Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 293 (2), 701-716

Autopolysialylation of Polysialyltransferases Is Required for Polysialylation and Polysialic Acid Chain Elongation on Select Glycoprotein Substrates

Affiliations

Autopolysialylation of Polysialyltransferases Is Required for Polysialylation and Polysialic Acid Chain Elongation on Select Glycoprotein Substrates

Gaurang P Bhide et al. J Biol Chem.

Abstract

Polysialic acid (polySia) is a large glycan polymer that is added to some glycoproteins by two polysialyltransferases (polySTs), ST8Sia-II and ST8Sia-IV. As polySia modulates cell adhesion and signaling, immune cell function, and tumor metastasis, it is of interest to determine how the polySTs recognize their select substrates. We have recently identified residues within the ST8Sia-IV polybasic region (PBR) that are required for neural cell adhesion molecule (NCAM) recognition and subsequent polysialylation. Here, we compared the PBR sequence requirements for NCAM, neuropilin-2 (NRP-2), and synaptic cell adhesion molecule 1 (SynCAM 1) for polysialylation by their respective polySTs. We found that the polySTs use unique but overlapping sets of PBR residues for substrate recognition, that the NCAM-recognizing PBR sites in ST8Sia-II and ST8Sia-IV include homologous residues, but that the ST8Sia-II site is larger, and that fewer PBR residues are involved in NRP-2 and SynCAM 1 recognition than in NCAM recognition. Noting that the two sites for ST8Sia-IV autopolysialylation flank the PBR, we evaluated the role of PBR residues in autopolysialylation and found that the requirements for polyST autopolysialylation and substrate polysialylation overlap. These data together with the evaluation of the polyST autopolysialylation mechanism enabled us to further identify PBR residues potentially playing dual roles in substrate recognition and in polySia chain polymerization. Finally, we found that ST8Sia-IV autopolysialylation is required for NRP-2 polysialylation and that ST8Sia-II autopolysialylation promotes the polymerization of longer polySia chains on SynCAM 1, suggesting a critical role for polyST autopolysialylation in substrate selection and polySia chain elongation.

Keywords: glycoprotein biosynthesis; glycosylation; glycosyltransferase; neural cell adhesion molecule; neuropilin-2; polysialic acid; polysialyltransferase; sialic acid; sialyltransferase; synaptic cell adhesion molecule.

Conflict of interest statement

The authors declare that they have no conflicts of interest with the contents of this article

