A selection and optimization strategy for single-domain antibodies targeting the PHF6 linear peptide within the tau intrinsically disordered protein.

The use of variable domain of the heavy-chain of the heavy-chain-only antibodies (VHHs) as disease-modifying biomolecules in neurodegenerative disorders holds promises, including targeting of aggregation-sensitive proteins. Exploitation of their clinical values depends however on the capacity to deliver VHHs with optimal physico-chemical properties for their specific context of use. We described previously a VHH with high therapeutic potential in a family of neurodegenerative diseases called tauopathies. The activity of this promising parent VHH named Z70 relies on its binding within the central region of the tau protein. Accordingly, we carried out random mutagenesis followed by yeast two-hybrid screening to obtain optimized variants. The VHHs selected from this initial screen targeted the same epitope as VHH Z70 as shown using NMR spectroscopy and had indeed improved binding affinities according to dissociation constant values obtained by surface plasmon resonance spectroscopy. The improved affinities can be partially rationalized based on three-dimensional structures and NMR data of three complexes consisting of an optimized VHH and a peptide containing the tau epitope. Interestingly, the ability of the VHH variants to inhibit tau aggregation and seeding could not be predicted from their affinity alone. We indeed showed that the in vitro and in cellulo VHH stabilities are other limiting key factors to their efficacy. Our results demonstrate that only a complete pipeline of experiments, here described, permits a rational selection of optimized VHH variants, resulting in the selection of VHH variants with higher affinities and/or acting against tau seeding in cell models.

Development of a V5-tag-directed nanobody and its implementation as an intracellular biosensor of GPCR signaling.

Protein-protein interactions (PPIs) form the foundation of any cell signaling network. Considering that PPIs are highly dynamic processes, cellular assays are often essential for their study because they closely mimic the biological complexities of cellular environments. However, incongruity may be observed across different PPI assays when investigating a protein partner of interest; these discrepancies can be partially attributed to the fusion of different large functional moieties, such as fluorescent proteins or enzymes, which can yield disparate perturbations to the protein's stability, subcellular localization, and interaction partners depending on the given cellular assay. Owing to their smaller size, epitope tags may exhibit a diminished susceptibility to instigate such perturbations. However, while they have been widely used for detecting or manipulating proteins in vitro, epitope tags lack the in vivo traceability and functionality needed for intracellular biosensors. Herein, we develop NbV5, an intracellular nanobody binding the V5-tag, which is suitable for use in cellular assays commonly used to study PPIs such as BRET, NanoBiT, and Tango. The NbV5:V5 tag system has been applied to interrogate G protein-coupled receptor signaling, specifically by replacing larger functional moieties attached to the protein interactors, such as fluorescent or luminescent proteins (∼30 kDa), by the significantly smaller V5-tag peptide (1.4 kDa), and for microscopy imaging which is successfully detected by NbV5-based biosensors. Therefore, the NbV5:V5 tag system presents itself as a versatile tool for live-cell imaging and a befitting adaptation to existing cellular assays dedicated to probing PPIs.

Inhibition of Tau seeding by targeting Tau nucleation core within neurons with a single domain antibody fragment.

Tau proteins aggregate into filaments in brain cells in Alzheimer's disease and related disorders referred to as tauopathies. Here, we used fragments of camelid heavy-chain-only antibodies (VHHs or single domain antibody fragments) targeting Tau as immuno-modulators of its pathologic seeding. A VHH issued from the screen against Tau of a synthetic phage-display library of humanized VHHs was selected for its capacity to bind Tau microtubule-binding domain, composing the core of Tau fibrils. This parent VHH was optimized to improve its biochemical properties and to act in the intra-cellular compartment, resulting in VHH Z70. VHH Z70 precisely binds the PHF6 sequence, known for its nucleation capacity, as shown by the crystal structure of the complex. VHH Z70 was more efficient than the parent VHH to inhibit in vitro Tau aggregation in heparin-induced assays. Expression of VHH Z70 in a cellular model of Tau seeding also decreased the aggregation-reporting fluorescence signal. Finally, intra-cellular expression of VHH Z70 in the brain of an established tauopathy mouse seeding model demonstrated its capacity to mitigate accumulation of pathological Tau. VHH Z70, by targeting Tau inside brain neurons, where most of the pathological Tau resides, provides an immunological tool to target the intra-cellular compartment in tauopathies.

Mammalian splicing factor SF1 interacts with SURP domains of U2 snRNP-associated proteins.

