**Purification of Marine Bacterial Sialyltransferases and Sialyloligosaccharides**

Toshiki Mine and Takeshi Yamamoto *Glycotechnology Business Unit, Japan Tobacco Inc., Japan* 

#### **1. Introduction**

68 Chromatography and Its Applications

Talasaz, A. H.; Abbasi, M. R.; Abkhiz, S. & Dahti-Khavidaki, S. (2010). *Tribulus terrestris*-

Toque, H. A.; Teixeira, C. E.; Lorenzetti, R.; Okuyama, C. E.; Antunes, E. & De Nucci, G. (2008).

Venhuis, B. J.; Zomer, G.; Hamzink, M.; Meiring, H. D.; Aubin, Y. & Kaste, D. (2011). The

Wang, J.; Jiang, Y.; Wang, Y.; Zhao, X.; Cui, Y. & Gu, J. (2007). Liquid chromatography tandem

Wollein, U.; Eisenreich, W. & Schramek, N. (2011). Identification of novel sildenafil-analogues in an adulterated herbal food supplement, *J. Pharm. Biomed. Anal.* 56:705-712. Zou, P.; Hou, P.; Low, M. Y. & Koh, H. L. (2006). Structural elucidation of a tadalafil analogue found as an adulterant of a herbal product, *Food Addit. Contam*. 23:446-451. Zou, P.; Hou, P.; Oh, S. S. Y.; Chong, Y. M.; Bloodworth, B. C.; Low, M. Y. & Koh, H. L. (2008).

supplement: A Viagra with a pop, *J. Pharm. Biomed. Anal.* 54:735-741.

3792-3793.

*Journal of Pharmacology*, 591:189-195.

plasma, *J. Pharm. Biomed. Anal.* 44:231-235.

supplements, *J. Pharm. Biomed. Anal.* 47:279-284.

induced severe nephrotoxicity in a young healthy male, *Nephrol Dial. Transplant*. 25:

Pharmacological characterization of a novel phosphodiesterase type 5 (PDE5) inhibitor lodenafil carbonate on human and rabbit corpus cavernosum, *European* 

identification of a nitrosated prodrug of the PDE-5 inhibitor aildenafil in a dietary

mass spectrometry assay to determine the pharmacokinetics of aildenafil in human

Isolation and identification of thiohomosildenafil and thiosildenafil in health

Sialic acids are important components of carbohydrate chains and are usually found at the terminal position of the carbohydrate moiety of glycoconjugates (Angata & Varki, 2002; Schauer, 2004). Sialyloligosaccharides of glycoconjugates play important roles in many biological processes (Gagneux & Varki, 1999; Varki, 1993). The transfer of sialic acids to carbohydrate chains is performed by specific sialyltransferases in the cell (Angata & Varki, 2002; Vimr et al., 2004). Thus, sialyltransferases are considered to be key enzymes in the biosynthesis of sialylated glycoconjugates. Detailed investigations of the biological functions of sialylated glycoconjugates require an abundant supply of the target compounds. To date, many sialyltransferases, and the genes encoding them, have been isolated from various sources including mammalian, bacterial, and viral sources (Schauer, 2004; Sujino et al., 2000; Yamamoto et al., 2006). During our research, we have isolated over 20 bacteria that produce sialyltransferase and have revealed the characteristics of these enzymes (Kajiwara et al., 2009; Yamamoto, 2010). In this chapter, we will introduce our research activities focusing on methods for (1) screening bacteria for glycosyltransferase activity; (2) purifying native sialyltransferases from marine bacteria; and (3) synthesizing and purifying sialyloligosaccharides produced by marine bacterial sialyltransferases.

Sialic acid is a family of acidic monosaccharides comprising over 50 naturally occurring derivatives of neuraminic acid (5-amino-3,5-dideoxy-D-*glycero*-D-*galacto*-2-nonulosonic acid or Neu) (Angata & Varki, 2002; Vimr et al., 2004). Structurally, sialic acid is one of the more complicated naturally occurring monosaccharides and is based on a skeleton of nine carbons (Schauer, 2004). *N*-acetylneuraminic acid (Neu5Ac), *N*-glycolylneuraminic acid (Neu5Gc), and 2-keto-3-deoxy-D-*glycero*-D-*galacto*-nonulosonic acid (deaminoneuraminic acid, KDN), are the three most common members of this family (Angata & Varki, 2002; Schauer, 2004). The structure of Neu, Neu5Ac, Neu5Gc and KDN are shown in Figure 1. Although sialic acid is widely distributed in higher animals and some classes of microorganisms, only Neu5Ac is ubiquitous (Angata & Varki, 2004). Usually, sialic acid exists in the carbohydrate moiety of glycoconjugates, including glycoproteins and glycolipids, and is linked to the terminal positions of the carbohydrate chains of the glycoconjugates. Many studies have been carried out to clarify the structure-function relationship of carbohydrate chains containing sialic acid. These studies have revealed that Neu5Ac is the most common sialic acid component of carbohydrate chains and sialylated carbohydrate chains of

Purification of Marine Bacterial Sialyltransferases and Sialyloligosaccharides 71

sialyloligosaccharides relies on sialyloligosaccharide synthesis by chemical or enzymatic methods. Although many methods for the synthesis of sialyloligosaccharides including chemical- and enzyme-based methods using glycosyltransferases have been developed, it is still difficult to synthesize large amounts of sialyloligosaccharides. Therefore, only a limited amount and only a few kinds of sialyloligosaccharides are currently available as research reagents. In this chapter, we introduce our results with regard to methods for screening for bacterial glycosyltransferases, purification of native sialyltransferases from bacteria, and the synthesis and purification of sialyloligosaccharides produced by marine bacterial sialyltransferases. The methodologies introduced in this chapter might to be useful for screening of bacteria that produce other types of glycosyltransferases, purification of

