Microangiopathy Diabetes And The Peripheral Nervous System Experimental Studies

Tuck and colleagues (22) initially reported that experimental diabetes of rats was associated with a decline of sciatic nerve blood flow and endoneurial hypoxia. Several other laboratories have reported similar findings and a variety of interventions have been reported to both correct nerve blood flow and diabetic electrophysiological abnormalities in tandem (see review [2]). A large number of such studies through their findings have consequently implied that reductions in nerve blood flow initiate the changes of diabetic neuropathy. Although this body of work has undoubtedly provided evidence of a linkage, cause and effect has not been proven. A number of the reports have arisen from relatively few experimental laboratories (33,34,64-97). Similarly, the spectrum of agents reported to correct blood flow and conduction slowing has been very wide. As such, this range of apparently beneficial interventions raises the strong possibility that they exert parallel benefits on separate, but not necessarily linked changes from diabetes. Most such studies have also not addressed the impact of interventional agents at the level of ganglia rather than nerve trunk.

Diabetic Neuropathy Damaged Axon

Fig. 1. Images of a L5 lumbar dorsal root ganglion and sural nerve from diabetic and nondi-abetic rats exposed 2 weeks earlier to local endothelin: (A) Vasoconstriction in diabetics resulted in downstream sural nerve axonal degeneration (inset shows higher power). (B) Karyoylsis and early formation of a nest of Nageotte from a dying neuron (below it is an intact neuron with a displaced nucleus, illustrating a cell body response to axotomy). (E,F) Intranganglionic axonal degeneration. In contrast there is relatively little damage in a nondiabetic ganglion and sural nerve (C,D respectively). Bar = 50 |im for A, C-F and 20 |im for B. (Reproduced with permission from ref. 63.)

Fig. 1. Images of a L5 lumbar dorsal root ganglion and sural nerve from diabetic and nondi-abetic rats exposed 2 weeks earlier to local endothelin: (A) Vasoconstriction in diabetics resulted in downstream sural nerve axonal degeneration (inset shows higher power). (B) Karyoylsis and early formation of a nest of Nageotte from a dying neuron (below it is an intact neuron with a displaced nucleus, illustrating a cell body response to axotomy). (E,F) Intranganglionic axonal degeneration. In contrast there is relatively little damage in a nondiabetic ganglion and sural nerve (C,D respectively). Bar = 50 |im for A, C-F and 20 |im for B. (Reproduced with permission from ref. 63.)

Several experimental interventions have corrected electrophysiological and structural abnormalities of diabetic axons without an action on microvessels. For example, both intermittent near nerve and chronic neuron perikaryal exposure to intrathecal insulin infusion that provided low subhypoglycemic doses, corrected axon conduction slowing in experimental diabetes (98,99). A direct receptor-mediated neuron cell support pathway is mediated by insulin that is unrelated to microvessels. In addition, intrathecal insulin reversed axon atrophy, whereas sequestration of endogenous intrathecal insulin by an anti-insulin antibody in nondiabetic rats generated abnormalities of conduction and caliber resembling those of diabetes (99). In another example, transgenic mice lacking axon neurofilaments, (but without vascular abnormalities) had electrophysiological and structural features of neuropathy that were accelerated (100).

Some of the studies examining nerve blood flow in experimental diabetes have had methodological weaknesses, such as failure to strictly maintain near nerve temperature, incorrectly described hydrogen microelectrodes, excessively large microelectrodes, and single uncontrolled LDF measures. Other laboratories, including ours, have failed to identify declines in nerve blood flow, using rigorously controlled HC or LDF in Streptozotocin (STZ) or BioBreeding Wistar (BBW) rat models of diabetes of varying durations (23,101-103). One laboratory consistently described contrary rises in nerve blood flow in their diabetes models (36-38,104). Moreover, despite more consistent findings of shifts toward lower levels of endoneurial oxygen in the very few studies that have addressed this change, these findings do not necessarily imply that blood flow need be reduced. Changes in erythrocyte oxygen delivery might develop in diabetes, depending on changes in perfusion. Careful morphometric studies, including those using unfixed tissues, have not identified reduced vessel calibers or numbers (27,28,105), excepting one study criticized for using nonvalidated measures of nerve perfusion (whole nerve laser doppler imaging and counts of fluorescein perfused vessels of whole nerve [106]). Morphological changes of axons, an expected consequence of ischemia, are usually very mild in diabetic rat models indicating axonal atrophy. In contrast, there might be loss of epidermal axons in the skin (without tissue breakdown or damage otherwise), a finding similarly difficult to link with nerve trunk or skin microangiopathy. Accurate measures of nerve and ganglion blood flow in mouse models of diabetes have also not been available to date for technical reasons.

