FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313305230> ?p ?o ?g. }

• W4313305230 abstract "Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Connections from the basolateral amygdala (BLA) to medial prefrontal cortex (PFC) regulate memory and emotion and become disrupted in neuropsychiatric disorders. The diverse roles attributed to interactions between the BLA and PFC may reflect multiple circuits nested within a wider network. To examine these circuits, we first used retrograde and anterograde anatomy to show that the rostral BLA (rBLA) and caudal BLA (cBLA) differentially project to prelimbic (PL) and infralimbic (IL) subregions of the mouse PFC. Using ex vivo whole-cell recordings and optogenetics, we then assessed which neuronal subtypes are targeted, showing that rBLA preferentially drives layer 2 (L2) cortico-amygdalar (CA) neurons in PL, whereas cBLA drives layer 5 (L5) pyramidal tract (PT) neurons in IL. We next combined in vivo silicon probe recordings and optogenetics to confirm that cBLA mainly influences IL L5, whereas rBLA primarily activates PL L2, but also evokes polysynaptic activity in PL L5. Lastly, we used soma-tagged optogenetics to explore the local circuits linking superficial and deep layers of PL, showing how rBLA can engage L2 CA neurons to impact L5 PT neuron activity. Together, our findings delineate how subregions of the BLA target distinct networks within the PFC and differentially influence output from PL and IL. Editor's evaluation This paper will be of interest to readers studying the neuronal circuit in general and those studying the basolateral amygdala (BLA), prefrontal cortex (PFC), and diseases associated with these regions. Using innovative circuit analysis techniques, this important paper shows that different subregions of the BLA form cell-type-specific connections with the subregion of PFC and engage specific local circuits within it. The key claim of the paper is supported by compelling data from multiple approaches. https://doi.org/10.7554/eLife.82688.sa0 Decision letter eLife's review process Introduction Connections from the basolateral amygdala (BLA) to the medial prefrontal cortex (PFC) regulate memory and emotion (Phelps and LeDoux, 2005; Etkin et al., 2011). Dysfunction of these connections contributes to neuropsychiatric disorders, including post-traumatic stress disorder, depression, and autism (Gilboa et al., 2004; Felix-Ortiz et al., 2016; McTeague et al., 2020). The BLA sends strong excitatory inputs to the PFC, which contact specific cell types to drive local activity and output to downstream brain areas (Sotres-Bayon et al., 2012; Little and Carter, 2013; Cheriyan et al., 2016; Burgos-Robles et al., 2017). However, previous studies have treated the BLA and PFC as monolithic entities, whereas each structure contains multiple subregions that may have distinct connections and functional roles (McDonald, 1991; Hintiryan et al., 2021). The PFC is subdivided along the dorsoventral axis into the prelimbic (PL) and infralimbic (IL) regions (Van De Werd et al., 2010; Van De Werd and Uylings, 2014), which play different functional roles in behavioral paradigms like threat conditioning (Sierra-Mercado et al., 2011). Similarly, the BLA is subdivided along the rostro-caudal axis into anterior (magnocellular) and posterior (parvocellular) regions (McDonald, 2003; O’Leary et al., 2020), which also play distinct roles in aversive and appetitive behaviors (Kim et al., 2017; Zhang et al., 2021). Anatomical studies using transgenic mice suggest one population of cells in anterior BLA targets PL, while a separate group in posterior BLA targets IL (Kim et al., 2016). However, recent work suggests a more complicated relationship, with multiple projections from cells distributed along the rostral to caudal axis of the BLA (Hintiryan et al., 2021), which have an unknown influence on activity and output of the PFC. In addition to contacting different subregions and layers, long-range