Comparative physiology and morphology of BLA-projecting NBM/SI cholinergic neurons in mouse and macaque

Abstract

Cholinergic projection neurons of the nucleus basalis and substantia innominata (NBM/SI) densely innervate the basolateral amygdala (BLA) and have been shown to contribute to the encoding of fundamental and life-threatening experiences. Given the vital importance of these circuits in the acquisition and retention of memories that are essential for survival in a changing environment, it is not surprising that the basic anatomical organization of the NBM/SI is well conserved across animal classes as diverse as teleost and mammal. What is not known is the extent to which the physiology and morphology of NBM/SI neurons have also been conserved. To address this issue, we made patch-clamp recordings from NBM/SI neurons in ex vivo slices of two widely divergent mammalian species, mouse and rhesus macaque, focusing our efforts on cholinergic neurons that project to the BLA. We then reconstructed most of these recorded neurons post hoc to characterize neuronal morphology. We found that rhesus macaque BLA-projecting cholinergic neurons were both more intrinsically excitable and less morphologically compact than their mouse homologs. Combining measurements of 18 physiological features and 13 morphological features, we illustrate the extent of the separation. Although macaque and mouse neurons both exhibited considerable within-group diversity and overlapped with each other on multiple individual metrics, a combined morpho-electric analysis demonstrates that they form two distinct neuronal classes. Given the shared purpose of the circuits in which these neurons participate, this finding raises questions about (and offers constraints on) how these distinct classes result in similar behavior.

Keywords: basal forebrain; cholinergic; morpho-electric physiology; mouse; nonhuman primate.

Fig. 1–1. Reconstruction of the microbead injection sites in mouse BLA for retrograde labeling of BLA-projecting neurons.

A. Representative slices showing major anatomical landmarks demarcating the BLA (green) and beads injection site (red). B. Beads target identification on slices in each animal. From 27 animals used for recordings, 16 animals were successfully scanned for site of beads injection, post recordings. BLA along the bregma is marked with Green. Each red dot represents the injection site from one animal.

Fig. 1–2. Approximate location of recorded, and relocalized BLA-projecting cholinergic and non-cholinergic neurons along bregma in mouse.

Schematic diagram of coronal views along bregma of NBM/SI regions (Left shown here) in which we found BLA projecting neurons. These cells comprise the sample for electrophysiological and morphological features. Approximate locations of all BLA-projecting, NBM/SI cholinergic neurons (n= 48) are shown in teal; the locations of BLA-projecting, non-cholinergic neurons (n= 46) are shown in grey.

Fig. 2–1. (related to Figs. 2, 4, 7 and 9) Physiological and morphological parameters used for morpho-electric profiling.

A. Schematic of an action potential with all electrophysiological parameters assayed at rheobase noted. Max firing rate (inset) was measured in response to a 200 pA x 500 mS pulse with the rate calculated as the spikes during the 500 ms pulse x 2, Hz. Adaption index was calculated as the number of spikes in the first vs the second half of the 500 ms pulse. CV ISI was assayed as the coefficient of variation of the interspike interval over the entire 500 mS x 200 pA. 10 pA steps from - 60 pA hyperpolarizing pulses to + 200 pA were applied. B. Schematic of phase plot showing the membrane potential (Vm) rate of change vs time (dV_m/dt) plotted vs membrane potential throughout the cycle of firing at rheobase. The basic features of a phase plot including upstroke, downstroke, etc are indicated. C. Schematic diagram of Sholl analysis and proximal dendritic parameters assayed following skeletonized representation of relocalized, neurobiotin stained neurons in Imaris. D. Illustration of convex hull analysis of proximal somatodendritic domain following skeletonized representation of relocalized, neurobiotin stained neurons in Imaris.

Fig. 2–2. Additional electrophysiological features that distinguish BLA-projecting, cholinergic from noncholinergic neurons and features that are shared in mouse NBM/SI.

