Harnessing the Circadian Clock to Boost Useful Bioproduction in Cyanobacteria
By: Carol A. Rouzer, VICB Communications
Published: November 20, 2013
Overexpression of a circadian clock protein increases expression of a majority of S. elongatus genes, a finding that may be exploited for industrial protein synthesis.
Due to their capacity for photosynthesis, cyanobacteria have become the focus of increasing interest as a potential source of renewable energy as well as biosynthetic products of medical or industrial value. Cyanobacteria have also attracted the attention of biochemists as an interesting model system for the study of circadian rhythms. The species Synechococcus elongatus (Figure 1), for example, is striking in that nearly all of its genes are expressed in an oscillating rhythm with the 24 hour periodicity characteristic of circadian regulation. Now, VICB member Carl Johnson and his laboratory harness their knowledge accumulated from years of study of the S. elongatus circadian clock to propose a way to increase the economic productivity of these abundant microorganisms [Y. Xu, et al. (2013) Curr. Biol., published online November 7, doi:10.1016/j.cub.2013.10.011].
Figure 1. Culture of cyanobacteria. Image reproduced from Wikimedia Commons, under the Creative Commons Attribution-Share Alike 30 Unported license.
Three proteins, KaiA, KaiB, and KaiC are primarily responsible for controlling the circadian clock in S. elongatus. KaiC, which possesses both autokinase and autophosphatase activities, undergoes a rhythmic cycle of hyperphosphorylation followed by dephosphorylation, aided by interactions with KaiA and KaiB. In solution, KaiC exists as a hexamer that, when hypophosphorylated, interacts with KaiA. This interaction stimulates KaiC’s autokinase activity, leading to phosphorylation first at threonine-432, then at serine-431. Gradually, KaiC becomes hyperphosphorylated and begins to interact with KaiB. Bound KaiB then associates with KaiA and prevents it from stimulating the autokinase activity of KaiC. As a result, the autophosphatase activity predominates, and KaiC is dephosphorylated at threonine-432 followed by serine-431. As KaiC becomes hypophosphorylated, it releases KaiA and KaiB, setting the stage for the beginning of another cycle (Figure 2). This robust series of oscillating reactions occurs in vitro with roughly 24 hour periodicity, requiring only the presence of the three proteins and ATP. In vivo, interaction of the oscillating complex with other proteins signals gene expression changes in the cell that result in the circadian regulation of nearly all of S. elongatus’s biochemical functions.
Figure 2. The KaiC protein (pink ovals) comprises a C1 and a C2 domain with C-terminal “tentacles”. It exists as a hexamer in solution. At the beginning of the cycling reaction, KaiA (blue octagon) repeatedly and rapidly interacts with KaiC’s C-terminal tentacles, initiating and propagating the phosphorylation phase. Phosphorylation occurs first at threonine-432 (yellow stars) and then at serine-431 (orange stars). When KaiC becomes hyperphosphorylated, it first binds KaiB (green hexagon) stably. Then, the KaiB•KaiC complex binds KaiA, sequestering it from further interaction with KaiC’s tentacles. At that point, KaiC initiates dephosphorylation, first at threonine-432, and then at serine-431. When KaiC is hypophosphorylated, it releases KaiB and KaiA, thereby launching a new cycle. Image adapted from C. H. Johnson, et al., (2011) Annu. Rev. Biophys., 40, 167.
S. elongatus’s proteins may be divided into two major classes. Class I proteins, also known as subjective dusk proteins, are encoded by genes that are turned on at dawn, increase expression throughout the day, and peak at dusk. In contrast, class II (subjective dawn) proteins are encoded by genes that begin expression at dusk and reach their peak at dawn. Prior work had shown that overexpression of the kaiC gene resulted in suppression of expression of class I genes with corresponding reductions in the levels of class I proteins. These changes were accompanied by increased expression of class II genes and levels of their encoded proteins. Together the results suggested that modulation of the clock proteins could result in global changes in gene expression, leading the Johnson lab to investigate ways to increase, rather than inhibit global expression of class I genes.
