Introduction As the closest algal relatives of land plants [ 1 – 6 ], the Zygnematophyceae have attracted considerable attention from the plant evo-devo community in recent years [ 7 – 9 ]. These efforts are underpinned by genome sequences of selected species [ 10 – 15 ]. The class Zygnematophyceae is the most species-rich of all streptophyte algae [ 16 ] and was established by Round [ 17 ] and taxonomically validated [ 18 ].
Traditionally, the class was divided into two orders, the filamentous Zygnematales and the unicellular Desmidiales, based on the morphology of vegetative cells [ 18 , 19 ]. Recently, based on a phylogenomic tree of 326 loci, Zygnematophyceae were divided into a five-order system [ 16 ] (Desmidiales, Spirogyrales, Zygnematales, Serritaeniales, and Spirogloeales). While each order was well-supported, the phylogenetic relationship among them remained open and orders were in some cases represented by low genetic diversity and/or limited taxon sampling.
To address this, we used a systematic sampling approach, aiming for covering increased phylodiversity in Zygnematophyceae across taxonomic groups, growth types, habitats, and culture collections. We obtained 43 strains from which we generated de novo transcriptomes. Combined with publicly available data, we assembled a dataset of 104 zygnematophyceaen algae, nine other streptophyte algae, eight land plants, and four chlorophyte algae as outgroup.
Building on our previously established methodological approaches [ 20 , 21 ] and using IQ-Tree v3 we computed a phylogenomic tree based on 2243 loci and a site-heterogeneous best-fit model [ 22 ] with Posterior Mean Site Frequency (PMSF) [ 23 ] profiling (LG + C60 + F + G-PMSF) ( Fig 1 ). 10.1371/journal.pbio.3004005.g001 Fig 1 A robust phylogenomic framework and deep genetic structure within Zygnematophyceae. (A) A maximum likelihood phylogenomic tree based on 2243 loci.
The Zygnematophyceae are divided into five orders: (i) Spirogloeales, (ii) Serritaeniales, (iii) Zygnematales, (iv) Spirogyrales, and (v) Desmidiales. Filled dots mark full support and the colors correspond to the labels of major monophyla. (B) Light micrographs showcasing the newly sequenced diversity: (i) Mesotaenium kramstae NIES-657; (ii) Zygnemopsis minutum ACOI-60; (iii) Mesotaenium caldariorum SAG 230-1; (iv) Mougeotia scalaris SAG 164.80; (v) Zygnema circumcarinatum SAG 698-1b; (vi) Zygnema sp.
SAG 698-4; (vii) Zygnema cf. cylindricum SAG 698-1a; (viii) Zygnema cylindricum SAG 698-2; (ix) Zygnema pseudogedeanum SAG 38.96; (x) Zygnema peliosporum SAG 698-3a; (xi) Zygnema frigidum ACOI-2542; (xii) Zygnema colinsianum ACOI-1666; (xiii) Zygnema sp. SAG 2418; (xiv) Zygnema sp. SAG 2675; (xv) Spirogyra majuscula NIES-4308; (xvi) Sirogonium sticticum UTEX 1985; (xvii) Spirogyra tenuissima ACOI-1925; (xviii) Spirogyra dentireticulata NIES-4304; (xix) Spirogyra corrugata NIES-4303; (xx) Spirogyra sp .
SAG 170.80; (xxi) Spirogyra chunkingensis NIES-4302; (xxii) Spirogyra sp . SAG 169.80; (xxiii) Spirogyra fennica ACOI-1668; (xxiv) Spirogyra pratensis MZCH-10213. All scalebars except for those in panels iii and xv are 50 µm.
Scalebar iii represents 20 µm and the one in xv represents 200 µm. The data and code needed to generate this Figure can be found in https://zenodo.org/records/18392035 as well as https://doi.org/10.5281/zenodo.22172294 on Zenodo. Bold font highlights the new transcriptomes generated in this study.
The phylogenomic framework of Zygnematophyceae is robust Our tree recovered the well-accepted gross topology of the Chloroplastida, with deeply split monophyletic chlorophytes and streptophytes, six monophyletic streptophyte algal classes, and the zygnematophytes as sister to land plants [ 1 – 5 ] ( Fig 1 ). Our sampling effort focused on increasing the representation especially for the orders Spirogyrales and Zygnematales to improve the phylogenomic resolution of our analyses. The traditional order Zygnematales was split into three orders as in ref [ 16 ]: Spirogyrales, Serritaeniales, and Zygnematales (in revised form), which included the transfer of unicellular taxa ( Mesotaenium Naegeli, Cylindrocystis Meneghini ex De Bary) to these new orders.
Most of the other unicellular taxa remained in the Desmidiales. The cells of these Desmidiales, characteristic of a desmid cell, are typically unicellular and composed of two symmetrical semicells joined at a central isthmus. Each semicell generally contains a chloroplast, and the cells exhibit diverse shapes, ranging from simple elongate or spherical forms to highly ornamented, lobed, or deeply constricted morphologies.
Interestingly, the fifth order Spirogloeales, remains represented by a single species, Spirogloea muscicola (De Bary) Melkonian, which was originally assigned as Spirotaenia bryophila (Brébisson) Lütkemüller [ 10 ]. Overall, our phylogenomic study confirmed the five-order system [ 16 ], but Zygnematales and Serritaeniales were recovered as sister groups in our analyses based on increased sampling of taxa and genes. Additional sampling of taxa such as members of the genera Ancylonema Berggren , Zygogonium Kützing [ 24 ] , and organisms that resemble Mesotaenium (which likely is a polyphyletic genus), closely related based on preliminary rbc L phylogeny and genomic signatures [ 25 – 27 ], could further change the topology of the orders Zygnematales and Serritaeniales and/or modify the orders as currently defined.
Using the phylogenomic tree as a backbone, we performed ancestral character state reconstruction (ACSR) to infer the evolution of growth types across Zygnematophyceae ( Fig 2 ). For ACSR, we considered the morphological capabilities of all organisms when coding the characters, although these characters primarily reflect the capabilities of the vegetative cells. In Zygnematophyceae in particular, the life cycle does not influence their overall growth capabilities, as it comprises only vegetative cells and zygospores.
Based on the current sampling, of the seven transitions to filamentous growth, only two are ancient: in the last common ancestor (LCA) of Spirogyrales and the genus Zygnema —to both of which we turn in the following. 10.1371/journal.pbio.3004005.g002 Fig 2 Ancestral character state reconstruction of body plan evolution across Zygnematophyceae. Growth types were coded as five-character state distributions to trace the evolution of different homologies of growth types: salmon = unicellular, yellow = sarcinoid, green = bona fide filamentous growth, blue = chain-like filaments (desmids), and orange = complex multicellular macrophytes.
The pie charts present the posterior probabilities (PP) for an inferred growth type at a node. The inferred ancestral unicellular state for the LCA of all Zygnematophyceae (PP of 0.988) is highlighted as well as the two transitions to filamentous growth types that lead to the LCAs of the genus Zygnema (PP of 0.970) and the order Spirogyrales (PP of 0.950). Note that the focus here was on the Zygnematophyceae character states and results on any other lineage should be interpreted with caution.
Furthermore, note that a PP of ≥ 0.95 is generally considered as the standard threshold for a robust or strongly supported ancestral state at a given node and values ≥0.99 represent very high or decisive support. The data and code needed to generate this Figure can be found in https://zenodo.org/records/18392035 as well as https://doi.org/10.5281/zenodo.22172294 on Zenodo. Pronounced genetic divergence before the emergence of extant Spirogyrales The order Spirogyrales currently consists of the single genus Spirogyra Link.
