[CyanoNews home page] [Contact CyanoNews]

International Symposium on Photosynthetic Prokaryotes Meeting Report
Control over Heterocyst Differentiation
by Jeff Elhai (U. Richmond)

I believe Noel Carr once posed the question, if there is a gradient of nitrogenous compounds emanating from heterocysts, then why don't vegetative cells distant from heterocysts grow more slowly than those close to heterocysts? Enrique Flores (U. Sevilla) proposed an answer to this question, starting from the premise that filaments of heterocystous strains are composed of connected cells with a common periplasm (there is a single continuous outer membrane). If nitrogenous compounds pass from heterocysts to vegetative cells through successive cytoplasms then, given the capacity of cyanobacteria to store amino acids, a gradient of nitrogen seems inescapable. Enrique supposed instead that amino acids are released from heterocysts into the periplasm, where they bind to amino acid soluble periplasmic transport proteins before they have a chance to diffuse away.

It follows from this scheme that any amino acid that serves as the medium of exchange must be able to support the growth of filaments... from the outside. With this in mind, Enrique's lab has systematically tested all 20 amino acids for their ability to support growth of Anabaena PCC 7120. Only a few amino acids were up to the task, most notably: arginine, glutamine, proline, and aspartate. Interestingly, two of these -- arginine and aspartate -- comprise the amino acid subunits of cyanophycin. The idea of a gradient of nitrogenous compounds emanating from heterocysts might be replaced by a cafeteria in reverse: the food passes by in the periplasm and each cell grabs what it can until it is satiated. There's plenty of food for all while it lasts, but the last cells in line go hungry.

Jim Golden (Texas A&M U.) surprised many with news of a gene, patS (previously called hetS), that suppresses heterocyst formation when present in multiple copy or overexpressed. More amazing yet, the gene encodes a protein only 17 amino acids long, but the open reading frame and specific amino acid sequence are essential for the effect. A patS knockout mutant sprouts heterocysts constitutively, despite the presence of a nitrogen source, and in its absence, filaments show LOTS of heterocysts, many contiguous.

Inspired by pentapeptide peptide signals by Bacillus subtilis [Lazazzera et al (1997) Cell 89:917] and the preponderance of mutations of patS in the five C-terminal amino acids, Jim's lab examined the effect of adding to the growth medium a synthetic pentapeptide based on the final five amino acids of PatS. Remarkably, this peptide completely inhibits heterocyst differentiation, an effect that disappears when the the peptide is hydrolyzed or its sequence changed. Summing up, PatS has many (but not all) of the characteristics one would expect of the diffusible inhibitory signal long postulated to be responsible for patterned heterocyst differentiation.

Paula Duggan (U. Leeds) presented her work showing found that ethionine at a level of 1 µM blocks heterocyst differentiation while having no effect on macromolecular synthesis. Ethionine, an analog of methionine, inhibits the production of S-adenosylmethionine, a substrate of many methylation reactions, suggesting a role for methylation in the regulation of differentiation. The inhibition of differentiation occurs prior to induction of hetR, which is the first gene known to be induced in differentiating cells. Interestingly, this is the same stage at which mutants defective in the histone-like protein HU fail [Khudyakov & Wolk (1996) J Bacteriol 178:3572]. HU is important in the initiation of DNA synthesis and (in E. coli) initiation is regulated by the methylation state of the DNA. Whether this conceptual connection is pertinent to Anabaena remains to be demonstrated.

Francisco Florencio (U. Sevilla) showed Westerns that indicated a high level of isocitrate dehydrogenase (IDH) in heterocysts but virtual absence of GOGAT. This means that there is a large capacity to make alpha-ketoglutarate (a-KG) in heterocysts but no obvious place for it to go, except out of the cell. Perhaps (as suggested to me by Sam Beale) there is an energetic reason why vegetative cells would make glutamate for export to heterocysts while heterocysts specialize in a-KG synthesis -- after all, the reductant required for glutamate synthesis is a valuable commodity in heterocysts, available only by import, while cyclic phosphorylation keep the ATP required by IDH high. On the other hand, the presumed high level of a-KG in heterocysts (and developing heterocysts?) may serve an informational role, telling the cell that it is starving for nitrogen despite the abundance of amino acids. The role of a-KG in regulating PII phosphorylation and dephosphorylation in Synechococcus PCC 7942 elucidated by Angelica Irmler and Karl Forchhammer fits well with this idea (see also Meeting Report on PII protein).

Protein phosphorylation may have another connection with heterocyst differentiation. Cheng-Cai Zhang (U. Strasbourg) reported cloning of a gene (prpA) encoding a serine/threonine phosphatase and a closely linked gene (pknE) encoding a serine/threonine protein kinase from Anabaena sp. PCC 7120. Two mutants of Anabaena, which lack either the prpA gene or the pknE gene, grow normally on ammonium- or nitrate-containing medium, but have severely impaired growth under nitrogen-fixing conditions. This Ser/Thr phosphatase, which from the sequence is classified as type PP1/PP2A/PP2B, seems to be different from the PII-P phosphatase of Synechocystis, which is a PP2C-type phosphatase.