inputs often make targeted connections onto specific cell types in the PFC (Anastasiades and Carter, 2021). Recent studies suggest that glutamatergic projections from the BLA may differentially activate specific networks in PL compared to IL. BLA inputs to PL target layer 2 (L2) cortico-amygdalar (CA) neurons, which in turn project back to the BLA, forming a direct, reciprocal circuit under the control of local inhibition (Little and Carter, 2013; McGarry and Carter, 2016). In contrast, BLA connections in IL can activate layer 5 (L5) pyramidal tract (PT) neurons, which project to the periaqueductal gray (PAG) and other brain regions to exert top-down control of behavior (Cheriyan et al., 2016). We hypothesize that these opposing findings reflect distinct connections from rostral BLA (rBLA) and caudal BLA (cBLA) onto specific cell types within PL and IL, reflecting two parallel networks that link these brain regions. In addition to their direct connections, long-range inputs can evoke polysynaptic excitation and inhibition in the PFC (Collins et al., 2018; Huang et al., 2019). Connections within and across layers are understudied in PFC, but well explored in other cortices that lack layer 4 (L4) (Hooks et al., 2011). In motor cortex, layer 2/3 (L2/3) neurons project to deeper layers and target L5 pyramidal neurons (Anderson et al., 2010; Hirai et al., 2012), whereas L5 PT neurons predominantly target nearby L5 PT neurons (Morishima and Kawaguchi, 2006). Similarly, inhibition mediated by local interneurons is transmitted within and across layers (Kätzel et al., 2011; Saffari et al., 2016; Tremblay et al., 2016). By activating specific cell types in different layers, the BLA can evoke complex polysynaptic activity, but it is unknown how rBLA and cBLA inputs engage these circuits. Here, we examine how the rBLA and cBLA target distinct subregions, layers, and cell types in the mouse PL and IL. We first use anatomy to show that rBLA primarily projects to PL, and cBLA mostly projects to IL, indicating distinct connections. We next use ex vivo recordings and optogenetics to assess cell type-specific connections, showing that rBLA preferentially activates PL L2 CA neurons and cBLA engages IL L5 PT neurons. We then combine in vivo recordings with optogenetics and find that while cBLA influences are restricted to IL L5, rBLA drives both PL L2 and evokes robust polysynaptic activity in PL L5. Lastly, using soma-tagged optogenetics, we establish that this local influence reflects connections from PL L2 CA neurons onto PL L5 PT neurons within PL. Together, our findings illustrate how rBLA and cBLA evoke activity in PL and IL, demonstrating how subregions of the BLA engage distinct cortical networks in the PFC. Results Distinct projections of rostral and caudal BLA to PL and IL PFC To confirm if subregions of the PFC receive distinct and spatially segregated inputs from the BLA, we first injected retrograde viruses (AAVrg-tdTomato and AAVrg-GFP) into both PL and IL of the same animals, at offset anterior–posterior coordinates to minimize potential overlap (Figure 1A). We observed largely separate populations of PL and IL-projecting neurons across the BLA, with a smaller population of dual-projection neurons (PL = 583 ± 59 cells, IL = 694 ± 78 cells, dual = 246 ± 30 cells, n = 3 animals) (Figure 1B). Labeled cells were distributed along a gradient in the rostro-caudal axis, with PL-projecting neurons biased toward the rostral BLA (rBLA), and IL-projecting neurons shifted toward the cBLA (Figure 1C). We also found a similar gradient in experiments in which we injected AAVrg-tdTomato into PL or IL at the same anterior–posterior coordinate (Figure 1—figure supplement 1). These results agree with studies using transgenic lines (Kim et al., 2016), and suggest PL and IL receive inputs from neurons residing in rBLA and cBLA, respectively. Figure 1 with 2 supplements see all Download asset Open asset Rostral basolateral amygdala (rBLA) and caudal