Synechocystis sp. PCC 6803 WT and a strain deficient in ORR, ORR1, were used for experimentation. The ORR1 strain was generated by knocking out the flv1 gene encoding flavodiiron 1, as described in (Smolinski et al., 2022). Synechocystis sp. PCC 6803 WT and ORR1 strains were grown in 5 mM N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES)-NaOH, pH 8.2 buffered BG-11 (Allen, 1968) media under continuous light (35 µmol photons m^-2 s^-1) at room temperature with 3% CO₂. For the fluctuating light condition, cells were first grown under constant light of 35 µmol photons m^-2 s^-1 for 3 days, then shifted to cycles of 35 µmol photons m^-2 s^-1 for 5 min, and 500 µmol photons m^-2 s^-1 for 30 s, for a total of 4 days of growth.

Thylakoid membranes (TMs) were isolated from the cells similar to Li et al (Li et al., 1991). Briefly, all procedures were performed at 4 °C. Cells were spun down at 5,000 x g for 10 min, and the pellet was resuspended in buffer containing 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-NaOH, pH 7.5; 10 mM CaCl₂; 10 mM MgCl₂ and 10 mM NaCl (buffer A). Cells were lysed by three passes through a French Press cell, while kept on ice. After the cell lysate was spun down at 5,000 x g for 10 min, the supernatant was collected and then spun at 208,000 x g for 1 h (Type-70Ti rotor, Beckman Coulter). The pellet containing TMs was resuspended in buffer A to approximately 1 mg chl ml^-1. TMs were incubated with 1% n-dodecyl-β-D-maltoside (DDM) (GoldBio) for 2 h in darkness, and the chlorophyll (chl) concentration was adjusted to 0.4-0.5 mg ml^-1. Solubilized TMs were centrifuged at 12,000 x g for 20 min to pellet non-solubilized material. The supernatant was then loaded on to 5-30% (w/vol) linear sucrose gradient with buffer A containing 0.01% DDM. The gradient was centrifuged at 110,000 x g, overnight. Each colored fraction was carefully collected from the top of the sucrose gradient. The PSI monomer/PSII and PSI trimer fractions were further concentrated by 10 kD cutoff filter (Amicon Ultra-15, 10 kD cutoff, Sigma), subsequently flash frozen in liquid nitrogen and stored in -80°C freezer.

CpcL-PBS was isolated from cells according to Hirose et al (Hirose et al., 2019). The following experiments were performed at room temperature (RT). The harvested cells were spun down at 5,000 x g for 10 min and then the pellets were resuspended in 0.85 M phosphate buffer, pH 7.0 twice. The cells were broken by passing through French press cell 3 times. The cell homogenates were centrifuged at 5,000 x g for 10 min. The supernatants were incubated with 2% (v/v) Triton X-100 and 0.5% (w/v) DDM in the dark for 45 min. After incubation, the mixtures were loaded onto 5-30% (w/vol) linear sucrose gradient with 0.85 M phosphate buffer, pH 7.0 containing 0.1% Triton and 0.1% DDM. The gradients were centrifuged at 110,000 x g at RT overnight. The colored fractions were carefully collected from the top of the sucrose gradient. The CpcL-PBS and CpcG-PBS were wrapped with foil and stored in 4°C fridge until further use.

The PBS fractions were concentrated and exchanged into low phosphate buffer by 3 kD cutoff filter (Amicon, Ultra-0.5, 3 kD cutoff, Sigma) @ room temperature. Before denaturing the samples, the amounts of PBS were normalized to the same PC absorption peak. Proteins were incubated on ice with 20% (v/v) trichloroacetic acid (TCA) for 1 h and then spun down @ 17,300 x g for 30 min. The pellets were resuspended in 10% (v/v) TCA once then acetone twice by spinning the samples down @ 17,300 x g for 10 min. The final pellets were resuspended with 62.5 mM Tris-HCl pH 6.8, 2% (w/v) sodium lauryl sulfate, 5% (w/v) sucrose, 0.005% (w/v) bromophenol blue and 100 mM dithiothreitol and the mixtures were incubated at 65°C for 3 min. The reaction mixtures were loaded onto a pre-made gel (Any kD™ Mini-PROTEAN^® TGX Stain-Free™ Protein Gels #4568124) and run for 30 min at 200 V.

The amounts of PSI were normalized to the same amount of chl concentration. Proteins were added to a 1:1 mixture with 2X SDS-PAGE sample buffer (Bio-Rad 2x Laemmli Sample Buffer #1610737) with 5% (w/v) 2-mercaptoethanol, and incubated for 10 minutes at 99°C. The sample mixtures were loaded onto a pre-made gel (Any kD™ Mini-PROTEAN^® TGX Stain-Free™ Protein Gels #4568124) and run for 65 min at 120 V.

