Midget pathways of the primate retina underly resolution
[General characteristics]
[Visual acuity]
[Cone to midget bipolar cell connections]
[Midget ganglion cells]
[Circuits in human retina]
[M and blue/yellow ganglion cells]
[References]
1. General characteristics.
The specialized cone pathways of the central fovea of human and monkey retinas
are designed to have the least convergence and the greatest resolution
capabilities of the visual system. This is accomplished by making the
connections as "private" as possible and narrowing them to a one to one
relationship in the so-called midget pathways. Fig. 1. Polyak's drawing of bipolars and ganglion cells in the foveal slope (98 K jpeg image)
2. Visual acuity starts with cone spacing and midget circuitry. Visual acuity is a measure of our ability to discriminate the finest detail: either by discriminating two parallel lines apart (Vernier acuity) or two tiny spots or stars apart (point discrimination). The primate and human
visual system is capable of acuity of 1 min of arc or 60 cycles/degree of
visual angle. One degree of visual angle is thought to cover approximately
280-300 um of retinal distance (Drasdo and Fowler, 1974). The area of the
retina where the cone photoreceptors are most closely packed in a more or less
hexagonal array, is, of course, the foveola and is about one degree across. So visual discrimination of 1 min of arc is about the center to center spacing (3 µm) of the cones of the central mosaic in the foveola.
 Fig. 2. Light micrograph of foveal mosaic (78 K jpeg image) |
 Fig. 3. Human foveal cone mosaic (59 K jpeg image) |
If the minimum center to center spacing of 3 µm for the cones of the fovea is a then the resolution limit, known as the Nyquist limit, is the square root of 3a. One may ask why the resolution is not better than 1 min of arc and the size of, or even less than the diameter of each cone, i.e. a or < a. The spatial interference of the packed cones would apparently cause Moire effects or aliasing if line separations of less than the square root of 3a are used, as shown below (i.e. under conditions of spacing of two lines either a or 2a) (Wässle and Boycott, 1991). It turns out that under normal conditions, the optics of the eye do not sharply image gratings of less than 60 cycles /degree (Cambell and Gubisch, 1966). So the optics of the eye minimize Moire effects in spatial sampling.
The image is also blurred by the optics of the eye to project onto about 10 cones. Lateral shifts of this image can then be compared over these 10 cones and their midget chains of neurons for later computation in the brain. This allows better discrimination (resolution) because of the possibility of imposing antagonism from neighboring cone systems, i.e. surrounds on the pathways and increasing contrast and discrimination between the several midget-single-cone originating pathways (Gouras, 1992).
Fig. 4. Schematic drawing of projected lines on the foveal cone mosaic. (98 K jpeg image)
We know that three classes of cone comprise the cone foveal mosaic, but we have long been curious as to how the three different spectral types of cone are arranged in the hexagonal mosaic. The S-cones (blue cones) are fairly easily picked out from the L-cones (red cones) and M-cone (green cones) on some distinguishing morphological features (see chapter on S-cone pathways). In fact the S-cones are seen as the larger profiles disrupting the hexagonal mosaic in the wholemount view of the foveal mosaic of Figure 2 (arrows point to S-cones). In the foveal region of the cone mosaic S-cones form from 8-12% of the cones (Ahnelt et al., 1987). The very center of the foveal pit is almost devoid of S-cones. So the remainder 88-92% of the cones are somehow divided into the L- and M-cone populations. In the human fovea, psychophysical measurements have suggested that red cones outnumber green by 2:1 (Cicerone and Nerger, 1989). However, direct measurements by microspectrophotometry of all cones in small patches of cones in the fovea of monkeys, has revealed that red and green cones occur in about equal proportion (Mollon and Bowmaker, 1992). Roorda and Williams (1999) made direct measurements of spectral sensitivity of foveal cones in the living human eye by a sophisticated technique of laser inferometry. They found that humans varied greatly in their proportionsof red cones to green cones: some individuals have almost eual proportions while others have a higher proportion of red cones to the extreme of 2 red cones to every green cone. The red and green cones lie randomly in the mosaic meaning that clusters of cones of the same spectral type will occur together as illustrated in the figure (Fig. 5, below) from Mollon and Bowmaker's paper. An interesting point can be deduced from these findings. It does appear possible then, that a multiheaded midget or diffuse cone bipolar with a small dendritic tree, could contact all the same spectral type of cones in the primate fovea.
