Transcriptional changes suggest a major involvement of Gibberellins in Trifolium pratense regrowth after mowing

Red clover (Trifolium pratense) is used worldwide as a fodder plant due its high nutritional value. In response to mowing, red clover exhibits specific morphological traits to compensate the loss of biomass. The morphological reaction is well described, but knowledge of the underlying molecular mechanisms are still lacking. Here we characterize the molecular genetic response to mowing of red clover by using comparative transcriptomics in greenhouse conditions and agriculturally used field. The analysis of mown and control plants revealed candidate genes possibly regulating crucial steps of the genetic network governing the regrowth reaction. In addition, multiple identified gibberellic acid (GA) related genes suggest a major role for GA in establishing the regrowth morphology of red clover. Mown red clover plants showing this regrowth morphology were partially “rescued” by exogenous GA application, demonstrating the influence of GA during regrowth. Our findings provide insights into the physiological and genetic processes of mowing red clover, to serve as a base for red clover yield improvement.


37
Facing today's challenges such as an increased demand on food production in an era of global 38 climate change together with the aim to solve these problems in an environmental friendly and 39 sustainable way requires improvement of forage crops like T. pratense [14,15]. T. pratense breeding 40 aims to offer genotypes with improved key agronomic traits (dry matter yield, high quality, 41 resistance to diseases and abiotic/biotic stress, persistency, [16]), while improving its regrowth version 3.0, http://plntfdb.bio.uni-potsdam.de/v3.0/) protein database with an e-value cutoff of 1e-140 20. The files contain the functional annotation description of all transcripts e-Appendix (Table S11).   (Table S6).  We were then interested to identify developmental processes in greater detail that are required for 242 the regrowth process. Thus, the results of the DEG analysis were restructured such that the DEG 243 were grouped in 16 descriptive classes by database and literature mining (Table S7 and Table S8).

244
Those classes describe major functional groups and serve to identify the potential role of a gene.      locations: more genes related to biotic stress processes and metabolism were upregulated in the 268 unmown locations (Fig. 1 B-D).

281
The largest group of differentially expressed genes is the one related to biotic stress with up to 38% 282 differentially expressed genes in one location (field b, Fig. 1

291
To find similarly regulated genes between the treatments and/or locations, a Venn diagram was 292 generated to compare the number of shared significantly DEG within the "mown" samples and the 293 "not mown" samples ( Fig. 1 E-F, Table S9). Within the "mown" samples we detected no overlap 294 between the groups with the exception of four genes that are differentially expressed and 295 upregulated in "mown" condition and are shared between the two field transcriptomes (FbM and 296 FaM (Fig. 1 E). Within the "not mown" samples also four genes are shared between the field 297 transcriptomes (FbNM and FaNM)) and one is shared between the field b and the greenhouse ( Fig. 1 298 F). No genes are shared between all three samples, neither in the "mown" treatment, nor in the "not 299 mown" treatment. The genes that were shared between the transcriptomes belong to the main 300 classes "growth", "phytohormone", "general cell functions", "biotic stress", "development" and 301 "transcription" (Table S9).

302
Two of the genes could not be annotated. The annotated genes include for example genes 303 tdn_60472 (shared between FaM/FbM, class: phytohormone), that was found to be the homolog of 304 the A. thaliana locus AT1G75750, describing a GA-responsive GASA1 protein homolog. Another A. DEGs) (Fig. 1 G). While ABA and SA are mainly involved in response to biotic and abiotic stresses, and 322 AUX is known to play a major role in growth and development, we identified GA as a novel candidate 323 phytohormone for regrowth response.

324
To learn more about the role of GA in the regrowth response, we identified 32 GA-related genes out 325 of 151 within the transcriptomes of the greenhouse and the field grown plants, matching our 326 selection criteria (TPM <5, involved in GA biosynthesis, signaling, GA responsive genes or catabolism, 327 displaying certain expression patterns Fig. 3 A) and classified them according to their function in the 328 GA biosynthesis and signaling processes (Table S13). Ranges of expression strength were calculated 329 and color coded to compare expression patterns (Fig. 3 A).  the regrowth process led to significant and specific changes in morphology (Fig. 3 B, C). Previous 365 work suggested that regrowing plants produce smaller and rounder leaflets with shorter petioles 366 than uncut plants [25]. Number of leaves, shoots and inflorescences, leaf area and the roundness of 367 leaflets were measured (Fig. 3 B, C, Suppl. Fig. 3).