- Email: [email protected]
Effects of tetrodotoxin treatment in LGN on neuromodulatory receptor expression in developing visual cortex
Developmental Brain Research 106 Ž1998. 93–99
Research report
Effects of tetrodotoxin treatment in LGN on neuromodulatory receptor expression in developing visual cortex Qiang Gu
a,b,)
, Yulin Liu
a,1
, Richard H. Dyck
a,2
, Virginia Booth
a,b
, Max S. Cynader
a,b
a
b
Department of Ophthalmology, UniÕersity of British Columbia, VancouÕer, BC, Canada, V5Z 3N9 Brain and Spinal Cord Research Center, UniÕersity of British Columbia, VancouÕer, BC, Canada V5Z 3N9 Accepted 29 October 1997
Abstract The expression and distribution patterns of transmitter receptors change dramatically during pre- and post-natal development of the visual cortex, but the factors that control these processes are largely unknown. We have tested the hypothesis that input activity from the lateral geniculate nucleus ŽLGN., one major input source to visual cortex, may contribute to the processes underlying transmitter receptor redistributions in the visual cortex during development. We found that a short period of tetrodotoxin ŽTTX. treatment in LGN retarded the developmental expression and age-dependent reorganization of neuromodulatory receptors, including muscarinic, serotonergic and adrenergic receptors, in kitten primary visual cortex. The visual cortices ipsilateral to the TTX infusion site displayed a ‘younger’ receptor pattern than that of their contralateral control counterparts in the same animals. The results suggest that active input from LGN regulates the expression profile of a broad range of receptors in the developing visual cortex. q 1998 Elsevier Science B.V. Keywords: Visual cortex; Lateral geniculate nucleus; Muscarinic M 1 receptor; Adrenergic a 1 receptor; Adrenergic a 2 receptor; Serotonergic 5-HT2c receptor; Postnatal development; Tetrodotoxin
1. Introduction The distribution patterns of transmitter receptors in the developing visual cortex have been extensively examined. Various techniques including autoradiography, immunocytochemistry, in situ hybridization, and electrophysiological recordings have been employed in these studies. The outcome of these studies has shown that the distribution patterns of transmitter receptors in the visual cortex change dramatically during early postnatal development. Several patterns of receptor expression as a function of age have been described for individual receptors with some canoni-
) Corresponding author. Department of Ophthalmology, University of British Columbia, 2550 Willow Street, Vancouver, British Columbia, Canada V5Z 3N9. Fax: q1-604-875-4663; E-mail: [email protected] 1 Present address: Department of Pathology, Allegheny Hospital, Medical College of Pennsylvania, Pittsburgh, PA 15211-4772, USA. 2 Present address: Department of Psychology, University of Lethbridge, Lethbridge, Alberta, Canada T1K 3M4.
0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 2 0 0 - 9
cal features observed. These include initial concentrations in layer IV, the cortical input layer, followed by later redistributions to other cortical layers w2,3x. Another frequently observed pattern involves initial concentrations in the subplate zone, a transiently expressed scaffold zone for the developing cortex w1x, followed by later expression elsewhere in the cortex w2,3x. While most studies on transmitter receptors in the developing visual cortex have concentrated on the distribution profiles, little is known about the mechanisms that control the developmental reorganization of these receptors. Are these genetically preprogrammed or dependent on some other factors? Although the mechanisms responsible for the reorganization process are not well understood, it has been proposed that extrinsic signals may actively drive the developmental expression of transmitter receptors w2,3x. The visual cortex receives a number of inputs w6,11x. The most important of these, the visual input, comes via the retino–geniculo–cortical pathway. Several lines of evidence suggest that visual input activity plays an important role in a number of key events in the developing visual
94
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
cortex. It may be possible that visual input activity is also crucial for the postnatal development and age-dependent reorganization of transmitter receptors in the visual cortex. In the present study, we have designed an experimental model and addressed the question of whether blocking neuronal activity in the lateral geniculate nucleus ŽLGN. can affect the distribution patterns of neuromodulatory receptors in kitten visual cortex.
