Cartilaginous Metabolomic Study Reveals Potential Mechanisms of

Subscriber access provided by University of Newcastle, Australia

Article

Cartilaginous metabolomic study reveals potential mechanisms of osteophyte formation in osteoarthritis Zhongwei Xu, Tingmei Chen, Jiao Luo, Shijia Ding, Sichuan Gao, and Jian Zhang J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00676 • Publication Date (Web): 07 Feb 2017 Downloaded from http://pubs.acs.org on February 12, 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 32

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

Journal of Proteome Research

Cartilaginous metabolomic study reveals potential mechanisms of osteophyte formation in osteoarthritis Zhongwei Xu 1, Tingmei Chen 2, Jiao Luo 3, Shijia Ding 2, Sichuan Gao 1, Jian Zhang 1* 1

Department of Orthopaedics, the First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China

2

Key Laboratory of Clinical Laboratory Diagnostics (Ministry of Education), College of Laboratory Medicine,

Chongqing Medical University, Chongqing 400016, China 3

West China School of Public Health, Sichuan University, Chengdu 610041, China

* Correspondence to Prof. Dr. Jian Zhang, Department of Orthopaedics, the First Affiliated Hospital of Chongqing Medical University, No.1 Youyi Road, Chongqing 400016, P. R. China. E-mail: [email protected]

Abstract Osteophyte is one of the inevitable consequences of progressive osteoarthritis with the main characteristics of cartilage degeneration and endochondral ossification. The pathogenesis of osteophyte formation is not fully understood to date. In this work, metabolomic approaches were employed to explore potential mechanisms of osteophyte formation by detecting metabolic variations between extracts of osteophyte cartilage tissues (n=32) and uninvolved control cartilage tissues (n=34), based on the platform of ultra-performance liquid chromatography tandem quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS), as well as the use of multivariate statistic analysis and univariate statistic analysis. The osteophyte group was significantly seperated from the control group by the orthogonal partial least-squares discriminant analysis (OPLS-DA) models, indicating that metabolic state of osteophyte cartilage had been changed. In total, 28 metabolic variations further validated by mass spectrum (MS) match, tandom mass spectrum (MS/MS) match and standards match mainly included amino acids, sulfonic acids, glycerophospholipids and fatty acyls. These metabolites were related to some specific physiological or pathological processes (collagen dissolution, boundary layers destroyed, self-restoration triggered, etc) which might be associated with the procedure of osteophyte formation. Pathway analysis showed phenylalanine metabolism (PI=0.168, p=0.004) was highly correlative to this degenerative process. Our findings provided a direction for targeted metabolomic study and an insight to further reveal the molecular mechanism of ostophyte formation.

Keywords Metabolomics, UPLC-MS/MS, Osteoarthritis, Osteophyte Formation, Endochondral Ossification

1

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

Page 2 of 32

Introduction Osteoarthritis of the knee (OAK) is one of the most common joint disorders in elders 1, 2. Approximately 8% individuals aged over 45 are suffering from symptomatic knee osteoarthritis in China 3. OAK is characterized by the imbalance between the breakdown and repair of articular cartilage, and is involved in several pathological changes including osteophyte formation, synovial inflammation, and subchondral bone sclerosis

4-6

that the chronic, progressive, degenerative process of OAK is also irreversible

7, 8

. The occurrence of osteophyte implies . Therefore the detection of osteophyte

could serve as a pre-radiographic biomarker for OAK development 9. Previous studies had showed that the preservation of osteophyte would result in poor clinical outcomes of several surgeries of the knee such as meniscus allograft transplantation (MAT) and total knee arthroplasty (TKA)

10, 11

, although osteophyte is not associated with knee pain 12, 13.

To date, there are not specific drugs or surgical methods which can prevent the progression of osteophyte formation. Endochondral ossification has been considered as the most important process in the formation of osteophytes

14, 15

.

