Structure of the β-Galactosidase Gene from Thermus sp. Strain T2: Expression in Escherichia coli and Purification in a Single Step of an Active Fusion Protein

Abstract

The nucleotide sequence of both the bgaA gene, coding for a thermostable β-galactosidase of Thermus sp. strain T2, and its flanking regions was determined. The deduced amino acid sequence of the enzyme predicts a polypeptide of 645 amino acids (M_r, 73,595). Comparative analysis of the open reading frames located in the flanking regions of the bgaA gene revealed that they might encode proteins involved in the transport and hydrolysis of sugars. The observed homology between the deduced amino acid sequences of BgaA and the β-galactosidase of Bacillus stearothermophilus allows us to classify the new enzyme within family 42 of glycosyl hydrolases. BgaA was overexpressed in its active form in Escherichia coli, but more interestingly, an active chimeric β-galactosidase was constructed by fusing the BgaA protein to the choline-binding domain of the major pneumococcal autolysin. This chimera illustrates a novel approach for producing an active and thermostable hybrid enzyme that can be purified in a single step by affinity chromatography on DEAE-cellulose, retaining the catalytic properties of the native enzyme. The chimeric enzyme showed a specific activity of 191,000 U/mg at 70°C and a K_m value of 1.6 mM with o-nitrophenyl-β-d-galactopyranoside as a substrate, and it retained 50% of its initial activity after 1 h of incubation at 70°C.

β-d-Galactosidase (EC 3.2.1.23) catalyzes the hydrolysis of β-1,4-d-galactosidic linkages. This enzyme is distributed in numerous microorganisms, plants, and animal tissues. The application of β-galactosidase to the hydrolysis of lactose in dairy products, such as milk and cheese whey, has received much attention (7, 21), and in this regard, thermostable β-galactosidases have attracted increasing interest because of their potential usefulness in the industrial processing of lactose-containing products (21). Thermostable enzymes have a number of generally recognized advantages in industrial applications, such as associated chemical resistance and reduced chances of microbial growth at high temperatures (15, 19). Nevertheless, relatively few studies have been conducted on β-galactosidases from thermotolerant or thermophilic bacteria, and as far as we know, only four genes encoding these enzymes have been cloned (5, 10, 11, 13, 16, 18).

An important property that has received little attention in the literature is the level of purity of commercial preparations of β-galactosidases, especially with regard to the presence of other enzymes, such as proteases. These contaminants could have a severe impact on the stability of the enzyme, leading to undesirable changes in dairy products during storage (21). To prevent these, a new method was developed to purify the β-galactosidase (LacZ) of Escherichia coli by fusing to its N terminus the choline-binding domain (ChBD) of the pneumococcal autolytic amidase LytA (23). This system allowed the purification of E. coli β-galactosidase in a single step by affinity chromatography on DEAE-cellulose (23). Thus, it appeared interesting to test whether this procedure could also be used in the purification of a thermostable enzyme in order to circumvent contamination problems.

This paper reports the molecular characterization of the bgaA gene, encoding the β-galactosidase (BgaA) of Thermus sp. strain T2, and describes the construction of a ChBD-BgaA chimera which retains the biochemical properties of the native enzyme and can be purified in a single chromatographic step.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

E. coli JM109 (28) was used for cloning and gene expression. Plasmid vectors used were pOS103 (13) (kindly provided by Y. Koyama), pIN-III-lpp^P5-A3 (12), and pCUZ1 (23). E. coli cells were cultured in Luria-Bertani (LB) broth (22) at 30°C with ampicillin (100 μg/ml).

Molecular cloning and sequencing procedures.

Recombinant DNA techniques were performed by conventional protocols (22). DNA sequencing was performed with a PRISM Ready Reaction DyeDeoxy Terminator Cycle sequencing kit (Applied Biosystems) and an automatic sequencer (model 377; Applied Biosystems), using double-stranded plasmids as templates and universal or specific oligonucleotides as primers. PCR amplification of the bgaA gene was performed with 0.5 U of Taq DNA polymerase (Perkin-Elmer), 10 ng of plasmid pOS103, a 0.25 μM concentration of each synthetic primer, a 25 μM concentration of each deoxynucleoside triphosphate, and 2 mM MgCl₂ in the buffer recommended by the manufacturer. Amplification was achieved with 30 cycles of 0.5 min of denaturation at 95°C, 0.5 min of annealing at 55°C, and 2.5 min of polymerase extension at 72°C, plus an additional extension at 72°C for 7 min with a Perkin-Elmer thermal cycler (GeneAmp PCR System 2400). The synthetic oligonucleotides used for PCR amplification were 5′BGT (5′-GCTCTAGAGGGTATTACATATGTCGACCGTTTGTTACTACCCTTAT-3′; the XbaI and SalI restriction sites are underlined) and NEBL 1212 (M13 forward primer from New England Biolabs).

