Wednesday, October 5, 2022
HomeBiologyHypDB: A functionally annotated web-based database of the proline hydroxylation proteome

HypDB: A functionally annotated web-based database of the proline hydroxylation proteome


Quotation: Gong Y, Behera G, Erber L, Luo A, Chen Y (2022) HypDB: A functionally annotated web-based database of the proline hydroxylation proteome. PLoS Biol 20(8):
e3001757.

https://doi.org/10.1371/journal.pbio.3001757

Tutorial Editor: Sui Huang, Institute for Programs Biology, UNITED STATES

Acquired: February 2, 2022; Accepted: July 13, 2022; Revealed: August 26, 2022

Copyright: © 2022 Gong et al. That is an open entry article distributed below the phrases of the Artistic Commons Attribution License, which allows unrestricted use, distribution, and replica in any medium, offered the unique writer and supply are credited.

Information Availability: All related knowledge are inside the paper and its Supporting Info recordsdata.

Funding: This work was supported by the Nationwide Institute of Well being (R35GM124896 to Y.C.). The funders had no function in research design, knowledge assortment and evaluation, determination to publish, or preparation of the manuscript.

Competing pursuits: The authors have declared that no competing pursuits exist.

Abbreviations::
ASA,
accessible floor space; ccRCC,
clear cell renal cell carcinoma; DDA,
data-dependent acquisition; DIA,
data-independent acquisition; HIF,
hypoxia-induced issue; Hyp,
proline hydroxylation; KNN,
k-nearest neighbor; PGD,
6-phosphogluconate dehydrogenase; P4HA,
prolyl 4-hydroxylase; PHD,
prolyl hydroxylase area; PTM,
posttranslational modification; RSA,
relative solvent accessibility

1. Introduction

Proline hydroxylation (Hyp), first found in 1902, is a vital protein posttranslational modification (PTM) pathway in mobile physiology and metabolism [14]. As a easy addition of a hydroxyl group to the imino aspect chain of proline residue, the modification is discovered to be evolutionarily conserved from micro organism to people. In mammalian cells, Hyp is essentially mediated by way of the enzymatic actions of two main households of prolyl hydroxylases—collagen prolyl 4-hydroxylases (P4HAs) [57] and hypoxia-induced issue (HIF) prolyl hydroxylase area (PHD) proteins [812], whereas there aren’t any identified enzymes able to eradicating protein-bound Hyp but. For the reason that exercise of prolyl hydroxylases is dependent upon the mobile collaboration of a number of co-factors, together with oxygen and iron, in addition to a number of metabolites, equivalent to alpha-ketoglutarate, succinate, and ascorbate, the Hyp pathway is a vital metabolic-sensing mechanism within the cells and tissues.

Probably the most well-characterized Hyp targets are collagen proteins and HIFα household of transcription components. Hyp on collagens mediated by P4Hs is essential to sustaining the triple-helical construction of the collagen polymer and enabling the correct protein folding after translation. Certainly, including an electronegative oxygen on the proline 4R place promotes the trans-conformation and stabilizes the secondary construction of collagen [1]. Inhibition of collagen Hyp destabilizes the collagen and prevents its export from the ER, due to this fact inducing cell stress and dying [1315]. HIFα transcription components are important to mediate hypoxia-response in mammalian cells [1618]. Hyp of HIFα proteins mediated by PHD proteins below normoxia situation is acknowledged by pVHL within the Cullin 2 E3 ligase complicated, which results in speedy ubiquitination and degradation of HIFα proteins [19,20]. Hypoxia situation inhibits HIFα Hyp and degradation, enabling the transcriptional activation of over 100 hypoxia-responding genes [2123].

Previously 2 a long time, quite a few research pushed by advances in mass spectrometry-based proteomics expertise have reported the identification and characterization of various new Hyp targets and the essential roles of the modification in physiological features [2429]. Hyp has been well-known to have an effect on protein homeostasis and the traditional instance is the PHD-HIF-pVHL regulatory axis. The same mechanism additionally regulates the turnover of various key transcriptional, metabolic, and signaling proteins, together with β2AR, NDRG3, ACC2, EPOR, G9a, and SFMBT1, and many others. [3034]. Along with pVHL-mediated protein degradation, Hyp additionally regulates substrate degradation by affecting its interplay with deubiquitinases. For instance, the hydroxylation of Foxo3a promotes substrate degradation by inhibiting the interplay with deubiquinase Usp9x, and hydroxylation of p53 enhances its interplay with deubiquitinases Usp7/Usp10 to stop its speedy degradation [35,36]. P4H-mediated Hyp has additionally been identified to manage the steadiness of various substrates together with AGO2 and Carabin [37,38]. Along with protein degradation, Hyp can even have an effect on protein–protein interplay to manage signaling and transcriptional actions. For instance, PKM2 hydroxylation promotes its binding with HIF1A for transcriptional activation, Hyp of AKT enhances the interplay with pVHL to inhibit the kinase exercise of AKT, and PHD1-mediated hydroxylation of Rpb1 is important for its translocation and phosphorylation [3942]. Extra lately, TBK1 hydroxylation was recognized and located to induce pVHL and phosphatase binding, which decreases its phosphorylation and enzyme exercise, whereas the lack of pVHL hyperactivates TBK1 and promotes tumor improvement in clear cell renal cell carcinoma (ccRCC) [27,43].

Regardless of these advances, there’s a lack of an built-in and annotated knowledgebase devoted for Hyp, which underappreciates the practical range and physiological significance of this evolutionarily conserved metabolic-sensing PTM pathway. To fill the information hole, we developed a publicly accessible Hyp database, HypDB (http://www.HypDB.web site) (S1 Fig). The event of the HypDB offers 3 major options—first, a classification-based algorithm for assured identification of Hyp substrates; second, built-in assets primarily based on exhaustive handbook literature mining, large-scale LC-MS evaluation, and curated public database; and third, a set of a giant spectral library for LC-MS-based site-specific identification from quite a lot of cell strains and tissues. Moreover, stoichiometry-based quantification of Hyp websites permits quantitative comparability of web site abundance throughout numerous proteins and tissues, and the extensively annotated Hyp proteome permits deep bioinformatic evaluation, together with community connectivity, structural area enrichment, and tissue-specific distribution research. The net database system permits the community-driven submission of LC-MS datasets to be included in HypDB annotation and the direct export of precursor and fragmentation with spectral library that allows the event of focused quantitative proteomics and data-independent evaluation workflow. We hope that the HypDB will present essential insights into the practical range and community of the Hyp proteome and support in additional mechanistic research on the physiological roles of the metabolic-sensing PTM pathway in cells and ailments.

