Mice
Experiments were performed on C57BL/6 male mice (N = 61; The Jackson Laboratory, Bar Harbor, ME), GlyT2-eGFP mice (N = 6; graciously provided by Dr. Manuel Miranda-Arango, University of Texas at El Paso, El Paso, TX), and GlyT2-Cre+/− mice (N = 18; graciously provided by Dr. Jack Feldman, University of California, Los Angeles). Litters were weaned at PND 21 and housed together until stereotaxic microinjections were performed at PND 70–84 (adult). Mice received food and water ad libitum in a 12-h light/dark cycle from 7:00 am to 7:00 pm. This age corresponds to the age of the animals used in the Paxinos and Franklin Mouse Brain Atlas, from which all the stereotaxic coordinates were derived, and cytoarchitectural boundaries delineated [52]. Following surgical procedures, mice were single-housed and monitored for the duration of the recovery period. Experiments were performed in accordance with and approved by the Institutional Animal Care and Use Committee of the University of Texas at El Paso (UTEP) and the University of Massachusetts Amherst (UMass).
Stereotaxic microinjections
Mice were sedated by inhaling 5% isoflurane vapors (Piramal Critical Care, Bethlehem, PA), then placed on a stereotaxic apparatus (model 900, David Kopf, Tujunga, CA) and immobilized using ear bars and a nose cone. Mice were maintained under 1.5–2% isoflurane throughout the duration of the surgical procedure. With bregma as a reference, the head of the mice were leveled on all 3 axes. A craniotomy was performed directly dorsal to the injection site. Then, using a microinjector (Stoelting Co., Wood Lane, IL) with a 5-μl Hamilton syringe (Hamilton Company Inc., Reno, NV) and a 32-gauge steel needle, unilateral injections of 50–80 nl of the retrograde neuronal tracer Fluoro-Gold (Molecular Probes, Eugene, OR, catalog# H22845, lot# 1611168) were infused into the PnC (coordinates from bregma: AP − 5.35 mm; ML + 0.5 mm, DV − 5.6 mm; N = 4 mice). The CAG-FLEX-tdTomato (Addgene# 28306-AAV1; lot# v16602) or rAAVDJ/Ef1α-DIO-eArch3.0-eYFP (Deisseroth Lab, virus# GVVC-AAV-055) viral vectors were injected (200 nl) in the PnC of mice expressing the CRE recombinase enzyme in GlyT2+ neurons (GlyT2-Cre mice; N = 10). In separate animal cohorts, 100–125 nl of AAV particles were unilaterally injected in the CeA (AP − 1.35 mm, ML + 2.66 mm, DV − 4.6 mm). For these viral injections, pAAV DJ-CamKIIα-eArch3.0-eYFP (Deisseroth Lab, # GVVC-AAV-053, lot# 1668 and 3605), pAAV DJ-CamKIIα-NpHR3.0-eYFP (Deisseroth Lab, #GVVC-AAV-057, lot#1378), pAAV DJ-CamKIIα-hChR2(H134R)-eYFP (Deisseroth Lab, #GVVC-AAV-037, lot#3150), pAAV DJ-CamKIIα-eYFP (Deisseroth Lab, #GVVC-AAV-8), or pAAV DJ-CamKIIα-mCherry (Deisseroth Lab, #GVVC-AAV-009) viral particles were used (4 × 1012 particles/mL; vectors were obtained from Dr. Karl Deisseroth’s Lab/Optogenetics Innovation Lab, Gene Vector and Virus Core, Stanford University, Palo Alto, CA, or through Addgene from Dr. Edward Boyden’s Lab plasmids from, Massachusetts Institute of Technology, Cambridge, Massachusetts). Fluoro-Gold and viral particles were delivered at a rate of 50 nL/min. The microinjection syringe was left in place for 10 min after infusion to limit spillover during needle retraction. Mice injected with Fluoro-Gold recovered for 5–7 days, to allow optimal Fluoro-Gold retrograde transport to occur. AAV-injected mice recovered for 3–5 weeks to allow sufficient time for maximal viral transduction.
