Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Apr 16;250(5):3460–3474. doi: 10.1111/nph.71167

The abundance of pollen coat small signaling proteins shows limited convergence between independent selfing transitions in Arabidopsis and Capsella

Ömer İltaş 1,2, Luciana Salomon 1, Martin Kovačik 1,3, Martin Čertner 1,4, Pavel Talacko 5, Karel Harant 5, Said Hafidh 6, Clément Lafon Placette 1,✉
PMCID: PMC13150313  PMID: 41992796

Summary

  • In plants, a key example of convergence is the repeated evolution of floral traits associated with the transition from outcrossing to self‐fertilization, often resulting in the ‘selfing syndrome’ (e.g. reduced flower size and loss of scent). However, potentially overlooked changes concern the pollen coat, which plays a role in different aspects of outcrossing strategy.

  • To investigate this, we compared pollen coat thickness and pollen proteomes between selfing and outcrossing species in the Arabidopsis and Capsella genera, representing three independent selfing transitions.

  • Pollen coat area diminished with the age of the selfing transition, with significant reductions in A. thaliana and Capsella rubella, but not in A. lyrata (recent transition, no selfing syndrome). Moreover, specific differentially abundant pollen coat proteins were shared across selfer–outcrosser contrasts more than expected by chance, even though the direction of changes was not consistent across the three comparisons. These specific proteins, while annotated as pathogen response due to their structure, are likely involved in pollen–pistil interactions instead.

  • Overall, our findings reveal limited convergent evolution of pollen coat thickness and protein composition following selfing transitions. Our work identified a group of pollen coat proteins that are largely understudied, but whose abundance appears to be a hotspot for evolutionary changes.

Keywords: convergence, pathogen response, pollen coat, pollen–pistil interactions, proteomics, selfing transition

Introduction

Sexual reproduction strategies in flowering plants display large variation, and this variation can have important consequences on selective processes, genome evolution, and reproductive traits evolution (Barrett, 2002; Sicard & Lenhard, 2011). The shift from outcrossing to self‐fertilization is one of the most widespread evolutionary transitions in plant reproduction, having occurred recurrently and independently across angiosperms (Stebbins, 1974; Igic et al., 2006; Sicard & Lenhard, 2011). Initially, such a transition can be advantageous because of reproductive assurance under pollinator scarcity (Kennedy & Elle, 2008). However, in the long term, selfing is considered an evolutionary dead end because of its fitness costs related to increased homozygosity, reduced effective population size, and weakened efficacy of selection (Barrett et al., 2014; Cutter, 2019). At the trait level, the transition to selfing is recurrently accompanied by characteristic changes in the morphology and function of floral traits (known as ‘selfing syndrome’; Sicard & Lenhard, 2011). In most cases, these changes are similar among independent taxa and lead to the vestigialization of floral traits involved in pollinator attraction, such as reduction in floral size and number, loss of odor and nectar production, reduced pollen‐to‐ovule ratios, and reduced herkogamy (Sicard & Lenhard, 2011; Tsuchimatsu & Fujii, 2022).

The transition to selfing is likely to impact pollen traits as well, even though these have been less studied. Pollen serves as the vital vehicle for delivering male gametes, ensuring the continuity of plant life through fertilization (Anderson & Minnaar, 2020). The journey of pollen grain from the anther to the receptive stigma is critical, as it must preserve its functional integrity to successfully initiate fertilization upon arrival. However, in nature, this journey is fraught with numerous stresses and challenges, from environmental pressures to potential pathogens before pollination. To navigate these challenges, pollen must be protected from herbivores and pathogens, shield against ultraviolet radiation, and maintain the components necessary for attracting and attaching to pollinators and interacting with the stigma (Pacini & Hesse, 2005; Rivest & Forrest, 2020). Pollen grains are coated with a specialized outermost layer, known as the pollen coat, which helps mitigate environmental and biotic stresses (Pacini & Hesse, 2005). It is rich in secretory elements and comprises a complex mixture of proteins, lipids, and pigments (Rejón et al., 2016). Considering these features, one would expect that pollen functions involved in survival of pollen grains and their successful transfer to the recipient stigma would be particularly important in outcrossing species due to the higher dependency on pollinator movement. By contrast, reproduction is assured in selfing species; thus, pollinator‐assisted pollen transfer is no longer needed, and the duration and distance of the pollen movement, within the same flower, is limited (Kalisz et al., 2004). Therefore, the importance of the pollen coat may be reduced in selfing species, along with the need for protection against environmental stresses, compared with outcrossing lineages. Although comparative studies on pollen coat variation between closely related species are limited, it has been shown that the transition from insect pollination to wind pollination leads to a corresponding decrease in pollen coat amount and density (Hesse, 1979). Moreover, few studies revealed a complex array of pollen coat proteins involved in defense against pathogens and herbivores, as well as in interactions with the pistil in model species like Arabidopsis and other plants (Gong et al., 2015; Rejón et al., 2016; Wang et al., 2023). However, it remains to be determined whether the transition to selfing affects pollen coat morphology and molecular composition.

The evolution of repeated similar phenotypical traits, known as convergence, is explained by a similar response of independent populations or species when faced with similar selective pressures (Stayton, 2015). Convergent evolution is widespread in nature, and has been widely observed among different taxa, like mammals, birds, insects and plants (Li et al., 2010; Woodhouse & Hufford, 2019; Zhang et al., 2020; Woźniak et al., 2024). At the genetic level, convergent evolution of similar traits can rely on mutations in the same gene (gene reuse) or different genes with similar functions (Stern, 2013). Recent studies revealed that the extent of gene reuse decreases with the divergence between the lineages showing convergent evolution, while convergence at the function level is more likely to happen between distant clades (Birkeland et al., 2020; Bohutínská et al., 2021). Regarding the convergent evolution of selfing syndrome, several studies have revealed the underlying molecular mechanisms. For example, flower scent evolution in Capsella revealed that the transition from outcrossing to self‐fertilization led to a reduction in fragrance, with certain compounds, such as β‐ocimene. The loss of β‐ocimene in the different selfing species appeared to have been caused by mutations in different genes (Woźniak et al., 2024). Besides mutations affecting protein sequence, convergent changes may be quantitative and affect mRNA and/or protein accumulation in a particular tissue. Such changes are likely to be hotspots for convergent evolution, as they have fewer pleiotropic effects, altering gene expression without necessarily changing the structure of the gene product (Stern, 2013). However, to what extent convergence happens at the protein accumulation level in the case of the transition to selfing remains to be tested.

The Brassicaceae family, including the Arabidopsis and Capsella genera, provides an excellent system for studying the effects of transition to self‐fertilization on repeated evolution of sexual traits. It exhibits a wide diversity of independent selfing transitions across different species, which are known to have occurred at various evolutionary timescales, ranging from recent, intermediate, and older. For instance, in Arabidopsis lyrata, the transition to selfing is a relatively recent event and is restricted to a few populations in North America and Siberia (Foxe et al., 2010; Kolesnikova et al., 2023), while the rest of the species is an obligate outcrosser (Schierup, 1998; Koch et al., 1999; Nasrallah et al., 2000). More specifically, in A. lyrata subsp. lyrata, multiple breakdowns of self‐incompatibility have been reported in the Great Lakes region c. 10 ka (Willi & Määttänen, 2010; Foxe et al., 2010). Due to this recent shift to selfing, these populations do not yet exhibit the characteristic reduction in floral traits associated with the selfing syndrome (Carleial et al., 2017). Selfing Capsella rubella, on the contrary, already exhibits strong reduction in floral traits caused by selfing syndrome, with its developmental and genetic bases well documented (Sicard & Lenhard, 2011). The breakdown of self‐incompatibility from its outcrossing ancestor and subsequent split from its sister outcrossing species Capsella grandiflora occurred c. 100 000–200 000 yr ago (Guo et al., 2009), making it a good model to study the effects of an intermediate timescale for transition to selfing. Finally, Arabidopsis thaliana is an annual, self‐compatible plant native to Europe, Asia, and Africa, and has been separated from other Arabidopsis species for c. 6 million years (Novikova et al., 2016). Recent studies suggest that the loss of self‐incompatibility in A. thaliana occurred c. 600 ka (Strütt et al., 2023). Arabidopsis thaliana exhibits a clear selfing syndrome. By contrast, its relative Arabidopsis arenosa is an obligate outcrossing species with an annual, biennial, or short‐lived perennial life cycle, widely distributed across Europe (Al‐Shehbaz & O'Kane, 2002).

In this study, using pairs of selfing and outcrossing diploid Brassicaceae lineages with recent, intermediate, and older selfing transition events, we aimed to address the following objectives: (1) to determine whether the transition to selfing similarly impacts the evolution of pollen coat thickness in independent lineages, (2) to assess whether these convergent changes in pollen coat thickness are accompanied by convergent changes in pollen coat protein composition, and (3) to investigate at what pace changes in pollen coat thickness and underlying proteins arise after the transition to selfing. To do so, we first compared pollen coat thickness between selfing and their closely related outcrossing lineage, representing three independent transitions to selfing in Arabidopsis and Capsella genera. We then generated protein profiles of pollen from these pairs and identified differentially abundant total pollen and pollen coat proteins between selfing and outcrossing species.

