Re: Why does RNA have a tertiary structure while this is rarely seen in DNA?

Dear K. R.,

Your question concerning differences in the tertiary structure of RNA vs. DNA has a relatively long answer. Hopefully you will appreciate that this is a much more complicated issue than can be properly addressed in this forum. In fact, it is still very much an area of active research, and is really more appropriately dealt with as an advanced college or graduate level course. Fortunately, I’ll be able to cheat a bit and refer you to a previously posed question on the Mad Scientists’ network, “Why are RNA duplexes more stable than DNA duplexes in vitro?”. Please read my response before continuing with the more complicated question of tertiary structure.

The basic differences between A-form and B-form duplex nucleic acid are dominated by the differences in base-stacking energetics, owing to slightly different overlap between nearest neighbor H-bonded base pairs. In B-form (DNA), the base pairs stack over one-another better with a relatively shallow tip-angle. By comparison, the base-pair stacking in A- form (RNA) overlaps less, placing the center of one pyridine or pyrimidines over the edge of the neighbor; in terms of van der Waals stacking interactions, this confers a higher stability to duplex A-form RNA over duplex B-form DNA. (I want to emphasize that the simple difference in backbone phosphodiester sugar conformations defines this topology; the energetic differences between conformations is a net sum of all contributing factors). Offset stacking is preferred owing to the slight differences in partial positive and negative charges due to pi- orbital ring currents, and the polar nature of pyridine and pyrimidine heterocycles.

Experimentally, we observe that short segments of RNA are more stable than the same duplex sequence of DNA, the persistence length of RNA is much longer than that of DNA (i.e. the RNA has higher stiffness than the same sequence of DNA), and that a hybrid duplex of DNA on one strand and RNA on the other results in an A-form structure. All of this really means that duplex DNA requires long uninterrupted regions of duplex in order to fold, while RNA duplexes can tolerate a mix of single-stranded and double- stranded regions. What we observe in nature are long segments of perfectly matched duplex DNA made up of two DNA strands; RNA’s are exclusively monomeric, and are observed to be a very heterogeneous mixture of double-stranded and single-stranded regions, containing hairpin loops, nucleotide bulges, three-and four-way junctions, and less commonly RNA triplexes and pseudo-knots. Additionally, any region of single-stranded nucleic acid that does not naturally have a base-pairing function is protected by single-stranded nucleic acid binding proteins. There are many hnRNP (heteronuclear Ribo-Nuclear Protiens) who function to coat mRNA, as well as many other specific proteins that coat RNA’s involved in splicing, in the ribosome, etc. When DNA is being replicated, it is protected from damage by SSB proteins (Single-Stranded Binding proteins).

There is a conspicuous lack of tertiary structure for DNA, while RNA molecules practically were born to knot up into complicated tertiary folds. I’ll assume you know enough about proteins to know that they also fold up into very well-defined native, folded conformations containing a mix of secondary and tertiary structures. For proteins, the secondary structures are only transiently stable, requiring the presence of tertiary interactions to adopt a final folded structure; for proteins, folding is highly cooperative, an “all or nothing” process. In the case of RNA folding, forming stable secondary structure (the basic double-helical duplex) is fast and spontaneous under physiological conditions. Formation of tertiary structure is comparatively slow, forming on the order of minutes as compared to the millisecond folding of duplex RNA. Tertiary structures themselves occur exclusively through non-duplex regions; for example, the non-base-paired hairpin loop regions may form hydrogen- bonding interactions with other non-base-paired regions. Base triplexes are commonly observed, which require a deformation of the normal A-form structure in the region of the triplex. Complicated pseudoknots are rightly considered to be tertiary structures, though these are primarily a complicated duplex structure. A large number of tertiary structures are observed to be stabilized by the presence of proteins bound to the RNA (notably, in group II introns and in the ribosome). I can give you a flavor of all of these, but you’ll better understand them by following up specific examples from research on ribozymes (catalytic intron RNAs) and on ribosome structures.

The wording of your question is very interesting to me...”why” are they different takes on some connotations of evolutionary considerations. This is obviously a much more difficult question to address. I’m going to cheat again and refer you to another question previously posed to this forum, “How is DNA better suited than RNA to carry genetic info?”. Please read my response before continuing.

To summarize, we believe that RNA evolved as the first prebiotic molecule. It is the only type of macromolecule that is a heteropolymer that can form spontaneously (given what is believed to be the prebiotic mix of building blocks), that has been demonstrated to perform a number of different catalytic reactions, including limited self-replication. The ability of RNA to form tertiary structures is at the heart of these complicated enzymatic activity. It is thought that RNA preceded DNA, and thus conferred an evolutionary function of “information storage” onto DNA. It’s job is to sit idly until an RNA, and ultimately both RNA and proteins, can perform either transcription or replication functions on the DNA.

Thanks for your interesting question.

Regards,

Dr. James Kranz
Johnson & Johnson Pharmaceutical Research & Development