Improved Production of Recombinant Human β-NGF in Escherichia coli – a Bioreactor Scale Study

NGF (Nerve Growth Factor) consists of three subunits, α, β and γ; among them β subunit is responsible for NGF biological activity. The β subunit is a dimer of two similar monomers containing cysteine knot motif, which is formed by three disulfide bonds (Wiesmann and De Vos 2001). β-NGF plays a fundamental role in development and survival of nervous system; so, it can be used as a therapeutic agent for the treatment of neurodegenerative diseases (Heese et al. 2006). NGF can be extracted from its natural source, the male mice submaxillar glands, by different methods of chromatography but it is unsuitable for clinical uses because contains heterogeneous mixtures of partially degraded dimmers (Bocchini and Angeletti 1969). Nowadays many studies have attempted to produce it as recombinant protein using different eukaryotic and prokaryotic hosts (Kurokawa et al. 2001; Choi and Lee 2004; Fan and Lou 2010).

Production of recombinant proteins in prokaryotic hosts is cost effective; among them E. coli is widely used expression host both in laboratory and industrial scale (Rosano and Ceccarelli 2014). Today many strategies have been introduced to increase the expression level in E. coli. In general, expression can be optimized by varying cultivation media, the proper selection of E. coli strain, induction conditions such as time and temperature of induction and the type of promoter and inducer concentration (Schumann and Ferreira 2004; Saez and Vincentelli 2014; Tegel et al. 2011). However, protein expression and production in industrial scale using a bioreactor requires special consideration. Previous studies have reported the significant effects of oxygen level, aeration and agitation rate on the growth rate in liquid medium of bioreactor (Zhong 2010; Wang et al. 2016). Agitation or shaking of culture is necessary to maintain a uniform environment without formation of bacterial aggregates throughout the bioreactor contents. Adequate agitation and aeration are also important factors in the success of protein expression (Thiry and Cingolani 2002; Lee et al. 2013).

The aim of this work was to optimize the fermentation parameters to obtain the highest β-NGF production in the bioreactor scale. Therefore, conditions of the fermentation process in the 5-l batch bioreactor were optimized to maximize the yield of the active protein. Response Surface Methodology (RSM) as the most common statistical method for optimization of various biochemical processes (Elibol and Ozer 2002; Wejse et al. 2003; Rui et al. 2009) was used to optimize the induction parameters and dissolved oxygen level (or different agitation rate) for β-NGF over production.

Effect of post induction time on β-NGF production. It has been widely reported that induction parameters can affect protein expression level in shake-flask culture (Azaman et al. 2010; Larentis et al. 2011; Savari et al. 2015; Gholami Tilko et al. 2016). In the present study for the first time, the effect of post-induction time on β-NGF production level was investigated in a 5-l bioreactor. Batch cultivation was performed using SOB medium with constant temperature (37°C), pH (7.0), agitation speed (300 rpm) and aeration speed (1vvm) throughout the experiment. As illustrated in Fig. 1, the maximum yield of β-NGF production (% recombinant protein expression level) was obtained at 2 h post induction time and recombinant protein expression level was reduced over extended periods of time. Therefore, post induction time of 2 h was selected and used for further experiments.

Fig. 1.

Optimization of post induction temperature and % dissolved oxygen (DO) using RSM. Post induction temperature is one of most important parameters affecting the recombinant protein expression level. Also, DO is a critical factor in the over expression of recombinant proteins, especially in the bioreactor scale. Therefore, to investigate the influence of DO and post induction temperature on β-NGF production, 10 experiments were designed using RSM method. These experiments were conducted by controlling DO (10, 20, 30, 40 and 50%) and post induction temperature (20, 24.25, 28.5, 32.75 and 37°C) (Tables I and II). Experimental error was evaluated at the central point conditions of the experimental plan (28.5°C and 30% DO). The constant cultivation temperature of 37°C was used before induction of promoter in each experiment. In all experiments, the promoter of pET39b::βNGF plasmid was induced by addition of 0.3 mM IPTG, when OD_{600 nm} reached the value of 0.6. After 2 h, the total proteins produced in recombinant bacteria were extracted. Simultaneously, cell growth (measured by absorbance at 600 nm) was monitored in each case. As the main purpose of this study was production of β-NGF, we used the western blot analysis in which an anti-his tag monoclonal antibody recognized the recombinant β-NGF with his-tag tail in the mixture of total proteins extracted from E. coli (Fig. 2A). To evaluate the effects of two variables (temperature and %DO) on the β-NGF expression level, the areas of each band detected on the nitrocellulose membrane were calculated by densitometry analysis (using Image J software) (Table II). Different combinations of post induction temperature (20–37°C) and dissolved oxygen (10–50%) yielded in the β-NGF production level from 3.8% ± 0.4 to 15.3% ± 1.2 as illustrated in Table II. The cell growth, which ranged from 1.02 ± 0.11 to 5.60 ± 0.29 (resulting in dry cell weight of 0.367 and 2.016 g/l) is also presented in Table II.

