Role of Selectins in Reperfusion Injury

A variety of common cardiovascular disorders (e.g., stroke, myocardial infarction, and hypovolemic shock) are characterized by a period of decreased or absent blood flow followed by a state of reperfusion, which occurs either by natural mechanisms or because blood flow has been restored by medical intervention. Despite rescue of the tissue from permanent anoxemic necrosis, the entry of leukocytes into the reperfused area can initiate a cascade of events that ultimately results in tissue damage (called reperfusion injury). P-Selectin on the activated endothelium in the reperfused area and/or L-selectin on leukocytes have vital roles in mediating the initial steps of this cascade (see Chapter 31). Substantial data in animal model systems indicate that blockade of this initial selectin-based recognition can ameliorate the subsequent tissue damage. A major goal of some pharmaceutical and biotechnology companies has been to make small-molecule inhibitors that can be used to achieve this blockade in human patients (see Chapters 49 and 50 regarding the synthesis of small glycan molecules designed to be selectin inhibitors). Interestingly, some forms of heparin currently used to effect anticoagulation under some of these conditions also have the ability to block P- and L-selectin at clinically tolerable doses.

Roles of Selectins, Glycosaminoglycans, and Sialic Acids in Atherosclerosis

High levels of low-density lipoprotein (LDL) cholesterol and decreased high-density lipoprotein (HDL) cholesterol are associated with an increased risk of atherosclerotic lesions in the large arteries, which are the major cause of heart attacks, strokes, and other serious diseases. The very earliest phase of the development of atherosclerotic lesions (the fatty streak) involves the entry of monocytes into the subendothelial regions of the blood vessels. There is evidence that this process involves the expression of P- and/or E-selectin on the endothelium, which recognizes P-selectin glycoprotein ligand-1 (PSGL-1) or sialyl-Lewis^x on circulating monocytes. Indeed, atherosclerotic lesions in atherosclerosis-prone mice showed delayed progression in a P-selectin-deficient background, and even slower progression occurs in a combined P- and E-selectin-deficient state. The induction of endothelial P-selectin expression may result from oxidized lipids that are present in LDL particles and/or the inflammatory process that occurs in the early atheromatous plaque. It remains to be seen whether it is feasible to intervene in this process, because early lesions probably develop very slowly and relatively early in life. The subsequent subendothelial retention of LDLs in the early plaque is thought to occur at least partly via their interactions with proteoglycans. The interaction is thought to cause irreversible structural alterations of LDL, potentiating oxidation and uptake by macrophages and smooth-muscle cells. At the molecular level, clusters of basic amino acids present in apolipoprotein B (the protein moiety of LDL) appear to bind the negatively charged glycosaminoglycans of proteoglycans. Meanwhile, heparan sulfate (HS) found in the liver may regulate the turnover of lipoprotein particles. Several reports also indicate a lowered overall sialylation of LDL in patients with coronary artery disease. The pathophysiological significance of this finding and the mechanism(s) involved remain unclear. One hypothesis is that the desialylated LDL is more prone to be taken up and incorporated into atheromatous plaques.

Role of Selectins in Inflammatory Skin Diseases

Several inflammatory skin diseases (e.g., atopic dermatitis and contact dermatitis) are characterized by the entry of leukocytes into the dermis, where they have a pathogenic role in recruiting other types of cells and in mediating tissue damage. These types of skin lesions are sometimes associated with the chronic persistent expression of E-selectin on the endothelial cells. Independent evidence indicates that E-selectin can recruit circulating lymphocytes carrying the cutaneous lymphocyte antigen (detected by the antibody HECA452), which appears to be a specific E-selectin ligand epitope carried on a subset of PSGL-1 molecules (see Chapter 31). There is also evidence that some T-helper-1 (Th1) lymphocytes can be recruited into the skin by virtue of their expression of the PSGL-1 ligand for P-selectin. Many of these observations have been made only in experimental models. The potential for therapeutic intervention in these selectin-mediated processes has not been fully pursued.

Pathogenesis and Complications of Diabetes Mellitus

Diabetes mellitus is a disease of dysregulated glucose metabolism, resulting from relative or absolute lack of insulin action. It is accompanied by characteristic long-term vascular and neurologic complications. One mechanism appears related to the high levels of free glucose in body fluids, which cause acceleration of a well-known nonenzymatic process in which the open-chain (aldehyde) form of the glucose reacts randomly with lysine residues on various proteins, resulting in reversible Schiff bases. With time, some of these adducts undergo the irreversible Amadori rearrangement. These then undergo a series of “browning” (Maillard) reactions, which eventually progress to advanced glycation end products (AGE). The resulting protein cross-links can damage cellular functions, and such adducts can also be recognized by receptors, for example, the receptor for advanced glycation end products (RAGE) and the macrophage scavenger receptor, perhaps participating in the process of atherogenesis. A current view is that this is a normal process of aging, which is accelerated in the setting of the chronic persistent hyperglycemia of uncontrolled diabetes mellitus. It is important to differentiate mechanistically and semantically between this nonenzymatic glycation (or “glucosylation”) process and enzymatic glycosylation that takes place in the endoplasmic reticulum (ER), Golgi apparatus, and cytoplasm, utilizing glycosyltransferases and nucleotide sugar donors.