Splicing factor 1 (SF1) recognizes the branch point sequence (BPS) at the 3' splice site during the formation of early complex E, thereby pre-bulging the BPS adenosine, thought to facilitate subsequent base-pairing of the U2 snRNA with the BPS. The 65-kDa subunit of U2 snRNP auxiliary factor (U2AF65) interacts with SF1 and was shown to recruit the U2 snRNP to the spliceosome. Co-immunoprecipitation experiments of SF1-interacting proteins from HeLa cell extracts shown here are consistent with the presence of SF1 in early splicing complexes. Surprisingly almost all U2 snRNP proteins were found associated with SF1. Yeast two-hybrid screens identified two SURP domain-containing U2 snRNP proteins as partners of SF1. A short, evolutionarily conserved region of SF1 interacts with the SURP domains, stressing their role in protein-protein interactions. A reduction of A complex formation in SF1-depleted extracts could be rescued with recombinant SF1 containing the SURP-interaction domain, but only partial rescue was observed with SF1 lacking this sequence. Thus, SF1 can initially recruit the U2 snRNP to the spliceosome during E complex formation, whereas U2AF65 may stabilize the association of the U2 snRNP with the spliceosome at later times. In addition, these findings may have implications for alternative splicing decisions.

Inhibition of autophagy by TAB2 and TAB3.

Autophagic responses are coupled to the activation of the inhibitor of NF-κB kinase (IKK). Here, we report that the essential autophagy mediator Beclin 1 and TGFß-activated kinase 1 (TAK1)-binding proteins 2 and 3 (TAB2 and TAB3), two upstream activators of the TAK1-IKK signalling axis, constitutively interact with each other via their coiled-coil domains (CCDs). Upon autophagy induction, TAB2 and TAB3 dissociate from Beclin 1 and bind TAK1. Moreover, overexpression of TAB2 and TAB3 suppresses, while their depletion triggers, autophagy. The expression of the C-terminal domain of TAB2 or TAB3 or that of the CCD of Beclin 1 competitively disrupts the interaction between endogenous Beclin 1, TAB2 and TAB3, hence stimulating autophagy through a pathway that requires endogenous Beclin 1, TAK1 and IKK to be optimally efficient. These results point to the existence of an autophagy-stimulatory 'switch' whereby TAB2 and TAB3 abandon inhibitory interactions with Beclin 1 to engage in a stimulatory liaison with TAK1.

The antiapoptotic protein AAC-11 interacts with and regulates Acinus-mediated DNA fragmentation.

The nuclear factor Acinus has been suggested to mediate apoptotic chromatin condensation after caspase cleavage. However, this role has been challenged by recent observations suggesting a contribution of Acinus in apoptotic internucleosomal DNA cleavage. We report here that AAC-11, a survival protein whose expression prevents apoptosis that occurs on deprivation of growth factors, physiologically binds to Acinus and prevents Acinus-mediated DNA fragmentation. AAC-11 was able to protect Acinus from caspase-3 cleavage in vivo and in vitro, thus interfering with its biological function. Interestingly, AAC-11 depletion markedly increased cellular sensitivity to anticancer drugs, whereas its expression interfered with drug-induced cell death. AAC-11 possesses a leucine-zipper domain that dictates, upon oligomerization, its interaction with Acinus as well as the antiapoptotic effect of AAC-11 on drug-induced cell death. A cell permeable peptide that mimics the leucine-zipper subdomain of AAC-11, thus preventing its oligomerization, inhibited the AAC-11-Acinus complex formation and potentiated drug-mediated apoptosis in cancer cells. Our results, therefore, show that targeting AAC-11 might be a potent strategy for cancer treatment by sensitization of tumour cells to chemotherapeutic drugs.

Isolation and Mass Spectrometry Identification of K48 and K63 Ubiquitin Proteome Using Chain-Specific Nanobodies.

Protein ubiquitylation is an essential mechanism regulating almost all cellular functions in eukaryotes. The understanding of the role of distinct ubiquitin chains in different cellular processes is essential to identify biomarkers for disease diagnosis and prognosis but also to open new therapeutic possibilities. The high complexity of ubiquitin chains complicates this analysis, and multiple strategies have been developed over the last decades. Here, we report a protocol for the isolation and identification of K48 and K63 ubiquitin chains using chain-specific nanobodies associated to mass spectrometry. Different steps were optimized to increase the purification yield and reduce the binding on nonspecific proteins. The resulting protocol allows the enrichment of ubiquitin chain-specific targets from mammalian cells.

