- Email: [email protected]
Effect of solvents on obligately anaerobic bacteria
ARTICLE IN PRESS
Anaerobe 14 (2008) 55–60 www.elsevier.com/locate/anaerobe
Physiology and microbial chemistry
Effect of solvents on obligately anaerobic bacteria Maria Fernanda Rodriguez Martineza, Niki Kelessidoua, Zoe Lawa, John Gardinerb, Gill Stephensa, a
School of Chemical Engineering and Analytical Science, University of Manchester, P.O. Box 88, Manchester M60 1QD, UK b School of Chemistry, University of Manchester, P.O. Box 88, Manchester M60 1QD, UK Received 23 August 2006; received in revised form 21 August 2007; accepted 21 September 2007 Available online 7 October 2007
Abstract Growth of Acetobacterium woodii and Clostridium sporogenes was studied in the presence of water-immiscible solvents. Nitrogen purging, vacuum distillation or distillation under nitrogen were all suitable as methods to remove oxygen from the solvents, since growth rates and yields of A. woodii were unaffected in the presence of tetradecane which had been degassed by these methods. Varying the solvent volume from 20% to 80% of the culture volume had little effect on growth rate of A. woodii. A.woodii was relatively sensitive to organic solvents since growth was inhibited by alkanes with log Poctanol/water values below 7.1. C. sporogenes was less solvent sensitive, since it grew without inhibition when the log P of the solvent was X6.6. Nevertheless, both A. woodii and C. sporogenes were more sensitive to solvent polarity than aerobic bacteria. r 2007 Elsevier Ltd. All rights reserved. Keywords: Solvent toxicity; Two-liquid phase culture systems; Acetobacterium woodii; Clostridium sporogenes
1. Introduction Obligately anaerobic bacteria are attracting increasing attention as a source of novel enzymes for biocatalytic reductions [1]. In particular, Acetobacterium woodii can reduce C–C and C–N double bonds [2–7], and the organism is also a useful biocatalyst for demethylation [8,9]. Similarly, the first example of a biocatalytic amide reduction was demonstrated using Clostridium sporogenes [10]. Many of the substrates are non-polar and some are very toxic [6,9,11]. Therefore, it would be useful to develop two-liquid phase reaction systems to deliver the waterinsoluble substrates and to protect the cells from substrate toxicity while obtaining efficient biotransformation [12–15]. In two-liquid phase reaction systems, a water-immiscible organic solvent is mixed with the aqueous reaction mixture, forming a second liquid phase. The toxic, non-polar substrate partitions preferentially into the organic phase so that the aqueous phase concentration is maintained at a Corresponding author. Tel.: +44 161 306 4377; fax: +44 161 306 4399.
E-mail address: [email protected] (G. Stephens). 1075-9964/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.anaerobe.2007.09.006
non-toxic level. High reaction rates can be maintained because the substrate equilibrates between the two phases as it is consumed by the cells suspended in the aqueous phase. However, developing these reaction systems depends on identifying suitable, non-toxic solvents. While solvent biocompatibility has been studied in great detail for aerobic biocatalysts [12,13,16,17], studies with anaerobes have been restricted to facultative species or the more oxygen-tolerant species, and rather few studies have been done with the phase ratios (vsolvent/vtotal) needed for biotransformation processes [18–23,29]. Therefore, we studied the sensitivity of A. woodii and C. sporogenes to organic solvents in a two-liquid phase reaction system to identify biocompatible solvents suitable for use in strictly anaerobic biotransformations. The simplest test for solvent toxicity is to test for growth inhibition [17]. However, it was first necessary to develop effective methods for deoxygenating the solvents prior to use, to avoid oxygen inhibition of this strict anaerobe. Although two-liquid phase reaction systems have been used previously with anaerobes for extractive fermentations [18–23] strictly anaerobic conditions were not required because the phase ratios were very low or because
ARTICLE IN PRESS 56
M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
the organisms could tolerate low concentrations of oxygen [24–28]. In contrast, the really strict anaerobes require completely oxygen-free culture media and reaction mixtures for biotransformations [1,9], particularly at the relatively high phase ratios needed for two-liquid phase biotransformations. Therefore, better methods were needed to deoxygenate the solvents. A vacuum degassing system has been used to prepare small quantities of anaerobic hydrocarbons for enrichment and isolation of alkane-degrading anaerobes [29]. However, this system is not really suitable for large-scale preparation of more volatile solvents since solvent losses would become substantial. Therefore, we assessed nitrogen purging, vacuum distillation and distillation under nitrogen, since these methods have been developed for degassing bulk solvents for use in organic chemistry [30]. Using these methods, we found that A. woodii and C. sporogenes are somewhat less tolerant to solvents than aerobic bacteria. 2. Materials and methods
a trace element solution containing MgCl2 6H2O (3.3 g l1), CaCl2 2H2O (4 g l1), (NH4)6Mo7O24 4H2O (1 g l1), FeSO4 7H2O (2.9 g l1) and either MnSO4 H2O (0.04 g l1) or MnSO4 4H2O (0.05 g l1) were prepared in the same way and added to the medium (both at 10 ml l1). A vitamin solution (p-amino benzoic acid, 0.08 g l1; 1 1 D-biotin, 0.004 g l ; riboflavin, 0.02 g l ) was sparged with N2 and then transferred to the anaerobic cabinet, filter sterilized and added to the medium (10 ml l1). The complete medium was dispensed into sterile universals in the anaerobic cabinet. Phenylalanine (121 mM) was prepared by suspending the required weight in distilled water, adding NaOH pellets until it dissolved, and then adjusting to pH 7.0 by adding 1 or 2 M HCl. The solution was purged with N2, transferred to the anaerobic cabinet, filter sterilized and added to the medium to the concentrations stated in the text. Pre-cultures were inoculated from stock cultures and grown in medium C for at least 10 h, and then aliquots of pre-cultures (0.5 ml) were used to inoculate the medium (11 ml).
