How to Produce D Tagatose by Fermentation Method？

D-tagatose is a six-carbon ketose and an isomer of D-galactose. Pure D-tagatose is not easy to find, as it is only found in the gum secreted by the Sterculia setigera tree. Small amounts of D-tagatose are also found in pasteurised milk, hot chocolate, cheese and cheese products. Since only about 20% of D-tagatose can be absorbed by the small intestine after consumption, it basically does not produce calories and does not cause blood sugar fluctuations. On the other hand, D-tagatose can prevent the formation of dental caries and can also be used by microorganisms in the intestine to promote the growth of certain probiotics [1]. In recent years, with the improvement of living standards and the prevalence of diseases such as obesity and diabetes, there have been increasing calls for D-tagatose to be used as a functional sweetener to replace sucrose, which has attracted a great deal of attention.

Beadle et al. [2] invented a process for producing D-tagatose from the isomeric D-galactose using Ca(OH)2 in 1991. D-galactose can be obtained by hydrolyzing lactose or the more affordable whey (a waste product from the production of dairy products), and the cost of the process is acceptable to the market. However, this method requires the use of large amounts of strong acid to neutralize the reaction solution, which is likely to cause wastewater pollution; moreover, there are many by-products of the reaction, which makes downstream separation difficult. Cheetham et al. [3] found that L-arabinose isomerase (L-AI enzyme, EC 5. 3. 1. 4) can catalyze the production of D-tagatose from D-galactose. In the following 20 years, through enzyme production screening, exogenous expression and characterization of different types of microorganisms, scholars have found a variety of L-AI enzymes suitable for the industrial production of D-tagatose L-AI enzymes suitable for the industrial production of D-tagatose; it has been found that they mainly have the characteristics of high conversion rate of D-galactose, good thermal stability of the enzyme, and the optimum catalytic pH of the enzyme tends to be acidic conditions suitable for industrial production (pH between 5.0 and 6.0) [4].

On the other hand, genetic engineering of the wild-type L-AI enzyme using molecular biology methods can also effectively enhance the industrial application potential of the wild enzyme; this is mainly reflected in the establishment of a structural model of the L-AI enzyme using computer molecular simulation technology and the rational design of important amino acid residues in the enzyme. In terms of catalytic processes, the batch conversion technology of immobilized cells/enzymes has achieved results; however, problems such as low substrate conversion rates, long reaction cycles, and short half-lives of enzyme activity remain bottlenecks in this technology. In addition, the urgent task of developing a safer food-grade L-AI enzyme exogenous expression host to replace recombinant Escherichia coli is imminent. Currently, the American company Spherix is conducting phase III clinical trials of D-tagatose as a treatment for type 2 diabetes [5].

If the trial is successful, D-tagatose will have broad prospects in the field of biomedicine and will also create an even more urgent demand for biological production processes. The author combines the research results of the research group in recent years on L-AI enzyme and the bioprocess production of D-tagatose to provide a review of the screening and application of L-AI enzyme, research on genetic engineering modification, and the application of immobilization technology, with the aim of providing a reference for the green production of new functional sweeteners such as D-tagatose.

1 Research on the source and properties of L-AI enzyme

Currently known L-AI enzymes have been obtained from prokaryotic microorganisms through cloning and foreign expression, including Bacillus subtilis [6], Lactobacillus plantarum [7], Acidothermus thermophilus [8], Geobacillus stearothermophilus [9] and Thermus thermophilus [10] (Table 1).

As can be seen from Table 1, it is possible that due to the very different living environments of these microorganisms, the catalytic properties of L-AI enzymes from different sources have also significantly diverged after long-term natural evolution.

1) The optimum temperature for the D-galactose isomerization reaction is different, and has been reported to range from 15 to 95 °C. Among them, L-AI enzymes from Shewanella sp. ANA-3 and Lactobacillus sakei 23K can stably catalyze at low temperatures of 4 °C [11-12], while Anoxybacillus The environment is very different. After long-term natural evolution, the catalytic properties of L-AI enzymes from different sources have also been significantly different.

1) The optimum temperature for the D-galactose isomerization reaction is different, and has been reported to range from 15 to 95 °C.

