Isolation and characterization of two thermophilic cellulolytic strains of Clostridium thermocellum from a compost sample



Zhongtang Yu, Department of Animal Sciences, The Ohio State University, 2029 Fyffe Road, Columbus, OH 43210-1095, USA. E-mail:



To isolate, identify and characterize new thermophilic cellulolytic bacterial strains from a compost sample.

Methods and Results

Two thermophilic and cellulolytic bacterial strains were isolated via enrichment on cellulose (milled filter paper) and characterized. Both strains, CS7 and CS8, were rod-shaped, Gram-positive and spore-forming bacteria, sharing the same optimal temperature (60°C) and pH (7·0) for growth. Both were highly cellulolytic and produced ethanol and acetate as the major fermentation products, but lacked xylanase activity. They only grew on cellulose (both filter paper and crystalline cellulose) and cellobiose and produced yellow pigment, without growing on other substrates including glucose. Based on 16S rRNA gene sequence analysis, CS7 and CS8 are closely related (99% sequence identity) to Clostridium thermocellum ATCC 27405. However, they had significantly higher specific cellulase activities and ethanol/acetate ratios than Cl. thermocellum ATCC 27405.


CS7 and CS8 are two new highly cellulolytic and ethanologenic Cl. thermocellum strains.

Significance and Impact of the Study

First report of applying the cloning-RFLP-sequencing approach for purity confirmation of the isolates beside conventional methods. Strains CS7 and CS8 might be of potential application in research and development of cellulosic bioconversion.


Bioethanol produced from starch (primarily corn in the USA) has been in the spotlight for the last few years, but it became evident that corn ethanol is not sustainable because corn is a major grain for both food and feed. The environmental benefits of corn ethanol have also been questioned (Landis et al. 2008). Because lignocellulosic biomass (e.g. wood chips, crop residues and energy crops) represents the most abundant renewable biomass available for conversion to biofuels (Demain et al. 2005), biofuel production from lignocellulosic biomass has attracted most of the research and development interest from both the scientific community and the industry in recent years. However, few cost-effective technologies can overcome the recalcitrance of lignocellulosic biomass, hindering utilization of this important renewable resource. A promising strategy is consolidated bioprocessing consisting of production of cellulolytic enzymes, hydrolysis of biomass and fermentation of resulting sugars to desirable products in a single process using a cellulolytic microorganism or consortium (Lynd et al. 2002). Among the known cellulolytic microbes, it has been recognized that thermophilic anaerobes with complexed cellulase systems (cellulosome) have greater cellulolytic activities than their mesophilic peers or aerobes with free cellulases, for example Trichoderma reesei (Lu et al. 2006; Báez-Vásquez and Demain 2008). Thus, thermophilic cellulolytic anaerobes are highly sought after. In this study, two new strains of thermophilic cellulolytic anaerobic bacteria were isolated and characterized with respect to their growth conditions, substrate spectrum and fermentation products. They were also compared with the model thermophilic cellulolytic strain, Clostridium thermocellum ATCC 27405.

Materials and methods

Isolation of cellulolytic strains

A compost sample was collected from a compost windrow at the Ohio Compost and Manure Management facility (Wooster, OH, USA) and inoculated into the EM medium (Champion et al. 1988) with pebble-milled filter paper as the sole carbon source to enrich cellulolytic microbes. The culture was incubated at 55°C and transferred into the same medium upon obvious degradation of the filter paper substrate for 20 times (1–2 days between transfers). In an anaerobic chamber, the enrichment culture was serially diluted in EM medium containing no carbon source; and 0·2 ml of the dilutions was mixed with 6 ml of soft EM agar (0·5% agar) containing milled filter paper. The soft agar mixtures were overlaid onto prepoured EM agar (2% agar) plates containing no carbon source. The agar plates were sealed with Parafilm and incubated at 50°C for 2 days. Individual colonies that produced clear halos due to cellulolytic activities were picked and inoculated into EM broth containing milled filter paper to confirm cellulose degradation.

