Chlorophyllide a oxygenase (CAO) gene duplication across the Viridiplantae

This project was supported by Natural Sciences and Engineering Research 22 Council of Canada Discovery Grants (NSERC DG) awarded to M.C. The authors are grateful for the 23 support from the Canada Foundation for Innovation (CFI) and University of Ottawa start-up funding. 24 M.P. was supported by Ontario Graduate Scholarship (OGS), NSERC Graduate Scholarship, and 25 Polar Knowledge Canada Antarctic Doctoral Scholarship. 26 Author’s contributions: M. Poirier and M. Cvetkovska conceptualized the work and designed the 27 experiments. Material preparation, data collection and analysis were performed by M. Poirier, R. 28 Wright and M. Cvetkovska. The first draft of the manuscript was written by M. Poirier and all authors 29 commented on all versions of the manuscript. All authors read and approved the final manuscript. 30 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 2

31 Viridiplantae, a diverse group of green plants and alga that have evolved from a common ancestor, 32 are unified in their ability to produce and use two types of chlorophyll (chlorophyll a and chlorophyll 33 b) to capture light energy. In addition to playing a role in light harvesting, chlorophyll b is required at 34 the appropriate level for the accumulation, assembly, and stability of light harvesting complexes 35 within the photosynthetic apparatus. Chlorophyll b is synthesized from chlorophyll a by the enzyme 36 chlorophyllide a oxygenase (CAO), a Rieske-type mononuclear non-heme iron oxygenase. A 37 regulatory degron sequence, described in detail only in land plants, regulates the stability of CAO 38 proteins based on the availability of chlorophyll b. Recent identification of CAO gene duplication in 39 bryophyte and green algal species, combined with expanded availability of sequenced genomes 40 within the Viridiplantae, prompted further investigation into the role of gene duplication in the 41 evolution of chlorophyll b biosynthesis. Examination of genomes from 246 plant and algae species 42 revealed independently occurring CAO duplications throughout the Viridiplantae, with a higher 43 prevalence of duplication in land plants compared to their algal relatives. Additionally, we 44 demonstrate that the degron sequence is poorly conserved in chlorophytes, but first appears as a 45 conserved sequence in charophytes, and is very highly conserved among the embryophytes. The 46 evolutionary history and functional role of CAO throughout the Viridiplantae lineage is discussed 47 based on these key observations, adding to our understanding of chlorophyll b biosynthesis and 48 the role of CAO in photosynthetic species. 49 50

51 Chlorophyll biosynthesis, chlorophyllide a oxygenase, gene duplication, degron, plants and algae 52 53 54 55 56 57 58 59 60 61 62 63 64 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 3

65 Photosynthetic organisms form the base of most food chains and provide the organic carbon 66 compounds that support life in our biosphere. The capture of light energy by chlorophyll is the first 67 step in the process of photosynthesis that fixes atmospheric CO2 into stable organic compounds. 68 Overwhelming evidence indicates that oxygenic eukaryotic photosynthesis that occurs in the 69 chloroplasts of modern-day plants and algae traces its origins from an endosymbiotic event with 70 cyanobacteria-like prokaryotes (Blankenship 2010; Cardona 2019; Sánchez-Baracaldo and 71 Cardona 2020). Viridiplantae (or green plants) have primary chloroplasts derived from an ancient 72 endosymbiosis, and encompass two major clades: the chlorophytes and the streptophytes. 73 Chlorophytes are a monophyletic group of marine, freshwater, and terrestrial green algae. With 74 ~8,000 described species (Guiry 2024), this group encompasses a large diversity of adaptations, 75 morphologies, and life histories (Leliaert et al. 2012). Streptophyta includes the charophytes, 76 ~5,500 species of largely freshwater green algae (Guiry 2024) and the embryophytes, or land 77 plants. Land plants are thought to have diverged from a single clade within the streptophytes (de 78 Vries and Archibald 2018) and number ~430,000 species (Ruggiero et al. 2015). One characteristic 79 that unifies members of the Viridiplantae is their ability to synthesize and use two types of 80 chlorophylls: chlorophyll a (Chl a) and chlorophyll b (Chl b). 81 The photosynthetic apparatus in the Viridiplatae is derived from the ancestral cyanobacterial 82 endosymbiont, which used chlorophyll as the main pigment in their photosystems (Xiong and 83 Bauer 2002; Sánchez-Baracaldo and Cardona 2020). Chl a is ubiquitously present in all oxygenic 84 photosynthetic eukaryotes, and functions in both energy capture in the antenna light harvesting 85 complexes (LHCs) and in driving electron transfer in the photosystem II (PSII) and photosystem I 86 (PSI) reaction centers. Chl b is specific to the Viridiplantae although it has also been detected in 87 prochlorophytes and Acaryochloris, unique photosynthetic prokaryotic groups that lack typical 88 cyanobacterial phycobilin light-harvesting pigments (Palenik and Haselkorn 1992; Roche et al. 89 1996; Tomitani et al. 1999; Partensky et al. 2018). Chl b resides predominantly in the peripheral 90 LHCs (Neilson and Durnford 2010) although it has been detected in the core complexes of some 91 deep water marine chlorophytes (Kunugi et al. 2016). 92 Strict Chl a/b stoichiometry is required for optimal energy transfer during photosynthesis. 93 Maintaining the correct Chl a/b ratios is a dynamic process and indicative of adaptation to different 94 light conditions (Tanaka and Tanaka 2007). In addition to playing a role in light harvesting, Chl b is 95 required at appropriate levels for the accumulation, assembly, and stability of LHCs in algae 96 (Bujaldon et al. 2017), bryophytes (Zhang et al. 2023), and angiosperms (Król et al. 1995; Reinbothe 97 et al. 2006; Kim et al. 2009; Nick et al. 2013). Under low light conditions, the amount of Chl b 98 relative to Chl a increases, leading to larger LHC antenna size, which maximizes the surface area 99 for light absorption (Kunugi et al. 2016; Kume et al. 2018; Ueno et al. 2019). Thus, the amount of 100 Chl b is a major regulator of light-harvesting capacity. 101 Chlorophyll biosynthesis and turnover occurs through a complex multistep pathway (reviewed in 102 Brzezowski et al. 2015; Willows 2020). Chl b is synthesized from Chl a, via the intermediate 7-103 hydromethyl chlorophyll a, by the action of a single enzyme: chlorophyllide a oxygenase (CAO) 104 (Tanaka et al. 1998; Espineda et al. 1999; Oster et al. 2000; Mueller et al. 2012). Chl b can be 105 converted back into Chl a by chlorophyll b reductase (CBR) and 7-hydroxymethyl reductase (HCAR) 106 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 4 to complete the cycle (Figure 1; reviewed in detail in Tanaka and Tanaka, 2019). The chlorophyll 107 cycle plays an essential role in many biological processes, including biogenesis of LHCs, antenna 108 size regulation during light acclimation, and chlorophyll degradation during senescence. While 109 many of the intricacies of how this pathway is regulated are still not fully understood (Tanaka and 110 Tanaka 2019), it has been suggested that CAO levels and activity are key determinants of Chl b 111 accumulation. 112 Figure 1. The chlorophyll cycle in land plants. Chlorophyll a is converted to chlorophyll b via the 113 intermediate 7-hydromethyl chlorophyll a, by the action of the enzyme chlorophyllide a oxygenase 114 (CAO). In a negative feedback loop, high levels of chlorophyll b trigger the binding of a stromal Clp 115 protease to the degron motif (QDLLTIMILH) and the subsequent destabilization of CAO (dashed red 116 line). Chlorophyll b is necessary for the stabilization of Light Harvesting Complexes (LHC) in plants 117 (dashed blue line). Excessive amounts of LHC or accumulation of energetically uncoupled non-118 functional LHCs (e.g., during senescence), triggers the accumulation of Chlorophyll b reductase 119 (CBR; blue line) and the conversion of chlorophyll b to chlorophyll a via the action of CBR and 7-120 hydroxymethyl reductase (HCAR). This regulatory mechanism of CAO and CBR enables the fine- 121 tuning and optimization of chlorophyll b and LHC levels (adapted from Tanaka and Tanaka 2019). 122 CAO is part of the Rieske-type mononuclear non-heme iron oxygenase group of proteins (Gray et al. 123 2004). It catalyzes the conversion of the methyl group of Chl a into a formyl group to form Chl b 124 through two oxygenation reactions, likely using ferredoxin as a reductant (Oster et al. 2000). CAO 125 has three main regions: an N-terminal regulatory domain and a C-terminal catalytic domain, which 126 are connected by a more variable linker region (Nagata et al. 2004). The catalytic domain contains a 127 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 5 Rieske cluster and a mononuclear iron-binding domain, both of which are necessary for the 128 enzymatic activity of CAO (Yamasato et al. 2005; Kunugi et al. 2013). The regulatory domain has 129 been characterized in detail only in land plants. In Arabidopsis, a conserved sequence 130 (QDLLTIMILH) termed a degron was demonstrated to regulate Chl b accumulation by affecting CAO 131 protein turnover. In a negative feedback loop, high levels of Chl b trigger the binding of a stromal 132 Clp protease to the degron motif leading to destabilization of the CAO protein; this, in turn, halts 133 Chl b biosynthesis (Sakuraba et al., 2009; Nakagawara et al., 2007). 134 The evolutionary origins of CAO can be traced back to prochlorophytes, which encode a CAO gene 135 homolog (Tomitani et al. 1999; Nagata et al. 2004). Notably, the prokaryotic CAO-like enzymes lack 136 the regulatory N-terminal domain, suggesting that the Chl b-dependent mechanism for controlling 137 CAO stability was acquired in eukaryotes. Recent work on 9 chlorophyte, 2 charophyte, and 3 138 embryophyte species suggested that the degron recognition sequence first appeared in 139 charophytes as an adaptation to high-light environments in shallow water (Kunugi et al. 2016). 140 These conclusions, however, were based on a very limited sample of representative species and 141 the presence of a degron in the CAO sequence in charophytes was inferred based on a single 142 species (Klebsormidinium flaccidum). This hypothesis requires further validation. 