Materials and Methods

Plant material and growth conditions

To compare the pollen coat structure and pollen cytoplasm from selfer and outcrosser species, 100 plants from each selfer species, including Arabidopsis lyrata ssp. lyrata (L.) (one population), Arabidopsis thaliana (L.) (two populations), and Capsella rubella (Reut.) (one population) were cultivated. For outcrosser species, we selected 25 plants each from Arabidopsis arenosa (L.) Lawalrée (two populations), Arabidopsis lyrata ssp. petrea (L.) O'Kane & Al‐Shehbaz (one population), and Capsella grandiflora (Fauché & Chaub.) (one population) (Supporting Information Table S1). Seeds of these species were germinated on agar plates containing Murashige and Skoog (MS) salts, MES hydrate, and 0.8% (w/v) plant agar, adjusted to a pH of 5.8. Then, plates were placed in a growth chamber set to a 12 h : 12 h, 23°C : 13°C, light : dark photoperiod and cultivated for 2 wk. Following germination, seedlings were transferred to small pots filled with garden substrate. For species requiring vernalization, such as A. arenosa and A. lyrata, seedlings were initially grown under the short‐day conditions (8 h of light per day at 4°C) for 8 wk to induce flowering. While these species were under the vernalization process, other species that did not require vernalization were grown simultaneously to perform experiments in batches. These species included Arabidopsis thaliana, Capsella rubella, and Capsella grandiflora. They were germinated under a cycle of 12 h : 12 h, 23°C : 13°C, light : dark for 3 wk before being moved to standard conditions. Finally, all seedlings were then repotted into 0.5‐liter pots and transferred to a growth chamber to grow under long‐day conditions (16 h : 8 h, 21°C : 15°C, light : dark), where they were maintained until ready for experimental use.

Pollen collection

An adaptation of the device described by Johnson‐Brousseau & McCormick (2004) was built to harvest pollen from individual flowers: One 200‐μl pipette tip was coupled with a 1000‐μl pipette tip, incorporating a 50‐μm mesh in between to collect unwanted plant tissue and debris; a 5‐μm mesh was attached to the wider section of the 1000‐μl tip, where the pollen would be collected, and this was inserted in a flexible plastic tube. All the structure was sealed with tape and the tube was connected to a mini‐Diaphragm vacuum pump. To remove the pollen from the 5‐μm mesh from the pipette tip setup, the filter mesh was removed, placed in weighted Eppendorf tubes (to know the final pollen quantity), and pelleted by high‐speed centrifugation for c. 15 min.

To avoid a bias on pollen coat and its proteome due to pollen development stages, we only collected mature pollen (open anthers), at which stage the pollen coat and its proteins are fully deposited already (Lyu & Liang, 2025). In addition, to avoid any bias related to the age of the plant that could impact the comparison between species, we regularly collected pollen across the lifespan of the plants, from the flowering onset to the end of the flowering period, capturing age variation for all species.

TEM microscopy

For transmission electron microscopy (TEM) observation, the collected pollen pellets in tubes were fixed by adding 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) in a ratio of 1 : 100 for 24 h. After this time, the pellet was washed with a washing buffer solution (WB, 0.1 M cacodylate with 1.5% glucose). Subsequently, the sample was concentrated in agarose as follows: equal volume of 4% low melting agarose in 0.1 M cacodyl buffer, liquid at 37°C, was added to sample suspension, carefully resuspended, and transferred at 4°C for 20 min. The samples were rinsed in the same buffer and postfixed in 2% osmium tetroxide for 2 h. After washing, pollen samples were dehydrated through an ascending ethanol and acetone series and embedded in Araldite – Poly/Bed® 812 mixture. Thin ultra‐sections (70–90 nm) were cut on a Reichert‐Jung Ultracut E ultramicrotome and stained using uranyl acetate and lead citrate. Sections were examined and photographed using the JEOL JEM‐1011 electron microscope. Fine structure measurements were performed using a Veleta camera and iTem v.5.1 software (Olympus Soft Imaging Solution GmbH, Münster, Germany). For visualization of conspicuous membranes of mitochondria and endoplasmic reticulum (ER) has been used this variety of postfixation: 2% OsO4 + 0.8% K4Fe (CN6) in buffer (0.5 ml 4%OsO4 + 0.5 ml PBS + 0.008 g K4Fe (CN)6). Pictures were processed using Gimp v.2.10 to obtain a clear differentiation between the pollen coat and the remaining structures from the pollen grain. Finally, images were analyzed using the ImageJ program (https://imagej.nih.gov/ij/). The pollen coat was determined as the outermost layer of the pollen wall filling between the sexine spaces (blue area in Fig. 1a). Total area was measured from cytoplasm area starting from the inner layer of the pollen wall (red area in Fig. 1a). Given that absolute values might be biased by putative differences in pollen grain size, relative pollen coat area was obtained by dividing the area of the pollen coat by the total pollen area for each pollen grain. All measurements were conducted in a double‐blind manner (Table S2). Differences in relative pollen coat area between outcrossers and selfers across three species comparisons (A. aren.–A. thal., C. gran.–C. rub., A. lyr. SI–A. lyr. SC) were tested using a two‐way ANOVA, followed by post hoc Tukey contrasts. The variable was first standardized to a unit mean by relating all measured values to the mean of the respective SI line (serving as a control); this was done separately for each of the three species comparisons, and allowed analysis on the combined dataset, unbiased by any putative differences in pollen grain size and ultrastructure between different taxonomic groups. Standardized Relative Pollen Coat Area was square‐root transformed to improve the normality of model residuals, and explanatory variables included Mating System, Species Comparison, and their interaction.

Fig. 1.

Fig. 1

Open in a new tab

Differences in pollen coat size between selfer and outcrosser pollen in three species comparisons: Arabidopsis arenosa (SI) vs Arabidopsis thaliana (SC), Capsella grandiflora (SI) vs Capsella rubella (SC), and. (a) Transmission electron microscopy pictures of outcrosser C. grandiflora pollen grains. The red color shows the regions measured as pollen coat area while the blue color shows what was measured as total pollen area. (b) Boxplots showing variation in relative pollen coat area (pollen coat area divided by total pollen area) for each species. Dots indicate individual measurements. Significance was determined using Tukey contrasts following a two‐way ANOVA. Asterisks indicate significant differences (*, P < 0.05; ***, P < 0.001). SC, self‐compatible; SI, self‐incompatible.

Protein digestion

Pollen grains were washed by hexane. Hexane fraction was dried and resuspended in 100 mM triethylammonium bicarbonate (TEAB), and the same volume of dichloromethane was added. After shaking, samples were centrifuged and TEAB fraction was further reduced and alkylated by 40 mM chloroacetamide, 10 mM Tris(2‐carboxyethyl) phosphine (TCEP) for 5 min at 95°C. Core of pollen grains left after extraction to hexane were mixed with 100 mM TEAB containing 2% sodium deoxycholate and abrasive glass beads and homogenized using FastPrep bead beating grinder. After homogenization samples were centrifuged and supernatants were further reduced and alkylated by 40 mM chloroacetamide, 10 mM TCEP for 5 min at 95°C. Samples were further processed using SP3 beads (Hughes et al., 2019) on the Thermo KingFisher Flex Automated Extraction and Purification System in a 96‐well plate. Briefly, 65 μl of sample was added to 65 μl of 100% ethanol and mixed with the SP3 beads. After binding, the beads were washed three times with 80% ethanol. After washing, samples were digested in 50 mM TEAB at 40°C with 1 μg of trypsin for 2 h, then another 1 μg of trypsin was added and digested overnight. After digestion, samples were acidified with trifluoroacetic acid to 1% final concentration and peptides were desalted using in‐house made stage tips packed with C18 disks (CDS Empore, Oxford, PA, USA) according to Rappsilber et al. (2007).

nLC–MS 2 analysis

Nano reversed‐phase column (Aurora Ultimate TS 25 × 75 C18 UHPLC column; Ion Opticks, Oxfordshire, UK) was used for liquid chromatography‐mass spectrometry (LC‐MS) analysis. Mobile phase buffer A was composed of water and 0.1% formic acid. Mobile phase B was composed of acetonitrile and 0.1% formic acid. Samples were loaded onto the trap column (C18 PepMap100, 5 μm particle size, 300 μm × 5 mm; Thermo Fisher Scientific, Waltham, MA, USA) for 4 min at 18 μl min−1; loading buffer was composed of water, 2% acetonitrile, and 0.1% trifluoroacetic acid. Peptides were eluted with Mobile phase B gradient from 4% to 35% B in 15 min. The elution time was chosen after optimization in a pilot experiment, which proved sufficient to achieve high‐quality data while keeping the overall experiment cost at a reasonable level. Eluting peptide cations were converted to gas‐phase ions by electrospray ionization and analyzed on a Thermo Orbitrap Ascend (Thermo Scientific) by data‐dependent approach. Survey scans of peptide precursors from 350 to 1400 m/z were performed in orbitrap at 120 K resolution (at 200 m/z) with a 100% ion count target. Tandem MS was performed by isolation at 1.6 Da with the quadrupole, higher‐energy collisional dissociation fragmentation with normalized collision energy of 30% and 10 ms activation time. Fragmentation spectra were acquired in ion trap with scan rate set to Rapid. The MS2 ion count target was set to 150%, and the max injection time was 75 ms. Only those precursors with charge state 2–6 were sampled for MS2. The dynamic exclusion duration was set to 30 s with a 10 ppm tolerance around the selected precursor and its isotopes. Monoisotopic precursor selection was turned on. We optimized the MS cycle time and decided for a shorter value of 1.5 s, which is appropriate for this shorter gradient. This ensured a sufficient number of data points across each chromatographic peak – at least four – allowing for accurate quantification.

