Skip to main content

Michigan Scientists Model Cell Signaling, Observe Bi-Directional Signal Transmission

Researchers at the University of Michigan and the Université de Nice Sophia-Antipolis in France have developed a model of an intracellular signaling cascade.
The investigators, who said they believe the model is the first of its kind, said that they have identified new properties of these signaling cascades. 
A key feature of this model is that a negative feedback is exerted between each cycle and its predecessor. The system displays damped temporal oscillations as a result of this negative feedback and, to the surprise of the researchers, propagates perturbations both forwards and backwards.
This bi-directionality challenges the widely held idea of unidirectionality in cellular signaling cascades, the researchers said.
The findings could also have significant implications in the understanding of cell signaling, the investigators said, because signaling pathways are often reconstructed from such experimental data.
The researchers published their work on March 21 in PLoS — Computational Biology.
Corresponding author Sofia Merajver, a professor of internal medicine at the University of Michigan, and scientific director of the breast cancer research program at the University of Michigan Comprehensive Cancer Center, spoke with Cell-Based Assay News this week about this new model and its applicability to drug discovery. 

Could you provide a little background on this work?
One of the problems in systems biology that people are talking about right now is the integration of cellular signaling pathways, and how to determine how these pathways interact with one another.
A few years ago, we got the idea that we wanted to understand how the pathways really transmit information. Why is the life of the cell organized around transmitting information from one cycle of covalent modification to another cycle to another cycle, et cetera, in a chain?
A covalent modification could be phosphorylation, for example, or any kind of modification that happens to a protein that instantly changes the state of the protein from inactive to active.
The cell finds itself having to probe the environment, and receive all of these signals from the environment. These signals interact at the cell surface, and then are transmitted to the interior of the cell.
From the point of view of a single molecule, no matter how large the molecule is, the distance between the surface of the cell and the nucleus is astronomical, because the molecule is so much smaller than the cell. Yet, all that information often has to be transmitted in a short period of time.
In a much deeper sense, we really do not understand the basic tenant of life, or “How is a single cell transmitting information?” So you have tiny molecules that are the only way that we know of for cells to transmit light. The only way that they transmit information is by a chain reaction or activation of subsequent molecules.
In the hopes of better understanding cellular pathways in cancer, I started a program in my lab to study how this information is transmitted. I am a quantitative scientist by training, so I was comfortable with attempting this on my own, but I hired a postdoctoral fellow, Alejandra Ventura, who is the first author on this paper.
She also worked with Jacques-Alexander Sepulchre in Nice, France. The three of us worked out the details of this signal transduction in a cellular pathway. What we found was surprising!
When information is transmitted forward from the cell surface to the nucleus, it is also transmitted backwards towards the surface. The point of origin begins to get information as well, regarding the fact that the information is moving forward. So the pathway is bi-directional.
The interesting thing is that now you have two cellular pathways talking to each other. For example, halfway through, at connection five out of ten, the two pathways interact. We call that cross-talk.
That cross-talk, like all stimuli that is transmitted by this type of a pathway, will also go both ways. No one has ever looked for that.
We are hoping that because of this work, researchers will return to their laboratories. I believe, and I know in my own lab that we have such data, that many labs worldwide may have such data, but it is not published.
It may look like bad data, like something that did not fit the hypothesis or did not fit the model, because researchers never looked at it that way. That may have something to do with why targeted therapies do not work as well as anticipated.
We never really had this kind of understanding about where it may be best to inhibit a pathway if it is one of several pathways that are linked together, as they often are in disease states, or in normal states but in disease states as well.
So you inhibit the pathway at a place where that inhibition is really not fully effective. It would take a lot more drug than should be necessary to inhibit the pathway. That would mean you have to give such a high dose of the inhibitor that it would be toxic.
This would lead, I hope when its development advances, to more of a roadmap of the complex connections between pathways and how to best inhibit them. We are working on this now.
How is this applicable to drug discovery?    
That will have to wait. That is the main application that I believe it will have in the immediate future. We will see how it works out in terms of when people start using it.
To give you a simple example: if you inhibit a pathway at say, step seven, you think that you will see the full effect of that inhibition. Then let’s say that there is another pathway that interacts with your pathway at say, step nine, so downstream of where you are inhibiting.
When that second pathway interacts with the first pathway, the information to further stimulate that pathway will be transmitted right back to the beginning of the first pathway.
If you inhibit downstream, then, it will take much more drug, because now you get the stimulation from the crosstalk that you were not expecting to go backwards.    
If the other pathway is stimulating, and that stimulation is being transmitted backwards and forwards, then it would take a lot of drug to inhibit that first pathway. This is something we see in the laboratory all the time.
We think a compound is going to inhibit a cellular signaling pathway, and it does, but it is not great inhibition. There must be a crosstalk pathway somewhere.
But we are always looking for a crosstalk pathway from above the point of inhibition, because we always thought that the information was transmitted uni-directionally.
What we know so far is that most crosstalk really does not happen far up in the pathway. When the information first starts traveling down the pathway towards the nucleus, it travels quite a bit of distance before it interacts with another pathway.
As they get closer and closer to the desired effect in the nucleus, the pathways become more and more entangled.
This is what we are working on right now — producing a whole bunch of models that exhibit crosstalk and the signals that are known to exist. We tell people, “If you try to inhibit the pathway here, it will take 10 times more drug than would be safe, but if you inhibit it here, it would take one-tenth of the safe amount of drug. Why don’t we go for the place where you are not going to have to give the patient so much drug?”
This is just the beginning of looking at the influence of crosstalk on the quantity of drug needed to inhibit a pathway. Qualitative reasons to inhibit the pathway at one place or another exist that we have not yet untangled.
What do you see as the next step in this research?
I think the next step would be to have more experimental data modeled in this manner, to see if we can explain it, and also to see if we can detect the backwards propagation of cellular signals along the pathways.
We are doing some of those experiments in our lab, and also in collaboration with other labs and I would hope that this would be undertaken in other labs to really see if there is a large repertoire of theoretical parameters by which this phenomena will be observable.
I think that many experiments can show this. The problem is that some of these experiments are technically challenging, so those who have the necessary equipment and resources, I hope, will do these experiments, while others would need to be set up.
That is the problem that we have in our own lab, which is why we are collaborating with another lab. It may be easier to do these experiments in yeast or in Escherichia coli, in simpler systems than mammalian systems. 
This work is described in the context of oncology. Would this model be applicable to drug discovery in other areas of research?
Yes, absolutely! All drug discovery depends on inhibiting something. Usually, it’s a pathway.
We did not publish this work in a cancer journal because it is of general interest — it is basic biology.

The Scan

UK Funds to Stay Ahead of Variants

The UK has announced a further £29.3 million to stay on top of SARS-CoV-2 variants, the Guardian reports.

Push for Access

In a letter, researchers in India seek easier access to COVID-19 data, Science reports.

Not as Cold

Late-stage trial results are expected soon for an RNA-based vaccine that could help meet global demand as it does not require very cold storage, the New York Times writes.

Genome Research Papers on Microbes' Effects on Host Transfer RNA, Honeybee Evolution, Single-Cell Histones

In Genome Research this week: influence of microbes on transfer RNA patterns, evolutionary relationships of honeybees, and more.