Ensemble fluorescence-emission spectroscopy

For the acquisition of fluorescence-emission spectra from thin-film ensembles of RC-LH1 complexes, the setup was used in the confocal mode. The detection unit, consisting of a spectrometer (SpectraPro-150, 300 lines/mm, Acton Research Cor-poration, Acton, USA) with a CCD camera (Luca-R 604M-OM, Andor Technology) attached to it, was positioned opposite to the CCD camera for wide-field imaging (Figure 3.2a). The emission signal from the sample was then deflected into the di-rection of the spectrograph by means of a flip mirror [82]. The sample was excited at 800 nm through an excitation bandpass filter (BP 805/60, AHF Analysetech-nik). The emission from the sample passed through a pair of longpass filters (LP 830, AHF Analysetechnik), was dispersed in the spectrometer and imaged onto the CCD chip. The spectral resolution of the combined spectrograph/camera system amounted to 1.5 nm (20 cm⁻¹) and all emission spectra have been corrected for the spectral sensitivity of the combined system. The exposure time for the mea-surement of the ensemble fluorescence-emission spectra amounted to 15s and the excitation intensity was 14kW/cm².

4 Spectroscopy on RC-LH1

complexes from Rps. acidophila

In the first part of this chapter the RC-LH1 complex fromRps. acidophila has been revisited for single-molecule spectroscopy. Thereby, the pigment-protein complexes were stabilized in the detergent buffer solution using a relatively mild detergent dodecyl-β-D-maltoside (DDM), instead of lauryldimethylamine N-oxide (LDAO) which was applied in a previous study [10]. This leads to a significant reduction of the fraction of broken/dissociated RC-LH1 complexes with respect to [10] and thus, it was possible to investigate a sufficiently large sample of individual RC-LH1 complexes. For most of the complexes the fluorescence-excitation spectra exhibit a narrow spectral feature at the red end of the spectrum. Analysis of the statistics of the spectral properties yields a close resemblance with the results obtained on RC-LH1 complexes fromRps. palustris for which a low-resolution x-ray structure is available. Based on this comparison the conclusion is gained that for both species the LH1 complex can be described by the same structural model, i.e. an overall elliptical assembly of pigments that features a gap.

In the second part of this chapter RC-LH1 complexes from Rps. acidophila sta-bilized in detergent buffer solution and reconstituted into a phospholipid bilayer have been investigated, and the results were compared with the outcome of the first part, conducted on RC-LH1 immobilized in polyvinyl alcohol (PVA). The aim of this study was to test whether the immobilization of the complexes in a PVA matrix might lead to a deterioration of the proteins and thereby limit the accessi-ble information that can be obtained from optical spectroscopy. It has been found that the complexes dissolved in a detergent buffer solution, without stabilization in a PVA matrix, are subject to fast spectral dynamics preventing any meaningful application of single-molecule spectroscopy. Furthermore, for the complexes recon-stituted into a lipid bilayer it is revealed that the reconstitution process results in a significantly larger fraction of broken complexes with respect to the preparation of the complexes in a PVA film. However, it is also found that for the intact com-plexes the statistics of the key spectral features, such as the spectral separations of the bands and the mutual orientation of their transition-dipole moments, show

4 Spectroscopy on RC-LH1 complexes from Rps. acidophila

no difference as a function of using either a bilayer or PVA as a matrix. Given the additional effort involved in the reconstitution process, the lower amount of intact RC-LH1 complexes, and, concerning the decisive spectral details, the iden-tical results with respect to embedding the complexes in a PVA matrix, led to the conclusion that the immobilization of these proteins in a PVA matrix is a good choice for conducting low-temperature experiments on individual light-harvesting complexes.

4.1 Spectroscopy on RC-LH1 complexes stabilized in the detergent DDM

4.1.1 The RC-LH1 complex from Rps. acidophila

Figure 4.1: 19 Å resolution projection map of negatively stained two-dimensional crystals of RC-LH1 complexes from Rps. acidophila, strain 7750 (from [2]). The length of the scale bar corresponds to12nm.

