Combined therapy (Rho-A-kinase inhibitor and chitosan/collagen porous scaffold) provides a supportive environment for endogenous regenerative processes after spinal cord trauma.

Due to the complexity of pathological processes in spinal cord injury (SCI), it is increasingly recognized that combined strategies are more effective than single treatments. The aim of the present study was to enhance neural tissue regeneration and axon regrowth using Rho-A-kinase inhibitor (Y-27632) in a rat SCI model (Th9 compression) and to bridge the lesion with a chitosan/collagen porous scaffold (ChC-PS) applied two weeks after SCI. In addition, to see the synergic effect of Y-27632 and ChC-PS, we combined these single therapeutic strategies to enhance the regenerative capacity of injured spinal cord tissue. The animals survived eight weeks. Application of Y-27632 modulated the inhibitory milieu by specifically targeting gray and white matter integrity, glial fibrillary acidic protein (GFAP)-immunoreactivity, and the outgrowth of neurofilaments and growth-associated protein-43 (GAP-43) immunoreactive axons across the lesion sites, leading to significant positive functional outcome from day 20 to 56. Compared to single treatments, combined Y-27632/ChC-PS therapy was more effective in neurofilaments and GAP-43 expression and GFAP immunoreactivity in the perilesional area of dorsal, lateral and ventral columns, and in enhancing the gray and white matter integrity throughout the cranio-caudal extent. The findings indicate that combined therapy provides a supportive environment for endogenous regenerative processes.

Survival and differentiation of human embryonic stem cell-derived neural precursors grafted spinally in spinal ischemia-injured rats or in naive immunosuppressed minipigs: a qualitative and quantitative study.

In previous studies, we have demonstrated that spinal grafting of human or rat fetal spinal neural precursors leads to amelioration of spasticity and improvement in ambulatory function in rats with spinal ischemic injury. In the current study, we characterize the survival and maturation of three different human embryonic stem (ES) cell line-derived neural precursors (hNPCs) once grafted into ischemia-injured lumbar spinal cord in rats or in naive immunosuppressed minipigs. Proliferating HUES-2, HUES-7, or HUES-9 colonies were induced to form embryoid bodies. During the nestin-positive stage, the rosettes were removed and CD184(+)/CD271(-)/CD44(-)/CD24(+) population of ES-hNPCs FAC-sorted and expanded. Male Sprague-Dawley rats with spinal ischemic injury or naive immunosuppressed Gottingen-Minnesota minipigs received 10 bilateral injections of ES-NPCs into the L2-L5 gray matter. After cell grafting, animals survived for 2 weeks to 4.5 months, and the presence of grafted cells was confirmed after staining spinal cord sections with a combination of human-specific (hNUMA, HO14, hNSE, hSYN) or nonspecific (DCX, MAP2, CHAT, GFAP, APC) antibodies. In the majority of grafted animals, hNUMA-positive grafted cells were identified. At 2-4 weeks after grafting, double-labeled hNUMA/DCX-immunoreactive neurons were seen with extensive DCX(+) processes. At survival intervals of 4-8 weeks, hNSE(+) neurons and expression of hSYN was identified. Some hSYN-positive terminals formed putative synapses with the host neurons. Quantitative analysis of hNUMA(+) cells at 2 months after grafting showed comparable cell survival for all three cell lines. In the presence of low-level immunosuppression, no grafted cell survival was seen at 4.5 months after grafting. Spinal grafting of proliferating pluripotent HUES-7 cells led to consistent teratoma formation at 2-6 weeks after cell transplantation. These data show that ES-derived, FAC-sorted NPCs can represent an effective source of human NPCs to be used in CNS cell replacement therapies.

Involvement of group I metabotropic glutamate receptors and glutamate transporters in the slow excitatory synaptic transmission in the spinal cord dorsal horn.

