Human-Derived Natural Antibodies as Potential Therapeutics
Human-Derived Natural Antibodies as Potential Therapeutics
How B cells are activated to secrete antibodies that recognize self-antigens of the nervous system remains unknown. The identification of human natural IgMs that recognize surface molecules of neural cells indicates that there exists a subset of B lymphocytes selected in response to self-antigens from the nervous system. As revealed by the specific self-antigens, the selection process may develop at the early embryonic stage and extend into adulthood.
A repertoire of human IgM antibodies was identified based on a screening protocol that selects human IgMs binding to fresh brain slices. By screening a serum bank originally established by Robert Kyle at Mayo Clinic, investigators discovered a group of natural human IgMs that bind to neural surface antigens. Two serum-derived human monoclonal IgMs (sHIgM) were isolated and tested in animal models mimicking distinctive neurological diseases. Serum-derived human IgM 22 (sHIgM22) that binds to mature oligodendrocytes promotes remyelination in the Theiler's Murine Encephalomyelitis Virus (TMEV) model of multiple sclerosis (MS); sHIgM12 binding to neurons supports neurite extension in vitro, and protects neurons when administrated in vivo either in the TMEV model of MS or in models of amyotrophic lateral schlerosis induced by expressing the genetic mutants of human superoxide dismutase 1 (SOD1) (Rodriguez laboratory unpublished observation). Recombinant forms of sHIgM22 and sHIgM12 (rHIgM22 and rHIgM12) manufactured in GMP quality have confirmed efficacy in promoting remyelination in animal models of MS (rHIgM22); supporting axonal outgrowth in neuronal cultures (rHIgM12); and improving neurological functions in mouse models of MS (rHIgM22 or rHIgM12) and amyotrophic lateral schlerosis (rHIgM12). A Phase I clinical trial of rHIgM22 in patients with MS is ongoing. NAbs binding to neural cells apparently function through a lipid-raft mechanism. Indeed, accumulating evidence has demonstrated that the IgM type of NAbs recognize lipid components of neural membranes.
Myelin components can trigger autoimmunity against the nervous system. Experimental autoimmune encephalomyelitis (EAE) can be induced by immunizing susceptible animal strains with myelin molecules such as myelin oligodendrocyte glycoprotein or myelin proteolipid protein. On the other hand, earlier studies in the Rodriguez laboratory have demonstrated that immune responses raised against self-antigens of CNS can induce repair of the demyelinated lesions. For example, immunizing mice with spinal-cord homogenates and incomplete Freund's adjuvant or adoptive transfer of the anti-sera from the immunized animals promoted remyelination in the recipient animals. Further studies to address the autoimmunity to antigens of the CNS have provided clues to solve the contradictions. A major milestone event was the demonstration of antigen-specific T cells in EAE pathogenesis. The finding provided strong evidence that the CNS auto-reactive T cells in the normal immune system form the basis of the autoimmune reaction. T-lymphocytes specific for myelin antigens start an inflammatory reaction in CNS, which ultimately leads to demyelination and subsequent axonal injury. These T cells need to be re-activated by their antigens in the context of major histocompatibility complex class II-bearing antigen-presenting cells (APCs) in order to recognize their target. This re-stimulation of T cells may occur in vessel-associated macrophages or dendritic cells within the brain. However, these APCs do not produce myelin-associated self-antigens per se. The myelin molecules have to be released (by injury to the myelin) and transferred to the APCs (an inflammation that facilitates antigen presentation). Thus the myelin-derived self-antigen is basically the bio-marker of myelin itself, which ensures myelin-specific immune activation.
However, the triggering factors leading to autoimmunity remain obscure. Both genetic and environmental factors contribute to autoimmunne diseases. Factors triggering the disease may not be identical to the elements that influence the severity or progression of the disorder. Human monozygotic twin studies and transgenic models demonstrate that genetic influences strongly determine whether one will develop autoimmunity. It is regarded that environmental factors function on a specific genetic background to disrupt the immune tolerance and trigger autoimmune reactivity. Genetic changes in the human genome which explain the locus of autoimmunity are scare. An increasing number of environmental components may be involved in autoimmunity in a world where new technologies are developed in agriculture, industry and pharmocology to generate chemicals used in foods, drugs and diets. The toxic factors released have had widespread impact on human health. Researches in clinical and experimental biology have shown that autoimmunity may be induced by chronic exposure to various chemicals. Organic compounds such as toluene diisocyanate, trimellitic anhydride, phthalic anhydride, bezoquinone, formaldehyde, ethylene oxide, dinitrochlorobenzene, picryl chloride, penicillins and D-penicillinamine have been shown to induce autoimmune responses. The reactive organic compounds can bind covalently to protein nucleophilic groups such as thiol, amino, and hydroxyl groups and mediate functional changes. Sensitizing metal ions can oxidize proteins or lipids to form stable metal chelate complexes. The dietary components also play a role in autoimmunity. For example, there is a link between gluten ingestion and gluten sensitive enteropathies. Low levels of vitamin D are related to MS and other autoimmune disorders. Stress has also been linked to autoimmunity. The toxic molecules may induce aberrant cell death making the hidden molecules exposed to and processed by antigen presenting cells thus forming neoantigens by converting self- into non-self molecules. In addition to the environmental factors, microbes including viruses, bacteria and eukaryotic parasites are also involved in host immune reaction. Millions of years of coevolution has forged pervasive host–microbiota interconnections, which are particularly apparent in immune responses to microbiota invasion. The abundant microbiota residence on the body surfaces like the digestive, respiratory and urogenital tracts poses immense health challenges. Opportunistic infection can induce inflammation and sepsis. The immune system has thus evolved adaptations to contain the microbiota and foster complex microbial communities for their metabolic benefits and avoid triggering autoimmunity.
Therefore, we propose that exposure to toxins from the environment, infectious agents and/or idiopathic cellular injuries in genetically predisposed individuals activates antigen-specific T cells that cross the blood–brain barrier (BBB). Reactivation of the T cells by CNS-resident APCs, taking the example of MS in which the myelin antigens are presented, recruits innate immune cells leading to further damage. During the process, immune cells infiltrate the CNS. The original triggers that initiate disease may be resolved and therefore are difficult to connect to the disease.
Furthermore, the T-cell response is necessary but insufficient to initiate an 'MS' pathology in the EAE model. The formation of large demyelinating lesions depends on the generation of myelin-specific antibodies by B lymphocytes. In MS patients, the identity of the antigens targeted by myelin-reactive antibodies has not been identified. There are only three known antigens, galactosyl ceramide, sulfatide and myelin oligodendrocyte glycoprotein, which can initiate a demyelinating AutoAb response in EAE, and the complete antigenic profile on the myelin surface is yet to be determined. The demyelination in response to the AutoAb binding is mediated by complement activation and subsequent inflammatory response that further stimulates and recruits effector cells into the lesions. The AutoAbs in normal healthy animals are usually non-pathogenic. More intriguingly, we have shown that O4, a mouse IgM that targets sulfatide, induces myelin repair in the TMEV model of MS, an effect similar to the human natural IgM, HIgM22. Therefore, the presence of CNS-reactive antibodies is generally harmless but, in certain situations, may induce demyelination or in contrast remyelination.
Although failure to sense auto-reactive B cells through central and peripheral tolerance mechanisms exists, the exchange of molecules and cells between CNS and other organs is highly restricted due to multiple barrier mechanisms including the BBB. Because of the relatively 'isolated' status, cells in CNS can express a large number of antigens or the neural-specific isoforms and/or novel post-translational modification(s) of a more widely expressed antigen can be induced in response to changes in the milieu. Only when a neural-specific immune memory develops and the BBB integrity is compromised can autoimmune-mediated neural pathology ensue.
Surface antigens of oligodendrocyte are markers that indicate their developmental status. For example, oligodendrocyte marker O4 is an antigen expressed on the surface of oligodendrocytes. The mouse monoclonal IgM antibody O4 binds to the O4 epitope carried by the myelin glycolipid, sulfatide, which is the earlier marker specific for oligodendrocyte lineage. The anti-A2ref-5 and -O1 antibodies are frequently used in combination with anti-O4 to define distinct stages of oligodendrocyte development. Generally, A2ref-5/O4 markers are expressed on the earlier stage oligodendrocytes; O4/O1 cells are intermediate; and O1+ defines the more mature stage oligodendrocytes. The human natural IgMs corresponding to the mouse A2ref-5, O1 and O4 have not been identified. We have shown that HIgM22 recognizes the surface antigen(s) on mature oligodendrocytes. Studies in cerebroside sulfotransferase (CST) knockout mice suggest that HIgM22 binds to a sulfated molecule(s), synthesized by CST enzyme activity, which catalyzes the 3′-sulfation of galactose residues in several glycolipids. The major product of CST in the mammalian nervous system is sulfatide, an essential component of myelin.
The fact that HIgM22 recognizes surface antigen(s) of mature oligodendrocytes/myelin suggests that the human immune system can react to the antigenic stimuli (a sulfated molecule) of the mature myelin. How HIgM22 is generated remains obscure. However, it is clear that the antigen(s) of IgM22 can be presented to activate natural IgM production without inducing autoimmunity. As previously mentioned, HIgM22 can mediate signaling to promote oligodendrocyte precursor cell proliferation in cultured mixed glia and accelerate myelin repair in vivo. This finding has provided evidence that, in addition to regulating the immune system, natural IgMs signal the neural cells directly and/or indirectly to mediate a novel CNS function distinctive to the traditional view of natural IgMs in maintaining homeostasis. This concept is further confirmed by the identification of HIgM12, another human natural IgM that recognizes antigens on neurons.
Cell-surface antigens are frequently glycosylated proteins or lipids or complex glycans. In contrast to the antigens of HIgM22 on myelin, the surface antigens to which HIgM12 binds are developmentally conserved molecules that can be traced from neural stem cells to mature neurons (Figure 1). During differentiation, HIgM12-positive neural stem cells tend to develop into the neuronal lineage (Figure 2). In primary cultured neurons, HIgM12 binding to neurons is sensitive to neuraminidase treatment, suggesting that the antigens recognized by HIgM12 are sialic acid-modified molecule(s). New evidence indicates that HIgM12 binds to antigens of both glycoprotein and glycolipid that are decorated with sialic acids. The glycolipid components are actually a group of glycosphingolipids, called complex gangliosides (Figure 3). Therefore, it is likely that the sialic acids participate to form the epitope(s) for HIgM12 binding. Based on their binding specificity, natural human IgMs and pathogens such as influenza virus may bind to similar cell surface glycans. Therefore, it is likely that the neural surface glycans recognized by HIgM12 can potentially affect the host immune system.
(Enlarge Image)
Figure 1.
HIgM12 recognizes antigen(s) on both neural stem cells and mature neurons. The neural stem cells were isolated and propagated from the cortex of E15 mice. The neurospheres in culture were harvested by quick spin, washed and stored in phosphate-buffered saline, pH 7.4 on ice. Both the neurospheres and the dissociated neural stem cells from the sphere were plated on poly-D-lysine/laminin coated coverslips and cultured in the CO2 incubator at 37°C in the presence of EGF and bFGF overnight. The adult hippocampal neurons were isolated from the brains of 1–2-month-old mice by Percoll gradient centrifugation and cultured in neurobasal media containing 2% B27 supplement. To stain surface molecules, live cells were kept on ice-bath and incubated with PBS containing 1 μg/ml of HIgM12. Cells were fixed with 4% paraformaldehyde in phosphate-buffered saline, permeablized with 0.1% Triton X-100, and stained with the corresponding antibodies for intracellular molecules. HIgM12 binds to (A) neurospheres, (B) dissociated neural stem cells and (C) adult neurons. (A1–C1) HIgM12, green; (A2 & B2) nestin, red; (C2) β3-tubulin, red; (A3–C3) merged picture of (A1–C1). The stem cells were also double stained with antibodies to nestin and glial fibrillary acidic protein and were positive to both markers (not shown here) [89]. Scale bar: 100 μm in (A & B); 200 μm in (C).
(Enlarge Image)
Figure 2.
HIgM12 binds to the neuronal lineage antigen(s). Neural stem cells were isolated and propagated from the cortex of one to two month old mice [90]. Neurospheres and neural stem cells were cultured and maintained as described in Figure 1. Dissociated neural stem cells were plated on PDL-LM -coated coverslips and differentiated in neurobasal media that contained retinoic acid and 2% B-27. The cells were collected at multiple time points and stained with HIgM12 ([A1–C1] green), anti-MAP2, a neuronal marker ([A2–C2] blue), and anti-GFAP, the astrocyte marker ([A3–C3] red). (A) Neural stem cells differentiated for 24 h. HIgM12 binds to the surface of cells that express high levels of GFAP, but low MAP2. (B) Neural stem cells differentiated for 48 h. Note that some cells that were mostly spindle in shape began to become polarized and presented with a round cell body bearing two processes. (C) At 1 week after plating in the differentiation media, many polarized cells were generated that were similar, but different to the ones observed in (B). Those cells expressed high levels of both HIgM12 epitopes and MAP2, but almost no expression of GFAP, which were designated as 'HIgM12MAP2GFAP cells'. It is not known whether those 'HIgM12MAP2GFAP cells' were differentiated from the neural stem cells originally plated or were actually propagated from a different repertoire of neural stem cells formed in the GFAP-positive cell niche, as the GFAP-positive cells carry a morphology different from that in (A) or (B). The MAP2GFAP phenotype indicated that the newly formed 'HIgM12MAP2GFAP cells' tended to develop into the neuronal lineage. Those 'HIgM12MAP2GFAP cells' stayed close with each other and had similar size, morphology and markers, suggesting that they are from the same ancestor(s). However, the possibility that the cells migrated to the current position in response to chemokine(s) cannot be ruled out. Scale bar: 50 μm.
(Enlarge Image)
Figure 3.
HIgM12 recognizes the epitope(s) on both proteins and lipids. (A) The hippocampal neurons isolated from E15 mice were cultured for 3 days in vitro (DIV3). For labeling neuronal surface sialic acids, the live neurons were first washed with PBS, pH 7.4 at 4°C and then treated with 2 mM of sodium periodate (NaIO4) for 15 min at 4°C, which created C7 aldehydes on the side chain of sialic acid. Finally, aminooxy-biotin with a final concentration of 250 μM was added to the neuronal cultures and incubated at room temperature for 2 h [91]. Neurons were rinsed with PBS and lyzed in RIPA buffer. The lysates from both control and treated neurons were separated on SDS-PAGE and transferred onto nitrocellulose membranes. Membranes were then cut into several pieces, with one incubated in 1% NaIO4, and another in 2 M acetic acids. Proteins separated were probed with HIgM12, anti-β3-tubulin or avidin, respectively. HIgM12 binds to the protein band(s) at about 250 kilodalton, whereas treatment with 1% NaIO4 or 2 M acetic acid, which cleaved either the total carbohydrate or sialic acids moieties, prevented HIgM12 binding. Probing with anti-β3-tubulin indicated that the treatment only destroyed the carbohydrate moieties, and the amino acid chains were not affected. The avidin blot confirmed that HIgM12 likely binds to the sialic acid epitope(s) that were cleaved by 2 M acetic acids. (B) Gangliosides GM1, GD1a and GT1b from bovine brain were separated on high performance thin layer chromatography plate and immune-blotted with HIgM12, which mainly recognized GD1a and trace of GT1b [92].
