Bioconversion of elephant grass (Pennisetum purpureum) acid hydrolysate to bacterial cellulose by Gluconacetobacter xylinus

Authors

  • X.-Y. Yang,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
    • These authors have the same contribution to this work.
  • C. Huang,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author
    • These authors have the same contribution to this work.
  • H.-J. Guo,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author
    • These authors have the same contribution to this work.
  • L. Xiong,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author
  • Y.-Y. Li,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author
  • H.-R. Zhang,

    1. Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    2. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author
  • X.-D. Chen

    Corresponding author
    1. Key Laboratory of Renewable Energy, Chinese Academy of Sciences, Guangzhou, China
    • Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou, China
    Search for more papers by this author

Correspondence

Xin-De Chen, No. 2 Nengyuan Road, Tianhe District, Guangzhou 510640, China. E-mail: cxd_cxd@hotmail.com

Abstract

Aims

To evaluate the possibility of elephant grass acid hydrolysate converting into bacterial cellulose (BC) produced by Gluconacetobacter xylinus CH001 and to characterize the morphology and structure of the cellulose produced.

Methods and Results

Acid-hydrolysed and detoxified elephant grass acid hydrolysate was inoculated with G. xylinus CH001. After 14 days of static fermentation, about 6·4 g l−1 of BC could be generated. Meanwhile, 60·4% (w/w) of BC yield on sugar consumption was obtained. Scanning electron micrographs illustrated that the network of cellulose fibres became denser, and the diameter changed with the growth. FT-IR spectra showed almost same results for all the BC samples collected on different culture time. X-ray diffractograms demonstrated that the crystalline form of BC was cellulose I, the crystallinity increased to 53·58%, and the crystallinity index reached up to 99%.

Conclusions

Elephant grass acid hydrolysate could be utilized efficiently for BC production by G. xylinus CH001. Structure analysis on the cellulose produced showed its potential of being excellent material for further application.

Significance and Impact of the Study

Our studies for the first time examined the bioconversion of low-cost elephant grass into high-value BC and the changes in its morphology and structure following the culture time.

Introduction

Bacterial cellulose (BC), synthesized by various bacterium, has great potential in the application on many fields such as food, biomedical material, papermaking and transducer diaphragms (Iguchi et al. 1988; Budhiono et al. 1999; Czaja et al. 2006). It was also reported the possible application of BC as a scaffold material (Dugan et al. 2013). Many factors would influence the BC productivity, and fermentation substrate cost is a significant one. To reduce its production cost, different substrates such as molasses (Bae and Shoda 2004), konjac powder hydrolysate (Hong and Qiu 2008), various fruit juices (Kurosumi et al. 2009) were utilized for BC production. However, it is worth noting that the sources and cost of these substrates are still problems restricting the scale up of BC production.

Compared with these substrates, lignocellulosic biomass has many advantages such as greatest availability, low cost and renewable characteristics, and thus, it is an optimal substrate for biorefinery (Rubin 2008). Based on this, it might be a potential substrate for BC production. In addition, this biochemical process could fulfil the conversion from low value-added plant cellulose to high value-added microbial cellulose (BC). However, to date, merely a few lignocellulosic biomasses such as wheat straw and cotton-based waste textiles were used for BC production (Hong et al. 2011, 2012).

It is worth noting that the nitrogen sources present in the common lignocellulosic hydrolysates were extremely low (Chen et al. 2012b; Huang et al. 2012a). This seems good for some bioconversion such as lipid production, but might be not suitable for BC production because nitrogen sources could stimulate the BC accumulation during fermentation.

Thus, screening optimal lignocellulosic biomass for BC production is critical and necessary. Elephant grass (Pennisetum purpureum) is a species from the Poaceae native to the tropical grasslands of Africa and now introduced into most tropical and subtropical countries (del Rio et al. 2012). The species is a robust grass with perennial stems, reaching over 3 m high, and is widely recognized as having the highest biomass productivity among herbaceous plants, attaining up to 45 Mg ha−1 year−1, and thus has been considered an excellent alternative feedstock to provide abundant and sustainable resources of lignocellulosic biomass for the biofuels production (Somerville et al. 2010). Different from common lignocellulosic biomass such as rice straw, wheat straw, corncob and bagasse, the protein amount of elephant grass is much bigger (Kabi et al. 2005). Therefore, more nitrogen sources would be generated during the hydrolysis of elephant grass, and thus, it could be a better substrate for BC production.

