Spatial connectivity matches direction selectivity in visual cortex

Abstract

The selectivity of neuronal responses arises from the architecture of excitatory and inhibitory connections. In the primary visual cortex, the selectivity of a neuron in layer 2/3 for stimulus orientation and direction is thought to arise from intracortical inputs that are similarly selective^{1,2,3,4,5,6,7,8}. However, the excitatory inputs of a neuron can have diverse stimulus preferences^{1,2,3,4,6,7,9}, and inhibitory inputs can be promiscuous¹⁰ and unselective¹¹. Here we show that the excitatory and inhibitory intracortical connections to a layer 2/3 neuron accord with its selectivity by obeying precise spatial patterns. We used rabies tracing^1,12 to label and functionally image the excitatory and inhibitory inputs to individual pyramidal neurons of layer 2/3 of the mouse visual cortex. Presynaptic excitatory neurons spanned layers 2/3 and 4 and were distributed coaxial to the preferred orientation of the postsynaptic neuron, favouring the region opposite to its preferred direction. By contrast, presynaptic inhibitory neurons resided within layer 2/3 and favoured locations near the postsynaptic neuron and ahead of its preferred direction. The direction selectivity of a postsynaptic neuron was unrelated to the selectivity of presynaptic neurons, but correlated with the spatial displacement between excitatory and inhibitory presynaptic ensembles. Similar asymmetric connectivity establishes direction selectivity in the retina^{13,14,15,16,17}. This suggests that this circuit motif might be canonical in sensory processing.

Extended data figures and tables

Extended Data Fig. 1 Targeted single-neuron electroporation with survival control in vivo.

a, Experimental pipeline: electroporation of the postsynaptic neuron, targeted by shadow-imaging and expression of GCaMP6; imaging of the postsynaptic neuron, labelled by dsRed; injection of the modified rabies virus; imaging and tracing of the presynaptic neurons marked by dsRed. b–d, Schematic of the electroporation technique, performed under a two-photon microscope on a transgenic mouse expressing GCaMP6 in cortical excitatory neurons. b, A pipette filled with DNA plasmids and Alexa Fluor 594 is targeted to a craniotomy; an 820-nm laser (red) excites Alexa Fluor 594 fluorescence (magenta) and GCaMP6 fluorescence (cyan). The latter is insensitive to neural activity because 820 nm is an isosbestic wavelength, at which calcium-bound and calcium-free isoforms fluoresce approximately equally. c, Upon electroporation, DNA plasmids and Alexa Fluor 594 are transferred into a neuron expressing GCaMP6. d, A healthy neuron maintains its GCaMP6 concentration and the resulting calcium-insensitive fluorescence (top), whereas a neuron with a damaged membrane bleeds indicator, gradually darkening and disappearing against the surrounding neuropil (bottom). e, Time lapse of an electroporation in layer 2/3 of mouse V1, using Alexa Fluor 594 negative contrast and calcium-insensitive GCaMP6 fluorescence imaging: approach (left), electroporation (middle) and pipette withdrawal (right). f, The same neuron as in e, imaged the next 1–3 days at 920 nm, expressing the electroporated genes for dsRed (red) and maintaining healthy GCaMP6 expression (green). g, Average fluorescence (mean ± s.e.m.) of somatic Alexa Fluor 594 (magenta) and GCaMP6 (cyan), relative to neuropil background (dashed line represents unity) before and after electroporation (arrow, n = 10 neurons). h, i, As in e, f, but with a slower time lapse of a neuron that survived the procedure. Images in h are 30-s-long averages acquired 0–1 min (left), 2–3 min (centre) and 5–6 min (right) after electroporation. j, k, As in h, i, for a neuron that did not recover from the electroporation. l,As in g, for GCaMP6 somatic fluorescence (mean ± s.e.m.) in neurons that did (cyan, n = 18) or did not (grey, n = 10) survive the procedure. Scale bar, 15 μm (for all fluorescence images).

