ARTICLES

Lateral competition for cortical space by

layer-specific horizontal circuits
Hillel Adesnik1 & Massimo Scanziani1

The cerebral cortex constructs a coherent representation of the world by integrating distinct features of the sensory
environment. Although these features are processed vertically across cortical layers, horizontal projections interconnecting
neighbouring cortical domains allow these features to be processed in a context-dependent manner. Despite the wealth of
physiological and psychophysical studies addressing the function of horizontal projections, how they coordinate activity
Dan O’Shea
among cortical domains remains poorly understood. We addressed this question by selectively activating horizontal
projection neurons in mouse somatosensory cortex, and determined how the resulting spatial pattern of excitation and
inhibition affects cortical activity. We found that horizontal projections suppress superficial layers while simultaneously
d-Lab Journal
activating Club
deeper cortical output layers. This layer-specific modulation does not result from a spatial separation of excitation
and inhibition, but from a layer-specific ratio between these two opposing conductances. Through this mechanism, cortical
17 May 2010
domains exploit horizontal projections to compete for cortical space.

The axons of pyramidal cells in layer 2/3 project vertically, across

layers, and expand horizontally, within layers, to contact neighbour-
a L2/3 b c
ing cortical domains (for example, orientation domains in visual L4
cortex or whisker representation domains in somatosensory L5 0.5 s 10 Light
cortex)1–7. Layer 2/3 pyramidal cells are thus poised to coordinate Light 8 Control

µV2 Hz–1
LFP 6
the activity of neighbouring domains with respect to their own8. The
relationship between the spatial extents of excitation and inhibition 4
2
generated by layer 2/3 pyramidal cells is likely to be instrumental in Unit
0
how these neurons regulate activity across and within layers. On one 200 µm
500 µV 0 20 40 60 80 100
100 µV
hand, the spatial extent of excitation generated by layer 2/3 pyramidal 100 ms Frequency (Hz)

cells across and within layers can be inferred from anatomical and
1–4,9,10 d e f
Overview

- Expression of ChR2 in L2/3 of whisker barrel cortex via in utero electroporation

- Photoinduced gamma activity in vivo and in vitro with light ramp

- Excitation and inhibition matched vertically across layers and horizontally

- Layer-specific ratio of excitation to inhibition leads to:

- feed-forward facilitation of L5 (high E/I from L2/3)

- lateral suppression within L2/3 (low E/I from L2/3)

 doi: 10.1038/nature08935
Expression of ChR2 in Layer 2/3
SUPPLEMENTARY
- in utero electroporation (E15-E16) of ChR2 and fluorophore DNA INFORM
- 23% ± 2% expression in L2/3 excitatory cells
matosensory cortex. We a
aramp of blue light, that
L2/3 b Adesnik and Scanziani,
L2/3
Supplementa
nsity (see Methods); we
L4 L4
id the fast desensitizing
L5
ary Fig. 1c) and moni- L5

ar electrode inserted in Light

200 µm L6
lurry of uncoordinated LFP
rhythmic oscillation NeuN
of
ncy (average 42 6 2 Hz, c
GFP-ChR2 EPSC d
1.2 16
the activity of simulta-

pA2 Hz–1
L4
(×103)

Unit L2/3 L6
the LFP (Fig. 1b). The 0.8 L5 Ivry 1996
Overview

- Expression of ChR2 in L2/3 of whisker barrel cortex via in utero electroporation

- Photoinduced gamma activity in vivo and in vitro with light ramp

- Excitation and inhibition matched vertically across layers and horizontally

- Layer-specific ratio of excitation to inhibition leads to:

- feed-forward facilitation of L5 (high E/I from L2/3)

- lateral suppression within L2/3 (low E/I from L2/3)

Ramped photostimulation of L2/3 elicits gamma oscillations in vivo
- 2 s light ramp 0.1-2 mW/s to maximum 2 mW
- peak oscillation frequency 42 ± 2 Hz
Ramped photostimulation of L2/3 elicits gamma oscillations in vitro
- peak oscillation frequency 33 ± 1 Hz
- abolished by NBQX+CPP (glutamate receptor antagonist) or gabazine (GABAA antagonist)
Gamma oscillation peak frequency increases with
Adesnik andlight ramp
Scanziani, slope inFigures
Supplementary vivo
!
!
!