2. Materials and methods The experimental design is illustrated in Fig. 1. Osmotic minipumps were used to infuse tetrodotoxin ŽTTX., a sodium-channel blocker, to abolish action potentials in the LGN of one hemisphere. Thus, the ipsilateral visual cortex received no active LGN input, while the contralateral visual cortex still received direct LGN input and served as an internal control within the same kitten. After two weeks of unilateral LGN impulse blockade, the density and distribution of several neuromodulatory receptors including muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors in the visual cortex of both hemispheres were studied using established autoradiographic techniques. Specifically, osmotic minipumps ŽAlza, model 2002. were implanted subcutaneously in three 23-day-old kittens. This age was chosen since the receptors studied here show evidence of dramatic changes in their laminar distribution patterns or density during the next two weeks of postnatal development w5,7,9x. The minipumps were connected to 30-G stainless steel cannulae via polyethylene tubing. The cannulae were inserted vertically into the brain by means
of a micromanipulator, and their tips were aimed just above the LGN at Horsley–Clarke stereotaxic coordinates: A 3.7, L 8, and H 11. The cannulae were stabilized on the skull first with tissue adhesive and then with acrylic dental cement. One minipump delivered TTX, dissolved in artificial cerebro-spinal fluid ŽACSF., at the rate of 50 pmolrh to the LGN in one hemisphere, while another pump delivered only ACSF at 0.5 m lrh onto the other LGN as a control. Previous studies have shown that this dose of TTX can totally block neuronal activity at least 2 mm from the infusion cannula tip w10x. Therefore, this dose of TTX should be sufficient to block all activity in the LGN on one side of the brain. Two weeks later, the kittens were sacrificed and their visual cortices were processed for receptor autoradiography. All kittens showed normal weight gain during the two-week infusion period. The animals were deeply anaesthetized and perfused transcardially with cold phosphate buffer Ž0.1 M, pH 7.4.. The brains were quickly removed, frozen at y208C. wH 3 xpirenzepine, wI 125 xDOI ŽŽ".-1-Ž2,5-dimethoxy-4- 125 Iiodophenyl . -2-am inopropane . , w H 3 x prasozin and wH 3 xrauwolscine were used to label muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors, respectively. The binding procedures we used and the specificity of each radioligand for individual receptors are based on ligand characterizations that have been described previously w5,7,9x. The brains were sectioned at a thickness of 15 m m with a cryostat. Before incubation with ligands, sections were equilibrated to room temperature for 10–20 min, washed for 30 min in buffer, and then incubated in the following solutions. For muscarinic M 1 receptors: 5 nM wH 3 xpirenzepine, 0.5 nM unlabelled 4-DAMP Ž4-diphenylacetoxy-N-methylpiperidine. and 20 mM HEPES
Fig. 1. Schematic illustration of the experimental design. The LGN of one hemisphere was treated with TTX Ž50 pmolrh. to abolish neuronal activity in the LGN. The other LGN was treated with ACSF Ž0.5 m lrh. as control. The treatment lasted for two weeks, from postnatal days 23 to day 37. During this period, the visual cortex in one hemisphere received no active input from its LGN, while the visual cortex in other hemisphere still received direct LGN input and served as an internal control within the same animal.
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
Tris–HCl buffer with 10 mM Mg 2q ŽpH 7.5. for 1 h at room temperature. For serotonergic 5-HT2 receptors: 0.5 nM wI 125 xDOI, 0.1% bovine serum albumin, 0.01% ascorbic acid and 50 mM Tris–Hcl buffer ŽpH 7.4. for 1 h at room temperature. For adrenergic a 1 and a 2 receptors: 0.75 nM wH 3 xprasozin and 0.6 nM wH 3 xrauwolscine, respectively, in 0.1 M phosphate buffer ŽpH 7.7. for 1 h at 48C. After incubation in the solutions, the slide-mounted sections were washed in cold buffer to remove unbound ligands, briefly dipped in cold distilled water to remove buffer salts, dried under a stream of cool air, and then apposed to X-ray film ŽAmersham.. To generate autoradiograms, the exposure times were 3 days for wI 125 xDOI, 4 weeks for wH 3 xpirenzepine, 3 months for wH 3 x-prasozin, and 3 months for wH 3 xrauwolscine, respectively. Autoradiographic images were captured and analyzed using a Macintosh-based computer image analysis system running IMAGE software ŽNIH, version 1.45.. For quantitative densitometry, we measured the optical density of the auto-
95
radiograms across the medial bank Žarea 17. where the visual cortex exhibits relatively constant thickness, and where there is a symmetry between the left and right visual cortex for direct comparison. The locations of cortical laminae were determined using adjacent Nissl-stained sections. Four sections were analyzed for each ligand.
