Genes cryIVA, cryIVD, and cryIVR, encoding the ‘-endotoxin from Bti, was subcloned in various combinations into a plasmid, pRL488p, carrying the Nostoc replicon pDU1, placed downstream from the strong promoter of psbA taken from Amaranthus hybridus. The recombinant plasmids were transferred into Anabaena PCC 7120 by conjugation. The resulting strains were tested for their abilities to kill larvae of the mosquito Aedes aegypti. Plasmids carrying cryIVA, with or without the other cry genes, killed 95% to 100% of the mosquito larvae, while cryIVD alone was much less efficient.
Many groups have expressed Bti toxin and related toxins in laboratory strains of cyanobacteria [Chungjatupornchai (1990) Curr Microbiol 21:283- 288; Murphy & Stevens (1992) Appl Environ Microbiol 58:1650-1655; Soltes- Rak et al (1993) Appl Environ Microbiol 59:2404-2410; Xu et al (1993) FEMS Microbiol Lett 107:247-250], but the question has remained whether these strains can persist in the field [Sangthongpitag et al (1996) Biotechnol Lett 18:175-180]. Wu and coworkers are working to move the plasmids they have tested to Anabaena siamensis, a strain originally isolated from a rice field in Thailand. The strain might be well suited to bring the toxin to the areas most affected by malaria.
The active substance was purified and found to be a glycine-rich peptide. The small peptide (molecular weight less than 12 kDa) had a UV absorption maximum at 228 nm. A 0.001% preparation of crude peptide killed 80% of a population of American boll worm (Helicoverpa armigera) after 84 h of treatment. Purified peptide gave a mortality rate half that of the crude preparation. A higher concentration (0.01%) was able to kill larvae of leaf rollers (Stylepta derogata) on cotton crops. The toxicity of the peptide to mammals has not been determined.
First, although Microcystis and Anabaena species generally dominate cyanobacterial blooms in the region, some of the incidents could be attributed to toxic Oscillatoria. In these cases, the toxin was identified as a hydrophobic microcystin that was toxic at significantly lower levels than hydrophilic microcystins. In one case, the level of the hydrophobic microcystin was only 71 mg/l, while in cases with hydrophilic microcystins, the level was typically around 1500 mg/l.
Second, the cyanobacterial source of the toxin was sometimes identified as mats on the wall of dams or cement drinking troughs, rather than buoyant scums. Clearly, the agricultural community in South Africa must now have heightened vigilance towards the appearance of toxic cyanobacteria and must not be lulled into a false sense of security by the absence of obvious blooms.
The names of most cyanobacteria are readily comprehensible. Naturally enough, most generic names describe how the organism looks. Some filamentous cyanobacteria are -thrixes ("hairlike"): Calo- ("beautiful"), Tolypo- ("wooly"), or Prochloro- ("primitive and green"). Others are "-nemas" ("threads"): Scyto- (leathery) or Plecto- ("twisted"). Some cyanobacteria are named after a father of cyanobacteriology: Fischerella (B. Fischer, 1852-1915) or Lyngbya (HC Lynbgye, 1782-1837). Bergey's Manual is a good source for such insights.
One genus stands out, however. Bergey's Manual throws up its hands when confronted with Nostoc ("origin uncertain", it says, Greek for "haven't a clue"). It is difficult to guess even from what language the name comes.
Malcolm Potts has recently proposed a solution to this dilemma [Internatl J Syst Bacteriol (1997) 47:584], tracing the origins of "Nostoc" back to the 15th century alchemist, Paracelsus. Paracelsus was a native German speaker and not at all the stuffy academic. He was impressed by the characteristic appearance of what we now call Nostoc commune. I won't tell all of Malcolm's tale, but suffice to say that Paracelsus combined common English and German to form a most graphic and human evocation of green slime.
Liu placed cDNA encoding TNF from rhesus monkey downstream from the strong psbA promoter on a plasmid capable of replicating in both E. coli and Anabaena PCC 7120. The plasmid, pDC-TNF, which expressed high levels of TNF (15%) in E. coli, was transferred into Anabaena to obtain a strain that produced a protein recognized by TNF-alpha monoclonal antibody. Expression of the protein had no measurable effect on the growth of the cyanobacterium, but transfer of energy from phycobilisomes was altered.
ZHOU Ruan-bao and ZHAO Jindong, hoping to shed some light on HetR function, have exploited antibodies raised against the protein. The hetR gene from Anabaena PCC 7120 was overexpressed in E. coli and the HetR protein purified to homogeneity. Sequencing of the N-terminus of the protein confirmed the identity of the protein and showed that the initial methionine residue was posttranslationally removed. Antibodies were raised in rabbit against purifed HetR and used for characterization of the native HetR protein in Anabaena.
Native HetR from Anabaena was compared with recombinant HetR overexpressed in E. coli. The two were found to have the same molecular mass, as judged by Western blotting, indicating that the start codon of the hetR gene assigned by Buikema and Haselkorn is correct. Although the native and recombinant proteins have approximately the same size, they differ in charge. Western blotting after isoelectrofucosing electrophoresis showed that HetR protein isolated from Anabaena starved for nitrogen exhibited an isoelectric point (pI) of approximately 3.5 while recombinant HetR exhibited a pI of 6.5. While Zhou and Zhao do not know the reason for the difference in charge, protein phosphorylation is one intriguing possibility.
