Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (14)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Sevigny, M. B.
Right arrow Articles by Troy, F. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sevigny, M. B.
Right arrow Articles by Troy, F. A., 2nd
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

`>
Glycobiology Pages 857-867  

Developmental expression and characterization of the [alpha]2,8-polysialyltransferase activity in embryonic chick brain
Introduction
Results
Discussion
Materials and methods
Acknowledgments
Abbreviations
References

Developmental expression and characterization of the [alpha]2,8-polysialyltransferase activity in embryonic chick brain

Developmental expression and characterization of the [alpha]2,8-polysialyltransferase activity in embryonic chick brain

Mary B.Sevigny, Jean Ye, Shinobu Kitazume-Kawaguchi2, Frederic A.Troy II1

Department of Biological Chemistry, University of California School of Medicine, Davis, CA 95616, USA

Received on October 31, 1997; revised on February 5, 1998; accepted on February 17, 1998

The [alpha]2,8-polysialyltransferases (polySTs) from embryonic chick brain catalyze the [alpha]2,8-specific polysialylation of endogenous neural cell adhesion molecules (N-CAMs). This posttranslation glycosylation decreases N-CAM-dependent cell adhesion and migration. The enzymatic properties of the membrane-bound form of the polyST activity was investigated in vitro. Our results show that the polyST activity was developmentally expressed with maximum specific activity appearing about 12 days after fertilization. This time shortly precedes maximal expression of the cognate polysialylated N-CAMs. Kinetic studies showed the KM and Vmax for CMP-Neu5Ac were 133 µM and 0.13 µM/h, respectively, at pH 6.1, 33°C. CMP-Neu5Gc was not a donor substrate. PolyST activity was increased 5- to 6-fold in the presence of 10 mM MnCl2, the preferred divalent cation, and 1 mM dithiothreitol (DTT). Heparin (3 kDa) was a noncompetitive inhibitor of polysialylation with a Ki of 9 µM. Based on the affinity of the enzyme for heparin, the polyST activity was partially purified (~30-fold) by heparin-Sepharose affinity chromatography, after differential solubilization with the zwitterionic detergent, CHAPS. DTT and chemical modification studies using the thiol-directed alkylating reagents, N-ethylmaleimide (NEM) and iodoacetamide (IAA), were used to show that at least one cysteinyl residue in the polyST was of critical importance for polysialylation, but of lesser importance for monosialylation, catalyzed by the [alpha]2,3-, [alpha]2,6-, and [alpha]2,8-monosialyltransferases (monoSTs). A sulfhydryl residue is implicated in chain initiation. Two important structural differences between the mono- and polySTs were revealed by sequence analyses. First, the polySTs contain heparin-like, positively charged amino acid clusters upstream of both sialylmotif L and S. Second, the polySTs contain a uniquely extended basic amino acid region (pI 11.6-12.0) of 31 residues immediately upstream of sialylmotif S. This extended, positively charged region may function in the processive mechanism of polymerization by allowing nascent polySia chains to remain bound to the polyST during the repetitive addition of each new Sia residue to the nonreducing termini of the growing chain. The importance of these studies is that they provide new information on the enzymatic basis of polysialylation. They also reveal that sulfhydryl residues and extended basic amino acid domains are two structural features unique to polysialylation, in contrast to monosialylation. Both may be important distinguishing features between the classes of distributive (monoSTs) and processive polysialyltransferases, which have not been previously described.

Key words: polysialic acid/polysialyltransferase/developmental expression/kinetic properties/cancer

Introduction

The [alpha]2,8-linked polysialic acid (polySia) glycotope covalently modifies surface glycoconjugates on cells as evolutionarily diverse as microbes and man (reviewed in Troy, 1992, 1995; Roth et al., 1993). These structurally unique, linear chains can contain more than 200 N-acetylneuraminic acid (Neu5Ac) residues (Rohr and Troy, 1980) that, interestingly, have a helical conformation similar to DNA (Michon et al., 1987; Yamasaki and Bacon, 1991; Brisson et al., 1992). PolySia was first discovered to be the capsular polysaccharide on neuropathogenic Escherichia coli K1 and Neisseria meningitidis Group B, and is a neurovirulence determinant associated with neonatal meningitis in humans (reviewed in Robbins et al., 1974; Troy, 1979). PolySia was later discovered to be present in the eggs of rainbow trout (Inoue and Iwasaki, 1978), in the jelly coat of sea urchin eggs (Kitazume et al., 1994a), in neural tissue of embryonic vertebrates (Finne, 1982; Vimr et al., 1984), and in some human cancers (Livingston et al., 1988; Roth et al., 1988; Grogan et al., 1994; Troy, 1995).

In vertebrates and in the sea urchin, polySia expression is developmentally regulated (McCoy and Troy, 1987; Lackie et al., 1993; Kitazume et al., 1994b; Cho et al., 1996). In vertebrates, two CMP-Sia:[alpha]2,8-polysialyltransferases (polySTs), designated STX and PST, have been cloned and shown to be involved in synthesis of the polySia chains (Livingston and Paulson, 1993; Eckhardt et al., 1995; Kojima et al., 1995; Nakayama et al., 1995; Scheidegger et al., 1995). Based on the deduced protein sequences, both polySTs appear to be type II integral membrane glycosyltransferases that are located in the Golgi complex. In vertebrates and sea urchins, polySia is maximally expressed during early organogenesis, and its expression is greatly reduced after embryonic development (McCoy and Troy, 1987; Kitazume et al., 1994b; Oka et al., 1995; Cho et al., 1996). Because of this pattern of temporal expression, when abnormally high levels of polySia were found expressed in some adult human tumors, e.g., neuroblastomas, nephroblastomas, and malignant lymphomas, it was concluded that polySia was an oncodevelopmental antigen (Livingston et al., 1988; Roth et al., 1988; Grogan et al., 1994).

When expressed on mammalian cell surfaces, polySia is most often attached to N-linked oligosaccharides on neural cell adhesion molecules (N-CAMs) and the [alpha]-subunit of the voltage-dependent sodium channel (Hoffman et al., 1982; Zuber et al., 1992). Its likely attachment to O-linked oligosaccharides on glycoproteins secreted by human breast and leukemia cells has also been reported (Martersteck et al., 1996). The extended polySia chains on N-CAM are postulated to prevent the intercellular homophilic binding that usually occurs between N-CAM-expressing cells (reviewed in Edelman, 1985; Rutishauser et al., 1988). As a consequence, highly polysialylated N-CAM is believed to play a central role in neurite fasciculation, neuromuscular interactions, and cell migration (Rosenberg et al., 1986; Rutishauser et al., 1988; Seki and Arai, 1993; Dubois et al., 1994; Tang et al., 1994; Wang et al., 1994), and also in the enhancement of the metastatic potential of certain human cancers (McCoy and Troy, 1987; Roth et al., 1988; Grogan et al., 1994; Scheidegger et al., 1994; Martersteck et al., 1996).

PST in hamster and human encodes a 359 amino acid (41.2 kDa) protein, while the human STX encodes a 375 amino acid (42.4 kDa) protein. PST and STX are members of a vertebrate multigene family consisting of known [alpha]2,3- and [alpha]2,6-monoSTs. Both contain the conserved sialylmotifs L (long) and S (short), and a hydrophobic stretch of 13 amino acids located within the N-terminal domain. In spite of these rapidly occurring advances in the cloning and sequencing of the polySTs, these transferases have not been studied extensively in their membrane environment, nor purified. Further, although several biochemical studies were carried out when the fetal rat brain polyST was first discovered (McCoy et al., 1985), and more recently on the chick brain polyST (Oka et al., 1995), there is a paucity of information on the enzymatic properties of these transferases in their native membrane environment. Also lacking is detailed structural information on the length of the polysialylated chains synthesized by the membrane-bound and soluble forms of the cloned polySTs, and the molecular mechanism whereby these enzymes catalyze polySia chain initiation, the processive chain polymerization and chain termination reactions. As a first step in understanding some of these important aspects of polysialylation, this study describes the developmental expression of polyST activity in embryonic chick brain, and the enzymatic properties of the polyST in its membrane environment, in vitro. We also report the results of chemical modification studies and gene sequence analyses that have provided new insights into the critical importance of cysteinyl residues and extended basic amino acid clusters in polysialylation. Finally, we report that the polyST can be differentially solubilized with CHAPS and partially purified by heparin-Sepharose affinity chromatography.