3. Cone to midget bipolar cell connections.
Polyak had already described two types of midget ganglion cell in the primate
retina in 1941. He had studied and drawn ganglion cells from vertical sections
of Golgi-stained monkey retina and noticed that the midget ganglion cells in
particular (although it is also true of the small and large parasol ganglion
cells) had dendrites branching at one of two levels in the IPL: close to the
ganglion cell bodies (the area we now know as sublamina b) or in
neuropil close to the amacrine cell bodies (the area we now know as sublamina
a). Fig. 6. Picture of Polyak (59 K jpeg image) He had also drawn and remarked on the fact that midget bipolar cells
came in long and short axon varieties to fit the two midget ganglion cell
varieties (Polyak, 1941).
Fig. 7. Drawing of midget bipolar cells by Polyak (59 K jpeg image) The important distinction between two types of midget bipolar cell was not
made until an electron microscope study was able to show differences
in their synaptic contacts with the cone pedicle (Kolb, 1970).
Fig. 8. Schematic drawing of a cone pedicle (59 K jpeg image)
Thus, invaginating midget bipolar cells (imb) were found to connect with cone
pedicles at central element, invaginating synapses at the cone pedicle ribbons.
Flat midget bipolar cells connect to cone pedicles at wide-cleft, basal
junctions (fmb), most often on either side of the invaginating dendrite of the
imb.
 Fig. 9. Invaginating midget bipolar cell (59 K jpeg image) |
 Fig. 10. Flat midget bipolar cell (98 K jpeg image) |
4. Midget ganglion cells.
Viewing Golgi stained wholemount primate retinas has been
particularly helpful in revealing the differences between the three
ganglion cell types that are involved with spatial and color vision: namely the midget ganglion cells, the blue/yellow ganglion cells and the parasol ganglion cells (Polyak, 1941; Kolb et al., 1992; Dacey and Lee, 1994) (Fig. 13). The midget ganglion cells are thought to be high acuity cells that also carry a color specific signal. They project to the parvocellular layers of the lateral geniculate nucles and are thus called P cells (Fig. 13) (Shapley and Perry, 1986). Midget ganglion cells come in high branching varieties that are probably OFF-center physiologically and low branching types that are ON-center physiologically.
 Fig. 12. Golgi staining of of human ganglion cells (39 K jpeg image) |
 Fig. 13. Primate ganglion cells (39 K jpeg image) |
Figure 12 shows two midget ganglion cells in comparison with a parasol ganglion cell (M cell) of peripheral human retina. The one midget ganglion cell has an 18 um diameter dendritic tree and the other has a larger dendritic field at 35 um in diameter. The midget ganglion cell with the smaller dendritic field comes into focus at a different level from the larger field cell. The former midget ganglion cell branches high in the IPL and so is probably an OFF center type physiologically. The lower branching midget ganglion cell with the larger dendritic tree is probably an ON center type. Why the OFF center type has a smaller dendritic tree compared with ON center type remains a mystery at present.
The connections between the midget bipolar and the midget ganglion cells have always been thought to be "private", i.e. one to one. Because of the similar size and branching level of the axon terminal of the bipolar and the dendritic arbor of the midget ganglion cell it has been assumed that they overlap and synapse with no room for convergence from more than one bipolar axon per ganglion cell (Fig. 14).
