Pattern of corticostriatal innervation in organotypic cocultures is dependent on the age of the cortical tissue

Abstract

The patch–matrix organization of the striatum is defined by the selective expression of neuronal markers and a semisegregated pattern of afferents and efferents that develops before birth in all mammals. Differential projections from ‘limbic’ and ‘somatomotor’ cortices contribute to the selective circuitry of patch (“striosome”) and matrix compartments. Organotypic cultures were used to determine the pattern of early corticostriatal innervation as a first step toward understanding the role of cortical innervation in the development of striatal patch–matrix organization. Perinatal striatum (E19–P4) was cocultured with the cortex obtained from same-age or different-age rats in the presence or absence of substantia nigra obtained from E14–15 fetuses. After 4–21 days in vitro, crystals of biocytin were placed directly onto the cortical piece to trace cortical projections into the striatal piece. Cortex obtained from fetuses (E19–22) or neonatal (P0–1) rats gave rise to a dense innervation of both prenatal and postnatal striatal slices; however, the pattern of biocytin-labeled fibers was found to be highly dependent on the age of the cortical tissue used. Cortex derived from rats between E20 and P1 gave rise to a heterogeneous distribution of fibers indicative of striatal patches when combined with striatal slices from same-age or younger (E18–19) fetuses. Cortex from E18–19 fetuses produced a homogeneous innervation even when cocultured with older striatal tissue in which the striatal patches were already present. The postnatal cortex (P2–P5) gave rise to little to no innervation of striatum of all ages. Similar findings were obtained with the use of either prelimbic or somatosensory cortex. In double- and triple-labeled cultures, the distribution of corticostriatal fibers overlapped substantially with patches of developing striatal neurons, as revealed by DARPP-32 immunocytochemistry. Dopaminergic innervation present when the substantia nigra was included in the cocultures also distributed preferentially to the developing patch compartment, but it did not substantially alter the pattern of corticostriatal innervation. These findings suggest that the cortex provides directive signals to the developing striatum rather than simply responding to the presence of patches that have already formed.

Introduction

The mammalian striatum (caudate–putamen in the human) is a region of immense complexity in terms of the richness of neuroactive substances and the number of anatomic circuits that interdigitate within this region. The compartmentalization of the striatum into irregularly shaped “patches” or “striosomes” within the surrounding “matrix” has been established from the substances expressed, which must be revealed by immunocytochemistry Gerfen, 1984, Gerfen and Wilson, 1996, Graybiel et al., 1981. The relative segregation of inputs and outputs among these compartments is an additional anatomic feature that contributes to the segregation of information flow through the striatum Donoghue and Herkenham, 1986, Gerfen, 1984, Gerfen, 1985, Kincaid and Wilson, 1996, Ragsdale and Graybiel, 1990.

Study of the factors that drive the formation of the patch and matrix compartments has been complicated by the fact that such compartmentalization develops before birth in all mammals (Johnston et al., 1990). Nigrostriatal dopaminergic innervation is also known to develop early, concentrating in developing patch regions by embryonic day (E) 20 in the rat Snyder-Keller, 1991, Voorn et al., 1988. Much less is known about the earliest projections from cortex, which are likely to make up the primary excitatory drive onto the developing striatum. Goldman-Rakic (1981) first described a transition from an initially homogeneous to a patchy distribution of terminals anterogradely traced from prefrontal cortex in the perinatal monkey. Similar anterograde-tracing studies performed in rats Christensen et al., 1999, Snyder-Keller et al., 2003a and mice Nisenbaum et al., 1998, Sheth et al., 1998 have demonstrated that single corticostriatal fibers are present before birth, although significant arborizations within patches are not present until a few days after birth.

To study the role of afferent innervation in the development of patch (striosome) and matrix compartments of the striatum, we have utilized organotypic cultures of the perinatal striatum. In our earlier studies, we demonstrated by immunocytochemical staining that slices taken from E19 fetuses initially exhibit a homogeneous distribution of many patch markers, which transition to a patchy distribution if cocultured with a source of afferent innervation. Most notably, patch formation in E19 striatal cultures was promoted by coculturing with dopamine (DA)-containing tissue from E14 substantia nigra, which gave rise to patchy DA innervation (Snyder-Keller et al., 2001). Although, in our initial studies, coculturing of striatum with cortex did not seem to promote patch formation, the wealth of evidence for a role of excitatory glutamatergic afferents in pattern formation in other regions led us to further pursue the issue of cortical influences. One critical problem with our earlier work was that we had no means of effectively visualizing the corticostriatal innervation that developed in organotypic cocultures, as we had for the dopaminergic innervation that resulted in nigral–striatal cocultures. Recently, Takuma et al. (2002) reported the use of biocytin crystals, inserted into a tissue piece grown in organotypic cocultures, as a useful means of anterograde labeling of projections from one piece to another. Thus, in the present study, we utilized this technique to label the corticostriatal projections that develop in organotypic cultures of cortex and striatum taken from fetuses either before or after in vivo patch formation, and we determined the pattern of ingrowth in relation to the developing patch compartment. The pattern of innervation revealed was found to change dramatically as a function of the age of the cortical tissue. Furthermore, the findings obtained were suggestive of an active role of cortical afferents in the process of striatal patch formation.