Another metabolic change of particular interest in diabetes mellitus is the increased production of UDP-GlcNAc caused by the conversion of excess glucose via the glucosamine:fructose aminotransferase (GFAT pathway). A current hypothesis is that this increase in cytoplasmic UDP-GlcNAc gives a secondary increase of O-GlcNAc levels on nuclear and cytoplasmic glycoproteins, secondarily altering the phosphorylation of the same proteins and their functions (see Chapter 18). Specific molecular mechanisms involving such altered O-GlcNAcylation have been defined in animal models for complications such as diabetic cardiomyopathy (increased O-GlcNAcylation of various nuclear proteins) and erectile dysfunction (O-GlcNAcylation of endothelial nitric oxide synthase). Interestingly, several of the cytoplasmic proteins involved in insulin receptor signaling and resulting nuclear transcription changes are themselves O-GlcNAcylated and are functionally altered in diabetes.

Nephropathy is a diabetic complication associated with high mortality. It begins with low levels of urinary albumin excretion (microalbuminuria), which progresses to frank macroalbuminuria. Ultimately, nephrotic syndrome and a concomitant decrease in glomerular filtration rate progress to end-stage renal disease. The proteinuria has been correlated with a reduction in the HS proteoglycan content of the glomerular basement membrane. The underlying mechanism may involve a reduction in HS synthesis by glomerular epithelial cells that may, in turn, be caused by the high glucose in the environment. One theory is that the resulting decrease in anionic change and loss of HS proteoglycan are thought to affect the porosity of the glomerular basement membrane. However, recent genetic evidence in mice questions this hypothesis. Interestingly, high glucose also mediates increased plasminogen activator inhibitor-1 (PAI-1) gene expression in renal glomerular mesangial cells, via O-GlcNAc-mediated alterations in Sp1 transcriptional activity.

Role of Gut Epithelial Glycans in Gastrointestinal Infections

Many gastrointestinal pathogens interact with the gut mucosa via recognition of glycan structures (see Chapter 34). Prominent examples include cholera toxin (which binds GM1 ganglioside) and Helicobacter pylori, the causative agent of peptic ulcer disease and gastritis (which binds Lewis type glycans in the stomach mucosa). Consideration is now being given to using orally administered soluble glycan inhibitors to impede the attachment of such pathogens in the gut. In this regard, it is interesting that a time-honored treatment for peptic ulcer disease was a combination of antacids and milk (which contains large amounts of free sialyloligosaccharides). In addition, the variety and high concentrations of free glycans found in human milk (especially in the early days after birth of the baby) are thought to impede the ability of gut pathogens to bind to the mucosa and initiate infections.

Autosomal Dominant Polycystic Liver Disease

Autosomal dominant polycystic liver disease is thought to arise by somatic mutations in individuals who already have a mutated allele in one of two different genes. One is SEC63, which facilitates recognition of proteins by ER chaperones. The other is PRKCSH (also known as hepatocystin or the β subunit of α-glucosidase II), which was previously described as a protein kinase C substrate (80K-H) and is localized on the cell surface, in intracellular vesicles, and in the ER. The products of both these genes are involved in the translocation, folding, and quality control of newly synthesized proteins. The cysts in the liver usually develop later in life, and outgrowth is restricted to the biliary epithelium, suggesting a specific impact on a regulator of their proliferation rather than a gross effect on all glycoproteins. The specific proteins affected have not been identified, but recent studies of genetically altered mice suggest that HS proteoglycans play a role.

