Natural variation in the glucose content of dilute sulfuric acid–pretreated rice straw liquid hydrolysates: implications for bioethanol production

Abstract

Rice straw is a promising resource for bioethanol production. Because the glucose content of pretreatment liquid hydrolysates is highly correlated with ethanol yield, the selection of appropriate rice cultivars is essential. The glucose content in liquid hydrolysates of pretreated rice straws of 208 diverse cultivars was evaluated in natural field in 2013 and 2014 using a novel high-throughput system. The glucose content of the rice straw samples varied across cultivars and was affected by environmental factors such as temperature and solar radiation. Several high-quality cultivars exhibiting high glucose content in both years were identified. The results of this study can aid in development of novel rice cultivars suitable as both feedstocks for bioethanol production and cooking.

Graphical abstract

The rice straw glucose content after pretreatment varied by cultivar. Thus selecting suitable cultivars is crucial for efficient ethanol production.

Key words:

• rice straw
• bioethanol
• diverse cultivars
• glucose content
• environmental factors

Modern society depends heavily on fossil fuels such as petroleum. This dependence may deplete fossil fuel resources and create environmental problems (e.g. climate change and acid rain). Increasing recycling is highly recommended as a means of lessening dependence on finite resources. The production of bioethanol, for example, enables the recycling of plant materials. Roy et al.¹⁾ and Koga and Tajima²⁾ claimed that increasing bioethanol production could reduce both fossil fuel consumption and the emission of greenhouse gases from soil.

First-generation bioethanol production processes relied on food materials such as maize (Zea mays) and the squeezed juice of sugarcane (Saccharum officinarum), but this resulted in increased food prices due to increased demand for harvests.³⁾ Unused biomass resources such as the straw of rice (Oryza sativa), maize stover, and sugarcane bagasse are promising alternatives because their use does not increase competition for food resources.⁴⁾ These agricultural residues are thus the primary resources for second-generation bioethanol production.

Rice straw, which is a lignocellulosic biomass, is composed primarily of glucan (starch and cellulose), xylan, and lignin⁵⁾ and is widely available in rice-cultivating regions. To produce bioethanol, the polysaccharides of rice straw must first be hydrolyzed into fermentable sugars.⁶⁾ However, lignin obstructs the enzymatic saccharification of the polysaccharides in rice straw.⁷⁾ Pretreatment is thus essential to break the lignin structure so that the polysaccharides are made available for the enzymes.⁸⁾

A number of methods for the pretreatment of rice straw have been proposed.⁹⁾ Dilute acid pretreatment is one of the most popular methods because it is inexpensive, convenient, and effective for many types of lignocellulosic biomass.^10,11) After pretreatment with dilute sulfuric acid, the resulting slurry is divided into a liquid hydrolysate and an acid-insoluble residue. The liquid hydrolysate contains glucose, xylose, and various byproducts.¹²⁾ A previous study reported that the glucose content in the liquid hydrolysate is highly correlated with the resulting ethanol yield,¹³⁾ suggesting that the glucose in the liquid hydrolysate is utilized during microbial fermentation. In addition, the rice straw liquid hydrolysate glucose content may vary widely (relative standard deviation: RSD = 75.9%).¹⁴⁾ Thus, this trait is an important factor to evaluate in the selection of rice cultivars most suitable for bioethanol production.

Very little information is available regarding varietal differences in the liquid hydrolysate glucose content. Two issues are particularly important in this regard. First, previous studies examined only a small number of cultivars. For example, Matsuda et al.¹³⁾ examined 8 cultivars, whereas Teramura et al.^12,14) examined 13 cultivars. More recently, Tanger et al.¹⁵⁾ evaluated 20 cultivars. Given that a large number of rice cultivars are available as potential bioenergy feedstocks, greater emphasis must be placed on identifying cultivars suitable for both food and bioenergy applications. Second, how environmental factors affect the liquid hydrolysate glucose content is largely unknown. A previous study reported that the liquid hydrolysate glucose content for a given cultivar may vary year to year,¹⁴⁾ but the precise relationship between glucose content and environmental factors remains unclear. Resolution of these issues will be necessary in order to more effectively utilize rice straw for biorefinery purposes.

