A dimer-monomer switch controls CHIP-dependent substrate ubiquitylation and processing.

The high substrate selectivity of the ubiquitin/proteasome system is mediated by a large group of E3 ubiquitin ligases. The ubiquitin ligase CHIP regulates the degradation of chaperone-controlled and chaperone-independent proteins. To understand how CHIP mediates substrate selection and processing, we performed a structure-function analysis of CHIP and addressed its physiological role in Caenorhabditis elegans and human cells. The conserved function of CHIP in chaperone-assisted degradation requires dimer formation to mediate proteotoxic stress resistance and to prevent protein aggregation. The CHIP monomer, however, promotes the turnover of the membrane-bound insulin receptor and longevity. The dimer-monomer transition is regulated by CHIP autoubiquitylation and chaperone binding, which provides a feedback loop that controls CHIP activity in response to cellular stress. Because CHIP also binds other E3 ligases, such as Parkin, the molecular switch mechanism described here could be a general concept for the regulation of substrate selectivity and ubiquitylation by combining different E3s.

A heterotypic assembly mechanism regulates CHIP E3 ligase activity.

CHIP (C-terminus of Hsc70-interacting protein) and its worm ortholog CHN-1 are E3 ubiquitin ligases that link the chaperone system with the ubiquitin-proteasome system (UPS). CHN-1 can cooperate with UFD-2, another E3 ligase, to accelerate ubiquitin chain formation; however, the basis for the high processivity of this E3s set has remained obscure. Here, we studied the molecular mechanism and function of the CHN-1-UFD-2 complex in Caenorhabditis elegans. Our data show that UFD-2 binding promotes the cooperation between CHN-1 and ubiquitin-conjugating E2 enzymes by stabilizing the CHN-1 U-box dimer. However, HSP70/HSP-1 chaperone outcompetes UFD-2 for CHN-1 binding, thereby promoting a shift to the autoinhibited CHN-1 state by acting on a conserved residue in its U-box domain. The interaction with UFD-2 enables CHN-1 to efficiently ubiquitylate and regulate S-adenosylhomocysteinase (AHCY-1), a key enzyme in the S-adenosylmethionine (SAM) regeneration cycle, which is essential for SAM-dependent methylation. Our results define the molecular mechanism underlying the synergistic cooperation of CHN-1 and UFD-2 in substrate ubiquitylation.

A peptidomimetic modulator of the CaV2.2 N-type calcium channel for chronic pain.

Transmembrane Cav2.2 (N-type) voltage-gated calcium channels are genetically and pharmacologically validated, clinically relevant pain targets. Clinical block of Cav2.2 (e.g., with Prialt/Ziconotide) or indirect modulation [e.g., with gabapentinoids such as Gabapentin (GBP)] mitigates chronic pain but is encumbered by side effects and abuse liability. The cytosolic auxiliary subunit collapsin response mediator protein 2 (CRMP2) targets Cav2.2 to the sensory neuron membrane and regulates their function via an intrinsically disordered motif. A CRMP2-derived peptide (CBD3) uncouples the Cav2.2-CRMP2 interaction to inhibit calcium influx, transmitter release, and pain. We developed and applied a molecular dynamics approach to identify the A1R2 dipeptide in CBD3 as the anchoring Cav2.2 motif and designed pharmacophore models to screen 27 million compounds on the open-access server ZincPharmer. Of 200 curated hits, 77 compounds were assessed using depolarization-evoked calcium influx in rat dorsal root ganglion neurons. Nine small molecules were tested electrophysiologically, while one (CBD3063) was also evaluated biochemically and behaviorally. CBD3063 uncoupled Cav2.2 from CRMP2, reduced membrane Cav2.2 expression and Ca2+ currents, decreased neurotransmission, reduced fiber photometry-based calcium responses in response to mechanical stimulation, and reversed neuropathic and inflammatory pain across sexes in two different species without changes in sensory, sedative, depressive, and cognitive behaviors. CBD3063 is a selective, first-in-class, CRMP2-based peptidomimetic small molecule, which allosterically regulates Cav2.2 to achieve analgesia and pain relief without negative side effect profiles. In summary, CBD3063 could potentially be a more effective alternative to GBP for pain relief.

Defining the Kv2.1-syntaxin molecular interaction identifies a first-in-class small molecule neuroprotectant.

