Stem cell factors rendezvous away from DNA

Sharing is caringShare on Facebook0Tweet about this on Twitter0Share on LinkedIn0

(Mol Syst Biol Oct 2013) Factors that are crucial for the regulation of embryonic stem cells control each other through protein-protein binding, adding a new layer of biological control in the maintenance of pluripotency in stem cells.

It is somewhat ironic that the great scientist Alexander Graham Bell would have said…

When one door closes, another opens; but we often look so long and so regretfully upon the closed door that we do not see the one which has opened for us.

Because in science, things often work the other way around… He could have said “When one door opens, we often look for answers behind it so excitedly and for so long, that we forget to open other ones”.

Take for instance the regulation of stem cells by the so-called . Since their discovery, scientists have focused most of their attention and efforts on how these factors regulate each other’s expression at the .

However, there may be more to their interaction than just transcriptional regulation, which moved Silvia Muñoz-Descalzo, Pau Rué and a team of collaborators led by Jordi Garcia-Ojalvo and Alfonso Martinez-Arias to start looking behind other doors. In a recent Molecular Systems Biology article, they presented a model of a protein interaction network involving key pluripotency factors that could provide new explanations as to how pluripotency is maintained in stem cells [1].

Interest in how pluripotency factor proteins interact with other proteins dates back about a decade [2], and Muñoz-Descalzo, Rué and their colleagues knew that the interaction between some of these pluripotency factors had already been documented. They also knew that not all cells express the same amount of these proteins when cells are cultured under conditions that promote stable pluripotency. In fact, quantifying their amounts on a cell by cell basis reveals that their amounts are strongly . And this was something that the existing models of reciprocal transcriptional regulation could not easily explain.

So the authors hit the drawing board, and modeled a protein interaction network connecting four of the most famous and heavily studied pluripotency factors: Oct4, Nanog, beta catenin and TCF3. The basic idea is simple: each of these proteins are made at a given rate, they can exist as single entities or become part of protein complexes with partners (e.g. Oct4 and Nanog, beta catenin and TCF3, Oct4 and beta catenin, or beta catenin-Oct4-Nanog), and they are all eventually destroyed (something that a protein could avoid if involved in an interaction with another protein). By assuming random production (or “stochastic expression”) of these factors and adjusting a few parameters, the model would predict the relative quantities of each of these proteins in a cell. The authors call it the TBON model (for its component initials), and it works really well. How do they know? Because the model simulates relative protein abundances and correlation profiles that are very similar to those observed experimentally, not only under normal conditions but also when one or more components of the model are taken out, from a piece of paper or a real cell.

In some ways, a good model can feel like a finish line. “There were many aspects of pluripotency that couldn’t be explained by a transcriptionally regulated network”, say Muñoz-Descalzo and Rué, “but that can be explained by our network of protein interactions”. But in most other ways, a good model is just the beginning, a compass guiding scientists to do more experiments at the bench. First of all, scientists often make a series of assumptions to make the numbers work, and confirming them can be reassuring and instructive. Like a missing bone from a dinosaur skeleton, that is replaced with a piece of plaster. No matter how certain we are about the shape and size of a missing bone, it is always nice to one day find the bone resting right next to our peace of mind. Muñoz-Descalzo and Rué recognize that validating some aspect of their model won’t be easy. “One of the many problems when trying to address direct (or indirect) protein interactions between pluripotency factors is that, experimentally, one should check for interactions in mutant backgrounds. However, when doing this, pluripotency is lost, the cells are in a different state, they are not pluripotent anymore.”

Other times, the “holes” in a model highlight important unanswered questions. “Our model can only reproduce the TCF3 mutant state if we tweak it to include experimental observations”, explain Muñoz-Descalzo and Rué, “suggesting that we currently do not know enough about TCF3’s role in the maintenance of pluripotency and, specifically, the interaction between TCF3 and beta-catenin”.

But of course, scientists always expect a model to tell them new things. For this team the new insight was realizing that the interactions among pluripotency proteins may help the stem cells to keep their levels of free Oct4 in check. “The idea behind our work is that Oct4, when bound to Nanog and/or beta-catenin, cannot promote the transcription of its own targets”. And why would that matter? Because based on their , the authors propose that many genes targeted by Oct4, but not Nanog, are associated with differentiation. In contrast, genes targeted by both Oct4 and Nanog are mainly associated with pluripotency. According to the authors, their model’s prediction that the amount of free Oct4 in stem cells is kept low by Oct4’s interactions with other pluritpotency factors could nicely explain previous experimental observations made by their team and other colleagues. “In Oct4 overexpression experiments, cells differentiate. But when we increase the levels of beta-catenin at the same time that we overexpress Oct4, the cells do not differentiate”, explain the authors. According to their new theoretical framework, in Oct4 overexpression experiments “there would be free Oct4 to induce the expression of differentiating genes”, whereas the simultaneous overexpression of beta catenin protects pluripotency because “the excess of free Oct4 is and cells do not differentiate”.

As it turns out, pluripotency factors could also dance away from DNA. Scientist will now have to watch this new choreography closely for more clues and answers.

A new door has just opened. Where is the next one?

1- Muñoz Descalzo S, Rué P, Faunes F, Hayward P, Jakt LM, Balayo T, Garcia-Ojalvo J, Martinez Arias A. A competitive protein interaction network buffers Oct4-mediated differentiation to promote pluripotency in embryonic stem cells. Mol Syst Biol. 2013 Oct 8;9:694. doi: 10.1038/msb.2013.49.

2- Wang J, Rao S, Chu J, Shen X, Levasseur DN, Theunissen TW, Orkin SH. A protein
interaction network for pluripotency of embryonic stem cells. Nature. 2006 Nov 16;444(7117):364-8.

Sharing is caringShare on Facebook0Tweet about this on Twitter0Share on LinkedIn0
Stem cells can give rise to any fully differentiated cell with a specific function in an organism (muscle, nerve, skin, gut, etc). But stem cells are pretty dull themselves; they can’t do any of the cool things that their descendants do (like contracting, or transmitting nerve impulses, or digesting food). Stem cells just divide and make more copies of themselves, remaining in a so-called “undifferentiated” or pluripotent ground state. Core pluripotency factors are a small group of “transcription factors” (proteins that can bind DNA and regulate the expression of genes) that are necessary to maintain stem cells in such undifferentiated state. If you take them out, stem cells lose their self-replicating potential, and spontaneously differentiate.
Genes are pieces of DNA carrying a specific sequence of “deoxyribonucleotides” (the famous A,G,C,T). But cells don’t use the information stored in DNA directly. Instead, they make a copy of the DNA sequence using “ribonucleotides”, and string together a very similar type of molecule, called RNA. Scientists aptly named this process of copying from DNA to RNA “transcription”. Therefore, saying that pluripotency factors regulate each other at the “transcriptional level” means that they regulate how much of each other’s genetic sequence is transcribed, which is the first and necessary step to make a given protein.
This means that the amount of one protein is associated with the amount of another. If positively correlated, it means that the more of protein A, the more of protein B. If negatively correlated, the more of protein A, the less of protein B (or vice versa). One thing that scientists are generally quick to remind themselves is that correlation doesn’t mean causation. Therefore, that more protein A = more protein B doesn’t necessarily imply that protein A functions to make protein B (or viceversa).
The term “bioinformatics” applies to a wide range of cases in which computers are used to analyze biological data. In this case, the authors compared very large data sets that had been generated by colleagues elsewhere and deposited in public databases. The authors wanted to find out what genes may be targets of Oct4 or Nanog, so they used previous studies that analyzed changes in gene expression following the removal or increase in the expression either protein. Then, they compared these results with those from yet another group, who looked at DNA regions bound by Oct4 or Nanog. Using this combined approach, the authors identified genes that could be direct targets of Oct4, Nanog or both.
Titration is the name of a laboratory technique that scientists use to determine the amount of a given compound X in a sample. This is how it works: if we know that compound X reacts with compound Y in some detectable manner, we can add slowly increasing amounts of Y to a sample and follow its reaction with X. At some point, when X has been fully consumed by Y, the reaction will stop. If we know how much of Y we needed to reach this point, then we can infer the original amount of X in the sample. By extension, scientists often use the term “titrate” in their daily jargon to indicate that a given entity (in this case the protein Oct4) is sequestered away or neutralized (as X was in its reaction with Y in the example above).

Leave a Reply

Your email address will not be published.