3. Results In general, all the receptors studied including muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors, showed clear differences in their binding profiles between the TTX-treated and control hemispheres. In the control hemispheres, each of the four receptors examined had its own unique distribution pattern, which was typical for kitten visual cortex at this age w5,7,9x. By contrast, the distributions of binding sites for each of the four receptors within the TTX-treated hemispheres were
Fig. 2. ŽA. The distribution patterns of muscarinic M 1 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of muscarinic M 1 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
96
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
quite different. The levels and distribution patterns of all four receptors were found instead to resemble the patterns of expression previously reported in the visual cortex of much younger kittens w5,7–9x. These results were consistent in all three kittens studied. Quantitative densitometric analyses of receptor densities in both control and TTXtreated hemispheres support the conclusions arrived at by visual inspection of the data. Nissl-stained serial sections that contain both sides of LGN indicate that the tips of the infusion cannulae were at the right location, just above LGN, and that there was no observable damage to the LGN. For muscarinic M 1 receptors, the control visual cortex exhibited high binding densities in the superficial layers, low levels in the deep layers, and the lowest levels in the middle layers ŽFig. 2A, left hemisphere.. By contrast, on the TTX-treated side, the binding density was high in superficial and middle layers, and lower in the deep layers ŽFig. 2A, right hemisphere.. Quantitative densitometric measurements showed that M 1 receptor binding density in
the control cortices was highest in layer II, followed by layers III, VI and V, and least in layer IV ŽFig. 2B, left panel., while in the TTX-treated hemisphere, it was highest in layer II, followed by layers IV, III and VI, and least in layer V ŽFig. 2B, right panel.. Comparing both hemispheres, one can clearly see that M 1 receptor binding sites in the TTX-treated hemispheres are more abundant in the middle and deep cortical layers than those in control hemispheres. For serotonergic 5-HT2 receptors, the highest densities in the control hemisphere occurred in the middle layers, with low levels in other layers ŽFig. 3A, left hemisphere.. On the TTX-treated side, the binding densities in the middle layers were lower than those of control hemisphere ŽFig. 3A, right hemisphere.. Quantitative densitometric measurements showed that 5-HT2 receptor binding density in the control cortices was highest in layers III and IV, followed by layer II, subplate layer ŽSP. and layer V, and least in layer VI ŽFig. 3B, left panel., while in the TTXtreated hemisphere, it was highest in layers II and III,
Fig. 3. ŽA. The distribution patterns of serotonergic 5-HT2 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of serotonergic 5-HT2 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
followed by subplate layer, layers IV and V, and least in layer VI ŽFig. 3B, right panel.. The largest difference in 5-HT2 receptor binding sites between the two hemispheres is the lack of dense receptor binding in layer IV in the TTX-treated hemisphere. For adrenergic a 1 receptors, binding sites were most strongly concentrated in the subplate zone and in the middle cortical layers, with lower densities of expression in the superficial layers, and the least binding in deep cortical layers in the control hemisphere ŽFig. 4A, left hemisphere.. On the TTX-treated side, the binding density was also high in the subplate cortical layer, but lower in the middle cortical layers than in the control hemisphere ŽFig. 4A, right hemisphere.. Quantitative densitometric measurements showed that a 1 receptor binding density in the control cortices was highest in subplate layer, followed by layers IV, III, II and V, and least in layer VI ŽFig. 4B, left panel., while in the TTX-treated hemisphere, it was
97
highest in subplate layer, followed by layers III, II, IV and V, and least in layer VI ŽFig. 4B, right panel.. a 1-receptor binding sites in the TTX-treated hemisphere showed much lower densities in the middle cortical layers compared to those in the control hemisphere. For adrenergic a 2 receptors, binding sites were mostly concentrated in the middle cortical layers and less concentrated in the superficial layers and deep layers in the control hemisphere ŽFig. 5A, left hemisphere.. In the TTX-treated side, the binding distribution pattern for a 2 receptors was similar to that of the control hemisphere, but the binding density was lower in the middle cortical layers ŽFig. 5B, right hemisphere.. Quantitative densitometric measurements showed that a 2 receptor binding density in the control cortices was highest in layers III and IV, followed by layers II and V, and least in layer VI ŽFig. 5B, left panel., while in the TTX-treated hemisphere, it was highest in layers III and IV, followed by layers II and V,
Fig. 4. ŽA. The distribution patterns of adrenergic a 1 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of adrenergic a 1 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
98
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
Fig. 5. ŽA. The distribution patterns of adrenergic a 2 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of adrenergic a 2 receptors across the visual cortices were measured by densitometric analyses. Right: TTX treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
and least in layer VI ŽFig. 5B, right panel.. The major difference of a 2 receptor binding sites between both hemispheres is seen within the middle cortical layers.
