When they found an transformant of R. eutropha that produced polymer, they extracted it from the cells to analyze the monomeric units: lyophilize the cells, extract the polymer with hot chloroform, precipitate with ethanol, then analyze methyl ester derivatives with a gas chromatograph.
Then they sequenced the gene they had found, responsible for this PHB synthesis, and compared it with other known synthases, from Pseudomonas, Ralstonia, and Burkholderia species; they also looked upstream and downstream from this gene for other relevant genes, and cloned and sequenced those that they found, comparing them to other species as before.
So then to make sure the phbC gene they found actually deserved its name (and was doing what they expected), they knocked it out in Azotobacter FA8 by inserting an antibiotic resistance marker in the middle of it. First they amplified a 578-bp section of gene, inserted that product into the pGEM-T Easy vector. Then they cut this piece open with PstI, inserted a kanamycin cassette from pUC4K, and transferred this whole construct into pAT18, which can replicate in E. coli but not Azotobacter. Finally they used conjugation to transfer the plasmid from E. coli to Azotobacter, where the latter recombined with it to transfer the kan cassette into its genome, thus becoming able to resist the kanamycin used to select for transconjugants. They confirmed the PHB-lacking phenotype using the GC method described above. To make extra-sure, they made another construct in pRK404 with an intact phbC gene, transferred that to Azotobacter, and observed that it restored the PHB-producing phenotype.
Finally they wanted to see if Azotobacter FA8 could make more kinds of polymers than just PHB. There are different kinds of PHA synthases that can incorporate different monomers with different carbon-chain lengths; some prefer short chains, some longer. So they grew the bacteria on media with different carbon sources, specifically glucose and/or octanoate or hexanoate. They also tried putting genes from Pseudomonas aeruginosa in the phbC-knockout strain, which they thought might help metabolize the longer chains better. And they tried putting Azotobacter genes in R. eutropha to see what polymers they could get from there.
What They Observed
Analyzing the polymer that R. eutropha produced with the Azotobacter gene, they found it was a homopolymer composed entirely of 3-hydroxybutyrate, so, basic PHB. The gene was closely related to PHB synthase genes from other related species, as mentioned, so they named it phbC in Azotobacter. Knocking the gene out and then restoring it by complementation confirmed that it was correctly named.
There was nothing interesting downstream of phbC, but upstream they found a couple other genes that were also very similar to PHB-related genes: phbA, that encodes β-ketoacyl-CoA thiolase, and phbB, which makes acetoacetyl-CoA reductase. Together these form the phb operon, apparently in Azotobacter as well as other species. They also found some consensus regions for σ70-dependent promoters upstream of phbB.
In their experiments with polymer production, Azotobacter could only grow when glucose was present, regardless of other genes or carbon sources. And when it produced polymer at all, which was only when it had its own phbC gene, it produced only PHB. There were similar results in R. eutropha possessing Azotobacter's phbC.
What This Means
Pettinari and colleagues seem to have identified the PHA synthase-encoding gene in Azotobacter FA8 and identified the product as a short-chain specific synthase, that doesn't do much longer than four carbons per monomer. So that's good to know.
Apparently around the time this paper was in publication, someone else published A. vinelandii phbB and phbC gene sequences; comparing them to the ones in this paper, they're over 90% similar, which makes sense. So it should be possible to use knowledge of one when studying the other.