Age-related alterations of membrane lipids in brain cell membranes together with high blood cholesterol are considered as major risk factors for Alzheimer's disease. Yet the molecular mechanisms by which these factors increase Alzheimer's risk are mostly unknown. In lipid raft domains of the plasma membrane, neurotoxic Alzheimer's beta-amyloid (Abeta) peptides interact with both cholesterol and ganglioside GM1. Recent data also suggested that cholesterol could stimulate the binding of Abeta to GM1 through conformational modulation of the ganglioside headgroup. Here we used a combination of physicochemical and molecular modeling approaches to decipher the mechanisms of cholesterol-assisted binding of Abeta to GM1. With the aim of decoupling the effect of cholesterol on GM1 from direct Abeta-cholesterol interactions, we designed a minimal peptide (Abeta5-16) containing the GM1-binding domain but lacking the amino acid residues involved in cholesterol recognition. Using the Langmuir technique, we showed that cholesterol (but not phosphatidylcholine or sphingomyelin) significantly accelerates the interaction of Abeta5-16 with GM1. Molecular dynamics simulations suggested that Abeta5-16 interacts with a cholesterol-stabilized dimer of GM1. The main structural effect of cholesterol is to establish a hydrogen-bond between its own OH group and the glycosidic-bond linking ceramide to the glycone part of GM1, thereby inducing a tilt in the glycolipid headgroup. This fine conformational tuning stabilizes the active conformation of the GM1 dimer whose headgroups, oriented in two opposite directions, form a chalice-shaped receptacle for Abeta. These data give new mechanistic insights into the stimulatory effect of cholesterol on Abeta/GM1 interactions. They also support the emerging concept that cholesterol is a universal modulator of protein-glycolipid interactions in the broader context of membrane recognition processes.

One of the most important lessons we have learned from sequencing the human genome is that not all proteins have a 3D structure. In fact, a large part of the human proteome is made up of intrinsically disordered proteins (IDPs) which can adopt multiple structures, and therefore, multiple functions, depending on the ligands with which they interact. Under these conditions, one can wonder about the value of algorithms developed for predicting the structure of proteins, in particular AlphaFold, an AI which claims to have solved the problem of protein structure. In a recent study, we highlighted a particular weakness of AlphaFold for membrane proteins. Based on this observation, we have proposed a paradigm, referred to as “Epigenetic Dimension of Protein Structure” (EDPS), which takes into account all environmental parameters that control the structure of a protein beyond the amino acid sequence (hence “epigenetic”). In this new study, we compare the reliability of the AlphaFold and Robetta algorithms’ predictions for a new set of membrane proteins involved in human pathologies. We found that Robetta was generally more accurate than AlphaFold for ascribing a membrane-compatible topology. Raft lipids (e.g., gangliosides), which control the structural dynamics of membrane protein structure through chaperone effects, were identified as major actors of the EDPS paradigm. We conclude that the epigenetic dimension of a protein structure is an intrinsic weakness of AI-based protein structure prediction, especially AlphaFold, which warrants further development.

Anandamide is a lipid neurotransmitter derived from arachidonic acid, a polyunsaturated fatty acid. The chemical differences between anandamide and arachidonic acid result in a slightly enhanced solubility in water and absence of an ionisable group for the neurotransmitter compared with the fatty acid. In this review, we first analyze the conformational flexibility of anandamide in aqueous and membrane phases. We next study the interaction of the neurotransmitter with membrane lipids and discuss the molecular basis of the unexpected selectivity of anandamide for cholesterol and ceramide from among other membrane lipids. We show that cholesterol behaves as a binding partner for anandamide, and that following an initial interaction mediated by the establishment of a hydrogen bond, anandamide is attracted towards the membrane interior, where it forms a molecular complex with cholesterol after a functional conformation adaptation to the apolar membrane milieu. The complex is then directed to the anandamide cannabinoid receptor (CB1) which displays a high affinity binding pocket for anandamide. We propose that cholesterol may regulate the entry and exit of anandamide in and out of CB1 by interacting with low affinity cholesterol recognition sites (CARC and CRAC) located in transmembrane helices. The mirror topology of cholesterol binding sites in the seventh transmembrane domain is consistent with the delivery, extraction and flip-flop of anandamide through a coordinated cholesterol-dependent mechanism. The binding of anandamide to ceramide illustrates another key function of membrane lipids which may occur independently of protein receptors. Interestingly, ceramide forms a tight complex with anandamide which blocks the degradation pathway of both lipids and could be exploited for anti-cancer therapies.

To assess the molecular epidemiology of human immunodeficiency virus type 1 (HIV-1), a screening method was developed for identification of non-B subtypes from sequence data obtained for resistance testing. The method is based on the evaluation of the percentage of divergence of a given sequence from the reference B subtype HXB2. Analysis of 1720 reverse-transcriptase (RT) and 1824 protease sequences stored in a database allowed for the determination of a threshold level of divergence from HXB2 above which a non-B subtype could be unambiguously characterized regardless of the pattern of resistance mutations (>8.6% for RT; >10.8% for protease). This conclusion was validated by phylogenetic analysis of RT, protease, and env genes. Overall, 72 (4.2%) and 73 (4.0%) non-B sequences were identified in the RT and protease coding regions, respectively. This method allows for the rapid detection of non-B subtypes among retrospective, recent, and future RT and/or protease sequence databases