Results

Developmental expression of the embryonic chick brain polysialyltransferase activity

To determine if polyST activity was developmentally expressed in embryonic chick nervous tissue, brains were isolated at various stages of embryonic development, ranging from 5 to 18 days postfertilization. PolyST activity was measured using CMP-[14C]Neu5Ac as substrate, as described under Materials and methods. To differentiate the Sia residues incorporated into polySia by the polyST from those incorporated into other glycoconjugates by the [alpha]2,3-, [alpha]2,6-, and [alpha]2,8-monosialyltransferases (monoSTs), the radiolabeled products were analyzed using the [alpha]2,8-specific reagent, endo-N-acylneuraminidase (Endo-N). This endosialidase is a diagnostic enzyme specific for measuring [alpha]2,8-polyST activity because it cleaves randomly only oligo- or polySia residues in [alpha]2,8 ketosidic linkages (Vimr et al., 1984; Ye et al., 1994). Thus, in incubation mixtures treated with Endo-N prior to chromatography, most of the polySia residues are depolymerized by Endo-N. The released sialyl oligomers are chromatographically mobile and do not remain at the origin. In contrast, Sia residues incorporated into glycoconjugates by the monoSTs are chromatographically immobile and remain at the origin. Therefore, the polyST activity was quantitated by determining the total amount of radioactivity at the origin (no Endo-N treatment) and subtracting the amount of radioactivity remaining after Endo-N treatment. The difference represents the polyST activity, as described under Materials and methods. Figure 1A shows the specific activity of the membrane-bound polyST at each developmental stage. The highest level of activity was reached at around the 12 day stage. This result correlates positively with maximal expression of the cognate polysialylated N-CAMs, as described below. These results are also in accord with our previous finding that expression of the rat brain polyST activity was developmentally regulated, and mostly restricted to an early stage in development (McCoy and Troy, 1987). Tsuji and colleagues (Tsuji et al., 1997) have recently shown that STX is the polyST activity that is preferentially expressed in the early postnatal developing mouse brain. No equivalent information is available as to if chick brain expresses homologs of STX or PST. Thus, the results reported in this article may represent contributions from one or more polyST activities. This is a potential advantage when seeking to determine activities of processive transferases in their native membrane environment because it more likely reflects functional correlates of the in vivo situation than parameters derived from using purified forms of detergent solubilized or soluble constructs of the same enzymes. Such baseline data will also be essential for comparing individual parameters when pure polySTs might eventually become available.


Figure 1. Developmental expression of the polyST activity in embryonic chick brains.(A) Fresh chick brain homogenates prepared from different embryonic stages of development were assayed for polyST activity, as described under Materials and methods. Incubations were carried out at 22°C prior to treatment with Endo-N. Results are expressed as specific activity (pmol [14C]Neu5Ac incorporated/mg protein per hour). (B) Immunoblot analysis of chick brain homogenates at different stages of embryogenesis was carried out using the anti-polySia monoclonal antibody, 12E3, as described previously (Ye et al., 1994). Each lane represents a stage after fertilization. Lane 1, day 5; lane 2, day 7; lane 3, day 9; lane 4, day 11; lane 5, day 14; lane 6, day 18; and lane 7, day 21.

To correlate the pattern of temporal expression of the polyST activity with expression of the polySia glycotope, chick brains from different stages of embryonic development were analyzed for the presence of polySia by SDS-PAGE and immunoblotting, using the monoclonal anti-polySia antibody, 12E3, as described previously (Ye et al., 1994). As shown inFigure 1B, the highest level of polySia expression occurred around day 14, which follows by 1-2 days the period of maximal expression of the polyST activity noted above. As shown in Figure 1, there was a significant decrease in expression of both the polyST and polySia after about day 16-18. On the basis of these results, we conclude that expression of the chick brain [alpha]2, 8-polyST activity is developmentally regulated, with maximal expression of the enzyme and its cognate polysialylated glycans occurring between the twelfth and fourteenth day of embryogenesis, respectively. This pattern of temporal expression is similar to the developmentally regulated expression of the rat brain [alpha]2,8-polyST responsible for polysialylation of N-CAM, as described previously (McCoy and Troy, 1987). It differs, however, from the results reported by Oka et al., who concluded that polyST activity in chick brain was maximal at the 5 day embryonic stage (Oka et al., 1995). These latter studies were carried out using Nonidet P-40 solubilized enzyme and N-CAM as an exogenous acceptor. As such, the apparent discrepancy may relate to detergent-induced differences in the activity of the soluble form of the enzyme, and/or to the fact that their specific activities were normalized to the amount of endogenous N-CAM present at each stage of development. This latter correction could err by underestimating the specific activity of the polyST in its membrane environment, because such a correction would decrease the specific activity as the concentration of N-CAM increases, which it does in the later stages of development (Figure 1B).

Characteristics and kinetic properties of the membrane-bound polysialyltransferase activity in embryonic chick brain

Effect of pH and buffers on polysialyltransferase activity.Optimum pH and buffer conditions were determined by first analyzing polyST activity over the pH range of 6.1-10.4. The Good buffers MES, PIPES, HEPES, TRICINE, CHES, and CAPS were used at pH values equal to their respective pKa values. These studies showed a nearly equal and optimum pH for the polyST activity at pH 6.1 in MES buffer, or pH 6.8 in PIPES buffer. To determine if there was a differential pH optimum for the mono- and polyST activities, the two different activities were measured in MES buffer, pH 5.5, 6.1, and 6.8; PIPES buffer, pH 6.8 and 7.5; and HEPES buffer, pH 7.5 and 8.1. While the monoST activity was also optimum at pH 6.1 in MES buffer, it was greatly reduced at pH 6.8 in PIPES buffer. Therefore, since the polyST to monoST activity ratio was substantially increased at pH 6.8 in PIPES buffer, compared to pH 6.1 in MES buffer, the former buffer and pH is the optimum one for determining polyST activity in embryonic chick brain membranes.

Effect of divalent cations on polysialyltransferase activity. Using the aforementioned polyST assay, the effect of various divalent cations on the polyST activity was determined. Various concentrations (0-10 mM) of Ca+2, Mg+2, and Mn+2 were tested separately for their effects on polyST activity. Calcium showed very little effect, and Mg+2 resulted in only a slight increase in activity. In contrast, Mn+2 at 10 mM stimulated polyST activity nearly 5-fold. Unless noted otherwise, all standard polyST incubation mixtures contained 10 mM MnCl2.

Kinetic parameters: effect of CMP-Neu5Ac concentration. The effect of CMP-Neu5Ac concentration on the membrane-bound form of the chick brain polyST activity is shown in Figure 2. The KM and Vmax for CMP-Neu5Ac was determined to be 133µM and 0.13µM/h, respectively, at 33°C, pH 6.1. This KM value is about 60% higher than the KM of the E.coli K1 polyST for CMP-Neu5Ac (81 µM), and the bacterial polyST is catalytically about three times as active, with a Vmax of 0.38 µM/h (Vijay and Troy, 1975). PolySia formation remained linear for up to 2-3 h of incubation at protein concentrations up to 2.8 mg/ml, a finding similar to the fetal rat brain polyST activity (McCoy and Troy, 1987). In contrast to the E.coli K1 and rainbow trout polySTs, the chick brain polyST activity did not recognize CMP-Neu5Gc as a sugar nucleotide donor substrate.


Figure 2. Effect of CMP-Neu5Ac concentration (S) on reaction velocity (V) for the polyST activity.A Lineweaver-Burk plot (1/v vs. 1/[substrate]) for CMP-Neu5Ac is plotted to determine the KM and Vmax values of the polyST activity.

Conserved cysteine residues and the effect of DTT and alkylation of these residues on polysialyltransferase activity.There are six conserved cysteine residues in the mammalian polyST, PST (residues 11, 142, 156, 169, 292, and 356; Eckhardt et al., 1995; Nakayama et al., 1995; Yoshida et al., 1995). Three of these residues are in sialylmotif L, the proposed CMP-Sia binding site. To determine if a sulfhydryl group was important for polysialylation, thiol reduction and alkylation studies were carried out using DTT and the thiol-directed alkylating reagents, NEM and IAA. While the addition of DTT to the incubation mixture was not obligatory for polysialylation (Figure 3), 1 mM DTT stimulated polyST activity over 2-fold, yet had little effect on monosialylation. Further, 80 mM DTT inhibited the monoST activities over 30%, but had no effect on polysialylation. The stimulatory effect of DTT on polysialylation was variable, ranging from little effect to over a 2-fold increase in polyST activity. This variation could reflect differences in the redox potential of the enzyme, depending on variations in the reducing environment of the membrane fraction, which may occur because of time differences between cell disruption and assay.


Figure 3. Effect of dithiothreitol (DTT) concentration on the monoST and polyST activities. Standard incubation mixtures were prepared containing thedifferent concentrations of DTT, as indicated. The monoST and polyST activities were measured as described under Materials and methods. Open circles, monoST activity; solid circles, polyST activity.

The central importance of cysteine residues for maximal polyST activity was confirmed in studies using the alkylating reagents, NEM and IAA. As shown in Figure 4A, preincubation of the polyST-enriched membrane fraction with 1 mM NEM and 1 mM DTT for 2 h at 25°C, prior to addition of CMP-[14 C]Neu5Ac, stimulated polyST activity 1.3-fold, yet had little effect on the monoST activities. Incubation with 5 mM NEM and 1 mM DTT, however, inhibited completely the polyST activity (Figure 4A). Similar results were obtained with IAA (1-10 mM), in either the presence or absence of 1 mM DTT, although complete loss of activity was not obtained at 10 mM. These modification studies were carried out in either the presence or absence of 1 mM DTT in MES buffer at pH 6.1 because this lower pH favors alkylation of most available sulfhydryl groups by NEM and IAA (Means and Feeney, 1971).


