2.4.2. The Hydrolysis of Maltose and Maltose-Like Sugars

We predicted *in silico* that *Ba*AG2 is either maltase or maltase-isomaltase (see Figure 1). To find out which of the predictions was correct, the purified *Ba*AG2 protein was reacted with a selection of 100 mM sugars and 1 mM *p*NPG that serves as a substrate for maltases, isomaltases and maltase-isomaltases (Figure 5). According to our assay, *Ba*AG2 could hydrolyze universal substrates (*p*NPG and sucrose), maltose and maltose-like sugars such as turanose, maltotriose, melezitose and maltulose. Isomaltose and isomaltose-like substrates palatinose and α-methylglucoside were not hydrolyzed (Figure 5). Therefore, *Ba*AG2 should be considered as a maltase.

**Figure 5.** Activity of *Ba*AG2 on 100 mM sugars and 1 mM *p*NPG. Universal substrates are indicated by green, maltose and maltose-like sugars by yellow, and isomaltose and isomaltose-like sugars by blue bars. The composition and linkage types of the tested substrates are indicated. SD values of three to five replicates on each substrate are indicated.

2.4.3. The Kinetic Parameters of Hydrolysis of Maltose, Maltose-Like and Universal Substrates

We studied the kinetics of the hydrolysis of *p*NPG, maltose, sucrose, maltotriose, melezitose, maltulose and turanose to calculate Km, Vmax, *k*cat and catalytic efficiency (*k*cat/Km) values for these substrates. Results are presented in Table 1.


**Table 1.** Km, Vmax, *k*cat and *k*cat/Km values of hydrolysis of *p*NPG and sugars by *Ba*AG2.

\* For the structure and linkage type of the substrates, see Figure 5. SD, standard deviation.

Table <sup>1</sup> shows that natural sugars, maltose and sucrose (α-d-Glc-(1→2)-β-d-Fru) were hydrolyzed by *Ba*AG2 with the highest catalytic efficiency (*k*cat/Km). Additionally, *Ba*AG2 had a high affinity and activity towards a synthetic substrate *p*NPG—the Km for *p*NPG was 0.76 mM and the Vmax over 750 U/mg. Affinity of *Ba*AG2 for maltose was slightly higher than for sucrose. From this aspect, *Ba*AG2 differs from the maltases of *S. cerevisiae* [23,24] and *Candida albicans* [41], and also from the maltase-isomaltase of *O. polymorpha* [16,40,42], for which affinity for sucrose is about twice higher than for maltose. In contrast, the maltase of *Schizosaccharomyces pombe* prefers maltose to sucrose [33]. Among the substrates tested, the affinity and catalytic efficiency of *Ba*AG2 was the lowest for melezitose. Thin layer chromatography (TLC) analysis showed that the glycosidic bond of turanose moiety of melezitose was hydrolyzed first, yielding sucrose and glucose as products (Figure S3). Similar mode of melezitose hydrolysis was earlier shown for the maltase-isomaltase of *O. polymorpha* [16]. Interestingly, isomelezitose was hydrolyzed by *Ba*AG2 with release of palatinose (Figure S3).

We would like to emphasize that the activity of *Ba*AG2 (Table 1) was higher compared to some of other studied maltases of yeasts and filamentous fungi. For example, its Vmax on maltose was 7.5 times higher compared to *S. cerevisiae* maltase MAL62 [23], and over two times higher compared to maltase-isomaltase of *O. polymorpha* [16]. On the other hand, α-glucosidases of *X. dendrorhous* and *A. niveus* preferred polysaccharides such as starch, amylopectin and glycogen, and their *k*cat values on maltose were respectively 25 and 28 times lower compared to the value of *Ba*AG2 [43,44]. Catalytic constant of α-glucosidase of *A. niger* on maltose was 144 1/s (2.6 times lower than *Ba*AG2), but the affinity towards maltose was very high (0.75 mM) [45]. Based on the literature, only one α-glucosidase of yeast and filamentous fungi has much higher *k*cat on maltose than that of *Ba*AG2—the extracellular α-glucosidase of *Schizosaccharomyces pombe* (*k*cat = 709 1/s) [46].

