Phytoplanktonic cells have an obligate requirement for light to perform photosynthesis. Yet in the marine environment, this energy source is highly variable quantitatively and qualitatively with depth but also along coast-offshore gradients, leaving aside daily oscillations (Kirk, 1994; Holtrop et al., 2021). This variability has triggered an extensive structural and pigment diversification of phytoplankton light-harvesting antennae, which enable cells to considerably enhance the wavelength range they can collect and therefore the number of photons reaching photosystems. Despite their apparent simplicity compared to algae and higher plants, cyanobacteria possess the most elaborated form of antennae known in oxyphototrophs, called phycobilisomes (PBS). PBS are huge water-soluble complexes composed of six to eight rods radiating around a central core. Both core and rods are constituted of phycobiliproteins that bind open-chain tetrapyrroles, called phycobilins (Sidler, 1994). While the PBS core is always made of allophycocyanin and is highly conserved, PBS rods display a very large structural flexibility since they can be made of phycocyanin (PC) only, or of PC and one or two phycoerythrin (PE) types, PE-I and PE-II (Ong and Glazer, 1991; Six et al., 2007). Additionally, each phycobiliprotein can bind up to three different kinds of phycobilins. The ultimate degree of sophistication is the capacity for some cyanobacterial cells to modify the composition of their PBS in response to changes in the ambient light color. This process called “chromatic acclimation” (CA)—the initial term was actually “chromatic adaptation,” but was recently replaced in order to best describe this physiological process (Shukla et al., 2012)—was first observed in the early 20th century in freshwater cyanobacteria shifted from red to green light (Engelmann, 1902; Gaidukov, 1903). CA was later on attributed to changes in the phycobiliprotein composition of PBS rods: in red light, rods are entirely composed of PC and cells look green, whereas in green light PC is restricted to the base of the rods and the distal part is made of PE, causing cells to exhibit a bright red color (Boresch, 1922). This type of complementary chromatic acclimation (CCA, also called type 3 chromatic acclimation or CA3) is one among the seven different types of CA known so far in cyanobacteria (Tandeau de Marsac, 1977; Hirose et al., 2019; Sanfilippo et al., 2019a). While CA2—a simple type of CA where PE production is induced in green light, generating longer PBS rods, and repressed in red light—and CA3 occur mainly in freshwater and brackish cyanobacteria, CA4 is the main type occurring in the open ocean and is specific to marine Synechococcus cyanobacteria (Palenik, 2001; Everroad et al., 2006; Humily et al., 2013). Contrary to CA2 and CA3, CA4 does not involve changes in the phycobiliprotein composition of PBS rods, but in their phycobilin composition. In response to shifts between green light (GL) and blue light (BL), chromatic acclimaters can indeed modify the relative amount of the two phycobilins bound to PE-I and PE-II in order to match the predominant ambient light color. More specifically, they exhibit in BL a high ratio of the BL-absorbing phycourobilin (PUB, λ_max ≈ 495 nm) to the GL-absorbing phycoerythrobilin (PEB, λ_max ≈ 545 nm), and vice-versa in GL (Everroad et al., 2006; Shukla et al., 2012). Variations in PUB and PEB cell content are generally assessed by measuring the relative ratio of whole cell fluorescence excitation at 495 and 545 nm (Exc_495:545) with emission set at 580 nm, which in chromatic acclimaters changes from 0.6–0.7 in GL to 1.6–1.7 in BL (Palenik, 2001; Everroad et al., 2006). In the nomenclature of Synechococcus pigment types (PTs) established by Six et al. (2007) and later modified by Humily et al. (2013), chromatic acclimaters are classified as “PT 3d” cells, meaning that they possess PBS rods made of PC, PE-I and PE-II, a feature shared by all PT 3 representatives, and can modify their Exc_495:545 ratio. In contrast, PTs 3a, 3b, and 3c display a constitutively low, medium and high Exc_495:545 ratio, respectively. For this reason, PT 3a strains are often referred to as “green light specialists” and PT 3c as “blue light specialists” (Grébert et al., 2018, 2022).