Figures

Figure 1.
Figure 1.
Schematic of ST8Sia-IV and ST8Sia-II PBR mutants used in this study. Shown are the conserved sequences in the polySTs, including the transmembrane region (TM), polybasic region (PBR), the large sialyl motif (SML), the polysialyltransferase domain (PSTD), the small sialyl motif (SMS), motif III (M3), and the very small sialyl motif (SMVS). The PBR sequences of ST8Sia-IV (amino acids 71–105) and ST8Sia-II (amino acids 86–120) are shown with each alanine replacement mutant used in this study shown. These mutations were made in the wild-type enzymes to test the role of each basic residue in polysialylation, as well as in inactive forms of each polyST (ST8Sia-IV H331K and ST8Sia-II H346K) for use in competition assays.
Figure 2.
Figure 2.
NCAM recognition and polysialylation requires Arg82 and Arg93 in the ST8Sia-IV PBR. A, left panels, V5-tagged NCAM was co-expressed with Myc-tagged ST8Sia-IV or its mutants in COS-1 cells. After 24 h, NCAM was recovered from cell lysates by immunoprecipitation and subjected to SDS-PAGE and immunoblotting with the 12F8 anti-polySia antibody to analyze the level of NCAM polysialylation (upper panel). Relative NCAM and ST8Sia-IV expression levels were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as % of wild-type enzyme polysialylation. B, left panels, V5-tagged NCAM was co-expressed in COS-1 cells with untagged wild-type ST8Sia-IV and Myc-tagged inactive ST8Sia-IV H331K or its PBR mutants in a ratio of 1:1:6. NCAM polysialylation was determined as described above (upper panel). The relative expression levels of NCAM and ST8Sia-IV H331K or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as fold recovery from competition with ST8Sia-IV H331K. A and B (right scatterplots), results from four different repeats were averaged, and standard deviation and significance were assessed using a one-way ANOVA test with a Dunnett's post hoc test, where **, 0.001 < p < 0.01; ***, 0.0001 < p < 0.001; ns, p > 0.05.
Figure 3.
Figure 3.
NRP-2 polysialylation requires Arg82 and Lys99 in the ST8Sia-IV PBR, but only Arg82 is required for recognition. A, left panels, V5-tagged NRP-2 was co-expressed with Myc-tagged ST8Sia-IV or its mutants in COS-1 cells. After 24 h, NRP-2 was recovered from cell lysates by immunoprecipitation using an anti-V5 antibody and subjected to SDS-PAGE and immunoblotting with the 12F8 anti-polySia antibody to analyze the level of NRP-2 polysialylation (upper panel). Relative NRP-2 and ST8Sia-IV expression levels were determined described under “Experimental procedures” (middle and bottom panels). Data are reported as % of wild-type enzyme polysialylation. B, left panels, V5-tagged NRP-2 was co-expressed in COS-1 cells with untagged wild-type ST8Sia-IV and Myc-tagged inactive ST8Sia-IV H331K or its PBR mutants in a ratio of 1:1:6. NRP-2 polysialylation was determined as described above (upper panel). The relative expression levels of NRP-2 and ST8Sia-IV H331K or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as fold recovery from competition with ST8Sia-IV H331K. A and B, right scatterplots, results from four and three repeats, respectively, were averaged, and standard deviation and significance were assessed using a one-way ANOVA test with a Dunnett's post oc test, where *, 0.01 < p < 0.05; **, 0.001 < p < 0.01; ***, 0.0001 < p < 0.001; ns, p > 0.05.
Figure 4.
Figure 4.
NCAM recognition and polysialylation requires Arg97, Lys108, Lys114, and Lys118 in the ST8Sia-II PBR. A, left panels, NCAM-Fc was co-expressed with V5-tagged ST8Sia-II or its PBR mutants in COS-1 cells. After 24 h, NCAM-Fc was recovered from cell medium using protein A-Sepharose beads and its polysialylation was assessed by SDS-PAGE and immunoblotting with the 12F8 anti-polySia antibody (upper panel). The relative expression levels of NCAM-Fc, ST8Sia-II, or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as % of wild-type enzyme polysialylation. B, left panels, NCAM-Fc was co-expressed in COS-1 cells with V5-tagged wild-type ST8Sia-II and Myc-tagged inactive ST8Sia-II H346K or its PBR mutants in a ratio of 1:1:6. NCAM-Fc polysialylation was determined as described above using the anti-polySia 735 antibody (upper panel). The relative expression levels of NCAM-Fc and ST8Sia-II H346K or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as fold recovery from competition with ST8Sia-II H346K. A and B, right scatterplots, results from six and five repeats, respectively, were averaged, and standard deviation and significance were assessed using a one-way ANOVA test with a Dunnett's post hoc test, where *, 0.01 < p < 0.05; ***, 0.0001 < p < 0.001; ns, p > 0.05.
Figure 5.
Figure 5.
