Supporting Information

Bero et al. 10.1073/pnas.1408378111 SI Materials and Methods isolate was (13.5 ± 0.5) × 106 reads, of which (9.3 ± 0.3) × 106 Animals. Wild-type (B6SJL; Taconic Farms) male mice were used reads were aligned. Aligned reads were mapped to the RefSeq at 2.5 ± 0.5 mo of age for experimentation. Mice were given ad database and counted (HTSeq). Differential expression analysis libitum access to food and water. Mice were experimentally naïve was performed using DESeq (Bioconductor) followed by Student before the onset of the study and housed (4–5 mice per cage) in t test to model the experimental and gene-specific dispersion, 12 h light/12 h dark conditions (light phase beginning at 7:00 respectively. Genes were considered differentially expressed if P ≤ AM). All animal work was approved by the Committee for 0.05. Ontological and network analyses of differentially expressed Animal Care of the Division of Comparative Medicine at the genes were performed using Ingenuity Pathway Analysis software. Massachusetts Institute of Technology (MIT). Quantitative RT-PCR. Quantitative RT-PCR was performed es- Behavioral Analysis. All behavioral experiments were conducted sentially as described (2). Briefly, total mRNA was extracted during the light cycle and performed essentially as described (1). from mPFC isolates (anterior cingulate cortex; bregma +1.2 mm Briefly, mice were exposed to the conditioning chamber (TSE to +0.6 mm) (3) using the RNeasy kit (Qiagen), reverse tran- Systems) for 3 min (habituation phase), after which they received scribed (Invitrogen), and quantitatively amplified using a ther- two unsignaled foot shocks (2-s duration, 0.8 mA, 30-s intershock mal cycler (Bio-Rad), SsoFast EvaGreen (Bio-Rad) and gene- C interval). Following the foot shocks, mice remained in the specific primers (Table S3). The comparative t method was chamber for an additional 15 s. Control groups of mice were used to determine differences in gene expression. Values were treated identically, but either did not receive a foot shock normalized to expression levels of Gapdh. (context control group) or received foot shocks in the absence of context exploration (shock control group). Long-term memory Electron Microscopy. Mice were transcardially perfused with chilled tests were performed 1 d (recent memory) and 30 d later (re- PBS (pH 7.4) followed by 2.5% glutaraldehyde/2.5% para- mote memory) by returning the mice to the experimental cham- formaldehyde in 0.1 M sodium cacodylate buffer (pH 7.4; ber for 2 min. Freezing, defined as an absence of all movement Electron Microscopy Sciences). Brains were removed and post- except respiration, was scored every 10 s by an experimenter fixed in 2.5% glutaraldehyde/2.5% paraformaldehyde in 0.1 M blind to condition. Movement velocity was recorded using au- sodium cacodylate buffer at 4 °C overnight. Following fixation, μ tomated procedures (TSE Systems). brain sections (100 m thick) were cut on a vibratome (Leica), washed in 0.1 M sodium cacodylate buffer, postfixed in 1% os- Immunohistochemistry. Immunohistochemical staining was per- miumtetroxide/1.5% potassium ferrocyanide, incubated in 1% formed essentially as described (2). Briefly, mice were trans- aqueous uranyl acetate, and dehydrated. The samples were em- cardially perfused with chilled PBS (pH 7.4) followed by 10% bedded in TAAB Epon (Marivac), polymerized, and sectioned paraformaldehyde under deep anesthesia (ketamine, xylazine). (80 nm thick) using a microtome (Reichert). Resultant sections Brains were removed and postfixed in 