2. Outcomes

2.1. Database building and evaluation workflow

To assemble a bioinformatic useful resource for metabolic-sensing Hyp targets, we developed HypDB, a MySQL-based relational database on a public-accessible net server (Figs 1 and S2). It was constructed primarily based on 3 major assets to comprehensively annotate human Hyp proteome (Fig 1). First, handbook curation of literature by way of PubMed (looking out time period: “proline hydroxylation” and time restrict between 2000 and 2021) was carried out by 2 impartial curators, which yielded 1,287 analysis journal articles. Website identification was extracted from every journal article, and its corresponding protein was mapped to UniProt protein ID if doable. Handbook curation of the analysis articles targeted on the websites that had been biochemically investigated with a number of proof together with mass spectrometry, mutagenesis, western blotting in addition to in vitro or in vivo enzymatic assays. Analyzed Hyp web site identifications had been then matched towards the prevailing knowledge within the database to scale back redundancy. Second, the database included in depth LC-MS-based direct proof of Hyp web site identifications primarily based on the built-in evaluation of over 100 LC-MS datasets of assorted human cell strains and tissues (see Experimental strategies). The datasets had been both downloaded from publicly accessible server or produced in-house. Every dataset was analyzed by way of a standardized workflow utilizing MaxQuant search engine, and the Hyp web site identifications had been filtered and imported into the HypDB with a streamlined bioinformatic evaluation pipeline laid out in particulars under. Our assortment of MS-based proof of Hyp identifications from cell strains and tissues possible revealed a good portion of Hyp websites that may be doubtlessly recognized by deep proteomic evaluation as evidenced by our statement that the speed of distinctive Hyp web site addition from every dataset decreased considerably regardless of the elevated assortment of datasets within the database (S2B Fig). Third, the HypDB additionally built-in Hyp identification annotated within the public UniProt database. For higher clarification, the database information point out whether or not the location was uniquely reported by the UniProt database or by each UniProt annotation and proof from large-scale LC-MS evaluation.

thumbnail

Fig 1. Workflow of creating HypDB database and webserver.

HypDB was constructed by way of deep proteome profiling evaluation of human tissues and cell strains, handbook literature mining, and integration with UniProt knowledge supply. Classification-based algorithm was utilized to extract assured identifications, and site-specific bioinformatic evaluation with stoichiometry-based quantification revealed the biochemical pathways concerned with human Hyp proteome. MS-based Hyp library additional enabled DIA-MS quantification of Hyp proteome in cells and tissues. DIA, data-independent acquisition; Hyp, proline hydroxylation.


https://doi.org/10.1371/journal.pbio.3001757.g001

We applied stringent standards for knowledge importing and classification from LC-MS-based identifications. To import knowledge into the HypDB, LC-MS-based identification of Hyp web site from database search evaluation was first analyzed by a classification-based algorithm to find out the boldness of Hyp web site identification and localization (Fig 2A). The classification was carried out utilizing the perfect scored MS/MS spectrum of a Hyp web site in every dataset evaluation. The algorithm categorised Hyp identifications that may be completely localized to proline residue primarily based on consecutive b- or y-ions as Class I websites. The algorithm categorised the Hyp identifications that can not be completely localized primarily based on MS/MS spectrum evaluation however might be distinguished from 5 frequent kinds of oxidation artifacts (methionine, tryptophan, tyrosine, histidine, phenylalanine) primarily induced throughout pattern preparation as Class II websites. Different Hyp identifications that had been reported by the MaxQuant database search software program (with 1% false-discovery charge on the site-level and a minimal Andromeda rating of 40) had been grouped as Class III websites. We additional developed a site-localization rating utilizing the relative intensities of key fragment ion to index the extent of confidence in web site localization with MS/MS spectrum evaluation for Class I and Class II websites (Experimental strategies). Every dataset was analyzed by the classification algorithm individually, and the perfect classification proof for every Hyp web site was chosen and reported on the HypDB web site to point the boldness of web site localization. The classification-based algorithm offers the specificity and reliability required for an precisely annotated database whereas sustaining all doable identifications as searchable information. And the localization credit score rating distribution of Class I and Class II websites had been proven in S2C and S2D Fig.

thumbnail

Fig 2. Substrate range of the human Hyp proteome.

(A) Illustration of classification-based algorithm to determine assured Hyp websites. (B)Venn diagram of Class I, II, III Hyp websites recognized from MS evaluation and manually curated UniProt websites. (C) PTM regulatory enzymes recognized as Hyp substrates. (D) Kinase tree classification displaying the distributions of kinases as Hyp substrates in numerous kinase households, together with AGC (named after PKA, PKG, PKC households), CAMK (leaded by calcium/calmodulin-dependent protein kinases), CK1 (cell kinase 1), CMGC (named after CDKs, MAPK, GSK, CLK households), STE (homologs of the yeast STE counterparts), TK (tyrosine kinases), and TKL (tyrosine kinase-like). (E) Hydroxyproline proteins that work together with EGLN1. (F) Hydroxyproline proteins that work together with P4HA2. Confer with Sheet A in S2 Desk and Sheet A–G in S3 Desk for the underlying knowledge of Fig 2B–2F. Hyp, proline hydroxylation; PTM, posttranslational modification.


https://doi.org/10.1371/journal.pbio.3001757.g002

To guage the site-specific prevalence of Hyp, a stoichiometry-based quantification technique was built-in into the evaluation workflow utilizing the beforehand established ideas [27,44]. Briefly, the Hyp stoichiometry was calculated by dividing the summed intensities of the peptides containing the Hyp web site identification with the entire intensities of the peptides containing the identical proline web site within the dataset. HypDB recorded all accessible site-specific Hyp stoichiometry evaluation from numerous cell strains and tissues, which allowed site-specific quantitative evaluation of modification abundance throughout cell and tissue varieties. And the median stoichiometry of all stoichiometry measurements for any particular web site was calculated and reported on the HypDB web site.

To additional discover the practical affiliation of Hyp proteome, a number of bioinformatic annotation methods had been built-in into the evaluation workflow as part of the information importing course of. These stand-alone workflows embrace evolutionary conservation evaluation, solvent accessibility evaluation, and protein–protein interface evaluation. Evolutionary conservation evaluation in contrast the conservation of Hyp websites with different proline websites on the identical protein and carried out a statistical check to find out if the Hyp web site is extra evolutionarily conserved than non-Hyp websites. Solvent accessibility evaluation analyzed the sequence of the substrate protein with DSSP package deal and calculated the chance of solvent accessibility for every Hyp websites. Protein–protein interplay interface evaluation extracted the area interplay residues from the 3DID database primarily based on PDB construction evaluation and matched them towards the Hyp web site within the database to determine the Hyp web site that’s localized within the interface and extra more likely to intervene with protein–protein interplay.

All info above was built-in into a number of tables and linked by way of overseas keys because the schema in S2A Fig. Full info on all Hyp websites was organized in 2 main tables together with a redundant web site desk (S1 Desk), which saved all Hyp websites recognized in numerous tissues and cell strains together with annotated MS/MS spectra, site-specific abundance and pattern supply info, and a nonredundant web site desk (Sheet A in S2 Desk), which merged the LC-MS-based proof from completely different sources on the site-specific stage and likewise built-in with the websites collected from UniProt and handbook curation of literatures.

2.2. Validation of the Hyp web site classification technique

To validate our classification-based technique for confidence Hyp web site identification, we carried out comparative evaluation of Hyp web site identifications from every class with manually curated UniProt Hyp identifications. Our evaluation confirmed that the Class I websites alone coated over 60% websites annotated within the UniProt, and a mixture of Class I and II websites coated about 63% of the UniProt websites, whereas only a few UniProt annotated websites overlapped with the Class III websites (Fig 2B), suggesting that our Hyp web site localization and classification algorithm allowed the gathering of extremely assured Hyp identification and considerably improved the reliability of LC-MS-based Hyp web site evaluation. To additional probe the present state of the Hyp proteome, we carried out in depth bioinformatic evaluation for practical annotation of the Hyp proteome primarily based on extra assured Hyp web site identifications in HypDB, which excluded Class III solely Hyp websites whose LC-MS proof can not distinguish them from potential oxidation artifacts.