Protein quantification and data analysis

All data were analyzed and quantified with the Maxquant software v.2.4.13.0 (Cox & Mann, 2008). The false discovery rate (FDR) was set to 1% for both proteins and peptides, and we specified a minimum peptide length of seven amino acids. The Andromeda search engine was used for the MS/MS spectra search. Samples from each species were searched separately against the organism specific database downloaded from Uniprot in April 2024 (Arabidopsis arenosa, containing 22 979 entries, Arabidopsis lyrata, containing 32 097 entries, Arabidopsis thaliana, containing 27 449 entries. Capsella grandiflora and Capsella rubella were searched together, databases contained 1865 and 26 721 entries, respectively). Enzyme specificity was set as C‐terminal to Arg and Lys, also allowing cleavage at proline bonds and a maximum of two missed cleavages. Carbamidomethylation of cysteine was selected as fixed modification and N‐terminal protein acetylation and methionine oxidation as variable modifications. The ‘match between runs’ feature of Maxquant was used to transfer identifications to other LC‐MS/MS runs based on their masses and retention time (maximum deviation 0.7 min), and this was also used in quantification experiments.

To test the efficiency of pollen coat isolation vs cytosol contamination, the abundance of proteins known to be localized either in the pollen coat or cytosol in A. thaliana was compared in pollen coat samples and total pollen samples. We used relative intensities (intensity of a particular protein/total intensity of all proteins detected in a given sample) to make abundance comparable between samples (Table S3). Overall, across the species, pollen coat proteins showed enrichment in pollen coat samples and depletion in total pollen samples (either numerical or binary detected/not detected). Vice versa, cytosolic proteins showed enrichment in total pollen samples and depletion in pollen coat samples (Table S4).

Proteomics data and Gene Ontology analyses

All proteomics analyses were conducted using Perseus software v.1.6.15.0 (Tyanova et al., 2016). For each dataset, reverse hits and potential contaminants were filtered out, and intensity values were converted to binary logarithms. To be able to compare protein profiles of different species with each other, we used A. thaliana homologues for each species to identify differentially abundant proteins. Blast v.2.14.1+ (Camacho et al., 2009) with default parameters was used to link results from proteomic analysis with specific genes in current genome annotation of different species. Specifically, protein blast of A. thaliana, A. lyrata, A. arenosa, and C. rubella UniProt IDs were performed on respective genome annotations (Lamesch et al., 2012; Slotte et al., 2013; Rawat et al., 2015). Identified proteins from A. lyrata, A. arenosa, and C. rubella were then reciprocally blasted on A. thaliana to get additional information from homologous genes. One‐to‐many match results were discarded in order to increase the robustness of orthologue identification. Only orthologous (one‐to‐one match) were used for differential protein abundance analyses. To identify differentially abundant proteins for each pairwise comparison, a two‐sample t‐test was used for statistical comparisons. These P‐values were subsequently used for permutation‐based FDR calculation (< 0.05). Only proteins meeting the 0.05 FDR threshold were considered significant. In addition, fold‐change limitation was applied using the S0 parameter, as implemented according to the method described by Tusher et al. (2001). Statistical analyses were performed using the Perseus software v.1.6.15 (Tyanova et al., 2016). To compare the protein sets identified by differential abundance analysis, Venn diagrams (Hulsen, 2022) were generated for each pairwise comparison. To test whether the observed protein overlap was significantly greater than expected by random chance, given the sample size, superexacttest package v.1.1.0 was implemented in R (Wang et al., 2015). The common overlapped gene set among the three groups specific to each Venn diagram was used as the total number of background elements to establish the reference for evaluating the statistical significance of the observed overlaps. Finally, biological roles of the obtained differentially abundant proteins from pairwise comparisons were identified with Gene Ontology (GO) enrichment analysis. The GO enrichment analysis was performed using the Plantregmap online tool (Tian et al., 2019). Enriched GO terms, corrected for multiple comparisons using the Benjamini and Hochberg method (FDR < 0.05), were considered significant. q‐values (q < 0.05) were selected and transformed into negative fold changes (−log10) for presentation in bar charts, which were generated using the SRplot online platform (Tang et al., 2023).

Results

A decrease in pollen coat size accompanies the selfing syndrome

To test whether the transition to selfing affects pollen coat size, we compared pollen coat area in two‐dimensional pictures (Fig. 1a) between selfing (self‐compatible, SC) lineages and their outcrossing (self‐incompatible, SI) close relatives. These comparisons included outcrossing Arabidopsis arenosa (SI) with selfing Arabidopsis thaliana (SC), outcrossing Capsella grandiflora (SI) with selfing Capsella rubella (SC), and outcrossing Arabidopsis lyrata ssp. petraea (SI) with selfing Arabidopsis lyrata ssp. lyrata (SC).

Mating system shifts from outcrossing to selfing generally led to a significant decrease in pollen coat development (F 1,204 = 54.4, P < 0.001; Table S5), but there were also differences between different species scenarios (F 2,204 = 26.3, P < 0.001), which resulted in a significant interaction of both factors (F 2,204 = 28.6, P < 0.001). The pollen coat was significantly more developed in the outcrossing species A. arenosa and C. grandiflora compared with their selfing counterparts A. thaliana and C. rubella (Tukey contrasts; P < 0.001 and P = 0.034, respectively; Fig. 1b). In contrast, the pollen coat area in the selfing A. lyrata population was not significantly different from that of its outcrossing counterpart (Tukey contrasts; P = 0.463).

Differentially abundant pollen coat proteins between selfers and outcrossers are associated with pathogen and defense response

To test whether pollen coat proteins show divergence in abundance levels across the three independent selfer–outcrosser contrasts, we characterized the protein profiles of pollen coat from the six lineages and identified proteins differentially abundant in outcrossers compared with selfers separately for each of the three contrasts (Table S6). Using a principal component analysis (PCA) from the pollen coat proteome abundance data, no clear species grouping according to phylogenetic relationships appeared (Fig. S1a). However, PC1 tended to separate selfing species from outcrossing ones, but between‐sample differences appeared as a large source of variation.

We then focused on identifying the pollen coat proteome that is common to outcrossing and selfing A. lyrata. A total of 393 pollen coat proteins were detected at least in either selfing or outcrossing A. lyrata. To gain insights into the functional characteristics of the pollen coat proteomes, we conducted GO enrichment analyses comparing the proteome against the whole set of A. lyrata predicted proteins. ‘Modification of morphology or physiology of other organism’, ‘disruption of cells of other organism’, ‘killing of cells of other organism’, ‘cell killing’, and ‘defense response to other fungus’ were some of the most overrepresented GO terms within the biological process (q‐value < 0.05; Fig. S2a). Of these 393 total pollen coat proteins, 56 proteins showed significant differential abundance between the outcrosser vs selfer, with 41 proteins being upregulated and 15 proteins downregulated in the pollen of outcrossing compared with selfing A. lyrata (Fig. 2a; q‐value < 0.05). These 56 proteins showed 10 GO terms that were significantly enriched compared with the background proteome (the 393 detected pollen coat proteins; q‐value < 0.05, Fig. 2a). These terms include ‘modification of morphology or physiology of other organism’, ‘disruption of cells of other organism’, ‘killing of cells of other organism’, ‘cell killing’, ‘regulation of biological quality’, and various terms related to defense responses to fungus, biotic stimuli, other organisms, or external stimuli (see details in Fig. 2a).

Fig. 2.

Fig. 2

Open in a new tab

Differential expression of pollen coat proteins between selfer and outcrosser and their respective enriched Gene Ontology (GO) terms. This was done in three comparisons: (a) Arabidopsis lyrata (SI) vs A. lyrata (SC); (b) Capsella grandiflora (SI) vs Capsella rubella (SC); and (c) Arabidopsis arenosa (SI) vs Arabidopsis thaliana (SC). The pie charts show the number of significantly up‐ and downregulated proteins (false discovery rate (FDR) < 0.05, marked with an asterisk) in blue and red, respectively. The number of proteins with no significant differences in expression are shown in beige. GO term enrichment analysis was performed on significantly differential proteins in pollen coat (up and down together). Significantly enriched GO terms are shown (q‐value < 0.05). Values of the bar charts are −log10 transformed q‐values. The size of each closed circle is proportional to the number of genes enriched in each category. BP, biological process; CC, cellular component; SC, self‐compatible; SI, self‐incompatible.

