Detergent-insoluble proteome analysis revealed ... - ACS Publications

Subscriber access provided by LAURENTIAN UNIV

Article

Detergent-insoluble proteome analysis revealed aberrantly aggregated proteins in human preeclampsia placentas Wanling Zhang, Xing Chen, Ziqi Yan, Yang Chen, Yizhi Cui, Bingjun Chen, Chujun Huang, Weiwen Zhang, Xingfeng Yin, Qing-Yu He, Fang He, and Tong Wang J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.7b00352 • Publication Date (Web): 01 Oct 2017 Downloaded from http://pubs.acs.org on October 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Detergent-insoluble proteome analysis revealed aberrantly aggregated proteins in human preeclampsia placentas

Wanling Zhang†1,2,3, Xing Chen†1, Ziqi Yan1, Yang Chen1, Yizhi Cui1, Bingjun Chen2, Chujun Huang2, Weiwen Zhang4, Xingfeng Yin1, Qing-Yu He1, Fang He*2,3, Tong Wang*1

1

Key Laboratory of Functional Protein Research of Guangdong Higher Education Institutes,

Institute of Life and Health Engineering, College of Life Science and Technology, Jinan University, 601 Huangpu Avenue West, Guangzhou, Guangdong 510632, China

Key Laboratory for Major Obstetric Diseases of Guangdong Province, 3Department of

2

Obstetrics and Gynecology, 4Department of Laboratory Medicine, The Third Affiliated Hospital of Guangzhou Medical University, 63 Duobao Road, Guangzhou, Guangdong 510150, China

1

ACS Paragon Plus Environment

Page 2 of 40

Page 3 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT

Preeclampsia (PE) is a placenta disease, featured by hypertension, proteinuria and other multi-organ dysfunctions, and its etiology is unclear. We and others have shown that intensive ER stress and unfolded protein response (UPR) occur in the PE placenta. In this study, we isolated detergent-insoluble proteins (DIPs) from human placenta tissues, which were enriched with protein aggregates, to characterize the placenta UPR in PE. With data-independent acquisition (DIA) mass spectrometry, we identified 2066 DIPs across all normal (n=10) and PE (n=10) placenta samples, among which 110 and 108 DIPs were significantly up- and down- regulated in PE, respectively. Per clustering analysis, differential DIPs could generally distinguish PE from normal placentas. We verified the MS quantitation of endoglin and vimentin by immunoblotting. In addition, we observed that PE placenta tissues have remarkable more endoglin the cytoplasm. Furthermore, we found that DIPs were evenly distributed across different chromosomes and could be enriched in diversified gene ontology terms, while differential DIPs avoided to distribute on X-chromosome. Significantly up-regulated DIPs in PE were focusing on the top functions of lipid metabolism, while 23 of these DIPs could form the top network regulating cellular movement, development, growth and proliferation. Our results implicate that human PE placenta have disease-relevant differential DIPs, reflecting aberrantly aggregated proteins of placental tissues. The mass spectrometry proteomics data have been deposited to ProteomeXchange consortium with the data set identifier PXD006654, and iProX database (accession number: IPX0000948000).

2



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

KEYWORDS: Detergent-insoluble proteins, preeclampsia, unfolded protein response, B/D-HPP, C-HPP.

3


Page 4 of 40

Page 5 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION

As a leading cause of maternal and neonatal mortality and morbidity, preeclampsia (PE) is prevalent in 3-5% pregnancies world-wide (1, 2). This disease is largely characterized by hypertension, proteinuria and other sever multi-organ dysfunctions, such as the hemolysis, elevated liver enzymes, and low platelet count (HELLP) syndrome (3, 4). For fetus, PE is also a major threat that is associated with fetal growth restriction, and preterm birth. In addition, Herms et al found that at 2.5 year post-delivery, 30% of the women who had PE had hypertension, and 25% had metabolic syndrome (5).

It is generally accepted that PE is a placenta disease, as most symptoms can be alleviated after the placenta is removed post-delivery (1). But, the etiology of PE is still unclear. It has been shown that PE-associated genes can be highly enriched in the regions of chromosomes 6p, 9q, 11p and 19q (6). In addition, genetic variations, such as coagulation factor V gene polymorphism rs6025 and prothrombin gene mutation G20210A, have close relevance to severe PE (7).

