Owing to its poly-anionic charge and large hydrodynamic volume, polysialic acid (polySia) attached to neural cell adhesion molecule regulates axon–axon and axon–substratum interactions and signalling, particularly, in the development of the central nervous system (CNS). Expression of polySia is spatiotemporally regulated by the action of two polysialyl transferases, namely ST8SiaII and ST8SiaIV. PolySia expression peaks during late embryonic and early post-natal period and maintained at a steady state in adulthood in neurogenic niche of the brain. Aberrant polySia expression is associated with neurological disorders and brain tumours. Investigations on the structure and functions, over the past four decades, have shed light on the physiology of polySia. This review focuses on the biological, biochemical, and chemical tools available for polySia engineering. Genetic knockouts, endo-neuraminidases that cleave polySia, antibodies, exogenous expression, and neuroblastoma cells have provided deep insights into the ability of polySia to guide migration of neuronal precursors in neonatal brain development, neuronal clustering, axonal pathway guidance, and axonal targeting. Advent of metabolic sialic acid engineering using ManNAc analogues has enabled reversible and dose-dependent modulation polySia in vitro and ex vivo. In vivo, ManNAc analogues readily engineer the sialoglycans in peripheral tissues, but show no effect in the brain. A recently developed carbohydrate-neuroactive hybrid strategy enables a non-invasive access to the brain in living animals across the blood–brain barrier. A combination of recent advances in CNS drugs and imaging with ManNAc analogues for polySia modulation would pave novel avenues for understanding intricacies of brain development and tackling the challenges of neurological disorders.

Introduction

All forms of living systems utilize four major classes of building blocks, namely nucleic acids, amino acids, lipids, and carbohydrates. A variety of carbohydrates are utilized to build mono-, di-, oligo-, and polysaccharide structures, collectively known as glycans, on proteins and lipids [1]. Glycans provide enormous diversity and hence a high degree of information content to enable spatiotemporal regulation of biological processes. Among monosaccharide building blocks utilized in mammalian glycosylation pathways, sialic acids are a unique family of nonulosonic acids [2]. Sialic acids are found usually at the termini of cell surface molecules such as glycoproteins (both N- and O-linked glycans) and gangliosides. N-acetyl-d-neuraminic acid (NeuAc) and N-glycolyl-d-neuraminic acid (NeuGc) are the most abundant sialic acids [3]. Biosynthesis of NeuAc begins from uridine 5′-phospho-N-acetyl-d-glucosamine (UDP-GlcNAc), which is converted to N-acetyl-d-mannosamine 6-phosphate (ManNAc-6-P) by the bifunctional enzyme UDP-GlcNAc C2-epimerase/ManNAc kinase (GNE/MNK, EC: 3.2.1.183/2.7.1.60) [4]. ManNAc-6-P is converted to NeuAc through an aldol reaction with phosphoenol pyruvate catalyzed by NeuAc synthase (EC: 2.5.1.57) to yield NeuAc-9-P, which is then dephosphorylated by NeuAc-9-P phosphatase (EC: 3.1.3.29) to yield NeuAc. NeuAc is activated to CMP-NeuAc in the nucleus by CMP-NeuAc synthetase (2.7.7.43) and transported to the Golgi apparatus via CMP-NeuAc transporters [5]. Sialyl- and polysialyl transferases utilize CMP-NeuAc as the donor and transfer NeuAc to glycoproteins or glycolipids terminated, respectively, with neutral glycans and to a pre-attached NeuAc. Action of sialyl transferases results in NeuAcα(2 → 3)Gal and NeuAcα(2 → 6)Gal/GalNAc structures, while polysialyl transferases yield NeuAcα(2 → 8)NeuAc, NeuAcα(2 → 9)NeuAc, or alternating NeuAcα(2 → 8)NeuAcα(2 → 9) linkages [6,7]. Polysialic acid (polySia) is a homopolymer of –[NeuAcα(2 → 8)]n– (n = ∼30–200 units) found attached to a restricted set of polypeptides, glycolipids, and as capsular polysaccharide in pathogenic bacteria [8,9]. Here, we review the biological, biochemical, and chemical methods that are available for probing the structure and functions of polySia both in vitro and in vivo (Figure 1).

Biological and chemical approaches for engineering of polySia.

Figure 1.
Biological and chemical approaches for engineering of polySia.

Neural cell adhesion molecules (NCAM) in neuronal cells are post-translationally decorated with polySia in two out of three N-glycans present on the fifth Ig domain. (Inset) Physiological modulation of polySia is regulated spatiotemporally through expression of NCAM, polysialyl transferases, sialidases, and proteases. Levels of polySia could be increased using exogenous polysialyl transferases, neurostimulants, and natural metabolic precursors (i–iii). Conversely, polySia levels could be reduced using the endo-neuraminidase (Endo-N), genetic deletion of polysialyl transferases, or non-natural ManNAc analogues acting as metabolic inhibitors (a–c).

Figure 1.
Biological and chemical approaches for engineering of polySia.

Neural cell adhesion molecules (NCAM) in neuronal cells are post-translationally decorated with polySia in two out of three N-glycans present on the fifth Ig domain. (Inset) Physiological modulation of polySia is regulated spatiotemporally through expression of NCAM, polysialyl transferases, sialidases, and proteases. Levels of polySia could be increased using exogenous polysialyl transferases, neurostimulants, and natural metabolic precursors (i–iii). Conversely, polySia levels could be reduced using the endo-neuraminidase (Endo-N), genetic deletion of polysialyl transferases, or non-natural ManNAc analogues acting as metabolic inhibitors (a–c).

Expression, structure, and functions of polySia in development and disease

PolySia biosynthesis depends on the expression of polysialyl transferases. In mammals, four polysialyl transferases are known of which only two are enzymatically active, namely ST8SiaII (EC: 2.4.99.-) and ST8SiaIV (EC: 2.4.99.-) (also known as STX and PST, respectively) [6]. Expression of polysialyl transferases has been shown to peak at late-embryonic and early post-natal stages in mice indicating the importance of polysialylation for development, learning, memory, and mental health [1012]. PolySia is found attached to both glycoproteins and glycolipids. Among the proteins, neural cell adhesion molecule (NCAM) is the most studied. NCAM is abundantly polysialylated in foetal and newborn animals compared with adults [13]. In adult mice, western blotting studies using anti-polySia (clone 12F8) antibody revealed that polySia on NCAM in adult mice is restricted to the brain, while peripheral tissues show mainly expression of NCAM (120 kDa) without polySia (Figure 2a) [14]. PolySia (12F8) levels on NCAM in mice brain and hippocampus were found to be maximal at 2 days post-natal and rapidly decreased as a function of age, while no major changes were observed with NCAM levels, using anti-NCAM (clone OB11) antibody which recognizes the C-terminal epitope (Figure 2b). Expression of polySia is mainly found in cells of the central nervous system (CNS) and immune system and is under tight spatiotemporal regulation during development [15]. Aberrant expression of polySia in brain tumours, including neuroblastoma, has been reported which helps in immune evasion as well as metastasis of tumour cells [16].

Tissue and developmental stage dependent expression of polySia in mice.

Figure 2.
Tissue and developmental stage dependent expression of polySia in mice.

Western blots using anti-polySia (12F8) and anti-NCAM (OB11) of tissue lysates from BALB/cByJ mice, showing expression (a) in various organs in adult mice and (b) in the brain and hippocampus as a function of post-natal age; blots shown are representative of two replicates [14]. K, kidney; S, spleen, Lu, lungs; Li, liver; H, heart; B, whole brain; B′, brain without hippocampus; H′, hippocampus.

Figure 2.
Tissue and developmental stage dependent expression of polySia in mice.

Western blots using anti-polySia (12F8) and anti-NCAM (OB11) of tissue lysates from BALB/cByJ mice, showing expression (a) in various organs in adult mice and (b) in the brain and hippocampus as a function of post-natal age; blots shown are representative of two replicates [14]. K, kidney; S, spleen, Lu, lungs; Li, liver; H, heart; B, whole brain; B′, brain without hippocampus; H′, hippocampus.

