As an undergrad, I majored in chemistry and worked in a lab doing organic synthesis. In chemistry, there are many things that can go wrong, but when they do, you usually find out pretty quickly, when your reaction workup geysers all over the inside of the fume hood, or your product turns to brown tar, or you recover only starting material. It may be difficult to discover why your experiment went wrong, but the fact that it’s not right is usually clear.
Not so with stem cells.
I began working with human ESCs and iPSCs during my master’s research, and in the years since I’ve gone from naïve to primed to full-on paranoid. And just this week, I learned about yet another way my tricky little cells are out to get me.
1. Contamination
This problem isn’t unique to stem cells, but it can certainly affect them. Contamination can mean that two different stem cell lines got swapped or mixed (which can easily happen if you mis-label something), or it could be with another microorganism. The most insidious contaminants are Mycoplasma bacteria, which don’t cause the cells to die, but subtly mess up your experimental results. In my master’s lab, three months into my project we discovered that all the cells in the lab had Mycoplasma contamination. And this was a well-established stem cell biology lab, where the researchers should have known better!
Although it’s sometimes possible to eliminate Mycoplasma by antibiotic treatment, the preferred solution is to nuke the site from orbit and start over. Hopefully you’ve been testing regularly and can restore your experiment from a clean frozen vial of cells.
The use of antibiotics (particularly penicillin/streptomycin, which don’t kill Mycoplasma) in routine cell culture can provide a false sense of security, and I definitely don’t recommend it. It’s better to practice good sterile technique. When I absolutely need antibotics (which is really only when isolating cells using the cell sorter) I use a broad-spectrum cocktail called Primocin.
2. Aneuploidy
During culture, cells can randomly lose, duplicate, or rearrange their chromosomes. Therefore, when establishing a stem cell line, it’s important to check the karyotype. There are various ways to do this; the traditional way is G-banding (staining chromosomes and looking under a microscope) but these days microarrays are another good option. On several occasions, I have had to discard aneuploid cells. Fortunately I knew to check early on, before starting important experiments, so I could simply use other cells that had a normal karyotype.
3. Incomplete reprogramming
This is mostly a problem with iPSC lines (not ESCs) but sometimes the reprogramming isn’t complete and the cells aren’t fully pluripotent. For example, this can show up when you’re trying to differentiate them into mesoderm but they only make ectoderm.
4. Weird mutations
Certain mutations can give a growth advantage in stem cell culture. For example, it’s known that inactivating p53 mutations recurrently arise in culture of ESCs and iPSCs. Fortunately this isn’t super common, but it is something to watch out for.
5. Weird mutations from gene editing
As part of my research, I have generated several iPSC lines with fluorescent reporter alleles. I used Cas9 to cut near the stop codon of the targeted gene, and homology-directed repair to insert a reporter allele. Since I wanted homozygous clones, I checked for the presence of the inserted allele and the absence of the wild-type allele. I knew Cas9 could also cause deletions in the wild-type allele, so I checked for those too. Finally, I also checked for integration of the plasmid backbone. Although I detected all of these anomalies at least once, I found a few clones where everything looked OK, and I’ve been using those for my experiments.
But now I saw a paper this week that scares me. It reports that Cas9 editing can cause not only deletions but also large insertions of DNA at the targeted site.
For example, in some cells mitochondrial DNA randomly got inserted. My screening methods weren’t designed to detect this, and the paper reports abnormalities in 1/3 of supposedly homozygous clones. Given the number of edited iPSC lines I’ve made, it’s likely that at least some have this problem.
The authors propose several methods for checking this, but I think at this point it’s probably worth it to get a full genome sequence for every iPSC line I make. It’s only a few hundred dollars1 per line and given the amount of trouble it could save, I think it’s definitely worth it. By looking at coverage across the genome, this could also detect aneuploidy. And it might also pick up bacterial contamination, if it’s present.
tl;dr: if you use an ESC or iPSC line for your research, get a whole genome sequence.
I’m being quoted $800 per cell line for a 15X coverage genome sequence. Meanwhile Nebula Genomics offers 30X coverage for $299. I wonder why sequencing for research purposes is so much more expensive. I also wonder if I could just submit the genomic DNA to Nebula.
That is a terrifying paper. Thank you for sharing your experience!