Liquid-Phase and Ultrahigh-Frequency-Acoustoﬂuidics-Based Solid-Phase Synthesis of Biotin-Tagged 6 / 3 -Sialyl- N -Acetylglucosamine by Sequential One-Pot Multienzyme System

: 6 (cid:48) / 3 (cid:48) -Sialylated N -acetyllactosamine (6 (cid:48) / 3 (cid:48) -SLN) is important for discrimination of the source (human or avian) of inﬂuenza virus strains. Biotinylated oligosaccharides have been widely used for analysis and quick detection. The development of e ﬃ cient strategies to synthesize biotin-tagged 6 (cid:48) / 3 (cid:48) -SLN have become necessary. In the current study, newly developed technology ultrahigh-frequency-acoustoﬂuidics (UHFA), which can provide a powerful source for e ﬃ cient microﬂuidic mixing, solid-phase oligosaccharide synthesis and one-pot multienzyme (OPME) system, were used to develop a new strategy for oligosaccharide synthesis. Firstly, biotinylated N-acetylglucosamine was designed and chemically synthesized through traditional approaches. Secondly, biotinylated 6 (cid:48) - and 3 (cid:48) -sialyl-N-acetylglucosamines were prepared in solution through two sequential OPME modules in with a yield of ~95%. Thirdly, 6 (cid:48) -SLN was also prepared through UHFA-based enzymatic solid-phase synthesis on magnetic beads with a yield of 64.4%. The current strategy would be potentially used for synthesis of functional oligosaccharides. modules: OPME 1 consisting of GalE and NmLgtB, OPME 2 including NmCSS and Pd2,6ST, and OPME 3 comprising NmCSS and PmST1. GalE: Escherichia coli UDP-galactose C4-epimerase; NmLgtB: Neisseria meningitides β1,4-galactosyltransferase; NmCSS: Neisseria meningitidis CMP-sialic acid synthetase; Pd2,6ST: Photobacterium damselae α2,6-sialyltransferase; PmST1 M144D: Pasteurella multocida α2,3-sialyltransferase. mg mL). reverse C 18 same an the stopped by the same volume of cold min, centrifuged rpm 30 precipitates

Compared to mammalian glycosyltransferases, bacterial ones are less sensitive to nucleotide inhibition, and show broad substrate specificity, and are available in recombinant and soluble form with high expression level. Therefore, bacterial glycosyltransferases are widely used in oligosaccharide synthesis. Considering the fact that effective and economic synthesis oligosaccharide is to combine the sugar nucleotide biosynthetic process with glycosyltransferase-catalyzed reactions, highly efficient bacterial glycosyltransferases based one-pot multienzyme (OPME) [21] methods starting from simple monosaccharides have been developed and used for synthesis of many structurally complicated oligosaccharides.
In comparison with liquid-phase oligosaccharide synthesis, solid-phase oligosaccharide synthesis (SPOS) offers advantages in several aspects: (1) only one purification step is needed in most cases at the end of the reaction; (2) unwanted reagents and side products can be easily removed by washing and filtering, and so excessive glycosyl donor can be used to ensure the high production yield. Enzymatic SPOS, combining advantages of OPME and SPOS in particular, is more desirable since it could offer a real simplification by combining the advantages of the OPME approach with those of the solid-phase method [22].
A micro-fabricated solid-mounted thin-film piezoelectric resonator (SMR) with a frequency of 1.54 GHz has been integrated into microfluidic systems. Experimental and simulation results showed that UHF (ultrahigh frequency)-SMR triggers strong acoustic field gradients to produce efficient and highly localized acoustic streaming vortices, providing a powerful source for microfluidic mixing [23]. Ultrahigh frequency (~2.5 GHz) piezoelectric resonators as acoustic micromixers were excited to produce turbulent flow in microdroplets for in situ, pumping-free, and highly efficient mixing [24]. This ultrahigh-frequency-acoustofluidics (UHFA) was successfully used for classic Diels-Alder reactions [25].
In this study, biotinylated 6 -and 3 -sialyl-N-acetyllactosamine were synthesized by the liquid-phase and SPOS-based OPME approach. Given the fact that efficient mixing is important for SPOS-based OPME synthesis, UHFA was applied for SPOS of biotinylated 6 -and 3 -sialyl-N-acetyllactosamine due to its highly localized acoustic streaming vortices.

