Generation of cell polarity in plants links endocytosis, auxin distribution and cell fate decisions

Abstract

Dynamically polarized membrane proteins define different cell boundaries and have an important role in intercellular communication-a vital feature of multicellular development. Efflux carriers for the signalling molecule auxin from the PIN family are landmarks of cell polarity in plants and have a crucial involvement in auxin distribution-dependent development including embryo patterning, organogenesis and tropisms. Polar PIN localization determines the direction of intercellular auxin flow, yet the mechanisms generating PIN polarity remain unclear. Here we identify an endocytosis-dependent mechanism of PIN polarity generation and analyse its developmental implications. Real-time PIN tracking showed that after synthesis, PINs are initially delivered to the plasma membrane in a non-polar manner and their polarity is established by subsequent endocytic recycling. Interference with PIN endocytosis either by auxin or by manipulation of the Arabidopsis Rab5 GTPase pathway prevents PIN polarization. Failure of PIN polarization transiently alters asymmetric auxin distribution during embryogenesis and increases the local auxin response in apical embryo regions. This results in ectopic expression of auxin pathway-associated root-forming master regulators in embryonic leaves and promotes homeotic transformation of leaves to roots. Our results indicate a two-step mechanism for the generation of PIN polar localization and the essential role of endocytosis in this process. It also highlights the link between endocytosis-dependent polarity of individual cells and auxin distribution-dependent cell fate establishment for multicellular patterning.

Figure 1. Endocytic recycling-based two-step mechanism generates PIN polarity

a, Targeting of newly synthesized PIN1–YFP to the plasma membrane after its complete photobleaching. Note PIN1–YFP localization (yellow arrowheads) at the plasma membrane in a non-polar manner (third panel) before becoming polar (fourth panel). b, Quantitative polarity index (ratio of polar to lateral PIN1–YFP intensity) for FRAP experiments. Data are mean and s.d.; n =15. c, Inducible overexpression of PIN1 in XVE-PIN1 epidermal cells shows PIN1 localization (yellow arrowheads) in the intracellular compartments and weak non-polar localization at the plasma membrane (first panel) after 1 h. At later time points, PIN1 shows gradually reduced intracellular signal and establishment of polar plasma membrane localization (third panel). d, A two-step mechanism for PIN polarity generation is shown. First, there is default non-polar secretion to the plasma membrane, followed by endocytic recycling establishing polarity. All are root cells. Scale bars are 5 μm.

Figure 2. Rab5-mediated endocytosis is required for PIN polarization

a, b, PIN1 (a) and PIN2 (b) polarity is defective in DN-Ara7 and atvps9a-2. Note the polar localization of PINs in wild type (WT; left panels, yellow arrowheads) and their largely non-polar localization in DN-Ara7 and atvps9a-2 (middle and right panels, yellow arrowheads). c, d, Quantitative evaluation of PIN1 (c) and PIN2 (d) polarity defects. Data are mean and s.d.; n =63 and 38 (c and d, respectively). e, Post-induction (3 h) XVE-PIN1 localization in wild type and in DN-Ara7. Note that in wild type (WT) induced PIN1 becomes polar (left panel, yellow arrowheads), whereas in DN-Ara7 seedlings it remains largely non-polar (right panel, yellow arrowheads). All are root cells. Scale bars are 5 μm.

Figure 3. Manipulation of the Rab5 pathway during embryogenesis leads to defects in PIN polarity, auxin response distribution and embryo development

a, PIN1 polarity (yellow arrowheads) defects in the heart-shaped embryos in RPS5A≫DN-Ara7 and in the embryo-lethal atvps9a-1 mutant as compared to control. Note that PIN1 is localized basally in control (left) and is largely non-polar in both RPS5A≫DN-Ara7 (middle) and atvps9a-1 (right) mutants. b, Distribution of auxin response (visualized by DR5rev::3×VENUS-N7) is altered in the RPS5A≫DN-Ara7 heart-shaped embryo. Note pronounced DR5 maxima at the root pole in control (left panel, yellow arrowhead) and RPS5A≫DN-Ara7 embryos and further strong DR5 maxima in cotyledons (right panel, yellow arrows) of RPS5A≫DN-Ara7 embryos. c, Embryonic defects from the 8-cell stage up to the heart-shaped stage in RPS5≫DN-Ara7 as compared to control. Note apical embryonic defects (bottom right, yellow arrowhead) in RPS5≫DN-Ara7. All are embryonic cells. Scale bars are 10 μm.

Figure 4. Manipulation of Rab5 pathway leads to homeotic leaf-to-root transformation

a, Phenotypic comparison of 6-day-old control and RPS5A≫DN-Ara7 seedlings. Insets show root-like structures emerging from the embryonic leaves (cotyledons) in RPS5A≫DN-Ara7 and in 35S::AtVps9a(RNAi) seedlings. Yellow arrows represent the root-like structures emerging from the embryonic leaves, whereas yellow arrowheads represent the main root. b, Two separate growing roots in RPS5A≫DN-Ara7 3-week-old seedling. One root emerges from the root pole and another from the position of cotyledon. Note that both roots have root hairs (root-specific epidermal structures highlighted by yellow arrowheads). c, d, SEM analysis of RPS5A≫DN-Ara7 and control. Stomata, a non-root cell type (red dotted circles, enlarged in insets) are absent from the root-like structure (yellow arrowhead, c). Epidermal cells of root-like structure (left panel, d), with narrow rectangular cell files, are clearly distinct from cotyledon pavement cells of control (right panel, d) and resemble main root epidermal cells of control (middle panel, d). e, In situ hybridization reveals ectopic expression of the root-specific transcription factor PLT1 and the root meristem-identity markers WOX5 and PIN3 in cotyledons of RPS5A≫DN-Ara7 heart-stage embryos. Note that expression of these markers is restricted to the root pole of control embryos. f, A schematic representation depicting enhanced auxin response maxima at the cotyledon regions resulted from non-polar PIN localization. This increased auxin response at the cotyledon regions leads to expression of auxin-induced root-forming regulators and triggers cell fate changes resulting in homeotic leaf-to-root transformation. Scale bars: 4 mm (a), 1 mm (b), 250 μm (c) and 10 μm (d).