The neuronal cell death-promoting loss of cytoplasmic K+ following injury is mediated by an increase in Kv2.1 potassium channels in the plasma membrane. This phenomenon relies on Kv2.1 binding to syntaxin 1A via 9 amino acids within the channel intrinsically disordered C terminus. Preventing this interaction with a cell and blood-brain barrier-permeant peptide is neuroprotective in an in vivo stroke model. Here a rational approach was applied to define the key molecular interactions between syntaxin and Kv2.1, some of which are shared with mammalian uncoordinated-18 (munc18). Armed with this information, we found a small molecule Kv2.1-syntaxin-binding inhibitor (cpd5) that improves cortical neuron survival by suppressing SNARE-dependent enhancement of Kv2.1-mediated currents following excitotoxic injury. We validated that cpd5 selectively displaces Kv2.1-syntaxin-binding peptides from syntaxin and, at higher concentrations, munc18, but without affecting either synaptic or neuronal intrinsic properties in brain tissue slices at neuroprotective concentrations. Collectively, our findings provide insight into the role of syntaxin in neuronal cell death and validate an important target for neuroprotection.

Electrophilic fatty acids impair RAD51 function and potentiate the effects of DNA-damaging agents on growth of triple-negative breast cells.

Homologous recombination (HR)-directed DNA double-strand break (DSB) repair enables template-directed DNA repair to maintain genomic stability. RAD51 recombinase (RAD51) is a critical component of HR and facilitates DNA strand exchange in DSB repair. We report here that treating triple-negative breast cancer (TNBC) cells with the fatty acid nitroalkene 10-nitro-octadec-9-enoic acid (OA-NO2) in combination with the antineoplastic DNA-damaging agents doxorubicin, cisplatin, olaparib, and γ-irradiation (IR) enhances the antiproliferative effects of these agents. OA-NO2 inhibited IR-induced RAD51 foci formation and enhanced H2A histone family member X (H2AX) phosphorylation in TNBC cells. Analyses of fluorescent DSB reporter activity with both static-flow cytometry and kinetic live-cell studies enabling temporal resolution of recombination revealed that OA-NO2 inhibits HR and not nonhomologous end joining (NHEJ). OA-NO2 alkylated Cys-319 in RAD51, and this alkylation depended on the Michael acceptor properties of OA-NO2 because nonnitrated and saturated nonelectrophilic analogs of OA-NO2, octadecanoic acid and 10-nitro-octadecanoic acid, did not react with Cys-319. Of note, OA-NO2 alkylation of RAD51 inhibited its binding to ssDNA. RAD51 Cys-319 resides within the SH3-binding site of ABL proto-oncogene 1, nonreceptor tyrosine kinase (ABL1), so we investigated the effect of OA-NO2-mediated Cys-319 alkylation on ABL1 binding and found that OA-NO2 inhibits RAD51-ABL1 complex formation both in vitro and in cell-based immunoprecipitation assays. The inhibition of the RAD51-ABL1 complex also suppressed downstream RAD51 Tyr-315 phosphorylation. In conclusion, RAD51 Cys-319 is a functionally significant site for adduction of soft electrophiles such as OA-NO2 and suggests further investigation of lipid electrophile-based combinational therapies for TNBC.

A combined computational and experimental approach reveals the structure of a C/EBPß-Spi1 interaction required for IL1B gene transcription.

We previously reported that transcription of the human IL1B gene, encoding the proinflammatory cytokine interleukin 1ß, depends on long-distance chromatin looping that is stabilized by a mutual interaction between the DNA-binding domains (DBDs) of two transcription factors: Spi1 proto-oncogene at the promoter and CCAAT enhancer-binding protein (C/EBPß) at a far-upstream enhancer. We have also reported that the C-terminal tail sequence beyond the C/EBPß leucine zipper is critical for its association with Spi1 via an exposed residue (Arg-232) located within a pocket at one end of the Spi1 DNA-recognition helix. Here, combining in vitro interaction studies with computational docking and molecular dynamics of existing X-ray structures for the Spi1 and C/EBPß DBDs, along with the C/EBPß C-terminal tail sequence, we found that the tail sequence is intimately associated with Arg-232 of Spi1. The Arg-232 pocket was computationally screened for small-molecule binding aimed at IL1B transcription inhibition, yielding l-arginine, a known anti-inflammatory amino acid, revealing a potential for disrupting the C/EBPß-Spi1 interaction. As evaluated by ChIP, cultured lipopolysaccharide (LPS)-activated THP-1 cells incubated with l-arginine had significantly decreased IL1B transcription and reduced C/EBPß's association with Spi1 on the IL1B promoter. No significant change was observed in direct binding of either Spi1 or C/EBPß to cognate DNA and in transcription of the C/EBPß-dependent IL6 gene in the same cells. These results support the notion that disordered sequences extending from a leucine zipper can mediate protein-protein interactions and can serve as druggable targets for regulating gene promoter activity.

