In the earliest diverging groups of terrestrial plants such as Anthocerotes, mucopolysaccharides, also called mucilage, are produced around organs and have dehydration protection roles during growth and reproduction (Renzaglia, Duff, Nickrent & Garbary 2000). In flowering plants, several kinds of mucilage can be secreted by a wide range of organs such as seeds, fruits, roots, leaves, and stems conferring impressive and diverse physical properties such as viscosity or adherence (Galloway, Knox & Krause 2019). The ability of species to release seed mucilage (SM) upon imbibition is called myxospermy. Its presence and the description of its morphology in several species, and its putative ecological functions, were reported early in a Charles Darwin’s letter (Weitbrecht, Müller & Leubner-Metzger 2011) and for over a century by the scientific community. Review of biochemical analysis of SM composition and of deep morphological characterization of the mucilage secretory cells (MSCs) has been performed more recently (Phan & Burton 2018). In addition, several recent reviews mainly focused on the molecular, biochemical and structural characterization of the most studied SM plant model species Arabidopsis thaliana have been published (Griffiths & North 2017; Golz et al. 2018; Phan & Burton 2018; Šola, Dean & Haughn 2019). SM represents 2-3% of the total seed mass with numerous ecological functions (reviewed in Western 2012; Yang, Baskin, Baskin & Huang 2012c; North, Berger, Saez-Aguayo & Ralet 2014). This gathered knowledge led to several applications in pharmaceutical and food industry taking advantage of mucilage specific physical and chemical properties as well as its easy extraction (Mirhosseini & Amid 2012; Soukoulis, Gaiani & Hoffmann 2018).

In the last 50 years, SM was described in a majority of flowering plants orders associated with the diversity of MSCs and polysaccharidic mixture produced (Phan & Burton 2018). However, within the studied plants orders, several species do not have mucilage or do not extrude it. This raises the question of the evolutionary origins of SM: is it an ancestral trait of flowering plants present in the common ancestor that subsequently greatly diverged or was lost, or has it appeared several times independently? The answer to this evolutionary question rests on solving SM function(s), on standardized morphological description and on molecular studies focused on phylogenetically selected plant species/families. So far, a deep characterization of molecular actors implicated in SM establishment was performed primarily in A. thaliana . Comparisons with other species should allow to trackback the evolutionary scenario. However, the occurrence of numerous genes in complex regulatory networks, unequal implications in the trait and their often pleiotropic functions, makes it difficult to clearly identify genes that are committed with the trait. As evolution should conserve the better compromise between selective advantage and metabolic cost, looking at the entire carbon partitioning in seed will help to better understand SM evolution. Seed formation represents a high metabolic cost for the mother plant leading to a tight regulation of carbon partitioning between e.g. storage lipids accumulated in the endosperm and embryo, seed coat flavonoid pigments and SM (Song et al. 2017). The embryo/endosperm lipids will provide the required nutriments for proper embryo development, while pigments will confer impermeability and radiation protection (Baroux & Grossniklaus 2019). In this review, we first present a quick emphasis on SM and MSC structural diversity. The current state of the literature concerning the evolutionary origin of underlying genes and of their integration with other seed traits are developed in a second part at both intra- and inter- species level. Finally, we present the recent insights on SM ecological functions facing abiotic and biotic constraints as well as SM impact on plant development.

To properly extrude the SM to the environment, MSCs correspond to the outermost seed coat epidermal cell layer. As exemplified in A. thaliana and Linum usitatissimum (flax), this specific cell layer is formed during seed development to become a dead layer at the end of the seed maturation (Figure 1; Western 2001; Miart et al.2019). These two models illustrate the obvious diversity in the MSC cell wall dynamics during seed development as well as in the various mucilage organizations and extrusion modes. During the MSC development, theA. thaliana SM is trapped, with no apparent sub-layering, between the periclinal primary wall and a volcano-shaped polarized secondary wall called columella, whereas in flax, a complex multilayered SM is sequentially deposited in the MSCs during seed development (Figure 1). Beyond these two models, the MSC morphological diversity also occurs among the various species studied over the years (Phan & Burton 2018) as exemplified by the pioneer morphological survey of mature MSCs conducted for 200 Brassicaceae species covering 90 genera (Vaughan & Whitehouse 1971). Upon seeds imbibition, the hydrophilic nature of the polysaccharide mixture absorbs water, becoming a hydrogel swelling outwards. As a result, the SM volume and mass increase 75-fold inCapsella bursa-pastoris (Deng, Jeng, Toorop, Squire & Iannetta 2012) and 5.5-fold in Henophyton deserti (Gorai, El Aloui, Yang & Neffati 2014), respectively. The differences also occur in SM extrusion modes despite being less studied (Figure 1). The swelling increase pressure within the MSCs that allows the rupture of primary wall domains occurring either simultaneously in all Arabidopsis MSCs or sequentially in adjacent flax MSCs (Figure 1). This organized explosion is carefully prepared previously during MSC seed development by cell wall polysaccharides differential deposition and modifications. The polysaccharidic-proteinaceous cell wall domain molecular scaffold enabling the A. thaliana primary wall domain loosening starts to be uncovered (Kunieda et al. 2013; Saez-Aguayo et al.2013; Francoz et al. 2019). In flax, the polysaccharidic composition and its internal organization are proposed to play a role in proper MSC opening (Miart et al. 2019). A convenient method to attest the presence of SM is pectin staining directly performed on hydrated seeds with ruthenium red or section staining with a generic and polychromatic stain such as toluidine blue (Western 2001). Further characterizations are obtained by immunolabeling with cell wall epitope antibodies directly performed on whole mount hydrated seeds or on sections (Ben-Tov et al. 2018), chemical analysis facilitated by the easy extraction of mucilage (Zhao, Qiao & Wu 2017; Poulain, Botran, North & Ralet 2019) or ultrastructural analysis by scanning electron microscopy and atomic force microscopy (Kreitschitz & Gorb 2018; Williams et al. 2020). In Arabidopsis, released SM is composed by an adherent layer bound to the seed and a non-adherent layer both enriched in poorly branches type I rhamnogalacturonan (RGI) pectin domains (Figure 1; Macquet et al. 2007; Poulain et al.2019) while in flax the SM is composed by four contrasted layers enriched in RGI, arabinoxylans and xyloglucans/cellulose, respectively (Figure 1; Kreitschitz & Gorb 2017; Miart et al. 2019)). This type of SM layering and variable composition also occurs for other species such as for example, Lepidium perfoliatum (Huang, Wang, Yuan, Cao & Lan 2015), Neopallasia pectinata (Kreitschitz & Gorb 2017) or Plantago ovata (Tucker et al. 2017; Yuet al. 2017). SM microstructure also displayed inter-species diversity (Kreitschitz & Gorb 2018), implicating polysaccharide-polysaccharide specific interactions (Yu et al.2018) making SM an excellent model for cell wall dynamics understanding for about 20 years.