Antibody against HetR was also used to study the regulation of hetR. Western blotting showed that Anabaena filaments grown in the presence of nitrate and ammonium contained detectable levels of HetR protein. Shifting the culture to a nitrogen-free medium resulted in an increase of HetR by a factor of about three. This small increase does not, however, reflect the true magnitude of induction, but rather an increase averaged over both vegetative cells and heterocysts. Hetrocysts alone contained about 20-fold more HetR than did vegetative cells grown with nitrate.
A clue as to HetR function may come from the observation that the purified protein is rapidly degraded in vitro. The degradation is blocked by phenylmethylsulfonyl fluoride (PMSF), a compound known to covalently bind to and inhibit serine proteases. PMSF was also found to react with a serine of HetR, and the sequence of amino acids around the binding site suggests that HetR is indeed a serine protease.
Overexpression of the protein offers the possibility of analyzing the structure of HetR. Recombinant HetR was found to form a homodimer in vitro upon removal of dithiothreitol from the solution. Circular dichroism spectrum taken of the protein indicated that the secondary structure of HetR expressed in E. coli contains 24% a-helices, 10% ‡-sheets, and 20% turns. Several salts were able to crystallize HetR. The crystals formed were mostly diamond shaped, and they were large enough (>0.25mm) for X-ray diffraction.
The maximal rate of light-induced movement of the alkalophyllic cyanobacterium, Spirulina platensis (Arthrospira) was observed when the medium had a pH between 10 and 12 and the sodium concentration was at least 10 mM. The pH for all further observations was set at 10.5.
If the proton gradient, Delta-Mu[H+], drives motility in Spirulina as it does in E. coli, then the proton ionophore carbonyl cyanide m-chlorophenyhydrazone (CCCP), a classical uncoupler, would be expected to block trichome movement. Brown and colleagues found that even at as high a concentration as 400 mM CCCP, Spirulina trichomes remained motile if 200 mM sodium were present. On the other hand, an 8-fold lower level of CCCP completely arrested motility in the presence of the sodium ionophore monensin. Monensin alone decreased motility only by 30%. The photosystem II inhibitor DCMU prevented CCCP-resistant movement of Spirulina trichomes that had been starved by extended preincubation in darkness.
Brown and his colleagues concluded that light-induced gliding of Spirulina is indeed driven by Delta-Mu[Na+] rather than by Delta-Mu[H+].
In these respects, the cyanobacterial gene product differs markedly from plant phytochrome, which does not fold correctly in E. coli. The resulting product was a chromoprotein, which behaved as a spectrally functional phytochrome after red/far red irradiation. A more detailed account of its properties has been published [Hughes et al (1997) Nature 386:683].
Their second step was finding a gene in Synechocystis that shows similarity to domains of phytochrome genes and to bacterial histidine kinases, including one from Calothrix [Kehoe & Grossman (1996) Science 273:1409-1412]. Knocking out this gene produced a mutant of Synechocystis that grows like the wildtype under red and far red light, slower than the wild-type under white light and, surprisingly, does not grow at all under blue light. This inability to grow under blue light could be overcome by addition of glucose to the medium. A report on this work has recently appeared [FEBS Lett (1997) 406:89-92].
The last speaker, Warwick Silvester, related some ways in which Nostoc punctiforme in association with Gunnera spp differs biochemically from free-living Nostoc. First, he showed data supporting the view that Nostoc within Gunnera do not have a functional photosystem II. Second, Nostoc leaks ammonia when provided with excess energy in the form of light (through photosystem I) . Finally, the activity of nitrogenase (as measured by acetylene reduction) is five-fold higher in associated Nostoc than in free-living isolates.
In between, Johanna Wouters and Birgitta Bergman presented a poster session describing a gene isolated from the previously mentioned subtractive cDNA library. The gene appears to encode an a-amylase and is apparently expressed during the infection process.
Cyanobacterial lichens were also duly represented. Eckhard Loos showed that the kinetics of glucose excretion by Nostoc sp. from Peltigera horizontalis decreased rapidly after isolation [Lichenologist (1996) 28:67-78]. That inspired him and his co-worker (R. Wastlhuber) to isolate a homologue to the glucose transporter (gtr) from Synechocystis, which they showed by reverse transcriptase-PCR to be expressed specifically in the lichen and in freshly isolated Nostoc.
Probably the oldest record of a cyanobacterial lichen was reported by Thomas Taylor, who found a 400 million years old fossil from the Lower Devonian Rhynie chert. This fossil shows remarkable resemblance to present day cyanolichens of the Lichinaceae family, containing unicellular Gloeocapsa-like cyanobionts.
One of the major themes in this congress was marine symbioses, and cyanobacterial associations were, on occasion, the center of attention. John Lee gave a plenary lecture on algal symbiosis in Foraminifera, reporting that the large Amphisora contains small unicellular cyanobacteria with conspicuous red pigmentation. In the poster session, John Lee and co-workers (S. Bacus and J. Morales) also reported that the giant protozoan, Marginopora vertebralis, contains in addition to dinoflagellates, two types of cyanobacteria, one unicellular and one with heterocysts (!). M. Sara concluded that the sponge Petrosia ficiformis contains both cyanobacteria and heterotrophic bacteria and that these affect the morphology and physiology of the sponge.
This congress had representatives from a broad span of disciplines, and many interesting systems were presented, even some not involving cyanobacteria (hopefully, these are being reported elsewhere). To hear more about them, try the 3rd International Congress on Symbiosis, to be held in the year 2000 in Marburg, Germany. For further information contact Hans Weber, FB Biology, Philips-University, Marburg, 35032, Germany.
Sven Janson and Johanna Wouters
Back to Cyanonews 13-1 Index
Back to Cyanosite