ATN and RSC send strong axonal projections to the superficial layers of dorsal Presubiculum

A, Retrograde labeling of cortical and subcortical regions projecting to the presubiculum (PrS) with retrobeads.

B, Ipsilateral anterodorsal and anteroventral thalamic nuclei are labeled with beads.

C, Ipsilateral granular (gRSC) and dysgranular (dRSC) retrosplenial cortex labeling. Scale bars B, C, 200 µm.

D, Anterograde labeling of thalamic and retrosplenial projections to PrS with AAV-Chronos-GFP.

E, F, Injection sites in ATN (E) and RSC (F). AD: anterodorsal thalamic nucleus, AV: anteroventral thalamic nucleus, CA1: field of the hippocampus, cc: corpus callosum, cg: cingulum, CPu: caudate putamen, DG: dentate gyrus, MD: thalamic medial dorsal nucleus, MEC: medial entorhinal cortex, PaS: parasubiculum, PVA: paraventricular thalamic nucleus, anterior, Rt: thalamic reticular nucleus, sm: stria medullaris, Sub: subiculum, VL: thalamic ventrolateral nucleus.

G, H, Five sequential 100-µm horizontal slices of the parahippocampal region show ATN (G) and RSC (H) axons expressing Chronos-GFP (green). Dashed lines show limits of the parahippocampal region to the left and of the dentate gyrus to the right. Numbers show dorsoventral level with respect to bregma. Scale bars 100 µm.

I, Normalized profiles of fluorescent intensity for ATN and RSC projections to the presubiculum, from white matter (WM) to pia (PIA). Mean (green) ± SEM (grey), n = 8 slices each.

J, Ventral (V) to dorsal (D) normalized distribution of ATN and RSC projections to the presubiculum, n = 3 mice.

K, ATN axon labeling (green) is segregated from calbindin labeling (white) in the PrS. Scale bar 100 µm.

See also Figure 1—figure supplement 1.

Brain regions providing input to the Presubiculum.

Semi-quantitative estimate of retrobeads labelling in serial coronal slices following injection in the presubiculum. Color intensity represents beads density (mouse #69).

  • Thalamic nuclei: anterodorsal (AD) +++, anteroventral (AV) +++, laterodorsal (LD) +++, lateroposterior ++, dorsal lateral geniculate ++, reuniens ++, anteromedial -, ventrolateral -

  • Retrosplenial cortex: dysgranular (dRSC) +++, granular +++, dysgranular contralateral ++

  • Visual cortex: primary (V1) ++, secondary (V2L) +++

  • Entorhinal cortex: lateral (LEnt) +++, medial (MEnt) +++

  • Hippocampus/parahippocampus: parasubiculum (PaS) +++, subiculum (Sub) +++, controlateral PrS ++, CA1 ++, CA3 -, dentate gyrus –

  • Other: claustrum ++, perirhinal cortex ++, temporal association cortex ++

    ADN: anterodorsal thalamic nucleus; AM: anteromedial thalamic nucleus; APT(V): anterior pretectal nucleus (ventral part); AVDM: anteroventral thalamic nucleus, dorsomedial part; AVVL: anteroventral thalamic nucleus, ventrolateral part; CA: cornu ammonis; cc: corpus callosum; Cl: Claustrum; Cg1,2: cingular cortex area 1,2; coll. sup: colliculus superior; CPu: caudate putamen (striatum); dhc: dorsal hippocampal commissure; DG: dentate gyrus; DLG: dorsal lateral geniculate nucleus; LEnt: lateral entorhinal cortex; LDDM: laterodorsal thalamic nucleus, dorsomedial part; LDVL: laterodorsal thalamic nucleus, ventrolateral part; LMN: lateral mammillary nucleus; LS: lateral septal nucleus; M1,2: primary, secondary motor cortex; MEnt: medial entorhinal cortex; MMN: medial mammillary nucleus; PAG: periaqueductal gray; PaS: Parasubiculum; PF: parafascicular thalamic nucleus; PH: posterior hypothalamic area; PMCo: posteromedial cortical amygdaloid nucleus; Po: posterior thalamic nuclear group; PRh: perirhinal cortex; PrS: presubiculum; PVA: paraventricular thalamic nucleus, anterior part; S1BF, S1FL, S1ULp, S1Tr: primary somatosensory cortex, barrel field, forelimb region, upper lip region, trunk region; S2: secondary somatosensory cortex; sm: stria medullaris of the thalamus; Sub: subiculum; Subst. Nigra: substantia nigra; Re/Reuniens: reuniens thalamic nucleus; RSA = dRSC, dysgranular retrosplenial cortex; RSC: retrosplenial cortex; RSG = gRSC: granular retrosplenial cortex; Rt: reticular thalamic nucleus; V1: primary visual cortex; V2M: secondary visual cortex, medial part; V2L: secondary visual cortex, lateral part; VA: ventral anterior thalamic nucleus; VDB: nucleus of the vertical limb of the diagonal band; VL: ventrolateral thalamic nucleus; VPL: ventral posterolateral thalamic nucleus; VPM: ventral posteromedial thalamic nucleus

Calbindin immunostaining and GFP-expressing ATN axons in horizontal sections.

A, Subicular (Sub), Parasubicular (PaS), medial entorhinal cortical (MEC), and presubicular (PrS) borders are outlined, as well as the dentate gyrus (DG). Patches of high density of ATN axonal ramifications in L3 are indicated as “ATN microdomains”. B, Territories devoid of ATN axons in PrS L2 stain positive for calbindin.

Responses of layer 3 presubicular cells to photoactivation of ATN and RSC fibers.

A, Expression of AAV5-Chronos in ATN or RSC.

B, Biocytin labeled layer 3 cells (white) and GFP positive axons (green) from the ATN or RSC. Scale bar 10 µm.

C, Firing pattern of two layer 3 cells receiving inputs from the ATN (cerulean) and the RSC (purple). Insets show current commands.

D, Cluster analysis of physiological parameters for cells tested by stimulating ATN (n = 29 cells /14 mice) or RSC (n = 43 cells/20 mice) fibers.

E, Representative EPSCs evoked in layer 3 cells by light stimulation (blue bar) of ATN or RSC inputs. Right, proportion of cells receiving ATN or RSC inputs.

F, Average amplitudes of ATN and RSC induced synaptic currents. Mann-Whitney U test revealed no significant difference (p = 0.95).

G, Latency of EPSCs evoked by light-stimulation of ATN (n = 24 cells/12 mice) or RSC fibers (n = 27 cells/14 mice). Mann-Whitney U test revealed no significant difference (p = 0.13).

H, EPSPs induced in layer 3 cells (single traces) by stimulating ATN or RSC inputs in absence and presence of 1 µM TTX and 100 µM 4-AP.

I, EPSCs induced by stimulating ATN or RSC fibers in the absence and presence of 100 µM APV and APV + 10 µM NBQX. Holding potential +40 mV.

J, Voltage-clamp responses of layer 3 cells to 20 Hz train stimulations of ATN (left) and RSC (right) inputs. Insets show EPSCs in response to the first two stimuli.

K, oEPSC amplitudes for 10 ATN or RSC fiber stimuli at 20 Hz (ATN, n = 15 cells/8 mice, RSC, n = 15 cells/11 mice). Short middle line, mean; Min to max and quartiles.

L, Paired-pulse ratio (PPR) and 10/1 ratio (ratio between 10th and 1st event amplitudes) for ATN or RSC inputs. Wilcoxon matched-pairs signed rank test: ATN PPR vs 10/1 **** p < 0.0001, RSC PPR vs 10/1 *** p = 0.0001.

M, Current clamp traces showing action potentials and EPSPs evoked by 10 stimuli at 20 Hz. N, Spiking probability during trains of 10 stimuli. ATN, n = 6 cells/4 mice, RSC, n = 12 cells/9 mice. Full line, median; short line, mean; Min to max and quartiles. In K and N, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 from Friedman’s test followed by Dunn’s post-hoc test.