The Johnson lab hypothesized that hyperphosphorylation of KaiC would be associated with increased expression of class I genes. They tested this hypothesis by overexpressing kaiA, anticipating that elevated levels of KaiA protein would promote KaiC autokinase activity. Microarray analysis showed that kaiA overexpression resulted in upregulation of the expression of about 20% and downregulation of about 12% of S. elongatus genes. Importantly, most of the upregulated genes were class I, while the downregulated genes were class II. Thus, kaiA overexpression had the opposite effect on global gene expression than kaiC overexpression.
The investigators further explored the effects of kaiA overexpression using a panel of luminescent reporter genes driven by a diverse range of bacterial promoters. These included class I and class II gene promoters from S. elongatus, in addition to heterologous reporters from Escherichia coli. Results showed that kaiA overexpression increased the expression of reporters driven by the clock gene promoters (kaiAp, and kaiBCp). Consistently, KaiC and KaiB protein levels were increased, as was the level of KaiC phosphorylation. In contrast, overexpression of kaiC resulted in decreased KaiC protein phosphorylation, presumably by flooding the pool with unphosphorylated protein monomers.
The luminescent reporter studies also showed that kaiA overexpression increased the expression of genes driven by two E. coli promoters. In the absence of kaiA overexpression, both of these reporter genes were expressed rhythmically in S. elongatus, in synch with endogenous class I genes. KaiA overexpression increased the level of expression of both reporter genes and abolished the rhythmic pattern in both cases.
Together, the results suggested that kaiA overexpression locked S. elongatus into a state comparable to the “dusk” phase of the circadian cycle, when class I genes are highly expressed. The investigators noted that, despite the major disruption of normal physiology, the bacteria demonstrated no reduction in growth under these conditions. This led the Johnson group to hypothesize that kaiA overexpression could be used to promote the expression of genes of medical or commercial value. Support for this hypothesis was found in the 3-fold increased levels of the [NiFe] hydrogenase present in kaiA overexpressing cells. The [NiFe] hydrogenase uses solar energy to produce hydrogen gas from water. If cyanobacteria could be induced to express high levels of this enzyme, they would be a valuable source of renewable energy in the form of hydrogen for use in fuel cells.
Although increased expression of the endogenous [NiFe] hydrogenase was encouraging, this enzyme may not be the best choice for industrial levels of hydrogen production, due to its sensitivity to oxygen damage. For this reason, bioengineers have focused on a [NiFe] hydrogenase from Alteromonas macleodii that is particularly oxygen damage-resistant. Prior attempts to express this complex enzyme in S. elongatus produced the RC41 strain, which exhibits low levels of expression and activity of the enzyme. These discouraging results are likely due to a requirement for a set of accessory proteins that facilitate assembly of the enzyme complex. The Johnson group showed that kaiA overexpression in the RC41 strain doubled hydrogen production by the bacteria. However, hydrogenase activity remained lower than was achievable with the native enzyme. Thus, kaiA overexpression can be used to increase hydrogen production by the RC41 strain, though further work will be required to make this system economically viable.
The Johnson lab further explored the utility of kaiA overexpression by evaluating its effect on expression of the human proinsulin protein driven by an E. coli promoter. Consistent with their other findings, kaiA overexpression increased the levels of the proinsulin protein produced by S. elongatus by up to three-fold, although the investigators were surprised to find maximal production under dark rather than light conditions.
Together the data revealed that kaiA and kaiC overexpression exert a “yin-yang” regulation of gene expression, with kaiA overexpression leading to activation of class I (dusk) genes, and kaiC overexpression promoting activation of class II (dawn) genes (Figure 3). The findings provide valuable insight into the fundamental biochemistry of circadian clock regulation in S. elongatus. In addition, they provide the foundation for the future practical exploitation of kaiA overexpression to increase the production of proteins that have industrial or medicinal value.
Figure 3. Mechanism of gene expression regulation by kaiA and kaiC overexpression. Increased levels of KaiA protein under conditions of kaiA overexpression stimulate the autokinase activity of KaiC, leading to KaiC hyperphosphorylation. In contrast, kaiC overexpression results in production of large quantities of KaiC monomer, which upsets the balance of KaiC in favor of hypophosphorylation. KaiC hyperphosphorylation promotes expression of dusk (class I) genes, whereas KaiC hypophosphorylation promotes expression of dawn (class II) genes. Thus, the two conditions work in a “yin-yang” fashion, to produce complementary effects on gene expression.