Four separate studies have identified reductions in lumbar dorsal root ganglion blood flow in experimental diabetes in rats: three in STZ rats using HC or [14C] iodoan-tipyrene distribution and one in BBW rats (63,102,103,107). In two of the studies, ganglion, but not nerve blood flow was reduced by diabetes. In the studies of BBW spontaneously diabetic female rats blood flow was normal in ganglia and nerve 2-3 months after the onset of diabetes, then declined to values less than nondiabetics by 4-6 months (Fig. 2). Oxygen tension measures trended toward lower values in nerve and ganglia (Fig. 3). In ganglia it might be that changes in blood flow and oxygen tension do not occur in parallel because there might also be decreases in oxygen extraction as blood flow declines.

A distinct feature of diabetic microangiopathy is its selective impact on vasodilation. The consequence of this alteration has been unopposed vasoconstrictor sensitivity both in large macrovessels and the microcirculatory bed (Fig. 4). Loss of vasodilation in diabetes might be secondary to vascular unresponsiveneness to NO (108-113), accelerated "quenching" of intraluminal NO by advanced glycosylation end products (114), or endothelial damage that fails to normally elaborate it (115,116). NO synthase isoforms (NOS) protein or messenger RNA (mRNA) were not altered by experimental diabetes in our laboratory beyond a slight decline in iNOS. Thus, deficits in NO vascular actions in diabetic nerves likely reflect failed action rather than a decline in its synthesis in diabetes.

Polarography Medical Applications

Fig. 2. Local measurements of nerve and ganglion blood flow using microelectrode HC polarography. On the left side of the bar chart, note that endoneurial blood flow measures were unaltered in BBW diabetic rats after 7-11 weeks of diabetes or 17-23 weeks of diabetes compared with nondiabetic littermates Control (Con). In L5 dorsal root ganglia, however, older diabetics did have a reduction in ganglion blood flow. Results are also compared with pooled values from normal rats (Lab con). (Reproduced with permission from ref. 103.)

Fig. 2. Local measurements of nerve and ganglion blood flow using microelectrode HC polarography. On the left side of the bar chart, note that endoneurial blood flow measures were unaltered in BBW diabetic rats after 7-11 weeks of diabetes or 17-23 weeks of diabetes compared with nondiabetic littermates Control (Con). In L5 dorsal root ganglia, however, older diabetics did have a reduction in ganglion blood flow. Results are also compared with pooled values from normal rats (Lab con). (Reproduced with permission from ref. 103.)

Diabetes Bar Chart

Fig. 3. Composite bar charts showing pooled oxygen tension histograms from experimental diabetes work carried out in the Zochodne lab. The values are constructed from multiple sampling in groups of diabetic rats (STZ or BBW) of diabetes duration 4 months or longer and nondiabetic littermates. Note that there are shifts toward lower oxygen tension levels in both the endoneurium and dorsal root ganglia of diabetic rats and that there are more lower level oxygen tension values in ganglia than endoneurium from nondiabetic rats.

Fig. 3. Composite bar charts showing pooled oxygen tension histograms from experimental diabetes work carried out in the Zochodne lab. The values are constructed from multiple sampling in groups of diabetic rats (STZ or BBW) of diabetes duration 4 months or longer and nondiabetic littermates. Note that there are shifts toward lower oxygen tension levels in both the endoneurium and dorsal root ganglia of diabetic rats and that there are more lower level oxygen tension values in ganglia than endoneurium from nondiabetic rats.

Neuropathy Peripheral Nervous System

Fig. 4. Local measurements of L5 ganglion blood flow using microelectrode HC polarogra-phy in diabetic rats and nondiabetic littermates before and after exposure to local endothelin. Baseline blood flow was reduced in diabetics. With endothelin exposure, diabetic ganglion blood flow declined to lower values and had a prolonged action (not shown) rendering ischemic damage (Fig. 1). (Reproduced with permission from ref. 63.)

Fig. 4. Local measurements of L5 ganglion blood flow using microelectrode HC polarogra-phy in diabetic rats and nondiabetic littermates before and after exposure to local endothelin. Baseline blood flow was reduced in diabetics. With endothelin exposure, diabetic ganglion blood flow declined to lower values and had a prolonged action (not shown) rendering ischemic damage (Fig. 1). (Reproduced with permission from ref. 63.)