A. Population scatter plus box plots of data for membrane properties of BLA-projecting, NBM/SI cholinergic neurons (n= 48) and their neighboring BLA-projecting, non-cholinergic neurons (n= 46; cholinergic: teal; non cholinergic: grey) B. Average time derivative vs time plots illustrate differences in action potential time course of BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey). Shading represents SEM. C. Population scatter plus box plots of data for the 3 membrane features that are shared between BLA-projecting, NBM/SI cholinergic neurons and their neighboring BLA-projecting, non-cholinergic neurons (cholinergic: teal; non-cholinergic: grey) C, cholinergic; NC, non-cholinergic

Fig. 3–1. Skeletonized renditions of relocalized BLA-projecting non-cholinergic neurons within NBM/SI along bregma in mouse.

Similar to cholinergic neurons, the proximal 100⁺ μm of the processes emanating from cholinergic somata were morphologically diverse regardless of location along bregma. All neurons were bi-polar or multi-polar.

Fig. 4–1. Most morphological features of somatodendritic domains are shared by neighboring NBM/SI neurons that project to the BLA for cholinergic and non-cholinergic in mouse.

A: Sholl intersections analysis as a function of sphere radius from cell soma. B: Ten additional aspects of proximal neuritic morphology in neighboring NBM/SI neurons that project to the BLA. Majority of cell shape and proximal neurite configuration parameters are shared amongst BLA-projecting cholinergic and non-cholinergic NBM/SI neurons in mouse, despite marked differences in electrophysiological properties. C, cholinergic (n = 31); NC, non-cholinergic (n = 44)

Fig. 6–1. Workflow for preparation of macaque ex vivo recording.

A. Beads injection Pre-operative structural MRIs were used to determine the stereotaxic injection coordinates for BLA fluorescently tagged microbeads injection. B. Tissue blocking About 4–6 weeks post op, a Br −4.0 to −8.0 section is surgically removed from ipsilateral macaque brain and is blocked into a small piece containing NBM/SI. C. Slicing and incubation The tissue block is glue to the vibratome platform and cut from medial to lateral. Slices are incubated at room temperature until being used for recording. D. Ex vivo slice recording Slices are examined under both DIC and fluorescent microscope for locating beads labeled neurons for recording.

Fig. 6–2. Reconstruction of the microsphere injection sites in macaque BLA for retrograde labeling of BLA-projecting neurons.

A: MRI guided BLA beads injection in two coronal plates separated by 1.5 – 2.0 mm; green lines on each image indicate the injection tracks. Three sites within each track were targeted along the dorsoventral axis, spaced 2 – 2.5 mm apart. 20 μL was injected at each site, for a total volume of 120 μL across 6 sites. B: Left, a representative bright field slice image showing major landmarks and the beads injection site. Right, same slice, stained with VAT antibody, shows high cholinergic terminal density and beads colocalization within BLA. C: Beads target identification on slices in each animal.

Fig.6–3. Approximate location of confirmed cholinergic BLA-projecting neurons that were subject to subsequent electrophysiological recording in rhesus macaque.

Schematic diagram of coronal views of NBM/SI regions. Approximate locations of all confirmed BLA-projecting, NBM/SI cholinergic (ChAT+) neurons (n= 52) are shown in purple; the locations of BLA-projecting, ChAT negative neurons (n= 11) are shown in grey. The confirmed ChAT+ re-localized neurons comprise the sample for the electrophysiological and morphological features described.

Fig.7–1. Additional electrophysiological features of BLA-projecting, cholinergic neurons in NBM/SI of rhesus macaque vs mouse.

A. Population scatter plus box plots of additional features that differ in the action potential time course of BLA-projecting, NBM/SI cholinergic neurons from macaque (n= 46; purple) vs. BLA-projecting, cholinergic neurons in mouse (n= 48; teal). Mouse data are the same as those in Figs. 2 and 2-2, presented here for ease of comparison. B. Time derivative vs time plots illustrate differences in action potential time course of ChAT + BLA-projecting, NBM/SI neurons from macaque (purple) compared with mouse (teal). Mouse data are the same as those in Fig. 2-2 for ease of comparison. C. Population scatter plus box plots showing features of AP width and Upstroke are distinct, but feature of downstroke is shared between BLA-projecting, NBM/SI cholinergic neurons from macaque (purple; n= 46) and mouse (teal; n= 48). Mouse data are the same as those in Figs. 2 and 2-2, presented here for ease of comparison. p value symbols used are ** ≤ 0.01; unlabeled comparisons are not statistically significantly different.

