Dynamic recruitment of PV interneurons in the somatosensory cortex induced by experience dependent plasticity

We thank Ms. Brenda Vasquez, Ms. Callie Liu, Ms. Sravya Gadepalli, and Mr. Noe Cazares Jr. for assistance with mouse breeding and colony management. We thank Drs. Aye Theint Theint, Sammy Alhassen, Nazim Kourdougli, Anand Suresh, and Carlos Portera-Cailliau for helpful feedback, constructive criticism, and advice. Conflict of Interest: Authors report no conflict of interest Funding Sources: This work was supported by American Heart Association grant 24POST1193716 to BC, National Institutes of Health grant 1K08NS114165-01A1 to WZ, and American Academy of Neurology Grant NRTS 2199 to WZ. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint

Adaptive circuit plasticity plays crucial roles in the brain during development, learning, sensory experience, and after injury. During chronic whisker trimming, a well-studied paradigm for inducing experience dependent plasticity, whisker representations in the somatosensory barrel cortex (S1BF) undergo remapping, with expansion of maps for spared whiskers and contraction of maps for trimmed whiskers. At the cellular level, excitatory pyramidal cells in Layer 2/3 shift their whisker tuning, increasing selectivity to spared whiskers and away from deprived whiskers. While these changes are well documented, the circuit mechanisms regulating experience- dependent plasticity remain incompletely characterized. Parvalbumin (PV) interneurons play important roles in regulating the spatial and temporal dynamics of sensory evoked activity in Pyr cells and have been implicated in the regulation of experience dependent plasticity in other cortical regions. However, there is little evidence as to how the sensory evoked activity of PV cells change in S1BF during whisker trimming or how those changes might affect cortical remapping. To address these questions, we used longitudinal in vivo two-photon (2P) calcium imaging of PV cells in S1BF before, during, and after inducing experience-dependent plasticity by whisker trimming. At baseline, we found that PV cells have spatially distributed responses to whisker deflections, responding best to the principal whisker of a given barrel and less frequently to surround whiskers in a distance-dependent manner. After whisker trimming, there is a substantial recruitment of PV cells responsive to the spared whisker in deprived, but not spared, barrels. Upon whisker regrowth, this recruitment is reversed, but changes in individual PV cell whisker selectivity can persist for weeks. To probe the potential casual effects of increased PV activity during whisker trimming, we used chemogenetics to acutely manipulate the activity of PV cells and found that modulating PV cell activity strongly affects sensory evoked responses in local Pyr and PV cells, as well as Somatostatin (SST) interneurons. In particular, increased PV cell activity strongly suppressed activity in all three cell types. Together, our results reveal dynamic changes in the spatial distribution and tuning of PV cells during experience dependent plasticity and suggest that increased PV cell activity could constrain the extent of potential cortical remapping in the adult S1BF. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint

Cortical circuits have the capacity to adapt and change through a process broadly termed plasticity. Cortical circuit plasticity can occur in many contexts, including development, learning, disease, and in response to sensory experience1–4. One of the best studied models for sensory experience dependent plasticity is the rodent whisker somatosensory system. Mice rely on whiskers as a primary sensory input and the whisker somatosensory area (S1BF) is one of the largest regions of the mouse cortex5. The S1BF has a strong somatotopic organization6. Thalamocortical inputs from the ventroposteriomedial nucleus (VPM) conveying information from individual whiskers project primarily to Layer 4 (L4) of corresponding columns, or “barrels”7,8. From L4, activity propagates vertically to supra- and infra-granular layers and then horizontally to adjacent barrels9–11. Within individual barrels, a plurality of excitatory pyramidal (Pyr) neurons respond best to deflection of the barrel’s corresponding principal whisker, but responses are overall very heterogenous, particularly in L2/3, with many Pyr cells responding best to surround whiskers12,13. Sensory deprivation via chronic whisker trimming is a classical paradigm for inducing experience dependent plasticity in the S1BF14. Whisker trimming leads to the expansion of the cortical representation of “spared” un-trimmed whiskers and the contraction of the cortical representation of “deprived” trimmed whiskers15–18. At the level of individual neurons, excitatory Pyr cells in L2/3 undergo dramatic changes in their whisker tuning, with shifts in selectivity for spared whiskers over trimmed whiskers in some neurons and recruitment of additional neurons becoming newly responsive to the spared whisker19. These changes likely arise from both Hebbian and homeostatic synaptic plasticity mechanisms20,21, but the precise circuit alterations regulating experience dependent plasticity remain incompletely characterized. Evidence suggests that parvalbumin (PV) interneurons may regulate experience dependent plasticity. PV cells are the most common cortical interneuron class, comprising ~40% of all cortical interneurons, depending on the cortical region and layer studied22–25. They exhibit key electrophysiological properties including short action potential duration, fast afterhyperpolarizations, and high frequency spiking with minimal adaptation26–28. Anatomically, PV cells synapse primarily onto peri-somatic regions of excitatory neurons29–31. PV cells receive both strong bottom-up inputs as well as intracortical inputs and participate in feed-forward and feedback inhibition. Functionally, PV cells help to sharpen the spatial and temporal precision of activity in excitatory cells32,33, contribute to gain modulation34, and regulate the generation of oscillatory activity35. Several lines of evidence suggest an important role for PV cells during cortical circuit plasticity36. In development, maturation of PV cells corresponds with the closure of critical periods in the visual cortex37. In adult animals, chemogenetic inhibitory modulation of PV cells potentiates experience dependent plasticity in both primary visual (V1)38 and primary auditory (A1) cortex39. Despite being a site of robust experience dependent plasticity, the precise role of PV cells in regulating S1BF circuit plasticity remains incompletely understood. Acutely after whisker trimming the intrinsic excitability of PV cells is reduced in deprived barrels, leading to disinhibition of Pyr cells and temporary preservation of firing rate. However, how PV cell activity and whisker selectivity change over longer time scales and in spared barrels is unknown. Here, we sought to address these gaps in knowledge using acute and longitudinal two-photon (2P) in vivo imaging and chemogenetic modulation of genetically defined PV cells in L2/3 of S1BF before, during, and after chronic whisker trimming. Before trimming, PV cells in S1BF exhibit spatially distributed responses to whisker deflections, with whisker receptive fields similar to those of pyramidal cells. During whisker trimming, PV cells undergo barrel-specific changes in whisker selectivity, with strong recruitment of PV cells responsive to the spared whisker in deprived, but not spared, barrels. On the other hand, after whisker regrowth, baseline whisker selectivity of PV interneurons is largely re-established in deprived barrels, but there are long- .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint lasting shifts in responsivity toward the spared whisker in the spared barrel. Acute modulation of PV cells with DREADDs (Designer Receptors Exclusively Activated by Designer Drugs) leads to changes in the activity of local pyramidal, PV, and somatostatin (SST) cells. In particular, activation of PV cells leads to strong suppression of sensory evoked activity in Pyr and SST cells, confirming a causal role for PV cells in regulating local microcircuit dynamics. Our results provide an important characterization of the spatial distribution of whisker selectivity in PV cells at baseline, how these responses change during and after sensory experience, and point to a causal role for PV-mediated inhibition in regulating experience dependent plasticity.

Materials: DREADD Agonist Compound 21 (C21) was obtained from HelloBio. Recombinant adeno-associated viruses (AAVs), including AAV1-syn-jGCaMP8s-WPRE40 (gift from GENIE Project, Addgene viral prep #162374-AAV1), AAV8-Ef1a-Coff/Fon-GCaMP6f41 (gift from Karl Deisseroth & INTRSECT 2.0 Project, Addgene viral prep #137124-AAV8), AAV9-hSyn-DIO- mCherry (gift from Bryan Roth, Addgene viral prep #50459-AAV9), AAV9-hSyn-DIO-hM4D(Gi)- mCherry42 (gift from Bryan Roth, Addgene viral prep #44362-AAV9), and AAV5-hSyn-DIO- hM3D(Gq)-mCherry42 (gift from Bryan Roth, Addgene viral prep #44361-AAV5) were obtained from Addgene. Experimental Animals: All experiments followed the U.S. National Institutes of Health guidelines for animal research, under an animal use protocol approved by the University of California Los Angeles Animal Research Committee (ARC). Male and female mice were used, beginning at 7- 10 weeks old at the time of cranial window surgery. All animals were housed in a vivarium with a 12 h light/dark cycle. For these experiments we used 68 heterozygous PV-Cre mice (B6.129P2- Pvalbtm1(cre)Arbr/J, Jax line 017320)43, 28 double transgenic PV-Cre:Ai162 mice (Ai162(TIT2L- GC6s-ICL-tTA2)-D, Jax line 031562)44, and 42 double transgenic PV-Cre:Sst-ires-Flp (B6J.Cg- Ssttm3.1(flpo)Zjh/AreckJ, Jax line 031629) mice. All transgenic lines were maintained on a C57BL/J6 background. Cranial Window Surgery: Cranial window implantation was performed according to previously published protocols45,46. Mice were deeply anesthetized with 5% isoflurane, with 1.5-2% isoflurane for maintenance. After removal of the scalp and periosteum, an ~4 mm diameter circular craniotomy, centered over the primary somatosensory cortex (~3 mm lateral to the midline and ~2 mm caudal to Bregma) was made using a pneumatic dental drill with a FG ¼ drill bit (Midwest Dental). For mice requiring injection of AAVs, AAVs were injected at a concentration from 1.8 to 2.1 x 10^13 GC/mL and volume of 75 µL directly into 4 sites in the cortex using a glass capillary nanoinjector (Neurostar). The craniotomy was sealed using either a single 5 mm #1 sterile glass coverslip (Harvard Apparatus), or a 4 mm coverslip glued to a 5 mm coverslip using an optical adhesive (Norland Products, #71) which was glued to the skull with cyanoacrylate glue (Krazy Glue) and dental acrylic (OrthoJet, Lang Dental). A stainless steel headbar for head fixation was embedded in dental acrylic. Carprofen (5 mg/kg, i.p., Zoetis) and dexamethasone (0.2 mg/kg, i.p., Vet One) were provided for pain relief and mitigation of edema on the day of surgery and daily for the next 48 h. Mice were allowed to recover from the surgery for at least 3 weeks before the first imaging session. Whisker Trimming: Animals were anesthetized with isoflurane (5% for induction, 1.5-2% for maintenance). To ensure whisker stimuli deflected only whiskers of interest, the whiskers directly adjacent to the those targeted for stimulation on the right side of the snout were trimmed using fine scissors to a length of ~15 to 20 mm immediately prior to imaging. For chronic whisker trimming experiments, all whiskers on the right side of the face except the spared whisker were trimmed flush with the vibrissal pad immediately following baseline imaging and .