The findings do not limit the utility of the cells for some types of research or for some future clinical applications, the researchers say, but draw attention to the need to closely monitor stem cell lines for genetic changes and to study how these alterations affect the cells' behavior. The researchers' work is described in the Sept. 4 online edition of Nature Genetics.
"This is just the first step," says Aravinda Chakravarti, Ph.D., one of the research team's leaders and professor and director of the McKusick-Nathans Institute of Genetic Medicine at Johns Hopkins. "While this is a snapshot of the genomic changes that can happen, it's certainly not everything going on. We still need comprehensive analyses of the changes and what they mean for the functions of embryonic stem cells."
"Embryonic stem cells are actually far more genetically stable than other stem cells, but our work shows that even they can accumulate potentially deleterious changes over time," adds Anirban Maitra, M.B.B.S., an assistant professor of pathology at Johns Hopkins who shares first authorship of the paper with Dan Arking, Ph.D., an instructor at Hopkins. Both are members of the McKusick-Nathans Institute of Genetic Medicine at Johns Hopkins. "Now it will be important to figure out why these changes occur, how they affect the cells' behavior and how time affects other human embryonic stem cell lines."
The researchers in the United States, Singapore, Canada and Sweden compared "early" and "late" batches of each of nine federally approved human embryonic stem cell lines. Twenty-nine human embryonic stem cell lines from seven different companies are approved by the United States National Institutes of Health under President George W. Bush's policy restricting federal funding of this research to cell lines in existence before his announcement of the policy at 9 p.m. ET, Aug. 9, 2001. The dozens of human embryonic stem cell lines developed since that announcement cannot be used in federally funded research.
Most of the "late" batches of stem cells -- those grown in the lab a year to three years longer than their early counterparts -- displayed gross changes in the number of copies of chromosomes or parts of chromosomes, in the marks that control whether a gene is used by the cell, or in the sequence of DNA found in the cell's mitochondria.
"The majority of the lines we tested had genetic changes over time," says Chakravarti. "Whenever you have something in a culture dish, it can change, and it will be important to identify, keep track of and understand these changes."
At this point, the precise effects of these changes on the cells aren't known, but some of the changes resemble those seen in cancerous cells. At any rate, the changes presumably became entrenched in a particular cell line because they conferred some advantage as the cells were grown in laboratory dishes. Whether the changes affect the stem cells' abilities to become other cell types is also unknown.
Although research with human embryonic stem cells is still in the lab -- not the clinic -- focusing on what the cells can do and how they are controlled, the hope is that in the future these cells might help replace or repair tissues lost to disease or injury. Because embryonic stem cells can become any type of cell found in the body, in theory they could replace certain pancreas cells in people with type I diabetes, or regenerate brain cells lost in a person with Parkinson's disease, for example.
The analyses of the embryonic stem cell lines and the computer comparisons of the mounds of resulting data required the efforts of scientists at four academic centers, two federal laboratories and three companies. Critical to the team's success was prescient support of cutting-edge technology development by the National Institutes of Health, support that enabled development of the technological infrastructure necessary for large-scale comparative research, particularly the Human Genome Project, says study co-author Mahendra Rao, M.B.B.S., Ph.D., of the Laboratory of Neurosciences at the National Institute on Aging.
The scientists used so-called GeneChip microarrays, or oligonucleotide arrays, to determine whether there were genetic differences between the early and the late batch of each of the stem cell lines, including whether any genes were present in extra copies. Depending on the gene affected, extra copies could lead to accelerated cell growth, increased cell death, or no measurable effect at all.
In addition to probing changes in the nuclear and mitochondrial DNA sequences and copy numbers, the researchers examined whether the cells' genetic material had shifts in marks that sit on genes and are passed from cell to cell during cell division. These so-called epigenetic marks -- in this case methyl groups on a gene region known as the promoter -- help control whether a gene is used by a cell to make proteins. The researchers determined the methylation status of 14 genes in each of the batches of stem cells; three of the genes did show different methylation patterns in late batches compared to early batches.
The scientists' analysis revealed that five of the nine cell lines had extra or fewer copies of at least one section of their genetic material in the late batch compared to the same cell line's early batch. Two of the nine lines had changes in their mitochondrial DNA over time, and all nine stem cell lines exhibited some shift in methylation of at least one of three genes. One of these genes, called RASSF1A, is also methylated in many cancers, but what effect the methylation has on the stem cells is unknown.
The team is already planning to conduct similar analyses of the remaining NIH-approved cell lines, but analysis of stem cell lines not available for use with federal funds will also be needed, the team members say.
The Johns Hopkins researchers were funded by the Henry J. Knott Professorship in Genetic Medicine, the Sol Goldman Pancreatic Cancer Research Center at Johns Hopkins, the National Cancer Institute, the Maryland Cigarette Restitution Fund and the Donald W. Reynolds Foundation Clinical Cardiovascular Research Center at Johns Hopkins.
Authors on the paper are Anirban Maitra, Dan Arking, Morna Ikeda, Keyaunoosh Kassauei, Guoping Sui, David Cutler and Aravinda Chakravarti of Johns Hopkins; Narayan Shivapurkar, Victor Stastny and Adi Gazdar of the University of Texas Southwestern Medical Center; Ying Liu and Mahendra Rao of the National Institute on Aging; Sandii Brimble and Thomas Schulz of BresaGen Inc., Athens, Ga.; Karin Noaksson, Johan Hyllner and Peter Sartipy of Cellartis AB, Goteburg, Sweden; Xianmin Zeng and William Freed of the National Institute on Drug Abuse; Alan Coleman of ES Cell International, Singapore; Sei-Ichi Matsui of Roswell Park Cancer Institute and the State University of New York at Buffalo; and Melissa Carpenter of Robarts Research Institute, Ontario, Canada.
The microarrays used in this work are the product of Affymetrix (Santa Clara, Calif.). Chakravarti is a paid member of the Affymetrix scientific advisory board. The terms of this arrangement are being managed by The Johns Hopkins University in accordance with its conflict of interest policies.
On the Web: