Promotion of Cell Migration by Neural Cell Adhesion Molecule (NCAM) Is Enhanced by PSA in a Polysialyltransferase-Specific Manner

Neural cell adhesion molecule 140 (NCAM-140) is a glycoprotein and always highly polysialylated in cancer. Functions of polysialic acid (PSA) that binds to N-glycan termini on NCAM remain unclear. ldlD-14 cells, a CHO cell mutant deficient in UDP-Gal 4-epimerase, are useful for structural and functional studies of Gal-containing glycoproteins because their abnormal glycosylation can be converted to normal status by exogenous addition of galactose (Gal). We cloned the genes for NCAM-140 and for polysialyltransferases STX and PST (responsible for PSA synthesis) from normal murine mammary gland epithelial (NMuMG) cells and transfected them into ldlD-14 and human breast cancer cells MCF-7. The effect of PSA on NCAM-mediated cell proliferation, motility, migration and adhesion was studied. We found that NCAM-140 significantly promoted cell proliferation, motility and migration, while polysialylation of NCAM-140 catalyzed by STX, but not by PST, enhanced NCAM-mediated cell migration, but not cell proliferation or motility. In addition, PSA catalyzed by different polysialyltransferases affected the adhesion of NCAM to different extracellular matrix (ECM) components.


Introduction
The neural cell adhesion molecule (NCAM), a member of the immunoglobulin superfamily, mediates both homophilic (NCAM to NCAM) and heterophilic binding (NCAM to sulfate proteoglycans or other collagens) during cellular interactions [1]. NCAM occurs in three isoforms: NCAM-180, NCAM-140, and NCAM-120. NCAM-140 and NCAM-180 contain a transmembrane and a cytoplasmic region, and are involved in early development and in guidance of migrating neurons. NCAM-120 is linked to the membrane via a glycosylphosphatidylinositol (GPI) anchor, and is up-regulated during differentiation [2,3].

Gene amplification and transfection
Immunofluorescence staining Cells (2x10 4 ) were cultured on 12-mm diameter glass cover slips in 24-well plates for 48 hr, washed with PBS, fixed with 4% fresh paraformaldehyde in PBS, blocked with 1% BSA/0.1% NaN 3 /PBS for 1hr, stained with primary antibody at 4°C overnight, and then with FITCconjugated secondary antibody for 1 hr at room temperature. Cell nuclei were stained with Hoechst 33342 (Invitrogen, Paisley, UK). Cells were mounted with Glycergel (Dako Cytomation, Carpinteria, CA) and observed by laser confocal fluorescence microscopy (model Eclipse Ti-U; Nikon, Tokyo, Japan) at 600× magnification.

Motility assay
Cell motility was determined by phagokinetic gold sol assay as described previously [22]. Cells (2x10 3 ) in complete culture medium were seeded onto gold sol-coated wells, incubated for 12-18 hr, and photographed under an inverted microscope. Tracking areas of 50 cells were measured using the ToupView imaging system [23] and expressed as square pixels.

Proliferation (MTT) assay
Cell proliferation was determined by the MTT assay as described previously [24]. Cells (4x10 3 / well) were seeded in 96-well plates and incubated for 12, 24, 48, 72 or 96 hr. Each well was added with 4 μL MTT solution (Cers, China) and incubated at 37°C for 4 hr. The reaction was quenched by addition of 100 μL DMSO, and absorbance at 595 nm was recorded immediately.

Transwell migration assay
Cells (5x10 4 ) were plated in an upper transwell insert (12/24-well transwell; 8 μm polycarbonate membrane; Costar, Corning, NY) in medium containing 0.1 ml of 0.2% BSA (Ruitaibio; Beijing, China). Medium (0.6 ml) supplemented with 5% FBS, serving as a chemo-attractant, was placed in the lower chamber. Cells were incubated for 16 hr at 37°C in 5% CO 2 atmosphere, washed with PBS, and fixed with cold 4% paraformaldehyde. Cells on the top surface of the insert filter were removed with a cotton wool tip and the filter was stained with crystal violet. Filters were rinsed with deionized water, air-dried, and photographed. Cells in five randomly selected optical fields were counted to determine migration [25,26].

Flow cytometry
Cells were plated in 24-well plates (2×10 5 cells/well) in Ham's F12 medium containing 5% FBS for 24 hr, and the medium was replaced by serum-free F12 containing ITS with or without 20 μM Gal. Cells were detached and either (i) incubated with primary antibody at 4°C for 2 hr, then with FITC-conjugated secondary antibody for 1 hr, or (ii) incubated with FITC-conjugated GSL-II at 4°C for 2 hr. Signals from cells were determined by flow cytometry (FACS Calibur, BD, San Jose, CA), with data acquisition and analysis by the FlowJo software program (Tree Star, San Carlos, CA).

