Evolution of Nitrogenases

[exchemist]

Following on from @Moontanman's thread on a new nitrogen-fixing organelle, I started wondering how biological nitrogen fixation first arose at the dawn of life. I found the linked paper, which I thought very interesting on the subject: https://www.sciencedirect.com/science/article/pii/S0966842X23000914 

The writers focus on the metal atoms (or metal/sulphide complexes) which are at the heart of nitrogenases, which can bind nitrogen, lower the strength of the triple bond and progressively add H+ and electrons to form eventually 2 molecules of ammonia. There are 3 variants of nitrogenase, one using just Fe, one using Fe and Mo (molybdenum) and one using Fe and V (vanadium). I was surprised to see these 2 transition metals have such a biologically important role, but there you go. 

It seems there is evidence the first nitrogenase appeared in the Archaean, before the Great Oxygenation Event (i.e. global-scale photosynthesis), which I suppose is not a surprise, seeing as a lot of life would be needed to  geo-engineer the planet, and that would require a lot of fixed nitrogen. They suggest that, before the GOE, there would have been a lot of Fe²⁺ in the oceans, whereas under oxidising conditions this would go to Fe³⁺, the salts of which tend to precipitate from aqueous solution, so would be less bio-available. So a system incorporating Fe is not hard to explain.

Curiously, though, phylogenetic analysis suggests that the version incorporating Mo as well as Fe was the first to appear, even though the concentration of Mo in the early ocean was apparently very low. That version has better kinetics, which may have favoured it, but it still leaves open the issue of where the Mo came from. They speculate that there may have been higher local concentrations in the zones where the first nitrogenase arose, perhaps in hydrothermal vents. But this is very much open-ended and needs further research.

By the way I found the chemistry of these nitrogenases really interesting. There seems to be some very unusual chemistry, involving bridged hydrides to supply the extra electrons  needed for the reduction. But that's another subject. 

It seems the evidence is that nitrogenases are an "evolutionary singularity", meaning this little family of 3 closely related variants, using the 3 metal combinations mentioned, seems to have evolved once only in the whole history of life on Earth. But absolutely vital to the whole enterprise of course. 

Edited April 29 by exchemist

[CharonY]

4 minutes ago, exchemist said:

It seems there is evidence the first nitrogenase appeared in the Archaean, before the Great Oxygenation Event (i.e. global-scale photosynthesis), which I suppose is not a surprise, seeing as a lot of life would be needed to  geo-engineer the planet, and that would require a lot of fixed nitrogen.

There is another hint: nitrogenase are sensitive to oxygen so they do not work well if too much oxygen is present. 

From what I remember, nitrogen is thought to be limiting during evolution of early life and the ability to fix nitrogen would have been a massive benefit. Nitrogenases and their cofactors (especially FeMoCo) have a massive body of lit (and I dabbled a bit with it as a grad student) so there is a lot to read on this topic :).

 

[exchemist]

4 minutes ago, CharonY said:

There is another hint: nitrogenase are sensitive to oxygen so they do not work well if too much oxygen is present. 

From what I remember, nitrogen is thought to be limiting during evolution of early life and the ability to fix nitrogen would have been a massive benefit. Nitrogenases and their cofactors (especially FeMoCo) have a massive body of lit (and I dabbled a bit with it as a grad student) so there is a lot to read on this topic :).

 

Yes I think there is a reference in the paper on the other thread about the means organisms use to keep oxygen away from the active centre. I can imagine that a metal site that can bind N2 might also bind O2 - even might prefer to do so - which would stop it working. (Reminiscent of how carbon monoxide blocks haemoglobin, though perhaps not an exact parallel.)  

In chemical terms it's really fascinating, since the N-N triple bond is so notoriously hard to break. 