#### 2.4.4. The Inhibition Studies of Acarbose, Tris, Isomaltose-Like Sugars and Glucose

Having seen that isomaltose and isomaltose-like sugars are not hydrolyzed by *Ba*AG2 (Figure 5), we measured inhibition of *p*NPG hydrolysis reaction by these sugars as in [16]. Because maltases are usually strongly inhibited by glucose and acarbose, and much less by fructose [16], these substrates were also assayed as potential inhibitors of *Ba*AG2. When testing the effect of different buffers during the experiments, we noticed that the activity of *Ba*AG2 was lost in Tris-HCl buffer. More precise assay of the effect of tris(hydroxymethyl)aminomethane (Tris) on *Ba*AG2 revealed strong inhibition of the enzyme by this compound (Table 2). All tested compounds inhibited *p*NPG hydrolysis competitively, and the most powerful inhibitors of *Ba*AG2 were acarbose, Tris and glucose.


**Table 2.** Ki values and inhibition mode for *Ba*AG2 inhibitors of *p*NPG hydrolysis.

\* These values are given in μM. SD, standard deviation.

We expected that binding of the substrates or competitive inhibitors of the enzyme should increase its thermostability. To confirm this, we conducted a differential scanning fluorimetry (DSF) assay of *Ba*AG2 in the presence and absence of competing inhibitors as in [16,47]. Acarbose, palatinose, Tris and glucose (strong inhibitors of *p*NPG hydrolysis by *Ba*AG2) and fructose that inhibited the reaction only weakly (see Table 2) were used as ligands. The melting temperature (Tm) values calculated from the DSF data are presented in Figure 6. The Tm of *Ba*AG2 without a ligand was similar to that of maltase-isomaltase MAL1 of *O. polymorpha*—51 ◦C [16]. Presence of acarbose (the strongest inhibitor of *Ba*AG2) increased the Tm of *Ba*AG2 by 11.4 ◦C. Presence of Tris raised the Tm value by 5.8 ◦C, and of glucose by 4.4 ◦C. Fructose and palatinose had only a minor effect on the Tm (Figure 6).

**Figure 6.** Thermostability of *Ba*AG2 in the presence and absence of indicated ligands. The Tm value of unliganded *Ba*AG2 (a grey bar) was 51 ◦C and the increase of Tm in the presence of a ligand is indicated above every bar. SDs of at least two independent experiments with two replicates at each condition are shown by error bars.

#### 2.4.5. The Hydrolysis of Malto-Oligosaccharides of DP 3 to 7

We have earlier shown that *O. polymorpha* maltase-isomaltase MAL1 could use maltotriose and maltotetraose as a substrate, while malto-oligosaccharides (MOS) of higher degree of polymerization (DP) were not hydrolyzed [16]. We showed thatMOS of DP3 (see Table 1), DP4, DP5, DP6 and DP7 served as substrates for *Ba*AG2 (Figure S4). Assay of reaction course indicated that exo-hydrolysis occurred—a glucose residue was stripped off from the oligomer. Activity on MOS was only moderate and a substantial proportion of it stayed unreacted even after extended (22 h) reaction time (see Figure S4). The MAL62 of *S. cerevisiae* that was assayed alongside could not hydrolyze MOS longer than maltotetraose (DP4) (Figure S4).

#### 2.4.6. The Hydrolysis of Amylose, Amylopectin and Glycogen

Surprisingly, we detected the ability of *Ba*AG2 to hydrolyze polysaccharides that is a rather exceptional feature among maltases. After we noticed that *Ba*AG2 could hydrolyze soluble starch with the release of glucose, we performed a more detailed assay testing the hydrolysis of a set of polymeric α-glucans: amylose and amylopectin from potato, glycogen from oysters and dextrans of Mw 20 and 110 kDa. Commercial amylolytic enzymes amyloglycosidase (glucoamylase) from *Aspergillus niger,* and α-amylase from *Aspergillus oryzae* and *S. cerevisiae* MAL62 were used as reference enzymes.