2.1. Microorganisms and growth
2.2. Growth in two-liquid phase cultures
A. woodii DSM 1030 was re-purified from the DSMZ stock culture, and its ability to grow on H2 plus CO2 and to ferment fructose by a homoacetate fermentation was confirmed. The pure culture was then maintained and grown in Balch medium as described previously [7] using fructose (11.1 mM) in the maintenance medium. Experimental cultures were grown with the fructose concentrations stated, after inoculation (3%) with pre-cultures which had been grown for 2–3 d. C. sporogenes DSM 795 was maintained in cooked meat medium prepared by dissolving two cooked meat medium tablets (Merck) in 10 ml distilled water overnight at 4 1C. A solution (8 ml) of yeast extract (Oxoid; 5 g l1), K2HPO4 (5 g l1), L-cysteine HCl (0.5 or 1 mg l1) and resazurin (1 mg l1 added from a 1%, w/v, stock solution) was added, together with 0.4 g agar and water to 20 ml. Bottles were sealed with a subaseal and sparged with N2 for 20 min before autoclaving and transferring to the anaerobic cabinet to pour the plates. Stock cultures were incubated for at least 24 h and then used to inoculate experimental pre-cultures. The stock cultures were subcultured every 10 d. Experimental cultures of C. sporogenes were grown at 30 1C in medium C [31] which was modified as follows. The basal medium solution (950 ml) contained proteose peptone (Difco; 20 g), yeast extract (Oxoid; 5 g) and resazurin (1 mg; added from a 1%, w/v, stock solution) and was made anaerobic by boiling and then sparging with N2 while cooling. The medium was autoclaved, cooled and transferred to the anaerobic cabinet. A salt solution containing K2HPO4 (131 g l1), KH2PO4 (34 g l1) and NaSeO3 5H2O (0.021 g l1) was sparged with N2, autoclaved, transferred to the anaerobic cabinet and added to the basal medium (20 ml l1). A sodium thioglycollate solution (30 g l1) and
Anaerobic solvents were prepared by nitrogen purging for 15–20 min, vacuum distillation or distillation in an inert atmosphere [30]. For N2 purging, solvents were placed in sterile universals sealed with subaseals, and purged and vented via syringe needles which had been sterilized by soaking in 70% methanol, then flaming immediately before use. Solvents were added to cultures of A. woodii to 20% of the culture volume (phase ratio, 0.2) unless stated otherwise, and were mixed as described in the text. Cultures of C. sporogenes contained solvents at a phase ratio of 0.179, and were stirred at 200 rpm on a magnetic stirrer. All cultures were sealed with subaseals. In all cases, the results are the means for at least two cultures. 2.3. Analytical methods Cultures were sampled by inverting them so that the aqueous phase settled over the subaseal, allowing it to be sampled directly using a sterile syringe. This minimized contamination of the aqueous phase with droplets of tetradecane, which interfered with OD readings and other analytical procedures. Growth was monitored routinely by measuring the OD at 660 nm using a Pye Unicam PU8600 UV–vis spectrophotometer. Samples were diluted 5-fold (A. woodii) or 10-fold (C. sporogenes) to further minimize interference from droplets of tetradecane. In some experiments with C. sporogenes, emulsification became more extensive as growth continued and this caused problems with OD readings. Therefore, all growth rates are initial growth rates, calculated from OD readings taken before the emulsification became a problem. Growth yields were determined from dry weight determinations made by collecting cells in aqueous phase samples by centrifugation,
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
washing twice in 20 mM potassium phosphate buffer pH 7.0, resuspending in the same buffer, collecting the cells on a dried preweighed filter and then drying to constant weight at 60 1C. Acetate and fructose concentrations in supernatants from aqueous phase samples were determined as described previously [7]. The extent of emulsification was assessed by transferring the cultures to measuring cylinders after incubation of the cultures had been completed. The phases were allowed to separate and the volume of the separated solvent phase was measured. This was compared with the initial volume of solvent added to the cultures. 3. Results 3.1. Development of methods to deoxygenate the solvents The first step was to test a range of methods to produce deoxygenated solvents for use with A. woodii. We tested for effective deoxygenation of the solvent by mixing it with culture medium and testing for growth inhibition. It was important to use a non-toxic solvent for this work, and we assumed that tetradecane would be suitable because the log Poctanol/water value is high (7.6) [16,32] and it has been used successfully in a wide range of whole cell biotransformations with aerobes, e.g. Ref. [33]. Initially, the tetradecane was degassed by purging with nitrogen. The absence of oxygen was confirmed by mixing the solvent with culture medium, which contains resazurin (a redox-sensitive dye). When the anaerobic solvent was added to the medium, the colour did not change and the medium remained yellow. However, when aerobic solvent was added, the colour changed to pink, indicating that the resazurin had been oxidized. This confirmed that nitrogen purging would remove oxygen from the solvent. The next step was to test A. woodii for growth in the presence of anaerobic tetradecane, by adding the solvent to 20% of the culture volume (phase ratio, 0.2). The cultures were stirred to ensure effective contact between the two immiscible liquid phases. The solvent had no significant effect on OD, growth rate, biomass yield, growth yield, culture pH or acetate production (Table 1). This indicated that nitrogen purging was an effective method for removing oxygen from the solvent, and also that tetradecane was not toxic to A. woodii. Similar results were obtained when A. woodii was grown in the presence of tetradecane which had been degassed by vacuum distillation or by distillation under nitrogen (Table 2). Since nitrogen purging is the simplest of the three methods tested, this method was used for solvent degassing in all subsequent experiments. We confirmed that A. woodii was unable to grow on tetradecane by demonstrating that there was no growth in two-liquid phase control cultures which did not contain fructose. Furthermore, the tetradecane was self-sterile, since there was no growth in uninoculated cultures incubated with fructose and tetradecane, and there was
57
Table 1 Growth of Acetobacterium woodii in the presence and absence of tetradecane Parameter
Without tetradecane
0.09970.022 Growth rate (h1) Final biomass concentration 0.8070.016 (gdry weight l1) Final acetate concentration (mM) 5571.8 Fructose consumed (mM) 2570.64 Growth yield (gcells gfructose 1) 0.1870 Acetate: fructose ratio (mol mol1) 2.270.014 Final pH 4.64
With tetradecane 0.08770.0061 0.9170.14 5372.5 2570.67 0.2170.021 2.270.028 4.67
A. woodii was grown on fructose (28 mM) in the presence and absence of degassed tetradecane. The solvent was degassed by purging with nitrogen and the phase ratio was 0.2. The cultures were stirred with a magnetic stirrer at 700 rpm. The data are the means and standard deviations for duplicate cultures, and are shown to two significant figures to reflect the accuracy of the assays.
Table 2 Effect of solvent degassing method on growth of A. woodii Degassing method
Growth rate (h1)
Nitrogen purging Vacuum distillation Distillation under nitrogen
0.14 0.14 0.12
A. woodii was grown on fructose (28 mM) in the presence of tetradecane. The solvent was degassed as indicated and the phase ratio was 0.2. The cultures were stirred with a magnetic stirrer at 700 rpm. The growth rates are calculated from the mean OD values in duplicate cultures. Variation in growth rate between Table 1 and this table is due to differences in the age of the inoculum.
no emulsification. Aerobic tetradecane inhibited growth. We also checked the effect of different phase ratios of nitrogen-purged tetradecane. Increasing the phase ratio progressively from 20% to 80% in increments of 20% had very little effect on the growth rate of A. woodii (results not shown). Next, we demonstrated that solvent degassing by nitrogen purging was suitable for a different species. For this, we grew C. sporogenes on phenylalanine (10.5 mM) in the presence and absence of nitrogen-purged tetradecane. The growth rate was 0.2 h1 for both sets of duplicate cultures. Therefore, the degassing method was also suitable for C. sporogenes and the solvent was not toxic. 3.2. Effect of different solvents on growth of A. woodii and C. sporogenes The toxicity of solvents to aerobic bacteria is related to the log Poctanol/water of the solvent, where log P is defined as the logarithm of the partition coefficient of the solvent in a standard octanol–water two-phase system [13,16,34]. Therefore, we examined the relationship between log P value and solvent toxicity for A. woodii by growing the organism on fructose in the presence of a series of
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
n-alkanes (6–14 carbon atoms) with log P values of 3.5–7.6 [32]. Although tetradecane and tridecane had no effect on growth, growth was inhibited progressively by the other solvents as the log P value decreased. Thus, the growth rate decreased as the chain length of the alkanes decreased to 10 carbon atoms (Fig. 1). Growth was inhibited completely in the presence of nonane or octane. Cyclohexane (log P ¼ 3.2) and propylbenzene (log P ¼ 3.6) were also tested, but there was very little growth with either solvent even after 335 h. However, some growth occurred in the presence of heptane (log P ¼ 4.0) and hexane (log P ¼ 3.5) but only after very prolonged incubation. Thus, the cultures did not reach stationary phase until 131 h with heptane and 335 h with hexane, compared with 47 h in control cultures grown without solvent. Heptane and hexane are extremely volatile, and the rubber seals in the culture bottles swelled during the experiment, indicating that the solvent vapour could penetrate the rubber. This suggests that the growth with heptane and hexane was due to loss of the solvents rather than inherent resistance to the solvents. An alternative explanation is that solventresistant mutants were selected, but this explanation is less attractive since no growth was observed with octane or nonane, even though mutant selection would be expected to occur with these hydrocarbons also, due to their similar structures and physical properties. There was extensive emulsification in cultures grown with the C6–C12 alkanes. The phases could not be separated at all after incubation with hexane, heptane or octane and incomplete phase separation was obtained in cultures containing nonane, decane, undecane and dode-
Growth rate (% control)
140 120 100 80 60 40 20 0 4
4.5
5
5.5
6
6.5
7
7.5
8
Solvent log P Fig. 1. Effect of solvent log P value on growth of A. woodii and C. sporogenes. A. woodii (’) and C. sporogenes (K) were grown on 25 mM fructose and phenylalanine (10.5 mM), respectively, in the presence of a series of n-alkanes at a phase ratio of 0.2. The solvents were octane (log P ¼ 4.5), nonane (log P ¼ 5.1), decane (log P ¼ 5.6), undecane (log P ¼ 6.1), dodecane (log P ¼ 6.6), tridecane (log P ¼ 7.1) and tetradecane (log P ¼ 7.6 [32]). The cultures were mixed using an IKA shaker at 500 rpm. Growth rates shown as a percentage of the growth rate in control cultures grown in the absence of solvent (0.11 h1 for A. woodii and 0.2 h1 for C. sporogenes), and are the mean values for duplicate cultures. Growth rates are given as initial growth rates because emulsification of the solvent interfered with OD readings progressively as growth continued.