2.3 Mapping human proline hydroxylation proteome

HypDB presently collected 14,413 nonredundant Hyp websites out of 59,436 Hyp web site information by way of large-scale deep proteomics evaluation of various tissue, cell strains, handbook curation of literatures, and integration with UniProt database. Amongst 14,413 nonredundant Hyp websites, 3,382 websites had been categorized as Class I websites, 4,335 websites had been categorized as Class II websites, and 6,432 had been categorized as Class III websites (Fig 2B). As well as, the database contained 55 websites from literature mining and 209 websites that had been built-in from the UniProt database. We utilized enrichment evaluation with Gene Ontology molecular operate annotation and located that Hyp substrates are broadly concerned in various mobile actions, from nucleotide binding and cell adhesion to enzymatic actions equivalent to oxidoreductase and ligases (S3 Fig and Sheet C in S4 Desk). Excluding Class III Hyp websites, we recognized a complete of 113 kinases (260 websites), 32 phosphatases (59 websites), 23 E3 ligases (47 websites), and 9 deubiquitinases (19 websites) as Hyp substrates (Fig 2C and Sheet A–D in S3 Desk). Statistical evaluation confirmed a selected enrichment of kinases in Hyp proteome (p = 0.037), suggesting a doubtlessly broad crosstalk between Hyp and kinase signaling pathways (Fig 2D and Sheet E in S3 Desk). Evaluating Hyp substrates with the interactome of prolyl hydroxylases in BioGRID [45], we recognized 22 Hyp proteins with 68 websites that had been identified to work together with EGLN1/PHD2, 17 Hyp proteins with 34 websites that had been identified to work together with EGLN2/PHD1, 416 Hyp proteins with 861 websites that had been identified to work together with EGLN3/PHD3, 58 Hyp proteins with 156 websites that had been identified to work together with P4HA1, 31 Hyp proteins with 296 websites that had been identified to work together with P4HA2, and 26 Hyp proteins with 66 websites that had been identified to work together with P4HA3 (Fig 2E and 2F and Sheet F and G in S3 Desk). The numbers of Class I, Class II Hyp websites, and Hyp proteins that work together with every prolyl hydroxylase had been collected in S4 Fig.

To find out if Hyp web site is extra accessible to solvent, we collected 3D buildings of proteins from PDBe and UniProt and calculated the relative solvent accessibility (RSA) of every proline residual on proteins with hydroxyproline websites with the DSSP package deal [46,47]. To look at if there’s an RSA distinction between Hyp websites and non-Hyp websites on protein with Hyp websites, we carried out a 2-tail t check and located no vital distinction within the distribution of solvent accessibility, suggesting that Hyp doesn’t essentially goal solvent accessible proline residues (S5A Fig and Sheet C in S2 Desk). To find out if Hyp targets proline websites which might be extra evolutionarily conserved, we carried out evolutionary conservation evaluation by way of in depth sequence alignment of protein orthologs throughout species primarily based on EggNOG database [48] and statistically in contrast the conservation of Hyp websites with the conservation of all proline on the identical protein. Our knowledge confirmed that about 49% websites had been evolutionarily conserved with statistical significance (p < 0.05) (S5B Fig). To find out if Hyp may play a possible function in area–area interactions, we analyzed knowledge of identified domain-based interactions of 3D protein buildings from HypDB nonredundant web site database. We recognized 168 distinctive Hyp websites that had been positioned on the interface of the interplay. These knowledge advised potential involvement of Hyp in straight regulating protein–protein interplay. For instance, Hyp at place 14 on Superoxide dismutase (SOD1) will type a hydrogen bonding with a neighboring chain Gln16 in a dimeric construction and doubtlessly promote the stabilization of the dimer (S5C Fig).

2.4. Useful options of proline hydroxylation proteins

We carried out GO enrichment assessments and different practical annotations on proteins that comprise Class I, II, literature, or UniProt websites (Fig 3A and Sheet A in S4 Desk). Our evaluation revealed that Hyp substrates are extremely enriched in metabolic processes equivalent to response to poisonous substances (p < 10−26) and natural cyclic compound catabolic course of (p < 10−14), mRNA splicing (p < 10−26) and structural features equivalent to NABA collagens (p < 10−35), supramolecular fiber group (p < 10−41), and cell morphogenesis concerned in differentiation (p < 10−18). To find out if the Hyp proteome prefers to be concerned in protein–protein interactions, we extracted a human protein interplay database from STRING with a cutoff rating of 0.7, after which, extracted all of the interactions containing 2 Hyp proteins primarily based on the STRING database. Primarily based on these knowledge, we carried out community connectivity evaluation by evaluating the variety of interactions of Hyp proteins with the distribution of the variety of interactions from randomly chosen human proteins with 10,000 occasions of repeats. Our knowledge confirmed that Hyp substrates are considerably concerned within the protein–protein interplay community (p < 0.0001) (Fig 3B and Sheet D in S2 Desk). We additional carried out protein complicated enrichment evaluation utilizing manually curated CORUM database, and our evaluation confirmed that Hyp proteome is considerably enriched with many identified protein complexes (S6 Fig and Sheet D in S4 Desk), equivalent to TNF-alpha/NF-kappa B signaling complicated 6 (S7A Fig and Sheet B in S4 Desk), TLE1 corepressor complicated (S7B Fig and Sheet B in S4 Desk), DGCR8 multiprotein complicated (S7C Fig and Sheet B in S4 Desk), Nop56p-associated pre-rRNA complicated (S7D Fig and Sheet B in S4 Desk), and PA700-20S-PA28 complicated (S7E Fig and Sheet B in S4 Desk), suggesting that Hyp targets proteins in a number of pathways that impacts signaling and gene expression. Utilizing MCODE clustering evaluation, we extracted considerably enriched clusters from Hyp proteome interplay community, and these extremely linked clusters of Hyp substrates advised that Hyp targets essential mobile actions together with regulation of mRNA splicing, hypoxia response, and focal adhesion (Fig 3C–3E and Sheet B in S4 Desk).

thumbnail

Fig 3. Gene enrichment and connectivity evaluation of HypDB.

(A) Interplay community of prime 20 enriched practical annotation clusters of HypDB proteins. (B) Bootstrapping-based evaluation of hydroxyproline protein interactions evaluating to a distribution of protein interactions from random samples with the identical variety of human proteins. (C) Hydroxyproline proteins enriched within the regulation of RNA splicing. (D) Hydroxyproline proteins enriched within the response to hypoxia. (E) Hydroxyproline proteins enriched in focal adhesion. Confer with Sheet A in S4 Desk, Sheet D in S2 Desk, and Sheet B in S4 Desk for the underlying knowledge of Fig 3.


https://doi.org/10.1371/journal.pbio.3001757.g003

2.5. Structural and motif options of proline hydroxylation websites

We analyzed the native sequence context round Hyp websites (excluding Class III websites) utilizing the MoMo software program instrument [49]. As we anticipated, Hyp websites with PG motif and GPPG motif had been extremely enriched (p < 10−10) which is attribute for collagen protein households (Figs 4A and S8A and Sheet A and B in S5 Desk). Along with collagen, we recognized 33 proteins with comparable motif to collagen, and these proteins could also be potential substrates of prolyl-4-hydroxylases. Apart from the collagen-like motif, we additionally recognized CP motif (p < 10−6) (Fig 4A and Sheet C in S5 Desk), and proteins containing CP motifs are extremely enriched in focal adhesion (FDR < 0.05). To take away the excessive background of web sites with collagen-like Hyp motifs, we filtered out websites with native sequence contexts in PG motif. Our re-analysis recognized that acidic amino acids had been enriched on the +1 place to type PD motif (Fig 4A and Sheet D in S5 Desk). PD motif containing proteins had been extremely enriched in metabolic pathways (FDR < 0.05). The quantity and proportion of Hyp websites represented within the HypDB proteome that appeared within the motifs above are proven in S8B Fig. As Hyp websites might have crosstalk with different protein, our evaluation revealed 2,386 phosphorylation websites and 535 ubiquitination websites which were recognized very near the Hyp websites (S8C Fig).

thumbnail

Fig 4. Motif and protein characteristic evaluation of HypDB.

