poglossi involved with the coordination of eye movements. At this brain stem level, the immunocytochemical distribution patterns for the enzymes from both pathways showed a remarkable degree of similarity. However, SpdS did not obviously highlight neuronal cell bodies when compared to the three other enzymes, which were also prominently expressed in giant neurons of the reticular Salvianic acid A web formation and in the inferior olive. Previous experiments analyzing the fine structural localization of Agm in rat cerebral cortex neuropil demonstrated that the diffuse signal observed with standard 24634219 immunoperoxidase light microscopy correlates with a punctate pattern of immunoreactivity obtained when virtual pre-embedding involving a CARD signal amplification was used. We therefore postulated that the neuropil labelling observed here with antibodies against Arg and ADC may similarly be displayed at higher resolution using the VirP procedure with light microscopy. Indeed, in the cerebral cortex and hippocampus, numerous punctate profiles were observed. Within the neuropil, some of these punctae were clearly associated with dendritic profiles, suggesting either a presynaptic or postsynaptic localization of the antigen. Besides this neuropil labeling, both antibodies also labeled neuronal cell bodies as observed with standard DAB light microscopy. With electron microscopy visualizing DAB-based immunoperoxidase reactivity, ADC was localized to dendrites and dendritic spines, whereas Arg was detected in presynaptic terminals in a fraction of cortical 
synapses. Within neuronal cell bodies, ADC-like immunoreactivity was associated with ER-like and vesicular structures, while Arg was distinctly observed in Golgi compartments. Rarely, also labeled disrupted mitochondria were visualized with anti-Arg antibodies. Arginase and Arginine Decarboxylase in Rat Brain Discussion From theoretical reasoning there are two nominal pathways leading to the formation of putrescine, the starting point in spermidine/spermine synthesis: the “classical”pathway via ornithine and the “alternative”pathway via the breakdown of the putative neurotransmitter agmatine. The relevance of both pathways, however, has been challenged, primarily based on data concerning the two arginine-converting enzymes Arg and ADC. Arg, existing in two isoforms, apparently is not necessary to maintain polyamine homeostasis in Arg1/Arg2 single and double knockout mice. However, Arg1 knockout animals die at the age of two weeks due to hyperammonemia resulting from an impaired urea cycle. Dietary spermidine/ spermine and also agmatine were discussed to account for the maintenance of blood and tissue polyamine levels. In the brain 25939886 the situation may be different due to the bloodbrain barrier. Apparently, spermidine/spermine only have a very limited capacity to cross the blood-brain barrier. By contrast, agmatine is readily transported into the cerebrospinal fluid. Furthermore, intravenously administered arginine rather takes effect by acting as a cerebrospinal fluid than blood precursor of agmatine. Thus, the brain seems to be able to locally synthesize the putative neurotransmitter agmatine. Hence, the systemic effects observed with impaired arginine metabolism may obscure any brain specific effects. With respect to Arg1 and Arg2 isoforms, our data indicate that both Arg isoforms are expressed in the brain. Furthermore, two groups of immunocytochemically differentiated hippocampal interneurons were clearly different wi