Interestingly, for species with seeds enclosed in a dry and non-dehiscent fruit called achene such as Salvia andArtemisia species, mucilage is extruded by the outermost fruit cell layer, namely the achene pericarp, and not by the seed integument. This parallel feature to myxospermy is called myxocarpy, both traits being regrouped under the term of myxodiaspory (Ryding 2001). Those are nice examples of evolutionary convergences leading to the similar differentiation into MSCs of different types of outer cells facing the environment. Indeed, the Salvia hispanica (chia) achene mucilage and MSCs show interesting parallels with A. thaliana and L. usitatissimum . Mucilage accumulates in the outer pericarp epidermal cells during seed development. After extrusion, the mucilage remains indirectly attached to the seed via the inner pericarp-seed tegument contact (Geneve, Hildebrand, Phillips, Al-Amery & Kester 2017). Finally, additional peculiarity may exist since in Medicago truncatula and M. orbicularis , the cell wall of the endosperm forms a mucilage gel between the seed coat and the embryo (Song et al. 2017). Therefore, the traits associated with myxospermy (or more generally with myxodiaspory) are numerous, including MSC structure, SM polysaccharidic composition and structural organization, and mucilage extrusion mode. However, there is not a simple clear-cut distribution of the myxodiasporic/non myxodiasporic traits along the Angiosperms families and even within families (Vaughan & Whitehouse 1971). For this reason, in the following part we will shed light on molecular mechanisms underlying this morphological diversity through inter-species and intra-species comparative studies for SM evolution understanding.

Twenty years of forward and reverse genetics together with more global approaches have allowed functional characterization of numerous genes involved in SM and MSC physiology in A. thaliana. The list currently contains between 82 (Phan & Burton 2018) and 58 genes (Sullivan et al. 2019) considering their indirect or direct involvement in SM, respectively. They constitute the continuously evolving MSC toolbox necessary for a proper SM production and release in the model plant (Francoz, Ranocha, Burlat & Dunand 2015; Voiniciuc, Yang, Schmidt, Günl & Usadel 2015). A majority of these genes are transcription factors including upstream master regulators that will be further discussed hereafter, and less characterized regulatory genes whose integration in the gene regulatory network is still puzzling (Golzet al. 2018). The other downstream genes of the toolbox are directly responsible of SM synthesis, assembly and secretion, or are involved in secondary cell wall synthesis and epidermal cell differentiation.

Understanding of SM evolution is easier while considering the intra-species rather than the inter-species natural variability because changes are still relatively recent on the evolutionary scale time and are scarcer. Indeed, the natural diversity occurring in L. usitatissimum cultivars or recombinant inbred lines from the cross of two varieties (Liu et al. 2016; Miart et al. 2019), or inA. thaliana natural ecotypes (Saez-Aguayo et al. 2014; Voiniciuc et al. 2016) shows gradient of SM abundance and release efficiency, reaching a complete loss of adherent mucilage extrusion forA. thaliana natural populations such as Sha (Macquet et al. 2007) or Rak-1 (Saez-Aguayo et al. 2014).Interestingly, in both cases the absence of adherent mucilage extrusion does not mean a lack of mucilage synthesis since Rak-1 releases even more non adherent mucilage than Col-0 (Saez-Aguayo et al. 2014). All natural mutants characterized for the myxospermy ability are related with polymorphisms concentrated on three downstream genes of the MSC tool box, namely PMEI6 for Dja, MUM2 for Sha, andPRX36 and MUM2 for Sk-1. These three genes encode enzymes necessary for proper SM extrusion and not SM synthesis or MSCs formation and PMEI6 and PRX36 functions are tightly related (Francoz et al . 2019). As non-myxospermic seeds have much better buyoancy efficiency and since Sha habitat is close to a river, seed dispersal by water run-off is one of the SM functions proposed to explain the loss of myxospermy (Macquet et al. 2007; Saez-Aguayo et al. 2014). Unfortunately, no clear association can be established between the natural population habitats and their mucilage phenotypes (Voiniciucet al. 2016).

More recent genome wide association studies (GWAS) conducted on A. thaliana (Fabrissin et al. 2019) and L. usitatissimum(Soto-Cerda et al. 2018) allowed identifying the statistically most relevant single nucleotide polymorphisms (SNPs) to explain the observed SM phenotype. In L. usitatissimum, all SNPs concerned orthologs of previously characterized A. thaliana MSC tool box genes that are either direct actors of SM synthesis or modification or regulatory genes that are not master regulators (except for LuTT8 having the lower E-value among the candidate genes) (Soto-Cerda et al.2018). In A. thaliana , through a very precise and molecular phenotyping, the GWAS revealed only 8 peaks significantly above the huge background of less implicated positions reminding that SM is an extremely polygenic trait (Fabrissin et al. 2019). Upon the 8 candidates genes two genes were identified and characterized, one already known to belong to MSC toolbox gene, and a newly implicated in SM (Fabrissin et al. 2019). These results highlight the fact that the MSC toolbox starts to be well characterized in A. thalianaand can be used to investigate whether functional orthologs are present in other species. It also suggests that the intra-species level, SM selective pressure does not act on the master regulators but rather on their downstream target genes of the toolbox. Reciprocally, the conservation of the top genes of the toolbox will allow to traceback the evolutionary origin of MSC toolbox genes across diverse families.

