Fate of GABAergic septohippocampal neurons after fimbria-fornix transection as revealed by in situ hybridization for glutamate decarboxylase mRNA and parvalbumin immunocytochemistry

Injection of Fluoro-Gold into the tibial nerve leads to prolonged but reversible functional deficits in rats

Tract tracing with neuronal tracers not only represents a straightforward approach to identify axonal projection connection between regions of the nervous system at distance but also provides compelling evidence for axonal regeneration. An ideal neuronal tracer meets certain criteria including high labeling efficacy, minimal neurotoxicity, rapid labeling, suitable stability in vivo, and compatibility to tissue processing for histological/immunohistochemical staining. Although labeling efficacy of commonly used fluorescent tracers has been studied extensively, neurotoxicity and their effect on neural functions remains poorly understood. In the present study, we comprehensively evaluated motor and sensory nerve function 2-24 weeks after injection of retrograde tracer Fluoro-Gold (FG), True Blue (TB) or Fluoro-Ruby (FR) in the tibial nerve in adult Spague-Dawley rats. We found that motor and sensory nerve functions were completely recovered by 24 weeks after tracer exposure, and that FG lead to a more prolonged delay in functional recovery than TB. These findings shed light on the long-term effect of tracers on nerve function and peripheral axonal regeneration, and therefore have implications in selection of appropriate tracers in relevant studies.

The integrity of cholinergic basal forebrain neurons depends on expression of Nkx2-1

The transcription factor Nkx2-1 belongs to the homeobox-encoding family of proteins that have essential functions in prenatal brain development. Nkx2-1 is required for the specification of cortical interneurons and several neuronal subtypes of the ventral forebrain. Moreover, this transcription factor is involved in migratory processes by regulating the expression of guidance molecules. Interestingly, Nkx2-1 expression was recently detected in the mouse brain at postnatal stages. Using two transgenic mouse lines that allow prenatal or postnatal cell type-specific deletion of Nkx2-1, we show that continuous expression of the transcription factor is essential for the maturation and maintenance of cholinergic basal forebrain neurons in mice. Notably, prenatal deletion of Nkx2-1 in GAD67-expressing neurons leads to a nearly complete loss of cholinergic neurons and parvalbumin-containing GABAergic neurons in the basal forebrain. We also show that postnatal mutation of Nkx2-1 in choline acetyltransferase-expressing cells causes a striking reduction in their number. These degenerative changes are accompanied by partial denervation of their target structures and results in a discrete impairment of spatial memory.

Axotomy-induced neurotrophic withdrawal causes the loss of phenotypic differentiation and downregulation of NGF signalling, but not death of septal cholinergic neurons

BACKGROUND

Septal cholinergic neurons account for most of the cholinergic innervations of the hippocampus, playing a key role in the regulation of hippocampal synaptic activity. Disruption of the septo-hippocampal pathway by an experimental transection of the fimbria-fornix drastically reduces the target-derived trophic support received by cholinergic septal neurons, mainly nerve growth factor (NGF) from the hippocampus. Axotomy of cholinergic neurons induces a reduction in the number of neurons positive for cholinergic markers in the medial septum. In several studies, the reduction of cholinergic markers has been interpreted as analogous to the neurodegeneration of cholinergic cells, ruling out the possibility that neurons lose their cholinergic phenotype without dying. Understanding the mechanism of cholinergic neurodegeneration after axotomy is relevant, since this paradigm has been extensively explored as an animal model of the cholinergic impairment observed in neuropathologies such as Alzheimer's disease.The principal aim of this study was to evaluate, using modern quantitative confocal microscopy, neurodegenerative changes in septal cholinergic neurons after axotomy and to assess their response to delayed infusion of NGF in rats.

RESULTS

We found that there is a slow reduction of cholinergic cells labeled by ChAT and p75 after axotomy. However, this phenomenon is not accompanied by neurodegenerative changes or by a decrease in total neuronal number in the medial septum. Although the remaining axotomized-neurons appear healthy, they are unable to respond to delayed NGF infusion.

CONCLUSIONS

Our results demonstrate that at 3 weeks, axotomized cholinergic neurons lose their cholinergic phenotype without dying and down-regulate their NGF-receptors, precluding the possibility of a response to NGF. Therefore, the physiological role of NGF in the adult septal cholinergic system is to support phenotypic differentiation and not survival of neurons. This evidence raises questions about the relationship between transcriptional regulation of the cholinergic phenotype by retrograde-derived trophic signaling and the transcriptional changes experienced when retrograde transport is impaired due to neuropathological conditions.

Ongoing expression of Nkx2.1 in the postnatal mouse forebrain: potential for understanding NKX2.1 haploinsufficiency in humans?