Heparan Sulfate Proteoglycans in the Pathogenesis of Protein-losing Enteropathy

Protein-losing enteropathy (PLE) is defined as the enteric loss of plasma proteins, which become life-threatening. The cellular and molecular mechanisms of this disease are not well understood, but it develops in some patients with congenital disorder of glycosylation type Ib (CDG-Ib) and CDG-Ic (see Chapter 42) or as a complication months to years following Fontan surgery to correct congenital heart malformations in patients with normal N-glycosylation. PLE appears to result from a collision of genetic insufficiencies and environmental stress. Impaired N-glycosylation, increased proinflammatory cytokines, and increased venous pressure all synergize to create PLE. In each of these pathologies, HS is specifically absent from the basolateral surface of intestinal epithelial cells during the episodes, and it returns when PLE subsides. The inflammatory cytokines tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ) both bind to HS, which may thus serve as a buffer. In vitro studies show that removal of HS from the basolateral surface increases cytokine signaling through their receptors, thereby loosening tight junctions that normally prevent protein leakage. Bacterial and viral infections increase these cytokines and often trigger PLE in patients. Some CDG-Ib and post-Fontan patients have increased venous pressure, which can down-regulate some of the genes involved in extracellular matrix biosynthesis. Pressure synergizes with the effects of cytokines and the localized loss of HS to create a downward spiral of disease. Traditional therapy for PLE includes treatment of the underlying conditions if possible, maintenance of nutritional state, and sometimes albumin infusions and steroid hormones or other anti-inflammatory drugs. The mother of a post-Fontan patient made the astute observation that her son’s PLE disappeared when he was given heparin injections as an anticoagulant prior to surgery. Later, in vitro studies showed that a few micrograms per milliliters of heparin or HS completely reverse the synergistic effects of epithelial cell HS loss and cytokine-induced breakdown of tight junctions. Nonanticoagulant heparin therapy may hold new promise for a variety of PLE patients.

Changes in Sialic Acid O-Acetylation in Ulcerative Colitis

Ulcerative colitis is an inflammatory disease typically affecting the superficial epithelial layer of the rectum and the distal colon. Although the primary cause of the disease is unknown, a large body of evidence suggests both genetic and environmental factors, and remissions and exacerbations are common. The sialic acids of the colonic mucosa, which are normally heavily O-acetylated, lose this modification in ulcerative colitis. Whether or not this is of pathogenic significance is uncertain, but these modifications normally do render the sialic acids more resistant to bacterial sialidases that are found in the gut. There have been conflicting claims about the efficacy of heparin treatments in improving the symptoms of this disease.

Clinical Use of Heparin as an Anticoagulant

Preparations of heparin (a highly sulfated form of HS; see Chapter 16) are routinely purified from animal tissues (particularly porcine intestines) and are used as a fast-acting and potent anticoagulant in a wide variety of diseases that involve thrombosis, in medical procedures such as dialysis, and in surgical procedures such as open heart surgery. As described in Chapter 16, the mechanism of anticoagulation is precise, involving a specifically sulfated heparin pentasaccharide that interacts with circulating antithrombin and markedly enhances its ability to inactivate coagulation factors Xa and IIa (thrombin). The use of “unfractionated heparin” is being partially supplanted by various forms of low-molecular-weight heparins that seem to be easier to use and are associated with fewer complications. One explanation is that the unfractionated heparin effects on factor IIa require a long chain that interacts both with the antithrombin and with the IIa itself, in a tripartite complex. In contrast, the shorter chains found in low-molecular-weight heparins only facilitate antithrombin inactivation of factor Xa. Thus, low-molecular-weight heparins affect Xa but not IIa levels. Most recently, a synthetic pentasaccharide that binds and facilitates antithrombin inactivation of factor Xa has been introduced as an alternative to heparin. Although these improvements have been valuable, it must be kept in mind that the original unfractionated heparin has a variety of other biological effects besides anticoagulation. Thus, other beneficial effects of heparin, such as the blockade of P- and L-selectin, are being reduced or even eliminated during the course of the switch to the low-molecular-weight heparins and the synthetic pentasaccharide.

A rare but feared complication of heparin treatment is heparin-induced thrombocytopenia. The pathogenesis appears to involve the formation of complexes between heparin and platelet factor-4 and the generation of antibodies against the complexes. These antibodies in turn then deposit on platelets, causing their aggregation and loss from circulation. Somewhat paradoxically, this process results in exaggerated thromboses, rather than bleeding. The incidence of this complication appears to be lower with the use of low-molecular-weight heparins, and it may be absent with the use of the pure pentasaccharide.

Hemolytic Transfusion Reactions

The invention of blood transfusion uncovered the existence of the ABO blood group system, which is dictated by different alleles of an α-Gal(NAc) transferase (for details, see Chapter 13). These and other less prominent glycan antigens are responsible for many of the hemolytic transfusion reactions that can occur when errors are made in blood typing. Active attempts have recently been under way to generate “universal donor” red blood cells, via enzymatic conversion of blood group A and B antigens to the O state.

Acquired Anticoagulation Due to Circulating Heparan Sulfate

Occasionally, patients with diseases such as cirrhosis and hepatocellular carcinoma spontaneously secrete a circulating anticoagulant and have an unusual coagulation test profile that makes it appear as if the patient has been treated with heparin. The anticoagulant activity can be purified from the plasma and has been identified as an HS glycosaminoglycan. The precise source of secretion has not been defined, and therapy is often difficult unless the underlying disease can be corrected or the liver transplanted.