[Moontanman]

31 minutes ago, exchemist said:

Following on from @Moontanman's thread on a new nitrogen-fixing organelle, I started wondering how biological nitrogen fixation first arose at the dawn of life. I found the linked paper, which I thought very interesting on the subject: https://www.sciencedirect.com/science/article/pii/S0966842X23000914 

The writers focus on the metal atoms (or metal/sulphide complexes) which are at the heart of nitrogenases, which can bind nitrogen, lower the strength of the triple bond and progressively add H+ and electrons to form eventually 2 molecules of ammonia. There are 3 variants of nitrogenase, one using just Fe, one using Fe and Mo (molybdenum) and one using Fe and V (vanadium). I was surprised to see these 2 transition metals have such a biologically important role, but there you go. 

It seems there is evidence the first nitrogenase appeared in the Archaean, before the Great Oxygenation Event (i.e. global-scale photosynthesis), which I suppose is not a surprise, seeing as a lot of life would be needed to  geo-engineer the planet, and that would require a lot of fixed nitrogen. They suggest that, before the GOE, there would have been a lot of Fe²⁺ in the oceans, whereas under oxidising conditions this would go to Fe³⁺, the salts of which tend to precipitate from aqueous solution, so would be less bio-available. So a system incorporating Fe is not hard to explain.

Curiously, though, phylogenetic analysis suggests that the version incorporating Mo as well as Fe was the first to appear, even though the concentration of Mo in the early ocean was apparently very low. That version has better kinetics, which may have favoured it, but it still leaves open the issue of where the Mo came from. They speculate that there may have been higher local concentrations in the zones where the first nitrogenase arose, perhaps in hydrothermal vents. But this is very much open-ended and needs further research.

By the way I found the chemistry of these nitrogenases really interesting. There seems to be some very unusual chemistry, involving bridged hydrides to supply the extra electrons  needed for the reduction. But that's another subject. 

It seems the evidence is that nitrogenases are an "evolutionary singularity", meaning this little family of 3 closely related variants, using the 3 metal combinations mentioned, seems to have evolved once only in the whole history of life on Earth. But absolutely vital to the whole enterprise of course. 

Very interesting, thank you for investigating this! I'd give you a thumbs up but I'm out for the day.

Did Ni ions play a role in this or am I thinking of another process that uses Ni? 

[exchemist]

Just now, Moontanman said:

Very interesting, thank you for investigating this! I'd give you a thumbs up but I'm out for the day.

Did Ni ions play a role in this or am I thinking of another process that uses Ni? 

I haven't seen any mention of Ni in the paper or the Wiki article on nitrogenases that I quickly read to try to understand more about them. 

But it is interesting I think to reflect on how many of these heavy metals play such critical roles in life. We often think in terms of H, C, N, O, S, and P, plus a handful of s-block cations, but actually a huge array of heavy elements gets pressed into service as well. Their multiple oxidation states and d orbitals turn out to be pretty important. 

[CharonY]

11 minutes ago, exchemist said:

But it is interesting I think to reflect on how many of these heavy metals play such critical roles in life. We often think in terms of H, C, N, O, S, and P, plus a handful of s-block cations, but actually a huge array of heavy elements gets pressed into service as well. Their multiple oxidation states and d orbitals turn out to be pretty important. 

Among those, Fe plays an outsized role for tons of redox reactions. But some of the rarer ones (including Mo and Va) have been utilized in rather critical enzymes and have not been replaced by more common metals, which in itself is interesting. If you are interested, there are whole fields on metalloenzymes, with recent approaches how various moieties in these large enzyme complexes might move during the various electron transfer processes. Not entirely my world, but it pops up frequently (and sometimes you get to work with folks on things like these).

And also the work with them is annoying, just getting your media and glass ware free of metals is a nightmare.

[TheVat]

Haven't read full paper yet but imagine they would consider hydrothermal ocean vents for sources of metals like moly.  From all I've gleaned over some years it seems like many early chemosynthesis roads lead back to the hot vents.  A lot of interesting metabolism-first prebiotic hypotheses go to the vents.

Hydrothermal vents/chimneys are pretty porous so there are large surface areas of potentially catalytic minerals, through which vent effluents circulate under high pressure and across large temperature gradients.  I can see N fixation happening there in those massive geochemistry reactors.

