Microbial Diversity for Biotechnology 2014View this Special Issue
Expression of Heterologous Cellulases in Thermotoga sp. Strain RQ2
The ability of Thermotoga spp. to degrade cellulose is limited due to a lack of exoglucanases. To address this deficiency, cellulase genes Csac_1076 (celA) and Csac_1078 (celB) from Caldicellulosiruptor saccharolyticus were cloned into T. sp. strain RQ2 for heterologous overexpression. Coding regions of Csac_1076 and Csac_1078 were fused to the signal peptide of TM1840 (amyA) and TM0070 (xynB), resulting in three chimeric enzymes, namely, TM1840-Csac_1078, TM0070-Csac_1078, and TM0070-Csac_1076, which were carried by Thermotoga-E. coli shuttle vectors pHX02, pHX04, and pHX07, respectively. All three recombinant enzymes were successfully expressed in E. coli DH5α and T. sp. strain RQ2, rendering the hosts with increased endo- and/or exoglucanase activities. In E. coli, the recombinant enzymes were mainly bound to the bacterial cells, whereas in T. sp. strain RQ2, about half of the enzyme activities were observed in the culture supernatants. However, the cellulase activities were lost in T. sp. strain RQ2 after three consecutive transfers. Nevertheless, this is the first time heterologous genes bigger than 1 kb (up to 5.3 kb in this study) have ever been expressed in Thermotoga, demonstrating the feasibility of using engineered Thermotoga spp. for efficient cellulose utilization.
Due to rising global energy demands, developing renewable forms of energy, such as solar, hydro-, and bioenergy, has become increasingly important. Traditionally, bioenergy is generated by fermenting the glucose derived from starch. Starch alone, however, accounts for too small a fraction of biomass to sustain a positive energy balance. The need to replace fossil fuels demands that cellulose, the most common and renewable organic material on Earth , must not be overlooked. Cellulose is a linear polymer of D-glucose units linked by 1,4-β-D-glycosidic bonds. It becomes useful as a food and energy source once it is broken down into soluble cellobiose (β-1,4 glucose dimer) and glucose, a process called hydrolysis because a water molecule is incorporated for each dissociated glycosidic bond. Effective hydrolysis of cellulose requires the cooperation of three enzymes, namely, endo-1,4-β-glucanase (EC 220.127.116.11), exo-1,4-β-glucanase (also called cellobiohydrolase) (EC 18.104.22.168), and β-glucosidase (EC 22.214.171.124) [2–5]. Endoglucanase randomly breaks down the β-1,4 linkages in the regions of low crystallinity, exoglucanase removes cellobiose units from the nonreducing ends of cellulose chains, and β-glucosidase converts cellobiose into glucose. In general, exoglucanases degrade cellulose more efficiently than endoglucanases.
Hyperthermophilic bacteria Thermotoga are attractive candidates for the production of biohydrogen and thermostable enzymes. Surveying of Thermotoga genomes in the CAZy database (http://www.cazy.org/) revealed dozens of carbohydrate-active enzymes, the molecular foundation that allows Thermotoga strains growing on a wide range of carbon sources such as glucose, xylose, semicellulose, starch, and carboxymethyl cellulose (CMC) [6–9]. Nevertheless, the apparent absence of exoglucanases suggests the limited ability of these organisms to use cellulose as their main carbon and energy source. Up to date, there are only two reports describing low levels of exoglucanase activities in Thermotoga [2, 4], a phenomenon probably caused by nonspecific reactions of endoglucanases [10–12].
