THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 13, Issue of May 5 , pp. 9886-9891, 1993 Printed in U. S. A.

Enzymatic Characterization of P-D-Galactoside a2,3-trans-Sialidase from Trypanosoma cruxi”

(Received for publication, November 19, 1992)

Peter ScudderlB, James P. Doomll, Marina Chuenkovall , Ian D. Manger*, and Miercio E. A. Pereira(1 From the SGlycosciences Group, Division of Protein Biochemistry, Monsanto Co., St. Louis, Missouri 63167, the Whysical Sciences Center, Monsanto Co., St. Louis, Missouri 63198, and the 11 Department of Medicine, Division of Geographic Medicine and Infectious Disease, New England Medical Center Hospitals, Boston, Massachusetts 021 11

The substrate specificity, physico-chemical, and ki- netic properties of the trans-sialidase from Trypano- soma cruzi have been investigated. The enzyme dem- onstrates activity towards a wide range of saccharide, glycolipid, and glycoprotein acceptors which terminate with a &linked galactose residue, and synthesizes ex- clusively an a2-3 sialosidic linkage. Oligosaccharides which terminate in Gal~l-4(Fucal-3)GlcNAc, GalB1- 3(Fucal-4)GlcNAc, or Galal- are not acceptor-sub- strates. The enzyme utilizes a2,3-linked sialic acid when the donor species is an oligosaccharide and can also transfer, at a low rate, sialic acid from synthetic a-sialosides such as p-nitrophenyl-a-N-acetylneura- minic acid, but NeuAca2-3Galj31-4(Fuca1-3)Glc is not a donor-substrate. The trans-sialidase has an ap- parent pH optimum of 7.9 and a temperature optimum of 13 “C. The kinetic properties of the enzyme suggest that the trans-sialylation reaction may occur via a rapid equilibrium random or steady-state ordered mechanism. A method for immobilizing the enzyme is described together with examples of its use for the synthesis of oligosaccharide and glycoprotein precur- sors of sialyl-Lewis” and sialyl-Lewis”.

Trypomastigotes of Trypanosoma cruzi express a develop- mentally regulated trans-sialidase (1-3) that is anchored to the plasma membrane of the parasite via a glycosylphospha- tidylinositol tail (4, 5 ) . Physiologically, the enzyme is thought to effect sialylation of parasite surface glycoproteins such as Ssp-3 (6), a process which is crucial for binding to, and infection of, target cells (3,6,7). The enzyme, which is unique to T. cruzi, is unusual in that it transfers sialic acid from an a2,3-sialylated sugar and not CMP-P-sialic acid (the donor- substrate for mammalian sialyltransferases), and host sial- oglycans are utilized exclusively as donors since T. cruzi is unable to synthesize sialic acid (8). The gene which encodes trans-sialidase has been isolated (9) and shown to be virtually homologous to the gene for T. cruzi neuraminidase (4), an enzyme that was previously shown to exert a negative control on the infectivity of T. cruzi (10, 11). It has now been con- cluded that these enzymes are the same gene product (9, 12)

* This work was supported by the Monsanto Corp. and National Institutes of Health Grant AI 18102 (to M. E. A. P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§ To whom correspondence should be addressed Glycosciences Group, Div. of Protein Biochemistry, Monsanto Co., 800 N. Lind- bergh Blvd., St. Louis, MO 63167. Tel.: 314-694-8976; Fax: 314-694- 8949.

and that the active site for both activities resides at the N terminus of the molecule, a region which is known to have homology with certain bacterial neuraminidases (4). The bio- chemical and enzymatic properties of the neuraminidase have been reported in some detail (11, 13), but only preliminary data have been reported for the trans-sialidase (3, 9). The studies described in this report examine the acceptor- and donor-substrate specificity, kinetic, and physico-chemical properties of the trans-sialidase.

EXPERIMENTAL PROCEDURES

Materials-LNT,’ LNNT, LNFPII, and LNFPIII were purified from human milk (14). 2,3-Sialyllactose (NeuAca2-3Galpl-4Glc) and 2,6-sialyllactose (NeuAca2-6Galp1-4Glc) were isolated from bovine colostrum using the method described by Veh et al. (15) and converted into the sodium salt by passsage over Dowex AG 5OW-X12 (Na+ form). The sialyl-Lewis’tetrasaccharide, NeuAcol2-3Gal~l-4(Fucal- 3)Glc, was prepared by the enzymatic fucosylation of 2,3-sialyllactose using partially purified human milk a-3,4-fucosyltransferase (16) and was isolated by chromatography on Dowex AG1-X2 (15) using an eluant of 40 mM pyridinium acetate, pH 5.0. The tetrasaccharide was judged homogeneous by ‘H NMR and HPAEC. Sialylparagloboside (NeuAca2-3Gal~l-4GlcNAc~l-3Galpl-4Glc~l-Cer) was isolated from bovine erythrocyte stroma using the method described by Hanf- land (17) and its structure was confirmed by fast atom bombardment- mass spectrometry. Paragloboside (Gal~l -4GlcNAc~l-3Gal~l - 4Glcpl-Cer) was isolated by extraction with chloroform/methanol/ water (3:2:1, v/v) following treatment of sialylparagloboside with Clostridium perfringens neuraminidase. 2-O-(p-Nitrophenyl)-N-ace- tyl-a-neuraminic acid was from Seikagaku Kogyo. LNFPV (Galpl- 3GlcNAcpl-3Galpl-4[Fuc~u1-3]Glc) and 2,3-sialyllactosamine (NeuAca2-3Galpl-4GlcNAc) were purchased from Oxford Glyco- Systems. C. perfringens neuraminidase, [‘4C]N-acetylIactosamine (55 Ci/mol), 2’-(4-methylumbelliferyl)-a-N-acetylneuraminic acid, and all other sugars were obtained from Sigma. Newcastle disease virus neuraminidase was from Genzyme.