5 4 3 OD
58
2 1 0 0
5
10
15
20
Time (h)
Fig. 2. Growth of C. sporogenes in the presence and absence of tetradecane with different concentrations of phenylalanine C. sporogenes was grown with phenylalanine provided at 5.25 mM (K, J) or 10.5 mM (’, &), in the presence (K, ’) or absence (J, &) of tetradecane (phase ratio 0.179). The OD values are the averages for duplicate cultures.
cane. The efficiency of phase separation increased with the alkane chain length and complete phase separation was obtained after growth with tridecane or tetradecane. The solvent sensitivity of C. sporogenes was also tested, by measuring growth on phenylalanine (10.5 mM) in the presence of decane, dodecane or tetradecane at a phase ratio of 0.179, after degassing the solvents by nitrogen purging. The solvents had little effect on growth when the OD was p1, but we noticed that the OD readings became much higher than the control values as growth continued. This was due to extensive emulsification of all three of the solvents, which increased progressively after the OD increased above 1. Since the OD readings were accurate during the initial stages of growth, when there was little emulsification (see Fig. 2), it was possible to calculate the initial growth rates using these data. The growth rates in the presence of dodecane and tetradecane were slightly higher than in the absence of the solvents (Fig. 1), indicating that they were non-toxic, and possibly stimulatory. There was a lag phase in the presence of decane (not shown), indicating that there was some inhibition, but the growth rate thereafter was similar to the solvent-free control. 3.3. Emulsification of solvents in cultures of C. sporogenes As noted above, the solvents were emulsified in cultures of C. sporogenes during the later phases of growth, when the OD was X1. Therefore, the OD readings were much higher in cultures grown on 10.5 mM phenylalanine with tetradecane than in the control (Fig. 2). When C. sporogenes was grown with a lower phenylalanine concentration (5.25 mM) in the presence of tetradecane, the biomass concentration was correspondingly lower than in cultures grown with 10.5 mM phenylalanine (Fig. 2). With the lower phenylalanine concentration, there was very little emulsification of the solvent, and the OD values were
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
similar to the solvent-free control. This confirms that growth of C. sporogenes was not affected by the anaerobic solvent. The emulsifying agent was produced by the cells, since there was no emulsification when tetradecane was incubated with uninoculated medium containing 5.25 or 10.5 mM phenylalanine. Since there was no emulsification in cultures grown with 5.25 mM phenylalanine or during the early stage of growth with 10.5 mM phenylalanine, we conclude that the cells can only produce the emulsifying agents once the biomass concentration reaches a certain critical level. 4. Conclusions Vacuum distillation, distillation under nitrogen and nitrogen purging were all effective methods for removing oxygen from tetradecane, since the degassed solvent had no effect on the growth or metabolism of A. woodii, whichever degassing method was used. We conclude that nitrogen purging is the method of choice for small-scale experiments, since it is much simpler than the distillation methods. This method seems to be applicable to other anaerobic biocatalysts, since growth of C. sporogenes was also unaffected by solvents prepared in this way, and we have also used this procedure with Peptostreptococcus productus (Enzyme Microb. Technol. in press). However, it would be necessary to use the distillation methods to prepare anaerobic solvents on a larger scale, since purging results in substantial solvent loss by evaporation. Tetradecane was non-toxic to A. woodii at all phase ratios tested, which indicates that this solvent would be a suitable choice for two-liquid phase biotransformations using this organism. Tridecane was also non-inhibitory, but growth was inhibited by n-alkanes with lower log Poctanol/water values (o7.1). C. sporogenes was more resistant to solvent polarity, since it could grow well with solvents with log P values as low as 5.6 (decane). However, both organisms seem to be more sensitive to solvents than aerobic bacteria, which generally tolerate solvents with log P values greater than 4–5 [16,32]. Solvent sensitivity in aerobic bacteria is partly related to cell surface composition, since Gram positive aerobes tend to be more solvent sensitive than Gram negative species [32,35–38]. A. woodii has a very unusual cell surface composition, and this may account for its extreme sensitivity to solvents. Thus, the cells are remarkably hydrophobic and show strong adherence to liquid hydrocarbons, particularly as the cells approach stationary phase [39]. Solvent toxicity depends on both molecular toxicity (general inhibition of cellular processes) and phase toxicity, which is due to interfacial effects [13,40]. The tendency of A. woodii to accumulate at the interface would exacerbate phase toxicity by increasing the contact between the microorganism and the solvent. In addition, solvent emulsification may have contributed to solvent toxicity, since stable emulsions were formed in cultures of A. woodii
59
with all solvents with log P values below 7.1, and both toxicity and the extent of emulsification increased as the log P value of the solvent decreased. Emulsification increases the interfacial area and this may also have contributed to phase toxicity. Therefore, it appears that the extreme sensitivity of A. woodii to the shorter chain alkanes may be due to a combination of adherence of the cells to the alkane phase and emulsification. Emulsification may also explain the increased sensitivity of C. sporogenes to solvents compared with aerobes. However, even the non-toxic solvents were emulsified by C. sporogenes cultures grown to high biomass concentration, but this did not have a significant effect on growth. Therefore, solvent emulsification does not seem to be a primary cause of growth inhibition, but may exacerbate other toxic effects, which remain to be identified. This study shows that A. woodii and C. sporogenes are more sensitive to solvent toxicity than aerobic bacteria. Nevertheless, non-polar solvents, such as tetradecane, are not toxic and are suitable for use in anaerobic, two-liquid phase reaction systems. Indeed, we have already begun to demonstrate the suitability of these solvents for reduction of C–C and C–N double bonds [3,6]. The solvents can be deoxygenated using any of the standard procedures used in organic chemistry. This work complements previous studies in which anaerobic biotransformations using relatively non-toxic substrates were intensified by using a water-miscible solvent system to deliver insoluble substrates [41]. Therefore, a range of anaerobic solvent systems is now available for use in biotransformations using anaerobic bacteria. Acknowledgement MFRM was supported by an EU Socrates scholarship. References [1] Simon H, Bader J, Guenther H, Neumann S, Thanos J. Chiral compounds via biocatalytic reductions. Angew Chem 1985;97: 541–55. [2] Tschech A, Pfennig N. Growth yield increase linked to caffeate reduction in Acetobacterium woodii. Arch Microbiol 1984;137:163–7. [3] Li H, Williams P, Micklefield J, Gardiner JM, Stephens G. A dynamic combinatorial screen for novel imine reductase activity. Tetrahedron 2004;60:753–8. [4] Davies ET, Stephens GM. Effect of growth substrate and electron donor on hydrogenation of carbon–carbon double bonds by Acetobacterium woodii. Enzyme Microbiol Technol 1998;23:129–32. [5] Davies ET, Stephens GM. Optimization of cell harvesting and assay procedures for reductive biotransformations in obligate anaerobes. Adv Bioprocess Eng 1994:495–9. [6] Foroughi F, Williams P, Stephens G. Reduction of carbon–carbon double bonds using Acetobacterium woodii: determination of the optimum inducer structure. Enzyme Microbiol Technol 2006;39: 1066–71. [7] Davies ET, Stephens GM. Efficient mediator-independent hydrogenation of carbon–carbon double bonds, using the obligate anaerobe, Acetobacterium woodii, with fructose as an electron donor. Appl Microbiol Biotechnol 1996;46:615–8.