So far structural information on the RC-LH1 complex from Rps. acidophila is very limited. There is only a projection map of negatively stained two-dimensional crys-tals of RC-LH1 complexes from the strain 7750 of Rps. acidophila available (in contrast to strain 10050 investigated in this work), however with a very poor res-olution of only 19 Å (Figure 4.1). From this projection map one could infer that the LH1 complexes from this bacterial species are circular and completely surround their corresponding RC. Then again the projection map originates from a time (1994), where such a type of LH1 model was dogma and actually the map was created by assuming a six-fold symmetry for a complete LH1 ring [2]. Therefore,

4.1 Spectroscopy on RC-LH1 complexes stabilized in the detergent DDM

given also the very low resolution of the projection map, the only information that can be extracted from it, is that the RC-LH1 complexes from Rps. acidophila are monomeric, but neither the precise shape (circular vs. elliptical) of the LH1 com-plex, nor if the LH1 aggregate has a gap or not, are revealed.

It is now well established that the spatial arrangement of the pigments determines to a large extent the spectroscopic features of the complexes and that in these sys-tems collective effects have to be considered in order to appropriately describe their electronically excited states [2, 5, 12, 14, 15, 61, 62, 64, 70, 83–88]. This leads to the so called Frenkel excitons, which arise from the interactions of the transition-dipole moments of the individual pigments, and which correspond to delocalized electron-ically excited states. The presence of a gap in the chromophore arrangement has severe consequences for the electronic structure of the complex and the photophys-ical parameters of the exciton states. Firstly, degeneracies between the exciton states are lifted and following common practice the exciton states are numbered k = 1,2, . . . , N [59], with N being the total number of pigments in the aggregate.

Secondly, significant oscillator strength is shifted to the lowest exciton state, i.e.

k = 1, in marked contrast to a closed-ring BChl a arrangement, as for example for LH2, where nearly all oscillator strength is accumulated in the next higher ex-citon states [11, 12, 14, 85]. Since the fluorescence lifetime of the k = 1 state is in the order of 600 ps [56, 89] which is rather long with respect to the relaxation of the higher exciton states, that takes place on a 100 f s timescale [27, 90], this leads to the occurrence of a relatively narrow spectral feature at the red-edge of the absorption spectrum. Exploiting single-molecule spectroscopy those features could be made visible and not only confirmed the presence of the gap in LH1 from Rps.

palustris, but also gave rise to a refinement of details of the x-ray structure [12, 15].

The first single-molecule experiments on RC-LH1 complexes were conducted by Ketelaars et al. [10] on those from the species Rps. acidophila. This study was a follow up of the experiments on LH2 complexes from the same species [11, 13, 14].

Unfortunately, the resulting spectra revealed a large degree of spectral heterogene-ity and it was argued that in some of the complexes the LH1 rings were not fully intact or might even be decomposed. The spectral signature of those LH1 com-plexes which were considered intact, was interpreted as originating from a closed ring structure. However, this interpretation was based upon the work performed on the LH2 complexes [13, 14] for which such a structural composition is well known.

This raises the question if the conclusions drawn from the early LH1 spectra for the buildup of their structures were really correct.

By means of simple room-temperature ensemble absorption experiments clear evi-dence was already found for a partial dissociation of the RC-LH1 complexes from Rps. acidophila if they were stabilized in the detergent lauryldimethylamine N-oxide (LDAO), as it was used in the previous study [10] (see following section 4.1.2). Thus, given the remaining structural uncertainties for the RC-LH1 com-plex from Rps. acidophila and the encouraging results obtained on RC-LH1 from

4 Spectroscopy on RC-LH1 complexes from Rps. acidophila

Rps. palustris [12, 15], in this work the RC-LH1 complex fromRps. acidophila was revisited for single-molecule spectroscopy, now using the milder detergent dodecyl-β-D-maltoside (DDM) for the stabilization of the complexes in the buffer solution.