Immunohistochemistry
Mice were perfused transcardially with 0.9% saline solution for 10 min followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer saline (PBS; pH 7.4) for 15 min, brains were then extracted and post-fixed overnight in 12% sucrose in PFA solution. After three 0.1 M PBS rinses (5 min each), brains were frozen in chilled hexanes for 1 min and stored at − 80 °C. Using a microtome, four 1-in-5 series of 30-μm coronal sections were cut and stored in cryoprotectant (50% 0.05 M phosphate buffer, 30% ethylene glycol, 20% glycerol) at − 20 °C. One of the series was rinsed three times (5mins each) with 0.1 M Tris-buffered saline (TBS; pH 7.4), mounted and coverslipped to visualize injection and projection sites. An adjacent series of brain sections was Nissl-stained to determine plane of section and delineate cytoarchitectural boundaries. The two remaining series were used for immunohistochemistry. For mice injected with Fluoro-Gold, coronal tissue sections at the level of the PnC, CeA, and PPTg were washed with 0.1 M TBS (5 washes, 5 min each) and incubated in blocking solution (2% normal donkey serum, 0.1% Triton X-100; in 0.1 M TBS) for 1–2 h at room temperature. PPTg sections were incubated with a goat anti-ChAT primary antibody (1:100, Millipore, catalog# AB144P-200UL, lot# 2854034, RRID:AB_90661) for 60 h at 4 °C, washed with TBS, and then incubated in a Cy3-conjugated donkey anti-goat secondary antibody (1:500, Jackson ImmunoResearch Laboratories, catalog# 705-165-147, lot# 115611, RRID:AB_2307351) for 4-5 h at room temperature. Tissue slices were then washed with TBS, mounted, and coverslipped. Similarly, for mice injected with viral particles, tissue sections containing the PnC, CeA, and the PPTg were incubated in a chicken anti-GFP primary antibody (1:1000, Abcam, catalog# ab13970, lot# GR236651-13, RRID:AB_300798), then incubated in an Alexa Fluor 488-conjugated donkey anti-chicken secondary antibody (1:500, Jackson ImmunoResearch Laboratories, catalog# 703-545-155, lot# 130357, RRID:AB_2340375), followed by incubation in NeuroTraceTM (640/660 deep red fluorescent nissl stain, 1:100 in TBS, Thermo Fisher, catalog# N21483, RRID:AB_2572212). NeuroTraceTM was alternatively used to determine plane of section and cytoarchitecture. Tissue sections at the level of the PnC of GlyT2-eGFP mice injected with pAAV DJ-CamKIIα-mCherry in the CeA (N = 6 mice) were incubated with a chicken anti-mCherry (1:1000, Abcam, catalog# ab205402, lot# GR225123-3, RRID:AB_2722769) and a rabbit anti-PSD95 (1:500, Abcam, catalog# ab12093, lot# GR317630-1, RRID:AB_298846) primary antibodies. Then, sections were incubated with a Cy3-conjugated donkey anti-chicken (1:500, Jackson ImmunoResearch Laboratories, catalog# 703-165-155, lot# 130328, RRID:AB_2340363) and a Cy5-conjugated donkey anti-rabbit (1:500, Jackson ImmunoResearch Laboratories, catalog# 705-545-147, lot#125100, RRID:AB_2336933) as described above.
In situ hybridization
Mice injected with pAAVDJ-CamKIIα-eYFP in the CeA (N = 3 mice) were anesthetized with inhaled isoflurane and rapidly decapitated. Brains were harvested, frozen in chilled isopentane, and stored at – 80 °C. Serial coronal sections (15 μm) at the level of the CeA were cut in a cryostat, directly mounted onto glass slides, and stored at – 80 °C. Tissue sections on slides were submerged in freshly prepared cold 4% PFA for 15 min, rinsed twice briefly with 0.1 M phosphate buffer (PB) and dehydrated in increasing ethanol solutions (50%, 70%, 100%, 100%; 5 min each at room temperature). Then, the RNAscope assay (Advanced Cell Diagnostics) started by incubating in hydrogen peroxide (H2O2) for 10 min in a humidified box, followed by protease III incubation for 15 min. RNA hybridization probes against genes encoding mouse VGLUT2 (319171-C1) and eYFP (312131-C2) were then incubated for 2 h at 40 °C. Antisense probes were also included as controls in a separate glass slide. Probe signals were then developed separately with Opal Dyes (opal 690 1:1.5 K, opal 520 1:750) and coverslipped with ProLong GoldTM with DAPI.
