Fig 1.
RUPO is expressed specifically in rice pollen.
(A) The genomic structure of RUPO and positions of T-DNA insertion site and CRISPR target sites (sg1 and sg2). RUPO is an intronless gene. The black rectangle represents the exon. (B) RT-PCR and (C) quantitative PCR analysis of RUPO transcripts in different tissues. 18S ribosomal RNA was used as an internal control. The relative expression is presented as mean ± s.e.. (D) Western blot analysis of RUPO protein in pollen. 20 μg of total proteins from pollen was loaded in each lane, resolved on 4~15% SDS-PAGE gel and detected with anti-RUPO polyclonal antibody. Asterisk indicates RUPO band. Upper panel: western blot; bottom panel: protein loading control stained with Coomassie blue. UNM, uninucleate microspore; BCP, bicellular pollen; TCP, tricellular pollen; MPG, mature pollen grain; GPG, germinated pollen grain.
Table 1.
Segregation analysis of heterozygous rupo+/- mutant.
Fig 2.
In vitro pollen germination and growth assay.
(A) Wild-type pollen. (B) rupo+/- pollen. Red arrowheads indicate ruptured pollen tubes. (C) Pollen germination rate and percentage of pollen tube integrity. The results are presented as mean±s.e. (D) Time-lapse images of wild-type pollen germinating on agarose medium. White arrows indicate the growth of a wild-type pollen tube. (E) Time-lapse images of bi-allelic pollen germinating on agarose medium. Note that rupo pollen tubes burst and released cell contents in seconds without forming any visible pollen tubes.Time is indicated in minutes:seconds format. Scale bars, 50 μm. (see also S1 and S2 Movies in the supplementary material).
Fig 3.
Homozygous bi-allelic mutant pollen tubes burst in pistils.
(A to H) Pollen tube growth in pistils of wild-type (A,E), heterozygous T-DNA mutant rupo+/- (B,F), mono-allelic mutant (C,G) and bi-allelic mutant (D,H). White arrowheads indicate the putative positions of micropyles. All ovules were targeted by pollen tubes except for ovules from the bi-allelic mutant. Inserts in (E,F,G,H) are close-ups of representative pollen tubes in wild-type (E), rupo+/- (F), mono-allelic (G) and bi-allelic (H) stigmas. Red arrows in (H) indicate callus spots. (I,J,K,L) Bright-field images of representative pistils without aniline blue staining from wild-type (I), rupo+/- (J), mono-allelic (K) and bi-allelic (L) plants. Scale bars, 100 μm. (M) The percentage of ovules targeted by pollen tubes. n = 109 for wild-type pistils, n = 106 for rupo+/- pistils, n = 95 for mono-allelic pistils and n = 163 for bi-allelic pistils. (N) Pollen number per pistil before or after aniline blue staining. n = 79 (before staining) or 75 (after staining) for wild-type pistils, n = 74 or 107 for rupo+/- pistils, n = 59 or 89 for mono-allelic pistils, and n = 80 or 145 for bi-allelic pistils. Results are presented as mean ± s.e..
Fig 4.
RUPO localizes to the plasma membrane and cytoplasmic vesicles of the pollen tube.
(A-F) Subcellular localization of the 35S::RUPOΔC-GFP fusion protein in onion epidermal cells. (A) Single confocal section and (B) bright-field image of the cell bombarded with 35S::RUPOΔC-GFP plasmid. (C,D) The same cell as in (A,B) was treated with 0.8 M mannitol to induce plasmolysis. (E) Single confocal section and (F) bright-field image of the cell bombarded with 35S::GFP. (G,H,I) Subcellular localization of the Ubi::RUPO-GFP fusion protein in lily pollen tubes. (G,H) RUPO or (I) GFP alone driven by ubiquitin promoter was transiently expressed in lily pollen tubes. Scale bars, 50 μm in (A) to (G), 5 μm in (H,I). (J-K) RUPO was enriched in membrane fraction. Homogenate from mature pollen grains was centrifuged at 10,000×g, and resultant supernatant (S10) was further centrifuged at 100,000×g to separate into supernatant (S100) and membrane fraction (P100). Proteins in each fraction were resolved by SDS-PAGE and probed with anti-RUPO antibody. 30 μg soluble protein (S10, S100) and 10 μg membrane protein (P100) were loaded. (J) Western blot image, (K) Protein loading control image stained with Coomassie brilliant blue. (L) Crude membrane vesicles were fractionized on a discontinuous sucrose density gradient, and proteins in individual fractions were probed with antibodies against RUPO, ER marker protein SMT1 and PM marker protein H+-ATPase.
Fig 5.
RUPO interacts with potassium transporters.
(A) The intracellular domain (without juxtamembrane region) of RUPO interacts with OsHAK1-C in yeast two-hybrid assay. (B) In vitro pull-down assay of RUPO and OsHAK1/19/20 interaction. GST-RUPO-C or GST was detected by anti-GST antibody. MBP-OsHAK-C or MBP was detected by anti-MBP antibody. (C) Co-immunoprecipitation assay of RUPO and OsHAK1/19/20 interaction. (D) Determination of K+ in MPGs from wild-type and bi-allelic plants by inductively coupled plasma atomic emission spectroscopy (ICP-OES). Results are presented as mean ± s.e.. Student’s test *p<0.05. (E) Effect of K+ on pollen tube integrity in vitro. (F) Effect of K+ on pollen germination in vitro. Each data point is the result of at least four replicates.
Fig 6.
Phosphorylation regulates RUPO-OsHAK interaction.
(A) Schematic diagram of RUPO used in the kinase assay and yeast-two hybrid assay. SP, signal peptide; TM, transmembrane domain; JM, juxtamembrane region. MBP, maltose binding protein. (B) The full-length RUPO (the 28 aa signal peptide is not included in this construct) is autophosphorylated, whereas its K543R point mutation version RUPO*, and JM-truncated intracellular region (MBP-RUPO515, and RUPO515) have no kinase activity. RUPO also is capable of phosphorylating MBP and MBP-RUPO515 but not RUPO515. (C) K543R point mutation abolishes the interaction of RUPO with OsHAK1 in yeast. JM and C-terminal region (788-845aa) are important elements affecting the interaction of RUPO and OsHAK1. (D) MS/MS identification of phosphorylation sites in the intracellular domain of RUPO. Phosphorylated residues are highlighted in bold red. JM region is underlined. (E) A representative MS/MS spectrum of identified phosphopeptides. The matched b and y ions are indicated.
Fig 7.
A proposed RUPO-potassium transporter signaling model to control pollen tube growth and integrity via regulating K+ homeostasis.
The PM-localized RUPO is autophosphorylated, interacts with OsHAKs, and negatively regulates the high-affinity K+ transporter activity, finally establishing a K+ homeostasis. As the pollen tube enters the receptive synergid, unknown signals may cause change in RUPO phosphorylation, the inactivated RUPO no longer interacts with OsHAKs. The released OsHAKs cause K+ influx and over-accumulation in pollen tubes. High level potassium-mediated increase in turgor pressure leads to tube discharge.