Technical field of invention

The present invention relates to glycosylating polyphenols via biocatalysis. More specifically the present invention discloses particular mutants of the thermostable sucrose phosphorylase derived from the microorganism Thermoanaerobacterium thermosaccharolyticum which efficiently glycosylate polyphenols such as resveratrol.

Background art Polyphenols are natural products with important applications in the cosmetic, pharmaceutical and food industry (Quideau et al., 2011; Espin et al., 2007). In particular, resveratrol (3,5,4'-trihydroxystilbene) is seen as a very interesting compound, since several studies reported it to be anti-carcinogenic, antiinflammatory, anti-microbial and even to expand man's life-span (Fremont, 2000; Novelle et al., 2015; Park and Pezzuto, 2015; Biagi and Bertelli, 2015). Glycosylation of resveratrol can dramatically improve its solubility and bioavailability, but the effect differs with the type of glycosidic bond (Thuan and Sohng, 2013; Torres et al., 2011; Ozaki et al., 2011). For example, the 3-a-glucoside of resveratrol was found to be more soluble than the 3- -glucoside (Torres et al., 2011). For stereo- and regiospecific glycosylation, the advantages of biocatalysis over chemical processes are obvious (de Roode et al., 2003; Desmet et al., 2012; Bornscheuer et al., 2013).

Most glycosyl transferases require nucleotide-activated donor substrates (e.g. UDP-GIc), which are rather expensive for large-scale applications (Bungaruang et al., 2013; Breton et al., 2012; Gantt et al., 2011; Palcic, 2011; Lairson et al., 2008). However, a few enzymes are able to catalyze transglycosylation reactions starting from simple sugars like maltodextrins (e.g. CGTase) or sucrose (e.g. glucansucrase) (Desmet et al., 2012). Similarly, sucrose phosphorylase (SP, EC 2.4.1.7) can transfer the glucose moiety of sucrose to a variety of acceptors through a double displacement mechanism (Puchart, 2015; Desmet and Soetaert, 2012; Goedl et al., 2008; Aerts, Verhaeghe, Roman, Stevens, Desmet and Soetaert, 2011a). It should be noted that sucrose not only is cheap but also very reactive, with an energy content that is comparable with that of nucleotide sugars (Monsan et al., 2010). High yields can thus be obtained, as has been reported for the production of 2-O-alpha-D-glucopyranosyl-s/i-glycerol, a moisturizing agent in cosmetics commercialized under the tradename Glycoin" (Goedl et al., 2008). Unfortunately, the activity of SP on (poly)phenolic acceptors is very low, even barely detectable in most cases (Aerts, Verhaeghe, Roman, Stevens, Desmet and Soetaert, 2011b; Kitao et al., 2014; De Winter et al., 2014; De Winter et al., 2013). It has been suggested that the main problem is the enzyme's low affinity for such compounds, which cannot be compensated by increasing the acceptor concentrations because of their low solubility (De Winter et al., 2013; Aerts, Verhaeghe, Roman, Stevens, Desmet and Soetaert, 2011a). Although increased transglycosylation rates have been reported upon addition of organic solvents (De Winter et al., 2014; De Winter et al., 2013). However, saturation of the enzyme has never been achieved and productivity has generally remained too low for practical applications. With resveratrol, for example, the best result has been obtained in 20% of the ionic liquid AMMOENG™ 101, but that only generated 3 mM of resveratrol 3-alpha-glucoside (De Winter et al., 2013). Therefore, there is an urgent need to find or design an SP capable of efficiently glycosylating polyphenols such as resveratrol.

Two (thermo)stable SP enzymes have so far been identified, i.e. one in Bifidobacterium adolescentis (BaSP, Cerdobbel et al., 2011; WO2011/124538) and one in Thermoanaerobacterium thermosaccharolyticum (TtSPP, WO2014/060452). The latter enzyme actually prefers sucrose 6'- phosphate as substrate (EC 2.4.1.329) but can still efficiently use sucrose as glycosyl donor (K _m = 77 mM) (Verhaeghe et al., 2014). The latter authors further disclosed a R134A variant of TtSPP, which was shown to play a role in the binding of sucrose-6'-phosphate or fructose-6-phosphate.

Brief description of figures

Figure 1. Comparing transglycosylation activity of BaSP, TtSPP and R134A on resveratrol and its building blocks resorcinol and orcinol.