Ctrl: Control.
The sialo-glycan moieties contribute important functional roles to the sialic-acid-bearing molecules. The sialic-acid epitopes are evolutionarily conserved cell surface structures containing a five-carbon core, an exocyclic glycerol side chain (C-7, C-8, C-9) and an N-acetyl or glycolyl moiety. Variations at C-5 determine the core sialic-acid structures; O-substitutions at C-4, C-7, C-8 and/or C-9 (including acetyl, methyl, lactyl, sulfate and phosphate groups) generate further diversity to sialic-acid structures, and the anionic C-1 carboxylate can form lactones and lactams that neutralize the charges. Some Gram-negative bacteria associated with human and/or animal hosts carry sialic acids and can cause serious illness. Neuroinvasive bacteria such as Escherichia coli k1 and Neisseria meningitides sero-group B express α2–8 and/or α2–9-linked polysialic acid. The size and charge of polysialic acid could facilitate bacterial dissemination and/or hindering host immune responses. The pathogenic bacteria, campylobacter jejuni, express sialo-glycans that attach to their lipooligosaccharides and are implicated to induce autoimmune neuropathies. The influenza C virus hemagglutinin binds specifically to the 9-O-acetyl bearing sialic acid; and the budding viruses use a Sialo-specific O-acetylesterase to release from the host cells.
When administrated in vivo, HIgM12 can mediate neuronal functions by directly targeting neurons/axons, which is distinct from the roles through regulating the immune system. The cell-surface sialic acids function by modifying the structures of oligosaccharides, larger glycans and multiglycan complexes to affect the protein or lipid cores carrying the sugar chains. At the cell membrane, glycolipids and glycoproteins organize into microdomains called lipid rafts. Sphingolipids, including gangliosides, have long, saturated carbohydrate chains on their ceramide moieties that associate with cholesterol and/or with selected membrane proteins to create the dynamic microdomains on cell surfaces. The sialic-acid-decorated molecules extend into the extracellular milieu and function as the antenna to interact with other cells and/or extracellular matrices. Accordingly, HIgM12 was shown to associate with lipid raft and participates in dynamic regulations of the raft microdomains. An immobilized layer of HIgM12 supports axonal differentiation and growth and can override the myelin-mediated neurite outgrowth inhibition, whereas the free HIgM12 molecules applied to cultured neurons regulate the cytoskeletal signaling cascade.
It remains to be determined whether HIgM12 induces the same or similar signaling cascade when administrated in vivo. Gangliosides function as receptors in axon–myelin interactions. Myelin-associated glycoprotein (MAG) is a transmembrane protein expressed by myelinating cells: oligodendrocytes in the CNS and Schwann cells in the PNS. MAG distributes on the innermost wrap of myelin, which directly contacts the axonal surface. MAG binds preferentially to the 'NeuAc(α2–3)-Gal(β1–3)-GalNAc' sugar chain, which is the shared terminal sequence of gangliosides GD1a and GT1b. The phenotypes of complex ganglioside-null mice support the roles for GD1a and GT1b in the function of MAG. Mice lacking complex gangliosides (ref-4galnt1-null) display the similar phenotypes as MAG-null mice including myelin anomalies, progressive axonal degeneration in the central and peripheral nervous systems, reduced neurofilament spacing, reduced axon diameters of myelinated axons and disrupted nodes of Ranvier. Behaviorally, both MAG-null and ref-4galnt1-null mice showed disruptions in hind limb reflexes and coordination as well as similar patterns of whole-body tremor and hyperactivity. The double-null mice all have the defects observed in each single-null strain (see review). Given that HIgM12 beneficially affects distinctive models of neurological diseases, it turns out that HIgM12 likely functions by inducing signal cascades through gangliosides and mediate glia–neuron interactions. Indeed, the signaling cross-talk(s) between HIgM12 and MAG were observed in cultured neurons when extending their neurites (unpublished observations). This supports a role for complex gangliosides in mediating protective signaling initiated either by MAG or HIgM12. The fact that HIgM12 also recognizes sialic acid-modified protein(s) may provide a redundant function and adds further complexity to HIgM12-induced signaling.
Neural Reactive Natural IgMs & the Nature of Neural Surface Antigens
How B cells are activated to secrete antibodies that recognize self-antigens of the nervous system remains unknown. The identification of human natural IgMs that recognize surface molecules of neural cells indicates that there exists a subset of B lymphocytes selected in response to self-antigens from the nervous system. As revealed by the specific self-antigens, the selection process may develop at the early embryonic stage and extend into adulthood.
Identification of Human Natural IgMs: The Potential Therapeutics to Treat Human Diseases
A repertoire of human IgM antibodies was identified based on a screening protocol that selects human IgMs binding to fresh brain slices. By screening a serum bank originally established by Robert Kyle at Mayo Clinic, investigators discovered a group of natural human IgMs that bind to neural surface antigens. Two serum-derived human monoclonal IgMs (sHIgM) were isolated and tested in animal models mimicking distinctive neurological diseases. Serum-derived human IgM 22 (sHIgM22) that binds to mature oligodendrocytes promotes remyelination in the Theiler's Murine Encephalomyelitis Virus (TMEV) model of multiple sclerosis (MS); sHIgM12 binding to neurons supports neurite extension in vitro, and protects neurons when administrated in vivo either in the TMEV model of MS or in models of amyotrophic lateral schlerosis induced by expressing the genetic mutants of human superoxide dismutase 1 (SOD1) (Rodriguez laboratory unpublished observation). Recombinant forms of sHIgM22 and sHIgM12 (rHIgM22 and rHIgM12) manufactured in GMP quality have confirmed efficacy in promoting remyelination in animal models of MS (rHIgM22); supporting axonal outgrowth in neuronal cultures (rHIgM12); and improving neurological functions in mouse models of MS (rHIgM22 or rHIgM12) and amyotrophic lateral schlerosis (rHIgM12). A Phase I clinical trial of rHIgM22 in patients with MS is ongoing. NAbs binding to neural cells apparently function through a lipid-raft mechanism. Indeed, accumulating evidence has demonstrated that the IgM type of NAbs recognize lipid components of neural membranes.
Myelin-derived Antigens & Myelin-reactive Antibodies
Myelin components can trigger autoimmunity against the nervous system. Experimental autoimmune encephalomyelitis (EAE) can be induced by immunizing susceptible animal strains with myelin molecules such as myelin oligodendrocyte glycoprotein or myelin proteolipid protein. On the other hand, earlier studies in the Rodriguez laboratory have demonstrated that immune responses raised against self-antigens of CNS can induce repair of the demyelinated lesions. For example, immunizing mice with spinal-cord homogenates and incomplete Freund's adjuvant or adoptive transfer of the anti-sera from the immunized animals promoted remyelination in the recipient animals. Further studies to address the autoimmunity to antigens of the CNS have provided clues to solve the contradictions. A major milestone event was the demonstration of antigen-specific T cells in EAE pathogenesis. The finding provided strong evidence that the CNS auto-reactive T cells in the normal immune system form the basis of the autoimmune reaction. T-lymphocytes specific for myelin antigens start an inflammatory reaction in CNS, which ultimately leads to demyelination and subsequent axonal injury. These T cells need to be re-activated by their antigens in the context of major histocompatibility complex class II-bearing antigen-presenting cells (APCs) in order to recognize their target. This re-stimulation of T cells may occur in vessel-associated macrophages or dendritic cells within the brain. However, these APCs do not produce myelin-associated self-antigens per se. The myelin molecules have to be released (by injury to the myelin) and transferred to the APCs (an inflammation that facilitates antigen presentation). Thus the myelin-derived self-antigen is basically the bio-marker of myelin itself, which ensures myelin-specific immune activation.
However, the triggering factors leading to autoimmunity remain obscure. Both genetic and environmental factors contribute to autoimmunne diseases. Factors triggering the disease may not be identical to the elements that influence the severity or progression of the disorder. Human monozygotic twin studies and transgenic models demonstrate that genetic influences strongly determine whether one will develop autoimmunity. It is regarded that environmental factors function on a specific genetic background to disrupt the immune tolerance and trigger autoimmune reactivity. Genetic changes in the human genome which explain the locus of autoimmunity are scare. An increasing number of environmental components may be involved in autoimmunity in a world where new technologies are developed in agriculture, industry and pharmocology to generate chemicals used in foods, drugs and diets. The toxic factors released have had widespread impact on human health. Researches in clinical and experimental biology have shown that autoimmunity may be induced by chronic exposure to various chemicals. Organic compounds such as toluene diisocyanate, trimellitic anhydride, phthalic anhydride, bezoquinone, formaldehyde, ethylene oxide, dinitrochlorobenzene, picryl chloride, penicillins and D-penicillinamine have been shown to induce autoimmune responses. The reactive organic compounds can bind covalently to protein nucleophilic groups such as thiol, amino, and hydroxyl groups and mediate functional changes. Sensitizing metal ions can oxidize proteins or lipids to form stable metal chelate complexes. The dietary components also play a role in autoimmunity. For example, there is a link between gluten ingestion and gluten sensitive enteropathies. Low levels of vitamin D are related to MS and other autoimmune disorders. Stress has also been linked to autoimmunity. The toxic molecules may induce aberrant cell death making the hidden molecules exposed to and processed by antigen presenting cells thus forming neoantigens by converting self- into non-self molecules. In addition to the environmental factors, microbes including viruses, bacteria and eukaryotic parasites are also involved in host immune reaction. Millions of years of coevolution has forged pervasive host–microbiota interconnections, which are particularly apparent in immune responses to microbiota invasion. The abundant microbiota residence on the body surfaces like the digestive, respiratory and urogenital tracts poses immense health challenges. Opportunistic infection can induce inflammation and sepsis. The immune system has thus evolved adaptations to contain the microbiota and foster complex microbial communities for their metabolic benefits and avoid triggering autoimmunity.
