Statistical Modeling of C-kit+ Progenitor Cell Extracellular Vesicles to Predict Clinical Trial Outcomes Open Access

Abstract

Congenital heart disease is the most common form of birth defect and affects nearly 1% of live births in the US, annually. Complex forms of congenital heart disease, including hypoplastic left heart syndrome, require surgical palliation and often lead to pediatric heart failure. Originally studied for use in adult populations, cardiac cell therapy is gaining traction for pediatric populations as a therapeutic strategy to address underlying damage and potentially repair and regenerate myocardium. Specifically, our group is involved in the CHILD clinical trial (NCT03406884), investigating the use of autologous cardiac-derived c-kit+ progenitor cells (CPCs). Unfortunately, cardiac cell therapy preclinical and clinical investigations have been hampered by mixed results, primarily too much variation in cell populations and patient outcomes. Finally, there has been a shift in our understanding of how cardiac cell therapy works: functional effects may be attributed to paracrine signaling and the release of extracellular vesicles (EVs), rather than direct cell engraftment and differentiation. The purpose of this dissertation is to understand sources of variability and determine the biological signals contributing to repair in CPCs and CPC-derived extracellular vesicles (CPC-EVs). We analyzed transcriptomic data (bulk and single cell) and used machine learning regression models to link our RNA-sequencing data to functional outcomes. The subsequent dissertation chapters are designed to explore (1) differences between neonate- and child-derived CPCs at the single cell level, (2) bulk transcriptomic differences between patient matched CPCs and CPC-EVs, and (3) the relationship between CPC-EV cargo and in vitro, cardiac-relevant outcomes. Overall, we uncovered a more heterogenous population in child CPCs, enriched in pro-fibrotic and inflammatory cell subpopulations. We determined that CPC-EVs contain different RNA cargo than EVs derived from other cell types, and CPC-EVs are particularly enriched in miRNAs involved in cardiac development and cell proliferation. Finally, we used machine learning models to link CPC-EV RNA cargo from the CHILD trial samples to in vitro functional outcomes for the purposes of building a predictive and informative clinical tool.

Table of Contents

Table of Contents

Abstract         vii

Acknowledgements   ix

Table of Contents      xii

List of Figures and Tables      xv

List of Abbreviations  xvi

1         Chapter 1. Introduction         1

1.1      Congenital heart disease       1

1.1.1   Forms 1

1.1.2   Etiology          3

1.1.3   Hypoplastic left heart syndrome       4

1.2      Cardiac cell therapy for congenital heart disease     7

1.2.1   Stem and progenitor cell types for treatment of congenital heart disease 9

1.2.2   CHILD clinical trial      11

1.3      Extracellular vesicles 13

1.3.1   Biogenesis      15

1.3.2   Structure and function          18

1.3.3   Role of stem/progenitor cell-derived extracellular vesicles in cardiac repair          19

1.4      Machine learning – unsupervised vs. supervised learning   21

1.4.1   Dimension reduction and clustering 22

1.4.2   Weighted correlation network analysis        23

1.4.3   Regularized regression models         24

1.4.4   Partial least squares regression         26

1.4.5   Random forest regression     27

1.5      Research objectives   29

2         Chapter 2: Single cell RNA sequencing reveals distinct c-kit+ progenitor cell populations  31

2.1      Abstract         31

2.2      Introduction   32

2.3      Methods        34

2.3.1   C-kit+ progenitor cell (CPC) culture and expansion  34

2.3.2   Cell sorting of CPC subpopulation     34

2.3.3   Computational Methods       35

2.4      Results 37

2.4.1   Clustering and compositional analysis reveal differences in neonate and child CPCs         37

2.4.2   Trajectory analysis identifies co-expressed genes within CPC populations  38

2.4.3   Cell cluster four is upregulated in cytokines 40

2.4.4   Cell cluster six is upregulated in several fibrosis-associated factors 41

2.4.5   Confirmation of cluster six surface proteins 42

2.5      Discussion      43

2.6      Supplemental Information    47

3         Chapter 3: Comparative computational RNA analysis of c-kit+ progenitor cells and their extracellular vesicles    49

3.1      Abstract         49

3.2      Introduction   50

3.3      Methods        52

3.3.1   Isolation and Culture of c-kit+ Progenitor Cells (CPCs)         52

3.3.2   Extracellular Vesicle (EV) Collection and Characterization   52

3.3.3   Next Generation Sequencing 53

3.3.4   RNA Sequencing Data Analysis          54

3.3.5   Data Mining   56

3.3.6   ceRNA Network Construction 56

3.4      Results 57

3.4.1   Characterization of EVs from neonate, infant, and child CPCs         57

3.4.2   CPCs retain ECM-related RNAs and export signaling pathway-related RNAs to EVs 58

3.4.3   CPC-EVs are enriched in miRNAs involved in cardiac development and cell signaling         62

3.4.4   CPC-EVs contain vesicle biosynthesis and cell cycle-related miRNAs           65

3.4.5   Construction of ceRNA network        66

3.5      Discussion      69

3.6      Supplemental Information    76

4         Chapter 4: Statistical Modeling of Extracellular Vesicle Cargo to Predict Clinical Trial Outcomes for Hypoplastic Left Heart Syndrome  81

4.1      Introduction   81

4.2      Methods        83

4.2.1   Isolation and Culture of c-kit+ Progenitor Cells (CPCs)         83

4.2.2   Extracellular Vesicle (EV) Collection  84

4.2.3   Tube Formation Assay           84

4.2.4   Mesenchymal Stromal Cell (MSC) Migration Assay  85

4.2.5   Fibroblast TGF‐β Stimulation Assay   85

4.2.6   Endothelial Cell TNF-α Stimulation Assay     86

4.2.7   Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)  86

4.2.8   Next Generation Sequencing 87

4.2.9   RNA Sequencing Data Analysis          88

4.2.10 WGCNA gene module detection       88

4.2.11 Regression Models    89

4.3      Results 90

4.3.1   CHILD clinical CPC samples release small EVs in cell culture 90

4.3.2   CPC-EV treatment affects recipient cell processes   93

4.3.3   Weighted gene co-expression network analysis (WGCNA) identifies clusters of co-expressed CPC-EV RNAs which correlate to in vitro outcomes       95

4.3.4   Partial least squares regression models predict CHILD CPC-EV in vitro outcomes   98

4.4      Discussion      100

4.5      Supplemental Information    105

5         Chapter 5: Discussion 108

5.1      Summary of results    108

5.2      Limitations and future directions      109

5.3      Conclusions    116

References     118

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Biological and Biomedical Sciences
Subfield / Discipline	Molecular & Systems Pharmacology
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Biology, Cell Biology, Bioinformatics
Keyword	RNA-sequencing extracellular vesicle cardiac cell therapy machine learning
Committee Chair / Thesis Advisor	Michael Davis, Emory University
Committee Members	Manu Platt, Emory University John Calvert, Emory University Roy Sutliff, National Institutes of Health Samantha Yeligar, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Statistical Modeling of C-kit+ Progenitor Cell Extracellular Vesicles to Predict Clinical Trial Outcomes ()	2022-07-06 11:21:17 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions