Novel β-galactosynthase–β-mannosynthase dual activity of β-galactosidase from Aspergillus oryzae uncovered using monomer sugar substrates

Abstract

A novel β-galactosynthase–β-mannosynthase dual-activity of β-galactosidase from Aspergillus oryzae was uncovered, for the first time, using free monosaccharide substrates. The enzyme successfully converted galactose and mannose monomer sugars efficiently into Gal-β(1–6)-Gal and Man-β(1–6)-Man, respectively. The discovery could potentially revolutionize chemoenzymatic glycan and non-digestible oligosaccharide (NDO) syntheses.

---

Beta galactosidases (β-gal) are enzymes known to catalyze hydrolysis of β(1–3) and β(1–4) galactosyl bonds by utilizing general acid hydrolysis.¹ These enzymes, also known as glycosyl hydrolases or lactase, belong to the general glycosyl hydrolase super family (GH-A). Members of this family are GH-1, GH-2, GH-35 and GH-42,^2,3 these sub-families are categorized either on the basis of their evolutionary relationships or based on sequence homology.^4,5 Aside from being well known in catalyzing reverse hydrolysis and trans-glycosylation reactions,^6–8 some glycosyl hydrolases were rarely reported to have activities and specificities on different substrate analogs.^9,10

Some glycosyl hydrolases have been reported to have activity towards structurally different substrates. Glycoside hydrolase family 31 has been reported to have a capacity to recognize α-glucosides and 2-deoxy derivatives,¹⁰ which rendered the change in 2-OH position of the sugar less important in substrate recognition, as in mannose and galactose.⁹

The β-gal from Aspergillus oryzae (Ao-β-gal) belongs to the GH-35 family and is often used, either as free or in immobilized form to produce a non-digestible oligosaccharide (NDO) of galactose, due to its thermostability and good yield.¹¹ Other β-gal such as Penicillium sp. (Psp-β-gal), and Trichoderma reesei (Tr-β-gal) also belong to the same glycosyl hydrolase family,³ with their respective crystal structures determined in 2004 ¹² and 2011,¹ bind and interact with the substrate galactose in almost the same manner. However, aside from galactose, the enzyme active site can also accommodate other substrates such as glycerol, isopropyl β-D-1-thiogalactopyranoside (IPTG) and 2-phenylethyl β-D-thiogalactoside (PETG).¹ The confirmed two conformations of the active site, which are both favourable in the native structure, enables the active site to slightly open or close as a consequence.¹ This open-close active site conformation was thought to be the reason that explains enzyme's recognition of various substrates.

The reported possible open–close structure rendered this particular glycosyl hydrolase family interesting to explore for possible new activities using different sugar substrates, other than its usual substrate. This could potentially lead to revolutionizing novel practical field of applications such as prebiotics and chemoenzymatic glycan synthesis by providing rare and practical source of synthetic oligosaccharides,¹³ in an exquisite stereo- and regio-selective fashion at cheaper cost, since this enzyme is cheaply available in the market.¹⁴

We reported herein, for the first time, the newly uncovered β-galactosynthase–β-mannosynthase dual activity of Ao-β-gal using free monosaccharides of galactose and mannose as substrates. The products obtained were structurally characterized using 1D and 2D NMR spectroscopy.

Our quest on finding different sugar substrates that can be recognized by Ao-β-gal was fuelled by the reported open and close conformations of Tr-β-gal active site, which was thought to be a contributing factor that gives Tr-β-gal the capacity to recognize and accommodate different substrates other than galactose.¹ The idea of possible multiple substrate specificity in Ao-β-gal is operating under the assumption that an open–close conformation of the enzyme active site might be a conserved characteristic within GH-35 sub-family and that Ao-β-gal might have multiple substrate specificity on different sugar substrate analogs.

We screened the enzyme activity using different sugar substrates such as N-acetyl-D-glucosamine, D-glucosamine, D-glucuronic acid, D-glucose, D-galactose, and D-mannose. The different substrates were co-incubated with Ao-β-gal for 24 hours at 40 °C in a phosphate buffer (50 mM KH₂PO₄, 1 mM MgCl₂, 5 mM 2-mercaptoethanol; pH 4.7). After saccharide enrichment using glycoblotting (Fig. 1) and MALDI-TOF/MS analysis, only D-galactose and D-mannose showed disaccharide products (Fig. 2) while all the other monosaccharides remained unchanged (results not shown).