Coordinated movements require the caudate-putamen and the globus pallidus, two nuclei belonging to the basal ganglia, to be intact and functioning properly. Many neurons populating these regions derive from the medial ganglionic eminence, a transient structure that expresses the transcription factor Nkx2.1 during prenatal development. Accordingly, the basal ganglia of Nkx2.1(-/-) mice are heavily affected and a substantial loss of several types of GABAergic interneurons has been observed. Interestingly, heterozygous mutation of the NKX2.1 gene in humans has been described as causing an unusual disorder from the second year of life onwards, which is mainly characterized by disturbances of motor abilities and delayed speech development. In the present study, we therefore investigated whether Nkx2.1 is still expressed in the young adult and aged mouse forebrain. After birth, the most intense immunolabeling for Nkx2.1 was detected in several components of the hypothalamic region, in the subventricular zone of the ventral tips lining the lateral ventricles, and in neighboring structures including the striatum, the globus pallidus and the various nuclei of the septal complex. Surprisingly, this staining pattern was substantially maintained into adulthood. Double immunocytochemistry for Nkx2.1 and various neuronal markers revealed that mainly parvalbumin-containing GABAergic neurons, but also cholinergic neurons, of the ventral forebrain express this protein. Moreover, in situ hybridization confirmed that these neurons maintain synthesis of Nkx2.1 throughout life. The robust expression of Nkx2.1 by these neurons points to a broad functional spectrum within the adult forebrain.

The role of cholinergic and GABAergic medial septal/diagonal band cell populations in the emergence of diencephalic amnesia

The septohippocampal pathway, which is mostly composed of cholinergic and GABAergic projections between the medial septum/diagonal band (MS/DB) and the hippocampus, has an established role in learning, memory and disorders of cognition. In Wernicke-Korsakoff's syndrome (WKS) and the animal model of the disorder, pyrithiamine-induced thiamine deficiency (PTD), there is both diencephalic damage and basal forebrain cell loss that could contribute to the amnesic state. In the current experiment, we used the PTD animal model to access both cholinergic (choline acetyltransferase [ChAT] immunopositive) and GABAergic (parvalbumin [PV]; calbindin [CaBP]) neuronal loss in the MS/DB in relationship to midline-thalamic pathology. In addition, to gain an understanding about the role of such neuropathology in behavioral dysfunction, animals were tested on a non-rewarded spontaneous alternation task and behavioral performance was correlated to neuropathology. Unbiased stereological assessment of neuronal populations revealed that ChAT-positive neurons were significantly reduced in PTD rats, relative to control pair-fed rats, and thalamic mass and behavioral performance correlated with ChAT neuronal estimates. In contrast, both the PV- and CaBP-positive neurons in the MS/DB were not affected by PTD treatment. These results support an interactive role of both thalamic pathology and cholinergic cell loss in diencephalic amnesia.

Altered neuronal responses and regulation of neurotrophic proteins in the medial septum following fimbria-fornix transection in CNTF- and leukaemia inhibitory factor-deficient mice

Degeneration of axotomized GABAergic septohippocampal neurones has been shown to be enhanced in ciliary neurotrophic factor (CNTF)-deficient mice following fimbria-fornix transection (FFT), indicating a neuroprotective function of endogenous CNTF. Paradoxically, however, the cholinergic population of septohippocampal neurones was more resistant to axotomy in these mutants. As leukaemia inhibitory factor (LIF) has been identified as a potential neuroprotective factor for the cholinergic medial septum (MS) neurones, FFT-induced responses were compared in CNTF(-/-), LIF(-/-) and CNTF/LIF double knockout mice. In CNTF(-/-) mice, FFT-induced cholinergic degeneration was confirmed to be attenuated as compared with wildtype mice. The expression of both LIF and LIF receptor beta was increased in the MS providing a possible explanation for the enhanced neuronal resistance to FFT in these animals. However, ablation of the LIF gene also produced paradoxical effects; following FFT in LIF(-/-) mice no loss of GABAergic or cholinergic MS neurones was detectable during the first postlesional week, suggesting that other efficient neuroprotective mechanisms are activated in these animals. In fact, enhanced activation of astrocytes, a source of neurotrophic proteins, was indicated by increased up-regulation of glial fibrillary acidic protein and vimentin expression. In addition, mRNA levels for neurotrophin signalling components (e.g. nerve growth factor, p75(NTR)) were differentially regulated. The positive effect on axotomized cholinergic neurones seen in CNTF(-/-) and LIF(-/-) mice as well as the increased up-regulation of astrogliose markers was abolished in CNTF/LIF double knockout animals. Our results indicate that endogenous CNTF and LIF are involved in the regulation of neuronal survival following central nervous system lesion and are integrated into a network of neurotrophic signals that mutually influence their expression and function.

Firing properties of anatomically identified neurons in the medial septum of anesthetized and unanesthetized restrained rats

Cholinergic and GABAergic neurons in the medial septum-diagonal band of Broca (MS-DB) project to the hippocampus where they are involved in generating theta rhythmicity. So far, the functional properties of neurochemically identified MS-DB neurons are not fully characterized. In this study, MS-DB neurons recorded in urethane anesthetized rats and in unanesthetized restrained rats were labeled with neurobiotin and processed for immunohistochemistry against glutamic acid decarboxylase (GAD), parvalbumin (PV), and choline acetyltransferase (ChAT). The majority of the 90 labeled neurons (75.5%) were GAD+. Among them, 34.0% were also PV+, but none were ChAT+. Only 8.8% of the labeled neurons were found ChAT+. Remaining neurons (15.5%) were not identified. In anesthetized rats, all of the PV/GAD+ and 65% of GAD+ neurons exhibited burst-firing activity at the theta frequency. PV/GAD+ neurons displayed higher discharge rate and longer burst duration compared with GAD+ neurons. At variance, all of the ChAT+ neurons were slow-firing. Cluster-firing and tonic-firing were observed in GAD+ and unidentified neurons. In unanesthetized rats, during wakefulness or rapid eye movement sleep with hippocampal theta, the bursting neurons were PV/GAD+ or GAD+, whereas all of the ChAT+ neurons were slow-firing. Across the sleep-wake cycle, the GABAergic component of the septohippocampal pathway was always more active than the cholinergic one. The fact that cholinergic MS-DB neurons do not display theta-related bursting or tonic activity but have a very low firing rate questions how acetylcholine exerts its activating role in the septohippocampal system.

Endogenous ciliary neurotrophic factor protects GABAergic, but not cholinergic, septohippocampal neurons following fimbria-fornix transection

Application of neurotrophic proteins including ciliary neurotrophic factor (CNTF) and leukemia inhibitory factor (LIF), members of the family of gp130-associated cytokines, can rescue CNS neurons from injury-induced degeneration. However, it is not clear so far if these effects reflect a physiological function of the endogenous cytokines. Using fimbria-fornix transection as a model, we examined whether responses of GABAergic and cholinergic septohippocampal neurons to axotomy are altered in mice lacking CNTF. In addition, we studied the cellular expression of CNTF, LIF and related cytokine receptor components in the septal complex following lesion. Degeneration of septohippocampal GABAergic neurons in the medial septum as indicated by the loss of parvalbumin-immunoreactive neurons was accelerated and permanently enhanced in CNTF(-/-) mice as compared to wild-type animals. Unexpectedly, the number of axotomized cholinergic MS neurons was significantly higher in CNTF-deficient mice during the first 2 weeks postlesion. Both in wild-type and in CNTF(-/-) mutants, expression of mRNA for the CNTF-specific alpha-subunit of the cytokine receptor complex was specifically upregulated in axotomized GABAergic septal neurons, whereas enhanced expression of the LIF-binding beta-subunit was specifically observed in axotomized cholinergic neurons. Following lesion, CNTF expression in wild-type mice was induced in activated astrocytes surrounding the axotomized neurons and at the lesion site. Expression of LIF mRNA was localized in the GABAergic and cholinergic septohippocampal neurons. These results strongly indicate that endogenous CNTF, supplied by reactive glia cells, acts as a neuroprotective factor for axotomized CNS neurons. In the septum, endogenous CNTF specifically supports lesioned GABAergic projection neurons, whereas LIF may play a similar role for the cholinergic counterparts.

Activation of STAT3 signaling in axotomized neurons and reactive astrocytes after fimbria-fornix transection

It is an open question to what extent neuroprotective mechanisms involving neurotrophic proteins are activated after central nervous system (CNS) lesions. Results from previous studies have indicated that ciliary neurotrophic factor (CNTF) and other members of the family of gp130-associated cytokines have neuroprotective effects on septohippocampal projection neurons axotomized by fimbria-fornix transection (FFT). Here we demonstrate that the transcription factor STAT3, a component of the primary cytokine signal transduction pathway, is transiently activated after FFT in the medial septum and in the lateral septum deafferented by the lesion. Immunocytochemical double-labeling showed nuclear signals for phosphorylated STAT3 in both types (GABAergic and cholinergic) of axotomized medial septal neurons around day 4 postlesion. This response temporally coincided with the cell-type-specific upregulation of the cytokine receptor components CNTF receptor alpha and LIF receptor beta in the same neurons. However, neuronal STAT3-activation was not abolished in CNTF- or LIF-deficient mouse mutants. Furthermore, lesion-induced STAT3 signaling was also found in reactive GFAP-positive astrocytes of the medial and lateral septum. Interestingly, basal GFAP expression was reduced but postlesional upregulation was markedly enhanced in CNTF(-/-) animals. These results demonstrate transient activation of cytokine signaling after CNS lesion and suggest that gp130-associated cytokines have multiple functions in the lesioned CNS by directly acting on axotomized neurons and on reactive astrocytes.

