Hypothermic neuroprotection of peripheral nerve of ...

Brain (1999), 122, 161–169

Hypothermic neuroprotection of peripheral nerve of rats from ischaemia–reperfusion injury Yoshiyuki Mitsui,1 James D. Schmelzer,1 Paula J. Zollman,1 Mikihiro Kihara2 and Phillip A. Low1 1Department

of Neurology, Mayo Clinic, Rochester, Minnesota, USA and 2Department of Neurology, Kinki University School of Medicine, Osaka, Japan

Correspondence to: Phillip A. Low, MD, Department of Neurology, Mayo Clinic, 811 Guggenheim Building, Rochester, MN 55905, USA E-mail: [email protected]

Summary sensory nerves was undertaken 1 week after surgical procedures. At that time, entire sciatic–tibial nerves were harvested and fixed in situ. Four portions of each nerve were examined: proximal sciatic nerve, distal sciatic nerve, mid-tibial nerve and distal tibial nerve. To determine the degree of fibre degeneration, each section was studied by light microscopy, and we estimated an oedema index and a fibre degeneration index. The groups treated at 36– 37°C underwent marked fibre degeneration, associated with a reduction in action potential and impairment in behavioural score. The groups treated at 28°C (for both 3 and 5 h) showed significantly less (P , 0.01; ANOVA, Bonferoni post hoc test) reperfusion injury for all indices (behavioural score, electrophysiology and neuropathology), and the groups treated at 32°C had scores intermediate between the groups treated at 36–37°C and 28°C. Our results showed that cooling the limbs dramatically protects the peripheral nerve from ischaemia–reperfusion injury.

Keywords: ischaemia; neuroprotection; rat; hypothermia; peripheral nerve Abbreviations: CMAP 5 compound muscle action potential; IFD 5 ischaemic fibre degeneration; IR injury 5 ischaemia–reperfusion injury; SAP 5 sensory action potential

Introduction Although the neuropathology of nerve ischaemia is well described, scant information is available on its pathophysiology and on peripheral nerve neuroprotection from ischaemic fibre degeneration (IFD). Low energy needs, a large energy reserve and the ability to utilize anaerobic metabolism are factors that result in a comparatively large safety factor preventing the development of IFD (Low et al., 1989). This greater safety factor preventing IFD could enhance the efficacy of neuroprotective measures. In this study, we evaluated the efficacy of hypothermia in the protection of nerves from IFD resulting from ischaemia– reperfusion injury, using an optimized model of nerve ischaemia induced by multiple arterial ligatures. © Oxford University Press 1999

Method Ligation model of ischaemia–reperfusion We used Sprague–Dawley male rats, weighing approximately 300 6 10 g at the beginning of the study. Forty-eight rats were divided into six groups. We used a modification of the arterial ligation previously used in our laboratory (Nagamatsu et al., 1996). Our original model involved ligation of the arteries supplying both hind limbs, and because the mesenteric supply was also ligated it resulted in high mortality. Our present model was produced by ligating the aorta and the arteries supplying the right hind limb. The following vessels were individually ligated: (i) the abdominal aorta proximal to the iliolumbar artery; (ii) the abdominal aorta distal to the

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Although there is much information on experimental ischaemic neuropathy, there are only scant data on neuroprotection. We evaluated the effectiveness of hypothermia in protecting peripheral nerve from ischaemia–reperfusion injury using the model of experimental nerve ischaemia. Forty-eight male Sprague– Dawley rats were divided into six groups. We used a ligation–reperfusion model of nerve ischaemia where each of the supplying arteries to the sciatic–tibial nerves of the right hind limb was ligated and the ligatures were released after a predetermined period of ischaemia. The right hind limbs of one group (24 rats) were made ischaemic for 5 h and those of the other group (24 rats) for 3 h. Each group was further divided into three and the limbs were maintained at 37°C (36°C for 5 h of ischaemia) in one, 32°C in the second and 28°C in the third of these groups for the final 2 h of the ischaemic period and an additional 2 h of the reperfusion period. A behavioural score was recorded and nerve electrophysiology of motor and

162

Y. Mitsui et al.

Fig. 1 Experimental paradigm. See text for details.