Samples of seawater, sea-sand, mud, seaweed, and small animals including various kinds of fishes and shells, were collected from various coastal locations in Japan. Bacteria that grew on marine agar 2216 or nutrient agar (Becton-Dickinson, Franklin Lakes, NJ, USA) that was supplemented with 2% NaCl at 15°C, 25°C, or 30°C were isolated from the samples. Aliquots of the bacteria were suspended in 10% glycerol and stored at −80°C. For each bacterial isolate, 6 mL of marine broth 2216 (Becton-Dickinson) in a 15-mL test tube was inoculated with bacteria and cultivated at 15°C, 25°C, or 30°C for 18 h on a rotary shaker (180 rpm). After the cultivation, bacteria were harvested from 2 to 4 mL of the culture broth by centrifugation and then suspended in 200 L of 20 mM sodium cacodylate buffer (pH 6.0) that contained 0.2% Triton X-100, lysed by sonication on ice, and measured immediately for sialyltransferase activity. Sialyltransferase activity was confirmed as follows: the reaction mixture (30 L) consisted of the bacterial lysate as the sample of enzyme, 120 mM lactose, 2.3 mM CMP-Neu5Ac (Nakarai Tesque, Kyoto, Japan), 4620 Bq CMP-[4,5,6,7,8,9-14C]- Neu5Ac (Amersham Biosciences, Little Chalfont, UK), 100 mM Bis–Tris buffer (pH 6.0), 0.5 M NaCl, and 0.03% Triton X-100. The reaction was carried out at 25°C for 2 h. The reaction mixture was then diluted with 5 mM sodium phosphate buffer (pH 6.8) to a final volume of 2 mL, and applied to a column (0.5 × 2 cm) of Dowex-1 × 8 (phosphate form, Bio-Rad Laboratories, Hercules, CA, USA). The eluate (2 mL) was collected directly into a scintillation vial for counting. The radioactivity of [4, 5, 6, 7, 8, 9-14C]-Neu5Ac that had transferred to the acceptor substrate in the eluate was measured by using a liquid scintillation counter, and the amount of Neu5Ac transferred was calculated. Unreacted CMP-Neu5Ac was not eluted in this buffer concentration. Using this procedure, we have isolated many bacteria that possess sialyltransferase activity. Many of the marine bacteria that produced sialyltransferases were classified in genus *Photobacterium* or the closely related genus *Vibrio*. For instance, *Photobacterium phosphoreum* JT-ISH-467 that showed 2,3 sialyltransferase activity was isolated from the outer skin of Japanese common squid, *Todarodes pacificus* (Tsukamoto et al. 2007); *Photobacterium damselae* JT0160 that expressed 2,6-sialyltransferase activity was isolated from seawater (Yamamoto et al., 1998); *Photobacterium* sp. JT-ISH-224 that contained both 2,3- and 2,6-sialyltransferase activities was isolated from the gut of Japanese barracuda, *Sphyraena pinguis* (Tsukamoto et al., 2008); and *Photobacterium leiognathi* JT-SHIZ-145 that expressed 2,6-sialyltransferase activity was isolated from the outer skin of Japanese squid, *Loliolus japonica* (Yamamoto et al., 2007).

membrane-binding proteins, and purification of oligosaccharides.

**2. Screening bacteria for sialyltransferase activity** 

**2.1 Basic screening method** 

glycoconjugates play significant roles in many biological processes including inflammation, glycoprotein clearance from circulation, cell-cell recognition, cancer metastasis, and virus infection (Kannagi, 2002; Paulson, 1989). Sialyltransferases commonly transfer Neu5Ac from cytidine 5'-monophospho-*N*-acetylneuraminic acid (CMP-Neu5Ac) to various acceptor substrates (Angata & Varki, 2002). Thus, sialyltransferases are thought to be one of the important enzymes in the biosynthesis of sialylated glycoconjugates.

(A) Neuraminic acid (Neu), (B) *N*-acetylneuraminic acid (Neu5Ac), (C) *N*-glycolylneuraminic acid (Neu5Gc), (D) deaminoneuraminic acid (KDN).

Fig. 1. Structures of sialic acids.

Among the biological phenomena described above, the relationship between the carbohydrate chain structure of the host cell and host cell recognition by influenza virus is one of the best investigated (Suzuki, 2005; Weis et al., 1988). Many reports have shown that influenza A and B viruses bind via viral hemagglutinin to host cell surface receptors that are Neu5Ac- or Neu5Gc-linked glycoproteins or glycolipids (Suzuki, 2005). Furthermore, these influenza viruses also recognize the carbohydrate chain structure of the host cell (Connor et al., 1994). Confirming evidence has shown that avian influenza viruses recognize Neu5Ac2-3Gal1-3/4GlcNAc structures, and that human influenza viruses recognize Neu5Ac2-6Gal1-3/4GlcNAc structures (Connor et al., 1994; Suzuki, 2005). The host cell specificities of the influenza A and B viruses are determined mainly by the linkage of Neu5Ac or Neu5Gc to the penultimate galactose residues and core structure of the host glycoproteins or glycolipids. For this reason, the distribution of Neu5Ac and Neu5Gc and their linkage patterns on the host cell surface are important determinants of host tropism.

A large variety of oligosaccharides exist in nature. For example, many kinds of sialyloligosaccharides, such as 3′-sialyllactose, 6′-sialyllactose, and sialyllacto-*N*-neotetraose, are contained in milk of various animals (Kunz et al., 2000); however, the purification and isolation of sialyloligosaccharides from natural sources is very difficult due to their structural complexity. Therefore, the research use and development of drugs that depend on sialyloligosaccharides relies on sialyloligosaccharide synthesis by chemical or enzymatic methods. Although many methods for the synthesis of sialyloligosaccharides including chemical- and enzyme-based methods using glycosyltransferases have been developed, it is still difficult to synthesize large amounts of sialyloligosaccharides. Therefore, only a limited amount and only a few kinds of sialyloligosaccharides are currently available as research reagents. In this chapter, we introduce our results with regard to methods for screening for bacterial glycosyltransferases, purification of native sialyltransferases from bacteria, and the synthesis and purification of sialyloligosaccharides produced by marine bacterial sialyltransferases. The methodologies introduced in this chapter might to be useful for screening of bacteria that produce other types of glycosyltransferases, purification of membrane-binding proteins, and purification of oligosaccharides.

