A bipartite molecular module controls cell death activation in the Basal cell lineage of plant embryos

Abstract

Plant zygote divides asymmetrically into an apical cell that develops into the embryo proper and a basal cell that generates the suspensor, a vital organ functioning as a conduit of nutrients and growth factors to the embryo proper. After the suspensor has fulfilled its function, it is removed by programmed cell death (PCD) at the late stages of embryogenesis. The molecular trigger of this PCD is unknown. Here we use tobacco (Nicotiana tabacum) embryogenesis as a model system to demonstrate that the mechanism triggering suspensor PCD is based on the antagonistic action of two proteins: a protease inhibitor, cystatin NtCYS, and its target, cathepsin H-like protease NtCP14. NtCYS is expressed in the basal cell of the proembryo, where encoded cystatin binds to and inhibits NtCP14, thereby preventing precocious onset of PCD. The anti-cell death effect of NtCYS is transcriptionally regulated and is repressed at the 32-celled embryo stage, leading to increased NtCP14 activity and initiation of PCD. Silencing of NtCYS or overexpression of NtCP14 induces precocious cell death in the basal cell lineage causing embryonic arrest and seed abortion. Conversely, overexpression of NtCYS or silencing of NtCP14 leads to profound delay of suspensor PCD. Our results demonstrate that NtCYS-mediated inhibition of NtCP14 protease acts as a bipartite molecular module to control initiation of PCD in the basal cell lineage of plant embryos.

Figure 1. Expression pattern of NtCYS.

(A) Classification of successive stages of embryogenesis in tobacco. DAP, days after pollination. Scale bars, 20 µm. (B) Semi-quantitative RT-PCR analysis of NtCYS in sperm cell (lane 1), egg cell (lane 2), zygote (lane 3), two-celled proembryo (stage 1 of embryogenesis; lane 4), eight-celled embryo (stage 3; lane 5), 32-celled embryo (stage 4; lane 6) and heart-shaped embryo (stage 8; lane 7). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a control. (C) RT-qPCR analysis of NtCYS in the embryos at successive developmental stages (1–9) and in both floral and vegetative tissues. The expression level of NtCYS in the proembryos at stage 1 was set to 1. (D–I) Localization of NtCYS-GFP at the early stages of embryogenesis in proNtCYS::NtCYS-GFP plants. (D) Two-celled proembryo (stage 1). (E) Three-celled proembryo (between stages 1 and 2). (F) Four-celled proembryo (stage 2). (G) Eight-celled embryo (stage 3). (H) 32-celled embryo (stage 4). (I) Early globular embryo (stage 5). Scale bars, 10 µm. Asterisks indicate the basal cell.

Figure 2. Dynamics of DNA fragmentation and plasma membrane permeabilization in the tobacco embryo-suspensor.

(A–E) Nuclear DNA fragmentation in the embryos at stages 4 to 8, respectively, as revealed by TUNEL. Scale bars, 10 µm. (F–K) Plasma membrane permeabilization and cell viability of in the embryos at stages 4 to 9, respectively, revealed by FDA and PI staining. Scale bars, 10 µm. (L) The frequency of suspensors containing indicated numbers of TUNEL-positive nuclei at stages 3 to 8. Data represent the mean ± SE from five independent experiments, with 30 embryos per stage analysed in each experiment (n = 150). (M) The frequency of suspensors containing at least one PI-positive cell at stages 6 to 9. Data represent the mean ± SE from four independent experiments, with 30 embryos per stage analysed in each experiment (n = 120). Asterisks indicate the basal cell.

Figure 3. Vacuolization of the cytoplasm, nuclear envelope disassembly, and caspase-like proteolytic activity during suspensor PCD.

(A–D) Morphology of the basal cell analysed by TEM in the embryos at stages 6 (A), 7 (B), 8 (C), and 9 (D). Scale bars, 2 µm in (A–C) and 1 µm in (D). n, nucleus; v, vacuole; nm, nuclear membrane; pm, plasma membrane. (E–G) In situ detection of active proteases with caspase 1-like (E), 3-like (F), and 6-like (G) specificity in the suspensor cells at stages 4 to 7, as revealed by staining with indicated caspase-specific fluorescent inhibitors. Scale bars, 20 µm. Asterisks denote the basal cell. (H) The frequency of suspensors with caspase-like activity at the developmental stages 3 to 7. Data represent the mean ± SE from four independent experiments, with 30 embryos per stage analysed in each experiment (n = 120).

Figure 4. Downregulation of NtCYS induces precocious cell death in the basal cell lineage and seed abortion.

(A) Reduced expression of NtCYS in RNAi lines, as measured by RT-qPCR. The expression level of NtCYS in the WT was set to l. (B) The frequency of the two-celled proembryos with PI-positive basal cells in WT and RNAi lines. Data represent the mean ± SE from five independent experiments, with 50 proembryos per line analysed in each experiment (n = 250). (C) Cell viability in the two-celled proembryos from WT and RNAi lines stained with FDA and PI. Scale bars, 10 µm. (D) Nuclear DNA fragmentation in the two-celled proembryos from WT and RNAi lines stained with TUNEL. Scale bars, 10 µm. (E) The frequency of two-celled proembryos with TUNEL-positive basal cells in WT and RNAi lines. Data represent the mean ± SE from three independent experiments, with 30 proembryos per line analysed in each experiment (n = 90). (F–K) Morphology of the apical (G, J) and basal (H, K) cells analysed by TEM in the two-celled proembryos from WT (F–H) and RNAi lines (I–K). Scale bars, 5 µm in (F) and (I), and 2 µm in (G, H, J and K). Ac, apical cell; Bc, basal cell; n, nucleus; v, vacuole; nm, nuclear membrane; cw, cell wall; pm, plasma membrane. (L) Seed abortion in RNAi lines. Asterisks indicate aborted seeds. Scale bars, 1 mm. (M) Aberrant cell division patterns in the apical cell lineage at the early stages of embryogenesis in RNAi lines. Scale bars, 10 µm. (N) The frequency of aborted seeds in WT and RNAi lines. Data represent the mean ± SE from three independent experiments, with 400 to 500 seeds per line analysed in each experiment. (O, P) Cathepsin-like (O) and caspase-like (P) activities in the two-celled proembryos from WT and NtCYS RNAi line L2-6. Data represent the mean ± SE, with 40 to 60 two-celled proembryos. ** indicates statistical difference compared to WT (t-test, p<0.01).