Purification and identification of isolates

After confirmation of cellulose degradation activities, the broth cultures were subjected to a purification step by dilution to extinction using EM broth containing cellobiose because the isolates could not form any colony on the surface of normal agar plates. The dilution to extinction was repeated twice, and only the isolates that remained cellulolytic after purification were further characterized. The purity of each isolate was examined by PCR amplification and cloning of the 16S rRNA gene, followed by RFLP and sequencing. The 16S rRNA genes were amplified using primers 27F and 1525R and sequenced as described previously (Yu et al. 2000). To further examine the purity and phylogenetically identify the isolates, the sequences were compared with GenBank using Blast and classified using the RDP Classifier (Wang et al. 2007). The pure isolates were also examined with respect to their cellular morphology and Gram staining.

Due to frequent contamination in cultures of thermophilic cellulolytic bacteria (Erbeznik et al. 1997), three additional criteria described previously (Freier et al. 1988) were also employed for the purity confirmation of the isolates: (i) absence of butyrate production (<0·1 mmol l−1) and a final culture pH above pH 5·9, (ii) no growth on glucose or cellobiose at 72 or 30°C and (iii) remaining cellulolytic after five consecutive subcultures on cellobiose.

Determination of optimal growth temperature and pH

The optimal temperature for growth and cellulose degradation was determined using EM broth (pH 6·7) containing milled filter paper as sole substrate at 25, 35, 45, 50, 55, 60, 65 and 75°C. Growth, as measured by protein yield, and cellulose degradation, as measured by residual cellulose, were determined during the incubation of each isolate from individual culture tubes terminated accordingly. All the cultures were done in three replicates for each time point. Each sample was equally divided into two aliquots. One aliquot was used to determine the residual cellulose using the anthrone assay (Dische 1962), while the other aliquot was analysed for cellular protein using a BCA Protein Assay kit (Pierce, Rockford, IL, USA) as described by Reveneau et al. (2003). The optimal pH for both growth and cellulose degradation was determined similarly by growing each isolate at their respective optimal temperatures and at pH 5·0, 6·0, 7·0, 7·5, 8·0, 9·0 and 10·0.

Substrate diversity, growth and cellulose degradation kinetics, and fermentation products

At the optimal temperature and pH, each isolate was tested for its ability to use common (poly)saccharides including xylan (birchwood), crystalline cellulose (Avicel, PH-101), glucose, cellobiose, xylose and carboxymethyl cellulose (CMC) in EM broth containing 1% (w/v) of each substrate. These cultures were incubated for 5 days, with growth on soluble substrates being determined by absorbance at 600 nm every 24 h, while growth on insoluble substrates being determined by substrate disappearance at the end of the incubation. For each substrate, three replicates and a control without inoculation were included.

Under optimal conditions, growth and cellulose degradation kinetics were determined for the isolates and Cl. thermocellum ATCC 27405 using EM broth containing milled filter paper in three replicates for each time point. As done in determining the optimal temperature and pH, three culture tubes of each strain were terminated at preset intervals (1·5, 3·0, 4·5, 6·0, 7·5, 9·0, 12·0, 15·0, 18·0 and 20·0 h). Each sample was equally divided into two aliquots and subjected to analysis for residual cellulose and cellular protein using the anthrone assay and the BCA Protein Assay kit, respectively.

The new isolates and Cl. thermocellum ATCC 27405 were cultured in EM broth containing 0·3% (w/v) milled filter paper under their optimal conditions. Each strain was inoculated into six replicate cultures, with three of them not incubated and the other three incubated until the substrate disappeared. Depletion of the substrate was confirmed by the anthrone assay with half of each culture, while the other half was used to analyse the major fermentation products [i.e. volatile fatty acid (VFA) and solvents], using gas chromatography as described previously (Han et al. 2011; Patra and Yu 2012).