### **2. Screening bacteria for sialyltransferase activity**

#### **2.1 Basic screening method**

70 Chromatography and Its Applications

glycoconjugates play significant roles in many biological processes including inflammation, glycoprotein clearance from circulation, cell-cell recognition, cancer metastasis, and virus infection (Kannagi, 2002; Paulson, 1989). Sialyltransferases commonly transfer Neu5Ac from cytidine 5'-monophospho-*N*-acetylneuraminic acid (CMP-Neu5Ac) to various acceptor substrates (Angata & Varki, 2002). Thus, sialyltransferases are thought to be one of the

(A) (B)

(C) (D) (A) Neuraminic acid (Neu), (B) *N*-acetylneuraminic acid (Neu5Ac), (C) *N*-glycolylneuraminic acid

**OH**

**HO**

**HO OH**

**HO**

**HO OH**

**O**

**HOOC**

**HO**

**HO**

**H3COCHN**

**HO**

**HOOC**

**OH**

**OH**

**O**

**OH**

**HOOC**

**HOH2COCHN <sup>O</sup>**

Among the biological phenomena described above, the relationship between the carbohydrate chain structure of the host cell and host cell recognition by influenza virus is one of the best investigated (Suzuki, 2005; Weis et al., 1988). Many reports have shown that influenza A and B viruses bind via viral hemagglutinin to host cell surface receptors that are Neu5Ac- or Neu5Gc-linked glycoproteins or glycolipids (Suzuki, 2005). Furthermore, these influenza viruses also recognize the carbohydrate chain structure of the host cell (Connor et al., 1994). Confirming evidence has shown that avian influenza viruses recognize Neu5Ac2-3Gal1-3/4GlcNAc structures, and that human influenza viruses recognize Neu5Ac2-6Gal1-3/4GlcNAc structures (Connor et al., 1994; Suzuki, 2005). The host cell specificities of the influenza A and B viruses are determined mainly by the linkage of Neu5Ac or Neu5Gc to the penultimate galactose residues and core structure of the host glycoproteins or glycolipids. For this reason, the distribution of Neu5Ac and Neu5Gc and their linkage patterns on the host cell surface are important determinants of host tropism.

A large variety of oligosaccharides exist in nature. For example, many kinds of sialyloligosaccharides, such as 3′-sialyllactose, 6′-sialyllactose, and sialyllacto-*N*-neotetraose, are contained in milk of various animals (Kunz et al., 2000); however, the purification and isolation of sialyloligosaccharides from natural sources is very difficult due to their structural complexity. Therefore, the research use and development of drugs that depend on

important enzymes in the biosynthesis of sialylated glycoconjugates.

(Neu5Gc), (D) deaminoneuraminic acid (KDN).

**HO**

**HO**

**HO OH**

**HO**

**H2N**

**HO**

**HO OH**

**O**

**HOOC**

Fig. 1. Structures of sialic acids.

Samples of seawater, sea-sand, mud, seaweed, and small animals including various kinds of fishes and shells, were collected from various coastal locations in Japan. Bacteria that grew on marine agar 2216 or nutrient agar (Becton-Dickinson, Franklin Lakes, NJ, USA) that was supplemented with 2% NaCl at 15°C, 25°C, or 30°C were isolated from the samples. Aliquots of the bacteria were suspended in 10% glycerol and stored at −80°C. For each bacterial isolate, 6 mL of marine broth 2216 (Becton-Dickinson) in a 15-mL test tube was inoculated with bacteria and cultivated at 15°C, 25°C, or 30°C for 18 h on a rotary shaker (180 rpm). After the cultivation, bacteria were harvested from 2 to 4 mL of the culture broth by centrifugation and then suspended in 200 L of 20 mM sodium cacodylate buffer (pH 6.0) that contained 0.2% Triton X-100, lysed by sonication on ice, and measured immediately for sialyltransferase activity. Sialyltransferase activity was confirmed as follows: the reaction mixture (30 L) consisted of the bacterial lysate as the sample of enzyme, 120 mM lactose, 2.3 mM CMP-Neu5Ac (Nakarai Tesque, Kyoto, Japan), 4620 Bq CMP-[4,5,6,7,8,9-14C]- Neu5Ac (Amersham Biosciences, Little Chalfont, UK), 100 mM Bis–Tris buffer (pH 6.0), 0.5 M NaCl, and 0.03% Triton X-100. The reaction was carried out at 25°C for 2 h. The reaction mixture was then diluted with 5 mM sodium phosphate buffer (pH 6.8) to a final volume of 2 mL, and applied to a column (0.5 × 2 cm) of Dowex-1 × 8 (phosphate form, Bio-Rad Laboratories, Hercules, CA, USA). The eluate (2 mL) was collected directly into a scintillation vial for counting. The radioactivity of [4, 5, 6, 7, 8, 9-14C]-Neu5Ac that had transferred to the acceptor substrate in the eluate was measured by using a liquid scintillation counter, and the amount of Neu5Ac transferred was calculated. Unreacted CMP-Neu5Ac was not eluted in this buffer concentration. Using this procedure, we have isolated many bacteria that possess sialyltransferase activity. Many of the marine bacteria that produced sialyltransferases were classified in genus *Photobacterium* or the closely related genus *Vibrio*. For instance, *Photobacterium phosphoreum* JT-ISH-467 that showed 2,3 sialyltransferase activity was isolated from the outer skin of Japanese common squid, *Todarodes pacificus* (Tsukamoto et al. 2007); *Photobacterium damselae* JT0160 that expressed 2,6-sialyltransferase activity was isolated from seawater (Yamamoto et al., 1998); *Photobacterium* sp. JT-ISH-224 that contained both 2,3- and 2,6-sialyltransferase activities was isolated from the gut of Japanese barracuda, *Sphyraena pinguis* (Tsukamoto et al., 2008); and *Photobacterium leiognathi* JT-SHIZ-145 that expressed 2,6-sialyltransferase activity was isolated from the outer skin of Japanese squid, *Loliolus japonica* (Yamamoto et al., 2007).

