Development and validation of a medium for recombinant endo-β-1,4-xylanase production by Kluyveromyces lactis using a statistical experimental design

Abstract

Kluyveromyces lactis is an excellent host for a high cell density culture, which allows high expression levels of recombinant enzymes. Nutrient composition and culture conditions affect the secretion, production level and stability of the recombinant host. Therefore, it is technologically important to formulate a medium that stimulates high cell density and enhances the desired enzyme production using K. lactis GG799. In this study, six media were initially compared, and a Plackett-Burman experimental design was employed to screen for important components and trace elements. Nitrogen sources such as ammonium sulfate and free amino acid (casamino acid) as well as compounds like MgSO₄∙7H₂O, Na₂SO₄, ZnSO₄∙6H₂O, MnSO₄∙4H₂O and KH₂PO₄ affected biomass concentrations (5.67 g/l) and recombinant endo-β-1,4-xylanase (Xyn2) production (49.73 U/ml). Optimum productivity was obtained at shorter incubation times (i.e., 6 h), making the medium suitable for use when seeking efficient production. Expression of recombinant Xyn2 by K. lactis GG799 in the designed medium resulted in satisfactory recombinant Xyn2 volumetric productivity (v_p) at 8.29 U/ml/h. When compared to the rich, non-selective YPD medium, the designed medium improved biomass output and recombinant Xyn2 production in K. lactis GG799 by approximately 9 and 22%, respectively.

Introduction

Xylanases (EC 3.2.1.8) play important roles in industrial processes, where they are used as additives to improve the quality of baked goods (Collins et al. 2006; Basinskiene et al. 2008) and animal feeds and to bleach kraft pulp. Xylanases also act synergistically with other hemicellulases to produce commercial xylooligosaccharides (Jiang et al. 2004). Most commercial xylanase preparations originate from microorganisms belonging to the genus Trichoderma and Aspergillus (Ahmed et al. 2009). However, extensive purification to isolate the pure enzyme from fungi is necessary, especially in large-scale production (Bauer et al. 2006). Therefore, cloning genes into a recombinant system is a prerequisite to obtain reproducible amounts of desired enzyme with specific properties (Bleve et al. 2008). One of the most promising hosts for gene cloning is Kluyveromyces lactis (Van Ooyen et al. 2006; Wamalwa et al. 2007). Kluyveromyces lactis is a unicellular organism that is phylogenetically close to Saccharomyces cerevisiae, as shown in comparison studies (Pain et al. 2004) subsequent to the whole genome sequencing of K. lactis (Sherman et al. 2004). The sequencing effort led to important breakthroughs in utilizing K. lactis specifically in biotechnological and industrial processes, as well as in regulatory network evolutions and their functional interactions (Bussereau et al. 2006; Rodicio et al. 2006). In addition, most K. lactis strains require a methanol-free media for growth. This requirement represents a major production advantage because explosion-proof fermentation equipment and high cost carbon sources are not needed, unlike in the large-scale growth of some methylotrophic yeasts (Mayer et al. 1999; Van Ooyen et al. 2006).

Despite the importance of the genetic background, the composition of cultivation media is crucial to the successful development of strains for large-scale industrial production of heterologous proteins (Hahn-Hagerdal et al. 2005). A cultivation medium is designed to reflect the biosynthetic capacity of microorganisms, which consequently affects strain analyses and performance in industrial applications. Chemically defined media are usually preferred in laboratory research because they permit one to determine specific requirements for growth and product formation (Hensing et al. 1995).

The classical procedure for optimizing fermentation media is by conducting one-dimensional searches with successive variable variations [or the one factor at a time (OFAT) method]. However, it is practically impossible for OFAT methods to achieve an optimal medium in a finite number of experiments (Ibrahim and Elkhidir 2011). Therefore, statistical designs are used to reduce experiment numbers despite large variable numbers. Statistical experimental designs are also important in distinguishing variables with significant effects on the desired response. For example, key variables can be revealed by employing fractional factorial designs, such as the Plackett-Burman method (Plackett and Burman 1946) and two-level factorial approach (Lundstedt et al. 1998). The Plackett-Burman statistical design is preferable to classical methods because it is rapid and reliable, highlights important nutrients and substantially reduces the number of experiments required, saving time, chemicals and manpower (Srinivas et al. 1994; Murthy et al. 2000).

