Biology subjects

Jensen, D.

Publications and source records attributed to Jensen, D..

11 recordsLinked to original sources

bioRxiv2026

Explaining how mutations affect AlphaFold predictions

Proteins can adopt many distinct conformations, yet the sequence determinants governing which structural states are accessible remain poorly understood. Using fold-switching proteins and AlphaFold as an analytical lens, we identify an organizational principle in which access to alternative structural states-and in some cases the folded state-is governed by surprisingly few amino acids, termed gating residues. Gating generalizes to single-fold proteins, indicating that conformational accessibility is often hierarchically organized around a small number of disproportionately influential residues. AlphaFold has implicitly learned this principle but amplifies and sometimes misapplies it, concentrating conformational control onto too few or incorrect residues. Guided by this principle, targeted MSA editing recovered a conformation AlphaFold confidently mispredicted, suggesting a path toward more accurate prediction of alternative conformations and mutational effects.

bioinformatics↗

bioRxiv2025

Mechanistic evidence that motif-gated domain recognition drives contact prediction in protein language models

Protein language models (pLMs) achieve state-of-the-art performance on protein structure and function prediction tasks, yet their internal computations re-main opaque. Sparse autoencoders (SAEs) have been used to recover sparse features, called latents, from pLM layer representations, whose activations cor-relate with known biological concepts. However, prior work has not established which model concepts are causally necessary for pLM performance on down-stream tasks. Here, we adapt causal activation patching to the pLM setting and perform it in SAE latent space to extract the minimal circuit responsible for accuracy in a contact prediction task for two case study proteins. We observe that preserving only a tiny fraction of latent-token pairs (0.022% and 0.015%) is sufficient to retain contact prediction accuracy in a residue unmasking experiment. Our circuit indicates a two-step computation in which early-layer motif detectors respond to short local sequence patterns, gating mid-to-late domain detectors which are selective for protein domains and families. Path-level ab-lations confirm the causal dependence of domain latents on upstream motif latents. To evaluate these components quantitatively, we introduce two diagnostics: a Motif Conservation Test and a Domain Selectivity Framework that supports hypothesis-driven tests. All candidate motif-detector latents pass the conservation test, and 18/23 candidate domain-detector latents achieve AUROC [≥]0.95. To our knowledge, this is the first circuits-style causal analysis for pLMs, pin-pointing the motifs, domains, and motif-domain interactions that drive contact prediction in two specific case studies. The framework introduced herein will enable future mechanistic dissection of protein language models. Code available at https://github.com/NainaniJatinZ/plm_circuits

bioinformatics↗

bioRxiv2025

Regulation of Steady State Ribosomal Transcription in Mycobacterium tuberculosis: Intersection of Sigma Subunits, Superhelicity, and Transcription Factors.

The regulation of ribosomal RNA (rRNA) is closely tied to nutrient availability, growth phase, and global gene expression, serving as a key factor in bacterial adaptability and pathogenicity. Mycobacterium tuberculosis (Mtb) stands out from other species with a single ribosomal operon controlled by two promoters: rrnAP3 and rrnAP1 and a high ratio of sigma ({sigma}) factors to genome size. While the primary {sigma} factor {sigma}A is known to drive ribosomal transcription, the alternative {sigma} factor {sigma}B has been proposed to contribute to the transcription of housekeeping genes, including rRNA under a range of conditions. However, {sigma}Bs precise role remains unclear. Here, we quantify steady-state rates in reconstituted transcription reactions and establish that {sigma}A-mediated transcription from rrnAP3 dominates rRNA production by almost two orders of magnitude with minimal contributions from {sigma}B holoenzymes and/or rrnAP1 under all conditions tested. We measure and compare the kinetics of individual initiation steps for both holoenzymes which, taken together with the steady-state rate measurements, lead us to a model where {sigma}B holoenzymes exhibit slower DNA unwinding and slower holoenzyme recycling. Our data further demonstrate that the transcription factors CarD and RbpA reverse or buffer the stimulatory effect of negative superhelicity on {sigma}A and {sigma}B holoenzymes respectively. Lastly, we show that a major determinant of {sigma}As increased activity is due to its N-terminal 205 amino acids. Taken together, our data reveal the intricate interplay of promoter sequence, {sigma} factor identity, DNA superhelicity, and transcription factors in shaping transcription initiation kinetics and, by extension, the steady-state rates of rRNA production in Mtb.

biochemistry↗

bioRxiv2025

Identification of a secreted protease from Bacteroides fragilis that induces intestinal pain and inflammation by cleavage of PAR2

Protease-activated receptor 2 (PAR2) is a central regulator of intestinal barrier function, inflammation and pain. Upregulated intestinal proteolysis and PAR2-signaling are implicated in inflammatory bowel diseases (IBDs) and irritable bowel syndrome (IBS). To identify potential bacterial regulators of PAR2 activity, we developed a functional assay for PAR2 processing and used it to screen conditioned media from a library of diverse gut commensal microbes. We found that multiple bacteria secrete proteases that cleave host PAR2. Using chemoproteomic profiling with a covalent irreversible inhibitor, we identified a previously uncharacterized Bacteroides fragilis serine protease Bfp1, and showed that it cleaves and activates PAR2 in multicellular and murine models. PAR2 cleavage by Bfp1 disrupts the intestinal barrier, sensitizes nociceptors, and triggers colonic inflammation and abdominal pain. Collectively, our findings uncover Bfp1-mediated PAR2-processing as a new axis of host-commensal-interaction in the gut that has the potential to be targeted for therapeutic intervention in IBD or IBS.

microbiology↗

bioRxiv2024

Human single-neuron activity is modulated by intracranial theta burst stimulation of the basolateral amygdala

Direct electrical stimulation of the human brain has been used for numerous clinical and scientific applications. At present, however, little is known about how intracranial stimulation affects activity at the microscale. In this study, we recorded intracranial EEG data from a cohort of patients with medically refractory epilepsy as they completed a visual recognition memory task. During the memory task, brief trains of intracranial theta burst stimulation (TBS) were delivered to the basolateral amygdala (BLA). Using simultaneous microelectrode recordings, we isolated neurons in the hippocampus, amygdala, orbitofrontal cortex, and anterior cingulate cortex and tested whether stimulation enhanced or suppressed firing rates. Additionally, we characterized the properties of modulated neurons, clustered presumed excitatory and inhibitory neurons by waveform morphology, and examined the extent to which modulation affected memory task performance. We observed a subset of neurons ([~]30%) whose firing rate was modulated by TBS, exhibiting highly heterogeneous responses with respect to onset latency, duration, and direction of effect. Notably, location and baseline activity predicted which neurons were most susceptible to modulation, although the impact of this neuronal modulation on memory remains unclear. These findings advance our limited understanding of how focal electrical fields influence neuronal firing at the single-cell level. HighlightsIndividual neurons in the human brain were responsive to theta burst stimulation Basolateral amygdala stimulation preferentially modulated neurons in the hippocampus, orbitofrontal cortex, and amygdala Excitatory and inhibitory neurons were modulated equally Neurons modulated by stimulation tended to have greater baseline firing rates Duration, onset, and valence of neuronal modulation were heterogeneous

neuroscience↗

bioRxiv2023

A Multi-Layered Computational Structural Genomics Approach Enhances Domain-Specific Interpretation of Kleefstra Syndrome Variants in EHMT1

