Its just a hypothesis we havn't seen evolution in real time. Scientist postulates these hypothesis in theory and what you are saying about the endosymbiont is accepted one and most probably true with new techniques of genoimic science, and all evidence leading to the acceptance of the theory and in my view they dont have to evolve or very will evolve at slow rate as their role are very fundamental and basic, but still the evolution will work as natures law leading to modification only IF NEEDED.

Well they've got bacterial ribosomes and codon usage, and a circular genome. Plus a lot of nuclear genes believed to be of mitochondrial genome origin are actually targeted to the mitos by crossing wholly into the matrix and then being re-exported to the inner membrane and IMS.

Finding an explanation that ISN'T basically the endosymbiont hypothesis is pretty tricky. 

As to evolution, they're basically stripped-down, streamlined organelles by this point, so while mutations still occur at more or less the same rate as they ever do (mitoDNA takes a pounding anyway, being that close to all that oxidative potential) there's very little that mutation can do that's actually beneficial, and quite a lot that is actively detrimental. We're very unlikely to see any further examples of gene transfer from mtDNA to the nucleus, either. The genes left on the mitogenome are presumably only those that must be produced in situ: either coding for things that cannot be targeted to the matrix (ribosomes, tRNAs) or that are too huge, complicated and unstable to be successfully imported (complex 1, ATP synthase etc). 

Well there are some mitochondria (I'm referring to mitochondria in all types of species, not just animals) that have linear genomes, a somewhat normal codon usage) and even organisms that don't have mitochondria (aside from prokaryotes). I would say the biggest evidence for the endosymbiont theory is that mito can't be produced de novo by the cell. However, many people point to the similar size/shape and membrane composition as bacteria, but in fact the size/shape are dramatically different, given that mitos are tubular and irregularly shaped and have major differences in membrane composition (lack of peptidoglycan, for example). While the endosymbiont theory is a good one, I feel the evidence for it is somewhat misinterpreted at times. Like how mitos replicate at the D-loop, and have multiple copies of DNA. This is not how bacterial genomes replicate at all and don't have multiple copies. Then, there are the phage similarities in mitos that you don't typically have in bacteria, haha. I feel the jury is still out, but many "experts" seem to have closed the case without explanation for the discrepancies

Several bacterial division mutants exhibit tubular structure and multiple genomes (you can stop them splitting into two, but they'll keep elongating and replicating their DNA), so "tubular and multigenomed" is not quite the zinger it might appear to be. Hell, eukaryotes can do this too. Division of the cell and replication of the genome usually synchronise (obviously), but they're not necessarily intrinsically linked processes.

It's also a lot easier to rationalise loss of features (peptidoglycan, for instance) than it is to explain gain of features. Organisms lose unnecessary functionality all the time, but gain of entirely novel function is rarer. And gain of a ribosome sequence and structure that is functionally prokaryotic is very hard to explain by any other means than "it came from a prokaryote".

The idea that a bacterial ancestor become an endosymbiont (especially given that invasion of eukaryotic cells is absolutely a function many bacteria possess) and then lost redundant or unnecessary features...is just a very neat explanation that ties up an awful lot of questions.

I'd be interested in reading further into it, though: do you have a reference for the phage similarities (and exactly what do you mean by this, anyway?). I'd imagine phage-like qualities would support the endosymbiont theory, given that phages are basically bacteria-specific viruses.

That's a fair point. I suppose the question would then be "is there evidence of a division mutation?"

That's true. I feel as scientists, though, we owe it to ourselves to find evidence rather than just assuming it lost these features. One point that people bring up with mtDNA movement to the nucleus, is that the nucleus has a different amino acid code for its products, and many gene products appear to be imported into the mitochondria with sequences. If these genes came from mitochondria...how did they pick up these sequences and how did the difference in coding happen? I've always thought that being at least partially originated from a plasmid helped explain many "weird" things about mitochondria.

So, the mito RNAP appears to be phage-like. You are correct, this is interpreted as being more evidence of bacterial origin given that they are bacteria-specific viruses.

Sure. I've attached a great article by Mike Gray that reviews the topic of mito evolution in-depth.

Thanks for that! Nice review: very interesting read.

On balance, that paper seems to be less concerned with whether the endosymbiont hypothesis is correct or not, rather taking that essentially as read (I would agree), and arguing over exactly what kind of endosymbiosis best explains the evidence. Not "are mitochondria of pseudo alpha-proteobacterial origin" but "how much of mitochondrial behaviour/function/etc is exclusively explained by their alpha-proteobacterial origin".

Which is a substantially different question. I'd be tempted to suggest that the events in question happened so long ago, in an environment so alien to current conditions, with organisms hugely less evolved, that distinctions of this sort rather lose any real meaning. All evidence suggests that every single mitochodriate organism (and quite a few essentially amitochondriate ones) can trace their ancestry back to a single mitochondriate progenitor, but beyond that we're sort of bottlenecked. 

Acquisition of a mitochondrion-like structure may have happened substantially before this hypothetical ancestor, and have been subject to all manner of changes, variations and evolutionary branchings, but we're unable to say for sure since everything we can study is descended from a single successful organism. We have no real idea how common horizontal gene transfer was back then, for instance. We can't even say "this only happened once", we can only say "it only happened once to the ancestor of all extant mitochondriate organisms" (though also, see chloroplasts).

It's important to keep in mind how different things were back when this was happening (everything was unicellular, for instance, and the world was a horrible anaerobic soup), so things like different codon usage, plasmid contributions, and gene transfer between mtDNA and host genome have to be viewed in that context. An awful lot can change in that amount of time, and the changes won't always leave detectable remains.

Still, very thought-provoking question.

Also, in case it didn't come across, the above is a very flowery way of me saying "I don't think I know enough to have a firmer opinion".

;)

Yes, I am aware of that article. He proposed about a decade later that phage elements  were adapted very early in the "engulfment" stage (see article).

You are correct about the codons. Plants and even physarum have a universal genetic code, so it's not hard to imagine loss of functions changing the code I suppose. I am aware of mountain of evidence supporting the theory (like mtio rRNA genes are in the same order as the bacterial rRNA operons), but am also aware of the evidence not supporting it, too. Membrane composition can vary, which is thought to have happened over time. Many mito targeting signals have magically appeared, the size and 3D shape are quite different, and then the age-old question of how proteobacterial genes can move and be expressed in a eukaryotic nucleus. I am not saying I don't support the endosymbiont theory, but I feel that the current body of evidence isn't fully supported by what we know, particularly amongst all mitochondria across the phylogenetic tree

http://www.sciencedirect.com/science/article/pii/S0168952505003367