Dawn illumination prepares desert cyanobacteria for dehydration.

Desert biological soil crusts (BSC), among the harshest environments on Earth, are formed by the adhesion of soil particles to polysaccharides excreted mainly by filamentous cyanobacteria (see [1] and references therein). These species are the main primary producers in this habitat where they cope with various stressors including frequent hydration-dehydration cycles. Water is mainly provided as early-morning dew, followed by dehydration with rising temperatures and declining relative humidity. Earlier studies focused on community structure and cyanobacterial activities in various BSCs [1,2]. They identified genes present in dehydration-tolerant, but not -sensitive cyanobacteria [3], and suggested that abiotic conditions during natural dehydration (Figure 1A) are critical for the recovery upon rewetting. Inability of Leptolyngbya ohadii, which is abundant in the BSC examined here, to recover after rapid desiccation (Figure 1B) [4] suggested that the cells must prepare themselves toward forthcoming dehydration, but the nature of the signal involved was unknown. We show here that the rising dawn illumination, perceived by photo-sensors, serves as the signal inciting BSC-inhabiting cyanobacteria to prepare for forthcoming dehydration. L. ohadii filaments were exposed to simulated natural conditions from the morning of October 14th 2009, using our environmental chamber that enables accurate reproduction of BSC environment [4] (Supplemental Figure S1A). Samples were withdrawn at specific time points (Figure 1A), followed by RNA extraction and global transcript profiling (accession PRJNA391854). Four hours of dehydration led to up-regulation of 567 genes and down-regulation of 1597 (by more than 2-fold). Since BSC-inhabiting organisms have not been used as genetic models, the functions of 3258 (43.5% of the 7487 L. ohadii genes [3]) are unknown. Nevertheless, a pronounced rise in transcript levels of genes involved in carbon metabolism, transport, osmolyte production, energy dissipation and other cellular activities was observed. On the other hand, a declining transcript abundance for genes involved in light harvesting, photosynthetic metabolism, protein biosynthesis, cell division and other pathways was detected. The analysis unraveled clear distinctions between early- and late-responding genes. Supplemental Table S1 lists the 40 strongest differentially expressed genes verified by RT-qPCR and used in further analyses.