O, oEPSP amplitudes for trains of 5 stimuli at 20 Hz. See also Figure 2—figure supplements 1 and 2.

Table showing the intrinsic electrophysiological physiological properties of layer 3 cells tested for responses to stimulating ATN or RSC fibers.

Comparison of evoked synaptic events in presubicular layer 3 neurons following photostimulations of ATN or RSC axons.

A, Variance of latency (jitter) in oEPSP onset following Chronos-expressing ATN or RSC fiber stimuli. RSC input stimulation gives similar variation in latency as ATN inputs. Mann-Whitney U test revealed no significant difference (p = 0.26). B, Jitter as a function of latency for ATN and RSC stimuli shows a similar coefficient of variation. C, Decay time constants of EPSCs evoked in layer 3 neurons by stimulating ATN or RSC inputs. Mann-Whitney U test revealed no significant difference (p = 0.87). D, E, First oEPSC amplitude as a function of oEPSC half-width (D) and oEPSC rise time (E) for both inputs. We found no clear segregation of oEPSC shapes according to the origin of afferent fibers. F, oEPSPs in a presubicular layer 3 cell, induced by blue or red light stimulation of Chronos-expressing ATN or Chrimson-expressing RSC afferents, respectively. G-J, Comparison of oEPSPs evoked by stimulating ATN or RSC axons, recorded from a same postsynaptic layer 3 presubicular neuron. G, Mean amplitude. H, Maximum rising slope of oEPSPs. I, oEPSPs rise time. p < 0.05, Wilcoxon test. J, oEPSPs decay time constants. oEPSPs following ATN and RSC fiber stimulation recorded from the same cell are connected by lines. Dashed lines indicate reversed experimental design (expressing Chronos in RSC and Chrimson in ATN). K, Responses to 10 light stimuli at 20 Hz for ATN and RSC inputs to the same cell. L, Amplitudes of oEPSPs evoked by trains of ATN (left) or RSC (right) fiber stimuli for 11 cells. Dashed lines indicate the reversed experimental design (expressing Chronos in RSC and Chrimson in ATN). M, Summary data from L. On average, the amplitudes of oEPSPs evoked by ATN fiber stimulation were smaller for the fifth than for the first stimulus. oEPSPs evoked by RSC fiber stimulation were similar across five stimuli. Full line, median; Short line, mean; Min to max and quartiles. * p < 0.05, Friedman’s test and post-hoc Dunn’s test comparing each stimulus with the first.

ATN and RSC axons converge in dorsal presubiculum and contact single layer 3 pyramidal neurons.

A, Expression of the blue light sensitive opsin Chronos in ATN and the red light sensitive opsin Chrimson in RSC.

B, Axons from ATN (GFP labeled, green) and RSC (tdTomato labeled, red) overlap in layer 1 and 3 of dorsal presubiculum. Layer 2 possesses patches of axon-dense and axon-poor zones. RSC fibers avoid axon-dense microstructures formed by ATN fibers in upper layer 3. Scale bar, 200 µm.

C, Independent dual wavelength optogenetic stimulation of light sensitive afferent fibers in presubicular slices.

D, Two biocytin labeled PrS layer 3 pyramidal cells surrounded by ATN (green) and RSC (red) axons.

E, Patch clamp recording from a layer 3 neuron shows optical EPSCs following photostimulation of ATN axons (blue light) and RSC axons (red light).

F, 76% of layer 3 pyramidal neurons tested (n = 17 cells/4 mice) received both ATN and RSC input.

G, Distributions of putative synaptic contacts from ATN (green) and RSC (red) on the dendrites of three layer 3 neurons. Scale bar 50 µm. The boxed area on Cell 2 is shown in panel I.

H, Normalized number of green and red spots for 6 neurons as a function of the distance from soma. Paired values are indicated by dotted lines for the 3 cells in G.