Design and Synthesis of Biotin-Tagged N-Acetylglucosamine (Biotin-GlcNAc)
The starting biotinylated monosaccharide acceptor-biotin-tagged N-acetylglucosamine (named as Biotin-GlcNAc in this study) was designed and synthesized through well-known procedures shown in Scheme 1 (Supplementary materials). Compound 1 was synthesized from p-hydroxymethyl phenol and 1,3-dibromo propane in the presence of K 2 CO 3 with medium yield. Compound 2 was prepared from compound with large excess of NaN 3 at 95% yield. Compound 3 was synthesized through four steps: glucosamine was first reacted with o-phthalic anhydride, then followed by peracetylation by acetic anhydride and selective hydrolysis of C1-acetate, and finally trichloroacetimidate glycosyl donor compound 4 was obtained stereo specifically. Compound 4 was produced through direct coupling between compounds 2 and 3. Compound 4 was then fully deprotected and followed by selective acetylation at the amino group to give compound 5. Compound 5 was reduced and condensed with N-hydroxysuccinimide (NHS)-activated biotin to afford the final product-Biotin-GlcNAc (compound 6), which was purified by HPLC (High Performance Liquid Chromatography) (Figures S1 and S2 in Supplementary materials). Compound 6 was characterized by NMR and MS ( Figure S3 in Supplementary materials).
Until now, biotin-labeled 6′/3′-sialyl-N-acetylglucosamine was only synthesized by Zeng et al. [20]. However, galactosidase, which led to lower yield of targeted galactosylation, and rat  Until now, biotin-labeled 6 /3 -sialyl-N-acetylglucosamine was only synthesized by Zeng et al. [20]. However, galactosidase, which led to lower yield of targeted galactosylation, and rat sialyltransferases were used. Bacterial galactosyltransferase and sialyltransferases were used in the current study instead, which led to highly efficient galactosylation and sialylation, respectively.
Catalysts 2020, 10, x FOR PEER REVIEW 6 of 12 sialyltransferases were used. Bacterial galactosyltransferase and sialyltransferases were used in the current study instead, which led to highly efficient galactosylation and sialylation, respectively.
Streptavidin magnetic beads (Dyna beads) from Invitrogen were used as solid-phase carriers. GlcNAc-Biotin (compound 6) or Compound I was attached to the surface of streptavidin magnetic beads through strong interaction between biotin and streptavidin. OPME 1-and 2-catalyzed reactions were performed on surface of magnetic beads on the UHFA platform. Both reactionslasted for 24 h. After enzymatic reactions were stopped, the product (Compound I or II) was released from magnetic beads by reversibly disrupting biotin-streptavidin interaction through heating at 70 C [27] on a metal bath, and analyzed by MS, demonstrating the feasibility of UHFA-based solid-phase sequential OPME synthesis of compound I and II (Figures 2 and 3). Unfortunately, the yields of two OPMEs-catalyzed reactions were only around 58.6-64.4% (Table 1). Despite attempts with increases in amounts of enzymes, UDP-Glc, sialic acid and CTP, or reaction time, the yields of UHFA-based synthesis of compound I and II through two OPMEs were not improved too much. In addition, even washing of magnetic beads after coupling and repeated coupling with fresh enzymes and nucleotide sugars (or "double-coupling") did not lead to better yields. The results were consistent with previous ones of solid-phase-based enzymatic synthesis of oligosaccharides. Obviously, there were unreactive acceptor sites present on the surface of magnetic beads. This is a well-known phenomenon in chemical solid-phase synthesis, caused by, e.g., capping by an Scheme 3. Ultrahigh-frequency-acoustofluidics (UHFA)-based solid-phase sequential OPME synthesis of Compound I (C I).
Catalysts 2020, 10, x FOR PEER REVIEW 6 of 12 sialyltransferases were used. Bacterial galactosyltransferase and sialyltransferases were used in the current study instead, which led to highly efficient galactosylation and sialylation, respectively.
