Theoretically, Could DNA Work With Different Nucleotides?

Factual Questions
1

The 4 nucleotides in DNA are AGCT (adenine, guanine, cytosine and thymine). These, as I understand it, are each composed of a phosphate group, a sugar, and a nitrogen base…or something like that.

I’ve pretty well just told you everything I know about DNA, so please excuse me if this is a really stupid question.

What I was curious about was, in theory, could some other nucleotides work just as well instead of just those four nucleotides? That is, could DNA originally have been constructed with a different nucleotide than adenine, for example?

To rephrase, could DNA composed of xGCT (you fill in the x) work just as well as our AGCT-based DNA? Could our life have worked if it had started out with a different nucleotide combination, or is there something special about these four?

2

I haven’t taken biology since the 9th grade, but apart from uracil which is the thymine analog in RNA, IIRC, what other nucleotides are there?

FWIW,
Rob

3

[QUOTE=LiveOnAPlane]
The 4 nucleotides in DNA are AGCT (adenine, guanine, cytosine and thymine). These, as I understand it, are each composed of a phosphate group, a sugar, and a nitrogen base…or something like that.

I’ve pretty well just told you everything I know about DNA, so please excuse me if this is a really stupid question.

What I was curious about was, in theory, could some other nucleotides work just as well instead of just those four nucleotides? That is, could DNA originally have been constructed with a different nucleotide than adenine, for example?

To rephrase, could DNA composed of xGCT (you fill in the x) work just as well as our AGCT-based DNA? Could our life have worked if it had started out with a different nucleotide combination, or is there something special about these four?
[/QUOTE]
Base pairing works based on hydrogen bonds. A binds to T because both adenine and thymine have two sites in just the right places that are suitable to form hydrogen bonds. Likewise C-G, but in their case, there’s three sites. There are other bases. First of all, you have uracil (U), which is used in RNA instead of T; there’s nothing stopping you from attaching uracil to deoxyribose instead of ribose. It’s actually quite capable of forming hydrogen bonds with most other bases. There’s also various methylated bases, such as methylcytosine and methyladenine. In humans, DNA methylation is done dynamically as a silencing signal, but I could imagine a world with methylated bases as the default. There’s also weirder things like glycosylated bases (you get those in kinetoplastids), and entirely different things like 2,4-diaminopyrimidine and xanthine. Oh, and caffeine!

Base pairing, by the way, does not have to work in the classic A-T, C-G way (this is also called Watson-Crick base pairing). You can also have Hoogstein base pairs, which are based on the interaction of different hydrogen bond sites. This can help form weird structures other than the standard double helix - triple helices and quadruplex structures have been observed.

The different bases I mentioned before are still purines and pyrimidines, just ones that aren’t used for coding in humans. Can you escape purine/pyrimidine chemistry entirely? Well, current theory states that life on earth began with self-catalyzing RNA that acted both as information storage AND enzyme simultaneously. The purines and pyrimidines came in right there, in the oldest theorized organisms, before any sort of cells existed to separate life from environment. You’d need to come up with some sort of molecule that could both store and transmit information (in other words, something regular that could act as a template to make a copy of itself) and catalyze chemical reactions, most importantly its own replication. You’d need to be able to get the raw material for this molecule out of the environment with minimum processing - we’re right at the beginning of life here, the raw components must be produced abiotically. We know you can make purines and pyrimidines abiotically. As for anything else? Well, maybe, but I don’t know how. :smiley:

4

[QUOTE=LiveOnAPlane]
To rephrase, could DNA composed of xGCT (you fill in the x) work just as well as our AGCT-based DNA? Could our life have worked if it had started out with a different nucleotide combination, or is there something special about these four?
[/QUOTE]

It’s been a while since I read it, and I don’t remember the details, but I recall the chapter “The Genetic Code: Arbitrary?” in Douglas Hofstadter’s “Metamagical Themas” explores this idea (among others).

5

Ah, you’ve all pretty much gotten the gist of what I was wondering about (which is sometimes uncommon)-- the replies are dead on to what I meant, so thanks so much. Crescend was a font of knowledge, and I will try to find the tome **John T. Conklin ** mentions, it sounds interesting.

6

A mighty chemist named Albert Eschenmoser has synthesised pyranosyl DNA, where the ribose sugars (5 membered rings) have been replaced with pyranoses (6 membered rings). It has been a while since I read the papers, but I think it formed a stable double helix - the 4 bases were the same.

You also have synthetic peptide nucleic acids - PNA. These molecules have been well studied - the phosphate ribose backbone of DNA is replaced with a peptide, with the side chains being occupied by the 4 nucleobases. They form extremely stable double helices.

I’m reading a fascinating book at the moment by a guy called Cairns-Smith on his theory of the crystalline origin of life. The storage and transmission of information, plus catalysis, can all be observed in rocks - specifically clay minerals. He posits these clays held the first genetic structures, which were superceded by an organic-based information system. In a moment of madness I started a thread on it in GD a few weeks back - the SD ear was not at my disposal :stuck_out_tongue:

7

[QUOTE=The Cocky Watchman]
A mighty chemist named Albert Eschenmoser has synthesised pyranosyl DNA, where the ribose sugars (5 membered rings) have been replaced with pyranoses (6 membered rings). It has been a while since I read the papers, but I think it formed a stable double helix - the 4 bases were the same.

You also have synthetic peptide nucleic acids - PNA. These molecules have been well studied - the phosphate ribose backbone of DNA is replaced with a peptide, with the side chains being occupied by the 4 nucleobases. They form extremely stable double helices.

