159 - Effect of chelating agents on hydrogenase in Azotobacter chroococcum: Evidence that nickel is required for hydrogenase synthesis

This study used chelating (metal-binding) compounds to study the cofactor of hydrogenase in Azotobacter chroococcum.

What They Saw
All the chelators they added (NTA, EDTA, etc.) decreased the hydrogenase activity in batch cultures, though to different extents. NTA was much stronger than EDTA. The effect was not inhibition of already-formed enzyme (since adding chelators to resting cells or extracts didn't affect activity), so it must be from preventing formation of additional enzyme.

They tried adding trace metal salts along with the chelators to see if pure excess of whatever was missing could restore activity. Copper, zinc, and manganese didn't really do anything. Cobalt helped if it was added in fairly large amounts, but the most helpful was nickel. Adding extra iron boosted this effect even more.

Monitoring nickel uptake by adding radioactive nickel, they saw that cyanide completely wiped out uptake (possibly by binding the nickel), but juglone and 2,4-dinitrophenol enhanced it (despite inhibiting respiration). Sodium azide didn't really affect either. The chelators they tested earlier generally seemed to inhibit nickel uptake too, generally in the same proportions as they had inhibited hydrogenase activity.

What This Means
Chelators seem to inhibit hydrogenase, but rather than acting on the enzyme directly, it seems to be by inhibiting its synthesis, and even this mechanism seems to be by inhibiting nickel uptake in most cases, rather than something more direct. Nickel is important for synthesizing the enzyme; it's a part of its essential cofactor. It seems like cobalt might be able to substitute for nickel somewhat though. I wonder if palladium would work too, since it has similar orbital arrangements. But apparently cobalt doesn't help in the absence of chelators or contaminating trace metals, so maybe it only helped here because it distracted the chelators away from nickel (so to speak).

Reference:

Partridge, C. D. & Yates, M. G. Effect of chelating agents on hydrogenase in Azotobacter chroococcum: Evidence that nickel is required for hydrogenase synthesis. Biochem. J. 204, 339–344 (1982).

071 - Two open reading frames (ORFs) identified near the hydrogenase structural genes in Azotobacter vinelandii, the first ORF may encode for a polypeptide similar to rubredoxins

This study looked at the genetics of the two open reading frames (ORFs) near the hydrogenase structural genes (hoxZM).

What They Saw
They sequenced the ORFs and compared to known genes. hoxZ seemed to have most homology with genes encoding proteins called rubredoxin from other species. These are typically small proteins that play roles in electron transport, which makes sense. And that's all.

Reference:

Chen, J. & Mortenson, L. Two open reading frames (ORFs) identified near the hydrogenase structural genes in Azotobacter vinelandii, the first ORF may encode for a polypeptide similar to rubredoxins. Biochim Biophys Acta 1131, 122–124 (1992).

070 - The hoxZ gene of the Azotobacter vinelandii hydrogenase operon is required for activation of hydrogenase

Here they wanted to look into the hoxZ gene more closely. Previous studies suggested that the product might be involved in electron transport for the hydrogenase.

What They Saw
They grew Azotobacter vinelandii DJ (an easy-to-transform strain) and knocked out hoxZ and hoxKG by transformation and screening for hydrogen production.

Comparing the hoxZ mutant to DJ (positive control) and the hoxKG mutant (negative control), they observed an intermediate rate of hydrogen oxidation, so there seemed to still be some activity. DJ consumed nearly all the hydrogen, and hoxKG consumed very little (the graph showed a decrease but it was apparently because gas leaked out of the vial, so it's a good thing they had good controls!).

Then they tried measuring short-term hydrogen oxidation with different electron acceptors: oxygen or methylene blue. DJ quickly oxidized all the hydrogen while reducing oxygen or methylene blue, as expected. Both mutants didn't show activity with either acceptor at first, despite the difference in the previous assay. But then they added sodium dithionite (a powerful reducer of oxygen) and more methylene blue, and the hoxZ mutant showed up to 80% of the activity of DJ. As far as I can tell, the hoxKG didn't show the same effect when they gave it the same treatment, but they don't say that explicitly. But it seems like the hydrogenase needs to be activated somehow, as by dithionite.