basolateral amygdala (cBLA) project to different subregions and layers of the prefrontal cortex (PFC). (A) Schematic for injections of AAVrg-tdTomato into prelimbic (PL) (magenta) and AAVrg-GFP into infralimbic (IL) (cyan). (B) Left, Images of injection sites in PL (+2.4 mm from bregma, magenta) and IL (+2.0 mm from bregma, cyan). Scale bar = 200 µm. Right, Images of retrogradely labeled PL-projecting (magenta) and IL-projecting neurons (cyan) across the AP axis of basolateral amygdala (BLA) (distances relative to bregma). Solid white lines mark white matter boundaries of the BLA. Scale bar = 200 µm. 4′,6-diamidino-2-phenylindole (DAPI) staining in gray. (C) Left, Average number of PL-projecting, IL-projecting, or dual-projecting cells along AP axis of BLA. Right, Percentage of PL-projecting, IL-projecting, and dual-projecting cells in the BLA (n = 3 animals). (D) Schematic for injections of AAV-YFP into rBLA (−1.1 mm from bregma; magenta) and AAV-mCherry into cBLA (−1.7 mm from bregma; cyan). (E) Left, Images of injection sites in rBLA (magenta) and cBLA (cyan). Scale bar = 250 µm. DAPI staining in gray. Right, Images of PL and IL (+2.2 mm from bregma), showing anterogradely labeled rBLA (magenta) and cBLA (cyan) axonal projections. Scale bar = 200 µm. (F) Top, Summary of fluorescence intensity from pia to white matter of rBLA (magenta) and cBLA (cyan) projections to PL, normalized to maximal fluorescence in each slice. Bottom, Similar for rBLA and cBLA projections to IL (n = 3 animals). See also Figure 1—figure supplements 1 and 2. To directly compare rBLA and cBLA projections to PFC, we next injected anterograde viruses (AAV-ChR2-eYFP and AAV-ChR2-mCherry) into rBLA and cBLA (−1.1 and −1.7 mm from bregma) in the same animals (n = 3) (Figure 1D). Care was taken to avoid injecting in the ventral hippocampus, which abuts the cBLA and also projects to the PFC (Liu and Carter, 2018; Liu et al., 2020). Consistent with our retrograde labeling, we observed rBLA axons primarily in PL and cBLA axons in IL, with distinct laminar targeting in each subregion (Figure 1E). Quantifying axon fluorescence across layers and subregions established that rBLA primarily targeted PL L2, whereas cBLA targeted IL L5 (Figure 1F). Together, these data indicate that rostral and caudal BLA form distinct anatomical connections across layers and the dorsal–ventral axis of the PFC. Lastly, to establish the presence of separate rBLA to PL and cBLA to IL projections, we next injected a combination of retrograde AAVrg-Cre into either PL or IL, followed by anterograde AAV-DIO-YFP into rBLA or cBLA, respectively, and quantified axonal fluorescence across layers and subregions of PFC (Figure 1—figure supplement 2). We again found that rBLA axons were prominent in PL L2 while cBLA axons were strongest in IL L5, consistent with our other anatomy and establishing that PL-projecting neurons in rBLA and IL-projecting neurons in cBLA do not collateralize within PFC. Rostral and caudal BLA target different cell types in PL and IL PFC Our anatomy showed that rBLA and cBLA project to different subregions and layers of PFC, but did not reveal which cell types are targeted within them. Previous work has shown that BLA inputs can selectively target L2 CA neurons in PL or L5 PT neurons in IL (Little and Carter, 2013; Cheriyan et al., 2016; McGarry and Carter, 2016). We next used whole-cell recordings and optogenetics to test whether rBLA and cBLA are responsible for engaging these different cell types in the PFC. We injected ChR2-expressing virus (AAV-ChR2-EYFP) into either rBLA or cBLA to visualize and stimulate axons in the PFC (Figure 2A; Little and Carter, 2013; McGarry and Carter, 2016). In the same mice, we also coinjected retrogradely transported, fluorescently tagged cholera toxin subunit B (CTB) into either rBLA or cBLA and the PAG to label CA and PT neurons, respectively (Figure 2A; Ferreira et al., 2015; Collins et al., 2018; Liu and Carter, 2018). After waiting for