Fig. 1 General protocol for glycoblotting-based high throughput analysis. The steps includes: (1) chemoselective capturing of reducing carbohydrates onto a hydrazide-functionalized bead, BlotGlyco H™, (2) washing and removal of any impurities, including denatured enzymes, and other debris, (3) on-bead trans-iminization with BOA, followed by elution with Milli Q water, and (4) quantification using MALDI-TOF/MS.

Fig. 2 Representative MALDI-TOF/MS spectra of BOA-tagged products enriched by glycoblotting method. Galactose formed galactobiose after co-incubation with Ao-β-gal for 24 hours at 40 °C (Left) and mannose formed mannobiose after co-incubation with Ao-β-gal for 24 hours at 45 °C (Right).

The usual production of galacto-oligosaccharides (GOS) described in a bunch of literatures is through utilization of lactose as the substrate.¹⁹ In this process, galactose binds to the free enzyme to make enzyme–galactose intermediate complex and then undergoes trans-galactosylation reaction with glucose or lactose as an acceptor.¹⁸ This results to the formation of GOS containing glucose in its reducing end.^17,19 It was also reported that galactose cannot bind to an enzyme–galactose complex, suggesting that galactose cannot be an acceptor for GOS production.¹⁷ Since galactose has been considered an inhibitor in galacto-oligosaccharide production,^16,17 no report, to our knowledge, has been published indicating that this enzyme can efficiently utilize free galactose in galacto-oligosaccharide synthesis.

The finding that galactose have a strong inhibitory effect on Ao-β-gal, but not when exogenously added at high concentration, during lactose hydrolysis and trans-galactosylation reaction,^16,17 draws our interest towards investigating the capacity of Ao-β-gal in recognizing free galactose monomer as its sole substrate.

In this report, galactose was co-incubated with Ao-β-gal to determine if galactose really can't bind to the Ao-β-gal–galactose complex. After enrichment using glycoblotting and MALDI-TOF/MS analyses, result revealed that D-galactose was converted to galactobiose, as indicated by the presence of a 470.187 Daltons peak that represented the sodium adduct of BOA-tagged galactobiose product (Fig. 2). The conversion of galactose to galactobiose suggested that, contrary to what was previously reported, galactose actually does bind to the Ao-β-gal-galactose complex and could be utilized by the enzyme as an acceptor in trans-galactosylation reaction. Thermodynamically, high monosaccharide concentration can cause equilibrium to shift towards condensation or reverse hydrolysis upon prolonged incubations.⁸ The production of galactobiose, after incubation of D-galactose with Ao-β-gal, was also proven to be high yielding as shown in Fig. 3. There was no appreciable amount of galactobiose formed up to 8 hours (9.49%). However, a drastic increase of almost four times (41.45%) was observed when it reached 24 hours of incubation time. This slow product formation behaviour has been reportedly observed when Ao-β-gal was co-incubated with 50 g L⁻¹ of galactose for 5 hours.¹⁹ This observation was only drawn from the 5 hour incubation period with the enzyme, but failed to account the drastic increase in the product formed after 24 hours (Fig. 3).¹⁹

Fig. 3 Amount of disaccharide products formed, in percent, after 24 hours of incubation at pH 4.7 catalyzed by free Ao-β-gal. Mannose and galactose were both 200 g L⁻¹ and incubated at 45 °C and 40 °C, respectively. The quantification was done by taking the peak area of the starting material and product peaks comparing them to the peak area of the internal standard tri-N-acetylchitotriose.¹⁵ Error bar represents ±SEM of the trials.

Isolation and purification of the galactobiose product was quite easy and straightforward. After acetylation of the mixture, peracetylated galactobiose was easily separated from the peracetylated galactose monomer in a simple flash column chromatography, due to its differential affinity, and was structurally characterized using 1D and 2D NMR.

The ¹H-NMR spectrum of the peracetylated galactobiose showed a coupling constant J = 7.94 of the anomeric proton at the nonreducing end, which is a characteristics of a β-linked glycoside (ESI Fig. S1†). The 2D ¹H–¹³C HSQC-HMBC correlation spectrum (Fig. 4) showed a strong interaction between the hydrogen attached to the C1 of galactose unit at the non-reducing end (2H1) and the hydrogens attached to the C6 of the galactose unit at the reducing end (1H6), which indicated a 1–6 linkage in peracetylated galactobiose. The results in 1D and 2D NMR structural characterization indicated that galactobiose is linked via β(1,6) glycosidic linkage.

