Why did evolution favor ATP and not GTP, TTP or CTP?

Your question is interesting but not easy to answer. Probably an evolutionarily 'young' enzyme used ATP as a co-factor and not another NTP by chance. With time orthologues of this hypothetical enzyme arose through gene duplication, for instance, so the use of ATP became more widespread. Probably ATP was the first NTP that could be synthesized by early cells...?

Best regards.

I think the chemical structure of adenine involving in the synthesis of ATP in the mitochondrial ATP synthase pump is more favourable than that of the quanine. This ATP synthesis is coupled with that of the oxidative phosphorylation and respiration functions of the mitochondria, which we learnt from biochemistry. But GTP, CTP and TTP syntheses were seldomly coupled with mitochondria as we learnt.

Wow what an interesting teleological question? It is a tough questions to ponder as well. I had many times wondered about it as well. Here are some thoughts:

1. Perhaps a chance factor?  Although it is easier to make pyrimidine ring/pyrimidines just with 2 or three components (Aspartate, HCO3, ATP) and Glutamine.

Now let us ask: Why Purine (Adenine)?

2. One possible factor may be inosine synthesis involves Glycine. As a simple aa perhaps there was more of Glycine in the prebiotic era that favored inosine<adenine<guanine synthesis. For making Inosine/adenine you need: 1C-folate, Glycine, Glutamate (can be easily obtaied by transamination of alpha-KG), Aspartate (can be easily obtained by transamination of OAA).

3. Now remember for making pyrimidine you need ATP not GTP.

4. In addition  many cofactor seems to have this handle that is made of Adenosine (NAD, FAD, CoA, PAPS etc...)

5. With all these put together it was easy to synthesize Adenosine/ATP first because of the available prebiotic ingredients such as Glycine, CO2/HCO3, Folate, Glutamine and Aspartate.  In all it might be just Glycine which started the whole sequence of events.

My thoughts lean more with the "chance" theory argument, presenting my points: 

1) I agree with both Christian and Venkatachalam in that a Chance event might have been the first choice! I strongly tend to support the hypothesis that the molecule came about first before we were able to evolve the proteins that synthesize it (therefore I tend not to go by Zhou's hypothesis where metabolism precedes small molecule evolution). This makes chance event a much simpler iteration towards ATP.... though if iteration is simply the answer, we cannot fathom as to why a single ring cytosine or thymine couldn't be used as against a double ring ATP/GTP (there's more than meets the eye). 

2) Especially if we look at the prebiotic soup and about its constituents, I like Venkatachalam's suggestion which is that Glycine could've started the whole process via via intermediates that led to ATP and other nucleosides. 

3) I think the theme of Gene Duplication (which Christian suggests as well), seems to be the best explanation as we observe time and again.... that a model once built (by chance) is repeated by biology over and over again (several examples of Domains being repeatedly used in varying combinations and so on), and thats how it came about to be the dominant energy storage of the cell. 

4) GTP did evolve as well eventually.... by alternate use of the enzymes/processes by which ATP came to be, and followed suite and hence is used in fewer enzyme processes. (energentically ATP and GTP may not be very different?) 

5) Lastly, The simpler model is to evolve TWO separate energy molecules:

ONE as a powerhouse (ATP) and

SECOND as a sensor (GTP) and

other minor ones like cAMP/cGMP and ddTTP and such.

This model gives better modulatory control over pathways, rather than have to intermingle and make it extremely confusing chemistry to figure out when ATP would be a sensor versus a powerhouse.  

would welcome added criticisms! Thanks! 

Good discussion! I will pass it to my students!

Dear All,

I do not believe in "By Chance Theory" at all, It was used only where people cannot come up with a good reasons of explanation.

