Esr1+ hypothalamic-habenula neurons shape aversive states

Excitatory projections from the lateral hypothalamic area (LHA) to the lateral habenula (LHb) drive aversive responses. We used patch-sequencing (Patch-seq) guided multimodal classification to define the structural and functional heterogeneity of the LHA–LHb pathway. Our classification identified six glutamatergic neuron types with unique electrophysiological properties, molecular profiles and projection patterns. We found that genetically defined LHA–LHb neurons signal distinct aspects of emotional or naturalistic behaviors, such as estrogen receptor 1-expressing (Esr1+) LHA–LHb neurons induce aversion, whereas neuropeptide Y-expressing (Npy+) LHA–LHb neurons control rearing behavior. Repeated optogenetic drive of Esr1+ LHA–LHb neurons induces a behaviorally persistent aversive state, and large-scale recordings showed a region-specific neural representation of the aversive signals in the prelimbic region of the prefrontal cortex. We further found that exposure to unpredictable mild shocks induced a sex-specific sensitivity to develop a stress state in female mice, which was associated with a specific shift in the intrinsic properties of bursting-type Esr1+ LHA–LHb neurons. In summary, we describe the diversity of LHA–LHb neuron types and provide evidence for the role of Esr1+ neurons in aversion and sexually dimorphic stress sensitivity.


State induction test by optogenetic stimulation
We adapted a conditioned place aversion (CPA) assay to test the presence and persistence of a behavioral state induced by LHA-LHb pathway stimulation. Mice were placed in a custom-made two compartment behavioral arena (50 × 25 × 25 cm black plexiglass, paired compartment walls had white on black stripes) and the behavior was recorded as in the real time place aversion. The first day optical fibers were connected to the animal but there was no light stimulation. Immediately after the habituation session, the animal was forced to spend time in the compartments paired with light stimulation (40 Hz, 5 ms pulse, 1 second on, one second off, 473 nm laser) for 10 minutes (state induction, Figure 3). Immediately after the state induction session, the animal was free to explore both compartments for 10 minutes (immediate state). The second day (24 h post induction) the animal was placed back to the same arena free to explore both compartments for 10 minutes (sustained state). For the habituation session, the immediate state test and sustained state test the optical fibers were connected to the animal but there was no light stimulation. We manually labeled the base of the tail, left and right hindlimbs and forelimbs in 1000 frames, sampled from all sessions. We based our movement analysis on the tail base point tracked by DLC. After running DLC on every video, we transformed the video coordinates (in pixels) to world coordinates (in cm) using a perspective transform matching the four corners of the box. During the 10 minutes state induction, discrete events of stop-backwards, sharp turns, digging and free rearing (number of events of standing on the hind limbs, far from the arena wall) where manually scored for each second during the stimulated and unstimulated epochs.

Open field test
Mice were placed in a custom-made open field (49×49 cm black plexiglass) for 10 minutes (TeLC silencing experiment, Extended Data Fig. 6ab) or 20 minutes (optogenetic stimulation experiment, Extended Data Fig.  6r). The behavioral arena was placed on a transparent plexiglass. The animal behavior was recorded with a camera placed below the arena. For the optogenetic experiment, the mouse performance was evaluated under alternating laser off/on epochs, starting with 5 minutes off (stimulated epoch: 40 Hz, 5 ms pulse, 1 second on 500 ms off for 5 minutes, 473 nm laser). We manually labeled the base of the tail, left and right hindlimbs and forelimbs in 1000 frames, sampled from all sessions. After running DLC on every video, we transformed the video coordinates (in pixels) to world coordinates (in cm) using a perspective transform matching the four corners of the box. The speed (cm/s) and stationary time were analyzed on the base of the tail tracked by DLC. For the optogenetic experiment, discrete events of wall rearing (number of events of standing on the hind limbs, with forepaws on the arena wall), free rearing (number of events of standing on the hind limbs, far from the arena wall) and discrete grooming events (number of events of mice in sitting position with licking of the fur, grooming with the forepaws, or scratching with any limb) where manually scored for each second during the stimulated and unstimulated epochs (See Extended Data Video 2).

Free-Access Caloric Consumption Assay
Mice were food-restricted to 85 to 90% of their initial body weight by administering one daily feeding of ∼2.5 to 3.0 g of standard grain-based chow (immediately following behavioral experiment, if performed). Water was provided ad libitum. All feeding-related behavioral experiments were conducted at the same time in the middle of the animals' dark cycle (at approximately 14:00). Food restricted mice were placed in a custom made 15 × 15 × 20 cm operant chamber with free access to a bottle containing a 15% sucrose reward for 40 min. Each lick (lick response) was detected, and reward was delivered (rewarded lick, 3 µL reward) with a 1 second timeout. Mice were habituated to the operant chamber and connected to the fibers with no light stimulation. Once stable licking was achieved with light-off sessions (three consecutive days of daily average lick responses within ± 10%), sucrose consumption was monitored for two consecutive days of light-off and two consecutive days of light-on stimulation (40 Hz, 5 ms pulse, 1 second on, one second off, 473 nm laser).

Quantification of cFOS in LHb
In order to evaluate the recruitment of LHb neurons we performed a cFOS IHC and quantified the cFOS+ neurons with the lateral habenula area, after confocal imaging of the LHb throughout the anterior-posterior axes. For optogenetic experiments (Extended Data Fig. 6o-p), mice were placed in a custom-made open field (49×49 cm black plexiglass) where they received 10 minutes of simulation protocol as in state induction test (40 Hz, 5 ms pulse, 1 second on, one second off, 473 nm laser). Mice were sacrificed 30 minutes after the start of the stimulation. For optogenetic experiments, the control group was implanted with optic fibers, placed in the same open field for 10 minutes and optical fibers were connected to the animal but there was no light stimulation. For TeLC experiments, mice were placed in a sound isolated fear-conditioning chamber where they received 5 inescapable, uncontrollable electric foot shocks, at 0.3 mA over 10 minutes with random shock duration ranging from 1 to 3 seconds and unpredictable inter-shock intervals (ITIs), from 1 to 15 seconds. Mice were sacrificed 30 minutes after the start of the foot shock protocol.

Stress induction protocol
In order to induce stress-like state in mice for electrophysiological recordings and behavioral tests, mice were exposed to a stress induction 'training protocol' for 3 days. Stress induction protocol was modified to 'mild foot shock' from what previously described in order to maximize the chance of detecting sexually dimorphic effect and avoiding ceiling effect. Mice were placed in a sound isolated fear-conditioning chamber where they received 360 inescapable, uncontrollable electric foot shocks at 0.3 mA over 1hour with random shock duration ranging from 1 to 3 seconds and unpredictable inter-shock intervals (ITIs), from 1 to 15 seconds. When possible, experiments were performed on pairs of littermates previously housed in the same cage. Control animals were placed in the shocking chamber for 1 h, without being shocked. 24 h after the last shocking protocol mice were assigned to slice electrophysiology or behavioral phenotyping. Animal identity was blinded to the researcher who performed the electrophysiological recordings and the scoring of the behavioral tests.

Stress Index
All behavioral experiments were consistently performed between 9 and 12 AM. The researcher who performed the test was blinded to the mice cohort. In order to build a reproducible index of stress level in mice we build a stress index, combining three behavioral tests. Tests were performed with an interval of 30 minutes, in an increasing level of aversiveness: all animals were first tested in the marble burying test (MBT), then in the looming test (LST) and finally in the forced swim test (FST). The stress index was built combining one parameter for each test using the Euclidean distance of z-scored normalized values. Parameters used: buried marbles (n, in the MBT), time spent in behaviors classified as aversive (s, in the LST), time spent immobile (s, in the FST).

Marble burying test (MBT)
The marble burying test was performed as previously described (39). Briefly, 20 clean glass marbles of diameter 1.5 cm and homogeneous color were disposed of in a 5x4 matrix on a 5 cm deep sawdust without food and water. One by one, animals were placed in the cage for 30 minutes. At the end of the due time, mice were returned to their cage, and the number of buried marbles was scored by two blinded experimenters. Marbles were considered buried if covered by sawdust by at least two-third. Before starting the testing of a new mouse, marbles were cleaned with 70% ethanol and placed in a newly prepared cage.

Looming stimulus test (LST)
30 minutes after the MBT, stressed and control mice performed a modified version of the looming test (40) in a custom designed 8-shaped arena. The looming arena, classically composed of one only squared arena, was here modified to an eight-shaped field, obtained by merging two round arenas, 30 cm diameter; 23.5 height, with black matt walls to prevent reflection of the stimulus. An opening between the two circular chambers allowed free exploration. No shelter was provided, as this arena design allowed for escape to the opposite compartment as a defensive strategy, where the animal behavior was monitored. A monitor was placed on the ceiling of the arena, providing dim lighting from the gray screen of the monitor. As for the rest of the behavioral test, infrared illumination, invisible to the mouse, was provided for video recording. The arena was placed on a transparent plexiglass. The animal behavior was recorded with a camera placed below the arena. Thanks to these modifications, the looming stimulus was reliably repeated three times. The looming stimuli was triggered by the experimenter once the animal was in the center of one of the arenas, as previously described (40). In brief, the stimulus was repeated 15 times with increasing diameter (from 2 to 20 degrees of visual angle) for the first 250 seconds, and then stable at 20 degrees for the remaining 250 ms. The next stimulus was presented when the animal was in the center of any of the two arenas, with a minimum of one-minute interval. Behaviors were scored from video recordings by two blind experimenters. A post-stimuli epoch of 60 seconds after each stimulus was analyzed. Each second was assigned to aversive (escape to opposite compartment, freezing, tail rattling, immobility, periphery) and non-aversive (normal walking, grooming, sniffing) behavior and reported as cumulative time spent in aversive behavior over the three trials.

Forced swim test (FST)
30 minutes after performing the looming test, stressed and control mice were individually placed in a transparent acrylic cylinder (height: 60 cm, diameter: 14 cm) containing 2 L of clear water at 25 ± 1° C for 6 min. The cylinder was placed on a transparent plexiglass. The animal behavior was recorded with a camera placed below the arena. Water was changed between subjects. Delay to immobility (delay in seconds to the first immobility) and time spent immobile (seconds spent floating passively in the water) were manually scored by a researcher blind to the animals' cohort.

Fear conditioning
Mice were placed into a sound isolated TSE Multi Conditioning System where they received five tones followed by mild foot shocks (5 second tone, 0.3 mA foot shock during the last 2 s of the tone) with randomized intershock intervals. The following day, conditioning to the tone was tested in the same TSE Multi Conditioning System, placing the mouse in a new context (arena shape and smell) where they received 10 tones (5 seconds tone) with randomized inter-tone intervals.