ARTICLE IN PRESS 60
M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
[8] Bache R, Pfennig N. Selective isolation of Acetobacterium woodii on methoxylated aromatic acids and determination of growth yields. Arch Microbiol 1981;130:255–61. [9] Kalil MS, Stephens GM. Catechol production by O-demethylation of 2-methoxyphenol using the obligate anaerobe, Acetobacterium woodii. Biotechnol Lett 1997;19:1165–8. [10] Dipeolu O, Gardiner J, Stephens G. Biocatalytic amide reduction using Clostridium sporogenes. Biotechnol Lett 2005;27:1803–7. [11] Schales O, Graefe HA. Arylnitroalkenes: a new group of antibacterial agents. J Am Chem Soc 1952;74:4486–90. [12] Salter GJ, Kell DB. Solvent selection for whole cell biotransformations in organic media. Crit Rev Biotechnol 1995;15:139–77. [13] Leon R, Fernandes P, Pinheiro HM, Cabral JMS. Whole-cell biocatalysis in organic media. Enzyme Microbiol Technol 1998;23:483–500. [14] Dervakos GA, Woodley JM, Keravnou ET, Washbrook J, Lilly MD. Development of a KBS [knowledge-based system] for biotransformation process design. AIChE Symp Ser 1989;114:283–91. [15] Schmid A, Kollmer A, Mathys RG, Witholt B. Developments toward large-scale bacterial bioprocesses in the presence of bulk amounts of organic solvents. Extremophiles 1998;2:249–56. [16] Laane C, Boeren S, Vos K, Veeger C. Rules for optimization of biocatalysis in organic solvents. Biotechnol Bioeng 1987. [17] Aono R, Tsukagoshi N, Miyamoto T. Evaluation of the growth and inhibition strength of hydrocarbon solvents against Escherichia coli and Pseudomonas putida grown in a two-liquid phase culture system consisting of a medium and organic solvent. Extremophiles 2001;5: 11–5. [18] Roffler SR, Blanch HW, Wilke CR. In situ recovery of butanol during fermentation. Part 2: fed-batch extractive fermentation. Bioprocess Eng 1987;2:181–90. [19] Playne MJ, Smith BR. Toxicity of organic extraction reagents to anaerobic bacteria. Biotechnol Bioeng 1983;25:1251–65. [20] Datta R. Acidogenic fermentation of corn stover. Biotechnol Bioeng 1981;23:61–77. [21] Setti L, Rossi M, Lanzarini G, Pifferi PG. Barrier and carrier effects of n-dodecane on the anaerobic degradation of benzothiophene by Desulfovibrio desulfuricans. Biotechnol Lett 1993;15:527–30. [22] Roffler SR, Blanch HW, Wilke CR. In situ recovery of fermentation products. Trends Biotechnol 1984;2:129–36. [23] Roffler SR, Blanch HW, Wilke CR. In situ recovery of butanol during fermentations. Part 1: batch extractive fermentations. Bioprocess Eng 1987;2:1–12. [24] O’Brien RW, Morris JG. Oxygen and the growth and metabolism of Clostridium acetobutylicum. J Gen Microbiol 1971;68:307–18. [25] Hewitt J, Morris JG. Superoxide dismutase in some obligately anaerobic bacteria. FEBS Lett 1975;50:315–8.
[26] Whittenbury R. Hydrogen peroxide formation and catalase activity in the lactic acid bacteria. J Gen Microbiol 1964;35:13–26. [27] Borch E, Molin G. The aerobic growth and product formation of Lactobacillus, Leuconostoc, Brochothrix and Carnobacterium in batch cultures. Appl Microbiol Biotechnol 1989;30:81–8. [28] Hansen TA. Metabolism of sulfate-reducing prokaryotes. Antonie van Leeuwenhoek J Microbiol Serol 1994;66:165–85. [29] Aeckersberg F, Bak F, Widdel F. Anaerobic oxidation of saturated hydrocarbons to CO2 by a new type of sulfate-reducing bacterium. Arch Microbiol 1991;156:5–14. [30] Shriver DF, Drezdzon MA. The manipulation of air-sensitive compounds. 2nd ed. New York: Wiley; 1986. [31] Giesel H, Simon H. On the occurrence of enoate reductase and 2-oxocarboxylate reductase in clostridia and some observations on the amino acid fermentation by Peptostreptococcus anaerobius. Arch Microbiol 1983;135:51–7. [32] Vermue M, Sikkema J, Verheul A, Bakker R, Tramper J. Toxicity of homologous series of organic solvents for the Gram-positive bacteria Arthrobacter and Nocardia sp. and the Gram-negative bacteria Acinetobacter and Pseudomonas sp. Biotechnol Bioeng 1993;42: 747–58. [33] Phumathon P, Stephens GM. Production of toluene cis-glycol using recombinant Escherichia coli strains in glucose-limited fed batch culture. Enzyme Microbiol Technol 1999;25:810–9. [34] Osborne SJ, Leaver J, Turner MK, Dunnill P. Correlation of biocatalytic activity in an organic–aqueous two-liquid phase system with solvent concentration in the cell membrane. Enzyme Microbiol Technol 1990;12:281–91. [35] Collins LD, Daugulis AJ. Characterization and optimization of a two-phase partitioning bioreactor for the biodegradation of phenol. Appl Microbiol Biotechnol 1997;48:18–22. [36] Harrop AJ, Hocknull MD, Lilly MD. Biotransformations in organic solvents: a difference between Gram-positive and Gram-negative bacteria. Biotechnol Lett 1989;11:807–10. [37] Inoue A, Horikoshi K. Estimation of solvent-tolerance of bacteria by the solvent parameter log P. J Ferment Bioeng 1991;71:194–6. [38] Rajagopal AN. Growth of Gram-negative bacteria in the presence of organic solvents. Enzyme Microbiol Technol 1996;19:606–13. [39] Donlon E, Colleran A. A comparison of different methods to determine the hydrophobicity of acetogenic bacteria. J Microbiol Methods 1993;17:27–37. [40] Bar R. Phase toxicity in multiphase biocatalysis. Trends Biotechnol 1986;4:167. [41] Tidswell EC, Salter GJ, Kell DB, Morris JG. Enantioselectivity of sulcatone reduction by some anaerobic bacteria. Enzyme Microbiol Technol 1997;21:143–7.
