All experiments were conducted in accordance with the Institutional Animal Care and Use Committee of the University of California, Irvine, Stanford University, the University of Arkansas for Medical Sciences, and Loma Linda University, as well as according to Hungarian Act of Animal Care and Experimentation (1998, XXVIII, Section 243/1998), which are in accordance with the European Communities Council Directive of November 24, 1986 (86/609/EEC; Section 243/1998).
To target PVINs for patch-clamp recordings, a PV-Cre line (The Jackson Laboratory stock # 008069) was crossed with a reporter line (The Jackson Laboratory stock # 007905) to produce mice expressing the red fluorescent protein tdTomato in PV + cells (PV-TOM mice). C57BL/6J mice (The Jackson Laboratory stock # 000664) were used in all other experiments.
Male C57BL/6J or PV-TOM mice at 2–3 months of age were exposed to whole body 150 MeV/n proton irradiation, and the irradiated group received 0.502 ± 0.004 Gy proton irradiation at a dose rate of 0.75 ± 0.06 Gy/min. Irradiation was performed at the James M. Slater Proton Therapy Treatment and Research Center (Loma Linda University Medical Center, Loma Linda, CA). Due to scattering foils and other beamline components, the proton energy at the target surface was reduced to 126 MeV and was associated with an LET of 0.62 keV/μm. Control mice were identically treated except that they were not exposed to proton irradiation.
In vitro electrophysiology
Coronal hippocampal slices (300 μm) were prepared from male PV-TOM mice or C57BL/6J mice 5–9 weeks after irradiation or sham treatment. Slices were incubated in sucrose-containing artificial CSF (ACSF) for an hour. ACSF contained, in mM (85 NaCl, 75 sucrose, 2.5 KCl, 25 glucose, 1.25 NaH2PO4, 4 MgCl2, 0.5 CaCl2, and 24 NaHCO3). After the initial incubation period, slices were transferred in the same ACSF solution used for recordings, which contained, in mM (126 NaCl, 2.5 KCl, 26 NaHCO3, 2 CaCl2, 2 MgCl2, 1.25 NaH2PO4, and 10 glucose). Patch pipettes had resistances of 3–5 MΩ. Signals were filtered at 3 kHz using a Bessel filter and digitized at 10 kHz with a Digidata 1440A analog–digital interface (Molecular Devices). Series resistances were carefully monitored, and recordings were discarded if the series resistance changed >20% or reached 20 MΩ. The recorded traces were analyzed using Clampfit 10.5 (Molecular Devices). Slices were visualized with an upright microscope (Olympus; BX61WI) with infrared–differential interference contrast (IR-DIC) optics. These microscopes were additionally equipped with a mercury lamp light source for epifluorescence. All electrophysiological recordings were made at 33 °C using a MultiClamp700B amplifier (Molecular Devices).
To examine interneuron to PC connections, whole-cell recordings in current-clamp configuration were obtained from the CB1BCs or PVINs with Vm adjusted to −60 mV. The interneuronal internal solution contained (in mM) 126 K-gluconate, 4 KCl, 10 HEPES, 4 ATP-Mg, 0.3 GTP-Na, 10 phosphocreatine, as well as 0.2% biocytin, with a pH of 7.2, and osmolarity of 290 mOsm. Superficial PCs were recorded in voltage-clamp at a holding potential of −70 mV. The PC internal solution contained (in mM) 40 CsCl, 90 K-gluconate, 1.8 NaCl, 1.7 MgCl2, 3.5 KCl, 0.05 EGTA, 10 HEPES, 2 Mg-ATP, 0.4 Na2GTP, and 10 phosphocreatine as well as 0.2% biocytin, with a pH 7.2, and osmolarity of 290 mOsm. Action potentials in presynaptic interneurons were induced in current clamp by injecting 2 ms square pulses of 2nA. In a subset of pairs, we tested the PC to interneuron connections as well (PC, voltage-clamp, holding potential, −70 mV; interneurons, voltage-clamp, holding potential, −60 mV). Action currents in presynaptic PCs were induced in voltage-clamp by injecting 2 ms square pulses from −70 to 30 mV.
