Samir Chitnavis: Optimizing photosynthetic light-harvesting under stars: simple and general antenna models

Of these, approximately 100 are deemed potentially habitable (Arecibo 2024; Méndez et al. 2021; Meadows and Barnes 2018; McKay 2014), though most of these orbit M-dwarf stars, stars that are significantly cooler and redder than the Sun. These include the planetary systems of TRAPPIST-1 (Gillon et al. 2017; Grimm et al. 2018), Proxima Centauri (Anglada-Escudé et al. 2016; Faria et al. 2022) and LHS 1140 (Dittmann et al. 2017), which are frequently discussed in terms of extra-solar life. M-dwarf planets are generally considered good astrobiological targets for the following reasons. Stellar formation models predict that planets occur more frequently around M-dwarfs than around larger/hotter stars (Mulders et al. 2015; Hsu et al. 2019, 2020). Moreover, the stable phase of their lifespan (the main sequence lifetime) lasts billions of years, meaning that complex biospheres would have time to evolve (Laughlin et al., 1997; Shields et al., 2016). Additionally, potentially habitable planets reside close to the star leading to more robust characterisation of their atmospheres (Shields et al. 2016). However, one downside is that, due to their low temperature (Ts∼2300 K), M-dwarf stars have very limited emissions in the region of the spectrum (400<λ<700 nm) traditionally considered as photosynthetically active radiation (PAR) (shown in Fig. 1A).

A biosphere requires energy input, with the parent star being the most obvious candidate (of course geothermal (Dodd et al. 2017) and geochemical (Ricci and Greening 2023) energy sources may also contribute), meaning it is reasonable to assume that exo-planetary biospheres will be founded on some form of photosynthesis. Oxygenic photosynthesis in particular is the focus of much astrobiological research for several reasons. It may have been a necessary prerequisite for the evolution of multicellular life on Earth (though there is considerable debate (e.g. Wood et al. 2020; Mills and Canfield 2014; Butterfield 2009; Cole et al. 2020; Bozdag et al. 2021), meaning it is a required basis for diverse and complex biospheres. More importantly, it may present the best chance of actually detecting an exo-biosphere in the near future (Kiang 2014), either via detection of O3 in atmospheric transmission spectra (Mendillo et al. 2018; Olson et al. 2018a, b; Lyons et al. 2014; Schwieterman et al. 2018), or by a strong depletion in the range 400–700 nm, in a surface reflectance signal from the planet (O’Malley-James and Kaltenegger 2018; Seager et al. 2005; Arnold et al. 2002; Battistuzzi et al. 2020). The latter is known as a vegetation red edge (VRE) and was first observed for Earth by the Galileo space probe (Sagan et al. 1993), an unmistakable signature of widespread, chlorophyll-rich vegetation. The question of where we should and shouldn’t look for signs of oxygenic photosynthesis is serious one. The James Webb Space Telescope (JWST) and the upcoming (2029) Ariel mission (Tinetti et al. 2020) will study the molecular composition of exo-planet atmospheres. Planned (next 10–20 years) observatories such a the Extremely Large Telescope (Bowens et al. 2021), the Habitable Worlds Observatory (Gaudi et al. 2019), and the Large Interferometer For Exoplanets (Quanz et al. 2022) will measure reflected light from the surface of rocky exo-planets. Of course, the huge resource investment, large measurement time and finite instrument lifetimes mean that candidate planets should be carefully targeted. This requires some method of ranking potential targets in terms of their potential to harbour detectable life based on bulk properties of the planet and parent star. A key part of this involves understanding the relationship between the feasibility and potential characteristics of photosynthesis and the irradiant spectral flux available in the environment.

