METHOD DETAILS

The constructs AAV-FLIM-AKAR, AAV-FLEX-FLIM-AKAR, AAV-FLEX-FLIM-AKAR^T391A and AAV-FLEX-PKIα-IRES-mRuby2 were described in (Chen et al., 2014) (Addgene #s 63058, 60445, 60466 and 63059). pCMV-SPORT6-adcy2 was an IMAGE clone from Invitrogen (IMAGE: 5367175 (Lennon et al., 1996)). pcDNA3.1-human M1R was from cDNA Resource Center (www.cdna.org). PKAc-mEGFP (pCAGGSS-mouse PKA-Calpha-mEGFP) was a gift from Haining Zhong (Addgene # 45528) (Zhong et al., 2009). pBS-β-actin-Cre was a gift from Susan Dymecki at Harvard Medical School. The Gq-DREADD hM3Dq construct (AAV-DIO-hM3Dq-mCherry) was a gift from Brad Lowell at Beth Israel Deaconess Medical Center (Armbruster et al., 2007; Krashes et al., 2011). The AAV-Cre-mCherry construct was from Matthew During at Ohio State University. AAV-DFI-Gcamp3.8 was cloned by replacing ChR2-mCherry in AAV-EF1α-DFI-ChR2-mCherry (Cardin et al., 2009) with Gcamp3.8.

For the construction of the mCherry-tagged PKA consensus substrate, the mEGFP and sReaCh regions of AAV-FLEX-FLIM-AKAR were each replaced by mCherry through gene synthesis and subcloning via ApaI and BglII sites. Subsequently, site-directed mutagenesis was performed to mutate the threonine to alanine at the phosphorylation site of the consensus substrate (Genscript).

In utero electroporation was used to deliver plasmids as described previously (Chen et al., 2014). During electroporation, the embryo head was held with a tweezertrode (5mm electrode diameter, Harvard Apparatus) and electric pulses were delivered five times (50 V, 50 ms pulse with 950ms interval) for hippocampus delivery of DNA (CUY21 electroporator, NEPA GENE, Japan).

For GCaMP3 imaging, AAV1-Cre-mCherry and AAV8-DFI-Gcamp3.8 were packaged at University of North Carolina Gene Therapy Center Virus Core Facility. For Figures 4B–D, P16–P21 pups with the genotype GCaMP3^f/f were anesthetized with isofluorane and placed on a small stereotaxic frame (David Kopf Instruments). To target the hippocampus, coordinates of posterior 2.8 mm and lateral 3.0 mm relative to Bregma, 2.2 mm from the pia were used. Unilateral injections of 1 μl of AAV1-Cre-mCherry (2×10¹² genome copy/ml) were made into the right hemispheres at a rate of 100 nl/min through a UMP3 micro syringe pump (World Precision Instruments). After injection, pups were returned to their home cage with their mother, weaned around P21–P23, and expression was allowed to occur for at least 11 days post infection. For Figures S2C–D, AAV1-Cre-mCherry (2×10¹² genome copy/ml) and AAV-DFI-Gcamp3.8 (1.39X10¹² virus molecules/ml) were injected into P0/P1 hippocampus according to previously published procedures (Lu et al., 2009), and mice were imaged at P15–P19.

Mice were anesthetized with isoflurane before being sacrificed. Their brains were rapidly dissected out.

For all the experiments except for those in Figures S5 and S8G–H, acute horizontal sections were sliced from the hippocampus with a Leica VT1000S vibratome (Leica Instruments) in cold sucrose cutting solution (containing in mM: 87 NaCl, 25 NaHCO3, 1.25 NaH2PO4 2.5 KCl, 75 sucrose, 25 glucose, 7.5 MgCl2). Slices were 300 μm thick. After sectioning, slices were transferred to ACSF (containing in mM: 127 NaCl, 2.5 KCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 2 CaCl₂, 1 MgCl₂, and 25 glucose; for nominal 0 Ca²⁺ experiments, ACSF contained no CaCl₂ and 3 MgCl₂). The slices were incubated for recovery at 34°C for 5–10 minutes (imaging experiments) or 30 minutes (electrophysiology recordings), and then kept in ACSF at room temperature. Slices were then transferred to a microscope chamber and ACSF was perfused at a flow rate of 2–4 ml/min.