Anaerobe 14 (2008) 55–60 www.elsevier.com/locate/anaerobe
Physiology and microbial chemistry
Effect of solvents on obligately anaerobic bacteria Maria Fernanda Rodriguez Martineza, Niki Kelessidoua, Zoe Lawa, John Gardinerb, Gill Stephensa, a
School of Chemical Engineering and Analytical Science, University of Manchester, P.O. Box 88, Manchester M60 1QD, UK b School of Chemistry, University of Manchester, P.O. Box 88, Manchester M60 1QD, UK Received 23 August 2006; received in revised form 21 August 2007; accepted 21 September 2007 Available online 7 October 2007
Abstract Growth of Acetobacterium woodii and Clostridium sporogenes was studied in the presence of water-immiscible solvents. Nitrogen purging, vacuum distillation or distillation under nitrogen were all suitable as methods to remove oxygen from the solvents, since growth rates and yields of A. woodii were unaffected in the presence of tetradecane which had been degassed by these methods. Varying the solvent volume from 20% to 80% of the culture volume had little effect on growth rate of A. woodii. A.woodii was relatively sensitive to organic solvents since growth was inhibited by alkanes with log Poctanol/water values below 7.1. C. sporogenes was less solvent sensitive, since it grew without inhibition when the log P of the solvent was X6.6. Nevertheless, both A. woodii and C. sporogenes were more sensitive to solvent polarity than aerobic bacteria. r 2007 Elsevier Ltd. All rights reserved. Keywords: Solvent toxicity; Two-liquid phase culture systems; Acetobacterium woodii; Clostridium sporogenes
1. Introduction Obligately anaerobic bacteria are attracting increasing attention as a source of novel enzymes for biocatalytic reductions [1]. In particular, Acetobacterium woodii can reduce C–C and C–N double bonds [2–7], and the organism is also a useful biocatalyst for demethylation [8,9]. Similarly, the first example of a biocatalytic amide reduction was demonstrated using Clostridium sporogenes [10]. Many of the substrates are non-polar and some are very toxic [6,9,11]. Therefore, it would be useful to develop two-liquid phase reaction systems to deliver the waterinsoluble substrates and to protect the cells from substrate toxicity while obtaining efficient biotransformation [12–15]. In two-liquid phase reaction systems, a water-immiscible organic solvent is mixed with the aqueous reaction mixture, forming a second liquid phase. The toxic, non-polar substrate partitions preferentially into the organic phase so that the aqueous phase concentration is maintained at a Corresponding author. Tel.: +44 161 306 4377; fax: +44 161 306 4399.
E-mail address: [email protected] (G. Stephens). 1075-9964/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.anaerobe.2007.09.006
non-toxic level. High reaction rates can be maintained because the substrate equilibrates between the two phases as it is consumed by the cells suspended in the aqueous phase. However, developing these reaction systems depends on identifying suitable, non-toxic solvents. While solvent biocompatibility has been studied in great detail for aerobic biocatalysts [12,13,16,17], studies with anaerobes have been restricted to facultative species or the more oxygen-tolerant species, and rather few studies have been done with the phase ratios (vsolvent/vtotal) needed for biotransformation processes [18–23,29]. Therefore, we studied the sensitivity of A. woodii and C. sporogenes to organic solvents in a two-liquid phase reaction system to identify biocompatible solvents suitable for use in strictly anaerobic biotransformations. The simplest test for solvent toxicity is to test for growth inhibition [17]. However, it was first necessary to develop effective methods for deoxygenating the solvents prior to use, to avoid oxygen inhibition of this strict anaerobe. Although two-liquid phase reaction systems have been used previously with anaerobes for extractive fermentations [18–23] strictly anaerobic conditions were not required because the phase ratios were very low or because
ARTICLE IN PRESS 56
M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
the organisms could tolerate low concentrations of oxygen [24–28]. In contrast, the really strict anaerobes require completely oxygen-free culture media and reaction mixtures for biotransformations [1,9], particularly at the relatively high phase ratios needed for two-liquid phase biotransformations. Therefore, better methods were needed to deoxygenate the solvents. A vacuum degassing system has been used to prepare small quantities of anaerobic hydrocarbons for enrichment and isolation of alkane-degrading anaerobes [29]. However, this system is not really suitable for large-scale preparation of more volatile solvents since solvent losses would become substantial. Therefore, we assessed nitrogen purging, vacuum distillation and distillation under nitrogen, since these methods have been developed for degassing bulk solvents for use in organic chemistry [30]. Using these methods, we found that A. woodii and C. sporogenes are somewhat less tolerant to solvents than aerobic bacteria. 2. Materials and methods
a trace element solution containing MgCl2 6H2O (3.3 g l1), CaCl2 2H2O (4 g l1), (NH4)6Mo7O24 4H2O (1 g l1), FeSO4 7H2O (2.9 g l1) and either MnSO4 H2O (0.04 g l1) or MnSO4 4H2O (0.05 g l1) were prepared in the same way and added to the medium (both at 10 ml l1). A vitamin solution (p-amino benzoic acid, 0.08 g l1; 1 1 D-biotin, 0.004 g l ; riboflavin, 0.02 g l ) was sparged with N2 and then transferred to the anaerobic cabinet, filter sterilized and added to the medium (10 ml l1). The complete medium was dispensed into sterile universals in the anaerobic cabinet. Phenylalanine (121 mM) was prepared by suspending the required weight in distilled water, adding NaOH pellets until it dissolved, and then adjusting to pH 7.0 by adding 1 or 2 M HCl. The solution was purged with N2, transferred to the anaerobic cabinet, filter sterilized and added to the medium to the concentrations stated in the text. Pre-cultures were inoculated from stock cultures and grown in medium C for at least 10 h, and then aliquots of pre-cultures (0.5 ml) were used to inoculate the medium (11 ml).
