Background and Rationale
The delicate balance between neuronal excitation and inhibition (E/I balance) represents a fundamental organizing principle of cortical microcircuits, and its disruption has emerged as a critical pathophysiological feature across neurodegenerative disorders, including Alzheimer's disease (AD). Parvalbumin-expressing (PV) interneurons constitute approximately 30-40% of all cortical GABAergic neurons and serve as the primary mediators of fast-spiking, feedforward and feedback inhibition onto pyramidal neurons. These cells are uniquely positioned to orchestrate gamma-band oscillations (30-80 Hz), which are essential for attention, memory encoding, and cortical information processing. The integrity of PV interneuron function depends upon precise molecular cues during development and continued signaling throughout adulthood for synaptic maintenance.
Neuregulin-1 (NRG1) and its cognate receptor tyrosine kinase ErbB4 form a crucial signaling axis governing multiple aspects of cortical circuit formation and function. NRG1 exists in multiple isoforms, with type III NRG1 being the predominant transmembrane form in the brain. Type III NRG1 undergoes proteolytic processing by beta-site amyloid precursor protein cleaving enzyme 1 (BACE1), which releases the extracellular domain to activate ErbB4 receptors on neighboring cells. This cleavage event is not merely a regulatory mechanism but a prerequisite for NRG1-ErbB4 signaling in specific neuronal populations. Notably, BACE1 exhibits particularly high expression in PV interneurons during critical periods of cortical development, suggesting a cell-type-specific role for BACE1-mediated NRG1 processing in these inhibitory neurons. The discovery that therapeutic BACE1 inhibitors, developed to reduce amyloid-beta production in AD, can impair NRG1-ErbB4 signaling has revealed an unexpected mechanism by which these agents may exert detrimental effects on cortical inhibition, potentially accelerating cognitive decline despite reducing amyloid burden.
Proposed Mechanism
The BACE1/NRG1 axis dysfunction hypothesis posits that BACE1-mediated cleavage of type III NRG1 is essential for proper PV interneuron maturation and the maintenance of their inhibitory output throughout life. Under physiological conditions, membrane-tethered type III NRG1 on pyramidal neuron axons undergoes constitutive BACE1-dependent shedding, generating a soluble NRG1 fragment that activates ErbB4 receptors concentrated on PV interneuron processes. This NRG1-ErbB4 signaling cascade triggers downstream phosphoinositide 3-kinase (PI3K)-AKT and mitogen-activated protein kinase (MAPK) pathways, promoting PV interneuron survival, dendritic arborization, and the formation of perisomatic inhibitory synapses onto pyramidal neurons.
BACE1 inhibition disrupts this signaling loop by preventing NRG1 ectodomain shedding, thereby reducing ErbB4 activation and its associated trophic and synaptic signaling. The resulting hypofunction of PV interneurons manifests as decreased GABA release from their characteristic basket synapses onto pyramidal neuron soma and initial axon segments. This PV interneuron impairment preferentially affects fast-spiking properties and the precise timing of inhibition required for gamma oscillation generation. The chandelier cells, another PV-positive subtype targeting pyramidal neuron axon initial segments, are similarly affected.
The disinhibition of pyramidal neurons creates a state of relative excitation that propagates through cortical microcircuits. Excessive glutamatergic drive onto PV interneurons normally provides a feedback mechanism to recruit inhibition, but when PV cells themselves are dysfunctional, this brake on excitation is lost. The resulting E/I imbalance disrupts gamma oscillations, which are generated through reciprocal connections between PV interneurons and pyramidal neurons organized in feedback inhibition circuits. Reduced gamma power and coherence compromise the temporal coordination of neural ensembles, fundamentally impairing the synchronization necessary for proper sensory processing, attention, and memory formation.
Supporting Evidence
Seminal work by Deja et al. and subsequent studies demonstrated that BACE1-null mice exhibit deficits in PV interneuron development, including reduced PV expression, abnormal dendritic morphology, and impaired inhibitory synapse formation. These mice show decreased GABA release probability and altered short-term plasticity at PV-to-pyramidal synapses. Electrophysiological recordings reveal disrupted gamma oscillations in the prefrontal cortex and hippocampus of BACE1 knockout animals, paralleling observations in human studies of BACE1 inhibition.
The critical role of NRG1-ErbB4 signaling in PV interneurons is further supported by conditional knockout studies. Mice lacking ErbB4 specifically in PV cells recapitulate key features of the BACE1 knockout phenotype, including reduced PV expression, abnormal inhibitory synapse function, and gamma oscillation deficits. Conversely, overexpression of the NRG1 intracellular domain or constitutive ErbB4 activation partially rescues BACE1-dependent phenotypes, demonstrating the specificity of this axis. Human genetic studies have identified NRG1 and ERBB4 polymorphisms associated with schizophrenia and other neuropsychiatric disorders characterized by E/I imbalance and gamma dysfunction, further supporting the translational relevance of this signaling pathway.
