A) Schematic of experimental setup. B) Image of loaded neurons in L4. Scale bar: 50μm. C) Mean imaging depths for the two groups. L2/3 mean depth from pia = 190±32 μm; L4 mean depth from pia = 348±46 μm. Average separation of imaging planes: 161μm. D) Left: Injections of latex microspheres at L4 imaging depth confirm imaging in mid-cortical layers. Right: Expression of the layer 4 specific marker Scnn1a (, ) indicates location of L4. We crossed a transgenic mouse expressing cre recombinase under the control of Scnn1a promoter [B6;C3-Tg(Scnn1a-cre)3Aibs/J; JAX strain 9613], with a mouse that expresses td Tomato in a cre-dependent manner [B6.Cg-Gt(ROSA)26Sortm9(CAG-tdTomato)Hze; Jax strain 7909].

A) Sound evoked fluorescence traces (ΔF/F) (mean ± s.e.m.) from two neurons in L2/3 (black) and L4 (red) imaging planes. Horizontal colored bars indicate sound frequency and sound duration. Vertical scale bars are 5%. B) Comparison of average maximum ΔF/F of all responding neurons in pairs of imaging fields. L4 fields showed larger maximum ΔF/F, p<0.03, paired t-test. C) Maximum ΔF/F response of all neurons in both layers; n_L2/3 = 310 responsive neurons; n_L4=341 responsive neurons. L4 neurons showed larger maximum ΔF/F, p<10⁻²³, Wilcoxon rank sum).

A) Fluorescence (ΔF/F) timecourse traces from 3 neighboring neurons in L2/3 (rows). Horizontal colored bars indicate sound frequency and duration (1s). Arrowheads indicate BF. B) Superimposed tuning curves of all responsive neurons in the imaged field. Filled circles indicate BF for each neuron. C) Spatial map of BFs for all imaged neurons in the field. Color indicates BF. White circles indicate unresponsive neurons (Methods). Scale bar: 25μm. D–F) Conventions as in A–C, D) Fluorescent timecourse traces (ΔF/F) from three neighboring neurons in L4 (red). E) Frequency tuning curves of all responsive neurons in the imaged field. F) Spatial map of BFs for all imaged neurons.

A second example of paired imaging fields. Conventions as in . A) Superimposed tuning curves of all responsive neurons in L2/3. B) Spatial map of BFs for imaged neurons in the field. C–D) Frequency tuning and spatial maps of BF for the corresponding L4 imaging field. Scale bar: 25μm.

A) BF variability (IQR_BF) in each paired imaging field. Each set of connected dots represents a pair of imaging fields. L4 fields showed lower variability (median_L2/3 = 0.47±1.3 octaves, median_L4 = 0.16±0.56 octaves, p=0.01, Wilcoxon sign rank) B) Cumulative distribution functions displaying local (<100μm radius) BF variability around each neuron in L2/3 (black dashed line) and L4 (thick gray line). L4 neurons show lower variability (p<10⁻⁶ Wilcoxon rank sum).

A) Scatter plot of median BF of all responsive neurons in each paired imaging plane. Median BF for each pair of imaging fields is similar. B) Percentage of single peaked neurons in each paired imaging field. Horizontal bars indicate the mean for each group. Mean values in each group were similar (L2/3: 59±7% sem and L4: 65±7% sem; p>0.3, paired t-test) C) Cumulative distribution functions displaying a measure of frequency tuning curve bandwidth (BW₇₅) for single peaked L2/3 (black) and L4 (red). Median_L2/3 = 0.23±0.18 iqr octaves; median_L4 = 0.28±0.18 iqr octaves, p<0.001, Wilcoxon rank sum). D) BF variability (IQR_BF) for only single peak neurons in paired imaging fields. Each set of connected dots represents the IQR_BF for all responding neurons within a pair of imaging fields, horizontal bars indicate the medians of each group (median_L2/3 = 0.31±0.97 iqr octaves, median_L4 = 0±0.25 iqr octaves. BFs across entire L4 fields are less variable (p<0.03, Wilcoxon sign rank). E) Cumulative distribution functions displaying local (<100μm radius) BF variability of only single peak neurons in L2/3 (black) or L4 (red). Local BF variability is less in L4 (median_L2/3 = 0.25±0.875 iqr octaves; median_L4 = 0±0.375 iqr octaves, p<10⁻⁸, Wilcoxon rank sum). F) Distributions of ratios of peak frequencies of multipeak neurons in L2/3 (gray) and L4 (red) is similar. Arrow indicates secondary peak representing octave spacing.