Specific cellulase activities

The new isolates and Cl. thermocellum ATCC 27405 were grown in EM broth in three replicates with 0·3% (w/v) milled filter paper as the sole substrate until substrate depletion. The cell biomass of each culture was harvested by centrifugation at 16 000 g at room temperature for 1 min, washed three times using a phosphate buffer (100 mmol l−1, pH = 7·0) and resuspended in 1/10 of the original culture volume with the same buffer (referred to as 10× cells). Sixty microlitres of each 10× cells was added into 540 μl of 1% milled filter paper, mixed and incubated aerobically at 60°C. In total, 27 such hydrolytic reactions were prepared for each strain so that hydrolysis data could be collected for nine time points: 0, 5, 10, 20, 30, 60, 120, 180 and 240 min, with three replicates for each time point. All the samples were subjected to analysis for reducing sugar using the DNS assay (Miller 1959). The total cellular protein concentration of the 10× cells of each strain was measured using the BCA Protein Assay kit. One unit of cellulase activity was defined as 1 μmol reducing sugar released per mg cellular protein per minute.

Statistical analysis

Cellulose degradation rate, growth rate and specific cellulase activity of each strain under different conditions (temperature and pH) were analysed by linear regression, using time, conditions, strains and their interactions as covariates. Time was treated as a continuous variable, and conditions/strains were treated as categorical variables. A t-test was used to make pairwise comparisons among means of ethanol/acetate ratios of fermentation products from different strains. In both cases, P-values <0·05 were defined as statistically significant. Standard deviations were also calculated for all the data.

Sequence accession number

The sequences of the 16S rRNA genes of CS7 and CS8 have been deposited in GenBank under the accession numbers JX912711 and JX912712, respectively.


Isolation, purification and identification of isolates

Sixteen colonies that exhibited a clear halo were obtained after incubation from the overlaid plates from the 1 : 10 diluted enrichment culture. The higher dilutions (10−2 to 10−5) of the enrichment culture did not yield any colony with a clear halo. During the three rounds of purification using dilution to extinction, 14 isolates either lost their cellulolytic activities or failed to produce a consistent RFLP pattern from their 16S rRNA genes. Two strains, referred to as CS7 and CS8, were obtained as pure cultures as indicated by a consistent RFLP pattern and sequencing results from multiple clones of their 16S rRNA genes (data not shown). Both strains were Gram-positive rods and produced intracellular terminal oval spores. Both strains CS7 and CS8 appeared to be closely related to Cl. thermocellum ATCC 27405, exhibiting about 99% 16S rRNA gene sequence identity (Fig. 1).

Figure 1.

A phylogenetic tree showing the taxonomic relatedness of strains CS7 and CS8 with other strains of Clostridium thermocellum. GenBank accession numbers are shown in parentheses.

Characterization of strains CS7 and CS8

Based on linear regression, strain CS7 grew and degraded cellulose significantly faster at 60°C than at the other tested temperatures (Fig. 2a,c), except for cellulose degradation at 55°C (P = 0·069) and 65°C (P = 0·188); strain CS8 grew and degraded cellulose significantly faster at 60°C than at the other tested temperatures (Fig. 2b,d), except for cellulose degradation at 55°C (P = 0·073). As a result, 60°C was selected as the optimal temperature for both strains CS7 and CS8. Growth and cellulose degradation rates of strain CS7 at pH 7·0 were the highest among all tested pH values (Fig. 3a,c) and significantly higher than the rates at pH 5·0 and 10·0. Similarly, growth and cellulose degradation rates of strain CS8 were the highest at pH 7·0 and significantly higher than the rates at pH 5·0 and 6·0 (Fig. 3b,d). Thus, pH 7·0 was selected as the optimal pH for both CS7 and CS8.

Figure 2.

Cellulose degradation (a and b) and growth (c and d) of strain CS7 (a and c) and strain CS8 (b and d) at different temperatures. (a) and (b) SD < 0·3; (c) and (d) SD < 6·0. Symbols for different temperatures: open and solid diamond, 25 and 35°C; open and solid square, 45 and 50°C; open and solid triangle, 55 and 60°C; open and solid circle, 65 and 75°C.

