bioRxiv Subject Collection: Neuroscience's Journal

In Vivo AGO-APP For Cell-Type- and Compartment-Specific miRNA Profiling in the Mouse Brain
AGO-APP through the expression of the T6B peptide permits the isolation of Ago-bound miRNAs. Here we present the generation and characterization of two transgenic mouse lines allowing to perform AGO-APP in vivo. First, we generated mice for CRE-dependent expression of T6B in the cytoplasm. Using this line we performed Ago-APP in olfactory bulb (OB) inhibitory interneurons and cerebral cortex excitatory neurons. Bioinformatic analysis validated the high reproducibility of the approach. It also demonstrated that, despite global miRNome conservation between the two cell types, a set of miRNAs including the miR-200 family and the miR-183/96/182 cluster, is massively enriched in OB interneurons which is consistent with previous observations. In the second mouse line T6B is fused to the postsynaptic protein PSD95. Isolation of T6B-PSD95 fractions from OB and cortical neurons identified specific sets of post-synapse enriched miRNAs. Gene ontology analyses confirmed that these miRNAs preferentially target mRNAs related to synaptic functions.

Expectations and Motor Signals Shape Auditory Evoked Responses in an Independent Manner
Voluntary actions are usually accompanied by sensory consequences that evoke neural responses in relevant sensory regions. These evoked responses are different when compared to those evoked by otherwise identical physical stimulation not associated with preceding actions (sensory attenuation). A common model suggests that sensory attenuations are caused by efference signals associated with the binding of actions with sensory outcomes. Nevertheless, the information encoded in such signals is debated, with some theories suggesting they convey predictive information while others suggest they convey global motor information regardless of specific outcome expectations. To address this debate, we recorded EEG data while participants (n=30) learned to associate motor or visual cues with corresponding tones. Following 8 repetitions of cue-tone learning, participants successfully learnt the association (>84% correct). At the neural level, the amplitude of auditory evoked responses (N100) decreased across repetitions, as the binding between cues and tones strengthened. In addition, N100 amplitude was attenuated when preceded by motor vs. visual cues. Most importantly, we did not find an interaction effect between repetition and cue type, suggesting a similar sensory attenuation effect across repetitions. Specifically, we show significant sensory attenuation even at the first repetition, when no expectation could be formed. Thus, our results suggest that the sensory attenuation effect does not change while forming experience with action-outcome contingency. These results have implications for contemporary models of sensorimotor control and predictive mechanisms.

Circuit inhibition promotes the dynamic reorganization of prefrontal task encoding to support cognitive flexibility
The mammalian prefrontal cortex encodes variables related to goal-directed behavior, and enables flexibility during environmental changes, making it critical to understand how the dynamic updating vs. stable maintenance of different encodings contribute to behavioral adaptation. We addressed this by comparing prefrontal encoding during successful adaptation vs. maladaptive perseveration. Specifically, we studied mutant (Dlx5/6+/-) mice, which have dysfunctional parvalbumin-expressing inhibitory interneurons and perseverate in a rule shifting task. We measured mPFC activity patterns using microendoscopic calcium imaging, then used linear classifiers and neural networks to compare representational geometries in wild-type mice and Dlx5/6+/- mutants before, during, and after benzodiazepine treatment, which persistently rescues their rule shift learning. The encoding of correct vs. incorrect trial outcomes rapidly shifts as mice successfully learn new cue-reward associations, but becomes more stable when mutant mice perseverate. We also find activity patterns that normally distinguish learning of the initial association vs. rule shift, but become diminished during perseveration. Finally, during perseveration, outdated representations are inappropriately reinstated, not just passively maintained. These results reveal prefrontal contributions to flexible behavior driven by the dynamic reorganization of abstract rule representations, rather than stable reinforcement signals.

Mesocorticolimbic reinforcement learning of reward representation and value provides an integrated mechanistic account for schizophrenia
Mesocorticolimbic dopamine projections are crucial for value learning, motivational control, and cognitive functions, but their precise neurocomputational roles remain elusive. Based on recent experimental and theoretical findings, we constructed a neural circuit model where dopamine neuronal populations receive differential inputs from individual rewards and encode heterogeneous reward prediction errors, which train cortical and striatal neurons to learn reward-associated state representation and value. Learning is achieved via simultaneous 'alignments' of the cortical and striatal downstream connections to the mesocorticolimbic dopamine projections, and inhibition-dominance in the cortical recurrent network is a key for successful learning. Excessive excitation, whether pre-existing or induced by manipulations, leads to aberrant activity, which disrupts the alignments and even causes anti-alignment. This impairs both reward-specific motivational control and credit assignment, potentially explaining the negative and positive symptoms of schizophrenia, respectively. Our model thus provides a mechanistic account for schizophrenia, integrating the different causes and symptoms, with testable predictions.