Its most iconic feature is its spiral chloroplast(s), undulating through the entire length of the cell. From the almost 400 species described, the genome of one species ( Spirogyra pratensis Transeau) has been sequenced [ 15 ]. This single genome [ 15 ]—bolstered by the confirmation of patterns from previous draft sequencing data [ 28 ]—uncovered conspicuous deviations from other zygnematophytes, evident by genomic streamlining and pronounced genetic divergence described in Goldbecker and colleagues [ 15 ]; the most palpable example being the absence of the plastid division machinery otherwise conserved across Archaeplastida [ 15 ].
We thus used our phylogenomic framework to contextualize this divergence. The only reliable way of distinguishing the different Spirogyra species based on morphology is by observing the formation of zygotes (zygospore) and their characteristics [ 29 ]. While the induction of the sexual reproduction has not been reported on all strains—and thus Spirogyra strains are unidentified at the species level—reducing nutrient availability has succeeded in many cases [ 30 , 31 ].
Takano and colleagues [ 30 ] identified 13 species based on experiments inducing sexual reproduction. They sequenced two plastid-coding markers ( rbc L and atp B) and discovered seven clades (I-VII). Our phylogenomic tree recovered seven highly supported clades of Spirogyra spp., which correspond to those established by Takano and colleagues [ 30 ] with one exception: clade VI is missing because no strain of this clade is available in public culture collections.
A subdivision into eight clades was found using SSU rDNA sequences (clades A-H; Chen and colleagues [ 32 ]). Unfortunately, no strain of this study is available and almost none was assigned to a species level because of the lack of zygospores in culture. Both of these studies confirmed that the genus Sirogonium Kützing 1843 belonged to Spirogyra (clade II in Takano and colleagues [ 30 ] or clade C in Chen and colleagues [ 32 ]).
This genus differed from Spirogyra by 2–10 ribbon-like chloroplasts, a feature that derives from spiral chloroplast(s) based on phylogenetic and phylogenomic studies. In addition, one new clade was uncovered by our phylogenomic study containing three strains (NIES-4303 Sp. corrugata Transeau, NIES-4314 Sp. punctata Cleve and SAG 169.80 Spirogyra sp.). Both NIES strains were not assigned to any clade in the study by Takano and colleagues [ 30 ] (2019).
In the future, our phylogenomic backbone can be integrated with phylodiverse and detailed analyses of zygospore morphology, cell wall chemistry, and three-dimensional architecture [ 33 ]; this will allow resolving the functional significance, evolutionary variability, and adaptive roles of zygnematophycean zygospore walls, an area of research that remains in its infancy despite recent methodological advances and emerging evidence for previously unrecognized cell wall complexity [ 34 ]. Overall, our phylogenomic tree included a phylodiversity of 25 different Spirogyra taxa. They form a monophylum, the order Spirogyrales, that branches off from the sister order Desmidiales.
The branch leading to extant Spirogyrales is remarkably long (6.2 times longer than the branch leading to Desmidiales). This pronounced branch length may reflect several, non-mutually exclusive evolutionary and phylogenetic scenarios, including substantial extinction within Spirogyrales, the existence of unsampled and potentially undescribed diversity, or accelerated rates of molecular evolution along the lineage. Long branches may also arise from heterogeneous selection pressures, compositional biases, or biases associated with the use of transcriptome-derived data.
Thus, the current phylogenetic pattern does not by itself allow us to distinguish among these alternative explanations. If the long branch primarily reflects extinction and/or unsampled diversity, it would imply that substantial evolutionary history is missing between the common ancestor of extant Spirogyrales and its living representatives. Under this scenario, the emergence of characteristic Spirogyrales traits—including spiral chloroplasts, genomic streamlining, and distinctive cell-division machineries—may have occurred prior to the LCA of extant Spirogyrales and subsequently been retained across the surviving lineages.
This interpretation is consistent with the ACSR, which reconstructs ancient filamentous growth forms within Spirogyrales ( Fig 2 ). However, given the alternative explanations for the observed long branch, these trait-evolutionary inferences should be regarded as hypotheses that will require broader taxon sampling and improved genomic resources to test. Deep genetic structure and recent bursts of radiations in the genus Zygnema The genus Zygnema C.
Agardh is characterized by two asteroid-shaped chloroplasts, each with a single central pyrenoid; we recover them as ancestrally filamentous ( Fig 2 ). One hundred thirty-seven species are described based on features of zygospores [ 29 ]. Phylogenetic analyses of rbc L sequences revealed that all thus far investigated specimens were distributed into three clades [ 35 , 36 ].
Our phylogenomic analyses confirmed this subdivision. Three groups were pinpointed: Z. circumcarinatum -, Z. cylindricum - and Z. peliosporum -groups. The first clade contains only two species Z. circumcarinatum Czurda (SAG 698-1b) and Z. tunetanum (Gauthier-Lièvre) Stancheva, J.D.
Hall, McCourt and Sheath (CPCC-136) the latter originally assigned to the genus Zygogonium . The second clade consists of several species, some not assigned at the species level. Interestingly, the strain SAG 698-1a, originally assigned as the “b” mating type of Z. circumcarinatum (now Z. sp.
‘ cylindricum ’) belongs to this clade, which indicated a past mistransfer in algal collections [ 12 ]. Most of the investigated strains are members of the third clade. Linking molecular data with zygospore characteristics remains challenging in the Zygnematophyceae.
Although mesosporium color has been proposed as a diagnostic trait [ 35 ], exceptions exist and color variation may partly reflect zygospore maturation. To date, detailed ultrastructural characterization of zygospores combined with molecular identification, including Focused Ion Beam Scanning Electron Microscope (FIB-SEM) reconstruction, has been reported for only a limited number of taxa [ 33 , 37 ]. For example, Zygogonium vaginatum SAG 2418 [ 36 ] and Z. peliosporum cluster within subgroup 1 ( Fig 1 ), an attractive pair where unequivocal molecular identification can be integrated with morphological, ultrastructural, and biochemical analyses of zygospores.
Publicly available strains with abundant zygospore production, such as the Svalbard Zygnemopsis material described by Pichrtová and colleagues [ 38 ], represent valuable resources for future studies, whereas access to some divergent filamentous genera (e.g., Temnogametum [ 39 ]) remains more limited. The sister group of Zygnematales in our phylogenomic study is the order Serritaeniales [ 16 ]. This order includes (i) several taxa assigned to the genus Serritaenia A.Busch and S.Hess [ 25 ], (ii) the type species of the genus Mesotaenium , M. endlicherianum Naegeli, and (iii) Mougeotiopsis calospora Palla.
Interestingly, Mougeotiopsis represents the only filamentous taxon within this order so far. The other specimens of this order have unicellular Mesotaenium -like morphologies. The genus Mesotaenium is polyphyletic as currently described.
For example, M. pyrenoidosum (P.A.Broady) Petlovany and M. kramstae Lemmermann belong to different lineages of the Zygnematales. The diversity of Mesotaenium -like organisms is much higher than expected and awaits taxonomic revisions [ 25 ]. Our phylogenomic data recovers the genus Zygnema as monophyletic.
In previous rbc L phylogenies [ 25 ], Zygnema was placed in an unsupported larger group consisting of other species of Cylindrocystis , different lineages of Mesotaenium (Meso-9 and Meso-10), and Zygnemopsis (Skuja) Transeau. Our phylogenomic study now robustly resolves their relationships; three additional, well-supported lineages beside the Zygnema subclade are recovered: (i) Zygnemopsis together with Cylindrocystis cushleckae A.J. Brook and Mesotaenium kramstae , (ii) Mougeotia C.
Agardh and Mesotaenium caldariorum (Lagerheim) Hansgirg (Meso-9), (iii) Cylindrocystis brebissonii , (iv) and Mesotaenium pyrenoidosum (Meso-4) sister to all other Zygnematales. Both genera Zygnema and Zygnemopsis cannot be discerned by looking at the vegetative filaments, but by the presence of mucilage surrounding the gametangia in Zygnemopsis . Compared to Zygnema and Spirogyra , Zygnemopsis is a small genus consisting of 42 species [ 29 ].
