Transgenerational changes in Arabidopsis smRNA expression and genomic methylation following exposure to CeO2 nanoparticles
On this page:
Crops and wild plants are being exposed to increasing amounts of engineered cerium oxide nanoparticles (CeO2-NPs) through soil, water and air. Recent studies showed that exposure to CeO2-NPs can alter the development and transcriptome profiles of plants. In addition, 1st generation exposures resulted in changes of physiology and nutrient profiles of 2nd generation plants. To further identify transgenerational phenological and molecular impacts from exposure, experimental groups of Arabidopsis thaliana plants were grown for three generations with 15 mL of 500 mg/L CeO2-NP +/- treatments occurring in the 1st and 2nd generations (C1C2 = control both generations; T1C2 = treated 1st generation; C1T2 = treated 2nd generation; T1T2 = treated both generations). No treatments were applied during the 3rd generation. Leaves from 28-day-old 2nd and 3rd generation plants were sampled (N = 5) for small RNA (smRNA) sequencing and reduced representation bisulfite sequencing (RRBS) respectively. In both cases, Illumina reads were assembled using the Arabidopsis GCF_000001735.4_TAIR10.1 genome as the template. Differentially expressed genes (DEGs) for smRNAs were identified by having ≥ 2-fold change, ≥ 10 signal: noise ratio, 95% confidence level, ≤ 0.0981 corrected P-value (T1C2 vs. C1C2), ≤ 0.05 corrected P-value (C1T2 vs. C1C2 and T1T2 vs. C1C2). During the 2nd generation, T1C2 plants had significantly fewer rosette leaves than C1C2 and C1T2s at day 30. Also, T1C2 plants had significantly smaller rosette diameters than C1T2s at all time points. DEGs were predominantly tRNA-derived small RNAs (tsRNA) that were cleaved within the anti-codon loop, which are capable of regulating translation. Altered expression of tsRNAs has been documented in other organisms in response to stressors and disease. For T1C2 plants there was significant up-regulation of 5′ tRH-Asp-GTCs, whereas C1T2 had up-regulation of 5′ tRH-Gly-GCCs. Interestingly, for T1T2 plants there was a switch to down-regulation of 5′ tRH-Gly-GCCs. Methods for identification of regulatory targets of these tsRNAs are in development.
By day-28 in the 3rd generation, significantly more T1C2 descendent plants had bolted than C1T2s but not more than C1C2s or T1T2s (P ≤ 0.1). 124,208 100-CpG probes were used to quantitate differential methylation at genes (gene bodies ± 2000 bp) across the plant genomes. Differential methylation was based on the intersection of probes with Chi-square corrected P ≤ 0.05, logistic regression corrected P ≤ 0.05 and EdgeR corrected P ≤ 0.05. The C1C2 cluster was most clearly resolved from C1T2 in both principle components and T-distributed stochastic neighbor embedding analyses. Notably, 84% of the differentially methylated genes were in the chloroplast genome, while only 3% of all 100-CpG probes were generated there. The most wide-spread enrichment of gene ontologies across treatments were those related to ATP synthesis, electron transport and photosystem II. The most frequently impacted biochemical pathway was also ATP synthesis. Overall, the results from both experiments point to
development of epigenetic memory of the particular CeO2-NP exposures that occurred in prior generations. The small RNA expression was more sensitive to exposures in just the 1st or 2nd generations (T1C2 and C1T2). By contrast, 2nd generation exposures (C1T2 and T1T2) had the largest effect on genomic DNA methylation of 3rd generation plants. Results from both experiments suggest that exposure during the 1st generation attenuated the responses to a follow-on exposure in the second generation. Additional research is needed to identify connections between these molecular changes and plant growth and development.