Purification of Marine Bacterial Sialyltransferases and Sialyloligosaccharides 73

reaction product was eluted with 1 mL of 70% ethanol. One millilitre of scintillation cocktail was added to the eluate, and the radioactivity of the mixture was measured by using a liquid scintillation counter. In this way, we could detect glycosyltransferase activities, comprising fucosyltransferase, galactosyltransferase and/or *N*-acetylglucosaminyltransferase activity, simultaneously in marine bacteria. To clarify which of the glycosyltransferase activities the bacteria displayed, the enzymatic reaction was performed independently with each of the donor substrates in turn. The two bacteria that showed glycosyltransferase activity in Figure 2

Lectins are sugar-binding proteins that are highly specific for their sugar moieties. The lectin *Sambucus sieboldiana* agglutinin (SSA) recognizes the Neu5Ac2-6Gal or Neu5Ac2-6GalNAc structure of sialyloligosaccharides in glycoconjugates (Shibuya et al., 1989). Kajiwara et al. carried out lectin staining of the cells of *P. damselae* JT0160, *P. leiognathi* JT-SHIZ-145, *P. phosphoreum* JT-ISH-467, and *Photobacterium* sp. JT-ISH-224, by using biotin-labeled SSA, and then examined the cells by using differential interference contrast (DIC) and fluorescence microscopy (Kajiwara et al., 2010). Lectin staining was carried out as follows: the 4 bacterial species described above were cultivated in nutrient broth supplemented with 2% (w/v) NaCl at 25°C for 18 h on a rotary shaker. The bacterial cells were collected by centrifugation (8,000*g*, 15 min, 4°C) and suspended in 25 mM Tris-HCl buffer 8 (pH 7.5). The suspensions were spotted onto glass slides, fixed with a 4% (w/v) paraformaldehyde solution at room temperature for 15 min, and blocked with a 5% (w/v) bovine serum albumin (BSA) phosphate-buffered saline (PBS) solution. After the glass slides were washed, biotin-labeled SSA (5 mg mL−1) was added and the cells were incubated at room temperature for 2 h. After the cells were washed 4 times with PBS, Alexa 594-labeled streptavidin (Invitrogen, Carlsbad, CA, USA) solution (5 mg mL−1) was added and the incubation was continued at room temperature for 1 h. After 5 washes with PBS, the cells were mounted using Prolong Gold antifade reagent (Invitrogen) and observed by using DIC and fluorescence microscopy. The SSA bound to *Photobacteriu*m sp. JT-ISH-224, *P. damselae* JT0160, and *P. leiognathi* JT-SHIZ-145. These *Photobacterium* strains produce 2,6-sialyltransferases, so the lectin staining indirectly detected 2,6-sialyltransferase-producing bacteria (Fig. 3). SSA did not bind to *P. phosphoreum* JT-ISH-467, which produces only 2,3-sialyltransferase. Therefore, the SSA lectin might be useful to screen for not only Neu5Ac2-6Gal and/or Neu5Ac2-6GalNAc structures on the bacterial cell surface but also to screen for the production of 2,6-sialyltransferase. We consider that this method would be applicable to the screening of other glycosyltransferases by changing the type of lectin used. We have confirmed that one of the two bacteria that showed fucosyltransferase activity, described in section 2.2, was detected by biotin-labeled *Aleuria aurantia* lectin (AAL, from Seikagaku Kogyo), which recognizes the fucose residue in

were shown to specifically produce fucosyltransferase.

carbohydrate chains (Kochibe & Furukawa, 1980).

**3. Purification of sialyltransferase from the native bacterium** 

For the purification of a protein, it is necessary and important to find the appropriate conditions for enzyme solubilization, including solubilization efficiency, and the most efficient combination of chromatography processes. Each process has a different separation mode, and it is crucial to conduct a detailed study of the conditions required for each process. Crude extracts are commonly used in such studies, but care must be taken to

**2.3 Screening by lectin staining** 

#### **2.2 Simultaneous measurement of several glycosyltransferases activities**

To assess the activities of various glycosyltransferases, not only sialyltransferases, we performed the enzyme assay using a mixture of the donor substrates of glycosyltransferases (GDP-fucose, the common donor substrate of fucosyltransferase; UDP-galactose, the common donor substrate of galactosyltransferase; and UDP-GlcNAc, the common donor substrate of *N*acetyl-glucosaminyltransferase), and a mixture of the acceptor substrates of glycosyltransferases (4-Nitrophenyl α-D-galactopyranoside {Gal-α-pNp}; 4-Nitrophenyl β-Dgalactopyranoside {Gal-β-pNp}; 4-Nitrophenyl *N*-acetyl-α-D-galactosaminide {GalNAc-αpNp}; 4-Nitrophenyl *N*-acetyl-β-D-galactosaminide {GalNAc-β-pNp}; 4-Nitrophenyl *N*-acetylα-D-glucosaminide {GlcNAc-α-pNp}; 4-Nitrophenyl *N*-acetyl-β-D-glucosaminide {GlcNAc-βpNp}; 4-Nitrophenyl α-D-glucopyranoside {Glc-α-pNp}; 4-Nitrophenyl β-D-glucopyranoside {Glc-β-pNp}; 4-Nitrophenyl -L-fucopyranoside {Fuc-α-pNp}; 4-Nitrophenyl β-Lfucopyranoside {Fuc-β-pNp}; 4-Nitrophenyl α-D-mannopyranoside {Man-α-pNp}; and 4- Nitrophenyl β-D-mannopyranoside {Man-β-pNp}), and bacterial lysate described in 2.1 as the enzyme sample. An example of results obtained by using this method is shown in Figure 2.

