- Email: [email protected]
Development of a microbioreactor for glycoconjugate synthesis
Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
Development of a microbioreactor for glycoconjugate synthesis Katsuji Haneda a, Takefumi Oishi b, Hiroshi Kimura b,c, Toshiyuki Inazu a,c,⇑ a
Department of Applied Chemistry, School of Engineering, Tokai University, Japan Department of Mechanical Engineering, School of Engineering, Tokai University, Japan c Micro/Nano Technology Center, Tokai University, 4-1-1 Kitakaname, Hiratsuka, Kanagawa 259-1292, Japan b
a r t i c l e
i n f o
Article history: Received 20 December 2017 Revised 6 March 2018 Accepted 7 March 2018 Available online 9 March 2018 Keywords: Endo-M Glycosynthase Microbioreactor Glycoconjugate synthesis Transglycosylation
a b s t r a c t A microbioreactor immobilized with a synthase-type mutant enzyme, Endo-M-N175Q (glycosynthase) of endo-b-N-acetylglucosaminidase derived from Mucor hiemalis (Endo-M), was constructed and used for glycoconjugate synthesis. The transglycosylation was performed with a reaction mixture containing an oxazoline derivative of sialo complex-type glycoside (SG), which was prepared from a sialo complex-type glycopeptide SGP derived from hen egg yolk, as a glycosyl donor and N-Fmoc-N-acetylglucosaminyl-Lasparagine [Fmoc-Asn(GlcNAc)-OH] as an acceptor. The reaction mixture was injected into a glycosynthase microbioreactor at a constant flow rate. Highly efficient and nearly stoichiometric transglycosylation occurred in the microbioreactor, and the transglycosylation product was eluted from the other end of the reactor. The glycosynthase microbioreactor was stable and could be used repeatedly for a long time. Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction Endo-b-N-acetylglucosaminidase derived from Mucor hiemalis (Endo-M) shows transglycosylation activity.1 Using the transglycosylation activity of Endo-M, a ‘‘chemoenzymatic method” for the synthesis of glycoconjugate from a natural glycosyl donor and synthetic acceptor substrates was developed2 and used for the synthesis of various complex glycoconjugates.3 As the Endo-M enzyme is originally a hydrolytic enzyme, it hydrolyzes both the glycosyl donor substrate and the transglycosylation product alongside the transglycosylation reaction; thus, the transglycosylation yield was moderate. The catalytic activity of the Endo-M enzyme is exhibited via an oxazolinium ion intermediate of N-acetylglucosamine (GlcNAc) as an active complex.4 Recently, a synthase-type mutant enzyme, Endo-M-N175Q (glycosynthase), was developed.5 This enzyme exhibits high transglycosylation activity using the oxazoline derivative of the reducing terminal GlcNAc of a glycosyl donor, but it scarcely shows hydrolytic activity. Therefore, highly efficient transglycosylation is possible using this enzyme. We intended to develop a bioprocess for glycoconjugate synthesis using a glycosynthase-immobilized microbioreactor. Immobilized glycosidases had been used for the synthesis of various oligosaccharides.6 Immobilized endoglycosidases were solely used for the de-glycosylation of glycoproteins and/or for the analysis of ⇑ Corresponding author at: Department of Applied Chemistry, School of Engineering, Tokai University, Japan https://doi.org/10.1016/j.bmc.2018.03.011 0968-0896/Ó 2018 Elsevier Ltd. All rights reserved.
oligosaccharides in glycoconjugates.7 Glycoconjugate synthesis using immobilized endoglycosidase, however, had scarcely been tried. Our concept of glycoconjugate synthesis using a glycosynthase microbioreactor is shown in Scheme 1. A complex glycoconjugate as the transglycosylation product is synthesized with a glycosynthase-immobilized microbioreactor using an N-glycan oxazoline derivative as a glycosyl donor and an appropriate acceptor substrate. In this paper, we will describe the construction of a glycosynthase-immobilized microbioreactor, and its application to a transglycosylation reaction in a dynamic flow system. 2. Materials and methods 2.1. Materials Sialo glycopeptide (SGP)8 was prepared from hen egg yolk. Sialo complex-type glycoside (SG) was prepared from SGP using Endo-M enzyme. The oxazoline derivative of SG (SG-oxazoline) was prepared from SG using CDMBI9 (2-chloro-1,3-dimethyl-1H-benzimidazol-3-ium chloride, Fushimi Pharmaceutical Co., Ltd.) as an oxazolination reagent. N-Fmoc (9-fluorenylmethyloxycarbonyl)N-acetylglucosaminyl-l-asparagine (Fmoc-Asn(GlcNAc)-OH) was synthesized as reported previously10 and prepared as a sodium salt. A carrier resin for enzyme immobilization, NHS-activated Sepharose 4 Fast Flow, was purchased from GE Healthcare Inc. Endo-M and glycosynthase (Endo-M-N175Q) were supplied by Tokyo Chemical Industry Co., Ltd.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
2093
Scheme 1. A process for neoglycoconjugate synthesis using glycosynthase microbioreactor.
2.2. Methods 2.2.1. Preparation of SG-oxazoline A sialo-complex-type glycoside SG was prepared by hydrolysis of SGP,8 a sialo complex-type glycosyl peptide derived from hen egg yolk, using Endo-M-immobilized bioreactor. The oxazoline derivative of SG (SG-oxazoline) was prepared using 2-chloro-1,3dimethyl-1H-benzimidazol-3-ium chloride (CDMBI) as reported by Noguchi et al.9 The oxazolination rates of the SG-oxazoline derivatives prepared were around 60% to 70% as analyzed by 1H NMR.