In this study, 208 diverse rice cultivars, including those utilized by Matsuda et al.¹³⁾ Teramura et al.^12,14) and Tanger et al.¹⁵⁾ were grown under natural field conditions for 2 years. The glucose content in the liquid hydrolysate was evaluated using a high-throughput method in order to identify cultivars suitable for use as a feedstock for bioethanol production and to enhance understanding of the factors affecting hydrolysate glucose content.

Materials and methods

Plant materials and weather conditions

A total of 208 rice cultivars (Oryza sativa) were grown in an experimental field located at the Kobe University, Food Resources Education and Research Center (Kasai City, Hyogo Prefecture, Japan) in 2013 and 2014 (Supplementary Table 1). Cultivars of the Japanese rice collection¹⁶⁾ and World rice collection¹⁷⁾ were included in the study. Three plants per cultivar were harvested 45 days after the flowering date and then desiccated by exposure to the sun for 3 days. The harvested plants were stored indoors until the following process. After the panicle was separated, 1 or 2 main straws were selected and powdered using a Shake Master Auto grinding apparatus (BioMedical Science Co., Ltd., Tokyo, Japan) operated at 1,100 rpm for 30 min. The weight of rice straw harvested in 2014 was measured by CJ-820 electronic balance (Shinko Denshi Co., Ltd., Tokyo, Japan). The weather conditions at the experimental field in 2013 and 2014 were observed by Vantage Pro complete weather station system (Davis Instruments Corp., California, USA).

Table 1. The analysis of variance table for two-way analysis of variance.

	Sum square	Df	F value	P value
Subspecies	7.38	1	1.1001	0.2949
Year	3.32	1	0.4942	0.4825
Subspecies:year	1.35	1	0.2007	0.6544
Residual	2,590.44	386

Notes: Df: Degree of freedom.

Pretreatment

In this study, a dilute-acid pretreatment method¹⁴⁾ with modifications for high throughput was employed. Powdered rice straw samples were dried at 80 °C overnight, and then 150 mg was removed from each sample using an XS105DU electric balance (Mettler Toledo, Greifensee, Switzerland) and suspended in 2 ml of 1% (v/v) sulfuric acid. The samples were then heated at 180 °C with agitation at 100 rpm using a rotating oven (RDV-TM2, SAN-AI Kagaku Co., Ltd., Aichi, Japan) programed to reach 180 °C over 60 min and maintain that temperature for another 60 min. After pretreatment, the entire sample was collected and centrifuged at 12,000 g for about 5 s to separate the liquid hydrolysate, which was then used in subsequent experiments.

Measuring the liquid hydrolysate glucose content

A 100-μl aliquot of each liquid hydrolysate was diluted with 900 μl of distilled water to be in the range of the calibration curve of glucose detection reagent (Glucose CII-Test Wako; Wako Pure Chemical Industries Ltd., Osaka, Japan; 0–2.5 g/L). Next, 10 μl of each diluted sample and glucose calibration solution was placed into each of 3 wells of a 96-well microplate. Each sample was mixed with 200 μl of glucose detection reagent and allowed to react at 37 °C for 15 min in an HBO-350B oven (AGC Techno Glass Co., Ltd., Shizuoka, Japan). The absorbance at 505 and 600 nm was then measured using a Viento XS microplate spectrophotometer (DS Pharma Biomedical Co., Ltd., Osaka, Japan). Based on the calibration curve, the glucose content in the pre-dilution liquid hydrolysate of each cultivar was calculated, and the average concentration was determined based on samples from 3 plants from each cultivar.

Statistical analyses

R software, version 3.1.1¹⁸⁾ was used for statistical analyses, which included the Wilcoxon signed rank test, Spearman’s rank correlation coefficient analysis, and two-way analysis of variance.

Results and discussion

High-throughput system for measuring glucose content

A high-throughput system is necessary for the analysis of a large number of samples¹⁹⁾; therefore, we developed a high-throughput analytical method for determining the glucose content of the liquid hydrolysates from the 208 rice straw cultivars examined in the present study. The 2-step method involved: (1) pretreatment (2 ml of 1% [v/v] sulfuric acid solution) of a small amount of rice straw (150 mg), and (2) efficient quantitative analysis using a glucose detection kit and microplate spectrophotometer. As compared with the previous method described by Teramura et al.,¹⁴⁾ the present system can accommodate a 3–15-fold greater number of samples per day with respect to pretreatment, and measuring the glucose content, respectively. The high-throughput method we developed is applicable to other plant species as well.