Figure 2. Acceptor promiscuity of TtSPP and R134A compared to BaSP. (A) Semi-quantitative comparison of transglycosylation of catechol, 4-methyl catechol and quercetin. (B) Qualitative survey of promiscuity on diverse mono- and polyphenolic acceptors. (C) Structural skeleton of natural polyphenols.

Figure 3. Time course of resveratrol glycosylation by variant R134A (1 g.L ¹ enzyme, 10 g.L ¹ resveratrol and 1M sucrose at pH 6.5 and 55°C) Description of invention

The present invention discloses improved variants, such as TtSPP_R134A, that can be used for the efficient glycosylation of polyphenols such as resveratrol at gram scales. Indeed, these new biocatalysts are able to (i) reach full and specific conversion of resveratrol in (ii) an aqueous system without the need for solvent additions and (iii) using sucrose as cheap and renewable glycosyl donor. However, it is clear that the variant's improved efficiency is not limited to resveratrol but is also demonstrated on several large acceptors, thereby revealing a substrate promiscuity that is useful for practical applications in various industrial sectors. Therefore, the present invention relates to an in vitro method to glycosylate a polyphenol comprising:

-contacting in vitro a sucrose phosphorylase comprising the amino acid sequence of SEQ I D N° 1 (= variant R134A of TtSPP), SEQ ID N° 2 (= variant R134V of TtSPP), SEQ I D N° 3 (= variant R134T of TtSPP), SEQ ID N° 4 (variant R134G of TtSPP) or SEQ I D N° 5 (= variant of TtSPP wherein R134 is deleted) with sucrose (or alternatively with alpha-glucose- 1-phosphate) and a polyphenol, and -glycosylating said polyphenol.

SEQ ID N° 1 is the following amino acid sequence (A134 is indicated in bold & underlined):

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSGDRGFAPITY SEIEPKFGTWYDIKKMAENFDILLDLMVNH VSRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFLARTLPYSTFKIEET GEEEKVWTTFGKTDPSEQIDLDVNSHL VREFLLEVFKTFSNFGVKIVRLDAVGYVIKKIGTSCFFVEPEIYEFLDWAKGQAASYGIE LLLEVHSQFEVQYKLAERGFLIYDFILPFTV LYTLINKSNEMLYHYLKNRPINQFTMLDCHDGIPVKPDLDGLIDTKKAKEVVDICVQRGA NLSLIYGDKYKSEDGFDVHQINCTYYS ALNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEGREINRHNYGLKEIE ESVQKNVVQRLLKLIRFRNEYEAFNGE FFIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

SEQ ID N° 2 is the following amino acid sequence (V134 is indicated in bold & underlined):

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSGDRGFAPITY SEIEPKFGTWYDIKKMAENFDILLDLMVNH VSRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFLVRTLPYSTFKIEET GEEEKVWTTFGKTDPSEQIDLDVNSHL VREFLLEVFKTFSNFGVKIVRLDAVGYVIKKIGTSCFFVEPEIYEFLDWAKGQAASYGIE LLLEVHSQFEVQYKLAERGFLIYDFILPFTV LYTLINKSNEMLYHYLKNRPINQFTMLDCHDGIPVKPDLDGLIDTKKAKEVVDICVQRGA NLSLIYGDKYKSEDGFDVHQINCTYYS ALNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEGREINRHNYGLKEIE ESVQKNVVQRLLKLIRFRNEYEAFNGE FFIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

SEQ ID N° 3 is the following amino acid sequence (T134 is indicated in bold & underlined):

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSGDRGFAPITY SEIEPKFGTWYDIKKMAENFDILLDLMVNH VSRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFLTRTLPYSTFKIEET GEEEKVWTTFGKTDPSEQIDLDVNSHL VREFLLEVFKTFSNFGVKIVRLDAVGYVIKKIGTSCFFVEPEIYEFLDWAKGQAASYGIE LLLEVHSQFEVQYKLAERGFLIYDFILPFTV LYTLINKSNEMLYHYLKNRPINQFTMLDCHDGIPVKPDLDGLIDTKKAKEVVDICVQRGA NLSLIYGDKYKSEDGFDVHQINCTYYS ALNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEGREINRHNYGLKEIE ESVQKNVVQRLLKLIRFRNEYEAFNGE FFIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

SEQ I D N° 4 is the following amino acid sequence (G134 is indicated in bold & underlined):

