Circuit-specific hippocampal ΔFosB underlies resilience to stress-induced social avoidance

The ventral hippocampus (vHPC) regulates emotional responses to stress and contributes to neuropsychiatric diseases such as depression. Glutamatergic vHPC neurons project to the nucleus accumbens (NAc) and basolateral amygdala (BLA), influencing reward, anxiety, and susceptibility to stress-induced social avoidance. Prior work shows that hyperexcitability of NAc-projecting vHPC neurons correlates with stress susceptibility and anhedonia, whereas activity patterns predict anxiety and social withdrawal. However, molecular mechanisms governing circuit-specific vHPC gene expression and activity during stress remain unclear. ΔFosB, a stable transcription factor induced by chronic neuronal activity, mediates stress responses in reward circuits, contributes to learning and excitability regulation in hippocampus, and is induced by both stress and antidepressants. This study tests the hypothesis that ΔFosB in defined vHPC projection neurons orchestrates gene expression and excitability changes that determine resilience or susceptibility to stress-induced social avoidance, using newly developed circuit-specific CRISPR gene editing and transcriptomic profiling approaches.

Background studies establish functional distinctions within hippocampus, with ventral components linked to anxiety and stress behaviors. Activity in vHPC-NAc projections regulates reward behavior and stress susceptibility, including sex-specific effects. ΔFosB has known roles in NAc and dorsal hippocampus in mediating stress, synaptic function, learning, neurogenesis, and neuronal excitability. Stress and antidepressants induce ΔFosB across hippocampal regions, with vHPC involvement in ketamine’s prophylactic effects. These findings motivated investigation of ΔFosB as a modulator of vHPC circuit function in stress-related behaviors and its projection-specific roles to NAc and BLA.

- Animals: Male C57BL/6J mice for most experiments; floxed FosB (FosBfl/fl) mice and Rosa26eGFP-L10a TRAP reporter line for circuit labeling; CD1 retired breeder mice as aggressors for social defeat. Standard housing conditions; IACUC-approved protocols. - Viral tools and surgeries: Stereotaxic injections targeted vHPC (vCA1/vSub), NAc, and BLA. Tools included AAV2 vectors (e.g., AAV2-CMV-GFP, AAV2-CMV-ΔJunD-GFP, AAV2-CMV-Cre-GFP) and HSV vectors (e.g., retrograde HSV-hEF1a-LSIL-myc-Cas9, HSV-hEF1a-LSIL-ΔFosB, HSV-hEF1a-Cre, HSV-IE4/5-TB-gRNA-eYFP-CMV-IRES-Cre). A novel dual-vector intersectional CRISPR-Cas9 system combined retrograde Cre-dependent Cas9 in projection targets (NAc or BLA) with local vHPC delivery of FosB gRNA and Cre to achieve circuit-specific FosB knockout (KO). For rescue, retrograde Cre-dependent ΔFosB co-expressed with Cas9 in targets while local gRNA+Cre in vHPC enabled simultaneous KO and ΔFosB re-expression in the same projection-defined neurons. Independent ΔFosB overexpression used retrograde Cre-inducible ΔFosB in targets and AAV-Cre in vHPC. - Stress paradigms: Chronic social defeat stress (CSDS) for 10 days with daily physical then sensory contact with novel CD1 aggressors. Subthreshold microdefeat (three 3–5 min encounters in one day) to probe stress sensitization. - Behavioral assays: Social interaction (SI) test with target absent/present to derive SI ratio and interaction time; elevated plus maze (EPM) for anxiety-like behavior; temporally dissociative passive avoidance (TDPA) and single-trial contextual avoidance for aversive learning; open field (OF) for locomotion and anxiety; von Frey for tactile allodynia; novelty-suppressed feeding; tail suspension. - Immunohistochemistry: Detection of FosB/ΔFosB, GFP, and α2A adrenergic receptor (α2AAR) in vHPC; confocal imaging and quantitative intensity analysis in blinded fashion. - Electrophysiology: Whole-cell patch-clamp recordings in vCA1 from GFP-labeled cells after ΔFosB overexpression or from identified vHPC-NAc projection neurons in FosB WT vs KO contexts; measured spike number across current steps, rheobase, I–V properties, AP waveform, and sEPSCs. - TRAP-Seq: Circuit-specific translating ribosome affinity purification in vHPC-NAc neurons of WT and FosBfl/fl mice after retrograde Cre; RNA isolation, library prep (SMARTer Stranded Total RNA-Seq), sequencing (TopHat to mm9), differential expression (HTSeq-count, DESeq2). Validation by qPCR in Neuro2a cells with ΔFosB overexpression or ΔJunD/Cas9+gRNA manipulations for candidate genes (e.g., Adra2a, Nefm, Prkcb, others). - CRISPR design/validation: gRNAs targeting FosB exon 2 designed via e-CRISP; screened in Neuro2a via Western blot and T7 endonuclease assay; in vivo validation by HSV single-vector CRISPR in dHPC and dual-vector intersectional system in vHPC projections. - Statistics: Alpha 0.05; appropriate t-tests, mixed two-way ANOVAs with Holm–Sidak corrections for within-factor designs; detailed parameters and sample sizes provided per assay; blinding for imaging analyses; replication across cohorts.