Enzymatic assays.

β-Galactosidase standard assays were performed with o-nitrophenyl-β-d-galactopyranoside (ONPG) at 70°C in Z buffer (0.1 M sodium phosphate [pH 7.0], 10 mM KCl, 1 mM MgSO₄, 5 mM mercaptoethanol) with permeabilized cells (17) or the purified enzyme. ONPG was dissolved in 0.1 M sodium phosphate buffer (pH 7.0) and used at a final concentration of 5 mM. The β-galactosidase activity of permeabilized cells was expressed in Miller units (17). One unit of purified β-galactosidase activity was defined as 1 nmol of o-nitrophenol released per min under the conditions described above. Activity on lactose was determined by quantitative analysis of glucose release, using glucose (Trinder) reagent from Sigma Diagnostics. These assays were carried out in 0.1 M sodium phosphate buffer (pH 7.0) at 70°C with 138 mM lactose as a substrate. The protein concentration was estimated by the method of Bradford (2), using bovine serum albumin as a standard, or by measuring the optical density at 280 nm (OD₂₈₀). In the latter case, the algorithm used for conversion was as follows: concentration (in micrograms per milliliter) = 0.4 × OD₂₈₀. The K_m value was calculated with the program Enzfitter (Elsevier-Biosoft).

Enzyme purification.

The ChBD-BgaA fusion protein overproduced in E. coli JM109(pBGT2) was purified on DEAE-cellulose essentially as previously described (23). Briefly, E. coli cells were grown in 0.5 liter of LB medium supplemented with ampicillin (100 μg/ml) at 30°C for 24 h with shaking, harvested by centrifugation, washed, resuspended in 20 mM sodium phosphate buffer (pH 7.0), and broken in a French pressure cell. The crude extract was clarified by centrifugation. A significant fraction of the enzyme (up to 50%, depending on the fermentation conditions) was recovered in the pellet. The total soluble protein (200 mg of protein; 380,000 U of activity) was loaded onto a DEAE-cellulose column equilibrated with 20 mM sodium phosphate buffer (pH 7.0). The column was extensively washed with 20 mM sodium phosphate buffer (pH 7.0), containing 1.5 M NaCl until no protein was detectable in the eluate by the method of Bradford (2) or by measuring the OD₂₈₀. ChBD-BgaA was eluted with 20 mM sodium phosphate buffer (pH 7.0), containing 0.1 M NaCl and 0.14 M choline hydrochloride. The protein contents of different active fractions were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (14). The fractions containing the highest protein concentration were pooled (1.5 mg of protein; 191,000 U/mg of protein). When the purified ChBD-BgaA enzyme was overloaded in SDS-PAGE, we observed faints bands of lower molecular masses (see Fig. 2) that reacted with anti-ChBD antibodies (data not shown) and hence could be retained in the matrix. These faint bands therefore appeared to be degradation products formed during fermentation or during the first step of the downstream processing, but they did not represent more than 5% of the total protein. Similar bands have also been observed in the case of the chimeric β-galactosidase from E. coli (23). This purification method showed a yield of about 75%, with a purification factor of 100-fold. Thus, we can estimate that the chimeric enzyme represented about 1% of the total soluble protein.

FIG. 2.

Analysis by SDS-PAGE of recombinant β-galactosidase. Cell extracts and purified enzyme were electrophoresed in a 10% (wt/vol) polyacrylamide gel and stained with Coomassie brilliant blue. Lane 1, molecular mass standards; lane 2, cell extract of E. coli JM109(pIN-III-lpp^P5-A3) (control); lane 3, cell extract of E. coli JM109(pBGT1); lane 4, cell extract of E. coli JM109(pBGT2); lane 5, 2 μg of purified chimeric ChBD-BgaA. Numbers at the left are molecular masses, in kilodaltons.

pH profile, optimal temperature, and enzyme thermostability.

The pH dependence of ChBD-BgaA activity was determined with citric acid-sodium phosphate buffer (pH range, 3.0 to 6.0) and sodium phosphate buffer (pH range, 7.0 to 8.0) (8). The dependence of ChBD-BgaA activity on temperature was determined by assaying aliquots of purified enzyme (380 ng) in Z buffer at temperatures ranging between 30 and 97°C. The thermal stability of ChBD-BgaA was studied by determining the remaining activity after incubating the enzyme in the absence of a substrate at 70, 80, and 90°C for different time periods. Activity assays were performed at 70°C for 10 min.