Predicting protein targets for drug-like compounds using transcriptomics.

An expanded chemical space is essential for improved identification of small molecules for emerging therapeutic targets. However, the identification of targets for novel compounds is biased towards the synthesis of known scaffolds that bind familiar protein families, limiting the exploration of chemical space. To change this paradigm, we validated a new pipeline that identifies small molecule-protein interactions and works even for compounds lacking similarity to known drugs. Based on differential mRNA profiles in multiple cell types exposed to drugs and in which gene knockdowns (KD) were conducted, we showed that drugs induce gene regulatory networks that correlate with those produced after silencing protein-coding genes. Next, we applied supervised machine learning to exploit drug-KD signature correlations and enriched our predictions using an orthogonal structure-based screen. As a proof-of-principle for this regimen, top-10/top-100 target prediction accuracies of 26% and 41%, respectively, were achieved on a validation of set 152 FDA-approved drugs and 3104 potential targets. We then predicted targets for 1680 compounds and validated chemical interactors with four targets that have proven difficult to chemically modulate, including non-covalent inhibitors of HRAS and KRAS. Importantly, drug-target interactions manifest as gene expression correlations between drug treatment and both target gene KD and KD of genes that act up- or down-stream of the target, even for relatively weak binders. These correlations provide new insights on the cellular response of disrupting protein interactions and highlight the complex genetic phenotypes of drug treatment. With further refinement, our pipeline may accelerate the identification and development of novel chemical classes by screening compound-target interactions.

Optimal affinity ranking for automated virtual screening validated in prospective D3R grand challenges.

The goal of virtual screening is to generate a substantially reduced and enriched subset of compounds from a large virtual chemistry space. Critical in these efforts are methods to properly rank the binding affinity of compounds. Prospective evaluations of ranking strategies in the D3R grand challenges show that for targets with deep pockets the best correlations (Spearman ρ ~ 0.5) were obtained by our submissions that docked compounds to the holo-receptors with the most chemically similar ligand. On the other hand, for targets with open pockets using multiple receptor structures is not a good strategy. Instead, docking to a single optimal receptor led to the best correlations (Spearman ρ ~ 0.5), and overall performs better than any other method. Yet, choosing a suboptimal receptor for crossdocking can significantly undermine the affinity rankings. Our submissions that evaluated the free energy of congeneric compounds were also among the best in the community experiment. Error bars of around 1 kcal/mol are still too large to significantly improve the overall rankings. Collectively, our top of the line predictions show that automated virtual screening with rigid receptors perform better than flexible docking and other more complex methods.

Extensive Citrullination Promotes Immunogenicity of HSP90 through Protein Unfolding and Exposure of Cryptic Epitopes.

Post-translational protein modifications such as citrullination have been linked to the breach of immune tolerance and clinical autoimmunity. Previous studies from our laboratory support this concept, demonstrating that autoantibodies targeting citrullinated isoforms of heat shock protein 90 (HSP90) are associated with rheumatoid arthritis complicated by interstitial lung disease. To further explore the relationship between citrullination and structural determinants of HSP90 immunogenicity, we employed a combination of ELISA-based epitope profiling, computational modeling, and mass-spectrometric sequencing of peptidylarginine deiminase (PAD)-modified protein. Remarkably, ELISAs involving selected citrullinated HSP90ß/α peptides identified a key epitope corresponding to an internal Arg residue (R502 [HSP90ß]/R510 [HSP90α]) that is normally buried within the crystal structure of native/unmodified HSP90. In vitro time/dose-response experiments reveal an ordered pattern of PAD-mediated deimination events culminating in citrullination of R502/R510. Conventional as well as scaled molecular dynamics simulations further demonstrate that citrullination of selected Arg residues leads to progressive disruption of HSP90 tertiary structure, promoting exposure of R502/R510 to PAD modification and subsequent autoantibody binding. Consistent with this process, ELISAs incorporating variably deiminated HSP90 as substrate Ag indicate a direct relationship between the degree of citrullination and the level of ex vivo Ab recognition. Overall, these data support a novel structural paradigm whereby citrullination-induced shifts in protein structure generate cryptic epitopes capable of bypassing B cell tolerance in the appropriate genetic context.