Isolation of trans-Sialidase by Immunoaffinity Chromatography- A crude preparation of trans-sialidase was obtained as described previously (10, 11). In brief, 2‘. cruzi trypomastigotes (Silvio X-10/4 clone) were grown in Vero cell monolayers for 3-5 days in a medium containing 2.5% NU serum (Collaborative Research) and, after har- vesting the parasites by centrifugation, the conditioned medium was filtered through nitrocellulose (0.22 pm). The trans-sialidase in 100

The abbreviations used are: LNT, lacto-N-tetraose (Galpl- 3GlcNAc~l-3Galpl-4G1c); LNNT, lacto-N-neotetraose (Galpl- 4GlcNAcpl-3Galp1-4Glc); LNFPII, lacto-N-fucopentaose I1 (Galpl- 3[Fuca1-4]GlcNAcp1-3Galpl-4Glc); LNFPIII, lacto-N-fucopen- taose I11 (Gal~l-4[Fucoll-3]G1cNAc~l-3Gal~l-4Glc); LNFPV, Gal~l-3GlcNAc~l-3Gal~l-4[Fucal-3]Glc; MeU-NeuAc, 2’44- methylumbellifery1)-a-N-acetylneuraminic acid; p-NP-NeuAc, 2-0- (p-nitropheny1)-N-acetyl-a-neuraminic acid; HPAEC, high perform- ance anion exchange chromatography; PED, pulsed electrochemical detection; HP-TLC, high performance-thin layer chromatography; BSA, bovine serum albumin; PBS, phosphate-buffered saline; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-l-propanesul- fonic acid.

9886

T. cruzi trans-Sialidase 9887

ml of the conditioned supernatant was isolated at 4 “C on an affinity column which contained 0.1 ml of TCN-2-Sepharose equilibrated with 50 mM PBS, pH 7.4. TCN-2 is a monoclonal antibody specific for the trans-sialidase (11) which was coupled to CNBr-activated Sepharose at a concentration of 2 mg/ml settled gel. The column was washed with 100 ml of PBS and the enzyme was eluted with PBS containing 10 mg/ml of the synthetic dodecapeptide, D-S-S-A-H-G- T-P-S-T-P-A, referred to as “TR.” This peptide corresponds to the eptitope recognized by antibody TCN-2 (18) which is present as a tandem repeat at the C terminus of the trans-sialidase (4). Enzyme- active fractions were pooled and TCN-2 immunoglobulin was re- moved by passage through a column containing Protein G-Sepharose. The purified enzyme was concentrated by ultrafiltration using a Centricon-10 microconcentrator (Amicon), and the peptide was re- moved by fast protein liquid chromatography on a Superose 12 column equilibrated with PBS.

trans-Sialidase Assay-Enzyme, 10 pl, was added to 40 pl of 50 mM cacodylate/HCl buffer, pH 6.9, containing 0.5 pmol of 2,3-sialyllac- tose, 0.25 pmol of [14C]N-acetyllactosamine (4.4 X 10‘ dpm), 50 pg of BSA, and 0.02% NaN3. After incubation at 37 “C for 30 min, the reaction mixture was diluted with 1 ml of distilled water and applied to a column containing 0.5 ml of Q-Sepharose Fast Flow (acetate- form) equilibrated with water. The [14C]N-acetyllactosamine was eluted by washing the column with 3 ml of distilled water and the sialylated [14C]N-acetyllactosamine then eluted with 2.5 ml of 0.8 M ammonium acetate. The sialylated product was quantitated by liquid scintillation counting. A similar procedure was used in assays which utilized p-NP-NeuAc or MeU-NeuAc as the donor-substrate. One unit of enzyme activity is the amount catalyzing the incorporation of 1 pmol of sialic acid into N-acetyllactosamine/min.

Assay of Acceptor-Substrate Activity by High Performance Anion Exchange Chromatography-Reactions were set up at 37 “C in a final volume of 50 p1 of 50 D M cacodylate/HCl buffer, pH 6.9, containing 65 microunits of trans-sialidase, 50 nmol of 2,3-sialyllactose, 50 nmol of acceptor, 50 nmol of glucuronic acid (internal standard), 50 pg of BSA, and 0.02% sodium azide. Aliquots (5 pl) were removed at zero time and after 45 min, and chromatographed (Dionex BioLC System) at ambient temperature on a CarboPac PA-1 (4 X 250 mm) column eluted with 150 mM NaOH, 100 mM NaOAc at a flow rate of 1 ml/ min. Carbohydrates were monitored by triple-pulsed electrochemical detection using the following pulse potentials and durations: 0.4 V (0.0-0.50 s), 0.9 V (0.51-0.59 s), -0.3V (0.60-0.65 s). Integration was from 0.3 to 0.5 s. The data were processed using Dionex AI-450 (version 3.21) data management software. The amount of sialylated product was calculated from the decrease in concentration of 2,3- sialyllactose. Under the incubation conditions used the amount of free sialic acid formed as a result of hydrolysis of either 2,3-sialyllac- tose or the sialylated product was negligible (see “Results”).