In A. thaliana , some of the master regulators belonging to the MSC toolbox also regulate the formation of trichomes and root hairs (Jones & Dolan 2012), the flavonoid biosynthesis (anthocyanidins and proanthocyanidins) (Lloyd et al. 2017) and are actors of the seed carbon partitioning regulation (Golz et al. 2018; Li, Zhang, Chen, Ji & Yu 2018; Chen & Wang 2019). The combinations of specific MYB and bHLH proteins together with TTG1, a WD40 domain repeats (WDR) protein allows the regulation of each of these traits. It constitutes the MYB-bHLH-WD40 repeat (MBW) regulatory complexes highly conserved across Angiosperms (Zhang & Hülskamp 2019; Zhang, Chopra, Schrader & Hülskamp 2019). It is important to note that (i) TTG1 is common to all aforementioned traits, (ii) each bHLH protein is involved in the regulation of two or more traits and (iii) each MYB mostly controls only one trait (Zhang et al. 2019; Figure 2A). In A. thaliana , the ttg1 mutant lacks root hair, trichome, has a reduced level of anthocyanins and proanthocyanins and does not accumulate mucilage in the MSCs (Western 2001). In A. thaliana , the phosphorylation of TTG1 by SHAGGY-like kinases 11/12 prevents its interaction with TT2, a MYB member of MBW complex, which decreases the transcription of the downstream regulator GL2 (Li et al. 2018; Figure 2A). The consequence for the seeds is the promotion of lipid storage in the embryo at the expense of mucilage and flavonoid pigment synthesis in the seed coat (Li et al. 2018; Figure 2A). This regulation is probably responsible of the differential balance between seed lipid and pigment/mucilage contents in two natural Medicagospecies that correlate with GL2 expression level (Song et al. 2017; Figure 2A). However, additional regulation mechanisms through interaction, competition, ubiquitination or epigenetics chromatin modifications may also occur (Xu, Dubos & Lepiniec 2015; Nguyen, Tran & Nguyen 2019). Within the three members of the WDR family existing in angiosperm species, TTG1 is the most recent one, appearing in the common ancestor of seed plants (Airoldi, Hearn, Brockington, Webb & Glover 2019) for the control of epidermal cells differentiation in essentially all organs of plants (Figure 2B). Interestingly, the serine 215 that can be phosphorylated by SK11/12 is conserved across seed plant TTG1 orthologs suggesting an ancestral function for this master regulator allowing switches in carbon flow between the seed coat and the embryo independently of the SM presence (Li et al. 2018).

It appears that some proteins belonging to the MBW complex were already present prior the emergence of the vascular plants (Doroshkov, Konstantinov, Afonnikov & Gunbin 2019), and that the control of flavonoid synthesis could be the most ancient trait controlled by this complex. This is coherent with the rescue of anthocyanin synthesis byPAP1 and PAP2 angiosperm MYB ortholog genes in theA. thaliana double mutant pap1pap2 and the absence of restoration of trichome differentiation in the A. thalianagl1 mutant by GL1 MYB ortholog genes coming from the same species (Zhang & Hülskamp 2019). Since bHLH and MYB proteins have been subjected to numerous and recent duplication events (Doroshkov et al. 2019; Sullivan et al. 2019), the different combinations of bHLHs and MYBs within MBW complexes have probably been co-opted to control the emergence of other biological process often linked to epidermal cells such as SM (Figure 2B). Indeed, the A. thalianaMBW regulatory complex controlling SM and seed coat pigments involves at least 2 bHLHs (TT8-EGL3), 1 MYB (TT2) together with TTG1 (reviewed in (Golz et al. 2018). For SM proper establishment there is a need of at least 3 additional MYBs (MYB5-MYB23-MYB61) (Figure 2A). Functional conservation of TTG1 in mucilage production and release has been demonstrated in the two Brassicaceae species Matthiola incana(Dressel & Hemleben 2009) and Arabis alpina (Chopra et al. 2014), and in M. truncatula (Pang et al. 2009). Orthologs of TTG1 from Camellia sinensis (Liu et al. 2018b) or even from the monocotyledonous species Setaria italica can restore mucilage wild type phenotype of A. thalianattg1 mutant through the recovery of GL2 and MUM4gene expression (Liu et al. 2017). Using a similar trans-complementation approach, the functional conservation of the two bHLHs EGL3 and TT8 has also been demonstrated in all tested angiosperm for all 5 traits (Zhang & Hülskamp 2019). MYB5 and TT2ortholog genes in M. truncatula have a conserved function of seed coat pigment and mucilage positive regulation (Liu, Jun & Dixon 2014) suggesting that the pathway regulation in seed through MBW complex dedicated to SM and seed coat pigments is conserved among the Rosids clade (Figure 2B). A. thaliana GL3 , another bHLH member of the MBW complex known to be implicated in anthocyanin, trichome and root hair, is unable to rescue SM in gl3 /egl3 /tt8 triple mutant. However, trans-complementation of the same triple mutant withA. alpina GL3 partially rescues SM (Zhang & Hülskamp 2019) indicating an intra- Brassicaceae divergency between the two proteins coherent with the contrasted morphology of MSCs compared to those of A. thaliana (Chopra et al. 2014). Similarly, in each Rosid family, few changes can be expected in the SM-related MBW complex and higher divergence in the downstream genes responsible of the observed morphological diversity of MSCs such as in Brassicaceae (Vaughan & Whitehouse 1971). As phylogeny of these multigenic family is difficult to solve and since their association in MBW complexes depend of non-binary competitive interaction (Zhang et al. 2019), more studies will be helpful to fully characterize the evolution of this protein complexes. A. thaliana NARS1, NARS2 and AP2 are also major regulators in A. thaliana controlling proper establishment of seed coat cells layers as well as embryogenesis (Kunieda et al. 2008). As they seem to act in parallel to TTG1 and GL2 and to control different downstream genes (Golz et al. 2018), investigation of their evolution in Angiosperms would be also very instructive.

Altogether, these results suggest that during seed plant evolution, TTG1 first appeared to balance carbon flow in seed tissues. It progressively interacted with bHLH members allowing more regulatory functions through a ternary complex modularity, with numerous and versatile recruitment of MYB members for deeper specialization and control of each traits in different seed zones such as SM in the MSCs.

Since SM is costly for mother plant metabolism, its presence implies that it displays major functions and that this trait is under a positive selection pressure in the myxospermic/myxodiasporic species. According to SM sticky and hydrophilic properties, the scientific community first investigated its influence in seed dispersal and germination (reviewed in Western 2012; Yang et al. 2012c) and more recently moved to the potential interactions between SM and the abiotic and biotic constraints.

