In Vivo Calcium Imaging with a Miniaturized Microscope in the Hypothalamus for Understanding Social Behaviors in Mice

Summary

This protocol details in vivo calcium imaging of hypothalamic neurons using a miniaturized microscope (miniscope) in freely behaving mice during social state manipulation (social isolation and social reunion). This method enables the imaging at single-neuron resolution of the hypothalamic activity that represents social need and social satiation, enabling mechanistic studies of social homeostasis.

Abstract

Social behavior in vertebrates is tightly regulated through hypothalamic circuits. Molecularly defined hypothalamic neuron types were recently identified to encode social need in a homeostatic manner. Here, we describe a step-by-step protocol for performing in vivo calcium imaging in the hypothalamic medial preoptic nucleus (MPN) using a head-mounted miniaturized microscope (miniscope) in freely behaving mice. This method permits longitudinal recordings of neuronal population dynamics during social isolation and reunion, thus enabling mechanistic studies of social need and social satiety at single-neuron resolution. Specifically, adult mice receive stereotaxic injections of an adeno-associated virus (AAV) expressing a genetically encoded calcium indicator into the MPN, followed by implantation of a gradient-index (GRIN) lens and a miniscope baseplate. After recovery, mice are habituated to the miniscope imaging setup and are subjected to scheduled social isolation and reunion, during which neuronal activities are imaged. Calcium signals are synchronized to animal behaviors for further analysis. This protocol provides comprehensive guidance for in vivo calcium imaging with miniscope in the mouse hypothalamus during social behaviors. It can be readily adapted for imaging in other deep brain regions and across different behavioral contexts. This approach bridges cellular-level neural activity and social behavior, advancing our understanding of the neural basis of complex behavior in freely moving animals.

Introduction

The ability to maintain social bonds is fundamental for animal survival^1,2,3. Prolonged social isolation induces anxiety, cognitive deficits, and metabolic changes, whereas reunion restores social engagement and social needs^4,5,6. The observed rebound in social behavior after short-term social isolation suggests the presence of a homeostatic mechanism⁷, analogous to the regulation of hunger^8,9 or thirst^10,11,12. However, the neural substrates mediating social homeostasis remain poorly understood.

Recent work has revealed that distinct hypothalamic neuronal populations encode social need and social satiety, respectively⁷. In vivo calcium imaging has shown that a population of genetically defined glutamatergic neurons in the medial preoptic nucleus (MPN) becomes active during isolation (MPN^Isolation neurons), whereas another population of GABAergic neurons becomes active upon reunion (MPN^Reunion neurons). Optogenetic activation of MPN^Isolation neurons mimics a social isolation state, and optogenetic activation of MPN^Reunion neurons attenuates social rebound, suggesting their opposing roles in regulating social need. These populations interact reciprocally and connect with brain-wide neural circuitry, forming a feedback loop that dynamically modulates social need.

Here, we present a detailed protocol with the overall goal of enabling researchers to record deep-brain calcium dynamics at single-neuron resolution in freely moving mice during ethologically relevant social behaviors such as social isolation and reunion¹³. The combination of this imaging technique with ethologically relevant social paradigms offers an unprecedented opportunity to investigate the neural representation of distinct social states. The rationale for developing this adapted method stems from the need to visualize how social need and satiety are dynamically represented in deep brain structures such as the MPN. While the use of miniscope has been previously described^14,15,16, its use during unrestricted social interactions, particularly those involving transitions between isolation and reunion requires specific technical and behavioral adaptations to maintain optical stability, minimize motion artifacts, and ensure naturalistic behavioral expression.

Compared to fiber photometry, the miniscope-based approach offers cellular resolution, allowing discrimination of distinct functional neuron types within the same region. It permits long-term recordings of freely interacting animals, which is essential for studying socially driven neural dynamics. Previous studies using one-photon microendoscopy have largely focused on the cortex or dorsal hippocampus during spatial or sensory tasks^14,15,16. Our protocol extends these advances to deeper subcortical regions during complex social behaviors, providing a template for investigating affective state circuits under naturalistic conditions.