143 The CAO protein is encoded by a single gene in most plants and algae (Kunugi et al. 2016; 144 Schumacher et al. 2022), with only a very few reported exceptions. Two CAO genes were detected 145 in rice (Lee et al. 2005) although only one of them is necessary for Chl b synthesis (Jung et al. 2021). 146 Recent work revealed CAO duplication in the bryophyte Physcomitrium patens, where both genes 147 contribute to Chl b accumulation (Zhang et al. 2023). CAO duplication was also reported in two 148 Antarctic Chlamydomonas species , where the CAO paralogs in each species appear to be derived 149 from independent duplications events rather than an ancestral one (Cvetkovska et al. 2019; Poirier 150 et al. 2025). 151 These insights prompted us to further examine the CAO gene content and the presence of the 152 regulatory domain across the diverse plant and algal groups. A systematic examination for the 153 presence of CAO across the Viridiplantae has not yet been preformed. In this work we took 154 advantage of the recent increase in publicly available plant and algal genomes to examine the 155 occurrence of CAO genes in diverse species across Viridiplantae and demonstrate a widespread 156 occurrence of CAO duplicates in this group. Furthermore, building on the work by Kunugi et al. 157 (2016), we have examined the occurrence of a degron sequence in 326 unique CAO sequences in 158 algae and plants. We demonstrate that this regulatory region is poorly conserved in chlorophytes, 159 compared to charophyte and embryophyte CAO sequences, suggesting a different mechanism for 160 Chl b regulation in early aquatic lineages. 161

162 CAO sequence identification and analyses 163 To identify CAO sequences and duplication events with high accuracy, we used publicly available, 164 high quality, and fully annotated genomes. Chlorophyte and streptophyte genomes were retrieved 165 from Phycocosm (Grigoriev et al. 2021), while embryophyte genomes were obtained from 166 Phytozome v13 (Goodstein et al. 2012). Putative CAO genes were identified with a tBLASTn search 167 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 6 with the Chlamydomonas reinhardtii (CrCAO; Cre01.g043350) and the Arabidopsis thaliana 168 (AtCAO; AT1G44446) full-length CAO peptides (Merchant et al. 2007; Lamesch et al. 2012). A strict 169 e-value cutoff of ≤e-30 was taken as a threshold, to avoid many unspecific hits due to the 170 conserved nature of the Fe-binding and Rieske cluster domains present among diverse plant and 171 algal proteins (Schmidt and Shaw 2001; Ferraro et al. 2005; Przybyla-Toscano et al. 2021). Putative 172 CAO sequences were manually curated and truncated or low-quality sequences were excluded 173 from downstream analyses. Curated CAO sequences were aligned using MUSCLE (Edgar 2004). 174 The degron sequence (QDLLTUMILH) was identified by multiple sequence alignment of the full-175 length CAO amino acid sequences and the regulatory N-terminal domain. In accordance with 176 Sakuraba et al. 2009, we identify the degron as “highly conserved” if 8 out of the 10 amino acids are 177 perfectly conserved, “moderately conserved” if 4 to 8 amino acids are conserved, and “poorly 178 conserved” if less than 4 amino acids are conserved. Sequence logos for each major group 179 (chlorophyte algae, charophyte algae, non-vascular plants and vascular plants) were generated by 180 Geneious Prime 2024.11 (Dotmatics), where the size of the letter reflects its frequency. To examine 181 the expression of CAO duplicate genes, we screened previously published transcriptomes from 16 182 species with representatives from the Chlorophytes (Arriola et al. 2018), Charophytes (Cheng et al. 183 2019), non-vascular (Perroud et al. 2018; Healey et al. 2023) and vascular plants (Zhu et al. 2019; 184 Wang et al. 2021; Liu et al. 2022b; Shang et al. 2023; Guo et al. 2025; He et al. 2025; Mascuñano et 185 al. 2025; Roy et al. 2025). Only transcriptomes that report FPKM, RPKM and TPM values obtained 186 from organisms cultivated at optimal conditions were considered (Supplementary Table S4). 187 Phylogenetic Inference 188 The analyses described above resulted in 374 individual CAO sequences (Supplementary Table 189 S1). To avoid inaccurate inference and over-representation, we manually removed heterodimeric 190 CAO proteins, where the Rieske domain and the Fe-binding domain are located on different genes 191 (Kunugi et al. 2013) (Supplementary Table S1). We also removed identical CAO sequences isolated 192 from algal cultures identified as the same species but belonging to different strains or culture 193 collections (Scenedesmus obliquus, Auxenochlorella protothecoides, Chlorella sorokiniana, 194 Ostreococcus tauri, Mesostigma viride, Zygnema circumcarinatum). This resulted in a 280 amino 195 acid alignment of 326 unique CAO sequences across 246 species. 196 These manually curated CAO genes were translated into amino acid sequences, aligned with 197 MUSCLE and trimmed to remove gaps and ambiguously aligned regions with Gblock (Castresana 198 2000; Talavera and Castresana 2007). Maximum likelihood trees were inferred in the CIPRES 199 Science Gateway (Miller et al. 2015) using RAxML v8.0 (Stamatakis 2014) with 1000 bootstraps with 200 the Whelan Goldman matrix for globular proteins (WAG), a gamma shape parameter, and empirical 201 estimation of invariable sites. The trees were rooted using prokaryotic CAO from Acaryochloris 202 thomasi (WP_110987895.1) and Prochlorotrix hollandica (WP_017713323.1) as an outgroup. 203 Tandem duplications were defined as those located on the same chromosome or contig within <10 204 genes of each other, based on the genomic coordinates retrieved from Phytozome v13 or 205 Phycocosm. To identify ancestral and lineage-specific duplications, we used Notung v2.9 (Stolzer 206 et al. 2012) and reconciled the CAO trees with a reference species tree derived from NCBI 207 taxonomy (Schoch et al. 2020) via the PhyloT v2 tool (https://phylot.biobyte.de). Putative ancestral 208 duplications were only reported when occurring on branches with bootstrap support ≥85%. All 209 trees were visualized in iTOL v6.8.2 (Letunic and Bork 2021). Previously inferred taxonomic 210 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 7 information for chlorophytes and streptophytes was confirmed via AlgaeBase, University of Galway 211 (www.algaebase.org), and for land plants via the Plants Of the World Online, Royal Botanic 212 Gardens, Kew (https://powo.science.kew.org). 213

and Discussion 214 Occurrences of multiple CAO genes are widespread among the Viridiplantae 215 Our analysis demonstrates a widespread occurrence of CAO gene duplication across the 216 Viridiplantae. Multiple gene copies are more prevalent among the angiosperms, where ~41% of 217 examined species encode for more than one CAO gene. In comparison, we detected two CAO gene 218 copies in only ~11% of chlorophyte and ~20% of charophyte species (Table 1; Table 2, 219 Supplementary Table S1; Supplementary Table S3). All 23 species where we detect more than two 220 CAO genes are members of the Embryophyta. The modern-day cultivar of sugarcane (Saccharum 221 sp. R570) encodes for 6 CAO copies, the largest number we report. This species is a result of inter-222 specific hybridization (Saccharum officinarum x spontaneum; Dumont et al. 2022), which has 223 resulted in a very large and highly redundant genome (~0.3 Gb, ~48,000 genes) (Healey et al. 2024). 224 It must be noted that these results are likely affected by the availability of many more angiosperm 225 genomes in public databases, while other plant and algal groups are not as well represented. As 226 more genomes are sequenced, these analyses will have to be revisited. 227 We also detected multiple occurrences of heterodimeric CAO genes, where the functional CAO 228 protein is encoded by two genes (Table 1, Supplementary Table S1). Previous work in the 229 prasinophyte Micromonas pusila demonstrated that in this species CAO was encoded by two 230 separate genes, one that possesses the mononuclear iron-binding motif (MpCAO1) and a second 231 one that encodes for the Rieske cluster (MpCAO2). It was shown that a combination of both gene 232 products was necessary to form an active heterodimeric CAO protein (Kunugi et al. 2013; Dey et al. 233 2023). We expand on this work by showing the presence of a heterodimeric CAO encoded by two 234 separate genes across the Mammiellophycea. This group within the prasinophytes contains M. 235 pusila and several of its relatives (Worden et al. 2009), the ecologically important picoalgae 236 Ostreococcus tauri, Ostreococcus lucimarinus and Bathycoccus prasinos (Palenik et al. 2007; 237 Moreau et al. 2012; Blanc-Mathieu et al. 2014), as well as the related Chloropicon primus 238 (Chloropicophyceae) (Lemieux et al. 2019). 239 We also show that heterodimeric CAO is not restricted only within the Prasinophytes. We detect a 240 similar CAO gene arrangement in the genome of the multinucleate green alga Caulerpa lentillifera 241 (Bryopsidales) (Arimoto et al. 2019) , as well as in in the genome of Bigelowiella natans 242 (Chlorarachniophyceae) (Curtis et al. 2012). Phylogenetic analysis based on plastid-encoded 243 protein sequences revealed that chlorarachniophytes acquired their secondary plastids by the 244 uptake of a filamentous green alga in the Bryopsidales (Suzuki et al. 2016). Thus, B. natans likely 245 gained a heterodimeric CAO from the Bryopsidales endophyte. Considering that CAO in all other 246 green lineages (Supplementary Table 1) and within the prochlorophytes (Nagata et al. 2004) is also 247 encoded by a single gene that contains both the Rieske and iron-binding domain it is likely that the 248 occurrence of two separate CAO genes is not an ancestral condition. Instead, it has been 249 postulated that CAO was encoded by a single gene in the common ancestor of green plants 250 (Kunugi et al. 2016). Indeed, it has been suggested that acquisition of two separate genes encoding 251 for key functional domains may not be an evolutionarily difficult process. CAO likely functions as a 252 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 8 trimer (Kunugi et al. 2013; Liu et al. 2022a), and the electron transfer from the Rieske to the 253 mononuclear iron occurs between two neighbouring CAO proteins. A functionally similar process 254 likely occurs in the CAO enzyme encoded by two separate genes. 