Diabetic neuropathies are a group of clinical syndromes that affect distinct regions of the nervous system, singly or combined, and markedly affect quality of life (1) and activities of daily living and increase morbidity and mortality (2). Therapy directed at the basic pathogenesis is sorely needed. Neurologic complications occur equally in type 1 and type 2 diabetes and additionally in various forms of acquired diabetes (3). Diabetic neuropathies may be diffuse somatic and involve proximal or distal nerves, occur as focal mononeuritides, and involve the autonomic nervous system (4,5).
The pathogenesis of diabetic neuropathies is multifactorial. Hyperglycemia causes nerve damage by inducing the activation of the polyol, protein kinase C, and hexosamine pathways and the accumulation of advanced glycation end products. Hyperglycemia also induces oxidative stress by enhancement of mitochondrial respiration, redox alteration, and uncoupling proteins, which leads to elevated superoxide anions (6,7). Oxidative stress depletes nitric oxide within the peripheral nerves and endothelium of the microvasculature by reducing endothelial nitric oxide synthase, altering nerve perfusion (8). In addition, there is deficiency of or a poor response to neurotrophic factors (9). However, there is now increasing evidence to suggest that autoimmunity has a role to play in the development and progression of diabetic neuropathies.
Fifty years ago, Waksman and Adams (10) suggested an autoimmune etiology of peripheral neuropathy when they injected rabbits with neuronal components to produce what they called “allergic neuritis. ” To be able to implicate autoimmunity as a causative factor of neuropathy, there would have to be a clear association between the antibody and the disease; neuropathy would have to be induced by introduction or development of antibodies, and there would have to be reversal of the disease with removal or neutralization of the antibodies. Peripheral nerves are normally protected against the immune system by tight capillary endothelial junctions and the perineurium. Nerves are also a rich source of glycoproteins, lipopolysachharides, and other lipoproteins that can potentially form active antigenic material. In the autoimmune onslaught against nerves, there is first damage to the protective sheath and then to the inner components. These can be brought on by viral or bacterial infections (e.g., polio, leprosy, Lyme’s disease), neoplasms, or connective tissue disorders, and often there is strong genetic predisposition, such as HLA DR-3 and -4, in type 1 diabetes. Table 1 illustrates the association of different types of antibodies with various neuropathy syndromes.
Neurons and pancreatic β-cells are neuroectodermal derivatives and therefore share common antigens, especially in the early stages of cellular evolution. Type 1 diabetes results from an autoimmune destruction of pancreatic β-cells. There also may be a direct destruction of neurons by the same autoimmune process in diabetes. The pancreatic islets of Langerhans are surrounded by a Schwann cell sheath. These cells form a tight cellular mantle that envelops the endocrine islet tissue. Components of the peri-islet Schwann cells include GAD (11). There is an early appearance of anti-GAD65–specific T-cells in type 1 diabetes. Anti-GAD65 antibody is a strong predictive marker for the onset of type 1 diabetes (12). Presence of this antibody in patients with recent-onset type 1 diabetes is associated with worse glycemic control and worse peripheral nerve function, suggesting a common mechanism for β-cell and neuronal damage (13). Patients with high GAD65 antibodies were shown to have positive correlation with motor nerve conduction velocities, F wave latencies, thermal threshold detection, and cardiovascular autonomic function (14). However, many studies have failed to show any significant relation of GAD antibodies to the development of neuropathy. These studies concluded that GAD antibodies had no effects on residual β-cell function or diabetic neuropathy (15). There is also no association between GAD antibodies or even islet-associated protein 2/islet cell antibody 512 with autoimmunity to nervous tissue structures or cardiac autonomic functions (16). Serum collected from type 1 diabetic patients is toxic to neuroblastoma cells of the N1E-115 cell line (17). About two-thirds of the toxicity is due to autoimmune serum factors. One of the components of this serum that mediates immune destruction of neuroblastoma cells in cultures was found to be Fas-specific IgG antibodies. These antibodies bind to Fas-ligand on the surface of N1E-115 neuroblastoma cells and induce apoptosis. Serum from patients with diabetic neuropathy contains an activator of Fas-regulated apoptosis that may contribute to the pathogenesis of diabetic neuropathy (18). There is no doubt that a variety of antibodies are present in the sera of diabetic patients with neuropathy and that the sera exert apoptotic effects on neurons grown in culture, but the missing link is the relation with clinical neuropathy and the potential for reversibility with immune therapy.
