Continuous production of n-butanol by Clostridium pasteurianum DSM 525 using suspended and surface-immobilized cells

Continuous production of n-butanol by Clostridium pasteurianum DSM 525 using suspended and surface-immobilized cells

Journal of Biotechnology 216 (2015) 29–35 Contents lists available at ScienceDirect Journal of Biotechnology journal homepage: www.elsevier.com/loca...

1MB Sizes 0 Downloads 19 Views

Journal of Biotechnology 216 (2015) 29–35

Contents lists available at ScienceDirect

Journal of Biotechnology journal homepage: www.elsevier.com/locate/jbiotec

Continuous production of n-butanol by Clostridium pasteurianum DSM 525 using suspended and surface-immobilized cells Alessandro Gallazzi a,b,c , Barbora Branska a , Flavia Marinelli b,c , Petra Patakova a,∗ a b c

Department of Biotechnology, University of Chemistry and Technology, Prague, Czech Republic Department of Biotechnology and Life Sciences, University of Insubria, Varese, Italy ¨ esearch Center, Politecnico of Milano, ICRM CNR Milano and University of Insubria, Varese, Italy ¨he Protein FactoryR T

a r t i c l e

i n f o

Article history: Received 27 April 2015 Received in revised form 6 October 2015 Accepted 9 October 2015 Keywords: Clostridium pasteurianum Glycerol fermentation n-Butanol Flow cytometry

a b s t r a c t For n-butanol production by Clostridium pasteurianum DSM 525, a modified reinforced Clostridium medium was used, where glucose was alternated with glycerol and two kinds of continuous fermentation were tested using suspended and surface immobilized cells on corn stover pieces. A steady state, with butanol productivity of 4.2 g/L h, was reached during the packed-bed continuous fermentation at a dilution rate of 0.44 h−1 . The average n-butanol concentration, yield and the ratio of n-butanol/liquid by-products were 10.4 g/L, 33 % and 2.5, respectively. Unexpectedly, during continuous fermentation with suspended cells, at a dilution rate of 0.01 h−1 , steady-state was not achieved and regular oscillations occurred in all measured variables, i.e. concentrations of glycerol, products and the number of cells stained with the fluorescent dyes carboxy fluorescein diacetate and propidium iodide. A possible explanation for oscillatory/steady-state behavior of suspended/surface-attached cells, respectively, may be specific butanol toxicity (toxicity per cell), which was higher/lower in respective cases, and which might be caused by lower/higher cell numbers respectively in both systems. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Fluctuating crude oil prices, unstable political situations in countries exporting the oil and a growing consciousness of environmental problems, such as global warming, attributed to the use of fossil fuels, has brought significant attention to the production of biofuels from biomass. In Europe, 20-20-20 targets, which aim, inter alia, to cover 20 % of energy consumption from renewable sources by the year 2020, resulted in increased production of biodiesel and bioethanol. Mostly, biodiesel is produced from triglycerides present in various plant oils, animal fats or waste oils, by transesterification with methanol using acidic, alkaline or enzymatic catalysts (Demirbas¸, 2003); this results in the formation of glycerol as an abundant by-product and, in general, lowers the refined glycerol world market price. However the actual glycerol price also depends on the price of oil plants (for recent fluctuations of refined glycerol price and the impact of biodiesel production on glycerol market, see Ciriminna et al., 2014). Nevertheless, glycerol has become an attractive fermentable substrate due to its availability, low cost and high

∗ Corresponding author at: Department of Biotechnology, University of Chemistry and Technology, Technicka 5, CZ 16628 Prague 6, Czech Republic. E-mail address: [email protected] (P. Patakova). http://dx.doi.org/10.1016/j.jbiotec.2015.10.008 0168-1656/© 2015 Elsevier B.V. All rights reserved.