Abnormal Glycosylation of Plasma Fibrinogen in Hepatoma and Liver Disorders

Plasma fibrinogen is heavily sialylated and the sialic acids are involved in binding calcium. Certain genetic disorders of fibrinogen are known to be associated with altered glycosylation of its N-glycans, which causes altered function in clotting. Patients with hepatomas and other liver disorders can also sometimes manifest increased branching and/or number of N-glycans, resulting in an overall increase in sialic acid content. This can present clinically as a bleeding disorder associated with a prolonged thrombin time. Patients with congenital genetic disorders affecting N-glycan biosynthesis (see Chapter 42) can also have thrombotic or bleeding disorders that may be partly explained by altered glycosylation of plasma proteins and/or platelets involved in blood coagulation.

Paroxysmal Nocturnal Hemoglobinuria

Paroxysmal nocturnal hemoglobinuria (PNH) is an unusual form of acquired hemolytic anemia (excessive destruction of red blood cells) that usually appears in adults. The defect arises through a somatic mutation in bone marrow stem cells that causes the production of one or more abnormal clones. The defect is an inactivation of the single active copy of the PIGA gene, an X-linked locus involved in the first stage of biosynthesis of glycosylphosphatidylinositol (GPI) anchors (for details on GPI anchor biosynthesis, see Chapter 11). Although several blood cell types show abnormalities, the red cell defect is the most prominent, being characterized by an abnormal susceptibility to the action of complement. This is now known to be due to the lack of expression of certain GPI-anchored proteins, such as decay-accelerating factor, that normally down-regulate complement activation on “self” surfaces. However, hypercoagulability also occurs, presumably due to loss of GPI-anchored proteins on other cells, such as monocytes. Interestingly, many of these patients later develop either bone marrow failure (aplastic anemia) or acute leukemia. It is now known that most normal humans already have a tiny fraction of circulating cells with the PNH defect. These presumably represent the products of one or more bone marrow stem cells that develop this acquired defect because of a single hit on the active X chromosome but then did not become prominent contributors to the total pool of circulating red blood cells. In this scenario, the independent occurrence of a process damaging other stem cells allows the “unmasking” of the PNH defect.

Paroxysmal Cold Hemoglobinuria

Patients with this rare disorder have a cold-induced intravascular destruction of red cells (hemolysis), which appears to be caused by a circulating IgG antibody directed against the red cell P blood group system. The pathogenesis of this disorder is unknown, but it tends to occur in the setting of some viral infections and in syphilis. The IgG antibody is demonstrated by the so-called “Donath–Landsteiner test,” where the patient’s serum is mixed either with the patient’s own red cells or with those from a normal person and chilled to 4°C. Hemolysis occurs after warming back to 37°C.

Cold Agglutinin Disease

This disease is caused by autoimmune IgM antibodies directed against glycan epitopes on erythrocytes. High titers of IgM agglutinins are present in serum and are maximally active at 4°C. This IgM is presumed to bind to erythrocytes that are circulating in the cooled blood of peripheral regions of the body. The antibody fixes complement, which then destroys the cells when they reach warmer areas of the body. There are several variants of the syndrome. One affects young adults and follows infection with Mycoplasma pneumoniae or Epstein-Barr virus (infectious mononucleosis). This antibody is typically directed against the so-called “i” antigen (poly-N-acetyllactosamine), is polyclonal, and is generally short-lived, disappearing when the infection subsides. Because M. pneumoniae is itself known to have a receptor that recognizes sialylated poly-N-acetyllactosamine, it is hypothesized that the autoimmune antibody results from a mirror-image, anti-idiotypic reaction to the initial antibody directed against the mycoplasma’s binding site. An idiopathic variant of cold agglutinin disease affects older individuals, involves a monoclonal IgM, and can be a precursor or an accompaniment to a lymphoproliferative disease such as Walden-ström’s macroglobulinemia, chronic lymphocytic leukemia, or other lymphomas. These antibodies are typically directed against the “I” antigen (β1-6-branched poly-N-acetyllactosamine) present on erythrocytes. Some less common variants of cold agglutinin disease involve antibodies directed against sialylated N-acetyllactosamines. In some patients on chronic hemodialysis, the syndrome occurs due to the formation of antibody directed against the sialylated blood group antigen N.