expression and transport, we observed CA neuron labeling in L2 and PT neuron labeling in L5 across PL and IL (Figure 2A, Figure 2—figure supplement 1). We then used wide-field illumination to activate rBLA or cBLA inputs, and measured synaptic responses at CA and PT neurons in TTX (1 µM), 4-AP (100 µM) and high extracellular Ca2+ (4 mM) to isolate monosynaptic connections (Petreanu et al., 2009; Mao et al., 2011). To control for variability in virus expression between slices and animals, we compared the ratio of responses referenced to a specific cell type (either PL L2 CA for rBLA or IL L5 PT for cBLA) (Figure 2B). Figure 2 with 1 supplement see all Download asset Open asset Rostral basolateral amygdala (rBLA) targets prelimbic (PL) L2 cortico-amygdalar (CA) neurons and caudal basolateral amygdala (cBLA) targets infralimbic (IL) L5 pyramidal tract (PT) neurons. (A) Left, Schematic for injections of AAV-ChR2-eYFP and CTB-647 (gray) into rBLA or cBLA and CTB-555 (red) into periaqueductal gray (PAG). Right, Example image of retrogradely labeled CA neurons (gray) and PT neurons (red) in prefrontal cortex (PFC). Scale bar = 200 µm. (B) Schematic of color-coded PL L2 CA, IL L2 CA, PL L5 PT, and IL L5 PT neurons recorded in the presence of TTX and 4-AP to isolate monosynaptic connections from either rBLA or cBLA. (C) Left, Average rBLA-evoked excitatory postsynaptic currents (EPSCs) recorded at L2 CA and L5 PT neurons in PL and IL. Magenta arrow = light stimulation. Right, Summary of EPSC amplitude ratios, comparing responses at PL L2 CA neurons with paired IL L2 CA (n = 8), IL L5 PT (n = 8), and PL L5 PT (n = 8) neurons (n = 5 animals). (D) Similar to (C) for cBLA-evoked EPSCs, comparing responses at IL L5 PT neurons with paired IL L2 CA (n = 7), PL L2 CA (n = 7), and PL L5 PT (n = 8) neurons (n = 6 animals). (E) Left, Schematic for sCRACM experiments, showing pseudorandom illumination (blue light) of 75 × 75 µm squares across L1 to L6 of PFC, while recording either rBLA-evoked EPSCs at PL L2 CA neurons or cBLA-evoked EPSCs and IL L5 PT neurons in the presence of TTX and 4-AP. Right, Example cBLA-evoked EPSCs recorded at a IL L5 PT neuron, including with stimulation at apical tuft, oblique, somatic, and basal dendrite locations. (F) Left, Average rBLA-evoked EPSC maps for PL L2 CA neurons (n = 4 cells, 2 animals), color coded by amplitude. Right, Average cBLA-evoked EPSC maps for IL L5 PT neurons (n = 6 cells, 4 animals). Triangle denotes typical cell body location. *p < 0.05. See also Figure 2—figure supplement 1. Taking this approach, we observed that rBLA-evoked excitatory postsynaptic currents (EPSCs) at −60 mV were much larger at PL L2 CA neurons compared to all other cell types in the different layers and subregions (EPSC ratio: IL L2 CA vs. PL L2 CA = 0.03, geometric standard deviation factor [GSD] = 12.3, p = 0.0003; n = 8 pairs; PL L5 PT vs. PL L2 CA = 0.02, GSD = 2.6, p = 0.0001; n = 8 pairs; IL L5 PT vs. PL L2 CA = 0.09, GSD = 7.7, p = 0.006; n = 8 pairs; 5 animals) (Figure 2C, Figure 2—figure supplement 1). In contrast, cBLA-evoked EPSCs were much larger at IL L5 PT neurons compared to all other cell types (EPSC ratio: PL L2 CA vs. IL L5 PT = 0.06, GSD = 9.3, p = 0.005; n = 7 pairs; IL L2 CA vs. IL L5 PT = 0.11, GSD = 10.5, p = 0.014; n = 7 pairs; PL L5 PT vs. IL L5 PT = 0.008, GSD = 3.1, p < 0.0001; n = 8 pairs; 6 animals) (Figure 2D, Figure 2—figure supplement 1). In separate experiments, we also confirmed that rBLA inputs target PL L2 CA neurons over L2 cortico-cortical (CC) neurons and IL L5 CA neurons (Figure 2—figure supplement 1) and that cBLA inputs target IL L5 PT over IL L5 CA neurons (Figure 2—figure supplement 1), consistent with previous results on non-specific BLA inputs. These findings show that subregions of BLA preferentially target different layers and cell types in the PFC, and further indicate that rBLA primarily synapses onto PL L2 CA neurons, and cBLA primarily contacts IL L5 PT neurons. Other long-range