Morphological reconstruction
Z-stacks from tissue sections of GlyT2-eGFP and GlyT2-Cre mice were obtained on a Nikon A1 Resonant Confocal microscope (Nikon Instruments Inc., Melville, NY) equipped with NIS-Elements High Content Analysis software (version 5.02, Nikon Instruments Inc., Melville, NY). Tissue sections containing labeled CeA and PnC neurons were first examined on a single Z-plane with the × 10 objective to survey the tissue section. Using a × 60 objective, an area (212.56 μm width × 212.56 μm height) within CeA and PnC sections was then sequentially scanned by the 488-, 561-, and 640-nm laser lines in 0.1 μm Z-steps throughout the 30-μm tissue section. Z-stacks were analyzed with NIS-Elements 5.0 Advanced Research software (version 5.02, Nikon Instruments Inc., Melville, NY). To visualize close appositions of CeA axons (labeled with mCherry) with GlyT2+ neurons (labeled with eGFP) in GlyT2-eGFP mice, a binary layer was configured to segregate putative synaptic contacts of > 50 nm in distance (due to technical limitations). These contacts were imaged in split-channels and orthogonal views. Then, z-stacks were reconstructed in three-dimension and volume was rendered.
Nissl stain
Series of tissue slices were mounted on gelatin-coated slides and air-dried overnight. Slides were immersed in deionized water, followed by ascending concentrations of ethanol (3 min each: 50%, 75%, 95%, and 100%), and then in xylenes (30 min). Brain slices were rehydrated in descending concentrations of ethanol and DI water, dipped 12–20 times in a thionin acetate solution, and then washed in DI water. Brain slices were dehydrated, and slides were then coverslipped with DPX and air-dried overnight.
Microscopy analysis
Tissue sections were analyzed with an Axio Observer.Z1 epifluorescence microscope (Carl Zeiss Inc., Thornwood, NY) equipped with Fluoro-Gold, GFP, Cy3 filters, × 10 and × 40 objectives, a motorized stage, and Axiovision Rel. 4.8 software (Carl Zeiss Inc., Thornwood, NY). To create photomontages, single Z-plane images were obtained with the MosaiX module of the Axiovision Rel. 4.8 software at × 10 for each fluorophore sequentially (1024 × 1024 pixel resolution). Images acquired for the intensity and quantification of eYFP fluorescence analysis were captured and processed using identical settings. A total of 836 images (fluorescence and bright-field) were analyzed for each brain region. Nissl-stained slices were imaged using bright-field microscopy, and boundaries were delineated using Adobe Illustrator (Adobe, San Jose, CA).
Electrophysiological recordings
Whole-cell recordings (N = 10 mice; n = 26 CeA neurons and n = 38 GlyT2+ neurons) were performed using glass pipettes (3–5 MΩ) filled with intracellular solution (in mM): KMeSO4 (125), KCl (10), HEPES (10), NaCl (4), EGTA (0.1), MgATP (4), Na2GTP (0.3), Phosphocreatine (10), Biocytin (0.1%) (pH = 7.3; osmolarity = 285–300 mosm). The glass microelectrode was mounted on a patch clamp headstage (Molecular Devices LLC, Sunnyvale, CA; catalog# CV-7B), which was attached to a multi-micromanipulator (Sutter Instrument, Novato, CA; catalog# MPC-200). Data were acquired with pClamp10 software using a MultiClamp™ 700B amplifier (Molecular Devices LLC, Sunnyvale, CA) and a Digidata 1550B digitizer (Molecular Devices LLC, Sunnyvale, CA). EYFP-expressing CeA cells and tdTomato-expressing GlyT2+PnC cells were imaged and targeted using NIS-Elements Basic Research software (version 5.11, Nikon Instruments Inc., Melville, NY). Only cells with an initial seal resistance greater than 1GΩ, a resting membrane potential between − 60 mV and − 70 mV, and a holding current within – 100 pA to 100 pA at resting membrane potential and overshooting action potentials were used.
In CeA slices, 15 pA depolarizing current steps were injected for 500 ms to induce action potentials in CeA neurons expressing CamKIIα-ChR2-eYFP, in the current clamp. Spontaneous EPSCs were recorded at a holding potential of − 70 mV, in the voltage clamp. Evoked EPSPs were also recorded in these CeA neurons held at − 70 mV, in response to a 1-ms blue light pulse. Blue light was delivered every 30 s using a 200-μm optic fiber mounted on a micromanipulator connected to a blue LED (473 nm; Plexon, Dallas, TX) and positioned in close proximity to the recorded neuron.