Methylation Analysis of Sialylated Enzyme Products-Reaction mixtures were set up as described above but in the absence of glucuronic acid and including 2,3-sialyllactose and the acceptor- substrate at 5 mM. After incubation for 18 h, the sialylated product was isolated by chromatography on Q-Sepharose Fast Flow and converted to the correspondingpermethylated monosaccharide alditol acetate as previously described (19). The derivatized samples were analyzed on a Hewlett-Packard 5890A capillary gas chromatogram equipped with a 0.25 mm X 30-m fused silica DB5 capillary column (J & W Scientific) which, after a 5-min hold time, was ramped from 150 to 230 “C at a rate of 2 “C/min. The instrument was interfaced to a 700 Series Finnigan MAT ITD mass spectrometer and the spectra obtained matched against a library of partially methylated alditol acetates.

Kinetics-Reaction mixtures of 50 p1 contained 50 mM cacodylate/ HCI buffer, pH 6.9, 0.1 milliunit of trans-sialidase, 50 pg of BSA, 0.02% NaN3, and varied amounts of the donor- and acceptor-sub- strates. The matrix consisted of five fixed concentrations of each substrate (1.07-12 mM 2,3-sialyllactose and 0.38-5.10 mM N-acetyl- lactosamine which included 0.2 pCi of [“CJN-acetyllactosamine) and was assayed in duplicate. Reactions were carried out for 60 min at 22 ‘C, and the sialylated product measured using the radioisotopic assay described above. Initial velocity data were fitted to the bisub- strate rate equation (GraFit software program, Erithacus Software Ltd., Staines, United Kingdom) after visual inspection of reciprocal plots indicated a sequential mechanism.

Immobilization of trans-Sialidase-1.0 ml of concanavalin A-Seph- arose 4B (Sigma, 14 mg of lectin/ml of gel) equilibrated with 10 mM cacodylate/HCl buffer, pH 6.9, containing 1 mM CaClz and 0.5 M NaCl was added to 5.0 ml of the partially purified preparation of

trans-sialidase (see above under “Isolation of trans-Sialidase”) con- taining 70 milliunits of enzyme and 1 mM CaClZ. The suspension was mixed for 30 min at 22 “C when the gel was recovered on a glass sinter and washed with 50 ml of 10 mM cacodylate/HCl buffer, pH 6.9, containing 1 mM CaC& and 0.02% sodium azide.

Generation of Lactose-BSA Conjugate-Lactose was first converted into a 1-N-glycyl intermediate (20), then to the corresponding isothi- ocyanate and finally coupled to BSA as previously described (21). The BSA conjugate had a lactose content of 24 mol/mol of protein as determined by phenol/sulfuric acid assay (22).

Sialylutwn of Gulp1 -3GlcNAc@1-3Gal~l-4Glc and Gal@l-4Glc-BSA Using Concanavalin A-immobilized trans-Sialidase-A 0.4-ml aliquot of concanavalin A-Sepharose-immobilized trans-sialidase (50 milliun- its/ml gel) was added to 0.9 ml of 50 mM cacodylate/HCl buffer, pH 6.9, containing 2 mM CaClZ and either 20 pmol of Galp-3GlcNAc@I- 3Galpl-4Glc and 40 pmol of 2,3-sialyllactose (Reaction 1) or 5 mg of lactose-BSA (1.74 pmol of lactose acceptor sites), 35.0 pmol of 2,3- sialyllactose, and 3 pmol of fucose as an internal standard (Reaction 2). The suspensions were mixed by inversion at 37 “C for 16 h and the course of the sialylation reactions were monitored by HPAECI PED as described above using eluants of 150 mM NaOH, 100 mM NaOAc and 150 mM NaOH for Reaction 1 and Reaction 2, respec- tively. Sialylation of the BSA conjugate was determined by measuring the conversion of 2,3-sialyllactose to lactose using fucose as an internal standard. The sialylated LNT and sialylated lactose-BSA were purified by gel filtration on a column (1.5 X 180 cm) of TSK- HW4O(S) equilibrated with 50 mM ammonium bicarbonate.

Determination of Protein-Protein was assayed using the method of Lowry et al. (23) using BSA as a standard.

Polyacrylamide Gel Electrophoresis and Immunoblotting-Reduc- tive SDS-polyacrylamide (7.5%) gel electrophoresis and immunoblot- ting of purified trans-sialidase was performed as described previously (10, 11).

High Performance-Thin Layer Chromatography-The products of the trans-sialylation reaction which utilized 2,3-sialyllactose and par- agloboside as substrates were separated by HP-TLC on Merck Silica Gel 60 plates developed with chloroform/methanol/water (6:4:1, v/v) and located with 0.2% orcinol in 2 M HzS04.

RESULTS

Purification and Properties of trans-Sialidase trans-Sialidase was purified 1000-fold to apparent homo-

geneity in a single immunoaffinity chromatography step using an immobilized monoclonal antibody, TCN-2, which recog- nizes a 12-amino acid tandem repeat (D-S-S-A-H-G-T-P-S- T-P-A, peptide TR) that is present at the C terminus of the enzyme (18). A synthetic peptide having this sequence was used to elute the trans-sialidase. The specificity of this peptide was demonstrated by the failure of a similar peptide, which lacked threonine at position 7 (peptide TR-t7), to displace trans-sialidase from the affinity column (see Fig. 1). After fast protein liquid chromatography (to remove the peptide ligand), the purified trans-sialidase had a specific activity of 100 units/

30000 I 1

4 0 5 10 1 5 20 25 30

Fractlon number

FIG. 1. Isolation of tram-sialidase by immunoaffinity chro- matography. Culture supernatants of T. cruzi were applied to the TCN-2-Sepharose column and eluted with the synthetic peptide TR- t7 (DSSAHGPSTPA), followed by TR (DSSAHGTPSTPA). Full details appear under “Experimental Procedures.”

