The distinctive features of axons and dendrites divide most neurons into two compartments. This polarity is fundamental to the ability of most neurons to integrate synaptic signals and transmit action potentials. It is not known, however, if the polarity of neurons in the adult mammalian nervous system is fixed or plastic. Following axotomy, some distal dendrites of neck motoneurons in the adult cat give rise to unusual processes that, at a light microscopic level, resemble axons (Rose, P.K. & Odlozinski, M., J. Comp. Neurol., 1998, 390, 392). The goal of the present experiments was to characterize these unusual processes using well-established ultrastructural and molecular criteria that differentiate dendrites and axons. These processes were immunoreactive for growth-associated protein-43 (GAP-43), a protein that is normally confined to axons. In contrast, immunoreactivity for a protein that is widely used as a marker for dendrites, microtubule-associated protein (MAP)-2a/b, could not be detected in the unusual distal arborizations. At the electron microscopic level, unusual distal processes contained dense collections of neurofilaments and were frequently myelinated. These molecular and structural characteristics are typical of axons and suggest that the polarity of adult neurons in the mammalian nervous system can be disrupted by axotomy. If this transformation in neuronal polarity is common to other types of neurons, axon-like processes emerging from distal dendrites may represent a mechanism for replacing connections lost due to injury. Alternatively, the connections formed by these axons may be aberrant and therefore maladaptive.

Following proximal axotomy, several types of neurons sprout de novo axons from distal dendrites. These processes may represent a means of forming new circuits following spinal cord injury. However, it is not know whether mammalian spinal interneurons, axotomized as a result of a spinal cord injury, develop de novo axons. Our goal was to determine whether spinal commissural interneurons (CINs), axotomized by 3-4-mm midsagittal transection at C3, form de novo axons from distal dendrites. All experiments were performed on adult cats. CINs in C3 were stained with extracellular injections of Neurobiotin at 4-5 weeks post injury. The somata of axotomized CINs were identified by the presence of immunoreactivity for the axonal growth-associated protein-43 (GAP-43). Nearly half of the CINs had de novo axons that emerged from distal dendrites. These axons lacked immunoreactivity for the dendritic protein, microtubule-associated protein2a/b (MAP2a/b); some had GAP-43-immunoreactive terminals; and nearly all had morphological features typical of axons. Dendrites of other CINs did not give rise to de novo axons. These CINs did, however, each have a long axon-like process (L-ALP) that projected directly from the soma or a very proximal dendrite. L-ALPs were devoid of MAP2a/b immunoreactivity. Some of these L-ALPs projected through the lesion and formed bouton-like swellings. These results suggest that proximally axotomized spinal interneurons have the potential to form new connections via de novo axons that emerge from distal dendrites. Others may be capable of regeneration of their original axon.

Following axotomy, morphologically unusual, distal processes (UDPs) emerge from motoneuron dendrites. These processes contain an axonal protein, growth-associated protein 43 (GAP-43) but lack immunostaining for the dendritic protein microtubule-associated protein 2a/b (MAP2a/b). Thus, it appears that neuronal polarity alters following axotomy. Our goal was to describe this change in neuronal polarity on a more detailed and quantitative level. We asked two questions: Following axotomy, where in the entire neuron does the immunoreactivity for MAP2a/b and GAP-43 change and do these changes reflect a transformation of dendrite to axon or growth from terminal dendrites? Using intracellular labeling and immunocytochemistry, changes in MAP2a/b and GAP-43 immunoreactivity were also found in processes with a morphology typical of terminal branches of intact motoneurons (called simple distal processes [SDPs]), as well as UDPs. Trajectories (the path from the soma to a single terminus) with UDPs and SDPs were longer than trajectories without these processes, and trajectories with UDPs were the longest. Trajectories without UDPs or SDPs were similar in length to trajectories from intact motoneurons. The distance from the soma to the point where MAP2a/b immunoreactivity became absent in trajectories with UDPs or SDPs was similar to the length of trajectories from intact motoneurons. Thus, following axotomy, two morphologically distinct types of axon-like processes emerge from dendrites. The formation of these processes does not involve a transformation of the original dendrite, but rather growth at the ends of dendrites.

At 8-12 weeks post axotomy, unusual distal processes (UDPs) with axon-like structural (uniform diameter, tortuous) and molecular (growth-associated protein [GAP]43, absence of microtubule-associated protein [MAP]2a/b immunoreactivity) features emerge from distal motoneuron dendrites (Rose et al. [2001] Eur J Neurosci 13:1166-1176). In this study, we determine the time course of molecular and morphological changes associated with the formation of axons from dendrites. Motoneurons innervating neck muscles in the adult cat were permanently axotomized for 2, 4, 20, or 35 weeks and intracellularly stained with Neurobiotin. Computer-assisted reconstructions were used to map the location of MAP2a/b and GAP-43 immunoreactivity. At 2 and 4 weeks post axotomy, all UDPs had short appendages, giving them an arboreal appearance. They were immunoreactive for GAP-43 and lacked immunostaining for MAP2a/b. Axon-like UDPs were not seen until 8-12 weeks post axotomy. By 20 and 35 weeks post axotomy, some axon-like UDPs acquired morphological features of axons with synaptic connections (right-angled branching, bouton-like specializations). GAP-43 immunoreactivity was not detected in any axotomized motoneurons by 20 weeks post axotomy, whereas all UDPs remained devoid of MAP2a/b immunoreactivity even at 35 weeks post axotomy. These molecular changes accompanied structural modifications to proximal regions of "dendrites" giving rise to UDPs. The distance from the ends of the UDPs to the soma did not change. Thus, all UDPs begin as simple, arboreal structures with molecular features of growing axons, but over a period of 35 weeks, some UDPs slowly acquire morphological and molecular features of motoneuron axons with synaptic connections. These results suggest a new modus operandi for axonal growth and the establishment of new synaptic connections after injury.

In the central nervous system, regeneration of injured axons and sprouting of intact axons are suppressed by myelin-derived molecules that bind to the Nogo receptor (NgR). We used a soluble form of the NgR (sNgR), constructed as an IgG of the human NgR extracellular domain, to manipulate plasticity of uninjured primary afferent and descending monoaminergic projections to the rat spinal cord following dorsal rhizotomy. Rats with quadruple dorsal rhizotomies were treated with intrathecal sNgR or saline, or were left untreated for 2 weeks. Rhizotomy alone resulted in sprouting of serotonergic axons and to a lesser extent, tyrosine-hydroxylase (TH)-expressing axons, while axons expressing dopamine-beta-hydroxylase (DbetaH) were unaffected. Human IgG immunohistochemistry revealed that sNgR infused into the intrathecal space penetrated approximately 300 microm into spinal white and grey matter. Separate axonal populations differed in their responses to intrathecal sNgR: TH-expressing and DbetaH-expressing axons responded most and least vigorously, respectively. Serotonergic axons were identified by serotonin (5-HT) or serotonin transporter (SERT) immunohistochemistry. Interestingly, a large increase in 5-HT compared to SERT-positive axons density in both saline and sNgR-treated rats indicated that serotonergic axons both sprouted and increased their transmitter content in response to rhizotomy and sNgR treatment. Calcitonin gene-related peptide-positive axons were largely depleted ipsilaterally by rhizotomy, and sNgR increased axon density only in deeper contralateral laminae (III-V). GAP-43 immunohistochemistry revealed a small increase in axon density following dorsal rhizotomy that was further augmented by sNgR treatment. These results reveal a differential effect of myelin antagonism on distinct populations of spinally projecting axons.