- Email: [email protected]
Bioconversion of glucose and fructose to sorbitol and gluconic acid by untreated cells of Zymomonas mobilis
Journal of Biotechnology 75 (1999) 99 – 103 www.elsevier.com/locate/jbiotec
Bioconversion of glucose and fructose to sorbitol and gluconic acid by untreated cells of Zymomonas mobilis Mauricio M. Silveira a, Elisabeth Wisbeck a, Claudia Lemmel a,b, Gilmar Erzinger a, Jose´ Paulo Lopes da Costa a, Marcelo Bertasso a, Rainer Jonas a,b,* a
b
Centro de Desen6ol6imento Biotecnolo´gico, Join6ille SC, Brazil Gesellschaft fu¨r Biotechnologische Forschung, Braunschweig, Germany
Received 31 August 1998; received in revised form 27 April 1999; accepted 7 May 1999
Abstract The bioconversion of glucose and fructose to gluconic acid and sorbitol, respectively, by the enzymes glucose – fructose oxidoreductase (GFOR) and glucono-d-lactonase (GL), contained in untreated cells of Zymomonas mobilis ATCC 29191, was investigated in batch runs with glucose plus fructose concentrations (S0) varying from 100 to 750 g l − 1 in equimolar ratio. When S0 was increased to 650 g l − 1, the yields were improved, reaching a maximum of 91% for both products, with productivities of 1.6 and 1.5 g g − 1 cell h − 1 for gluconic acid and sorbitol, respectively. Above this level (S0 =750 g l − 1), no further improvement in yields was observed and productivities decreased due to the longer process time. The high yields of bioconversion runs with S0 ] 650 g l − 1 are a consequence of the sequential inhibition of the normal metabolism of Z. mobilis by substrates and products, resulting in preferential utilization of substrates via the GFOR/GL system. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Zymomonas mobilis; Sorbitol; Gluconic acid; Glucose – fructose oxidoreductase
1. Introduction The ethanologenic bacterium Zymomonas mobilis is able to produce glucono-lactone and sorbitol from a mixture of glucose and fructose through reactions catalyzed by glucose – fructose * Corresponding author. Present address: Gesellschaft fu¨r Biotechnologische Forschung mbH, Mascheroder Weg 1, D38124 Braunschweig, Germany. Tel.: + 49-531-6181-503; fax: +49-531-181-575. E-mail address: [email protected] (R. Jonas)
oxidoreductase (GFOR; EC 1.1.1.99), an enzyme with tightly coupled NADP. Subsequently, glucono-d-lactone is hydrolyzed by glucono-d-lactonase (GL; EC 3.1.1.17) to gluconic acid (Zachariou and Scopes, 1986). Biological processes for the production of sorbitol and gluconic acid, using previously grown and concentrated Z. mobilis cells, have been proposed. In these processes, the concentrated cell mass is treated with permeabilizing agents such as toluene (Chun and Rogers, 1988) or detergents
0168-1656/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 1 4 9 - 2
100
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
(Rehr et al., 1991) to release essential soluble co-factors necessary for the conversion of gluconic acid to ethanol and other catabolic products via the Entner–Doudoroff pathway. Yields for sorbitol and gluconic acid of 94 – 95% (toluenetreated cells) and 98 – 99% (detergent-treated cells), in batch processes with free cells were reported. In the same article, Rehr et al. (1991) described a two-stage continuous process with k-carrageenan immobilized cells, in which yields over 98% for both products were achieved. Previously, Silveira et al. (1994) proposed a process in batch mode with non-permeabilized cells of the strain Z. mobilis CP1, in which high yields for both products were obtained when the initial concentration of glucose plus fructose was increased to 600 g l − 1. Similar results were found by Wisbeck et al. (1997) with the flocculent strain Z. mobilis Z1-81. The reason why the utilization of non-permeabilized cells in this process was searched was to avoid operational difficulties and reduce costs in a large-scale production of sorbitol and gluconic acid. The present work examines the influence of the concentration of glucose and fructose on the process performed with non-permeabilized cells of Z. mobilis in a more detailed way. An explanation for the high yields achieved in these conditions is proposed.
2. Materials and methods Z. mobilis ATCC-29191 was obtained from the American Type Culture Collection (ATCC) and maintained in semi-synthetic (SS) medium containing 20 g l − 1 glucose, at 4°C. The SS medium had the following composition (g l − 1): (NH4)2SO4, 2.0; MgSO4·7H2O, 1.0; FeSO4, 0.01; KH2PO4, 3.5; sodium citrate.2H2O, 0.2; yeast extract, 5.0. Glucose solutions were prepared separately and added to the medium prior to inoculation to reach the desired concentration for each case. Both medium and glucose solutions were sterilized at 121°C for 20 min. Cell cultivation was performed in the SS medium containing 150 g l − 1 glucose in a B. Braun Biotech Biostat B bioreactor at an impeller
speed of 300 min − 1. The temperature was kept at 30°C and pH was controlled at 5.5 with 3 M NaOH. After growth, the cell mass was centrifuged at 4000× g for 20 min and resuspended in distilled water to approximately 95 g l − 1. Bioconversion of glucose and fructose to sorbitol and gluconic acid was carried out in a 300 ml reactor, with 30 g l − 1 of untreated cells, at different concentrations of substrates, at 39°C. Agitation was provided by magnetic stirring. The pH was automatically controlled at 6.5 with 14 M NaOH. The effect of sorbitol and gluconic acid concentrations on the metabolism of Z. mobilis was evaluated in the 300 ml reactor. The SS medium containing 30 g l − 1 glucose, with a variety of equimolar concentrations of sorbitol and sodium gluconate, was used in these experiments. The initial cell concentration was 30 g l − 1. The pH was controlled at 5.5 with 3 M NaOH and the temperature was kept at 30°C. Cell growth was determined by measuring the optical density of cell suspensions at 560 nm. These measurements gave a linear relationship with dry cell mass concentration. In some bioconversion experiments, the viable cell number was estimated after incubation in the SS medium (20 g l − 1 glucose), at 30°C, for 48 h. Glucose, fructose and sorbitol were determined by high performance liquid chromatography (Merck-Hitachi), and gluconic acid by the enzymatic Boehringer– Mannheim method. All experiments were performed at least twice, and the results presented here correspond to the average of the values obtained. The results of corresponding experiments showed no significant differences.
3. Results and discussion To investigate the influence of the initial concentration of substrates (S0) on the production of gluconic acid and sorbitol, batch runs with equimolar glucose plus fructose concentrations varying from 100 to 750 g l − 1 were carried out. As can be seen from Table 1, increasing product yields were found for S0 up to 650 g l − 1. Above
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
101
Table 1 Yields and productivities for gluconic acid and sorbitol with different initial glucose plus fructose concentrations (cell concentration, 30 g l−1)a Glucose+fructose (g l−1)
Gluconic acid yield (%) Sorbitol yield (%) Specific gluconic acid productivity (g g−1 h−1) Specific sorbitol productivity (g g−1 h−1)
100
300
400
500
600
650
750
7.5 35 0.24 0.80
40 83 1.3 1.7
58 74 1.5 1.6
80 79 2.2 1.8
83 86 1.6 1.5
91 91 1.6 1.5
90 92 1.3 1.3
a Calculations made when one of the substrates was completely depleted; the specific productivity is referred to grams of product obtained per grams of cell mass per hour.
this level (S0 =750 g l − 1), no further improvement in yields was observed, and productivities decreased due to the longer process time. The analysis of the results of the experiments with S0 =100, 300, and 650 g l − 1 (Figs. 1 – 3) can explain the results presented in Table 1. With S0 = 100 g l − 1 (Fig. 1), glucose and fructose were simultaneously converted to gluconic acid and sorbitol, respectively, in the initial minutes of the run. The relative high concentrations of gluconic acid when compared with those for sorbitol after nearly 20 min may be due to the fact that part of the gluconic acid formed, was produced via glucose dehydrogenase (Zachariou and Scopes, 1986; Strohdeicher et al., 1988). After this time, however, one could observe a shift to the normal metabolism of carbohydrates of Z. mobilis, characterized by the following aspects: (i) preferential consumption of glucose, (ii) strong reduction of the fructose consumption rate, (iii) interruption of sorbitol formation, and (iv) degradation of gluconic acid. With S0 =300 g l − 1 (Fig. 2), similar behavior, at different times, was noticed. In this case, however, the predominance of the metabolic phase was not so evident as in the previous run, excepting the rapid decrease of gluconic acid concentration, after the total depletion of glucose. Increasing S0 to 650 g l − 1 (Fig. 3) resulted in a nearly complete bioconversion of substrates to sorbitol and gluconic acid, with final concentrations of 300 and 320 g l − 1, respectively. Thus, the yields presented in Table 1 are due to the prevalent form of utilization of substrates, bioconversion or their catabolic degradation, determined by S0.
In Table 1, it is shown that the maximum specific productivities were achieved with S0 = 500 g l − 1. According to Silveira et al. (1995), GFOR/ GL contained the highest activities in intact cells of Z. mobilis ATCC 29191, when glucose plus fructose concentrations between 400 and 550 g l − 1 are used. In this condition, substrates remained in favorable concentrations for a large part of the run and, therefore, the process rate was improved. Moreover, with the shorter process time, compared with runs with higher S0, the product formation rates were less influenced by the loss of GFOR/GL activity during the bioconversion. Nevertheless, due to the better yields, an initial glucose plus fructose concentration close to 650 g l − 1 is more convenient for this process. The enhancing yields of bioconversion with increasing S0 were probably linked to the concen-
Fig. 1. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 100 g l − 1.
102
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
Fig. 2. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 300 g l − 1.
trations of both substrates and products present in the medium through two possible effects: (i) loss of cell viability owing to osmotic pressure and (ii) inhibition of the metabolism of carbohydrates by Z. mobilis. The effect of the high concentration of dissolved substances on Z. mobilis viability was examined by cell counting before starting and at the end of bioconversion runs with S0 of 400 and 650 g l − 1. At the beginning of each experiment, cell concentrations around 2×109 cfu ml − 1 were measured. In bioconversion with S0 =400 g l − 1, after 3 h of process, cell concentration was reduced to approximately 2× 104 cfu ml − 1. With S0 = 650 g l − 1, the final counting, after 6 h of
Fig. 3. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 650 g l − 1.
process, was 8×103 cfu ml − 1. These results confirm that increasing substrates concentration led to considerable losses of cell viability in bioconversion runs, although viable cell concentration was still large enough to metabolize glucose and gluconic acid. Thus, osmotic pressure cannot be considered as the principal reason for the results shown in Table 1, which is in accordance to the protecting effect of sorbitol in cells of Z. mobilis at high sugar concentrations (Sahm et al., 1995). Inhibition of Z. mobilis by high glucose concentrations is well known. Erzinger (1996) reported a significant reduction of growth and ethanol production rates when initial glucose concentrations above 150 g l − 1 were used in batch cultures. Additionally, our results show that bioconversion products also inhibit the consumption of both glucose and gluconic acid by this microorganism. When equimolar concentrations of sorbitol and gluconic acid added to the SS medium raised from 0 to 328 and 353 g l − 1, respectively, decreasing glucose consumption rates and increasing residual concentrations of both glucose and gluconic acid were measured (Table 2). In the presence of 255 g l − 1 sorbitol and 274 g l − 1 gluconic acid, for example, some 89% of the initial glucose and 96% of the added gluconic acid remained in the medium. Therefore, the main reason for the high yields of bioconversion runs with S0 ] 650 g l − 1 is the sequential inhibition of the normal metabolism of the microorganism by its substrates and products, resulting in preferential utilization of the substrates via the glucose–fructose oxidoreductase/glucono-d-lactonase system. Under such conditions, gluconic acid consumption is negligible and the yields for both bioconversion products are similar since GFOR converts equimolar amounts of glucose and fructose to produce glucono-d-lactone and sorbitol. In the present work, the best yields achieved for sorbitol and gluconic acid (91%) were lower than those reported in the literature for permeabilized Z. mobilis, showing some advantage in using treated cells (Chun and Rogers, 1988; Rehr et al., 1991). On the other hand, sorbitol and gluconic acid have a relatively low market price, and, therefore, any reduction in the production costs is
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
103
Table 2 Effect of gluconic acid and sorbitol on the metabolism of glucose and gluconic acid by Z. mobilis (initial glucose concentration, 30 g l−1) Initial sorbitol/gluconic acid (g l−1)
Glucose consumption rate (g l−1 h−1) Residual glucose (%)a Residual gluconic acid (%)a a
0/0
55/59
146/157
255/274
328/353
92 0 –
25 0 0
11 0 27
0.40 89 98
0.14 96 99
After 8 h of process.
desirable. For this reason, their economical production using permeabilized cells should consider the reutilization of the GFOR-containing biomass in long-term operation processes as the two-stage continuous method described by Rehr et al. (1991). In that process, performed with Z. mobilis cells permeabilized with cetyltrimethylammonium bromide (CTAB) and immobilized in k-carrageenan, hardened with glutaraldehyde and polyethylenimine, a decrease of only 3.5% of the original enzymatic activity was observed after 75 days. On the other hand, with a feed concentration (glucose +fructose) of 200 g l − 1, relative low specific productivities (product per biomass hour) of 0.19 and 0.21 g g − 1 h − 1 for sorbitol and gluconic acid, respectively, were obtained. These low productivities were due to the concentration of glucose and fructose in the second stage of that bioprocess ( B10 g l − 1 of each substrate), considerably smaller than the Km of 77.2 g l − 1 fructose calculated by these authors for permeabilized and immobilized cells. The continuous process has as an additional problem: the low concentration of products obtained (B 100 g l − 1) means increasing costs with the removal of water in the downstream processing. Furthermore, considering that the most important commercial application of sorbitol is as a non-cariogenic sweetener, the contamination with fermentable sugars in the outflow is undesirable and would imply difficulties to carry out the downstream processing. Therefore, the use of free-untreated cells of Z. mobilis in a batch process for the biological production of sorbitol and gluconic acid, despite the lower yields obtained with them, become an attractive option. An inconvenient aspect of this
alternative would be the need for cell cultivation in each run. The cost of cell growth, however, could be compensated for by the ethanol produced during this phase. References Chun, U.H., Rogers, P.L., 1988. The simultaneous production of sorbitol and gluconic acid by Zymomonas mobilis. Appl. Microbiol. Biotechnol. 29, 19 – 24. Erzinger, G.S., 1996. Influeˆncia da concentrac¸a˜o de glicose e de etanol sobre a atividade de glicose – frutose oxidoredutase em Zymomonas mobilis ATCC 29191, Master thesis, Faculty of Pharmaceutical Sciences, University of Sa˜o Paulo, Brazil, 99 pp. Rehr, B., Wilhelm, C., Sahm, H., 1991. Production of sorbitol and gluconic acid by permeabilized cells of Zymomonas mobilis. Appl. Microbiol. Biotechnol. 35, 144 – 148. Sahm, H., Loos, H., Rehr, B., Sprenger, G., 1995. Production of sorbitol and gluconic acid by Zymomonas mobilis. Meded. Fac. Landbouwwet. Rijksuniv. Gent 59 (4b), 2403 – 2410. Silveira, M.M., Lopes da Costa, J.P.C., Jonas, R., 1994. Processo de produc¸a˜o e recuperac¸a˜o de sorbitol e a´cido glucoˆnico ou gluconato, Patente de Invenc¸a˜o PI 9403981-0, INPI, Brazil. Silveira, M.M., Sales, R., Lemmel, C., Jonas, R., 1995. Glucose – fructose activity in six strains of Zymomonas mobilis. Arq. Biol. Tecnol. 38, 619 – 622. Strohdeicher, M., Schmitz, B., Bringer-Meyer, S., Sahm, H., 1988. Formation and degradation of gluconate by Zymomonas mobilis. Appl. Microbiol. Biotechnol. 27, 378 – 382. Wisbeck, E., Silveira, M.M., Ninow, J., Jonas, R., 1997. Evaluation of the flocculent strain Zymomonas mobilis Z1-81 for the production of sorbitol and gluconic acid. J. Basic Microbiol. 37, 445 – 449. Zachariou, M., Scopes, R.K., 1986. Glucose – fructose oxidoreductase, a new enzyme isolated from Zymomonas mobilis that is responsible for sorbitol production. J. Bacteriol. 167, 803 – 809.
Bioconversion of glucose and fructose to sorbitol and gluconic acid by untreated cells of Zymomonas mobilis Mauricio M. Silveira a, Elisabeth Wisbeck a, Claudia Lemmel a,b, Gilmar Erzinger a, Jose´ Paulo Lopes da Costa a, Marcelo Bertasso a, Rainer Jonas a,b,* a
b
Centro de Desen6ol6imento Biotecnolo´gico, Join6ille SC, Brazil Gesellschaft fu¨r Biotechnologische Forschung, Braunschweig, Germany
Received 31 August 1998; received in revised form 27 April 1999; accepted 7 May 1999
Abstract The bioconversion of glucose and fructose to gluconic acid and sorbitol, respectively, by the enzymes glucose – fructose oxidoreductase (GFOR) and glucono-d-lactonase (GL), contained in untreated cells of Zymomonas mobilis ATCC 29191, was investigated in batch runs with glucose plus fructose concentrations (S0) varying from 100 to 750 g l − 1 in equimolar ratio. When S0 was increased to 650 g l − 1, the yields were improved, reaching a maximum of 91% for both products, with productivities of 1.6 and 1.5 g g − 1 cell h − 1 for gluconic acid and sorbitol, respectively. Above this level (S0 =750 g l − 1), no further improvement in yields was observed and productivities decreased due to the longer process time. The high yields of bioconversion runs with S0 ] 650 g l − 1 are a consequence of the sequential inhibition of the normal metabolism of Z. mobilis by substrates and products, resulting in preferential utilization of substrates via the GFOR/GL system. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Zymomonas mobilis; Sorbitol; Gluconic acid; Glucose – fructose oxidoreductase
1. Introduction The ethanologenic bacterium Zymomonas mobilis is able to produce glucono-lactone and sorbitol from a mixture of glucose and fructose through reactions catalyzed by glucose – fructose * Corresponding author. Present address: Gesellschaft fu¨r Biotechnologische Forschung mbH, Mascheroder Weg 1, D38124 Braunschweig, Germany. Tel.: + 49-531-6181-503; fax: +49-531-181-575. E-mail address: [email protected] (R. Jonas)
oxidoreductase (GFOR; EC 1.1.1.99), an enzyme with tightly coupled NADP. Subsequently, glucono-d-lactone is hydrolyzed by glucono-d-lactonase (GL; EC 3.1.1.17) to gluconic acid (Zachariou and Scopes, 1986). Biological processes for the production of sorbitol and gluconic acid, using previously grown and concentrated Z. mobilis cells, have been proposed. In these processes, the concentrated cell mass is treated with permeabilizing agents such as toluene (Chun and Rogers, 1988) or detergents
0168-1656/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 1 4 9 - 2
100
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
(Rehr et al., 1991) to release essential soluble co-factors necessary for the conversion of gluconic acid to ethanol and other catabolic products via the Entner–Doudoroff pathway. Yields for sorbitol and gluconic acid of 94 – 95% (toluenetreated cells) and 98 – 99% (detergent-treated cells), in batch processes with free cells were reported. In the same article, Rehr et al. (1991) described a two-stage continuous process with k-carrageenan immobilized cells, in which yields over 98% for both products were achieved. Previously, Silveira et al. (1994) proposed a process in batch mode with non-permeabilized cells of the strain Z. mobilis CP1, in which high yields for both products were obtained when the initial concentration of glucose plus fructose was increased to 600 g l − 1. Similar results were found by Wisbeck et al. (1997) with the flocculent strain Z. mobilis Z1-81. The reason why the utilization of non-permeabilized cells in this process was searched was to avoid operational difficulties and reduce costs in a large-scale production of sorbitol and gluconic acid. The present work examines the influence of the concentration of glucose and fructose on the process performed with non-permeabilized cells of Z. mobilis in a more detailed way. An explanation for the high yields achieved in these conditions is proposed.
2. Materials and methods Z. mobilis ATCC-29191 was obtained from the American Type Culture Collection (ATCC) and maintained in semi-synthetic (SS) medium containing 20 g l − 1 glucose, at 4°C. The SS medium had the following composition (g l − 1): (NH4)2SO4, 2.0; MgSO4·7H2O, 1.0; FeSO4, 0.01; KH2PO4, 3.5; sodium citrate.2H2O, 0.2; yeast extract, 5.0. Glucose solutions were prepared separately and added to the medium prior to inoculation to reach the desired concentration for each case. Both medium and glucose solutions were sterilized at 121°C for 20 min. Cell cultivation was performed in the SS medium containing 150 g l − 1 glucose in a B. Braun Biotech Biostat B bioreactor at an impeller
speed of 300 min − 1. The temperature was kept at 30°C and pH was controlled at 5.5 with 3 M NaOH. After growth, the cell mass was centrifuged at 4000× g for 20 min and resuspended in distilled water to approximately 95 g l − 1. Bioconversion of glucose and fructose to sorbitol and gluconic acid was carried out in a 300 ml reactor, with 30 g l − 1 of untreated cells, at different concentrations of substrates, at 39°C. Agitation was provided by magnetic stirring. The pH was automatically controlled at 6.5 with 14 M NaOH. The effect of sorbitol and gluconic acid concentrations on the metabolism of Z. mobilis was evaluated in the 300 ml reactor. The SS medium containing 30 g l − 1 glucose, with a variety of equimolar concentrations of sorbitol and sodium gluconate, was used in these experiments. The initial cell concentration was 30 g l − 1. The pH was controlled at 5.5 with 3 M NaOH and the temperature was kept at 30°C. Cell growth was determined by measuring the optical density of cell suspensions at 560 nm. These measurements gave a linear relationship with dry cell mass concentration. In some bioconversion experiments, the viable cell number was estimated after incubation in the SS medium (20 g l − 1 glucose), at 30°C, for 48 h. Glucose, fructose and sorbitol were determined by high performance liquid chromatography (Merck-Hitachi), and gluconic acid by the enzymatic Boehringer– Mannheim method. All experiments were performed at least twice, and the results presented here correspond to the average of the values obtained. The results of corresponding experiments showed no significant differences.
3. Results and discussion To investigate the influence of the initial concentration of substrates (S0) on the production of gluconic acid and sorbitol, batch runs with equimolar glucose plus fructose concentrations varying from 100 to 750 g l − 1 were carried out. As can be seen from Table 1, increasing product yields were found for S0 up to 650 g l − 1. Above
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
101
Table 1 Yields and productivities for gluconic acid and sorbitol with different initial glucose plus fructose concentrations (cell concentration, 30 g l−1)a Glucose+fructose (g l−1)
Gluconic acid yield (%) Sorbitol yield (%) Specific gluconic acid productivity (g g−1 h−1) Specific sorbitol productivity (g g−1 h−1)
100
300
400
500
600
650
750
7.5 35 0.24 0.80
40 83 1.3 1.7
58 74 1.5 1.6
80 79 2.2 1.8
83 86 1.6 1.5
91 91 1.6 1.5
90 92 1.3 1.3
a Calculations made when one of the substrates was completely depleted; the specific productivity is referred to grams of product obtained per grams of cell mass per hour.
this level (S0 =750 g l − 1), no further improvement in yields was observed, and productivities decreased due to the longer process time. The analysis of the results of the experiments with S0 =100, 300, and 650 g l − 1 (Figs. 1 – 3) can explain the results presented in Table 1. With S0 = 100 g l − 1 (Fig. 1), glucose and fructose were simultaneously converted to gluconic acid and sorbitol, respectively, in the initial minutes of the run. The relative high concentrations of gluconic acid when compared with those for sorbitol after nearly 20 min may be due to the fact that part of the gluconic acid formed, was produced via glucose dehydrogenase (Zachariou and Scopes, 1986; Strohdeicher et al., 1988). After this time, however, one could observe a shift to the normal metabolism of carbohydrates of Z. mobilis, characterized by the following aspects: (i) preferential consumption of glucose, (ii) strong reduction of the fructose consumption rate, (iii) interruption of sorbitol formation, and (iv) degradation of gluconic acid. With S0 =300 g l − 1 (Fig. 2), similar behavior, at different times, was noticed. In this case, however, the predominance of the metabolic phase was not so evident as in the previous run, excepting the rapid decrease of gluconic acid concentration, after the total depletion of glucose. Increasing S0 to 650 g l − 1 (Fig. 3) resulted in a nearly complete bioconversion of substrates to sorbitol and gluconic acid, with final concentrations of 300 and 320 g l − 1, respectively. Thus, the yields presented in Table 1 are due to the prevalent form of utilization of substrates, bioconversion or their catabolic degradation, determined by S0.
In Table 1, it is shown that the maximum specific productivities were achieved with S0 = 500 g l − 1. According to Silveira et al. (1995), GFOR/ GL contained the highest activities in intact cells of Z. mobilis ATCC 29191, when glucose plus fructose concentrations between 400 and 550 g l − 1 are used. In this condition, substrates remained in favorable concentrations for a large part of the run and, therefore, the process rate was improved. Moreover, with the shorter process time, compared with runs with higher S0, the product formation rates were less influenced by the loss of GFOR/GL activity during the bioconversion. Nevertheless, due to the better yields, an initial glucose plus fructose concentration close to 650 g l − 1 is more convenient for this process. The enhancing yields of bioconversion with increasing S0 were probably linked to the concen-
Fig. 1. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 100 g l − 1.
102
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
Fig. 2. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 300 g l − 1.
trations of both substrates and products present in the medium through two possible effects: (i) loss of cell viability owing to osmotic pressure and (ii) inhibition of the metabolism of carbohydrates by Z. mobilis. The effect of the high concentration of dissolved substances on Z. mobilis viability was examined by cell counting before starting and at the end of bioconversion runs with S0 of 400 and 650 g l − 1. At the beginning of each experiment, cell concentrations around 2×109 cfu ml − 1 were measured. In bioconversion with S0 =400 g l − 1, after 3 h of process, cell concentration was reduced to approximately 2× 104 cfu ml − 1. With S0 = 650 g l − 1, the final counting, after 6 h of
Fig. 3. Time course of a typical bioconversion run with initial glucose plus fructose concentrations of 650 g l − 1.
process, was 8×103 cfu ml − 1. These results confirm that increasing substrates concentration led to considerable losses of cell viability in bioconversion runs, although viable cell concentration was still large enough to metabolize glucose and gluconic acid. Thus, osmotic pressure cannot be considered as the principal reason for the results shown in Table 1, which is in accordance to the protecting effect of sorbitol in cells of Z. mobilis at high sugar concentrations (Sahm et al., 1995). Inhibition of Z. mobilis by high glucose concentrations is well known. Erzinger (1996) reported a significant reduction of growth and ethanol production rates when initial glucose concentrations above 150 g l − 1 were used in batch cultures. Additionally, our results show that bioconversion products also inhibit the consumption of both glucose and gluconic acid by this microorganism. When equimolar concentrations of sorbitol and gluconic acid added to the SS medium raised from 0 to 328 and 353 g l − 1, respectively, decreasing glucose consumption rates and increasing residual concentrations of both glucose and gluconic acid were measured (Table 2). In the presence of 255 g l − 1 sorbitol and 274 g l − 1 gluconic acid, for example, some 89% of the initial glucose and 96% of the added gluconic acid remained in the medium. Therefore, the main reason for the high yields of bioconversion runs with S0 ] 650 g l − 1 is the sequential inhibition of the normal metabolism of the microorganism by its substrates and products, resulting in preferential utilization of the substrates via the glucose–fructose oxidoreductase/glucono-d-lactonase system. Under such conditions, gluconic acid consumption is negligible and the yields for both bioconversion products are similar since GFOR converts equimolar amounts of glucose and fructose to produce glucono-d-lactone and sorbitol. In the present work, the best yields achieved for sorbitol and gluconic acid (91%) were lower than those reported in the literature for permeabilized Z. mobilis, showing some advantage in using treated cells (Chun and Rogers, 1988; Rehr et al., 1991). On the other hand, sorbitol and gluconic acid have a relatively low market price, and, therefore, any reduction in the production costs is
M.M. Sil6eira et al. / Journal of Biotechnology 75 (1999) 99–103
103
Table 2 Effect of gluconic acid and sorbitol on the metabolism of glucose and gluconic acid by Z. mobilis (initial glucose concentration, 30 g l−1) Initial sorbitol/gluconic acid (g l−1)
Glucose consumption rate (g l−1 h−1) Residual glucose (%)a Residual gluconic acid (%)a a
0/0
55/59
146/157
255/274
328/353
92 0 –
25 0 0
11 0 27
0.40 89 98
0.14 96 99
After 8 h of process.
desirable. For this reason, their economical production using permeabilized cells should consider the reutilization of the GFOR-containing biomass in long-term operation processes as the two-stage continuous method described by Rehr et al. (1991). In that process, performed with Z. mobilis cells permeabilized with cetyltrimethylammonium bromide (CTAB) and immobilized in k-carrageenan, hardened with glutaraldehyde and polyethylenimine, a decrease of only 3.5% of the original enzymatic activity was observed after 75 days. On the other hand, with a feed concentration (glucose +fructose) of 200 g l − 1, relative low specific productivities (product per biomass hour) of 0.19 and 0.21 g g − 1 h − 1 for sorbitol and gluconic acid, respectively, were obtained. These low productivities were due to the concentration of glucose and fructose in the second stage of that bioprocess ( B10 g l − 1 of each substrate), considerably smaller than the Km of 77.2 g l − 1 fructose calculated by these authors for permeabilized and immobilized cells. The continuous process has as an additional problem: the low concentration of products obtained (B 100 g l − 1) means increasing costs with the removal of water in the downstream processing. Furthermore, considering that the most important commercial application of sorbitol is as a non-cariogenic sweetener, the contamination with fermentable sugars in the outflow is undesirable and would imply difficulties to carry out the downstream processing. Therefore, the use of free-untreated cells of Z. mobilis in a batch process for the biological production of sorbitol and gluconic acid, despite the lower yields obtained with them, become an attractive option. An inconvenient aspect of this
alternative would be the need for cell cultivation in each run. The cost of cell growth, however, could be compensated for by the ethanol produced during this phase. References Chun, U.H., Rogers, P.L., 1988. The simultaneous production of sorbitol and gluconic acid by Zymomonas mobilis. Appl. Microbiol. Biotechnol. 29, 19 – 24. Erzinger, G.S., 1996. Influeˆncia da concentrac¸a˜o de glicose e de etanol sobre a atividade de glicose – frutose oxidoredutase em Zymomonas mobilis ATCC 29191, Master thesis, Faculty of Pharmaceutical Sciences, University of Sa˜o Paulo, Brazil, 99 pp. Rehr, B., Wilhelm, C., Sahm, H., 1991. Production of sorbitol and gluconic acid by permeabilized cells of Zymomonas mobilis. Appl. Microbiol. Biotechnol. 35, 144 – 148. Sahm, H., Loos, H., Rehr, B., Sprenger, G., 1995. Production of sorbitol and gluconic acid by Zymomonas mobilis. Meded. Fac. Landbouwwet. Rijksuniv. Gent 59 (4b), 2403 – 2410. Silveira, M.M., Lopes da Costa, J.P.C., Jonas, R., 1994. Processo de produc¸a˜o e recuperac¸a˜o de sorbitol e a´cido glucoˆnico ou gluconato, Patente de Invenc¸a˜o PI 9403981-0, INPI, Brazil. Silveira, M.M., Sales, R., Lemmel, C., Jonas, R., 1995. Glucose – fructose activity in six strains of Zymomonas mobilis. Arq. Biol. Tecnol. 38, 619 – 622. Strohdeicher, M., Schmitz, B., Bringer-Meyer, S., Sahm, H., 1988. Formation and degradation of gluconate by Zymomonas mobilis. Appl. Microbiol. Biotechnol. 27, 378 – 382. Wisbeck, E., Silveira, M.M., Ninow, J., Jonas, R., 1997. Evaluation of the flocculent strain Zymomonas mobilis Z1-81 for the production of sorbitol and gluconic acid. J. Basic Microbiol. 37, 445 – 449. Zachariou, M., Scopes, R.K., 1986. Glucose – fructose oxidoreductase, a new enzyme isolated from Zymomonas mobilis that is responsible for sorbitol production. J. Bacteriol. 167, 803 – 809.