The feasibility of oxygenic photosynthesis under M-dwarf light was recently and conclusively demonstrated by the La Rocca group in Padua. Battistuzzi et al. (2023a) showed that cyanobacteria could comfortably grow and produce O2 under simulated M-dwarf light, even with a simulated atmosphere which attenuated the irradiance. This was extended to several eukaryotic macroalgae, the breophyte moss Physcomitrium patens, and the model vascular plant Arabidopsis thaliana, though the latter showed signs of shade avoidance syndrome (Battistuzzi et al. 2023b). As important as these results are, this does not mean analogues of these organisms would evolve in such light environments, with vascular organisms in particular seeming poorly suited. Complementary to this approach are theoretical models that characterize some simple relationships between incident light and photosynthetic strategies. Kiang et al. (2007b) derived an empirical rule-set via an exhaustive review of different organisms. They concluded that photoautotrophs will evolve to have their absorption maximum close to the local irradiance maximum, will evolve secondary pigments that absorb at shorter wavelengths, and will have a reaction centre (RC) that operates close to the red edge of the irradiance window. Applied to M-dwarf light, this implies anoxygenic organisms that absorb in the 930<λ<2500 nm (Kiang et al. 2007a). Björn (1976) and later Marosvölgyi and van Gorkom (2010) proposed that photosynthetic structures evolve to maximize light absorption while minimising the metabolic cost of synthesizing pigments, neatly predicting the absorption spectra of both higher plants and purple bacteria. Lehmer et al. (2021) applied this model to a range of stellar irradiances, concluding that M-dwarf light would select for organisms that utilize near-infrared (λ∼1000 nm) light. Arp et al. (2020) argued that photosynthetic antennae evolve to be robust against a noisy energy input, which necessarily selects for two sub-populations of pigments with similar (but not identical) absorption maxima, that absorb on the edge of the irradiance window (e.g. Chl a and b in the case of plants). A modified version of Arp’s principles were applied by Duffy et al. (2023) to various stellar spectra, concluding that M-dwarf stars may preferentially select for anoxygenic photoautotrophs. Lingam et al. (2021) arrived at a similar conclusion based on the pi-electron conjugation length that pigments would need to efficiently absorb light in different stellar fluxes. Finally, Hall et al. (2023) considered how photosynthetic feasibility would intersect with habitability in general, predicting that oxygenic photosynthesis may be restricted to hotter and bluer stars like the Sun. However, a notable feature of these works is that, while they consider the overlap between the light-harvesting antenna and the spectral irradiance, they do not consider the structure, size and overall efficiency.

Oxygenic photosynthesis can and does exist in niches on Earth with very limited PAR, due to the evolution of the antenna-RC architecture (Wolfe et al. 1994; Fleming et al. 2012). While there is considerable diversity in antenna structures, they all function in the same way: a large, modular assembly of pigment-protein complexes captures light and directs the resulting excitation energy into a much smaller central RC (sketched in Fig. 1B). They are generally also adaptable structures, with antenna size changing as organisms acclimate to high or low light, maximizing light input in the latter and mitigating photo-damage in the former (Sanfilippo et al. 2019; Lokstein et al. 2021). Still, there are differences between how plants and cyanobacteria harvest light, which may be relevant to why plants struggle in M-dwarf light and cyanobacteria appear to thrive (Battistuzzi et al. 2023b). Photosystem II of higher plants (PSII, the oxygen-producing photosystem) has a transmembrane antenna composed of structurally similar antenna sub-units binding energetically similar pigments (Su et al. 2017). The supercomplex formed of the PSII RC core (RCII), several minor/monomeric LHCII antenna complexes and several bound trimeric LHCII, sit in a wider, disordered pool of peripheral LHCII (see Fig. 1C). It is this modularity that enables plant PSII to rapidly adapt to changes in the light environment (Vialet-Chabrand et al. 2017; Ruban 2016; Ruban and Wilson 2021). Cyanobacteria and some red algae possess the phycobilisome (PBS) antenna which sits out of the plane of the membrane and has a hierarchical rather than modular structure. A core of allophycocyanin (APC) proteins form a hub from which rods of bluer phycocyanin (PC) and phycoerythrin (PE) radiate (Zheng et al. (2021), see Fig. 1D). How do these two light-harvesting systems fundamentally differ? Is one more likely to evolve under M-dwarf light than the other?

In this work we construct a simple and general model of a RC-antenna light-harvesting system. It is simple in the sense that it considers only the basic thermodynamics of light-harvesting, and it is general in the sense that it presupposes no (or very little) molecular detail. Our aim is to predict what type of antenna structures are best suited to a range of irradiant fluxes from a range of different star types. The null hypothesis is that, so long there is some flux in the PAR region, a sufficiently large antenna will facilitate oxygenic photosynthesis. The alternative hypothesis is that oxygenic photosynthesis under an M-dwarf star will require a more hierarchical (PBS-like) antenna than hotter stars. Harvesting light, that is, absorbing photons over a large area and concentrating that energy into a small RC, is a process that reduces entropy (see Fig . 1 F). This results in steeply diminishing returns in light-harvesting efficiency with increasing antenna size. To overcome this and make light-harvesting thermodynamically favourable (or less unfavourable) antennae adopt, to greater or lesser degrees, an enthalpy funnel structure, with excitations transferred from higher to lower-energy pigments. We argue that the fundamental difference between the cyanobacterial PBS and the plant PSII antenna is how and to what extent they compensate for this entropy penalty. The former has a steeper funnel than the later, along with multiple independent antenna branches. We show that the PBS is inherently much more adaptable to the type of limited PAR light produced by a M-dwarf star.

Read full publication here