Secretion of VGF relies on the interplay between LRRK2 and post-Golgi v-SNAREs.

The neuropeptide VGF was recently proposed as a neurodegeneration biomarker. The Parkinson's disease-related protein leucine-rich repeat kinase 2 (LRRK2) regulates endolysosomal dynamics, a process that involves SNARE-mediated membrane fusion and could regulate secretion. Here we investigate potential biochemical and functional links between LRRK2 and v-SNAREs. We find that LRRK2 directly interacts with the v-SNAREs VAMP4 and VAMP7. Secretomics reveals VGF secretory defects in VAMP4 and VAMP7 knockout (KO) neuronal cells. In contrast, VAMP2 KO "regulated secretion-null" and ATG5 KO "autophagy-null" cells release more VGF. VGF is partially associated with extracellular vesicles and LAMP1+ endolysosomes. LRRK2 expression increases VGF perinuclear localization and impairs its secretion. Retention using selective hooks (RUSH) assays show that a pool of VGF traffics through VAMP4+ and VAMP7+ compartments, and LRRK2 expression delays its transport to the cell periphery. Overexpression of LRRK2 or VAMP7-longin domain impairs VGF peripheral localization in primary cultured neurons. Altogether, our results suggest that LRRK2 might regulate VGF secretion via interaction with VAMP4 and VAMP7.

Chr21 protein-protein interactions: enrichment in proteins involved in intellectual disability, autism, and late-onset Alzheimer's disease.

Down syndrome (DS) is caused by human chromosome 21 (HSA21) trisomy. It is characterized by a poorly understood intellectual disability (ID). We studied two mouse models of DS, one with an extra copy of the <i>Dyrk1A</i> gene (189N3) and the other with an extra copy of the mouse Chr16 syntenic region (Dp(16)1Yey). RNA-seq analysis of the transcripts deregulated in the embryonic hippocampus revealed an enrichment in genes associated with chromatin for the 189N3 model, and synapses for the Dp(16)1Yey model. A large-scale yeast two-hybrid screen (82 different screens, including 72 HSA21 baits and 10 rebounds) of a human brain library containing at least 10⁷ independent fragments identified 1,949 novel protein-protein interactions. The direct interactors of HSA21 baits and rebounds were significantly enriched in ID-related genes (<i>P</i>-value &lt; 2.29 × 10^-8). Proximity ligation assays showed that some of the proteins encoded by HSA21 were located at the dendritic spine postsynaptic density, in a protein network at the dendritic spine postsynapse. We located HSA21 DYRK1A and DSCAM, mutations of which increase the risk of autism spectrum disorder (ASD) 20-fold, in this postsynaptic network. We found that an intracellular domain of DSCAM bound either DLGs, which are multimeric scaffolds comprising receptors, ion channels and associated signaling proteins, or DYRK1A. The DYRK1A-DSCAM interaction domain is conserved in <i>Drosophila</i> and humans. The postsynaptic network was found to be enriched in proteins associated with ARC-related synaptic plasticity, ASD, and late-onset Alzheimer's disease. These results highlight links between DS and brain diseases with a complex genetic basis.

Transportin-SR2 imports HIV into the nucleus.

BACKGROUND: The human immunodeficiency virus type 1 (HIV-1) and other lentiviruses have the capacity to infect nondividing cells like macrophages. This requires import of the preintegration complex (PIC) through the nuclear pore. Although many cellular and viral determinants have been proposed, the mechanism leading to nuclear import is not yet understood. RESULTS: Using yeast two-hybrid and pull-down, we identified and validated transportin-SR2 (TRN-SR2) as a bona fide binding partner of HIV-1 integrase. We confirmed the biological relevance of this interaction by RNAi. Depletion of TRN-SR2 interfered with the replication of HIV-1 and HIV-2 but not MoMLV in HeLaP4 cells. Knockdown of TRN-SR2 in primary macrophages likewise interfered with HIV-1 replication. Using Q-PCR, we pinpoint this block in replication to the early steps of the viral lifecycle. A reduction in 2-LTR formation suggests a block in PIC nuclear import upon siRNA-mediated knockdown. Different lines of evidence clearly proved that the late steps of viral replication are not affected. In an in vivo nuclear-import assay using labeled HIV-1 particles, the defect in nuclear import after depletion of TRN-SR2 was directly visualized. In comparison with control cell lines, the great majority of siRNA-treated cells did not contain any PIC in the nucleus. CONCLUSION: Our data clearly demonstrate that TRN-SR2 is the nuclear-import factor of HIV.