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint re-trimmed as needed to remove any whisker re-growth, approximately three times weekly, for a total of 18-21 days, depending on the experiment. Intrinsic optical signal imaging (IOSI) Acquisition: IOSI was performed as previously described18. Animals were sedated with chlorprothixene (~3 mg/kg, i.p.), lightly anesthetized with ~0.5-0.7% isoflurane, and head-fixed. The cortical surface was illuminated with 525 nm light to capture an image of the superficial vasculature. The microscope was then focused 300 µm below the cortical surface and illuminated with 625 nm light to record intrinsic signals. Frames were collected at 10 Hz (100 ms exposure time) using an 8 megapixel CCD camera (Thorlabs, 8051M-USB) during thirty trials of whisker deflections (100 Hz sine wave, 1.5 seconds long). Whisker stimuli were generated in MATLAB, amplified, and delivered using a glass capillary affixed to a piezoelectric bending actuator (Bimitech Python PBA6014-5H200). Cortical whisker maps were generated as previously described18. Briefly, stimulus-evoked change in reflectance values (ΔR/R) were calculated and binarized by thresholding for ΔR/R values below a Z-score of -3. Binarized images were then pseudocolored and overlaid onto images of the vasculature. To quantify map area for single whisker evoked maps, the medfilt2 function in MATLAB was used to apply a median filter with a 3x3 pixel neighborhood size to binarized maps to remove noise and the area of thresholded pixels was calculated. The mean value of all thresholded pixels in the map area was calculated to quantify map intensity. Two-photon Calcium Imaging: In vivo calcium imaging was performed using a Thorlabs Bergamo II microscope with fast galvo-resonant scanning mirrors, 14° collection optics with two high-sensitivity GaAsP photomultiplier tubes, and a 16x/0.8 NA objective (Nikon), coupled to an Insight X3 dual output Ti:Sapphire laser (Spectra Physics). For anesthetized imaging, mice were lightly sedated with chlorprothixene (~3 mg/kg, i.p.) and isoflurane (0.7-0.9%) and kept warm with a heating blanket. For awake imaging, mice were habituated to head-fixation under the microscope in a cylindrical tube prior to imaging. Stimulation of the whiskers (20 stimuli, 1 s duration 10 Hz square wave, with a pseudorandomized 3-6 s interstimulus interval) was delivered to a single whisker by a glass capillary tube affixed to a piezoelectric actuator (Bimitech Python PBA6014-5H200). Imaging data and whisker stimulation data were synchronized using ThorSync software. Whole field 2x zoom 512 x 512 pixel images (typically 422.16 x 422.16 μm) were acquired with bidirectional scanning at ~30 Hz, with averaging of 3 frames for a final image acquisition rate of ~10 Hz. The imaging field-of-view (FOV) was determined according to IOSI maps. For each FOV, ~150 s of spontaneous activity data was collected prior to imaging whisker evoked activity. For experiments in which multiple whiskers on an individual mouse were stimulated, imaging was performed serially for each whisker – for example, imaging was performed during stimulation of one whisker, then the whisker affixed to the stimulator was switched and imaging repeated. For experiments involving the use of heterozygous PV-Cre mice injected with AAV1-syn-jGCaMP8s we recorded an average of ~14 PV cells and ~123 Pyr cells per FOV. For experiments involving the use of double transgenic PV-Cre:Ai162 mice and PV-Cre:Sst-ires-Flp, we recorded on average ~19 PV cells and ~5 SST cells per FOV, respectively. Motion correction of movies was performed using the motion correction module from EZcalcium47. Fluorescence traces (DF/F) of neuronal calcium transients were extracted using custom-written semi-automated MATLAB routines as previously described48 (from PV-Cre:Ai162 and PV-Cre:Sst-ires-Flp double transgenic mice), or using Suite2P49 (from PV-Cre mice injected with GCaMP8s) automated cell segmentation with manual refinement. For Suite2P extracted traces, 0.7 times the neuropil fluorescence signal was subtracted from somatic fluorescence signals. We calculated a modified Z score vector from the fluorescence traces for each neuron as previously described48. Modified Z-scores for each of the 20 individual whisker stimuli were .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint aligned to the onset of the stimulus and a mean stimulus evoked trace was calculated for each cell. A modulation index (MI) indicating the strength of the whisker evoked response for each cell was calculated as: 𝑀𝐼 = 𝑚𝑒𝑎𝑛(𝑍!"#$ * − 𝑚𝑒𝑎𝑛(𝑍!%& ) Where Zpost is the modified Z-score trace from stimulus onset until 2 seconds later and Zpre is the modified Z-score trace 1 second prior to stimulus onset. To determine if neurons were responsive to stimulation of a specific whisker we used a probabilistic bootstrapping method to correlate calcium transients with epochs of stimulation, shuffling the stimulus trace48. Neurons that had significant correlations between whisker stimulation and fluorescence transients (p values < 0.01) and had whisker stimulation modulation indices greater than 1 were considered whisker responsive. The selectivity index (SI) of individual cells was calculated using the MI to the C1 and D1 whiskers as follows: 𝑆𝐼 = 𝑀𝐼'( − 𝑀𝐼)( 𝑀𝐼'( + 𝑀𝐼)( Negative MI values were set to 0 to constrain SI values between -1 and 1, with SI=-1 indicating perfect selectivity for the D1 whisker and SI=1 indicating perfect selectivity for the C1 whisker. For spontaneous activity, amplitude and frequency of calcium transients were detected using the “findpeaks” function in MATLAB with settings of “MinPeakHeight” = 4, “MinPeakDistance” = 0.5 s, and “MinPeakProminence” = 3. We calculated the mean Z-score and area-under-the- curve (AUC) by quantifying the mean and trapezoidal integral of the entire modified Z-scored calcium trace, respectively. Principal components analysis to predict changes in responsivity of PV cells over time was performed using the “pca” function in MATLAB. A matrix was created with individual cells as rows and column vectors including the barrel in which cells were located (C1 or D1), MI to C1 and D1 whisker deflection, SI, and the mean Z-score of the spontaneous activity trace for each cell. The same data was used to fit a generalized linear mixed effects (GLME) model with a binomial distribution to predict changes in responsivity of PV cells over time. Receiver operating characteristic analysis of the fitted model was then performed using the “perfcurve” function in MATLAB, with 1000 bootstrap replicas. Tuning curves for PV cells in the C1 barrel were calculated by sorting MI to the C1, D1, B2, and E3 whiskers in descending order, with negative MI set to zero. Responses to each whisker were then normalized to the preferred whisker (the whisker with the largest MI) and the data was fit with a single-term exponential decay function. Decay constants were then used as a measure of tuning width, with larger decay constants indicating more narrow tuning. The AUC of the mean stimulus evoked trace was quantified as the trapezoidal integral of the modified Z-score vector over 3.1 s after stimulus onset. For each neuron we also calculated characteristics of whisker stimulus evoked calcium transients using the “findpeaks” function in MATLAB, with settings of “MinPeakHeight” = 4, “MinPeakDistance” = 0.5 s, and “MinPeakProminence” = 3. For each of the 20 individual whisker stimuli delivered in a given imaging session we quantified the presence (or absence) of a detectable peak within 2 s of stimulus onset, the amplitude of that peak, and the latency to peak. For peak amplitude and latency, the mean response was calculated only for stimulus epochs with a detected peak; stimulus epochs without a detected peak were ignored. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint DREADD Agonist Administration: For acute chemogenetic manipulations, mice were imaged as described above to obtain baseline activity measures. The DREADD ligand C2150 (1-5 mg/kg, i.p.), or saline, was then administered and isoflurane anesthesia was turned off. Mice remained on the microscope stage on a heated blanket for 25 minutes. Anesthesia was then resumed as in the baseline imaging session, with brief ~1 minute 5% isoflurane induction followed by maintenance at ~0.7% isoflurane for 4 minutes. Imaging was then repeated exactly as performed in the baseline session. Statistical Analyses: All data are plotted as mean +/- standard error of the mean, unless otherwise stated. Sample sizes were not based on a priori power calculations but are consistent with other studies in the field using similar techniques, including our own48,51. Statistical analyses were performed in MATLAB. Imaging files were blinded to condition during cell selection. Data were tested for normality using the Lilliefors test and analyzed using parametric or non-parametric statistical tests, as indicated in the figure legends. To account for nested data (measuring multiple cells nested within individual mice) we used mixed effects models52. For experiments measuring percentage of whisker-responsive cells (Fig. 1d; 2d; 4a,d,g; 5a,c,e; S1a; S2a; S3a), we used GLME models with a binomial distribution, with individual mouse ID modeled as a random effect. For analysis of MI, SI, AUC, peak amplitude, peak frequency, and peak latency, we used linear mixed effects (LME) models. When individual cells were tracked longitudinally, cells were assigned a unique index and these were nested within individual mice as a random effect. For models with more than one fixed effect, ANOVA was used to test for overall significance of fixed effects. For fixed effects that achieved significance at the level of the ANOVA, p values for individual coefficients from the model were corrected for multiple hypothesis testing using the Benjamini & Hochberg procedure53. In the figures, results are presented with p values for overall fixed effects shown as an inset on respective graphs, with p values for individual coefficients depicted adjacent to the respective data points where appropriate. Full statistical model specification can be found in the supplementary data.

