Introduction

Sorangium cellulosum was named as such because of its ability to grow with only cellulose as the carbon source; this organism is a species of the myxobacterium [16]. Next to Actinomyces and Bacillus, S. cellulosum is the microorganism that can produce the most number of bioactive molecules [4].

S. cellulosum has become a medicinal microorganism because of its secondary metabolites with indefinite quantity and miscellaneous species [6]. Epothilone is a secondary metabolite produced by S. cellulosum; this substance is a cytotoxic macrolide that demonstrates antineoplastic activity and resembles taxol [8]. Currently, more than six epothilones or their derivatives, such as patupilone (Epothilone B, EPO906), ixabepilone (BMS-247550), sagopilone (ZK-EPO), BMS-310705, KOS 862 (Epothilone D), KOS 1584, and KOS 193, have been evaluated in clinical trials; moreover, these drugs have already been launched in the market [22, 23]. However, the production and development of epothilone drugs are seriously limited by various difficulties, including a relatively long doubling time, limited number of molecular tools, and cells secreting various exopolysaccharides, which lead to the formation of bacterial clumps and uneven cell growth under liquid-state fermentation condition [13]. Our previous work has shown that S. cellulosum can grow well on the surface of the agar plate and stably produce epothilone when cultured in the solid medium with filter paper as substrate (unpublished data). In the present study, we added immobilization materials (i.e., porous ceramics) into the epothilone fermentation broth, which allowed S. cellulosum attachment onto the surface of the solid materials, as well as steady growth on these solid materials. Subsequently, these processes can finally lead to the production of epothilone under submerged fermentation condition, which is similar to the possible event on the agar plate. In this regard, we may address the problem of unstable epothilone production, which is caused by agglomerative growth, and simultaneously improve the fermentation yield of epothilone.

Immobilization technology has been widely used in the fields of enzymatic reaction and biofermentation because of its advantages of a convenient separation process and excellent biocompatibility [17]. At present, several studies have discussed information regarding the increase in antibiotic production by using the technology of immobilized cells, such as thienamycin [5], penicillin [19], and tylosin [26]. Porous ceramic, which exhibits a controllable porous structure, is the most commonly used material in the field of biofermentation [10]. The combination between micropores and macropores is favorable to the adsorption and immobilization of cells, as well as the transmission of organic matter [12]. Furthermore, the diatomite-based porous ceramics, which consist of SiO2 with micropores, show a more extensive surface and excellent biocompatibility. In the current study, we developed a diatomite-based porous ceramic and applied this innovation in liquid fermentation. Furthermore, we found an apparent improvement in epothilone production.

Materials and Methods

Microorganism and Medium

The Sorangium cellulosum SoF5-09 strain, which had been isolated from soil and genetically modified by genome shuffling, was preserved in our laboratory [11]. The strain was refrigerated for convenience [15]. The SoF5-09 cells were routinely inoculated on M26 agar [20] and subsequently cultured in liquid M26 at 30℃ while shaking at 200 rpm.

The culture medium (pH 7.2) for epothilone production contained 0.3% dextrin, 0.07% sucrose, 0.02% glucose, 0.17% beancake powder, 0.17% MgSO4·7H2O, 0.3% CaCl2, 1.0 ml/l EDTA-Fe 3+(v/v), 2% Amberlite XAD-16 resin (Rohm and Haas), and 1.0 ml/l trace element solution (i.e., per liter of the trace element solution containing 100 mg MnCl2·4H2O, 20 mg CoCl2, 10 mg CuSO4, 10 mg Na2MoO4·2H2O, 20 mg ZnCl2, 5 mg LiCl, 5 mg SnCl2·2H2O, 10 mg H3BO3, 20 mg KBr, 20 mg KI, and 1,000 ml distilled water) [21]. In this initial fermentation system without porous ceramics, the maximum epothilone yield of S. cellulosum was 23.3 mg/l.

Preparation and Characterization of Diatomite-Based Porous Ceramics

The powder diatomite (Haiyun Fitter Company, Henan, China), kilned diatomite (pre-sintered at 1,000℃), and paraffin were mixed at a mass ratio of 35:45:2. The mixture, which was used as raw material for the diatomite-based porous ceramics, was then ball-milled for 30 min. The varisized saw dust (20 to 80 orders), which was used as a pore-forming agent, was added into the powder mixture with the mass ratio of 1:5. The homogeneous mixture was then compressed into tablets (Φ10 mm × 2 mm) at different pressures (5 to 20 MPa) and subsequently calcined in the resistance furnace for 3 h at 1,000℃.

The microstructure of the specimens was examined under a Hitachi S-4800 scanning electron microscope (SEM) [1]. The pore size distribution and the specific surface area of the specimens were measured by a Micromeritics IV 9500 automatic mercury porosimeter. The mechanical strength of the specimens was measured using a Baoda universal testing machine (1036 PC). The open porosity of the specimens was determined by the Archimedes method with water as the liquid medium [27].

