Evolution of the thermostability of actin-depolymerizing factors enhances the adaptation of pollen germination to high temperature

Abstract

Double fertilization in many flowering plants (angiosperms) often occurs during the hot summer season, but the mechanisms that enable angiosperms to adapt specifically to high temperatures are largely unknown. The actin cytoskeleton is essential for pollen germination and the polarized growth of pollen tubes, yet how this process responds to high temperatures remains unclear. Here, we reveal that the high thermal stability of 11 Arabidopsis (Arabidopsis thaliana) actin-depolymerizing factors (ADFs) is significantly different: ADFs that specifically accumulate in tip-growing cells (pollen and root hairs) exhibit high thermal stability. Through ancestral protein reconstruction, we found that subclass II ADFs (expressed specifically in pollen) have undergone a dynamic wave-like evolution of the retention, loss, and regeneration of thermostable sites. Additionally, the sites of AtADF7 with high thermal stability are conserved in ADFs specific to angiosperm pollen. Moreover, the high thermal stability of ADFs is required to regulate actin dynamics and turnover at high temperatures to promote pollen germination. Collectively, these findings suggest strategies for the adaptation of sexual reproduction to high temperature in angiosperms at the cell biology level.

IN A NUTSHELL

Background: Double fertilization of most angiosperms occurs in the heat of summer; understanding how the male gametophytes (pollen grains) germinate properly and pollen tubes grow normally at high temperature is essential to elucidating the evolution of angiosperm reproduction. Pollen tubes transport immobile sperm cells to ovules to complete double fertilization, thereby starting seed development. Rapid pollen tube growth requires dynamic remodeling of the actin cytoskeleton, which is finely regulated by numerous actin-binding proteins (ABPs). Therefore, understanding of ABP-mediated actin regulation of pollen germination and pollen tube growth at high temperature will help shed light on the evolution of angiosperms.

Question: How does the actin cytoskeleton respond to high temperatures and what is the mechanism by which it regulates pollen germination and pollen tube growth at high temperature?

Findings: Angiosperm actin-depolymerizing factors (ADFs), a family of ABPs, have been identified as limiting factors in rapid actin turnover. Here, we revealed that the evolution of the high thermal stability of ADFs played a pivotal role in the origin and evolution of adaptation to high temperature in Arabidopsis (Arabidopsis thaliana). Specifically, (i) 11 AtADFs show substantially different thermal stabilities, with ADFs that are expressed specifically in polar cells (pollen and root hairs) showing high thermal stability. (ii) The Sub II ADFs AtADF7 and AtADF10, which are expressed specifically in pollen, originated from the ancestral protein ancADF-A with high thermal stability, but their thermostable sites differ from those of ancADF-A, which underwent a wave-like evolution of the loss of ancestral thermostable sites and regeneration of new sites. (iii) The high thermal stability and thermostable sites of AtADF7 are conserved in ADFs specific to angiosperms. (iv) The high thermal stability of ADFs is necessary for regulating actin turnover to promote pollen germination at high temperature.

Next steps: A series of ABPs act together to regulate the dynamics of the intracellular actin cytoskeleton; as next steps, we will explore and determine which other ABPs with high thermal stability are involved in regulating pollen germination and pollen tube growth at high temperatures.

Introduction

Flowering plants (angiosperms) are the most diverse group of land plants. Angiosperms undergo a unique double fertilization process (Dresselhaus and Franklin-Tong 2013; Hafidh and Honys 2021) in which pollen germination and pollen tube growth are key steps. Pollen tubes are the fastest-growing cells in angiosperms and exhibit polarized growth. Immobile sperm cells rely on precisely regulated pollen tube growth for rapid growth and transport to ovules to complete double fertilization, thereby initiating seed development (Cheung and Wu 2001; Dresselhaus and Franklin-Tong 2013). High temperature affects all stages of plant growth and development, particularly the reproductive process of angiosperms (Peng et al. 2004; Zhang et al. 2017). Pollination and fertilization in most flowering plants occur in the heat of summer, during which pollen germination and rapid pollen tube elongation also mainly occur during the hot part of the day. Male gametophytes are more sensitive to high-temperature stress than female gametophytes (Sato et al. 2006; Yang et al. 2009; Zinn et al. 2010; Barba-Montoya et al. 2018). High temperature leads to abnormal pollen development, reduced pollen viability, a reduced pollen germination rate, and inhibited pollen tube growth (Dupuis and Dumas 1990; Yang et al. 2009; Zinn et al. 2010; Muhlemann et al. 2018). However, it is currently unclear how angiosperm pollen evolved to adapt to high temperatures.

Pollen germination and pollen tube growth require the polar transport of cell wall materials, cell membrane materials, and essential functional proteins to the apical growth point. The proper morphology and dynamic changes in the actin cytoskeleton are indispensable for this process (Samaj et al. 2006; Cheung and Wu 2008; Hepler et al. 2013). The organization and dynamics of actin are finely regulated by numerous actin-binding proteins (ABPs; Gibbon et al. 1999; Vidali et al. 2001; Cheung and Wu 2008; Li et al. 2015). High temperature seriously affects the dynamics of the actin cytoskeleton and actin-mediated vesicle and substance polar transport (Parrotta et al. 2016), which may also be a key reason for high temperature–induced pollen abortion.

Actin-depolymerizing factors (ADFs)/cofilins comprise a family of highly conserved ABPs in eukaryotes that promote the dynamic turnover of actin filaments by severing and depolymerizing actin filaments, allowing them to participate in various cellular biological processes. ADF/cofilin activities are regulated by different signals and proteins (Maciver and Hussey 2002; Andrianantoandro and Pollard 2006; Bamburg and Bernstein 2008). The ADF/cofilin family of proteins is therefore referred to as homeostatic regulators or “functional nodes” in cell biology (Bernstein and Bamburg 2010). Plant ADFs are the most important ABPs that regulate the dynamic turnover of actin filaments and are involved in various aspects of plant growth, development, and stress responses (Henty et al. 2011; Zheng et al. 2013; Jiang et al. 2017; Zhu et al. 2017; Qian et al. 2019; Zhang et al. 2021; Bi et al. 2022; Jiang et al. 2022; Wang et al. 2023). High-temperature stress can cause irreversible protein inactivation. Some ABPs in animals and yeast tolerate high-temperature stress and remain active at high temperatures (Pivovarova et al. 2007; Johnson et al. 2018), but the thermal stability of plant ADFs is unclear.

In this study, we revealed the evolutionary history and formation mechanism of the high thermal stability of ADFs in flowering plants and the mechanism by which the actin cytoskeleton regulates pollen germination in Arabidopsis (Arabidopsis thaliana) at high temperatures. In addition, we propose that the high thermal stability of ADFs that are expressed specifically in pollen might represent a strategy for the evolution of sexual reproduction in angiosperms to help the plant adapt to high temperatures at the cell biology level.