To understand how the activity of PV cells changes with experience, we first sought to define the spatial distribution of whisker evoked responses in PV cells. To record neuronal activity from PV cells, we crossed PV-Cre mice with mice expressing the Cre-dependent genetically encoded calcium indicator GCaMP6s44. We placed chronic cranial windows over the S1BF and used intrinsic optical signal imaging (IOSI) to localize the cortical C1 and D1 barrel representations. We then performed 2P in vivo calcium imaging from L2/3 PV cells in fields-of-view (FOV) centered on the C1 or D1 barrel while individually deflecting the contralateral C1, D1, B2, or E3 whiskers at 10 Hz for 1 second, delivering 20 trials total for each whisker (Fig 1a-b). We tested different waveforms for whisker deflection (square wave or sine wave) and found that faster rise times for square wave deflections and larger amplitude deflections for sine wave deflections led to higher numbers of whisker-responsive PV cells (Fig. S1a). We chose a square wave deflection with 1 ms rise time for all further experiments as this led to robust activation of PV cells with clear whisker evoked calcium transients (Fig. 1c). There were no significant differences in the number of whisker-responsive cells or the magnitude of whisker evoked responses between awake and lightly anesthetized animals using this protocol (Fig. S1). For that reason, and to reduce variability across animals, we performed all subsequent recordings in lightly anesthetized animals. We next quantified whisker evoked activity of PV cells in FOV centered on the C1 barrel. We found that a majority of cells (85.7% +/- 2.5%) responded with time-locked responses to deflections of the C1 whisker (Fig. 1d). Fewer cells responded to surround whiskers, with the .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint number of whisker responsive neurons decreasing as the distance of the surround whisker increased from C1 (46.1% +/- 4.1%, 34.1% +/- 14.0%, and 5.6% +/- 5.6% for the D1, B2, and E3 whiskers, respectively) (Fig. 1d). Mean whisker evoked responses averaged across all trials for all individual PV cells showed that evoked responses were qualitatively substantially higher for C1 whisker deflections compared to other whiskers (Fig. 1e). Given that significantly more cells were responsive to the C1 whisker than other whiskers, this could account for the substantially higher whisker evoked responses when averaged across all cells. Therefore, we classified PV cells into groups of either “responders” or “non-responders” based on their evoked activity to each whisker to compare whisker evoked responses across these groups. As expected, clear whisker-evoked activity was seen for cells classified as responders (Fig. 1f), whereas mean evoked traces were relatively flat for cells classified as non-responders (Fig. 1g). We then calculated a modulation index (MI) to quantify the strength of whisker evoked responses in each PV cell. In responders, whisker evoked activity was greatest for the C1 whisker (MI=8.3 +/- 0.6) and significantly lower for the D1 (MI=3.9 +/- 0.6) and B2 (MI=3.1 +/- 0.3) whiskers (Fig. 1h); whisker evoked activity for the E3 whisker (MI=4.7 +/- 1.2) was also lower, but this result was not significant due to the low number of E3 responsive PV cells in the C1 barrel. In non-responders, whisker evoked activity was low across all whiskers, as expected (Fig. 1i). Finally, we plotted the position of all responsive cells (relative to the FOV centered on the C1 barrel) imaged across all mice to see if there is spatial clustering of PV cells tuned to specific whiskers. C1 whisker responsive PV cells were distributed relatively homogenously across the FOV, whereas D1 whisker responsive PV cells tended to cluster in the anteromedial portion of the FOV, closest to the D1 barrel (Fig. 1j). B2 and E3 responsive PV cells were much rarer and were distributed throughout the FOV (Fig. 1j). We repeated the same experiment for the D1 barrel, focusing just on the C1 and D1 whiskers, and found similar results: More cells were responsive to the D1 whisker in the D1 barrel; whisker evoked responses in responders were higher for the D1 whisker compared to the C1 whisker; and C1 responsive PV cells were clustered in the posterolateral portion of the imaging FOV, closest to the C1 barrel (Fig. S2). Having defined the spatial distribution of PV cell responses to different whiskers in the adult S1BF, we next sought to determine how PV cell activity changes during experience. We implemented chronic whisker trimming, a well-studied paradigm for inducing experience- dependent plasticity. Macroscopically, chronic whisker trimming leads to expansion of cortical map representations of spared whiskers and contraction of maps for deprived whiskers15–18, and this is accompanied by shifts in selectivity of individual Pyr cells to the spared whisker and away from the deprived whisker19. One day after whisker trimming, intrinsic excitability of PV cells in deprived barrels is reduced54. This results in disinhibition of Pyr cells and increased whisker evoked responses to trimmed whiskers acutely, but over time as trimming is maintained Pyr cell responses to the trimmed whisker become depressed55. However, how PV responses change in spared barrels and during more chronic phases of trimming remains unknown. To investigate this, we trimmed all contralateral whiskers except D1 and longitudinally imaged individual PV cells in the C1 and D1 barrels over 18 days of whisker trimming followed by 28 days of regrowth (Fig. 2a). Prior to whisker trimming, the selectivity of individual PV cells in the C1 and D1 barrels were clearly separable, with PV cells in each barrel more selective for the principal whisker of that barrel (SI in D1=-0.37; SI in C1=0.46; Fig. 2b). In the D1 barrel, the number of D1 whisker responsive neurons was high and did not change significantly over 18 days of trimming or after whisker regrowth (Baseline=72.8% +/- 2.8%; Day 18=71.8% +/- 5.2%; Day 46=75.3% +/- 3.9%; Fig. 2c). Mean whisker evoked responses of D1 responsive PV cells in the D1 barrel increased slightly after 18 days of whisker trimming and returned to baseline levels after regrowth (MI: Baseline=6.9 +/- 0.6; Day 18=8.2 +/- 0.7; Day 46=7.8 +/- 1.0; Fig. 2d). In the C1 barrel, the number of D1 whisker responsive cells .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint significantly increased after 12 and 18 days of whisker trimming and was slightly reduced compared to baseline after whisker regrowth (Baseline=46.1% +/- 4.1%; Day 12=59.4% +/- 6.7%; Day 18=59.2% +/- 4.5%; Day 46=37.2% +/- 5.2%; Fig. 2c). Mean whisker evoked responses of D1 responsive PV cells in the C1 barrel did not change significantly over time (MI: Baseline=3.9 +/- 0.6; Day 18=3.5 +/- 0.3; Day 46=3.2 +/- 0.4; Fig. 2d). In control animals that we imaged without whisker trimming, we did not see any recruitment of D1 responsive PV cells in either barrel and there was no change in mean D1 whisker evoked responses (Fig. S3). These results strongly mirror the recruitment of Pyr cells in spared barrels to the spared whisker previously observed during whisker trimming19. We also quantified changes in the spontaneous activity of PV cells over time during whisker trimming. In the D1 barrel, total spontaneous activity (measured as the area-under-the-curve [AUC] across ~150 s of activity) of PV cells changed minimally during trimming but was reduced after whisker regrowth (Fig. S4a). More significant changes were seen in the C1 barrel, where spontaneous activity was reduced during whisker trimming and remained below baseline levels after whisker regrowth (Fig. S4a). These changes were driven primarily by reductions in the amplitude of calcium transients, as calcium transient peak frequency did not change over time (Fig. S4b-c). In addition to population level changes, we also quantified changes in responsivity of individual PV cells longitudinally over time. In the D1 barrel, 7.9% of cells that were non-responsive to the D1 whisker at baseline became responsive after 12 days of whisker trimming (Fig. 2e). This was balanced by 5.8% of cells switching from responsive at baseline to non-responsive after 12 days, resulting in a similar number of responsive PV cells overall (Fig. 2e). In the C1 barrel, 6.2% of cells switched from responsive at baseline to non-responsive after 12 days, but 22.4% of cells switched from non-responsive to responsive (Fig. 2f), consistent with a recruitment of initially non-responsive cells to the D1 whisker after whisker trimming. We next sought to determine if baseline neuronal activity could predict which cells might change their responsivity to the D1 whisker between baseline and day 12. At baseline in the D1 barrel, PV cells that were D1 responsive at baseline and stayed D1 responsive at day 12 (R-R cells) had the highest evoked responses to both D1 and C1 whisker stimulation (Fig. 2g-h). R-R responses were significantly higher than both cells that were non-responsive to D1 at baseline and stayed non- responsive to D1 at day 12 (NR-NR) and cells that switched from non-responsive to D1 at baseline to responsive to D1 at day 12 (NR-R) (Fig. 2g-h). PV cells that switched from responsive to D1 at baseline to non-responsive to D1 at day 12 (R-NR), had responses that trended lower than R-R cells, but this was not significant (Fig. 2g-h). Selectivity of cells in the D1 barrel was biased to the D1 whisker across all classes of cells, as expected, with no significant differences among classes (Fig. 2i). In the C1 barrel, trends in response magnitude to D1 whisker stimulation were similar across classes to the trends seen in the D1 barrel, with R-R cells having higher response magnitudes than NR-NR cells to the C1 and D1 whisker (Fig. 2j-k). Selectivity of all classes of cells in the C1 barrel were biased to the C1 whisker, as expected (Fig. 2l). Compared to baseline D1 non-responders (NR-NR cells and NR-R cells), R- R cells in the C1 barrel had the lowest selectivity for the C1 whisker at baseline (Fig. 2l). These data suggested that individual activity metrics were poor predictors of which cells might change responsivity over time, especially for baseline non-responders to the D1 whisker (NR- NR and NR-R cells). To see if multiple metrics better predict which non-responders switch classes over time, we performed a principal components analysis of non-responder cells using the barrel location, response magnitude to C1 and D1 whisker deflection, selectivity index, and the mean Z-score of the spontaneous activity trace for each cell. NR-NR and NR-R cells were poorly separable by PCA (Fig. 2m). The first principal component explained 89% of the variance and was heavily weighted toward response magnitude to C1 whisker stimulation (coefficient = 0.998). We next tried fitting a generalized linear mixed effects (GLME) model to .