Encyclopedia
Please note this is a comparison between Version 2 by Dean Liu and Version 1 by Goh Kian Mau.
Starch and pullulan degrading enzymes are essential industrial biocatalysts. Pullulan-degrading enzymes are grouped into pullulanases (types I and type II) and pullulan hydrolase (types I, II and III).
- amylopullulanase
- alpha-amylase-pullulanase
- carbohydrate-active enzyme
1. Classification and Action of Pullulan-Degrading Enzymes
Pullulan-degrading enzymes are classified into two major groups: pullulanases (type I and type II) and pullulan hydrolases (types I, II and III) (Figure 1). Type I pullulanase (EC 3.2.1.41) acts on α-1,6 glucosidic bonds in pullulan to produce maltotriose (Figure 1a). The enzyme is also active against short branches (G1–G7) of starch, amylopectin, glycogen and β-limit dextrin, producing reducing sugars (i.e., glucose, maltose and maltotriose). Moreover, this enzyme group does not hydrolyse α-1,4 glucosidic bonds in the polysaccharides mentioned earlier [16], except the enzyme from Bacillus sp. AN-7 and Lactococcus lactis IBB 500 that possess the additional ability to hydrolyse α-1,4 glucosidic bonds in pullulan and form maltose [43,44]. Type I pullulanase is widely distributed in bacteria such as psychrophiles (i.e., Paenibacillus and Shewanella spp.), mesophiles (i.e., Bacillus, Exiguobacterium and Klebsiella spp.), thermophiles (i.e., Anoxybacillus, Geobacillus and Thermus spp.) and hyperthermophiles (i.e., Fervidobacterium and Thermotoga spp.) [45,46]. The optimum temperatures of type I pullulanases varied. For instance, the type I pullulanases from psychrophiles (i.e., Paenibacillus and Shewanella spp.) and mesophiles (i.e., Bacillus and Exiguobacterium spp.) are active at low temperatures of 35–50 °C. In contrast, those from thermophiles (i.e., Anoxybacillus, Geobacillus and Thermus spp.) and hyperthermophiles (i.e., Fervidobacterium and Thermotoga spp.) exhibit a higher optimum temperature in the range of 60–90 °C. From the industrial perspective, thermostable type I pullulanases are more feasible than heat-sensitive ones.
Figure 1. The action pattern of (a) pullulanases (types I and II) on pullulan, starch or other polysaccharides and (b) pullulan hydrolases (types I, II and III) on pullulan.
Type II pullulanase (EC 3.2.1.1/41) can be divided into two subgroups: (i) amylopullulanase and (ii) α-amylase-pullulanase. These enzymes hydrolyse α-1,6 glucosidic bonds in pullulan. In addition, type II pullulanase can act on both α-1,4 and α-1,6 glucosidic bonds in starch and other polysaccharides [21]. Amylopullulanases from the following genera have been reported: thermophilic bacteria (i.e., Alkalilimnicola, Anoxybacillus, Geobacillus, Thermoanaerobacter, Thermoanaerobacterium and Thermus spp.) and hyperthermophilic archaea (i.e., Caldivirga, Desulfurococcus, Pyrococcus, Staphylothermus, Sulfolobus, Thermococcus and Thermofilum spp.)[47,48,49]. Other than that, a small number of currently known enzymes originated from mesophilic bacteria (i.e., Bacillus, Cohnella, Lactobacillus and Streptomyces spp.) and halo-mesophilic archaeon (i.e., Halorubrum sp.) [50,51]. The molecular mass of amylopullulanases falls in 45–184 kDa. Some amylopullulanases exhibit high molecular mass (>200 kDa), such as those from Anoxybacillus sp. SK3-4 (225 kDa) that ouresearcher's group studied earlier [48]. Other examples are enzymes from Bacillus circulans F-2 (220 kDa), Bacillus sp. XAL601 (224 kDa), Geobacillus stearothermophilus TS-23 (220 kDa) and Thermoanaerobacterium thermosulfurigenes EM1 (205 kDa). Most of the currently known amylopullulanases are active at high temperatures (60–105 °C) and acidic (pH 3.0–6.5).
Another subgroup of type II pullulanase is α-amylase-pullulanase. At present, only six of these enzymes have been characterised. They all originated from mesophilic bacteria (i.e., Bacillus sp. KSM-1378, Bifidobacterium breve UCC2003, Bifidobacterium adolescentis P2P3, Lactobacillus plantarum L137, Streptococcus suis P1/7 and Alkalibacterium sp. SL3 [18,52,53,54,55]. This subgroup of enzymes exhibited a high molecular mass of approximately 200 kDa, and is active at 40–50 °C. The action of α-amylase-pullulanases on polysaccharides (i.e., pullulan, starch, amylopectin, amylose and glycogen) generate reducing sugars of various lengths (G1–G6).
Pullulan hydrolases are further classified into three groups (Figure 1). Type I (neopullulanase, EC 3.2.1.135) hydrolyses α-1,4 glucosidic bonds of pullulan to form panose. In contrast, pullulan hydrolase type II (isopullulanase, EC 3.2.1.57) cleaves α-1,4 glucosidic bonds in pullulan and produces isopanose [23]. Pullulan hydrolase type III attacks α-1,4 and α-1,6 glucosidic bonds in pullulan and forms panose, maltotriose, maltose and glucose [30]. Furthermore, pullulan hydrolases can degrade starch, liberating reducing sugars such as glucose, maltose and maltotriose [21]. Researchers have described neopullulanases from Paenibacillus, Bacillus, Bacteroides, Lactobacillus, Alicyclobacillus, Anoxybacillus and Geobacillus spp. The biochemical properties for isopullulanases from Bacillus sp. US149 and Aspergillus brasiliensis ATCC 9642T were reported in these articles [38,39,40,41,56]. Apart from that, only two pullulan hydrolases type III from Thermococcus spp. have been characterised [30,32].
Almost all pullulan hydrolases are recombinant enzymes, and exhibit a molecular mass of 53–200 kDa. The optimum temperature of pullulan hydrolases falls in the range of 37–70 °C, except for pullulan hydrolase type III, which they are active at 95 °C. These pullulan hydrolases favour acidic condition of pH 3.5–6.0. However, the neopullulanases from Bacillus sp. KSM-1876 and Micrococcus halobius OR-1 function at a slightly alkaline pH of 7.5 and 8.0, respectively [20,57].
Carbohydrate-Active enZYmes (CAZy) database classified pullulan-degrading enzymes in glycoside hydrolase (GH) families GH13, GH49 or GH57 [58]. Type I pullulanase, type II pullulanase, neopullulanase and pullulan hydrolase type III are members of GH13. Members within this family have consensus key features such as (i) they cleave the α-glucosidic bonds, (ii) the amino acid sequences contain four conserved regions, (iii) the enzymes possess a TIM barrel structure and (iv) Asp, Glu and Asp are the enzymes catalytic residues [59]. Among the type II pullulanases, some of the characterised enzymes from hyperthermophilic archaea (i.e., Caldivirga, Pyrococcus, Thermococcus, Sulfolobus and Staphylothermus spp.) have been classified in family GH57 [34]. Isopullulanases are categorised in family GH49 [11].
2. Domains, Structures and Properties of Pullulan-Degrading Enzymes
Pullulan-degrading enzymes are multi-domain proteins. Generally, these enzymes contain several domains such as carbohydrate-binding module (CBM), catalytic domain, C-terminal domain (i.e., domain C or amyC domain) and fibronectin type III (FnIII) domain [11]. CBM, a non-catalytic ancillary domain, assists protein attachment to polysaccharide’s surface (i.e., pullulan or starch). CBM also facilitates the degradation process by distorting the conformation or the packing of the polysaccharides. Based on the CAZy database (November 2021), CBMs are currently grouped into 88 primary sequence-based families [58]. CBMs with affinity for starch (CBM20, CBM21, CBM25, CBM26, CBM34, CBM41, CBM45, CBM48, CBM53, CBM58, CBM68, CBM69, CBM82 and CBM83) are commonly known as starch-binding domains (SBDs) [60]. Figure 2 shows the comparison of domains and structure architecture for each group of pullulan-degrading enzymes.
Figure 2.
Schematic representation of conserved domains in pullulan-degrading enzymes.
Initially, all type II pullulanases were named amylopullulanase. However, after more sequences were available, the differences between amylopullulanase and α-amylase-pullulanase were noticed. Now, amylopullulanase refers to a group of enzymes with a single catalytic domain (domain A). In contrast, α-amylase-pullulanase referred to enzymes with dual catalytic domains: (i) an α-amylase catalytic domain for α-1,4 bonds and (ii) type I pullulanase catalytic domain for α-1,6 glucosidic bonds (Figure 2) [11,18,52,53,54,55]. A recent article reported the role of CBM and the dual catalytic domains [61]. α-amylase-pullulanase (denoted as PulP) from Bifidobacterium adolescentis P2P3 underwent truncation study. Deletion of either catalytic domain affected the catalytic activity and thermostability. The removal of CBM41 and CBM25, located somewhere in the middle of the protein sequence, led to the loss of binding affinity and hydrolytic capability on raw and soluble starches. In a separate recent publication, the author preferred to name it pullulanase PulSL3 instead of α-amylase-pullulanase PulSL3 [55]. PulSL3 originated from Alkalibacterium sp. SL3 contains the typical dual catalytic domains explained earlier. Interestingly, the recombinant enzyme exhibited pullulanase activity, but lacked α-amylase activity. Huang et al. [55] created domain-truncated mutants and examined the function of each domain. For some reason, probably due to certain unfavourable residues, the N-terminal α-amylase domain is missing the α-1,4 hydrolytic ability.
As illustrated in Figure 2, pullulan hydrolases exhibit a less complicated architecture. Neopullulanase (pullulan hydrolase type I) and pullulan hydrolase type III have three domains (Domains N, A and C) [39]. Domains N and A function as starch-binding domain (CBM34) and catalytic domain, respectively. The C-terminal of these enzymes is composed of the amyC domain that orientates the substrate’s glucose chains so the hydrolytic action can take place [62]. Isopullulanase (pullulan hydrolase type II) possess only two domains, where domains N and C act as CBM34 and catalytic domain, respectively [39].
3. Single Immobilisation of Pullulan-Degrading Enzyme
In laboratories and industrial setups, enzymes immobilisation can increase enzyme activity or improve other catalytic features such as stability and specificity (selectivity); facilitate the recovery or the reusability of biocatalysts and enhance biochemical properties [63,64,65]. The methods for protein immobilisation are divided into three main categories: (i) the Entrapment (encapsulation) approach, a method based on the inclusion of the enzyme in a polymer network such as calcium alginate, polyacrylamide, silica, hollow fibre membrane or microcapsule [63], which often requires synthesis of the polymeric matrix together with the enzyme [66]; (ii) second, the cross-linked immobilisation strategy, which often employs linker reagents (i.e., glutaraldehyde) to produce carrier-free immobilised enzymes particles denoted as cross-linking enzymes crystals (CLECs) and cross-linking enzymes aggregates (CLEAs) [67,68]; and (iii) the binding approach, where protein is immobilised to the pre-fabricated solid support (carrier) via physical adsorption, ionic bonding or covalent attachment [69].
Some conventional immobilisation techniques, such as calcium alginate, could be an excellent example for students’ training; however, it is not feasible on an industrial scale. Choosing the ideal support for the enzyme is essential for a successful immobilisation process. One shall carefully examine the properties of the carriers; among the many, these are the critical selection criteria: particle size, pore size, surface area and hydrophilicity/hydrophobicity of the particle surface. Readers that would like to learn more on these topics may refer to several excellent review articles [63,68,70,71,72,73,74,75]. To date, many ready-to-use polymer supports are available at an affordable price. These include the epoxide or amino-epoxide-activated supports such as Sepabeads (Resindion S.R.L.), ReliZyme (Resindion S.R.L.), Immobead (ChiralVision) and Purolite (Purolite). Acrylic-based macroporous supports offer high mechanical stability, high resistance towards microbial contamination and low swelling effects in water, making them suitable for laboratory-, pilot- or industrial applications [76,77]. Furthermore, the material used in producing such supports is generally safe in food processing.
Unlike many immobilisation research on GH13 enzymes (i.e., α-amylase and α-glucosidase), studies on immobilisation of pullulan-degrading enzymes are rather scarce. The collective findings on single immobilised pullulan-degrading enzymes. Most of the immobilised enzymes were type I pullulanases that originated from several mesophilic bacteria such as Bacillus, Klebsiella and Fontibacillus spp. A small number of immobilisation studies used other types of pullulan-degrading enzymes. Examples include type II pullulanase (amylopullulanase) and neopullulanase from Anoxybacillus sp. WB42 and Geobacillus stearothermophilus TRS40, respectively [78,79].
