Evolving circuitries in plant signaling cascades Free

Land plants (embryophytes; see Glossary) are the most species-rich group of photosynthetic eukaryotes (Corlett, 2016). Their evolution drastically changed the face of our planet by accumulating an astounding biomass (Bar-On et al., 2018) and raising atmospheric oxygen concentration to the current levels (Bowles et al., 2023; Lenton et al., 2016). All of this diverse flora can be traced back to an evolutionary singularity; thus, it is not surprising that the evolutionary origin of land plants has attracted much research interest from various angles (Bowman, 2022; Delaux and Schornack, 2021; Delwiche and Cooper, 2015; Donoghue et al., 2021; Fürst-Jansen et al., 2020; Graham et al., 2000; McCourt et al., 2023; Strother and Foster, 2021). In the past 15 years, major advancements have been made in the field of plant evolutionary biology. The first to highlight is the establishment of a robust phylogenetic framework of streptophytes (Streptophyta, the clade encompassing land plants and streptophyte algae; see Glossary) that permitted subsequent comparative analyses. This started with the early recognition (Wodniok et al., 2011) and phylogenomics-based reinforcement that Zygnematophyceae (see Glossary), a class of streptophyte algae, are the closest algal relatives to land plants (One Thousand Plant Transcriptomes, 2019; Puttick et al., 2018; Wickett et al., 2014). These results ended a long search for which of the three algal classes within the Phragmoplastophyta – Charophyceae, Coleochaetophyceae or Zygnematophyceae (see Glossary) – is most closely related to land plants (Fig. 1A). Likewise, the recovery of the bryophyte (see Glossary) monophyly with phylogenomic datasets has also changed our view of the last common ancestor (LCA) (see Glossary) of land plants (Puttick et al., 2018), revealing that bryophytes and tracheophytes have taken very different evolutionary routes, including both profound inflation and shrinkage of gene families (Harris et al., 2022). For reconstructing the LCA of land plants, their phylodiverse descendants, which carry disparate and mosaic set of genes – the result of these ancient genetic events (Harris et al., 2022) – must be sampled. For the purpose of this Review, this specifically means that a variation in response or regulation found in bryophytes as compared to tracheophytes pinpoints two divergent possibilities that emerged from the LCA of land plants, which could have had either of the two variants. Specific examples of such cases follow below, particularly in the phytohormone section.

The close affinity of Zygnematophyceae to land plants is surprising; among phragmoplastophytic algae, Zygnematophyceae have the least complex body plans. Indeed, this species-rich lineage has, throughout its evolutionary history, shifted multiple times between unicellular and multicellular body plans (Hess et al., 2022). Despite this superficial lack of land plant-like bodies, Zygnematophyceae provide a rich source of information on the evolutionary development of the molecular programs shared between land plants and streptophyte algae. These programs, which were present in common ancestors that lived long before plants had conquered land, include key genes for symbiotic interactions, specialized metabolism, phyotormone biosynthesis and generally elaborate signaling networks (Cheng et al., 2019; Dadras et al., 2023a; de Vries et al., 2018; Delaux et al., 2015; Feng et al., 2024; Jiao et al., 2020; Rieseberg et al., 2024 preprint; Sun et al., 2019; Wang et al., 2015). Combining streptophyte algae data with functional genomics-based surveys and knowledge from decades of work on land plant model systems, researchers have identified a range of molecular programs that likely share a long history across almost a billion years of streptophyte evolution. In the following section, we highlight a set of such ancient molecular programs salient to plant signaling today.

Research into how streptophytes other than model land plants perceive their environment is currently limited by the means and tools through which we can manipulate biological processes. In streptophyte algae, so far, this has predominantly involved altering the environment (including exogenous application of substances) to trigger signaling processes, the effects of which can be assessed using high-throughput functional genomic approaches, foremost RNA sequencing (Dadras et al., 2023a; de Vries et al., 2018, 2020; Fürst-Jansen et al., 2022; Rieseberg et al., 2023; Rieseberg et al., 2024 preprint; Rippin et al., 2017, 2019; Serrano-Perez et al., 2022). These approaches are mainly a means to study effects; however, through the application of advanced computational biological approaches, including statistical modeling and machine learning, such correlative data can be used to yield predicted genetic hierarchies (Rieseberg et al., 2024 preprint). Nonetheless, targeted investigations on selected cascades have been carried out, particularly on phytohormone signaling cascades, one of the primary topics of interest in the field of plant evolutionary biology.