Figure 4. ffect of N-ethylmaleimide (NEM) on monoST and polyST activities. (A) Standard incubation mixtures were prepared containing 1 mM DTT and the different concentrations of NEM, as indicated. The membrane fraction was preincubated for 2 h at 25°C prior to addition of CMP-[14C]Neu5Ac. Specific activities were determined 2 h after addition of substrate. Open circles, monoST activity; solid circles, polyST activity. (B) To determine the effect of DTT on the rate of inactivation of the polyST by NEM, 2.5 mM NEM was added to the polyST incubation mixture in either the presence or absence of 1 mM DTT. The incubation mixtures were incubated at 25°C for 0, 0.5, 1, 2, or 4 h prior to addition of CMP-[14C]Neu5Ac. Control incubations were run without NEM. Because the polyST activity was essentially eliminated at 1 h, the 2 and 4 h time points are not shown. Solid square, control, no NEM plus 1 mM DTT; solid circle, control, no NEM and no DTT; open squares, 2.5 mM NEM plus 1 mM DTT; open circles, 2.5 mM NEM alone.

To determine the effect of DTT on the rate of inactivation of the polyST by NEM, the membranous enzyme fractions were preincubated with 2.5 mM NEM for 0, 0.5, 1, 2, and 4 h at 25°C. Incubations were carried out in the presence and absence (control) of 1 mM DTT, prior to addition of CMP-[14C]Neu5Ac. As shown in Figure 4B,whenthe enzyme was incubated with 2.5 mM NEM and 1 mM DTT without preincubation, there was a 36% decrease in polyST activity (squares). In contrast, in the absence of DTT, the activity was reduced by nearly 75% (circles). Also, there was nearly a complete loss in polyST activity after a 1 h preincubation with NEM, in either the presence or absence of DTT, prior to addition of substrate. On the basis of these results, we conclude that one or more of the cysteine residues in the polyST must remain in the reduced sulfhydryl state for optimum polyST activity. This unexpected finding is consistent with the possibility that at least one cysteinyl residue may be required for polySia chain initiation, and/or for facilitating the processive mechanism of chain polymerization, as discussed below. These two processes are unique to polysialylation, in contrast to monosialylation, and may be an important distinguishing feature between the classes of distributive (monoSTs) and processive polysialyltransferases, which have not been previously described.

Basic amino acid clusters unique to the polySTs and their relationship to heparin inhibition of polysialylation. Our analyses of the deduced protein sequences of PST and STX revealed that both polySTs contained three potential heparin-like binding sites (Sevigny and Troy, 1997), based on the presence of the conserved consensus recognition sequences for heparin binding, [-X-B-B-X-B-X-] and [-X-B-B-B-X-X-B-X-], where B represents basic amino acid residues (Cardin and Weintraub, 1989). Our analyses further revealed that in the hamster and human PST one of these motifs, [135-KNRRFK-140], was located immediately upstream and contiguous with the sialyl motif L and that a second, [246-KNKLKVR-252], was located just 24 amino acid residues upstream of sialyl motif S. Neither of these two basic amino acid clusters, with pIs of 12.5 and 11.7, respectively are present in the human [alpha]2,3 or [alpha]2,8 monoSTs. The human [alpha]2,6 monoST, however, has three positively charged clusters [27-KEKKK-31; pI 10.51], [67-PHRGRQ-72; pI 12.51], and [268-YRKLH-272; pI 10.34], but none of these sequences are contiguous with sialyl motif L or just upstream of sialylmotif S, as in the polySTs.

A second distinctive structural feature unique to PST and STX, and also absent in the monoSTs, is an extended basic amino acid region of 31 amino acids (residues 246-277 in PST; pI 11.6 and residues 261-292, pI 12.0 in STX). This region is located immediately upstream and contiguous with sialylmotif S, and includes the second domain of the positively charged amino acids noted above. Thus, this region in the polySTs can be considered as constituting an extended sialylmotif S, wherein the distal 31 amino acids have a pI of 11.6-12.0, while the proximal 24 amino acids (sialylmotif S) have a pI of 4.2-4.6. On the basis of these new findings, we hypothesized that one or more of these positively charged clusters may function in the binding of the growing polySia chain to facilitate the processive mechanism of chain polymerization, a mechanism unique to polysialylation in contrast to monosialylation, and in the binding of CMP-Neu5Ac residues. To determine if the heparin-like binding sites were functionally important in polysialylation, we tested various concentrations of heparin (3 kDa and 6 kDa) for their potential inhibitory effects on the embryonic chick brain polyST. Our initial studies showed that heparin strongly inhibited polysialylation. Further studies were carried out to determine the nature of this inhibition. Increasing concentrations of heparin were tested with increasing concentrations of CMP-Neu5Ac to determine: (1) if heparin inhibition was competitive, noncompetitive, or uncompetitive; and (2) the Ki for heparin. Figure 5A, shows a Henri-Michaelis-Menten plot for the polyST activity in the absence and presence of 15 µM and 30 µM heparin (3 kDa). Both concentrations showed a greatly reduced Vmax compared to control (no heparin added), whereas the KM remained the same for all three concentrations. A Dixon plot (1/V vs. [Inhibitor]) for heparin in the presence of different fixed concentrations of CMP-Neu5Ac was constructed to determine the Ki for heparin (Figure 5B). These data show that heparin is a noncompetitive inhibitor of the chick brain polyST activity, with a Ki of 9 µM (Figure 5B).Comparing this Ki value with the KM for CMP-Neu5Ac (133 µM), it appears that the polyST has a 14-fold greater affinity for the polyanionic heparin polymer than its affinity for CMP-Neu5Ac. That the heparin inhibition was noncompetitive is consistent with our supposition that one or more of the heparin binding sites on the polyST may be involved in interactions with the polySia chain, and that the enzymes high affinity for heparin may mirror its affinity for the polySia chain. In this case, it would appear that CMP-Neu5Ac binding, presumably to the positively charged amino acid residues within or contiguous with sialyl motif L (Datta and Paulson, 1995) may be more rate-limiting than the processive functions of the enzyme during polySia chain polymerization. Heparin is also an effective inhibitor (Ki 9 µM) of in vitro polysialylation in neuroinvasive E.coli K1, catalyzed by the bacterial polyST (NeuS) complex.


Figure 5. Effect of heparin on the kinetics of polysialylation.(A) Henri-Michaelis-Menten plots (v vs. [CMP-Neu5Ac]) for the polyST in the presence and absence of 3kD heparin. Circles, no heparin added (control); triangles, 15µM heparin; squares, 30 µM heparin. (B) Dixon plot (1/v vs. [Inhibitor]) for heparin in the presence of different fixed concentrations of CMP-Neu5Ac was constructed to determine the Ki for heparin.

Two other anionic compounds, the glycosphingolipid, GD3, and dolichylphosphate also inhibited the chick brain polyST activity. This is of particular interest because, even though bacteria do not contain gangliosides, the E.coli K1 polyST can use a number of gangliosides, including GD3, as an exogenous acceptor substrate, at an optimum concentration of 325 µM (Cho and Troy, 1994). In the present experiments, various concentrations of GD3 (143-571 µM) were tested to determine if the chick brain polyST could recognize GD3 as an exogenous acceptor substrate. Surprisingly, as shown in Figure 6, GD3 inhibited polyST activity in a concentration dependent manner.


Figure 6. Effect of the ganglioside, GD3, on the polyST activity. Various concentrations of GD3 (143-570 µM) were added to the polyST incubation mixtures prior to addition of CMP-[14C]Neu5Ac. The polyST activity was then determined as described under Materials and methods.

Dolichylphosphate plays a major role in synthesis of N-linked oligosaccharides in the rough endoplasmic reticulum (Hanover and Lennarz, 1982). Although localization of this polyisoprenol is believed to be restricted to the ER, considerable evidence showing the presence of dolichol in the Golgi has been presented (Eggens et al., 1983). Therefore, to determine if the activity of the Golgi-localized polyST was influenced by the exogenous addition of this glycosyl carrier lipid, different concentrations of dolichylphosphate (2-100 µM) in 0.017% Triton CF-54 were added to the polyST incubation mixture. These studies showed that dolichylphosphate at a concentration of 20 µM or higher inhibited the polyST activity by about 40%.

Detergent solubilization and partial purification of the embryonic chick brain polysialyltransferase activity.Despite the recent cloning of the human, hamster, and mouse polyST genes, and expression of soluble constructs of the polySTs they encode, there remains a dearth of information about their mechanism of catalysis, since their catalytic properties in the membrane have not been well characterized. Further, none of these cloned polyST activities have been purified. Because the chick polyST has not been purified, we have undertaken a biochemical approach to isolate the chick brain enzyme to begin a comparison of its catalytic properties with that of the membrane-associated form of the same enzyme. This comparison is particularly important to understand if soluble forms of the enzyme, either detergent solubilized or soluble constructs expressed in mammalian cell lines, catalyze the same processive mechanism of chain polymerization as does the membranous form of the polyST. There is ample evidence that soluble forms of the distributive [alpha]2,3- and [alpha]2,6-monoSTs retain their linkage specificity (Weinstein et al., 1987).However,there are no detailed published studies that we are aware of which show that the soluble polyST constructs actually catalyze the transfer of multiple Sia residues via a processive mechanism of synthesis, in contrast to catalyzing the transfer of only one or a few Sia residues to preexisting [alpha]2,8-oligo- or polySia chains (distributive mechanism of synthesis). The lack of this information leads to a potential flaw in some of the published studies claiming polysialylation of exogenous N-CAM by either detergent solubilized or soluble constructs of cloned polySTs. These studies have used Endo-N as the diagnostic indicator of polysialylation, but ignore the possibility that transfer of a few, or even one [14C]Sia residue to preexisting polySia chains, will give the same appearance of radiolabeled polysialylated structures that are sensitive to Endo-N, and with the expected subsequent loss of high Mr signals in Western blots probed with anti-N-CAM or anti-polySia antibodies. Thus, ambiguity concerning studies to determine 'how many enzymes does it take to polysialylate" will not be resolved until, minimally, structural studies are carried out to determine the degree of polymerization of polySia chains synthesized in vitro by soluble and membrane-bound forms of the same enzyme.