We used a similar approach to analyze the pollen coat proteins of outcrossing C. grandiflora vs selfing C. rubella. We identified a total of 255 pollen coat proteins present in at least one of the two species. The biological function of these proteins was investigated with GO enrichment analysis, and highly represented GO terms within the biological process were like those observed in A. lyrata (q‐value < 0.05; see details in Fig. S2a). A total of 140 proteins were found to be significantly upregulated in C. grandiflora compared with C. rubella (Fig. 2b). To further evaluate the biological functions of the 140 significantly upregulated proteins, a GO enrichment analysis was performed using whole Capsella pollen coat proteins as background. Similar to the results obtained for outcrossing A. lyrata, the proteins with higher abundance in C. grandiflora were significantly enriched for GO terms such as ‘disruption of cells of other organism’, ‘modification of morphology or physiology of other organism’, ‘killing of cells of other organism’, and ‘cell killing’ (q‐value < 0.05; Fig. 2b).

Finally, we identified pollen coat proteins that were differentially abundant between the outcrosser A. arenosa and the selfer A. thaliana. A total of 721 pollen coat proteins were detected in at least one of the two species. ‘Disruption of cells of other organism’, ‘killing of cells of other organism’, and ‘cell killing’ were some of the significantly represented GO terms within the biological process (q‐value < 0.05; see other terms in Fig. S2a). Among these, 222 proteins were found to be significantly differentially abundant, with 122 being upregulated and 100 downregulated in A. arenosa compared with A. thaliana (Fig. 2c). GO enrichment analysis of the 222 differentially abundant proteins (including both upregulated and downregulated) using whole A. arenosa and A. thaliana pollen coat proteome as background revealed significant enrichment for 10 GO terms in the biological process category. These terms included ‘disruption of cells of other organism’, ‘killing of cells of other organism’, ‘cell killing’, ‘modification of morphology or physiology of other organism’, ‘multi‐organism process’, and various terms related to defense responses to fungi, biotic stimuli, other organisms, or external biotic stimuli (see other related terms in Fig. 2c). Furthermore, two GO terms were significantly enriched in the cellular component category, specifically ‘extracellular region’ and ‘lipid particle’ (Fig. 2c; q‐value < 0.05).

In the rest of the pollen, differentially abundant proteins in selfer–outcrosser comparisons tend to be part of the secretome

We then investigated the rest of the pollen proteome (excluding the coat during the extraction protocol, see the Materials and Methods section) and tested whether it also showed signatures associated with the transition to selfing in the three independent comparison pairs (Table S7). First, using a PCA from the pollen proteome abundance data, we found again that PC1 captured between‐samples variation (Fig. S1b). In addition, PC2 clustering was consistent with the phylogenetic relationship between species (A. lyrata SI and SC closest to each other, Arabidopsis and Capsella species forming two separate groups).

The number of detected proteins for each selfer–outcrosser contrast was 1367 proteins in A. lyrata–A. lyrata, 1543 proteins in C. grandiflora–C. rubella, and 2034 proteins in A. arenosa–A. thaliana (Fig. 3a). Similar enriched GO terms in the biological process category were found across the three groups, including ‘response to metal ions’, ‘organonitrogen compound metabolic process’, ‘response to cadmium ion’, ‘translation’, and ‘peptide biosynthetic process’ (q‐value < 0.05; see other terms in Fig. S2b). Of the 1367 total pollen proteins identified between outcrossing and selfing A. lyrata, 119 showed significant differential abundance, with 72 upregulated and 47 downregulated proteins (Fig. 3a; q‐value < 0.05). In the comparison between outcrossing C. grandiflora and selfers C. rubella, 36 proteins were significantly upregulated, while 35 showed significantly lower abundance in C. grandiflora compared with C. rubella. Lastly, among the 200 significant total pollen proteins identified in A. arenosa and A. thaliana, 124 were significantly upregulated and 76 were downregulated in A. arenosa compared with A. thaliana (Fig. 3a).

Fig. 3.

Fig. 3

Open in a new tab

Differential expression of pollen proteins (excluding pollen coat) between selfer and outcrosser and enriched Gene Ontology (GO) terms. (a) This was done in three comparisons: Arabidopsis lyrata (SI) vs A. lyrata (SC), Capsella grandiflora (SI) vs Capsella rubella (SC), and Arabidopsis arenosa (SI) vs Arabidopsis thaliana (SC). The pie charts show the number of significantly up‐ and downregulated proteins (false discovery rate (FDR) < 0.05, marked with an asterisk) in blue and red, respectively. The number of proteins with no significant differences in expression are shown in beige. (b) Significantly enriched GO terms (q‐value < 0.05) from the significantly differential pollen proteins (both up and down taken together) between the indicated species. Values of the bar charts are −log10 transformed q‐values. The size of each closed circle is proportional to the number of genes enriched in each category. CC, cellular component; MF, molecular function; SC, self‐compatible; SI, self‐incompatible.

To further explore the biological functions of differentially abundant pollen proteins between outcrossing and selfing species across three independent comparison groups, we performed a GO enrichment analysis. Significantly up‐ and downregulated proteins were included in the analysis, using a background list of whole pollen proteins from species involved in each respective pairwise comparison to correct for overrepresented functions in pollen. Among the 119 differentially abundant proteins in the A. lyrata (SI) vs A. lyrata (SC) comparison, four GO terms were significantly enriched (q‐value <0.05; Fig. 3a). These included ‘anchored component of membrane’ and ‘anchored component of plasma membrane’ for cellular component, and ‘hydrolyzing O‐glycosyl compounds, hydrolase activity’ and ‘acting on glycosyl bond’ for molecular function (Fig. 3b). For the 200 differentially abundant proteins in the A. arenosa (SI) vs A. thaliana (SC) comparison, significant enrichment of GO terms associated with protein secretion and extracellular proteome was observed in cellular component functions, such as ‘cell periphery’, ‘vacuole’, ‘symplast’, ‘cell junction’, ‘plasmodesma’, ‘extracellular region’, and ‘apoplast’ (Fig. 3b; q‐value < 0.05). No significantly enriched GO term was found for the comparison between C. rubella and C. grandiflora.

Specific pollen coat proteins are preferential targets of abundance changes across the three transitions to selfing

To investigate whether pollen coat and total pollen proteins abundance repeatedly diverges among independent selfer–outcrosser systems, we compared the differentially abundant proteins identified in the three comparisons: A. arenosa (SI) vs A. thaliana (SC), A. lyrata (SI) vs A. lyrata (SC), and C. grandiflora (SI) vs C. rubella (SC). As a reminder, a total of 222 significantly differentially abundant pollen coat proteins were identified in the comparison between A. arenosa (SI) and A. thaliana (SC), 56 in the comparison between A. lyrata (SI) and A. lyrata (SC), and 140 in the comparison between C. grandiflora (SI) and C. rubella (SC). Among these, 11 proteins (3.2%) were found to overlap across all three independent groups, which was more than expected by chance (Fig. 4a; superexacttest P = 2.37e‐04). To test whether specific pathways undergo protein abundance changes between the three transitions to selfing, we overlapped a total of 12 significantly enriched GO terms for A. arenosa (SI) vs A. thaliana (SC), 10 for A. lyrata (SI) vs A. lyrata (SC), and four for C. grandiflora (SI) vs C. rubella (SC). The overlap of GO terms among the three groups was greater than the overlap observed from proteins, with four common GO terms shared across the three comparisons being all enriched GO terms for Capsella, half for A. lyrata and one third for the A. arenosa–A. thaliana contrast (Fig. 4a; superexacttest P = 1.09e‐05). These four GO terms were all related to pathogen response.

Fig. 4.

Fig. 4

Open in a new tab

Commonly differential pollen coat proteins across three independent selfer–outcrosser contrasts. (a) Venn diagram showing the overlap of differential pollen coat proteins (left) and enriched Gene Ontology (GO) terms (right) between the comparisons of Arabidopsis lyrata (SI) vs A. lyrata (SC), Capsella grandiflora (SI) vs Capsella rubella (SC), and Arabidopsis arenosa (SI) vs Arabidopsis thaliana (SC). The colors used in GO terms Venn diagrams are consistent with gene ID diagrams. Differential proteins and enriched GO terms are the same as those shown in Figs 2 and 3. (b) Venn diagram showing the overlap between the three contrasts of upregulated pollen coat proteins in outcrossers vs selfers and their enriched GO terms. No GO term was significantly enriched in A. arenosa upregulated proteins vs A. thaliana. Asterisks indicate that the number of overlapping proteins/GO terms is significantly greater than expected by chance (superexacttest, P < 0.01). (c) Detailed description and statistical analysis (false discovery rate (FDR) < 0.05) of the two overlapping pollen coat proteins identified as upregulated in outcrossers across the three selfer–outcrosser systems. Gene descriptions were obtained from the A. lyrata v.2.1 annotation file (Rawat et al., 2015). SI, self‐incompatible; SC, self‐compatible.