These results were confirmed by observing methylene blue color change too; DJ quickly started oxidizing hydrogen, but the hoxZ mutant did too after a longer lag period.

When they isolated membrane-bound hydrogenase from cells (still embedded in membranes), even DJ needed activation with dithionite. hoxKG mutants had no activity in any case, of course. But hoxZ mutant had more activity in the soluble supernatant portion than DJ did, at least when membranes were isolated aerobically; it seemed like lack of hoxZ led to more soluble enzyme. But it had low activity in general so this conclusion was uncertain. Though membrane-bound activity in general was higher when isolated anaerobically, and they didn't measure soluble activity in that case for some reason. So HoxZ may help stabilize hydrogenase in the presence of oxygen.

The increased presence of detached hydrogenase in the mutant was not confirmed by Western blot, so it seems like an artifact.

What This Means
HoxZ seems to have a role in shuttling electrons between hydrogenase and oxygen, though there may be other components involved in this path. It's possible that when HoxZ is missing, another acceptor can take the electrons, but isn't as good at it.

It also may be involved in activating the enzyme (which requires removing oxygen and providing reduction); somehow hydrogen is not enough for this. And it may help stabilize the hydrogenase to keep oxygen from inactivating it, maybe also using its role as electron transporter.

Reference:

Sayavedra-Soto, L. A. & Arp, D. J. The hoxZ gene of the Azotobacter vinelandii hydrogenase operon is required for activation of hydrogenase. J. Bacteriol. 174, 5295–5301 (1992).

067 - Analysis of a gene region required for dihydrogen oxidation in Azotobacter vinelandii

This study looks more closely into the genes discovered in previous studies (518,050,051,052) to be required for hydrogenase in Azotobacter vinelandii.

What They Saw
They knocked out different genes (or the whole operon) in the hyp operons of A. vinelandii CA by inserting resistance + lacZ cassettes, then tested these strains for hydrogen oxidation and expression of the genes (via beta-galactosidase activity). The medium they used had a lot more trace elements than typical Burk medium, including nickel.

They found that when lacZ was inserted in the same direction as the gene, they saw expression in all cases, whether fixing nitrogen or not, but if it was inserted in the opposite direction, they didn't. So apparently the genes are expressed to some extent even when not fixing nitrogen.

When they measured hydrogen oxidation, there was about 6x more when fixing nitrogen though, in the wild-type, and none in the mutants. The growth rates of the mutants were similar to the wild-type though, or so they claim without reporting any details.

Finally they tried growing the strains with extra added nickel, because the hydrogenase is a nickel-containing enzyme. This didn't have much effect on the wild-type, but the hypB mutant actually showed some hydrogenase activity in nitrogen-fixing conditions with the extra nickel, and activity was higher with more nickel added.

What This Means
It seems like extra added nickel can substitute for the lack of HypB, so maybe the enzyme has a role in nickel cofactor processing somehow. This makes sense considering the multiple histidines it contains. It's interesting that nickel didn't help activity in non-fixing conditions; maybe there are multiple hydrogenases, active in different conditions, and hypB is required for all but nickel doesn't help some of them.

Reference:

Chen, J. C., Mortenson, L. E. & Seefeldt, L. C. Analysis of a gene region required for dihydrogen oxidation in Azotobacter vinelandii. Curr. Microbiol. 30, 351–355 (1995).

056 - Properties of Hydrogenase from Azotobacter vinelandii

Another study tries to purify the uptake hydrogenase from Azotobacter vinelandii.

What They Saw
They were unable to get pure, soluble enzyme; they couldn't separate it from the membrane, so it remained in insoluble particles.

They observed that hydrogenase activity of these particles didn't decrease after 18 days of exposure to air, unlike particles from other species. They did lose activity within a few weeks though.

They also observed that the enzyme reduces cytochrome C, possibly as part of its electron transport role. They found that carbon monoxide and cyanide inhibit the enzyme. Azide does not, at least not at pH 8.

What This Means
Some of the observations here may be due to the presence of other membrane-bound proteins, not just hydrogenase.

Reference:

Hyndman, L. A., Burris, R. H. & Wilson, P. W. Properties of Hydrogenase from Azotobacter vinelandii. J. Bacteriol. 65, 522–531 (1953).