Our experiments demonstrate a novel role for group I metabotropic glutamate receptor (mGluR) subtypes 1 and 5 in generating a long-lasting synaptic excitation in the substantia gelatinosa (SG) and deep dorsal horn (DH) neurons of the rat spinal cord. In the present study we have investigated a slow excitatory postsynaptic current (EPSC), elicited by a brief high intensity (at Adelta/C fiber strength) and high frequency (20 or 100 Hz) stimulation of primary afferent fibers (PAFs) using whole-cell patch-clamp recordings from neurons located in the DH (laminae II-V) in spinal cord slices of young rats and wild-type and gene-targeted mice lacking mGluR1 subtype. The results shown here suggest that the activation of both mGluR1 and mGluR5 along with NK1 receptors, may be involved in the generation of the slow EPSC in the spinal cord DH. Inhibition of glial and neuronal glutamate transporters by DL-threo-beta-benzyloxyaspartate (TBOA) enhanced the group I mGluR-dependent slow EPSC about eightfold. Therefore, we conclude, that glutamate transporters strongly influence the group I mGluR activation by PAFs possibly at sensory synapses in the DH. Overall these data indicate that stimulus trains can generate a sustained and widespread glutamate signal that can further elicit prolonged EPSCs predominantly mediated by the group I mGluRs. These slow excitatory synaptic currents may have important functional implications for DH cell firing and synaptic plasticity of sensory transmission, including nociception.

Development of GABA-sensitive spasticity and rigidity in rats after transient spinal cord ischemia: a qualitative and quantitative electrophysiological and histopathological study.

Transient spinal cord ischemia may lead to a progressive degeneration of spinal interneurons and subsequently to increased hind limb motor tone. In the present work we sought to characterize the rigidity and spasticity components of this altered motor function by: i) tonic electromyographic activity measured in gastrocnemius muscle before and after ischemia, ii) measurement of muscle resistance during the period of ankle flexion and corresponding changes in electromyographic activity, iii) changes in Hoffmann reflex, and, iv) motor evoked potentials. In addition the effect of intrathecal treatment with baclofen (GABAB receptor agonist; 1 microg), nipecotic acid (GABA uptake inhibitor; 300 microg) and dorsal L2-L5 rhizotomy on spasticity and rigidity was studied. Finally, the changes in spinal choline acetyltransferase (ChAT) and vesicular glutamate transporter 2 and 1 (VGLUT2 and VGLUT1) expression were characterized using immunofluorescence and confocal microscopy. At 3-7 days after ischemia an increase in tonic electromyographic activity with a variable degree of rigidity was seen. In animals with modest rigidity a velocity-dependent increase in muscle resistance and corresponding appearance in electromyographic activity (consistent with the presence of spasticity) was measured during ankle rotation (4-612 degrees /s rotation). Measurement of the H-reflex revealed a significant increase in Hmax/Mmax ratio and a significant loss of rate-dependent inhibition. In the same animals a potent increase in motor evoked potential amplitudes was measured and this change correlated positively with the increased H-reflex responses. Spasticity and rigidity were consistently present for a minimum of 3 months after ischemia. Intrathecal treatment with baclofen (GABA B receptor agonist) and nipecotic acid (GABA uptake inhibitor) provided a significant suppression of spasticity, rigidity, H-reflex or motor evoked potentials. Dorsal L2-L5 rhizotomy significantly decreased muscle resistance but had no effect on increased amplitudes of motor evoked potentials. Confocal analysis of spinal cord sections at 8 weeks-12 months after ischemia revealed a continuing presence of ChAT positive alpha-motoneurons, Ia afferents and VGLUT2 and VGLUT1-positive terminals but a selective loss of small presumably inhibitory interneurons between laminae V-VII. These data demonstrate that brief transient spinal cord ischemia in rat leads to a consistent development of spasticity and rigidity. The lack of significant suppressive effect of dorsal L2-L5 rhizotomy on motor evoked potentials response indicates that descending motor input into alpha-motoneurons is independent on Ia afferent couplings and can independently contribute to increased alpha-motoneuronal excitability. The pharmacology of this effect emphasizes the potent role of GABAergic type B receptors in regulating both the spasticity and rigidity.

A simple and reproducible model of spinal cord injury induced by epidural balloon inflation in the rat.

This paper describes a modification of a balloon-compression technique to produce spinal cord injury in adult rats. A 2-French Fogarty catheter is inserted into the dorsal epidural space through a small hole made in T10 vertebral arch, advanced cranially to T8-9 spinal level, and inflated for 5 min. Spinal cord damage is graded by increasing the volume of saline used to inflate the balloon. Quantitative neurological and histopathological outcomes are presented with three different volumes (10, 15, and 20 microl of saline) to characterize the gradation of injury. Volume of 15 microl produced complete paraplegia followed by gradual recovery, finally reaching approximately the middle of the scale used to quantitate the locomotor performance. With these animals, after 4 weeks, the center of the lesion shows complete loss of grey matter and partial sparing of the white matter. We conclude that 15 microl volume produced submaximal injury that will be useful for studying the pathophysiology and effects of protective therapies with this compression-injury model.

A rare case of a somatosensory evoked potentials recording during cardiopulmonary resuscitation.