In PnC slices, electrical properties of the GlyT2+ neurons were first recorded in the voltage clamp. Spontaneous excitatory post-synaptic currents (sEPSC) were recorded for 5 min at − 70 mV, and inhibitory post-synaptic currents (IPSCs) were recorded for 5 min at 0 mV. Then, in the current clamp mode, 15 pA depolarizing current steps (from – 150 pA to 150 pA) were injected for 500 ms to analyze the spiking properties of GlyT2+ cells. Pulses of blue light (1 ms), applied every 30 s, were used to photo-stimulate CeA excitatory fibers in PnC slices. The photo-stimulation elicited EPSPs and EPSCs in GlyT2+ neurons, held at − 70 mV. Paired light pulses with 50 and 100 ms ISI were also delivered to characterize short-term plasticity. GlyT2+ neurons were then held at 0 mV, to record light-evoked inhibitory post-synaptic current (IPSCs) or potentials (IPSPs). The NMDA receptor antagonist AP5 (50 μM) and the AMPA receptor antagonist DNQX (25 μM) were freshly diluted prior to use. At synapses between CeA excitatory cells and GlyT2+ neurons, synaptic events were recorded for 10 min in aCSF. Then, 20 min after the bath application of glutamate receptor antagonists, synaptic events were recorded during 10 min in the presence of the antagonists. This was followed by a 20-min washout period, and synaptic events were recorded during the following 10 min, in aCSF.
At the end of all whole-cell recordings, the cell membrane was sealed by forming an outside-out patch. The glass microelectrode was slowly retracted, and as the series resistance increased, the membrane potential was clamped at − 40 mV. The 300-μm-thick acute brain slices containing the recorded cells (CeA or PnC) were immersed in 4% PFA solution overnight. Following overnight PFA fixation, these brain slices were rinsed with PBS (3 times, 5 min each). Slices were then incubated in anti-RFP and/or anti-GFP antibodies and in complementary secondary antibodies to enhance the fluorescence of the viral vectors used. Following PBS rinses, slices were incubated with Cy5-conjugated streptavidin (a biotin-binding protein) diluted in PBS (with 0.1% Triton X-100) at room temperature for 4–5 h or overnight at 4 °C. Slices were then rinsed with PBS, mounted on glass slides, coverslipped, and sealed with ProLongTM Gold antifade reagent (Invitrogen by Thermo Fisher Scientific, Waltham, MA, catalog# P36934, lot# 1943081), and air-dried overnight in the dark.
Behavioral testing
Three to four weeks after the viral injection in the CeA, non-injected WT control mice and WT mice injected with a viral vector were sedated by inhaling 5% isoflurane vapors, placed in a stereotaxic apparatus, and immobilized using ear bars and a nose cone. Mice were maintained under anesthesia (1.5–2% isoflurane), and the head was leveled in all three axes. With bregma as a reference, a craniotomy was drilled directly dorsal to the implantation site, at the PnC level. A cannula guide with a 200-μm core optical fiber (Thorlabs, Newton, NJ) was then implanted over the PnC (AP − 5.35 mm, ML + 0.5 mm, DV − 5.3 mm), and cemented to the skull with dental cement (Parkell, Edgewood, NY). Mice recovered for 7 days post-surgery before behavioral testing. Mice underwent the PPI task in a startle response system (PanLab System, Harvard Apparatus, Holliston, MA). Behavioral testing trials were designed, and data were recorded using PACKWIN V2.0 software (Harvard Apparatus, Holliston, MA). Sound pressure levels were calibrated using a standard SPL meter (model 407730, Extech, Nashua, NH). Mice were placed on a movement-sensitive platform. Vertical displacements of the platform induced by startle responses were converted into a voltage trace by a piezoelectric transducer located underneath the platform. Startle amplitude was measured as the peak to peak maximum startle magnitude of the signal measured during a 1-s window following the presentation of the acoustic stimulation. Prior to any testing session, animals were first handled and acclimatized to the testing chamber, where the mice were presented to a 65-dB background noise, for 10 min. This acclimatization period was used to reduce the occurrence of movement and artifacts throughout testing trials. Following the acclimatization period, an input/output (I/O) assay was performed to test startle reactivity. This I/O test began with the presentation of a 40-ms sound at different intensities (in dB: 70, 80, 90, 100, 110, and 120) every 15 s, in a pseudorandomized order. Background noise (65 dB) was presented during the 15 s between sounds. A total of 35 trials (i.e., 7 sound intensities, each sound presented 5 times) were acquired and quantified. Startle reactivity, derived from this I/O assay, allowed the gain of the movement-sensitive platform to be set. This gain allowed the startle responses to be detected within a measurable range. Once determined, the gain for each experimental subject was kept constant throughout the remaining of the experiment. Following a 1-h resting period, mice were presented with seven startle-inducing 120 dB (40 ms) sounds called “pulse-alone” stimulations. These 120 dB sounds were presented every 29 s (interspersed with 65 dB background noise) and were used to achieve a stable baseline startle response. The following PPI test consisted of two different conditions as follows: (1) startling pulse-alone stimulations (for baseline startle amplitude), and (2) combinations of a prepulse (75 dB noise; 20 ms) followed a 120-dB startling pulse (40 ms) at 8 different interstimulus intervals (in ms): 10, 30, 50, 100, 200, 300, 500, and 1000 (end of prepulse to onset of startle pulse). The inter-trial interval of these two conditions was 29 s.