Fig. 9–1. Additional morphological parameters assessed in BLA projecting, NBM/SI, cholinergic neurons of rhesus macaque compared with those in mouse.

A. Sholl intersections analysis as a function of sphere radius from cell soma; B. Seven additional aspects of proximal somatodendritic morphology were assessed in NBM/SI, BLA projecting cholinergic, neurons from macaque (purple n=52) compared with those from mouse (n = 31, teal). Mouse data are from the same set of morphological parameters as presented in Figs. 4 and 4-1.

Fig. 9–2. Reconstructed images of relocalized BLA-projecting cholinergic neurons within NBM/SI outlined by fitted convex hulls along bregma in mouse.

Mouse neurons are smaller than macaque neurons in volume and occupy relatively small three-dimensional physical space.

Fig. 9–3. Reconstructed images of relocalized BLA-projecting cholinergic neurons within NBM/SI outlined by fitted convex hulls along bregma in macaque.

In comparison to mouse neurons, macaque neurons are larger in volume and occupy more three-dimensional physical space.

Fig. 1. Workflow for morpho-electric profiling of NBM/SI BLA-projecting cholinergic and non-cholinergic neurons in mouse.

See Extended Data Figs. 1-1 and 1-2. A. Surgical back-labeling Schematic for injection of fluorescently tagged microbeads into bilateral BLA of Chat-Tau-GFP mice (gift from S. Vijayaraghavan, Univ. Colorado, Grybko et al., 2011). Stereotactic delivery of 200~300 nl of red fluorescence tagged, microbeads (FluoSpheres^™ Carboxylate-Modified Microspheres, F8793, Invitrogen) at approximately −1.1 mm bregma (L/M, +/− 3.25 mm & D/V −4.15 mm from dura) of a 6-week-old ChAT-tau-GFP mouse to retrograde-label BLA projecting neurons. (Below: RHS) schematic of coronal section illustrating approximate stereotactic positioning of injection pipette and ipsilateral back labeling of cholinergic (red + green, yellow) and non-cholinergic (red) neurons within the NBM/SI region examined (~−0.7– 0.9 mm Bregma). Green symbols represent Chat-Tau-GFP cell bodies that were not beads labeled. B. Live imaging & recording About 1 week post op, mice are prepared for ex vivo, slice recording. (top left) Photo micrograph of sample injection site of red beads into the BLA of a ChAT-tau-GFP mouse. Yellow and green signal in BLA derives from the extensive (green fluorescent) cholinergic terminal fields within this area. (Top right) photomicrographs taken during the recording session in the region of the NBM/SI. Cells were first identified in DIC (top); ChAT-tau-GFP labeling was readily detected in live imaging of NBM/SI cholinergic neurons. BLA-projecting neurons were readily detectable in live imaging due to beads from the BLA injection. If the neuron was labeled with both ChAT GFP and the red beads it was classified as BLA projecting & cholinergic. BLA projecting, non-cholinergic neurons are defined as bead labeled, without ChAT-Tau-GFP label. All neurons were injected with Neurobiotin (SP-1120, Vector Laboratories) during electrophysiological recording. C. Relocalization Neurons were relocalized based on coordinates and Neurobiotin staining with streptavidin, ChAT and/or beads labeling. Confirmed images were imported into Imaris for morphological parameter assessment.

Fig. 2. Electrophysiological features consistent with the lower excitability of mouse cholinergic BLA-projecting NBM/SI neurons compared with neighboring non-cholinergic BLA-projecting NBM/SI neurons.