Mesenchymal cells can proliferate and differentiate into hypertrophic cells, and the latter ones secrete collagen, matrix metalloproteinase and vascular endothelial growth factor, and then angiogenesis and calcification of surrounding matrix lead to the replacement of cartilage by bone 16. Animal experiments showed numerous factors could aggravate osteophyte formation, such as bone morphogenetic protein 2 (BMP2), transforming growth factor beta (TGF-β) and Alarming S100A8/A9 17, 18. However, the mechanisms of this complex process still remain unclear to date. As one of the most important parts of systems biology, metabolomics has been employed to look for biomarkers and study mechanisms of diseases by identifying and quantifying all metabolic variations and analyzing pathways in biological systems

19

. Metabolomics performed based on human blood, urine or diverse tissue extracts can provide a

panoramic view of the physiologic or disease state

20-22

. Ultra-performance liquid chromatography-tandem mass

spectrometry (UPLC-MS) technique, with high sensitivity, good reproducibility of retention time, wide dynamic range,

2


Page 3 of 32

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


and coverage of a wide chemical diversity 23, has been proved appropriate for metabolomic study. In this study, UPLC-MS based cartilaginous metabolomic analysis concerning knee osteophyte was carried out to explore the pathogenesis of osteophyte formation in patients with OAK. We collected cartilage tissues from osteophytes and uninvolved lateral posterior femoral condyle of patients after TKA, identified metabolic disturbances between the 2 kinds of cartilage tissues, and analyzed the probable metabolic pathways relevant to osteophyte formation.

Materials and methods Chemicals and materials

Methanol and dichloromethane were purchased from Tedia (Fairfield, OH, USA); HPLC-grade acetonitrile, formic acid, ammonium acetate, and acetic acid were purchased from Sigma-Aldrich (St.Louis, MO, USA); distilled water was acquired by a Milli-Q Ultra-pure water system (Millipore, Billerica, USA); available reference standards were purchased from MCGBW (MCGBW, Beijing, China). Materials included pressure blowing concentrator MTN-2800D (AUTO&SCIENCE, Tianjing, China), high-throughput tissue grinder SCIENTZ-48(SCIENTZ, Zhejiang, China), knife blades (Jinhuan, Shanghai, China), and rongeurs (Depuy, Johnson Medicine, NJ, USA).

Study participants

The study subjects consisted of 34 postmenopausal women who were all hospitalized in the Department of Orthopedics of the First Affiliated Hospital of Chongqing Medical University between February and August, 2015. The diagnosis of OAK was according to the American College of Rheumatology (ACR) criteria for knee osteoarthritis 24, and the grading of osteophyte was according to histological classification

25

. All of the recruited patients underwent TKA. Cartilage

samples got from osteophytes were assigned to the experimental group (n=32), and those from lateral posterior femoral condyle (LPC) were assigned to the control group (n=34) since LPC is highly similar to healthy cartilage which was 3



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

Page 4 of 32

unavailable in this study 26 (Figure 1f). The leading difference between the 2 kinds of samples was whether the cartilage tissue was involved in endochondral ossification which is the major process of osteophyte formation. Inclusion criteria included osteophytes observed on plain radiography, histologically classified type A-C

25

and more than 100 mg of

cartilage tissue available. Exclusion criteria included traumatic, infectious or rheumatic arthritis accompanied, or any history of metabolic bone diseases such as rickets or Graves’ disease. Two more samples were recruited in control group because there were two osteophyte samples classified type D and then excluded but the paired LPC samples were both eligible. This study was approved by the Ethics Committee of Chongqing Medical University (approval number: 2015-70). Written informed consents were obtained from all participants. The work was in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans.

Samples collection

Resected tibial plateau and LPC were placed on ice immediately. After peeling perichondrium off with a surgical knife blade, cartilage samples (100 mg for each) were quickly taken out from osteophytes or LPC with 2 independent rongeurs and then weighed. Histological analysis (H&E staining) was performed for both osteophytes and LPC. Before metabolite extraction, all of the cartilage samples were transferred into microtubes and preserved in liquid nitrogen at an average temperature lower than -196 degree centigrade 27.

