Light-induced changes of far-red excited chlorophyll fluorescence: further evidence for variable fluorescence of photosystem I in vivo

Recently, the long-standing paradigm of variable chlorophyll (Chl) fluorescence (Fv) in vivo originating exclusively from PSII was challenged, based on measurements with green algae and cyanobacteria (Schreiber and Klughammer 2021, PRES 149, 213-231). Fv(I) was identified by comparing light-induced changes of Fv > 700 nm and Fv < 710 nm. The Fv(I) induced by strong light was about 1.5 × larger in Fv > 700 nm compared to Fv < 710 nm. In the present communication, concentrating on the model green alga Chlorella vulgaris, this work is extended by comparing the light-induced changes of long-wavelength fluorescence (> 765 nm) that is excited by either far-red light (720 nm, mostly absorbed in PSI) or visible light (540 nm, absorbed by PSI and PSII). Polyphasic rise curves of Fv induced by saturating 540 nm light are measured, which after normalization of the initial O-I1 rises, assumed to reflect Fv(II), display a 2 × higher I2-P transient with 720 nm excitation (720ex) compared with 540ex. Analysis of the Fo(I) contributions to Fo(720ex) and Fo(540ex) reveals that also Fo(I)720ex is 2 × higher than Fo(I)540ex, which supports the notion that the whole I2-P transient is due to Fv(I). The twofold increase of the excitation ratio of F(I)/F(II) from 680 to 720 nm is much smaller than the eight–tenfold increase of PSI/PSII known from action spectra. It is suggested that the measured F > 765 nm is not representative for the bulk chlorophyll of PSI, but rather reflects a small fraction of far-red absorbing chlorophyll forms (“red Chls”) with particular properties. Based on the same approach (comparison of polyphasic rise curves measured with 720ex and 540ex), the existence of Fv(I) is confirmed in a variety of other photosynthetic organisms (cyanobacteria, moss, fern, higher plant leaves). Supplementary Information The online version contains supplementary material available at 10.1007/s11120-022-00994-9.


S2
In panel a the data between 660 and 715nm of Schreiber and Vidaver (1974) are replotted. The PS II value for 715 nm was estimated by extrapolation. Panel b shows the derived ratio spectrum of PS I/PS II action.
Above 680nm the ratio of PSI/PSII action increases in two approximately equal steps, with the 1. step ending at about 695nm, i.e. where PSI action is still high (92%), whereas PSII action has already dropped to 20%. During the 2. step both the PSI and the PSII actions decline, but PSII action more steeply than PSI action.

(2) Estimation of the effective background signal and of Fo in measurements of 720nm excited fluorescence under the conditions of the experiment of Fig.4 described in the main text
Measurements of 720nm excited fluorescence changes were carried out in the optical geometry described in Figs. 1-3 (main text). Due to the weak signals of 720nm excited chlorophyll fluorescence >765nm (720ex), high ML intensity and gain settings had to be applied with 720ex to obtain similar signal amplitudes as with 540ex, for which low ML intensity and gain settings were used. Therefore, the unavoidable background signal (i.e. the signal not consisting of chlorophyll fluorescence) was much larger with 720ex than with 540ex. While in the case of 540ex the background signal was negligibly low, it contributed significantly to the overall signal with 720ex, thus complicating quantitative assessment of Fo(720ex The background signal with 720ex was composed of optical and non-optical components. The latter consisted of an electrical "pick-up" signal that could be readily determined by shielding the detector (black cardboard in front of photodiode), amounting to 145mV. The optical background signal was heterogeneous, one part being due to 720nm ML reflected from the cuvette walls and another part due to 720nm ML scattered by the Chlorella cells towards the detector. The reflectance part was revealed by a "blank" measurement (cuvette filled with suspension medium, without Chlorella): The overall blank signal amounted to 270mV, being composed of the 145mV electrical pick-up signal and the reflectance signal. For estimation of the fluorescence signal caused by scattered 720nm ML, first the amount of freshly precipitated BaSO 4 was determined that was required to obtain the same scattering signal as with the Chlorella suspension (dotted green line in figure S3). For this purpose, the RG780 filter in front of the detector was replaced by a set of neutral density filters attenuating the signal by a factor of 800. 5µl aliquots of a 50mM BaCl 2 solution were added to the cuvette filled with 1300µl BG11 suspension medium enriched with 2mM SO 4 2-. 25µl of the BaCl 2 solution gave the same scattering signal as the Chlorella suspension.

S3
Titration of 720nm scattering signal induced by stepwise precipitation of BaSO 4 in the suspension medium. The broken green line indicates the scattering signal measured under the same conditions with a suspension of Chlorella, as used in the experiment of Fig.4 (main text). Measurement carried out in the optical geometry described in Fig.1 (main text), with the RG780 filter in front of the detector being replaced by a set of neutral density filters attenuating the signal by a factor of 800. 5µl aliquots of a 50mM BaCl 2 solution were added to the cuvette filled with 1300µl BG11 suspension medium enriched with 2mM SO 4 2-. After addition of 25µl of the BaCl 2 solution a similar signal was reached as with the Chlorella suspension.
The supplementary figure S4 shows the increase of the background fluorescence signal upon addition of 25µl BaCl 2 solution, as measured in presence of the RG780 filter at the same sensitivity as used in the experiment of Fig. 4 (main text) with the Chlorella suspension.

