Terry, M. I. | bioRxiv Science

bioRxiv2020

Automatic classification of constitutive and non-constitutive metabolites with gcProfileMakeR

Data analysis in non-targeted metabolomics is extremely time consuming. Genetic factors and environmental cues affect the composition and quantity of present metabolites i.e. the constitutive and non-constitutive metabolites. We developed gcProfileMakeR, an R package that uses standard output files from GC-MS for automatic data analysis using CAS numbers. gcProfileMakeR produces three outputs: a core or constitutive metabolome, a second list of compounds with high quality matches that is non-constitutive and a third set of compounds with low quality matching to MS libraries. As a proof of concept, we defined the floral scent emission of Antirrhinum majus using wild type plants, the floral identity mutants deficiens and compacta as well as RNAi lines of AmLHY. Loss of petal identity was accompanied by appearance of aldehydes typical of green leaf volatile profiles. Decreased levels of AmLHY caused a major increase in volatile complexity, and activated the synthesis of benzyl acetate, absent in WT. Furthermore, some volatiles emitted in a gated fashion in WT such as methyl 3,5-dimethoxybezoate or linalool became constitutive. Using sixteen volatiles of the constitutive profile, all genotypes were classified by Machine Learning with 0% error. gcProfileMakeR may thus help define core and pan-metabolomes. It enhances the quality of data reported in metabolomic profiles as text outputs rely on CAS numbers. This is especially important for FAIR data implementation. One sentence summarygcProfileMakeR allows the automatic annotation of the core metabolome and non-constitutive metabolites, increasing speed and accuracy of non-targeted metabolomics.

bioinformatics↗

bioRxiv2019

Molecular and transcriptional structure of the petal and leaf circadian clock in Petunia hybrida

The plant circadian clock coordinates environmental signals with internal processes. We characterized the genomic and transcriptomic structure of the Petunia hybrida W115 clock in leaves and petals. We found three levels of evolutionary differences. First, PSEUDO-RESPONSE REGULATORS PhPRR5a, PhPRR5b, PhPRR7a, PhPRR7b, and GIGANTEA PhGI1 and PhGI2, differed in gene structure including exon number and deletions including the CCT domain of the PRR family. Second, leaves showed preferential day expression while petals tended to display night expression. Under continuous dark, most genes were delayed in leaves and petals. Importantly, photoperiod sensitivity of gene expression was tissue specific as TIMING OF CAB EXPRESSION PhNTOC1 was affected in leaves but not in petals, and PhPRR5b, PhPRR7b and the ZEITLUPE ortholog CHANEL, PhCHL, were modified in petals but not leaves. Third, we identified a strong transcriptional noise at different times of the day, and high robustness at dawn in leaves and dusk in petals, coinciding with the coordination of photosynthesis and scent emission. Our results indicate multilayered evolution of the Petunia clock including gene structure, number of genes and transcription patterns. The major transcriptional reprogramming of the clock in petals, with night expression may be involved in controlling scent emission in the dark.\n\nHighlightThe petunia leaf circadian clock shows maxima during the day while petal clock does it during the night. Reaction to dark is organ specific.

plant biology↗