Neural grafting in a rat model of Huntington's disease: striosomal-like organization of striatal grafts as revealed by acetylcholinesterase histochemistry, immunocytochemistry and receptor autoradiography

Receptor characteristics and recovery of function following kainic acid lesions and fetal transplants of the striatum. I. Cholinergic systems

The relationship between striatal muscarinic cholinergic receptor development and locomotor activity/T-maze alternation behavior in adult female rats with kainic acid lesions (kal) and fetal transplants of the striatum (str) was examined. Kal led to a number of deficits under conditions of spontaneous locomotion, including: (1) decreased stereotypical and increased horizontal movements during spontaneous overnight locomotion, (2) decreased spontaneous alternation on a T-maze, and (3) deficits on a sensorimotor neurological exam. Lesion-induced deficits following injection with cholinergic agonists (pilocarpine)/antagonists (scopolamine) included: (1) hypoactivity on vertical activity and stereotypical activity following scopolamine injection, and (2) increased stereotypical activity and decreased horizontal activity following pilocarpine injection. Transplants differentially affected the different types of behavioral deficits. Transplants reversed some of the deficits under conditions of spontaneous locomotion, including the hyperactivity noted during the night period, but only partially reversed the sensorimotor neurological exam and had no effect on spontaneous alternations in the T-maze. The transplants did not reverse the lesion-induced deficits following scopolamine injection, but partially reversed the lesion-induced changes in locomotion following pilocarpine injection. The striatal transplants had reduced numbers of M1 but increased numbers of M2 muscarinic cholinergic receptors. Cholinergic receptor density correlated with scores on the sensorimotor functioning and alternation tasks, but not with the locomotor measures. Conversely, the cross-sectional area of the str correlated strongly with the transplant-induced recovery in the lesion group. These results suggest that the development of cholinergic receptor systems within the transplants proceeds abnormally, and that the abnormal development of the transplant may impact on the transplant's ability to remediate lesion-induced deficits.

Connectivity of striatal grafts implanted into the ibotenic acid-lesioned striatum--I. Subcortical afferents

Subcortical afferents to transplants of fetal striatal tissue, implanted into the excitotoxically lesioned striatum of adult recipient rats, were studied with retrograde and anterograde axonal tracers and immunohistochemistry. One week after a striatal ibotenic acid lesion, involving most of the head of the caudate-putamen, a suspension of fetal striatal tissue (embryonic day 14-15) was injected into the lesioned area. In one group of rats, the ibotenic acid lesion was preceded (10 days) by large intrastriatal injections of True Blue, with injection sites matching the area to be lesioned. This was done to retrogradely pre-label the host brain afferents to the area of the striatum later to be lesioned and grafted. At 3 or 6 months post-transplantation, small injections (50 nl) of rhodamine-labelled latex beads were made into the striatal grafts. In animals where the injections were confined to the graft, retrogradely labelled host brain neurons were found in the thalamus, the substantia nigra, amygdala and dorsal raphe nucleus. Double-labelling analysis revealed that the vast majority of the rhodamine bead-labelled neurons also contained True Blue, which indicates that the host afferents to the graft, to a large extent, were derived from the neurons which normally project to the area of the caudate-putamen which was lesioned by the ibotenic acid injection. To further substantiate these observations a second group of lesioned and grafted animals received unilateral wheatgerm agglutinin-horseradish peroxidase injections into the ipsilateral host thalamus at 4 months post-transplantation in order to anterogradely label the host thalamostriatal axons. In a third group of animals serotonin immunocytochemistry was performed in order to detect possible afferents from the raphe nuclei. In contrast to the serotonin-containing fibers, which were fairly evenly distributed throughout the graft tissue, the peroxidase-labelled thalamic afferents were most prominent in the peripheral zones of the grafts and they were densely aggregated at the graft-host interface. The combined results provide evidence that the intrastriatal grafts receive afferents from the host substantia nigra, thalamus, amygdala and dorsal raphe nucleus, but with different distributions. The afferents from the substantia nigra, amygdala and raphe nuclei seem to distribute throughout the grafted tissue, although they are most dense in the peripheral parts, whereas the thalamic afferents are largely confined to the peripheral areas of the transplants and to the graft-host interface.