Selective solubilization of the membrane-associated polyST activity from embryonic chick brain using CHAPS. Fourteen-day embryonic chick brain was used as the source of the polyST because both the level of enzyme and endogenous polysialylated N-CAM acceptors were nearly maximal at this stage of development (Figure 1). Frozen rather than fresh tissue was used because the monoST activities were selectively reduced about 3.6-fold after one freeze-thaw cycle, compared to the polyST activity. Freeze-thawing resulted in less than a 10% reduction in polyST activity, thus increasing the ratio of polyST to monoST activity.

To screen a number of detergents for their potential efficacy for solubilization, detergents were added to standard incubation mixtures at various concentrations, ranging from 0 to 2% (w/v), and their effect on the activity of the membrane-bound polyST was determined. The detergents tested included CHAPS, octyl [beta]-glucoside, saponin, and Triton CF-54. The concentration dependency of CHAPS at 0.5% (w/v), which represents a detergent to protein molar ratio of 62:1, was the only detergent tested that stimulated polyST activity (~1.7-fold). Subsequent studies confirmed the importance of the 62:1 molar ratio of CHAPS to protein, and also showed that the specific activity of the enzyme was increased 3-fold when the membrane fraction was preincubated with CHAPS at 4°C for 1 h, prior to centrifugation (Table I). Preincubation at 33°C for 1 h, or doubling the CHAPS concentration, resulted in only a 1.9-fold and 1.8-fold increase in activity, respectively. These studies also revealed that the polyST activity was selectively solubilized relative to the monoST activity, after a 1 h preincubation at 4°C with CHAPS. Under these conditions, the solubilized form of the polyST accounted for 86% of the total sialyltransferase activity, whereas in the absence of CHAPS, it accounted for only 19% of the total activity.

Partial purification of the polyST activity by heparin-Sepharose affinity chromatography. The CHAPS solubilized form of the polySTactivity from 14-day-old embryonic chick brain was partially purified (about 30-fold) by heparin-Sepharose affinity chromatography, as described under Materials and methods. This value likely underestimates the fold purification because of a loss in activity associated with having to dialyze the enzyme free of NaCl before assay. The relative instability of the partially purified enzyme has hindered subsequent purification procedures.Comparative properties of the polysialyltransferases and monosialyltransferases. Table II summarizes the properties of the membrane-bound [alpha]2,8 polyST activity from 14-day-old embryonic chick brain, and compares these properties with that of the monoSTs. This comparison revealed four important differences between these two classes of sialyltransferases. First, the most notable functional difference was the requirement for at least one sulfhydryl residue for optimum polyST activity. This was observed in the differential effect on activity by both DTT and NEM. For example, 1 mM DTT and 1 mM NEM stimulated the polyST activity 2-fold and 30%, respectively, yet neither reagent at 1 mM had much effect on the monoST activities. And while 80 mM DTT inhibited the monoST activity over 30%, it had no effect on polyST activity. There were also some differences between the effect of metal ions on mono- and polyST activities, Ca2+ being the most notable. Second, sequence analysis showed significant structural differences in the basic amino acid clusters upstream and contiguous with sialylmotif L, and the stretch of 24 amino acids upstream of sialmotif S in the polySTs. Similar sequences were not present in the [alpha]2,3- and [alpha]2,8- monoSTs. Third, sequence analysis also revealed an extended basic amino acid region that consisted of 31 amino acids that was present in both PST and STX, but absent in the monoSTs. The calculated pI of this region, 11.6 in PST and 12.0 in STX, likely contributes to the distinctly higher isoelectric points of the polySTs (pI 9.8-10.2) compared with the monoSTs (pI 8.3-9.8). This overall difference in pI values suggests a correlation between function and pI. The functional correlate may relate to the binding by the polySTs of the growing polySia chain to facilitate the processive mechanism of chain polymerization, an event unique to polysialylation, and which is fundamentally distinct from the distributive mechanism of synthesis catalyzed by the monoSTs. Fourth, the monoST activities were found to be considerably more sensitive to freeze-thawing of the embryonic chick brain than the polySTs. Thus, over 70% of the monoST activity and less than 10% of the polyST activity was lost after one freeze-thaw cycle. This finding was used to differentially enhance the polyST activity relative to the monoST activities.

Table I. Solubilization of membrane-associated polysialyltransferase activity from 14-day-old embryonic chick brain with CHAPS
Treatment Specific activity
Solubilized polyST activity
(supernatant; pmol/mg/4 h)		 PolyST activity remaining
in membrane (pmol/mg/4 h)
None (Control, no CHAPS) 148.1 94.1
CHAPS (62:1 molar ratio)a;
4°C preincubation; 1 hb		 459.6 52
CHAPS (62:1 molar ratio);
33°C preincubation; 1 h)		 277.8 107.9
CHAPS (124:1 molar ratio);
4°C preincubation; 1 h		 270.3 132.1
CHAPS (124:1 molar ratio);
33°C preincubation; 1 h		 226.9 30.8
The zwitterionic detergent, CHAPS, was added to the Golgi-enriched membrane fraction at either a detergent to protein molar ratio of 62:1 or 124:1 and incubated for 1 h at either 4°C or 33°C. The incubation mixtures were then centrifuged at 105,000 × g for 1 h at 4°C. The pellet fraction was resuspended in MES buffer, pH 6.1, and the polyST activity in both the supernatant and pellet fractions was determined as described in Materials and methods. As a control, the supernatant and pellet fractions from membranes not treated with CHAPS were also tested.
aMolar ratios refer to the moles of CHAPS to the moles of protein (assuming an average Mr of 40,000).
bThe T° represents the T° at which the membranes were incubated with the CHAPS for 1 h prior to centrifugation at 105,000 × g to isolate the solubilized polyST.

Table II. Summary of the properties of the membrane-bound [alpha]2,8-polyST from 14-day old embryonic chick brain and comparison with properties of the monoST activities
Property PolySTs MonoSTs
Subfamilies [alpha]2,8- PolySTs (PST and STX) [alpha]2,3-ST [alpha]2,6-ST and [alpha]2,8-STs (GD3 and GT3)
Proposed function Polysialylation of N-CAM or sodium channel [alpha] subunit; Important in development and oncogenesis Monosialylation of N- and O- linked glycoproteins and gangliosides; Important in development and oncogenesis
Kinetic properties
   KM for CMP-Neu5Ac 133 µM 50 µM ([alpha]2,6- ST)
   Vmax for CMP-Neu5Ac 0.13 µM/h 0.59 µM/h
Optimum pH and buffer pH 6.1 (MES) and pH 6.8 (PIPES) pH 6.1 (MES) only
Metal ion requirements
   Mn2+ 10 mM increases activity ~5-fold 10 mM increases activity ~7-fold
   Mg2+ 10 mM increases activity ~2-fold 10 mM increases activity ~3-fold
   Ca2+ Little effect 10 mM increases activity ~3-fold
Effect of one freeze-thaw cycle of embryonic chick brain <10% reduction in activity >70% reduction in activity
Effect of DTT and NEM
   1 mM DTT Increases activity 2-fold No effect
   80 mM DTT No effect Inhibits activity over 30%
   1 mM NEM Increases activity 30% No effect
   5 mM NEM Completely inhibits activity 60% reduction in activity
Deduced pI of enzymes 9.8-10.2 8.3-9.8
Heparin-like binding motifs 3 None in [alpha]2,3- or [alpha]2,8-monoSTs; potentially 3 in [alpha]2,6-monoST
Extended basic amino acid region Present (31aa; pI 11.6-12.0) Absent
Mechanism of synthesis Processive Distributive

Discussion

The polysialic acid glycotope plays a central role in embryonic development in species as evolutionarily distinct as sea urchin and man, and may also be a critical factor in facilitating neuroinvasive tumor metastasis (reviewed in Troy, 1995). Evidence for this latter function is largely correlative, based on finding a positive correlation between polySia expression and tumorigenesis. Thus, a better understanding of the molecular mechanism regulating synthesis and surface expression of polySia may provide greater insight into the biological function of this structurally novel glycotope. Such information could possibly aid in the development of therapeutics that could inhibit polysialylation, and thus potentially alter the malignant potential of some human tumors and meningitidis caused by neuroinvasive bacteria. The present studies were therefore initiated to characterize the membrane-associated form of the polyST activity from embryonic chick brain with respect to developmental expression, kinetic parameters, and inhibitors. We chose the embryonic chick brain as the tissue of choice because of the extensive information available on regulation of neural development in the chick, and because the tissue is abundant, relatively inexpensive and easy to isolate. Our studies have established that expression of the polyST activity is developmentally regulated, with the highest level of activity occurring at about the twelfth day of development. The optimum conditions for polyST activity were determined to include 10 mM MnCl2, 1 mM DTT and 50 mM MES buffer at pH 6.1 or 50 mM PIPES at pH 6.8. The latter buffer and pH maximized the ratio of polyST to monoST activities.