Hydrolysis of the polymers was evaluated according to the release of glucose and by TLC analysis of reaction products. The commercial enzymes hydrolyzed amylose, amylopectin and glycogen rapidly and with the expected pattern of products (Figure 7). Dextrans were hydrolyzed only by the amyloglycosidase, and the release of glucose was minimal. *Ba*AG2 and *Sc*MAL62 had no activity on dextrans. However, *Ba*AG2 exhibited moderate, but clearly detectable and recordable exo-hydrolysis of amylose, amylopectin and glycogen. The activity was the highest with amylopectin, and the lowest with amylose (Figure 7). From amylopectin (5 g/L), 0.1 g/L of glucose was released by 24 h, and 0.3 g/L by 72 h of the reaction. In the case of *Sc*MAL62, no hydrolysis of amylose was detected, and hydrolysis of glycogen and amylopectin became detectable only by 72 h of the reaction (Figure 7).

**Figure 7.** Hydrolysis of polysaccharides (5 g/L) by *Ba*AG2, *Sc*MAL62 and amyloglycosidase of *A. niger* (AG) and α-amylase of *A. oryzae* (α-AM). Samples withdrawn at indicated time points were analyzed using TLC. Reaction mixtures were spotted on TLC plates alongside with reference sugars (M): Glc (30 mM glucose); Mal (10 mM maltose); MalTri (10 mM maltotriose); DP4 (10 mM maltotetraose). The same marker sugars were used in all assays but are marked only on TLC plate of glycogen degradation. C—control sample without the enzyme but containing 5 g/L bovine serum albumin (BSA) incubated at the same conditions for 74/96 h. Glucose release was quantified enzymatically. See Materials and Methods, paragraphs 4.3 and 4.6 for details. SDs of two to three replicates are shown by error bars.

#### 2.4.7. The Transglycosylation of Maltose

Many α-glucosidases can transglycosylate and produce short oligosaccharides, especially at high concentration of the substrate [17–21]. We assayed this possibility by incubating *Ba*AG2 with 250 and 500 mM maltose up to 72 h and analyzed the reaction products by HPLC. The maltase MAL62 of *S. cerevisiae* was used as a reference. Figure 8 and Table S3 show that already within 2 h at 250 mM (85.6 g/L) maltose concentration, *Ba*AG2 produced maltotriose (4.2 g/L) and panose, <sup>α</sup>-d-Glc-(1→6)-α-d-Glc-(1→4)-d-Glc (1.6 g/L), in addition to a maltose hydrolysis product—glucose. By 72 h of reaction, the maltotriose content was decreased and panose content increased to 2.6 g/L (Figure 8, Table S3). The *Sc*MAL62 produced only maltotriose under the same conditions, and its amount was considerably lower than in the case of *Ba*AG2—only 2.0 g/L produced by 2 h of reaction (Figure 8, Table S3). Transglycosylation of maltose was enhanced at 500 mM (171.2 g/L) concentration: up to 13.3 g/L of maltotriose was produced by 2 h and 10.4 g/L of panose by 72 h of reaction. Notably, a new transglycosylation product, isomaltose, emerged. It was produced by both enzymes, but its concentration was certainly higher in the case of *Ba*AG2—its concentration in the 72-h reaction sample was 5.2 g/L (Table S3). By 24 h of reaction with 500 mM maltose, the amount of transglycosylation products was 12.6% of total sugars in the reaction mixture, while the respective value for the MAL62 protein was about three times less—4.5% (see Table S3).

**Figure 8.** Transglycosylation of maltose by *Ba*AG2 and *Sc*MAL62. 20 μg/mL of the enzyme was reacted with 250 mM or 500 mM of maltose. Samples from the reaction mixtures were withdrawn at designated time points, heated to terminate the reaction and analyzed for sugar composition by HPLC as described in Materials and Methods, paragraphs 4.5. and 4.6. Total amount of detected saccharides at each time point was equalled to 100%. Products were identified using glucose, maltose, isomaltose, maltotriose and panose as references. SDs of two to three HPLC measurements at each time point were up to 20%.

We tested panose hydrolysis by *Ba*AG2 and conclude that it accumulated in transglycosylation reaction since it was not hydrolyzed by the enzyme even during extended (22 h) reaction time (Figure S3). In contrast, maltotriose was hydrolysed by *Ba*AG2 (Figure 5, Table 1), and due to that, its content decreased at prolonged transglycosylation reaction (Figure 8, Table S3).