Physical and genetic interactions of yeast Cwc21p, an ortholog of human SRm300/SRRM2, suggest a role at the catalytic center of the spliceosome.

In Saccharomyces cerevisiae, Cwc21p is a protein of unknown function that is associated with the NineTeen Complex (NTC), a group of proteins involved in activating the spliceosome to promote the pre-mRNA splicing reaction. Here, we show that Cwc21p binds directly to two key splicing factors-namely, Prp8p and Snu114p-and becomes the first NTC-related protein known to dock directly to U5 snRNP proteins. Using a combination of proteomic techniques we show that the N-terminus of Prp8p contains an intramolecular fold that is a Snu114p and Cwc21p interacting domain (SCwid). Cwc21p also binds directly to the C-terminus of Snu114p. Complementary chemical cross-linking experiments reveal reciprocal protein footprints between the interacting Prp8 and Cwc21 proteins, identifying the conserved cwf21 domain in Cwc21p as a Prp8p binding site. Genetic and functional interactions between Cwc21p and Isy1p indicate that they have related functions at or prior to the first catalytic step of splicing, and suggest that Cwc21p functions at the catalytic center of the spliceosome, possibly in response to environmental or metabolic changes. We demonstrate that SRm300, the only SR-related protein known to be at the core of human catalytic spliceosomes, is a functional ortholog of Cwc21p, also interacting directly with Prp8p and Snu114p. Thus, the function of Cwc21p is likely conserved from yeast to humans.

A functional network involved in the recycling of nucleocytoplasmic pre-60S factors.

Eukaryotic pre-ribosomes go through cytoplasmic maturation steps before entering translation. The nucleocytoplasmic proteins participating in these late stages of maturation are reimported to the nucleus. In this study, we describe a functional network focused on Rei1/Ybr267w, a strictly cytoplasmic pre-60S factor indirectly involved in nuclear 27S pre-ribosomal RNA processing. In the absence of Rei1, the nuclear import of at least three other pre-60S factors is impaired. The accumulation in the cytoplasm of a small complex formed by the association of Arx1 with a novel factor, Alb1/Yjl122w, inhibits the release of the putative antiassociation factor Tif6 from the premature large ribosomal subunits and its recycling to the nucleus. We propose a model in which Rei1 is a key factor for the coordinated dissociation and recycling of the last pre-60S factors before newly synthesized large ribosomal subunits enter translation.

Characterization of single chain antibody targets through yeast two hybrid.

BACKGROUND: Due to their unique ability to bind their targets with high fidelity, antibodies are used widely not only in biomedical research, but also in many clinical applications. Recombinant antibodies, including single chain variable fragments (scFv), are gaining momentum because they allow powerful in vitro selection and manipulation without loss of function. Regardless of the ultimate application or type of antibody used, precise understanding of the interaction between the antibody's binding site and its specific target epitope(s) is of great importance. However, such data is frequently difficult to obtain. RESULTS: We describe an approach that allows detailed characterization of a given antibody's target(s) using the yeast two-hybrid system. Several recombinant scFv were used as bait and screened against highly complex cDNA libraries. Systematic sequencing of all retained clones and statistical analysis allowed efficient ranking of the prey fragments. Multiple alignment of the obtained cDNA fragments provided a selected interacting domain (SID), efficiently narrowing the epitope-containing region.Interactions between antibodies and their respective targets were characterized for several scFv. For AA2 and ROF7, two conformation-specific sensors that exclusively bind the activated forms of the small GTPases Rab6 and Rab1 respectively, only fragments expressing the entire target protein's core region were retained. This strongly suggested interaction with a non-linear epitope. For two other scFv, TA10 and SF9, which recognize the large proteins giantin and non-muscle myosin IIA, respectively, precise antibody-binding regions within the target were defined. Finally, for some antibodies, secondary targets within and across species could be revealed. CONCLUSIONS: Our method, utilizing the yeast two-hybrid technology and scFv as bait, is a simple yet powerful approach for the detailed characterization of antibody targets. It allows precise domain mapping for linear epitopes, confirmation of non-linear epitopes for conformational sensors, and detection of secondary binding partners. This approach may thus prove to be an elegant and rapid method for the target characterization of newly obtained scFv antibodies. It may be considered prior to any research application and particularly before any use of such recombinant antibodies in clinical medicine.