Adsorption Experiment of Diatomite-Based Porous Ceramics

We immersed the diatomite-based porous ceramics that were prepared previously into the bacterial suspension of S. cellulosum at a ratio of 5 g:100 ml. Subsequently, the mixtures were placed into the shaker (30℃, 100 rpm) to conduct the adsorption experiment. After 5 h, we took a sample to determine the adsorbing capacity of diatomite-based porous ceramics through the dry weight method [2].

Modification of Diatomite-Based Porous Ceramics

We attempted to modify the diatomite-based porous ceramics using HCl and FeCl3 as modifiers.

In the first method, steeping the porous ceramics into the HCl solution (0.2 mol/l) for 24 h was a vital step. Afterward, pouring out the acid and washing the porous ceramics with distilled water until the pH became neutral were necessary.

For the second method, the porous ceramics and FeCl3 liquid were mixed at a ratio of 1:4. Subsequently, the mixture was dried in a vacuum oven and stirred once every 30 min at 110℃. The porous ceramics were then placed into a muffle burner for 3 h of calcination [18]. Finally, the naturally cooling porous ceramics were washed several times with sufficient distilled water and dried again in a vacuum oven at 110℃ for 24 h [18].

Immobilized Fermentation of Sorangium cellulosum

We cultured the activated S. cellulosum in a shaking incubator (30℃, 200 rpm) for 3 days after inoculating these organisms into a 300 ml flask, which contained 50 ml of M26 liquid medium. The seed liquids were ready to be inoculated into another 300 ml beaker flask with fermentation medium, as well as the porous ceramics contained in the solution. Once the fermentation started during the spinning process, the addition of extra porous ceramics was unnecessary. The technological conditions (e.g., solid/liquid ratio, temperature, rotary shaker speed, inoculation amount, initial pH, fluid volume, and fermentation time) for immobilized fermentation were optimized by orthogonal experiment based on epothilone yield. Subsequently, we made the metabolic curve of immobilized S. cellulosum under optimal fermentation conditions.

Determination of Metabolic Curve

During the fermentation period, we obtained samples from the culture solution at 1-day intervals and determined the following parameters: contents of reducing sugar and total sugar, cell dry weight (i.e., the sum of free cells and immobilized cells), and epothilone yield [3]. These factors were measured by the DNS method, lost heavy method, and HPLC method, respectively [3].

Results

Preparation of Diatomite-Based Porous Ceramics

In addition to pressure during the forming process, the absorptive capacity of porous ceramics was also affected by the contents and sizes of the pore former sawdust (Fig. 1). Fig. 1 shows a high level of absorptive capacity in the following conditions: 20 orders of sawdust size, 2.5% mass fraction of its content, and 5 Mp forming pressure. The orthogonal optimization experiment based on these factors was scheduled according to table L9 (34). Finally, the results showed that the diatomite-based porous ceramics demonstrated good performance (i.e., 5 µm pore diameter, 23.55 ㎡/g specific surface, 32% porosity, 10.2 MPa mechanical strength, and 28.4mg/g adsorbance) under 30 orders of sawdust size, 2.5% mass fraction of its content, and 7 Mp forming pressure.

Fig. 1.Effects of sawdust size and forming pressure on adsorption quantity.

Fig. 2B presents the SEM photographs of the diatomitebased porous ceramics without carrying the immobilized S. cellulosum. The average pore size of diatomite-based porous ceramics was approximately 5 µm, whereas the length of S. cellulosum was approximately 3 µm to 5 µm (Fig. 2A). As the immobilization carrier for S. cellulosum, the porous ceramics satisfied the demands of size in that over 70% of the aperture should be greater than the minimum cell and should be smaller than five times the size of the maximum cell [7]. Figs. 2C and 2D show the porous ceramics, which have adsorbed S. cellulosum. The porous ceramics could adsorb the S. cellulosum efficiently and thus can be utilized as a growth and immobilized carrier.

Fig. 2.SEM photographs. (A) SEM photograph of Sorangium cellulosum. (B) SEM photograph of diatomite-based porous ceramics before adsorption. (C) SEM photograph of diatomite-based porous ceramic-adsorbed cell. (D) SEM photograph of diatomite-based porous ceramic-adsorbed cell.

Performance Analyses of Modified Porous Ceramics

Different modification conditions can improve the surface adsorption property of porous ceramics at different degrees. As shown in Fig. 3A, the adsorption quantity of modified porous ceramics was increased significantly with the modifier; in addition, the modifier FeCl3 (34.7 mg/g) was superior to HCl, based on its ability of improving epothilone scale. By contrast, different FeCl3 concentrations showed different effects on the modification results (Fig. 3B). The appropriate concentration for the use of the modifier FeCl3 was 1.5 mol/l. The porous ceramics modified in this condition could increase epothilone yield by 3.8 times within 7 days compared with the fermentation system without porous ceramics. By adding the modified porous ceramics into the fermentation liquid, the final epothilone content in the liquid reached 83.6 mg/l.

Fig. 3.Effects of different factors on adsorption quantity. (A) Effect of different treatments on adsorption quantity. (B) Effect of different modifier concentrations on adsorption quantity.