Results

Arabidopsis ADF proteins exhibit diverse levels of thermostability

To explore the protein stability of AtADFs at high temperatures, we analyzed the high-temperature tolerance of 11 members of the Arabidopsis ADF protein family (Fig. 1A; Nan et al. 2017). As shown in Fig. 1B and Supplemental Fig. S1, after 45 °C treatment for 15 min, approximately half of the neofunctionalized Sub III AtADF proteins (ADF5/9) with bundling activity were denatured, while only 20% of the 3 other typical ADF subfamily members were denatured, even after treatment for 45 min at 45 °C; after treatment for 45 min at 60 °C, more than half of the Sub I AtADF proteins (ADF1 to ADF4) were denatured. However, more than 80% of the Sub II AtADF proteins (ADF7/8/11) remained undenatured (except for AtADF10, of which more than 60% remained). When the proteins were placed in a boiling water bath (the boiling point of water in Lanzhou, Gansu, China, is 95 °C) for 20 min, more than 80% of Sub I AtADFs were denatured, 25% of AtADF6 was not denatured, and more than 70% of AtADF7 and AtADF11 were not denatured. The above results show that the thermostability of Arabidopsis ADF family proteins is significantly divergent. Among these proteins, Sub II AtADFs, which are expressed specifically in tip-growing cells (root hairs and pollen tubes; Ruzicka et al. 2007), had the highest thermostability, followed by Sub IV AtADFs; however, the thermal stability of Sub I AtADFs was lower, and that of Sub III AtADFs was the lowest. Since Sub II AtADFs are expressed specifically in tip-growing cells (Ruzicka et al. 2007; Bou Daher et al. 2011; Daher and Geitmann 2012), we reasoned that the high thermostability of Sub II AtADFs might be required for the rapid tip growth of cells in angiosperms at high temperatures.

Figure 1.

Analysis of the thermal stability of AtADF and ancADF proteins. A) Evolution of AtADFs. AtADF family proteins can be divided into 4 subfamilies (I, II, III, and IV). Nodes A to J indicate the recent common ancestral ADF proteins (ancADFs) modified from our previous report. The scale bar corresponds to 0.1 estimated amino acid substitutions per site. B, C) Recombinant AtADF proteins B) and ancADFs C) were subjected to 45 °C for different durations (0, 15, 30, and 45 min), to 60 °C for different durations (0, 15, 30, and 45 min), and to 95 °C for different durations (0, 5, 10, and 20 min). Shown are the results of statistical analysis of the percentage of undenatured proteins in the supernatant among total proteins through densitometry. The error bars represent the Sds from 3 biological replicates.

Evolution of ADF protein thermostability

The ADF gene family has undergone continuous duplication throughout the process of plant evolution, especially in angiosperms (Nan et al. 2017); this family has evolved into a multigene family and has differentiated in terms of tissue-specific expression and protein function (Ruzicka et al. 2007; Nan et al. 2017). Although Sub I AtADFs and Sub II AtADFs have large differences in thermostability, their sequence similarity is relatively high. Through the swapping of domains (Fig. 2), we found that different sites between AtADF7 and AtADF4 are mainly distributed with the 40 N-terminal amino acids, and that 4 key and specific amino acids (11E/27Y/Δ37KΔ39K; Supplemental Fig. S2) are the main sites contributing to the differences in thermostability between AtADF7 and AtADF4. We reconstructed the structures of AtADF4 and AtADF7 by homology modeling (Supplemental Fig. S2D), which revealed that these 4 amino acid sites indeed contribute to the structural differences between AtADF4 and AtADF7. Specifically, the β-hairpin between the β1 and β2 strands is destabilized in ADF4 due to the presence of 2 additional lysines (37K and 39K). Furthermore, in AtADF7, Tyr27 (27Y) and Glu11 (11E) potentially establish hydrogen bonds with Lys17 (17K) and Gln38 (38Q), respectively, while in AtADF4, the corresponding residues do not appear to participate in hydrogen bonding. These structural characteristics potentially contribute to the enhanced heat stability observed in AtADF7 compared to AtADF4.

Figure 2.

The 40 N-terminal amino acids of AtADF7 comprise the key domains for thermostability. A) Schematic diagram of the swapped domains of AtADF4 (139 amino acids, AA) and AtADF7 (137 AA). The different rectangles represent the amino acid sequence of ADF7 or the amino acid sequence of AtADF4, respectively. B) AtADF4 and AtADF7 domain-swapped recombinant proteins were subjected to 45 °C for different durations (0, 15, 30, and 45 min), 60 °C for different durations (0, 15, 30, and 45 min), and 95 °C for different durations (0, 5, 10, and 20 min). The undenatured proteins in the supernatant (S) and the denatured proteins in the pellet (P) were analyzed by SDS–PAGE. C) Statistical results of the experiments shown in B). The error bars represent the Sds from 3 biological replicates.

To explore the evolutionary process of the thermostability of ADF proteins, we constructed a phylogenetic tree and analyzed the sequences of the ADF family in mosses, ferns, gymnosperms, and angiosperms. The thermostable sites of AtADF7 are not conserved in mosses or gymnosperms (Supplemental Figs. S3A to S6). Furthermore, we explored the protein thermostability of ADFs of these species and found that their thermostability varied widely, with no apparent linear evolutionary relationship (Supplemental Fig. S3, B and C). Therefore, we utilized the most recent common ancestor (MRCA) genes of ADFs (ancADFs) to model the evolutionary process of their thermostability via coalescent theory (Nan et al. 2017).

We analyzed the thermostability of ADF ancestral proteins (Fig. 1, A and C; Supplemental Fig. S7) and found that the oldest ADF ancestor protein, ancADF-A, had high thermal stability. This protein evolved into ancADF-B and I, where ancADF-I has high-temperature stability characteristics similar to those of ancADF-A, while the high-temperature stability of ancADF-B significantly decreased. ancADF-B lost part of the thermostable site (Fig. 3, A and C; Supplemental Fig. S8, A, E, and F) of ancADF-A. ancADF-B further evolved into 2 subclades: ancADF-C/D/E and Sub I AtADFs are expressed in vegetative organs, and ancADF-F/G/H and Sub II AtADFs are specifically expressed in tip-growing cells (pollen and root hairs). The thermal stability of ancADF-F is higher than that of ancADF-B. We found that this stability is mainly caused by the mutation of His at position 34 to Tyr and the mutation of Lys at position 46 to Gln in ancADF-B (Fig. 3, B and C; Supplemental Fig. S8, B, E, and F), suggesting that during the evolution from ancADF-B to ancADF-F, the improvement in thermostability may have been retained by adapting to the environment at that time.

Figure 3.

Resurrected ancestral ADF proteins reveal the evolution of the high thermal stability of AtADFs. A, B, D, and E) Recombinant proteins were subjected to 60 °C for different durations (0, 15, 30, and 45 min). Shown are the results of statistical analysis of the percentage of undenatured proteins in the supernatant among total proteins (relative content) through densitometry. The error bars represent the Sds from 3 biological replicates. Student's t test, **P < 0.01. Experiments were repeated 3 times with similar results. A) ancADF-A-m1 refers to the mutation of 36Y59S125H126Y of ancADF-A to 36F59T125Q126V, and ancADF-B-m1 refers to the mutation of 36F59T125Q126V of ancADF-B to 36Y59S125H126Y. B) ancADF-B-m2 refers to the mutation of 34H46K of ancADF-B to 34Y46Q, and ancADF-F-m1 refers to the mutation of 34Y46Q of ancADF-F to 34H46K. C) Tracking the evolution of AtADF7 thermostability. Different indicator bands represent the percentages of undenatured proteins in the supernatant among total proteins after being subjected to 60 °C treatment for 45 min. Indicator bands from top to bottom indicate 80% to 100%, 60% to 79%, 40% to 59%, 20% to 39%, and 0% to 19%, respectively. The numbers below the protein labels represent the relative content after being subjected to 45 °C for 45 min, 60 °C for 45 min, and 95 °C for 20 min. Mutations from 8M9A10N to 8L9K10T occurred in ancADF-I to ancADF-J. D) ancADF-F-m2 refers to the mutation of 18N45A of ancADF-F to 18E45G, and ancADF-G-m1 refers to the mutation of 18E45G of ancADF-G to 18N45A. E) ancADF-G-m3 refers to the deletion of 7 amino acids from the N-terminus of ancADF-G, the loss of 44K, and the mutation of 84A to 84D; AtADF7-m2 refers to the addition of 7 amino acids from the N-terminus of ancADF-G to the N-terminus of ADF7, the insertion of an amino acid K at position 37, and the mutation of 75D to 75A.

