrophoresis (1D electrophoresis) may differ from the protein folding process in a eukaryotic organism, since important organelles for protein synthesis are lacking in prokaryotes [10]. The aim of this study was to express three diagnostic target proteins from A. cantonensis, ES-7, Lec-5, and 14-3-3, in two widely used mammalian expression systems, Chinese hamster ovary (CHO) cells and human embryonic kidney (HEK) cells [11]. A preliminary evaluation for sensitivity and specificity of these antigens was done using sera samples from individuals diagnosed with angiostrongyliasis, Echinococcosis and NHS. All recombinant proteins were subjected to 1D electrophoresis with Mini-PROTEAN® TGX™ Precast Gels (15%, Bio-Rad Laboratories). Sample buffer [10% SDS, 6% glycerol, and tracking dye (50 mg bromophenol blue, 8 ml glycerol, 1 ml 0.5 M Tris HCL (pH 8.0), 1 ml distilled water)] with or without 100 mM dithiothreitol (DTT). Protein separation was achieved at a voltage of 75 V for 25 min with the BioRad MiniProtean electrophoresis system. The amount of protein separated with 1D electrophoresis included: 0.12 ug of ES-7 produced in HEK cells, 0.15 ug of ES-7 produced in CHO cells, 0.07 ug of Lec-5, and 0.1 ug of 14-3-3. Detection of IgG4 was performed as a loading control. 2. Materials and methods 2.4. Two-dimensional (2D) electrophoresis 2.1. Recombinant proteins Each recombinant protein was desalted with a 2-D Clean-Up Kit (GE Healthcare, Piscataway, NJ, USA) and then subjected to solubilization with DeStreak Rehydration Solution (GE Healthcare) containing 66 mM DTT (Sigma-Aldrich, Canada) and 0.5% carrier ampholytes. The samples were in-gel rehydrated on 11 cm pH 3–11 NL IPG strips (GE Healthcare) overnight and then isofocalized with the following stepwise increases in voltage: 500 V for 500 V h, a linear gradient from 500 to 8000 V for 6500 V h, followed by a hold at 6000 V for 22,000 V h. After that, the strips were soaked for 15 min in fresh equilibration buffer (20% v/v glycerol, 6 M urea, 1% DTT, and 2% SDS). IPG strips were run in the second dimension on Criterion XT Precast Gels 4–12% Bis-Tris IPG (Bio-Rad Laboratories) and then were stained with colloidal Coomassie blue or transferred to nitrocellulose membranes for the immunological assay. The amount of protein separated with 2D electrophoresis included: 20 ug of ES-7 produced in HEK cells, 0.5 ug of ES-7 produced in CHO cells, 14 ug of Lec-5, and 1.1 ug of 14-3-3. Detection of IgG4 was performed as a loading control. Cloning of recombinant proteins was performed by GenScript (Piscataway, NJ, USA). All three proteins were cloned into a pUC57 plasmid as fusion proteins with glutathione S-transferase (GST). Expression and purification of the recombinant proteins were performed according to the following protocol. Briefly, 100 ml suspensions of CHO cells (approximately 2 × 107 cells/ml) in serum-free FreeStyle™ TM CHO Expression Medium (Life Technologies, Carlsbad, CA, USA) and HEK 293-6E cells (approximately 2 × 107 cells/ml) in serum-free FreeStyle™ 293 Expression Medium (Invitrogen) were grown in Erlenmeyer flasks on an orbital shaker at 37 °C under 5% CO2. Twentyfour hours later, DNA and polyethylenimine (PEI) (Polysciences, Eppelheim, Germany) were mixed at an optimal ratio of 5:1 (PEI:DNA) and then were added into each flask. Recombinant plasmids encoding ES-7, 14-3-3, and Lec-5 were each transfected into the CHO cell cultures. In parallel, the recombinant plasmid encoding ES-7 was transfected into a 100 ml suspension of HEK 293-6E cells, since ES-7 was previously shown to have great potential as a diagnostic target (Morassutti et al. [7]). After 6 days, all of the cell culture supernatants were centrifuged, filtered, and added to Glutathione Sepharose 4 HP resin columns (5 ml for CHO supernatants and 3 ml for HEK-6E supernatants, GE Healthcare, Uppsala, Sweden, Cat. no. 17-5279-01) at 1 ml/ min. After approximately 4 h, the resins were washed and then eluted with an appropriate buffer. The purified proteins were then analyzed by SDS-PAGE and Western blotting according to standard protocols and molecular weights, yields, and purities, were recorded. Protein quantification was obtained by Thermo Scientific™ NanoDrop™ UV–vis spectrophotometers. To confirm the recombinant expression, a conjugated anti-GST-horseradish peroxidase primary antibody was used for Western blot analyses (GenScript, Cat. no. A00866). Moreover, the identities of all three proteins were confirmed by mass spectrometry analysis. 2.5. Western blot analysis Resolved proteins were transferred onto nitrocellulose membranes with a semi-dry trans-blot apparatus (Bio-Rad) and tested for specificity against Angiostrongylus-positive sera, NHS, and a specificity control. The membranes were blocked with 5% skim milk for 1 h at room temperature and then were incubated for 2 h with a pool of positive or negative serum or a specificity control diluted at 1:100. After three washes with PBS Tween (0.05%), the membranes were incubated with a secondary peroxidase-conjugated anti-human IgG antibody diluted at 1:8000 and an IgG4 antibody diluted at 1:1000 for 1 h at room temperature. Bands were visualized with after the strips were incubated with 0.05% diaminobenzidine (DAB; Sigma-Aldrich), 0.015% H2O2, in PBS, pH 7.4. 2.2. Serum samples 2.6. N-glycosidase F (PNGase F) treatment for deglycosylation To test the antigenicity of the purified recombinant proteins, Western blot analyses were performed with Angiostrongylus-positive sera (which included sera from 20 individuals and a pool of 5 selected individual positive sera, normal human sera (NHS), which was pooled from 20 individuals that had no exposure to A. cantonensis and a pool of Echinococcus granulosus-positive sera, as a specificity control. All of the sera were diluted 1:100 when used. The serums obtained from Angiostrongylus-infected patients originated from biobanks from Pontifical Catholic University of Rio Grande do Sul (PUCRS). The use of these sera for this study was approved by the ethics committee of our institution. In addition, the specificity controls used were obtained from the serum bank from the Immunochemistry Laboratory of the Division of Parasitic Diseases and Malaria, Centers for Disease Control and Prevention (CDC). PNGase F treatment was performed according to the manufacturer’s instructions (5 × 105 U/mL; BioLabs, United Kingdom). To investigate the antigenicity of the proteins in the absence of carbohydrates, PNGase F was mixed with each of the four recombinant proteins (Lec-5, 14-3-3, ES-7-HEK, and ES-7-CHO and PBS as a control) at the concentrations listed in Section 2.4. Then, each of the samples were incubated overnight at 37 °C with 500 U of PNGase F. Recognition of the carbohydrate moieties by positive and negative sera was tested by dot blot. 2.7. Dot blot Recombinant proteins were blotted onto nitrocellulose membranes at concentrations of 1.4 ug, 5.5 ug, 0.5 ug, and 2.5 ug for Lec-5, 14-3-3, ES-7-HEK, and ES-7-CHO, respectively. Each membrane was treated as described for Western blot analysis. 33 Molecular & Biochemical Parasitology 221 (2018) 32–35 B.B. Cognato et al. Fig. 1. Expression of recombinant ES-7 protein in mammalian cells. A representative Western blots is shown for the detection of ES-7 recombinant protein (arrow) expressed by CHO cells with Angiostrongylus-positive serum (a) and normal human serum (b), and by HEK cells with Angiostrongylus-positive serum (c) and normal human serum (d). M – molecular weight marker in kDa. 3. Results Table 1 Recombinant proteins ES-7, Lec-5, and 14-3-3 in CHO cells, as well as ES-7 in HEK cells, were analyzed in 1D and 2D electrophoresis and Western blots using pooled or individual serum samples from infected individuals with Angiostrongylus (A-positive), normal human serum (NHS) and pooled serum samples from individuals infected with Echinococcus (Epositive). 