Biology subjects

Rumbell, T. H.

Publications and source records attributed to Rumbell, T. H..

3 recordsLinked to original sources

bioRxiv2026

The lateral septum orchestrates state-dependent modulation of associative threat memory dynamics across the ovarian hormone cycle

Cycling ovarian hormones orchestrate neural states that greatly impact brain function and guide behavior. Here, we discover that cued threat memory processes are regulated across the mouse estrous cycle in a state-dependent manner. We show that low hormone states lead to overexpression of threat memory in females. However, high hormone states confer protection against this overexpression via state-dependent recruitment and reengagement of the lateral septum (LS) to the memory ensemble. Chemogenetic manipulations confirmed the LS as necessary and sufficient to suppress female memory overexpression. Single nucleus sequencing revealed a novel LS neuron population, defined by coexpression of neurotensin and somatostatin, selectively recruited to the female memory ensemble during high hormone states. This estradiol-sensitive population exhibits a female bias in projection strength to the nucleus accumbens and displays unique calcium dynamics associated with state-dependent memory suppression. Altogether, we uncover a sexually divergent neural mechanism whereby cycling ovarian hormones modulate memory expression.

neuroscience↗

bioRxiv2023

Inferring parameters of pyramidal neuronexcitability in mouse models of Alzheimer'sdisease using biophysical modeling and deeplearning

Alzheimers disease (AD) is believed to occur when abnormal amounts of the proteins amyloid beta and tau aggregate in the brain, resulting in a progressive loss of neuronal function. Hippocampal neurons in transgenic mice with amyloidopathy or tauopathy exhibit altered intrinsic excitability properties. We introduce a novel parameter inference technique, deep hybrid modeling (DeepHM), that combines deep learning with biophysical modeling to map experimental data recorded from hippocampal CA1 neurons in transgenic AD mice and age-matched wildtype littermate controls to the parameter space of a conductance-based CA1 model. Although mechanistic modeling and machine learning methods are by themselves powerful tools for approximating biological systems and making accurate predictions from data, when used in isolation these approaches suffer from distinct shortcomings: model and parameter uncertainty limit mechanistic modeling, whereas machine learning methods disregard the underlying biophysical mechanisms. DeepHM addresses these shortcomings by using conditional generative adversarial networks (cGANs) to provide an inverse mapping of data to mechanistic models that identifies the distributions of mechanistic modeling parameters coherent to the data. Here, we demonstrate that DeepHM accurately infers parameter distributions of the conductance-based model and outperforms a Markov chain Monte Carlo method on several test cases using synthetic data. We then use DeepHM to estimate parameter distributions corresponding to the experimental data and infer which ion channels are altered in the Alzheimers mouse models compared to their wildtype controls at 12 and 24 months. We find that the conductances most disrupted by tauopathy, amyloidopathy, and aging are delayed rectifier potassium, transient sodium, and hyperpolarization-activated potassium, respectively.

neuroscience↗

bioRxiv2020

Neuronal Population Models Reveal Specific Linear Conductance Controllers Sufficient to Rescue Preclinical Disease Phenotypes

Objective During the preclinical phase of drug development, potential drug candidates are often screened for their ability to alleviate certain in vitro electrophysiological features among neurons. This ability is assessed by measuring treatment outcomes using the population mean, both across different cells and different animals. The go/no-go decision for progressing a drug to a clinical trial is then based on ‘average effects’, yet these measures may not be sufficient to mitigate clinical end point risk. Population-based modeling is widely used to represent the intrinsic variability of electrophysiological features among healthy, disease and drug treated neuronal phenotypes. We pursued a method for optimizing therapeutic target design by identifying a single coherent set of ion channel targets for recovery of the healthy (Wild type) cellular phenotype simultaneously across multiple measures. Specifically, we aimed to determine the set of target modulations that best recover a heterogeneous Huntington’s disease (HD) population of model neurons into a multivariate region of phenotypic measurements corresponding to the healthy excitability profile of a heterogenous Wild type (WT) population of model neurons.Methods Our approach combines mechanistic simulations with populations modeling of striatal neurons using evolutionary algorithms for population optimization to design ‘virtual drugs’. We introduce efficacy metrics to score population of model outcomes and use these to rank our virtual candidates.Results We found that virtual drugs identified using heuristic approaches performed better than single target modulators and those derived from standard classification methods. We compare a real drug to the virtual candidates and demonstrate a novel in silico triaging method.Competing Interest StatementThe authors have declared no competing interest.View Full Text

pharmacology and toxicology↗