Dopamine release and dopamine-related gene expression in the amygdala are modulated by the gastrin-releasing peptide in opposite directions during stress-enhanced fear learning and extinction

The study addresses how dopamine contributes to fear extinction and PTSD-related behaviors, focusing on identifying molecular and circuit mechanisms within the amygdala circuitry. Fear extinction forms a new memory that suppresses conditioned fear when the unconditioned stimulus is omitted. Deficits in extinction contribute to PTSD. Dopamine has been implicated in signaling the omission of expected aversive outcomes and in modulating extinction through VTA and substantia nigra projections to the amygdala, but the specific molecular pathways and cell types remain unclear. The gastrin-releasing peptide (Grp) gene is enriched in excitatory neurons of the BLA and in inputs/outputs including the VTA, where it co-localizes with dopaminergic markers. GRP enhances BLA inhibition via GRPR on interneurons, whereas dopamine suppresses inhibition, suggesting an opposing relationship. The authors hypothesize that GRP regulates dopamine function in the BLA during fear learning and extinction, predicting that loss of GRP alters dopaminergic signaling and extinction outcomes under stress.

Prior work established that fear extinction relies on interactions among amygdala, hippocampus, mPFC, NAc, and VTA. Dopamine has a recognized role in extinction learning across species by signaling prediction error for omitted aversive outcomes, with VTA/SNc projections interfacing with fear circuits. Within the BLA, GRP is released from excitatory neurons and activates GRPR on GABAergic interneurons to increase inhibition, and GRPR loss enhances LTP, fear memory, and impairs extinction. Dopamine in the amygdala gates LTP by suppressing feedforward inhibition, supporting an opponent relationship with GRP. Dopaminergic subpopulations in VTA co-express Grp, and some VTA neurons co-transmit dopamine and glutamate and participate in salience signaling. Despite circuit insights, intracellular pathways by which dopamine modulates extinction and how peptides like GRP interact with dopamine signaling were unclear.

- Animals: Generation of Grp knockout mice by deleting most of exon 1 and inserting GFP; validation of expression patterns. Crosses to GRPR knockout for double KO. Adult BLA interneuron ablation using bombesin-saporin to target GRPR-expressing cells. - Neuroanatomy: Immunohistochemistry for GFP and tyrosine hydroxylase (TH); rAAV2-retro-CaMKII-tdTomato retrograde tracing from TE3, LA, BA, and mPFC to map GRPergic projections; mapping VTA-BLA dopaminergic and GRP-positive fibers. - Behavior: Standard cued and contextual fear conditioning (single or two tone-shock pairings), short-term (1 h) and long-term (24 h) memory tests. Stress-Enhanced Fear Learning (SEFL): 2 h restraint stress or handling, 7-day delay, cued FC, two days of extinction 4 days after FC, recall test 24 days later; classification into stress-susceptible (SS) and stress-resilient (SR) based on post-shock freezing. Anxiety tests (EPM, OF, LD) and shock sensitivity assessments. - Photometry: AAV-dLight1.2 expression and optic fiber implantation in BA; in vivo fiber photometry during SEFL training and extinction to record dopamine dynamics. - Optogenetics/electrophysiology: AAV-hSyn-ChR2-EGFP injection into VTA; ex vivo whole-cell recordings from BLA neurons to measure sEPSPs and sIPSPs with and without light stimulation of VTA-BLA terminals; intrinsic properties assessed. - Molecular assays: qRT-PCR on BLA tissue 30 min after SEFL recall for dopamine-related genes (Th, Nurr1, Drd1, Drd2, Grp) normalized to Gapdh/β-actin; Western blot for TH in naïve tissues; RNA-seq of BLA after recall, DESeq2 differential expression, TPM comparisons, and GO enrichment analyses. - Statistics: Two-way and three-way ANOVA (repeated measures where appropriate), post-hoc tests (Tukey, Holm-Sidak), t-tests, correlation (R^2).

