Faculty Bibliography

Spatially modulated grid cells have been recently found in the rat secondary visual cortex (V2) during active navigation. However, the computational mechanism and functional significance of V2 grid cells remain unknown. To address the knowledge gap, we train a biologically inspired excitatory-inhibitory recurrent neural network to perform a two-dimensional spatial navigation task with multisensory input. We find grid-like responses in both excitatory and inhibitory RNN units, which are robust with respect to spatial cues, dimensionality of visual input, and activation function. Population responses reveal a low-dimensional, torus-like manifold and attractor. We find a link between functional grid clusters with similar receptive fields and structured excitatory-to-excitatory connections. Additionally, multistable torus-like attractors emerged with increasing sparsity in inter- and intra-subnetwork connectivity. Finally, irregular grid patterns are found in recurrent neural network (RNN) units during a visual sequence recognition task. Together, our results suggest common computational mechanisms of V2 grid cells for spatial and non-spatial tasks.

Background: The ability to maintain balance is an evolutionarily-conserved behavior that is frequently disrupted found in patients with neurodegenerative diseases. One of the most prominent balance disorders is found in patients with progressive supranuclear palsy (PSP), a primary tauopathy pathologically characterized by tau over-representation in the brainstem vestibulospinal (VS) nucleus, where they frequently exhibit accidental-backward falls starting from the early stage of the disease. Although pathological features of PSP correlate well with its clinical phenotype, how tau aggregation affects neuronal and circuit functions, which eventually leads to behavioral deficits, remains unclear. Method: To dissect disease mechanisms across molecular, cellular, circuitry, and behavioral levels, we generated tau fish by expressing human 0N/4R-Tau in zebrafish VS nucleus. Tau expression and phosphorylation were validated using immunohistochemistry staining with PHF-1 antibody. To examine the effect of tau on balance behavior, we measured postural control of free-swimming tau fish and compared to that of tau-negative siblings. Moreover, we tested response of VS neurons to nose-down and nose-up tilt stimulus using 2-photon calcium imaging. Result: Ttau-expressing zebrafish exhibit impaired balance control while maintaining normal locomotor ability. Interestingly, we did not observe any neuronal death in the VS nucleus. Functional imaging of the VS nucleus shows impaired directional tuning in tau-expressing neurons in response to tilt stimulus. We also found ectopic accumulation of acidic organelles in the cell bodies of tau-positive neurons, suggesting abnormal lysosomal function. Conclusion: Our results demonstrate how molecular abnormalities disrupt specific behavior in tauopathies before neurodegeneration appeared.

Background: Astrocytes can have helpful or harmful effects on neuron health and brain function in disease. While they normally provide trophic support to neurons during development and normal functioning, in response to many stimuli their heterogeneous "˜reactive"™ responses can alter these functions drastically. Changes in astrocyte function depends on their "˜reactive"™ sub-state. Understanding when and where sub-states of reactive astrocytes occur, and how these altered functions contribute to disease will pave the way for novel strategies to protect neurons. Method: We performed combined 10x genomics single cell and spatial transcriptomics in wildtype and Alzheimer"™s disease (AD) model mice, combined with single nuclei RNA sequencing of human postmortem non-symptomatic and AD patient brains. Results: With improved capture rates and subsequent powering of astrocyte sequencing we highlight lowly abundant, biologically important, reactive astrocyte sub-states that are positioned in strategic locations throughout the brain "“ namely at sites of entry for peripheral immune cells (e.g. adjacent to penetrating vessels in layer I of the cortex, and around the ventricles). Further, we integrate our datasets with previously published scRNAseq and snRNAseq datasets to confirm these small populations exist in other patient populations. Most surprising was that interferon-responsive reactive astrocytes were present early progression of pathology in the 5xFAD mouse AD model, but not at later time points "“ suggesting important early (possibly protective) roles for astrocytes early in AD. Additionally, when comparing mouse and human datasets we find most disease pathology-associated reactive astrocytes are located around strategic points of entry to the brain, and express many inflammation-responsive transcripts. Probing for "˜modules"™ of genes associated with inflammation-response and reactive sub-states of microglia and other immune cells highlights putative interactions likely integral for feedback between these two cell types. Conclusion: Optimization of astrocyte capture for single cell/nuclei sequencing combined with integration of previously published datasets increased the size of datasets for analysis and power of our analysis. Our data highlight several novel reactive astrocyte sub-states that warrant additional functional characterization and further investigation.