10% paraformaldehyde at were placed onto copper grids, stained with lead citrate, and 4 °C overnight. Following fixation, serial coronal brain sections imaged using a transmission electron microscope (1200EX; JEOL). (40 μm thick) were cut on a vibratome (Leica) and collected Micrographs of mPFC synapses (anterior cingulate cortex; bregma from frontal cortex to caudal hippocampus. Sections were per- +1.2 mm to +0.6 mm) (3) were recorded using a CCD camera meabilized and blocked with 10% FBS/0.3% Triton-X 100 in (Advanced Microscopy Techniques). Synapses, active zones, post- PBS and incubated overnight with the following primary anti- synaptic densities, and docked synaptic vesicles were identified bodies: Zif268/Egr-1 (1:100; Santa Cruz, sc-101033), synapto- and defined as described (4, 5) and quantitative analysis of physin (1:1,000; Sigma, S5768) or CaMKIIα (1:100; Millipore, synapse ultrastructure was performed using ImageJ (NIH) by an MAB8699). Primary antibodies were visualized with a Cy5-con- experimenter blind to condition. jugated secondary antibody (1:400; Jackson ImmunoResearch, 115-175-146) and nuclei were visualized with Hoechst 33342 Golgi–Cox Impregnation. Golgi–Cox staining was performed es- (Invitrogen). A mouse brain atlas (3) was used to identify medial sentially as described (1). Briefly, mice were transcardially prefrontal cortex (mPFC) (anterior cingulate cortex; bregma + perfused with chilled PBS (pH 7.4) followed by 10% para- 1.2 mm to +0.6 mm), hippocampal area cornu ammonis 1 (CA1; formaldehyde under deep anesthesia (ketamine, xylazine). bregma −1.8 mm to −2.3 mm), and entorhinal cortex (bregma −2.2 Brains were removed and postfixed in 10% paraformaldehyde at mm to −2.5 mm), and images were acquired using a confocal mi- 4 °C overnight. Following fixation, whole-brain Golgi–Cox im- croscope (LSM 510; Zeiss) at identical settings for each condition. pregnation was performed using the Rapid GolgiStain kit (FD Images were quantified using ImageJ (National Institutes of Health, Neurotechnologies) as per the manufacturer’s instructions and NIH) by an experimenter blind to condition groups, whenever serial brain sections (60 μm thick) were cut on a vibratome possible. (Leica). Using a confocal microscope (LSM 510; Zeiss), den- dritic spines present on visually unobstructed apical and basal Genome-Wide RNA Sequencing. RNA sequencing was performed dendrites of layer II/III mPFC pyramidal neurons (anterior essentially as described (1). Briefly, total mRNA was extracted cingulate cortex; bregma +1.2 mm to +0.6 mm) (3) were imaged from mPFC isolates (anterior cingulate cortex; bregma +1.2 mm by an experimenter blind to condition. To capture the morpho- to +0.6 mm) (3) using the RNeasy kit (Qiagen). Purified mRNA logical extent of all spines on each dendritic segment of interest, was quality controlled (Agilent Bioanalyzer), poly-A purified, a z-stack image series was acquired from the anterior to poste- and converted to cDNA using the Illumina TruSeq protocol as rior limits of each spine set (incremental z-step distance, 0.25–0.5 per the manufacturer’s instructions. High-throughput sequencing μm). Quantitative morphological analysis of individual dendritic was performed using the Illumina HiSeq 2000 platform at the spines was performed using ImageJ (NIH) and spines were classi- MIT BioMicro Center. Sequence reads were aligned to the fied as thin, mushroom, or stubby using previously described cri- mouse mm9 genome using TopHat. The mean yield per mPFC teria (6). Briefly, thin spines lack a definable head or possess