For combined optogenetic manipulations, animals injected with either control viral vectors or vectors containing ChR2 (N = 8 mice), Arch3.0 (N = 16 mice), NpHR3.0 (N = 8 mice), or the control vector (pAAV DJ-CamKIIα-eYFP; N = 8 mice) were tested in the startle chamber. These animals were closely monitored to ensure that they were comfortably tethered to an optic fiber, which exited through a small opening from the roof of the startle chamber. The optic fiber (200 μm diameter, Thorlabs, Newton, NJ) was connected to the animal’s head via a cannula implanted on the head of the mouse with a zirconia sleeve (Thorlabs, Newton, NJ). Animals were tethered ~ 15 min before testing and allowed to move freely, exploring their home cages before being transferred to the startle chamber. Optogenetic stimulation was triggered by a signal from the Packwin software (PanLab System; Harvard Apparatus, Holliston, MA), which was transformed into a TTL pulse. This TTL pulse triggered a waveform generator (DG1022, Rigol Technologies), which was used to modulate light stimulation. Photo-stimulation was delivered using a blue 473-nm laser (Opto Engine LLC, Midvale, UT) for ChR2 activation. Photo-inhibition was delivered using a yellow 593.5-nm laser (Opto Engine LLC, Midvale, UT) for NpHR3.0 activation or a green 532-nm LED (Plexon, Dallas, TX) for Arch3.0 activation. During PPI trials paired with optogenetic inhibition, a train of light stimulation (1 ms light ON, 200 ms light OFF) was delivered at 5 Hz and was either (1) delivered 500 ms prior to and concurrent to the pulse-alone stimulation, or (2) delivered 500 ms before the prepulse, lasting the entire ISI. During PPI trials with optical stimulation used as a prepulse, a 5-Hz or 20-Hz stimulation train (3 pulses of 15 ms) was delivered at various ISI (10, 30, 50, 100, 200, 300, 500, and 1000 ms; end of prepulse to onset of startle pulse) prior to the startling pulse. Blue light stimulation was paired with pulse-alone stimulations in a subset of mice. At the end of each experiment, histological analyses were performed to confirm that (1) the injected viral particles were confined to the CeA, and (2) the cannula guide placement was successfully aimed at the PnC. If these criteria were not met, the subject was excluded from the study.
Statistical analysis
Cell counting of EYFP-labeled or VGLUT2-expressing somata within the CeA was performed in a tissue slice series of 6 slices spanning levels 40 to 44 of the Paxinos and Franklin Mouse Brain Atlas [52]. Imaging was performed as outlined in the “Microscopy analysis” section. Percentages of labeled somata were calculated as EYFP+/Neurotrace-labeled cells (Fig. 3) or VGLUT2+/EYFP+ (Fig. 4). Statistical analyses were performed using SigmaPlot (Systat Software, Inc., San Jose, CA). Normality and equal variance of the data were first tested, and data transformations were made before performing further statistical analyses. We determined the significance of the interaction between the factors assessed using ANOVA. For the results of whole-cell patch clamp recordings with receptor antagonists, one-way repeated-measures (RM) ANOVA and Tukey post hoc testing were used to assess the effect of the receptor antagonists on the light-evoked events. For PPI in vitro results, one-way ANOVAs and Tukey post hoc testing were used to reveal if at any ISI the electrically evoked fEPSPs were significantly attenuated by the optical stimulation of CeA-PnC excitatory synapses. PPI was defined and measured as [1–(startle amplitude during “Prepulse+Pulse” trials/startle amplitude during “Pulse” trials)] × 100. Two-way RM ANOVA was used to assess the effect of the vector used, light, sound intensity/ISI, and light interaction and the interaction among groups. Then Tukey testing was applied for post hoc comparisons. For optical stimulation experiments where the photo-stimulation of CeA fibers was used as a prepulse in vivo, two-way RM ANOVA was used to assess the effect of the stimulation modality/frequency used, ISI, ISI, and stimulation modality/frequency interaction and the interaction among groups. Then, Tukey testing was applied for post hoc comparisons. A confidence level of p < 0.05 was considered statistically significant. Sample sizes were chosen based on expected outcomes, variances, and power analysis. Data are presented as means ± SEM. N indicates total number of animals; n indicates total number of brain slices or testing trials. Adobe Illustrator was used to create figures.