Immobilized Clostridium acetobutylicum P262 Mutants for

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1985, p $02.00O0 Copyright 1985, American Society for Microbiology Vol. 50, No. 2 Immobilized Clostridium acetobutylicum P22 Mutants for Solvent Production SUSAN T. LARGIER, SUSAN LONG, JOSEPH D. SANTANGELO, DAVID T. JONES, AND DAVID R. WOODS* Department of Microbiology, University of Cape Town, Rondebosch, 7700, South Africa Received 27 December 1984/Accepted 29 April 1985 The production of acetone, butanol, and ethanol by two immobilized, sporulation-deficient (spo) Clostridium acetobutylicum P22 mutants which were held in the solventogenic phase was investigated. The spoa2 mutant, which was an early-sporulation mutant and did not form a forespore septum, produced higher solvent yields than did the spob mutant which was a late-sporulation mutant and was blocked at a stage after forespore septum formation. The spoa2 mutant was also granulose and capsule negative. In a conventional batch fermentation, the wild-type strain produced g of solvents per liter after 50 h at a productivity of 7.41 g of solvents per liter per day. The spoa2 mutant produced g of solvents per liter at a productivity of 72.4 g of solvents per liter per day, with a retention time of 2.4 h in a continuous immobilized cell system employing a fluidized bed reactor. This represents a major advance, since the immobilization of wild-type cells showed similar increases in productivity but a ca. fivefold reduction in final product concentrations. Although studies on acetone and butanol synthesis by immobilized Clostridium acetobutylicum cells suggest that the application of immobilized cells may be advantageous compared with existing technologies, solvent yields are still very low (5-7, 9). In the conventional industrial batch process, which takes 30 to 48 h, solvent yields of 15 to 20 g/liter are obtained, and the acetone-butanol-ethanol (ABE) ratio is :3:1 (8, 1). In studies with immobilized C. acetobutylicum cells in repeated batch or continuous column operations, the total solvent yields varied between 1.44 and 4.53 g/liter (5-7, 9). The conventional ABE fermentation carried out by C. acetobutylicum is a two-phase batch process characterized by an initial acidogenic phase followed by a solventogenic phase (8, 11, 12). Studies on the physiological and developmental changes associated with growth, solvent production, and sporulation indicated that the clostridial stage (swollen, granulose-containing, phase bright, cigar-shaped cells surrounded by an extracellular polysaccharide capsule) which precedes endospore formation is responsible for solvent production (8, 11; S. Long, Ph.D. thesis, University of Cape Town, Rondebosch, South Africa, 1984). The isolation of cls mutants which were unable to form a clostridial stage and did not produce solvents or endospores confirmed the physiological studies linking the clostridial stage with solvent production (8, 12). Since the clostridial stage is important for solvent production, an important class of spo (sporulation deficient) mutants was isolated. These spo mutants formed the clostridial stage, and as they were unable to sporulate and were held in the clostridial stage, they produced slightly higher levels of solvents than did the wild-type strain (8, 12). Solvent-producing spo mutants offer the possibility of immobilizing a C. acetobutylicum culture in which 100% of the cells are held in the solventogenic phase. A major problem with previous attempts to produce solvents with immobilized C. acetobutylicum cells is that the immobilized systems were not homogeneous and contained differentiating cells (5-7), or they contained predominantly nonsolventproducing mature spores (9). The dynamic immobilized systems were obtained by heat activation of spores within * Corresponding author. 477 gel beads, and although growth was interrupted by washing off nutrients when solvent production started, the cultures were not synchronous and contained nonsolvent-producing germinating spores, vegetative cells, and spores as well as solvent-producing clostridial stages. We have investigated the immobilization of two spo mutants which are held in the solventogenic phase. A comparison of solvent production by these sporulation-deficient mutants and the wild-type P22 strain was carried out by a conventional batch fermentation system, a batch immobilized cell system, and a continuous immobilized cell system with a fluidized bed reactor. We report the production of high levels of solvents by an immobilized spo mutant. MATERIALS AND METHODS Bacterial mutants. The C. acetobutylicum P22 strain which has been described previously was used (1, 2, 8, 18). The P22 strain differs from the type culture strain, ATCC 824, in that it exhibits clearly defined morphological and cytological changes associated with the various phases of fermentation (8, 10). The isolation and characterization ofspo mutants, granulose-negative mutants, and capsule-negative mutants have been described previously (8, 12). The spoa2 mutant was an early-sporulation mutant which did not form a forespore septum and was also granulose negative and capsule negative. Typical clostridial stages were not observed, but the slightly refractile, phase-gray, solventproducing cells were morphologically distinct from the phase-dark vegetative rods. The spob mutant was a late-sporulation mutant and was blocked after forespore septum formation. Granulose, capsules, and forespore septa were produced, but mature spores were not formed. Growth, media, and maintenance. Since the spo mutants were unable to sporulate, they were maintained in beef liver medium. The maintenance and activation of the wild-type strain has been described previously (8, 11). The bacteria were grown at 34 C in the molasses fermentation medium (MFM) of Barber et al. (2) and the buffered Clostridium basal medium of O'Brien and Morris (14). All manipulations were carried out under stringent anaerobic conditions in an anaerobic glove cabinet (Forma-Scientific Inc., Marietta, Ohio). For cell immobilization, the strains were inoculated

2 478 LARGIE]k ET AL. into MFM and grown until ca. 100% of the cells had reached the clostridial stage. Total bacterial counts and clostridialstage counts were determined with a Thoma counting chamber (Weber Scientific International, Lancing, England) and a Zeiss photomicroscope fitted with phase- and interferencecontrast optics. Growth was also monitored by viable plate counts on Clostridium basal medium agar plates and by dry weight determinations. Cell immobilization. Clostridial stages were harvested from 500-ml MFM cultures by centrifugation at 5,000 x g for 15 min. The pellets were suspended in 50 ml of sterile distilled water. A 50-ml solution of sodium alginate (Manucol LD, Kelcol Ail International Ltd, Girvan, Ayrshire) in sterile distilled water was made with an alginate concentration which was double that required (4 to 1% sodium alginate). The clostridial-stage cells and sodium alginate were mixed to form a viscous suspension and extruded - dropwise through a hypodermic needle into CaCl2 2H20 (20 g/liter) with gentle stirring. All manipulations during immobilization were performed under anaerobic conditions. - The beads were collected, washed in CaC12 2H20 (5 g/liter), and suspended in MFM containing acetic acid (2 g/liter) and butyric acid (1 g/liter) (MFM + A medium) at a of 5.1. Solvent production. The production of solvents by the immobilized cells was investigated in batch and continuous column operations. The continuous fermentation was carried out in a liquid-fluidized bed reactor (3, 4, 15). The reactor was assembled outside the anaerobic cabinet and sparged with N2 before the addition of the immobilized cells suspended in MFM + A. The reactor and the substrate reservoir were sparged with N2 before being set into operation. The substrate was pumped upwards through the column. Immediately before entering the reactor, the substrate was mixed by a rapid recycle stream (153 ml/min). This inflow stream, together with a slow N2 flow, ensured the fluidization of the bed in the reactor. The total volume of the reactor including the recycle tubing, was 98 ml. The flow rate was 0.4 liters/day, and the temperature was maintained at 34 C by a water jacket. Acetate, butyrate, acetone, butanol, and ethanol were measured by gas chromatography as described previously (11). Sucrose was determined as total invert sugar (13). Productivity is defined as the production rate per unit reactor volume (17) and was calculated by the following equation: Qp = FCpIv; Qp is productivity (grams of product per liter of