Tn Polyagglutinability Syndrome

Tn polyagglutinability syndrome is an acquired condition in which the blood cells made by the bone marrow express the Tn antigen (O-linked N-acetylgalactosamine, GalNAcα-O-Ser/Thr) and sialyl-Tn (Siaα2-6GalNAcα-O-Ser/Thr), thus becoming susceptible to hemagglutination by the naturally occurring anti-Tn antibodies present in most normal human sera. The defect tends to be incomplete, with some circulating cells expressing the more complete sialylated tri- and tetrasaccharide O-glycans as well. These observations are best explained as an acquired stem-cell-based loss of expression of the O-glycan core-1 β1-3 galactosyltransferase activity (also called the T synthase). This in turn has now been explained by the acquired inactivation of Cosmc, a chaperone required for the biosynthesis of the T synthase. As with PNH, the existence of the COSMC gene on the X chromosome allows a single hit on the active X chromosome to cause a glycosylation defect in a single bone marrow stem cell. Patients with this syndrome show a wide range of symptoms. Some are picked up simply because the polyagglutinability of their red blood cells is detected when blood typing is done for a possible transfusion. Others have varying degrees of hemolytic anemia and/or decreases in other blood cell types. Some of these patients can subsequently progress into frank leukemia. It is unclear how the primary syndrome predisposes to the development of the malignancy. As with PNH, the possibility exists that an underlying bone marrow disorder simply allows the “unmasking” of preexisting minor stem cell clones with the defect. In keeping with this, the leukemic clones that arise later need not necessarily have the same defect.

Changes in IgG Glycosylation in Rheumatoid Arthritis

The IgG class of circulating immunoglobulins carry N-glycans, and those in the constant (CH2 or Fc) region of human IgG are reported to have several unusual properties. First, the glycans are buried between the folds of the two constant regions. Second, they are often sufficiently immobilized by carbohydrate–protein interactions that can be seen in the crystal structure of the protein (most glycans are not visible in crystal structures). Third, although processed into biantennary complex type-N glycans, they are hardly ever completed into fully sialylated molecules. Instead, most of the molecules remain with one or two terminal β-linked galactose residues (so-called G1 and G2 molecules, respectively). It was previously noted that in patients with a chronic systemic disease called rheumatoid arthritis, a major fraction of the serum IgG molecules have decreased galactosylation of N-glycans, some carrying no galactose at all (so-called G0 molecules). The severity of the disease tends to correlate with the extent of the glycosylation change, and the spontaneous improvement that occurs during pregnancy is correlated with a restoration in galactosylation. One function attributed to the Fc N-glycans is to maintain the conformation of the Fc domains as well as the hinge regions. These structural features are necessary for effector functions such as complement binding and Fc-dependent cytotoxicity. Nuclear magnetic resonance (NMR) studies have shown that the G0 N-glycans have an increased mobility resulting from the loss of interactions between the glycan and the Fc protein surface. Thus, it is thought that regions of the protein surface that are normally covered by the glycan are exposed in rheumatoid arthritis. In addition, some studies suggest that the more mobile G0 N-glycan may be recognized by the circulating mannose-binding protein, which can activate complement directly. Rheumatoid arthritis is also characterized by circulating immune complexes consisting of antibody molecules (called rheumatoid factor) that seem to be directed against the Fc region of other IgG molecules. However, the epitopes involved here do not seem to be glycan-related. Another possibility being considered is that the altered glycosylation changes interactions with Fc receptors. With regard to the mechanism for the underglycosylation, some groups have reported lowered activities of β-galactosyltransferase activity in lymphocytes from patients with rheumatoid arthritis. It remains an open question whether the altered glycosylation of IgG has a primary pathogenic role in rheumatoid arthritis, because the appearance of G0 molecules is a general feature of other unrelated chronic granulomatous diseases, for example, Crohn’s disease and tuberculosis. Furthermore, the glycan change is also seen to a lesser extent in osteoarthritis, a form of chronic degenerative arthritis with a completely different pathogenesis. Overall, the change in IgG glycosylation in rheumatoid arthritis remains an interesting phenomenon whose precise significance and pathogenic role need further study.

Secondary Changes in the O-Glycans of CD43 in Wiskott–Aldrich Syndrome

This inherited genetic disease is characterized by skin eczema, altered cellular immune responses, and low platelet counts, symptoms that appear in childhood. Early studies suggested that the disease was associated with an absence of CD43 (also called leukosialin or sialophorin), the major O-glycosylated protein of lymphocytes. However, in retrospect, it is clear that this polypeptide is still expressed normally, but it has changed gel mobility because of markedly increased branching of O-glycans. Recent data indicate that the primary defect in this syndrome is not in glycosylation but in a transcription factor. However, the glycan changes seen in resting T cells of these patients are exactly the same as those that can be induced upon activation of normal T cells. Thus, it remains possible that some aspects of the immune disorders in this disease are due to a secondary change in glycosylation.