degree of reduction. Use of glycerol can be also considered a model example of functioning biorefinery concept idea where biodiesel waste is consumed in the following production of other valuable compound, butanol. Biobutanol, a promising biofuel candidate, can be produced by acetone–butanol–ethanol (ABE) fermentation by selected species of the genus Clostridium. Butanol can be blended both with gasoline and diesel and due to several advantages over ethanol, such as higher energy content, lower water miscibility and lower corrosivity, has attracted considerable attention in recent years (Dürre, 2008). The strain Clostridium pasteurianum DSM 525 differs from the most well-known Clostridium acetobutylicum ATCC 824 strain in its fermentation pattern and can ferment glycerol to n-butanol and 1,3-propanediol as main products together with ethanol, butyric and lactic acids, hydrogen and carbon dioxide as by-products (Dabrock et al., 1992; Biebl, 2011). This or closely related strains deposited in other culture collections i.e., ATCC 6013 and MTCC 116 have already been used for glycerol fermentation (Dabrock et al., 1992; Biebl, 2011; Moon et al., 2011; Jensen et al., 2012; Khanna et al., 2013; Gallardo et al., 2014) but not using packed-bed continuous fermentation with corn stover pieces as immobilization support material. This work is focused on a comparison of continuous n-butanol production from glycerol by C. pasteurianum DSM 525 using suspended and surface (on corn stover) immobi-

30

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35

lized cells. In addition, double fluorescent staining with propidium iodide (PI) and carboxy fluorescein diacetate (CFDA), followed by flow cytometry (FC), were used to monitor the physiological state of cells during fermentation. To the best of our knowledge, neither a comparison of different types of continuous fermentation nor flow cytometric data have been used previously for this Clostridium strain. 2. Materials and methods 2.1. Microorganism and culture conditions The strain C. pasteurianum DSM 525 was stored frozen at −70 ◦ C and an inoculum was grown in the modified reinforced clostridium medium (RCM) (containing in g/L: tryptone (Merck, Germany) 10, meat extract (Merck, Germany) 10, yeast extract (Merck, Germany) 3, NaCl 5, sodium acetate 3, glucose 15), at 37 ◦ C in an anaerobic chamber (Sony Technology Centre, Glamorgam, UK) for 18–24 h in Erlenmeyer flasks.

size. A laboratory pH meter (WTW, Germany) was used for online pH measurements in a through-flow cell at the bioreactor outlet. The feed rate was calculated from weight loss, as measured by a laboratory balance (KERN, Germany). The bioreactor was filled with immobilization material and 200 mL of the RCM culture medium with glycerol as a carbon source prior to sterilization, and was then sterilized in an autoclave at 121 ◦ C for 20 min. After cooling, the bioreactor was placed in a sterile inoculation box and inoculated with 20 mL of the seed culture. After inoculation, the cell culture was grown batch-wise for 18 h, with an initial glycerol concentration of 54 g/L, and with recirculation of the medium allowing colonization of the bacteria on the corn stover. The steady-state working volume was kept constant at 180 mL and the medium feed was set to reach the required dilution rate (D). Dilution rates adopted were: 0.13 h−1 during the first 140 h and after that, 0.44 h−1 until the end of the fermentation. The concentrations of glycerol in the feeding solutions were: 45 g/L for D = 0.13 h−1 and 35 g/L for D = 0.44 h−1 . 2.4. Analyses

2.2. Batch and continuous production with suspended culture Multifors (Infors HT, Switzerland) bioreactors were used for batch and continuous fermentations with suspended cells. Both batch and steady-state working volumes in the bioreactor were kept at 400 mL and cultivation parameters were maintained at 200 rpm and 37 ◦ C. Bioreactors with suspended cultures were run in duplicate. The carbon source, originally glucose in the modified RCM, was replaced with pure glycerol (Penta, Czech Republic) and its initial concentration in the medium was 40 or 45 g/L. Prior to inoculation, the bioreactors were bubbled with CO2 for 10 min to establish an anaerobic atmosphere and the pH was adjusted to 6.6–7.0. During fermentation, pH was not regulated. After inoculation (ratio 1:10), the cultivation was operated in batch mode for 18 h followed by continuous fermentation wherein regular samplings were carried out. Dilution rate during the continuous period was kept constant, at 0.01 h−1 , and the glycerol concentration in the feed medium was 60 g/L. 2.3. Continuous production with the culture immobilized on corn stover Immobilized cultivations were carried out in a glass, in-house made, jacketed, packed-bed bioreactor (250 mL total volume and 200 mL max. working volume) heated by water to 37 ◦ C. A schematic of the bioreactor system is shown in Fig. 1. The bioreactor was filled with cut corn stover pieces of 1 cm3 approximate

Growth was followed as an optical density (OD) measured at 600 nm using a spectrophotometer (Varian, Spain). Concentrations of carbon source, glycerol, and products (nbutanol, 1,3 propanediol, lactic acid, butyric acid and ethanol) were determined by HPLC analysis (Agilent Series 1200HPLC; Agilent, Spain) using a Polymer IEX H+ column (Watrex, Czech Republic) heated to 60 ◦ C with a mobile phase of 5 mM H2 SO4 (flow rate 0.5 mL/min) and equipped with refractive index detection. 2.5. Flow cytometric analysis and fluorescent staining Cell fluorescence was monitored by microscopic observation as described previously (Linhova et al., 2010) and by flow cytometric analysis. FC analyses were conducted after fluorescent staining with two probes, CFDA and PI; both purchased from Sigma Aldrich. The samples were filtered through a 30 ␮m filter, centrifuged (2000 × g, 3 min), washed twice in sterile 0.8% (w/v) NaCl and diluted in the same solution to an OD of 0.20 ± 0.01. The fluorescent probes, PI and CFDA, were added to the cell suspension to a final concentration of 10.0 ␮g/mL. The cultures were incubated in the dark, at room temperature, for 10 min and subsequently analyzed. Stained suspensions were analyzed with an Accuri C6 cytometer (BD Accuri Cytometer Inc., USA) equipped with an argon laser (488 nm). FSC and SSC signals were used as a trigger, and green (FL1; 515–565 nm) and red (FL3; >605 nm) fluorescence were used to measure the percentage of PI or CFDA stained cells. The sample flow rate was set to 10 ␮L/min for all experiments. 2.6. Electron microscopy SEM microphotographs of the material (corn stover) with attached cells were performed using SEM Hitachi S-4700 (Hitachi, Japan) device. Prior the microscopy, the samples were mounted using double stick carbon tape, dried in thermostat at 50 ◦ C, and coated by Au/Pd (approx. 15 nm). 3. Theory/calculation

Fig. 1. Schema of the packed-bed bioreactor system.

From an industrial point of view, the preferred method for producing n-butanol is continuous production. However, common drawbacks of continuous production with solventogenic clostridia are culture instability, strain degeneration and alternating growth and sporulation phases that cause a decrease or loss of solvent production. This behavior has been studied in detail for C. acetobutylicum ATCC 824 (Clarke et al., 1988). However in the case of C.

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35 Table 1 Parameters of the batch process. Initial glyce1rol concentration Final glycerol concentration Batch fermentation time Glycerol consumption rate in exponential growth phase Specific growth rate in exponential phase Final n-butanol concentration Butanol yield from consumed glycerol Total liquid products yield from consumed glycerol Butanol productivity in production perioda Ratio n-butanol/by-products

(g/L) (g/L) (h) (g/L h) (h−1 ) (g/L) (%) (%) (g/L h)

45 5 20 2 0.26 9.0 22 48 0.7 0.9

Average data from two fermentations are presented. a Butanol was produced in exponential and stationary growth phases for 12 h.