This was positively proven to be the case in the parafoveal area by an electron microscope study and reconstruction of serially sectioned midget ganglion cells and their input midget bipolar cell axons in parafovea, about 2 mm from the foveal pit (Fig 16) (Kolb and DeKorver, 1991). More recently we have also performed a serial section electron microscope study of foveal midget bipolar axons and their connections with the dendrites of foveal midget ganglion cells (Fig. 15) (Kolb and Marshak, 2003)
The respective midget bipolar cells are almost solely synaptic upon single midget ganglion cells, except in the very central fovea where a few midget bipolar synapses are shared with neighboring midget ganglion cells because of the crowding of neurons and neuropil. Since we know (Nelson et al., 1978) that ganglion cells branching in sublamina a will be OFF center and those branching in sublamina b will be ON center we can be sure that midget ganglion cells branching close to the amacrine cells will be OFF center and those branching close to ganglion cell layer will be ON center midget ganglion cells. This ON and OFF midget ganglion cell organization in primate has now been proved conclusively by Dacey and coauthors (2000). So it appears that in the foveal region and out to the borders of the central retina (about 4 mm from the fovea center) the midget pathways of the human fovea are organized in the following manner: 1 cone to 2 midget bipolar cells (ON- and OFF-center bipolar types) to 2 midget ganglion cells (ON- and OFF-center ganglion cell types).
What we find more difficult to say is what spectral type of cone is connected in the midget bipolar to midget ganglion cell chain. M- and L-cones are indistinguishable anatomically. Where we can recognize S-cone (blue) pedicles on unusual morphological features (Ahnelt et al., 1987, 1990; Kolb et al., 1997) (see chapter on S-cone pathways), the M- and L-cones look the same. There is some evidence that M- and L-connected midget bipolar cells have different numbers of synaptic ribbons associated with their axons and therefore different quantities of output sites to their midget ganglion cell in the fovea (Calkins et al., 1994): but which axon terminal/ganglion cell pair is associated with M- or L-cone pathways is still a moot question.
Four to five millimetres beyond the fovea, in near periphery, the midget bipolar cells become 2- and 3-headed connecting to 2 and 3 cones respectively (Polyak, 1941; Kolb et al., 1992; Boycott and Hopkins, 1991) (Fig. 14). Midget ganglion cells are found throughout the near and mid peripheral retina (Polyak, 1941; Rodieck et al., 1985; Dacey and Peterson, 1992; Dacey, 1993; Kolb et al., 1992) and their dendritic trees increase in size but are often composed of several domains of dendrites (Fig. 14). The further peripheral it gets, the more domains of dendrites compose the dendritic tree of midget ganglion cells. So it seems likely, that more than one midget bipolar cell will synapse upon each midget ganglion cell, probably each bipolar occupying one of the domains of the dendritic tree.
Serial section electron microscopy and reconstruction of axons and dendrites of midget ganglion cells at the edge of the central area (i.e. 4 mm from the foveal center) and from peripheral retina (12 mm from the foveal center) indicates that the prediction of multiple cone to midget bipolar and multiple midget bipolar inputs to the midget ganglion cells is correct in peripheral retina (Figs. 17 and 18). The most midget bipolars synapsing on a single ganglion cell is 3 at 12 mm from the fovea (Fig. 17) (Kolb and Marshak, 2003). Even more convergence of midget bipolar cells to midget ganglion cells occurs in far peripheral retina (Fig. 18).
In the fovea where a single cone connects with its midget system, the pathway will of necessity be carrying a single color in its receptive field center. Beyond the fovea where the midget pathway is 2 and 3 channeled, due to connection with 2 or 3 cones it remains to be seen whether these will be the same or a mixture of chromatic types. However, as mentioned above, due the patchy nature of the red and green cone mosaic, it could quite likely be that multi-branched midget bipolars can retain a committment to one spectral class of cone.
5. Circuits underlying red and green color opponency in the human retina.
Midget ganglion cells of the monkey retina (presumably also in human retina) are known to respond to light with an opponent chromatic organization (Gouras, 1968, Gouras, 1991 for a review). That is, midget ganglion cells when recorded from electrophysiologically, have the smallest receptive fields and are organized as L-cone ON or OFF center/M-cone OFF or ON surround or vice versa (Fig. 19).