excitatory inputs to the PFC can evoke unique responses by selectively targeting the apical dendrites of pyramidal neurons (Anastasiades et al., 2021). Because our anatomy showed prominent BLA axons in superficial layers, we also tested the possibility that BLA inputs contact the apical dendrites of L5 PT neurons using subcellular Channelrhodopsin-Assisted Circuit Mapping (sCRACM) in the presence of TTX and 4-AP (Petreanu et al., 2009), stimulating across a grid aligned to the pia and soma (Anastasiades et al., 2021; Figure 2E). We found strong rBLA and cBLA input to the basal and perisomatic dendrites of PL L2 CA and IL L5 PT neurons, respectively, and no rBLA input to either the apical or basal dendrites of PL L5 PT neurons (Figure 2F, Figure 2—figure supplement 1). Together, these findings indicate that rBLA and cBLA primarily make connections close to the soma of pyramidal neurons in the PFC, and do not target the apical dendrites of deep layer neurons. Rostral and caudal BLA activate different networks in PL and IL PFC Having established which cell types are targeted by rBLA and cBLA, we next assessed how they are engaged during more physiological conditions and activity patterns. We stimulated rBLA or cBLA inputs with the repetitive trains (5 pulses of 473 nm light for 2 ms at 20 Hz) and recorded EPSCs at −60 mV (EGABA) and inhibitory postsynaptic currents (IPSCs) at +15 mV (EAMPA) in the presence of CPP (10 µM) to block N-methyl-D-aspartate (NMDA) receptors. For rBLA inputs, we recorded pairs of PL L2 CA neurons and IL L5 PT neurons (Figure 3A). We found that rBLA-evoked EPSCs were again largest at PL L2 CA neurons (EPSC1: PL L2 CA = −401 ± 63 pA; IL L5 PT = −133 ± 55 pA; PL L2 CA vs. IL L5 PT, p = 0.0078, n = 8 pairs, 5 animals) (Figure 3B). In contrast, rBLA-evoked IPSCs were of comparable amplitude at PL L2 CA and IL L5 PT neurons (IPSC1: PL L2 CA = 450 ± 86 pA; IL L5 PT = 521 ± 136 pA; PL L2 CA vs. IL L5 PT, p = 0.8125, n = 7 pairs, 5 animals) (Figure 3B). We found differences in excitation/inhibition (E/I) ratios, with rBLA inputs onto PL L2 CA neurons showing higher values than IL L5 PT neurons (rBLA E/I ratio: PL L2 CA = 0.99, GSD = 1.9; IL L5 PT = 0.15, GSD = 2.7; PL L2 CA vs. IL L5 PT, p = 0.0156, n = 7 pairs, 5 animals) (Figure 3B). For cBLA inputs, we recorded pairs of IL L5 PT neurons and PL L2 CA neurons (Figure 3C). We found that both EPSCs and IPSCs were strongly biased onto IL L5 PT neurons (EPSC1: PL L2 CA = −31 ± 10 pA; IL L5 PT = −415 ± 139 pA; PL L2 CA vs. IL L5 PT, p = 0.0156, n = 7 pairs, 4 animals) (IPSC1: PL L2 CA = 31 ± 28 pA; IL L5 PT = 858 ± 149 pA; PL L2 CA vs. IL L5 PT, p = 0.0156, n = 7 pairs, 4 animals) (cBLA E/I ratio; IL L5 PT = 0.4, GSD = 2.3; n = 7 pairs, 4 animals) (Figure 3D). In a subset of recordings, we also recorded from PL L5 PT and IL L2 CA neurons in addition to PL L2 CA or IL L5 PT neurons and found similar targeting biases from rBLA and cBLA, with the exception of prominent inhibition on PL L5 PT neurons evoked by rBLA stimulation (Figure 3—figure supplement 1). These results show that rBLA and cBLA directly excite their targets and evoke feed-forward inhibition that could prevent firing of action potentials (APs). Figure 3 with 1 supplement see all Download asset Open asset Rostral basolateral amygdala (rBLA) and caudal basolateral amygdala (cBLA) activate distinct projection neurons in prelimbic (PL) and infralimbic (IL). (A) Average rBLA-evoked excitatory postsynaptic currents (EPSCs) (bottom, recorded at EGABA) and IPSCs (top, recorded at EAMPA) at PL L2 cortico-amygdalar (CA) (left, black) and IL L5 pyramidal tract (PT) neurons (right, blue) (n = 8 pairs, 5 animals). Diamonds denote light pulses. (B) Left, Summary of rBLA-evoked EPSC1 amplitudes for PL L2 CA and IL L5 PT neurons. Middle, Similar for IPSC1 amplitudes. Right, Similar for EPSC1/IPSC1 ratios. Gray lines denote individual pairs of neurons. (C, D) Similar to (A, B) for cBLA inputs onto PL L2 CA and IL L5 