Saturday, August 30, 2014

Life's Deepest Divergence

Why do the membrane chemistries of Archea and Bacteria differ so much?

Hypotheses for the origin of life increasingly tend away from quiet ponds & soups and towards the more dramatic undersea vents with their billowing pumes of geochemicals. Life needs energy, and these have energy in abundance, in the form of chemicals rather than light, which took much longer to harness. Perhaps the leading current hypothesis speculates that a constant stream of highly reduced and alkaline brew of sulfides and carbon compounds allowed nascent chemical complexity to build up in this roiling, rocky boundary zone.


That may form the basis of another post. But one aspect of this theory is that membranes were relatively late to the party. Semi-isolated nodules / holes could occur in the rocky matrix that allowed just the right amount of flow of small metabolic chemicals while anchoring the larger organic, (pre-biotic) macromolecules. Only the transition to living free in the ancient seas necessitated increasingly tight membranes, cell walls, etc.

A recent paper uses this theory as the springboard for a theory for the divergence which is the deepest among currently living organisms- that between Archaea and Bacteria. Eukaryotes were an immensely complicated fusion of Archaeal and Bacterial cells which happened much later on, and is another story altogether. Archaea and Bacteria share a vast majority of core systems such as DNA-based genetics, RNA-based transcription and translation, ribosomes, circular DNA, lack of internal organelles, and most of the basic, carbon-based metabolism, with phosphate energy carriers on ATP and its relatives. They differ in their use of RNA polymerases (Archaea have three, as do Eukaryotes), in their transcription factors and histones (Archaea have histones, and more complex transcription factor system, simlar to that of Eukaryotes), and in details of DNA replication. Their cell walls have different chemistries, and most oddly, their membranes have quite different chemistries. It is noteworthy that Archaea tend to inhabit the most extreme environments of temperature, salinity, acidity, etc., suggesting that they may reflect their ancient heritage in ecological terms more so than do Bacteria.

The chemistries of membrane lipids are strikingly different between Archaea and Bacteria. The lipid, head group, and linkages on either end of the glycerol core are each distinct.

Membranes have magical properties. In modern organisms, they keep all large molecules, and many small ones, out of the cell using only the thinnest layer of two molecules- the bilayer. This bilayer molecule has a charged or polar headgroup (the P and glycerol / 3-C backbone above) which keeps water happy on one side. And it has fatty tails, which is to say long (CH2)n chains, which make water very unhappy, and form the stable center of the bilayer sandwich, which repells all sorts of charged ions as well as water.

Since such a membrane seals out ions, which are the life-blood of metabolism and of life in general, such a membrane presupposes a large cast of (protein) ion channels which allow selected ions in/out, or in advanced cases, pump them actively. Thus the well-sealed membrane can not have been a terribly early event in the story of life. The current authors propose that early membranes were quite different, and quite leaky, establishing the sort of partial, controlled traffic that the earliest cells needed to replicate or supplement their early rocky homes. Thus the transition to tightly sealed membranes occurred later on, after life had gotten quite far along, and after the Archaeal / Bacterial split.

What are these reactions that could happen in a leaky cell, at a sea-floor vocanic plume? To introduce this requires take a brief detour into the chemiosmotic theory- one of the most elegant and significant theories in biology, after those of evolution and DNA structure. ATP had long been known to be the basic energy currency of biological organisms, being converted to ADP and AMP in a constant cycle of re-use. But in 1961 Peter Mitchell proposed another biological energy currency- electricity in the form of ionic differences around membranes. The issue was where the cell's ATP charging capacity comes from. It seemed localized to mitochondria in Eukaryotes, but the mechanism was unknown.

An ATPase in the mitochondrial membrane diligently manufactures ATP from the chemical burning of food that happens in the mitochondrial matrix, but how the energy from the one process fuels the other was quite mysterious. And when this ATPase was studied more closely, the mysteries only piled up. It is an ion transporter, of H+, of all things, and is present in all cells, indeed conserved from the original ancestor of all life. It can break up ATP (giving it the "-ase" name) in the lab, when spinning freely with no H+ gradient, but in real life it is tightly stuck and oriented in the (inner) mitochondrial membrane, using the significant H+ gradient across the membrane as its fuel to run in the opposite direction, synthesizing ATP. And that is the heart of the story. The mitochondrion acts like a battery, in that its ATP production driven by H+ pumping is indirectly coupled the H+ production that is a product of glycolysis and the Krebs cycle elsewhere. In this way, the Krebs cycle can do its thing, at its own rate, and build up the fuel of high H+ outside without having to be physically linked to the ATP-producing enzyme complex. This concept also applies to chloroplasts and to all non-Eukaryotic cells. As weird as it seemed for cells to be spending their hard-earned fuels just pumping protons willy-nilly into the outside (i.e. into the ocean for single-celled organisms), the energetics work out. The cell (or mitochondrion for Eukaryotes) is a tiny battery.