255 Table 1: Number of species across the Viridiplantae with fully sequenced and annotated genomes 256 available through Phytozome (Total), compared to the number of species that encode for multiple 257 CAO genes (Multiple CAO) and number of species that encode for two CAO genes that form 258 heterodimeric complex (Heterodimeric CAO). 259 260 Group Total Heterodimeric CAO Multiple CAO Chlorophytes* 70 11 8 Charophytes 10 0 2 Non-vascular plants 6 0 3 Lycophytes 3 0 2 Gymnosperms 1 0 0 Angiosperms 151 0 62 *The Chlorophyte group also includes Bigelowiella natans (Chlorarachniophyceae) that has gained 261 secondary plastids by the uptake of a green alga. 262 All examined streptophytes encode for at least one CAO gene (Supplementary Table 1) signifying 263 the importance of Chl b for plant physiology, but we were not able to identify a homolog of CAO in 264 several chlorophyte algal species. Not surprisingly, we did not detect a CAO homolog in the 265 invertebrate parasite Helicosporidium sp., a colorless protist that belongs within the 266 trebouxiophycean green algal clade (Tartar et al. 2003). As a consequence of its parasitic lifestyle, 267 Helicosporidium has lost nearly all genes associated with light harvesting, photosynthesis and 268 chlorophyll biogenesis (Pombert et al. 2014), including CAO. 269 We were also not able to conclusively identify a CAO gene in the genomes of the endolithic coral 270 holobiont Ostreobium quekettii (Bryopsidales) (Iha et al. 2021), the aquaculture species 271 Tetraselmis striata (Chlorodendrophyceae) (Steadman Tyler et al. 2019) and the halotolerant 272 Picocystis sp. ML (Picocystophyceae) (Junkins et al. 2019). All identified sequences within the 273 genomes of these species had e-values higher than the cut-off (≥e-15) used in this work and high 274 similarity to pheophorbide a oxygenase (PAO), a Rieske-type protein involved in chlorophyll 275 degradation (Pružinská et al. 2003). The lack of CAO genes in these algae is surprising and could be 276 an artifact of genome assembly, as Chl b has been detected in all three species (Roesler et al. 277 2002; Massé et al. 2020; Conlon et al. 2024). The genome of O. queketti has a low BUSCO score 278 (60.7%) (Iha et al. 2021) and the BUSCO score is not reported for the draft genomes of T. striata and 279 P. sp. ML (Junkins et al. 2019; Steadman Tyler et al. 2019). The presence of CAO should be 280 experimentally examined in these species. 281 It is tempting to speculate that these species may have a different mechanism for Chl b 282 biosynthesis or a CAO sequence significantly different than most algae and plants. For instance, 283 Ostreobium is an endolithic coral holobiont adapted to extreme shading, and has lost many genes 284 associated with photoprotection and light perception due to evolution in a light-limited 285 environment (Iha et al. 2021). This alga is also the only known eukaryote that has lost the gene 286 encoding light-dependent protochlorophyllide oxidoreductase (LPOR), a key enzyme that catalyzes 287 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 9 the conversion of protochlorophyllide to chlorophyllide a, a necessary step towards the synthesis 288 of Chl a (Reinbothe et al. 2010). Instead, this alga fully depends on dark operative POR (DPOR) to 289 synthesize Chl a, likely as an adaptation to extreme shading (Iha et al. 2021). The mechanism of 290 chlorophyll biosynthesis has not been examined in the species in detail, but it is clear that the 291 endolithic lifestyle has resulted in a unique physiology. Further work will shed light on the role of 292 CAO in this and other shade-tolerant species. 293 We used a very conservative method to identify putative CAO sequences with both a Rieske and 294 Fe-binding domains, but this strict approach may have prevented the identification of duplicates 295 that may have diverged or become pseudogenes. Indeed, using a less conservative cutoff (≤e-5) 296 when screening the genomes of A. thaliana and C. reinhardtii (Wang et al. 2022; Craig et al. 2022) 297 reveals that PAO and Translocon at the inner envelope membrane of chloroplasts 55 (TIC55) may 298 be paralogous to CAO (Supplementary Table S2). PAO and TIC55 encode for enzymes with a Rieske 299 and Fe-binding domains but share a very low sequence similarity with CAO and each other 300 (Supplementary Table S2, Supplementary Figure S1). Both PAO and TIC55 are involved in 301 chlorophyll breakdown and allow for the degradation of the highly phototoxic pheophorbide a and 302 its export from the chloroplast (recently reviewed in Kuai et al. 2018). Evolutionary analysis of the 303 genes involved in chlorophyll degradation revealed that this pathway was already present in the 304 common ancestor of land plants, suggesting that PAO and TIC55 may share their origins with CAO 305 (Schumacher et al. 2022). This hypothesis has not yet been experimentally supported as ancestral 306 monooxygenase have been difficult to distinguish without detailed knowledge on their biochemical 307 activity and function (Schumacher et al. 2022). This possibility raises interesting questions about 308 the ancient evolutionary origins of chlorophyll metabolism within the Viridiplantae. 309 Was CAO gene duplication an ancestral event? 310 To determine whether the CAO gene duplication was an ancestral event, we performed a 311 phylogenetic analysis of all high-quality CAO amino acid sequences (except heterodimeric CAO). 312 Phylogenetic patterns typically match previously shown relationships among major Viridiplantae 313 groups (Figure 2), although some relationships are poorly resolved (particularly among the algal 314 groups and the Rosids; Supplementary Figure 2). The presence of two CAO genes has been 315 previously reported in rice (OsCAO1 and OsCAO2) where the two genes were positioned in tandem 316 on Chromosome 10, taken as evidence for a recent gene duplication in this species (Lee et al. 317 2005). A tandem CAO duplication, however, appears to be the exception rather than the rule. In 318 addition to rice, we only observed CAO duplicates located in proximity to each other in the 319 genomes of a handful of species: the green alga Volvox reticuliferus, the lycophyte Diphastrium 320 complanatum, and the magnoliid Cinnamomum kanehirae (Table 2). 321 There are several instances, mainly in very closely related species, where the duplication may have 322 originated in a recent shared ancestor. There are several possible cases in the Trebouxiophyceae 323 and Chlorophyceae class, where CAO duplicates are detected in closely related Chlorella and 324 Scenedesmus species (Supplementary Figure 2A). This insight must be interpreted with caution as 325 phylogenetic resolution within the chlorophytes is still weak due to limited gene and taxon 326 sampling (Li et al. 2021). Within the bryophytes, Spaghnum magellanicum and Sphagnum falax, 327 species that share a very close evolutionary relationship (Piatkowski and Shaw 2019; Bell et al. 328 2020), both have two CAO genes possibly arising from an ancestral duplication. Two CAO genes 329 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 10 are also encoded in the genome of P. patens (Zhang et al. 2023). This raises the possibility of a CAO 330 duplication in a recent shared ancestor (Supplementary Figure 2B), but the bryophytes are 331 represented by only 5 species in our analysis. Additional genomic sequencing and a deeper 332 phylogenetic analysis will be needed to clarify the evolution of CAO within these groups. 333 Figure 2: A phylogenetic tree based on CAO amino acid sequences inferred from using maximum 334 likelihood analysis, using 326 unique CAO sequences across 246 species color-coded to represent 335 the major Viridiplantae groups. Scale bar indicates amino acid substitutions per site. Species with 336 more than one detected CAO gene are highlighted in red 337 nob ct ri h oro h t h ro h t r o h t co h t nos r s no ii s onocots Ast ri s s osi s bi s i s .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 11 Species that encode for multiple CAO genes within their genomes are more prevalent among 338 angiosperms. Within the monocots, CAO duplicates arising from a shared ancestral gene are likely 339 present in the Asparagaceae, specifically among the closely related Agave tequiliana, Yucca 340 filamentosa and Yucca aloifolia (Ji et al. 2023) , although we were only able to detect a single CAO 341 copy in Asparagus officinalis (Supplementary Figure 2C). We detected two CAO gene copies in two 342 Brachypodium species (B. mexicanum and B. hybridum) but not in their relatives from the same 343 genus (B. arbuscula, B. stacei, B. sylvaticum and B. distachyon), raising the possibility of an 344 ancestral duplication followed by gene loss. All phylogenetic studies in Brachypodium support a 345 rapid and recent divergence from a common ancestor (Catalan et al. 2016), supporting this 346 possibility. 347 CAO duplication may be ancestral in the two closely related Saxifragales Kalanchoe laxiflora and 348 Kalanchoe fedtschenkoi (Han et al. 2024). We detected two CAO genes in K. fedtschenkoi and four 349 in K. laxiflora, possibly arising in a recent ancestor, followed by a second duplication within K. 350 laxiflora (Supplementary Figure 2D). Within the Fabids, four Populus species and their relative Salix 351 purpurea, each encode for two CAO gene copies that may be a result of an ancestral duplication. 352 The cultivated soybean Glycine max and its wild relative Glycine soja each encode for four CAO 353 gene copies (Supplementary Figure 2D) that are closely related to their homologs between the two 354 species. The phylogenetic separation of CAO1/2 and CAO3/4 from G. max and G. soja is puzzling, 355 as the two species are the closest relatives within the Glycine genus (Zhuang et al. 2022). All 356 Gossypium species examined here encode for four CAO genes, except G. raimondii which encodes 357 for two copies (Supplementary Figure 2E). The only other member of the Malvaceae examined, 358 Theobroma cacao, encodes for a single CAO gene, and although it clusters with CAO from 359 Gossypium sp., the relationship is poorly supported. Finally, there is a possibility of ancestral 360 duplications within the Brassicaceae (Supplementary Figure 2E), but this family has a poor 361 resolution due to factors such as incomplete sampling, rapid radiation, and hybridization (Huang et 362 al. 2016; Hendriks et al. 2023). 363 The genomes of multiple species across the Viridiplantae encode for more than one CAO gene, but 364 are these gene duplications related to plant evolutionary patterns? The green lineage is 365 characterized by a dynamic and complex evolutionary history marked by multiple whole genome 366 (WGD) and segmental duplication events, gene family expansions, and lineage-specific gene 367 losses (Panchy et al. 2016; Kersey 2019; Leebens-Mack et al. 2019; Soltis and Soltis 2021). Such 368 tremendous diversity in genome content can be a challenge for elucidating CAO evolutionary 369 trajectories even withing a single family. For instance, the Brassicaceae have been shaped by 370 multiple well-established WGD events (Mabry et al. 2020), but we detected only a single CAO gene 371 in the genome of A. thaliana and its closest relatives. In contrast, the genomes of several other 372 Brassica species encode for two or more CAO genes (Figure 2). This highlights the importance of 373 considering not only gene duplication but also gene retention dynamics when interpreting the 374 evolution of CAO. Future work that incorporates in-depth analyses, such as relative dating of CAO 375 duplication events and genome-wide synteny, will further clarify the evolutionary trajectory of CAO 376 within the Viridiplantae. 377 We show that CAO gene duplication is not a rarity among the Viridiplantae, but are all CAO 378 duplicates expressed and do they encode for active proteins? In our work, we focused only on CAO 379 genes with highly conserved catalytic domains likely to encode for active enzymes, but the 380 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 12 functional role of CAO duplication may be complex. A survey of published transcriptomes from 16 381 representative species from all major groups (Chlorophytes, Charophytes, non-vascular and 382 vascular plants) suggests that all CAO duplicate genes within a species are expressed, albeit at 383 different levels. Typically, one CAO gene has a higher expression than its paralog within the same 384 genome (Supplementary Table 4) suggesting differential regulation of CAO expression. 385 Significantly, all experimental studies to date suggest that the expression of CAO duplicates may 386 be differently regulated by light availability. In rice, the expression of OsCAO1 is light-inducible, 387 while OsCAO2 accumulates during periods of darkness and is expressed mainly in non-388 photosynthetic tissues (Lee et al. 2005). In P. patens, both PpCAO1 and PpCAO2 are expressed in 389 both light and dark conditions, but only PpCAO1 is upregulated with prolonged light exposure. 390 Similarly, in two Chlamydomonas species that harbour CAO duplicate genes (Chlamydomonas 391 priscui and Chlamydomonas sp. ICE-MDV), only one CAO homolog is upregulated in response to 392 increases in light intensity, while the second one is constitutively expressed (Poirier et al. 2025). 393 The role of light in regulating CAO expression has been also demonstrated previously in 394 Arabidopsis and tobacco, species with a single CAO gene copy (Pattanayak et al. 2005; Tanaka and 395 Tanaka 2005; Biswal et al. 2012, 2024). 396 Additionally, the mechanisms behind the control of Chl b synthesis may be species specific. 397 Mutational analysis in rice revealed that only OsCAO1 was necessary for Chl b biosynthesis. 398 OsCAO1 knockout mutants exhibited low Chl b levels, disordered thylakoid membrane 399 arrangements, membrane damage due to excessive ROS accumulation and decreased chilling 400 tolerance (Lee et al. 2005; Jung et al. 2021; Xiong et al. 2024), but these effects were absent in 401 OsCAO2 knockouts (Lee et al. 2005). In contrast, knocking out either of the two CAO homologs in 402 P. patens lead to significantly reduced Chl b levels suggesting a dependency on two active CAO 403 enzymes for the appropriate control of Chl b synthesis in this species (Zhang et al. 2023). 404 It is tempting to speculate that species that encode for multiple CAO genes have a more robust 405 capacity for Chl b biosynthesis through an increase in gene dosage or a finer control over Chl b 406 accumulation though gene subfunctionalization or differential expression (Kondrashov 2012; 407 Panchy et al. 2016). Two gene copies with different expression patterns or activity levels could 408 provide a fine-tuned control of Chl b levels and photosynthetic performance under varying 409 environmental conditions, but additional experimental work is needed to substantiate this 410 hypothesis. 411 412 413 414 Table 2: Distribution of CAO duplicate genes across Viridiplantae. Only genes with a confirmed 415 Riske conserved cluster, Fe-binding domain (all species), and degron sequence (Streptophytes 416 only) are listed. A checkmark (✓) indicates that the CAO genes are positioned in a tandem on the 417 same chromosome within the genome, while a cross (x) indicates the genes are located on 418 different chromosomes. Accession numbers can be found in Supplementary Table 1 (continued on 419 next page). 420 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 13 Taxon Species #CAO genes Tandem Chlorophyta Chlorophyceae Chlamydomonas priscui 2 x Chlorophyceae Volvox reticuilferus 2 ✓ Chlorophyceae Tetradesmus deserticola 2 x Trebouxiophyceae Chlorella sp. A99 2 x Trebouxiophyceae Chlorella sorokiniana 2 x Trebouxiophyceae Chlorellaceae sp. 2 x Trebouxiophyceae Micractinium conductrix 2 x Pedinophyceae Pedinomonas minor 2 x Charophyta Zygnematophyceae Mesotaenium endlicherianum 2 x Zygnematophyceae Spirogloea muscicola 2 x Bryophyta Bryopsida Physcomitrium patens 2 x Sphagnopsida Sphagnum fallax 2 x Sphagnopsida Sphagnum magellanicum 3 x Lycophyta Equisetopsida Diphasiastrum complanatum 4 ✓ Equisetopsida Ceratopteris richardii 2 x Angiosperms Laurales Cinnamomum kanehirae 2 ✓ Nymphaeales Nymphaea colorata 2 x Asparagales Agave tequilana 2 x Asparagales Yucca aloifolia 3 x Asparagales Yucca filamentosa 3 x Zingiberales Musa acuminata v1 2 x Alismatales Spirodela polyrhiza v2 2 x Poales Typha latifolia 2 x Poales Eleusine coracana 2 x Poales Oryza sativa 2 ✓ Poales Thinopyrum intermedium 3 x Poales Brachypodium hybridum 2 x Poales Brachypodium mexicanum 2 x Poales Zea mays 2 x Poales Andropogon gerardi 3 x Poales Miscanthus sinensis 2 x Poales Panicum virgatum 2 x Poales Saccharum officinarum 6 x Caryophyllales Chenopodium quinoa 2 x Gentianales Coffea arabica geisha 2 x .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 14 Table 2:(continued) Taxon Species #CAO genes Tandem Apiales Hydrocotyle leucocephala 2 x Boraginales Ehretia anacua 2 x Asterales Helianthus annuus 2 x Lamiales Olea europaea 2 x Saxifragales Kalanchoe fedtschenkoi 2 x Saxifragales Kalanchoe laxiflora 4 x Fabales Arachis hypogaea 2 x Fabales Chamaecrista fasciculata 2 x Fabales Glycine max 4 x Fabales Glycine soja 4 x Fabales Lupinus albus 2 x Fabales Phaseolus vulgaris 2 x Fagales Carya illinoinensis 2 x Rosales Fragaria x ananassa 3 x Rosales Malus domestica 2 x Malpighiales Linum usitatissimum 4 x Malpighiales Populus deltoides 2 x Malpighiales Populus nigra x maximowiczii 2 x Malpighiales Populus tremula x alba 2 x Malpighiales Populus trichocarpa 2 x Malpighiales Salix purpurea 2 x Sapindales Anacardium occidentale 3 x Malvales Gossypium barbadense 4 x Malvales Gossypium darwinii 4 x Malvales Gossypium hirsutum 4 x Malvales Gossypium mustelinum 4 x Malvales Gossypium raimondii 2 x Malvales Gossypium tomentosum 4 x Brassicales Alyssum linifolium 2 x Brassicales Caulanthus amplexicaulis 2 x Brassicales Iberis amara 2 x Brassicales Lepidium sativum 2 x Brassicales Stanleya pinnata 2 x Brassicales Brassica juncea 3 x Brassicales Brassica oleracea 2 x Brassicales Cakile maritima 2 x Brassicales Crambe hispanica 2 x Brassicales Eruca vesicaria 2 x Brassicales Sinapis alba 3 x Brassicales Camelina sativa 3 x Brassicales Isatis tinctoria 4 x 421 422 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 15 Evolutionary implications of the CAO degron sequence 423 In addition to light-dependant regulation at the level of transcription, the main regulation of CAO 424 occurs post-translationally. Early work in A. thaliana demonstrated that this regulation is governed 425 by the presence of a degron sequence (QDLLTIMILH) in the protein’s N-terminal A domain that acts 426 as a docking site for a chloroplast Clp protease (Yamasato et al. 2005; Nakagawara et al. 2007; 427 Sakuraba et al. 2009). It has been hypothesized that the degron sequence is “hidden” within the A 428 domain when there are insufficient amounts of Chl b, but accumulation of this product causes a 429 conformational change in the CAO protein, exposing the degron, and giving access to the protease 430 (Tanaka and Tanaka 2011). The evolution of this negative-feedback regulatory loop has not been 431 examined in detail, but it has been suggested that the degron was acquired early in the 432 Viridiplantae evolution as a mechanism for high light adaptation (Kunugi et al. 2016). 433 Here we show that the degron sequence is highly conserved among all land plants (86% identity), 434 moderately conserved among charophytes and non-vascular plants (68% and 66% identity 435 respectively), and very poorly conserved in chlorophytes (39% identity) (Figure 3; Supplementary 436 Figure 3). Land plants (both vascular and non-vascular) have several highly conserved sites across 437 all species (Asparagine at position 2, Isoleucine at position 6, Histidine at position 10), whereas 438 these positions are only moderately conserved in charophytes, and poorly conserved in 439 chlorophytes. In contrast, the Rieske cluster [CXH(X)15-17CXXH] exhibits an 84% pairwise identity 440 among all species, where the two histidine (positions 3 and 23) and cysteines (positions 1 and 20) 441 that coordinate the metallocentre [2Fe-2S] (Przybyla-Toscano et al. 2021) are present in all 442 examined sequences. Nearly all residues are conserved among the Viridiplantae in the C-terminal 443 Fe-binding domain (96% pairwise identity). Overall, these results suggest that the mechanism of 444 Chl b biosynthesis involving the Rieske and Fe-binding domains is ancestral and conserved among 445 all photosynthetic species, but this is not the case for the regulatory degron domain. 446 The lack of a highly conserved degron sequence may have several implications in the evolution of 447 efficient light harvesting. For instance, the N-regulatory domain (including the degron) is 448 completely absent in marine eukaryotic Mammiellophycea that encode for a heterodimeric CAO 449 gene, as well as in their deep-water relatives from the order Palmophyllales. These species 450 typically have low Chl a/b ratios and, unlike land plants, can incorporate Chl b into their core 451 antennae. This was interpreted as a necessary adaptation for efficient light harvesting in deep 452 marine environments characterized by low levels of blue-green light (Kunugi et al. 2016). 453 454 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 16 Figure 3: Key domains in the CAO protein across the Viridiplantae. (A) Sequence logos of the 455 conserved motifs identified in the CAO protein sequences. N and C represent the N-terminus and 456 C-terminus, respectively. All sequence logos are superimposed on a graphical representation of 457 plant CAO sequences (not to scale) showing the regulatory degron motif, conserved Rieske domain 458 cluster and Fe-binding domain (Sakuraba et al., 2009; Liu et al., 2022). The logos were generated 459 for CAO proteins from chlorophyte algae, charophyte algae, non-vascular plants, and vascular 460 plants. The sequences on top of each logo are the consensus sequence with the most common 461 base in each sequence alignment. (B) Comparison of the CAO amino acid sequences 462 corresponding to the degron and its surrounding region in the N-terminal regulatory domain in 463 representative species from Chlorophyte, Charophyte, non-vascular and vascular land plants. The 464 region that corresponds to the degron sequence as described in A. thaliana is emphasized with a 465 box. Amino acids that are identical with the A. thalina sequence are colored, and those that are 466 different are shown in white. 467 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 17 It has been suggested that the degron sequence evolved in shallow-water algae and land plants as 468 a means to regulate Chl b accumulation and limit Chl b incorporation into the reaction center in an 469 environment characterized with high light intensities (Kunugi et al. 2016). Indeed, such species 470 exclusively harbour Chl a in their core complexes, but this appears to be controlled by pigment 471 availability rather than the chemical nature of the two chlorophyll types. The binding affinity of Chl 472 a and b to the PSI complex is similar (Ikegami et al. 2007) and cyanobacteria that do not naturally 473 produce Chl b successfully incorporated this pigment into their reaction centers when transformed 474 with plant CAO (Satoh et al. 2001; Xu et al. 2001). Incorporation of Chl b into the reaction center 475 accompanied by increased Chl b accumulation was also observed in Arabidopsis when full length 476 CAO was overexpressed or replaced with a prokaryotic CAO that lacked a degron sequence 477 (Hirashima et al. 2006; Sakuraba et al. 2009). 478 The chlorophyll cycle has not been examined in detail in the core chlorophytes (Chlorophyceae, 479 Trebouxiophyceae, Ulvophyceae), a diverse group that inhabits marine, freshwater and terrestrial 480 habitats (Leliaert et al. 2012). Although many chlorophytes thrive in high light, others are adapted to 481 light limited environments, including deep waters. As a result, these species have developed 482 adaptations allowing for efficient light harvesting in low light. For instance, Chlamydomonas priscui 483 is endemic to the depths of the Antarctic Lake Bonney, an environment characterized by extreme 484 shading and dominated by blue-green wavelengths (Neale and Priscu 1995). This alga displays a 485 host of adaptations to extreme shading including a large LHC antennae size (reviewed in 486 Cvetkovska et al. 2017; Hüner et al. 2022). As suggested for marine prasinophytes (Kunugi et al. 487 2016), such low-light adapted chlorophytes could benefit from increased Chl b levels and could 488 thrive without a Chl b-dependent feedback mechanism. 489 Despite the lack of a conserved degron sequence to regulate CAO protein levels (Figure 3), 490 chlorophytes have maintained the ability to alter their Chl a/b ratio in response to light (Leliaert et 491 al. 2012). Transcriptional control of CAO levels was postulated to be a minor aspect of adjusting 492 Chl b levels in plants (Tanaka and Tanaka 2019), but this process could play a bigger role in 493 chlorophytes. It is also possible that a different, yet unidentified, mechanism exists for Chl b 494 turnover independent of a degron sequence, allowing these species to modify their Chl a/b ratios. 495 Finally, we can not rule out that the region corresponding to a degron in Chlorophyte CAO may still 496 have a similar function despite poor sequence conservation with land plants. The mechanism of 497 regulating protein degradation via degron motifs is well understood in plants (Isono et al. 2024), but 498 insights on algal degrons is lacking. A detailed experimental investigation on the control and 499 mechanism of Chl b biosynthesis is clearly needed among the Chlorophytes. For instance, targeted 500 mutagenesis of the N-terminus of algal CAO will reveal the importance and role of this regulatory 501 region in chlorophyll turnover. 502

and Drawbacks 503 Our work suggests that CAO gene duplication is widespread among the Viridiplantae, likely 504 originating from multiple independent duplication events throughout plant evolution. While we 505 show that the presence of multiple CAO gene copies is highly prevalent in land plants, these 506

must be interpreted with caution. Many more angiosperm genomes are available in public 507 databases, while other plant and algal groups are not as well represented. Thus, the conclusions of 508 our study must be re-visited once more genomes are sequenced. Furthermore, many genome 509 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 18 sequencing initiatives are focused on plants of economic and agricultural importance, in which 510 domestication and selective breeding may have affected the genomic content (Turner-Hissong et 511 al. 2020). While the exact physiological impact of CAO duplication is unknown, a study in 512 transgenic tobacco overexpressing CAO (Biswal et al. 2012) showed that an increase in Chl b may 513 be a strategy to enhance CO2 assimilation, delay senescence, and enhance productivity. Such 514 traits would be attractive in crop breeding programs (Voitsekhovskaja and Tyutereva 2015), and the 515 high prevalence of CAO duplication seen in crop plants in this study may have been selected for 516 during plant breeding. Indeed, several crop plants (including corn, soybean, and olives, among 517 others) appear to express all CAO genes encoded in their genome (Supplementary Table 4). The 518 functional role of these duplicates in plant physiology, photosynthetic efficiency, and productivity 519 warrants further investigation. 520 Our understanding of the prevalence and evolution of CAO at the genetic level will be improved in 521 future years, as more algal and non-vascular plant genomes become available. Furthermore, there 522 are still numerous complex and outstanding questions regarding the control of Chl b accumulation 523 in photosynthetic species. In particular, the chlorophytes are an understudied group when it 524 comes to Chl b regulating mechanism. Detailed genomic and experimental studies will shed light 525 on the role and regulation of Chl b in these species and will deepen our understanding on this 526 crucial photosynthetic pigment. 527 Supplementary Information 528 Supplementary Table S1: A summary of genes encoding for chlorophyllide a oxygenase (CAO) 529 across the Viridiplantae used in this work. All sequences were obtained from Phycocosm 530 (Chlorophytes, Euglenophytes, Cercozoa, Charophytes) or Phytozome v13 (Bryophytes, 531 Marchantiophytes, Lycophytes, Gymnosperms, Angiosperms). Accession numbers represent the 532 unique IDs assigned to each gene (Phycocosm: Gene portal|Protein ID|Gene Model; Phytozome: 533 Gene Model). E-values are based on a tBLASTn search with the Chlamydomonas reinhardtii CAO 534 (Cre01.g043350) for species within the Chlorophyta and Charophyta, and the Arabidopsis thaliana 535 CAO (AT1G44446) for all other species 536 Supplementary Table S2: BLAST results obtained by screening the genome of Arabidopsis thaliana 537 and Chlamydomonas reinhardtii for CAO homologs. Accession numbers were obtained from 538 Phytozome v13 (Goodstein et al. 2012). The functional domains according to InterPro (Blum et al. 539 2025) that are shared among all proteins are highlighted with an asterisk (Rieske Domain; InterPro: 540 IPR017941; Pheophorbide a Oxygenase Domain; InterPro: IPR013626). 541 Supplementary Table S3: A summary of the occurrence of CAO duplicate genes across the major 542 groups within Viridiplantae. The number in brackets represents the proportion of species which 543 encode for multiple copies of the CAO gene (Mutiple CAO) or encode for two CAO genes that form 544 heterodimeric complex (Heterodimeric CAO). 545 Supplementary Table S4: CAO gene expression in species whose genome encodes for more than 546 one CAO gene. All data was obtained from previously published work. Gene expression is 547 presented as FPKM, RPKM or TPM averages with standard deviation. In all cases, gene expression 548 ratios were determined by dividing all expression values by the highest CAO expression level within 549 the species. 550 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 19 Supplementary Figure 1: An alignment of the CAO protein from (A) Arabidopsis thaliana and (B) 551 Chlamydomonas reinhardtii with all other proteins identified using the CAO amino acid sequence 552 as a query. The color of the residue depends on the degree of conservation among proteins (blue 553 fully conserved; green 80–100% similar; yellow 60–80% similar; white < 60% similar). The important 554 domains that are shared among all proteins are highlighted in green (Rieske Domain; InterPro: 555 IPR017941), yellow (Fe-binding domain) and blue (Pheophorbide a Oxygenase Domain; InterPro: 556 IPR013626). 