It is essential to understand that during early embryonic development, the size of the embryo does not change while the number of cells increase exponentially. That means that at a very early stage (1-8 cells) all cells are identical and can adopt any fate. At one point not all cells can interact with the environment in the same way, as the increase in numbers means that some are left "inside" (with no contact with the extraembryonic environment) and others are left "outside" (in contact with the extramebryonic environment and also with the cells "inside". At this point, the signals that the inside and outside cells receive are different and thus the first steps in differentiation start. The inside cells activate a different genetic programme than the ones outside.

From the Plos article: "For example, a cell on the hand expresses a set of genes that locate the cell on the top half of the body (anterior) and another set of genes that locates the cell as being far away from the body or distal and a third set of genes that identifies the cell on the outside of the body (not internal). "

The obvious question is if a set of expressed genes provides an "address code", what determines which genes are expressed that confer this positional identity?

Yeah, but as you know, the proteins (including transcription factors) are themselves encoded in the genome. Moreover, the asymmetry in the concentration of signalling proteins that you refer to still needs to be set up in the first place. The article that Joachim cited above says that "positional identities are relative to major anatomic axes." Great, but how are the anatomic axes themselves established? Do you see where this leads? A vicious causal regress. From wikipedia: http://en.wikipedia.org/wiki/Spatiotemporal_gene_expression "Because current expression patterns depend strictly on previous expression patterns, there is a regressive problem of explaining what caused the first differences in gene expression."

Yes, cells can sense and receive positional cues from chemical, electrical and mechanical forces. But that still doesn't address why cells, with identical genomes, express their genes differentially to begin with, particularly in the early stages of development when anatomical axes have not been fully established.

Already before the first division, asymetric localisation of mRNAs and proteins in the egg suffices to explain a certain degree of asymetry. Bicoid, for example, or /par/ genes are such factors. Localisation cues are encoded in the mRNA (3'UTR I think) or the protein amino-acids. Still, the proteins that anchor these mRNA / proteins in a specific part of the egg (e.g. anterior) have to be themselves asymetrically localised! This is supposed to be maternal, since it happens in the egg cytoplasm.

I think we have an "egg or chicken" issue here...

It is simple to answer. Of course, ll epigenetic and other transcriptional regulation events determine the fate of the cell. These are as important as the whole genome. The heterogeneity in cell responses is also evident in cultured cells of the same origin.

Beston Nore

It is essential to understand that during early embryonic development, the size of the embryo does not change while the number of cells increase exponentially. That means that at a very early stage (1-8 cells) all cells are identical and can adopt any fate. At one point not all cells can interact with the environment in the same way, as the increase in numbers means that some are left "inside" (with no contact with the extraembryonic environment) and others are left "outside" (in contact with the extramebryonic environment and also with the cells "inside". At this point, the signals that the inside and outside cells receive are different and thus the first steps in differentiation start. The inside cells activate a different genetic programme than the ones outside.

Asymmetry is not the answer for mammalian development. It is the way drosophila, xenopus, chicken, an so on begin their development. Our cells, rat cells and so on (mammalian cells) do not have such a significant differential deposition of maternal RNA and proteins to originate asymmetric cells upon the first division. Therefore our first cells can be separated and each one can originate a identical individual - a monozygotic twin, a clone, to say. Which is not possible in the asymmetric development (where the first two cells divide inequality the amount of proteins and RNA the original zygote already had from maternal differential deposition in the egg, therefore originating two cells with different content). There something will be missing if the two cells are separated and the resulting individual will either be smaller, or not develop or develop partially. In initial mammalian development we have the zygote that divide 2 to 3 times (blastomeres, all identical non tightly adhered cells) and at the stage of about 16 cells all start to express adhesion molecules (cadherin-e). Until then they were equal. However, from that moment on, these cells aggregate very tightly (morula stage) and there are cells which have direct contact with the outside environment (the Zona Pellucida) and cells that are inside the cellular "ball", which only contact and receive input though the outer cells. That is the first moment of differential gene expression - you know, signal transduction pathways do maintain the cell informed about what happens outside - and that is the first obvious moment where different signals are sent to the genomes of inner and outer cells. Therefore they begin to respond differently, activate the transcription of different regulatory proteins and very soon express different set of genes. From there on the story is relatively easy to understand - at least in conception. The outer cells bomb fluid into the "cell ball", then we have the blastocyst, where again different layers of cells get distinct inputs from the surrounding environment. The inner cell mass will be on one side of the cavity and some cells will have contact with the fluid of the cavity and the inner cells, others will be in the middle, only having contact with inner cells, other will have only contact with inner and outer cells, and so on. Gene regulation responds finely to the close environment and the cellular contacts.

Ooops, where I wrote bomb fluids, please read pump fluids - I have some problems with english... In portuguese pumping is bombear...

all the cells are NOT "programmed" with the same genetic instructions,.. time and space (spatio-temporal) regulation is the what makes a cell different from another one.

That is a fundamental question in biology...now get to work!!!!