That is the closely coupled system with an internal H+ generation system and a tightly sealed membrane. But suppose we are at an earlier stage, when energy didn't come from a well-worked out internal food-burning Krebs cycle, but from a kind of arbitrage on outside chemical gradients? Then having a sealed membrane would be counter-productive. The scenario the authors envision is where a proto-cell is lodged in the rocky vent matrix, with geochemical fluids passing on one side, at, say pH 10, and sea water on the other side at, say, pH 7. A three pH unit difference is very large; enough free energy to do a great deal of work, if harnessed to an ATPase that runs off the H+ gradient.

Author's model for their simulations, where the lower flow is the alkaline geochemical vent product, and the upper half is sea water, more or less. The wide H+ gradient between them provides energy to the green ATPase that produces ATP from ADP.

In this scenario, the membrane needs to be semi-permeable to allow all the ions to pass. Its only real role is to tether the ATPase, which the authors assume still conducts H+ orders of magnitude more readily (while generating ATP) than the semipermeable membrane does. The authors run numerous simulations of permeabilities, pH gradients, ATPase concentration, and of ancillary ion transporters. For example, as proton permeability declines, the usable gradient declines to zero, since even as the ATPase uses the protons coming in, they have no where to go back out of the cell, nor can OH- ions come in to neutralize them. (Run your own simulations using their software!)

"However, 1%–5% [surface area of the membrane covered by] ATPase in a leaky membrane (10−3 cm/s) retains a −ΔG of close to 20 kJ/mol. With 3–4 protons translocated per ATP synthesized, this gives a −ΔG for ATP hydrolysis of 60 to 80 kJ/mol, similar to modern cells and sufficient to drive intermediary biochemistry, including aminoacyl adenylation in protein synthesis."

As membrane permeability declines (colored cases), the energy available via the simple H+ gradient (Y-axis) drops to zero with time through the simulation (X-axis).

The next innovation is to introduce a Na+ / H+ antiporter, which is a protein in the membrane that exchanges sodium for protons 1:1. This is electronically and typically energetically neutral, but has dramatic effects on the ability of this protocell to manage its permeability and use the H+ gradient. Sodium has much greater difficulty getting across even a leaky membrane than the smaller H+. I should note that the authors assume that the ATPase can use Na+ as well as H+, which has some plausibility given the primitiveness of the system. They also assume that all their protein transactions only happen on the acid (sea water) side of the cell, which is much less plausible. Given the high H+ gradient from the acid side of the cell, it drags Na+ out, creating a supplementary gradient of high Na+ outside to inside, which the ATPase can use, in addition, to the protons, to generate ATP.

The net effect of all this is three-fold. It immediately raises the available energy of the H+ gradient by about 50%. It also sends the cell on a selective trajectory towards sealing its membrane, since the ion flows can now be managed entirely through the proteins in the membrane, and the H+ gradient yields more energy the more H+ are funneled through the ATPase. Lastly, it favors the generation of active H+ and Na+ pumps that expel these ions under some conditions, such as metabolic energy from light or from eating other life forms. Naturally this sets the stage for freeing the nascent cells from the vent ecosystem, if they can find another source for H+ gradients, i.e. food.  It also incidentally explains the universal property of our cells having very low Na+ concentrations, though our ionic levels otherwise approximate those of sea water.

When a Na+ / H+ antiporter (SPAP) is present, the energetics of the H+ gradient improve markedly in the author's simulation. But the effect is available only at highly alkaline conditions (graphs B and C).
"Crucially, SPAP [sodium / proton anti-porter] is also a necessary preadaptation for the active pumping of protons, and for decreasing membrane permeability towards modern values. Whereas pumping H+ in the absence of SPAP gives no sustained benefit in terms of −ΔG, the presence of SPAP in a leaky membrane allows pumping of H+ to pay dividends. −ΔG now markedly increases with decreasing permeability, for the first time giving a sustained selective advantage to higher levels of pumping and tighter membranes."
With the Na+ / H+ anti porter present, and with an additional pump (powered by some kind of novel metabolism) that exports H+ or Na+, dropping the permeability of the membrane (X-axis) pays consistent dividends (blue).

Needless to say, if tightening the permeability of the cell membrane was a later development after so many other critical mechanisms (genetic coding, enzyme production, leaky membrane maintenance, crude energy metabolism) had developed, then it stands to reason that the principal chemical components of the modern biological membrane might differ between forms of life that had already diverged into what became the two earliest domains of life.

The paper is a bit unclear, though reading the methods helps tremendously, supplying needed detail and organization. The overall scenario for the origin of life in these very dynamic and energy-rich settings is reasonably persuasive, and it is good to see people taking the next step to figure out how nascent cells might have gotten over some of the notable humps of the process.

"Our findings allow us to propose a new and tightly constrained bioenergetic route map leading from a leaky LUCA [last universal common ancestor] dependent on natural proton gradients, to the first archaea and bacteria with highly distinct ion-tight phospholipid membranes. These bioenergetic considerations give striking insights into the nature of LUCA, and the deep divergence between archaea and bacteria."