Surprising human mosaicism means not all your cells have the same DNA

The field of induced pluripotent stem cells (iPSC) has really shaken up science over the last half dozen years reversing our notion that differentiated cells are stuck in that fate, when in reality they are plastic and can “turn back the clock” to become embryonic stem cell-like cells.

Now we have an iPSC paper in Nature, Abyzov et al, which while saying interesting things about iPSC, more generally has shaken up biology by suggesting that humans being are mosaics. Yeah, including you reading this article.

What the heck?

A mosaic is a mixture of course and when we think of mosaics we may more often than not being thinking of a piece of art (see above mosaic of Ulysses) made from a bunch of different small pieces…..but human beings as mosaics?

What does this mean?

Mosaic human beings, which this paper suggests most if not all of us are, are animals whose cells do not all have precisely the same DNA. Reading most textbooks of biology, one would get the strong impression that all of our cells, with the exceptions of gametes and cancers, have exactly the same DNA sequences.

It turns out that this is not true. Imagine something in a textbook being wrong? Shocker, huh?

This new iPSC paper from a collaborative team of Yale and Stanford scientists analyzed the genomes of a bunch of iPSC lines and then compared them to the “parental” fibroblast cells from which they had been made. Because iPSC lines are clonal (i.e. arising from a single cell), in a way studying a group of iPSC is like doing single cell assays since each line came from a single “mother” or “parental” fibroblast.

The authors were interested in mutations (manifesting as something called copy number variations or CNVs) that arose specifically during the iPSC production process. What they found is that very few such mutations arise, perhaps as few as two. Interesting. In fact most of the CNVs observed in the iPSC were present in the parental fibroblasts to start with. Also interesting. However, CNVs varied from one cell to another even though all the cells came from a single person.

What this means is that within our bodies our cells are not all created equal at the genomic level. In other words, we are mosaics.

Wait a minute, you say, could this be wrong? And if so, how?

One alternative possibility contrary to the notion of inherent mosaicism is that the apparent mosaicism is not actually intrinsic to people but arose during the short period of time the fibroblasts used to make iPSCs were cultured prior to the iPSC production process. This possibility is discounted by the authors since the fibroblasts were only passaged a few times, however it remains a possible explanation.

If anything it would seem likely that this paper underestimates mosaicism since they mainly focused on CNVs 2kb or larger….so smaller changes including point mutations and SNPs would not have shown up in their studies.

It is also notable that other studies in the past have pointed to human mosaicism. For example, check out the references from the Abyzov paper:

21. Rehen, S. K. et al. Constitutional aneuploidy in the normalhuman brain. J.Neurosci. 25, 2176–2180 (2005).

22. Youssoufian, H. & Pyeritz, R. E. Mechanisms and consequences of somatic mosaicism in humans. Nature Rev. Genet. 3, 748–758 (2002).

23. Piotrowski, A. et al. Somatic mosaicism for copy number variation in differentiated human tissues. Hum. Mutat. 29, 1118–1124 (2008).

24. Mkrtchyan, H. et al. Early embryonic chromosome instability results in stable mosaic pattern in human tissues. PLoS ONE 5, e9591 (2010).

However, the Abyzov paper really is a more powerful demonstration of this phenomenon.

OK, backup, what does this paper mean for iPSCs and the stem cell field?

Well, on the one hand, the fact that tons of genomic changes do NOT seem to occur during iPSC formation itself seems like good news. However, they did find on average two large (10kb or bigger) CNVs per iPSC cell line that had occurred during the production of those iPSC so that’s still potentially a problem. In addition, the fact that human cells are so mosaic to begin with also presents a challenge because it means that if you generate many iPSC lines from a single patient, they are likely to be genomically heterogeneous. This could have huge implications for issues such as biological function of the iPSC and their derivatives, but also for immunologic recognition of iPSC therapies by a seemingly autologous recipient.