Visual cortex in mammals is composed of many distinct areas that are linked by reciprocal connections to form a multilevel hierarchy. Ascending information is sent via forward connections from lower to higher areas and is thought to contribute to the emergence of increasingly complex receptive field properties at higher levels. Descending signals are transmitted via feedback connections from higher to lower areas and are believed to provide information about the context in which a stimulus appears, to contribute to modulation of visual responses by attention, and to play a role in memory processes. To determine whether forward and feedback pathways in rat visual cortex constitute distinct intracortical circuits, we have studied the distribution of reciprocal corticocortical inputs to pyramidal cells and gamma-aminobutyric acid (GABA)ergic interneurons. For this purpose, we chose forward and feedback connections between primary visual cortex and the secondary extrastriate lateromedial (LM) area as a model system. Pathways were traced with the axonal marker phaseolus vulgaris-leucoagglutinin. Labeled terminals were identified in the electron microscope, and GABA immunocytochemistry was used to identify the postsynaptic dendritic shafts of GABAergic interneurons. In both pathways, inputs to pyramidal cells were directed preferentially to dendritic spines and not to shafts. In the forward pathway, 90% of labeled inputs were distributed to pyramidal cells and 10% to interneurons. This proportion was similar to that of nearby unlabeled connections in the neuropil, indicating that forward connections are not selective for pyramidal cells or interneurons. In sharp contrast, feedback connections were significantly different from the unlabeled connections and supplied almost exclusively pyramidal cells (98%). Feedback inputs to GABAergic neurons were five times weaker (2%) relative to the forward direction. These structural differences suggest that disynaptic GABAergic inhibition is much stronger in forward than in feedback pathways. Recent physiological experiments have confirmed this prediction (Shao et al. [1995] Soc. Neurosci. Abstr., 21:1274) and we, therefore, conclude that relatively small anatomical differences in the microcircuitry can have important functional consequences. It remains an open question whether generally reciprocal interareal circuits at all levels of the cortical hierarchy are organized in similar fashion.

Parcellation of the mammalian cerebral cortex into distinct areas is essential for proper cortical function; however, the developmental program that results in the genesis of distinct areas is not fully understood. We examined the expression of members of the EphA family-the EphA receptor tyrosine kinases and the ephrin-A ligands-within the developing mouse cerebral cortex, with the aim of characterizing this component of the molecular landscape during cortical parcellation. We found that specific embryonic zones, such as the ventricular, subventricular, intermediate, subplate, and marginal zones, as well as the cortical plate, were positive for particular EphA genes early in corticogenesis (E12-E15). Along with this zone-selective expression, several genes (EphA3, EphA4, EphA5) were evenly expressed along the axes of the developing cortex, whereas one family member (EphA7) was expressed in a distinct anteroposterior pattern. Later in corticogenesis (E16-E18), other EphA family members became selectively expressed, but only within the cortical plate: EphA6 was present posteriorly, and ephrin-A5 was expressed within a middle region. At birth, patterning of EphA gene expression was striking. Thus, we found that the expression of a single EphA gene or a combination of family members can define distinct embryonic zones and anteroposterior regions of the neocortex during development. To examine whether cellular context affects the patterning of EphA expression, we examined gene expression in embryonic cortical cells grown in vitro, such that all cellular contacts are lacking, and in Mash-1 mutant mice, in which thalamocortical connections do not form. We found that the expression patterns of most EphA family members remained stable in these scenarios, whereas the pattern of ephrin-A5 was altered. Taken together, this work provides a comprehensive picture of EphA family expression during mouse corticogenesis and demonstrates that most EphA expression profiles are cell intrinsically based, whereas ephrin-A5 is plastically regulated.

Feedback connections from extrastriate cortex to primary visual cortex (V1) in the primate may provide "top-down" information that plays a role in visual attention and object recognition. Our work in a rodent model of corticocortical circuitry demonstrates that feedback pathways synapse preferentially with pyramidal cells in V1 (Johnson and Burkhalter, 1996) and favor excitation over inhibition in cortical microcircuits .To investigate the polysynaptic circuits activated by feedback inputs, we studied chains of neurons postsynaptic to feedback connections using a combination of axonal tract tracing and anterograde degeneration. This approach enabled independent labeling of local collaterals of forward-projecting neurons in V1 and feedback connections from extrastriate lateromedial (LM) visual area to V1. Postsynaptic targets were identified in the electron microscope after retrograde transport of biotinylated dextran amine (BDA) to identify dendrites of forward-projecting neurons (i.e., from V1 to LM) and postembedding immunogold labeling to identify GABAergic interneurons.The results show that feedback connections provide strong monosynaptic input to forward-projecting neurons in V1. These neurons in turn make local connections that preferentially form synapses with other pyramidal cells (ϳ97%), many of which were identified as forward-projecting neurons. This indicates that feedback pathways provide input directly to neurons which make the reciprocal forward connection, and that feedbackrecipient forward-projecting neurons are strongly interconnected. The function of these excitatory networks within V1 may be to amplify feedback activity and provide a circuit for modulation of striate cortical activity by top-down influences.

No abstract