Peroxidase activation of cytoglobin by anionic phospholipids: Mechanisms and consequences.

Cytoglobin (Cygb) is a hexa-coordinated hemoprotein with yet to be defined physiological functions. The iron coordination and spin state of the Cygb heme group are sensitive to oxidation of two cysteine residues (Cys38/Cys83) and/or the binding of free fatty acids. However, the roles of redox vs lipid regulators of Cygb's structural rearrangements in the context of the protein peroxidase competence are not known. Searching for physiologically relevant lipid regulators of Cygb, here we report that anionic phospholipids, particularly phosphatidylinositolphosphates, affect structural organization of the protein and modulate its iron state and peroxidase activity both conjointly and/or independently of cysteine oxidation. Thus, different anionic lipids can operate in cysteine-dependent and cysteine-independent ways as inducers of the peroxidase activity. We establish that Cygb's peroxidase activity can be utilized for the catalysis of peroxidation of anionic phospholipids (including phosphatidylinositolphosphates) yielding mono-oxygenated molecular species. Combined with the computational simulations we propose a bipartite lipid binding model that rationalizes the modes of interactions with phospholipids, the effects on structural re-arrangements and the peroxidase activity of the hemoprotein.

Structure Guided Chemical Modifications of Propylthiouracil Reveal Novel Small Molecule Inhibitors of Cytochrome b5 Reductase 3 That Increase Nitric Oxide Bioavailability.

NADH cytochrome b5 reductase 3 (CYB5R3) is critical for reductive reactions such as fatty acid elongation, cholesterol biosynthesis, drug metabolism, and methemoglobin reduction. Although the physiological and metabolic importance of CYB5R3 has been established in hepatocytes and erythrocytes, emerging investigations suggest that CYB5R3 is critical for nitric oxide signaling and vascular function. However, advancement toward fully understanding CYB5R3 function has been limited due to a lack of potent small molecule inhibitors. Because of this restriction, we modeled the binding mode of propylthiouracil, a weak inhibitor of CYB5R3 (IC50 = ∼275 µM), and used it as a guide to predict thiouracil-biased inhibitors from the set of commercially available compounds in the ZINC database. Using this approach, we validated two new potent derivatives of propylthiouracil, ZINC05626394 (IC50 = 10.81 µM) and ZINC39395747 (IC50 = 9.14 µM), both of which inhibit CYB5R3 activity in cultured cells. Moreover, we found that ZINC39395747 significantly increased NO bioavailability in renal vascular cells, augmented renal blood flow, and decreased systemic blood pressure in response to vasoconstrictors in spontaneously hypertensive rats. These compounds will serve as a new tool to examine the biological functions of CYB5R3 in physiology and disease and also as a platform for new drug development.

Restricted protein phosphatase 2A targeting by Merkel cell polyomavirus small T antigen.