The protocol is optimized for brain regions up to ~5 mm deep, using GRIN lens implantation and genetically encoded calcium indicators. Key considerations include surgical precision, lens placement, behavioral paradigm design, and related data analysis. This method can be adapted to other deep brain areas and behavioral contexts. This protocol guides users through all critical stages: surgical preparation, viral delivery, lens implantation, behavioral design, miniscope attachment, imaging, and data analysis (Figure 1A-E). We anticipate that this accessible, detailed workflow will empower the study of deep-brain neural populations across a wide range of naturalistic behaviors.

Protocol

All procedures were approved by the animal ethics committee at Westlake University with the animal protocol #25-076-LD-9. All experiments were performed in accordance with the guidelines of the Laboratory Animal Resources at Westlake University. All equipment and materials used in this protocol are listed in the Table of Materials.

1. Virus injection surgery

1. Administer carprofen (5 mg/kg) subcutaneously for analgesia and place the mouse in an anesthesia induction chamber with 2-3% isoflurane.
   NOTE: A successful induction of anesthesia is confirmed by the absence of a toe-pinch reflex and the presence of steady, deep breathing.
2. After 2-3 min, or once the righting reflex is lost, quickly transfer the animal to the stereotaxic instrument. Secure the nose in the anesthesia nose cone and maintain anesthesia with 1-1.5% isoflurane for the duration of surgery (Figure 1B). The oxygen flow rate is typically set between 0.8-1.2 L/min.
   NOTE: Monitor the animal's respiratory rate and other physiological signs. If the animal's respiratory rate increases and it shows whiskering and other signs of waking, increase the isoflurane concentration to 2-3%.
3. Secure the mouse's incisor teeth through the stereotaxic instrument's incisor bar. Tighten the nose clamp to fix the head.
4. Carefully insert two ear bars into the external auditory canals until they securely rest, ensuring a stable and rigid head fixation (Figure 1B).
   NOTE: (1) Improper placement that compresses neck tissues could cause respiratory distress or unstable fixation. (2) If the animal shows signs of labored breathing or struggle during fixation, immediately release and reposition the ear bars, as this indicates potential airway compression. (3) Ensure both ear bars are at the same height. This is essential for subsequent precise leveling of the skull.
5. Apply ophthalmic ointment to both eyes of the mouse to prevent the eyes from drying out.
6. Remove head fur using hair removal cream. Wipe the area thoroughly with 75% ethanol wipes to remove residual cream and hair.
7. Disinfect the exposed skin with 75% ethanol wipes and betadine. Using fine surgical scissors to make a midline sagittal incision through the scalp. Retract the skin to fully expose the skull over the target brain region. Gently clean the skull surface with a cotton swab to remove the periosteum.
   NOTE: Exposing a sufficiently large area of the skull is necessary to provide a robust anchor for the dental cement, ensuring long-term stability of the implant.
8. Mount a glass pipette onto the stereotaxic holder for skull surface measurement. Under a stereomicroscope, locate the pipette tip to the bregma (the intersection of the sagittal and coronal sutures) and set it as the origin (X = 0, Y = 0, Z = 0, Figure 1C).
9. Move the pipette tip to positions ±2 mm lateral to bregma and adjust the skull level until the height difference between these two points is within ±0.03 mm.
10. Then, move the pipette to the lambda (the intersection of the sagittal and lambdoid sutures) and adjust the anteroposterior tilt until the height difference between bregma and lambda is within ±0.03 mm (Figure 1C).
    NOTE: (1) A preceding visual assessment and adjustment of the skull could significantly streamline the fine leveling process. (2) Drying the surface of the skull with compressed air before leveling improves the visibility of the cranial sutures.
11. Use a handheld drill with a 0.8 mm drill head to make a craniotomy over the target brain region (e.g., for C57BL6/J mice. The coordinates are at anterior-posterior (AP) 0 mm, medial-lateral (ML) 0.3 mm, and dorsal-ventral (DV) -4.8 mm to target the MPN region; for FVB/NJ mice, the craniotomy is made at AP +0.4 mm-+0.5 mm, ML 0.3 mm, and DV -4.8 mm) (Figure 1C).
    NOTE: In case of bleeding, temporarily increase the isoflurane concentration to 2-2.5%. Use a sterile cotton tip or a hemostatic sponge to absorb the blood, then immediately apply sterile saline. Return the isoflurane to the maintenance level (1-1.5%) after hemostasis and remove excess fluid.
12. Following craniotomy, perform stereotaxic injection of the virus expressing the calcium indicator in the targeted brain region.
13. Dilute the AAV-hSyn-jGCaMP7f, an adeno-associated virus that is used to express a genetically encoded calcium indicator in neurons and to allow the imaging of neuronal activity via calcium signals, to a final titer of approximately 5 × 10¹² viral genomes per milliliter before injection.
14. Fill a glass pipette with mineral oil first, and then the viral solution, ensuring absence of air bubbles. Position the pipette at the target coordinates on the surface of the skull and slowly lower it to the target depth.
15. Inject the virus unilaterally at a slow, constant rate (e.g., 30 nL/min), with a total volume of 400 nL⁷, to ensure the infection of enough neurons in the MPN region. This amount can be adjusted according to specific sizes of brain regions and whether further constraints, such as the Cre-loxP system, exist.
16. Upon completion, leave the pipette in place for 10-15 min to facilitate adequate viral diffusion and minimize reflux along the injection track, before being slowly withdrawn.
    NOTE: (1) Viral aliquots with appropriate and consistent titers are critical for the successful and consistent expression across animals. Excessively high titer may cause neurotoxicity, while insufficient titer leads to weak expression and poor signal-to-noise ratio for imaging. (2) Slow injection speed and enough waiting time before withdrawing the pipette are critical to minimize tissue damage and prevent viral reflux, ensuring precise and robust viral expression in the target area.
17. After virus injection, sew the skin above the skull with absorbable sutures.
18. Subcutaneously inject carprofen (5 mg/kg) immediately after surgery for analgesia, with continued monitoring and extra administration over the following 3 days.
19. House the mouse individually after virus injection surgery for at least 2-3 weeks to allow sufficient animal recovery and viral expression.
    NOTE: GRIN lens implantation (Section 2) can be performed on the same day as the viral injection or 2-3 weeks after the viral injection. In this protocol, we describe the two-step method, which allows the experimenter to assess viral expression during GRIN lens implantation.

