The olfactory cortex is the only part of the vertebrate forebrain to receive a direct sensory input. Present in even the most primitive fish, it retains its place and form throughout the vertebrate series, suggesting that it is a core element in the basic plan of the vertebrate forebrain. Since olfaction is the dominant sensory modality in most vertebrate species, an understanding of olfactory cortical mechanisms can yield insight into basic behavioral patterns underlying much of mammalian and primate behavior. The olfactory system is also one of the first sensory systems to differentiate and become functional during fetal life.
An understanding of the olfactory cortex requires a clear view of its place in the olfactory pathway. This pathway and its constituent neurons were first revealed by the use of the Golgi stain in the later part of the nineteenth century. The pathway consists of three main parts (see Figure 1). First is the olfactory sensory epithelium in the nose, containing the olfactory sensory neurons, which transduce the stimulating odor molecules into impulses that are sent over their axons in the olfactory nerve to the olfactory bulb, the second main structure of the olfactory pathway. Here the axons extend synapses onto the dendrites of relay neurons, the large mitral cells and smaller tufted cells. These cells interact with interneurons in the olfactory bulb and send the processed information by means of impulse discharges in their axons in the lateral olfactory tract (LOT) on the ventrolateral surface of the brain. The axons give rise to numerous collaterals, which terminate in the third main region, the olfactory cortex, to make synapses on the dendrites of cortical pyramidal neurons.
The olfactory cortex is the area of cortex in the vertebrate forebrain that receives direct input form the olfactory bulb. In most mammals there are several areas of this cortex: The main region is the piriform cortex (PC) (meaning "pear-shaped"—it was originally termed "prepyriform" and sometimes spelled "pyriform"). In most mammalian brains it extends over much of the ventrolateral surface of the brain dorsal to the LOT, as shown in Figure 2. The piriform cortex sends its output axons to several areas. One target area is the mediodorsal nucleus of the thalamus, which has connections to the prefrontal areas of the neocortex. This pathway is believed to be involved in conscious perception of odors. Depending on the species, the piriform cortex also sends fibers to other cortical areas, such as the insula, where odor information may be combined with taste information to give the overall perception of flavor. Subcortical connections are made to parts of the limbic system, including the hypothalamus, hippocampus, and basal ganglia.
The other olfactory cortical areas include the following: The anterior olfactory nucleus (AON) is a major station for integrating activity of the olfactory bulb with that of olfactory cortical areas on both sides of the brain via connections through the anterior commissure. The olfactory tubercle (OT) in rodents lies medial to the LOT; it is notable for receiving a heavy input of dopaminergic axons from the mid-brain. Because of the implication of dopamine systems in various types of mental disorders (depression, sleep disturbances, schizophrenia), the olfactory tubercle has been studied intensively in rodents for its possible role in these types of disorders. Another target for olfactory input is the amygdala. The accessory olfactory bulb (AOB) receives input from the vomeronasal organ in the nose about chemical signals involved in mating in many vertebrate species and projects to the corticomedial nuclei within the amygdalar complex; these, in turn, project to the hypothalamus, where they presumably activate some of the behavioral patterns in mating. Finally, there is the lateral entorhinal cortex (LEC), a major region for multimodal integration of olfactory, visual, auditory, and somatosensory inputs; the output of this region is carried in fibers of the perforant pathway to the hippocampus, which is a critical region for storage and retrieval of information of behavioral significance.
Human Olfactory Cortical Areas
The regions of the rodent brain described above are similar in the brains of other mammals and some lower primates; in higher primates, including humans, there are a number of differences. In the monkey brain, the olfactory pathway appears much reduced in relative terms compared with the large forebrain, seemingly reflecting the reduced importance of the sense of smell. However, absolute numbers of neurons or fibers are not a reliable guide to the behavioral importance of a particular system; for example, the LOT contains many more fibers than the auditory nerve, yet no one would say that hearing is unimportant in human life.