UNLABELLED: Merkel cell polyomavirus (MCV) is a newly discovered human cancer virus encoding a small T (sT) oncoprotein. We performed MCV sT FLAG-affinity purification followed by mass spectroscopy (MS) analysis, which identified several protein phosphatases (PP), including PP2A A and C subunits and PP4C, as potential cellular interacting proteins. PP2A targeting is critical for the transforming properties of nonhuman polyomaviruses, such as simian virus 40 (SV40), but is not required for MCV sT-induced rodent cell transformation. We compared similarities and differences in PP2A binding between MCV and SV40 sT. While SV40 sT coimmunopurified with subunits PP2A Aα and PP2A C, MCV sT coimmunopurified with PP2A Aα, PP2A Aß, and PP2A C. Scanning alanine mutagenesis at 29 sites across the MCV sT protein revealed that PP2A-binding domains lie on the opposite molecular surface from a previously described large T stabilization domain (LSD) loop that binds E3 ligases, such as Fbw7. MCV sT-PP2A interactions can be functionally distinguished by mutagenesis from MCV sT LSD-dependent 4E-BP1 hyperphosphorylation and viral DNA replication enhancement. MCV sT has a restricted range for PP2A B subunit substitution, inhibiting only the assembly of B56α into the phosphatase holoenzyme. In contrast, SV40 sT inhibits the assembly of B55α, B56α and B56ε into PP2A. We conclude that MCV sT is required for Merkel cell carcinoma growth, but its in vitro transforming activity depends on LSD interactions rather than PP2A targeting. IMPORTANCE: Merkel cell polyomavirus is a newly discovered human cancer virus that promotes cancer, in part, through expression of its small T (sT) oncoprotein. Animal polyomavirus sT oncoproteins have been found to cause experimental tumors by blocking the activities of a group of phosphatases called protein phosphatase 2A (PP2A). Our structural analysis reveals that MCV sT also displaces the B subunit of PP2A to inhibit PP2A activity. MCV sT, however, only displaces a restricted subset of PP2A B subunits, which is insufficient to cause tumor cell formation in vitro. MCV sT instead transforms tumor cells through another region called the large T stabilization domain. The PP2A targeting and transforming activities lie on opposite faces of the MCV sT molecule and can be genetically separated from each other.

Choosing the Optimal Rigid Receptor for Docking and Scoring in the CSAR 2013/2014 Experiment.

The 2013/2014 Community Structure-Activity Resource (CSAR) challenge was designed to prospectively validate advancement in the field of docking and scoring receptor-small molecule interactions. Purely computational methods have been found to be quite limiting. Thus, the challenges assessed methods that combined both experimental data and computational approaches. Here, we describe our contribution to solve three important challenges in rational drug discovery: rank-ordering protein primary sequences based on affinity to a compound, determining close-to-native bound conformations out of a set of decoy poses, and rank-ordering sets of congeneric compounds based on affinity to a given protein. We showed that the most significant contribution to a meaningful enrichment of native-like models was the identification of the best receptor structure for docking and scoring. Depending on the target, the optimal receptor for cross-docking and scoring was identified by a self-consistent docking approach that used the Vina scoring function, by aligning compounds to the closest cocrystal or by selecting the cocrystal receptor with the largest pocket. For tRNA (m1G37) methyltransferase (TRMD), ranking a set of 31 congeneric binding compounds cross-docked to the optimal receptor resulted in a R(2) = 0.67; whereas, using any other of the 13 receptor structures led to almost no enrichment of native-like complex structures. Furthermore, although redocking predicted lower RMSDs relative to the bound structures, the ranking based on multiple receptor structures did not improve the correlation coefficient. Our predictions highlight the role of rational structure-based modeling in maximizing the outcome of virtual screening, as well as limitations scoring multiple receptors.

Optimal strategies for virtual screening of induced-fit and flexible target in the 2015 D3R Grand Challenge.

Induced fit or protein flexibility can make a given structure less useful for docking and/or scoring. The 2015 Drug Design Data Resource (D3R) Grand Challenge provided a unique opportunity to prospectively test optimal strategies for virtual screening in these type of targets: heat shock protein 90 (HSP90), a protein with multiple ligand-induced binding modes; and mitogen-activated protein kinase kinase kinase kinase 4 (MAP4K4), a kinase with a large flexible pocket. Using previously known co-crystal structures, we tested predictions from methods that keep the receptor structure fixed and used (a) multiple receptor/ligand co-crystals as binding templates for minimization or docking ("close"), (b) methods that align or dock to a single receptor ("cross"), and (c) a hybrid approach that chose from multiple bound ligands as initial templates for minimization to a single receptor ("min-cross"). Pose prediction using our "close" models resulted in average ligand RMSDs of 0.32 and 1.6 Å for HSP90 and MAP4K4, respectively, the most accurate models of the community-wide challenge. On the other hand, affinity ranking using our "cross" methods performed well overall despite the fact that a fixed receptor cannot model ligand-induced structural changes,. In addition, "close" methods that leverage the co-crystals of the different binding modes of HSP90 also predicted the best affinity ranking. Our studies suggest that analysis of changes on the receptor structure upon ligand binding can help select an optimal virtual screening strategy.