Rapid growth of host afferents into fetal thalamic transplants

Fetal cell suspension grafts grow and differentiate when implanted into adult rat CNS areas previously neuron-depleted using an excitotoxin. There is some controversy in the literature concerning the timetable of establishment and possible extent of host-graft connections in these experimental conditions. The present study was undertaken to analyze the development of adult host monoaminergic afferents into a transplant formed by fetal thalamic neurons in the previously excitotoxically lesioned thalamus. It is demonstrated that both norepinephrin- and serotonin-immunoreactive fibers are present in the transplant as soon as 8 days after grafting. At those times, immunoreactive fibers exhibit morphological characteristics typically associated with immature stages. After longer survival time, up to 4 weeks after grafting, immunoreactive fibers are numerous in the transplant and exhibit morphological features comparable to those observed in the adult thalamus. These results demonstrate the rapid ingrowth of some fiber systems of the adult host into the transplant and suggest that grafted fetal cells can be functionally integrated into the host circuitry as soon as a few weeks after grafting.

Localization of immunoreactive GABA and enkephalin and NADPH-diaphorase-positive neurons in fetal striatal grafts in the quinolinic-acid-lesioned rat neostriatum

Fetal striatal tissue grafts have been shown to partially reverse the biochemical and behavioral deficits induced by excitotoxic lesions. To determine if grafted striatal neurons contain neurochemical markers similar to those in neurons in the caudate nucleus and to establish the morphological characteristics and relative frequency of labeled neurons in the grafts, the localization of immunoreactive GABA and leucine-enkephalin (ENK) and of NADPH-diaphorase (NADPH-d) activity was examined in fetal striatal grafts at the light and electron microscopic levels. Striatal tissue from 17-day fetuses was grafted into the caudate nucleus of adult rats 1 week after intracaudate injections of either a low or high dose of quinolinic acid. At the light microscopic level, immunoreactive GABA and ENK and NADPH-d-positive neurons, processes, and punctate structures were present within adjacent sections of the same grafts. The frequency and morphological features of these labeled cell populations were similar in grafts placed into either minimally or extensively lesioned striata. Immunoreactive GABA and ENK neurons in the grafts constituted 28% and 13.5%, respectively, of the neuronal population of the graft and their mean diameters were 22 and 14% larger, respectively, than neostriatal neurons that contained the same chemical markers. NADPH-d-positive neurons in the grafts formed 3.5% of total grafted neurons and exhibited characteristics of neostriatal NADPH-d-containing aspiny cells, including medium-sized somata, indented nuclei, and varicose dendrites. At the electron microscopic level most GABA-positive neurons in the grafts contained indented nuclei and most immunoreactive ENK somata had unindented nuclei. Dendrites and dendritic spines with GABA or ENK immunoreactivity were present in the grafts where they were postsynaptic to unlabeled axons. Immunoreactive GABA and ENK axon terminals formed synapses with unlabeled neuronal profiles in the grafts. These findings demonstrate that fetal striatal grafts contain chemically defined neuronal populations that form synaptic connections within the graft and share some features with corresponding cell groups in the neostriatum. These results provide an anatomical basis for the graft-induced recovery from behavioral and biochemical deficits caused by instrastriatal lesions reported in other studies.

Fetal homotypic transplant in the excitotoxically neuron-depleted thalamus: light microscopy

One month after an in situ injection of kainic acid into the ventrobasal thalamic complex (VB), the lesioned area is totally depleted of neurons. The present study has been undertaken to determine the cytoarchitecture and connectivity of the nucleus constructed by fetal thalamic neurons implanted into the excitotoxically lesioned area. Adult rats received an injection of kainic acid inducing a total neuronal depletion of the right lateral thalamus (including both the nucleus reticularis thalami and the lateral portion of the ventrobasal complex). One month later, homotypic neurons were taken from the dorsal thalamic primordium of rat embryos (gestational age 15-16 days), dissociated, and injected into the lesioned area as a cell suspension. After 2-4-month survival, the cytoarchitecture of the neonucleus formed by the grafted neurons within the previously neuron-depleted area was analyzed. Additionally, connectivity was analyzed in seven rats in which dorsal column nuclei and/or cortical projections to the area were labeled anterogradely with either 3H-leucine or wheat-germ agglutinin conjugated to HRP, and the animals were perfused and processed following various histological procedures (Nissl staining, autoradiographic processing, and histochemistry for visualization of peroxidase). Fetal neurons grew, differentiated, and progressively occupied the previously neuron-depleted area of the adult host CNS. They organized themselves into a neonucleus with particular cytoarchitectural features including 1) the existence of two concentric zones--a central zone containing neurons and glial cells and a marginal zone only filled with a band of glial cells, 2) an increase in cellular density compared to the intact thalamus, 3) the grouping of neurons in spherical clusters, and 4) apparent polymorphism of neuronal somata. Lemniscal and corticothalamic afferents originating from the host were observed in the neonucleus when the fetal neurons had been implanted correctly into the lesioned area but not when they had been misplaced into either normal thalamic tissue or the internal capsule. The afferents labeled from either the dorsal column nuclei or the somatosensory cortex were, however, less dense in the neonucleus than in the normal thalamus. These results are discussed with regard to the normal cytoarchitecture and connectivity of the ventrobasal complex of the rat thalamus.