Bero et al. www.pnas.org/cgi/content/short/1408378111 1of6 a head whose diameter is less than 120% of the stem diameter. Comparative Medicine at MIT. Mice were anesthetized under Mushroom spines exhibit a bulbous, mushroom-shaped head volatile isoflurane (0.5–2%) with the head secured in a stereo- whose diameter is equal to or greater than 120% of the stem taxic frame (Leica). The skin and periosteum were removed to diameter. Stubby spines lack a definable neck and their length expose the skull, and either the AAV5-CaMKIIα-eNpHR3.0- approximately equals their width. Spine analysis was performed EYFP or the control AAV5-CaMKIIα-EYFP vector was ste- by an experimenter blind to condition. reotaxically injected into the bilateral mPFC (1.0 μL hemi- − − sphere 1; flow rate, 0.075 μL min 1; bregma +1.0 mm, 0.35 mm Electrophysiology. For whole-cell patch-clamp recordings, coronal lateral to midline, 1.3 mm below the dura) (3) using a motorized brain slices (250 μm thick) containing the mPFC (anterior cin- injector (Stoelting). Following viral injection, injection needles gulate cortex; bregma +1.2 mm to +0.6 mm) (3) were prepared in ice-cold dissection buffer containing (in mM) 211 sucrose, 3.3 remained in place for 5 min and were then slowly withdrawn. Optic fibers (300-μm diameter, NA = 0.39; Thorlabs) held in KCl, 1.3 NaH2PO4, 0.5 CaCl2, 10 MgCl2, 26 NaHCO3, and 11 μ glucose using a vibratome (VT1000S; Leica). Slices were re- a stainless steel ferrule (330- m diameter; Precision Fiber Products) were implanted into the unilateral mPFC (counter- covered in a submerged chamber with 95% O2–5% CO2-satu- rated artificial cerebrospinal fluid (aCSF) consisting of (in mM) balanced for hemisphere) as close as possible to the midline + 124 NaCl, 3.3 KCl, 1.3 NaH2PO4, 2.5 CaCl2, 1.5 MgCl2,26 (bregma 1 mm, 0.2 mm lateral to midline, 0.8 mm below the NaHCO3, and 11 glucose for 1 h at 28–30 °C. Intracellular re- dura) and secured to the skull using dental cement (Stoelting). cordings were made from layer II/III mPFC pyramidal neurons After drying, the incision was closed and sealed with tissue ad- using recording pipettes (3–5MΩ) filled with an internal solu- hesive (Vetbond). Animal body temperature was maintained at tion containing (in mM) 145 CsCl, 5 NaCl, 10 Hepes-CsOH, 10 37 °C until fully recovered from anesthesia. All experiments were EGTA, 4 MgATP, and 0.3 Na2GTP. For miniature excitatory conducted 4 wk following surgery. postsynaptic current (mEPSC) measurements, tetrodotoxin (1 μM) and picrotoxin (50 μM) were added to the perfusion In Vivo Laser Delivery. To permit optogenetic inhibition of mPFC solution (aCSF) and cells were held at −70 mV. For ex vivo neurons in vivo, a 200-mW 593-nm DPSS laser (Opto Engine) was optical inhibition experiments, 20 mW of constant optical (593 connected to a patch cord with a fiber channel/physical contact nm) stimulation was generated by a 200-mW 593-nm diode- connector at each end (Doric Lenses). To allow free movement of pumped solid-state (DPSS) laser (Opto Engine) and delivered the mouse during optical stimulation, the patch cord was con- during current injection (500-ms duration, 25-pA step increment nected to a fiber-optic rotary joint (Doric Lenses), and separate from −25 to 150 pA). A MultiClamp 700B amplifier and a Digidata ends of a second patch cord were connected to the rotary joint and 1440A analog-to-digital converter (Axon Instruments) were used the implanted fiber optic. During contextual fear conditioning, 8 for data aquisition and data were analyed with pClamp10 (Axon mW (∼28.28 mW/mm2 at the tip of the fiber) of constant optical Instruments). (593 nm) stimulation was laser-generated and delivered to the mPFC through the patch cord, mating sleeve, and implanted fiber Virus Construction and Packaging. Adeno-associated viral (AAV) optic. Laser output was measured using an optical power and vectors were serotyped with AAV5 coat proteins and packaged by the University of North Carolina Vector Core. Viral titers were energy meter (Thorlabs) and manipulated using a Master-8 12 −1 pulse stimulator (A.M.P.I.). 3 × 10 particles mL for the AAV5 encoding eNpHR3.0 halorhodopsin fused to the enhanced yellow fluorescent pro- 2+ Statistical Analysis. Statistical significance was determined by two- tein (EYFP) under control of the Ca /calmodulin-dependent tailed t test if two groups were compared. When variance dif- protein kinase IIα (CaMKIIα) promoter (AAV5-CaMKIIα- × 12 −1 fered significantly between groups, Welch’s t test was used. One- eNpHR3.0-EYFP), and 6 10 particles mL for the AAV5 ’ encoding EYFP alone under control of the CaMKIIα promoter way ANOVA followed by Tukey s post hoc test for multiple comparisons was used when more than two groups were com- (AAV5-CaMKIIα-EYFP). The maps for the above constructs are available online at www.optogenetics.org. pared. Correlations were determined using the Pearson product- moment correlation test. Statistical analyses were performed Stereotaxic Surgery. All surgeries were performed under aseptic using Prism version 5.0 for Macintosh (GraphPad). Values were conditions in accordance with the guidelines of the Division of accepted as significant if P ≤ 0.05.