Among them, L-AI enzymes from Shewanella sp. ANA-3 and Lactobacillus sakei 23K can stably catalyze at low temperatures of 4 °C [11-12], while Anoxybacillus flavithermus L-AI enzymes can maintain enzyme activity at extreme temperatures of 95 °C [17]. In general, L-AI enzymes can be divided into three categories according to their optimum reaction temperature: mesophilic L-AI enzymes, thermophilic L-AI enzymes and hyperthermophilic ). Their optimum reaction temperatures are less than 60 °C, between 60 and 80 °C, and between 80 and 95 °C, respectively [18]. Among them, the thermophilic type is considered more suitable for the industrial production of D-tagatose, mainly because the Gibbs free energy required for the isomerization reaction from D-galactose to D-tagatose is high (4.96 kJ/mol), a relatively high catalytic temperature is required to obtain a high conversion rate. However, an excessively high temperature will instead lead to browning reactions, which will affect the quality of the product. Therefore, most scholars believe that a reaction temperature of 60 to 70 °C is most suitable [1, 4].

2) The optimum reaction pH varies, and has been reported to range from 5.0 to 10.5. Most L-AI enzymes have an optimum catalytic pH of between 7.5 and 8.5, but industrial production requires an optimum catalytic pH in the acidic range to avoid non-specific side reactions caused by alkaline reaction conditions, and to match the acidic pH required for lactose hydrolysis, avoiding repeated adjustments and simplifying the production process [1, 4].

3) Different requirements for metal ions. Most L-AI enzymes require metal ions to exert enzymatic activity and maintain thermal stability. Among them, Mn2 + and Co2 + ions are the most common. However, Co2 + cannot be used in the food industry, so L-AI enzymes that are highly dependent on Co2 + have relatively limited industrial applications [10]. Interestingly, the enzyme activity of L-AI from Bacillus halodurans and Bacillus stearothermophilus US100 was not reduced after EDTA treatment. The possible reasons are that these two enzymes do not depend on metal ions, or the active center of the enzyme has a strong binding ability with metal ions, and the metal ions are not easily detached [16].

4) There are significant differences in substrate specificity. For example, L-AI enzymes derived from Bacillus subtilis and Bacillus licheniformis only have the ability to catalyze L-arabinose [6, 13] and cannot catalyze D-galactose, which is very special in the L-AI enzyme family; they will be suitable for studying the substrate selectivity of L-AI enzymes.

4) Different catalytic efficiency for D-galactose. The catalytic efficiency (kcat/Km value) of almost all L-AI enzymes for L-arabinose is much higher than that for D-galactose, so the catalytic efficiency of L-AI enzymes for D-galactose is generally low. However, Cheng et al. [15] found that L-AI enzyme derived from Acidothermus cellulolytics ATCC43068 has a relatively high catalytic efficiency (kcat/Km = 9.3 mmol/(L·min)) for D-galactose. Rhimi et al. [12] obtained L-AI enzyme from Lactobacillus sakei 23K strain that can catalyze D-galactose at low temperatures and in acidic environments. Its kcat/Km value for D-galactose reaches 10.3 mmol/(L·min), which is the highest value reported in the literature so far.

For the industrial production of D-tagatose, if a thermophilic, acid-tolerant, metal-ion-independent L-AI enzyme with high catalytic efficiency for D-galactose can be screened, it will greatly facilitate technological progress. The research group to which I belong has screened a strain of Lactobacillus fermentum that produces L-AI enzyme from traditional pickles [19-20]. The L-AI enzyme encoding gene contained in this strain was found to be exogenously expressed, and the recombinant L-AI AI enzyme catalyzes D-galactose at an optimum temperature of 65 °C and an optimum pH of 6.5, with a catalytic efficiency of 9.02 mmol/(L·min), demonstrating certain potential for industrial application [14, 21].