#### **3. Discussion**

Utilization of α-glucosidic sugars by maltases and isomaltases has earlier been thoroughly studied in *S. cerevisiae* because metabolism of these sugars is crucial in brewing and baking [22,48]. However, transport and intracellular hydrolysis of α-glucosidic sugars have also been investigated in *Ogataea polymorpha* [25] and *Schizosaccharomyces pombe* [33,49]. A maltase has been characterized from *C. albicans* [41] and four maltase-isomaltases from *Sche*ff*ersomyces stipitis* [12]. Genes potentially encoding for either maltases, isomaltases or maltase-isomaltases were recently discovered in the genomes of many non-conventional yeasts [12]. Yeast species with deep phylogeny were expected to possess ancient-like α-glucosidases [12]. Figure 9 shows the phylogram of selected yeast species and *A. oryzae* based on sequence analysis of D1/D2 domains of large subunit ribosomal RNA to illustrate the evolutionary relationships between the species. *B. adeninivorans* and *Lipomyces starkeyi* belong to basal group of Saccharomycotina [50]. Based on phylogenetic analysis of orthologous proteins [4], this group diverged from *S. cerevisiae* lineage 200 to 400 million, and from the CTG clade 200 million years ago. The basal group is considered very heterogeneous, its most studied member is *Yarrowia lipolytica* (not presented in Figure 9) and the most basal lineage to this group and all Saccharomycotina is *Lipomyces starkeyi* [50].

**Figure 9.** Phylogram based on analysis of D1/D2 domains of large subunit ribosomal RNA (rRNA) gene sequences of ten yeast species and of a filamentous fungus *Aspergillus oryzae*. The bootstrap values (1000 replicates) are shown at the nodes. The scale bar shows the number of base substitutions per site. The Saccharomycotina clades according to [50] are designated by background coloring.

On the phylogram (Figure 9), *B. adeninivorans* clusters with *Lipomyces starkeyi*. Genes for eight intracellular α-glucosidases (five maltases and three isomaltases) were predicted in the genome of *L. starkeyi.* In the phylogram of α-glucosidases from non-conventional yeasts, all eight *Lipomyces* enzymes clustered with those of *B. adeninivorans* [12]. However, these putative α-glucosidases of *L. starkeyi* have not been cloned for protein expression and characterization. In Kelly et al. an extracellular α-glucosidase from *L. starkeyi* was biochemically characterized [51]. Unfortunately, the sequence data of this protein is not available. The above-mentioned enzyme of *L. starkeyi* had equally high activity on maltose and isomaltose, but it also hydrolyzed maltotriose, isomaltotriose, panose, amylopectin and starch. Its activity with starch and amylopectin comprised 31 and 42% of that with maltose and isomaltose, and due to this feature, the authors considered it more similar to fungal rather than yeast enzymes [51]. Interestingly, this enzyme did not hydrolyze sucrose and had a quite low activity on *p*NPG. Typically, yeast α-glucosidases hydrolyze *p*NPG more rapidly than maltose. or sucrose [16,24] (Table 1) whereas the opposite is true for maltases of bacteria and archaea [52–54].

Our study showed that *Ba*AG2 is a maltase. *Ba*AG2 efficiently hydrolyzed maltose and maltose-like sugars, but could not hydrolyze isomaltose, palatinose nor α-MG that are specific substrates for yeast isomaltases (Figure 5, Table 1). Isomaltose, isomaltose-like sugars as well as acarbose and glucose competitively inhibited *p*NPG hydrolysis by *Ba*AG2 and increased thermostability of the enzyme (Table 2; Figure 6). We also discovered a very strong inhibition of *Ba*AG2 by Tris with the Ki of 70.5 μM (Table 2). Tris also considerably increased thermostability of *Ba*AG2 in a DSF assay (Figure 6). As thermostability of *S. cerevisiae* isomaltases was also elevated in the presence of Tris [55], it may bind to the active site of these enzymes. Literature mining revealed Tris as a competitive inhibitor of *Bacillus brevis* maltase with the Ki of 14.5 mM [56]. In a yeast *Brettanomyces bruxellensis* (former name *Br. lambicus*)*,* both extra- and intracellular maltases were inhibited not only by acarbose (Ki values between 28.5 and 57 μM) but also by Tris (Ki values between 7.45 and 15.7 mM) [57]. Compared to *Br. bruxellensis* and *Bacillus brevis* maltases, *Ba*AG2 was much more sensitive to Tris. The pH optimum of *Ba*AG2 was in a moderate acidic region as in the case of other yeast maltases [23,24]. The temperature optimum for *Ba*AG2 was between 40–50 ◦C, with maximum activity (530 U/mg) achieved at 45 ◦C. Thermostability of the enzyme was not high—after keeping the enzyme at 45 ◦C for 30 min, the enzyme's activity dropped to 46% from the initial. Incubation for 30 min at 50 ◦C totally inactivated the enzyme (Figure 4). According to literature data, thermolability has been reported for some other yeast α-glucosidases. For example, the four isomaltases of *S. cerevisiae* had all low thermostability. Of those, the IMA1 was most stable, and IMA5 the least stable [55].