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSGDRGFAPITY SEIEPKFGTWYDIKKMAENFDILLDLMVNH VSRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFLGRTLPYSTFKIEET GEEEKVWTTFGKTDPSEQIDLDVNSHL VREFLLEVFKTFSNFGVKIVRLDAVGYVIKKIGTSCFFVEPEIYEFLDWAKGQAASYGIE LLLEVHSQFEVQYKLAERGFLIYDFILPFTV LYTLINKSNEMLYHYLKNRPINQFTMLDCHDGIPVKPDLDGLIDTKKAKEVVDICVQRGA NLSLIYGDKYKSEDGFDVHQINCTYYS ALNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEGREINRHNYGLKEIE ESVQKNVVQRLLKLIRFRNEYEAFNGE FFIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

SEQ ID N°5 is the following amino acid (amino acid 134 of TtSPP is deleted):

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSGDRGFAPITYSEI EPKFGTWYDIKKMAENFDILLDLMVNH VSRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFLRTLPYSTFKIEETG EEEKVWTTFGKTDPSEQIDLDVNSHLV REFLLEVFKTFSNFGVKIVRLDAVGYVIKKIGTSCFFVEPEIYEFLDWAKGQAASYGIEL LLEVHSQFEVQYKLAERGFLIYDFILPFTVL YTLINKSNEMLYHYLKNRPINQFTMLDCHDGIPVKPDLDGLIDTKKAKEVVDICVQRGAN LSLIYGDKYKSEDGFDVHQINCTYYSA LNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEGREINRHNYGLKEIEE SVQKNVVQRLLKLIRFRNEYEAFNGEF FIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

The term polyphenol' relates to hydroxylated molecules that contain at least two benzene rings. The latter polyphenols comprise the flavonoids and stil benoids. The structural formula of the latter polyphenols and specific, representatives is/are the following:

Flavones: variing hydroxyl groups at ring A and/or B

I 3 [e.g. apigenine = 4',5,7-thhydroxyflavone]

O Flavanones: flavone-structure lacking double bond in ring C Π )

[e.g. hesperetin = 3',5,7-trihydro>¾/-4'-rnethoxyflavanone] Flavonols: flavone-structure with a hydroxylated ring C (2)

[e.g. quercetin = 3,3',4',5,7-pentahydroxyflavone]

Catechins: flavone-structure lacking the keto-group (3) and

double bond ( 1 ) in ring C but holding a hydroxyl group at (2) [e.g.epicatechin = (-)-3 ,3',4',5,7-flavanpentol]

Structural skeleton of STILBENOIDS:

2 phenyl rings (A and B) connected via an ethene double bond (structure known as stilbene) being polyhydroxylated at the phenyl rings

main representative: resveratrol = frans-3 ,4' ,5-trihydroxystilbene wherein R is at least one H and/or attached glycosides.

Therefore, the present invention further specifically relates to an in vitro as indicated above method wherein said polyphenol is a stilbenoid, an in vitro method as indicated above wherein said stilbenoid is resveratrol, an in vitro method as indicated above wherein said polyphenol is a flavonoid, and an in vitro method as indicated above wherein said flavonoid is apigenine, hesperetin, quercetin or

epicatechin.

Moreover the present invention relates to an in vitro as indicated above wherein said polyphenol is a phenoxyphenol having the following structural formula:

wherein R is at least one H and/or attached

glycosides The present invention further relates to an in vitro method as described above wherein said sucrose phosphorylase is a fragment of SEQ ID N° 1, 2, 3, 4 or 5, wherein said fragment comprises, in addition to having the amino acid A,V,T or G at position 134 of SEQ ID N° 1-4, respectively, or having a deletion at position 134 as shown by SEQ ID N° 5, at least the 16 catalytic amino acids on the amino acid positions as shown in SEQ ID N° 6.

SEQ ID N° 6 is thus the amino acid sequence of TtSPP indicating in bold the 16 catalytic amino acids of TtSPP in addition to X on position 134 of SEQ ID N° 1-4, wherein X = the amino acid A,V,T or G at position 134 of SEQ ID N° 1-4 , or is deleted as shown by SEQ ID N° 5 :

MALKNKVQLITYPDSLGGNLKTLNDVLEKYFSDVFGGVHILPPFPSSG|D|RG|F|A PITYSEIEPKFGTWYDIKKMAENFDILLDLMVN |^SRRSIYFQDFLKKGRKSEYADMFITLDKLWKDGKPVKGDIEKMFL TLPYSTFKIEETGEEEKVWTT GKTDPSE§IDLDVNS HLVREFLLEVFKTFSNFGVKIV[^L[DA]VG^VIKKIGTSCFFVEPEIYEFLDWAKGQAA SYGIELLL \/|H|SQFEVQYKLAERGFLIYDFIL PFTVLYTUNKSNEMLYHYLKNRPINQFTMLDC[HD|GIPVKPDLDGLIDTKKAKEVVDIC VQRGANLSLIYGDKYKSEDGF[D|VH^|IN CTYYSALNCDDDAYLAARAIQFFTPGIPQVYYVGLLAGVNDFEAVKKTKEG|R|EINRHN YGLKEIEESVQKNVVQRLLKLIRFRNEYE AFNGEFFIEDCRKDEIRLTWKKDDKRCSLFIDLKTYKTTIDYINENGEEVKYLV