Recognition of Glycans by Bacterial Adhesins, Toxins, and Viral Hemagglutinins

A wide variety of pathogens initiate infection by specifically recognizing cell-surface glycans (see Chapter 34). In some instances, the differences in infection rates between individuals can be attributed to variations in the expression of the glycan target. For example, adhesion of certain pathogenic strains of Escherichia coli to cells in the urinary tract can be mediated by P fimbriae, involving a specific glycan receptor on the P blood group antigens. Infections do not occur in individuals who are P negative. P fimbriae also appear to be important in determining the propensity for bacterial bloodstream invasion from the kidney.

Desialylation of Blood Cells by Circulating Microbial Sialidases during Infections

Several microorganisms produce sialidases (classically called neuraminidases) that are involved in the pathogenesis of the diseases that they cause. In most instances, this enzyme remains localized to the site of infection. However, in some severe cases, for example, Clostridium perfringens-mediated gas gangrene, a sufficient amount of the sialidase is produced so that it can appear in the plasma. In this situation, the surface of circulating blood cells can become desialylated, resulting in enhanced clearance and anemia. The detection of the circulating sialidase has been proposed to have diagnostic and prognostic significance. Some cases of hemolytic-uremic syndrome are also associated with sialidase-producing Streptococcus pneumoniae infections. It is possible that blocking the sialidase with appropriate inhibitors could have therapeutic value in these situations.

Loss of Glomerular Sialic Acids in Nephrotic Syndrome

Nephrotic syndrome occurs when the kidney glomerulus fails to retain serum proteins during the initial filtration of plasma and these proteins then leak into the urine. The epithelial/endothelial mucin molecule called podocalyxin, which is present on the foot processes (pedicles) of glomerular podocytes, is thought to have a role in maintaining pore integrity and in excluding large molecules, such as proteins, from the glomerular filtrate. The sialic acid residues of podocalyxin molecules are believed to be critical in this process. Loss of glomerular sialic acid is seen in spontaneous minimal-change renal disease in children and in the nephrotic syndrome that follows some bacterial infections. Several animal models seem to mimic this situation. Proteinuria and renal failure develop in a dose-dependent manner after a single inoculation of Vibrio cholerae sialidase, and this correlates with loss of sialic acids from the glomerulus. This was also accompanied by the effacement of foot processes and the apparent formation of tight junctions between podocytes. The anionic charge returned to endothelial and epithelial sites within two days of sialidase inoculation, but the foot process loss remained. Another model is termed aminonucleoside nephrosis, and it is induced in rats by injection of puromycin. Again, defective sialylation of a podocalyxin and glomerular glycosphingolipids has been detected in this model.

Changes in the O-Glycans in IgA Nephropathy

In humans, only IgA1 and IgD contain O-glycans in the hinge regions, whereas all immunoglobulin classes contain N-glycans in the Fc domain. Aggregation of the IgA1 molecule is thought to be involved in a form of nephrotic syndrome called IgA nephropathy. Studies of the O-glycans on serum IgA1 showed glycan truncations in the IgA nephropathy group compared with a negative control group. One of the functions of the IgA1 O-glycan chains is thought to be to stabilize the three-dimensional structure of the molecule. Studies of heat-induced aggregation support the notion that the altered glycosylation on the hinge region of IgA1 results in a loss of conformational stiffness, perhaps explaining the aggregation phenomenon. Removal of glycans from the IgA1 molecule also results in non-covalent self-aggregation and a significant increase in adhesion to the extracellular matrix proteins. It is therefore suggested that the underglycosylation of the IgA1 molecule found in IgA nephropathy is involved in the nonimmunologic glomerular accumulation of IgA1. The primary mechanism of underglycosylation remains unknown. A likely scenario is a defect in the COSMC gene, similar to that found in the Tn polyagglutinability syndrome (see above). The difference is that instead of affecting a bone marrow stem cell, the defect would involve a clone of B cells that specifically expresses IgA.

Heparan Sulfate in Systemic Lupus Erythematosus

Systemic lupus erythematosus (SLE) is an autoimmune disorder characterized by deposition of antigen-antibody complexes in various organs, especially the skin and the kidney. The precise initiating mechanisms of SLE are unknown, but it appears that the resulting pathology may involve both cytokines and HS. Amounts of HS are reduced on the glomerular basement membrane, and this was thought to result from masking of HS by complexes of nucleosomes and antinuclear antibodies, but the actual situation is likely to be more complex. Even though anti-double-stranded DNA antibodies are the hallmark of SLE, circulating antibodies to HS strongly correlate with disease activity. In some studies, HS injections into dogs induce SLE symptoms within several weeks. Elevated HS is found in the urine of SLE patients, especially in severe cases. SLE patients also have systemic elevated TNF-α. Anticytokine therapy against systemic and locally expressed cytokines suggests that blocking the proinflammatory cascade may be of significant value. Glomeruli have locally increased TNF-α, which acts to induce interleukin-1β (IL-1β) and IL-6. IFN-γ is thought to induce TNF-α, precipitating glomerulonephritis due to local inflammation. Some SLE patients also develop PLE, perhaps as a consequence of misplaced or degraded HS and elevated cytokines, creating the appropriate environment for PLE (see above).