Several papers have been published on the development and optimization of media for the production of recombinant proteins in hosts such as Pichia pastoris (Chang et al. 2006), S. cerevisiae (Zhang et al. 2007; Larsson et al. 2001) and Escherichia coli strains (Beshay et al. 2003). However, optimal medium conditions have not been developed and published for K. lactis. Therefore, this study was undertaken to develop a medium for the high expression level of a Trichoderma reesei Xyn2 in K. lactis GG799 strain (a wild-type industrial isolate from DSM Food Specialties, Delft, Netherlands) using a statistical experimental approach. The results presented below should provide a firm basis for developing an industrial process of recombinant enzyme production in K. lactis GG799.

Materials and methods

Recombinant strain and construction of expression vector

Kluyveromyces lactis GG799 strain (see K. lactis Protein Expression Kit Instruction Manual; New England Biolabs, USA) was used as a host for the expression of recombinant Xyn2 from T. reesei ATCC 58350. Prior to cloning the gene, reverse transcription polymerase chain reaction (RT-PCR) amplifications were carried out using primers designed from an mRNA sequence of T. reesei strain QM6a (NCBI accession number: U24191). Oligonucleotide primers were designed as follow: XynF 5′ AATGTCGACCAGACGATTCAGCCCGGCACGGGCTACAAC-3′ and XynR 5′-AATGCGGCCGCTTTAGCTTAGCTGACGGTGATGGAAGCAGA-3′ supplemented with SalI and NotI restriction sites, respectively. The endo-β-1,4-xylanase gene (xyn2) was cloned into vector pKLAC1 (9,091 bp) located downstream of the K. lactis α-mating factor (α_MF) signal sequence. The vector is based on the P_LAC-PBI variant that has mutations in the Pribnow box-like sequence (PBI) containing the strong LAC4 promoter (P_LAC4) for expression of endoxylanase. The vector also contained the Xyn2 and the ADH2 promoter (P_ADH2), which drives expression of an acetamidase selectable marker gene (amdS). The constructed pKLAC1xyn2 vector was linearized with SacII before being integrated into the K. lactis genome at the LAC4 promoter locus. Transformant selection was carried out by growing recombinant cultures harboring pKLAC1xyn2 on a nitrogen-free minimal medium with a yeast carbon base (YCB) containing 5 mM acetamide.

Growth medium and conditions

Recombinant K. lactis GG799 was routinely cultured in 100 ml of yeast extract peptone dextrose (YPD) medium containing 20 g/l of glucose as a carbon source (Table 1); the medium was buffered to pH 5.0 ± 0.2 using a 50 mM phosphate-citrate buffer. Media and glucose preparations were autoclaved separately at 121°C for 15 min. The same initial glucose concentration and pH conditions were used throughout the study. All chemicals were purchased from Sigma-Aldrich (USA) and Merck (Germany) unless otherwise stated. Cells were incubated at 30 ± 0.5°C with shaking at 250 rpm until growth reached the mid-exponential phase (i.e., A₆₆₀ of 17). Thereafter, cultures were centrifuged at 6,000 rpm for 10 min and washed twice with sterile distilled water. Cultures were resuspended and concentrated to 20–30 ml prior to use in inoculations.