In our analyses, the investigated strains are separated from Zygnema confirming that this morphological feature seems apt for differentiation of both genera if sexual stages are present. Unfortunately, no sexual reproduction was observed for the isolates included in the tree and therefore, no species assignment was possible. Based on the topology of our phylogenomic tree, we made two observations for the extant diversity represented by the genus Zygnema .
First, the clades within the genus are deeply divergent from one another—aligning with the molecular clock analysis of Feng and colleagues [ 12 ] that dated the divergence between Zygnema circumcarinatum and Zygnema cylindricum Transeau (see also [ 40 ]) to have occurred 236 ± 5 million years ago. Second, each clade experienced pronounced radiation, giving rise to the diversity captured by our dataset. Order-specific occurrence of signature genes shared by land plants and zygnematophytes Traditionally, other streptophyte algal lineages than Zygnematophyceae were considered as closest algal relatives of land plants.
This was owed to single-gene phylogenies and the complexity of body plans found in genera like Chara L. and Coleochaete Brébisson; yet, both in the case of Charophyceae and the discoidal Coloechaete species, the complexity in body plans are likely the result of secondary gains [ 21 , 41 ]; for a recent review on the multiple evolutionary emergences of (sometimes very complex) multicellular streptophyte bodies, see Darienko and colleagues [ 42 ]. The genomic revolution has not only phylogenomically established the sister group relationship of Zygnematophyceae and embryophytes [ 1 – 3 , 43 ] but comparative genomics pinpointed to several genetic synapomorphies in their ancestor [ 44 ]. Abscisic acid (ABA) is the foremost phytohormone in stress signaling, mounting a myriad of responses that adjust the biology of land plants to terrestrial challenges [ 45 ].
The genes homologous to those coding for the canonical ABA receptor [ 46 – 52 ] PYRABACTIN RESISTANCE1/PYR1-LIKE/REGULATORY COMPONENTS OF ABA RECEPTOR (PYR/PYL/RCAR). Homologous genes of PYR/PYL/RCAR were once considered to occur only among land plants; yet, with the advent of transcriptome [ 44 ] and genome [ 10 , 12 ] sequencing of zygnematophytes PYR/PYL/RCAR genes were found in these closest algal relatives of land plants and thus emerged as a potential molecular synapomorphy defining the clade of zygnematophytes and land plants. Among zygnematophytes, they are found in the genomes of Mesotaenium [ 10 ] and Zygnema [ 12 , 44 ] but not in that of the desmids Penium Brébisson ex Ralfs [ 11 ] and Closterium Nitzsch ex Ralfs [ 13 ].
Considering this patchy recovery of signature genes shared by land plants and zygnematophytes, coding capacity in genome space of the Zygnematophyceae has been insufficiently covered only by one or two genomes per order. We here made use of our largely extended sampling to explore the occurrence of PYR/PYL/RCAR homologs in zygnematophytes. Our maximum likelihood phylogeny recovers that the characterized [ 53 , 54 ] PYL homologs of Mesotaenium endlicherianum SAG 12.97 and Zygnema circumcarinatum SAG 698-1b fall into monophyla of PYL homologs of Serritaeniales and Zygnematales, respectively ( Fig 3A )—including homologs in, e.g., Zygnemopsis , which is within the order Zygnematales and distantly related to the model strains of Zygnema .
Yet, despite our vastly extended dataset, we did not detect homologs in any other order, suggesting potential order-specific losses of PYL (one in Spirogloeales and one in the common ancestor of Desmidiales and Spirogyrales); confirming this absence will require broader taxonomic sampling, supported by contiguous genome assemblies rather than relying largely on transcriptomic data. That being said, the lineage-specific occurrence of PYL in zygnematophytes aligns with the idea that extensive rewiring in the homologous chassis of PYL–PP2C–SnRK2 occurred—and that this likely does not constitute the core cascade [ 55 , 56 ]. This is further in line with our view on the modular assembly of other phytohormone signaling systems as in the case of auxin [ 57 – 63 ].
10.1371/journal.pbio.3004005.g003 Fig 3 Order-specific radiations in land plant heritage genes. Maximum likelihood phylogenies of protein-coding genes homologous to signature regulators of the molecular biology of land plants. (A) Phylogeny of PYRABACTIN RESISTANCE1/PYR1-LIKE/REGULATORY COMPONENTS OF ABA RECEPTOR ( PYR/PYL/RCAR ) homologs, radiating independently of the land plants subfamilies (cf.
Park and colleagues [ 50 ] and Zimran and colleagues [ 54 ]) in Zygnematophyceae and limited to the orders Serritaeniales and Zygnematales. (B) Phylogeny of CINNAMYL ALCOHOL DEHYDROGENASE ( CAD ), showing several order-specific clades and radiations; homologs were recovered across orders. (C) Phylogeny of histidine kinase homologs and their domain structure.
PFAM domains were annotated using interproscan and consolidated per clade. PF00072: Response regulator receiver domain (Response_reg), PF00512: His Kinase A (phosphor-acceptor) domain (HisKA), PF02518: Histidine kinase-, DNA gyrase B-, and HSP90-like ATPase (HATPase_c), PF03924: CHASE domain, PF24896: AHK4/CRE1/WOL first receiver domain (Receiver_CRE1), PF05231: MASE1, PF13185: GAF domain, PF00989: PAS fold (PAS), PF08447: PAS fold (PAS_3), PF08448: PAS fold (PAS_4), PF13188: PAS domain (PAS_8), PF13426: PAS domain (PAS_9). For the fully labeled trees, see S1 Fig .
The data and code needed to generate this Figure can be found in https://zenodo.org/records/18392035 as well as https://doi.org/10.5281/zenodo.22172294 on Zenodo. Another set of land plant heritage genes concerns the chassis of the phenylpropanoid pathway. The phenylpropanoid pathway is key to a myriad of aspects of embryophyte biology, leading to the production of compounds crucial for biotic and abiotic stress response [ 64 , 65 ] as well as structural compounds such as lignin [ 66 , 67 ].
Tracing its evolutionary origin to a chassis that emerged in algae is thus a long-standing question in the field [ 68 – 73 ]. Streptophyte algae not only have genes homologous to those coding for a series of enzymes that have assembled in land plants into the phenylpropanoid pathway, but in some cases these homologs have even a radiated repertoire [ 74 ]. CINNAMYL ALCOHOL DEHYDROGENASE ( CAD ) is a point in case [ 74 ].
CAD is the last step in monolignol biosynthesis, reducing diverse cinnamyl aldehydes into their corresponding cinnamyl alcohols [ 75 , 76 ]. Divergent homologs of CAD occur even in chlorophytes but underwent massive radiation in streptophytes [ 74 ]. Homologs of this NADP(H)-dependent enzyme have been classified into three to five different groups or classes.
Generally, bona fide [ 77 ] or class I CADs show a clear response to stress [ 78 – 81 ] with the highest catalytic activity [ 78 , 82 , 83 ] and with the full catalytic motif set [ 84 , 85 ]. The other groups or classes describe CAD-like isoforms with generally a lower (Class II) [ 78 , 82 , 83 , 86 ] to an even lower or no catalytic activity (Class III) [ 82 , 86 ] or other functions. We here use the classification from Saballos and colleagues [ 87 ], where group IV was described as monocot-specific.
None of the CAD homologs of Zygnematophyceae fall within the embryophyte classes, but instead form several monophyla, each containing a rich diversity ( Fig 3B ). This bolsters the notion that CAD was present in the LCA of zygnematophytes and land plants [ 73 , 74 , 88 ]. But as the function and structures of CAD homologs vary between the classes, it remains to be experimentally determined which substrates the zygnematophyte CAD homologs accept and under which kinetics they catalyze them.