A seed adaptation such as SM is expected to have an effect on seed dispersal and germination (Figure 3A). However, between closely relative species, these roles can be completely different making difficult to extend the concept to all myxodiasporic species. Counter-intuitively, SM can be a negative regulator of seed germination in Leptocereus scopulophilus (Barrios, Flores, González-Torres & Palmarola 2015) and also for the achene mucilage of Artemisia monosperma (Huang & Gutterman 1999a). In Blepharis persica , the SM could block oxygen transfer under water excess and then prevent germination (Witztum, Gutterman & Evenari 1969). This function was regularly re-emphasized (last time in Gorai et al. 2014) though never fully demonstrated. However, mucilage can as supposed also improve germination (Figure 3A-1). The pioneer most cited SM function in A. thaliana was a positive role during germination under osmotic stress conditions considering the polyethylene glycol (PEG)-dependent decrease of germination rate observed for myb61 , gl2 and ttg1mutants (Penfield 2001). On another MSC toolbox downstream gene mutant that presents a defect in SM extrusion, germination is also delayed suggesting a positive effect of mucilage on germination efficiency rather than on germination rate (Arsovski et al. 2009). However, no such phenotypes were obtained in the mum2 or myb61mutants (Saez-Aguayo et al. 2014). Thus, the role of SM inA. thaliana germination deserves to be deeply explored. ForS. hispanica , the intact myxocarpic achenes germinates clearly better than the achenes without mucilage (Geneve et al. 2017). Interestingly, SM-dependent better germination phenotypes are obtained with PEG application and not with salt at equal osmotic potential (Geneve et al. 2017). Upon five desert species (Lavandula subnuda, Lepidium aucheri, Boerhavia elegans, Plantago ciliata and Plantago amplexicaulis), SM presence increased water uptake but SM removal led to contrasted germination effects (Bhatt, Santo & Gallacher 2016). Therefore, mucilage function in germination seems to be related to water uptake and/or seed permeability to water and possibly to gazes.

Another interesting approach is the study of contrasted SM adaptation occurring within the same species on two seed morphotypes (myxospermic and non-myxospermic seeds). Those species use these dimorphic seeds to improve persistence and dispersion (Liu, Wang, Tanveer & Song 2018a). For three Brassicaceae species having characterized dimorphic seeds for myxospermy, namely Diptyocharpus strictus, Capsella bursa-pastorisi and Aethionema arabicum, the seed morphotype with higher dormancy does not have SM (Lu, Tan, Baskin & Baskin 2010; Tooropet al. 2012; Arshad et al. 2019) suggesting that myxospermic seeds should germinate without delay. The co-occurrence of SM and wings on seeds of D. strictus (Lu et al. 2010) andHenophyton deserti (Gorai et al. 2014) questions whether antitelochory and anemochory are opposite or can be complementary. Antitelochory prevents seed dispersion far from the mother plant while anemochory favors wind-driven dissemination under dry conditions until the seed encounters water and stops its dispersion. Combination of both traits can give a powerful advantage by an efficient dispersal until reaching an optimal place for hydric conditions (Figure 3A-2). Interestingly, Lunaria annua which possesses non-myxospermic but flattened and winged seeds shows a surprising use of mucilage by secreting it from the fruit to keep the seeds sticked, waiting for windy conditions to improve long distance dispersal of seed-carrying dry fruits by the wind (Leins, Fligge & Erbar 2018).

As SMs constitute highly hydrophilic gels, it is tempting to propose that SM may provide water for the embryo. For A. thaliana , SM takes a large amount of water from the environment but sequesters it due to ionic linkages with galacturonic acid residues (Figure 3B-3; Saez-Aguayo et al. 2014). Indeed, mutants that have a lack of mucilage, imbibed their seed faster than wild type or mutants with un-released mucilage (Saez-Aguayo et al. 2014). However, SM allows fast seed sinking in water as compared to non myxospermic seeds that can float on water surface for longer time and even germinate on it. Achenes of Artemisia sphaerocephala germinate and float better when their mucilage was removed (Huang & Gutterman 1999b). Interestingly, adhesive and frictional properties of SM can change according to its hydration level influencing at least its dispersal properties (Kreitschitz, Kovalev & Gorb 2015, 2016). In A. arabicum , the SM thick fibers emerging upon imbibition are able to conserve their structures and size upon dehydration (Lenser et al. 2016) allowing a trade-off of dispersal efficiency for the dry and the imbibed seeds, with wind, water run-off, and buoyancy dissemination ways, respectively (Arshad et al. 2019). A more complete understanding of the complex roles of SM in water management would necessitate investigating the SM role in more complex/natural situation such as succession of wetting and drying cycles, or flooding following a long drying period.

Soil physical properties can have a major impact on water availability and root penetration (Figure 3B-4). SM ofArtemisia sphaerocephala enhances seedling emergence in its sandy environment and reduces plant mortality (Yang, Baskin, Baskin, Liu & Huang 2012a). By adding SM extracted from C. bursa pastoris , soil rheological properties are modified particularly for hydraulic conductivity retaining water for longer time (Deng et al. 2014). A similar effect is provoked by S. hispanica SM addition which links soil particles to increase aggregate stability for at least 30 days in diverse kinds of soil (Figure 3B-4; Di Marsico et al.2018). So, SM could improve the soil rheological local environment in agreement with the non-disseminating lifestyle of species such asA. thaliana excreting non-adherent mucilage.

Similarly to root border cell mucilage (Knee et al. 2001), non-adherent SM constitutes an impressive amount of polysaccharides released in the environment. Taking into consideration the importance of micro-organisms in plant physiology along their development and their omnipresence around plant organs, this metabolic investment could indicate an involvement of SM on biotic constraints (Figure 3C). Involvement of SM on the influence of microbial community for the plant was shown in A. thaliana using the bacteria Streptomyces lividans that inhibits germination and growth of the pathogenic fungusVerticillium dahlia that causes the verticillium wilt. When both micro-organisms are co-inoculated on seeds, S. lividans has a better proliferation in SM in comparison to V. dahliaconsiderably reducing the plant disease symptoms (Meschke & Schrempf 2010). This “selective media effect” (Figure 3C-5) promoting microbial hyphae development was further illustrated with Salvia hispanicaachene mucilage and Colletotrichum graminicola fungi (Geneveet al. 2017). The desert plant Artemisia sphaerocephalaachene mucilage was shown to be degraded by microorganisms, providing CO₂ and soluble sugars to promote seedling establishment (Yang, Baskin, Baskin, Zhang & Huang 2012b). This promotion was recently explained in the same species through the mucilage positive effect on soil microbial community composition and diversity to favor fungal-bacterial interaction and soil enzyme activities, protecting young seedling from drought and pathogens (Hu et al. 2019b). Glomeromycota is one of the groups of fungal species positively impacted by achene mucilage. However, these fungi responsible of arbuscular mycorrhizal symbiosis do not significantly interact with mucilage at the seedling stage, acting probably independently to enhance seedling establishment (Hu et al. 2019a). Active stress-associated enzymes, such as nuclease, protease, and chitinase, are secreted from the seed coat of several species even in seeds several decades old (Raviv et al. 2017). However, this ability to secrete proteins is conserved in the non myxospermic species Raphanus sativus or in the A. thaliana gl2 mutant deprived of SM suggesting that protein secretion and SM are independent (Raviv et al. 2017).

Nematodes are in close interaction with plants and can have major pathogenic impact on plant development. A. thaliana SM contributes to attract root-knot nematodes (Figure 3C-6) with the additional requirement of seed-surface carbohydrates and proteins (Tsaiet al. 2019). Considering the parasitic nature of those nematodes, this attraction is more probably due to the nematode adaptation rather than a plant adaptation. Conversely, this nematode could be a predator of an even more dangerous organism for the plant.Capsella bursa-pastoris, a closely relative species of A. thaliana, has myxospermic seeds also able to attract nematodes (Roberts, Warren & Provan 2018). Surprisingly, it seems to be a case of protocarnivory because the massive death of trapped nematodes in SM increases plant development from germination to young seedling establishment, especially under low nutrient level (Roberts et al. 2018). Thus, SM involvement in biotic interactions starts to be uncovered promising new astonishing functions that could impact the plant development in an unexpected manner than the previously characterized functions

Myxodiaspory is a very diverse trait among Angiosperms that shows great variability both at the intra-species and inter-species levels. This astonishing morphological diversity makes it difficult to trace back the evolutionary origin of SM based on this sole trait. The available deeply characterized A. thaliana MSC toolbox genes now allows comparison of these molecular actors within a species or between species starting to shed light on this mysterious evolutionary story. The intense selection pressure that undergoes mucilage establishment is applied mainly on downstream genes of the MSC toolbox contrary to the SM-related MBW master regulatory complex that appears to be majoritarily conserved across Angiosperms. From this regulatory complex, mucilage may have evolved several times independently as a highly diverse trait allowing a wide range of ecological functions for the seed of each species facing contrasted environments. The biotic constraints are the least studied point in this field and probably the most promising track to uncover new SM integrated functions related to particular environments.

ACKNOWLEDGMENTS

The authors are thankful to Université Paul Sabatier-Toulouse III (France) and CNRS for supporting their research work. S.V. benefited from a PhD scholarship funded by the University Paul Sabatier-Toulouse III. This work was also supported by the French Laboratory of Excellence project “TULIP” (ANR-10-LABX-41; ANR-11-IDEX-0002-02) and the French National Research Agency project “MicroWall” (ANR-18-CE20-0007). We would like to thank Charles Uhlmann for the english editing.

FIGURE LEGENDS:

Figure 1: Arabidopsis and flax as model species for seed mucilage secretory cell (MSC) illustrating the diversity of development, intracellular polysaccharidic mucilage organization, and extrusion mechanisms upon water imbibition. Kinetics of MSC development in the historical model Arabidopsis thaliana on the left (adapted from Western 2001; Francoz et al. 2019), and in the emerging modelLinum usitatissimum on the right (adapted from Miart et al. 2019). The 30-50 µm wide cells from both species are drawn at similar scale, the released mucilage layers are not drawn to scale. Note the numerous differences between both models: Major features in A. thaliana : presence of a volcano-shaped columella, simultaneous rupture of primary cell wall domain in each MSC cells and distinction between adherent mucilage (am) and non-adherent mucilage layers (nam). Major features in L. usitatissimum : sequential synthesis of four highly different mucilage layers (m1 to m4) and sequential rupture of primary wall domains expending from cell to cell. DAP=day after pollination

Figure 2: Evolution of MYB-bHLH-WDR (MBW) complexes regulating spatiotemporal carbon partitioning in A. thaliana seeds. (A) The conserved ancestral master regulator TTG1 together with a bHLH and MYB modularity enables to regulate the spatiotemporal specificity of various seed traits including SM production and release (Golz et al.2018; Li et al. 2018; Chen & Wang 2019). (B) Simplified phylogenetic tree of land plants giving an overview of the sequential evolution of the A. thaliana MSC toolbox from upstream MBW complex members until downstream directs actors of SM morphology so far characterized in A. thaliana (Liu et al. 2014, 2017; Liet al. 2018; Airoldi et al. 2019; Zhang & Hülskamp 2019; Zhang et al. 2019).

Figure 3: Global overview of seed mucilage (SM) major functions facing environmental constraints. (A) SM may influence plant development through (1) positive or negative impact on germination upon inappropriate condition, depending on the species, or (2) seed dispersal in relation to conferred seed physical properties such as sinking ability or soil anchoring (B) SM is modified by abiotic conditions in close environment through (3) regulation of water flux and water availability as well as (4) soil rheological remodelling properties. (C) SM is involved in biotic interactions though (5) direct or indirect influence on microbial community establishment around the seed and the future plant and (6) attraction ability of nematodes.

Vincent Burlat https://orcid.org/0000-0002-0897-6011

Christophe Dunand https://orcid.org/0000-0003-1637-4042

Airoldi C.A., Hearn T.J., Brockington S.F., Webb A.A.R. & Glover B.J. (2019) TTG1 proteins regulate circadian activity as well as epidermal cell fate and pigmentation. Nature Plants 5 , 1145–1153.

Arshad W., Sperber K., Steinbrecher T., Nichols B., Jansen V.A.A., Leubner-Metzger G. & Mummenhoff K. (2019) Dispersal biophysics and adaptive significance of dimorphic diaspores in the annualAethionema arabicum (Brassicaceae). New Phytologist221 , 1434–1446.

Arsovski A.A., Popma T.M., Haughn G.W., Carpita N.C., McCann M.C. & Western T.L. (2009) AtBXL1 encodes a bifunctional  -D-Xylosidase/ -L-Arabinofuranosidase required for pectic arabinan modification in arabidopsis mucilage secretory cells. Plant Physiology 150 , 1219–1234.

Baroux C. & Grossniklaus U. (2019) Seeds—An evolutionary innovation underlying reproductive success in flowering plants. Current Topics in Developmental Biology 131 , 605–642.

Barrios D., Flores J., González-Torres L.R. & Palmarola A. (2015) The role of mucilage in the germination of Leptocereus scopulophilus(Cactaceae) seeds from Pan de Matanzas, Cuba . Botany93 , 251–255.

Ben-Tov D., Idan-Molakandov A., Hugger A., Ben-Shlush I., Günl M., Yang B., … Harpaz-Saad S. (2018) The role of COBRA-LIKE 2 function, as part of the complex network of interacting pathways regulatingArabidopsis seed mucilage polysaccharide matrix organization.Plant Journal 94 , 497–512.

Bhatt A., Santo A. & Gallacher D. (2016) Seed mucilage effect on water uptake and germination in five species from the hyper-arid Arabian desert. Journal of Arid Environments 128 , 73–79.

Chen S. & Wang S. (2019) GLABRA2, a common regulator for epidermal cell fate determination and anthocyanin biosynthesis in Arabidopsis .International Journal of Molecular Sciences 20 , 1–13.

Chopra D., Wolff H., Span J., Schellmann S., Coupland G., Albani M.C., … Hülskamp M. (2014) Analysis of TTG1 function in Arabis alpina . BMC Plant Biology 14 , 16.

Deng W., Hallett P.D., Jeng D.S., Squire G.R., Toorop P.E. & Iannetta P.P.M. (2014) The effect of natural seed coatings of Capsella bursa-pastoris L. Medik. (shepherd’s purse) on soil-water retention, stability and hydraulic conductivity. Plant and Soil387 , 167–176.

Deng W., Jeng D.S., Toorop P.E., Squire G.R. & Iannetta P.P.M. (2012) A mathematical model of mucilage expansion in myxospermous seeds ofCapsella bursa-pastoris (shepherds purse). Annals of Botany 109 , 419–427.

Doroshkov A. V., Konstantinov D.K., Afonnikov D.A. & Gunbin K. V. (2019) The evolution of gene regulatory networks controlling Arabidopsis thaliana L. trichome development. BMC Plant Biology 19 , 53.

Dressel A. & Hemleben V. (2009) Transparent Testa Glabra 1 (TTG1) and TTG1-like genes in Matthiola incana R. Br. and related Brassicaceae and mutation in the WD-40 motif. Plant Biology11 , 204–212.

Fabrissin I., Cueff G., Berger A., Granier F., Sallé C., Poulain D., … North H.M. (2019) Natural variation reveals a key role for rhamnogalacturonan I in seed outer mucilage and underlying genes.Plant Physiology , 181 , 1498-1518

Francoz E., Ranocha P., Burlat V. & Dunand C. (2015) Arabidopsis seed mucilage secretory cells: Regulation and dynamics. Trends in Plant Science 20 , 515–524.

Francoz E., Ranocha P., Le Ru A., Martinez Y., Fourquaux I., Jauneau A., … Burlat V. (2019) Pectin demethylesterification generates platforms that anchor peroxidases to remodel plant cell wall domains.Developmental Cell 48 , 261-276.e8.

Galloway A.F., Knox P. & Krause K. (2020) Sticky mucilages and exudates of plants: putative microenvironmental design elements with biotechnological value. New Phytologist 225 , 1461-1469.

Geneve R.L., Hildebrand D.F., Phillips T.D., Al-Amery M. & Kester S.T. (2017) Stress influences seed germination in mucilage-producing chia.Crop Science 57 , 2160–2169.

Golz J.F., Allen P.J., Li S.F., Parish R.W., Jayawardana N.U., Bacic A. & Doblin M.S. (2018) Layers of regulation – Insights into the role of transcription factors controlling mucilage production in theArabidopsis seed coat. Plant Science 272 , 179–192.

Gorai M., El Aloui W., Yang X. & Neffati M. (2014) Toward understanding the ecological role of mucilage in seed germination of a desert shrubHenophyton deserti : Interactive effects of temperature, salinity and osmotic stress. Plant and Soil 374 , 727–738.

Griffiths J.S. & North H.M. (2017) Sticking to cellulose: exploitingArabidopsis seed coat mucilage to understand cellulose biosynthesis and cell wall polysaccharide interactions. New Phytologist 214 , 959–966.

Hu D., Baskin J.M., Baskin C.C., Wang Z., Zhang S., Yang X. & Huang Z. (2019a) Arbuscular mycorrhizal symbiosis and achene mucilage have independent functions in seedling growth of a desert shrub.Journal of Plant Physiology 232 , 1–11.

Hu D., Zhang S., Baskin J.M., Baskin C.C., Wang Z., Liu R., … Huang Z. (2019b) Seed mucilage interacts with soil microbial community and physiochemical processes to affect seedling emergence on desert sand dunes. Plant Cell and Environment 42 , 591–605.

Huang D., Wang C., Yuan J., Cao J. & Lan H. (2015) Differentiation of the seed coat and composition of the mucilage of Lepidium perfoliatum L.: A desert annual with typical myxospermy. Acta Biochimica et Biophysica Sinica 47 , 775–787.

Huang Z. & Gutterman Y. (1999a) Water absorption by mucilaginous achenes of Artemisia monosperma : Floating and germination as affected by salt concentrations. Israel Journal of Plant Sciences47 , 27–34.

Huang Z. & Gutterman Y. (1999b) Germination of Artemisia sphaerocephala (Asteraceae), occurring in the sandy desert areas of Northwest China. South African Journal of Botany 65 , 187–196.

Jones V.A.S. & Dolan L. (2012) The evolution of root hairs and rhizoids. Annals of botany 110 , 205–212.

Knee E.M., Gong F.C., Gao M., Teplitski M., Jones A.R., Foxworthy A., … Bauer W.D. (2001) Root mucilage from pea and its utilization by rhizosphere bacteria as a sole carbon source. Molecular Plant-Microbe Interactions 14 , 775–784.

Kreitschitz A. & Gorb S.N. (2017) How does the cell wall ‘stick’ in the mucilage? A detailed microstructural analysis of the seed coat mucilaginous cell wall. Flora: Morphology, Distribution, Functional Ecology of Plants 229 , 9–22.

Kreitschitz A. & Gorb S.N. (2018) The micro- and nanoscale spatial architecture of the seed mucilage—Comparative study of selected plant species. PLoS ONE 13 , e0200522..

Kreitschitz A., Kovalev A. & Gorb S.N. (2015) Slipping vs sticking: Water-dependent adhesive and frictional properties of Linum usitatissimum L. seed mucilaginous envelope and its biological significance. Acta Biomaterialia 17 , 152–159.

Kreitschitz A., Kovalev A. & Gorb S.N. (2016) “Sticky invasion” – the physical properties of Plantago lanceolata L. seed mucilage .Beilstein Journal of Nanotechnology 7 , 1918–1927.

