Abstract Mechanisms linking CD33 variants to Alzheimer Disease (AD) are poorly defined. Here, we combine structural, cellular, and genetic analyses to delineate how the CD33 M splice isoform, upregulated in carriers of CD33 risk alleles, modulates microglial function. We show that CD33 M ectodomain dimerizes, enabling binding of large multi-sialylated molecules. We demonstrate that another AD risk protein - clusterin (CLU) ± Aβ oligomers (but not ApoE) binds with nanomolar avidity to CD33 M , but not CD33 m . We show that in human monocytes CD33 M :CLU binding induces CD33 M ITIM phosphorylation, recruits SHP-1, suppresses Aβ phagocytosis, and impairs clearance of amyloid plaques. We identify a soluble CD33 M ectodomain fragment (sCD33 M ) - absent from CD33 m -expressing cells - which could contribute to the role of CD33 M in AD. Genetic analyses confirm that CD33:CLU interaction modulates amyloid burden, cognition, and disease risk. These findings define a mechanistic CLU:Aβ:CD33 M axis, highlighting CD33 M dimerization and ligand-binding sites as potential therapeutic targets.

The sialic-acid binding immunoglobulin-like lectin 3 receptor (Siglec-3 / CD33) is one of the highly associated AD risk genes. Previous studies revealed that the non-coding AD-risk alleles (rs3865444 and rs12459419) are associated with increased total levels of CD33 expression and a higher relative expression of the long CD33 M splice form. However, the molecular basis of the immuno-inhibitory function of CD33 remains unclear. To confirm the presence of CD33 dimers, we conducted multiple experiments. Blue Native Gel electrophoresis and co-immunoprecipitation assays were used to detect CD33 bands corresponding to the expected molecular weights. Flow cytometry with specific antibodies was performed to quantify cell-surface CD33. Additionally, single-molecule fluorescence resonance energy transfer (smFRET) combined with TIRF imaging was employed to visualize CD33 dimers on the cell surface. Furthermore, Western Blotting of phosphorylation of CD33 and its downstream molecule were performed to verify if the dimers were functional. Biochemical analyses demonstrated that CD33 M and CD33 m can form homodimers or heterodimers. Flow cytometry confirmed that CD33 M isoforms are selectively trafficked to the cell surface, while smFRET imaging verified the presence of dimers on the cell surface. The elevation of the CD33 pathway following stimulation with CD33-specific ligands provided evidence that CD33 M homodimers are functional. This study reveals the critical role of CD33 M in AD pathology by elucidating its molecular mechanisms. We provide direct evidence that CD33 M and CD33 m isoforms can form both homodimers and heterodimers. However, only CD33 M isoforms are preferentially trafficked to the cell surface and form functional dimers. These findings advance our understanding of the molecular basis of CD33's immuno-inhibitory function and offer new insights into its involvement in AD risk, potentially paving the way for the development of targeted therapeutic strategies.

We report the results of structural, functional and genetic studies on the CD33 sialic acid- binding receptor that reveal how non-coding variants in CD33 alter risk for Alzheimer's disease (AD). The full-length CD33 M isoform, whose expression is upregulated by non-coding AD-risk alleles, preferentially forms dimers at the cell surface, where they interact with AD-related proteins (clusterin and Aβ). This interaction induces CD33 M inhibitory signalling and downregulates protective microglial functions including phagocytic removal of amyloid plaques. Human brain expression quantitative trait loci (eQTL) and causal mediation analyses confirm that quantitative interactions between CLU and CD33 genotypes modulate AD phenotypes and suggest that genotypes at these loci might be used to personalise future therapeutic approaches. Our work also highlights several other unexpected aspects of CD33 biology, including a soluble shed extracellular fragment of CD33 M and a similar soluble secreted product arising from a truncating mutation in the CD33 extracellular domain (CD33 MΔ4bp ).

