Nitrogen fixation is limited to prokaryotes. Some eubacteria and a few archebacteria can fix nitrogen - but no eukaryotic cells can do this. Some N-fixing bacteria are free-living whereas other form symbiotic associations with plants.

Free Living Nitrogen Fixing Bacteria:

Since nitrogenase is inactivated by O2, the fixation of N2 must occur under conditions which are anaerobic at least locally. For anaerobes there is no problem. Facultative organisms such as purple photosynthetic bacteria or Klesbsiella fix N2 only when anaerobic. Other organisms have protective mechanisms. In Azotobacter, an obligate aerobe, the O2 concentration inside the cell is held down by partial uncoupling of a highly active respiratory chain. This wastes carbohydrate, but if growth is limited by absence of nitrogen compounds then this is justifiable. In cyanobacteria O2 is actually generated by photosynthesis. Fixation of N2 occurs in special cells known as heterocysts which do NOT photosynthesize but are devoted solely to N2 fixation.

Symbiotic Nitrogen Fixing Bacteria:

Symbiotic bacteria are protected from oxygen by inhabiting a plant host. Bacteria of the genus Rhizobium and Bradyrhizobium inhabit the root nodules of leguminous plants (e.g. peas, beans, clover, alfalfa, soybeans). Other symbiotic associations occur but are less important. Anabaena azollae, a nitrogen fixing cyanobacterium, lives in pores on the fronds of a water fern called Azolla. This symbiotic partnership is used to enrich rice paddies with organic nitrogen.

Rhizobium is also found free in the soil but only fixes N2 when inside the root nodules of its host plant, in a strictly controlled microaerophilic environment. Oxygen is required to generate sufficient respiratory energy to drive N2 fixation. But too much oxygen inactivates nitrogenase.

In root nodules the O2 level is regulated by a special hemoglobin - leghemoglobin. The globin protein is encoded by plant genes but the heme cofactor is made by the symbiotic bacteria. This is produced only when the plant is infected with Rhizobium. The plant root cells convert sugar to organic acids which they supply to the bacteroids. In exchange, the plant receives amino-acids (rather than free ammonia).

Specific strains of bacteria are found inhabiting specific plant species. For example, a carbohydrate binding protein (lectin) on the surface of root cells of clover (Trifolium) specifically binds to lipopolysaccharide of Rhizobium trifolii which contains 2-deoxyglucose. The bacteria then enter and produce cytokinins (a type of plant hormone) which promote the division of plant cells to form nodules. The bacteria lose their outer membranes and become irregular in shape - "bacteroids".

Structure and Operation of Nitrogenase

Nitrogenase contains the two proteins molybdoferredoxin and azoferredoxin. These must be supplied with reducing equivalents by other proteins that vary. Here we consider nitrogenase from Klebsiella, a close relative of E. coli where the accessory proteins are flavodoxin and pyruvate flavodoxin reductase. In most bacteria electrons are passed from NAD(P)H or pyruvate to ferredoxin, an FeS protein. If iron is in short supply ferredoxin is replaced by flavodoxin, a flavoprotein. In Klebsiella there is no ferredoxin and flavodoxin (NifF protein) is used all the time. Azoferredoxin transfers electrons from reduced flavodoxin (or ferredoxin) to molybdoferredoxin.

Molybdoferredoxin is an alpha2/beta2 tetramer. The alpha and beta subunits are similar but distinct and are encoded by genes nifK and nifD. Each tetramer contains 2 Mo and several FeS groups. The molybdenum is part of a low molecular weight cofactor containing Mo bound to an Fe7S8 cluster and to homocitrate. This MoFe cofactor is unique to nitrogen fixation and distinct from the Mo-pterin cofactor of other Mo proteins (e.g. nitrate reductase, xanthine oxidase). Azoferredoxin is a dimer of identical subunits encoded by nifH and contains a single Fe4S4 group per dimer. Azoferredoxin is modified by the NifM protein. Molybdoferredoxin from one genus can often interact with azoferredoxin from another genus to give active enzyme. These two proteins have several alternative names:

Molybdoferredoxin = component I, MoFe protein, or "nitrogenase"

Azoferredoxin = component II, Fe protein, or nitrogenase reductase

Nitrogenase is not very fast (the turnover number is around 50 moles/min per mole of Mo) and so about 2-5% of the total cell protein is nitrogenase. The reaction N2 + 3H2 Æ 2NH3 actually releases energy. However, the activation energy needed to break the NºN triple bond is very high and in practice energy, as ATP, is consumed by NifH protein (azoferredoxin). If there is an excess of azoferredoxin then ATP tends to be wasted. In Klebsiella nifHDK form an operon that keeps the ratio of components constant.

The Nif (nitrogen fixation) proteins are often referred to by their gene names:

NifJ = pyruvate flavodoxin reductase

nifF = flavodoxin

nifH = azoferredoxin

nifM = processing of NifH protein

nifK,D = molybdoferredoxin

nifB,N,E,V,W,Z = MoFe cofactor synthesis

nifY = MoFe cofactor insertion

nifQ = molybdenum uptake

nifA,L,R = regulation

nifU,S = metal center biosynthesis

nifX,T = function unknown (not necessary, at least under normal conditions)

Mechanism of Nitrogenase

Nitrogenase will reduce many small molecules with triple bonds in addition to nitrogen. Oxygen, which is triple-bonded inactivates nitrogenase. Carbon monoxide, another triply bonded molecule is a competitive inhibitor.

