Harnessing the potential of chloroplast-derived expression elements for enhanced production of cellulases in Escherichia coli

Bacterial strains used and provided cell culturing conditions

Escherichia coli strains namely Top10 and/or DH5α were chosen for the purpose of cloning our target gene and construction of the vector plasmid owing to its very high and well reported transformation efficiency. The strains were propagated in the widely used Luria Bertani (LB) broth/agar (1% (w/v) tryptone, 0.5% (w/v) yeast extract, 1% (w/v) NaCl with addition of 1.5% (w/v) agar) for solid media at 37 °C. The E. coli strain BL-21 (DE3) was used for expression studies.

Vector construction, transformation and identification of transformants

The nucleotide sequences encoding T. maritima β-glucosidase (EC 3.2.1.21; GenBank accession number: X74163.1) and endoglucanase (EC 3.2.1.4; GenBank accession number: CP004077.1) were retrieved from the NCBI GenBank database. Since the target enzymes are intended for expression in higher plant chloroplasts, to enhance their expression, codon optimization was performed to match the codon usage preferences of the chloroplast genome. This optimization was carried out using the Kazusa Codon Usage Database (https://www.kazusa.or.jp/codon/), ensuring improved translational efficiency in the target expression system.

For the construction of expression cassettes, we used the Golden Gate cloning strategy, with pKP9 serving as the Level 1 vector (Zhou et al., 2008). β-glucosidase and endoglucanase genes were cloned into this vector to create the Level 2 constructs. To facilitate their use in the Golden Gate system, both the target genes and vector backbone were modified to incorporate PaqCI restriction sites using a set of primers (Table S1). This system utilizes Type IIS restriction enzymes, which generate non-palindromic overhangs, allowing for directional and order-specific assembly of DNA fragments. This approach significantly minimizes the risk of insert misorientation and enhances cloning accuracy (Engler, Kandzia & Marillonnet, 2008). The cloning vector, pKP9, was provided by Prof. Ralph Bock at the Max Planck Institute of Molecular Plant Physiology, Potsdam, Germany (Zhou et al., 2008). Briefly, the coding regions were linked to a strong constitutive 16S ribosomal RNA promoter, Prrn. A six-histidine tag (His₆-tag) was added to N-terminus of expressed sequences for protein detection and quantification. A terminator of rps16 (Trps16) was used to halt the ongoing transcription via polymerase and stabilize the encrypted strands (Fig. 1 for details).

The Golden Gate reaction was set up in a total volume of 10 µl, consisting of 3 nM of each insert, 1 nM of vector DNA, 1 µl of 10× ligase buffer, 0.5 µl of T4 DNA ligase, 0.5 µl of Type IIS restriction enzyme PaqCI, and 7.15 µl of dH₂O. The Golden Gate assembly protocol was as follows: an initial incubation at 37 °C for 15 min, followed by 30 cycles of 37 °C for 5 min and 16 °C for 5 min. This was then followed by a final incubation at 37 °C for 15 min and 65 °C for 20 min. The E. coli strain DH5α was transformed with the ligation mixture and plated on LB agar. Bacterial transformants were selected on 100 µg/ml ampicillin/carbenicillin and screened by colony polymerase chain reaction (PCR) using Prrn-F (5′-AGCTCGGTACCCCAAAGCTCCCCCG-3′) as forward primer and HT494 (5′-TACATTTGTACGGCTCCG-3′) as a reverse primer. The resulting expression vectors pKP9-BETA and pKP9-ENDO were confirmed through enzyme digestion, PCR, and sequencing. PCR amplification was conducted using specific primer sets to confirm each construct. For pKP9-BETA, confirmation was achieved using the primers BETA-For (5′-CAGGAAGGGTGAATCAGAAA-3′) and BETA-Rev (5′-TTGAGTATTCTGCGAACCAA-3′). Similarly, pKP9-ENDO was confirmed with the primers ENDO-For (5′-GACTGAGGCATCTATTGGAG-3′) and ENDO-Rev (5′-CGAGCCATAATTCCCATGTA-3′). DNA integrity and successful modifications were verified on 1% (w/v) agarose gel electrophoresis, stained with ethidium bromide/SYBR safe, and visualized under UV light.

