There are a couple of generally accepted suggestions: 1) it consumes all the oxygen around it quickly by respiration decoupled from energy production, a process called respiratory protection; or 2) it can deactivate nitrogenase temporarily when overwhelmed with oxygen, and reactivate it when oxygen levels are under control again.
But the authors of the current study, Kerstin Linkerhägner and Jürgen Oelze, question the first of these mechanisms for several reasons (respiration rates and oxygen levels don't always correlate, oxygen can enter the cell without inactivating nitrogenase, etc), and propose their own: that nitrogenase can protect itself from oxygen by reducing it to water, as long as it has the energy needed to do so in the form of ATP. This they call autoprotection.
To test this hypothesis, they grew A. vinelandii wild-type and other strains in chemostats with different levels of oxygen and other nutrients. I'll try to describe each experiment and its possible interpretations.
First the wild-type was grown in carbon-limited conditions, with different levels of oxygen and different dilution rates (D). They measured rates of respiration per cell and concentrations of protein per volume of culture (as a proxy for biomass).
What they saw was, at the lowest levels of dissolved oxygen, biomass increased slightly over lower D values, when cells were growing more slowly, but then leveled off; at higher levels of oxygen, biomass increased constantly over the range of D values. In all cases, though, higher oxygen meant lower biomass.
Respiration increased over the range of D values also, at all concentrations of oxygen, but higher oxygen meant higher respiration.
Glucose was determined to be fully and equally consumed in all these conditions, so it was truly limiting. Nitrogen fixation (as measured by fixed nitrogen per cell) stayed constant over different levels of oxygen, and increased as D increased.
At the highest level of oxygen, levels of ATP per cell increased with increasing D, while levels of ADP and AMP (spent ATP) stayed pretty constant.
Linkerhägner and Oelze's interpretation: Generally at higher dilution rates, cells' use of energy becomes more efficient, but high levels of oxygen inhibits this in A. vinelandii. So at the lowest level, its efficiency maxed out and the biomass stopped increasing, but at higher oxygen levels, efficiency never reached its maximum potential.
My interpretation: Since oxygen must have some manner of inhibitory effect on biomass production, probably by diverting resources away from growth. The increased respiration must not be providing extra resources for growth. It cannot be determined whether the extra respiration removes oxygen directly, or generates extra ATP for nitrogenase to use for oxygen removal. Since increased respiration did mean increased consumption of oxygen in this case, it seems to go against what L and O suggest in their introduction (that respiration and oxygen levels don't correlate). The possibility of extra ATP for use by nitrogenase is not supported, but I suppose not ruled out yet.
They didn't measure change in adenine nucleotides over different oxygen levels for some reason, so it's hard to say how that changed. Over increasing dilution rates at the highest level of oxygen, both ATP (but not ADP) and fixed nitrogen increased; I'm not sure what this means.
Then A. vinelandii was grown in phosphate-limited or phosphate-sufficient conditions, with two different levels of oxygen, to control the maximum possible amounts of ATP present.
As supplied phosphate increased, biomass increased (not surprising); respiration rates dropped; and levels of adenosine nucleotides increased. The same trends were observed as levels of dissolved oxygen decreased while holding phosphate levels constant. Nitrogen-fixing remained constant over all conditions.
Linkerhägner and Oelze's interpretation: Somehow these results support their hypothesis.
My interpretation: With sufficient phosphate, despite increases in respiration at higher levels of oxygen, levels of ATP dropped (ADP was fairly constant). With limiting phosphate, higher oxygen meant higher respiration but constant ATP and ADP (presumably the cells were making all they could manage given the limited phosphate). But since nitrogen fixation seemed equal in all conditions, limited phosphate and lower ATP didn't really seem necessary for its protection from oxygen.
ATP per Nitrogen Fixed
When L and O plotted amounts of ATP per cell against fixed nitrogen per cell over all these conditions, the points all ended up in a rather linear relationship:
|Figure 3, Linkerhägner and Oelze 1997|
ATP Regeneration per Oxygen Consumed
Supposedly it is possible to calculate ATP regeneration rate by multiply dilution rate by cellular ATP content. When L and O did this and plotted the values against the corresponding oxygen consumption rates, the points from the glucose-limited cultures showed linear relationships, but from phosphate-limited cultures the ATP regeneration seemed constant over different levels of oxygen consumption.
For the glucose-limited points, cells in low-oxygen conditions regenerated ATP much faster at lower rates of oxygen consumption than those in high-oxygen conditions. Phosphate-limited cells regenerated ATP pretty slowly even when consuming lots of oxygen.
Nitrogen fixed vs. ATP Regeneration
Finally they plotted amount of rate of fixing nitrogen per cell (D times nitrogen concentration) over rates of ATP regeneration in each condition, and the linear relationship incorporated all the points from all conditions, including with the mutant strains as in Figure 3. So higher rates of nitrogen fixation correlate with higher rates of ATP regeneration.
In carbon-limited cultures, energy is limited. How this affects things is complex, since everything in the cell requires energy, not just growth, and growth requires other processes than just energy generation (for example, nitrogen fixation).
Interpretation of these results seems to depend a whole lot on previous studies that are likely to be as complex as this one.
At this point, my brain kinda hurts. Physiology is complicated. I'm not sure I buy their conclusions though; I think if I want to understand it better, I'll need to read more about what results to expect from oxygen sensitivity.
Citation: Linkerhägner, K. & Oelze, J. Nitrogenase activity and regeneration of the cellular ATP pool in Azotobacter vinelandii adapted to different oxygen concentrations. Journal of Bacteriology 179, 1362–1367 (1997).