Heart disease is the number one cause of death in the United States each year for both women and men. Although significant advances have been made in conventional drug and device therapies in recent years, they have not been able to reverse the loss of functional myocardium. Regeneration of the heart may someday be possible with the ability to derive functional cardiomyocytes from human stem cells. However, before these cells can be used for cardiac repair, more must be known about their electrophysiology and their likelihood to seamlessly integrate with native cardiac tissue. In particular, it is of critical importance to establish the electrophysiological compatibility of these cells with host myocardium to minimize the risk of arrhythmia. Despite this critical need, the classification of the electrophysiological phenotypes has been largely subjective for all cell types that have been studied so far, relying mainly on parameters related to action potential shape. The overall goal of this project is to develop a new, analytical and automated method to classify newly differentiated cardiac cells, based on techniques developed for machine learning.
The specific aims are first, to use optical mapping and microelectrode recordings to generate datasets of functional electrophysiological characteristics of human embryonic stem cell-derived cardiomyocytes (hESC-CMs), and second, to use machine learning techniques to classify the phenotypes of the hESC-CMs, based on parametric descriptions. These techniques will involve linear and nonlinear dimensionality reduction algorithms, and clustering and classification algorithms. Cells at different stages of differentiation will be evaluated at different pacing and pharmacological conditions that will help to establish the functional properties of the cells. In summary, the proposed research will enable the classification of cardiomyocytes that are derived from human stem cells. The ability to classify and identify cardiomyocyte phenotypes will permit a quantitative assessment of the batch-to-batch variability in cell cultures, the effect of different differentiation procedures, and the evolution of phenotypes during cardiomyocyte differentiation and maturation. These are critically important issues for the future clinical application of these cells to regenerate cardiac tissue.

Public Health Relevance

Stem cell therapy holds great potential for treating diseased and failing hearts, but still faces significant challenges. This project addresses the electrophysiological aspects these cells, and the goal is to develop methods that can classify and identify the variety of heart cells that are derived from different kinds of stem cells.

Agency
National Institute of Health (NIH)
Institute
National Heart, Lung, and Blood Institute (NHLBI)
Type
Exploratory/Developmental Grants (R21)
Project #
5R21HL108210-02
Application #
8259042
Study Section
Biodata Management and Analysis Study Section (BDMA)
Program Officer
Buxton, Denis B
Project Start
2011-04-20
Project End
2014-03-31
Budget Start
2012-04-01
Budget End
2014-03-31
Support Year
2
Fiscal Year
2012
Total Cost
$205,000
Indirect Cost
$80,000
Name
Johns Hopkins University
Department
Biomedical Engineering
Type
Schools of Medicine
DUNS #
001910777
City
Baltimore
State
MD
Country
United States
Zip Code
21218
Gorospe, Giann; Zhu, Renjun; Millrod, Michal A et al. (2014) Automated grouping of action potentials of human embryonic stem cell-derived cardiomyocytes. IEEE Trans Biomed Eng 61:2389-95