2.1. Microorganisms and growth
2.2. Growth in two-liquid phase cultures
A. woodii DSM 1030 was re-purified from the DSMZ stock culture, and its ability to grow on H2 plus CO2 and to ferment fructose by a homoacetate fermentation was confirmed. The pure culture was then maintained and grown in Balch medium as described previously [7] using fructose (11.1 mM) in the maintenance medium. Experimental cultures were grown with the fructose concentrations stated, after inoculation (3%) with pre-cultures which had been grown for 2–3 d. C. sporogenes DSM 795 was maintained in cooked meat medium prepared by dissolving two cooked meat medium tablets (Merck) in 10 ml distilled water overnight at 4 1C. A solution (8 ml) of yeast extract (Oxoid; 5 g l1), K2HPO4 (5 g l1), L-cysteine HCl (0.5 or 1 mg l1) and resazurin (1 mg l1 added from a 1%, w/v, stock solution) was added, together with 0.4 g agar and water to 20 ml. Bottles were sealed with a subaseal and sparged with N2 for 20 min before autoclaving and transferring to the anaerobic cabinet to pour the plates. Stock cultures were incubated for at least 24 h and then used to inoculate experimental pre-cultures. The stock cultures were subcultured every 10 d. Experimental cultures of C. sporogenes were grown at 30 1C in medium C [31] which was modified as follows. The basal medium solution (950 ml) contained proteose peptone (Difco; 20 g), yeast extract (Oxoid; 5 g) and resazurin (1 mg; added from a 1%, w/v, stock solution) and was made anaerobic by boiling and then sparging with N2 while cooling. The medium was autoclaved, cooled and transferred to the anaerobic cabinet. A salt solution containing K2HPO4 (131 g l1), KH2PO4 (34 g l1) and NaSeO3 5H2O (0.021 g l1) was sparged with N2, autoclaved, transferred to the anaerobic cabinet and added to the basal medium (20 ml l1). A sodium thioglycollate solution (30 g l1) and
Anaerobic solvents were prepared by nitrogen purging for 15–20 min, vacuum distillation or distillation in an inert atmosphere [30]. For N2 purging, solvents were placed in sterile universals sealed with subaseals, and purged and vented via syringe needles which had been sterilized by soaking in 70% methanol, then flaming immediately before use. Solvents were added to cultures of A. woodii to 20% of the culture volume (phase ratio, 0.2) unless stated otherwise, and were mixed as described in the text. Cultures of C. sporogenes contained solvents at a phase ratio of 0.179, and were stirred at 200 rpm on a magnetic stirrer. All cultures were sealed with subaseals. In all cases, the results are the means for at least two cultures. 2.3. Analytical methods Cultures were sampled by inverting them so that the aqueous phase settled over the subaseal, allowing it to be sampled directly using a sterile syringe. This minimized contamination of the aqueous phase with droplets of tetradecane, which interfered with OD readings and other analytical procedures. Growth was monitored routinely by measuring the OD at 660 nm using a Pye Unicam PU8600 UV–vis spectrophotometer. Samples were diluted 5-fold (A. woodii) or 10-fold (C. sporogenes) to further minimize interference from droplets of tetradecane. In some experiments with C. sporogenes, emulsification became more extensive as growth continued and this caused problems with OD readings. Therefore, all growth rates are initial growth rates, calculated from OD readings taken before the emulsification became a problem. Growth yields were determined from dry weight determinations made by collecting cells in aqueous phase samples by centrifugation,
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
washing twice in 20 mM potassium phosphate buffer pH 7.0, resuspending in the same buffer, collecting the cells on a dried preweighed filter and then drying to constant weight at 60 1C. Acetate and fructose concentrations in supernatants from aqueous phase samples were determined as described previously [7]. The extent of emulsification was assessed by transferring the cultures to measuring cylinders after incubation of the cultures had been completed. The phases were allowed to separate and the volume of the separated solvent phase was measured. This was compared with the initial volume of solvent added to the cultures. 3. Results 3.1. Development of methods to deoxygenate the solvents The first step was to test a range of methods to produce deoxygenated solvents for use with A. woodii. We tested for effective deoxygenation of the solvent by mixing it with culture medium and testing for growth inhibition. It was important to use a non-toxic solvent for this work, and we assumed that tetradecane would be suitable because the log Poctanol/water value is high (7.6) [16,32] and it has been used successfully in a wide range of whole cell biotransformations with aerobes, e.g. Ref. [33]. Initially, the tetradecane was degassed by purging with nitrogen. The absence of oxygen was confirmed by mixing the solvent with culture medium, which contains resazurin (a redox-sensitive dye). When the anaerobic solvent was added to the medium, the colour did not change and the medium remained yellow. However, when aerobic solvent was added, the colour changed to pink, indicating that the resazurin had been oxidized. This confirmed that nitrogen purging would remove oxygen from the solvent. The next step was to test A. woodii for growth in the presence of anaerobic tetradecane, by adding the solvent to 20% of the culture volume (phase ratio, 0.2). The cultures were stirred to ensure effective contact between the two immiscible liquid phases. The solvent had no significant effect on OD, growth rate, biomass yield, growth yield, culture pH or acetate production (Table 1). This indicated that nitrogen purging was an effective method for removing oxygen from the solvent, and also that tetradecane was not toxic to A. woodii. Similar results were obtained when A. woodii was grown in the presence of tetradecane which had been degassed by vacuum distillation or by distillation under nitrogen (Table 2). Since nitrogen purging is the simplest of the three methods tested, this method was used for solvent degassing in all subsequent experiments. We confirmed that A. woodii was unable to grow on tetradecane by demonstrating that there was no growth in two-liquid phase control cultures which did not contain fructose. Furthermore, the tetradecane was self-sterile, since there was no growth in uninoculated cultures incubated with fructose and tetradecane, and there was
57
Table 1 Growth of Acetobacterium woodii in the presence and absence of tetradecane Parameter
Without tetradecane
0.09970.022 Growth rate (h1) Final biomass concentration 0.8070.016 (gdry weight l1) Final acetate concentration (mM) 5571.8 Fructose consumed (mM) 2570.64 Growth yield (gcells gfructose 1) 0.1870 Acetate: fructose ratio (mol mol1) 2.270.014 Final pH 4.64
With tetradecane 0.08770.0061 0.9170.14 5372.5 2570.67 0.2170.021 2.270.028 4.67
A. woodii was grown on fructose (28 mM) in the presence and absence of degassed tetradecane. The solvent was degassed by purging with nitrogen and the phase ratio was 0.2. The cultures were stirred with a magnetic stirrer at 700 rpm. The data are the means and standard deviations for duplicate cultures, and are shown to two significant figures to reflect the accuracy of the assays.
Table 2 Effect of solvent degassing method on growth of A. woodii Degassing method
Growth rate (h1)
Nitrogen purging Vacuum distillation Distillation under nitrogen
0.14 0.14 0.12
A. woodii was grown on fructose (28 mM) in the presence of tetradecane. The solvent was degassed as indicated and the phase ratio was 0.2. The cultures were stirred with a magnetic stirrer at 700 rpm. The growth rates are calculated from the mean OD values in duplicate cultures. Variation in growth rate between Table 1 and this table is due to differences in the age of the inoculum.
no emulsification. Aerobic tetradecane inhibited growth. We also checked the effect of different phase ratios of nitrogen-purged tetradecane. Increasing the phase ratio progressively from 20% to 80% in increments of 20% had very little effect on the growth rate of A. woodii (results not shown). Next, we demonstrated that solvent degassing by nitrogen purging was suitable for a different species. For this, we grew C. sporogenes on phenylalanine (10.5 mM) in the presence and absence of nitrogen-purged tetradecane. The growth rate was 0.2 h1 for both sets of duplicate cultures. Therefore, the degassing method was also suitable for C. sporogenes and the solvent was not toxic. 3.2. Effect of different solvents on growth of A. woodii and C. sporogenes The toxicity of solvents to aerobic bacteria is related to the log Poctanol/water of the solvent, where log P is defined as the logarithm of the partition coefficient of the solvent in a standard octanol–water two-phase system [13,16,34]. Therefore, we examined the relationship between log P value and solvent toxicity for A. woodii by growing the organism on fructose in the presence of a series of
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
n-alkanes (6–14 carbon atoms) with log P values of 3.5–7.6 [32]. Although tetradecane and tridecane had no effect on growth, growth was inhibited progressively by the other solvents as the log P value decreased. Thus, the growth rate decreased as the chain length of the alkanes decreased to 10 carbon atoms (Fig. 1). Growth was inhibited completely in the presence of nonane or octane. Cyclohexane (log P ¼ 3.2) and propylbenzene (log P ¼ 3.6) were also tested, but there was very little growth with either solvent even after 335 h. However, some growth occurred in the presence of heptane (log P ¼ 4.0) and hexane (log P ¼ 3.5) but only after very prolonged incubation. Thus, the cultures did not reach stationary phase until 131 h with heptane and 335 h with hexane, compared with 47 h in control cultures grown without solvent. Heptane and hexane are extremely volatile, and the rubber seals in the culture bottles swelled during the experiment, indicating that the solvent vapour could penetrate the rubber. This suggests that the growth with heptane and hexane was due to loss of the solvents rather than inherent resistance to the solvents. An alternative explanation is that solventresistant mutants were selected, but this explanation is less attractive since no growth was observed with octane or nonane, even though mutant selection would be expected to occur with these hydrocarbons also, due to their similar structures and physical properties. There was extensive emulsification in cultures grown with the C6–C12 alkanes. The phases could not be separated at all after incubation with hexane, heptane or octane and incomplete phase separation was obtained in cultures containing nonane, decane, undecane and dode-
Growth rate (% control)
140 120 100 80 60 40 20 0 4
4.5
5
5.5
6
6.5
7
7.5
8
Solvent log P Fig. 1. Effect of solvent log P value on growth of A. woodii and C. sporogenes. A. woodii (’) and C. sporogenes (K) were grown on 25 mM fructose and phenylalanine (10.5 mM), respectively, in the presence of a series of n-alkanes at a phase ratio of 0.2. The solvents were octane (log P ¼ 4.5), nonane (log P ¼ 5.1), decane (log P ¼ 5.6), undecane (log P ¼ 6.1), dodecane (log P ¼ 6.6), tridecane (log P ¼ 7.1) and tetradecane (log P ¼ 7.6 [32]). The cultures were mixed using an IKA shaker at 500 rpm. Growth rates shown as a percentage of the growth rate in control cultures grown in the absence of solvent (0.11 h1 for A. woodii and 0.2 h1 for C. sporogenes), and are the mean values for duplicate cultures. Growth rates are given as initial growth rates because emulsification of the solvent interfered with OD readings progressively as growth continued.