Streptavidin magnetic beads (Dyna beads) from Invitrogen were used as solid-phase carriers. GlcNAc-Biotin (compound 6) or Compound I was attached to the surface of streptavidin magnetic beads through strong interaction between biotin and streptavidin. OPME 1-and 2-catalyzed reactions were performed on surface of magnetic beads on the UHFA platform. Both reactionslasted for 24 h. After enzymatic reactions were stopped, the product (Compound I or II) was released from magnetic beads by reversibly disrupting biotin-streptavidin interaction through heating at 70 C [27] on a metal bath, and analyzed by MS, demonstrating the feasibility of UHFA-based solid-phase sequential OPME synthesis of compound I and II (Figures 2 and 3). Unfortunately, the yields of two OPMEs-catalyzed reactions were only around 58.6-64.4% (Table 1). Despite attempts with increases in amounts of enzymes, UDP-Glc, sialic acid and CTP, or reaction time, the yields of UHFA-based synthesis of compound I and II through two OPMEs were not improved too much. In addition, even washing of magnetic beads after coupling and repeated coupling with fresh enzymes and nucleotide sugars (or "double-coupling") did not lead to better yields. The results were consistent with previous ones of solid-phase-based enzymatic synthesis of oligosaccharides. Obviously, there were unreactive acceptor sites present on the surface of magnetic beads. This is a well-known phenomenon in chemical solid-phase synthesis, caused by, e.g., capping by an Scheme 4. Ultrahigh-frequency-acoustofluidics (UHFA)-based solid-phase sequential OPME synthesis of Compound II (C II).
Streptavidin magnetic beads (Dyna beads) from Invitrogen were used as solid-phase carriers. GlcNAc-Biotin (compound 6) or Compound I was attached to the surface of streptavidin magnetic beads through strong interaction between biotin and streptavidin. OPME 1-and 2-catalyzed reactions were performed on surface of magnetic beads on the UHFA platform. Both reactionslasted for 24 h. After enzymatic reactions were stopped, the product (Compound I or II) was released from magnetic beads by reversibly disrupting biotin-streptavidin interaction through heating at 70 • C [27] on a metal bath, and analyzed by MS, demonstrating the feasibility of UHFA-based solid-phase sequential OPME synthesis of compound I and II (Figures 2 and 3). Unfortunately, the yields of two OPMEs-catalyzed reactions were only around 58.6-64.4% (Table 1). Despite attempts with increases in amounts of enzymes, UDP-Glc, sialic acid and CTP, or reaction time, the yields of UHFA-based synthesis of compound I and II through two OPMEs were not improved too much. In addition, even washing of magnetic beads after coupling and repeated coupling with fresh enzymes and nucleotide sugars (or "double-coupling") did not lead to better yields. The results were consistent with previous ones of solid-phase-based enzymatic synthesis of oligosaccharides. Obviously, there were unreactive acceptor sites present on the surface of magnetic beads. This is a well-known phenomenon in chemical solid-phase synthesis, caused by, e.g., capping by an undesired chemical group during the coupling procedure or steric factors. In our case, it was reasonable to assume that the low yields were caused not by capping but rather by steric factors. The relatively short length of the linker (11 atoms, approximately Catalysts 2020, 10, 1347 7 of 12 13 Å in the most extended conformation) connecting the acceptor monosaccharide or disaccharide to biotin and the size of enzymes such as glycosyltransferases (galactosyltransferase or sialyltransferase) (diameter approximately 60 Å, assuming a globular form) should make at least some acceptor sites "unapproachable" by the enzymes or lead to steric interference between enzymes and magnetic beads and also less conformational flexibility [28]. Moreover, the binding of biotin to streptavidin on the surface of magnetic beads would make that worse [29]. In addition, we think the low yields were possibly due to heterogeneous reactions catalyzed by two OPMEs.
Catalysts 2020, 10, x FOR PEER REVIEW 7 of 12 undesired chemical group during the coupling procedure or steric factors. In our case, it was reasonable to assume that the low yields were caused not by capping but rather by steric factors. The relatively short length of the linker (11 atoms, approximately 13 Å in the most extended conformation) connecting the acceptor monosaccharide or disaccharide to biotin and the size of enzymes such as glycosyltransferases (galactosyltransferase or sialyltransferase) (diameter approximately 60 Å , assuming a globular form) should make at least some acceptor sites "unapproachable" by the enzymes or lead to steric interference between enzymes and magnetic beads and also less conformational flexibility [28]. Moreover, the binding of biotin to streptavidin on the surface of magnetic beads would make that worse [29]. In addition, we think the low yields were possibly due to heterogeneous reactions catalyzed by two OPMEs.  Continuous UHFA-based solid-phase sequential synthesis of biotinylated 6′-sialyl-N-acetylglucosamine catalyzed by OPME 1 and OPME 2 was not attempted due to the observation that yield of OPME 1-catalyzed UHFA-based solid-phase synthesis of Compound I was below 50%.