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint these data to classify cells into NR-NR or NR-R. Performance of this model as measured by receiver-operating characteristic curve was modest (AUC = 0.77, confidence bounds 0.69-0.83) (Fig. 2n). In the model, only the coefficient for the response magnitude to D1 whisker stimulation was significant (p<0.001). These analyses suggest that response magnitude of cells to whisker stimulation has some predictive value, but that a substantial portion of the variability in which cells will switch from non-responders to responders with whisker trimming is unexplained by our studied measures of baseline activity. We next examined changes in D1 whisker responsivity in the C1 and D1 barrels after whisker regrowth. Individual cells exhibited complex trajectories over time, with all possible changes in responsivity represented across baseline, day 12, and day 46 timepoints (Fig. 3a-b). In the D1 barrel, changes in responsivity were generally balanced, such that the overall proportion of D1- responsive and non-responsive cells was maintained at each timepoint (Fig. 2c, 3a). In the C1 barrel, the increase in D1 responsive cells seen at day 12 with whisker trimming reversed at day 46 after regrowth, with fewer D1 responsive cells compared to baseline (Fig. 2c, 3b). In addition to the binary outcome of whether a cell was responsive to the D1 whisker or not, we also compared the selectivity of cells for the C1 and D1 whiskers between baseline and day 46. In the D1 barrel, although we did not observe any change in the percentage of D1 responsive cells over time, selectivity of cells was even more biased to the spared D1 whisker after trimming and regrowth compared to baseline (Fig. 3c). In the C1 barrel, despite significant shifts in the population of D1 responsive cells during whisker trimming, cell selectivity changed only slightly between baseline and day 46, shifting subtly toward the D1 whisker (Fig. 3d). We also calculated a measure of tuning width for cells in the C1 barrel across all four stimulated whiskers (C1, D1, B2, and E3), with evoked response magnitudes normalized to the response of the most preferred whisker (Fig. S5a-b). Tuning width of PV cells in the C1 barrel shifted slightly between baseline and day 46, becoming more narrow, or selective for the cell’s preferred whisker (Fig. S5c). To better understand these persistent changes in selectivity after whisker regrowth, we calculated the change in the percentage of cells responsive to either the C1 or D1 whisker in each barrel between baseline and day 46 for each mouse. In the D1 barrel there was no significant change in the percentage of D1 responsive neurons (2.5 ± 3.4% increase), but the percentage of C1 responsive neurons decreased by 17.9 ± 6.5% on average across mice (Fig. 3e, left). We also compared changes in the magnitude of whisker evoked responses and found a small but significant increase in evoked response over time to the D1 whisker compared to the C1 whisker (ΔMI=0.8 ± 0.7 for D1; -1.2 ± 0.2 for C1; Fig. 3f, left). In the C1 barrel, the percentage of C1 responsive neurons decreased between baseline and day 46 (-12.1 ± 3.1% across mice), but the percentage of neurons responsive to the D1, B2, and E3 whiskers also decreased (-8.9 ± 6.0%, -32.2 ± 13.7%, and -5.6 ± 5.1%, respectively; Fig. 3e, right). The magnitude of whisker evoked responses was slightly reduced for the D1, B2, and E3 whiskers (ΔMI=-0.8 ± 0.2, -1.3 ± 0.2, and -0.4 ± 0.1, respectively) but was most strongly reduced for the C1 whisker (ΔMI=-2.5 ± 0.4; Fig. 3f, right). These changes suggest that the persistent shift in selectivity toward the D1 whisker in the D1 barrel is driven by a reduction in C1 responsivity, rather than potentiation of D1 responsivity. In the C1 barrel, reductions in C1 responsivity are partially balanced by reductions in responsivity to other whiskers, leading to relatively preserved cell selectivity for C1 after whisker regrowth compared to baseline. Our longitudinal imaging data show that PV cells in the deprived C1 barrel are recruited to become responsive to the spared D1 whisker. These changes could be homeostatic, in response to recruitment of Pyr cells to the spared whisker; driven by increased activity in bottom-up inputs shared with Pyr cells; and/or casual, facilitating or limiting cortical map .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint plasticity. To understand how changes in PV cell activity might causally affect activity in different components of the S1BF microcircuit, we used chemogenetic DREADDs to manipulate the activity of PV cells. We expressed GCaMP8s in all neurons using an AAV with a pan-neuronal synapsin promoter and recorded activity of PV cells (expressing mCherry-tagged DREADDs) and presumptive Pyr cells (mCherry negative) before and after injection of the DREADD ligand C2150. We quantified both spontaneous activity and activity evoked by deflection of the imaged barrel’s principal whisker. In animals expressing the inhibitory DREADD hM4D(Gi) in PV cells, C21 did not change the number of whisker responsive PV cells, but did increase the number of whisker responsive Pyr cells (Fig. 4a). Mean whisker evoked responses were smaller in responsive PV cells after C21, confirming the expected inhibitory effect (Fig. 4b). After C21, responsive Pyr cells showed a sharp, fast rise time of calcium fluorescence, but also faster decay, for overall reduced mean whisker evoked responses (Fig. 4c). Total spontaneous activity was unchanged after C21 in PV cells, though there were small increases in the frequency and amplitude of calcium transients (Fig. S6a-c). Total spontaneous activity in Pyr cells increased after C21 (Fig. S6a), with a shift toward more frequent calcium transients with lower peak amplitudes (Fig. S6b-c). Changes were more substantial in animals expressing the activating DREADD hM3D(Gq) in PV cells. After C21, the number of whisker responsive cells was strongly reduced for both PV and Pyr cells (Fig. 4d). Mean whisker evoked responses were strongly reduced for Pyr cells (Fig. 4f), as expected, but were also reduced for responsive PV cells (Fig. 4e). Spontaneous activity was unchanged in PV cells after C21, but total activity, transient frequency, and transient amplitude were all decreased in Pyr cells (Fig. S6d-f). Controls expressing mCherry in PV cells did not show any changes in the number of whisker responsive cells or mean whisker evoked responses in PV or Pyr cells after C21 (Fig. 4g-i). Total spontaneous activity was unchanged in PV cells (Fig. S6g). Total spontaneous activity and transient frequency in Pyr cells showed very slight increases (Fig. S6g-i), but these were in the opposite direction of the changes in spontaneous activity of Pyr cells seen in animals expressing hM3D(Gq) in PV cells. Some of the changes observed in the activity of PV and Pyr cells after modulation of PV cell activity were unexpected. Therefore, we hypothesized that compensatory changes in other microcircuit components, particularly SST interneurons, might occur following modulation of PV cell activity. To test this, we injected S1BF of PV-Cre:Sst-ires-Flp double transgenic mice with AAVs to express Cre-dependent DREADDs in PV cells and Flp-dependent GCaMP6f in SST cells. We then recorded spontaneous and whisker evoked activity from SST cells. In mice expressing the inhibitory DREADD hM4D(Gi) in PV cells, we found a strong increase in the number of whisker responsive SST cells after C21 (Fig. 5a). Mean evoked traces from whisker responsive SST cells showed a shorter latency to the peak of evoked activity and less overall whisker evoked activity after C21 (Fig. 5b). Spontaneous activity of SST cells was also increased, driven mainly by increased frequency of calcium transients (Fig. S7a-c). Qualitatively, this resembles the changes seen in Pyr cells (Fig. 4c), and suggests that when early, fast somatic inhibition provided by PV cells is reduced, early SST-mediated inhibition may compensate. In animals expressing the activating DREADD hM3D(Gq) in PV cells, the number of whisker responsive SST cells and the mean evoked activity from responsive SST cells were strongly reduced (Fig. 5c-d), similar to the changes observed in Pyr cells (Fig. 4d,f). Spontaneous activity of SST cells was also decreased, including reductions in both the frequency and amplitude of calcium transients (Fig. S7d-f). Controls expressing mCherry in PV cells did not show any changes in the number of whisker responsive cells or mean whisker evoked responses in SST cells after C21 (Fig. 5g-i). Likewise, there was no change in total spontaneous activity or calcium transient amplitude, with only a slight increase in peak frequency observed (Fig. S7g-i). Together, these data demonstrate that modulation of PV cell .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint activity acutely alters microcircuit dynamics across cell types in the S1BF and suggests PV cells may play an important casual role in regulating experience dependent plasticity.

PV cells are the most common type of inhibitory interneuron in the S1BF. Until now, the specific response properties of PV cells to whisker stimulation and the spatial distribution of those responses have not been well studied. Our data define for the first time the spatial distribution of whisker evoked PV cell responses within and across barrels. In a given barrel, deflection of the principal whisker evokes responses in the greatest number of PV cells, with fewer PV cells responding as distance of the whisker increases from the principal whisker of the barrel (Fig. 1d, Fig. S2a). This pattern is largely similar to the well characterized “salt-and-pepper”-like distribution of Pyr cell responses in L2/3 of S1BF, where principal whisker selective cells are intermixed with other cells that respond best to surround barrel whiskers12,13. The number of PV cells responsive to whisker stimulation were higher than we, and others, have reported for Pyr cells, both for the principal whisker (70-86% for PV cells here [Fig. 1d, Fig. S2a] vs. 25-52% for Pyr cells) and surround whiskers (up to 37-46% for PV cells 1 row/arc distant here [Fig. 1d, Fig. S2a] vs. 2-20% for Pyr cells)12,13,51. We also found that many PV cells are responsive to multiple whiskers (Fig. S5). Although no studies have reported responses of PV cells to multiple whiskers in mice, our results are largely in agreement with earlier recordings from suspected interneurons in L2/3 of rabbit S1BF, where putative interneurons responded on average to ~5.5 whiskers56. Although PV cells often responded to multiple whiskers, one whisker, typically the principal whisker, evoked substantially larger responses compared to other whiskers. For example, the response magnitude of PV cells for the principal whisker was ~2-fold greater compared to surround whiskers (mean MI 2.2-fold for C1>D1 in the C1 barrel, or 1.7-fold for D1>C1 in the D1 barrel, Fig. 1h, Fig. S2d). This was reflected in relatively high selectivity indices for most PV cells for the principal whisker vs an adjacent surround whisker (SI=~0.37- 0.46 for PV cells, Fig. 2b)19. The high selectivity of these cells may be a result of convergent bottom up inputs from L4, which make up the majority of pre-synaptic inputs onto PV cells57. In contrast, smaller differences in selectivity indices and magnitude of response evoked by deflection of the principal whisker compared to surround whisker have been reported for Pyr cells12,13,19,51. These results, suggesting relatively narrow tuning of PV cells for specific whiskers, stand in contrast to the more broad orientation tuning of PV cells compared to Pyr cells in V158– 60. On the other hand, while S1BF PV cells may be narrowly tuned to whisker identity, they may be more broadly tuned to other properties of whisker deflection, such as direction61,62. We next determined how PV cell responsivity changes with experience by implementing a chronic whisker trimming paradigm. Whisker trimming is a well-established method for inducing experience dependent plasticity in S1BF. During whisker trimming, L2/3 Pyr cells shift their selectivity toward spared whiskers, with potentiation of responses to the spared whisker and reductions in response to the deprived whisker17,19. These changes are mediated by 1) strongly reduced responses for the trimmed whisker coupled with more modest reductions in responses for the spared whisker in cells with high whisker responsivity at baseline; and 2) small increases in responses to spared whiskers in cells with low whisker responsivity at baseline17,19. In deprived barrels, L4 PV-mediated feed-forward inhibition is acutely reduced for the trimmed whisker, mediating disinhibition and transient, homeostatic firing rate stabilization of L2/3 Pyr cells54,55. We hypothesized that a similar reduction in L2/3 PV-mediated inhibition might facilitate tuning changes in Pyr cells toward the spared whisker after trimming. However, we found that after trimming all whiskers except D1, the number of D1 responsive PV cells increased significantly in the deprived C1 barrel (Fig. 2c). This effect was barrel specific, as the number of D1 responsive PV cells did not change in the spared D1 barrel (Fig. 2c), although there was a .