Temperature-dependence of protonation dynamics

In this subsection, the protonation dynamics depending on temperature shall be explored.

BR in native purple membrane was used first to evaluate how Asp85 protonation and Schiff base deprotonation dynamics depend on temperature. According to Arrhenius -law, the "kinetic effect" of faster dynamics for higher temperatures could be demonstrated.

BR reconstituted into DMPC liposomes was then investigated in a temperature range below and above the phase transition temperature of DMPC. In this case, the "kinetic effect" and the underlying "lipid phase transition effect" were studied.

5.1. Bacteriorhodopsin

BR in native purple membrane

Figure 5.13 illustrates the protonation dynamics of the primary proton transfer step for BR in native purple membrane for 5^◦C, 20^◦C, 30 ^◦C and 50 ^◦C.

Figure 5.13.: Protonation dynamics of the primary proton transfer step of BR in purple membrane monitored at 1760 cm^-1(left) and 1188 cm^-1 (right) for 5^◦C, 20^◦C, 30 ^◦C and 50 ^◦C (from top to bottom).

5. Results and Discussion

At 5 ^◦C protonation dynamics of the primary proton transfer step occur very slowly.

The minimum at 1188 cm^-1occurs only around 3 ms. Furthermore, M state is not decaying until the end of the acquired time range of 19 ms monitored at 1760 cm^-1. With increasing temperature, the minimum at 1188 cm^-1, maximum at 1760 cm^-1 respectively shift to earlier times. At 20^◦C, Schiff base deprotonation (decay), Schiff base reprotonation (rise) and final thermal relaxation of the retinal (decay) becomes visible within the acquired time range at 1188 cm^-1 as well as rise and decay dynamics of the M intermediate at 1760 cm^-1. At 30 ^◦C, M state decay and thermal relaxation of the retinal are fully accomplished within the recorded time range and for 50^◦C dynamics are further enhanced with a minimum at 1188 cm^-1, maximum at 1760 cm^-1 located around 100 µs. The

"kinetic effect" of faster protonation dynamics with rising temperature could be clearly demonstrated.

Figure 5.14 shows the corresponding Arrhenius-plot. Rate constants for Schiff base deprotonation and Asp85 protonation correlate well except for the higher difference of the rate constants observed at 50 ^◦C. An activation energy of 46.4 kJmol^-1 for Schiff base deprotonation could be calculated which fits well to reported activation energies for the primary proton transfer step based on theoretical calculations [101]. For Asp85 protonation, a lower activation energy of 31 kJmol^-1 was calculated, mainly due to the lower rate constant determined at 50 ^◦C.

0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 8.0

Figure 5.14.: Arrhenius-plot of the temperature-dependent Schiff base deprotonation and Asp85 protonation dynamics of BR in purple membrane.

5.1. Bacteriorhodopsin

BR reconstituted into DMPC liposomes

Apart from this "kinetic effect" it was expected that a "lipid phase transition effect"

should occur additionally for the protonation dynamics of BR in a temperature range below and above the phase transition temperature of the lipids. The lipid DMPC was chosen for this investigation due to its convenient phase transition temperature of 24^◦C.

Below the phase transition temperature, DMPC lipids are in a gel phase, while above the phase transition temperature the lipids are in a liquid phase. Reconstitution of membrane proteins can affect the phase transition properties of the lipids. The gauche-to-trans ratio of the lipids changing along the phase transition can be monitored by a frequency shift of the asymmetric CH₂ stretching vibration. For increasing gauche content of the lipid alkyl chains indicating the more fluid phase, the frequency shifts to higher wavenumbers. A peak position below 2920 cm^-1 is indicative of lipids in the gel phase, a peak position above 2923 cm^-1 is indicative of lipids in the liquid phase. Figure 5.15 shows the determined phase transition temperatures for a pure DMPC lipid film and for a BR reconstituted into DMPC lipid film.

