Biology subjects

Fantini, J.

Publications and source records attributed to Fantini, J..

9 recordsLinked to original sources

bioRxiv2026

Leucine-rich glioma inactivated 1 is a ganglioside-binding protein

In LGI1-linked animal models of inherited autosomal dominant lateral temporal lobe epilepsy, increased neuronal excitability is accompanied by modifications in the AMPA/NMDA receptor ratio and a large decrease in Kv1 type potassium channels. However, the mechanism which links the absence of LGI1 to reduced expression of key neuronal ion channels is unknown. We observed multiple conserved canonical ganglioside-binding domains (GBDs) within human LGI1, mainly located in the EPTP domain. We show that GT1b is co-captured from native rat brain extracts by human LGI1 antibodies and, using SPR analysis, that recombinant full length LGI1 interacted with liposomes containing GT1b and GM1, but not GM3, lyso-lactosylceramide, phosphatidylserine or phosphatidylcholine. The ganglioside binding capacity of GBD peptide sequences exposed at the surface of LGI1 were confirmed using SPR and Langmuir film balance. Our data suggest that LGI1 interacts with gangliosides and may be involved in organizing lipid membrane platforms accommodating functional protein complexes. The loss of LGI1 could destabilize these platforms and contribute to reduced expression of key ion channels in Lgi1-/- mice.

neuroscience↗

bioRxiv2025

Temporal Perturbation Scanning: AI-Driven Deconstruction of Universal Biomolecular Recognition Mechanisms

Understanding molecular interactions at the atomic level remains a central challenge across biology, medicine, and engineering. We introduce Perturbation Scanning (PS), an interrogative AI framework that actively deconstructs molecular interfaces. PS integrates graph-based representations of structures derived from molecular dynamics trajectories or Protein Data Bank files to systematically probe the electrostatic, hydrophobic, and steric contributions of each residue. To translate these insights into actionable design, we introduce the Intelligent Interface Optimization Scanner (IIOS), a standalone tool that generates energy-scored mutation proposals from interface maps. Together, PS and IIOS provide an integrated platform for dissecting and rationally engineering molecular interactions by resolving force-specific and stage-dependent contributions that are not directly accessible with existing computational approaches. Unlike traditional alanine scanning or free-energy methods such as MMPBSA--which provide only static or ensemble-averaged measures--PS delivers stage-resolved, force-wise decomposition of binding interfaces via a suite of targeted physicochemical perturbations applied to its AI model, directly quantifying each residues mechanistic role.

bioinformatics↗

bioRxiv2025

In Silico Study of Membrane-Inserted Synaptotagmin Conformational Changes in Complex with Botulinum Toxin B1

Synaptotagmins 1 and 2 (SYT1 and SYT2) are essential Ca2+ sensors in neurotransmission and the functional receptors of botulinum neurotoxin B1 (BoNT/B1). While crystallographic models have defined key contacts, they neglect membrane constraints. Using molecular dynamics simulations in lipid rafts, we uncover how the membrane environment reshaped synaptotagmin conformation and enables critical contacts with BoNT/B1s lipid-binding loop (LBL). Notably, ganglioside GT1b bridges BoNT/B1 and SYT1, stabilizing their interface through lipid-mediated interactions. This mechanism escapes AlphaFold prediction, which generates non-physiological complexes with steric clashed, revealing a fundamental limitation of current AI methods for membrane-constrained interfaces. Our study demonstrates that lipid rafts create functional binding modes through synergistic protein-lipid interactions, highlighting the epigenetic dimension of protein structure where environment dictates conformation.

biochemistry↗

bioRxiv2025

Long-Term Intra-Host Evolution of SARS-CoV-2 in an Immunocompromised Patient: Recombination and Within-Host Mutations Driving Viral Adaptation

An immuno-compromised patient with lymphoma experienced a prolonged SARS-CoV-2 infection lasting 14 months, initially infected with B.1.160, followed by B.1.1.7. This study focused on intra-host single nucleotide variants (iSNVs) and single nucleotide polymorphisms (SNPs), distinguishing their origin as either within-host mutations or parental. The whole genome analysis revealed accelerated evolution, with positive selection detected in key genes such as Spike, N, ORF9b, and nsp13, all involved in viral replication and immune evasion. Of the two evolutionary mechanisms involved, host-driven mutation has played a dominant role in this evolutionary story. C>T transitions emerged as the most widespread mutational signature, consistent with host-driven RNA editing mechanisms. Two host-internal mutations, A28271T in the translation initiation region of the N Kozak gene and C26858T in the M gene, were highly shared among the deposited SARS-CoV-2 sequences. Recombination with parental lineages played a major role particularly in Spike, ORF3a and M genes. In Spike, the B.1.1.7 sequence was selected, wild-type ORF3a and M were restored, suggesting a selective advantage in returning to an ancestral sequence. In addition, the temporary emergence of intra-host convergent mutations in Spike, notably L5F, D796H and T572I, underlines the strong selective pressures exerted on this gene. A 126-nucleotide deletion in ORF8 resulted in a truncated protein, reinforcing its uselessness for viral replication, as observed in circulating variants such as B.1.1.7. The present case highlights the complex interplay between viral recombination and mutations within the host in chronic infections and further underscores the remarkable evolutionary plasticity of SARS-CoV-2 and its potential to generate highly adapted viral strains.