2.2.2. Preparation of glycosynthase microbioreactor Glycosynthase microbioreactor M135: N-Hydroxysuccinimideactivated Sepharose resin (NHS-activated Sepharose 4 Fast Flow, GE Healthcare, 17–0906-01) was used for enzyme immobilization. Resin (150 lL, 50% suspension in 2-propanol) was washed with cold 1 mM HCl (1 M = 1 moldm3). Then, the resin slurry (60 lL) was immediately packed into a Teflon tube (/1.6 mm 4 cm) whose ends were plugged with 5 mm-length cotton filters. A solution of glycosynthase (20 lL, 20 mU) was mixed with 20 lL of a coupling buffer CB (0.1 M, pH 7.0 phosphate buffer (PB) and 0.5 M NaCl), and then this solution was immediately injected into a Teflon tube packed with the resin. After standing for 1.5 h at room temperature, the tube was washed with CB, and this was exchanged with 100 lL of a blocking buffer BB (0.2 M 2-aminoethanol, 0.08 M, pH 7.0 PB, and 0.4 M NaCl). After standing for 0.5 h at room temperature, the tube was washed with CB, and this was exchanged with a neutral buffer NB (0.1 M, pH 7.0 PB). Then, the glycosynthase microbioreactor named M135 (Glycosynthase 20 mU, resin 60 lL) was prepared. Both ends were sealed with union seals, and it was stored in a refrigerator after exchanging with the same buffer containing 0.1% NaN3. Glycosynthase microbioreactor M161: NHS-activated Sepharose 4 Fast Flow resin 50% solution (100 lL in 2-propanol) was placed in a 0.5 mL centrifuge filter tube (SpinTrap, GE Healthcare), then washed and activated with cold 1 mM HCl. Glycosynthase
solution (40 lL, 40 mU) mixed with 20 lL of double concentration coupling buffer 2 CB (0.2 M, pH 7.0 PB, and 1 M NaCl) was placed in the SpinTrap tube packed with the resin. This resin slurry containing the enzyme was agitated by rolling for 1.5 h at room temperature. The resin was then washed with CB, exchanged with blocking buffer BB, and stood for 0.5 h. The resin was then washed with CB and neutral buffer NB. Then, this resin slurry (60 lL) immobilized with glycosynthase was packed into a Teflon tube (/1.6 mm 4 cm). Both ends of the tube were plugged with cotton filter (5 mm-length each) and sealed with union seals. The glycosynthase microbioreactor named M161 (glycosynthase 40 mU, resin 60 lL) was prepared, as shown in Fig. 1. 2.2.3. Transglycosylation reaction using a glycosynthase microbioreactor in a dynamic flow system Glycosynthase microbioreactor M135 or M161 was used for the transglycosylation reaction. Before use, it was initialized by injecting a 60 mM, pH 7.0 PB solution. A reaction mixture containing 2.45 mg (1.2 lmol as weight basis) of SG-oxazoline as a glycosyl donor, 8 lL of 50 mM Fmoc-Asn(GlcNAc)-OH (400 nmol) in DMSO as an acceptor, 12 lL of 0.2 M, pH 7.0 PB, and 20 lL of distilled water in a total volume of 40 lL was injected into the glycosynthase microbioreactor using a microsyringe and a syringe pump with a constant flow rate at room temperature (25 °C). The flow rate was set at 1 lL/min in microbioreactor M135 immobilized with 20 mU glycosynthase, or 2 lL/min in microbioreactor M161 immobilized with 40 mU glycosynthase. After injecting the reaction mixture, 60 mM, pH 7.0 PB solution (220 lL) was successively injected at the same flow rate. The effluents eluted from the other end of the microbioreactor were collected every 20 lL, and the transglycosylation product and the remaining acceptor in each effluent fraction were analyzed by HPLC. A typical reaction system is shown in Fig. 2. 2.2.4. HPLC analysis of the transglycosylation product Ten lL of the reaction mixture, which was diluted 20 times with distilled water, was applied to analytical HPLC using a
2094
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
Fig. 1. Immobilization of glycosynthase on a carrier resin and preparation of a glycosynthase microbioreactor.
Fig. 2. Transglycosylation reaction using a glycosynthase microbioreactor in a dynamic flow system.
4.6 mm 250 mm ODS column (ODS-3, GL Science Co., Ltd.) and the HPLC system equipped with a Delta 600 series interface (Waters Inc.). Elution was performed with a linear increase of 25% to 50% acetonitrile in 0.1% TFA over 15 min at a flow rate of 1 mL/min at 40 °C. A single peak of the transglycosylation product of Fmoc-Asn(GlcNAc)-OH with SG [Fmoc-Asn(SG)-OH] was found at 9.5 min in front of the remaining acceptor [Fmoc-Asn(GlcNAc)OH] at 14.8 min by monitoring at 265 nm based on the absorbance of the Fmoc moiety. The transglycosylation yield was calculated from the molar ratio of Fmoc-Asn(SG)-OH to the total, [Fmoc-Asn (SG)-OH plus residual Fmoc-Asn(GlcNAc)-OH]. 3. Results 3.1. Preparation of glycosynthase microbioreactor A microbioreactor for synthesizing the glycoconjugate was prepared with glycosynthase (Endo-M-N175Q) immobilized on an Nhydroxysuccinimide-activated Sepharose resin (NHS-activated Sepharose 4 Fast Flow, GE Healthcare), which was packed into a small Teflon tube. Two microbioreactors were prepared by chang-
ing the order of immobilization of glycosynthase on a resin and packing of carrier resin into a column. Microbioreactor M135 was prepared with 20 mU of glycosynthase by injecting the enzyme into a Teflon tube (/1.6 mm 4 cm (packed resin length 30 mm)) previously packed with 60 lL of a carrier resin. Microbioreactor M161 was prepared by packing the resin slurry (60 lL) previously immobilized with 40 mU of glycosynthase into a Teflon tube (same size) (Fig. 1). By using either procedure, the enzyme was immobilized completely, and both microbioreactors showed equally favorable transglycosylation activity as shown in the next paragraph. Either procedure for the preparation of a microbioreactor was available. The procedure used for the preparation of microbioreactor M161, however, had better performance for application, as its procedure could freely settle the amount of glycosynthase immobilized and the form or size of the reactor. Double amount of glycosynthase (40 mU) was immobilized on microbioreactor M161, and it had twice specific activity (0.66 mU/lL) compared with microbioreactor M135 (0.33 mU/lL). Then, the procedure used for the preparation of microbioreactor M161 was recommended, and several other microbioreactors were prepared using this procedure. Those microbioreactors showed good performance.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
3.2. Transglycosylation reaction using a glycosynthase microbioreactor We constructed the dynamic flow reaction system using a glycosynthase microbioreactor (Fig. 2). SG-oxazoline,9 the oxazoline derivative of a sialo-complex-type glycoside SG, which was prepared by the hydrolysis of SGP,8 a sialo complex type glycosyl peptide derived from hen egg yolk, using Endo-M-immobilized bioreactor, was used as a glycosyl donor. Fmoc-Asn(GlcNAc)-OH (sodium salt) was used as an acceptor, (Scheme 2). The reaction mixture was injected into a glycosynthase microbioreactor at a constant flow rate. The transglycosylation of SG-oxazoline to Fmoc-Asn(GlcNAc)-OH occurred in the microbioreactor to produce Fmoc-sialo-complex-type glycosyl-l-asparagine, [Fmoc-Asn(SG)OH]. A typical reaction mixture for the transglycosylation reaction was composed of 2.45 mg (1.2 lmol) of SG-oxazoline, 8 lL (400 nmol) of 50 mM Fmoc-Asn(GlcNAc)-OH in DMSO solution, 12 lL of 0.2 M, pH 7.0 PB, and 20 lL of distilled water in a total volume of 40 lL. SG-oxazoline (1.2 lmol, 3 equivalents to the acceptor on a weight basis) (oxazolination ratio of about 60%–70%) was used as a glycosyl donor. Fmoc-Asn(GlcNAc)-OH was used as an acceptor at a final concentration of 10 mM. The reaction mixture was injected into the bioreactor at a constant flow rate using a syringe
2095
pump, and then 60 mM, pH 7.0 PB was passed successively at the same flow rate. The reaction was performed at room temperature (24 °C to 25 °C). The reaction mixture eluted from the other end of the reactor was collected every 20 lL. The transglycosylation product Fmoc-Asn(SG)-OH and the remaining acceptor substrate Fmoc-Asn(GlcNAc)-OH in each eluate were analyzed by HPLC using an ODS column. The transglycosylation yield was calculated from the molar ratio of Fmoc-Asn(SG)-OH to the total of Fmoc-Asn (SG)-OH and residual Fmoc-Asn(GlcNAc)-OH. A typical result of the transglycosylation reaction is shown in Fig. 3. A (left) is the result using microbioreactor M135 at the flow rate of 1 lL/min. B (right) is that using microbioreactor M161 at the flow rate of 2 lL/min. As the specific activity of microbioreactor M161 was twice of microbioreactor M135, the twice flow rate was settled on the reaction using microbioreactor M161. In Fig. 3, the relative amounts of the transglycosylation product Fmoc-Asn(SG)-OH, the remaining acceptor Fmoc-Asn(GlcNAc)OH, and the transglycosylation yield in each eluate fraction are shown as percentages. The solid lines show their cumulative values, and the dotted lines show the differential values of each eluate fraction. The transglycosylation proceeded quite well using each microbioreactor. Almost all acceptor substrates were converted to the
Scheme 2. Transglycosylation reaction in a glycosynthase microbioreactor system using SG-oxazoline as a glycosyl donor and Fmoc-Asn(GlcNAc)-OH as an acceptor.