3.1. ES-7 Recombinant ES-7 protein was purified from HEK (ES7-HEK) and CHO (ES7-CHO) cells at concentrations of 0.4 mg/ml and 0.5 mg/ml, respectively, with approximately 90% purity. The molecular weights varied from below 37–50 kDa in CHO cells and above 50–75 kDa in HEK cells and with different number and shape of spots (Fig. 1a and c). For both analyses, the isoelectric points varied between 3 and 7, resulting in a mean isoelectric point of 5. In addition, both ES7-HEK and ES7-CHO were recognized by a pool of positive A. cantonensis sera, yet were not recognized by NHS (Fig. 1a–d). However, when 0.12 μg of ES7-CHO and or ES7-HEK (0.15 μg) were subjected to 1D electrophoresis, recognition was observed in all tested sera (either by positive, NHS and specificity control, (Table 1). After the expressed proteins were treated with PNGase, no reaction with the Angiostrongylus-positive control sera was observed. Protein ES-7 CHO ES-7 HEK Lec-5 CHO 14-3-3 CHO Pooled serum Specificity serum Individual serum E-Positive A-Positive NHS A-Positive NHS 1D 2D 1D 2D 1D 2D 1D 1D + + + + + + + + + + − + − − − + + + + + − − − + 20/20 11/20 11/20 0/20 17/20 20/20 9/20 1/20 in an average size of 66 kDa (with 26 kDa representing GST). In 2D electrophoresis, eleven spots were identified with isoelectric points varying between 3 and 7. Lec-5 was recognized by a pool of A. cantonensis infected sera, while very weak recognition was observed with the pool of NHS. However, it was recognized by 11 of the 20 individual Angiostrongylus-infected serum samples and by nine of NHS individual serum (Table 1). Again after PNGase treatment no reaction with positive sera was observed. 3.2. Lec-5 Recombinant Lec-5 protein was purified from CHO cells at a concentration of 0.28 mg/ml with approximately 85% purity. The molecular size of the purified protein varied from 50 kDa to 75 kDa, resulting 34 Molecular & Biochemical Parasitology 221 (2018) 32–35 B.B. Cognato et al. specifically recognized by all of the serums tested, demonstrating that this protein cannot serve as specific targets. 3.3. 14-3-3 Recombinant 14-3-3 protein was purified from CHO cells at a concentration of 0.11 mg/ml with approximately 90% purity. It was detected at a predicted molecular weight of 54 kDa (28 kDa + 26 kDa GST) in both 1D and 2D electrophoresis and it had an isoelectric point of 4.83. The 14-3-3 protein was recognized by all 20 positive serums for A. cantonensis, as well as all 20 NHS samples (Table 1). Therefore, 14-33 was not specifically recognized. Also, after the enzymatic removal of glycans by PNGase from 14 to 3-3 protein did not react with Angiostrongylus- positive controls. 5. Conclusion Glycosylation is a very important posttranslational modification, especially for recombinant proteins. Here, the importance of eukaryotic cell systems, specifically CHO and HEK mammalian cells, for the expression of three antigenic proteins from Angiostrongylus cantonensis: ES-7, Lec-5 and 14-3-3, was demonstrated. These proteins were recognized by Angiostrongylus-positive serum, thereby resulting in a more specific reaction than previously observed when these proteins were expressed in prokaryotic cells, thus, seem to be essential for efficient biding by the antibody. While specific detection of 14-3-3 protein was not achieved in the present study, the recombinant expression of ES-7 protein by CHO/HEK and Lec-5 by CHO cells provided very promising antigens to improve a 2D-based immunodiagnosis of angiostrongyliasis. 4. Discussion Currently, the primary diagnostic antigen for angiostrongyliasis is a complex of glycoproteins that have been isolated from female worms of A. cantonensis and exhibit an average molecular weight of 31 kDa in Western blots [6,7]. Veríssimo and collaborators [12] have shown that the antigenicity presented by the 31 kDa antigen and by the total soluble extract of A. cantonensis is dependent on the presence of carbohydrate structures. This accounts for the lack of recognition that was observed when these proteins were expressed in prokaryotic cells [8]. Here, the importance of using eukaryotic cell systems to obtain glycosylated forms of heterologous expressed proteins was demonstrated since immunogenicity of these recombinant proteins was shown to be dependent on the presence of glycan moieties, since recognition by Angiostrongylus-positive sera was absent following PNGase treatment of all three proteins. In the 2D electrophoresis analysis of the ES7-HEK, ES7-CHO, and Lec-5 proteins, several spots were observed (Fig. 1a). Proteins generally present more than one isoform, with each having the same function, yet the isoforms can be encoded by a different gene and contain small differences in their peptide sequences. While this type of diversity has been observed in naturally synthesized proteins, it does not explain the number of spots that can be observed in Fig. 1a, especially since the recombinant construction method most likely results in clones with the same sequence. A possible explanation for this would be the glycidic moieties incorporated into a nascent protein which may cause a great variation on isoelectric point of the protein [13]. In addition, ES-7 purified from HEK cells showed six well defined spots, while the ES-7 purified from CHO cells produced more elongated and blurred spots proximal to the acidic isoelectric point (Fig. 1a). These results were consistent with those of other studies where expressed proteins in CHO and HEK cells presented differences in glycosylation patterns. CHO cells have consistently been shown to contain higher levels of sialic acid [14] and to produce distinct isoform patterns [15–17]. Another interesting observation in the present study was the specific recognition of ES-7 by Angiostrongylus-positive serum only after 2D electrophoresis (Table 1). These results suggested that at 1D electrophoresis many proteins with possible common glycan epitopes are comigrating with the ES-7 protein at same molecular weight, probably contaminants from GST-binding purification. These proteins were possibly carrying common glycidic moieties, causing cross-reactivity, because PGNase treatment abolished recognition. Based on the possible differences between HEK and CHO cells as described above, and the lack of antigenicity observed for ES-7 when it was expressed in prokaryotic systems [8], the present findings support the use of HEK cells for the heterologous expression of Angiostrongylus-derived antigens and the necessity of using wide separation electrophoresis to avoid crossreactivity. The recognition of expressed Lec-5 protein (at 0.28 μg/μL) by Angiostrongylus-positive serum, and not by NHS and specificity controls, in 2D Western blots, clearly demonstrated the usefulness of the Lec-5 protein for an immunodiagnosis of angiostrongyliasis, however it was also necessary to use 2D separation of the antigen, since it was recognized by 9 of 20 NHs serum at 1D. The 14-3-3 protein was non- Funding This work was supported by the Conselho Nacional de Pesquisa e Desenvolvimento Tecnológico do Brasil (grant numbers CNPq,401904/ 2013-0, 2013 and CNPq PQ1D307005/2014-3, 2014), CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) (grant number Edital 32, 2010). References [1] Q.P. Wang, D.H. Lai, X.Q. Zhu, X.G. Chen, Z.R. Lun, Human angiostrongyliasis, Lancet Infect. Dis. 8 (2008) 621 e 630. [2] A.L. Morassutti, S.C. Thiengo, M. Fernandez, K. Sawanyawisuth, C. Graeff-Teixeira, Eosinophilic meningitis caused by Angiostrongylus cantonensis: an emergent disease in Brazil, Mem. Inst. Oswaldo Cruz 109 (4) (2014) 399–407. [3] D.Y. Kim, T.B. Stewart, R.W. Bauer, M. Mitchell, Parastrongylus (=Angiostrongylus) cantonensis now endemic in Louisiana wildlife, J. Parasitol. 