Supplemental Figure 2: "#$#%&#%'#!()!*+#!(,'-../*-(%!)0#12#%'3!(%!*+#!,.($#!()!*+#!$+(*(4,*-52.2,!in

vivo6 a, In vivo! 789! :;/%&! $/,,! )-.*#0#&! ;#*<##%! =4=>>! ?@A! 0#'(0&-%B! )0(5! ./3#0! CDE! -%! 0#,$(%,#! *(! /!
$+(*(4,*-52.2,! <+(,#! ,.($#! </,! F2,*! ,.-B+*.3! ,*##$#0! *+/%! <+/*! ,2))-'-#%*! *(! #.-'-*! /%! (,'-../*-(%! :>6GE!
5HD,I!;./'JA!(0!-%!0#,$(%,#!*(!/!$+(*(4,*-52.2,!*+/*!*0-BB#0#&!*+#!5/K-5/.!(,'-../*-(%!)0#12#%'3!:>6LM!
5HD,I! B0/3A6! b,! N(05/.-@#&! $(<#0! ,$#'*0/! ()! *+#! 0#,$(%,#! *(! *+#! *<(! $+(*(4,*-52.-6! N(*#! *+#! ,+-)*! -%!
)0#12#%'36!c,!9.(*!()!*+#!$#/J!)0#12#%'3!:;#*<##%!C>4O>!?@A!/B/-%,*!$+(*(4,*-52.2,!,.($#!)(0!M!,-5-./0!
#K$#0-5#%*,6!P00(0!;/0,!/0#!Q,6#656!!!!
Overview

- Expression of ChR2 in L2/3 of whisker barrel cortex via in utero electroporation

- Photoinduced gamma activity in vivo and in vitro with light ramp

- Excitation and inhibition matched vertically across layers and horizontally

- Layer-specific ratio of excitation to inhibition leads to:

- feed-forward facilitation of L5 (high E/I from L2/3)

- lateral suppression within L2/3 (low E/I from L2/3)

L2/3 driven excitation and inhibition are matched vertically
L2/3 driven excitation and inhibition are matched horizontally

- patch L2/3 or L5 pyramidal, voltage clamp at -70 mV or 0 mV

- photostimulate L2/3 in adjacent barrel columns
itiated 400 mm away in the adjacent (10 6 10% increase two columns away, P 5 0.02; Fig. 4c). Pyra
6 7% (n 5 12) in the parent barrel cells in both sublayers 5A and 5B were facilitated, and this facili
n andL2/3
0.004driven excitation
for inhibition; Fig.
IPSC
and inhibition
3d). occurred areamatched
over similar horizontally
horizontal distance from the initiation

- patch L2/3 or L5 pyramidal, voltage clamp at -70 mV or 0 mV

- photostimulate L2/3 inEPSC
adjacent–3barrel columns–2 –1 0 1

b Charge Power (20–60 Hz) cd Charge

1.0 L2/3 1.0L2/3 1 1.0 L5

Charge (norm.)
s
0.8 pA
500
Charge (norm.)