As a phenotypic explanation, the ischemic-reperfusion derived oxidative stress represents another important cause of PE (8). When placenta extravillous cytotrophoblast (EVT) cells fail to invade the spiral artery, high-resistant vessels will be formed, which causes dysfunction of maternal circulation of the placenta (9, 10). Consequently, endoplasmic reticulum stress will be initiated that leads to PE symptoms (11, 12). In this regard, we have shown that human PE placentas have significantly up-regulated molecular chaperons and almost all endoplasmic reticulum stress sensors, including inositol-requiring enzyme-1 (Ire-1), protein 4



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

kinase RNA-like endoplasmic reticulum kinase (Perk), activating transcription factor-6 (ATF6) and translation initiation factor 2A (eIF2α), which are accompanied by apoptotic cell death (13).

Such a typical endoplasmic reticulum stress implicates that unfolded protein accumulation and the unfolded protein response (UPR) exist in PE placentas (14). Favorable to this notion, Buhimschi et al have found that misfolded serine protease inhibitor A1 (SERPINA1) aggregates are exclusively deposited in the perivascular of human PE placentas (15), and urine samples (16). With a mouse model, Kalkunte et al identified the transthyretin aggregate as an apoptosis inducer of the PE placental cells (17). Hence, chromosome-based and/or biology-based protein findings of the proteome carried by protein aggregates should be

of closely clinical relevance to define the PE UPR in a systems biology view.

To enrich the protein aggregates, we have developed a method to isolate the detergent-insoluble protein (DIP) fraction of cell lysates from cancer cell lines (18). In the framework of Chromosome-Centric Human Proteome Project (C-HPP), such a strategy has led to the discovery of 23 missing proteins (18), which are protein evidence 2-4 (PE2-4) proteins as defined by the neXtProt knowledgebase (19-22). Of these 23 missing proteins, 10 have survived the standardized reanalysis by PeptideAtlas and neXtProt and are now ranked as PE1 proteins. In addition, we found that DIPs in cancer cells were enriched on chromosome 2 and 12, and tended to be cell/tissue type specific (18). This led us to hypothesize that DIPs in the placenta can also differentiate PE from normal conditions due to their biology/disease relevance. Addressing this question is of potential importance to the 5


Page 6 of 40

Page 7 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


field of PE as the proteome level evaluation of unfolded protein rich subcellular fraction will provide novel views on the UPR features in PE placenta. Therefore, in this study, we used data-independent acquisition (DIA) mass spectrometry to analyze the DIP fraction of human normal and PE placentas. With bioinformatics, chromosome-centric analysis and biological verifications, we provided evidence favorable the clinical importance of placental DIPs to feature PE by aberrantly aggregated proteins.

MATERIALS AND METHODS Human placental tissue samples The scientific and ethics review committee of The Third Affiliated Hospital of Guangzhou Medical University approved this study, and written informed consents were obtained from all participants. PE was diagnosed according to the criteria published by the International Society for the Study of Hypertension in Pregnancy (ISSHP) in 2014 (4). The Inclusion criteria included, 1) de novo hypertension (blood pressure > 140/90 mmHg) after 20 weeks gestation; 2) proteinuria (>300 mg/24 h); 3) other systemic dysfunctions, such as glomerular endothelial injury, elevated liver enzymes and fetal growth restriction. Exclusion criteria included: 1) maternal age > 42 yr; 2) history of recurrent early history of recurrent early pregnancy loss; 3) drug abuse; 4) pre-gestational diabetes; 5) HIV-1, Hepatitis B/C infection or other sexually transmitted diseases; 6) congenital thrombophilia. The placenta tissue samples were acquired via biopsy upon delivery at the Third Affiliated Hospital of Guangzhou Medical University. For each biopsy, ~20 g tissue was dissected from the central area of the placenta, while the calcified area was avoided. All samples were rinsed with sterile 6



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

0.5% saline and aliquoted into ~1 g tissue per tube for the -80 °C storage until use. In total, we collected 17 normal and 15 PE placenta tissue samples. Clinical information of all participants and placenta samples used for different experiments could be found in Supplementary Table S1.