Nucleotide sequence accession number.

The GenBank accession number for the sequence reported here is Z93773.

RESULTS AND DISCUSSION

Nucleotide sequence of the β-galactosidase gene (bgaA) and its flanking regions.

To determine the structure of the β-galactosidase-encoding gene of Thermus sp. strain T2, we sequenced the whole insert of plasmid pOS103 (13). The complete nucleotide sequence had 3,129 bp and revealed the existence of four open reading frames (orf genes). We found an orf1 gene from nucleotides 1 to 275 encoding a polypeptide of 91 amino acids that appears to be the C-terminal region of a protein related to several membrane proteins involved in sugar transport. For example, the product of orf1 was highly similar to the lactose permease from Bacillus subtilis (33% identity and 68% similarity; GenBank no. Z99107) and the lactose transport system permease (LacF) from a Synechocystis sp. (30% identity and 68% similarity; GenBank no. D90905).

The orf2 gene shows an ATG start codon which overlaps the TGA stop codon of orf1 and which is preceded by a putative ribosome-binding site (RBS) (GGAGG). The deduced amino acid sequence of the protein encoded by orf2 (274 amino acids long) is highly similar to membrane proteins that are also involved in sugar transport, such as the lactose transport system permease (LacG) from Agrobacterium radiobacter (30% identity and 64% similarity; GenBank no. X66596).

The orf3 gene shows an ATG start codon located six nucleotides downstream of the TGA stop codon of orf2, and it is preceded by a putative RBS (AGGA) that overlaps the 3′ end of orf2. It encodes a protein of 645 amino acids (M_r, 73,595) having a calculated pI of 5.6. Interestingly, its deduced amino acid sequence exhibits great similarity to several β-galactosidases, such as those from Haloferax alicantei (41% identity and 65% similarity; GenBank no. U70664), Bacillus stearothermophilus (BgaI [32% identity and 58% similarity; GenBank no. P19668]), Bacillus circulans (BgaA [30% identity and 59% similarity; GenBank no. L03424] and BgaB [30% identity and 59% similarity; GenBank no. L03425]), Clostridium perfringens (product of open reading frame 80; 30% identity and 58% similarity; GenBank no. O49537), and Arthrobacter sp. strain B7 (23% identity and 42% similarity; GenBank no. U17417).

These similarities allowed us to identify orf3 as the β-galactosidase gene contained in plasmid pOS103. A multiple sequence alignment of the bgaA product (BgaA) with the mentioned β-galactosidases (data not shown) revealed that the two glutamic acid residues that are involved in the catalytic mechanism of the β-galactosidase from B. stearothermophilus (9) are also conserved in BgaA (E141 and E312). This result suggests that BgaA and the other proteins we examined could be considered new members of family 42 of glycosyl hydrolases, which has been represented only by the β-galactosidase of B. stearothermophilus so far (9). Hence, we propose the name bgaA for orf3, according to the previous nomenclature established for B. stearothermophilus (11).

Finally, there is a truncated orf4 gene downstream from the bgaA gene that starts at an ATG codon which overlaps the stop codon of bgaA, and it is preceded by the putative RBS GGAGG. The encoded polypeptide of 33 amino acids showed similarity to cellulases from Thermomonospora fusca (GenBank no. L20094) and Ruminococcus flavefaciens (GenBank no. S55178).

The overlapping organization of the four orf genes contained in the insert of plasmid pOS103 and the absence of putative long stem-loop structures that might act as transcription terminators strongly suggest that the four orf genes might constitute an operon involved in the metabolism of sugars. Nevertheless, an experimental demonstration of this organization will require further analyses.

Expression of the β-galactosidase gene in E. coli.