Fig. 2.

As illustrated in Table II and Fig. 2, high β-NGF expression levels were observed in Runs 2 and 9 (15.3% ± 1.2 and 14.6% ± 1.2, respectively). Run 10 was a repetition of Run 9 and resulted in the same value. In these runs, the cell growth (or dry cell weight) was not at maximum level. In other words, the increased cell growth was not necessarily accompanied by an increase in recombinant production level. It is necessary to mention that the highest value of β-NGF production was obtained at post induction temperature of 32.75°C and % DO of 20.

In contrast, the lowest value of β-NGF production level (3.8% ± 0.4) was obtained at post induction temperature of 28.5°C and 50% DO (Run 8). When comparing the runs 8 and 9 with constant post induction temperature (28.5°C) and different %DO (also runs 2 and 4 with constant post induction temperature of 32.75°C and different %DO) it could be observed that the recombinant β-NGF production was significantly reduced at high degree of DO. It can be a result of high agitation speed because higher speed of stirring could cause a breakdown of the cells (Banerjee et al. 2009).

Analysis of variance. The recombinant β-NGF production was the most important response measured in this study. Interaction of two parameters examined in this study (post induction temperature and %DO) had pronounced effect on the β-NGF production indicating the importance of these factors for enhancement of the recombinant protein production. The response under different combinations was analyzed by ANOVA technique (Table III). The model F-value of 120.36 from ANOVA of the recombinant protein expression level indicated that this model was significant. The p values were used as a tool to test the significance of each coefficient. The smaller the magnitude of p value, the corresponding coefficient was more significant. Also, a p value less than 0.05 indicated that the model terms were significant. The p values of terms presented in Table III indicated that both terms (A and B) had a significant influence on β-NGF production. In addition, the model is significant because p value is less than 0.05 (probe > F is 0.0002).

The other response was cell growth measured by absorbance at 600 nm and the effects of two parameters (post induction temperature and %DO) on this parameter were also examined. The model F-value of 6.82 from ANOVA of cell growth indicated that this model was also significant. The ANOVA indicated that this model and term of A (post induction temperature) were significant (p value < 0.05) but %DO did not significantly influence cell growth.

Model diagnostics. The normal (%) probability plot of the studentized residuals is a major tool to detect that errors are normally distributed and independent of each other (Luzier 1992). It can be seen from Fig. 3 that the errors of the models for the recombinant β-NGF production were normally distributed and insignificant.

Fig. 3.

Optimization of the process. The response surface plots and their corresponding contour plots described by the second-order polynomial equation were illustrated in Fig. 4. Each contour curve represents an infinite number of combinations of two tested variables. Fig. 4A and 4B indicate that the interaction of post induction temperature and %DO was significant. The β-NGF production increased with both post induction temperature and %DO till certain point and then decreased. The best post induction temperature and %DO for the maximum β-NGF production were 32°C and 20%, respectively. Fig. 4C and 4D show that post induction temperature had positive effect on cell growth but the effects of % DO on cell growth was not significant.

Fig. 4.

Large scale purification of β-NGF and bioactivity test. It must be noted that the β-NGF produced by pET39b::βNGF vector has his-tag tail; therefore it was purified using an affinity chromatography column (Ni⁺²-NTA agarose). To evaluate whether the purified β-NGF is functional or not, differentiation of PC12 cell line in the presence of the purified β-NGF was studied. Our data showed that the purified β-NGF was capable of differentiate PC12 cell line to the nerve cells similarly to the standard β-NGF (data not shown) (Hajihassan et al. 2017).

