G-quadruplexes and their unexpected ability to fold proteins

How do proteins, DNA, and RNA fold? This question has remained one of the most fascinating one in Biology and Chemistry. Research in folding addresses three major questions: Given a primary sequence, what is the final structure of a protein, a DNA, or an RNA in their folded state? Some say that alphafold has solved this issue, at least for protein folding. The other two questions remain open: The first involves the question of how cells balance the equilibrium between benign folding and aberrant folding, leading to misfolded, aggregated states. Misfolded protein states are known as causative agents inducing not only neurodegenerative diseases but also diabetes type II; they accumulate in an aging society. The second question is linked to the first and is concerned with the rates of folding. These rates are typically determined in sophisticated biophysical experiments. But these in vitro folding rates are often too slow and not consistent with the cellular needs for rapid turn-over of biomolecular folds. In many cellular machineries, proteins and RNAs have to fold for a short period of time, but then need to unfold or refold. In cells, these rates are often faster than those determined for the isolated systems. In fact, various protein chaperones have been identified that solve both questions: Protein chaperones are known to help proteins to avoid misfolding. The necessity to prevent misfolded species has also been reported for RNA, and consequently, RNA chaperones, e.g., RNA helicases have been found. In addition, protein chaperones are known to accelerate the unfolding of double-stranded RNA and other sophisticated RNA and DNA structures. Again, helper proteins help complicated RNAs and DNAs to (re)fold. Huang et al. now extend this repertoire of chaperones: They show that G- quadruplexes, a special form of DNA and RNA, can help proteins to fold properly. Their findings reverse the role of the chaperone helper and the system that needs help. DNA G4s help out proteins, a finding that adds a new cellular function to DNA and presumably also for RNA (1). The findings of Huang et al. open up a new understanding of the role of RNA and DNA quadruplex structures in cells. During translation, the coding messenger RNA can form RNA G4 prior to being read-off by the ribosome or after translation. Up to now, such RNA G4 formation was considered to function as roadblock, inhibiting protein translation and leading to premature translation abortion but not as chaperones to promote protein folding.

What is special about RNA and DNA quadruplexes? Within the diversity of biomolecular structures, RNA and DNA guanosine quadruplexes (G4) stand out as structures that form by pairing of four strands of guanosine- or deoxyguanosine-rich strands, respectively. Thus, G-quadruplexes (G4s) are noncanonical RNA/DNA structures that contain nucleobase tetrads stabilized by Hoogsteen hydrogen bonds. At least two tetrads stack onto each other, and monovalent cations including K⁺ or Na⁺ can bind between the tetrad in the center of the tetrad (2, 3). DNA G4s are found in genomic regions (4), including telomeres (5, 6), oncogene promoters (7–10), and replication origins. Their function includes transcriptional regulation (11, 12), genome stability (13), and replication (14). DNA G4s exhibit structural polymorphism with diverse topologies influenced by sequence (15–18) and cation (19) type due to their complex folding energy landscapes (20).

The findings of Huang et al. open up a new understanding of the role of RNA and DNA quadruplex structures in cells.

The work reported here builds on previous work (21, 22) that had already documented the general chaperoning role of G4. The work published in this edition of the Proceedings provides a kinetic mechanistic understanding of the phenomenon by characterizing the effect of DNA G4 with different topologies to the kinetics of folding of a model protein. Determining the changes in kinetics allows deciphering the impact of G4 on the protein folding landscape. The work stands out due to the clever design of different DNA G4 and of the protein model system. The authors show that pending on the topology of the four stranded DNA G4 that adopt either parallel, antiparallel, or 3 + 1 hybrid strand topologies, different folding rates are observed (Fig. 1A). Interestingly, the parallel strand orientation accelerates the most, followed by the 3 + 1 hybrid strand topology. It is worth noting that RNA G4 also adopt a parallel topology. As protein, the authors chose the model protein TagRFP675 (21, 23). TagRFP675 has been shown to populate at least one fluorescent intermediate; thus, TagRFP675 is known to fold as a three state-folder. While the unfolded state of TagRFP675 is not fluorescent, the excitation/emission wavelengths of the intermediate (388/510 nm) and the folded state (598/675 nm) both have a large Stoke shift but differ substantially in their excitation and emission wavelengths. For proteins, various folding mechanisms are perceivable: With a single intermediate, a sequential folding pathway can be populated, in the case of two intermediates; the transient states I₁ and I₂ can be on- or off-pathway, branched or triangular. Given the high signal-to-noise of the fluorescence data, the experimental data can be compared to predictions that allow differentiation between different protein folding pathways. In the absence of the DNA G4 chaperone, two intermediates I₁ and I₂ accumulate along the folding pathway of TagRFP675. The off-pathway intermediate I₂ interconverts with both, the unfolded state U and the on-pathway intermediate I₁, but not with the native state N (Fig. 1B). The mechanism for TagRFG675 is similar but not identical to folding pathways detected, e.g., for the paradigm folding model proteins lysozyme and α-lactalbumin (24, 25).

Fig. 1.

(A) Schematic representation of possible G4 topologies and examples of corresponding DNA/RNA structures taken from PDB. (B) Modulation of the kinetic protein folding pathways by DNA G4. In the absence of G4, the model protein adopts a branched folding pathway, in which to intermediate I₁ and I₂ interact with the unfolded state U, but not with the native state N. G4 interacts with the unfolded state and stabilizes the on-pathway intermediate I₁. This stabilization suppresses the formation of the off-pathway intermediate I₂.

The authors then investigated the impact of the addition of the parallel G4 Seq 576 on the folding pathway of the model protein TagRFP675. They show that the triangular pathway is reduced to a sequential pathway populating the unfolded protein (U), the on-pathway intermediate I_1, and the folded state N. The DNA G4 rapidly binds to the unfolded protein and stabilizes the transient G4-U and the G4-I₁ complexes. From measuring the temperature dependence of folding, including the temperature dependence of Δc_p, the authors can show that the acceleration of folding and unfolding by a factor of 5 is due to an increase in the enthalpic barrier that is small compared to the decrease in entropic barrier. For the acceleration of the transition from U to I₁, the kinetics require a change in heat capacity.

As pointed out by the authors, with the new function identified here, also RNA G4, accumulating under stress, could help fold slow-folding proteins that have not been previously identified as clients for Hsp60. More thorough in-cell kinetic investigations will need to be conducted to further fill this new chapter for G4 DNA/RNA function. As often, finding function that can be assigned to RNAs or DNAs as opposed to be reserved for proteins only fuels into the concept of the RNA world that was capable of carrying out previously unknown function. Low complexity, G-rich RNAs and DNAs could act as chaperones before more complex protein complexes have evolved. It remains to be seen how much of this ancient function is still operational in more highly evolved cells.