Yeast two-hybrid detection of integrase-host factor interactions.

Here we describe methods developed based on systematic yeast two-hybrid screenings that allowed us to identify several binding partners of HIV-1 integrase. We have developed an efficient strategy to perform large comprehensive screenings with different highly complex cDNA libraries derived both random- and oligo-dT primed reactions. A very efficient mating procedure was used for screening in yeast, allowing genetic saturation of positive clones. This importantly leads with confidence to the determination of the regions within the participating proteins responsible for the interactions. Several additional tools were used that allowed us to assess the specificity of the interactions detected, including rebound screens with cellular co-factors as baits performed against a library of random fragments of HIV-1 proviral DNA. For some of the identified cell factors, we have generated and characterized loss of affinity mutants of integrase, which, when combined with viral functional assays, validated the involvement of human lens epithelium-derived growth factor (LEDGF/p75) in the integration step of the HIV-1 replication cycle. All tolled, our studies identified LEDGF/p75, Transportin-SR2 (TNPO3), von Hippel-Lindau binding protein 1 (VBP1), and sucrose non-fermenting 5 (SNF5) as cellular binding partners of HIV-1 integrase.

Filaggrin and filaggrin 2 processing are linked together through skin aspartic acid protease activation.

Skin aspartic acid protease (SASPase) is believed to be a key enzyme involved in filaggrin processing during epidermal terminal differentiation. Since little is known about the regulation of SASPase function, the aim of this study was to identify involved protein partners in the process. Yeast two hybrid analyses using SASPase as bait against a human reconstructed skin library identified that the N-terminal domain of filaggrin 2 binds to the N-terminal fragment of SASPase. This interaction was confirmed in reciprocal yeast two hybrid screens and by Surface Plasmon Resonance analyses. Immunohistochemical studies in human skin, using specific antibodies to SASPase and the N-terminal domain of filaggrin 2, showed that the two proteins partially co-localized to the stratum granulosum. In vitro enzymatic assays showed that the N-terminal domain of filaggrin 2 enhanced the autoactivation of SASPase to its 14 kDa active form. Taken together, the data suggest that the N-terminal domain of filaggrin 2 regulates the activation of SASPase that may be a key event upstream of filaggrin processing to natural moisturizing factors in the human epidermis.

Fr-PPIChem: An Academic Compound Library Dedicated to Protein-Protein Interactions.

Protein-protein interactions (PPIs) mediate nearly every cellular process and represent attractive targets for modulating disease states but are challenging to target with small molecules. Despite this, several PPI inhibitors (iPPIs) have entered clinical trials, and a growing number of PPIs have become validated drug targets. However, high-throughput screening efforts still endure low hit rates mainly because of the use of unsuitable screening libraries. Here, we describe the collective effort of a French consortium to build, select, and store in plates a unique chemical library dedicated to the inhibition of PPIs. Using two independent predictive models and two updated databases of experimentally confirmed PPI inhibitors developed by members of the consortium, we built models based on different training sets, molecular descriptors, and machine learning methods. Independent statistical models were used to select putative PPI inhibitors from large commercial compound collections showing great complementarity. Medicinal chemistry filters were applied to remove undesirable structures from this set (such as PAINS, frequent hitters, and toxic compounds) and to improve drug likeness. The remaining compounds were subjected to a clustering procedure to reduce the final size of the library while maintaining its chemical diversity. In practice, the library showed a 46-fold activity rate enhancement when compared to a non-iPPI-enriched diversity library in high-throughput screening against the CD47-SIRPα PPI. The Fr-PPIChem library is plated in 384-well plates and will be distributed on demand to the scientific community as a powerful tool for discovering new chemical probes and early hits for the development of potential therapeutic drugs.

Single Domain Antibody Fragments as New Tools for the Detection of Neuronal Tau Protein in Cells and in Mice Studies.