BACKGROUND: Neuronal damage is a possible complication of cardiac surgery. To reduce the potential risk of postoperative neurological deficit, the functional state of affected central nervous system pathways is monitored intraoperatively by recording evoked potentials (EPs). Apart from animal research, there is little clinical evidence of EPs recording during cardiac arrest and cardio-pulmonary resuscitation (CPR). METHODS: Both scalp (SCEPs) and spinal (SSEPs) short-latency somatosensory EPs were recorded as the response to the electrical stimulation of the right median nerve during mitral valve replacement surgery. Evoked potentials were recorded before, during, and after sudden ventricular fibrillation followed by CPR. RESULTS: Preoperative control recordings of both SCEPs and SSEPs were in the normal ranges. During the first 4 min of cardiac arrest and resuscitation, all SCEPs waves disappeared, while the spinal component of the SSEPs was still recognizable. After CPR, all waves of both EPs recordings recovered completely. The patient woke from anesthesia without neurological deficits. CONCLUSIONS: As expected, scalp-recorded EPs are more sensitive to the cardiac arrest than spinal EPs. Rapid and almost complete recovery of postoperative EPs, namely SCEPs, correlated well with normal neurological recovery.

Technique of selective spinal cord cooling in rat: methodology and application.

In a number of interventions, it is desirable to be able to produce a rapid but readily reversible change in spinal cord temperature (SCT) without altering general body temperature and to maintain this selective spinal cord hypothermia stable for an extended interval. To accomplish this, we developed a technique of subcutaneous perfusion cooling in rat. This was accomplished by constructing a copper heat exchanger which was readily implanted into subcutaneous space overlying the upper thoracic to upper sacral spinal segments. The heat exchanger was then perfused with fluid from an external temperature bath maintained at (8 degrees C) at a perfusion rate of 100 ml/min. The temperature of the heat exchanger was controlled by regulating the pump with a feed back controller driven by a thermocouple placed percutaneously into the paraspinal musculature. A series of studies were performed to demonstrate the characteristics and utility of this cooling technique. Lowering the pump set point to 24 degrees C resulted in a fall in the intrathecal temperature (ITT) to 27 +/- 0.3 degrees C within 15 min with no significant changes observed in rectal temperature (37.5- > 37.2 degrees C). Change in intrathecal temperature showed a highly significant correlation with changes in paravertebral muscle temperature (r = 0.977). The hypothermic state could be readily maintained for extended intervals up to 5 h and an underbody heating pad was used to maintain rectal temperature between 35-36.5 degrees C. Lowering the ITT from 37 degrees C-27 degrees C evoked a temperature-dependent increase in the latency of precooling spinal somatosensory evoked potentials (SSEPs) with the highest sensitivity observed in postsynaptic components. Returning the set point temperature back to 37 degrees C produced a rapid recovery of the SSEPs latencies. Consistent with previously published data, selective spinal cord hypothermia (27 degrees C) provided complete protection against otherwise injurious interval of normothermic ischemia produced by balloon occlusion of the descending aorta. This technique provides a simple, relatively non-invasive and reliable experimental tool for studying the effect of selective, acute and/or prolonged spinal cord hypothermia.

Changes in spontaneous unit activity in lumbar spinal cord after reversible aortic occlusion in the rat.

Lumbar dorsal horn neuronal multiunit activity was recorded in the halothane anesthetized rat before, during, and after a 30 min interval of spinal cord ischemia induced by inflation of a balloon catheter inserted through the femoral artery. After initiation of ischemia, there was typically a brief burst of neuronal activity, followed by electrical silence. During reperfusion, there was a progressive increase in discharge activity over the ensuing 30 min. Spontaneous activity was significantly increased, as measured by the number of discriminable cells and frequency of discharges. Post-ischemic hyperactivity reached maximal values typically after 60-90 min of reperfusion. The character and time course of recorded activity are consistent with the behavioral sequelae of transient spinal ischemia.

Epidural perfusion cooling protects against spinal cord ischemia in rabbits. An evaluation of cholinergic function.

