Single-molecule trapping and spectroscopy reveals photophysical heterogeneity of phycobilisomes quenched by Orange Carotenoid Protein

The Orange Carotenoid Protein (OCP) is a cytosolic photosensor that is responsible for non-photochemical quenching (NPQ) of the light-harvesting process in most cyanobacteria. Upon photoactivation by blue-green light, OCP binds to the phycobilisome antenna complex, providing an excitonic trap to thermally dissipate excess energy. At present, both the binding site and NPQ mechanism of OCP are unknown. Using an Anti-Brownian ELectrokinetic (ABEL) trap, we isolate single phycobilisomes in free solution, both in the presence and absence of activated OCP, to directly determine the photophysics and heterogeneity of OCP-quenched phycobilisomes. Surprisingly, we observe two distinct OCP-quenched states, with lifetimes 0.09 ns (6% of unquenched brightness) and 0.21 ns (11% brightness). Photon-by-photon Monte Carlo simulations of exciton transfer through the phycobilisome suggest that the observed quenched states are kinetically consistent with either two or one bound OCPs, respectively, underscoring an additional mechanism for excitation control in this key photosynthetic unit.


Supplementary Note 1: Bulk measurements show a low percentage of unquenched complexes in the quenched samples
The bulk measurements made in this study (lifetime, emission spectrum) are mutually consistent with a very low percentage of unquenched complexes initially present in the quenched samples (~3.7%). The fluorescence decay of the PB+OCP bulk sample was fit with a bi-exponential decay, which prior to convolution with the IRF and addition of the expected background can be simply written as: That is, this bulk lifetime decay shows that <5% of complexes are unquenched. We can then calculate the expected total brightness ratio of this quenched sample to an unquenched sample, again assuming that quenched complexes are 6% as bright as unquenched complexes: 96.3% * 0.06 + 3.7% * 1 = 9.5% We expect that the bulk sample from the lifetime decay shown in Figure 1e will be about 9.5% as bright as the unquenched sample. This is very close to the ratio of the integrated areas under the spectra shown in Figure 1d, which shows that the quenched sample is 9% as bright as the unquenched sample.
Put another way, we could say that the ratio of emission spectra brightnesses (9%) shown in Figure 1d implies that only about 3% of complexes could be in state U in the quenched sample, if we assume that quenched complexes are 6% as bright as unquenched complexes.
Additional complexities including the presence of some Q1 complexes and/or the presence of decoupled rods are not included here, and indeed might alter these numbers somewhat. Nevertheless, the close agreement of these calculations to the measurements serves to demonstrate that the bulk fluorescence lifetime data and the bulk fluorescence emission spectrum data are mutually consistent, and indicate that very few unquenched complexes are present in the quenched sample.
Moreover, these numbers also imply that the bulk data are consistent with the reported singlemolecule quenching levels (6% for Q2), as well as with our suggestion that most complexes are quenched into the Q2 state upon (saturated ratio) binding of OCP.

Supplementary Note 2: The electric fields of the ABEL trap do not alter the photophysical properties of analyte molecules
In the trap, events from quenched complexes are photostable and many seconds in duration.
Unquenched complexes (at lower illumination powers) are also photostable. Moreover, the lifetime and emission spectrum of the unquenched complex is identical to bulk. Therefore, it is evident that the ABEL trap does not substantively alter the photophysical properties of the phycobilisome upon trapping. Additionally, the topic of sample perturbation has been addressed in several previous ABEL trap studies -we find that the rotational diffusion, 1 binding and unbinding kinetics, 2 structural conformation of biopolymers, 3 and fluorescence parameters 4-6 are unchanged from their bulk values.

