Activating transcription factor 3 induction in sympathetic neurons after axotomy: response to decreased neurotrophin availability.

Activating transcription factor 3 (ATF3) is induced in a high proportion of axotomized sensory and motor neurons after sciatic nerve transection. In the present study, we looked at the expression of this factor in the superior cervical ganglion (SCG) after axotomy and after other manipulations that induce certain aspects of the cell body response to axotomy. Sympathetic ganglia from intact rats and mice exhibit only a very occasional neuronal nucleus with activating transcription factor 3-like immunoreactivity (ATF3-IR); however, as early as 6 h and as late as 3 weeks postaxotomy, many of the neurons showed intense ATF3-IR. A second population of cells had smaller and generally less intensely stained nuclei, and at least some of these cells were satellite cells. Lesions distal to the SCG induced by administration of 6-hydroxydopamine or unilateral removal of the salivary glands produced increases in ATF3-IR similar to those seen after proximal axotomy, indicating that this response is not strictly dependent on the distance of the lesion from the cell body. Two proposed signals for triggering ATF3 expression were examined: reduction in nerve growth factor (NGF) availability and induction of the cytokine leukemia inhibitory factor (LIF). While administration of an antiserum raised against NGF to intact animals induced ATF3-IR, induction of ATF3-IR after axotomy was not reduced in LIF null mutant mice. Since axotomy, 6-hydroxydopamine, and sialectomy are known to decrease the concentration of NGF in the SCG, our data suggest that these decreases in NGF lead to increases in ATF3-IR. Furthermore, since the number of neurons in the SCG expressing ATF3-IR was greater after axotomy than after antiserum against NGF treatment, this raises the possibility that decreased NGF is not the only process regulating ATF3 expression after axotomy.

Galanin expression increases in adult rat sympathetic neurons after axotomy.

Changes in neuropeptide expression occur in sensory, motor, and sympathetic neurons following axotomy. The particular pattern of peptide changes that occurs varies among the three cell types. We have studied the regulation in the rat superior cervical ganglion of the expression of galanin, a peptide previously shown to increase in axotomized sensory and motor neurons. While normally only an occasional neuron exhibiting galanin-like immunoreactivity is found in this ganglion, at two days after transection of the postganglionic internal and external carotid nerves, immunostaining can be observed in many neurons throughout the ganglion. Similar changes are found when ganglia are placed in organ culture for two days. The distribution of immunostained neurons after section of only one of the postganglionic trunks suggests that changes in galanin-like immunoreactivity occur only within neurons whose axons are transected. None the less, even when both nerve trunks are transected, only about half of the neurons in the ganglion exhibit galanin-like immunoreactivity, indicating that only a proportion of the axotomized neurons exhibit a detectable response. The few immunostained neurons seen after section of the cervical sympathetic trunk may also represent axotomized neurons. Galanin-like immunoreactivity extracted from the ganglion co-chromatographs with authentic galanin, and the level of this immunoreactivity increases dramatically after axotomy and explantation, and modestly after decentralization. These same manipulations produce parallel increases in the level of galanin messenger RNA. Together, the findings indicate that the expression of galanin increases in sympathetic neurons after axotomy. Galanin is thus the first neuropeptide whose expression has been shown to increase after transection of all three types of peripheral axons that have been studied.

The conditioning lesion effect on sympathetic neurite outgrowth is dependent on gp130 cytokines.

Sympathetic neurons, like sensory neurons, increase neurite outgrowth after a conditioning lesion. Studies in leukemia inhibitory factor (LIF) knockout animals showed that the conditioning lesion effect in sensory neurons is dependent in part on this cytokine; however, similar studies on sympathetic neurons revealed no such effect. Comparable studies with sensory neurons taken from mice lacking the related cytokine interleukin-6 (IL-6) have yielded conflicting results. LIF and IL-6 belong to a family of cytokines known as the gp130 family because they act on receptors containing the subunit gp130. In sympathetic ganglia, axotomy leads to increases in mRNA for four of these cytokines (LIF, IL-6, IL-11, and oncostatin M). To test the role of this family of cytokines as a whole in the conditioning lesion response in sympathetic neurons, mice in which gp130 was selectively eliminated in noradrenergic neurons were studied. The postganglionic axons of the SCG were transected, and 7days later the ganglia were removed and neurite outgrowth was measured in explant and dissociated cell cultures. In both systems, neurons from wild type animals showed enhanced growth after a conditioning lesion. In contrast, no enhancement occurred in neurons from mutant animals. This lack of stimulation of outgrowth occurred despite an increase in expression of activating transcription factor 3 (ATF3) in the mutant mice. These studies demonstrate that stimulation of enhanced growth of sympathetic neurons after a conditioning lesion is dependent on gp130 cytokine signaling and is blocked in the absence of signaling by these cytokines in spite of an increase in ATF3.

Chemical sympathectomy and postganglionic nerve transection produce similar increases in galanin and VIP mRNA but differ in their effects on peptide content.