The results of UHFA-based solid-phase sequential OPME synthesis of biotinylated 6′-sialyl-N-acetylglucosamine were also compared with those of traditional solid-phase sequential OPME synthesis, which was carried out in Eppendorf tubes by using a tube rotator at 37 °C (Figures S11 and S12 in Supplementary materials). It seems that yields of UHFA-based solid-phase approach were about 15% higher than those of Eppendorf tube-based one (Table 1).

Figure 2. MS analysis of UHFA-based solid-phase sequential OPME synthesis of Compound I (C I).
Catalysts 2020, 10, x FOR PEER REVIEW 7 of 12 undesired chemical group during the coupling procedure or steric factors. In our case, it was reasonable to assume that the low yields were caused not by capping but rather by steric factors. The relatively short length of the linker (11 atoms, approximately 13 Å in the most extended conformation) connecting the acceptor monosaccharide or disaccharide to biotin and the size of enzymes such as glycosyltransferases (galactosyltransferase or sialyltransferase) (diameter approximately 60 Å , assuming a globular form) should make at least some acceptor sites "unapproachable" by the enzymes or lead to steric interference between enzymes and magnetic beads and also less conformational flexibility [28]. Moreover, the binding of biotin to streptavidin on the surface of magnetic beads would make that worse [29]. In addition, we think the low yields were possibly due to heterogeneous reactions catalyzed by two OPMEs.  Continuous UHFA-based solid-phase sequential synthesis of biotinylated 6′-sialyl-N-acetylglucosamine catalyzed by OPME 1 and OPME 2 was not attempted due to the observation that yield of OPME 1-catalyzed UHFA-based solid-phase synthesis of Compound I was below 50%.
The results of UHFA-based solid-phase sequential OPME synthesis of biotinylated 6′-sialyl-N-acetylglucosamine were also compared with those of traditional solid-phase sequential OPME synthesis, which was carried out in Eppendorf tubes by using a tube rotator at 37 °C (Figures S11 and S12 in Supplementary materials). It seems that yields of UHFA-based solid-phase approach were about 15% higher than those of Eppendorf tube-based one (Table 1). Continuous UHFA-based solid-phase sequential synthesis of biotinylated 6 -sialyl-Nacetylglucosamine catalyzed by OPME 1 and OPME 2 was not attempted due to the observation that yield of OPME 1-catalyzed UHFA-based solid-phase synthesis of Compound I was below 50%.
The results of UHFA-based solid-phase sequential OPME synthesis of biotinylated 6 -sialyl-Nacetylglucosamine were also compared with those of traditional solid-phase sequential OPME synthesis, which was carried out in Eppendorf tubes by using a tube rotator at 37 • C (Figures S11 and S12 in Supplementary materials). It seems that yields of UHFA-based solid-phase approach were about 15% higher than those of Eppendorf tube-based one (Table 1).

General Information
Unless otherwise stated, chemicals were purchased and used without further purification. Gel filtration chromatography was performed using a column (100 cm × 2.5 cm) packed with Bio-Gel P-2 Fine resins (Bio-Rad, Hercules, CA, USA). 1 H and 13 C NMR spectra were recorded on Bruker AVANCE-500 or Bruker AVANCE-400 spectrometer (Brucker, Bremen, Germany) at 25 • C. Low and high ESI (Electrospray ionization) (Thermo Fisher Scientific, Waltham, MA, USA) and MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time of Flight) (Brucker, Bremen, Germany) mass spectra were obtained at Institute of Process Engineering, Chinese Academy of Sciences.
The supernatant was concentrated, and compound III (C III) was purified by BioGel P-2 column (eluted with H 2 O). Compound III (C III)(yield: 94.5%), white solid after lyophilization. 1   3.6. Ultrahigh-Frequency-Acoustofluidics (UHFA)-Based Solid-Phase Sequential OPME Synthesis of Biotinylated 6 -sialyl-N-Acetylglucosamine The fabrication process of the hypersonic device-Solid Mount Resonator (SMR) was performed according to a published procedure [36]. The experimental set-up was very similar to the published one [25,36].