Protein structure predictions hint that streptophyte algal CADs are more like sinapyl alcohol dehydrogenases (SADs) [ 70 ]. However, given that (a) CAD homologs are differentially regulated in diverse zygnematophytes exposed to stressors [ 89 , 90 ] and (b) radiated in zygnematophytes, CAD seems to be a potential key player contributing to the chemodiversity of zygnematophytes. Phosphorelay-based signaling is key in molecular physiological adjustments of land plants.
In this response system, histidine kinases homologous to the ARABIDOPSIS HISTIDINE KINASES (AHK) are major players, acting in cytokinin-based signaling [ 91 – 93 ], their connections to stress response [ 94 ], and frequent responders in transcriptome data of stressed zygnematophytes [ 90 , 95 – 97 ]. Histidine kinases are ancient signaling proteins that originated in bacteria as part of two-component systems, which typically consist of a sensor histidine kinase and a response regulator. The diversity of histidine kinases expanded in streptophytes, including streptophyte algae; for example, analyses of the Klebsormidium nitens genome revealed a surprisingly large repertoire of HKs (e.g., ~67 in K. nitens ) [ 98 ].
While histidine kinase family underwent significant radiation and functional specialization, marked by the emergence of distinct subfamilies with specialized roles such as cytokinin perception (CHKs), ethylene and ABA Signaling (ETR-HKs), many HK groups critical and potentially unique to land plants, such as AHK1, CKR, and ETR, have homologs in streptophyte algae. Our maximum likelihood phylogenetic analysis recovers massive expansions in zygnematophyte AHKs in a lineage-dependent manner ( Fig 3C ). All clades in the gene tree contain proteins with Response regulator receiver domains (Response_reg), His Kinase A (phosphor-acceptor) domains (HisKA) and Histidine kinase-, DNA gyrase B-, and HSP90-like ATPase Histidine Kinase domains (HATPase_c).
Yet, we also recover AHK1 and AHK5 as clear orthologous monophyla, whilst there are several co-orthologs to AHK2,3,4. In Arabidopsis , AHK2, 3 and 4 bind cytokinin via their extracellular CHASE domain [ 92 , 99 , 100 ]. Next to the land plant-specific clade there are large expansions in zygnematophyte histidine kinases containing CHASE domains.
We identified several zygnemtatophyte histidine kinases that contain a PAS fold or domain. These domains are widespread across the tree of life and play a role in light sensing in plants [ 101 ]. PAS domain-containing histidine kinases that are not phytochromes also exist in Physcomitrium patens , other non-seed plants as well as in Klebsormidium nitens ; in Physcomitrium they are regulated by light and play a role in gametophore development [ 102 ].
We recovered three of the five Klebsormidum PAS-kinases identified by Ryo and colleagues [ 102 ] in our dataset within the zygnematophyte PAS kinase clade. Our analysis shows that PAS-domain-containing histidine kinases are abundant and likely to be important in zygnematophytes. We also identified histidine kinases with PAS domains in the clade containing AHK5.
The majority of the PAS domains correspond to the zygnematophyte sequences within that clade. On balance, Zygnematophyceae have a rich and diversified repertoire of AHKs, likely speaking of their extensive intracellular signaling. Overall, our data capture the shared occurrence of what has been coined [ 41 ] ‘land plant heritage genes’—genes that constitute a toolkit key for signature traits of land plants.
However, this shared molecular toolkit is, considering loss and divergence, only graspable by using a phylodiverse dataset [ 14 ]. With our vastly extended sampling of functional genomic space, several order-dependent patterns of loss and expansion can be gleaned—simultaneously bolstering the need for further whole-genome sequencing efforts into zygnematophyte diversity. Concluding remarks: Dynamic trait evolution in zygnematophytes We inferred a robust phylogenomic framework for zygnematophytes, the closest algal relatives of land plants that pays homage to their species richness.
While this represents a comprehensive dataset, it still captures a fraction of the diversity of more than 4,000 species, including hyperdiverse Desmidiales and unculturable extremophilic zygnematophytes [ 103 ]. That said, our framework recovers pronounced deep genetic divergences within Zygnematophyceae. This is not only evident from the genetic distances in the tree, but through the order-specific occurrence and within-order radiations, of genes for signature traits, and strong reductive evolutionary forces acting on the genomes [ 14 ].
This is mirrored by the diversity in cellular features and growth patterns, including repeated shifts between uni- and multicellular body plans [ 12 , 16 ], division patterns [ 104 ], plastid morphology [ 15 ], and more. Our data bolster the concept that the LCA of Zygnematophyceae was an ancestral unicell ( Fig 2 ), which our analyses recover with a posterior probability of 0.988 (the body plan of the LCA of land plants and Zygnematophyceae could not be robustly inferred). Transitions to multicellularity occurred several times independently, with a few ancient transitions in the LCAs of classical filamentous genera such as Spirogyra , Zygnema , and Mougeotia ( Fig 2 ).
Such multiple transitions to multicellularity suggest that they are underpinned by modes of rapid evolution—conceivably through the rewiring of existing components, facilitated by changes in regulation of gene expression and protein–protein interaction network [ 42 ]. This is further complicated by zygnematophytes not only including bona fide filaments but also the chain-like filaments of desmid cells that are brought about by diverse and complex modes of division [ 105 ]; such chain-like filaments of desmid cells consist of individual, typically unicellular desmid cells arranged end-to-end, forming elongated, unbranched chains. Adjacent cells remain connected following cell division, while retaining their characteristic desmid morphology, often with distinct semicells separated by a median constriction.
The homology of the cytokinetic mechanisms that bring about this diverse cell biology remains obscure. Overall, we infer seven transitions to filamentous body plans, to which broader sampling of diversity will likely add more. A comparative and complementing phylodiverse approach that accounts for loss and lability [ 14 ] (some of which might be environmentally inducible [ 106 ]) is thus needed to reliably define synapomorphies of the LCA of algae and all land plants.