Regulation of limb temperature The experimental paradigm is shown in Fig. 1. The supplying arteries to the right sciatic–tibial nerve were ligatured at room temperature. Temperature control was maintained for 2 h prior to release of ligatures (reperfusion) and maintained for 2 h of reperfusion. For the three groups subjected to 3 h of ischaemia (IR337; IR3-32; IR3-28; IR 5 ischaemia–reperfusion), the right lower extremity was maintained at 37, 32 and 28°C, respectively for the last 2 h of ischaemia and the first 2 h of reperfusion. The silk ties were released after 3 h of ischaemia. For the three groups subjected to 5 h of ischaemia (IR536; IR5-32; IR5-28), the temperature was maintained at 36, 32 and 28°C, respectively, for the last 2 h of ischaemia and the first 2 h of reperfusion. The silk ties were released after 5 h of ischaemia. The limb temperature was monitored at the mid-thigh level with an intramuscular thermistor probe. The right lower

extremity was temperature-controlled using a specially designed thermal garment (HAAKE® KT2, Germany) containing fine tubes through which temperature-controlled water was circulated. Deep rectal temperature, measured by inserting a rectal thermistor probe 7 cm into the rectum, was also monitored with a thermistor probe and maintained at 36.5 6 1°C. Body temperature was maintained by the use of an infrared lamp with the temperature clamped by servo control at 36.5°C. The same nerves were studied for electrophysiology and histology. Behavioural measurements were done on all animals. The studies were approved by the Mayo Institutional Animal Care and Utilization Committee and fully complied with their guidelines.

Behavioural score The function of the limb was scored with the observer blinded to the status of the rats. The score was based on gait, grasp, paw position and reaction to pinch. Gait was scored from 0 (no function) to 3 (normal function), with 1 and 2 for very and slightly impaired, respectively. Similarly, grasp was graded from 0 (none) to 3 (normal). Paw position varied from 0 (paw contracted) to 3 (normal). Withdrawal from pinch was scored either as present (2) or absent (0). Increasing function was indicated by a larger score. From the foregoing, a composite score, 0 (no function) to 11 (normal function), reflecting increasing limb function was derived.

Electrophysiology We used techniques that are standard for our laboratory (Low and Tuck, 1984; Parry and Kozu, 1990). We measured the amplitude of compound muscle action potential (CMAP) of the sciatic–tibial nerve and sensory nerve action potential in the digital nerve using fine stainless steel near-nerve stimulating and recording electrodes. The CMAP was recorded from the dorsum of the hind paw while stimulating at the level of the sciatic notch and ankle. The sensory nerve action potential was recorded from the ankle while stimulating the digital nerve at the tip of the digit. Recordings were done at 35°C and amplified 1000-fold, stored on computer disk,

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

iliolumbar artery but proximal to the inferior mesenteric artery; (iii) the common iliac artery; (iv) the hypogastric trunk; (v) the superior gluteal artery; (vi) the internal pudendal artery; (vii) the external iliac artery; (viii) the deep circumflex iliac artery; (ix) the femoral artery; (x) the superficial circumflex iliac artery; (xi) the saphenous artery. The rat was anaesthetized with intraperitoneal pentobarbital (60 mg/kg) and supplemented with additional doses of pentobarbital if anaesthesia lightened during the experiment. Silk ties with slip-knots were used for ready release and reperfusion. Ischaemia was maintained for 3 or 5 h and was followed by reperfusion for 1 week prior to removal of the nerve for pathological studies. On the day of tissue harvest, the function of the limb was evaluated using a behavioural score, and electrophysiological studies were done. Half of 48 rats were used for a study in which the duration ischaemia was 3 h, and the other half were used for a study in which ischaemia lasted for 5 h. Each group was further divided into three groups of eight rats for treatment at different temperatures: 37°C (3 h ischaemic group); 36°C (5 h ischaemic group); 32 and 28°C.

Hypothermic neuroprotection of nerve

163

and analysed off-line using a Nicolet digital oscilloscope (Nicolet Instruments, Madison, Wis., USA). For each animal, identical recordings were done on the contralateral nerve, which served as control.