Pathogenic Autoimmune Antibodies Directed against Neuronal Glycans

A variety of diseases are associated with circulating antibodies directed against specific glycan molecules that are enriched in the nervous system. These patients suffer from symptoms related to autoimmune neural damage. Such antibodies can arise via at least three distinct pathogenic mechanisms. In the first situation, patients with benign or malignant B-cell neoplasms (e.g., benign monoclonal gammopathy of unknown significance [MGUS], Waldenstrom’s macroglobulinemia, and plasma cell myeloma) secrete monoclonal IgM or IgA antibodies that are highly specific for either ganglio-series gangliosides or, more commonly, for sulfated glucuronosyl glycans (the so-called HNK-1 epitope). These antibodies react with the glycolipid bearing this epitope 3-O-SO₃- GlcAβ1-4Galβ1-4GlcNAcβ1-3Galβ1-4Glc-Cer (3′sulfoglucuronosylparagloboside) and against the N-glycans on a variety of CNS glycoproteins (MAG, P0, L1, N-CAM) that bear the same terminal sequence (3-O-SO₃-GlcA β1-4Galβ1-4GlcNAcβ1-). The resulting peripheral demyelinating neuropathy can sometimes be more damaging than the primary disease itself. Therapy consists of attempts to treat the primary disease with chemotherapy or to remove the immunoglobulin by plasmapheresis. Both approaches are usually unsuccessful at lowering the immunoglobulin to a level sufficient to diminish the symptoms. The second situation is an immune reaction to the molecular mimicry of neural ganglioside structures by the lipooligosaccharides of bacteria such as Campylobacter jejuni. Following an intestinal infection with such organisms, circulating cross-reacting antibodies against gangliosides such as GM1 and GQ1b appear in the plasma. These are typically associated with the onset of symptoms of a demyelinating neuropathy involving the peripheral and central nervous systems (the Guillain–Barré and Miller–Fisher syndromes, respectively). The third situation is a human-induced disease arising from recent attempts to treat patients with disease such as stroke using intravenous injections of mixed bovine brain gangliosides. Although some evidence exists that this treatment may be beneficial for the primary disease, several cases of Guillain–Barré syndrome have been reported as a likely side effect. One suggested scenario is that the presence of small amounts of gangliosides with the nonhuman sialic acid N-glycolylneuraminic acid facilitates the formation of antibodies that cross-react with ganglioside containing human sialic acid N-acetylneuraminic acid.

Role of Glycans in the Histopathology of Alzheimer’s Disease

Alzheimer’s disease is a common primary degenerative dementia of humans, with an insidious onset and a progressive course. The ultimate diagnosis is made by postmortem histological examination of brain tissue, which shows characteristic amyloid plaques with neurofibrillary tangles that are associated with neuronal death. Two types of glycans have been implicated in the histopathogenesis of the lesions: O-GlcNAc and HS glycosaminoglycans. Paired helical filaments are the major component of the neurofibrillary tangle. These are primarily composed of the microtubule-associated protein Tau, which is present in a hyperphosphorylated state. This abnormally hyperphosphorylated Tau no longer binds microtubules and self-assembles to form the paired helical filaments that may contribute to neuronal death. Normal brain Tau is known to be multiply modified by Ser(Thr)-linked O-GlcNAc, the dynamic and abundant posttranslational modification that is often reciprocal to Ser(Thr) phosphorylation (see Chapter 18). The hypothesis currently being investigated is that site-specific or stoichiometric changes in O-GlcNAc addition may modulate Tau function and may also play a part in the formation of paired helical filaments by allowing excessive phosphorylation. The hyperphosphorylated Tau in Alzheimer’s disease brain is found in association with HS proteoglycans. Nonphosphorylated Tau isoforms with three microtubule-binding repeats form paired helical-like filaments under physiological conditions in vitro when incubated with HS. Heparin prevents Tau from binding to microtubules and promotes microtubule disassembly. These findings, together with previous evidence that heparin stimulates Tau phosphorylation by protein kinases, have been used to argue that sulfated glycosaminoglycans may be a critical factor in the formation of the neurofibrillary tangles. However, no significant difference was noted between the detailed structure of HS obtained from control brains and that from Alzheimer’s disease brains. Furthermore, the topological separation of Tau (which occurs in the cytoplasm) from glycosaminoglycans (which are extracellular) indicates that this physical association can only occur after cell death. On the other hand, HS proteoglycans may also have an important role in amyloid plaque deposition. Investigators have demonstrated high-affinity binding between HS proteoglycans and the amyloid precursor, as well as with the A4 peptide derived from the precursor. In addition, a specific vascular HS proteoglycan found in senile plaques bound with high affinity to two amyloid protein precursors. Overall, the data indicate that HS chains may have a significant role in the pathogenesis of the histological lesions.

Role of Selectins, Siglecs, and Mucins in Bronchial Asthma

Asthma is a disease that is characterized by a hyperresponsiveness of the tracheobronchial tree to various stimuli, resulting in widespread narrowing of the airways, and it changes in severity, either spontaneously or as a result of therapy. The two dominant pathological features of asthma are airway wall inflammation and luminal obstruction of the airways by inflammatory exudates, consisting predominantly of mucins. Most cases are due to the presence of antigen-specific IgE antibodies, which then bind to mast cells as well as to basophils and certain other cell types. Subsequently, antigen can cross-link adjacent IgE molecules, triggering an explosive release of vasoactive, bronchoactive, and chemotactic agents from mast cell granules into the extracellular milieu. Eosinophils also contribute to the pathogenesis of asthma in several ways, by synthesizing leukotrienes, stimulating histamine release from mast cells and basophils, providing a positive feedback loop, and releasing major basic protein, a granule-derived protein that has toxic effects on the respiratory epithelium. Underlying all this, it appears that CD4⁺ Th2 cells are responsible for orchestrating the responses of other cell types. Recent evidence indicates that the selectins are intimately involved in the recruitment of eosinophils and basophils (and possibly T lymphocytes) into the lung, raising the hope that small-molecule inhibitors of selectin function and/or heparin can be used to treat the early stages of an asthmatic attack. Likewise, chemokine interactions with HS are important in leukocyte trafficking. Recent evidence from Siglec-F knockout mice also suggests that the functionally equivalent human paralog Siglec-8 is a good target for reducing the contributions of eosinophils to the pathology (see Chapter 32). Finally, the large increase in mucus production is at least partly mediated by an up-regulation of synthesis of mucin polypeptides, under the influence of various cytokines that stimulate the goblet cells of the airway epithelium.

Role of Selectins in Acute Respiratory Distress Syndrome

Acute respiratory distress syndrome is a serious pathophysiological process that is the final common pathway of lung injury arising from a variety of events, such as shock, trauma, or sepsis. It is characterized by diffuse pulmonary endothelial injury, progressing to pulmonary edema, which results from a marked increase in capillary permeability. Selectins and integrins help circulating neutrophils to adhere to the endothelium and release injurious oxidants, proteolytic enzymes, and arachidonic acid metabolites, resulting in endothelial cell dysfunction and destruction. The presence of many neutrophils and secretory products in bronchoalveolar lavage liquid emphasizes the critical role of the underlying inflammatory response. Again, the hope is that small molecule selectin inhibitors and/or the right kind of heparin can be used in the early stages of this syndrome, before it progresses to extensive lung damage and respiratory failure.

Altered Glycosylation of Epithelial Glycoproteins in Cystic Fibrosis

Cystic fibrosis is a very common genetic disorder caused by a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR). This causes defective chloride conduction across the apical membrane of involved epithelial cells. Cystic fibrosis is associated with increased accumulation of viscous mucins in the pancreas, gut, and lungs, which leads to many of the symptoms of the disease. In the lung airways, there are known to be widespread increases in sialylation of secreted proteins and increases in the sulfation and fucosylation of mucus glycoproteins. One possible explanation is that the primary CFTR defect allows a higher Golgi pH, resulting in the abnormalities in glycosylation: however, there is currently some controversy about this conclusion. Curiously, the CFTR is mainly expressed within nonciliated epithelial cells, duct cells, and serous cells of the tubular glands, but it is not highly expressed in the goblet cells and mucous glands of the acinar cells, which are the cells that synthesize respiratory mucins. Thus, the CFTR mutation may indirectly affect mucin glycosylation through the generation of inflammatory responses. Another major cause of morbidity in the disease is the colonization of respiratory epithelium by an alginate-producing form of Pseudomonas aeroginosa. Certain glycolipids and mucin glycans have been suggested to be the Pseudomonas receptors that help to maintain the colonization. The changes in glycolipid and mucin glycosylation could enhance the production of potential binding targets for organ colonization. The presence of bacterial products is also a proinflammatory condition, since the bacterial capsular polysaccharides may activate Toll receptors and lead eventually to neutrophil accumulation and organ damage.