Metabolites extraction

Each cartilage sample was taken out, protected and embrittled by using 5-10-mL liquid nitrogen, and then ground into powder in a mortar. After liquid nitrogen was volatilized, cartilage powder was transferred into a 2-mL microtube rapidly. Extraction steps were as follows

28

: (1) 1 mL methanol-water (1:1) was added to precipitate proteins and dissolve

4


Page 5 of 32

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


metabolites contained within the powder with the aid of a bead-based high-throughput tissue grinder (40Hz, 5min); (2) the mixture was centrifuged at 12,000 rpm for 10 minutes and then separated into supernatant and precipitate which were further transferred into different micro-tubes; (3) 1 mL dichloromethane-methanol (3:1) was added to extract water-insoluble organics of the precipitate; (4) the mixture from step 3 was centrifuged at 12,000 rpm for 10 minutes again; (5) 0.75 mL supernatant from step 2 and another 0.75 mL supernatant from step 4 were mixed and transferred to a 2-mL micro-tube; (6) Supernatants were blow-dried by a continuous flow of nitrogen gas and the solid left was redissolved with 100 µL acetonitrile-water (1:1); (7) the solution was centrifuged at 12,000 rpm for 10 minutes for the third time; (8) 60 µL of supernatant was transferred to an auto-sampler vial; (9) a mixture of supernatants of all samples (10µL for each) was transferred to a quality control (QC) vial. Samples were kept on ice throughout the procedure unless centrifuged. Ratios above were all according to volume.

UPLC-MS/MS analysis

UPLC-QTOF-MS/MS analysis was performed with a Shimadzu UFLC-equipped AB-Sciex Triple TOF 4600 in both positive and negative ionization mode using the Turbo V ESI ion source. Samples were injected into a Kinetex XB-C18 column (100 mm × 2.1 mm, 2.6 µm, Phenomenex) or a Kinetex HILIC column (100 mm × 2.1 mm, 2.6 µm, Phenomenex) with a flow rate of 0.35 mL/min. After 10 µL aliquot of 10 QCs successively injected to adjust the system consistency, 10 µL aliquot of each experimental sample was injected to analyze in turn. After each 5 samples injected, one QC was used to estimate the system reproducibility, and one blank (prepared by using identical protocols in “extracting” ultra-pure water instead of experimental tissues) was used to flush the column

29

. The ion spray voltage was set to 5500 V for

positive ion and -4500 V for negative ion mode. The nebulizer gas (air) and turbo gas (air) were set to 55 psi, and the heater temperature was 600°C. The curtain gas (nitrogen) was set at 25 psi, and the rolling collision energy (CE) was set at (40 ± 15) V for positive ion mode and (-40 ± 15) V for negative ion mode. The mobile phase for reversed phase liquid

5



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

chromatography (RPLC) methods consisted of 0.1% formic acid in purified water and 0.1% formic acid in purified acetonitrile, and the latter was changed by time (held constant at 5% for 1 minute, increased from 5% to 85% by 8 minutes, held constant at 85% by 12 minutes, decreased to 5% by 12.1 minutes, and held constant at 5% by 15 minutes). The mobile phase for hydrophilic interaction liquid chromatography (HILIC) methods consisted of acetonitrile and ammonium acetate/acetic acid, and the latter was changed by time (held constant at 5% for 1 minute, increased from 5% to 30% by 2 minutes, increased from 30% to 40% by 10 minutes, returned to 5% by 12.1 minutes, and held constant at 5% by 15 minutes). Full scan analysis was performed in the electrospray ionization mass spectrometry mode using electrospray ionization technique with coverage of mass ranged from 50 to 1000 Da by using scan rate of 0.25 second, and the MS/MS screening was accomplished in the combinational mode of Information Dependent Acquisition (IDA) with a scan rate of 0.1 second. To reduce inter-day errors, feature capturing of samples in one mode was performed continuously, and QC samples were analyzed throughout all analytic days.

Data analysis

All of the data from UPLC-QTOF-MS/MS were imported to MarkerView software (version 1.2.1, ABScix) for peak finding, alignment, filtering, and numeric data exporting. The 80% rule

30

was used to remove missing values, and the

remaining data described by mass-to-charge ratio (m/z), retention time (RT), and peak intensity were imported to SIMCA-P software (version 13.0, Umetrics) for multivariate statistical analysis. After pareto scaling, an overview of the 2 groups was offered by using unsupervised principal component analysis (PCA). For further analysis, supervised orthogonal partial least-squares discriminant analysis (OPLS-DA) was carried out to discriminate features with differences in peak intensity between the 2 groups. R2 and Q2 of OPLS-DA models, both scoring more than 0.5 accepted, were adjusted repeatedly to obtain the most appropriate and predicable models and to avoid chance findings to the most extent

31

. In addition, the 7-fold cross-validation was performed to avoid overfitting of the models, and the permutation

6


Page 6 of 32

Page 7 of 32

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


test (999 permutations) was further performed to validate the OPLS-DA models using SIMCA-P 14.1 software

31

.

