Sigrun I. Korsching

The neurotrophic factor concept: a reexamination

The neurotrophic factor concept in its basic form envisages that innervated tissues produce a signal for the innervating neurons for the selective limitation of neuronal death occurring during development (Purves, 1986; Oppenheim, 1991). This concept arose several decades ago on the basis of the observation that experimental manipulation of the amount of target tissue could modulate the size of neuronal populations. By making the survival of neurons dependent on their target, nature would provide a means to match neuron and target cell populations. NGF, discovered in the 1950s, represents the first known molecular realization of the neurotrophic factor concept. NGF was found to regulate survival, neurite growth, and neurotransmitter production of a particular neuronal type, the sympathetic neurons of the PNS. NGF produced by target cells is specifically bound and internalized by sympathetic neurons, followed by retrograde axonal transport of NGF to the cell soma, where NGF exerts its effects via the cotransported receptor molecule (Levi-Montalcini, 1987; Thoenen et al., 1987). Strictly speaking, increased neurite growth and neurotransmitter production are not trophic effects; however, I will use the term “neurotrophic” in the extended meaning of enhancing neuronal differentiation as well as neuronal survival. It was expected that these results could be generalized to a model of multiple, mutually independent, retrograde trophic messengers, which are synthesized in distinct target areas and act on restricted neuronal types (Fig. 1). This assumption leads to a conceptually simple way to arrange and maintain a variety of neuronal subsystems. One might call this a modular approach to the construction of the nervous system. The hypothesis of multiple retrograde signals has gained widespread experimental support in recent years. Originally proposed for the PNS, the model could be extended to the CNS, in which target neurons synthesize trophic factors for their afferent neurons (Ernfors et al., 1990b). In addition to NGF, a family of NGF-related molecules (now commonly called neurotrophins), which are thought to exert retrograde trophic influences (DiStefano et al., 1992), has been identified.

Combinatorial and Chemotopic Odorant Coding in the Zebrafish Olfactory Bulb Visualized by Optical Imaging

Odors are thought to be represented by a distributed code across the glomerular modules in the olfactory bulb (OB). Here, we optically imaged presynaptic activity in glomerular modules of the zebrafish OB induced by a class of natural odorants (amino acids [AAs]) after labeling of primary afferents with a calcium-sensitive dye. AAs induce complex combinatorial patterns of active glomerular modules that are unique for different stimuli and concentrations. Quantitative analysis shows that defined molecular features of stimuli are correlated with activity in spatially confined groups of glomerular modules. These results provide direct evidence that identity and concentration of odorants are encoded by glomerular activity patterns and reveal a coarse chemotopic organization of the array of glomerular modules.

Chemotopic, Combinatorial, and Noncombinatorial Odorant Representations in the Olfactory Bulb Revealed Using a Voltage-Sensitive Axon Tracer

Odor information is first represented in the brain by patterns of input activity across the glomeruli of the olfactory bulb (OB). To examine how odorants are represented at this stage of olfactory processing, we labeled anterogradely the axons of olfactory receptor neurons with the voltage-sensitive dye Di8-ANEPPQ in zebrafish. The activity induced by diverse natural odorants in afferent axons and across the array of glomeruli was then recorded optically. The results show that certain subregions of the OB are preferentially activated by defined chemical odorant classes. Within these subregions, “ordinary” odorants (amino acids, bile acids, and nucleotides) induce overlapping activity patterns involving multiple glomeruli, indicating that they are represented by combinatorial activity patterns. In contrast, two putative pheromone components (prostaglandin F2α and 17α,20β-dihydroxy-4-pregnene-3-one-20-sulfate) each induce a single focus of activity, at least one of which comes from a single, highly specific and sensitive glomerulus. These results indicate that the OB is organized into functional subregions processing classes of odorants. Furthermore, they suggest that individual odorants can be represented by “combinatorial” or “noncombinatorial” (focal) activity patterns and that the latter may serve to process odorants triggering distinct responses such as that of pheromones.