The sialic-acid binding immunoglobulin-like lectin 3 receptor (Siglec-3 / CD33) expressed on microglia, regulates immune functions relevant to Alzheimer's disease (AD). Clusterin (CLU) and apolipoprotein E (ApoE) are soluble, sialylated proteins implicated in AD pathogenesis through genetic associations and their interactions with amyloid-beta (Aβ). However, the role of these proteins as potential CD33 ligands remains unclear. This study explores whether CLU and/or ApoE bind CD33 and examines the functional impact of these interactions on Aβ uptake and amyloid plaque clearance. The binding of CD33 to CLU and ApoE was assessed through co-immunoprecipitation using U937 cells (endogenously expressing CD33) and HEK293 cells (expressing exogenous CD33). Quantitative bio-layer interferometry (BLI) and microscale thermophoresis determined binding affinities, focusing on the role of CD33's Arg119 sialic acid binding site. In situ proximity ligation assays (PLA) and co-immunoprecipitation from AD and control human brain lysates validated in vivo interactions. Functional assays examined Aβ uptake and amyloid plaque clearance in monocytes and U937 cells, with or without CLU treatment. The quantitative binding assay revealed that CLU, but not ApoE, was a sialylation-dependent ligand for CD33, binding with high affinity (Kd = 28.9 ± 10.3 nM). Binding required an intact Arg119 residue and dimeric CD33 structure. PLA and co-immunoprecipitation studies demonstrated colocalization of CD33 and CLU on microglia in AD brains, especially near amyloid plaques. Functionally, sialylated CLU inhibited Aβ uptake in monocytes from CD33 “CC” risk allele carriers and reduced amyloid plaque clearance in U937 cells. Desialylated CLU showed no significant effects. Notably, CLU + Aβ oligomers induced stronger CD33 ITIM signaling than CLU alone, enhancing phosphorylation and SHP-1 recruitment. This study identifies CLU as a specific CD33 ligand and highlights its role in modulating microglial functions via CD33 ITIM signaling. Sialylated CLU inhibits Aβ uptake and amyloid plaque clearance, suggesting a potential mechanism underlying microglial dysfunction in AD. These findings underscore the therapeutic potential of targeting the CD33-CLU axis to restore microglial homeostasis and enhance amyloid clearance in AD.

Computational free energy-based methods have the potential to significantly improve throughput and decrease costs of protein design efforts. Such methods must reach a high level of reliability, accuracy, and automation to be effectively deployed in practical industrial settings in a way that impacts protein design projects. Here, we present a benchmark study for the calculation of relative changes in protein-protein binding affinity for single point mutations across a variety of systems from the literature, using free energy perturbation (FEP+) calculations. We describe a method for robust treatment of alternate protonation states for titratable amino acids, which yields improved correlation with and reduced error compared to experimental binding free energies. Following careful analysis of the largest outlier cases in our dataset, we assess limitations of the default FEP+ protocols and introduce an automated script which identifies probable outlier cases that may require additional scrutiny and calculates an empirical correction for a subset of charge-related outliers. Through a series of three additional case study systems, we discuss how protein FEP+ can be applied to real-world protein design projects, and suggest areas of further study.

Chemical inducer of dimerization (CID) modules can be used effectively as molecular switches to control biological processes, and thus there is significant interest within the synthetic biology community in identifying novel CID systems. To date, CID modules have been used primarily in engineering cells for in vitro applications. To broaden their utility to the clinical setting, including the potential to control cell and gene therapies, the identification of novel CID modules should consider factors such as the safety and pharmacokinetic profile of the small molecule inducer, and the orthogonality and immunogenicity of the protein components. Here we describe a CID module based on the orally available, approved, small molecule simeprevir and its target, the NS3/4A protease from hepatitis C virus. We demonstrate the utility of this CID module as a molecular switch to control biological processes such as gene expression and apoptosis in vitro, and show that the CID system can be used to rapidly induce apoptosis in tumor cells in a xenograft mouse model, leading to complete tumor regression.

