Introduction

Pollen tube growth is an essential step leading to reproductive success in flowering plants. After germination on compatible stigmatic tissues, pollen tubes penetrate stigma papilla cell wall, enter the complex stylar tissues, grow into the ovary along the transmitting track, turn toward the micropyle, and finally enter the embryo sac for sperm release (Dresselhaus and Franklin-Tong 2013; Johnson et al. 2019). Pollen tube growth involves a full spectrum of intracellular activities, including but not limited to calcium gradient, levels of reactive oxygen species, vesicular trafficking, wall deposition, dynamic actin microfilaments and so on (Qin and Yang 2011; Craddock et al. 2012; Hao et al. 2022).

Vesicle-mediated membrane dynamics is critical for pollen tube growth and mutations of trafficking regulators often result in defective pollen tube growth. Mutations at Arabidopsis Sec1/Munc18 (SM) proteins caused abnormal accumulation of secretory vesicles in pollen tubes and reduced tube growth (Beuder et al. 2022). Similarly, mutations at Arabidopsis RabA4d, a small GTPase critical for secretion, compromised pollen tube growth (Szumlanski and Nielsen 2009). Functional loss of several CONSERVED OLIGOMERIC GOLGI (COG) components in Arabidopsis resulted in abnormal Golgi morphology and reduced pollen tube growth (Johnson et al. 2016; Rui et al. 2020). Defective vacuolar trafficking by compromising either Adaptor Protein-3 (AP-3) or Rab5 GTPases resulted in defective pollen tube growth in vitro and in vivo (Feng et al. 2017; Hao et al. 2023). Mutations affecting endocytic trafficking, such as in the mutants of the clathrin-mediated endocytosis (CME) pathway, also compromised pollen tube growth (Geitmann et al. 2019).

Vesicular trafficking from endoplasmic reticulum (ER) to the Golgi apparatus is mediated by the coat protein complex II (COPII) while secretion-associated Ras-related GTPase 1 (Sar1) is an essential component of the COPII complex. The Arabidopsis genome encodes 5 Sar1 isoforms (Kirchhausen 2000; Russell and Stagg 2010; Brandizzi 2017), which play distinct yet sometimes redundant roles in various cellular and developmental processes (Zeng et al. 2015, 2021; Chung et al. 2016; Li et al. 2021), especially in reproduction (Liang et al. 2020, 2023). Arabidopsis Sar1b is essential for sporophytic control of pollen development while Sar1b and Sar1c are critical for gametophytic control of pollen development (Liang et al. 2020). Because functional loss of Sar1b and Sar1c resulted in pollen abortion, whether they influence pollen tube growth was unclear.

Here we demonstrate that Sar1b mediates pollen tube growth, in addition to its role in pollen development. Although functional loss of Sar1b does not affect pollen germination, it causes a significant reduction in male transmission and pollen tube penetration of the style. The pollen tube defects of sar1b are rescued not only by Sar1b but also by Sar1c. We further show that membrane dynamics at the apex of pollen tubes are compromised by Sar1b loss-of-function, suggesting Sar1b is critical for membrane homeostasis during pollen tube growth.

Results and discussion

Functional loss of Sar1b results in reduced pollen tube growth

We previously demonstrated that the homozygous sar1b mutants failed to produce viable pollen (Liang et al. 2020), indicating the role of Sar1b in sporophytic control of pollen development. Although pollen development was not abnormal in the heterozygous sar1b, i.e. sar1b/+ (Liang et al. 2020), we hardly obtained homozygous plants from self-fertilized sar1b/+. These results suggested a role of Sar1b in male transmission. Indeed, by reciprocal crosses and segregation ratio analysis, we determined that functional loss of Sar1b alone severely compromised male transmission (Table 1). In comparison, functional loss of Sar1c did not affect male transmission (Table 1). Introducing UBQ10: GFP-Sar1b or UBQ10: GFP-Sar1c rescued the reduced male transmission of sar1b (Table 1), indicating that Sar1c could substitute Sar1b for this process.