The ancADF-F family evolved into 2 branches and differentiated in terms of function and thermal stability: the ancADF-H branch retained high thermostability and ultimately evolved into AtADF8 and AtADF11, with high thermal stability; these proteins are specifically expressed in root hairs (Figs. 1 and 3). In another branch, ancADF-G protein has undergone both amino acid mutations (the mutation of Asn at position 18 to Glu and Ala at position 45 to Gly in ancADF-F), resulting in a decrease in thermal stability (Fig. 3, C and D; Supplemental Fig. S8, C, E, and F). However, ancADF-G has undergone amino acid mutations and retentions. The pollen-specific expression of AtADF7 and AtADF10 evolved, of which AtADF10 acquired only partial thermostable site mutations (loss of 7 N-terminal amino acids [N7] of ancADF-G and Lys at position 44 in ancADF-G). The thermal stability of AtADF10 is higher than that of ancADF-G, while AtADF7 acquired multiple thermostable site mutations (loss of N7 of ancADF-G and Lys at position 44 in ancADF-G, Ala at position 84 in ancADF-G to Asp at position 75 in AtADF7), causing very high thermostability (Fig. 3, C and E; Supplemental Fig. S8, D to F). When we mutated the 75-position amino acid Ala in AtADF10 to Asp to produce AtADF10m1, the thermal stability of AtADF10m1 significantly improved (Supplemental Figs. S8D and S9), further demonstrating that Lys-75 (75D) is a key site for the high thermal stability of AtADF7. AtADF10m2 (the addition of N7 of ancADF-G to the N-terminus of AtADF10) and AtADF10m3 (the addition of N7 of ancADF-G to the N-terminus of AtADF10 and the insertion of Lys-37 [37K] in AtADF10) exhibited lower thermal stability than AtADF10 (Fig. 3C; Supplemental Figs. S8D and S9), suggesting that the absence of N7 and the loss of Lys-37 (37K) are key to the increased thermal stability of AtADF10.

In addition, our previous work revealed that a functional transition from actin filament–severing activity to actin filament–bundling activity occurred in ancADF-I to ancADF-J (Nan et al. 2017). Therefore, we further analyzed whether mutations at this site affect the thermostability of ADF proteins. The mutation of sites (the mutations of Met, Ala, and Asn at positions 8, 9, and 10 to Leu, Lys, and Thr, respectively, in ancADF-I) of this transition resulted in reduced thermal stability of ancADF-J and Sub III AtADFs (Figs. 1 and 3; Supplemental Fig. S10, A to C), whereas AtADF6 retained a partial heat-resistant site from ancADF-A and still showed high thermal stability (Figs. 1 and 3; Supplemental Fig. S10A). During the evolution of ancADF-B to produce Sub I AtADFs, amino acid mutations occurred continually, resulting in poor thermal stability of Sub I AtADFs (Figs. 1 and 3; Supplemental Fig. S10D).

Taken together, the above results indicate that the thermostable sites of ancADF-A were partially retained only in Sub IV AtADF6 but were completely lost during the evolution of other subfamilies. However, during the process of evolution to produce Sub II AtADFs, ancADFs accumulated several new thermostable sites. Instead of linear increases in protein thermostability, the process reflected a wave-like evolution, resulting in the evolution of very high heat stability of Sub II AtADFs expressed specifically in tip-growing cells of angiosperms.

The thermostability of ADFs is conserved in angiosperms

To analyze whether the thermostable sites of AtADF7 are conserved and ubiquitous in angiosperms, we analyzed the homologous protein sequences of AtADF7s (Fig. 4A) of 12 angiosperms, finding that the thermostable site is highly conserved among these proteins across different species. The homologous AtADF7 proteins in 11 angiosperms had high thermal stability after 60 °C treatment for 45 min (Fig. 4, B and C). In addition, analysis of information from published reports and databases showed that many homologous AtADF7 proteins are also mainly expressed in pollen (Kim et al. 1993; Lopez et al. 1996; Allwood et al. 2002; Li et al. 2010; Khatun et al. 2016; Huang et al. 2020). The above results suggest that the key thermostable sites and evolution of AtADF7's high thermal stability are well conserved in angiosperms.

Figure 4.

Conservation of thermostable sites and thermostability of ADF7 homologous proteins in angiosperms. A) Amino acid sequence alignment of AtADF7 homologous proteins in different angiosperms (Al-500, AL7G27500, Arabidopsis lyrata; Cc-890, Cc02_g02890, coffee (Coffea canephora); Cs-600, Cucsa.129600, cucumber (Cucumis sativus); Os-470, LOC_Os02g44470, rice (Oryza sativa ssp. japonica); Zm-714, Zm00001d002714, maize (Z. mays); Tca-173, TCA.TCM_015173, cocoa (Theobroma cacao); Man-800, Manes.06G147800, cassava (Manihot esculenta); Soly-980, Solyc06g035980.2, tomato (S. lycopersicum); Hbr-010, HBR0911G010, rubber tree (Hevea brasiliensis); Cpa-58, Cpa.g.sc107.58, papaya (Carica papaya); Ccaj-195, C.cajan_13195.g, pigeon pea (Cajanus cajan); and NtADF7, XM_016605039, tobacco (Nicotiana tabacum). The asterisks and dashed boxes indicate the conserved amino acid sites. B) Recombinant ADF7 homologous proteins were subjected to 45 °C for different durations (0, 15, 30, and 45 min), 60 °C for different durations (0, 15, 30, and 45 min), and 95 °C for different durations (0, 5, 10, and 20 min), and the undenatured proteins in the supernatant (S) and denatured proteins in the pellet (P) were analyzed by SDS–PAGE. C) Statistical results of the experiments shown in B). Shown are the results of statistical analysis of the percentage of undenatured proteins in the supernatant among the total proteins through densitometry. The error bars represent the Sds from 3 biological replicates.

Sub II AtADFs (AtADF7 and AtADF10) promote pollen germination at high temperature

Pollen germination and the polarized growth of pollen tubes depend on the actin cytoskeleton and actin-mediated vesicular transport (Gibbon et al. 1999; Samaj et al. 2006; Cheung and Wu 2008; Hepler et al. 2013). To verify the role of the high thermal stability of pollen-expressed Sub II AtADFs (AtADF7 and AtADF10) in sexual reproduction in Arabidopsis, we analyzed pollen germination and pollen tube growth at different temperatures. At the normal culture temperature of 22 °C, the pollen germination of adf7 was not significantly different from that of the wild type (WT), while its pollen tubes were shorter and pollen tube growth rate was lower than that of the WT (Figs. 5, A and B, and 6, A and B; Supplemental Fig. S11), which is consistent with previous results (Zheng et al. 2013; Jiang et al. 2017). When the temperature was increased to 37, 38.5, or 40 °C, the pollen germination rate of the WT decreased, while the pollen germination rate of adf7 was significantly lower than that of WT (Fig. 5, A and C). However, the sensitivity of adf7 pollen tube growth to high temperatures was similar to that of WT (Figs. 5, A and B, and 6, A and B; Supplemental Fig. S11), suggesting that AtADF7 might not be involved in regulating pollen tube growth in response to high temperatures. Complementation with AtADF7 rescued the high-temperature sensitive phenotype of adf7 pollen (Fig. 7, C and D; Supplemental Fig. S12), indicating that AtADF7 is required for pollen germination at high temperatures.

