THE JOURNAL OF COMPARATIVE NEUROLOGY 372~327-342 (1996) Development of Connections Within and Between Areas

We have investigated the development of intrinsic and interareal connections in areas V1 and V2 of the macaque monkey using postmortem transport of the lipophilic fluorescent tracer diI, applied to brains fixed at different preand postnatal ages. Intrinsic connections in the deep layers of V1 are evident on embryonic day 108 (ElOS), but are not robust in the superficial layers until around E118, when migration is largely complete. Both intrinsic horizontal projections and extrinsic projections to V2 initially have a continuous distribution. Patchy projections are first evident in V1 around E145, the same age at which cytochrome oxidase blobs appear, presumably signaling the differentiation of the blob-dominated and interblob-dominated streams in the primary visual cortex. The magnocellular-dominated stream becomes distinct at earlier stages (by E122), as judged by connectional and histochemical criteria. In area V2, intrinsic connections initially (at E108) involve only deep layer cells and do not have a clustered organization. By E130, superficial layer cells are involved and the V2 intrinsic connections have a patchy distribution; by E145, an adult-like pattern is present. The projection from V2 to V l passes through an early stage (up to E133) of originating principally from deep layer cells, and thereafter originating from superficial as well as deep layers. We found evidence for changes in dendritic morphology during development. Most notably, at E118, many neurons in layer 6 which are involved in intrinsic or interareal connections have dendrites that extend well into the superficial layers, even into layer 1, a characteristic not reported in the adult. G? 1996 Wiley-Liss, Inc. Indexing terms: modularity, lamination, intrinsic connections The intricate circuitry of the mammalian neocortex is established over an extended period of prenatal and postnatal development. It involves a complex sequence of cell migration, elaboration of dendritic and axonal arbors in specific laminar patterns, and selective pruning of inappropriate connections. These processes have been most intensively studied in visual and somatosensory cortex, particularly in non-primate species (see Katz and Callaway, 1992; O’Leary et al., 1994). In the macaque monkey and other primates, visual cortex contains a number of unique anatomical features that are of interest from a developmental viewpoint. The lamination of area V1 is more highly elaborated than in other cortical areas, with layer 4 containing at least four distinct sublaminae (4A, 4B, 4Ca, 4Cp) that are related to the pattern of geniculocortical inputs (Hubel and Wiesel, 1977). In the tangential domain (i.e., parallel to the cortical surface), both V1 and V2 have a modular organization that is manifested, in part, by an array of blobs and interblobs in V1 and an array of thick stripes, thin stripes, and interstripes in V2, all of which are visualizable using cytochrome oxidase (CO) histochemistry (Horton, 1984; Livingstone and Hubel, 1984; Tootell et al., 1985). There is a high degree of specificity in the intrinsic connections within each area (Livingstone and Hubel, 1984; Blasdel et al., 1985; Fitzpatrick et al., 1985; Lachica et al., 1992; Yoshioka et al., 1994). In addition, the connections between V1 and V2 involve specific pathways between corresponding types of modules (Livingstone and Hubel, 1987a,b). Previous studies of the fetal development of visual cortex in the macaque have described several maturational changes in architecture and connectivity. The basic architectonic distinction between areas V1 and V2 in the macaque first becomes discernible around embryonic day 93-100 (E93ElOO), out of a 165-day gestation (Kostovic and Rakic, 1984; Rakic, et al., 1991). This is near the end of cortical Accepted December 7,1995. Address reprint requests to David C. Van Essen, Box 8108, Washington Univ. School Medicine, 660 S. Euclid, St. Louis, MO 63110. o 1996 WILEY-LISS, INC. 328 T.A. COOGAN AND D.C. VAN ESSEN TABLE 1. Number of Injections by Age and Location Case determination Age landmarks injections' Illustrations injections' Illustrations Species, age Developmental v1 v2 1 112 Fig. 4 313 Fig. 10 2 M . fascicularis; est. =E l18 313 Fig. 5,6, 17 111 3 M. mulatta, timed El18 212 Oil 4 M. mulatta; timed El22 V2 AChE stripes O i l 111 5 M . mulatta; timed El26 111 Fig. 16A 012 7 M fascicularis; est. =E l30 112 111 8 M. nemestrina; timed El33 314 3!3 Figs. 8, 12 11 M. mulatta; timed El59 212 Fig. 16B 212 13 M. fascicularis P7 111 212 14 M. fascicularis P21 111 111 15 M. fasczcularis P3mo. O i l 112 M nemrstrina; timed El08 Last neurons horn 6 M nemestrina, timed El30 112 213 Fig. 11 9 M . fascicularis; est. = El45 CO blobs present 212 Fig. 7A 213 Figs. 13, 14 10 M. mulalta; timed El55 212 Fig. 7B 313 Fig. 15 12 M. mulatta PO 313 Fig. 18 415 Fig. 9