Identification of Socially-activated Neurons.

Determining the neuronal circuitry responsible for specific behaviors is a major focus in the field of neurobiology. Activity-dependent immediate early genes (IEGs), transcribed and translated shortly after neurons discharge action potentials, have been used extensively to either identify or gain genetic access to neurons and brain regions involved in such behaviors. By using immunohistochemistry for the protein product of the IEG c-Fos combined with retrograde labeling of specific neuronal populations, precise experimental timing, and identical data acquisition and processing, we present a method to quantitatively identify specific neuronal subpopulations that were active during social encounters. We have previously used this method to show a stronger recruitment of ventral hippocampal neurons that project to the medial prefrontal cortex, compared to those that project to the lateral hypothalamus, following social interactions. After optimization of surgeries for the injection of retrograde tracers, this method will be useful for the identification and mapping of neuronal populations engaged in many different behaviors.

. Accumulation of c-Fos protein begins approximately 45 min after action potential discharge, reaching peak levels after 1-2 h ( Barros et al., 2015). While organic Ca 2+sensitive dyes and genetically encoded Ca 2+ indicators (such as GCaMP) provide a more precise temporal estimation of neural activity in the millisecond-to-second timescale, dyes and sensors require invasive surgical procedures, and current attempts to co-register activity to specific cell subtypes is cumbersome and time-consuming.
Several genetic c-Fos-based strategies have been developed to tag active neurons during specific temporal windows using the rapid transcription and translation of IEGs, including the Fos-tTa and FosTRAP approaches. Fos-tTa uses the c-Fos promoter to drive the expression of the doxycyclinerepressible tetracycline transactivator. In the absence of doxycycline (Dox), a reporter tags the cells with LacZ, which can then be visualized and quantified. However, the slow metabolism of Dox results in a long labeling window in the time frame of several days, which leads to low signal-to-noise (DeNardo and Luo, 2017). An improvement on temporal resolution was achieved with FosTRAP, which reduced the labeling window to ~12 h by using the Cre recombinase system, and a benefit of this system is that it allows genetic access to the labeled neurons (DeNardo and Luo, 2017). However, both of these genetic techniques reduce the temporal specificity of c-Fos itself. Due to their low temporal resolution and signalto-noise ratio, quantification of relative neuronal activity is difficult and unreliable. An alternative to genetic approaches for labeling IEG-determined neuronal subsets is to perform immunohistochemistry for c-Fos protein after sacrificing the animal. This technique is commonly used to identify neurons and classifying them as either c-Fos positive or negative cells based on user-defined thresholds. However, immunohistochemical labeling of c-Fos has the potential to be used in a more quantitative manner by ensuring a precise timing interval between behavioral interactions and animal sacrifice, as well as ensuring identical tissue processing and imaging techniques. Quantification of c-Fos levels provides a more accurate depiction of neuronal activity, revealing relative activity of neurons based on subtype, projection, and/or experience, as well as reducing bias introduced by user-defined arbitrary cutoffs to score a neuron as a c-Fos positive cell.
Here, we describe a method that utilizes a precise timing between behavioral interactions and immunohistochemical detection of c-Fos that allows for reliable quantification of relative neuronal activity in different neuron subsets defined by their projection target regions. Fluorescent retrograde tracers are intracranially injected into projection targets of our primary brain region of interest. Retrograde tracers are taken up by axon terminals (Riddle et al., 1995) and transported back to the cell bodies, marking these neurons based on their projection targets. In our study, we injected green fluorescent microbeads (Retrobeads TM ) in the medial prefrontal cortex (mPFC) and red Retrobeads TM in the lateral hypothalamus (lHYP) for the identification of projection neurons in the CA1 region of the ventral hippocampus (vCA1) (Phillips et al., 2019). This technique can be widely applied to other brain regions 3 www.bio-protocol.org/e3744 or by using alternative techniques to label specific neuronal subtypes. After allowing sufficient time for the retrograde transfer of the Retrobeads TM and recovery from surgery (approximately 2 weeks), mice are subject to sequential interactions with either two conspecifics (social condition) or two fake toy mice (object condition). Mice are anesthetized and fixed by transcardiac perfusion at precisely 45 min after the conclusion of these interactions, and c-Fos levels, revealed by immunohistochemistry on sections from the ventral hippocampus, are used as a surrogate for neuronal activity during the interactions. By using a precise timing, we ensure the highest possible resolution and selectivity for neuronal activity driven by behavioral interactions. By comparing c-Fos levels within neurons with different projection targets, and between neurons from mice in either the social or object conditions, we determine the specificity of neuronal activity by projection region, as well as the selectivity for social encounters (Phillips et al., 2019). While this protocol has previously been used to gauge neuronal activation following social versus object interaction, the test mouse can be exposed to different stimuli during the behavioral paradigm to assess neural activation to other social cues such as male/female odor or social novelty.

