**3.4 Triplet quenching in HPCP**

It is of interest to investigate whether the same "triplet sharing" occurs in H-PCP as observed in A-PCP. In A-PCP, we previously observed the coexistence of Per and Chl-*a* carbonyl modes during the Per triplet lifetimes of 13 and 42 µs (Alexandre, Luhrs et al. 2007), which led us to conclude that in this triplet state, 3Chl-*a* and 3Per are mixed. Here, we investigated the H-PCP complex by step scan time-resolved FTIR spectroscopy to assess the nature of its triplet state. Upon excitation of peridinin at 530 nm, the H-PCP triplet decay is satisfactorily fitted with a single time constant of 10 µs. Excitation of H-PCP at 670 nm gave essentially the same result (data not shown). Figure 8 shows the DADS with a 10 µs lifetime in H-PCP (black line), plotted along with the 13 µs DADS observed in A-PCP (gray line), reproduced from ref. (Alexandre, Luhrs et al. 2007). Note that in A-PCP, also a 42 µs decay component with a distinctly different infrared signature was observed in the step-scan FTIR experiment (Alexandre, Luhrs et al. 2007). The H-PCP triplet state is spectrally very similar to the triplet state of A-PCP that decays in 13 µs. In both complexes, specific Per C=O lactone conformers are involved in the triplet state, of which the 1745(-)/~1720(+) cm-1 is the principal signature. In H-PCP, the bands at 1745(-)/1724(+) are assigned to the Per C=O lactone, while the 1700(-)/1667(+) cm-1 shift is attributed to the Chl-*a* C=O 9-keto (Breton, Nabedryk et al. 1999), showing that in the H-PCP triplet IR spectrum, explicit 3Chl-*a* modes are present. This result indicates that some Chl-*a*/Per specific conformations and interactions are conserved in both A-PCP and H-PCP to achieve the efficient photoprotective TEET. This is consistent with the similar T-S ODMR spectra obtained for both complexes (Carbonera, Giacometti et al. 1996). Thus we conclude that both in A-PCP and in H-PCP the triplet state is shared by Per and Chl-*a* in a similar fashion.

In the 10 µs DADS of H-PCP (Figure 8), at least two C=O lactone conformers (1745 and 1710 cm-1 in a relative stoichiometry of about 80 and 20%) can be identified for H-PCP, independent of the excitation wavelength. For A-PCP one or two conformers at 1745 and 1745-1725 cm-1 depending on the excitation wavelength have been observed, i.e. 670 and 480-530 nm for the 13 µs component (Alexandre, Luhrs et al. 2007). In the two components of A-PCP, three Per conformers have been identified with lactone carbonyl vibrations at 1745, 1741 and 1725 cm-1. This is consistent with the observation that the Per bound to the protein in H-PCP and A-PCP adopts multiple conformations (Hofmann, Wrench et al. 1996). Excitation of A-PCP yielded two triplet decay components of 13 and 42 µs (Alexandre, Luhrs et al. 2007). The differences between A-PCP and H-PCP may be related to the different protein structure, i.e., H-PCP is a 15.5 kDa homo-dimer and A-PCP is a 32 kDa trimer. The 32 kDa unit of A-PCP is asymmetric, composed of two different N-terminal and C-terminal units that share 56% of identity (Hofmann, Wrench et al. 1996). In contrast, H-PCP is a symmetric homo-dimer. The asymmetry within the A-PCP monomer is responsible for the Chl-*a* singlet equilibration on a ps timescale (Kleima, Hofmann et al. 2000; Salverda 2003). Possibly, the asymmetry in A-PCP leads to the observed triplet populations with different lifetimes.

