Chronic chemogenetic activation of forebrain excitatory neurons in postnatal life evokes long-lasting changes in mood-related behavior

Early adversity is a key risk factor for the development of adult psychopathology, including anxiety, depression and schizophrenia. Rodent models of early adversity program persistent behavioral, molecular, metabolic, and neurophysiological changes. Perturbed signaling via forebrain Gq-coupled neurotransmitter receptors is a common feature across multiple models of early adversity. We addressed whether enhanced Gq-mediated signaling in forebrain excitatory neurons during postnatal life can evoke long-lasting mood-related behavioral changes. Excitatory hM3Dq DREADD-mediated chemogenetic activation of CamKIIα-positive forebrain excitatory neurons during postnatal life (P2-14) increased anxiety- and despair-like behavior, and evoked sensorimotor gating deficits in adulthood. In contrast, chronic chemogenetic hM3Dq DREADD activation of forebrain excitatory neurons in the juvenile or adult window did not evoke any mood-related behavioral alterations, highlighting the criticality of the postnatal temporal window. The enhanced anxiety-, despair- and schizophrenia-like behavioral changes evoked by chronic chemogenetic activation of forebrain excitatory neurons in postnatal life, was accompanied by an increased cortical and hippocampal metabolic rate of glutamatergic and GABAergic neurons in adulthood. Furthermore, animals with a history of postnatal hM3Dq activation exhibited a decline in the expression of activity-dependent and plasticity-associated markers within the hippocampus, along with perturbed hippocampal excitatory and inhibitory currents in adulthood. These results indicate that Gq signaling mediated activation of forebrain excitatory neurons during the critical postnatal window is sufficient to program altered mood-related behavior, as well as metabolic and neurophysiological changes in forebrain glutamate and GABA systems, recapitulating specific aspects of the consequences of early adversity.


Introduction
Early life experience plays a crucial role in the maturation and fine-tuning of neurocircuitry that drives emotional behavior in adulthood [1][2][3][4][5][6] . Both clinical and preclinical evidence indicates that early life adversity serves as a key risk factor for the development of adult psychopathology, increasing susceptibility to psychiatric disorders like anxiety, major depression and schizophrenia 5,[7][8][9][10] . Stressful experiences in adulthood can produce behavioral alterations that are often transient in nature, however perturbations in the vulnerable perinatal 'critical window' can program lasting changes in emotional behavior [11][12][13] . Several animal models have been used to study the persistent behavioral changes caused by early life perturbations, and have been instrumental in understanding specific underlying neural mechanisms involved in the programming of adult emotional behavior 10,11,[14][15][16][17][18][19] .
The prenatal period and the first few weeks after birth are marked by the establishment and functional maturation of several neurocircuits in rodent models, representing a critical period in which these circuits are particularly amenable to modification by environmental stimuli 2,5 . Rodent models of early life perturbations encompass those based on gestational stress 10 , maternal immune activation 8,9 , disruption of dam-pup interaction 20,21 or pharmacological treatments 22-24 , and exhibit both distinct and overlapping behavioral and physiological effects in adulthood 5,11 . Strikingly, a commonality noted across these animal models is the fact that multiple molecular, cellular, functional and behavioral changes often persist throughout the animal's lifespan 5,11,25 . Amongst the underlying mechanisms implicated in the establishment of such long-lasting changes in response to early life perturbations are a dysregulation of the hormonal stress response pathway 22,26-30 , serotonergic system 31-33 , and emergence of excitation-inhibition balance within key cortical neurocircuits 34,35 .
Common across several rodent models of early life perturbations are alterations in G protein-coupled receptor (GPCR) signaling, including via the serotonin 1A (5-HT 1A ) receptor 36-38 , serotonin 2A (5-HT 2A ) receptor 39-43 , metabotropic glutamate receptors 1 and 5 (mGluR1/5) 44,45 , muscarinic acetylcholine receptor 1 (M1) 46 and α 1 adrenergic receptors 47 . The emergence of aberrant emotional behavior in these animal models has been suggested to involve a key role for both excitatory Gq-coupled and inhibitory Gi-coupled GPCRs, in particular an appropriate balance of signaling between the Gq-coupled 5-HT 2A receptor and the Gi-coupled 5-HT 1A receptor in the forebrain has been hypothesized to be a critical determinant of the establishment of emotional behavior [48][49][50][51] . Enhanced signaling via the cortical 5-HT 2A receptor is thought to be one of the common features noted across distinct models of early life perturbations, including maternal separation 39,40 , postnatal fluoxetine 48  observation that the same perturbation performed in the juvenile window or in adulthood has no effect on mood-related behavior. Our findings provide evidence in support of the hypothesis that enhanced Gq signaling within forebrain excitatory neurons during the critical postnatal window is sufficient to evoke perturbed mood-related behavior in adulthood, recapitulating the enhanced vulnerability to psychopathology associated with early adversity.  (Table 1). We did note a trend towards a depolarizing shift in the resting membrane potential (RMP) in CA1 pyramidal neurons of the PNCNO treatment group as compared to their vehicle-treated controls ( Table 1, p = 0.06).