054 - Nitrogen fixation by Azotobacter vinelandii in tungsten-containing medium

This study investigates how the alternative (vanadium) nitrogenase interacts with tungsten.

What They Saw
They grew Azotobacter vinelandii UW (aka CA) or a tungsten-tolerant mutant with tungsten and Mo or V, then extracted the nitrogenase.

This mutant was derived by growing UW with W and ammonium over several passages to try to remove all the Mo the cells might be storing, and then plating the cells on plates without ammonium, so they would have to fix nitrogen in the presence of W and absence of Mo to survive. Stuff that grew was tungsten-tolerant: strain LM2.

Then to get nitrogenase, they grew LM2 with tungsten and UW with tungsten and ammonium (not sure why it would produce nitrogenase in that condition, though I guess it might when it ran out of fixed nitrogen).

Both UW and LM2 could grow on medium with Mo, though LM2 was sometimes slower. With W, LM2 grew faster or slower depending on amount of W, while UW did not grow.

They used something called rocket immunoelectrophoresis to measure how much MoFe nitrogenase components each strain produced with W. Each produced about the same amount of dinitrogenase reductase in all conditions, but UW with 1mM W produced about 56% the amount of dinitrogenase as it produced with Mo, and LM2 with 10mM W produced about 7% of the amount that UW produced with Mo. I wonder how accurate this technique is. But they also note that the specific acetylene reduction activity of crude extracts was about 6-8% for each strain grown in W compared to UW in Mo. Does this include alternative nitrogenase activity? They say no. But overall, it seems that both strains produce less Mo nitrogenase with W, LM2 less than UW, but LM2's is relatively more active.

I think what happened next was that they couldn't isolate nitrogenase well from LM2, so they studied it from UW. They had two kinds: typical Mo nitrogenase, and Mo/W nitrogenase that had one FeW-cofactor and one FeMo-cofactor. This latter showed less activity in every way: hydrogen under argon, nitrogen, or acetylene atmospheres, nitrogen reduction, and acetylene reduction. This fits with at least one previous study (045). There is still activity though, even with nitrogen reduction, so it's not clear why the cells can't grow in W; maybe its interaction with other important proteins?

What This Means
It's possible that instead of each molecule of Mo/W protein containing one W and one Mo, half the protein could have all Mo and the other half all W; this would give the same results, but seems a less likely explanation. Of course, both possibilities are pretty weird and confusing.

According to some chemistry stuff they did, it seems like the enzyme can't reduce FeW-cofactors, which could reduce the possible electron flux by half, I think. There are a lot of mysteries here.

Reference:

Hales, B. J. & Case, E. E. Nitrogen fixation by Azotobacter vinelandii in tungsten-containing medium. J. Biol. Chem. 262, 16205–16211 (1987).

045 - Tungsten incorporation into Azotobacter vinelandii nitrogenase

Tungsten is known to cause problems for molybdenum-containing enzymes. This report looks into its effect on Azotobacter vinelandii's Mo nitrogenase.

What They Saw
They grew A. vinelandii OP (aka CA) in Burk without Mo, with added ammonium phosphate. Because it's really hard to get rid of every little bit of Mo, they added lots of tungsten (W) to make sure that they could see it if it got incorporated into enzymes. Some of the W was radioactive.

W didn't inhibit growth when ammonia was present, which makes sense. But it did inhibit it, about the same, with N2, nitrate, or urea. The enzymes that use these N sources all need Mo. When Mo was about 0.1 μM, it took 20 μM W to inhibit growth 50%; when Mo was 10 μM, it took 4 mM W. When just a little ammonia was added, it took about 5000 times more W than Mo to stop growth.

When they purified nitrogenase from these W-grown cells, they actually did see acetylene reduction activity, though not nearly as much as with normal Mo nitrogenase. The W content of extracts was very high, though it seemed to be easily removable. Specifically purifying Fe-W protein and comparing to the Fe-Mo version, all activities seemed relatively low: acetylene reduction, hydrogen production, and ATP hydrolysis.

What This Means
A. vinelandii might treat W the same as it treats Mo: taking up as much as it can and storing what it doesn't incorporate. But it does seem to incorporate some into the Mo nitrogenase. This seems to result in a poorly functional enzyme, but is that enough to stop cells from growing entirely? Maybe W's effects on other enzymes cause some problems too.

