Axioms fall

Figure 1

Once upon a time, I believed in various axioms about the body, in particular about the brain and its all important neurons, one of these last being illustrated above.

Axiom 1: humans are made up of cells and pretty much everything which happens inside a cell is, ultimately, under the collective control of the 23 pairs of chromosomes which reside in its nucleus, that is to say the nuclear DNA. Here leaving aside the business of the much smaller  amount  of mitochondrial DNA which resides in mitochondria rather than in nuclei. Here leaving aside the business of cells which do not carry DNA, for example blood cells.

Axiom 2: unlike most less specialised cells in the body, neurons cannot divide, cannot produce new neurons. All the neurons in your brain have been hatched by around the time you were born. After that, it is all downhill. With the consolation that you started with a lot more neurons than you actually need.

Axiom 3: all the cells in your body, including your brain, carry in their chromosomes a more or less exact copy of the chromosomes put in place at the time the egg that became you was fertilised. The mixture of paternal and maternal genetic material that is you.

Corollary: yes, there are accidents; sometimes there are mistakes in the copying. But such mistakes are usually fatal and bad copies do not usually survive for very long.

Axiom 4: a cell keeps the chromosomes with which it is born. They do not change during its life.

Axiom 5: nothing that happens in normal life disturbs the chromosomes which you carry in a way that can be handed down to your descendants. That was fixed at the time of your fertilisation. Although, to be fair, which bits of which of your chromosomes make it to your descendants is up for grabs on each and every occasion. 

This understanding, these axioms, are summarised in the figure above, with the usual convention that an arrow marks a many to one relationship. So, for example, many cells share the same place of usual residence, reasonably coarsely classified. Our opening position is that the two boxes for chromosome set, bottom left, are redundant because all cells are built using just the one chromosome template, top left. As far as that goes, all cells are the same.

We start with the single cell, the fertilized egg, dark blue in the middle. Which by a complicated process of splitting turns into the trillions of cells – that is to say thousands of billions of cells – of the adult human body. With neurons amounting to getting on for 100 billion of them. Some of these cells retain the ability to split, to spawn new neurons; some of them do not: reproducing cells left and terminal cells right, respectively.

To the right we have some cell properties. From the top, the time, date, place and type of birth. Time because some cells do not last very long. Usual residence reminding us that cells may migrate from their place of birth, although not usually very far. Type of birth reminding us that a cell may split into two either symmetrically or asymmetrically and that a child cell may or may not have a different function, a different destination from the parent cell. It may or may not be terminal – remembering here that all neurons are terminal. We might be interested in the number of generations there are between the cell in hand and the original cell, the fertilized egg. And we are quite likely to be interested in what sort of cell we are looking at. Does it belong down among the intestines or does it belong up in the head? Is it fish, flesh or good red herring?

We are agnostic on the question of whether the parent lives on as one of the children in the case of an asymmetrical birth. Figure 2 caters for both possibilities. A question of theology rather than biology?

Provocation

This post was provoked by reading an article in the MIT Technology Review (reference 1) – a magazine I have recently fallen for, at least for a year – a  magazine which comes with a regular barrage of emails – and I now find that the second and third axioms fall. And I am not very confident about the fourth and the fifth. Provocation which included poking around in Wikipedia – and in various papers, for example references 3 and 4 – which last were mostly more or less inaccessible, but which did provide some flavour.

Falling to the extent that lots of cells in the brain don’t even have the right number of chromosomes. And lots more have various other copying errors. The term for all this being genomic mosaicism or GM.

One example of GM is copy number variation. A chromosome might contain a segment which is about building some particular protein. It might usually contain some particular small number of copies of that segment, one after the other. But that number can vary, giving rise to what is called copy number variation. In some cases, it has been possible to show that the amount of protein building that goes on is a function of the number of copies of the relevant segment.

Cancerous cells are particularly likely to exhibit GM. The cells of older people are more likely to exhibit GM.

Estimates for the proportion of neurons which have some significant variation or other appear to range up to about half. Of which only some small proportion will be pathological. Estimates which increase as you add in smaller and smaller variations – with detection at the single cell level getting hard as the variation gets small. Which gives us another counting problem, another statistical problem. Does it help, is it appropriate, to count everything?

Part of the motivation for this work is analogy with the immune system, where chromosome copying errors serve to help generate the huge variety of immune cells needed to combat all the potential threats out there.

Another part is the failure to tie many if any mental disorders to specific errors in specific genes. Maybe we would do better to look at copying errors in aggregate. Perhaps it would turn out that too many copying errors of this sort in that part of the brain was the root cause of some or other mental disorder of interest – with there being plenty of people out there looking for the root causes of the unfortunately common complaints of autism and schizophrenia.

Maybe it is not the copying error itself that matters. Maybe it does not matter so much if the chromosome has three copies or four (say), but maybe it does matter, it does cause trouble, if half the relevant cells have three and the other half have four.

And the opportunity is provided by the emerging techniques & technology for sequencing the genome, at least after a fashion, of a single cell.

Very much work in progress. And in the meantime, I shall have to be careful with my axioms.

Figure 3

PS 1: part of one of the figures which I did not properly get to grips with in reference 4.

PS 2: one wonders whether old-speak sequencing, using a lot more material than a single cell, is actually producing some kind of an average where there is variation. More work in progress, I dare say.

PS 3: the next day: after some poking around, I have turned up where I came across some of this sort of thing before at reference 5, notice of a talk by Paul Davies at the Royal Institution. Must take another look at his ‘The Demon in the Machine’ and at Gerald Edelman’s ‘Second Nature’. Lots of Darwinism Mark II in both. Sometimes alternatively known as Neural Darwinism.

References

Reference 1: The quest to learn if our brain’s mutations affect mental health: For years scientists have tried to find a gene for conditions like schizophrenia, Alzheimer’s and autism. But the real source could lie in a much more complex genetic puzzle – Roxanne Khamsi/MIT Technology Review – 2001.

Reference 2: https://www.technologyreview.com/. 

Reference 3: Genomic mosaicism in the developing and adult brain – Rohrback S, Siddoway B, Liu CS, Chun J – 2018.

Reference 4: Mosaic copy number variation in human neurons – McConnell MJ, et al. – 2013.

Reference 5: https://psmv4.blogspot.com/2019/02/paul-davies.html.