In vitro crosslinking experiments were performed similar to Hartmann et al (Hartmann et al., 2020). Isolated CpcL-PBS or CpcG-PBS were incubated with 500 µM 1,4-butanediol diglycidyl ether (BDDE) at 4°C for 3 h in the dark. The excess BDDE was removed by overnight dialysis (14 kD cutoff) to avoid potential crosslinking among RCs. Under this condition, only PBS was crosslinked with RCs. The concentrations of CpcL-PBS and CpcG-PBS from different strains growing under different conditions were normalized to 0.21 µM PC. The concentrations were determined by UV-Vis spectroscopy using an extinction coefficient for PC of ~3 × 10⁵ M⁻¹ cm⁻¹ (Niedzwiedzki et al., 2019).

PSI monomer/PSII and PSI trimer fractions were normalized to the same estimated P₇₀₀ content similar to Smolinski and Lubner et al (Smolinski et al., 2022). Fluorescence emission spectra at 77 K were measured with a Fluorolog-3 (HORIBA Scientific) using 450 nm excitation wavelength (excitation slit of 5 nm) and emission at 730 nm (emission slit of 1 nm), and isolated PSI samples at a chl concentration of 3 μg mL^-1. Chl concentration was based on absorbance at 663 nm by diluting 10 μL sample in 990 μL methanol, A₆₆₃ x (100/82). Maximum fluorescence emission intensity at approximately 718 nm and 685 nm were used to determine the ratio of chl in PSI and PSII. Estimated P₇₀₀ was based on the amount of chl in the sample attributable to PSI. In the diluted PSI trimer samples, P₇₀₀ (nmol) was determined by: (chl concentration of diluted sample ÷ 893.51) ÷ 95. P₇₀₀ in the diluted PSI monomer samples (nmol) was determined by: [(chl concentration of diluted sample x (ratio for PSI: PSII x 95)) ÷ ((ratio for PSI: PSII x 95) + (ratio value for PSII, which is 1 x 35) ÷ 893.51] ÷ 95. The number of chl molecules in PSI monomeric unit: 95. The number of chl molecules in PSII: 35. The molecular weight of chl: 893.51 g mol^-1. Estimated P₇₀₀ in the stock PSI monomer or trimer sample (nmol) was based on the amount of chl, as determined by: [(nmol P₇₀₀+ in diluted sample for fluorescence emission spectra samples) ÷ (3 μg mL^-1 chl)] x (chl concentration of non-diluted PSI sample). In the samples for cross-linking analysis, the final chl concentration of PSI was 0.21 µM (i.e. estimated PSI content or chl_PSI).

PBS and RC fractions were mixed in a molar ratio of PC-to-chl_PSI equaling to 1:1. The mixture was incubated for more than 10 min at room temperature and used for further analysis.

Fluorescence excitation and emission spectra at 77 K were measured with a Fluorolog-3 (HORIBA Scientific). Whole cells (2.5 µg chl ml^-1) or crosslinked PBS-RC samples (at concentrations indicated above) were transferred into a quartz EPR tube with 200 µL volume and slowly frozen in liquid nitrogen. Excitation wavelengths were 450 nm or 580 nm (slit width of 5 nm) to excite chl or PBS, respectively. The emission spectra were recorded between 600 nm and 850 nm or 750 nm (slit width of 1 nm) for whole cells or crosslinked samples, respectively.

ORR1 strains cultured under standard photon flux (GL, 35 µmol photons m^-2 s^-1) conditions exhibited WT-like growth. However, under fluctuating light (FL, 35 µmol photons m^-2 s^-1 for 5 min with 500 µmol photons m^-2 s^-1 for 30 s) the growth rate of the mutant was attenuated, consistent with a mismatch between photon and electron flux (Smolinski et al., 2022). Whole cell 77 K fluorescence emission spectra suggested more energy directed toward PSII in ORR1 under either light regime ( Supplementary Figure S1 ), indicating potential differences in PBS preferences however the reasons for this were not apparent from this data alone. A change in the composition of the PBS forms across the strains was observed by sucrose density gradients ( Supplementary Figure S2 ), which shifted to an increased amount of CpcL-PBS in ORR1 GL and FL ( Table 1 ). This observation is consistent with our hypothesis of CpcL-PBS having a greater role in managing photon and electron flow balance in this strain.

To measure the potential changes of PBS complexes from WT and ORR1 strains, we used sucrose gradients to isolate CpcG- and CpcL-PBS in cells cultured under GL and FL conditions ( Supplementary Figure S2 ). The absorption spectra of CpcL-PBS showed a PC peak at around 620 nm ( Figure 2A ). The 77 K fluorescence emission spectra of CpcL-PBS ( Figure 2B ), excited at 580 nm, displayed the typical CpcL-PBS spectral signatures at 650 nm, which is the blue PC rod, and 670 nm, which is the typical terminal emitter of CpcL-PBS. Aside from intensity changes, the data exhibit no obvious changes in the spectral characteristics of CpcL-PBS in ORR1 cells cultured under either light condition. SDS-PAGE analysis of isolated CpcL-PBS did not provide evidence of substantial compositional changes in ORR1 compared to WT ( Supplementary Figure S3 ).