- Stress and antidepressant induction of ΔFosB in vHPC: CSDS increased ΔFosB+ neurons in vHPC dentate gyrus; fluoxetine induced ΔFosB across all vHPC subregions (Fig. S1). In NAc-projecting vHPC neurons labeled via TRAP after CSDS, FosB/ΔFosB transcripts showed >4-fold increase (Fig. S1f). Quantified immunostaining in vHPC-NAc neurons confirmed increased ΔFosB after stress (P=0.0048; Fig. 1a,b). - Necessity of ΔFosB activity in vHPC for stress resilience: Transcriptional inhibition of ΔFosB (ΔJunD) in vHPC, but not dHPC, converted subthreshold microdefeat into social avoidance: reduced SI ratio and interaction time (P=0.0125 for SI ratio; interaction time P<0.0001; Fig. 1e). dHPC ΔJunD had no effect (Fig. 1f). Baseline anxiety/locomotion unaffected (Fig. S3). - Circuit-specific FosB/ΔFosB roles via dual-vector CRISPR: vHPC-NAc FosB KO enhanced CSDS-induced social avoidance (P=0.0092; Fig. 2b) with small locomotion decrease (Fig. S6a), but no effect on passive avoidance or baseline anxiety (Fig. 2c,d). In contrast, vHPC-BLA FosB KO did not affect social avoidance (Fig. 2f) but impaired avoidance learning (trend P=0.0614; Fig. 2g) and decreased anxiety-like behavior (more open arm time and entries: P=0.0079 and P=0.0022; Fig. 2h), without locomotor change (Fig. S6b). - Causality via ΔFosB rescue and sufficiency via overexpression: In vHPC-NAc, ΔFosB rescue in FosB KO neurons reversed heightened social avoidance (SI ratio P=0.0161; interaction time P=0.0431; Fig. 3a,b). In vHPC-BLA, ΔFosB rescue reversed anxiolysis and avoidance learning deficits (EPM time P=0.0042; TDPA P=0.0459, 0.0060, <0.0001 across days; Fig. 3c–e). Independent ΔFosB overexpression in vHPC-NAc produced a resilient phenotype after stress (increased SI ratio and interaction; P<0.05; Fig. 3f,g), while overexpression in vHPC-BLA did not alter anxiety or avoidance learning (Fig. 3h–j). - Electrophysiological mechanisms: ΔFosB overexpression in vCA1 neurons reduced excitability: fewer spikes across 150–300 pA steps (multiple P<0.05 across steps; total spikes P=0.0108), trend to higher rheobase (P=0.0680), altered AP properties and decreased sEPSC frequency (Fig. 4a–c; Fig. S10). Conversely, FosB KO specifically in vHPC-NAc projection neurons increased excitability (more spikes total P=0.0418; Fig. 4g–h) with complementary AP and synaptic changes (Fig. S12); rheobase unchanged (Fig. 4i). These opposite phenotypes support ΔFosB-driven hypoexcitability in vHPC-NAc as a resilience mechanism. - Circuit-specific TRAP-Seq identifies ΔFosB targets: TRAP captured neuron-enriched transcripts from vHPC-NAc neurons; FosB transcripts were nearly absent in FosBfl/fl TRAP (Fig. 5b). Hundreds of differentially enriched mRNAs were identified (Tables S2, S3). In Neuro2a validation, ΔFosB regulated several candidates, including repression of Adra2a (P=0.0240) and trends for Nefm and Prkcb. ΔJunD or CRISPR of FosB increased Adra2a (P=0.0133; Fig. 5c,d). In vivo, α2AAR immunoreactivity was increased in vHPC-NAc neurons after FosB KO (P<0.0001; Fig. 5e,f), supporting ΔFosB-mediated repression of Adra2a. - Model: In resilient mice, stress robustly induces ΔFosB in vHPC-NAc neurons, triggering gene programs that decrease projection excitability and protect against social avoidance; insufficient ΔFosB in susceptible mice permits high vHPC-NAc excitability and social withdrawal (Fig. 5g).

