Biology subjects

Gerwert, K.

Publications and source records attributed to Gerwert, K..

8 recordsLinked to original sources

bioRxiv2026

A fast and accurate calculation method for light induced isomerization of retinal proteins in real time

Retinal is a chromophore covalently bound to various photoreceptors. Its photo-induced isomerization triggers a series of structural changes named photocycle, leading to diverse biological functions. Despite tremendous advances in structural biology and artificial intelligence-driven structure prediction, it remains challenging to analyze all photocyclic intermediates. Here, we present an optimized computational approach to calculate RSBH+ isomerization and its induced structural changes based on a classical molecular mechanics approach using quantum mechanically improved retinal force field parameters. Isomerization is induced by an excited state restraint which is subsequently relaxed to allow the return to the electronic ground state. We applied this approach to the key protein of optogenetics, Channelrhodopsin-2 from Chlamydomonas reinhardtii (CrChR2). Besides the reformation of the alltrans/CN-anti ground state, we observed the production of a mixture of two isomeric states 13-cis/CN- anti and 13-cis/CN-syn. These findings agree with the previously found branched photocycle model based on experimental data. Our calculations show an asymmetric potential energy landscape of the excited state leading to a corresponding isomerization state distribution. Unlike earlier publications, our procedure describes the retinal photoisomerization on the natural timescale of 500 fs. As our newly derived retinal force field parameter set precisely relies on quantum biological knowledge, it assists to improve the refinement of experimental structure biological data. Our readily customizable strategy provides mechanistic insights at high spatio-temporal resolution, which permits accurate structural predictions of early photocycle intermediates. These insights will stimulate the rational design of optogenetic tools thus providing improved diagnostic and therapeutic approaches for neuronal and other diseases. HighlightsO_LIuniversal method to study molecular mechanism of optogenetic tools C_LIO_LIretinal photo-isomerization calculation in real time C_LIO_LIprediction of branched photo cycle agrees with experimental IR spectroscopic results C_LIO_LIdetected asymmetric excited state potential energy landscape C_LIO_LIassists to improve structural model refinement of retinal proteins C_LI Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=116 HEIGHT=200 SRC="FIGDIR/small/707937v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@71f7fdorg.highwire.dtl.DTLVardef@503482org.highwire.dtl.DTLVardef@1a77120org.highwire.dtl.DTLVardef@1f410a0_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2025

IR spectroscopy: from experimental spectra to high-resolution structural analysis by integrating simulations and machine learning

Understanding biomolecular function at the atomic scale requires detailed insight into the structural changes underlying dynamic processes. Vibrational infrared (IR) spectroscopy--when paired with biomolecular simulations and quantum-chemical calculations--determines bond length variations on the order of 0.01 [A], providing insights into these structutral changes. Here, we address the forward problem in IR spectroscopy: predicting high-accuracy vibrational spectra from known molecular structures identified by biomolecular simulations. Solving this problem lays the groundwork for the inverse problem: inferring structural ensembles directly from experimental IR spectra. We evaluate two computational approaches, normal mode analysis and Fourier-transformed dipole autocorrelation, against experimental IR spectra of N-Methylacetamide, a prototypical model for peptide bond vibrations. Spectra are derived from simulation models at multiple levels of theory, including hybrid quantum mechanics/molecular mechanics, machine-learned and classical molecular mechanics approaches. Our results highlight the capabilities and limitations of current theoretical biophysical approaches to decode structural information from experimental vibrational spectroscopy data. These insights underscore the potential of future artificial intelligence (AI)-enhanced models to enable direct IR-based structure determination. For example, resolving the so far experimentally inaccessible structures of toxic oligomers involved in neurodegenerative diseases, enabling improved disease diagnostics and targeted therapies. TOC Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC="FIGDIR/small/665767v1_ufig1.gif" ALT="Figure 1"> View larger version (24K): org.highwire.dtl.DTLVardef@6603d3org.highwire.dtl.DTLVardef@1852c85org.highwire.dtl.DTLVardef@2ddaccorg.highwire.dtl.DTLVardef@77b554_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2025

Replacement of a single residue in an antibody completely abolishes cognate antigen binding, as predicted by theoretical methods