Tau is a neuronal protein linked to pathologies called tauopathies, including Alzheimer's disease. In Alzheimer's disease, tau aggregates into filaments, leading to the observation of intraneuronal fibrillary tangles. Molecular mechanisms resulting in tau aggregation and in tau pathology spreading through the brain regions are still not fully understood. New tools are thus needed to decipher tau pathways involved in the diseases. In this context, a family of novel single domain antibody fragments, or VHHs, directed against tau were generated and characterized. Among the selected VHHs obtained from screening of a synthetic library, a family of six VHHs shared the same CDR3 recognition loop and recognized the same epitope, located in the C-terminal domain of tau. Affinity parameters characterizing the tau/VHHs interaction were next evaluated using surface plasmon resonance spectroscopy. The equilibrium constants KD were in the micromolar range, but despite conservation of the CDR3 loop sequence, a range of affinities was observed for this VHH family. One of these VHHs, named F8-2, was additionally shown to bind tau upon expression in a neuronal cell line model. Optimization of VHH F8-2 by yeast two-hybrid allowed the generation of an optimized VHH family characterized by lower KD than that of the F8-2 wild-type counterpart, and recognizing the same epitope. The optimized VHHs can also be used as antibodies for detecting tau in transgenic mice brain tissues. These results validate the use of these VHHs for in vitro studies, but also their potential for in-cell expression and assays in mouse models, to explore the mechanisms underlying tau physiopathology.

LPS-Induced Inflammation Abolishes the Effect of DYRK1A on IkB Stability in the Brain of Mice.

Down syndrome is characterized by premature aging and dementia with neurological features that mimic those found in Alzheimer's disease. This pathology in Down syndrome could be related to inflammation, which plays a role in other neurodegenerative diseases. We previously found a link between the NFkB pathway, long considered a prototypical proinflammatory signaling pathway, and the dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A). DYRK1A is associated with early onset of Alzheimer's disease in Down syndrome patients. Here, we sought to determine the role of DYRK1A on regulation of the NFkB pathway in the mouse brain. We found that over-expression of Dyrk1A (on a C57BL/6J background) stabilizes IκBα protein levels by inhibition of calpain activity and increases cytoplasmic p65 sequestration in the mouse brain. In contrast, Dyrk1A-deficient mice (on a CD1 background) have decreased IκBα protein levels with an increased calpain activity and decreased cytoplasmic p65 sequestration in the brain. Taken together, our results demonstrate a role of DYRK1A in regulation of the NFkB pathway. However, decreased IκBα and DYRK1A protein levels associated with an increased calpain activity were found in the brains of mice over-expressing Dyrk1A after lipopolysaccharide treatment. Although inflammation induced by lipopolysaccharide treatment has a positive effect on calpastatin and a negative effect on DYRK1A protein level, a positive effect on microglial activation is maintained in the brains of mice over-expressing Dyrk1A.

Downregulated Wnt/ß-catenin signalling in the Down syndrome hippocampus.

Pathological mechanisms underlying Down syndrome (DS)/Trisomy 21, including dysregulation of essential signalling processes remain poorly understood. Combining bioinformatics with RNA and protein analysis, we identified downregulation of the Wnt/ß-catenin pathway in the hippocampus of adult DS individuals with Alzheimer's disease and the 'Tc1' DS mouse model. Providing a potential underlying molecular pathway, we demonstrate that the chromosome 21 kinase DYRK1A regulates Wnt signalling via a novel bimodal mechanism. Under basal conditions, DYRK1A is a negative regulator of Wnt/ß-catenin. Following pathway activation, however, DYRK1A exerts the opposite effect, increasing signalling activity. In summary, we identified downregulation of hippocampal Wnt/ß-catenin signalling in DS, possibly mediated by a dose dependent effect of the chromosome 21-encoded kinase DYRK1A. Overall, we propose that dosage imbalance of the Hsa21 gene DYRK1A affects downstream Wnt target genes. Therefore, modulation of Wnt signalling may open unexplored avenues for DS and Alzheimer's disease treatment.

HENA, heterogeneous network-based data set for Alzheimer's disease.

Alzheimer's disease and other types of dementia are the top cause for disabilities in later life and various types of experiments have been performed to understand the underlying mechanisms of the disease with the aim of coming up with potential drug targets. These experiments have been carried out by scientists working in different domains such as proteomics, molecular biology, clinical diagnostics and genomics. The results of such experiments are stored in the databases designed for collecting data of similar types. However, in order to get a systematic view of the disease from these independent but complementary data sets, it is necessary to combine them. In this study we describe a heterogeneous network-based data set for Alzheimer's disease (HENA). Additionally, we demonstrate the application of state-of-the-art graph convolutional networks, i.e. deep learning methods for the analysis of such large heterogeneous biological data sets. We expect HENA to allow scientists to explore and analyze their own results in the broader context of Alzheimer's disease research.