4. Discussion The present results demonstrate that treatment of LGN with TTX disturbed the developmental redistribution of muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors in kitten visual cortex. It is unlikely that the observed unilateral effects were due to any possible lesion of the LGN, and hence associated degeneration of thalamo–cortical projections, because anatomical reconstruction indicated that the tips of infusion cannulae were all placed just above LGN, and there were no obvious abnormalities in the TTX-treated LGN as examined histologically. It also seems unlikely that TTX diffused into the ipsilateral visual cortex, and therefore blocked neuronal activity in the visual cortex, because based on the location of infusion cannula, TTX would diffuse first into the LGN,
and then into the ventricle. So, it would be mixed with cerebral fluid and thereafter affect both cerebral hemispheres. Our results thus suggest that active input from LGN contributes to the reorganization and expression of these receptors in the developing visual cortex. Our results did not indicate an overall reduction of receptor expression in the visual cortex following LGN-input deprivation, since, for example, the muscarinic M 1 receptors in layer IV ŽFig. 2. and adrenergic a 1 receptors in the subplate layer ŽFig. 4. remained highly dense in the TTX-treated hemispheres. Our results suggest rather that the lack of active LGN-input ‘freezes’ the receptor distribution pattern at a ‘younger’ stage. Since only one time point has been examined in this study, it would be interesting to further investigate whether these effects in the visual cortex that were caused by TTX treatment in the LGN are temporary or permanent. We do not know if after TTX infusion stops, the developmental redistribution of these receptors in the visual cortex resumes, and eventually ‘catches up’ with the normal developmental progression, or whether the distribution of these receptors in the visual
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
cortex remains in an immature state permanently. This is a distinct possibility, since the visual cortex may already have passed the time window or critical period during which receptor reorganization can occur. Available evidence indicates that LGN input is mediated through an excitatory amino acid pathway w12x rather than through cholinergic, serotonergic or noradrenergic inputs. We have found previously that interruption of modulatory inputs to visual cortex, induced by surgically severing the modulatory pathways that arise from the basal forebrain Žcholinergic., locus coeruleus Žnoradrenergic. and raphe nuclei Žserotonergic., respectively, at an early age, had no discernible effects on the density or pattern of receptor development w4x. Our results therefore also suggest that during development, the LGN input that contributes to the overall activity of visual cortical neurons is more important than the receptors’ ‘own’ inputs Žcholinergic, serotonergic, or noradrenergic. in determining these receptors’ expression and developmental fate. Little is known about how the LGN input could regulate the expression patterns of neuromodulatory receptors in the developing visual cortex. One plausible hypothesis is that LGN input activates cortical neurons and triggers an elevation of intracellular calcium concentrations in cortical neurons, and that an increase of calcium concentration in turn activates numerous calcium-dependent signal transduction pathways that ultimately control the profile of receptor binding. Since the receptor subtypes examined here are linked with different second messenger systems, either IP3rDAG ŽM 1 , 5-HT2 and a 1 receptor. or cAMP Ž a 2 receptor., blocking LGN input not only affected the redistribution of these receptors, but also the second messenger levels in cortical cells which would express these receptors when receiving normal LGN input. This may subsequently influence the maturation processes of particular cortical layers, and eventually affect the development and plasticity of the entire visual cortex. In the present study, we have examined only a few subtypes of neuromodulatory receptors. Whether LGN in-
99
put also contributes to the expression pattern of other major transmitter receptors ŽGABAergic or glutamatergic. in the developing visual cortex will need further exploration. References w1x K.L. Allendoerfer, C.J. Shatz, The subplate, a transient neocortical structure: Its role in the development of connections between thalamus and cortex, Annu. Rev. Neurosci. 17 Ž1994. 185–218. w2x M. Cynader, C. Shaw, F. Van Huizen, G. Prusky, Transient receptor expression in visual cortex development and the mechanisms of cortical plasticity, in: B.L. Finley, G. Innocenti, H. Scheich ŽEds.., The Neocortex, Plenum, New York, 1990, pp. 245–253. w3x M. Cynader, C. Shaw, G. Prusky, F. Van Huizen, Neural mechanisms underlying modifiability of response properties in developing cat visual cortex, in: B. Cohen, I. Bodis-Wollner ŽEds.., Vision and the Brain, Raven Press, New York, 1990, pp. 85–108. w4x M.S. Cynader, V. Booth, Y.L. Liu, R. Dyck, W.G. Jia, Input-dependent transient expression of receptor populations in kitten visual cortex development, Soc. Neurosci. Abstr. 16 Ž1990. 331.9. w5x R.H. Dyck, M.S. Cynader, Autoradiographic localization of serotonin receptor subtypes in cat visual cortex: Transient regional, laminar, and columnar distributions during postnatal development, J. Neurosci. 13 Ž1993. 4316–4338. w6x C.D. Gilbert, Microcircuitry of the visual cortex, Annu. Rev. Neurosci. 6 Ž1983. 217–247. w7x W.W.-G. Jia, Y. Liu, F. Lepore, M. Ptito, M. Cynader, Development and regulation of alpha adrenoceptors in kitten visual cortex, Neuroscience 63 Ž1994. 179–190. w8x Y. Liu, W. Jia, Q. Gu, M. Cynader, Involvement of muscarinic acetylcholine receptors in regulation of kitten visual cortex plasticity, Dev. Brain Res. 79 Ž1994. 63–71. w9x G. Prusky, M. Cynader, The distribution of M 1 and M 2 muscarinic acetylcholine receptor subtypes in the developing cat visual cortex, Dev. Brain Res. 56 Ž1990. 1–12. w10x H.O. Reiter, D.M. Waitzman, M.P. Stryker, Cortical activity blockade prevents ocular dominance plasticity in the kitten visual cortex, Exp. Brain Res. 65 Ž1986. 182–188. w11x R.W. Rodieck, Visual pathways, Annu. Rev. Neurosci. 2 Ž1979. 193–225. w12x H. Tamura, T.P. Hicks, Y. Hata, T. Tsumoto, A. Yamatodani, Release of glutamate and aspartate from the visual cortex of the cat following activation of afferent pathways, Exp. Brain Res. 80 Ž1990. 447–455.
Research report
Effects of tetrodotoxin treatment in LGN on neuromodulatory receptor expression in developing visual cortex Qiang Gu
a,b,)
, Yulin Liu
a,1
, Richard H. Dyck
a,2
, Virginia Booth
a,b
, Max S. Cynader
a,b
a
b
Department of Ophthalmology, UniÕersity of British Columbia, VancouÕer, BC, Canada, V5Z 3N9 Brain and Spinal Cord Research Center, UniÕersity of British Columbia, VancouÕer, BC, Canada V5Z 3N9 Accepted 29 October 1997
Abstract The expression and distribution patterns of transmitter receptors change dramatically during pre- and post-natal development of the visual cortex, but the factors that control these processes are largely unknown. We have tested the hypothesis that input activity from the lateral geniculate nucleus ŽLGN., one major input source to visual cortex, may contribute to the processes underlying transmitter receptor redistributions in the visual cortex during development. We found that a short period of tetrodotoxin ŽTTX. treatment in LGN retarded the developmental expression and age-dependent reorganization of neuromodulatory receptors, including muscarinic, serotonergic and adrenergic receptors, in kitten primary visual cortex. The visual cortices ipsilateral to the TTX infusion site displayed a ‘younger’ receptor pattern than that of their contralateral control counterparts in the same animals. The results suggest that active input from LGN regulates the expression profile of a broad range of receptors in the developing visual cortex. q 1998 Elsevier Science B.V. Keywords: Visual cortex; Lateral geniculate nucleus; Muscarinic M 1 receptor; Adrenergic a 1 receptor; Adrenergic a 2 receptor; Serotonergic 5-HT2c receptor; Postnatal development; Tetrodotoxin
1. Introduction The distribution patterns of transmitter receptors in the developing visual cortex have been extensively examined. Various techniques including autoradiography, immunocytochemistry, in situ hybridization, and electrophysiological recordings have been employed in these studies. The outcome of these studies has shown that the distribution patterns of transmitter receptors in the visual cortex change dramatically during early postnatal development. Several patterns of receptor expression as a function of age have been described for individual receptors with some canoni-
) Corresponding author. Department of Ophthalmology, University of British Columbia, 2550 Willow Street, Vancouver, British Columbia, Canada V5Z 3N9. Fax: q1-604-875-4663; E-mail: [email protected] 1 Present address: Department of Pathology, Allegheny Hospital, Medical College of Pennsylvania, Pittsburgh, PA 15211-4772, USA. 2 Present address: Department of Psychology, University of Lethbridge, Lethbridge, Alberta, Canada T1K 3M4.
0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 2 0 0 - 9
cal features observed. These include initial concentrations in layer IV, the cortical input layer, followed by later redistributions to other cortical layers w2,3x. Another frequently observed pattern involves initial concentrations in the subplate zone, a transiently expressed scaffold zone for the developing cortex w1x, followed by later expression elsewhere in the cortex w2,3x. While most studies on transmitter receptors in the developing visual cortex have concentrated on the distribution profiles, little is known about the mechanisms that control the developmental reorganization of these receptors. Are these genetically preprogrammed or dependent on some other factors? Although the mechanisms responsible for the reorganization process are not well understood, it has been proposed that extrinsic signals may actively drive the developmental expression of transmitter receptors w2,3x. The visual cortex receives a number of inputs w6,11x. The most important of these, the visual input, comes via the retino–geniculo–cortical pathway. Several lines of evidence suggest that visual input activity plays an important role in a number of key events in the developing visual
94
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
cortex. It may be possible that visual input activity is also crucial for the postnatal development and age-dependent reorganization of transmitter receptors in the visual cortex. In the present study, we have designed an experimental model and addressed the question of whether blocking neuronal activity in the lateral geniculate nucleus ŽLGN. can affect the distribution patterns of neuromodulatory receptors in kitten visual cortex.