Table 1 Functional loss of Sar1b impairs male gametophytic transmission
Full size table

To determine at which process after development sar1b pollen was defective, we first performed in vitro pollen germination assays using UBQ10: GFP-Sar1b/−; sar1b-2 or UBQ10: GFP-Sar1c/−; sar1b-2. The single-copy heterozygous transgene expressed GFP-Sar1 fusions in half of pollen grains, thus would provide an easy way to distinguish GFP-Sar1; sar1b (complemented) pollen grains/tubes from the sar1b pollen grains/tubes (Fig. S1).

There was no significant difference between GFP-Sar1; sar1b (complemented) pollen grains and sar1b pollen grains regarding to pollen germination ratio (Fig. S2), suggesting that Sar1b was not critical for the germination process. However, the length of sar1b pollen tubes was significantly reduced compared to that of GFP-Sar1; sar1b pollen tubes or that of wild type (Fig. 1c, e–g). On the other hand, sar1b pollen tubes were slightly but significantly thinner than those of GFP-Sar1; sar1b pollen tubes or of wild-type pollen tubes (Fig. 1d).

Fig. 1
figure 1

Sar1b is critical for pollen tube growth. a, b Histochemical GUS staining of a pollen tube from the pSar1b:GUS plants (a) or a wild-type pistil pollinated with pollen grains from the pSar1b:GUS plants (b). c, d Quantification of pollen tube length (c) or width (d) in in vitro germination assays. Pollen from wild type (WT), two independent lines of UBQ10: GFP-Sar1b/−; sar1b-2 (Sar1b#1; sar1b and Sar1b#2; sar1b), or two independent lines of UBQ10: GFP-Sar1c/−; sar1b-2 (Sar1c#1; sar1b and Sar1c#2; sar1b) were incubated on pollen germination medium (PGM) for 3.5 h before imaging. Dark-green columns indicate GFP-positive tubes (complemented) whereas grey columns indicate GFP-negative tubes (sar1b). Results shown are means ± standard errors (SE, n = 4). Every biological replicate includes more than 50 tubes for every genotype. Different letters indicate significantly different groups (OneWay ANOVA, Tukey’s multiple comparisons test, P < 0.01). e, g Representative in vitro germination assays using pollen from wild type (e), from the UBQ10: GFP-Sar1b/−; sar1b-2 plants (f), or from the UBQ10: GFP-Sar1c/−; sar1b-2 plants (g). Images shown are merges of the transmission and GFP channels. Non-fluorescence tubes were pseudo-colored in pink. h, i Length (h) or percentage (i) of pollen tubes growing out of the styles. Pistils were cut at the style after 30 min of pollination; styles were placed on PGM for 10 h before imaging and quantification. Results shown are means ± standard deviation (SD). For every genotype, more than 10 pistils were quantified in semi in vivo assays. Dark-green columns indicate GFP-positive tubes (complemented) whereas grey columns indicate GFP-negative tubes (sar1b). j, k A representative semi in vivo assay. The wild-type pistil was either pollinated with pollen from the UBQ10: GFP-Sar1b/−; sar1b-2 plants (j) or from the UBQ10: GFP-Sar1c/−; sar1b-2 plants (k). Non-fluorescence tubes were pseudocolored in pink. Bars = 50 μm for (a, b, e, f, g); 100 μm for (j, k)

Full size image

We next determined whether Sar1b was important for pollen tube growth in vivo by using semi in vivo assays, which allowed us to distinguish GFP-Sar1; sar1b (complemented) pollen tubes from the sar1b pollen tubes (Fig. S1). A more severe defect of sar1b pollen tubes was detected in semi in vivo assays such that more than 80% pollen tubes growing out of the style were GFP-positive tubes, i.e. GFP-Sar1; sar1b pollen tubes and only 20% pollen tubes growing out of the style were sar1b pollen tubes (Fig. 1i–k). In addition, sar1b pollen tubes growing out of the style were significantly reduced in length compared to either GFP-Sar1; sar1b pollen tubes or wild-type pollen tubes (Fig. 1h, j, k), suggesting a greater role of Sar1b in pollen tube growth in vivo. The results were consistent with the fact that Sar1b was expressed in pollen tubes either in vitro (Fig. 1a) or in vivo (Fig. 1b).