A switch in kappa opioid receptor signaling from inhibitory to excitatory induced by stress in a subset of cortically-projecting dopamine neurons
The kappa opioid receptor (KOR) has shown potential as a therapeutic target for several neuropsychiatric disorders including major depressive disorder, pain, and substance use disorder. In vivo signaling of G protein coupled receptors like the KOR is generally thought to change in magnitude but not sign in such behavior states. Here we investigated KOR modulation of ventral tegmental area (VTA) neurons following an acute, behaviorally aversive manipulation. We found this switches KOR signaling from inhibitory to excitatory in a subset of VTA dopamine neurons. Brief corticotrophin releasing factor (CRF) exposure ex vivo rapidly induces a similar switch in KOR signaling, specifically in dorsal medial prefrontal cortex (dmPFC) projecting neurons, but not nucleus accumbens or basolateral amygdala projecting neurons. These KOR mediated excitations depend on G protein activation, but where somatodendritic VTA KORs activate a K+ conductance to hyperpolarize dopamine neurons in control conditions, depolarizations require HCN channel function. One behavioral impact of this change is a loss of the aversiveness of intra-VTA KOR activation, providing direct evidence that rapid changes in GPCR signaling pathways can be triggered by activity at other GPCRs and significantly alter behavioral responses driven by neuromodulators.

Attention-related sampling of targets rhythmically alternates with increased susceptibility to co-occurring distractors
The Rhythmic Theory of Attention proposes that visual spatial attention is characterized by alternating states that promote either sampling at the present focus of attention or a higher likelihood of shifting attentional resources to another location. While theta-rhythmically (4-8 Hz) occurring windows of opportunity for shifting attentional resources might provide cognitive flexibility, these windows might also make us more susceptible to distractors. Here, we used EEG in humans to test how frequency-specific neural activity phasically influences behavioral performance and visual processing when high-contrast distractors co-occur with low-contrast targets. For trials with and without distractors, perceptual sensitivity at the cued target location depended on pre-stimulus theta phase (~7 Hz) recorded at central electrodes. For trials with distractors, there was a greater increase in false alarm rates at the same theta phase associated with lower hit rates (i.e., during the proposed 'shifting state'), confirming theta-rhythmically occurring windows of increased susceptibility to distractors. In addition to these phase-behavior effects at central electrodes, we observed phase-behavior effects at frontocentral and occipital electrodes that (i) only occurred on trials with distractors, (ii) peaked in the alpha-frequency range (~9-10 Hz) and (iii) were strongest at occipital electrodes that were contralateral to distractors. Alpha phase at these electrodes was also associated with fluctuations in the amplitude of distractor-evoked visual responses, consistent with an alpha-mediated gating of distractors. The present findings thus provide evidence for distinct theta- and alpha-mediated mechanisms of spatial attention that phasically modulate the influence of distractors on task performance.

Tau drives cell specific functional isolation of the hippocampal formation
A major challenge in understanding Alzheimer's disease is linking changes that occur across different biological scales. For example, how do changes in individual neurons build into widespread network disruptions? To address this, we used flexible mesh electronics to record neuronal activity for six months in ThyTau22 mice, a model of tauopathy that accumulates mutant human tau with age. Electrophysiology was recorded simultaneously from the hippocampus and entorhinal cortex of awake, behaving mice. At all ages we observed neuron-level, tau-driven silencing including ages without detectable tangles or cell-death. We found an unexpected phenomenon: neurons silenced by tau spontaneously recover individual firing patterns, yet these neurons fail to regain normal network interactions. Thus, as the animals age, disrupted network-level activity emerges. Specifically, we observe a global decrease in excitatory interactions and a breakdown in gamma-band coherence, which is particularly disrupted between the entorhinal cortex and hippocampus. These observations reveal a temporal relationship between neuronal silencing and impaired network connectivity, which also contributes to a progressive disruption in the excitatory/inhibitory balance. This ultimately disconnects viable entorhinal-hippocampal connections, physiologically isolating the hippocampus. Importantly, this network dysfunction is not driven by neuron loss, but by the failure of neurons to re-establish proper network interactions after silencing. This reveals a previously unrecognized mechanism by which mutant tau can destabilize neural systems. Further, these experiments indicate that a therapeutic window may exist where neuronal function and network activity might still be restored prior to irreversible degeneration.