This study aimed at introducing heterologous exoglucanase activities into Thermotoga by genetic engineering. T. sp. strain RQ2 was selected as the host strain, because its genome encodes the largest repertoire of carbohydrate-active enzymes among all published Thermotoga genomes (Table 1). Moreover, T. sp. strain RQ2 has recently been discovered to be naturally transformable, making the transformation procedure straightforward . The selection of candidate cellulases was focused on Caldicellulosiruptor saccharolyticus, a Gram-positive anaerobe growing optimally at 70°C and can use cellulose as a sole carbon source . Csac_1076 (CelA) [15, 16] and Csac_1078 (CelB) [17, 18] of C. saccharolyticus DSM 8903 have been experimentally characterized as multidomain proteins with both endo- and exoglucanase activities and are suitable candidates to be introduced into T. sp. strain RQ2. However, Caldicellulosiruptor are Gram-positives and Thermotoga are Gram-negatives; CelA and CelB are unlikely to be secreted properly in T. sp. strain RQ2. The Thermotoga host would not benefit much from the heterologous cellulases unless the enzymes can be secreted to the extracellular environment. Signal peptides with a Thermotoga origin would be required to guide the transportation of foreign proteins in T. sp. strain RQ2. A literature search revealed that T. maritima TM1840 (amylase A, AmyA) [19–21] and TM0070 (xylanase B, XynB)  have been experimentally confirmed to be secretive proteins. The former is anchored on the “toga” part with catalytic domain facing outward, and the latter is secreted into the environment after the cleavage of its signal peptide. Therefore, the promoter regions and the signal peptide sequences of TM1840 (amyA) [19–21] and TM0070 (xynB)  were chosen to control the expression and transportation of the Caldicellulosiruptor cellulases in T. sp. strain RQ2.
2. Materials and Methods
2.1. Strains and Cultivation Conditions
The bacterial strains and vectors used in this study are summarized in Table 2. All E. coli strains were cultivated in Luria-Bertani (LB) medium (1% tryptone, 1% NaCl, 0.5% yeast extract) at 37°C. Thermotoga strains were cultivated at 77°C, 125 rpm in SVO medium . SVO plates were made with 0.25% (w/v) gelrite . Thermotoga plates were put into Vacu-Quik Jars (Almore International Inc., Portland, OR, USA) filled with 96 : 4 N2-H2 and 4 g palladium catalyst (to remove oxygen) and incubated at 77°C for 48 h. When needed, ampicillin was supplemented into LB medium to a final concentration of 100 μg mL−1, and kanamycin was added into liquid SVO medium and SVO plates to a final concentration of 150 and 250 μg mL−1, respectively.
2.2. Construction of Vectors
All vectors were constructed by following standard cloning methods and verified by restrictive digestions. Primers used in this study are summarized in Table 3. The Thermotoga-E. coli shuttle vector pDH10 was used as the parent vector. Inverse PCR was performed with pDH10  using primers DBs F and DBs R, and the amplicon was digested with BsaI followed by self-ligation to give rise to pDH26. With the same approach, pDH27 was generated based on pDH26 using primers DNd F and DNd R, and pHX01 was created from pDH27 using primers DLZ F and DLZ R. Compared to pDH10, pHX01 is 342 bp shorter and is free of the BsaI and NdeI recognition sites, which makes it a better cloning vector than pDH10. Based on pHX01, intermediate vectors pHX02.1, pHX04.1, and pHX07.1 were constructed. Vector pHX02.1 carries the promoter and signal peptide region of TM1840 (amyA), which was inserted immediately upstream of the Apr gene (Figure 1(a)); primers AmP F and Amp R were used to amplify the desired region from T. maritima chromosome, and the amplicon was digested with NotI and SacI. Vectors pHX04.1 and pHX07.1 both carry the promoter and signal peptide region of TM0070 (xynB), but one has the insert upstream of the Apr gene (Figure 1(b)) and the other has it downstream of the ori region (Figure 1(c)). Because the two insertions sites were recognized by different restriction enzymes, primers XyBPB F and XyBPB R were used to amplify the insert for pHX04.1, and the amplicon was digested with NotI and SacI; primers XyBPA F and XyBPA R were used to prepare the insert for pHX07.1, and the amplicon was digested by XhoI and PstI. As a result, a BsaI site was introduced immediately after the signal peptide sequence in each vector to facilitate the insertion of the coding regions of the C. saccharolyticus cellulases.