Extended Data Fig. 2 Photo-ablation of supernumerary postsynaptic neurons before rabies injection.

a, Cartoon of the protocol. The target neuron expresses both GCaMP6 (green) and dsRed (red), whereas surrounding neurons express only GCaMP6. The day after electroporation (day 1), the neuron is targeted with steady two-photon illumination at 820 nm, focused with intensity of approximately 150 mW, for 10–20 s (top). The target neuron is ablated and by the next day (day 2) it has disappeared (bottom). b, Time-lapse imaging during the photo-ablation of two neighbouring neurons, lasting approximately 10 min. Imaging at 920 nm shows calcium-sensitive fluorescence of GCaMP6 (top) and fluorescence of dsRed (bottom). Each of the neurons was targeted with two photo-ablation pulses (red triangles). Photo-ablation pulses lasted 20 s, and each neuron was imaged for 30 s afterwards. Each pulse increases cellular damage: localized photo-bleaching after the first pulse; elevated intracellular calcium and cell swelling after the second pulse. Neighbouring cells, not expressing dsRed, resist the photo-damage. c, The successful elimination of the target neurons (red, see insets) is confirmed the day following the photo-ablation procedure: by day 2, the target neuron has either disappeared (yellow arrow, top neuron) or gone in apoptosis (yellow arrow, bottom neuron). The surrounding tissue is unaffected, as shown by the normal activity detected in neighbouring neurons using Suite2p (blue-green ROIs and traces; scale bar, 30 s and 10 s.d.). d, Effectiveness of photo-ablation as a function of cortical distance from the photo-ablation target neuron (n = 9 attempts, mean ± s.e.m.).

Extended Data Fig. 3 Time course of rabies tracing and recordings.

a, Viability of postsynaptic neurons as a function of day after the rabies virus injection, based on n = 17 injections. b, Count of observed presynaptic neurons traced over the same period from n = 17 postsynaptic neurons (mean ± s.e.m.). c, Fraction of viable presynaptic neurons (red) over the total traced (grey), and worst-case-scenario population mortality of presynaptic neurons (dashed), estimated assuming the viability of each newly labelled presynaptic neuron degrades at the same rate as the viability of the postsynaptic neurons measured in a. d, Average distribution across mice (n = 17, mean ± s.e.m.) of the imaging sessions used to record the responses of presynaptic neurons, split by cortical layer. The time of imaging did not systematically change across layer (black triangle, median, first and third quartiles), and most of the data were acquired before presynaptic neurons suffered from the toxicity of the rabies virus. e, Distribution of responsivity of the presynaptic neurons across days (red violin plot with black median). Responsivity was measured as the maximum average stimulus-triggered response. To compare across sessions, presynaptic responsivity was normalized to the median responsivity of the local population (shaded line). f, Longitudinal imaging of presynaptic neurons identified before (left, black ROI) and 5–12 days after (right, red ROI) rabies virus infection. Scale bar, 25 μm. g, Average responses to drifting gratings of the same presynaptic neurons before (black) and after (red) the rabies virus infection. Responses (4-s long) were z-scored with the respect to blank trials. Scale bar, 1 s.d. h, Preferred orientation of presynaptic neurons recorded before and after the rabies virus infection. At the time of recording used in this study, the preferred orientation of presynaptic neurons is unperturbed by the rabies virus infection (n = 51 neurons from n = 4 mice, circular correlation 0.76, P_r = 1.6 × 10⁻⁶, Z test).

Extended Data Fig. 4 Classification of excitatory and inhibitory presynaptic neurons.