BACKGROUND:Epithelial damage, repair and remodelling are critical features of chronic airway diseases including chronic obstructive pulmonary disease (COPD). Interleukin (IL)-33 released from damaged airway epithelia causes inflammation via its receptor, serum stimulation-2 (ST2). Oxidation of IL-33 to a non-ST2-binding form (IL-33ox) is thought to limit its activity. We investigated whether IL-33ox has functional activities that are independent of ST2 in the airway epithelium.METHODS:In vitro epithelial damage assays and three-dimensional, air-liquid interface (ALI) cell culture models of healthy and COPD epithelia were used to elucidate the functional role of IL-33ox. Transcriptomic changes occurring in healthy ALI cultures treated with IL-33ox and COPD ALI cultures treated with an IL-33-neutralising antibody were assessed with bulk and single-cell RNA sequencing analysis.RESULTS:We demonstrate that IL-33ox forms a complex with receptor for advanced glycation end products (RAGE) and epidermal growth factor receptor (EGFR) expressed on airway epithelium. Activation of this alternative, ST2-independent pathway impaired epithelial wound closure and induced airway epithelial remodelling in vitro. IL-33ox increased the proportion of mucus-producing cells and reduced epithelial defence functions, mimicking pathogenic traits of COPD. Neutralisation of the IL-33ox pathway reversed these deleterious traits in COPD epithelia. Gene signatures defining the pathogenic effects of IL-33ox were enriched in airway epithelia from patients with severe COPD.CONCLUSIONS:Our study reveals for the first time that IL-33, RAGE and EGFR act together in an ST2-independent pathway in the airway epithelium and govern abnormal epithelial remodelling and muco-obstructive features in COPD.

Background: Remodelling of the airway epithelium is a key feature of chronic obstructive pulmonary disease (COPD). Interleukin (IL)-33 is a multifunctional cytokine that drives COPD pathology. Upon tissue damage epithelial cells release IL-33, which binds its receptor ST2 on immune cells leading to airway inflammation. Oxidation of IL-33 via disulphide bond formation, results in conformational changes that disrupt ST2 binding. Objective: To identify whether oxidised IL-33 (IL-33ox) has functional activities that are independent of ST2 in the airway epithelium. Methods: We used in vitro epithelial damage assays and 3D cell culture models of healthy and COPD epithelia to elucidate the role of IL-33ox. Tozorakimab (MEDI3506), a high-affinity human IgG1 monoclonal antibody, was used to inhibit IL-33 signalling. Results: IL-33ox activates an alternative, ST2-independent pathway through a signalling complex of receptor for advanced glycation end products (RAGE) and epidermal growth factor receptor (EGFR) expressed on the airway epithelium. Activation of the IL-33ox–RAGE/EGFR pathway in healthy airway epithelial cells impaired epithelial wound closure and led to mucus hypersecretion, mimicking the phenotype of COPD-derived cells. In COPD epithelia, inhibition of IL-33ox signalling reduced mucus synthesis and secretion thereby reversing key pathogenic features of COPD. Gene set enrichment analysis showed that gene signatures defining the effects of IL-33ox were enriched in airway epithelia from patients with severe COPD. Conclusion: Our data reveal a previously unknown IL-33ox–RAGE/EGFR epithelial signalling pathway, which governs key features of COPD and could be an attractive therapeutic target.

TREM2 is a pattern recognition receptor, expressed on microglia and myeloid cells, detecting lipids and Aβ and inducing an innate immune response. Missense mutations (e.g., R47H) of TREM2 increase risk of Alzheimer's disease (AD). The soluble ectodomain of wild-type TREM2 (sTREM2) has been shown to protect against AD in vivo, but the underlying mechanisms are unclear. We show that Aβ oligomers bind to cellular TREM2, inducing shedding of the sTREM2 domain. Wild-type sTREM2 bound to Aβ oligomers (measured by single-molecule imaging, dot blots, and Bio-Layer Interferometry) inhibited Aβ oligomerization and disaggregated preformed Aβ oligomers and protofibrils (measured by transmission electron microscopy, dot blots, and size-exclusion chromatography). Wild-type sTREM2 also inhibited Aβ fibrillization (measured by imaging and thioflavin T fluorescence) and blocked Aβ-induced neurotoxicity (measured by permeabilization of artificial membranes and by loss of neurons in primary neuronal-glial cocultures). In contrast, the R47H AD-risk variant of sTREM2 is less able to bind and disaggregate oligomeric Aβ but rather promotes Aβ protofibril formation and neurotoxicity. Thus, in addition to inducing an immune response, wild-type TREM2 may protect against amyloid pathology by the Aβ-induced release of sTREM2, which blocks Aβ aggregation and neurotoxicity. In contrast, R47H sTREM2 promotes Aβ aggregation into protofibril that may be toxic to neurons. These findings may explain how wild-type sTREM2 apparently protects against AD in vivo and why a single copy of the R47H variant gene is associated with increased AD risk.