Variable important in the projection (VIP) values of all features were exported from the validated OPLS-DA models. Furthermore, signed-rank test was performed to ensure the univariate statistical significance (p < 0.05) of the changed features between the 2 groups by SPSS 17.0 software (IBM, NY, USA) with the 2 unpaired controls excluded 32. Features with both VIP > 1 and p < 0.05 were considered significantly different between osteophyte group and LPC group 33-35.

Metabolite identification

Mass spectrum (MS) data of features with VIP >1 and p < 0.05 were primarily matched by searching Human Metabolome Database (HMDB, http://www.hmdb.ca/), and then the data of tandom mass spectrum (MS/MS) exported from MarkerView software were uploaded to Massbank (http://www.massbank.jp) for library standards match. Mol files from HMDB were downloaded for MS/MS match by using Peakview software (version 1.2, Umetrics) as well

36, 37

. Several

features with both MS and MS/MS matched were further putatively identified by comparing features of tentative metabolites with those of available authentic standards (neat compounds) run under the same experimental conditions. Receiver operating characteristic (ROC) curve analysis and discriminatory plots were performed by using biomarker analysis at MetaboAnalyst (http://www.metaboanalyst.ca/). Pathway analysis including all metabolomic variations was performed by searching KEGG database (http://www.kegg.jp/) and MetaboAnalyst.

Results

Clinical characteristics

All of the 34 subjects were women and underwent TKA. The average age of the patients was 65.88 ± 9.45, and the average duration of knee pain was about 9 years. Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels of all patients were within the normal range before surgery. Celecoxib Capsules (100 mg) and Paracetamol and

7



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

Tramadol Hydrochloride Tablets (325 mg + 37.5 mg) were used for preemptive analgesia. Radiographs of patients’ knees, photographs of sampling sites, and histological sections of cartilage tissues were shown in Figure 1.

Stability of UPLC-MS/MS

The stability and repeatability of the instruments were assessed by using identical QC samples throughout the process of experimental samples injection. PCA analysis was performed to evaluate the variation of QC samples which were supposed to show identical metabolic profiles under same experimental condition. In Figure 2, all QC samples during capturing features of experimental samples were included to perform PCA (Fig. 2a, intra-day QCs; Fig. 2b, inter-day QCs). All of QC samples clustering together indicated a satisfactory stability of the instruments. Besides, 7.13% was the maximum value of relative standard deviations (RSDs) of a group of QC samples in a specific ion mode of 4 modes (7.13% in HILIC (-) ion mode). Both results above indicated the data were reliable and reproducible.

Statistical analysis

There were 8,313 features captured in total. After the 80% rule used, there were 156 features left in HILIC negative mode, 225 features in HILIC positive mode, 67 features in RPLC negative mode, and 72 features in RPLC positive mode. Multivariate analysis including PCA and OPLS-DA was performed to recognize metabolic variations of the two groups. In Figure S1, PCA for 4 modes provided an overview that unsupervised analysis did not distinguish metabolic variations adequately between the 2 groups, but a moderate intra-group clustering was observed. In Figure 3, osteophyte group was separated from control group by OPLS-DA models which were characterized by clustered intra-group samples and scattered inter-group ones. With R2 and Q2 both scoring more than 0.5, OPLS-DA models of the 4 modes were proved appropriate and predicable. Further permutation tests for the 4 models indicated that they were all valid (Figure S2). Subgroup analysis showed that no marked difference was found among osteophyte samples of different histological

8


Page 8 of 32

Page 9 of 32

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


grades (Figure S3). Subsequently, features with VIP>1 were exported. Signed-rank test was carried out to delete features without univariate statistical significance (p1 and p0.1 (enrichment analysis) and raw p

Cartilaginous Metabolomic Study Reveals Potential Mechanisms of

Recommend Documents