Horizontal propagation of visual activity in the synaptic integration field of area 17 neurons

Science

Volumn 283, Issue 5402, 1999, Pages 695-699

  Bringuier, Vincent ^a,b   Chavane, Frédéric ^a   Glaeser, Larry ^a   Frégnac, Yves ^a  

^a CNRS   (France)

^b MEDICAL RESEARCH COUNCIL   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

ACTION POTENTIAL; ANIMAL EXPERIMENT; ARTICLE; BRAIN CELL; CAT; DEPOLARIZATION; EVOKED VISUAL RESPONSE; LATENT PERIOD; NONHUMAN; PRIORITY JOURNAL; RECEPTIVE FIELD; RETINA; SPIKE WAVE; SYNAPTIC TRANSMISSION; VISUAL CORTEX; VISUAL STIMULATION;

ACTION POTENTIALS; ANIMALS; AXONS; BRAIN MAPPING; CATS; EVOKED POTENTIALS, VISUAL; PATCH-CLAMP TECHNIQUES; PHOTIC STIMULATION; SYNAPSES; VISUAL CORTEX; VISUAL FIELDS; VISUAL PATHWAYS;

ANIMALIA; FELIS CATUS;

EID: 0033614035     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.283.5402.695     Document Type: Article

References (42)

1. D. H. Hubel and T. N. Wiesel, J. Physiol. 160, 106 (1962).
2. G. Orban, Stud. Brain Funct. 11, 365 (1984).
3. R. J. Douglas et al., Science 269, 981 (1995); R. C. Reid and J.-M. Alonso, Nature 378, 281 (1995).
4. R. J. Douglas et al., Science 269, 981 (1995); R. C. Reid and J.-M. Alonso, Nature 378, 281 (1995).
5. P. O. Bishop et al., J. Physiol. (London) 219, 625 (1971); L. Maffei and A. Fiorentini, Vision Res. 16, 1131 (1976); G. C. De Angelis et al., J. Neurophysiol. 71, 347 (1994).
6. P. O. Bishop et al., J. Physiol. (London) 219, 625 (1971); L. Maffei and A. Fiorentini, Vision Res. 16, 1131 (1976); G. C. De Angelis et al., J. Neurophysiol. 71, 347 (1994).
7. P. O. Bishop et al., J. Physiol. (London) 219, 625 (1971); L. Maffei and A. Fiorentini, Vision Res. 16, 1131 (1976); G. C. De Angelis et al., J. Neurophysiol. 71, 347 (1994).
8. C. Y. Li and W. Li, Vision Res. 34, 2337 (1994).
9. C. D. Gilbert and T. N. Wiesel, J. Neurosci. 3, 1116 (1983).
10. P. Salin and J. Bullier, Physiol. Rev. 75, 107 (1995).
11. O. Creutzfeldt and M. Ito, Exp. Brain Res. 6, 324 (1968); R. J. Douglas and K. Martin, J. Physiol (London) 440, 735 (1991); L. J. Borg-Graham et al., Nature 393, 369 (1998).
12. O. Creutzfeldt and M. Ito, Exp. Brain Res. 6, 324 (1968); R. J. Douglas and K. Martin, J. Physiol (London) 440, 735 (1991); L. J. Borg-Graham et al., Nature 393, 369 (1998).
13. O. Creutzfeldt and M. Ito, Exp. Brain Res. 6, 324 (1968); R. J. Douglas and K. Martin, J. Physiol (London) 440, 735 (1991); L. J. Borg-Graham et al., Nature 393, 369 (1998).
14. X. Pei et al., J. Neurosci. 14, 7130 (1994).
15. Cells in the primary visual cortex of anesthetized (Althesin) and paralyzed kitten and adult cats were recorded intracellularty for periods ranging from 30 to 615 min. Three-millimeter artificial pupils were used. Sharp electrode recordings (46 cells) were performed in bridge mode with 50 to 90 megohm glass pipettes filled with 2 M potassium methyl sulfate. The average resting membrane potential was -65.8 mV (±8.7 mV), input resistance ranged between 10 and 70 megohm, and spike height ranged between 35 and 70 mV, Whole-cell recordings (20 cells) were done with 2 to 5 megohm glass patch electrode filled with 140 mM potassium-gluconate [L. J. Borg-Graham et al., Nature 393, 369 (1998)]. Data processing and visual stimulation protocols used in-house software (G. Sadec, Acquis1 Biologic CNRS-ANVAR). Because the different measurements (the spatial extent of the synaptic integration field, ASHP) were found to be independent of age, data obtained from kittens older than 6 weeks of age and adult cats were pooled.
16. 2). (ii) Light or dark bars were randomly flashed ON and OFF for 500 ms or 1 s, at various eccentricities from the MDF (10 positions along the width axis, 20 trials), (iii) Medium-contrast gratings were adjusted for optimal spatial and temporal frequencies and presented either moving or counterphased through annular windows whose inner width and length were preset multiples of the dimensions of the MDF (120 and 360%). Each protocol sequence was repeated 10 to 50 times. Discharge field center locations varied between 1 and 16° of retinal eccentricity, but most receptive fields were located within 10° around the area centralis. Because no significant differences in our measurements were found between simple cells (62% of our sample) and complex cells (38%), both types of receptive fields were pooled.
17. Estimates of the ratio of D-field size over MDF size are based on similar statistical tests performed at the subthreshold and spiking level The MDF was determined as the distance between the positions that elicited a significant increase of the spike discharge (one-tailed t test, P < 0.01) when compared with the shuffled discharge elicited for a similar duration period. The visually evoked subthreshold response strength was measured by the integral of the depolarizing component of the stimulus-locked waveform over a 300-ms period starting at the stimulus onset, and compared with that obtained from similar durations of shuffled activity (t test, P < 0.01). A linear analysis, relating response strength with the eccentricity from the MDF center, was restricted to the regions where significant responses were evoked, and did not take into account "silent" regions. For each cell, the spatial sensitivity hill profile was fitted by the best pair of lines (least-mean-square criteria), whose intersection with the threshold level of significant response defined the outer diameter of the depolarizing subthreshold field.
18. C. Blakemore and E. A. Tobin, Exp. Brain Res. 15, 439 (1972); B. Gulyas et al., J. Neurophysiol. 57, 1767 (1987).
19. C. Blakemore and E. A. Tobin, Exp. Brain Res. 15, 439 (1972); B. Gulyas et al., J. Neurophysiol. 57, 1767 (1987).
20. A. M. Sillito et al., Nature 378, 492 (1995).
21. r = -65.8 ± 8.7 mV) than the equilibrium potential for chloride {-77 ± 3.9 mV in vitro [B. W. Connors et al., J. Physiol. 406, 443 (1988)] and -75 mV in vivo [D. E. Shulz et al., Soc. Neurosci. Abstr. 19, 638 (1993)]}, events of positive and negative polarity, when occurring in isolation, may be equated, respectively, with dominating excitatory and inhibitory postsynaptic events. This first-order approximation does not take into account nonlinear interaction between excitation and inhibition [L. J. Borg-Graham et al., Nature 393, 369 (1998)]. Furthermore, assuming a reversal potential close to 0 mV for excitatory conductances, detection of glutamate-mediated events is largely favored with respect to GABA (γ-aminobutyric acid)- mediated ones. H-fields are therefore underestimated in our study. The qualitative subthreshold D-field extents reported by Pei et al. (9) are difficult to compare with our own results because of a limited sample size. In addition, their recordings were performed at more depolarized membrane potentials (-30 or -50 mV), where inhibitory hyperpolarizing responses may mask partially the D-field responses, and response amplitudes were measured at a fixed latency.
22. r = -65.8 ± 8.7 mV) than the equilibrium potential for chloride {-77 ± 3.9 mV in vitro [B. W. Connors et al., J. Physiol. 406, 443 (1988)] and -75 mV in vivo [D. E. Shulz et al., Soc. Neurosci. Abstr. 19, 638 (1993)]}, events of positive and negative polarity, when occurring in isolation, may be equated, respectively, with dominating excitatory and inhibitory postsynaptic events. This first-order approximation does not take into account nonlinear interaction between excitation and inhibition [L. J. Borg-Graham et al., Nature 393, 369 (1998)]. Furthermore, assuming a reversal potential close to 0 mV for excitatory conductances, detection of glutamate-mediated events is largely favored with respect to GABA (γ-aminobutyric acid)- mediated ones. H-fields are therefore underestimated in our study. The qualitative subthreshold D-field extents reported by Pei et al. (9) are difficult to compare with our own results because of a limited sample size. In addition, their recordings were performed at more depolarized membrane potentials (-30 or -50 mV), where inhibitory hyperpolarizing responses may mask partially the D-field responses, and response amplitudes were measured at a fixed latency.
