MY RESEARCH PROPOSAL
Background
The K pathway and its roles in visual perception
The visual system processes different aspects of visual information and sends signals in parallel from the retina to cortex. In primates there are three parallel visual pathways: the magnocellular (M), parvocellular (P), and koniocellular (K) pathways, which project via separate layers of the lateral geniculate nucleus (LGN) to the primary visual cortex (V1) (Casagrande & Kaas, 1994).
The M and P pathways have been studied extensively and they have been proposed to carry motion and steoropsis, color and form signals, respectively, via separate V1 layers and compartments to higher visual areas (Livingstone & Hubel, 1988). However, much less is known about the function of the K pathway.
Little attention has been paid to the K pathway until relatively recently. Anatomical studies have basically demonstrated that the LGN of primates contains a third class of relay cells in addition to P and M cells. These very small cells lie within the interlaminar zone or intercalated layers, as well as within specialized layers, the superficial layers of simian primates and the kiniocellular layers of prosimians (Casagrande, 1994). Hereafter in the paper these relay cells are generally referred to as K cells. In simian primates like macaque monkeys, the cells belonging to LGN K pathway are difficult to study because these cells are more scattered between and below the more densely packed cells within the M and P layers (Kaas et al, 1978). The origin of retinal input to K LGN cells had been debated for a long time in primates, especially in macaque monkeys. Some researchers reported that a third class of retinal ganglion cells did not project to the LGN in macaque monkeys but projected to the superior colliculus (Perry & Cowey, 1984). Other investigators, nevertheless, supported the existence of a distinct class of ganglion cells that project to macaque K LGN cells (Leventhal et al., 1981; Conley & Fitzpatrick, 1989). More recent results also showed that K LGN cells can be as numerous as M LGN cells in macaque monkeys (Hendry, 1995). In other primates, especially in bush babies, studies clearly show the existence of a third class of retinal ganglion cells projecting to K LGN cells (Lachia & Casagrande, 1988; 1993). Now accumulative studies in primates have indicated the K cells in the LGN could provide substantial input to the supragranular layers of V1 (Livingstone & Hubel, 1982; Hendry & Yoshioka, 1994; Ding & Casagrande, 1996, 1998 ).
Generally, several aspects of the anatomy, neurochemistry, and physiology of the K pathways to V1 in different primate species now indicate that this pathway can play an important role in vision (Casagrande, 1994; Casagrande & Kaas, 1994).
1.Characteristics of the K pathway
To date, the most detailed studies of K LGN cells have been done in the prosimian primate, bush baby (Casagrande, 1994). Many of the morphological and neurochemical characteristics of K LGN cells identified for bush babies have also been identified for the K pathway in simian primates. K LGN cells in the bush baby are quite different from P and M cells in their manner of projections to V1. That is, K cell axons terminate directly within single CO blobs in layer III and more broadly within layer I, while M and P axons terminate principally within layer IV and to a lesser extent in the lower and upper tiers of layer VI (Lachica & Casagrande, 1992; Casagrande & Kaas, 1994). The K axons terminate within CO blobs in layer III, no axons are found to terminate within interblob zones. Most K axon arbors branch in layer III B and some can be located in either layer III A or III C. So it would be possible that the K arbors are in a position to directly influence the cells in the major pathways that project to extrastriate cortex via layers III A or III C. Moreover, some K arbors extend branches to layer I where arbors spread tangentially over a broad zone. The major projection of the LGN K1 layer cell axons, as well as collaterals of some K3 axons of owl monkeys ascend to cortical layer I (Ding & Casagrande, 1997). Thus, K arbors could have transcompartmental influence via contacts with the apical dendrites extending into layer I.