Cell adhesion assay
Adhesion assays were performed as described previously [30]. In brief, 96-well plates were coated with fibronectin (1 μg/well), collagen IV (1 μg/well), matrigel (40 μg/well) or laminin (1 μg/ well) at 37°C for 2 hr. Wells were rinsed and blocked with 1% BSA in HBSS at 37°C for 1 hr. Cells were harvested with trypsin and plated at 40 000 cells per coated well. After 30 min incubation at 37°C, unattached cells were gently removed with HBSS. Adherent cells were fixed with 4% paraformaldehyde for 10 min and stained with 0.1% crystal violet (in 20% MeOH) for 10 min. Excess dye was washed off with PBS, crystal violet in the cells was dissolved in 100 μl 10% acetic acid, and absorbance was measured at 595 nm [31]. Bars represent mean absorbance +/-SEM of each condition tested in triplicates.

Data analysis
Data were statistically analyzed using the Prism 5 software program [32]. Differences between means were evaluated by Student's t-test, and p-values <0.05 were considered significant.

PSA on NCAM is easily modified in ldlD/N140 cells
To investigate the function of NCAM-140 and PSA-NCAM, the NCAM-140 gene from NMuMG cells was cloned. NCAM levels in parental ldlD-14 cells were undetectable, whereas transfected cells (ldlD/N140) showed a strong NCAM-140 signal in all tested conditions (Fig 1A and 1B).
Modification of glycosylation in ldlD-14 cells by addition of Gal was confirmed by MAL-DI-TOF/TOF-MS/MS. Since sialic acids are easily lost during mass spectrometric ionization, the glycans released from ldlD/N140 cells cultured with ITS or ITS+Gal were purified, derivatized and analyzed. Proposed N-glycan structures and their molecular weights are shown in Fig 2A and 2B. Cells cultured with ITS had no Gal or sialic acid residues on any of the annotated N-glycans (Fig 2A). In contrast, cells cultured with ITS+Gal presented two unique structures (m/z values 2465.120, 3147.336), annotated as N-glycans with terminal Gal and sialic acid (Fig 2B). This result indicated that ldlD/N140 cells can express NCAM under ITS condition, and express PSA-NCAM under ITS +Gal condition.
In ldlD/N140 cells cultured with ITS, NCAM had a low molecular weight because of the lack of PSA synthesis ( Fig 2C). In cells cultured with ITS+Gal, the salvage pathway for Gal incorporation provided an alternative source of UDP-Gal as described previously [33] and the glycan on NCAM was elongated, resulting in a higher molecular weight of NCAM ( Fig 2C).
GSL-II, a lectin with N-acetyl glucosamine (GlcNAc)-binding specificity [34], was used to distinguish N-glycan termini of ldlD/N140 cells cultured with ITS vs. ITS+Gal. ITS-cultured cells displayed terminal GlcNAc and a high GSL-II-FITC binding signal, whereas ITS+Galcultured cells had almost no fluorescent signal (Fig 2D-2Da). The NCAM pattern was the same under the two culture conditions (Fig 2D-2Db). The PSA-NCAM fluorescent signal was slightly stronger for ITS+Gal-cultured than for ITS-cultured cells (Fig 2D-2Dc).
NCAM-140 enhances cell motility, proliferation and migration ldlD/N140 cells, in comparison with ldlD-14 cells, were larger and lost the spindle-like morphology ( Fig 3A). They displayed significantly enhanced cell motility ( Fig 3B) and proliferation under both ITS and ITS+Gal culture (Fig 3C and 3D). Up-regulation of NCAM-140 resulted in increased migration of both ITS-cultured and ITS+Gal-cultured cells (Fig 3E), suggesting a role of NCAM-140 in metastasis. However, no significant difference of cell motility and migration in ITS-cultured vs. ITS+Gal-cultured cells was observed (Fig 3E, right panel). The proposed reason was that the expression of PSA was not significant enough to make the difference (Fig 1B-1l and Fig 2D-2Dc).
Proliferation of ldlD/N/S (but not ldlD/N/P) cells was slightly higher than that of ldlD/ N140 cells, which had low PSA content (Fig 4D). Cell motility was not affected in ldlD/N/S and ldlD/N/P cells, indicating that NCAM may be the main factor responsible for cell motility (Fig 4E). Interestingly, cell migration was significantly enhanced in ldlD/N/S cells, but not in ldlD/N/P cells (Fig 4F). In order to confirm the above result, a noninvasive breast cancer cell line, MCF-7, was used [36,37]. Stable overexpression of NCAM (MCF7/N140), and transient overexpression of STX or PST (MCF7/N/S and MCF7/N/P) in MCF-7 cells were constructed (Fig 5A and 5B). MCF7/ N/S and MCF7/N/P cells showed a strong signal of PSA-NCAM, compared with control cells (Fig 5C). Overexpression of NCAM resulted in slightly increased proliferation, while STX expression, but not PST expression, increased proliferation further in comparison to MCF7/ N140 cells (Fig 5D). Similarly, increased migration of MCF7/N140 cells was further enhanced by overexpression of STX, but not PST (Fig 5E). These findings suggested that PSA catalyzed by STX vs. PST has a different effect on cell proliferation and migration, presumably because of a differing degree of polymerization of PSA.