 Fig. 19. Color-opponent units as described in monkey retina by Gouras (1968). (78 K jpeg image) |
 Fig. 20. Summary diagram of the center organization of midget ganglion cells (39 K jpeg image) |
The color of the center response in the midget ganglion cells is clearly coming from the midget bipolar's connection with a single cone type (either an L-cone or an M-cone). Thus midget ganglion cells must come in the four varieties: L-cone ON and L-cone OFF, M-cone ON and M-cone OFF (Figs. 19 and 20).
The parasol varieties of ganglion cells are probably the non-color opponent phasic varieties of ganglion cell recorded in the retina (Gouras, 1968) (Fig. 20), and called M cells projecting to the lateral geniculate nucleus (Shapley and Perry, 1986; Wiesel and Hubel, 1966; Gouras, 1992 for a review) (Fig. 24). A smaller field, tonic and not clearly color opponent ganglion cell type is also recorded in the retina and parvocellular layers of the geniculate nucleus (Gouras, 1992). Whether these are extrafoveal midget ganglion cells that are no longer color specific or some other as yet unidentified ganglion cell type still remains to be sorted out.
The opponent color surrounds recorded in midget ganglion cells (Gouras, 1968) provide color and spatial opponency. This antagonistic surround has to come from circuitry involving interneurons at the outer and inner plexiform layers. i.e. from horizontal and amacrine cells respectively.
Fig. 21. Contribution of horizontal cells to surround responses of bipolar cells (59 K jpeg image)
The horizontal cells are thought to contribute to the surround response of bipolar cells in submammalian species and particularly in creatures with good color vision, the horizontal cells have been shown to be color opponent in response. For example, horizontal cells are known to be red-green opponent or yellow-blue opponent in fish and turtle retinas (Stell and Lightfoot, 1975; see Kamermans and Spekreijse, 1995 or Kolb and Lipetz, 1991 for reviews). The situation may be different in primate retina where color opponency has been difficult to detect in intracellular recordings of monkey horizontal cells (Dacey et al., 1996) despite the anatomical findings on spectrally selective connectivity (Ahnelt and Kolb, 1994) (Fig. 21).
The role of amacrine cells may be more important in the formation of surrounds at ganglion cells in primates than in the fish and turtle retinas. Suffice it to say that in terms of color opponent surrounds for midget ganglion cells in primate, an additional circuit may be added by small-field amacrine cells such as illustrated in Fig. 22.
We believe that L- and M-midget ganglion cell responses are organized in the manner shown in figure 23. An L-cone would contact two L-cone midget bipolars and through them, two L-cone midget ganglion cells. For example, an L-cone ON-center midget bipolar would contact a single L-cone at invaginating synapses in the OPL, and a single L-cone ON-center midget ganglion cell in sublamina b of the IPL. The same L-cone would also be contacted by a single OFF-center midget bipolar at basal junctions, and the bipolar would, in turn, contact a single OFF-center midget ganglion cell in sublamina a of the the IPL (Fig. 23). Thus L-cone ON- and L-cone OFF-center receptive field ganglion cell types are generated (red + center and orange - center, Fig. 23, left). M-cones would be connected in a similar manner to two midget bipolars (ON and OFF types) and two midget ganglion cells which would also be ON and OFF-center receptive field types (dark green + center and light green - center, Fig. 23, right).
Two of the three types of horizontal cell could make M-cone surrounds to L-cone center bipolars, and L-cone surrounds to M-cone center bipolar cells (HII and HIII). In addition, L-cone surrounds for a M-cone center ganglion cell might be provided by an amacrine that collects information from a cluster of L-cone midget bipolar cell axons to pass to a M-cone driven midget bipolar/midget ganglion cell synaptic complex (red and orange a). The opposite of a M-cone driven amacrine cell providing M-cone surrounds to L-cone center bipolar and ganglion cells could also occur.