pasteurianum DSM 525, sporulation has not been studied in depth and correlated with n-butanol production. The main goal of the present work was to compare the culture behavior and n-butanol production in continuous cultivation using suspended and immobilized cells. Continuous production with immobilized cells offers an option to implement a dilution rate higher than the maximum growth rate, with the possibility of achieving higher productivity without risk of cell wash-out from the system. Flow cytometry is a technique enabling analysis of heterogeneous populations at the level of individual cells, discriminating between viable and non-viable cells if combined with fluorescent staining (Müller and Nebe von Caron, 2010). Flow cytometry has been already used for assessment of C. acetobutylicum ATCC 824 ˜ et al., 2015; Tracy et al., 2008, 2010). populations (González-Penas Among six different fluorescent probes tested for viability estimation of C. pasteurianum NRRL B-598 and Clostridium beijerinckii CCM 6218 (Linhova et al., 2012), only propidium iodide reflected viability changes for both species. Subsequently, double fluorescent staining with a combination of fluorescent dyes based on different principles (i.e., PI staining cells with impaired cell integrity and CFDA staining cells exhibiting esterase activity) was demonstrated to be an effective tool for estimating population viability in C. beijerinckii CCM 6218 and Clostridium tetanomorphum DSM 4473 (Patakova et al., 2014). The same method, based on double fluorescent staining and flow cytometry, was applied for viability assessment of C. pasteurianum DSM 525 cells in this paper. 4. Results Growth and production characteristics were calculated from batch fermentation parameters, and specific data are summarized in Table 1. Continuous cultivations with suspended and immobilized cells have been carried out in order to optimize process conditions using different selection criteria such as process productivity, yield, concentration of n-butanol and total amount of by-products. During the continuous fermentations, viability of the Clostridium cells was followed by double fluorescent staining and FC analysis. 4.1. Continuous cultivation with suspended culture Based on batch cultivation (Table 1), the glycerol consumption rate in exponential growth phase was calculated to be 2 g/L h. Continuous production was then carried out with a constant dilution rate of 0.01 h−1 and a glycerol concentration in the feeding solution of 60 g/L, see Fig. 2a, b. During the initial batch period of cultivation (duration 18 h), there was a gradual decrease in pH, associated mainly with butyric acid production (see Fig. 2a, b). Amplitudes of butyric acid concentration oscillations coincided with pH changes. Furthermore, there was an association between butyrate production and growth of the production organism, and about 20 h after the maximum butyrate concentration, the maximum n-butanol

31

concentration was achieved. Total by-products, i.e., sum of butyric acid, lactic acid, 1,3-propanediol, and ethanol followed the trend of butyric acid concentration and pH oscillation, corresponding with the fact that the predominant component (60% or more) of total by-products comprised butyric and lactic acids (Fig. 2a). 18 h after inoculation (start of continuous cultivation), culture optical density reached 8.6 with 9% PI and 89 % CFDA stained cells (see Fig. 2b). An association between trends of the two types of fluorescent viability staining (PI versus CFDA) is clear, with opposite amplitudes for PI and CFDA. This correlates with optical density (a measure of population growth) that reaches maxima at the point of highest viability (maximum CFDA and minimum PI stained cells). During fermentation with suspended cells, steady state was not achieved because of oscillatory behavior in the cell culture, which was distinct in terms of cell density, n-butanol, glycerol and butyric acid concentrations. From a careful observation of the oscillatory behavior, it was possible to identify a periodic cycle, each of approximately 75 h (3 days). No correlation between sporulation and oscillatory behavior was found by microscopic examination of samples; only a few spores were observed during the whole cultivation and their occurrence was random i.e. not associated with any specific event such as a metabolic switch. When continuous cultivation was repeated with higher dilution rates (0.05 and 0.07 h−1 ) comparable oscillatory behavior was observed (data not shown). In addition, fermentation performance of the bioreactor with suspended cells did not reach the assumed level (set according to the exponential phase of batch fermentation) as shown by a comparison of estimated (2 g/L h) and actual (0.25 g/L h) glycerol consumption rates (see Table 2). 4.2. Continuous production with corn stover immobilization During continuous fermentation in a packed-bed bioreactor, two dilution rates were tested and two steady-states exhibiting almost constant values for all variables were reached. Fig. 3a, b shows the results collected during continuous production in packed-bed bioreactor. After the first 18 h of batch growth we observed a consumption of 16 g/L of glycerol and a production of 2 g/L of n-butanol, showing slower adaptation and growth when compared with a suspended culture. After 18 h, the highest production of butyric acid, 2.6 g/L, was associated with a decrease in pH to 5.3. Both dilution rates, 0.13 and 0.44 h−1 revealed a transient state of about 52 h prior to achievement of steady state. At a dilution rate of 0.13 h−1 , pH stabilized at 5.9, which correlated with the stabilization of butyric acid concentration at 0.45 g/L and n-butanol production reached a maximum of 10.8 g/L and an average of 10.5 g/L (see Fig. 3a, b). During the first steady-state, the average unconsumed glycerol concentration was 16.4 g/L and an average total by-products concentration was 4.5 g/L, giving a ratio n-butanol/by-products of 2.3 (see Table 2). Changing the dilution rate to 0.44 h−1 , after 28 h, there was a visible increase in butyric acid concentration to 1.1 g/L associated with a decrease in pH to 5.6 and a decrease in n-butanol and ethanol concentrations, to 7.1 g/L and 0.7 g/L, respectively (see Fig. 3a). Similarly, as in case of a lower dilution rate, we observed a stabilization of butyric acid concentration and pH at values of 0.63 g/L and 5.72, respectively, after 52 h. Furthermore, n-butanol production reached a maximum of 10.4 g/L and an average of 9.5 g/L. During the second steady-state, the average unconsumed glycerol concentration was 6.2 g/L and the average total by-products concentration was 3.9 g/L, giving a n-butanol/by-products ratio of 2.5 (see Table 2). Optical density, together with fluorescence parameters, were measured in the effluent stream and therefore gave information predominantly about suspended cells released from the biofilm. After 18 h from inoculation (start of continuous cultivation), optical density reached 3.8 with 12 % of PI and 88 % of CFDA stained

