Phenotypic analysis of mature pollen from wild type (Ws-2 and Col-0), <i>gpt1</i>^<i>+/-</i> heterozygous, and <i>wrky2 wrky34</i> double mutant plants.

(A) Reduced viability of pollen grains from gpt1+/- plants in comparison to wild-type control (Ws-2) based on PI viability staining. Dead pollen grains show red fluorescence. (B) In vitro pollen germination of Ws-2 and gpt1+/- plants. (C) Quantitation of pollen viability of Ws-2 and gpt1+/- heterozygous mutant plants. (D) Quantitation of pollen germination rates of Ws-2 and gpt1+/- heterozygous mutant. (E) Lipid staining of pollen from gpt1+/-, wrky2 wrky34, and their respective wild-type controls. Lipid-containing pollen is stained blue by Nile blue A. (F) Quantitation of Nile blue A-stainable pollen from gpt1+/-, wrky2 wrky34, and their respective wild-type control. (G) BODIPY 505/515 staining of lipid bodies in pollen of gpt1, wrky2 wrky34, and their respective wild-type controls. White arrows indicate smaller pollen grains, possibly of gpt1 genotype. (H) Quantitation of fluorescence intensity in pollen of gpt1, wrky2 wrky34, and their respective wild-type controls. Intensity of fluorescence was quantified by ImageJ, and normalized to that in their respective wild-type controls, which was set as 100%. Only the smaller round pollen grains, which are likely of gpt1 genotype, from gpt1+/- plants were included in the quantification. In C, D, and F, at least 90 pollen grains were counted in each repeat. Error bars indicate SD (n = 3 in C, D, and F, and ≥ 10 in G). **P ≤ 0.01. Bar = 50 μm (A, B, and E) or 10 μm (G).