Protein extraction, SDS-PAGE and western blot analysis

Both the transformed and untransformed cells were cultured in 10 ml of LB broth and incubated at 37 °C for 16–18 h until the OD₆₀₀ reached 0.4–0.6. Subsequently, 3 ml of the bacterial culture was centrifuged at 14,000 rpm at 4 °C for 2 min. Following centrifugation, the separated supernatant was saved to analyse extracellular proteins, and the cell pellet was resuspended in 150 μl of 150 mM Tris-HCl buffer (pH 7.5), with the addition of 1.5 μl of 100 mM PMSF and 15 μl of 10 mg/mL lysozyme. The cell pellet was thoroughly mixed by vortexing. The microfuge tube containing the cellular lysate was then incubated on ice for 30 min, gently thawed and mixed by vortexing every 5 min. After 30 min, the lysate was centrifuged at 10,000 rpm, 4 °C for 10 min, resulting in the separation of two protein fractions. The supernatant, containing soluble proteins, was transferred to a new sterile microfuge tube and kept at 4 °C. The pellet, containing insoluble proteins, was resuspended in 1.5 ml of 150 mM Tris-HCl buffer (pH 7.5) by vortexing and stored at 4 °C. Protein quantification was carried out using Bradford assay (Bradford, 1976).

To visualise the expression of targeted enzymes in E. coli cells, extracted proteins were separated by SDS-PAGE using 12% (w/v) gels, and visualised by staining with Coomassie blue. After the proteins were separated on gel, they were transferred (or “blotted”) onto a nitrocellulose membrane and incubated at 4 °C overnight before incubation with secondary antibody. The membrane was then probed with anti-His antibodies at 1:3,000 dilution to detect the target protein. The secondary antibody used was anti-mouse IgG alkaline phosphatase conjugate (Sigma, St. Louis, MO, USA) at a dilution of 1:10,000. Alkaline phosphatase substrate (Thermo Fisher Scientific, Waltham, MA, USA, 1 Step^TM NBT/BCIP) was used for visualizing the bands.

Enzyme activity assays

Thermotolerance assays

To assess the thermostability of β-glucosidase and endoglucanase enzymes, the isolated enzyme samples were divided into microcentrifuge tubes and subjected to their respective determined optimum temperatures using a thermal cycler or incubator. Incubation was carried out every 5 min interval for up to 60 min to monitor the enzyme activity. Enzyme activity assays were conducted as described in previous section (See pNPG assay for β-glucosidase quantification and Enzyme assay for endoglucanase quantification). The enzyme activity was determined using a spectrophotometer, and the data was analysed to generate a First Order Plot by plotting enzyme activity against time for both enzymes.

Kinetics of substrate hydrolysis

The Michaelis-Menten kinetic constants (V_max and K_m) for β-glucosidase and endoglucanase were determined by assaying the enzymes on varied concentrations of substrates (pNPG & CMC) as described earlier. The kinetic constants for substrate hydrolysis were determined by direct fit method (non-linear regression). Briefly, kinetic constants of β-glucosidase for pNPG hydrolysis at 90 °C, pH 4.8 were determined by incubating the fixed amount of β-glucosidase extract (27.85 μg) with varied concentrations of pNPG ranging from 0.02 to 1.0 mM as described (Javed et al., 2009). Stock solution of pNPG (5.0 mM) was prepared in 50 mM sodium acetate buffer, pH 4.8 and required concentrations of pNPG were prepared by diluting the stock with the buffer.

The endoglucanase was assayed on varied CMC concentrations, ranging from 0.05–3.0% (w/v), which were prepared by using CMC stock solution (5% w/v). The endoglucanase activity was performed at 90 °C, pH 4.0 and fixed amount of endoglucanase extract (22.53 μg) was used per assay. The required concentration was achieved by using the following equation:

$${\mathrm{M}}_1{\mathrm{V}}_1={\mathrm{M}}_2{\mathrm{V}}_2$$where, M₁ = stock concentration, M₂ = required concentration, V₁ = volume of stock solution, V₂ = volume of solution to be made.

Statistical analyses

The data obtained were subjected to various statistical analyses utilizing R Environment 4.1.3 for Windows. The data for enzyme activity were tested for significance using one–way analysis of variance (ANOVA) keeping α = 0.05 followed by Tukey’s Honest Significant Difference (HSD) post-hoc test for pairwise comparison of means between groups.

Structural analysis of enzymes for thermostability

Three–dimensional (3D) structures of endoglucanase (PDB codes: 3AMH and 3AMM, Cheng et al., 2011) and β-glucosidase (PDB codes: 1UZ1, Vincent et al., 2004; 2CBU, Gloster, Madsen & Davies, 2006; 2WBG, Aguilar-Moncayo et al., 2009) from T. maritima were accessed from the Protein Data Bank (PDB) (Berman et al., 2000). Structural features relevant to thermostability of the afore-mentioned enzymes were compared with that found in homologous structures from thermophilic and mesophilic organisms. PyMOL (The PyMOL Molecular Graphics System, Version 1.3, Schrödinger, LLC) was used for visualizing 3D structures.