The protective effect of regional epidural spinal cord cooling was evaluated in a rabbit spinal cord ischemia model. Hypothermia was performed by the continual perfusion of 2-4 degrees C cold saline in the epidural space around the ischemic lumbar segments, 4 min before and during ischemia. The spinal cord was deeply hypothermic (21 degrees C) throughout the whole ischemic period. Ischemia was induced by the occlusion of the abdominal aorta for 40 min under normothermic or hypothermic conditions. Recovery of motor and sensory functions, spinal cord-evoked potentials, and motor-evoked potentials were then evaluated up to 24 h postischemia. After this period, choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) activities were measured, in particular, zones of the lumbar spinal cord. AChE was also investigated histochemically. Animals in the normothermic group displayed fully developed spastic paraplegia with near complete loss of spinal somatosensory and motor-evoked potentials. AChE histochemistry showed extensive necrotic changes affecting lumbosacral gray matter. These changes corresponding with the pronounced losses of ChAT and AChE activities indicated irreversible injury of the spinal cord. In contrast, after hypothermic ischemia, animals survived without any sign of neurological impairment with almost full recovery of the spinal cord-evoked potentials. ChAT and AChE activities in the gray matter showed near control values corresponding with histochemical analysis of fully preserved gray matter. Hypothermia under the present experimental conditions efficiently protected the spinal cord against ischemic injury.

Unilateral grafting of fetal neocortex into a cortical cavity improves healing of a symmetric lesion in the contralateral cortex of adult rats.

Fetal neocortical tissue (ED 14) was grafted unilaterally into a cortical cavity made bilaterally in the sensorimotor cortex of adult rats. Transplantation was done immediately after the lesion (group TR0, n = 8) or with 14-day delay (group TR14, n = 8). Six rats served as lesion only controls (group LES). After long-term survival (up to 15 months) the brains were photographed and surface areas of transplant and contralateral cavities were measured by means of a graphic tablet. The results show that (a) the presence of a transplant in one lesion cavity in the cortex decrease the size of a similar cavity in the contralateral cortex and that (b) the better host transplant integration there is, the greater the effect on the contralateral lesion. No correlation between the size of the transplant and the size of the symmetric traumatic lesion was found. The ameliorating effect of the transplant on the contralateral cortical lesion size is most likely related the long-term influence of growth of trophic factors released by transplanted cells which lead to the healing of the symmetric lesion.

Sciatic nerve stimulation increases the degree of histopathological damage in lumbosacral segments after short lasting spinal cord ischemia in rabbit.

To characterize the influence of primary afferent activation on the development of histopathological changes in the spinal cord after reversible ischemia, the left sciatic nerve was stimulated at the intensity of myelinated fibers before, during and for 1 h after 10 min of abdominal aortic ligation in halothane anesthetized rabbits. In control animals, only 10 min of spinal cord ischemia or sciatic nerve stimulation was employed. One hour after reperfusion all animals were perfusion fixed with 4% paraformaldehyde. Histopathological analysis using the suppressive Nauta method revealed significantly higher number of argyrophilic neurons in dorsal horns and in the intermediate zone in animals in which spinal cord ischemia was combined with sciatic nerve stimulation in comparison with the animals with spinal cord ischemia, but without stimulation. These histopathological changes corresponded with signs of irreversible damage analyzed on the ultrastructural level. Stimulation of sciatic nerve, but without ischemia did not evoke any detectable neuronal changes. The data from the present study suggest that increased activity of spinal cord neurons evoked by peripheral nerve stimulation can be an important factor in determining the extent of irreversible damage after short lasting ischemia.

Ischemia-induced delayed-onset paraplegia is accompanied by an unusual form of synaptic degeneration in the lumbosacral segments: an experimental light and electron microscopic study in dogs.

We studied the effect of high thoracic aorta cross-clamping, complete transverse section of the spinal cord at Th6 level, and combined hemisection at Th6 level followed later by high thoracic aorta cross-clamping upon the morphology and number of identified presynaptic knobs in lumbosacral segments in dogs. In animals surviving 48-72 hours after high thoracic aorta cross-clamping the occurrence of an unusual form of boutons accompanied by periboutonal halo in L3-S1 segments was found. According to the bouton size and light as well as electron microscopic appearance, four types, i.e., light giant (T1), dark enlarged (T2), light giant with periboutonal halo (T3), and giant disintegrating (T4) boutons were detected after 48 and 72 hour reperfusion. The appearance of four boutonal types in the lumbosacral segments is caused by spinal cord ischemia secondary to high thoracic aorta cross-clamping followed by 48 or 72 hour reperfusion. At the end of the sixth reperfusion day no signs of enlarged and giant boutons were detected in L3-S1 segments. A statistically significant increase of enlarged and giant boutons was noted at the end of the third reperfusion day in comparison with 48 hour survival. After spinal cord transection at midthoracic (Th6) level, followed by 72 hour survival, no such unusual synaptic knobs could be found in L3-S1 segments. The laminar distribution pattern of T1-T4 types based on light microscopic analysis and confirmed electron microscopically is characteristic and strictly bound to those spinal cord gray matter layers which serve as main termination sites of the descending cortical, brain stem, as well as long propriospinal projections in the lumbosacral segments (laminae V-VII). A statistically significant increase of enlarged and giant boutons was found in the intermediate zone (lamina VII). Hemisection at midthoracic level (Th6) followed later by 30 minute high thoracic aorta cross-clamping and 48 hour reperfusion caused a marked decrease of enlarged and giant boutons in L3-S1 segments on the hemisectioned side in comparison with the contralateral one. Large amounts of irregularly arranged round vesicles and tubular profiles were disclosed in the boutonal matrix of T1, T3, and T4 types in L3-S1 segments of animals subjected to 30 minute high thoracic aorta cross-clamping followed by 72 hour reperfusion. Accumulation of tubular and membranous materials was invariably seen in the bulbous enlargement of the terminal axonal branch.(ABSTRACT TRUNCATED AT 400 WORDS)