(A) Motif enrichment evaluation with the flanking sequences of Hyp websites recognized PG, GPPG and CP motifs (adj p < 10−6) and repeated evaluation with the flanking sequences of Hyp websites after filtering out PG motif sequences recognized PD motif (adj p < 10−6). (B) Secondary construction enrichment of Hyp websites primarily based on PDB protein buildings (*p < 0.05). (C) Useful area enrichment evaluation of Hyp websites primarily based on area localizations on proteins in UniProt (***p < 0.001). (D) Useful area enrichment evaluation of Hyp websites primarily based on area localizations on proteins in UniProt (***p < 0.001, **p < 0.01). Confer with the S5 and S6 Tables for the underlying knowledge of Fig 4.


https://doi.org/10.1371/journal.pbio.3001757.g004

To find out the structural options of Hyp websites, we extracted all Hyp proteins with identified secondary buildings. These proteins comprise 2,279 Hyp websites and 27,159 non-Hyp websites on sequences which have experimentally decided PDB construction. We then categorised construction options into helix, sheet, flip, and non-structure areas and carried out statistical evaluation to match the secondary construction options of Hyp websites and non-Hyp websites. We discovered that Hyp certainly preferentially targets proline residues which might be localized within the helix (p < 0.05) and switch secondary buildings (p < 0.05) (Fig 4B left panel). Accordingly, we noticed a depletion of Hyp websites exterior of a secondary construction characteristic (Fig 4B proper panel).

As secondary buildings might not totally characterize practical structural options, we developed an identical statistical evaluation technique to find out the site-specific enrichment of Hyp websites on practical domains or structural areas. In distinction to the normal area enrichment evaluation utilizing Pfam or Interpro for protein-level evaluation, our technique enabled site-specific enrichment evaluation of domains or areas primarily based on UniProt annotation. Software of this technique revealed various identified and novel structural options that had been extremely enriched with Hyp, such because the triple-helical area, which is attribute for collagen protein household (Fig 4C). Along with the triple-helical area, our evaluation revealed greater than 10 practical areas and domains that had been extremely enriched with Hyp, together with p-domain (p < 10−6), NBD area (p < 10−6), thioredoxin area (p < 10−2), and ferritin-like area (p < 10−6) (Fig 4C and 4D). These knowledge revealed beforehand sudden function of Hyp concentrating on practical domains in various mobile pathways.

2.6. Website-specific stoichiometric quantification of Hyp proteome

Evaluating to relative quantification, stoichiometry evaluation measures the prevalence and dynamics of the modification in a physiologically significant method [27,50,51]. Our mass spectrometry-based deep proteome profiling permits site-specific quantification of Hyp stoichiometries throughout a number of tissues and cell strains. Our knowledge confirmed that site-specific abundance of Hyp varies broadly from under 1% to almost 100% with an general median stoichiometry of seven.89% (Fig 5A and Sheet A in S8 Desk). Certainly, a bulk portion of the Hyp websites have both very low or very excessive stoichiometries. To analyze the practical variations between websites with completely different stoichiometry, we divided proteins into 5 quantiles primarily based on common stoichiometry measurement for a similar web site throughout all cells and tissues (Fig 5B and Sheet B in S8 Desk). The 4 cutoffs 5%, 20%, 80%, and 95% had been chosen so that every quantile contained an identical variety of Hyp websites. We then carried out GO enrichment and practical annotation on the 5 quantiles respectively and carried out hierarchical clustering with correlation coefficient. Our knowledge confirmed that proteins in immune response and neutrophil activation pathways are enriched with low to medium stoichiometry, and proteins in cell adhesion and system improvement are enriched with medium to excessive stoichiometry (Fig 5B). We additionally noticed a major enrichment of proteins concerned in chromatin meeting and RNA processing however the stoichiometry of hydroxylation on these proteins gave the impression to be very low (Fig 5B). Combining site-specific practical characteristic annotation and stoichiometry evaluation, we carried out stoichiometry-based clustering of Hyp-targeted practical domains. Our knowledge confirmed that ODD area that’s identified to manage hydroxylation-mediated protein degradation of HIFα was enriched with medium stoichiometry, and triple-helical area on collagen, whose hydroxylation is required for its maturation, was enriched with excessive stoichiometry (Fig 5C and Sheet C in S8 Desk). Moreover, our evaluation revealed stoichiometry-based enrichment of kinase domains at medium stoichiometry, GATA1 interplay domains at excessive stoichiometry, nucleotide-binding domains at low to medium stoichiometry, and histone-binding domains at low stoichiometry (Fig 5C).

thumbnail

Fig 5. Stoichiometry-based practical enrichment evaluation of the Hyp proteome.

(A) Stoichiometry distribution of the Hyp websites divided into 5 quantiles—Q1, Q2, Q3, This fall, and Q5, from low to excessive stoichiometry with 4 cutoffs of 5%, 20%, 80%, and 95% respectively. (B, C) Hierarchical clustering of GO organic processes enrichment of Hyp proteins (B) and practical area enrichment of Hyp websites on proteins in UniProt (C) throughout the 5 quantiles. Confer with Sheet A–C in S8 Desk for the underlying knowledge of Fig 5.


https://doi.org/10.1371/journal.pbio.3001757.g005

2.7. Tissue-specific distribution of Hyp proteome

The gathering of mass spectrometry-based identification of Hyp proteome enabled cross-tissue comparative evaluation (Sheet A in S8 Desk). Certainly, at particular person protein stage, we noticed a large distribution of Hyp abundance for a similar web site and between completely different websites throughout completely different tissue (Figs 6A and S9). For instance, Fibrillin-1 (FBN1) was recognized with 22 Hyp websites of which 17 had been Class I or II websites. Hyp1090 on EGF_CA repeat confirmed constant excessive Hyp stoichiometry (71% to 96%) throughout 4 completely different tissues (testis, colon, coronary heart, and rectum), whereas Hyp1453 on one other EGF_CA repeat confirmed various Hyp stoichiometry (3% to 50.5%) throughout the identical 4 tissues (testis, colon, coronary heart, and rectum) (Fig 6A). In one other instance, 6-phosphogluconate dehydrogenase (PGD) was recognized with 8 Hyp websites with half of them belonging to Class I or II websites. Hyp169 on the NAD-binding area confirmed comparatively low stoichiometries in coronary heart, liver, and ovary (7.6% to 11.6%) however a lot increased stoichiometries in intestine and B cell (21.9% and 75.6%) (S9B Fig). We carried out pathway enrichment evaluation of Hyp and clustering of the enrichment throughout the tissues. Our knowledge confirmed that Hyp proteome various dramatically when it comes to pathway and abundance amongst tissues (Fig 6B and 6C and Sheet D in S8 Desk). For instance, in lung, the Hyp proteome is especially concerned in collagen synthesis and tissue improvement, and it has comparatively low portion of distinctive Hyp websites, however in liver, the Hyp proteome is closely concerned in various metabolic and translational processes with many liver-specific Hyp targets (Fig 6B and 6C). Apparently, clustering evaluation confirmed that tissues sharing comparable physiological features are inclined to share comparable Hyp profiles and are due to this fact clustered collectively. Testis and ovary, for instance, have comparable enrichment of Hyp proteins associated to chromosome group, DNA restore, and different DNA-related metabolic processes (Fig 6D and Sheet E in S8 Desk). Hyp proteomes in urinary bladder and prostate are co-enriched in regulation of proteolysis and morphogenesis of various tissues. CD4 T cells and CD8 T cells are enriched with Hyp proteins associated to chromatin transforming and immune system improvement. Liver confirmed a particular enrichment sample evaluating to different tissues, and its Hyp proteome is strongly enriched in numerous metabolic and catabolic processes. In the meantime, 4 of those tissues: ovary, testis, liver, and prostate, co-enriched in neutrophil activation concerned in immune response (Fig 6D).

thumbnail

Fig 6. Hyp proteome distributions in numerous tissues.

