Mechanisms of axonal dysfunction in diabetic and uraemic neuropathies

Highlights

• Nerve excitability techniques have provided important insights into the mechanisms underlying axonal dysfunction in diabetic and uraemic neuropathy.

• Excitability studies in diabetes have suggested that axonal ion channel dysfunction may contribute to the development of neuropathic symptoms.

• Membrane depolarization due to hyperkalemia may underlie the development of uraemic neuropathy.

Abstract

The global burden imposed by metabolic diseases and associated complications continue to escalate. Neurological complications, most commonly peripheral neuropathy, represent a significant cause of morbidity and disability in patients with diabetes and chronic kidney disease. Furthermore, health care costs are substantially increased by the presence of complications making investigation into treatment a matter of high priority. Over the last decade nerve excitability techniques have entered the clinical realm and enabled in vivo assessment of biophysical properties and function of peripheral nerves in health and disease. Studies of excitability in diabetic neuropathy have demonstrated alteration in biophysical properties, including changes in Na⁺ conductances and Na⁺/K⁺ pump function, which may contribute to the development of neuropathic symptoms. Interventional studies have demonstrated that these changes are responsive to pharmacological agents. Excitability studies in patients with chronic kidney disease have demonstrated prominent changes that may contribute to the development of uraemic neuropathy. In particular, these studies have demonstrated strong correlation between hyperkalaemia and the development of nerve dysfunction. These studies have provided a basis for future work assessing the benefits of potassium restriction as a therapeutic strategy in this condition.

Introduction

The prevalence and burden of metabolic disorders continues to escalate worldwide, largely due to the increasing rates of diabetes (Atkins and Zimmet, 2010). The global prevalence of diabetes is estimated to climb from 366 million in 2011 to 552 million in 2030 (Whiting et al., 2011). Furthermore, the presence of complications in diabetes has been demonstrated to result in a substantial increase to health costs and to have a marked impact on patient quality of life (Boulton et al., 2005, Gordois et al., 2003, Happich et al., 2008). In addition to causing macrovascular disease, diabetes is associated with significant microvascular complications, including neuropathy, nephropathy and retinopathy. From a neurological perspective, neuropathy remains the most prevalent long-term complication of diabetes, affecting >50% of patients with long standing diabetes (Dyck et al., 1993, Pirart, 1978). In addition to causing neuropathy, diabetes is now the leading cause of chronic kidney disease in both developed and developing nations (Atkins and Zimmet, 2010). The implications of this are significant from a neurological perspective, given that chronic kidney disease on its own causes peripheral neuropathy, also known as uraemic neuropathy, in >90% of patients with severe chronic kidney disease (Laaksonen et al., 2002, Tilki et al., 2009, Van den Neucker et al., 1998).

Metabolic polyneuropathies due to diabetes and chronic kidney disease share similar clinical features, typically manifesting as a progressive, symmetrical, length dependent neuropathy that commences in lower limb nerves. In its initial stages, patients present with sensory symptoms with motor involvement occurring in more severe cases (Krishnan et al., 2009b). Consequently, early symptoms include pain, paraesthesia and numbness in the distal lower limbs, and clinical examination reveals loss of distal sensory function and a reduction in ankle deep tendon reflexes. With disease progression distal motor involvement may develop resulting in muscle atrophy and weakness.

Neurophysiological techniques have provided important insights into the mechanisms that underlie nerve dysfunction in metabolic neuropathies. Early neurophysiological studies utilised nerve conduction methods to assess important nerve parameters, including action potential amplitude and latency. While these methods continue to provide important diagnostic information regarding neuropathy, they provide only limited information regarding pathophysiological mechanisms that may underlie the development of axonal dysfunction in these disorders.

Over the last decade, nerve excitability methods have been developed as clinical techniques for the investigation of the pathophysiological processes involved in the development of peripheral neuropathy. These methods enable the indirect assessment of axonal membrane potential and the activity of ion channels, energy dependent pumps and ion exchange processes involved in impulse conduction in peripheral nerve axons (Bostock et al., 1998, Krishnan et al., 2009a). In recent years these techniques have enabled the assessment of biophysical properties and function of peripheral nerves in health and disease (Krishnan et al., 2009a). While studies of excitability have been undertaken in inflammatory, toxic, metabolic and degenerative disorders of peripheral nerve (Cheah et al., 2012, Krishnan et al., 2006b, Park et al., 2009), it may be argued that the greatest clinical mileage has been gained in the investigation of metabolic neuropathies, particularly those relating to diabetes and chronic kidney disease. This review will focus on recent studies of nerve excitability that have been undertaken in diabetic and uraemic neuropathies. The Review will address the potential mechanisms of axonal dysfunction in these disorders and the clinical relevance of these findings in terms of the development of future diagnostic methods and novel treatments.