23. r = -65.8 ± 8.7 mV) than the equilibrium potential for chloride {-77 ± 3.9 mV in vitro [B. W. Connors et al., J. Physiol. 406, 443 (1988)] and -75 mV in vivo [D. E. Shulz et al., Soc. Neurosci. Abstr. 19, 638 (1993)]}, events of positive and negative polarity, when occurring in isolation, may be equated, respectively, with dominating excitatory and inhibitory postsynaptic events. This first-order approximation does not take into account nonlinear interaction between excitation and inhibition [L. J. Borg-Graham et al., Nature 393, 369 (1998)]. Furthermore, assuming a reversal potential close to 0 mV for excitatory conductances, detection of glutamate-mediated events is largely favored with respect to GABA (γ-aminobutyric acid)-mediated ones. H-fields are therefore underestimated in our study. The qualitative subthreshold D-field extents reported by Pei et al. (9) are difficult to compare with our own results because of a limited sample size. In addition, their recordings were performed at more depolarized membrane potentials (-30 or -50 mV), where inhibitory hyperpolarizing responses may mask partially the D-field responses, and response amplitudes were measured at a fixed latency.
24. J. P. Jones and L A. Palmer, J. Neurophysiol. 58, 1187 (1987).
25. Reverse correlation analysis: A d.c. offset was applied to the high-pass filtered voltage records (1 Hz, elliptic filter) such as to center the mode of the membrane potential distribution on a null value. Successive local extrema were defined as being separated by at least 0.5-mV amplitude. Depolarizing events were selected as local positive maxima. Hyperpolarizing events were defined as local minima whose amplitude remained below -1 mV. To assess the statistical significance of the maps, we carried out similar procedures of correlation counts (established over a 25-ms window) in the reverse and forward directions. The reverse-correlation delay [varied from -200 to 0 (retrograde) and 0 to 200 ms (forward) with a 1-ms step] was optimized independently for each visual field pixel such as to maximize the z score given by [number of counts (reverse) - mean (forward)]/SD (forward). The forward procedure allowed us to empirically establish the distribution of the maximal count obtained in a given pixel, under the null hypothesis of independent output and input processes, and define a one-tailed rejection threshold (statistical significance of 1% ). The maps presented here are thus without any temporal significance and focus on the spatial structure of the receptive field. Measures of the spatial extent of maps independent of their patchiness were expressed by the diameter of the disk covering the same total area, expressed in degrees of visual angle.
26. 2 = 0.77 with 2D impulse-like input and 0.88 with long bars), the intersection of which defined the retinal position of the latency basin center (Fig. 4, B and C). This center was superimposed on or close to the MDF center (relative eccentricity: 0.85 ± 0.7°).
27. Despite the fact that both the latency and the strength of the response were highly correlated with the relative eccentricity from the MDF center, both features of the postsynaptic response were weakly correlated together, thus reinforcing the hypothesis that the causative independent variable is the relative eccentricity itself.
28. Z. F. Kisvarday and U. T. Eysel, Eur. J. Neurosci. 5, 1558 (1993).
29. A similar range of ASHP values is found in the adult cat if one takes into account the dependency of the cortical magnification factor on retinal eccentricity [K. Albus, Exp. Brain Res. 24, 181 (1975); R. J. Tusa et al., J. Comp. Neurol. 177, 213 (1978)]. Furthermore, the ASHP values derived from individual latency basins are not significantly linked to the absolute retinal eccentricity of the discharge field of the recorded cells.
30. A similar range of ASHP values is found in the adult cat if one takes into account the dependency of the cortical magnification factor on retinal eccentricity [K. Albus, Exp. Brain Res. 24, 181 (1975); R. J. Tusa et al., J. Comp. Neurol. 177, 213 (1978)]. Furthermore, the ASHP values derived from individual latency basins are not significantly linked to the absolute retinal eccentricity of the discharge field of the recorded cells.
31. A. Grinvald et al., J. Neurosci. 14, 2545 (1994).
32. F. Chavane, V. Bringuier, Y. Frégnac, data not shown.