2 Genetic engineering of L-AI enzyme

Given that there is room for improvement in the enzymatic properties of the L-AI enzyme currently used in D-tagatose production, efforts have been made to rationally modify it at the molecular level. In 2006, the crystal structure of the L-AI enzyme from Escherichia coli was successfully resolved (PDB database accession number: 2AJT/2HXG), which laid the structural biology foundation for the genetic engineering of the L-AI enzyme [22]. In recent years, directed evolution technology and site-directed mutagenesis guided by computer homology modeling have become effective means for L-AI enzyme modification. As important indicators of industrial application, the modification of enzyme activity, optimal reaction temperature, optimal reaction pH, and substrate selectivity have received more attention (Table 2).

Studies on the role of key amino acid residues have mainly focused on L-AI enzymes from Bacillus stearothermophilus. Rhimi et al. [23] performed a molecular simulation of the spatial structure of B. stearothermophilus US100 L-AI enzyme and site-directed mutated the 175th amino acid . The mutant (N175H) thus obtained a wider optimal reaction temperature range (50 to 65 °C) than the wild-type enzyme, and compared with the wild-type enzyme (80 °C), which has a higher optimal reaction temperature, the mutant is beneficial for avoiding browning side reactions during production. Oh et al. [27] results as the basis for further studies on site-directed mutations, and found that when the 408th amino acid of the G. sterothermophilus L-AI enzyme is changed from a neutral amino acid to a polar or basic amino acid, the optimal reaction pH of the mutant strain is significantly reduced (from 8.5 to 7.5).

It is speculated that this is because this position is close to the active center of L-AI enzyme, so the change in the polarity of the amino acid may lead to a change in the charge distribution of the microenvironment of the catalytic center, thereby enhancing the ability of L-AI enzyme to catalyze the substrate under acidic conditions. Oh et al. [28] found that the benzene ring of amino acid 164 and the size of the side chain of amino acid 475 have a significant effect on the ability of L-AI enzyme to catalyze D-galactose by site-directed mutagenesis of multiple important amino acid residues of G. thermodenitrificans L-AI enzyme. lactose. Its double mutant C450S-N475K has a 5-fold increase in catalytic efficiency for D-galactose, and the conversion rate of D-galactose increased from 46% to 58%. In 2010, P rabhu et al. [25] used homology modeling to analyze the amino acid residues in and near the catalytic center of B. licheniformis L-AI enzyme, and selectively mutated multiple sites. The results showed that the mutant (Y333A) completely lost the ability to catalyze L-arabinose. The author believes that the distance between the key amino acid residues in the catalytic active center and the substrate determines the catalytic ability. Therefore, the conserved amino acid residues near the catalytic center are important for the substrate specificity of the L-AI enzyme. This provides ideas for modifying the substrate specificity of the L-AI enzyme.

3 Production of D-tagatose using immobilization technology

As the most effective biological method for producing D-tagatose to date, the immobilized cell/enzyme technology for producing D-tagatose has been studied in depth, and a variety of production processes have been established (Table 3).

3.1 Production of D-tagatose using immobilized cell technology

The method of using sodium alginate to immobilize recombinant E. coli cells to produce D-tagatose has long been in the field of vision of scholars. This method is highly operable, reliable, and inexpensive, and has the potential for industrial application. Generally, after inducible expression, the recombinant E. coli cells can be used for the catalysis of D-galactose after being embedded with sodium alginate and treated with CaCl2. Hong et al. [35] established a stable and inexpensive technology for the production of D-tagatose using sodium alginate-immobilized recombinant E. coli cells. coli cells to produce D-tagatose.

It has a half-life of 43 days and produces a high-purity (>99%) D-tagatose after being separated by an ion exchange column. However, using Escherichia coli as a production strain for D-tagatose has safety risks; therefore, the development of a safer immobilized cell production technology using food-grade microorganisms could be a future research direction. The author used the food-grade Lactobacillus fermentum CGMCC2921 strain to produce immobilized cells for the production of D-tagatose by the method of sodium alginate embedding and glutaraldehyde cross-linking. Under the condition of adding boric acid, the conversion rate of D-galactose to D-tagatose can reach 60%, with a yield of 11.1 g/(L·h). There was no significant decrease in conversion rate during the eight conversion batches (totaling 192 h) [37]. However, in order to achieve the cost requirements of industrial production, in-depth research must be carried out on high-density cultivation and efficient expression of the production strain.