Selective expression and activation of hM3Dq DREADD in
We next sought to parcellate the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons in the postnatal window on excitatory and inhibitory neurotransmission. Whole-cell patch clamp analysis was carried out to measure sEPSCs and sIPSCs in CA1 pyramidal neurons in acute hippocampal slices derived from bigenic mouse pups treated with CNO (1 mg/kg) or vehicle ( Figure 1N).
We observed a significant difference in sEPSC amplitude in CA1 pyramidal neurons of PNCNO-treated mouse pups as compared to vehicle-treated controls, as revealed by a small but significant decrease in low amplitude events (< 100 pA), and a significant increase in large amplitude events characterized by the presence of a long-tail in sEPSC amplitude event cumulative distribution ( Figure 1O, Q, p < 0.0001). CA1 pyramidal neurons in hippocampal slices from PNCNO-treated mouse pups displayed large sEPSC events characterized by compound negative peaks as compared to vehicle-treated controls (Fig 1O; bottom traces). We also noted a significant decline in the cumulative probability of sEPSC interevent intervals in CA1 pyramidal neurons from the PNCNO treatment group ( Figure 1R, p < 0.0001). Further, we observed a significant reduction in the cumulative probability of sIPSC amplitude ( Figure   1P, S, p < 0.0001) and an increase in the cumulative probability of sIPSC interevent intervals ( Figure 1T, p < 0.0001) in CA1 pyramidal neurons from PNCNO-treated mouse pups.  Figure   2M, p = 0.04) and a trend towards a decrease in the time spent in the light box ( Figure 2N, p = 0.07). We then evaluated the influence of chronic CNO-mediated hM3Dq DREADD activation of forebrain excitatory neurons in the early postnatal window on despair-like behavior in adulthood using the FST ( Figure 2O). We observed increased despair-like behavior in CamKIIα-tTA::TetO-hM3Dq bigenic adult male mice with a history of PNCNO treatment, as noted by a significant increase in the time spent immobile in the FST ( Figure 2P, p = 0.02).
Taken together, these results indicate that chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window results in long-lasting increases in anxiety-and despair-like behavior in adult male mice.
Following this, we sought to ascertain whether chronic CNO-mediated hM3Dq Taken together, these results indicate that chronic CNO-mediated hM3Dq DREADD activation of forebrain excitatory neurons during the early postnatal window results in long-lasting increases in both anxiety-and despair-like behavior in adult male mice, whereas it evokes a persistent increase in anxiety-, but not despair-like behavior, in adult female mice. Henceforth, all our studies to assess the behavioral, metabolic, molecular and electrophysiological influence of chronic CNO-mediated hM3Dq DREADD activation of forebrain excitatory neurons during the early postnatal window have been restricted to male mice.  Collectively, these control experiments indicate postnatal CNO administration does not evoke off-target effects that influence anxiety-and despair-like behavior.
Chronic chemogenetic activation of CamKIIα-positive forebrain excitatory neurons during the juvenile window or in adulthood does not evoke any long-lasting changes in anxiety-and despair-like behavior Given we observed that chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons in the early postnatal window can program persistent changes in anxiety-and despair-like behavior, we next sought to ascertain whether the temporal window in which this perturbation is performed is critical to the establishment of these long-lasting behavioural changes. To address this question, we used chronic CNO-  Figure 4A). In order to assess for sensorimotor gating, we subjected CamKIIα-tTA::TetO-hM3Dq bigenic adult male mice with a history of PNCNO treatment to the pre-pulse inhibition (PPI) test ( Figure 4B). We did not observe any significant alterations in the basal startle response across treatment groups ( Figure 4C). Strikingly, we noticed a significant PPI deficit at all prepulse tones, with a decline in percent prepulse inhibition to tone (120 dB) following a pre-pulse of + 4 dB (69 dB; Figure   4D, p = 0.024), + 8 dB (73 dB; Figure 4D, p = 0.017), and + 16 dB (81 dB; Figure 4D Dysregulation of glutamatergic and GABAergic neurotransmission within forebrain neurocircuitry, including the hippocampus and several cortical regions, is thought to causally contribute to the pathogenesis of several mood-related disorders including anxiety, major depression, and schizophrenia [61][62][63][64] . In particular, metabolic dysfunction of glutamate and GABA systems are considered to be important endophenotypes of mood-related disorders [65][66][67][68] . Hence, we next sought to investigate the effects of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on the metabolic activity in glutamatergic and GABAergic neurons in the hippocampus and cortex in adulthood. We orally administered CNO (PNCNO; 1 mg/kg) or vehicle to CamKIIα-tTA::TetO-hM3Dq bigenic mouse pups once daily from P2-P14, and performed metabolic analysis in adulthood using a trace approach by infusing [1,6-13 C 2 ]glucose ( Figure 5A; Figure   5 -figure supplements 1, 2A, 3A). [1,6-13 C 2 ]Glucose is transported and metabolized in the brain to Pyruvate C3 via glycolysis. The pyruvate C3 is subsequently oxidized through the TCA cycles of glutamatergic and GABAergic neurons, and astrocytes to produce 13 C labelled metabolites ( Figure 5 -figure supplement 1). The 13 C labeling of brain metabolites was measured in 1 H-[ 13 C]-NMR spectra of brain tissue extracts. The metabolic rate of glucose oxidation in excitatory and inhibitory neurons was determined by using the three-compartment metabolic rate model [69][70][71] .
First, we measured the concentration of different metabolites in the hippocampus and cortex from the non-edited 1 H-[ 13 C]-NMR spectrum using [2-13 C]glycine as the reference ( Figure 5 -figure supplement 2B). We did not observe any significant difference in the levels of glutamate, GABA, glutamine, aspartate, N-acetylaspartate, lactate, inositol, taurine, choline and creatine in the hippocampus and cerebral cortex of PNCNO-treated bigenic adult male mice as compared to their vehicle-treated controls ( Table 2). We observed a significant decline in the levels of alanine in the hippocampus (p = 0.047), but not in the cortex, of bigenic adult male mice with a history of PNCNO treatment (Table 2).
Further, we measured the 13 C labeling of amino acids from [1,6-13 C 2 ]glucose TCA from the 13 C edited spectrum ( 13 C only) ( We next sought to distinguish the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on hippocampal excitatory and inhibitory neurotransmission in adulthood. We performed whole-cell patch clamp analysis to measure sEPSCs and mEPSCs in CA1 pyramidal neurons in acute hippocampal slices derived from bigenic adult male mice with a history of PNCNO treatment ( Figure 7A). We noted a significantly altered sEPSC amplitude

Discussion
The major finding of our study is that chronic chemogenetic hM3Dq DREADD  90 . We noted a significant impairment in sensorimotor gating indicated by prepulse inhibition (PPI) deficits, but no change in object memory or stereotypic behavior, in adult mice with a history of PNCNO treatment. Deficient PPI is considered to be a behavioral deficit associated with schizophrenia-like behavior in both genetic or environmental perturbation based animal models 87,[91][92][93][94] . Preclinical genetic models targeting signaling pathways downstream to Gq (PLC-β1 -/mice) exhibit enhanced schizophrenia-like behavior 95 . Further, loss of function of the Gq-coupled mGluR5 receptor in parvalbumin-positive interneurons increased both compulsive behavior and aberrant sensorimotor gating 96 . It is important to note that PPI deficits are also common across various other neuropsychiatric conditions, in addition to schizophrenia 97,98 . Several reports indicate that early adversity during the perinatal window results in PPI impairments in adulthood 87,[99][100][101] . However, both the intensity and timing of the early stressor could program differing outcomes on PPI 87,101,102 . For example, severe maternal deprivation evokes robust PPI deficits, whereas short duration maternal separation has no effect on PPI 87,102 . In this regard, our findings that chemogenetic activation of forebrain excitatory neurons produces an entire spectrum of mood-related behavioral changes, namely enhanced anxiety-, despair-and schizophrenia-like behaviors is suggestive of behavioral endophenotypes noted with the more severe of early stress models 98,103,104 .
Our results that the timing of the chronic chemogenetic activation of forebrain excitatory neurons is central to determining consequent changes in mood-related behavior underscores the key importance of 'critical' periods for programming emotionality 105,106 . We Associated with the long-lasting behavioral changes programmed by chronic DREADD activation of CamKIIα-positive forebrain excitatory neurons, we noted persistent dysregulation of glutamate and GABA neurotransmitter metabolism, a decline in the expression of neuronal activity-and plasticity-related markers, as well as alterations in hippocampal spontaneous excitatory and inhibitory currents. The dysregulation of both glutamate and GABA systems is amongst the key factors in the pathophysiology of several psychiatric disorders including anxiety, depression, and schizophrenia 61,62,64,114,115 . Neuroimaging studies on human subjects with mood-disorders and schizophrenia demonstrate altered volume and resting state functional activity in several forebrain regions, including hippocampus, sensory and frontal cortices [116][117][118] . A major endophenotype that reflects persistent alterations in neuronal activity in moodrelated disorders is the levels and neurometabolic activity of glutamate and GABA, the major excitatory and inhibitory neurotransmitters respectively 61,62,[64][65][66][67][68] . Although 1 H-MRS has been widely used to examine the levels of these neurotransmitters in both human patients and rodents 119,120 , there has been a scarcity of studies to investigate neurometabolic activity, which represent a functional readout of metabolic dynamics in these neurocircuits 121,122 . We

