CYP704B1 is a long-chain fatty acid omega-hydroxylase essential for sporopollenin synthesis in pollen of Arabidopsis

Abstract

Sporopollenin is the major component of the outer pollen wall (exine). Fatty acid derivatives and phenolics are thought to be its monomeric building blocks, but the precise structure, biosynthetic route, and genetics of sporopollenin are poorly understood. Based on a phenotypic mutant screen in Arabidopsis (Arabidopsis thaliana), we identified a cytochrome P450, designated CYP704B1, as being essential for exine development. CYP704B1 is expressed in the developing anthers. Mutations in CYP704B1 result in impaired pollen walls that lack a normal exine layer and exhibit a characteristic striped surface, termed zebra phenotype. Heterologous expression of CYP704B1 in yeast cells demonstrated that it catalyzes omega-hydroxylation of long-chain fatty acids, implicating these molecules in sporopollenin synthesis. Recently, an anther-specific cytochrome P450, denoted CYP703A2, that catalyzes in-chain hydroxylation of lauric acid was also shown to be involved in sporopollenin synthesis. This shows that different classes of hydroxylated fatty acids serve as essential compounds for sporopollenin formation. The genetic relationships between CYP704B1, CYP703A2, and another exine gene, MALE STERILITY2, which encodes a fatty acyl reductase, were explored. Mutations in all three genes resulted in pollen with remarkably similar zebra phenotypes, distinct from those of other known exine mutants. The double and triple mutant combinations did not result in the appearance of novel phenotypes or enhancement of single mutant phenotypes. This implies that each of the three genes is required to provide an indispensable subset of fatty acid-derived components within the sporopollenin biosynthesis framework.

Figure 1.

zebra mutants have defective exine architecture. Confocal images of exine surface from the wild-type (A) and mutant (B–I) pollen grains that exhibit zebra phenotypes. Pollen was stained with the fluorescent dye auramine O and visualized using fluorescein isothiocyanate settings. Pollen from the following lines are shown: SAIL_1149_B03 (B), 28-3-1 (C), 37-2-3 (D), 119-1-1 (E), 131-3-1 (F), 155-4-1 (G), 174-154-2 (H), and 135-2-3 (I). All images are to the same magnification. Bar = 10 μm.

Figure 2.

One zebra complementation group maps to the At1g69500 gene. A, The exon-intron map of the At1g69500 gene. White boxes represent coding regions, and the gray box represents the untranslated region. Positions of the identified point mutations are indicated by asterisks, and the position of the T-DNA insertion is indicated by a triangle. The dashed line indicates the region that could not be PCR amplified in line 37-2-3; however, the exact nature of this mutation was not defined. B, At1g69500 and MS2 gene expression in stage 9 to 10 buds from At1g69500 and ms2 T-DNA mutants was assessed with RT-PCR. Line SAIL_1149_B03 was used for the At1g69500 mutant, and SAIL_92_C07 was used for ms2. Actin1 was used as a control. WT, Wild type.

Figure 3.

zebra mutants have severe defects in exine structure. A to D, Scanning electron micrographs of the surface structure of pollen grains from the wild type (A and C) and the At1g69500 mutant (allele SAIL_1149_B03; B and D). Samples were prepared without fixation (A and B) or with fixation (C and D). Bars = 5 μm. E to H, Transmission electron micrographs of pollen grain sections from the wild type (E) and zebra mutants At1g69500 allele SAIL_1149_B03 (F), At1g69500 allele 119-1-1 (G), and ms2 allele 135-2-3 (H). Bars = 1 μm. b, Baculae; i, intine; P, pollen grain cytoplasm; pc, pollen coat; t, tectum. I and J, zebra mutants are sensitive to acetolysis. Wild-type (I) and At1g69500 allele SAIL_1149_B03 (J) pollen grains mock treated with 50 mm Tris-HCl, pH 7.5, have similar morphology. K, Wild-type pollen grains treated with the acetolysis mixture remain intact. L, At1g69500 mutant pollen is completely lysed after the acetolysis treatment. Arrows point to the remnants of five pollen grains. Bars = 20 μm.

Figure 4.

At1g69500 is expressed in the developing anthers. A, RT-PCR analysis shows flower-specific expression of At1g69500. Different tissues were collected from flowering wild-type plants and analyzed for the presence of At1g69500 transcript. Primers were designed to exclude the amplification of genomic DNA. Actin1 was used as a control gene. B to I, The At1g69500pr∷GUS promoter fusion construct is expressed in anthers. Expression starts in buds of stage 9 (C), is most prominent at stages 9 to 11 (D–F), and fades during stage 12 (G and H). No expression is visible before stage 9 (B) and during or after late stage 12 (H and I). All images are to the same magnification. Bar = 200 μm.