PROXIMAL NEUROPATHIES
Perhaps the clearest link between autoimmunity and neuropathy has been the demonstration of an 11-fold increased likelihood of chronic inflammatory demyelinating polyneuropathy, multiple motor polyneuropathy, vasculitis, and monoclonal gammopathies in diabetes (19). These are proximal neuropathies presenting with pain in the buttocks and thighs, fasciculation, and weakness with inability to rise from the sitting position or when kneeling on the floor. They may be the presenting symptom in many autoimmune vasculitides and celiac disease, a multigenetic, T-cell–mediated autoimmune disorder that results from a loss of tolerance to gluten (20). In support of an autoimmune mechanism for proximal neuropathies is the salutary response to intravenous Ig and immunotherapy (21).
SOMATIC NEUROPATHIES
The situation with somatic neuropathies is less clear. Several different autoantibodies in human sera have been reported that can react with epitopes in neuronal cells. Prominent among them are the gangliosides, and antibodies to GD1a, GD1b, GM1, GM2, GalNAc-GD1a, etc., are not uncommon. Other antibodies include anti-sulfatide, anti–myelin-associated glycoprotein, anti-Hu (associated with neuropathy in paraneoplastic syndromes), perinuclear anti-neutrophilic cytoplasmic antibodies, and cytoplasmic anti-neutrophilic cytoplasmic antibodies. We have reported a 12% incidence of a predominantly motor form of neuropathy in patients with diabetes associated with monosialoganglioside antibodies (22). Furthermore, we previously found that sera with high titers of phospholipase antibody inhibited the growth and differentiation of neuroblastoma cells in culture (23). Unfortunately, this is so commonplace that the issue has been raised that phospholipase antibodies do not directly contribute to nerve damage and that they are formed as a result of antigen release from tissue damage. Pittenger et al. (24) reported on neurotoxicity of sera from 39 patients with diabetic neuropathy. Neurotoxicity was assessed using the NIE neuronal cell line (adrenal medulla and ventral spinal cord 4.1, a motor cell line). Neurotoxicity correlated with vibration detection thresholds and sera from patients with motor neuropathy were highly toxic to the VSC 4.1 line, indicating that there was a relationship between the specific nerve fiber function and the type of neuronal cell killed by the serum factors. Unfortunately, there have been no trials on immunotherapy for somatic neuropathies to confirm or refute the importance of these findings.
AUTONOMIC NEUROPATHY
The relationship between autoimmunity and development of autonomic neuropathy has historically been stronger than somatic neuropathies and was first suggested in the early 1980s with the report of coincident autoimmune iridocyclitis and diabetic autonomic neuropathy (25). Autoantibodies against autonomic structures are frequently found in diabetes, although rare in type 2 diabetes (26). However, whether these antibodies lead to autonomic dysfunction is not clearly known. Retesting of neural and adrenal antibodies in diabetic autonomic neuropathy demonstrated that once present, these antibodies normally persist in these individuals; most patients who were negative at the beginning remained negative. We have previously identified the frequent occurrence of phospholipid antibodies in diabetic patients and demonstrated the positive correlation of these antibodies to the extent of neuropathy (23). Furthermore, we also reported that autoimmune neuronal destruction may contribute to the development of autonomic neuropathy in type 1 diabetes (17). Anti-sympathetic and -parasympathetic antibodies are relatively specific for type 1 diabetes, and there is evidence to suggest that these antibodies can be associated with dysautonomia. At the same time, there are many reports that dispute this and claim that antibodies against autonomic nervous system antigens are an inconsistent feature of diabetes (27). Hypoglycemia unawareness in the presence of anti–adrenal medullary antibodies (28) and diminished catecholamine output with orthostasis (29) in individuals with anti–sympathetic nervous system antibodies provide some evidence in support of the pathogenic role of autoantibodies. Furthermore, it is believed that autoimmune nerve destruction may be involved in diabetic neuropathy, even in type 2 diabetic patients, as parasympathetic nerve antibodies were found to be related to the severity of parasympathetic neuropathy in these patients (30,31). Surprisingly, the frequency of sympathetic nerve antibodies was low in type 1 diabetic patients (31). The finding of similar frequencies of Ig binding to adrenal medulla in both type 1 and type 2 diabetic patients, as well as in normal control subjects, argues against specificity of these autoantibodies (32). No association was demonstrated between anti–vagus nerve, anti–sympathetic ganglion, and anti-adrenal autoantibodies with retinopathy, peripheral somatic neuropathy, or nephropathy, even though they were frequently present in type 1 diabetes (33).