According to Hasegawa et al., the MalT protein of *A. oryzae* that has 51% of sequence identity to *Ba*AG2 (Table S2) exhibited *p*NPG-hydrolyzing activity, its expression was induced when grown on maltose, and thereby the MalT was defined by the authors as a maltase [36]. To date no additional data on MalT protein is available. The intracellular maltase (MAL1) protein of *Schizosaccharomyces pombe*, with 43.2% of sequence identity to *Ba*AG2 hydrolyzed *p*NPG and maltose, but had also activity on soluble starch, and some activity on sucrose [33].

Intriguingly, *Ba*AG2 had a detectable hydrolytic activity on MOS with DP up to 7, glycogen, amylose and amylopectin (Figure 7 and Table S4). We assume that the ability to hydrolyze MOS and to cleave polymeric α-glucans, at least to some extent, may be characteristic to maltases of early-diverged yeasts. As *Ba*AG2 is an intracellular enzyme, and this yeast possesses a secreted glucoamylase [9], the maltase *Ba*AG2 has most probably no role in starch degradation. However, yeasts synthesize glycogen as a reserve polysaccharide [58]. Considering that *Ba*AG2 had a remarkable activity on glycogen, the enzyme may contribute to glycogen catabolism in *B. adeninivorans.* We hypothesize that the ability to degrade malto-oligosaccharides and α-glycosidic polysaccharides may be characteristic for the maltases of yeasts with deep phylogeny. Isolation and study of α-glucosidases of the most basal lineages to Saccharomycotina, for example *Lipomyces starkeyi*, should verify the correctness of this hypothesis.

Several GHs of yeasts and filamentous fungi exhibit transglycosylating activity. For example, α-glucosidase of a yeast *X. dendrorhous* (syn. *Pfa*ffi*a rhodozyma*) produced from maltose a large proportion of transglycosylation products with α-1,6 linkage, among which panose was the most abundant [17]. It has been shown that the *S. cerevisiae* maltase also produced panose from maltose, yet transglycosylating activity of the *Saccharomyces* enzyme was more than three times smaller compared to the *X. dendrorhous* enzyme. Both enzymes also synthesized isomaltose and maltotriose, but the latter was rapidly used by the enzymes and was therefore not present among the final products [17]. We showed that maltotriose was produced from maltose also by *Ba*AG2 (Figure 8, Table S3), but as it serves as a substrate for the enzyme (Table 1), it was eventually hydrolyzed. α-Glucosidases of filamentous fungi

have been suggested as feasible catalysts for transglycosylation. Thus, an *Aspergillus* enzyme with high transglycosylating activity was reported to produce panose and isomaltose from maltose [18,59,60]. When an *A. nidulans* α-glucosidase with strong transglycosylating activity was reacted with 5 g/L maltose during 6 h, approximately 50% of maltose was converted to transglycosylation products, 60% of which was isomaltose [60]. Notably, in addition to maltotriose and panose, isomaltose was also detected among the transglycosylation products produced from maltose by *Ba*AG2 (Figure 8, Table S3). Isomaltose synthesis from maltose was also confirmed for *Sc*MAL62, even though the content of it was only minimal (Figure 8, Table S3).

### **4. Materials and Methods**

### *4.1. Yeast and Bacterial Strains, Cultivation of B. adeninivorans*

*Blastobotrys* (*Arxula*) *adeninivorans* LS3 (CBS 8244) was kindly provided by Assoc. Prof. V. Passoth (SLU, Uppsala, Sweden). The yeast strain was grown on solid YPD medium (20 g/L peptone, 20 g/L glucose, 10 g/L yeast extract, 20 g/L agar) at 30 ◦C 24 h for harvesting the cells for genomic DNA extraction. *E. coli* DH5α (Thermo Fisher Scientific, Waltham, MA, USA) was used for DNA cloning and plasmid production. *E. coli* BL21 (DE3) [61] was used for heterologous expression of *Ba*AG2. The ability of *B. adeninivorans* to grow on sugars was assayed as in [12]. Yeast cells grown overnight on BD Difco YNB medium (Thermo Fisher Scientific, Waltham, MA, USA) without amino acids containing 2 g/L glucose were used as inoculum. The YNB medium was supplemented with 2 g/L of a filter-sterilized sugar (glucose, maltose, maltotriose, isomaltose, maltulose, α-methylglucoside, sucrose, raffinose, melibiose, turanose, palatinose, melezitose or trehalose). The cells were incubated on Greiner 96-well flat-bottom transparent polystyrene microplates (Greiner Bio-One, Frickenhausen, Germany) in 200 μl under agitation for 24 h at 37 ◦C. Optical density of the culture at 600 nm was measured against inoculated medium without sugar at the beginning and end of the experiment using an Infinite M200 PRO microplate reader (Tecan Group Ltd., Männedorf, Switzerland) equipped with Tecan i-control v. 1.7 software from the same provider. Two independent experiments with two parallel measurements were conducted.

#### *4.2. Cloning, Heterologous Expression and Protein Purification*

Genomic DNA of *B. adeninivorans* was extracted using PowerMax Soil DNA Isolation Kit (MoBio Laboratories, Carlsbad, CA, USA) and the standard protocol by manufacturer. The oligonucleotide primers Ba20130g\_PURICter\_Fw (5 **TAACTTTAAGAAGGAGATATACAT** *ATGGTTCTAGGATTTTTCAAAAAG* 3 ) and Ba20130g\_PURICter\_Rev (5 **GCTATTAATGATGATG ATGATGAT***GGATTTCATAGATGACTGCCTCCA* 3 ), designed according to the gene sequence of *rna\_ARAD1D20130g*, were applied to amplify a 1789 bp fragment that represented the coding sequence of *Ba*AG2 [12]. In the primers, the nucleotides annealing with the pURI3Cter vector [62] are shown in bold and those annealing with the *BaAG2* gene sequence are shown in italics. The ATG start codon and the stop codon are underlined in the primer sequences. Recombinant *Pfu* polymerase (Thermo Fisher Scientific, Waltham, MA, USA) and 2.6 ng per μL of reaction mixture of genomic DNA were used in amplification. The PCR product was cloned into a pJET vector from CloneJET PCR Cloning Kit (Thermo Fisher Scientific, Waltham, MA, USA), yielding pJET-BaAG2. To produce the *Ba*AG2 protein with a C-terminal His6-tag, an expression plasmid pURI3-AG2Cter was constructed by cloning the *BaAG2* gene into pURI3-Cter vector similarly as in [63]. Insertion of the *BaAG2* gene into the vector was confirmed by DNA sequencing. The cloning procedure was conducted with recombinant *Pfu* polymerase (Thermo Fisher Scientific, Waltham, MA, USA). DNA Clean & Concentrator-5 Kit (Zymo Research, Irvine, CA, USA) was used for purification and concentration of PCR products. Plasmid DNA was purified with FavorPrep Plasmid Extraction Mini Kit (Favorgen Biotech Corp., Ping-Tung, Taiwan). The *MAL62*-containing plasmid pURI-ScMAL62Cter [12] was used to produce the *S. cerevisiae* maltase protein that was analyzed as a reference.

