Cyanosite webserver:

Summerfield et al, (2008); Summerfield et al, (2011a); and Summerfield et al, (2011b)

A. Anaerobic Growth

Research: Tina Summerfield, with Jorg Toepel and Sowmya Nagarajan

It has become clear that the evolution of oxygenic photosynthesis and of cyanobacteria took place primarily in an anaerobic environment. The sequencing of genomes of numerous photosynthetic bacteria has led to the proposal that the first phototrophs were procyanobacteria (anoxygenic ancestors of the extant cyanobacteria). Since these cyanobacterial ancestors were responsible for adding oxygen to the environment, the external environment was still low in oxygen until around 2.4 billion years ago (Bya). In addition, the level of O2 in the atmosphere has varied during this time period to a high of nearly 30% at some times, but often closer to 10%, especially during the past 400 million years (3). Thus, it would seem likely that cyanobacteria had developed a regulatory system capable of inducing important genes under low oxygen and/or anaerobic conditions. Despite the importance of understanding this situation, very little information is available for the way in which cyanobacteria respond to low oxygen conditions. Our recent work, outlined in the 3 papers attached to this section, represented the first solid evidence in this area and show that a series of important gene clusters were induced in a series of cyanobacteria, both diazotrophic and non-diazotrophic.

Before oxygenic photosynthesis was developed in cyanobacteria, the atmosphere was much more reducing in nature. Indeed, there was a glut of reducing equivalents on earth's surface. During this period, electrons were made available from copious quantities of donors such as H2, H2S and CH4. The electrons could be donated easily, because of a relatively small energy differential, to acceptors such as CO2 and, to a lesser extent, SO4. Because a relatively large energy was needed to oxidize water, the biggest electron donor pool, H2O, remained inaccessible until the advent of oxygenic photosynthesis. Although the exact date for the development of oxygenic photosynthesis and the cyanobacteria is somewhat elusive, the dates from different types of evidence are reaching some congruence. It is very likely that the evolution of cyanobacteria occurred prior to 2.4 Bya because their metabolism is required, at least in part, to explain the appearance and the rise of environmental oxygen at that time. The most recent evidence indicates the period for the rise of atmospheric O2 of between approximately 2.4 and 2.3 Bya and that this was followed by the onset of an extreme ice age. Indeed, some have argued the case for a late origin of photosystem II (PSII) evolution and suggested that oxygenic photosynthesis could have evolved close in geological time to the Makganyene Snowball Earth Event and suggest a causal link between these two important events. Based on this analysis, higher levels of oxygen began to pollute the atmosphere approximately 2.2 Bya. Although our work does not directly impinge on such dates, we are comfortable with the situation in which atmospheric O2 increased rather substantially soon after the development of the water splitting function.

We have recently studied the impact of low-O2 growth on global transcription in the unicellular, nondiazotrophic cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis). We used our Synechocystis microarray system to compare gene expression between aerobically grown and these low-O2 incubated cells. These experiments led to some surprising findings, not the least of which was the induction of psbA1, the third copy of the genes encoding the main Photosystem II reaction center protein. In addition, an alternative Rieske iron-sulfur protein (petC2) was located in the same gene cluster, suggesting that changes in the cytochrome b6f complex might also occur during low-O2 growth. We discovered related changes in Cyanothece sp. ATCC 51142 and Anabaena sp. PCC 7120, unicellular and heterocystous diazotrophs, respectively. In these experiments, we grew cultures in a 6 L bioreactor in which we could bubble 99.9% N2/0.1% CO2 for specific periods of time. After growing cells aerobically until late log phase, we then bubble this gas mixture to produce low-O2 conditions for 1, 2 and 6 hours (O2 levels of ~0.001 normal). We used our Synechocystis microarray system to compare gene expression between aerobically grown and these low-O2 cells. Although most gene sets showed a decrease in transcription during the low-O2 phase, five gene clusters were strongly upregulated under these conditions, including slr1181-slr1185, a cluster that contains psbA1 and petC2 (encoding alternative isoforms of the PSII protein, D1, and the Rieske iron-sulfur center protein, respectively). In addition, the entire bidirectional hydrogenase cluster (the hox genes) was upregulated as were: the sll0217-sll0221 cluster which included genes encoding flavoproteins Flv2 and Flv4; sll1875 and sll1876 that encode proteins involved in pigment biosynthesis; and an important gene cluster involved with regulation (hik31 and rre34). This was a most surprising and exciting result and we first concentrated on an understanding of the large increase in psbA1transcript levels. This gene had been considered cryptic, in that it was normally transcribed at very low levels and was not induced under any other condition, although many attempts have been made. We first wanted to determine if this transcription was of functional importance and to determine if the D1 protein was made and inserted into the photosynthetic membranes. This we did quite successfully as described in our recent publication in Biochemistry. We found that the protein is made and inserted in quantities large enough to provide virtually 100% of the oxygen evolution found in the aerobically grown Synechocystis cells

We then went on to show that similar induction events occurred during low O2 incubation in at least two other diazotrophic cyanobacteria. These included the unicell, Cyanothece ATCC 51142, and the filamentous heterocystous, Anabaena sp. PCC 7120. Both demonstrated a similar induction of psbA1, but with somewhat different kinetics. In Cyanothece, there was a sharp increase at 1 hour, a decline from that level at 2 hours and a return to the basal level by 6 hours. In Anabaena, there was a slightly slower increase than found in Synechocystis, but transcript levels remained high at 6 hours. In addition, Cyanothece also had an alternative Rieske Fe-S protein in the same cluster and this also was strongly upregulated under low-O2 conditions. On the other hand, there was no Rieske Fe-S center associated with the alternative psbA gene in Anabaena and none of the Rieske Fe-S center proteins were upregulated under these conditions. Thus, Synechocystis and Cyanothece showed many similarities, including the upregulation of similar gene clusters to those mentioned above for Synechocystis.

We have investigated the response of the cyanobacterium Synechocystis sp. PCC 6803 during growth at very low O2 concentration (bubbled with 99.9% N2/0.1% CO2). Significant transcriptional changes upon low O2 growth included up-regulation of an operon that includes the two-component regulatory histidine kinase, Hik31. This regulatory cluster is of particular interest, since there are virtually identical copies on both the chromosome and on plasmid pSYSX. We used a knockout mutant in the chromosomal copy of hik31 and studied differential transcription during the aerobic-anaerobic transition in this ΔHik31 strain and the wild type. We observed two distinct responses to this transition, one Hik31 dependent, the other Hik31 independent. The Hik31 independent responses included the psbA1 induction and genes involved in pigment biosynthesis. In addition, there were changes in a number of genes that may be in involved in assembling or stabilizing PS II, and the hox operon and the LexA-like protein (Sll1626) were up-regulated during low O2 growth. This family of responses mostly focused on PS II and overall redox control. There was also a large set of genes that responded differently in the absence of the chromosomal Hik31. In the vast majority of these cases, Hik31 functioned as a repressor and transcription was enhanced when Hik31 was deleted. Genes in this category included both core and peripheral proteins for PS I and PS II, the main phycobilisome proteins, chaperones, the ATP synthase cluster and virtually all of the ribosomal proteins. These findings, coupled with the fact that Δhik31 grew somewhat better than the wild-type under low O2 conditions, suggested that Hik31 helped to regulate growth and overall cellular homeostasis. We detected changes in the transcription of other regulatory genes that may compensate for the loss of Hik31. We conclude that Hik31 regulates an important series of genes that relate to energy production and growth and that helps to determine how Synechocystis responds to changes in O2 tension.

Back to Sherman Lab Homepage