Figure 3.

Cellulose degradation (a and b) and growth (c and d) of strain CS7 (a and c) and strain CS8 (b and d) at different pH values. (a) and (b) SD < 0·3; (c) and (d) SD < 8·0. Symbols for different pH values: open and solid diamond, pH 5·0 and 6·0; open and solid triangle, pH 7·0 and 7·5; open and solid circle, pH 8·0 and 9·0; solid square, pH 10·0.

Under optimal conditions (60°C and pH 7·0), both CS7 and CS8 grew on cellulose (filter paper and crystalline cellulose) and cellobiose and produced yellow affinity substance (YAS) as Cl. thermocellum (Ljungdahl et al. 1983), but did not grow on CMC, glucose, xylan or xylose even after prolonged (5 days) incubation. Strains CS7, CS8 and Cl. thermocellum ATCC 27405 exhibited similar cellulose degradation rates (Fig. 4a) and all consumed 0·3% cellulose substrate within 15 h under the determined optimal conditions. The corresponding growth rate of Cl. thermocellum ATCC 27405 was the highest among the three strains (Fig. 4b), significantly higher than that of CS7 (Table 1), and also higher than that of CS8 (P = 0·089). The generation times determined on milled filter paper (0·3%) at 60°C and pH 7·0 were 1·33, 1·27 and 1·20 h for strains CS7, CS8 and Cl. thermocellum ATCC 27405, respectively.

Table 1. Growth rate, cellulose degradation rate and specific cellulase activities of CS7, CS8 and Clostridium thermocellum ATCC 27405 under optimal conditions
 ATCC 27405CS7CS8
  1. Cellulose degradation and growth rates were calculated with the data from 0 to 12 h in Fig. 4a,b; specific cellulase activities were calculated with the data from 0 to 4 h in Fig. 4c; units of cellulose degradation rates, growth rates and specific cellulase activities were expressed as g glucose equivalent l−1 h−1, mg protein l−1 h−1 and μmol glucose equivalent mg cellular protein−1 min−1, respectively. Values within each row with different superscripts were significantly different (P < 0·05).

Cellulose degradation rate0·210·190·21
Growth rate7·96a5·34b6·25a
Specific cellulase activity0·14a0·21b0·21b
Figure 4.

Cellulose degradation (a), growth (b), and specific reducing sugar released by cell-surface associated cellulase (c) of strain CS7 (diamond), strain CS8 (square) and Clostridium thermocellum ATCC 27405 (triangle). (a) SD < 0·25; (b) SD < 7·0; (c) SD < 3·0.

Using 10× cells harvested from cultures grown under optimal conditions, the cell surface-associated cellulase activities were determined by quantifying reducing sugar release from milled filter paper substrate (Fig. 4c). Based on linear regression, specific cellulase activities of CS7 and CS8 were both significantly higher (by 50·0%) than that of Cl. thermocellum ATCC 27405 (Table 1).

Fermentation products were analysed for VFA including acetate, (iso)butyrate, (iso)propionate, and (iso)valerate, and solvents including ethanol, acetone and butanol. Acetate was the only detected VFA, while ethanol was the only detected solvent (Table 2) at the end of 20 h incubation on 0·3% (w/v) milled filter paper under the optimal conditions (60°C and pH 7·0). Based on t-test, the ethanol/acetate ratio of CS7 was significantly higher than that of Cl. thermocellum ATCC 27405, while the ethanol/acetate ratio of CS8 was also higher than that of Cl. thermocellum ATCC 27405 (P = 0·118). No significant difference in ethanol/acetate ratio was observed between strains CS7 and CS8 (Table 2).