In other words, the present invention relates to an in vitro method to glycosylate a polyphenol comprising:

-contacting in vitro a sucrose phoshorylase with sucrose or alpha-glucose-1 phosphate, and, a polyphenol, and,

-glycosylating said polyphenol, wherein said sucrose phoshorylase is a fragment of SEQ ID N° 1, 2, 3, 4 or 5 wherein said fragment comprises, in addition to having the amino acid A,V,T or G at position 134 of SEQ ID N° 1-4, respectively, or having a deletion at position 134 as shown by SEQ ID N° 5, at least the following 16 catalytic amino acids of SEQ ID N° 6: D49, F52, H87, F157, Q165, R195, D197, A198, Y201, E238, H240, H295, D296, D342, Q345 and R399.

The terms 'a fragment of SEQ ID N° 1, 2, 3, 4 or 5' refer to a part of SEQ ID N° 1, 2, 3, 4 or 5 containing less amino acids than SEQ ID N° 1, 2, 3, 4 or 5 and that has 100% sequence identity with SEQ ID N° 1, 2, 3, 4 or 5.

More specifically, the present invention relates to an in vitro method to glycosylate a polyphenol comprising:

-contacting in vitro a sucrose phoshorylase with sucrose or alpha-glucose-1 phosphate, and, a polyphenol, and, -glycosylating said polyphenol, wherein said sucrose phoshorylase is a fragment of SEQ I D N° 1, 2, 3, 4 or 5 that has at least 90% sequence identity with SEQ ID N° 1, 2, 3, 4 or 5 and wherein said fragment comprises, in addition to having the amino acid A,V,T or G at position 134 of SEQ I D N° 1-4, respectively, or having a deletion at position 134 as shown by SEQ ID N° 5, at least the following 16 catalytic amino acids of SEQ I D N° 6: D49, F52, H87, F157, Q165, 195, D197, A198, Y201, E238, H240, H295, D296, D342, Q345 and R399.

The terms 'at least 90% sequence identity' refers to a sequence identity of 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ I D N° 1, 2, 3, 4 or 5.

The present invention further relates to a variant of the thermostable sucrose phosphorylase derived from the microorganism Thermoanaerobacterium thermosaccharolyticum comprising SEQ ID N° 2, 3, 4 or 5, or, comprising a fragment as defined above. Examples

Materials & Methods

Chemicals. All chemicals were of analytical grades and purchased either from Sigma-Aldrich or Carbosynth.

Cloning, expression and mutagenesis. Vector constructs for heterologous expression of BaSP- and TtSPP-gene sequences, as well as the designed variants, were inserted into the pCXP34h vector and continuously expressed in E. coli BL21 cells as previously described (Verhaeghe et al, 2014). Site-specific mutations were introduced using phosphorylated primers for whole plasmid PCRs and successive ligation. Generation of the single site-saturation library of position R134 was done likewise using a whole plasmid amplification method with N N K-degeneracy primers. Successful randomization and final plasmids were checked by sequencing.

Protein extraction and purification. Recombinant proteins were produced and heat-purified as previously described (Cerdobbel et al, 2010). Crude protein extracts for screening purposes were generated by exposing frozen cell pellets from production cultures to lysis buffer (50 mM MOPS-buffer, 50 mM Na ₂ S0 ₄ , 4 mM MgS0 ₄ , 1 mM EDTA, 1 gL ¹ lysozyme, 100 μΜ PMSF, pH 7.5) for 30 min at 37°C. Supernatants after centrifugation were assigned as crude protein extracts. Protein contents were determined according to Bradford using BSA as standard. All proteins were stored at 4°C until further usage.