If you talk about the genome and the `instructions` it contains, equal in all cells, is at the level of the DNA. For instance the proteins a cell produces depends on transcriptional regulation. The transcriptome of between cells can be very different. That depends a lot on their position and their identity, which in turn also affects their identity. Furthermore, hormones, which can work as morphogenes, also affect cell identity. As mentioned already, in initial steps, for instance causing asymmetric cell division, maternal mRNAs play a role.

I had asked me this question many many years ago when I attended at school of medicine. I never done more to me this question. This witness to the stupidity and aging. I have read the answers that were given, and these seem to me very reasonable, in particular the observations of Angelica. I imagined that in recent years scientists have reached many more knowledges about the topic. I wonder if, in addition to what has already been said, there could be also a kind of "binary" process. A gene begins to transcribe in the first cell. Proteins and/or RNA stimulate the transcription of other (two?) genes that will be different in the two cells (this is the event most likely after a few cell divisions). After that has occurred for the first time this asymmetry, the different microenvironments in the cells will determine the activation of different genes.

http://doursat.free.fr/docs/Doursat_2008_devo_NZ_slides.pdf

Great question and some great answers above! However I feel a few components are still missing to get the full picture (and none of us has it today although some of us seem to think so): We know from assymetric bacteria (Caulobacter) that DNA gets hemimethylated when one half of the unicellular organism is organized completely different from the other half during cell division. We also know that DNA hemimethylation plays a key role in regulation of gene expression in higher organisms. And we know that both histone methylation and microRNAs influence gene expression. In addition, lipids of different kinds come in clusters thus making the cellular membrane non-uniform, They all play a role in making cells asymmetric before division takes place giving rise to two different progeny.

Now will anyone link all the above answers plus my two cents together and tell us the complete answer to this fundamental question? It will take an entire book and the writer will deserve the Nobel price.

Part of this identical genome is dedicated to just this remarkable event. Certain genes code for RNA's and proteins that direct the process. The real root of the trick is the fact that as cells divide there develops a difference in gene expression in the different cell types of the developing organism. This is even true sometimes for cells of the same type, like cells in different regions of the bones in a developing arm. In this case the differences between the same cell types are small, but the cells have small differences that inform about their location within the arm. Differential gene expression means that there are differences in how much, when and if a gene, one of the approximately 23,000 different genes each cell possesses, is expressed. For this to occur correctly, information is required. For example, information about how many times the cells have divided since the original zygote.

Here is an interesting example. If you separate a two-cell human embryo's blastomeres, you will form two different embryos, each which will develop normally. If you had not separated the embryo, it would have formed a blastocyst, on, say, day 5. A blastocyst is a stage of development where there is a balloon of cells in a single layer with a clump of cells attached to the inside. The human will develop from the clump and the placenta from the balloon of cells. (For you real experts, I am simplifying here). If you separated it into two blastomeres, both embryos would still form a blastocyst on day five. These two would be half the size an un-separated embryo of the same stage and would have about half the number of cells. Here, you can see that the cells of the embryo "know" how many times they have divided. The proportion of cells in the clump of the smaller embryos would be fewer, and the number of cells in the trophoblast (the balloon) would be more initially. This is because the cells on the outer regions of the ball of cells are influenced by there interaction with the solution environment, whereas the cells within the ball of cells are influenced differently. This is information about where the cells are within the developing embryo. Gene expression will already be different for these two different types of cells. As the smaller embryos develop, there will be a time when it adjusts the cell numbers to be even with an whole embryo. You can even put two or more embryos together before they form blastocysts and get on normal ~ individual to form. Here, if two are put together, the blastocyst will be about twice as large and have more than twice the number of cells inside the balloon and a few less than twice the number forming the balloon. This will adjust later to normalize to the number of the typical embryo. (Joke: Maybe making children with 6 or 8 parents might make college affordable).

For mammals, birds and some other vertebrate embryos, the next stage has to do with forming tissues from the cells in the clump. Cells closest to the cells of the balloon, because of their location and the information provided by that location become amnion. Below this, another single layer or tissue like structure forms, which is the early ectoderm, the outer part of the body. Initially it is called an epiplast.

Cells form local structures, express genes based on their genetic code, which includes instructions on how to develop, get informed and behave appropriately, and inform others. It is a truly information rich system, where things talk to each other, react correctly and get things done.

Paradise compared to....

Charles, I think Hans Driesch realized the point you are making but with sea-urchin blastulae over a hundred years ago. He would kill no more than 3/4 of the cells and the rest would always develop into very small, but complete, organisms. However, he attributed this ability to compensate for the loss of parts by claiming that it was due to a "mind-like" force, which he called "entelechy", that brought development towards its end-goal. This is, of course, a neo-vitalist rather than a mechanist explanation.Driesch was convinced there was no way the cells could "know" that some of them had been kiilled by way of a chemical or physical signal and respond accordingly.

I apologize for the term "knowing" if you thought it to be anthropomorphic. I used it as an analogy to human knowing. Of course it is molecular-based. It seems to me that someone smart enough to have thought that it was anthropomorphic would have certainly realize I was simplifying molecular principles. It was clear from the question the answer would be more useful in general terms. The "" were to let you know I did not actually mean know. However, your contribution informed and I enjoyed it. A clip of me making identical twins can be seen on youtube: station wwwcanbeorg.

