Field of Science

Bacterial evolution: negative to positive?

ResearchBlogging.orgThe most commonly used distinction in bacterial populations is that of Gram negative and Gram positive bacteria. They are named after the method used to distinguish them: the Gram stain (developed by Hans Christian Gram). Gram positive bacteria have larger peptidoglycan cell walls, and therefore retain the crystal violet stain, whereas Gram negative bacteria have two membranes with a thin peptidoglycan wall between them, do not retain the crystal violet stain, and pick up the safranin counter-stain. The practical upshot of this is that you end up squinting at small blotched shapes under the microscope, trying to work out whether they look more pink or purple:

Gram +ve on the left, Gram -ve on the right.

As well as hinting that Mr Gram was one of those people who knows what shade 'fuchsia' is, the Gram stain is also one of the most important ways of telling what kind of bacteria you're dealing with. Despite being seemingly arbitrary, the composition of the cell wall plays a major role in determining behaviour. Gram negative bacteria (small cell wall, two cell membranes, see the picture below) tend to be motile, opportunistic, and able to colonise a wider range of environments. Gram positives on the other hand (big cell wall) are not so motile, but tend to have a huge range of excretory proteins to make up for this; almost all known antibiotics come from Gram positive bacteria.
Again, Gram +ve on the left, -ve on the right. Image from soil microbiology webpage

One thing that I've never really considered before is which one of them evolved from which. I haven't done much taxonomy, and the only time I really covered bacteria (unrelated to lab work) was in my Pathology course, which didn't seem too concerned about where different types of bacteria had come from, only what they were currently up to. The few times I did vaguely think about this though, I would have gone for the positive to negative direction. After all, surely you start with one cell membrane, and move on to two.

I recently came across a paper that came to the complete opposite conclusion, and therefore was too interesting not to read. The thing about bacterial taxonomy is that a lot of the major changes to morphology took place in Deep Time, and bacteria leave precious few fossils. Bacteria (and archaea...) had somewhere in the region of over one billion years to evolve before eukaryotic-things even started to be considered. That's a lot of time to try and sort out. To put that into context, one billion years ago from now things were just about starting to think about going multicellular. No dinosaurs, no plants even; the most complex form of life was something resembling a sofa cushion.

So how to sort out what was going on in that billion years or so? There are four main ways of going about it:
  • Paleontological evidence. Bacteria don't form a huge number of fossils, but they can occasionally leave some physical evidence of their presence. For example, bacteria that eat iron will leave behind little fossilised iron cases; those that eat rocks can leave microscopic drilling holes. These provide temporal evidence for changes in structure and metabolism.
  • Transition analysis. This is used to polarise major changes by turning them into a simple before-or-after question, and uses comparative, developmental, and selective arguments for determining answers. For example: did legs or wings develop first? Or, in bacterial cases: Which came first, Gram negative or Gram positive?
  • Congruence testing. This searches for similarities across whole evolutionary trees, enabling loss or gain of evolutionary abilities (wings, feathers, second membranes etc) to be identified and polarised. As this is a comparison of many species, it allows potential mistakes from the arguments made in transition analysis to be found.
  • Sequence trees. Sequence trees are ... problematic, but at the same time indispensably useful. They are formed by taking DNA sequences from a range of organisms and then using algorithms to tell the 'relatedness' between sequences and using these 'relatedness' levels to make evolutionary trees. They tend to be biased towards your sample distribution, undirectional, unable to properly account for generation times, and go somewhat screwy when you try to introduce horizontal gene transfer. Nevertheless they were instrumental data in showing that archaea and bacteria are two very distinct super-kingdoms (and I will freely admit that most of my distrust for them occurs because I can't get the damn things to work whenever I try them)
So...using these techniques can we get a clearer idea what was happening with bacterial membranes during those 1 billion-odd years before the arrival of eukaryotes? On the face of it; positive-to-negative seems to make more sense: start with one membrane, gain a second, possibly by gene duplication.

However like many evolutionary stories, that one falls apart a little when closer examined. Because Gram positive bacteria are not simply 'one cell membrane' they also have a massive cell wall surrounding them. Developing a second cell membrane on top of that seems absurd. And then why would the cell wall shrink? And how would anything get through this suddenly developed cell membrane. Transport proteins for the outer membrane tend to form a protein structure called a beta-sheet, while those for the inner membrane form an alpha-helix. That's a whole new system of protein folding that has to evolve pretty quickly, because otherwise the bacteria will starve, nothing can get through its outer membrane (which is balancing precariously on top of the huge cell wall...)