Neuronal acetylcholine receptor (AchR) antibodies are considered a novel serologic marker of neurologic autoimmunity, but the pathogenicity of neuronal AchR autoantibodies in autonomic neuropathy has not been established (34). It has been shown that patients with orthostatic hypotension and prominent cholinergic dysautonomia are most likely to be seropositive for ganglionic AchR antibodies (35) and that higher antibody titers correlated with greater autonomic dysfunction and more frequent cholinergic dysautonomia (36).
Immune responses driven by distinct neuronal AChR (ganglionic nicotinic AChR) subtypes expressed in small-cell carcinomas account for autoimmune autonomic neuropathy, as well as seizures, dementia, and movement disorders (37). Antibodies to this receptor can also interfere with ganglionic neurotransmission and produce autoimmune autonomic neuropathy (38).
Among other antibodies, autoantibodies against amphiphysin I and II have been associated with sensory motor neuropathy (39). The anti–Sc 170 and anti-U1snRNP antibodies are associated with esophageal motor dysfunction and cardiovascular autonomic neuropathy (40). Autoantibodies that activate smooth muscle l-type calcium channels produced specifically by type 1 diabetic patients may mediate gastrointestinal and autonomic dysfunction in these patients (41). Autoantibodies to nerve growth factor may play a role in diabetic autonomic neuropathy and may be a feature of evolving but not established neuropathy (42).
In this issue of Diabetes Care, Granberg et al. (43) provide epidemiological data to support the implication of autoimmunity in autonomic neuropathy. They examined 41 patients for a period of 14 years and assessed, among other measures, heart rate variability, vasoconstrictor response to cooling, and acceleration of the brake index, which was measured as the heart rate reaction to postural change at intervals over 14 years. Fifty-six percent of patients had antibodies to autonomic nervous system of some sort: sympathetic ganglion, vagus nerve, or adrenal medulla. What the authors show is that an index of autonomic neuropathy is 7.5 times more likely to become abnormal in patients who are autoantibody positive than those who are autoantibody negative. The authors did not, however, determine whether patients had autonomic nervous system antibodies at their first visit and whether this correlated to the progression of autonomic neuropathy, and they did not show the appearance and disappearance of these antibodies over time. In certain subtypes of type 1 diabetes, antibodies that are present in the beginning subsequently disappear. It is difficult to argue post hoc ergo propter hoc because antibodies may be secondary to damage to the autonomic nervous system that is occurring for other reasons. Nonetheless, this is a very interesting report suggesting that antibodies predict the evolution and development of autonomic nervous system dysfunction. The authors fall short of demonstrating that the course of autonomic neuropathy can be reversed or abrogated by therapies directed at autoimmune neuropathy and, for this reason, do not fulfill the criteria for cause and effect now evident with proximal neuropathies.
Autoimmunity in diabetic neuropathy has always been a bridesmaid but never a bride. The article by Granberg et al. (43) suggests a predictive association. The ultimate proof of the relevance of circulating antibodies to neuronal structures will rest with identification of the specific antigen and reversal of diabetic neuropathies with neutralization of the antibody to the antigen.
| Antibodies . | Associated syndrome/medical condition/symptoms . | Anatomical structure/target . | Author or related article . |
|---|---|---|---|
| Muscarinic ganglionic AChR antibodies | Orthostatic hypotension without tachycardia, cholinergic dysautonomia, abnormal blood pressure and pulse rate response to valsalva manuever, dry eyes and mouth, abnormal pupillary response, upper and lower gastrointestinal symptoms, Sicca complex, neurogenic bladder, thymoma | Ganglionic AChR | Sandroni et al. (35), Goldstein et al. (38), Klein et al. (36), Vernino et al. (44) |