Figure 5.

CYP704B1 metabolizes long-chain fatty acids present in yeast. Microsomes obtained from yeast transformed with an empty vector (control) and from yeast expressing CYP703A2, CYP704B1, or CYP703A2/CYP704B1 were incubated with lauric acid (10 nmol) in the presence of NADPH. Metabolites were analyzed by LC-MS. CYP703A2-expressing microsomes produced three in-chain monohydroxylated lauric acids (black arrows). Three novel peaks (C16:1-OH, C16:0-OH, and C18:1-OH; gray arrows) unrelated to lauric acid were produced by the CYP704B1-expressing microsomes. The elution position of the substrate (lauric acid) is shown by black triangles.

Figure 6.

CYP704B1 metabolizes long-chain fatty acids. Microsomes were incubated with 100 μm radioactive substrates (C12:0, lauric acid; C16:0, palmitic acid; C18:1, oleic acid; C18:2, linoleic acid; C18:3, linolenic acid; epox C18, 9,10-epoxystearic acid) in the absence or presence of NADPH, and product formation was monitored by TLC. The signal generated in the absence of NADPH was subtracted from the signal generated in the presence of NADPH. All values are means ± sd of experiments performed in triplicate.

Figure 7.

CYP704B1 does not hydroxylate phenolic compounds. Activities of CYP704B1-expressing microsomes toward five phenolic compounds (cinnamate [A], p-coumaroyl shikimate [C], p-coumarate [E], sinapate [F], and ferulate [G]) were measured using UPLC-MS/MS. Reactions were performed in the absence or presence of NADPH. In each panel, the top chromatogram corresponds to the predicted mass of hydroxylated product, while the bottom chromatogram corresponds to the substrate. None of the phenolic substrates was metabolized by CYP704B1. Two other CYPs (CYP73A5 and CYP98A3), known to act as hydroxylases of cinnamate and p-coumaroyl shikimate, respectively, served as positive controls in the reactions with the corresponding substrates (B and D).

Figure 7.

CYP704B1 does not hydroxylate phenolic compounds. Activities of CYP704B1-expressing microsomes toward five phenolic compounds (cinnamate [A], p-coumaroyl shikimate [C], p-coumarate [E], sinapate [F], and ferulate [G]) were measured using UPLC-MS/MS. Reactions were performed in the absence or presence of NADPH. In each panel, the top chromatogram corresponds to the predicted mass of hydroxylated product, while the bottom chromatogram corresponds to the substrate. None of the phenolic substrates was metabolized by CYP704B1. Two other CYPs (CYP73A5 and CYP98A3), known to act as hydroxylases of cinnamate and p-coumaroyl shikimate, respectively, served as positive controls in the reactions with the corresponding substrates (B and D).

Figure 7.

CYP704B1 does not hydroxylate phenolic compounds. Activities of CYP704B1-expressing microsomes toward five phenolic compounds (cinnamate [A], p-coumaroyl shikimate [C], p-coumarate [E], sinapate [F], and ferulate [G]) were measured using UPLC-MS/MS. Reactions were performed in the absence or presence of NADPH. In each panel, the top chromatogram corresponds to the predicted mass of hydroxylated product, while the bottom chromatogram corresponds to the substrate. None of the phenolic substrates was metabolized by CYP704B1. Two other CYPs (CYP73A5 and CYP98A3), known to act as hydroxylases of cinnamate and p-coumaroyl shikimate, respectively, served as positive controls in the reactions with the corresponding substrates (B and D).

Figure 8.

Pollen surfaces in the novel ms2 alleles, null alleles of cyp703a2, and the double and triple mutant combinations of cyp703a2, ms2, and cyp704b1 all show zebra phenotypes. Confocal images of auramine O-stained exine surfaces of pollen grains of the following genotypes: ms2 T-DNA insertion alleles SAIL_92_C07 (A) and SAIL_75_E01 (B), cyp704b1 ms2 double mutant (C), cyp703a2 point mutants lap4-1 (D) and 40-4-1 (E), cyp703a2 T-DNA insertion alleles SALK_119582 (F) and SLAT N56842 (G and H), cyp703a2 cyp704b1 double mutant (I), cyp703a2 ms2 double mutant (J), and cyp703a2 cyp704b1 ms2 triple mutant (K and L). In K, the cyp703a2 allele has a point mutation, lap4-1, and in L, the cyp703a2 allele is the insertion line SALK_119582. Point mutants of cyp703a2 retained a reticulate exine pattern (D and E); however, their exine was very thin, and bare exineless patches were often present. Insertion mutants of cyp703a2 sometimes had remnants of reticulate structure on portions of the grains (visible in H). Bar = 10 μm.