DIP and total protein extraction from placental tissues For DIP extraction, ~1 g placental sample was washed by pre-chilled PBS at 4°C, and treated with a hypotonic buffer [10 mM Tris-HCl, 1.5 mM MgCl2, 10 mM KCl, (pH 7.3)] supplemented with protease inhibitor (PI) cocktail (cOmplete™ mini tablet, Roche, Shanghai, China; final concentration, 1 tablet in 10 mL solution) and 1 mM phenylmethanesulfonyl fluoride (PMSF) at 10 mL per tube. Placental tissues were then sheared by surgical scissors on ice, followed by 5 min homogenization and 30 min incubation on ice. Tissue lysates were subjected to debris removal by centrifugation (2,000 × g, 4 °C, 15 min), and the supernatant, which contained the cytoplasmic proteins fraction, was collected. The cytoplasmic protein fraction was mixed with the cold Buffer A [10 mM tripotassium phosphate, 1 mM EDTA•2Na, 2% (v:v) PI cocktail and 1 mM PMSF, pH 6.5] at a volume ratio of 1:3. With 30-min centrifugation at 15,000 × g at 4 °C, DIPs were pelleted, while the soluble proteins (SP) fraction remained in the supernatant. The pellet was resuspended with 500 µL Buffer A under sonication (40 w, 20 s on ice), and re-precipitated by another 30-min centrifugation (15,000 × g, 4 °C). To purify the DIPs, the pellet was resuspended by 500 µL of Buffer A under sonication; after the addition of 2% (v:v) NP-40, DIPs were re-pelleted with centrifugation. Such a wash step was repeated twice to obtain the purified DIPs in the 7


Page 8 of 40

Page 9 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


pellet. DIPs were then resuspended by a lysis buffer [7 M Urea, 2 M Thiourea, 2% (v:v) CHAPS, supplemented with cOmplete™ PI cocktail and 1 mM PMSF] for fresh use. Protein concentration was determined with a Bradford protein assay kit (Beyotime, Nanjing, China) according to the manufacturer’s instructions. For the total protein extraction, tissues were rapidly homogenized with a SDS lysis buffer (Beyotime), supplemented with cOmplete™ PI cocktail and 1 mM PMSF, by a high-speed benchtop homogenizer FastPrep®-24 (MP Biomedicals, Irvine, California, USA). Tissue lysates were centrifuged at 17,000 × g at 4 °C for 30 min. Supernatants were collected as the total protein fraction. The protein concentration was determined by a BCA kit (ThermoFisher Scientific, Shanghai, China).

Immunoblotting analysis The immunoblotting (IB) analysis was performed as we previously reported with minor modification (18, 23). In this study, we used the BlotTM Mini Gel Tank system with the Mini Blot Module (Thermo Scientific). Precast polyacrylamide gels of Invitrogen NuPAGE 10% Bis-Tris Gels (1.5mm × 15well, Thermo Scientific) were employed. Equal amounts of proteins were analyzed by IB, and Coomassie Brilliant Blue stained gels were used as loading controls. The primary antibodies and their working dilution are listed as follows: mouse anti-human vimentin mAb (1:5000; Santa Cruz, Shanghai, China, Cat: sc-32322), rabbit anti-human BIP mAb (1:1000; Cell Signaling Technology, Shanghai, China, Cat: #3177), rabbit anti-human γ-tubulin pAb (1:1000; CST, Cat: #5886), rabbit anti-human endoglin mAb (1:1000; CST, Cat: #4335). HRP-conjugated secondary Abs included goat anti-mouse IgG 8



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(1:2000; Bioworld, Nanjing, China, Cat: BS12478) and goat anti-rabbit IgG (1:2000; Bioworld, Cat: BS13278). The IB quantitation was performed by using the Quantity One software version: 4.6.2 (Bio-Rad Laboratories, Guangzhou, China). Here, we used the volume tools with the local background subtraction method for volume adjustment.

Protein digestion The protein digestion was performed as we previously described (18, 23-25). Briefly, the DIPs were reduced (8 M urea and 50 mM DTT at 37 °C, 1h) and alkylated (100 mM IAA, at room temperature, 30 min) in 1.5 mL tubes. Samples were then transferred into 30 kDa ultracentrifugal filters (Sartorius Stedim Biotech, Shanghai, China; Cat: UFC5030BK) that were fitted on top of 1.5 mL collection tubes. Proteins were then mixed with trypsin at a mass ratio of 30:1, and digested for 8 h at 37 °C. Finally, peptides were centrifuged to the collection tube, and dried with a cold-trap speed vacuum.