Two genetically different types of chromatic acclimaters have been described so far: PTs 3dA and 3dB (Humily et al., 2013). Both possess a small genomic island involved in the CA4 process, yet the CA4-A and CA4-B islands differ genetically and structurally (Humily et al., 2013; Sanfilippo et al., 2019b; Grébert et al., 2021). Although long overlooked, CA4 appears to be an ecologically important process since chromatic acclimaters were shown to account for more than 40% of the whole marine Synechococcus population along the Tara oceans expedition transect (Grébert et al., 2018). Moreover, PTs 3dA and 3dB were found to be equally abundant (22.6% and 18.9%, respectively) but distributed in complementary ecological niches in the field. The former was indeed predominant in cold, nutrient-rich and highly productive waters at high latitude, as well as in other vertically mixed environments, while the latter was mostly found in nitrogen and phosphorus-poor oceanic areas and appeared to be more abundant at depth. The emergence and maintenance of two CA4 types over the course of evolution, as well as the differential distribution of PTs 3dA and 3dB in the environment (Grébert et al., 2018), strongly suggest that they may not be as phenotypically equivalent as previously thought (Humily et al., 2013). To check this hypothesis, we acclimated three representatives of each PT 3dA and 3dB in batch culture under two conditions of temperature, two light irradiances and five light colors in order to compare their growth rates and PBS properties. Furthermore, we performed shifts from BL to GL (and vice-versa) at two irradiances to compare the CA4 kinetics between five strains of each PT. These experiments demonstrated that PT 3dA and 3dB strains actually differ in the blue-to-green light ratio necessary to trigger the CA4 process, and revealed some significant discrepancies in their acclimation pace.

The marine environment was recently shown to shelter five distinct spectral niches, based on the absorption properties of water molecules as well as the variable concentrations of colored dissolved organic matter and non-algal particles (Holtrop et al., 2021). Thanks to their specific antenna complexes binding divinyl derivatives of Chl a and b (Goericke and Repeta, 1992), cells of the tiny cyanobacterium Prochlorococcus appear to be well adapted to the violet niche (401–449 nm), which encompasses the central oceanic gyres. In contrast, Synechococcus BL specialists (PT 3c), i.e., cells possessing a high content in PUB (λ_max ≈ 495 nm), are best suited for the blue niche (449–514 nm) that comprises most other open ocean zones. As concerns Synechococcus GL specialists (PT 3a), i.e., cells possessing a high content in PEB (λ_max ≈ 545 nm), they preferentially thrive in the green niche (514–605 nm) that is essentially found in coastal and upwelling areas (Holtrop et al., 2021). In this context, Synechococcus cells capable of CA4, i.e., to match their Exc_495:545 ratio to the ambient light color in order to optimize photon collection, expectedly colonize both blue and green niches, where they can constitute a large part of the whole Synechococcus population, especially at high latitude (Xia et al., 2017; Grébert et al., 2018). The occurrence of two genetically distinct types of chromatic acclimaters colonizing different habitats in the field (Humily et al., 2013; Grébert et al., 2018, 2021, 2022) however made us wonder whether they displayed phenotypic differences that may partly explain their different spatial distributions.

Here, we looked at the interplay between light quality, light quantity and temperature, and we managed to unveil subtle but significant differences between PTs 3dA and 3dB. Comparisons of cultures acclimated to various light colors ranging from 100% BL to 100% GL at two temperatures and two irradiance levels revealed that, in intermediate blue-green light conditions, PT 3dB strains displayed significantly higher Exc_495:545 ratios than their PT 3dA counterparts (Figure 4). The observed differences are notably due to a more progressive decrease of the Exc_495:545 ratio for PT 3dB than PT 3dA representatives from BL to GL. The latter were indeed more frequently found in the GL-acclimated state with some of them, such as RS9916 in most conditions, even shifting their Exc_495:545 ratio to the BL-acclimated state only in 100% BL (Supplementary Figure S5). In contrast, PT 3dB cells tended to remain longer in the BL-acclimated state, even when the proportion of GL in the incident light was important, an extreme case being A15-62 at 25°C and LL that shifted to the GL-acclimated state only in 100% GL. Based on the phenotypes of knock-out mutants of genes involved in the CA4 process, it has been previously hypothesized that the PT 3dA genotype may have derived from a former GL specialist having acquired the CA4 capacity by integrating a CA4-A island, while the PT 3dB genotype may have been derived from a former BL specialist having integrated a CA4-B island (Sanfilippo et al., 2019b; Grébert et al., 2021). Interestingly, our results are in good agreement with this hypothesis since they suggest that PT 3dA strains need a large proportion of blue photons to induce the CA4-A response, while on the contrary PT 3dB cells require a larger proportion of green photons to induce the CA4-B response. In other words, the two PTs seemingly differ in the blue-to-green ratio necessary to trigger the CA4 process, even though there is some strain-to-strain variability.

