Mosaicism: indicates the presence of two or more populations of cells with different genotypes in one individual who has developed from a single fertilized egg .
For some time, layperson and scientist alike have been under the misapprehension that, excepting in cases of distinct pathology and a handful of immune cells, the genome in a given individual is homogenous across cell types. That certainly simplifies things: you study one cell type of an individual, generally serum-derived, you genotype those leukocytes, and ba-da-bing! The genome for every cell in that person’s body.
But is it?
It has traditionally been assumed in research and medicine that somatically observed mutations are originally borne from the germ-line. But evidence has been building over the last decade which suggests that this is not always the case and somatic mosaicism is more common than once thought. As Piotrowski et al. (2008) summarize: “The majority of described [copy number variants (CNV) have been] previously shown to be polymorphic between unrelated subjects, suggesting that some CNVs previously reported as germline might represent somatic events, since in most studies of this kind, only one tissue is typically examined and analysis of parents for the studied subjects is not routinely performed” (p. 1118).
Somatic mosaicism is believed to be involved in intercellular genetic diversity, prenatal development, different human diseases such as cancer, aging, partial or whole uniparental disomies, and aneuploidy . In a study published in 2010 in The American Journal of Human Genetics, researchers reported somatic mutations in approximately 1.7% of samples across 1991 subjects . Mutations included “23 segmental uniparental disomies, 8 complete trisomies, and 11 large (1.5–37 Mb) copy-number variants”. It will be interesting to determine in future whether smaller copy number variants are also more common given the recent development in higher-resolution comparative genomic hybridization (CGH) technology .
Mosaicism is apparently even more common during the pre- and perinatal periods, particularly that of aneuploidy, and is found in approximately half of all preimplantation embryos (an occurrence which is likely involved in spontaneous abortion), 1% of chorionic villus samples (CVS), 0.2-0.3% of amniotic samples, and 0.1% of newborns . Vanneste et al. (2009) also found that segmental duplications, deletions, and amplifications were very common in cleavage-stage embryos. These researchers likewise suggest that these results explain the comparatively low rates of human fertility and common chromosomal disorders, although further animal studies will be necessary in order to bear out these conclusions. Yet with talk of growing rates of infertility in our modern age combined with grievous concerns over man-made teratogens and mutagens such as synthetic endocrine disruptors, it will be interesting– and perhaps frightening– to determine whether such environmental agents may be promoting greater genomic instability [5, 6].
In the study of mosaicism, mobile elements (otherwise known as transposable elements) have received more recent attention, particularly that of the long interspersed elements, LINE1 or L1, the only active LINE within the human genome . Mobile elements are believed to comprise approximately 50% of the primate genome, including our own, and are stretches of DNA within the larger genome which, if not transpositionally dormant through mutation, are either capable of direct transposition utilizing a cut-and-paste methodology or indirectly through retrotransposition as in the case of retroviruses . As one might imagine, having segments of chromosome which can jump around the genome can not only be potentially disruptive but on some occasions may also be adaptive. As such, certain cases of transposition may have been advantageously harnessed by cells and integrated into their regulation. And in fact, one study found that up to 25% of regulatory elements generally contain remnants of mobile elements . What with the ENCODE Consortium’s recent publications detailing RNA transcription on a much broader scale than once assumed, it is also possible that RNA products derived from segments of dormant transposable elements, while incapable of retrotransposition, may instead play ribozymal or structural roles within the cell. Another possibility is that the transpositionally-dormant sequences themselves provide a certain level of DNA flexibility which is vital to the potentially adaptable nature of the genome, as may be the case with the C4b gene and its resident HERV-KC4 extinct retrovirus which is believed to promote variations in haplotype amongst humans within that gene .
Interest in the roles of mobile elements in brain development and disease susceptibility have garnered additional attention of late, with some research suggesting that L1 transposition could promote somatic mosaicism amongst neurons. Singer et al. (2010) proposes that, “L1-induced variation could affect neuronal plasticity and behavior . . . [such that it may provide] a mechanism for generating diversity in the brain [which] could broaden the spectrum of behavioral phenotypes that can originate from any single genome” (p. 345). In a recent study, however, Evrony et al. (2012) counter with the finding that within 300 single neurons profiled, L1 retrotranspositions occur on average less than 0.6 times per neuron, while most neurons sampled exhibited no detectable retrotranspositions. While 300 neurons may not provide enough power, it does suggest that L1 retrotransposition isn’t a mechanism utilized in the development of every neuron.
Having studied some of these LINE elements, I would predict that utilizing retrotransposition as a mechanism for maturational regulation would be an unreliable method for large scale transcriptional modulation. While such elements no doubt target predictable areas for insertion, many such areas exist across the genome, which would potentially prevent specificity for something as refined as neuronal differentiation. However, the concept of genetic instability providing a means for change between generations of cells is not a bad one and the method may instead lie within localized duplications and deletions rather than more unpredictable reinsertions. In such a circumstance, mobile elements, whether active or extinct, could instead effect the local topology of the DNA, leading to greater instability which may promote localized deletions and duplications and subsequently change gene transcription. The Huntington’s locus which contains a separate copy of the HERV-KC4 extinct retrovirus shared by C4b exhibits a similar instability and likewise houses trinucleotide CAG repeats whose expansion leads to somatic mosaicism and ultimately to Huntington’s disease [10, 11].
While altering the genome is risky, what uses may it serve? One question which I could conceive of it addressing is “What are the intrinsic mechanisms by which a progenitor cell knows when to proliferate and when to stop?” Research has shown that telomere length is one means by which we may measure cell age [12, for example]. And telomeres, like certain unstable intrachromosomal repeat sequences tend to form hairpin structures (like the header image posted with this entry). If cellular intergenerational changes in telomeres can provide a timer for the decline of totipotent and multipotent cells, then perhaps select intrachromosomal changes in the form of deletions, duplications, and other rearrangements may likewise provide some form of internal “clock” for the cell, leading to changes underlying differentiation. This is not to suggest that all such unstable regions are adaptive, but it is feasible that some may have eventually been exapted for such a purpose.
While it is still too early to tell how extensive somatic mosaicism may be and whether it plays adaptive roles during development, research in areas such as embryology, aging, cancer, Huntington’s, Fragile X, and schizophrenia all point towards its importance. Should genetic dynamism be proven a fundamental player in cellular development, an entirely new and different view of the genome will arise. One which will make our current model of DNA as seemingly static and stagnant as a brackish puddle of water.
This is a very nice review of the subject.
Actually, though perhaps widely forgotten, the idea of cancer and other disorders being due to somatic mutation is rather old. In an earlier version, the immunologist Macfarlane Burnet referred to ‘forbidden clones’ and there were other ideas about cancer in particular, in the 1970s and early 1980s that led to some of the major advances in cancer genetics, such as ‘loss of heterozogyosity’ work by Vogelstein and others. Several of us wrote articles suggesting that somatic mosaicism can be important, as long as you have a mechanism that I have referred to as phenotype amplification, for making somatic change manifest at the organ or organismal level.
But despite the fact that it is obvious that somatic mutations occur, and data on many diseases have existed for some time, it is not convenient for those who want to do ‘genomics’ in the quick-and-easy GWAS way by getting blood or cheek samples from individuals, as you said in your post. Even those cells are not genomically identical, for the same reasons you describe. So, maybe with a push from many, like you, there will be more of a recognition of the importance of somatic mutation–many interesting issues arise about how their effects may work.
Thank you for your insights! I agree with everything you’ve said. And like you say, somatic mosaicism certainly throws a monkey wrench in the GWAS works. But I truly look forward to a change in paradigm, even if it’s slowly but surely. I suspect since mosaicism is a result of the intricate workings of the genome, the more we learn of the latter the more it will be impossible to continue to ignore the former. 🙂
Pingback: Links 1/3/13 | Mike the Mad Biologist·
I know nothing of this subject so this may be a stupid question, but could the existence of “mosaicism” cast doubt on the accuracy of DNA evidence in court cases?
I’m definitely no forensic scientist, so please take my educated guess with a grain of salt and don’t necessarily quote me on it. However, I should imagine that despite that mosaicism may be far more common than was once assumed, our own cells still have more genetically in common with each other than to another human being’s cells, so the overall result would probably still confirm identity. Plus in collecting DNA from a crime scene, while the sample may be small, it’s still more than just DNA from a single cell. In the case of blood, I’d imagine the majority of blood cells share almost identical sequence homology and those that don’t would get washed away by the majority result.
Well, that was my stab at it. I’m sure a forensic scientist would undoubtedly have answered that more succinctly since I’m genuinely not familiar with the methods used for genotyping in forensics. If they’re at all similar to methods I have to apply in my daily work, then as I said, the majority genotype would carry the lot.