a, Three example excitatory presynaptic neurons showing similar expression of dsRed (left) and decreasing expression of GCaMP6 (middle). The somatic masks obtained from dsRed and the GCaMP6 signal were used to compute a map of phase correlation in the 5-μm-radius annulus around the somatic centre (right). In the first two examples, the stronger GCaMP6 expression in the soma compared to the surrounding neuropil results in positive peaks of the phase correlation. b, As in a, for three example inhibitory neurons. The lack of GCaMP6 expression in the soma results in negative peaks of the phase correlation. c, For each presynaptic neuron, the central peak of phase correlation is plotted against the s.d. of the phase correlation within the 5-μm annulus centred on the soma (top). Darker dots indicate neurons the identity of which was confirmed by immunostaining. Open dots represent example neurons shown in a, b. A classification boundary was used to split the clusters of excitatory (red) and inhibitory neurons (blue). This boundary was identified independently for each presynaptic ensemble with a bilinear fit. The average fit across datasets is shown (black line). The histogram (bottom) summarizes the classification across experiments, with the average classification boundary. d, Average expression of dsRed (left), GCaMP6 (middle) and map of phase correlation (right) for presynaptic neurons classified as excitatory neurons (top, n = 584) or inhibitory neurons (bottom, n = 426). e, Number of presynaptic neurons classified as inhibitory versus excitatory in experiments in CaMK2a–GCaMP6 mice (red upward triangles, n = 13) and GAD2–NLS–mCherry mice (red downward triangles, n = 4). A linear fit (blue, r = 0.6, P_r = 2.6 × 10⁻³, linear correlation, F-test) shows that the fraction of traced presynaptic neurons that are inhibitory tends towards 30% as the yield of tracing increases; yet, owing to the positive intercept, the fraction of inhibitory neurons may appear as high as 60% in experiments with low tracing yield. In control experiments in which the red marker tdTomato was expressed only in excitatory neurons (AAV-CaMK2-tomato in CaMK2a–GCaMP6 mice, grey circles), the fraction of neurons incorrectly classified as inhibitory was below 5%.

Extended Data Fig. 5 Immunohistochemical verification of presynaptic neuron classification method.

a, Coronal maximum projection from a z-stack acquired in vivo, showing a section through a presynaptic network labelled with RV-dsRed (red) in a CaMK2a–GCaMP6s transgenic mouse (green). Scale bar, 50 μm. b, Coronal brain slice matching the section in a. We could match n = 94 out of the n = 105 neurons traced in vivo. c, Inset from b (dashed) showing the immunostaining for the inhibitory neurons marker GAD (blue). d, ROC curve illustrating the sensitivity and specificity of the neuron classification method based on somatic GCaMP6 fluorescence (Extended Data Fig. 4) against the ground-truth measurements obtained by immunostaining against GAD, for a range of classification boundaries. The fitted classification boundary was optimal (blue dot), yielding a specificity of 94% and sensitivity of 91%. e, Two example presynaptic neurons classified as excitatory (n = 49) from the z-stack in a. f, Slice immunostaining of the two neurons in e. Neurons were confirmed as excitatory if they expressed GCaMP6 and were GAD-negative (n = 45). Scale bar, 40 μm. g, h, As in e, f, for two example presynaptic neurons classified as inhibitory. Neurons were confirmed as inhibitory if they did not express GCaMP6 and were GAD-positive (n = 42).

Extended Data Fig. 6 Spectral unmixing of dsRed and mCherry fluorescence.

a, Emission spectra of GCaMP6, dsRed and mCherry. Shaded areas indicate the emission band captured by the green (G) and red (R) channels of the microscope. The G channel collects mostly GCaMP6 fluorescence, and the R channel captures a mixture of mCherry, dsRed and GCaMP6 fluorescence. Emission spectra were normalized to their peak. b, Two-photon action cross-section of dsRed (top, red) and mCherry (bottom, purple). c, Ratio between the two-photon action cross-sections of dsRed and mCherry (red) and its inverse (purple). Shaded bands in b, c indicate the excitation wavelengths used for imaging: 780, 890, 970 and 1,020 nm. The four wavelengths were chosen to maximize the SNR of each fluorophore while ensuring the ratio between the two signals was maximal. d, Example field of view imaged at the four excitation wavelengths in the G channel. GCaMP6 was expressed in all neurons with an AAV2.1-Syn-GCaMP6s in a GAD–NLS–mCherry mouse; a subset of presynaptic neurons was traced with a dsRed rabies virus. e, As in d, for the R channel. f, The fluorescence in the R channel (R_λ) plotted against the fluorescence in the G channel (G_λ) for each pixel, and for each excitation wavelength λ. Because the GCaMP6 labelling was dense and both the dsRed and the mCherry signals were sparse, and because the contribution of dsRed and mCherry to G_λ was minimal, the GCaMP6 signal contributing to R_λ could be recovered by piecewise robust linear regression (αG_λ + β). g, The image representing the linear mix of dsRed or mCherry signals, M_λ, was recovered by subtracting the scaled G_λ from the R_λ. h, An iterative algorithm was used to linearly unmix the two source images. Each unmixing iteration was constrained to minimize the quadratic reconstruction error over the data and return maximally uncorrelated sources R_dsRed and R_mCherry. i, j, The two source images R_dsRed and R_mCherry for the example field of view in d, e. The unmixing procedure correctly recovers the nuclear localization of mCherry without any prior. Scale bars, 50 μm. Similar results were obtained for all GAD–NLS–mCherry mice (n = 4).