557 Supplementary Figure 2: A phylogenetic tree of CAO genes inferred from using maximum 558 likelihood analysis. The node labels show the bootstrap support, with dashed lines indicating <50 559 bootstrap support. Scale bar indicates amino acid substitutions per site. Species with more than 560 one detected CAO gene are highlighted in red. Cases where CAO duplication may have arisen in a 561 common shared ancestor are marked with a blue arrow, and only those located on branches with a 562 strong bootstrap support (>85%) are shown. The inset shows the outline of the full CAO 563 phylogenetic tree inferred from using maximum likelihood analysis, using 326 unique CAO 564 sequences across 246 species color coded to represent the major Viridiplantae groups. The 565 position of the chlorophytes (green) is highlighted with a black box. (A) Chlorophytes; (B) 566 Charophytes, bryophytes (including marachantiophytes), and lycophytes; (C) Gymnosperms, 567 magnoliids, and monocots; (D) Basal eudicots, asterids and fabids; (F) Malvids. 568 Supplementary Figure 3: Multiple sequence alignment of the predicted CAO amino acid 569 sequences from Chlorophytes (A), Charophytes (B), non-vascular plants (C) and vascular plants 570 (D). Global alignments were performed with MUSCLE. The colour of the residue depends on the 571 degree of conservation among species (blue, fully conserved; green, 80–100% similar; yellow, 60–572 80% similar; gray, < 60% similar). As a reference, the CAO sequence from Arabidopsis thaliana is 573 present in all alignments (top), with the key domains annotated (green, Degron sequence; blue, 574 conserved Rieske cluster; purple, Fe-binding domain) 575

576 Arimoto A, Nishitsuji K, Higa Y, et al (2019) A siphonous macroalgal genome suggests convergent 577 functions of homeobox genes in algae and land plants. DNA Research 26:183–192. 578 https://doi.org/10.1093/dnares/dsz002 579 Arriola MB, Velmurugan N, Zhang Y, et al (2018) Genome sequences of Chlorella sorokiniana UTEX 580 1602 and Micractinium conductrix SAG 241.80: implications to maltose excretion by a 581 green alga. The Plant Journal 93:566–586. https://doi.org/10.1111/tpj.13789 582 Bell D, Lin Q, Gerelle WK, et al (2020) Organellomic data sets confirm a cryptic consensus on 583 (unrooted) land-plant relationships and provide new insights into bryophyte molecular 584 evolution. Am J Bot 107:91–115. https://doi.org/10.1002/ajb2.1397 585 Biswal AK, Pattanayak GK, Pandey SS, et al (2012) Light Intensity-Dependent Modulation of 586 Chlorophyll b Biosynthesis and Photosynthesis by Overexpression of Chlorophyllide a 587 Oxygenase in Tobacco. Plant Physiol 159:433–449. https://doi.org/10.1104/pp.112.195859 588 Biswal AK, Pattanayak GK, Ruhil K, et al (2024) Reduced expression of chlorophyllide a oxygenase 589 (CAO) decreases the metabolic flux for chlorophyll synthesis and downregulates 590 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 20 photosynthesis in tobacco plants. Physiol Mol Biol Plants 30:1–16. 591 https://doi.org/10.1007/s12298-023-01395-5 592 Blanc-Mathieu R, Verhelst B, Derelle E, et al (2014) An improved genome of the model marine alga 593 Ostreococcus tauri unfolds by assessing Illumina de novo assemblies. BMC Genomics 594 15:1–12. https://doi.org/10.1186/1471-2164-15-1103 595 Blankenship RE (2010) Early Evolution of Photosynthesis1. Plant Physiol 154:434–438. 596 https://doi.org/10.1104/pp.110.161687 597 Blum M, Andreeva A, Florentino LC, et al (2025) InterPro: the protein sequence classification 598 resource in 2025. Nucleic Acids Research 53:D444–D456. 599 https://doi.org/10.1093/nar/gkae1082 600 Brzezowski P, Richter AS, Grimm B (2015) Regulation and function of tetrapyrrole biosynthesis in 601 plants and algae. Biochimica et Biophysica Acta - Bioenergetics 1847:968–985. 602 https://doi.org/10.1016/j.bbabio.2015.05.007 603 Bujaldon S, Kodama N, Rappaport F, et al (2017) Functional Accumulation of Antenna Proteins in 604 Chlorophyll b-Less Mutants of Chlamydomonas reinhardtii. Molecular Plant 10:115–130. 605 https://doi.org/10.1016/j.molp.2016.10.001 606 Cardona T (2019) Thinking twice about the evolution of photosynthesis. Open Biology 9:180246. 607 https://doi.org/10.1098/rsob.180246 608 Castresana J (2000) Selection of Conserved Blocks from Multiple Alignments for Their Use in 609 Phylogenetic Analysis. Molecular Biology and Evolution 17:540–552. 610 https://doi.org/10.1093/oxfordjournals.molbev.a026334 611 Catalan P, López-Álvarez D, Díaz-Pérez A, et al (2016) Phylogeny and Evolution of the Genus 612 Brachypodium. In: Vogel JP (ed) Genetics and Genomics of Brachypodium. Springer 613 International Publishing, Cham, pp 9–38 614 Cheng S, Xian W, Fu Y, et al (2019) Genomes of Subaerial Zygnematophyceae Provide Insights into 615 Land Plant Evolution. Cell 179:1057–1067. https://doi.org/10.1016/j.cell.2019.10.019 616 Conlon T, Parkes R, Fierli D, Touzet N (2024) Comparative pigment and fatty acid profiling of marine 617 species within the chlorophyte genus Tetraselmis. Food Bioscience 58:103660. 618 https://doi.org/10.1016/j.fbio.2024.103660 619 Craig RJ, Gallaher SD, Shu S, et al (2022) The Chlamydomonas Genome Project, version 6: 620

assemblies for mating type plus and minus strains reveal extensive structural 621 mutation in the laboratory. The Plant Cell koac347. https://doi.org/10.1093/plcell/koac347 622 Curtis BA, Tanifuji G, Burki F, et al (2012) Algal genomes reveal evolutionary mosaicism and the fate 623 of nucleomorphs. Nature 492:59–65. https://doi.org/10.1038/nature11681 624 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 21 Cvetkovska M, Hüner NPA, Smith DR (2017) Chilling out: the evolution and diversification of 625 psychrophilic algae with a focus on Chlamydomonadales. Polar Biol 40:1169–1184. 626 https://doi.org/10.1007/s00300-016-2045-4 627 Cvetkovska M, Orgnero S, Hüner NPA, Smith DR (2019) The enigmatic loss of light‐independent 628 chlorophyll biosynthesis from an Antarctic green alga in a light‐limited environment. New 629 Phytol 222:651–656. https://doi.org/10.1111/nph.15623 630 de Vries J, Archibald JM (2018) Plant evolution: landmarks on the path to terrestrial life. New 631 Phytologist 217:1428–1434. https://doi.org/10.1111/nph.14975 632 Dey D, Tanaka R, Ito H (2023) Structural Characterization of the Chlorophyllide a Oxygenase (CAO) 633 Enzyme Through an In Silico Approach. J Mol Evol 91:225–235. 634 https://doi.org/10.1007/s00239-023-10100-9 635 Dumont T, Barau L, Thong-Chane A, et al (2022) Sugarcane Breeding in Reunion: Challenges, 636 Achievements and Future Prospects. Sugar Tech 24:181–192. 637 https://doi.org/10.1007/s12355-021-00998-0 638 Edgar RC (2004) MUSCLE: a multiple sequence alignment method with reduced time and space 639 complexity. BMC Bioinformatics 5:113. https://doi.org/10.1186/1471-2105-5-113 640 Espineda CE, Linford AS, Devine D, Brusslan JA (1999) The AtCAO gene, encoding chlorophyll a 641 oxygenase, is required for chlorophyll b synthesis in Arabidopsis thaliana. Proceedings of 642 the National Academy of Sciences of the United States of America 96:10507–10511. 643 https://doi.org/10.1073/pnas.96.18.10507 644 Ferraro DJ, Gakhar L, Ramaswamy S (2005) Rieske business: Structure–function of Rieske non-645 heme oxygenases. Biochemical and Biophysical Research Communications 338:175–190. 646 https://doi.org/10.1016/j.bbrc.2005.08.222 647 Goodstein DM, Shu S, Howson R, et al (2012) Phytozome: a comparative platform for green plant 648 genomics. Nucleic Acids Research 40:D1178–D1186. https://doi.org/10.1093/nar/gkr944 649 Gray J, Wardzala E, Yang M, et al (2004) A small family of LLS1-related non-heme oxygenases in 650 plants with an origin amongst oxygenic photosynthesizers. Plant Molecular Biology 54:39–651 54 652 Grigoriev IV, Hayes RD, Calhoun S, et al (2021) PhycoCosm, a comparative algal genomics 653 resource. Nucleic Acids Research 49:D1004–D1011. https://doi.org/10.1093/nar/gkaa898 654 Guiry MD (2024) How many species of algae are there? A reprise. Four kingdoms, 14 phyla, 63 655 classes and still growing. Journal of Phycology 60:214–228. 656 https://doi.org/10.1111/jpy.13431 657 Guo Q, Zhu S, Lai T, et al (2025) A Phosphate-Starvation Enhanced Purple Acid Phosphatase, 658 GmPAP23 Mediates Intracellular Phosphorus Recycling and Yield in Soybean. Plant, Cell 659 and Environment 1–18. https://doi.org/10.1111/pce.15400 660 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 22 Han S, Zhang S, Yi R, et al (2024) Phylogenomics and plastomics offer new evolutionary 661 perspectives on Kalanchoideae (Crassulaceae). Annals of Botany 133:585–604. 662 https://doi.org/10.1093/aob/mcae017 663 He Y, Deng Z, He S, et al (2025) Transcriptome and co-expression network analysis reveal the 664 genetic basis of cell wall components in maize stalks. BMC Genomics 26:. 665 https://doi.org/10.1186/s12864-025-11816-2 666 Healey AL, Garsmeur O, Lovell JT, et al (2024) The complex polyploid genome architecture of 667 sugarcane. Nature 628:804–810. https://doi.org/10.1038/s41586-024-07231-4 668 Healey AL, Piatkowski B, Lovell JT, et al (2023) Newly identified sex chromosomes in the Sphagnum 669 (peat moss) genome alter carbon sequestration and ecosystem dynamics. Nature Plants 670 9:238–254. https://doi.org/10.1038/s41477-022-01333-5 671 Hendriks KP, Kiefer C, Al-Shehbaz IA, et al (2023) Global Brassicaceae phylogeny based on filtering 672 of 1,000-gene dataset. Current Biology 33:4052-4068.e6. 673 https://doi.org/10.1016/j.cub.2023.08.026 674 Hirashima M, Satoh S, Tanaka R, Tanaka A (2006) Pigment Shuffling in Antenna Systems Achieved 675 by Expressing Prokaryotic Chlorophyllide a Oxygenase in Arabidopsis. The Journal of 676 Biological Chemistry 281:15385–15393. https://doi.org/10.1074/jbc.M602903200 677 Huang C-H, Sun R, Hu Y, et al (2016) Resolution of Brassicaceae Phylogeny Using Nuclear Genes 678 Uncovers Nested Radiations and Supports Convergent Morphological Evolution. Molecular 679 Biology and Evolution 33:394–412. https://doi.org/10.1093/molbev/msv226 680 Hüner NPA, Smith DR, Cvetkovska M, et al (2022) Photosynthetic adaptation to polar life: Energy 681 balance, photoprotection and genetic redundancy. Journal of Plant Physiology 268:153557. 