Discovery of an a 2,9-PolyNeu5Ac Glycoprotein in C-1300 Murine Neuroblastoma (Clone NB41A3)*

only in capsules of neuroinvasive Neisseria meningitidis group C bacteria. Here we report the unexpected discovery of a 2,9-polyNeu5Ac in a new cell adhesion-related glycoprotein on the membrane of C-1300 murine neuroblastoma cells (clone NB41A3). We also report the expression of a 2,9-polyNeu5Ac was affected by cell growth and retinoic acid-induced differentiation. Occurrence of the linkage isomer of a 2,8-polyNeu5Ac has left unrecognized by conventional methods using biological diagnostic probes for a 2,8-polyNeu5Ac. Thus, our discovery may change contemporary views of polysialobiology and pathology and open new avenues for the development of anti-neural tumor drugs.


INTRODUCTION
Polysialic acid (PSA) is a unique cell surface homopolymer that is expressed by clinically important serogroups B and C Neisseria meningitidis in α2,8and α2,9-linked isomers respectively (1) (Fig. 1). These molecules encapsulate the organisms, which are responsible for dominant portion of all cases of meningococcal meningitis, and are critical to their pathogenesis. One of these linkage isomers, α2, , which also forms the capsule of neuroinvasive E. coli K1, is poorly immunogenic (2), probably due to the fact that it is a by guest on March 24, 2020 http://www.jbc.org/ Downloaded from self-antigen occurring as an oncodevelopmentally regulated mammalian antigen in the neural cell adhesion molecule (NCAM) and certain tumors. In contrast α2,9-polyNeu5Ac is immunogenic enough to be used as a vaccine against meningitis caused by group C meningococci and is expressed by these organisms in both O-acetylated and O-deacetylated forms (2). Since the 1980s, the molecular biology, structural biology, physiological and pathophysiological functions of α2,8-polyNeu5Ac on NCAM, have already been the subject of major reviews (3)(4)(5)(6)(7). Usually detection and identification of α2,8-PSA in NCAM studies are made using monoclonal antibodies specific to α2, . Although such immunological reagents are useful for studies of developmentally regulated dynamic expression of α2,8-polyNeu5Ac, they cannot be used to define DP, or to detect either changes in the distribution of glycoforms differing in DP, or the presence of PSA differing in sequences or inter-residue linkages. Prior to the identification of α2,8-polyNeu5Ac-NCAM, we discovered α2,8-polyNeu5Gc by chemical and biochemical methods which was the first example of animal PSA (8,9). Later we also unveiled divergent forms of PSA such as an α2,8-poly(Neu5Ac, Neu5Gc) copolymer, α2,8-polyKDN, and α2,5-Oglycolyl-linked oligo/polyNeu5Gc (10,11). These PSA-glycoproteins were shown to play biological functions in fertilization and early embryogenesis, and until now, except for the last example found in sea urchin eggs, all inter-residue linkages in PSA chains found in vertebrate sources were α2,8.

Insert Figure 1 near here.
In view of the fact that the presence of a large number of glycoforms for a single functional protein is important for the delicate and fine regulation of 'protein-based functions' in higher animals, we thought that the NCAM of neuroblastoma cell lines would be an excellent model system to demonstrate how the DP of PSA chains and the structure of the core N-glycan, are modulating the adhesive and migratory behavior of neural cells. To define DP, and the change in distribution of glycoforms differing in DP, we developed an ultrasensitive fluorescent labeling method using a fluorogenic probe DMB (1,2-diamino-4,5-methylenedioxybenzene), which selectively reacts with any type of oligo/poly-αulosonic acids to introduce a strong fluorescent group at the reducing termini of these chains (DMB/HPLC-FD method) (12)(13)(14)(15)(16)(17). This method permits us to achieve the same level of sensitivity as that attained by using antibodies.
Here we report the serendipitous discovery of α2,9-linked polyNeu5Ac in a new cell adhesion-related glycoprotein on the membrane of C-1300 murine neuroblastoma cells (clone NB41A3). α2,9-PolyNeu5Ac has long been thought, until now, to be a homopolymer only found as capsular polysaccharide of N. meningitidis group C bacteria.

EXPERIMENTAL PROCEDURES
Antibodies Used ( Sources for rat anti-mouse embryonic PSA-NCAM (CD56) mAb (12F8, IgM) and anti-rat NCAM mAb (5B8, IgG) were given previously (16). Serum pooled mouse polyclonal antisera against de-O-acetyl GCMP-tetanus toxoid conjugate Ab against α2,9-polyNeu5Ac was prepared as described previously (18). Taiwan. Cells were cultured in Ham's F-10 medium, supplemented with 15% horse serum, 2.5% FBS and kept in a humidified atmosphere containing 5% CO 2 at 37°C. Cells were transferred upon reaching confluence and those within 5 passages after purchasing from ATCC were used for analysis, and harvested by trypsination after washing with phosphate-

Analysis of PSA and Total Neu5Ac in the Solubilized Glycoproteins ( PSA and total
Neu5Ac in the glycoproteins solubilized with Triton X-100 were analyzed by DMB/HPLC-FD method (16). Peaks of oligo/polysialic acids were monitored with a fluorescence detector set at 372 nm for excitation and 456 nm for emission. Because each PSA can acquire only one DMB molecule, the total peak area for DMB-oligo/polysialic acid peaks (DP > 4) represents the total number of PSA chains and used as a value comparing the level of PSA in the cells.
ELISA was also applied to evaluate quantitatively the PSA expression level in the cells by guest on March 24, 2020 http://www.jbc.org/ Downloaded from using mouse polyclonal antisera against de-O-acetylated GCMP (α2,9-polyNeu5Ac)tetanus toxoid conjugate Ab and alkaline phosphatase-conjugated goat antimouse IgM, as the primary and secondary antibody, respectively. The absorbance at 405 nm (p-nitrophenol) of the reaction mixture represents the amount of PSA epitope. All procedures used in SDS-PAGE and Western blot were essentially similar to those previously described (16). The products were hydrolyzed in 0.1 M trifluoroacetic acid at 80°C for 4 h, derivatized with DMB for 2.5 h at 55°C and analyzed for C7 and C9 nonulosonates by HPLC/FD as described elsewhere (17).