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint small increase in response magnitude in the D1 barrel after 18 days of trimming (Fig. 2d). Although there is significant representational drift in the barrel cortex13,63, the increase in D1 responsive PV cells is unlikely due to random drift as we did not see any increase in control animals without whisker trimming (Fig. S3). We tried to determine what factors predict which PV cells will maintain or switch responsivity after trimming. However, we found that these populations largely overlap in terms of baseline activity metrics using PCA (Fig. 2m), and a GLME model incorporating baseline neural activity metrics achieved only modest performance (Fig. 2n). Baseline response magnitude to the C1 and D1 whiskers contributed most to the PCA and GLME, respectively, but a significant portion of the variability remains unexplained. Prior studies have found that horizontal intracortical connections from excitatory pyramidal cells in spared barrels to deprived barrels increase after whisker trimming64–66. This could account for the increase in PV cells responsive to the spared whisker we observed in the deprived barrels, though future work will be required to directly test this. After whisker regrowth, expanded whisker cortical map representations contract back to pre- trimming sizes15. Less is known about changes in whisker evoked neuronal activity after whisker regrowth, but selectivity of Pyr cells for spared and trimmed whiskers also appears to return to pre-trimming levels19. For PV cells, we found changes in neuronal activity that persisted even after 28 days of whisker regrowth. In the deprived C1 barrel, selectivity was largely maintained, with only a small shift in SI toward the D1 whisker after regrowth (Fig. 3d). However, the number of responsive PV cells and evoked response magnitude were reduced across all trimmed whiskers and the spared D1 whisker compared to baseline (Fig. 3e-f). This suggests that during regrowth a reduction in the number of D1 responsive PV cells might be a compensatory response to restore neuronal whisker selectivity to the C1 whisker. On the other hand, in the spared D1 barrel, there was a strong, persistent shift in neuronal selectivity toward the D1 whisker (Fig. 3c). The percentage of D1 responsive neurons and response magnitude to D1 whisker stimulation were similar compared to baseline, but the number of C1 responsive neurons was strongly reduced. This suggests the persistent shift in selectivity is driven more strongly by reductions in responses to the trimmed whiskers rather than persistent potentiation of responses to the spared whisker. Reductions in PV activity may be necessary to facilitate cortical map restoration after whisker regrowth, but future manipulation studies will be required to test a potential causal role for PV activity in this process. The recruitment of PV cells responsive to the spared whisker during whisker trimming could reflect either a passive homeostatic response to changes in the activity of local Pyr cells and/or bottom-up inputs. Alternatively, the recruitment of PV cells could contribute causally to experience dependent cortical remapping, either by facilitating or limiting this process. In S1BF, brief optogenetic inhibition of PV cells increases whisker evoked activity of Pyr cells, locally and in surround barrels32,33. However, few studies have reported how more prolonged changes in PV cell activity affect sensory evoked responses and experience dependent plasticity. To test this, we first expressed chemogenetic DREADDs in PV cells and recorded the acute effects on the activity of local PV and Pyr cells. Acute modulation of PV cells led to substantial changes in the number of whisker responsive Pyr cells (Fig. 4a, d). As expected, sensory evoked responses of responder Pyr cells were strongly reduced by activation of PV cells (Fig. 4f). Interestingly, the number of responder PV cells and the magnitude of sensory evoked responses in those cells was also decreased after chemogenetic activation of PV cells (Fig. 4d-e). This may be due to the strong reciprocal connectivity between individual PV cells and local Pyr cells. For example, activation of PV cells could reduce Pyr->PV excitation or increase PV->PV inhibition67,68, leading to the modest reductions in PV cell activity we observed. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Although inhibition of PV cells increased the number of whisker responsive Pyr cells as expected, whisker evoked responses in those responder Pyr cells was reduced overall, with a shorter latency to peak onset (Fig. 4a-c). This shift in the temporal profile of these responses suggests engagement of compensatory inhibition from other interneuron subtypes. Compared to PV cells, SST cells have longer latency responses to whisker touches69. PV to SST connectivity has not been studied in S1BF, but monosynaptic inputs from PV to SST cells have been identified in other cortical regions70,71. Therefore, we hypothesized that changes in PV activity might modulate the activity of SST cells, which would in turn affect Pyr cell activity. Consistent with this idea, we found that activation of PV cells strongly suppressed SST cell activity (Fig. 5c- d). Conversely, inhibition of PV cells led to a substantial increase in the number of whisker responsive SST cells (Fig. 5a). The magnitude of whisker evoked responses in SST cells was smaller and the latency to peak was shifted earlier, similar to the effects observed in Pyr cells (Fig. 4c, 5b). Thus, inhibition of PV cells leads to a larger pool of more weakly responding SST and Pyr cells. In visual cortex, coordinated activity of groups of SST cells is required for suppression of stimulus evoked responses72. This suggests that in the setting of reduced PV cell activity SST cells may be coordinately recruited to compensate for reduced PV-mediated inhibition. Our study has some limitations. We used currently available genetically encoded calcium indicators, including GCaMP673 and jGCaMP840 to monitor neuronal activity. These indicators, imaged with framerates used in this study, likely cannot discriminate individual action potentials during high firing rates as in fast-spiking PV cells. However, even without single-spike resolution, our work here, and multiple prior studies, demonstrate these indicators can reliably detect changes in neuronal activity in PV cells74–78. We focused on L2/3 in this study as supragranular layers maintain capacity for cortical plasticity into adulthood16. It is likely there may be other layer-specific changes in PV cell activity that contribute to experience-dependent plasticity. This question will need to be explored further in future studies. Stimulus evoked Pyr cell responses for deprived and spared whiskers are differentially affected by trimming19,79, so it is likely PV responses to the deprived whisker would be different compared to the spared whisker. We saw such differential responses to whisker regrowth, but in our chronic whisker trimming experiment, we were only able to stimulate the spared whisker during trimming as we kept deprived whiskers trimmed flushed with the vibrissal pad. Our DREADD manipulation experiments demonstrate that changes in PV cell activity strongly influence the activity of SST interneurons. Direct PV-to-SST connectivity has been demonstrated in other areas of the cortex, but future paired electrophysiology recordings will be required to determine if effects of PV cells on SST cells are monosynaptic or indirect. Despite these limitations, our work reveals, for the first time, how PV cell activity changes dynamically during sensory evoked experience dependent plasticity. PV cells responsive to the spared whisker are recruited over time in a spatially-specific manner, parallelling changes previously observed in Pyr cells. This recruitment can be reversed as sensory experience changes, though alterations in individual PV cell selectivity can persist for weeks after whisker regrowth. Our chemogenetic manipulation experiments demonstrate how changes in PV cell activity influence the activity of other local PV and Pyr cells. These changes can be difficult to predict, highlighting the reciprocal connectivity of PV and Pyr cells in local microcircuits and underscoring the importance of directly measuring the effects of chemogenetic manipulations on circuit activity. These experiments also revealed a novel and strong functional connectivity between PV cells and SST cells in S1BF. Interestingly, reducing SST cell activity in deprived barrels has recently been shown to block whisker trimming induced map expansion of the spared whisker80. Together with our results, we speculate that recruitment of PV cells to the spared whisker during whisker trimming might constrain remapping, both through direct .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint increases in PV-mediated inhibition of Pyr cells, but also indirectly through inhibition of SST cells. Future experiments will be required to test these possibilities directly. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figures Figure 1. The spatial distribution of whisker evoked responses of PV cells in the S1BF. a. Schematic of the imaging FOV and stimulated whiskers (C1, magenta; D1, blue; B2, green; E3, orange). b. Average projection image of a representative FOV of PV cells expressing GCaMP6s. Scale bar=100 µm. c. Z-scored calcium fluorescence traces from PV cells during whisker stimulation. Gray bars indicate epochs of whisker deflections (10 Hz, 1 s). Scale Y-axis, .