Methods Resource table 10.1371/journal.pbio.3004005.t001 REAGENT or RESOURCE SOURCE IDENTIFIER Critical Commercial Assays DNAse I Thermo Fisher, Waltham, MA, USA N/A Sigma’s Spectrum Plant Total RNA Kit Sigma, Saint Louis, MO, USA N/A Deposited Data Amborella trichopoda *only for gene family analysis Amborella Genome Project [ 107 ] https://www.ncbi.nlm.nih.gov/bioproject/212863 Anthoceros agrestis BONN Li and colleagues [ 108 ] https://www.hornworts.uzh.ch/static/jbrowse/?data=a_agr_bonn Anthoceros agrestis OXF Li and colleagues [ 108 ] https://www.hornworts.uzh.ch/static/jbrowse/?data=a_agr_oxford Anthoceros punctatus *only for gene family analysis Li and colleagues [ 108 ] https://www.hornworts.uzh.ch/en.html Arabidopsis thaliana genome TAIR V10 TAIR [ 109 ] http://www.arabidopsis.org Arabidopsis lyrata *only for gene family analysis Hu and colleagues [ 110 ] https://phytozome-next.jgi.doe.gov/info/Alyrata_v2_1 Azolla filiculoides Galgenwaard genome Li and colleagues [ 111 ] https://www.fernbase.org Bambusina borreri CCAC 0045 1KP Code QWFV Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Bathycoccus prasinos Moreau and colleagues [ 113 ] https://bioinformatics.psb.ugent.be/orcae/overview/Bathy Brachypodium distachyon Bd21 The International Brachypodium Initiative [ 114 ] https://phytozome-next.jgi.doe.gov/info/Bdistachyon_v3_1 Brassica oleracea *only for gene family analysis Liu and colleagues [ 115 ] https://phytozome-next.jgi.doe.gov/info/Boleraceacapitata_v1_0 Brassica rapa *only for gene family analysis Goodstein and colleagues [ 116 ] https://phytozome-next.jgi.doe.gov/info/BrapaFPsc_v1_3 Capsella grandiflora *only for gene family analysis Slotte and colleagues [ 117 ] https://phytozome-next.jgi.doe.gov/info/Cgrandiflora_v1_1 Carica papaya *only for gene family analysis Ming and colleagues [ 118 ] https://phytozome-next.jgi.doe.gov/info/Cpapaya_ASGPBv0_4 Ceratopteris richardii *only for gene family analysis Marchant and colleagues [ 119 ] https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA729743 Chaetosphaeridium globosum SAG 26.98 Cooper and Delwiche [ 120 ] https://www.ncbi.nlm.nih.gov/sra/SRR16849197 Chara braunii S276 Nishiyama and colleagues [ 41 ] https://bioinformatics.psb.ugent.be/orcae/overview/Chbra Chlamydomonas reinhardtii CC-503 Merchant and colleagues [ 121 ] Blaby and colleagues [ 122 ] https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Creinhardtii Chlorella vulgaris CCAP 211/11P Cecchin and colleagues [ 123 ] https://www.ncbi.nlm.nih.gov/Traces/wgs/wgsviewer.cgi?val=SIDB00000000 Chlorokybus cerffii SAG 34.98 de Vries and colleagues [ 44 ] Irisarri and colleagues [ 124 ] https://datadryad.org/stash/dataset/doi:10.5061/dryad.0gb5mkm25 Chlorokybus melkonianii CCAC 0220 *only for gene family analysis Wang and colleagues [ 125 ] https://db.cngb.org/search/assembly/CNA0002352/ Closterium lunula M2156 1KP Code DRFX Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Closterium peracerosum-strigosum-littorale NIES-64 (mt-) Kawaguchi YQ and colleagues [ 126 ] https://www.ncbi.nlm.nih.gov/sra/?term=ERR10800206 - ERR10800208 Closterium peracerosum-strigosum-littorale NIES-65 (mt+) Kawaguchi YQ and colleagues [ 126 ] https://www.ncbi.nlm.nih.gov/sra/?term=ERR10800221 - ERR10800223 Closterium peracerosum–strigosum–littorale NIES-67 *only for gene family analysis Sekimoto and colleagues [ 13 ] https://www.ncbi.nlm.nih.gov/datasets/genome/GCA_027943415.1/ Closterium peracerosum–strigosum–littorale NIES-68 *only for gene family analysis Sekimoto and colleagues [ 13 ] https://www.ncbi.nlm.nih.gov/datasets/genome/GCA_027943385.1/ Coccomyxa subellipsoidea NIES-2166 Blanc and colleagues [ 127 ] https://www.ncbi.nlm.nih.gov/Traces/wgs/wgsviewer.cgi?val=AGSI00000000.1 Coleochaete scutata UTEX 2567 (mislabeled as Coleochaete orbicularis UTEX 2651) [ 21 ] Cooper and Delwiche [ 120 ] https://figshare.com/articles/dataset/Green_algal_transcriptomes_for_phylogenetics_and_comparative_genomics/1604778 Coleochaete scutata SAG 50.90 transcriptome de Vries and colleagues [ 44 ] https://www.ncbi.nlm.nih.gov/Traces/wgs/wgsviewer.cgi?val=GFXZ00000000 Cosmarium broomei CCAC 0143 1KP Code HIDG Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium crenatum MZCH-561 Mundt and colleagues [ 128 ] https://www.ncbi.nlm.nih.gov/sra/?term=ERR3005824 and ERR3005825 Cosmarium granatum CCAC 0137 1KP Code MNNM Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium ochthodes CCAC 1205 1KP Code HJVM Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium ochthodes M1384 1KP Code STKJ Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium subtumidum CCAC 3067 1KP Code WDGV Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium tinctum CCAC 0218 1KP BHBK Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cosmarium turpinii UTEX 733 Jackrel and colleagues [ 129 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR3478572 , SRR3478573, SRR3478589 Cosmocladium cf. constrictum CCAC 0217 1KP WDGV Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cylindrocystis brebissonii CCAC 0211 1KP YOXI Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cylindrocystis brebissonii CCAC 2853 1KP YLBK Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cylindrocystis brebissonii SAG 615-1 This Study Cylindrocystis cushleckae CCAC 0213 1KP JOJQ Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Cylindrocystis sp. CCAC 0212 1KP code VAZE Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Desmidium aptogonum CCAC 2329 1KP DFDS Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Euastrum humerosum var. affine CCAC 1175 1KP GYRP Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Gnetum montanum *only for gene family analysis Wan and colleagues [ 130 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA339497/ Gonatozygon brebissonii GB138 Su and colleagues [ 131 ] SRR12975768 Gonatozygon kinahani CCAC 0216 Carpenter and colleagues [ 112 ] ERR2041121 Gossypium hirsutum v2.1 *only for gene family analysis Li and colleagues [ 132 ] Chen and colleagues [ 133 ] https://phytozome-next.jgi.doe.gov/info/Ghirsutum_v2_1 Isoetes taiwanensis Kuo4500 Wickell and colleagues [ 134 ] https://genomevolution.org/coge/GenomeInfo.pl?gid=61511 and https://www.ncbi.nlm.nih.gov/bioproject/735564 Klebsormidium nitens NIES-2285 Hori and colleagues [ 135 ] https://phycocosm.jgi.doe.gov/Klenit1_1/Klenit1_1.home.html Marchantia polymorpha genome v3.1 Bowman and colleagues [ 136 ] https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Mpolymorpha Maximum likelihood phylogeny ( Fig 1 ) This Study https://zenodo.org/records/18392035 Mesostigma viride CCAC 1140 and NIES-296 Wang and colleagues [ 125 ] https://db.cngb.org/search/assembly/CNA0002352/ Mesotaenium caldariorum SAG 150.80 This Study Mesotaenium caldariorum SAG 230-1 This Study Mesotaenium caldariorum SAG 648-1 1KP Code HKZW Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Mesotaenium endlicherianum SAG 12.97 Gene models V2 Cheng and colleagues [ 10 ] Dadras and colleagues [ 90 ] Rieseberg and colleagues [ 137 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA541331/ https://www.ncbi.nlm.nih.gov/sra/?term=SRS17311714 , SRS17311718, SRS17311720-SRS17311725 Mesotaenium kramstae UTEX 1025 1KP Code NBYP Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Mesotaenium pyrenoidosum SAG 1.88 This Study Micrasterias fimbriata CCAC 1188 1KP Code MCHJ Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Micromonas pusilla *only for gene family analysis Worden and colleagues [ 138 ] https://phycocosm.jgi.doe.gov/MicpuC3v2/MicpuC3v2.home.html Micromonas sp. *only for gene family analysis Worden and colleagues [ 138 ] https://phycocosm.jgi.doe.gov/MicpuN3v2/MicpuN3v2.home.html Mougeotia scalaris SAG 164.80 Cooper and Delwiche [ 120 ] https://figshare.com/articles/dataset/Green_algal_transcriptomes_for_phylogenetics_and_comparative_genomics/1604778 Mougeotia scalaris SAG 164.80 (Labeled wrongly as SAG 164.90) Cooper and Delwiche [ 120 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR16849218 Mougeotia sp. CCAC 0197 Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Mougeotia sp.