evolutionary biology↗

bioRxiv2023

Role of SARS-CoV-2 mutations in the evolution of the COVID-19 pandemic

RNA viruses, including SARS-CoV-2, evolve by mutation acquisition, or by hybridization between viral genomes. The SARS-CoV-2 pandemic provided an exceptional opportunity to analyze the mutations that appeared over a three-year period. In this study, we analysed the type of mutations and their epidemic consequences on the thousands of genomes produced in our laboratory. These were obtained by next-generation sequencing from respiratory samples performed for genomic surveillance. The frequencies of mutations were calculated using Nextclade, Microsoft Excel, and an in-house Python script. In total, 61,397 genomes matching 483 Pangolin lineages were analyzed; 22,225 nucleotide mutations were identified, and of them 220 (1.0%) were each at the root of at least 836 genomes, a frequency threshold classifying mutations as "hyperfertile". Two of these seeded the pandemic in Europe, namely a mutation in the RNA-dependent RNA polymerase associated with an increased mutation rate (P323L) and one in the spike protein (D614G), which plays a particular role in virus fitness. Most of these 220 "hyperfertile" mutations occurred in areas not predicted to be associated with increased virulence. Their number was 8{+/-}6 (0-22) per 1,000 nucleotides on average per gene. They were 3.7 times more frequent in accessory than informational genes (14 versus 4; p= 0.0037). Particularly, they were 4.1 times more frequent in ORF8 than in the gene encoding RNA polymerase. Interestingly, stop codons were present in 97 positions, almost only in six accessory genes including ORF7a (25 per 100 codons) and ORF8 (21). Furthermore, 1,661 mutations (16.3%) were associated with a lower number of "offspring" (50-835) and classified as "fertile". In conclusion, except for two initial mutations that could predict a change in the dynamics of the epidemic (mutation rate and change in the virus attachment site), most of the "hyperfertile" mutations did not predict the emergence of a new epidemic form. Significantly, some mutations were in non-coding areas and some consisted of stop codons, indicating that some genes (particularly ORF7a and ORF8) were rather "non-virulence genes" at a given stage of the epidemic, which is an unusual concept for viruses.

microbiology↗

bioRxiv2022

SARS-CoV-2 Spike Protein Induces Hemagglutination: Implications for COVID-19 Morbidities and Therapeutics and for Vaccine Adverse Effects

Experimental findings for SARS-CoV-2 related to the glycan biochemistry of coronaviruses indicate that attachments from spike protein to glycoconjugates on the surfaces of red blood cells (RBCs), other blood cells and endothelial cells are key to the infectivity and morbidity of COVID-19. To provide further insight into these glycan attachments and their potential clinical relevance, the classic hemagglutination (HA) assay was applied using spike protein from the Wuhan, Alpha, Delta and Omicron B.1.1.529 lineages of SARS-CoV-2 mixed with human RBCs. The electrostatic potential of the central region of spike protein from these four lineages was studied through molecular modeling simulations. Inhibition of spike protein-induced HA was tested using the macrocyclic lactone ivermectin (IVM), which is indicated to bind strongly to SARS-CoV-2 spike protein glycan sites. The results of these experiments were, first, that spike protein from these four lineages of SARS-CoV-2 induced HA. Omicron induced HA at a significantly lower threshold concentration of spike protein than for the three prior lineages and was much more electropositive on its central spike protein region. IVM blocked HA when added to RBCs prior to spike protein and reversed HA when added afterwards. These results validate and extend prior findings on the role of glycan bindings of viral spike protein in COVID-19. They furthermore suggest therapeutic options using competitive glycan-binding agents such as IVM and may help elucidate rare serious adverse effects (AEs) associated with COVID-19 mRNA vaccines which use spike protein as the generated antigen.