Striatal grafts in rats with unilateral neostriatal lesions--I. Ultrastructural evidence of afferent synaptic inputs from the host nigrostriatal pathway

Tyrosine hydroxylase immunocytochemistry, in combination with Golgi impregnation, has been used to study the dopaminergic afferent input to striatal suspension grafts implanted into the previously ibotenic acid-lesioned striatum in adult recipient rats. The rats were perfused for combined light- and electron microscopy at 10-11 months after transplantation, at the end of a series of behavioural experiments and a study of in vivo GABA release, reported in the two accompanying papers. A tyrosine hydroxylase-positive fibre network occurred within the grafts in all eight specimens analysed. The tyrosine hydroxylase-positive fibres had a distinct "patchy" distribution, throughout the graft tissue, and within these patches the terminal density was similar to that of the normal intact striatum. Ultrastructurally, the tyrosine hydroxylase-positive fibres were seen to make abundant synaptic contacts with neuronal elements within the grafts. As in the normal striatum, they were all of the symmetric type and dendritic shafts and spines were the most usual postsynaptic targets. Sections from three of the grafted animals were taken for combined Golgi-impregnation and immunostaining. Only cells of the medium-sized densely spiny type were impregnated in this material. Six of them, which had portions extending into the immunostained neuropil, were drawn using a camera lucida and processed for electron microscopy. Tyrosine hydroxylase-positive boutons were seen to make symmetrical synaptic contacts onto the shafts and spines of the impregnated dendrites, and in one case also with the perikaryon. The results indicate that the medium-sized densely spiny neuron type (which is a predominant target for the dopaminergic afferents in the normal striatum) is abundant in the grafted tissue, and that these neurons represent a synaptic target also for the tyrosine hydroxylase-positive innervation of the striatal grafts.

Striatal grafts in rats with unilateral neostriatal lesions--III. Recovery from dopamine-dependent motor asymmetry and deficits in skilled paw reaching

This study investigated the functional capacity of intrastriatal grafts of embryonic striatal tissue in rats with unilateral ibotenic acid lesions of the neostriatum. The group of grafted rats was compared with lesion-alone and control groups for motor bias, as assessed by tests of rotation induced by dopaminergic, cholinergic and GABAergic drugs, and of skilled paw reaching. Unilateral striatal lesions induced marked ipsilateral turning to apomorphine and methamphetamine, which was substantially ameliorated in the grafted rats. Atropine induced similar rates of moderate (but non-significant) ipsilateral turning in the lesion and graft groups, whereas muscimol and gamma-acetylenic GABA induced no turning bias in any group. The lesioned rats showed a strong bias in their preference to use the paw ipsilateral to the lesion when reaching for food pellets, and a decline in reaching success with both paws. The grafts did not influence the ipsilateral paw preference in this task, but did provide a substantial improvement in the animal's reaching accuracy and ability to retrieve food with either paw. The results indicate that striatal grafts can provide a substantial amelioration of motor impairments induced by striatal lesions. Moreover they suggest that the graft's influence on the host brain is itself under the functional regulation of an afferent dopaminergic input from the host brain.

Striatal grafts in rats with unilateral neostriatal lesions--II. In vivo monitoring of GABA release in globus pallidus and substantia nigra