Table 1 Medium composition for preliminary screening phase. All media were supplied with 20 g/l glucose as carbon source

Preliminary screening

A preliminary screening experiment was carried out to verify mineral components needed for growth and secretion of the recombinant Xyn2. A 10% K. lactis GG799 inoculum, having an A₆₆₀ of 5–6 (unless otherwise mentioned), was cultivated under sterilized conditions in 50 ml of media as shown in Table 1. Cultures were incubated at 30 ± 0.5°C with shaking at 250 rpm. Variable components in the Plackett-Burman design were chosen based on results of biomass and xylanase activity. Software by Design Expert (Statistics Made Easy, v. 6.0.4, Stat-Ease, Minneapolis, MN, USA, 2001) was used to generate the experimental design.

Screening of nutrient components using the Plackett-Burman method

Eleven independent variables in twelve combinations were organized according to the Plackett-Burman design matrix (Table 2). Variables and concentrations were obtained from the preliminary experiment performed. Inoculum with a final A₆₆₀ of 10 was seeded in 250 ml flasks containing 50 ml of media and incubated at 30°C ± 0.5°C on a rotary shaker at 250 rpm. Samples of 3 ml were collected and centrifuged at 6,000 rpm for 10 min, and the resulting supernatant was used for the xylanase assay.

Table 2 Plackett-Burman experimental design combinations of factors and levels for recombinant Xyn2 and biomass production using K. lactis GG799

Experimental design and statistical analysis

A Plackett-Burman experimental design was used to determine the most important media components for recombinant Xyn2 expression by K. lactis GG799. Design Expert software was used to fit the first-order model of a linear approach with $Y\left( {x,\beta } \right) = {\beta_{\text{o}}} + \sum\limits_{{i = 1}}^n {{\beta_i}{x_i}}$ where Y(x,β) is the estimated response (i.e., xylanase produced), β_o and β_i are constant regression coefficients of the model, and x_i represents independent variables (i.e., media components). Results of a regression analysis (ANOVA) were used to determine significance levels and screen out non-significant media components.

Determination of biomass concentration

The biomass concentration was determined by turbidity measurements at an absorbance of 660 nm using a UV/visible spectrophotometer (Ultrospec 1100 Pro; Amersham Bioscience, Sweden). Measured values were correlated to dry weight from duplicate samples (2 ml taken every 2 h during cultivation) that were centrifuged at 6,000 rpm for 10 min, washed twice with distilled water and dried for at least 24 h at 80°C. The equation to calculate dry mass was as follows:

$$ {\text{dry mass}}\left( {{\text{g}}/1} \right) = 0.{379} \times {{\text{A}}_{{{66}0}}} $$

Xylanase assay

Xylanase activity was measured using a dinitrosalicylic acid method (Miller 1959) with 0.01 g/l of beechwood xylan (Sigma) as the substrate at 50°C. Appropriate dilutions in 50 mM sodium citrate buffer (pH 5.0 ± 0.2) were used as the enzyme source. The amount of enzyme was determined by measuring the release of reducing sugars. One unit of xylanase was defined as the amount of enzyme required to liberate one μmole of xylose from xylan per minute under assay conditions.

Validation of the Plackett-Burman experimental model

The mathematical model generated from Plackett-Burman implementation was validated by conducting an experiment with a predicted medium set for maximum responses. The experiment was performed in the same manner as the screening phase.

Results and discussion

Expression of recombinant Xyn2 in K. lactis GG799

A cDNA fragment encoding T. reesei Xyn2 and 191 amino acid residues was obtained by RT-PCR (data not shown). The amino acid sequence is identical to T. reesei strain QM6a xyn2 cDNA sequence (GenBank accession number U24191). The integrative plasmid pKLAC1xyn2 produced polypeptides consisting of an N-terminal signal peptide of the K. lactis α_MF and the family 11 xylanase. Transformation of K. lactis GG799 was achieved by a chemical method that used transformation reagents supplied with the NEB yeast transformation kit. Colonies of recombinant K. lactis GG799 cells with the pKLAC1xyn2 construct containing the amdS gene were isolated on yeast carbon base (YCB agar medium that lacked a nitrogen source but contained acetamide). Acetamidase expression in transformed cells allows for cell growth because acetamidase metabolizes acetamide to ammonia, which can be utilized by cells as a nitrogen source. The recombinant strain was then grown in YPD medium at 30°C in shaking flasks.