Long-term neuronal effects and disposition of ectopic preproNGF gene transfer into the rat septum

Although NGF gene therapy has been proposed to treat age- or disease-related brain cholinergic decline, little is known about the ectopic expression or function of this trophic factor after transduction in the brain especially over long intervals. The neuron-targeting, recombinant adeno-associated virus serotype 2 (rAAV2) vector was used to express mouse NGF with C-terminal myc-tag in septum using a full-length preproNGF sequence. While the predominant form of endogenous NGF immunoreactivity in septum was 31 kd of proNGF, almost all of the ectopic NGF-immunoreactivity attributable to the rAAV2-mediated transduction in this region was recovered as mature NGF. Transgene expression was found in both cholinergic and GABAergic neurons, with the number of transduced neurons dependent on vector dose. To determine the long-term effects of this NGF-expression on neuron function, fimbria-fornix (FF) lesions were conducted 6 months after NGF gene transfer. NGF gene transfer attenuated the lesion-induced loss of septal cholinergic but not GABAergic neurons, indicating that long-term expression did not eliminate this response, which has been noted over short intervals. The effects and dose dependency of NGF gene delivery on neuroprotection and neurotrophism were also examined. NGF transduction increased cholinergic cell size in the septum, but required a higher vector dose than neuroprotection. These results reveal potential long-term benefits as well as concerns for genetically modifying septal NGF gene expression to preserve neuronal viability and function.

Effects of fimbria-fornix transection on calpain and choline acetyl transferase activities in the septohippocampal pathway

The ability of fimbria-fornix bilateral axotomy to elicit calpain and caspase-3 activation in the rat septohippocampal pathway was determined using antibodies that selectively recognize either calpain- or caspase-cleaved products of the cytoskeletal protein alphaII-spectrin. Radioenzymatically determined choline acetyl transferase (ChAT) activity was elevated in the septum at day 5, but reduced in the dorsal hippocampus at days 3, 5 and 7, after axotomy. Prominent accumulation of calpain-, but not caspase-3-, cleaved spectrin proteolytic fragments was observed in both the septum and dorsal hippocampus 1-7 days after axotomy. ChAT-positive neuronal cell bodies in the septum also displayed calpain-cleaved spectrin indicating that calpain activation occurred in cholinergic septal neurons as a consequence of transection of the septohippocampal pathway. Calpain-cleaved alphaII-spectrin immunoreactivity was observed in cholinergic fibers coursing through the fimbria-fornix, but not in pyramidal neurons of the dorsal hippocampus, suggesting that degenerating cholinergic nerve terminals were the source of calpain activity in the dorsal hippocampus following axotomy. Accumulation of calpain-cleaved spectrin proteolytic fragments in the dorsal hippocampus and septum at day 5 after axotomy was reduced by i.c.v. administration of two calpain inhibitors. Calpain inhibition partially reduced the elevation of ChAT activity in the septum produced by transection but failed to decrease the loss of ChAT activity in the dorsal hippocampus following axotomy. These findings suggest that calpain activation contributes to the cholinergic cell body response and hippocampal axonal cytoskeletal degradation produced by transection of the septohippocampal pathway.

Up-regulation of growth-associated protein 43 mRNA in rat medial septum neurons axotomized by fimbria-fornix transection

Transection of septohippocampal fibres is widely used to study the response of CNS neurons to axotomy. Septohippocampal projection neurons survive axotomy and selectively up-regulate the transcription factor c-Jun. In the present study we investigated whether these cells concomitantly up-regulate the growth-associated protein-43 (GAP-43), a potential target gene of c-Jun implicated in axonal growth and regeneration. Using in situ hybridization histochemistry (ISHH) it was demonstrated that postlesional c-jun mRNA expression is accompanied by an increased expression of GAP-43 mRNA in the medial septum 3 days following fimbria-fornix transection (FFT). The increase reached a maximum at 7 days and gradually declined thereafter (17 days, 3 weeks). Retrograde prelabeling with Fluoro-Gold followed by axotomy and ISHH revealed that GAP-43 mRNA was up-regulated in septohippocampal projection neurons. Colocalization of GAP-43 mRNA and choline acetyltransferase protein showed that GAP-43 mRNA was expressed by cholinergic medial septal neurons after axotomy. Selective immunolesioning of the cholinergic component of the septohippocampal projection with 192 IgG-saporin followed by FFT demonstrated that GAP-43 mRNA was also synthesized by axotomized GABAergic neurons. These results demonstrate an up-regulation of GAP-43 mRNA in axotomized septohippocampal projection neurons independent of their transmitter phenotype which is closely correlated with c-Jun expression. Because the GAP-43 gene contains an AP-1 site, we hypothesize a c-Jun-driven up-regulation of GAP-43 in lesioned medial septal neurons that may contribute to their survival and regenerative potential following axotomy.