Background: Alzheimer"™s disease (AD) is a neurodegenerative illness characterized by progressive accumulation of amyloid beta (Aβ) and neurofibrillary tangles, with cognitive impairment and altered neural activity. Hyperexcitability in the early stages of AD contribute to Aβ accumulation and cognitive impairment, aggravating the progression of AD. However, the hyperexcitability origin is not clear. This study aimed to test whether mossy cells (MCs), an excitatory cell of the hippocampal dentate gyrus, show increased excitability at early stages of AD and contribute to the increased network excitability generation. Indeed, alterations of MCs contribute to hyperexcitability and cognitive impairment in epilepsy. However, the role of MCs in AD has not been substantially explored. Methods: Intrinsic and synaptic properties of MCs and granule cells (GCs) from WT and Tg2576 mice at early ages (1-2 m.o.) were characterized by whole-cell patch-clamp recordings. Synaptic properties included the frequency and amplitude of spontaneous excitatory postsynaptic potentials (EPSPs) and excitatory and inhibitory postsynaptic currents (EPSCs and IPSCs). Deterioration in MCs morphology was evaluated using Nissl staining and GluR2/3 labeling by light- and confocal microscopy. Aβ deposition was evaluated using the McSA1 antibody. Results: Tg2576 GCs did not have any significant difference in their intrinsic properties, as we shown previously in mice ∼3 m.o. However, an enhanced excitatory and inhibitory input to GCs, depicted by augmented IPSC (7.16 vs 14.04 events/s) and NMDA-mediated EPSC frequencies (0.81 vs 1.41 events/s) were found. Interestingly, Tg2576 MCs had an augmented EPSP frequency (5.75 vs 9.44 events/s), and their intrinsic properties showed a depolarized RMP (-72.88 vs -58.36 mV), and reduced rheobase (145.56 vs 47.14 pA), AP amplitude (98.14 vs 76.66 mV), time-to-peak (552.75 vs 266.16 ms) and maximum rise (171.44 vs 88.68 mV/ms) and decay slopes (-61.17 vs -42.38 mV/ms). The correlation between #APs and current injected showed Tg2576 MCs fired significantly more APs (SEZD = 0.34; z = 2.48). Tg2576 MCs showed robust intracellular Aβ aggregation without any significant morphological change. Conclusions: MCs changes in excitability and early accumulation of Aβ suggest that MCs could be the cause of increased excitability occurring later in GCs. In this manner, MCs could be an important contributor to AD.

Beyond detecting brain lesions or tumors, comparatively little success has been attained in identifying brain disorders such as Alzheimer"™s disease (AD), based on magnetic resonance imaging (MRI). Many machine learning algorithms to detect AD have been trained using limited training data, meaning they often generalize poorly when applied to scans from previously unseen scanners/populations. Therefore, we built a practical brain MRI-based AD diagnostic classifier using deep learning/transfer learning on a dataset of unprecedented size and diversity. A retrospective MRI dataset pooled from more than 217 sites/scanners constituted one of the largest brain MRI samples to date (85,721 scans from 50,876 participants) between January 2017 and August 2021. Next, a state-of-the-art deep convolutional neural network, Inception-ResNet-V2, was built as a sex classifier with high generalization capability. The sex classifier achieved 94.9% accuracy and served as a base model in transfer learning for the objective diagnosis of AD. After transfer learning, the model fine-tuned for AD classification achieved 90.9% accuracy in leave-sites-out cross-validation on the Alzheimer"™s Disease Neuroimaging Initiative (ADNI, 6,857 samples) dataset and 94.5%/93.6%/91.1% accuracy for direct tests on three unseen independent datasets (AIBL, 669 samples / MIRIAD, 644 samples / OASIS, 1,123 samples). When this AD classifier was tested on brain images from unseen mild cognitive impairment (MCI) patients, MCI patients who converted to AD were 3 times more likely to be predicted as AD than MCI patients who did not convert (65.2% vs. 20.6%). Predicted scores from the AD classifier showed significant correlations with illness severity. In sum, the proposed AD classifier offers a medical-grade marker that has potential to be integrated into AD diagnostic practice.