2096
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
Fig. 3. Transglycosylation reaction in a dynamic flow system using a glycosynthase microbioreactors M135 or M161. A: microbioreactor M135, B: microbioreactor M161. d, s: Transglycosylation rate (%); ◆, }: Product Fmoc-Asn(SG)–OH; ▲, 4: Remaining acceptor Fmoc-Asn(GlcNAc)-OH. 40 lL of a reaction mixture containing 30 mM SGoxazoline, 10 mM Fmoc-Asn(GlcNAc)-OH, 60 mM, pH 7.0 PB, and 20% DMSO were injected into microbioreactor M135 at a flow rate of 1 lL/min or into microbioreactor M161 at a flow rate of 2 lL/min at room temperature (around 25 °C). After injection of the reaction mixture, 60 mM pH 7.0 PB was passed at the same flow rate. The reaction mixture eluted from the reactor was collected and analyzed by HPLC. The relative amounts of the transglycosylation product Fmoc-Asn(SG)-OH and the remaining acceptor Fmoc-Asn(GlcNAc)-OH in the eluate fractions are shown as percentages. The transglycosylation rate indicates the percentage of the transglycosylation product to the acceptor added. Dotted lines show the differential values of each fraction, and solid lines show their cumulative values.
transglycosylation product Fmoc-Asn(SG)-OH, and the transglycosylation occurred almost quantitatively. Microbioreactor M135 was used for the establishment of the fundamental reaction conditions. As microbioreactor M161 had better performance for application, this microbioreactor was used to examine the operation conditions of a glycosynthase microbioreactor. 3.3. Several factors affecting the transglycosylation reaction using a glycosynthase microbioreactor in the dynamic flow system 3.3.1. Effect of glycosyl donor/acceptor molar equivalency In the preliminary experiment in the dynamic flow system, almost quantitative transglycosylation was achieved. The effects of several factors affecting the reaction were then examined. When the glycosyl donor to acceptor molar ratio was three equivalents on a weight basis, almost quantitative transglycosylation occurred. (Fig. 3) The effect of the glycosyl donor/acceptor (D/ A) molar equivalency on the transglycosylation was examined. The transglycosylation yields at D/A molar equivalencies of two and three on a weight basis are shown in Fig. 4. As the oxazolination rate of SG-oxazoline used in this experiment was 70%, the net D/ A molar ratios were 1.4 and 2.1. Almost quantitative transglycosylation may occur at a net D/A molar equivalency of 1.5. As the real oxazolination ratios of the SG-oxazoline preparations fluctuated around 60% to 70%, the D/A molar equivalency was set at three on a weight basis in the following experiments. 3.3.2. Effect of flow rate In the following experiments, microbioreactor M161, in which 40 mU of glycosynthase was immobilized, was used. The volume of the reaction mixture was 40 lL, and the flow rate was set at 2 lL/min as a standard condition. When the reaction mixture was passed through microbioreactor M161 at a flow rate of 2 lL/ min, the transglycosylation proceeded quite well. The retention time of the reaction mixture in the microbioreactor (volume 60 lL) was 30 min. Practically, the faster reaction may be desirable.
Fig. 4. Effect of glycosyl donor (D) versus acceptor (A) molar equivalency on the transglycosylation reaction. Transglycosylation reactions were performed under donor versus acceptor (D/A) molar equivalencies (mol/mol) of two (dotted lines) or three (solid lines) using microbioreactor M135. All lines show the cumulative values. Other reaction conditions were the same as those in Fig. 3. The symbols are the same as in Fig. 3.
Then, the flow rate was raised to 4 or 8 lL/min. The reactions at the flow rate of 2 or 4 lL/min proceeded well and almost quantitative transglycosylation occurred as shown in Fig. 5. The flow rate of 2 or 4 lL/min was favorable. At the flow rate of 4 lL/min, the retention time of the reaction mixture in the microbioreactor (15 min) was shortened to a half of that (30 min) at the flow rate of 2 lL/min. The flow rate of 4 lL/min was considered to be practically more favorable. At the flow rate of 8 lL/min, however, nearly good reactivity was achieved, but the transglycosylation yield was a little lower and the reaction tended to be unstable.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
2097
Fig. 5. Effect of the flow rate of the reaction mixture on the transglycosylation reaction. Transglycosylation reactions were performed with 40 lL of the reaction mixtures at flow rates of 2 (solid line), 4 (dotted line), or 8 lL/min (broken line) using microbioreactor M161. All lines show the cumulative values. Other reaction conditions and the symbols in the figure are the same as those in Fig. 3.
At more than this flow rate, the retention time (7.5 min) of the reaction mixture in the microbioreactor might be too short to make the reaction complete. The retention time of about 15 min was considered to be necessary for the quantitative transglycosylation in this microbioreactor. 3.3.3. Effect of volume of the reaction mixture Using 40 lL of reaction mixture, almost quantitative transglycosylation occurred under the standard reaction conditions. Then, the continuation of the reactivity of the transglycosylation reaction in the microbioreactor was examined when the reaction mixture was supplied continuously. In this experiment, a double (80 lL) or triple amount (120 lL) of the reaction mixture was injected into the reactor at the same flow rate (2 lL/min). As shown in Fig. 6, almost quantitative transglycosylation continued, and no apparent decrease in the reactivity was observed. Therefore, quantitative transglycosylation may occur as long as the reaction mixture is supplied continuously.
Fig. 6. Effect of the reaction mixture volumes on the transglycosylation reaction. Transglycosylation reactions were performed with 40 (solid line), 80 (dotted line) or 120 lL (broken line) of reaction mixture using microbioreactor M161. The flow rate was 2 lL/min. All lines show the cumulative values. Other reaction conditions and the symbols in the figure are the same as those in Fig. 3.