The molecular basis of the difference in Exc_495:545 ratio between chromatic acclimaters fully acclimated to either 100% BL or 100% GL has been well documented, and was found to be the same in the PT 3dA model strain RS9916 (Shukla et al., 2012; Sanfilippo et al., 2016, 2019b) and the PT 3dB model strain A15-62 (Grébert et al., 2021). Both CA4-A and -B processes indeed consist in an exchange of one out of the five phycobilins bound to the α-PE-I subunit (Cys-139) and two out of the six phycobilins bound to the α-PE-II subunit (Cys-83 and Cys-140), with PUB molecules being bound to these three positions in BL, and PEB in GL. However, the observation of intermediate Exc_495:545 ratios in blue-green light mixes (Figure 4; see also Sanfilippo et al., 2019b) remains difficult to interpret as it may translate different, but not mutually exclusive, sources of variability. More precisely, this observation may be explained by: (i) heterogeneous populations of Synechococcus cells with PBS either fully acclimated to BL (PUB-rich) or to GL (PEB-rich); (ii) homogeneous populations of Synechococcus cells all having different phycobilins at the three swing sites in a given light color; (iii) individual cells containing PBS with different chromophorylation states; and/or (iv) heterogeneity in phycobiliprotein chromophorylation within single PBS, i.e., PBS having rods with different chromophorylation states. Unfortunately, there is currently no simple experimental way to demonstrate which of these hypotheses is most likely.

In contrast to Exc_495:545 ratios, light quality had no significant effect on growth rates, while the latter varied with PT, temperature and light quantity. In HL, the growth rates of all six chromatic acclimaters used for acclimation experiments were significantly higher at 25°C than 18°C, the former being an optimal growth temperature for most marine Synechococcus strains tested so far (Figure 1; Mackey et al., 2013; Pittera et al., 2014; Breton et al., 2020; Doré et al., 2022; Ferrieux et al., 2022). The impact of increasing temperature on growth rate was however much less important at LL, confirming that temperature and irradiance have a synergistic effect on growth, as previously observed in the model strain WH7803 (Guyet et al., 2020). Of note, whatever the temperature, all cells were also able of typical photoacclimation, i.e., to adjust the surface of their photosynthetic membranes to reduce the incoming photon flux (Kana and Glibert, 1987; Moore et al., 1995; Six et al., 2004), as shown by the decrease of both Chl a and PE fluorescence signals between LL and HL (Figures 2, 3). Interestingly, both parameters also slightly decreased in all representatives in response to progressive changes in light quality from 100% BL to 100% GL (Figures 2, 3; Supplementary Figures S3, S4), suggesting that chromatic acclimaters used GL more efficiently than BL, at least in our experimental setup. Consequently, all strains needed to slightly adjust their thylakoid surface (and thus the number of both photosystems and PBS) in BL in order to maximize the collection of available photons and maintain similar growth rate (Supplementary Figure S2).

Another parameter that differentiated PTs 3dA from 3dB in the present study is the Em_560:650 fluorescence emission ratio, which is often interpreted as a proxy of the PE:PC ratio. This ratio exhibited higher values and tended to decrease from 100% BL to 100% GL in PT 3dA strains, while it was more stable in PT 3dB representatives (Figure 5). Since PE fluorescence per cell decreased in PT 3dA from BL to GL (Figure 3) and the PC:TA ratio was somewhat constant (Supplementary Figure S7), this may indicate a lower within-rod photon energy transfer efficiency in PT 3dA cells in GL compared to BL. Another possibility is that PBS rod length decreased by disconnection of the distal PUB-rich PE-II hexamers, as previously observed in Synechococcus sp. WH8102 as a result of photoacclimation (Six et al., 2004). In PT 3dB representatives, the PE:PC ratio stability, associated with decreasing PE fluorescence, rather suggests that other kinds of structural changes occurred, such as a reduction of the PBS number per cell or of the thylakoidal surface area.