An important problem in neuroscience is to understand how brains extract relevant signals from mixtures of unknown sources, i.e., perform blind source separation. To model how the brain performs this task, we seek a biologically plausible single-layer neural network implementation of a blind source separation algorithm. For biological plausibility, we require the network to satisfy the following three basic properties of neuronal circuits: (i) the network operates in the online setting; (ii) synaptic learning rules are local; and (iii) neuronal outputs are nonnegative. Closest is the work by Pehlevan et al. (Neural Comput 29:2925-2954, 2017), which considers nonnegative independent component analysis (NICA), a special case of blind source separation that assumes the mixture is a linear combination of uncorrelated, nonnegative sources. They derive an algorithm with a biologically plausible 2-layer network implementation. In this work, we improve upon their result by deriving 2 algorithms for NICA, each with a biologically plausible single-layer network implementation. The first algorithm maps onto a network with indirect lateral connections mediated by interneurons. The second algorithm maps onto a network with direct lateral connections and multi-compartmental output neurons.

ATP-sensitive K⁺ (K_ATP) channel couples membrane excitability to intracellular energy metabolism. Maintaining K_ATP channel surface expression is key to normal insulin secretion, blood pressure and cardioprotection. However, the molecular mechanisms regulating K_ATP channel internalization and endocytic recycling, which directly affect the surface expression of K_ATP channels, are poorly understood. Here we used the cardiac K_ATP channel subtype, Kir6.2/SUR2A, and characterized Rab35 GTPase as a key regulator of K_ATP channel endocytic recycling. Electrophysiological recordings and surface biotinylation assays showed decreased K_ATP channel surface density with co-expression of a dominant negative Rab35 mutant (Rab35-DN), but not other recycling-related Rab GTPases, including Rab4, Rab11a and Rab11b. Immunofluorescence images revealed strong colocalization of Rab35-DN with recycling Kir6.2. Rab35-DN minimized the recycling rate of K_ATP channels. Rab35 also regulated K_ATP channel current amplitude in isolated adult cardiomyocytes by affecting its surface expression but not channel properties, which validated its physiologic relevance and the potential of pharmacologic target for treating the diseases with K_ATP channel trafficking defects.

Decades of work propose that hippocampal activity supports internal representation of learned experiences and contexts, allowing individuals to form long-term memories and quickly adapt behavior to changing environments. However, recent studies insinuate hippocampal representations can drift over time, raising the question: how could the hippocampus hold stable memories when activity of its neuronal maps fluctuates? We hypothesized that task-dependent hippocampal maps set by learning rules and structured attention stabilize as a function of behavioral performance. To test this, we imaged hippocampal CA1 pyramidal neurons during learning and memory recall phases of a new task where mice use odor cues to navigate between two reward zones. Across learning, both orthogonal and overlapping task-dependent place maps form rapidly, discriminating trial context with strong correlation to behavioral performance. Once formed, task-selective place maps show increased long-term stability during memory recall phases. We conclude that memory demand and attention stabilize hippocampal activity to maintain contextually rich spatial representations.

Despite its increasing recognition and extensive research, there is no unifying hypothesis on the pathophysiology of the postural tachycardia syndrome. In this cross-sectional study, we examined the role of fear conditioning and its association with tachycardia and cerebral hypoperfusion upon standing in 28 patients with postural tachycardia syndrome (31 ± 12 years old, 25 women) and 21 matched controls. We found that patients had higher somatic vigilance (p = 0.0167) and more anxiety (p < 0.0001). They also had a more pronounced anticipatory tachycardia right before assuming the upright position in a tilt-table test (p = 0.015), a physiologic indicator of fear conditioning to orthostasis. While standing, patients had faster heart rate (p < 0.001), higher plasma catecholamine levels (p = 0.020), lower end-tidal CO2 (p = 0.005), and reduced middle cerebral artery blood flow velocity (p = 0.002). Multi-linear logistic regression modeling showed that both epinephrine secretion and excessive somatic vigilance predicted the magnitude of the tachycardia and the hyperventilation. These findings suggest that the postural tachycardia syndrome is a functional psychogenic disorder in which standing may acquire a frightful quality, so that even when experienced alone, it elicits a fearful conditioned response. Heightened somatic anxiety is associated with and may predispose to a fear-conditioned hyperadrenergic state when standing. Our results have therapeutic implications.