Animals
The CamKIIα-tTA transgenic mice 146  Following subsequent washes, blots were exposed to HRP conjugated goat anti-rabbit secondary antibody (1:6000, Cat. No. AS014, Abclonal Technology, United States) for one hour. Signal was visualized using a GE Amersham Imager 600 (GE life sciences, United States) with a western blotting detection kit (WesternBright ECL, Advansta, United States).
Densitometric quantitative analysis was performed using ImageJ software.

Immunofluorescence
HA-tagged hM3Dq DREADD expression in the hippocampus and cortex of CamKIIα-tTA::TetO-hM3Dq bigenic mice (P7) was visualized using immunofluorescent staining for the HA epitope. Pups single-positive for either CamKIIα-tTA or TetO-hM3Dq were used as the genotype-controls. Mice were sacrificed by transcardial perfusion with 4 % paraformaldehyde, and 40 µm thick serial coronal sections were obtained using a vibratome (Leica, Germany). test performed on day 15, and the light-dark avoidance test carried out on day 22, with the treatment paradigm being carried out from day 1 to day 13. To assay for persistent alterations in emotional behavior following chronic CNO-mediated hM3Dq DREADD activation, CamKIIα-tTA::TetO-hM3Dq bigenic male mice treated chronically with either vehicle/CNO as described above, were given a wash-out period of three months and then were assayed for anxiety-and depressive-like behavior, followed by the prepulse inhibition test for sensorimotor gating.

Reflex behaviors
To assess the influence of chronic CNO-mediated hM3Dq DREADD activation of

Light-dark avoidance test
The light-dark box consisted of a rectangular box with a light chamber (25 cm x 25 cm) and a dark chamber (15 cm x 25 cm) which were connected via a passage (10 cm x 10 cm).
The mice were released into the behavioral arena facing the light box following which behavior was recorded for 10 minutes using an overhead camera, digitized at 25 fps. The time spent and number entries in the light box were then manually assessed by an experimenter blind to the experimental treatment groups.

Despair-like behavior
In order to test the effect of postnatal, juvenile, and adult treatments on despair-like behavior, forced swim test (FST) was performed with the PNCNO (males and females), JCNO (males), and ACNO (males) treatment groups. Further, FST was also performed on vehicle/PNCNO-treated genotype-control and C57BL6/J male mice.

Forced swim test
The forced swim test was performed in a transparent cylindrical chamber (30 cm height, outer diameter: 15 cm, inner diameter: 14 cm) filled with 25°C water to a height of 30 cm from the base. The mice were released into the water and behavior was recorded for 6 minutes using a side-mounted webcam (Logitech, Switzerland). The time spent immobile was calculated for a duration of five minutes, with the first minute discarded from the analysis by an experimenter blind to the experimental treatment groups.