Therefore, we propose that exposure to toxins from the environment, infectious agents and/or idiopathic cellular injuries in genetically predisposed individuals activates antigen-specific T cells that cross the blood–brain barrier (BBB). Reactivation of the T cells by CNS-resident APCs, taking the example of MS in which the myelin antigens are presented, recruits innate immune cells leading to further damage. During the process, immune cells infiltrate the CNS. The original triggers that initiate disease may be resolved and therefore are difficult to connect to the disease.
Furthermore, the T-cell response is necessary but insufficient to initiate an 'MS' pathology in the EAE model. The formation of large demyelinating lesions depends on the generation of myelin-specific antibodies by B lymphocytes. In MS patients, the identity of the antigens targeted by myelin-reactive antibodies has not been identified. There are only three known antigens, galactosyl ceramide, sulfatide and myelin oligodendrocyte glycoprotein, which can initiate a demyelinating AutoAb response in EAE, and the complete antigenic profile on the myelin surface is yet to be determined. The demyelination in response to the AutoAb binding is mediated by complement activation and subsequent inflammatory response that further stimulates and recruits effector cells into the lesions. The AutoAbs in normal healthy animals are usually non-pathogenic. More intriguingly, we have shown that O4, a mouse IgM that targets sulfatide, induces myelin repair in the TMEV model of MS, an effect similar to the human natural IgM, HIgM22. Therefore, the presence of CNS-reactive antibodies is generally harmless but, in certain situations, may induce demyelination or in contrast remyelination.
Although failure to sense auto-reactive B cells through central and peripheral tolerance mechanisms exists, the exchange of molecules and cells between CNS and other organs is highly restricted due to multiple barrier mechanisms including the BBB. Because of the relatively 'isolated' status, cells in CNS can express a large number of antigens or the neural-specific isoforms and/or novel post-translational modification(s) of a more widely expressed antigen can be induced in response to changes in the milieu. Only when a neural-specific immune memory develops and the BBB integrity is compromised can autoimmune-mediated neural pathology ensue.
Myelin Surface Antigens Recognized by HIgM22
Surface antigens of oligodendrocyte are markers that indicate their developmental status. For example, oligodendrocyte marker O4 is an antigen expressed on the surface of oligodendrocytes. The mouse monoclonal IgM antibody O4 binds to the O4 epitope carried by the myelin glycolipid, sulfatide, which is the earlier marker specific for oligodendrocyte lineage. The anti-A2ref-5 and -O1 antibodies are frequently used in combination with anti-O4 to define distinct stages of oligodendrocyte development. Generally, A2ref-5/O4 markers are expressed on the earlier stage oligodendrocytes; O4/O1 cells are intermediate; and O1+ defines the more mature stage oligodendrocytes. The human natural IgMs corresponding to the mouse A2ref-5, O1 and O4 have not been identified. We have shown that HIgM22 recognizes the surface antigen(s) on mature oligodendrocytes. Studies in cerebroside sulfotransferase (CST) knockout mice suggest that HIgM22 binds to a sulfated molecule(s), synthesized by CST enzyme activity, which catalyzes the 3′-sulfation of galactose residues in several glycolipids. The major product of CST in the mammalian nervous system is sulfatide, an essential component of myelin.
The fact that HIgM22 recognizes surface antigen(s) of mature oligodendrocytes/myelin suggests that the human immune system can react to the antigenic stimuli (a sulfated molecule) of the mature myelin. How HIgM22 is generated remains obscure. However, it is clear that the antigen(s) of IgM22 can be presented to activate natural IgM production without inducing autoimmunity. As previously mentioned, HIgM22 can mediate signaling to promote oligodendrocyte precursor cell proliferation in cultured mixed glia and accelerate myelin repair in vivo. This finding has provided evidence that, in addition to regulating the immune system, natural IgMs signal the neural cells directly and/or indirectly to mediate a novel CNS function distinctive to the traditional view of natural IgMs in maintaining homeostasis. This concept is further confirmed by the identification of HIgM12, another human natural IgM that recognizes antigens on neurons.
Neuron Surface Antigens Recognized by HIgM12
Cell-surface antigens are frequently glycosylated proteins or lipids or complex glycans. In contrast to the antigens of HIgM22 on myelin, the surface antigens to which HIgM12 binds are developmentally conserved molecules that can be traced from neural stem cells to mature neurons (Figure 1). During differentiation, HIgM12-positive neural stem cells tend to develop into the neuronal lineage (Figure 2). In primary cultured neurons, HIgM12 binding to neurons is sensitive to neuraminidase treatment, suggesting that the antigens recognized by HIgM12 are sialic acid-modified molecule(s). New evidence indicates that HIgM12 binds to antigens of both glycoprotein and glycolipid that are decorated with sialic acids. The glycolipid components are actually a group of glycosphingolipids, called complex gangliosides (Figure 3). Therefore, it is likely that the sialic acids participate to form the epitope(s) for HIgM12 binding. Based on their binding specificity, natural human IgMs and pathogens such as influenza virus may bind to similar cell surface glycans. Therefore, it is likely that the neural surface glycans recognized by HIgM12 can potentially affect the host immune system.
(Enlarge Image)
Figure 1.
HIgM12 recognizes antigen(s) on both neural stem cells and mature neurons. The neural stem cells were isolated and propagated from the cortex of E15 mice. The neurospheres in culture were harvested by quick spin, washed and stored in phosphate-buffered saline, pH 7.4 on ice. Both the neurospheres and the dissociated neural stem cells from the sphere were plated on poly-D-lysine/laminin coated coverslips and cultured in the CO2 incubator at 37°C in the presence of EGF and bFGF overnight. The adult hippocampal neurons were isolated from the brains of 1–2-month-old mice by Percoll gradient centrifugation and cultured in neurobasal media containing 2% B27 supplement. To stain surface molecules, live cells were kept on ice-bath and incubated with PBS containing 1 μg/ml of HIgM12. Cells were fixed with 4% paraformaldehyde in phosphate-buffered saline, permeablized with 0.1% Triton X-100, and stained with the corresponding antibodies for intracellular molecules. HIgM12 binds to (A) neurospheres, (B) dissociated neural stem cells and (C) adult neurons. (A1–C1) HIgM12, green; (A2 & B2) nestin, red; (C2) β3-tubulin, red; (A3–C3) merged picture of (A1–C1). The stem cells were also double stained with antibodies to nestin and glial fibrillary acidic protein and were positive to both markers (not shown here) [89]. Scale bar: 100 μm in (A & B); 200 μm in (C).
(Enlarge Image)
Figure 2.
HIgM12 binds to the neuronal lineage antigen(s). Neural stem cells were isolated and propagated from the cortex of one to two month old mice [90]. Neurospheres and neural stem cells were cultured and maintained as described in Figure 1. Dissociated neural stem cells were plated on PDL-LM -coated coverslips and differentiated in neurobasal media that contained retinoic acid and 2% B-27. The cells were collected at multiple time points and stained with HIgM12 ([A1–C1] green), anti-MAP2, a neuronal marker ([A2–C2] blue), and anti-GFAP, the astrocyte marker ([A3–C3] red). (A) Neural stem cells differentiated for 24 h. HIgM12 binds to the surface of cells that express high levels of GFAP, but low MAP2. (B) Neural stem cells differentiated for 48 h. Note that some cells that were mostly spindle in shape began to become polarized and presented with a round cell body bearing two processes. (C) At 1 week after plating in the differentiation media, many polarized cells were generated that were similar, but different to the ones observed in (B). Those cells expressed high levels of both HIgM12 epitopes and MAP2, but almost no expression of GFAP, which were designated as 'HIgM12MAP2GFAP cells'. It is not known whether those 'HIgM12MAP2GFAP cells' were differentiated from the neural stem cells originally plated or were actually propagated from a different repertoire of neural stem cells formed in the GFAP-positive cell niche, as the GFAP-positive cells carry a morphology different from that in (A) or (B). The MAP2GFAP phenotype indicated that the newly formed 'HIgM12MAP2GFAP cells' tended to develop into the neuronal lineage. Those 'HIgM12MAP2GFAP cells' stayed close with each other and had similar size, morphology and markers, suggesting that they are from the same ancestor(s). However, the possibility that the cells migrated to the current position in response to chemokine(s) cannot be ruled out. Scale bar: 50 μm.
(Enlarge Image)
Figure 3.
HIgM12 recognizes the epitope(s) on both proteins and lipids. (A) The hippocampal neurons isolated from E15 mice were cultured for 3 days in vitro (DIV3). For labeling neuronal surface sialic acids, the live neurons were first washed with PBS, pH 7.4 at 4°C and then treated with 2 mM of sodium periodate (NaIO4) for 15 min at 4°C, which created C7 aldehydes on the side chain of sialic acid. Finally, aminooxy-biotin with a final concentration of 250 μM was added to the neuronal cultures and incubated at room temperature for 2 h [91]. Neurons were rinsed with PBS and lyzed in RIPA buffer. The lysates from both control and treated neurons were separated on SDS-PAGE and transferred onto nitrocellulose membranes. Membranes were then cut into several pieces, with one incubated in 1% NaIO4, and another in 2 M acetic acids. Proteins separated were probed with HIgM12, anti-β3-tubulin or avidin, respectively. HIgM12 binds to the protein band(s) at about 250 kilodalton, whereas treatment with 1% NaIO4 or 2 M acetic acid, which cleaved either the total carbohydrate or sialic acids moieties, prevented HIgM12 binding. Probing with anti-β3-tubulin indicated that the treatment only destroyed the carbohydrate moieties, and the amino acid chains were not affected. The avidin blot confirmed that HIgM12 likely binds to the sialic acid epitope(s) that were cleaved by 2 M acetic acids. (B) Gangliosides GM1, GD1a and GT1b from bovine brain were separated on high performance thin layer chromatography plate and immune-blotted with HIgM12, which mainly recognized GD1a and trace of GT1b [92].
Ctrl: Control.
The sialo-glycan moieties contribute important functional roles to the sialic-acid-bearing molecules. The sialic-acid epitopes are evolutionarily conserved cell surface structures containing a five-carbon core, an exocyclic glycerol side chain (C-7, C-8, C-9) and an N-acetyl or glycolyl moiety. Variations at C-5 determine the core sialic-acid structures; O-substitutions at C-4, C-7, C-8 and/or C-9 (including acetyl, methyl, lactyl, sulfate and phosphate groups) generate further diversity to sialic-acid structures, and the anionic C-1 carboxylate can form lactones and lactams that neutralize the charges. Some Gram-negative bacteria associated with human and/or animal hosts carry sialic acids and can cause serious illness. Neuroinvasive bacteria such as Escherichia coli k1 and Neisseria meningitides sero-group B express α2–8 and/or α2–9-linked polysialic acid. The size and charge of polysialic acid could facilitate bacterial dissemination and/or hindering host immune responses. The pathogenic bacteria, campylobacter jejuni, express sialo-glycans that attach to their lipooligosaccharides and are implicated to induce autoimmune neuropathies. The influenza C virus hemagglutinin binds specifically to the 9-O-acetyl bearing sialic acid; and the budding viruses use a Sialo-specific O-acetylesterase to release from the host cells.
HIgM12 Regulates Neuronal Functions
When administrated in vivo, HIgM12 can mediate neuronal functions by directly targeting neurons/axons, which is distinct from the roles through regulating the immune system. The cell-surface sialic acids function by modifying the structures of oligosaccharides, larger glycans and multiglycan complexes to affect the protein or lipid cores carrying the sugar chains. At the cell membrane, glycolipids and glycoproteins organize into microdomains called lipid rafts. Sphingolipids, including gangliosides, have long, saturated carbohydrate chains on their ceramide moieties that associate with cholesterol and/or with selected membrane proteins to create the dynamic microdomains on cell surfaces. The sialic-acid-decorated molecules extend into the extracellular milieu and function as the antenna to interact with other cells and/or extracellular matrices. Accordingly, HIgM12 was shown to associate with lipid raft and participates in dynamic regulations of the raft microdomains. An immobilized layer of HIgM12 supports axonal differentiation and growth and can override the myelin-mediated neurite outgrowth inhibition, whereas the free HIgM12 molecules applied to cultured neurons regulate the cytoskeletal signaling cascade.
It remains to be determined whether HIgM12 induces the same or similar signaling cascade when administrated in vivo. Gangliosides function as receptors in axon–myelin interactions. Myelin-associated glycoprotein (MAG) is a transmembrane protein expressed by myelinating cells: oligodendrocytes in the CNS and Schwann cells in the PNS. MAG distributes on the innermost wrap of myelin, which directly contacts the axonal surface. MAG binds preferentially to the 'NeuAc(α2–3)-Gal(β1–3)-GalNAc' sugar chain, which is the shared terminal sequence of gangliosides GD1a and GT1b. The phenotypes of complex ganglioside-null mice support the roles for GD1a and GT1b in the function of MAG. Mice lacking complex gangliosides (ref-4galnt1-null) display the similar phenotypes as MAG-null mice including myelin anomalies, progressive axonal degeneration in the central and peripheral nervous systems, reduced neurofilament spacing, reduced axon diameters of myelinated axons and disrupted nodes of Ranvier. Behaviorally, both MAG-null and ref-4galnt1-null mice showed disruptions in hind limb reflexes and coordination as well as similar patterns of whole-body tremor and hyperactivity. The double-null mice all have the defects observed in each single-null strain (see review). Given that HIgM12 beneficially affects distinctive models of neurological diseases, it turns out that HIgM12 likely functions by inducing signal cascades through gangliosides and mediate glia–neuron interactions. Indeed, the signaling cross-talk(s) between HIgM12 and MAG were observed in cultured neurons when extending their neurites (unpublished observations). This supports a role for complex gangliosides in mediating protective signaling initiated either by MAG or HIgM12. The fact that HIgM12 also recognizes sialic acid-modified protein(s) may provide a redundant function and adds further complexity to HIgM12-induced signaling.
Source...