Regards.

Ps. I think all knowing is molecular.

Spacial/temporal Chromatin variation can change how the genome is expressed during development. Can lead to persistent to changes to gene expression etc long after transient developmental signals. Often people call it epigenomics or Something.

Charles, so what is the molecular basis by which cells regulate development in response to the destruction of other cells at the earliest stages like cleavage?

After cleavage, the cells of these early structures are a size determined by their stage and state of development. To compensate, they divide more than embryos that have not been damaged, until the size of the structure, for example, the balloon (trophoblast) and cell numbers are correct.

It is fascinating that some species are this adjustable. Others, like nematodes, are not that adjustable. If early cells are destroyed, there are examples of them missing parts.

Yes, but what is the molecular mechanism by which they sense that their sister cells have been destroyed and adjust the developmental process accordingly by dividing more? Driesch claimed this was only explainable in terms of some collective intelligence.

These kinds of informational issues are typically resolved by molecular mechanisms that you can get a sense of by reviewing the rule of interpolation. Search this plus molting and cockroach. This explains the short-term signalling systems and how they can be perturbed. If there was a need for a unseen hand, the hand can be easily fooled. Also, you will learn how to make cockroaches with legs of extraordinary length.

Driesch was brilliant for his day, but would not believe his own hypotheses today. His life was before DNA was known to be the information molecule, before the DNA structure was known, before micro-array chips existed, and such. We will too be wrong tomorrow. You can count on it.

Signing out, Charlie

Thanks for the non-answer, Charles.

I guess what is known is that maternal mRNA starts the proccess by giving polarity and correctly localized transcripts, what follows is that each set of genes are then induced sequentially and their product interact with the existing proteome which composition is dependent on position of the cells. This was found by studies of drosophila development. The system obviously has many layers and one level is DNA methylation, then the "histone code".

This is because all cells in the body have the same genome but do not have the same epi-genome which controls gene expression. This will make cells to have different gene expression profiles which eventually will make cells to take different fates.

Having the same genome does not mean that all cells are equal... The whole developmental process is not programmed within the genome. Cells are submitted to different environmental cues such as inner side, outer side for a morula, for example, and also to signals coming from signaling centers (secreted molecules such as TGFb like Nodal, FGFs, Wnts...). These different signals have a word to say on which genes will be or will not be activated and will impact on the fate of the cells. It is like having a book of novels and having someone telling you at some points, read novel10, but not the other ones... Usually, most of the cells will not read their whole genome throughout their life, although globally, the cells that derive from the egg will, again, as a whole, use it.

Stephane, what cues exactly? If you mean biochemical ones, like diffusing signal proteins, that does not help because they have to ulitmately be genetic in origin. So all cells should be sending and receiving the same biochemical secretions if they have the same genome. The real question is how cells begin to express their genes differentially to allow for a difference in the concentration gradient of chemicals. It all becomes very circular and viciously regressive.

WARNING!

If you look for previous threads in questios asked by Joseph, he wants a full and detailed mechanism with all nuts and bolts. He is looking for absolut truth. otherwise he will refuse to accept that:

1. Cells which are genetically identical can take on different developmental states by differential gene expression. His alternative: entelechy, a misterious force that somehow directs development. No details about nature, origin and mechanism for entelechy.

2. Genes are responsible for the developmental program. His alternative is entelechy again.

3. There have been ancestors common to modern species. Unless an unequivocal fossil is found for every transitional or ancestral species. His alternative: creation by God.

4. That there has been a Holocaust (Google his name) unless names and places of death are provided for all alleged six million victims. Accoridng to him it´s all a Western conspiracy.

Unfortunately he is less demanding of his own mechanisms, interpretations, etc. It is enough for him to say he cannont envisage any alternative.

During development the genome is imprinted with epigenetic marks that limit expression to active or accessible regions and silence the non-expressed regions. These marks accumulate sequentially and are the result of positional signals that cells are exposed to in the embryo. Positional information is generally specified by secreted or contact dependent signals like WNTs, Fgfs, hedgehog, BMPs, etc.

Embryonic stem cells are pluripotent and have few suppressive epigenetic marks. However at gastrulation, Polycomb and Trithorax proteins begin to modify histones in a locus and lineage specific manner to restrict cell fate and gene expression patterns. It is not entirely clear how they are able to recognize the right genes at the right time, but this is likely to require both secreted signals and DNA binding proteins that recruit epigenetic complexes to chromatin. Once established, the epigenome is now different and remains so in the separate lineages.

Epigenetic gene regulation is the basis of cell-type specificity.

OK, so we now have many respondents here claiming that epigenetic marks, due to DNA methylation and histone modification, may explain differential gene expression. But I am told that the basis for this differential epigenomics is due to positional signals sent by diffusing morphogens and transcription factors...which are mostly gene products. So the argument becomes circular since all cells have the same genome and therefore should transcribe the same set of genes during the course of development! The only difference that I can see is that some cells are created by division before others which may lead to them producing the same proteins, but not at the same time. There may also be a difference in mechanical pressures for cells further from the centre of the morula or bastula which could affect gene transcription.

Another answer; first genome may contain all necessary information about "timing & space" of differentiation, related to number of division and environment. But we don't know yet this digital information inside the genome. What do IVF physicians think?

Without going into any specific mechanistic detail or full analysis of what happens during early development... I believe the answer to this dilemma how cells with identical genome can adopt different fates is in principle relatively simple (which doesn't mean that it is fully understood in mechanistic detail). But as mentioned above until certain point all cells are identical. Once their environment becomes non-equal for example due to encapsulation within cell mass vs cells at the outer layer the differential gene regulation could be triggered. Epigenetic regulation (where gene products regulate expression of genes) provides means to allow cell to "remember" where they were so gene expression is by no means independent of the environment or changes thereof. For this reason if for example a once encapsulated cells is positioned back to the surface it does not need to respond the same way as a cell that was at the surface all the time because they have been differently programmed.

Aki, you seem to imply that there is a physico-mechanical reason for differential gene expression, with cells at the fringe/outer layer coming under different degrees of pressure and tension than cells drawn towards the centre. Is that your basic idea?

Well, not only that. But, yes mechanical cues are known to play a role (and have been shown to regulate gene expression - same cells seeded onto stiff or soft substrate have different gene expression profile or even cellular fate of differentiation - similarly cells +/- flow has been shown to induce differences in gene expression). But you have to appreciate that there is a remarkable repertoire of microenvironmental variables in developing embryo when you look at it at cellular level. Even in very early embryo there are other cues such as gradient of nutrients and gases (O2, for example) that will develop especially once cell start to make tight junctions (compaction). Then cells inside may see many nutrients only once they have passed through (and been /modified metabolized) by the cells on the outer layer.

As for the biochemical cues - if thinking about just the inner cell mass in isolation (there are many more cues possible if thinking about the whole blastocyst but for clarity lets simplify) it appears that (all of the) the cell on the outmost layer facing the blastocoel lumen will start secreting laminin-rich ECM (the trigger of this event is not well understood but it might simply be the free surface establishing an apical domain that the inner cells don't have) that will form a specific type of "basal lamina" to the opposite side of the free surface facing the outside. Formation of this basal lamina below the primitive endoderm now generates a cue for the second layer of cells which epithelialize and form tight junctions isolating the cells which remain inside the newly formed apical lumen of the second layer (now beneath two layers of polarized epithelium). This in turn leads to cavitation where most of these "isolated" cells are eliminated leaving cell debris inside. During all these phases the cells can "match" the number (meaning they can divide more if there is space or be eliminated or stop growing if there were excess - relating to question how divided or combined blastomeres give rise to normal embryos - size matching - however, the mechanisms underlying the "final size determination" are poorly understood).

Most importantly stem cells having what could be considered "open chromatin" are all the time being programmed to form different versions of "epigenetically packaged/regulated" genomes which tune cellular responses in sequential multistep manner. Thus cells are being programmed stepwise during the dynamic process of development. This programming can be reversible but to large extent it is "irreversible" (in normal development that is - I leave out the iPS/reprogramming business for now). This is the basis of generating cell populations with possibly very differently responsive "transcriptomes" even to same cues (be them gene products, metabolites, gases or nutrients).

Aki, I have never heard of an oxygen-gradient at cell junctions affecting gene expression, but that is an intriguing idea. The problem with biochemical cues, and concentration gradients in general, is that all of the cells should be secreting the same chemicals if they have the same DNA and genetic programs. Hence, establishing such a gradient (i.e a difference in concentration) requires differential expression to begin with! Therefore, concentrations of diffusible chemicals can't be the root cause of differential transcription because they, in fact, require it. In any case, positional cues by chemical signals is a very messy and imprecise process and even the likes of Wolpert, who first proposed it 40 years ago, have now said it is unreliable: https://www.researchgate.net/publication/47660566_Positional_information_and_patterning_revisited

Article Positional information and patterning revisited

Well, you would need to explain to me again why do you think that cells are the source of all cues? The cells of the embryo themselves are certainly not the only source of even "gene-derived" cues - there is also the "host" right? Also I see no problem in the logic of generating unequal cells by divisions and there by forming inside and outside cells. You need to be more specific which part of the logic you do not understand. In the developmental scenario it does not matter that cells have same DNA. Different expression profile you can get in many ways. Just browse some articles regarding micro environmental control of gene expression. There are lot of examples of how MICROenvironmental cues regulate gene expression.

The first lineage decision in the morula is made between the cells on the outside (trophectoderm) and those on the inside of the aggregate (inner cell mass) to make the blastocyst. There are already differences in the distribution of critical proteins, such as cell adhesion molecules and nuclear proteins like Yap1 (Hippo pathway), see the paper from the Rossant lab in Current biology 23, p1195. Once an initial axis is specified, outside to inside, differences in gene expression can be easily modified and stabilized by cell-cell contact through Notch signaling, for example. These very early events are just now being addressed at the molecular level. Positional information, gradients, morphogens, etc. are all much easier to invoke once the embryo has established some differential patterning.