CLICK HERE to see an animation of midget pathways in the human retina (1.3 MB quicktime movie)
6. M and blue/yellow ganglion cells.
The large field parasol cells of the primate retina (Fig. 13) are thought to convey achromatic information concerning low spatial acuity and movement and luminance messages (Merigan, 1989; Merigan and Maunsell, 1990). Parasol cells come in ON- and OFF-center varieties like the midget ganglion cells and project to the magnocellular layers of the lateral geniculate nucleus (Fig. 24): hence they are also called M cells (Shapley and Perry, 1986; Kaplan and Shapley, 1986) The blue yellow ganglion cells (Fig. 13) are a bistratified type and thought to be driven by blue-specific bipolar cells, and other midget or diffuse types to give them their yellow response (Dacey and Lee, 1994). They project to a special layer of the geniculate nucleus called the koniocellular layer or K layer (Irwin et al., 1993; Calkins et al., 1997, 1998) (Fig. 24). We shall deal with the short wavelength sensitive system in another chapter and so will say little more about these ganglion cells here.
7. References.
Ahnelt, P. and Kolb, H. (1994) Horizontal cells and cone photoreceptors in
human retina: a Golgi-electron microscopic study of spectral connectivity. J. Comp. Neurol. 343, 406-427.
Ahnelt, P.K., Kolb, H. and Pflug, R. (1987) Identification of a subtype of cone photoreceptor, likely to be blue sensitive, in the human retina. J. Comp.
Neurol. 255, 18-34.
Ahnelt, P. K., Keri, C. and Kolb, H. (1990) Identification of pedicles of
putative blue sensitive cones in human and primate retina. J. Comp. Neurol.
293, 39-53.
Boycott, B. B. and Hopkins, J. M. (1991) Cone bipolar cells and cone synapses
in the primate retina. Vis. Neurosci. 7, 49-60.
Calkins, D.J., Schein, S.J., Tsukamoto, Y. and Sterling, P. (1994) M and L
cones in macaque fovea connect to midget ganglion cells by different numbers of excitatory synapses. Science 371, 70-72.
Calkins, D.J., Tsukamoto, Y. and Sterling, P. (1998) Microcircuitry and mosaic of a blue-yellow ganglion cell in the primate retina. J. Neurosci. 18, 3373-3385.
Calkins, D.J., Meszler, L.B. and Hendry, S.H.C. (1997) Multiple ganglion cell types express the a subunit of type II calmodulin-dependent protein kinase in the primate retina. Soc. for Neurosci. Abstr. 23p. 729.
Cambell, F.W. and Gubisch, F.W. (1966) Optical quality of the human eye. J.
Physiol. (Lond.) 186, 556-578.
Cicerone, C. M. and Nerger, J. L. (1989) The relative numbers of long-wavelength-sensitive to middle-wavelength-sensitive cones in the human fovea centralis. Vision Res. 29, 115-128.
Dacey, D.M. (1993) The mosaic of midget ganglion cells in the human retina. J. Neurosci. 13, 5334-5355.
Dacey, D.M. and Lee, B.B. (1994) The 'blue-on' opponent pathways in primate retina originates from a distinct bistratified ganglion cell. Nature 367, 731-73.
Dacey, D.M., Lee, B.B., Stafford, D.K., Pokorny, J. and Smith, V.C. (1996)
Horizontal cells of the primate retina: cone specificity without spectral
opponency. Science 271, 656-659.
Dacey, D., Packer, O.S., Diller, L., Brainard, D., Peterson, B. and Lee, B. (2000) Center surround receptive field structure of cone bipolar cells in primate retina. Vision. Res. 40, 1801-1811.
Gouras, P. (1968) Identification of cone mechanisms in monkey ganglion cells.
J. Physiol. (Lond.) 199, 533-547.
Gouras, P.(1991) Precortical physiology of colour vision. In :Vision and
Visual Dysfunction volume 6: "The Perception of Colour". (Ed. Gouras. P.)