Reference:

Benemann, J. R., Smith, G. M., Kostel, P. J. & McKenna, C. E. Tungsten incorporation into Azotobacter vinelandii nitrogenase. FEBS Lett 29, 219–221 (1973).

053 - Purification to homogeneity of Azotobacter vinelandii hydrogenase: a nickel and iron containing αβ dimer

This study purified and characterized the uptake hydrogenase from Azotobacter vinelandii.

What They Saw
As a single unit, not denatured, the hydrogenase ran as a single band on a gel. On a denaturing gel it was two bands though, 31 and 67 kDa, indicating that it had two subunits, a larger and a smaller (HoxG and HoxK). The whole thing is about 98.6 kDa. This contrasts with previous reports that it was a single subunit of 60 kDa (049), but it's not really clear why there's this discrepancy.

They also measured metal contents, and found 6.6 mol iron and 0.68 mol nickel per mol hydrogenase, so they rounded up to 1 mol Ni and 10 mol Fe, 1:10 ratio.

They also contradicted the previous result that the hydrogenase couldn't donate to acceptors with negative potential. But they confirmed the possibility that hydrogenase produces hydrogen when paired with a very reduced donor, except this peaked at a higher pH (6 to 8.5).

Reference:

Seefeldt, L. C. & Arp, D. J. Purification to homogeneity of Azotobacter vinelandii hydrogenase: a nickel and iron containing αβ dimer. Biochimie 68, 25–34 (1986).

518 - Cloning, sequencing and characterization of the [NiFe]hydrogenase-encoding structural genes (hoxK and hoxG) from Azotobacter vinelandii

To continue our journey back in time through the discovery of the uptake hydrogenase genetics in Azotobacter vinelandii, this study describes the sequencing of the hydrogenase structural genes, hoxKG.

What They Saw
They probed A. vinelandii DNA with a chunk from A. chroococcum with its hydrogenase genes, and sequenced a fragment they found. This fragment contained three full open reading frames (ORFs) and possibly the beginning of a fourth. The first two encode the structural subunits, HoxK and HoxG (for hydrogen oxidation), based on comparison to sequences in other organisms.

HoxK seems to undergo some processing, since the final protein is smaller than that predicted by the gene sequence; there might be a signal sequence that localizes the protein to the membrane or something. The structural ORFs overlap each other a little, indicating that they're probably transcribed as a unit (in the same operon) which helps the proportions come out right.

The third ORF, ORF3 (which we know now is hoxZ) was also similar to other organisms, and seemed to be a membrane protein of some sort. Possibly part of the hydrogenase electron transport chain.

Reference:

Menon, A. L., Stults, L. W., Robson, R. L. & Mortenson, L. E. Cloning, sequencing and characterization of the [NiFe]hydrogenase-encoding structural genes (hoxK and hoxG) from Azotobacter vinelandii. Gene 96, 67–74 (1990).

052 - Nucleotide sequences and genetic analysis of hydrogen oxidation (hox) genes in Azotobacter vinelandii

This study looks at the hydrogenase-related genes in Azotobacter vinelandii in between the structural genes and the hyp genes.

What They Saw
They already knew about the structural genes, hoxKG. So they got some overlapping fragments containing those and sequences downstream of them, and sequenced them. They found 5 new open reading frames (ORFs), ORF3-7. They named these hoxZMLOQ. There was also part of an eighth at the end. All of them were homologous to hydrogenase genes in other organisms.

Then they knocked out each gene by inserting kanamycin or kan+lacZ cassettes, though probably some of these insertions affected other genes (polar effects) so it wasn't possible to study each gene individually. Each insertion abolished hydrogen oxidation. All 5 new ORFs seem to be involved in subunit processing, as the unprocessed form was present in higher proportion (or alone) in their knockouts.

The knockouts didn't seem to be impaired in growth, even when fixing nitrogen, though they might show more effect in certain conditions (like carbon limitation).

HoxZ seems to be a membrane protein, possibly a cytochrome, so part of the electron transport chain. HoxL seems like it might be involved in metal-binding somehow.