Figure 5.

Sub II ADFs (AtADF7 and AtADF10) promote pollen germination at high temperature. A) Representative images of pollen germination in WT, adf7, and adf10 plants after 1.5 h at different temperatures (22, 37, 38.5, and 40 °C). The bar represents 100 µm. B) Statistical analysis of pollen tube length in WT, adf7, and adf10 plants after 1.5 h of pollen germination at different temperatures (22, 37, 38.5, and 40 °C). C) Statistical analysis of the germination rate after 1.5 h of growth of pollen from WT, adf7, and adf10 plants at different temperatures (22, 37, 38.5, and 40 °C). D) Arrangement of actin filaments labeled by LAT52-lifeact-EGFP in pollen grains after 30 min of pollen germination from WT, adf7, and adf10 plants at different temperatures (22, 37, and 40 °C). The bar represents 10 µm. E) Statistical analysis of average fluorescence intensity of pollen grains after 30 min of WT, adf7, and adf10 pollen germination at different temperatures (22, 37, and 40 °C). F) Statistical analysis of the correlation coefficients of changes in actin filament arrangement in WT, adf7, and adf10 at 37 °C. G) Representative images of pollen germination from WT, adf7, and adf10 plants after 1.5 h at 37 °C on medium with different concentrations of LatB. The bar represents 100 µm. H) Statistical analysis of the germination rate after 1.5 h of growth of pollen from WT, adf7, and adf10 plants at 37 °C on medium with different concentrations of LatB. The asterisks indicate significant differences compared with the values of WT. The error bars represent the Ses, according to Student's t test: *P < 0.05, **P < 0.01, and ***P < 0.001. B)n = 200 pollen tubes. C)n = 200 pollen grains. E)n = 30. F)n = 10. H)n = 200 pollen grains.

Figure 6.

The fertility of adf7 and adf10 is more sensitive to high temperature compared to the WT. A) Representative images of aniline blue staining of pistils that grew for 8 h at different temperatures (22 and 37 °C) after pollination. The arrows indicate the front of pollen tube growth. Scale bar: 100 μm. B) Statistical analysis of in vivo pollen tube lengths of WT, adf7, and adf10 plants after 8 h of pollination at different temperatures (22 and 37 °C). The error bars represent the Ses, according to t tests: *P < 0.05 and **P < 0.01 (n > 20 pistils). C) High temperature induces silique abortion. Representative images of WT, adf7, and adf10 plants that were recovered at 22 °C for 1 wk after treatment at 37 °C for 6 or 8.5 h. The arrow indicates that the silique is sterile. Bar, 1 cm. D, E) Statistical analysis of the number of aborted siliques and the lengths of the 3 shortest aborted siliques in C). The error bars represent the Ses, according to Student's t test: *P < 0.05 and **P < 0.01 (n = 54 seedlings).

Figure 7.

ADF7 lacking critical thermostable sites cannot compensate for the high temperature sensitivity of adf7 pollen germination. A) Representative images of pollen germination for 1.5 h under different LatB concentrations (0 and 2.5 nM) from WT, adf7, and lines of adf7 complemented by AtADF7, AtADF7m, AtADF8, ancADF-A, and PpADF. The bar represents 100 µm. B) Statistical results of the experiments shown in A). All experiments were repeated 3 times, and the results were consistent. The error bars represent the Ses (n = 200 pollen grains) according to t tests: **P < 0.01. C) Representative images of pollen germination for 1 h and 40 min at different temperatures (22 and 37 °C) from WT, adf7 plants, and lines of adf7 complemented by AtADF7, AtADF7m, AtADF8, ancADF-A, and PpADF. The bar represents 100 µm. D) Statistical results of the experiments in C). All experiments were repeated 3 times, and the results were consistent. The error bars represent the Ses (n = 200 pollen grains) according to Student's t test: **P < 0.01.

We also analyzed the pollen germination of adf10 in response to high temperatures. Like adf7, pollen germination in adf10 was also more sensitive to high temperatures than the WT, while the sensitivity of adf10 pollen tube growth to high temperatures was similar to that of WT (Figs. 5, A and B, and 6, A and B; Supplemental Fig. S11). Furthermore, we analyzed the silique filling rates of adf7 and adf10 at both 22 and 37 °C. At high temperature, the number of aborted siliques was higher in adf7 and adf10 than in WT, and adf7 and adf10 siliques were shorter than those of WT (Fig. 6, C to E). Moreover, we carried out limited pollination experiments with WT and adf7. There was no significant difference in fruit set between the adf7 mutant and WT at normal temperature (22 °C), but the abortion rate of the adf7 mutant was significantly higher than that of WT at high temperature (37 °C; Supplemental Fig. S13), suggesting that the more severe abortion of adf7 plants grown at high temperatures is indeed due to the greater sensitivity of adf7 pollen germination to high temperatures. These results indicate that pollen-expressed Sub II AtADFs are important for pollen germination and sexual reproduction at high temperatures.

To determine whether AtADF7 and AtADF10 promote pollen germination at high temperatures by regulating the dynamic turnover of actin filaments, we analyzed the arrangement of actin filaments in WT pollen grains at different temperatures by introducing the actin filament marker LAT52-lifeact-EGFP into WT plants and examining fluorescent signals. Figure 5, D and E, shows that with increasing temperatures, the fluorescence intensity of actin filaments in WT pollen grains decreased, indicating that the inhibition of pollen germination by high temperatures is closely related to the arrangement of the actin cytoskeleton in pollen grains. We also analyzed the dynamic arrangement of actin filaments in adf7 and adf10 pollen grains. With increasing temperature, the fluorescence intensity of actin filaments in adf7 and adf10 pollen grains decreased to a significantly smaller extent than that of WT (Fig. 5, D and E). Moreover, statistical analysis of the correlation coefficients of changes in actin filament arrangement found that at high temperatures, the overall dynamic correlation coefficient of actin filaments in adf7 and adf10 pollen grains was higher than that in WT pollen grains (Fig. 5F).

In addition, we used latrunculin B (LatB), an inhibitor of actin polymerization, to analyze whether the sensitivity of adf7 and adf10 pollen germination to high temperatures could be partially restored. The hypersensitivity of adf7 and adf10 pollen germination to high temperatures was restored in response to low concentrations of LatB (Fig. 5, G and H), indicating that adf7 and adf10 pollen germination was indeed more sensitive to high temperatures due to the slowdown in actin turnover. Taken together, the above results suggest that Sub II AtADFs of Arabidopsis regulate the dynamic turnover of actin filaments at high temperatures to promote pollen germination.