Indexing Volumetric Shapes with Matching and Packing.

We describe a novel algorithm for bulk-loading an index with high-dimensional data and apply it to the problem of volumetric shape matching. Our matching and packing algorithm is a general approach for packing data according to a similarity metric. First an approximate k-nearest neighbor graph is constructed using vantage-point initialization, an improvement to previous work that decreases construction time while improving the quality of approximation. Then graph matching is iteratively performed to pack related items closely together. The end result is a dense index with good performance. We define a new query specification for shape matching that uses minimum and maximum shape constraints to explicitly specify the spatial requirements of the desired shape. This specification provides a natural language for performing volumetric shape matching and is readily supported by the geometry-based similarity search (GSS) tree, an indexing structure that maintains explicit representations of volumetric shape. We describe our implementation of a GSS tree for volumetric shape matching and provide a comprehensive evaluation of parameter sensitivity, performance, and scalability. Compared to previous bulk-loading algorithms, we find that matching and packing can construct a GSS-tree index in the same amount of time that is denser, flatter, and better performing, with an observed average performance improvement of 2X.

Discovery of a Potent Allosteric Kinase Modulator by Combining Computational and Synthetic Methods.

The rational design of allosteric kinase modulators is challenging but rewarding. The protein kinase PDK1, which lies at the center of the growth-factor signaling pathway, possesses an allosteric regulatory site previously validated both in vitro and in cells. ANCHOR.QUERY software was used to discover a potent allosteric PDK1 kinase modulator. Using a recently published PDK1 compound as a template, several new scaffolds that bind to the allosteric target site were generated and one example was validated. The inhibitor can be synthesized in one step by multicomponent reaction (MCR) chemistry when using the ANCHOR.QUERY approach. Our results are significant because the outlined approach allows rapid and efficient scaffold hopping from known molecules into new easily accessible and biologically active ones. Based on increasing interest in allosteric-site drug discovery, we foresee many potential applications for this approach.

Shape-based virtual screening with volumetric aligned molecular shapes.

Shape-based virtual screening is an established and effective method for identifying small molecules that are similar in shape and function to a reference ligand. We describe a new method of shape-based virtual screening, volumetric aligned molecular shapes (VAMS). VAMS uses efficient data structures to encode and search molecular shapes. We demonstrate that VAMS is an effective method for shape-based virtual screening and that it can be successfully used as a prefilter to accelerate more computationally demanding search algorithms. Unique to VAMS is a novel minimum/maximum shape constraint query for precisely specifying the desired molecular shape. Shape constraint searches in VAMS are particularly efficient and millions of shapes can be searched in a fraction of a second. We compare the performance of VAMS with two other shape-based virtual screening algorithms a benchmark of 102 protein targets consisting of more than 32 million molecular shapes and find that VAMS provides a competitive trade-off between run-time performance and virtual screening performance.

ZINCPharmer: pharmacophore search of the ZINC database.

ZINCPharmer (http://zincpharmer.csb.pitt.edu) is an online interface for searching the purchasable compounds of the ZINC database using the Pharmer pharmacophore search technology. A pharmacophore describes the spatial arrangement of the essential features of an interaction. Compounds that match a well-defined pharmacophore serve as potential lead compounds for drug discovery. ZINCPharmer provides tools for constructing and refining pharmacophore hypotheses directly from molecular structure. A search of 176 million conformers of 18.3 million compounds typically takes less than a minute. The results can be immediately viewed, or the aligned structures may be downloaded for off-line analysis. ZINCPharmer enables the rapid and interactive search of purchasable chemical space.

PocketQuery: protein-protein interaction inhibitor starting points from protein-protein interaction structure.

PocketQuery (http://pocketquery.csb.pitt.edu) is a web interface for exploring the properties of protein-protein interaction (PPI) interfaces with a focus on the discovery of promising starting points for small-molecule design. PocketQuery rapidly focuses attention on the key interacting residues of an interaction using a 'druggability' score that provides an estimate of how likely the chemical mimicry of a cluster of interface residues would result in a small-molecule inhibitor of an interaction. These residue clusters are chemical starting points that can be seamlessly exported to a pharmacophore-based drug discovery workflow. PocketQuery is updated on a weekly basis to contain all applicable PPI structures deposited in the Protein Data Bank and allows users to upload their own custom structures for analysis.