Assessing the remarkable morphological diversity and transcriptomic basis of leaf shape in Ipomoea batatas (sweetpotato)

Leaf shape is a spectacularly diverse trait that influences various aspects of plant physiology, and is even correlated with crop yield and quality in multiple species. However, only a few genetic dissections of leaf shape have been accomplished at a species-wide level. Here, we perform an initial characterization of leaf shape variation in Ipomoea batatas, the sweetpotato, at multiple scales of analysis. We use a transcriptomic survey to identify gene expression changes associated with two commonly studied leaf shape traits--circularity and aspect ratio using 19 individuals (accession) of sweetpotato. We comprehensively describe the remarkable morphological diversity in leaf shape in sweetpotato, and identify 147 differentially regulated genes associated with circularity and aspect ratio, providing an initial set of hypotheses regarding the genetic basis of leaf shape in this species. To obtain an initial set of candidate genes responsible for leaf shape traits (circularity and aspect ratio), total mRNA was extracted from 19 individuals (unique accessions) of sweetpotato. These plants were grown in the greenhouse to reduce gene expression variation due to environmental noise. Accessions displaying very high and low lobedness, and likewise accessions exhibiting high/low aspect ratio were chosen. Eleven individuals were used to assess variation in circularity (six entire (PI566613, PI585054, PI585108, PI592993, PI595868, PI612701), five lobed (PI573318, PI585061, PI585067, PI599387, PI634379)) and 9 individuals were chosen to assess variation in aspect ratio (five high AR (PI208886, PI344123, PI566617, PI566638, PI585061), four low AR (PI318848, PI531122, PI564154, PI599392)); one accession (PI585061) was common for circularity and AR. RNA was extracted from three to five leaves that were in P3-P5 stage of growth, from multiple branches of each individual accession, and combined replicate leaves per individual to increase the coverage of the transcriptome.