Screening assays. Heat-purified proteins (for tests on different acceptors) or crude protein extracts (for library screening) were incubated to a final concentration of 1 gL ¹ with 10 gL ¹ acceptor and 1 M sucrose in 50 mM MOPS-buffer at pH 6.5. Assays were performed at 37°C in microtiterplates shaking at 200 rpm. The product amounts were evaluated after 16h by semi-quantitative TLC-analysis: 1 μΙ of each sample solution was spotted on a Silica-plate (TLC-Silica gel 60 F254, Merck Millipore) and run in a solvent system from ethylacetate:methanol:water (30:5:4, v/v/v). Plates were oxidized with 10% sulphuric acid and spot intensities were determined using ImageJ. Enzyme kinetics. Kinetics of enzyme variants on resveratrol were determined in an aqueous system containing 20% ionic liquid AMMOENG 101 (kindly provided by Evonik Industries AG) to dissolve the acceptor in a range of 0 to 200 mM. At this concentration of ionic liquid, the kinetic parameters of the enzyme for its natural substrates are not affected. Reactions were started combining 8.5 μΜ heat purified enzymes, 1 M sucrose and the respective acceptor-concentration at 55°C. To determine the initial velocity, samples were taken every 15 minutes for one hour, and inactivated by ten-fold dilution in pure ethanol. Product formation was again quantified by TLC and ImageJ analysis, using purified resveratrol 3-alpha-glucoside as TLC-standard and a respective standard curve for calculations. Kinetic parameters were calculated by non-linear regression of the Michaelis-Menten equation with SigmaPlot vl3.0. All experiments were performed in triplicate to obtain kinetic parameters as reproducible mean values ± standard errors.

Production and purification of resveratrol 3-alpha-glucoside. Up-scaling of the transglycosylation reactions on resveratrol was performed up to 400 ml volumes. Reactors were stirred continuously at 55°C for 24h, and contained 1 M sucrose, 1 mgL ¹ enzyme and 10 gL ¹ of resveratrol (suspension) in 50 mM MOPS-buffer pH 6.5. Thanks to the quantitative conversion of resveratrol, its glucoside could be simply isolated by repeated ethylacetate extractions to a purity of >90%. The solvent was evaporated and the dried product analyzed by HPLC to determine purity and yield. To verify the identity of the compound by NMR and to produce the standards (TLC, HPLC), further purification was performed by column chromatography on silica gel (EtOAC:MeOH:H ₂ 0; 30:5:4; v/v/v) as described earlier (De Winter et al, 2013). HPLC analysis. Reactions were followed for 24h. Samples were taken at several time points by inactivating aliquots at 100°C for 10 min, followed by centrifugation. Supernatants were filtered through a 0.2 μιτι polystyrol filter, diluted in DMSO and analyzed by HPLC using a polar Waters XBridge Amide column (250 x 4.6 mm, 3.5 μιτι) and an Alltech 2000ES ELSD-detector (40°C, gas flow 1.5 Lmin ^"1 ), supplied with a gradient mixture of eluent A (acetonitrile, 0.2% triehtylamine) and eluent B (0.2% triethylamine): 90% solvent A (6 min), 10-25% solvent B (stepwise, 3% min ^"1 ), 25% solvent B (20 min) and 90% solvent A (10 min). Concentrations of ingredients and respective yields were calculated according to standards of fructose, glucose and resveratrol 3-alpha-glucoside.

NMR analysis of the resveratrol 3-alpha-glucoside. The chemical structure of the purified resveratrol glycoside was inferred from the combination of ID NMR ( ^X H NMR and ¹³ C NMR) and 2D NMR (gCOSY, gHSQC and gHMBC) spectroscopic analysis; J values are given in Hz. δ _Η (400 MHz; CD ₃ OD) 7.39 (2 H, d, J = 8.6, 2'-H and 6'-H), 7.04 (1 H, d, J = 16.3, 8-H), 6.87 (1 H, d, J = 16.3, 7-H), 6.86 (1 H,s, 2-H),6.80 (2 H, d, J = 8.6, 3'-H and 5'-H), 6.62 (1 H, s, 6-H), 6.54 (1 H, s, 4-H), 5.50 (1 H, d, J = 3.5, -Η), 3.87 (1 H, dd, J = 9.7 and 8.9, 3"-H), 3.76 (2 H, m, 6"-Ha, 6"-Hb), 3.72 (1 H, ddd, J = 9.7, 4.2 and 2.5, 5"-H), 3.60 (1 H, dd, J = 9.7 and 3.6, 2"-H) and 3.48 (1 H, dd, J = 9.7 and 8.9,

4"-H).