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Z-score; X-axis, time (seconds). d. Percentage of whisker responsive cells to the indicated whiskers (n=11 mice for C1 and D1 whiskers; n=6 mice for B2 and E3 whiskers). GLME binomial model, fixed effect of whisker, with significance for individual coefficients compared to C1, corrected using Benjamini and Hochberg’s method, indicated over corresponding data points (*, p<0.05; **, p<0.01; ***, p<0.001). e-g. Mean evoked calcium traces for all cells (e, n=210 cells for C1 and D1, n=100 cells for B2 and E3), responders (f, n=183 cells for C1, 101 cells for D1, 39 cells for B2, and 5 cells for E3), and non-responders (g, n=27 cells for C1, 109 cells for D1, 61 cells for B2, and 95 cells for E3). h-i. Modulation index (MI) calculated for all responders (h, n=183 cells for C1, 101 cells for D1, 39 cells for B2, and 5 cells for E3) and non- responders (i, n=27 cells for C1, 109 cells for D1, 61 cells for B2, and 95 cells for E3) by whisker stimulated. LME model, fixed effect of whisker. Significance for individual coefficients compared to C1, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (*, p<0.05; **, p<0.01; ***, p<0.001). j. Spatial distribution of all PV cells plotted according to relative position in the FOV centered on the C1 barrel. Responders are colored according to the indicated whisker, non-responders are colored in gray. Size of the circle corresponds to the MI of the cell for the indicated whisker. Scale bar=50 µm. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure 2. PV cell responses to the spared whisker are increased in the deprived barrel during whisker trimming. a. Schematic of the imaging FOV, stimulated whisker, and trimming timeline. b. Histogram of the selectivity index of cells in the C1 (magenta) or D1 (blue) barrel at baseline (n=189 cells for D1 and 210 cells for C1). LME model, fixed effect of barrel (***, p<0.001). c. Percentage of D1 responsive PV cells in the C1 (magenta) or D1 (blue) barrel over time (n=11 mice). GLME binomial model for the D1 barrel, fixed effect of timepoint (p=0.41). GLME binomial model for the C1 barrel, fixed effect of timepoint (p<0.001). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (*, p<0.05; **, p<0.01). d. MI of D1 responsive PV cells in the C1 (magenta) or D1 (blue) barrel over time (n=139/101, 154/83, 149/116, 143/135, 141/126, and 141/70 at day 0, 3, 6, 12, 18, and 46 for the D1/C1 barrels, respectively). LME for the D1 barrel, fixed effect of timepoint (p=0.03). LME for the C1 barrel, fixed effect of timepoint (p=0.36). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (*, p<0.05). e-f. Sankey diagrams showing responsiveness of individual PV cells to the D1 whisker in the D1 (e, n=189 cells) or C1 (f, n=210 cells) barrels between days 0 and 12. g. MI of PV cells in the D1 barrel to the D1 whisker at day 0. Cells are grouped by switch category: non-responsive to the D1 whisker at day 0 and day 12 (NR-NR, n=35); switched from non-responsive to responsive (NR- R, n=15); switched from responsive to non-responsive (R-NR, n=11); or responsive on both day 0 and day 12 (R-R, n=128). Kruskal-Wallis test (**, p<0.01; ***, p<0.001). h. MI of PV cells in the D1 barrel to the C1 whisker at day 0, by switch category (n=35, 15, 11, and 128 cells for NR-NR, NR-R, R-NR, and R-R, respectively). Kruskal-Wallis test (**, p<0.01; ***, p<0.001). i. Selectivity index of PV cells in the D1 barrel at day 0, by switch category (n=35, 15, 11, and 128 cells for NR-NR, NR-R, R-NR, and R-R, respectively). Kruskal-Wallis test. j. MI of PV cells in the C1 barrel to the D1 whisker at day 0, by switch category (n=62, 47, 13, and 88 cells for NR-NR, NR-R, R-NR, and R-R, respectively). Kruskal-Wallis test (**, p<0.01; ***, p<0.001). k. MI of PV cells in the C1 barrel to the C1 whisker at day 0, by switch category (n=62, 47, 13, and 88 cells for NR-NR, NR-R, R-NR, and R-R, respectively). Kruskal-Wallis test (*, p<0.05; ***, p<0.001). i. Selectivity index of PV cells in the C1 barrel at day 0, by switch category (n=62, 47, 13, and 88 cells for NR-NR, NR-R, R-NR, and R-R, respectively). Kruskal-Wallis test (**, p<0.01; ***, p<0.001). m. Plot of the first and second principal components from analysis of non-responder PV cells at day 0. Cells are group by switch category (NR-NR, dark gray; NR-R, light gray). n. Receiver-operating characteristic curve for a GLME model of switch category (NR-NR or NR-R) of day 0 non-responder PV cells (AUC=0.77). The data for PCA and the GLME model included barrel location, response magnitude to C1 and D1 whisker deflection, selectivity index, and the mean Z-score of the spontaneous activity trace for each. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure 3. PV cell selectivity shifts toward the spared whisker in the spared barrel even after whisker regrowth. a-b. Sankey diagrams showing responsiveness of individual PV cells to the D1 whisker in the D1 (a) or C1 (b) barrels between days 0, 12, and 46. c-d. Histogram of the selectivity index of cells in the D1 (c, n=189 cells) or C1 (d, n=210 cells) barrel at day 0 (gray) or day 46 (blue or magenta, respectively). LME model, fixed effect of timepoint (*, p<0.05; ***, p<0.001). e. Change in the percentage of PV cells responsive to the C1 (magenta), D1 (blue), B2 (green), or E3 (orange) whisker in the D1 (left) or C1 (right) barrel between day 0 and day 46 (n=11 mice for C1 and D1 whiskers; n=6 mice for B2 and E3 whiskers). D1 barrel, GLME binomial model, ANOVA for fixed effects of whisker (C1 vs. D1 whisker at day 0, p<0.001), timepoint (day 0 vs day 46 for C1, ###, p<0.001), and whisker*timepoint interaction (**, p<0.01). C1 barrel, GLME binomial model, ANOVA for fixed effects of whisker (C1 vs. other whiskers at day 0, p<0.001), timepoint (day 0 vs day 46 for C1, ##, p<0.01), and whisker*timepoint interaction (**, p<0.01), with significance for individual coefficients for whisker*timepoint interactions, corrected using Benjamini and Hochberg’s method, indicated over corresponding data points (**, p<0.01). f. Change in the MI of PV cells responsive to the C1 (magenta), D1 (blue), B2 (green), or E3 (orange) whisker in the D1 (left) or C1 (right) barrel between day 0 and day 46 (D1 barrel: n=189 cells for C1 and D1 whiskers. C1 barrel: n=210 cells for C1 and D1 whiskers; n=100 cells for B2 and E3 whiskers). D1 barrel, GLME binomial model, ANOVA for fixed effects of whisker (C1 vs. D1 whisker at day 0, p<0.001), timepoint (day 0 vs day 46 for C1, p<0.07), and whisker*timepoint interaction (*, p<0.01). C1 barrel, LME model, ANOVA for fixed effects of whisker (C1 vs. other whiskers at day 0, p<0.001), timepoint (day 0 vs day 46 for C1, ###, p<0.001), and whisker*timepoint interaction (*, p<0.05), with significance for individual coefficients for whisker*timepoint interactions, corrected using Benjamini and Hochberg’s method, indicated over corresponding data points (**, p<0.01). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure 4. Acute chemogenetic modulation of PV cells affects whisker evoked activity of local PV and Pyr cells. a. Percentage of PV and Pyr cells responsive to principal whisker stimulation in mice (n=8) expressing hM4D(Gi) in PV cells before and after C21 (5 mg/kg, i.p.). Paired t-test comparing mice before and after C21 (PV, p=0.93; *, Pyr, p<0.05). b. Mean evoked calcium trace for responder PV cells before (n=108 cells) and after (n=108 cells) C21. LME model for the MI, fixed effect of condition (***, p<0.001). c. Mean evoked calcium trace for responder Pyr cells before (n=546 cells) and after (n=674 cells) C21. LME model for the MI, fixed effect of condition (***, p=0.001). d. Percentage of PV and Pyr cells responsive to principal whisker stimulation in mice (n=7) expressing hM3D(Gq) in PV cells before and after C21. Paired t-test comparing mice before and after C21 (PV, **, p<0.01; Pyr, ***, p<0.001). e. Mean evoked calcium trace for responder PV cells before (n=75 cells) and after (n=35 cells) C21. LME model for the MI, fixed effect of condition (***, p<0.001). f. Mean evoked calcium trace for responder Pyr cells before (n=377 cells) and after (n=79 cells) C21. LME model for the MI, fixed effect of condition (***, p<0.001). g. Percentage of PV and Pyr cells responsive to principal whisker stimulation in mice (n=6) expressing mCherry in PV cells before and after C21. Paired t-test comparing mice before and after C21, (PV, p=0.44; Pyr, p=0.66). h. Mean evoked calcium trace for responder PV cells before (n=59 cells) and after (n=55 cells) C21. LME model for the MI, fixed effect of condition (p=0.19). i. Mean evoked calcium trace for responder Pyr cells before (n=277 cells) and after (n=313 cells) C21. LME model for the MI, fixed effect of condition (p=0.89). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure 5. Acute chemogenetic modulation of PV cells affects whisker evoked activity of local SST cells. a. Percentage of SST cells responsive to principal whisker stimulation in mice (n=16) expressing hM4D(Gi) in PV cells before and after C21 (5 mg/kg, i.p.). Paired t-test comparing mice before and after C21, (**, p<0.01). b. Mean evoked calcium trace for responder SST cells before (n=44 cells) and after (n=97 cells) C21. LME model for the MI, fixed effect of condition (**, p<0.01). c. Percentage of SST cells responsive to principal whisker stimulation in mice (n=14) expressing hM3D(Gq) in PV cells before and after C21. Wilcoxin signed rank test comparing mice before and after C21 (***, p<0.001). d. Mean evoked calcium trace for responder SST cells before (n=84 cells) and after (n=22 cells) C21. LME model for the MI, fixed effect of condition (*, p<0.05). e. Percentage of SST cells responsive to principal whisker stimulation in mice (n=12) expressing mCherry in PV cells before and after C21. Paired t-test comparing mice before and after C21 (p=0.98). f. Mean evoked calcium trace for responder SST cells before (n=40 cells) and after (n=40 cells) C21. LME model for the MI, fixed effect of condition (p=0.72). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S1. Whisker evoked responses in PV cells are similar in awake and lightly anesthetized mice. a. Percentage of whisker responsive PV cells in the D1 barrel to D1 whisker deflection in lightly anesthetized (gray) or awake (red) mice (n=6). Whisker deflection stimuli were delivered at 10 Hz for 1 second for a total of 20 trials using either square (Sq.) wave stimuli with 0.01 or 0.001 s rise times or sine wave stimuli with voltage amplitudes of 2, 4, 6, or 8 V. GLME binomial model, ANOVA for fixed effects of anesthesia condition (C, p=0.09), stimulus type (S, ***, p<0.001), and condition*stimulus interaction (C*S, #, p<0.05), with significance for individual coefficients for stimulus type and condition*stimulus interactions, corrected using Benjamini and Hochberg’s method, indicated over corresponding data points (S: *, p<0.05; ***, p<0.001. C*S: ##, p<0.01). b. Mean evoked calcium trace for responder PV cells to square wave whisker deflections with 0.001 s rise time in lightly anesthetized (gray, n=61 cells) or awake (red, n=75 cells) mice. c. MI of responder PV cells to square wave whisker deflections with 0.001 s rise time in lightly anesthetized (gray, n=61 cells) or awake (red, n=75 cells) mice. LME model for the MI, fixed effect of condition (p=0.64). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S2. The spatial distribution of whisker evoked responses of PV cells in the D1 barrel. a. Percentage of whisker responsive cells to the indicated whiskers (n=11 mice). GLME binomial model, fixed effect of whisker (***, p<0.001). b. Mean evoked calcium traces for responders (n=67 cells for C1, 139 cells for D1). c. Mean evoked calcium traces for non- responders (n=122 cells for C1, 50 cells for D1). d. MI calculated for responders (n=67 cells for C1, 139 cells for D1). LME model, fixed effect of whisker (**, p<0.01). e. MI calculated for non- responders (n=122 cells for C1, 50 cells for D1). LME model, fixed effect of whisker (p=0.47). f. Spatial distribution of all PV cells plotted according to relative position in the FOV centered on the D1 barrel. Responders are colored according to the indicated whisker, non-responders are colored in gray. Size of the circle corresponds to the MI of the cell for the indicated whisker. Scale bar=50 µm. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S3. PV cell responses do not increase in control animals without whisker trimming. a. Percentage of D1 responsive PV cells in the D1 (blue) or C1 (magenta) barrel over time (n=5 mice). GLME binomial model for the D1 barrel, fixed effect of timepoint (p=0.29). GLME binomial model for the C1 barrel, fixed effect of timepoint (**, p<0.01). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (*, p<0.05). b. MI of D1 responsive PV cells in the D1 (blue) or C1 (magenta) barrel over time (n=92/49, 90/24, and 98/49 at day 0, 18, and 46 for the D1/C1 barrels, respectively). LME for the D1 barrel, fixed effect of timepoint (p=0.05). LME for the C1 barrel, fixed effect of timepoint (p=0.14). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S4. Spontaneous activity changes in PV cells during whisker trimming and regrowth. a. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from PV cells recorded in the D1 (blue) or C1 (magenta) barrels over time (n=189 cells for D1 and 210 cells for C1 from 11 mice). LME for the D1 barrel, fixed effect of timepoint (###, p<0.001). LME for the C1 barrel, fixed effect of timepoint (***, p<0.001). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (###, p<0.001; *, p<0.05; **, p<0.01; ***, p<0.001). b. Frequency of calcium transient peaks in the D1 (blue) or C1 (magenta) barrels over time (n=189 cells for D1 and 210 cells for C1 from 11 mice). LME for the D1 barrel, fixed effect of timepoint (##, p<0.01). LME for the C1 barrel, fixed effect of timepoint (p=0.06). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (#, p<0.05). c. Mean amplitude of calcium transient peaks in the D1 (blue) or C1 (magenta) barrels over time (n=189 cells for D1 and 210 cells for C1 from 11 mice). LME for the D1 barrel, fixed effect of timepoint (###, p<0.001). LME for the C1 barrel, fixed effect of timepoint (***, p=0.001). Significance for individual coefficients compared to day 0, corrected using Benjamini and Hochberg’s method, are indicated over corresponding data points (*, p<0.05; ***, p<0.001). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S5. Changes in tuning width of PV cells before and after whisker trimming and regrowth. a. Tuning curves of individual PV cells (n=100 cells) in the C1 barrel to the C1, D1, B2, and E3 whiskers at baseline. Modulation indices for each whisker were normalized to the response magnitude of the preferred whisker, sorted in descending order, and fit with a single- term exponential decay. b. Tuning curves of individual PV cells (n=100 cells) in the C1 barrel to the C1, D1, B2, and E3 whiskers at day 46 after whisker trimming and regrowth. c. Histogram of tuning curve decay constants from PV cells (n=100 cells) in the C1 barrel at day 0 (gray) or day 46 (magenta). LME model, fixed effect of timepoint (*, p<0.05). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S6. Acute chemogenetic modulation of PV cells affects spontaneous activity of local PV and Pyr cells. a. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from PV and Pyr cells in mice expressing hM4D(Gi) in PV cells before and after C21 (n=132/1017 and 132/910 PV/Pyr cells before and after C21, respectively, from 8 mice). LME model for PV cells, fixed effect of timepoint (p=0.41). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). b. Frequency of calcium transient peaks from PV and Pyr cells before and after C21 (n=132/1017 and 132/910 PV/Pyr cells before and after C21, respectively, from 8 mice). LME model for PV cells, fixed effect of timepoint (***, p<0.001). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). c. Mean amplitude of calcium transient peaks from PV and Pyr cells before and after C21 (n=132/1017 and 132/910 PV/Pyr cells before and after C21, respectively, from 8 mice). LME model for PV cells, fixed effect of timepoint (***, p<0.001). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). d. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from PV and Pyr cells in mice expressing hM3D(Gq) in PV cells before and after C21 (n=92/774 and 92/638 PV/Pyr cells before and after C21, respectively, from 7 mice). LME model for PV cells, fixed effect of timepoint (p=0.31). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). e. Frequency of calcium transient .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint peaks from PV and Pyr cells before and after C21 (n=92/774 and 92/638 PV/Pyr cells before and after C21, respectively, from 7 mice). LME model for PV cells, fixed effect of timepoint (p=0.06). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). f. Mean amplitude of calcium transient peaks from PV and Pyr cells before and after C21 (n=92/774 and 92/638 PV/Pyr cells before and after C21, respectively, from 7 mice). LME model for PV cells, fixed effect of timepoint (p=0.78). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). g. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from PV and Pyr cells in mice expressing mCherry in PV cells before and after C21 (n=71/596 and 71/673 PV/Pyr cells before and after C21, respectively, from 6 mice). LME model for PV cells, fixed effect of timepoint (p=0.71). LME model for Pyr cells, fixed effect of timepoint (***, p<0.001). h. Frequency of calcium transient peaks from PV and Pyr cells before and after C21 (n=71/596 and 71/673 PV/Pyr cells before and after C21, respectively, from 6 mice). LME model for PV cells, fixed effect of timepoint (p=0.67). LME model for Pyr cells, fixed effect of timepoint (**, p<0.01). i. Mean amplitude of calcium transient peaks from PV and Pyr cells before and after C21 (n=71/596 and 71/673 PV/Pyr cells before and after C21, respectively, from 6 mice). LME model for PV cells, fixed effect of timepoint (*, p<0.05). LME model for Pyr cells, fixed effect of timepoint (p=0.24). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint Figure S7. Acute chemogenetic modulation of PV cells affects spontaneous activity of local SST cells. a. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from SST cells in mice expressing hM4D(Gi) in PV cells before and after C21 (n=173 cells from 16 mice). LME model, fixed effect of timepoint (**, p<0.01). b. Frequency of calcium transient peaks before and after C21 (n=173 cells from 16 mice). LME model, fixed effect of timepoint (***, p<0.001). c. Mean amplitude of calcium transient peaks before and after C21 (n=173 cells from 16 mice). LME model, fixed effect of timepoint (p=0.1). d. Total area-under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from SST cells in mice expressing hM3D(Gq) in PV cells before and after C21 (n=137 cells from 14 mice). LME model, fixed effect of timepoint (*, p<0.05). e. Frequency of calcium transient peaks before and after C21 (n=137 cells from 14 mice). LME model, fixed effect of timepoint (***, p<0.001). f. Mean amplitude of calcium transient peaks before and after C21 (n=137 cells from 14 mice). LME model, fixed effect of timepoint (***, p<0.001). g. Total area- under-the-curve (AUC) of spontaneous activity (no whisker deflection) calcium traces from SST .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint cells in mice expressing mCherry in PV cells before and after C21 (n=80 cells from 12 mice). LME model, fixed effect of timepoint (p=0.33). h. Frequency of calcium transient peaks before and after C21 (n=80 cells from 12 mice). LME model, fixed effect of timepoint (*, p<0.05). i. Mean amplitude of calcium transient peaks before and after C21 (n=80 cells from 12 mice). LME model, fixed effect of timepoint (p=0.68). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint

1. Hooks, B. M. & Chen, C. Circuitry Underlying Experience-Dependent Plasticity in the Mouse Visual System. Neuron 106, 21–36 (2020). 2. Barth, A. L. & Ray, A. Progressive Circuit Changes during Learning and Disease. Neuron 104, 37–46 (2019). 3. Chéreau, R., Williams, L. E., Bawa, T. & Holtmaat, A. Circuit mechanisms for cortical plasticity and learning. Semin Cell Dev Biol 125, 68–75 (2022). 4. Hübener, M. & Bonhoeffer, T. Neuronal plasticity: beyond the critical period. Cell 159, 727– 737 (2014). 5. Diamond, M. E. & Arabzadeh, E. Whisker sensory system - from receptor to decision. Prog. Neurobiol. 103, 28–40 (2013). 6. Woolsey, T. A. & Van der Loos, H. The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res 17, 205–242 (1970). 7. Petersen, C. C. H. Sensorimotor processing in the rodent barrel cortex. Nat. Rev. Neurosci. 20, 533–546 (2019). 8. Oberlaender, M. et al. Cell type-specific three-dimensional structure of thalamocortical circuits in a column of rat vibrissal cortex. Cereb Cortex 22, 2375–2391 (2012). 9. Petersen, C. C. H., Grinvald, A. & Sakmann, B. Spatiotemporal dynamics of sensory responses in layer 2/3 of rat barrel cortex measured in vivo by voltage-sensitive dye imaging combined with whole-cell voltage recordings and neuron reconstructions. J Neurosci 23, 1298–1309 (2003). 10. Narayanan, R. T. et al. Beyond Columnar Organization: Cell Type- and Target Layer-Specific Principles of Horizontal Axon Projection Patterns in Rat Vibrissal Cortex. Cereb Cortex 25, 4450–4468 (2015). 11. Fox, K., Wright, N., Wallace, H. & Glazewski, S. The Origin of Cortical Surround Receptive Fields Studied in the Barrel Cortex. J Neurosci 23, 8380–8391 (2003). 12. Clancy, K. B., Schnepel, P., Rao, A. T. & Feldman, D. E. Structure of a single whisker representation in layer 2 of mouse somatosensory cortex. J. Neurosci. 35, 3946–3958 (2015). 13. Wang, H. C., LeMessurier, A. M. & Feldman, D. E. Tuning instability of non-columnar neurons in the salt-and-pepper whisker map in somatosensory cortex. Nat Commun 13, 6611 (2022). 14. Fox, K. Anatomical pathways and molecular mechanisms for plasticity in the barrel cortex. Neuroscience 111, 799–814 (2002). 15. Polley, D. B., Chen-Bee, C. H. & Frostig, R. D. Two directions of plasticity in the sensory- deprived adult cortex. Neuron 24, 623–637 (1999). 16. Fox, K. A critical period for experience-dependent synaptic plasticity in rat barrel cortex. J Neurosci 12, 1826–1838 (1992). 17. Glazewski, S. & Fox, K. Time course of experience-dependent synaptic potentiation and depression in barrel cortex of adolescent rats. J. Neurophysiol. 