Indeed, overall NOS enzyme activity was increased, perhaps indicating a compensatory rise in its production in the setting of accelerated consumption (117). Diabetes might also be associated with unresponsiveness to peptide vasodilatation (118), loss of perivascular peptidergic axons innervating vasa nervorum (119), increased sensitivity to angiotensin II (120), increased adrenergic tone (121), or decreased vasodilation from prostacyclin (122) (Fig. 5).

If not a "triggering" event in the development of diabetic polyneuropathy, microangiopathy does account for some unique properties of diabetic nerves. As discussed earlier, diabetic peripheral nerve trunks exposed to a mild ischemic insult unexpectedly develop severe damage to axons. Such damage can occur in spite of axon resistance to ischemic conduction failure (RICF), a property whereby ischemic axons have a delayed loss of excitability. RICF is an invariable, though not exclusive, electrophysiological property of diabetic axons that is linked to chronic hypoxia and excess glucose substrate (123). RICF might be thought of as a mechanism of protecting diabetic axons, yet it fails to protect them from ischemia lasting longer than the duration of even prolonged RICF (its duration is approximately 25-35 minutes after complete ischemia and is defined most often as the time for a 50% decline in motor axon excitability). For longer durations of ischemia, diabetic axons are more likely to undergo axonal degeneration than nondiabetic axons when exposed to ischemia. In nondiabetic nerves, complete ischemia of 1-3 hours is required to damage axons. In diabetic nerves, pre-existing hypoxia and loss of compensatory vasodilation are both likely to contribute to their vulnerability.

DIABETES, REGENERATION, AND THE MICROCIRCULATION

Microangiopathy likely impacts the regenerative microenvironment of an injured diabetic peripheral nerve trunk. Focal lesions or mononeuropathies are common and

Neuropathy Peripheral Nervous System

Fig. 5. Local measurements of sciatic nerve blood flow using microelectrode HC polarogra-phy in nondiabetic rats (left) and diabetic rats (right) before (pre) and after (post) exposure to topical capsaicin. Capsaicin generates acute release of vasoactive peptides from perivascular sensory terminals after topical application. In nondiabetic rats, there are prompt rises in endoneur-ial blood flow from acute activation of peptidergic epineurial vasodilatation. The rise was inconsistent and insignificant in diabetic rats.

Fig. 5. Local measurements of sciatic nerve blood flow using microelectrode HC polarogra-phy in nondiabetic rats (left) and diabetic rats (right) before (pre) and after (post) exposure to topical capsaicin. Capsaicin generates acute release of vasoactive peptides from perivascular sensory terminals after topical application. In nondiabetic rats, there are prompt rises in endoneur-ial blood flow from acute activation of peptidergic epineurial vasodilatation. The rise was inconsistent and insignificant in diabetic rats.

disabling in diabetes. Examples are carpal tunnel syndrome, intercostal neuropathies, and lumbosacral plexopathies. These focal peripheral nerve lesions regenerate more slowly in diabetics than nondiabetics (124,125) and regeneration from them might be incapable of restoring function in many patients. If clinical therapy to arrest polyneuropathy is developed, attention will shift toward understanding how diabetic nerves with failed regeneration might be resurrected.

Ischemic peripheral nerve lesions also have impaired regeneration (126). What level of perfusion is required to sustain or enhance regenerative activity involving axons, Schwann cells, and other cellular constituents is unknown. "Normal" rises in local blood flow, that develop when peripheral nerves are injured, are robust, persistent, and probably important in supporting nerve repair. Despite early work suggesting otherwise, most focal injuries of peripheral nerves do not involve ischemia. For example, at the site of an acute nerve crush, there might be expectations of local vessel disruption, local microthrombosis, and endoneurial edema. However, direct blood flow measures of blood flow at the site of experimental crush in rats, indicate that injured nerve trunks compensate for such microvascular disruptions with preserved and enhanced local blood flow (53,54). Moreover, by 24-48 hours, endoneurial and epineurial blood flow rises substantially after peripheral nerve injury, a development linked to local actions of CGRP and NO (54,127,128). Their elaboration within this microenvironment is part of an interesting and complex story involving the formation of axonal endbulbs and accumulation of peptides and enzymes within them. In the case of NO, accumulation of NOS within endbulbs allows diffusion of the vasodilator to nearby microvessels. Within a few days following nerve injury, angiogenesis, especially prominent in the epineurial plexus, ensues and is associated with local rises in vascular endothelial growth factor (VEGF) mRNA (129,130). Neovascularization develops concurrently with other events within the nerve trunk involving axon outgrowth and Schwann cell (SC) proliferation.