9888 T. cruzi trans-Sialidase

mg protein and migrated by reducing SDS-PAGE as an im- munoreactive doublet having an apparent M, = 200,000 (see Fig. 2). The recovery of isolated enzyme was approximately 10 ng/ml conditioned medium.

The purified enzyme (protein concentration, 0.5 pg/ml 50 mM cacodylate/HCl buffer, pH 6.9, containing 0.02% NaN3) was extremely stable and no loss of activity could be detected after storage at 4 “C for 1 month. In the presence of 50 mM cacodylate/HCl buffer, pH 6.9, the enzyme was unaffected by rapid freezing in liquid nitrogen and could be freeze-dried without detectable loss of activity in the presence of a carrier protein such as 2 mg/ml BSA.

The enzyme was active over a broad range of pH with an apparent pH optimum of 7.9. At pH 5.4 and 8.9, the activity of the trans-sialidase was approximately 30% of maximal (see Fig. 3A). The temperature optimum of the trans-sialidase was

A B C 200+ - 92+ 69+

FIG. 2. Reductive SDS-polyacrylamide gel (7.5%) electro- phoresis of affinity-purified tram-sialidase. Electrophoresis and immunoblotting of the affinity-purified trans-sialidase was per- formed as described under “Experimental Procedures.” The arrows indicate the position of molecular weight markers. A, silver stain of conditioned medium supernatants, 1 pg of protein/lane. B, silver stain of the affinity-purified trans-sialidase, 0.05 pg of protein/lane. C, immunoblot of the purified trans-sialidase, 0.05 pg of protein/lane, developed with TCN-2 followed by a goat anti-mouse antibody con- jugated with alkaline phosphatase.

0 2 4 6 8 1 0 1 2 0 10 20 30 4 0

PH Temperature (“C) FIG. 3. Effect of pH and temperature on the activity of

tram-sialidase. A, enzyme, 0.25 milliunits, was assayed as described under “Experimental Procedures” over the pH range 4.4-8.9. The buffers used were as follows: pH 4.4-5.4, 0.2 M citric acid/Na,HPO,; pH 5.4-7.4, 0.25 M cacodylate/HCb pH 7.4-8.9, 0.25 M Tris/HCl. Results are expressed as a percentage of the maximum activity which was seen at pH 7.9. B, enzyme, 0.25 milliunits, was assayed (see “Experimental Procedures”) at various temperatures from 2 to 37 “C. Results are expressed as a percent of the activity seen at 13 “C, the

determined to be 13 “C (see Fig. 3B), with 55% of the maximal activity being seen at 2 and 37 “C.

Acceptor-Substrate Specificity HPAEC was used to monitor the ability of various oligo-

saccharides to act as acceptor-substrates for the trans-siali- dase. Conditions were optimized to restrict consumption of 2,3-sialyllactose to less than 15%, and the extent of sialylation was monitored by following the decrease in concentration of 2,3-sialyllactose using an internal standard of glucuronic acid. The comparative rates of sialic acid incorporation into the various substrates and the HPAEC retention times of the products are given in Table I. Also included are previously published (24) specificity data for rat liver 8-galactoside a2,3- sialyltransferase. As can be seen, trans-sialidase catalyzed the transfer of sialic acid to the majority of oligosaccharides tested which terminated with a nonreducing &linked galactose res- idue. The exceptions were lactobionic acid (Galpl-4gluconic acid), LNFPII, and LNFPIII which, even after incubation with the enzyme for 18 h, showed no detectable formation of sialylated product. Disaccharides which terminated in an a- linked galactose residue or the monosaccharides Gal, GalNAc, and Man were not acceptor-substrates.

Methylation Analysis of Sialylated Products-The sialylated products were isolated by anion exchange chromatography and subjected to methylation analysis. In each case, an alditol acetate corresponding to 2,4,6-tri-O-methylgalactose was identified, clearly establishing that the trans-sialidase cata- lyzes the formation of a 2-3 linkage between sialic acid and galactose (as expected, an additional alditol acetate corre- sponding to 2,3,6-tri-O-methylgalactose was identified when Gal81-4Galfi1-4Glc was used as a substrate). Each product was completely desialylated by incubation with C. perfringens a-neuraminidase (determined by TLC analysis, results not shown) establishing that the linkage between sialic acid and galactose was in the a-anomeric configuration.

Sialylatwn of Glycosphingolipid Substrates-Paragloboside was tested and shown to be an acceptor-substrate for trans- sialidase. Activity was maximal in the presence of nonionic detergents such as Triton X-100 or Tween 80, but was reduced by 40% in the presence of 0.5% CHAPS, and was totally inhibited by 1% sodium deoxycholate. Fig. 4 shows an HP- TLC of the reaction products formed using 2,3-sialyllactose as the donor. The structure of the glycolipid product was deduced to be NeuAca2-3Galfi1-4GlcNAcfil-3Gal~l- 4Glcj3l-Cer by, (i) co-chromatography with authentic sialyl- paragloboside and (ii) susceptibility to Newcastle disease virus a-neuraminidase, an enzyme known to hydrolyze NeuAca2- 3, but not NeuAca2-6, linkages (25).

In the acceptor-specificity studies described above, sialyl- lactose and the acceptor oligosaccharide were each present at a concentration of 1 mM. At equilibrium, the concentration of the sialyllated product ranged from 0.45 to 0.55 mM, but the thermodynamics were such that the reaction could be driven to virtual completion by using an excess of 2,3-sialyl- lactose over the acceptor-substrate. For example, by using a 20 mM concentration of 2,3-sialyllactose, 95% of N-acetyllac- tosamine (initial concentration, 1 mM) could be converted to 2,3-sialyllactosamine (see Fig. 5).

Donor-Substrate Specificity In agreement with an earlier study (3), 2,6-sialyllactose was

found to be completely inactive as a donor-substrate for the trans-sialidase when N-acetyllactosamine or lactose were used

apparent temperature optimum. as acceptors. Hence, it appears that the enzyme has a strict

T. cruzi trans-Sialidase 9889

TABLE I Oligosaccharide acceptor-substrate specificity of T. cruzi trans-sialidase

Reaction mixtures containing trans-sialidase, 2,3-sialyllactose (donor-substrate), glucuronic acid (internal standard), and various saccharide acceptors were set up and analyzed using HPAEC/PED as described under “Experimental Procedures.” Incorporation of sialic acid into each acceptor has been expressed relative to the incorporation into Galfl1-4GlcNAcfll-3GaIfll-4Glc, which has been arbitrarily set at 100. The HPAEC retention times of 2,3-sialyllactose and glucuronic acid were 7.1 and 12.9 min, respectively.

Acceptor-substrate

Galfll-4GlcNAcfll-3Galfll-4Glc Galfll-4GlcNAc Galfl1-3GlcNAc~l-3GaIfll-4Glc Galfll-3GlcNAc Galfll-3GlcNAcfll-3Ga1fl1-4(Fucal-3)Glc Galfll-6GlcNAc Galpl-4Galfll-4Glc Galfll-4Man Galfll-3GalNAc Galfll-6Gal Galfl1-4(Fucal-3)GlcNAc~l-3Gal~l-4Glc Gal~l-3(Fucal-4)GlcNAcfll-3Galfll-4Glc Galfll-4gluconic acid, Gala1-4Ga1, Galal-GGlc, Gal. GalNAc. Man

Relative rate

Trans-sialidase n2,3-ST” Retention time

of product

100 58 35 4

35 45 41 25 35 30 0 0 0 0 0

35* 32 85

100 ND’

1 ND ND

7 ND

0 0 0

ND ND

rnin 5.9 6.2 9.2 8.1 5.5 6.7

11.7 5.5 6.1 6.1

a2,3-ST, a2,3-sialyltransferase. ‘Taken from Weinstein et al. (24).

ND, not determined.

PG

SPG

Lac 2,3-SL

0 0. N

0.

Tlme (mm)

FIG. 5. Effect of 2,3-sialyllactose concentration on the for- mation of 2,3-sialyllactosamine. Reaction mixtures in 50 mM cacodylate/HCI buffer, pH 6.9, containing 1 mM N-acetyllactosamine, 32 milliunits/ml trans-sialidase, 1 mg/ml BSA, and concentrations of 2,3-sialyllactose ranging from 1 to 20 mM were incubated a t 37 “c. Aliquots equivalent to 3 nmol of 2,3-sialyllactose were removed at the time intervals indicated and analyzed by HPAEC/PED (see “Exper- imental Procedures”). 0, rate curve a t 1 mM 2,3-sialyllactose; 0, 5 mM; A, 10 mM; A, 20 mM.

FIG. 4. HP-TLC of the trans-sialidase reaction products ob- tained using 2,3-sialyllactose and paragloboside as substrates. Paragloboside, 60 nmol, was dissolved in 60 pl of 0.1 M cacodylate/ HCI buffer, pH 6.9, containing 60 nmol of 2,3-sialyllactose, 0.3 milliunits of trans-sialidase, 30 pg of BSA, 0.1% Triton X-100, and 0.02% NaN3. The mixture was incubated at 37 “C for 18 h and the reaction products (aliquots equivalent to 2 nmol of paragloboside) analyzed by HP-TLC using a solvent of chloroform/methanol/water (6:4:1, v/v). Lane I, reaction mixture a t zero time. Lane 2, reaction mixture after 18 h. Lane 3, reaction mixture after 18 h treated with Newcastle disease virus neuraminidase (0.5 units/ml for 2 h a t 37 “C). Lane S, standards of paragloboside ( E ) , sialylparagloboside (SPG), lactose (Lac), and 2,3-sialyllactose (2,3-SL). An arrow marks the point of sample application.

specificity for donor-substrates having an a2,3-linkage be- tween sialic acid and galactose.

The standard assay (see “Experimental Procedures”) and an HPAEC/PED assay were used to test whether or not the tetrasaccharide, NeuAca2-3Gal~l-4(Fucal-3)Glc, was a do-

nor-substrate for trans-sialidase using lactose as the acceptor. Under the conditions used (substrates at 2 mM, incubation for 18 h at 22 “C with 10 milliunits/ml enzyme), neither assay could detect a sialylated product, indicating that fucosylation of the monosaccharide residue adjacent to NeuAca2-3Gal is refractory to the activity of trans-sialidase.

