If some science fiction writer doesn't pick up on this idea for a story about really alien forms of life, he's missing a good bet.
All life we know of on Earth depends on RNA or DNA, the long, ladder-like molecules that hold the sequences of three-letter words (each spelled with the four-base alphabet A-G-C-T) that serve as code for the 20 amino-acid building blocks of our proteins. No one knows how such a code developed in the first place, or why all earthly life uses essentially the same code. No one knows why all life strings the code along a ladder built of the sugar called ribose (the R in RNA) or its slightly altered cousin, deoxyribose (the D in DNA). Was it just the structure that fell into place first, like the QWERTY keyboard, and everyone kept using it from then on? There's no obvious reason why the ladder couldn't be built of other sugars. For that matter, there's no obvious reason why the alphabet employed by our genetic code for protein synthesis couldn't choose other letters from the unknown number of potential nucleotide bases of which our familiar A, G, C, and T comprise only four (five, if you count the U that substitutes for T in RNA). Further, there's no obvious reason why the code should limit itself to three-letter words and resulting vocabulary of 4 to the third power, or 64 words.
Obviously a 64-word vocabulary is sufficient to spell out some amazing complexity. There's no limit in principle to the sentences you can form with 64 words. You don't even need four letters to spell a lot of words, as binary computers attest. The point is, the code our DNA uses is not the only way to skin a cat. For instance, the amino acids that line up like pearls on a string to form our long-chain protein molecules number 22 in total, but of those only 20 are assigned a three-letter "code word" in our DNA. (The other two get added in by separate enzymes at a later stage in the protein synthesis process not directly controlled by a DNA transcription.) What's more, a number of other amino acids have specialized uses other than as beads in the protein string, such as for neurotransmitters or steps in metabolic pathways, but they are not assigned a three-letter DNA code word. So our genetic code has more words than it needs for the amino acids we use, but it uses up the redundancy in synonyms for some of them, while having no word for others.
So, back to the article I linked to above. The guys who fool around with this stuff are beginning to synthesize genetic molecules they call "XNA." These are still identifiable as nucleic acids, using a sugar and a phosphate for the ladder backbone and the familiar bases A, G, C, and T for the rungs, but they use different sugars from the usual ribose or D-ribose. Some of the alternative sugars turn out to be more structurally sound, standing up unusually well, for instance, under the stress of voracious enzymes and extreme pH levels.
Intrepid experimenters are even adding a couple of new bases to the usual A-G-C-T quartet, thus vastly expanding the code's vocabulary. I'll be very interested to learn how (and if) the surrounding cell mechanism learns to "read" the new words. I've never quite been able to understand even how the old words are read. Some sources I've read suggest that the shapes of the A-G-C-T code words are in some way physical cookie-cutter templates for the corresponding amino acids, but my impression is that that part is not well understood, and in any event I certainly don't understand it. It's one of my favorite mysteries.
All this work still sticks pretty close to Earth-style genetic molecules, of course, using a sugar-phosphate ladder backbone with bases for rungs. And yet sugars surely aren't the only way to construct a ladder, nor ladders the only possible structure on which to string a series of letters, nor a linear string of letters the only way to express and preserve a code. What works here needn't be what works best under different conditions. So I'm really curious to see how experiments in synthetic genetics come out. Because of my abiding interest in the origins of life, I'd love to find out more about how ordinary molecules could possibly have developed into active metabolism from dead-end equilibrium, and from there into replicating systems that take resources from the outside world and use energy to restructure them according to their own pattern. If nothing else, I'd like to see a better understanding develop of what kind of proto-molecules could possibly have developed into RNA, which, as primitive as it may be, is still an extremely complex structure and very, very far removed from the kind of chemical gunk you can generate from experiments designed to mimic primordial conditions.
I often hear casual statements to the effect that conditions on such-and-such a planet are "too extreme" to support life. I don't find a statement like that meaningful. Even on Earth in recent decades, we've found microbes thriving in extremely hot, cold, or poisonous conditions we'd confidently have called impossible until they were discovered. The assumption in the 1950s that life originated in shallow seas is giving way to the notion that it may have started in deep-sea thermal vents or in venues sporting other extremes of heat and pressure. We'd have to know considerably more about how life originated here before we could make any sensible statements about what it needs to get started universally, or about what sorts of forms life might take besides the ones we're used to.