To examine the intrinsic properties of CB1BCs or PVINs, we recorded interneurons using an internal solution containing (in mM) 126 K-gluconate, 4 KCl, 10 HEPES, 4 m ATP-Mg, 0.3 GTP-Na, 10 phosphocreatine, as well as 0.2% biocytin, with a pH 7.2, and osmolarity of 290 mOsm. Interneurons were held at their resting membrane potentials. Action potential discharges were evoked by current injections (1 s-long step currents, from 0 to +500 pA with 50 pA increments). To determine the input resistance of interneurons, they were held at their resting membrane potentials and voltage responses to small current pulses (1 s-long current steps from −100 to +100 pA with 50 pA increments) were measured at steady state (0.8–1.0 s from the start of 1 s-long current steps).
For recordings from presynaptic CB1BCs, multipolar neurons located in the striatum radiatum were targeted using infrared differential interference contrast microscopy and filled with biocytin for post hoc cell identification. All putative CB1+ interneurons were identified post hoc as CB1BCs by their characteristic axonal arborization and histological immunopositivity for CB1 (see below for details). PVINs were targeted based on tdTomato fluorescence. The somata of all recorded CA1 PCs were located within the superficial sublayer of the stratum pyramidale (distance from the pyramidale-radiatum border: 0–20 μm, dorsal hippocampus; 0–50 µm, ventral hippocampus; see Lee et al. 2014 for the rationale and discussion of the superficial sublayer).
Reconstruction of CB1BCs and analysis of their axonal/dendritic morphology
CB1BCs were filled with biocytin in hippocampal slices from C57BL/6J mice a mean of 8 weeks (range 7–9 weeks) post-irradiation. Slices were resectioned into 40–70 μm thin sections. Cells not processed for STORM imaging, were processed for CB1 immunopositivity (CB1-GP-Af530-1; 1:5000, guinea pig; Frontiers Science). A secondary antibody conjugated to Alexa Fluor 488, raised in donkey against guinea pig (Invitrogen), was used to detect the location of the primary antibody. Biocytin was visualized DyLight594-labeled streptavidin (Jackson, 1:1000). See below for STORM imaging of CB1. Confocal z-stacks containing the entire recovered arborization of the filled neurons were collected on a C2 confocal system (Nikon). For the visualization of representative cells, two cells were manually reconstructed in Neurolucida software (MBF Bioscience). For the analysis of dendritic morphology, image stacks were processed with ImageJ software (NIH), using identical parameters across all images. All visible branches of the dendrites were then manually reconstructed in 3D using Neuronstudio software (Rodriguez et al. 2006). Sholl analysis was performed in Neuronstudio to measure the length of dendrites traversing a concentric series of spherical shells with 1 μm increments in radius. Then, the measured dendritic lengths were averaged to form bins with 25 μm increments. Cells were pooled by treatment group, and the differences in length values in each bin were tested using a two-sample Kolmogorov–Smirnov test. For the quantitative analysis of axonal morphology, the bouton distribution index (BDI) was calculated for each cell as described previously (Dudok et al. 2015). Briefly, the positions of axonal varicosities were recorded as a distance from the borders of the pyramidal layer. The laminar distribution of the varicosities was then visualized as the histogram of relative distances (where 0 is the center and 1 is the thickness of the pyramidal layer), and the BDI was calculated from the descriptive statistics of the distribution. Cells with BDI >1 were included in the study as perisomatically targeting interneurons, while all other cells were excluded.