Covalent attachment on a solid carrier is a favoured approach. Lenders and Crichton were among the earliest groups that immobilised type I pullulanase using the covalent attachment on aldehyde-activated amylose conjugate [80]. The immobilisation construct increased the enzymes optimum temperature and thermostability. Other favourable covalent attachments materials for type I pullulanases include polyacrylate, agar, silica, calcium alginate, Duolite XAD761 and magnetic chitosan beads. Interestingly in one report, immobilised enzyme on carboxyl-activated calcium alginate beads enhanced thermostability by 50 times than its free enzyme counterpart at 60 °C [81].
Besides covalent attachment, pullulan-degrading enzymes were also immobilised using ionic binding. For example, the amylopullulanase from Anoxybacillus sp. WB42 was immobilised using ionic binding on magnetic carboxymethyl cellulose (CMC) nanoparticles [82]. The reusable immobilised enzyme was 27 cycles at 60 °C, and it had similar biochemical characteristics to the free enzyme. Entrapment (encapsulation), an aged immobilisation concept, can increase the enzymes’ thermostability. For instance, the type I pullulanase entrapped in silica-magnetic nanoparticles had a half-life activity of 5 h at 60 °C, while the free enzyme was only stable for an hour at an identical temperature [83].
On a pilot and industrial scale, immobilising using purified enzymes is costly. From a research (laboratory scale) point of view, immobilising using purified enzymes is better than crude proteins extracted from the whole cells. By doing so, wresearchers can eliminate the influences of non-targeted background proteins [67]. For instance, any changes to the enzymes biochemical characteristics after immobilisation is genuinely contributed by the effect of the immobilisation carrier or the chosen technique, rather than an uncertain reason such as complex interactions of the proteins and other factors [84]. However, a significant amount of enzymes immobilisation research on a lab-scale have been carried out using the crude enzyme [85].
The lab investigated the effects of single enzyme-immobilisation on the product specificity of type I pullulanase (PulASK) from Anoxybacillus sp. SK3-4 [86]. Four derivatives of PulASK were prepared by immobilising the purified enzyme using the covalent attachment on three epoxides (i.e., ReliZyme EP403/M, Immobead IB-150P and Immobead IB-150A) and an amino-epoxide (i.e., ReliZyme HFA403/M)-activated support. The collected data indicates that the free and immobilised PulASKs were optimally active at 60 °C and pH 6.0. However, the product specificities of free and immobilised enzymes were dissimilar [86]. Free PulASK was reactive towards pullulan and β-limit dextrin, but it could not hydrolyse the short branches in starch. Hence, no reducing sugars were formed. In contrast, the immobilised PulASKs were able to hydrolyse starch at both short and long branches producing reducing sugars (G3 and G2) and oligosaccharides (≥G8), respectively. Moreover, the ratio of reducing sugars and oligosaccharides produced by each PulASK derivative using pullulan, starch, β-limit dextrin, glycogen and amylopectin were dissimilar. These phenomena were hypothesised to be associated with changes in the enzyme binding pocket influenced by the carrier surface properties (hydrophobic or hydrophilic) and the spacer arms lengths between the supports and proteins.
4. Co-Immobilisation of Pullulan-Degrading Enzymes
Bioconversion processes using enzymes such as starch degradation industries require multi-enzymatic reactions in a cascade manner. For simplicity and cost-effective purpose, industrial players may prefer a one-pot reaction concept using a cocktail of several enzymes. In the same practice, researchers have developed the idea of co-immobilised enzymes [87], where various types of enzymes are immobilised together or separately but applied together [88]. In a report, Talekar et al. [89] co-immobilised type I pullulanase, glucoamylase and α-amylase by cross-linked enzyme aggregates (combi-CLEAs) method. The one-pot starch hydrolysis shortened the process from 35 h (using free enzymes mixture) to 2 h (using co-immobilised enzymes). The co-immobilisation strategy exhibits multiple advantages [90]; however, some already known drawbacks are limited substrate diffusion and other issues [84].
Both enzymes were individually purified to be at least 98% purity. ReliZyme HFA403/M support was chosen for co-immobilisation. Ouresearcher's experimental setups contained:
- (i)(i) free enzymes mixture (PulASK+TASKA)
-
free enzymes mixture (PulASK+TASKA)
- (ii) individual immobilised enzymes mixture (HFA403/M–PulASK+ HFA403/M–TASKA)
- (ii)(iii) co-immobilised enzymes (PulASK–HFA403/M–TASKA).
-
individual immobilised enzymes mixture (HFA403/M–PulASK+ HFA403/M–TASKA)
- (iii)
-
co-immobilised enzymes (PulASK–HFA403/M–TASKA).The accumulated data from that setup indicated that the total amount and the generated spectrum of reducing sugars using (iii) co-immobilised enzymes were significantly different from either the (i) free enzymes mixture or (ii) the individual immobilised enzymes mixture [86]. Wresearchers thought that structural changes in PulASK and TASKA may alter the substrate-enzyme-product dynamic interaction and thus drastically modify their product specificities. Nevertheless, due to random covalent attachment, that work could not pinpoint the exact residues bound to the carrier, or how these residues affect the product specificity of the bound enzyme.
5. Protein Engineering for Improving the Performance of Pullulan-Degrading Enzymes
Protein engineering, also known as protein mutagenesis, is an approach that alters or improves enzyme biochemical features (i.e., substrate specificity, product specificity, specific activity, reducing inhibition effects or thermostability). The strategies include (i) directed evolution (error-prone PCR, staggered extension PCR, DNA shuffling, etc.) and (ii) rational design or site-directed mutagenesis (mainly mega primer PCR and overlap extension PCR methods) [117]. As expected, there are relatively fewer attempts for pullulan hydrolases engineering [37,118,119]. One central mutagenesis theme for pullulan-degrading enzymes is understanding the role of a residue at a particular position, specific amino acids stretch, motif or domain. For instance, Kim et al. [61] and Li et al. [126] performed CBM truncation to understand starch saccharification function. In another work, Kahar et al. elucidated that a short loop of the native protein prevents the native enzyme from generating reducing sugars from short linear or branched oligosaccharides [135]. In one particular study, Chen et al. shifted optimum pH from 5.0 to 4.0, and the mutant pullulanase exhibited an increased tolerance against acid denaturation [136]. Using pullulanase as a role model, the concept of ECSM (evolutionary coupling saturation mutagenesis) seems promising in cutting the time in identifying the amino acid positions for mutations [137]. The ECSM is an extension concept of an open-source software EVcouplings that perform computational predictions based on evolutionary sequence covariation (https://evcouplings.org). Improving activity half-life or structural stability is another central theme for pullulanase. For instance, among the six constructed variant proteins, pullulanase mutant G692M from Geobacillus thermoleovorans had a two-fold improvement in the activity half-life [127]. In the past decade, several computation predictors were created to assist researchers in doing in silico protein structure stabilisation analysis before starting to construct variants on the lab bench. Most of these predictors calculate ΔΔG (changes in the Gibbs free energy) between in silico native and variant proteins. In early times, these predictors applied the thermodynamic information recorded in the ProTherm database [140]. Approximately 20 years ago, ProTherm contained less than 6000 data entries and in 2013, the database updated to >25,000 thermodynamic data. Structure-based predictors, including FoldX, I-Mutant series, CUPSAT, PoPMuSiC, STRUM and etc., may outperform sequence-based predictors [141]. For the mutant pullulanase G692M that wresearchers stated at the beginning of this paragraph, the researchers utilised FoldX, I-Mutant 3.0 and dDFIRE [127]. From structural biologists’ perspective, X-ray protein structure is more accurate than homology-modelling predicted models. One may face drawbacks if the input files for structure-based predictors are the protein models instead of high-resolution X-ray structures. Crystallography is costly and the state-of-the-art AlphaFold may be an option to generate a better model structure.More recent predictors apply protein structural datasets and evolutional information to train machine learning. A few years ago, a research team noticed several problems with ProTherm, and they later cleaned the database before developing the PON-tstab predictor [148]. To overcome other limitations of ProTherm, the developers for DDGun and DDGun3D used a method that combines anti-symmetric features for predicting the ΔΔG upon variation, and claimed that this untrained and straightforward method has excellent performance. So far, only a few predictors address the anti-symmetric issue, and examples of these programs include PROTS-RF, INPS, SDM, ProTstab and a few others [149]. Predictor DeepDDG, a neural network approach, was trained using ProTherm, and manually curated literature data. The developer suggested that DeepDDG outperformed the other eleven commonly used predictors [146]. The identical research group that created PON-tstab later proposed a new platform, ProTstab, which used a >3500 proteolysis and mass spectrometry (LiP-MS) dataset to train the gradient-boosting based machine learning platform. Different algorithms may suggest contradictory suggestions, and it is time-consuming to try one predictor at a time. One may consider iStable 2.0, an integrated platform consisting of eleven structure- and sequence-based tools [147]. Readers can refer to thise excellent review that compares the limitations and challenges of each predictor from the angle of future applications in precision medicine [149]. Wresearchers observed that not many pullulanase protein engineers use the tools mentioned above; therefore, keywords stated in this subsection will not appear in bibliometric records (Figure 1c).