Variation in liquid hydrolysate glucose content

The liquid hydrolysate glucose content was determined for rice straw samples from a total of 208 diverse rice cultivars. The frequency distributions showed similar trends in average glucose content in both 2013 and 2014 (Fig. 1, Supplementary Fig. 1). Our analyses indicate that the liquid hydrolysate glucose content varies among the rice cultivars examined in this study, in agreement with previous reports.^13,14) The average liquid hydrolysate glucose content was slightly higher in 2014 than in 2013, but the difference was not significant (Wilcoxon signed rank test; p = 0.094). The differences between the cultivars exhibiting the highest and lowest glucose content were very large: 25-fold in 2013 (Manamusume, 13.27 g/L; Kamenoo4, 0.538 g/L) and 22-fold in 2014 (Tachisuzuka, 17.17 g/L; Yukihikari, 0.79 g/L) (Supplementary Table 1). The glucose content also exhibited a high degree of variation (RSDs of 55.1 and 50.8% in 2013 and 2014, respectively). These data indicate that the liquid hydrolysate glucose content varies considerably in a large number of the cultivars examined, highlighting the importance of selecting appropriate cultivars to maximize bioethanol yield.

Fig. 1. Frequency distributions of the liquid hydrolysate glucose content in 208 rice cultivars in 2013 (a) and 2014 (b).

Glucose content was also analyzed with respect to rice cultivar subspecies (japonica [n = 145] and indica [n = 50]; 13 forage rice cultivars were excluded because they were derived from hybrids of japonica and indica rice). No significant differences were observed with respect to subspecies, year, or interaction between subspecies × year based on two-way analysis of variance (Fig. 2, Table 1; p > 0.05). Furthermore, the subspecies RSDs were similar: 53.5% (japonica in 2013); 51.9% (indica in 2013); 46.2% (japonica in 2014); and 55.2% (indica in 2014). The japonica and indica subspecies exhibit a number of distinct differences in traits such as grain shape²⁰⁾; however, this study revealed no significant difference between these subspecies in terms of liquid hydrolysate glucose content (Table 1; p = 0.2949). Therefore, subspecies suitable for cultivation in particular regions could be selected for producing biofuels from rice straw.

Fig. 2. Frequency distributions of the liquid hydrolysate glucose content in (a) japonica rice cultivars in 2013, (b) indica rice cultivars in 2013, (c) japonica rice cultivars in 2014, and (d) indica rice cultivars in 2014.

Annual variation in glucose content and identification of candidate rice cultivars suitable for biofuel production

The liquid hydrolysate glucose content in the 208 rice cultivars in 2013 and 2014 was significantly correlated (Fig. 3(a); ρ = 0.382, p < 0.001, Spearman’s rank correlation coefficient). The correlation coefficient was not high, however, as several cultivars exhibited quite different values in 2013 vs. 2014. These results suggest that although the overall average liquid hydrolysate glucose content was similar in 2013 and 2014 (Figs. 1 and 2), the glucose content was affected by environmental factors in some cultivars. A previous study reported that the glucose content in the rice straw of some cultivars varies based on environment.¹⁵⁾ In our study, the weather conditions in the summers of 2013 and 2014 were quite different at the experimental field in Kasai City, Hyogo, Japan; there were many hot, sunny days in 2013 but a number of cold and rainy days in 2014 (Fig. 4). In particular, the temperature and solar radiation in early August 2014, when many cultivars were in the flowering or ripening period, were lower than during the same period in 2013. Because carbohydrates accumulate in organs such as the culm and leaf before flowering and are then transported to the panicles to fill grains after flowering,²¹⁾ these weather differences in 2013 and 2014 could explain the differences in glucose content between some cultivars. Therefore, when selecting cultivars, suitable for biofuel production, annual variations in environmental factors should be considered.