NCAM is expressed on the cell surface in three alternately spliced isoforms, namely 180, 140, and 120 kDa. Both 180 and 140 kDa forms are polypeptides, respectively, with long and short length cytoplasmic domains, while the 120 kDa form is attached to glycosylphosphatidyl inositol (GPI) anchors on the plasma membrane. Extracellular domain of NCAM1 contains five Ig domains at the N-terminus and two fibronectin-III (FNIII) domains proximal to the plasma membrane. PolySia is attached specifically to two out of three N-glycans present on the Ig5 domain (Figure 1). Such specificity is mediated mainly by the geometrical constraints imposed by the dimensions of Golgi membrane-anchored polysialyl transferases and accessibility to their active site [10,17]. Recent studies have revealed other polypeptide candidates carrying polySia including SynCAM-1 on polydendrocytes (NG2) cells in mouse brain, sodium channels in rat brain, CD36 in murine and human milk, neuropilin-2 on dendritic cells (DC), macrophages, and microglia, C–C chemokine receptor type 7 (CCR7) on DC, E-selectin ligand-1 on macrophages and microglia, and polysialyl transferases [18]. Sperm of sea urchin has been reported to carry polySia attached to O-glycans as well as on gangliosides [19]. Pioneering studies by Rutishauser and co-workers explored the function of polySia as a permissive steric barrier that regulated cell–cell contacts and facilitated axon–axon contacts more than axon–muscle cell (substratum) contacts [20,21]. NCAM with polySia has a high propensity to display homotypic (NCAM-NCAM) and heterotypic (NCAM–L1; NCAM-FGFR, NCAM–ECM, etc.) interactions with other receptors, in a cis- (intramembrane side-on interactions) and trans- (intercellular contacts) fashions [10].

PolySia expression is essential for normal CNS development, learning, memory, and mental health. Aberrant polySia expression is associated with neurodegenerative disorders, autism, psychiatric disorders, addiction, and tumours of the brain. PolySia expression is generally thought to be limited to neurogenic niches such as hypothalamus, cortex, and hippocampus, in both rodent and human brain. Recent studies using human brain sections have revealed that polySia is expressed in many loci other than neurogenic niches, particularly on precursor neuronal cells [22]. PolySia expression was widespread, yet in specific regions, throughout the adult human brain, particularly in caudate nucleus and cerebellum; by contrast, caudate nucleus and cerebellum in rodent brain are polySia negative which highlights species-specific differences. PolySia levels were found to be markedly reduced in the entorhinal cortex in brain tissue from Alzheimer's disease patients, compared with controls, which affect neuronal plasticity. PolySia levels were selectively reduced in NeuN+ interneurons, but not in calbindin+ or calretinin+ interneurons, in the entorhinal cortex of Alzheimer's disease brain [23]. The presence of PolySia has been shown to affect the efficacy of antidepressant drugs (e.g. selective serotonin re-uptake inhibitors) and facilitate neurogenesis [24]. The role of polySia in the specific context of human brain development, disorders, and diseases remains very much underexplored.

PolySia contains multiple chemical functional groups to facilitate interactions with a variety of cell surface molecules. The large number of hydroxyl groups enables retention of water molecules through hydrogen bonding resulting in a large hydrodynamic volume. The carboxylate group at C-2 facilitates salt–bridge formation with proteins and retention of cations. The N-acetyl group at C-5 provides hydrophobic interaction surfaces at regular intervals. X-ray crystallographic analysis of pentameric polySia (NeuAc5) co-crystallized with an inactive mutant of endo-sialidase-NF (from bacteriophage K1F) revealed the flexibility of helical conformation of polySia. PolySia was found bound to three distinct locations, namely the active site, β-barrel, and the stalk domains. Helical pitch of polySia is flexible in such a way that it forms either four- or five-sialic acid residues spanning a one-turn length, respectively, of 12.4 and 14.7 Å in order to complement the polypeptide-binding surface [protein data bank (PDB) accession numbers 3GVK and 3GVJ] [10,25]. Structural diversity, in terms of variable polymer length, poly-anionic nature, hydrodynamic volume, retention of essential metal ions, plasticity of secondary structures, and dynamic conformations, endows polySia with the unique ability to interact with various polypeptides in a context and microenvironment dependent manner and fine-tune brain development.

Tools for the engineering of polySia

The development of versatile, robust, highly selective, and reliable tools is essential to understand the structure and functions of polysialylation under physiological and pathological conditions [10,26]. Multiple avenues are available for manipulation of polysialic acids in vitro in cell lines, ex vivo in primary neurons and tissue slices, and in vivo. Complex processes govern the post-translational attachment of polySia on substrate proteins. Expression of polySia on proteins depends on genes for the specific polypeptide (e.g. NCAM), N-glycan assembly, trimming, and elaboration, sialylation machinery, polysialyl transferases, endogenous sialidases, and proteases. The currently available tools for polySia engineering include enzymes, antibodies, genetic knockout and transgenic mice and cells, small molecules, and non-natural metabolic precursors. Biological, biochemical, and chemical tools for polySia engineering have their unique advantages and limitations depending on the system and context under study as illustrated below.

Biological tools for the engineering of polySia

One of the earliest tools for the manipulation of polysialic acid is the Endo-N, an endo-neuraminidase identified from a bacteriophage against an Escherichia coli strain which expresses polysialic acid (also known as colominic acid) on its capsule. Endo-N cleaves the innermost NeuAcα(2 → 8)NeuAc linkage on polySia and its application has provided significant insights in the functions of polySia in neuronal development [21,27,28] (Figure 1). Treatment of embryonic chick neurons with Endo-N leads to increased fasciculation with thinner branch diameters, whereas the control neurons displayed less fasciculation with the ability to follow defined tracts to form natural wing patterns. The phenotype resulting upon Endo-N treatment was also captured using antibodies against polySia, NCAM, and L1 which confirmed the regulation of axon outgrowth, axon path finding, and axon arborization by polySia [27].

Several studies have been reported on the phenotype of mice knockouts of ST8SiaII, ST8SiaIV, and NCAM1 as either single knockout or a combination of these genes (ST8SiaII–/– ST8SiaIV–/– double knockout and a triple knockout including NCAM1) [10]. Mice with ST8SiaII–/– or ST8SiaIV–/– showed phenotypes indicating partial redundancy in their activity; at the same time, unique brain regions were identified wherein there was no compensatory polysialylation. ST8SiaII–/– mice, in particular, showed defects in cognition and memory with a schizophrenia-like phenotype [29]. It might be concluded that both ST8SiaII and ST8SiaIV are needed for normal brain development. Compromised activities, even to a minor degree, could have consequences on axonal targeting, axon path finding, and spatial memory. The double knockout mice (ST8SiaII–/–, ST8SiaIV–/–) showed drastic alterations to brain morphology, particularly in rostral migratory stream and olfactory bulb interneurons. Interestingly, this phenotype is reversed in the triple knockout mice (ST8SiaII–/–, ST8SiaIV–/–, and NCAM–/–) with apparently normal developmental patterns. These studies showed that when NCAM is expressed but not decorated with polySia, results in untimely interactions with other receptors leading to haphazard organization of neural networks and connections [10]. Studies in embryonic and post-natal mice have shown that polySia expression is necessary in enteric neurons to reach their target sites [30]. In general, polySia on cells seems to act as an engine that could extend the neurite projections to their destined target locations and get detached once the appropriate connections were achieved. The absence of ST8SiaII or ST8SiaIV seems to affect diverse types of cross-connections involving neurons (e.g. corticothalamal and thalamocortical neurons). Steric hindrance or barrier functions alone were insufficient to explain the phenotype and behaviour of knockout mice. Subsequent studies showed that polySia on NCAM regulates downstream intracellular signalling events by directly controlling NCAM binding to receptors on the neighbouring cells through trans-interactions.

Gene knockout studies in mice have highlighted not only the importance of polySia for brain development, but also the spatiotemporal control of polySia expression and its regulation. However, biological approaches to polySia do have limitations in terms of the complexity of resulting phenotypes, irreversibility, lack of control on dose–response, and direct applications for therapeutic development.

Chemical and pharmacological tools for the engineering of polySia

Various chemical tools are currently available for the modulation, manipulation, and inhibition of polysialic acid both in vitro and in vivo. The advent of metabolic glycan engineering (MGE) has enabled the investigations of complex cellular physiology of glycosylation in general [31]. MGE exploits the permissivity of enzymes involved in various glycan biosynthetic pathways for processing non-natural monosaccharide analogues. MGE provides exquisite control in terms of modulation of biological properties, ease of application in vivo, dose–response control, reversibility, and potential for drug development. Depending on the specific chemical structure of the metabolic precursor employed, multiple outcomes, such as modulation of cell–cell and cell–pathogen interactions, cell migration, cell tagging and tumour imaging in vivo, cell attachment to engineered surfaces, modulation of immunogenicity, inhibition of biosynthesis, and modulation of carbohydrate–lectin interactions, could be achieved [32]. ManNAc analogues are the most suited for engineering of sialic acids as ManNAc is a committed precursor and it enables bypassing of the feedback inhibition of GNE/MNK by CMP-NeuAc to achieve increased flux. ManNAc analogues with modifications at the N-acyl moiety carrying alkyl, ketone, azidoacetyl, alkynyl, glycolyl, thioglycolyl, arylalkyl, and diazirine have been reported for various biological applications [3337]. MGE has become a prominent tool for investigation of glycoconjugates both in vitro and in vivo as discussed below.