Retrograde tracing with Fluoro-Gold: different methods of tracer detection at the ultrastructural level and neurodegenerative changes of back-filled neurons in long-term studies

Among the available retrograde fluorescent tracers Fluoro-Gold (FG) is particularly advantageous because it (1) is not only detectable by fluorescence microscopy but also immunocytochemically, resulting in an almost complete staining of the dendritic arbor, (2) is visible in lysosome-like structures allowing for the identification of projection neurons at the ultrastructural level, and (3) remains in the labeled neurons for extended periods of time. Photoconversion and immunostaining for FG, respectively, result in a stable, electron-dense reaction product. Thus, the retrogradely labeled cells can be analyzed quantitatively in the light- and electron microscope for their structural characteristics and input synapses. Long-term studies of back-filled neurons provided evidence for neurotoxic effects of FG in these cells.

Retrograde and anterograde tracing combined with transmitter identification and electron microscopy

Fiber tracts in the brain are formed by neurochemically heterogeneous neuron populations. To distinguish between the different neurons that contribute to a fiber tract it is necessary to combine anterograde and retrograde tracing techniques with immunocytochemistry. In this article, we describe two techniques which allow for the neurochemical identification of retrogradely labeled neurons and anterogradely labeled axons on the ultrastructural level. The identification of the neurotransmitter identity of retrogradely labeled neurons is achieved by combining retrograde Fluoro-Gold tracing with preembedding immunocytochemistry, while the neurotransmitter identity of anterogradely labeled axons can be revealed by combining anterograde Phaseolus vulgaris-leucoagglutinin (PHAL) tracing and postembedding immunostaining.

An analysis of atrophy in the medial mammillary nucleus following hippocampal and fornix lesions in humans and nonhuman primates

Lesions of the hippocampal formation or transections of the fornix are followed by shrinkage of the medial mammillary nucleus (MMN). We determined whether the shrinkage of this nucleus was due to loss and/or shrinkage of neurons in addition to the loss of neuropil. We examined the MMN in a patient (KB) with an infarct that led to marked atrophy of the left hippocampus and subiculum, leaving the right MMN intact. Unbiased, stereological measurement techniques were used to compare the total cell number and individual neuronal cross-sectional areas in both left and right MMN in this patient and in two control human brains. We also analyzed the MMN in four macaque monkeys that underwent experimental unilateral transections of the fornix. The volume of the MMN on the lesioned side in KB was 55% of the unlesioned side (2.8 mm(3) vs 5.1 mm(3)); the MMN in the monkey cases were reduced to 47-58% of the volume of the nonlesioned side. Neurons in the deafferented MMN of KB and of the monkey subjects were decreased in cross-sectional area (16-20%, P < 0.0001). There was a trend toward decreased cell numbers (11-15%) on the lesioned side in all cases. We have estimated that the loss in cell number and shrinkage of remaining cells contribute negligibly to the 45% reduction in MMN volume. Therefore, the loss of neuropil (dendrites and afferent and efferent axons) appears to be the major contributor to the change in MMN volume.

Axotomy-induced c-JUN expression in young medial septal neurons is regulated by nerve growth factor

In the present study we investigated the axotomy-induced expression of the proto-oncogene c-jun in young rat medial septal neurons and its regulation by nerve growth factor. First, medial septal neurons were retrogradely labelled by Fast Blue injection into the hippocampus at postnatal day 1 (P1). Rats of different developmental ages (P6, P9, P14, P21, P28 and P42) were then subjected to bilateral fimbria-fornix transection resulting in the axotomy of septohippocampal projection neurons. After the lesion, c-JUN immunoreactivity was observed in the nuclei of axotomized medial septal neurons of all stages examined, suggesting that c-JUN induction is an age-independent feature of axotomized medial septal neurons. Double immunolabelling for choline acetyltransferase and c-JUN or parvalbumin and c-JUN, respectively, revealed that both cholinergic and GABAergic septohippocampal projection neurons express c-JUN after axotomy. In addition, a co-localization of immunostaining for c-JUN and the neuropeptide galanin was found after lesion, as both proteins were induced in the same medial septal neurons following fimbria-fornix transection. Next, the regulation of c-JUN expression in axotomized medial septal neurons was studied in organotypic cultures of the medial septum. Axotomized medial septal neurons in culture did not express c-JUN in contrast to the in vivo situation. With the concept that nerve growth factor suppresses c-JUN expression, slice cultures of the medial septum were treated with antibodies against nerve growth factor. This treatment caused a dose-dependent increase in c-JUN-positive cells in these slice cultures. Simultaneous addition of nerve growth factor and antibodies against nerve growth factor resulted in the reversal of this effect. These data suggest an age-independent induction of c-JUN in axotomized medial septal neurons and its regulation by nerve growth factor.