Unlike P and M LGN cells, K LGN cells in primates receive plenty of projections from the superficial layers of the superior colliculus (Harting et al., 1991). Some studies in axon reconstruction and ultrastructure of bush babies suggest that collicular inputs are important to K cell function (Lachica & Casagrande, 1993; Harting et al., 1991). In addition, K LGN cells also receive input from the parabigeminal nucleus, V1 and several extrastriate visual areas (Harting et al., 1991; Lachica & Casagrande, 1993, Casagrande, 1994).
Neurochemical research also showed that K pathway is distinct from the P and K pathways in Ca2+-binding protein content and CO activity. In primates like bush babies, calbindin immunoactivity is dense, but CO staining is light and parvalbumin immunoactivity is almost absent in K LGN layers; while parvalbumin and CO immunoactivity is dense in the P and M layers, where little calbindin immunoactivity is present (Johnson & Casagrande, 1995; Casagrande, 1994). For macaque monkey, the K LGN cells were also immunoreactive for the ( subunit of calmodulin-dependent (CaM II) kinase (Hendry & Yoshioka, 1994).
In spite of the common features that define K cells across primates, there is also evidence of differences. The relative concentration of K cells within the different interlaminar zones of the LGN, as defined by calbindin immunocytochenistry, shows considerable interspecies variability (Hendry & Casagrande, 1996). Data in squirrel monkeys, owl monkeys and macaque monkeys indicate that different populations of LGN K cells may project to different supragranular layers of V1 (Fitzpatrick et al., 1983; Hendry & Yoshioka, 1994, Ding & Casagande, 1997). In the owl monkey, the axon arbors from different K layers are morphologically distinct; axons from LGN layer K1 mainly project to cortical layer I, while axons from LGN layer K3 mainly terminate in cortical layer III (Ding & Casagrande, 1997). Some studies have reported that a subset of K LGN cells in simians projects to extrastriate visual areas including V2, MT and inferior temporal cortex (ITC) (Kennedy & Bullier, 1985; Lysakowski et al., 1988). A single K axon was reconstructed that projected to V2 in an owl monkey and this axon showed a distinct pattern of termination from the K axons projecting to V1 (Ding & Casagrande, 1997). Therefore, it seems that the K pathway may actually consist of more than one subpathway.
2. Functional considerations of the K pathway and their possible significance
Aspects of the K pathway characteristics suggest functions. The direct projections of K axons to the CO blobs of V1 suggest that they may involve the function of these compartments. In all primates examined (nocturnal and diurnal), the cells in CO blobs project to two extastriate areas: the second visual area (V2) and the dorsal medial visual area (DM). The projection to V2 is always to the V2 compartments that in turn send signals to the dorsolateral area(DL); these signals end up in temporal lobe areas concerned with object identification. In contrast, DM projects to parietal areas concerned with motion and object location (Goodale & Milner, 1992). Thus, the direct projection of the K pathway to the CO blob could suggest that K cells support: 1) color perception, 2) object vision, 3) motion or spatial vision, or 4) some combination. Some physiological studies have suggested the role of K cells or the K pathway in visual perception.
In the bush baby, physiological studies have shown that K cells are a heterogenous population that share some features with cat LGN W-cells (Norton & Casagrande, 1982; Irvin et al., 1986; Norton et al., 1988; Irvin et al., 1993). Compared with M and P cells, the K cells in bush babies generally have larger receptive fields, longer latencies to chiasm and cortical (antidromic) stimulation, longer visual onset (but not peak) latencies, and are more sensitive to non-visual (auditory and somatosensory). Studies in bush babies also suggest that a subpopulation of K cells could contribute to visual resolution based upon their contrast-sensitivity functions. The K cells (those could be driven by grating stimuli) have peak spatial frequency cut-offs and contrast sensitivities intermediate between those of the M and P cells (Norton et al., 1988). Recently, in the new world monkey marmoset, some studies (Martin, et., 1997; White, et al., 1998) found out that one group of color opponent cells, the blue-ON ganglion cells in the retina, were largely segregated to the interlaminar zone (K layer equivalent) in the LGN. These results suggested that color information might be processed in the K pathway.