Supplementary Note 3: Transitions among quenched and unquenched states of the phycobilisome are infrequently observed in the ABEL trap
In this work we do not claim to definitively identify transitions among the various states we have identified. This is for three main reasons: 1. Regarding the expectation of observing transitions during a trapping event: Our timedependent unbinding data ( Figure 4) indicates that the PB-OCP complex is relatively stable, and OCP unbinds over the course of hours. Here, we trap each complex for no more than a few seconds. Therefore, we expect that we would only rarely observe transitions during trapping events.
2. We do occasionally observe trapped objects that appear to change their state (a couple of examples of this are evident in the extended raw data set presented in SI Figure S3). However, these events are rare. This is consistent with the expectation described above for a stable PB+OCP complex, but it is also possible that those rare transitions could be replacement events in the ABEL trap, where a second complex (which might be in a different state from the trapped object) enters the trapping area and by chance replaces the initially trapped object. While we work at very low concentration to minimize this possibility, it cannot be discounted in the case of rare events. We therefore do not speculate in the manuscript as to the nature of these events.
3. Moreover, the active observation of binding or unbinding of OCP to the phycobilisome is more relevant to its binding kinetics than to the photophysical states present in the quenched complex, which is the topic of this work.

Supplementary Note 4: Alternative combinations of binding sites and non-degenerate binding sites
In our Discussion, we present simulation results for two OCPs bound at the four pairs of rotationally symmetric compartments. We excluded non-rotationally symmetric combinations and more than two OCPs from consideration because these conditions would generally predict more than two quenched populations except under specific conditions. For example, we expect that three distinct populations would be observed for binding at two asymmetric sites: 1) the doubly-quenched population, 2) one of the single-quenching sites, and 3) the other single-quenching site. These would collapse into fewer than two populations only under special conditions, including cooperative binding of the two OCPs (most likely producing just one population) or if the two non-symmetric sites just happened to produce spectroscopically identical states (two populations).
For more than two sites (for example, for three bound) you would expect to see 1) three bound OCPs, 2) at least one (but likely more than one) quenching level that represents two bound OCPs and 3) at least one quenching level that represents one bound OCP (again, likely more than one). So, for three bound OCPs, we would expect to see at least three or more distinct populations. Again, this possibility cannot be completely eliminated because it is conceivable that strong cooperative binding or similar anomalous factors might prevent some of these combinatorically possible states from being populated. However, in our view the simplest possible explanation for seeing precisely two populations is that two OCPs can bind at symmetric sites.
To illustrate what happens if more than two OCPs are bound, we simulated a three-quencher scenario that includes two symmetric sites (therefore producing some degeneracy and lowering the number of expected states). We selected the quencher strength by requiring that the triply-quenched complex produce the Q2 photophysical parameters ( Supplementary Fig. 11a), and then used this quencher strength to simulate the various possible combinations of two and one quenchers ( Supplementary Fig.  11b). As described above, the results predict that this scenario would produce at least 4-5 mutually distinguishable quenched populations.

Supplementary Figures
Supplementary Figure 1

: Brightness of unquenched CB-PB is linear with excitation intensity
The initial brightness of trapped CB-PB complexes is plotted as a function of excitation intensity in the trap. Nonlinear effects are observed above excitation intensities of ~100 W cm -2 , which may be due to either singlet-singlet annihilation, or an extremely brief initial brightness state that is not resolved here. At low excitation intensities on this scale (< 20 W cm -2 ), little or no photodamage is observed for the unquenched phycobilisome. At higher excitation intensities (> 10 W cm -2 ), progressive photodamage is apparent in trapped complexes (see Supplementary Figure 2), although the brightness of the initial state continues to be linear with excitation intensity up to ~100 W cm -2 , as shown here. Dark blue circles indicate data taken with pulsed excitation (Mira OPO / 80 MHz @ 594 nm), while cyan markers indicate data taken with CW excitation on three different days with realignment in between. Least-squares linear fits are shown with dotted lines.

Supplementary Figure 2: Anti-Brownian trapping of single unquenched phycobilisomes under high excitation intensity
Raw trapping data for all parameters (Br, FPol, τ, λCM, and Em(λ)) for the CB-PB complex in the absence of OCP under high incident excitation intensity (50 W cm -2 ) shows progressive photodegradation and blinking of the phycobilisome. Data is plotted in either 800-photon groups (markers) or 20-ms bins. Notably, while photodamaged CB-PBs can reach the brightness levels of Q1 and Q2, the lifetime remains at or above ~0.3 ns even for large amounts of photodamage. This is clearly distinct from the lifetimes observed for Q1 and Q2, τQ1 = 0.21 ns and τQ2 = 0.09 ns.

Supplementary Figure 3: Scatter density plots for unquenched phycobilisome at high excitation power
Scatter heatmap of the photophysical states (Udamage) observed for the unquenched CB-PB phycobilisome excited at high power (25 W cm -2 ), shown in Br-τ, Br-λCM, and Br-FPol projections (left, right, and center, respectively). Each point represents 250 photons, colored according to the local density of points. The location of state B, very rarely observed here, is also indicated for reference. Despite the heterogeneity of photophysical states observed, Q1 and Q2 are not observed in this sample (see inset). Colorbar indicates scatter plot coloration according to normalized local density of data points within each panel.

Supplementary Figure 4: (Previous page) Extended raw data trace for ABEL trapping of single unquenched phycobilisomes
Raw trapping data is shown for Br and τ for OCP-quenched CB-PB complexes (40:1 OCP:CB-PB) over a representative 5-minute data set. Most events contain a single state, and those that do not (e.g.: at 41 sec, at 147 sec, and at 293 sec) may represent replacement events rather than actual transitions among states (see also SI Note S1 and SI Fig. S9). Photophysical parameters within individual states are generally stable over time. Only three trapping events are likely to represent unquenched CB-PB complexes (at 47 sec, 125 sec, and 147 sec), and several free C-PC hexamers are observed (at 24 sec, 67 sec, 103 sec, 137 sec, 232 sec, and 293 sec). States with very dim brightness but long lifetime may represent free trimers or monomers of C-PC or APC (92 sec, 109 sec, 111 sec, 157 sec, 187 sec, and 295 sec).

Supplementary Figure 5: Histograms for state U
Gaussian fits for all parameters, a) Br, b) τ, c) λCM, and d) FPol, for the unquenched CB-PB phycobilisomes (state U), created from the 500-photon groups shown in the scatter plots for Figure 3a.
Most likely values and standard deviations in each dimension are also presented in Table 1.  Table 1.

Supplementary Figure 9: Fluorescence lifetime and brightness of C-PC rods in the ABEL trap
C-PC rods from the ΔAB mutant of Synechocystis PCC 6803 7 do not include the phycobilisome core. The expected rod structure for this mutant is three hexamers associated face-to-face via the LR33, LR30, and LR9 (rod linker proteins) and CpcG2 rod core linker protein 7,8 . Data was analyzed using the same workflow described in Methods for the CB-PB and CB-PB + OCP data. Some rods dissociate in solution, producing three primary populations: 3-hexamer rods, 2-hexamer rods, and single hexamers. a) Fluorescence lifetime decay histogram for single C-PC rods observed in the ABEL trap. The observed lifetime, 1.63 ns, which is well-fit with a single-exponential decay, closely matches the lifetime observed for state B. b) Histogram and 3-Gaussian fit of brightness data from ABEL trapping of C-PC rods. The above fit was performed under the constraint that Br3Hex = 3*Br1Hex and Br2Hex = 2*Br1Hex to determine the expected brightness and standard deviation for a single C-PC hexamer as Br1Hex = 1505 ± 400 cts s -1 µW -1 . The brightness of a single C-PC hexamer from this sample is similar to the observed brightness of state B observed in the quenched CB-PB + OCP sample and occasionally observed in the unquenched CB-PB sample (1630 cts s -1 µW -1 ). Further characterization of the CpcG2-PBS mutant will be reported elsewhere 8 . (3) Two bound quenchers at a'e or d'e, which happen to produce nearly indistinguishable photophysical states, shown in gray-blue, (4) a single bound quencher at a' or d', which are symmetric and therefore produce degenerate photophysical states, shown in purple, and (5) a single bound quencher at e, shown in magenta. As described above in SI Note 4, the results predict that this scenario would produce 4-5 distinct quenched populations, rather than the two that were observed.

Supplementary Tables
Supplementary Table 1: Connectivity of compartmental model Rate matrix showing the connectivity of the compartmental model and the associated rates of forward and backward transfer for each connected pair of compartments. All rates are in ns -1 . Quenching via OCP during different simulations is turned on (given a baseline rate of 54 ns -1 ) or off (given a rate of 0 ns -1 ) to test different quenching sites and different combinations of sites. Which compartments have an active quencher for a given simulation is stated in the main text.