Nowadays, recombinant production of functional proteins is in high demand in modern biotechnology. E. coli is the most widely used expression host employed to produce many recombinant proteins. Optimization of culture conditions and induction parameters allows us to achieve the high quantity of the interest protein in this valuable host. Most of the studies in this area are in a shake-flask culture (Azaman et al. 2010; Larentis et al. 2011; Papaneophytou and Kontopidis 2012; Papaneophytou et al. 2013; Savari et al. 2015; Gholami Tilko et al. 2016). For example, in a survey recently conducted by our laboratory optimization of recombinant β-NGF production in shake-flask culture was done (Gholami Tilko et al. 2016). In the above study, the optimum post induction time and temperature were 2 h and 25°C, respectively. However, cell growth and protein production in the bioreactor scale are influenced by different parameters affecting each other (Zhong 2010). Since very few studies have been done to optimize protein production, especially recombinant proteins in industrial scale, our aim in this study was to find out the optimal conditions for high level production of recombinant β-NGF in E. coli grown in a 5-l bioreactor. Among the most important factors influencing the recombinant protein production level, the post induction time, post induction temperature and %DO play critical roles. Optimization of these parameters results in higher productivity and reduces cost of fermentation. It is of interest to note that the best induction time can be established for each recombinant protein by a trial and error. However, in many reports the expression is reduced over extended induction times, and this phenomenon is likely due to the recombinant protein degradation by bacterial proteinases (Joshi and Puri 2005; Choi et al. 2006; Papaneophytou et al. 2013; Gholami Tilko et al. 2016). In this study to produce high level of recombinant β-NGF in the bioreactor scale, the post induction times between 2 to 12 hours was examined. According to our results, the post induction time of 2 h is optimal for improved recombinant β-NGF production level.

Another factor that influences the expression level is post induction temperature. As the best cultivation temperature for E. coli is between 35–37°C, lowering the temperature causes a lower growth rate. On the other hand, in the case of expression of complex heterologous proteins, usually a decrease in the temperature during protein synthesis especially after IPTG induction and, consequently, slower growth rate of bacteria favor the production of the active protein form. It is also important to note that in the literature different post induction temperatures have been reported for overproduction of different heterologous proteins in the bioreactor scale or shake-flask cultures (Azaman et al. 2010; Papaneophytou and Kontopidis 2012; Morowvat et al. 2015; Savari et al. 2015; Su et al. 2015; Gholami Tilko et al. 2016). Therefore, the evaluation of the best post induction temperature for β-NGF production on bioreactor scale was one of the goals of this study. Also %DO is a critical factor in the over expression of recombinant proteins, especially in the bioreactor scale, because it can affect directly cell growth (Ram et al. 2015; Wang et al. 2016). Previous studies showed that bacterial growth and protein production were dependent on the degree of oxygenation; distribution of oxygen to the bacteria in the bioreactor was related to the intensity of aeration and stirring or agitation (Çalik et al. 2004; Kaya-Çeliker et al. 2009). On the other hand, at high degree of DO as a result of high agitation speed break down of the cells takes place (Banerjee et al. 2009). Recently, Zaslona et al. (2015) showed that production of recombinant 1,3-β-glucanase in E. coli increased when moderate stirring and oxygenation was applied. Therefore, to evaluate the effects of %DO and post induction temperature on β-NGF expression level and cell growth rate, we used response surface methodology (RSM) as a high effective method (Elibol and Ozer 2002; Wejse et al. 2003; Rui et al. 2009). Thus, various combinations of %DO (10–50) and post induction temperature (20–37°C) as two independent variables were examined (Tables I and II). Our results indicated that the highest β-NGF production in the bioreactor scale was obtained at post induction temperature and %DO of 32.75°C and 20%, respectively (Fig. 1). On the other hand, in this run the maximum cell growth or dry cell weigh were not obtained. It can be concluded that increasing cell growth does not necessarily increase recombinant protein expression level. The most important result obtained here was that the highest β-NGF production was achieved at moderate %DO (20%) and post induction temperature of 32°C. These results are in agreement with the previous data presented by others on the production of various proteins in E. coli in the bioreactor scale (Zaslona et al. 2015).

In summary, the response surface methodology was successfully used for the optimization of recombinant β-NGF production by varying the induction parameters and %DO. Our results could be beneficial for industrial production of β-NGF.