Nevertheless, strictly speaking, convergence between the three selfing transitions requires abundance changes to go in the same direction. To test whether this was the case, we compared proteins which underwent abundance changes in the same direction (up or downregulated between outcrossers and selfers), and their enriched GO terms. Two proteins were commonly upregulated in outcrossers vs selfers across the three comparisons, which was not significantly different from chance (Fig. 4b,c; superexacttest P = 0.49). By contrast, we found a significant overlap of upregulated proteins between the A. lyrata SI‐SC and C. grandiflora/C. rubella comparisons (superexacttest P = 0.00684). Enriched GO terms among the upregulated proteins also showed significant overlap between these two contrasts (Fig. 4b; hypergeometric test, P = 0.008519), the four same GO terms as in Fig. 4(a) related to pathogen response. No significantly enriched GO term was found among upregulated proteins in A. arenosa compared with A. thaliana. We looked in more detail at the upregulated proteins in A. lyrata SI vs SC and C. grandiflora vs C. rubella flagged with the GO term ‘cell killing’ (Table 1). These 26 proteins belonged to three gene families only: Defensin‐like proteins, low‐molecular‐weight cysteine‐rich (LCR) and SCR‐like proteins (proteins homologous to the male component of the self‐incompatibility system, SCR). None but one protein is documented for mutant/genetic characterization (source: TAIR website), PLANT DEFENSIN‐LIKE 1.3. Upon transient expression in tobacco leaves, PLANT DEFENSIN‐LIKE 1.3 triggered a marginal HR response, unlike other genes from the same family, of which expression provoked more significant symptoms (Omidvar et al., 2021).

Table 1.

List of upregulated proteins in outcrossers compared with selfers, flagged with the Gene Ontology (GO) term ‘cell killing’.

Gene ID Comparison Name/description Mutant phenotype/genetic characterization
AT1G69825 Up in C. grandiflora vs C. rubella PLANT DEFENSIN‐LIKE 1.3 (PDFL1.3) GUS staining showing strongest PDFL1.3 expression in stigma and anthers; upon transient expression (tobacco infiltration), PDFL1.3 scarcely triggers HR/cell death, unlike many other PDFLs, which trigger significant HR responses (Omidvar et al., 2021)
AT2G22941 Up in C. grandiflora vs C. rubella Encodes a defensin‐like (DEFL) family protein. Not available
AT4G10603 Up in C. grandiflora vs C. rubella Encodes a defensin‐like (DEFL) family protein. Not available
AT4G09647 Up in A. lyrata SI vs SC Encodes a defensin‐like (DEFL) family protein. Not available
AT5G39365 Up in A. lyrata SI vs SC Encodes a defensin‐like (DEFL) family protein. Not available
AT4G10595 Up in C. grandiflora vs C. rubella LCR2 (low‐molecular‐weight cysteine‐rich 2) Not available
AT3G25265 Up in A. lyrata SI vs SC LCR4 (low‐molecular‐weight cysteine‐rich 4) Not available
AT2G28355 Up in C. grandiflora vs C. rubella LCR5 (low‐molecular‐weight cysteine‐rich 5) Not available
AT2G27145 Up in C. grandiflora vs C. rubella + A. lyrata SI vs SC LCR9 (Low‐molecular‐weight cysteine‐rich 9) Not available
AT2G15535 Up in C. grandiflora vs C. rubella LCR10 (low‐molecular‐weight cysteine‐rich 10) Not available
AT2G25344 Up in C. grandiflora vs C. rubella LCR14 (low‐molecular‐weight cysteine‐rich 14) Not available
AT4G11760 Up in C. grandiflora vs C. rubella LCR17 (low‐molecular‐weight cysteine‐rich 17) Not available
AT4G29305 Up in C. grandiflora vs C. rubella LCR25 (low‐molecular‐weight cysteine‐rich 25) Not available
AT1G28335 Up in C. grandiflora vs C. rubella LCR31 (low‐molecular‐weight cysteine‐rich 31) Not available
AT4G09984 Up in A. lyrata SI vs SC LCR34 (low‐molecular‐weight cysteine‐rich 34) Not available
AT1G61070 Up in A. lyrata SI vs SC LCR66 (low‐molecular‐weight cysteine‐rich 66) Not available
AT2G02100 Up in C. grandiflora vs C. rubella LCR69 (low‐molecular‐weight cysteine‐rich 69) Not available
AT2G02140 Up in C. grandiflora vs C. rubella + A. lyrata SI vs SC LCR72 (low‐molecular‐weight cysteine‐rich 72) Not available
AT4G10457 Up in A. lyrata SI vs SC SCRL1 (SCR‐Like 1) Not available
AT1G65113 Up in C. grandiflora vs C. rubella + A. lyrata SI vs SC SCRL2 (SCR‐Like 2) Not available
AT1G08695 Up in C. grandiflora vs C. rubella SCRL3 (SCR‐Like 3) Not available
AT3G23715 Up in C. grandiflora vs C. rubella SCRL13 (SCR‐Like 13) Not available
AT2G06983 Up in A. lyrata SI vs SC SCRL16 (SCR‐Like 16) Not available
AT4G33465 Up in C. grandiflora vs C. rubella SCRL22 (SCR‐Like 22) Not available
AT4G14785 Up in C. grandiflora vs C. rubella SCRL23 (SCR‐Like 23) Not available
AT4G32717 Up in C. grandiflora vs C. rubella SCRL24 (SCR‐Like 24) Not available
Open in a new tab

This GO term is commonly enriched among upregulated proteins in both the C. grandiflora–C. rubella and A. lyrata SI‐SC comparisons (Fig. 4b). Information regarding the gene description and genetic characterization obtained from arabidopsis.org. SC, self‐compatible; SI, self‐incompatible.

Three downregulated proteins were shared between the A. lyrata SI‐SC and A. arenosa‐A. thaliana comparisons (the Capsella comparison had none; Fig. 2b), which was not significantly different from chance (Fig. 4b; hypergeometric test, P = 0.15). Pathogen defense‐related GO terms were enriched among downregulated proteins between A. arenosa and A. thaliana (Fig. S3a). In fact, we observed a reverse trend: a significant overlap between downregulated proteins in A. arenosa vs A. thaliana, and upregulated proteins in outcrossers of the two other comparisons (Fig. S3b).

We also tested whether the total pollen proteome shows signals of convergent evolution between the three transitions to selfing. A total of 200 significantly differentially abundant total pollen proteins were identified in the comparison between A. arenosa (SI) and A. thaliana (SC), 119 in the comparison between A. lyrata (SI) and A. lyrata (SC), and 71 in the comparison between C. grandiflora (SI) and C. rubella (SC). The number of shared proteins among the three groups was relatively low, with only two proteins overlapping across all comparisons (Fig. 5a; superexacttest P = 0.127). Additionally, the overlap of significantly enriched GO terms for total pollen proteins was null. Focusing on protein abundance changes going in the same direction in the three comparisons, we found none, neither for upregulated nor downregulated proteins in outcrossers compared with selfers (Fig. 5b). In addition, for the three outcrosser vs selfer comparisons, there were no significantly enriched GO terms among up‐ or downregulated proteins.

Fig. 5.

Fig. 5

Open in a new tab

Commonly differential total pollen proteins across three independent selfer–outcrosser contrasts. (a) Venn diagram showing the overlap of differential total pollen proteins and enriched Gen Ontology (GO) terms between the three comparisons (Arabidopsis lyrata (SI) vs A. lyrata (SC), Capsella grandiflora (SI) vs Capsella rubella (SC), and Arabidopsis arenosa (SI) vs Arabidopsis thaliana (SC)). Differential proteins and enriched GO terms are the same as those shown in Figs 2 and 3. The colors used in GO terms Venn diagrams are consistent with gene ID diagrams. (b) Venn diagram showing the overlap between the three contrasts of up‐ and downregulated total pollen proteins in outcrossers vs selfers. Asterisks indicate that the number of overlapping proteins/GO terms is significantly greater than expected by chance (superexacttest, P < 0.01). SC, self‐compatible; SI, self‐incompatible.

Discussion

Reduction in pollen coat area in selfing A. thaliana and C. rubella species aligns with characteristic of selfing syndrome

Vestigialization of ancestral outcrossing floral traits following the transition to selfing has largely been documented in the literature (Sicard & Lenhard, 2011; Tsuchimatsu & Fujii, 2022). The loss of traits in selfing species often involves the reduction in characteristics important for pollinator attraction associated with prepollination processes which are no longer needed due to reproductive assurance. In addition to the reduced reliance on pollinators, other changes accompany the transition to selfing, such as the re‐allocation of resources to male and female gametes (reduced pollen‐ovule ratio; Barrett, 2003), or the enhancement of the efficiency of self‐pollination (Shimizu & Tsuchimatsu, 2015; Toräng et al., 2017). Although certain traits are well known to evolve with the selfing syndrome, it remains to be elucidated whether additional traits are associated with it. The pollen coat plays a vital role in pollinator‐mediated sexual reproduction and thus we hypothesize that the pollen coat becomes less crucial in selfing species, for example as pollen is not transported over long distances anymore, protection against pathogens and other environmental stressors is less needed. Our findings showed that independent transitions to selfing in A. thaliana and C. rubella, both of which display a strong selfing syndrome, resulted in a significant reduction in pollen coat area, suggesting convergent evolution of the pollen coat (Fig. 1). Although comparative studies on pollen coat variation between closely related species are limited, it has been shown that the transition from insect pollination to wind pollination leads to a corresponding decrease in pollen coat amount and density (Hesse, 1979). The transition from outcrossing to selfing may resemble similar effects, as pollinator dependence gradually decreases in selfing species. In this context, a reduction in pollen coat area in independently evolved selfing species exhibiting strong selfing syndromes, in our study A. thaliana and Capsella rubella, could represent pollen coat as another trait associated with the selfing syndrome. Whether this reduction is a result of neutral decay over evolutionary time, or it is beneficial as it may save resources, remains unclear. A study has found that an allele decreasing the number of pollen grains in A. thaliana is under positive selection (Tsuchimatsu et al., 2020), suggesting that the vestigialization of floral traits (including pollen coat) associated with the selfing syndrome might have benefits.

