Correspondence: Jon H. Kaas, Department of Psychology, Vanderbilt University, 301 Wilson Hall, 111 21st Ave. South, Nashville, TN 37240 USA. Tel: (615) 322±6029; Fax: (615) 343±4342 E-mail: jon.kaas@vanderbilt.edu Inverted pyramidal neurons in chimpanzee sensorimotor cortex are revealed by immunostaining with monoclonal antibody SMI-32 HUI-XIN QI 1 , NEERAJ JAIN 1 , TODD M. PREUSS 2 and JON H. KAAS 1 1 Department of Psychology, Vanderbilt University, Nashville, TN 37240; 2 Division of Behavioral Biology, University of Southwestern Louisiana, New Iberia Research Center New Iberia, LA 70560, USA Abstract We used the monoclonal antibody SMI-32 to label pyramidal cells of sensorimotor cortex in two chimpanzees. The majority of the pyramidal cells had typical vertically oriented apical dendrites that extended towards the pial surface. A small population of pyramidal cells varied from this orientation, so that the apical dendrites were 20 Æ or more from radial, and were often inverted, extending away from the pial surface. When numbers of non-inverted and inverted pyramidal cells were compared, less than 1% were found to be inverted. Key words: Anthropoid primate , Pan, somatosensory cortex, immunohistochemistry Introduction Pyramidal neurons in the neocortex of mammals normally extend their apical dendrites towards the pial surface. Van der Loos (1965), however, descri- bed cells that deviate from this typical configuration. In a Golgi study of the visual cortex in rabbits, he estimated that some 15±20% of pyramidal cells had apical dendrites that deviated 20 Æ or more from the normal orientation, with completely inverted ori- entations the most frequent of the atypical cells. Small numbers of non-upright pyramidal cells were also recognized in Golgi preparations from rats, cats, and monkeys. Van der Loos argued that the non- upright pyramidal cells were errors rather than a morphological adaptation to a functional role. Thus, he concluded that ªthe existence of disorientated pyramids may be explained by assuming that, in the turmoil of migration to, and alignment of the cells in, their definitive cortical sites, a few cells, perhaps not surprisingly so, become misaligned.º In support of this hypothesis, larger proportions of inverted pyr- amidals are described in certain pathological studies that affect cell migration, such as the reeler mutant mouse (Landrieu and Goffinet, 1981). In addition, inverted pyramids are thought to be more common in the deeper layers (Ferrer et al. , 1986; van Brederode and Snyder, 1992; Einstein and Fitzpa- trick, 1991), which develop the earliest. Others indicate that the functional significance of inverted pyramidals is unknown, holding out the possibility that they may play a unique functional role (van Brederode and Snyder, 1992). One type of unusual pyramidal neuron, the Martinotti cell, with atypically oriented apical dendrites and vertically ascending axons (T È ombÈ ol, 1984; Ferrer et al. , 1986), may well be a specialized type of neuron (Prieto et al. , 1994). While inverted pyramidals have been observed by a number of investigators in mice, rats, rabbits, cats, and even humans (see Feldman 1984), little is known about the frequency of their occurrence, the locations in cortex where they occur, and the prevalence among different species. Most reports suggest that they are less frequent than indicated by Van der Loos (1965), with Parnavelas (1977) esti- mating that they represent approximately 1% of pyramidal cells in the sensorimotor cortex of rats. The rarity of observations on inverted pyramidal cells is undoubtedly related to the difficulty of obtaining suitable Golgi material from large amounts of cortex. This difficulty can now be circumvented by staining pyramidal cells immunohistochemically. A monoclonal antibody to neurofilament protein, SMI- 32, labels the cells body and dendrites of a large subset of pyramidal cells in a Golgi-like manner (Campbell and Morrison, 1989, Hof et al. , 1995a; 1995b; 1996; Nimchinsky et al. , 1996; Preuss et al. , 1997). In such preparations, it is relatively easy to examine the orientations of the apical dendrites of a large number of SMI-32 immunoreactive neurons. We did this in sections of cortex from the brains of two chimpanzees that were processed for SMI-32 0899±0220/99/010049±08 $9.00  1999 Taylor & Francis Ltd Somatosensory & Motor Research 1999; 16(1): 49±56