What This Means
Knowing more of the sequence of the operon helps to study it. The whole thing, hox and hyp, is pretty big, kinda surprising for an enzyme that doesn't seem to help the organism in some conditions. Probably in nature, the likelihood of encountering carbon-limited circumstances is much higher, so having an uptake hydrogenase is important.

Reference:

Menon, A. L., Mortenson, L. E. & Robson, R. L. Nucleotide sequences and genetic analysis of hydrogen oxidation (hox) genes in Azotobacter vinelandii. J Bacteriol 174, 4549–4557 (1992).

051 - The hypE Gene Completes the Gene Cluster for H2-oxidation in Azotobacter vinelandii

As a followup to 050, this study completed the hyp operon by sequencing hypE. They suspected its existence because other sequenced operons always had a hypE after the hypD, so they purified a fragment from Azotobacter vinelandii that hybridized hypD and contained some downstream region.

What They Saw
The open reading frame (ORF) in the downstream region had good homology to hypE genes in other organisms. A few bases of hypD and hypE genes overlap, suggesting that they're translated together so there should be equal amounts present so they can work together equally.

They knocked out hypE by inserting a lacZ and kanamycin resistance cassette, which also let them detect expression (the product of lacZ can break down certain compounds to make color, so color intensity can be measured as a proxy for enzyme activity). It seemed like transcription of all the hox and hyp genes was in the same direction, and knocking out hypE meant that A. vinelandii couldn't oxidize hydrogen anymore.

They also noticed that there was a higher proportion of unprocessed hydrogenase structural component, so HypE might be involved in processing; this seems true in other organisms. It also affects the localization of hydrogenase, soluble or membrane-bound. It may affect pre-protein folding.

What This Means
This study completes the discovery of the whole uptake hydrogenase gene set in A. vinelandii. It's pretty similar to the gene sets in other organisms, including A. chroococcum, and sequencing A. vinelandii's genome confirmed the results.

Reference:

Garg, R. P., Menon, A. L., Jacobs, K., Robson, R. M. & Robson, R. L. The hypE Gene Completes the Gene Cluster for H₂-oxidation in Azotobacter vinelandii. J. Mol. Biol. 236, 390–396 (1994).

050 - Identification of six open reading frames from a region of the Azotobacter vinelandii genome likely involved in dihydrogen metabolism

This study looked into which genes in Azotobacter vinelandii are required for its uptake hydrogenase other than the structural genes, hoxKG.

What They Saw
They sequenced almost 6 kilobases somewhat downstream of the structural genes and found 6 open reading frames (ORFs), the last incomplete, so they sequenced some more to complete it. All 8 seemed to be a single operon. These genes are now called hoxV and hypABFCD (based on homology to E. coli genes).

They predicted that the middle 3 and last ORF (now called hypABF and hypD) would produce proteins good at binding Fe-S clusters. HypB has a histidine-rich region that might be good for binding nickel, and HypF has a zinc finger-like region. They have good homology to genes related to hydrogen metabolism in other organisms. They could be related to hydrogenase regulation or assembly, especially related to Ni and Fe; some (HypBF) have homology to other Ni-related proteins, such as urease.

Reference:

Chen, J. C. & Mortenson, L. E. Identification of six open reading frames from a region of the Azotobacter vinelandii genome likely involved in dihydrogen metabolism. Biochim. Biophys. Acta 1131, 199–202 (1992).

049 - Purification and properties of membrane-bound hydrogenase from Azotobacter vinelandii

Azotobacter has a hydrogenase. Some organisms have hydrogenases that produce hydrogen, or sometimes produce and sometimes break down (reversible). Azotobacter's had not been observed to break down hydrogen (so was considered unidirectional). This study purified the membrane-bound uptake hydrogenase from A. vinelandii and tested its properties.

What They Saw
They measured activity of purified enzyme with an electrode measuring hydrogen oxidation in the presence of methylene blue dye (an electron acceptor).

They measured stability of the enzyme in the presence of oxygen, and found that crude extracts were very stable (and could go for weeks without losing activity), but the more pure the preparation, the less oxygen-tolerant it was: the most pure lost half its activity in 20 minutes at 20% oxygen. This inactivation was irreversible.

