ANALYTICALBIOCHEMISTRY

185,112-117

(1990)

An Enzyme-Linked lmmunosorbent N-Acetylglucosaminyltransferase-V’

Assay for

Suzanne C. Crawley,* Ole Hindsgaul,-fp2 Gordon Alton,? Michael Pierce,+ and Monica M. Palcic*y2 *Department of Food Science, University of Alberta, Edmonton, Alberta, Canada T6G 2P5, ~Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2, and #Department of Cell Biology and Anatomy, Univers& oi Miami School of Medicine, Miami, Florida 33;Ol

Received

August

25,1989

The development of an enzyme-linked immunosorbent assay (ELISA) for uridine 5’-diphospho-N-acetylglucosamine:amannoside j?l + 6 N-acetylglucosaminyltransferase (GnT-V) is reported. The assay quantitates the enzymatic conversion of the specific synthetic GnT-V acceptor GlcNAc@l + 2Mancul -P 6Man&R (5) to the product GlcNAcBl + 2[GlcNAc81+ G]Mancwl + 6Man&R (6) when these oligosaccharide structures were covalently attached to bovine serum albumin which was then coated on microtiter wells. Conversion of 5 to 6 was detected using a polyclonal antiserum raised against the product 6 and from which antibodies cross-reacting with acceptor 6 had been removed by affinity adsorption. GnT-V activity detected by ELISA was linearly proportional to both enzyme concentration and time under appropriate experimental conditions where 50-300 fmol of product was formed per microtiter well. GnT-V activity could be measured by ELISA in Triton X-100 extracts of hamster kidney acetone powder and in human serum. The twofold increase in GnT-V activity which is known to accompany Rous sarcoma virus transformation of baby hamster kidney cells could also be quantitated using the ELISA. Q 1090 academic press, IW.

UDP-GlcNAc3cw-mannoside /3(1 -W 6)-N-acetylglucosaminyltransferase (N-acetylglucosaminyltransferase-V, 1 This study was supported by a joint Strategic Grant from the Natural Sciences and Engineering Research Council of Canada to O.H. and M.M.P. and the Alberta Heritage Foundation for Medical Research. M.P. was supported by the National Institutes of Health Grant CA 35377 and is a recipient of the American Cancer Society Faculty Research Award. ’ To whom correspondence should be addressed. 3 Abbreviations used: GnT-V, N-acetylglucosaminyltransferase-V, ELISA, enzyme-linked immunosorbent assay; GlcNAc, N-acetyl-Dglucosamine; UDP-GlcNAc, uridine 5’-diphospho-N-acetylglucos112

GnT-V, EC 2.4.1.155) is a key enzyme involved in the branching of asparagine-linked oligosaccharides (1,2). Interest in the development of assays for GnT-V activity stems from several observations that transformation of cells by tumor viruses (3,4) or oncogenes (5) results in increased GnT-V activity and concomitantly altered cell-surface glycosylation. Increases in intracellular GnT-V activity have also been shown to correlate with the metastatic potential of both rodent and human tumor cells (6-8). Biosynthetically, GnT-V transfers a GlcNAc residue from UDP-GlcNAc to acceptors having the minimum oligosaccharide structure 1, converting it to structure 2 which bears the additional branch at the g-position of the Mancvl --* 6 arm (1,2) (see Scheme 1). Early assays for GnT-V involved quantitation of the transfer of radiolabeled GlcNAc to 1 or its derivatives isolated from natural sources, to produce labeled 2 (1,3,4). After separation of UDP-GlcNAc and its degradation products, 2 was then structurally characterized by its retention volume on Bio-Gel P-4 and/or reactivity toward glycosidases or lectins. Later assays involved HPLC techniques for separation of products (9,lO). Advantage was also taken in these assays of the observation that GnT-V was the only N-acetylglucosaminyltransferase which did not have an absolute requirement for Mn2+ and was therefore active in the presence of EDTA. Subsequently, the synthetic trisaccharide 3 was prepared and shown to be an acceptor specific for GnT-V, which converted it to the tetrasaccharide 4 (11-13). In radioactive assays, labeled product 4 could be readily quantitated after adsorption onto C-18 sample-preparation cartridges by virtue of the

amine; BSA, bovine serum albumin; BHK, baby RS-BHK, Rous sarcoma virus-transformed BHK, buffered saline; AU, absorbance unit; Mes, ethanesulfonic acid.