Data-independent acquisition MS The tryptic peptide was reconstituted by deionized water containing 0.1% (v:v) formic acid. A pooled tryptic peptide mixture for the ion library construction was prepared by collecting the same amount of tryptic peptides from all samples. To allow retention time calibration, the iRT-standard provided by the iRT-Kit (Biognosys, Schlieren, Switzerland) was added into the pooled sample at 1/10 by volume. Samples were then analyzed in a data-dependent acquisition (DDA) mode by the LC-MS/MS, equipped with an EASY-nLC 1200 (Thermo) HPLC system and Orbitrap Fusion Lumos (Thermo) mass spectrometer. For LC seperation, 9


Page 10 of 40

Page 11 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


tryptic peptides were sequentially injected into an AcclaimTM PepMapTM 100 C18 column (100µm × 2cm, 5 µm, Thermo, P/N：164564) and an AcclaimTM PepMapTM 100 C18 column (50µm × 15cm, 2 µm, Thermo, P/N: 164943). Mobile phases included Buffer A [0.1% (v:v) formic acid] and Buffer B (80% acetonitrile with 0.1% formic acid). With a flow rate of 300 nL/min, the gradients included: 15 min equilibration with 100% of Buffer A; Buffer B, 4-6%, 0-4 min; Buffer B, 6-28%, 4-131 min; Buffer B, 28%-90%, 131-146 min; wash with 90% Buffer B, 4 min.

Four runs were performed to construct the precursor ion library in DDA mode. Global setting included: ion source type, nanospray ionization; spray voltage, 2.1 kV (positive); ion transfer tube temp, 320 °C; default charge state, 2. MS parameters: detector type, orbitrap; orbitrap resolution, 60,000 at 400 m/z; mass range, normal; scan range, 400-1500 m/z; S-lens RF level, 30%; AGC (automatic gain control) target, 4.0e5; maximum injection time, 50 ms; include charge state(s), 2-7; data dependent mode, top speed; dynamic exclusion, 90 s with a 10 ppm tolerance around the selected precursor. 3) MS/MS parameters: signal intensity threshold for parent ion selection, 5.0e4; MS/MS isolation mode, quadrupole; isolation window, 1.6; activation type, HCD; HCD collision energy, 30%; detector type, orbitrap; orbitrap resolution, 15,000 at 400 m/z; AGC target, 5.0e4; maximum injection time, 30 ms.

For the DIA MS analysis, individual tryptic peptide samples were mixed with the iRT-standard (1/10 by volume), and analyzed by the same LC-MS/MS system. The parameters are listed as follows. 1) Global settings: ion source type, nanospray ionization; spray voltage, 2.1 kV (positive); ion transfer tube temp, 320 °C; default charge state, 2; 10



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

full-MS scan in the scan cycle included 37 of MS/MS scans. 2) MS parameters: detector type, orbitrap; MS resolution, 60,000 at 400 m/z; mass range, normal; scan range, 350-1200 m/z; S-lens RF level, 30%; MS AGC target, 4.0e5; maximum injection time, 30 ms. 3) MS/MS parameters: MS/MS isolation mode, quadrupole; activation type, HCD; HCD collision energy, 30%; detector type, orbitrap; MS/MS resolution, 30,000 at 400 m/z; mass range, normal; scan range, 200-2000 m/z; MS/MS AGC target, 5.0e5; maximum injection time, 50 ms; enabling of inject for all parallelizable time, no; enabling of MS/MS centroiding, no; the MS/MS isolation windows could be found on Supplementary Table S2.

All MS raw data are available in both iProX (accession number: IPX0000948000), and ProteomeXchange consortium (identifier: PXD006654).

Database searches We employed the Sequest HT engine of the Proteome Discoverer v2.1 (Thermo) to search the DDA raw files against a combined database of neXtProt (the 2017-01-23 release, 20159 entries; reverse sequence was used to generate the decoy database) and the iRT standard peptides sequence. Common contaminants in the database included trypsin and keratins. Search parameters included: MS tolerance, 10 ppm; number of trypsin missed cleavage: 2; Keil rule, applied; Fragment Mass tolerance, 0.02 Da; enzyme, trypsin; static modifications, carbamidomethyl (Cys, +57.021 Da); dynamic modifications, oxidation (+15.995 Da) of methionine, deamination (+0.984 Da) of Gln and Asn, and acetyl (+42.011 Da) of the N-terminus. Confident protein identifications should meet the following criteria: 1) protein 11


Page 12 of 40

Page 13 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


level FDR ≤ 1%; 2) unique peptides ≥ 2; 3) peptide length ≥ 9 aa. We next used scaffold software version 4.7.1 (Proteome Software, Portland, OR) for controlling the PSM, peptide and protein level FDR < 1% as we previously described (25). For the DDA MS, the FDR at different levels, and the expected true and false positives were included in Supplementary Table S3.