Epidural perfusion cooling protection against protracted spinal cord ischemia in rabbits.

The protective effect of a modified epidural cooling technique was assessed in a rabbit spinal cord ischemia model. The epidural space around the lumbar segments with induced ischemia was continually perfused with cold (5 degrees C) isotonic saline via two communicating spinal canal openings. This procedure allowed the spinal cord to be kept deeply hypothermic (< 15 degrees C within central gray matter) during the ischemic period. The animals were subjected to either normothermic ischemia (Group A) or hypothermic ischemia (Group B). Each group contained three subgroups of animals undergoing 20, 40, or 60 minutes of aortic ligation. Their neurological outcomes were evaluated up to 48 hours postischemia, and the intergroup differences were compared. Two days postischemia, all of the animals were sacrificed by transcardial perfusion-fixation and their lumbar segments were processed for histopathological examination. In addition, in animals with 60-minute ischemia, spinal somatosensory evoked potentials were recorded during surgical intervention and again after 48 hours. In the normothermic animals, a high incidence of paraplegia was detected: in 40% after 20 minutes of ischemia, in 75% after 40 minutes, and in 100% after 60 minutes. In contrast, all of the hypothermic animals exhibited full neurological recovery even after 60 minutes of ischemia. Both electrophysiological and histological observations clearly correlated with the neurological findings. The results suggest that deep spinal cord hypothermia produced by epidural perfusion cooling provides effective protection against protracted spinal cord ischemia in rabbits.

Panmyelic epidural cooling protects against ischemic spinal cord damage.

The neuroprotective effect of epidural cooling before and during spinal cord ischemia on the neurological, neurophysiological, and histopathological outcome was evaluated after 40 min of proximal and distal thoracic aorta crossclamping in dogs. In the normothermic group (n = 12), no attempt was made to change the spinal cord temperature. Four of eight animals showed complete paraplegia and four had partial recovery. The N3 component of spinal somatosensory-evoked potentials recovered to only 11.7 +/- 1.4% after 2 hr of recirculation and to 45% of control value after 2 days of survival. In the transverse sections taken from L1-L7 segments, apparent interneuronal damage in the intermediate zone was found after 2 hr of reperfusion followed by a heavy loss of interneurons after 2 days of survival and functionally defined as fully developed paraplegia. In the hypothermic group (n = 12), the spinal cord temperature was lowered 3 min before aortic crossclamping with a bolus of epidurally administered 0.9 N saline solution (8 ml/kg at 5 degrees C) to 28.5 +/- 1.3 degrees C and was maintained throughout the crossclamping time with the additional infusion of the same solution (20 ml/kg/40 min) using a peristaltic pump. Seven of eight animals had no neurological deficit and one animal showed partial recovery, which was significantly better than the motor score for the normothermic group (P < 0.05). The SSEP revealed 55% of postsynaptic (N3) wave recovery after 2 hr of recirculation and 92% recovery after 2 days survival, which was significantly higher than those for the normothermic animals (P < 0.05). Histological analysis showed almost full protection of interneurons and A-motoneurons verified after 2 hr and 2 days, respectively. We conclude that spinal cord epidural cooling has a highly protective effect against ischemic spinal cord damage under experimental conditions of high thoracic aorta crossclamping in dogs.

Post cardiac arrest hyperoxic resuscitation enhances neuronal vulnerability of the respiratory rhythm generator and some brainstem and spinal cord neuronal pools in the dog.