reactor volume per day); F is flow rate (liters per day); Cp is concentration of product (grams per liter); and V is reactor volume (liters). Biomass/solvent ratios and yield coefficients (grams of product per gram of substrate) were also calculated (17). RESULTS Fermentation proffles in the conventional batch process. The fermentation profiles (, total cell count, percentage of clostridial forms, and total solvents) of the wild-type, spoa2, and spob mutants were investigated in MFM (Fig. 1). The wild-type strain differed from the mutant strains in that it grew more rapidly and did not have an extensive lag growth phase. The wild-type and spoa2 strains were characterized by V-shaped profiles with pronounced breakpoints. The breakpoint of the spob mutant was less well marked, and the did not increase after the breakpoint. The maximum total cell count of the three cultures varied between 5.00 x 108 and 3.75 x 109 cells per ml. The percentage of clostridial forms produced by the three strains was APPL. ENVIRON. MICROBIOL. relatively similar and ranged between 93 and 98%. However, the total solvent and butanol yields differed, and the spoa2 mutant produced higher levels of total solvents and butanol than did the wild-type strain (Table 1). The spob mutant produced approximately half the concentration of solvents and butanol as did the wild type; spoa mutants were similar, but the spob mutant was markedly less productive (Table 1). The ABE ratios of the three strains were comparable, but the spoa2 mutant produced more acetone (Table 1). Optimization of immobilization. The effect of CaCl2 * 2H20, bead size, and alginate concentration on solvent production was determined. The addition of 5 g of CaCl2 * 2H20 per liter did not affect the production of solvents; 14.5 and 15.5 g of solvents per liter were produced in the control culture and a culture containing 5 g of CaCl2 * 2H20 per liter, respectively. The highest solvent levels were obtained with the smallest beads (diameter of ca. 2 mm). Although the concentration of alginate in the beads (between 2 and 8%) did not affect the final solvent levels, there was an increase in the lag time before solvent production with the higher alginate concentrations. Since it was important that the maximum number of active, stable solvent-producing cells was immobilized, the effect of the time of immobilization of the spoa2 mutant on solvent production was investigated (Fig. 2). Immobilization of spoa2 cells from an MFM culture at various time intervals between 18.7 and 2.5 h did not affect either the lag before solvent production or the solvent yields in the immobilized batch system. Immobilized batch system. The performance of immobilized wild-type, spoa2, and spob cells was determined in a batch system. The highest levels of total solvents and butanol were obtained with the spoa2 mutant, whereas the spob mutant produced the lowest solvent yields (Table 1). The productivity of the spoa2 mutant was markedly better than the other two strains (Table 1). The ABE ratios of the three strains were comparable, but the spoa2 mutant produced slightly more acetone (Table 1). The yield coefficients of the spoa2 mutant for total solvents and butanol were 0.44 and 0.2 g of solvents per g of sucrose, respectively. Continuous solvent production by immobilized spoa2 mutant cells. Since the immobilization of the spoa2 mutant gave the best solvent yields and highest productivity in the batch experiments, it was utilized in the continuous column operations. An example of a continuous fermentation profile with immobilized spoa2 clostridial stages is shown in Fig. 3. Similar profiles were obtained in four separate experiments, and total solvent and butanol levels of >15 and >8 g/liter, respectively, were routinely obtained with retention times of 2.4 and 4.9 h. The productivities of total solvents and butanol were 72.4 and g of solvent per liter per day, respectively (Table 1). The ABE ratio was 3::0.4. The ratios of biomass/total solvents and biomass/butanol were 3.54 and.52, respectively. The yield coefficients for total solvents and butanol were 0.5 and 0.3 g of solvents per g of sucrose, respectively. Although 3% alginate beads remained intact in batch fermentations with MFM + A, the 3% beads disintegrated in the continuous fermentation column after ca. 40 h. However, 4.5% alginate beads remained intact for at least 2 weeks in the continuous column, but the total solvent levels decreased from >15 to 10 g/liter. DISCUSSION The maximum levels of solvents previously reported for immobilized C. acetobutylicum cells in batch and continuous