Missense mutations (e.g. R47H) of the microglial receptor TREM2 increase risk of Alzheimer’s disease (AD), and the soluble ectodomain of wild-type TREM2 (sTREM2) appears to protect in vivo, but the underlying mechanisms are unclear. We show that Aβ oligomers bind to TREM2, inducing shedding of sTREM2. Wild-type sTREM2 inhibits Aβ oligomerization, fibrillization and neurotoxicity, and disaggregates preformed Aβ oligomers and protofibrils. In contrast, the R47H AD-risk variant of sTREM2 is less able to bind and disaggregate oligomeric Aβ, but rather promotes Aβ protofibril formation and neurotoxicity. Thus, in addition to mediating phagocytosis, wild-type TREM2 may protect against amyloid pathology by Aβ-induced release of sTREM2 that blocks Aβ aggregation and neurotoxicity; while R47H sTREM2 promotes Aβ aggregation into neurotoxic forms, which may explain why the R47H variant gene increases AD risk several fold.

Partial loss-of-function variants in the TREM2 immune receptor are associated with increased risk for Alzheimer's disease (AD) and other forms of neurodegenerative disease, but the molecular bases for these connections are unknown. Three new structures of WT and R47H mutant TREM2 immunoglobulin-like (Ig-like) domain now reveal that R47 functions to correctly position elements of the ligand-binding surface. Intriguingly, the authors also demonstrate a disruption of receptor oligomerization by the R47H mutation, suggesting a role for ligand-induced clustering in receptor signaling and resultant plaque clearance.

We have characterised the proteolytic cleavage events responsible for the shedding of triggering receptor expressed on myeloid cells 2 (TREM2) from primary cultures of human macrophages, murine microglia and TREM2-expressing human embryonic kidney (HEK293) cells. In all cell types, a soluble 17 kDa N-terminal cleavage fragment was shed into the conditioned media in a constitutive process that is inhibited by G1254023X and metalloprotease inhibitors and siRNA targeting ADAM10. Inhibitors of serine proteases and matrix metalloproteinases 2/9, and ADAM17 siRNA did not block TREM2 shedding. Peptidomimetic protease inhibitors highlighted a possible cleavage site, and mass spectrometry confirmed that shedding occurred predominantly at the H157-S158 peptide bond for both wild-type and H157Y human TREM2 and for the wild-type murine orthologue. Crucially, we also show that the Alzheimer’s disease-associated H157Y TREM2 variant was shed more rapidly than wild type from HEK293 cells, possibly by a novel, batimastatand ADAM10-siRNA-independent, sheddase activity. These insights offer new therapeutic targets for modulating the innate immune response in Alzheimer’s and other neurological diseases.

Hypoxia-inducible transcription factors (HIFs) control adaptation to low oxygen environments by activating genes involved in metabolism, angiogenesis, and redox homeostasis. The finding that HIFs are also regulated by small molecule metabolites highlights the need to understand the complexity of their cellular regulation. Here we use a forward genetic screen in near-haploid human cells to identify genes that stabilize HIFs under aerobic conditions. We identify two mitochondrial genes, oxoglutarate dehydrogenase (OGDH) and lipoic acid synthase (LIAS), which when mutated stabilize HIF1α in a non-hydroxylated form. Disruption of OGDH complex activity in OGDH or LIAS mutants promotes L-2-hydroxyglutarate formation, which inhibits the activity of the HIFα prolyl hydroxylases (PHDs) and TET 2-oxoglutarate dependent dioxygenases. We also find that PHD activity is decreased in patients with homozygous germline mutations in lipoic acid synthesis, leading to HIF1 activation. Thus, mutations affecting OGDHC activity may have broad implications for epigenetic regulation and tumorigenesis.

Inherited variants in multiple different genes are associated with increased risk for Alzheimer's disease (AD). In many of these genes, the inherited variants alter some aspect of the production or clearance of the neurotoxic amyloid β-peptide (Aβ). Thus missense, splice site or duplication mutants in the presenilin 1 (PS1), presenilin 2 (PS2) or the amyloid precursor protein (APP) genes, which alter the levels or shift the balance of Aβ produced, are associated with rare, highly penetrant autosomal dominant forms of Familial Alzheimer's Disease (FAD). Similarly, the more prevalent late-onset forms of AD are associated with both coding and non-coding variants in genes such as SORL1, PICALM and ABCA7 that affect the production and clearance of Aβ. This review summarises some of the recent molecular and structural work on the role of these genes and the proteins coded by them in the biology of Aβ. We also briefly outline how the emerging knowledge about the pathways involved in Aβ generation and clearance can be potentially targeted therapeutically. This article is part of Special Issue entitled "Neuronal Protein".