Currently, the phytohormone that appears to be the clearest case for deep conservation in streptophytes is ethylene. Investigations into ethylene signaling were first prompted by the discovery that the Zygnematophyceae Spirogyra has promising homologs for genes coding for putative biosynthesis of ethylene, its receptor and its signaling proteins (Ju et al., 2015). Spirogyra was found to react to exogenous application of ethylene through cell elongation (Ju et al., 2015) and the upregulation of stress response-associated genes (Van de Poel et al., 2016). Additionally, Spirogyra appears to produce ethylene in a 1-aminocyclopropane-1-carboxylate (ACC)-dependent manner. In seed plants, ACC is the direct precursor of ethylene (Adams and Yang, 1979) and the substrate of the ethylene producing ACC oxidase (ACO; Hamilton et al., 1990). ACO belongs to a large gene family, and although there are homologs of ACO in streptophyte algae, a clear ortholog is missing; therefore, a land plant-typical ACO might first have emerged in seed plants (Van de Poel and de Vries, 2023). Furthermore, ACC can act as a signaling molecule on its own accord (Li et al., 2020), raising additional questions about whether streptophyte algae convert ACC into ethylene. Hence, the evolutionary history of ethylene production is not yet resolved. Much greater resolution is available to us when it comes to the evolution of its signaling components. The ethylene signaling cascade involves the ethylene receptor ETHYLENE RESPONSE 1 (ETR1), which, upon binding ethylene, inactivates its negative regulator CONSTITUTIVE TRIPLE RESPONSE 1 (CTR1), leading to the stabilization of ETHYLENE INSENSITIVE 2 (EIN2). EIN2 then facilitates the translocation of the transcription factor EIN3 into the nucleus, where EIN3 activates ethylene-responsive genes, thus mediating the effects of ethylene on plant growth and development. Interestingly, the Spirogyra homolog of ETR1 can partially rescue Arabidopsis etr triple mutants (Ju et al., 2015). Meanwhile, EIN2, which moves from the membrane of the endoplasmic reticulum to the nucleus upon ethylene treatment (Qiao et al., 2009, 2012), shows ethylene-dependent trafficking to the nucleus (Ju et al., 2015). Finally, Arabidopsis mutants deficient in the transcription factor EIN3 can be partially restored to wild type through the expression of Spirogyra EIN3 (Ju et al., 2015). Several of these genes salient to ethylene signaling were also recovered in a predicted network of gene–gene regulatory relationships shared across Zygnematophyceae and land plants (Rieseberg et al., 2024 preprint). Thus, although genetic evidence is still missing due to the current methodological limitations in streptophyte algae, it is fair to say that ethylene induces quantifiable alterations in Spirogyra and that all data point towards a conserved molecular chassis for the ethylene-triggered response cascade.

When it comes to other phytohormones, the situation is less straightforward than for ethylene. A phytohormone that has received much attention in recent years is jasmonic acid (JA) and its role in bryophytes, as exemplified by studies in the moss and liverwort model systems. This started with the finding that the model moss Physcomitrium patens lacks JA (Stumpe et al., 2010). Work that focused on the JA receptor coronatine-insensitive protein 1 (COI1) in the liverwort M. polymoprha (Mp) revealed that instead of binding the bona fide ligand known from angiosperms, MpCOI1 instead binds dinor-cis-12-oxo-phytodienoic acid (OPDA) and dinor-iso-OPDA (Monte et al., 2018), compounds that are present in several streptophyte algae (Schmidt et al., 2024). Full downstream responses are, however, most likely only achieved with a ligand diversity that also includes a dinor-OPDA-like compound (Kneeshaw et al., 2022). Despite the difference in ligand preference, the downstream response cascades appear to be conserved across land plants (Monte et al., 2019; Penuelas et al., 2019). However, the origin of COI1 is placed at the LCA of land plants, where it emerged together with TIR1 from a duplication event of an F-box protein. These F-box proteins are involved in phytohormonal-activated targeting for proteosomal degradation of transcription factors. The F-box protein is part of the Skp1-cullin-F-box (SCF)-type E3 ubiquitin ligase complex and determines substrate specificity for both the target and the phytohormone in the case of COI1 and TIR1 (Blázquez et al., 2020). Streptophyte algae possess a co-ortholog to the duplicated COI1 and TIR1 F-box receptors, but these algal co-orthologs differ in key residues and both phytohormonal and degradation targets are still unknown (Bowman et al., 2019, 2021; Carrillo-Carrasco et al., 2023; Feng et al., 2024). Nevertheless, a conserved COI1-independent signaling mechanism to integrate environmental response to temperature has been identified across streptophytes, pointing to an existence of JA or dinor-OPDA based signaling in streptophytes (Monte et al., 2020). Moreover, these compounds have been detected in land plants and streptophyte algae (Chini et al., 2023; Schmidt et al., 2024).

A similar story emerges for four other phytohromones – auxins, gibberellic acid (GA), strigolactones and karrikins. Analogously to JA signaling, these corresponding phytohormonal signaling cascades are SCF-mediated (Blázquez et al., 2020). In the case of auxin, most key components of the transcriptional response module were likely already present in an ancient phramgoplastophytic streptophyte algal ancestor that lived more than 650 million years ago (Carrillo-Carrasco et al., 2023). This is best exemplified by the AUXIN RESPONSE FACTORS (ARFs). Class C ARFs are present across streptophytes and likely emerged in their LCA about one billion years ago. Conversely, a class A and B co-ortholog first emerged in the common ancestor of Coleochaetophyceae, Zygnematophyceae and Embryophyta (Flores-Sandoval et al., 2018; Mutte et al., 2018). The potentially more ancient C-ARFs have been shown to possess different binding affinities that set them apart from the A- and B-ARFs; this C-ARF-characteristic binding even occurs in the single ARF of the more distantly related streptophyte alga Chlorokybus (Martin-Arevalillo et al., 2019), which share with land plants a LCA that lived about one billion years ago (for analyses of recent timings, see Bierenbroodspot et al., 2024; Bowles et al., 2024). Together, this means that the A- and B-ARFs might have arisen from C-ARFs (Martin-Arevalillo et al., 2019). However, the other components of the nuclear auxin response module (the TIR1- and AFB-like family of proteins, and the AUX- and IAA-like family of proteins) occur in a less diversified manner and lack functionally important residues (Carrillo-Carrasco et al., 2023). That being said, a deep conservation across streptophytes appears to exist regarding the much swifter phosphorylation-based responses (Kuhn et al., 2024). The phytohormone auxin is present throughout the streptophyte lineage and was likely present in its LCA. However, its role and biosynthetic pathway in this ancestor remain unresolved. The canonical auxin biosynthetic enzymes YUCCA and TAR are not found in streptophyte algae (Bowman et al., 2019, 2021). Given that auxins are secreted into the external medium, it has been hypothesized that they might have mediated quorum sensing (sensing of cell densities) in the LCA of streptophytes (Schmidt et al., 2024; Vosolsobě et al., 2020 preprint).

In the case of abscisic acid (ABA), the main response module consists of the cytosolic PYRABACTIN RESISTANCE1 (PYR1), PYR1-LIKE (PYL) and REGULATORY COMPONENTS OF ABA RECEPTOR (RCAR) family, the PROTEIN PHOSPHATASE 2CA (PP2CA) proteins and the SUCROSE NONFERMENTING 1-RELATED PROTEIN KINASE 2 (SnRK2) (Cutler et al., 2010). In brief, when ABA levels increase, an ABA–PYL–PP2CA complex is more likely to form, which inhibits the activity of the formerly unbound PP2CA (Melcher et al., 2009; Miyazono et al., 2009; Park et al., 2009; Santiago et al., 2009; Yin et al., 2009); consequently, SnRK2s are not further inhibited by PP2CA via dephosphorylation and can activate downstream targets (Soon et al., 2012). Zygnematophyceae have a full homologous set of components for the bona fide ABA signaling module components (Cheng et al., 2019; de Vries et al., 2018; Feng et al., 2024). Functional characterization has shown that the interaction between PYL and PP2CAs, as well as the action of SnRK2s, have been conserved for hundreds of millions of years of streptophyte evolution (Lind et al., 2015; Shinozawa et al., 2019; Sun et al., 2019). However, importantly, characterization of the interaction between a Zygnema PYL homolog has shown that its inhibitory action on PP2CA occurs in an ABA-independent manner (Sun et al., 2019). A plausible scenario is that after the lineage of land plants split from that of Zygnematophyceae algae, an ancient cascade from PP2Cas, to SnRKs2 and downstream transcriptional regulators was brought under the regime of ABA by changes in the formation of the PYL–PP2CA complexes. Gaining an ABA dependency for complex formation hence yielded the bona fide phytohormone-dependent cascade known from land plants. However, several unknowns remain regarding the evolution of the ABA signaling cascade – given that a PYL protein is present in Zygnema and that it negatively regulates the PP2CA phosphatases, it is fair to assume that something must in turn modulate that regulation, be it a molecule different from ABA or another regulatory protein. Regardless, it is a valid working hypothesis that the PP2CA–SnRK2 cascade could have been part of molecular processes involved in the acclimation to adverse environmental conditions and only later became controlled under the regime of ABA. This is further supported by the checkered detection of the canonical biosynthetic enzymes ABA2 and 9-cis-epoxycarotenoid dioxygenase (NCED) in streptophyte algae (de Vries et al., 2018; Feng et al., 2024). It should be noted, however, that ABA has been detected in some of these algae, suggesting that ABA could be produced via alternative pathways (Schmidt et al., 2024), for which there are several options even in land plants (Jia et al., 2022). Overall, the question of whether ABA is a physiologically relevant molecule in streptophyte algae remains open.

All of the aforementioned examples speak of the versatility of plant signaling cascades during evolution. However, in most cases, portions of the response-determining steps seem conserved. Next, we will explore how conserved tracks that bear out of regulatory loops might have assembled into important signaling cascades.

In the previous section, we highlighted homology-guided approaches that identified the conservation of signaling cascades that are known from land plants. Such approaches limit our understanding of the full evolutionary picture, as they cannot identify additional proteins that might be at play in non-model streptophyte algae. Proteomics, including high-throughput protein–protein interaction experiments, has the power of identifying the actual proteins involved in the signaling cascades. However, such studies are currently often limited to Arabidopsis or a few other model species, which hampers our understanding of signaling cascades across phylodiverse organisms and thereby the inference of their evolution. There is some evidence that within a given signaling pathway, the interaction between proteins varies across species. Consider, for example, the ABA-independent cascade in Zygnema (see above) or the regulation of the ion-channel slow anion channel associated 1 (SLAC1) by SnRK2 in the ABA signaling pathway. The latter interaction is well established in Arabidopsis, but their respective homologs in Klebsormidium do not exhibit the same interaction (Lind et al., 2015). In the gibberellin SCF-mediated pathway, the transcriptional regulator DELLA acts as an interactor with transcription factors as well as the target for proteosomal degradation. The interaction between DELLA and transcription factors is well documented in Arabidopsis, and several of these interactions are conserved; however, it appears that DELLA binds promiscuously with transcription factors, and the establishment of specificity during evolution remains an open question (Briones-Moreno et al., 2023).

Proteins participating in a signaling cascade that have been lost or have evolved another function in the model species investigated (as compared to other representatives from the lineage to which the model organism belongs), are very hard to integrate into quantifiable evolutionary models, as they are built on sequence-based determination of phylogenetic relationships. Additionally, there is no straightforward approach to give weight to a few (or even a single) amino acid substitutions that can completely change key biochemical properties, such as binding capacity, substrate preference and the interaction with crucial downstream components. Simply put, solely looking at a phylogenetic tree of a protein family cannot capture how biological functions have changed; owing to the interwoven nature of molecular programs, such changes can lead to cascading changes in other components (a ‘domino effect’). Here, researchers are limited to conceptualization based on evolutionary scenarios that are often subjectively inferred; this is exciting and humbling, but it also biases our view. A common consequence is that the complexity of a signaling cascade diminishes the further one moves away from the model organism, largely due to fewer homologous signaling components being found through model-based homology searches (see Fig. 1B). There are undoubtedly additional, yet unknown, components in the signaling cascades of streptophyte algae acting together with ‘canonical’ proteins known from land plants (orthologs). Obvious candidates are paralogous genes that originated from gene duplication a long time ago, resulting in differential recruiting of paralogs in land plants and the various streptophyte algae. Paralogous genes often share protein domains and active sites, although these might have also diverged (including functionally), or even evolved convergently. Gene family analyses in streptophytes often show the independent diversification of some paralogs in streptophyte algae along with the orthologs of well-known Arabidopsis components. However, the functionality of these paralogs in streptophyte algae is in most cases not understood. The analysis of protein structure and the conservation of functionally relevant amino acid residues that might be known from experimental studies in model species such as Arabidopsis might help us establish hypotheses of shared – and thus potentially ancient – functions. In addition to paralogs, other proteins might also be involved in the lesser-studied signaling cascades of streptophyte algae.