3.3.4. Stability of the glycosynthase microbioreactor For a bioreactor using immobilized enzyme, it is required that the reactor can be used repeatedly for a long time without loss of enzyme activity. Microbioreactor M135 was used for a series of experiments. After being used for one experiment, the microbioreactor was washed with 60 mM, pH 7.0 PB containing 0.1% NaN3 and reserved in a refrigerator. This microbioreactor was then, used repeatedly for subsequent experiments. As shown in Fig. 7, microbioreactor M135 could maintain almost full enzyme activity for one year. The glycosynthase-immobilized microbioreactor was stable and could be used repeatedly for the transglycosylation reaction. 4. Discussion A synthase-type mutant Endo-M enzyme, Endo-M-N175Q (glycosynthase), can effectively transglycosylate N-glycan oxazoline derivatives to appropriate receptors, and the highly efficient synthesis of complex glycoconjugates was possible. We then constructed a process for chemoenzymatic synthesis of complex glycoconjugates using a glycosynthase microbioreactor (Scheme 1).
Fig. 7. Stability of the glycosynthase microbioreactor. Microbioreactor M135 was used repeatedly for transglycosylation reactions for one year. The reaction conditions for monitoring the reactivity of the reactor were the same as those in Fig. 3. The reactivity of the reactor is represented by the transglycosylation yield for each reaction. The first reaction on the fourth day was performed in the static state.
2098
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
microbioreactor with only a carrier resin in which no enzyme was immobilized. When the reaction mixture was passed through this reactor, no transglycosylation occurred. Therefore, the transglycosylation reaction in the glycosynthase microbioreactor really was enzymatic and not chemical. We are planning to use this glycosynthase microbioreactor system for the synthesis of various neoglycoconjugates. Acknowledgments The authors acknowledge Tokyo Chemical Industry Co. Ltd. (TCI) for kindly providing Endo-M, Endo-M N175Q (glycosynthase), and SGP. We also thank Technology Joint Management Office in our university for technical assistance of instrumental analyses, especially Dr. Yoshiki Oda for his support of NMR measurement. A part of this work was financially supported by Manufacturing Technology Association of Biologics. Fig. 8. Comparison of the transglycosylation yields between the reactions in static and dynamic states.
For this purpose, the bioreactor requires a higher reactivity and stability. 4.1. Reactivity of glycosynthase microbioreactor In the dynamic reaction system with flow using a glycosynthase-immobilized microbioreactor (Fig. 2), a higher reactivity was achieved compared with the static reaction system without flow, where the reaction mixture that was injected into the microbioreactor was stood for 2 h at 30 °C (Fig. 8). The transglycosylation yield of the static reaction was no more than 90%. The glycosynthase-immobilized gel carrier packed into a small column (diameter 1.6 mm, resin length 30 mm) was composed of micro-sized particles (diameter 45–165 nm, mean 100 nm). In the glycosynthase microbioreactor, the reaction mixture may flow through narrow paths constructed from gaps or surfaces of gel particles performing a microfluidic state, and efficient transglycosylation can be performed. 4.2. Stability of glycosynthase microbioreactor Microbioreactor M135 was shown to be a highly stable glycosynthase-immobilized microbioreactor (Fig. 7). Another microbioreactor M161 also exhibited enough activity after ten-times of use during six month and had been highly stable. These glycosynthase microbioreactors lost scarcely any enzyme activity and could be used repeatedly for a long time. From the extreme stability of the glycosynthase microbioreactor, a suspicion arose as to whether this reaction was really enzymatic. The SG-oxazoline derivative as a glycosyl donor is a compound with a high energy level that is speculated to be the reaction intermediate.4 The transglycosylation reaction from SGoxazoline to an acceptor in the microreactor as a microfluidic state might occur chemically rather than enzymatically. We prepared a
References 1. (a) Yamamoto K, Kadowaki S, Watanabe J, Kumagai H. Biochem Biophys Res Commun. 1994;203:244–252; (b) K. Yamamoto, Endoglycosidases in: Endo, M.; Hase, S.; Yamamoto, K.; Takagaki, H. (eds.), Kodansha: Tokyo; 2006.; (c) Fairbanks AJ. Chem Soc Rev. 2017;46:5128–5146. 2. Haneda K, Inazu T, Yamamoto K, Kumagai H, Nakahara Y, Kobata A. Carbohydr Res. 1996;292:61–70. 3. (a) Mizuno M, Haneda K, Iguchi R, et al. J Am Chem Soc. 1999;121:284–290; (b) Yamamoto K, Fujimoto K, Haneda K, Mizuno M, Inazu T, Kumagai H. Carbohydr Res. 1998;305:415–422; (b) Haneda K, Inazu T, Mizuno M, et al. Biochim Biophys Acta. 2001;1526:242–248; (d) Haneda K, Tagashira M, Yoshino E, et al. Glycoconjugate J. 2004;21:377–386; (e) Haneda K, Takeuchi M, Tagashira M, et al. Carbohydr Res. 2006;341:181–190; (f) Tomabechi Y, Odate Y, Izumi R, Haneda K, Inazu T. Carbohydr Res. 2010;345:2458–2463; (g) Tomabechi Y, Inazu T. Tetrahedron Lett. 2011;52:6504–6507. 4. Fujita M, Shoda S-I, Haneda K, Inazu T, Takegawa K, Yamamoto K. Biochim Biophys Acta. 2001;1528:9–14. 5. (a) Umekawa M, Huang W, Li B, et al. J Biol Chem. 2008;283:4469–4479; (b) Umekawa M, Li C, Higashiyama T, et al. J Biol Chem. 2010;285:511–521; (c) Umekawa M, Higashiyama T, Koda Y, et al. Biochim Biophys Acta. 1800;2010:1203–1209. 6. (a) Nilsson KGI. Carbohydr Res. 1987;167:95–103; (b) Ajisaka K, Fujimoto H. Carbohydr Res. 1990;199:227–234; (c) Trincone E, Improta R, Nucci R, Rossi M, Gambacorta A. Biocatalysis. 1994;10:195–210; (d) Gibson RR, Dickinson RP, Boons G-J. J Chem Soc Perkin Trans. 1997;1:3357–3360. 7. (a) Toyo’oka T, Kurihara T, Kato M, et al. Biomed Chromatogr. 2007;21:852–860; (b) Krenkova J, Lacher NA, Svec FJ. Chromatog A. 2009;1216:3252–3259; (c) Belhaj-ben RI, Fendri A, Frikha F, Gargouri A, Belghith H. Int J Biol Macromol. 2012;51:892–900; (d) Tradtrantip L, Ratelade J, Zhang H, Verkman AS. Ann Neurol. 2013;73:77–85; (e) Ren X, Bai H, Pan Y, et al. Anal Methods. 2014;6:2518–2525; (f) Karav S, Cohen JL, Barile D, de Moura B, Juliana MLN. Biotechnol Prog. 2017;33:101–112. 8. Seko A, Koketsu M, Nishizono M, et al. Biochim Biophys Acta. 1997;1335:23–32. 9. Noguchi M, Fujieda T, Huang WC, Ishihara M, Kobayashi A, Shoda S-I. Helv Chim Acta. 2012;95:1928–1936. 10. (a) Inazu T, Kobayashi K. Synlett. 1993;869–870; (b) Mizuno M, Muramoto I, Kobayashi K, Yaginuma H, Inazu T. Synthesis. 1999;162–165.