MZCH 240/ SVCK-240 de Vries and colleagues [ 89 ] Fürst-Jansen and colleagues [ 139 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA543475/ Mougeotiopsis calospora MZCH-580 Hess and colleagues [ 16 ] https://www.ncbi.nlm.nih.gov/sra/SRR19751296 Netrium digitus CCAC 0148 1KP Code FFGR Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Nicotiana tabacum TN90 *only for gene family analysis Sierro and colleagues [ 140 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA208209 Nitella mirabilis S040 Ju and colleagues [ 141 ] https://www.ncbi.nlm.nih.gov/Traces/wgs/wgsviewer.cgi?val=GBST01&search=GBST01000000&display=scaffolds Nucleotaenium eifelense CCAC 3006 1KP Code KMNX Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Onychonema laeve CCAC 0151 1KP Code GGWH Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Oryza sativa *only for gene family analysis Ouyang and colleagues [ 142 ] https://phytozome-next.jgi.doe.gov/info/Osativa_v7_0 Ostreococcus lucimarinus *only for gene family analysis Palenik and colleagues [ 143 ] https://phycocosm.jgi.doe.gov/Ost9901_3/Ost9901_3.home.html Penium exiguum CCAC 0142 1KP Code YSQT Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Penium margaritaceum SAG 22.82 Cooper and Delwiche [ 120 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR16849209 Penium margaritaceum Skd-8 = SAG 2640 Jiao and colleagues [ 11 ] http://bioinfo.bti.cornell.edu/cgi-bin/Penium/download.cgi Phymatodocis nordstedtiana SAG 47.89 1KP Code RPQV Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Physcomitrium patens Gransden Lang and colleagues [ 144 ] Bi and colleagues [ 145 ] https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Ppatens Picea abies *only for gene family analysis Nystedt and colleagues [ 146 ] https://www.treegenesdb.org/org/Picea-abies Planotaenium ohtanii CCAC 0202 1KP Code SNOX Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Pleurotaenium trabecula CCAC 0163 1KP Code MOYY Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Salvinia cucullata *only for gene family analysis Li and colleagues [ 111 ] https://fernbase.org/ftp/Salvinia_cucullata/ Selaginella moellendorffii *only for gene family analysis Banks and colleagues [ 147 ] https://phytozome-next.jgi.doe.gov/info/Smoellendorffii_v1_0 Serritaenia sp. CCAC 0155 1KP Code WSJO Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Serritaenia sp. SAG 2655 This Study Sirogonium sticticum UTEX 1985 This Study Solanum lycopersicum ITAG4.0 *only for gene family analysis The Tomato Genome Consortium [ 148 ] https://phytozome-next.jgi.doe.gov/info/Slycopersicum_ITAG4_0 Sphagnum fallax *only for gene family analysis Healey and colleagues [ 149 ] https://phytozome-next.jgi.doe.gov/info/Sfallax_v0_5 Spirogloea muscicola CCAC 0214 Cheng and colleagues [ 10 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA541068/ Spirogloea muscicola CCAC 0214 (OneKP) 1KP Code TPHT Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Spirogyra chenii NIES-4312 This Study Spirogyra chungkingensis NIES-4302 This Study Spirogyra corrugata NIES-4303 This Study Spirogyra dentireticulata NIES-4304 This Study Spirogyra fennica ACOI-1668 This Study Spirogyra hopeiensis NIES-4305 This Study Spirogyra longata NIES-4306 This Study Spirogyra longata NIES-4307 This Study Spirogyra majuscula NIES-4308 This Study Spirogyra majuscula NIES-4309 This Study Spirogyra minuticrassoidea NIES-4310 This Study Spirogyra mirabilis NIES-4311 This Study Spirogyra pratensis MZCH-10213 de Vries and colleagues [ 89 ] Goldbecker and colleagues [ 15 ] https://www.ncbi.nlm.nih.gov/bioproject/PRJNA543475/ Spirogyra pratensis UTEX 921 Cooper and Delwiche [ 120 ] https://figshare.com/articles/dataset/Green_algal_transcriptomes_for_phylogenetics_and_comparative_genomics/1604778 Spirogyra pratensis UTEX 928 Ju and colleagues [ 141 ] https://www.ncbi.nlm.nih.gov/Traces/wgs/wgsviewer.cgi?val=GBSM01000000 Spirogyra pseudomaxima NIES-4313 This Study Spirogyra punctata NIES-4314 This Study Spirogyra semiornata NIES-4315 This Study Spirogyra sp.
EDC SP-AU1 Cooper and Delwiche [ 120 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR16849204 Spirogyra sp. CCAC 0205 1KP Code TPHT Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Spirogyra sp. SAG 169.80 This Study Spirogyra sp.
SAG 170.80 This Study Spirogyra tenuissima ACOI-1925 This Study Spirogyra varians NIES-4316 This Study Spirotaenia minuta CCAC 2661 1KP Code NNHQ Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Staurastrum punctulatum SAG 679-1 Jackrel and colleagues [ 129 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR3478579 -SRR3478581 Staurastrum sebaldii CCAC 1129 1KP Code ISHC Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Staurastrum tetracerum SAG 7.94 Paul and colleagues [ 150 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR1174746 Staurodesmus convergens CCAC 2558 1KP Code WCQU Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Staurodesmus omearae CCAC 0751 Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Theobroma cacao *only for gene family analysis Argout and colleagues [ 151 ] Motamayor and colleagues [ 152 ] https://phytozome-next.jgi.doe.gov/info/Tcacao_v1_1 Triticum aestivum *only for gene family analysis The International Wheat Genome Sequencing Consortium [ 153 ] https://plants.ensembl.org/Triticum_aestivum/Info/Index Ulva mutabilis Føyn De Clerck and colleagues [ 154 ] https://bioinformatics.psb.ugent.be/orcae/overview/Ulvmu Volvox carteri Prochnik and colleagues [ 155 ] https://phycocosm.jgi.doe.gov/phycocosm/home/releases?flt=volvox+carteri Xanthidium antilopaeum CCAC 1229 1KP Code GBGT Carpenter and colleagues. https://db.cngb.org/onekp/ Zygnema cylindricum SAG 698-1a_XF *only for gene family analysis Feng and colleagues [ 12 ] https://phycocosm.jgi.doe.gov/Zygcyl6981a_1/Zygcyl6981a_1.home.html Zygnema circumcarinatum SAG 698-1b Feng and colleagues [ 12 ] Rieseberg and colleagues [ 137 ] https://phycocosm.jgi.doe.gov/Zygcir6981b_2/Zygcir6981b_2.home.html https://www.ncbi.nlm.nih.gov/sra/?term=SRR24166998 - SRR24167004, SRR24167011, SRR24167014-SRR24167023 Zygnema circumcarinatum UTEX 1559 *only for gene family analysis Feng and colleagues [ 12 ] https://phycocosm.jgi.doe.gov/Zygcir1559_1/Zygcir1559_1.home.html Zygnema circumcarinatum UTEX 1560 *only for gene family analysis Feng and colleagues [ 12 ] https://phycocosm.jgi.doe.gov/Zygcir1560_1/Zygcir1560_1.home.html Zygnema collinsianum ACOI-1666 This Study Zygnema collinsianum ACOI-2544 This Study Zygnema cylindricum SAG 698-2 This Study Zygnema extenue UTEX 1740 This Study Zygnema frigidum ACOI-2542 This Study Zygnema irregulare SAG 2674 This Study Zygnema irregulare SAG 2675 This Study Zygnema peliosporum SAG 698-3a This Study Zygnema peliosporum SAG 698-3b This Study Zygnema pseudogedeanum CCAC 0199 Hess and colleagues [ 156 ] https://www.ncbi.nlm.nih.gov/sra/?term=SRR8869038 Zygnema pseudogedeanum SAG 38.96 This Study Zygnema sp. Alga2O-2499 This Study Zygnema sp. SAG 2418 This Study Zygnema sp.