The fact that the K pathway receives much input from the superior colliculus, suggests a role in perception during eye movements (Casagrande, 1994). K cells could be visually driven either directly from the retina or indirectly from strong input from the colliculus (Harting et al., 1991; Lachica & Casagrande, 1993; Feig & Harting, 1994). The superior colliculus has been shown to influence the response of the LGN cells in ways that suggest this pathway could be involved with eye movements (Xue et al., 1994). It is interesting that some LGN cells in cats showed changes in firing in response to passive eye movement (Lal & Friedlander, 1989).
Another suggestion for a role of the K pathway is that it plays a modulatory role (Casagrande, 1994). Thalamic pathways that target cortical layer I are often suggested to be modulatory pathways given the physiological evidence that cortical layer I play such a role (Vogt, 1991). The major projection of the LGN K1 layer cell axons, as well as collaterals of some K3 axons of owl monkeys to cortical layer I, could be supportive of such a proposal (Ding & Casagrande, 1997).
Research proposal
Although we have made some progress toward understanding the K pathway in recent years, many questions remain to be unraveled about the function of the K pathway. It is essential to decide the functional subclasses of the K pathway in simian primates. We also need to further explore the role of K pathway in visual perception, especially about its input to V1 layer III cortical cells and about its relation with behavior relevant activity. Following is our long-term plans for future research.
First, considering the known morphological subclasses in the K pathway, we hypothesize that the K pathway in prosimian primates is made up of some different functional subclasses, which should have different physiological characteristics. We will use owl monkeys for the proposed experiments because they have a simple LGN structure with two P layers and two M layers and relatively large and easily accessible K3 and K1 layers, which is quite helpful to our study of the K pathway. Also, the previous studies about the K cell properties were mostly done in the prosimian bush babies (Casagrande, 1994), so it is necessary to study the K cells in some simian species. The purpose of this study is to define the subclasses of K LGN cells within different K layers in the owl monkey LGN, and distinguish the K LGN geniculocortical afferents segregated in the V1 cortical layers. The latter will be done after silencing the cortex with muscimol blockade by adopting the protocol described as Chapman et al. (1991) did in ferrets.
Second, the fact that K cells in the LGN provide substantial input to the supragranular layers of V1, terminating within the CO blobs of cortical layer IIIB and in layer I, suggests some important roles in visual perception. We hypothesize that K pathway can contribute to some important response properties of V1 cortical cells and it can involve some important aspects in visual information processing. So far it is not clear at all that what specific attributes of V1 cortical cells would be related with the input of K LGN cells and what the K pathway can contribute to visual information processing in striate and extrastriate levels. We will try to find out the contribution of K pathway to V1 cell properties in bush babies and owl monkey, by studying their response properties with and without K LGN inputs. In these experiments we will investigate the influence of LGN inputs, especially K input, by reversibly inactivating ("blocking") the neural activity of the particular LGN layers with iontophoretic application of (-aminobutyric acid (GABA) while recording from upper layer cells in the cortex. Our lab had successfully used the technique of iontophoretic application of GABA in the study about the influence of input from the lower cortical layers on the orientation tuning of upper layer V1 cell in bush babies (Allison et al, 1995). Therefore, it would be useful to perform experiments for future related research.
Third, we know the K pathway has strong input from the superior colliculus and also some previous studies have shown the influence of eye movements on the LGN K cells. We hypothesize that K cells could carry information related to eye movements, eye position, or behavioral relevance, and we would determine the relationship of LGN K cell activity to eye movements or to behavioral relevance. The purpose of this aim is to test the relationship of LGN (especially K cell) activity with eye movements and behavioral relevance. We try to know if eye movement like saccades can affect the cell activity and if behavioral relevance can affect the cell activity. We will work on awake monkeys that have received behavior training. Together with collaboration from Dr. Schall's lab and our electrophysiological experience, our lab can adapt some of their techniques to accomplish our studies.
EXIT