Experimental designs and procedures

We investigated the mPFC–MBH and mPFC–LH pathways from the two hypothalamic sites by using cholera toxin β subunit (CTB) and Fluorogold (FG), respectively, as retrograde tracers. All surgeries were performed in a stereotaxic frame (Kopf Instruments) under deep ketamine (50 mg/kg; Medicus Partner), xylazine (20 mg/kg; Medicus Partner), and pipolphen (0.2 ml/kg; Egis) anesthesia. After a midline incision, a hand drill was used to expose the brain surface above the target site. Retrograde tracers and viral vectors were microinjected through a glass pipette (the diameter of the tip was 20–30 μm) by using a MicroSyringe Pump Controller System (World Precision Instruments). In Experiment 1, we injected either 50 nl of 1% CTB into the MBH [anteroposterior (AP), −1.6 mm; ML, −1.2 mm; DV, −9.2 mm; CTB, List Biological Laboratories] or 20 nl of 2% Fluorogold into the LH (AP, −2,3 mm; ML, −2,1 mm; DV, −8,5 mm; FG, Fluorochrome). In this experiment, rats did not undergo behavioral testing. Two weeks after the injections, rats were anesthetized and transcardially perfused with 150 ml of ice-cold saline, followed by 300 ml of 4% paraformaldehyde in 0.1 m PBS, pH 7.4. Brains were removed and postfixed in the same fixative overnight, and cryoprotected in 30% sucrose in PBS at 4°C for 2 d before sectioning (30 μm coronal sections) on a sliding microtome. Only animals with injection sites limited to the MBH (n = 4), LH (n = 4), or both in the case of double injection (n = 3) were included in the analysis. CTB-positive and FG-positive cell bodies in the infralimbic cortex (IL) and prelimbic cortex (PrL) were labeled by immunohistochemical methods as follows. A previously described immunohistochemical protocol was used (Biro et al., 2017) with slight modifications as indicated in Table 1. Briefly, after several rinses in PBS, sections (90 μm apart) were incubated in PBS containing 0.3% Triton X-100 (Sigma-Aldrich) and 0.3% H₂O₂ for 30 min. Then, after several washes in PBS, sections were incubated in PBS containing either 5% normal donkey serum or 10% normal goat serum for 90 min, which was followed by incubation in primary antibody solution (in PBS containing 2% normal donkey serum) for 48–72 h at 4°C. Primary antibodies were detected by either biotin- or fluorescent-conjugated IgG antibodies (1:200–1000; Jackson ImmunoResearch Europe) using a 1–2 h incubation time. All specific primary and secondary antibodies and their applications are summarized in Table 1. In case of biotin-conjugated antibodies, labeling was amplified by avidin–biotin complex (1:1000; Vector Laboratories) by incubation for 1 h at room temperature. The peroxidase reaction was developed in the presence of diaminobenzidine tetrahydrochloride (0.2 mg/ml), nickel–ammonium sulfate (0.1%), and hydrogen peroxide (0.003%) dissolved in Tris buffer. Sections were mounted onto gelatin-coated slides, dehydrated, and coverslipped with DPX Mountant (Sigma-Aldrich). Regions of interest were digitalized by an Olympus DP70 Light Microscope and CCD camera system at 60–250× magnification, and analyzed by an experimenter blind to treatments. Sections labeled with fluorescent-conjugated antibodies were mounted on glass slides and coverslipped using Mowiol4–88 fluorescent mounting medium and analyzed using C2 Confocal Laser-Scanning Microscope (Nikon Europe). Two-dimensional overview images (tiles) of mPFC were acquired by 488, 561, and 642 nm lasers. Images were taken of three sections (at +3.7, +3.2, and +2.7 levels from bregma). For the FG/CTB coexpression quantification, single-labeled and double-labeled neurons were counted manually using NIS Elements Software (Nikon Europe; RRID:SCR_014329), while NeuN-positive cell counting was performed using ImageJ Software (version 1.41; National Institutes of Health; RRID:SCR_003070).

Details of the immunohistochemical reagents used in the study

ABC, avidin–biotin complex.