Materials and Reagents
A. Surgery to behavioral acclimation. On days 1-3, mice are acclimated to the testing room. On days 4-6, mice are handled (to reduce their anxiety) and the test mice are acclimated to the testing chamber. On day 7, mice are exposed to the behavioral paradigm for labeling and sacrificed.

Notes:
a. For the behavioral paradigm, you will need one non-surgical mouse per cage acting as a known sentinel. In group housed males, it is recommended that the groups including the sentinel are littermates and that the housing is kept to no more than 5 mice per cage. Additional non-surgical novel sentinel mice (age-, sex-, and size-matched, also group housed) should be made available for the behavioral testing. b. Test mice will be injected with Retrobeads TM 2 weeks prior to final behavioral exposure and sacrifice, which should be considered in the case of age-or disease-related studies ( Figure 1). 7 www.bio-protocol.org/e3744 Bio-protocol 10 (17) 1. Aliquot Retrobeads TM into small PCR tubes, adding one volume more than will be used to account for hub loss (250 nl x number of injection sites per mouse x # of mice + 250 nl).
2. Insert Hamilton needle into syringe, priming with sterile saline, ensuring no bubbles are in the needle.
3. Anesthetize mice in induction chamber using 4-5% vaporized isoflurane diluted with 100% oxygen at a constant flow rate of 2.5 L/min. 4. Give mouse subcutaneous injection of analgesics (e.g., 2 mg/kg Ketaprofen). 5. Shave the head of the mouse using standard clippers, sterilizing the scalp with alcohol wipes.
6. Position the mouse in the stereotactic frame. Reduce isoflurane to ~2%, adjusting throughout the surgery as necessary.
7. Apply ophthalmic ointment to the eyes.
8. Subcutaneously inject mouse with analgesics and antibiotics.
9. Using micro-scissors, make an incision in the scalp that is long enough to see both bregma and lambda. 10. Ensure the head is leveled and straight using the arms of the stereotactic frame. Tighten ear bars to secure mouse, being careful to avoid changing the orientation of the head.  25. Table 1 shows the stereotactic coordinates used to target the mPFC and the lHYP during injections of Retrobeads TM in adult C57/Bl6 mice.  3 min by picking them up and allowing them to sit in your hands. After 3 min, or when the mice have calmed down (whichever is longest), transfer the mice to the interaction chamber. The sentinel mice should be handled, but not moved to the interaction chamber.
3. After 10 min of acclimation, remove the mice from the interaction chamber and return to the home cage. Remove the used bedding and replace with clean. At the end of each day, clean the interaction chamber with 70% isopropyl alcohol.
4. On labeling day (day 7), remove sentinels from home cages and place singly in clean cages for the duration of the testing. Be sure to mark the cages well for ease in determining which mice will be the familiar and novel sentinel.
5. Use a random number generator to separate surgical mice into 2 groups: those that will interact with other mice, and those that will interact with inanimate objects (e.g., toy mice). Be sure to fill the bottom of the interaction chamber with clean bedding before each interaction session.
6. For test mice that will interact with other mice, place the test mouse in the interaction chamber and quickly add the familiar sentinel (cage mate). Allow to interact for 10 min before returning to respective cages ( Figure 3). 7. One hour following the initial exposure, return the test mouse to the interaction chamber. Quickly add the novel sentinel (non-cage mate) and allow them to interact for 10 min.
8. For test mice interacting with objects, place one of the fake toy mice in the corner of the interaction chamber prior to adding the test mouse. Allow to interact for 10 min before returning the test mouse to the home cage. 9. After 1 h, place the second fake toy mouse in an alternative corner. Add the test mouse and allow to interact for 10 min before returning the test mouse to the home cage. 7. Dissect the brain, being careful not to create nicks in the tissue. Immerse the brain in 4% paraformaldehyde, and store at 4 °C overnight.
8. Replace 4% paraformaldehyde with brain storage solution (see Recipe 5). 9. Be sure to clear final perfusion line of paraformaldehyde using 1x PBS before next perfusion. D. Sectioning 1. Soak vibratome cutting blades in acetone for ~5 min prior to use (to remove shipping glue and oil).
2. Cut brain with large blocking cuts using a razor blade; make sure to save injection sites for confirmation. Make straight cuts, leaving ample room for the brain to be glued to the cutting chuck.
3. Prepare agar block in a way that the sides are cut straight and it is slightly wider and taller than the brain. Glue this side to the back of the cutting chuck using cyanoacrylate glue (e.g., Super Glue TM ). 12 www.bio-protocol.org/e3744 Bio-protocol 10 (17) 4. Quickly remove excess brain storage solution using tissue paper, and glue brain directly in front of (closer to the blade) the agar block on the cutting chuck. 5. Position the chuck in the vibratome, fill vibratome well with 1x PBS. 6. Cut brain sections at 30 µm thickness for c-Fos immunohistochemistry, and 100 µm thickness for visualization of injection sites. Store serial sections in order within 12-well plates containing brain storage solution. Seal the well plates with Parafilm TM , and store at 4 °C and covered from light until staining.