This study investigates the functional significance of assorted variants of uncertain significance (VUS) in euchromatic histone lysine methyltransferase 1 (EHMT1), which is critical for early development and normal physiology. EHMT1 mutations cause Kleefstra syndrome and are linked to various human cancers. However, accurate functional interpretation of these variants are yet to be made, limiting diagnoses and future research. To overcome this, we integrate conventional tools for variant calling with computational biophysics and biochemistry to conduct multi-layered mechanistic analyses of the SET catalytic domain of EHMT1, which is critical for this protein function. We use molecular mechanics and molecular dynamics (MD)-based metrics to analyze the SET domain structure and functional motions resulting from 97 Kleefstra syndrome missense variants within this domain. Our approach allows us to classify the variants in a mechanistic manner into SV (Structural Variant), DV (Dynamic Variant), SDV (Structural and Dynamic Variant), and VUS (Variant of Uncertain Significance). Our findings reveal that the damaging variants are mostly mapped around the active site, substrate binding site, and pre-SET regions. Overall, we report an improvement for this method over conventional tools for variant interpretation and simultaneously provide a molecular mechanism of variant dysfunction.

genomics↗

bioRxiv2023

High-throughput, fluorescent-aptamer-based measurements of steady-state transcription rates for Mycobacterium tuberculosis RNA polymerase

The first step in gene expression is the transcription of DNA sequences into RNA. Regulation at the level of transcription leads to changes in steady-state concentrations of RNA transcripts, affecting the flux of downstream functions and ultimately cellular phenotypes. Changes in transcript levels are routinely followed in cellular contexts via genome-wide sequencing techniques. However, in vitro mechanistic studies of transcription have lagged with respect to throughput. Here, we describe the use of a real-time, fluorescent-aptamer-based method to quantitate steady-state transcription rates of the Mycobacterium tuberculosis RNA polymerase. We present clear controls to show that the assay specifically reports on promoter-dependent, full-length RNA transcription rates that are in good agreement with the kinetics determined by gel-resolved, -32P NTP incorporation experiments. We illustrate how the time-dependent changes in fluorescence can be used to measure regulatory effects of nucleotide concentrations and identity, RNAP and DNA concentrations, transcription factors, and antibiotics. Our data showcase the ability to easily perform hundreds of parallel steady-state measurements across varying conditions with high precision and reproducibility to facilitate the study of the molecular mechanisms of bacterial transcription. Significance StatementRNA polymerase transcription mechanisms have largely been determined from in vitro kinetic and structural biology methods. In contrast to the limited throughput of these approaches, in vivo RNA sequencing provides genome-wide measurements but lacks the ability to dissect direct biochemical from indirect genetic mechanisms. Here, we present a method that bridges this gap, permitting high-throughput fluorescence-based measurements of in vitro steady-state transcription kinetics. We illustrate how an RNA-aptamer-based detection system can be used to generate quantitative information on direct mechanisms of transcriptional regulation and discuss the far-reaching implications for future applications.

biochemistry↗

bioRxiv2021

Mycobacterium tuberculosis DNA-repair helicase UvrD1 is activated by redox-dependent dimerization via a conserved 2B domain cysteine.

Mycobacterium tuberculosis (Mtb) causes Tuberculosis and, during infection, is exposed to reactive oxygen species (ROS) and reactive nitrogen intermediates (RNI) from the host immune response that can cause DNA damage. UvrD-like proteins are involved in DNA repair and replication and belong to the SF1 family of DNA helicases that use ATP hydrolysis to catalyze DNA unwinding. In Mtb, there are two UvrD-like enzymes where UvrD1 is most closely related to other family members. Previous studies have suggested that UvrD1 is exclusively monomeric, however it is well-known that E. coli UvrD and other UvrD-family members exhibit monomer-dimer equilibria and unwind as dimers in the absence of accessory factors. Here, we reconcile these incongruent studies by showing that Mtb UvrD1 exists in monomer, dimer, and tetramer oligomeric forms where dimerization is regulated by redox potential. We identify a 2B domain cysteine, conserved in many Actinobacteria, that underlies this effect. We also show that UvrD1 DNA unwinding activity correlates specifically with the dimer population and is thus titrated directly via increasing positive (i.e. oxidative) redox potential. Consistent with the regulatory role of the 2B domain and the dimerization-based activation of DNA unwinding in UvrD-family helicases, these results suggest that UvrD1 is activated under oxidizing conditions when it may be needed to respond to DNA damage during infection.