Having established that there is an enrichment of CpcL-PBS in ORR1, we utilized isolated CpcL-PBS and RC proteins (PSI monomer/PSII and PSI trimer) ( Supplementary Figure S4 ) to assess whether there are changes in the preference for energy transfer between CpcL-PBS and RCs (Wilson et al., 2007). SDS-PAGE analysis of the isolated PSI monomer/PSII and PSI trimer fractions, did not indicate substantial compositional changes between WT and ORR1 ( Supplementary Figure S5 ). Furthermore, our previous in vitro studies of ORR1 PSI showed no evidence of PSI photoinhibition (Smolinski et al., 2022). The ability of CpcL-PBS to bind and transfer energy to RCs was tested by in vitro crosslinking of PBS and RCs from the same strains, using normalized component levels (i.e., chl_PSI and PC) ( Supplementary Figure S6 ). We used 77 K fluorescence spectroscopy with excitation at 580 nm to detect CpcL-PBS excitation energy transfer to RCs as a relative intensity difference (Wilson et al., 2007). Because fluorescence yield can vary depending on the chemical environment of the components (Niedzwiedzki et al., 2019), we conducted all experimentation in the same buffer composition to limit artifactual intensity changes. The relative intensity differences in ORR1 versus WT that are observed in the 77 K fluorescence spectra ( Figure 3 ) are consistent with a change in the CpcL-PBS preference for excitation energy transfer from PSI to PSII in ORR1 ( Figure 3 , Table 2 ). This change is further enhanced in ORR1 cultured under FL conditions.

Our previous study found that the ratio of PSI: PSII remains similar for WT and ORR1 for both light regimes (Smolinski et al., 2022), demonstrating that the relative abundance of PSI and PSII does not change in whole cells. PSI content, however, shifts from primarily trimeric in WT (14% monomeric), to slightly increased monomeric in ORR1 GL (18%), to significantly increased monomeric forms in ORR1 FL (34%). This increase in PSI monomerization observed for ORR1 FL translates to an enrichment of PSI monomers relative to PSII in the PSI monomer/PSII fraction from ORR1 FL. Despite the increase in PSI monomer amount, we still observed a shift toward more energy transfer with PSII, displaying that the preference shift is independent of the total amount of RC present in the sample and rather depends on changes specific to the CpcL-PBS, RCs, and/or their interactions. Further supporting this interpretation is that WT and ORR1 GL, which have similar ratios of PSI monomers to PSII, display opposite behaviors with WT exhibiting a preference for PSI monomers as reported previously (Kondo et al., 2005; Kondo et al., 2007) and ORR1 GL preferring PSII. Furthermore, analysis of CpcG-PBS showed the energy transfer preference for PSII that is observed for WT was maintained in ORR1, under both light regimes ( Supplementary Figure S7 , Supplementary Table S1 ). Thus, the combined effects of variation in light and electron utilization on energy transfer during photosynthesis are isolated to changes in the CpcL-PBS interactions with RCs.

Excitation of crosslinked CpcL-PBS and PSI trimers isolated from ORR1 shows the peak PSI fluorescence emission intensities are similar to WT, but slightly decreased ( Figure 4 , Table 3 ). When compared to PSI monomers ( Figure 3 , Table 2 ), the lower peak intensities for PSI trimers could imply favored energy transfer from CpcL-PBS to PSI monomers than to PSI trimers.

The differences in the observed excitation energy distribution in ORR1 versus WT strains might be due to modifications to the PBS or RC components in ORR1 that change preferences. To test this, we performed experiments where the PBS or RC component of ORR1 was crosslinked with the corresponding component from the WT strain and measured energy transfer as the fluorescence intensity. When WT CpcL-PBS was crosslinked with ORR1 PSI monomer/PSII ( Figure 5 , black dashed line), or ORR1 CpcL-PBS was crosslinked with WT PSI monomer/PSII ( Figure 5 , blue dashed line), there was a slight preference for PSII over PSI monomer ( Table 4 ). The results are consistent with the preference observed for ORR1-only crosslinking experiments where CpcL-PBS preference is for PSII in ORR1 ( Figure 5 , blue solid line) but is reversed from the WT-only crosslinking experiment which shows preference for PSI ( Figure 5 , black solid line). These results demonstrate that determinants from both CpcL-PBS and PSI determine the preference change in ORR1. There was a greater amount of uncoupled CpcL-PBS ( Figure 5 , PBS peak at 650 nm) for cross-strain versus same-strain crosslinked reactions, further supporting that modifications to both components underpin the energy transfer activity.

The preference of CpcL-PBS to transfer energy to PSII over PSI monomers in ORR1 was also consistent when ORR1 CpcL-PBS was crosslinked to ORR1 RCs from the different light regimes (GL and FL) ( Supplementary Figure S8 , Supplementary Table S2 ). The crosslinking analysis of CpcG-PBS with RCs demonstrated that the preference for PSII over PSI was maintained for all strains under all light conditions ( Supplementary Figure S9 , Supplementary Table S3 ).