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35

50

10

a

D = 0.01 h-1

Glycerol (g/l)

9 8

40

7

Products (g/l)

32

6

30

5 4

20

3 2

10

1 0

0 50

100

150

200

Glycerol (g/l) Butyric acid (g/l)

Butanol (g/l) Total by-products (g/l)

300 Time (h)

1b

100 Number of stained cells (%)

250

90

D = 0.01

9

b

h-1

8

80 70

7

60

6

pH, OD

0

50 40

5

30

4

20 3

10 0

2

0

50

100

150

200

250

300 Time (h)

PI (%)

CFDA(%)

OD

pH

Fig. 2. (a, b) Continuous fermentation with suspended cells. A vertical line marks the start of continuous cultivation. In all cases, average values from two replicates are presented and standard deviations did not exceed 5 %.

Table 2 Overview of fermentation parameters. Type of continuous fermentation −1

Dilution rate (h ) Feed glycerol concentration (g/L) Average unconsumed glycerol concentration (g/L) Average glycerol consumption rate (g/L h) Maximum n-butanol concentration (g/L) Average n-butanol concentration (g/L) Average total by-products concentration (g/L) Butanol productivity (g/L h) Total by-products productivity (g/L h) Ratio n-butanol/by-products Average n-butanol yield from consumed glycerol (%) Average total liquid products yield (%)

Suspended cells

Immobilized cells

Immobilized cells

0.01 60.0 35.0 0.25 8.6 7.6 2.5 0.1 0.03 3.0 40 52

0.13 45.0 16.4 3.7 10.8 10.1 4.5 1.3 0.6 2.3 35 51

0.44 35.0 6.2 12.7 10.4 9.5 3.9 4.2 1.7 2.5 33 47

cells (see Fig. 3b). During the first phase of continuous fermentation, at D = 0.13 h−1 , there was a decrease in active cells to a value of 56%, associated with an increase in dead cells to a value of 31 %; the optical density of released cells stabilized at 1.9. As in the fermentation with suspended cells, spores were observed only randomly and rarely. With the change in dilution rate to 0.44 h−1 , there was an increase in CDFA positive cells to 85%, and a correlated decrease in PI stained cells to 12 %. The change in viability in the population after dilution rate change, was associated with an increase in optical density to a value of 6. We assume that the positive increment in cell viability was associated with a higher rate of cell detachment and cell proliferation, allowing achievement of a higher cell density, together with a transient decrease in the concentration of toxic products, particularly n-butanol and ethanol. After approximately

4 days from the change in dilution rate, the proportion of stained cells stabilized (see Fig. 3b); about 50% PI positive cells and 40 % CFDA positive cells were constant during the second steady state. During steady state in both types of cultivation, the sum of stained cells (PI + CFDA) was less than 100 %; this was predominantly due to the presence of debris from dead cells. A proportion of non-stained particles also represent cells with insufficient metabolic activity to produce adequate fluorescent carboxy fluorescein but still maintain membrane integrity, therefore they are neither stained by PI or CFDA. Distribution of fluorescence in such populations is represented by four distinct subpopulations with different staining patterns (see Fig. 4). Exponential phases and transient states with high rates of cell proliferation contain very low amounts of inactive cells and dead cell residues, thus the sum of PI and CFDA stained cells approaches 100%.