Metabolic engineering of polySia in vitro using ManNAc analogues

ManNAc analogues were employed for polySia engineering either as free monosaccharides, usually, millimolar concentrations, or as peracetylated derivatives, in micromolar concentrations. Peracetylation improves the cellular uptake by diffusion and processing by an order of magnitude (Figure 3) [38]. To achieve metabolic incorporation into polySia, ManNAc analogues must be processed through several enzymatic steps of the sialic acid biosynthetic pathway (Figure 4). Biosynthesis of polySia on polypeptide carriers, its chain length, and polydispersity are governed by a multitude of factors. Gene expression, protein quantities, intracellular localization of proteins, and activities of all the enzymes involved in the sialic acid biosynthesis, CMP-NeuAc transporters, and (poly)sialyltransferases could affect the expression and quantity of polySia (Figure 4). More importantly, intracellular concentrations of the metabolic intermediates (small molecules) govern the flux of polySia biosynthesis and impose limiting conditions.

Metabolic sialic acid engineering (MSE).
Figure 3.
Metabolic sialic acid engineering (MSE).

Chemical structure of N-acetyl-d-mannosamine (ManNAc) analogues employed for the engineering of polySia. Earlier studies employed (a) free monosaccharides in millimolar concentrations; (b) peracetylation improved the efficiency effecting MSE at micromolar concentrations. (c) The CNHs, Ac3ManNAc-Nic, Ac3ManNAz-Nic, and Ac3ManNBut-Nic were able to modulate polySia in brain across the BBB. The non-hybrids Ac4ManNAc, Ac4ManNBut, and Ac4ManNAz showed little or no effect in the brain of living animals.

Figure 3.
Metabolic sialic acid engineering (MSE).

Chemical structure of N-acetyl-d-mannosamine (ManNAc) analogues employed for the engineering of polySia. Earlier studies employed (a) free monosaccharides in millimolar concentrations; (b) peracetylation improved the efficiency effecting MSE at micromolar concentrations. (c) The CNHs, Ac3ManNAc-Nic, Ac3ManNAz-Nic, and Ac3ManNBut-Nic were able to modulate polySia in brain across the BBB. The non-hybrids Ac4ManNAc, Ac4ManNBut, and Ac4ManNAz showed little or no effect in the brain of living animals.

Metabolic processing of non-natural ManNAc analogues.
Figure 4.
Metabolic processing of non-natural ManNAc analogues.

The sialic acid biosynthetic pathway is shown taking Ac4ManNBut as an example. Processing of endogenous ManNAc results in polySia with n = 30–200 units (A). Ac4ManNBut is uptaken by cells by diffusion (pinocytosis), de-acetylated by non-specific esterases, phosphorylated, and coupled with phosphoenol pyruvate to yield SiaBut-9-P. Dephosphorylation of SiaBut-9-P results in SiaBut which is activated to CMP-SiaBut in the nucleus and then transported to the Golgi through the CMP-NeuAc transporters. Sialyl and polysialyl transferases resident in the Golgi membrane utilize CMP-SiaBut for decoration on to sialoglycoproteins. It is remarkable that CMP-SiaBut competes with endogenous concentrations of CMP-NeuAc to induce inhibition of polySia (B). It is necessary that non-natural analogues are processed by each of the enzymes (shown in italics) involved in the pathway.

Figure 4.
Metabolic processing of non-natural ManNAc analogues.

The sialic acid biosynthetic pathway is shown taking Ac4ManNBut as an example. Processing of endogenous ManNAc results in polySia with n = 30–200 units (A). Ac4ManNBut is uptaken by cells by diffusion (pinocytosis), de-acetylated by non-specific esterases, phosphorylated, and coupled with phosphoenol pyruvate to yield SiaBut-9-P. Dephosphorylation of SiaBut-9-P results in SiaBut which is activated to CMP-SiaBut in the nucleus and then transported to the Golgi through the CMP-NeuAc transporters. Sialyl and polysialyl transferases resident in the Golgi membrane utilize CMP-SiaBut for decoration on to sialoglycoproteins. It is remarkable that CMP-SiaBut competes with endogenous concentrations of CMP-NeuAc to induce inhibition of polySia (B). It is necessary that non-natural analogues are processed by each of the enzymes (shown in italics) involved in the pathway.

The application of MGE to polysialic acid was first reported by Bertozzi and co-workers using N-levulinoyl-d-mannosamine (ManNLev) [39]. ManNLev was biosynthetically converted to N-levulinoyl-d-neuraminic acid (SiaLev) in retinoic acid-differentiated NT2 (human embryonic carcinoma) cells cultured either on matrigel or on primary astrocytes. SiaLev was found to be incorporated into polySia-NCAM in a dose-dependent manner. SiaLev expression on polySia-NCAM was confirmed by exploiting the biotin-linker-hydrazide reaction specific for the ketone functional group. Three antibodies were employed for characterization of polySia-NCAM, namely clone 12F8 and mAb735 (both recognizing native polySia), and clone OB11 which recognizes the C-terminus of NCAM. SiaLev was found to be resistant to hydrolysis by Clostridium perfringens sialidase, but could be hydrolyzed by Endo-NE. At concentrations greater than 1.0 mM of ManNLev, lower levels of 12F8 and OB11 epitopes were found indicating partial inhibition. However, no direct evidence for engineering of polySia was presented. These results indicated that CMP-SiaLev was able to compete with endogenous CMP-NeuAc for utilization by polysialyl transferases.

Subsequently, it was reported that N-butanoyl-d-mannosamine (ManNBut) was an efficient inhibitor of polysialylation in NT2 neurons [40]. Incubation with ManNBut, but not ManNAc, abolished expression of 12F8 epitope in a dose-dependent manner with maximal effect at 3.0 mM. Treatment with ManNProp was found to have an intermediate effect. To address reversibility of polySia inhibition, cells were incubated with 5.0 mM ManNBut for 2 days and then the analogue was removed. Time course immunoprecipitation studies showed that the effect was sustained until 3 days after removal of ManNBut and the wild-type polysialylation resumed partly by day 4 and fully by day 5. Estimation of total sialic acids, estimated using the periodate-resorcinol assay, revealed that ManNBut did not inhibit overall sialylation, but rather had a selective inhibitory effect on polySia biosynthesis. In another study, ManNBut, but not ManNProp, has been shown to inhibit polySia in enteric neurons from foetal rat gut and prevent BMP-4-induced neuronal clustering [30].

Inhibition of polySia-NCAM biosynthesis by ManNBut could be explained through at least two distinct mechanisms, namely (a) inhibition and slow turnover rate of CMP-SiaBut utilization and (b) inhibition and slow turnover rate for the addition of CMP-NeuAc to a primed polySia-containing SiaBut. Studies on enzymatic sialylation of NCAM-Fc chimeric protein by ST8SiaII employing synthetic CMP-NeuAc, CMP-SiaProp, or CMP-SiaBut as the donor revealed slower kinetics for the transfer of SiaBut compared with NeuAc. SiaProp exhibited intermediate transfer rates; similar, albeit moderate, inhibitory results were observed for polysialylation of NCAM-Fc by ST8SiaIV. Similarly, when polySia-NCAM-Fc primed with NeuAc, SiaProp, or SiaBut were employed as acceptors for transfer of CMP-NeuAc by ST8SiaII, slower rates were observed for SiaBut-terminated acceptors compared with NeuAc-terminated acceptors [40]. Hence, the drastic inhibition of polySia on NCAM observed in NT2 neurons induced by ManNBut could be the result of operation of both the mechanisms.

In one of the earlier approaches to cancer immunotherapy, Jennings and co-workers employed ManNProp for metabolic modulation of polySia in RBL-2H3 (rat) and RMA (mice) leukaemia cell lines [41]. This approach relies on selective metabolism of ManNProp resulting in the biosynthesis of polySia-containing SiaProp and its subsequent binding by specific antibodies (mAb 13D9 generated against poly-SiaProp) to initiate tumour cell death processes. The expressions of native polySia (mAb 735) and N-propanoyl-polySia (mAb 13D9) were found to be inversely correlated. Growth of RMA tumours in mice was inhibited by 50% of control (PBS) when treated with a combination of 13D9 and ManNProp. It was proposed that metabolic incorporation of ManNProp in polySia results in unique epitopes that react with 13D9, but not with native polySia. Treatment of NT2 neurons with ManNProp or ManNBut resulted in the expression of 13D9 epitopes indicating the cross-reactivity of 13D9 with N-butanoyl-polySia [42].