There is no convenient radioactive isotope of N2 and the reduction of nitrogen-15, a heavy but non-radioactive isotope, to NH3 is difficult to measure and requires a mass spectrometer. In practice nitrogenase is usually assayed by its ability to reduce acetylene, CHºCH, to ethylene, CH2=CH2, which is easily detected by gas chromatography. The FeMo cofactor is the active site, and it can reduce acetylene in the absence of the protein if provided with a good chemical reducing agent such as borohydride.

The overall Delta G for N2 + 3H2 = 2NH3 is about -8 kcal/mole. However the first step, opening up the triple bond, is extremely unfavorable:

N2 + H2 = N2H2 Delta G° = +50 kcal/mole (approximately)

N2 + 2e- + 2H+ = N2H2 Eo = –1200 mV (approximately)

Thus nitrogenase has to carry out a single step which needs a reductant with a redox potential of –1200 mV. Alterations in solvation, local pH etc. could bring this down to about –1000 mV but even so this is much more negative than any other biological redox potential.

The redox potential of azoferredoxin is –290mV. [When ATP binds to this protein its redox potential is lowered to –400mV.] The ATP must be hydrolyzed for reduced azoferredoxin to reduce molybdoferredoxin. Two ATP are hydrolyzed per electron transferred or 4ATP/2e. 4ATP yields approximately –30 kcal which is equivalent to 750 mV per pair of electrons. Adding this 750 mV to the Eo of azoferredoxin (–290mV) just over 1000 mV negative - the correct value. This suggests that ATP is used to generate reducing power. The mechanism is unknown, but remember that cells convert reducing power to ATP during respiration.

N2 is reduced at the MoFe cofactor site on the molybdoferredoxin. The intermediates N2H2 and N2H4 (hydrazine) are assumed to exist. Although N2H4 has been detected, N2H2 is very unstable and tends to decompose back to N2 + H2.

Hydrogen is always produced when nitrogenase reduces N2 to NH3. There are two views on this. The first is that this is a side reaction - nitrogenase is such a powerful reductant that conversion of H2O to H2 inevitably occurs. The second view is that under optimum conditions one H2 is evolved per N2 fixed suggesting that H2 evolution is an integral part of the enzyme mechanism. Furthermore, reduction of acetylene to ethylene is not accompanied by H2 evolution. Since the Eo for acetylene is +320 mV it is possible that nitrogenase reduces acetylene when only partly activated and H2 evolution is not necessary in this case.

Proposed Steps In Nitrogenase Mechanism:

The mechanism is largely based on work with non-protein MoFe complexes, some of which will fix N2 chemically (but very inefficiently).

This mechanism also explains why acetylene, C2H2, is a non-competitive inhibitor of N2 fixation. Acetylene reduction discharges nitrogenase before it ever reaches full activation. Although N2 fixation wastes reducing power when H2 is evolved, most N2 fixing bacteria contain hydrogenase which uses gaseous H2 to reduce NAD(P). Hence they recycle the hydrogen at least partly.

Regulation of Nitrogen Fixation

All the nif genes in Klebsiella are clustered and coordinately regulated. E. coli to which the nif genes of Klebsiella have been transferred can fix N2. In both the original Klebsiella and the E. coli nitrogenase is expressed only in the absence of both O2 and NH3 in the growth medium. Other organic N-sources will also repress nitrogenase. The better the N-source the greater the repression.

The nif genes are regulated by the nifLA operon. The nitrogen regulators NtrC (= GlnG), and NtrB determine whether or not the nifLA operon is expressed (depending on the presence of ammonia or organic nitrogen). In the absence of ammonia or organic nitrogen the NtrC protein is phosphorylated by the NtrB protein. NtrC-P then binds to the upstream region of the nifLA operon and activates transcription.

NtrA (= GlnF = RpoN = s54) is the nitrogen sigma factor, which is needed for expression of the nifLA operon and the nif structural genes. NtrA is an alternative sigma factor used by RNA polymerase to recognize many genes involved in nitrogen metabolism which are not recognized by the standard sigma factor. The nifA gene encodes a protein required for switching on all of the nif genes except the regulatory genes nifLA themselves. if NifA protein is made, its function is to activate the other nif genes. The nifL gene is required for O2 repression. In the absence of NifL protein, nitrogenase is made in the presence of O2 (but is inactivated by O2). When oxygen is present, the NifL protein binds to NifA and prevents it from activating the other nif genes.

Alternative Nitrogen Fixation Systems

When Mo is absent some N-fixing bacteria, such as Azotobacter, make an alternative nitrogenase in which vanadium is used instead of Mo. This is encoded by a duplicate set of vnf genes which make VFe cofactor as well as the corresponding nitrogenase proteins. Mo, if available represses the vnf system which is less efficient. If vanadium is also absent Azotobacter can make a third nitrogenase which uses only iron - the even less efficient anf system. The sequences of the nif, vnf and anf genes are very similar.