It is of interest to investigate whether the same "triplet sharing" occurs in H-PCP as observed in A-PCP. In A-PCP, we previously observed the coexistence of Per and Chl-*a* carbonyl modes during the Per triplet lifetimes of 13 and 42 µs (Alexandre, Luhrs et al. 2007), which led us to conclude that in this triplet state, 3Chl-*a* and 3Per are mixed. Here, we investigated the H-PCP complex by step scan time-resolved FTIR spectroscopy to assess the nature of its triplet state. Upon excitation of peridinin at 530 nm, the H-PCP triplet decay is satisfactorily fitted with a single time constant of 10 µs. Excitation of H-PCP at 670 nm gave essentially the same result (data not shown). Figure 8 shows the DADS with a 10 µs lifetime in H-PCP (black line), plotted along with the 13 µs DADS observed in A-PCP (gray line), reproduced from ref. (Alexandre, Luhrs et al. 2007). Note that in A-PCP, also a 42 µs decay component with a distinctly different infrared signature was observed in the step-scan FTIR experiment (Alexandre, Luhrs et al. 2007). The H-PCP triplet state is spectrally very similar to the triplet state of A-PCP that decays in 13 µs. In both complexes, specific Per C=O lactone conformers are involved in the triplet state, of which the 1745(-)/~1720(+) cm-1 is the principal signature. In H-PCP, the bands at 1745(-)/1724(+) are assigned to the Per C=O lactone, while the 1700(-)/1667(+) cm-1 shift is attributed to the Chl-*a* C=O 9-keto (Breton, Nabedryk et al. 1999), showing that in the H-PCP triplet IR spectrum, explicit 3Chl-*a* modes are present. This result indicates that some Chl-*a*/Per specific conformations and interactions are conserved in both A-PCP and H-PCP to achieve the efficient photoprotective TEET. This is consistent with the similar T-S ODMR spectra obtained for both complexes (Carbonera, Giacometti et al. 1996). Thus we conclude that both in A-PCP and in

H-PCP the triplet state is shared by Per and Chl-*a* in a similar fashion.

In the 10 µs DADS of H-PCP (Figure 8), at least two C=O lactone conformers (1745 and 1710 cm-1 in a relative stoichiometry of about 80 and 20%) can be identified for H-PCP, independent of the excitation wavelength. For A-PCP one or two conformers at 1745 and 1745-1725 cm-1 depending on the excitation wavelength have been observed, i.e. 670 and 480-530 nm for the 13 µs component (Alexandre, Luhrs et al. 2007). In the two components of A-PCP, three Per conformers have been identified with lactone carbonyl vibrations at 1745, 1741 and 1725 cm-1. This is consistent with the observation that the Per bound to the protein in H-PCP and A-PCP adopts multiple conformations (Hofmann, Wrench et al. 1996). Excitation of A-PCP yielded two triplet decay components of 13 and 42 µs (Alexandre, Luhrs et al. 2007). The differences between A-PCP and H-PCP may be related to the different protein structure, i.e., H-PCP is a 15.5 kDa homo-dimer and A-PCP is a 32 kDa trimer. The 32 kDa unit of A-PCP is asymmetric, composed of two different N-terminal and C-terminal units that share 56% of identity (Hofmann, Wrench et al. 1996). In contrast, H-PCP is a symmetric homo-dimer. The asymmetry within the A-PCP monomer is responsible for the Chl-*a* singlet equilibration on a ps timescale (Kleima, Hofmann et al. 2000; Salverda 2003). Possibly, the asymmetry in A-PCP leads to the observed triplet populations with

pathways.

different lifetimes.

**3.4 Triplet quenching in HPCP** 

dynamics shows a strong wavelength dependence suggesting increased population of this Per conformer upon direct Per excitation. These spectral changes are accompanied by an overall increase of signal amplitude observed from 670 to 480 and to 530 nm. This wavelength dependence indicates that Per triplet formation proceeds via different

Fig. 8. Decay-Associated Difference Spectra (DADS) that result from a global analysis of step-scan FTIR data of H-PCP and A-PCP. The black line represents the 10 µs component in H-PCP. The gray line represents the 13 µs DADS in A-PCP.

The second component in A-PCP has been observed below 200 K by Carbonera *et al*. (Carbonera, Giacometti et al. 1999) and at 77K by Kleima *et al*., while only a 10 µs component was present at room temperature in the latter work (Kleima, Wendling et al. 2000). We reproduced Kleima's result at room temperature under aerobic and anaerobic conditions, by measuring the triplet decay in the visible using a diluted A-PCP solution. We found only the 10 µs component under aerobic and anaerobic conditions, which excludes an effect related to oxygen (data not shown). Thus, it seems that the appearance of the ~40 µs component is likely related to a protein conformational change induced by cooling and/or high concentrations used for FTIR, rather than a temperature effect on the equilibration among the triplet sublevels (Carbonera, Giacometti et al. 1999).