PT neurons, with E/I ratios only for IL L5 PT due to the lack of inhibition onto PL L2 CA (n = 7 pairs, 4 animals). (E) Left, Example rBLA-evoked EPSPs and APs from pairs of PL L2 CA and IL L5 PT neurons recorded in current-clamp at resting membrane potential (RMP). Right, Summary of rBLA-evoked AP probability for PL L2 CA and IL L5 PT neurons. Gray lines denote individual pairs of neurons (n = 8 pairs, 5 animals). (F) Similar to (E) for cBLA-evoked EPSPs and APs from pairs of PL L2 CA and IL L5 PT neurons (n = 7 pairs, 3 animals). Note that all gray lines overlap with the black line. *p < 0.05. See also Figure 3—figure supplement 1. We then tested how rBLA and cBLA inputs drive cell type-specific firing of the main recipient cell types in PL and IL. Projection neurons in the PFC have distinct intrinsic properties (Ferreira et al., 2015), with CA neurons resting at very hyperpolarized potentials (McGarry and Carter, 2016), and PT neurons displaying robust h-current (Dembrow et al., 2010; Anastasiades et al., 2018). Taking these properties into account, we recorded responses in current-clamp at resting membrane potential (RMP). We found that rBLA inputs evoked significantly more action potential firing of PL L2 CA neurons compared to IL L5 PT neurons in the same slices (firing probability: PL L2 CA = 0.88 ± 0.05, IL L5 PT = 0.23 ± 0.15, p = 0.016; n = 8 pairs, 5 animals) (Figure 3E). In contrast, cBLA inputs activated only IL L5 PT neurons and never activated PL L2 CA neurons (firing probability: PL L2 CA = 0, IL L5 PT = 0.98 ± 0.02, p = 0.03; n = 7 pairs, 3 animals) (Figure 3F). Together, these findings indicate rBLA and cBLA engage distinct projection neuron networks in the PFC, supporting the hypothesis that BLA consists of functionally distinct rostral and caudal divisions. Rostral and caudal BLA evoke distinct activity in PL and IL PFC To test these predictions, and assess differences in the impact of rBLA and cBLA inputs on the PFC, we next combined in vivo optogenetics and Neuropixels (NP) recordings (Jun et al., 2017). We first injected AAV-ChR2-EYFP into either rBLA or cBLA and implanted an optical fiber above the injection site (Figure 4A). After waiting for expression, we acutely recorded light-evoked activity in the PFC of awake, head-fixed mice. The thin dimensions and high density of NP probes allowed us to target L2 or L5 of PL and IL in the same animals (Figure 4B). We confirmed the locations of each NP probe by registering to the Allen common coordinate framework (CCF) (Wang et al., 2020; Figure 4—figure supplement 1). Our stimulus protocol (5 × 2 ms pulses @ 20 Hz) was similar to previous studies on BLA to PFC connections (Floresco and Tse, 2007; Felix-Ortiz et al., 2016). Light-evoked activity was recorded across the dorsal–ventral axis of the PFC and analyzed from NP probes located in either L2 (n = 9 insertions from 6 mice) or L5 (n = 8 insertions from 6 mice) of PL and IL (Figure 4C). Figure 4 with 2 supplements see all Download asset Open asset Rostral basolateral amygdala (rBLA) and caudal basolateral amygdala (cBLA) stimulation evoke distinct responses in prelimbic (PL) and infralimbic (IL). (A) Schematic of AAV-ChR2-eYFP injections and fiber optic placement into either rBLA (magenta) or cBLA (cyan), along with Neuropixels (NP) recordings in prefrontal cortex (PFC). (B) Left, Schematic of NP probe insertions into L2 and L5 of IL and PL. Right, Example of NP probe tracts from sequential insertions into L2 and L5 of PFC (red = DiI), along with cBLA axons (cyan = eYFP). DAPI staining is shown in gray. Scale bar = 500 µm. (C) Left, Example of rBLA stimulation and NP recording in L2 of PFC. Graph shows the average change in spiking activity (z-scored) evoked by rBLA stimulation across each channel of the NP probe. Arrows denote timing of light-emitting diode (LED) stimulation. Right, Similar for cBLA stimulation and NP recording in L5 of PFC. (D) Left, Example single unit with significant basolateral amygdala (BLA)-evoked activation, with average response (top) and raster plot of peri-stimulus activity for 40 trials (bottom). Right, Similar for single unit with significant BLA-evoked suppression. (E) Percentage of units responsive to rBLA (left) or cBLA (right) inputs across different layers and subregions of the PFC. (F) Similar to (E) for units activated (left) or suppressed (right) by rBLA inputs. (G) Similar to (E) for units activated (left) or suppressed (right) by cBLA inputs. *p < 0.05. See also Figure 4—figure supplements 1 and 2. To characterize BLA-evoked activity, we used semi-automated spike-sorting to distinguish single units (rBLA-evoked recordings = 979 single units/1897 total; cBLA-evoked recordings = 596 single units/1152 total). We found that rBLA and cBLA evoked a variety of responses across the PFC, leading to significantly activated or suppressed units (Figure 4D). rBLA stimulation led to more responsive units in PL L2, PL L5, and IL L5 compared to IL L2 (PL L2 = 20.1%; IL L2 = 9.8%; PL L5 = 20.0%; IL L5 = 23.9% of single units in subregion), whereas cBLA stimulation mainly drove responses in IL L5 (PL L2 = 2.6%; IL L2 = 7.5%; PL L5 = 12.9%; IL L5 = 26.6% of single units in subregion) (Figure 4E). The highest proportion of rBLA-activated units were found in PL L2 (PL L2 = 18.2%; IL L2 = 5.5%; PL L5 = 7.9%; IL L5 = 9.9% of single units in subregion), whereas rBLA-suppressed units were enriched in L5 of PL and IL (PL L2 = 1.9%; IL L2 = 4.3%; PL L5 = 12.4%; IL L5 = 14% of single units in subregion) (Figure 4F, Figure 4—figure supplement 2). In contrast, cBLA-activated units were greatest in IL L5 (PL L2 = 1.3%; IL L2 = 5%; PL L5 = 4%; IL L5 = 17.7% of single units in subregion), with no significant differences in the proportion of cBLA-suppressed units (PL L2 = 1.3%; IL L2 = 2.5%; PL L5 = 8.9%; IL L5 = 8.9% of single units in subregion) (Figure 4G, Figure 4—figure supplement 2). We observed no significant differences in latency to spike after each light-emitting diode (LED) pulse in activated units (Figure 4—figure supplement 2). Together, these findings indicate that rBLA and cBLA drive distinct activity in PL and IL, with rBLA primarily activating PL L2, but also unexpectedly generating mixed activation and suppression across L5, and cBLA predominantly activating IL L5, with few responses elsewhere. Rostral BLA evokes polysynaptic excitation and inhibition of PL L5 PT neurons While our in vivo recordings indicate that rBLA inputs primarily target PL L2, we also observed a substantial number of activated and suppressed units in PL L5 and IL L5. Previous studies also found that BLA can influence the firing of PL L5 PT neurons to shape PFC output (Huang et al., 2019). Because we found no monosynaptic connections onto PL L5 PT neurons, we hypothesized a role for local connections. To test this idea, we first combined current-clamp recordings of PL L2 CA neurons with voltage-clamp recordings of PL L5 PT neurons (Figure 5A). We illuminated rBLA inputs to PL L2 with trains that evoked subthreshold or suprathreshold activity at PL L2 CA neurons. In the same slices, we recorded rBLA-evoked EPSCs at EGABA and IPSCs at EAMPA from PL L5 PT neurons. We found no responses at subthreshold intensities, but robust responses at suprathreshold intensities, suggesting polysynaptic activity from L2 to L5 is involved in rBLA-evoked responses at PL L5 PT neurons (EPSC: sub = −9 ± 4 pA, supra = −113 ± 24 pA, p = 0.02; n = 8 pairs; IPSC: sub = 17 ± 5 pA, supra = 508 ± 101 pA, p = 0.008; n = 8 pairs; 4 animals) (Figure 5B, C). In contrast, we observed no responses in IL L5 PT neurons at either subthreshold or suprathreshold intensities, suggesting activity does not spread from PL L2 to IL L5 in the slice, and that rBLA instead projects directly to IL L5 (EPSC: sub = −4 ± 3 pA, supra = −7 ± 4 pA, p = 0.055; n = 8 pairs; IPSC: sub = −7 ± 2 pA, supra = −11 ± 3 pA, p = 0.64; n = 8 pairs; 3 animals) (Figure 5D–F). In order to confirm that cBLA