Sensorimotor gating
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on sensorimotor gating behavior, vehicle and PNCNO-treated CamKIIα-tTA::TetO-hM3Dq bigenic male mice were assayed on the prepulse inhibition test performed using a startle and fear conditioning apparatus (Panlab, Spain). In addition, CamKIIα-tTA::TetO-hM3Dq bigenic mice treated with vehicle or CNO during the juvenile (JCNO) or adult (ACNO) window were also assayed for PPI to observe the long-term influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during different time epochs on sensorimotor gating behavior. The apparatus comprised of a soundproof chamber with metal grid flooring, a strain gauge coupled to load cells to transduce rapid load change during startle behavior, a load cell amplifier, and a control/interface unit connected to the computer. The load cell was calibrated prior to behavioral testing using a 20 g standard weight by setting the load cell amplifier in DC mode at a gain of 1000. The mouse was placed inside a restrainer to limit spatial location with respect to sound and habituated to the apparatus for four days, followed by habituation for three days with 65 dB background white noise. On the test day, the load cell amplifier was set in AC mode at a gain of 5000. Packwin software (Panlab, Spain) was used to program the protocol and acquire the data. A digital gain of 8 was applied while acquiring the data. The mouse was first habituated to the box for 5 min with 65 dB background white noise which was followed by the first block in which ten tone pulses (120 dB, 1s) were presented to measure basal startle response. In the second block, the mouse was presented with either only tone (120 dB, 1s; x10) or a 100 ms pre-pulse which was either + 4 dB, + 8 dB or + 16 dB higher than the background noise (69/73/81 dB, 1s; x5) which coterminated with a 1s, 120 dB tone. The percent prepulse inhibition was calculated using the following formula: Percent prepulse inhibition = 100 × (average startle response with only tone − average startle response with the prepulse) ⁄ average startle response with only tone.

Novel object recognition
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on object recognition memory, vehicle and PNCNO-treated CamKIIα-tTA::TetO-hM3Dq bigenic male mice were subjected to the novel object recognition test. Mice were first habituated in the open field box (40 cm x 40 cm x 40 cm) for five minutes, twice a day for three days. On the fourth day, they were exposed to two identical objects (either sand-filled jugs or marble-filled transparent bottles) for ten minutes. Objects were interchangeably used as familiar or novel to avoid any object bias. On the fifth day, the object recognition memory test was performed. The mice were exposed to the familiar objects for five minutes which was replaced by a novel object for the next five minutes. Behavior was recorded using an overhead analog camera, digitized at 25 fps using an analog to digital converter (Startech, UK), and then tracked online using the automated behavior analysis software Ethovision XT 11. The discrimination index was calculated as the ratio of time spent exploring the novel object to the total time spent exploring both the novel and familiar object.

Marble burial
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of

Neurometabolic analysis in hippocampus and cerebral cortex
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on neurometabolism in the hippocampus and cerebral cortex of vehicle and PNCNO-treated CamKIIα-tTA::TetO-hM3Dq bigenic male mice, the 13 C labeling of brain metabolites were measured in tissue extracts using 1 H-[ 13 C]-NMR spectroscopy following infusion of [1,6-13 C 2 ]glucose.

Infusion of [1,6-13 C 2 ]Glucose
Vehicle and PNCNO-treated adult CamKIIα-tTA::TetO-hM3Dq bigenic male mice were subjected to 6 hours of fasting, and briefly restrained for the cannulation of the tail-vein to infuse 13 C labeled [1,6-13 C 2 ]glucose. [1,6-13 C 2 ]Glucose (ISOTEC, Miamisburg, United States) dissolved in deionized water (0.225 mol/L) was first administered as a bolus followed by an exponentially decreasing infusion rate for 2 min. Blood was withdrawn via retro-orbital bleeding under mild chloroform anesthesia just before the end of the experiment, centrifuged, and the plasma was collected and frozen in liquid N 2 , and subsequently stored at -80°C. Exactly

Preparation of brain extract
Metabolites were extracted from brain tissue using a modified protocol described previously 147 . In brief, the frozen hippocampal and cortical tissue samples were homogenized in 3X volume/weight of 0.1 mol/L HCl dissolved in methanol using a motorized homogenizer. volume/weight of 90% ice-cold ethanol was added and the tissue was homogenized, which was then centrifuged at 16,000g for 45 min at 4°C. The supernatant was collected and passed through a custom made Chelex column (Biorad, USA). The pH of the extract was adjusted to 7.0, which was followed by lyophilization. The lyophilized powder was dissolved in phosphate-buffered deuterium oxide containing sodium trimethylsilylpropionate (0.25 mmol/L).

