A question about cell specialization, plus endless twins

Factual Questions
1

Do we have an understanding of the mechanism behind cell specialization? That is, how do all these cells from the original division “know” when to start becoming an arm?

During cell division, shortly after the sperm cell has penetrated the egg cell and division starts, but before any cell specialization occurs, can we seperate a cell off of that grouping and have a twin? Is there a limit to how many times we could do this (if we assume, for example, that we could grow people in vats rather than in the female uterus)?

2

WAG regarding the first question: cell differentiation is controlled (as you probably know) by HOX genes (or homeobox), which are the cell’s coordinates in 3 dimensions. These will tell cells when and where to become a foot or hand or arm. As for “knowing”, my guess would be that integrins and other receptors are a major factor. Integrins cross cell membranes and bind to the extracellular matrix or other cells, and are linked inside the cell to hundreds of different pathways as well as actin, a cell’s skeleton. They allow signals to be transmitted from the inside of the cell out, or outside in (through ligand binding to integrins, or integrin opening/closing which lets it detach from the extracellular matrix). I think zygotes go through a certain number of divisions before they differentiate, so you can imagine a small ball of cells each in contact with one another. At some point a cell on the surface will only have a certain number of integrins bound to neighboring cells, and this could affect transcription in the nucleus of certain HOX genes, which say “hey, we’re on the outside of a sphere, lets start making an arm”. Again, just a guess!

As for your second question, if you split the cells up early enough (before differentiation) a single cell will lack the binding to it’s neighbors and probably figure it is the original copy, which would affect transcription/translation. But I do not know if zygotes contain telomerase which might be required for endless perfect copies to be made (since telomeres - the tips of DNA - shorten with each cell division).

3

This is basically what I am doing for my PhD (cell specialization), or more precisely how undifferentiated cells become neurons.

To the first question: Bob55’s answer has many elements of truth, but it is a little more general than that. HOX genes and integrins are only part of it, albeit in my field a rather large part of it.

In developmental biology, there is a concept of a morphogen – that is something (not necessarily a gene or gene product, but it can be) which has a concentration gradient across a field of cells. Examples of this may be simple things like calcium, retinoic acid, or light. More complicated things may be the ribonucleoproteins determining anteroposterior axes in the fruitfly embryo. HOX genes can certainly be included for some developmental processes. This morphogen will activate or repress different genes at different concentrations. This sets up distinct domains of gene expression, which relatively easy to subdivide as the tissue expands and starts to develop. Each subdivision triggers a new cascade of gene expression and eventual cellular specialization. Eventually, it can be subdivided to a cellular level.

Polarity is another issue. This is usually determined at very early stages. Usually, by the time you have a few cells, they know which way is up, down, left, right, front, and back. This is usually a stochastic event between a pair of cells. Each produce a receptor on its membrane and a ligand for the receptor on the neighboring cell. There is also feedback – the activated receptor means more receptor and less ligand is made. One cell, by chance, usually produces more receptor, and wins the repression battle. That one cell ends up producing all of receptor and the other cell all of the ligand.

In my field (sorry to get longwinded here, but it is late and I get excited to see questions like this), a group of cells in the embryo is first cordoned into an eye precursor by morphogens and subsequent gene cascades. This leads to the expression of a homeotic gene, eyeless and its partners, during embryonic development, which specifies the cells as retinal cells (they now will no longer form leg cells). This packet of cells goes on to form a flat disc of cells in the larva, where all of the interesting development starts to happen. Gradients are set up to determine dorsal from ventral and anterior from posterior (the cells know top from bottom already). Starting at the posterior of the disc, cells differentiate first into neuronal photoreceptors in a wavelike fashion. These recruit other cells in a very orderly fashion to be photoreceptors and accessory cells like pigment cells and lens-secreting cone cells. During pupal life, the cells start to adopt adult morphology to form the precisely structured 800 facet compound eye.

To answer the second question, until the process of gastrulation, which occurs at different times early in development depending on the organism, all of the cells of the embryo have totipotency, that is the potential to form any tissue (or another whole organism). So, yes it is possible to take one of those cells and make a twin. After gastrulation, most cells lose totipotency, but may still retain pluripotency, that is the ability to make many (but not all tissues). In mammalian embryos after gastrulation the inner cell mass of the gastrula are called embryonic stem cells (or ES cells). These cells have pluripotency to form any tissue in the embryo, but not the placenta or other associated structures.

These cells in mice have been successfully cultured for years now. You can (basically) grow them in a petri dish for as long as you need. While it is impractical to make each one into a separate mouse, it is pretty easy to introduce them into another gastrula, thus forming a chimeric mouse, which can be bred out using genetic strategies.

Could human ES cells be cultured? Absolutely, and they already are. It is not my field, but it is a big political argument about how many different lines, their exact pluripotency, their exact culture conditions, and a host of other things. You may remember Bush put some strict limitations on federal funding for research on stem cells, and that is proving to be a rather large hinderance on this kind of research.

4

Something I found incredibly interesting appeared in Nature recently…I’d link, but you have to have a subscription to log in. It seems that in zebra fish, tail cells will ONLY form new tails when implanted in embryonic fish. Other cells, taken from the head, body, fin, etc., will go on to form a new body, resulting in conjoined twin zebra fish, with only one tail. Implanting tail cells results in one fish with two tails, however.

5

Whew, after this thing dropped like a stone yesterday I was worried it wouldn’t get answered.

Bob, it was the telomere question which plagued me in a way, and I’m still not clear on the implications of this. IT would seem the telomerase at the end of the DNA strand would be there from the beginning to me, and that, if this is truly the underlying mechanism for aging, we’d not escape it by simply keeping a few cells on hand that have been reproduced but not allowed to specialize.

edwino, happy to oblige, and thanks for your incredible answer. Since neurons, AFAIK, do not reproduce, is it the idea that introducing such unspecialized or partly-specialized cells into the proper environment could, say, replace damaged neural connections?

I do recall the limitations on that research. But it is still allowed to go on, provided there is funding, right?

6

It is certainly possible to use stem cells to treat some neuronal disease. There already has been some success in treating Parkinson Disease in exactly the way you propose. The problem with regenerating neurons is that the most important thing is for those cells to make the proper connections. As of right now, we have very little way of doing that.

I don’t want to push this into GD territory in order to answer this. It is a sore spot for the American scientific community. The facts, as I know them. Bush met with his Bioethics Council and (allegedly) a team of scientists. This led him to go on national television and state that stem cell research was permitted, but he was limiting federal funding to the 60 stem cell lines in existence at the time of the speech. Federal funding is the absolute lifeline of biomedical research in the US. Without it, it becomes very hard to do meaningful academic science.

It sounded like a good compromise until those in the field started doubting the existence and availability of most of those lines. I believe around 11 of those lines are actually available for research. Second problem is those lines were derived in culture conditions that were suboptimal. Newer and better culture conditions are now used, and that has led to better stem cell lines. But now, a scientist who wants to research those lines needs to use private funding sources. If she were caught using federal money, it would be a crime. Those are the facts. It is messy. Luckily, it ain’t my field.

Let me just cite a pretty good essay about it and leave it there lest it carries into GD. Feel free to start a thread there.
http://www.washingtonmonthly.com/features/2003/0307.thompson.html

To take a crack at your question to Bob, one of the properties of stem cells is that they don’t senesce (age). That’s how come we still have skin, gut lining, glandular tissue, hair, and lung lining when we are elderly. Germ cells have unaffected telomeres as well. The problem hasn’t been isolating these cells, though. It has been keeping them in a culture condition in which they don’t differentiate and start to age.