(A) An instance displaying various stoichiometries of Hyp websites throughout various kinds of tissue for FBN1 with protein domains labeled in coloured bins. (B) Correlation plot of Hyp proteins in 5 completely different tissues: coronary heart, liver, lung, ovary, and urinary bladder with the dimensions of arc reveals relative quantity and the purple curved strains displaying overlap proteins (C) Warmth map of the highest 20 enriched practical annotations of the Hyp proteins in 5 tissues. (D) GO organic course of enrichment warmth map of the Hyp proteins throughout 7 tissues. Confer with Sheet A in S2 Desk, Sheet D-E in S8 Desk for the underlying knowledge of Fig 6B–6D. CD4, CD4 T cells; CD8, CD8 T cells; FBN1, Fibrillin-1; Hyp, proline hydroxylation; P, prostate; UB, urinary bladder.


https://doi.org/10.1371/journal.pbio.3001757.g006

2.8. Information-independent acquisition (DIA) evaluation of Hyp targets with HypDB-generated spectral library

DIA has been developed prior to now 10 years as a strong technique for dependable and environment friendly quantification of proteins and PTM websites [5260]. Our in depth assortment of the MS-based proof for human Hyp websites offered a great useful resource to ascertain a DIA workflow for world, site-specific quantification of Hyp targets in cells and tissues. To this finish, our net server has built-in features for the direct export of annotated MS/MS identification of Hyp websites for chosen proteins, cell line, tissue, or at a proteome scale. The Export operate offered 2 choices—exporting the peptide precursor m/z solely or exporting formatted MS/MS spectra. The previous possibility can generate goal m/z checklist that can be utilized as an inclusion checklist for focused quantification of Hyp websites on chosen proteins or websites. The latter possibility can straight generate spectral library used for DIA evaluation. Utilizing the Export operate, the present HypDB allowed the era of a complete Hyp spectral library within the NIST Mass Search format (msp) consisting of 6,000 precursor ions, 5,307 peptides, representing 7,717 Class 1 and a pair of websites from 3,022 proteins. The webserver was additionally built-in with the varied choices for selective exporting. To show the applicability of our useful resource in DIA evaluation workflow, we analyzed 2 lately printed large-scale DIA evaluation datasets [55,56]. Each datasets utilized DIA evaluation to quantify protein dynamics within the a number of replicates of paired regular and tumor samples.

The research by Kitata and colleagues analyzed world protein profiles of lung most cancers with 5 pairs of tumor and regular tissues in triplicate evaluation for a complete of 30 DIA-based LC-MS runs [55]. As a routine process in DIA evaluation, we first carried out database looking out of data-dependent acquisition (DDA) knowledge within the dataset. Then, utilizing the spectral library generated from the DDA knowledge in the identical research, we carried out DIA evaluation of all tumor and regular tissues with replicates. The evaluation quantified 1,339 Class 1 and a pair of Hyp websites from Kitata and colleagues research (1% FDR). Subsequent, we utilized the HypDB-generated spectral library and repeated the DIA evaluation. Our end result confirmed that utilizing the HypDB-generated spectral library led to greater than double the entire variety of Hyp websites utilizing a DDA-based spectral library with 3,015 Hyp websites recognized whereas protecting greater than 83% of the nonredundant Hyp websites recognized utilizing the two spectral libraries, suggesting that the applying of the HypDB-generated spectral library was ample to cowl majority of the Hyp identifications and considerably elevated the sensitivity of Hyp proteome protection (Fig 7A). DIA evaluation with a mixed library generated by each HypDB and DDA recognized 3,651 Hyp websites and 1,249 Hyp proteins (1% FDR). To find out the reproducibility of the quantification, we calculated the distribution of the share of coefficient variance (%CV) for DIA evaluation of Hyp websites. Our knowledge confirmed that %CV various between 2% and 15% with a median worth round 5% (Fig 7B), just like the %CV distribution noticed within the DIA evaluation of proteins and phosphoproteins [55]. Given the excessive reproducibility of the quantification, we filtered the Hyp websites with a worldwide 1% q-value cutoff (2,283 websites) and carried out hierarchical clustering evaluation of Hyp websites quantified with normalized depth in tumor and regular lung tissues (Fig 7C). Our knowledge clearly confirmed that site-specific Hyp quantification was ample to cluster and distinguish tumor versus regular tissue. To determine considerably up- or down-regulated Hyp websites in tumor tissues, we carried out a 2-sample t check and analyzed the information within the volcano plot (Fig 7D). The evaluation allowed us to determine 142 Hyp websites that had been considerably up-regulated and 178 Hyp websites that had been considerably down-regulated in tumor tissue (5% permutation-based FDR). The dynamically regulated Hyp websites confirmed sturdy traits that had been distinct between tumor and regular tissue. Apparently, we noticed subtype-dependent Hyp dynamics on collagen proteins. Collagen subtypes IV and VI confirmed considerably down-regulated Hyp stage throughout a number of websites in tumor samples, whereas collagen subtype X confirmed considerably elevated Hyp (Fig 7D). Since Hyp promotes the structural stability of collagens, such adjustments possible indicated a considerably enhance in stability for collagen X and reduce in stability for collagen IV and VI in lung most cancers tissue in comparison with the conventional tissue. Our discovering agreed effectively with a really current publication indicating a pro-metastatic function of up-regulated collagen X in lung most cancers development [61]. As well as, we additionally recognized vital up-regulation of Hyp on glycolysis enzymes pyruvate kinase (PKM), enolase (ENO1), and autophagy protein Parkin (PARK7) in tumor tissue (Fig 7D). P4HB, a member of the collagen prolyl 4-hydroxylase enzyme, additionally confirmed vital enhance in Hyp (Fig 7D), possible resulting from elevated prolyl 4-hydroxylase exercise in lung most cancers [62].

thumbnail

Fig 7. Label-free quantification of the Hyp proteome in lung most cancers with DIA evaluation.