| Neuronal nicotinic AChR antibodies | Autonomic neuropathy, seizures, dementia, movement disorder, carcinomas, dementia, sensory neuropathy, gastrointestinal hypomotility, dilated pupils with impaired light response, distended bladder, subacute autonomic neuropathy and related syndromes, Eaton-Lambert myasthenic syndrome | Neuronal nicotinic AChR | Lennon et al. (37), Vernino et al. (34) |
| Antibodies against muscle AChR | Myasthenia gravis | Muscle AChR (all subjects) | Vernino et al. (44) |
| Neuronal AChR antibodies | Myasthenia gravis, thymoma | Brain, peripheral nerves, serum and cerebrospinal fluid, neuronal ganglionic AChRs | Bogousslavsky et al. (46), Vernino et al. (34) |
| Ganglionic receptor–binding antibodies | Decreased salivation, idiopathic gastrointestinal dysmotility and constipation; dry skin; orthostatic intolerance; diabetic, idiopathic, or paraneoplastic autonomic neuropathy; postural tachycardia syndrome | Ganglionic receptors | Vernino et al. (44) |
| Antibodies to l-type calcium channel, P/Q-type Ca2+ channel antibodies, n-type Ca2+ channels, anti-VGCC antibodies | Type 1 diabetes, Eaton-Lambert myasthenic syndrome | Smooth muscle l-type calcium channels at the dihydropyridine binding site, P/Q-type Ca2+ channel, n-type Ca2+ channel, solubilized calcium channel–ω-conotoxin complexes, VGCC, small cells of the lung | Jackson et al. (41), O’Suilleabhain et al. (47), Lennon et al. (48), Kaiser (49) |
| Anti-CV2 antibodies | Paraneoplastic syndrome, sensory or sensory motor neuropathies | Peripheral nerves | Antoine and Camdessanche (50) |
| Anti-Hu antibodies | Subacute sensory neuropathy, demyelinating neuropathy, rapidly developing sensory neuropathy or peripheral neuropathy, early-onset dysautonomia, symptoms of Encephalomyelitis, Eaton-Lambert myasthenic syndrome | Type 1 antineuronal nuclear antibody, small-cell lung cancer, thymoma | Antoine and Camdessanche (50), O’Suilleabhain et al. (47), Camdessanche et al. (51), Winkler et al. (52), Lucchinetti et al. (53), Kusunoki and Kanazawa (54), Dalmau and Clouston (55), Anderson et al. (56), Vernino and Lennon (57) |
| Anti-neuronal antibodies (50 kDa) | Distal myasthenia gravis, sensory neuropathy | Dorsal root ganglia neurons, Purkinje cells | Uncini et al. (58) |
| Anti–amphiphysin I and II antibodies | Sensory motor neuropathy | Amphiphysin I, amphiphysin II | Perego et al. (39) |
| Anti-Sc170 antibodies | Systemic sclerosis | Esophagus | Stacher et al. (40) |
| Anti-U1snRNP antibodies | Mixed connective tissue disease | Esophagus | Stacher et al. (40) |
| Phospholipase antibodies | Diabetic neuropathy | Cell membrane phospholipid | Vinik et al. (23) |
| Compliment-fixing antibodies | Autonomic neuropathy in type 1 diabetes | Neurons | Pittenger et al. (24) |
| Anti-GAD antibodies | Type 1 diabetes, cerebellar ataxia, peripheral neuropathy, thymoma | GAD | Hoeldtke et al. (14), Vernino and Lennon (59) |
| Anti–sympathetic ganglia antibodies | Type 1 diabetes, neuropathy | Sympathetic ganglia | Zanone et al. (42), Brown et al. (29) |
| Anti–adrenal medullary antibodies | Hypoglycemia unawareness in type 1 diabetes | Adrenal medulla | De Riva (28) |
| Anti–vagus nerve antibodies | Type 1 diabetic neuropathy, parasympathetic neuropathy in type 2 diabetes | Vagus nerve | Zanone et al. (60), Sundkvist et al. (31) |
| Anti–GM1 ganglioside antibodies | Childhood-onset neuropathy, melanoma, motor dominant neuropathy, motor neuron disease | GM1 ganglioside gangliosides | Antoine et al. (61), Milesevic et al. (22), Kusunoki et al. (62) |
| Anti-MAG antibodies, anti-Po antibodies, anti–sulphated glucuronyl glycolipid antibodies | Demyelinating neuropathy | Peripheral nerve | Kusunoki et al. (62) |
| Anti-VGKC antibodies | Neuromyotonia, thymoma | VGKC | Kaiser (49), Vernino and Lennon (59) |
| Anti-recoverin antibodies | Cancer-associated retinopathy | Recoverin | Kaiser (49) |
| Antibodies . | Associated syndrome/medical condition/symptoms . | Anatomical structure/target . | Author or related article . |
|---|---|---|---|
| Muscarinic ganglionic AChR antibodies | Orthostatic hypotension without tachycardia, cholinergic dysautonomia, abnormal blood pressure and pulse rate response to valsalva manuever, dry eyes and mouth, abnormal pupillary response, upper and lower gastrointestinal symptoms, Sicca complex, neurogenic bladder, thymoma | Ganglionic AChR | Sandroni et al. (35), Goldstein et al. (38), Klein et al. (36), Vernino et al. (44) |