Kunieda T., Mitsuda N., Ohme-Takagi M., Takeda S., Aida M., Tasaka M., … Hara-Nishimura I. (2008) NAC Family Proteins NARS1/NAC2 and NARS2/NAM in the Outer Integument Regulate Embryogenesis in Arabidopsis.The Plant Cell 20 , 2631–2642.

Kunieda T., Shimada T., Kondo M., Nishimura M., Nishitani K. & Hara-Nishimura I. (2013) Spatiotemporal secretion of PEROXIDASE36 is required for seed coat mucilage extrusion in Arabidopsis .The Plant Cell 25 , 1355–1367.

Leins P., Fligge K. & Erbar C. (2018) Silique valves as sails in anemochory of Lunaria (Brassicaceae). Plant Biology20 , 238–243.

Lenser T., Graeber K., Cevik Ö.S., Adigüzel N., Dönmez A.A., Grosche C., … Leubner-Metzger G. (2016) Developmental control and plasticity of fruit and seed dimorphism in Aethionema arabicum . Plant Physiology 172 , 1691–1707.

Li C., Zhang B., Chen B., Ji L. & Yu H. (2018) Site-specific phosphorylation of TRANSPARENT TESTA GLABRA1 mediates carbon partitioning in Arabidopsis seeds. Nature Communications9 , 571.

Liu C., Jun J.H. & Dixon R.A. (2014) MYB5 and MYB14 play pivotal roles in seed coat polymer biosynthesis in Medicago truncatula .Plant Physiology 165 , 1424–1439.

Liu J., Shim Y.Y., Shen J., Wang Y., Ghosh S. & Reaney M.J.T. (2016) Variation of composition and functional properties of gum from six Canadian flaxseed (Linum usitatissimum L.) cultivars.International Journal of Food Science and Technology 51 , 2313–2326.

Liu K., Qi S., Li D., Jin C., Gao C., Duan S., … Chen M. (2017) TRANSPARENT TESTA GLABRA 1 ubiquitously regulates plant growth and development from Arabidopsis to foxtail millet (Setaria italica ). Plant Science 254 , 60–69.

Liu R., Wang L., Tanveer M. & Song J. (2018a) Seed heteromorphism: an important adaptation of halophytes for habitat heterogeneity.Frontiers in Plant Science 9 , 1515.

Liu Y., Hou H., Jiang X., Wang P., Dai X., Chen W., … Xia T. (2018b) A WD40 repeat protein from Camellia sinensis regulates anthocyanin and proanthocyanidin accumulation through the formation of MYB–bHLH–WD40 ternary complexes. International Journal of Molecular Sciences 19 , E1686.

Lloyd A., Brockman A., Aguirre L., Campbell A., Bean A., Cantero A. & Gonzalez A. (2017) Advances in the MYB-bHLH-WD Repeat (MBW) pigment regulatory model: Addition of a WRKY factor and co-option of an anthocyanin MYB for betalain regulation. Plant and Cell Physiology 58 , 1431–1441.

Lu J., Tan D., Baskin J.M. & Baskin C.C. (2010) Fruit and seed heteromorphism in the cold desert annual ephemeral Diptychocarpus strictus (Brassicaceae) and possible adaptive significance.Annals of Botany 105 , 999–1014.

Macquet A., Ralet M.-C., Loudet O., Kronenberger J., Mouille G., Marion-Poll A. & North H.M. (2007) A naturally occurring mutation in anArabidopsis accession affects a β-d-galactosidase that increases the hydrophilic potential of rhamnogalacturonan I in seed mucilage .The Plant Cell 19 , 3990–4006.

Di Marsico A., Scrano L., Labella R., Lanzotti V., Rossi R., Cox L., … Amato M. (2018) Mucilage from fruits/seeds of chia (Salvia hispanica L.) improves soil aggregate stability.Plant and Soil 425 , 57–69.

Meschke H. & Schrempf H. (2010) Streptomyces lividans inhibits the proliferation of the fungus Verticillium dahliae on seeds and roots of Arabidopsis thaliana. Microbial Biotechnology3 , 428–443.

Miart F., Fournet F., Dubrulle N., Petit E., Demailly H., Dupont L., … Pageau K. (2019) Cytological approaches combined with chemical analysis reveals the layered nature of flax mucilage. Frontiers in Plant Science 10 , 684.

Mirhosseini H. & Amid B.T. (2012) A review study on chemical composition and molecular structure of newly plant gum exudates and seed gums. Food Research International 46 , 387–398.

Nguyen C.T., Tran G.B. & Nguyen N.H. (2019) The MYB–bHLH–WDR interferers (MBWi) epigenetically suppress the MBW’s targets.Biology of the Cell 111 , 284–291.

North H.M., Berger A., Saez-Aguayo S. & Ralet M.C. (2014) Understanding polysaccharide production and properties using seed coat mutants: Future perspectives for the exploitation of natural variants. Annals of Botany 114 , 1251–1263.

Pang Y., Wenger J.P., Saathoff K., Peel G.J., Wen J., Huhman D., … Dixon R.A. (2009) A WD40 repeat protein from Medicago truncatula is necessary for tissue-specific anthocyanin and proanthocyanidin biosynthesis but not for trichome development.Plant Physiology 151 , 1114–1129.

Penfield S. (2001) MYB61 ss required for mucilage deposition and extrusion in the Arabidopsis seed coat. The Plant Cell13 , 2777–2791.

Phan J.L. & Burton R.A. (2018) New insights into the composition and structure of seed mucilage. In Annual Plant Reviews . pp. 1–41. John Wiley & Sons, Ltd, Chichester, UK.

Poulain D., Botran L., North H.M. & Ralet M.C. (2019) Composition and physicochemical properties of outer mucilage from seeds ofArabidopsis natural accessions. AoB PLANTS 11 , plz031.

Raviv B., Aghajanyan L., Granot G., Makover V., Frenkel O., Gutterman Y. & Grafi G. (2017) The dead seed coat functions as a long-term storage for active hydrolytic enzymes. PLoS ONE 12 , e0181102.

Renzaglia K.S., Duff R.J., Nickrent D.L. & Garbary D.J. (2000) Vegetative and reproductive innovations of early land plants: Implications for a unified phylogeny. Philosophical Transactions of the Royal Society B: Biological Sciences 355 , 769–793.

Roberts H.R., Warren J.M. & Provan J. (2018) Evidence for facultative protocarnivory in Capsella bursa-pastoris seeds. Scientific Reports 8 , 10120.

Ryding O. (2001) Myxocarpy in the Nepetoideae (Lamiaceae) with Notes on Myxodiaspory in General. Systematics and Geography of Plants71 , 503–514.