Once the regulation and transportation regions were in place, the cellulases genes from C. saccharolyticus were inserted into pHX02.1, pHX04.1, and pHX07.1 to give rise to pHX02, pHX04, and pHX07, respectively (Figure 1). The total DNA of C. saccharolyticus DSM 8903 was used as the template. Primers CelB F and CelB R were used to amplify Csac_1078 (celB), and the amplicon was digested with BsaI and SacI. Primers CelA F and CelA R were used to amplify Csac_1076 (celA), and the amplicon was digested with BbsI and PstI.
The transformation of E. coli was done with standard calcium chloride method, and the transformation of Thermotoga was done by natural transformation, as described previously . The DNA substrates used to transform Thermotoga were in vitro methylated by methylase M. TneDI [26, 27].
2.3. Detection of Endoglucanase Activity with CMC Plates
Endoglucanase activities were evaluated using Congo red assays . For preliminary screening, E. coli transformants were inoculated on CMC plates (1% NaCl, 0.5% yeast extract, 0.2% CMC, 1.5% agar) and incubated at 37°C for overnight. The plates were then kept at 77°C for 8 h, stained with 0.1% Congo red (dissolved in water) at room temperature for 15 min, and washed with 1 M NaCl until clear halos showed up. After that, the plates were rinsed with 1 M HCl, which changed the background color into blue, providing a better contrast for the halos. To test liquid cultures, 40 μL of each normalized overnight culture was directly loaded onto CMC plates. To minimize the sizes of the loading spots, the liquid cultures were loaded through 8 times with 5 μL in each loading. The next round of loading only happened when the liquid from the previous loading had been completely absorbed. To localize the expression of the recombinant proteins, supernatants were collected from 1 mL of normalized overnight culture by centrifugation. Meanwhile, the cells were washed once with fresh medium and resuspended in 1 mL of the same medium.
2.4. Detection of Endoglucanase Activities with Zymogram
Native polyacrylamide gel electrophoresis was modified from a previous report . SDS (sodium dodecyl sulfate) was omitted from the gel (10%, w/v), but CMC was added to a final concentration of 0.08% (w/v). Protein samples were prepared in the absence of SDS, reducing agents, and the heat treatment. After electrophoresis, gels were rinsed with deionized water for 3 times prior to immersion in 0.25 M Tris-HCl (pH 6.8) for 8 h at 77°C. Following the enzymatic reaction, gels were visualized with Congo red, as detailed above.
2.5. Detection of Exoglucanase Activity
MUC (4-methylumbelliferyl β-D-cellobioside) agar was used to detect exoglucanase activity [17, 30]. Under the hydrolysis of exoglucanase, MUC is converted to cellobiose and MU (4-methylumbelliferone), which shows fluorescence under ultraviolet light. Forty microliters of normalized overnight culture of each Thermotoga transformant was spotted on MUC plates, incubated at 77°C for 8 h, and examined under UV light. The formation of fluorescent halos surrounding the loading spots indicates the activity of exoglucanase.
3. Results and Discussion
3.1. Expression and Localization of the Chimeric Enzymes in E. coli
In the endoglucanase activity screening experiment, all tested DH5α/pHX02 (Figure 2), DH5α/pHX04, and DH5α/pHX07 strains showed clear halos surrounding the overnight colonies, indicating functional expression of the recombinant cellulases in E. coli. To determine the localizations of the recombinant cellulases, cultures were normalized and supernatants and cell suspensions were tested separately. Because pHX07 and pHX04 share the same localization signal, experiments were carried out with just pHX04 and pHX02 transformants. On CMC plates, the pHX04 transformant demonstrated a higher endoglucanase activity than the transformants of pHX02 (Figure 3). For both constructs, most of the endoglucanase activity was associated with the cell suspensions (Figure 3). This is not surprising for pHX02, because its fusion protein is designed to be anchored on the outer membrane of a Gram-negative host. As for pHX04, since its fusion protein is meant to be released into the medium, these data suggest that the signal peptide of TM0070 (XynB) is not functional in E. coli, even though its promoter is.