hamster kidney; PBS, phosphate2-[morpholinol-

0003-2697/90

$3.00

Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

ELISA

FOR

hydrophobic nature imparted to it by its lipid-like 8-methoxycarbonyloctyl aglycone (14). We report here the development of an enzyme-linked immunosorbent assay (ELISA) for GnT-V which utilizes a polyclonal rabbit antiserum for detection and quantitation of enzymatically produced 4 when it is covalently attached to BSA. The advantages of this ELISA over the previously reported radiochemical assays lie in the simultaneous structural identification of the product, its high sensitivity, and its amenability to extensive automation allowing the rapid measurement of GnT-V activity in a large number of samples. MATERIALS

AND

113

N-ACETYLGLUCOSAMINYLTRANSFERASE-V GlcNAcpb-ZManal-6, Ma@-4GlcNAc@-4GlcNAc$-Asn

1:

I f Fuca(l-6)

UDP-GlcNAc

Mxn~l-4GlcNAc~l-4GlcNAc~-Asn

2:

I f Fuca(l-6)

METHODS

Materials GlcNAc@l-2Manal-6,

3: R=OMe 5: R=BSA

Hamster kidney acetone powder extract was prepared as described for N-acetylglucosaminyltransferase I (15)) except that 2-[N-morpholinolethanesulfonic acid (Mes 50 mM, pH 6.5) was used for the wash and extraction steps instead of cacodylate buffer, and protease inhibitors (0.05 mg/ml of each of soybean trypsin inhibitor, aprotinin; and 0.1 mM phenylmethylsulfonyl fluoride) were incorporated during homogenization and in subsequent steps. Human serum used as a source of GnT-V was prepared by allowing blood to clot at room temperature for 2 h, refrigerating overnight at 4”C, and centrifuging to remove red blood cells. Serum samples were made 0.01% (w/v) in NaN3 and stored frozen at -20°C. Microsomal extracts of BHK and RS-BHK cells were prepared and assayed as previously described (12,13). Alkaline phosphatase conjugate of goat anti-rabbit IgG (whole molecule, adsorbed with human serum proteins) and alkaline phosphatase substrate tablets containing 5 mg p-nitrophenyl phosphate were from Sigma. [6-3H] UDP 3-N-acetylglucosamine (26.8 Ci/mmol) was from New England Nuclear and ACS scintillation cocktail was from Amersham. Removable flat-bottomed wells of Immulon 2 were from Dynatech. Chromosorb P was from Johns Manville. The following buffers were used: PBS, 7.8 mM Na2HP04, 2.2 mM KH2P04, 0.9% NaCl, and 15 mM NaN3, pH 7.4; and PBST, PBS with 0.05% Tween 20.

of 3 was covalently attached to silylaminated Chromosorb P beads as previously described (19) to provide an affinity matrix (20) with an incorporation of 0.35 gmol/ g. The rabbit antisera (1.0 ml) were added to 1.5-ml microfuge tubes containing 0.2 g of the affinity support derived from 3 and rotated end over end at 4°C for 4 h. After centrifugation, the supernatants were removed and used directly in the ELISA assays. One of the adsorbed antisera showed background binding to wells coated with 5 (Ado5 = 0.082 after 1 h of development) and was used in all subsequent GnT-V ELISAs.

Preparation

Plate Coating

of BSA Conjugates 5 and 6

Compounds 3 and 4 (11,12) were coupled to BSA via their acyl azides as described by Pinto and Bundle (16). The carbohydrate contents of 5 and 6 were determined using the phenol-sulfuric acid assay (17) with 3 and 4, respectively, as reference standards. Incorporations of 11-13 oligosaccharides per BSA molecule were achieved.