I, Examples of ATN-labeled (left), RSC-labeled (middle) and both (right) synapses closely apposed to dendrites of a biocytin-filled layer 3 pyramidal cell. Scale bar 20 μm. Insets show representative high-magnification images. Scale bar 2 µm.

See also Figure 3—figure supplement 1.

Calibration of blue and red light stimulation of Chronos and Chrimson.

A, Blue light stimulation of ATN Chronos-expressing fibers evoked oEPSPs in presubicular layer 3 neurons at an intensity of 0.25mW, and 0.5 ms duration of illumination. B, Blue light stimulation at the same intensity and duration did not evoke oEPSPs from RSC Chrimson-expressing fibers. C, D, Response probabilities of presubicular layer 3 cells to stimulation of fibers expressing Chronos or Chrimson with red (C) or blue (D, left) light. E, Experimental design. Injection of AAV5-Chronos-GFP (green) in the ATN, followed by photostimulation combined with patch recording from a presubicular layer 3 cell. F, Amplitudes of oEPSCs evoked in layer 3 cells by blue (0.5 ms, 0.25 mW) or red light (2 ms, 0.6 mW) stimuli. G, Representative responses to light stimulation of ATN inputs. H, Injection of AAV5-Chrimson-tdTomato (red) in the RSC, followed by photostimulation combined with patch recording from a presubicular layer 3 cell. I, Amplitudes of oEPSCs evoked in layer 3 cells by blue (0.5 ms, 0.25 mW) or red light (2 ms, 0.6 mW) stimuli. J, Representative responses to light stimulation of RSC inputs K, Values and statistical analysis for different stimulation configurations (Panels F and I).

Supralinear summation of EPSPs and action potential firing following photostimulation of ATN and RSC axons.

A, Optical EPSPs in layer 3 neurons in response to blue light activation of ATN axons (cerulean traces), red light activation of RSC axons (purple traces). Supralinear summation of EPSPs following coincident activation of both ATN and RSC axons (black traces). The pink broken line indicates the calculated linear sum of oEPSPs evoked by stimulation of either ATN or RSC axons. Records from cell 1, Fig 3D.

B, Amplitudes of dual ATN and RSC oEPSPs plotted as a function of the sum of separate ATN and RSC stimulations (left). Each circle is a cell (n = 11 cells/4 mice), Pink line (± 10%) indicates linearity. Pie charts give the number of tested layer 3 neurons with supralinear (grey), linear (pink) or sublinear (white) summation. oEPSPs normalized to linear sum (bottom right). Solid circle, Chronos in ATN/Chrimson in RSC, empty circle, Chronos in RSC/Chrimson in ATN. * p < 0.05, ** p < 0.01, from Wilcoxon test.

C, As in B, for dual ATN and RSC oEPSP integrals.

D, oEPSPs induced by 20 Hz stimulation of ATN (cerulean), RSC (purple) or both (black). The pink broken line indicates the calculated linear sum of oEPSPs evoked by separate ATN and RSC photostimulation.

E, Amplitudes of dual oEPSPs compared to those of the sum of ATN and RSC oEPSPs show that supralinearity increases across 5 stimulations. Two-way ANOVA, p = 0.0117. * p < 0.05, **** p < 0.0001, Šidák’s multiple comparison test. # p < 0.05, ## p < 0.01, Friedman’s and post-hoc Dunn’s test.

F, As in E, for dual ATN and RSC oEPSP integrals. Two-way ANOVA, p = 0.0128.

G, Excitatory postsynaptic responses to photostimulation of ATN (blue light, cerulean traces) or RSC axons (red light, purple traces) or both (blue and red light, black traces) at resting membrane potential (-65 mV) or at a depolarized holding potential (-55 mV). Synaptic excitation led to action potentials when dual ATN and RSC stimuli reached firing threshold.

H, Action potential (AP) probability for either or both stimuli at -65 and -55 mV. Data are presented as mean ± SEM.

I, Action potentials were induced in presubicular layer 3 neurons by near coincident activation of ATN axons (red light) and RSC axons (blue light). Time delays varied from -50 to +50 ms.