Recently, Parodi et al. (6) reported that, at a low micromolar concentration of N-acetyllactosamine, both 2’-(4-methylum- bellifery1)-a-N-acetylneuraminic acid and 2,3-sialyllactose were equally efficient as donor-substrates for trans-sialidase. In the present study, trans-sialidase was assayed against MeU-NeuAc and p-NP-NeuAc a t a near saturating concen- tration (2.5 mM, 2 K,) of N-acetyllactosamine. Table 11 shows that, compared with 2,3-sialyllactose, both synthetic a-sialo- sides were extremely poor sialic acid donors compared with 2,3-sialyllactose. In each case, the 14C-labeled sialylated prod- uct co-eluted (retention time of 6.2 min) with authentic 2,3- sialyllactosamine when analyzed by HPAEC/PED using an eluant of 150 mM NaOH, 100 mM NaOAc.

9890 T. cruzi trans-Sialidase

TABLE I1 Synthetic a-sialosides as donor-substrates for trans-sialidase

trans-Sialidase, 0.2 milliunits, was incubated for 60 min at 22 “C with 10 mM 2,3-sialyllactose, 2’-(4-methylumbelliferyl)-a-N-acetyl- neuraminic acid (MeU-a-sialoside) or 2-O-(p-nitrophenyl)-N-acetyl- a-neuraminic acid (p-NP-a-sialoside) and assayed against 2.5 mM [14C]N-acetyllactosamine as described under “Experimental Proce- dures.”

Donor-substrate (10 mM) Product

nmollh 2,3-Sialyllactose 16.9 MeU-a-sialoside 0.66 p-NP-a-sialoside 0.69

“ l

”.- I I

0.34 / I

- 0.21 II / I

m 0.0 4

0 20 4 0 6 0 8 0 100 120

Time (min)

FIG. 6. 12.ans-sialidase catalyzed hydrolysis of 2,3-sialyl- lactose in the presence and absence of the acceptor-substrate N-acetyllactosamine. Two incubation mixtures (0.1 ml) containing 50 mM cacodylate/HCl buffer, pH 6.9, 0.1 pmol of 2,3-sialyllactose, 0.3 milliunits of trans-sialidase, 1 mg/ml BSA, and including or lacking 0.1 pmol of N-acetyllactosamine were incubated at 37 “C. Aliquots equivalent to 3 nmol of 2,3-sialyllactose were removed at the times indicated and assayed by HPAEC/PED (see “Experimental Procedures”) for the formation of free sialic acid and 2,3-sialyllacto- samine. A, rate curve for the formation of 2,3-~ialyllactosamine; 0, rate curve for the release of free sialic acid in the presence of 1 mM N-acetyllactosamine; 0, rate curve for the release of free sialic acid in the incubation which lacked N-acetyllactosamine.

Neuraminidase Versus trans-Sialidase Actiuity trans-Sialidase is known to display neuraminidase activity

(1) which can be monitored using substrates such as 4-meth- ylumbelliferyl-a-sialoside (10). In the present study, HPAEC/ PED was used to monitor the rate of sialic acid release from 2,3-sialyllactose in the presence and absence of stoichiometric amounts of N-acetyllactosamine as the acceptor-saccharide. Fig. 6 shows that the rate of release of free sialic acid was very low (less than 3%) compared with that transferred to N- acetyllactosamine but was increased (to about 6%) in the absence of the acceptor. No neuraminidase activity could be detected using 2,6-sialyllactose as a substrate.

Kinetic Analysis of trans-Sialidase trans-Sialidase substrate kinetics were determined by ob-

taining initial rate data from a matrix of 2,3-sialyllactose and N-acetyllactosamine concentrations. The double reciprocal plots (see Fig. 7) indicate a bisubstrate sequential mechanism. Michaelis constants for 2,3-sialyllactose and N-acetyllacto- samine were determined to be 4.3 and 1.3 mM, respectively, with a corresponding Vmax of 400 pmol. min-’ . mg protein-’.

Immobilization of trans-Sialidase and Synthesis of NeuAca2- 3Galfi1-3GlcNAcfi1-3Galfi1-4Glc and NeuAca2-3Galfil-4Glc-

BSA Immobilization of trans-sialidase was achieved by incuba-

tion of the enzyme with concanavalin A-Sepharose in the

-1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0

IiSialyllactose (mM ”)

FIG. 7. Reciprocal plots of rate data with 2,3-sialyllactose as the varied substrate at different fixed concentrations of N- acetyllactosamine. Initial rates of sialic acid transfer to N-acetyl- lactosamine were measured as described under “Experimental Pro- cedures.’’ The fixed concentrations of N-acetyllactosamine were 0.38 (e), 0.58 (O), 0.98 (M), 1.78 (01, and 5.10 (A) mM.

0 4 8 1 2 1 6 20 2 4

Time (h)

FIG. 8. Sialylation of Gal@1-3GlcNAc/31-3Galj31-4Glc and GaU1-4Glc-BSA using concanavalin A-immobilized trans- sialidase. Gal~l-3GlcNAc~l-3Gal~l-4Glc and GalP1-4Glc-BSA were sialylated as described under “Experimental Procedures” and the course of each reaction monitored by HPAEC/PED. Results have been expressed as the percentage sialylation of the available galactose acceptor sites. 0, rate curve for the sialylation of lactose-BSA 0, rate curve for the sialylation of LNT.

presence of 2 mM CaC12 (see “Experimental Procedures”). Binding of the enzyme to the lectin was complete within 30 min but was accompanied by a 25% reduction in activity. This loss could not be attributed to any difference in temper- ature or pH optimum of the immobilized enzyme. The im- mobilized trans-sialidase showed no detectable loss of activity following storage for 6 months at 4 “C in the presence of 0.02% sodium azide.

Experiments were performed to investigate the potential of immobilized trans-sialidase as a synthetic reagent. Fig. 8 shows rate curves for the sialylation of lactose-BSA and LNT to generate products that are precursors of sialyl-Lewis’ and sialyl-Lewis”, respectively. Sialylation of lactose-BSA pro- ceeded to 95% of the theoretical maximum and this was achieved by using a 20-fold molar excess of 2,3-sialyllactose over the available galactose acceptor sites on the neoglycopro- tein. The product, which had a sialic acid content of 22 mol/ mol BSA, could be completely desialylated by Newcastle disease virus neuraminidase (results not shown), indicating that the structure of the enzymatically synthesized oligosac- charide linked to BSA was NeuAcaZ-3Galfi1-4Glc. In the reaction which included LNT, a %fold molar excess of 2,3- sialyllactose was used which resulted in 66% (13.2 pmol) of LNT being converted to NeuAca2-3Gal~l-3GlcNAcfi1- 3Galfil-4Glc.

T. cruzi trans-Sialidase 9891

DISCUSSION

This study shows that the trans-sialidase of T. cruzi dem- onstrates activity towards a wide range of oligosaccharide, glycolipid, and glycoprotein acceptors which terminate in a p- linked galactose residue. Direct and conclusive evidence, using methylation analysis, was also provided to establish that the enzyme exclusively catalyzes the synthesis of an a2,3-sialos- idic linkage. In many respects, therefore, T. cruzi trans- sialidase resembles rat liver a2,3-~ialyltransferase (25). Both enzymes are able to sialylate unsubstituted type-1 and type-2 lactosaminoglycans, but the corresponding fucosylated Lewis" and Lewis" structures are not substrates. In addition, however, the trans-sialidase is able to sialylate galactosides having subterminal @1-4Man, P1-4Ga1, Pl-GGal, and P1-6GlcNAc residues. Another notable difference between these two en- zymes is the ability of the trans-sialidase to synthesize a product where the a-anomeric configuration of the sialic acid is retained following transfer from the donor-substrate. In the case of the sialyltransferase, transfer of sialic acid from CMP- &sialic acid to the acceptor is accompanied by an inversion of anomeric configuration.

The present kinetic data suggest that a ping pong (enzyme substitution) mechanism which involves a sialylated enzyme intermediate is unlikely. However, the results do not differ- entiate between a random mechanism (which has been dem- onstrated to occur in the case of porcine submaxillary gland a2,3-sialyltransferase (26) and bovine colostrum a2,6-sialyl- transferase (27)) or other possible mechanisms which involve the ordered addition of the donor- and acceptor-substrates.

The ability of the trans-sialidase to function as a neuramin- idase has already been noted (11, 13). The results described above demonstrate that, in the presence or absence of a saccharide which is an acceptor-substrate for the trans-siali- dase, the hydrolase activity of the enzyme is extremely low and may therefore be of little physiological significance. How- ever, the removal of sialic acid from host cell glycans and the concomitant exposure of cryptic P-galactose residues, which occurs as a result of the trans-sialidase reaction, is of physio- logical significance to the parasite. Thus, sialylation of para- site surface glycans by the trans-sialidase was proposed to be a requirement for the parasite to become infective for cells in culture (3). Alternatively, the enzyme may sialylate host cell glycoconjugates to generate receptors used by the trypano- some to adhere to and penetrate target cells. Results with sialic acid-deficient mutants of Chinese hamster ovary cells support this hypothesis. For example, one such mutant, Lec2 (28), is a very poor host for adherence and invasion by T. cruzi when compared to parental K1 cells. However, after Lec2 cells are sialylated by the trans-sialidase, they support adherence and growth of the parasite to nearly the same extent as K1 cells.2

The unique ability of the trans-sialidase to utilize a2,3- sialylated sugars as donor-substrates prompted experiments to investigate the potential of the enzyme to transfer sialic acid from two synthetic a-sialosides, MeU-NeuAc and p-NP- NeuAc. The trans-sialidase was able to utilize either corn- pound to synthesize a product which corresponded to 2,3- sialyllactosamine. However, compared with 2,3-sialyllactose, the synthetic a-sialosides were 95% less active as sialic acid donors and are therefore likely to be of little practical value as substrates for the the large-scale synthesis of sialoglycans.