75, 1714–1729 (1996). 18. Vasquez, B., Campos, B., Cao, A., Theint, A. T. & Zeiger, W. High-Sensitivity Intrinsic Optical Signal Imaging Through Flexible, Low-Cost Adaptations of an Upright Microscope. eNeuro 10, ENEURO.0046-23.2023 (2023). 19. Margolis, D. J. et al. Reorganization of cortical population activity imaged throughout long- term sensory deprivation. Nat. Neurosci. 15, 1539–1546 (2012). 20. Gainey, M. A. & Feldman, D. E. Multiple shared mechanisms for homeostatic plasticity in rodent somatosensory and visual cortex. Philos. Trans. R. Soc. Lond., B, Biol. Sci. 372, (2017). 21. Margolis, D. J., Lütcke, H. & Helmchen, F. Microcircuit dynamics of map plasticity in barrel cortex. Curr Opin Neurobiol 24, 76–81 (2014). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint 22. Xu, X., Roby, K. D. & Callaway, E. M. Immunochemical characterization of inhibitory mouse cortical neurons: three chemically distinct classes of inhibitory cells. J Comp Neurol 518, 389–404 (2010). 23. Gonchar, Y., Wang, Q. & Burkhalter, A. Multiple distinct subtypes of GABAergic neurons in mouse visual cortex identified by triple immunostaining. Front Neuroanat 1, 3 (2007). 24. Tamamaki, N. et al. Green fluorescent protein expression and colocalization with calretinin, parvalbumin, and somatostatin in the GAD67-GFP knock-in mouse. J Comp Neurol 467, 60–79 (2003). 25. Celio, M. R. Parvalbumin in most gamma-aminobutyric acid-containing neurons of the rat cerebral cortex. Science 231, 995–997 (1986). 26. Kawaguchi, Y., Katsumaru, H., Kosaka, T., Heizmann, C. W. & Hama, K. Fast spiking cells in rat hippocampus (CA1 region) contain the calcium-binding protein parvalbumin. Brain Res 416, 369–374 (1987). 27. Kawaguchi, Y. & Kubota, Y. Correlation of physiological subgroupings of nonpyramidal cells with parvalbumin- and calbindinD28k-immunoreactive neurons in layer V of rat frontal cortex. J Neurophysiol 70, 387–396 (1993). 28. Kawaguchi, Y. Physiological subgroups of nonpyramidal cells with specific morphological characteristics in layer II/III of rat frontal cortex. J Neurosci 15, 2638–2655 (1995). 29. Katsumaru, H., Kosaka, T., Heizmann, C. W. & Hama, K. Immunocytochemical study of GABAergic neurons containing the calcium-binding protein parvalbumin in the rat hippocampus. Exp Brain Res 72, 347–362 (1988). 30. Czeiger, D. & White, E. L. Comparison of the distribution of parvalbumin-immunoreactive and other synapses onto the somata of callosal projection neurons in mouse visual and somatosensory cortex. J Comp Neurol 379, 198–210 (1997). 31. Di Cristo, G. et al. Subcellular domain-restricted GABAergic innervation in primary visual cortex in the absence of sensory and thalamic inputs. Nat Neurosci 7, 1184–1186 (2004). 32. Yeganeh, F. et al. Effects of optogenetic inhibition of a small fraction of parvalbumin-positive interneurons on the representation of sensory stimuli in mouse barrel cortex. Sci Rep 12, 19419 (2022). 33. Yang, J.-W. et al. Optogenetic Modulation of a Minor Fraction of Parvalbumin-Positive Interneurons Specifically Affects Spatiotemporal Dynamics of Spontaneous and Sensory- Evoked Activity in Mouse Somatosensory Cortex in Vivo. Cereb. Cortex 27, 5784–5803 (2017). 34. Atallah, B. V., Bruns, W., Carandini, M. & Scanziani, M. Parvalbumin-expressing interneurons linearly transform cortical responses to visual stimuli. Neuron 73, 159–170 (2012). 35. Cardin, J. A. et al. Driving fast-spiking cells induces gamma rhythm and controls sensory responses. Nature 459, 663–667 (2009). 36. Rupert, D. D. & Shea, S. D. Parvalbumin-Positive Interneurons Regulate Cortical Sensory Plasticity in Adulthood and Development Through Shared Mechanisms. Front Neural Circuits 16, 886629 (2022). 37. Reh, R. K. et al. Critical period regulation across multiple timescales. Proc Natl Acad Sci U S A 117, 23242–23251 (2020). 38. Kuhlman, S. J. et al. A disinhibitory microcircuit initiates critical-period plasticity in the visual cortex. Nature 501, 543–546 (2013). 39. Cisneros-Franco, J. M. & de Villers-Sidani, É. Reactivation of critical period plasticity in adult auditory cortex through chemogenetic silencing of parvalbumin-positive interneurons. Proc Natl Acad Sci U S A 116, 26329–26331 (2019). 40. Zhang, Y. et al. Fast and sensitive GCaMP calcium indicators for imaging neural populations. Nature (2023) doi:10.1038/s41586-023-05828-9. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint 41. Fenno, L. E. et al. Comprehensive Dual- and Triple-Feature Intersectional Single-Vector Delivery of Diverse Functional Payloads to Cells of Behaving Mammals. Neuron 107, 836- 853.e11 (2020). 42. Krashes, M. J. et al. Rapid, reversible activation of AgRP neurons drives feeding behavior in mice. J Clin Invest 121, 1424–1428 (2011). 43. Hippenmeyer, S. et al. A developmental switch in the response of DRG neurons to ETS transcription factor signaling. PLoS Biol 3, e159 (2005). 44. Daigle, T. L. et al. A Suite of Transgenic Driver and Reporter Mouse Lines with Enhanced Brain-Cell-Type Targeting and Functionality. Cell 174, 465-480.e22 (2018). 45. Holtmaat, A. et al. Long-term, high-resolution imaging in the mouse neocortex through a chronic cranial window. Nat Protoc 4, 1128–1144 (2009). 46. Mostany, R. & Portera-Cailliau, C. A craniotomy surgery procedure for chronic brain imaging. J Vis Exp (2008) doi:10.3791/680. 47. Cantu, D. A. et al. EZcalcium: Open Source Toolbox for Analysis of Calcium Imaging Data. bioRxiv 2020.01.02.893198 (2020) doi:10.1101/2020.01.02.893198. 48. He, C. X. et al. Tactile Defensiveness and Impaired Adaptation of Neuronal Activity in the Fmr1 Knock-Out Mouse Model of Autism. J. Neurosci. 37, 6475–6487 (2017). 49. Pachitariu, M. et al. Suite2p: beyond 10,000 neurons with standard two-photon microscopy. 061507 Preprint at https://doi.org/10.1101/061507 (2017). 50. Thompson, K. J. et al. DREADD Agonist 21 Is an Effective Agonist for Muscarinic-Based DREADDs in Vitro and in Vivo. ACS Pharmacol. Transl. Sci. (2018) doi:10.1021/acsptsci.8b00012. 51. Zeiger, W. A. et al. Barrel cortex plasticity after photothrombotic stroke involves potentiating responses of pre-existing circuits but not functional remapping to new circuits. Nat Commun 12, 3972 (2021). 52. Yu, Z. et al. Beyond t test and ANOVA: applications of mixed-effects models for more rigorous statistical analysis in neuroscience research. Neuron 110, 21–35 (2022). 53. Benjamini, Y. & Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological) 57, 289–300 (1995). 54. Gainey, M. A., Aman, J. W. & Feldman, D. E. Rapid Disinhibition by Adjustment of PV Intrinsic Excitability during Whisker Map Plasticity in Mouse S1. J. Neurosci. 38, 4749–4761 (2018). 55. Li, L., Gainey, M. A., Goldbeck, J. E. & Feldman, D. E. Rapid homeostasis by disinhibition during whisker map plasticity. Proc. Natl. Acad. Sci. U.S.A. 111, 1616–1621 (2014). 56. Swadlow, H. A. Efferent neurons and suspected interneurons in S-1 vibrissa cortex of the awake rabbit: receptive fields and axonal properties. J Neurophysiol 62, 288–308 (1989). 57. Hafner, G. et al. Mapping Brain-Wide Afferent Inputs of Parvalbumin-Expressing GABAergic Neurons in Barrel Cortex Reveals Local and Long-Range Circuit Motifs. Cell Rep 28, 3450- 3461.e8 (2019). 58. Kerlin, A. M., Andermann, M. L., Berezovskii, V. K. & Reid, R. C. Broadly tuned response properties of diverse inhibitory neuron subtypes in mouse visual cortex. Neuron 67, 858– 871 (2010). 59. Liu, B. et al. Visual receptive field structure of cortical inhibitory neurons revealed by two- photon imaging guided recording. J Neurosci 29, 10520–10532 (2009). 60. Niell, C. M. & Stryker, M. P. Highly selective receptive fields in mouse visual cortex. J Neurosci 28, 7520–7536 (2008). 61. Guy, J., Möck, M. & Staiger, J. F. Direction selectivity of inhibitory interneurons in mouse barrel cortex differs between interneuron subtypes. Cell Rep 42, 111936 (2023). 62. Swadlow, H. A. & Gusev, A. G. Receptive-field construction in cortical inhibitory interneurons. Nat Neurosci 5, 403–404 (2002). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint 63. Dobler, Z. et al. Adapting and facilitating responses in mouse somatosensory cortex are dynamic and shaped by experience. Curr Biol 34, 3506-3521.e5 (2024). 64. Finnerty, G. T., Roberts, L. S. & Connors, B. W. Sensory experience modifies the short-term dynamics of neocortical synapses. Nature 400, 367–371 (1999). 65. Marik, S. A., Yamahachi, H., McManus, J. N. J., Szabo, G. & Gilbert, C. D. Axonal dynamics of excitatory and inhibitory neurons in somatosensory cortex. PLoS Biol 8, e1000395 (2010). 66. Albieri, G. et al. Rapid Bidirectional Reorganization of Cortical Microcircuits. Cereb Cortex 25, 3025–3035 (2015). 67. Selten, M. et al. Regulation of PV interneuron plasticity by neuropeptide-encoding genes. Nature (2025) doi:10.1038/s41586-025-08933-z. 68. Moore, A. K., Weible, A. P., Balmer, T. S., Trussell, L. O. & Wehr, M. Rapid Rebalancing of Excitation and Inhibition by Cortical Circuitry. Neuron 97, 1341-1355.e6 (2018). 69. Yu, J., Hu, H., Agmon, A. & Svoboda, K. Recruitment of GABAergic Interneurons in the Barrel Cortex during Active Tactile Behavior. Neuron 104, 412-427.e4 (2019). 70. Campagnola, L. et al. Local connectivity and synaptic dynamics in mouse and human neocortex. Science 375, eabj5861 (2022). 71. Pfeffer, C. K., Xue, M., He, M., Huang, Z. J. & Scanziani, M. Inhibition of inhibition in visual cortex: the logic of connections between molecularly distinct interneurons. Nat Neurosci 16, 1068–1076 (2013). 72. Ms, S., J, M.-Z., H, H. & T, T. Parvalbumin-expressing interneurons can act solo while somatostatin-expressing interneurons act in chorus in most cases on cortical pyramidal cells. Scientific reports 7, (2017). 73. Chen, T.-W. et al. Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499, 295–300 (2013). 74. Goel, A. et al. Impaired perceptual learning in a mouse model of Fragile X syndrome is mediated by parvalbumin neuron dysfunction and is reversible. Nat. Neurosci. 21, 1404– 1411 (2018). 75. Kourdougli, N. et al. Improvement of sensory deficits in fragile X mice by increasing cortical interneuron activity after the critical period. Neuron 111, 2863-2880.e6 (2023). 76. Znamenskiy, P. et al. Functional specificity of recurrent inhibition in visual cortex. Neuron 112, 991-1000.e8 (2024). 77. Garcia-Junco-Clemente, P., Tring, E., Ringach, D. L. & Trachtenberg, J. T. State-Dependent Subnetworks of Parvalbumin-Expressing Interneurons in Neocortex. Cell Rep 26, 2282- 2288.e3 (2019). 78. Khan, A. G. et al. Distinct learning-induced changes in stimulus selectivity and interactions of GABAergic interneuron classes in visual cortex. Nat Neurosci 21, 851–859 (2018). 79. House, D. R. C., Elstrott, J., Koh, E., Chung, J. & Feldman, D. E. Parallel regulation of feedforward inhibition and excitation during whisker map plasticity. Neuron 72, 819–831 (2011). 80. Dobrzanski, G., Zakrzewska, R., Kossut, M. & Liguz-Lecznar, M. Impact of somatostatin interneurons on interactions between barrels in plasticity induced by whisker deprivation. Sci Rep 12, 17992 (2022). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 28, 2025. ; https://doi.org/10.1101/2025.06.27.662081doi: bioRxiv preprint