0 10 20 30 40 50 60 70

Figure 5.15.: Temperature-dependent frequency shift of the asymmetric CH₂ stretching vi-bration of the DMPC lipids in the BR in DMPC film (blue squares) and for a DMPC lipid film without protein (black triangles). Respective Boltzmann fits are depicted in red. Details on the evaluation of the phase transition temperature are given in the main text.

5. Results and Discussion

For the pure DMPC lipid film, a phase transition temperature of 24 ^◦C comparable to the literature value of 23.8 ^◦C could be monitored. Unexpectedly, a huge rise of the phase transition temperature for the BR in DMPC film was observed (T_m = 37 ^◦C).

The phase transition temperatures were determined by fitting the measured wavenumber peak position data points depending on temperature with a Boltzmann fit function and calculating the point of inflection of the respective fit. Moreover, the transition becomes broader, when incorporating the protein. For aqueous samples of BR in DPPC and pure DPPC, it was reported that the reconstitution of the protein led only to a broadening of the change of frequency associated with lipid phase transition without changing the T_m [102]. A possible explanation for the observed differences could be the different sample form used and therefore different water content. The lipids in rehydrated protein-lipid films might probably have a higher degree of order due to tight packing which would explain the observed higher phase transition temperature.

Time-resolved step-scan FTIR measurements were then performed from 0 ^◦C to 60 ^◦C in 5^◦C steps in order to observe the protonation dynamics of the primary proton transfer step below and above the phase transition temperature for BR in DMPC. Figure 5.16 shows the monitored time traces at 1760 cm^-1 for the different temperatures.

5.1. Bacteriorhodopsin

Figure 5.16.: Temperature-dependent time traces for BR in DMPC (14:0) liposomes at 1760 cm^-1 showing the protonation dynamics of primary proton acceptor Asp85.

Time-resolved step-scan FTIR measurements were performed from 0^◦C to 60^◦C in 5 ^◦C steps. Blue shadows indicate how the maximal absorbance changes shift to earlier times for increasing temperatures, as well as how the life-time of the M intermediate (indicated by the plateau of the transients) also changes.

5. Results and Discussion

At 0 ^◦C, the time trace at 1760 cm^-1 reveals the rise of the M state equivalent to the protonation dynamics of the primary proton transfer step from the Schiff base to Asp85 with a maximum around 2 ms. No decay of the M state is then observed to the end of our acquisition time range of 18 ms at this low temperature. With increasing temperatures, the maximum shifts to earlier times, meaning accelerated primary proton transfer dynamics. The following decay of the time traces which becomes visible within the acquisition range for higher temperatures, is indicative of the subsequent decay of the M state. E.g. at 30^◦C, the time trace shows the rise of the M state with a maximum around 0.3 ms and following decay of the M state. In the temperature range from 0 ^◦C to 35 ^◦C (DMPC lipids in the gel phase) and 40^◦C (transition phase), the time traces show a different curve profile than the time traces from 45^◦C to 60 ^◦C (DMPC lipids in the liquid phase). The time trace at 45 ^◦C shows a very fast rise of the M state, followed by a longer-lived plateau in the time trace with a following delayed decay of the M intermediate starting later than 1 ms. For 50^◦C and higher temperatures, Asp85 is already protonated at the beginning of the acquisition time range of 10µs. The time traces show also a delayed decay of the M intermediate. From 15^◦C to 40 ^◦C, the decay of the M intermediate directly follows the rise of the M intermediate without a plateau in the time traces.

The corresponding temperature-dependent time traces monitored at 1188 cm^-1 reveal similar altered dynamics below and above the phase transition temperature (Figure 5.17).

Starting at 45 ^◦C for the liquid lipid phase, the Schiff base remains deprotonated for a longer time period correlating with the observed longer-lived M intermediate, whereas between 25^◦C to 40 ^◦C deprotonation (decay) and reprotonation (rise) of the Schiff base and final thermal relaxation of the retinal (decay) can be observed within the time range of acquisition.

These measurements further indicate that the liquid lipid phase induces faster protonation of Asp85 followed by a delayed decay of the M intermediate as observed similarly for BR in DOPC and BR in DMoPC at 20 ^◦C (where both lipids are in the liquid phase) in subsection 5.1.3.

5.1. Bacteriorhodopsin

10 100 1000 10000

time / µs 0°C

5°C 10°C 15°C 20°C 25°C

/a.u.

30°C 35°C

40°C 45°C 50°C 55°C 60°C

Figure 5.17.: Corresponding temperature-dependent time traces for BR in DMPC (14:0) liposomes at 1188 cm^-1 showing the protonation dynamics of the Schiff base.

In order to assign the dynamics around the phase transition temperature more quantita-tively, anArrhenius-plot for the primary proton transfer step dynamics was attempted (Figure 5.18). A change of the slope is expected to take place at the phase transition temperature. However, only from 0 ^◦C to 45 ^◦C, rate constants could be calculated revealing correlated faster protonation dynamics for Schiff base deprotonation and Asp85 protonation for higher temperatures ("kinetic effect"). Time traces above 45 ^◦C did

5. Results and Discussion

not yield temperature-dependent rate constants for theArrhenius-plot since Asp85 is already protonated within the time-resolution of 10 µs. Therefore rate constants for the liquid lipid phase above 45^◦C could not be evaluated and hindered a more quantitative analysis of the suspected "lipid phase transition effect". Activation energies of 47.8 kJmol^-1 (Schiff base deprotonation) and 50.4 kJmol^-1 (Asp85 protonation) were determined in

this case for BR in DMPC.

0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 0.0037 8.0

Figure 5.18.: Arrhenius-plot of the temperature-dependent Schiff base deprotonation and Asp85 protonation dynamics of BR in DMPC for the time traces from 0 ^◦C to 45^◦C in 5 ^◦C steps. Time-traces from 50^◦C on had to be excluded from the Arrhenius-plot since Asp85 was already protonated at 10 µs. Therefore, not enough rate constants for the liquid lipid phase could be determined to observe an expected change of slope at the phase transition temperature (T_m = 37^◦C).

Time constants were obtained by fitting the time traces at 1188 cm^-1 from the beginning to the minimum and the time traces at 1760 cm^-1 from the beginning to the maximum, both with a monoexponential function.

The "lipid phase transition effect" was therefore further evaluated by comparing protonation dynamics of BR in DMPC (liquid phase, T_m = 37 ^◦C) and BR in DSPC (gel phase, T_m = 66 ^◦C), respectively at 50^◦C (Figure 5.19). Both lipids are saturated and only the effect of the lipid phase is evaluated in this case. BR in DMPC shows corroboratively faster protonation of Asp85 and longer-lived M intermediate as compared to BR in DSPC, supporting the observed differences for gel phase and lipid phase

5.1. Bacteriorhodopsin

dependent dynamics. The elevated phase transition temperature of T_m = 66 ^◦C for BR in DSPC was evaluated identically to the phase transition temperature of BR in DMPC by measuring the frequency shift of the asymmetric CH₂ stretching vibration (Figure B.6 in Appendix).

10 100 1000 10000

0.0 0.2 0.4 0.6 0.8 1.0

A(normalized)/a.u.

time / µs

Figure 5.19.: Protonation dynamics of Asp85 monitored at 1760 cm^-1 for BR in DSPC (red) and BR in DMPC (black) at 50 ^◦C.

In summary, protonation dynamics of the primary proton transfer step were analyzed below and above the measured phase transition temperature of T_m = 37 ^◦C. Protona-tion dynamics above 40 ^◦C for the lipid liquid phase are altered and indicate a faster protonation of Asp85 followed by a longer-lived M intermediate which is in accordance with measured altered dynamics for DOPC and DMoPC liquid lipid phase environments at 20^◦C (subsection 5.1.3). It is therefore concluded that the unsaturation as well as the lipid liquid phase have the same effect on the protonation dynamics. Unsaturated bonds or an increase ofgauche conformations in the lipid environment (lipid liquid phase) lead to a higher degree of conformational freedom for the protein which influences the proton translocation and functionality of the photoreceptor.

5. Results and Discussion