(A) Venn diagram of DIA-based Hyp web site identifications utilizing HypDB-generated library and the library generated by the DDA in Kitata and colleagues research. (B) Distribution of %CV for Hyp websites quantified with HypDB-generated library, DDA-generated library, or the hybrid library that mixed each sources. (C, D) Hierarchical clustering (C) and volcano plot (D) of considerably up- or down-regulated Hyp websites in regular (blue) and tumor (purple) tissues within the DIA evaluation. (E, F) Considerably enriched GO organic processes amongst up-regulated (E) and down-regulated (F) Hyp proteins in tumor with not less than 1-fold change after normalizing with protein abundance adjustments. Confer with Sheet A–E in S9 Desk for the underlying knowledge of Fig 7B–F. DDA, data-dependent acquisition; DIA, data-independent acquisition; Hyp, proline hydroxylation.


https://doi.org/10.1371/journal.pbio.3001757.g007

In one other research, Guo and colleagues utilized DIA evaluation to quantitatively profile kidney most cancers proteome and the dataset consisted of an evaluation of 18 regular tissues and 18 tumor tissues [56]. Following the identical workflow, we first carried out DDA evaluation after which utilized DDA-generated Hyp library to quantify Hyp substrates in tissues. The DDA library-based evaluation solely quantified 387 Hyp websites from all replicate evaluation. Software of the HypDB-generated spectral library elevated the variety of Hyp web site quantifications by greater than 5 occasions, figuring out 2,510 websites (S10A Fig). Our end result confirmed that HypDB-generated library drastically elevated the Hyp sequence protection and evaluation sensitivity. DIA evaluation with a mixed library generated by each HypDB and DDA evaluation recognized 2,556 Hyp websites and 981 Hyp proteins (1% FDR). To check the reproducibility amongst replicate tissues, we carried out a correlation matrix evaluation utilizing the corrplot package deal in R. Our knowledge confirmed that quantitative evaluation of Hyp substrates allowed environment friendly clustering and segregation of tumor versus regular tissues (S10B Fig). After world q-value filtering and depth normalization, we analyzed 1,160 Hyp websites throughout all samples with pair-wise t check, and our evaluation recognized 12 up-regulated websites and 24 down-regulated Hyp websites in tumor (5% permutation-based FDR) (S10C Fig).

To know whether or not the differential abundance of Hyp websites between the conventional and tumor tissues was resulting from adjustments within the abundance of corresponding proteins, we in contrast the log2 reworked common web site ratios to the log2 reworked common protein ratios for each Kitata and colleagues and Guo and colleagues datasets (S11A and S11B Fig and S9 and S10 Tables). We discovered that greater than 82% of the Hyp websites in Kitata and colleagues dataset and not less than 37% of the Hyp websites in Guo and colleagues dataset might be quantified with the corresponding protein abundance (S9 and S10 Tables). From the correlative evaluation between web site ratios and protein ratios, we observed a sure diploma of linearity, suggesting the adjustments within the abundance of some Hyp websites had been certainly pushed by the adjustments within the abundance of corresponding proteins (S11A and S11B Fig). We additionally observed that a good portion of Hyp web site dynamics didn’t correlate with protein abundance adjustments. To this finish, we calculated 95% confidence interval alongside the bisector correlation strains that characterize equal ratios of Hyp web site and protein abundance adjustments for all Hyp websites with corresponding protein quantification ratios (S9 and S10 Tables). Our evaluation confirmed that 78% of the Hyp websites in Kitata and colleagues dataset and 35% of the Hyp websites in Guo and colleagues dataset confirmed vital deviation in web site abundance adjustments from the corresponding protein abundance adjustments (S11A and S11B Fig). The correlation evaluation due to this fact recognized Hyp substrates that confirmed differential adjustments in abundances evaluating to the corresponding protein abundance adjustments. We additional extracted solely the considerably up- or down-regulated Hyp websites primarily based on DIA evaluation and in contrast their dynamics with corresponding protein abundance adjustments (S11C–S11F Fig). Notably, in Kitata and colleagues dataset, the protein abundance of COL1A2 and COL14A1 was comparable between tumor and regular tissues, whereas the abundance of the Hyp websites on every of these proteins had been effectively above or under the 95% confidence interval (S11C and S11E Fig). The correlation evaluation additionally confirmed the down-regulation of Hyp abundance on collagen subtypes IV and VI in Kitata and colleagues lung most cancers dataset with the protein-level normalization, whereas displaying that the up-regulation of Hyp abundance on collagen subtype X in tumor was because of the up-regulation of the protein abundance (S11C and S11E Fig). In Guo and colleagues dataset, considerably modified Hyp websites confirmed good correlation with corresponding protein dynamics, whereas the Hyp websites of CRK and TPI1 confirmed a lot bigger enhance or lower in abundance in comparison with these of their complete proteins, suggesting differential actions of the Hyp pathways for every substrate (S11D and S11F Fig).

To disclose the practical significance of up-regulated or down-regulated Hyp substrates in each datasets, we carried out practical annotation enrichment evaluation with Hyp substrates whose web site ratios confirmed not less than 1-fold enhance or lower with protein abundance normalization. Evaluation of Kitata and colleagues dataset confirmed that the organic processes associated to homotypic cell–cell adhesion, coagulation, cell redox homeostasis, response to interleukin-12, and angiogenesis had been considerably enriched amongst up-regulated Hyp substrates (Fig 7E), whereas processes associated with regulation of gene expression, neutrophil-mediated immunity, carbohydrate catabolism, collagen metabolic course of, and response to interleukin-7 had been considerably enriched amongst down-regulated Hyp substrates (Fig 7F) (BH corrected FDR < 0.05). The evaluation of Guo and colleagues dataset confirmed that Hyp proteins up-regulated in kidney most cancers had been strongly enriched in KEGG pathways together with ECM-receptor interplay, focal adhesion, glyoxylate/dicarboxylate metabolism, and tryptophan metabolism (S10D Fig), whereas pathways together with biosynthesis of amino acids, fructose/mannose metabolism, pathgenic E. coli an infection and PI3K-Akt signaling had been considerably enriched amongst down-regulated Hyp proteins in tumor tissue (S10E Fig) (BH corrected FDR < 0.05).

3. Conclusions

A grand problem in practical evaluation of PTM pathways is the shortage of annotation assets to profile modification substrates and annotate enzyme-target relationships. Hyp is a key oxygen and metabolic-sensing PTM that governs the mobile applications in response to the hypoxia microenvironment and micronutrient stress. Earlier research of Hyp primarily targeted on its function in structural stability and maturation of cytoskeletal proteins equivalent to collagens. Previously a number of a long time, in depth biochemical research on HIF pathways in addition to different new Hyp substrates means that Hyp is broadly concerned in regulating protein–protein interplay, protein stability, sign transduction, metabolism, and gene expression. Rising proof has additionally advised that particular Hyp pathways play essential roles in most cancers improvement, metastasis, coronary heart illness, and diabetes. Systematic categorization and practical annotation of Hyp proteome will present complete understanding and essential physiological insights into Hyp-regulated mobile pathways in addition to potential therapeutic methods concentrating on metabolic-sensing Hyp pathways in ailments.

To handle this want, we developed HypDB, an built-in on-line portal and publicly accessible server for practical evaluation of Hyp substrates and their interplay networks. HypDB collected numerous knowledge sources for complete protection of Hyp proteome, together with handbook curation of printed literature, deep proteomics evaluation of tissues, and cell strains, in addition to integration with annotated UniProt database. The location-localization and classification algorithm enabled environment friendly extraction of assured Hyp substrate identification from LC-MS evaluation. Our identification of extremely assured Hyp substrates expanded the present annotation of human Hyp targets in UniProt by over 40-fold. Streamlined knowledge processing and stoichiometry-based Hyp quantification allowed site-specific comparative evaluation of Hyp abundance throughout 26 human organs and fluids in addition to 6 human cell strains. We collected 14,413 Hyp websites from numerous origins, and 86% of the highest 500 Hyp websites with essentially the most repeat identifications in numerous MS datasets had been structural proteins, which matched effectively with one in every of its most essential molecular operate.