An intriguing result from our study is the observation that selfing A. lyrata did not show the same trend. This finding may reflect the relatively recent breakdown of self‐incompatibility in A. lyrata populations (Foxe et al., 2010), with no evidence of selfing syndrome yet (Carleial et al., 2017). Consistently, the number of differentially abundant pollen coat proteins between A. lyrata selfer and outcrosser was substantially lower than for the two other selfer–outcrosser comparisons (Fig. 2a).

Pollen coat protein composition evolves quickly, in contrast with the proteome of the rest of the pollen

The evolution of reproductive proteins is likely driven by strong positive selection, influenced by sexual selection pressures, such as cryptic female choice (Eberhard, 1996) and sperm competition (Clark, 2002). Beyond sexual selection, before pollination, male gametophytes face challenges, such as water loss and defense against pathogens, which are critical for survival of pollen grains present on anther (Rivest & Forrest, 2020). These pressures may contribute to the rapid evolution of reproductive proteins, as selection favors advantageous modifications that improve pollen grain survival. This rapid evolution may happen at the protein sequence level, protein abundance, or gene gain/loss level, which we could not test as we focused on proteins for which we found an orthologue in all species (using a selfer, A. thaliana, as a reference). In this work, we primarily focused on divergence at the protein abundance level. In all the comparisons, we performed between species (independent of the mating system), we generally found a higher proportion of differentially abundant proteins between lineages in the pollen coat compared with the rest of the pollen (Figs 2, 3). In addition, the PCAs showed a clear phylogenetic signal on the total pollen proteome, while this signal was less clear for the pollen coat proteome (Fig. S1). This suggests that the pollen coat proteome evolves faster than the proteome of the rest of the pollen. The core pollen proteome may be genetically constrained, likely due to its involvement in some of the most basic and general reproductive processes (Hafidh et al., 2016; Scholz et al., 2020). By contrast, the pollen coat may be more susceptible to evolutionary changes due to the constantly changing ecological demands, as it is directly involved in the interactions with the ambient environment (Rivest & Forrest, 2020). Consistently, while there was virtually no convergence for the total pollen proteome in all the comparisons, we performed between lineages, similar proteins and functions recurrently showed differential abundance in the pollen coat, and these were related to pathogen response. This could be related to the challenges faced by outcrossers but not selfers, namely the transport of pollen over long distances and its need to survive until its deposition on the stigma (Rivest & Forrest, 2020). Nevertheless, it is important to note that the proteins we identified as recurrently differential and annotated as pathogen response all belonged to families of small signaling peptides, such as cysteine‐rich, SCR‐like, or defensin‐like proteins (Table 1), being potentially connected with pathogen defense but also pollen–pistil (stigma) cell–cell communication and gamete interaction in flowering plants (Lionetti et al., 2012; Takeuchi & Higashiyama, 2012; Wormit & Usadel, 2018; Hafidh & Honys, 2021). In fact, the pollen tube growth across the transmission tract resembles a pathogen response at the transcriptome level (Elleman & Dickinson, 1999; Dresselhaus & Márton, 2009; İltaş et al., 2024). The SCR‐like proteins, homologous to the SCR protein responsible for the recognition of pollen by the stigma in the SI system, particularly suggest that these proteins are rather involved in sexual interactions than pathogen response. Unfortunately, the large majority of these proteins are uncharacterized, with no genetic study available except for one, PDFL1.3, which showed no role in inducing a pathogen response (Omidvar et al., 2021). It could thus be that the selective pressure leading to the fast pollen coat proteome divergence between lineages is sexual selection rather than pathogen response. Future works should aim at disentangling the two, especially by investigating the functional role of the proteins we identified, which seem to have been overlooked by research efforts so far.

Independent transitions to selfing are accompanied by limited signals of convergent changes in pollen coat proteins

As discussed above, the selfing syndrome may be accompanied by changes in pollen traits, such as pollen coat thickness. These changes may display features of convergent evolution, as selfing emerged independently in separate lineages and similar selective pressures are at play. As discussed in the previous paragraph, we did see abundance changes recurrently associated with the three transitions to selfing for specific pollen coat protein families. These are small proteins flagged with GO terms related to pathogen response, but might be involved in pollen–pistil interactions instead. Recurrent abundance changes for these proteins in response to selfing transitions may thus reflect the evolution of the molecular crosstalk between the pollen and pistil, shifting from an interaction with multiple foreign pollen donors to one main donor (self).

However, convergence sensu stricto requires protein abundance changes happening in the same direction. According to this criterion, convergence across the three selfing transitions we studied was rather limited, with only a few pollen coat proteins showing an increased abundance in outcrossers compared with selfers. The A. lyrata and Capsella selfing transitions significantly converged in terms of protein abundance and associated functions (Fig. 4b), with again ‘pathogen response’ functions being overrepresented among upregulated proteins in outcrossers, alongside other proteins, such as pectin methylesterase inhibitors, important enzymes for pollen tube growth (Scholz et al., 2020). The same proteins also showed significant abundance changes in the A. arenosa–A. thaliana comparison but instead of going down in the selfer, their abundance was higher in A. thaliana compared with A. arenosa (Fig. S3). This result suggests that the pollen coat proteome composition is not prone to convergence strictly speaking in response to selfing; it would nevertheless be important to study more selfer–outcrosser transitions to rule out that the A. thaliana–A. arenosa case is not an outlier. In particular, these two species are relatively far apart (6 Ma; Novikova et al., 2016) compared with the two other pairs (Capsella split 100–200 ka, Guo et al., 2009; A. lyrata SI‐SC divergence 30 ka, Schmickl et al., 2010). It is possible that additional evolutionary factors impact the pollen coat proteome divergence between A. arenosa and A. thaliana which are not directly related to the mating system. The fact that a significant portion of proteins showing abundance changes was the same across the three selfing transitions remains striking nevertheless, suggesting a hotspot molecular pathway for male gametophyte evolution. It could of course be that these proteins are preferential targets of abundance divergence between lineages unrelated to mating systems; again, increasing the design for further studies with more populations and more selfer–outcrosser pairs should help disentangle this issue.

Competing interests

None declared.

Author contributions

CLP and LS conceptualized the study. Öİ and LS performed the experiments. Öİ, SH, MK, MČ, PT and KH analyzed the data. Öİ and CLP wrote the manuscript with contributions from all authors.

Disclaimer

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.

Supporting information

Fig. S1 General protein expression patterns of pollen coat and total pollen among the species used in this study.

Fig. S2 GO enrichment analysis of the proteome of separated pollen coat, and pollen excluding the coat identified in each comparison between outcrosser and selfer species.

Fig. S3 Reverse changes in pollen coat protein abundance in the A. thaliana–A. arenosa comparison vs the two other selfer–outcrosser contrasts.

Table S1 Population index of Brassicaceae lineages used in this study.

NPH-250-3460-s002.pdf (467KB, pdf)

Table S2 Details of the pollen phenotypic data used in this study.

NPH-250-3460-s004.xlsx (29.5KB, xlsx)

Table S3 Relative intensities for detected proteins in pollen coat and total pollen for each species.

NPH-250-3460-s001.xlsx (5.1MB, xlsx)

Table S4 Assessment of cytosol contamination and pollen coat enrichment using pollen coat and cytosol markers.

Table S5 Summary of a two‐way ANOVA model testing how the extent of pollen coat development is affected by mating system transitions across three different species comparisons within the Brassicaceae family.

NPH-250-3460-s003.xlsx (22.6KB, xlsx)

Table S6 Identified and quantified pollen coat proteins for each comparison.

NPH-250-3460-s006.xlsx (1MB, xlsx)

Table S7 Identified and quantified total pollen proteins for each comparison.

Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

NPH-250-3460-s005.xlsx (3.3MB, xlsx)

Acknowledgements

We are very grateful to Adrien Sicard and Vojtech Cermak for providing seeds of Capsella and A. thaliana species, respectively. This work, Öİ, LS, and CLP were supported by the Charles University Research Centre program (grant no.: PRIMUS/19/SCI/02 to CLP). MČ was supported by a long‐term research development project no. RVO 67985939 of the Czech Academy of Sciences. The authors acknowledge the Viničná Microscopy Core Facility (VMCF of the Faculty of Science, Charles University), an institution supported by the MEYS CR (LM2023050 Czech‐BioImaging); special thanks to Miroslav Hyliš for his support and assistance in this work. LC‐MS analyses were performed in the Laboratory of Mass Spectrometry at Biocev research center, Faculty of Science, Charles University. Computational resources were provided by the e‐INFRA CZ project (ID: 90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. AI tools were not used for any part of this work. Open access publishing facilitated by Univerzita Karlova, as part of the Wiley ‐ CzechELib agreement.

Data availability

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez‐Riverol et al., 2024) partner repository with the dataset identifier PXD061508. All data are available in Tables [Link], [Link], [Link], [Link], [Link], [Link].

References

  1. Al‐Shehbaz IA, O'Kane SL. 2002. Taxonomy and phylogeny of Arabidopsis (Brassicaceae). The Arabidopsis Book/American Society of Plant Biologists 1: e0001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson B, Minnaar C. 2020. Illuminating the incredible journey of pollen. American Journal of Botany 107: 1323–1326. [DOI] [PubMed] [Google Scholar]
  3. Barrett SCH. 2002. The evolution of plant sexual diversity. Nature Reviews Genetics 3: 274–284. [DOI] [PubMed] [Google Scholar]
  4. Barrett SCH. 2003. Mating strategies in flowering plants: the outcrossing‐selfing paradigm and beyond. Philosophical Transactions of the Royal Society, B: Biological Sciences 358: 991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barrett SCH, Arunkumar R, Wright SI. 2014. The demography and population genomics of evolutionary transitions to self‐fertilization in plants. Philosophical Transactions of the Royal Society, B: Biological Sciences 369: 20130344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Birkeland S, Gustafsson ALS, Brysting AK, Brochmann C, Nowak MD. 2020. Multiple genetic trajectories to extreme abiotic stress adaptation in Arctic Brassicaceae. Molecular Biology and Evolution 37: 2052–2068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bohutínská M, Alston M, Monnahan P, Mandáková T, Bray S, Paajanen P, Kolář F, Yant L. 2021. Novelty and convergence in adaptation to whole genome duplication. Molecular Biology and Evolution 38: 3910–3924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, Madden TL. 2009. BLAST+: architecture and applications. BMC Bioinformatics 10: 421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Carleial S, van Kleunen M, Stift M. 2017. Small reductions in corolla size and pollen: ovule ratio, but no changes in flower shape in selfing populations of the North American Arabidopsis lyrata . Oecologia 183: 401–413. [DOI] [PubMed] [Google Scholar]
  10. Clark AG. 2002. Sperm competition and the maintenance of polymorphism. Heredity 88: 148–153. [DOI] [PubMed] [Google Scholar]
  11. Cox J, Mann M. 2008. MaxQuant enables high peptide identification rates, individualized p.p.b.‐range mass accuracies and proteome‐wide protein quantification. Nature Biotechnology 26: 1367–1372. [DOI] [PubMed] [Google Scholar]
  12. Cutter AD. 2019. Reproductive transitions in plants and animals: selfing syndrome, sexual selection and speciation. New Phytologist 224: 1080–1094. [DOI] [PubMed] [Google Scholar]
  13. Dresselhaus T, Márton ML. 2009. Micropylar pollen tube guidance and burst: adapted from defense mechanisms? Current Opinion in Plant Biology 12: 773–780. [DOI] [PubMed] [Google Scholar]
  14. Eberhard WG. 1996. Female control: sexual selection by cryptic female choice. New Jersey, USA: Princeton University Press. doi: 10.2307/j.ctvs32rx1. [DOI] [Google Scholar]
  15. Elleman CJ, Dickinson HG. 1999. Commonalities between pollen/stigma and host/pathogen interactions: calcium accumulation during stigmatic penetration by Brassica oleracea pollen tubes. Sexual Plant Reproduction 12: 194–202. [Google Scholar]
  16. Foxe JP, Stift M, Tedder A, Haudry A, Wright SI, Mable BK. 2010. Reconstructing origins of loss of self‐incompatibility and selfing in North American Arabidopsis lyrata: A population genetic context. Evolution 64: 3495–3510. [DOI] [PubMed] [Google Scholar]
  17. Gong F, Wu X, Wang W. 2015. Diversity and function of maize pollen coat proteins: from biochemistry to proteomics. Frontiers in Plant Science 6: 199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Guo YL, Bechsgaard JS, Slotte T, Neuffer B, Lascoux M, Weigel D, Schierup MH. 2009. Recent speciation of Capsella rubella from Capsella grandiflora, associated with loss of self‐incompatibility and an extreme bottleneck. Proceedings of the National Academy of Sciences, USA 106: 5246–5251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hafidh S, Fíla J, Honys D. 2016. Male gametophyte development and function in angiosperms: a general concept. Plant Reproduction 29: 31–51. [DOI] [PubMed] [Google Scholar]
  20. Hafidh S, Honys D. 2021. reproduction multitasking: the male gametophyte. Annual Review of Plant Biology 72: 581–614. [DOI] [PubMed] [Google Scholar]
  21. Hesse M. 1979. Ultrastruktur und Verteilung des Pollenkitts in der insekten‐ und windblütigen GattungAcer (Aceraceae). Plant Systematics and Evolution 131: 277–289. [Google Scholar]
  22. Hughes CS, Moggridge S, Müller T, Sorensen PH, Morin GB, Krijgsveld J. 2019. Single‐pot, solid‐phase‐enhanced sample preparation for proteomics experiments. Nature Protocols 14: 68–85. [DOI] [PubMed] [Google Scholar]
  23. Hulsen T. 2022. Deepvenn – a web application for the creation of area‐proportional Venn diagrams using the deep learning framework Tensorflow.js . doi: 10.48550/ARXIV.2210.04597. [DOI]
  24. Igic B, Bohs L, Kohn JR. 2006. Ancient polymorphism reveals unidirectional breeding system shifts. Proceedings of the National Academy of Sciences, USA 103: 1359–1363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. İltaş Ö, Čertner M, Lafon Placette C. 2024. Natural variation in sexual traits and gene expression between selfing and outcrossing Arabidopsis lyrata suggests sexual selection at work. Plant and Cell Physiology 66: 581–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Johnson‐Brousseau SA, McCormick S. 2004. A compendium of methods useful for characterizing Arabidopsis pollen mutants and gametophytically‐ expressed genes. The Plant Journal 39: 761–775. [DOI] [PubMed] [Google Scholar]
  27. Kalisz S, Vogler DW, Hanley KM. 2004. Context‐dependent autonomous self‐fertilization yields reproductive assurance and mixed mating. Nature 430: 884–887. [DOI] [PubMed] [Google Scholar]
  28. Kennedy BF, Elle E. 2008. The inbreeding depression cost of selfing: importance of flower size and population size in Collinsia parviflora (Veronicaceae). American Journal of Botany 95: 1596–1605. [DOI] [PubMed] [Google Scholar]
  29. Koch M, Bishop J, Mitchell‐Olds T. 1999. Molecular systematics and evolution of Arabidopsis and Arabis. Plant Biology 1: 529–537. [Google Scholar]
  30. Kolesnikova UK, Scott AD, Van de Velde JD, Burns R, Tikhomirov NP, Pfordt U, Clarke AC, Yant L, Seregin AP, Vekemans X et al. 2023. Transition to self‐compatibility associated with dominant S‐allele in a diploid siberian progenitor of allotetraploid Arabidopsis kamchatica Revealed by Arabidopsis lyrata genomes. Molecular Biology and Evolution 40: msad122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lamesch P, Berardini TZ, Li D, Swarbreck D, Wilks C, Sasidharan R, Muller R, Dreher K, Alexander DL, Garcia‐Hernandez M et al. 2012. The Arabidopsis Information Resource (TAIR): improved gene annotation and new tools. Nucleic Acids Research 40(D1): D1202–D1210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Li Y, Liu Z, Shi P, Zhang J. 2010. The hearing gene Prestin unites echolocating bats and whales. Current Biology 20: R55–R56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lionetti V, Cervone F, Bellincampi D. 2012. Methyl esterification of pectin plays a role during plant‐pathogen interactions and affects plant resistance to diseases. Journal of Plant Physiology 169: 1623–1630. [DOI] [PubMed] [Google Scholar]
  34. Lyu B, Liang C. 2025. Insights into the molecular basis of pollen coat development and its role in male sterility. International Journal of Molecular Sciences 26: 7036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nasrallah ME, Yogeeswaran K, Snyder S, Nasrallah JB. 2000. Arabidopsis species hybrids in the study of species differences and evolution of amphiploidy in plants. Plant Physiology 124: 1605–1614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Novikova PY, Hohmann N, Nizhynska V, Tsuchimatsu T, Ali J, Muir G, Guggisberg A, Paape T, Schmid K, Fedorenko OM et al. 2016. Sequencing of the genus Arabidopsis identifies a complex history of nonbifurcating speciation and abundant trans‐specific polymorphism. Nature Genetics 48: 1077–1082. [DOI] [PubMed] [Google Scholar]
  37. Omidvar R, Vosseler N, Abbas A, Gutmann B, Grünwald‐Gruber C, Altmann F, Siddique S, Bohlmann H. 2021. Analysis of a gene family for PDF‐like peptides from Arabidopsis. Scientific Reports 11: 18948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Pacini E, Hesse M. 2005. Pollenkitt – its composition, forms and functions. Flora ‐ Morphology, Distribution, Functional Ecology of Plants 200: 399–415. [Google Scholar]
  39. Perez‐Riverol Y, Bandla C, Kundu DJ, Kamatchinathan S, Bai J, Hewapathirana S, John NS, Prakash A, Walzer M, Wang S et al. 2024. The PRIDE database at 20 years: 2025 update. Nucleic Acids Research 53: D543–D553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rappsilber J, Mann M, Ishihama Y. 2007. Protocol for micro‐purification, enrichment, pre‐fractionation and storage of peptides for proteomics using StageTips. Nature Protocols 2: 1896–1906. [DOI] [PubMed] [Google Scholar]
  41. Rawat V, Abdelsamad A, Pietzenuk B, Seymour DK, Koenig D, Weigel D, Pecinka A, Schneeberger K. 2015. Improving the annotation of Arabidopsis lyrata using RNA‐seq data. PLoS ONE 10: e0137391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rejón JD, Delalande F, Schaeffer‐Reiss C, Alché JD, Rodríguez‐García MI, Van Dorsselaer A, Castro AJ. 2016. The pollen coat proteome: at the cutting edge of plant reproduction. Proteomes 4: 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Rivest S, Forrest JRK. 2020. Defence compounds in pollen: why do they occur and how do they affect the ecology and evolution of bees? New Phytologist 225: 1053–1064. [DOI] [PubMed] [Google Scholar]
  44. Schierup MH. 1998. The effect of enzyme heterozygosity on growth in a strictly outcrossing species, the self‐incompatible Arabis petraea (Brassicaceae). Hereditas 128: 21–31. [Google Scholar]
  45. Schmickl R, Jørgensen MH, Brysting AK, Koch MA. 2010. The evolutionary history of the Arabidopsis lyrata complex: a hybrid in the amphi‐Beringian area closes a large distribution gap and builds up a genetic barrier. BMC Evolutionary Biology 10: 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Scholz P, Anstatt J, Krawczyk HE, Ischebeck T. 2020. Signalling pinpointed to the tip: the complex regulatory network that allows pollen tube growth. Plants 9: 1098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Shimizu K, Tsuchimatsu T. 2015. Evolution of selfing: recurrent patterns in molecular adaptation. Annual Review of Ecology, Evolution, and Systematics 46: 593–622. [Google Scholar]
  48. Sicard A, Lenhard M. 2011. The selfing syndrome: a model for studying the genetic and evolutionary basis of morphological adaptation in plants. Annals of Botany 107: 1433–1443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Slotte T, Hazzouri KM, Ågren JA, Koenig D, Maumus F, Guo YL, Steige K, Platts AE, Escobar JS, Newman LK et al. 2013. The Capsella rubella genome and the genomic consequences of rapid mating system evolution. Nature Genetics 45: 831–835. [DOI] [PubMed] [Google Scholar]
  50. Stayton CT. 2015. The definition, recognition, and interpretation of convergent evolution, and two new measures for quantifying and assessing the significance of convergence. Evolution 69: 2140–2153. [DOI] [PubMed] [Google Scholar]
  51. Stebbins GL. 1974. Flowering plants: evolution above the species level. London, UK: Harvard University Press. [Google Scholar]
  52. Stern DL. 2013. The genetic causes of convergent evolution. Nature Reviews Genetics 14: 751–764. [DOI] [PubMed] [Google Scholar]
  53. Strütt S, Sellinger T, Glémin S, Tellier A, Laurent S. 2023. Joint inference of evolutionary transitions to self‐fertilization and demographic history using whole‐genome sequences. eLife 12: e82384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Takeuchi H, Higashiyama T. 2012. A species‐specific cluster of defensin‐like genes encodes diffusible pollen tube attractants in Arabidopsis. PLoS Biology 10: e1001449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Tang D, Chen M, Huang X, Zhang G, Zeng L, Zhang G, Wu S, Wang Y. 2023. SRplot: a free online platform for data visualization and graphing. PLoS ONE 18: e0294236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Tian F, Yang DC, Meng YQ, Jin J, Gao G. 2019. PlantRegMap: charting functional regulatory maps in plants. Nucleic Acids Research 48: gkz1020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Toräng P, Vikström L, Wunder J, Wötzel S, Coupland G, Ågren J. 2017. Evolution of the selfing syndrome: anther orientation and herkogamy together determine reproductive assurance in a self‐compatible plant. Evolution 71: 2206–2218. [DOI] [PubMed] [Google Scholar]
  58. Tsuchimatsu T, Fujii S. 2022. The selfing syndrome and beyond: diverse evolutionary consequences of mating system transitions in plants. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 377: 20200510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Tsuchimatsu T, Kakui H, Yamazaki M, Marona C, Tsutsui H, Hedhly A, Meng D, Sato Y, Städler T, Grossniklaus U et al. 2020. Adaptive reduction of male gamete number in the selfing plant Arabidopsis thaliana . Nature Communications 11: 2885. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Tusher VG, Tibshirani R, Chu G. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proceedings of the National Academy of Sciences, USA 98: 5116–5121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Tyanova S, Temu T, Sinitcyn P, Carlson A, Hein MY, Geiger T, Mann M, Cox J. 2016. The Perseus computational platform for comprehensive analysis of (prote)omics data. Nature Methods 13: 731–740. [DOI] [PubMed] [Google Scholar]
  62. Wang L, Lau YL, Fan L, Bosch M, Doughty J. 2023. Pollen coat proteomes of Arabidopsis thaliana, Arabidopsis lyrata, and Brassica oleracea reveal remarkable diversity of small cysteine‐rich proteins at the pollen‐stigma interface. Biomolecules 13: 157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang M, Zhao Y, Zhang B. 2015. Efficient test and visualization of multi‐set intersections. Scientific Reports 5: 16923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Willi Y, Määttänen K. 2010. Evolutionary dynamics of mating system shifts in Arabidopsis lyrata . Journal of Evolutionary Biology 23: 2123–2131. [DOI] [PubMed] [Google Scholar]
  65. Woodhouse MR, Hufford MB. 2019. Parallelism and convergence in post‐domestication adaptation in cereal grasses. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 374: 20180245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wormit A, Usadel B. 2018. The multifaceted role of pectin methylesterase inhibitors (PMEIs). International Journal of Molecular Sciences 19: 2878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Woźniak NJ, Sartori K, Kappel C, Tran TC, Zhao L, Erban A, Gallinger J, Fehrle I, Jantzen F, Orsucci M et al. 2024. Convergence and molecular evolution of floral fragrance after independent transitions to self‐fertilization. Current Biology 34: 2702–2711. [DOI] [PubMed] [Google Scholar]
  68. Zhang Y‐H, Ravi V, Qin G, Dai H, Zhang HX, Han FM, Wang X, Liu YH, Yin JP, Huang LM et al. 2020. Comparative genomics reveal shared genomic changes in syngnathid fishes and signatures of genetic convergence with placental mammals. National Science Review 7: 964–977. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig. S1 General protein expression patterns of pollen coat and total pollen among the species used in this study.