Extended Data Fig. 7 Distributions of individual excitatory and inhibitory presynaptic ensembles.

a, Density of excitatory presynaptic neurons around the postsynaptic neuron (black triangle) as a function of cortical depth and radial distance from the postsynaptic neuron. Density was normalized to its maximum for display purposes. Depth marginals are shown on the left of each density map. b, As in a, for inhibitory presynaptic neurons. c, Overlay of excitatory and inhibitory densities shown in a, b. Hue indicates the relative proportion of excitatory (red) and inhibitory (blue) inputs, and saturation indicates maximum normalized neuronal density. Data are shown for the 13 experiments that included substantial recordings from L4.

Extended Data Fig. 8 L2/3 neurons receive presynaptic excitation preferentially tuned to their orientation preference.

a, Average tuning across the postsynaptic neurons responding to drifting gratings (n = 16, mean ± s.e.m.), after alignment of their preferred stimulus direction. b, Distribution of preferred direction relative to the postsynaptic preferred direction, for excitatory presynaptic ensembles across all layers (n = 15, median ± m.a.d.). c, As in b, for excitatory presynaptic ensembles within L2/3 (top) and within L4 (bottom). d, As in b, for inhibitory presynaptic ensembles. e, Average orientation tuning across the postsynaptic neurons responding to drifting gratings (n = 16, mean ± s.e.m.), after alignment of their preferred stimulus orientation. f, Average distribution of preferred orientation relative to the postsynaptic preferred orientation, for excitatory presynaptic neurons pooled across all layers (n = 15, median ± m.a.d.). Presynaptic ensembles tuning for orientation (P_KW = 5 × 10⁻⁸, two-sided one-way Kruskal–Wallis test) was significantly stronger than expected from random samples of the surrounding population (grey, median ± m.a.d.). g, As in f, for presynaptic ensembles within L2/3 (top, n = 15, P_KW = 10⁻⁷) and within L4 (bottom, n = 13, P_KW = 2 × 10⁻⁵). h, As in f, for inhibitory presynaptic ensembles (n = 4). i, The tuning of the distributions of preferred orientation of excitatory presynaptic ensembles across layers plotted against the preferred orientation of their postsynaptic neuron. The co-tuning for orientation (r = 0.75, circular correlation, P_r = 4 × 10⁻³, Z-test; P_V = 9 × 10⁻⁸ circular V-test) was stronger than expected from random samples of the local population (P_{r_rand} < 10⁻⁴ for circular correlation, P_{V_rand} < 10⁻⁴ for V statistic). Upwards triangles represent experiments in CaMK2a–GCaMP6 mice (n = 11); downward triangles indicate experiments in GAD2–NLS–mCherry mice (n = 4). j, As in i, for excitatory presynaptic ensembles within L2/3 (top, r = 0.92, P_r = 5 × 10⁻³, P_V = 10⁻⁵, P_{r_rand} = 2 × 10⁻⁴, P_{V_rand} < 10⁻⁴, n = 15) and within L4 (bottom, r = 0.76, P_r = 2 × 10⁻², P_V = 3 × 10⁻³, P_{r_rand} = 0.18, P_{V_rand} = 3 × 10⁻³, n = 13). k, As in i, for inhibitory presynaptic ensembles. Inhibitory presynaptic ensembles were weakly biased to the orientation preference of the postsynaptic neuron (P_V = 0.05, circular V-test, n = 4).