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
Development of a microbioreactor for glycoconjugate synthesis Katsuji Haneda a, Takefumi Oishi b, Hiroshi Kimura b,c, Toshiyuki Inazu a,c,⇑ a
Department of Applied Chemistry, School of Engineering, Tokai University, Japan Department of Mechanical Engineering, School of Engineering, Tokai University, Japan c Micro/Nano Technology Center, Tokai University, 4-1-1 Kitakaname, Hiratsuka, Kanagawa 259-1292, Japan b
a r t i c l e
i n f o
Article history: Received 20 December 2017 Revised 6 March 2018 Accepted 7 March 2018 Available online 9 March 2018 Keywords: Endo-M Glycosynthase Microbioreactor Glycoconjugate synthesis Transglycosylation
a b s t r a c t A microbioreactor immobilized with a synthase-type mutant enzyme, Endo-M-N175Q (glycosynthase) of endo-b-N-acetylglucosaminidase derived from Mucor hiemalis (Endo-M), was constructed and used for glycoconjugate synthesis. The transglycosylation was performed with a reaction mixture containing an oxazoline derivative of sialo complex-type glycoside (SG), which was prepared from a sialo complex-type glycopeptide SGP derived from hen egg yolk, as a glycosyl donor and N-Fmoc-N-acetylglucosaminyl-Lasparagine [Fmoc-Asn(GlcNAc)-OH] as an acceptor. The reaction mixture was injected into a glycosynthase microbioreactor at a constant flow rate. Highly efficient and nearly stoichiometric transglycosylation occurred in the microbioreactor, and the transglycosylation product was eluted from the other end of the reactor. The glycosynthase microbioreactor was stable and could be used repeatedly for a long time. Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction Endo-b-N-acetylglucosaminidase derived from Mucor hiemalis (Endo-M) shows transglycosylation activity.1 Using the transglycosylation activity of Endo-M, a ‘‘chemoenzymatic method” for the synthesis of glycoconjugate from a natural glycosyl donor and synthetic acceptor substrates was developed2 and used for the synthesis of various complex glycoconjugates.3 As the Endo-M enzyme is originally a hydrolytic enzyme, it hydrolyzes both the glycosyl donor substrate and the transglycosylation product alongside the transglycosylation reaction; thus, the transglycosylation yield was moderate. The catalytic activity of the Endo-M enzyme is exhibited via an oxazolinium ion intermediate of N-acetylglucosamine (GlcNAc) as an active complex.4 Recently, a synthase-type mutant enzyme, Endo-M-N175Q (glycosynthase), was developed.5 This enzyme exhibits high transglycosylation activity using the oxazoline derivative of the reducing terminal GlcNAc of a glycosyl donor, but it scarcely shows hydrolytic activity. Therefore, highly efficient transglycosylation is possible using this enzyme. We intended to develop a bioprocess for glycoconjugate synthesis using a glycosynthase-immobilized microbioreactor. Immobilized glycosidases had been used for the synthesis of various oligosaccharides.6 Immobilized endoglycosidases were solely used for the de-glycosylation of glycoproteins and/or for the analysis of ⇑ Corresponding author at: Department of Applied Chemistry, School of Engineering, Tokai University, Japan https://doi.org/10.1016/j.bmc.2018.03.011 0968-0896/Ó 2018 Elsevier Ltd. All rights reserved.
oligosaccharides in glycoconjugates.7 Glycoconjugate synthesis using immobilized endoglycosidase, however, had scarcely been tried. Our concept of glycoconjugate synthesis using a glycosynthase microbioreactor is shown in Scheme 1. A complex glycoconjugate as the transglycosylation product is synthesized with a glycosynthase-immobilized microbioreactor using an N-glycan oxazoline derivative as a glycosyl donor and an appropriate acceptor substrate. In this paper, we will describe the construction of a glycosynthase-immobilized microbioreactor, and its application to a transglycosylation reaction in a dynamic flow system. 2. Materials and methods 2.1. Materials Sialo glycopeptide (SGP)8 was prepared from hen egg yolk. Sialo complex-type glycoside (SG) was prepared from SGP using Endo-M enzyme. The oxazoline derivative of SG (SG-oxazoline) was prepared from SG using CDMBI9 (2-chloro-1,3-dimethyl-1H-benzimidazol-3-ium chloride, Fushimi Pharmaceutical Co., Ltd.) as an oxazolination reagent. N-Fmoc (9-fluorenylmethyloxycarbonyl)N-acetylglucosaminyl-l-asparagine (Fmoc-Asn(GlcNAc)-OH) was synthesized as reported previously10 and prepared as a sodium salt. A carrier resin for enzyme immobilization, NHS-activated Sepharose 4 Fast Flow, was purchased from GE Healthcare Inc. Endo-M and glycosynthase (Endo-M-N175Q) were supplied by Tokyo Chemical Industry Co., Ltd.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
2093
Scheme 1. A process for neoglycoconjugate synthesis using glycosynthase microbioreactor.
2.2. Methods 2.2.1. Preparation of SG-oxazoline A sialo-complex-type glycoside SG was prepared by hydrolysis of SGP,8 a sialo complex-type glycosyl peptide derived from hen egg yolk, using Endo-M-immobilized bioreactor. The oxazoline derivative of SG (SG-oxazoline) was prepared using 2-chloro-1,3dimethyl-1H-benzimidazol-3-ium chloride (CDMBI) as reported by Noguchi et al.9 The oxazolination rates of the SG-oxazoline derivatives prepared were around 60% to 70% as analyzed by 1H NMR.
2.2.2. Preparation of glycosynthase microbioreactor Glycosynthase microbioreactor M135: N-Hydroxysuccinimideactivated Sepharose resin (NHS-activated Sepharose 4 Fast Flow, GE Healthcare, 17–0906-01) was used for enzyme immobilization. Resin (150 lL, 50% suspension in 2-propanol) was washed with cold 1 mM HCl (1 M = 1 moldm3). Then, the resin slurry (60 lL) was immediately packed into a Teflon tube (/1.6 mm 4 cm) whose ends were plugged with 5 mm-length cotton filters. A solution of glycosynthase (20 lL, 20 mU) was mixed with 20 lL of a coupling buffer CB (0.1 M, pH 7.0 phosphate buffer (PB) and 0.5 M NaCl), and then this solution was immediately injected into a Teflon tube packed with the resin. After standing for 1.5 h at room temperature, the tube was washed with CB, and this was exchanged with 100 lL of a blocking buffer BB (0.2 M 2-aminoethanol, 0.08 M, pH 7.0 PB, and 0.4 M NaCl). After standing for 0.5 h at room temperature, the tube was washed with CB, and this was exchanged with a neutral buffer NB (0.1 M, pH 7.0 PB). Then, the glycosynthase microbioreactor named M135 (Glycosynthase 20 mU, resin 60 lL) was prepared. Both ends were sealed with union seals, and it was stored in a refrigerator after exchanging with the same buffer containing 0.1% NaN3. Glycosynthase microbioreactor M161: NHS-activated Sepharose 4 Fast Flow resin 50% solution (100 lL in 2-propanol) was placed in a 0.5 mL centrifuge filter tube (SpinTrap, GE Healthcare), then washed and activated with cold 1 mM HCl. Glycosynthase
solution (40 lL, 40 mU) mixed with 20 lL of double concentration coupling buffer 2 CB (0.2 M, pH 7.0 PB, and 1 M NaCl) was placed in the SpinTrap tube packed with the resin. This resin slurry containing the enzyme was agitated by rolling for 1.5 h at room temperature. The resin was then washed with CB, exchanged with blocking buffer BB, and stood for 0.5 h. The resin was then washed with CB and neutral buffer NB. Then, this resin slurry (60 lL) immobilized with glycosynthase was packed into a Teflon tube (/1.6 mm 4 cm). Both ends of the tube were plugged with cotton filter (5 mm-length each) and sealed with union seals. The glycosynthase microbioreactor named M161 (glycosynthase 40 mU, resin 60 lL) was prepared, as shown in Fig. 1. 2.2.3. Transglycosylation reaction using a glycosynthase microbioreactor in a dynamic flow system Glycosynthase microbioreactor M135 or M161 was used for the transglycosylation reaction. Before use, it was initialized by injecting a 60 mM, pH 7.0 PB solution. A reaction mixture containing 2.45 mg (1.2 lmol as weight basis) of SG-oxazoline as a glycosyl donor, 8 lL of 50 mM Fmoc-Asn(GlcNAc)-OH (400 nmol) in DMSO as an acceptor, 12 lL of 0.2 M, pH 7.0 PB, and 20 lL of distilled water in a total volume of 40 lL was injected into the glycosynthase microbioreactor using a microsyringe and a syringe pump with a constant flow rate at room temperature (25 °C). The flow rate was set at 1 lL/min in microbioreactor M135 immobilized with 20 mU glycosynthase, or 2 lL/min in microbioreactor M161 immobilized with 40 mU glycosynthase. After injecting the reaction mixture, 60 mM, pH 7.0 PB solution (220 lL) was successively injected at the same flow rate. The effluents eluted from the other end of the microbioreactor were collected every 20 lL, and the transglycosylation product and the remaining acceptor in each effluent fraction were analyzed by HPLC. A typical reaction system is shown in Fig. 2. 2.2.4. HPLC analysis of the transglycosylation product Ten lL of the reaction mixture, which was diluted 20 times with distilled water, was applied to analytical HPLC using a
2094
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
Fig. 1. Immobilization of glycosynthase on a carrier resin and preparation of a glycosynthase microbioreactor.
Fig. 2. Transglycosylation reaction using a glycosynthase microbioreactor in a dynamic flow system.
4.6 mm 250 mm ODS column (ODS-3, GL Science Co., Ltd.) and the HPLC system equipped with a Delta 600 series interface (Waters Inc.). Elution was performed with a linear increase of 25% to 50% acetonitrile in 0.1% TFA over 15 min at a flow rate of 1 mL/min at 40 °C. A single peak of the transglycosylation product of Fmoc-Asn(GlcNAc)-OH with SG [Fmoc-Asn(SG)-OH] was found at 9.5 min in front of the remaining acceptor [Fmoc-Asn(GlcNAc)OH] at 14.8 min by monitoring at 265 nm based on the absorbance of the Fmoc moiety. The transglycosylation yield was calculated from the molar ratio of Fmoc-Asn(SG)-OH to the total, [Fmoc-Asn (SG)-OH plus residual Fmoc-Asn(GlcNAc)-OH]. 3. Results 3.1. Preparation of glycosynthase microbioreactor A microbioreactor for synthesizing the glycoconjugate was prepared with glycosynthase (Endo-M-N175Q) immobilized on an Nhydroxysuccinimide-activated Sepharose resin (NHS-activated Sepharose 4 Fast Flow, GE Healthcare), which was packed into a small Teflon tube. Two microbioreactors were prepared by chang-
ing the order of immobilization of glycosynthase on a resin and packing of carrier resin into a column. Microbioreactor M135 was prepared with 20 mU of glycosynthase by injecting the enzyme into a Teflon tube (/1.6 mm 4 cm (packed resin length 30 mm)) previously packed with 60 lL of a carrier resin. Microbioreactor M161 was prepared by packing the resin slurry (60 lL) previously immobilized with 40 mU of glycosynthase into a Teflon tube (same size) (Fig. 1). By using either procedure, the enzyme was immobilized completely, and both microbioreactors showed equally favorable transglycosylation activity as shown in the next paragraph. Either procedure for the preparation of a microbioreactor was available. The procedure used for the preparation of microbioreactor M161, however, had better performance for application, as its procedure could freely settle the amount of glycosynthase immobilized and the form or size of the reactor. Double amount of glycosynthase (40 mU) was immobilized on microbioreactor M161, and it had twice specific activity (0.66 mU/lL) compared with microbioreactor M135 (0.33 mU/lL). Then, the procedure used for the preparation of microbioreactor M161 was recommended, and several other microbioreactors were prepared using this procedure. Those microbioreactors showed good performance.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
3.2. Transglycosylation reaction using a glycosynthase microbioreactor We constructed the dynamic flow reaction system using a glycosynthase microbioreactor (Fig. 2). SG-oxazoline,9 the oxazoline derivative of a sialo-complex-type glycoside SG, which was prepared by the hydrolysis of SGP,8 a sialo complex type glycosyl peptide derived from hen egg yolk, using Endo-M-immobilized bioreactor, was used as a glycosyl donor. Fmoc-Asn(GlcNAc)-OH (sodium salt) was used as an acceptor, (Scheme 2). The reaction mixture was injected into a glycosynthase microbioreactor at a constant flow rate. The transglycosylation of SG-oxazoline to Fmoc-Asn(GlcNAc)-OH occurred in the microbioreactor to produce Fmoc-sialo-complex-type glycosyl-l-asparagine, [Fmoc-Asn(SG)OH]. A typical reaction mixture for the transglycosylation reaction was composed of 2.45 mg (1.2 lmol) of SG-oxazoline, 8 lL (400 nmol) of 50 mM Fmoc-Asn(GlcNAc)-OH in DMSO solution, 12 lL of 0.2 M, pH 7.0 PB, and 20 lL of distilled water in a total volume of 40 lL. SG-oxazoline (1.2 lmol, 3 equivalents to the acceptor on a weight basis) (oxazolination ratio of about 60%–70%) was used as a glycosyl donor. Fmoc-Asn(GlcNAc)-OH was used as an acceptor at a final concentration of 10 mM. The reaction mixture was injected into the bioreactor at a constant flow rate using a syringe
2095
pump, and then 60 mM, pH 7.0 PB was passed successively at the same flow rate. The reaction was performed at room temperature (24 °C to 25 °C). The reaction mixture eluted from the other end of the reactor was collected every 20 lL. The transglycosylation product Fmoc-Asn(SG)-OH and the remaining acceptor substrate Fmoc-Asn(GlcNAc)-OH in each eluate were analyzed by HPLC using an ODS column. The transglycosylation yield was calculated from the molar ratio of Fmoc-Asn(SG)-OH to the total of Fmoc-Asn (SG)-OH and residual Fmoc-Asn(GlcNAc)-OH. A typical result of the transglycosylation reaction is shown in Fig. 3. A (left) is the result using microbioreactor M135 at the flow rate of 1 lL/min. B (right) is that using microbioreactor M161 at the flow rate of 2 lL/min. As the specific activity of microbioreactor M161 was twice of microbioreactor M135, the twice flow rate was settled on the reaction using microbioreactor M161. In Fig. 3, the relative amounts of the transglycosylation product Fmoc-Asn(SG)-OH, the remaining acceptor Fmoc-Asn(GlcNAc)OH, and the transglycosylation yield in each eluate fraction are shown as percentages. The solid lines show their cumulative values, and the dotted lines show the differential values of each eluate fraction. The transglycosylation proceeded quite well using each microbioreactor. Almost all acceptor substrates were converted to the
Scheme 2. Transglycosylation reaction in a glycosynthase microbioreactor system using SG-oxazoline as a glycosyl donor and Fmoc-Asn(GlcNAc)-OH as an acceptor.