The study demonstrates that ΔFosB induction in vHPC projection-defined neurons constitutes a molecular brake on circuit excitability that confers resilience to stress-induced social avoidance. Using intersectional CRISPR tools, the authors reveal that ΔFosB’s behavioral effects are circuit-specific: in vHPC-NAc neurons it is necessary and sufficient for resilience to social defeat, whereas in vHPC-BLA neurons it is necessary for fear and anxiety expression. Electrophysiological data show that ΔFosB reduces intrinsic excitability and synaptic drive in vHPC neurons, aligning with previous evidence that reduced vHPC-NAc activity supports resilience, while enhanced activity promotes susceptibility. Circuit-specific TRAP-Seq identifies candidate downstream effectors, including ΔFosB-dependent repression of Adra2a (α2AAR), which could influence excitability and stress responses. The findings integrate molecular, cellular, and circuit levels to explain how stress alters hippocampal output pathways to shape distinct behavioral outcomes, and they suggest that targeting ΔFosB-regulated genes in defined circuits could provide therapeutic avenues for symptoms such as social withdrawal and anxiety in depression.

This work establishes that ΔFosB in ventral hippocampal projection neurons drives circuit-specific behavioral adaptations to stress. In the vHPC-NAc pathway, ΔFosB is both necessary and sufficient for resilience to social defeat-induced social avoidance by reducing neuronal excitability; in the vHPC-BLA pathway, ΔFosB is necessary for anxiety and aversive learning. Novel intersectional CRISPR approaches, complemented by rescue and overexpression designs, validate causality and minimize off-target concerns. Circuit-specific TRAP-Seq highlights downstream targets, including Adra2a, offering potential druggable pathways. Future research should systematically validate additional ΔFosB targets, dissect their mechanistic contributions to excitability and synaptic plasticity within defined circuits, and explore translational strategies to modulate these pathways for treating mood disorder symptoms.

- The floxed FosB/GFP-L10a approach used for electrophysiology is not circuit-specific, limiting its use for certain behavioral assays (e.g., social avoidance) and potentially including off-circuit neurons. - While rescue experiments mitigate concerns about CRISPR off-target effects, comprehensive off-target assessment was not exhaustively reported. - Most downstream targets from TRAP-Seq remain to be validated functionally; evidence for Adra2a is initial and requires further mechanistic exploration in vivo. - Translation to humans is uncertain; differences in circuit coherence cannot currently be mechanistically interrogated in human studies, underscoring preclinical constraints.