biochemistry↗

bioRxiv2022

All-atom Molecular Dynamics Simulation of botulinum neurotoxin serotype A1 in complex with human SV2A or SV2C receptors in a lipid raft environment

Botulinum neurotoxin A1 (BoNT/A1) is the most potent serotype in humans with the highest clinical duration. BoNT/A1 interacts with synaptic vesicle glycoprotein 2 (SV2) and gangliosides to be taken up by neurons. In this study, we present three molecular dynamics simulations in which BoNT/A1 is in complex with singly or doubly glycosylated SV2C or singly glycosylated SV2A, in a ganglioside rich (lipid raft) context. Our computational data suggest that the N-glycan at position 480 (N480g) in the luminal domain of SV2C (LD-SV2C) indirectly enhanced the contacts of the neurotoxin surface with the second N-glycan at position 559 (N559g) by acting as a shield to prevent N559g to interact with residues of LD-SV2C. The N-glycosylation at the position N573 (N573g) in the luminal domain of SV2A has a slightly lower affinity for the surface of BoNT/A1 compared to 559g because of possible intermolecular contacts between N573g and residues of the luminal domain of SV2A (LD-SV2A). In addition to the ganglioside binding site (GBS) conserved across serotypes B, E, F and G, the lipid-raft associated GT1b interacted with a structure we coined the ganglioside binding loop (GBL) which is homologous to the lipid binding loop (LBL) in serotypes B, C, D, D/C and G. Finally, we proposed a global model in which BoNT/A1 interacts with its glycosylated protein receptor, one molecule of GT1b interacting in the GBS and five molecules of GT1b interacting with the GBL and residue Y1133. These data solved the puzzle generated by mutational studies that could be only partially understood with crystallographic data that lack both a biologically relevant membrane environment and a full glycosylation of SV2. Brief statementWe propose a full molecular description of the initial binding of a microbial toxin (Botulinum neurotoxin A1) to the surface of neural cells. Our model includes a protein receptor (SV2) in its native environment, i.e. the periphery of a cluster of gangliosides belonging to a membrane microdomain (lipid raft). A major outcome of our study is the elucidation of the role of the full length glycans (previously resolved by MS spectroscopy) covalently attached to the protein receptor. These data solved the puzzle generated by mutational studies that could be only partially understood with crystallographic data that lack both a biologically relevant membrane environment and a full glycosylation of SV2.

neuroscience↗

bioRxiv2021

The protein-binding pocket of Botulinum neurotoxin B accommodates a preassembled synaptotagmin / ganglioside complex

Botulinum neurotoxin serotype B (BoNT/B) uses two separate protein and polysialoglycolipid-binding pockets to interact with synaptotagmin 1/2 and gangliosides. However, an integrated model of BoNT/B bound to its neuronal receptors in a native membrane topology is still lacking. Using a panel of in silico and experimental approaches, we present here a new model for BoNT/B binding to neuronal membranes, in which the toxin binds to a preassembled synaptotagmin-ganglioside GT1b complex and a free ganglioside allowing a lipid-binding loop of BoNT/B to interact with the glycone part of the synaptotagmin-associated GT1b. Furthermore, our data provide molecular support for the decrease in BoNT/B sensitivity in Felidae that harbor the natural variant synaptotagmin2-N59Q. These results reveal multiple interactions of BoNT/B with gangliosides and support a novel paradigm in which a toxin recognizes a protein/ganglioside complex.

pharmacology and toxicology↗

bioRxiv2021

Long-lasting reduction of intrinsic excitability in O-LM interneurons is mediated by endocannabinoid-dependent up-regulation of Kv7 channels

KCNQ-Kv7 channels are found at the axon initial segment of pyramidal neurons where they control cell firing and membrane potential. In oriens lacunosum moleculare (O-LM) interneurons, these channels are mainly expressed in the dendrites, suggesting a peculiar function of Kv7 channels in these neurons. The physiology of Kv7 channels is well characterized today but the precise contribution of these channels to neuronal plasticity is still unknown. Here, we show that Kv7 channel activity is up-regulated following induction of presynaptic long-term synaptic depression (LTD) in O-LM interneurons, thus resulting in a synergistic long-term depression of intrinsic neuronal excitability (LTD-IE). Both LTD and LTD-IE involve endocannabinoid (eCB) biosynthesis for their induction. Molecular modeling shows strong interaction of eCBs with Kv7.2/3 channel, suggesting a persistent action of these lipids on Kv7 channel activity. Our data thus unveil a major role for eCB synthesis in triggering both synaptic and intrinsic depression in O-LM interneurons.

neuroscience↗