Dynamics of unfolded and [alpha]-helical polypeptide chains

The aim of this thesis was the investigation of the dynamics of elementary steps in
protein folding. During folding, the polypeptide chain explores the free energy surface
and first interactions are established. These interactions lead to formation of
secondary structure elements which then can assemble to form protein structure.
Contact formation between different residues in the polypeptide chain limits the rate
with which a protein can explore its conformational space and sets an upper limit for
the speed of folding.
We studied loop closure reactions in polypeptides and dynamics of α-helices applying
the method of triplet-triplet energy transfer (TTET). Triplet excitation is transferred
by a two electron exchange mechanism from a xanthone (Xan) donor moiety
to a naphthalene (Nal) acceptor upon van der Waals contact. The transfer reaction is
diffusion controlled allowing direct determination of rate constants for loop formation
by observing the decay of xanthone triplet absorption or the concomitant increase
in naphthalene triplets.
By introducing the triplet labels into unfolded model peptides we wanted to test the
effect of polypeptide chain length, amino acid sequence and solvent conditions.
In a short host-guest loop Xan-Ser-Xaa-Ser-Nal-Ser-Gly-OH the local effect of different
amino acids on loop formation was probed by introducing the guest amino
acids Xaa = Ser, Gly, Pro, Ala, Ile, Glu, Arg and His. We observed similar kinetics
for all amino acids except glycine and proline, although amino acids with a Cβ-atom
showed slightly slower loop closure rate constants. In glycine containing peptides the
contact formation rate constant was found to be faster because of the increased flexibility
of the glycine residue. The introduction of proline however led to double exponential
kinetics. A slow phase corresponding to trans-Pro showed the slowest kinetics
of all peptides due to the rigid structure of the proline residue, whereas cis-Pro
showed the fastest kinetics due to the introduction of a kink and thus smaller end-toend
distances.
To probe chain dynamics in a natural loop derived from a protein we introduced the
TTET labels into an unstructured 18-residue loop from carp muscle β-parvalbumin.
This allowed us to compare loop formation rate constants obtained in a natural sequence
to values obtained in model peptides. The kinetics in the parvalbumin loop
corresponded well to values obtained for poly-Ser chains. This showed that a small
amount of glycine in the sequence can compensate for the slowing down of chain
dynamics induced by large amino acids.
During protein folding, interactions are mostly established between amino acids in
the interior of the chain. Thus, the effect of additional tails on the loop closure rate
constants has to be taken into account. Three types of loops can be distinguished.
Type I loops denote end-to-end loops, type II-loops are end-to-interior loops, while
type III-loops denote interior-to-interior loops.
We measured loop formation rate constants in type II and type III-loops depending
on the size of the additional tails. It was observed that the loop formation rate constant
is decreased with increasing size of the tail. For different type II-loops this effect
was found to reach a limit when the tail dimensions are about three times larger
than the loop size. In this limit, loop closure rates are decreased by a factor of 2.3.
For type III-loops the effect of additional tails was found to be stronger than in the
type II case. However, the limiting value could not be determined. Assuming that the
limit for type III-loop formation is reached at the same tail size as in the type II case
this would result in a decrease in loop closure rates by a factor of four.
All observed rate constants in TTET experiments were found to be on the nanosecond
time scale. However, faster reaction could not be ruled out. To study reactions
on a shorter time scale using TTET it is necessary to understand the photophysics of
xanthone triplet formation and energy transfer to naphthalene in detail. Thus, femtosecond
timescale experiments were performed to determine the timescale of xanthone
intersystem crossing and triplet-triplet energy transfer. It was found that xanthone
triplet formation proceeds on the 2 ps timescale and TTET to naphthalene occurs
below 2 ps. This allows to observe contact formation processes with time constants
of 5-10 ps.
As analysis of initial amplitudes in TTET experiments suggested fast reactions that
cannot be observed in nanosecond time resolution experiments, we studied contact
formation in small peptides applying femtosecond laserflash spectroscopy. Two fast
processes were detected. A faster decay with a time constant of 3-4 ps and 15% amplitude
was followed by a slower process on the 100’s of picosecond time scale
which accounted for 30-40% of the xanthone triplet absorption decay. These fast
reactions result from motions of a subpopulation of peptides within a conformational
substate on the free energy surface that allows contact without major barrier crossing.
The remaining population of molecules has to sample the free energy surface which
leads to exponential kinetics in the nanosecond time range at room temperature.
To gain more insight into the properties of the free energy surface we tested conditions
where barrier crossing between the local minima on the free energy surface is
slow. At low temperature or high viscosity the kinetics for loop formation were
found to deviate from exponential behaviour and could be described by a stretched
exponential decay. These results show that the concept of conformational substates
which was initially developed to describe protein motions is valid for unstructured
polypeptides.
During protein folding, initial contacts lead to formation of local structure. α-helices
represent the most abundant and most local secondary structure element. We studied
global and local stability and dynamics in α-helices applying TTET to alanine based
helical peptides. We introduced TTET labels at different positions and could thus
obtain information about local and global helix unfolding and refolding kinetics. We
observed that α-helices exhibit higher stability and slower kinetics at central positions,
whereas the termini were frayed and showed fluctuations on a faster timescale.
These results show that α-helix formation is a complex process and its kinetics are
position dependent.
To observe elementary reactions of α-helix formation as nucleation steps or propagation
reactions, kinetic experiments have to start with an unfolded ensemble. We developed
a synthetic access to a peptide system where an α-helical peptide is cyclised
by a photocleavable crosslinker moiety based on a p-hydroxyphenacyl moiety which
prevents helix formation. This system can now be used to monitor helix formation
upon a fast release of the peptide by nanosecond laser irradiation.