Neuropathology Nerves were fixed in situ for 30 min using 4% glutaraldehyde in 0.1 mol/l phosphate buffer (pH 7.4). The entire length of the sciatic–tibial nerve was removed, and the following segments were separately studied: the proximal sciatic; distal sciatic; mid-tibial; and distal tibial. These nerve segments were separately osmicated, dehydrated, infiltrated and embedded in epoxy resin. Transverse sections of 0.5 µm were stained with 1% toluidine blue. Under 4003 magnification, these sections were graded for oedema and IFD using a modification of previously described methods (Kihara et al., 1993). In brief, all slides were coded and read by blinded observers. Fibres were considered to be undergoing ischaemic degeneration if axonal changes were visible. The axon may be swollen or shrunken, watery and light, or dark and shrunken. Secondary myelin changes are typically seen, including attenuation, collapse or breakdown. For each section the percentage of fibres undergoing IFD was estimated. Oedema was graded semiquantitatively as follows: 0 5 normal; 1 5 mild oedema; 2 5 moderate oedema; 3 5 severe oedema. No distinction was made as to endoneurial, perivascular or subperineurial oedema. Figure 2 shows representative transverse sections at increasing severities of endoneurial oedema.

Statistics All experiments were done on groups of eight rats and values are expressed as mean 6 SEM. Statistical analysis was done

Fig. 3 Behavioural score for 3 h (A) and 5 h (B) of ischaemia. ***P , 0.001.

using ANOVA (analysis of variance) with Bonferoni post hoc analysis. Significance was accepted at P , 0.05.

Results Behavioural score The behavioural score for groups IR3–37, IR3–32 and IR3– 28 was 1.4 6 0.4, 5.4 6 0.8 and 10.1 6 0.4, respectively (Fig. 3A). The behavioural score for groups IR5–36, IR5– 32 and IR5–28 was 2.7 6 0.5, 4.3 6 0.4 and 5.8 6 0.6, respectively (Fig. 3B). Behavioural scores were consistently normal (11.0) for the contralateral side.

Electrophysiology In the present study, the results of electrophysiology are expressed as percentages of values for the left side. The mean sensory action potential (SAP) amplitude of the right digital nerve was 6.2 6 4.1, 95.8 6 16.5 and 126.9 6 36.9% for groups IR3–37, IR3–32 and IR3–28, respectively (Fig. 4A). The mean CMAP amplitude of the right sciatic–tibial

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Fig. 2 Representative transverse sections from distal sciatic nerve showing oedema grade 0 (5 h ischaemia, 28°C), grade 1 (5 h ischaemia, 28°C), grade 2 (3 h ischaemia, 32°C) and grade 3 (5 h ischaemia, 36°C). Scale bar 5 100 µm.

164

Y. Mitsui et al.

nerve was 0, 5.1 6 1.4 and 38.3 6 12.8% for groups IR3– 37, IR3–32 and IR3–28, respectively (Fig. 4B). The SAP amplitude of the right digital nerve was 0.6 6 0.6, 36.1 6 11.0 and 57.5 6 16.9% for groups IR5–36, IR5–32 and IR5–28, respectively (Fig. 4C). The mean CMAP amplitude of the right side, expressed as a percentage of the left side, was 0.5 6 0.5, 8.5 6 4.8 and 22.7 6 12.9% for groups IR5–36, IR5–32 and IR5–28, respectively (Fig. 4D).

Neuropathology The frequency of fibre degeneration varied with the duration of ischaemia, the segment of peripheral nerve and temperature. The pathological changes consisted of fibre degeneration associated with nerve oedema. Axonal changes involving swollen watery axons, dark axons and shrunken axons, were accompanied by different degrees of myelin breakdown. Ischaemic fibre degeneration was most pronounced at the mid-tibial and distal tibial level in the IR3 and IR5 groups. The most profound pathological alterations were seen in the IR-36 group, in which nearly all myelinated fibres had undergone axonal damage. Axons with very thin myelin were also observed, but these were always accompanied by axonal changes. In contrast, IR3–28 was almost completely protected from ischaemia–reperfusion

injury. The pathology of group IR5 was consistently more severe than that of group IR3 at the same temperature (Figs 5 and 6). Figure 5 shows representative transverse sections of the mid-tibial nerve after 3 h of ischaemia followed by 1 week of reperfusion. Degenerated fibres were predominant for the 37°C treatment, and small number of degenerated fibres were seen for the 32°C treatment. Most fibres looked normal after treatment at 28°C. In Fig. 6, which shows representative transverse sections of the mid-tibial nerve after 5 h of ischaemia–reperfusion, most fibres are degenerating; normal fibres were seldom seen after treatment at 36°C. Degenerated fibres constituted approximately half of all fibres after treatment at 32°C; in contrast, few fibres were degenerated after treatment at 28°C. The IFD and oedema indices are summarized in Figs 7 and 8. After 3 h of ischaemia, IFD was confined to the tibial nerve. For both tibial nerve levels, excellent neuroprotection was seen in the 28°C group. For the mid-tibial level, there was temperature-dependent neuroprotection from IFD from 37 to 32 to 28°C, each step being significant (P , 0.001). For 5 h of ischaemia, neuroprotection was less at each temperature than for 3 h of ischaemia. Changes were also confined to the tibial nerve. For the mid-tibial nerve there was temperature-dependent neuroprotection, nerves treated at 36°C showing significantly more IFD than those treated

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Fig. 4 Electrophysiology for 3 h (A and B) and 5 h (C and D) of ischaemia. The amplitudes SAP (A and C) and CMAP (B and D) are expressed as percentages of the contralateral (non-ischaemic) side. *P , 0.05; **P , 0.01.

Hypothermic neuroprotection of nerve

165

Fig. 6 Light-microscope findings in transverse sections of the mid-tibial nerve after 5 h of ischaemia. Scale bar 5 100 µm.

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Fig. 5 Representative transverse sections of the mid-tibial nerve after 3 h of ischaemia–reperfusion. Most fibres degenerated when treated at 37°C; fewer degenerated fibres were seen after treatment at 32°C. Most fibres were normal after treatment at 28°C. Scale bar 5 100 µm.

166

Y. Mitsui et al.

at 28°C (P , 0.001) or 32°C (P , 0.01). Hypothermia at 28°C conferred neuroprotection over 32°C (P , 0.05). For the distal tibial nerve, there was also temperature-dependent neuroprotection, reaching statistical significance in the group treated at 28°C (P , 0.01). The degree of endoneurial oedema was more pronounced after 5 h of ischaemia (Fig. 8B) than after 3 h (Fig. 8A). A temperature-dependent reduction in oedema was observed for the 3 h treatment, and 28°C conferred significant reduction of oedema in both the mid-tibial (P , 0.01) and the distal tibial (P , 0.001) nerve. For the 5 h ischaemia groups, hypothermia did not confer any consistent benefits in the reduction of endoneurial oedema (Fig. 8B).

Discussion The main finding of the present study is that hypothermia (28°C) provided dramatic neuroprotection of peripheral nerve from ischaemia–reperfusion injury in the 3 h ischaemia model, returning all behavioural, electrophysiological and pathological indices to essentially normal function. The greater stress of 5 h of ischaemia resulted in more fibre degeneration, oedema, functional impairment and neurophysiological deficit and less complete hypothermic neuroprotection. Mild hypothermia (32°C) also protected peripheral nerve from ischaemia–reperfusion injury, but less completely than hypothermia at 28°C for both 3 and 5 h of ischaemia. For studies of nerve salvage from IFD, it was necessary to develop a model of significant but reversible ischaemia. Although a number of animal models of peripheral nerve

ischaemia have been developed (Korthals and Wisniewski, 1975; Hess et al., 1979; Parry and Brown, 1981; Sladky et al., 1985; Nukada et al., 1993), these models were primarily developed to evaluate the pathology of ischaemia rather than of neuroprotection, and the reproducibility of the model was not of primary concern. Our model of ligation with splitknot ties of the supplying arteries of the sciatic–tibial nerve, permitting ready release and reperfusion, provides an optimal model of ischaemia–reperfusion injury of peripheral nerves. We have demonstrated that reperfusion following nerve ischaemia results in a large increase in nerve hydroperoxides (Nagamatsu et al., 1996) and breakdown of the blood–nerve barrier (Schmelzer et al., 1989). The time point of 7 days was chosen following the results of our previous study (Kihara et al., 1993, 1996; Nagamatsu et al., 1996). We have found that this time point is optimal in demonstrating the neuropathological alterations, and that nerve blood flow alterations at this time-point correlate well with neuropathology (Michenfelder and Theye, 1968). Instead of 37°C, 36°C was chosen for 5 h of ischaemia because the combination of 5 h of ischaemia and 37°C caused a high level of mortality. At 36°C, all rats but one survived for 1 week. In the IR3 model, hypothermia (28°C) resulted in dramatic salvage from both IFD (IFD index) and blood–nerve barrier disruption (oedema index). With the longer duration of ischaemia in the IR5 model, the IFD but not the oedema index showed significant improvement, indicating that IFD could be prevented even when the nerve had been subjected to the major insult of 5 h of ischaemia. The dissociation of the IFD and oedema indices is of interest, indicating that

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Fig. 7 Ischaemic fibre degeneration index for 3 h (A) and 5 h (B) of ischaemia–reperfusion. *P , 0.05; **P , 0.01; ***P , 0.001.

