FEATURE ESSAY 3.1  Protecting nitrogenase from oxygen

Printer-friendly version

W.B. Silvester

While dinitrogen (N2) fixation occurs in many free-living bacteria in soil, it is only when bacteria such as Rhizobium species and Frankia enter symbiotic relationships with a plant that really large quantities of nitrogen are fixed. Crop, livestock and wool production in Australia and New Zealand are highly dependent on plant protein derived from N2 fixation in nodules of legumes (e.g. lupins, lucerne and clovers). In these systems, solar energy drives both photosynthesis and biological reduction of N2 thereby providing a basis for low-energy, efficient production.

Biochemical conversion of N2 to ammonia (NH3) by bacteria is energetically expensive and extremely sensitive to O2 (Section 3.5). Both proteins that make up the nitrogenase complex are irreversibly denatured by even the slightest ‘sniff’ of O2 and all systems that support nitrogenase have evolved mechanisms of creating a low O2 environment at the site of N2 fixation. Nodule respiration exerts a major influence on O2 partial pressure (pO2) in nodules because it consumes O2; an interesting affirmation of the central role of O2 is provided by the problem of cold lability of nodules.

It is common practice in physiological work to harvest tissues onto ice as a way of slowing down metabolism, preventing loss of substrates and retaining tissue integrity. However, during the late 1960s, Esam Moustafa, working at Palmerston North in New Zealand, showed that chilled legume nodules lost most of their activity after warming up, compared to control nodules which were kept at air temperature during harvest. This phenomenon of cold lability defied explanation until the work of Fraser Bergerson in Canberra and others demonstrated the pivotal role of O2 and respiratory O2 uptake in modulating nodule function. When nodules are chilled, respiration slows down (Q10 about 1.8) but O2 continues diffusing into the nodule (Q10 about 1.1), so that lower temperatures disproportionately affect respiration. The build up of O2 due to lowered respiration and higher solubility of O2 at lower temperatures conspire to destroy nitrogenase enzymes in chilled nodules.

Meanwhile, we were working with a group of non-leguminous nodulated plants which were physiologically and geographically distinct from legumes. These actinorhizal plants are found in temperate or cool temperate areas and have representatives in a wide variety of angiosperm families. They include Casuarina (sheoke) in Australia; Coriaria (tutu) in New Zealand and Alnus (alder). They have major roles as N2 fixers and while the nodules look superficially like those of legumes, they form through a symbiosis with a different bacterium, Frankia.

Frankia was first isolated in 1978 and shown by John Tjepkema soon after to fix N2 in air which contains 21% O2. He also showed that actinorhizal nodules are highly aerated structures. This is in strong contrast to legume nodules which maintain a very low pO2 within the infected zone and to Rhizobium which in pure culture will only fix N2 at very low pO2.

We became interested in the mechanisms whereby Frankia and actinorhizal nodules are able to cope with atmospheric levels of O2. Frankia is a rather complex bacterium, which produces both sporangia and a highly specialised cell called a vesicle (Figure 1). We imagined the vesicle might be the site of N2 fixation and therefore provide a clue to O2 protection in this bacterium. Working with two graduate students, Richard Parsons and Sharon Harris, we probed the possibility of modifying the ability of Frankia to cope with O2.


Figure 1 Dark-field microscopy showing Frankia (stain CcI3) grown at (a) 3 kPa O2 and (b) 60 kPa O2. Frankia consists of fine hyphae (c. 1 µm diameter) and when induced to fix N2 it produces rounded terminal vesicles 2-3 µm in diameter. Vesicles respond dramatically to O2 level by producing a thickened envelope

When we grew Frankia in stirred cultures gassed with various pO2, we found that the bacteria were very sensitive to transfer from one O2 level to another. Abrupt O2 shocks killed them. If cultures were not shocked, it was possible to grow Frankia at almost any O2 level. Thus, when cultures are grown at low O2 (e.g. 5 kPa), optimum pO2 for nitrogenase is 5 kPa; likewise it is possible to grow cultures at 70 kPa O2 (over three times atmospheric levels) and nitrogenase activity optimises at around 70 kPa O2.

This mechanism for exquisite adaptation to extreme O2 levels, and ability to express nitrogenase at very high pO2, only became clear by using some interesting optical techniques. John Torrey and colleagues had shown that the vesicle is surrounded by a thick envelope of multilayered lipid which is assumed to be an O2 diffusion barrier, and we proposed that this envelope must adapt to the O2 level at which Frankia is growing. We hit upon dark-field microscopy as a way to visualise the vesicle walls. This technique relies on light from a steep angle outside the field of view in a microscope being refracted off surfaces such as cell walls. Lipid material is highly refractive and shows up particularly well under dark-field. Vesicles from Frankia cultures grown at 3 kPa O2 are very thin walled, but at high pO2 dark-field images show a massive wall around the vesicles (Figure 1).


Figure 2 Freeze-fracture electron micrograph of a vesicle of Frankia grown at 40 kPa O2. Note multiple lipid layers in the vesicle envelope. Scale bar = 1µm

Lipids are difficult to visualise under most light and electron microscopes and the only way in which the envelope could be seen was to use freeze-fracture electron microscopy. We collaborated with Stan Bullivant at Auckland University to obtain freeze-fracture views of the envelope and confirmed that at high pO2 the vesicle may have over 100 layers of lipid tightly packed into a thick envelope (Figure 2).

While the vesicle envelope obviously responds to O2 levels in free-living Frankia, what happens in root nodules? Is the nodule an essential part of the O2 protection mechanism as it is in legumes? We gassed root systems of a variety of nodulated actinorhizal plants with various pO2 and showed that Frankia nodules could likewise adjust to ambient pO2. However, unlike legume nodules, there seem to be several different mechanisms of O2 protection. In Alnus nodules, for example, vesicle envelopes are the important site of adjustment, while in some other nodules, vesicles do not form and we postulate that the host cell wall is an important O2 barrier. Just to make matters more complex, some Frankia nodules, particularly Casuarina, also have leghaemoglobin and function more like legume nodules. We now accept that actinorhizal nodules run the full range of physiologies: Casuarina resembles legumes, relying on low internal pO2 and oxygenated leghaemoglobin, whereas Alnus has very prominent vesicles which apparently provide much of the O2 protection.

No symbiotic mechanism for O2 protection of nitrogenase has evolved as the most efficient. If, in an Alnus symbiosis, a vesicle is proven to be the prime site of O2 protection, this has some implications for developing novel symbioses. Protection against O2 in legume nodules is largely host provided but a bacterially derived O2 protection mechanism as in Frankia might one day be transferable to legumes. A much-simplified nodule could then be envisaged to sustain biological N2 fixation.

Further reading

Schwintzer, C.R. and Tjepkema, J.D. (1990). The Biology of Frankia and Actinorhizal Plants, Academic Press: New York.

Benson, D.R. and Silvester, W.B. (1993). ‘Biology of Frankia strains, actinomycete symbionts of actinorhizal plants’, Microbiology Reviews, 57, 293–319.