Multiple transcription factors mediate acclimation of Chlamydomonas to light stress

Abstract

Light as a substrate for photosynthesis may be a boon or a bane. To thrive, photosynthetic organisms must constantly respond to changing light and CO2 conditions by balancing energy harvest and consumption in a highly dynamic way. Two major safeguard measures of photoacclimation, that is photoprotection and carbon concentrating mechanism, underlie tight transcriptional control, leading to expression changes under high light and limited CO2 with different dynamics for both systems. Here, by using a consensus gene regulatory network inferred by employing a compendium of 1,869 RNA-seq datasets, we identified and validatedin vivoeight candidate transcription factors (TFs) that contribute to photoacclimation inChlamydomonas reinhardtii. Target gene analyses indicate that the TFs act individually in associated pathways but also influence each other in expression, and function as network parts with partial redundancy with respect to photoprotection. The analyses unveil that stress responses inChlamydomonasare mediated by a complex, interconnected network of TFs rather than a hierarchical system where multiple regulators can influence each other and target gene expression and thereby mitigate the effects of loss.

Comparative analysis of gene regulatory networks identifies conserved regulators in seed plants

“…The gene regulatory network of A. thaliana and C. reinhardii were both validated by mutant and overexpressor analyses in previous studies (Halpape et al, 2023; Wulf et al, 2023). Therefore, they act as a good indicator if the network inference of the other networks was similarly successful.…”

“…For C. reinhardtii , the enrichment analyses with submodules of photosynthesis identified no candidate regulators for the CBBc and PR (Figure 3B). When mutants of TFs predicted to be relevant for photoprotection and for photosynthesis were analyzed, a role for a subset of TFs in regulating CBBc and PR was shown (Wulf et al, 2023). The use of a significance cut-off for candidate identification (p < 1e-10) may contribute to the problem.…”

“…Gene Ontology (GO) terms were transferred based on the best blastp (Altschul et al, 1990) hit from A. thaliana to the other species with a blastp (e < 1e-5) to exclude annotation artefacts introduced by different GO annotations. Genes with a weight greater than 0.005 were considered as target of the transcription factor (Arend et al, 2023; Halpape et al, 2023; Wulf et al, 2023). For each target list of the transcription factors a GO term enrichment was performed with topGO (v2.52.0) (Alexa and Rahnenführer, 2023) using the classic fisher method for the ontology biological process with a significance cut-off at p-value < 1e-10 (Table S5).…”

“…Network inference with the random forest decision tree-based algorithm GENIE3 (Huynh-Thu et al, 2010) was used to infer a gene regulatory network (GRN) in wheat (Ramírez-González et al, 2018) and analyze polyploid evolution (Almeida-Silva and Van de Peer, 2023). Recently, GENIE3 was used for identification and validation of regulators involved in photosynthesis in A. thaliana (Halpape et al, 2023) and in the regulation of photoprotection and the carbon concentrating mechanism in Chlamydomonas reinhardtii (Arend et al, 2023; Wulf et al, 2023) making it a promising candidate to compare gene regulatory networks across species. By performing GRN inference based on expression data comparability of individual samples is not necessary, because transcription factors are examined in relation to all other genes.…”

“…The split between BEP and PACMAD can be roughly dated at 55 million years ago (Christin et al, 2014). By including previously published and validated gene regulatory networks of A. thaliana (Halpape et al, 2023) and C. reinhardii (Wulf et al, 2023) the analysis is extended to dicots and algae to include more distant species at 150 million years between eudicots and monocots and 950 million years between the chlorophyta and the streptophyta (Magallón et al, 2013; Hedges et al, 2004). We hypothesized that the comparison of networks will identify conserved and divergent regulatory events.…”