Publications

Detailed Information

Metabolomics study to explore pathophysiology and discover diagnostic biomarkers of idiopathic inflammatory myopathy : 대사체학 기반 특발성 염증성 근육병증의 기전 탐색 및 진단 바이오마커 발굴

DC Field Value Language
dc.contributor.advisor조주연-
dc.contributor.author강지현-
dc.date.accessioned2023-11-20T04:43:01Z-
dc.date.available2023-11-20T04:43:01Z-
dc.date.issued2023-
dc.identifier.other000000178110-
dc.identifier.urihttps://hdl.handle.net/10371/197097-
dc.identifier.urihttps://dcollection.snu.ac.kr/common/orgView/000000178110ko_KR
dc.description학위논문(박사) -- 서울대학교대학원 : 의과대학 의과학과, 2023. 8. 조주연.-
dc.description.abstract서 론: 특발성 염증성 근육 병증은 자가면역 질환으로, 여러가지 임상 증상, 치료에 대한 반응, 예후가 다양하게 나타난다. 또한, 근생검 없이는 정확한 진단이 어렵다. 본 연구에서는 특발성 염증성 근육 병증 환자를 진단할 수 있는 대사체 패널을 식별하고 환자의 혈청과 마우스 모델의 근육에서 특발성 염증성 근육 병증 특이적인 대사체 특징을 탐색하고자 하였다.
방 법: 특발성 염증성 근육 병증 환자 50명, 강직성 척추염 환자 30명, 건강인 10명의 혈청과, 마우스 모델의 근육조직 및 혈청을 수집하였다. 모든 시료는 액체크로마토프/질량분석기 기반 표적대사체 분석으로 진행되었고 일부 환자시료와 건강인 시료로 비표적 대사체 분석도 진행하였다. 단변량, 다변량 분석을 통하여 대사체들의 변화를 확인하고 3가지 머신러닝 방법인 로지스틱 회기분석 (LR), support vector machine (SVM), 및 random forest (RF)를 이용하여 특발성 염증성 근육병증을 예측할 수 있는 모델을 구축하였다.
결 과: 37개의 IIM 특이적인 대사체 중 후진 소거법을 이용하여 7개의 예측 바이오마커를 발굴하였다. 구축된 모델은 5-fold 교차 검증을 통해 3가지 머신러닝 방법으로 평가되었다. 각 분류 모델에 대해 ROC curve의 AUC 값은 0.955 (LR), 0.908 (RF) and 0.918 (SVM)로 산출되었다. 추가적으로 특발성 염증성 근육 병증 환자 군에서 옥시리핀이 유의하게 증가한 것을 확인하였다. 마우스 모델 조직 시료에서는, CIM 군에서 총 68개의 대사체가 유의하게 변화하였다. 타우린을 제외한 모든 대사체들은 클래스별로 CIM 쥐에서 동일한 양상으로 변화하였다. 특히, 경로 분석을 통해 근육 염증에 기여하는 가장 영향력 있는 경로는 polyamine 대사 경로와 beta-alanine 대사 경로인 것을 확인하였다.
결 론: 다양한 시료와 대사체방법론을 이용하여 특발성 염증성 근육 병증의 질병 기전이 노화 및 염증 관련 특징이 있는 것을 확인하였고 추가적으로 진단 가능성있는 대사체 바이오마커 패널을 제시하였다.
-
dc.description.abstractINTRODUCTION: Idiopathic inflammatory myopathy (IIM) is a diverse set of autoimmune diseases with various clinical symptoms, responses to treatment, and prognoses. The diagnosis for IIM can be challenging without conducting muscle biopsy. In this study, a panel of metabolites were identified through a metabolomics approach in serum samples for IIM detection. The metabolomic signature of IIM was investigated using both human serum samples and tissue samples from a mouse model.
METHODS: Serum samples from 50 IIM patients, 30 ankylosing spondylitis (AS) patients, and 10 healthy volunteers as well as muscle tissue samples from C-protein-induced myositis (CIM) which is a murine model of IIM were collected. All samples were subjected to a targeted liquid chromatography–mass spectrometry-based metabolomic approach, and part of human serum samples underwent an untargeted metabolomics approach. Three machine learning methods, namely logistic regression (LR), support vector machine (SVM), and random forest (RF), were applied to build prediction models for IIM patients. In addition, univariate and multivariate statistical analyses, as well as pathway enrichment analysis, were performed on serum and tissue samples to identify metabolic alterations.
