Biology subjects

Miranda, L.

Publications and source records attributed to Miranda, L..

8 recordsLinked to original sources

bioRxiv2026

Machine learning-guided discovery of a conserved plasmid proteomic signature enables MALDI-TOF MS detection of pOXA-48-carrying Enterobacterales

OXA-48 carbapenemases are among the most widespread and important resistance mechanisms in Enterobacterales. Yet detecting carbapenemases by conventional workflows necessitates additional testing, thus delaying optimization of therapy and implementation of infection control measures. Here, we present a machine learning approach that identifies the conserved pOXA-48 plasmid directly from routine MALDI-TOF spectra acquired for species identification. The model detects pOXA-48 carriers with an AUROC of 0.96-0.98 across two independent hospital cohorts and instrument platforms, indicating near-perfect discrimination. Using bottom-up proteomics, plasmid conjugation, and plasmid curing, we link the discriminative MALDI-TOF spectral features to proteins encoded on pOXA-48, with DUF1496 domain-containing protein producing the most discriminative spectral feature. Our approach reframes the resistance prediction task from inferring a resistance phenotype to detecting a conserved plasmid through its expressed proteomic signature and has the potential to enable rapid MALDI-TOF MS-based diagnostics for a wide range of plasmid-based resistance determinants.

microbiology↗

bioRxiv2025

Loss of noradrenergic Fkbp5 disrupts social behavior and norepinephrine dynamics in the basolateral amygdala

Social dysfunction is common in depression and varies with stress exposure and genetic risk. The current study identifies a cell-type specific role for Fkbp5, a glucocorticoid receptor co-chaperone, in noradrenergic neurons engaged during social stress. Acute social stress upregulated Fkbp5 in the locus coeruleus (LC), whereas repeated exposure attenuated this effect. Noradrenergic Fkbp5 deletion (Fkbp5Nat) increased pro-social behavior exclusively in male mice. In the basolateral amygdala (BLA), social interaction reduced norepinephrine (NE) turnover in wild-type but not Fkbp5Nat mice. Consistently, proteomics revealed mitochondrial/energy and synapse-related remodeling in BLA neurons. Miniscope imaging showed that behavior-locked NE transients in BLA were selectively blunted in Fkbp5Nat mice during interaction with outbred CD1 conspecifics, while same-strain C57BL/6N encounters preserved NE dynamics. Together, this study indicates that Fkbp5 tunes LC-BLA output to social salience in a sex- and context-dependent manner, suggesting a circuit-specific route to normalize social salience without broadly suppressing noradrenergic function.

neuroscience↗

bioRxiv2025

Hyperdominance and rarity in Amazonian secondary forests

Recent studies have shown that a few tree species dominate abundance and carbon storage in undisturbed Amazonian forests. However, this pattern remains poorly understood in forests regrowing after deforestation. Understanding dominance patterns in secondary forests is important as it could improve insights into their functioning and guide management and restoration efforts. We used unweighted and novel plot-weighted approaches to assess hyperdominance in abundance and carbon storage of trees and palms during secondary forest recovery, using data from 102 plots across four regions in the eastern Amazon, separated by an average distance of 600 km (range: 150-1,000 km). Secondary forests show extreme hyperdominance in all approaches, with just 16-25 species ([~]3.6-5.6%) accounting for over 50% of abundance and carbon storage. A small subset of species ([~]9%) were found across all regions and successional stages, highlighting their key role in ecosystem structure and function during forest regrowth. Annona exsucca, Cecropia palmata, Tapirira guianensis, Inga alba, Attalea maripa, Amphiodon effusus, Banara guianensis, and Vismia guianensis consistently emerged as hyperdominant across stem size classes, approaches, and metrics, jointly accounting for roughly 16-27% of total abundance and carbon in Amazonian secondary forests. For larger stems >10cm DBH, hyperdominant species exhibited lower wood density compared to non-hyperdominant species across all successional stages, suggesting these specialist pioneers have distinct ecological strategies from the rest of the community. Our findings indicate that secondary forest functioning can be largely predicted by a relatively limited set of species, providing a framework for future research ranging from ecophysiology to remote sensing of canopy traits and restoration planning.

ecology↗

bioRxiv2025

Early detection of ampicillin susceptibility in Enterococcus faecium with MALDI-TOF MS and machine learning

BackgroundEnterococcus faecium can cause severe infections and is often resistant to the first-line antibiotic ampicillin. Consequently, clinicians usually prescribe broad-spectrum antibiotics, promoting the selection of multidrug-resistant bacteria. In this study, we investigate the application of machine learning techniques to detect ampicillin susceptibility directly from MALDI-TOF mass spectrometry. This technique could enable an earlier optimised treatment in infections with ampicillin-susceptible E. faecium. MethodsTwo datasets of clinical E. faecium MALDI-TOF spectra and their resistance phenotype were analysed: our own Technical University of Munich (TUM) dataset and the publicly available MS-UMG dataset. We tested logistic regression (LR) and LightGBM models on each dataset via nested cross-validation and explored transferability on the respective other dataset. ResultsLightGBM demonstrated slightly better performance than LR in identifying susceptible isolates in the TUM dataset (area under the precision-recall curve (AUPRC) 0.907 {+/-} 0.016 vs 0.902 {+/-} 0.030) as well as in the MS-UMG dataset (AUPRC 0.902 {+/-} 0.029 vs 0.899 {+/-} 0.054). External validation demonstrated good model transferability (AUPRC of 0.784 {+/-} 0.039 when trained on MS-UMG; 0.804 {+/-} 0.013 when trained on TUM). SHAP analysis consistently identified a top-ranked spectral feature corresponding to a peak at an m/z of 5091 in resistant isolate spectra. ConclusionThis study demonstrates that LR and LightGBM models can identify ampicillin-susceptible E. faecium isolates from MALDI-TOF spectra and generalise well to unseen datasets. While clinical implementation currently still requires confirmatory testing, the addition of larger datasets in the future will support the development of more robust machine learning models.

microbiology↗

bioRxiv2025

Single cell spatial proteomics maps human liver zonation patterns and their vulnerability to fibrosis

Understanding protein distribution patterns across tissue architecture is crucial for deciphering organ function in health and disease. Here, we applied single-cell Deep Visual Proteomics to perform spatially-resolved proteome analysis of individual cells in native tissue. We combined this with a novel strategic cell selection pipeline and a continuous protein gradient mapping framework to investigate larger clinical cohorts. We generated a comprehensive spatial map of the human hepatic proteome by analyzing hundreds of individual hepatocytes from 18 individuals. Among more than 2,500 proteins per cell about half exhibited zonated expression patterns. Cross-species comparison with mouse data revealed conserved metabolic functions and human-specific features of liver zonation. Analysis of fibrotic samples demonstrated widespread disruption of protein zonation, with pericentral proteins being particularly susceptible. Our study provides a comprehensive resource of human liver organization while establishing a broadly applicable framework for spatial proteomics analyses along tissue gradients.

molecular biology↗

bioRxiv2024

Sex-specific fear acquisition following early life stress is linked to amygdala glutamate metabolism