2. GRIN lens implantation

1. Repeat steps 1.1-1.10 to reopen the craniotomy. Scratch the skull surface with a scalpel for better grasping of dental cement in step 2.8. For the new craniotomy, the diameter should be slightly larger than the GRIN lens (typically ~1 mm) to allow for smooth implantation. Meticulously clear all bone debris after drilling.
2. To mitigate tissue compression during the GRIN lens implantation, insert a glass pipette into the target brain region and slowly move the pipette mediolaterally to create an incision. This micro-cut is ~1 mm long to create space for implanting the GRIN lens (Figure 2A).
   NOTE: This step is critical for the success of the whole protocol.
3. Before implanting the GRIN lens, disinfect the lens with 75% ethanol, then wash it with sterile saline, and gently wipe it clean with lens tissue.
4. Hold the GRIN lens with a lens holder and attach the miniscope to the top of the lens.
5. Position the entire assembly at the bregma and set all coordinates to zero (i.e., X = 0, Y = 0, Z = 0). Then, position it to the target X-Y coordinates above the craniotomy.
6. Moisten the skull surface with sterile saline. Very slowly lower the GRIN lens into the brain tissue at a rate of approximately 100 µm/min until it reaches the target (Figure 2B).
   NOTE: (1) If the GRIN lens implantation is performed 2-3 weeks after viral injection, during the lens insertion, turn on the miniscope to observe real-time fluorescence changes. As the lens approaches the target region, the overall fluorescent signal in the field of view becomes brighter, and this change can be directly read from the signal histogram in real time using the imaging software (Figure 2C, D). (2) Be careful to avoid bleeding during implantation, as blood clots significantly degrade signal quality. (3) If bleeding is inevitable and severe, withdraw the GRIN lens, clean the lens, and re-implant it after the bleeding stops.
7. After the lens implantation, dry the skull surface thoroughly by absorbing residual liquid and blowing with compressed air.
8. Carefully apply dental cement to anchor the GRIN lens base to the skull and build a full head cap that covers the exposed skull (Figure 2B). Be careful to avoid the cement from contacting the surrounding muscle tissues. Then carefully loosen the lens holder and remove the holder together with the miniscope.
   NOTE: (1) Mix dental cement in a cooled slab on ice to slow down the solidification process and provide more working time. (2) A head bar is recommended to be added behind the lens for holding the mouse during the miniscope connection. (3) Be very cautious not to apply the cement to the lens holder or miniscope.
9. Prepare the two-part silicone adhesive by mixing its two liquids together. Apply it to the top of the GRIN lens to form a protective seal against dust and scratches during the recovery period.
10. After the GRIN lens implantation, house the mouse individually to avoid potential damage to the lens by its cage mates during recovery. Administer carprofen (5 mg/kg) subcutaneously for 3 days after surgery for analgesia.

