Cyclic nucleotide-gated channel 18 is an essential Ca2+ channel in pollen tube tips for pollen tube guidance to ovules in Arabidopsis

Abstract

In flowering plants, pollen tubes are guided into ovules by multiple attractants from female gametophytes to release paired sperm cells for double fertilization. It has been well-established that Ca(2+) gradients in the pollen tube tips are essential for pollen tube guidance and that plasma membrane Ca(2+) channels in pollen tube tips are core components that regulate Ca(2+) gradients by mediating and regulating external Ca(2+) influx. Therefore, Ca(2+) channels are the core components for pollen tube guidance. However, there is still no genetic evidence for the identification of the putative Ca(2+) channels essential for pollen tube guidance. Here, we report that the point mutations R491Q or R578K in cyclic nucleotide-gated channel 18 (CNGC18) resulted in abnormal Ca(2+) gradients and strong pollen tube guidance defects by impairing the activation of CNGC18 in Arabidopsis. The pollen tube guidance defects of cngc18-17 (R491Q) and of the transfer DNA (T-DNA) insertion mutant cngc18-1 (+/-) were completely rescued by CNGC18. Furthermore, domain-swapping experiments showed that CNGC18's transmembrane domains are indispensable for pollen tube guidance. Additionally, we found that, among eight Ca(2+) channels (including six CNGCs and two glutamate receptor-like channels), CNGC18 was the only one essential for pollen tube guidance. Thus, CNGC18 is the long-sought essential Ca(2+) channel for pollen tube guidance in Arabidopsis.

Keywords: Arabidopsis; CNGC18; Ca2+ channel; Ca2+ gradient; pollen tube guidance.

Fig. S1.

CNGC7, CNGC8, CNGC9, CNGC10, and CNGC16 are all Ca²⁺-permeable channels. (A and B) Typical whole-cell recordings (A) and average current-voltage curves (B) for CNGC7. Nine HEK293T cells were tested for CNGC7, CNGC7 + 0 mM Ca²⁺, CNGC7 + 10 mM Ba²⁺, and CNGC7 + 100 μM 8Br-cAMP; 6 cells for CNGC7 + 100 μM Gd³⁺; and 5 cells for CNGC7 + 100 μM 8Br-cGMP. (C and D) Typical whole-cell recordings (C) and average current-voltage curves (D) for CNGC8. Ten HEK293T cells were tested for CNGC8, 8 cells for CNGC8 + 0 mM Ca²⁺, 6 cells for CNGC8 + 10 mM Ba²⁺, 3 cells for CNGC8 + 100 μM Gd³⁺, 7 cells for CNGC8 + 100 μM 8Br-cGMP, and 6 cells for CNGC8 + 100 μM 8Br-cAMP. (E and F) Typical whole-cell recordings (E) and average current-voltage curves (F) for CNGC9. Seven HEK293T cells were tested for CNGC9, 6 cells for CNGC9 + 0 mM Ca²⁺, 4 cells for CNGC9 + 10 mM Ba²⁺, 6 cells for CNGC9 + 100 μM Gd³⁺, and 5 cells for CNGC9 + 100 μM 8Br-cGMP and CNGC9 + 100 μM 8Br-cAMP. (G to H) Typical whole-cell recordings (G) and average current-voltage curves (H) for CNGC10. Eight HEK293T cells were tested for CNGC10, 6 cells for CNGC10 + 0 mM Ca²⁺, 7 cells for CNGC10 + 10 mM Ba²⁺, 6 cells for CNGC10 + 100 μM Gd³⁺, 5 cells for CNGC10 + 100 μM 8Br-cGMP, and 3 cells for CNGC10 + 100 μM 8Br-cAMP. (I and J) Typical whole-cell recordings (I) and the average current-voltage curves (J) for CNGC16. Eleven HEK293T cells were tested are for CNGC16, 9 cells for CNGC16 + 0 mM Ca²⁺, 5 cells for CNGC16 + 10 mM Ba²⁺, 6 cells for CNGC16 + 100 μM Gd³⁺, 9 cells for CNGC16 + 100 μM 8Br-cGMP, and 8 cells for CNGC16 + 100 μM 8Br-cAMP. C7, C8, C9, C10, and C16 denote CNGC7, CNGC8, CNGC9, CNGC10, and CNGC16, respectively. Nineteen HEK293T cells were tested for mock control, and the same control data were used to plot the average current-voltage curve in B, D, F, H, and J. Error bars depict means ± SEM. *Significant difference (P < 0.05) and **significant difference (P < 0.01), relative to cyclic nucleotide-free conditions.