Here we investigated the arborization of mPFC neurons in the LH and MBH, and checked the presence of axon terminals in these hypothalamic fields. The study was performed in those rats, which were used in the behavioral experiments presented below. As anterograde tracers, we used adeno-associated virus (AAV)-mediated gene transfer carrying the gene for channelrhodopsin-2 (ChR2) fused to enhanced yellow fluorescent protein (eYFP). We injected an AAV vector that encoded ChR2 [AAV2.5.hSyn.hChR2(H134R)eYFP.WPRE.hGH; 1.3e13 GC/ml titer; catalog #26973, Addgene; Penn Vector Core] into the mPFC (AP, 2.8 mm; ML, 0.6 mm; DV, 4.4 and 5.0 mm from bregma) by means of a MicroSyringe Pump Controller system (World Precision Instruments). To cover both regions of interest, we first injected 50–50 nl/cc virus into the IL and PrL. Five weeks after injections, rats were perfused under deep anesthesia, and brains were processed for confocal microscopy as described above. The extension of virus expression in the mPFC was investigated in serial sections covering the entire area. The rostrocaudal distance between the analyzed sections was 90 μm. The fluorescence of eYFP was amplified with anti-GFP labeling. For coexpression analysis of eYFP and vesicular GABA transporter (vGAT)/vesicular glutamate transporter 1 (vGLUT1) in the hypothalamus, high-resolution (0.1 μm/pixel) z-stack images (z-step size, 0.5 μm) were taken from 10 μm of the slice. Confocal images were acquired using identical pinhole size, gain level, axial section (z) depth, and laser intensity settings for all hypothalamic slices. NIS Elements Software was used to quantify eYFP/vGLUT1 and eYFP/vGAT double-labeled axon varicosities in the LH and MBH. A 210 μm × 210 μm counting frame was used in both regions on three slices (at −1.80, −1.88, and −2.12 from bregma). The antibodies used in the experiment are presented in Table 1.

This study aimed at verifying whether the photostimulation of mPFC axon terminals at the level of hypothalamus can reliably affect postsynaptic neuronal activity in this area. The above-mentioned virus construct was injected into the mPFC of young male rats on postnatal day 30 (P30) to P35 under deep anesthesia, using the same protocol as described above, except that they were not submitted to behavioral testing. Animals were given 4 weeks to recover and express ChR2 in axon terminals before recording. The brain was removed rapidly and immersed in ice-cold sodium-free solution (in mm: saccharose 205.0, KCl 2.5, NaHCO₃ 26.0, CaCl₂ 1.0, MgCl₂ 5.0, NaH₂PO₄ 1.25, and glucose 10) bubbled with a mixture of 95% O₂ and 5% CO₂. Acute 250-μm-thick coronal slices containing MBH were then prepared with a VT-1000S Vibratome (Leica) in the sodium-free solution. The slices were transferred into artificial CSF (aCSF; in mm: NaCl 130.0, KCl 3.5, NaHCO₃ 26.0, CaCl₂ 2.5, MgSO₄ 1.2, NaH₂PO₄ 1.25, and glucose 10) saturated with O₂/CO₂, and kept in it for 1 h to equilibrate. The initial temperature of aCSF was 33°C, which was left to cool to room temperature during equilibration. Electrophysiological recording, during which the brain slices were oxygenated by bubbling the aCSF with O₂/CO₂ gas, was performed at 33°C. An Axopatch 200B patch-clamp amplifier, a Digidata-1322A Data Acquisition System, and pCLAMP version 10.4 software (Molecular Devices; RRID:SCR_011323) were used for recording. Cells were visualized with a BX51WI IR-DIC Microscope (Olympus) located on a S'Table antivibration table (Supertech). The patch electrodes (outer diameter, 1.5 mm, thin wall; Hilgenberg) were pulled with a Flaming-Brown P-97 puller (Sutter Instrument) and polished with an MF-830 Microforge (Narishige). MBH was identified under microscopic control. Exit of the glass fiber of the 473 nm emission wavelength IKE-473–100-OP Laser (Ikecool) was set onto the surface of the brain slice, at the hypothalamic attack area of the MBH. Then a neuron was patch clamped in the close vicinity (in 200–300 μm) of the end of the glass fiber. Whole-cell patch-clamp measurements were performed to record postsynaptic currents (PSCs). The neurons were voltage clamped at a pipette holding potential of −70 mV. Pipette resistance was 1–2 MΩ, and resistance of the gigaseal was 2–3 GΩ. The pipette solution contained the following (in mm): K-gluconate 130, NaCl 10, KCl 10, MgCl₂ 0.1, HEPES 10, EGTA 1, Mg-ATP 4, and Na-GTP 0.3 (pH 7.3 with KOH). Osmolarity was adjusted to 295–300 mOsm with sorbitol. Duration of a laser pulse was 10 ms with 2.5 mW power. A train of laser pulses was applied at 0.2 Hz (60 pulses totally). For analysis, records of the 60 pulses were averaged. In total, analysis contained 12 recorded cells from three animals. Recordings were stored and analyzed off-line using the Clampfit module of the PClamp version 10.4 software (Molecular Devices; RRID:SCR_011323).

Here we investigated the behavioral effects of axonal photostimulation in the MBH and LH, respectively. The two experiments were highly similar except for the target of the stimulation (Experiment 4, MBH; Experiment 5, LH). Viral infection of the mPFC was performed as described for Experiment 2. Two weeks later, rats were implanted with an optic fiber (diameter, 250 μm; flat tip) either in the MBH (AP, 1.8 mm; ML, −1 mm; DV, −8.8 mm) or LH (AP, 2.3 mm; ML, 2 mm; DV, 7.4 mm). We used custom optic fibers (AFS 105/125Y; numerical aperture, 0.22; low OH) that were obtained from Thorlabs; light transmission through the optic fiber was checked for each rat. Implants were secured to the skull by screws and acrylic resin (Duracryl Plus, SpofaDental). Behavioral testing started 2 weeks after optic fiber implantation. Virus injection sites and optic fiber locations were verified using immunohistochemistry as described above. No rat was excluded from the study except for those where virus expression extended beyond the mPFC or where the tip of the optic fiber was outside the target hypothalamic areas (n = 8 in the case of MBH; n = 10 in the case of LH).

Territorial aggressive behavior was studied in the RI test performed in Plexiglas cages measuring 40 × 25 × 25 cm. Three days before the RI test, resident animals were transferred into test cages designed for photostimulation. The 10-min-long resident/intruder test started with the placement of a smaller conspecific (300–350 g) into the test cage. Concomitantly, photostimulation was initiated and maintained for 3 min (473 nm light delivery, 20 mW output power, 20 Hz with 10 ms pulses). The same rats underwent the test four times in 2 d intervals. Photostimulation was applied in two of the trials. In the other two trials, rats were connected to optic fibers, but no light was delivered. Stimulation and no-stimulation trials followed according to a randomized crossover design. The test was performed in the early phase of the dark period under dim red illumination provided by two 40 W red bulbs placed on the ceiling of the experimental room. Behavior was video recorded, and biting attacks were later analyzed in detail by an experimenter who was blind to treatment conditions.

Behavioral analysis focused on the patterns of biting attacks, namely on attack targeting and the relationship between offensive threats and attacks (Tóth et al., 2008). Attack episodes were analyzed at low speed (frame-by-frame if necessary) for identifying attack targets and the context of attack. An attack was considered vulnerable if it targeted the head (areas anterior to the ears), the throat (the ventral area below the ears), the belly (ventral areas between legs), or the paws. Dorsal and lateral areas (posterior to the ears and dorsal to the legs) were considered nonvulnerable targets. An attack was considered not signaled if it was not preceded by an offensive threat, and signaled if it was performed in the context of an offensive threat. Both vulnerable and nonsignaled attacks, respectively, were expressed as the percentage of total attacks according to the following formulas: vulnerable attack counts/total attack counts × 100; and not signaled attack counts/total attack counts × 100. The frequency and duration of the following behavioral variables were also assessed: exploration (sniffing directed toward the environment); social investigation (sniffing directed toward the opponent's flank, nasal, or anogenital region); grooming (self-grooming with forepaws and scratching with hindlegs); offense (aggressive grooming, lateral threat, offensive upright posture, mounting, and chasing taken together); defense (defensive upright, defensive kick, fleeing, and freezing taken together); dominant posture (keeping down the opponent while he is laying on his back); and submissive posture (laying on back while being kept down by the opponent).

This study aimed at investigating the ChR2 specificity of behavioral changes induced by photostimulation. All experimental conditions were the same, and rats were prepared as described for Experiments 4 and 5, with the exception that the virus vector did not contain the gene for ChR2. We used the adeno-associated virus vector AAV-hSyn-EYFP (titer, 3 × 10¹²; Addgene; University of North Carolina, Chapel Hill, NC).

The experiment was performed to check possible confounds resulting from the backward propagation of hypothalamic excitation into the mPFC. Rats were prepared as described for Experiments 4 and 5, and were stimulated in their home cage under resting conditions. The stimulation protocol was the one described above. Stimulation was followed by transcardial perfusion for c-Fos immunohistochemistry (see above and Table 1). c-Fos-positive cell numbers were counted in mPFC to measure potential retrograde propagation-induced neuronal activation.

This experiment was performed to verify whether the effects on aggression were explained by changes in sociability. Rats used in Experiment 4 were submitted to a modified three-chamber sociability test 2 d after the last resident/intruder test. Subjects were placed in an open field arena measuring 100 × 100 × 40 cm. After a 5-min-long habituation period, when animals could explore the arena freely, a smaller male conspecific (age, 8–9 weeks; weight, 300–350 g) was presented in a plastic perforated cylinder in the corner of the arena, whereas an empty cylinder was placed in another corner. Immediately after placing the cylinders, the photostimulation protocol described for the resident/intruder tests was initiated. We recorded the total distance covered by means of Noldus EthoVision software (locomotor activity) and the time spent with investigating cylinders by means of an H77 event-recorder software. Social motivation was evaluated by means of the following formula: 100 × the time spent investigating conspecific cylinder/(time investigating conspecific cylinder + time investigating empty cylinder).