(nice, btw, to suddenly find more threads that are not on fundamental physics - not that FP isn't wonderful, but my comfort level is higher in biochemistry and biology generally)(at least I can follow the equations better, haha)

4 hours ago, Moontanman said:

Very interesting, thank you for investigating this! I'd give you a thumbs up but I'm out for the day.

I will donate one, in your name.

[CharonY]

9 hours ago, exchemist said:

Yes I think there is a reference in the paper on the other thread about the means organisms use to keep oxygen away from the active centre. I can imagine that a metal site that can bind N2 might also bind O2 - even might prefer to do so - which would stop it working. (Reminiscent of how carbon monoxide blocks haemoglobin, though perhaps not an exact parallel.)  

I believe the iron clusters got oxidized and destabilized under oxygen. CO is also an inhibitor and it actually binds to nitrogenases like haemoglobin, IIRC. As a side note, oxidizing and often destabilizing iron cores is one of the ways many organisms sense oxygen and quite a few regulatory factors related to oxidative stress (and also iron metabolism) are using that.  

4 hours ago, TheVat said:

(nice, btw, to suddenly find more threads that are not on fundamental physics - not that FP isn't wonderful, but my comfort level is higher in biochemistry and biology generally)(at least I can follow the equations better, haha)

Well, you could have started them! I am just sitting here, looking at physics threads and pretend to understand what swansont is explaining.

[exchemist]

3 hours ago, CharonY said:

I believe the iron clusters got oxidized and destabilized under oxygen. CO is also an inhibitor and it actually binds to nitrogenases like haemoglobin, IIRC. As a side note, oxidizing and often destabilizing iron cores is one of the ways many organisms sense oxygen and quite a few regulatory factors related to oxidative stress (and also iron metabolism) are using that.  

Well, you could have started them! I am just sitting here, looking at physics threads and pretend to understand what swansont is explaining.

This has prompted me to revise the bonding scheme for transition metals with carbon monoxide: https://en.wikipedia.org/wiki/Metal_carbonyl

This involves a dative σ-bond from the lone pair on C to the metal and a π back bond from occupied d orbitals on the metal to low-lying π* antibonding orbitals on CO. So  donation of an electron pair in both directions, preserving neutrality overall. The effect of the involvement of the antibonding orbital is to weaken the C≡O bond (nominally triple in the free CO molecule) and strengthen the M-C bond.  

This I think gives a clue as to how such metal atoms can weaken the N≡N triple bond (preparatory to adding H atoms), N≡N being isoelectronic with C≡O. It starts to become clearer........

N.B. The need for an electron pair in a metal d-orbital to make the back-bond requires the metal to be in a low oxidation state. 

 

Edited April 30 by exchemist

[exchemist]

11 hours ago, TheVat said:

Haven't read full paper yet but imagine they would consider hydrothermal ocean vents for sources of metals like moly.  From all I've gleaned over some years it seems like many early chemosynthesis roads lead back to the hot vents.  A lot of interesting metabolism-first prebiotic hypotheses go to the vents.

Hydrothermal vents/chimneys are pretty porous so there are large surface areas of potentially catalytic minerals, through which vent effluents circulate under high pressure and across large temperature gradients.  I can see N fixation happening there in those massive geochemistry reactors.

(nice, btw, to suddenly find more threads that are not on fundamental physics - not that FP isn't wonderful, but my comfort level is higher in biochemistry and biology generally)(at least I can follow the equations better, haha)

I will donate one, in your name.

I admit I've slightly lost the plot on where we are these days on the various scenarios for abiogenesis. I had the idea thermal vents had gone out of fashion a bit for some reason, but I agree these heavy metal biochemistries suggest something like that. The use of sulphide sulphur is also a bit suggestive.  

[Moontanman]

7 hours ago, CharonY said:

Well, you could have started them! I am just sitting here, looking at physics threads and pretend to understand what swansont is explaining.

I know, if not for @swansont I know I would have ridden "certain" threads off into the desert to be lost forever by now!  

[sethoflagos]

47 minutes ago, exchemist said:

I admit I've slightly lost the plot on where we are these days on the various scenarios for abiogenesis. I had the idea thermal vents had gone out of fashion a bit for some reason, but I agree these heavy metal biochemistries suggest something like that. The use of sulphide sulphur is also a bit suggestive.  