Distal stumps of severed peripheral nerves eventually, become inhospitable to new axon entry. Several factors contribute to the hostility of long-term denervated nerve stumps including changes in their microvascular supply. Distal nerve stumps that remained unconnected for 6 months had a decline in their epineurial (LDF) and endoneurial (HC) blood flow by more than 50% (129). Although new blood vessels formed soon after injury were retained in these stumps, their caliber and number gradually declined as did VEGF mRNA levels. Overall the remodeling of the microvascular supply with prolonged denervation is interesting and highlights a close relationship between axons and microvessels in the peripheral nerve. These changes also render an unsuitable ischemic microenvironment for late regeneration, should the nerve be reconnected. In injured diabetic nerves, distal nerve stumps that have not been reinnervated because of slowed regeneration suffer such problems that then impose further barriers to regrowth. In addition to ischemia, long-term denervated nerve trunks undergo other changes that make them less hospitable including Schwann cell atrophy, loss of growth factor synthesis (e.g., Glial cell line-derived neurotrophic factor [GDNF]), and perhaps excessive collagen elaboration with fibrosis (131-133).

In diabetic nerve trunks that are transected, rises in local blood flow in both the proximal and distal stumps of the nerve are attenuated and angiogenesis is dampened. Kennedy and Zochodne (134) described microvascular changes associated with transected sciatic peripheral nerve trunks in rats with experimental diabetes (STZ) of 8 month duration using quantitative HC measures of endoneurial flow and multiple LDF epineurial plexus sampling. The findings were correlated with quantitation of endoneurial and epineurial vessels by India ink perfusion using unfixed tissues. As described earlier, nondiabetic rats exhibit substantial rises in local nerve blood flow, or hyperemia in both proximal and distal nerve stumps that is evident at both 48 hours and 2 weeks after injury. Although diabetic nerve trunks had normal, unaltered blood flow after sham exposure without injury, epineurial rises in blood flow following injury were almost completely dampened. Rises in blood flow within the endoneurium also failed to develop in diabetic nerves at 48 hours, but began to rise by 2 weeks (Fig. 6). There were rises in the numbers of epineurial vascular profiles in nondiabetic, but not in diabetic rats in the distal and proximal stumps. Similarly, there were substantial rises in mean vascular luminal areas of both endoneurial and epineurial vessels of the proximal and distal stumps after transection that were almost completely attenuated in diabetic vessels (Figs. 7 and 8). Overall the findings suggested that diabetic functional microangiopathy imposes severe limitations on the capability of vasa nervorum to respond to injury. These limitations involve both failed epineurial and endoneurial vascular dilatation and a somewhat later failure of compensatory angiogenesis.

Additional work indicates that diabetic nerves also fail to upregulate NOS activity at the site of a peripheral nerve injury. Normally, iNOS expression and activity rise following peripheral nerve injury concomitant with SC phagocytic activity and macrophage invasion. Both SCs and macrophages express iNOS (135). Rises of iNOS activity in turn, are important in facilitating the process of Wallerian degeneration through clearance of axon and probably more importantly, myelin components that inhibit regeneration. As such, effective clearance is a prerequisite for later successful axonal ingrowth. Nondiabetic

Axon Affected Diabetic Neuropathy

Proximal Distal

Days

Fig. 6. Local measurements of erythrocyte flux (LDF) made in rats with experimental diabetes of 8 months duration and littermates, proximal and distal to a sciatic nerve transection. Note the prominent rises in blood flow after injury in nondiabetic sciatic nerves that are attenuated in diabetics. There are no changes in flux of intact nerves in diabetic rats (sham surgery). (*p < 0.05 between diabetics and nondiabetics). (Reproduced with modifications and permission from ref. 134.)

Proximal Distal

Days

Fig. 6. Local measurements of erythrocyte flux (LDF) made in rats with experimental diabetes of 8 months duration and littermates, proximal and distal to a sciatic nerve transection. Note the prominent rises in blood flow after injury in nondiabetic sciatic nerves that are attenuated in diabetics. There are no changes in flux of intact nerves in diabetic rats (sham surgery). (*p < 0.05 between diabetics and nondiabetics). (Reproduced with modifications and permission from ref. 134.)

mice lacking iNOS have delays in the progress of Wallerian degeneration and subsequent regeneration (136).