The mechanisms by which mutations in FUS and other RNA binding proteins cause ALS and FTD remain controversial. We propose a model in which low-complexity (LC) domains of FUS drive its physiologically reversible assembly into membrane-free, liquid droplet and hydrogel-like structures. ALS/FTD mutations in LC or non-LC domains induce further phase transition into poorly soluble fibrillar hydrogels distinct from conventional amyloids. These assemblies are necessary and sufficient for neurotoxicity in a C. elegans model of FUS-dependent neurodegeneration. They trap other ribonucleoprotein (RNP) granule components and disrupt RNP granule function. One consequence is impairment of new protein synthesis by cytoplasmic RNP granules in axon terminals, where RNP granules regulate local RNA metabolism and translation. Nuclear FUS granules may be similarly affected. Inhibiting formation of these fibrillar hydrogel assemblies mitigates neurotoxicity and suggests a potential therapeutic strategy that may also be applicable to ALS/FTD associated with mutations in other RNA binding proteins.

The presenilin genes were first identified as the site of missense mutations causing early onset autosomal dominant familial Alzheimer's disease. Subsequent work has shown that the presenilin proteins are the catalytic subunits of a hetero-tetrameric complex containing APH1, nicastrin and PEN-2. This complex (variously termed presenilin complex or gamma-secretase complex) performs an unusual type of proteolysis in which the transmembrane domains of Type I proteins are cleaved within the hydrophobic compartment of the membrane. This review describes some of the molecular and structural biology of this unusual enzyme complex. The presenilin complex is a bilobed structure. The head domain contains the ectodomain of nicastrin. The base domain contains a central cavity with a lateral cleft that likely provides the route for access of the substrate to the catalytic cavity within the centre of the base domain. There are reciprocal allosteric interactions between various sites in the complex that affect its function. For instance, binding of Compound E, a peptidomimetic inhibitor to the PS1 N-terminus, induces significant conformational changes that reduces substrate binding at the initial substrate docking site, and thus inhibits substrate cleavage. However, there is a reciprocal allosteric interaction between these sites such that prior binding of the substrate to the initial docking site paradoxically increases the binding of the Compound E peptidomimetic inhibitor. Such reciprocal interactions are likely to form the basis of a gating mechanism that underlies access of substrate to the catalytic site. An increasingly detailed understanding of the structural biology of the presenilin complex is an essential step towards rational design of substrate- and/or cleavage site-specific modulators of presenilin complex function.

Presenilin-mediated endoproteolysis of transmembrane proteins plays a key role in physiological signaling and in the pathogenesis of Alzheimer disease and some cancers. Numerous inhibitors have been found via library screens, but their structural mechanisms remain unknown. We used several biophysical techniques to investigate the structure of human presenilin complexes and the effects of peptidomimetic γ-secretase inhibitors. The complexes are bilobed. The head contains nicastrin ectodomain. The membrane-embedded base has a central channel and a lateral cleft, which may represent the initial substrate docking site. Inhibitor binding induces widespread structural changes, including rotation of the head and closure of the lateral cleft. These changes block substrate access to the catalytic pocket and inhibit the enzyme. Intriguingly, peptide substrate docking has reciprocal effects on the inhibitor binding site. Similar reciprocal shifts may underlie the mechanisms of other inhibitors and of the “lateral gate” through which substrates access to the catalytic site.