Bioinformatic evaluation of the primary draft of human Hyp proteome supply essential insights into the practical and structural range of the modification substrates. The evaluation not solely revealed various mobile pathways enriched with Hyp proteins together with mRNA processing, metabolism, cell cycle, and signaling, but in addition demonstrated for the primary time that Hyp preferentially targets protein complexes and protein–interplay networks, indicating essential roles of Hyp in fine-tuning protein structural options and mediating protein–protein interactions. Certainly, evaluation of the expanded Hyp proteome with site-level secondary construction enrichment evaluation indicated a major enrichment of Hyp websites on the alpha-helix, whereas site-level enrichment evaluation of practical domains and areas revealed novel protein area options which might be preferentially focused by Hyp, equivalent to P-domain, NBD area, ferritin-like area, and thioredoxin. These findings advised doubtlessly essential roles for Hyp-mediated regulation of area stability or exercise which might be worthy of additional biochemical investigation.

MS-based evaluation of Hyp proteome permits the stoichiometry-based quantification of Hyp abundance on the site-specific stage. By classifying Hyp substrates primarily based on stoichiometry dynamics, we revealed the enrichment of practical domains and exercise with very excessive stoichiometry, indicating that Hyp on these domains could also be required for the protein operate, which has similarities to collagen. As compared, the oxygen-sensing ODD area was enriched with median stoichiometry and nucleotide or histone-binding domains had been enriched with low stoichiometry. Such distinction might recommend differential actions of prolyl hydroxylases concentrating on numerous practical domains. Comparative evaluation of Hyp stoichiometry throughout tissues additionally indicated variations in modification abundance on the site-specific stage. Such variation could also be attributed to the differential metabolic and gene expression profiles in numerous tissues.

The gathering of MS-based identification of Hyp proteome in HypDB established an annotated spectral library for Hyp-containing peptides that had been recognized and web site localized with excessive confidence. Such in depth spectral library enabled dependable and delicate evaluation of deep proteomic evaluation of human cells and tissues with DIA. Software of the HypDB-generated spectral library in DIA evaluation demonstrated wonderful knowledge reproducibility, considerably improved the protection of Hyp proteome in most cancers proteome evaluation and revealed novel enrichment of Hyp websites that had been considerably up-regulated or down-regulated in most cancers tissues.

Though the present version of HypDB (v1.0) is restricted to the human proteome, future improvement of HypDB will embrace Hyp proteome in different species. Comparative evaluation of Hyp targets from various species will permit evolutionary conservation evaluation of Hyp websites and determine functionally essential Hyp targets in protein construction and exercise. Additional utility of the HypDB-generated spectral library in tissue evaluation will allow the invention of novel Hyp targets in illness animal fashions or affected person samples and doubtlessly result in the event of clinically related therapeutic methods.

4. Experimental strategies

4.1. MS uncooked knowledge evaluation

We collected MS knowledge from the human proteome draft [63], deep proteome evaluation of human cell strains [64], PHD interactome evaluation [44,65], and Hyp proteome evaluation [27] in addition to IP-MS evaluation of Flag-tagged HIF1A. All MS uncooked knowledge collected above had been searched with MaxQuant (model 1.5.3.12) towards the UniProt human database whereas having carbamidomethyl cystine as mounted modification and protein N-terminal acetylation, methionine oxidation, and Hyp as variable modification. Many of the uncooked knowledge had trypsin because the digestion enzyme, whereas a couple of samples used different digestion enzymes, for instance, LysC and GluC, primarily based on the experimental process of authentic tasks. Most lacking cleavage quantity was set to 2 and the identification threshold was set at 1% false discovery charge for concatenated reversed decoy database search at protein, peptide, and web site ranges.

4.2. Website localization classification and scoring

To filter out low confidence websites, we developed the location localization classification algorithm. Primarily based on the expertise that websites are localized extra precisely when extra ion fragments are present in corresponding MS2 spectra serving to to localize the modification mass shift, our algorithm divided websites into 3 courses in accordance with their modification localization confidence: unique localized websites in Class I, websites nonexclusive however distinguishable from comparable modifications in Class II, and the remainder in Class III (Fig 2A).

For a web site to be categorised as Class I web site, a pair of b-ions or y-ions separating the proline from different amino acids should be discovered to localize it completely. On this manner, a mass shift attributable to hydroxylation can solely happen on that particular proline. And we gave credit to that ion pair within the scoring operate for Class I websites as follows:

the place CS stands for credit score rating, I stand for depth of various ion fragments, for instance, stands for the depth of bm-ion, and l stands for peptide size. We gave credit score to the pair of b-ions and y-ions that localizes hydroxylation completely. The one with decrease depth inside the pair will likely be chosen, and we calculate the credit score rating primarily based on the ratio of their intensities to common ion depth on the identical peptide.

Hydroxylation that can not be completely localized however distinguishable from occurring on different prion-to-oxidize amino acid residuals are categorised as Class II as a result of we are able to infer that hydroxylation happens on proline on this case. As all ions that separate proline from nearest amino acid might get oxidized simply, we gave credit to all ions that assist to separate them within the scoring operate for Class II websites as follows:

the place ll and lr for distance between hydroxylated proline and nearest prion-to-oxidation amino acid residual on the left aspect and proper aspect. As a substitute of solely giving credit score to the pair subsequent to the aspect, for Class II websites, we gave credit to all ions that contributed to separate Hyp with different prion-to-oxidation amino acid residues. We require that Hyp web site accommodates not less than 1 fragment ion on each left and proper flanking sequences excluding terminal fragment ions. After that, we additionally calculate the ratio between the common depth of chosen ions and all ions on either side, and the credit score rating is decided by the weaker aspect.

Websites that belong to neither Class I nor Class II are categorised as Class III websites. There are possibilities that Class III websites are Hyp on different positions or different modifications which might be recognized falsely. As a consequence of their low credibility, we don’t rating them and solely use extra assured Hyp identifications, which embrace Class I, Class II, UniProt, and literature websites for bioinformatic analyses.

4.5. Motif enrichment analysis

The protein sequences of the proteins represented in HypDB were downloaded from the UniProt database. In-house Python scripts were written to extract peptides that contained Hyp sites that passed our stringent filtering criteria. These peptides were extended to the length of 27 amino acids and centered around the hydroxylated proline residue. The prealigned peptides were uploaded to the MoMo (version 5.4.1) web application [49]. All protein sequences that had been obtained from the UniProt database had been set because the background for the evaluation. Inside the MoMo net utility, the motif-x algorithm was chosen. The minimal variety of occurrences for a motif was set to twenty. The sequence logos had been generated by the MoMo net utility.

4.9. Protein–protein interface evaluation

The interacting area pairs and cases of area–area interactions of 3D protein buildings had been downloaded from 3DID (https://3did.irbbarcelona.org/index.php). In-house Python scripts had been developed to research the variety of Hyp websites interacting with one other residue and the variety of Hyp websites inside 3 residues of an interacting residue.

4.10. Evolutionary conservation evaluation

Evolutionary conservation evaluation of Hyp websites was carried out utilizing EggNOG ortholog database (v5.0) and EggNOG-mapper on-line portal [48]. Briefly, first, utilizing EggNOG-mapper, Hyp proteins had been mapped to the corresponding ortholog teams. Subsequent, Hyp websites and non-Hyp proline websites on Hyp proteins had been aligned to ortholog sequences utilizing MAFFT algorithm [66]. The variety of matches a Hyp web site or non-Hyp proline web site to a proline for a similar positions in ortholog sequences and the entire variety of sequences within the ortholog group had been recorded. Lastly, HyperG check was carried out for every Hyp web site primarily based on normalized variety of matches to proline residues in ortholog sequences for Hyp websites and non-Hyp websites, in addition to the entire variety of any amino acid residues in ortholog sequences for a similar place because the Hyp websites or non-Hyp websites.

4.11. Improvement of web site and MySQL database

The web site serves as a front-end interactive interface of the database. It was developed utilizing HTML, CSS, Javascript, and PHP and works on a Linux-Apache-MySQL-PHP (LAMP) server structure. The front-end was designed utilizing the Bootstrap framework. Related protein knowledge are fetched utilizing APIs from a number of sources. Protein sequences, identifiers, and descriptions are fetched from entries within the UniProtKB/Swiss-Prot knowledgebase [67], protein secondary construction knowledge are fetched from PDBe [68], and domains are fetched from Pfam [69]. The protein sequences are displayed on the web site utilizing neXtProt Sequence Viewer (https://github.com/calipho-sib/sequence-viewer). The spectral graphs on the web site are visualized utilizing d3.js (https://d3js.org/). The backend of the web site makes use of PHP to interface with a MySQL database that accommodates the information as proven in S2A Fig.

4.12. Transfection and immunoprecipitation of HIF1A

Transfection and overexpression of Flag-tagged HIF1A was carried out following a process as beforehand described [70]. Flag-tagged HIF1A plasmid (Sino Organic) was transfected into 293T cells with polyethylenimine. Cells had been handled with 10 μm proteasome inhibitor MG-132 (Apexbio) for 4 hours previous to harvesting. Roughly 24 hours after transfection, cells had been washed with chilly PBS buffer and lysed in lysis buffer (150 mM NaCl, 50 mM Tris-HCL, 0.5% NP-40, 10% glycerol (pH 7.5), protease inhibitor cocktail (Roche)) on ice for 15 to twenty minutes. Then, the cell lysates had been clarified by centrifugation previous to the incubation with anti-FLAG M2 affinity gel (Sigma) for six hours at 4°C. After incubation, the M2 gel was washed with wash buffer (cell lysis buffer with 300 mM NaCl) for 3 occasions after which eluted with 3× Flag peptide (ApexBio). The eluate had been blended with 4× SDS loading buffer and boiled, after which, loaded onto home made SDS-PAGE gel and stained with Coomassie blue (Thermo Fisher).

4.13. In-gel digestion and LC-MS evaluation of HIF1A

A big gel piece protecting a large MW vary above 100 kDa was minimize out and subjected to discount/alkylation and in-gel digestion with trypsin (Promega) as beforehand described [51]. Tryptic peptides had been desalted with home made C18 StageTip and resuspended in HPLC Buffer A (0.1% formic acid) earlier than being loaded onto a capillary column (75 μm ID and 20 cm in size) in-house full of Luna C18 resin (5 μm, 100 Å, Phenomenex). The peptides had been separated with a linear gradient of seven% to 35% HPLC Buffer B (0.1% formic acid in 90% acetonitrile) at a stream charge of 200 nl/min on Dionex Final 3000 UPLC and electrosprayed right into a high-resolution Orbitrap Lumos mass spectrometer (Thermo Fisher). Peptide precursor ions had been acquired in Orbitrap with a decision of 120,000 at 200 m/z, and peptides had been fragmented with Electron Switch/Excessive Vitality Collision Dissociation (EThcd) with calibrated charge-dependent ETD parameters and ETD Supplemental Activation and purchased in Top12 data-dependent mode type by highest cost state and lowest m/z as precedence settings. Uncooked knowledge had been analyzed by Maxquant software program following the identical process and parameter setting as beforehand printed dataset as described above.

4.14. Utilization of HypDB web site

A devoted web site with built-in MySQL database was established to host the HypDB service. The database schema consists of 4 tables representing redundant Hyp web site identifications, nonredundant Hyp web site identifications, interplay interface evaluation, evolutionary conservation evaluation, and solvent accessibility evaluation. Every document within the web site identification desk is assigned a novel HypDB web site ID. The web site was designed with the Bootstrap framework (v4.1.3) and options a number of key features together with a Search bar, Protein info web page, Website info web page, Database abstract, Add/contribute web page, and Obtain/export web page.

The Search bar permits the person to enter a UniProt accession quantity or Gene title of the protein of curiosity, and the server will use the data to extract and show a ranked checklist of most comparable entries in actual time. Clicking on an entry will convey the person to the protein info web page the place protein identifiers, description, and protein sequence are displayed. All Hyp websites are recognized on the protein sequence in addition to identified acetylation and phosphorylation websites from PhosphoSitePlus database [71] are highlighted by completely different colours. The checklist of Hyp websites is additional displayed under the sequence within the desk that features the location properties together with localization class, localization rating, stoichiometry, solvent accessibility, and evolutionary conservation info. Hyp web site desk is adopted by properties of Hyp proteins together with protein–protein interplay, secondary construction, practical domains, and area–area interactions. Hyp websites recognized with MS/MS proof within the HypDB have a “Particulars” button displayed for every web site within the web site desk on the protein info web page. Clicking on the Particulars button will convey the person to the peptide info web page the place the perfect recognized MS/MS spectrum for the location is displayed with annotations of fragment ions.

The Contribute/Add web page permits the neighborhood to contribute uncooked MS/MS identifications to the HypDB by way of an embedded Google Type. Info relating to the uncooked knowledge sort, location, pattern sort, database looking out parameters in addition to person info will likely be entered into database. Uncooked knowledge will likely be downloaded and processed utilizing the identical streamlined workflow. The information will cross by way of the classification and site-localization evaluation course of and annotated with the bioinformatic workflows as described above. The ultimate knowledge will likely be deposited into the HypDB to share with the analysis neighborhood.

The Export/Obtain web page permits the neighborhood to obtain the entire dataset deposited within the HypDB together with each redundant and nonredundant modification web site tables. As well as, the Export operate permits customers to pick an inventory of proteins, tissues of pursuits, filter websites primarily based on localization credit score class, MS fragmentation sort, proteolytic enzyme utilized in proteomics evaluation, in addition to specify the precursor ion m/z of Hyp proteins for export to arrange focused quantification methodology when buying knowledge or export the collected spectral libraries of Hyp websites from the chosen Hyp proteins to carry out database looking out with DIA.

4.16. DIA knowledge evaluation

DIA knowledge had been analyzed utilizing DIA-NN (v1.8) [72]. The default workflow for evaluation utilizing a spectral library was adopted (https://github.com/vdemichev/diann). The DIA knowledge from Kitata and colleagues and Guo and colleagues research had been analyzed individually with DIA-NN. FDR (q-value) for protein teams and Hyp web site identification was set at 1.0%. The evaluation of the DIA knowledge from every research was carried out with spectral library from numerous sources: HypDB Library, Research-Particular DDA-based Library, and Mixed Library generated by each HypDB and Research-Particular DDA Evaluation for each Hyp peptide identifications and non-Hyp peptide identifications. DIA-NN additional utilized world q-value filtering and depth normalization to generate Hyp web site matrix output for Hyp websites that had been confidently quantified throughout all samples. Python scripts developed in-house to course of the output from DIA-NN to be Hyp web site nonredundant. The matrix output from every research with nonredundant Hyp web site quantification was used for clustering, annotation enrichment evaluation, and visualization utilizing the Perseus software program platform [73]. Lacking values had been imputed utilizing a traditional distribution, and the information had been hierarchically clustered. The processed site-nonredundant Hyp depth knowledge from DIA-NN was additionally analyzed and visualized utilizing R. Lacking values had been imputed utilizing the k-nearest neighbor (KNN) methodology within the NAguideR instrument [74].

RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular

Recent Comments