NMR analysis of plasma and brain extract
Blood plasma (100 µL) was mixed with deuterium oxide (450 µl) containing sodium formate (1 µmol/L) and then passed through a 10 KDa cutoff centrifugal filter (VWR, Radnor, PA, United States). 1 H-NMR spectra were acquired using a 600 MHz spectrometer (Bruker AVANCE II, Karlsruhe, Germany). Percent 13 C enrichment of glucose-C1 was calculated by dividing the area of the 13 C-coupled satellites by the total area 1 H area ( 12 C + 13 C) observed at 5.2 ppm. 1 H-[ 13 C]-NMR spectra of hippocampal and cortical tissue extracts were recorded as described previously 122,148,149 . Briefly, two spin-echo 1 H NMR spectra were recorded with an OFF/ON 13 C inversion pulse. Free induction decays (FIDs) were zero-filled, apodized to Lorentzian line broadening, Fourier transformed, and phase-corrected. C-13 edited NMR was obtained by subtracting the sub-spectrum obtained with 13 C inversion pulse from that acquired without inversion. The concentration of metabolites was calculated by using [2-13 C]glycine as the relative standard. Percentage 13 C enrichment of the desired metabolites was determined as the ratio of the peak areas of the 1 H-[ 13 C]-NMR difference spectrum ( 13 C only) to the nonedited spectrum ( 12 C + 13 C). This was further corrected for the natural abundance (1.1%) of 13 C.

Estimation of metabolic rate of glucose oxidation
The metabolic rates of excitatory and inhibitory neurons were calculated using 13 C label trapped into brain metabolites as described previously 69,72 . The metabolic rate of glucose oxidation by glutamatergic neurons (MR Glu ) was calculated using the following formula:

Quantitative PCR
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on persistent changes in gene expression within the hippocampus, hippocampi derived from vehicle and PNCNO-treated CamKIIα-tTA::TetO-hM3Dq bigenic adult male mice were subjected to qPCR analysis. Vehicle and PNCNO-treated adult CamKIIα-tTA::TetO-hM3Dq bigenic mice were anesthetized by CO 2 inhalation and sacrificed by rapid decapitation. The hippocampi were then dissected in ice-cold PBS, snap-frozen in liquid N 2 , and stored at −80°C. RNA extraction was performed using Trizol (TRI reagent, Sigma-Aldrich, USA). The RNA was quantified using a Nanodrop (Thermo Scientific, USA) spectrophotometer followed by reverse transcription reaction to produce cDNA using PrimeScript RT Reagent Kit (Takara, Clonetech, Japan). Specific primers against the genes of interest (Table 3) were designed and qPCR was performed to amplify the genes of interest using the CF96X Real Time System (BioRad, USA).
The qPCR data were analyzed using the ΔΔCt method as described previously 150 . Ct value for a particular gene was normalized to the endogenous housekeeping gene GAPDH (Glyceraldehyde 3-phosphate dehydrogenase), which was unchanged across treatment groups.

c-Fos immunohistochemistry and cell counting analysis
To determine the influence of chronic CNO-mediated hM3Dq DREADD activation of CamKIIα-positive forebrain excitatory neurons during the early postnatal window on persistent changes in neuronal activity within the hippocampus, brain sections were subjected to c-Fos immunohistochemistry and cell counting analysis. Vehicle and PNCNO-treated CamKIIα-tTA::TetO-hM3Dq bigenic adult male mice that were naïve for behavioral testing, were sacrificed by transcardial perfusion with 4% paraformaldehyde. Coronal sections of 40 µm thickness were obtained using the vibratome (Leica, Germany). Sections were then blocked at room temperature for 2 hours in 10% horse serum with 0.3% TritonX-100 (made in 0.1M Phosphate buffer) following which they were incubated with rabbit anti-c-Fos antibody continuously with 95% O 2 and 5% CO 2 . The brain was then dissected in a petri dish containing an ice-cold cutting solution, the cerebellum was removed, and glued onto a brain holder before placing in a container filled with ice-cold cutting solution. Horizontal sections (300 µm) were obtained using a vibrating microtome (Leica, VT-1200, Germany) and transferred to a petri dish containing aCSF (124 mM NaCl, 3 mM KCl, 1 mM MgSO 4 .7H 2 O, 1.25 mM NaH 2 PO 4 , 10 mM D-glucose, 24 mM NaHCO 3 , and 2 mM CaCl 2 ) at room temperature. The part of the sections containing the hippocampus was gently dissected out and transferred to a nylon mesh chamber containing aCSF that was continuously bubbled with 95% O 2 and 5% CO 2 at 37°C.
Following a recovery period of 45-60 min at 37°C to ensure a stable electrophysiological baseline response, slices were kept at room temperature and subsequently transferred to the recording chamber as required.