Confirmation of expression vectors based on chloroplast-derived expression elements

To express β-glucosidase and endoglucanase in E. coli from the chloroplast gene expression elements (promoter, untranslated regions (UTRs) and terminator), the tobacco chloroplast expression vector pKP9 was used. This vector contains the strongest known promoter in the chloroplast genome, 16S rRNA Promoter-Prrn and a Trps16 terminator. The coding sequences of β-glucosidase and endoglucanase were cloned in pKP9 as PaqCI fragments. In addition, a hexa-histidine-tag (His₆-tag) was N-terminally fused for identification through immunoblotting (Fig. 1). The constructed plasmids pKP9-ENDO and pKP9-BETA were restricted using SpeI enzyme, resulting in the release of expected fragment sizes of 1.9 and 1.3 kbp, respectively (Fig. 1). In addition to restriction digestion, the constructs were also validated through PCR using gene specific primers. The amplification of desired length of fragments confirmed the correct integration of the transgenes at the desired position (Fig. 2).

Enhanced protein expression in E. coli

The recombinant vectors after being constructed and confirmed to contain the desired sequences were then transformed in commercially available E. coli strain BL21(DE3). Upon reaching an OD₆₀₀ between 0.4 and 0.6, the soluble and insoluble proteins were isolated both from BL21(DE3)-BETA (BL21(DE3) cells expressing β-glucosidase), BL21(DE3)-ENDO (BL21(DE3) cells expressing endoglucanase) and untransformed cells and separated through SDS–PAGE. Upon staining with Coomassie Brilliant Blue, an additional fragment equivalent to the theoretical size of β-glucosidase (~52 kDa) and endoglucanase (~30 kDa) were observed in the soluble fraction of transformed cells while no additional band could be detected in any fraction of the untransformed cells (Fig. 3A). To confirm the identity of the additional fragments observed, the resolved proteins on the gel were then carefully transferred to a nylon-based membrane sheet and probed with anti-His antibodies. This immunoblotting confirmed the accumulating proteins were β-glucosidase and endoglucanase. The detection of enzyme in the insoluble fraction as a weaker band suggests that the proteins were predominately accumulating in the soluble fractions (Fig. 3B).

Physiological activity of $\text{β}$-glucosidase

After confirmation of accumulation of recombinant enzymes in bacteria, we then determined whether the expressed proteins are physiologically active as well. The activity of β-glucosidase was measured due to its sensitivity and specificity. The pNPG assay is a widely accepted method for quantifying β-glucosidase activity because it provides a clear and quantifiable colorimetric readout that directly correlates with enzyme activity. Upon hydrolysis of pNPG by β-glucosidase, p-nitrophenol is released, producing a yellow colour measurable at 410 nM. This method allows for precise kinetic analyses and is highly effective in assessing the enzyme’s activity in various conditions (Rajoka & Malik, 1997). The pNPG assay revealed that the enzyme was active in both the soluble and insoluble fractions of intracellular extracts, with no enzymatic activity detected in control cells. This enzymatic activity was evidenced by a colorimetric change, indicative of the enzyme’s capacity to hydrolyze p-nitrophenyl-β-D-glucopyranoside (Fig. S1). Quantitative assay results manifested a significant increase in absorbance at 410 nM, consistent with the enzymatic release of p-nitrophenol. These findings demonstrate the enzyme’s functionality across various intracellular compartments (Strahsburger et al., 2017).

Absence of endoglucanase secretion into the extracellular medium

For endoglucanase, we utilized the carboxymethyl cellulose (CMC) plate assay, which is particularly suitable for detecting and visualizing endoglucanase activity on a solid substrate. The CMC plate assay involves the application of the enzyme onto a plate containing CMC embedded in an agarose gel. Upon incubation, areas where endoglucanase has hydrolyzed the CMC become clear against a dark background when stained with Congo red. This assay is chosen for its ability to provide a spatial visualization of enzyme activity, which is crucial for comparative analyses of enzyme action across different samples. Additionally, it is a straightforward and cost-effective method to assess the activity of endoglucanases, particularly suitable for preliminary screenings in our experimental design.

The clear zones formed due to hydrolysis around the wells on the agar plates showed that expressed endoglucanase is in an active conformation (Fig. 4). The absence of halozones in extracellular extract suggests that the enzyme was not secreted in the medium. Both the soluble and insoluble fractions of intracellular extract showed visible activity by forming halozones. The presence of these clear zones indicates that our enzyme produced is active and can actively degrade CMC substrate, leading to the formation of soluble sugars and a visible hydrolysis pattern.

Optima temperature requirement of recombinant enzymes

Optimum temperatures of recombinant β-glucosidase were determined by testing enzymatic activity from 50 °C to 110 °C using the pNPG assay. The enzyme showed optimal activity at 90 °C (Figs. 5A, 5B). Earlier studies on the heterologous expression of T. maritima β-glucosidase in E. coli reported an optimal activity at pH of 5.0–7.0 and at temperatures from 80–100 °C (Mehmood et al., 2014).

Likewise, the degree of thermotolerance for endoglucanase was determined using a DNS assay. The enzyme showed optimal activity at 100 °C whereas further increase in temperature showed slight change in activity (Figs. 5C, 5D). The optimal temperature requirement for recombinant endoglucanase was in line with earlier results (Bok, Yernool & Eveleigh, 1998).

Thermostability of recombinant enzymes

To examine thermal stability, the enzymes were preheated at their respective optimum temperature for 1 h at an interval of 5 min. Both the enzymes were found to be stable after heating and showed significant activity even after 60 min incubation (Fig. 6). In the analysis of the enzyme-catalysed hydrolysis of pNPG/CMC using the pseudo-first-order kinetics model, a plot of the natural logarithm of the enzyme activity against time (t) was generated. This approach allows for the quantification of enzymatic reaction rates and determining rate constant (k) governing the reaction kinetics.

Both enzymes showed excellent thermal stability after treatment at high temperature above 80 °C. The graph showed that the enzymes β-glucosidase and endoglucanase retained 85% (P = 2.2 × 10⁻⁴) and 77% (P = 6.4 × 10⁻⁶) of activity, exhibiting a half-life of 123 and 167 min, respectively.

Enzyme kinetics of recombinant $\text{β}$-glucosidase and endoglucanase

The kinetics of pNPG hydrolysis by β-glucosidase at 90 °C, pH 4.8 were determined by applying the direct fit model (Fig. 7A). The specific activity of intra cellular β-glucosidase for pNPG hydrolysis by T. maritima gene expressed in E. coli was 283.2 U mg⁻¹(Table 1). The Michaelis Menten constant (K_m) for pNPG was 0.136 mM. The expression protocol used in our study was very efficient as compared to the previous report, which deals with the heterologous expression of β-glucosidase gene (bglA) from T. maritima in E. coli M15 (Mehmood et al., 2014). However, in their report the kinetic constants for pNPG hydrolysis were determined at 80 °C. We found that the affinity of pNPG to the active site (K_m) was about 4-fold higher, which confirmed that the pNPG was highly specific to our β-glucosidase as compared to the bglA gene expressed in E. coli M15.

The β-glucosidase of Thermoanaerobacterium thermosaccharolyticum DSM 571 worked optimally at 60 °C. The specific activity of the purified enzyme and the kinetic constants K_m and k_cat for pNPG hydrolysis were 64 U mg⁻¹, 0.64 mM and 67.7 s⁻¹, respectively, which confirmed that the enzyme has higher catalytic efficiency, and pNPG has higher affinity and specificity to the active site (Pei et al., 2012). Furthermore, Amouri & Gargouri (2006) reported kinetics of pNPG hydrolysis by β-glucosidase of Stachybotrys sp. at 50 °C. The specific activity was 78 U mg⁻¹ and K_m 0.2 mM, which highlighted that our β-glucosidase was kinetically highly efficient.

Kinetics of CMC hydrolysis by endoglucanase was also determined by applying the direct fit method (Fig. 7B). The specific activity of intracellular endoglucanase for CMC hydrolysis by T. maritima gene expressed in E. coli at 90 °C, pH 4.0 was 16.10 U mg⁻¹. The Michaelis Menten constant (K_m) for CMC was 9.64 mg ml⁻¹.

Structural basis for thermostability of $\text{β}$-glucosidase from T. maritima

Like thermophilic endoglucanases, structural stability of β-glucosidase from T. maritima could also be contributed by ionic interactions. More ionic interactions are found in the thermophilic β-glucosidases than in the mesophilic β-glucosidases. Some of these ionic interactions are common, located at corresponding positions in the structure, in both the thermophilic β-glucosidases from T. maritima (PDB codes: 1UZ1, Vincent et al., 2004; 2CBU, Gloster, Madsen & Davies, 2006; 2WBG, Aguilar-Moncayo et al., 2009) and T. neapolitana (PDB code: 5IDI; Kulkarni et al., 2017) and mesophilic β-glucosidases from Lactococcus lactis (PDB code: 1PBG, Wiesmann et al., 1995), and Arabidopsis thaliana (PDB code: 7F3A, Horikoshi et al., 2021). Importantly, there are ionic interactions that are only observed in the case of β-glucosidases from thermophilic bacteria but are missing in β-glucosidases from mesophiles (Fig. 8, Table S2). Thus, higher thermostability of β-glucosidase from T. maritima could be due to the presence of ‘additional’ ionic interactions not found in mesophilic β-glucosidases.

Structural basis for thermostability of endoglucanase from T. maritima

The structural analysis of endoglucanase from T. maritima (PDB codes: 3AMH and 3AMM; Cheng et al., 2011) revealed that the protein predominantly (>59%) consists of beta (β)-sheets. Notably, a high content (>55%) of β-sheets is also found in endoglucanases from other organisms Thermococcus sp. 2319x1 (PDB code: 7S8K), Pyrococcus furiosus (PDB code: 3VGI, Kim, Kataoka & Ishikawa, 2012), Streptomyces lividans (PDB code: 2NLR, Sulzenbacher et al., 1999) and Streptomyces sp. 11AG8 (PDB code: 1OA4, Sandgren et al., 2003) as given in Table S3. In endoglucanases, β-sheets are arranged as convex outer β-sheets (Sheet A) and concave inner β-sheets (Sheet B). It has been suggested that outer β-sheets (Sheet A) are important for stability as they play a structural role, whereas the inner β-sheets (Sheet B) harbour the active site (Cheng et al., 2011). Importantly, the number of residues involved in the formation of Sheet A are more in thermophilic endoglucanases as compared to that in mesophilic endoglucanases (Fig. S2, Table S4). Longer β-sheets (Sheet A) in thermophilic endoglucanases may stabilize the structure and could contribute to thermal stability of the protein. Furthermore, the number of ionic interactions found in thermophilic endoglucanases is also more than that in mesophilic endoglucanases. Importantly, some ionic interactions are only found in thermophilic endoglucanases and are absent at structurally equivalent positions in mesophilic endoglucanases. Thus, higher thermostability of endoglucanase from T. maritima could also be due to the presence of ionic interactions not found in mesophilic endoglucanases (Fig. 9, Table S5).

Conclusion and future perspectives

In conclusion, this research achieved successful heterologous expression of β-glucosidase and endoglucanase enzymes, native to the hyper-thermophilic bacterium Thermotoga maritima, in the industrially favoured E. coli BL21 using chloroplast genetic machinery elements. By employing a combination of a constitutive 16S rRNA promoter, rps16 terminator (Trps16), and 5′ UTRs, we attained high expression levels for endoglucanase and β-glucosidase. Notably, the recombinant enzymes maintained their high thermotolerance, indicating the potential for producing thermotolerant cellulolytic enzymes at high levels. This high-level expression of thermotolerant cellulases holds promise for making the bioconversion of lignocellulosic biomass commercially feasible.

While this study successfully focused on heterologous expression in E. coli, these findings provide a foundation for future applications in plant-based systems. The E. coli expression system serves as a critical preliminary step in demonstrating the feasibility of utilizing chloroplast genetic elements to achieve high-level expression of thermostable enzymes. Building upon these results, future research can explore the extension of this approach to plant systems, where plastid transformation has already shown significant potential in enhancing transgene expression, reducing enzyme production costs, and enabling large-scale production. Notably, a parallel study is currently being conducted to apply this strategy to plant systems, and the initial results are promising. However, this research is ongoing, and the corresponding manuscript has yet to be submitted for publication.

Future research will focus on optimizing expression levels, maximizing enzyme activity under varying conditions, and exploring diverse applications of these enzymes in biotechnology and industrial settings. This study represents a significant step toward harnessing the potential of thermophilic enzymes for biotechnological applications. By leveraging the unique capabilities of plastid genetic elements, this study would lead to unlock the potential for cost-effective and sustainable production of various products, contributing to the advancement in developing sustainable biomanufacturing platforms to produce specially biomolecules for diverse industrial applications.