RESULTS AND DISCUSSION
The expression of PSA and NCAM in the solubilized membrane fractions was analyzed for the murine neuroblastoma, by DMB/HPLC-FD and by Western blot analysis using anti-  (Table I). NB41A3 cells were α2,8-polyNeu5Ac-negative on Western blots using 12F8 (Fig. 2C) but PSA-positive on the DMB/HPLC-FD analysis ( Fig. 2A and B). PSA residues in NB41A3 cells were also shown to be completely hydrolyzed to unsubstituted Neu5Ac monomer by treatment with Structural evidence for the presence of α2,9-polyNeu5Ac was also obtained from the results of the analysis of C7 and C9 ulosonates obtained after periodate oxidation of the sample, followed by borohydride reduction and mild acid hydrolysis. Oxidation of the α2,9-polyNeu5Ac chain with periodate resulted in its depolymerization and the destruction of all Neu5Ac residues to form their C7 analogs. The time-course of reaction was examined for by guest on March 24, 2020 http://www.jbc.org/ Downloaded from three samples, NB41A3 PSA-glycopeptide, authentic α2,9-polyNeu5Ac, and α2,8-polyNeu5Ac, by monitoring C7 and C9 monomers using the DMB/HPLC-FD method ( Fig.   3A-C). The reaction proceeded slowly (19), and the NB41A3 PSA-glycopeptide and authentic α2,9-polyNeu5Ac samples gave the C7 analog of Neu5Ac as the only product ( Fig. 3B and C) whereas α2,8-polyNeu5Ac remained unoxidized and gave C9 Neu5Ac (Fig.   3A). These results prove that the interresidue linkage of polyNeu5Ac in NB41A3 PSA-gp is α2,9.

Insert Figure 3 near here.
Supporting evidence for the presence of α2,9-polyNeu5Ac was obtained, using the DMB/HPLC-FD method, by co-injecting NB41A3-derived PSA-gp with each of the authentic samples, which resulted in a subtle retention time difference between homologous α2,8-oligo/polyNeu5Ac and α2,9-oligo/polyNeu5Ac labeled with fluorescent quinoxalinone at their reducing termini. The results are reproduced in Fig. 3D and E, and, as expected, NB41A3 PSA exhibited a simple profile when cochromatographed with authentic standard α2-9-polyNeu5Ac (Fig. 3D) whilst, when mixed with α2,8-polyNeu5Ac, a series of doublet peaks was seen on the co-chromatogram (Fig. 3E).
Interestingly, while α2,9-polyNeu5Ac has been known for well over a couple of decades as a constituent of the capsule of N. meningitidis group C (1), its occurrence in animal glycoproteins had yet to be identified. Indeed, to the best of our knowledge, there has been only one report concerning the finding of an α2, 9- Fig. 5A and B. The amount of total protein-bound Neu5Ac (ng/mg of wet cells) was also included. In the absence of RA the PSA level increased with cell culture time and reached the maximum values at confluence (day 7) (Fig. 5A). By contrast, in the presence of RA the level of PSA was high on the day 5 and gradually decreased during cell growth until it reached the lowest value at day 7 (Fig. 5B). We were not able to determine the PSA expression level at stages before the day 5 post-inoculation time, simply because the cells grew so slowly that not enough of them could be harvested.

Insert Figures 4 and 5 near here.
Cellular adhesive properties were examined in cultured NB41A3 cells. When NB41A3 cells were incubated in the absence of RA, they formed aggregates at the initial stages of proliferation (day 5 post-inoculation incubation time; Fig. 5C) and underwent dissociation of cell aggregates at later stages (day 7 ; Fig. 5C). These changes were associated with a parallel increase in the level of PSA per unit weight of wet cells (Fig. 5A). In sharp contrast to these observations, in the presence of 10 µM RA, NB41A3 cells did not form aggregates at the initial stages of incubation (day 5; Fig. 5D) and a majority of the cells differentiated to a neuronal phenotype. They extended long neuritic processes during incubation, and finally on day 7 of culture a large portion of the cells formed large aggregates (Fig. 5D). Similar morphological changes were correlated with changes in the expression level of α2,8-polyNeu5Ac chains in human neuroblastoma IMR-32 cells and pheochromocytoma PC-12 cells during growth in the absence and presence of RA (16). Since cell aggregation can be correlated with change in the expression level of PSA, the newly identified α2,9-polyNeu5Ac-expressing glycoprotein may be considered as a molecule closely related to NCAM. α2,8-PolyNeu5Ac chain was suggested to be associated with the voltage-sensitive Na-channels in adult rat brain merely by immunohistochemical and Western blot data (21).   Chromatographic data were acquired at the times of periodate oxidation indicated. (A) α2,8-polyNeu5Ac, the inner Neu5Ac residues are unattacked by periodate and only Neu5Ac was detected on hydrolysis following the periodate oxidation/reduction procedure; (B) α2, 9-polyNeu5Ac, as expected, underwent periodate oxidation to result in the destruction of the α2,9-polyNeu5Ac chain to form C7-ulosonic acid upon subsequent reduction/hydrolysis. Thus,