About a year ago, we wrote about the retraction of a paper in Oral Oncology that highlighted a big issue in oncology research: Widespread contamination of cancer cell lines by other lines, making findings difficult to interpret.
One of the common contaminants is HeLa cells. HeLa, of course, stands for Henrietta Lacks, the subject of Rebecca Skloot’s bestseller The Immortal Life of Henrietta Lacks. As we noted last year,
scientists began realizing that Lacks’s cells, now known as the HeLa cell line and used in labs around the world, were so good at proliferating that they had taken over many other cell lines researchers use to study human disease.
Now, in a study published in the Journal of Urology, a group of researchers has taken stock of a supposed bladder cancer cell line, KU7, and confirmed that it, too, is actually HeLa.
Some background on KU7 from the paper:
One of the most popular bladder cancer cell lines has been KU7, which was isolated from a patient with low-grade papillary bladder cancer at the Keio University (KU) in 1980. KU7 has been widely used due to its robust growth in vitro, its amenability to molecular manipulation, and its reliable growth characteristics in xenograft models.
As the paper’s title, “Hiding in plain view,” suggests, there have been clues, the authors write:
Interestingly, KU7 has behaved over many years as a highly invasive and rapidly growing cell line, which is quite disparate to its original description as being derived from a low grade papillary tumor. This discrepancy, however, was not recognized until we had already determined the true identity of KU7. This highlights the importance of bearing the disease context of cell lines in mind when using them for pre-clinical modeling. The widespread contamination of KU7 clones determined here makes us believe that all KU7 in the urologic literature since at least 1984 is likely HeLa.
And that contamination happened at Keio, they conclude:
Our analysis identified that a cross-contamination of KU7 with HeLa occurred prior to 1984 at the source institution. All KU7 clones in the urologic literature should be considered HeLa and the experimental results should be viewed in this light. Our results emphasize the need to authenticate cell lines in oncologic research.
We asked lead author Peter Black how many papers might be affected, and whether his team’s find would mean any retractions:
There are dozens. I have not attempted to quantify them. Most of our own work, for example, has included KU7 as one in a panel of cell lines, which is not particularly important. Others, especially those focusing on orthotopic animal models are more dependent on this particular cell line. I have taken the stance that it is just a model in the first place, and we have evidence now that the model is a bad one – but that nothing should be retracted.
Black raises a good point. Still, what about noting the mixed-up cell line on the papers that have used KU7?
So how many of the papers compare KU7 with HeLa and get different results?
And I think an important question is how many of the papers make conclusions about bladder cancer based on this cell line? Some of those might warrant retraction.
If the falseness of the result is propagated through new false results, then the paper that propagates that false result merits retraction. If the falseness is simply the basis for an initial hypothesis, and the results either confirm or disprove what is believed to be true, or if the false results are simply refered to as part of the discussion, then this is just basic and pure science taking its natural course. In the latter case, artificial, human-interfered retractions are not required. We must take care about how the power of retraction can be abused, even when there is no veritable reason for retraction. Sometimes, an online erratum is sufficient.
The thing may be simply that KU7 evolved in culture at some point into highly invasive, etc, cell line, and it became very much like HeLa. The question is, of course, what tests were used to say it’s not KU7? Now, the tests will be also difficult because in the process of evolution of a benign line into malignant line (the transformation point) a lot of chromosomal changes occur. Was KU7 a benign tumor at first, did it have stable caryotype?
“The question is, of course, what tests were used to say it’s not KU7?”
The link to the article gives the answer. The tests are straightforward and uncomplicated. They used STR typing to show that the putative KU7 cells were an exact match for HeLa. STR typing is the same process used for forensic identity testing. They followed up with comparative genomic hybridization to confirm.
Cell line identity and mycoplasma contamination are two things that keep me up at night. Consequently, I freeze cells down on a quarterly basis, and also when I make a new passage for a particularly critical experiment. STR testing isn’t particularly expensive considering its value, and an increasing number of commercial outfits are beating the ATCC on price and turnaround. The only thing one really needs is time and liquid N2 dewar space.
There’s little excuse for not running these tests.
First, I am not familiar with the test. However, on the level of caryotype, malignant cells are not identical to normal cells, they, as far as it was known, contain deletions and may be duplications. I. e. not exactly the same DNA complement. Would this test show the difference? Would this difference be greater or smaller than one between individuals, probably much greater? Was it ever tried? If the test really shows minor differences in individual genes, as needed in forensic expertise, it should show the difference between normal cells and malignant cells containing deletions, shouldn’t it?
The test needed to distinguish normal somatic cells from cancer cells from the same individual are fairly involved, probably complete genome sequencing. Not impossible, and getting cheaper all the time, but still.
However, the test needed to distinguish two different people (an anonymous Japanese bladded cancer patient and Mrs. Henrietta Lacks) are straightforward and quite simple. As easy if not easier than any routine paternity test.
Thanks! I imagined something like this.
It’s a pretty straightforward process to compare a given cell line with archived material (a la ATCC).
http://faf.grcf.jhmi.edu/str.html
“cell lines came become altered by contamination from another cell line or, due to their instability, from spontaneous genetic change. General recommendations are to authenticate:
-when a new line is established or acquired to determine an identity for the cell line.
-before freezing.
-every two months that the culture is actively growing.
-if the performance of the line is not consistent or results are unexpected.
-before publication.”
Interestingly, note the small number of journals that require authentication: http://www.promega.com/products/pm/cell-authentication/journals/
This is karma. Those who propagated Henrietta Lacks’ cancer cells and reused them without her permission nor acknowledgement, much less payment (to her family?), are reaping what they have sown. There are dozens, if not hundreds, of publications dependent on cell lines taken over by HeLa cells that will need to be re-evaluated and possibly even retracted.
Apparently, HeLa cells, aggressive and invasive, are doing what they do best: surviving.
Gene functions have been identified and validated; dominant negative cell lines were generated using HeLa cells for important discoveries. Sometimes, only one cell line i.e. HeLa cells was used to discover mechanisms of a gene function. It appeared that reviewers did not even ask for validation of these mechanisms in a normal cell type or in any other cancer cell line.
The problem with culturing HeLa cells is they get everywhere. In all labs I’ve been in we always refuse to allow those culturing this line to be incubated with our cells or even use the same cabinets during passage/feeding. It is known they get contaminate even culture hoods after sloppy work and I would imagine there are many other cell lines, believed to be pure, who have been unknowingly contaminated with this line. They soon outgrow less aggressive lines and then outnumber them after just a few passages.
You may be confusing this with mycobacteria – which do have somewhat mystical powers of contamination – HeLa cells are unable to contaminate cultures hoods or crawl through incubators or anything else. The rare cases of contamination that have been identified almost certainly came from sharing media bottles/reagents and/or pipettes between cell lines.
Or maybe you have just worked in very superstitious labs.
LGR….I wish you were right.
Try putting a new PhD student with a flask of Helas with another new PhD student with a flask of fibroblasts.
Check the flasks after two weeks.
HeLas are rather resistant to dying unfortunately.
There is a lot of information about the STR testing process and service for human cell lines on http://www.celllineauthentication.com.