With a very good electron donor (methyl viologen), the hydrogenase could produce hydrogen. The highest rate they saw was 3.4 μmol hydrogen per minute per mg protein, which peaked at a fairly low pH (around 4). The rate was almost 0 closer to neutral. Also, the presence of hydrogen in the environment inhibits its production by the enzyme.

Even when electron acceptors were not present, the enzyme could combine one deuterium from D₂ with one hydrogen from water to make HD.

The enzyme seemed to be good at donating electrons to acceptors with positive mid-point potentials but not to negative ones, so it seems to have a higher potential than reversible hydrogenases.

Reference:

Kow, Y. W. & Burris, R. H. Purification and properties of membrane-bound hydrogenase from Azotobacter vinelandii. J Bacteriol 159, 564–569 (1984).

256 - Construction of a recF deletion mutant of Azotobacter vinelandii and its characterization

So we saw that Azotobacter vinelandii seems to have multiple copies of its chromosome in each cell (253,255), but we also seemed to see that it doesn't exhibit the behavior expected of a polyploid organism, at least sometimes (445). Other times it might. One proposed explanation is that it has some mechanism by which it homogenizes its genotype, a type of homologous recombination or "homogenotization."

A possible mechanism for this is the RecF system found in E. coli, which is involved in repairing the genome after recombination. A. vinelandii has a homologous system, so this study investigated the phenotype of cells with it knocked out.

What They Saw
They knocked out recF, recA, or both from A. vinelandii UW (aka CA), by inserting tetracycline resistance cassettes. This was confirmed by Southern blotting. Actually they already had a recA knockout, so they just made that a double-knockout. That was somewhat tricky, since the recombination frequency is much lower than in the wild-type; that's not very surprising, since RecA is probably pretty important for recombination.

They found that knocking out recF seemed to impair recombination too (also not surprising), though not as much as lack of recA. With both gone, the recombination proficiency was even lower. They saw similar results with UV sensitivity (and thus ability to repair DNA damage).

Finally, the question of homogenotization: to test this, they knocked out the nifLA genes (which regulate/activate nitrogenase) by inserting a kanamycin resistance cassette, expecting that if RecA or RecF were involved in homogenotization, the lack of them would mean that it would be easy to isolate cells with kan resistance in some chromosome copies and yet capable of nitrogen fixation (because wild-type nifLA is still present in other copies to turn on the nitrogenase). This approach, again, seems questionable to me due to the presence of multiple nitrogenases. I'm not sure if NifLA are necessary to regulate all three versions, or just the Mo one.

However, they didn't find any transformants from any of the strains (wild-type or recA or recF or double-knockouts) that were both kan-resistant and nitrogen-fixing. So they conclude that these deleted genes aren't involved in so-called homogenotization, since their absence didn't make a difference. Is it possible that the homogenotization happened when cells divided and only those with kan resistance survived? I'm not sure that question was addressed. It doesn't seem known if progeny get just one copy and then make more, or if copies are equally partitioned. I suppose the latter makes more sense, since it is binary fission, as far as we know.

Also, they did say that antibiotics weren't necessary to prevent nitrogen fixation phenotypes, so I guess that addresses my question. And they did more Southern blotting to show that there was a cassette inserted in the nifLA locus in all the transformants, with no wild-type alleles visible.

I'm not convinced that homogenotization is a real thing, but I'm not yet sure how to explain these data. Somehow it seems like A. vinelandii has many copies of its genome, but it's just as possible to transform all of them the same way as to transform just some. Guess I'll keep reading.

Reference:

Badran, H., Sohoni, R., Venkatesh, T. V. & Das, H. K. Construction of a recF deletion mutant of Azotobacter vinelandii and its characterization. FEMS Microbiology Letters 174, 363–369 (1999).

255 - Segregation characteristics of multiple chromosomes of Azotobacter vinelandii

Again on the subject of multiple chromosome copies in Azotobacter vinelandii, they wanted to see if the larger amount of DNA in the cells compared to E. coli actually meant that many more copies of each gene.

