Hi Tom,

There are a lot of ways cells can manage this problem. You're right, feedback loops are common. At every step in the production of a protein (transcription, translation, processing, etc.) there can be inhibition or activation depending on certain conditions - one example is if the product of some enzyme A binds another protein that blocks the transcription of the gene for enzyme A, so that when the cell has enough activity of this enzyme it stops making more. It would be really hard to monitor every single protein individually, as you can imagine.

One general way cells control protein levels is just that there's a lot of machinery in the cell to degrade proteins, either specifically recognizing 'defective' ones, or just constantly degrading anything, to prevent them from becoming defective. A really simple model (with no specific feedback at all) could say a protein is being produced at a constant rate and it's being degraded at the same rate - the cell will end up with a constant level of protein, and any single protein will be around for some average lifetime before getting degraded.

Check out this wiki for more details and references: http://en.wikipedia.org/wiki/Gene_regulatory_network

The regulation of protein activity extends beyond just expression though. You can also have regulation at the protein level. For example, blocking the action of the protein without degrading it (because it took a lot of energy to make). This could be through allosteric regulation for example, where a small molecule binds to the protein and makes it work or stop working.

As for detecting when proteins are broken (or more broadly when they should be disposed of). Sometimes it is because proteins may have a specific half life depending on their sequence and structure. Thus if more isn't made, it will all be disposed of eventually. Sometimes they can be specificaly targeted for modification by attachment of other chemicals like ubiquitin or oxidation due to some environmental stimulus. Once modified, the protein is immeditely targeted for disposal.

It is an immensely complicated but interesting area to be interested in.

Barabasi, among others, has addressed the problems of understanding the complicated nature of network equilibrium in several of his papers by mathematical modeling and systems treatment of metabolism. He shows that metabolic networks have topological scaling properties and organization similar to complex non-biological systems. He also shows that metabolic networks, made of proteins and metabolites, follow the power-law relation and hence are robust and resemble error-tolerant scale-free networks. To start with, you can refer his paper on "The large-scale organization of metabolic networks "http://www.nature.com/nature/journal/v407/n6804/full/407651a0.html"

In my basic knowledge under general biophysical concept, we would easily think of this question in terms of both protein and cell altogether.

Generally, the universe, especially each of species, has a principle that within the limitation of the reactant and product there has steady increase in entropy. Focusing on the regulation of protein synthesis in a cell, we can say that protein synthesis occurs continuously if there has plenty amounts of ingredient, reactants, and certain conditions for reaction.

Of various conditions, I would argue that the maximum range or density of cell dimension, to maintain the highest effieciency of protein synthesis, is important for regulation of protein synthesis. According to this condition, we need to care about the available amount of nutrient or equipments for translation within the limiting cellular dimension. If we consider that the dimension of cellular surface affects proportionally the probability of reaction rate(i.e the number of encounter between reactant and substrate), the entropy of universe(system + surround) increases when the dimension of surface is relatively more.

Still, as you mentioned, delicate feedback system works worthly in order to regulate the synthesis in any part. Here, as an additional way, I do suggest the point of interaction between subject and environment.

I hope this letter would be faithful opinion for your brilliant focus.

Sincerely,

Chung Mingin

@Mingin,

Thanks for your answer. But I don't think it addresses my question. Reaction kinetics is one aspect of the problem, but not the main one, for cell physiology.

What I am interested is to figure out the Informational Logic in protein concentration control. The binding affinity to promoter sight is an issue, but that can be set by the information of binding affinity (how strong the binding is, that is within the sequence).  But that will lead to a steady-state of a protein concentration.  But that is not sufficient. For example, if I increase the temperature, and this protein is destroyed by heat, how can the cell compensate that? How the information gets into the gene expression level? Which protein is counting the protein concentration within the cytosol, and then tell the promoter to start binding and induce production again. Cell is by most an information system. Without a signal, there will be no gene expression.

I am looking for solid example of gene pathways that achieve this FINE regulation of protein concentration, with gene names, and regulation mechanism.

Tom, I'm am not aware of a single resource that can give you all the information you require here. I suspect if one exists it is a textbook, which means out of date by the time it is published. Re: specifically the issue you raise with temperature. Damage to a protein, for example through heat, can be compensated in two main ways. The damaged protein can be disposed of and more made to replace it, or the protein can be unfolded and re-folded using heat shock response proteins. I am afraid specifics of genes and flux control, etc etc (i.e enough to build a model of) you might need to to start looking for people working in systems biology and artificial life projects.

Also, I am generally not aware of a specific 'sensor' protein. Protein concentrations and activities are generally regulated by balancing between opposing forces, rather than specifically a sensor unit. Although I'm sure there are exceptions and I am stepping well outside my field in saying this.

@Jason Smith: "Protein concentrations and activities are generally regulated by balancing between opposing forces, rather than specifically a sensor unit. "  I am afraid I cannot agree. We don't know how it is regulated, that does not mean we cannot figure it out. I am thinking on that. A cell has to be smart enough to know how to control the concentration of each protein at each precise moment.  Synthetic life is a good way to start on the problem. If you are a programmer, and you have to write the code/operating system/genome for a synthetic cell, how are you planning to make the cell behave correctly? I mean you need to have the right amount of different proteins in a concerted way. If you come up with a good answer, then we may be not that far from how a cell does it.

@Tom, I say that because the concentration of the protein is not generally relevent to the cell's function (in my opinion), but rather the level of the protein's activity. Therefore the 'sense' portion of the equation is at the 'effect' level of the protein, rather than at the specific molecular concentration level. I certainly leave room for their to be a specific molecular recognition sensor present to monitor the level of a particular protein. But that sensor must also be regulated, so I wonder if we delve deep enough into the regulation of that sensor, whether it would also simply be regulated by virtue of the level of activity of the protein it is ultimately trying to sense. I would certainly be interested to learn of a regulation system that monitors the concentration of a protein independtly of its level of activity. I would assume not an enzyme but perhaps a transmembrane signal transducer or a structural protein.

Although it is very complex, we do know a great deal about gene regulation, and systems biology is partly dedicated to the study of regulatory circuits of genes and proteins. For instance, in bacteria, you will find several genes for proteins in the same metabolic pathway lined up in sequence on the chromosome, forming what is called an operon. Binding of a transcription factor (TF) to an upstream operator triggers transcription of all these genes. The lac-operon, which contains the machinery for lactose metabolism, has been extensively studied for decades and is a model system for gene regulation in bacteria. One of the proteins in the operon, LacI, acts as a lactose sensor, and in absence of lactose it represses transcription of the operon by binding to an upstream promotor. As such, it forms a negative feedback loop when no lactose is present. In presence of lactose, LacI has little or no affinity for the operator and the operon is transcribed. In this example the expression level is determined by the amount of lactose present (which determines the fraction of active LacI) and the inherent affinity for the operator (which determines the "leakiness" of the operon, that is, how often the operon is transcribed in the absence of lactose). See Jacob and Monod's work from the early sixties, briefly summarised in this Nature piece (http://www.nature.com/milestones/geneexpression/milestones/articles/milegene02.html)

The regulation is far more complex in eucaryotes, but what is important to realise is that the protein concentration need not to be sensed directly, but in many cases something else can be used as a proxy, for instance a metabolite. In addition, the stability of the promotor-TF complex leads to a certain macroscopic transcription rate, which at a protein certain concentration equals the degradation rate of the protein, yielding a steady state concentration of the protein. As such, the promotor-TF complex can be tuned by evolution to give a concentration that matches the needs of the cell.

 Eukaryotic cells use signalling pathways to regulate protein concentration/ activity.  Proteins can be regulated by a number of effector molecules.  For example ubiquitin tags proteins for destruction (ubiquination), kinases generally activate proteins via phosphorylation, phosphatases deactivate/ inhibit (generally) via dephosphorylation, and there are a bunch of other pathways as well.  Usually enzyme/ protein activity are affected by the molecules they are catalyzing/ synthesizing.  For example the enzymes involved in glycolysis and glyconeogenesis are intimately influenced by the surrounding levels of ATP, ADP, and cAMP.  Genes are regulated genetically (through the mediation of different activation, enhancement, etc sites) epigenetically (through cytosine methylation and histone acetylation, and remodeling), and by transcription factors. There are great examples of regulatory genetic expression as it relates to protein concentrations in circadian pathways.   Here is an awesome paper that I think hits exactly what your question is asking. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2841896/). I apologize for my response not being very developed, I'm kinda strapped for time.  I hope it helps. 

@Erik Marklun: LacI, yes! That is a great example. Many thanks! 

That is one mode of the regulation, and it is in bacteria. Do you know for higher animals, if there are good examples? If it is a general rule, then we can expand. Our goal would be to find general rule of cellular behavior concerning protein level/gene expression regulation.

 Some relevant subjects:

Lac operon

Typ operon

Cholesterol metabolism

renin/angiotensin path ways.

The last 2 being "traditional" feedback loops, and the first 2 being slightly unique. I would also suggest looking into retinoic acid metabolism by CYP p450s, and is nuclear receptors RXR, RARa, RARb