Hypothermic neuroprotection of nerve

167

hypothermia protected nerve fibres from breaking down, even though the blood–nerve barrier was disrupted. This observation suggests that hypothermia may operate on mechanisms distal to the entry of plasma products. Scant information is available on the molecular mechanisms of nerve ischaemia. Hypoxia/ischaemia results in an increase in tissue reducing equivalents (Neely and Feuvray, 1981; Low et al., 1985) and an increase in tissue xanthine (Jennings and Reimer, 1981). Ischaemia or hypoxia has been suggested to activate a calcium-dependent protease (possibly calpain) which converts cytosolic xanthine dehydrogenase to xanthine oxidase (Battelli et al., 1972). This important reaction occurs in the capillary endothelium (Jarasch et al., 1981). The above three conditions create an oxygen-free radical-generating system. Furthermore, reperfusion accelerates the formation of reduced oxygen species by introducing O2 into a system primed and generating reduced oxygen species (McCord, 1985). Using our model, we were able consistently to produce ischaemic conduction failure within 30 min of ischaemia, with prompt recovery of impulse transmission following 1 h of ischaemia but persistent

conduction failure and conduction block following 3 h of ischaemia. Reperfusion resulted in an increase in the permeability 3 surface area product, indicating a breakdown of the blood–nerve barrier that was evident in peripheral nerve after 1 and, especially, 3 h of ischaemia (Schmelzer et al., 1989). Nerve blood flow was not completely restored with reperfusion and there was a suggestion of progressive reduction in nerve blood flow with increasing duration of reperfusion (Schmelzer et al., 1989). Reperfusion also resulted in an increase in lipid hydroperoxides in the sciatic nerve (Nagamatsu et al., 1996). Ischaemia results in a massive increase in cellular calcium (Cheung et al., 1986), leading to the suggestion that a key pathogenetic mechanism of ischaemia is cytosolic calciummediated activation of phospholipases and the production of free fatty acids and lysophospholipids. Calcium may potentiate the toxicity of reduced oxygen species, and the beneficial effects of calcium channel blockers may be mediated by blocking this potentiating effect and by reducing catecholamine-induced vasoconstriction (Cheung et al., 1986). There is some evidence that similar ischaemic

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Fig. 8 Oedema index for 3 h (A) and 5 h (B) of ischaemia. The changes were seen mainly in the tibial nerves. *P , 0.05; **P , 0.01; ***P , 0.001.

168

Y. Mitsui et al. ischaemic non-reperfused nerves that had been rendered ischaemic with microemboli to nerve (Kihara et al., 1996). Although reperfusion is deleterious, its role appears to be most important in mild or moderate ischaemia. When ischaemia is severe or protracted, as occurs in the tibial nerve, oedema and fibre degeneration is as severe in nonreperfused as in reperfused nerves (Nagamatsu et al., 1996). A related observation is that, although reperfusion is injurious, the beneficial effects of oxygenation, applied early, outweigh the effects of reperfusion, as demonstrated in our study of nerve neuroprotection using hyperbaric oxygenation (Kihara et al., 1995). Peripheral nerve may have several advantages over brain in neuroprotection. Our study revealed that a deeper degree of hypothermia (28°C) had a greater protective effect against ischaemia–reperfusion injury than mild hypothermia (32°C). Mild hypothermia (around 30–34°C) was usually chosen for studies of hypothermia in brain ischaemia or closed injury, because a deeper degree of hypothermia caused cardiac electrical irritability, electrolyte imbalance and pulmonary infections. Technically, limb cooling is relatively easier than hypothermia of the brain, so our method and deeper hypothermia might have clinical applications. Secondly, the therapeutic window may be significantly larger in nerve than in brain, since nerve has lower energy demands, functions well on anaerobic metabolism and has relatively high energy stores. In our experiments, SAP was generally better preserved than CMAP. Since CMAP is affected not only by nerve function but also by muscle function, this difference probably reflects the greater susceptibility of muscle than nerve to ischaemia (Zollman et al., 1991) when subjected to ischaemia–reperfusion. Hypothermia is known to have a protective effect on contractile force in skeletal muscles (Bolognesi et al., 1996), so that the improved behavioural score and CMAP reflects, in part, the salvage of muscle function. In summary, peripheral nerve subjected to ischaemia– reperfusion undergoes a breakdown in the blood–nerve barrier and IFD. Ischaemic fibres can be rescued from IFD by hypothermia. Deep hypothermia is more effective than mild hypothermia. Even after 5 h of ischaemia, peripheral nerve can be rescued. Limb cooling, a relatively easy technique, might be a potent therapeutic method for the treatment of ischaemia–reperfusion injury of peripheral nerves.