Role of NGF in axotomy-induced c-Jun expression in medial septal cholinergic neurons

The extent of neuronal regeneration after axotomy largely depends on the survival capacity of the injured cell. It has been shown for a long time that nerve fiber transection results in retrograde changes in the parent neuronal cell body, and that these changes may eventually lead to neuronal degeneration. At present, little is known about the sequence of events initiated in a nerve cell body by the transection of its axonal process. In this report, we will focus on an interaction of nerve growth factor (NGF) with the transcription factor c-Jun in intact and axotomized septohippocampal projection neurons.

Electron microscopic evidence for a cholinergic innervation of GABAergic parvalbumin-immunoreactive neurons in the rat medial septum

The presence of interconnections between cholinergic and parvalbumin (PARV)-containing gamma aminobutyric acid (GABA)ergic septohippocampal projection neurons is still a matter of debate. To search for contacts of cholinergic collateral axon terminals in the septal-diagonal band region the immunotoxin 192IgG-saporin was applied, which was proved to selectively destroy cholinergic basal forebrain neurons. Seven and 10 days after administration of the immunotoxin, choline acetyltransferase immunoreactivity had disappeared, and numerous neuronal somata and dendrites as well as axonal terminals revealed characteristics of electron-lucent degeneration. Electron-dense degeneration was never observed in dendrites and synaptic boutons. Degenerating terminals were found in contact with PARV-immunopositive and PARV-negative neurons. Because only cholinergic cells were degenerating, the terminals should be collaterals from cholinergic neurons. In addition to such contacts, PARV-immunoreactive boutons were seen in contact with PARV-positive and PARV-negative cells, but were not identified at degenerating postsynaptic profiles. As suggested in other studies, cholinergic boutons contacting GABAergic PARV-containing septal projection cells may influence hippocampal theta activity. Furthermore, multiple synaptic connections of both neuronal populations forming the septohippocampal pathway may contribute to their high rate of survival after fimbria-fornix transection.

Ultrastructural analysis of the progression of neurodegeneration in the septum following fimbria-fornix transection

The fimbria-fornix transection paradigm has been used as a model of retrograde neurodegeneration within the medial septal nucleus and anterograde degeneration of axon terminals within the lateral septal nucleus. Because the maintenance and survival of neurons may depend on the integrity of both efferents and afferents, the ultrastructure of neurons in the medial septal nucleus and dorsolateral septal nucleus was analysed at three, seven, 14, 30 days, and six months following unilateral transection of the fimbria-fornix in adult rats. Degeneration of axonal and somatodendritic compartments occurred in both nuclei on the side ipsilateral to fimbria-fornix transection. Degeneration of axons and terminals was present by three days and dissipated thereafter, although degenerating axodendritic and axosomatic terminals were still detected at 14-30 days postlesion. Dendrosomal alterations in both septal nuclei manifested as redistribution of organelles, dispersion and loss of rough endoplasmic reticulum, formation of membrane-bound vacuolar cisternae and membranous inclusions, loss of cytoplasmic matrix, and dispersion of chromatin throughout the nucleoplasmic matrix. These changes occurred in the absence of apparent ultrastructural damage to mitochondria and condensation of the nucleus. Dendritic pathology in both the medial and dorsolateral septal nuclei was most prominent at 14-30 days postlesion, but the neuropil recovered to control appearance by six months postlesion. In contrast, the cytoplasmic rarefaction and vacuolation of neuronal cell bodies were persistent in both the medial septal nucleus and the dorsolateral septal nucleus. We conclude that, following disconnection from the hippocampus, ultrastructural abnormalities occur within neurons in both the medial and lateral septal nuclei. The characteristics and time-course for these changes are similar in both nuclei. The neuropilar degeneration was transient, in contrast to the neuronal cell body injury which was persistent and was morphologically consistent with long-term neuronal atrophy.

Parvalbumin immunoreactive neurons in the rat septal complex have substantial glial coverage and receive few direct contacts from catecholaminergic terminals

Our previous studies have demonstrated that septohippocampal neurons in the rat septal complex have substantial glial coverage and have a number of synaptic associations with catecholaminergic terminals. While similar ultrastructural characteristics are observed for septal cholinergic neurons, the morphology and synaptic relations of catecholaminergic terminals with septal GABAergic neurons is largely unknown. Since the GABAergic septohippocampal neurons colocalize the calcium-binding protein, parvalbumin (PVA), the present study examined the ultrastructural relations of PVA neurons with catecholaminergic terminals in the septal complex. Single sections were dually labeled with antibodies to PVA and either tyrosine hydroxylase (TH) or dopamine-beta-hydroxylase (DBH). By light microscopy, processes with TH- and DBH- (TH/DBH) immunoreactivity were near PVA-labeled neurons. By electron microscopy, PVA-labeled perikarya had an average diameter of 14.9+/-6 microm and were ovoid or elongated. PVA-labeled perikarya (n = 124) had a large amount of astrocytic coverage (75+/-14%) and a low amount of terminal coverage (15+/-12%). PVA-labeled perikarya and dendrites mostly were contacted by terminals lacking immunoreactivity for either PVA or TH/DBH (82% of 1,663). Of the TH/DBH terminals or axons near PVA somata and dendrites, few (3% of 1,663) directly contacted them while the majority abutted adjacent glial or neuronal profiles. Some TH/DBH- and PVA-labeled terminals contacted the same dendrites; a few of these contained immunoreactivity for PVA. The results demonstrate that PVA-containing GABAergic septal neurons, like cholinergic neurons, are mostly surrounded by astrocytes and have very little terminal coverage. However, in contrast to cholinergic neurons, PVA-containing neurons are contacted primarily by non-catecholaminergic terminals suggesting that any functional interactions would be indirect. These findings further support the functional diversity of subpopulations of septohippocampal neurons.