RESULTS: ANOVA revealed 37 IIM-specific metabolites and a set of 7 predictive metabolites was calculated by backward stepwise selection. The discrimination model for IIM was evaluated within 5-fold cross-validation by using three machine learning algorithms. The model produced area under the receiver operating characteristic curve values of 0.955 (LR), 0.908 (RF), and 0.918 (SVM). Additionally, the analysis of subset of human serum samples revealed a significant increase in oxylipins in the IIM group. In mouse tissue samples, a total of 68 metabolites were significantly changed in CIM mice. The metabolic profiles of CIM mice showed a consistent pattern among all metabolites from various classes, except for taurine. Notably, the polyamine pathway and the beta-alanine pathway were identified as the pivotal pathways implicated in the inflammatory response of muscle tissue in CIM mice.
CONCLUSION: A metabolomics-based approach was employed to identify potential biomarkers of IIM and uncover the relevant metabolic pathways involved in the underlying pathological processes of IIM.
-
dc.description.tableofcontentsIntroduction 1
Methods 4
1. Study participants 4
2. Induction of C protein-induced myositis 6
3. Histological Analysis 7
4. Mass spectrometry-based analyses 8
4.1. Mass spectrometry-based targeted metabolomics analysis 8
4.2. Mass spectrometry-based untargeted metabolomics analysis 9
5. Raw data processing and normalization 12
5.1. Targeted metabolomics raw data processing using MetIDQ 12
5.2. Untargeted metabolomics raw data processing 12
6. Protein Extraction and Western Blot Analysis 14
7. Enzyme-linked immunosorbent assay 16
8. Statistical analysis 17
8.1. Normalized targeted metabolomics data 17
8.2. Untargeted metabolomics data 18
Results 19
1. Clinical characteristics of the participants 19
2. Metabolic profiling of healthy control, AS and IIM patients 22
3. Predictive Biomarker and Machine Learning Algorithm Optimization for Distinguishing IIM 35
4. Bile acid levels in human serum 42
5. Oxylipin profiles from untargeted metabolomic analysis 45
6. Serum metabolites change after corticosteroid treatment 49
7. Metabolic profiling in the C-protein-induced myositis mouse model 52
8. Metabolic pathway associated with the muscle of CIM mice and expression of ODC-1 and SMOX 66
Discussion 72
1. IIM-specific metabolites based on targeted metabolomics analysis 72
2. Diagnostic biomarker discovery of IIM 76
3. Muscle inflammation induces alteration of oxylipins 77
4. Pathophysiology of IIM in mouse muscle tissue 79
5. Study limitations 82
Conclusion 83
References 85
국문 초록 92
-
dc.format.extentix, 93-
dc.language.isoeng-
dc.publisher서울대학교 대학원-
dc.subject특발성 염증성 근육병증-
dc.subject대사체학-
dc.subjectC protein-induced myositis-
dc.subject바이오마커-
dc.subject.ddc610.72-
dc.titleMetabolomics study to explore pathophysiology and discover diagnostic biomarkers of idiopathic inflammatory myopathy-
dc.title.alternative대사체학 기반 특발성 염증성 근육병증의 기전 탐색 및 진단 바이오마커 발굴-
dc.typeThesis-
dc.typeDissertation-
dc.contributor.AlternativeAuthorJihyun Kang-
dc.contributor.department의과대학 의과학과-
dc.description.degree박사-
dc.date.awarded2023-08-
dc.identifier.uciI804:11032-000000178110-
dc.identifier.holdings000000000050▲000000000058▲000000178110▲-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share