Protein folding is a unique biological problem that single molecule techniques have had success in tackling to gain detailed insight into the folding landscape of purified proteins in solution. The in vitro post-translational refolding pathway, however, may not accurately recreate the pathway of folding when a protein is being synthesized by the ribosome. Folding on the surface of the ribosome is difficult to probe since the ribosome itself is composed of both RNA and proteins, so is sensitive to many of the same techniques used to probe the nascent chain. Our lab previously developed a technique to use optical tweezers to watch the nascent chain fold at the single molecule level with minimal perturbation to the ribosome, which showed that the ribosome slows down folding kinetics. Continuing this experimental setup, in this work we explore the folding of a multi-domain calcium-binding protein, calerythrin, on the ribosomal surface both in static stalled complexes and in real-time elongation conditions.
Our results show that off the ribosome, the refolding pathway proceeds via the C-domain, while N-domain folding is the rate-limiting step to the folded state. In contrast, an alternate intermediate that is a misfolded, unproductive state is accessible in a stalled ribosome complex. When we further investigated this misfolded state, we found that it is accessible to partially synthesized, truncated proteins off the ribosome as well. The ribosome decreases the folding rate, as previously reported for native folds, but it also acts to increase the unfolding rate, a combined effect that promotes an escape from misfolding.
Although this effect is interesting, the kinetics of folding, even when delayed, are still quite fast relative to synthesis. These different timescales have been interpreted in the literature as implying that folding during real-time elongation occurs at equilibrium. However, to directly test if this assumption is in fact relevant, we conducted measurements of folding while real-time elongation was ongoing. This gave us the unexpected result that folding on the ribosome during active protein synthesis is not at equilibrium despite the disparate timescales involved. In fact, there is a non-equilibrium relaxation time of 71 seconds prior to the onset of misfolding. In this time, the protein remains unfolded and is incapable of fully exploring its energy landscape. After equilibration occurs, the folding observed matches stalled complexes and is completely reversible and at equilibrium despite further chain elongation or changes to the force. This long delay before folding provides a mechanism for reducing the probability of misfolding significantly by allowing synthesis to terminate (likely for small proteins) or by providing additional time for chaperone recruitment (likely for larger proteins). This study gives us further insight into the importance of the ribosome not just for synthesis but also for regulating protein function, and opens up interesting questions about further interplay between elongation and protein folding.
We also developed a modified experimental design that provides insight into folding through two parallel channels– optical tweezers and fluorescence. In this assay, fluorophores within the protein can report on local conformational changes independent of the signal from the optical tweezers via FRET, allowing us to correlate global and local changes without using indirect measures of folding intermediates such as mutations. We have conducted proof-of-concept experiments for this design using calmodulin, a eukaryotic signaling protein, that aim to understand the order of events in early folding and to identify any correlation in secondary structure formation between the two domains. Our preliminary results show that the design is feasible and suggest promising avenues for future development to definitively answer the questions posed. In the future, this assay could also be combined with our experiments on ribosome-bound nascent chains to look at dynamics of the unfolded state.