Fig. 2. Screening assay for glycosyltransferase activities.

High levels of radioactivity were observed in the eluates from the reaction mixture of samples #1 and #8, respectively, when the reaction was performed in the presence of acceptor substrate. From this result, it was strongly expected that lysates prepared from bacteria number #1 and #8 contained fucosyltransferase, galactosyltransferase and/or *N*acetylglucosaminyltransferase. NC; negative control, +AC; containing acceptor substrate mixture in the reaction mixture, -AC; no acceptor substrate mixture in the reaction mixture.

The reaction mixture (50 L) consisted of the following: a sample of enzyme, a mixture of 0.5 mM acceptor substrates consisting of 4-nitrophenyl compounds, as described above, a mixture of 0.5 mM donor substrates consisting of sugar-nucleotides as described above, 4620 Bq UDP-[U-14C]-galactose, 4620 Bq UDP-*N*-acetyl-D-[U-14C]-glucosamine, 4620 Bq GDP-[U-14C]-fucose (Amersham Biosciences, Little Chalfont, UK), 100 mM bis-Tris buffer (pH 6.0), 10 mM MnCl2, and 3 mM ATP. The reaction was carried at 25°C for 16 to 18 h. After the reaction, 100 L of water was added to the reaction mixture, and the mixture was applied to a Sep-Pac Vac 50cc column (Waters, Milford, MA, USA) that was conditioned with ethanol and equilibrated with water. The column was washed twice with 1 mL of water and the reaction product was eluted with 1 mL of 70% ethanol. One millilitre of scintillation cocktail was added to the eluate, and the radioactivity of the mixture was measured by using a liquid scintillation counter. In this way, we could detect glycosyltransferase activities, comprising fucosyltransferase, galactosyltransferase and/or *N*-acetylglucosaminyltransferase activity, simultaneously in marine bacteria. To clarify which of the glycosyltransferase activities the bacteria displayed, the enzymatic reaction was performed independently with each of the donor substrates in turn. The two bacteria that showed glycosyltransferase activity in Figure 2 were shown to specifically produce fucosyltransferase.

#### **2.3 Screening by lectin staining**

72 Chromatography and Its Applications

To assess the activities of various glycosyltransferases, not only sialyltransferases, we performed the enzyme assay using a mixture of the donor substrates of glycosyltransferases (GDP-fucose, the common donor substrate of fucosyltransferase; UDP-galactose, the common donor substrate of galactosyltransferase; and UDP-GlcNAc, the common donor substrate of *N*acetyl-glucosaminyltransferase), and a mixture of the acceptor substrates of glycosyltransferases (4-Nitrophenyl α-D-galactopyranoside {Gal-α-pNp}; 4-Nitrophenyl β-Dgalactopyranoside {Gal-β-pNp}; 4-Nitrophenyl *N*-acetyl-α-D-galactosaminide {GalNAc-αpNp}; 4-Nitrophenyl *N*-acetyl-β-D-galactosaminide {GalNAc-β-pNp}; 4-Nitrophenyl *N*-acetylα-D-glucosaminide {GlcNAc-α-pNp}; 4-Nitrophenyl *N*-acetyl-β-D-glucosaminide {GlcNAc-βpNp}; 4-Nitrophenyl α-D-glucopyranoside {Glc-α-pNp}; 4-Nitrophenyl β-D-glucopyranoside {Glc-β-pNp}; 4-Nitrophenyl -L-fucopyranoside {Fuc-α-pNp}; 4-Nitrophenyl β-Lfucopyranoside {Fuc-β-pNp}; 4-Nitrophenyl α-D-mannopyranoside {Man-α-pNp}; and 4- Nitrophenyl β-D-mannopyranoside {Man-β-pNp}), and bacterial lysate described in 2.1 as the enzyme sample. An example of results obtained by using this method is shown in Figure 2.

High levels of radioactivity were observed in the eluates from the reaction mixture of samples #1 and #8, respectively, when the reaction was performed in the presence of acceptor substrate. From this result, it was strongly expected that lysates prepared from bacteria number #1 and #8 contained fucosyltransferase, galactosyltransferase and/or *N*acetylglucosaminyltransferase. NC; negative control, +AC; containing acceptor substrate mixture in the reaction mixture, -AC; no acceptor substrate mixture in the reaction mixture. The reaction mixture (50 L) consisted of the following: a sample of enzyme, a mixture of 0.5 mM acceptor substrates consisting of 4-nitrophenyl compounds, as described above, a mixture of 0.5 mM donor substrates consisting of sugar-nucleotides as described above, 4620 Bq UDP-[U-14C]-galactose, 4620 Bq UDP-*N*-acetyl-D-[U-14C]-glucosamine, 4620 Bq GDP-[U-14C]-fucose (Amersham Biosciences, Little Chalfont, UK), 100 mM bis-Tris buffer (pH 6.0), 10 mM MnCl2, and 3 mM ATP. The reaction was carried at 25°C for 16 to 18 h. After the reaction, 100 L of water was added to the reaction mixture, and the mixture was applied to a Sep-Pac Vac 50cc column (Waters, Milford, MA, USA) that was conditioned with ethanol and equilibrated with water. The column was washed twice with 1 mL of water and the

**2.2 Simultaneous measurement of several glycosyltransferases activities** 

Fig. 2. Screening assay for glycosyltransferase activities.

