Cloning and overproduction of gibberellin 3-oxidase in hybrid aspen trees. Effects on gibberellin homeostasis and development

Abstract

To broaden our understanding of gibberellin (GA) biosynthesis and the mechanism whereby GA homeostasis is maintained in plants, we have investigated the degree to which the enzyme GA 3-oxidase (GA3ox) limits the formation of bioactive GAs in elongating shoots of hybrid aspen (Populus tremula x Populus tremuloides). We describe the cloning of a hybrid aspen GA3ox and its functional characterization, which confirmed that it has 3beta-hydroxylation activity and more efficiently converts GA9 to GA4 than GA20 to GA1. To complement previous studies, in which transgenic GA 20-oxidase (GA20ox) overexpressers were found to produce 20-fold higher bioactive GA levels and subsequently grew faster than wild-type plants, we overexpressed an Arabidopsis GA3ox in hybrid aspen. The generated GA3ox overexpresser lines had increased 3beta-hydroxylation activity but exhibited no major changes in morphology. The nearly unaltered growth pattern was associated with relatively small changes in GA1 and GA4 levels, although tissue-dependent differences were observed. The absence of increases in bioactive GA levels did not appear to be due to feedback or feed-forward regulation of dioxygenase transcripts, according to semiquantitative reverse transcription polymerase chain reaction analysis of PttGA20ox1, PttGA3ox1, and two putative PttGA2ox genes. We conclude that 20-oxidation is the limiting step, rather than 3beta-hydroxylation, in the formation of GA1 and GA4 in elongating shoots of hybrid aspen, and that ectopic GA3ox expression alone cannot increase the flux toward bioactive GAs. Finally, several lines of evidence now suggest that GA4 has a more pivotal role in the tree hybrid aspen than previously believed.

Figure 1.

Competitive assay for GA 3β-hydroxylation with cell lysates from recombinant E. coli expressing PttGA3ox1 in pGEX-4T-2. Cell lysates were incubated with cofactors and a mixture of equal amounts of the substrates GA₉ and GA₂₀ at varying concentrations for 1 h at 20°C, and the corresponding formation of GA₄ and GA₁ was monitored by GC/MS-SRM using deuterated GAs as internal standards.

Figure 2.

Northern analysis of AtGA3ox1 expression in young expanding leaf tissue of WT and 35S-AtGA3ox1 transgenic lines (4, 5, 12, 13, and 17). Twenty micrograms of total RNA was loaded per lane and hybridized under stringent conditions at 65°C to a full-length AtGA3ox1 probe. A ubiquitin-like EST, PttUBQ2, was used as loading control.

Figure 3.

Relatively minor changes observed in height growth increment of 35S-AtGA3ox1 transgenic plants. Actively growing hybrid aspen plants cultivated in the greenhouse under LD conditions were monitored during a 45-d period. The depicted growth pattern is based on averages per line, n = 9 except for line 17 (n = 5). Student's t test analysis of the significance of differences between each line and WT based on the last data point per genotype showed that line 5 (P < 0.01) and line 17 (P < 0.05) were shorter and line 12 taller (P < 0.01) than WT.

Figure 4.

GA content of apical, internode, and leaf tissue in 35S-AtGA3ox1 transgenic plants and WT. Tissue from nine individuals was pooled per genotype and tissue type. GAs from 200 mg fresh weight of pooled tissue were purified and analyzed by GC/MS-SRM using ²H₂-GAs as internal standards. Data presented are the means of three technical replicates of pooled sample ±sd. GA₈ levels in line 5 were not measured. Boxed GAs depict bioactive GAs.

Figure 5.

GA dioxygenase expression levels, as determined by semiquantitative RT-PCR in various tissues of WT and the 35S-AtGA3ox lines 5 and 12. Different gene-specific and intron-spanning primer pairs were used together with a primer pair for the internal standard 18S. One cDNA source was used per sample that originated from pooled plant tissues (nine plants per genotype). A, The expression of PttGA20ox1 and PttGA3ox1 in various tissues of WT and line 5 and 12. B, Quantification results of two GA genes were normalized to the internal standard 18S. The average of 3 to 4 independent RT-PCRs were calculated and results plotted; bars depict se. C, The PttGA2ox1 expression was analyzed first in WT and then compared to line 5 and 12 in apical tissue. D, PttGA2ox2 transcript levels were determined in WT and compared to line 5 and 12 in young expanding internodes. Int, internodes; NC, negative control—no cDNA added to the PCR mixture; M, DNA size marker.

Figure 6.

A, PCA-score plot from the analysis of GAs in five WT and five line 12 plants. All variables were log-transformed, centered, and scaled to unit variance. Each point in the plot corresponds to a separate sample type, i.e. internode tissue of different developmental stages. Internode A, circles; internode B, triangles; and internode C, boxes. Line 12 is represented by white symbols, and WT is depicted in black. A clear separation can be observed between all samples, suggesting developmental regulation of GA levels. B, PLS-loading plots from the first loading vector (w) in the analysis of GA levels and internode length from internodes at different developmental stages of five WT plants. The height of the bars shows the relative correlation between GAs and internode length. If the confidence interval (calculated with jackknifing; Efron, 1986; Martens and Martens, 2000) does not include 0, a variable is considered significant. Bioactive GAs are boxed, and the plot shows that GA₄ levels, and not GA₁ levels, are correlated with internode length.

Figure 7.

Sensitivity of P. tremula seedlings to GA₄ and GA₁. Hypocotyl length of 2.5-d-old seedlings are presented as an average of 40 seedlings ±se, except for the 10-μm treatment (30 ± se). Letters indicate statistically significant differences, according to Student's t test, in sensitivity to GA₄ as compared to GA₁; a, P < 0.001; b, P < 0.0001; and c, P < 0.01. At GA₄ concentrations greater than 100 nm and GA₁ concentrations higher than 1 μm the seedlings grew significantly taller than controls, P < 0.00001. The experiment was repeated once with P. tremuloides seedlings with very similar results.