Aki, what I am specifically referring to are "biochemical cues" between cells in the developing embryo. It makes no sense to say that concentration gradients of a diffusible chemical are responsible for differential transcription when the gradient itself requires differential production of the particular chemical/morphogen to begin with. That is circular reasoning and logic, would you not agree? I can see why concentration gradients might become useful later on in development, but not at the earliest stages.

Dear Joseph, Your question is about cell differentiation – one of the most fundamental problems of the biomedical sciences. After 50 years research in developmental biology of reproductive system, fertilization and embryogenesis, molecular biology of the genes (replication, transcription and translation) applying a broad palette of methods, finally I managed to see through the secrets of the epigenesis (please see “Morphogens:composition and function” full text in my RG profile). It explains the production of the factors responsible for selective gene unlocking. In differentiated cells the active genes are euchromatic (the histons of the gene in the sense chromatid are replaced by nonhiston (HOX) proteins. All other genes are heterochromatic. The genome is the same, but the histones keep these genes inactive.

The discovery of the morphogens and their functions elucidates the differentiation, chromatin remodeling, the plurypotency of the blastomeres and finally the problem with the lncRNAs .I added my “Correspondens” in RG. It shows, that the transcript of each gene might be coding or non-coding, depending on the proteins it binds.

Your question is "How is it that all cells have identical genomes and yet have non-identical fates in development?"

Besides many reasons listed above (such as inside-outside of the cell mass, environmental influences, etc), one should always consider the availability of energy and at what energy level available to the developing cells are important. It's because all of the molecules inside each cell function identical or not, or function at what levels of efficiencies, depending on the above mentioned energy levels. Then the issue of Mitochondria comes in. Are they identical in quality and quantity, and genomics and proteomics and energetics? Further, another issue of variations in quantum mechanics of each cell system might be different. Many of the above issues are still beyond the present knowledge of a cell system.

Even within supposedly homogeneous populations of cells one can find discrete differences that emerge as major functional phenotypic distinctions if one happens to be looking carefully at the function in question. Working mainly on postnatal myogenic cells, I have come across two clear examples of this in quite different contexts. One is the spontaneous emergence in the mdx dystrophic mouse, of clonal myogenic populations that give rise to muscle fibres expressing the missing protein dystrophin. They appear to do this by excluding exons from the transcript in such a way as to restore open reading frame. Each colony appears individual in terms of the number of axons it excludes and the mechanism does not correspond well to a simple mutation.

A second is that when one grafts myogenic cells into a dystrophic muscle, the majority of cells die and only a minor subset survive and contribute to myogenesis. Retroviral marking showed that, even within a clonal population, a small number of clones arise that are capable of survival and myogenic differentiation when grafted. Both of these examples suggest that cells may diversify, probably by epigenetic means, within what would appear to be a homogeneous environment. I suspect that what we see is only the tip of the iceberg - we are quite used to the idea of individual specificities of lymphocytes in terms of their antibody paratopes - it may be that they are not that special - only that they have been more susceptible to detailed investigation.

Xenopus has been used as a model organism to study early development and its way of division helps to elucidate this problem: one DNA, a lot of fates. The first cleavage that the egg undergoes divide the embryos in left and right, the second in Dorsal and ventral and the third, perpendicular to the first two, in Ectoderm and mesoderm/endoderm. The idea is that the cytoplasm is not uniform and daughter cells inherit a different set of proteins and RNAs. In Xenopus the site of sperm entry in the oocyte determines were dorsal will developed. Other organisms have different ways of division, but asymmetric inheritance of protein and RNAs and the different environment that individual cells perceive can explain how the cell fates get different and restricted with time. There us still a lot of things in this process that scientist still do not fully understand.

Joseph, ECM is structural cell-associated scaffold that does not diffuse anywhere and instead can absorb and store soluble factors and thereby regulate their diffusion and activation. Biology today is well beyond the phase when cells were though to be sack of enzymes which function as relatively passive responders to the signals from their environment. Cells are remarkably complex compartmentalized dynamic entities that actively modify their environment and thereby can also themselves generate polarity in homogenous environment! Although formation of the above-mentioned ECM could be explained already just by generating different microenvironments upon cell divisions I am sure that there are a lot more variables contributing to the actual mechanisms regulating this process.

As indicated by others there are stochastic cellular/molecular mechanisms that could explain differences between two daughter cells immediately after division. It is true that there are lot that we don't know but you can be convinced that both intrinsic and extrinsic cues can regulate and cause differential gene/protein expression even in a single cell let alone a cluster of cells. I know it is a taunting (impossible?) task to explore all the related literature but there are already a lot of good sources introduced above that would help you where to start studying these issues further.

An interesting aspect (about what I know very little myself but am interested to try to understand) was raised by Dr. Zhou concerning the "rules" of mechanisms at molecular level most likely obey laws of quantum physics rather than those of the "engineer physics" taught at school. I believe "quantum biology" might at some point be able to explain and model the complex seemingly stochastic molecular events taking place in cells. But we may have to wait for a while until we are there :)

Aki, I totally agree that Dr. Zhou's reference to the quantum mechanics and indeterminacy is likely to reveal much about the biophysics of the cells and will surely bexome an important avenue of research in the future. But I seriously doubt that quantum events in the cell are sufficient to cause substantial differential transcription. Biochemical signals, on the other hand, are known to interact with transcription factors and that is why they are given much attention.It is not a matter of there being *some* physico-chemical differences between cells, but rather those that are enough to cause them to behave differently by expressing specific sets of genes and not others. Cell traction forces (with respect to cytoskeletal tension homeostasis) need to be fairly significant and their effect needs to be very precise.

The cells making up the zygote result from successive divisions of a unique cell, then they are genetically identical. As they progress, a coordinated set of signals govern the differentiation and the type of cellular specification of stem and progenitor cells into different directions.

Thus, the fate of cells is determined by qualitative and quantitative gene expression changes, that result from changes in endogenous and exogenous signaling.

The regulation of these changes are transcription factors, that are influenced by environmental factors.

The cells making up the zygote result from successive divisions of a unique cell, then they are genetically identical. As they progress, a coordinated set of signals govern the differentiation and the type of cellular specification of stem and progenitor cells into different directions.

Thus, the fate of cells is determined by qualitative and quantitative gene expression changes, that result from changes in endogenous and exogenous signaling.

The regulators of these changes are transcription factors, that are influenced by cellular micro-environmental factors.

Kalthoum, I agree that transcription factors determine which genes are expressed and which are not. The problem is explaining what the source of the "coordinated set of signals" you refer to is, and also what are the "micro-environmental factors" that are sufficient to get cells to differentiate in the way that they do.

If you imagine and magnify a cell into a huge industrial manufacturing factory which possesses numerous and various assembly and production lines (with all those transcriptional factors and microenvironment factors parts there), but the availabilty of ATP (energy) at different lines and parts are vary, then the manufacturing and final packaging products coming out are different. Of course, the final production products of 2 identical cells (factories) but with varying levels of energy available to activate each active sites of numerious enzymes and molecular interactions including transportors or receptors etc,, will be different.

I hope the following info can initiate our thinking and understanding what I have explained and mentioned in above.

"Three investigators with university appointments in the United States and France were named co-winners of the 2013 Nobel Prize in Chemistry this morning for laying the groundwork behind today’s computer models for understanding and predicting chemical processes—models whose applications include drug discovery.

Martin Karplus, Ph.D., of Harvard University and France’s University of Strasbourg, Michael Levitt, Ph.D., of Stanford University School of Medicine, and Arieh Warshel, Ph.D., of the University of Southern California were cited “for the development of multiscale models for complex chemical systems,” namely, models that marry the classical physics of Isaac Newton with quantum physics in modeling processes that involve chemical reaction."

There is an interesting basic system you might be interested in. With stem cell differentiation, where one cell remains undifferentiated and the other continues on a path of differentiation, the one that remains undifferentiated has keeps the old centrioles, the one that differentiates gets newly replicated centrioles.

I guess the part with one cell remains undifferentiated, is having low level of energy supplyl then the old centriole would not duplicate.

Production of cells with different cell fates can occur by cell polarization followed by asymmetric division, or in multicellular structures, by exposure of the daughter cells to a different environment.

Cell polarity induced by antagonism between polarity determinants (i.e. PAR proteins as mentioned by Vincent) seems to be a common mechanism in various organisms (see these two nice reviews: Goehring and Grill 2013-Trends Cell Biol-, Thompson 2013-Development-). The models proposed explain how polarity can be originated intrinsically, or by environmental/external stimuli.

In plants, the Fucus zygote is a well-studied system of cell polarity establishment induced by environmental conditions (i.e. light, see Bogaert et al. 2013 http://link.springer.com/protocol/10.1007/978-1-62703-221-6_6#). Initial stimuli direct formation of F-actin patches, and later Ca2+ signaling and polar secretion operate to fix and stabilize the polar axis.

How the initial signals are transduced remain the core research question. Changes in the physical properties of the membrane and/or the establishment of chemical gradients are good candidate mechanisms.

As pointed by Javier, I think that in both animals and plants the differentiation of inner-outer cells is decisive for embryo histodifferentiation. In this case the question is how cells sense their position within the cell cluster.

In plants, a very interesting line of research shows that a calpain protein which is highly conserved, is necessary for L1 layer differentiation and later embryo histodifferentiation (see Prof. Olson’s group webpage: http://www.plantbiology-ipm.net/index.php/members/81-research-projects/170-dek1-project).

So, different mechanisms have evolved and are most likely integrated to make cell fate decisions. I think it is worth to look at it from an evolutionary perspective.

I like to introduce students to the mechanisms for morphogenesis, cell differentiation and differential gene expression by using the Drosophila Toll signaling pathway for the establishment of the dorsal/ventral axis. A similar pathway is present in other multicellular organisms for other events. A good description can be found at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2773625/

Key and most interesting to me is how continuous molecular concentration gradients are converted to step responses.

Gene expression

I saw the magic word mentioned above.Stem cells. Though I'm much farther away from the basic sciences and molecular biology in particular than the rest of you in this discussion, stem cells and their potential to differentiate along multiple lineages is pertinent to this discussion. Why do stem cells stem cells differentiate into the cells we want them to? Somehow it goes back to the environment they are exposed to. Maybe too simplistic an explanation but how do you explain why cells with the same genetic potential differentiate into such a variety of cellular elements with specific functions? "Field effect"?

These comments all bring up fascinating aspects of the process, but, the original question is THE big question behind the study of differentiation during development, and we are far from knowing the answer. A lot of scientists dedicate their lives to answering some small part of it, plus the answer depends on what life form you are studying. There is certainly no short answer (unless the question is much more specific than I understand). I think the best way to get a good overview of current thinking would be to take a Developmental Biology course.

One of the mechanisms of diversification within a homogeneous system is the generation of metastable states, where whether you fall on one side or the other of the dividing line is dependent on what are, to the observer, trivial differences. Such metastable states can be generated within linear systems; they occur, for instance when a tube subjected to longitudinal pressure suddenly collapses. This is described within catastrophe theory. Within biological systems, the interaction of large numbers of variables would commonly produce cusps, where the transition to one or other of two alternate states would not be attributable to any measurable quantitative difference. In addition, many, probably all, biological variables are non-linear. The upshot is that you do not necessarily need to receive outside stimuli to generate diversity of cell fate - it can arise within an apparently homogeneous system.

Selective cloning of stem cell and understanding of gene expression provides best option.

Genes are the first level of genetic code. Transcription into RNA is different, that is the second level of difference. short life of RNA is different and regulated by micro RNAs . This is the third level of genetic code. Then there is RNA editing. Protein synthesis is controlled by RNA poly A length and the translatibility of proteins is dependent on many RNA binding proteins and the localization of RNA determines another level of complexity. Further there is post translational modifications that determine function of protein and its degradation rate, thus this leads to another level of complexity. That is how we end up with several types of cells. All this is determined by location, signaling and modifications in expresibility of genes by histone modifying enzymes which determine when and what genes can be transcribed, etc

It is a good question Joseph, and a question science has been researching for a long time. As I read though some of these great responses, it seems we have come a long way to a better understanding why cells with identical genetic material differentiate. However, we should also remember the role of the epigenome in cell differentiation. During development, certain groups of cells receive some chemical/ hormonal signals that cause certain regions of the genomic DNA to methylate (in simple terms) and bind tightly to the histone complex effectively silencing those genes. The development of silenced genes cause cells to differentiate. Of course the question remains as to the original source of these chemical signals. This may help explain why stem cells differentiate into specific cells particularly for repair. Perhaps the damaged cell sends out the chemical signals causing the stem cell to differentiate.

Cells differentiate without division. Stem cells do not differentiate. However, stem cells have to divide in order to generate cells prone to differentiation. Do we know at all if a dividing stem cell either produces two identical daughter cells both prone to differentiation,or produces one daughter cell prone to differentiation and one copy of itself?

Just what I always thought Mark! Asymmetric division is the result of temporal and spacial gene regulation and it cannot be explained by epigenetic and / or environmental signals alone. They both must play their role in this process, but there is still a lot of research required to fully understand this process. I suspect that lipid domains being sequestered to one side of the stem cell must be an early feature well before asymmetric division. This then will cause asymmetric distribution of membrane proteins......

The short answer is that cells do NOT have identical epigenomes. It is clear that during development, as pluripotent ES cells or cells of the Epibalst (prior to gastrulation), start to take on specialized roles, mesoderm neural ecotderm etc., their genomes are modified by specific histone methylation marks and DNA methylation. This epigenetic code acts to silence some genes and keep other poised or active. In essence, epigenetic modifications compartmentalize the genome into active and silent domains. The process of iPS reprogramming in the lab acts to erase those epigenetic marks to push cells back into a pluripotent state.

The question addresses one of the most fundamental problems of cell biology. One of possible answers is presented in Progress in Biophysics and Molecular Biology 2012 (110): 87-96. The of the paper abstacts is below. A pdf of its draft manucript is attached.

Clonal cells are known to display stochastically varying interdivision times (IMT) and stochastic choices of cell fates. These features are suggested in the present paper to stem from discrete transitions of genes between different modes of their engagement in transcription. These transitions are explained by stochastic events of assembly/disassembly of huge ensembles of transcription factors needed to built-up gene-specific transcription preinitiation complexes (PIC). The time required to assemble a PIC at a gene promoter by random collisions of numerous proteins may be long enough to be comparable with the cell cycle. Independently published findings are reviewed to show that active genes may display discontinuous patterns of transcriptional output consistent with stochastically varying periods of PIC presence or absence at their promoters, and that these periods may reach several hours. This timescale matches the time needed for synchronised clonal cells to pass the restriction point (RP) of the cell cycle. RP is suggested to correspond to cell state where cell fate is determined by such competing discrete transcriptional events. Cell fate choice depends on the event that, by chance, has outpaced other events able to commit the cell to alternative fates. Simple modelling based on these premises is consistent with general features of cell kinetics, including RP passage dependance on mitogenic stimulation, IMT distributions conformance to exponentially modified Gaussian, the limited proliferative potential of untransformed cells, relationships between changes in cell proliferation and differentiation, and bimodal distributions of cells over expression levels of genes involved in stem cell differentiation.

F1000 Developmental Biology

Anming Meng

Tsinghua University, Haidian District, Beijing, China.

DOI: 10.3410/f.718338044.793495033

It is well known that the dorsal organizer of vertebrate embryos, e.g. the Spemann organizer of amphibians, the embryonic shield of the zebrafish and the Hensens' node of birds, has the ability to organize an embryonic body axis through gastrulation. In amphibians and fish gastrulas, the dorsal organizer, which is located on the dorsal side, is essential for generating the dorsal to ventral gradient of Nodal proteins. On the other hand, the bone morphogenetic protein (BMP) activity gradient, running from high to low levels along the ventral to dorsal axis, is established by synthesis of Bmp ligands mainly on the ventral side and antagonism of the organizer-derived Bmp antagonists. During the past 20 years, several signaling pathways and dozens of their modulators have been found to participate in the formation and function of the dorsal organizer, and many BMP signaling modulators have also been identified. An impression one would perceive is that inducing an entire secondary embryonic axis with a few molecules is an impossible mission.

In this paper, Xu et al. demonstrated that an ectopic embryonic body could be induced simply by imposing Nodal and Bmp morphogen gradients on the animal-pole side of the zebrafish blastula. They microinjected ndr2 mRNA into one animal-pole blastomere and bmp2b and bmp7 mRNAs into another animal-pole blastomere of the same 128-cell stage blastula, which were usually separated from each other by one uninjected blastomere. This manipulation resulted in the formation of a complete secondary embryonic axis in more than 50% of injected embryos at later stages, while the primary embryonic axis was formed as normal. This finding indicates that the opposing gradients of Nodal and BMPs meet a minimal requirement for evoking downstream molecular mechanisms and cellular behaviors in pluripotent cells to organize an embryonic body. The experimental strategy used in this study may be adopted to regenerate tissues or organs using pluripotent cells in vitro or in vivo.

It is a summary / review by "F1000 Developmental Biology"

Having learned my embryology in the 1960s, I do seem to remember that vertebral axis formation described by Maurizio could be initiated by non-specific stimuli such as a sliver of glass. While it is likely that this acted via the mechanisms he describes, it should be considered in the light of a more general idea that biological systems tend to react to almost any minor perturbation by generating diversity within cell populations.

Perhaps one of the oldest articles addressing the issue of diversification within a clonal population of cells is that of my old mentor Michael Abercrombie.

Abercrombie, M. & Stephenson, E. M. (1969) Observations on chromocentres in cultured mouse cells, Nature. 222, 1250-3.

It demonstrates a progressive discrepancy between the patterns of chromatinization in daughter cells over a series of cell divisions. A surprisingly up-to-date vision derived from a very simple analysis of data from a simple technology.

This actually touches the most fundamental question in developmental biology, that how the first asymmetry begins, especially for biological systems like mammalian zygotes where there is no obvious asymmetric distribution of molecular substances. Our recent paper aims to answer this question by analyzing single-cell RNA-seq data from mouse and human embryos. The paper entitled "Dynamic transcriptional symmetry-breaking in pre-implantation mammalian embryo development revealed by single-cell RNA-seq" is deposited in researchgate: https://www.researchgate.net/publication/282128777_Dynamic_transcriptional_symmetry-breaking_in_pre-implantation_mammalian_embryo_development_revealed_by_single-cell_RNA-seq

Article Dynamic transcriptional symmetry-breaking in pre-implantatio...

http://www.ncbi.nlm.nih.gov/pubmed/20426623

Irreversible differentiation (change of morphogenetic status) and programmed death (apoptosis) are observed only in somatic cells, and cell division is the only way by which the morphogenetic status of the offspring cells may be modified. It is known that there is a fixed limit to the number of possible cell divisions, the so-called Hayflick limit. Existing links between cell division, differentiation, and apoptosis make it possible to conclude that all of these processes could be controlled by a single self-reproducing structure. Potential candidates for this replicable structure in a somatic cell are the chromosomes, mitochondria (both contain DNA), and centrioles. Centrioles (a diplosome, or pair of centrioles) are the most likely unit that can fully regulate the processes of irreversible differentiation, determination, and modification of the morphogenetic status. Centrioles may contain differently encoded RNA molecules stacked in a definite order, and during mitosis, these RNA molecules are released one by one into the cytoplasm. In the presence of reverse transcriptase and endonuclease, processing of this RNA presumably changes the status of repressed and potentially active genes and, subsequently, the morphogenetic status of a cell.

The notion of irreversible differentiation was revoked, mainly by the evolving concept of cell plasticity, didifferenciation, metaplasia, and transdifferentioation.