3.2. Detection of Endoglucanase Activities in Thermotoga
As our main purpose is to express the recombinant cellulases in Thermotoga, we next transformed these 3 expression vectors into T. sp. strain RQ2. Eleven T. sp. strain RQ2/pHX02, eleven RQ2/pHX04, and eight RQ2/pHX07 transformants were isolated and tested with Congo red assays. Wild type T. sp. strain RQ2 and C. saccharolyticus DSM 8903 were used as the negative and positive controls. Almost all transformants showed enhanced endoglucanase activities compared with the wild type strain (Figure 4), indicating the successful transformation and expression of the recombinant enzymes. Next we set out to validate the transformants with PCR and restriction digestions.
3.3. Validating Thermotoga Transformants
Three RQ2/pHX02 transformants (#2, #3, and #4) and three RQ2/pHX04 transformants (#3, #4, and #5) were picked up from SVO plates and grown overnight in liquid SVO medium with 150 μg kanamycin mL−1. Plasmid extracts were prepared and used as the templates to amplify the exoglucanase domain of recombinant celB gene with primers celBV F and celBV R. One RQ2/pHX02 isolate (#4) and two RQ2/pHX04 isolates (#4 and #5) showed bands with the expected size (in addition to some nonspecific bands) (Figure 5). The positive bands from RQ2/pHX02 #4 and RQ2/pHX04 #5 were gel-purified, reamplified using the same primers, and digested with HaeIII (Figure 6). Both amplicons developed the expected digestion profile, demonstrating the authenticity of the two transformants, which were then selected to be further characterized in later studies. The attempts to amplify either exo- or endoglucanase domain from the pHX07 transformants failed. Since pHX07 carries the celA gene, instead of the celB as found in pHX02 and pHX04, further optimization of PCR conditions and/or primer selections may eventually allow one to validate the transformants of this vector. Nevertheless, pHX07 #2 was selected for further studies, because it at least displayed strong endoglucanase activity on CMC plates.
3.4. Detection of the Exoglucanase Activity in Thermotoga
The exoglucanase activity of the Thermotoga transformants was tested with MUC plates (Figure 7). Compared to the host strain, RQ2/pHX02 and RQ2/pHX04 demonstrated greatly enhanced exoglucanase activities, as bright fluorescent light emitted under UV light from the spots where their overnight cultures were loaded. This suggests that the exoglucanase domain of Caldicellulosiruptor CelB was successfully expressed and fully functional in T. sp. strain RQ2. However, the fluorescence level displayed by RQ2/pHX07 was at about the same level to the wild type strain. Because the recombinant enzymes carried by pHX04 and pHX07 share the same promoter and signal peptide, the low level of exoglucanase activity presented by RQ2/pHX07 indicates the exodomain of Caldicellulosiruptor CelA was either lost (which echoes the PCR results above) or not functional in T. sp. strain RQ2.
3.5. Localization of the Recombinant Cellulases in Thermotoga
Localization of the recombinant enzymes was carried with T. sp. strain RQ2/pHX02 #4, pHX04 #5, and pHX07 #2 by comparing the endoglucanase activity in supernatants versus cell suspensions. Unlike what happened in E. coli where the majority of the activity was associated with cell suspensions, the Thermotoga transformants had about half of the endoglucanase activity found in supernatants (Figure 8(a)). The endoglucanase activity in the supernatants was double-checked with native polyacrylamide gels followed by Congo red assay. All supernatants showed brighter bands than the wild type strain, indicating enhanced cellulases activities (Figure 8(b)). No protein bands were detectable in the supernatants by Coomassie brilliant blue staining. The cell suspensions retained the other half of the enzymatic activities, probably because of the cytoplasmic proproteins of the chimeric enzymes (Figure 8(a)).
3.6. Stabilities of the Recombinant Strains
Stabilities of the E. coli and T. sp. strain RQ2 recombinant strains were tested by consecutively transferring corresponding cultures under the selection of antibiotics. After four transfers, all shuttle vectors were readily detected in the E. coli transformants (Figure 9(a)). The endo- and exodomains of celA and celB were also successfully amplified from the plasmid DNA extracts (data not shown). Congo red assays with DH5α/pHX02 showed that the enzyme was as active as before and the expression level of the enzyme was not affected by the inclusion of 0.25% starch in the medium [21, 25] (Figure 9(b)). These results indicate that, in E. coli DH5α, the constructed vectors are stably maintained and the enzymes are constitutively expressed.
Unfortunately, in T. sp. strain RQ2, the vectors seemed to be gradually lost, as indicated by decreasing enzyme activities with each transfer. After the 3rd transfer, the activities of both endo- (Figure 10) and exoglucanase (data not shown) were at the same level as the negative control. Trying to induce the cultures with 0.25% starch or 0.25% xylose [21, 25, 31] did not result in improved expression of the enzymes, suggesting the diminishing of enzyme activities was a result of loss of genes rather than a lack of expression. Our previous study demonstrated that the Thermotoga-E. coli shuttle vector pDH10 is stably maintained in both E. coli and Thermotoga . However, the vectors constructed in this study, which are derived from pDH10, were only stable in E. coli, but not in Thermotoga. This might be due to different genetics of the Thermotoga hosts. In the previous study, T. sp. strain RQ7 and T. maritima were used, and in this study, the host was T. sp. strain RQ2. Cryptic miniplasmids pRQ7 and pMC24 have been found in T. sp. strain RQ7 and T. maritima [32, 33], but no natural plasmids have ever been seen in T. sp. strain RQ2. Plasmids pRQ7 and pMC24 are only 846 bp in length and encode just one apparent protein, which seems to play a role in plasmid replication but lacks the site-specific nuclease activity typical to a fully functional replication protein . It is possible that the genomes of T. sp. strain RQ7 and T. maritima encode gene(s) essential to the replication of pRQ7-like plasmids, allowing the survival of pRQ7/pMC24-based vectors, whereas T. sp. strain RQ2 may lack such gene(s). As about half of the Thermotoga genomes encode uncharacterized proteins, finding such gene(s) requires thorough functional genomics studies and will be the future direction of our work.
This work demonstrated that it is possible to functionally express large heterologous proteins in Thermotoga. Transformed with the recombinant Caldicellulosiruptor cellulases, T. sp. strain RQ2 displayed increased endoglucanase activity with the expression of all three engineered enzymes, namely, TM1840 (AmyA)-Csac_1078 (CelB), TM0070 (XynB)-Csac_1078 (CelB), and TM0070 (XynB)-Csac_1076 (CelA). Exoglucanase activity was also improved significantly in T. sp. strain RQ2 transformants expressing the chimeric enzymes TM1840 (AmyA)-Csac_1078 (CelB) and TM0070 (XynB)-Csac_1078 (CelB). However, the Thermotoga transformants lost their recombinant genes after three consecutive transfers. This study represents an important milestone in the effort of using Thermotoga to produce biohydrogen directly from cellulosic biomass. Future studies should be focused on improving the stability of the transformants.
Conflict of Interests
The authors declare that they have no competing interests.
This work was supported by the BGSU Commercialization Catalyst Award, the Building Strength Award, and the Katzner Award.
K. Bronnenmeier, A. Kern, W. Liebl, and W. L. Staudenbauer, “Purification of Thermotoga maritima enzymes for the degradation of cellulosic materials,” Applied and Environmental Microbiology, vol. 61, no. 4, pp. 1399–1407, 1995.View at: Google Scholar
L. D. Ruttersmith and R. M. Daniel, “Thermostable cellobiohydrolase from the thermophilic eubacterium Thermotoga sp. strain FjSS3-B.1: purification and properties,” Biochemical Journal, vol. 277, no. 3, pp. 887–890, 1991.View at: Google Scholar
M. Balk, J. Weijma, and A. J. M. Stams, “Thermotoga lettingae sp. nov., a novel thermophilic, methanol-degrading bacterium isolated from a thermophilic anaerobic reactor,” International Journal of Systematic and Evolutionary Microbiology, vol. 52, no. 4, pp. 1361–1368, 2002.View at: Publisher Site | Google Scholar
S. Belkin, C. O. Wirsen, and H. W. Jannasch, “A new sulfur-reducing, extremely thermophilic eubacterium from a submarine thermal vent,” Applied and Environmental Microbiology, vol. 51, no. 6, pp. 1180–1185, 1986.View at: Google Scholar
J.-D. Bok, D. A. Yernool, and D. E. Eveleigh, “Purification, characterization, and molecular analysis of thermostable cellulases CelA and CelB from Thermotoga neapolitana,” Applied and Environmental Microbiology, vol. 64, no. 12, pp. 4774–4781, 1998.View at: Google Scholar
F. A. Rainey, A. M. Donnison, P. H. Janssen et al., “Description of Caldicellulosiruptor saccharolyticus gen. nov., sp. nov: an obligately anaerobic, extremely thermophilic, cellulolytic bacterium,” FEMS Microbiology Letters, vol. 120, no. 3, pp. 263–266, 1994.View at: Publisher Site | Google Scholar
V. Zverlov, S. Mahr, K. Riedel, and K. Bronnenmeier, “Properties and gene structure of a bifunctional cellulolytic enzyme (CelA) from the extreme thermophile ‘Anaerocellum thermophilum’ with separate glycosyl hydrolase family 9 and 48 catalytic domains,” Microbiology, vol. 144, no. 2, pp. 457–465, 1998.View at: Publisher Site | Google Scholar
D. J. Saul, L. C. Williams, R. A. Grayling, L. W. Chamley, D. R. Love, and P. L. Bergquist, “celB, a gene coding for a bifunctional cellulase from the extreme thermophile ‘Caldocellum saccharolyticum’,” Applied and Environmental Microbiology, vol. 56, no. 10, pp. 3117–3124, 1990.View at: Google Scholar
A. L. VanFossen, I. Ozdemir, S. L. Zelin, and R. M. Kelly, “Glycoside hydrolase inventory drives plant polysaccharide deconstruction by the extremely thermophilic bacterium Caldicellulosiruptor saccharolyticus,” Biotechnology and Bioengineering, vol. 108, no. 7, pp. 1559–1569, 2011.View at: Publisher Site | Google Scholar
W. Liebl, I. Stemplinger, and P. Ruile, “Properties and gene structure of the Thermotoga maritimaα-amylase amyA, a putative lipoprotein of a hyperthermophilic bacterium,” Journal of Bacteriology, vol. 179, no. 3, pp. 941–948, 1997.View at: Google Scholar
S. G. N. Grant, J. Jessee, F. R. Bloom, and D. Hanahan, “Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants,” Proceedings of the National Academy of Sciences of the United States of America, vol. 87, no. 12, pp. 4645–4649, 1990.View at: Publisher Site | Google Scholar
R. M. Teather and P. J. Wood, “Use of Congo red-polysaccharide interactions in enumeration and characterization of cellulolytic bacteria from the bovine rumen,” Applied and Environmental Microbiology, vol. 43, no. 4, pp. 777–780, 1982.View at: Google Scholar
K. A. Laderman, B. R. Davis, H. C. Krutzsch et al., “The purification and characterization of an extremely thermostable α- amylase from the hyperthermophilic archaebacterium Pyrococcus furiosus,” Journal of Biological Chemistry, vol. 268, no. 32, pp. 24394–24401, 1993.View at: Google Scholar
C. Winterhalter and W. Liebl, “Two extremely thermostable xylanases of the hyperthermophilic bacterium Thermotoga maritima MSB8,” Applied and Environmental Microbiology, vol. 61, no. 5, pp. 1810–1815, 1995.View at: Google Scholar
O. T. Harriott, R. Huber, K. O. Stetter, P. W. Betts, and K. M. Noll, “A cryptic miniplasmid from the hyperthermophilic bacterium Thermotoga sp. strain RQ7,” Journal of Bacteriology, vol. 176, no. 9, pp. 2759–2762, 1994.View at: Google Scholar
J.-S. Yu and K. M. Noll, “Plasmid pRQ7 from the hyperthermophilic bacterium Thermotoga species strain RQ7 replicates by the rolling-circle mechanism,” Journal of Bacteriology, vol. 179, no. 22, pp. 7161–7164, 1997.View at: Google Scholar