Our Speckled Brains

It’s not exactly true to say that each of us has our own genome. We have genomes. Some of us, known as chimeras, have genomes from more than one person. The cells of children linger behind in their mothers; in the womb, cells from twins can intermingle. The rest of us non-chimeras can trace our genomes to one origin–the fertilized egg from which we developed. But as the cells in our bodies divided, they sometimes mutated, creating a panoply of genetic variation known as mosaicism.

I wrote about chimeras and mosaics in September in the New York Times. My article was a status report of sorts. Scientists have known about our many genomes for decades. But with the advent of single-cell genome sequencing, they’re now learning some surprising things about our genetic multitudes. As a status report, my story was far from the final word. And now, just a couple months later, a new study has come out that sheds more light on a place where our mosaic nature can have huge consequences: our brains.

For a long time, scientists who study mosaicism have focused their attention on its dark side. In the 1960s, for example, scientists recognized that cancer cells were the result of our mosaic nature. Mutations arose in a line of cells, and eventually those mutations drove the cells to grow quickly and develop into tumors. And since then mosaicism research has continued to revolve around diseases. A number of rare diseases such hemimegalencephaly–in which one side of the brain is bigger than the other–have been traced to mutations that arise in developing cells.

This is important research, but it risks providing a lopsided view of our mosaic nature. We are left to wonder how many genomes a healthy person can have. Scientists have started to shift their attention from disease to health, and they’re finding that we can have a surprisingly large amount of variation with no apparent ill effect. In the latest issue of Science, Fred Gage of Salk Institute for Biological Studies and his colleagues provide a deep look into the mosaic nature of healthy brains.

First, they watched the brain’s mosaic emerge. They grew three colonies of human stem cells, rearing each of them in a broth of nutrients. Mixed into that broth where chemicals that coaxed the stem cells to develop into neurons. The scientists then plucked out 40 of these neurons and analyzed their genomes. Thirteen of the 40 cells had changed markedly from their ancestors. Some had accidentally gained an extra copy of a chromosome, while others had copies of smaller chunks of DNA. In other neurons, chunks of DNA had been chopped out. The changes were never the same, which meant that they had originated separately.

The scientists then turned their attention to real brains. They took tissue samples from three healthy people who had died in their twenties in accidents. From those samples, the scientists isolated 110 neurons and surveyed their genomes.  In those neurons, they found a similar pattern to the one they saw in their dishes of stem cells. Forty-five out of the 110 neurons had either extra copies of DNA or missing segments. Again, none of the neurons shared the same mutations. That finding means it’s unlikely that the neurons share mutations that arose in a single neuron early in development. Instead, new mutations kept emerging as the brains matured and neurons divided.

Far from being a rare, dangerous fluke, in other words, mosaic neurons turn out to be abundant in our brains. The figure at the bottom of this post shows how this new study expands our understanding of how we become mental mosaics.

With so much mutating going on in our brains, it may be hard to believe that our brains can work at all. In a commentary accompanying the paper, Evan Macosko and Steven McCarroll of Harvard sketch out some defenses our brains may have against this genomic messiness. For one thing, mutations tend to emerge in the parts of the genome that a cell uses least. So many of the mutations that Gage and his colleagues have discovered may affect genes that don’t matter in the brain anyway.

Even if a mosaic neuron does turn out to be defective, the brain may have ways to prevent it from causing much trouble. When the brain develops, it starts by producing an abundance of connections between its neurons. Only later does it then prune many of those connections back. The brain may take its pruning shears to defective mosaic neurons with particular vigor, cutting them off from conversations with other cells.

It’s even possible that those misfit neurons can let our brains perform in new ways, Macosko and McCarroll suggest. The brain may not just tolerate diversity. It may depend on it.

A: If a cell mutates very early in development, its descendants will be found across much of the body. B: A mutation that arises later in the brain and causes cells to proliferate may be easily detected. C: A subtler mosaic forms when neurons experience unique, late-developing mutations. From Macosko & McCarroll, Science 2013
A: If a cell mutates very early in development, its descendants will be found across much of the body. B: A mutation that arises later in the brain and causes cells to proliferate may be easily detected. C: A subtler mosaic forms when neurons experience unique, late-developing mutations. From Macosko & McCarroll, Science 2013

11 thoughts on “Our Speckled Brains

  1. I would think that since brain mosaicism is so widespread it must be advantageous. Otherwise it would have been selected against a long time ago in the process of Evolution.

  2. What is very interesting is that no two of the cells had the same mutation. That strongly implies that the deletions occurred during the last cell division.

    In other words, what ever triggered the cells to stop dividing also triggered reduced fidelity copying. or triggered the ending of high fidelity proof-reading. It may even be that the trigger to stop dividing is the accumulation of these types of errors. The very high error cells have probably been deleted via apoptosis in utero or later (these were 20 year olds).

    There are signals to stop the cell cycle when DNA is not replicating correctly. These are probably the archetypal regulators of the cell cycle. Humans only recently evolved our gigantic brains. It is likely that what limits brain size in utero is not something “neat and tidy” but rather nerve cells dividing until they crash into some limit and can’t divide any more because of random DNA fidelity issues. This is probably something that cannot be modified.

    If this result holds up, then the genetics of neuropsychiatric effects just got impossibly complex. There isn’t one genome in the brain, there are ~10^8 or more. If this is correct, then genetic testing for neuropsychiatric effects will not be successful except for serious and syndromic conditions.

  3. Wow. The different experiments suggest that between 30 to 40% of our brain cells carry unique mutations. The human mutation rate is described to be in the range of X 10^-8 mutations per genome per generation. I am not really sure how comparable these figures are since they are different ways of measuring.

  4. “I would think that since brain mosaicism is so widespread it must be advantageous. Otherwise it would have been selected against a long time ago in the process of Evolution.”

    You can only select against something if it is a germ-line variant. Since the mutations that produce mosaicism occur later in development and are not part of the germline, they cannot be inherited. What is not heritable cannot evolve.

    The null hypothesis that brain mosaicism is a non-heritable secondary side effect of environmental and in-born factors that impact the overall mutation rate seems more likely to me.

    1. If the human genome is like other self-modifiable program code, then it is quite possible that part of it contains code to permit or inhibit (constrained) mutations in other parts of it. The mutations themselves would not be inherited, but the capacity to produce them could be. If that capacity proved advantageous to survival and reproduction, then it would be selected for.

      I have read about self-correcting genetic code, so it seems likely that there could also be code which ‘chooses’ not to correct in some cases.

  5. amphiox, that is not correct. If brain mosaicism was detrimental, there would be evolutionary pressure to select against it, or rather to select against germ cells that produce individuals with brain mosaicism (which is the same thing). Evolution works on populations, not individuals.

    That brain mosaicism seems to have happened during the last neuron division, most divisions and essentially all previous neuron precursor divisions are essentially error-free.

    That most neuron divisions can be error-free means that there is nothing inherent about neuronal division that requires CNV errors. To produce 10^11 nerve cells requires ~37 divisions. If half the 10^11 cells are error-free, that means that only in the ~last division did errors occur. That last division does occur in utero. If that division caused errors that were sufficiently detrimental, that individual would never reproduce.

  6. In a sense this is not all that surprising. We’ve known for quite a while now that each mitotic cell division creates a possibility of mutation, and that sperm cells show levels of mutation that increase steadily with age. It would be surprising if the same principles did not apply to stem cells in the brain.

    At a philosophical level, we now see that the question of nature versus nurture is incomplete — the adult human brain is the product of nature plus nurture plus a remarkable amount of randomness.

  7. Thank you for discussing that it begins in the UTERUS.

    That would certainly account for the story in our newspaper today:

    While police were investigating a burglary at a woman’s residence, the burglar left his cell phone behind on the counter and his MOM called him while the police were investigating. The phone stated “Mom” when it rang. Obviously they caught the 19 year old burglar and the apple didn’t fall far from the tree it appears.

  8. For me there was no shadow of a doubt, that our brains are using all instruments of genetic variations to orchestrate our brain activity. I extrapolated this from my basic research with plants and cell lines. However it has to be questioned in how far epigenetics can be located in a single cell. The brain is modelling the outer world through the conversion of signals in multidimensional though patterns, on which we even walk in mind. By surpassing a perception area either in time or space and time the brain is beforehand of the present moment, having in mind the past. Perhaps some know Peter Thosh’s Song: I’m a man of the past, living in the presence, stepping into future. Maybe he made a mindwalking to see this beforehand. But we should walk, not run. From the archaic perspective running is for hunting, but we should not go on a hunt of our species. Or flight, but also for this I DON’T see areas left on this planet. We can cross the world, and we will come back to the same point were decisions can’t be avoided, but not about life in the hands of the sosociety, this is dangerous for mankin, because the society is following blind a few opinion makers. The society has to care for the evolutionionary branch on which we are sitting, and therefore it has to be a viscous process.
    The evoltion of mind was only possible through those observations we made on higher automobile organisms. And in the antropogenic world, which physically shapes it’s environment, the major evolutive pressure is on the brain, as the master organ for modelling the environment. Thinking is influenced by all signalling pathways and a memory is not restricted to brain neuronal cell rather than to all cells of the entire orchestra of all intra and interacting cell systems, which shape Homo sapiens. This is autopoesis but it is not a construction of mind like drawn by arts, it is coconstructed by the infrastructure of mind, which is the entire body in it’s environment.
    Due to that we should not follow linear thinking when talking about consequences of advanced diagnostic tools such as PCR of blood cellservice. We need a monitoring for that because it is possible, that we cut our own species evolution. I am pretty much in favor a general omic diagnosis of our children, to monitor this beauty of evolution and this would serve for new jobs in the future. A future in which a new and advanced thinking of the so far unthinkable will be shaped computer aided. Internet is already a mega cortex of human intelligence and even when a lot of data are not reliable, it is just a mirror of our human interaction based on hiding, cheating and discrimination for profit and growth. These are biological principles, nothing is wrong with it but our planet cried already for those who understand the language of complex problems in closed system, even in the didimension of a planet. IT’S up to us to form the basis of a future, which should be as less invasive in our genetic code as possible, as long as we have in fact no clue about it. Computers can’t model the adaptation on antropogenic species development. I am sure Steven Hawkins would agree on that, although I am not so pessimistic regarding artificial intelligence, this will soon be able to take over the role of the cerebellum in the world, for example in intelligent Prothese. Step by step societies will become more social again because we all are sitting in one boat.

    Molecular medicine is something we should focus more on, than on negative selection. I have a daughter with Trisomie 21 and when I am deeply listening to her, than I realize that she seems to see the world with different eyes. So we should leave the decision at the parents site and don’t blame them for medical costs or social care wichave is negligible in comparison to the penalty we pay when we shorten the evolutionionary capacity.

    Perhaps we can observe an allopathic species evolving in the Antrozoicum – the world shaped by the mind of a species living on this planet – the Homo antrozoicum. genetic is the Lego brick in life but epigenetics is already known as a part of the orchestration. Lamarck wasn’t wrong and as far as I know Darwin never criticized him, since most probably both were aware of modifications.

Leave a Reply

Your email address will not be published. Required fields are marked *