For the experiments in Figure S5, acute visual cortical slices (300 μm) were prepared as described (Seol et al., 2007). Briefly, slices were sectioned in ice-cold cutting buffer (containing in mM: 212.7 sucrose, 5 KCl, 1.25 NaH₂PO₄, 10 MgCl₂, 0.5 CaCl₂, 26 NaHCO₃, 10 dextrose). The slices were transferred to ACSF (containing in mM: 124 NaCl, 5 KCl, 1.25 NaH₂PO₄, 1 MgCl₂, 2 CaCl₂, 26 NaHCO₃, 10 dextrose) for recovery at 30°C for 30 minutes then at room temperature for at least 30 minutes prior to recording.

All solutions were continuously bubbled with carbogen (95% O2, 5% CO2), and the experiments were performed at 30°C–34 °C.

For images in Figure 1B, mice were deeply anesthetized at P30 with isoflurane and perfused transcardially with 4% paraformaldehyde in phosphate buffered saline (PBS). Brains were postfixed overnight, washed in PBS and 40 μm parasagittal sections were cut with a Leica VT1000S vibratome (Leica Instruments). They were then mounted on superfrost slides, dried and covered with ProLong antifade reagent containing DAPI (Molecular Probes) followed by a coverslip. Whole sections were imaged with an Olympus VS120 slide-scanning microscope.

Two photon imaging was achieved by a custom-built microscope with a mode-locked Ti-sappire laser source (Carter and Sabatini, 2004; Chen et al., 2014) (Chameleon Vision II, 80 MHz, Coherent). Photons were collected with fast photomultipler tubes (PMTs) (H7422-40MOD, Hamamatsu). A 60X (NA1.1) objective (Olympus) was used. Image acquisition was performed using a custom-written software ScanImage that ran in Matlab (Chen et al., 2014; Pologruto et al., 2003).

For Ca²⁺ imaging, 910nm was used as the excitation wavelength, and 128×128 pixel images were collected by frame scan at 4Hz. CA1 neurons with basal GCaMP3 signals excluded from the nucleus were used for experiments since neurons with labelled nuclei were previously shown to have impaired calcium homeostasis and GCaMP3 function (Tian et al., 2009). At the end of all Ca²⁺ imaging experiments, 50mM KCl was applied to activate voltage-gated calcium channels as a positive control for cell health.

FLIM was performed as described previously (Chen et al., 2014). The FLIM board SPC-150 (Becker and Hickl GmbH) was used, and time-domain single photon counting was performed in 256 time channels. 920nm excitation wavelength was used to excite the donor fluorophore mEGFP in FLIM-AKAR.

Figure 1F was constructed from the maximum projections along the z dimension of two regions of interests acquired with a two-photon microscope, at 3 frames per z slice and 1 μm per z step. All experiments were performed in the presence of 1 μM DPCPX to inhibit adenosine receptors, and 1 μM tetrodotoxin to block action potentials.

Each region of interest (ROI) corresponding to a single neuronal cell body was manually selected. The fluorescence signal for all pixels in a given ROI was averaged and plotted against time. The ΔF/F₀ was calculated as (F−F₀)/F₀, where F₀ is the fluorescence signal averaged over the entire baseline period (usually 2.5–3 minutes).

Fluorescence lifetime curve fitting and the calculation of average lifetime over a particular region of interest were performed as described previously (Chen et al., 2014; Harvey et al., 2008; Yasuda et al., 2006). Instrument response curve (IRF) was measured with double harmonic generation of urea crystals and used to deconvolve the fluorescence decay curve. The time constant for the free donor lifetime (τ_free) was determined by transfecting the donor alone into HEK293T cells, and was determined to be 2.14ns. To determine the time constant for donors that have undergone FRET (τ_FRET), FLIM-AKAR was transfected into HEK 293T cells, and the best double exponential fit was performed with τ_free fixed at 2.14ns. τ_FRET was determined to be 0.69ns. These values of τ_free and τ_FRET were then used for double exponential fitting of lifetime distribution curves during all experiments.

Only images with photon count rates between 40,000 photons per second (40KHz) and 1.3MHz were used. The lower limit ensures accurate lifetime estimation based on our simulation, and gives a standard deviation of lifetime estimate of 0.0035ns. The upper limit ensures that we do not run into dead time and pile-up issues of the FLIM board.

For ROI analysis, somatic cytoplasm, nucleus and dendrite were segmented via a semiautomated software written in Matlab that utilizes both intensity and lifetime data. The image segmentation was visually inspected and revised by the experimenter post automation.

The amplitudes of lifetime changes were quantitated as follows (Figure S1A):

BaselineStart = lifetime measurements averaged over the first minute of baseline;

BaselineEnd = lifetime measurements averaged over the last minute of baseline;

BaselineMin = minimum lifetime measurement during baseline;

TreatmentMin = minimum lifetime measurement after a particular drug flow-in, before the next drug flow in.