+ 25µl 50mM BaCl 2 S4
Increase of apparent fluorescence signal of cuvette filled with 1300µl BG11 suspension medium enriched with 2mM SO 4 2upon addition of 25µl of a 50mM BaCl 2 solution. 720nm ML switched on at 10s. Optical geometry as described in Fig. 1 (main text), with 3mm RG780 filter in front of detector. The broken red line indicates the contribution of an electrical offset to the overall signal. The signal measured before addition of BaCl 2 is composed of the electrical offset and reflectance signals. The overall background signal measured after addition of 25µl BaCl 2 solution amounts to 430mV (green dotted line).
It appears reasonable to assume that a similar overall background signal as obtained in presence of freshly precipitated BaSO 4 in the experiment of figure S4 (i.e. approximately 430mV) does also apply to the Chlorella data in Fig.4 (main text). While the non-optical components per se are identical, the scattering components were made equal via the BaSO 4 titration and the reflectance components may be assumed to be close to equal, as the color of the Chlorella cells should not affect the reflectance of wavelengths >765nm.

The thus estimated background signal can be subtracted from the 720ex response measured with
Chlorella so that a plausible estimate for the dark fluorescence level, Fo(720ex) = 1000mV can be derived, as shown in figure S5.

S5
Polyphasic fluorescence rise upon onset of strong actinic illumination measured with darkadapted Chlorella using 720nm pulse-modulated excitation (as also shown in Fig.4, main text) under the same conditions as the measurement of the background signal in figure S4. The green dotted line indicates the corrected baseline, accounting for the background signal estimated in the experiment of S4. The corrected Fo amplitude is shown. This Fo, which amounts to 1000mV, is composed of contributions of Fo(I)720ex and Fo(II)720ex. Fig. 4

(main text)
The initial fluorescence yield, Fo, in green C3 photosynthetic organisms may be assumed to be composed of about 35% Fo(I) and 65% Fo(II), when fluorescence is excited with visible light and measured at wavelengths >700nm. In figure S6 below, this information is applied to the 540ex data of Fig.4 (main text). At the given ordinate scaling, the resulting Fo(I) and Fo(II) values with 540ex amount to 260mV and 480mV, respectively.
As outlined above (figures S3-S5), in the case of 720ex, quantitative determination of Fo is complicated by an unavoidable, relatively large background signal, for which, however, a plausible estimate could be derived under the conditions of the measurements in Fig.4 (main text). After correction for this background signal, the resulting Fo(720ex) can be readily deconvoluted into its Fo(I) and Fo(II) components. For this purpose, it may be assumed that because of O-I 1 equalization of the 720ex and 540ex responses, Fo(II)720ex is equal to Fo(II)540ex, based on the rationale that O-I 1 is a specific PSII response and that when the O-I 1 amplitudes are equal also the amplitudes of all other PSII responses, including Fo, should be equal. The resulting deconvolution of Fo(720ex) is shown in figure S7.

S7
Polyphasic fluorescence rise upon onset of strong actinic illumination measured with darkadapted Chlorella using 720nm pulse-modulated excitation, as also shown in figure S5, with deconvolution of the corrected Fo into the contributions of Fo(I) and Fo(II), based on the assumption that Fo(II)720ex = Fo(II)540ex = 480mV (see figure S6).
With overall Fo(720ex) amounting to 1000mV and Fo(II)720ex = Fo(II)540ex = 480mV, it follows that Fo(I)720ex = 520mV, which happens to be twice the amplitude of Fo(I) 540ex = 260mV in figure S6. As due to the O-I 1 equalization F(II) excitation is equal, this means that with 720ex two times more F(I) is excited compared with 540ex, i.e. F(I)720ex/F(I)540ex = 2, which holds for both Fo(I) and Fv(I). It should be noted that the numerical value of 2 for this excitation ratio relies on the tentative assumption that Fo(540ex)>765nm in Chlorella under the given conditions contains 35% Fo(I). While the exact excitation ratio is not known, in the following section the possible influence of variations of this ratio on deconvolution of Fv(I) and Fv(II) is investigated.

(4) Considering variations in the F(I)/F(II) excitation ratio with 540ex
When O-I 1 -equalized responses with 720ex and 540ex are analyzed, as in Figs. 6-7 (main text), the excitation ratio F(I)720ex/F(I)540ex corresponds to the factor by which the difference signal Fv(720ex) -Fv(540ex), i.e. the "extra Fv(I)", has to be multiplied in order to obtain Fv(I)720ex. This factor depends on the assumed F(I)/F(II) excitation ratio with 540ex, which determines the Fo(I)540ex/Fo(540ex). In figure S8 a plot of F(I)720ex/F(I)540ex versus the assumed Fo(I) contribution to the total Fo(540ex) is presented for the same original data as in figures S5-S7 (and