6. Influence of Next-Generation Metagenome Sequencing on Novel Pullulan-Degrading Enzymes Discovery
Next-generation sequencing (NGS) is a powerful tool in natural science and biotechnology in the current era. Metagenome sequencing with computation pipelines can generate a tremendous amount of novel macromolecule sequences. Interested readers in this topic can refer to these publications for more insights [150,151,152,153,154]. Wresearchers searched using the keyword ‘metagenome and X’ (where X is cellulase, lipase, amylase or pullulanase), the hit of research articles in the Scopus database is 180, 137, 60, 5, respectively. The articles related to pullulanase are listed in the references [155,156,157,158,159]. No relevant documents for pullulan hydrolase types I, II and III were found.A type I pullulanase derived from Reshi hot spring metagenome dataset was reported in 2021 [157]. The enzyme, named PulM, has very low sequence similarity to currently characterised pullulanases. For instance, PulM shares only 42.9% identity to a type I pullulanase from Geobacillus stearothermophilus. PulM had an optimum catalytic temperature at 40 °C. Interestingly, the enzyme remained 50% active at 4 °C, a relatively higher activity than other pullulanases obtained from isolated bacteria sampled from hot springs [157]. In another recent article, a novel type I pullulanase (PersiPul1) was established from cow rumen metagenomic data [156]. The enzyme is active from 30–80 °C and retained close to 60% of its original activity at 80 °C. In the same study, the researchers blended a cocktail of type I pullulanase (PersiPul1) and α-amylase PersiAmy2 (also derived from metagenome strategy, [160]). They used the enzyme cocktail to improve bread physical and sensory properties [156].Amylopullulanase PulSS4 was a type II pullulanase derived via the metagenome approach [159]. The gene encoding this enzyme was identified in a gut metagenome (clonal functional library approach) of Hermetia illucens (black soldier fly). PulSS4 closest counterpart (51%) is a protein from the anaerobic bacterium Amphibacillus xylanus. The authors claimed that PulSS4 had the maximum activity at pH 9.0 and exhibited a broad pH tolerance [159].The laboratory recently performed Illumina shotgun sequencing using environmental DNA extracted from biofilms attaching to plant litters socked in a local hot spring. Using the assembled contigs (Figure 3c), researchers have mined >10,000 sequences related to various CAZymes and identified close to 40 sequences for pullulan-degrading enzymes. All of them are type I or type II pullulanases, except one was detected as neopullulanase. This may suggest that pullulan hydrolases are rarely encoded in the genomes of thermophiles. Nevertheless, this approach enables uresearchers to find novel pullulanase sequences from rare prokaryotes, for instance, Candidatus Sericytochromatia and Candidatus Roseilinea. Ouresearcher's team is currently analysing the high-quality genomes generated by metagenome-assembled genomes (MAGs) using short-reads from Illumina. Besides, Oxford Nanopore and HiFi PacBio long-reads sequencing will soon become a common practice for MAGS. The MAGs strategy might extend ourresearcher's ability to mine novel pullulan-degrading enzymes.
Figure 3. Research flow chart to find novel active and thermostable pullulan-degrading enzymes.---------------------------------------------------------------------Research flow chart to find novel active and thermostable pullulan-degrading enzymes.Suppose searching for active and thermostable pullulan-degrading enzymes is the aim of ouresearcher's research. Despite the feasibility of protein engineering approach to design enzymes with improved thermostability and longer half-life, especially among the thermolabile origins, the engineered enzymes might be far from the temperature threshold required by the starch- and pullulan processing industries (Figure 3a). Harness native biocatalyst from a large pool of metagenome-derived enzymes, particularly from samples such as high-temperature hot springs, is perhaps a better route than the protein engineering approach (Figure 3c). Biologists may be unacquainted with the advanced bioinformatic tools summarised in Figure 3c. In addition, from ouresearcher's own experience and the limited examples of pullulan-degrading enzymes [155,156,157,158,159], metagenome-derived enzymes may exhibit low sequence similarity to patented sequences.