| Neuronal nicotinic AChR antibodies | Autonomic neuropathy, seizures, dementia, movement disorder, carcinomas, dementia, sensory neuropathy, gastrointestinal hypomotility, dilated pupils with impaired light response, distended bladder, subacute autonomic neuropathy and related syndromes, Eaton-Lambert myasthenic syndrome | Neuronal nicotinic AChR | Lennon et al. (37), Vernino et al. (34) |
| Antibodies against muscle AChR | Myasthenia gravis | Muscle AChR (all subjects) | Vernino et al. (44) |
| Neuronal AChR antibodies | Myasthenia gravis, thymoma | Brain, peripheral nerves, serum and cerebrospinal fluid, neuronal ganglionic AChRs | Bogousslavsky et al. (46), Vernino et al. (34) |
| Ganglionic receptor–binding antibodies | Decreased salivation, idiopathic gastrointestinal dysmotility and constipation; dry skin; orthostatic intolerance; diabetic, idiopathic, or paraneoplastic autonomic neuropathy; postural tachycardia syndrome | Ganglionic receptors | Vernino et al. (44) |
| Antibodies to l-type calcium channel, P/Q-type Ca2+ channel antibodies, n-type Ca2+ channels, anti-VGCC antibodies | Type 1 diabetes, Eaton-Lambert myasthenic syndrome | Smooth muscle l-type calcium channels at the dihydropyridine binding site, P/Q-type Ca2+ channel, n-type Ca2+ channel, solubilized calcium channel–ω-conotoxin complexes, VGCC, small cells of the lung | Jackson et al. (41), O’Suilleabhain et al. (47), Lennon et al. (48), Kaiser (49) |
| Anti-CV2 antibodies | Paraneoplastic syndrome, sensory or sensory motor neuropathies | Peripheral nerves | Antoine and Camdessanche (50) |
| Anti-Hu antibodies | Subacute sensory neuropathy, demyelinating neuropathy, rapidly developing sensory neuropathy or peripheral neuropathy, early-onset dysautonomia, symptoms of Encephalomyelitis, Eaton-Lambert myasthenic syndrome | Type 1 antineuronal nuclear antibody, small-cell lung cancer, thymoma | Antoine and Camdessanche (50), O’Suilleabhain et al. (47), Camdessanche et al. (51), Winkler et al. (52), Lucchinetti et al. (53), Kusunoki and Kanazawa (54), Dalmau and Clouston (55), Anderson et al. (56), Vernino and Lennon (57) |
| Anti-neuronal antibodies (50 kDa) | Distal myasthenia gravis, sensory neuropathy | Dorsal root ganglia neurons, Purkinje cells | Uncini et al. (58) |
| Anti–amphiphysin I and II antibodies | Sensory motor neuropathy | Amphiphysin I, amphiphysin II | Perego et al. (39) |
| Anti-Sc170 antibodies | Systemic sclerosis | Esophagus | Stacher et al. (40) |
| Anti-U1snRNP antibodies | Mixed connective tissue disease | Esophagus | Stacher et al. (40) |
| Phospholipase antibodies | Diabetic neuropathy | Cell membrane phospholipid | Vinik et al. (23) |
| Compliment-fixing antibodies | Autonomic neuropathy in type 1 diabetes | Neurons | Pittenger et al. (24) |
| Anti-GAD antibodies | Type 1 diabetes, cerebellar ataxia, peripheral neuropathy, thymoma | GAD | Hoeldtke et al. (14), Vernino and Lennon (59) |
| Anti–sympathetic ganglia antibodies | Type 1 diabetes, neuropathy | Sympathetic ganglia | Zanone et al. (42), Brown et al. (29) |
| Anti–adrenal medullary antibodies | Hypoglycemia unawareness in type 1 diabetes | Adrenal medulla | De Riva (28) |
| Anti–vagus nerve antibodies | Type 1 diabetic neuropathy, parasympathetic neuropathy in type 2 diabetes | Vagus nerve | Zanone et al. (60), Sundkvist et al. (31) |
| Anti–GM1 ganglioside antibodies | Childhood-onset neuropathy, melanoma, motor dominant neuropathy, motor neuron disease | GM1 ganglioside gangliosides | Antoine et al. (61), Milesevic et al. (22), Kusunoki et al. (62) |
| Anti-MAG antibodies, anti-Po antibodies, anti–sulphated glucuronyl glycolipid antibodies | Demyelinating neuropathy | Peripheral nerve | Kusunoki et al. (62) |
| Anti-VGKC antibodies | Neuromyotonia, thymoma | VGKC | Kaiser (49), Vernino and Lennon (59) |
| Anti-recoverin antibodies | Cancer-associated retinopathy | Recoverin | Kaiser (49) |
MAG, myelin-associated glycoprotein; VGCC, voltage-gated calcium channel; VGKC, voltage-gated potassium channel.