High thermal stability of ADFs contributes to pollen germination at high temperatures

To analyze the importance of the thermostable sites of Sub II AtADFs to pollen germination at high temperatures, we complemented the adf7 mutant with AtADF7m, encoding AtADF7 with a mutation of its thermostable site (referring to the addition of 7 amino acids from the N-terminus of ancADF-G to the N-terminus of AtADF7, the insertion of 2 amino acids Lys-37 and Lys-39, and the mutation of Asp-75 to Ala-75). We carefully analyzed the filament-severing and depolymerizing ability between AtADF7m and AtADF7 through kinetic actin assembly and disassembly assays and performed direct visualization of individual actin filaments by total internal reflection fluorescence (TIRF) light microscopy (Zheng et al. 2013; Jiang et al. 2017; Zou et al. 2021; Jiang et al. 2022). As shown in Supplemental Fig. S14, both AtADF7m and AtADF7 depolymerized actin filaments, showing similar activities in high-speed F-actin cosedimentation experiments (Supplemental Fig. S14, A and B). AtADF7m and AtADF7 showed similar capabilities in quenching the fluorescence of both 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (NBD)-ADP-G-actin and NBD-ATP-G-actin, which indicates that AtADF7m and AtADF7 exhibit similar binding affinities to ATP-G-actin and ADP-G-actin (Supplemental Fig. S14, C to E). Tracing the kinetic actin depolymerization process indicated that AtADF7m and AtADF7 had similar effects on actin depolymerization (Supplemental Fig. S14F); AtADF7m had a similar capability to AtADF7 in inhibiting actin nucleotide exchange (Supplemental Fig. S14G). TIRF analysis showed that AtADF7m and AtADF7 had similar capabilities in severing filaments and enhancing monomer dissociation from the pointed ends of actin filaments (Supplemental Fig. S14, H to J). The above results indicate that mutating the key thermostable sites of AtADF7 did not affect its severing or depolymerization of actin filaments.

Genetic complementation experiments showed that these mutation sites rescued the LatB-insensitive phenotype of adf7 in terms of pollen germination (Fig. 7, A and B; Supplemental Fig. S12). However, AtADF7m failed to rescue the sensitivity of adf7 pollen germination to high temperatures (Fig. 7, C and D). Taken together, these results indicate that the thermostable sites of AtADF7 are required for pollen germination at high temperatures. Finally, to verify the specificity of ADF7's high thermal stability for pollen germination at high temperatures, we examined the compensatory effects of different ADFs with high thermal stability on the high-temperature sensitivity of adf7 pollen germination. AtADF8, ancADF-A, and Physcomitrium patens ADF (PpADF; encoding proteins with high thermal stability) rescued the high-temperature sensitivity of adf7 pollen germination (Fig. 7, C and D). These results suggest that the high thermal stability of the ADF protein is important for sexual reproduction in angiosperms but that its function is not site specific.

Discussion

The double fertilization employed by angiosperms represents the most efficient form of sexual reproduction within the plant kingdom (Dresselhaus and Franklin-Tong 2013; Hafidh and Honys 2021), which provided a key factor in angiosperms flourishing during plant evolution. It is noteworthy that a majority of angiosperms have evolved to bloom in the hot summer seasons for successful reproduction (Magallon 2010; Dresselhaus and Franklin-Tong 2013) and can germinate at high temperatures, but when temperatures become more extreme, pollen germination rates decrease and pollen tubes become shorter (Sato et al. 2006; Yang et al. 2009; Zinn et al. 2010; Barba-Montoya et al. 2018). This has also been observed in plants in our study, such as maize (Zea mays B73), tomato (Solanum lycopersicum cv. Micro-Tom), and lily (Lilium Siberia; Supplemental Fig. S15). Therefore, deciphering the underlying mechanism of such processes under high-temperature conditions is crucial for gaining deeper insights into the sexual reproduction of angiosperms.

ADFs are responsible for the turnover of actin/actin dynamics (Henty et al. 2011; Zheng et al. 2013; Jiang et al. 2017; Nan et al. 2017; Zhu et al. 2017; Qian et al. 2019; Zhang et al. 2021; Bi et al. 2022; Jiang et al. 2022; Wang et al. 2023). In the current study, we observed remarkable differences in the thermostability of ADF proteins in Arabidopsis (Figs. 1 and 3). Our findings reveal that subclass I ADFs (AtADF1/2/3/4) and subclass III ADFs (AtADF5/9) exhibit the lowest thermal stability, whereas subclass II ADFs (AtADF7/8/10/11) and subclass IV ADF (AtADF6) display high thermal stability (Figs. 1 and 3). Moreover, AtADF7 and AtADF10 are highly expressed specifically in pollen cells (Ruzicka et al. 2007; Bou Daher et al. 2011; Daher and Geitmann 2012). Unlike other types of plant cells, pollen tubes grow very quickly, primarily via tip growth (Steer and Steer 1989). This rapid growth process strictly relies on highly dynamic and ordered vesicle transport mediated by the actin cytoskeleton to facilitate the continuous transport of proteins and substances related to cell growth and cell wall synthesis to the apex of the pollen tube (Samaj et al. 2006; Cheung and Wu 2008; Hepler et al. 2013). This growth process is highly temperature sensitive (Boavida and McCormick 2007).

Our study found that pollen germination of Arabidopsis loss-of-function mutants adf7 and adf10 was more vulnerable to high temperatures, with the turnover of actin filaments occurring much more slowly than in WT under high-temperature conditions (Fig. 5). Genetic complementation tests demonstrated that AtADF7 protein carrying mutations in its key thermostable sites was unable to restore the defective pollen germination of adf7 at high temperature. However, the other ADF proteins with high thermal stability partially complemented the high-temperature sensitivity of pollen germination in adf7 (Fig. 7). These results suggest that the thermostability of ADF proteins is vital for pollen germination at high temperatures.

We previously revealed that members of the Sub III ADFs have undergone neofunctionalization during evolution, evolving from the typical severing and depolymerizing of actin filaments to bundling and stabilizing actin filaments, in which AtADF5 has also been shown to be involved in regulating pollen germination (Nan et al. 2017; Zhu et al. 2017). Additionally, our experiments indicated that this novel function that evolved in Sub III ADFs was obtained by the acquisition of mutations in their heat-tolerant sites during evolution (Fig. 3; Supplemental Fig. S10); therefore, the acquisition of this novel function occurred at the expense of the loss of thermal stability. Taken together, our results suggest that the high thermal stability that evolved in Sub II ADF family members, such as AtADF7 and AtADF10, might represent a strategy for enhancing sexual reproduction in angiosperms in response to high temperature (Fig. 8). AtADF7 and AtADF10 are also expressed during meiosis in the process of microspore development (Bou Daher et al. 2011; Daher and Geitmann 2012), and meiosis is also very sensitive to high temperature. Therefore, at high temperatures, AtADF7 and AtADF10 may also play an important role in the meiotic process of microspore development, which needs to be further studied (Fig. 8). In addition, the dynamic equilibrium of the actin cytoskeleton is coregulated by a series of ABPs, suggesting there may be other heat-stable ABPs involved in regulating pollen germination at high temperature (Fig. 8).

Figure 8.