Recording Rig and data acquisition
The slice recording chamber was continuously circulated with aCSF at 1-2 mL min -1 using a combination of a peristaltic pump (BT-3001F, longer precision pump Co. Ltd., China) and gravity feed. The aCSF was pre-heated to 34°C using a single channel temperature

Whole-cell patch clamp recording
Somata of CA1 pyramidal neurons were patched in order to perform whole-cell patch clamp recording. A small positive pressure was applied to the patch pipette filled with appropriate intracellular recording solution using a mouth-operated syringe which was attached to the pipette holder using air-tight tubing. After setting current and voltage offset to zero, the resistance of the electrode was noted by applying a test voltage step and measuring the resulting current according to Ohm's law. When the pipette tip touched the cell surface which could be confirmed both by the appearance of "dimple shape" under the microscope and a change in resistance, the positive pressure was released. A subsequent negative pressure was applied by a gentle suction leading the pipette to form a tight seal with the somata, indicated by a gigaohmseal (> 1 GΩ). The slow and fast capacitance were offset followed by the rupturing of the cell membrane by applying gentle suction. The cells with membrane potential more depolarized than -55 mV in adulthood and -45 mV at P7 were not considered. In addition, only cells having a series resistance in the range of 5-25 MΩ during the course of recording were considered.
Recording of spiking activity of CA1 pyramidal neurons following CNO administration at P7 was performed by holding cells in a current clamp mode. The identity of CA1 pyramidal neurons was confirmed qualitatively using the shape of action potential (characterized by the presence of an after-depolarization potential) by injecting a current of up to 2 nA for 2 ms through the patch electrode. A five-minute baseline was recorded followed by bath application of 1 µM for two minutes. The slices were then washed in aCSF and spiking activity was recorded. In order to measure intrinsic membrane properties and input-output characteristics, both at P7 and adulthood, a 500 ms, 7-step hyperpolarizing or depolarizing current (ranging from -100 pA to 180 pA) was injected through the patch electrode at an inter-sweep interval of

Statistics
All experiments had two treatment groups and were subjected to a two-tailed, unpaired Student's t-test using GraphPad Prism (Graphpad Software Inc., USA). One-sample Kolmogorov-Smirnov test was performed to confirm normality. All graphs were plotted using GraphPad Prism (Graphpad Software Inc., USA). Data are expressed as mean ± standard error of the mean (S.E.M) and statistical significance was set at p < 0.05. To account for type I errors, the qPCR data were further subjected to the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli method to calculate false discovery rate (FDR) at 5%. Vehicle and PNCNO treatment groups were subjected to linear regression followed by ANCOVA in order to compare input-output curves and statistical significance was set at p < 0.05. For the analysis of spontaneous current data, amplitudes and inter-event intervals of events recorded from vehicle or PNCNO treatment groups were converted to corresponding cumulative probability distributions and then subjected to Kolmogorov-Smirnov two-sample comparison. Statistical significance was set at p < 0.001.

Acknowledgements
We are grateful to Prof. Rishikesh Narayan (Molecular Biophysics Unit, Indian Institute of Science, Bangalore) for his valuable inputs on the manuscript. We thank all members of the Vaidya and Clement Lab for their technical help. We thank Monalisa Ghosh and Manish Biyani from the Patel Lab for their technical support during NMR experiments. We thank TIFR animal Facility personnel and Dr Sachin Atole for technical support.                      Schematic of 13    of PNCNO or vehicle administration, were subjected to fasting for 6 hours, following which [1,6-13 C 2 ]glucose was infused via the tail-vein. Blood plasma was collected and mice were sacrificed 6 minutes following glucose infusion, followed by dissection of hippocampus and