Structural insights into the interaction between antibodies and antigens at the atomic level are pivotal for understanding the molecular mechanisms of antigen binding. Despite the availability of structural models generated by recent artificial intelligence advancements, computational predictions require experimental validation to confirm their accuracy. Here, we demonstrate a novel approach that combines computational protein modeling with spectroscopic experiments to validate antibody-antigen interactions. As a case example we use solanezumab, a monoclonal antibody that targets amyloid-beta (A{beta}), whose misfolding is the main factor responsible for Alzheimers disease. For this antibody we predicted a single mutation, G95AHC, within the paratope of the heavy chain to disrupt antigen binding. This mutation, referred to as a "dead mutant", was experimentally validated using an immuno-infrared biosensor (iRS). Our results confirmed that the mutation abolished antigen binding without affecting the native structure of the antibody. The use of dead mutants enables precise differentiation between specific and nonspecific binding, which is particularly important in medical diagnostics. We applied this approach to analyze the binding of solanezumab to synthetically produced A{beta} variants and A{beta} catched by the iRS functionalized surface from cerebrospinal fluid, showcasing its utility in Alzheimers disease diagnostics. These findings highlight the value of computational modeling and experimental validation in understanding antigen-antibody interactions, with significant implications for diagnostic and therapeutic applications.

biophysics↗

bioRxiv2021

Microsecond resolved infrared spectroscopy on non-repetitive protein reactions by applying caged-compounds and quantum cascade laser frequency combs

Infrared spectroscopy is ideally suited for the investigation of protein reactions at the atomic level. Many systems were investigated successfully by applying Fourier transform infrared (FTIR) spectroscopy. While rapid-scan FTIR spectroscopy is limited by time resolution (about10 ms with 16 cm-1 resolution), step-scan FTIR spectroscopy reaches a time-resolution of about 10 ns but is limited to cyclic reactions that can be repeated hundreds of times under identical conditions. Consequently, FTIR with high time resolution was only possible with photoactivable proteins that undergo a photocycle. The huge number of non-repetitive reactions, e.g. induced by caged compounds, were limited to the ms time domain. The advent of dual comb quantum cascade laser allows now for a rapid reaction monitoring in the s time domain. Here we investigate the potential to apply such an instrument to the huge class of G-proteins. We compare caged-compound induced reactions monitored by FTIR and dual comb spectroscopy, respectively, by applying the new technique to the subunit of the inhibiting Gi protein and to the larger protein-protein complex of Gi with its cognate regulator of G-protein signaling (RGS). We observe good data quality with 4 s time resolution with a wavelength resolution comparable to FTIR. This is more than three orders of magnitude faster than any FTIR measurement on G-proteins in the literature. This study paves the way for infrared spectroscopic studies in the so far unresolvable s time regime for non-repetitive biological systems including all GTPases and ATPases.

biophysics↗

bioRxiv2019

Deep representation learning for domain adaptable classification of infrared spectral imaging data

MotivationApplying infrared microscopy in the context of tissue diagnostics heavily relies on computationally preprocessing the infrared pixel spectra that constitute an infrared microscopic image. Existing approaches involve physical models, which are non-linear in nature and lead to classifiers that do not generalize well, e.g. across different types of tissue preparation. Furthermore, existing preprocessing approaches involve iterative procedures that are computationally demanding, so that computation time required for preprocessing does not keep pace with recent progress in infrared microscopes which can capture whole-slide images within minutes.\n\nResultsWe investigate the application of stacked contractive autoencoders as an unsupervised approach to preprocess infrared microscopic pixel spectra, followed by supervised fine-tuning to obtain neural networks that can reliably resolve tissue structure. To validate the robustness of the resulting classifier, we demonstrate that a network trained on embedded tissue can be transferred to classify fresh frozen tissue. The features obtained from unsupervised pretraining thus generalize across the large spectral differences between embedded and fresh frozen tissue, where under previous approaches seperate classifiers had to be trained from scratch.\n\nAvailabilityOur implementation can be downloaded from https://github.com/arnrau/SCAE_IR_Spectral_Imaging\n\nContactaxel.mosig@bph.rub.de

bioinformatics↗

bioRxiv2018

A unifying photocycle model for light adaptation and temporal evolution of cation conductance in Channelrhodopsin-2