Table 2. Major fermentation products of strains CS7, CS8 and Clostridium thermocellum ATCC 24705
StrainAcetate (mmol l−1)Ethanol (mmol l−1)Ethanol/acetate ratio
  1. Data were presented as the means ± SD of three replicate cultures containing 3·21 g cellulose (glucose equivalent) per litre of culture.

  2. P-values for pair comparisons were 0·039 between strains CS7 and ATCC 24705; 0·118 between strains CS8 and ATCC 24705; and 0·179 between strains CS7 and CS8.

CS75·49 ± 0·0817·10 ± 3·573·11 ± 0·62
CS87·28 ± 1·1317·93 ± 5·092·44 ± 0·36
ATCC 247059·17 ± 0·7717·37 ± 4·221·88 ± 0·33


Highly cellulolytic bacteria have attracted great interest in the pursuit of cellulosic biofuels, and thermophilic cellulolytic bacteria are of particular interest due to potentially improved cellulolytic activities at increased temperatures. However, thermophilic cellulolytic bacteria are hard to isolate than their mesophilic peers due to severe evaporation of agar media during incubation, which reduces culturability on solid agar and increases chances of contamination. From an enrichment culture on cellulose at thermophilic temperature, two strains were obtained from the initial 16 colonies that exhibited cellulolytic activities. The other 14 isolates failed to be recovered into pure cellulolytic cultures probably due to difficulties of separating them from noncellulolytic bacteria during the purification by dilution to extinction. It should be pointed out that cellulolytic colonies were only obtained from the 1 : 10 dilution from the enrichment culture. This suggests most cellulolytic bacteria in the enrichment culture failed to grow on the solid agar medium. Other isolation techniques that do not require colony formation (e.g. direct single cell sorting) or minimize the evaporation (e.g. roll-tube) may be more efficient in isolating thermophilic cellulolytic bacteria.

Strains CS7 and CS8 are both closely related to Cl. thermocellum ATCC 27405. Beside their highly similar 16S rRNA genes, CS7 and CS8 were also very similar in terms of optimal growth conditions, growth and cellulolytic activities, substrate spectrum, fermentation products and production of YAS when grown on cellulose and cellobiose. Unlike other Cl. thermocellum strains (Freier et al. 1988; Tachaapaikoon et al. 2012), CS7 and CS8 exhibited no growth on glucose, CMC or xylose. They also showed no xylanase activities even after extended incubation, while most Cl. thermocellum strains have xylanase activities and some Cl. thermocellum strains can even grow on xylan. Thus, both CS7 and CS8 should be considered new strains of the Cl. thermocellum species within the new Clostridium III genus. Strains of Clthermocellum have not been isolated from compost even though its occurrence in compost has been detected by sequencing of a DGGE band (Ueno et al. 2001; Lu et al. 2012). This study suggests that compost may be a source of new strains of Cl. thermocellum, and this species may play an important role in cellulose degradation during composting process.

Filter paper cellulose contains several molecular structures ranging from amorphous regions to crystalline fibres, which can have different recalcitrance to degradation (Mandels et al. 1976). The data of reducing sugar release from the milled filter paper substrate appeared to be biphasic, with the most obvious change point at 20 min (Fig. 4c) reflecting probable change in degradation rate of the filter paper substrate. However, residual analysis showed that all assumptions of our linear model were satisfied. In addition, a high R2 value of 0·982 indicated strong fitting of the model, so the usage of the linear model over the entire time course was warranted. According to statistical analysis, specific cellulase activities of CS7 and CS8 were both significantly higher than that of Cl. thermocellum ATCC 27405. The ethanol/acetate ratios of CS7 and CS8 were also higher than that of Cl. thermocellum ATCC 27405. These two highly cellulolytic and ethanologenic new strains might be used in future studies and development of cellulase enzymes for efficient lignocellulose conversion and as candidates for metabolic engineering for enhanced cellulosic ethanol production.


This study was partially supported by a DOE grant (award number: DE-FG36-05GO85010) and a North East SUN grant (award number: 52110-8512) awarded to Z.Y.