A) Significant pairwise signal cross-correlations in L2/3 for an example imaging field. Light gray circles show correlations obtained from a randomized version of the data (Methods, n = 91 pairs, p=10⁻²⁴, paired t-test) B) Significant pairwise signal cross-correlations in L4 (n = 136 pairs, p=10⁻⁵⁹, paired t-test) C) Comparison of distributions of signal correlation values in L2/3 (black) and L4 (red). Vertical lines indicate means of distributions (asterisks indicate p<10⁻⁰⁹, unpaired t-test). Inset: Scatter plots comparing SC values and SC-NC values. Each symbol represents SC value (x-axis) or SC values corrected for the contributions of noise correlations (y-axis) for a pair of neurons in L2/3 (black) or L4 (red) D, E) Significant pairwise noise correlations in L2/3 and L4 in example field (p=10⁻³⁹ and 10⁻²³ respectively). F) Comparison of distributions of noise correlation values in L2/3 (black) and L4 (red). Vertical lines indicate means of distributions, p<10⁻³⁰, unpaired t-test.

A) Significant pairwise signal cross-correlations in L2/3 for a second example field. Light gray circles show correlations obtained from a randomized version of the data; n = 78 pairs, p<0.001, paired t-test B) Significant pairwise signal cross-correlations in L4. n = 300 pairs, p=10⁻⁶⁷, paired t-test C) Comparison of distributions of signal correlation values in L2/3 (black) and L4 (red). Vertical lines indicate means of distributions; asterisks indicate p<10⁻⁰⁷, unpaired t-test. Inset: Scatter plots comparing SC values and SC-NC values. Each symbol represents SC value (x-axis) or SC values corrected for the contributions of noise correlations (y-axis) for a pair of neurons in L2/3 (black) or L4 (red) D, E) Significant pairwise noise correlations in L2/3 and L4 in example field (p<0.001 and 10⁻¹¹ respectively). F) Comparison of distributions of noise correlation values in L2/3 (black) and L4 (red). Vertical lines indicate means of distributions p=0.08, Wilcoxon rank sum).

A, B) Significant pairwise signal cross- correlations (SC) in the population of L2/3 (black) and L4 (red). Light gray circles represent correlations obtained from a randomized version of the data; black or red circles indicate correlation values from the actual data. C) Distributions of SC values from L2/3 (black) and L4 (red) from all imaged fields (2,968 L2/3 neuron pairs; 3,988 L4 neuron pairs). Top: SC values; Bottom: SC values corrected for the contribution of noise correlations (SC-NC). Vertical lines indicate medians of distributions, asterisks indicate p<10⁻¹⁹ (SC) and p<10⁻²⁶ (SC-NC), Wilcoxon rank sum test. D) Scatter plots comparing SC values and SC-NC values. Each symbol represents SC value (x-axis) or SC values corrected for the contributions of noise correlations (y-axis) for a pair of neurons in L2/3 (top) or L4 (bottom), E–F) Significant pairwise noise correlations in L2/3 and L4 layers. Light gray circles represent correlations obtained from a randomized version of the data; black or red circles indicate correlation values from the actual data. G) Comparison of distributions of noise correlation values in L2/3 (black) and L4 (red). Vertical lines indicate medians of distributions p= 0.009, Wilcoxon rank sum.

A) Noise and signal correlation for individual cell pairs are correlated. Data was normalized for comparison across animals and imaged regions by computing z-scores (see Methods). Slope_L2/3 = 0.52; Slope_L4 = 0.45; p<0.002. B) Display of cell densities in the two imaging planes (L2/3: black; L4: red). C) Signal correlations in L2/3 and L4 as function of cell-pair distance. Data was grouped into ~10μm bins and all correlations values occurring within that distance increment were plotted. Data from the two groups (L2/3 and L4) were offset for clarity. Lines (black and red) represent the median values in each distance bin. D) Conventions as in C) Noise correlations in L2/3 and L4 as function of cell-pair distance.