See Extended Data Figs. 2-1 and 2-2. A. Sample traces at rheobase (left) and at maximum current injection (right; 200 pA) are shown for typical BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey) B. Average phase plots illustrate differences in action potential kinetics comparing BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey) C. Non dimensional (UMAP) plot of all 18 electrophysiological features comparing BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey) D. Population scatter plus box plots of all data for the 12 features that most strongly distinguish BLA-projecting, NBM/SI cholinergic neurons (n = 48) from their neighboring BLA-projecting, non-cholinergic neurons (n = 46; cholinergic: teal; non cholinergic: grey) p value symbols used in this and all subsequent figures are * ≤0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001. C, cholinergic (n = 48); NC, non-cholinergic (n = 46)

Fig.3. Skeletonized renditions of relocalized BLA-projecting cholinergic neurons within NBM/SI along all bregma in mouse.

See Extended Data Fig. 3-1. The proximal 100⁺ μm of the processes emanating from cholinergic somata were morphologically diverse and independent of location along bregma. Most neurons were multi-polar although fairly simple in morphology (n = 31).

Fig.4. BLA-projecting, NBM/SI neurons in mouse, whether cholinergic or noncholinergic, differ in 3 of the 13 morphological parameters assessed.

See Extended Data Fig. 4-1. A. Confocal images of a representative BLA-projecting cholinergic neuron in mouse B. Confocal images of a representative BLA-projecting non-cholinergic neuron in mouse C. Non-dimensional (UMAP) plot of the morphological features of BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey) D. Population scatter plus box plots of data for the 3 morphological features that are differed significantly between BLA-projecting, NBM/SI cholinergic neurons (n = 31) and their neighboring BLA-projecting, non-cholinergic neurons (n = 44; cholinergic: teal; non cholinergic: grey) “C” stands for cholinergic, and “NC” stands for non-cholinergic.

Fig.5. Linear Discriminant Analysis (LDA) strongly distinguishes cholinergic from non-cholinergic BLA projecting NBM/SI mouse neurons based on electrophysiological, but not on morphological features.

A. Linear discriminant analysis was applied to all electrophysiological features for BLA-projecting, NBM/SI neurons (cholinergic: teal, n = 48; non cholinergic: grey, n = 46). There is clear separation in clustering of the two populations from one another but not from the distribution of shuffled data (shown in inset), consistent with the many electrophysiological features that distinguish between BLA-projecting cholinergic vs. non-cholinergic neurons in mouse. B. Plot of the absolute values of the weighted differences in electrophysiological features between BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey). C. Linear discriminant analysis was applied to all morphological features for BLA-projecting, NBM/SI neurons (cholinergic: teal, n = 31; non cholinergic: grey, n = 44). The two populations don’t separate from one another nor is the distribution of morphological features very different from the shuffled data (shown in inset). D. Plot of the absolute values of the weighted differences in morphological features between BLA-projecting, NBM/SI neurons (cholinergic: teal; non cholinergic: grey) of mouse.

Fig. 6. Workflow for morpho-electric profiling of BLA-projecting NBM/SI cholinergic neurons in macaque.

See Extended Data Figs. 6-1, 6-2 and 6-3. A. Surgical retrograde-labeling (top) Schematic for fluorescently tagged microbeads injection into the basolateral amygdala (BLA) for retrograde labeling of BLA projecting basal forebrain neurons in rhesus macaque. MRI guided and stereotactic delivery of 120 μl of fluorescence tagged microbeads into the region of the BLA (Bregma −6.75 to −7.75 mm at approx. L/M 7.5 mm x D/V 34.0 mm in either right or left hemisphere) of ~7 – 13-year-old rhesus macaques to back-label BLA projecting neurons (see supplementary for more details). (Below: RHS) Schematic of coronal section illustrating approximate track and positioning of bead injection needle. (Bottom: LHS) Schematic of coronal view to illustrate target of ipsilateral retrograde labeling within the NBM/SI region examined (~−5.4 mm Bregma). B. Live imaging & recording: About 4–6 weeks post op, a Br −4.0 to −8.0 mm section is surgically removed from side ipsilateral to the beads injection of the macaque brain and is prepared for ex vivo, slice recording (Top left and see Fig. 6-1B). Top right: Photomicrographs taken during the recording session in the region of the NBM SI, Cells were identified in DIC (top) and BLA-projecting neurons detected in live imaging by microbeads from the BLA injection. Bead positive neurons were injected with Neurobiotin tracer during electrophysiological recording. C. Relocalization: Neurons were relocalized based on coordinates and neurobiotin staining. ChAT labeling was confirmed by immunostaining and high-power images of the cell body and proximal dendrites were imported into Imaris for morphological parameter assessment.