Selective neuronal vulnerability of the motor cortex, basal ganglia, brainstem, medulla, cerebellum, C6 and L6 segments of the spinal cord were studied after 15 min of cardiac arrest followed by 1 h of normoxic or hyperoxic resuscitation using the suppressive Nauta method in dogs. Hyperoxic resuscitation causes characteristic somatodendritic argyrophilia of the interneuronal pool in the spinal cord and lower medulla. Cuneate, lateral reticular, supraspinal, and caudal trigeminal nuclei as well as the dorsal and ventral respiratory neuronal groups were heavily involved. Similarly, the Purkinje cells, neurons in the middle and deep portions of the mesencephalic tectum, perirubral, pretectal, posterior commissure, middle-sized striatal and giant pyramidal (Betz's) neurons in the motor cortex became argyrophilic. Hyperoxic resuscitation versus normoxic resuscitation causes statistically significant somatodendritic argyrophilia of the dorsal respiratory group, cuneate, dorsal lateral geniculate and thalamic reticular nuclei.

Integration of neocortical embryonal grafts with the neocortex of host rats examined by Leao's spreading cortical depression.

Cortical spreading depression (SD) was used to assess the density and organization of neural elements in neocortical transplants and their connectivity with the host brain. Embryonal neocortex (E14) was transplanted into cavities in the frontoparietal cortex of 3-month-old rats. SD elicited in the cortex of anesthetized host rats (n = 12) 3 to 8 months after transplantation did not penetrate into the grafts. SD could be elicited in large transplants but did not propagate to the surrounding host neocortex. Spontaneous unit activity in the transplants was affected by SD elicited in the neocortex of the host rats anesthetized with urethane. Most units (n = 49) displayed excitatory-inhibitory (52%) or inhibitory (29%) reactions, whereas purely excitatory reactions were less frequent (8%). The results suggest that the packing density of neurons in the transplant can support SD but that the conditions at the graft-host boundary (glial scar, scarcity of neurons) stop SD propagation. High reactivity of the graft neurons to SD in the host neocortex indicates that afferentation from the host brain represents an important, predominantly excitatory contribution to the spontaneous activity of the transplant.

Integration of neocortical embryonal grafts with the neocortex of host rats examined by leao's spreading cortical depression

Cortical spreading depression (SD) was used to assess the density and organization of neural elements in neocortical transplants and their connectivity with the host brain. Embryonal neocortex (E14) was trasplanted into cavities in the frontoparietal cortex of 3-month-old rats. SD could be elicited in large transplants but did not propagate to the surroundling host neocortex. Spontaneous unit activity in the transplants was affected by SD elicited in the neocortex of the host rats anesthetized with urethane. Most units (n=49) displayed excitatory-inhibitory (52 percent) or inhibitory (29 percent) reactions, whereas purely excitatory reactions were less frequent (8 percent). The results suggest that the packing density of neurons in the transplant can support SD but that the conditions at the graft-host boundary (glial scar, scarcity of neurons) stop SD propagation. High reactivity of the graft neurons to SD in the host brain represents an important, predominantly excitatory contribution to the spontaneous activity of the transplant(AU)

Functional recovery after olfactory bulbectomy in rats: effect of embryonal brain grafts.

Regeneration of olfactory receptor neurones after bulbectomy can lead to formation of extrabulbar synapses, the functional significance of which remains controversial. Adult hooded rats (n = 27) were bilaterally bulbectomized under pentobarbital anaesthesia. Small pieces of brain tissues (1-2 mm3; OB: olfactory bulb; St: corpus striatum) were obtained from embryos of the same strain and placed into the bulbectomy-produced cavity in contact with the exposed brain surface. Smell was tested at 2- to 3-week intervals from 7 weeks to 7 months after the operation. The latency to find hidden food gradually improved and attained the intact control level in bulbectomized rats without grafts, but remained poor in the OB and St transplanted groups. Seven to ten months after transplantation, spontaneous unit activity and unit reactions to amyl acetate vapours were examined with a carbon fibre microelectrode. Unit responses in the transplants resembled those in the normal OB, but were less pronounced. Morphological examination of the transplant and of its connections with the olfactory receptor neurones and with the host brain suggested that functional recovery was mediated by the connections of the olfactory axons with the remnants of the OB, with the anterior olfactory nucleus and/or with the frontal cortex. The adverse effect of OB and St transplants was probably due to their poor connectivity with the host brain which prevented the regenerating olfactory axons from reaching higher olfactory centres.