We previously proposed that the Per conformers involved in the photoprotective mechanism were likely Per 612/622 and Per 614/624. In a recent TR-EPR study (Di Valentin, Ceola et al. 2008), participation of the Per 612/622 pair in the 3Chl-*a* quenching was considered unlikely on the basis of the similarity of the 3Per triplet spectra in Main Form A-PCP (MFPCP) and High Salt A-PCP (HSPCP), since the latter does not bind the Per 612/622 pair. Comparison between experimental and calculated EPR spectra led the authors to conclude that the triplet was mainly (~80%) localized on the Per 614/624 pair, which has the shortest center-to-center distance to the Chl-*a*. This conclusion is consistent with the result of MD calculations showing that in the triplet state, the highest spin density is localized in the center of the Per backbone (Di Valentin, Ceola et al. 2008). Thus, it seems likely that in the step-scan FTIR experiments, the main Per conformer at 1745(-)/~1720(+) cm-1 corresponds to Per 614, Per 624, or both. Hence, we conclude that in A-PCP and H-PCP, Per 614 and/or 624 likely constitute the principal 3Chl-*a* quenchers, and that their specific interaction with Chl-*a* promotes the mixing of the 3Chl-*a* and 3Per states during the lifetime of 3Per.

Time-Resolved FTIR Difference Spectroscopy Reveals the Structure and Dynamics

**4. Comparison of higher plant and purple bacteria triplet state** 

Resonance Raman is a very sensitive and selective technique which allows access to the vibrational modes of molecules via inelastic scattering. In this section we describe the use of Resonance Raman to confirm the triplet delocalisation observed by step scan spectroscopy. In solution, upon triplet state formation, the frequency of stretching modes of the conjugated C=C's of the molecule dramatically downshifts from 1522 to 1494 cm-1 for ßcarotene in tetrahydrofuran (THF; (Hashimoto, Koyama et al. 1991), (Fujiwara, Yamauchi et al. 2008)), or from 1529 to 1500 cm-1 for *all-trans* spheroidene in hexane (Mukai-Kuroda, Fujii et al. 2002). This reflects the localization of the triplet throughout the conjugated system (triplet high spin density "hot spot" is mainly in the centre of the conjugated system; ((Mukai-Kuroda, Fujii et al. 2002), (Di Valentin, Ceola et al. 2008)), which causes a reduction in the C=C bond order. The frequencies of the 1 resonance Raman bands observed in protein-bound carotenoid triplet spectra are reported in Table 2 and compared to those observed for triplet states of other carotenoids, including ß-carotene in solution and spheroidene bound to the bacterial reaction center. The downshift depends on the carotenoid configuration, and it is always larger in *cis* configurations ((Hashimoto, Koyama et al. 1991; Mukai-Kuroda, Fujii et al. 2002)). For rhodopin glucoside in LH2, this downshift of the band from 1517 to 1493 cm-1 is very similar to that observed for all-trans ß-carotene and spheroidene in solution (24 cm-1 *vs.* 23-25 cm-1). In contrast, the observed downshifts for this band for both luteins in LHCII trimers, LHCII monomers, CP29 and CP43 are much smaller and in all cases close to 18 cm-1, *i.e.* approximately 75% of the 23-25 cm-1shift that is observed for all-*trans* carotenoid triplets in solution (see Table 2). Such a reduction of the 1

observed for higher plants.

of Carotenoid and Chlorophyll Triplets in Photosynthetic Light-Harvesting Complexes 247

difference spectrum (DADS) normalized to the contribution of the keto group at 1653 cm-1. Taking into account the time-resolved absorption data displayed above, the first decayassociated spectrum must be assigned to the carotenoid triplet (Fig. 9A, black spectrum and termed DADS1). In view of its long time the second one is attributed to unquenched Chl, *i.e.*  a small proportion of triplet chlorophyll states which have not been transferred to the carotenoid molecules (Fig. 9A, blue spectrum and termed DADS2). It has already been observed that a small fraction of chlorophyll triplet may not be quenched by the carotenoid molecules (Mozzo, Dall'Osto et al. 2008). Considering the results of time-resolved absorption, DADS1 should contain positive contributions of the carotenoid triplet state and negative contributions of its ground state. Although such contributions clearly appear in the difference spectrum (region termed luteins), additional bands are obviously present in this spectrum. Indeed, no intense contribution in the higher frequency region is expected from lutein molecules (see Fig. 9B). On the contrary, in this region, DADS1 is typical of the spectrum of a chlorophyll triplet in solvent which is plotted in Fig. 9C for comparison (Breton, Nabedryk et al. 1999; Bonetti, Alexandre et al. 2009). The negative contributions around 1700 cm-1 represent bleaching of bands arising from the stretching modes of conjugated keto carbonyl groups. These groups, when conjugated with the Chl macrocycle, experience large downshifts upon triplet formation, which results in positive contributions at lower frequencies. Nevertheless, DADS1 has a lifetime characteristic of carotenoid triplets, and so we conclude that in LHCII these chlorophyll infrared modes decay with the same lifetimes as carotenoid modes. Then "sharing" of the triplet wavefunction is also