Fig. 3. Liquid hydrolysate glucose content of 208 rice cultivars in 2013 and 2014 (a) and relationships of liquid hydrolysate glucose content and total glucose content in rice straw in 2014 (b).

Fig. 4. Temperature and solar radiation at the Kobe University, Food Resources Education and Research Center (Kasai, Hyogo Prefecture, Japan). Lines show the temperature in 2013 and 2014. Bars show the difference in solar radiation (the value in 2014 minus that in 2013).

Total glucose content of whole rice straw harvested in 2014 was estimated from liquid hydrolysate glucose content and straw weight to validate whether liquid hydrolysate glucose content in rice main straw reflects the glucose amount of total straw. The total glucose content was highly correlated with liquid hydrolysate glucose content in main rice straw (Fig. 3(b); ρ = 0.904, p < 0.001, Spearman’s rank correlation coefficient). Therefore, the glucose content measured by our high-throughput method is enough to account for the glucose content in total straw. The high correlation is attributed to the positive correlation between liquid hydrolysate glucose content and straw weight (ρ = 0.532, p < 0.001, Spearman’s rank correlation coefficient).

The cultivars in the highest and lowest glucose-content quartiles in both 2013 and 2014 are shown in Fig. 5. Despite the environmental variations described above, several cultivars exhibited stable, high liquid hydrolysate glucose content in both years, indicating that they are ideal choices for use in bioethanol production. Tachisuzuka, Leaf star, and Momiroman are Japanese forage cultivars that exhibited high glucose content in both years. Of the 208 cultivars examined in the study in 2013 and 2014, Tachisuzuka ranked 11th and 1st; Leaf star ranked 18th and 3rd; and Momiroman ranked 31st and 33rd, respectively. A number of reports have indicated that these cultivars accumulate a considerable amount of sugar in their straw.^14,22,23) Forage rice cultivars are also known for their high biomass yield and lodging resistance.²⁴⁾ However, the grain quality and yield of forage rice cultivars are generally low, making them poorly suited for use as cooking rice. The cooking rice cultivars Norin22 (2nd and 24th in glucose content in 2013 and 2014, respectively) and Natsuhikari (38th and 30th glucose content in in 2013 and 2014, respectively) are thus candidates for use in bioethanol production in Japan. Another bioethanol production candidate cultivar, Yamadanishiki, which exhibited high glucose content in both 2013 and 2014, is the most popular cultivar for producing Japanese rice wine, or ‘sake’.²⁵⁾ Among 8 cultivars examined in 2009, Matsuda et al.¹³⁾ reported that the Yamadanishiki cultivar exhibited the highest glucose content, and its rank in the present study was 48th in 2013 and 17th in 2014. Since the high-glucose cultivars also showed high total glucose content in 2014 (Fig. 3(b)), the candidate cultivars were prominent for efficient bioethanol production. Despite the small number of repetition (n = 3) using 1 or 2 rice main straws and the environmental variations in 2013 and 2014 (Fig. 5), the present result generally agreed with the previous reports referring to high glucose cultivars.^13,14) Consequently, we propose that liquid hydrolysate glucose content by our high-throughput method is used as indicator for screening the cultivars suitable for bioethanol feedstocks from large population.

Fig. 5. Upper and lower quartiles of glucose content in 2013 and 2014. Error bars show standard deviation.

In addition to their suitability as feedstocks for bioethanol production, Norin22, Natsuhikari, and Yamadanishiki produce high-quality grain. Given that their straws can be used for bioethanol production, the efficient utilization of both the straw and grain of these cultivars may be realized.

Although a number of other cultivars are not popular in Japan for disagreement with Japanese palatability and cultivation, they could serve as genetic resources for the development of new cultivars for use as bioethanol feedstocks. Tanger et al.¹⁵⁾ used the world’s rice collection: IR64 (a popular and widely adaptable cultivar worldwide) and Nipponbare (widely used in rice genome analyses).²⁶⁾ In the present study, these cultivars ranked relatively low in 2013 and 2014 in terms of liquid hydrolysate glucose content: 115th and 53rd for IR64, respectively, and 126th and 109th for Nipponbare, respectively. To provide information that may lead to more efficient utilization of both of rice straw and grain, the large-scale screening of the glucose content of a diverse array of rice cultivars undertaken in the present study was worthwhile. The data generated here should prove valuable in the selection and use of candidate cultivars for bioethanol production.