Rabbit Immunizations

and Refining

of Antisera

Three rabbits were immunized with 6 following the protocol described by Lemieux et al. (18). The acyl azide

Man@-O(CHZ)gCOR

GnT-V

UDP-GlcNAc

4: R=OMe 6: R=BSA

7:

Ma@-O(CH2)gCOR

GlcNAc@-2[6dcoxy-Mlal-6 ’ SCHEME

Man@-O(CH&CHj

I

Microtiter plates were coated as previously described (21) by incubation with 100 ~1 of synthetic BSA-glycoconjugate 5 or 6 (20 pg/ml) in 50 mM sodium phosphate buffer, pH 7.5, containing 5 mM MgClz and 15 mM NaN3 for 16 h at ambient temperature. This solution was then removed by aspiration and replaced with 5% BSA in PBS (200 ~1). After 4 h, this solution was removed and wells were washed three times with PBS (200 ~1) and once with Hz0 (200 pl), air-dried for 1 h, and stored at 4°C. Plates were washed again with Hz0 (200 ~1) immediately before use.

114

CRAWLEY

ET

AL.

all subsequent work. Sufficient antiserum was obtained from this one rabbit to perform 4 X 10” ELISAs for GnT-V assays were performed by adding buffer and GnT-V. enzyme directly to microtiter plates coated with 5 and Microtiter plates were coated with acceptor conjugate initiating the reaction by addition of UDP-GlcNAc or, 5 in admixture with O.l-0.8% of product 6 in order to alternatively, by combining buffer, enzyme, and UDPsimulate the plate that would result from the action of GlcNAc (final volume 100 ~1) and adding this mixture GnT-V on immobilized 6. In Fig. lA, product detection to the coated wells to initiate the enzyme reaction. All by the affinity-purified antiserum, amplified by alkaline assays, except the radiochemical, were carried out in phosphatase-conjugated goat anti-rabbit IgG, is seen to triplicate with the buffer composition, concentrations of be linearly proportional to the amount of coated product enzyme, and nucleotide donor noted in the figure legends up to about 0.5%. The slope of this standard curve varied for the different enzyme sources. The microtiter plate over the range 2.0 f 0.4 (Ados/% product 6) when new was incubated at 37°C for 60 or 90 min. The reaction batches of plates were coated and new solutions of mixture was removed by aspiration; wells were washed ELISA reagents were prepared. ELISA response, how(2X 200 ~1 Hz0 and 1X 200 ~1 PBST) and then incubated ever, always remained linear up to 0.5% product. The for 2 h at ambient temperature with the refined rabbit saturation at concentrations greater than 0.5% is not antiserum (100 ~1 of l/8000 dilution, in 1% BSA/ due to nonlinear microplate reader response since satuPBST). Wells were aspirated, washed with 5X 200 ~1 of ration was observed for readings taken at 40 and 50 min PBST, and then incubated with alkaline phosphatasewith lower absolute values. Figure 1B shows that, as exconjugated goat anti-rabbit IgG (100 ~1 of l/1000 dilupected, antibody detection of 6 is inhibited by the solution in 1% BSA/PBST) for 2 h. This solution was aspible hapten 4. The antiserum was therefore concluded to rated and wells were washed (3X 200 ~1 PBST, 1X 200 specifically recognize the product tetrasaccharide struc~1 HzO, and 1X 300 ~1 HzO) before addingp-nitrophenyl ture present in 6. phosphate substrate solution (1.0 mg/ml in 1 M diethaA Triton X-100 extract of hamster kidney acetone nolamine-HCl buffer, pH 9.8, containing 1% BSA and powder was used as a known source (14) of GnT-V in the 500 pM MgClJ. Increase in absorbance at 405 nm was experiments summarized in Figs. 2 and 3. Addition of monitored over time using a Bio-Tek EL-309 or EL-311 this GnT-V-containing extract and unlabeled UDP-GlcEIA plate reader. Data were acquired at lo-min intervals NAc to microtiter plates coated with 6 resulted in the and readings reported .are for 60 min, except where production of immobilized 6 as detected by the antisenoted. Both ELISA plate readers exhibited linear rerum. The amount of enzymatically formed product is sponse to an absorbance of 2.8. The experiments for the seen to be proportional to enzyme concentration in Fig. quantitation of radiolabeled GlcNAc transfer onto wells 2. In Fig. 3, the formation of product is shown to be linwere carried out at reduced nucleotide donor concentraear with time up to about 2 h. In order to obtain an estitions and included 0.8 &i [3H]UDP-GlcNAc in the inmate of the absolute amount of product being generated cubation mixtures. After color development, wells were by GnT-V, and detected in ELISA by the antiserum, the counted in 10 ml ACS cocktail with a Beckman LS1801 experiments summarized in Fig. 3 were performed using scintillation counter. The average background translabeled [3H]UDP-GlcNAc. After the absorbances of the ferred onto the microtiter wells for the radiochemical explate were measured using the ELISA reader, the radioperiments was 49 f 5 dpm, while enzyme incubations activity of individual wells was then quantitated by liqgave 45-204 dpm above background. A unit of activity is uid scintillation counting. The amount of radioactivity defined as the quantity of enzyme producing 1 nmol/h of incorporated onto each well is seen in Fig. 3 to parallel product. the product formation detected spectrophotometrically by ELISA. The ELISA response was found to be linear with the enzymatically generated product in the range RESULTS AND DISCUSSION of 50-300 fmol/well, corresponding to an absorbance of Acceptor 3 and product 4 were covalently attached to 0.2-1.2 at 405 nm after a l-h color development. The BSA to provide the synthetic glycoconjugates 6 and 6 linear range is dependent on the amount and source of enzyme, as well as the concentration of UDP-GlcNAc having 11-13 mol of carbohydrate per mole of BSA. Three rabbits were immunized with 6 resulting in the used and must be operationally defined for a given production of antisera containing antibodies which de- sample. The ELISA also detected GnT-V activity in human tected both 6 and 6 immobilized on microtiter plates. serum (Fig. 4A) where product formation was proporFor one of these antisera, the cross-reacting antibodies binding acceptor structure 6 could be completely re- tional to enzyme in the 0.2-0.7 AU response range for a moved by adsorption with an affinity matrix prepared by l-h ELISA development. Additional evidence that seimmobilization of trisaccharide 3 on Chromosorb P, a rum GnT-V was indeed responsible for the production of antibody-detected 6 was obtained by performing the calcined diatomaceous earth. The refined, monospecific antiserum thus obtained was used at l/8000 dilution in incubation in the presence of the deoxy-trisaccharide 7,

