Layer 4 Gates Plasticity in Visual Cortex Independent of a Canonical Microcircuit

Abstract

Disrupting binocular vision during a developmental critical period can yield enduring changes to ocular dominance (OD) in primary visual cortex (V1). Here we investigated how this experience-dependent plasticity is coordinated within the laminar circuitry of V1 by deleting separately in each cortical layer (L) a gene required to close the critical period, nogo-66 receptor (ngr1). Deleting ngr1 in excitatory neurons in L4, but not in L2/3, L5, or L6, prevented closure of the critical period, and adult mice remained sensitive to brief monocular deprivation. Intracortical disinhibition, but not thalamocortical disinhibition, accompanied this OD plasticity. Both juvenile wild-type mice and adult mice lacking ngr1 in L4 displayed OD plasticity that advanced more rapidly L4 than L2/3 or L5. Interestingly, blocking OD plasticity in L2/3 with the drug AM-251 did not impair OD plasticity in L5. We propose that L4 restricts disinhibition and gates OD plasticity independent of a canonical cortical microcircuit.

Keywords: amblyopia; cortical circuit; critical period; experience-dependent plasticity; myelin; ocular dominance; reticulon receptor; visual cortex.

Figure 1.. Selective loss of ngr1 in forebrain is sufficient to retain OD plasticity in adult mice

(A) Contralateral Bias Index (CBI) scores for non-deprived adult WT mice (WT, n=6), juvenile WT mice following 4 days of monocular deprivation (4d MD) (WT CP 4d MD, n=8), adult non-deprived ngr1 −/− mice (KO, n=6), adult ngr1 −/− mice following 4 d MD (KO 4d MD, n=6), adult non-deprived ngr1 flx/flx; L2–6-Cre mice (L2–6 Cre, n=4), and adult ngr1 flx/flx; L2–6-Cre mice following 4d MD (L2–6 Cre 4d MD, n=6). Individual mice are represented as circles. The bar represents the mean of each group. The range of typical CBI values for non-deprived adult WT mice are demarcated by the grey rectangle. KW test comparing non-deprived and 4d MD for each genotype. (B) Cumulative distributions of units for non-deprived adult ngr1 flx/flx; L2–6-Cre mice and following 4d MD for units in L2/3 (73, 91), L4 (62, 84), L5 (56, 50). MD yields a significant shift in the distribution of recorded units for each layer (P<.0001, KW test comparing non-deprived and 4d MD for each layer). See also Figure S1.

Figure 2.. L4 gates OD plasticity in visual cortex through ngr1

(A) Layer selective activity of Cre recombinase revealed with immunofluorescent staining with antibodies directed against GFP. Coronal sections from adult ngr1 flx/flx mice (flx) receiving in utero electroporation of pCAG: CRE-GFP at E15.5 (L2/3-Cre), as well as in combination with several Cre driver lines: Scnn1a-Tg3-Cre (L4-Cre), Rpb4a-Cre (L5-Cre), and Ntsr1-Cre (L6-Cre). Scale bar, 100 microns. Relative positions of each cortical layer are indicated on the left. (B) CBI scores for adult, ngr1 flx/flx; L2/3-Cre mice (L2/3-Cre, n=7), ngr1 flx/flx; L4-Cre mice (L4-Cre, n=8), ngr1 flx/flx; L5-Cre mice (L5-Cre, n=8), and ngr1 flx/flx; L6-Cre mice (L6-Cre, n=8) following 4 days of MD, as well as non-deprived adult L4-Cre mice (n= 5). The ngr1 flx/flx; L2–6-Cre mice (L2–6 Cre, n=6) from Figure 1 are shown for reference. CBI scores for L4-Cre mice following 4d MD are significantly lower than both flx alone mice after MD (P=.005) as well as non-deprived L4-Cre mice (P=.006), whereas flx alone is not significantly different from L2/3-Cre, L5-Cre or L6-Cre (KW test). The bar represents the mean of each group. The range of typical CBI values for non-deprived adult WT mice from Figure 1 are demarcated by the grey rectangle. (C-E) Cumulative distributions of units for the genotypes and deprivation conditions in (B) classified according to layer (n=units). (C) L2/3: flx (96), L2/3-Cre (80), L4-Cre (86), L5-Cre (88), and L6-Cre (85) (D) L4: flx (61), L2/3-Cre (60), L4-Cre (73), L5-Cre (70), and L6-Cre 66) (E) L5: flx (29), L2/3-Cre (54), L4-Cre (50), L5-Cre (40), and L6-Cre (45). L4-Cre is significantly different from flx for each layer (P=.004 or lower, KW test). See also Figure S2.

Figure 3.. Deletion of ngr1 in L4 with a distinct Cre driver also sustains OD plasticity in adult mice

(A) Coronal sections of adult mouse brain from HDC-Cre (left) and NR5a-Cre (right) in combination with the Cre reporter Ai14 (tdTomato). Cell soma labeled with red fluorescence are evident in the lateral geniculate nucleus (LGN) in HDC-Cre mice and in L4 of Nr5a-Cre mice. Scale bar = 0.5mm (B) Higher magnification images of coronal sections of visual cortex from these same genotypes. Scale bar, 100 microns. Relative positions of each cortical layer are indicated on the left. HDC-Cre section is oversaturated to highlight thalamocortical axons. (C) CBI scores for ngr1 flx/flx; HDC-Cre (n=4 mice) and ngr1 flx/flx; Nr5a-Cre (n=4 mice) following 4 days of MD and non-deprived ngr1 flx/flx; Nr5a-Cre (n=5 mice). The range of typical CBI values for non-deprived adult WT mice from Figure 1 are demarcated by the grey rectangle. CBI scores for ngr1 flx/flx; Nr5a-Cre mice following MD are significantly lower than those of ngr1 flx/flx; HDC-Cre mice after MD and non-deprived ngr1 flx/flx; Nr5a-Cre mice (P = .022 and .0391, respectively, KW test). (D) Cumulative distributions of units for these same mice (n=units): HDC-Cre 4d MD(120), Nr5a-Cre 4d MD (111), Nr5a-Cre non-deprived (166). Nr5a-Cre 4d MD is significantly different from HDC-Cre 4d MD and Nr5a-Cre non-deprived (both P<.0001, KW test). (E) The layer of Cre expression, the estimated percentage of excitatory cortical neurons expressing Cre, and whether Cre expression permitted (+) or did not permit (−) OD plasticity in adult mice in each layer for mice with deletion of ngr1 restricted by IUEP or a Cre driver line. OD plasticity by layer for NR5a-Cre is data not shown. See also Figure S3.