Δlifetime (baseline) = BaselineMin - BaselineStart;

Δlifetime (Treatment) = TreatmentMin - BaselineEnd.

For Figure S3B, paired whole-cell voltage clamp recordings were taken from transfected hippocampal CA1 pyramidal neurons, identified by mRuby2 epifluorescence for PKIα-IRES-mRuby2 expressing cells, and neighboring untransfected control cells. Slices were perfused in ACSF containing 100 μM picrotoxin. Results from two conditions (10 μM NBQX or 10 μM (R)-CPP) were pooled for analysis. Whole-cell access to recorded neurons was made using 3–5 MΩ glass pipettes filled with internal solution containing (in mM): 135 CsMeSO4, 8 NaCl, 10 HEPES, 0.3 EGTA, 5 QX-314, 4 Mg-ATP, 0.3 Na-GTP, 0.1 spermine, at 293 mOsm and pH 7.24. Voltage-clamp was performed using a Multiclamp 700B amplifier (Molecular Devices) with a 3 KHz Bessel filter, digitized at 10 KHz using a National Instruments data acquisition board, and recorded and analyzed using ScanImage. Input resistance and capacitance were calculated following whole-cell break-in by fitting a 5 mV test pulse with an exponential decay, and holding current measured when neurons were voltage-clamped to −70 mV.

For Figure S5, visualized whole-cell current-clamp recordings were made from layer II/III regular-spiking pyramidal cells using MultiClamp 700A amplifier (Molecular Devices). Borosillicate glass recording pipettes (4–6 MΩ) were filled with intracellular solution containing (in mM): 130 K-Gluconate, 10 KCl, 0.2 EGTA, 10 HEPES, 4 Mg-ATP, 0.5 Na-GTP, 10 Na-Phosphocreatine, with or without 0.01 myristoylated PKI 14–22 amide, at 280–290mOsm and pH 7.25. Only cells with membrane potentials more negative than −65 mV, series resistance <20 MΩ (8–18 MΩ, compensated at 80%), and input resistance larger than 100 MΩ were studied. Cells were excluded if input resistance changed > 15% over the entire experiment, with the exception of changes during bath application of the agonists. Data were filtered at 2 kHz and digitized at 5 kHz using Igor Pro (WaveMetrics Inc. Lake Oswego, Oregon). Synaptic responses were evoked every 15 seconds by stimulating layer IV with 0.2 ms pulses delivered through theta glass pipettes filled with ACSF. Intensity was adjusted to evoke a 4–6 mV response. Synaptic strength was quantified as the initial slope (the first 2 ms) of the EPSP. One cell per slice was used.

Unless otherwise noted, all chemicals were applied via bath perfusion: they were either spiked into the perfusion reservoir, or pre-made buffers with the specified chemical concentrations were switched from one to another via a custom-made solution exchanger. Lifetime was allowed to stabilize before a new chemical was added; when there was no clear lifetime changes, 10 minutes were given before the addition of another chemical. Whenever possible, perturbation experiments were performed with interleaved brain slices, and the perfusion tubing was washed and changed for different drug conditions. The final concentrations of chemicals are specified in brackets: acetylcholine (100 μM) was from Sigma; (+)-muscarine-iodide (10 μM), forskolin (50 μM), isoproterenol (1 μM), clozapine N-oxide (10 μM), scopolamine hydrobromide (10 μM), cyclopiazonic acid (30 μM, pre-incubation at room temperature for at least 35 minutes), phorbol 12, 13-dibutyrate (1 μM), GF 109203X (2 μM, and pre-incubation at room temperature for at least 35 minutes), (S)-3,5-DHPG (50 μM), DPCPX (1 μM), picrotoxin (100 μM), NBQX (10 μM) and (R)-CPP (10 μM), PKI 14–22 amide, myristoylated (10 μM, included in the internal solution in the patch pipette) were from Tocris Bioscience; tetrodotoxin citrate (1 μM) was from Abcam and Tocris; YM-254890 (1 μM, and pre-incubation at 32°C for at least 10 minutes) was from Wako; and 77-LH-28-1 (10 μM) was a gift from Eli Lilly and company.

For ACh puffing experiments, ACh (200 μM in ACSF) was puffed onto an apical or basal dendrite from a glass patch pipette via a picospritzer (Parker) at 2 psi, and approximately 30 μm from the dendrite.