A proposed model for the evolution of ADF thermal stability in the adaptation of Arabidopsis pollen germination to high temperature. The Earth's geological temperature has fluctuated significantly over the past 375 MYA, and the thermal stability of ADF proteins has exhibited corresponding fluctuations, as ADFs undergo a dynamic wave-like evolution of the retention, loss, and regeneration of thermostable sites. Interestingly, AtADF7 and AtADF10 diverged around 114 MYA, which is roughly consistent with the continuous warming of the Earth's environment and the rise of flowering plants in the Early Cretaceous period. In Arabidopsis, at high temperature, pollen expressing ADF7 and ADF10 with highly dynamic actin can germinate normally and complete double fertilization, while pollen expressing ADF7m with less dynamic actin, which is not heat tolerant, germinates poorly, resulting in abortion. Global temperature represents the Earth's geological climate changes: High represents high temperature, and Low represents low temperature. The pie chart on the left shows the 11 key amino acid (AA) sites related to thermal stability during the evolution of ADFs (N7 is 7 N-terminal amino acids); the remaining pie charts (from left to right) refer to ancADF-A, ancADF-B, ancADF-F, ancADF-G, AtADF7, and AtADF10, respectively; 18E, 36F, 45G, 59T, 125Q, and 126V represent sites with thermal instability; 34Y, 46K, and 75D represent heat-resistant sites, and Δ44K and ΔN7 indicate that the sites were lost and the thermal stability of ADF increased. The height of the column is proportional to the thermal stability. In the diagram on the right, the thermometer and the sun represent plants at high temperature (HT). The solid circles in pollen represent AtADF7m, AtADF7, AtADF10 or other possible heat-stable ABPs, respectively.

The ADF gene family has undergone multiple duplication events during plant evolution, particularly during the evolution of angiosperms (Nan et al. 2017), thus evolving into a multigene family. During evolution, in addition to the mutations in the coding regions of ADF genes, mutations in cis-elements of their promoters have also occurred, leading to both functional divergence and expression pattern diversification of ADF family members (Ruzicka et al. 2007; Nan et al. 2017). Our findings demonstrate that subfamily II (AtADF7 and AtADF10) has gained the function to regulate the heat tolerance of pollen germination (Figs. 3, 6, and 7). Based on the phylogenetic tree of ancestral ADFs (Nan et al. 2017), by tracing the origin and evolution of the heat-tolerant–related sites of Sub II ADFs, we found that these sites have undergone a series of evolutionary events, including loss, acquisition, and retention, leading to a fluctuation in the heat tolerance function of ADF, including a decrease, followed by an increase, decrease, and another increase (Figs. 3 and 8).

Using K_s (synonymous [silent] substitution rates) as the proxy, we previously estimated the evolutionary frequencies of ADF genes (Nan et al. 2017) and found that the divergence of Sub I ADFs and Sub II ADFs occurred at around 270 million years ago (MYA), during the Carboniferous–Permian Ice Age. During this geological period, the Earth was cold, and heat-tolerant sites were lost in ancADF-B after its divergence from ancADF-A (Figs. 3 and 8). Subsequently, from 270 MYA to 212 MYA, the Earth underwent a transition from glacial to warm environments, with ancADF-F emerging after its divergence from ancADF-B and acquiring new heat-tolerant sites (Figs. 3 and 8). Notably, ancADF-G diverged from ancADF-F and lost some heat-tolerant sites; this finding is in line with the notion that a period of decreasing Earth's temperature occurred before the rise of angiosperms (Figs. 3 and 8). During the evolution of ancADF-G to the angiosperm Sub II ADFs, some heat-tolerant sites were retained and new heat-tolerant sites were generated (Fig. 3; Supplemental Figs. S8 and S9). AtADF7 and AtADF10 diverged around 114 MYA, which is roughly consistent with the continuous warming of the Earth's environment (180 MYA to 100 MYA) and the rise of flowering plants in the Early Cretaceous period (145 MYA to 100 MYA; Magallon 2010; Herendeen et al. 2017; Barba-Montoya et al. 2018; Li et al. 2019; Ramirez-Barahona et al. 2020).

Our study also showed that the heat-tolerant sites of AtADF7 are highly conserved in angiosperms (Fig. 4). In addition, we analyzed the amino acid sequences of AtADF7 proteins of A. thaliana accessions from different latitudes and found that these thermostable sites showed no latitude-dependent thermotolerance and are conserved among Arabidopsis accessions (Supplemental Fig. S16). These findings suggest that the diversification of ancADF-G to AtADF7 and AtADF10 occurred during the explosion of angiosperms, perhaps to ensure that angiosperms can complete sexual reproduction at higher temperatures. Collectively, our findings suggest that the wave-like evolution of the thermal stability of proteins from ancADF-A to ancADF-F, ancADF-G, and finally Sub II ADFs (AtADF7 and AtADF10) in angiosperms is closely associated with plant adaptation to temperature changes on Earth over the past few hundred million years (Fig. 8).

Materials and methods

Plant materials and growth conditions

The WT A. thaliana used in this study was the Col-0 ecotype. The adf7 (salk_024576; Zheng et al. 2013) T-DNA insertion mutant was obtained from the Arabidopsis Biological Resource Center (ABRC) of Ohio State University, United States. The adf10 (ADF10 TALEN line #10) mutant and LAT52-Lifeact-EGFP/adf10, which were kindly provided by Dr. Shanjin Huang (Tsinghua University), were described previously (Jiang et al. 2017). The transgenic materials used in this study include plants harboring the pollen actin marker LAT52-Lifeact-GFP (Zhu et al. 2017). The Arabidopsis seeds were sterilized by soaking in 15% sodium hypochlorite for 10 min, washed with sterile water 6 times, vernalized at 4 °C for 3 d, sown on MS medium, incubated for approximately 7 d, and transplanted in nutrient-enriched soil for growth in a greenhouse. The culture conditions in the greenhouse were 22 ± 1 °C, long-day conditions (16-h light/8-h dark cycle), a light intensity of 80 to 100 μmol/m²/s, and a relative humidity of 60%.

Vector construction and gene synthesis

The vectors used in this study were generated by enzyme digestion and ligation and by using the Gateway recombination system. To study the function of ADFs, we amplified their coding DNA sequences (CDSs) with forward and reverse primers and cloned them into the pET30a vector with the corresponding restriction endonuclease (Nan et al. 2017). All genes with point mutations were generated by de novo DNA synthesis by Synbio Technologies (Suzhou, China) and verified by DNA sequencing. For the complementation constructs, the promoter sequence of ADF7 (including the first exon) was inserted into a pBIB vector to generate pBIB-ADF7p-GWR. Different ADF CDSs, including the attB1 and attB2 sites, were amplified via PCR, and the amplified fragment was cloned into the donor vector using BP Clonase II enzyme mix (Life Technologies) to generate pDonr-ADFs. Point mutation sequences (AtADF7m), AtADF7, AtADF8, predicted ancADF-A, and homologous genes in the moss P. patens by gene synthesis were amplified by PCR to supplement the attB sites, and the amplified fragment was cloned into the donor vector using BP Clonase II enzyme mix, yielding pDonr-AtADF7m, AtADF7, AtADF8, ancADF-A, and PpADF. These entry clones were used in a MultiSite Gateway LR recombination reaction, yielding pBIB-ADF7p-AtADF7, AtADF7m, AtADF8, ancADF-A, and PpADF. All primer sequences used for cloning are listed in Supplemental Table S1. All complementation constructs were transformed into adf7 plants by the floral dip method using the Agrobacterium tumefaciens strain GV3101 (Clough and Bent 1998). Homozygous lines were selected from among plants of the T₃ generation and used for further experiments.