ELISA

for GnT- V

ELISA

FOR

115

N-ACETYLGLUCOSAMINYLTRANSFERASE-V 2.5

B

1.82 gc

1.5-

g %

1.2-

g s

$

0.9-

8 5

f e8

0.6 -

2

2.0

C

1.5

i 2 fl

0.3-

0.0’ 0.0

I

I

I

I

0.2

0.4

0.6

0.8

% immobilized

Product

0.5 1.0 i

o.ol0.0

0.2

6

0.4

0.6

Soluble Product

0.8

1.0

1.2

4 (mM)

FIG. 1. (A) Standard curve for ELISA response of wells coated with increasing ratios of product 6 and substrate 5. Absorbance was measured after a 60-min incubation withp-nitrophenyl phosphate after incubating with refined antiserum at l/8000 dilution, then alkaline phosphataseconjugated IgG antibodies as described in ELISA methods. (B) Competition of soluble product 4 with the bound BSA product conjugate 6 for antibody in the antibody-binding step of the ELISA. Assays were performed as described under Materials and Methods except that wells coated with 100% 6 were used, and in the primary antibody-binding step, 100 ~1 of l/8000 dilution of a5nity-purified antiserum was replaced with 50 ~1 of 11 to 109 nmol of 4 in 1% BSA/PBST plus 50 pl of l/4000 dilution of antibody.

an inhibitor specific for GnT-V (22). The amount of product formed using serum as the enzyme source decreased with increasing inhibitor concentration (Fig. 4B), as would be expected for the GnT-V-catalyzed reaction. The ELISA-detected serum GnT-V activity was also not inhibited by up to at least 20 mM EDTA, a characteristic property of the enzyme (1). The KM of the serum enzyme for UDP-GlcNAc was estimated, by ELISA, to be 0.3 mM (data not shown). To our knowl-

edge this is the first report of GnT-V activity in human serum. Finally, Rous sarcoma virus transformation of BHK cells has been shown (4) to result in a twofold increase in the specific activity of GnT-V, detected using radiochemical assays. Figure 5 shows that this increase was also detected by the ELISA using BHK and RS-BHK microsomes as the sources of GnT-V. The specific activ-

1.4 r

0.8 -

0.6 -

0.4 -

0.2 -

0.0 ’ 0.0

I

1

I

1

0.2

0.4

0.6

0.8

I

1.0

mU GnT-V FIG. 2. Effect of enzyme concentration on product formation. Enzyme incubations contained 0.22 to 0.88 mU of GnT-V enzyme from hamster kidney acetone powder extract (0.17 to 0.68 pg protein) and 563 nmol UDP-GlcNAc in 30 mM Mes buffer, pH 6.5, with 1% Triton X-100. Enzyme incubations were carried out at 37°C for 60 min. After aspiration, ELISA was carried out as described under Materials and Methods, with color development measured at 70 min.

z

1.0 -

-

0.8-

: $j

0.6-

b jcj a

0.4-

n

0.2I 60

Incubation

I 120

I 180

Time (min)

FIG. 3. Dependence of the ELISA response on time of GnT-V reaction (m). Wells coated with 5 were incubated with 20 mU of hamster kidney acetone powder extract, 2.8 nmol UDP-GlcNAc, and 0.8 &i [3H]UDP-GlcNAc, for the indicated times at 37°C. Absorbance was measured after a 70-min incubation with p-nitrophenyl phosphate. (0) After absorbance measurements, the amount of [3H]GlcNAc transferred to the wells was quantitated by liquid scintillation counting.

116

r

CRAWLEY

ET

AL.

0.6 -

$ $ 5 e s 2

0.4 -

0.2 -

0.0. 0.0

I 0.1

I 0.2

I 0.3

I 0.4

I 0.5

I 0.6

0.4 0.0

Serum Protein (mg)

I 0.5

Concentration

I 1.0

of GnT-V Inhibitor

I 1.5

I 2.0

7 (mM)

FIG. 4.

(A) Effect of serum concentration on GnT-V activity. Enzyme incubations contained 0.5 to 7 ~1 of serum (37 to 520 rg protein) and 11 nmol UDP-GlcNAc and were carried out in 75 mM sodium cacodylate buffer with 3.5 mu MnCl,, pH 7.2, for 60 min at 37’C. (B) Effect of the specific GnT-V inhibitor, 7, on apparent enzyme activity. Incubations contained 4 ~1 serum (298 pg protein), 352 nmol UDP-GlcNAc, and increasing concentrations of 7, in a total volume of 100 pl in 40 mM sodium cacodylate buffer, pH 7.2. After reaction at 37’C for 90 min, solutions were aspirated and ELISA was carried out as described under Materials and Methods.

ity of GnT-V in the RS-BHK cells was estimated by ELISA to be twofold higher than the activity in BHK cells using absorbance readings in the linear range approximately 0.2-0.8 AU. In conclusion, GnT-V activity can be specifically detected by ELISA in crude biological samples and the sen-

0.01 0

I 2

I 4

I 6

I 8

Protein (pg) FIG. 5.

Comparison of GnT-V activities of Rous sarcoma virustransformed (m) and non-transformed parental BHK cells (0) using microsomal extracts. Enzyme incubations contained 1.9 to 5.7 pg protein and 1.0 to 7.4 gg protein for RS-BHK and BHK extracts, respectively, and 11 nmol UDP-GlcNAc in 38 mM Mes buffer, pH 6.5, with 0.8% Triton X-100. Incubations were carried out for 60 min at 37”C, solutions were removed by aspiration, and ELISA was carried out as described under Materials and Methods, with absorbance readings taken at 65 min.