3. Baseplate installation

1. Two to three weeks after the GRIN lens implantation, baseplate installation is performed. Anesthetize and fix the mouse in the stereotaxic instrument as described above.
2. Remove the silicone seal from the top of the GRIN lens and gently clean the lens surface with a piece of lens tissue.
3. Attach the baseplate to the miniscope. Mount the entire microscope-baseplate assembly onto the stereotaxic manipulator arm via a holder.
4. Connect the miniscope to the data acquisition box and turn on the live imaging.
5. Align the miniscope objective and the implanted GRIN lens by finely adjusting the manipulator to achieve precise optical co-axial alignment between the miniscope and the GRIN lens.
6. Position the miniscope objective to be parallel to the GRIN lens upper surface and make sure that the border of the whole GRIN lens is included in the imaging window with a perfect circular look, and the center of the lens is at the center of the imaging window.
7. Lower the miniscope to an optimal working distance to acquire clear images of neurons in the target focal plane.
8. Use dental cement to bridge and encapsulate the baseplate onto the existing cement cap (Figure 2B).
   NOTE: (1) A critical step is to raise the miniscope by a small, empirically determined distance (approximately 50-100 µm) before the final fixation to compensate for potential focal plane shift caused by cement shrinkage during solidification. (2) Be extremely cautious to prevent the dental cement from contacting any parts of the miniscope that are to be detached after baseplate installation.
9. After the dental cement is fully solidified, loosen the holder, carefully disconnect the microscope from the baseplate, and immediately screw a cap onto the baseplate to protect the optical interface.
10. Individually house the mouse for 3-5 days after the baseplate installation and then reunite the mouse with its cage mates for at least 1 week before social state manipulation (e.g., social isolation).

4. Habituation and imaging parameter setting

1. Habituate the implanted mice to temporary head restraint for miniscope attachment and to the imaging setup for 2-3 times prior to formal imaging sessions.
2. To mount the miniscope to the baseplate, restrain the head of the implanted mice, remove the baseplate cap, and dock the miniscope into the baseplate through a magnet interface. Then, tighten the side screw to secure the miniscope. Perform calcium imaging for 10 min, during which the animal is freely moving in the recording arena or cage.
3. Introduce stimuli, such as a gentle touch and a new mouse, during this pilot imaging session. Analyze the data (see details in Section 6) to check whether single-neuron signals could be extracted. During habituation, determine the optimal imaging settings by adjusting the miniscope focal plane across its working distance to maximize the number of neurons observed and image sharpness.
4. Set the illumination power to ~10% of the maximum value and the sensor gain to ~10-20% of the maximum value to get clear signals without saturation. Image at the 1280 x 800 spatial resolution with a 20 Hz sampling rate. Save these settings for each animal and reuse the settings during the following imaging sessions.

