 |
|
Mahendra Rao
Senior Investigator Stem Cell Group, National Institute of Aging |
At a Glance
Name: Mahendra Rao
Title: Senior Investigator, Stem Cell Group, National Institute of Aging
Professional Background: 2001 — present, Senior Investigator, Stem Cell Group, Laboratory of Neurosciences, National Institute of Aging; 2001 — present, associate professor of neurosciences, Johns Hopkins University Medical School; 1999 — 2001, associate professor, University of Utah Medical School, 1994 — 1999, assistant professor, University of Utah Medical School.
Education: 1991 — PhD, California Institute of Technology in developmental neurobiology; 1983 — MBBS, Bombay University, India.
A research team comprising investigators from the National Institute of Aging and Johns Hopkins University announced earlier this month that eight of the 22 lines of human embryonic stem cells that are currently available to researchers have undergone enough mutation as to invalidate studies that are based on them.
Publishing their results in this month's online edition of Nature Genetics, the investigators, led by Mahendra Rao, who heads the Stem Cell Group at the NIA's Laboratory of Neurosciences, and Aravinda Chakravarti, the director of the McKusick-Nathans Institute of Genetic Medicine at Johns Hopkins, used Affymetrix microarrays as well as PCR-based methylation to determine that existing stem cell lines are mutating at rates that may render them useless within one to three years.
While the National Institutes of Health is aware of the results published in the paper, it is still undecided about instituting the monitoring regime suggested by Rao and co-author Aravinda Chakravarti. Jim Battey, the director of National Institute on Deafness and Other Communication Disorders who is familiar with the study, told BioArray News last week in an e-mail that "the NIH has not yet made a decision about the follow-up plans in light of this new information." Chakravarti said that he knew of no organized effort to monitor the embryonic stem cell lines but that Hopkins "might organize to do it."
To learn more about this discovery on which the Nature Genetics paper is based, and what can be done to preserve the quality of the existing lines, BioArray News spoke with NIA's Mahendra Rao last week.
Maybe you can just give me a little background on what the stem cell group does at the National Institute on Aging.
We study stem cells and our focus is on both embryonic and neural stem cells, and our longterm goal is to understand why stem cells are a model and don't age, while all other cells do.
How did you get access to study these eight stem cell lines for this paper?
These are on the National Institutes of Health federal registry. And we talked to the providers who originally generated the lines to get the lines.
Why did you go for eight? If there are 14 others, why were eight a sufficient number for your study?
Well, it's sort of a practical issue on how many we could reasonably do for the cost that entails. You know, you have to culture cells and you have to propagate them for a certain number of passages and it takes one person pretty much to handle a couple of lines. So doing eight meant three or four people were occupied pretty much full time. So, physically it is costly all at the same time. If I could do it sequentially that's a lot cheaper.
So, what prompted you to look at them in this manner? Were you looking for copy number changes or were you investigating something else and found out that they'd mutated?
We've worked with cell cultures for a long time, and we know that cells change in culture. On the other hand, for stem cells, we really want them to stay unchanged. So we want to figure out a way to monitor them to make sure that changes don't occur. So that was what we were looking to do. [The question was] "How do we set up a good monitoring system that is reliable, reproducible, and substantive, which we can use to assess the state of the cells?" So we weren't saying, "We are sure that the cells are abnormal," we just wanted to describe the baseline state of cells and make sure there was a monitoring system in place.
To date, have they been monitoring them?
Not with a sufficient level of sensitivity — it's hard to do.
And that's why you brought in these different arrays to take a look at them. According to the paper, you had an Affymetrix Mapping 100K array and Affy's Mitochondrial Resequencing Array 2.0, and then you used methylation. What were the different tools used for?
The logic was the following. There are two genomes in a cell — your chromosomes, which carry DNA material, and, in addition, you have your mitochondria, which carry a small number of genes. Either one of them, if it changes, can affect cell behavior. The mitochondria are really critical for metabolic activity and mutations in mitochondria cause several muscle diseases and brain diseases. So we thought, "Let's look for those changes in case they have occurred" — one thing that happens as cells age. The other thing we thought was "Well, let's look at genomic changes." So that was the second reason. And we knew that, apart from gene expression, what is really important for stem cells that have come from in vitro fertilization studies is the epigenetic pattern. If it's abnormal the cells, when they differentiate, don't differentiate appropriately. So we thought that that was a good sensitive indicator. People have already shown that might change; we have to figure out a way to be able to do it.
We evaluated various technologies and there are many ways to look at mitochondria, and we were looking for sensitivity, and it turned out that using this mitochip was the best way to go, and it turned out the SNP array was the most effective way to go. And then we looked at methylation. We used methylation-based PCR, rather than an array, because we couldn't find a good array format that gave us results reliably and reproducibly.
Did any of the manufacturers play an advisory role in your study?
Affymetrix has some nice software that they had developed for their SNP arrays which was quite useful, and we also looked at the Illumina arrays and they were very helpful with providing advice and troubleshooting because they had more experience than we had on their BeadArray technology.
Did you use BeadArray?
Not in this paper, but we did test it and we are now using it because we like it as well. And for the mitochip, it turned out our partners at Johns Hopkins had helped pioneer that whole methodology. We took advantage of their expertise.
So what did you find?
What we found was that changes do occur. We compared cells that had been non-passaged for any length of time and cells that had been passaged for some length of time in culture. And we asked, 'Are there changes in any of these three major parameters?' And there were changes. The important thing was that the changes were different and were not dependent on one line — it wasn't like any one line was good or bad. It's just a feature of cells when they grow. Surprisingly, these cells seemed to be a lot more stable than other cell populations that we have looked at. But they are not perfect. If you monitor you can select a stable subset of clones, but if you don't monitor you can lose your population because cells that have a growth advantage from these changes might take over.
You recommend monitoring in the paper, but what does "periodic monitoring" mean?
My recommendation is that one should do it every tenth passage. And there's a rationale for that sort of thing depending on how quickly one can accumulate these changes given the rate of cell division. And I think if you look at the tenth passage and you see an early sign of a change, then you will be able to discard it and go back to an earlier passage and select a healthier population. That way you can maintain a line for a much longer period of time.
You have also warned that the repository is in danger of depleting its usefulness over time. How long will this current set of lines be useful for researchers?
I think it all depends on how well people monitor. Right now, arguably, there are not many people monitoring on a routine basis. So, if you don't monitor routinely it's going to quicker rather than later and our estimate is that it will be between one to three years. But if people monitor it could be 20 years.
How receptive do you think the people that are caring for these lines will be to your paper?
Several of the [paper's authors] actually derived the lines. And they were very cooperative about doing this because initially when we found this sort of preliminary result, we really said, 'Look, we don't know how to grow cells, so why don't you grow them in the best possible conditions?' They were all very willing.
Will it be cost effective for them to do it?
Yes, once they are set up and they know what to do and they have a baseline, I do think it will be very cost effective, especially if they are doing it in one big batch, every tenth passage.