Further variations in streptophyte algae signaling cascades can also go unnoticed. Environmental stimuli can trigger similar response cascades, but the transduction might occur through a different signaling molecule or pathway. A noteworthy example is the far-red light response in the liverwort Marchantia, which is triggered by GA-like compounds, even though bryophytes lack canonical GA receptors (Sun et al., 2023). Here, ent-kaurenoic acid and GA₁₂ seem to be precursors of the main, currently unknown, GA(-related) compound (Sun et al., 2023). This highlights how chemodiverse ligands can funnel into conserved cascades, co-opting them for a regulatory regime ascribed to that ligand and simultaneously reinforcing the involvement of chemodiversity in regulation (Dadras et al., 2023b). This can be a source of increasing complexity of molecular networks – using established genetic substrates (see also Dhabalia Ashok et al., 2024) – whose ramifications in plants and algae can only be grasped through studying diversity.

A single component plugged into an established cascade can change the regulatory regime under which the whole cascade operates (Sun et al., 2019; Fürst-Jansen et al., 2020). All cascades are most likely intertwined with basic functions that any cell must accomplish, such as photosynthesis, growth and cell division. These ancient cascades tend to be highly interconnected, increasing the likelihood of new connections being formed and yielding complex systems. A study that exemplified this concept investigated the conservation of phosphorylation-based signaling in auxin responses (Kuhn et al., 2024). This study demonstrated that auxin triggers both fast and slow cellular responses across land plants and algae. Although the nuclear auxin pathway mediates gene expression and developmental control in land plants, this pathway is incomplete in algal sister groups. Despite this differences and the potential confounding factors that comparing gametophytic and sporophytic tissues (and the subfunctionalization therein; see Glossary) impose, a conserved rapid proteome-wide phosphorylation response to auxin across five land plant and algal species has been identified. This response converges on a core group of shared phosphorylation targets, mediated by RAF-like protein kinases. These kinases, identified as central mediators of auxin-triggered phosphorylation, connect fast auxin signaling to cellular processes, such as cytoplasmic streaming (Friml et al., 2022). Indeed, these findings echo that conserved responses of the cell systems are at the heart of the evolution of signaling networks, because the conserved fast auxin response system pinpointed by Kuhn et al. (2024) pertains to general responses like membrane polarization. Therefore, an effective approach to studying signaling networks involves examining the central signaling hubs and their interconnected environmental framework (interacting proteins, metabolites and/or biological cues). By focusing on these key components, we can minimize errors and streamline our analysis. Although molecular micro-evolutionary events – such as changes in alternative splicing, mutations in regulatory elements and changes in enzyme activity – are disregarded in this conceptual framework, the emphasis on central components allows for a more efficient understanding of the system. Attempting to explain molecular macro-evolutionary events, such as the emergence of new gene families, co-option of existing pathways or loss of a pathway, through a single or a few gradual events, would be akin to searching for a needle in a haystack. However, it is worth noting that single or just a few mutations in just one or a few genes can have strong effects and also be the trigger of drastic rearrangement, including the aforementioned domino effect. Utilizing comparative methods offers a distinct advantage in this context, as central signaling components are evolutionarily robust, meaning they undergo minimal loss or replacement over time. This stability enables researchers to trace the evolution of signaling networks across significant time scales more effectively.

We argue that to identify conservation in signaling, the focus should shift towards determining central hub genes (see Glossary) rather than attempting to find one-to-one orthologs of proteins from known land plant cascades. Protein kinases often form these central hubs in stress response cascades because they can integrate and amplify signals (Zulawski et al., 2014; see Fig. 2). The majority of protein kinases belong to the large family of eukaryotic-type protein kinases and are shared among many eukaryotes beyond the streptophyte lineage (Leonard et al., 1998). The two other groups are the histidine kinases, likely acquired through endosymbiotic gene transfer around 1 billion years ago (Mount and Chang, 2002), and the phosphoinositide-3-OH-kinase-related kinases (PIKKs) (Templeton and Moorhead, 2005). Protein kinases are vital in rapidly regulating essential cellular functions, making evolutionary changes such as the expansion and appearance of new subfamilies through domain shuffling highly informative. For instance, various receptor-like kinases (RLKs) arose from a merger between a duplicated kinase domain and a receptor-like protein (RLP), such as the RLP LysM yielding CERK1, and the RLP malectin yielding FERONIA, which predated their corresponding RLK existence (Ngou et al., 2024).