Bear in mind that for its first 2 billion years, the earth was a temperate water world with much reduced continents, and a mildly reducing atmosphere of N₂ and CO₂ perhaps with minor CO and CH₄. Sediments from that period (unlike later deposits) generally feature pyrite, pitchblende and siderite indicative of reducing conditions throughout the ocean basins. 

Levels of volcanic activity were considerably higher than today, including forms like kimberlite pipes that recycled material from very deep within the earth's mantle, including large amounts of the heavier transition metals. However, it is possible that modern plate tectonics didn't really get going until after the major chemosynthesis and retinal based photosynthesis processes had developed so this happened in a somewhat different geological setting than we see today.

Thermal vents must surely have played a part in the development of life - too good an opportunity to miss, but with many of the key micronutrients widely distributed in maybe more soluble reduced forms throughout the oceans, there seems less need for all the ingredients for abiogenesis to be concentrated at a single source location. Splitting up the task between a diverse range of mineralised environments, that may or may not have been closely spaced massively improves the odds in favour. 

[exchemist]

1 hour ago, sethoflagos said:

Bear in mind that for its first 2 billion years, the earth was a temperate water world with much reduced continents, and a mildly reducing atmosphere of N₂ and CO₂ perhaps with minor CO and CH₄. Sediments from that period (unlike later deposits) generally feature pyrite, pitchblende and siderite indicative of reducing conditions throughout the ocean basins. 

Levels of volcanic activity were considerably higher than today, including forms like kimberlite pipes that recycled material from very deep within the earth's mantle, including large amounts of the heavier transition metals. However, it is possible that modern plate tectonics didn't really get going until after the major chemosynthesis and retinal based photosynthesis processes had developed so this happened in a somewhat different geological setting than we see today.

Thermal vents must surely have played a part in the development of life - too good an opportunity to miss, but with many of the key micronutrients widely distributed in maybe more soluble reduced forms throughout the oceans, there seems less need for all the ingredients for abiogenesis to be concentrated at a single source location. Splitting up the task between a diverse range of mineralised environments, that may or may not have been closely spaced massively improves the odds in favour. 

Sure. If you read the link, this is all gone into, but apparently there was very little Mo in the early oceans, hence the interesting question of how the first nitrogenase (or nitrogenase cofactor) was able to employ it. 
 

Reading more about all this, it seems that iron-sulphur complexes are widely prevalent in biochemistry, which is certainly suggestive of volcanic origins, though in the reducing environment pre-GOE, Fe(II) at least would apparently have been available in the oceans.

[sethoflagos]

1 hour ago, exchemist said:

... but apparently there was very little Mo in the early oceans...

Far be it from me to question this, yet the paper itself offers strong evidence of the early appearance of Mo based BNF which implies that Mo was definitely available to life in the oceans. Something of a paradox.

Molybdenum is not a rare element, especially around hydrothermal sites. Perhaps its low apparent oceanic concentration at that time was a reflection of an intense biological demand that kept it locked up in biomass. Just a thought.

[exchemist]

1 hour ago, sethoflagos said:

Far be it from me to question this, yet the paper itself offers strong evidence of the early appearance of Mo based BNF which implies that Mo was definitely available to life in the oceans. Something of a paradox.

Molybdenum is not a rare element, especially around hydrothermal sites. Perhaps its low apparent oceanic concentration at that time was a reflection of an intense biological demand that kept it locked up in biomass. Just a thought.

Hmm, I don't think I buy that explanation, for the simple reason that before the evolution of nitrogenase, life on Earth must have very sparse, constrained by........... the scarcity of fixed nitrogen. So the low Mo in the Archaean oceans can't be attributed to lifeforms scooping it all up. Life would have proliferated only after nitrogenase appeared.  So for that hypothesis to work, one would expect to see a fall in Mo at a certain point. That would be dramatic evidence for the onset of nitrogenase-exploiting organisms, but it is not what they report. Mind you, it isn't clear to me exactly how they have deduced a low level of Mo in the Archaean oceans.  