The XRD patterns of porous ceramics were examined with (and without) FeCl3 modification (Fig. 4). Moreover, the XRD patterns indicated that if the porous ceramics were modified by 1.5 mol/l FeCl3, the silica (SiO2) and Fe2O3 will be the main phases. The results indicated that Fe2O3 paint coat was on the porous ceramic surface. Fe2O3 paint coat can raise the isoelectric point of porous ceramics, thereby enhancing the adsorbability of porous ceramics to microorganisms.

Fig. 4.XRD patterns of porous ceramics. (A) Porous ceramics modified by FeCl3. (B) Porous ceramics not modified by FeCl3.

Fermentation Experiments of Immobilized Sorangium cellulosum

The modified diatomite-based porous ceramics were added to the S. cellulosum liquid fermentation system, and inspections of the immobilized fermentation conditions (i.e., solid-to-liquid, temperature, speed shaking, inoculation dosage, liquid volume, and fermentation time) were performed using the single-factor experiment (Fig. 5). The optimization experiments were designed through an orthogonal method [L9 (44)] based on significant factors (i.e., temperature, shaking speed, inoculation dosage, and liquid volume). The results showed that the optimal conditions of immobilized fermentation involved 3:50 as the solid-toliquid ratio, 30℃ fermentation temperature, 220 rpm shaking speed, 10% inoculation dosage, 45 ml of liquid volume in a 300 ml flask, and 8 days of fermentation. Finally, the epothilone production yield was 90.2 mg/l.

Fig. 5.Effects of different factors on epothilone production. (A) Effect of solid-to-liquid ratio on epothilone production. (B) Effect of temperature on epothilone production. (C) Effect of shaking speed on epothilone production. (D) Effect of inoculation dosage on epothilone production. (E) Effect of loading volume on epothilone production. (F) Effect of fermentation time on epothilone production.

The metabolizing curve of immobilized S. cellulosum was determined under optimum fermentation condition. The curves showed that cells will accumulate at the stage of bacterial growth; moreover, the sugar content declined during S. cellulosum catabolism and epothilone production (Fig. 6A). The epothilone production speed increased apparently from the fifth day, and the dry cell weight reached the maximum value on the sixth day (Fig. 6B). At day 7, the total glucose content was almost utilized, and the speed of epothilone production increased relatively slowly. At the end of fermentation, the dry cell weight declined sharply with bacterial autolysis because of the exhaustion of nutrition.

Fig. 6.Fermentation curves of immobilization Sorangium cellulosum. (A) Glucose metabolism curve of immobilization Sorangium cellulosum fermentation. (B) Epothilone production and dry cell weight curve during different fermentation stages.

Discussion

At present, several studies have reported on epothilone production by different strains, as well as the methods of improving fermentation yield. Gerth et al. [9] obtained 22 mg/l epothilone A and 11 mg/l epothilone B from the fermentation broth of S. cellulosum strain Soce90. Lau et al. [14] increased epothilone D production from 0.16 mg/l to 23 mg/l by optimizing the fermentation conditions for the genetically modified strain Myxococcus xanthus K111-40-1. Li et al. [15] improved the epothilone B production of S. cellulosum strain So 0157-2 from 0.8 mg/l to 104mg/l by genome shuffling. Sim et al. [24] showed that by in situ removal of ammonium using cation exchange resin, epothilone production of S. cellulosum Soce90 can be increased by 2.6 times; they also found that the production of epothilone B that was obtained from S. cellulosum Soce90 can be increased to 5.03 mg/l/day by alginate-immobilized cells, which was three times higher than that of free cells [26].

We added the modified diatomite-based porous ceramics into the submerged culture system of S. cellulosum to improve the production of epothilone. We also optimized the fermentation conditions for the immobilized cells. Finally, epothilone production was improved from 23.3 mg/l to 90.2 mg/l. To the best of our knowledge, S. cellulosum will accumulate and form bacterial clumps in the liquid medium; these processes hindered the internal cells from obtaining nutrients (unpublished data). Therefore, cell growth was inhibited under liquid-state fermentation conditions, and epothilone production was less than that of the solid-state cultivation method (unpublished data). After adding the porous ceramics into the fermentation broth, S. cellulosum can grow along the hole surface of porous ceramics, and the bacterial clumps will no longer form (Fig. 2D). This result demonstrated that the abnormal growth behavior of S. cellulosum can be removed by cellimmobilized material porous ceramics, thereby addressing the problem of unstable epothilone production.

Furthermore, the optimal condition for preparing porous ceramics involved the following: 2.5% (mass fraction) as contents of 30 orders of sawdust (pore-making agent) and 7 MPa as modeling pressure. The efficiency of porous ceramics was confirmed by the electron micrographs, and the results showed that cells dispersed over the porous ceramics. The optimal fermentation conditions for immobilized S. cellulosum were 3 g:50 ml solid-to-liquid ratio, 45 ml of liquid in a 300 ml flask, 10% inoculation quantity of seed liquids, 30℃ fermentation temperature, 220 rpm shaking speed, and 8 days of fermentation time. Under the optimal conditions described above, the final epothilone production was increased by 3.87 times compared with free-cell fermentation.