6c(100 MHz; CD ₃ OD) 160.23 (C-3), 159.87 (c-5), 158.73 (C-4'), 141.75 (C-l), 130.63 (C-l'), 130.25 (C- 8), 129.18 (C-2' and C-6'), 127.01 (C-7), 116.83 (C-3' and C-5'), 108.70 (C-6), 107.85 (C-2), 104.88 (C- 4), 99.61 (C-l"), 75.28 (C-3"), 74.63 (C-5"), 73.64 (C-2"), 71.81 (C-4"), 62.63 (C-6"). According to the COSY experiment, the proton spectrum displays a -CH=CH- group, AA'BB' and AMX spin systems of para-disubstituted and 1,3,5-trisubstituted aromatic rings, and one saccharide unit. The set of extracted vicinal constants (d, J = 3.5 Hz) identified a-glucose, while the magnitude of - _Η , 8- _Η (16.3 Hz) proved the trans-conformation of resveratrol. The position of the glycosylation was confirmed by the HMBC contact of the C-3 with 1"-H.

Results

Two (thermo)stable SP enzymes have so far been identified, i.e. one in Bifidobacterium adolescentis (BaSP) and one in Thermoanaerobacterium thermosaccharolyticum (TtSPP). ^[9] The latter actually prefers sucrose 6'-phosphate as substrate (EC 2.4.1.329) but can still efficiently use sucrose as glycosyl donor (Km = 77 mM). ^[9bl Neither of these enzymes display significant activity on resveratrol or other polyphenolic compounds as acceptor.

We tested the corresponding R134A variant for transglycosylation activity on resveratrol and smaller building blocks (Figure 1). Effectively, the gradual decrease in activity from resorcinol > orcinol > resveratrol observed with the wild-type enzyme was almost balanced with the variant, which retained more than half of its activity on resveratrol. Site-saturation mutagenesis of position 134 did not yield variants that were better than R134A, but other small amino acids like valine and threonine also offered increased activities compared to the wild-type arginine. This confirmed our hypothesis that a bulky side chain limits binding of large acceptors in the active site of SP. However, this only was valid for position 134, since mutating other residues at the active-site entrance (e.g. K301 or H344) to alanine did not have a similar effect.

Interestingly, the unbolted R134 variants enabled us to determine kinetic parameters for the glycosylation of resveratrol (Table 1), which has never been possible with the wild-type enzyme. These measurements revealed that the affinity for resveratrol is quite reasonable, with K _m -values in the range of 50-200 mM. As expected, the improvements in accessibility strongly influence the k _cat -value, which gradually increases to 2.4 s ¹ when smaller side chains are introduced at position 134. The highest catalytic efficiency (12.8 M ¹ s ^"1 ) is observed with variant R134A, which thus is the enzyme of choice for application in glycoside synthesis.

Table 1. Kinetic parameters of TtSPP variants for resveratrol

glycosylation' ³¹ enzyme K _m [mM] kcajs ¹ ] k _C at/K _m [M ¹ s ^"1 ] 134A 185 ± 32 2.36 ± 0.25 12.8 ± 2.6

R134V 110 ± 9 0.45 ± 0.02 4.1 ± 0.4

R134T 56 ± 8 0.08 ± 0.004 1.4 ± 0.2

^[a] Using as co-substrate 1M sucrose at pH 6.5 and 55°C

The improved accessibility of the active site should be not exclusive for resveratrol but should also be noticeable with other large acceptors. Therefore, variant R134A was tested on a variety of (poly)phenols to evaluate its full promiscuity and potential as biocatalyst. Similar to the meta- benzenediol series (resorcinol - orcinol - resveratrol) shown in Figure 1, a gradual decrease in activity with the ori/jo-benzenediol series (catechol > 4-methyl catechol > quercetin) was again observed for the wildtype enzymes (Figure 2 A). Also in this case, TtSPP showed higher transglycosylation activities than BaSP on the smaller compounds, but neither had activity on quercetin, a bulky representative of the flavonol class. In contrast, significant amounts of transglycosylation product could be detected with variant R134A, which confirms the enzyme's improved ability to bind larger acceptors thanks to its unbolted active site.

When comparing these three enzymes on more phenolic acceptors, the increased promiscuity of R134A towards large molecules became even more obvious (Figure 2 B). Indeed, the latter displayed activity on structures from all five major classes of flavonoids (apigenin, epicatechin, quercetin and hesperetin) but also on artificial polyphenols like phenoxyphenols (Figure 2 C) in contrast to TtSPP and BaSP, respectively. This screening experiment demonstrates that variant R134A is a promiscuous biocatalyst that can be used for the glycosylation of several polyphenolic compounds of industrial relevance. Hence, this new enzyme is a very promising target for further engineering as it has opened up the potential to discover and exploit novel identities, properties and applications of glycosylated products.