Quantitative demonstration of the retrograde axonal transport of endogenous nerve growth factor

The level of endogenous nerve growth factor (NGF) in rat sciatic nerve was determined using a highly sensitive two-site enzyme immunoassay. After crushing this nerve NGF accumulated linearly distal to the crush during the first 12 h to reach levels 13-fold higher than in the uncrushed contralateral side. In contrast, proximal to the crush NGF levels approached or were below the detection limit of the assay. The asymmetrical distribution of NGF on the two sides of a crush is direct evidence for the retrograde axonal transport of endogenous NGF.

Cholinergic denervation of the rat hippocampus by fimbrial transection leads to a transient accumulation of nerve growth factor (NGF) without change in mRNANGF content

Unilateral stereotaxic fimbrial transection was used to disrupt the cholinergic pathway from the medial septum to the hippocampus in adult rats. The resulting alterations in hippocampal nerve growth factor (NGF) and mRNANGF levels were then analysed by two-site enzyme immunoassay and quantitative Northern blots, respectively. NGF levels did not change up to one week after lesion but had increased to 145% of contralateral control values by two weeks. In contrast, mRNANGF levels were not changed at this time point. Neither bilateral fimbrial transection, nor removal of the sympathetic innervation of hippocampal vasculature by excision of the superior cervical ganglia, led to more pronounced increases in NGF levels. These results are consistent with the concept that the magnocellular cholinergic neurons in the medial septum are constantly reducing hippocampal NGF levels by retrograde axonal transport.

Developmental changes of nerve growth factor and its mRNA in the rat hippocampus: comparison with choline acetyltransferase

Previous experiments have demonstrated that in the septo-hippocampal system choline acetyltransferase (ChAT) is induced by nerve growth factor (NGF) (Gnahn et al. (1983) Dev. Brain Res. 9, 45-52) and that hippocampal NGF and mRNANGF levels are correlated with the density of cholinergic innervation (Korsching et al. (1985) EMBO J. 4, 1389-1393). In the present investigation we have compared the developmental changes of ChAT, NGF, and mRNANGF levels in this system. During the postnatal development of the hippocampus the time courses of NGF and ChAT were well correlated including the most rapid increase between P12 and P14. This increase in hippocampal NGF was preceded by a corresponding increase in mRNANGF. The developmental changes in hippocampal NGF levels were also closely reflected by corresponding changes in the septum. This, together with previous observations (Korsching et al., 1985) that the adult septum, in spite of relatively high NGF levels, does not contain measurable quantities of mRNANGF, suggests that the NGF levels in the septum are determined by the quantity of NGF transported retrogradely from the field of innervation rather than by local synthesis. During the prenatal period hippocampal NGF levels were relatively high, whereas the mRNANGF was below the level of detection. Since the ingrowth of septal fibers, and with that also the removal of NGF by retrograde transport, begins around birth, the relatively high prenatal NGF levels probably result from an accumulation produced by a small copy number of mRNANGF prior to the removal of NGF by retrograde axonal transport. It is concluded that the correlation of the developmental changes in NGF and mRNANGF with the ChAT activity in the hippocampus further supports the concept of a physiological role of NGF in the central nervous system.

The role of nerve growth factor in the CNS

Abstract In contrast to the peripheral nervous system, catecholaminergic neurons in the central nervous system do not respond to nerve growth factor (NGF). However, the cholinergic magnocellular neurons of the basal forebrain do respond to exogenous NGF by producing increased levels of the enzyme choline acetyltransferase (CAT). In addition, levels of endogenous NGF in brain are highest in those regions where the cell bodies or terminals of magnocellular cholinergic neurons are located. NGF mRNA is only found in the regions to which these neurons project, indicating that their target areas synthesize NGF that is then taken up and transported retrogradely to the magnocellular perikarya. NGF may also support other CNS neurons, such as the striatal cholinergic interneurons. Reduction in NGF levels may possibly be involved in the development of certain dementias such as Alzheimers disease.

