Among the findings of our 2010 Nucleic Acids Research (NAR) article titled The RNA backbone plays a crucial role in mediating the intrinsic stability of the GpU dinucleotide platform and the GpUpA/GpA miniduplex, the key is identifying the O2′(G)…O2P(U) H-bond (see figure below). As noted in a previous post What’s special about the GpU dinucleotide platform?, it was an accidental observation while I was preparing a figure for our 2008 3DNA Nature Protocols paper. Trained as a chemist, after scrutinizing the many occurrances of the GpU platforms in the large ribosomal subunit of Haloarcula marismortui (PDB entry 1jj2), I had no doubt that it is an H-bond. Yet, behind the scene, things were never that straightforward: if it is indeed an H-hbond as we’ve claimed, how could it have been missed altogether by the RNA structural biology community?

Anticipating the potential questions that could be raised by the reviewers, we were extremely careful in characterizing the O2′(G)…O2P(U) H-bond:

• It is formed between the hydroxyl group (donor) of G and a non-bridging phosphate oxygen atom (O2P, acceptor) of U.
• The distance between O2′(G) and O2P(U), 2.68 ± 0.14 Å, is perfect for an H-bond.
• I queried the Cambridge Structure Database for hydroxyl-phosphate H-bonds with similar relative geometry and chemical identity. We found a case in the phospholipid lysophosphatidyl-ethanolamine, where this type of H-bond is highlighted in the abstract: The free glycerol hydroxyl group forms an intramolecular hydrogen bond with a phosphate oxygen and thus affects the conformation and orientation of the head group.
• I also performed a survey of potential O2′(i)…O2P(i+1) H-bonds within dinucleotides regardless of platform configuration, and detected 1186 such pairwise interactions within a distance cutoff of 3.3 Å in RNA crystal structures of 2.5 Å or better resolution.

Careful as we were, we still failed to convince reviewer #3 of our manuscript, which was originally submitted to the RNA journal and finally rejected following the second round of review. Here is an excerpt related to the O2′(G)…O2P(U) H-bond from reviewer #3’s comment:

The first main concern is that the “new” H-bond interaction that the authors propose as an explanation for the greater occurrence of GU platforms versus di-nucleotide combinations does not make much sense on a fundamental chemical and stereo-chemical point of view. Unless the whole community of chemists and biochemists agree to redefine what an H-bond is, the fact that the 2’OH (i) atom is at 2.68 Å from the O2P atom cannot be the only criteria for an H-bond. In fact, if the authors are the first to mention this H-bond, it is because none of the scientists working in RNA structural biology would have considered this to be an H-bond interaction at the first place! H-bonds are known to be very directional. The O2’-H bond should be aligned with one of the electron doublets of O2P to be able to form a proper H-bond. Acceptable variation could be 20° to 30° degree with respect of a straight H-bond interaction, not 90°! The unique paper that the authors cite for justifying their claim cannot be used as a reference. If the authors want to justify that the close proximity of the 2’OH(i) and O2P is the important factor that contributes to preference of GU platforms versus other platforms, they should undergo quantum mechanics calculations to demonstrate it.

This review is so critical that I saw no point in arguing with it — I certainly have neither the power to “redefine what an H-bond is” nor the expertise to perform quantum mechanics (QM) calculations to validate the O2′(G)…O2P(U) H-bond or otherwise. What is compelling to me about the GpU story from the very beginning is that once this sugar-phosphate H-bond is acknowledged, every other parts of our NAR paper follow naturally and logically. Leaving the chicken or the egg issue alone, our work provides a novel perspective about GpU platform’s predominance, the formation of the bulged-G or loop-E motif, the evolutionary co-occurrence of GpUpA and GpA in the GpUpA/GpA miniduplex, and the extreme conservation of GpU observed at most 5′-splice sites. Put another way, we connect the dots to form a coherent picture that is easily understandable to biologists and chemists.

Luckily, after being re-submitted to NAR, the paper was quickly accepted for publication and even selected as a featured article! As another nice surprise, shortly after it was available online as an Advance Access paper, I received an email from Jiri Sponer. Thereafter, we collaborated on a follow-up paper titled Understanding the Sequence Preference of Recurrent RNA Building Blocks Using Quantum Chemistry: The Intrastrand RNA Dinucleotide Platform. While not unexpected, the results of the state-of-the-art QM calculations were nevertheless reassuring:

The mixed-pucker sugar–phosphate backbone conformation found in most GpU platforms, in which the 5′-ribose sugar (G) is in the C2′-endo form and the 3′-sugar (U) in the C3′-endo form, is intrinsically more stable than the standard A-RNA backbone arrangement, partially as a result of a favorable O2′···O2P intraplatform interaction. Our results thus validate the hypothesis of Lu et al. (Lu, X.-J.; et al. Nucleic Acids Res. 2010, 38, 4868–4876) that the superior stability of GpU platforms is partially mediated by the strong O2′···O2P hydrogen bond. …… In contrast, we show that the dinucleotide platform is not properly described in the course of atomistic explicit-solvent simulations. Our work also gives methodological insights into QM calculations of experimental RNA backbone geometries. Such calculations are inherently complicated by rather large data and refinement uncertainties in the available RNA experimental structures, which often preclude reliable energy computations.

So, the O2′(G)…O2P(U) H-bond is more than likely to be real; at least some other scientists working in RNA structural biology do share our view.

See also: What’s special about the GpU dinucleotide platform?