SynCAM 1 polysialylation primarily requires Lys102, Lys114, and Lys118 in the ST8Sia-II PBR with Lys102 and Lys114 making the most significant contributions to recognition. A, left panels, SynCAM-Fc was co-expressed with V5-tagged ST8Sia-II or its PBR mutants in COS-1 cells. After 24 h, SynCAM-Fc was recovered from cell medium using protein A-Sepharose beads, and its polysialylation was assessed by SDS-PAGE and immunoblotting with the 12F8 anti-polySia antibody (upper panel). Relative expression levels of SynCAM-Fc, ST8Sia-II, or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as % of wild-type enzyme polysialylation. B, left panels, SynCAM-Fc was co-expressed in COS-1 cells with V5-tagged wild-type ST8Sia-II and Myc-tagged inactive ST8Sia-II H346K or its PBR mutants in a ratio of 1:1:6. SynCAM-Fc polysialylation was determined as described above (upper panel). The relative expression levels of SynCAM-Fc and ST8Sia-II H346K or its mutants were determined as described under “Experimental procedures” (middle and bottom panels). Data are reported as fold recovery from competition with ST8Sia-II H346K. A and B, right scatterplots, results from three repeats were averaged, and standard deviation and significance were assessed using a one-way ANOVA test with a Dunnett's post hoc test, where *, 0.01 < p < 0.05; **, 0.001 < p < 0.01; ***, 0.0001 < p < 0.001; ns, p > 0.05.
Figure 6.
Figure 6.
Summary of the roles of polyST PBR residues in substrate recognition and polysialylation and polyST autopolysialylation. To make comparisons between competition assays that varied in their fold recovery of polysialylation, we set the highest fold recovery to 10 for each enzyme/substrate pair and adjusted the other numbers to that scale. We then grouped residues impacting polysialylation and/or competition into three groups. Residues colored green are those that when replaced led to substrate polysialylation of between 0 and 50% that seen with the wild-type enzyme and rated between 5 and 10 on the loss of competition/recovery of polysialylation scale. For these residues, loss of polysialylation matches the loss of substrate recognition. Residues colored red are those that when replaced led to substrate polysialylation of between 51 and 100% of that seen with the wild-type enzyme and rated between 5 and 10 on the loss of competition/recovery of polysialylation scale. For these residues, loss of recognition/competition was greater than observed loss in polysialylation. Residues colored blue are those that when replaced led to substrate polysialylation of between 0 and 50% that seen with the wild-type enzyme and rated between 0 and 4.99 on the loss of competition/recovery of polysialylation scale. For these residues loss of polysialylation was greater than the observed loss of recognition/competition. Those residues that impact autopolysialylation when replaced with alanines are indicated by a *.
Figure 7.
Figure 7.
Structural model of ST8Sia-IV based on the X-ray crystal structure of ST8Sia-III (Protein Data Bank 5BO6) (38), showing the relationship of the PBR sequences to sites of enzyme autopolysialylation. The PBR region is depicted in wheat and the PSTD is depicted in orange. Sialyl motifs are depicted in cyan (large sialyl motif), marine (SMS), magenta (M3), and red (SMVS). Asparagine attachment sites of the autopolysialylated N-glycans are shown as green spheres, and Leu151 is marked as cyan sphere. This model was created using SWISS-MODEL server (39–42).
Figure 8.
Figure 8.
Arg82 and Lys99 in the ST8Sia-IV PBR region and Lys114 and Lys118 in the ST8Sia-II PBR region are crucial for polyST autopolysialylation. A, Myc-tagged ST8Sia-IV PBR mutants were expressed in COS-1 cells. After 24 h, ST8Sia-IV PBR mutants were recovered from cell lysates by immunoprecipitation using an anti-Myc antibody and subjected to SDS-PAGE and immunoblotting with the 12F8 anti-polySia antibody to analyze the level of ST8Sia-IV autopolysialylation (upper panel). An aliquot of cell lysate was boiled with Laemmli sample buffer to remove polySia, subjected to SDS-PAGE, followed by immunoblotting using anti-Myc antibody (lower panel). B, autopolysialylation of Myc-tagged ST8Sia-II and its PBR mutants was analyzed as described above. A and B, lower scatterplots, results from four repeats were averaged, and S.D. and significance were assessed using a one-way ANOVA test with a Dunnett's post hoc test, *, 0.01 < p < 0.05; ***, 0.0001< p < 0.001; ns, p > 0.05.
Figure 9.
Figure 9.
polyST autopolysialylation is self- and not cross-polysialylated and is required for NRP-2 polysialylation and optimal SynCAM 1 polysialylation. A, V5-tagged ST8Sia-IV or non-autopolysialylated ST8Sia-IV mutants (mut2.3 and L151A) were expressed in COS-1 cells. B, Myc-tagged, inactive ST8Sia-IV H331K was co-expressed with V5-tagged mut2.3 or L151A catalytically active but non-autopolysialylated enzyme mutants in COS-1 cells. Myc-tagged ST8Sia-IV was expressed alone as an autopolysialylation control. C, V5-tagged ST8Sia-IV and its non-autopolysialylated mutants mut2.3 and L151A were co-expressed in COS-1 cells with Myc-tagged NCAM or NRP-2. D, Fc-tagged NCAM and SynCAM 1 were co-expressed separately with V5-tagged ST8Sia-II or its non-autopolysialylated mutant (ST8Sia-II mut2.4.5) in COS-1 cells. For all panels, protein polysialylation and relative protein expression levels were assessed by immunoblotting with either the anti-polySia 12F8 antibody (A–C) or the anti-polySia 735 antibody (D) and appropriate anti-tag antibodies as described under “Experimental procedures.” All experiments were performed at least three times, but quantitation was not performed because under the conditions tested there was no polysialylation (A–C) or the absence of a specific polysialylated form (D).

Similar articles

See all similar articles

Cited by 1 article

Publication types

MeSH terms

Associated data

LinkOut - more resources

Feedback