Although a protein-kinase-centered approach is valuable, considering the role of proteasomal degradation in phytohormonal signaling pathways – an F-box protein-centered perspective – is also worthwhile. To infer the evolutionary history of a stress signaling cascade, we should first examine the evolution of its central components. Then, we should attempt to identify conserved interactors and stressors. Any cell must have feedback mechanisms to sense its primary functions, be it cell wall homeostasis (Elliott et al., 2024; Gonneau et al., 2018; Wolf et al., 2014), sugar status (Li and Sheen, 2016; Rolland et al., 2006; Smeekens et al., 2010; Xiong et al., 2013), or reactive oxygen species (ROS) and redox feedback from photosynthesis (Foyer and Hanke, 2022; Laohavisit et al., 2020; Wu et al., 2020). These feedback loops, linked to sensing cell homeostasis and metabolic status, as well as specific inputs from (a)biotic stressors, all converge in the central hubs. Furthermore, we must also consider additional proteins or signaling network variations occurring in the diverse species in order to fully reconstruct the evolution of these networks. Failure to account for this signaling diversity might lead to erroneously inferring a continuous increase in complexity during evolution up to the canonical network known from land plants. We thus highlight the need to move away from such a biased Arabidopsis-centric view of signaling pathway evolution.

It is fully conceivable that the earliest land plants had a branching filamentous body plan resembling a protonemal culture of moss, additionally developing rhizoids (filamentous root-like structures of one or a few cells that anchor plants to their substrates). Akin to the morphological versatility that we see in Coleochaetophyceae, the morphogenetic cytokenesis that results in branching could have entailed both anticlinal and periclinal cell divisions (see Glossary), resulting in parenchymatous thalli (Delwiche et al., 2002; Dupuy et al., 2010). With the above, this brings together two key aspects – responses to environmental cues and development, both of which hinge on signaling cascades. Since the LCA of land plants, the elaboration of the plant body likely increased, going hand in hand with specialized tissues; this likely also coincided with whole genome duplication (WGD) events, leading to and accompanied by rampant gene duplications followed by neo- and sub-functionalization (i.e. the generation of paralogs that are functionally divergent or differentially regulated across tissues, respectively; see Glossary). This situation is what has been studied in land plant model systems. How did plants balance (1) the need for specific signaling processes in a particular tissue of their many diversified tissue types with (2) the existence of ‘general cascades’, as discussed in the previous section?

Significant evolutionary changes in two eukaryotic protein kinase families provide valuable insights towards addressing this balance. First, there has been a notable expansion in Ca²⁺-dependent protein kinases (CDPKs), which play crucial roles in Ca²⁺ signaling (Lehti-Shiu and Shiu, 2012). Plants utilize Ca²⁺ for rapid signal transmission over long distances. The diverse array of CDPKs, each with varying Ca²⁺ affinities, allows different cells to respond uniquely to Ca²⁺ waves travelling through the plant. Secondly, large expansions and emergences of new subfamilies within the receptor-like kinase (RLK) and Pelle family have been inferred at various points along the trajectory of streptophyte evolution (Gong and Han, 2021; Ngou et al., 2024). RLK and Pelle proteins initiate phosphorylation cascades and often sense signals on the cell surface (Gish and Clark, 2011; Ngou et al., 2024). This expansion likely facilitated the development of more complex body plans by enabling tailored responses of each cell to specific signals. Moreover, this diversification might have allowed for fine-tuning interactions with microbes by enabling the recognition of a broader range of microbial elicitors through the development of sensors for various peptide signals (Ngou et al., 2024).