An advantage of K. lactis GG799 that makes this yeast especially attractive as a host system is that the recombinant plasmid has good stability in shake-flasks and pilot-scale cultures (Merico et al. 2004). As stated by Van Ooyen et al. (2006), a major benefit in using integrative vectors is the increased of genetic stability of the transformed strain compared to episomal vectors. Other than that, selection of transformant using acetamide influences the frequency of targeted tandem integration of a vector into the chromosome of K. lactis GG799 cells. According to Read et al. (2007), acetamide selection not only almost completely enriches populations for strains harboring multiple tandem vector integrations but also increases vector copy numbers compared to antibiotic selection. Transformation using expression vector pKLAC1 routinely yields strains bearing two to six integrated copies in more than 90% of the transformant on medium containing acetamide (Van Ooyen et al. 2006). In this study, whole-cell PCR strategies are used to detect targeted integration of pKLAC1xyn2 in K. lactis GG799 chromosome (as recommended by the manufacturer). The PCR results exhibited multicopy integration in K. lactis genome (data not shown).

Preliminary experiment

It is essential to have extensive knowledge when formulating a medium so as to include important minerals and growth factors. However, not all nutrients enhance cell growth and protein production when present above a certain concentration (Lee 1996). To determine which basic components are vital for cell growth and production, six different media with different components and concentrations were tested before implementing the Plackett-Burman design. The effects of the media towards recombinant Xyn2 production are shown in Fig. 1. The highest recombinant Xyn2 activity of 48.10 U/ml was obtained from YHP medium with a biomass concentration of 4.23 g/l. In contrast, recombinant Xyn2 activity attained in the YPD medium was the lowest at 40.85 U/ml, but this medium yielded the highest biomass concentration at 5.19 g/l. In yeast extract-containing mediums, such as YPD, sugars and lactate serve as auxiliary carbon sources and strongly affect yeast cultivation rates (Hahn-Hagerdal et al. 2005). Variable selections for the Plackett-Burman design were based on these media that produced the highest levels of recombinant Xyn2, regardless of biomass concentrations. In addition, the free amino acid casamino acid (Amresco, Ohio, USA) was included in the process design because Jahic et al. (2006) reported that it had a significant positive effect on growth and protein production in P. pastoris. Further medium improvement was carried out by combining casamino acid with components of YHP and PCT media because these two media have produced high expression of recombinant Xyn2 in K. lactis GG799, as shown in Fig. 1 and described in the initial screening process.

Fig. 1

Effects of six different media on recombinant Xyn2 production by K. lactis GG799 and biomass concentrations. The inoculum was cultured in 50 ml of each media with an A₆₆₀ of 5 to 6 and maintained at 30°C with continuous shaking at 250 rpm (see “Materials and methods”). Each bar represents () recombinant xylanase activity produced and () biomass concentrations in the media as the mean ± standard error (n = 3)

Statistical analysis by ANOVA

By using a Plackett-Burman experimental design (Table 2), wide variations in recombinant Xyn2 production were observed, ranging from 19.49 ± 1.49 to 45.56 ± 0.26 U/ml. These wide variations in recombinant Xyn2 activity were coupled with narrow variations in biomass that ranged from 4.08 ± 0.18 to 5.85 ± 0.03 g/l (Table 2). The wide variation in recombinant Xyn2 production reflected the significance of factors on the enzyme and biomass production (Purama and Goyal 2008) and vice versa. Comparisons among these eleven trials showed that the medium in trial 4 produced the highest recombinant Xyn2 activity, while the medium in trial 10 produced the highest biomass. However, influences of these media compositions were further analyzed using ANOVA.

Regression analyses data for recombinant Xyn2 and biomass production are shown in Tables 3 and 4, respectively. The model for recombinant Xyn2 production in K. lactis GG799 has a coefficient of determination (R ²) = 0.9976, meaning that 99.76% of the data variability can be elucidated effectively.