The concept of nerve excitability assessment as a method of investigating axonal function was pioneered by the work of Bergmans, who demonstrated that changes in the ‘threshold’ or current required to elicit a target response reflected physiological alterations of the axon (Bergmans, 1970). Bergmans discovered that by measuring changes in the threshold of axons induced by impulse activity or artificial polarisation, considerable information about axonal physiology could be inferred (Bergmans, 1970, Bostock et al., 1998). The recent development of automated excitability software has enabled the rapid acquisition of multiple excitability parameters in a short period of time (Kiernan et al., 2000). These developments have resulted in significant clinical translation, with studies of excitability having been conducted in a range of neurological disorders (Krishnan et al., 2009a).

Automated excitability protocols consist of four discrete testing paradigms; strength-duration time constant, threshold electrotonus, the current-threshold relationship and the recovery cycle.

Strength-duration behaviour is calculated from the relationship between the stimulus intensity and its duration (Fig. 1A). The strength duration properties of the nerve reflect the properties of the nodal membrane (Mogyoros et al., 1996). Specifically, strength-duration time constant provides an assessment of nodal persistent Na⁺ currents (Bostock and Rothwell, 1997). Though these currents only contribute to ∼2.5% of the total Na⁺ current, these conductances have an important role in repetitive and spontaneous activity, making this measurement a clinically relevant excitability parameter (Kiernan et al., 2000). However, the strength-duration time constant may also be influenced by passive membrane properties and membrane potential; as such measurements of latent addition may help separate persistent Na⁺ current from the passive nodal time constant (Nodera and Kaji, 2006). Latent addition uses subthreshold polarising currents of short duration to enable an in depth assessment of axonal strength duration properties (Bostock and Rothwell, 1997).

Threshold electrotonus (TE) is assessed by measuring the change in threshold at multiple time points during and after prolonged (100 ms) subthreshold, depolarizing (TEd) and hyperpolarizing (TEh) conditioning currents (Fig. 1B). Although these currents are not sufficient to produce an action potential, they enable a slow spread of current under the myelin sheath into the internode, consequently altering the membrane potential and activating a number of internodal accommodative conductances (Baker et al., 1987, Bostock et al., 1998, Burke et al., 2001).

Current/threshold relationship (I/V) assesses the effects of longer duration polarising currents by measuring the change in threshold following polarising currents of 200 ms duration, ranging from +50% (depolarising) to −100% (hyperpolarising) of the unconditioned threshold intensity (Fig 1C). This parameter is particularly useful for examining the rectifying properties of the axon (Kiernan et al., 2000). Strong depolarising currents result in outward rectification achieved by activation of fast and slow K⁺ channels (Kiernan and Bostock, 2000). Strong hyperpolarising currents result in activation of inwardly rectifying (I_H) conductances (Kiernan et al., 2000, Pape, 1996).

Recovery cycle (RC) measures the changes in threshold that occur over 200 ms following a supramaximal stimulus (Fig. 1D). These parameters provide insights into the behaviour of nodal and juxtaparanodal Na⁺ and K⁺ channels. With short conditioning-test intervals, a supramaximal stimulus causes a brief period (0.5–1 ms) of absolute refractoriness, during which the axon is completely inexcitable (Hodgkin and Huxley, 1952). This period is followed by ∼3 ms of relative refractoriness while Na⁺ channels recover from inactivation. (Hodgkin and Huxley, 1952). During this time the axon is less excitable and an action potential can only be generated with greater than usual stimuli. Following the recovery of Na⁺ channel inactivation, a passive capacitive charge stored in the internodal membrane produces a depolarising afterpotential, effectively increasing membrane excitability (Barrett and Barrett, 1982, Kiernan et al., 1996). This is referred to as superexcitability and peaks between 5–7 ms following impulse conduction. Finally, there is a second period of reduced excitability due to activation of slow K⁺ channels at the node of Ranvier during depolarisation (Baker et al., 1987, Schwarz et al., 1995). These channels are slow to deactivate and produce hyperpolarisation as the final stage of the recovery cycle known as subexcitability.

Current excitability software enables the acquisition of these parameters from a single nerve in a ten minute period. This technological advance has enabled detailed assessment of axonal function in a clinical setting, thus providing a valuable tool for clinical translation.