3.2 Production of D-tagatose using immobilized enzyme technology

Production of D-tagatose using immobilized enzyme technology can achieve high production intensity, but requires a cumbersome enzyme purification process. Kim et al. [31] used sodium alginate to encapsulate thermophilic L-AI enzyme to produce D-tagatose, with a yield of 13.3 g/(L·h). Ryu et al. [30] used a sodium alginate embedding method to immobilize the purified G. stearothermophilus L-AI enzyme, and then used a packed-bed bioreactor to produce D-tagatose. After investigating factors such as the optimal height-to-diameter ratio of the reaction bed, they successfully obtained a production rate of 54 g/(L·h), which is the highest value reported in the literature to date. Jung et al. [ 34] believe that the reason for the high D-tagatose yield obtained using immobilized enzyme technology is that after purification, the L-AI enzyme is not interfered with by other host proteins after immobilization, and the biocatalysis can proceed with the most ideal reaction trend; while the immobilized cells contain a large amount of host proteins, the presence of these proteins has a considerable steric hindrance effect, resulting in a much lower D-tagatose yield when using immobilized cells.

In addition, immobilized L-AI enzyme technology can also effectively improve the properties of L-AI enzyme. Liang et al. [38] used sodium alginate to embed L-AI enzyme derived from Tan. mathranii, and the optimal reaction temperature of the immobilized enzyme after embedding was significantly improved (from 60 °C to 75 °C). 5 to 8. 0 can maintain more than 80% of enzyme activity, and these data are much better than free enzymes. Jebors et al. [39] first used Noria / NoriaPG polymer materials to immobilize Lactobacillus sakei L-AI enzyme, and the stability of this enzyme at high temperatures and low pH conditions was improved compared to free enzymes.

4. Improving the yield of D-tagatose by changing the chemical reaction equilibrium

In the reaction of the L-AI enzyme isomer D-galactose, some substances with high affinity for the product D-tagatose (such as boric acid) can coordinate with it to dissociate it from the reaction, resulting in a decrease in the product in the solution and a shift in the chemical equilibrium towards product formation. Lim et al. [40 first used boric acid to increase the conversion rate of D-galactose to D-tagatose using purified thermophilic L-AI enzyme to 77.4%, which is the highest conversion rate obtained using L-AI enzyme to prepare D-tagatose. The D-tagatose-boric acid complex formed is soluble in methanol, and the boric acid can be removed by evaporation without affecting the purity of the product. Therefore, this method has great application prospects. Further research has shown that the catalytic efficiency of L-AI enzyme on the substrate is significantly improved in the presence of boric acid. Li et al. [17] reported that the catalytic efficiency (kcat/Km) of Anoxybacillus flavithermus L-AI enzyme on D-galactose increased from 5.16 mmol/(L·min) to 9.47 mmol/(L·min) with the addition of boric acid, an increase of 83.5%.

5 Production of rare sugar mixtures containing D-tagatose through the combined action of L-AI enzyme and commercial enzyme

Lactose is hydrolyzed to produce an equimolar mixture of glucose and D-galactose. Glucose is not involved in the production of D-tagatose and can therefore be used to produce other rare sugars. Jorgensen et al. [32] were the first to report the use of lactose as a raw material to produce a mixed syrup containing D-tagatose and fructose using immobilized thermophilic L- AI enzyme, β-galactosidase and commercial glucose isomerase to produce a mixed syrup containing D-tagatose and fructose from lactose, in order to fully utilize glucose to increase the added value of the product. Rhimi et al. [41] used sodium alginate to encapsulate genetically engineered bacteria co-expressing the thermostable L-AI enzyme and glucose isomerase, and successfully produced D-tagatose and fructose in a continuous batch using a biological packed bed with lactose hydrolysate as the raw material. However, the separation of the components in the mixed syrup still requires the support of downstream technology research, and is currently only in the exploratory stage.