Large changes in neuronal gene expression occur in adult peripheral neurons after axonal transection. In the rat superior cervical ganglion, for example, neurons that do not normally express vasoactive intestinal peptide (VIP) or galanin do so after postganglionic nerve transection. These effects of axotomy could result from a number of aspects of the surgical procedure. To test the idea that the important variable might be the disconnection of axotomized neuronal cell bodies from their target tissues, we examined the effects of producing such a disconnection by means of the compound 6-hydroxydopamine (6-OHDA), a neurotoxin that causes degeneration of sympathetic varicosities and avoids many of the complications of surgery. Two days after 6-OHDA treatment, VIP and galanin immunoreactivities had increased two- and 40-fold, respectively. Nevertheless, these increases were substantially smaller than the 30- and 300-fold changes seen after surgical axotomy. When expression of VIP and galanin was examined at the mRNA level, however, comparable increases were found after either procedure. The results indicate that chemical destruction of sympathetic varicosities produces an equivalent signal for increasing VIP and galanin mRNA as does axonal transection. The differences in the neuropeptide levels achieved suggests that peptide expression after nerve transection is regulated both at the mRNA and protein levels.

Phenotypic plasticity in adult sympathetic ganglia in vivo: effects of deafferentation and axotomy on the expression of vasoactive intestinal peptide.

The expression of neurotransmitters/neuromodulators in sympathetic neurons is regulated by anterograde and retrograde mechanisms. We have examined the role of such mechanisms in the regulation of the neuropeptide vasoactive intestinal peptide (VIP). The adult rat superior cervical ganglion (SCG) contains low levels of peptide-like immunoreactivity (IR) and mRNA for VIP. Some VIP-IR nerve processes, but only a few VIP-IR cell bodies, are detectable. Previous evidence demonstrates, however, that after the SCG is placed in organ culture for 48 hr, the level of VIP-IR and VIP mRNA and the number of VIP-IR cell bodies and fibers increase considerably. Two of the possible causes for these changes in peptide expression in sympathetic neurons are deafferentation and axotomy, both of which occur when the SCG is placed in culture. To determine the importance of deafferentation, the preganglionic cervical sympathetic trunk was cut and the ganglion left in situ. Forty-eight hours later, VIP-IR increased twofold. A corresponding increase in the number of VIP-IR nerve processes was seen, but there was no detectable change in the number of VIP-IR cell bodies. The content of VIP/PHI mRNA also increased by 1.8-fold. The effect of axotomy on VIP-IR was examined by cutting the postganglionic internal and external carotid nerves and leaving the ganglion in situ. Forty-eight hours later, the level of VIP-IR increased 22-fold, many immunostained neurons were found, and the content of VIP mRNA increased over fivefold. After either deafferentation or axotomy, changes in VIP-IR were accompanied by comparable changes in the related molecule peptide histidine isoleucine amide (PHI)-IR. Neuropeptide Y-IR, on the other hand, decreased after deafferentation and increased only twofold after axotomy. The results indicate plasticity in the expression of VIP- and PHI-IR in adult sympathetic neurons in vivo, and suggest that the changes previously seen in organ culture were primarily a response to axotomy.

Phenotypic plasticity in adult sympathetic neurons: changes in neuropeptide expression in organ culture.

Vasoactive intestinal peptide (VIP)-like immunoreactivity is present at low levels in the superior cervical ganglion of the adult rat, where immunostained neural processes, but only an occasional immunostained cell body, are found. However, when ganglia are maintained for 24 or 48 hr in organ culture, their content of VIP-like immunoreactivity increases 6- or 31-fold, respectively. When examined at 24 hr, the increase in VIP-like immunoreactivity is totally blocked by an inhibitor of RNA or protein synthesis. Many neuronal cell bodies and processes with immunoreactivity for VIP and the related peptide histidine isoleucine amide (PHI) are seen in cultured ganglia. In addition, VIP/PHI mRNA is abundant in cultured ganglia but only barely detectable in ganglia prior to culture. Under the same culture conditions, neuropeptide Y-like immunoreactivity increases to a small extent, and tyrosine hydroxylase activity and total ganglion protein remain unchanged. These results support the idea that adult sympathetic neurons exhibit plasticity in neuropeptide expression and that this plasticity, in the case of VIP, depends on changes in gene expression.

Changes in neuropeptide phenotype after axotomy of adult peripheral neurons and the role of leukemia inhibitory factor.

Adult peripheral neurons undergo dramatic shifts in gene expression following axotomy that are collectively referred to as the cell body reaction. Changes in neuropeptide expression are a prominent feature of these axotomized neurons. For example, while sympathetic, sensory, and motor neurons do not normally express the neuropeptides galanin and vasoactive intestinal peptide, they begin to do so within days after axotomy. In contrast, the expression of other peptides, which these neurons normally express, such as neuropeptide Y in sympathetic neurons and substance P in sensory neurons, is decreased. Recent studies in sympathetic neurons have demonstrated that leukemia inhibitory factor plays an important role in triggering these changes in neuropeptide phenotype in adult neurons. Future studies will be directed at determining to what extent LIF triggers the many other changes in gene expression after sympathetic axotomy and whether this cytokine plays a similar role in sensory and motor neurons.