Table 3 Analysis of ANOVA regression model for recombinant xylanase activity

Table 4 Analysis of ANOVA regression model for biomass production

Each component was screened at the 95% confidence level regardless of its effects, either positive or negative. Among the 11 variables evaluated, only 8 factors, (NH₄)₂SO₄, MgSO₄∙7H₂O, Na₂SO₄, casamino acid, CuSO₄∙5H₂O, CaCl₂∙2H₂O, K₂SO₄ and MnSO₄∙4H₂O, had significant model terms based on the confidence level (Table 3). The remaining components studied including FeSO₄∙7H₂O, ZnSO₄∙6H₂O and KH₂PO₄ were less than the 95% confidence level and therefore considered insignificant. In addition, apart from the percentage of the confidence level, main effects can be ranked based on the mean square. Table 3 demonstrates that casamino acid had the highest mean square value, supporting that it had a major impact on recombinant Xyn2 expression.

A positive coefficient intervals (CIs) (95% confidence level) show that the studied component will produce incremental responses if added and vice versa. If the influence of the component is equal to or greater than the 95% confidence level, and its coefficient values (CI low and CI high) are both negative, then the component is effective at enhancing recombinant Xyn2 production, but the concentration required is lower than the signified low (−) concentration (Gohel et al. 2006). However, if the CI low and CI high coefficient values are positive and negative, respectively, then the component has no effect on recombinant Xyn2 production. The medium components CuSO₄∙5H₂O, MnSO₄∙4H₂O and Na₂SO₄ displayed positive coefficient values, suggesting that the concentrations of these three components should be increased to improve recombinant Xyn2 production. Conversely, casamino acid, (NH₄)₂SO₄, MgSO₄∙7H₂O, CaCl₂∙2H₂O and K₂SO₄ exhibited negative coefficients and thus should be reduced to improve recombinant Xyn2 production. Ignoring insignificant factors, the linear regression for recombinant Xyn2 production can be written as:

$$ {\text{RecombinantXyn2activity}}\left( {{\text{U}}/{\text{ml}}} \right) = - {3}.{26}{X_1} - {4}.{31}{X_2} + {3}.{57}{X_3} - {4}.{46}{X_5} + {4}.{33}{X_6} + {2}.{18}{X_8} - {2}.{51}{X_9} - {2}.{81}{X_{{10}}} $$

where

$$ {X_1} = {\left( {{\text{N}}{{\text{H}}_{{4}}}} \right)_{{2}}}{\text{S}}{{\text{O}}_{{4}}};{X_2} = {\text{MgS}}{{\text{O}}_{{4}}} \bullet {7}{{\text{H}}_{{2}}}{\text{O}};{X_3} = {\text{N}}{{\text{a}}_{{2}}}{\text{S}}{{\text{O}}_{{4}}}{;}{X_5} { =}{\text{asamino acid}};{X_6} = {\text{CuS}}{{\text{O}}_{{4}}} \bullet {5}{{\text{H}}_{{2}}}{\text{O}};{X_8} = {\text{MnS}}{{\text{O}}_{{4}}} \bullet {4}{{\text{H}}_{{2}}}{\text{O}};{X_9} = {\text{CaC}}{{\text{l}}_{{2}}} \bullet {2}{{\text{H}}_{{2}}}{\text{O and}}{X_{{10}}} = {{\text{K}}_{{2}}}{\text{S}}{{\text{O}}_{{4}}}. $$