2. Materials and methods The experimental design is illustrated in Fig. 1. Osmotic minipumps were used to infuse tetrodotoxin ŽTTX., a sodium-channel blocker, to abolish action potentials in the LGN of one hemisphere. Thus, the ipsilateral visual cortex received no active LGN input, while the contralateral visual cortex still received direct LGN input and served as an internal control within the same kitten. After two weeks of unilateral LGN impulse blockade, the density and distribution of several neuromodulatory receptors including muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors in the visual cortex of both hemispheres were studied using established autoradiographic techniques. Specifically, osmotic minipumps ŽAlza, model 2002. were implanted subcutaneously in three 23-day-old kittens. This age was chosen since the receptors studied here show evidence of dramatic changes in their laminar distribution patterns or density during the next two weeks of postnatal development w5,7,9x. The minipumps were connected to 30-G stainless steel cannulae via polyethylene tubing. The cannulae were inserted vertically into the brain by means
of a micromanipulator, and their tips were aimed just above the LGN at Horsley–Clarke stereotaxic coordinates: A 3.7, L 8, and H 11. The cannulae were stabilized on the skull first with tissue adhesive and then with acrylic dental cement. One minipump delivered TTX, dissolved in artificial cerebro-spinal fluid ŽACSF., at the rate of 50 pmolrh to the LGN in one hemisphere, while another pump delivered only ACSF at 0.5 m lrh onto the other LGN as a control. Previous studies have shown that this dose of TTX can totally block neuronal activity at least 2 mm from the infusion cannula tip w10x. Therefore, this dose of TTX should be sufficient to block all activity in the LGN on one side of the brain. Two weeks later, the kittens were sacrificed and their visual cortices were processed for receptor autoradiography. All kittens showed normal weight gain during the two-week infusion period. The animals were deeply anaesthetized and perfused transcardially with cold phosphate buffer Ž0.1 M, pH 7.4.. The brains were quickly removed, frozen at y208C. wH 3 xpirenzepine, wI 125 xDOI ŽŽ".-1-Ž2,5-dimethoxy-4- 125 Iiodophenyl . -2-am inopropane . , w H 3 x prasozin and wH 3 xrauwolscine were used to label muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors, respectively. The binding procedures we used and the specificity of each radioligand for individual receptors are based on ligand characterizations that have been described previously w5,7,9x. The brains were sectioned at a thickness of 15 m m with a cryostat. Before incubation with ligands, sections were equilibrated to room temperature for 10–20 min, washed for 30 min in buffer, and then incubated in the following solutions. For muscarinic M 1 receptors: 5 nM wH 3 xpirenzepine, 0.5 nM unlabelled 4-DAMP Ž4-diphenylacetoxy-N-methylpiperidine. and 20 mM HEPES
Fig. 1. Schematic illustration of the experimental design. The LGN of one hemisphere was treated with TTX Ž50 pmolrh. to abolish neuronal activity in the LGN. The other LGN was treated with ACSF Ž0.5 m lrh. as control. The treatment lasted for two weeks, from postnatal days 23 to day 37. During this period, the visual cortex in one hemisphere received no active input from its LGN, while the visual cortex in other hemisphere still received direct LGN input and served as an internal control within the same animal.
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
Tris–HCl buffer with 10 mM Mg 2q ŽpH 7.5. for 1 h at room temperature. For serotonergic 5-HT2 receptors: 0.5 nM wI 125 xDOI, 0.1% bovine serum albumin, 0.01% ascorbic acid and 50 mM Tris–Hcl buffer ŽpH 7.4. for 1 h at room temperature. For adrenergic a 1 and a 2 receptors: 0.75 nM wH 3 xprasozin and 0.6 nM wH 3 xrauwolscine, respectively, in 0.1 M phosphate buffer ŽpH 7.7. for 1 h at 48C. After incubation in the solutions, the slide-mounted sections were washed in cold buffer to remove unbound ligands, briefly dipped in cold distilled water to remove buffer salts, dried under a stream of cool air, and then apposed to X-ray film ŽAmersham.. To generate autoradiograms, the exposure times were 3 days for wI 125 xDOI, 4 weeks for wH 3 xpirenzepine, 3 months for wH 3 x-prasozin, and 3 months for wH 3 xrauwolscine, respectively. Autoradiographic images were captured and analyzed using a Macintosh-based computer image analysis system running IMAGE software ŽNIH, version 1.45.. For quantitative densitometry, we measured the optical density of the auto-
95
radiograms across the medial bank Žarea 17. where the visual cortex exhibits relatively constant thickness, and where there is a symmetry between the left and right visual cortex for direct comparison. The locations of cortical laminae were determined using adjacent Nissl-stained sections. Four sections were analyzed for each ligand.
3. Results In general, all the receptors studied including muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors, showed clear differences in their binding profiles between the TTX-treated and control hemispheres. In the control hemispheres, each of the four receptors examined had its own unique distribution pattern, which was typical for kitten visual cortex at this age w5,7,9x. By contrast, the distributions of binding sites for each of the four receptors within the TTX-treated hemispheres were
Fig. 2. ŽA. The distribution patterns of muscarinic M 1 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of muscarinic M 1 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