sitivity of this assay exceeds that of conventional radiochemical methods. GnT-V activity measured in this manner is proportional to enzyme concentration and incubation time over appropriate experimentally determined absorbance ranges. This ELISA is currently in routine use in our laboratories both for monitoring column fractions for GnT-V activity in the course of purification of the enzyme and in screening clinical serum samples for alterations in GnT-V activity. The approach described here using synthetic glycoconjugates as acceptor-product pairs for the development of an ELISA for GnT-V, coupled with earlier reports (21,23) of similar assays for both a galactosyltransferase and a fucosyltransferase, suggests the general applicability of ELISAs for the detection and quantitation of glycosyltransferase activities. A major impediment to the development of ELISA assays for glycosyltransferases, as described above, is the requirement of acceptor-product pairs of oligosaccharides which must generally be prepared by multistep chemical synthesis. The increasing commercial availability of synthetic oligosaccharides should, however, allow the development of such assays outside synthetic chemical laboratories. ACKNOWLEDGMENTS We thank Dr. P. Sporns for the use of the ELISA reader, Chembiomed for a gift of silylaminated Chromosorb P, and M. Shoreibah for the preparation of microsomes from BHK and RS-BHK cells. REFERENCES 1. Cummings,

R. D., Trowbridge,

I. S., and Kornfeld,

Biol. CVwn. 267,13,421-13,427. 2. Schachter, H. (1966) Bidsem. Cell. Bid. 64,163-181.

S. (1982)

J.

ELISA 3. Yamashita, K., Tachibana, Chem. 260,3963-3969. 4. Arango,

J., and Pierce,

5. Dennis, J. W., Kosh, Oncogene 4,853-860.

Y.,

and

M. (1988) K., Bryce,

Kobata,

D.-M.,

7. Dennis,

J. W., and Laferte,

8. Dennis,

J. W. (1988)

0. (1986)

O., Tahir, S. H., Srivastava, Res. 173,263-272.

13. Pierce, M., Arango, J., Tahir, them. Biophys. Res. Commun.

M. L. (1989) M.

L., and

Res. 49,945-950.

7,573-595. P. W., and Van den Eijnden,

J. P., and Schachter,

S. H., and Hindsgaul,

12. Hindsgaul, Carbohydr.

Suru.

J. Biol.

37,225-231.

C., Breitman, Cancer

A. H. L., Wijermans, FEBS Lett. 222,42-46.

10. Brockhausen, I., Carver, Cell. Biol. 66,1134-1151. 11. Tahir, 1780.

S. (1989)

Cancer

A. (1985)

and Breitman,

J. W., Laferte, S., Waghorne, R. S. (1987) Science 236,582-585.

Cunad.

117

iV-ACETYLGLUCOSAMINYLTRANSFERASE-V

J. Cell. Biochem.

6. Dennis, Kerbel,

9. Koenderman, D. H. (1987)

FOR

H. (1988)

Biochem.

J. Chem.

64,1771-

O., and Pierce,

S. H., and Hindsgaul, 146,679-684.

M. (1988)

0. (1987)

Bio-

14. PaIcic, M. M., Heerze, L. D., Pierce, Glycoconjugate J. 5,49-63. 15. Oppenheimer, 799-804.

C. L., and Hill,

M., and Hindsgaul,

R. L. (1981)

J. Biol.

0. (1988) Chem.

256,

16. Pinto, B. M., and Bundle, D. R. (1983) Carbohydr. Res. 124,313318. 17. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., and Smith, F. (1956) Anal. Chem. 3 1,296-305. 18. Lemieux, R. U., Bundle, D. R., and Baker, D. A. (1977) Conad. J. Biochem. 55.507-512. 19. Boullanger, P. H., Nagpurkar, A., Noujaim, A. A., and Lemieux, R. U. (1978) Canad. J. Biochem. 56,1102-1108. 20. Lemieux, R. U., Baker, D. A., Weinstein, W. M., and Switzer, C. M. (1981) Biochemistry 20,199-205. 21. Palcic, M. M., Ratcliffe, R. M., Lamontage, L. R., Good, A. H., Alton, G., and Hindsgaul, 0. (1989) Carbohydr. Res., 195, in press. 22. Palcic, M. M., Kaur, K. J., Shoreibah, M., Ripka, J., Hindsgaul, O., and Pierce, M. (1989) manuscript submitted for publication. 23. Stults, C. L. M., Wilbur, B. J., and Macher, B. A. (1988) Anal. Biochem. 174,151-156.