In view of this, the schematic seen on the right starts to make a little more sense (figure taken from the reference below). 'Murein' means 'peptidoglycan cell wall' and the cytoplasm denotes the inside of the cell. In this scenario, the double-membraned proto-bacteria (which has spend the last half-a-billion years or so evolving a well adjusted double membrane system) suddenly looses the outer membrane. A very simple genetic change would lead to a massively overgrown cell wall, which would rip the outer membrane away. The cell looses all it's outer membrane porins, and signal systems, but in return gains a highly protective cell wall, which potentially allows it to survive in different niches. How these aspects are lost genetically is another matter, and the paper rather hand-waves away by saying that unused genes tend to get lost eventually. Which is true in bacteria, they have such a small genome they don't want it getting filled up with unnecessary genes, but I have a feeling genes tend to leave something behind. Even so, the question of where the now-unnecessary genes go is possibly one of the weaker parts of this arguments (to my untrained student eyes at least.).

One thing that would really support this hypothesis would be to show that Gram positive bacteria formed a 'mono-clade' i.e came from a single universal common ancestor. Unfortunately this data is proving hard to pin down, not helped by the bacterial trick of swapping DNA around with all and sundry. Another confounding factor is the sheer space of time. Trying to determine whether a range of different modern bacteria all came from the same blob several million years ago is a daunting task. You can sort of get RNA sequence trees that support the mono-cladal Gram positives, but only if you close one eye and squint, which is not generally accepted scientific practise.

I don't think I'll ever end up going into taxonomy, even of bacteria. but it does produce fascinating ways to look at the world; how it changed, how it evolved, and how it finally turned into the way it is now. Orwell wrote, fairly famously, "He who controls the past commands the future", and when you're trying to figure out how bacterial resistance works, and preferably how to stop them getting it, that phrase takes on a whole new meaning beyond the political.

(It's not a perfect quote for this post. "Understands the past" would work better. But I'm not quite pretentious enough to go trawling through the quote archives to find something better. Any suggestions would be appreciated :p )


Cavalier-Smith T (2006). Rooting the tree of life by transition analyses. Biology direct, 1 PMID: 16834776


Lucas Brouwers said...

Very interesting hypothesis! If it were possible to artificially recreate the mutation(s) that lead to this sudden outgrowth of the inner cell wall in Gram-negative bacteria (and see if the little fellow would be able to survive a little), this would really lend credibility to this idea.
I'm not that well versed into bacterial biochemistry, but if it indeed was a relatively simple event, multiple origins of 'gram positivism' could also be considered, explaining the funny taxonomic distribution of Gram (+)'s..

Psi Wavefunction said...

Great post, nice to hear about this stuff from an actual prokaryotic microbiologist. I've always been curious about bacteria, but our bacteriology courses are taught by the Department of Microbiology and IMMUNOLOGY, so unfortunately we actually don't have ANY classes on microbiology from the microbes' perspective (I don't particularly care how they interact with the human body, tbh). I've looked, and there's nothing: ALL are medical, and ALL are cram-for-MCAT-oriented. Grrr.

That said, TC-S does stress what, in the eyes of a cell biologist, is a very crucial and oft-missed point: genes are not the only basis for cell structure/organisation. There IS extranuclear inheritance (and thus, evolution), and perhaps it plays a larger role than molecular biologists like to think. Of course, evolution is mostly studied by ecologists, population geneticists and molecular biologists, so cells get mostly ignored in the process. And yes, I've got serious beef with that, because I think cytoplasmic and cortical inheritance/evolution may be a very important factor in evolution, and help resolves issues where genes lie powerless. After all, once again, an organism is more than just its genome.

In his 2001 J Mol Evol "obcell" paper, Tom starts off with: "Compared with nucleic acids and proteins, lipids must seem third-class citizens of the molecular world [...]" And has a very good point: of course you hear about endless arguments over RNA, DNA and protein evolution, with lipids casually thrown in as something trivial that just happened, somehow. Membranes are boring to geneticists.

Thus, the loss, gain or modification of some membrane structure does not actually require genetic change to be heritable (and thus conserved). In fact, it can happen entirely on its own, and add a sprinkle of adaptive advantage to it, and it will actually become fixed in a population, much like a gene would. There's nothing inherently 'magical' about DNA that makes it the only structure capable of evolution. It's a wonderful substrate for that, perhaps the best possible, but not the only one.

Sure a membrane gain or loss is likely to cause changes in the genome. For example, if you need special pore structures in the outer membrane to allow stuff to go through, you'd no longer need them if you lose said membrane, PROVIDED THAT those genes aren't involved elsewhere. My thus far very limited experience working with cellular molecular biology already shows that it is very rare for a protein to ONLY function in a single pathway. Cellular pathways are a fucking mess, you can trust me on that. "My" gene seems to sleep around with a rather ridiculous number of various pathways and processes, and is thus unlikely to go away if you remove on or even several.

