Though as many as one in five people with diabetes experience painful diabetic neuropathy, the causes of this severe complication have remained unclear, leaving patients with few treatment options. Now, two studies pinpoint novel mechanisms—one in the periphery, and one in the central nervous system—driving the pain. The studies link painful diabetic neuropathy to the pathophysiology of both diabetes and neuropathic pain, and suggest new targets for treatment.
In one paper, Peter Nawroth, Angelika Bierhaus, Thomas Fleming, and colleagues at University Hospital Heidelberg, Germany, in collaboration with Peter Reeh and colleagues at Friedrich-Alexander-University Erlangen-Nuremberg in Germany, and other groups, report that diabetic patients with pain have elevated plasma levels of the glucose metabolite methylglyoxal, which they show heightens the activity of peripheral sensory neurons by post-translationally modifying the voltage-gated sodium channel Nav1.8. Nawroth and colleagues further show that, in mice with streptozotocin (STZ)-induced diabetes, treatment with a peptide scavenger of methylglyoxal reduces pain.
How the metabolic disturbances of diabetes lead to nervous system dysfunction is not well understood, and the study, which appeared online May 13 in Nature Medicine, is notable for tracing painful changes in sensory neurons back to a specific harmful metabolite.
“I have patients with devastating pain, and I have patients without any pain, and I have no idea what the difference is,” Nawroth said. “Now we have a molecule that perhaps can help us differentiate.”
In a separate study, researchers present evidence that, as for other forms of neuropathic pain, alterations in central nervous system structure and function contribute to pain in rats with diabetic neuropathy. Using the STZ diabetes model, Stephen Waxman, Andrew Tan, and colleagues at Yale University School of Medicine, New Haven, Connecticut, US, and the Veterans Affairs Connecticut Healthcare System, West Haven, US, find that painful neuropathy is associated with changes in the morphology of dendrites of spinal cord neurons involved in pain transmission. Reversing those changes with an inhibitor of the GTPase Rac1, which has an important role in dendritic spine plasticity, decreases neuronal firing and reduces pain, they show. The study appeared May 16 in the Journal of Neuroscience.
A menace among metabolites
High blood glucose (hyperglycemia) causes many of the complications of diabetes, and the mainstay of diabetes treatment is glucose control. However, lowering glucose levels is often unsuccessful in treating neuropathic symptoms, including pain, especially in type 2 diabetes (for a just-published review, see Callaghan et al., 2012). Instead, there is evidence that toxic byproducts of glucose metabolism—reactive dicarbonyl compounds that modify lipids, nucleic acids, and proteins—are important causative factors. One of those reactive glucose byproducts is methylglyoxal.
In their study, Nawroth, Bierhaus, and colleagues measured plasma methylglyoxal in 20 patients with type 2 diabetes, and found that all of the patients had higher levels than did healthy controls. For patients with pain, levels were even higher than for those without pain. Neuropathy deficit scores were equivalent in the two diabetes groups, indicating that methylglyoxal was not a general marker of diabetic neuropathy, but was specifically associated with pain.
“There’s a threshold [of plasma methylglyoxal concentration] above which we see pain—and below which, we don’t,” Nawroth said. He noted, though, that the threshold effect might be an artifact of the small number of patients, so larger studies are needed.
Further experiments revealed that methylglyoxal was not only associated with pain, but also caused it by direct actions on neurons. The metabolite heightened excitability of dorsal root ganglia (DRG) sensory neurons in culture. In mice, administration of methylglyoxal rendered the animals hypersensitive to heat and touch stimuli, and induced other changes including release of the pro-inflammatory neuropeptide calcitonin gene-related peptide (CGRP) in skin tissue. Increasing endogenous methylglyoxal levels by knockdown or inhibition of the methylglyoxal-metabolizing enzyme glyoxylase-1 (GLO1) had similar effects.
Nawroth said that he and his colleagues expected methylglyoxal’s excitatory effects on sensory neurons to involve methylglyoxal-modified proteins that interact with the receptor for advanced glycation endproducts (RAGE)—a pathway involved in a host of diabetic complications (for a recent review, see Ramasamy et al., 2011). Instead, their experiments led them to Nav1.8, a sodium channel that is involved in initiation and propagation of action potentials in sensory neurons. They discovered that Nav1.8 in sciatic nerve tissue from patients with diabetes, and in DRG from STZ mice, bore methylglyoxal modifications. Electrophysiological experiments indicated that the modifications increased Nav1.8 activity by inhibiting channel closing. Nav1.8 was required for methylglyoxal-induced hyperalgesia and CGRP release, as both effects were absent in Nav1.8-deficient mice.
In a step towards therapy, the researchers designed a peptide to scavenge methylglyoxal, and found that it decreased methyglyoxal- and diabetes-induced hyperalgesia when administered systemically in mice. Nawroth says he is now keen to develop the compound for clinical testing.
The study also revealed that methylglyoxal inactivated a related sodium channel, Nav1.7, but the impact of those changes is not yet known. Furthermore, Nawroth says his group is exploring whether additional glucose byproducts alter Nav1.8 activity. “Is it only methylglyoxal, or do other reactive metabolites also inhibit closing of this channel and induce pain?” he said.
The Nawroth group’s findings join other new results in implicating methylglyoxal and additional toxic metabolites in diabetic neuropathy. Recently, researchers reported that mouse strains expressing higher levels of GLO1 in DRG neurons were protected from STZ-induced sensory neuropathy (Jack et al., 2012). Together, the studies to date suggest that there is much to learn about glucose metabolites, and their detoxifying systems, in diabetic neuropathy and pain.
Shaping pain memories
In the brain, experience alters the structure of spines—the tiny synapse-bearing projections on dendrites—to modulate synaptic strength and help store memories. Previously, the Waxman lab reported that similar changes occur in spinal cord dorsal horn neurons in animal models of spinal cord injury or peripheral nerve damage, and that this plasticity supports a pain “memory” that makes the cells hyper-responsive to future signals (reviewed in Tan and Waxman, 2011, and Melemedjian and Price, 2011). The structural changes in dorsal horn dendrites “are a physical engram” that “locks in the pain,” Tan said, just as dendritic remodeling locks in memories in the brain.
The new study indicates that those changes in the dorsal horn also underlie diabetic neuropathic pain. Specifically, four weeks after STZ induction of diabetes in rats, Tan found that dendritic spines on dorsal horn neurons were longer and had larger heads, giving them a mushroom shape. In addition to the structural changes, the neurons displayed heightened excitability and enlarged receptive fields on the skin, at approximately the same time as the animals became hyper-responsive to mechanical stimuli. Treating the rats with an inhibitor of Rac1—an enzyme involved in reorganization of the actin cytoskeleton which supports the spines—countered all of the changes, normalizing (at least partially) spine shape, neuronal activity, and mechanical hypersensitivity.
Rac1 inhibitors are already being developed as potential cancer therapeutics. The new results suggest that targeting Rac1, or other molecules that control dendritic spine morphology, “might be therapeutic approaches for reducing neuropathic pain either in diabetes or other neuropathic pain conditions,” Tan said.
Top image: Methylglyoxal.
Angelika Bierhaus, lead author of the study on methylglyoxal, died in April. To learn more about her life and work, see Nawroth, 2012.
A discussion on drug development, biomarkers, and animal models in diabetic neuropathic pain has been initiated in the PRF Forums section. We invite you to add your ideas by commenting either on this news story or at the Forum discussion.