- Behavior and extinction susceptibility: - Grp−/− mice show enhanced long-term fear memory (LTM) for contextual (p=0.031) and trend for cued; short-term memory unaffected. GRP/GRPR double KOs further enhanced LTM (contextual p=0.018; cued p=0.015). - Adult BLA ablation of GRPR interneurons increased cued LTM freezing and reduced GRPR cell counts (~50% reduction; Two-way ANOVA interaction F(1,58)=6.45, p=0.0138; Tukey p=0.0062). - In SEFL, genotype significantly affected extinction freezing (Genotype F1,64=18.137, p<0.001). Recall showed Genotype×Stress interaction (F1,64=4.269, p=0.042); KO-SEFL froze more than KO-FL (p=0.0196), WT-SEFL (p=0.0003), and WT-FL (p=0.001). - WT-SEFL animals segregated into SR and SS groups with distinct extinction and recall (extinction Susceptibility F1,18=11.380, p=0.003; recall p=0.008), whereas KO-SEFL showed no SR vs SS differences, indicating generalized susceptibility in KOs. - Amygdala activity and anxiety/pain controls: - Increased c-Fos and Arc mRNA in BLA 30 min after conditioning in Grp−/−; no differences in naïve state. Anxiety and shock sensitivity were normal, indicating memory-specific effects. - Dopamine dynamics (fiber photometry in BA): - Both genotypes showed increased dLight to tone and shock during training (F(2,49)=19.47, p<0.0001). - Grp−/− had larger shock-evoked dopamine during training (Group×Event F(2,48)=4.60, p<0.05; KO vs WT shock p=0.0007; another analysis: Group×Epoch F(2,20)=3.515, p=0.0492; shock p=0.0187). - During extinction, KO had larger tone-evoked dopamine on day 1 (Main effect of Group F(1,16)=6.054, p=0.026; Extinction 1 tone KO vs WT p=0.0439) but not day 2 (p=0.7228). - Tone responses during extinction correlated with prior shock-evoked responses during training (R^2=0.70, p<0.001), not with tone responses during training (R^2=0.096, p=0.33). - Circuit mapping and VTA-BLA connectivity: - GRP-positive projections mapped to indirect MGm/PIN→TE3→LA and vHP→BLA pathways; GRPergic BLA neurons did not project to mPFC. VTA dopaminergic projections primarily innervated BA; GRP and TH co-expressed in subsets of VTA neurons. - Optogenetics/electrophysiology: In WT, VTA terminal stimulation increased sEPSP frequency; in Grp−/−, baseline sEPSP frequency matched WT-on and did not further increase with light (Genotype F(1,21)=19.49, p=0.0003; Optogenetic F(1,21)=8.14, p=0.010; WT OFF vs WT ON p=0.023). sIPSP frequency was higher at baseline in Grp−/− and unaffected by light (Genotype F(1,21)=46.09, p≈2.3×10−?; multiple Tukey comparisons significant). Resting membrane potential was ~4 mV more depolarized and spike activity increased in Grp−/−, indicating presynaptic dysregulation and hyperexcitability. - Gene expression (qPCR/RNA-seq after SEFL recall): - qPCR: In BLA, Th, Nurr1, Drd1 downregulated in KO-SEFL with significant Genotype×Stress interactions (two-way ANOVA); Drd1 decreased also in KO-FL vs WT. No differences in naïve VTA/BLA. - RNA-seq: Lower TPM for Th and Drd1 in KO-SEFL vs WT-SEFL; Nurr1 lower in SEFL vs FL (significant in KO). Additional decreases in Drd2 and Grik2 in KO-SEFL vs WT-SEFL. Ppm1f also reduced in KO-SEFL. - Differential expression: KO-SEFL vs KO-FL (452 genes, q<0.05) and KO-SEFL vs WT-SEFL (336 genes, q<0.05). GO terms showed downregulation of synaptic transmission regulation and upregulation of microtubule/cilium assembly. - Overall: Loss of GRP enhances dopamine release to aversive stimuli and cues during learning/early extinction, but suppresses dopamine-pathway gene transcription after remote extinction memory recall, linking GRP to dopaminergic control of extinction and stress susceptibility.

Findings demonstrate that GRP is a key molecular regulator of dopaminergic control over fear learning and extinction within the VTA-BLA circuit. In Grp−/− mice, dopamine release is amplified during fear conditioning and early extinction, likely increasing salience and strengthening initial fear memory traces, contributing to SEFL susceptibility. Subsequently, after long-term extinction memory recall, dopamine-related gene expression (Th, Drd1, Drd2, Nurr1) is reduced in BLA, suggesting compensatory downregulation following earlier dopamine hyperrelease or stress- and learning-dependent plasticity. Optogenetic and electrophysiological data indicate presynaptic dysregulation of VTA inputs to BLA in Grp−/−, with elevated baseline excitatory and inhibitory synaptic event frequencies and occluded responses to VTA terminal activation, supporting a model where GRP maintains appropriate gain of dopaminergic signaling onto BLA networks. Circuit tracing shows GRP selectively labels the indirect auditory thalamo-cortico-amygdala pathway, highlighting its role in processing learned fear cues. Together, these results connect peptide signaling (GRP) to dopamine dynamics across phases of fear learning and extinction and propose GRP as a functional biomarker of learned fear processing and potential contributor to PTSD vulnerability.

This study identifies GRP as a molecular link that modulates dopaminergic signaling in opposite directions across fear learning and extinction: enhancing dopamine release during conditioning and early extinction while reducing dopamine-related gene transcription after remote extinction recall. Loss of GRP augments fear memory, increases SEFL susceptibility, disrupts VTA-BLA presynaptic connectivity, and downregulates dopamine pathway genes in the BLA after recall. GRP thus emerges as the first gene shown to regulate dopaminergic control of fear extinction and as a potential biomarker and therapeutic target for extinction-impaired states such as PTSD. Future work should dissect cell type– and projection-specific mechanisms of GRP action in VTA and BLA, determine developmental versus adult contributions using conditional models, and test whether enhancing GRP or dopamine-pathway gene expression can improve extinction and resilience in stress-exposed subjects.

- Global, constitutive Grp knockout may introduce developmental or systemic effects; adult GRPR interneuron ablation in BLA partially addresses but does not fully resolve this. - Other neurotransmitter systems (e.g., glutamate) likely contribute to observed phenotypes and were not comprehensively examined. - Photometry and optogenetic manipulations may affect behavior (e.g., mobility due to fibers), potentially blunting behavioral differences during extinction. - RNA-seq/qPCR analyses capture BLA bulk tissue; cell type–specific transcriptional changes were not resolved. - Mixed housing of WT and KO could influence behavior across groups. - Temporal dynamics of transcriptional changes relative to learning phases were sampled at a single post-recall timepoint; causal links require further testing.