Conformational changes of proteins involved in the chaperone mediated protein folding cycle measured by spFRET

Heat shock proteins, like Hsp70, Hsp90 and Hsp60, act as chaperones. This class of proteins assist nascent unfolded proteins in reaching their final functional structure and conformation in the appropriate location within the cell. For the development of treatment for various disorders such as Alzheimer or Parkinson’s disease, it is essential to understand the underlying mechanisms of chaperone-assisted protein folding. Here, we investigated multiple chaperone systems according to their conformation, which can be one of the first steps in understanding the functional mechanisms. Therefore, we studied different chaperone systems related to their conformation upon the addition of interaction partners or with respect to the structural changes they introduce in a substrate protein. To monitor distance changes on a nanometer scale, the proteins were labeled with two dyes, the donor and acceptor dye. By exciting the donor, parts of the energy are transferred to the acceptor. This process is called Förster Resonance Energy Transfer (FRET) and is distance dependent. Furthermore, we chose single-pair FRET (spFRET) instead of ensemble due to the fact that multiple conformations of a protein can be monitored, which has been shown to be impossible with ensemble measurements. Two different types of experiments were performed. On the one side, solution based experiments using a Multiparameter Fluorescence Detection (MFD) setup combined with Pulsed Interleaved Excitation (PIE) were performed to detect snapshots of single proteins. On the other side, total internal reflection fluorescence (TIRF) microscopy was used to study the dynamics of a single protein over time. The Hsp70 in the endoplasmic reticulum (ER) is called BiP. It consists of two domains that are connected by a short linker, the nucleotide binding domain (NBD) and the substrate binding domain (SBD). The SBD contains a flexible, alpha helical lid that can open or close the substrate binding pocket of the SBD. When a client protein enters the ER, its charged regions are initially protected by BiP to avoid non-specific interactions. One of the regulating factors in this process is the BiP associated protein, BAP, a nucleotide exchange factor (NEF) of BiP. BAP controls the binding and release of the nascent proteins from BiP by accelerating the ATP/ADP exchange. To analyze the conformational changes BAP introduces in BiP, BiP was fluorescently labeled with a donor and acceptor dye. Three different mutants were used. The first one is the interdomain-mutant, which gives information about the distance between NBD and SBD. The second mutant is the lid-mutant, which monitors the lid opening and closing, and the last mutant is the combined-mutant, which has labels on the NBD and the lid to observe the distance between NBD and the C-terminal end of the lid. The results show that BAP can stably bind to BiP in the ADP and the nucleotide free conformation, but only transient interaction were detected in the presence of ATP. This transient interaction is mediated by the N-terminal domain of BAP. Together with the lid of BiP, this domain of BAP was identified to keep the NBD and SBD of BiP apart from each other. Furthermore, it was found that BAP speeds up the nucleotide cycle and, thus, act as a NEF. From our results, the idea arises that BAP mediates an open nucleotide binding pocket, which promotes the release of the nucleotide. In addition, BAP was identified to compete with a nature substrate and is kicked out of the complex when the substrate binds or the other way around. Thus, our results clarify the interaction of BAP with BiP and show that BAP can mediate the nucleotide cycle of BiP. In another set of experiments, we analyzed how a substrate, which has been caught by Hsp70, is handed over to Hsp90, the next interaction partner in the folding machinery. The cochaperone, which is known to act as a kind of scaffold in this process, is called Sti1. Sti1 was labeled with a FRET pair to study the conformational changes and dynamics introduced by binding of Hsp70 and Hsp90. Sti1 consists of two modules. The first one has an aspartate and proline rich (DP) domain and a tetratricopeptide repeats (TPR) domain. The TPR domain can interact with Hsp70. This module is connected by a flexible linker with the second module, which consists of a TPR2A, TPR2B and a DP2 domain. Hsp90 can interact with the TPR2A domain and Hsp70 has a second interaction domain, the TPR2B. Hsp70 was found to bind preferably to the TPR1 domain, when Hsp90 is bound to Sti1. In addition, it was detected by spFRET that Sti1 is already dynamic on its own but, upon binding of Hsp70, Hsp90 or both, the number of dynamic molecules increases. These dynamics are important for bringing the two modules closer together and, thus, mediate the transfer of the substrate from Hsp70 to Hsp90. Afterwards, Hsp70 is transferred back to the TPR1 binding domain and Sti1 opens up. Due to the low affinity of Hsp70 to the TPR1 binding domain, Hsp70 is released and the cycle can start over again. Thus, Sti1 has the important role of binding to Hsp70 and Hsp90 and dynamically connects the two chaperones as a scaffold and assists in substrate transfer. Another important chaperone is the Hsp60. In bacteria, it is called GroEL and consists of two identical rings, which stick together. Each of them form a cavity. Inside the cavity, substrate proteins can be trapped. Binding of ATP to GroEL mediates the binding of GroES, a co-chaperone for GroEL and the lid for the cavity. Inside the cage, the substrate protein can fold to its final conformation. To study the effect of GroEL on a substrate, the maltose binding protein (MBP) was chosen. From a previous investigation it is known that a double mutant of MBP (DM-MBP) is not a native substrate for GroEL. It can fold by its own. However, the interaction of GroEL and GroES with DM-MBP makes the folding kinetics faster. To monitor the changes on the timescale of folding, the N-terminal domain of DM-MBP was labeled with ATTO 532 and Alexa Fluor 647 and a home-built microfluidic add-on for the MFD-PIE setup was implemented. Microfluidic provides the possibility to mix different buffers or interaction partners directly on top of the microscope and, thus, makes it possible to measure the conformational changes within a time window of 33 ms to 1.07 s. DM-MBP was found to fold spontaneously on this timescale to an intermediate state, but not to the final conformation. The timing of this folding process was found to depend only on the chaotropic agent concentration, in our case GuHCl, but not on the concentration of other salts. By adding GroEL, it was found that DM-MBP gets partly unfolded on a timescale of 343 ms. The other fraction of molecules shows only slight conformational changes, which are related to a more compact conformation and occurs on a timescale between 343 ms and 512 ms. Furthermore, the addition of GroEL and GroES speeds up the folding process. From the unfolded to the intermediate state, the conformational changes gets 5 to 6 times faster. Afterwards, the DM-MBP gets more and more compact and after 1.07 s a fraction of molecules reach their final folded conformation. Taken together, the timescales of DM-MBP folding that was monitored with microfluidics depends on the GuHCl concentration. The addition of GroEL alone unfolds DM-MBP within the first 512 ms of interaction but GroEL together with GroES speeds up the folding kinetics. In summary, it was possible to study different chaperone systems with respect to their conformational changes by using spFRET. Our spFRET results reveal new insights that could not be detected before by ensemble methods. Furthermore, MFD-PIE and TIRF microscopy offers the opportunity to detect conformations under equilibrium conditions. Microfluidic widen up the range of questions that can be answered by these methods. It makes it possible to mix different components of the sample direct ontop of the microscope and, thus, enables to detect fast kinetics and conformational changes on short time scales