[sethoflagos]

3 minutes ago, exchemist said:

... Mind you, it isn't clear to me exactly how they have deduced a low level of Mo in the Archaean oceans.  

Me neither. Unaltered sediments from those far distant times are exceedingly rare. Examples from just the right place and right moment are likely long gone.

43 minutes ago, exchemist said:

... before the evolution of nitrogenase, life on Earth must have very sparse, constrained by........... the scarcity of fixed nitrogen. 

And yet life definitely started therefore there was sufficient fixed nitrogen around.

So both the 'no molybdenum' and 'no fixed nitrogen' claims are belied by the undisputable evidence of our existence. It seems that somebody somewhere is extrapolating a small and questionable data set way beyond its scope of applicability.

[exchemist]

1 hour ago, sethoflagos said:

Me neither. Unaltered sediments from those far distant times are exceedingly rare. Examples from just the right place and right moment are likely long gone.

And yet life definitely started therefore there was sufficient fixed nitrogen around.

So both the 'no molybdenum' and 'no fixed nitrogen' claims are belied by the undisputable evidence of our existence. It seems that somebody somewhere is extrapolating a small and questionable data set way beyond its scope of applicability.

Well obviously life did not have to wait for nitrogenase before it could start. No one is making a claim of no fixed nitrogen, just that there was much much less and not enough to sustain a biosphere on anything like the scale of later epochs.

Edited April 30 by exchemist

[sethoflagos]

8 hours ago, exchemist said:

Well obviously life did not have to wait for nitrogenase before it could start. No one is making a claim of no fixed nitrogen, just that there was much much less and not enough to sustain a biosphere on anything like the scale of later epochs.

I can buy this. Nitrogen fixation is energetically expensive, so the evolution of BNF would not be favoured unless abiotic processes failed to satisfy the biological demand for fixed nitrogen. 

However, the claimed unavailability of molybdenum simply does not square with an early appearance of Mo based BNF. The two are mutually contradictory.  

The reference quoted for this low Mo condition is "Proterozoic ocean chemistry and evolution: a bioinorganic bridge?", Anbor & Knoll,  Science 297 (2002). I've not seen the full paper, but the abstract describes the limited availability of (by inference) molybdenum under the Proterozoic (543 - 2,500 mya) conditions of oxic surface waters and anoxic depths. This is one and a half billion years after the period we are discussing and a very different chemical redox environment. Our present interest is in the availability of Mo³⁺ and Mo⁴⁺ in the anoxic surface waters of the early Archaean around 4 billion years ago, not Mo⁶⁺ in oxic waters of a couple of billion years later. Having now invested a bit of time in this, I'm not seeing any relevant evidence for the claim of low Mo availabilty during the period of interest. 

Edited May 1 by sethoflagos

[exchemist]

3 hours ago, sethoflagos said:

I can buy this. Nitrogen fixation is energetically expensive, so the evolution of BNF would not be favoured unless abiotic processes failed to satisfy the biological demand for fixed nitrogen. 

However, the claimed unavailability of molybdenum simply does not square with an early appearance of Mo based BNF. The two are mutually contradictory.  

The reference quoted for this low Mo condition is "Proterozoic ocean chemistry and evolution: a bioinorganic bridge?", Anbor & Knoll,  Science 297 (2002). I've not seen the full paper, but the abstract describes the limited availability of (by inference) molybdenum under the Proterozoic (543 - 2,500 mya) conditions of oxic surface waters and anoxic depths. This is one and a half billion years after the period we are discussing and a very different chemical redox environment. Our present interest is in the availability of Mo³⁺ and Mo⁴⁺ in the anoxic surface waters of the early Archaean around 4 billion years ago, not Mo⁶⁺ in oxic waters of a couple of billion years later. Having now invested a bit of time in this, I'm not seeing any relevant evidence for the claim of low Mo availabilty during the period of interest. 

OK I take your point. I had not looked into the sources they are relying on for the claim of low Mo in the pre-GOE oceans.