12

a

50

10 40 8

D = 0.44 h-1

D = 0.13 h-1 30

6

20

4

10

2 0

0 0

50

100

150

Glycerol (g/l) Total by-products (g/l)

Number of stained cells (%)

Products (g/l)

33

200

250

300 Time (h)

Butyric acid (g/l) Butanol (g/l)

b

100

D = 0.13

90

h-1

D = 0.44

8

h-1 7

80

pH, OD

Glycerol (g/l)

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35

6

70

60

5

50

4

40

3

30

2

20

1

10 0

0 0

50 PI (%)

100

150 CFDA (%)

200 OD

250 pH

300 Time (h)

Fig. 3. (a, b) Continuous fermentation with immobilized cells. Vertical lines mark the start of continuous cultivation and change of dilution rate. In all cases, average values from two replicates are presented and standard deviations did not exceed 5 %.

A sample of corn stover was withdrawn from continuously growing culture for examination of cell colonization of the material surface using scanning electron microscopy. The carrier material is shown in Fig. 5A, while bacterial cells attached to the surface are shown in Fig. 5B. 4.3. Suspended versus immobilized cells in continuous fermentation A comparison of selected fermentation parameters achieved during different kinds of continuous fermentation is provided in Table 2. The observed high production of n-butanol, above the maximum growth rate obtained in batch fermentation, (0.26 h−1 ) confirms the possibility to operate with higher dilution rates, allowing the achievement of high productivity 4.2 g/L h when compared with a suspended culture where a productivity of only 0.1 g/L h was reached. 5. Discussion

Fig. 4. FC dot-plot diagram of continuous culture of Clostridium pasteurianum DSM 525 from a packed bed bioreactor (D = 0.13 h−1 ), double stained with CFDA and PI: upper left—dead i.e. red stained cells; lower right—vital i.e. green stained cells; upper right—doublets or injured double stained cells and lower left—non-stained particles.

Surprisingly, it was not possible to attain steady state during fermentation with suspended cells but instead, periodic, quite regular oscillations were observed. Neither Biebl (2011) nor any other author working with the same strain have described such a phenomenon. Similar oscillatory behavior during continuous cultivation has not been observed for C. pasteurianum but it was described (Clarke et al., 1988) for C. acetobutylicum P2 cultivated in a chemostat using a suspended cell culture. For this strain, oscillations were associated with alternating growth and sporulation phases, which resulted in changes in growth rate, substrate

34

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35

Fig. 5. (A, B) SEM photographs of the material (corn stover) surface structure (A) and clostridia attached on the surface of the material (B) during the continuous cultivation.

consumption and production of both acids and solvents. As in the case of C. pasteurianum DSM 525, negligible sporulation was observed,the nature of the observed oscillations might be caused by a different phenomenon, most probably n-butanol toxicity. Oscillations caused by product toxicity were described by Mulchandani and Volesky (1994) for C. acetobutylicum ATCC 824. In this case, the explanation lies in periodic accumulation and washing-out of n-butanol, which may induce a periodical decrease and increase in culture growth, viability, substrate consumption and product formation. This assumption corresponds well with our findings, where culture viability follows the oscillation of solvents with the typical pattern showing a decrease in viability after reaching the maximal concentration of n-butanol. An interdependance of PI and CFDA staining for estimation of population viability, mainly during exponential growth phase, has already been proven for different solventogenic species i.e., C. beijerinckii CCM 6218 and C. tetanomorphum DSM 4473 (Patakova et al., 2014). The present findings confirm that this type of double fluorescent staining reflects culture conditions, as in the case of the continual cultivation of C. pasteurianum DSM 525. Immobilization of the cells on the surface of corn stover pieces resulted in better maximum and average n-butanol concentrations and productivity, namely 6 times higher productivity for D = 0.44 h−1 compared to batch cultivation and 42 times in comparison with cultivation of suspended cells, while maintaining similar yields of liquid products of approx. 50 %. This could be