For biomass production of K. lactis GG799, the model presented an R ² = 0.9180, explaining 91.80% of the validity in the response. Similarly, components were screened at a 95% confidence level (Table 4). Among the 11 variables evaluated, only 4 (casamino acid, MgSO₄∙7H₂O, Na₂SO₄ and CaCl₂∙2H₂O) appeared to have significant model terms. The remaining components studied and not included in Table 4 [(NH₄)₂SO₄, FeSO₄∙7H₂O, K₂SO₄, CuSO₄∙5H₂O, MnSO₄∙4H₂O and KH₂PO₄] had no significant effect on biomass production. Casamino acid (mean square of 1.47) appeared to be the most important factor in biomass production based on the relatively high mean square estimate. Specifically, casamino acid allowed for complete use of the carbon source and, consequently, enhanced cell growth and heterologous protein production (Merico et al. 2004). In addition, cultivation medium is composed of casamino acid, supplementing free amino acid requirements for heterologous protein production in recombinant yeast and relieving the metabolic burden of cells (Hahn-Hagerdal et al. 2005).

Na₂SO₄ was the only significant variable that had a negative coefficient value, while the other three variables (casamino acid, MgSO₄∙7H₂O and CaCl₂∙2H₂O) showed positive coefficients. This outcome suggested that the Na₂SO₄ concentration should be decreased while concentrations of casamino acid, MgSO₄∙7H₂O and CaCl₂∙2H₂O should be increased to improve biomass production. The linear regression obtained for biomass production was:

$$ {\text{Biomass}}\left( {{\text{g}}/{1}} \right) = 0.{22}{X_2} - 0.{15}{X_3} + 0.{35}{X_5} + 0.{17}{X_9} $$

where

$$ {X_2} = {\text{MgS}}{{\text{O}}_{{4}}} \bullet {7}{{\text{H}}_{{2}}}{\text{O}};{X_3} = {\text{N}}{{\text{a}}_{{2}}}{\text{S}}{{\text{O}}_{{4}}};{X_5} = {\text{casamino acid}};{X_9} = {\text{CaC}}{{\text{l}}_{{2}}} \bullet {2}{{\text{H}}_{{2}}}{\text{O}} $$

Validation of screening results

An experiment was performed to validate the efficiency of the Plackett-Burman-designed medium. Recombinant Xyn2 activity and biomass production were set at the maximum response level with the intention that the suggested medium would maximize both factors in K. lactis GG799. Highest predicted values of recombinant Xyn2 activity and biomass production were 45.8 U/ml and 5.25 g/l, respectively, in a medium containing 5 g/l (NH₄)₂SO₄, 1.69 g/l MgSO₄∙7H₂O, 0.18 g/l Na₂SO₄, 3.93 g/l casamino acid, 0.24 g/l ZnSO₄∙6H₂O, 1.0 g/l MnSO₄∙4H₂O and 7.66 g/l KH₂PO₄. The recombinant Xyn2 activity attained was 49.73 U/ml, and biomass concentration was 5.67 g/l. Figure 2 represents the recombinant Xyn2 activity and biomass production profiles. Experimental and predicted values closely agreed with percent deviations of 7.9% for recombinant Xyn2 activity and 7.4% for biomass production. This outcome validated our findings for components screened under the Plackett-Burman design.

Fig. 2

Production of recombinant Xyn2 and cell growth of K. lactis GG799 profile. A validation experiment was performed to verify the efficiency of the optimal design medium in which both responses were set at their maximum level using statistical software. The method involves pulling together both first-order models and determining a set of conditions that is a compromise to the stated goals. Both () production of recombinant xylanase and () cell growth of K. lactis GG799 were obtained using a designed medium after screening by Plackett-Burman methods. Each line represents the mean ± standard error (n = 3). The recombinant xylanase production started in the early exponential growth phase and increased to its maximum activity and later decreased when glucose concentrations became a limiting growth factor