Streptavidin magnetic beads were washed 4 times with buffer 1 (PBS containing 0.01% Tween 20), and were diluted to 5 mg/mL with buffer 1 in a 2 mL Eppendorf tube. Compound 6 or compound I was added, and tube was rotated up and down for 30 min at room temperature to ensure that compound 6 or compound I would be bound to streptavidin, which is equivalent to compound 6 or compound I was immobilized onto magnetic beads. Then supernatant was separated from magnetic beads with a magnet, and magnetic beads were washed 4 times with buffer 1. Magnetic beads were then resuspended in buffer 1 containing 10 mg/mL BSA, and transferred into a plastic chamber which was immobilized onto the top of the device. The components for OPME 1 or OPME 2-catalyzed reaction as above excluding compound 6 or compound I were premixed, and added into the plastic chamber. Finally, the chamber was covered with a lid, and 200 mW power was applied to the resonator. The enzymatic reactions were lasted for 24 h at 37 • C. Magnetic beads were separated from supernatant with a magnet, and washed 4 times with PBS. Then magnetic beads were responded in 20 µL deionized water, and heated on a metal bath at 70 • C and 1000 rpm for 5 min. Finally, deionized water was separated from magnetic beads, and used for MALDI-TOF-MS analysis.
3.7. Optimization of UHFA-Based Solid-Phase Sequential OPME Synthesis of Biotinylated 6 -sialyl-N-Acetylglucosamine To improve the yields of UHFA-based solid-phase OPME synthesis of biotinylated 6 -sialyl-Nacetylglucosamine, the following conditions were tested: (1) increasing amounts of enzymes, UDP-Glc, sialic acid and CTP; (2) extending reaction time to 48 h; (3) washing magnetic beads after 24 h, and adding a new batch of enzymes and substrates.

Traditional Solid-Phase Sequential OPME Synthesis of Biotinylated 6 -sialyl-N-Acetylglucosamine
For comparison, traditional solid-phase sequential OPME synthesis of biotinylated 6 -sialyl-sacetylglucosamine was also carried out as above except OPME 1 and OPME 2-catalyzed reactions were done in 2 mL of Eppendorf tubes by using a tube rotator at 37 • C for 24 h.

Conclusions
In conclusion, using biotin-labeled N-acetylglucosamine (Biotin-GlcNAc) synthesized by chemical method as the glycosyl receptor, the one-pot multi-enzyme (OPME) synthesis strategy was successfully adopted to achieve the liquid-phase enzymatic synthesis of biotinlylated 6 /3 -sialyl-N-acetylglucosamine (6 /3 -SLN). Biotinylated 6 -sialyl-N-acetylglucosamine (6 -SLN) was also prepared through ultrahigh-frequency-acoustofluidics (UHFA)-based solid-phase sequential OPME system on magnetic beads. The biotinylated 6 /3 -sialyl-N-acetylglucosamine synthesized here would be used to identify whether influenza viruses can infect humans or the source (human or avian) of influenza virus strains. Alternatively, in order to reuse enzymes or to reduce production cost of enzymes, immobilization of glycosyltransferases and/or related enzymes, which has been successfully for oligosaccharide synthesis [37,38], could be used for synthesis of 6 /3 -SLN too. All in all, this novel UHFA-based solid-phase synthetic strategy could be potentially applied toother organic and enzymatic synthesis.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4344/10/11/1347/ s1, Synthetic procedures for biotin-tagged N-acetylglucosamine (Biotin-GlcNAc). Figure S1. Purification of biotinylated N-acetylglucosamine (Biotin-GlcNAc) by reverse HPLC. Figure S2. HPLC analysis of purified biotinylated N-acetylglucosamine (Biotin-GlcNAc). Figure S3. MS analysis of purified Biotin-GlcNAc. Figure S4. SDS PAGE analysis of purified GalE. Figure S5. SDS PAGE analysis of purified NmLgtB. Figure S6. SDS PAGE analysis of purified NmCSS. Figure S7. SDS PAGE analysis of purified Pd26ST. Figure S8. SDS PAGE analysis of purified PmST1 (M144D). Figure S9. MS analysis of purified Compound I (C I). Figure S10. HPLC analysis of OPME 2-catalyzed synthesis of compound II (C II). Figure S11. MS analysis of purified Compound II (C II). Figure S12. HPLC analysis of OPME 3-catalyzed synthesis of compound III (C III). Figure S13. MS analysis of purified Compound III (C III). Figure S14. MS analysis of traditional solid-phase sequential OPME synthesis of Compound I (C I). Figure S15. MS analysis of traditional solid-phase sequential OPME synthesis of Compound II (C II). Figure S16. 1 H NMR analysis of purified Compound I (C I). Figure S17. 13 C NMR analysis of purified Compound I (C I). Figure S18. 1 H NMR analysis of purified Compound II (C II). Figure S19. 13 C NMR analysis of purified Compound II (C II). Figure S20. 1 H NMR analysis of purified Compound III (C III). Figure S21. 13 C NMR analysis of purified Compound III (C III).