The differentiation of human pluripotent stem cells into cardiomyocytes is now relatively commonplace, and a heterogeneous mixture of electrophysiological phenotypes is generally obtained. The relative proportions of the 3 major phenotypes (nodal-like, atrial-like and ventricular-like) have been widely reported for various differentiation methods and cell lines, and is crucial information in determining their suitabilityfor any given application. However, such assessments are generally obtained by subjective evaluation, and in the few instances where quantitative measures have been used, classification was based on ad-hoc criteria using only a few parameters. To significantly improve the accuracy and reproducibility of the phenotypic analysis, we will utilize the entire action potential shape and employ advanced image analysis and machine learning methods in an automated computational algorithm. This algorithm will be used to classify the electrophysiological phenotypes of human embryonic stem cell-derived cardiomyocytes (hESC-CMs) grown in culture, and optimized by using molecular identifiers.
The specific aims of the project are to: (1 obtain archetypal electrophysiological datasets of human ventricular, atrial and nodal cells, and (2) use image analysis and machine learning methods to classify hESC-CM phenotypes. To achieve these aims, we will use computational models of the 3 major subtypes of the heart as reference points for our classification algorithm, multi-site optical measurements of action potentials, molecular and pharmacological indicators of subtype, and metamorphosis distance as a quantitative measure of the dissimilarity between action potentials. In summary, this project will result in a standardized, objective and automated computational method to classify the relative amounts of the 3 major electrophysiological phenotypes in populations of hESC-CMs, as well as the development of an action potential ontology.

Public Health Relevance

Exciting progress has been made in the creation of human cardiomyocytes from stem cell sources, and they are known to develop into the major subtypes found in the heart. However, quantitative and automated measures of their relative proportions are lacking, and will be developed in this project using optical measurements and advanced computational algorithms. This information is vital in determining the suitability of the cells for any given application.

Agency
National Institute of Health (NIH)
Institute
National Heart, Lung, and Blood Institute (NHLBI)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21HL122881-01A1
Application #
8824780
Study Section
Biodata Management and Analysis Study Section (BDMA)
Program Officer
Wang, Lan-Hsiang
Project Start
2014-11-15
Project End
2016-10-31
Budget Start
2014-11-15
Budget End
2015-10-31
Support Year
1
Fiscal Year
2015
Total Cost
$243,000
Indirect Cost
$93,000
Name
Johns Hopkins University
Department
Biomedical Engineering
Type
Schools of Medicine
DUNS #
001910777
City
Baltimore
State
MD
Country
United States
Zip Code
21218
Zhu, Renjun; Millrod, Michal A; Zambidis, Elias T et al. (2016) Variability of Action Potentials Within and Among Cardiac Cell Clusters Derived from Human Embryonic Stem Cells. Sci Rep 6:18544
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