Chapter 6 – Conclusions and

Future Work

It was the intention of this work to investigate the feasibility of spin-labelling proteins and peptides for ESR experiments that study folding events on the real timescale. Furthermore experience was gained in preparing and synthesising the chemicals and exploiting the CD and ESR techniques to investigate the labelled proteins and peptides.

Haemoglobin was successfully labelled, though with a low efficiency (measured using spin calibration standards and doubly integrating the ESR spectrum), and various experiments were performed on the protein. These experiments included adding a radical quenching agent to investigate whether there were any free spin label molecules in the system, increasing the viscosity of the solution using sucrose, and denaturing the protein using urea. The commercial haemoglobin was found to be methaemoglobin (denatured) using electronic absorption spectroscopy. This finding led to the investigation of oxy-haemoglobin separated from freshly drawn human blood, which has a known structure. The ESR spectra for haemoglobin have not been fully analysed using computer simulations, but the spectra were typical of a spin label in a restricted rotational motion environment, e.g. a pocket within the protein structure.

A 14 amino acid peptide sequence from the human alpha-lactalbumin protein was chosen as a single cysteine (at position 6) peptide for labelling experiments. The peptide was found to have a CD spectrum typical of an alpha-helical structure in an equal mixture of water and tfe and to have the random coil structure in pure water. The custom-made peptide was successfully labelled with all three labels used in this work. The peptide was mutated at the 13^th residue to form a double cysteine sequence that was then spin labelled. For future work, a review of the labelling process may be necessary so that HPLC, and not dialysis, is used to terminate the labelling reaction and remove unreacted spin labels. ESR experiments using the doubly-labelled peptide and the corresponding singly-labelled forms were performed in mixtures of water and tfe to represent different stages of folding for the peptide (the graduation from the random coil to the helical conformation was observed using CD).

The ESR experiments were analysed using a computer simulation program. The width of the hyperfine lines was found to be narrower in water than in the tfe solutions; this behaviour was simulated successfully by changing the rotational correlation time constant of the spin. The label and peptide were found to rotate faster in water than in tfe; the conclusion was that this could be due to the label being more restricted in the helical conformation than when the peptide is a random coil. The doubly-labelled peptide seemed to have an additional broadening, particularly in the folded form; this was thought to be due to a dipolar coupling between the two spins which was not fully averaged to zero in the constrained system. However, there are no computer simulation programs commercially available for studying the dipolar interaction for CW X-band ESR spectra at room temperature; the design of such a program is envisaged for the future.

The spectra were also seen to shift to higher fields as the concentration of tfe increased. This shift was thought to be a consequence of the change in solvent conditions as tfe has a lower dielectric constant than water. The polarity of the system affects the g- and A-tensors of the nitroxide spin. Simulations showed that the expected relative alterations of the g- and A-values that would occur with the addition of tfe, shift the spectra to higher fields.

The simulations performed in order to analyse the labelled peptide could not model the system exactly, since actual values for the g- and A-values in the system, information about the orientation of the label with respect to the peptide, the angle between the rotation axes and the magnetic axes of the spin, and the exact nature of the inhomogeneous line broadening were not known. Low temperature experiments were performed in this project, but the temperature used was too high to give reliable information on the g- and A-tensors. Molecular modelling and two-dimensional NMR will be used in the future to give the likely orientation of the label to the peptide. Future ESR measurements will be taken at high fields (W-band ESR) or on deuterated spin labels in order to decrease the inhomogeneous line broadening effect. Possible line broadening by the spin-spin exchange interaction will be investigated through alteration of the concentration of the sample since Heisenberg exchange broadening is proportional to the spin concentration.

The next step in this project will be a major research proposal for a high field W-band ESR spectrometer. Collaboration with Professor T. Prisner of the University of Frankfurt has already begun and will lead to the development of a stopped-flow cavity for the ESR investigations of real time folding events of spin-labelled peptides and proteins.

The overall conclusion is that this work has provided this laboratory with the knowledge to progress with studying the real time folding events of spin-labelled proteins and peptides.