Figure 5. NtCYS interacts with and inhibits cathepsin H-like protease NtCP14.

(A) BiFC analysis of the interaction between NtCYS and six cathepsins in tobacco epidermal cells. Scale bars, 50 µm. (B) Proteolytic activity of recombinant NtCP14 against different substrates. The activity of NtCP14 against each substrate is expressed as the percentage of NtCP14 activity against substrate FVR-AMC. Data represent the mean ± SE of three independent experiments. (C) Co-IP of NtCYS with NtCP14-GFP. GFP was used as negative control. Immunoblotting was performed with anti-NtCYS for NtCYS and anti-GFP for NtCP14-GFP or GFP. (D) NtCYS concentration-dependent inhibition of proteolytic activity of recombinant NtCP14 towards Bz-FVR-AMC. Data represent the mean ± SE of three independent experiments. (E) Cytoplasmic localization of NtCYS-GFP and NtCP14-RFP in the onion epidermal cells. Scale bars, 50 µm.

Figure 6. Constitutive overexpression of NtCP14 leads to cell death at the two-celled proembryo stage.

(A) Enhanced expression of NtCP14 under the control of proZC1 promoter in transgenic lines, as measured by RT-qPCR. The expression level of NtCP14 in the WT was set to l. (B) Cathepsin H-like proteolytic activity towards substrate Bz-FVR-AMC in the two-celled proembryos from WT and NtCP14-overexpressing line L14-4. Data represent the mean ± SE, with 45 to 65 two-celled proembryos in WT and line L14-4, respectively. (C) The frequency of aborted seeds in WT and NtCP14-overexpressing lines. Data represent the mean ± SE from three independent experiments, with 300 to 400 seeds per line analysed in each experiment. (D) Cell viability in the two-celled proembryos from WT and NtCP14-overexpressing line analysed by staining with FDA and PI. Scale bars, 10 µm. (E) The frequency of two-celled proembryos with PI-positive apical and basal cells in WT and NtCP14-overexpressing lines. Data represent the mean ± SE from four independent experiments, with 50 proembryos per line analysed in each experiment (n = 200). (F) Nuclear DNA fragmentation in the two-celled proembryos from WT and NtCP14-overexpressing line L14-4 stained with TUNEL. Scale bars, 10 µm. (G) The frequency of two-celled proembryos with TUNEL-positive apical and basal cells in WT and NtCP14-overexpressing line L14-4. Data represent the mean ± SE from three independent experiments, with 30 proembryos per line analysed in each experiment (n = 90). (H) Caspase-like activities in the two-celled proembryos from WT and NtCP14-overexpressing line L14-4. Data represent the mean ± SE, with 40 to 65 two-celled proembryos in WT and L14-4, respectively. (I) Cathepsin H-like activity of NtCP14 in the WT embryo proper versus suspensor at developmental stages 3 and 4. Data represent the mean ± SE, with 85 to 100 cells of embryo proper or suspensor analysed at stages 3 and 4. (J) Cathepsin H-like activity of NtCP14 in the basal cell lineage at successive stages of embryogenesis in WT plants. Data represent the mean ± SE, with 85 to 100 suspensor cells. * and ** indicate statistical difference compared to WT or as shown otherwise (t-test, p<0.05 or 0.01, respectively).

Figure 7. Upregulation of NtCYS or downregulation of NtCP14 delays the onset of suspensor PCD.

(A) Enhanced expression of NtCYS in overexpression lines, as measured by RT-qPCR. The expression level of NtCYS in the WT was set to l. (B) Decreased expression of NtCP14 in RNAi lines, as measured by RT-qPCR. The expression level of NtCP14 in the WT was set to l. (C–G) Representative examples of TUNEL-stained embryos in the NtCYS-overexpressing lines at stages 4 to 8, respectively. Scale bars, 20 µm. (H) The frequency of suspensors containing indicated numbers of TUNEL-positive nuclei in NtCYS-overexpressing line L-5 at stages 4 to 8. Data represent the mean ± SE from five independent experiments, with 30 embryos per stage analysed in each experiment (n = 150). (I,J) Plasma membrane permeabilization and cell viability in stage-8 (I) and stage-9 (J) embryos revealed by FDA and PI staining. Scale bars, 20 µm. (K) The frequency of FDA-positive suspensors at stages 8 and 9 in NtCYS-overexpressing line L-5. Data represent the mean ± SE from three independent experiments, with 30 embryos per stage analysed in each experiment (n = 90). (L) The frequency of suspensors containing indicated numbers of TUNEL-positive nuclei in NtCP14 RNAi line L2-8 at stages 4 to 8. Data represent the mean ± SE from three independent experiments, with 30 embryos per stage analysed in each experiment (n = 90). Asterisks indicate the basal cell.

Figure 8. The model for the regulation of suspensor PCD by the antagonistic action of NtCYS and NtCP14.