To date, conversion of plant cellulose to microbial cellulose by chemical hydrolysis and later biological fermentation is one relatively new topic in the field of biomacromolecules, and the process, especially the evolution of BC accumulation and its structure, was still not clear on different lignocellulosic hydrolysates. In this work, for the first time, BC fermentation was carried out on the elephant grass acid hydrolysate. And the process of BC accumulation including the evolution of its yield and structure was measured and analysed.

Methods and materials

Elephant grass and its composition analysis

Elephant grass was obtained from Guangdong Province (China). Benzene–alcohol extract, nitric acid–alcohol cellulose and lignin content were determined according to the method described by Liu (2004). In detail, benzene–alcohol extract was measured after extraction of biomass with benzene and ethanol at 90°C for 6 h, and the content of benzene–alcohol extract was the original biomass dry weight minus the biomass dry weight after extraction. Nitric acid–alcohol cellulose was measured after extraction of biomass with nitric acid and ethanol, and the content of nitric acid–alcohol cellulose was the original biomass dry weight minus the biomass dry weight after extraction. Lignin content was measured by using modified Kalson methods (Klason 1910; Hussain et al. 2002) using 80% sulfuric acid as extraction solvent. Hemicellulose content was detected by 2 mol l−1 hydrochloric acid method (Boyin et al. 1981). In detail, the biomass was firstly treated with 2 mol l−1 HCl at 100°C and then neutralized by NaOH, filtrated and washed. Finally, the hemicellulose content was obtained by calculation of sugar concentration in filtrate and washing solution using DNS method.

Preparation of elephant grass acid hydrolysate

Elephant grass was hydrolysed by ZHONGKE New Energy Co., Ltd (Ying-Kou, China). According to ZHONGKE New Energy Co., Ltd, the hydrolysis of elephant grass was carried out based on Chen's patent (Chen et al. 2012a), and the hydrolysis of elephant grass was carried out in an 8-m3 bioreactor with sulfuric acid (2·5% w/v) at 135°C for 1 h. The resulted hydrolysate was detoxified by overliming and absorption according to the previous work (Huang et al. 2012b).

Micro-organism, medium and fermentation condition

Gluconacetobacter xylinus CH001 (Laboratory of Energy and Biochemical Engineering, Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences) was used as the micro-organism for BC production in this work.

The preculture was performed on precultivation medium (g l−1, mannitol 25, peptone 5, yeast extract 3, initial pH 6·6) at 28°C and 150 rpm for 48 h. Then, 8% (v/v) seed culture was translated into the elephant grass acid hydrolysate, and the static fermentation was carried out at 28°C for 14 days.

Harvest of BC and determination of its dry weight

After fermentation, the fermentation broth and BC were separated by vacuum filtration. Then, the BC was treated with 1·5% (w/v) NaOH at 80°C for 2 h. After that, the BC was repeatedly washed with distilled water. Finally, the BC was dried to constant weight at 105°C and weighed.

Measurement of sugars metabolism of Gluconacetobacter xylinus

Two flasks containing fermentation broth (50 ml) were withdrawn periodically for parallel measurements. After filtration, the filtrate was used for analysing the residual sugars and nitrogen. Sugar concentrations (d-glucose, d-xylose and l-arabinose) in elephant grass hydrolysate were analysed by HPLC (Waters 2685 systems; Waters Corp., Milford, CT, USA), with a RI detector (Waters 2414) and on Aminex HPX-87H column (300 mm × 7·8 mm; Bio-Rad Corp., Philadelphia, PA, USA) using 5 mmol l−1 H2SO4 solution at a flow rate of 0·55 ml min−1 at 50°C.

Analysis of the physical and chemical structure of BC

Scanning electron microscopy

The morphologies of vacuum-dried BC samples were observed by a Hitachi S-4800 high-resolution field emission scanning electron microscope (FE-SEM; Hitachi, Tokyo, Japan) operated at 2·0 kV and 10 μA.

FT-IR spectroscopy

BC samples were mixed with spectroscopic-grade potassium bromide powder (1% w/w), and Fourier transform infrared spectroscopy (FTIR) was carried out in the range of 400–4000 cm−1 wavelength with a Perkin-Elmer Spectrum one FTIR Spectrometer (US). The kind of crystallite allomorph was determined with shifting FTIR bands related to specified groups and bonds (Oh et al. 2005; Phisalaphong and Jatupaiboon 2008; Shen et al. 2009).