682 https://doi.org/10.1016/j.jplph.2021.153557 683 Iha C, Dougan KE, Varela JA, et al (2021) Genomic adaptations to an endolithic lifestyle in the coral-684 associated alga Ostreobium. Current Biology 31:1393-1402.e5. 685 https://doi.org/10.1016/j.cub.2021.01.018 686 Ikegami I, Satoh S, Aoki M (2007) Binding Affinity of Chl b for the Chl a -Binding Sites in PSI Core 687 Complexes. Plant Cell Physiology 48:1092–1097. https://doi.org/10.1093/pcp/pcm078 688 Isono E, Li J, Pulido P, et al (2024) Protein degrons and degradation: Exploring substrate recognition 689 and pathway selection in plants. The Plant Cell 36:3074–3098. 690 https://doi.org/10.1093/plcell/koae141 691 Ji Y, Landis JB, Yang J, et al (2023) Phylogeny and evolution of Asparagaceae subfamily Nolinoideae: 692 new insights from plastid phylogenomics. Annals of Botany 131:301–312. 693 https://doi.org/10.1093/aob/mcac144 694 Jung YJ, Lee HJ, Yu J, et al (2021) Transcriptomic and physiological analysis of OsCAO1 knockout 695 lines using the CRISPR/Cas9 system in rice. Plant Cell Rep 40:1013–1024. 696 https://doi.org/10.1007/s00299-020-02607-y 697 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 23 Junkins EN, Stamps BW, Corsetti FA, et al (2019) Draft Genome Sequence of Picocystis sp. Strain 698 ML, Cultivated from Mono Lake, California. Microbiol Resour Announc 8:e01353-18. 699 https://doi.org/10.1128/MRA.01353-18 700 Kersey PJ (2019) Plant genome sequences: past, present, future. Current Opinion in Plant Biology 701 48:1–8. https://doi.org/10.1016/j.pbi.2018.11.001 702 Kim EH, Li XP, Razeghifard R, et al (2009) The multiple roles of light-harvesting chlorophyll a/b-703 protein complexes define structure and optimize function of Arabidopsis chloroplasts: A 704 study using two chlorophyll b-less mutants. Biochimica et Biophysica Acta - Bioenergetics 705 1787:973–984. https://doi.org/10.1016/j.bbabio.2009.04.009 706 Kondrashov FA (2012) Gene duplication as a mechanism of genomic adaptation to a changing 707 environment. Proceedings of the Royal Society B: Biological Sciences 279:5048–5057. 708 https://doi.org/10.1098/rspb.2012.1108 709 Król M, Spangfort MD, Huner NPA, et al (1995) Chlorophyll a/b-binding proteins, pigment 710 conversions, and early light-induced proteins in a chlorophyll b-less Barley mutant. Plant 711 Physiology 107:873–883. https://doi.org/10.1104/pp.107.3.873 712 Kuai B, Chen J, Hörtensteiner S (2018) The biochemistry and molecular biology of chlorophyll 713 breakdown. Journal of Experimental Botany 69:751–767. https://doi.org/10.1093/jxb/erx322 714 Kume A, Akitsu T, Nasahara KN (2018) Why is chlorophyll b only used in light-harvesting systems? 715 Journal of Plant Research 131:961–972. https://doi.org/10.1007/s10265-018-1052-7 716 Kunugi M, Satoh S, Ihara K, et al (2016) Evolution of Green Plants Accompanied Changes in Light-717 Harvesting Systems. Plant and Cell Physiology 57:1231–1243. 718 https://doi.org/10.1093/pcp/pcw071 719 Kunugi M, Takabayashi A, Tanaka A (2013) Evolutionary changes in chlorophyllide a oxygenase 720 (CAO) structure contribute to the acquisition of a new light-harvesting complex in 721 micromonas. Journal of Biological Chemistry 288:19330–19341. 722 https://doi.org/10.1074/jbc.M113.462663 723 Lamesch P, Berardini TZ, Li D, et al (2012) The Arabidopsis Information Resource (TAIR): improved 724 gene annotation and new tools. Nucleic Acids Res 40:D1202-1210. 725 https://doi.org/10.1093/nar/gkr1090 726 Lee S, Kim JH, Eun SY, et al (2005a) Differential regulation of chlorophyll a oxygenase genes in rice. 727 Plant Molecular Biology 57:805–818. https://doi.org/10.1007/s11103-005-2066-9 728 Lee S, Kim J-H, Yoo ES, et al (2005b) Differential regulation of chlorophyll a oxygenase genes in rice. 729 Plant Mol Biol 57:805–818. https://doi.org/10.1007/s11103-005-2066-9 730 Leebens-Mack JH, Barker MS, Carpenter EJ, et al (2019) One thousand plant transcriptomes and 731 the phylogenomics of green plants. Nature 574:679–685. https://doi.org/10.1038/s41586-732 019-1693-2 733 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 24 Leliaert F, Smith DR, Moreau H, et al (2012) Phylogeny and Molecular Evolution of the Green Algae. 734 Critical Reviews in Plant Sciences 31:1–46. https://doi.org/10.1080/07352689.2011.615705 735 Lemieux C, Turmel M, Otis C, Pombert J-F (2019) A streamlined and predominantly diploid genome 736 in the tiny marine green alga Chloropicon primus. Nat Commun 10:4061. 737 https://doi.org/10.1038/s41467-019-12014-x 738 Letunic I, Bork P (2021) Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display 739 and annotation. Nucleic Acids Res 49:W293–W296. https://doi.org/10.1093/nar/gkab301 740 Li X, Hou Z, Xu C, et al (2021) Large Phylogenomic Data sets Reveal Deep Relationships and Trait 741 Evolution in Chlorophyte Green Algae. Genome Biol Evol 13:evab101. 742 https://doi.org/10.1093/gbe/evab101 743 Liu J, Knapp M, Jo M, et al (2022a) Rieske Oxygenase Catalyzed C–H Bond Functionalization 744 Reactions in Chlorophyll b Biosynthesis. ACS Cent Sci 8:1393–1403. 745 https://doi.org/10.1021/acscentsci.2c00058 746 Liu Y, Liu J, Kong Z, et al (2022b) Transcriptomics and metabolomics analyses of the mechanism of 747 flavonoid synthesis in seeds of differently colored quinoa strains. Genomics 114:138–148. 748 https://doi.org/10.1016/j.ygeno.2021.11.030 749 Mabry ME, Brose JM, Blischak PD, et al (2020) Phylogeny and multiple independent whole-genome 750 duplication events in the Brassicales. Am J Bot 107:1148–1164. 751 https://doi.org/10.1002/ajb2.1514 752 Mascuñano B, Coto-Elena J, Guerrero-Sánchez VM, et al (2025) Transcriptome analysis of wild olive 753 (Olea Europaea L. subsp. europaea var. sylvestris) clone AC18 provides insight into the role 754 of lignin as a constitutive defense mechanism underlying resistance to Verticillium wilt. 755 BMC Plant Biology 25:. https://doi.org/10.1186/s12870-025-06301-7 756 Massé A, Tribollet A, Meziane T, et al (2020) Functional diversity of microboring Ostreobium algae 757 isolated from corals. Environmental Microbiology 22:4825–4846. 758 https://doi.org/10.1111/1462-2920.15256 759 Merchant SS, Prochnik SE, Vallon O, et al (2007) The Chlamydomonas genome reveals the 760 evolution of key animal and plant functions. Science 318:245–250. 761 https://doi.org/10.1126/science.1143609 762 Miller MA, Schwartz T, Pickett BE, et al (2015) A RESTful API for Access to Phylogenetic Tools via the 763 CIPRES Science Gateway. Evol Bioinform Online 11:43–48. 764 https://doi.org/10.4137/EBO.S21501 765 Moreau H, Verhelst B, Couloux A, et al (2012) Gene functionalities and genome structure in 766 Bathycoccus prasinos reflect cellular specializations at the base of the green lineage. 767 Genome Biol 13:1–16. https://doi.org/10.1186/gb-2012-13-8-r74 768 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 25 Mueller AH, Dockter C, Gough SP, Lundqvist U (2012) Characterization of Mutations in Barley fch2 769 Encoding Chlorophyllide a Oxygenase. Plant and Cell Physiology 53:1232–1246. 770 https://doi.org/10.1093/pcp/pcs062 771 Nagata N, Satoh S, Tanaka R, Tanaka A (2004) Domain structures of chlorophyllide a oxygenase of 772 green plants and Prochlorothrix hollandica in relation to catalytic functions. Planta 773 218:1019–1025. https://doi.org/10.1007/s00425-003-1181-6 774 Nakagawara E, Sakuraba Y, Yamasato A, et al (2007) Clp protease controls chlorophyll b synthesis 775 by regulating the level of chlorophyllide a oxygenase. The Plant Journal 49:800–809. 776 https://doi.org/10.1111/j.1365-313X.2006.02996.x 777 Neale PJ, Priscu JC (1995) The Photosynthetic Apparatus of Phytoplankton from a Perennially Ice-778 Covered Antarctic Lake: Acclimation to an Extreme Shade Environment. Plant and Cell 779 Physiology 36:253–263. https://doi.org/10.1093/oxfordjournals.pcp.a078757 780 Neilson JAD, Durnford DG (2010) Structural and functional diversification of the light-harvesting 781 complexes in photosynthetic eukaryotes. Photosynthesis Research 106:57–71. 782 https://doi.org/10.1007/s11120-010-9576-2 783 Nick S, Meurer J, Soll J, Ankele E (2013) Nucleus-encoded light-harvesting chlorophyll a/b proteins 784 are imported normally into chlorophyll b-free chloroplasts of arabidopsis. Molecular Plant 785 6:860–871. https://doi.org/10.1093/mp/sss113 786 Oster U, Tanaka R, Tanaka A, Rüdiger W (2000) Cloning and functional expression of the gene 787 encoding the key enzyme for chlorophyll b biosynthesis (CAO) from Arabidopsis thaliana. 788 Plant Journal 21:305–310. https://doi.org/10.1046/j.1365-313X.2000.00672.x 789 Palenik B, Grimwood J, Aerts A, et al (2007) The tiny eukaryote Ostreococcus provides genomic 790 insights into the paradox of plankton speciation. Proceedings of the National Academy of 791 Sciences 104:7705–7710. https://doi.org/10.1073/pnas.0611046104 792 Palenik B, Haselkorn R (1992) Multiple evolutionary origins of prochlorophytes, the chlorophyllb-793 containing prokaryotes. Nature 355:265–267. https://doi.org/10.1038/355265a0 794 Panchy N, Lehti-Shiu M, Shiu S-H (2016) Evolution of Gene Duplication in Plants. Plant Physiol 795 171:2294–2316. https://doi.org/10.1104/pp.16.00523 796 Partensky F, Six C, Ratin M, et al (2018) A novel species of the marine cyanobacterium 797 Acaryochloris with a unique pigment content and lifestyle. Sci Rep 8:9142. 798 https://doi.org/10.1038/s41598-018-27542-7 799 Pattanayak GK, Biswal AK, Reddy VS, Tripathy BC (2005) Light-dependent regulation of chlorophyll 800 b biosynthesis in chlorophyllide a oxygenase overexpressing tobacco plants. Biochemical 801 and Biophysical Research Communications 326:466–471. 802 https://doi.org/10.1016/j.bbrc.2004.11.049 803 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 26 Perroud PF, Haas FB, Hiss M, et al (2018) The Physcomitrella patens gene atlas project: large-scale 804 RNA-seq based expression data. Plant Journal 95:168–182. 805 https://doi.org/10.1111/tpj.13940 806 Piatkowski BT, Shaw AJ (2019) Functional trait evolution in Sphagnum peat mosses and its 807 relationship to niche construction. New Phytologist 223:939–949. 808 https://doi.org/10.1111/nph.15825 809 Poirier MC, Fugard K, Cvetkovska M (2025) Light quality affects chlorophyll biosynthesis and 810 photosynthetic performance in Antarctic Chlamydomonas. Photosynth Res 163:9. 