partly explained by the fact that clostridia are originally soil bacteria that prefer an “attached way of life” rather than “swimming in a medium”, together with probably a higher cell concentration in a packed-bed bioreactor. The higher cell density may result in a lower specific butanol production (butanol production per cell) and thus mitigate the stress effect. This probable explanation can be also supported by an absence of the oscillatory phenomenon, which may correlate with a non-inhibiting internal butanol concentration in cells attached to the corn stover pieces. He and Chen (2013) also observed an improvement in butanol production by C. acetobutylicum ATCC 824 attached to treated wheat straw pieces, in comparison with suspended cells, and explained this phenomenon by partial adsorption of solvents onto the support, together with an increased cell concentration in the bioreactor, possibly alleviating butanol inhibition. A beneficial impact of supporting material such as bonechar or clay brick on butanol production was also mentioned in the comprehensive review by Qureshi et al. (2005). Although the flow cytometric data demonstrated a slightly lower viability of cells in the immobilized system, it must be taken into account that only detached cells are analyzed in the effluent stream; nevertheless, they can still provide valuable information about behavior of the whole system. An increase in dilution rate resulted in a higher percentage of dead cells (50 % compared to 32 %) in the product stream but at the same time, a significant increase in optical density. This means there was a much higher number of live cells in total, which is in agreement with higher glycerol consumption and production rates. To compare our results with previously published data is difficult even for studies using the same strain i.e., C. pasteurianum DSM 525 (or C. pasteurianum ATCC 6013 which should be identical or closely related) because neither experiment was performed under the same cultivation conditions. A number of authors (Taconi et al., 2009; Jensen et al., 2012; Khanna et al., 2013; Gallardo et al., 2014) used “crude” glycerol obtained from different steps of biodiesel production. However, “crude” glycerol obtained immediately after transesterification cannot be used for fermentation because of its toxicity caused by high methanol (23–37 wt%) and salt (up to 5 wt%) concentrations (Ciriminna et al., 2014). Therefore methanol is usually removed by distillation (an inexpensive operation) but the crude glycerol still contains traces of methanol, free fatty acids and salts which may inhibit fermentation. As it is difficult and expensive to remove these compounds, differently treated, crude glycerol is often used for fermentation. However, fermentation parameters achieved with crude glycerol were mostly lower than those obtained using pure glycerol (Taconi et al., 2009; Jensen et al., 2012; Khanna et al., 2013; Gallardo et al., 2014). The only study, in which C. pasteurianum MTCC 116 (this strain should be identical to strain ATCC 6013) was immobilized on Amberlite using cell cross-linking with glutaraldehyde, was published by Khanna et al. (2013). The published results are in agreement with our data that modified RCM medium, in which glucose was alternated with glycerol, was the best choice for n-butanol production. The highest n-butanol productivity, 7.8 g/L h, using glycerol as a substrate, was reached with a mutant strain derived from C. pasteurianum ATCC 6103 during high cell density continuous fermentation (Malaviya et al., 2012). Although productivity achieved in our study was about half of this value, 4.2 g/L h, the n-butanol concentration and yield were higher than the values published by Malaviya et al., (2012).

6. Conclusion Results presented here unambiguously demonstrate that using packed-bed bioreactors for continuous production of n-butanol by

A. Gallazzi et al. / Journal of Biotechnology 216 (2015) 29–35

C. pasteurianum has great potential; so-far the highest n-butanol productivity reached with this unmodified strain was 4.2 g/L h. Further, the combination of fluorescent dyes, CFDA and PI, functioning on different principles, proved to be a useful tool for evaluation of population viability, also for the first time in this strain. However, there are still some gaps in our knowledge of C. pasteurianum DSM 525; regulation of fermentation pathways is still unclear and the signal necessary to initiate sporulation, unlike the pH drop for C. acetobutylicum ATCC 824, is unknown. Acknowledgement The work was supported by the project BIORAF No. TE 01020080 awarded by the Technological Agency of the Czech Republic. References Biebl, H., 2011. Fermentation of glycerol by Clostridium pasteurianum, batch and continuous culture studies. J. Ind. Microbiol. Biotechnol. 27, 18–26. Ciriminna, R., Pina, C.D., Rossi, M., Pagliaro, M., 2014. Understanding the glycerol market. Eur. J. Lipid Sci. 116, 1–8. Clarke, K.G., Hansford, G.S., David, T.J., 1988. Nature and significance of oscillatory behavior during solvent production by Clostridium acetobutylicum in continuous culture. Biotechnol. Bioeng. 32, 538–544. Dabrock, B., Bahl, H., Gottschalk, G., 1992. Parameters affecting solvent production by Clostridium pasteurianum. Appl. Environ. Microbiol. 58, 1233–1239. Demirbas¸, A., 2003. Biodiesel fuels from vegetable oils via catalytic and non-catalytic supercritical alcohol transesterifications and other methods: a survey. Energy Convers. Manage. 44, 2093–2109. Dürre, P., 2008. Fermentative butanol production: bulk chemical and biofuel. Ann. N. Y. Acad. Sci. 1125, 353–362. Gallardo, R., Alves, M., Rodrigues, L.R., 2014. Modulation of crude glycerol fermentation by Clostridium pasteurianum DSM 525 towards the production of butanol. Biomass Bioenergy 71, 134–143. ˜ González-Penas, H., Lu-Chau, T.A., Moreira, M.T., Lema, J.M., 2015. Assesment of morphological changes of Clostridium acetobutylicum by flow cytometry during acetone/butanol/ethanol extractive fermentation. Biotechnol. Lett. 37, 577–584.