#### **3.5 Triplet state in higher plants**

It has been recently reported that the TEET in LHCII takes place in less than 4 ns and is characterised by the lack of accumumaltion of Chl a triplet state. In order to understand the mechanisms underlying the very efficient triplet-triplet transfer, we investigated LHCII by step-scan time-resolved FTIR spectroscopy. Upon excitation of LHCII carotenoids at 475 nm, the triplet decay was satisfactorily fitted with two components only using global analysis. The first component decays with a time constant of 20 µs, while the second component, which does not account for more than 10% of the signal, does not decay within the time window of the measurement (about 320 µs). Figure 9A displays the first decay-associated

(A) The Decay-Associated Difference Spectra (DADS) have a 20 µs component (black trace, termed DADS1), with an amplitude of 90% (Lutein-Chls shared triplet) and a non-decaying component (blue trace termed DADS2) which has an amplitude of 10% (unquenched Chls). For clarity, the spectra have been normalized to the keto modes. For comparison, the FTIR of lutein (B) and Chl *a* (C) triplet (redrawn from (Bonetti, Alexandre et al. 2009)) in THF are also plotted

Fig. 9. Global analysis of step-scan FTIR data of LHCII excited at 475 nm showing the Lutein-Chls shared triplet state.

It has been recently reported that the TEET in LHCII takes place in less than 4 ns and is characterised by the lack of accumumaltion of Chl a triplet state. In order to understand the mechanisms underlying the very efficient triplet-triplet transfer, we investigated LHCII by step-scan time-resolved FTIR spectroscopy. Upon excitation of LHCII carotenoids at 475 nm, the triplet decay was satisfactorily fitted with two components only using global analysis. The first component decays with a time constant of 20 µs, while the second component, which does not account for more than 10% of the signal, does not decay within the time window of the measurement (about 320 µs). Figure 9A displays the first decay-associated

(A) The Decay-Associated Difference Spectra (DADS) have a 20 µs component (black trace, termed DADS1), with an amplitude of 90% (Lutein-Chls shared triplet) and a non-decaying component (blue trace termed DADS2) which has an amplitude of 10% (unquenched Chls). For clarity, the spectra have been normalized to the keto modes. For comparison, the FTIR of lutein (B) and Chl *a* (C) triplet

Fig. 9. Global analysis of step-scan FTIR data of LHCII excited at 475 nm showing the

(redrawn from (Bonetti, Alexandre et al. 2009)) in THF are also plotted

Lutein-Chls shared triplet state.

**3.5 Triplet state in higher plants** 

difference spectrum (DADS) normalized to the contribution of the keto group at 1653 cm-1. Taking into account the time-resolved absorption data displayed above, the first decayassociated spectrum must be assigned to the carotenoid triplet (Fig. 9A, black spectrum and termed DADS1). In view of its long time the second one is attributed to unquenched Chl, *i.e.*  a small proportion of triplet chlorophyll states which have not been transferred to the carotenoid molecules (Fig. 9A, blue spectrum and termed DADS2). It has already been observed that a small fraction of chlorophyll triplet may not be quenched by the carotenoid molecules (Mozzo, Dall'Osto et al. 2008). Considering the results of time-resolved absorption, DADS1 should contain positive contributions of the carotenoid triplet state and negative contributions of its ground state. Although such contributions clearly appear in the difference spectrum (region termed luteins), additional bands are obviously present in this spectrum. Indeed, no intense contribution in the higher frequency region is expected from lutein molecules (see Fig. 9B). On the contrary, in this region, DADS1 is typical of the spectrum of a chlorophyll triplet in solvent which is plotted in Fig. 9C for comparison (Breton, Nabedryk et al. 1999; Bonetti, Alexandre et al. 2009). The negative contributions around 1700 cm-1 represent bleaching of bands arising from the stretching modes of conjugated keto carbonyl groups. These groups, when conjugated with the Chl macrocycle, experience large downshifts upon triplet formation, which results in positive contributions at lower frequencies. Nevertheless, DADS1 has a lifetime characteristic of carotenoid triplets, and so we conclude that in LHCII these chlorophyll infrared modes decay with the same lifetimes as carotenoid modes. Then "sharing" of the triplet wavefunction is also observed for higher plants.