To ascertain the function of the bgaA product, this gene was isolated from pOS103 by PCR amplification with the primers 5′BGT and NEBL 1212, cloned in an expression vector, and expressed in E. coli JM109 (Fig. 1). The primer 5′BGT introduces a novel SalI restriction site at the 5′ end of the coding sequence of bgaA to facilitate the construction of a fusion protein (see below). The presence of this restriction site generates a mutation: whereas the wild-type β-galactosidase starts at its N terminus with the amino acid sequence MLGVCYY, the protein encoded by the amplified fragment starts with the sequence MSTVCYY (the two modified residues are underlined). The PCR product was purified, digested with the restriction enzymes XbaI and HindIII, and ligated to the E. coli expression vector pIN-III-lpp^P5-A3, which had been previously digested with the same restriction enzymes (Fig. 1). Interestingly, when the resulting plasmid, pBGT1 (10.1 kb), was transformed into E. coli JM109, we observed that the recombinant cells showed a blue phenotype in LB plates containing 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal). Moreover, E. coli JM109(pBGT1) cells were also able to grow on minimal agar plates with lactose as the sole carbon source. In agreement with these observations, permeabilized E. coli JM109(pBGT1) cells exhibited high β-galactosidase activity (38,500 Miller units) when assayed at 70°C. Moreover, by SDS-PAGE, we observed that E. coli JM109(pBGT1) cell extracts contained a novel overproduced protein of approximately 70 kDa, which corresponds to the expected size of BgaA (Fig. 2). These results strongly supported the previous assumption that the bgaA gene encodes a β-galactosidase from Thermus sp. strain T2 and demonstrated that this thermostable enzyme could be overproduced in its active form in a mesophilic microorganism.

FIG. 1.

Construction of plasmids pBGT1 and pBGT2. Triangles indicate the positions of the lpp^P-5 and lac^PO promoters. The gene encoding the β-galactosidase from Thermus sp. strain T2 (bgaA) and the region encoding the ChBD of the lytA product (chbD) are indicated. Abbreviations: Ap^R, ampicillin resistance; H, HindIII; S, SalI; X, XbaI. lacI is a lactose repressor gene.

Construction of a chimeric thermostable β-galactosidase.

To improve the biotechnological applicability of the β-galactosidase from Thermus sp. strain T2, we decided to create a tagged β-galactosidase protein which could facilitate the purification of the enzyme by a single and low-cost chromatographic step. It is well established that a major drawback of fusion approaches can be a decrease in or the complete inactivation of enzyme activity, induced either by folding problems or by the unpredictable conformational changes caused by novel interactions with the fused polypeptide. In addition, these changes could affect the multimeric structure of the proteins. Nevertheless, a method to purify the tetrameric β-galactosidase of E. coli, generating a fusion with the ChBD coded for by the 3′-terminal region of the pneumococcal autolytic amidase lytA gene, has been reported (23). Therefore, we decided to test whether this system could also be used to purify a thermostable β-galactosidase. The high affinity of the ChBD for choline or choline structural analogs, such as DEAE, allows the purification of chimeric proteins by affinity chromatography in DEAE-cellulose (24). To carry out this experiment, we constructed the plasmid pBGT2, encoding a chimeric protein made from the fusion of the ChBD to the N terminus of BgaA (ChBD-BgaA) (Fig. 1). The permeabilized E. coli JM109 cells harboring plasmid pBGT2 exhibited high β-galactosidase activity (13,900 Miller units) at 70°C. In addition, by SDS-PAGE it was observed that the E. coli JM109(pBGT2) cell extracts contained a novel protein of about 88 kDa, which fits with the expected size of the chimeric ChBD-BgaA enzyme (Fig. 2). These results suggested that the chimeric protein is produced in an active form that retains the thermostability of the native enzyme.

Biochemical characterization of the chimeric ChBD-BgaA enzyme.

We analyzed several biochemical properties of the purified protein to determine whether the chimeric ChBD-BgaA enzyme retained the properties of the native β-galactosidase. The pure ChBD-BgaA enzyme was stable for several months at 4°C or at −20°C in the presence of 10% glycerol. It has been reported that the native β-galactosidase of Thermus sp. strain T2 also remains stable through freezing and thawing and can be routinely stored at −10°C (26).

The chimeric enzyme showed a specific activity of 191,000 or 13,000 U/mg when assayed under standard conditions with ONPG or lactose, respectively, as a substrate. However, the maximum specific activity of ChBD-BgaA (435,000 U/mg) was achieved at pH 5 and 70°C with ONPG as a substrate. The specific activity determined for the ONPG hydrolysis of the partially purified wild-type β-galactosidase from Thermus sp. strain T2 (ca. 2,000 U/mg) (26) was about 2 orders of magnitude lower than that of the chimeric enzyme. Although the specific activities reported for the partially purified β-galactosidases from Thermus aquaticus YT1 (ca. 6,000 U/mg) (1) and Thermus sp. strain 4-1A (ca. 63,000 U/mg) (4) were higher than that of Thermus sp. strain T2 (26), they were still lower than that of the chimera reported in this work. Moreover, the pure β-galactosidase from the archaebacterium Sulfolobus solfataricus showed a specific activity of 116,000 U/mg on ONPG (20). All these data indicate that fusion with ChBD does not appear to alter dramatically the structure of the β-galactosidase, since it showed very high specific activity compared to the wild-type enzyme as well as to the other thermophilic β-galactosidases described. Consequently, to the best of our knowledge, the tagged β-galactosidase constructed here presents the highest specific activity so far reported.