2096
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
Fig. 3. Transglycosylation reaction in a dynamic flow system using a glycosynthase microbioreactors M135 or M161. A: microbioreactor M135, B: microbioreactor M161. d, s: Transglycosylation rate (%); ◆, }: Product Fmoc-Asn(SG)–OH; ▲, 4: Remaining acceptor Fmoc-Asn(GlcNAc)-OH. 40 lL of a reaction mixture containing 30 mM SGoxazoline, 10 mM Fmoc-Asn(GlcNAc)-OH, 60 mM, pH 7.0 PB, and 20% DMSO were injected into microbioreactor M135 at a flow rate of 1 lL/min or into microbioreactor M161 at a flow rate of 2 lL/min at room temperature (around 25 °C). After injection of the reaction mixture, 60 mM pH 7.0 PB was passed at the same flow rate. The reaction mixture eluted from the reactor was collected and analyzed by HPLC. The relative amounts of the transglycosylation product Fmoc-Asn(SG)-OH and the remaining acceptor Fmoc-Asn(GlcNAc)-OH in the eluate fractions are shown as percentages. The transglycosylation rate indicates the percentage of the transglycosylation product to the acceptor added. Dotted lines show the differential values of each fraction, and solid lines show their cumulative values.
transglycosylation product Fmoc-Asn(SG)-OH, and the transglycosylation occurred almost quantitatively. Microbioreactor M135 was used for the establishment of the fundamental reaction conditions. As microbioreactor M161 had better performance for application, this microbioreactor was used to examine the operation conditions of a glycosynthase microbioreactor. 3.3. Several factors affecting the transglycosylation reaction using a glycosynthase microbioreactor in the dynamic flow system 3.3.1. Effect of glycosyl donor/acceptor molar equivalency In the preliminary experiment in the dynamic flow system, almost quantitative transglycosylation was achieved. The effects of several factors affecting the reaction were then examined. When the glycosyl donor to acceptor molar ratio was three equivalents on a weight basis, almost quantitative transglycosylation occurred. (Fig. 3) The effect of the glycosyl donor/acceptor (D/ A) molar equivalency on the transglycosylation was examined. The transglycosylation yields at D/A molar equivalencies of two and three on a weight basis are shown in Fig. 4. As the oxazolination rate of SG-oxazoline used in this experiment was 70%, the net D/ A molar ratios were 1.4 and 2.1. Almost quantitative transglycosylation may occur at a net D/A molar equivalency of 1.5. As the real oxazolination ratios of the SG-oxazoline preparations fluctuated around 60% to 70%, the D/A molar equivalency was set at three on a weight basis in the following experiments. 3.3.2. Effect of flow rate In the following experiments, microbioreactor M161, in which 40 mU of glycosynthase was immobilized, was used. The volume of the reaction mixture was 40 lL, and the flow rate was set at 2 lL/min as a standard condition. When the reaction mixture was passed through microbioreactor M161 at a flow rate of 2 lL/ min, the transglycosylation proceeded quite well. The retention time of the reaction mixture in the microbioreactor (volume 60 lL) was 30 min. Practically, the faster reaction may be desirable.
Fig. 4. Effect of glycosyl donor (D) versus acceptor (A) molar equivalency on the transglycosylation reaction. Transglycosylation reactions were performed under donor versus acceptor (D/A) molar equivalencies (mol/mol) of two (dotted lines) or three (solid lines) using microbioreactor M135. All lines show the cumulative values. Other reaction conditions were the same as those in Fig. 3. The symbols are the same as in Fig. 3.
Then, the flow rate was raised to 4 or 8 lL/min. The reactions at the flow rate of 2 or 4 lL/min proceeded well and almost quantitative transglycosylation occurred as shown in Fig. 5. The flow rate of 2 or 4 lL/min was favorable. At the flow rate of 4 lL/min, the retention time of the reaction mixture in the microbioreactor (15 min) was shortened to a half of that (30 min) at the flow rate of 2 lL/min. The flow rate of 4 lL/min was considered to be practically more favorable. At the flow rate of 8 lL/min, however, nearly good reactivity was achieved, but the transglycosylation yield was a little lower and the reaction tended to be unstable.
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
2097
Fig. 5. Effect of the flow rate of the reaction mixture on the transglycosylation reaction. Transglycosylation reactions were performed with 40 lL of the reaction mixtures at flow rates of 2 (solid line), 4 (dotted line), or 8 lL/min (broken line) using microbioreactor M161. All lines show the cumulative values. Other reaction conditions and the symbols in the figure are the same as those in Fig. 3.
At more than this flow rate, the retention time (7.5 min) of the reaction mixture in the microbioreactor might be too short to make the reaction complete. The retention time of about 15 min was considered to be necessary for the quantitative transglycosylation in this microbioreactor. 3.3.3. Effect of volume of the reaction mixture Using 40 lL of reaction mixture, almost quantitative transglycosylation occurred under the standard reaction conditions. Then, the continuation of the reactivity of the transglycosylation reaction in the microbioreactor was examined when the reaction mixture was supplied continuously. In this experiment, a double (80 lL) or triple amount (120 lL) of the reaction mixture was injected into the reactor at the same flow rate (2 lL/min). As shown in Fig. 6, almost quantitative transglycosylation continued, and no apparent decrease in the reactivity was observed. Therefore, quantitative transglycosylation may occur as long as the reaction mixture is supplied continuously.
Fig. 6. Effect of the reaction mixture volumes on the transglycosylation reaction. Transglycosylation reactions were performed with 40 (solid line), 80 (dotted line) or 120 lL (broken line) of reaction mixture using microbioreactor M161. The flow rate was 2 lL/min. All lines show the cumulative values. Other reaction conditions and the symbols in the figure are the same as those in Fig. 3.