96
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
quite different. The levels and distribution patterns of all four receptors were found instead to resemble the patterns of expression previously reported in the visual cortex of much younger kittens w5,7–9x. These results were consistent in all three kittens studied. Quantitative densitometric analyses of receptor densities in both control and TTXtreated hemispheres support the conclusions arrived at by visual inspection of the data. Nissl-stained serial sections that contain both sides of LGN indicate that the tips of the infusion cannulae were at the right location, just above LGN, and that there was no observable damage to the LGN. For muscarinic M 1 receptors, the control visual cortex exhibited high binding densities in the superficial layers, low levels in the deep layers, and the lowest levels in the middle layers ŽFig. 2A, left hemisphere.. By contrast, on the TTX-treated side, the binding density was high in superficial and middle layers, and lower in the deep layers ŽFig. 2A, right hemisphere.. Quantitative densitometric measurements showed that M 1 receptor binding density in
the control cortices was highest in layer II, followed by layers III, VI and V, and least in layer IV ŽFig. 2B, left panel., while in the TTX-treated hemisphere, it was highest in layer II, followed by layers IV, III and VI, and least in layer V ŽFig. 2B, right panel.. Comparing both hemispheres, one can clearly see that M 1 receptor binding sites in the TTX-treated hemispheres are more abundant in the middle and deep cortical layers than those in control hemispheres. For serotonergic 5-HT2 receptors, the highest densities in the control hemisphere occurred in the middle layers, with low levels in other layers ŽFig. 3A, left hemisphere.. On the TTX-treated side, the binding densities in the middle layers were lower than those of control hemisphere ŽFig. 3A, right hemisphere.. Quantitative densitometric measurements showed that 5-HT2 receptor binding density in the control cortices was highest in layers III and IV, followed by layer II, subplate layer ŽSP. and layer V, and least in layer VI ŽFig. 3B, left panel., while in the TTXtreated hemisphere, it was highest in layers II and III,
Fig. 3. ŽA. The distribution patterns of serotonergic 5-HT2 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of serotonergic 5-HT2 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
followed by subplate layer, layers IV and V, and least in layer VI ŽFig. 3B, right panel.. The largest difference in 5-HT2 receptor binding sites between the two hemispheres is the lack of dense receptor binding in layer IV in the TTX-treated hemisphere. For adrenergic a 1 receptors, binding sites were most strongly concentrated in the subplate zone and in the middle cortical layers, with lower densities of expression in the superficial layers, and the least binding in deep cortical layers in the control hemisphere ŽFig. 4A, left hemisphere.. On the TTX-treated side, the binding density was also high in the subplate cortical layer, but lower in the middle cortical layers than in the control hemisphere ŽFig. 4A, right hemisphere.. Quantitative densitometric measurements showed that a 1 receptor binding density in the control cortices was highest in subplate layer, followed by layers IV, III, II and V, and least in layer VI ŽFig. 4B, left panel., while in the TTX-treated hemisphere, it was
97
highest in subplate layer, followed by layers III, II, IV and V, and least in layer VI ŽFig. 4B, right panel.. a 1-receptor binding sites in the TTX-treated hemisphere showed much lower densities in the middle cortical layers compared to those in the control hemisphere. For adrenergic a 2 receptors, binding sites were mostly concentrated in the middle cortical layers and less concentrated in the superficial layers and deep layers in the control hemisphere ŽFig. 5A, left hemisphere.. In the TTX-treated side, the binding distribution pattern for a 2 receptors was similar to that of the control hemisphere, but the binding density was lower in the middle cortical layers ŽFig. 5B, right hemisphere.. Quantitative densitometric measurements showed that a 2 receptor binding density in the control cortices was highest in layers III and IV, followed by layers II and V, and least in layer VI ŽFig. 5B, left panel., while in the TTX-treated hemisphere, it was highest in layers III and IV, followed by layers II and V,
Fig. 4. ŽA. The distribution patterns of adrenergic a 1 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of adrenergic a 1 receptors across the visual cortices were measured by densitometric analyses. Right: TTX-treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
98
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
Fig. 5. ŽA. The distribution patterns of adrenergic a 2 receptor binding sites in kitten visual cortex after two weeks of unilateral TTX treatment within the LGN. The LGN of the right hemisphere was treated with TTX, while the LGN of the left hemisphere was treated with ACSF Žcontrol.. ŽB. The relative binding densities of adrenergic a 2 receptors across the visual cortices were measured by densitometric analyses. Right: TTX treated hemisphere; left: control hemisphere. The curves represent mean densities from eight measurements at different locations across the medial bank.
and least in layer VI ŽFig. 5B, right panel.. The major difference of a 2 receptor binding sites between both hemispheres is seen within the middle cortical layers.
