Afferent and efferent connections of the secondary general visceral sensory nucleus in goldfish.

The secondary general visceral sensory nucleus (SVN) receives ascending fibers from the commissural nucleus of Cajal (NCC), or the primary general visceral sensoru in the medulla oblongata of teleosts. However, the full set of fiber connections of the SVN have been studied only in the Nile tilapia. We have investigated the connections of the SVN in goldfish by tracer injection experiments to the nucleus. We paid special attention to the possible presence of spinal afferents, since the spinal cord projects to the lateral parabrachial nucleus, or the presumed homologue of SVN, in mammals. We found that the SVN indeed receives spinal projections. Spinal terminals were restricted to a region ventrolaterally adjacent to the terminal zone of NCC fibers, suggesting that the SVN can be subdivided into two subnuclei: the commissural nucleus-recipient (SVNc) and spinal-recipient (SVNsp) subnuclei. Tracer injections to the SVNc and SVNsp as well as reciprocal injections to the diencephalon revealed that both subnuclei project directly to diencephalic structures, such as the posterior thalamic nucleus and nucleus of lateral recess, although diencephalic projections of the SVNsp were rather sparse. The SVNsp appears to send fibers to more wide-spread targets in the preoptic area than the SVNc does. The SVNc projects to the telencephalon, while the SVNsp sends scarce or possibly no fibers to the telencephalon. Another notable difference was that the SVNsp gives rise to massive projections to the dorsal diencephalon (ventromedial thalamic, central posterior thalamic, and periventricular posterior tubercular nuclei). These differential connections of the subnuclei may reflect discrete functional significances of the general visceral sensory information mediated by the medulla oblongata and spinal cord.

Retrograde Transgene Expression via Neuron-Specific Lentiviral Vector Depends on Both Species and Input Projections.

For achieving retrograde gene transfer, we have so far developed two types of lentiviral vectors pseudotyped with fusion envelope glycoprotein, termed HiRet vector and NeuRet vector, consisting of distinct combinations of rabies virus and vesicular stomatitis virus glycoproteins. In the present study, we compared the patterns of retrograde transgene expression for the HiRet vs. NeuRet vectors by testing the cortical input system. These vectors were injected into the motor cortex in rats, marmosets, and macaques, and the distributions of retrograde labels were investigated in the cortex and thalamus. Our histological analysis revealed that the NeuRet vector generally exhibits a higher efficiency of retrograde gene transfer than the HiRet vector, though its capacity of retrograde transgene expression in the macaque brain is unexpectedly low, especially in terms of the intracortical connections, as compared to the rat and marmoset brains. It was also demonstrated that the NeuRet but not the HiRet vector displays sufficiently high neuron specificity and causes no marked inflammatory/immune responses at the vector injection sites in the primate (marmoset and macaque) brains. The present results indicate that the retrograde transgene efficiency of the NeuRet vector varies depending not only on the species but also on the input projections.

A note on retrograde gene transfer efficiency and inflammatory response of lentiviral vectors pseudotyped with FuG-E vs. FuG-B2 glycoproteins.

Pseudotyped lentiviral vectors give access to pathway-selective gene manipulation via retrograde transfer. Two types of such lentiviral vectors have been developed. One is the so-called NeuRet vector pseudotyped with fusion glycoprotein type E, which preferentially transduces neurons. The other is the so-called HiRet vector pseudotyped with fusion glycoprotein type B2, which permits gene transfer into both neurons and glial cells at the injection site. Although these vectors have been applied in many studies investigating neural network functions, it remains unclear which vector is more appropriate for retrograde gene delivery in the brain. To compare the gene transfer efficiency and inflammatory response of the NeuRet vs. HiRet vectors, each vector was injected into the striatum in macaque monkeys, common marmosets, and rats. It was revealed that retrograde gene delivery of the NeuRet vector was equal to or greater than that of the HiRet vector. Furthermore, inflammation characterized by microglial and lymphocytic infiltration occurred when the HiRet vector, but not the NeuRet vector, was injected into the primate brain. The present results indicate that the NeuRet vector is more suitable than the HiRet vector for retrograde gene transfer in the primate and rodent brains.

High-Speed and Scalable Whole-Brain Imaging in Rodents and Primates.

Subcellular resolution imaging of the whole brain and subsequent image analysis are prerequisites for understanding anatomical and functional brain networks. Here, we have developed a very high-speed serial-sectioning imaging system named FAST (block-face serial microscopy tomography), which acquires high-resolution images of a whole mouse brain in a speed range comparable to that of light-sheet fluorescence microscopy. FAST enables complete visualization of the brain at a resolution sufficient to resolve all cells and their subcellular structures. FAST renders unbiased quantitative group comparisons of normal and disease model brain cells for the whole brain at a high spatial resolution. Furthermore, FAST is highly scalable to non-human primate brains and human postmortem brain tissues, and can visualize neuronal projections in a whole adult marmoset brain. Thus, FAST provides new opportunities for global approaches that will allow for a better understanding of brain systems in multiple animal models and in human diseases.

The use of an optimized chimeric envelope glycoprotein enhances the efficiency of retrograde gene transfer of a pseudotyped lentiviral vector in the primate brain.

Lentiviral vectors have been used not only for various basic research experiments, but also for a wide range of gene therapy trials in animal models. The development of a pseudotyped lentiviral vector with the property of retrograde infection allows us to introduce foreign genes into neurons that are localized in regions innervating the site of vector injection. Here, we report the efficiency of retrograde gene transfer of a recently developed FuG-E pseudotyped lentiviral vector in the primate brain by comparing its transduction pattern with that of the parental FuG-C pseudotyped vector. After injection of the FuG-E vector encoding green fluorescent protein (GFP) into the striatum of macaque monkeys, many GFP-immunoreactive neurons were found in regions projecting to the striatum, such as the cerebral cortex, thalamus, and substantia nigra. Quantitative analysis revealed that in all regions, the number of neurons retrogradely transduced with the FuG-E vector was larger than in the FuG-C vector injection case. It was also confirmed that the FuG-E vector displayed explicit neuronal specificity to the same extent as the FuG-C vector. This vector might promote approaches to pathway-selective gene manipulation and provide a powerful tool for effective gene therapy trials against neurological disorders through enhanced retrograde delivery.

Connections of the commissural nucleus of Cajal in the goldfish, with special reference to the topographic organization of ascending visceral sensory pathways.

The primary general visceral nucleus of teleosts is called the commissural nucleus of Cajal (NCC). The NCC of goldfish has been divided into the medial (NCCm) and lateral (NCCl) subnuclei that receive inputs from subdiaphragmatic gastrointestinal tract and the posterior pharynx, respectively. Fiber connections of the NCC were examined by tract-tracing methods in the goldfish Carassius auratus. Tracer injections into the NCC suggested that the NCC projects directly not only to the secondary visceral sensory region in the rhombencephalic isthmus and other brain stem centers, but also to the forebrain, similar to the situations in mammals, birds, and the Nile tilapia. Although fiber connections of the NCCm and NCCl were basically similar, the NCCm was the more important source of ascending general visceral fibers to the forebrain. Topographic organization was recognized regarding projections to the isthmic secondary visceral sensory zone; input from the NCCm is represented in the secondary general visceral sensory nucleus, while input from the NCCl in the lateral edge of the secondary gustatory nucleus. Moreover, specific injections into different regions of the vagal lobe revealed that the dorsomedio-ventrolateral axis of the lobe is represented in the lateromedial axis of the secondary gustatory nucleus. These observations suggest fine topographic organization of ascending visceral sensory pathways to the isthmic secondary centers. It should also be noted that the reception of primary afferents from the posterior pharynx and projections to the secondary gustatory nucleus suggest that the NCCl may be regarded as a gustatory rather than a general visceral sensory structure.