GABA release was recorded in vivo by push-pull perfusion from the globus pallidus and substantia nigra of control rats, rats with unilateral ibotenic acid lesions of the neostriatum, and rats with embryonic striatal tissue grafts implanted in the lesioned striatum. The lesions reduced baseline levels of GABA release to 5% of control levels in the globus pallidus and to 13% of control levels in the substantia nigra pars reticulata. GABA release was substantially restored in both the globus pallidus and substantia nigra of the grafted rats, to 34 and 60%, respectively. Peripheral injection of the dopaminergic stimulant methamphetamine induced a short (lasting approximately 20 min) 4-5 fold increase in GABA release in the intact globus pallidus and a longer (lasting longer than 80 min) increase in the substantia nigra. The stimulatory effect of methamphetamine on GABA release was completely abolished in both sites by the strial lesions, suggesting that the effect was mediated via a direct or indirect dopaminergic action on striatal output neurons. The grafts reinstated methamphetamine-induced stimulation of GABA release in striatal output targets to a level (as a proportion of baseline) that was similar to that seen in the control rats. The results support the view that activation of the dopaminergic inputs to the striatum is functionally excitatory on the major striatal output projections to the globus pallidus and substantia nigra pars reticulata. The results also support the hypothesis that striatal grafts have the capacity to become functionally incorporated by reciprocal graft-host connections into the neural circuitry of the host brain.

Comparison of growth and reinnervation properties of cholinergic neurons from different brain regions grafted to the hippocampus

Grafts of five different types of central cholinergic neurons, from the septal-diagonal band region, the nucleus basalis magnocellularis region (NBM), the striatum, the pontomesencephalic tegmentum of the brainstem, and the spinal cord, were compared with respect to their ability to grow and to reinnervate the cholinergically denervated hippocampal formation of adult rats. The areas were dissected from 14 to 15-day-old rat fetuses, and the same number of viable cells (35 X 10(4) from each of the different regions were stereotaxically injected as cell suspensions into the hippocampus of rats subjected to a transection of the intrinsic septo-hippocampal cholinergic pathways. At 17-19 weeks after transplantation, the various graft types differed considerably in their volume, the total amount of acetylcholinesterase (AChE)-positive fiber outgrowth, and the innervation pattern and morphology of the AChE-positive fibers growing into the host hippocampus. On average the NBM and spinal cord grafts had grown to become three to four times larger than the septal and the brainstem grafts, and 15-20 times larger than the striatal grafts. By contrast, the total ingrowth score of AChE-positive fibers in the host hippocampus from the septal grafts was about twice that of the NBM and brainstem grafts, about five times greater than the striatal grafts, and about six times greater than that of the spinal cord grafts. The large NBM grafts thus exhibited similar fiber outgrowth to the much smaller brainstem grafts, and the AChE-positive neurons of the grafted spinal cord grew very poorly into the hippocampus despite the fact that they survived very well. The innervation pattern and morphological features of the ingrowing AChE-positive fibers in the host hippocampus proper and in the dentate gyrus resembled those of normal rats in animals with grafts from any of the three forebrain regions (i.e., septum, NBM, or striatum), whereas ingrowth from the brainstem and spinal cord grafts were markedly abnormal with respect to both innervation pattern and fiber morphology. These results provide further evidence that the overall survival, growth, and fiber outgrowth of intracerebral neural grafts depend on interactions with the surrounding host tissue. Since the ability to reinnervate the previously denervated host target was greatest for the neuron type normally innervating that area, i.e., the septal-diagonal band neurons, we conclude that neuronal properties beyond the transmitter type are essential for the optimal performance of implanted neurons in intracerebral grafting experiments.

Minimal connectivity between neostriatal transplants and the host brain

The present study sought to determine if axonal connectivity is established between neostriatal transplants and the host brain during the first two months of graft development. Cell suspensions of embryonic neostriatum were transplanted into the adult rat neostriatum lesioned previously by kainic acid. After 1-2 months, injections of horseradish peroxidase conjugated with wheat germ agglutinin (HRP) were made either within the graft, into adjacent host neostriatum or the host ventral midbrain. In animals with HRP injection sites restricted to the graft no retrograde or anterograde label was found in the host brain. However, both anterograde axon label and retrogradely labelled neurons were found in areas within the transplant but distal to the injection site. Neither ventral midbrain nor host neostriatal HRP injections resulted in any significant anterograde or retrograde label within the graft. These results demonstrate a lack of connectivity between neostriatal grafts and the host brain 1-2 months post-transplantation but an ability of grafted neurons to project to different locations within the transplant. Therefore, transplanted neostriatal neurons develop for the first two months in the absence of normal neostriatal afferent and efferent connections.

Quantitative analysis of dendrites from transplanted neostriatal neurons

Dendritic spine density was determined quantitatively in 35-day-old neostriatal transplants and age-matched control tissue. Transplanted spiny I neurons showed significant decreases in spine density and in number of proximal dendrites. These differences may be due to aberrant maturation of transplanted neurons.