Early life stress (ELS) adversely affects physiological and behavioral outcomes, increasing the vulnerability to stress-related disorders, such as post-traumatic stress disorder (PTSD). PTSD prevalence is significantly higher in women and is partially mediated by genetic risk variants. Understanding how sex influences the interaction of PTSD risk genes, such as FKBP5, with trauma-related behaviors is crucial for uncovering PTSDs neurobiological pathways. The development of in-depth behavioral analysis tools using unsupervised behavioral classification is thereby a crucial tool to increase the understanding of the behavioral outcomes related to stress-induced fear memory formation. The current study investigates the sex-specific effects of ELS exposure by using the limited bedding and nesting (LBN) paradigm. The LBN exposure disrupted different facets of the hypothalamic-pituitary-adrenal (HPA) axis in a sex-specific manner directly after stress and at adult age. Moreover, freezing was altered by LBN exposure in both the acquisition and the retrieval of fear in a sex-dependent manner. Unsupervised behavioral analysis revealed a higher active fear response after LBN exposure during fear acquisition in females, but not in males. The regulation of the HPA axis is closely intertwined with cellular metabolism and core regulatory cascades. To investigate the impact of LBN exposure on tissue-specific metabolism, a metabolomic pathway analysis in the basolateral amygdala revealed a specific sex- and stress-dependent effect on purine, pyrimidine, and glutamate metabolism. The present study highlights the intricate interplay between metabolic pathways and the neurobiological substrates implicated in fear memory formation and stress regulation. Overall, these findings highlight the importance of considering sex-specific metabolic alterations in understanding the neurobiological mechanisms underlying stress-related disorders and offer potential avenues for targeted interventions.

animal behavior and cognition↗

bioRxiv2022

Automatically annotated motion tracking identifies a distinct social behavioral profile following chronic social defeat stress

Severe stress exposure is a global problem with long-lasting negative behavioral and physiological consequences, increasing the risk of stress-related disorders such as major depressive disorder (MDD). An essential characteristic of MDD is the impairment of social functioning and lack of social motivation. Chronic social defeat stress is an established animal model for MDD research, which induces a cascade of physiological and social behavioral changes. The current developments of markerless pose estimation tools allow for more complex and socially relevant behavioral tests, but the application of these tools to social behavior remains to be explored. Here, we introduce the open-source tool "DeepOF" to investigate the individual and social behavioral profile in mice by providing supervised and unsupervised pipelines using DeepLabCut annotated pose estimation data. The supervised pipeline relies on pre-trained classifiers to detect defined traits for both single and dyadic animal behavior. Subsequently, the unsupervised pipeline explores the behavioral repertoire of the animals without label priming, which has the potential of pointing towards previously unrecognized motion motifs that are systematically different across conditions. We here provide evidence that the DeepOF supervised and unsupervised pipelines detect a distinct stress-induced social behavioral pattern, which was particularly observed at the beginning of a novel social encounter. The stress-induced social behavior shows a state of arousal that fades with time due to habituation. In addition, while the classical social avoidance task does identify the stress-induced social behavioral differences, both DeepOF behavioral pipelines provide a clearer and more detailed profile. DeepOF aims to facilitate reproducibility and unification of behavioral classification of social behavior by providing an open-source tool, which can significantly advance the study of rodent individual and social behavior, thereby enabling novel biological insights as well as drug development for psychiatric disorders.

neuroscience↗

bioRxiv2021

Unique role and vulnerability of EP300 KIX domain in small-cell lung cancer

EP300 (E1A binding protein p300) is a versatile transcription co-activator important in cell proliferation and differentiation. The gene EP300 is frequently mutated in diverse cancer types, including small-cell lung cancer (SCLC). While it is widely believed that these mutations result in loss of EP300 function, the impact on SCLC pathogenesis remains largely unknown. Here we demonstrate that mutant EP300 variants lacking histone acetyltransferase (HAT) domain accelerate tumor development in autochthonous mouse models of SCLC. However, unexpectedly, complete knockout of Ep300 suppresses tumor development and inhibits proliferation of both human and mouse SCLC cells. Genetic dissection of EP300 domains identifies kinase-inducible domain (KID)-interacting (KIX) domain, specifically its interaction with transcription factors such as CREB1 and MYB, as the determinant of pro-tumorigenic activity. Blockade of the KIX-mediated protein interactions using a small molecule and a recombinant peptide mimicking the KIX-binding sequences of EP300-interacting partners inhibits the growth of SCLC cells. These findings identify domain-specific roles of EP300 in SCLC and unique vulnerability of the EP300 KIX domain to potential therapeutics.

cancer biology↗