Developmental changes of nerve growth factor levels in sympathetic ganglia and their target organs

The predominant source of nerve growth factor (NGF) used by mature sympathetic neurons originates in their target organs (Heumann, R., Korsching, S., Scott, J., and Thoenen, H. (1984), EMBO J. 3, 3183-3189; Korsching, S., and Thoenen, H. (1985), J. Neurosci. 5, 1058-1061). We have determined the NGF content of two sympathetically innervated mouse organs, submandibular gland and heart ventricle, and of sympathetic ganglia from mouse and rat between embryonic Day 12 (E12) and adulthood. NGF levels were measured by a two-site enzyme immunassay (Korsching, S., and Thoenen, H. (1983), Proc. Natl. Acad. Sci. USA 80, 3513-3516). In heart ventricle and submandibular gland, NGF first became detectable around the time of initial innervation by sympathetic neurons (E12 and E13, respectively) and increased respectively 14- and 7-fold in the following 2 days, to reach adult levels already at E14 for heart ventricle (1.4 +/- 0.2 ng NGF/g wet wt). NGF in the superior cervical ganglion (SCG) was first detected at the same time as in its target organ, the submandibular gland. NGF content in the SCG then increased 6-fold during the next 2 days and continued to increase until the end of the third postnatal week, when adult levels were reached. Although the levels of NGF in the adult mouse submandibular gland are sexually dimorphic and six orders of magnitude higher than those in other sympathetic target organs, no sex difference in the NGF content was found in either developing submandibular gland or SCG until the end of the third postnatal week. Moreover, the steep NGF increase observed in the male submandibular gland after postnatal Day 18 (250-fold within the following 3 days and up to the 55,000-fold in the next 7 days) was not reflected in a corresponding increase in the NGF content of the male SCG. These data indicate that, in accordance with earlier findings (see Levi-Montalcini, R., and Angeletti, P. U. (1968), Physiol. Rev. 48, 534-569), SCG neurons do not have access to the large amounts of NGF synthesized during and after adolescence in the mouse submandibular gland. Our results support the concept that initial fiber outgrowth of sympathetic neurons is neither dependent on NGF nor mediated by it. The time course of NGF levels in the SCG is consistent with the concept that sympathetic neurons are provided with NGF by means of retrograde axonal transport from the innervated organs already early in development.

Nerve growth factor supply for sensory neurons: Site of origin and competition with the sympathetic nervous system

Nerve growth factor (NGF) levels in the sensory nervous system were measured by a highly sensitive two-site enzyme immunoassay for NGF. Dorsal root ganglia and the adjacent spinal nerves contained 2.8 +/- 0.3 and 1.7 +/- 0.4 ng NGF/g wet wt., respectively, whereas no NGF was detectable in dorsal roots and spinal cord (less than 0.05 ng NGF/g wet wt.). It is concluded that sensory neurons are supplied with NGF exclusively from their peripheral and not from their central field of projection. Two days after treatment with 6-hydroxydopamine, which destroys sympathetic nerve terminals and thereby prevents the removal of NGF by sympathetic neurons, the NGF content of dorsal root ganglia and trigeminal ganglia increased to 285% and 161% of control, respectively. This indicates that in peripheral target tissues sensory and sympathetic neurons compete for NGF.

Olfactory maps and odor images

The understanding of neuronal processing of olfactory stimuli has been furthered by genetic studies and specialized imaging of particular neuronal populations. Selective optical imaging of odor-induced presynaptic and postsynaptic glomerular activity in the olfactory bulb/antennal lobe has visualized odorant-responsive receptor repertoires and shown a more confined odor image at the level of projection neurons compared to their olfactory receptor neuron input. Genetic tracing of projection neurons connected to particular glomeruli has revealed a somewhat dispersed spatial map of termination areas for these neurons both in insects and in vertebrates. Modifications of the glomerular odor map have resulted in altered percepts of the corresponding odors.