4. Discussion The present results demonstrate that treatment of LGN with TTX disturbed the developmental redistribution of muscarinic M 1 , serotonergic 5-HT2 , and adrenergic a 1 and a 2 receptors in kitten visual cortex. It is unlikely that the observed unilateral effects were due to any possible lesion of the LGN, and hence associated degeneration of thalamo–cortical projections, because anatomical reconstruction indicated that the tips of infusion cannulae were all placed just above LGN, and there were no obvious abnormalities in the TTX-treated LGN as examined histologically. It also seems unlikely that TTX diffused into the ipsilateral visual cortex, and therefore blocked neuronal activity in the visual cortex, because based on the location of infusion cannula, TTX would diffuse first into the LGN,
and then into the ventricle. So, it would be mixed with cerebral fluid and thereafter affect both cerebral hemispheres. Our results thus suggest that active input from LGN contributes to the reorganization and expression of these receptors in the developing visual cortex. Our results did not indicate an overall reduction of receptor expression in the visual cortex following LGN-input deprivation, since, for example, the muscarinic M 1 receptors in layer IV ŽFig. 2. and adrenergic a 1 receptors in the subplate layer ŽFig. 4. remained highly dense in the TTX-treated hemispheres. Our results suggest rather that the lack of active LGN-input ‘freezes’ the receptor distribution pattern at a ‘younger’ stage. Since only one time point has been examined in this study, it would be interesting to further investigate whether these effects in the visual cortex that were caused by TTX treatment in the LGN are temporary or permanent. We do not know if after TTX infusion stops, the developmental redistribution of these receptors in the visual cortex resumes, and eventually ‘catches up’ with the normal developmental progression, or whether the distribution of these receptors in the visual
Q. Gu et al.r DeÕelopmental Brain Research 106 (1998) 93–99
cortex remains in an immature state permanently. This is a distinct possibility, since the visual cortex may already have passed the time window or critical period during which receptor reorganization can occur. Available evidence indicates that LGN input is mediated through an excitatory amino acid pathway w12x rather than through cholinergic, serotonergic or noradrenergic inputs. We have found previously that interruption of modulatory inputs to visual cortex, induced by surgically severing the modulatory pathways that arise from the basal forebrain Žcholinergic., locus coeruleus Žnoradrenergic. and raphe nuclei Žserotonergic., respectively, at an early age, had no discernible effects on the density or pattern of receptor development w4x. Our results therefore also suggest that during development, the LGN input that contributes to the overall activity of visual cortical neurons is more important than the receptors’ ‘own’ inputs Žcholinergic, serotonergic, or noradrenergic. in determining these receptors’ expression and developmental fate. Little is known about how the LGN input could regulate the expression patterns of neuromodulatory receptors in the developing visual cortex. One plausible hypothesis is that LGN input activates cortical neurons and triggers an elevation of intracellular calcium concentrations in cortical neurons, and that an increase of calcium concentration in turn activates numerous calcium-dependent signal transduction pathways that ultimately control the profile of receptor binding. Since the receptor subtypes examined here are linked with different second messenger systems, either IP3rDAG ŽM 1 , 5-HT2 and a 1 receptor. or cAMP Ž a 2 receptor., blocking LGN input not only affected the redistribution of these receptors, but also the second messenger levels in cortical cells which would express these receptors when receiving normal LGN input. This may subsequently influence the maturation processes of particular cortical layers, and eventually affect the development and plasticity of the entire visual cortex. In the present study, we have examined only a few subtypes of neuromodulatory receptors. Whether LGN in-
99
put also contributes to the expression pattern of other major transmitter receptors ŽGABAergic or glutamatergic. in the developing visual cortex will need further exploration. References w1x K.L. Allendoerfer, C.J. Shatz, The subplate, a transient neocortical structure: Its role in the development of connections between thalamus and cortex, Annu. Rev. Neurosci. 17 Ž1994. 185–218. w2x M. Cynader, C. Shaw, F. Van Huizen, G. Prusky, Transient receptor expression in visual cortex development and the mechanisms of cortical plasticity, in: B.L. Finley, G. Innocenti, H. Scheich ŽEds.., The Neocortex, Plenum, New York, 1990, pp. 245–253. w3x M. Cynader, C. Shaw, G. Prusky, F. Van Huizen, Neural mechanisms underlying modifiability of response properties in developing cat visual cortex, in: B. Cohen, I. Bodis-Wollner ŽEds.., Vision and the Brain, Raven Press, New York, 1990, pp. 85–108. w4x M.S. Cynader, V. Booth, Y.L. Liu, R. Dyck, W.G. Jia, Input-dependent transient expression of receptor populations in kitten visual cortex development, Soc. Neurosci. Abstr. 16 Ž1990. 331.9. w5x R.H. Dyck, M.S. Cynader, Autoradiographic localization of serotonin receptor subtypes in cat visual cortex: Transient regional, laminar, and columnar distributions during postnatal development, J. Neurosci. 13 Ž1993. 4316–4338. w6x C.D. Gilbert, Microcircuitry of the visual cortex, Annu. Rev. Neurosci. 6 Ž1983. 217–247. w7x W.W.-G. Jia, Y. Liu, F. Lepore, M. Ptito, M. Cynader, Development and regulation of alpha adrenoceptors in kitten visual cortex, Neuroscience 63 Ž1994. 179–190. w8x Y. Liu, W. Jia, Q. Gu, M. Cynader, Involvement of muscarinic acetylcholine receptors in regulation of kitten visual cortex plasticity, Dev. Brain Res. 79 Ž1994. 63–71. w9x G. Prusky, M. Cynader, The distribution of M 1 and M 2 muscarinic acetylcholine receptor subtypes in the developing cat visual cortex, Dev. Brain Res. 56 Ž1990. 1–12. w10x H.O. Reiter, D.M. Waitzman, M.P. Stryker, Cortical activity blockade prevents ocular dominance plasticity in the kitten visual cortex, Exp. Brain Res. 65 Ž1986. 182–188. w11x R.W. Rodieck, Visual pathways, Annu. Rev. Neurosci. 2 Ž1979. 193–225. w12x H. Tamura, T.P. Hicks, Y. Hata, T. Tsumoto, A. Yamatodani, Release of glutamate and aspartate from the visual cortex of the cat following activation of afferent pathways, Exp. Brain Res. 80 Ž1990. 447–455.