5 4 3 OD
58
2 1 0 0
5
10
15
20
Time (h)
Fig. 2. Growth of C. sporogenes in the presence and absence of tetradecane with different concentrations of phenylalanine C. sporogenes was grown with phenylalanine provided at 5.25 mM (K, J) or 10.5 mM (’, &), in the presence (K, ’) or absence (J, &) of tetradecane (phase ratio 0.179). The OD values are the averages for duplicate cultures.
cane. The efficiency of phase separation increased with the alkane chain length and complete phase separation was obtained after growth with tridecane or tetradecane. The solvent sensitivity of C. sporogenes was also tested, by measuring growth on phenylalanine (10.5 mM) in the presence of decane, dodecane or tetradecane at a phase ratio of 0.179, after degassing the solvents by nitrogen purging. The solvents had little effect on growth when the OD was p1, but we noticed that the OD readings became much higher than the control values as growth continued. This was due to extensive emulsification of all three of the solvents, which increased progressively after the OD increased above 1. Since the OD readings were accurate during the initial stages of growth, when there was little emulsification (see Fig. 2), it was possible to calculate the initial growth rates using these data. The growth rates in the presence of dodecane and tetradecane were slightly higher than in the absence of the solvents (Fig. 1), indicating that they were non-toxic, and possibly stimulatory. There was a lag phase in the presence of decane (not shown), indicating that there was some inhibition, but the growth rate thereafter was similar to the solvent-free control. 3.3. Emulsification of solvents in cultures of C. sporogenes As noted above, the solvents were emulsified in cultures of C. sporogenes during the later phases of growth, when the OD was X1. Therefore, the OD readings were much higher in cultures grown on 10.5 mM phenylalanine with tetradecane than in the control (Fig. 2). When C. sporogenes was grown with a lower phenylalanine concentration (5.25 mM) in the presence of tetradecane, the biomass concentration was correspondingly lower than in cultures grown with 10.5 mM phenylalanine (Fig. 2). With the lower phenylalanine concentration, there was very little emulsification of the solvent, and the OD values were
ARTICLE IN PRESS M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
similar to the solvent-free control. This confirms that growth of C. sporogenes was not affected by the anaerobic solvent. The emulsifying agent was produced by the cells, since there was no emulsification when tetradecane was incubated with uninoculated medium containing 5.25 or 10.5 mM phenylalanine. Since there was no emulsification in cultures grown with 5.25 mM phenylalanine or during the early stage of growth with 10.5 mM phenylalanine, we conclude that the cells can only produce the emulsifying agents once the biomass concentration reaches a certain critical level. 4. Conclusions Vacuum distillation, distillation under nitrogen and nitrogen purging were all effective methods for removing oxygen from tetradecane, since the degassed solvent had no effect on the growth or metabolism of A. woodii, whichever degassing method was used. We conclude that nitrogen purging is the method of choice for small-scale experiments, since it is much simpler than the distillation methods. This method seems to be applicable to other anaerobic biocatalysts, since growth of C. sporogenes was also unaffected by solvents prepared in this way, and we have also used this procedure with Peptostreptococcus productus (Enzyme Microb. Technol. in press). However, it would be necessary to use the distillation methods to prepare anaerobic solvents on a larger scale, since purging results in substantial solvent loss by evaporation. Tetradecane was non-toxic to A. woodii at all phase ratios tested, which indicates that this solvent would be a suitable choice for two-liquid phase biotransformations using this organism. Tridecane was also non-inhibitory, but growth was inhibited by n-alkanes with lower log Poctanol/water values (o7.1). C. sporogenes was more resistant to solvent polarity, since it could grow well with solvents with log P values as low as 5.6 (decane). However, both organisms seem to be more sensitive to solvents than aerobic bacteria, which generally tolerate solvents with log P values greater than 4–5 [16,32]. Solvent sensitivity in aerobic bacteria is partly related to cell surface composition, since Gram positive aerobes tend to be more solvent sensitive than Gram negative species [32,35–38]. A. woodii has a very unusual cell surface composition, and this may account for its extreme sensitivity to solvents. Thus, the cells are remarkably hydrophobic and show strong adherence to liquid hydrocarbons, particularly as the cells approach stationary phase [39]. Solvent toxicity depends on both molecular toxicity (general inhibition of cellular processes) and phase toxicity, which is due to interfacial effects [13,40]. The tendency of A. woodii to accumulate at the interface would exacerbate phase toxicity by increasing the contact between the microorganism and the solvent. In addition, solvent emulsification may have contributed to solvent toxicity, since stable emulsions were formed in cultures of A. woodii
59
with all solvents with log P values below 7.1, and both toxicity and the extent of emulsification increased as the log P value of the solvent decreased. Emulsification increases the interfacial area and this may also have contributed to phase toxicity. Therefore, it appears that the extreme sensitivity of A. woodii to the shorter chain alkanes may be due to a combination of adherence of the cells to the alkane phase and emulsification. Emulsification may also explain the increased sensitivity of C. sporogenes to solvents compared with aerobes. However, even the non-toxic solvents were emulsified by C. sporogenes cultures grown to high biomass concentration, but this did not have a significant effect on growth. Therefore, solvent emulsification does not seem to be a primary cause of growth inhibition, but may exacerbate other toxic effects, which remain to be identified. This study shows that A. woodii and C. sporogenes are more sensitive to solvent toxicity than aerobic bacteria. Nevertheless, non-polar solvents, such as tetradecane, are not toxic and are suitable for use in anaerobic, two-liquid phase reaction systems. Indeed, we have already begun to demonstrate the suitability of these solvents for reduction of C–C and C–N double bonds [3,6]. The solvents can be deoxygenated using any of the standard procedures used in organic chemistry. This work complements previous studies in which anaerobic biotransformations using relatively non-toxic substrates were intensified by using a water-miscible solvent system to deliver insoluble substrates [41]. Therefore, a range of anaerobic solvent systems is now available for use in biotransformations using anaerobic bacteria. Acknowledgement MFRM was supported by an EU Socrates scholarship. References [1] Simon H, Bader J, Guenther H, Neumann S, Thanos J. Chiral compounds via biocatalytic reductions. Angew Chem 1985;97: 541–55. [2] Tschech A, Pfennig N. Growth yield increase linked to caffeate reduction in Acetobacterium woodii. Arch Microbiol 1984;137:163–7. [3] Li H, Williams P, Micklefield J, Gardiner JM, Stephens G. A dynamic combinatorial screen for novel imine reductase activity. Tetrahedron 2004;60:753–8. [4] Davies ET, Stephens GM. Effect of growth substrate and electron donor on hydrogenation of carbon–carbon double bonds by Acetobacterium woodii. Enzyme Microbiol Technol 1998;23:129–32. [5] Davies ET, Stephens GM. Optimization of cell harvesting and assay procedures for reductive biotransformations in obligate anaerobes. Adv Bioprocess Eng 1994:495–9. [6] Foroughi F, Williams P, Stephens G. Reduction of carbon–carbon double bonds using Acetobacterium woodii: determination of the optimum inducer structure. Enzyme Microbiol Technol 2006;39: 1066–71. [7] Davies ET, Stephens GM. Efficient mediator-independent hydrogenation of carbon–carbon double bonds, using the obligate anaerobe, Acetobacterium woodii, with fructose as an electron donor. Appl Microbiol Biotechnol 1996;46:615–8.