Fig. 7. Both passive and active membrane properties of macaque BLA-projecting, NBM/SI cholinergic neurons are consistent with higher excitability than those of mouse.

See Extended Data Fig. 7-1. A. Sample traces from BLA projecting, cholinergic NBM/ SI macaque neurons at rheobase (left) and at maximum current injection (right; 200 pA) are shown (macaque: purple; mouse: teal) Mouse data (from same data base of BLA-projecting cholinergic neurons as in Fig. 2A) are shown here again for comparison purposes. B. Average phase plots illustrate differences in action potential kinetics between macaque, BLA-projecting, NBM/SI cholinergic neurons (purple) and mouse (teal); Mouse data are the same as those in Fig. 2B, presented here for comparison purpose. C. Non dimensional (UMAP) plot of all 18 electrophysiological features comparing BLA-projecting, NBM/SI cholinergic neurons from macaque (purple) with comparable samples from mouse (teal). Mouse data are the same as those in Fig. 2C, presented here for ease of comparison. D. Population scatter plus box plots of data for 12 features that distinguish BLA-projecting, NBM/SI cholinergic neurons of the rhesus macaque (n = 46; purple) from BLA-projecting, NBM/SI cholinergic neurons from mouse (n = 48; teal). Mouse data are the same as those in Fig. 2D presented here for ease of comparison.

Fig.8. Skeletonized renditions of relocalized BLA-projecting cholinergic neurons within NBM/SI along bregma in macaque.

The proximal 100⁺ μm of the processes emanating from cholinergic somata of macaque were morphologically diverse regardless of location along bregma. Most neurons were multi-polar, yet fairly simple (n = 52).

Fig. 9. BLA projecting, NBM/SI, cholinergic neurons in macaque differ from those in mouse in 6 of the 13 morphological parameters assessed.

See Extended Data Figs. 9-1, 9-2 and 9-3. A. (LHS) Approximate location in hemi-coronal diagram of macaque basal forebrain region at ~−6.75 Brega. (RHS) A volume rendering of proximal neuritic domain of a representative NBM/SI, BLA projecting, cholinergic neuron from macaque B. Approximate location in hemi coronal section of mouse basal forebrain and volume rendering of proximal neurite domain of a representative NBM/SI, BLA projecting, cholinergic neuron from mouse C. Non dimensional (UMAP) plot of all 13 morphological features of BLA-projecting, NBM/SI, cholinergic neurons from macaque (purple, n = 52) vs mouse (teal, n = 31). Mouse data are from the same set of morphological parameters as presented in Figs. 4 and 4-1. D. Population scatter plus box plots of data for the 6 set of morphological properties that differ between BLA-projecting, NBM/SI cholinergic macaque (n = 52; purple) compared with mouse (n = 31; teal) neurons. Mouse data are from the same set of morphological parameters as presented in Figs. 4 and 4-1.

Fig. 10. Examination of combined morpho-electric features distinguish macaque vs mouse basal forebrain neurons despite common projection target, anatomical location, and cholinergic phenotype.

A. Non dimensional (UMAP) plot of all 18 electrophysiological and 13 morphological features of BLA-projecting, NBM/SI, cholinergic neurons from macaque (purple, n = 46) vs mouse (teal, n = 27). B. Linear discriminant analysis was applied to all 18 electrophysiological and 13 morphological features of BLA-projecting, NBM/SI, cholinergic neurons from macaque (purple, n = 46) vs mouse (teal, n = 27). Note the clear separation in clustering of the two populations from one another and its distinct nature from the distribution of shuffled data (shown in inset). C. Plot of the absolute values of the weighted differences for LDA in all 18 electrophysiological and 13 morphological features of BLA-projecting, NBM/SI, cholinergic neurons from macaque (purple, n = 46) vs mouse (teal, n = 27).