3 VOL. 50, 1985 SOLVENT PRODUCTION BY C. ACETOBUTYLICUM MUTANTS lo 25- o107p 5 ol lor. 75 [ o L FIG. 1. Conventional batch fermentation profile of the spob mutant (A), P22 wild-type strain (B), and spoa2 mutant (C). Total cell counts (A), (O), percentage of clostridial stages (0), and total solvents (0) were determined. column operations varied between 1.44 and 4.53 g/liter (5-7, 9). The production of >15 g of solvents per liter by immobilized spoa2 clostridial stages in a continuous column represents a threefold improvement on the previous systems. The yields obtained with immobilized spoa2 cells are comparable with the yields obtained in the conventional industrial process (1), but the fermentation time is reduced from ca. 40 to 2.4 h in the continuous reactor. Immobilization of the spoa2 mutant was markedly more successful than either the spob or wild-type strains. Since the spoa2 mutant was an early-sporulation mutant, it is suggested that it is important to utilize mutants which are unable to form a forespore septum. Further advantages of the spoa2 mutant are that it is unable to produce granulose or capsules, and more substrate is therefore available for conversion to solvents. The absence of a slimy polysaccharide capsule enhanced immobilization and bead formation and eliminated a diffusion barrier between the immobilized cells and the substrate. The productivity of the spoa2 mutant improved from 7.89 g of solvent per liter per day and 4.92 g of butanol per liter per day in the conventional batch fermentation to 72.4 g of solvents per liter per day and g of butanol per liter per day in the continuous immobilized system. The ABE ratio remained the same in the conventional and immobilization systems. Although productivities of 57 to 7 g of butanol per 5 liter per day have been reported previously (, 9), the total solvent and butanol yields were very low (2.2 g/liter and 2.05 g/liter, respectively). The yields obtained with the immobilized spoa2 mutant in the continuous system were TABLE 1. Comparison of wild-type (P22) and mutant (spoa2 and spob) strains of C. acetobutylicum in conventional and immobilized fermentation systems Concn (g/liter) of: Productivity' Fermentation Strain ABE system sototal Butanol ratio Solvents Butanol solvents Conventional P :: batch spoa :: spob :: Immobilized P :: batch spoa :: spob :: Immobilized spoa :: continuousb a Productivity was measured in grams of solvents per liter of reactor volume per day. bflow rate = 0.4 liter/day.

4 480 LARGIER ET AL. APPL. ENVIRON. MICROBIOL co 10 z 'u.i 8 0 Co FIG. 2. The effect of the time of immobilization on solvent production in an immobilized batch system. spoa2 clostridial-stage cells from MFM were immobilized at 18 h 40 min (0), 20 h 25 min (0), 22 h 15 min (U), 24 h 15 min (O), and 2 h 30 min (A) g of butanol per liter and g of total solvents per liter. It is important when comparing processes for product formation to consider not only the productivity but also the final product concentration leaving the fermentor and entering the product recovery system (17). This is particularly relevant to the ABE fermentation in which the cost of solvent recovery is a major factor in the financial success of the fermentation. The continuous fermentation with immobilized spoa2 mutants represents a major advance, as the final product concentrations are comparable with conventional batch processes, but the productivity has been increased ca. 10-fold. The previously reported continuous immobilized processes showed similar increases in productivity but ca. fivefold reduction in final product concentrations. Yield coefficients of 0.17 to g of butanol per g of glucose for immobilized batch and continuous systems have been reported previously (5, 7). Comparable yield coefficients of 0.2 to 0.3 g of butanol per g of sucrose and 0.44 to 0.5 g of total solvents per g of sucrose were obtained with the spoa2 mutant in the batch and continuous systems. Haggstrom (5) reported a biomass/butanol ratio of 2% in a continuous immobilized process, and the ABE