6 Separation and purification of D-tagatose

The goal of downstream technology research is to establish an efficient process for the separation and extraction of D-tagatose. D-tagatose in the conversion solution can be separated by extraction with an organic solvent. Xu Guiqiang et al. [42] used phenylboronic acid-quaternary ammonium ionic liquid solvent extraction to separate D-tagatose, with an extraction rate of 65.7% and a recovery rate of 92.3%. However, considering the potential food safety risks and environmental pollution problems associated with organic solvent systems, this method may not be suitable for large-scale use. Since D-tagatose and D-galactose are different forms of aldopentoses, they can be separated using ion exchange resins. Huang Wenxia et al. [43] used Ca2+-type ion exchange resin to purify D-tagatose synthesized by chemical method, obtaining D-tagatose with a purity of 98% and a recovery rate of 83%. The desalted D-tagatose can be crystallized with ethanol to obtain pure D-tagatose. This method is simple to operate and highly feasible.

7 Operational methods and equipment requirements for the biological production of D-tagatose

The biological production of D-tagatose mainly uses whey as the raw material. A membrane separation device is used to remove impurities such as protein and salt and to concentrate the whey. A mixture of D-galactose and glucose is obtained by hydrolyzing the concentrated whey with immobilized lactase. The glucose in the mixture is then fermented by microorganisms (such as brewer's yeast), and the ethanol is distilled off. The remaining D-galactose is converted to D-tagatose by immobilized L-AI enzyme (or genetically engineered bacteria containing L-AI enzyme). The D-galactose that has not been completely converted can be separated and recovered using a cation exchange column. Evaporation and concentration of the conversion solution containing only D-tagatose can be used for product crystallization and drying. The equipment involved in the entire process includes membrane filtration devices, fermenters, centrifuges (or plate and frame filter presses), distillation equipment, ion exchange resins, crystallization equipment, dryers, etc. Taking Kraft Foods Inc.'s D-tagatose bioprocess production method as an example [44], the specific operating methods and equipment usage are shown in Figure 1.

8 Problems and development prospects of the biological production of D-tagatose

As a green and low-carbon production technology, the biological production process of D-tagatose still cannot replace the existing chemical catalytic method, mainly for the following reasons:

1) High process cost. In terms of current technology, the highest yield of D-tagatose is obtained by using immobilized L-AI enzyme, but the enzyme purification process requires sophisticated purification equipment and complex process control, which is expensive to maintain and labor-intensive, and the technical threshold is relatively high. The production of D-tagatose using immobilized recombinant E. coli cells is simple to operate and the method is mature. The disadvantage is that the yield is low, and there are hidden dangers in the food safety of the E. coli host. However, this method still has the most industrial prospects.

2) The reaction cycle of the biological method for preparing D-tagatose is relatively long. The time required to produce D-tagatose using batch conversion is generally 24 h/batch, or even up to 168 h/batch, and a high temperature environment needs to be maintained throughout the process. Compared with a simple chemical catalytic method, this has no advantages in terms of energy consumption or reaction cycle. Although the production of D-tagatose using the room temperature L-AI enzyme is relatively inexpensive in terms of energy consumption, it is limited by the low conversion rate (20% to 30%), and has little practical application value.

3) Lack of development and utilization of food-grade L-AI enzyme expression hosts. Since the main future application of D-tagatose is in the food and pharmaceutical fields, it is necessary to develop food-grade (e.g., GRAS certified) microorganisms as biocatalytic vectors. Potential candidates include Bacillus subtilis, Gluconobacter oxydans, Saccharomyces cerevisiae, etc. Kim et al. [45] expressed the Geobacillus sp. L-AI enzyme in a GRAS-certified microbial host to produce D-tagatose, but the yield is unknown. It is worth mentioning that Rhimi et al. [46] expressed the B. stearothermophilus US100 L-AI enzyme in the probiotic L. bulgaricus and S. thermophilus strains. The new strains were able to produce D-tagatose during the fermentation of milk, which resulted in yogurt that maintained its flavor and reduced calories.

4) There is insufficient understanding of the catalytic mechanism of L-AI enzyme, making it difficult to effectively improve the catalytic properties of the enzyme. Due to the very limited structural data of L-AI enzyme, it is difficult to use structural biology to analyze the catalytic mechanism and attempt enzyme engineering modification; as a more accurate means, protein crystallization technology is of great importance for the study of the substrate catalytic process of L-AI enzyme. Therefore, the crystal structure analysis of L-AI enzyme will be an important topic in the future.