Heparin was shown to be a noncompetitive inhibitor of polysialylation, with a Ki of 9 µM. We were able to exploit this finding to purify the polyST activity ~30-fold by heparin-Sepharose affinity chromatography, after selective detergent solubilization with CHAPS. Our analysis of the deduced amino acid sequences of the mammalian polySTs revealed two potential heparin-like binding sites upstream of sialylmotifs L and S. The proposed function of the L motif is to bind CMP-Sia, while the function of the S motif is unknown (Datta and Paulson, 1995). On the basis of our new findings, we propose that the function of these two positively charged binding sites and the extended basic amino acid cluster of 31 amino acids, also upstream of sialyl motif L in the polySTs, may be to participate cooperatively in CMP-Neu5Ac binding and in the binding of the nascent polySia chain. For the processive mechanism of polymerization to occur it seems plausible that a polymer binding site, in addition to CMP-Neu5Ac binding sites, may be required for the multiple addition of new Sia residues to the growing nascent chain. This suggests that the basic amino acid clusters contiguous with sialyl motif L (pI 12.5) and upstream of sialylmotif S (pI 11.7) may be spatially in close proximity with each other in the 3-D structure of the enzyme, thus allowing a site or pocket for binding and catalysis to occur without releasing the polymeric substrates between successive catalytic steps.

The finding that dolichylphosphate and the ganglioside, GD3, also inhibited the chick brain polyST activity illustrates a major difference between the prokaryotic and eukaryotic polySTs, since the addition of undecaprenylphosphate, the prokaryotic equivalent of dolichylphosphate, stimulated polyST activity in E.coli K1 (Vijay and Troy, 1975). In polySia biosynthesis in E.coli K1, undecaprenylphosphate functions as an intermediate carrier of sialyl residues (Vijay and Troy, 1975). It is interesting to note that while there is no significant primary sequence homology between the mammalian and E.coli K1 polySTs, both catalyze synthesis of structurally identical [alpha]2,8-linked polySia chains, and both are inhibited by heparin with a Ki of 9 µM.

Inhibition of the chick brain polyST by GD3 demonstrates another functional difference between the prokaryotic and mammalian polySTs, as previous studies revealed that this ganglioside was an effective exogenous acceptor substrate for the E.coli K1 polyST (Cho and Troy, 1994). Because a developmentally regulated [alpha]2,8-polyST in embryos of the sea urchin Lytechinus pictus also catalyzed the polysialylation of GD3, this indicates that the sea urchin enzyme appears to be functionally more similar to the E.coli K1 polyST than to the embryonic chick brain polyST.

The second aspect of this work focused on thiol reduction experiments with dithiothreitol and chemical modification studies using the thiol-directed alkylating reagents, N-ethylmaleimide and iodoacetamide to assess the importance of cysteine residues in catalysis. These studies revealed that at least one thiol group on the polyST was critical for polysialylation, but of lesser importance for monosialylation. Of the six conserved cysteine residues in the mammalian polyST, PST, one is in the transmembrane spanning domain, three are in sialylmotif L, one in sialylmotif S, and one is located near the C-terminal domain of the polypeptide chain. Our findings show that the sulfhydryl group of at least one of these cysteine residues is a 'reactive thiol" that may play a role in polySia chain initiation and/or polymerization. Our 'reactive thiol" model postulates the formation of a reactive covalent adduct between CMP-Neu5Ac and a conserved cysteinyl residue in the active site of the polyST, possibly within the sialylmotif L. Thus, synthesis of polySia may be initiated when the active site cysteine residue, acting as an 'acceptor" nucleophile, binds to the C-2 carbon of the donor substrate, CMP-Neu5Ac, leading to the release of CMP. Chain polymerization could take place when the oxygen atom on C-8 of the enzyme-bound sialyl residue serves as a nucleophile to attack the phosphoryl group of a second CMP-Neu5Ac residue bound to the enzyme. Polymerization would continue with the processive addition of sialyl residues to the nonreducing termini of the growing nascent polySia chain by multiple repetition of this reaction sequence.

Our results from the thiol modification and reduction experiments are supportive of this model, as both showed that one or more of the cysteinyl residues of the polyST must be in the reduced form for maximal activity. For example, if all thiol groups were in the oxidized disulfide state, then the addition of NEM or IAA, in the absence of DTT, should have little or no effect on polysialylation. However, as shown in this study (Figure 4A,B) NEM greatly decreased polyST activity, indicating that one or more of the sulfhydryl groups essential for catalytic activity may have been modified. In addition, the 2-fold stimulation in polyST activity by 1 mM DTT (Figure 3) further confirmed the importance of a thiol group for optimum polyST activity. Addition of 1 mM DTT would not only prevent oxidation of a free thiol group, but would reduce disulfide bonds that may induce a more favorable conformation of the polyST required for the processive activity of the enzyme. Site-directed mutagenesis of select cysteine residues will now be necessary to substantiate this model.

Materials and methods

Reagents

The following reagents and materials were purchased from the sources indicated: CMP-[14C]Neu5Ac, 267.5- 324.1 mCi/mmol (DuPont NEN Research Products, Boston, MA); CMP-Neu5Ac, aprotinin, and leupeptin (Boehringer Mannheim, Indianapolis, IN); ammonium acetate, calcium chloride, 3-[cyclohexylamino]-1-propanesulfonic acid (CAPS), 2-[N-cyclohexylamino]ethanesulfonic acid (CHES), Coomassie brilliant blue G, dithiothreitol, dolichol phosphate, GD3, glycerol, heparin (3 kDa and 6 kDa), N-[2-hydroxyethyl]piperazine-N[prime]-[2-ethanesulfonic acid] (HEPES), iodoacetamide, magnesium chloride, manganese chloride, 2-[N-morpholino]ethanesulfonic acid (MES), N-ethylmaleimide, phenylmethylsulfonyl fluoride (PMSF),piperazine-N,N[prime]-bis[2-ethanesulfonic acid) (PIPES), saponin, sodium azide, and N-tris[hydroxymethyl]methylglycine (TRICINE), Triton CF-54 (Sigma); BCA protein assay kit, 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS detergent), and Slide-A-Lyzer 10K Dialysis Cassettes (Pierce); HiTrap heparin-Sepharose affinity columns (Pharmacia Biotech Inc., Piscataway, NJ); ammonium sulfate, methanol, sodium chloride, and Whatman 3MM chromatography paper (Fisher Scientific); Duracryl (high tensile strength, 30% acrylamide,.8% N, N-methlenebisacrylamide) (Millipore, Bedford, MA); and Ready Safe Liquid Scintillation Cocktail (Beckman).

Preparation of the embryonic chick brain [alpha]2,8-polysialyltransferase-enriched membrane fraction

Intact brains were isolated from 14 day embryonic chicks (except for the developmental studies), frozen on dry ice, and stored at -80°C. Brains were homogenized in 3 volumes of 50 mM MES buffer, pH 6.1, containing 10% glycerol, 500 kiu/ml aprotinin, 40 µg/ml leupeptin, 1 µg/ml pepstatin A, and 1 mM PMSF in a Köntes glass homogenizer(20-25 strokes). The homogenates were centrifuged for 10 min at 1000 × g, and the supernatant was removed and centrifuged for 30 min at 150,000 × g. This Golgi-enriched pellet (P2) was resuspended in one-quarter or one-half volume of the 50 mM MES buffer, pH 6.1, described above. After resuspension, the protein concentration was determined spectrophotometrically using the BCA Protein Assay. All procedures were carried out at 4°C.