Studies by Horstkorte and co-workers have focused on the potential of ManNAc analogues as a novel therapeutic avenue for neuroblastoma in combination with anti-cancer drugs [43]. SH-SY5Y (human neuroblastoma) cells were employed as an in vitro model system for polySia engineering and its functional consequences on migration and invasion, the processes that drive cancer metastasis. Upon incubation with 10 mM of ManNProp or ManNPent for 72 h, there was an ∼90% reduction in the mAb 735 epitopes on the SH-SY5Y cell surface indicating the reduction in the biosynthesis of native polySia. On the other hand, incubation with 10 mM ManNAc leads to an increase in 15% in mAb 735 epitopes over untreated control cells indicating increased flux of polySia biosynthesis. PolySia chains containing polydispersity of n = 8–22 were found in cells treated with ManNAc, ManNProp, and controls. The relative amount of each chain length was consistently higher by ∼35% for ManNAc treatment and lower by ∼60% for ManNProp treatment. Interestingly, ManNPent treatment inhibited biosynthesis of polySia with chain length of eight units or higher. Total sialic acid levels, studied by HPLC using 1,2-diamino-4,5-methylenedioxybenzene (DMB) derivatization of sialic acids, indicated that non-natural ManNAc analogues did interfere with sialic acid biosynthesis and sialylation in general. Migration assays of SH-SY5Y cells pre-treated with ManNAc analogues showed that ManNProp and ManNPent reduced cell migration, respectively, by 25 and 60% of controls. Invasion assays through an artificial extracellular matrix mixture showed reduction in the number of invading cells by 30 and 55% with respect to controls, respectively, upon treatment with ManNProp and ManNPent.

It has been hypothesized that cancer cells display hyper-glycosylation, including expression of mucins and polySia, which blocks the uptake of cancer drugs and aids the development of drug resistance. Thus, reduction in polySia on neuroblastoma cell surface might enable better uptake of anti-cancer drugs and thus enhance their efficacy. Pre-treatment of SH-SY5Y cells with ManNProp or ManNPent, followed by incubation with 5-fluorouracil (5-FU), improved the IC50 values by 1.5-fold and 1.7-fold, respectively, compared with treatment with 5-FU alone [43]. Pre-treatment with ManNPent, but not with ManNProp, improved the cytotoxic activity of cisplatin towards SH-SY5Y cells by 20% at low doses. Notably, ManNPent treatment, not ManNProp, improved the susceptibility of neuroblastoma cells to γ-radiation. Direct comparisons of application of ManNAc analogues for enhancement of drug sensitivity against cancer cells could not be arrived at as ManNBut was not included in these studies. However, recent studies have shown that polySia (12F8) on NCAM levels in SH-SY5Y cells was inhibited by Ac4ManNBut at 100 μM [44]. These results suggested that inhibition of polySia levels using small molecules might have synergistic and beneficial effects for combinatorial cancer chemotherapy.

To elucidate differential inhibitory effects of ManNAc analogues between ST8SiaII and ST8SiaiV, ectopic expression systems in vitro were investigated. Studies using HeLa cells stably transfected with NCAM (140 kDa) in combination with either ST8SiaII or ST8SiaIV showed that ST8SiaII was inhibited by ManBut more efficiently compared with ST8SiaIV [45]. Whole cell ELISA experiments for detection of polySia using 12F8 antibody showed that ManNAc increased polySia production for both ST8SiaII and ST8SiaIV transfectants. ManNBut showed an IC50 of 2.7 and 2.0 mM for ST8SiaII and ST8SiaIV, respectively; ManNPent also inhibited polySia levels with IC50 values of 4.1 and 5.9 mM for ST8SiaII and ST8SiaIV, respectively. Western blotting using mAb 735 revealed that the polySia levels were much higher in HeLa-NCAM-ST8SiaII cells compared with HeLa-NCAM-ST8SiaIV cells, indicating that ST8SiaIV might need prior action by ST8SiaII for elongation of polySia. Drastic reduction in mAb 735 epitopes was observed at 3.0 mM of ManNBut, while the ManNProp effects were apparent at 10 mM.

Inhibition of polySia by ManNBut was predominantly mediated through inhibition of ST8SiaII, rather than ST8SiaIV. Studies in HL-60 cells (non-transfected), PC12 cells stably transfected with ST8SiaIV, NT-2 cells (non-transfected), and CHO cells stably expressing either ST8SiaII or ST8SiaIV revealed that the differential inhibition by ManNBut was amplified under in vitro cellular conditions [46]. Western blotting using clone mAb 735 (which recognizes native polySia) and clone 13D9 (which recognizes polySia acids carrying elongated N-acyl hydrocarbon chains but not the native N-acetyl moiety) revealed that PSA-NCAM expression was unaffected in HL-60 cells upon treatment with 5.0 mM of ManNAc, ManNProp, ManNBut, or ManNPent with concomitant expression of polySia carrying SiaProp and SiaBut. HPLC measurement of total native and non-natural sialic acid levels from HL-60 cells treated with ManNAc analogues showed expression of 7.7, 5.3, and 17.2%, respectively, of SiaProp, SiaBut, and SiaPent, indicating better utilization of ManNPent. This is attributable to limited steric preference for derivatives of ManNProp and ManNBut by pathway enzymes; for ManNPent, it is possible that the N-pentanoyl side chain is too long and might be exposed to solvents while binding to the enzymes. However, detailed structural investigations on N-acyl chain length and complementary interactions with sialic acid pathway enzymes are not available currently. Similar results were found in PC12 cells stably expressing ST8SiaIV. In contrast, NT-2 cells showed significant inhibition of PSA-NCAM upon treatment with the non-natural metabolic precursors ManNProp, ManNBut, and ManNPent, when probed with a mixture of anti-polySia antibodies (mAb 735 and 13D9) to detect both native and engineered polySia. The contrasting effect of ManNAc analogues on polySia biosynthesis in HL-60 and NT-2 neurons was attributed to the differential gene expression of ST8SiaII and ST8SiaIV. HL-60 cells expressed mRNA for ST8SiaIV exclusively with no detectable ST8SiaII levels. Although NT-2 cells expressed both the polysialyl transferases, the mRNA levels of ST8SiaII were found to be 10-fold higher than ST8SiaIV. Thus, HL-60 and NT-2 cells served as complementary systems to study differential effects of ManNAc analogues [46]. These results are consistent with the earlier studies on enzymatic reactions wherein ST8SiaII showed a higher preference to utilize CMP-SiaProp, CMP-SiaBut compared with ST8SiaIV [40].

Further evidence of differential effect of ManNAc analogues on polysialyl transferases was obtained in CHO cells through exogenous expression. Incubation of CHO-2A10-ST8SiaII cells — but not of CHO-2A10-ST8SiaIV — with ManNProp, ManNBut, or ManNPent resulted in drastic reduction in polySia (mAb 735) levels [4648]. At first look, this suggests that ST8SiaII does not accept unnatural SiaProp, SiaBut, or SiaPent as substrates. However, in the absence of detailed kinetic studies on these enzymes, these results remain open to alternative interpretations. Considering that mAb 735 needs at least eight units of sialic acids and binds only to native polySia, the reduction observed in CHO-2A10-ST8SiaII could be attributed to efficient utilization and incorporation of non-natural Sia donors in polySia. On the contrary, it might be interpreted that ST8SiaIV is the more stringent enzyme and prefers CMP-NeuAc over non-natural donors. Such interpretation would be consistent with other reports. Apparent reduction in polySia could occur due to epitope masking or epitope loss. The dependence of antibodies, such as mAb 735 for native and 13D9 for non-natural polySia, for specific chain lengths, chemical structures, and conformations should be kept in mind during interpretation.

It is known that during brain development, the presence or absence of polysialic acid on cells of the substratum could favour either axon–axon or axon–matrix interactions [27]. To investigate the consequence of polySia inhibition on neuronal behaviour, HeLa-NCAM-ST8SiaII cells were first incubated without or with 10 mM ManNBut for 3 days and used as a substratum layer for co-culture of neurons isolated from chick dorsal root ganglia [45]. HeLa-NCAM-ST8SiaII cells showed a 35% increase in neurite length compared with non-transfected cells; upon ManNBut treatment, the gain in neurite length induced by ST8SiaII was abrogated thus indicating the major role played by polysialic acid in facilitating neurite outgrowth. These results confirmed the ability of ManNBut to selectively inhibit polySia synthesis in a reversible and dose-dependent manner.

Gene mutations that abrogate allosteric CMP-NeuAc-binding result in overproduction of sialic acid as found in the case of sialuria and associated mental retardation [49,50]. The possibility that polySia levels are dependent on the intracellular concentrations of NeuAc was illustrated elegantly using CHO-2A10-ST8SiaII and CHO-2A10-ST8SiaIV cells transfected with GNE (wild-type) or GNE-sialuria (R263L mutant) genes [47]. The polySia-NCAM (mAb 735) levels were increased by 2.5-fold in cells expressing GNE-sialuria, while there was no gain in GNE wild-type transfectants. Interestingly, the polySia levels were increased by 3.0-fold in cells transfected with GNE wild type when treated with 10 mM ManNAc thus confirming the ability to bypass the natural CMP-NeuAc feedback inhibition. Notably, ManNAc rescued polysialylation in embryonic stem cells that were either heterozygous (+/−) or homozygous (−/−) knockouts of GNE. Mice with transgenic expression of sialuria mutant GNE (R263L) showed accumulation of sialic acid in brain cytoplasm and higher levels of polySia on NCAM compared with wild-type mice [51]. The converse scenario is observed in the case of hereditary inclusion body myopathy (HIBM), wherein the mutations in GNE/MNK affect sialic acid production and hence biosynthesis of sialoglycoconjugates including polySia [52]. The role of polySia in the onset of HIBM and other myopathies is currently not known.

In general, depending on the specific chemical structure of ManNAc analogues, multiple scenarios could be visualized for the effect on polySia (Figure 5). Chain length of polySia could be unaffected with or without incorporation of the non-natural moieties, reduction in chain length with or without metabolic incorporation of non-natural moieties, or a total replacement of natural NeuAc with non-natural moieties. ManNBut appears to induce inhibition via metabolic incorporation followed by reduction in polySia chain length.

Effects of modulation of polySia through metabolic sialic acid engineering (MSE).
Figure 5.
Effects of modulation of polySia through metabolic sialic acid engineering (MSE).

Metabolism of ManNAc analogues could result in various outcomes depending on their specific chemical structure and preference of the pathway enzymes. These possibilities are (a) wild-type, (b) truncation without metabolic incorporation (classical inhibition of polysialyl transferases), (c) metabolic incorporation without truncation, (d) truncation with metabolic incorporation, and (e) uniform replacement of NeuAc with NeuR (analogues). Only the N-glycan structures are shown without polypeptide for simplicity. Magenta diamond, NeuAc; red diamond, NeuAc analogues (e.g. SiaProp and SiaBut).

Figure 5.
Effects of modulation of polySia through metabolic sialic acid engineering (MSE).

Metabolism of ManNAc analogues could result in various outcomes depending on their specific chemical structure and preference of the pathway enzymes. These possibilities are (a) wild-type, (b) truncation without metabolic incorporation (classical inhibition of polysialyl transferases), (c) metabolic incorporation without truncation, (d) truncation with metabolic incorporation, and (e) uniform replacement of NeuAc with NeuR (analogues). Only the N-glycan structures are shown without polypeptide for simplicity. Magenta diamond, NeuAc; red diamond, NeuAc analogues (e.g. SiaProp and SiaBut).

Other small-molecule-based approaches for engineering of polySia

Apart from the ManNAc analogues mentioned above, other complementary small-molecule-based approaches for polySia modulation have been reported. Anaesthetic alkaloids such as curare have been shown to increase polySia levels and facilitate nerve sprouting [27]. Increased levels of polySia have been reported in heroin addicts and blood levels of morphine positively correlated with increased polySia-containing neurons [53]. Tegaserod, a small-molecule mimetic of polySia obtained through phenotypic screens of in vitro neurite cultures, has been shown to improve recovery in mouse models of spinal cord injury [54]. Efforts towards the development of CMP-NeuAc analogues [55] and inhibitors of (poly)sialyl transferases [56] have also been reported which could potentially be employed to alter polySia levels in biological systems.

Metabolic engineering of polySia in vivo using ManNAc analogues

Studies discussed above illustrate the versatility of ManNAc analogues to engineer polySia. However, these studies were restricted to enzymatic reactions, in vitro in cells with and without exogenous expression, and ex vivo in primary neurons. To develop applications to pre-clinical, clinical, and biomedical situations, it is imperative to study these approaches in animals. The expression of non-physiological sialic acids was first shown by Reutter in 1992 in vivo in rats using ManNProp [57]. Bertozzi and co-workers showed engineering of peripheral tissues in mice using Ac4ManNAz and Ac4GalNAz, respectively, for sialoglycoconjugates and mucin-type glycoproteins [58,59]. Horstkorte and co-workers employed Ac4ManNProp for engineering of various tissues in mice [60]. Brindle, Chen, and co-workers have employed Ac4ManNAz for selective imaging of tumours in animals taking advantage of higher metabolic and growth rate of cancer cells [61,62]. Recently, Chen and co-workers have employed 9-azido-N-acetyl-d-neuraminic acid (Sia9Az) encapsulated in liposomes for engineering of various tissues in mice [63].

However, very few studies are reported to date on direct non-invasive modulation of polysialic acid in the CNS in animals (Table 1) [64]. Applications that seek to investigate physiological roles of polySia in the development of CNS should be able to achieve delivery of ManNAc analogues across the blood–brain barrier (BBB). Sialic acid biosynthesis in the brain largely depends on the supply of glucose, which is imported through glucose transporters. There are no known transporters for N-acetyl-d-hexosamines (HexNAc). Studies in mice using Ac4ManNAz, Ac4GalNAz, and Ac4ManNProp (also as ManNProp in rats) showed facile expression on glycoconjugates in the form, respectively, of SiaNAz, GalNAz, and SiaProp in peripheral organs [5860]. Notably, in all cases, very little or no expression was found in the brain thus supporting the argument that the peracetylated HexNAc derivatives are not transported across BBB and their bioavailability in the brain, if any, is minimal. Indeed, very little replacement of NeuAc with SiaProp was detected in the brain in mice treated intraperitoneally with Ac4ManNProp at 200 mg/kg/mice with twice a day administration for 13 days [60]. However, western blotting for polySia-NCAM using mAb 735 showed reduction by 60% by day 13, while mAb 13D9 epitopes were found to increase slightly from day 4 onwards. It was concluded that 13D9 epitopes are not restricted only to polySia chains and that the minute quantities of Ac4ManNProp reaching the brain were sufficient to engineer polySia. Recently, a liposome-assisted bioorthogonal reporter (LABOR) strategy has been reported, wherein 9-azido-N-acetyl-d-neuraminic acid (9-Az-NeuAc) was encapsulated in liposomes and intravenously administered in mice, once daily, for 7 days [63]. Expression of 9-Az-NeuAc was observed in the brain as confirmed by in vivo imaging using strain promoted azide-alkyne cycloaddition. However, it is not known if polySia is engineered in the brain with this approach.

Table 1
Summary of hexosamine analogues used in vivo in mice/rats, targeted disease models, and their effects on the brain tissue
S. No. ManNAc analogues Effect on sialoglycoconjugates and application in vivo in mice or rats Biological disease models employed Effect on polysialic acid in brain tissue in vivo Reference 
ManNProp Metabolic sialic acid engineering (MSE) of peripheral organs in rats and tumour growth arrest in mice Rat basophilic leukaemia, Morris 7777 hepatoma, and mouse leukaemia Not reported [41,57
Ac4ManNProp MSE of peripheral organs; very little expression in the brain C57BL/6j mice as a model for systemic engineering of multiple tissues Yes; mild reduction in polySia mAb 735 and increase in 13D9 epitopes [60
Ac4ManNAz MSE of peripheral organs and selective imaging of tumours; very little or no expression in the brain Lewis lung carcinoma, melanoma, and adenocarcinoma No expression [58,61,62
Ac4GalNAz MSE of peripheral organs for mucin-type O-glycosylation; very little or no expression in the brain Model for systemic engineering of multiple tissues No expression [59
Ac4ManNBut Not reported in peripheral organs Control for testing BBB permeability in BALB/cByJ and C57BL./6j mice No change in polySia-NCAM levels [44
Ac4ManNAc Not reported in peripheral organs Control for testing BBB permeability in BALB/cByJ and C57BL./6j mice No change in polySia-NCAM levels [44
9-Az-NeuAc (in liposomes) MSE of various organs including the brain Liposome-mediated delivery to the brain Effects on polySia unknown [63
Ac3ManNAz-Nic MSE of various organs including the brain Exploiting natural carrier-mediated transport systems for delivery to the brain Yes; transported across the BBB and polySia engineered. [44
Ac3ManNBut-Nic MSE of various organs including the brain Exploiting natural carrier-mediated transport systems for delivery to the brain Yes; transported across the BBB and reduced polySia levels in adult brain [44
S. No. ManNAc analogues Effect on sialoglycoconjugates and application in vivo in mice or rats Biological disease models employed Effect on polysialic acid in brain tissue in vivo Reference 
ManNProp Metabolic sialic acid engineering (MSE) of peripheral organs in rats and tumour growth arrest in mice Rat basophilic leukaemia, Morris 7777 hepatoma, and mouse leukaemia Not reported [41,57
Ac4ManNProp MSE of peripheral organs; very little expression in the brain C57BL/6j mice as a model for systemic engineering of multiple tissues Yes; mild reduction in polySia mAb 735 and increase in 13D9 epitopes [60
Ac4ManNAz MSE of peripheral organs and selective imaging of tumours; very little or no expression in the brain Lewis lung carcinoma, melanoma, and adenocarcinoma No expression [58,61,62
Ac4GalNAz MSE of peripheral organs for mucin-type O-glycosylation; very little or no expression in the brain Model for systemic engineering of multiple tissues No expression [59
Ac4ManNBut Not reported in peripheral organs Control for testing BBB permeability in BALB/cByJ and C57BL./6j mice No change in polySia-NCAM levels [44
Ac4ManNAc Not reported in peripheral organs Control for testing BBB permeability in BALB/cByJ and C57BL./6j mice No change in polySia-NCAM levels [44
9-Az-NeuAc (in liposomes) MSE of various organs including the brain Liposome-mediated delivery to the brain Effects on polySia unknown [63
Ac3ManNAz-Nic MSE of various organs including the brain Exploiting natural carrier-mediated transport systems for delivery to the brain Yes; transported across the BBB and polySia engineered. [44
Ac3ManNBut-Nic MSE of various organs including the brain Exploiting natural carrier-mediated transport systems for delivery to the brain Yes; transported across the BBB and reduced polySia levels in adult brain [44