Expression of recombinant Xyn2 by K. lactis GG799 using the designed medium in a shake-flask culture provided satisfactory recombinant Xyn2 volumetric productivity (v_p). The maximum productivity for recombinant Xyn2 in K. lactis GG799 was 8.29 U/ml/h, which was eight-fold higher than productivity levels obtained by Wamalwa et al. (2007) from recombinant xylanase in K. lactis CBS 1065. The productivity level of recombinant Xyn2 from T. reesei produced by K. lactis was also seven-fold higher than levels obtained from S. cerevisiae (La Grange et al. 2001) and twice as high as production from the P. pastoris expression system (He et al. 2009). Moreover, the highest production of recombinant Xyn2 was reached within a relatively short time. Studies of xylanase production with K. lactis systems required more than 3 days using a complex medium (Faraco et al. 2008). Percent improvements in biomass concentration and recombinant Xyn2 production were approximately 9 and 22%, respectively, with use of the designed medium versus the YPD medium. Because the designed medium is superior to the YPD medium, we investigated the effects of different glucose concentration using the designed medium. Our findings revealed that higher glucose concentrations (5–50 g/l) resulted in increased cell biomass. However, at elevated glucose concentrations, the time required to reach maximum recombinant Xyn2 production was longer than at the original glucose concentration of 20 g/l glucose (data not shown).

An experiment was performed to understand relationships over the incubation time between glucose levels, recombinant xylanase expression and biomass production. Samples of cultured K. lactis GG799 were taken at 2, 4 and 8 h of fermentation during the validation phase. Supernatant from the centrifugation process was used to measure recombinant Xyn2 activity and glucose concentrations. The influence over time of glucose on recombinant Xyn2 production and biomass concentrations is shown in Fig. 3. Results show that Xyn2 activity decreased along with glucose concentration while biomass concentrations increased slightly. These data indicated that reductions of recombinant Xyn2 production in K. lactis GG799 may be linked to glucose depletions in the medium. As mentioned in “Materials and methods”, the constructed vector pKLAC1xyn2 was integrated at the LAC4 locus of the genome. Glucose depletions can cause low Xyn2 expression levels because the LAC4 expression system is mediated by glucose. Hsieh and Da Silva (2000) observed that glucose is an important inducer of recombinant enzyme production and also serves to provide carbon, energy and cell structure and cell maintenance. Thus, glucose depletions would reduce the synthesis of recombinant Xyn2 by K. lactis GG799. Conversely, feeding glucose into the designed medium of bioreactors may further increase recombinant Xyn2 production levels using K. lactis GG799.

Fig. 3

The influence over time of glucose on recombinant Xyn2 production and biomass concentrations. An experiment was conducted to understand the relationship over time between glucose levels, recombinant xylanase expression and biomass production. The effects of () glucose concentration on () recombinant xylanase and () biomass production were investigated. Samples of cultured K. lactis GG799 were taken after fermenting for at 2, 4 and 8 h of fermentation. Each bar and line represents the mean ± standard error (n = 3)

The reduction in recombinant Xyn2 activity could also be due to proteolysis by proteases because K. lactis secretes proteases in the presence of glucose, galactose and glycerol (Madinger et al. 2009). Proteolytic degradation has been an ongoing concern for the production of recombinant proteins in other hosts (Zhou and Zhang 2002), such as P. pastoris (Sinha et al. 2005) and S. cerevisiae (Rao et al. 1999). On the other hand, a question arises on the stability of plasmid implicated in the production of Xyn2. In the present study, relatively, the integrated plasmid should be stable as showed by other K. lactis system published before. This is because chromosomal integration is a more stable option when compared to episomal maintenance of foreign DNA (Romanos et al. 1992). However, further studies are needed to confirm the integrated plasmid stability in expression of Xyn2 in K. lactis GG799.

Conclusion

Improved recombinant Xyn2 production in K. lactis was achieved using a statistical experimental design. These findings demonstrate that medium components play a significant role in the growth of K. lactis GG799 and recombinant Xyn2 production. Results of this study will facilitate the large-scale, economically viable production of xylanase using K. lactis. Moreover, enhanced recombinant Xyn2 productivity supports the use of the designed medium, especially because this enzyme is widely used and commercially important. Bioreactor studies are needed to further improve the production of recombinant Xyn2 in K. lactis GG799.