To build the ion spectral library, we imported the PD software generated pdResult files into the Spectronaut™ software v10 (Biognosys); the parameters of BGS Factory Settings in Load Experiment. Details are listed as follows. 1) calibration: calibration mode, automatic; iRT calibration strategy, non-liner iRT calibration. 2) Identification: p-value estimator, kernel density estimator; precursor q-value cutoff, 0.01. 3) Workflow: default labeling type, label; profiling strategy, none; unify peptide peaks, false. 4) Quantification: interference correction, true; major (protein) grouping, by protein-group id; minor (peptide) grouping, by stripped sequence; only proteotrypic peptides, false; minor group top n, true; min, 1; max, 3; minor group quantity, average precursor quantity; major group top n, true; min, 1; max, 3; major group quantity, average peptide quantity; quantity MS-level, MS2; quantity type, area; data filtering, q-value; cross run normalization, true; row selection, q-value sparse. 5) Reporting: scoring histograms, true; generate SNE File, false; pipeline report schema, protein quant; pipeline reporting unit, experiment. 6) Protein inference: protein inference workflow, automatic. 7) Data extraction: MS1 mass tolerance strategy, dynamic; correction factor, 1; MS2 mass tolerance strategy, dynamic; correction faction, 1. 8) Post analysis: differential abundance grouping, major group (quantification settings); smallest quantitative unit, 12



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

precursor ion (summed fragment ions); use top n selection, false. 9) XIC extraction: XIC RT extraction window, dynamic; correction factor, 1. To analyze individual DIA MS files, the DIA raw files were converted into HTRMS files, and analyzed with the review section of Spectronaut by choosing the matched database fasta file and spectral library. In the report section of Spectronaut, protein report files containing the protein identification and quantitation information were exported for further analyses.

Physical-chemical property comparisons The physical-chemical property analysis was performed as we previously described with minor modifications (18, 23). In brief, we obtained the DIP and PE1 protein sequences from neXtProt. The protein length in amino acids, the isoelectric point (pI), and instability index (26) of each protein were computed with the MATLAB 2015a software (MathWorks, Natick, Massachusetts). The bootstrap analysis was performed by comparing the DIPs with the PE1 proteins.

Chromosome enrichment analysis

We followed our published procedures for this analysis (18, 25). The gene chromosome location was obtained from neXtProt (2017-01-23 release). Chromosomal distribution of DIPs and differential DIPs were statistically compared with the background (PE1-4 proteins), using the two-tailed Fisher’s exact test.

13


Page 14 of 40

Page 15 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Gene Ontology analysis

ClueGO + CluePedia (version 2.3.3 and 1.3.3) (27) was used for the gene ontology (GO) analysis, as we previously described with minor modification (25, 28). The parameters for biological process (BP) GO enrichment included: Ontologies Biological Process (BP; date: 23.02.2017); evidence code, ALL_Experimental_(EXP, IDA, IPI, IMP, IGI, IEP); GO term fusion, applied; pathway p-value cut-off, P ≤ 0.05; and p-value correction algorithm, Benjamini-Hochberg. The parameters for the cellular component (CC) GO analysis included: Ontologies Cellular Component (CC; date: 23.02.2017); evidence code, ALL; GO term fusion, applied; pathway p-value cut-off, P ≤ 0.05; and p-value correction algorithm, Benjamini-Hochberg. The parameters for the Wikipathway GO analysis included: Ontologies WikiPathway (date: 01.03.2017); GO term fusion, applied; pathway p-value cut-off, P ≤ 0.001; and p-value correction algorithm, Benjamini-Hochberg.

Ingenuity pathway analysis (IPA)

The dataset containing gene names and the mean fold changes of scientifically up-regulated DIPs was loaded into IPA (www.ingenuity.com, QIAGEN, Shanghai, China). A core analysis was then performed as we described previously (25, 28-31). When analyzing nodes with a certain molecular and cellular functions, nodes and functions were imported in the My Pathway module; and the Connect tool was used to build networks.

14



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Immunohistochemistry (IHC)

IHC was performed as we previously described with minor modifications (30). In brief, 20 sections (10 from normal and 10 from PE placentas) were sequentially treated with the primary mouse anti-endoglin mAb (1:100, CST, #14606), and the EnVisionTM Flex Mouse High Ph (Link) detection kit (Agilent, Beijing, China; Cat: K8002) was used for secondary antibody staining and visualization.

Statistics

We used the power law global error model (PLGEM) (32) to determine significantly differential DIPs as we previous described (28, 33). For inter-group comparisons, Student’s t-test or Kolmogorov-Smirnov test (KS-test) were used based on the data normal distribution test with the Jarque-Bera test (JB-test). Data were shown as mean ± s.e.m. Statistical significance was accepted when P < 0.05. GraphPad Prism version 6.0 (GraphPad Software, Inc., San Diego, CA, USA) and MATLAB 2015a software were used. Logistic regression analysis was performed by using the SPSS software version 22.0 (IBM, New York, USA).

RESULTS MS identifications of placental DIPs and SPs For MS analysis, we used 10 normal and 10 PE placental tissues. The DIA MS analysis was following the workflow shown in Figure 1. The clinical features of the participants could be 15


Page 16 of 40

Page 17 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


found in Table 1. Specifically, both systolic and diastolic blood pressures of the PE subjects were significantly higher than the normal group, and their urinary protein levels were all greater than 300 mg/24 h (Table 1). PE women had significantly less gestation weeks and fetal birth weights (Table 1). In addition, the peripheral blood uric acid (UA) level was also significantly higher in the PE women (Table 1). With the univariate logistic regression analysis on the clinical features shown in Table 1, we found that that none of them had statistical significance to the binary outcome of PE or Normal (Supplementary Table S4). When using multivariate logistic regression, no model could be fitted with these variables. These results suggested that these clinical features of participants had no significant confounding effects on our proteomic analyses.

Post extraction of the DIP and SP fractions from the placenta, we evaluated the purity of DIPs by our previously reported method (18). We found that the DIPs and SPs had remarkable different protein band distribution patterns, especially at the 50-75 kDa (SP-rich) and 37-50 kDa (DIP-rich) regions (Fig. 2A). We next used IB to determine the expression difference of 78 kDa glucose-regulated protein (Bip), vimentin and γ-tubulin between DIPs and SPs, according to the following rationales. It is known that Bip is a molecular chaperon that binds to the exposed hydrophobic portion of mis-/unfolded proteins. (34). When protein aggregate forms, hydrophobic regions tend be blocked by each other per aggregation effects. When comparing with soluble proteins that have the same amount of exposed hydrophobic regions, we should expect that less Bip will bind to the protein aggregates. In addition, vimentin-cage is an important structure to form aggresomes (35), while γ-tubulin is a major component of centrosome and microtubule (36), as well as autophagosomes and ubiquitinated 16



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

conjugates (37). With IB, we observed that Bip was enriched in the SP fraction, while the DIP fraction had considerably more vimentin and γ-tubulin (Fig. 2B). Hence, these results suggested DIPs were extracted from human placenta tissues with acceptable purity.

When normalized to the cytoplasmic protein fraction, we found that the relative abundance of DIPs had no statistical difference between the normal and the PE groups (Fig. 2C). With DIA MS analysis, we identified 2162 DIPs from all normal subjects, and 2100 DIPs from all PE subjects, among which 2066 DIPs could be identified across all 20 placental tissues (Fig. 2C). The protein quantification values output by the Spectronaut software were included in Supplementary Table S5. All MS2 spectrum-level quantification information for DIPs is included in Supplementary Table S6. Although the median difference was minor, the PE samples had significantly less DIP protein number than the normal samples (Fig. 2E). As compared with the background PE1 proteins, we found that placental DIPs tended to have significantly longer in aa length, and more acidic and stable (Fig. 2F).

Differential DIPs are PE-relevant

We previously found that PLGEM algorithm worked properly in the SWATH MS analysis, which was also a DIA MS (33). We here found that DIP abundances quantified by DIA MS and their standard deviation (SD) could be linearly fitted by PLGEM either, with the adjusted r2 value of 0.993 and Pearson r of 0.927 (Fig. 3A). The residuals distributed evenly and were independent from the rank of mean abundances (Fig. 3B). In addition, we observed that the

17


Page 18 of 40

Page 19 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


residuals followed a normal-like distribution (Fig. 3C). These results showed that PLGEM could nicely fit the placental DIP abundances.

Per PLGEM, we determined 110 and 108 significantly up- and down- regulated DIPs in the PE placentas, respectively (P

Detergent-insoluble proteome analysis revealed ... - ACS Publications

Recommend Documents