Shift experiments in both LL and HL conditions from 100% BL to 100% GL, and vice-versa, confirmed previous observations by Humily et al. (2013) that PT 3dA cells generally exhibit a more variable acclimation kinetics than PT 3dB cells. Two PT 3dA strains belonging to the CRD1 clade (BIOS-U3-1 and MITS9220) were indeed stuck at an Exc_495:545 ratio of around 1.2 in HBL (corresponding to “phenotypic group 2” in Humily et al., 2013). Moreover, MITS9220 and another PT 3dA strain (BL107) showed a delay in the initiation of CA4 from LBL to LGL, but not in the reverse condition, as previously reported by these authors for BL107, a behavior they designed as “phenotypic group 3.” Some PT 3dA representatives common to both studies (RS9916 and WH8020) exhibited different behaviors. Both strains indeed reached a significantly lower Exc_495:545 ratio in HBL than LBL in the previous study but not in the present one, strengthening the idea that PT 3dA cells display a more variable acclimation phenotype than PT 3dB cells, at least in some conditions. In contrast, most PT 3dB representatives exhibited a “typical” CA4 dynamics in both studies, designed by Humily et al. (2013) as “phenotypic group 1.” The only exceptions are strains such as WH8103 that have a typical CA4-B region but, for a yet unknown reason, have completely lost their CA4 ability and thus display a fixed Exc_495:545 ratio. Interestingly in this context, the WH8109 strain that was reported to display this fixed phenotype in Humily et al. (2013) was found to have recovered its CA4 ability in a more recent study (Lovindeer et al., 2021). The wider phenotypic variability observed in PT 3dA representatives might be attributed to the erratic genomic localization of the CA4-A island, which can be found virtually anywhere in the genome of PT 3dA strains, including the 5′-end of the PBS region, while the CA4-B island is systematically located in the middle of the PBS rod region in PT 3dB cells (Humily et al., 2013; Grébert et al., 2022).

The most striking outcome of our shift experiments was certainly the discrepancy in acclimation paces between the two types of chromatic acclimaters in three out of the four tested conditions. Indeed, after the LGL to LBL and HGL to HBL shifts, the PT 3dA representatives acclimated faster than their PT 3dB counterparts, while the opposite pattern was observed after the HBL to HGL shift (Figure 7; Supplementary Figures S8, S9). Altogether, our results showed that while PT 3dA and PT 3dB cells preferentially remain in their respective basal state (i.e., low and high Exc_495:545 ratio, respectively) when grown in a blue-green light mix (Figure 4), they can reach the acclimation state opposite to their basal state faster than the other PT once the CA4 process has been triggered (Figure 7; Supplementary Figures S8, S9). These distinct acclimation paces might be due to differences in the organization and gene content of the CA4-A and CA4-B islands. While both share two genes in tandem encoding the regulatory proteins FciA and FciB (Sanfilippo et al., 2016), as well as a gene encoding a protein of unknown function (Unk10), they also contain specific genes. A third putative regulatory gene, fciC, is indeed only found in PT 3dA representatives, possibly explaining why the two CA4 processes are opposite, CA4-A being activated in BL and CA4-B in GL (Shukla et al., 2012; Humily et al., 2013; Sanfilippo et al 2019b; Grébert et al., 2021). Additionally, the CA4-A and CA4-B islands encode distinct enzymes, the PEB lyase-isomerase MpeZ and the PEB lyase MpeW, respectively. Both compete with another enzyme, encoded in the main PBS genomic region, for binding either a PUB in BL or a PEB in GL at Cys-83 of the α-PE-II subunit (Sanfilippo et al., 2019b; Grébert et al., 2021). Functional studies are needed to determine whether these mechanistic discrepancies are responsible for the phenotypic differences between PT 3dA and 3dB strains reported in the present study.

In conclusion, PTs 3dA and 3dB cells exhibit subtle but significant phenotypic differences that may explain why, in a spectral niche encompassing both blue and green light, they can coexist not only with BL and/or GL specialists, but also with their CA4-able counterpart (Huisman et al., 2002; Stomp et al., 2004; Grébert et al., 2018; Luimstra et al., 2020; Holtrop et al., 2021). Future studies should allow one to confirm this hypothesis, either using co-cultures of different PTs grown in various light colors, or by correlation analyses of the variations of the relative abundance of the different PTs with changes in the underwater light field.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

LD: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Writing – original draft, Writing – review & editing, Software, Visualization. LG: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Writing – review & editing, Funding acquisition, Project administration, Supervision. BG: Investigation, Methodology, Writing – review & editing. JC: Investigation, Methodology, Writing – review & editing. MR: Investigation, Methodology, Resources, Writing – review & editing. FP: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Supervision, Validation, Writing – review & editing, Funding acquisition, Project administration, Visualization.