Gene expression profiling in response to ultraviolet radiation in maize genotypes with varying flavonoid content

Abstract

Microarray hybridization was used to assess acclimation responses to four UV regimes by near isogenic maize (Zea mays) lines varying in flavonoid content. We found that 355 of the 2,500 cDNAs tested were regulated by UV radiation in at least one genotype. Among these, 232 transcripts are assigned putative functions, whereas 123 encode unknown proteins. UV-B increased expression of stress response and ribosomal protein genes, whereas photosynthesis-associated genes were down-regulated; lines lacking UV-absorbing pigments had more dramatic responses than did lines with these pigments, confirming the shielding role of these compounds. Sunlight filtered to remove UV-B or UV-B plus UV-A resulted in significant expression changes in many genes not previously associated with UV responses. Some pathways regulated by UV radiation are shared with defense, salt, and oxidative stresses; however, UV-B radiation can activate additional pathways not shared with other stresses.

Figure 1.

A, Absorption spectra of extracted flavonoids in acidic methanol. Extract from B, Pl plants, ▴ in red; extract from Mutator plants, ▪ in blue; extract from bz2 plants, ♦ in green. The b, pl green plants were used as the blank, to normalize spectra among the samples. B, UV transmittance spectra of the filters and comparison with the solar spectrum. UV solar spectrum, • in green; CA filter (full sunlight), ▴ in red; PE filter (no UV-B), ▪ in blue; LE filter (no UV), ♦ in yellow. C, Irradiance of the supplementary UV-B lamps and comparison with the UV-B solar spectrum and with the UV-B spectrum of the UV-B lamps covered with PE. The supplementary lamps were used in a greenhouse, which has about 20% of the summer solar irradiation. Solar spectrum, ♦ in blue; UV-B lamps, ▴ in red; UV-B lamps covered with PE (no UV-B), • in green. D, Plant height after the different UV treatments. The values are the average of plants from the different plots after 3 weeks under the CA filter (six plots, green), PE filter (six plots, blue), or LE filter (four plots, red). Error bars are ses. There is a statistically significant difference between the values of plants under full UV-B compared with values under no UV-B or no UV-A+B for all of the lines, independently of the line used (two-way factorial ANOVA test, significance level P = 0.05; PT (effect of the treatment) < 0.05, PG (effect of the genotype) > 0.50; PT × G > 0.25).

Figure 2.

Scatter plot comparing ratios of signal values from two replicates on microarray hybridizations with mRNA from leaves of b, pl plants under full sunlight and under no UV-B labeled with Cy3-dUTP and Cy5-dUTP. Data from images of dye-swapping experiments were plotted as the mean intensity after normalization of expressed sequence tags (ESTs) spotted in triplicate. Additional examples are provided in Supplementary Figure 2.

Figure 3.

Summary of the number of ESTs responsive to UV-B exclusion or supplementation.

Figure 4.

Venn diagrams of comparisons between UV-B-responsive genes in lines with different levels of flavonoids and anthocyanins. A, Venn diagrams of comparisons between UV-B-responsive genes in b, pl (green) and B, Pl (purple) plants in UV-B exclusion experiments for 3 weeks in the field (Exclusion, SL UV-B/no SL UV-B); 1 d after UV-B restoration in plants grown for 3 weeks without UV-B in the field (Restoration, 1 d SL UV-B after 3 weeks without UV-B/SL UV-B); and after 10-fold UV-B supplementation for 8 h in the greenhouse (Supplementation, 10-fold UV-B for 8 h in the greenhouse/no UV-B in the greenhouse). B, Venn diagrams of comparisons between UV-B-responsive genes in bz2 (beige) and Mutator (pink) plants in UV-B exclusion experiments for 3 weeks in the field (Exclusion, SL UV-B/no SL UV-B); and after 10-fold UV-B supplementation for 8 h in the greenhouse (Supplementation, 10-fold UV-B for 8 h in the greenhouse/no UV-B in the greenhouse). Up-regulated genes are in red; down-regulated genes are in green. Sets of genes were selected using the criteria described in “Materials and Methods.”

Figure 5.

Cluster analysis of transcripts from b, pl (green) and B, Pl (purple) plants after 3 weeks under UV-B exclusion in the field, 1 d after UV-B restoration in plants grown for 3 weeks without UV-B in the field, and after 10-fold UV-B supplementation for 8 h in the greenhouse. Clustering was performed according to Eisen et al. (1998). The color saturation reflects the magnitude of the log2 expression ratio for each transcript. Transcripts are grouped into patterns A to E accordingto their expression profiles. b, pl (in green), Low flavonoid plants; B, Pl (in purple), high flavonoid and anthocyanin plants. Exclusion, Plants grown under full UV-B sunlight/plants grown without UV-B in the field; Restoration, plants 1 d after removing the PE filters/plants grown under full sunlight; Supplementation, plants with supplementary UV-B for 8 h in the greenhouse/plants grown without UV-B in the greenhouse. Last column, b, pl/B, Pl under full UV-B sunlight. Red, Higher transcript levels than the reference; green, lower transcript levels than the reference. Gray, Flagged ESTs that had signals similar to the background in some condition and were eliminated during the analysis. The color log2 scale is provided at the bottom of the Figure. For clustering analysis, the same data from microarray experiments shown in Venn diagrams in Figure 4A were used.

Figure 6.

RNA gel-blot analysis to confirm microarray data. Lanes contained 10 μg of total RNA extracted from plants after different UV-B treatments. Several identical gels were prepared and blotted. Each blot was hybridized with ³²P-labeled elongation factor 1α (A), ribosomal protein QM (B), MRP33 (C), or MRP47 (D) probes. b, pl, Low flavonoid plants; B, Pl, high flavonoid and anthocyanin plants. Exclusion, Plants grown without UV-B in the field; Restoration, plants 1 d after removing the PE filters; Sunlight, plants grown under full sunlight. Figure 4E shows an ethidium bromide-stained gel as a check for equal loading. The log2 ratio was calculated as for microarray experiments by quantification of hybridization signals and ethidium bromide-stained bands using Kodak ds 1D Digital Science, as described in “Materials and Methods,” and is provided at the bottom of each blot, using as a reference RNA from plants that were grown under natural levels of UV-B (listed as 0 change).

Figure 7.

Cluster analysis of transcripts from Mutator (Mu, pink); bz2 (bz2, beige); and B, Pl (B, purple) plants in sunlight. RNA from b, pl green plants in sunlight was used as the reference for microarray hybridization. Clustering was performed according to Eisen et al. (1998). The color saturation reflects the magnitude of the log2 expression ratio for each transcript. Red color means higher transcript levels than the reference, whereas green means lower transcript levels than the reference. The log2 scale of color values is provided at the bottom of the figure.