Figure 6. Two parallel circuits in superficial granular retrosplenial cortex.

(A) Left: Schematics of dual injections into anterior thalamus (ADAV) and claustrum (CLA; top) or anterior cingulate (ACC; bottom). Right: Resulting dual expression of ADAV (green) and CLA (red; top) or ACC (magenta; bottom) axons with zoomed-in confocal view of layers, demarcated by white lines. (B) Left: Projection density of anterior thalamus (ADAV; green; n = 5), claustrum (CLA; red; n = 6), anterior cingulate (ACC: magenta; n = 4), and dorsal subiculum (DS; yellow; n = 4; mean ± SEM shaded for all) axons and terminal arbors in RSG. Note distinct sublaminar distribution of axons from ADAV, ACC, and CLA in layer 1. Right: Projection density of LR (purple; n = 10) and RS (blue; n = 5; mean ± SEM shaded for all) neurons. Note the distinct difference between LR and RS dendrite lamination across the layers. (C) Correlation matrix of means for all axon-axon, axon-dendrite, and dendrite-dendrite comparisons. Note that LR dendrites are significantly positively correlated with ADAV and DS axons but negatively correlated with CLA and ACC axons, while RS dendrites are significantly positively correlated with CLA and ACC axons but negatively correlated with ADAV and DS axons. (D) Summary schematic showing selective control of LR neurons by inputs from ADAV and DS, including head direction signals. In contrast, RS neurons are preferentially controlled by inputs from the CLA and ACC. This precise organization of principal cell dendrites and afferent axons forms two parallel circuits in superficial RSG.

Figure 6—figure supplement 1. Contralateral RSG projections drive RS, but not LR, neurons.

(A) Left: Schematic of contralateral RSG (cRSG) ChR2-mCherry injection (opposite hemisphere compared to recordings). MCherry is pseudo colored as gray. Middle: Schematic of ipsilateral ADAV eYFP injection (same hemisphere as recordings). Right: Schematic of target recording region in RSG. (B) Confocal image of dual-injection expression with cRSG axons and terminal arbors in gray and ADAV axons and terminal arbors in green. Layers are demarcated by black lines. Note the precise targeting of L1c, L2, and L5 by cRSG axons and terminal arbors (gray) in contrast to the L1a and L3 targeting of anterior thalamic arbors (green). (C) Schematic showing cRSG inputs to patched LR and RS neurons being stimulated at L1/2 boundary. (D) Example traces of two sequential pairs of RS (blue) and LR (purple) neuron responses to the high power LED pulse. (E) Significantly larger EPSP amplitudes in RS (n = 5) cells compared to LR (n = 7) cells (p=0.0051; Wilcoxon rank sum test). Error bars are SEM.

Figure 6—figure supplement 2. Sublaminar differences in thalamic vs claustral projections to medial prefrontal cortex.

Each panel shows overview of expression in a cortical area (left) and its layer 1 only (right). Images of axonal projection patterns were obtained from the Allen Mouse Brain Connectivity Atlas (https://connectivity.brain-map.org; Oh et al., 2014). The following images, as numbered in high resolution image viewer, were used: for TC, ventromedial nucleus (VM; experiment #127797441), image #69 for injection and #36 for PrL/IL; for CLA (experiment #485903475), image #56 for injection and #31 for PrL/IL. Abbreviations: PrL - prelimbic cortex and IL - infralimbic cortex. Scale bar is 70 μm. (A) Axonal projection expression in infralimbic and prelimbic cortices from anterograde injection into ventromedial thalamus. Projections show strong preference for superficial parts of layer 1. (B) Axonal projection expression in infralimbic and prelimbic cortices from anterograde injection into claustrum. Higher density of projections can be observed at deep layer 1, while very few axons reach superficial-most parts.