Crow's Pick For Wed Sept. 29~ Deep wattsr1--molecular, even!

[Crow]

My background in science is thin and spotty. Kinda like the film of unpleasant bacteria and filth that cloaks Thirsty Possum....

I can tell you all sorts of weird and useless details about possum and snake gestation but I only have a nodding aquaintance with the endosymbiont theory (yeah, that's it over there..I think) and that is all.

My hat's off to Roland for taking this subject and working it around so that even I can follow what he's saying.

Hi All,


In this thread, I hope, from time to time, to post essays dealing with molecular evolution - the study of evolution at the level of the molecule - be it the gene or the protein.

First off is an essay dealing with the evolution of a molecular machine. In this case the machine is made up of a group of proteins that sit within the inner membrane of the mitochondria and are involved in pulling proteins into the organelle from the surrounding cellular cytoplasm.

As is usual, this is a layman's perspective and all errors are mine.


So here goes.


THE EVOLUTION OF A MOLECULAR MACHINE

This essay deals with the proof in principle evolution of a tiny complex molecular machine housed in the membrane of mitochondria and is based on this piece of research:-

Abigail Clements, Dejan Bursac, Xenia Gatsos, Andrew J. Perry, Srgjan Civciristov, Nermin Celik, Vladimir A. Likic, Sebastian Poggio, Christine Jacobs-Wagner, Richard A. Strugnell, and Trevor Lithgow ”The reducible complexity of a mitochondrial molecular machine” PNAS 2009 106 (37) 15791-15795

Molecular machines drive essential biological processes, with the component parts of these machines each contributing a partial function or structural element. Mitochondria are organelles of eukaryotic cells, and depend for their biogenesis on a set of molecular machines for protein transport. How these molecular machines evolved is a fundamental question. Mitochondria were derived from an α-proteobacterial endosymbiont, and we identified in α-proteobacteria the component parts of a mitochondrial protein transport machine. In bacteria, the components are found in the inner membrane, topologically equivalent to the mitochondrial proteins. Although the bacterial proteins function in simple assemblies, relatively little mutation would be required to convert them to function as a protein transport machine. This analysis of protein transport provides a blueprint for the evolution of cellular machinery in general.

Introduction

Our cells are a kind of cell called “eucaryote”.

There are two kinds of cells - eucaryote cells and prokaryote cells:-

a. Eucaryotes have their DNA (genes etc) located within a membrane enclosed nucleus within the cell.

b. Procaryotes have no nucleus thus their DNA kind of sloshes around within the whole cell.

Animals, plants and fungi have eucaryotic cells.

Bacteria are prokaryotic cells. (Note, bacteria are single celled organisms).

Mitochondria are small organelles existing within animal and fungi eucaryotic cells, outside of the nucleus. (Plants have organelles called chloroplasts).

There is a lot of evidence showing that these organelles originated as bacteria that were engulfed by another (predatory?) bacterium some 1.5 billion years ago, and managed to avoid being digested.

Mitochondria now provide the cell with energy while the cell provides them with shelter and nourishment.

In order to take on this role as the energy provider for the eucaryotic cells in animals and fungi, the mitochrondria needed to develop protein transport machines within their membranes so that they could exchange proteins with the surrounding cellular cytoplasm. This is because bacteria do not have protein transport machines within their membranes, having the ability to make the required proteins themselves.

The research I am dealing with concerns itself with the evolution of one of these machines.

The introduction to the research paper

Clements et al. are currently investigating the functions and evolution of transport machines in mitochondria. The machines of interest in this research are those that lie in the membranes of mitochondria allowing an organelle to take molecules in from the surrounding cytoplasm or to put molecules out into the cytoplasm. The cytoplasm is the goo (in which the mitochondria reside with all kinds of other things) surrounding the cell’s nucleus.

This kind of research “provides an excellent, and perhaps unique, system to provide evidence that a sophisticated molecular machine can evolve from simple components, in a process strictly adhering to principles of Darwinian evolution.”. That is, they wanted to see if they could explain the origin of these protein transport machines via Darwinian evolution (as opposed to other forms of evolution.)

Mitochondria are essential to the cell that houses them because they provide the cell with its energy. Hence, mitochondria must be constantly reproduced so that the host cell can grow and divide.

They probably evolved by a process known as “endosymbiosis” - see my last post.

