Skip to main content

4.1: Excitation energy transfer in light harvesting complex LH2 of purple bacteria

Photosynthesis is the process taking place in plants, some bacteria and algae, during which almost all the energy used by all living organisms (including us) is captured and stored in the form of sugars. This process starts when light quanta are absorbed and their energy is used to separate charges across biological lipid membrane. Later, the energy of separated charges is used for synthesizing ATP and NADPH. These are further used in Calvin cycle where carbohydrates are made out of CO2 [9]. Energy transfer and charge separation must compete with the lifetime of excited states in the photosynthetic pigments. Therefore, when the photon is absorbed, it takes only several tens of picoseconds for the energy to reach the reaction centre protein, where primary charge separation takes place. Overall quantum efficiency of photosynthesis is close to unity [10].

The photosynthetic apparatus of purple bacteria called Rhodobacter sphaeroides and Rhodopseudomonas acidophila consists of two types of light harvesting antennae and reaction center. The task of pigment-protein complexes called light harvesting antennae is to absorb light and transfer the energy of the excitation into the reaction center, where the energy is utilized for charge separation across the membrane of the cell. One of the antennae called LH2 (light-harvesting complex-2) has a three dimensional structure shown in Fig. 19. This structure is determined from x-ray diffraction on the crystal of LH2 protein [11]. Figure 19 also contains the absorption spectrum of this complex. As obvious from the figure, LH2 consists of two sets of bacteriochlorophyll pigments and a set of carotenoid pigments arranged in concentric circles. Thus the entire complex consists of nine circularly arranged subunits, each containing three bacteriochlorophylls and one carotenoid molecule. The lowest absorption band[1] of bacteriochlorophylls in the closely packed ring (green color in Fig. 19) is at 850 nm, therefore these bacteriochlorophylls are called B850. Loosely packed bacteriochlorophyll ring absorbs 800 nm light and is called B800. The Qx transitions of all the bacteriochlorophylls correspond to the absorption band at 590 nm. Carotenoids (elongated pigments shown in orange) absorb in the visible spectral range at 450-550 nm. They supplement the absorption spectrum of the photosynthetic apparatus and absorb the wavelengths, to which bacteriochlorophylls are transparent. Time-resolved fluorescence experiments performed by R.Jimenez and coworkers [12] helped to elucidate how energy is transferred between bacteriochlorophylls with identical and different absorption maxima.



Fig. 1) Three-dimensional structure (A) [11] and absorption spectrum (B) of LH2 of Rhodopseudomonas acidophila. Carotenoind molecules shown in orange absorb at 450-550 nm, blue bacteriochlorophylls (B800) – at 800 nm and green bacteriochlorphylls (B850) at 850 nm. The protein is shown as helices.


Energy transfer from B800 to B850 can be probed by exciting B800 pigments and monitoring the fluorescence at 900 nm. The energy will initially be located on B800 pigments, and should be transferred downhill to B850 pigments. Therefore, initial fluorescence intensity at 900 nm, where B850 pigments dominate will initially be low, and, as energy is received from B800, it should increase.

The results of such experiment are shown in Fig. 20 A. As expected, immediately after the excitation of B800 pigments, the fluorescence is virtually zero (signal at t=0). However within 650 fs the signal rises and reaches its maximum. This corresponds to the energy transfer time from B800 ring of bacteriochlorophylls to B850.


Fig. 2) Fluorescence upconversion experiments on the light harvesting antenna LH2 of purple bacteria. A: Fluorescence kinetics at 940 nm after excitation of 800 nm absorption band. B: Polarized fluorescence experiment with the excitation in B850 bacteriochlorphylls. Top curve is the measurement with detection and excitation polarizations in parallel, bottom curve perpendicular. Kinetics of fluorescence anisotropy calculated using Eq. (4.14.5). Reproduced from [12].



So it is easy to measure energy transfer B800 à B850 in this fashion (well, not exactly easy keeping in mind all experimental difficulties and hitches that are usually left out of the scientific papers). But at least it is clear how this can be done. On the other hand, observation of energy transfer within B850 ring poses a greater challenge, because the energies of all pigments are all the same (corresponding to the energy of 850 nm photon) and energy transfer cannot be observed by selecting an appropriate detection wavelength as it was done in B800 à B850 transfer. R.Jimenez and coworkers have solved this problem by measuring the depolarization of fluorescence, i.e. exciting the B850 bacteriochlorophylls and recording the fluorescence polarized parallel and perpendicular to the polarization direction of the excitation beam (Fig. 21). Fluorescence anisotropy is defined as follows:



Fig. 3) Measurement of fluorescence parallel and perpendicular to the excitation polarization



It can be shown that in the cases when fluorescence occurs from the same state that was originally excited, and the molecules are not rotating, the ensemble of randomly oriented pigments, yields fluorescence anisotropy equal to 0.4 [13]. In the context of time-resolved fluorescence, it means that initial anisotropy is allways equal to 0.4, because initially the emission will occur from the same pigment that was originally excited. Later different reasons may determine the depolarization of fluorescence (i.e. the directions of radiating dipoles may become different from the absorbing ones). These reasons include

  • Rotational diffusion of molecules,
  • Molecular changes (photoreactions) in the excited state,
  • Radiationless relaxation to a different molecular state with a differently oriented transition dipole moment,
  • Energy transfer,
  • Etc.

We are interested in the depolarization due to energy transfer, because this is what we would like to watch in a time-resolved fluorescence experiment. It is obvious that after the hopping of excitation from one pigment to another, with differently oriented transition dipole moment, fluorescence depolarization will occur, even if the wavelength of fluorescence remains the same. Depolarization of fluorescence recorded in LH2 B850 ring is shown in Fig. 20B, C. Panel B shows the fluorescence signal with LH2 excitation at 850 nm and detection of parallel[2] (top curve) and perpendicular (bottom curve) components at 940 nm. First, we must note that none of the curves shows ultrafast dynamics with characteristic times of 100 fs or less. However, if the measured curves are plugged into the Eq. and anisotropy is calculated, femtosecond processes come unraveled. The anisotropy change is plotted in Fig. 20C. It is obvious that depolarization of B850 fluorescence is lighning-fast: instrument response function (160 fs in this experiment) is too slow to capture the initial value of 0.4, and within 200 fs, the final polarization value is established, equal to 0.06. This is an indication that the energy transfer rate in B850 ring is extremely fast.

Modelling of the fluorescence anisotropy performed by R. Jimenez and coworkers has revealed that a single hop of energy between the pair of B850 bacteriochlorophylls must be roughly 100 fs. As obvious from the final value of anisotropy in Fig. 20C, the fluorescence remains partly polarized even after a large number of energy hops. The reason for this is that a single LH2 ring contains just 18 bacteriochlorophylls, in a rigid cylindrical arrangement (see Fig. 19A), and some information about the direction of initially excited transition is retained even after the excitation is completely randomized within the ring. Losely speaking, this can be interpreted as the excitation sometimes returning to the originally excited pigment, which will re-polarize the fluorescence signal.

To conclude, polarized fluorescence experiments have answered the question that could not be addressed by isotropic experiments: they allow detecting energy transfer between isoenergetic pigments (pigments with identical transition energies), which in case of B850 ring in LH2 is 100 fs.



[1] In bacteriochlorophylls and other porphyrin type pigments, the lowest energy transition is called Qy, the second-lowest is Qx, and the transitions corresponding to the blue-violet absorption are called Soret.

[2] Terms parallel and perpendicular are defined with respect to the polarization of excitation beam.