Acknowledgements We wish to thank Ms Anita Zeller for excellent secretarial assistance. This research was supported in part by grants from NINDS (NS22352), MDA, and Mayo funds. References Battelli MG, Corte ED, Stirpe F. Xanthine oxidase type D (dehydrogenase) in the intestine and other organs of the rat. Biochem J 1972; 126: 747–9.

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

mechanisms are operative in peripheral nerves. Calcium ionophore will cause vesicular disruption of myelin (Schlaepfer, 1977b) and axonal degeneration (Schlaepfer, 1977a), and nerve reconnection in a calcium-free medium resulted in improved functional recovery (de Medinaceli et al., 1983). Excitotoxic injury, considered important in the brain, is unlikely to be important in nerve trunks, which lack NMDA (N-methyl-D-aspartate) receptors. We have proposed that the molecular mechanisms of nerve ischaemia are similar to those of tissues like heart, gut and brain, but modified in terms of their threshold to ischaemic and reperfusion damage (Schmelzer et al., 1989). Ischaemia would result in phospholipase activation and phospholipid breakdown, liberating arachidonic acid with its cascade producing the prostaglandins and leucotrienes. Reduced oxygen species would be generated during ischaemia and especially during reperfusion. The formation of lipid hydroperoxides (Nagamatsu et al., 1996) would inhibit prostacyclin synthetase (Low et al., 1989). The increased biosynthetic rate of TxA2 coupled with the ischaemia-related inhibition of prostacyclin production by endothelial cells would result in vasoconstriction, aggravating the ischaemic insult (Low et al., 1989). Our results suggest that reperfusion injury may happen in nerves if the ischaemia is nearly total and of long duration, since it occurs following 3 h but not 1 h of occlusion. The generation of reduced oxygen species damages the endothelial barrier (Schmelzer et al., 1989), resulting in endoneurial oedema (Nagamatsu et al., 1996) and, subsequently, IFD. While reperfusion assures nerve tissues an ample resupply of oxygen and nutrients, our previous study strongly indicated that reperfusion resulted in a large increase in lipid hydroperoxides (Nagamatsu et al., 1996) and breakdown of the blood–nerve barrier (Schmelzer et al., 1989). Hypothermia reduced the severity of endoneurial oedema in nerves subjected to 3 h of ischaemia–reperfusion (IR3), suggesting that hypothermia will reduce the generation of reduced oxygen species. Part of the explanation may relate to the effect of temperature on enzymatic activity, including nitric oxide synthase and the generation of superoxide radicals. Peripheral nerve is unique metabolically. It has a very low metabolic rate and the ability to function relatively well on anaerobic metabolism (Low et al., 1989). Compared with rat brain, nerve tissue has ~10% of the oxygen requirement but similar energy stores. When an ample supply of glucose is coupled with a reduction in temperature, peripheral nerve can conduct for many hours (Fink and Cairns, 1982). Hypothermia, which reduces oxygen consumption and glucose utilization, could result in significant neuroprotection, especially if nerve energy substrates are intact. The efficacy of hypothermia in IR5 nerves when oedema is unaffected but neuroprotection is provided suggests that mechanisms beyond an impairment of the blood–nerve barrier may be operative. Additional support derives from an earlier study where we demonstrated hypothermic neuroprotection in

Hypothermic neuroprotection of nerve Bolognesi MP, Chen LE, Seaber AV, Urbaniak JR. Protective effect of hypothermia on contractile force in skeletal muscle. J Orthop Res 1996; 14: 390–5. Cheung JY, Bonventre JV, Malis CD, Leaf A. Calcium and ischemic injury. [Review]. N Engl J Med 1986; 314: 1670–6. de Medinaceli L, Wyatt RJ, Freed WJ. Peripheral nerve reconnection: mechanical, thermal, and ionic conditions that promote the return of function. Exp Neurol 1983; 81: 469–87. Fink BR, Cairns AM. Differential tolerance of mammalian myelinated and unmyelinated nerve fibers to oxygen lack. Reg Anesth 1982; 7: 2–6. Hess K, Eames RA, Darveniza P, Gilliatt RW. Acute ischaemic neuropathy in the rabbit. J Neurol Sci 1979; 44: 19–43.