does not engage PL L5 via IL L5, we also activated cBLA inputs in IL L5, and observed no EPSCs in PL L5 PT neurons at either subthreshold or suprathreshold intensities, and only minimal IPSCs at suprathreshold intensities, suggesting polysynaptic activity also does not spread from IL L5 to PL L5 (EPSC: sub = −4 ± 2 pA, supra = −13 ± 5 pA, p = 0.055; n = 9 pairs; IPSC: sub = 7 ± 2 pA, supra = 54 ± 22 pA, p = 0.027; n = 9 pairs; 3 animals) (Figure 5G–I). These findings suggest that rBLA, but not cBLA, influences PL L5 PT neurons via local connections across layers of the PFC. Figure 5 with 1 supplement see all Download asset Open asset Rostral basolateral amygdala (rBLA) but not caudal basolateral amygdala (cBLA) engages prelimbic (PL) L5 pyramidal tract (PT) neurons via local networks. (A) Schematic for recording rBLA-evoked responses at PL L2 cortico-amygdalar (CA) neurons (black) and polysynaptic responses at PL L5 PT neurons (red). (B) Left, Example rBLA-evoked EPSPs and APs from PL L2 CA neurons recorded in current-clamp at resting membrane potential (RMP) with subthreshold (bottom, gray) and suprathreshold (top, black) stimulation. Right, Average rBLA-evoked excitatory postsynaptic currents (EPSCs) (bottom, at EGABA) and IPSCs (top, at EAMPA) recorded at PL L5 PT neurons at the same subthreshold (gray) and suprathreshold (red) stimulation. (C) Summary of rBLA-evoked EPSC1 and IPSC1 amplitudes at PL L5 PT neurons. Gray lines show subthreshold and suprathreshold responses in the same neuron (n = 8 cells, 4 animals). (D–F) Similar to (A–C) for rBLA-evoked EPSPs and APs at PL L2 CA neurons (gray and black) and lack of polysynaptic responses at infralimbic (IL) L5 PT neurons (right, blue) (n = 8" @default.
• W4313305230 created "2023-01-06" @default.
• W4313305230 creator A5033400284 @default.
• W4313305230 date "2022-09-29" @default.
• W4313305230 modified "2023-09-30" @default.
• W4313305230 title "Editor's evaluation: Rostral and caudal basolateral amygdala engage distinct circuits in the prelimbic and infralimbic prefrontal cortex" @default.
• W4313305230 doi "https://doi.org/10.7554/elife.82688.sa0" @default.
• W4313305230 hasPublicationYear "2022" @default.
• W4313305230 type Work @default.
• W4313305230 citedByCount "0" @default.
• W4313305230 crossrefType "peer-review" @default.
• W4313305230 hasAuthorship W4313305230A5033400284 @default.
• W4313305230 hasBestOaLocation W43133052301 @default.
• W4313305230 hasConcept C15744967 @default.
• W4313305230 hasConcept C169760540 @default.
• W4313305230 hasConcept C169900460 @default.
• W4313305230 hasConcept C2779144063 @default.
• W4313305230 hasConcept C2779934784 @default.
• W4313305230 hasConcept C2780451725 @default.
• W4313305230 hasConcept C2781195155 @default.
• W4313305230 hasConceptScore W4313305230C15744967 @default.
• W4313305230 hasConceptScore W4313305230C169760540 @default.
• W4313305230 hasConceptScore W4313305230C169900460 @default.
• W4313305230 hasConceptScore W4313305230C2779144063 @default.
• W4313305230 hasConceptScore W4313305230C2779934784 @default.
• W4313305230 hasConceptScore W4313305230C2780451725 @default.
• W4313305230 hasConceptScore W4313305230C2781195155 @default.
• W4313305230 hasLocation W43133052301 @default.
• W4313305230 hasOpenAccess W4313305230 @default.
• W4313305230 hasPrimaryLocation W43133052301 @default.
• W4313305230 hasRelatedWork W1573440693 @default.
• W4313305230 hasRelatedWork W1977502547 @default.
• W4313305230 hasRelatedWork W1987895660 @default.
• W4313305230 hasRelatedWork W2069807169 @default.
• W4313305230 hasRelatedWork W2148791245 @default.
• W4313305230 hasRelatedWork W2767851849 @default.
• W4313305230 hasRelatedWork W2952322062 @default.
• W4313305230 hasRelatedWork W41800005 @default.
• W4313305230 hasRelatedWork W4310938028 @default.
• W4313305230 hasRelatedWork W4313305213 @default.
• W4313305230 isParatext "false" @default.
• W4313305230 isRetracted "false" @default.
• W4313305230 workType "peer-review" @default.