Spatial visualization of differential metabolites in SCs by DESI/PI-MSI
We next wondered if the three polyacetylenes indeed highly accumulatedin situ in the red SCs. We made efforts on JAR samples since they have two distinct SC types, of which the cortex SCs have red secretory cavities. We attempted to apply DESI-MSI to spatially visualize the major differential metabolites in frozen sections of JAR. However, detection of those non-polar compounds was not successful, likely due to low ionization efficiency and ion suppression. We next succeed in performing DESI/PI-MSI (that involves in a two-step ionization process), which showed less polarity discrimination and had a better sensitivity on the differential metabolites (Figure S5). With DESI/PI-MSI, we detected compounds corresponding to the three polyacetylenes out of the multivariate data analysis (Figures 3I). Spatial visualization of m/z 183.0807 or 287.1279 revealed that these molecules were highly correlated with the coloured SCs in JAR (Figures 4A-C). We also detected compounds with m/z 231.1392 and 233.1530, corresponding to atractylenolide I and atractylenolide II, respectively (Figures 4D-E). These sesquiterpene compounds were near-equally detected in SCs of all tissue types, in line with the multivariate statistical analysis that they are not differential metabolites related to red SCs. Notably, sesquiterpene atractylone (m/z 217.1594) is also highly correlated with red cortex SCs in JAR (Figure 4F), it however was not identified as differential metabolite between red and non-red SCs among natural accessions. Altogether, our results strongly indicate that the highly accumulated atractylodin, 1Z-atractylodin & 3E,5E,11E-tridecatriene-7,9-diyne-1,2-diacetate are likely the causal compounds underlying the cinnabar spots in ARs.