I’m reading a fascinating book at the moment by a guy called Cairns-Smith on his theory of the crystalline origin of life. The storage and transmission of information, plus catalysis, can all be observed in rocks - specifically clay minerals. He posits these clays held the first genetic structures, which were superceded by an organic-based information system. In a moment of madness I started a thread on it in GD a few weeks back - the SD ear was not at my disposal :stuck_out_tongue:
[/QUOTE]
Ah, right, catalytic clays. I know that RNA-replicating ribozymes have been made in the lab, and I know that certain clays have been demonstrated to catalyze RNA polymerization, but I don’t know how they would have encoded information. I could imagine that they’d act as a non-specific polymerization catalyst to boostrap RNA chain formation, but then the RNAs would compete on their ability to store information and catalyze their own polymerization. Then membranes would appear, and you’d get bags of molecules competing with other bags of molecules for resources, and next thing you know, you’re walking down to the corner store to buy some bread and jam.

Ah, progress.

8

Hydrogen bonds are geometric beasts. Typically, in DNA, the nucleotides in a sequence are linked through the 3’ and 5’ linkages of the attached sugars. Christopher Switzer has done some experimentation with changing the linkages to the 2’ and 5’ positions. Mixed results have been encountered from this minor modification. This change has produced mixed, but not completely unsuccesful results. For one thing, the resulting DNA can form both double and triple stranded helices. I can’t imagine that the triple stranded form is particularly valuable in coding protein sequences or whatever. (You will notice my biochemistry starts to run out here, so feel free to correct.)

IMO In the overall scheme of an abiogenic synthesis of life, I can’t imagine such an ability to be advantageous. In my opinion, at that level of evolution, the more thermodynamic stability to the molecule the better. A molecule that can order itself any which way is not very thermodynamically stable. Of course there are other factors, but the fact is DNA that is ordered with the 3’, 5’ connection is the one that made it. Nothing else did as far as I know. Even at the virus level.

My feeling is, that changing DNA even a little bit, results in something less stable and even less likely than what we have now.

9

I am getting some fascinating information here…thanks and feel free to add anything else to further my education on this.

10

[QUOTE=Crescend]
Base pairing works based on hydrogen bonds. A binds to T because both adenine and thymine have two sites in just the right places that are suitable to form hydrogen bonds. Likewise C-G, but in their case, there’s three sites. There are other bases. First of all, you have uracil (U), which is used in RNA instead of T; there’s nothing stopping you from attaching uracil to deoxyribose instead of ribose. It’s actually quite capable of forming hydrogen bonds with most other bases. There’s also various methylated bases, such as methylcytosine and methyladenine. In humans, DNA methylation is done dynamically as a silencing signal, but I could imagine a world with methylated bases as the default. There’s also weirder things like glycosylated bases (you get those in kinetoplastids), and entirely different things like 2,4-diaminopyrimidine and xanthine. Oh, and caffeine!
(…)
[/QUOTE]

Caffeine? You mean, it’s theoretically possible to have caffeine-based life form? Ie. with caffeine as one of the nucleotides? Now, that’s beyond cool.

11

The proposition that caffiene could be a suitable nucleotide comes from the basic realization that two of the essential nucleotides (Adenine and guanine) are based on the same purine structure as caffiene. If you want to extend things, you could actually say that the pyrimidine bases (thymine, cytosine, and uracil) have similar structures.

What is notably absent in any of the nucleotide bases are methyl groups on the nitrogens of the pyrimidine ring. These methyl groups will have severely detrimental effect on the hypothetical nucleotides ability to hydrogen bond effectively. Since the double helix is dependent on hydrogen bonding at or near these positions, caffiene is not suitable as a nucleotide.

12

[QUOTE=puppygod]
Caffeine? You mean, it’s theoretically possible to have caffeine-based life form? Ie. with caffeine as one of the nucleotides? Now, that’s beyond cool.
[/QUOTE]

I imagine such life would move fast and not sleep much.

13

[QUOTE=WarmNPrickly]
The proposition that caffiene could be a suitable nucleotide comes from the basic realization that two of the essential nucleotides (Adenine and guanine) are based on the same purine structure as caffiene. If you want to extend things, you could actually say that the pyrimidine bases (thymine, cytosine, and uracil) have similar structures.

What is notably absent in any of the nucleotide bases are methyl groups on the nitrogens of the pyrimidine ring. These methyl groups will have severely detrimental effect on the hypothetical nucleotides ability to hydrogen bond effectively. Since the double helix is dependent on hydrogen bonding at or near these positions, caffiene is not suitable as a nucleotide.
[/QUOTE]
Yeah, you’re right. Maybe you could have some sort of alternative base-pairing mechanism here that relies on hydrogen bonding via the imidazole ring and a non-double-helix structure…okay, I’m grasping at straws. Thanks for catching that error. :smiley:

14

Meh, I tend to be a bit of a wet blanket around here when people start speculating about how evolution/abiogenesis coulda shoulda woulda happened. If only nature had SciFinder and Hartree Fock, any number of concoctions could’ve worked. The fact is, the only nucleotides we know that worked naturally are the only ones that we find naturally. Anything else is pretty wild speculation IMO.

[QUOTE=The Cocky Watchman]
In a moment of madness I started a thread on it in GD a few weeks back - the SD ear was not at my disposal
[/QUOTE]

Too bad for me, it’s just the sort of thread I like to lurk over and a subject I’d like to learn more about without actively reading. I don’t get involved in GD though so I’d let it sink.