Polysialyltransferase assay

The [alpha]2,8-polyST activity was determined in complete incubation mixtures containing the MES buffer, pH 6.1, described above. Each incubation mixture contained the following components in 350 µl: (1) polyST-enriched membrane fraction (P2; ca. 1 mg protein); (2) 10 mM MnCl2; and (3) 200 µM CMP-[14C]Neu5Ac (1.11 × 106 d.p.m.) and included 1 mM dithiothreitol (DTT). When testing a specific reagent for possible stimulatory or inhibitory effects, the reagent was added to the incubation mixture just prior to the addition of CMP-[14C]Neu5Ac, unless otherwise indicated. Reaction mixtures were incubated at 22°C or 33°C for 2 h or longer, as indicated in the figure caption. At various times, 30 µl aliquots were removed and spotted 2 1/2 inches down from the top edge of a 46 × 57 cm sheet of Whatman 3MM chromatography paper (origin). At the designated times, a 100 µl aliquot was removed from each incubation mixture and centrifuged for 30 min at 150,000 × g, 4°C, to remove any excess substrate. The pellets were resuspended in 100 µl of 20 mM Tris-HCl, pH 7.4, and divided into 2 equal aliquots. Endo-N-acylneuraminidase (Endo-N), a bacteriophage-derived endo-sialidase that specifically cleaves [alpha]2,8- linked polySia chains, was added to one of the two samples, and both samples were incubated at 37°C for 1 h (Ye et al., 1994). Thirty microliters of each sample were subsequently spotted on Whatman 3 MM paper, as described above. Descending paper chromatography was carried out overnight with ethanol: 1 M ammonium acetate, pH 7.5 (7:3) as the developing solvent. The papers were then allowed to dry. A 1 1/2 × 1 inch2 area encompassing each origin was cut from the paper, placed in vials containing scintillation fluid, and the radioactivity quantitated by counting in either a Wallac 1410 scintillation counter (Pharmacia Biotech, Piscataway, NJ) or a Beckman LS6000IC counter.

Determination of optimum pH and buffers

The polyST-enriched membrane fraction was obtained as previously described, but for these studies, the membrane pellets were resuspended in one of the following buffers: (1) 50 mM MES, pH 5.5, 6.1, and 6.8; (2) 50 mM PIPES, pH 6.8 and 7.5; (3) 50 mM HEPES, pH 7.5 and 8.1; (4) 50 mM TRICINE, pH 8.1; (5) 50 mM CHES, pH 9.3; and (6) 50 mM CAPS, pH 10.4. Each buffer contained 10% glycerol, 500 kiu/ml aprotinin, 40 µg/ml leupeptin, 1 µg/ml pepstatin A, and 1 mM PMSF. PolyST assays were then carried out using the designated buffers.

Determination of the relative amount of monoST and polyST activities

The total amount of [14C]Neu5Ac incorporated from CMP-[14C]Neu5Ac into different sialyl polymers represents the amount incorporated by both the mono- and polySTs. The difference in radioactivity between the Endo-N treated samples and the untreated (control) samples is the amount of [14C]Neu5Ac incorporated into polySia by the polyST activity. The amount of radioactivity remaining at the origin after Endo-N treatment represents the amount of [14C]Neu5Ac incorporated by the monoSTs, as described previously (Ye et al., 1994). The polyST value slightly underestimates the extent of polysialylation because Endo-N does not cleave all [alpha]2,8-sialyl linkages, leaving one or two residues attached to the penultimate Sia residues terminating each chain (Weisgerber and Troy, 1990).

Effect of DTT and chemical modification studies on polysialyltransferase activity

To determine the optimum concentration of dithiothreitol (DTT) that resulted in maximal polyST activity, 0-10 mM DTT was added to the polyST incubation mixture described previously. Stock solutions of DTT (100 mM) were prepared in double distilled water and stored at -20°C. Various concentrations of the cysteinyl alkylating reagent, N-ethylmaleimide (NEM), were tested for their effect on polyST activity. Stock solutions of NEM (10 mM and 100 mM) were prepared fresh in double distilled water. NEM was added either directly to the polyST incubation mixture just prior to addition of CMP-[14C]Neu5Ac, or was preincubated with the enzyme for 0.5, 1, 2, or 4 h at 25°C before addition of substrate. Iodoacetamide (IAA) was also tested at various concentrations for its effect on the polyST activity. A stock solution of 100 mM IAA was prepared in double distilled water, and this cysteinyl alkylating reagent was also preincubated with the polyST-enriched membrane fraction for 2 h at 33°C before addition of CMP-[14C]Neu5Ac.

Detergent solubilization of the polysialyltransferase activity

A number of detergents includingCHAPS, octyl [beta]-glucoside, saponin, and Triton CF-54 were tested for their efficacy in solubilizing the polyST activity from the membrane fraction (P2), described above. The zwitterionic detergent, CHAPS, was found to be the most effective for differential solubilization of the polyST activity. For these experiments, the detergent was incubated with the membrane fraction under the following conditions: (1) 62:1 detergent to protein molar ratio at 4°C for 0, 1, or 2 h; (2) 124:1 detergent to protein molar ratio at 4°C for 1 h; (3) 62:1 detergent to protein molar ratio at 33°C for 1 h; or (4) 124:1 detergent to protein molar ratio at 33°C for 1 h. The incubation mixtures were then centrifuged at 105,000 × g for 1 h in a Beckman Optima TLX ultracentrifuge, and the supernatant fractions removed and saved. The protein concentration and polyST activity in both the supernatant and pellet fractions were determined as described above with the following modifications: (1) in place of the centrifugation step to remove excess substrate prior to the addition of Endo-N, the polyST incubations were 'quenched" by adding a 7.5-fold molar excess of unlabeled CMP-Neu5Ac to the 100 µl sample (4.8 µl of a 4.07 mM stock solution); and (2) the Endo-N incubations were carried out for 2 h in MES buffer, pH 6.1 rather than Tris-HCl, pH 7.4.

Purification of the polysialyltransferase activity by heparin-Sepharose affinity chromatography

The polyST activity was partially purified from frozen 14 day embryonic chick brain, after solubilization with CHAPS, by heparin-Sepharose affinity chromatography as follows. The membrane fraction (P2) was prepared from ~3.2 g of tissue, except that 25 mM PIPES, pH 6.8, was used in place of 50 mM MES, pH 6.1. This fraction was resuspended in 25 mM PIPES, pH 6.8, to one-half the original volume, and protein concentration was determined spectrophotometrically using the BCA protein assay. An aliquot of freshly prepared CHAPS (190-200 mg/ml stock) was added to the membrane fraction to achieve a 62:1 detergent to protein molar ratio. After mixing by gently inverting the tube, the solution was incubated on ice for 1 h to solubilize the polyST activity. The detergent-membrane solution was then centrifuged at 105,000 × g for 1 h. The supernatant was removed and applied to a 1 ml HiTrap Heparin-Sepharose affinity column (Pharmacia Biotech Inc.). The column was washed at room temperature using 25 mM PIPES, pH 6.8, 10% glycerol, 0.1% CHAPS, 0.2% sodium azide, and protease inhibitors (described above) as the starting buffer. The column was eluted with a step-gradient using the starting buffer containing increasing concentrations of NaCl, ranging from 0.25, 0.50, 0.75, 1.0, and 2.0 M. One milliliter fractions were collected at a flow rate of ~1 ml/min.

Absorbance (280 nm) of each fraction was determined using a UVIKON 941 spectrophotometer (Kontron Elektronik Corp., Newport Beach, CA). Because higher concentrations of NaCl can inhibit polyST activity, selected fractions from the column were pooled and dialyzed against 25 mM PIPES, pH 6.8, using Slide-A-Lyzer Dialysis Cassettes (Pierce). A 200 µl aliquot of the dialyzed fractions were then assayed for polyST activity after the addition of: (1) 40 µl of assay buffer, which contained of 25 mM PIPES, pH 6.8, 10% glycerol, 87.5 mM MnCl2, and 8.75 mM DTT; (2) 100 µl of chick brain homogenate, which had been previously heated at 50°C for 15 min to inactivate endogenous polyST activity, and which served as the exogenous N-CAM acceptor; and iii) 200 µM CMP-[14C]Neu5Ac (1.11 × 106 d.p.m.). The reaction was then incubated at 33 °C for 2 h and spotted on Whatman 3MM chromatography paper. After 2 h, a 100 µl aliquot was removed and the reaction was 'quenched" by the addition of unlabeled CMP-Neu5Ac, as described above. The sample was then divided into 2 equal volumes, and one was treated with Endo-N. Both were incubated at 37°C for at least 3 h, and the rest of the assay was carried out as described above (Ye et al., 1994).

Fractions were also subjected to SDS-PAGE. Aliquots of each fraction (60 µg protein) were loaded onto 20 cm, 7% polyacrylamide gels, 1 mm thick. Electrophoresis was carried out using the PROTEAN II Slab Cell (Bio-Rad, Hercules, CA) using procedures outlined by Laemmli (Laemmli, 1970). Gels were then rinsed in double distilled water and proteins stained with Coomassie brilliant blue G (CBB G-250), which consisted of 20% methanol in 80% stock staining solution (0.1% w/v CBB G-250, 2% w/v phosphoric acid, and 10% w/v ammonium sulfate) (Neuhoff et al., 1988). Gels were subsequently destained in 35% methanol, and then dried using BioDesign Gel Wrap and Frame (BioDesign Inc. of New York, Carmel, NY).

Acknowledgments

These studies were supported in part by a Hibbard E. Williams Research Grant from the U.C. Davis School of Medicine (F.A.T.), Research Grant AI-09352 from the NIH (F.A.T.), and by a Grant-in Aid for the International Research Program: Joint Research (0404405 to Dr. Yasuo Inoue, University of Tokyo) from the Ministry of Education, Science, and Culture of Japan for the support of Dr. Kitazume-Kawaguchi. The excellent technical assistance of Steven Hatfield in carrying out some of the kinetic studies is also gratefully acknowledged.