Recovery of ChAT immunoreactivity in axotomized rat cholinergic septal neurons despite reduced NGF receptor expression

Previous studies have suggested that target-derived nerve growth factor (NGF) is essential for the survival of cholinergic basal forebrain neurons. Thus, axotomy of septohippocampal neurons in adult rats resulting in the withdrawal of target-derived NGF caused a dramatic loss of choline acetyltransferase (ChAT)-immunoreactive neurons in the medial septum-diagonal band complex. We have recently shown that this loss of immunolabelled neurons does not indicate cell death, since many septohippocampal cholinergic neurons recover their immunoreactivity for ChAT after a long survival time despite disconnection from target-derived neurotrophins. One possibility would be that these surviving ChAT-immunoreactive neurons have gained access to other, probably local, NGF sources. Here we provide evidence that the recovery of ChAT immunoreactivity after axotomy is not accompanied by a similar recovery of NGF receptor expression in these neurons. In situ hybridization for p75NTR mRNA and trkA mRNA 6 months after bilateral fimbria-fornix transection revealed a substantial loss of labelled cells. In addition, there was a persisting loss of p75NTR-immunoreactive and NGF-immunoreactive medial septal neurons. Cholinergic neurons in controls did not express NGF mRNA, but were heavily immunostained for NGF protein due to receptor-mediated uptake. These data suggest that at least some cholinergic septohippocampal neurons re-express ChAT either independently of NGF or with a reduced need for NGF.

192 IgG-saporin-induced loss of cholinergic neurons in the septum abolishes cholinergic sprouting after unilateral entorhinal lesion in the rat

After unilateral lesion of the entorhinal cortex, cholinergic septohippocampal fibres are believed to sprout in the denervated outer molecular layer of the rat dentate gyrus. This cholinergic sprouting has been demonstrated by acetylcholinesterase (AChE) histochemistry, a method said selectively to label cholinergic septohippocampal fibres in the hippocampus. However, a recent report has questioned this concept, suggesting that AChE may not be an adequate marker to monitor cholinergic sprouting and that other, non-cholinergic axons sprouting after entorhinal cortex lesion cause the dense AChE-positive band in the denervated outer molecular layer. In order to determine the contribution of cholinergic septohippocampal fibres to the dense AChE band appearing after entorhinal cortex lesion, the neurotoxin 192 IgG-saporin, known to destroy cholinergic neurons in the basal forebrain selectively, was used. Rats received bilateral injections of 192 IgG-saporin into the lateral ventricles 3 weeks before entorhinal cortex lesion, simultaneously with entorhinal cortex lesion, or 8 weeks after entorhinal cortex lesion. Immunocytochemistry for choline acetyltransferase (ChAT) and in situ hybridization for ChAT mRNA demonstrated the loss of cholinergic neurons in the medial septum and diagonal band after 192 IgG-saporin treatment. The cholinergic sprouting response in the molecular layer, as visualized with AChE histochemistry, was abolished in all animals treated with immunotoxin. These data indicate that the dense AChE band forming after entorhinal cortex lesion represents the sprouting of cholinergic septohippocampal fibres.

Development of cholinergic and GABAergic neurons in the rat medial septum: effect of target removal in early postnatal development

During normal development of the nervous system, the target fields influence the survival and differentiation of projection neurons, but the factors regulating this interaction remain obscure. In the present study, we have raised the question whether the target region is essential for the postnatal development and maintenance of two different types of central projection neurons, cholinergic and GABAergic septohippocampal cells. In early postnatal rats (P5, P10), the hippocampus was eliminated by unilateral intrahippocampal injections of the excitotoxin N-methyl-D-aspartate. After a long survival time (at P70), we have immunostained serial sections of the septal region with antibodies against choline acetyltransferase (ChAT), the acetylcholine-synthesizing enzyme, or the calcium-binding protein parvalbumin (PARV) which is known to be contained in GABAergic septohippocampal neurons. In the medial septum ipsilateral to the lesioned side, about 60% of ChAT-immunoreactive neurons and 62% of PARV-immunoreactive neurons were found in adulthood even after complete elimination of the hippocampus. Some immunoreactive cells appeared heavily shrunken, but electron microscopic analysis revealed ultrastructural characteristics typical for medial septal neurons obtained from controls. Our results indicate that target elimination during development affected both types of projection cells, although only the cholinergic cells are known to be responsive to target-derived factors.