Surgical Removal of Visceral Adipose Tissue has Therapeutic Benefit in Male APPNL-F Mice
Purpose: Visceral white adipose tissue (vWAT) accumulation causes systemic inflammation, insulin resistance, metabolic syndrome, and senescent cell accumulation that are risk factors for Alzheimer's disease (AD). Visceral fat removal (VFR) improves metabolism and reduces proinflammatory cytokines. We hypothesized that VFR removal in AD mice would improve metabolism and cognition. Methods: Male and female APPNL-F mice underwent sham or vWAT surgical resection (epididymal and perirenal) at 4 (pre-symptomatic) and 16 (symptomatic) months of age to understand interventional and therapeutic effects, respectively. At 18 months of age, glucose metabolism and novel object recognition (NOR) memory were assayed followed by assessment of body composition and tissue-specific markers of metabolism, cell senescence, inflammation, or amyloid accumulation. Results: Male and female APPNL-F mice showed distinct VFR responses. In pre-symptomatic males, increased vWAT lipolysis and hepatic lipogenesis led to ectopic liver lipid accumulation, with reduced adiponectin and leptin, elevated visfatin, and impaired glucose metabolism. Symptomatic males showed reduced vWAT lipogenesis, enhanced hepatic lipolysis, glycolysis, and glycogenesis, lowering liver lipids and improving insulin sensitivity. Only symptomatic males improved NOR, linked to elevated hippocampal learning and memory markers. Female vWAT reaccumulation was due to increased lipogenesis and lower lipolysis. Pre-symptomatic females had lower hepatic lipogenesis, while glycolysis and glycogenesis declined with disease progression. Hippocampal senescence and inflammation were elevated early in the disease that persisted symptomatically. Conclusions: Sex-specific differences in glucose and lipid metabolism and lipid accumulation underlie the divergent responses to VFR in APPNL-F mice, with symptomatic males showing the only beneficial outcomes in metabolism and cognition.

Modulation of repopulating microglia in multiple sclerosis models with implications for neuroprotection
Microglia play a critical role in central nervous system (CNS) pathologies including multiple sclerosis (MS), and their modulation offers therapeutic potential especially during progressive disease courses. Using cell culture and experimental autoimmune encephalomyelitis (EAE) models, we investigated microglial dynamics during depletion and repopulation (MGrepo) and their modulation using siponimod (sipo), an established CNS penetrating MS medication. Repopulating microglia exhibited a transient reactive state (CD86, MHC-II, Il1b, Tnf). Sipo modulated microglia populations, increasing CD163+, CD206+, and CX3CR1+ while reducing CD86+MHC-II+ cells accompanied by a reduction of neuronal damage. Proteomic spinal cord analysis revealed protein expression alterations by MGrepo and sipo linked to inflammation, myelination, and neuronal structural organization, supported by RNA sequencing of the spinal cord. The neuroinflammation attenuating role of sipo could be linked to cell maintenance and myelin formation associated processes. These findings highlight the capacity of pharmacological interventions to modulate microglia, offering new insights into therapeutic strategies targeting microglial activity in neuroinflammatory diseases.

Increased neuronal activity restores circadian functionin Drosophila models of C9orf72-ALS/FTD
Circadian rhythm disruptions are common across neurodegenerative diseases, but their link to amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) remains unclear. The C9orf72 hexanucleotide repeat expansion is the most prevalent genetic cause of ALS/FTD. Here, we used Drosophila models expressing toxic arginine-rich dipeptides (PR or GR) or GGGGCC hexanucleotide repeats to investigate circadian deficits in C9orf72-ALS/FTD. We found that circadian rhythmicity and period length were disrupted in a repeat number-, dosage-, and age-dependent manner. Additionally, we observed lower levels of the neuropeptide PDF, a key regulator of free-running circadian rhythms, as well as decreased projection complexity and reduced neuronal activity in PDF-expressing neurons. Importantly, increases in neuronal activity significantly restored circadian function under select conditions. These results implicate reduced neuronal activity in C9orf72-ALS/FTD circadian deficits, underscoring the importance of precisely tuned, circuit- and stage-specific interventions.