35

He, Q., Chen, H.Z., 2013. Improved efficiency of butanol production by absorbent fermentation with a renewable carrier. Biotechnol. Biofuels 6, 121. Jensen, T.O., Kvist, T., Mikkelsen, M.J., Christensen, P.V., Westermann, P., 2012. Fermentation of crude glycerol from biodiesel production by Clostridium pasteurianum. J. Ind. Microbiol. Biotechnol. 39, 709–717. Khanna, S., Goyal, A., Moholkar, V.S., 2013. Production of n-butanol from biodiesel derived crude glycerol using Clostridium pasteurianum immobilized on Amberlite. Fuel 112, 557–561. Linhova, M., Patakova, P., Lipovsky, J., Fribert, P., Paulova, L., Rychtera, M., Melzoch, K., 2010. Development of flow cytometry technique for detection of thinning of peptidoglycan layer as a result of solvent production by Clostridium pasteurianum. Folia Microbiol. 55, 340–344. Linhova, M., Branska, B., Patakova, P., Lipovsky, J., Fribert, P., Rychtera, M., Melzoch, K., 2012. Rapid flow cytometric method for viability determination of solventogenic clostridia. Folia Microbiol. 57, 307–311. Malaviya, A., Jang, Y.S., Lee, S.Y., 2012. Contiunous butanol production with reduced byproducts formation from glycerol by a hyper producing mutant Clostridium pasteurianum. Appl. Microbiol. Biotechnol. 93, 1485–1494. Moon, C., Lee, C.H., Sang, B.-I., Um, Y., 2011. Optimization of medium compositions favoring butanol and 1,3-propandiol production from glycerol by Clostridium pasteurianum. Bioresour. Technol. 102, 10561–10568. Mulchandani, A., Volesky, B., 1994. Production of acetone–butanol–ethanol by Clostridium acetobutylicum using a spin filter perfusion bioreactor. J. Biotechnol. 34, 51–60. Müller, S., Nebe von Caron, G., 2010. Functional single-cell analyses: flow cytometry and cell sorting of microbial populations and communities. FEMS Microbiol. Rev. 34, 554–587. Patakova, P., Linhova, M., Vykydalova, P., Branska, B., Rychtera, M., Melzoch, K., 2014. Use of fluorescent staining and flow cytometry for monitoring physiological changes in solventogenic clostridia. Anaerobe 29, 113–117. Qureshi, N., Annous, B.A., Ezeji, T.C., Karcher, P., Maddox, I.S., 2005. Biofilm reactors for industrial bioconversion processes: employing potential of enhanced reaction rates. Microb. Cell Fact. 4, 24. Taconi, K.A., Ventakataramanan, K.P., Johnson, D.T., 2009. Growth and solvent production by Clostridium pasteurianum ATCC 6013, utilizing biodiesel-derived crude glycerol as the sole carbon source. Environ. Prog. Sustain. Energy 28, 100–110. Tracy, B.P., Gaida, S.M., Papoutsakis, E.T., 2008. Development and application of flow-cytometric techniques for analyzing and sorting endospore-forming clostridia. Appl. Environ. Microbiol. 74, 7497–7506. Tracy, B.P., Gaida, S.M., Papoutsakis, E.T., 2010. Flow cytometry for bacteria: enabling metabolic engineering, synthetic biology and the elucidation of complex phenotypes. Curr. Opinion Biotechnol. 21, 85–99.