As a result of the overgrowth of the forebrain in primates, the olfactory pathway is limited to the ventral surface of the brain. The olfactory bulbs give rise to a long LOT, which divides into three roots as it enters the brain. The lateral root goes to a cortical area at the junction of the frontal and temporal cortexes, which seems to be the homologue of the piriform cortex. From here there are connections to the prefrontal cortex, called the orbital cortex because it is on the surface of the brain facing the orbit of the eye. The medial root dives into a small area that, being pock-marked with many penetrating blood vessels, is called the perforated substance, a homologue of the olfactory tubercle. The medial root includes fibers that project toward or into the hypothalamus, from which there is a projection to the orbital cortex complementary to that from the piriform cortex. There are thus several routes over which information can be processed, through different olfactory cortical regions, to both cortical and subcortical regions.
The Basic Cortical Circuit
The main regions and pathways described above are like cities and motor routes on a map. An understanding of the basis of information processing requires an analysis of the structures within each region. A traditional way to characterize the organization of the olfactory cortex is by its layers. The olfactory cortex is the prototypical three-layer cortex, consisting of an outer molecular layer of incoming fibers and apical dendrites; a middle layer of pyramidal cell bodies; and an inner plexiform or polymorphic layer of basal dendrites, fibers, and interneurons. It shares this three-layer organization with the hippocampus and contrasts with the six-layer construction of the neocortex.
Researchers have clarified the synaptic circuits of these cortical regions. A basic circuit is made up of the minimum types of input fibers, output neurons, intrinsic neurons, and their connections sufficient to represent the most important information-processing functions of a given region. For the olfactory cortex, the main input elements are the fibers from the LOT, which make excitatory synapses on the spines of the distal dendrites of the pyramidal neurons. The main output elements are the pyramidal neurons, each consisting of apical and basal dendrites, receiving excitatory synapses on their dendritic spines and inhibitory synapses on their dendritic shafts and cell bodies. The main intrinsic elements are two types of interneurons that make the inhibitory connections onto the pyramidal neurons. There are two main types of intrinsic circuits: a reexcitatory feedback circuit through long axon collaterals of the pyramidal neurons and inhibitory circuits for feedforward and feedback inhibition of the pyramidal neurons.
This basic circuit not only summarizes the main anatomical circuits within the piriform cortex but also, with minor variations, applies to the other olfactory cortical areas; moreover, it is similar to the other circuits: those of the hippocampus; the neocortex, particularly its superficial layers; and the "canonical circuit" proposed by some researchers for the neocortex. Thus, the basic circuit for the olfactory cortex is a useful model for correlating the general properties of all types of forebrain cortices.
Researchers seek to characterize the functional properties of each type of cortical element and synaptic circuit and to identify the stages of expression of different circuit components and types of excitatory, inhibitory, and modulatory neurotransmitters. Through such studies they seek correlations with the developing behavior of the fetus and the neonate. Studies of plasticity have indicated that the properties of the excitatory synapses made by the sensory input fibers from the LOT are different from those of the intrinsic reexcitatory collateral system; the latter give evidence of n-methyl-D-aspartate (NMDA) receptors and are thus strong candidates for mediating the long-term potentiation in the piriform cortex. Computational networks based on the basic circuit model suggest that sensory inputs and reexcitatory feedback interact over extensive areas of the olfactory cortex to provide a distributed parallel system; such analyses shed light on the mechanisms of odor discrimination and may constitute a model for related types of pattern recognition by other cortical systems. In particular, the excitatory feedback between pyramidal cells within the piriform cortex may mediate autoassociative memory function. During initial encoding, synaptic modification of these recurrent connections in the piriform cortex could store associations between neurons activated by a given odor stimulus. Subsequent encounters with similar stimuli could activate a subset of these neurons, and activity could spread across the modified connections to complete the previously encoded pattern of activity. This pattern completion could serve as a retrieval of the previously encoded memory for the odor. Further knowledge of the structural organization of the olfactory cortex should provide a better basis for understanding these and other important aspects of olfactory function and behavior.
Ramón y Cajal, S. (1894). Les nouvelles idées sur la structure du système nerveux chez l'homme et chez les vertébrés. Paris: Reinwald.
Shepherd, G. M. (1989). A basic circuit for cortical organization. In M. C. Gazzaniga, ed., Perspectives on memory research. Cambridge, MA: MIT Press.
Shepherd, G. M., and Stewart, W. B. (1985). The chemical senses: Taste and smell. In M. Swash and C. Kennard, eds., Scientific basis of clinical neurology. London: Churchill Livingstone.
Revised byMichael E.Hasselmo