What They Saw

They mutagenized A. vinelandii UW (aka CA) with transposons and tried to see if these insertions were present in as many copies as they had observed for other genes (30-40). They tried to find auxotroph mutants by conjugating with E. coli carrying a transposon. They didn't find any A. vinelandii that failed to grow on plates lacking certain amino acids, but they did find some that didn't grow very well unless the amino acids were present. This poor growth got less poor over time in successive generations though, but the cells were still resistant to the selective marker (ampicillin). They interpret this as being related to the proportion of genome copies with an insertion vs. without.

They also saw that mutagenized cells mostly couldn't grow with ampicillin when plated directly, though they could grow on antibiotic-free Burk medium. Then when cells from each of these conditions were transferred again to plates with antibiotic, only some from selective plates grew, while all from non-selective plates grew. The idea is that growing on selective plates, the ability to grow without added amino acids would be lost (since the transposon knocked out that ability in some copies, and those copies would be higher in proportion because of the selection). They should've tried growing cells from selective medium on non-selective medium to see if they got the same result.

And gaining the ability to resist the antibiotic? Shouldn't they have had that from the beginning? This whole set of experiments is unclear.

They also mutagenized cells with a different transposon that conferred tetracycline resistance, isolated DNA from them after growing on different amounts of tet, and probed with radioactive probes, correlating radioactivity to number of copies. They observed more radioactivity in cells grown with greater selective pressure, implying that there were multiple alleles and selection increased the proportion of resistant allele in a population.

I'm still somewhat dubious, but it seems like the data might be fairly solid.

Reference:

Phadnis, S. H., Dimri, G. P. & Das, H. K. Segregation characteristics of multiple chromosomes of Azotobacter vinelandii. J. Genet. 67, 37–42 (1988).

445 - Segregation pattern of kanamycin resistance marker in Azotobacter vinelandii did not show the constraints expected in a polyploid bacterium

This is a paper similar to the last one, by the same authors (254), investigating Azotobacter vinelandii's seeming ability to possess multiple copies of its chromosome, and thus exhibit polyploidy (multiple genotypes in the same locus).

What They Saw
They used A. vinelandii ATCC12837 and knocked out the nifY gene (whose product makes the central Mo cofactor) with a kanamycin resistance insertion, then grew it with or without nitrogen or kanamycin. It shouldn't be able to fix nitrogen with Mo present, though maybe with Mo absent.

Looking at the methods, I couldn't find one of the restriction sites they claim to have used, but it's possible that it was present in their strain and not in mine. All the others seem to be there.

Actually, they transformed cells, plated them out to obtain single colonies on non-selective plates, and then screened these colonies for kanamycin resistance. The idea was that if each cell had multiple copies of its chromosome, only some of the offspring of a resistant cell would have kan resistance, because some would inherit the resistant version of the genome and others wouldn't.

But when they grew more colonies from originally resistant colonies and tested their resistance, all of them were resistant that came from resistant colonies. So once the kan-resistant phenotype was present in a cell, it got passed to all offspring; no evidence of polyploidy.

This study is lacking in some ways: I would've liked to see confirmation of the locus of insertion by sequencing, to make sure it's in the right place. And it would be interesting to see what happened if they selected for nitrogen fixation, as in 254. So the question isn't quite settled.

Reference:

Pulakat, L., Efuet, E. T. & Gavini, N. Segregation pattern of kanamycin resistance marker in Azotobacter vinelandii did not show the constraints expected in a polyploid bacterium. FEMS Microbiology Letters 160, 247–252 (1998).

254 - Isolation and characterization of nifDK::kanamycin and nitrogen fixation proficient Azotobacter vinelandii strain, and its implication on the status of multiple chromosomes in Azotobacter

Others seemed to find that Azotobacter vinelandii had many copies of its chromosome (253). This was done partly by comparison to E. coli, but it seems that A. vinelandii's genome size is similar to E. coli's. And probing for specific genes seemed to show high copy numbers. A. vinelandii cells seem larger (about 12.5x), but is this enough to accommodate 40-80x more DNA? Also, it's possible to knock out genes from the species, which would be difficult if they had many copies of a gene that could substitute for each other. This study investigated.

What They Saw
They grew A. vinelandii OP (aka CA) and did genetic transformations with it, knocking out the nifDK genes (which encode the Mo dinitrogenase) by inserting a kanamycin resistance cassette by homologous recombination.