Delivery of anti-cancer drugs, therapeutic antibodies, and imaging contrast agents to brain has been a long-time challenging problem in biomedicine [6567]. Several approaches such as chemical delivery systems, exploitation of receptor-mediated endocytosis, and osmotic pressure modulation (e.g. d-mannitol) of BBB have been reported. Several transport systems such as ion channels, ATP-transport, cationic amino acids transporter, fatty acid transporters, and vitamin transporters are expressed abundantly, for nutrient import, waste export, and maintenance of brain homeostasis. Inspired by these transporter mechanisms, we recently designed the carbohydrate-neuroactive hybrid (CNH) strategy for engineering of polySia in the brain in living mice (Figure 6) [44]. CNH strategy involves piggybacking of hexosamine analogues on known neuroactive molecules (e.g. CNS acting drugs, vitamins, and neurotransmitters) via a biodegradable linkage in a pro-drug-like approach (Figure 3). Conjugates of ManNAc analogues with vitamin B3 (nicotinate) and valproate (a known anti-epileptic agent) connected via an ester bond were synthesized and the corresponding non-hybrids were employed as controls. Using expression of SiaAz in sialoglycoproteins as a read-out, it was shown that CNH hybrids achieved facile expression in the brain at concentrations that the non-hybrid molecules could not. Intravenous administration of Ac3ManNAz-Nic, but not Ac4ManNAz, at a dosage of 0.26 mmol/kg (once daily for 3 days) resulted in robust expression of NeuAz carrying sialoglycoproteins in mouse brain. The ability to metabolically tag sialoglycoproteins and gangliosides of the brain in vivo with NeuAz using CNH strategy and subsequent bioorthogonal ligations provides a promising tool for enumeration of sialoglycoconjugates in brain development as well as enable quantitative differential glycoproteomic studies in CNS disease models.

CNH strategy for non-invasive access to the brain.
Figure 6.
CNH strategy for non-invasive access to the brain.

Piggybacking of ManNAc analogues on neuroactive molecules, through biodegradable covalent linkage, enables transport across the BBB. Nicotinate conjugate of N-butanoyl-d-mannosamine (Ac3ManNBut-Nic), but not Ac4ManNBut, reduced polySia levels significantly on NCAM in living mice upon intravenous administration.

Figure 6.
CNH strategy for non-invasive access to the brain.

Piggybacking of ManNAc analogues on neuroactive molecules, through biodegradable covalent linkage, enables transport across the BBB. Nicotinate conjugate of N-butanoyl-d-mannosamine (Ac3ManNBut-Nic), but not Ac4ManNBut, reduced polySia levels significantly on NCAM in living mice upon intravenous administration.

The observation that piggybacking of ManNAc analogues on neuroactive molecules results in facile transport across BBB provides new avenues for manipulation of brain glycosylation in living animals. Similarly, administration of Ac3ManNBut-Nic, but not Ac4ManNBut (Figure 3), at a dosage of 173 mg/kg (0.36 mmol/kg) on alternate days with a total of four injections, resulted in a significant reduction in polySia on NCAM to 60% of control levels in the mouse brain (Figure 6) [44]. These results confirmed that the CNH strategy readily facilitated transport of ManNAc analogues across BBB. Several studies have hypothesized connection of polySia levels to autism spectrum disorders, psychiatric disorders, and disability in learning and memory [49]. However, causal relationships have not been established, to date, due to lack of tools that allow probing under physiological conditions. In this context, the CNH strategy provides a promising platform to access brain in a non-invasive manner for physiological and behavioural studies in animal models in the near future.

Conclusions

The intricate and complex biology of polySia has been a topic of intense research for the last four decades for neuroscientists and glycobiologists. In parallel, great strides have been achieved in the fields of neuroscience, mental health, development of CNS acting drugs, and imaging [68]. The biological, biochemical, and chemical approaches discussed here showcase the versatility of the modern tools. Especially, in the context of worldwide efforts on brain mapping projects [69,70], non-invasive tools for in vivo modulation of polySia are vital for unravelling the hidden secrets of the brain and mind. Combining the tools of CNS acting drugs with polySia engineering tools is likely to shed more light on the complex biology of development, disorders, and diseases of the brain.

Summary
  • Polysialic acid (polySia) governs major neuro-developmental signalling owing to its poly-anionic nature, size, hydrodynamic volume, and interaction with cell surface molecules both in a permissive and in a restrictive manner.

  • Expression of polySia could be manipulated in vitro and in vivo using antibodies, enzymes, and small molecules, particularly through the metabolic glycan engineering methodology.

  • Synthetic ManNAc analogues could either enhance or inhibit polySia expression depending on the specific chemical structure.