Fig. S2.

CNGC7, CNGC8, CNGC9, CNGC10, and CNGC16 exhibit no obvious permeability to K⁺ relative to the typical inward K⁺ channel KAT1. (A and B) Typical whole-cell recordings (A) and the average current-voltage curves (B). Eight HEK293T cells were tested for mock control, 10 cells for CNGC7, 9 cells for CNGC8, 7 cells for CNGC9, 8 cells for CNGC10, 9 cells for CNGC16, and 13 cells for KAT1. (C) The amplitudes of average steady-state whole-cell currents at −160 mV as shown in B. Error bars depict means ± SEM. **Significant difference from mock control (P < 0.01).

Fig. 1.

cngc18-17 and cngc18-22 show severe male sterility. (A) Schematic model showing the cngc18-17 (R491Q) and cngc18-22 (R578K) point mutations. N, TM, and CNBD denote the N-terminal domain, transmembrane domains, and cyclic nucleotide binding domain, respectively. (B, C, and E) Images of plants (B), siliques (C), and in vitro-grown pollen tubes (E). (Scale bars: B, 2 cm; C, 0.5 cm; and E, 100 μm.) (D, F, and G) Average seed number per silique (D), pollen germination rate (F), and average pollen tube length (G). Error bars depict means ± SEM. **Significant differences from WT (P < 0.01).

Fig. S3.

RT-PCR to demonstrate that cngc7 (CS870639), cngc8 (CS409627), cngc9 (Salk_026086), cngc10 (Salk_015952C), cncg16 (Salk_065792C), glr1.2 (Salk_031498), and glr3.7 (Salk_103942C) are T-DNA insertion knock-out mutants.

Fig. S4.

qRT-PCR to assess relative gene expression levels in transgenic Arabidopsis lines. Expression of proCNGC18:eGFP in transgenic line proCNGC18-eGFP/cngc18-17 was set to 1.0.

Fig. 2.

The R491Q and R578K point mutations in CNGC18 resulted in severe pollen tube growth defects whereas the T-DNA insertion mutations in the other seven Ca²⁺ channels had no pollen tube growth defects. (A) In vivo aniline blue staining and scanning electron microscopy images, showing typical on-target pollen tubes of WT, COM1, and COM2, examples of passing-by pollen tubes of cngc18-17 and cngc18-22, and an example of branched pollen tubes of cngc18-17. Pollen tubes and micropyles are indicated by white arrows and arrowheads, respectively. (Scale bars: 20 µm.) (B) Percentages of passing-by and branched pollen tubes. A “0” indicates that no branched pollen tubes were observed. **Significant difference from WT (P < 0.01). Error bars depict means ± SEM.

Fig. S5.

Pollen tube growth analysis. (A and B) Aniline blue staining images showing typical on-target pollen tubes of WT and seven channel mutants (cngc7, -8, -9, -10, and -16 and glr1.2 and -3.7) (A) and a typical on-target pollen tube of WT and off-target pollen tubes of transgenic lines expressing either of the five CNGCs (CNGC7, -8, -9, -10, and -16) in the cngc18-17 background (B). Pollen tubes are indicated using white arrows in A and B. (C) The percentages of total off-target pollen tubes, including passing-by and branched pollen tubes. Pollen tubes and micropyles were indicated by white arrows and arrowheads, respectively. (Scale bars: 20 µm.) proC18:C7-eGFP denotes proCNGC18:CNGC7-eGFP, and the others are similarly represented.

Fig. S6.

Schematic for domain swapping between CNGC18 and CNGC16. NT, TM, and CT denote N-terminal domain, transmembrane domains, and C-terminal domain, respectively.

Fig. S7.

The chimeric channels CNGC16-18C, CNGC16-18N, CNGC16-18TM, and CNGC18-16TM are all Ca²⁺-permeable channels and can be activated by cyclic nucleotide cGMP. (A and B) Typical whole-cell recordings (A) and the average current-voltage curves (B) for CNGC16-18CT. Seven HEK293T cells were tested for CNGC16-18CT and CNGC16-18CT + 100 μM 8Br-cGMP. (C and D) Typical whole-cell patch-clamping recordings (C) and the average current-voltage curves (D) for CNGC16-18NT. Eight HEK293T cells were tested for CNGC16-18N and 6 cells for CNGC16-18NT + 100 μM 8Br-cGMP. (E and F) Typical whole-cell recordings (E) and the average current-voltage curves (F) for CNGC16-18TM. Ten HEK293T cells were tested for CNGC16-18TM and 6 cells for CNGC16-18TM + 100 μM 8Br-cGMP. (G and H) Typical whole-cell recordings (G) and the average current-voltage curves (H) for the CNGC18-16TM. Eleven HEK293T cells were tested for CNGC18-16TM and 8 cells for CNGC18-16TM + 100 μM 8Br-cGMP. (I) The amplitudes of average steady-state whole-cell currents recorded at −180 mV as shown in B, D, F, H, Fig. S1J, and Fig. 4B. Seven HEK293T cells were tested for Mock and Mock + 8Br-cGMP, and the same Mock and Mock + 8Br-cGMP data were used to plot the average current-voltage curves in B, D, F, and H. Error bars depict means ± SEM. *Significant difference relative to cyclic nucleotide-free conditions (P < 0.05); **significant difference relative to cyclic nucleotide-free conditions (P < 0.01).

Fig. S8.

The transmembrane domains of CNGC18 are essential for pollen tube growth and guidance. (A) Aniline blue staining (Upper) and scanning electron microscopy observation (Lower) of pollen tubes attracted by ovules, showing that the pollen tubes of qrt, CNGC18/cngc18-1 (−/−), and CNGC16-18TM/cngc18-1 (−/−) precisely entered micropyles. Pollen tubes and micropyles are indicated by white arrows and arrowheads, respectively. (Scale bars: 20 µm.) (B) The proportion of pollen tubes that precisely entered an ovule. Three replicates were performed. Error bars depict means ± SEM. Note that the background of cngc18-1 is qrt and that the qrt mutant was used as a positive control, as reported (37).

Fig. 3.

The CNGC18 point mutations R491Q and R578K impair activation of CNGC18 by cGMP. (A and B) Typical whole-cell recordings (A) and the average current-voltage curves (B) of whole-cell currents recorded in HEK293T cells. (C and D) Typical whole-cell recordings (C) and the average current-voltage curves (D) of whole-cell currents recorded in Arabidopsis pollen tube protoplasts. Extracellular 8Br-cGMP and intracellular cGMP treatments were used to activate channel currents in HEK293T cells and pollen tube protoplasts, respectively. C18, R491Q, R578K, c18-17, and c18-22 denote WT CNGC18, CNGC18-R491Q, CNGC18-R578K, cngc18-17, and cngc18-22, respectively. Error bars depict means ± SEM. **Significant difference from WT (P < 0.01).

Fig. 4.

CNGC18 point mutations R491Q and R578K lead to irregular drifts of basal [Ca²⁺]_cyt levels in Arabidopsis pollen tube tips. (A) Time course of typical [Ca²⁺]_cyt oscillation in the pollen tube tips of WT, cngc18-17, and cngc18-22. (B) Fractions of pollen tubes showing irregular drifts of basal [Ca²⁺]_cyt levels. Thirty-five pollen tubes were tested for WT, 33 for cngc18-17, and 39 for cngc18-22. Dashed lines indicate the expected relatively stable basal [Ca²⁺]_cyt levels. Error bars depict means ± SEM. **Significant difference from WT (P < 0.01).