J, Firing probability was maximal for short delays between -2 to +5 ms (RSC preceding ATN).

EPSP amplifications in layer 3 pyramidal neurons.

A, Optical EPSPs evoked in layer 3 pyramidal neurons following the stimulation of ATN axons (blue light, cerulean traces), RSC axons (red light, purple traces) or both (blue and red light, black traces). Recording pipette contained a cesium gluconate based internal solution and the Na+ channel blocker QX-314. A large all–or-none EPSP amplification occurred for dual stimuli at 20 Hz, on some trials.

B, In the presence of the GABAA receptor antagonist gabazine (10 µM), dual EPSP amplification occurred earlier in the train.

C, The additional presence of the NMDA receptor antagonist APV (100 µM) abolished dual EPSP amplification (black trace). EPSP amplification was partially restored by increasing red light intensity x2 (pale pink traces).

Presubicular LMN-projecting layer 4 neurons avoid thalamo-recipient layer 3 and receive little direct input from ATN and RSC.

A, Expression of Chronos-GFP in ATN and retrograde labeling of neurons that target LMN.

B, Thalamic axons (green) in superficial layers 1 and 3 of presubiculum. Retrobeads label cell bodies of presubicular layer 4 cells (red).

C, Retrograde rAAV2-tdTomato label cell bodies and dendrites of layer 4 LMN projecting neurons (red). Apical dendrites of layer 4 pyramidal neurons avoid layer 3 where thalamic axons ramify.

D, Presubicular slice containing two layer 3 and two layer 4 neurons filled with biocytin (white) and GFP expressing thalamic axons (green). Scale bar 100 µm. Inset, retrobeads (red) in the soma of a biocytin filled LMN-projecting layer 4 neuron. Scale bar 10 µm.

E, Layer 3 neurons are regular spiking and layer 4 neurons are burst firing, initially and at rebound, in reponse to current injection. Black trace, rheobase.

F, T-type Ca2+ channel blocker TTA-P2 (1 μM) suppressed burst firing in presubicular layer 4 neurons, while single action potentials were preserved.

G. Representative oEPSCs in layer 3 (left) and layer 4 (right) pyramidal cells, in response to stimulation of ATN (cerulean) or RSC (purple) inputs.

H, oEPSC latencies in layer 3 and layer 4 cells, for ATN inputs (left, cerulean), or RSC inputs (right, purple). Each dot is a cell. Same layer 4 cells are indicated by connecting lines, to show the difference in latency for direct and indirect synaptic responses.

I, oEPSPs in layer 4 neurons in response to stimulation of ATN (cerulean) or RSC (purple) inputs in control and in the presence of TTX (1 µM) and 4-AP (100 µM). Dashed lines indicate the timing of the large disynaptic component of the responses (i, indirect), and the small monosynaptic response (d, direct), isolated in TTX-4AP.

See also Figure 6—figure supplements 1 and 2.

Apical dendrites of presubicular layer 4 neurons avoid the thalamorecipient layer 3.

A, Retrograde AAV2retro-tdTom was injected in LMN. B, PrS L4 neuron cell bodies and dendrites were labeled. Scale bar, 50 µm. Full arrowheads, L4 neurons apical dendrites crossing L3; Empty arrowheads, L4 neurons apical dendrites skirting L3. C, Anterograde AAV5-Chronos-GFP was injected in ATN to label thalamic axons, delimiting PrS layer 3. D, Examples of biocytin filled layer 4 presubicular neurons. (1) The apical dendrite of this neuron avoided the densely packed ATN afferents and it didn’t show responses to thalamic stimulations. (2) This neuron’s apical dendrite crossed through a short stretch of layer 3 containing ATN afferents, and it responded with low amplitude indirect events. (3) The apical dendrite of this neuron crossed layer 3 and the zone containing ATN afferents, and was directly recruited. Scale bar 10µm. E, AAV-Chronos-GFP injected in ATN, and AAV-Chrimson-tdTom injected in RSC. F, Image from Figure 7 panel D. Scale bar 100µm. G, (Left) Enlarged view of dotted area in F, with two labelled L4 neurons. The apical dendrite of top neuron makes a U-turn to ramify in deep layers and seems to cross path with the axon of the bottom neuron, potentially forming a synaptic contact. (Right) The dendrite of the top neuron and axon of bottom neuron are overlayed with pink and green lines respectively. Scale bar 20µm.

Electrophysiological passive and active intrinsic properties of layer 3 vs. layer 4 neurons.

Data for layer 3 are summarized in Figure 2—figure supplement 1 and are shown here for comparison. Statistical differences are indicated as levels of p values (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001) obtained from Mann-Whitney tests between layer 4 (n = 19) and layer 3 (n = 33) for each parameter. Box plots indicate maximum and minimum values, quartiles, full line is the median, short line is the mean.

Cross-laminar activation of LMN-projecting layer 4 bursting neurons.

A, Simultaneous records of a layer 4 and a layer 3 neuron during photostimulation of ATN afferents. EPSP onset was delayed in the layer 4 neuron. Right panel, expanded view of the boxed area.

B, Simultaneous records of a layer 4 and a layer 3 neuron during photostimulation of RSC afferents. As in A, layer 4 neurons responded with a delay. Right panel, expanded view of the boxed area.

C, Latencies of synaptic activation indicated by the dotted lines in A, B (ATN stimulation, n = 4 cells; RSC stimulation, n = 1). EPSP onset and AP peak from layer 3 neurons (n = 5). Dotted lines link to the EPSP onset in simultaneously recorded layer 4 neurons (n = 5). p < 0.05, Kruskal-Wallis multiple comparison test.

D, Biocytin labeled layer 3 and layer 4 pyramidal neurons, in a presubicular slice containing thalamic (green) and retrosplenial (red) axons. The apical dendrite of one layer 4 neuron makes a U-turn (arrowhead), away from layer 3, where the thalamic axons ramify. The neighboring layer 4 neuron’s apical dendrite crosses the thalamo-recipient layer 3 for a short distance before arborizing outside of ATN targeted area, towards the subiculum, on the right. Another biocytin filled layer 4 neuron’s dendrites extend toward the deep layers. Scale bar, 100µm.

E, Simultaneous records from a layer 3 and a layer 4 cell to ATN input stimulation (top, 0.25 mW, blue light), ATN and RSC input stimulation (middle, 0.25 mW blue and 0.25 mW red light; light intensities compatible with independent photostimulation) or non-specific ATN and RSC input stimulation (bottom, 3.25 mW blue and 0.25 mW red light). ATN fibers expressed Chronos-GFP (green) and RSC fibers expressed Chrimson-tdTomato (red).

F, Top, oEPSPs and bursts of action potentials in a layer 4 neuron, evoked by dual wavelength stimulation of ATN and RSC afferents at 20 Hz. Amplifications of dual oEPSPs led to firing. Bottom, the NMDA receptor antagonist APV (100 µM) reduced EPSP amplification and prevented action potential firing.

G. Layer 4 bursting neurons are sensitive to the acetylcholine receptor agonist carbachol (10 µM). Action potential firing in response to step current injections in control (black), in the presence of carbachol (grey), and after wash-out (black). Bottom graph, membrane potential depolarization during a 2-minute carbachol application. The number of action potentials increased during the depolarizing steps (i) and on the baseline (ii).

See also Figure 7—figure supplement 1.

Regular firing layer 3 vs. intrinsically bursting layer 4 neurons responded to high-intensity light stimulations of ATN or RSC afferents.

Layer 3 neurons respond with single spikes following each light stimulation while layer 4 neurons respond with bursts of action potentials at the beginning of the stimulation train.

Schematic models for landmark anchoring of HD signals.

A, different scenarios with varying degrees of divergence and convergence of the two input sources. We note that matching directional with visual landmark information does not necessarily require synaptic plasticity, and might rely on temporal coincidence. B, subcellular arrangements of synaptic contacts are important. C, Inhibitory connectivity is important.