Toward a systems-level understanding of gene regulatory, protein interaction, and metabolic networks in cyanobacteria

Abstract

Cyanobacteria are essential primary producers in marine ecosystems, playing an important role in both carbon and nitrogen cycles. In the last decade, various genome sequencing and metagenomic projects have generated large amounts of genetic data for cyanobacteria. This wealth of data provides researchers with a new basis for the study of molecular adaptation, ecology and evolution of cyanobacteria, as well as for developing biotechnological applications. It also facilitates the use of multiplex techniques, i.e., expression profiling by high-throughput technologies such as microarrays, RNA-seq, and proteomics. However, exploration and analysis of these data is challenging, and often requires advanced computational methods. Also, they need to be integrated into our existing framework of knowledge to use them to draw reliable biological conclusions. Here, systems biology provides important tools. Especially, the construction and analysis of molecular networks has emerged as a powerful systems-level framework, with which to integrate such data, and to better understand biological relevant processes in these organisms. In this review, we provide an overview of the advances and experimental approaches undertaken using multiplex data from genomic, transcriptomic, proteomic, and metabolomic studies in cyanobacteria. Furthermore, we summarize currently available web-based tools dedicated to cyanobacteria, i.e., CyanoBase, CyanoEXpress, ProPortal, Cyanorak, CyanoBIKE, and CINPER. Finally, we present a case study for the freshwater model cyanobacteria, Synechocystis sp. PCC6803, to show the power of meta-analysis, and the potential to extrapolate acquired knowledge to the ecologically important marine cyanobacteria genus, Prochlorococcus.

Keywords: cyanobacteria; meta-analysis; metabolic pathways; networks; systems biology.

Figure 1

Overview of the workflow from different “omic” methods to different systems-level networks. Only technologies discussed in this review are shown. For more details, the manually-curated meta-database OMICtools (http://omictools.com/) provides tools and platforms for multi-omic data analysis. Note: 2-D PAGE, two-dimensional polyacrilamide gels; DIGE, difference gel electrophoresis; NMR, nuclear magnetic resonance; PPI, protein-protein interaction.

Figure 2

World map showing genomic records for cyanobacteria generated using an interface powered by google maps, available on the Genomes Online Database (GOLD) website (http://genomesonline.org/cgi-bin/GOLD/index.cgi). Red labels indicate the original location of a specifically sequenced strain. Labels direct the user to information on the organism, genome characteristics (i.e., GC content, size), sequencing method used, specific coordinates of the origin of the strain, as well as links to external databases (as shown for Prochlorococcus marinus NATL1A). The GOLD also describes the status of each record in tabular format.

Figure 3

Correlation network for functional categories (as defined by Falkowski, 2012) based on the expression of “core” genes in Synechocystis sp. PCC 6803 under multiple environmental conditions extracted from CyanoEXpress. Nodes represent Gene Ontology (GO) pathways, colored based on their average differential expression in the dark such that gradients of red indicate induction, while green indicate repression. Only categories with an absolute Spearman correlation value greater than 0.95 were connected by an edge.

Figure 4

Pie chart showing records of expression data for different cyanobacteria that are currently available in the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/). The bias toward Synechocystis 6803 can be clearly seen in this species breakdown of transcriptomic data. Data from April 2014.

Figure 5

Iron regulatory networks based on the differential expression upon iron limitation of genes conserved (identity larger than 20%) between Synechocystis sp. PCC6803 and Prochlorococcus MED4. The ferric uptake regulator (Fur) and a hypothetical ncRNA were set as central regulatory elements (in blue) following the well-described iron regulatory network of E. coli. Circular nodes were colored using a gradient from green to red, reflecting differential expression upon iron depletion at 72 h for Synechocystis 6803 (A) and 53 h for Prochlorococcus MED4 (B). Red edges indicate a putative repression by the regulatory element. The apparent differential regulation of cmpA/nrtC is discussed in more detail in the case study. (C) CINPER-generated network created using the keyword set to Fur, with Synechocystis 6803 as the reference organism, and Prochlorococcus MED4 as the target.