Experimental Approach
Testing this hypothesis requires a multi-modal approach combining genetic, pharmacological, and electrophysiological methods in relevant model systems. Primary mouse models would include PV-Cre;Rosa26-tdTomato mice for cell-type-specific manipulation and visualization. BACE1 flox/flox mice crossed with PV-Cre animals would enable PV neuron-specific BACE1 deletion, distinguishing direct effects on interneurons from secondary consequences of amyloid reduction. Rescue experiments using virally delivered constitutively active ErbB4 or cleavage-resistant NRG1 constructs would establish causality.
Acute brain slice electrophysiology using optogenetic tools would assess PV-to-pyramidal synapse function. Channelrhodopsin-assisted circuit mapping specifically targets PV cell inputs while recording from pyramidal neurons, measuring IPSC amplitude, kinetics, and short-term plasticity. In vivo recordings using silicon probes in freely behaving animals would quantify gamma oscillation power, coherence, and phase-amplitude coupling during behavior. Two-photon imaging of genetically encoded calcium indicators (GCaMP6) in PV cells enables longitudinal monitoring of interneuron activity during learning paradigms.
Human iPSC-derived neurons from healthy donors and AD patients, differentiated into cortical organoids containing PV interneurons, provide a human cellular platform. CRISPR-based editing of BACE1 or NRG1 genes in these systems would test the mechanism directly in human neurons. Biochemical assays including ErbB4 phosphorylation, PI3K-AKT signaling readouts, and NRG1 cleavage products in tissue and cell lysates would quantify pathway activity.
Clinical Implications
Understanding the BACE1/NRG1/ErbB4 axis in PV interneurons carries substantial therapeutic implications for neurodegenerative disease treatment. The failed clinical trials of BACE1 inhibitors for AD (verubecestat, atabecestat, lanabecestat) may partly result from on-target effects on cortical inhibition. These findings suggest that future BACE1-targeted approaches must achieve sufficient selectivity for APP over NRG1 processing, or be combined with strategies to preserve NRG1-ErbB4 signaling. Allosteric BACE1 modulators that spare NRG1 cleavage, or intermittent dosing regimens that allow NRG1 processing recovery, represent potential refinements.
Conversely, augmenting NRG1-ErbB4 signaling in PV interneurons may represent a therapeutic strategy for restoring E/I balance in neurodegeneration. ErbB4 agonists, NRG1 mimetics, or positive allosteric modulators could potentially enhance PV interneuron function and gamma oscillations. Such approaches might be particularly valuable in early AD, where PV interneuron dysfunction may precede overt neuronal loss. Combination therapies targeting amyloid pathology while preserving inhibitory circuit function could optimize therapeutic outcomes.
Challenges and Limitations
Several challenges complicate investigation and therapeutic translation of this hypothesis. First, BACE1 exhibits broad substrate specificity, with over 100 identified substrates beyond APP and NRG1, making it difficult to isolate NRG1-dependent effects from other BACE1 functions. Conditional genetic approaches and substrate-specific rescue constructs help address this issue but add complexity. Second, the temporal dynamics of BACE1/NRG1 axis dysfunction remain uncertain—whether acute BACE1 inhibition in adult animals recapitulates developmental deficits or produces distinct phenotypes requires systematic investigation.
The translatability of findings from rodent models to human cortical circuits remains a persistent concern. Human cortical organization differs from rodents, with expanded prefrontal regions and potentially distinct PV interneuron subtypes. The expression patterns and functional roles of BACE1 and NRG1/ErbB4 in human PV interneurons require direct study using postmortem tissue and iPSC-derived systems. Furthermore, neurodegeneration itself may alter BACE1/NRG1 axis function through multiple mechanisms, including altered expression, cellular mislocalization, or proteostatic stress. Determining whether BACE1/NRG1 dysfunction is a cause or consequence of neurodegenerative processes will require longitudinal studies in appropriate animal models and human cohorts.
Competing hypotheses regarding BACE1 inhibitor failure include off-target effects, excessive amyloid reduction, and pharmacokinetic factors. Distinguishing among these possibilities requires careful biomarker analysis in clinical trial samples. The development of NRG1 cleavage-specific biomarkers and PV neuron activity markers could enable patient stratification and treatment monitoring in future trials. Ultimately, addressing these challenges will require integrated approaches combining basic neuroscience, translational research, and clinical investigation to fully elucidate the role of BACE1/NRG1 axis dysfunction in neurodegeneration and develop safe, effective therapeutics.