Cell-free synthesis and purification of CelJ
To analyze the activity of full-length CelJ for various polysaccharides, we constructed plasmids encoding full-length CelJ and the N- and C-terminal segments of CelJ, Cel9D and Cel44A, respectively, which contain the GH9 and CBM30 modules and the GH44 and CBM44 modules, respectively (Fig. 1a). In previous studies, the CelJ deletion derivatives Cel9D and Cel44A were purified to homogeneity (Ahsan et al., 1997; Arai et al., 2003; Najmudin et al., 2006), although the purification of full-length CelJ from recombinant E. coli cells was unfeasible as a result of endogenous proteolysis (Ahsan et al., 1996). However, here, we attempted to synthesize full-length CelJ and its deletion derivatives as GST-fusion proteins by in vitro translation using a wheat germ cell-free protein synthesis system. The synthesized fusion proteins were recovered in soluble form, purified to homogeneity by glutathione affinity chromatography, and cleaved to yield the full-length and deletion derivatives (Fig. 1b). Purification of the full-length form of CelJ, which was 176 kDa in size, was confirmed by SDS-PAGE (Fig. 1b) and Western blotting analysis using an antibody against the C-terminal FLAG tag (Fig. 1c). As estimated by SDS-PAGE and densitometric analyses, the amount of CelJ purified from 1 mL of translation mixture was 10 μg.
Substrate specificity of CelJ
The enzymatic properties of Cel9D and Cel44A, such as optimal pH and temperature, and substrate specificities have been reported in detail (Ahsan et al., 1997; Arai et al., 2003; Najmudin et al., 2006). Cel9D has a high activity for CMC and a low activity for acid-swollen cellulose (ASC) and Avicel, but displays no detectable activity towards oat spelt xylan (Arai et al., 2003). In contrast, although Cel44A also exhibits high and low activities for CMC, and ASC and Avicel, respectively, it has significant activity for oat spelt xylan (Ahsan et al., 1997; Najmudin et al., 2006). These results are reasonably consistent with their observed substrate binding properties; the carbohydrate-binding module of Cel9D (CBM30) binds to CMC, ASC, and Avicel, but not to oat spelt or birch wood xylan (Arai et al., 2003; Najmudin et al., 2006), whereas that of Cel44A (CBM44) binds to CMC, ASC, Avicel, and oat spelt xylan (Najmudin et al., 2006). Moreover, Cel44A displays a high activity for xyloglucan (a highly branched β-1,4-glucan), which was shown to bind CBM44 (Najmudin et al., 2006). Although the xyloglucanase activity of Cel9D has not been reported, CBM30 binds to xyloglucan with lower affinity than that of CBM44 (Najmudin et al., 2006). The substrate specificity of full-length CelJ was previously investigated using a crude extract obtained from CelJ-expressing recombinant E. coli cells, and both CMCase and xylanase activities were detected by zymogram analysis of the cell lysate (Ahsan et al., 1996). Prior to the present study, however, the activity of purified CelJ had not been quantified.
Here, we examined the substrate specificities of full-length CelJ and its deletion derivatives by measuring the hydrolytic activities of these purified proteins for various cellulosic substrates, including Avicel, PASC, CMC, xylan, and xyloglucan. In addition, we analyzed the synergy of CelJ for the degradation of these cellulosic substrates by comparing the specific activities per μmol between CelJ and an equimolar mixture of Cel9D and Cel44A (Cel9D+Cel44A) (Table 1). Cel9D showed a high activity for CMC and xyloglucan, but no detectable activity for Avicel, PASC, or xylan, whereas Cel44A exhibited relatively high activity for CMC, xylan, and xyloglucan, but had no detectable activity towards Avicel or PASC. Although our assay system showed a high background level (2.5 U μmol−1), these results were comparable to the previous findings for Cel9D and Cel44A, as described above (Ahsan et al., 1997; Arai et al., 2003; Najmudin et al., 2006). In contrast, CelJ showed a high activity for CMC, xylan, and xyloglucan, a low activity for PASC, and no detectable activity for Avicel. Notably, the equimolar mixture of Cel9D and Cel44A had no detectable activity for PASC (< 2.5 U μmol−1), in contrast to full-length CelJ, which showed at least fourfold higher activity for this substrate (10 U μmol−1), indicating that the fusion of Cel9D and Cel44A results in synergy for PASC degradation. Similarly, the fusion of Cel9D and Cel44A also generated synergy for xylan degradation.