Factors affecting liquid hydrolysate glucose content and future prospects

A previous study reported that the glucose present in rice straw liquid hydrolysates after dilute sulfuric acid pretreatment is liberated primarily from starch rather than cellulose.¹⁴⁾ Similar to our results, Arai-Sano et al.²⁷⁾ reported that the starch content of rice straw varies between cultivars. In the present study, the cultivars bred in Hokkaido (the northernmost island of Japan) demonstrated a low glucose content (Fig. 5; e.g. cultivars Hayamasari, Eiko, Akage, Nanatsuboshi, Yukimaru, and Kirara397). The flowering of these cultivars occurred too early for the plants to grow well due to photoperiod insensitivity and shorter day length (our experimental field was at a lower latitude than Hokkaido). Thus, several cultivars grown in other regions could be candidates for efficient bioethanol production. Some Japanese glutinous rice cultivars (Toyohatamochi, Koganemochi, and Hiyokumochi) harboring mutations affecting grain starch composition²⁸⁾ exhibited low glucose content in the present study (Fig. 5). Furthermore, the high-glucose cultivars Yamadanishiki and Tachisuzuka also have unique grains; the former has a white core grain, which is a kind of chalky grain and is important in sake brewing,²⁹⁾ and the latter has a short panicle resulting from mutation.²²⁾ These observations suggest that the type of rice grain may play a role in the accumulation of glucose. As there were no significant differences between the japonica and indica subspecies in terms of the liquid hydrolysate glucose content (Table 1; p = 0.2949), the relationship between grain characteristics and glucose content in rice straw should be investigated further.

The present study of a large number of rice cultivars revealed that the glucose content in the straw varies considerably between cultivars. Therefore, selecting the most suitable cultivars and improving the characteristics of less-suitable cultivars are critical in the production of bioethanol from rice straw. Because we found that some high-quality rice cultivars are more suitable as bioethanol feedstocks, cultivars more suitable for use in both cooking and as biomass resources for bioethanol production could be developed through future breeding efforts. In addition, the prominent cultivars should be investigated in field-scale level about other straw properties such as straw weight and the content of cellulose, xylose, and fermentation inhibitors since low transport cost of rice straw and high bioethanol production efficiency are important as much as abundant fermentation substrate to realize commercial production.

Author contributions

TG, HT, KK, CO, and MY conceived the experiments. TG, HT, MS, and MY performed the experiments and analyzed the data. TG, HT, KK, CO, and MY discussed the results. KK, HK, CO, AK, and MY contributed reagents, materials, and analyze tools. TG and MY wrote the paper.

Funding

This work was supported by the Special Coordination Fund for the Promotion of Science and Technology, Creation of Innovation Centers for Advanced Interdisciplinary Research Areas (Innovative Bioproduction Kobe), MEXT, Japan

Supplemental materials

Supplemental material for this article can be accessed at http://dx.doi.org/10.1080/09168451.2015.1136882.

Disclosure statement

No potential conflict of interest was reported by the authors.

Acknowledgments

We are grateful to National Agriculture and Food Research Organization: National Agricultural Research Center, National Agricultural Research Center for Hokuriku Research Center, National Institute of Crop Science, National Agricultural Research Centers for Hokkaido Region, Tohoku Region, Western Region, and Kyushu Okinawa Region; National Institute of Agrobiological Sciences, Kyushu University, and Prefecture Agricultural Experiment Centers or Stations: Hokkaido, Aomori, Miyagi, Akita, Yamagata, Ibaraki, Gunma, Chiba, Niigata, Ishikawa, Fukui, Gifu, Aichi, Okayama, Yamaguchi, Fukuoka, Kumamoto, Miyazaki, and Kagoshima for rice seeds. We thank Dr. Thomas Lübberstedt (Department of Agronomy, Iowa State University, Ames, IA, USA) for advice regarding the publication of this paper. We are grateful to Kentaro Masaki and Yuko Watanabe (Food Resources Education and Research Center, Graduate School of Agricultural Science, Kobe University) for giving us weather data.