Functional loss of Sar1b compromises membrane homeostasis at the apex

Sar1 as the COPII component mediates vesicular trafficking, which would influence membrane dynamics especially at the apex. To test this idea, we first applied the lipophilic dye FM4-64, which enters the cell through endocytosis. By examining the uptake of FM4-64 in GFP-Sar1; sar1b pollen tubes versus sar1b pollen tubes at different time points, we determined that sar1b pollen tubes showed a significantly reduced uptake of FM4-64 (Fig. 2a, h). Fluorescence signals of FM4-64 were clearly detected at the apical and subapical cytoplasm of GFP-Sar1; sar1b pollen tubes as soon as 5 min uptake (Fig. 2a, g, h). In comparison, fluorescence signals of FM4-64 were mostly associated with the PM of sar1b pollen tubes at the same time point (Fig. 2b, g, h). Although FM4-64 uptake occurred in sar1b pollen tubes over time (Fig. 2c, f), consistently reduced signals were detected in sar1b versus GFP-Sar1; sar1b in pollen tubes (Fig. 2g, h), indicating compromised endocytosis in pollen tubes by Sar1b loss-of-function.

Fig. 2
figure 2

Functional loss of Sar1b affects membrane homeostasis at the apical PM of pollen tubes. a, f FM4-64 uptake assays. Growing GFP-Sar1b;sar1b-2 (a, c, e) or sar1b-2 pollen tubes (b, d, f) were stained with FM4-64 and CLSM imaging was performed at 5 min (a, b), 30 min (c, d), or 60 min (e, f). For (a, c, e), both RFP and GFP channel images are shown. g, h Intensity of endocytosed FM4-64 at the apical cytoplasm (g) or at the subapical cytoplasm (h) after 5 min, 30 min, or 60 min uptake. Region of Interest (ROI) was defined as within 5 µm toward the very apex of pollen tubes (g) or between 5 to 10 µm toward the very apex (h); a.u., arbitrary fluorescence unit. Results are means ± SD (n > 30). P values are shown on top (t-test). i, j Quantification of fluorescence intensity at the apical plasma membrane (PM) for Calcofluor white-stained (i) or S4B-immunostained (j) pollen tubes. ROI was defined as the apical PM; a.u. indicates arbitrary fluorescence units. Results are means ± SD (n > 30). P values are shown on top (t-test). k, n Representative CLSM of a GFP-Sar1b;sar1b-2 (k, m) or sar1b-2 pollen tube (l, n) stained with Calcofluor white (k, l) or immune-stained with S4B (m, n) for cellulose. For (k, l), pseudo-colors cover the full range of measured values within each dataset (High to Low, H to L). The pseudo-color display represents the linear value. Bars = 5 µm

Full size image

Endocytosis is balanced with exocytosis to maintain membrane homeostasis of growing pollen tubes. To determine whether exocytosis was also affected by Sar1b loss-of-function, we examined the deposition of cellulose, whose asymmetric patterning is an indicator of polar exocytosis in pollen tubes (Mollet et al. 2013). By Calcofluor white (Chebli et al. 2012) and by Fast Scarlet 4B (S4B) staining (Wang et al. 2011; Hao et al. 2023), we determined that strong fluorescence signals were detected at the apex of sar1b pollen tubes (Fig. 2l, n) in comparison to that of GFP-Sar1; sar1b pollen tubes (Fig. 2k, m). The significantly increased deposition of cellulose at the apex of sar1b pollen tubes (Fig. 2i, j) correlated with the reduced pollen tube growth (Fig. 1c) and compromised endocytosis (Fig. 2g, h).