Table 1. Enzymatic activities of CelJ and its deletion derivatives for various cellulosic substrates
|Substrate||Specific activity (U μmol−1)|
Among the examined substrates, CelJ showed the highest specific activity for xyloglucan degradation, a finding that is attributable to the substantial synergy generated by the fusion of Cel9D and Cel44A. Specifically, the apparent enzymatic activities of Cel9D+Cel44A for CMC and xyloglucan were 2.9- and 1.3-fold higher, respectively, than the theoretical activities, which were determined by summing the individual activities of Cel9D and Cel44A (480 vs. 166 U μmol−1 for CMC, and 1018 vs. 770 U μmol−1 for xyloglucan). The activities of CelJ were 1.6- and 3.0-fold higher for CMC and xyloglucan, respectively, than those of Cel9D+Cel44A (770 vs. 480 U μmol−1 for CMC, and 3070 vs. 1018 U μmol−1 for xyloglucan). These results indicate that the combination of Cel9D and Cel44A generated more synergy for CMC degradation than for xyloglucan degradation, whereas the physical fusion of Cel9D and Cel44A generated more synergy for xyloglucan degradation than for CMC degradation.
Based on our analyses, CelJ has 4.0-fold higher specific activity for xyloglucan (3070 U μmol−1) than for CMC (770 U μmol−1) (Table 1). However, CelJ cannot be classified as a specific xyloglucanase, such as C. thermocellum Xgh74A (Zverlov et al., 2005; Martinez-Fleites et al., 2006), because specific xyloglucanases are defined as enzymes that exhibit at least 10-fold higher activity for xyloglucan than for CMC (Grishutin et al., 2004), and CelJ did not meet this criterion. Furthermore, although CelJ was previously identified by proteomic analysis as a major enzymatic subunit of the C. thermocellum ATCC27405 cellulosome obtained from cells grown on crystalline cellulose (Gold & Martin, 2007; Raman et al., 2009), the proportion of CelJ increased when the carbon source was changed from crystalline cellulose (i.e. homogeneous cellulose) to pretreated switchgrass (i.e. highly heterogeneous polysaccharides), as is commonly observed for hemicellulases (e.g. xylanases or xyloglucanases) (Raman et al., 2009). Thus, the broad substrate specificity exhibited by CelJ may play an important role in the degradation of plant cell walls composed of highly heterogeneous polysaccharides.
Although it remains unclear why synergy for the degradation of CMC and xyloglucan in particular is generated upon the fusion of Cel9D and Cel44A, the different modes of Cel9D and Cel44A activities upon hydrolysis may be an underlying factor. CelJ is predicted to interact with CMC and xyloglucan through CBM44-mediated binding, because C-terminal CBM44 shows 10-fold higher binding affinity to these substrates than N-terminal CBM30 (Najmudin et al., 2006). The GH44 module adjacent to CBM44 cleaves CMC and xyloglucan at the internal β-1,4-glucosidic bond (Ahsan et al., 1997; Najmudin et al., 2006). Following cleavage of this bond, Cel44A introduces the neighboring Cel9D to the reducing end of this cleavage site, and the introduced Cel9D proceeds to hydrolyze the chain from the reducing end. Cel9D is proposed to function as a semi-processive endoglucanase because the enzyme reduces the viscosity of a CMC solution at a slower rate than C. thermocellum CelC, a typical endoglucanase (Arai et al., 2003). This finding suggests that Cel9D cleaves CMC at the internal β-1,4-glucosidic bond and proceeds to hydrolyze the substrate from the reducing end of the cleavage site with low processivity. In fact, the combination of Cel9D and CelC generated clear synergy for ASC degradation (Arai et al., 2003), supporting the notion that Cel9D and Cel44A act synergistically. Moreover, a similar synergistic effect on cellulolytic activity generated by the fusion of endo- and exo-glucanases has been reported (Riedel & Bronnenmeier, 1998); the artificial fusion of GH9 processive endoglucanase CelZ and GH48 exoglucanase CelY from Clostridium stercorarium showed twofold synergy for the degradation of Avicel as compared with the equimolar mix of these cellulases, termed ‘intramolecular synergism’. Further studies on the mode of CelJ action are needed to reveal the mechanism underlying the synergistic action between Cel9D and Cel44A in full-length CelJ.