SAG 2419 This Study Zygnema sp. SAG 2642 This Study Zygnema sp. SAG 698-4 This Study Zygnema sp.
ZygnemaMat Rippin and colleagues [ 157 ] https://www.ncbi.nlm.nih.gov/sra?linkname=bioproject_sra_all&from_uid=498913 Zygnema tunetanum CPCC-136 This Study Zygnemopsis minutum ACOI-60 This Study Zygnemopsis sp. CCAP 699/1 1KP Code MFZO Carpenter and colleagues [ 112 ] https://db.cngb.org/onekp/ Experimental Models: Organisms/Strains Coleochaete scutata UTEX 610 Bierenbroodspot and colleagues [ 20 , 21 ] Software and Algorithms ApplyPPPFormat (apply Phylopypruner format) Bierenbroodspot and colleagues [ 20 , 21 ] https://github.com/mjbieren/ApplyPPPFormat BUSCO v.5.0.0 Seppey and colleagues [ 158 ] https://busco.ezlab.org ClipKIT Steenwyk and colleagues [ 159 ] https://github.com/JLSteenwyk/ClipKIT FASTQC Babraham Institute www.bioinformatics.babraham.ac.uk/projects/fastqc FilterPPPResult Bierenbroodspot and colleagues [ 20 , 21 ] https://github.com/mjbieren/FilterPPPResult GPDS (Get Positive DataSet) Bierenbroodspot and colleagues [ 20 , 21 ] https://github.com/mjbieren/GPDS Interproscan v5.77 Jones and colleagues [ 160 ] https://interproscan-docs.readthedocs.io/en/v5/index.html IQ-Tree v1.5.5 (for gene family analysis) Nguyen and colleagues [ 161 ] http://www.iqtree.org IQ-Tree2 v2.2.2.7 (pre-analysis) Minh and colleagues [ 162 ] IQ-Tree3 V3.0.1 (for final tree interference) Wong and colleagues [ 163 ] MAFFT v7.490 Katoh and Standley [ 164 ] https://mafft.cbrc.jp/alignment/software/ ModelFinder Kalyaanamoorthy and colleagues [ 22 ] http://www.iqtree.org/ModelFinder/ OrthoFinder v3.1.1 Emms and Kelly [ 165 , 166 ] https://github.com/OrthoFinder/OrthoFinder OSG (Orthogroup Sequence Grabber) Bierenbroodspot and colleagues [ 20 , 21 ] https://github.com/mjbieren/OrthoGroup_Sequence_Grabber PAML 4 Yang [ 167 ] http://abacus.gene.ucl.ac.uk/software/#phylogenetic-analysis-by-maximum-likelihood-paml Phytools Revell [ 168 ] https://cran.r-project.org/web/packages/phytools/index.html Posterior Mean Site Frequency Profiles Wang and colleagues [ 169 ] Implemented in IQ-Tree http://www.iqtree.org PREQUAL Whelan and colleagues [ 170 ] https://github.com/simonwhelan/prequal Tracer v1.7.1 Rambaut and colleagues [ 171 ] https://github.com/beast-dev/tracer/releases/tag/v1.7.1 Transcdecoder v.5.5.0 Brian J. Haas https://github.com/TransDecoder/TransDecoder/releases Trimal v1.4.rev15 Capella-Gutierrez and colleagues [ 172 ] http://trimal.cgenomics.org Trimmomatic v0.36 Bolger and colleagues [ 173 ] http://www.usadellab.org/cms/?page=trimmomatic Resource availability Lead contact.
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jan de Vries ( devries.jan@uni-goettingen.de ). Materials availability. This study did not generate new unique reagents.
Experimental model and subject details Algal strains. Strains were obtained from the Culture Collection of Algae at Göttingen University (SAG) [ 174 ], the microbial culture collection of the National Institute for Environmental Studies in Japan (NIES), the Coimbra Collection of Algae (ACOI), the Microalgae and Zygnematophyceae Collection Hamburg (MZCH), and the culture collection of algae at the University of Texas at Austin (UTEX). Method details Algal culturing.
All cultures were maintained for six weeks in 3NBBM or MiEB12 liquid medium (medium 26a in Schlösser [ 175 ], and medium 7 in Schlösser [ 176 ]) under standardized laboratory conditions (18 °C; 14:10 h light:dark cycle; 25–35 μmol photons m −2 s −1 ) using full-spectrum fluorescent lighting. Light microscopy. High-resolution images of the studied strains were obtained with an Olympus BX-60 microscope (Olympus, Japan) with DIC equipped with a ProgRes C14plus camera and the ProgRes CapturePro Software (version 2.9.01) (JENOPTIK AG, Jena, Germany).
All investigated strains were examined in week six of cultivation. RNA extraction. For the RNA extraction of the strains, 50 mL of 6-week-old liquid culture were centrifuged for 5 min at 20 °C and 11,000 rpm and the supernatant was removed.
Pellets were transferred into 1.5 ml BioMasherII tubes (Nippi, Japan) and each sample was disrupted using PowerMasherII (Nippi, Japan) after freezing in liquid nitrogen. RNA extraction was done using the Spectrum Plant Total RNA Kit (Sigma-Aldrich Chemie GmbH, Germany) according to the manufacturer’s instructions. DNAse I treatment (Thermo Fisher, Waltham, MA, USA) was applied to the RNA samples, and their integrity and concentration were assessed using a 1% agarose gel with a Midora green stain, and Nanodrop (Thermo Fisher), respectively.
The RNA samples were shipped on dry ice to Novogene (Munich, Germany). RNA sequencing, transcriptome assembly, and transcriptome data. At Novogene (Munich, Germany), RNA quality was verified using a Bioanalyzer (Agilent Technologies, USA).
Libraries were prepared using poly-A enrichment and directional mRNA library construction protocols. After the libraries were quality-checked, sequencing was conducted on an Illumina NovaSeq 6000 platform using dual-indexed Novogene adapters: 5′- AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATT-3′ for read 1 and 5′- GATCGGAAGAGCACACGTCTGAACTCCAGTCACGGATGACTATCTCGTATGCCGTCTTCTGCTTG-3′. We conducted transcriptome sequencing on 43 isolates belonging to different orders of the Zygnematophyceae.
All transcriptomes were assembled de novo with Trinity [ 177 ] v2.15.1, following adapter trimming with Trimmomatic [ 173 ] (--trimmomatic “ILLUMINACLIP:novogene_adapter_sequences.fa:2:30:10:2:keepBothReads LEADING:3 TRAILING:3 MINLEN:36″) and default settings. Splicing isoforms were collapsed into SuperTranscripts [ 178 ] using Trinity’s implementation. Completeness was assessed with BUSCO [ 158 ] v5.4.3 using the ‘eukaryota_odb10′ dataset, yielding an average completeness of 94.84%.
Protein-coding sequences were predicted with TransDecoder v5.5.0, retaining only the longest open reading frame per transcript (--single_best_only). Phylogenomic analysis: Preamble. Phylogenomic analyses and ancestral character state reconstruction were conducted following the framework outlined in Bierenbroodspot and colleagues [ 20 ], with the following methodological refinements.
For orthogroup selection, we utilized only the Phylogenetic Hierarchical Orthogroups at the tree root (N0.tsv file produced by OrthoFinder v3.1.1 [ 165 , 166 ]) in conjunction with Orthogroup Sequence Grabber (OSG) [ 20 , 21 ]. A streamlined version of the COGS software [ 20 , 21 ] was implemented, excluding one redundant MAFFT [ 164 ]/IQ-TREE [ 161 – 163 ] alignment step. Similarly, the final ClipKit [ 159 ] trimming step was omitted, as it did not enhance alignment quality under the revised workflow.