Of course, the development of an industrial process for the production of D-tagatose has great potential. Compared to chemical production methods, the advantages of biological methods include the ease of product extraction, high catalytic specificity and socially accepted food safety. The blueprint for the development of the core technology has also been revealed, which is to find an L-AI enzyme that can efficiently produce D-tagatose and establish an inexpensive and safe production process that can efficiently produce D-tagatose. It is believed that with the continuous deepening of relevant research, the biological process of using L-AI enzyme to prepare D-tagatose can mature.

Reference:

［1 ］  Oh  D  K．Tagatose : properties，applications，and  biotechnological processes［J］．Appl Microbiol Biotechnol，2007，76 : 1-8.

［2 ］  Beadle  J R，Saunder J P，Wajada T J．Process for manufacturing tagatose : US，500261［P］．1991-3-26.

［3 ］  Cheetham P  S J，Wootton  A N．Bioconversion of D-galactose into D-tagatose［J］．Enzyme Microb Technol，1993，15 : 105-108.

［4 ］  Boudebbouze   S ，Maguin   E ，Rhimi   M．Bacterial   L-arabinose isomerases : industrial  application  for  D-tagatose  production［J］.  Recent Pat DNA Gene Seq，2011，5 (3) : 194-201.

［5 ］  Lu Y，Levin  G V，Donner T  W．Tagatose，a new antidiabetic and obesity control  drug［J］．Diabetes  Obes Metab，2008，10  ( 2 ) :

109-134.

［6 ］  Kim  J  H，Prabhu  P，Jeya   M，et   al．Characterization  of  an  L- arabinose isomerase  from  Bacillus  subtilis［J］．Appl   Microbiol  Biotechnol，2009，85 (6) : 1839-1847.

［7 ］  Chouayekh  H，Bejar W，Rhimi M，et al．Characterization of an L- arabinose isomerase from the Lactobacillus plantarum NC8  strain  showing pronounced  stability  at  acidic pH［J］．FEMS  Microbiol  Lett，2007，277 :260-267.

［8 ］  Lee  S  J，Lee   D  W，Choe  E  A，et  al．Characterization  of  a thermoacidophilic   L-arabinose   isomerase    from   Alicyclobacillus acidocaldarius : role of Lys-269 in pH optimum［J］．Appl Environ Microbiol，2005，71 :7888-7896.

［9 ］  Lee  D W，Choe  E A，Kim S B，et al．Distinct metal dependence  for catalytic and structural functions in the L-arabinose isomerase  from  the   mesophilic  Bacillus  halodurans   and  the  thermophilic  Geobacillus stearothermophilus［J］．Arch Biochem Biophys，2005， 434 : 333-343.

［10］  Lee  D  W，Jang  H  J，Choe   E  A，et  al．Characterization  of  a thermostable  L-arabinose    ( D-galactose )   isomerase   from    the hyperthermophilic  eubacterium  Thermotoga  maritima［J］．Appl Environ Microbiol，2004，70 : 1397-1404.

［11］  Rhimi  M，Bajic   G，Ilhammami  R，et  al．The  acid-tolerant  L- arabinose isomerase from the mesophilic Shewanella sp.  ANA-3 is  highly active at low temperatures［J］．Microb Cell Fact，2011，10 :

96．doi : 10．1186 /1475-2859-10-96 .

［12］  Rhimi  M，Ilhammami  R，Bajic  G，et  al．The  acid  tolerant  L- arabinose isomerase from the food grade Lactobacillus sakei 23K is  an attractive D-tagatose  producer［J］．Bioresour  Technol，2010， 101 :9171-9177.

［13］  Prabhu  P，Tiwari M K，Jeya M，et al．Cloning and characterization of a novel L-arabinose isomerase from Bacillus licheniformis［J］.  Appl Microbiol Biotechnol，2008，81 :283-290.