The feasibility to use variant R134A as biocatalyst for the glycosylation of resveratrol in an aqueous system was then demonstrated by mixing the enzyme and 1 M sucrose with a saturated suspension of 10 g.L ¹ resveratrol. After 24h incubation at 55°C, complete conversion of the acceptor was achieved (Figure 3) and 17.1 ± 0.6 g. L ¹ of resveratrol 3-alpha-glucoside was produced. The structure of the product was determined by NM R spectroscopy (see supplementary information) to confirm the strict regioselectivity of the enzyme. This production process not only is selective and efficient, but can also be easily implemented at larger, commercial scales.

References

Aerts, D., Verhaeghe, T. F., Roman, B. I., Stevens, C.V., Desmet, T. and Soetaert, W. (2011b).

Transglucosylation potential of six sucrose phosphorylases toward different classes of acceptors.

Carbohydr. Res.346:1860-1867.

Biagi, M. and Bertelli, A.A.E. (2015). Wine, alcohol and pills: What future for the French paradox? Life

Sci.

Bornscheuer, U.T., Huisman, G.W., Kazlauskas, R.J., Lutz, S., Moore, J.C. and Robins, K. (2013).

Engineering the third wave of biocatalysis. Nature.485:185-194.

Breton, C, Fournel-Gigleux, S. and Palcic, M. M. (2012). Recent structures, evolution and mechanisms of glycosyltransferases. Curr. Opin. Struct. Biol.1-10.

Bungaruang, L, Gutmann, A. and Nidetzky, B. (2013). Leloir Glycosyltransferases and Natural Product

Glycosylation: Biocatalytic Synthesis of the C-Glucoside Nothofagin, a Major Antioxidant of

Redbush Herbal Tea. Adv. Synth. Catal.355:2757-2763.

Cerdobbel A, De Winter K, Desmet T, Soetaert W (2010) Sucrose phosphorylase as cross-linked enzyme aggregate: Improved thermal stability for industrial applications Biotechnol J, 5:1-6.

Cerdobbel, A., De Winter, K., Aerts, D., Kuipers, R., Joosten, H.J., Soetaert, W. and Desmet, T. (2011).

Protein Eng., Des. Sel.24:829-834.

Chaloupkova, R., Sykorova, J., Prokop, Z., Jesenska, A., Monincova, M., Pavlova, M., Tsuda, M., Nagata,

Y. and Damborsky, J. (2003). Modification of Activity and Specificity of Haloalkane Dehalogenase from Sphingomonas paucimobilis UT26 by Engineering of Its Entrance Tunnel. J. Biol. Chem.

278:52622-52628.

de Roode, B. M., Franssen, M.C.R., van der Padt, A. and Boom, R. M. (2003). Perspectives for the Industrial Enzymatic Production of Glycosides. Biotechnol. Prog.19:1391-1402.

De Winter, K., Desmet, T., Devlamynck, T., Van Renterghem, L, Verhaeghe, T., Pelantova, H., Kren, V. and Soetaert, W. (2014). Biphasic Catalysis with Disaccharide Phosphorylases: Chemoenzymatic Synthesis of a- d-Glucosides Using Sucrose Phosphorylase. Org. Process Res. Dev.18:781-787.

De Winter, K., Verlinden, K., Kren, V., Weignerova, L, Soetaert, W. and Desmet, T. (2013). Green Chem.

15:1949-1955.

Desmet, T. and Soetaert, W. (2012). Process Biochemistry. Process Biochem.47:11-17. Desmet, T., Soetaert, W., Bojarova, P., Kfen, V., Dijkhuizen, L, Eastwick-Field, V. and Schiller, A. (2012).

Enzymatic Glycosylation of Small Molecules: Challenging Substrates Require Tailored Catalysts.

Chem. Eur. J.18:10786-10801.

Espin, J.C., Garcia-Conesa, M.T. and Tomas-Barberan, F.A. (2007). Nutraceuticals: Facts and fiction.

Phytochemistry.68:2986-3008.

Fremont (2000). Biological Effects of Resveratrol. Life Sci.66:663-673.

Gantt, R.W., Peltier-Pain, P., Cournoyer, W.J. and Thorson, J.S. (2011). Using simple donors to drive the equilibria of glycosyltransferase-catalyzed reactions. Nat. Chem. Biol.7:685-691.

