**4. Conclusions**

The present study is the first large-scale gene expression investigation on dormancy termination process in wild cardoon. Transcriptome patterns associated with the imbibition at constant temperature include upregulation of ABA biosynthesis genes, ABA-responsive genes, as well as other genes previously related to physiological dormancy and inhibition of germination. While expression patterns stimulated at alternating temperatures comprised ethylene and ROS signaling and metabolism together with ABA degradation and cell wall loosening. Physiological assays support molecular data showing that ethylene is necessary for dormancy termination at alternating temperatures, even if its presence does not imply the completion of germination. Similarly, ROS is needed for dormancy termination since its depletion hampers this process, but ROS donors cannot overcome dormancy completely. These findings sugges<sup>t</sup> an important role of both ethylene and ROS in dormancy termination at alternating temperatures, most probably as a fine-tuned mechanism for environmental sensing. This can be a very useful system for effectively achieving dormancy termination once environmental conditions are suitable for germination in highly disturbed habitat in which this plant lives. Moreover, our results may have applications in naturalization efforts using wild cardoon for the naturalization of highly disturbed habitats impacted by human activity (e.g., sowing seeds at the correct environmental temperature regimes).

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2223-7747/9/9/1225/s1, Table S1. qRT-PCR primers designed on selected transcripts identified within the transcriptome assembly and used for RNA-seq data validation. Table S2. Expression levels, reported as log2-transformed TPM value, of 4737 DEGs identified in this study. Table S3. Complete list of correlated transcripts present in the co-expression network analysis. Figure S1. In silico functional annotation of identified transcripts. Gene Ontology (GO) terms are reported for each GO category. Figure S2. Enrichment analysis of seed transcriptome annotation compared to the *C. cardunculus* transcriptome including several phenological stages. Figure S3. Amount of transcription factor families identified in this study. Figure S4. qRT-PCR analysis for RNA-Seq data validation. A: Relative expression of wild cardoon homologous transcripts associated with regulation of ethylene, ROS and ABA. The relative expression ratio is expressed as the fold increase relative to dry achenes. The error bars represent the standard error of the mean of three biological replicates. Letters indicate significantly different values according to ANOVA (*p*-value ≤ 0.05). B: Correlation of gene expression results obtained from real-time PCR analysis and RNA-Seq (TPM) for 6 selected genes across samples. The correlation of determination (R2) was 0.64. Figure S5. Hierarchical cluster analysis of differentially expressed genes in dry achenes, imbibed achenes at a constant temperature (15 ◦C) and imbibed achenes at alternating temperatures (20/10 ◦C). The color scale represents the log2-transformed TPM value. Figure S6. Enrichment analysis of the 764 highly correlated DEGs. Figure S7. Expression heatmap of the 764 highly correlated DEGs. DS: dry achenes; IDS: imbibed achenes at constant temperature (15 ◦C); INDS: imbibed achenes at alternating temperatures (20/10 ◦C). The color scale represents the log2-transformed TPM value. Figure S8. The layout of all the correlation-based co-expression networks obtained using of the 764 highly correlated DEGs, in which are functional GO categories.

**Author Contributions:** G.D.P., H.R.H. and S.A.R. conceived and designed the study; G.D.P. and H.R.H. collected the samples; G.D.P. and H.R.H. conducted the laboratory experiments; A.D.P. and G.D.P. conducted the bioinformatic analyses; G.D.P., H.R.H. and A.D.P. wrote the manuscript. S.A.R. contributed to the experimental design and writing of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financed with funding provided by Catania section of the Istituto per i Sistemi Agricoli e Forestali del Mediterraneo (CNR-ISAFOM) and by (i) LomasCyT IV (FCA-025), Universidad Nacional de Lomas de Zamora, Llavallol, Argentina; (ii) PICT 2015-3087 from ANPCyT, Argentina. Bioinformatic analyses were supported by St. Petersburg University (project ID #51555422).

**Acknowledgments:** We thank Pietro Calderaro for his technical assistance in the field operations and Elisa Miraglia for her contribution in bioinformatic analyses. We acknowledge the support of ELIXIR-IT, CINECA for the provision of computational resources service.

**Conflicts of Interest:** The authors declare no conflict of interest.