  • PolySia expression in mouse brain could be altered in vivo using the CNH strategy in a non-invasive manner.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Varki
,
A.
and
Gagneux
,
P
. (
2015
) Biological functions of glycans. In
Essentials of Glycobiology
(
Varki
,
A.
,
Cummings
,
R.D.
,
Esko
,
J.D.
,
Freeze
,
H.H.
,
Stanley
,
P.
,
Hart
,
G.W.
et al. eds), pp.
77
88
,
Cold Spring Harbor
,
NY
2
Varki,
A.
,
Schnaar,
R.L
and
Schauer,
R.
(
2015
) Sialic acids and other nonulosonic acids. In
Essentials of Glycobiology
(
Varki
,
A.
,
Cummings
,
R.D.
,
Esko
,
J.D.
,
Freeze
,
H.H.
,
Stanley
,
P.
,
Hart
,
G.W.
et al. eds), pp.
179
195
,
Cold Spring Harbor
,
NY
3
Angata
,
T.
and
Varki
,
A.
(
2002
)
Chemical diversity in the sialic acids and related alpha-keto acids: an evolutionary perspective
.
Chem. Rev.
102
,
439
469
4
Lucka
,
L.
,
Krause
,
M.
,
Danker
,
K.
,
Reutter
,
W.
and
Horstkorte
,
R.
(
1999
)
Primary structure and expression analysis of human UDP-N-acetyl-glucosamine-2-epimerase/N-acetylmannosamine kinase, the bifunctional enzyme in neuraminic acid biosynthesis
.
FEBS Lett.
454
,
341
344
5
Tanner
,
M.E.
(
2005
)
The enzymes of sialic acid biosynthesis
.
Bioorg. Chem.
33
,
216
228
6
Nakayama
,
J.
,
Angata
,
K.
,
Ong
,
E.
,
Katsuyama
,
T.
and
Fukuda
,
M.
(
1998
)
Polysialic acid, a unique glycan that is developmentally regulated by two polysialyltransferases, PST and STX, in the central nervous system: from biosynthesis to function
.
Pathol. Int.
48
,
665
677
7
McGowen
,
M.M.
,
Vionnet
,
J.
and
Vann
,
W.F.
(
2001
)
Elongation of alternating alpha 2,8/2,9 polysialic acid by the Escherichia coli K92 polysialyltransferase
.
Glycobiology
11
,
613
620
8
Pon
,
R.A.
,
Lussier
,
M.
,
Yang
,
Q.-L.
and
Jennings
,
H.J.
(
1997
)
N-Propionylated group B meningococcal polysaccharide mimics a unique bactericidal capsular epitope in group B Neisseria meningitidis
.
J. Exp. Med.
185
,
1929
1938
9
Stein
,
D.M.
,
Robbins
,
J.
,
Miller
,
M.A.
,
Lin
,
F.-Y.
and
Schneerson
,
R.
(
2006
)
Are antibodies to the capsular polysaccharide of Neisseria meningitidis group B and Escherichia coli K1 associated with immunopathology?
Vaccine
24
,
221
228
10
Schnaar
,
R.L.
,
Gerardy-Schahn
,
R.
and
Hildebrandt
,
H.
(
2014
)
Sialic acids in the brain: gangliosides and polysialic acid in nervous system development, stability, disease, and regeneration
.
Physiol. Rev.
94
,
461
518
11
Wang
,
B.
(
2009
)
Sialic acid is an essential nutrient for brain development and cognition
.
Annu. Rev. Nutr.
29
,
177
222
12
van Karnebeek
,
C.D.
,
Bonafé
,
L.
,
Wen
,
X.-Y.
,
Tarailo-Graovac
,
M.
,
Balzano
,
S.
,
Royer-Bertrand
,
B.
et al. 
(
2016
)
NANS-mediated synthesis of sialic acid is required for brain and skeletal development
.
Nat. Genet.
48
,
777
784
13
Finne
,
J.
,
Bitter-Suermann
,
D.
,
Goridis
,
C.
and
Finne
,
U.
(
1987
)
An IgG monoclonal antibody to group B meningococci cross-reacts with developmentally regulated polysialic acid units of glycoproteins in neural and extraneural tissues
.
J. Immunol.
138
,
4402
4407
PMID:
[PubMed]
14
Shajahan
,
A.
(
2014
)
Application of Metabolic Glycan Engineering (MGE) to Neurobiology Using Carbohydrate-Neuroactive (CH-NA) Hybrid Molecules
,
Jawaharlal Nehru University
15
Drake
,
P.M.
,
Stock
,
C.M.
,
Nathan
,
J.K.
,
Gip
,
P.
,
Golden
,
K.P.
,
Weinhold
,
B.
et al. 
(
2009
)
Polysialic acid governs T-cell development by regulating progenitor access to the thymus
.
Proc. Natl Acad. Sci. U.S.A.
106
,
11995
12000
16
Fukuda
,
M.
(
1996
)
Possible roles of tumor-associated carbohydrate antigens
.
Cancer Res.
56
,
2237
2244
PMID:
[PubMed]
17
Angata
,
K.
and
Fukuda
,
M.
(
2003
)
Polysialyltransferases: major players in polysialic acid synthesis on the neural cell adhesion molecule
.
Biochimie
85
,
195
206
18
Galuska
,
C.E.
,
Lutteke
,
T.
and
Galuska
,
S.P.
(
2017
)
Is polysialylated NCAM not only a regulator during brain development but also during the formation of other organs?
Biology
6
,
1
17
PMID:
[PubMed]
19
Miyata
,
S.
,
Yamakawa
,
N.
,
Toriyama
,
M.
,
Sato
,
C.
and
Kitajima
,
K.
(
2011
)
Co-expression of two distinct polysialic acids, α2,8- and α2,9-linked polymers of N-acetylneuraminic acid, in distinct glycoproteins and glycolipids in sea urchin sperm
.
Glycobiology
21
,
1596
1605
20
Rutishauser
,
U.
(
2008
)
Polysialic acid in the plasticity of the developing and adult vertebrate nervous system
.
Nat. Rev. Neurosci.
9
,
26
35
21
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
22
Murray
,
H.C.
,
Low
,
V.F.
,
Swanson
,
M.E.
,
Dieriks
,
B.V.
,
Turner
,
C.
,
Faull
,
R.L.
et al. 
(
2016
)
Distribution of PSA-NCAM in normal, Alzheimer's and Parkinson's disease human brain
.
Neuroscience
330
,
359
375
23
Murray
,
H.C.
,
Swanson
,
M.E.V.
,
Dieriks
,
B.V.
,
Turner
,
C.
,
Faull
,
R.L.M.
and
Curtis
,
M.A.
(
2018
)
Neurochemical characterization of PSA-NCAM(+) cells in the human brain and phenotypic quantification in Alzheimer's disease entorhinal cortex
.
Neuroscience
372
,
289
303
24
Wainwright
,
S.R.
,
Barha
,
C.K.
,
Hamson
,
D.K.
,
Epp
,
J.R.
,
Chow
,
C.
,
Lieblich
,
S.E.
et al. 
(
2016
)
Enzymatic depletion of the polysialic acid moiety associated with the neural cell adhesion molecule inhibits antidepressant efficacy
.
Neuropsychopharmacology
41
,
1670
1680
25
Schulz
,
E.C.
,
Schwarzer
,
D.
,
Frank
,
M.
,
Stummeyer
,
K.
,
Mühlenhoff
,
M.
,
Dickmanns
,
A.
et al. 
(
2010
)
Structural basis for the recognition and cleavage of polysialic acid by the bacteriophage K1F tailspike protein EndoNF
.
J. Mol. Biol.
397
,
341
351
26
Hildebrandt
,
H.
and
Dityatev
,
A.
(
2015
)
Polysialic acid in brain development and synaptic plasticity
.
Top. Curr. Chem.
366
,
55
96
27
Rutishauser
,
U.
and
Landmesser
,
L.
(
1991
)
Polysialic acid on the surface of axons regulates patterns of normal and activity-dependent innervation
.
Trends Neurosci.
14
,
528
532
28
Rutishauser
,
U.
(
1992
)
NCAM and its polysialic acid moiety: a mechanism for pull/push regulation of cell interactions during development?
Dev. Suppl.
116
,
99
104
PMID:
[PubMed]
29
Kröcher
,
T.
,
Malinovskaja
,
K.
,
Jürgenson
,
M.
,
Aonurm-Helm
,
A.
,
Zharkovskaya
,
T.
,
Kalda
,
A.
et al. 
(
2015
)
Schizophrenia-like phenotype of polysialyltransferase ST8SIA2-deficient mice
.
Brain Struct. Funct.
220
,
71
83
30
Faure
,
C.
,
Chalazonitis
,
A.
,
Rhéaume
,
C.
,
Bouchard
,
G.
,
Sampathkumar
,
S.-G.
,
Yarema
,
K.J.
et al. 
(
2007
)
Gangliogenesis in the enteric nervous system: roles of the polysialylation of the neural cell adhesion molecule and its regulation by bone morphogenetic protein-4
.
Dev. Dyn.
236
,
44
59
31
Prescher
,
J.A.
and
Bertozzi
,
C.R.
(
2006
)
Chemical technologies for probing glycans
.
Cell
126
,
851
854
32
Sletten
,
E.M.
and
Bertozzi
,
C.R.
(
2009
)
Bioorthogonal chemistry: fishing for selectivity in a sea of functionality
.
Angew. Chem. Int. Ed. Engl.
48
,
6974
6998
33
Campbell
,
C.T.
,
Sampathkumar
,
S.-G.
and
Yarema
,
K.J.
(
2007
)
Metabolic oligosaccharide engineering: perspectives, applications, and future directions
.
Mol. Biosyst.
3
,
187
194
34
Du
,
J.
,
Meledeo
,
M.A.
,
Wang
,
Z.
,
Khanna
,
H.S.
,
Paruchuri
,
V.D.
and
Yarema
,
K.J.
(
2009
)
Metabolic glycoengineering: sialic acid and beyond
.
Glycobiology
19
,
1382
1401
35
Collins
,
B.E.
,
Fralich
,
T.J.
,
Itonori
,
S.
,
Ichikawa
,
Y.
and
Schnaar
,
R.L.
(
2000
)
Conversion of cellular sialic acid expression from N-acetyl- to N-glycolylneuraminic acid using a synthetic precursor, N-glycolylmannosamine pentaacetate: inhibition of myelin-associated glycoprotein binding to neural cells