PSA catalyzed by STX vs. PST has a different effect on the role of NCAM in cell adhesion to different extracellular matrix substrates
Literature shows that altered cell adhesion is involved in tumor metastasis and progression, and that the extracellular matrix (ECM) plays an important role in the regulation of cell adhesion [38]. We further examined the effect of NCAM and PSA on the attachment of ldlD-14 cells to the following ECM components: fibronectin (FN), laminin (LN), collagen IV and  (Fig 6). ldlD-14 cells were more adherent to LN and collagen IV than FN and matrigel. In ITS condition, NCAM-140 overexpressing cells (ldlD/N140, ldlD/N/S and ldlD/N/P) were highly adherent to FN and poorly adherent to LN, collagen IV and matrigel, compared with ldlD-14 cells. These different adhesive properties became more pronounced in ITS+Gal condition. ldlD/N/S and ldlD/N/P cells presented decreased adhesion to FN and collagen IV, and increased adhesion to matrigel, compared to ldlD/N140 cells. Interestingly, adhesion to LN was not changed in ldlD/N/S cells, but greatly increased in ldlD/N/P cells. These results suggested that PSA catalyzed by different polysialyltransferases affected the adhesion of NCAM to ECM components.

Discussion
Cell adhesiveness and migration are key factors in the metastatic process, and modulation of tumor cell adhesion molecules is crucial in control of the metastatic cascade [39]. There is increasing evidence that NCAM functions as a surface marker for human small cell lung cancer [40] and other types of cancer [41]. NCAM function is affected by attached PSA molecules, which are expressed in malignant cancer cells and associated with high metastasis [39].
We used an ldlD-14 cell model to elucidate the role of PSA in NCAM function. The UDP-Gal 4-epimerase deficiency characteristic of ldlD-14 cells results in low internal pools of UDP-Gal when cells are grown in the absence of Gal [33]. The derived cell line ldlD-14 is a useful model for studying the functions of glycans in glycoproteins and glycolipids [18,42]. NCAM model in ldlD-14 cells was proposed when cultured in ITS only condition. In this situation, incomplete N-glycan on ldlD-14 cells lacked the factors of Gal for the attachment of PSA ( Fig  7B). To study the role of PSA-NCAM, addition of Gal on N-glycan termini is the essential step for PSA synthesis on NCAM (Fig 7C). Using transfected ldlD/N140 cells, we were able to restore normal cell phenotype by addition of Gal and study glycan patterns by flow cytometry and MALDI-TOF/TOF-MS/MS.
In ldlD/N140 cells cultured in serum-free medium supplemented with ITS or ITS+Gal, NCAM-140 strikingly enhanced cell proliferation, motility, and migration without modification of PSA on N-glycan termini. To examine the effect of PSA on NCAM function, we constructed ldlD/N/S and ldlD/N/P cell lines that expressed heavily polysialylated NCAM-140. The polysialyltransferases STX and PST had differing effects on ldlD/N140 cells: PSA catalyzed by STX, but not by PST, significantly enhanced cell migration. Similar results were observed in cancer cells MCF-7.
The hallmarks of tumor malignancy, which refer to invasion and metastasis, frequently coincide with the loss of cell-cell adhesion [43]. High levels of PSA have different effects on cell adhesion to ECM components, such as LN, heparin and matrigel [31]. In our study, cell adhesion to FN was strongly promoted, and cell adhesion to matrigel was reduced by NCAM-140, while these phenomenon was reversed by the presence of PSA. However, the decrease in adhesion to collagen IV by NCAM-140 was further reduced in ldlD/N/S and ldlD/N/P cells. In a previous study, STX and PST differentially directed PSA synthesis in temporal-and spatialspecific manners [44]. STX and PST show differing postnatal expression in various organs and catalyze different length of PSA on NCAM [45]. These properties may account for the differing effects we observed on NCAM function.
In future studies, we will use other cell lines that naturally express NCAM and PSA-NCAM to clarify how PSA enhances metastasis through up-regulation of NCAM, and the differential pathways directed by STX-vs. PST-catalyzed PSA on NCAM.
Some important findings of this study are: (i) ldlD-14 cells provide a useful model for studying and comparing the functions of NCAM and PSA-NCAM; (ii) in our model, up-regulation of NCAM-140 enhanced cell proliferation, motility, and migration, which are characteristic processes of metastatic tumors; (iii) NCAM bearing PSA was associated with enhanced cell proliferation, migration and aberrant adhesion, suggesting that PSA overexpression promotes metastasis; (iv) different polysialyltransferases (STX vs. PST) have differing effects on cell proliferation, adhesion and migration.