They plated transformant colonies on plate with or without kanamycin and with or without fixed nitrogen (so four different kinds of plate). The wild-type of course could grow on either plate without kanamycin but neither plate with it. They also saw two kinds of mutant phenotype: one that could grow with fixed nitrogen either with or without kanamycin, and another that could grow on all plates, with or without antibiotic or fixed nitrogen.

They tried to confirm this using PCR across the insertion, and restriction digestions followed by electrophoresis. The primers they report seem to be appropriate for getting the Mo nitrogenase genes.

Trying to figure out what they did, this paper seems to have lots of problems with reporting exactly what kind of digestions/cloning they did... sites they claim to have used don't seem to exist in my copy of the genome, or aren't in the right place, or are in too many places. So who knows what's actually going on genetically with this strain they isolated.

The gel they show from the PCR shows two bands in the kan-resistant, N-fixing mutant, and the bands seem to be the right size to correspond to sequences with and without the resistance cassette insertion. They claim to have sequenced the region but don't report the sequence in the paper.

What This Means
The authors conclude that A. vinelandii is exhibiting behavior suggesting multiple copies of a single gene locus (both with kan resistance and with nitrogen fixation). While it does seem to exhibit both phenotypes (assuming no contamination with multiple strains), I'm not sure the interpretation is clear. A. vinelandii has multiple different enzymes capable of fixing nitrogen, so a mutation in their regulation could explain that phenotype, though it wouldn't explain the bands on the gel. Still, it would be good to see this study replicated; unfortunately, the methods are explained poorly.

Reference:

Suh, M.-H., Pulakat, L. & Gavini, N. Isolation and characterization of nifDK::kanamycin and nitrogen fixation proficient Azotobacter vinelandii strain, and its implication on the status of multiple chromosomes in Azotobacter. Genetica 110, 101–107 (2000).

253 - Multiple chromosomes of Azotobacter vinelandii

Previous studies and some preliminary data supposedly showed that Azotobacter vinelandii cells had up to 40 times as much DNA material as Escherichia coli cells, so this study looks at the form that this excess of DNA takes. Is it all one molecule or are there multiple chromosomes or copies of the same chromosome?

What They Saw
They extracted DNA from A. vinelandii and E. coli and determined the amount per cell (by counting number of cells), finding that E. coli had about 3.4 femtograms per cell (3.4 * 10^-15 g) and A. vinelandii had 135 femtograms. Assuming E. coli's genome is about 4 megabases (actually 4.6, though it depends on the strain) and it has only one copy per cell, A. vinelandii should have about 160 megabases-worth of DNA per cell. A. vinelandii's genome is only about 5.4 megabases, so that's about 30 copies of the genome per cell.

However, they made an artificial mixture of known numbers of copies of a certain gene in A. vinelandii and measured the intensity of radioactivity for the correct band of a Southern blot when probed with a probe labeled with radioactive phosphorus. The amount of radioactivity for DNA extracted from cells was about twice the amount seen with 40 copies of the gene, so they concluded there must be about 80 copies of the gene present. Does this mean there are 80 copies of the chromosome per cell? Not necessarily; there could be multiple copies of the gene per chromosome. However, I don't think that is the case.

They also tried with nitrogenase nifDK genes and got similar amounts of radioactivity. nifH gave multiple bands (presumably because of the alternative nitrogenases), but the main band gave a similar brightness.

To distinguish between a giant chromosome with 80 copies of each gene, and 80 copies of a smaller chromosome (or something in between), they integrated a resistance marker in a particular place in the genome. If it were a single large genome, the marker would probably only integrate once or a few times, whereas after several generations, the cell would make copies of a smaller genome with an integrated marker such that the amount of marker would increase over time. They observed the latter result, suggesting single copies of genes on a chromosome with many copies.

What This Means
The copy number of Azotobacter chromosomes was about 30-40 compared to E. coli but about 80 in terms of specific gene copies. A possible way to reconcile this is that the E. coli cells might actually have had more than one copy too. In any case, this many copies of the genome might make it difficult to stably transform the organism.

Reference:

Nagpal, P., Jafri, S., Reddy, M. A. & Das, H. K. Multiple chromosomes of Azotobacter vinelandii. J. Bacteriol. 171, 3133–3138 (1989).