I highly recommend Stoltzfus 1999 Mol Biol Evol and Lynch 2007 PNAS for some background on neutral evolution of complexity and genetic networks. It helps explain why gene interactions are such a fucking MESS. *glances at own research notes* *whimper*

Psi Wavefunction said...

Anyway, you would expect that perhaps there may be some gene loss events universal to 'posibacteria' (single membraned proks), perhaps related to proteins hanging out on the outer membrane in negibacteria. Or perhaps some extreme divergence. You're right that it is a huge problem with the long duration, high rate and more stringent selection in the world of prokaryotic evolution -- it's harder and harder to make the trees speak. But this is where it becomes important to weigh in multiple factors, not just trees -- and molecular biologists generally suck at that.

Before I get lynched, some are good at recognising that, including a few in our department. *looks around nervously* Actually, the ones I have in mind do involve a lot of morphology work, so they're safe. *phew* But still, most of the field gets by on intense sequence worship... which is why Tom comes out with his 'hyperbolic assertions', and tries to piss them off into working on his hypotheses. Very effective strategy, in my opinion =P

Oh, and important detail: Tom thinks the early prok cenancestor was double membraned, as a result of a fusion of two 'obcells' to form a double-membraned first cell (2001 JME). In fact, some guy liked that in a Nature Review, if I can find it...ah, here (free access), so Tom's not alone in that insanity.

Would indeed be interesting if posibacteria are polyphyletic. I forget what Tom thinks about that, but he may well consider some to be independently outer-membrane-less...

Sorry for typing up a mini Annual Review here... (but I really LOVE!!! this stuff!)


PS: "but only if you close one eye and squint, which is not generally accepted scientific practise."

Oh come on, you can be honest here =P

Ok it was whining about length; how about we carry this out by email? Mine's up on my blog/profile...

Captain Skellett said...

I think the Orwell quote worked. 1984 explains everything, including bacterial heritage. I've never wondered which came first, the negative or the positive, but the theory they put together makes more sense then the spontaneous second membrane and smaller wall theory.

I think we should take a whole bunch of bacteria and try to recreate the conditions they would have evolved in, a la Miller's Experiment. It can't be THAT hard, right?

Lab Rat said...

Thanks so much for all the comments guys!

@Lucas: This is in the realms of pure speculation, but i get the feeling that while is was a relatively simple event *genetically* survival of the resulting organism would have been quite difficult. it would have to have been in a very specialised niche.

That would be a very awesome experiment to try, but Gram -ve's have done quite a lot of evolving since then, and the outer membrane might now be rather more vital than it used to be (i'm thinking particularly of flagella here). probably worth an attempt nevertheless.

@Psi: "genes are not the only basis for cell structure/organisation" I am TOTALLY with you on this point! Unfortunatly for (mostly historical) reasons heretability seems to be mostly studied by geneticists, which means epigenetics is only just beginning to be appreciated. I'd never thought of it before but for bacteria cell wall/membrane componant inheritance would be vital and when the cell splits in half the two daughter cells will both get different bits (what happens to the occasionally-mentioned 'nanobrain' at this point!)

I'll check out the Stoltzfus paper, and the nature review sounds cool :)

@Skellet: I think one of the main problems with carryin g out that experiment is a) we'd need a researcher with a billion years worth of funding and b) noone is quite sure what those conditions are! At some point in the whole mess the oxygen explosion happened as well which must have messed everything up for a little bit :)

Shawn said...

Wow! It is wonderful. Thanks for sharing your precious information. We do work for Pest control. We also believe in pest control. You can visit us for more information.
Pest Control Services NJ

Charles Kennett said...

There was an interesting Nature article by James Lake of UCLA that gram negatives may have evolved from endosymbiosis of an Actinobacterium and a Clostridium. It's only one paper, but very interesting and compelling. We know that for eukaryotes symbiosis was very important (i.e. prokaryotic homologues of mitochondria and chloroplasts) for their evolution, so why not prokaryotes? At least I don't think we can ignore the possibility.

Lake, JA. 2009. Evidence for an early prokaryotic endosymbiosis. Nature 460:20 p.967-71.

Anonymous said...

Genetic information and analysis might actually support this or positive to negative. The thick peptidoglycan can be as a result of the hash environment that these bacteria had to live in millions of years ago. Then a bacteria possibly could have swallowed another and reduced its peptidoglycan layer as a result of the presence of another membrane. Just a theory though. Good stuff either way

Lab Rat said...

@Charles: Thanks for the paper! Well definitely look it up :)

@anon: I'd come across various theories for bacterial membrane evolution, this was just the first I'd ever heard suggesting the negative to positive direction, which was why I found it so interesting.