Phylogenomic analysis: Dataset construction for phylotranscriptomics. To eliminate contamination from de novo transcriptome assemblies, sequence similarity searches were conducted using the MMseqs2 tool [ 179 ] against a custom database containing proteins from the axenic strains MZCH-10213, SAG 1.88, SAG 12.97, SAG 615-1, SAG 698-1a, and SAG 698-1b as positive control, and potential contaminants including RefSeq representative genomes of bacteria (43,588), fungi (1,225), invertebrate animals (7,213), virus (13,134), archaea (1,120), protists (893), mitochondrial genes (240,117), plasmids (4,710), and plastids (12,307) (downloaded July 11, 2023). An iterative approach was used for MMseqs2 (--start-sens 1 --sens-steps 3 -s 7 --alignment-mode 3 --max-seqs 10), and decontamination was completed with GPDS [ 20 , 21 ].
Only sequences whose best match was against the combined positive set were retained. Potential contaminant-matching sequences were archived and are available via Zenodo https://doi.org/10.5281/zenodo.18392035 Phylogenomic analysis: Phylotranscriptomic analysis. Orthogroup inference was performed using OrthoFinder [ 166 ] v3.1.1 with a species tree following the framework of the One Thousand Plant Transcriptomes Initiative [ 2 ], including chlorophytes, Chlorokybus cerffii SAG 34.98, Mesostigma viride NIES-296, and Klebsormidium nitens NIES-2285, alongside multiple Phragmoplastophyta and Coleochaetophyceae (see “09_Orthofinder.zip” on Zenodo).
A total of 60,932 orthogroups were identified. Two orthogroup subsets were generated using OSG [ 20 , 21 ]: one enriched for ingroup taxa (13,409 orthogroups, filter: 2 of 4 groups) and another for outgroups (7,253 orthogroups, filter: 10 of 17 groups). All data are available at Zenodo (“10_OSG.zip”).
Multiple sequence alignments were performed with MAFFT [ 164 ] v7.304 under default settings. Maximum likelihood trees were constructed using the IQ-TREE 2 software [ 162 ] v2.2.2.7 with automated model selection via ModelFinder [ 22 ] (selection of best-fit models was based on Bayesian Information Criterion), using SH-like aLRT (-alrt 1000) and ultrafast bootstrap replicates (-bb 1000). Tree files were reformatted for PhyloPyPruner v1.2.4 using ApplyPPPFormat ( https://github.com/mjbieren/ApplyPPPFormat ).
PhyloPyPruner ( https://github.com/fethalen/phylopypruner ) was used to remove paralogs. For the orthogroup subset biased towards the ingroup, we used the following settings: --mask pdist --prune MI --min-taxa 3 --trim-lb 5 --min-support 0.75 --threads 50 --trim-divergent 1.25 --jackknife, yielding 93,280 ortholog alignments. For the outgroup-centered subset, we used: --min-taxa 10 --min-gene-occupancy 0.1, yielding 5,960ortholog alignments.
After taxonomic filtering to ensure a balanced taxonomic representation in alignments (FilterPPPResult, -t 10 (of 110)), we retained 17,588 ingroup-rich and (FilterPPPResult, -t 2) 5,923 outgroup-rich loci. These were merged using COGS [ 21 ] and then pruned to 5,225 unique orthogroups. A final PhyloPyPruner run produced 80,973 loci; after filtering (FilterPPPResult, -t 30), 2,243 loci remained.
These were masked using PREQUAL [ 170 ] v1.02 and re-aligned with MAFFT (G-INS-i, ‘--allowshift --unalignlevel 0.8′). Gappy columns (>75% gaps) were removed using ClipKIT [ 159 ] v2.0.1. The final concatenated matrix was created using phyx [ 180 ] comprising 126 taxa, 2,243 loci, and 579,537 aligned amino acid sites.
Phylogenies were reconstructed with IQ-TREE 3 V3.0.1, using LG+C60+F+G-PMSF, 1000 replicates of each SH-aLRT and UFBoot [ 181 ]. Ancestral character state reconstruction. Ancestral character state reconstruction (ACSR) estimates the probability of character states at ancestral nodes from the observed states at the terminal taxa, the phylogenetic tree, including branch lengths, and an explicit model of character evolution.
We used an equal-rates model, assigning an identical transition rate of 1 to all possible state transitions, i.e., analyses assumed symmetric transition rates between unordered states. ACSR was conducted using Phytools [ 168 ], employing Yang’s re-rooting method [ 182 ]. Traits were coded with a 5-state model: (1) unicellular, (2) multicellular, (3) Package-like, (4) Bona fide filaments, and (5) Chain-like filaments.
Gene family analysis. For gene family analysis, we worked with a dataset of predicted proteins derived from (i) the transcriptomes generated in this study and two available ones and (ii) published genomes of phylodiverse Chloroplastida (see Resource table ). We searched for homologs of the genes of interest using representative sequences from Arabidopsis as a query in a BLASTp search against the full protein dataset.
An e-value cutoff of 10 -7 was applied; in case of large gene families, the number of species was reduced. All sequences were aligned using MAFFT [ 164 ] v7.490 (L-INS-i) and a maximum likelihood phylogeny was computed using IQ-TREE [ 161 ] with 1000 UFBoot [ 183 ] pseudoreplicates and best-fit models chosen according to Bayesian Information Criterion by ModelFinder [ 22 ]. The AHK gene family tree was annotated using PFAM domains predicted by InterProScan [ 160 ] v5.77-108.0.
The domains were consolidated per clade. A clade was annotated to contain a domain when it appeared in several proteins within that clade. Outlier domains, e.g., only predicted for 1 protein within the whole clade, were ignored as an observation based on a single protein ( n = 1) provides very limited support for drawing conclusions about the broader evolutionary pattern within that clade.
We therefore adopted a conservative filtering criterion and retained only labels for clades in which the domain was detected in more than one protein, thereby reducing the influence of potentially spurious individual observations on the overall analysis. We summarized PFAM PAS fold and PAS domain annotations (PF00989, PF08447, PF08448, PF13188, PF13426) with one symbol. Overview of the quantifications and statistical analysis.
The final maximum likelihood phylogenomic tree was inferred using the LG + C60 model with 1000 SH-aLRT and 1000 ultrafast bootstrap replicates. Additional analyses incorporating site-heterogeneous PMSF models were conducted using a guide tree inferred with LG + C60 + F + Γ. These consistently recovered the same topology as the original LG + C60-based tree.
ModelFinder [ 22 ] identified LG + C60 as the best -fitting model for protein evolution compared to the more standard empirical amino acid substitution model with site-rate heterogeneity LG + I + G4, as indicated by improved likelihood scores and a lower Bayesian Information Criterion (LG + C60: Log(Likelihood) = −15,486,259.27; Bayesian Information Criterion = 30,975,836.031; LG + I + G4: Log(Likelihood) = −15,683,898.390; Bayesian Information Criterion = 31,371,127.541). The maximum likelihood phylogenomic tree was inferred under the LG + C60 model using 1,000 SH-aLRT and 1,000 ultrafast bootstrap replicates. To account for site heterogeneity, an additional analysis was performed using the PMSF approximation with LG + C60 + F + Γ, guided by the previously inferred tree.
This analysis consistently recovered the same tree topology, reinforcing the robustness of the phylogenetic inference across model frameworks. For gene family analyses, 144 protein models were tested using ModelFinder [ 22 ] and Bayesian Information Criterion was used to determine the best-fit model. 1000 UFBoot [ 183 ] pseudoreplicates were computed per gene tree.
Supporting information S1 Fig Fully labeled phylogenetic trees of signature gene families. Related to Fig 3 . Fully labeled phylogenies shown in Fig 3 .
The data and code needed to generate this Figure can be found in https://zenodo.org/records/18392035 as well as https://doi.org/10.5281/zenodo.22172294 on Zenodo. (PDF)