Fig. 4 Anomeric region of the 2-D ¹H–¹³C HSQC-HMBC spectrum of galactobiose showing strong interaction between the hydrogen attached to the C1 of second galactose unit (2H1) and the hydrogen attached to the C6 of the first galactose unit (1H6). Red arrows indicate 1H6's interact with 2H1 while green arrows indicate the interaction of 2H1 with 1H6s.

Structurally, mannose is different from galactose in its hydroxyl groups in the second and fourth carbon positions. The reported open–close conformations of the active site of β-gal, from the same family, and its ability to recognize different substrates¹ prompted us to use mannose as a substrate of Ao-β-gal. After co-incubation for 24 hours and enrichment using glycoblotting, mannobiose was produced as revealed by MALDI-TOF/MS analysis (Fig. 2). The mannobiose product was confirmed by the presence of the 470.246 Daltons peak, which represented the mass of the sodium adduct of BOA-tagged mannobiose. The formation of the mannobiose was gradually increased from 6.18% to 16.8% in 12 hours and 24 hours, respectively. This finding revealed the novel mannosynthase activity of Ao-β-gal, which has not been reported.

Structural characterization using ¹H-NMR spectroscopy revealed that the peracetylated mannobiose showed very small coupling constants on both anomeric hydrogens, for non-reducing and reducing anomeric carbons, with J = 1.5 Hz and J = 1.9 Hz, respectively (ESI Fig. S2†). The small coupling constants are characteristic of mannose that makes anomeric deduction ambiguous.²⁰ The 2-D ¹H–¹³C HSQC-HMBC (Fig. 5A) correlation showed that the anomeric hydrogen attached to the C1′ of the mannose (non-reducing end) unit (H1′) has a strong interaction with the hydrogens attached to the C6 position of the mannose (reducing end) unit (H6A and H6B), suggesting that the two mannose residues were connected via 1–6 linkage. The anomeric configuration using 2D ¹H–¹H NOESY-TOCSY correlation spectrum (Fig. 5B) revealed that the hydrogen in C1 (H1′) has a strong interaction with the hydrogens in C3 (H3′) and C5 (H5′) within the mannose residue (non-reducing end), suggesting an overall β-conformation resulting to a β-(1,6) linked mannobiose.

Fig. 5 NMR spectra of the mannobiose at 600 MHz showing the (A) anomeric region of 2-D ¹H–¹³C HSQC-HMBC spectrum of mannobiose indicating strong interaction between anomeric hydrogen of C1 (H1′) and the hydrogens of C6 (H6A and H6B). Red arrows indicate the interaction of H6A and H6B with H1′. (B) Anomeric region of the 2-D ¹H–¹H NOESY-TOCSY spectrum of galactobiose revealing strong interaction between H1′ of C1 and hydrogens of C3 (H3′) and C5 (H5′), suggesting a β-(1,6) linked mannobiose. Red arrows indicate the interaction of H1′ with H3′ and H5′, blue arrows indicate the interaction of H5′ with H1′ and H3′. Green arrows indicate the interaction of H3′ with H1′ and H5′.

Although the exact interaction of mannose with the active site of Ao-β-gal is yet to be determined, one might predict that mannose recognition by this enzyme, like the other substrates, could likely be due to the possible conserved open-close conformation, which enabled this substrate to fit into the active site during open state of the site and that the mannose–enzyme complex might likely be stabilized by the mobile Trp side-chain¹ in the active site, through side-chain extension, to pack the hydrophobic side of the mannose ring. The similar regio- and stereo-chemistries of both products might suggest similar interaction of the monosaccharides in the enzyme's active site.

The new activity of this enzyme could have a very important significant application in the fields of glycan synthesis and prebiotics. This enzyme is cheaply available in the market compared to those expensive glycosyl transferases that require expensive sugar substrates for glycan syntheses. In the production of prebiotics, this enzyme utilizes a monosaccharide substrate and converting it to non-digestible oligosaccharides, this process only has the unconverted starting materials and the oligosaccharide product in the solution, which could make this process favourably economical in terms of product purification and that companies could highly save time and money.