169

failure and energy metabolism in experimental diabetic neuropathy. Am J Physiol 1985; 248: E457–62. Low PA, Lagerlund TD, McManis PG. Nerve blood flow and oxygen delivery in normal, diabetic, and ischemic neuropathy. [Review]. Int Rev Neurobiol 1989; 31: 355–438. McCord JM. Oxygen-derived free radicals in postischemic tissue injury. [Review]. N Engl J Med 1985; 312: 159–63. Michenfelder JD, Theye RA. Hypothermia: effect on canine brain and whole-body metabolism. Anesthesiology 1968; 29: 1107–12. Nagamatsu M, Schmelzer JD, Zollman PJ, Smithson IL, Nickander KK, Low PA. Ischemic reperfusion causes lipid peroxidation and fiber degeneration. Muscle Nerve 1996; 19: 37–47. Neely JR, Feuvray D. Metabolic products and myocardial ischemia. [Review]. Am J Pathol 1981; 102: 282–91. Nukada H, Powell HC, Myers RR. Spatial distribution of nerve injury after occlusion of individual major vessels in rat sciatic nerves. J Neuropathol Exp Neurol 1993; 52: 452–9.

Jennings RB, Reimer KA. Lethal myocardial ischemic injury. [Review]. Am J Pathol 1981; 102: 241–55.

Parry GJ, Brown MJ. Arachidonate-induced experimental nerve infarction. J Neurol Sci 1981; 50: 123–33.

Kihara M, Zollman PJ, Schmelzer JD, Low PA. The influence of dose of microspheres on nerve blood flow, electrophysiology, and fiber degeneration of rat peripheral nerve. Muscle Nerve 1993; 16: 1383–9.

Parry G, Kozu H. Piroxicam may reduce the rate of progression of experimental diabetic neuropathy. Neurology 1990; 40: 1446–9.

Kihara M, McManis PG, Schmelzer JD, Kihara Y, Low PA. Experimental ischemic neuropathy: salvage with hyperbaric oxygenation. Ann Neurol 1995; 37: 89–94.

Schlaepfer WW. Vesicular disruption of myelin simulated by exposure of nerve to calcium ionophore. Nature 1977b; 265: 734–6.

Kihara M, Schmelzer JD, Kihara Y, Smithson IL, Low PA. Efficacy of limb cooling on the salvage of peripheral nerve from ischemic fiber degeneration. Muscle Nerve 1996; 19: 203–9. Korthals JK, Wisniewski HM. Peripheral nerve ischemia. Part 1. Experimental model. J Neurol Sci 1975; 24: 65–76. Low PA, Tuck RR. Effects of changes of blood pressure, respiratory acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J Physiol (Lond) 1984; 347: 513–24. Low PA, Ward K, Schmelzer JD, Brimijoin S. Ischemic conduction

Schlaepfer WW. Structural alterations of peripheral nerve induced by the calcium ionophore A23187. Brain Res 1977a; 136: 1–9.

Schmelzer JD, Zochodne DW, Low PA. Ischemic and reperfusion injury of rat peripheral nerve. Proc Natl Acad Sci USA 1989; 86: 1639–42. Sladky JT, Greenberg JH, Brown MJ. Regional perfusion in normal and ischemic rat sciatic nerves. Ann Neurol 1985; 17: 191–5. Zollman PJ, Awad O, Schmelzer JD, Low PA. Effect of ischemia and reperfusion in vivo on energy metabolism of rat sciatic–tibial and caudal nerves. Exp Neurol 1991; 114: 315–20. Received June 17, 1998. Revised August 20, 1998. Accepted September 7, 1998

Downloaded from http://brain.oxfordjournals.org/ by guest on December 23, 2016

Jarasch ED, Grund C, Bruder G, Heid HW, Keenan TW, Franke WW. Localization of xanthine oxidase in mammary-gland epithelium and capillary endothelium. Cell 1981; 25: 67–82.