1. Gräff J, et al. (2014) Epigenetic priming of memory updating during reconsolidation to 5. Schikorski T, Stevens CF (1997) Quantitative ultrastructural analysis of hippocampal attenuate remote fear memories. Cell 156(1–2):261–276. excitatory synapses. J Neurosci 17(15):5858–5867. 2. Gräff J, et al. (2012) An epigenetic blockade of cognitive functions in the neurodegenerating 6. Chen Y, et al. (2013) Impairment of synaptic plasticity by the stress mediator CRH brain. Nature 483(7388):222–226. involves selective destruction of thin dendritic spines via RhoA signaling. Mol 3. Franklin KB, Paxinos G (1996) The Mouse Brain in Stereotaxic Coordinates (Academic, Psychiatry 18(4):485–496. San Diego). 4. Couteaux R, Pécot-Dechavassine M (1970) [Synaptic vesicles and pouches at the level of “active zones” of the neuromuscular junction]. C R Acad Sci Hebd Seances Acad Sci D 271(25):2346–2349.

Bero et al. www.pnas.org/cgi/content/short/1408378111 2of6 Fig. S1. Neuronal activity profile of the mPFC upon associative memory encoding. (A) Representative immunohistochemical images depicting expression of the activity-regulated immediate early gene, zif268, in the mPFC 1 h following exposure to foot shock alone (shock), context exploration (context), or con- textual fear conditioning. (Scale bar, 50 μm.) (B) Quantification of A (n = 4–6 mice per group). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. Values represent mean ± SEM.

Bero et al. www.pnas.org/cgi/content/short/1408378111 3of6 Fig. S2. Ultrastructural analysis of individual mPFC synapses upon associative memory encoding. (A–C) Histograms depicting the frequency distribution of active zone length (A), postsynaptic density length (B), and docked synaptic vesicle number (C) of individual mPFC synapses upon control exposure (CTL) or fear conditioning (FC) assessed by transmission electron microscopy (n = 69–92 synapses from four mice per group). Blue and red bars indicate mPFC synapses from CTL and FC groups, respectively. Note the rightward shift in each frequency distribution induced by FC. (D) Active zone length was correlated with postsynaptic density length across treatment groups (n = 69–92 synapses from four mice per group; Pearson r = 0.9533; P ≤ 0.0001). (E) The number of docked synaptic vesicles was correlated with postsynaptic density length across treatment groups (n = 69–92 synapses from four mice per group; Pearson r = 0.7539; P ≤ 0.0001). Blue and red circles indicate synapses in CTL and FC groups, respectively.

Fig. S3. Context exploration is not sufficient to increase synaptophysin expression in the mPFC. Quantification of synaptophysin immunoreactivity in mPFC 1h following exposure to foot shock alone (shock) or context exploration (context). n = 4 mice per group. n.s., not significant. Values represent mean ± SEM.

Bero et al. www.pnas.org/cgi/content/short/1408378111 4of6 Fig. S4. Total synapse density in the mPFC remains unchanged upon associative memory encoding. (A) Representative transmission electron micrographs of the mPFC depicting total synapse density upon CTL or FC. Arrows indicate a subset of synapses. (Scale bar, 500 nm.) (B) Quantification of A (n = 1,010–1,016 synapses from four mice per group). (C) Quantification of total density of Golgi–Cox impregnated spines on mPFC pyramidal neurons upon CTL or FC (n = 716–753 synapses from three to four mice per group). n.s., not significant. Values represent mean ± SEM.

Fig. S5. Neuronal activity profiles of the entorhinal cortex and hippocampal area CA1 upon associative memory encoding. (A and B) Representative im- munohistochemical images depicting expression of the activity-regulated immediate early gene, zif268, in the entorhinal cortex (EC; A) and hippocampal area CA1 (CA1; B) 1 h following exposure to foot shock alone (shock), context exploration (context), or contextual fear conditioning. (Scale bars, 50 μm.) (C and D) Quantification of A and B, respectively (n = 4–6 mice per group). *P ≤ 0.05; **P ≤ 0.01; n.s., not significant. Values represent mean ± SEM.

Bero et al. www.pnas.org/cgi/content/short/1408378111 5of6 Fig. S6. Bilateral stereotaxic injection sites in mice that received optical (593 nm) mPFC inhibition. Gray and blue circles indicate stereotaxic mPFC injection sites in mice that received AAV5-CaMKIIα-eYFP and AAV5-CaMKIIα-eNpHR3.0-eYFP vectors, respectively. Animals that exhibited viral expression outside of the mPFC were excluded from analysis. Values to Right of coronal sections indicate anteroposterior coordinates relative to bregma.

Other Supporting Information Files

Table S1 (DOCX) Table S2 (DOCX) Table S3 (DOCX)

Bero et al. www.pnas.org/cgi/content/short/1408378111 6of6