The effect of substrate concentration on the velocity of the enzyme reaction was determined at 70°C, using ONPG at concentrations between 0.25 and 15 mM. The process followed Michaelis-Menten kinetics, with a calculated K_m value of 1.6 mM, as determined by nonlinear regression analysis. This value is comparable to the K_ms reported for the native enzyme of Thermus sp. strain T2 (K_m = 2 to 4 mM) (26) and the β-galactosidase of T. aquaticus 4-1A (K_m = 5 mM) (4) and T. aquaticus YT1 (K_m = 6 mM) (1). This result emphasizes the idea that the structure of the catalytic site probably remains unaltered in the chimeric ChBD-BgaA enzyme.

The effect of pH on ChBD-BgaA activity is shown in Fig. 3. The optimum pH is approximately 5.0 with citrate-phosphate buffer. This value and the pH activity profile match those previously reported for the native enzyme (26).

FIG. 3.

Effect of pH on activity of the purified ChBD-BgaA enzyme.

Figure 4A illustrates the relationship between temperature and activity of the chimera when assayed between 30 and 97°C. The activity increased continuously from 30 to 90°C and dropped dramatically at higher temperatures. This behavior is typical of the β-galactosidases isolated from other thermophilic microorganisms (1, 4, 26). Figure 4B is an Arrhenius plot showing two different slopes with a breakpoint around 52°C. The Arrhenius activation energies were 76 kJ/mol between 30 and 52°C and 24 kJ/mol between 52 and 90°C. These results are comparable to those previously described for the β-galactosidase of S. solfataricus, which were interpreted as indications of a conformational change in the enzyme (20). A lower value for the activation energy at higher temperatures is consistent with an effective adaptation of this enzyme to a thermophilic environment.

FIG. 4.

Dependence of ChBD-BgaA activity on temperature. (A) Effect of temperature on activity of the purified chimeric ChBD-BgaA enzyme. (B) Arrhenius plot for the ONPG hydrolysis reaction by the chimeric ChBD-BgaA enzyme.

Finally, we studied the thermal inactivation of the ChBD-BgaA enzyme. Figure 5 shows that the enzyme retained 90 and 50% of activity after 30 min of incubation at 70 and 80°C, respectively. However, the residual activity of the chimeric enzyme decreased to an undetectable level after heating for 30 min at 90°C. The thermal stability of the purified ChBD-BgaA enzyme appears to be similar to that of its partially purified wild-type counterpart (26). Surprisingly, an increase in activity was observed after 10 min of treatment at 70°C. This activation effect has been also observed in the thermostable β-galactosidase of S. solfataricus (20). Possible causes of a heat activation effect have been pointed out elsewhere (3).

FIG. 5.

Thermal stability of ChBD-BgaA at various temperatures. Enzyme solutions were incubated at the indicated temperatures. Aliquots of the enzyme were withdrawn from the incubation mixtures at the indicated times and assayed at 70°C under standard conditions.

Taken together, these results demonstrate that the β-galactosidase produced in E. coli retains its catalytic properties and thermostability. They also provide conclusive evidence that neither growth temperature nor the tag fused to the enzyme is essential for acquiring a correct conformation. Interestingly, the enzyme still maintains significant activity at 30°C, a finding that explains the blue phenotype displayed by the recombinant E. coli cells on X-Gal-containing plates and the growth in minimal medium with lactose as the sole carbon source at this temperature. This residual activity could have facilitated, several billion years ago, the moving of hyperthermophilic bacteria into cooler niches (18).

In summary, the results presented here not only provide molecular and kinetic evidence of a thermostable enzyme but also illustrate an alternative approach for producing a new thermostable tagged β-galactosidase, facilitating the development of a rapid and inexpensive method for its purification on the industrial scale. The biochemical properties of the purified chimeric enzyme do not appear to be essentially different from those of the wild-type enzyme. In addition to its basic research interest and to its potential application to the hydrolysis of lactose in dairy products (21), this enzyme could be adequate for other biotechnological applications, such as the design of integrators to quantify thermal processing (27), the synthesis of oligosaccharides (6), or the nonradioactive labeling of nucleic acids (25).