Very low amplitude muscle activity increases probability of motor evoked potentials in healthy individuals and in ALS
Background In many clinical and research settings, transcranial magnetic stimulation (TMS) intensities are standardised based on resting motor threshold (RMT). It is well-established that contraction of the target muscle increases motor evoked potential (MEP) amplitude and correspondingly decreases RMT. As such, when estimating RMT it is crucial to ensure the target muscle is relaxed. Typically trials in which baseline electromyographic (EMG) amplitude exceeds a specified threshold are rejected. The influence of motor activity below typical rejection thresholds on MEP amplitudes has yet to be established. Methods We retrospectively analysed TMS-EMG data collected during RMT measurement in 45 healthy controls (1761 datapoints) and 35 people with amyotrophic lateral sclerosis (ALS, 1238 datapoints). Trials where root mean squared (RMS) baseline EMG amplitude exceeded 10V were rejected. Generalised linear mixed-effects models were used to assess effects of muscle activity below this rejection threshold on probability of evoking an MEP with peak-to-peak amplitude [≥]50V. Results Greater sub-rejection-threshold activity significantly increases MEP probability in controls (p<0.0004) and people with ALS (p=0.0010). Models predicted a 31-38% increase in MEP probability when baseline RMS-EMG amplitude increased from 1V to 9V. Sub-rejection-threshold baseline activity was significantly greater in ALS than controls (p=0.0055). Conclusions We have shown for the first time that higher EMG amplitudes below a typical rejection threshold markedly increase probability of evoking an MEP with peak-to-peak amplitude [≥]50V. Researchers should take measures to account for effects of sub-rejection-threshold activity on RMT, particularly in populations where baseline activity may be elevated, such as in ALS.

Inflammation increases the penetrance of behavioral impairment in Shank3 haploinsufficiency mice -- can it explain the behavioral regression in Autism?
Behavioral regression occurs in ~40% of SHANK3-associated autism spectrum disorder (ASD). We previously reported that significant behavioral regression in a small cohort with SHANK3 haploinsufficiency patients, triggered by subclinical infections, responded to immunomodulator treatments. We hypothesize that behavioral regression results from the interplay between SHANK3 deficiency and neuroinflammation. Using Shank3 exon 4-22 deletion heterozygous mutant (Sh3+/-) mouse, which shows no significant behavior impairments, we established a preclinical model -- Shank3 haploinsufficiency mouse undergoing systemic inflammation challenge via intraperitoneal injection of lipopolysaccharides (LPS). We found that, two weeks after LPS challenge, wild-type mice (WT) recovered but Sh3+/- mice exhibited motor impairment, anxiety-like behaviors, and excessive grooming, similar to Shank3 exon 4-22 deletion homozygous mutants. Anti-inflammatory treatment partially reversed LPS-induced behavioral changes. Transcriptomic analysis revealed upregulation of neuroinflammation-related genes and downregulation of synaptic function-related genes in Sh3+/- mice in response to LPS. Especially, pro-inflammatory genes and microglia markers were overly activated that may result from the increased Toll-Like Receptor 4 (TLR4) expression in microglia in Sh3+/- mice. Our findings indicate that neuroinflammation increases the penetrance of behavioral impairment in Shank3 haploinsufficiency mice and support a potential mechanism for the behavioral regression in human SHANK3 disorders for future investigations.

A Cell-Autonomous Role for the Vitamin B6 Metabolism Gene PNPO in Drosophila GABAergic Neurons
In animals, the enzyme pyridox(am)ine 5'-phosphate oxidase (PNPO) is critical for synthesizing the active form of vitamin B6 (VB6), pyridoxal 5'-phosphate (PLP), from inactive vitamers. PLP is a required cofactor for many enzymatic reactions, including the synthesis of GABA and the monoamines. PNPO disruption in humans is associated with an array of epilepsy syndromes, while Drosophila harboring mutations in the sole PNPO ortholog, sugarlethal (sgll), display spontaneous seizures and short lifespans. These phenotypes are suppressed by PLP supplementation and are exacerbated by restriction of dietary B6 vitamers. In the context of PNPO deficiency, it remains to be resolved what the specific contributions by cellular subpopulations in the nervous system are to neurological phenotypes. We addressed this question in sgll mutants by expressing human PNPO (hPNPO) cDNA in cholinergic, glutamatergic, and GABAergic neurons as well as glia and measuring changes in survival and seizure phenotypes. We found hPNPO expression in GABAergic neurons largely restored lifespan and attenuated seizure activity, while glial expression also improved sgll phenotypes albeit to a lesser degree. In contrast, hPNPO expression in either cholinergic or glutamatergic neurons, accounting for most neurons in the fly brain, did not appreciably alter sgll phenotypes. We contrasted these observations with changes in sgll mutants induced by feeding GABA receptor modulators. The GABAB agonist SKF-97541 reduced mortality, while GABA or GABAA receptor modulators did not improve survival. Together, our data establish a cell-autonomous role for PNPO in GABAergic neurons to support brain function, especially under VB6-restricted conditions.