Stochastic optical reconstruction microscopy (STORM) super-resolution imaging
After imaging the developed cells as described above, immunostaining, correlated confocal and STORM microcopy, and image analysis were carried out according to the previously described protocol (Barna et al. 2016). Slices containing filled cells were embedded in agarose and resectioned to 20 micron thickness. Immunostaining was performed using a previously validated primary raised in Guinea pig against the C terminus of CB1 (1:1000 in TBS) (Fukudome et al. 2004; Dudok et al. 2015) and Alexa 647-conjugated secondary antibodies (Jackson ImmunoResearch, 2 µg/mL in TBS). Before imaging, sections were covered with Dulbecco’s phosphate buffered saline containing 5% (wt/vol) glucose, 0.1 M 2-Mercaptoethylamine (Sigma), 1 mg/mL glucose oxidase (Sigma), and 1500 U/mL catalase (Sigma). Images were collected using Nikon C2 confocal microscope. Then, astigmatic 3D-STORM (Huang et al. 2007) images of the immunostaining were recorded in continuous dSTORM mode (Heilemann et al. 2008) for 5000 frames at 31 Hz. Confocal image stacks were deconvolved using Huygens software (SVI). STORM images were processed for peak detection using the N-STORM module in NIS-Elements software (Nikon). The average lateral localization precision was 9.6 ± 0.9 nm. Correlated pairs of confocal images and molecule lists were loaded in VividSTORM software (Barna et al. 2016), aligned, and regions of interests (ROI) containing labeled axon terminals were defined using unbiased active contour algorithm. The perimeter of the ROI, as well as the STORM NLP within the ROI, were determined for each axon terminal. Density values were calculated as NLP over perimeter. As the Kruskal–Wallis test detected significant difference between the cells within groups, cells were not pooled but the average values were calculated from 12 ± 3 axon terminals per cell. These values were then compared between treatment groups using Mann–Whitney U test. The internalization of CB1 was assessed by calculating the internalization index as reported earlier, where this measure could readily detect increased CB1 internalization upon in vivo THC exposure (Dudok et al. 2015). All images were recorded, processed and analyzed with identical settings.
Liquid chromatography/mass spectrometry analyses
Whole hippocampi were collected from C57BL/6J mice 7 weeks after irradiation, quickly frozen on dry ice, and stored at −70 °C until the lipid analyses. Amounts of 2-AG in the dissected hippocampi were determined as described (Astarita and Piomelli 2009). Briefly, frozen tissue samples were homogenized in cold methanol (1 ml) containing 2-arachidonyl glycerol-d8 (2-AG-d8; Cayman Chemical, Ann Arbor, MI) as an internal standard. Protein concentration was determined in the homogenate to normalize samples using the bicinchinonic acid (BCA) protein assay (Pierce, Rockford, IL, USA). Lipids were extracted by adding chloroform and water (2:1) and fractionated through open-bed silica gel columns (60-Å 230–400 Mesh ASTM; Whatman, Clifton, NJ) by elution with 1 ml of chloroform/methanol (9:1). Eluates were dried under N2, and reconstituted in chloroform/methanol (1:3 μl).
We used an Agilent 1100-LC system coupled to a 1946A-MS detector equipped with an electrospray ionization interface (Agilent Technologies, Inc., Palo Alto, CA). Lipids were separated on a reversed-phase XDB Eclipse C18 column (50 × 4.6 mm i.d., 1.8 μm, Zorbax, Agilent Technologies). They were eluted with a gradient of methanol in water (from 85 to 90% methanol in 2.0 min and 90–100% in 3.0 min) at a flow rate of 1.5 ml/min. Column temperature was kept at 40 °C. Mass spectrometry detection was in the positive ionization mode, capillary voltage was set at 3 kV and fragmentor voltage was 120 V. N2 was used as drying gas at a flow rate of 13 l/min and a temperature of 350 °C. Nebulizer pressure was set at 60 PSI. For quantification purposes, we monitored the sodium adducts of the molecular ions [M+Na]+ in the selected ion-monitoring mode, using 2-AG-d8 (mass-to-charge ratio for 2-AG-d8: m/z = 409) as internal standards.
Paired or unpaired two-tailed Student’s t tests were used when the data showed a normal distribution; otherwise Wilcoxon’s signed rank (paired data) or Mann–Whitney tests (unpaired data). Pearson’s Chi-squared tests were used for the connection probability. Other statistical tests are noted in the text. Data are presented as mean ± SEM. A p value <0.05 was considered significant. Statistical analyses were performed using Origin Pro 2015 (OriginLab Corporation, Northampton, MA, USA), STATISTICA 10 (Dell Statistica, Tulsa, OK, USA), and GraphPad QuickCalcs (GraphPad Software, La Jolla, CA, USA).