Protein expression and purification

To investigate the thermal stability of ADFs, ADF cDNA was amplified using the corresponding primers and subcloned into the pET30a vector using a specific restriction endonuclease, and the plasmids were transformed into Escherichia coli Rosetta cells (Nan et al. 2017). The E. coli cells were cultured with LB medium grown at 37 °C and shaken at 200 rpm in an incubator until they reached an optical density at 600 nm (OD₆₀₀) of 0.4 to 0.6. The culture was then supplemented with 0.3 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and further grown at 16 °C for 16 h. His-tagged recombinant proteins were purified via a previously described method (Nan et al. 2017). The purified proteins were subsequently dialyzed overnight in TBS buffer. Before treatment, the dialyzed proteins were centrifuged at 14,500 × g for 30 min.

High-speed cosedimentation assays

To detect the actin filament-severing and depolymerization activity of different recombinant proteins, a high-speed cosedimentation experiment was performed as described previously (Xiang et al. 2007; Nan et al. 2017). Before the experiment, all recombinant proteins were centrifuged at 100,000 × g for 1 h. The actin used in this study was purified from rabbit skeletal muscle acetone powder via a previous method (Pardee and Spudich 1982). The relative actin contents in the supernatant and pellet were analyzed by SDS–PAGE. ImageJ software was used to analyze the gray values of the target protein bands in the supernatant and pellet.

Analysis of the thermal stability of proteins

The concentrations of the dialyzed proteins were measured. The proteins were diluted to 1 mg/mL, after which they were subjected to heat treatment in water baths with different temperatures (45, 60, and 95 °C) for different durations, immediately removed, and placed into an ice bath for cooling. The heat-treated proteins were subsequently centrifuged at 14,500 × g for 30 min at 4 °C, and the supernatant and pellet were obtained. The undenatured proteins in the supernatant and the denatured proteins in the pellet were analyzed by SDS–PAGE. ImageJ software was used to analyze the gray values of the target protein bands in the supernatant and pellet.

In vitro pollen germination and analysis of pollen tube growth rate

In vitro pollen germination assays were performed according to a previous report (Zhu et al. 2017). To analyze the pollen germination rate at different temperatures, pollen grains were cultured on solid germination medium under different temperature conditions for a certain period of time. Images were taken under a Zeiss microscope (Axio Imager.Z2), and the corresponding germination rate was analyzed. To analyze the effect of LatB on pollen germination, different concentrations of LatB (Abcam, ab144291) were added to the solid medium. To analyze pollen tube growth, pollen tubes growing for 1.5 h on solid medium were imaged, and the pollen tube length was analyzed.

Pollen tube growth rates at different temperatures were analyzed as previously described (Zheng et al. 2013; Zhu et al. 2017) with the following modifications: images of elongating pollen tubes that germinated on solid germination medium in 22 and 37 °C incubators for 110 and 130 min were captured, respectively. The length of pollen tubes was measured with ImageJ, and the average elongation rate at different temperatures was calculated according to the following equation: (L130 − L110)/20 min. L130 represents the average lengths of the pollen tubes at 130 min, and L110 represents the average lengths of the pollen tubes at 110 min. More than 200 pollen tubes per genotype at 22 °C and more than 100 pollen tubes at 37 °C were analyzed. Four independent repeated experiments were performed.

Quantification of actin filaments in pollen grains

To observe the distribution of actin filaments in pollen grains, we introduced the marker LAT52-lifeact-EGFP for actin filaments into WT plants and obtained LAT52-lifeact-EGFP lines in the adf7 background by genetic crossing. Images of the actin cytoskeleton in pollen grains were collected using a spinning disk confocal microscope (Andor, Cell Observer) equipped with a constant temperature chamber (22, 37, and 40 °C). The excitation wavelength of EGFP was set to 488 nm, and the emission wavelength was 520 nm. Images were taken with a 63× oil lens, and the z-step was set to 0.3 µm. To quantify actin dynamics in pollen grains, only half of the pollen grains were imaged in multiple planes (usually 20) at 3-s intervals for 3 min, which were integrated and constructed into serial images. The relevant parameters were analyzed by ImageJ.

Binding of ADF to G-actin

The G-actin binding activity of ADF was determined by the NBD-actin binding assay as described previously (Detmers et al. 1981; Zheng et al. 2013). NBD-ATP-G-actin and NBD-ADP-G-actin were prepared as described previously (Chaudhry et al. 2007). Briefly, to prepare NBD-ATP-G-actin, 20 µM NBD-G-actin was incubated with 1 mM ATP for 70 min on ice in the dark. To prepare NBD-ADP-G-actin, 20 µM NBD-actin was incubated with 20 units/mL hexokinase (Sigma-Aldrich) and 1 mM glucose (Sigma-Aldrich) for 3 h on ice in the dark. Then, 0.2 µM NBD-ATP-G-actin or NBD-ADP-G-actin was incubated with increasing concentrations of AtADF7 or AtADF7m in 100 mM KCl and 1 mM MgCl₂. The NBD (excitation at 475 nm and emission at 530 nm) fluorescence change was monitored with a FluoroMax-4 spectrofluorometer (HORIBA Jobin Yvon) for 20 s. Data were fitted using KaleidaGraph (Synergy Software, version 3.6) as previously reported (Zheng et al. 2013).

Nucleotide exchange analysis

The nucleotide exchange rate of ADP-G-actin was determined by measuring the increase in fluorescence upon binding of 1,N⁶-ethenoadenosine 5′-triphosphate (ɛ-ATP, Jena Bioscience, Jena, Germany) as described previously (Chaudhry et al. 2010; Jiang et al. 2017). The fluorescence was monitored using a FluoroMax-4 spectrofluorometer (HORIBA Jobin Yvon) for 10 min (excitation at 350 nm and emission at 410 nm). The data were fitted to a single exponential function using GraphPad Prism 5.0 (GraphPad, Boston, MA, United States) as described previously.

Actin filament depolymerization assay

To determine the depolymerizing activity of AtADF7 or AtADF7m, a dilution-mediated actin depolymerization assay was performed as described previously (Chaudhry et al. 2010; Zhu et al. 2017). The change in NBD fluorescence was monitored for 10 min with a Fluoro-Max-4^R spectrofluorometer (HORIBA Jobin Yvon; excitation at 350 nm and emission at 410 nm).

Direct observation of actin filament severing and monomer dissociation by TIRF microscopy

Single actin filament–severing events and monomer dissociation were directly observed by TIRF microscopy using a flow cell prepared as previously described (Shi et al. 2013; Zou et al. 2021). Monomeric actin labeled by Oregon-green was polymerized by the addition of 1 × KMEI (50 mM KCl, 1 mM MgCl₂, 1 mM EGTA, and 10 mM imidazole, pH 7.0). Actin filaments (15% to 50% Oregon-green labeled, 0.125 to 1 mM) were mixed with 2 × TIRF buffer (2 mM MgCl2, 100 mM KCl, 2 8 mM EGTA, 10 mM DTT, 0.4 mM ATP, 30 mM glucose, 200 µg/mL glucose oxidase, 40 µg/mL catalase, 1% methylcellulose, and 20 mM imidazole [pH 7.4]), and the mixture was transferred to a microscope flow cell for imaging at room temperature. To determine the severing activity of AtADF7 and AtADF7m, different concentrations of these 2 proteins were introduced after placing the chamber on the microscope stage. TIRF images were captured at 1- to 2-s intervals using ELYRA 7 (Carl Zeiss). The filament-severing frequency (breaks/µm/s) was quantified as previously reported (Andrianantoandro and Pollard 2006; Zheng et al. 2013). More than 20 actin filaments with lengths >10 µm were selected for quantification in each experiment. The monomer dissociation rate (subunits/s) was determined as previously described (Zhang et al. 2016).