Abbreviations

Sia,sialic acid (Neu5Ac); polySia, polysialic acid(s), [alpha]2,8-linked homopolymers of Neu5Ac residues; polyST, [alpha]2,8-polysialyltransferases (CMP-Neu5Ac:poly-[alpha]2,8-sialosyl sialyltransferase; PST and STX, designation for two cloned genes encoding [alpha]2,8-polysialyltransferase activities; sialylmotif L (long) and sialylmotif S (short); Endo-N, poly-[alpha]2,8-endo-N-acylneuraminidase (endo-sialidase); N-CAM, neural cell adhesion molecules; DTT, dithiothreitol; NEM, N-ethylmaleimide; IAA, iodoacetamide.

References

Brisson ,J.R., Baumann,H., Imberty,A., Perez,S. and Jennings,H.J. (1992) Helical epitope of the group B meningococcal alpha(2-8)-linked sialic acid polysaccharide. Biochemistry, 31, 4996-5004. MEDLINE Abstract

Cardin ,A.D. and Weintraub,H.J. (1989) Molecular modeling of protein-glycosaminoglycan interactions. Arteriosclerosis, 9, 21-32. MEDLINE Abstract

Cho ,J.-W. and Troy,F.A. (1994) Polysialic acid engineering: synthesis of polysialylated neoglycosphingolipid by using the polysialyltransferase from neuroinvasive Escherichia coli K1. Proc. Natl. Acad. Sci. USA, 91, 11427-11431. MEDLINE Abstract

Cho ,J.-W., Troy,F.A., Inoue,S., Inoue,Y. and Lennarz,W.J. (1996) A developmentally regulated a2,8-polysialyltransferase in embryos of the sea urchin Lytechinus pictus. Dev. Growth Differ., 38, 477-487.

Datta ,A.K. and Paulson,J.C. (1995) The sialyltransferase 'sialylmotif" participates in binding the donor substrate CMP-NeuAc. J. Biol. Chem., 270, 1497-1500. MEDLINE Abstract

Dubois ,C., Figarella-Branger,D., Pastoret,C., Rampini,C., Karpati,G. and Rougon,G. (1994) Expression of NCAM and its polysialylated isoforms during mdx mouse muscle regeneration and in vitro myogenesis. Neuromuscul. Disord., 4, 171-182. MEDLINE Abstract

Eckhardt ,M., Muhlenhoff,M., Bethe,A., Koopman,J., Frosch,M. and Gerardy-Schahn,R. (1995) Molecular characterization of eukaryotic polysialyltransferase-1. Nature, 373, 715-718. MEDLINE Abstract

Edelman ,G. M. (1985) Cell adhesion and the molecular processes of morphogenesis. Annu. Rev. Biochem.,54, 135-169. MEDLINE Abstract

Eggens ,I., Chojnacki,T., Kenne,L. and Dallner,G. (1983) Separation, quantitation and distribution of dolichol and dolichyl phosphate in rat and human tissues. Biochim. Biophys. Acta, 751, 355- 368. MEDLINE Abstract

Finne ,J. (1982) Occurrence of unique polysialosyl carbohydrate units in glycoproteins of developing brain. J. Biol. Chem.,257, 11966-11970. MEDLINE Abstract

Grogan ,T., Guptil,J., Mullen,J., Ye,J., Hanneman,E., Vela,L., Frutiger,Y., Miller,T. and Troy,F. (1994) Polysialylated NCAM as a neurodeterminant in malignant lymphoma (ML). Lab. Invest., 70, 110A (abstract 637).

Hanover ,J.A. and Lennarz,W.J. (1982) Transmembrane assembly of N-linked glycoproteins. Studies on the topology of saccharide lipid synthesis. J. Biol. Chem.,257, 2787-2794. MEDLINE Abstract

Hoffman ,S., Sorkin,B.C., White,P.C., Brackenbury,R., Mailhammer,R., Rutishauser,U., Cunningham,B.A. and Edelman,G.M. (1982) Chemical characterization of a neural cell adhesion molecule purified from embryonic brain membranes. J. Biol. Chem., 257, 7720-7729. MEDLINE Abstract

Inoue ,S. and Iwasaki,M. (1978) Isolation of a novel glycoprotein from the eggs of rainbow trout: occurrence of disialosyl groups on all carbohydrate chains. Biochem. Biophys. Res. Commun., 83, 1018-1023. MEDLINE Abstract

Kitazume ,S., Kitajima,K., Inoue,S., Troy,F.A., Cho,J.W., Lennarz,W.J. and Inoue,Y. (1994a) Identification of polysialic acid-containing glycoprotein in the jelly coat of sea urchin eggs. Occurrence of a novel type of polysialic acid structure. J. Biol. Chem., 269, 22712-22718. MEDLINE Abstract

Kitazume ,S., Kitajima,K., Inoue,S., Inoue,Y. and Troy,F.A.I. (1994b) Developmental expression of trout egg polysialoglycoproteins and the prerequisite [alpha]2,6-, and [alpha],8-sialyl and [alpha]2,8-polysialyltransferase activities required for their synthesis during oogenesis. J. Biol. Chem., 269, 10330-10340. MEDLINE Abstract

Kojima ,N., Yoshida,Y., Kurosawa,N., Lee,Y.C. and Tsuji,S. (1995) Enzymatic activity of a developmentally regulated member of the sialyltransferase family (STX): evidence for alpha 2,8-sialyltransferase activity toward N-linked oligosaccharides. FEBS Lett.,360, 1-4. MEDLINE Abstract

Lackie ,P.M., Zuber,C. and Roth,J. (1993) Expression patterns of polysialic acid during vertebrate organogenesis. In Roth,J., Rutishauser,U. and Troy,F.A. (eds.), Polysialic Acid: From Microbes to Man. Birkhauser Verlag, Basel, pp. 263-278.

Laemmli ,U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685. MEDLINE Abstract

Livingston ,B.D. and Paulson,J.C. (1993) Polymerase chain reaction cloning of a developmentally regulated member of the sialyltransferase gene family. J. Biol. Chem.,268, 11504-11507. MEDLINE Abstract

Livingston ,B.D., Jacobs,J.L., Glick,M.C. and Troy,F.A. (1988) Extended polysialic acid chains (n greater than 55) in glycoproteins from human neuroblastoma cells. J. Biol. Chem., 263, 9443-9448. MEDLINE Abstract

Martersteck ,C.M., Kedersha,N.L., Drapp,D.A., Tsui,T.G. and Colley,K.J. (1996) Unique [alpha]2, 8-polysialylated glycoproteins in breast cancer and leukemia cells. Glycobiology, 6, 289-301. MEDLINE Abstract

McCoy ,R.D. and Troy,F.A. (1987) CMP-Neu5Ac:poly-alpha-2,8-sialosyl sialyltransferase in neural cell membranes. Methods Enzymol., 138, 627-37. MEDLINE Abstract

McCoy ,R.D., Vimr,E.R. and Troy,F.A. (1985) CMP-Neu5Ac:poly-[alpha]-2,8-sialosyl sialyltransferase and the biosynthesis of polysialosyl units in neural cell adhesion molecules. J. Biol. Chem.,260, 12695-12699. MEDLINE Abstract

Means ,G.E. and Feeney,R.E. (1971) Alkylating and similar reagents. In Chemical Modification of Proteins. Holden-Day, San Francisco, pp. 105-114.

Michon ,F., Brisson,J.R. and Jennings,H.J. (1987) Conformational differences between linear [alpha](2,8)-linked homosialooligosaccharides and the epitope of the group B meningococcal polysaccharide. Biochemistry,26, 8399-8405. MEDLINE Abstract

Nakayama ,J., Fukuda,M.N., Fredette,B., Ranscht,B. and Fukuda,M. (1995) Expression cloning of a human polysialyltransferase that forms the polysialylated neural cell adhesion molecule present In embryonic brain. Proc. Natl. Acad. Sci. USA,92, 7031-7035. MEDLINE Abstract

Neuhoff ,V., Arold,N., Taube,D. and Ehrhardt,W. (1988) Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie brilliant blue G-250 and R-250. Electrophoresis,9, 255-262. MEDLINE Abstract

Oka ,S., Bruses,J.L., Nelson,R.W. and Rutishauser,U. (1995) Properties and developmental regulation of polysialyltransferase activity in the chicken embryo brain. J. Biol. Chem., 270, 19357-19363. MEDLINE Abstract

Robbins ,J.B., McCracken,G.H.,Jr., Gotschlich,E.C., Orskov,F., Orskov,I. and Hanson, L.A. (1974) Escherichia coli K1 capsular polysaccharide associated with neonatal meningitis. N. Engl. J. Med., 290, 1216-1220. MEDLINE Abstract

Rohr ,T.E. and Troy,F.A. (1980) Structure and biosynthesis of surface polymers containing polysialic acid in Escherichia coli. J. Biol. Chem., 255, 2332-2342. MEDLINE Abstract

Rosenberg ,J., Ellis,L., Troy,F. and Kayalar,C. (1986) The 5B4 antigen expressed on sprouting neurons contains [alpha]-2,8-linked polysialic acid. Brain Res., 395, 262-267. MEDLINE Abstract

Roth ,J., Zuber,C., Wagner,P., Taatjes,D.J., Weisgerber,C., Heitz,P.U., Goridis,C. and Bitter-Suermann,D. (1988) Reexpression of poly(sialic acid) units of the neural cell adhesion molecule in Wilms tumor. Proc. Natl. Acad. Sci. USA, 85, 2999-3003. MEDLINE Abstract

Roth ,J., Rutishauser,U. and Troy,F.A. (1993) Polysialic Acid: From Microbes to Man. Birkhauser Verlag, Basel and Boston.