5. Social behavioral assays for miniscope calcium imaging

1. Social reunion assay (Figure 3A)
   1. Isolate the miniscope-implanted mouse for 1, 3, or 5 days prior to imaging. Perform calcium imaging during baseline, reunion and re-isolation.
   2. Baseline: Connect the mouse to the miniscope and record for 10 min during which the mouse is kept isolated.
   3. Reunion: Add a former cagemate to the recording chamber and record for 10 min.
   4. Re-isolation: Remove the cagemate after the 10 min reunion, and image for another 10 min during re-isolation.
2. Social isolation assay (Figure 3B)
   1. Isolate the miniscope-implanted mouse for 6 h, and record calcium signals at the start of every hour for 15 min.
   2. Add a former cagemate to the recording chamber and record for 10 min at the end of the entire imaging session to identify neurons that are significantly modulated by social reunion (See details in Sections 6 and 7).
      NOTE: (1) This specific imaging schedule is employed to minimize signal photobleaching and enable the monitoring of neuronal activity during the first 6 h of social isolation. (2) The miniscope is connected throughout the whole procedure. (3) This approach enables the tracking of the same neuronal ensemble over the entire imaging session. (4) Food and hydrogel are provided during imaging.

6. Data preprocessing

1. Import raw imaging videos into the data processing software.
2. Spatially and temporally down-sample the original video when necessary to reduce the computational load for analyzing the data.
   NOTE: For this analysis, raw imaging data is captured at 1280 x 800 spatial resolution at 20 Hz; spatially down-sample the video by a factor of 4 and temporally down-sample it by a factor of 2.
3. Apply a spatial bandpass filter with the global mean subtraction to reduce the high-frequency noise and the low-frequency background drift.
4. Correct motion artifacts using the motion correction algorithm in the data processing software. Visually inspect the motion-corrected movie to ensure stable neuronal somata and minimal residual motion artifacts.
5. Click the "Identify Cells" button and choose "CNMFe" in the list to extract calcium activity traces of individual neurons with the CNMF-E¹⁷ (Constrained Non-negative Matrix Factorization for microEndoscopic data) algorithm.
6. Click the ruler button in the data processing software and measure the soma diameters (in pixels) from representative neurons. Input the average size in the "Average cell diameter (pixels)" box. Then click the Apply button to run CNMF-E.
7. Adjust CNMF-E initialization settings when cell detection is insufficient, excessive non-cell components appear, or over-segmentation occurs.
8. Inspect all extracted components and accept/reject them based on cell morphology and biologically plausible calcium dynamics. Exclude components with irregular and spiky shapes. Exclude noise or motion artifact-dominated traces.
   NOTE: (1) Figure 2E and 2F show two accepted neurons with clear morphology of neurons as well as the typical calcium transients. (2) Figure 2G and 2H show two rejected neurons with pronounced baseline fluctuations and lacking typical calcium transients.
9. Export fluorescence traces of accepted neurons from the data processing software for downstream identification of social state modulated neurons (Section 7).

7. Identify social state modulated neurons

1. Use MATLAB or Python to further analyze the preprocessed data.
2. Align calcium activity traces of the accepted neurons to behavioral timestamps.
3. Find the time point of social reunion from the behavior video. For each neuron, segment activity into a 300 s pre-reunion social isolation period and a 300 s social reunion period.
4. For each neuron, quantify the calcium activity distributions during the isolation and reunion periods.
5. Perform the Receiver Operating Characteristic (ROC) analysis¹⁸ to evaluate the ability of neuronal activity to discriminate between isolation and reunion states, yielding an AUC value that calculates the area under the ROC curve.
6. Assess the statistical significance using a permutation test. For each neuron, calcium activity values are circularly shuffled across behavioral states in 1 s time bins for 1,000 times to generate a null distribution of AUC values. Neurons with observed AUC values exceeding the 95th percentile of their shuffled null distribution are considered significantly modulated by the social state.
7. Classify the significantly modulated neurons based on the direction of modulation. Classify the neurons inhibited by reunion as the Isolation Neurons and activated by reunion as the Reunion Neurons (Figure 4).

8. Post-experiment verification

1. Perfuse the mice and do a standard histology to verify the GRIN lens track and viral expression (see an example histology imaging in Figure 2I).
2. Use the 4% PFA to fix the whole mouse head together with the implant, then remove the GRIN lens to visualize a clear lens track.
   NOTE: Only include the data from the animals with correct targeting.

9. Troubleshooting & practical tips

1. Secure a head-bar with dental cement behind the miniscope to facilitate the miniscope mounting and removal.
   NOTE: The head-bar should not be too long and could be placed in a slightly tilted angle to prevent the interference with social behavior.
2. Individually house the implanted mice during the recovery period following the surgery for at least one week. This provides enough time for the mice to recover and prevents the damage of the implant by other mice.
3. Reunite the implanted mice to their previous cagemates once they are fully recovered. Regularly check and tighten the cover caps. Prepare some extra cover caps and replace them when they are lost during animal maintenance.
4. Check the calcium signals 3-4 weeks after the GCaMP virus is injected. If the signals are too weak or unclear, a longer period of recovery sometimes improves the quality of signals.
5. Be careful of the lens surfaces at either the GRIN lens side or the miniscope side. Keep lenses clean and free from scratching. Cover the lens when not using.
6. Before imaging, adjust the length of the signal cable to be long enough for the animal to reach corners of the imaging arena and short enough to allow for smooth social interactions.
7. Protect the cable with a layer of tough material when necessary to prevent biting from other mice during social interactions. Check the signal cable regularly to monitor if it is twisted by intense social interaction.
8. To identify the same neuronal ensemble from consecutive imaging sessions (for example, in Figure 3B), the raw imaging videos could be concatenated to extract calcium traces.
9. To synchronize multiple signals, for instance, the behavioral videos and the calcium signals, common TTL inputs and synchronized flashlights are usually used to align different recording files.

Results

Identification of neuronal populations
We expressed the GCaMP virus and implanted a GRIN lens in the MPN of a C57BL/6J mouse. Following the surgery, the mouse recovered individually and was reunited with its cage mates for at least one week. The mouse was then singly housed for 3 days prior to the calcium imaging session. During imaging, a 10 min isolation baseline was first recorded, followed by a 10 min social reunion with a former cagemate. Calcium activity recorded before and after the social reu...

Discussion

This protocol details an integrated approach combining in vivo miniscope calcium imaging with social behavioral paradigms to assess the activity profile of hypothalamic neurons. It enables long-term, cellular-resolution calcium imaging from deep brain structures during natural social interactions. The following sections discuss the critical steps, methodological significance, limitations, and future potential of this approach.

The success of this protocol hinges on a series of meticul...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
AAV2/9-hSyn-jGCaMP7f-WPRE-pA	Taitool	S0587-9	Viral vectors, Surgery
Absirbable Haemostatic Gelatin Sponge	FUKANGSEN	FKS-A	Surgery
Betadine solution (10% iodinatedPovidone) 500mL	HYNAUT	Q13702855HNW002	Surgery
Camera for behavioral video	DAHENG	MER2-160-227-U3C	Behavior
Carprofen	MCE	HY-B1227	Surgery
Digital Stereotaxic Instruments	RWD	68803	Surgery
Ethanol (75%)	Sinopharm chemical reagent co.,Ld	801769610	Surgery
Eye Ointment	FODU	H37022025	Surgery
Inscopix Data Processing Software (IDPS)	Inscopix	1000-006551	Data Analysis
Isoflurane	RWD	R510-22-10	Surgery
Kwik-Sil adhesive pack of 2	World Precision Instruments	KWIK-SIL	Surgery
Microdrill	RWD	78001	Surgery
MP-1000 Micropipette Puller	RWD	MP-1000	Surgery
Multi-function Animal Anesthesia Solutions	RWD	TAIJI-IE	Surgery
Nanoliter Microinjection Pump	RWD	R-480	Surgery
nVista 3.0 miniscope	Inscopix	1000-004323	Imaging
Opthalmic forcep	RWD	F12006-10	Surgery
Operating Scissors	RWD	S14001-15	Surgery
ProView Implant Kit	Inscopix	1000-004238	Surgery and Imaging
Resin cement (Super-bond)	Sun Medical	Super bond C&B	Surgery
Set of 5 baseplate covers	Inscopix	1050-004639	Imaging
Set of 5 baseplates	Inscopix	1050-004638	Imaging
Set of 5 GRIN lenses, 0.6mm diameter, 7.3mm length	Inscopix	1050-004626	Surgery and Imaging
Stereotaxis Microscope	Murzider	MSD205A	Surgery
Sterile saline	QIDUKELIN	R1B25071902	Surgery