［14］  Xu Z，Qing  Y J，Li  S，et  al．A novel L-arabinose isomerase from Lactobacillus fermentum  CGMCC2921  for D-tagatose  production : gene cloning，purification and characterization［J］．J Mol Catal B : Enzy，2011，70 : 1-7.

［15］  Cheng  L，Mu  W，Zhang  T，et  al．An  L-arabinose  isomerase  from  Acidothermus   cellulolyticus    ATCC43068 :  cloning ，expression， purification，and characterization［J］．Appl Microbiol Biotechnol， 2010，86 : 1089-1097.

［16］  Rhimi    M ，Bejar   S． Cloning ，purification    and   biochemical  characterization of metallic-ions independent and thermoactive L- arabinose  isomerase  from  the  Bacillus  stearothermophilus  US100  strain［J］．Biochim Biophys Acta，2006，1760 : 191-199.

［17］  Li  Y J，Zhu Y，Liu A，et al．Identification and characterization of a  novel   L-arabinose   isomerase    from   Anoxybacillus   flavithermus  useful    in    D-tagatose    production［J］． Extremophiles ，2011， 15 :441-450.

［18］  Hong  Y  H，Lee  D  W，Pyun  Y  R，et  al．Creation  of  metal- independent     hyperthermophilic     L-arabinose      isomerase     by  homologous recombination［J］．J  Agric  Food   Chem，2011，59 :  12939-12947.

[19] Wang Furong, Xu Hong, Li Sha, et al. Screening and identification of D-tagatose-producing strains [J]. Food and Fermentation Industry, 2009, 35 (8): 15-19.

[20] Xu Hong, Wang Furong, Li Sha, et al. A method for preparing D-tagatose using a fermentable lactobacillus and the lactobacillus: China, 200910025982. 9 [P]. 2010-12-29.

[21] Xu Hong, Xu Zheng, Zhu Hongyang, et al. A high-temperature-resistant L-arabinose isomerase and its application: China, 201010153253. 4 [P]. 2011-08-31

［22］Manjasetty  B A，Chance M R．Crystal structure of Escherichia coli L-arabinose isomerase ( ECAI) ，the  putative  target  of  biological tagatose production［J］．J Mol Biol，2006，360 :297-309.

［23］  Rhimi  M，Aghajari  N，Juy  M，et  al．Rational  design of  Bacillus stearothermophilus    US100     L-arabinose      isomerase :  potential applications for D-tagatose production［J］．Biochimie，2009，91 : 650-653.

［24］  Rhimi  M，Juy M，Aghajari N，et al．Probing the essential catalytic residues  and   substrate   affinity    in   the   thermoactive  Bacillus stearothermophilus  US100  L-arabinose  isomerase  by  site-directed mutagenesis［J］．J Bacteriol，2007，189 : 3556-3563.

［25］  Prabhu P，Jeya  M，Lee  J  K．Probing the molecular  determination for the catalytic efficiency of L-arabinose isomerase from Bacillus licheniformis［J］．Appl Environ Microbiol，2010，76 : 1653-1660.

［26］  Kim  H J，Kim J  H，Oh  H J，et  al．Characterization  of a  mutated  Geobacillus     stearothermophilus     L-arabinose     isomerase      that  increases the production rate of D-tagatose［J］．J Appl Microbiol， 2006，101 :213-221.

［27］  Oh  D K，Oh H J，Kim H J，et al．Modification of optimal pH in L- arabinose  isomerase  from   Geobacillus   sterothermophilus  for  D- galactose  isomerization［J］．J   Mol   Cata   B : Enzym，2006，43 :  108-112.

［28］  Oh  H J，Kim H J，Oh D K．Increase in D-tagatose production rate by   site-directed   mutagenesis    of   L-arabinose    isomerase    from Geobacillus  thermodenitrificans［J］．Biotehnol   Lett，2006，28 : 145-149.

[29] Jiang Bo, Mu Wanmeng, Cheng Lifang. Mutant enzyme L20A with high yield of D-tagatose and its mutant method: China, 201010112337. 3[P]. 2012-05-23.

[30] Ryu  S  A，Kim  C  S，Kim  H  J，et   al．Continuous  D-tasgatose production by immobilized thermostable L-arabinose isomerase in a   packed-bed    bioreactor［J］． Biotechnol    Prog ，2003 ， 19 : 1643-1647.