biochemistry↗

bioRxiv2020

Multimodal brain-age prediction and cardiovascular risk: The Whitehall II MRI sub-study

Brain age is becoming a widely applied imaging-based biomarker of neural aging and potential proxy for brain integrity and health. We estimated multimodal and modality-specific brain age in the Whitehall II MRI cohort using machine learning and imaging-derived measures of gray matter morphology, diffusion-based white matter microstructure, and resting state functional connectivity. Ten-fold cross validation yielded multimodal and modality-specific brain age estimates for each participant, and additional predictions based on a separate training sample was included for comparison. The results showed equivalent age prediction accuracy between the multimodal model and the gray and white matter models (R2 of 0.34, 0.31, and 0.31, respectively), while the functional connectivity model showed a lower prediction accuracy (R2 of 0.01). Cardiovascular risk factors, including high blood pressure, alcohol intake, and stroke risk score, were each associated with more apparent brain aging, with consistent associations across modalities.

neuroscience↗

bioRxiv2018

Opioids for breathlessness: Psychological and neural factors influencing variability in response

Effective management of distressing bodily symptoms (such as pain and breathlessness) is an important clinical goal. However, extensive variability exists in both symptom perception and treatment response. One theory for understanding variability in bodily perception is the Bayesian Brian Hypothesis, whereby symptoms may result from the combination of sensory inputs and prior expectations. In light of this theory, we explored the relationships between opioid responsiveness, behavioural/physiological symptom modulators and brain activity during anticipation of breathlessness. MethodsWe utilised two existing opioid datasets to investigate the relationship between opioid efficacy and physiological/behavioural qualities, employing hierarchical cluster analyses in: 1) a clinical study in chronic obstructive pulmonary disease, and 2) a functional magnetic resonance brain imaging study of breathlessness in healthy volunteers. We also investigated how opioid efficacy relates to anticipatory brain activity using linear regression in the healthy volunteers. ResultsConsistent across both datasets, diminished opioid efficacy was more closely associated with negative affect than with other physiological and behavioural properties. Furthermore, in healthy individuals, brain activity in the anterior cingulate and ventromedial prefrontal cortices during anticipation of breathlessness were inversely correlated with opioid effectiveness. ConclusionDiminished opioid efficacy for relief of breathlessness may be associated with high negative affective qualities, and was correlated with the magnitude of brain activity during anticipation of breathlessness. Clinical implicationsNegative affect and symptom expectations may influence perceptual systems to become less responsive to opioid therapy.

neuroscience↗

bioRxiv2018

Offline impact of transcranial focused ultrasound on cortical activation in primates

To understand brain circuits it is necessary both to record and manipulate their activity. Despite increased availability of techniques for manipulating neural activity in rodents, manipulating neural activity in primates remains difficult. Here we show that a minimally invasive technique, transcranial focused ultrasound stimulation (TUS), induced offline changes to activity of circumscribed brain regions in the macaque. Applying TUS to the supplementary motor area or frontal polar cortex resulted in spatially specific patterns of activity change measurable with functional magnetic resonance imaging. In each case changes reflected each areas known network of interactions with the rest of the brain. Independent of these specific neural effects, TUS over these regions also induced widespread signal changes that might have a non-neuronal origin possibly mediated by the cerebral spinal fluid compartment. Although sustained for more than one hour beyond the 40s stimulation period, TUS effects were reversible and not associated with histological changes.\n\nHighlights O_LIMany studies of ultrasound neuromodulation focus on online effects in rodents.\nC_LIO_LIWe use fMRI connectivity to investigate its offline impact in the primate brain.\nC_LIO_LI40 s of ultrasound leads to a sustained, specific, reversible neural modulation.\nC_LIO_LIUltrasound caused a sharpening of the stimulated regions connectivity profile.\nC_LI\n\nIn BriefA new application of focused ultrasound safely modulates brain activation in primates for up to 2 hours after 40 seconds of stimulation. Ultrasound caused the stimulated area to interact more selectively with the rest of the brain.

neuroscience↗