Figure 4.. Deletion of ngr1 in L4 permits intracortical disinhibition with 1-day MD in adult mice

(A) Schematic of the recording configuration. PV interneurons expressing GFP directed are patched in the whole-cell configuration. A UV laser directs the focal release of glutamate over the soma of excitatory neurons distributed throughout the tissue section. Glutamate uncaging drives the firing of APs by neurons under the region of brief UV illumination. (B) An example of the 16×16 grid (aqua dots) and the position of a recorded PV interneuron on L2/3 (red circle). (C) An example of the current induced by direct somatic stimulation (1, red trace, upper) of the recorded PV interneuron, and excitatory synaptic currents (2, blue trace, lower). (D) LSPS mapping of excitatory synaptic inputs onto PV interneurons in L2/3 of adult (P55-P65) ngr1 flx/flx; L4-Cre; GAD67-GFP mice (non-deprived (ND) n=10 cells, 1-day (1d) MD n=16 cells). Relative positions of each cortical layer are indicated on the right. These mice display a loss of excitatory drive with 1d MD similar to during the critical period. (E) Average excitatory synaptic input to L2/L3 PV interneurons by layer for non-deprived mice and after 1d MD. Synaptic current per layer is plotted as mean ± SEM. Synaptic current is significantly lower from L2/3 (P=.025), L4 (P=.036), and L5 (P <.001) following MD (Two-way RM ANOVA, Sidak’s multiple comparison test). See also Figure S4.

Figure 5.. Thalamocortical E/I ratio and mEPSC amplitude are unaltered by 1-day MD

(A) Schematic of the recording configuration. Pyramidal excitatory neurons in L4 of visual cortex of juvenile (P26–30) wild-type HDC-Cre; Ai32 (ChR2-YFP) mice are patched in the whole-cell configuration. Thalamic axons projecting into V1 express ChR2-YFP. (B) Example traces of direct responses recorded at a holding potential of −50mV; disynaptic inhibitory currents are recording at a holding potential of 5mV. Synaptic release is evoked with a 5ms pulse of blue light (blue line). (C) Neurons from non-deprived (ND) mice (n = 16) and mice after 1 day (1d) of MD (n=20) display similar E/I ratios (P=.147, MW test). (D) Example traces optogenetic-evoked putative thalamocortical mEPSCs desychronized by the presence of Sr²⁺. The light pulse is indicated by the blue line. Gold asterisks identify mEPSCs. (E) Neurons from non-deprived (ND) mice (n = 14) and mice after 1d of MD (n=17) display similar mEPSC amplitudes (P=.89, MW test). See also Figure S5.

Figure 6.. OD plasticity is first detectable in L4 in juvenile WT and adult ngr1 flx/flx; L4-Cre mice

(A) Cumulative distributions of ODI values for units in L2/3 (left), L4 (middle), and L5 (right), for non-deprived critical period (CP) WT mice (ND, dashed line) and following 2 days of MD (2d MD, light grey line) and 4 days of MD (4d MD, dark grey). Units per layer in parentheses, L2/3: ND (36), 2d (73), 4d (39); L4: ND (39), 2d (60), 4d (40); L5: ND (29), 2d (38), 4d (38). Statistical comparison is a KW test comparing all combinations of ND, 2d MD, and 4d MD for each layer. (B) Cumulative distributions of ODI values for units in L2/3 (left), L4 (middle), and L5 (right), for non-deprived adult ngr1 flx/flx; L4-Cre mice (ND, dashed line) and following 2d MD (2d MD, light green) and 4d MD (4d MD, dark green). Units per layer in parentheses, L2/3: ND (57), 2d (63), 4d (86); L4: ND (44), 2d (58), 4d (73); L5: ND (19), 2d (51), 4d (50) Statistical comparison is a KW test comparing all combinations of ND, 2d MD, and 4d MD for each layer. (C) The percentage of the total OD shift per layer between non-deprived mice and after 4d of MD achieved by 2d of MD for both CP WT mice and adult ngr1 flx/flx; L4-Cre mice. See also Figure S6.

Figure 7.. OD plasticity in L5 does not require OD plasticity in L2/3

Cumulative distributions for units per layer for juvenile WT mice either non-deprived (ND) or after 4 days (4d) of MD during treatment with AM-251 or vehicle (n=units). (A) AM-251 blocks OD plasticity in L2/3 as AM-251 treated (58) is not significantly different from ND (73) (P>.99) while vehicle is significantly shifted towards the open eye (36) (P=.0002, KW test). (B) AM-251 does not affect OD plasticity in L4 as both AM-251 treated (75) and vehicle treated (40) are significantly different than ND (64) (P=.002, P=.01, respectively, KW test). (C) AM-251 does not affect OD plasticity in L5 as both AM-251 treated (74) and vehicle treated (31) are significantly different than ND (62) (P=.0001, P=.0021, respectively, KW test). See also Figure S7.