Please look at and compare the chemical structures of adenine and ATP vs guanine and GTP and also consider the evolution of adenine and guanine molecules at many billion years ago on our Earth. At that time and before that, the oxygen level on Earth was below 0.1% as compared to 20% nowaday. Therefore, adenine was far earlier than guanine appearred on Earth by billions of years on early Earth. It is because, adenine has no oxygen atom in its structue, whereas guanine has 1 oxygen atom in its structure. Therefore, the amount of adenine molecule was then (pre-historic of our Earth) and  is now (oxygen 20% and nitrogen 50%) far more than that of guanine.

When more oxygen and phosphorus were formed billions years later, then ATP and GTP were formed; but their amounts were a matter of several orders of magnitude different between the two, that ATP is a lot more than GTP.

Very interesting question indeed. I am considering two folds, namely, the prokaryotic and eukaryotic cells. The "power house", the mitochondria, of Eukaryotic cells was proposed to come from the endosymbiosis of an early prokaryotic cell. The mitochondrion is responsible for generating majority of the ATPs.  The endosymbiosis process of such prokaryotic cell is for sure a probabilistic event, which depends on the metabolism of the symbiotic prokaryote. For prokaryote energy metabolism at the time before eukaryogenesis, due to the diversity of ecological adaptations of prokaryotes there are possibly both types existing, but the majority is ATP using type. So the question finally rests at: why at the time before eukaryogenesis the majority type is ATP using? As some other comments mentioned, is it due to the earth environment at the time?

@MH Zhou Didn't the earliest forms of life (the RNA world) use RNA, which contains all four base pairs, not just adenine? So how could adenine have come before guanine?

I think now we should think out the problem from the protein perspective. RNA came first: Which means during early RNA world we already had ATP, GTP, CTP and UTP to make RNA.

As I had suggested earlier the ATP binding protein fold was easier/most suited to have been made during the course evolution.

For e.g. tRNA synthetase I from bacteria uses ATP and splits it between alpha-beta position to yield ~7.6 kcals of energy. In addition the Ubiquitous  pyrophosphatase would cleave the PPi into 2Pi releasing additional energy. 

Then came the protein fold that split ATP between beta-gamma position.

So I am tempted to say that ATP for energetics and GTP for signaling (came about due to shear chance of protein folding). After all with the trial and error, some sort of evolutionary consensus came about which at the end resulted in ATP for energetics and GTP for signaling.

CTP, TTP and UTP being monocyclic was not selected because of loose amino acid interaction that is required for binding and would perhaps incur non-specificity and poor binding.

I guess we have to give more importance to the time factor, trial and error, chance event, reaching out on consensus and eventually reaching the ultimate universal perfection. Once the perfection is made, that motif is common among all the organisms from archaea to eukaryotes.

The decrease in entropy at the microcosmic level?

Am I making sense? 

Uracil has two oxygen atoms in its chemical structure, therefore, its evolution had to occur a lot later than Adenine, which has no oxygen atom at all. And the amount of Adenine molecule was the most abundant one amongst U, G, T, and C, during the earliest period of Life Evolution. I am kind of agree with the idea that pyrimidine has only one ring in its structure and is not specific enough as compared to 2-rings-purine to develop signalling bonding abilities.

I don't know when, how or why one is chosen over the other but it's good that we have two different NTPs, one for energy and one for signalling. If we solely depended on ATP/GTP for everything that goes on in our cells, we would run out of it quickly I think. So it's good that we have separate stocks and one can be converted into the other when necessary..

thank uu to alll

It is interesting to note from Filiz Korkmaz above, that the evolution of ATP was related to energy first before life development in the earliest time at 4.2 billion years ago, while the evolution of GTP was later because it needed oxygen atom in the GTP molecule (remember, the oxygen level in the atmosphere was very low at that earlier time!). That means the signalling function of cells was developed much later than the energy production in the mitochondria.

Both ATP and GTP act as both energy currency and signal molecules (i.e. there are signal pathways that form cyclic AMP to carry signals, and signal pathways that form cyclic GMP, etc.). Structurally, these two molecules perform each other's functions in many different circumstances.

As to why ATP ended up heavily favored as an energy currency as opposed to GTP, which is more favored in common signaling pathways, I can only offer two observations that might be factors:

The body needs to have some way to use energy for signalling and to use energy for other purposes. Since GTP and ATP are both present at the same time, it makes sense to specialize one for one task and the other for the other.

In DNA and RNA, Adenine forms only two hydrogen bonds with the pyrimidines T and U whereas Guanine forms three with C. Therefore, if you simply had to choose which nucleotide to use for bulk energy transactions between A and G, it would make sense to choose A because there is slightly less energy required to separate it from its binding partner in structures (which would be a benefit when you want to isolate it and use it as an energy molecule).

ATP - is Bitcoin! ATP is considered as cellular currency for a long time. I realized that ATP is Bitcoin - cellular cryptocurrency, along with other NTPs (GTP, CTP, UTP, TTP), few others cryptocurrencies. And it uses same network and blockchain technology, this technology is used in nature for millions of years! By comparing to biological blockchain new features and possibilities of cryptocurreny can be revealed.  😃

It's a very nice discussion. Each expert's own perspective provides the closest results to the truth. The behavior of the molecules, a little bit of chance, enough time, and the result is incredibly molecular organization in this catastrophic environment. Very complex structural and functional relationship networks. I think ATP had started a step ahead in this area.

Clearly the relative solubility of the nucleobases adenine and guanine is relevant! For life to focus on adenylates as the primary energy transducers, the solubility of adenine has to be tractable. Guanine is practically water-insoluble and so is a dubious choice for a base constituent of the central metabolic pool.

This is soo interesting..But then why GTP is utilized and not just ATP in Kreb's cycle,protein synthesis,gluconeogenesis,signal transduction in G-proteins etc..

Nirali, ATP can vary over larger magnitudes of concentration than GTP. So the inference is probably because, if a model where metabolically highly variant energy sources (ATP) versus a less variant signalling source (GTP), if separated in utility, provides a more robust/responsive system for

regulation. I must say that this inference is more speculative.

That different NTPs involved in different processes - something is to do with energy distribution and thermodynamic independence. Say, GTP/GDP ratio controls protein synthesis independent of changes in ATP/ADP ratio. Energy of GTP hydrolysis (deltaG GTP) apparently is higher than ATP, such energy is required for efficient protein synthesis. In signaling systems there are GTP-sensitive G-proteins as well as independent ATP-sensitive - A-proteins (e.g. K-ATP channel). Similarly with other NTPs - they provide energy to specific processes which can be regulated independently and have thermodynamics sovereignty. Number of phosphotransfer enzymes regulate delivery of specific NTPs to distant process. Presence of phosphotransfer enzymes in the vicinity of signaling or synthetic enzymes guarantee rapid regeneration of spent NTPs maintaining high local level. It is known that NDPK regulates GDP-GTP conversion and G-protein signaling. In cancer cells energy distribution is uncontrollable, as they shift to glycolytic metabolism, which can spread ATP everywhere in the cell including nucleus, and have deranged phosphotransfer system.

I agree with Petras: different pathways use different energy carriers (not only nucleotides, but also for example pyrophosphate and phosphoenolpyruvate). One could imagine that these pathways developed independently, and came together when the modern cell arose from more primitive predecesors by endosymbiont formation and horizontal gene transfer. The possibility to regulate these pathways independently is then an added bonus. Of course, that is speculation as we weren't there to observe 4 billion years ago.

Wow this is a very informative debate about the favourable use of ATP over other NTPs.

Here I want to say two things:

First the free atmospheric oxygen has no role to give a preference to ATP then GTP. Because we need to keep in mind that oxygen atom of Inosine, the precursor of Guanine and Adenine can come from CO2 or dissolved CO2 (HCO3) or can be contributed by H2O.

Second, if we think about biosignaling here also ATP out competed other NTP in the form of cAMP, which can sense the status of energy in the cell and at the same time, can play a significant role in signal transduction. But other molecule including Ca2+ , cGMP, IP3 have a complimentary role for signalling.