Lectins are sugar-binding proteins that are highly specific for their sugar moieties. The lectin *Sambucus sieboldiana* agglutinin (SSA) recognizes the Neu5Ac2-6Gal or Neu5Ac2-6GalNAc structure of sialyloligosaccharides in glycoconjugates (Shibuya et al., 1989). Kajiwara et al. carried out lectin staining of the cells of *P. damselae* JT0160, *P. leiognathi* JT-SHIZ-145, *P. phosphoreum* JT-ISH-467, and *Photobacterium* sp. JT-ISH-224, by using biotin-labeled SSA, and then examined the cells by using differential interference contrast (DIC) and fluorescence microscopy (Kajiwara et al., 2010). Lectin staining was carried out as follows: the 4 bacterial species described above were cultivated in nutrient broth supplemented with 2% (w/v) NaCl at 25°C for 18 h on a rotary shaker. The bacterial cells were collected by centrifugation (8,000*g*, 15 min, 4°C) and suspended in 25 mM Tris-HCl buffer 8 (pH 7.5). The suspensions were spotted onto glass slides, fixed with a 4% (w/v) paraformaldehyde solution at room temperature for 15 min, and blocked with a 5% (w/v) bovine serum albumin (BSA) phosphate-buffered saline (PBS) solution. After the glass slides were washed, biotin-labeled SSA (5 mg mL−1) was added and the cells were incubated at room temperature for 2 h. After the cells were washed 4 times with PBS, Alexa 594-labeled streptavidin (Invitrogen, Carlsbad, CA, USA) solution (5 mg mL−1) was added and the incubation was continued at room temperature for 1 h. After 5 washes with PBS, the cells were mounted using Prolong Gold antifade reagent (Invitrogen) and observed by using DIC and fluorescence microscopy. The SSA bound to *Photobacteriu*m sp. JT-ISH-224, *P. damselae* JT0160, and *P. leiognathi* JT-SHIZ-145. These *Photobacterium* strains produce 2,6-sialyltransferases, so the lectin staining indirectly detected 2,6-sialyltransferase-producing bacteria (Fig. 3). SSA did not bind to *P. phosphoreum* JT-ISH-467, which produces only 2,3-sialyltransferase. Therefore, the SSA lectin might be useful to screen for not only Neu5Ac2-6Gal and/or Neu5Ac2-6GalNAc structures on the bacterial cell surface but also to screen for the production of 2,6-sialyltransferase. We consider that this method would be applicable to the screening of other glycosyltransferases by changing the type of lectin used. We have confirmed that one of the two bacteria that showed fucosyltransferase activity, described in section 2.2, was detected by biotin-labeled *Aleuria aurantia* lectin (AAL, from Seikagaku Kogyo), which recognizes the fucose residue in carbohydrate chains (Kochibe & Furukawa, 1980).

#### **3. Purification of sialyltransferase from the native bacterium**

For the purification of a protein, it is necessary and important to find the appropriate conditions for enzyme solubilization, including solubilization efficiency, and the most efficient combination of chromatography processes. Each process has a different separation mode, and it is crucial to conduct a detailed study of the conditions required for each process. Crude extracts are commonly used in such studies, but care must be taken to

Purification of Marine Bacterial Sialyltransferases and Sialyloligosaccharides 75

cell lysis after cultivation. For instance, almost no sialyltransferase activity was detected in crude extract prepared from cryopreserved cells of *P. damselae*. The procedure for crude

1. After cultivation, *P. damselae* JT0160 cells were harvested from the culture by

2. The harvested cells were suspended in 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100 and 1 M NaCl, and were sonicated immediately (<4°C) until the absorbance at 660 nm reached 30% or less of that of the original cell

3. The sonicated solution was centrifuged (100,500*g*, 60 min) and the supernatant was dialyzed, using cellulose tubing, against 20 mM sodium cacodylate buffer (pH 6.0)

4. After dialysis, the precipitate was removed by centrifugation (100,500*g*, 60 min) to

Sialyltransferase produced by *P. damselae* was then purified from the crude extract by a combination of 4 steps of column chromatography. The conditions and method used for

1. Q-Sepharose column chromatography. A column of Hi-Load 26/10 Q Sepharose HP (ø 2.6 × 10 cm; GE Healthcare Science, Buckinghamshire, UK) was equilibrated with 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100. Clarified extract was applied to the column, and the column was washed with 150 mL of the same buffer. Enzyme fractions were eluted with a linear gradient of 0 to 1 M NaCl in the buffer. The fractions exhibiting sialyltransferase activity ("active" fractions) were collected and pooled. Desalting of the "active" fractions was performed by dialysis, using cellulose tubing, against 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100. 2. Hydroxyapatite column chromatography. A column of hydroxyapatite (ø 2 × 10 cm; Bio-Rad Laboratories) was equilibrated with 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100. After the application of the enzyme solution obtained in step 1, the column was washed with the same buffer. The enzyme fraction was eluted with a gradient of 0 to 0.35 M potassium phosphate. The "active" fractions were collected and pooled, and then concentrated by ultrafiltration using Molecut L

3. Gel–filtration column chromatography. A column of Hi-Load 26/60 Sephacryl S-200 HE (ø 2.6 × 60 cm; GE Healthcare Science, Buckinghamshire, UK) was equilibrated with 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100 and 0.1 M NaCl. The enzyme solution obtained in step 2 was applied to the column and eluted with the same buffer. The "active" fractions were collected and pooled. Desalting of these fractions was performed by dialysis, using cellulose tubing, against 20 mM sodium

4. CDP–hexanolamine-agarose column chromatography. A column of CDP– hexanolamine-agarose (ø 1 × 3 cm) was equilibrated with 20 mM sodium cacodylate buffer (pH 6.0) containing 0.2% Triton X-100. The enzyme solution (4 mL) obtained in step 3 was applied to the column. The column was washed with 8 mL of the same buffer. The enzyme was eluted with 6 mL of 2 M NaCl, and "active" fractions were

**3.1.2 Purification of sialyltransferase by using column chromatography** 

(exclusion molecular mass, 10 kDa ; Millipore, Billerica, MA, USA).

cacodylate buffer (pH 6.0) containing 0.2% Triton X-100.

collected and pooled.

extract preparation was as described below.

centrifugation (6,000 *g*, 20 min).

containing 0.2% Triton X-100.

each chromatography step are described below.

obtain the clarified extract.

suspension.