Extended Data Fig. 9 Mapping retinotopy using individual neurons versus wide-field signals.

a, The stimulus used for retinotopic mapping was a sparse random pattern of white and black squares on a grey background (top). The fluorescence time-course from the entire field of view was used to compute a global stimulus-triggered average response elicited by changes in luminance at each position. The centre of mass of this global receptive field (RF) was used to constrain the fits of wide field and neuronal receptive field to the appropriate retinotopic region (dashed rectangle). b, Maximal projection from an example field of view. In this example, the field of view was subsampled in a grid of 9 × 9 regions of interest (ROIs, red squares) to compute wide-field receptive fields. Scale bar, 100 μm. c, The wide-field receptive fields calculated for the ROIs in b, normalized to their maximum. The wide-field receptive field centres from the grid of ROIs were interpolated to estimate a retinotopic map, assigning a wide-field receptive field to each cortical location, whether it contained a responsive neuron, an unresponsive neuron or neuropil. d, Estimation of neuronal receptive fields. ON (red) and OFF (blue) receptive fields were estimated by regularized smooth pseudoinverse regression using either streams of white (ON) or black stimuli (OFF) as predictors, and assuming a common response kernel across neurons. ON and OFF subfields were then combined to estimate the receptive field centre (green dot). Receptive fields were considered significant if the cross validated correlation coefficient between predicted (red trace) and actual response (black trace) was greater 0.2. e, Azimuth of neuron receptive field centre versus wide-field receptive field centre for all excitatory presynaptic neurons (black dots, n = 113, r_pre = 0.89, P_{r_pre} = 2.8 × 10⁻³⁹, linear correlation, F-test) and surrounding excitatory neurons (red density, n = 25,677, r_all = 0.88, P_{r_all} < 10⁻³⁰⁸, linear correlation F-test) across experiments. f, As in e for elevation (r_pre = 0.80, P_{r_pre} = 9 × 10⁻²⁷; r_all = 0.85, P_{r_all} < 10⁻³⁰⁸). g, h, As in e, f, for inhibitory presynaptic neurons (black dots, n = 37, r_pre = 0.92, P_{r_pre} = 1.3 × 10⁻¹⁵, for azimuth; r_pre = 0.71, P_{r_pre} = 9.7 × 10⁻⁷, for elevation) and all inhibitory neurons (n = 1,963, r_all = 0.95, P_{r_all} < 10⁻³⁰⁸, for azimuth; r_all = 0.74, P_{r_all} < 10⁻³⁰⁸, for elevation).

Extended Data Fig. 10 Spatial connectivity accords with direction selectivity in cortex and across layers.

a, Distribution of excitatory and inhibitory presynaptic neurons in cortex, pooled across experiments (n = 17), and polar tuning curves for each postsynaptic neuron (top). The colour hue indicates the average fraction of local excitatory (red) or inhibitory (blue) presynaptic neurons; the colour saturation indicates the maximum normalized input density, averaged across experiments. b, c, As in a, plotting the excitatory and inhibitory presynaptic neurons separately. d, As in a, after rotating each presynaptic cortical distribution to align the postsynaptic preferred direction (n = 16). The cortical angle of rotation, corresponding to the postsynaptic preferred direction, was calculated from the local retinotopic gradient at the postsynaptic location. After the alignment, the postsynaptic preferred orientation approximately maps to a line at the postsynaptic location (dashed line). e, As in d, for excitatory presynaptic neurons. f, As in d, for inhibitory presynaptic neurons. g, As in d, for the distribution of L1 and L2/3 excitatory and inhibitory presynaptic neurons in visual space, pooled across experiments after alignment to the preferred direction across postsynaptic neurons (n = 16). h, i, As in g, distinguishing between excitatory and inhibitory presynaptic neurons. j–l, As in g–i, for presynaptic neurons in L4 and in superficial L5. In all panels, upwards triangles and circles represent CaMK2a–GCaMP6 datasets; downward triangles and squares indicate GAD2–NLS–mCherry datasets.