Saez-Aguayo S., Ralet M.-C., Berger A., Botran L., Ropartz D., Marion-Poll A. & North H.M. (2013) PECTIN METHYLESTERASE INHIBITOR6 promotes Arabidopsis mucilage release by limiting methylesterification of homogalacturonan in seed coat epidermal cells.The Plant Cell 25 , 308–323.

Saez-Aguayo S., Rondeau-Mouro C., Macquet A., Kronholm I., Ralet M.C., Berger A., … North H.M. (2014) Local evolution of seed flotation in Arabidopsis . PLoS Genetics 10 , e1004221.

Šola K., Dean G.H. & Haughn G.W. (2019) Arabidopsis seed mucilage: a specialised extracellular matrix that demonstrates the structure–function versatility of cell wall polysaccharides.Annual Plant Reviews 2 , 1085–1116.

Song Y., He L., Wang X.-D., Smith N., Wheeler S., Garg M.L. & Rose R.J. (2017) Regulation of carbon partitioning in the seed of the model legumeMedicago truncatula and Medicago orbicularis : A Comparative Approach. Frontiers in Plant Science 8 , 2070.

Soto-Cerda B.J., Cloutier S., Quian R., Gajardo H.A., Olivos M. & You F.M. (2018) Genome-wide association analysis of mucilage and hull content in flax (Linum usitatissimum l.) seeds.International Journal of Molecular Sciences 19 , E2870.

Soukoulis C., Gaiani C. & Hoffmann L. (2018) Plant seed mucilage as emerging biopolymer in food industry applications. Current Opinion in Food Science 22 , 28–42.

Sullivan A.M., Arsovski A.A., Thompson A., Sandstrom R., Thurman R.E., Neph S., … Queitsch C. (2019) Mapping and dynamics of regulatory DNA in maturing Arabidopsis thaliana siliques. Frontiers in Plant Science 10 , 1434.

Toorop P.E., Campos Cuerva R., Begg G.S., Locardi B., Squire G.R. & Iannetta P.P.M. (2012) Co-adaptation of seed dormancy and flowering time in the arable weed Capsella bursa-pastoris (shepherds purse).Annals of Botany 109 , 481–489.

Tsai A.Y.L., Higaki T., Nguyen C.N., Perfus-Barbeoch L., Favery B. & Sawa S. (2019) Regulation of root-knot nematode behavior by seed-coat mucilage-derived attractants. Molecular Plant 12 , 99–112.

Tucker M.R., Ma C., Phan J., Neumann K., Shirley N.J., Hahn M.G., … Burton R.A. (2017) Dissecting the genetic basis for seed coat mucilage heteroxylan biosynthesis in Plantago ovata using gamma irradiation and infrared spectroscopy. Frontiers in Plant Science8 , 326.

Vaughan J.G. & Whitehouse J.M. (1971) Seed structure and the taxonomy of the Cruciferae. Botanical Journal of the Linnean Society64 , 383–409.

Voiniciuc C., Yang B., Schmidt M.H.W., Günl M. & Usadel B. (2015) Starting to gel: How Arabidopsis seed coat epidermal cells produce specialized secondary cell walls. International Journal of Molecular Sciences 16 , 3452–3473.

Voiniciuc C., Zimmermann E., Schmidt M.H.-W., Günl M., Fu L., North H.M. & Usadel B. (2016) Extensive natural variation in Arabidopsisseed mucilage structure. Frontiers in Plant Science 7 , 803.

Weitbrecht K., Müller K. & Leubner-Metzger G. (2011) First off the mark: Early seed germination. Journal of Experimental Botany62 , 3289–3309.

Western T.L. (2001) Isolation and characterization of mutants defective in seed coat mucilage secretory cell development in Arabidopsis .Plant Physiology 127 , 998–1011.

Western T.L. (2012) The sticky tale of seed coat mucilages: Production, genetics, and role in seed germination and dispersal. Seed Science Research 22 , 1–25.

Williams M.A.K., Cornuault V., Irani A.H., Symonds V.V., Malmstrom J., An Y., … North H.M. (2020) Polysaccharide structures in the outer mucilage of Arabidopsis seeds visualised by AFM.Biomacromolecules . DOI: 10.1021/acs.biomac.9b01756.

Witztum A., Gutterman Y. & Evenari M. (1969) Integumentary Mucilage as an oxygen barrier during germination of Blepharis persica .Botanical Gazette 130 , 238–241.

Xu W., Dubos C. & Lepiniec L. (2015) Transcriptional control of flavonoid biosynthesis by MYB-bHLH-WDR complexes. Trends in Plant Science 20 , 176–185.

Yang X., Baskin C.C., Baskin J.M., Liu G. & Huang Z. (2012a) Seed mucilage improves seedling emergence of a sand desert shrub. PloS one 7 , e34597..

Yang X., Baskin C.C., Baskin J.M., Zhang W. & Huang Z. (2012b) Degradation of seed mucilage by soil microflora promotes early seedling growth of a desert sand dune plant. Plant, Cell and Environment35 , 872–883.

Yang X., Baskin J.M., Baskin C.C. & Huang Z. (2012c) More than just a coating: Ecological importance, taxonomic occurrence and phylogenetic relationships of seed coat mucilage. Perspectives in Plant Ecology, Evolution and Systematics 14 , 434–442.

Yu L., Lyczakowski J.J., Pereira C.S., Kotake T., Yu X., Li A., … Dupree P. (2018) The patterned structure of galactoglucomannan suggests it may bind to cellulose in seed mucilage. Plant Physiology178 , 1011–1026.

Yu L., Yakubov G.E., Zeng W., Xing X., Stenson J., Bulone V. & Stokes J.R. (2017) Multi-layer mucilage of Plantago ovata seeds: Rheological differences arise from variations in arabinoxylan side chains. Carbohydrate Polymers 165 , 132–141.

Zhang B., Chopra D., Schrader A. & Hülskamp M. (2019) Evolutionary comparison of competitive protein-complex formation of MYB, bHLH, and WDR proteins in plants. Journal of Experimental Botany70 , 3197–3209.

Zhang B. & Hülskamp M. (2019) Evolutionary analysis of MBW function by phenotypic rescue in Arabidopsis thaliana . Frontiers in Plant Science 10 , 375.

Zhao X., Qiao L. & Wu A.M. (2017) Effective extraction ofArabidopsis adherent seed mucilage by ultrasonic treatment.Scientific Reports 7 , 40672.