X-ray measurement

Structure, size and percentage of crystals of BC samples were analysed by D/max-RA X-ray diffractometer (Rigaku, Tokyo, Japan) with Cu Kα radiation (λ = 0·154 nm) operated at 40 kV and 100 mA. Samples were scanned from 10 to 50° (2θ range) at a scan speed of 0·008°/step. To determine the sample crystallinity, profile analysis was carried out with a peak fitting program using Gaussian line shapes. The crystal size of cellulose was calculated using Debye–Scherrer's equation (Shezad et al. 2010). Crystallinity index (CI) was calculated from the related intensity data by the Segal method [eqn (1)] (Cheng et al. 2009).

display math(1)

where I200 is the maximum intensity of crystalline region at 2θ (about 22·8°) and Iam is the intensity of the amorphous region of the wide-angle X-ray diffraction curves at 2θ (about 18°).

Results

Elephant grass acid hydrolysate preparation

The aim of this work is to fulfil the bioconversion from the low value-added plant cellulose to high value-added microbial cellulose (BC). To achieve this, hydrolysis of elephant grass into fermentable sugars is necessary.

The composition of elephant grass was analysed before the hydrolysis process. Generally, the composition of elephant grass mainly contained 43·03% cellulose, 18·44% hemicellulose, 22·63% lignin, 10·56% benzene–alcohol extractive and 5·34% ash. Overall, the relatively high ratio of cellulose and hemicellulose makes this raw material to be a potential substrate for BC production.

Considering the hydrolysis technology for lignocellulosic biomass, enzymatic hydrolysis seems more attractive than acid hydrolysis in the future (Rubin 2008; Huang et al. 2012a); however, the enzyme used for the hydrolysis is still high for the industrial-scale production. In contrast, acid hydrolysis requires merely low concentration of low-cost inorganic acids (usually sulfuric acid; Huang et al. 2009; Chen et al. 2012b). Thus, it seems more potential for industrialization at today's situation. However, it is worth noting that xylose is the main component in the lignocellulosic acid hydrolysate; thus, the capability of using xylose as carbon source for fermentation is critical for the bioconversion of lignocellulosic acid hydrolysate (Huang et al. 2009; Chen et al. 2012b). This would be considered at the following parts.

The hydrolysis of elephant grass was fulfilled by ZHONGKE New Energy Co., Ltd in their 8-m3 reactor. By their hydrolysis methods, common lignocellulosic biomass such as corncob could be hydrolysed efficiently for later fermentation (Chen et al. 2012b). It is worth noting that detoxification is necessary before fermentation on the lignocellulosic acid hydrolysate (Hong et al. 2011; Chen et al. 2012b; Huang et al. 2012b). And in this work, two classical, low-cost detoxification methods namely overliming and absorption (Chen et al. 2012b; Huang et al. 2012b) with activated carbon were applied. After the treatment by both methods, the resulted elephant grass acid hydrolysate could be utilized as the substrate for BC production without adding any other compounds. The sugar composition (g l−1, glucose 12·0 ± 0·6, xylose 20·3 ± 0·8, arabinose, 2·3 ± 0·1) of resulted elephant grass acid hydrolysate was analysed by HPLC. As the HPLC analysis depicted, xylose was the main sugar present in the elephant grass acid hydrolysate, and glucose and arabinose were also found in this substrate. To make the sugar concentration suitable for BC fermentation, the elephant grass acid hydrolysate was diluted about three times, and finally, the total sugar concentration of elephant grass acid hydrolysate used for fermentation was 11·8 g l−1.

BC fermentation of Gluconacetobacter xylinus

The BC fermentation was carried out on the diluted elephant grass acid hydrolysate mentioned above. Then, the evolution of BC dry weight was measured, and the results are shown in Fig. 1. In the beginning 3 days of fermentation, no obvious BC pellicle (a film covered the surface of the medium) could be observed in the flask. The initial pH value in this work is 6·0, and during the first 3 days of fermentation, the pH value was stable at round 6·0 (Fig. 1). The BC pellicle begun to be seen from the 4th day to 5th day, and BC dry mass was about 2·0 g l−1 after 5 days of fermentation. Interestingly, the pH value did not decrease when the BC accumulation begun (Fig. 1), which was in contrast with most BC production whose process accompanies with decreasing pH value. Interestingly, from the 5th day to the 10th day, the BC dry mass increased slowly, and after that, the BC concentration increased faster, and finally, a BC yield of 6·4 g l−1 could be obtained after 14 days of static fermentation (Fig. 1). At the same time, it is found that the pH value of fermentation broth continued to increase throughout the fermentation process. Finally, the pH value increased to about 7·0 (Fig. 1).

Figure 1.