ratio was 2:7:1. A higher butanol/biomass ratio of.52% was obtained with the spoa2 mutant which had an ABE ratio of 3::0.4. The substrate utilized by Haggstrom (5-7) contained higher concentrations of butyrate than did the substrate in the spoa2 continuous immobilized system. It is unlikely that variations in the acid concentrations of the substrate will affect the ABE ratios and biomass/butanol ratios in a continuous process. The time of immobilization of the spoa2 mutant is not a critical factor. This is important for an industrial process and J 5.0 A ~~~~~14 0o TIE ()10 -'10 8 _1 LU U ~~~~ 1- ~~~~~ FIG. 3. Continuous fermentation profile by immobilized spoa2 clostridial stages. Total cell count (A), (Ol), total acids (U), total solvents (0), and butanol concentration (0) were determined. will facilitate the preparation of large amounts of immobilized cells. Furthermore, the cells were pregrown in a conventional industrial MFM, and the immobilized cells were suspended in the same medium. It was not necessary to wash off nutrients to obtain solvent-producing cells. The successful immobilization of the spoa2 mutant and the production of high solvent yields indicate that a continuous fermentation system is feasible for ABE production by C. acetobutylicum. Single-stage continuous cultures of C. acetobutylicum vegetative cells have not been successful for ABE production. However, a two-stage continuous fermentation is now possible with cls and spoa2 mutants. In the first stage, the cls mutant could be utilized for the production of acids from molasses in a continuous culture of vegetative cells. The molasses and acids would then form the substrate for the spoa2 mutant in the continuous immobilized system. LITERATURE CITED 1. Allcock, E. R., S. J. Reid, D. T. Jones, and D. R. Woods Autolytic activity and an autolysis-deficient mutant of Clostridium acetobutylicum. Appl. Environ. Microbiol. 42: Barber, J. M., F. T. Robb, J. R. Webster, and D. R. Woods Bacteriocin production by Clostridium acetobutylicum in an industrial fermentation process. Appl. Environ. Microbiol. 37: Black, G. M., C. Webb, T. M. Matthews, and B. Atkinson Practical reactor systems for yeast immobilisation using biomass support particles. Biotechnol. Bioeng. 2: Cho, G. H., and C. Y. Choi Continuous ethanol production by immobilised yeast in a fluidised reactor. Biotechnol. Lett. 3: Forberg, C., S. 0. Enfors, and L. Haggstrom Control of immobilised non-growing cells for continuous production of 2

5 VOL. 50, 1985 SOLVENT PRODUCTION BY C. ACETOBUTYLICUM MUTANTS 481 metabolites. Eur. J. Appl. Microbiol. Biotechnol. 17: Haggstrom, L., and S. Enfors Continuous production of butanol with immobilised cells of Clostridium acetobutylicum. Appi. Biochem. Biotechnol. 7: Haggstrom, L., and N. Molin Calcium alginate immobilised cells of Clostridium acetobutylicum for solvent production. Biotechnol. Lett. 2: Jones, D. T., A. van Westhuizen, S. Long, E. R. Alicock, S. J. Reid, and D. R. Woods Solvent production and morphological changes in Clostridium acetobutylicum. Appl. Environ. Microbiol. 43: Kobt, F. B Immobilised cells for solvent production. Process Biochem. 19: Long, S., D. T. Jones, and D. R. Woods Sporulation of Clostridium acetobutylicum P22 in a defined medium. Appl. Environ. Microbiol. 45: Long, S., D. T. Jones, and D. R. Woods Initiation of solvent production, clostridial stage and endospore formation in Clostridium acetobutylicum P22. Eur. J. Appl. Microbiol. Biotechnol. 20: Long, S., D. T. Jones, and D. R. Woods The relationship between sporulation and solvent production in Clostridium acetobutylicum P22. Biotechnol. Lett. : Mann, F. G., and B. C. Saunders Practical organic chemistry, 4th ed. Longmans, Green and Co. Ltd. London. 14. O'Brien, R. W., and J. G. Morris Oxygen and the growth and metabolism of Clostridium acetobutylicum. J. Gen. Microbiol. 8: Scott, C. D Fermentation in a fluidised bed reactor. Chemtech 13: Spivey, M. J The acetone/butanol/ethanol fermentation. Process Biochem. 13: Wang, D. I. C., C. L. Cooney, A. L. Demnain, P. Dunnill, A. E. Humphrey, and M. D. Lilly Fermentation and enzyme technology. John Wiley & Sons, Inc. 18. Webster, J. R., S. J. Reid, D. T. Jones, and D. R. Woods Purification and characterization of an autolysin from Clostridium acetobutylicum. Appl. Environ. Microbiol. 41: