Co-expression of l -arabinose isomerase and d -glucose isomerase in E. coli and development of an efficient process producing simultaneously d -tagatose and d -fructose

Enzyme and Microbial Technology 40 (2007) 1531–1537

Co-expression of l-arabinose isomerase and d-glucose isomerase in E. coli and development of an efficient process producing simultaneously d-tagatose and d-fructose Moez Rhimi, Ezzedine Ben Messaoud, Mohamed Ali Borgi, Khalifa Ben khadra, Samir Bejar ∗ Laboratoire d’Enzymes et de M´etabolites des Procaryotes, Centre de Biotechnologie de Sfax BP “K”3038 Sfax, Tunisie Received 31 May 2006; received in revised form 25 October 2006; accepted 30 October 2006

Abstract To develop a feasible enzymatic process for the concomitant d-tagatose and d-fructose production, the thermostable l-arabinose isomerase of Bacillus stearothermophilus US100 (l-AI US100) and the mutant d-glucose isomerase obtained from that of Streptomyces SK (SKGI-A103G) were successfully co-expressed in Escherichia coli HB101 strain. The recombinant cells were immobilized in alginate beads and showed, similarly to the free cells, optimal temperatures for d-galactose and d-glucose isomerisation of 80 and 85 ◦ C, respectively. The two isomerases were optimally active at pH 7.5. Cell entrapment significantly enhanced the acidotolerance of the two isomerases, as well as their stability at high temperatures. To perform simultaneous isomerisation of d-galactose and d-glucose at 65 ◦ C and pH 7.5 in packed-bed bioreactor, cells concentration, dilution rate, productivity and bioconversion rate were optimized to be 32 g/l, 2.6 h−1 , 3 g/l h and 30%, respectively. © 2006 Elsevier Inc. All rights reserved. Keywords: l-Arabinose isomerase; Glucose isomerase; Co-expression; Immobilization; d-Fructose; d-Tagatose

1. Introduction The d-tagatose, an isomer of d-galactose, is a novel natural ketohexose having a taste and physical properties similar to sucrose [1]. It is also an anti-biofilm agent, which can be used as a cytoprotective supplement for the storage of organs to reduce the reperfusion injury [2,3]. Interestingly, d-tagatose is an antihyperglycemiant factor with a very low calorie carbohydrate and bulking agent [4–6]. It was the subject of recent interests in food and drug industries and was considered as safe and low calorie substrate in the United States [7]. The d-fructose ketohexose is twice as sweet as d-glucose playing an important role as a diabetic sweetener, according to its slow intestinal absorption. Therefore it does not influence the glucose level in blood [8]. d-fructose production cost, in HFCS using xylose isomerase commonly known as glucose isomerase Abbreviations: B, Bacillus; l-AI, l-arabinose isomerase; GI, glucose isomerase; HPLC, high performance liquid chromatography; HFCS, high fructose corn syrup ∗ Corresponding author. Tel.: +216 74 44 04 51; fax: +216 74 44 04 51. E-mail address: [email protected] (S. Bejar). 0141-0229/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.enzmictec.2006.10.032

(EC 5.3.1.5), was 10–20% lower than that of sucrose. Additionally, this natural sweetener is preferred in the food industry since it does not provoke the crystallisation problem as it is the case with sucrose [8]. The use of the d-tagatose was limited due to its high cost. Recently, an enzymatic procedure of d-tagatose production has been developed as an easy feasible and environmentally clean procedure. This process was based on the use of a combined immobilized ␤-glycosidase derived from Sulfolobus solfataricus and l-arabinose isomerase (EC 5.3.1.4) of Thermoanaerobacter mathranii, in the same reactor, allowing the direct conversion of lactose into d-tagatose and d-glucose [9]. In fact, lactoserum is considered as cheaper source for the d-galactose production at industrial scale due to its large abundance in the by-milk products such as: cheese and whey. Actually, the d-tagatose production procedure from lactoserum leads to a mixture of d-tagatose and d-glucose followed by a separation step of d-tagatose from d-glucose. Alternatively, it will be more profitable to convert the residual d-glucose into d-fructose using the xylose isomerase, which is a commercial enzyme with high relevance, having the ability to convert dglucose into d-fructose [10,11]. Therefore, the co-expression

1532

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537 was cloned into the linearized pMR5/SalI/BamHI plasmid. This leads to pMR20 plasmid, containing the araA US100 gene under the control of the inducible Plac promoter and the xylA-A103G gene downstream of its own constitutive promoter. E. coli HB101 strain was transformed by pMR20 plasmid and recombinant cells were selected on McConkey agar plates.

2.3. Cell crude extract preparation

Fig. 1. Schematic diagram of enzymatic tagatose and fructose production process: (1), feed tank containing substrates solution; (2), thermostated column reactor containing the alginate beads; (3), peristaltic pump sucking up the fluid through the column; (4) products recuperation tank.

of these two isomerases offers an excellent opportunity toward innovative aspect for the concomitant production of d-fructose and d-tagatose (Fig. 1). In a previous work, we have reported the gene cloning and the characterization of the l-arabinose isomerase extracted from Bacillus stearothermophilus US100 strain (LAI-US100). This thermoactive metallic ions independent enzyme had an optimal activity at 80 ◦ C and pH 7.5 [12]. In addition to the study of the structure-function relationship of the glucose isomerase (SKGI) extracted from thermophile Streptomyces SK strain, we have obtained the SKGI-A103G mutant enzyme [13]. The SKGI-A103G showed approximately the same biochemical characteristics, compared to the LAI-US100, with a maximal activity at pH 7.5 and 85 ◦ C [12,13]. The present paper concerns the construction of a recombinant Escherichia coli strain, efficiently co-expressing the l-AI US100 and the SKGI-A103G genes. A comparative study between free and immobilized cells was investigated, as well as the optimization of operating conditions for the production of d-tagatose and d-fructose, using immobilized cells in packed-bed bioreactor. 2. Materials and methods 2.1. Bacterial strains, plasmids and media E. coli HB101 (F− hsdS20 ara-1 recA13 proA12 lacY1 galK2 rpsL20 mtl1 xyl-5) was used in this work as host strain. Culture of E. coli strains was done in Luria Bertani (LB) medium [14]. McConkey agar medium from Sigma (Steinheim, Germany) was used for the identification of colonies having lAI and GI activities [15]. These media were supplemented, when necessary, with ampicillin (100 ␮g/ml) and IPTG (isopropyl-␤-d-thiogalactopyranoside) at 160 ␮g/ml. pBMA5 is the plasmid harboring the xylA-A103G gene encoding the mutated glucose isomerase SKGI-A103G under the control of its own promoter [13]. pMR5 contains the encoding gene of l-AI US100 downstream of the Plac promoter [12]. pMR20 construction, carrying both genes encoding the l-AI US100 and SKGI-A103G is described in this work.

2.2. DNA manipulation and co-expression procedures Plasmid DNA preparation, digestion with restriction endonucleases and separation of fragments by agarose gel electrophoresis were performed as described by Sambrook et al. [14]. The DNA purification was carried out using the GFXTM PCR DNA and Gel Band Purification Kit (Amersham Bioscience), on the basis of the manufacturer’s instructions. In order to co-express the two enzymes, the pMR5 plasmid was digested by SalI and BamHI located downstream of the l-AI US100 encoding gene. Taking advantage of the compatibility of SalI and XhoI restriction enzymes, the XhoI-BamHI DNA fragment, harboring the xylA-A103G gene from pBMA5,

E. coli HB101/pMR20 strain was grown in LB medium with ampicillin (100 ␮g/ml) and IPTG (160 ␮g/ml). Cells were harvested by centrifugation (7500 × g, for 10 min) and the pellets were suspended in 100 mM 3-[Nmorpholino] propanesulfonic acid (MOPS) buffer (pH 7.5) supplemented, where indicated, with 1 mM MnSO4 and 0.2 mM CoCl2 . Then, cells suspension was incubated for 1 h on ice in presence of 5 mg/ml lysozyme, 100 mM PMSF (phenylmethane-sulfonyl fluoride) and 2 ␮g/ml pepstatin A. Cell disruption was carried out by sonication at 4 ◦ C for 6 min (pulsations of 3 s, amplify 90) using a Vibra-CellTM 72405 Sonicator and cell debris were removed by centrifugation (30,000 × g, for 30 min at 4 ◦ C). The obtained supernatant constitutes the cytoplasmic crude enzyme extract.

2.4. Preparation of cells and immobilization method The immobilization was carried out by mixture of 10 ml of cells with 10 ml of 6.0% (w/v) sodium alginate (Fluka). Then, the obtained mixture was dropped slowly within a cooled and stirred 200 mM barium chloride (BaCl2 ) solution (200 ml). Integrity of beads was improved by renewing the BaCl2 bath for an overnight incubation at 4 ◦ C. Then, beads were washed twice in 100 mM MOPS Buffer (pH 7.5) at 4 ◦ C for 1 h. The average bead size was approximately 0.7 mm.

2.5. Activities assays In standard conditions, free l-arabinose isomerase activity was measured by the determination of the amount of formed d-tagatose. The reaction mixture contained 50 ␮l of the crude enzyme preparation at a suitable dilution and 10 g/l of d-galactose in a final volume of 1 ml MOPS buffer 100 mM (pH 7.5). The reaction mixture was incubated at 80 ◦ C throughout 10 min. The free d-glucose isomerase activity was assayed in the same MOPS buffer using 10 g/l d-glucose as substrate and the reaction was incubated during 30 min at 85 ◦ C in presence of 5 mM Mg2+ . Activities in immobilized and free cell preparations were determined in 100 mg of alginate beads or 50 mg of cells, respectively suspended in 100 mM MOPS (pH 7.5). Isomerization was performed in 2 ml reaction mixture under the same assay conditions described above during 30 min. Samples were then cooled in ice to stop the reactions. The generated ketoses (d-tagatose or d-fructose) were determined by cysteine carbazole sulfuric-acid method, and the absorbance was measured at 560 nm [16]. One unit of glucose isomerase or l-arabinose isomerase activity is defined as the amount of enzyme needed to produce 1 ␮mol of d-fructose or d-tagatose, respectively, per min under the assay conditions. In the case of immobilized cells in packed-bed bioreactor, the d-glucose and d-galactose are prepared in the same feed solution. The total produced ketoses (d-tagatose and d-fructose) were determined by means of the cysteine carbazole sulfuric-acid method. The amount of d-fructose was deduced by measuring the residual d-glucose, after isomerisation, by glucose-oxidase (GOD-PAP, kit Biomaghreb). Taking account of the known formed d-fructose quantity, and the total generated ketoses, the amount of the produced d-tagatose was then determined. Conversion rate represent the percentage ratio between the formed ketose concentration obtained at stationary functioning and the feed substrates concentration.

2.6. Cell concentration, temperature, pH and metallic ions effects To optimize the operating conditions, the effect of cell concentrations, temperature, pH, and metal ion effects were investigated. The cells concentration inside the alginate capsules were varied between 20 and 50 g/l and the GI and l-AI activities were measured. Thermoactivity profiles were studied using free and immobilized cells at temperatures comprised between 50 and 90 ◦ C. pH

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537

1533

profiles were obtained by measuring activities at pH ranging from 6.0 to 9.0 (6.0–8.0 with MOPS buffer and 8.0–9.0 with glycine buffer). Diverse divalent ions were supplemented to a final concentration of 0.1 mM, in order to explore their influence on activities. In fact, each enzyme was pre-incubated during 30 min in presence of metallic ions (MgCl2 , CoCl2 , CaCl2 , MnSO4 , CuCl2 and BaCl2 ) followed by the determination of l-AI and GI activities under standard conditions.

2.7. Thermal and pH stabilities The thermal stability of isomerase activities was monitored after incubation of free and immobilized enzymes at temperatures comprised between 65 and 80 ◦ C. The pH stabilities were studied by incubation of enzymes at different pH ranged between 6.0 and 8.5 at 65 ◦ C (5–8 with MOPS buffer and 8–11 with glycine–NaOH buffer). Samples were then withdrawn, at periodic intervals, for the determination of the residual activity under standard conditions.

2.8. Operating in a packed-bed bioreactor Beads were packed in a thermostat column, with a diameter/height of 2.7/24 cm, used for d-tagatose and d-fructose production. An equimolar dgalactose and d-glucose solution with 10 g/l each in 100 mM MOPS buffer (pH 7.5) containing 5 mM Mg2+ , was fed into the bioreactor at a dilution rate of 1.2 h−1 using a peristaltic pump (Gilson-MINIPLUS3). The converted d-tagatose and d-fructose were monitored from the effluent. The reaction temperature was maintained at 65 ◦ C using a water circulator.

2.9. Optimization of the operating conditions in column A solution of d-galactose and d-glucose at 10 g/l each, in 100 mM MOPS buffer (pH 7.5) was fed into the bioreactor maintained at 65 ◦ C. The effect of the dilution rate on the conversion level and productivity of d-tagatose and d-fructose was explored.

2.10. Analysis of the reaction products by HPLC The d-tagatose and d-fructose production was also confirmed by highperformance liquid chromatography analysis (HPLC) using Polypore CA column (250 mm × 4.6 mm). The products were separated by isocratic elution with water at a flow rate of 0.3 ml/min and detected with a refractive index detector (SHIMADZU, RID-10A). A Solution of d-galactose, d-glucose, d-fructose and d-tagatose at 10 g/l each were used as standards.

3. Results and discussion 3.1. Co-expression of the l-AI US100 and the SKGI-A103G Recombinant HB101/pMR20, obtained as described in Section 2, co-expressing the two isomerases was used for a liquid culture followed by the monitoring of the l-AI and GI activities in crude extract. This new constructed strain displayed a high l-AI and GI activities of 41 and 21 U/mg, respectively, confirming their efficient co-expression. This result is approximately similar to those obtained separately with HB101/pMR5 and HB101/pBMA5 having respectively 43 U/mg of l-AI activity and 22 U/mg of GI activity. 3.2. Cells concentration, temperature, pH and divalent ions effects The effect of cell concentration, inside the immobilized capsules, on the d-galactose or d-glucose bioconversion was

Fig. 2. Effect of cell concentration on the relative activities of l-AI US100 () and SKGI-A103G (). Assays were carried out in standard conditions, and optimal activities were defined as 100%.

studied. As it is shown in the Fig. 2, the effect of cell concentration on the l-AI US100 and the SKGI-A103G activities were similar. Indeed, the production of d-tagatose and d-fructose increased proportionally to the cell concentration, until a maximum of 32 g/l, above which the drop of ketoses production is assigned to steric hindrance that may occur at high cell concentration. As it is shown in Fig. 3 (A and B), the maximal activities of l-AI US100 and SKGI-A103G were reached at 80 and 85 ◦ C, respectively. These optimal temperatures were not affected by immobilization, which contrast with previous work that reported thermoactivity improvement of entrapped l-AI and GI [17,18]. The investigation of the pH effect on the isomerase activities showed that the two enzymes were optimally active at pH 7.5. The l-AI US100 in immobilized cells was more acidotolerant, comparing to that in free state. Thus, the enzyme retained 85% for the immobilized cells against only 72% for free cells, of its activity at pH 6.0 (Fig. 4A). This phenomenon is more noticeable in the case of SKGI-A103G, which retained 60 and 40% of its activity at pH 6.0, respectively for the immobilized and free forms (Fig. 4B). This behaviour towards acidic pH takes more importance when one considers that isomerisation under alkaline conditions generates undesirable compounds such as dpsicose and d-mannose [19,20]. Hence, immobilization appears as promising tool to improve physico-chemical features of this kind of isomerases, to be more suitable for biotechnological uses. The l-AI US100 was previously described as an independent metallic ions enzyme for its activity and requires only 0.2 mM Co2+ and 1 mM Mn2+ for its optimal thermostability [12]. Immobilization does not provoke any change in the enzymes independency toward divalent ions, as it is illustrated in the (Fig. 5A). Only, the Cu2+ ion appears to be an inhibitor of the l-AI US100. This interesting l-AI US100 behaviour towards metallic ions for its activity contrast with that of Geobacillus stearothermopilus T6, which showed an increase of activity after

1534

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537

Fig. 3. Effect of temperature on the activities of l-AI US100 (A) and SKGIA103G (B) in free () and immobilized () cells. Assays were carried out in standard conditions, and optimal activities were defined as 100%.

Mn2+ addition [17]. This fact can be explained by the absolute dependency of this earlier enzyme for metallic ions for its activity when compared to the l-AI US100 [12,21]. The SKGIA103G enzyme requires only Mg2+ for its activity; in fact the activity in presence of this ion was increased to 145% (Fig. 5B), since this divalent ion is implicated in the mechanism of glucose isomerization [22]. We have also confirmed the inhibitor role of Ca2+ on the SKGI-A103G activity (Fig. 5B). 3.3. Thermal and pH stabilities The stability is a crucial criterion, which influences the use of the enzymes at industrial scale. The immobilization is a mean, which offers an excellent opportunity to improve enzyme features and their effective reuse. The study of the l-AI US100 and SKGI-A103G demonstrates that comparatively to free cells, the immobilized cells were significantly more thermostable. Indeed, the half-life durations at 80 ◦ C were 1 and 11 h for the SKGI-A103G in the free and immobilized states, respectively. At the same temperature the l-AI US100 had half-life times of 0.5 and 9 h for the free and immobilized cells, respectively (Table 1). These results provide evidence that the immobilization clearly improves the thermal stability of the two isomerases. As it is illustrated on the Table 1, the two isomerases display a decrease on their thermal stability at temperatures over 65 ◦ C. This fact is explained by the absence of metallic ions (Co2+ and Mn2+ ) implicated on the thermostability enhancement of the l-

Fig. 4. Effect of pH on the activities of l-AI US100 (A) and SKGI-A103G (B) in free () and immobilized () cells. Assays were carried out in standard conditions, and optimal activities were defined as 100%.

AI US100 and GI-SK A103G [12,13]. These cofactors when introduced during the isomerization process should be removed from the end products, which considerably decrease the production cost-effectiveness. Consequently, it is of interest to operate under conditions where these enzymes are independent towards these metallic ions for their thermal stability. The effect of pH on free and immobilized cell enzymes stability was investigated. This study showed that the half-life durations corresponding to the l-AI US100 activity at pH 6 Table 1 Half-lives (t1/2 , h) of the l-AI US100 and the SKGI-A103G in the free and immobilized cells at different temperatures Temperature (◦ C) 65

70

75

80

Half-lives (h) l-AI US100 Free cells Immobilized cells

70 123

5 16

0.75 11

0.5 9

SKGI-A103G Free cells Immobilized cells

102 143

12 21

5 15

Initial activities were defined as 100%.

1 11

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537

1535

Fig. 6. Effect of the dilution rate on the productivity (䊉) and the bioconversion rate () in packed bed-bioreactor. Assays were carried out at 65 ◦ C and pH 7.5 using 10 g/l d-galactose and d-glucose each.

However, at the same pH, the l-AI US100 displayed half-lives of only 26 and 36 h in the free and immobilized systems. Consequently the immobilization in alginate beads appears to be efficient on the stabilisation of the two isomerases at extreme conditions of pH and temperature, which could enhance their industrial application. 3.4. Study of the operating parameters and performance in the packed-bed bioreactor

Fig. 5. Effect of different metallic ions on the activities of l-AI US100 (A) and SKGI-A103G (B) in free () and immobilized () cells. Assays were carried out in standard conditions without addition of Mg2+ for the SKGI-A103G. (C) Denotes the control assays corresponding to the activities measured in absence of metallic ions.

and 6.5 of the immobilized cells were approximately two- and three-fold greater than those corresponding to free cells, respectively (Table 2). For the immobilized SKGI-A103G activity, the half-life durations were 1.2-fold higher than that in free state at pH 6.0 and 6.5 (Table 2). At pH 8.5, the SKGI-A103G was stable since the enzyme showed half-life durations evaluated to 52 and 89 h under the free and immobilized state, respectively. Table 2 Half-lives (t1/2 , h) of the l-AI US100 and the SKGI-A103G in the free and immobilized cells at different pH pH 6.0

6.5

7.0

7.5

8.0

8.5

Half-lives (h) l-AI US100 Free cells Immobilized cells

12 28

21 69

39 105

70 123

68 121

26 36

SKGI-A103G Free cells Immobilized cells

32 39

38 47

56 92

102 143

73 123

52 89

Incubation temperature was 65 ◦ C. The initial activities were defined as 100%.

In order to study the operating conditions in the bioreactor, the conversion rate and productivity were investigated with various dilution rates (0.6–6.0 h,−1 ) at 65 ◦ C and pH 7.5. As it is illustrated in the Fig. 6, a dilution rate of 0.6 h−1 allowed the highest conversion rate, evaluated to 44% (including the two ketohexoses), and the lowest productivity of 0.85 g/l h. A dilution rate of 6 h−1 had the smallest conversion rate, estimated to 18%, combined to a maximal productivity of 4.32 g/l h. Consequently, the decrease of the dilution rate allows the enhancement of the conversion rate, which is explained by the augmentation of the sojourn time course in the column, providing greater duration of contact between substrates and enzymes. However, the productivity increased as soon as the dilution rate augmented. This is explained by the decrease of the cycle time as the dilution rate increase. Hence, the increase of flow rate was generally limited by both velocities of mass transfer and isomerization reaction. The most favourable dilution rate providing optimal productivity and optimal conversion rate coincide with the intersection of the two curves of the Fig. 6. Hence the optima of dilution rate, productivity and conversion rate were evaluated to be 2.6 h−1 , 3.0 g/l h and 30%, respectively. The monitoring of the d-glucose and d-galactose bioconversion under previously determined optimal conditions showed that the formed ketoses were composed of nearly 10% and 20% of d-fructose and d-tagatose, respectively. These results are in accordance with spectra obtained through HPLC analysis, as it is shown in the Fig. 7. 3.5. Reuse stability The reuse stability of immobilized enzymes is one of the mainly essential factors influencing the use of an immobilized

1536

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537

This recombinant strain was able to convert simultaneously the d-glucose and the d-galactose into d-fructose and d-tagatose, respectively. The newly constructed strain, co-expressing the two isomerases, is certainly of importance due to its potential use with a ␤-galactosidase activity, to produce a d-tagatose and d-fructose containing syrup from lactoserum, an abundant and cheap disaccharide. Acknowledgement This research was supported by the Tunisian Government “Contrat Programme CBS-LEMP”. References

Fig. 7. HPLC analysis of the isomerization products. (A) Standard solution of d-galactose, d-glucose, d-fructose and d-tagatose at 10 g/l each in MOPS buffer 100 mM (pH 7.5); (B) reaction medium at initial time; (C) isomerization products after 8 h of reaction at 65 ◦ C. 1: d-glucose with a retention time of 6.1 min, 2: d-galactose with a retention time of 6.6 min, 3: d-fructose with a retention time of 7.3 min and 4: d-tagatose with a retention time of 8.6 min.

enzyme system. The immobilized enzymes were recycled and used several times, and then the isomerases activities were monitored. After each cycle the beads were washed thoroughly with 100 mM MOPS buffer (pH 7.5) until there was no trace of ketoses. The Fig. 8 shows the effect of the repeated use of immobilized E. coli HB101/pMR20 cells, showing that after eight cycle of use more than 70% of the initial activity was retained. In this paper we report, for the first time, the efficient coexpression of an l-AI and a GI in the HB101 E. coli strain.

Fig. 8. Reuse stability of the l-AI US100 () and SKGI-A103G () activities immobilized in alginate beads. Initial activities were defined as 100%.

[1] A.W. Mazur. Functional sugar substitutes with reduced calories. EP 341062 (1989). [2] G.V. Levin, Y. Lu. d-Tagatose as an anti-biofilm agent. US patent 20,030,103,912 (2003). [3] Paterna JC, Boess F, Staubli A, Boelsterli UA. Antioxidant and cytoprotective properties of d-tagatose in cultured murine hepatocytes. Toxicol Appl Pharm 1998;148:117–25. [4] Livisey G, Brown JC. d-Tagatose is a bulk sweetener with zero energy determined in rats. J Nutr 1996;126:1601–9. [5] Levin GV, Zehner LR, Saunder JP, Beadle JR. Suguar substitutes: their energy values, bulk characteristics and potential health benefits. Am J Clin Nutr 1995;62:1161–8. [6] L.R. Zehner, G.V. Levin, J.P. Saunder, J.R. Beadle. d-Tagatose as antihyperglycemiant agent. US patent 5,356,879 (1994). [7] Lee DW, Jang HJ, Choe EA, Kim BC, Lee SJ, Kim SB, et al. Characterization of a Thermostable l-arabinose (d-galactose) isomerase from the hyperthermophilic eubacterium Thermotoga maritima. Appl Environ Microb 2004;70:1397–404. [8] Antrim RL, Colilla W, Schnyder BJ. Glucose isomerase production of high fructose syrups. Appl Biochem Bioeng 1979;2:97–155. [9] Jorgensen F, Hansen OC, Stougaard P. Enzymatic conversion of d-galactose to d-tagatose: heterologous expression and characterisation of a thermostable l-arabinose isomerase from Thermoanaerobacter mathrnaii. Appl Microbiol Biotechnol 2004;64:816–22. [10] Boeshale HS, Rao MB, Deshpande VV. Molecular and industrial aspects of glucose isomerase. Microb Rev 1996;60:280–300. [11] Bandlish RK, Hess JM, Epting KL, Vieille C, Kelly RM. Glucose tofructose conversion at high temperatures with xylose (glucose) isomerases from Streptomyces murinus and two Thermotoga species. Biotechnol Bioeng 2002;80:185–94. [12] Rhimi M, Bejar S. Cloning, purification and biochemical characterization of metallic-ions independent and thermoactive l-arabinose isomerase from the Bacillus stearothermophilus US100 strain. Biochim et Biophys Acta 2006;1760:191–9. [13] Borgi MA, Srih BK, Ben Ali M, Mezghani M, Tranier S, Haser R, et al. Glucose isomerase of the Streptomyces sp. SK strain: purification, sequence analysis and implication of alanine 103 residue in the enzyme thermostability acidotolerance. Biochimie 2004;86:561–8. [14] Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, New York, USA: Cold Spring Harbor Laboratory Press; 1989. [15] Sedlak M, Ho NWY. Expression of E. coli araBAD operon encoding enzymes for metabolizing l-arabinose in Saccharomyces cerevisiae. Enzyme Microb Technol 2001;28:16–24. [16] Dishe Z, Broenfreund E. A new spectrophotometric method for the detection of keto sugars and trioses. J Biol Chem 1951;192:583–7. [17] Jung ES, Kim HJ, Oh DK. Tagatose production by immobilized recombinant Escherichia coli cells containing Geobacillus stearothermophilus l-arabinose isomerase mutant in a packed-bed bioreactor. Biotechnol Prog 2005;21:1335–40.

M. Rhimi et al. / Enzyme and Microbial Technology 40 (2007) 1531–1537 [18] Converti A, Del Borghi M. Simultaneous effects of immobilization and substrate protection on the thermodynamics of glucose isomerase activity and inactivation. Enzyme Microb Technol 1997;21:511–7. [19] Lee SJ, Lee DW, Choe EA, Hong YH, Kim SB, Kim BC, et al. Characterization of a Thermoacidophilic l-arabinose isomerase from Alicyclobacillus acidocaldarius: role of lys-269 in pH optimum. Appl Environ Microbiol 2005;72:7888–96. [20] Bucke C. Glucose transforming enzymes. In: Fogarty W, editor. Microbial Enzymes and Biotechnology. London: Applied Science Publishers; 1983. p. 93–127.

1537

[21] Lee DW, Choe EA, Kim SB, Eomc SH, Hong YH, Lee SJ, et al. Distinct metal dependence for catalytic and structural functions in the l-arabinose isomerases from the mesophilic Bacillus halodurans and the thermophilic Geobacillus stearothermophilus. Arch Biochem Biophys 2005;434:333–43. [22] Asboth B, Naray-Szabo G. Mechanism of action of d-xylose isomerase. Curr Protein Pept Sci 2000;1:237–54.