811 https://doi.org/10.1007/s11120-024-01127-0 812 Pombert J-F, Blouin NA, Lane C, et al (2014) A Lack of Parasitic Reduction in the Obligate Parasitic 813 Green Alga Helicosporidium. PLoS Genet 10:e1004355. 814 https://doi.org/10.1371/journal.pgen.1004355 815 Pružinská A, Tanner G, Anders I, et al (2003) Chlorophyll breakdown: Pheophorbide a oxygenase is 816 a Rieske-type iron–sulfur protein, encoded by the accelerated cell death 1 gene. 817 Proceedings of the National Academy of Sciences 100:15259–15264. 818 https://doi.org/10.1073/pnas.2036571100 819 Przybyla-Toscano J, Boussardon C, Law SR, et al (2021) Gene atlas of iron-containing proteins in 820 Arabidopsis thaliana. The Plant Journal 106:258–274. https://doi.org/10.1111/tpj.15154 821 Reinbothe C, Bakkouri ME, Buhr F, et al (2010) Chlorophyll biosynthesis: spotlight on 822 protochlorophyllide reduction. Trends in Plant Science 15:614–624. 823 https://doi.org/10.1016/j.tplants.2010.07.002 824 Reinbothe C, Bartsch S, Eggink LL, et al (2006) A role for chlorophyllide a oxygenase in the regulated 825 import and stabilization of light-harvesting chlorophyll a/b proteins. Proceedings of the 826 National Academy of Sciences of the United States of America 103:4777–4782. 827 https://doi.org/10.1073/pnas.0511066103 828 Roche JL, van der Staay GWM, Partensky F, et al (1996) Independent evolution of the 829 prochlorophyte and green plant chlorophyll a/b light-harvesting proteins. Proceedings of 830 the National Academy of Sciences 93:15244–15248. 831 https://doi.org/10.1073/pnas.93.26.15244 832 Roesler CS, Culbertson CW, Etheridge SM, et al (2002) Distribution, production, and ecophysiology 833 of Picocystis strain ML in Mono Lake, California. Limnology and Oceanography 47:440–452. 834 https://doi.org/10.4319/lo.2002.47.2.0440 835 Roy P, Sethi S, New J, et al (2025) Transcriptomic Response of White Lupin Roots to Short-Term 836 Sucrose Treatment. Plants 14:14–17. https://doi.org/10.3390/plants14030381 837 Ruggiero MA, Gordon DP, Orrell TM, et al (2015) A Higher Level Classification of All Living 838 Organisms. PLoS One 10:e0119248. https://doi.org/10.1371/journal.pone.0119248 839 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 27 Sakuraba Y, Tanaka R, Yamasato A, Tanaka A (2009) Determination of a chloroplast degron in the 840 regulatory domain of chlorophyllide a oxygenase. Journal of Biological Chemistry 841 284:36689–36699. https://doi.org/10.1074/jbc.M109.008144 842 Sánchez-Baracaldo P, Cardona T (2020) On the origin of oxygenic photosynthesis and 843 Cyanobacteria. New Phytologist 225:1440–1446. https://doi.org/10.1111/nph.16249 844 Satoh S, Ikeuchi M, Mimuro M, Tanaka A (2001) Chlorophyll b Expressed in Cyanobacteria 845 Functions as a Light-harvesting Antenna in Photosystem I through Flexibility of the Proteins. 846 The Journal of Biological Chemistry 276:4293–4297. 847 https://doi.org/10.1074/jbc.M008238200 848 Schmidt CL, Shaw L (2001) A Comprehensive Phylogenetic Analysis of Rieske and Rieske-Type 849 Iron-Sulfur Proteins. J Bioenerg Biomembr 33:9–26. 850 https://doi.org/10.1023/A:1005616505962 851 Schoch CL, Ciufo S, Domrachev M, et al (2020) NCBI Taxonomy: a comprehensive update on 852 curation, resources and tools. Database (Oxford) 2020:baaa062. 853 https://doi.org/10.1093/database/baaa062 854 Schumacher I, Menghini D, Ovinnikov S, et al (2022) Evolution of chlorophyll degradation is 855 associated with plant transition to land. The Plant Journal 109:1473–1488. 856 https://doi.org/10.1111/tpj.15645 857 Shang S, Zhang Z, Li L, et al (2023) Transcriptome analysis reveals genes expression pattern of 858 Spirodela polyrhiza response to heat stress. International Journal of Biological 859 Macromolecules 225:767–775. https://doi.org/10.1016/j.ijbiomac.2022.11.139 860 Soltis PS, Soltis DE (2021) Plant genomes: Markers of evolutionary history and drivers of 861 evolutionary change. PLANTS, PEOPLE, PLANET 3:74–82. 862 https://doi.org/10.1002/ppp3.10159 863 Stamatakis A (2014) RAxML version 8: a tool for phylogenetic analysis and post-analysis of large 864 phylogenies. Bioinformatics 30:1312–1313. https://doi.org/10.1093/bioinformatics/btu033 865 Steadman Tyler CR, Hovde BT, Daligault HE, et al (2019) High-Quality Draft Genome Sequence of 866 the Green Alga Tetraselmis striata (Chlorophyta) Generated from PacBio Sequencing. 867 Microbiology Resource Announcements 8:10.1128/mra.00780-19. 868 https://doi.org/10.1128/mra.00780-19 869 Stolzer M, Lai H, Xu M, et al (2012) Inferring duplications, losses, transfers and incomplete lineage 870 sorting with nonbinary species trees. Bioinformatics 28:i409–i415. 871 https://doi.org/10.1093/bioinformatics/bts386 872 Suzuki S, Hirakawa Y, Kofuji R, et al (2016) Plastid genome sequences of Gymnochlora stellata, 873 Lotharella vacuolata, and Partenskyella glossopodia reveal remarkable structural 874 conservation among chlorarachniophyte species. J Plant Res 129:581–590. 875 https://doi.org/10.1007/s10265-016-0804-5 876 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 28 Talavera G, Castresana J (2007) Improvement of phylogenies after removing divergent and 877 ambiguously aligned blocks from protein sequence alignments. Syst Biol 56:564–577. 878 https://doi.org/10.1080/10635150701472164 879 Tanaka A, Ito H, Tanaka R, et al (1998) Chlorophyll a oxygenase (CAO) is involved in chlorophyll b 880 formation from chlorophyll a. Proceedings of the National Academy of Sciences of the 881 United States of America 95:12719–12723. https://doi.org/10.1073/pnas.95.21.12719 882 Tanaka A, Tanaka R (2019) The biochemistry, physiology, and evolution of the chlorophyll cycle, 1st 883 edn. Elsevier Ltd. 884 Tanaka R, Tanaka A (2007) Tetrapyrrole biosynthesis in higher plants. Annual Review of Plant 885 Biology 58:321–346. https://doi.org/10.1146/annurev.arplant.57.032905.105448 886 Tanaka R, Tanaka A (2005) Effects of chlorophyllide a oxygenase overexpression on light 887 acclimation in Arabidopsis thaliana. Photosynth Res 85:327–340. 888 https://doi.org/10.1007/s11120-005-6807-z 889 Tanaka R, Tanaka A (2011) Chlorophyll cycle regulates the construction and destruction of the 890 light-harvesting complexes. Biochimica et Biophysica Acta (BBA) - Bioenergetics 1807:968–891 976. https://doi.org/10.1016/j.bbabio.2011.01.002 892 Tartar A, Boucias DG, Becnel JJ, Adams BJ (2003) Comparison of plastid 16S rRNA (rrn16) genes 893 from Helicosporidium spp.: evidence supporting the reclassification of Helicosporidia as 894 green algae (Chlorophyta). Int J Syst Evol Microbiol 53:1719–1723. 895 https://doi.org/10.1099/ijs.0.02559-0 896 Tomitani A, Okada K, Miyashita H, et al (1999a) Chlorophyll b and phycobilins in the common 897 ancestor of cyanobacteria and chloroplasts. Nature 400:159–162. 898 https://doi.org/10.1038/22101 899 Tomitani A, Okada K, Miyashita H, et al (1999b) Chlorophyll b and phycobilins in the common 900 ancestor of cyanobacteria and chloroplasts. Letters to nature 400:159–162 901 Turner-Hissong SD, Mabry ME, Beissinger TM, et al (2020) Evolutionary insights into plant breeding. 902 Current Opinion in Plant Biology 54:93–100. https://doi.org/10.1016/j.pbi.2020.03.003 903 Ueno Y, Aikawa S, Kondo A, Akimoto S (2019) Adaptation of light-harvesting functions of unicellular 904 green algae to different light qualities. Photosynthesis Research 139:145–154. 905 https://doi.org/10.1007/s11120-018-0523-y 906 Voitsekhovskaja OV, Tyutereva EV (2015) Chlorophyll b in angiosperms: Functions in 907 photosynthesis, signaling and ontogenetic regulation. Journal of Plant Physiology 189:51–908 64. https://doi.org/10.1016/j.jplph.2015.09.013 909 Wang B, Yang X, Jia Y, et al (2022) High-quality Arabidopsis thaliana Genome Assembly with 910 Nanopore and HiFi Long Reads. Genomics Proteomics Bioinformatics 20:4–13. 911 https://doi.org/10.1016/j.gpb.2021.08.003 912 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint 29 Wang W, Wang L, Wang L, et al (2021) Transcriptome analysis and molecular mechanism of linseed 913 (Linum usitatissimum L.) drought tolerance under repeated drought using single-molecule 914 long-read sequencing. BMC Genomics 22:1–23. https://doi.org/10.1186/s12864-021-915 07416-5 916 Willows RD (2020) Biosynthesis of Chlorophyll and Bilins in Algae. In: Larkum AWD, Grossman AR, 917 Raven JA (eds) Advances in Photosynthesis and Respiration, 45th edn. Springer Nature, pp 918 83–103 919 Worden AZ, Lee J-H, Mock T, et al (2009) Green Evolution and Dynamic Adaptations Revealed by 920 Genomes of the Marine Picoeukaryotes Micromonas. Science 324:268–272. 921 https://doi.org/10.1126/science.1167222 922 Xiong J, Bauer CE (2002) COMPLEX EVOLUTION OF PHOTOSYNTHESIS. Annual Review of Plant 923 Biology 53:503–521. https://doi.org/10.1146/annurev.arplant.53.100301.135212 924 Xiong J, Wen G, Song J, et al (2024) Knockout of the Chlorophyll a Oxygenase Gene OsCAO1 925 Reduces Chilling Tolerance in Rice Seedlings. Genes 15:721. 926 https://doi.org/10.3390/genes15060721 927 Xu H, Vavilin D, Vermaas W (2001) Chlorophyll b can serve as the major pigment in functional 928 photosystem II complexes of cyanobacteria. Proceedings of the National Academy of 929 Sciences of the United States of America 98:14168–14173. 930 https://doi.org/10.1073/pnas.251530298 931 Yamasato A, Nagata N, Tanaka R, Tanaka A (2005) The N-Terminal Domain of Chlorophyllide a 932 Oxygenase Confers Protein Instability in Response to Chlorophyll b Accumulation in 933 Arabidopsis. Plant Cell 17:1585–1597. https://doi.org/10.1105/tpc.105.031518 934 Zhang L, Yang C, Liu C (2023) Revealing the significance of chlorophyll b in the moss 935 Physcomitrium patens by knocking out two functional chlorophyllide a oxygenase. 936 Photosynthesis Research 158:171–180. https://doi.org/10.1007/s11120-023-01044-8 937 Zhu X, Tai X, Ren Y, et al (2019) Genome-wide analysis of coding and long non-coding RNAs 938 involved in cuticular wax biosynthesis in cabbage (Brassica oleracea L. var. capitata). 939 International Journal of Molecular Sciences 20:. https://doi.org/10.3390/ijms20112820 940 Zhuang Y, Wang X, Li X, et al (2022) Phylogenomics of the genus Glycine sheds light on polyploid 941 evolution and life-strategy transition. Nat Plants 8:233–244. 942 https://doi.org/10.1038/s41477-022-01102-4 943 944 945 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 18, 2025. ; https://doi.org/10.1101/2025.02.10.637544doi: bioRxiv preprint