ARTICLE IN PRESS 60
M.F. Rodriguez Martinez et al. / Anaerobe 14 (2008) 55–60
[8] Bache R, Pfennig N. Selective isolation of Acetobacterium woodii on methoxylated aromatic acids and determination of growth yields. Arch Microbiol 1981;130:255–61. [9] Kalil MS, Stephens GM. Catechol production by O-demethylation of 2-methoxyphenol using the obligate anaerobe, Acetobacterium woodii. Biotechnol Lett 1997;19:1165–8. [10] Dipeolu O, Gardiner J, Stephens G. Biocatalytic amide reduction using Clostridium sporogenes. Biotechnol Lett 2005;27:1803–7. [11] Schales O, Graefe HA. Arylnitroalkenes: a new group of antibacterial agents. J Am Chem Soc 1952;74:4486–90. [12] Salter GJ, Kell DB. Solvent selection for whole cell biotransformations in organic media. Crit Rev Biotechnol 1995;15:139–77. [13] Leon R, Fernandes P, Pinheiro HM, Cabral JMS. Whole-cell biocatalysis in organic media. Enzyme Microbiol Technol 1998;23:483–500. [14] Dervakos GA, Woodley JM, Keravnou ET, Washbrook J, Lilly MD. Development of a KBS [knowledge-based system] for biotransformation process design. AIChE Symp Ser 1989;114:283–91. [15] Schmid A, Kollmer A, Mathys RG, Witholt B. Developments toward large-scale bacterial bioprocesses in the presence of bulk amounts of organic solvents. Extremophiles 1998;2:249–56. [16] Laane C, Boeren S, Vos K, Veeger C. Rules for optimization of biocatalysis in organic solvents. Biotechnol Bioeng 1987. [17] Aono R, Tsukagoshi N, Miyamoto T. Evaluation of the growth and inhibition strength of hydrocarbon solvents against Escherichia coli and Pseudomonas putida grown in a two-liquid phase culture system consisting of a medium and organic solvent. Extremophiles 2001;5: 11–5. [18] Roffler SR, Blanch HW, Wilke CR. In situ recovery of butanol during fermentation. Part 2: fed-batch extractive fermentation. Bioprocess Eng 1987;2:181–90. [19] Playne MJ, Smith BR. Toxicity of organic extraction reagents to anaerobic bacteria. Biotechnol Bioeng 1983;25:1251–65. [20] Datta R. Acidogenic fermentation of corn stover. Biotechnol Bioeng 1981;23:61–77. [21] Setti L, Rossi M, Lanzarini G, Pifferi PG. Barrier and carrier effects of n-dodecane on the anaerobic degradation of benzothiophene by Desulfovibrio desulfuricans. Biotechnol Lett 1993;15:527–30. [22] Roffler SR, Blanch HW, Wilke CR. In situ recovery of fermentation products. Trends Biotechnol 1984;2:129–36. [23] Roffler SR, Blanch HW, Wilke CR. In situ recovery of butanol during fermentations. Part 1: batch extractive fermentations. Bioprocess Eng 1987;2:1–12. [24] O’Brien RW, Morris JG. Oxygen and the growth and metabolism of Clostridium acetobutylicum. J Gen Microbiol 1971;68:307–18. [25] Hewitt J, Morris JG. Superoxide dismutase in some obligately anaerobic bacteria. FEBS Lett 1975;50:315–8.
[26] Whittenbury R. Hydrogen peroxide formation and catalase activity in the lactic acid bacteria. J Gen Microbiol 1964;35:13–26. [27] Borch E, Molin G. The aerobic growth and product formation of Lactobacillus, Leuconostoc, Brochothrix and Carnobacterium in batch cultures. Appl Microbiol Biotechnol 1989;30:81–8. [28] Hansen TA. Metabolism of sulfate-reducing prokaryotes. Antonie van Leeuwenhoek J Microbiol Serol 1994;66:165–85. [29] Aeckersberg F, Bak F, Widdel F. Anaerobic oxidation of saturated hydrocarbons to CO2 by a new type of sulfate-reducing bacterium. Arch Microbiol 1991;156:5–14. [30] Shriver DF, Drezdzon MA. The manipulation of air-sensitive compounds. 2nd ed. New York: Wiley; 1986. [31] Giesel H, Simon H. On the occurrence of enoate reductase and 2-oxocarboxylate reductase in clostridia and some observations on the amino acid fermentation by Peptostreptococcus anaerobius. Arch Microbiol 1983;135:51–7. [32] Vermue M, Sikkema J, Verheul A, Bakker R, Tramper J. Toxicity of homologous series of organic solvents for the Gram-positive bacteria Arthrobacter and Nocardia sp. and the Gram-negative bacteria Acinetobacter and Pseudomonas sp. Biotechnol Bioeng 1993;42: 747–58. [33] Phumathon P, Stephens GM. Production of toluene cis-glycol using recombinant Escherichia coli strains in glucose-limited fed batch culture. Enzyme Microbiol Technol 1999;25:810–9. [34] Osborne SJ, Leaver J, Turner MK, Dunnill P. Correlation of biocatalytic activity in an organic–aqueous two-liquid phase system with solvent concentration in the cell membrane. Enzyme Microbiol Technol 1990;12:281–91. [35] Collins LD, Daugulis AJ. Characterization and optimization of a two-phase partitioning bioreactor for the biodegradation of phenol. Appl Microbiol Biotechnol 1997;48:18–22. [36] Harrop AJ, Hocknull MD, Lilly MD. Biotransformations in organic solvents: a difference between Gram-positive and Gram-negative bacteria. Biotechnol Lett 1989;11:807–10. [37] Inoue A, Horikoshi K. Estimation of solvent-tolerance of bacteria by the solvent parameter log P. J Ferment Bioeng 1991;71:194–6. [38] Rajagopal AN. Growth of Gram-negative bacteria in the presence of organic solvents. Enzyme Microbiol Technol 1996;19:606–13. [39] Donlon E, Colleran A. A comparison of different methods to determine the hydrophobicity of acetogenic bacteria. J Microbiol Methods 1993;17:27–37. [40] Bar R. Phase toxicity in multiphase biocatalysis. Trends Biotechnol 1986;4:167. [41] Tidswell EC, Salter GJ, Kell DB, Morris JG. Enantioselectivity of sulcatone reduction by some anaerobic bacteria. Enzyme Microbiol Technol 1997;21:143–7.