The pURI3-BaAG2Cter and pURI-ScMAL62Cter containing the *BaAG2* gene or *MAL62* gene, respectively, were electroporated into *E. coli* BL21 (DE3) for heterologous expression. A simplified autoinduction medium as in [64] was used for protein overproduction: the LB-based medium (20 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl) was supplemented with 25 mM phosphate buffer (Na2HPO4/KH2PO4; pH 7.2) and 3 g/L glycerol to which filter-sterilized 0.25 g/L glucose and 1 g/L lactose were added. Medium for transformant selection contained ampicillin (Amp, 150 mg/L) for plasmid preservation. Bacterial cells were grown overnight in LB-Amp medium at 37 ◦C and diluted 100 times in autoinduction medium. At first, the cultures were incubated for 2 h at 37 ◦C followed by 20-h incubation at 22 ◦C. The cells were harvested by centrifugation (2400× *g*, 10 min) at 4 ◦C and stored at −20 ◦C until further use. Cells were disrupted by sonication with Ultrasonic Homogenizer (Cole-Parmer Instrument Company, Vernon Hills, IL, USA) in 100 mM K-phosphate buffer (pH 6.5) with the cOmplete, EDTA-free Protease Inhibitor Cocktail (Roche Diagnostics, Mannheim, Germany), and centrifuged 30 min at 2400 × *g* at 4 ◦C. The resulting supernatants were syringe-filtered (pore size 0.45 μm) and loaded onto an IMAC HisTrap FF column coupled with an ÄKTAprime plus chromatography system (GE Healthcare, Uppsala, Sweden). Further purification steps were conducted as described in [63]. Proteins were quantified by measuring the absorbance at 280 nm. The respective extinction coefficients of *Ba*AG2 [133,855 1/(M × cm)] and *Sc*MAL62 [148,990 1/(M × cm)] were computed at ExPASy Proteomics Server (http://expasy.org). Purified proteins were maintained in 100 mM K-phosphate buffer (pH 6.5) at 4 ◦C.

#### *4.3. Determination of Substrate Specificity, Kinetic Parameters and Inhibition*

Hydrolysis of *p*NPG was assayed as in [16,40] according to the release of *p*-nitrophenol. 100 mM K-phosphate buffer (pH 6.5) was used and reactions were conducted at 30 ◦C if not stated otherwise. The purified enzyme was unstable if diluted therefore 5 g/L bovine serum albumin (BSA) was added to the dilution buffer as a protein stabilizer to retain its full catalytic activity. For preliminary assay of substrate specificity of *Ba*AG2, the enzyme was reacted with 100 mM concentration of various potential substrates: maltose, sucrose, maltotriose, isomaltose, melezitose, maltulose, turanose, palatinose or α-MG. At fixed time points, aliquots were withdrawn, combined with three volumes of 200 mM Tris buffer (pH 8.3) and subsequently heated at 96◦C for 5 min to stop the reaction. The content of the released glucose was determined spectrophotometrically as in [16,47]. The activities (μmoles of glucose released per minute per mg of protein; U/mg) were calculated from initial velocities of the reaction. For kinetic analysis, initial rates of *p*-nitrophenol or glucose release from substrates were measured at four to seven concentrations ranging from 0.1–3.0 mM for *p*NPG and 2.5–250 mM for diand trisaccharides. At least three independent measurements for each substrate and concentration were made. *Ba*AG2 content in reaction mixtures ranged from 0.02 to 3.5 μg/mL. The initial velocity data analysis with enzyme kinetics module of the SigmaPlot (Systat Software, San Jose, CA, USA) yielded kinetic parameters (Km, Vmax) for the enzyme. *k*cat and *k*cat/Km were calculated from these data. The theoretical Mw value of 67,901 Da for the *k*cat calculation was computed in ExPASy Proteomics Server (http://expasy.org).

Inhibition of *Ba*AG2 was studied by incubating a suitable amount of enzyme (0.035–0.13 μg/mL) with 0.2–2.0 mM *p*NPG in the presence of following potential inhibitors: 5 μM acarbose, 0.1 mM Tris, 5 mM palatinose, 10 mM isomaltose, 10 mM glucose, 100 mM α-methylglucoside, 100 mM trehalose or 400 mM fructose. The Ki values were calculated using enzyme kinetics module of the SigmaPlot (Systat Software, San Jose, CA, USA).

Differential scanning fluorimetry (DSF) was used to evaluate the thermostability of *Ba*AG2 in the presence and absence of ligands. The concentration of *Ba*AG2 was used with 2 μM and for ligands: 100 mM fructose, 50 mM palatinose, 50 mM glucose, 5 mM Tris and 5 mM acarbose. The reaction was conducted in 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) (HEPES) buffer (pH 7.0) with 150 mM NaCl. The experiment was based on [16,47,65] with above-mentioned modifications. At least two independent experiments were performed with two technical replicates.