Macmillam Press Ltd. England 6, 163-178.
Gouras, P. (1992) Retinal circuitry and its relevance to diagnostic
psychophysics and electrophysiology. Curr. Opin. Ophthal. 3, 803-812.
Irvin, G.E., Casagrande, V.A. and Norton, T.T. (1993) Center/surround relationships of magnocellular, parvocellular, and koniocellular relay cells in primate lateral geniculate nucleus. Visual Neurosci. 10, 363-373.
Kamermans, M and Spekreijse, H. (1995) Spectral behavior of cone-driven
horizontal cells in teleost retina. Prog. Ret. and Eye Res. 14,
313-360.
Kaplan, E. and Shapley, R.M. (1986) The primate retina contains two types of ganglion cells, with high and low contrast sensitivity. Proc. Natl. Acad. Sci. USA 88, 2755-2757.
Kolb, H. (1970) Organization of the outer plexiform layer of the primate
retina: electron microscopy of Golgi-impregnated cells. Phil. Trans. R. Soc. B258, 261-283.
Kolb, H. and DeKorver, L. (1991) Midget ganglion cells of the parafovea of the human retina: A study by electron microscopy and serial section
reconstructions. J. Comp. Neurol. 303, 617-636.
Kolb, H. and Lipetz, L. E. (1991) The anatomical basis for colour vision in the vertebrate
retina. In :Vision and Visual Dysfunction volume 6: "The Perception of Colour". (Ed. Gouras, P.) Macmillan Press Ltd., London pp. 128-145
Kolb, H., Linberg, K. A. and Fisher, S. K. (1992) The neurons of the human
retina: a Golgi study. J. Comp. Neurol. 318, 147-187.
Kolb, H., Goede, P., Roberts, S., McDermott, R. and Gouras, P. (1997) The
unique morphology of the S-cone pedicle n the human retina and its consequences for color processing. J. Comp. Neurol. 386, 443-460.
Kolb, H. and Marshak, D. (2003) The Midget Pathways of the Primate Retina. Doc. Ophtal. In press.
Merigan, W.H. and Maunsell, J.H.R. (1990) Macaque vision after magnocellular lateral geniculate lesions. Vis. Neurosci. 90, 347-352.
Nelson, R., Famiglietti, E. V. and Kolb, H. (1978) Intracellular staining reveals different levels of stratification for on-center and off-center ganglion cells in the cat retina. J. Neurophysiol. 41, 427-483.
Mollon, J. D. and Bowmaker, J. K. (1992) The spatial arrangement of cones in the primate fovea. Nature 360 677-679.
Polyak, S. L. (194) The Retina. University of Chicago Press, Chicago.
Rodieck, R.W., Binmoeller, K.F. and Dineen, J.T. (1985) Parasol and midget
ganglion cells of the human retina. J. Comp. Neurol. 233, 115-132.
Roorda, A., & Williams, D. R. (1999). The arrangement of the three cone classes in the living human eye. Nature, 397, 520-522.
Shapley, R. and Perry, V. H. (1986) Cat and monkey retinal ganglion cells and their visual functional roles. TINS 9, 229-235.
Stell, W.K. and Lightfoot, D.O. (1975) Color-specific interconnections of cones and horizontal cells in the retina of the goldfish. J. Comp. Neurol.
159, 473-501.
Wässle, H. and Boycott, B.B. (1991) Functional architecture of the
mammalian retina. Physiol. Rev. 71, 447-480.
Wiesel, T. N. and Hubel, D. H. (1966) Spatial and chromatic interactions in the
lateral geniculate body of the rhesus monkey. J. Neurophysiol. 29,
1115-1156.
[General characteristics]
[Visual acuity]
[Cone to midget bipolar cell connections]
[Midget ganglion cells]
[Circuits in human retina]
[M and blue/yellow ganglion cells]
[References]
Updated September, 2004