Charge (norm.)
0.5 0.5

Power (norm.)
125 pA 0
0.6 ms
100 1.0 EPSC

L2/3 0.0 –1 0.0

–4 –2 0 2 4 Barrel co

Charge (norm.)
0.4 Barrel column
0.8 Adjacent
–0.5 EPSC –0.5
IPSC 0.2 0.6
IPSCinv
–1.0 0.0 0.4
–1.0
–4 –2 0 2 4 –4 –2 0 2 4 –4
0.2 –2 0 2
–1 0 1
Barrel column
2 Barrel column
3 Barrel column
0.0
3 | Horizontal 0 200 400 600
Power (20–60 Hz) c
Figure Chargematch of excitation and(20–60
Power inhibition
Hz) within layers. 2/3 pyram
Distance from cell (µm
0 a, The recording
1.0 L5 configuration is shown 2000
1.0 L5 at the top left. Blue and red traces (n 5 12).
L2/3 1.0 IPSC
indicate IPSCs s and EPSCs, respectively, recorded in a layer 2/3 pyramidal Middle a
8 0.8 Barrel co

Charge (norm.)
Charge (norm.)

0.5response to focal light stimulation translated tangentially across 0.8 barrel

Power (norm.)
cell in microme
Adjacent
6 columns. Numbers below traces 0.6
indicate distance (in barrel columns)
0.6 of the when the
4 L5 0.0
light-stimulated barrel column with 0.4respect to the home barrel column
0.4 somatose
2
–0.5
(position 0). The inset to the right0.2 shows normalized charge (excitation
0.2
in (n 5 10);
red and inhibition in blue) plotted against distance. b, Plots of average are 6s.e.
0 –1.0 0.0 0.0
–4 –2 0 2 4 charge –4 (left)–2and power
0 2between
4 20–60–4 Hz–2(right)
0 against
2 distance
4 in0 layer
200 400 600
Barrel column Barrel column Barrel column Distance from cell (µm

tation and inhibition within layers. ©2010

2/3 pyramidal cells (n 5 19). Macmillan
c, Same as b butPublishers
for layer 5Limited. All r
pyramidal
Overview

- Expression of ChR2 in L2/3 of whisker barrel cortex via in utero electroporation

- Photoinduced gamma activity in vivo and in vitro with light ramp

- Excitation and inhibition matched vertically across layers and horizontally

- Layer-specific ratio of excitation to inhibition from L2/3 leads to:

- feed-forward facilitation of L5 - high E/I

- lateral suppression within L2/3 - low E/I

Photostimulation increases L5 firing rate, suppresses L2/3 firing rate in vitro
- simultaneously patch L2/3 and L5 pyramidals
- inject current step
- photostimulate same or adjacent barrel with light ramp
- compare L2/3 and L5 cell firing rates with vs. without light

firing rate with / without light

Photostimulation increases L5 firing rate, suppresses L2/3 firing rate in vivo

- whole cell recording from L2/3, inject current step

- multiunit recording from L5
- compare L2/3 and L5 cell firing rates with vs. without light
Excitation / inhibition ratio higher in L5 relative to L2/3
- patch L2/3 and L5 pyramidals, voltage clamp at -70 mV or 0 mV
- compare excitatory / inhibitory charge ratio across all simultaneous pairs
- higher E/I ratio in L5 due to increased inhibition

each dot represents one

simultaneously patched L2/3, L5 pair
Larger excitation / inhibition ratio in L5 neurons with higher firing rates
- cell attached then whole cell recording of L5 pyramidals
- voltage clamp at -70 mV or 0 mV
- compare excitatory / inhibitory charge ratio vs. firing rate during light ramp
Replay of inhibitory waveforms from other layer reverses firing rate change
- record EPSCs in one layer simultaneously with IPSCs in other layer
- current clamp to inject same layer excitatory waveform plus current step
- dynamic clamp to impose inhibitory conductance waveform from other layer

L5 inhibition onto L2/3

leads to increased firing

L2/3 inhibition onto L5

leads to suppressed firing
Overview

- Expression of ChR2 in L2/3 of whisker barrel cortex via in utero electroporation

- Photoinduced gamma activity in vivo and in vitro with light ramp

- Excitation and inhibition matched vertically across layers and horizontally

- Layer-specific ratio of excitation to inhibition from L2/3 leads to:

- feed-forward facilitation of L5 - high E/I

- lateral suppression within L2/3 - low E/I