3.3.4. Stability of the glycosynthase microbioreactor For a bioreactor using immobilized enzyme, it is required that the reactor can be used repeatedly for a long time without loss of enzyme activity. Microbioreactor M135 was used for a series of experiments. After being used for one experiment, the microbioreactor was washed with 60 mM, pH 7.0 PB containing 0.1% NaN3 and reserved in a refrigerator. This microbioreactor was then, used repeatedly for subsequent experiments. As shown in Fig. 7, microbioreactor M135 could maintain almost full enzyme activity for one year. The glycosynthase-immobilized microbioreactor was stable and could be used repeatedly for the transglycosylation reaction. 4. Discussion A synthase-type mutant Endo-M enzyme, Endo-M-N175Q (glycosynthase), can effectively transglycosylate N-glycan oxazoline derivatives to appropriate receptors, and the highly efficient synthesis of complex glycoconjugates was possible. We then constructed a process for chemoenzymatic synthesis of complex glycoconjugates using a glycosynthase microbioreactor (Scheme 1).
Fig. 7. Stability of the glycosynthase microbioreactor. Microbioreactor M135 was used repeatedly for transglycosylation reactions for one year. The reaction conditions for monitoring the reactivity of the reactor were the same as those in Fig. 3. The reactivity of the reactor is represented by the transglycosylation yield for each reaction. The first reaction on the fourth day was performed in the static state.
2098
K. Haneda et al. / Bioorganic & Medicinal Chemistry 26 (2018) 2092–2098
microbioreactor with only a carrier resin in which no enzyme was immobilized. When the reaction mixture was passed through this reactor, no transglycosylation occurred. Therefore, the transglycosylation reaction in the glycosynthase microbioreactor really was enzymatic and not chemical. We are planning to use this glycosynthase microbioreactor system for the synthesis of various neoglycoconjugates. Acknowledgments The authors acknowledge Tokyo Chemical Industry Co. Ltd. (TCI) for kindly providing Endo-M, Endo-M N175Q (glycosynthase), and SGP. We also thank Technology Joint Management Office in our university for technical assistance of instrumental analyses, especially Dr. Yoshiki Oda for his support of NMR measurement. A part of this work was financially supported by Manufacturing Technology Association of Biologics. Fig. 8. Comparison of the transglycosylation yields between the reactions in static and dynamic states.
For this purpose, the bioreactor requires a higher reactivity and stability. 4.1. Reactivity of glycosynthase microbioreactor In the dynamic reaction system with flow using a glycosynthase-immobilized microbioreactor (Fig. 2), a higher reactivity was achieved compared with the static reaction system without flow, where the reaction mixture that was injected into the microbioreactor was stood for 2 h at 30 °C (Fig. 8). The transglycosylation yield of the static reaction was no more than 90%. The glycosynthase-immobilized gel carrier packed into a small column (diameter 1.6 mm, resin length 30 mm) was composed of micro-sized particles (diameter 45–165 nm, mean 100 nm). In the glycosynthase microbioreactor, the reaction mixture may flow through narrow paths constructed from gaps or surfaces of gel particles performing a microfluidic state, and efficient transglycosylation can be performed. 4.2. Stability of glycosynthase microbioreactor Microbioreactor M135 was shown to be a highly stable glycosynthase-immobilized microbioreactor (Fig. 7). Another microbioreactor M161 also exhibited enough activity after ten-times of use during six month and had been highly stable. These glycosynthase microbioreactors lost scarcely any enzyme activity and could be used repeatedly for a long time. From the extreme stability of the glycosynthase microbioreactor, a suspicion arose as to whether this reaction was really enzymatic. The SG-oxazoline derivative as a glycosyl donor is a compound with a high energy level that is speculated to be the reaction intermediate.4 The transglycosylation reaction from SGoxazoline to an acceptor in the microreactor as a microfluidic state might occur chemically rather than enzymatically. We prepared a
References 1. (a) Yamamoto K, Kadowaki S, Watanabe J, Kumagai H. Biochem Biophys Res Commun. 1994;203:244–252; (b) K. Yamamoto, Endoglycosidases in: Endo, M.; Hase, S.; Yamamoto, K.; Takagaki, H. (eds.), Kodansha: Tokyo; 2006.; (c) Fairbanks AJ. Chem Soc Rev. 2017;46:5128–5146. 2. Haneda K, Inazu T, Yamamoto K, Kumagai H, Nakahara Y, Kobata A. Carbohydr Res. 1996;292:61–70. 3. (a) Mizuno M, Haneda K, Iguchi R, et al. J Am Chem Soc. 1999;121:284–290; (b) Yamamoto K, Fujimoto K, Haneda K, Mizuno M, Inazu T, Kumagai H. Carbohydr Res. 1998;305:415–422; (b) Haneda K, Inazu T, Mizuno M, et al. Biochim Biophys Acta. 2001;1526:242–248; (d) Haneda K, Tagashira M, Yoshino E, et al. Glycoconjugate J. 2004;21:377–386; (e) Haneda K, Takeuchi M, Tagashira M, et al. Carbohydr Res. 2006;341:181–190; (f) Tomabechi Y, Odate Y, Izumi R, Haneda K, Inazu T. Carbohydr Res. 2010;345:2458–2463; (g) Tomabechi Y, Inazu T. Tetrahedron Lett. 2011;52:6504–6507. 4. Fujita M, Shoda S-I, Haneda K, Inazu T, Takegawa K, Yamamoto K. Biochim Biophys Acta. 2001;1528:9–14. 5. (a) Umekawa M, Huang W, Li B, et al. J Biol Chem. 2008;283:4469–4479; (b) Umekawa M, Li C, Higashiyama T, et al. J Biol Chem. 2010;285:511–521; (c) Umekawa M, Higashiyama T, Koda Y, et al. Biochim Biophys Acta. 1800;2010:1203–1209. 6. (a) Nilsson KGI. Carbohydr Res. 1987;167:95–103; (b) Ajisaka K, Fujimoto H. Carbohydr Res. 1990;199:227–234; (c) Trincone E, Improta R, Nucci R, Rossi M, Gambacorta A. Biocatalysis. 1994;10:195–210; (d) Gibson RR, Dickinson RP, Boons G-J. J Chem Soc Perkin Trans. 1997;1:3357–3360. 7. (a) Toyo’oka T, Kurihara T, Kato M, et al. Biomed Chromatogr. 2007;21:852–860; (b) Krenkova J, Lacher NA, Svec FJ. Chromatog A. 2009;1216:3252–3259; (c) Belhaj-ben RI, Fendri A, Frikha F, Gargouri A, Belghith H. Int J Biol Macromol. 2012;51:892–900; (d) Tradtrantip L, Ratelade J, Zhang H, Verkman AS. Ann Neurol. 2013;73:77–85; (e) Ren X, Bai H, Pan Y, et al. Anal Methods. 2014;6:2518–2525; (f) Karav S, Cohen JL, Barile D, de Moura B, Juliana MLN. Biotechnol Prog. 2017;33:101–112. 8. Seko A, Koketsu M, Nishizono M, et al. Biochim Biophys Acta. 1997;1335:23–32. 9. Noguchi M, Fujieda T, Huang WC, Ishihara M, Kobayashi A, Shoda S-I. Helv Chim Acta. 2012;95:1928–1936. 10. (a) Inazu T, Kobayashi K. Synlett. 1993;869–870; (b) Mizuno M, Muramoto I, Kobayashi K, Yaginuma H, Inazu T. Synthesis. 1999;162–165.