33. The typical speed of action potential propagation along horizontal excitatory connections is 0.15 to 0.50 m/s in T. Murakoshi et al. [Neurosci. Lett. 163, 211 (1993)], m/s in D. A. Nelson and L. C. Katz [Neuron 15, 23 (1995)], and 0.35 m/s in J. A. Hirsch and C. D. Gilbert [J. Neurosci. 11, 1800 (1991)]. A similar estimate is found for inhibitory connection s: 0.06 to 0.20 m/s in P. Salin and D. A. Prince [J. Neurophysiol. 75, 1589 (1996)].
34. The typical speed of action potential propagation along horizontal excitatory connections is 0.15 to 0.50 m/s in T. Murakoshi et al. [Neurosci. Lett. 163, 211 (1993)], m/s in D. A. Nelson and L. C. Katz [Neuron 15, 23 (1995)], and 0.35 m/s in J. A. Hirsch and C. D. Gilbert [J. Neurosci. 11, 1800 (1991)]. A similar estimate is found for inhibitory connection s: 0.06 to 0.20 m/s in P. Salin and D. A. Prince [J. Neurophysiol. 75, 1589 (1996)].
35. The typical speed of action potential propagation along horizontal excitatory connections is 0.15 to 0.50 m/s in T. Murakoshi et al. [Neurosci. Lett. 163, 211 (1993)], m/s in D. A. Nelson and L. C. Katz [Neuron 15, 23 (1995)], and 0.35 m/s in J. A. Hirsch and C. D. Gilbert [J. Neurosci. 11, 1800 (1991)]. A similar estimate is found for inhibitory connection s: 0.06 to 0.20 m/s in P. Salin and D. A. Prince [J. Neurophysiol. 75, 1589 (1996)].
36. The typical speed of action potential propagation along horizontal excitatory connections is 0.15 to 0.50 m/s in T. Murakoshi et al. [Neurosci. Lett. 163, 211 (1993)], m/s in D. A. Nelson and L. C. Katz [Neuron 15, 23 (1995)], and 0.35 m/s in J. A. Hirsch and C. D. Gilbert [J. Neurosci. 11, 1800 (1991)]. A similar estimate is found for inhibitory connection s: 0.06 to 0.20 m/s in P. Salin and D. A. Prince [J. Neurophysiol. 75, 1589 (1996)].
37. The divergence of the terminal field of thalamocortical fibers extends over a 1- to 2-mm radius on average in the plane of cortical layers [A. L. Humphrey et al., J. Comp. Neurol. 233, 159 (1985)]. This value, replotted in visual field coordinates, is on the order of 1° to 2° of visual angle for central representation of the visual field and remains too low to explain the size of the D-field, which was found to be 4 to 15 times as large. Furthermore, the conduction velocity of X and Y thalamocortical axons is 10 to 100 times as last as that derived from our measurements; the average ASHP value derived from our recordings is on the order of 0.15 m/s, whereas the conduction velocities of X and Y thalamic axons are, respectively, on the order of 8 and 20 m/s in H. P. Hoffmann and J. Stone [Brain Res. 32, 460 (1971)].
38. The divergence of the terminal field of thalamocortical fibers extends over a 1- to 2-mm radius on average in the plane of cortical layers [A. L. Humphrey et al., J. Comp. Neurol. 233, 159 (1985)]. This value, replotted in visual field coordinates, is on the order of 1° to 2° of visual angle for central representation of the visual field and remains too low to explain the size of the D-field, which was found to be 4 to 15 times as large. Furthermore, the conduction velocity of X and Y thalamocortical axons is 10 to 100 times as last as that derived from our measurements; the average ASHP value derived from our recordings is on the order of 0.15 m/s, whereas the conduction velocities of X and Y thalamic axons are, respectively, on the order of 8 and 20 m/s in H. P. Hoffmann and J. Stone [Brain Res. 32, 460 (1971)].
39. A. C. Rosenquist, Cereb. Cortex 3, 81 (1985).
40. Using identical stimuli, we obtained LGN (n = 6) and optic radiation (n = 3) discharge fields of restricted size (1.3° ± 0.6°). These controls preclude possible lateral excitation by light scatter in the retina.
41. A. Das and C. D. Gilbert, Nature 375, 780 (1995).
42. We thank P. Godement, K. Grant, and P.-M. Lledo for helpful comments. We are indebted to L. Borg-Graham and C. Monier for patch electrode data, to D. E. Shulz and V. Ego for LGN recordings, and to P. Baudot, who participated in some of the experiments. This work was supported by CNRS, PROGRES-IN-SERM, AFIRST, GIS-Cognisciences, and Human Frontier Science Program (RG0103/1998-B) grants to Y.F.