ISH224, *Photobacterium* sp. JT-ISH-224; Pd0160, *P. damselae* JT0160; SHIZ145, *P. phosphoreum* JT-SHIZ-145; SSA, fluorescence microscopy of cells stained with *Sambucus sieboldiana* agglutinin (SSA); DIC, differential interference contrast microscopy of the cells shown in the SSA panels.

Fig. 3. Lectin staining of *Photobacterium* strains by *Sambucus sieboldiana* agglutinin.

minimize protease activity, which may decompose the target enzyme. Furthermore, it is necessary to consider the pH of the buffers used in the purification steps as well as the temperature employed during the preparation of the extracts and the purification process. For details of general procedures and methods for protein purification, we recommend that you refer to other textbooks (e.g., Deutscher, 1990; Scopes,.1982). Here, we describe examples of the purification of sialyltransferase from marine bacteria.

#### **3.1 Purification of 2,6-sialyltransferase from** *P. damselae* **JT0160**

#### **3.1.1 Preparation of the crude extract from** *P. damselae* **cells**

The first step in the purification of a protein is the preparation of an extract containing the protein in a soluble form. During the purification of sialyltransferase from *P. damselae*, we examined in detail the conditions for preparing a crude extract containing the target enzyme in a soluble form (Yamamoto et al., 1998). The method that we established was deemed appropriate for the preparation of a crude extract containing sialyltransferase because no decrease in sialyltransferase activity was detected during the procedure. We determined that the most important factor in the preparation of the crude extract was the timing of the

ISH224, *Photobacterium* sp. JT-ISH-224; Pd0160, *P. damselae* JT0160; SHIZ145, *P. phosphoreum* JT-SHIZ-145; SSA, fluorescence microscopy of cells stained with *Sambucus sieboldiana* agglutinin (SSA); DIC,

minimize protease activity, which may decompose the target enzyme. Furthermore, it is necessary to consider the pH of the buffers used in the purification steps as well as the temperature employed during the preparation of the extracts and the purification process. For details of general procedures and methods for protein purification, we recommend that you refer to other textbooks (e.g., Deutscher, 1990; Scopes,.1982). Here, we describe

The first step in the purification of a protein is the preparation of an extract containing the protein in a soluble form. During the purification of sialyltransferase from *P. damselae*, we examined in detail the conditions for preparing a crude extract containing the target enzyme in a soluble form (Yamamoto et al., 1998). The method that we established was deemed appropriate for the preparation of a crude extract containing sialyltransferase because no decrease in sialyltransferase activity was detected during the procedure. We determined that the most important factor in the preparation of the crude extract was the timing of the

differential interference contrast microscopy of the cells shown in the SSA panels.

examples of the purification of sialyltransferase from marine bacteria.

**3.1 Purification of 2,6-sialyltransferase from** *P. damselae* **JT0160** 

**3.1.1 Preparation of the crude extract from** *P. damselae* **cells** 

Fig. 3. Lectin staining of *Photobacterium* strains by *Sambucus sieboldiana* agglutinin.

cell lysis after cultivation. For instance, almost no sialyltransferase activity was detected in crude extract prepared from cryopreserved cells of *P. damselae*. The procedure for crude extract preparation was as described below.


#### **3.1.2 Purification of sialyltransferase by using column chromatography**

Sialyltransferase produced by *P. damselae* was then purified from the crude extract by a combination of 4 steps of column chromatography. The conditions and method used for each chromatography step are described below.


Purification of Marine Bacterial Sialyltransferases and Sialyloligosaccharides 77

4. Mono Q column chromatography (pH 7.0). The enzyme solution obtained in step 3 was applied to a column of Mono Q 10/100 GL equilibrated with 20 mM bis-Tris buffer (pH 7.0) containing 0.3% Triton X-100. After the column was washed with 20 mM bis-Tris buffer (pH 7.0) containing 0.3% Triton X-100, the enzyme was eluted with a linear

gradient of 0 to 1 M NaCl in the same buffer. The "active" fractions were pooled. 5. Superdex 200 column chromatography. The enzyme solution obtained in step 4 was loaded onto Hi-Load 16/60 Superdex 200 pg (ø 1.6 × 60 cm; GE Healthcare Science) that was equilibrated with 20 mM bis-Tris buffer (pH 7.0) containing 0.3% Triton X-100 and 0.2 M NaCl and eluted with the same buffer. The "active" fractions were collected and

Crude extract 3155 6159 8.40 1.4 100 1 DEAE 410 932 3.10 3.4 37 3 Hydroxyapatite 264 153 1.30 8.2 15 6 Mono Q (pH 6.0) 12 24 0.96 39 11 29 Mono Q (pH 7.0) 1.5 1.7 0.52 315 6.2 29 Superdex 200 1.5 0.2 0.10 457 1.2 333

The enzyme was purified 333-fold, with a yield of 1.2%. Because no affinity chromatography step was used, the yield of the protein purification in this case was very low. Therefore, preparing affinity gels with the appropriate ligand for the target enzyme is very important

The results for the purification of the enzyme are summarized in Table 2.