The crystal structure of an archaeal Rce1 protein has been determined; this protein represents a novel type of intramembrane protease, with a distinct architecture and catalytic site. The signalling functions of CAAX proteins, which include Ras and Rho small GTPases, are determined by their correct location at cellular membranes. Membrane localization is controlled by post-translational modifications of their C-terminal CAAX motifs involving cysteine prenylation, endoproteolysis and methylation of the carboxyl-prenylated cysteine residue. Here David Barford and colleagues present the crystal structure of an intramembrane CAAX protease (Ras and a-factor converting enzyme 1 or Rce1) from the archaeon Methanococcus maripaludis. The structure of this novel intramembrane protease reveals a catalytic mechanism distinct from previously described examples such as rhomboids. This work could contribute to the development of antagonists of CAAX motif processing with the potential to disrupt Ras signalling pathways. CAAX proteins have essential roles in multiple signalling pathways, controlling processes such as proliferation, differentiation and carcinogenesis1. The ∼120 mammalian CAAX proteins function at cellular membranes and include the Ras superfamily of small GTPases, nuclear lamins, the γ-subunit of heterotrimeric GTPases, and several protein kinases and phosphatases2. The proper localization of CAAX proteins to cell membranes is orchestrated by a series of post-translational modifications of the carboxy-terminal CAAX motifs3 (where C is cysteine, A is an aliphatic amino acid and X is any amino acid). These reactions involve prenylation of the cysteine residue, cleavage at the AAX tripeptide and methylation of the carboxyl-prenylated cysteine residue. The major CAAX protease activity is mediated by Rce1 (Ras and a-factor converting enzyme 1), an intramembrane protease (IMP) of the endoplasmic reticulum4,5. Information on the architecture and proteolytic mechanism of Rce1 has been lacking. Here we report the crystal structure of a Methanococcus maripaludis homologue of Rce1, whose endopeptidase specificity for farnesylated peptides mimics that of eukaryotic Rce1. Its structure, comprising eight transmembrane α-helices, and catalytic site are distinct from those of other IMPs. The catalytic residues are located ∼10 Å into the membrane and are exposed to the cytoplasm and membrane through a conical cavity that accommodates the prenylated CAAX substrate. We propose that the farnesyl lipid binds to a site at the opening of two transmembrane α-helices, which results in the scissile bond being positioned adjacent to a glutamate-activated nucleophilic water molecule. This study suggests that Rce1 is the founding member of a novel IMP family, the glutamate IMPs.

Background Signal peptide peptidase (SPP), a member of the presenilin-like intra-membrane cleaving aspartyl protease family, migrates on Blue Native (BN) gels as 100 kDa, 200 kDa and 450 kDa species. SPP has recently been implicated in other non-proteolytic functions such as retro-translocation of MHC Class I molecules and binding of misfolded proteins in the endoplasmic reticulum (ER). These high molecular weight SPP complexes might contain additional proteins that regulate the proteolytic activity of SPP or support its non-catalytic functions. Results In this study, an unbiased iTRAQ-labeling mass spectrometry approach was used to identify SPP-interacting proteins. We found that vigilin, a ubiquitous multi-KH domain containing cytoplasmic protein involved in RNA binding and protein translation control, selectively enriched with SPP. Vigilin interacted with SPP and both proteins co-localized in restricted intracellular domains near the ER, biochemically co-fractionated and were part of the same 450 kDa complex on BN gels. However, vigilin does not alter the protease activity of SPP, suggesting that the SPP-vigilin interaction might be involved in the non-proteolytic functions of SPP. Conclusions We have identified and validated vigilin as a novel interacting partner of SPP that could play an important role in the non-proteolytic functions of SPP. This data adds further weight to the idea that intramembrane-cleaving aspartyl proteases, such as presenilin and SPPs, could have other functions besides the proteolysis of short membrane stubs.

The posttranslational modification of C-terminal CAAX motifs in proteins such as Ras, most Rho GTPases, and G protein gamma subunits, plays an essential role in determining their subcellular localization and correct biological function. An integral membrane methyltransferase, isoprenylcysteine carboxyl methyltransferase (ICMT), catalyzes the final step of CAAX processing after prenylation of the cysteine residue and endoproteolysis of the -AAX motif. We have determined the crystal structure of a prokaryotic ICMT ortholog, revealing a markedly different architecture from conventional methyltransferases that utilize S-adenosyl-L-methionine (SAM) as a cofactor. ICMT comprises a core of five transmembrane a helices and a cofactor-binding pocket enclosed within a highly conserved C-terminal catalytic subdomain. A tunnel linking the reactive methyl group of SAM to the inner membrane provides access for the prenyl lipid substrate. This study explains how an integral membrane methyltransferase achieves recognition of both a hydrophilic cofactor and a lipophilic prenyl group attached to a polar protein substrate.