Conclusions

The novel β-galactosynthase–β-mannosynthase dual-activity of Ao-β-gal was uncovered, for the first time, using free monosaccharides galactose and mannose. The monosaccharides were efficiently converted to Gal-β(1–6)-Gal (41.45%) and Man-β(1–6)-Man (16.18%), respectively. The regio- and stereo-chemistries of both products might suggest similar interaction of substrates with the enzyme's active site. However, the possible similar interactions of mannose and galactose with the enzyme active site may conclusively be determined through conformational and comprehensive analysis of the enzyme–substrate complex crystal structure, which we will communicate shortly. The discovery of this efficient and highly selective dual-activity of the enzyme could potentially revolutionize important applications in chemoenzymatic glycan synthesis and production of non-digestible oligosaccharides (NDOs) as prebiotics.

Acknowledgements

This work was supported by JSPS KAKENHI Grant Number 25220206, JSPS Core-to-Core Program “B. Asia-Africa Science Platform,” and a program “Innovation COE program for future drug discovery and medical care” from the Ministry of Education, Culture, Science, and Technology, Japan.

References

1. M. Maksimainen, N. Hakulinen, J. M. Kallio, T. Timoharju, O. Turunen and J. Rouvinen, J. Struct. Biol., 2011, 174, 156–163.
2. M. Hidaka, S. Fushinobu, N. Ohtsu, H. Motoshima, H. Matsuzawa, H. Shoun and T. Wakagi, J. Mol. Biol., 2002, 322, 79–91.
3. B. L. Cantarel, P. M. Coutinho, C. Rancurel, T. Bernard, V. Lombard and B. Henrissat, Nucleic Acids Res., 2009, 37, D233–D238.
4. G. Davies and B. Henrissat, Structure, 1995, 3, 853–859.
5. B. Henrissat and G. Davies, Curr. Opin. Struct. Biol., 1997, 7, 637–644.
6. H. Fujimoto, M. Miyasato, Y. Ito, T. Sasaki and K. Ajisaka, Glycoconjugate J., 1998, 15, 155–160.
7. K. Pocedicova, L. Curda, D. Misun, A. Dryakova and L. Diblikova, J. Food Eng., 2010, 99, 479–484.
8. M. Maksimainen, S. Paavilainen, N. Hakulinen and J. Rouvinen, FEBS J., 2012, 279, 1788–1798.
9. K. Yamamoto and B. G. Davis, Angew. Chem., Int. Ed., 2012, 51, 7449–7453.
10. T. Nishio, W. Hakamata, A. Kimura, S. Chiba, A. Takatsuki, R. Kawachi and T. Oku, Carbohydr. Res., 2002, 337, 629–634.
11. D. F. M. Neri, V. M. Balcão, R. S. Costa, I. C. A. P. Rocha, E. M. F. C. Ferreira, D. P. M. Torres, L. R. M. Rodrigues, J. R. L. B. Carvalho and J. A. Teixeira, Food Chem., 2009, 115, 92–99.
12. A. L. Rojas, R. A. P. Nagem, K. N. Neustroev, M. Arand, M. Adamska, E. V. Eneyskaya, A. A. Kulminskaya, R. C. Garratt, A. M. Golubev and I. Polikarpov, J. Mol. Biol., 2004, 343, 1281–1292.
13. L. X. Wang and W. Huang, Curr. Opin. Chem. Biol., 2009, 13, 592.
14. M. Faijes, M. Saura-Valls, X. Perez, M. Conti and A. Planas, Carbohydr. Res., 2006, 341, 2055.
15. T. Furukawa, M. Arai, F. Garcia-Martin, M. Amano, H. Hinou and S.-I. Nishimura, Glycoconjugate J., 2013, 30, 171–182.
16. S. Himakshi and C. Martin, Enzyme Microb. Technol., 1993, 15, 297–299.
17. C. S. Kim, E. S. Ji and D. K. Oh, Biochem. Biophys. Res. Commun., 2004, 316, 738–743.
18. C. Vera, C. Guerrero and A. Illanes, Carbohydr. Res., 2011, 346, 745–752.
19. N. Albayrak and S. T. Yang, Biotechnol. Prog., 2002, 18, 240–251.
20. W. A. Bubb, Concepts Magn. Reson., Part A, 2003, 19(1), 1–19.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental section. See DOI: 10.1039/c6ra08060j

---