.
Glycobiology
10
,
11
20
36
Choi
,
J.Y.
,
Park
,
M.
,
Cho
,
H.
,
Kim
,
M.-H.
,
Kang
,
K.
and
Choi
,
I.S.
(
2017
)
Neuro-compatible metabolic glycan labeling of primary hippocampal neurons in noncontact, sandwich-type neuron-astrocyte coculture
.
ACS Chem. Neurosci.
8
,
2607
2612
37
Cheng
,
B.
,
Xie
,
R.
,
Dong
,
L.
and
Chen
,
X.
(
2016
)
Metabolic remodeling of cell-surface sialic acids: principles, applications, and recent advances
.
ChemBioChem
17
,
11
27
38
Kim
,
E.J.
,
Sampathkumar
,
S.-G.
,
Jones
,
M.B.
,
Rhee
,
J.K.
,
Baskaran
,
G.
,
Goon
,
S.
et al. 
(
2004
)
Characterization of the metabolic flux and apoptotic effects of O-hydroxyl- and N-acyl-modified N-acetylmannosamine analogs in jurkat cells
.
J. Biol. Chem.
279
,
18342
18352
39
Charter
,
N.W.
,
Mahal
,
L.K.
,
Koshland
, Jr,
D.E.
and
Bertozzi
,
C.R.
(
2000
)
Biosynthetic incorporation of unnatural sialic acids into polysialic acid on neural cells
.
Glycobiology
10
,
1049
1056
40
Mahal
,
L.K.
,
Charter
,
N.W.
,
Angata
,
K.
,
Fukuda
,
M.
,
Koshland
, Jr,
D.E.
and
Bertozzi
,
C.R.
(
2001
)
A small-molecule modulator of poly-alpha 2,8-sialic acid expression on cultured neurons and tumor cells
.
Science
294
,
380
381
41
Liu
,
T.
,
Guo
,
Z.
,
Yang
,
Q.
,
Sad
,
S.
and
Jennings
,
H.J.
(
2000
)
Biochemical engineering of surface alpha 2–8 polysialic acid for immunotargeting tumor cells
.
J. Biol. Chem.
275
,
32832
32836
42
Pon
,
R.A.
,
Zou
,
W.
and
Jennings
,
H.J.
(
2011
)
Polysialic acid bioengineering of cancer and neuronal cells by N-acyl sialic acid precursor treatment
.
Adv. Exp. Med. Biol.
705
,
679
688
43
Gnanapragassam
,
V.S.
,
Bork
,
K.
,
Galuska
,
C.E.
,
Galuska
,
S.P.
,
Glanz
,
D.
,
Nagasundaram
,
M.
et al. 
(
2014
)
Sialic acid metabolic engineering: a potential strategy for the neuroblastoma therapy
.
PLoS ONE
9
,
e105403
44
Shajahan
,
A.
,
Parashar
,
S.
,
Goswami
,
S.
,
Ahmed
,
S.M.
,
Nagarajan
,
P.
and
Sampathkumar
,
S.-G.
(
2017
)
Carbohydrate-neuroactive hybrid strategy for metabolic glycan engineering of the central nervous system in vivo
.
J. Am. Chem. Soc.
139
,
693
700
45
Charter
,
N.W.
,
Mahal
,
L.K.
,
Koshland
, Jr,
D.E.
and
Bertozzi
,
C.R.
(
2002
)
Differential effects of unnatural sialic acids on the polysialylation of the neural cell adhesion molecule and neuronal behavior
.
J. Biol. Chem.
277
,
9255
9261
46
Horstkorte
,
R.
,
Mühlenhoff
,
M.
,
Reutter
,
W.
,
Nöhring
,
S.
,
Zimmermann-Kordmann
,
M.
and
Gerardy-Schahn
,
R.
(
2004
)
Selective inhibition of polysialyltransferase ST8SiaII by unnatural sialic acids
.
Exp. Cell Res.
298
,
268
274
47
Bork
,
K.
,
Reutter
,
W.
,
Gerardy-Schahn
,
R.
and
Horstkorte
,
R.
(
2005
)
The intracellular concentration of sialic acid regulates the polysialylation of the neural cell adhesion molecule
.
FEBS Lett.
579
,
5079
5083
48
Bork
,
K.
,
Gagiannis
,
D.
,
Orthmann
,
A.
,
Weidemann
,
W.
,
Kontou
,
M.
,
Reutter
,
W.
et al. 
(
2007
)
Experimental approaches to interfere with the polysialylation of the neural cell adhesion molecule in vitro and in vivo
.
J. Neurochem.
103
,
65
71
49
Sato
,
C.
,
Hane
,
M.
and
Kitajima
,
K.
(
2016
)
Relationship between ST8SIA2, polysialic acid and its binding molecules, and psychiatric disorders
.
Biochim. Biophys. Acta
1860
,
1739
1752
50
Dwyer
,
C.A.
and
Esko
,
J.D.
(
2016
)
Glycan susceptibility factors in autism spectrum disorders
.
Mol. Aspects Med.
51
,
104
114
51
Kreuzmann
,
D.
,
Horstkorte
,
R.
,
Kohla
,
G.
,
Kannicht
,
C.
,
Bennmann
,
D.
,
Thate
,
A.
et al. 
(
2017
)
Increased polysialylation of the neural cell adhesion molecule in a transgenic mouse model of sialuria
.
ChemBioChem
18
,
1188
1193
52
Marini
,
M.
,
Ambrosini
,
S.
,
Sarchielli
,
E.
,
Thyrion
,
G.D.
,
Bonaccini
,
L.
,
Vannelli
,
G.B.
et al. 
(
2014
)
Expression of sialic acids in human adult skeletal muscle tissue
.
Acta Histochem.
116
,
926
935
53
Weber
,
M.
,
Modemann
,
S.
,
Schipper
,
P.
,
Trauer
,
H.
,
Franke
,
H.
,
Illes
,
P.
et al. 
(
2006
)
Increased polysialic acid neural cell adhesion molecule expression in human hippocampus of heroin addicts
.
Neuroscience
138
,
1215
1223
54
Pan
,
H.-C.
,
Shen
,
Y.-Q.
,
Loers
,
G.
,
Jakovcevski
,
I.
and
Schachner
,
M.
(
2014
)
Tegaserod, a small compound mimetic of polysialic acid, promotes functional recovery after spinal cord injury in mice
.
Neuroscience
277
,
356
366
55
Miyazaki
,
T.
,
Angata
,
K.
,
Seeberger
,
P.H.
,
Hindsgaul
,
O.
and
Fukuda
,
M.
(
2007
)
CMP substitutions preferentially inhibit polysialic acid synthesis
.
Glycobiology
18
,
187
194
56
Ehrit
,
J.
,
Keys
,
T.G.
,
Sutherland
,
M.
,
Wolf
,
S.
,
Meier
,
C.
,
Falconer
,
R.A.
et al. 
(
2017
)
Exploring and exploiting acceptor preferences of the human polysialyltransferases as a basis for an inhibitor screen
.
ChemBioChem
18
,
1332
1337
57
Kayser
,
H.
,
Zeitler
,
R.
,
Kannicht
,
C.
,
Grunow
,
D.
,
Nuck
,
R.
and
Reutter
,
W.
(
1992
)
Biosynthesis of a nonphysiological sialic acid in different rat organs, using N-propanoyl-d-hexosamines as precursors
.
J. Biol. Chem.
267
,
16934
16938
PMID:
[PubMed]
58
Prescher
,
J.A.
,
Dube
,
D.H.
and
Bertozzi
,
C.R.
(
2004
)
Chemical remodelling of cell surfaces in living animals
.
Nature
430
,
873
877
59
Dube
,
D.H.
,
Prescher
,
J.A.
,
Quang
,
C.N.
and
Bertozzi
,
C.R.
(
2006
)
Probing mucin-type O-linked glycosylation in living animals
.
Proc. Natl Acad. Sci. U.S.A.
103
,
4819
4824
60
Gagiannis
,
D.
,
Gossrau
,
R.
,
Reutter
,
W.
,
Zimmermann-Kordmann
,
M.
and
Horstkorte
,
R.
(
2007
)
Engineering the sialic acid in organs of mice using N-propanoylmannosamine
.
Biochim. Biophys. Acta
1770
,
297
306
61
Neves
,
A.A.
,
Stöckmann
,
H.
,
Harmston
,
R.R.
,
Pryor
,
H.J.
,
Alam
,
I.S.
,
Ireland-Zecchini
,
H.
et al. 
(
2011
)
Imaging sialylated tumor cell glycans in vivo
.
FASEB J.
25
,
2528
2537
62
Xie
,
R.
,
Dong
,
L.
,
Huang
,
R.
,
Hong
,
S.
,
Lei
,
R.
and
Chen
,
X.
(
2014
)
Targeted imaging and proteomic analysis of tumor-associated glycans in living animals
.
Angew. Chem. Int. Ed. Engl.
53
,
14082
14086
63
Xie
,
R.
,
Dong
,
L.
,
Du
,
Y.
,
Zhu
,
Y.
,
Hua
,
R.
,
Zhang
,
C.
et al. 
(
2016
)
In vivo metabolic labeling of sialoglycans in the mouse brain by using a liposome-assisted bioorthogonal reporter strategy
.
Proc. Natl Acad. Sci. U.S.A.
113
,
5173
5178
64
Sampathkumar
,
S.-G.
,
Li
,
A.
and
Yarema
,
K.J.
(
2006
)
Sialic acid and the central nervous system: perspectives on biological functions, detection, imaging methods and manipulation
.
CNS Neurol. Disord. Drug Targets
5
,
425
440
65
Pavan
,
B.
,
Dalpiaz
,
A.
,
Ciliberti
,
N.
,
Biondi
,
C.
,
Manfredini
,
S.
and
Vertuani
,
S.
(
2008
)
Progress in drug delivery to the central nervous system by the prodrug approach
.
Molecules
13
,
1035
1065
66
Tsuji
,
A.
(
2005
)
Small molecular drug transfer across the blood-brain barrier via carrier-mediated transport systems
.
NeuroRx
2
,
54
62
67
Pardridge
,
W.M.
(
2012
)
Drug transport across the blood-brain barrier
.
J. Cereb. Blood Flow Metab.
32
,
1959
1972
68
Lindsley
,
C.W.
(
2014
)
Statistics for global prescription medications: CNS therapeutics maintain a leading position among small molecule therapeutics
.
ACS Chem. Neurosci.
5
,
250
251
69
Insel
,
T.R.
,
Landis
,
S.C.
and
Collins
,
F.S.
(
2013
)
The NIH BRAIN initiative
.
Science
340
,
687
688
70
D'Angelo
,
E.
(
2012
)
The human brain project
.
Funct. Neurol.
27
,
205
PMID:
[PubMed]