Genome-wide fitness assessment during diurnal growth reveals an expanded role of the cyanobacterial circadian clock protein KaiA

Abstract

The recurrent pattern of light and darkness generated by Earth's axial rotation has profoundly influenced the evolution of organisms, selecting for both biological mechanisms that respond acutely to environmental changes and circadian clocks that program physiology in anticipation of daily variations. The necessity to integrate environmental responsiveness and circadian programming is exemplified in photosynthetic organisms such as cyanobacteria, which depend on light-driven photochemical processes. The cyanobacterium Synechococcus elongatus PCC 7942 is an excellent model system for dissecting these entwined mechanisms. Its core circadian oscillator, consisting of three proteins, KaiA, KaiB, and KaiC, transmits time-of-day signals to clock-output proteins, which reciprocally regulate global transcription. Research performed under constant light facilitates analysis of intrinsic cycles separately from direct environmental responses but does not provide insight into how these regulatory systems are integrated during light-dark cycles. Thus, we sought to identify genes that are specifically necessary in a day-night environment. We screened a dense bar-coded transposon library in both continuous light and daily cycling conditions and compared the fitness consequences of loss of each nonessential gene in the genome. Although the clock itself is not essential for viability in light-dark cycles, the most detrimental mutations revealed by the screen were those that disrupt KaiA. The screen broadened our understanding of light-dark survival in photosynthetic organisms, identified unforeseen clock-protein interaction dynamics, and reinforced the role of the clock as a negative regulator of a nighttime metabolic program that is essential for S. elongatus to survive in the dark.

Keywords: circadian clock; cyanobacteria; diurnal physiology; photosynthesis; transposon sequencing.

Fig. 1.

Using RB-TnSeq to assess genetic contributions to fitness in LDC. A dense pooled mutant library containing unique known bar-code identification sequences linked to each individual gene insertion was used to estimate the fitness of each loss-of-function mutant grown under alternating light–dark conditions. The library was grown under continuous light and sampled to determine the baseline abundance of each strain in the population before transfer to photobioreactors which were then either exposed to continuous light or alternating 12-h light–12-h dark regimes. Bar-code quantification of the library population in both conditions was normalized to the baseline and compared against each other to estimate the fitness consequences of loss of function of each gene specific to growth in LDC.

Fig. 2.

LDC Rb-TnSeq screen results breakdown. (A) Of the 2,723 genes in the genome, 1,872 genes were analyzed in the library population. Of those, data from mutants with loss of function of 1,420 genes fell below our false discovery threshold. Of the remaining 452 genes, 220 had moderate to strong fitness effects. (B) A volcano plot highlighting genes of interest. The red-shaded region indicates an estimated fitness decrease in LDC and the green-shaded region indicates a fitness increase. (C) Functional category composition of genes that gave strong fitness scores. Categories based on Clusters of Orthologous Groups (COG) classification or Gene Ontology ID. Green corresponds to strong fitness increases; red corresponds to strong fitness decreases.

Fig. 3.

Connection between essential nighttime metabolism and RpaA regulation. (A) A metabolic map showing the reactions (red) that are controlled by genes that are essential in LDC. Reactions that are nonessential (solid arrows) or essential (dashed arrows) in CLC are are also indicated. Stars mark NADPH-generating reactions. *Gene was not revealed in the screen but is known to be required in LDC; ^#minor LDC decrease measured, likely due to fitness contribution in CLC which affects calculation. (B) LDC-sensitive mutant population enriched for genes known to be positively regulated by RpaA.

Fig. 4.

kaiA mutant growth and cikA-based intervention in LDC. (A) Dilution spot-plate growth of strains in CLC and LDC. A representation of the genotype of each strain is portrayed in SI Appendix, Fig. S1. (B) cikA mutant fitness increase in competition with WT in LDC and CLC. (C) Growth of liquid cultures in replicate photobioreactors in LDC. Vertical shaded areas represent dark conditions. Data points are mean (SD); n = 2.

Fig. 5.

Bioluminescence levels from kaiBC reporter in mutant backgrounds. The intensity of P_kaiBC::luc output signal correlates with LDC sensitivity. Rhythms were measured in CLC conditions driven by the kaiBC promoter after exposure to 72 h of entraining LDC that includes a low-light intensity that is permissive for the mutants. Shaded areas indicate the SEM of four biological replicates.

Fig. 6.

RpaA status correlates with growth success in LDC. (A) RpaA phosphorylation at dusk (after 12 h of light). Box plot above the representative Phos-tag immunoblot shows values for replicate densitometry analysis (n ≥ 4) of P-RpaA/total RpaA. (B) Immunoblot shows levels of KaiC from cells taken at dusk. (C) Phos-tag immunoblot shows RpaA phosphorylation across the light–dark transitions with samples taken just before the lights turning on (dawn) or off (dusk). (D) H₂DCFDA fluorescence over a 12-h dark period during an LDC, indicating total cellular ROS in WT, kaiA^del, and kaiA^delcikA. The shaded area indicates the period of darkness following a 12-h light period. P-RpaA denotes phosphorylated protein.

Fig. 7.

CikA-activating complex formation measured by fluorescence anisotropy. (A) Kinetics of 6-IAF–labeled KaiB binding to WT KaiC (Upper Right) or phosphomimetic KaiCs (KaiC-AE, or KaiC-EA, Left) in the absence (red) or presence (blue) of CikA, with 0.5 mM ATP + 0.5 mM ADP. Controls are shown for KaiB-only (black) and for KaiB and CikA without KaiC (green, Lower Right). (B) Binding of KaiB to WT KaiC (Right) or KaiC-AE (Left) in the absence and presence of CikA with 1.0 mM ATP, using the same color codes as for A.

Fig. 8.

Model for LDC sensitivity due to circadian clock disfunction. (A) Conventionally, the role of KaiA is primarily to bind to the A-loops of the CII domain of KaiC, stimulating its autophosphorylation and consequently increasing the activity of the kinase SasA; this activity in turn results in accumulation of phosphorylated RpaA throughout the day. P-RpaA initiates the expression of circadian-controlled nighttime class 1 genes at dusk. Yellow and red dots on KaiC represent phosphorylation at S431 and T432, respectively. A red dot also represents phosphorylation on RpaA. The rare fold switch of KaiB is represented by different orange shapes. The major conformational change of KaiA bound to the CII A-loops versus the KaiB ring and the exposure of the KaiB binding site on the CI ring of KaiC are also depicted. (B) Without KaiA, KaiC phosphorylation is suppressed and, thus, SasA kinase activity is diminished, reducing the pools of SasA-mediated P-RpaA. A small fraction of the KaiC pool that is phosphorylated even in the absence of KaiA can bind KaiB. CikA has access to these KaiC-bound KaiB proteins that are usually occupied by KaiA and becomes hyperactive; excessive CikA phosphatase activity extinguishes intracellular pools of P-RpaA and eliminates the cell’s ability to express genes needed for survival in LDC. Modified from ref. , with permission from AAAS. The dotted line in the B graph represents the WT levels of P-RpaA from A.