The immobilization of trans-sialidase on concanavalin A- Sepharose is rapid and simple, and presumably occurs via

* Ming, M., Chuenkova, M., Ortega-Barria, E., and Pereira, M. E. A. (1993) Mol. Biochem. Parasitol.. in Dress.

binding to oligomannosidic glycans which correspond to one or both of the two asparagine-linked glycans predicted to be present on the enzyme (4). The small reduction in activity which invariably accompanies immobilization of the enzyme is probably the result of steric or conformational constraint imposed upon the trans-sialidase as a result of the binding with concanavalin A. The suitability of the lectin-bound trans-sialidase for large-scale synthetic experiments was dem- onstrated by synthesizing an oligosaccharide precursor of sialyl-Lewis" and a glycoprotein precursor of sialyl-Lewis'. In each case, an excess of 2,3-sialyllactose was used to drive the equilibrium reaction in the direction of the desired product. No release of enzyme from the matrix could be detected following prolonged incubation at 37 "C during the course of the sialylation reactions (results not shown), demonstrating the high affinity and stability of the carbohydrate-lectin in- teraction. At present, the only practical limitation to using immobilized trans-sialidase for Gram scale synthesis is the availability of 2,3-sialyllactose which is currently isolated from bovine colostrum. The use of simple synthetic a-sialo- sides may solve this problem by providing sialic acid donors that are not substrates for the reverse reaction, i.e. compounds that, following transfer of sialic acid, do not act as acceptor- substrates.

Acknowledgments-We are grateful to Prof. Anne Dell for perform- ing fast atom bombardment-mass spectrometry analysis on the pu- rified sialylparagloboside, Drs. Gary Jacob and Stanton Dotson for helpful discussions during the course of this study, and Jean Rotsaert for skillful assistance in preparing the manuscript.

REFERENCES

2. Previato, J. O., Andrade, A. F. B., Pessolani, M. C. V., and Mendonca- 1. Pereira, M. E. A. (1983) Science 219,1444-1447

3. Schenkman, S., Jiang, M-S., Hart, G. W., and Nussenzweig, V. (1991) Cell

4. Pereira, M. E. A.. Santiago Meiia, J.. Orteaa-Barria. E.. Matzilevich. D..

Previato, L. (1985) Mol. Biochem. Purasitol. 16, 85-96

6 5 , 1117-1125

and Prioli, R. P. (1991)2. Exp. Me$. 174; 179-192 '

Mol. Biochem. Purasitol. 4 6 , 303-306

I ,

5. Rosenberg, I., Prioli, R. P., Ortega-Barria, E., and Pereira, M. E. A. (1991)

6. Andrew!, N. W., Hong, K. S., Robbins, P. W., and Nussenzweig, V. (1987) h'xp. Purasitol. 64,474-484

76-86

E. A. (1983) Hoppe Seyler's 2. Physiol. Chem. 364 , 1053-1057

O., and Frasch, A. C. C. (1992) EMBO J. 1 1 , 1705-1710

7. Schenkman, S., Diaz, C., and Nussenzweig, V. (1991) Exp. Purasitol. 72 ,

8. Schauer, R., Reuter, G., Muhlpfordt, H., Andrade, A. F., and Pereira, M.

9. Parodi, A. J., Pollevick, G. D., Mautner, M., Buschiazzo, A., Sanchez, D.

10. Cavallesco, R., and Pereira, M. E. A. (1988) J. Immunol. 140 , 617-625 11. Prioli, R. P., Santiago Mejia, J., and Pereira, M. E. A. (1990) J. Immunol.

12. Schenkman, S., de Carvalho, L. P., and Nussenzweig, V. (1992) J. Exp.

13. Libby, P., Ahoy, J., and Pereira, M. E. A. (1986) J. Clin. Invest. 77, 127-

144 , 4384-4391

Med. 175,567-575

135 14. Donald, A. S. R., and Feeney, J. (1988) Curbohydr. Res. 1 7 8 , 79-91 15. Veh, R. W., Michalski, J-C., Corfield, A. P., Sander-Wewer, M., Gies, D.,

16. Scudder, P., Neville, D. C. A., Butters, T. D., Fleet, G. W. J. Dwek, R. A. and Schauer, R. (1981) J. Chromatogr. 212,313-322

Rademacher, T. W., and Jacob, G. S. (1991) J. Biol. Chem: 266,16472- 16477

17. Hanfland, P. (1975) Chem. Phys. Lipids 16 , 105-124 18. Prioli, R. P., Ortega-Barria, E., Megia, J. S., and Pereira, M. E. A. (1992)

19. Scudder, P., Lawson, A. M., Hounsell, E. F., Carruthers R. A,, Childs, R.

20. Manger, I. D., Rademacher, T. W., and Dwek, R. A. (1992) Biochemistry

21. Smith, D. F., Zopf, D. A., and Ginsburg, V. (1978) Methods Enzymol. 5 0 ,

22. Dubois, M., Gilles, K. A., Rebers, P. A., and Smith, F. (1956) Anal. Chem.

Mol. Biochem. Purasitol. 62,85-96

A., and Feizi, T. (1987) Eur. J. Biochem. 168,585-59i

3 1 , 10724-10732

169-171

2 ~ . x n - ! m 23. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J.

24. Weinstein, J., de Souza-e-Silva, U., and Paulson, J. C. (1982) J. Biol. Chem.

25. Paulson, J. C., Weinstein, J. Dorland L., van Halbeek, H., and Vliegen-

26. Rearick, J. I., Sadler, J. E., Paulson, J. C., and Hill, R. L. (1979) J. Biol.

","" ""

Biol. Chem. 193,265-275

2 5 7 , 13845-13853

thart, J. F. G. (1982) J. d l . Chem. 2 5 7 , 12734-12738

Chem. 2 6 4 . 4444-4451 27. Paulson, J.-C., Rearick, J. I., and Hill, R. L. (1977) J. Biol. Chem. 252 ,

, ~~~~ ""

2363-2.171 28. Stanley, P. (1984) Annu. Reu. Genet. 1 8 , 525-552

~~~