In vivo pollen tube growth and seed setting rate assay

Mature pistils from 5-wk-old WT Arabidopsis plants were pollinated with pollen grains from WT, adf7, or adf10 plants. The plants were placed in constant temperature incubators (22 or 37 °C) for 8 h. The pollinated pistils were collected and immersed in fixation solution (ethanol:acetic acid = 3:1) for 2 h. Fixed pistils were sequentially treated with 70%, 50%, and 30% ethanol and ddH₂O for 10 min each and treated with 8 M NaOH overnight. The next day, the pistils were washed 3 times with ddH₂O for 10 min each and stained (protected from light) for 12 h with aniline blue solution (0.1% aniline blue in 100 mM K₃PO4 buffer, pH 8.0; Yang et al. 2022). A fluorescence microscope (Olympus BX53, Japan) was used to observe the phenotypes of the pistils. Five-week-old Arabidopsis plants in the flowering period were placed in constant temperature incubators (22 or 37 °C) with a light intensity of 80 to 100 μmol/m²/s (bulb type, TCL, TCLMY-28) and relative humidity of 60% for different times and allowed to recover for 7 d in a greenhouse with a constant temperature of 22 ± 1 °C. The seed setting rates and lengths of the treated siliques were analyzed.

Reverse transcription quantitative PCR

To measure the expression of different ABPs in the adf7 background, floral tissue was obtained from different complementation transgenic lines. RNA was extracted from the tissue with an RNA extraction kit (Takara) and reverse transcribed into cDNA with a reverse transcription kit (Takara). Gene expression was analyzed by reverse transcription quantitative PCR (RT-qPCR), and EF4A was used as the internal reference gene. The relevant primer sequences used for amplification are listed in Supplemental Table S2.

Phylogenetic analysis

ADF protein sequences from different plant species were obtained from the PPG database (https://www.plabipd.de/plant_genomes_pa.ep). All ADF protein sequences were aligned using MAFFT v7.4.02 (Katoh and Standley 2013). RAxML (version 8.0.26; Stamatakis 2014) was used to construct a maximum likelihood tree with no outgroup setting. A total of 1,000 bootstrap analyses were executed to test the robustness of each node, and the bootstrap values are labeled beside the branches. The amino acid sequence alignments are provided in Supplemental File 1.

Amino acid sequence alignment and structural modeling

To explore the sequence similarity of the ADFs, BioEdit was used for amino acid sequence alignment and subsequent analysis. The homology models of AtADF4 and AtADF7 were constructed using SWISS-MODEL based on the crystal structure of ADF1 (PDB ID code1F7S; Bowman et al. 2000). PyMOL (version 1.4.1; DeLano Scientific) and VMD software (Humphrey et al. 1996) were used to illustrate the structures.

Statistical analysis

Experiments were performed at least 3 times independently. The error bars show the Sds or Ses (as indicated). SPSS software was used for statistical analysis, and one-way ANOVA or Student's t test was used to analyze significant differences. Differences were considered significant at *P < 0.05, **P < 0.01, and ***P < 0.001 (Supplemental Data Set 1).

Accession numbers

Sequence data from this article can be found in the GenBank/EMBL libraries under the following accession numbers: AtADF4, AT5G59890; AtADF7, AT4G25590; AtADF8, AT4G00680; and AtADF10, AT5G52360.

Acknowledgments

We appreciate Dr. Shanjin Huang (Tsinghua University) for sharing seeds of the adf10 mutant and Lifeact-EGFP/adf10. We thank Dr. Guangpeng Ren (Lanzhou University) for his help with the ADF sequence download. We thank the Core Facility of the School of Life Sciences, Lanzhou University, for technical assistance.

Author contributions

Conceived and designed the experiments: Y.X. and D.Q. Performed the experiments: D.Q., T.L., S.C., C.Z., Jia.L., Z.S., Jie.L., J.S., Ying.N., H.L., and M.W. Provided theoretical contributions to the project: L.A., Y.H., and D.W. Analyzed the data and wrote the paper: Y.X., D.Q., T.L., Y.N., and Y.Y.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. SDS–PAGE analysis of AtADF recombinant proteins after heat treatment.

Supplemental Figure S2. Identification of key sites underlying the differences in thermal stability of the 40 N-terminal amino acids between AtADF4 and AtADF7.

Supplemental Figure S3. Alignment of ADF sequences and SDS–PAGE analysis of ADF recombinant proteins from mosses to angiosperms after heat treatment.

Supplemental Figure S4. Phylogenetic tree of ADFs from different species.

Supplemental Figure S5. Amino acid sequence alignment of ADF proteins in different ferns.

Supplemental Figure S6. Amino acid sequence alignment of ADF proteins in different gymnosperms.

Supplemental Figure S7. SDS–PAGE analysis of ancADF recombinant proteins after heat treatment.

Supplemental Figure S8. Identification of differentially thermostable sites between different recombinant proteins.

Supplemental Figure S9. Identification of thermostable sites in AtADF10.

Supplemental Figure S10. Identification of differentially thermostable sites between ancADF-I and ancADF-J and between ancADF-B and ancADF-C/D/E.

Supplemental Figure S11. The pollen tube elongation rate is significantly reduced in adf7 and adf10 plants.

Supplemental Figure S12. The corresponding gene expression levels determined by RT-qPCR.

Supplemental Figure S13. The limited pollination of adf7 is more sensitive to high temperature compared to the WT.

Supplemental Figure S14. AtADF7m and AtADF7 have comparable activity in enhancing actin turnover in vitro.

Supplemental Figure S15. Pollen germination and pollen tube growth of different species at high temperature.

Supplemental Figure S16. Sequence alignment of ADF7 putative orthologous proteins in different ecotypes of A. thaliana.

Supplemental Table S1. Primers used for plasmid construction: pBIB-ADF7p-ADF7, ADF7m, ADF8, ancADF-A, and PpADF.

Supplemental Table S2. Primers used for RT-qPCR.

Supplemental File 1. Alignment data of ADFs from different species.

Supplemental Data Set 1. Statistical analysis in this study.

Funding

This work was supported by National Natural Science Foundation of China grants (32170331, 31970195, 32170330, and 31700161), a project from Hainan Yazhou Bay Seed Lab (B23YQ1510), and Fundamental Research Funds for the Central Universities from Lanzhou University (lzujbky-2022-ey06 and lzujbky-2023-eyt02).

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• ADF1 Gramene: AT3G46010

• ADF1 Araport: AT3G46010

• ADF4 Gramene: AT5G59890

• ADF4 Araport: AT5G59890

• ADF7 Gramene: AT4G25590

• ADF7 Araport: AT4G25590

• ADF10 Gramene: AT5G52360

• ADF10 Araport: AT5G52360

• pollen AmiGo: PO:0025281

• ATP CHEBI: CHEBI:15422

Author notes

Dong Qian, Tian Li and Shuyuan Chen contributed equally.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plcell/) is: Yun Xiang ([email protected]).

Conflict of interest statement. The authors declare no competing interests.