Adding further layers of complexity, the establishment of a multicellular body in land plants created the need for signals to travel across greater distances. Some of the mechanisms might have already been established as ways of communication between members of the same species. For example, in case of ethylene, one could question the relevance of a gaseous signaling molecule for aquatic organisms (although several streptophyte and chlorophyte algae are aeroterrestrial). However, a multicellular body of a single individual has different needs, most palpable by the division between source and sink cells in respect to the sugar status of a plant body, where sugar transport is involved in long-distance signaling (Jamsheer et al., 2019; Li and Sheen, 2016; Lin et al., 2014). Additionally, hormones and peptides that emerged during diversification of embryophytes are used to transfer signals over long distances. Take, for example, FERONIA, a malectin RLK that can sense changes in the cell wall, but is most famous for its role in sensing the signalling peptide rapid alkalinization factor (RALF). FERONIA pre-dated the emergence of the first embryophyte; however, RALF has a more complicated evolutionary history, appearing limited to land plants but also present in heterotrophic microbes (Furumizu and Shinohara, 2024; Liu et al., 2017; Zhang et al., 2020).

We can assume that the aforementioned long-distance signaling was established when the first elaborate streptophyte bodies arose. Current evidence suggests that the LCA of Zygnematophyceae and land plants, as well as the ancestor of all Phragmoplastophyta, had a more complex body than the Zygnematophyceae ancestor (Bierenbroodspot et al., 2024; Bowles et al., 2024; Feng et al., 2024; Fürst-Jansen et al., 2020; Hess et al., 2022; Wodniok et al., 2011). Indeed, Zygnematophyceae have seemingly enriched signaling cascades (Feng et al., 2024), which one could provocatively interpret as being overly complicated when species in this class only have a few cell types, more often than not being unicellular. Assuming that the earlier ancestor shared with land plants was multicellular and potentially with complex tissues (Bierenbroodspot et al., 2024; Bowles et al., 2024; Feng et al., 2024; Hess et al., 2022; Wodniok et al., 2011), it is plausible that the elaborate signaling pathways in Zygnematophyceae are a relic from a time before this reduction in cell complexity. If so, why did they remain? After a gain of more signaling components because of (or causing) an increase in the number of cell types, it might not be possible to suddenly decrease the signaling complexity; it would disentangle entire networks. But can we say that these signaling pathways do not function in Zygnematophyceae or that they are not needed? There might be additional pathways that depend on some of the components, but also occasions when they might need such networks independent of any signaling function. Similar to concepts in constructive neutral evolution (see Glossary) (Covello and Gray, 1993; Stoltzfus, 1999; Gray et al., 2010), these signaling components might have become entrenched in the biology of the organism.

A lot has been learned by focusing on homology- and orthology-based bioinformatic analyses to find deeply conserved ‘one-to-one matches’ in components of land plant signaling cascades. This research highlights the evolutionary continuity of key signaling components from the identified ethylene, JA, auxin and ABA signaling cascades in Arabidopsis, underscoring the deep roots of plant signaling networks and their sophisticated molecular programs preserved over hundreds of millions of years. Ethylene signaling demonstrates a highly conserved molecular framework, with components like ETR1 and EIN2 in Spirogyra showing functional conservation with their Arabidopsis counterparts. Phytohormones, such as JA and auxin, exhibit complex evolutionary trajectories, indicating that a sophisticated regulatory system evolved to meet diverse physiological needs. The ABA signaling cascade showcases evolutionary adaptation, with the conservation of core components like the SnRK2 family and PP2CA proteins highlighting ancient stress response pathways that have been refined in land plants.

Homology-based approaches have limitations, particularly in identifying novel components specific to non-model streptophyte algae. Protein–protein interaction studies in non-model species could reveal additional players in these signaling cascades, providing a more comprehensive understanding of their evolution. To charter the path forward, another meaningful step might be finding recurrence in what is being sensed (e.g. metabolic feedback or photophysiological status) and which type of interactions in a cascade are robust across a long evolutionary time. Here, working around the hubs, such as protein kinases and F-box proteins, that have emerged through various comparisons, might be an apt approach. Establishing evolutionary ‘anchors’ and investigating recurring environmental stimuli can reveal patterns of robustness in signaling cascades. Expanding functional genomics and protein interaction studies to non-model streptophyte algae will offer a complete picture of signaling evolution, crucial for reconstructing the full evolutionary history of plant signaling pathways. Ultimately, this allows for an inference of the evolutionary substrate in the chain of ancestors spanning millions of years of evolution from the land plants and algae of today to the first streptophyte ancestors that molded the complex signaling networks of plants.

We thank the three anonymous reviewers for very constructive feedback and valuable suggestions that have markedly sharpened our manuscript.