MICROANGIOPATHY, DIABETES, AND THE PERIPHERAL NERVOUS SYSTEM

Human Investigations

There have been a limited number of human investigations of nerve trunk microan-giopathy in diabetes. The earliest and most convincing data are rigorous morphological studies of epineurial and endoneurial blood vessels in human sural nerve biopsies. These have identified microthrombosis and microvessel occlusion in diabetic nerves, endothe-lial duplication, smooth muscle proliferation, endoneurial capillary closure, basement membrane thickening, pericyte degeneration, and other changes (105,137-146). The loss of axons in a multifocal pattern in such biopsies has also suggested an ischemic or microvascular etiology (4,140,141,147). However, it is difficult to ascribe cause and effect to morphological changes of peripheral nerve biopsies because they cannot tell us whether the changes in vessels parallel or follow loss of axons or other abnormalities. In earlier investigations, such biopsies were harvested from older patients with already existing neuropathy. More recently, however, Malik and colleagues (148) have demonstrated that microvascular changes might develop early in patients with only mild neuropathy, findings that suggest a role in promoting axonal damage. Similarly recent work by Tesfaye and colleagues (149) has linked the development of neuropathy with risk factors for cardiovascular disease, as well as implying a close relationship between macrovessels and neuropathy.

Risk Factors For Diabetic Neuropathy

Fig. 7. Total vascular luminal areas of vasa nervorum from nerves of 8 month diabetic rats or littermates proximal (above) or distal (below) to a sciatic nerve transection. The rats were perfused with India ink and both endoneurial and epineurial vascular profiles were counted and sized using unfixed tissues. (Units are in |m2. Bars are diabetics [open], nondiabetics [solid], diabetic sham [hatched], and nondiabetic sham [squares]). Note that in nondiabetics there is a rise in endoneurial vascular area at the early 48 hour time-point but a more prominent rise in the epineurium at both time-points. These rises are attenuated in diabetics. (Reproduced with permission from ref. 134.)

Fig. 7. Total vascular luminal areas of vasa nervorum from nerves of 8 month diabetic rats or littermates proximal (above) or distal (below) to a sciatic nerve transection. The rats were perfused with India ink and both endoneurial and epineurial vascular profiles were counted and sized using unfixed tissues. (Units are in |m2. Bars are diabetics [open], nondiabetics [solid], diabetic sham [hatched], and nondiabetic sham [squares]). Note that in nondiabetics there is a rise in endoneurial vascular area at the early 48 hour time-point but a more prominent rise in the epineurium at both time-points. These rises are attenuated in diabetics. (Reproduced with permission from ref. 134.)

Direct approaches to measure blood flow in human nerve have been fewer and less definitive. Such approaches have included fluorescein transit times that are delayed in patients with diabetes with advanced neuropathy (29,150). Endoneurial oxygen tensions in patients with established neuropathy were reduced when measured with direct micro-electrode recordings (151). Theriault et al. (152) measured human sural nerve blood flow using multiple epineurial LDF measures of erythrocyte flux in patients in the operating room (with a micromanipulator and theater lights turned off) undergoing sural

Fig. 8. Examples of sciatic nerves from nondiabetic (left) and diabetic (right) rats perfused with India ink to outline vasa nervorum in the distal nerve stump 2 weeks following transection. Note the large number of perfused vessels in epineurial area of the nondiabetics, but fewer in diabetics. (Bar = 1 mm) (Reproduced with permission from ref. 134.)

nerve biopsy as part of a clinical trial for therapy in diabetic neuropathy. In the trial, subjects with relatively mild diabetic polyneuropathy (presence of a sural nerve sensory nerve action potential) underwent a biopsy on one leg at the outset of the trial then on the contralateral leg one year later. Ultimately, the trial agent (acetyl-L-carnitine) did not improve neuropathy, but the LDF measures yielded interesting insights. Measures were compared with results from "disease control" subjects with nondiabetic neuropathies. These controls provided proof of principal, because patients with necrotizing vasculitis of nerve had expected reductions in blood flow as did one subject in whom the surgeon applied the vasoconstrictor epinephrine over the epineurial plexus before biopsy. Overall diabetic subjects had no evidence of lowered blood flow, but trends toward higher values (Fig. 9). Interestingly, there were similar trends toward higher blood flow values in the same subjects when studied serially after one year, and patients with lower sural nerve sensory nerve action potentials (that correlated very closely with lowered myelinated fiber density) also trended toward higher, rather than lower blood flow values. These findings therefore did not link alterations in blood flow, as assessed using an epineurial LDF probe, with progressive diabetic axon loss and neuropathy.

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook


Responses

Post a comment