Although Channelrhodopsin (ChR) is a widely applied light-activated ion channel, important properties such as light-adaptation, photocurrent inactivation, and alteration of the ion selectivity during continuous illumination are not well-understood from a molecular perspective. Herein, we address these open questions using single turn-over electrophysiology, time-resolved step-scan FTIR and Raman spectroscopy of fully dark adapted ChR2. This yields a unifying parallel photocycle model explaining all data: in dark-adapted ChR2, the protonated Schiff base retinal chromophore (RSBH+) adopts an all-trans,C=N-anti conformation only. Upon light activation, a branching reaction into either a 13-cis,C=N-anti or a 13-cis,C=N-syn retinal conformation occurs. The anti-cycle features sequential H+ and Na+ conductance in a late M-like state and an N-like open-channel state. In contrast, the 13-cis,C=N-syn isomer represents a second closed-channel state identical to the long lived P480-state, which has been previously assigned to a late intermediate in a single photocycle model. Light excitation of P480 induces a parallel syn-photocycle with an open channel state of small conductance and high proton selectivity. E90 becomes deprotonated in P480 and stays deprotonated in the C=N-syn-cycle and we show that deprotonation of E90 and successive pore hydration are crucial for late proton conductance following light-adaptation. Parallel anti- and syn-photocycles explain inactivation and ion selectivity changes of ChR2 during continuous illumination, fostering the future rational design of optogenetic tools.\n\nSignificance statementUnderstanding the mechanisms of photoactivated biological processes facilitates the development of new molecular tools, engineered for specific optogenetic applications, allowing the control of neuronal activity with light. Here, we use a variety of experimental and theoretical techniques to examine the precise nature of the light-activated ion channel in one of the most important molecular species used in optogenetics, channelrhodopsin-2. Existing models for the photochemical and photophysical pathway after light absorption by the molecule fail to explain many aspects of its observed behavior including the inactivation of the photocurrent under continuous illumination. We resolve this by proposing a new branched photocycle explaining electrical and photochemical channel properties and establishing the structure of intermediates during channel turnover.

biophysics↗

bioRxiv2018

Detailed structural orchestration of Lewy pathology in Parkinson’s disease as revealed by 3D multicolor STED microscopy

Post-translational modifications of alpha-synuclein (aSyn), particularly phosphorylation at Serine 129 (Ser129-p) and truncation of its C-terminus (CTT), have been implicated in Parkinsons disease (PD) pathology. To gain more insight in the relevance of Ser129-p and CTT aSyn under physiological and pathological conditions, we investigated their subcellular distribution patterns in normal aged and PD brains using highly-selective antibodies in combination with 3D multicolor STED microscopy. We show that CTT aSyn localizes in mitochondria in PD patients and controls, whereas the organization of Ser129-p in a cytoplasmic network is strongly associated with pathology. Nigral Lewy bodies show an onion skin-like architecture, with a structured framework of Ser129-p aSyn and neurofilaments encapsulating CTT aSyn in their core, which displayed high content of proteins and lipids by label-free CARS microscopy. The subcellular phenotypes of antibody-labeled pathology identified in this study provide evidence for a crucial role of Ser129-p aSyn in Lewy body formation.

neuroscience↗

bioRxiv2017

Lewy pathology in Parkinson’s disease consists of a crowded organellar membranous medley

Parkinsons disease, the most common age-related movement disorder, is a progressive neurodegenerative disease with unclear etiology. Key neuropathological hallmarks are Lewy bodies and Lewy neurites, which are neuronal inclusions that are immunopositive for the protein -synuclein. In-depth ultrastructural analysis of this Lewy pathology is crucial to understanding pathogenesis and progression of the disease. Using correlative light and electron microscopy/tomography on brain tissue from five Parkinsons disease brain donors, we identified -synuclein immunopositive Lewy pathology and could show that the majority of these features including Lewy bodies and Lewy neurites primarily consists of a crowded membranous medley of vesicular structures and dysmorphic organelles. Only a small fraction of observed Lewy bodies contained predominant proteinaceous filaments, as previously described. The crowding of organellar components was confirmed by STED- based super-resolution microscopy, and high lipid content within the -synuclein immunopositive inclusions was corroborated by confocal imaging, CARS/FTIR imaging and lipidomics. Applying this correlative high-resolution imaging and biophysical approach, we discovered in the postmortem brain of Parkinsons patients a subcellular protein-lipid compartmentalization not previously described in Lewy pathology.

neuroscience↗