Transient up-regulation of ciliary neurotrophic factor receptor-alpha mRNA in axotomized rat septal neurons

Using non-radioactive in situ hybridization we investigated the effect of fimbria-fornix transection on the expression of ciliary neurotrophic factor receptor alpha (CNTFR alpha) mRNA in axotomized septohippocampal neurons of the rat septal complex. Whereas CNTFR alpha expression was undetectable in the medial septal nucleus/diagonal band complex (MSDB) of control animals, specific up-regulation was observed in MSDB neurons after fimbria-fornix transection. CNTFR alpha expression was maximal 7-10 days after the lesion and had returned to control levels after 3 weeks. Following unilateral fimbria-fornix transection, CNTFR alpha up-regulation was restricted to the MSDB ipsilateral to the lesion. When cholinergic septal neurons were selectively eliminated by immunolesioning with 192 IgG-saporin prior to fimbria-fornix transection, the lesion-induced expression of CNTFR alpha was still observed in many medial septal nucleus neurons. These results demonstrate that after fimbria-fornix transection CNTFR alpha expression is transiently induced in axotomized, non-cholinergic neurons of the medial septal nucleus, suggesting a postlesion function of locally supplied CNTF.

Survival, regeneration and sprouting of central neurons: the rat septohippocampal projection as a model

The septohippocampal projection was used to study the survival following axotomy, axonal regeneration, and sprouting of a defined group of central neurons. Septohippocampal projection neurons in adult rats were axotomized by bilateral lesions of the fimbria-fornix. Using prelabeling prior to axotomy, intracellular staining, electron microscopy, and immunocytochemical and in situ hybridization techniques, we were able to demonstrate that the majority of septohippocampal neurons survived after axotomy. At least in young postnatal rats, these axotomized neurons have the capacity to regenerate an axonal process that reinnervates its appropriate target tissue, the hippocampus. We demonstrated this by axotomizing young septohippocampal neurons and co-culturing them with sections of hippocampus. Septohippocampal neurons appear to retain their capacity for axonal growth in adulthood, since they are able to sprout within hippocampal layers partially denervated by removing entorhinal afferents. In this paradigm the terminals of septohippocampal neurons themselves were not lesioned. Our results point to a previously underestimated capacity of septohippocampal neurons for survival following axotomy, regeneration, and sprouting.

Multiple projections are unlikely to account for the survival of rat medial septal neurons after axotomy

Previous studies have demonstrated that numerous septohippocampal neurons survive axotomy disconnecting them from target-derived neurotrophins. One reason for survival could be that these neurons have additional projections to other targets supplying them with neurotrophic factor(s). We show that anterograde tracing indeed labeled additional targets of the medial septum (MS) in controls as well as in experimental animals. However, only very few MS neurons could be double-labeled by retrograde tracer injections into the hippocampus and these targets. Thus, for the majority of septohippocampal neurons multiple projections cannot account for their survival following axotomy.

Regeneration of the GABAergic septohippocampal projection in vitro

The formation of the GABAergic septohippocampal projection was studied in vitro. Slice cultures of the septal complex from young postnatal rats were prepared and co-cultivated with hippocampal slices for up to four weeks. Then, the anterogradely transported tracer Phaseolus vulgaris leucoagglutinin was injected into the septal culture and the labeled fibers were traced into the hippocampal culture. Some fibers were identified as originating from GABAergic septal cells by double-labeling with an antiserum against GABA using the postembedding immunogold procedure. Our results showed that double-labeled terminals of GABAergic septohippocampal neurons established symmetric synapses exclusively with GABA-positive dendrites in one out of five co-cultures, but also contacted numerous GABA-negative structures in the remaining four co-cultures. These findings, together with light microscopic data from sections double-stained for Phaseolus and parvalbumin, indicate that the high target selectivity of the GABAergic septohippocampal pathway for GABAergic interneurons in vivo is lost in most cases, at least under the present in vitro conditions. It is hypothesized that this may be due to an immaturity of the connection, the lack of axon-guiding factors or an expansion of the septohippocampal GABAergic fibers in the absence of many extrinsic afferents, including GABAergic fibers. The simultaneous occurrence of anterogradely labeled, but GABA-negative, septohippocampal terminals in the hippocampal target culture also suggests that the septohippocampal cholinergic projection developed in vitro, as was shown before in other studies. Since most septohippocampal neurons have to be axotomized for culture preparation, the present results indicate that GABAergic septohippocampal neurons from young postnatal rats survive axotomy and are capable of regenerating a septohippocampal projection, including the formation of characteristic GABAergic synapses on co-cultured hippocampal neurons. However, the characteristic target selectivity is rarely preserved.