Goedl, C, Sawangwan, T., Mueller, M., Schwarz, A. and Nidetzky, B. (2008). A High-Yielding Biocatalytic

Process for the Production of 2- O-(a-D-glucopyranosyl)- sn-glycerol, a Natural Osmolyte and

Useful Moisturizing Ingredient. Angew. Chem. Int. Ed.47:10086-10089.

Gora, A., Brezovsky, J. and Damborsky, J. (2013). Gates of Enzymes. Chem. Rev.113:5871-5923. Kingsley, L.J. and Lill, M.A. (2015). Substrate tunnels in enzymes: Structure-function relationships and computational methodology. Proteins: Struct., Funct., Bioinf.83:599-611.

Kitao, S., Ariga, T., Matsudo, T. and Sekine, H. (2014). The Syntheses of Catechin-glucosides by

Transglycosylation with Leuconostoc mesenteroidesSucrose Phosphorylase. Biosci., Biotechnol.,

Biochem.57:2010-2015.

Lairson, L. L., Henrissat, B., Davies, G.J. and Withers, S.G. (2008). Glycosyltransferases: Structures,

Functions, and Mechanisms. Annu. Rev. Biochem.77:521-555.

Maldonado OS, Lucas R, Comelles F, Gonzalez MJ, Parra J L, Medinac I, Morales JC (2011). Synthesis and characterization of phenolic antioxidants with surfactant properties: glucosyl- and glucuronosyl alkyl gallates. Tetrahedron 67: 7268-7279.

Mirza, O., Skov, L. K., Sprogoe, D., van den Broek, L.A. M., Beldman, G., Kastrup, J.S. and Gajhede, M.

(2006). Structural Rearrangements of Sucrose Phosphorylase from Bifidobacterium adolescentis during Sucrose Conversion. J. Biol. Chem.281:35576-35584.

Monsan, P., Remaud-Simeon, M. and Andre, I. (2010). Transglucosidases as efficient tools for oligosaccharide and glucoconjugate synthesis. Curr. Opin. Microbiol.13:293-300.

Nestl, B. M. and Hauer, B. (2014). Engineering of Flexible Loops in Enzymes. ACS Catal.4:3201-3211. Novelle, M.G., Wahl, D., Dieguez, C, Bernier, M. and de Cabo, R. (2015). Resveratrol supplementation:

Where are we now and where should we go? Ageing Res. Rev.21:1-15.

Ozaki, S.-l., Imai, H., Iwakiri, T., Sato, T., Shimoda, K., Nakayama, T. and Hamada, H. (2011).

Regioselective glucosidation of trans-resveratrol in Escherichia coli expressing glucosyltransferase from Phytolacca americana. Biotechnol. Lett.34:475-481.

Palcic, M. M. (2011). Glycosyltransferases as biocatalysts. Curr. Opin. Chem. Biol.1-8.

Park, E.-J. and Pezzuto, J. M. (2015). The pharmacology of resveratrol in animals and humans. Biochim. Biophys. Acta, Mol. Basis Dis. 1852:1071-1113.

Puchart, V. (2015). Glycoside phosphorylases: Structure, catalytic properties and biotechnological potential. Biotechnol. Adv.1-16.

Quideau, S., Deffieux, D., Douat-Casassus, C. and Pouysegu, L. (2011). Plant Polyphenols: Chemical

Properties, Biological Activities, and Synthesis. Angew. Chem. Int. Ed. 50:586-621.

Sprog0e, D., van den Broek, L.A.M., Mirza, O., Kastrup, J.S., Voragen, A.G.J. , Gajhede, M. and Skov, L.K.

(2004). Crystal Structure of Sucrose Phosphorylase from Bifidobacterium adolescentis ^† .

Biochemistry. 43:1156-1162.

Thuan, N.H. and Sohng, J.K. (2013). Recent biotechnological progress in enzymatic synthesis of glycosides. J. Ind. Microbiol. Biotechnol. 40:1329-1356.

Torres, P., Poveda, A., Jimenez-Barbero, J., Parra, J.L, Comelles, F., Ballesteros, A.O. and Plou, F.J.

(2011). Enzymatic Synthesis of a-Glucosides of Resveratrol with Surfactant Activity. Adv. Synth.

Catal. 353:1077-1086.

Verhaeghe, T., Aerts, D., Diricks, M., Soetaert, W. and Desmet, T. (2014). The quest for a thermostable sucrose phosphorylase reveals sucrose 6'-phosphate phosphorylase as a novel specificity. Appl. Microbiol. Biotechnol. 98:7027-7037.