［31］  Kim  H J，Ryu S A，Kim P，et al．A feasible enzymatic process for  D-tagatose production by an immobilized thermostable L-arabinose  isomerase in a packed-bed bioreactor［J］．Biotechnol Prog，2003， 19 :400-404.

［32］  Jorgensen  F，Hansen  O  C，Stougaard  P．Enzymatic  conversion  of  D-galactose to D-tagatose : heterologous  expression  and characterisation of  a  thermostable  L-arabinose  isomerase  from  Thermoanaerobacter mathranii［J］．Appl  Microbiol  Biotechnol， 2004，64 : 816-822.

［33］  Lim  B C，Kim H J，Oh D K．Tagatose production with pH control in  a   stirred   tank   reactor   containing   immobilized   L-arabinose isomerase   from    Thermotoga    neapolitana［J］． Appl    Biochem Biotechnol，2008，149 :245-253.

［34］  Jung E S，Kim H J，Oh D K．Tagatose production  by immobilized recombinant    Escherichia     coli    cells     containing     Geobacillus stearothermophilus L-arabinose isomerase mutant in a packed-bed bioreactor［J］．Biotechnol Prog，2005，21 : 1335-1340.

［35］  Hong Y  H，Lee  D  W，Lee  S J，et  al．Production  of D-tagatose  at high   temperatures    using   immobilized    Escherichia    coli    cells expressing  L-arabinose  isomerase  from Thermotoga  neapolitana ［J］．Biotechnol Lett，2007，29 :569-574.

[36] Fu Fenggen, Xu Zheng, Li Guixiang, et al. Production of D-tagatose using immobilized recombinant Escherichia coli cells [J]. Chinese Journal of Bioengineering, 2011, 31 (7): 85-90.

[37] Xu Z，Li  S，Fu  F  G，et  al．Production  of D-tagatose，a  functional sweetener，utilizing  alginate  immobilized Lactobacillus fermentum CGMCC2921  cells［J］．Appl   Biochem   Biotechnol，2012，166 (4) : 961-973.

［38］  Liang  M，Chen M，Liu X，et al．Bioconversion of D-galactose to D- tagatose : continuous  packed  bed  reaction  with  an  immobilized  thermostable L-arabinose  isomerase  and  efficient  purification by  selective  microbial  degradation［J］．Appl   Microbid  Biotechnol， 2012，93 : 1469-1474.

［39］  Jebors S，Tauran Y，Aghajari N，et al．Supramolecular stabilization of acid  tolerant  L-arabinose   isomerase  from Lactobacillus sakei ［J］．Chem Commun，2011，47 : 12307-12309.

［40］  Lim  B C，Kim H J，Oh D K．High production of D-tagatose by the addition of boric acid［J］．Biotechnol Prog，2007，23 : 824-828.

［41］  Rhimi  M，Messaoud  E  B，Borgi  M  A，et  al．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［J］．Enzyme  Microb Technol，2007，40 :  1531-1537.

[42] Xu Guiqiang, Feng Biao, Jiang Bo, et al. Separation of tagatose by phenylboronic acid/quaternary ammonium ionic liquid solvent extraction [J]. Food Industry Science and Technology, 2010(8): 294-297.

[43] Huang Wenxia, Mu Wanmeng, Jiang Bo. Separation and purification of D-tagatose [J]. Food and Fermentation Industry, 2008, 34(6): 168-171.

［44］  Ibrahim  O O，Spradlin J E．Process for manufacturing D-tagatose : US，6057135［P］．2000-05-02.

［45］  Kim S  B，Lee  Y M，Park S  W，et al．Food grade thermophilic  arabinose    isomerase    expressed     from    GRAS ， and    tagatose  manufacturing method by using it : US，20080124771［P］．2008- 05-29.

［46］  Rhimi  M，Chouayekh  H，Gouillouard  I，et al．Production of  D- tagatose，a low caloric sweetener during milk fermentation using L- arabinose    isomerase［J］． Bioresour     Technol ， 2011 ， 102 :  3309-3315.