Table 2. Purification of sialyltransferase from cell lysate of *P. phosphoreum.*

**4. Enzymatic synthesis and purification of sialyloligosaccharides** 

**4.1 Synthesis of sialyloligosaccharides by recombinant sialyltransferases from** 

Chemoenzymatic synthesis of various sialyloligosaccharides by mammalian sialyltransferases, and the purification of the product, has been reported (Sabesan & Paulson, 1986). However, mass-production of sialyloligosaccharides by using mammalianderived sialyltransferases remains problematic because the enzymes are unstable and difficult to produce as recombinant proteins in *Escherichia coli*. In comparison to mammalian sialyltransferases, bacterial sialyltransferases are generally more stable and productive in *E. coli* protein-expression systems (Tsukamoto et al., 2007, 2008; Yamamoto et al., 2006), and they show a broader acceptor substrate specificity (Izumi & Wong, 2001; Yu et al., 2005). Here, we report the methods that we developed to use recombinant sialyltransferases from marine bacteria to successfully produce large quantities of 6′-sialyllactose and synthesize

buffer (pH 7.0) containing 0.3% Triton X-100.

pooled.

Purification step

in the purification process.

various sialyloligosaccharides.

**marine bacteria** 

eluted with a linear gradient of 0 to 1 M NaCl in the same buffer. The "active" fractions were pooled, and then diluted to three times the original volume with 20 mM bis-Tris

> Volume Total protein Total activity Specific activity Yield Purification (mL) (mg) (U) (mU/mg) (%) (fold)


The purity and yield of the enzyme at each step is summarized in Table 1.

Table 1. Purification of sialyltransferase from cell lysate of *Photobacterium damselae.*

The enzyme was purified 688-fold, with a yield of 19%. The purified enzyme migrated as a single polypeptide with a molecular mass of 61 kDa by SDS-polyacrylamide gel electrophoresis under denaturing conditions.

#### **3.2 Purification of 2,3-sialyltransferase from** *P. phosphoreum* **JT-ISH-467**

#### **3.2.1 Preparation of the crude extract from** *P. phosphoreum* **cells**

The crude extract containing sialyltransferase from *P. phosphoreum* cells was prepared by the method described in section 3.1.1, with slight modifications (Tsukamoto et al., 2007), and then crude extract containing the soluble form of the enzyme was prepared.

#### **3.2.2 Purification of sialyltransferase from** *P. phosphoreum* **by using column chromatography**

Sialyltransferase produced by *P. phosphoreum* was purified from the crude extract by a combination of 5 steps of column chromatography. The conditions and method used for each of the column chromatography steps are described below.


Crude extract 760 2584 21.1 0.008 100 1 Q Sepharose 240 552 12.4 0.022 59 2.8 Hydroxyapatite 120 85 8 0.094 38 11.8 Sephacryl S-200 30 20.1 6.7 0.3 32 37.5 CDP-hexanolamine-agarose 15 0.75 4.1 5.5 19 687.5

Table 1. Purification of sialyltransferase from cell lysate of *Photobacterium damselae.*

**3.2 Purification of 2,3-sialyltransferase from** *P. phosphoreum* **JT-ISH-467** 

then crude extract containing the soluble form of the enzyme was prepared.

**3.2.2 Purification of sialyltransferase from** *P. phosphoreum* **by using column** 

**3.2.1 Preparation of the crude extract from** *P. phosphoreum* **cells** 

each of the column chromatography steps are described below.

electrophoresis under denaturing conditions.

6.0) containing 0.3% Triton X-100.

The enzyme was purified 688-fold, with a yield of 19%. The purified enzyme migrated as a single polypeptide with a molecular mass of 61 kDa by SDS-polyacrylamide gel

The crude extract containing sialyltransferase from *P. phosphoreum* cells was prepared by the method described in section 3.1.1, with slight modifications (Tsukamoto et al., 2007), and

Sialyltransferase produced by *P. phosphoreum* was purified from the crude extract by a combination of 5 steps of column chromatography. The conditions and method used for

1. DEAE column chromatography. The clarified crude extract was applied to a Hi-Prep 16/10 DEAE FF column (ø 1.6 × 10 cm; GE Healthcare Science) equilibrated with 20 mM bis-Tris buffer (pH 6.0) containing 0.3% Triton X-100. After the column was washed with the same buffer, the enzyme was eluted with a linear gradient of 0 to 1 M NaCl in the same buffer. The fractions with sialyltransferase activity were pooled and then diluted to three times the original volume with 20 mM potassium phosphate buffer (pH

2. Hydroxyapatite column chromatography. The enzyme solution obtained in step 1 was applied to a hydroxyapatite column (ø 1.5 × 11.3 cm; Bio-scale CHT20-I; Bio-Rad Laboratories) that was equilibrated with 20 mM potassium phosphate buffer (pH 6.0) containing 0.3% Triton X-100. After the column was washed with the same buffer, the enzyme was eluted with a linear gradient of 20 to 500 mM potassium phosphate. The "active" fractions were pooled, and then diluted to two times the original volume with

20 mM potassium phosphate buffer (pH 6.0) that contained 0.3% Triton X-100. 3. Mono Q column chromatography (pH 6.0). The enzyme solution obtained in step 2 was loaded onto a column of Mono Q 10/100 GL (ø 1 × 10 cm; GE Healthcare Science) that was equilibrated with 20 mM potassium phosphate buffer (pH 6.0) containing 0.3% Triton X-100. After the column was washed with the same buffer, the enzyme was

Volume Total protein Total activity Specific activity Yield Purification (mL) (mg) (U) (U/mg) (%) (fold)

The purity and yield of the enzyme at each step is summarized in Table 1.

Purification step

**chromatography** 

eluted with a linear gradient of 0 to 1 M NaCl in the same buffer. The "active" fractions were pooled, and then diluted to three times the original volume with 20 mM bis-Tris buffer (pH 7.0) containing 0.3% Triton X-100.



The results for the purification of the enzyme are summarized in Table 2.

Table 2. Purification of sialyltransferase from cell lysate of *P. phosphoreum.*

The enzyme was purified 333-fold, with a yield of 1.2%. Because no affinity chromatography step was used, the yield of the protein purification in this case was very low. Therefore, preparing affinity gels with the appropriate ligand for the target enzyme is very important in the purification process.