Sound-evoked auditory neurophysiological signals are a window into prodromal functional differences in a preclinical model of Alzheimer's Disease
Hearing has been identified as the largest modifiable mid-life risk factor for Alzheimer's Disease (AD), despite that its link to dementia remains unclear. Here we identify a biomarker of AD risk in an auditory neural signal using the non-invasive, rapidly acquired, and clinically translatable auditory brainstem response (ABR) in normal hearing knock-in rats (Swedish familial AD risk variant to Amyloid precursor protein, AppS; male and female). While ABR morphology has been proposed as a biomarker for AD, we report a novel biomarker derived from multidimensional parametric feature extraction on the distribution statistics of repeated single traces of the ABR that increases its potential for clinical utility with sensitivity and specificity. We report accurate prediction of AD genetic risk in both young and aged animals: AppS rats separate clearly from humanized controls in both sex- and age-dependent manners. Notably, auditory learning in young adulthood normalized the ABR signature in AppS rats with maintained effects into older age, altogether supporting a central neural generator of auditory dysfunction related to AD risk. These preclinical findings show how ABRs could provide a very early biomarker for detection of AD risk and lay the groundwork to test the synergy of auditory and cognitive functions in human dementia.

Neural Circuit Remodeling Underlying Enhanced Feeding During Pregnancy in Mice
Pregnancy represents a unique physiological state marked by substantial adaptations influencing both behavior and metabolism. Alterations have been observed across multiple levels, from changes in cellular morphology, synaptic density, and neuroglial interactions to a brain-wide reduction in gray matter volume. While these adaptations are thought to support offspring survival and promote maternal physiological and psychological well-being, their functional significance remains largely unresolved. In particular, linking localized synaptic plasticity to a broader brain-wide circuit dynamics has posed a major challenge. Using virus-mediated screening in mice, we investigated the pregnancy-induced remodeling of the presynaptic landscape of paraventricular hypothalamic (PVH) oxytocin neurons, a central hub of maternal physiology. Here we show a selective and reversible elimination of excitatory synaptic inputs from the medial preoptic nucleus to PVH oxytocin neurons, displaying region-, cell-type-, and pathway-specificity. This reduction in excitatory drive attenuated the anorexigenic activity of oxytocin neurons during feeding, thereby promoting the increased food intake characteristic of pregnancy. These data demonstrate that the selective suppression of anorexigenic PVH oxytocin neuron activity plays a critical role in appetite regulation during pregnancy. More broadly, our data identify a pregnancy-associated remodeling of hypothalamic neural circuits with direct physiological relevance.

Reverse Predictivity: Going Beyond One-Way Mapping to Compare Artificial Neural Network Models and Brains
A major goal in systems neuroscience is to build computational models that capture the primate brain's internal representations. Standard evaluations of artificial neural networks (ANNs) emphasize forward predictivity -- how well model features predict neural responses -- without testing whether model representations are themselves recoverable from neural activity. Here we develop the reverse predictivity metric, which quantifies how well macaque inferior temporal (IT) cortex responses predict ANN unit activations. This two-way framework reveals a striking asymmetry: models with high forward predictivity (~50% variance explained) often contain units unpredictable from neural activity, reflecting biologically inaccessible dimensions. In contrast, monkey-to-monkey mappings are symmetric, confirming that the asymmetry reflects genuine representational mismatch. Reverse predictivity isolates common ANN units -- shared with IT, behaviorally relevant, and generalizing across species -- and unique units lacking such alignment. Influenced by feature dimensionality, training objectives, and adversarial robustness, reverse predictivity offers a principled benchmark for guiding next-generation ANNs toward both high task performance and genuine biological plausibility.