Evaluation of [bacterial cellulose (BC)] dry weight, residual sugars and pH value of broth throughout the fermentation process. (▽) BC dry weight; (○) residual sugars and (□) pH value.

Sugar metabolism of Gluconacetobacter xylinus

In this work, the evolution of different sugars during fermentation was measured and is shown in Fig. 2. As it depicted, during the first 3 days of fermentation, the sugar utilization including glucose, xylose and arabinose was slow, indicating that an obvious lag phase existed during the beginning of fermentation. After that, the sugars were consumed much more quickly. Interestingly, the glucose and xylose were utilized simultaneously by G. xylinus. Glucose was exhausted after 8 days of fermentation. The rate of sugar consumption became slower after 5 days of fermentation. Correspondingly, the increasing rate of BC yield was also slower after 5 days of fermentation (Fig. 1). It is worth noting that the utilization of arabinose by G. xylinus was much slower than the other two sugars throughout the fermentation process (Fig. 2). Finally, the xylose remaining in the fermentation broth was merely about 1·1 g l−1. Overall, the BC yield on sugars consumed was about 60·4%. This value was excellent for the bioconversion process on lignocellulosic hydrolysates.

Figure 2.

Utilization of different sugars by Gluconacetobacter xylinus during [bacterial cellulose (BC)] fermentation. (□) glucose; (○) xylose and (△) arabinose.

Change in BC morphology and structure during fermentation

FE-SEM

The morphologies of BC samples collected on different static fermentation time are showed in Fig. 3. On the 4th day of fermentation, the BC showed as a dense pellicle with little visible pores. With the prolongation of fermentation time, a large number of glucose molecules linked together with -1, 4-glycosidic bond into polydextrose, which then secreted from the eyelet of cell wall. On the 8th day, some cellulose microfibrils were formed from the polydextrose and parallelly extended forward as the increase in secretion. Then, microfibrils were connected by hydrogen bonding to form microfibrils bundle with a diameter of 20–100 nm on the 10th day. These bundles winded each other into microfibrils ribbon with a diameter of 15–100 nm on the 12th day. At last, these microfibrils ribbons interweaved each other into a porous network structure, exhibiting a layer of transparent gel film to the naked eye.

Figure 3.

FE-SEM pictures of [bacterial cellulose (BC)] samples collected on different static fermentation time.

FT-IR

Full-range FT-IR spectra of BC samples collected on different static fermentation time were carried out to detect the kind of crystallite allomorph with shifting peaks related to specified groups and bonds, which are showed in Fig. S1. The characteristic band appeared at 3392–3421 cm−1 for the stretching vibration of hydroxyl groups (-OH), at 2922–2926 cm−1 for the asymmetric stretching vibration of C-H, at 1456 cm−1 for the asymmetric deformation vibration of C-H, at 1047 cm−1 for C-O-C and C-O-H stretching vibration of sugar ring and at 880 cm−1 for γ (COC) in plane, symmetric stretching (Colom and Carrillo 2002). As shown in Fig. S1, these BC samples collected on different fermentation time showed similar structure except the tiny shift in the range of 2800–3600 cm−1. With the prolongation of fermentation time, the characteristic bands of BC became much wider and stronger. This might be due to the enhanced intermolecular hydrogen bonding effect resulted from the simultaneous utilization of glucose and xylose by G. xylinus (Fig. 2), which could bring about the change in O-H stretching vibration band (Sturcova et al. 2004).

X-ray diffraction

X-ray diffraction (XRD) patterns obtained from BC samples are shown in Fig. S2. They showed diffraction profile characteristics of cellulose I, with peaks at 2θ angles of 14·7, 17·2, 22·8 and 34·3° corresponding to the (1Ī0), (110), (200) and (004) crystallographic planes (Table S1), respectively. The diffraction peaks in the range of 25–34 and 35–50° were attributed to the inorganic salt introduced during the process of BC production, and they had been deducted to avoid the effect on the calculation of crystallinity. The crystallinity of BC samples increased from 23·15 to 53·58% after 2 weeks of static fermentation (Table S2). Furthermore, the crystallite size of (200) plane decreased from 102 to 87 Å (Table S2), and the CI of BC samples reached up to 99%. The peak of (200) lattice plane showed the highest intensity in the diffraction patterns of our produced BC, indicating that the hydrogen bond role during crystallization promoted the preferred orientation growth to (200) crystal plane and leaded to higher crystallinity and smaller crystallite size. According to the XRD analysis (Fig. S1), the effective utilization of glucose and xylose by G. xylinus enhanced the intermolecular hydrogen bonds and promoted the aggregation and crystallization of subelementary fibrils, while the processes of orientation and crystallization in cellulose depended largely on the intermolecular hydrogen bonds formed by the hydroxyl groups. Namely, the major influential factors affecting the crystallinity of BC could be ascribed to intermolecular hydrogen bonds.