88 (2002) 1024–1026. [4] T.J. Slom, M.M. Cortese, S.I. Gerber, R.C. Jones, T.H. Holtz, A.S. Lopez, C.H. Zambrano, R.L. Sufit, Y. Sakolvaree, W. Chaicumpa, B.L. Herwaldt, S. Johnson, An outbreak of eosinophilic meningitis caused by Angiostrongylus cantonensis in travelers returning from the Caribbean, N. Engl. J. Med. 346 (9) (2002) 668–675. [5] T. Pincay, L. García, E. Narváez, O. Decker, L. Martini, J.M. Moreira, Angiostrongyliasis due to Parastrongylus (Angiostrongylus) cantonensis in Ecuador. First report in South America, Trop. Med. Int. Health 14 (Suppl. 2) (2009) S37. [6] P. Eamsobhana, A. Tungtrongchitr, D. Wanachiwanawin, et al., Characterization of 31 kda specific antigen from Parastrongylus cantonensis (Nematoda: Metastrongylidae), Int. Med. Res. J. 2 (1988) 9–12. [7] A.L. Morassutti, K. Levert, A. Perelygin, A.J. Da silva, P. Wilkins, C. Graeff-Teixeira, The 31-kDa antigen of angiostrongylus cantonensis comprises distinct antigenic glycoproteins, Vector-Borne Zoon. Dis. 12 (2012) 961–968. [8] A.L. Morassutti, A. Perelygin, K. Levert, S.C. Lin, Y.M. Lee, A.J. da Silva, P.P. Wilkins, C. Graeff-Teixeira, Expression of recombinant antigenic proteins from Angiostrongylus cantonensis: a brief report, Hawaii J. Med. Public Health 72 (6 Suppl. 2) (2013) 58–62. [9] A.L. Morassutti, K. Levert, P.M. Pinto, A.J. da Silva, P. Wilkins, C. Graeff-Teixeira, Characterization of Angiostrongylus cantonensis excretory-secretory proteins as potential diagnostic targets, Exp. Parasitol. 130 (1) (2012) 26–31. [10] A. Dell, A. Galadari, F. Sastre, P. Hitchen, Similarities and differences in the glycosylation mechanisms in prokaryotes and eukaryotes, Int. J. Microb. (2010) 148178. [11] K.F. Suen, M.S. Turner, F. Gao, B. Liu, A. Althage, A. Slavin, W. Ou, E. Zuo, M. Eckart, T. Ogawa, M. Yamada, T. Tuntland, J.L. Harris, J.W. Trauger, Transient expression of an IL-23R extracellular domain Fc fusion protein in CHO vs. HEK cells results in improved plasma exposure, Protein Exp. Purif. 71 (1) (2010) 96–102. [12] C.M. Veríssimo, A.L. Morassutti, M. von itzstein, G. Sutov, L. Hartley-Tassell, S. McAtamney, A. Dell, S.M. Haslam, C. Graeff-Teixeira, Characterization of the N-glycans of female Angiostrongylus cantonensis worms, Exp. Parasitol. 166 (2016) 137–143. [13] E.P. Go, H.X. Liao, S.M. Alam, D. Hua, B.F. Haynes, H. Desaire, Characterization of hostcell line specific glycosylation profiles of early transmitted/founder HIV-1 gp120 envelope proteins, J. Proteome Res. 12 (3) (2013) 1223–1234. [14] A. Croset, L. Delafosse, J.P. Gaudry, C. Arod, L. Glez, C. Losberger, D. Begue, A. Krstanovic, F. Robert, F. Vilbois, L. Chevalet, B. Antonsson, Differences in the glycosylation of recombinant proteins expressed in HEK and CHO cells, J. Biotechnol. 161 (3) (2012) 336–348. [15] T. Shinkawa, K. Nakamura, N. Yamane, E. Shoji-Hosaka, Y. Kanda, M. Sakurada, K. Uchida, H. Anazawa, M. Satoh, M. Yamasaki, N. Hanai, K. Shitara, The absence of fucose but not the presence of galactose or bisecting N-acetylglucosamineof humanIgG1 complex-type oligosaccharides shows the critical role of enhancing antibody-dependent cellular cytotoxicity, J. Biol. Chem. 278 (5) (2003) 3466–3473. [16] A. Zeck, G. Pohlentz, T. Schlothauer, J. Peter-Katalinić, J.T. Regula, Cell type-specific and site directed N-glycosylation pattern of Fc(RIIIa, J. Proteome Res. 10 (7) (2011) 3031–3039. [17] J. Ye, V. Kober, M. Tellers, Z. Naji, P. Salmon, J.F. Markusen, High-level protein expression in scalable CHO transient transfection, Biotechnol. Bioengy 103 (3) (2009) 542–551. 35