Most nucleic acid species have specific structures (or a limited number of alternative structures) which may involve a greater or lesser amount of double helix through complementary base pairing. The extent of such helical components depends primarily on the availability of sequences of complementary bases, which is obviously greatest for the two distinct strands of a double-stranded DNA or RNA genome, but also occurs in a specific manner within RNA species where base-pairing has been evolutionarily conserved in particular regions.
Two ways of interpreting “why?”
When responding to questions about biomolecular processes and structure, I think it important to differentiate between two meanings of “why”, and I shall distinguish these by appropriate numbering in my answer.
1. The mechanistic “why”
This is generally easier to answer objectively, as it is request for a physico-chemical explanation for the formation of a structure (RNA or RNA in this case) or operation of a process.
It is worth emphasizing that just as different macromolecules of the class “protein” can adopt different particular structures — globular or fibrous with many variants on each — so can different RNA, and even DNA, species, as described below.
2. The functional “why”
By this I mean “what function(s) can a particular structure serve?”.
This is more difficult and beset with dangers. Whereas, we can say that the structure of myoglobin allows it to bind oxygen and release it as the partial pressure in tissues falls, with some macromolecules the function of their structural features may be less clear. The dangers here are in thinking that suggestions of function are facts, and that a particular function can only be accomplished or is best accomplished by a particular structure or structural feature.
The physico-chemical principles governing structure of macromolecules
Put simply, the structure adopted by a macromolecule in a particular environment is that in which it has the lowest thermodynamic energy (Gibbs Free Energy) in that environment. This is the structure in which the sum of the individual interaction energies (generally hydrogen bonds and Van der Walls interactions) is greatest.
Thus, a macromolecule is able to adopt alternative structures when these structures have similar energies, and the equilibrium between such structures can be changed by action of or interaction with other molecules.
Genomic DNA Structure(s)
Double-stranded DNA (dsDNA) genomes
1. Double-stranded genomic DNAs form an anti-parallel helical duplex structure (double-helix) primarily because they consist of two perfect complementary sequences that allow maximum hydrogen bonding of A–T and G–C base-pairs. The helical structure of the duplex allows the maximize the energetic π– π interactions of the stacked bases.
2. The function of genomic DNA is to maintain the genetic information of an organism and allow it to be transmitted to daughter cells. The double-stranded structure protects the bases from external modifying influences, to some extent, and allows the semi-conservative replication of the strands when a cell divides.
Single-stranded DNA (ssDNA) genomes
Some small viruses of bacteria and eukaryotes have single-stranded DNA genomes, illustrating that this particular function can be served by different structures.
1. Although the replication of ssDNA viruses involves production of a strand complementary to that of the genome, the genome is not double-helical — negating the generalization in the question. The simple reason for this is that the replication mechanism generates many more copies of the genomic DNA strand than the other: complementary nucleic acid sequences are required to form a double-helix! One could imagine a ssDNA having local regions that could base pair to one another, but this is prevented by the compact supercoiled structure that they adopt, which is a consequence of their closed-circular nature.
2. The compact super-coiled structure of such ss genomic DNAs (e.g. φX174) is suitable for encasing in a viral capsid (although ds circular viral genomes also supercoil).
The question of why some viruses have ssDNA genomes whereas most have ssDNA genomes is neither a mechanistic nor a functional “why”. It may be worth discussing, but falls into the realm of hypothesis.
Genomic RNA structures
1. To a large extent, the remarks above about the structure of genomic DNA apply to the genomic RNA of viruses. Genomes of dsRNA viruses have an anti-parallel helical duplex structure because two perfect complementary sequences are available in equal proportions from replication. Genomes of ssRNA viruses do not have this structure because replication generates primarily one type of strand. The ssRNA viruses, unlike the ssDNA viruses, have linear, rather than circular genomes.
2. I have nothing to say about the relation of these structures to function. Both are packaged into viral particles. Both can be replicated.
Transfer RNA (tRNA)
1. The single-strand of tRNA has a clover-leaf stem-loop structure with three short double-helical stems formed by Watson–Crick base pairing, folded into a three-dimensional L-shape by other hydrogen bonds. The sequence of bases in tRNA determines that this self hydrogen bonding can only occur in a manner to produce the particular structure.
2. The purpose of tRNA is to bring an amino acid onto the ribosome in response to a particular mRNA codon. The non-hydrogen bonded loop at one end of the folded molecule contains the anti-codon, allowing it to interact with a complementary mRNA codon; the free 3′-end at the other extremity is covalently attached to an appropriate amino acid. The compact hydrogen-bonded ‘L’ is of a shape that can be accommodated in the A or P site of the ribosome.
Ribosomal RNA (rRNA)
1. The single strands of the two major rRNAs are folded into a complex of stem-loops formed by short stretches of internal double helices involving either Watson–Crick (A–U, G–C) or single hydrogen-bond G–U base pairs. There is conservation of base pairs in the helical regions, without necessary conservation of particular bases. This determines that a specific structure is formed.
2. The double-helical stems are thought to play a structural role in the ribosome. The unpaired regions are free to interact with the various substrates of protein synthesis on the ribosome, and the catalytic centre is constituted of such unpaired bases. In this respect the ribosome represents catalytic RNA in which double-stranded regions have a functional role but the catalytic bases need to be unpaired.
Messenger RNA (mRNA)
1. mRNA is generally thought to lack hydrogen bonded secondary structure or have a limited amount of such structure, depending on the particular mRNA. This presumably reflects a general lack of complementary stretches of nucleotides in mRNAs. In eukaryotic mRNAs there is significant secondary structure at the 5′ end in the untranslated region. Phage RNAs (and perhaps other mRNAs) do have significant double-stranded regions.
2. This can be rationalized as arising from the fact that the primary role of the bases in mRNA is to specify the sequence of a protein, and the mRNA must be capable of being translated by the ribosome. The secondary structure regions that do occur limit this, and may allow the control of translation. This latter point illustrates RNA flexibility — the potential for a switch from one structure to an alternative. This can only occur if the alternative becomes more energetically favourable, e.g. by unwinding of the helical region at the 5′ end of eukaryotic mRNAs.′