Discussion

Usually, static BC fermentation is a long-time process, and lag phase is existed in such bioconversion. In the first 3 days of fermentation, no obvious BC pellicle was observed, indicating that lag phase was still existing even if elephant grass acid hydrolysate was the substrate for BC production. The phenomenon during the 4th day to 5th day showed that elephant grass acid hydrolysate had some buffer capacity or the environment of elephant grass acid hydrolysate could stimulate G. xylinus to secrete some alkalic materials when accumulated BC. Also, it was possible that G. xylinus could utilize the organic acids such as acetic acid in the elephant grass acid hydrolysate, and thus, the pH value of medium increased. Undoubtedly, both could overcome the fatal disadvantage of BC fermentation that more and more acidic environment would completely inhibit the BC accumulation process. The results after the 5 days of fermentation showed that it was possible that the growth of G. xylinus and its BC accumulation was in a stepwise manner and that its growth would be prior to the BC accumulation. Thus, the BC yield would increase faster after 10 days of fermentation.

The BC yield on this work was further compared with previous work with other substrates (Table 1). Corn steep liquor was used as substrate for BC production early (Noro et al. 2004). Later, pineapple juice was proved to be a promising substrate for BC production (Kongruang 2008). More recently, fruit juice was used as substrate for BC production, and the highest BC yield of this work was close to 6 g l−1 on the orange-based medium (Kurosumi et al. 2009). However, it is worth noting that using expensive fruit juice might unfortunately increase the cost of BC production. Compared with these raw materials, lignocellulosic biomass has the greatest availability in nature. Moreover, much of lignocellulosic biomass is the residues from agriculture and industry such as rice straw, bagasse, corncob. Thus, the cost of lignocellulosic biomass was low for BC production. As shown in Table 1, the BC yield on elephant grass acid hydrolysate was close or better to that on other substrates. Thus, using elephant grass acid hydrolysate has great potential for BC production in future.

Table 1. Comparison of bacterial cellulose (BC) production on different substrates
StrainCarbon sourceBC yield (g l−1)References
Gluconacetobacter xylinum BPR2001Corn steep liquor8·0Noro et al. (2004)
G. xylinum TISTR 893Pineapple juice576 (wet weight)Kongruang (2008)
G. xylinum ATCC 23770Konjac powder hydrolysate2·1Hong and Qiu (2008)
G. xylinum IFO 15606Xylose0·1Ishihara et al. (2002)
G. xylinum BPR2001Molasses5·3Bae and Shoda (2004)
G. xylinum NBRC 13693Fruit juice6·0Kurosumi et al. (2009)
G. xylinum CH001Elephant grass acid hydrolysate6·4This work

As mentioned above, xylose was the main sugar present in elephant grass acid hydrolysate, and thus, the possibility of this bioconversion depended on the capacity to utilize xylose for BC production. In fact, to evaluate the possibility of using lignocellulosic hydrolysates as substrate for BC production, xylose had been used as carbon source for BC production already (Ishihara et al. 2002). However, it was shown that xylose might be not an optimal carbon source for BC production when compared with glucose in that work. Although adding D-xylulose into medium could stimulate the xylose utilization, but this effect was not obvious (Ishihara et al. 2002). However, the final result in this work was completely contrast to previous work that xylose was not suitable for BC production (Ishihara et al. 2002). Obviously, the environment of elephant grass acid hydrolysate could stimulate the capacity of xylose utilization of G. xylinus and finally made this bioconversion feasible. In contrast, another pentose, namely arabinose, could hardly be utilized by G. xylinus for BC production. However, arabinose could be recovered from water after fermentation and make profits for its relatively high market price.

According to the FE-SEM (Fig. 3), FT-IR (Fig. S1) and XRD (Fig. S2) results of the BC produced from elephant grass acid hydrolysate, it could be concluded that the microstructure of BC, namely crystallinity and average degree of polymerization at the molecular level had taken place, changes corresponding (Table S2 and Fig. S2) with the increase in fermentation time, thus leading to desired structural properties for practical application.

Acknowledgements

The authors acknowledge the financial support of Natural Science Foundation of Guangdong Province (S2012040007546), the Support Plan Project of National Science and Technology (2012BAD32B07), National Natural Science Foundation of China (U1261116) and Foundation of Director of Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences (y107rf1001).

Ancillary