Degradation of glucose polymers, i.e., amylopectin-free amylose from potato, amylopectin from potato, glycogen from oysters and dextrans of Mw 20 kDa and 110 kDa, was assayed in K-phosphate buffer (100 mM, pH 6.5) containing 0.2 g/L Na-azide. Polysaccharide concentration in the reaction mixture was 5 g/L, and 20 μg/mL of the enzyme was used. At desired time points (2 h, 24 h, 74/96 h) aliquots were withdrawn and heated to stop the reaction. A negative control containing 20 μg/mL BSA instead of the enzyme was incubated alongside. Concentration of released glucose was measured as described above and 3 μl of the samples were analysed on TLC. Maltase from *S. cerevisiae* (*Sc*MAL62), amyloglycosidase from *Aspergillus niger* (Sigma-Aldrich, Merck, Darmstadt, Germany) and α-amylase from *Aspergillus oryzae* (Sigma-Aldrich, Merck, Darmstadt, Germany) were reacted with tested glucose polymers for comparison.

To assay the hydrolysis of malto-oligosaccharides (DP 4-7), panose and melezitose by *Ba*AG2 and *Sc*MAL62, 50 mM of the sugar was reacted with 2.6 μg/mL of the enzyme in K-phosphate buffer (100 mM, pH 6.5) containing 0.1 g/L of Na-azide and samples were withdrawn at fixed time points. 20 mM isomelezitose (transglycosylation product from sucrose of *Ogatae polymorpha* MAL1) was also tested as a potential substrate. The reaction samples were analyzed using TLC.

#### *4.4. Determination of pH and Temperature Optima, Evaluation of Thermostability*

Initial velocity of 1 mM *p*NPG hydrolysis by *Ba*AG2 was measured at 30 ◦C in BSA-supplemented buffers of varied pH (from 3.8 to 8.5) using McIlvaine's buffer (Na-phosphate/citrate buffer) [66] and 100 mM K-phosphate buffer to cover respective pH-interval, and data were plotted against the pH to determine the pH optimum. Hydrolysis of 1 mM *p*NPG was measured at varied temperatures from 20 to 65 ◦C. Initial velocity data were plotted against the temperature to reveal the temperature optimum. For thermostability determination, *Ba*AG2 was incubated in 100 mM K-phosphate buffer (pH 6.5) buffer containing 5 g/L BSA for 30 min at temperatures 10, 20, 30, 37, 45 and 50 ◦C. After cooling the samples on ice, residual activity of the enzyme was determined according to the hydrolysis of 1 mM *p*NPG at 30 ◦C. Every temperature point was assayed in triplicate. The activity measured after incubation at 10 ◦C was taken as 100%.

#### *4.5. Study of Transglycosylation*

20 μg/mL of the enzyme (*Ba*AG2 or *Sc*MAL62) was incubated in 100 mM K-phosphate buffer (pH 6.5) with 0.2 g/L Na-azide and 250 mM or 500 mM maltose at 30 ◦C up to 72 h. The samples with *Ba*AG2 also contained 5 g/L BSA. Samples were withdrawn at fixed intervals, heated at 95 ◦C to stop the reaction and analysed on TLC and by HPLC.

#### *4.6. Chromatography of Substrates and Reaction Products*

To visualize hydrolysis and polymerization products, the TLC analysis was conducted as in [16] on Silica Gel 60 F254 plates with concentrating zone (Merck, Darmstadt, Germany). 0.5 μl of the stopped reaction mixture were spotted onto the plate and sugars were separated with two runs in chloroform:acetic acid:water (6:7:1, v:v:v) [67]. For the analysis of products of polysaccharide degradation, 3 μl of the reaction mixtures were spotted. Sugars were visualized by immersion of the plates in aniline-diphenylamine reagent and subsequent heating of the dried plates at 100 ◦C [68].

HPLC analysis was performed similarly as in [65]. Glucose and fructose were used to calibrate the column. Fructose, maltose, sucrose, isomaltose, palatinose, turanose, maltotriose, panose and melezitose were used as reference sugars.

#### *4.7. Alignment of RNA and Protein Sequences and Construction of the Phylogram*

Gene sequences from domains 1 and 2 (D1/D2) of large subunit ribosomal RNA were aligned to build a neighbor-joining phylogenetic tree [69] of yeasts using MEGA v. 7.0 [70]. The maximum composite likelihood model [71] with 1000 bootstrap replicates was applied. Protein sequences were aligned using Clustal Omega [72] to calculate identity values of the proteins.