We report here a novel function of Arabidopsis Sar1b in reproduction, i.e. in the penetrative growth of pollen tubes through the style, which explains the reduced male transmission of sar1b (Table 1). Similar to the role of Sar1b in tapetum (Liang et al. 2020), compromised pollen tube growth of sar1b could be restored by exogenous Sar1c (Fig. 1), indicating that these two isoforms were inter-changeable on the protein level. Functional loss of Sar1b inhibited endocytosis at the apical PM as determined by the uptake of FM4-64 (Fig. 2), supporting the idea that membrane homeostasis requires a dynamic balance for pollen tube growth. Indeed, functional loss of vacuolar trafficking regulator Rab5 GTPases was recently demonstrated to cause the failure of pollen tube penetration of style partially by inhibiting endocytosis at the apical PM of pollen tubes (Hao et al. 2023). Extensive studies on COPII components have demonstrated their critical roles in pollen development and germination (Conger et al. 2011; Tanaka et al. 2013; Aboulela et al. 2018; Liu et al. 2021). Whether specific isoforms of different COPII components participate with Sar1b for style penetration of pollen tubes remains to be explored.

Materials and methods

Plant materials and growth conditions

Arabidopsis thaliana ecotype Columbia-0 was used as wild type. Plant materials including sar1b, UBQ10: GFP-Sar1b/−; sar1b-2, UBQ10: GFP-Sar1c/−; sar1b-2, and pSar1b:GUS were either described or derived from described materials (Liang et al. 2020). Seeds were vernalized in darkness at 4 °C for 3 days before being transferred to a greenhouse with a 16-h-light/8-h-dark cycle at 22 °C. Determination of genotypes in reciprocal crosses was performed as described (Liang et al. 2020).

GUS histochemistry

For in vitro growing pollen tubes, pollen from pSar1b:GUS were germinated on pollen germination medium (PGM) for 3.5 h and dropped with 5-bromo-4-chloro-3-indolyl-D-GlcUA (X-Gluc) (1 mg/ml). For in vivo growing pollen tubes, pollen from pSar1b: GUS were placed onto emasculated wild-type stigma and pistils were stained with X-Gluc (1 mg/ml) at 5 h after pollination (HAP). Imaging of GUS-stained pollen tubes was captured with an Olympus BX53 microscope.

Phenotypic analyses and imaging of pollen tubes

For in vitro pollen germination and tube growth, mature pollen was placed onto PGM at 28 °C for 3.5 h before quantification, as described, which was described previously (Feng and Zhang 2017). For semi in vivo assays, wild-type pistils were cut at the style at 30 min after saturated pollination, placed on PGM for 10 h before imaging.

For FM4-64 uptake assays, in vitro growing pollen tubes were dropped with liquid PGM supplemented with 5 µM FM4-64 (Invitrogen), incubated at 28 °C in the dark; imaged after 5 min, 30 min, or 60 min. For cellulose staining assays, in vitro growing pollen tubes were dropped with liquid PGM supplemented with 0.01% S4B (Direct Red 23; Sigma) or 0.01% Calcofluor white (Sigma) as described (Sheng et al. 2012; Hao et al. 2023).

Fluorescent imaging was performed with a Zeiss LSM880 (Zeiss). The excitation/emission for GFP-fusions are 488 nm/505–550 nm; for FM4-64 and S4B staining are 561 nm/575–650 nm; for Calcofluor white staining are 458 nm/463 to 500 nm.

Statistical analysis

Statistical analysis for male transmission was performed using χ2 analysis. Other quantification data were analyzed using GraphPad Prism 6.02 (www.graphpad.com/scientificsoftware/prism/). Statistical analysis was performed with Student’s t-tests or OneWay ANOVA.