The GlmS Riboswitch: A very unique addition to the catalytic RNA family!

The GlmS Riboswitch: A very unique addition to the catalytic RNA family!

Figure 1. Riboswitch folding! (People in White Coats)

We have all heard at some point that our genetic information is encoded in our DNA. Perhaps you have also heard that RNA is transcribed from DNA, and that proteins are synthesized from the information carried by RNA. This is known as the Central Dogma of Molecular Biology. From the Central Dogma, one would think that RNA only serves as a messenger, relaying instructions from DNA to protein, but that is not the case! RNA molecules are capable of participating in a wide array of reactions! They can even perform catalysis—previously thought to be exclusively performed by proteins. These catalytic RNA molecules have been dubbed “ribozymes.” Another interesting set of RNA molecules are the riboswitches. Riboswitches regulate gene expression. They are found in the sequence immediately prior to the gene that they regulated, and, when stimulated, they fold up into shapes that either block transcription or translation (the information transfer processes from DNA to RNA and from RNA to protein, respectively). Of these riboswitches, the GlmS riboswitch (Figure 2) is especially cool because it is also a ribozyme!

Figure 2. Tertiary structure of the GlmS ribozyme from Bacillus anthracis in ribbons format. The regions of the riboswitch are color coded, and the small gray ball and stick structure represents GlcN6P. The surrounding blue helix (P2.2) is the active site.

Figure 3. Schematic of the reduction in glmS gene expression in the presence of GlcN6P.

Like other riboswitches, the GlmS riboswitch is encoded in the segment of DNA slightly ahead of the gene that it regulates. This gene encodes a protein called glucosamine-6-phosphate (GlcN6P) synthetase. This is an enzyme involved in biosynthesis of the cell wall that, as the name suggests, synthesizes GlcN6P. GlcN6P then goes on to be a building block of the cell wall in Gram-positive bacteria! The GlmS riboswitch is normally inactive, meaning that synthetase production continues uninterrupted. When the GlmS riboswitch binds the downstream gene product GlcN6P, though, the riboswitch reduces synthetase production (Figure 3). Riboswitches often regulate transcription by folding to trigger RNA polymerase (RNAP) release to stop transcription, or they regulate translation by folding to block the ribosome binding site (RBS) to stop translation. Most riboswitches undergo a conformational change (i.e., they fold differently) when they bind their ligand (a small molecule target, like GlcN6P) and this conformational change is what permits them to regulate transcription or translation (Figure 4). Many thought that the GlmS riboswitch did the same, but as it turns out, GlmS is not like most riboswitches! The GlmS riboswitch represents a separate class in which gene expression is brought to a halt by GlmS self-cleavage in the presence of GlcN6P.

Figure 4. Schematic for gene regulation at different levels via typical riboswitches. A) The ligand-free riboswitch allows transcription to proceed as normal, but the ligand binding causes a structural change resulting in a terminator hairpin. This blocks transcription. B) When no ligand is present, the ribosome binds the ribosome binding site (RBS) and begins translation. When the ligand binds the riboswitch and causes conformational change, the RBS becomes unavailable and translation is inhibited. 

Ken Hampel and Melissa Tinsley of the University of Vermont sought to understand the extent of conformational change undergone by the GlmS riboswitch upon binding GlcN6P. They purified the RNA and studied it in a test tube! First, they conducted a “hydroxyl radical footprinting assay.” In this process, the GlmS RNA is radioactively labeled so it can be visualized later on. It is then allowed to fold in the presence of magnesium ions (Mg²⁺). Mg²⁺ ions are in high abundance within the cell and stabilize RNA through and after folding. Next, the folded riboswitch is treated with hydroxyl radicals. These cleave particular bonds in the sugar backbone of RNA, but only when the sugar is positioned outside of the molecule. Because the riboswitch is a folded structure, the building blocks that make up its core are protected from hydroxyl radicals while the exterior portions are exposed. Consequently, the cleavage patterns on the RNA can be used to determine the structure of the nascent folded riboswitch.

To determine whether the ligand-bound GlmS riboswitch is conformationally different from the ligand-free riboswitch, Hampel and Tinsley repeated the footprinting assay but first added GlcN6P to the GlmS RNA. If ligand binding causes a conformational change, then different portions of the riboswitch structure would be exposed to the hydroxyl radicals and a different cleavage pattern would result. Much to their surprise, Hampel and Tinsley found that the cleavage patterns did not differ between the ligand-bound and ligand-free GlmS riboswitches! This suggests that no structural change occurs when the riboswitch binds GlcN6P. This is unlike any of the other riboswitches scientists had studied! Before they made any conclusions, however, further experimentation was required to ensure that this wasn’t a fluke!

The next experiment Hampel and Tinsley conducted was photo-cross-linking, which relies on ultraviolet (UV) light to lock the building-blocks of RNA together where they make contact. It is unhealthy for your skin when UV-light does this to DNA, which is why it is important to protect your skin with sunscreen! Photo-cross-linking allows researchers to identify points of direct interaction and compare them for the GlmS riboswitch with and without GlcN6P. To execute photo-cross-linking, they incubated the GlmS riboswitches with Mg²⁺ either with or without GlcN6P and then exposed the samples to the UV-light. The locations of the cross-linkages were examined. Once again, they were the same regardless of whether GlcN6P was present! The photo-cross-linking data further suggested that the GlmS riboswitch folds independently of its ligand, GlcN6P. 

If the riboswitch binding site pre-folds and doesn’t change its structure after GlcN6P binds, how does GlcN6P activate the GlmS riboswitch? GlcN6P has been found to function as a coenzyme. This means that GlcN6P completes the active site in the GlmS riboswitch to enable regulation. The GlmS riboswitch regulates transcription by undergoing self-cleavage, which severs a bond in the sugar backbone of the RNA. In the GlmS riboswitch active site, there are hydrogen bonds between different functional groups such that the piece that executes self-cleavage is all tied up. GlcN6P hydrogen bonds with the same part of the GlmS active site as the self-cleaving group, so its presence frees up the self-cleaving group and stimulates cleavage! Because GlcN6P promotes GlmS self-cleavage and GlmS self-cleavage stops gene expression, increased GlcN6P leads to decreased gene expression.

The ability to essentially pre-form a binding pocket to avoid extensive conformational change upon ligand binding makes the GlmS riboswitch wholly unique among its fellow riboswitches. At least, unique from the ones we’ve discovered so far. The GlmS riboswitch is a common feature of many Gram-positive bacteria and it regulates the production of cell wall building blocks, which together make it an interesting target for antibiotics. As we have yet to unlock all of its secrets, we leave it to future researchers to explore the possibilities for this singular riboswitch and ribozyme!

 References

1.              Bingaman, J. L., Gonzalez, I. Y., Wang, B., and P. C. Bevilacqua. (2017). Activation of the glmS ribozyme nucleophile via overdetermined hydrogen bonding. Biochemistry, 56: 4313-4317. doi:10.1021/acs.biochem.7b00662

2.              Cochrane, J. C., Lipchock, S. V., & Strobel, S. A. (2007) Structural Investigation of the GlmS Ribozyme Bound to Its Catalytic Cofactor. Chemistry & Biology (14): 97-105. doi:10.1016/j.chembiol.2006.12.005

3.              Hampel, K. J., & Tinsley, M. M. (2006). Evidence for preorganization of the glmS ribozyme ligand binding pocket. Biochemistry, 45(25): 7861-7871.