Fig. S2 GO enrichment analysis of the proteome of separated pollen coat, and pollen excluding the coat identified in each comparison between outcrosser and selfer species.

Fig. S3 Reverse changes in pollen coat protein abundance in the A. thaliana–A. arenosa comparison vs the two other selfer–outcrosser contrasts.

Table S1 Population index of Brassicaceae lineages used in this study.

NPH-250-3460-s002.pdf (467KB, pdf)

Table S2 Details of the pollen phenotypic data used in this study.

NPH-250-3460-s004.xlsx (29.5KB, xlsx)

Table S3 Relative intensities for detected proteins in pollen coat and total pollen for each species.

NPH-250-3460-s001.xlsx (5.1MB, xlsx)

Table S4 Assessment of cytosol contamination and pollen coat enrichment using pollen coat and cytosol markers.

Table S5 Summary of a two‐way ANOVA model testing how the extent of pollen coat development is affected by mating system transitions across three different species comparisons within the Brassicaceae family.

NPH-250-3460-s003.xlsx (22.6KB, xlsx)

Table S6 Identified and quantified pollen coat proteins for each comparison.

NPH-250-3460-s006.xlsx (1MB, xlsx)

Table S7 Identified and quantified total pollen proteins for each comparison.

Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

NPH-250-3460-s005.xlsx (3.3MB, xlsx)

Data Availability Statement

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez‐Riverol et al., 2024) partner repository with the dataset identifier PXD061508. All data are available in Tables [Link], [Link], [Link], [Link], [Link], [Link].

Articles from The New Phytologist are provided here courtesy of Wiley

RESOURCES