Rutishauser ,U., Acheson,A., Hall,A.K., Mann,D.M. and Sunshine,J. (1988) The neural cell adhesion molecule (NCAM) as a regulator of cell-cell interactions. Science, 240, 53-57. MEDLINE Abstract

Scheidegger ,E.P., Lackie,P.M., Papay,J. and Roth,J. (1994) In vitro and in vivo growth of clonal sublines of human small cell lung carcinoma is modulated by polysialic acid of the neural cell adhesion molecule. Lab. Invest., 70, 95-106. MEDLINE Abstract

Scheidegger ,E.P., Sternberg,L.R., Roth,J. and Lowe,J.B. (1995) A human STX cDNA confers polysialic acid expression in mammalian cells. J. Biol. Chem., 270, 22685-22688. MEDLINE Abstract

Seki ,T. and Arai,Y. (1993) Distribution and possible roles of the highly polysialylated neural cell adhesion molecule (NCAM-H) in the developing and adult central system. Neurosci. Res., 17, 265-290. MEDLINE Abstract

Sevigny ,M. and Troy,F.A. (1997) Enzymatic properties, solubilization, and chemical modification of the [alpha]2,8-polysialyltransferase from embryonic chick brain. FASEB J.,11, A1248.

Tang ,J., Rutishauser,U. and Landmesser,L. (1994) Polysialic acid regulates growth cone behavior during sorting of motor axons in the plexus region. Neuron, 13, 405-414. MEDLINE Abstract

Troy ,F.A. (1979) The chemistry and biosynthesis of selected bacterial capsular polymers. Annu. Rev. Microbiol.,33, 519-560. MEDLINE Abstract

Troy ,F.A. (1992) Polysialylation: from bacteria to brains. Glycobiology, 2, 5-23. MEDLINE Abstract

Troy ,F.A. (1995) Sialobiology and the polysialic acid glycotope: Occurrence, structure, function, synthesis and glycopathology, In Rosenberg,A. (eds.), Sialic Acid: Molecular, Regulatory, and Abnormal Biology. Plenum, New York, pp. 95-144.

Tsuji ,S., Kojima,Y., Yoshida,Y. and Kurdsawa,N. (1997) Characterization of two polysialic acid synthases. XIV Intl. Symp. on Glycoconjugates Abst. S25. Zurich, Switzerland.

Vijay ,I.K. and Troy,F.A. (1975) Properties of membrane-associated sialyltransferase of Escherichia coli. J. Biol. Chem., 250, 164-170. MEDLINE Abstract

Vimr ,E.R., McCoy,R.D., Vollger,H.F., Wilkison,N.C. and Troy,F.A. (1984) Use of prokaryotic-derived probes to identify poly(sialic acid) in neonatal neuronal membranes. Proc. Natl. Acad. Sci. USA,81, 1971-1975. MEDLINE Abstract

Wang ,C., Rougon,G. and Kiss,J.Z. (1994) Requirement of polysialic acid for the migration of the O-2A glial progenitor cell from neurohypophyseal explants. J. Neurosci., 14, 4446-4457. MEDLINE Abstract

Weinstein ,J., Lee,E.U., McEntee,K., Lai,P.H. and Paulson,J.C. (1987) Primary structure of beta-galactoside alpha 2,6-sialyltransferase. Conversion of membrane-bound enzyme to soluble forms by cleavage of the NH2-terminal signal anchor. J. Biol. Chem., 262, 17735-17743. MEDLINE Abstract

Weisgerber ,C. and Troy,F.A. (1990) Biosynthesis of the polysialic acid capsule in Escherichia coli K1. The endogenous acceptor of polysialic acid is a membrane protein of 20 kDa. J. Biol. Chem., 265, 1578-1587.

Yamasaki ,R. and Bacon,B. (1991) Three-dimensional structural analysis of the group B polysaccharide of Neisseria meningitidis 6275 by two-dimensional NMR: the polysaccharide is suggested to exist in helical conformations in solution. Biochemistry, 30, 851-857. MEDLINE Abstract

Ye ,J., Kitajima,K., Inoue,Y., Inoue,S. and Troy,F.A. (1994) Identification of polysialic acids in glycoconjugates. Methods Enzymol., 230, 460-484. MEDLINE Abstract

Yoshida ,Y., Kojima,N. and Tsuji,S. (1995) Molecular cloning and characterization of a third type of N-glycan alpha 2,8-sialyltransferase from mouse lung. J. Biochem. (Tokyo),118, 658-664. MEDLINE Abstract

Zuber ,C., Lackie,P.M., Catterall,W.A. and Roth,J. (1992) Polysialic acid is associated with sodium channels and the neural cell adhesion molecule N-CAM in adult rat brain. J. Biol. Chem., 267, 9965-9971. MEDLINE Abstract

1To whom correspondence should be addressed
2Present address: Molecular Glycobiology, Frontier Research Program, RIKEN, Hirosawa 2-1, Wako-shi, Saitama, 351-01 Japan

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 22 Jul 1998
Copyright©Oxford University Press, 1998.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
D. A. Foley, K. G. Swartzentruber, and K. J. Colley
Identification of Sequences in the Polysialyltransferases ST8Sia II and ST8Sia IV That Are Required for the Protein-specific Polysialylation of the Neural Cell Adhesion Molecule, NCAM
J. Biol. Chem., June 5, 2009; 284(23): 15505 - 15516.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
Y. Guerardel, L.-Y. Chang, E. Maes, C.-J. Huang, and K.-H. Khoo
Glycomic survey mapping of zebrafish identifies unique sialylation pattern
Glycobiology, March 1, 2006; 16(3): 244 - 257.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
D. Nakata and F. A. Troy II
Degree of Polymerization (DP) of Polysialic Acid (PolySia) on Neural Cell Adhesion Molecules (N-CAMs): DEVELOPMENT AND APPLICATION OF A NEW STRATEGY TO ACCURATELY DETERMINE THE DP OF polySIA CHAINS ON N-CAMS
J. Biol. Chem., November 18, 2005; 280(46): 38305 - 38316.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. M. Steenbergen and E. R. Vimr
Functional Relationships of the Sialyltransferases Involved in Expression of the Polysialic Acid Capsules of Escherichia coli K1 and K92 and Neisseria meningitidis Groups B or C
J. Biol. Chem., April 18, 2003; 278(17): 15349 - 15359.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
S. Inoue, S.-L. Lin, Y. C. Lee, and Y. Inoue
An ultrasensitive chemical method for polysialic acid analysis
Glycobiology, September 1, 2001; 11(9): 759 - 767.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
D. S. Rickman, M. P. Bobek, D. E. Misek, R. Kuick, M. Blaivas, D. M. Kurnit, J. Taylor, and S. M. Hanash
Distinctive Molecular Profiles of High-Grade and Low-Grade Gliomas Based on Oligonucleotide Microarray Analysis
Cancer Res., September 1, 2001; 61(18): 6885 - 6891.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
C. S. K. Mayanil, D. George, B. Mania-Farnell, C. L. Bremer, D. G. McLone, and E. G. Bremer
Overexpression of Murine Pax3 Increases NCAM Polysialylation in a Human Medulloblastoma Cell Line
J. Biol. Chem., July 21, 2000; 275(30): 23259 - 23266.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M. Windfuhr, A. Manegold, M. Muhlenhoff, M. Eckhardt, and R. Gerardy-Schahn
Molecular Defects That Cause Loss of Polysialic Acid in the Complementation Group 2A10
J. Biol. Chem., October 13, 2000; 275(42): 32861 - 32870.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. Inoue, S.-L. Lin, and Y. Inoue
Chemical Analysis of the Developmental Pattern of Polysialylation in Chicken Brain. EXPRESSION OF ONLY AN EXTENDED FORM OF POLYSIALYL CHAINS DURING EMBRYOGENESIS AND THE PRESENCE OF DISIALYL RESIDUES IN BOTH EMBRYONIC AND ADULT CHICKEN BRAINS
J. Biol. Chem., September 22, 2000; 275(39): 29968 - 29979.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. Kitazume-Kawaguchi, S. Kabata, and M. Arita
Differential Biosynthesis of Polysialic or Disialic Acid Structure by ST8Sia II and ST8Sia IV
J. Biol. Chem., May 4, 2001; 276(19): 15696 - 15703.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
K. Angata, T.-Y. Yen, A. El-Battari, B. A. Macher, and M. Fukuda
Unique Disulfide Bond Structures Found in ST8Sia IV Polysialyltransferase Are Required for Its Activity
J. Biol. Chem., April 27, 2001; 276(18): 15369 - 15377.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (14)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Sevigny, M. B.
Right arrow Articles by Troy, F. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sevigny, M. B.
Right arrow Articles by Troy, F. A., 2nd
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?