Phenotypic diversity and altered environmental plasticity in Arabidopsis thaliana with reduced Hsp90 levels

Abstract

The molecular chaperone HSP90 aids the maturation of a diverse but select set of metastable protein clients, many of which are key to a variety of signal transduction pathways. HSP90 function has been best investigated in animal and fungal systems, where inhibition of the chaperone has exceptionally diverse effects, ranging from reversing oncogenic transformation to preventing the acquisition of drug resistance. Inhibition of HSP90 in the model plant Arabidopsis thaliana uncovers novel morphologies dependent on normally cryptic genetic variation and increases stochastic variation inherent to developmental processes. The biochemical activity of HSP90 is strictly conserved between animals and plants. However, the substrates and pathways dependent on HSP90 in plants are poorly understood. Progress has been impeded by the necessity of reliance on light-sensitive HSP90 inhibitors due to redundancy in the A. thaliana HSP90 gene family. Here we present phenotypic and genome-wide expression analyses of A. thaliana with constitutively reduced HSP90 levels achieved by RNAi targeting. HSP90 reduction affects a variety of quantitative life-history traits, including flowering time and total seed set, increases morphological diversity, and decreases the developmental stability of repeated characters. Several morphologies are synergistically affected by HSP90 and growth temperature. Genome-wide expression analyses also suggest a central role for HSP90 in the genesis and maintenance of plastic responses. The expression results are substantiated by examination of the response of HSP90-reduced plants to attack by caterpillars of the generalist herbivore Trichoplusia ni. HSP90 reduction potentiates a more robust herbivore defense response. In sum, we propose that HSP90 exerts global effects on the environmental responsiveness of plants to many different stimuli. The comprehensive set of HSP90-reduced lines described here is a vital instrument to further examine the role of HSP90 as a central interface between organism, development, and environment.

Figure 1. Targeting HSP90 with double-strand RNA constructs.

The four cytosolic HSP90s are depicted in their genomic context. Introns are in black, exons in white. Regions cloned to create the double-strand RNA (RNAi) constructs are indicated in green (RNAi-A), blue (RNAi-B), red (RNAi-C), and yellow (RNAi-D) bars. Below each isoform the percent identity and the longest consecutive number of bases of perfect identity to the cloned region are noted for each construct.

Figure 2. Similar morphological phenotypes of seedlings with reduced HSP90 function by RNAi or pharmacological means (GDA).

(a) and (b): purple pigment accumulation; (c) and (d): organ number defect; (e) and (f): narrowly-shaped deformed true leaves; (g) and (h): twisted rosettes; (i) and (j): lobed cotyledon. RNAi plants are T₃ generation with from line RNAi-A3. Size bar 2 mm for b and g–i, 1 mm for a, c–f, and 3 mm for normal phenotype. (b) and (f) originally published in .

Figure 3. Examples of standardized seedling morphological phenotypes for quantitative assessment.

(a) no scored phenotype; (b) extra leaf, concave cotyledons, aerial organs on growth medium; (c) missing leaf; (d) narrow leaves, delayed development; (e) shoot meristemless, delayed development; (f) necrotic spots on leaves; (g) purple pigmentation in cotyledons, delayed development; (h) curled hypocotyl with root extruded from medium; (i) no hypocotyl, aerial organs on growth medium, delayed development. All seedlings are ten days old. Size bar 2 mm.

Figure 4. Quantitation of selected standard seedling phenotypes.

(a) comparison of phenotypic frequency in percent between five HSP90-RNAi lines and three control lines (control data merged). (b) similar comparison between three HSP90-TDNA lines and Col-0. Error bars represent standard error. *p<0.01, **p<0.001; χ² test.

Figure 5. Response of selected seedling phenotypes to growth temperature.

(a) narrow leaf phenotype; (b) shoot meristemless phenotype; (c) purple pigment accumulation; and (d) one of first true leaves missing. Phenotypes for two HSP90-RNAi lines and one control were measured at 17, 22, and 27°C. Error bars represent standard error.

Figure 6. Examples of altered morphologies observed in adult HSP90-reduced plants.

(a) control inflorescence; (b and c) fasciated inflorescence; (d) normal morphology with single primary inflorescence; (e) multiple simultaneous developing inflorescences. White arrows in d and e denote primary inflorescences. Scale bar 5 mm for a–c, 1 cm for d,e.

Figure 7. Developmental progression of common altered shoot meristem development phenotype.

(a–f): Seedling with RNAi-A construct exhibiting delayed development of true leaves and shoot meristem organization defects resulting in overproliferation of leaves. (g–l): Seedling with control construct with normal morphology. (a and g) 7 days after germination; (b and h) 10 days; (c and i) 14 days; (d and j) 18 days; (e and k) 21 days; (f and l) 25 days. Scale bar 1 mm for a–c, g–h, 2 mm for i, and 1.5 cm for d–f, j–l. These seedlings are of the Shahdara ecotype and are used to illustrate the phenotype; the developmental progression of affected Col-0 seedlings is similar.

Figure 8. Response to herbivore attack differs in HSP90-reduced lines.

Response to herbivory was measured by T. ni caterpillar dry weight. Least-squares means of caterpillar weight are plotted for each genotype and treatment (see Methods). Top: Basal and induced herbivore responses. For each pair, lighter colors denote basal response, darker colors indicate response after prior herbivore attack. Bottom: Herbivore response after treatment with different concentrations of Me-JA. Control in shades of green; hsp90.2-3 in shades of blue. Lighter colors indicate higher Me-JA concentration. Statistical analysis by pairwise Student's t-tests; the response of lines sharing any letter is not statistically different. Separate statistical analysis conducted for top and bottom panels.