For the engulfed bacteria to become mitochondria, it required the evolution of protein transport machines (among other things) given that bacteria don’t need sophisticated transport machines in their membranes. (They make all their own proteins internally.)
The author’s proposal was as follows:-

a. First, simple core machines evolved from existing bacterial proteins that previously provided other functions to the bacterial cell. This evolution was via a process called cooption, something first proposed by Darwin. The notion is a common one, whereby a system or structure serving one function is switched to provide a different one, somewhat like a policeman directing you to temporarily abandon your trip home to deliver a new TV set and direct traffic while she attends to a nearby crash.

b. Secondly, in a step by step process (i.e. classical Darwinian), these core machines would have been refined by the addition of other molecules to enhance their efficiency.

Their proposal has 3 findings from other research that suggested their could find evidence in support of their hypothesis:-

a. Proteins found in other bacteria look somewhat like those found in mitochondria transport machines. That is, they could indeed have been coopted.

b. However, these proteins are not used in protein transport systems - underscoring the cooption idea.

c. Some ‘primitive’ organisms found today have protein transport systems that function with only one or a few component parts. This supports the idea of core machines which are functional and can be refined by the addition of more components. That is, in the early stages of evolution, it may only be required that something work. Improvements can be added over time afterwards.

The authors introduce the protein transport machinery. Here is a picture of the majority of the system:-

http://www.google.com/imgres?imgurl=...ed=0CCgQ9QEwAw

In the paper the researchers focus on the TIM23 machine and its evolution.

Their description of the mitochondrial protein import system

It consists of 4 machines embedded in the mitochondrial membrane - TOM and SAM in the outer membrane and TIM22 and TIM23 in the inner membrane. (The above picture only shows TOM and the two TIMs)

Each may be composed of up to 8 protein subunits.

Bacteria do not import proteins across their membranes (they have no need to) and TOM and TIM23 are not found in bacteria. So the question is, how did TIM23 arise? And what evidence could be produced to demonstrate any ideas concerning this?

TIM23 transports proteins across the inner mitochondrial membrane. (Generally both TOM and TIM23 act together to transport proteins across from the cytoplasm to the mitochondrial matrix space in one go. However they can act independently of each other. That is, TOM can pass a protein from the cytoplasm to the space between the inner and outer membranes and then later TIM23 can pick that protein up and pass it through to the matrix inside the mitochondria. TIM23 is also involved in inserting inner membrane proteins into the inner membrane itself).

TIM23 has been extensively studied in the past in yeast, humans and plants. The protein units that make it up are conserved across those groups of organisms. That is, they have not really changed much through evolutionary history.

And three of the TIM23 subunits have been found in enough different eucaryotic organisms to suggest that they, and TIM23 are common to all eucaryotes and their mitochondria.

These three subunits are called by the researchers:-

a. Tim23. This is a simple protein that forms a channel allowing a protein to pass from the space between the inner and outer membranes to the inner matrix of the mitochondria.

b. Tim44. This subunit is found on the inside face of the inner membrane, and it interacts with both Tim23 and the other subunit Hsp70. Hsp70 is a ‘motor’ that pulls the protein through the Tim23 channel and so Tim44 acts as an anchor to the Hsp70 motor.

c. Tim14/Pam18. This subunit interacts with several other protein subunits of the TIM23 system, stimulating energy production of the Hsp70 system thereby switching on and driving the motor in order to undertake the protein transport.

(The Hsp70 subunit belongs to the same family of proteins that chaperone the migrating protein both prior to its insertion into the outer membrane machine TOM, and its subsequent pulling through the inner membrane by its machine, TIM23)

The researchers then show that alphaproteobacteria have:-

a. a protein of the same family as Tim44 which functions in membrane quality control, and

b. a Tim14/Pam18 protein that functions in a completely different (but undefined) process.

Why the alphaproteobacteria? Well this is the class of bacteria thought to have given rise to mitochondria in the first place, all those 1.5 billion years ago. Here is a link describing alphaproteobacteria:-

http://en.wikipedia.org/wiki/Alphaproteobacteria

The idea is that the above two components, already existing and operational, but for different functions, came together with the LivH amino acid transporter to form the TIM23 complex. That is, these three components “would have provided ‘pre_adaptation’ to bacteria ahead of a need for protein import”.

So what is the evidence they produced, showing that various parts of TIM23 can be found in the extant bacteria whose ancestors long ago became the initiates of the mitochondria. Those parts still exist as functioning but independent proteins.


The research

Tim44, the protein that sits on the underside of the membrane linking Tim23 to the Hisp70 motor is found in all eucaryotes (and hence alphaproteobacteria). It varies in mass from 25kDa to 50kDA, where a “kDa” is a “kiloDalton” where a “Dalton” is a measure of atomic mass, specifically half the mass of a carbon 12 atom. So these proteins are roughly 12.5 thousand to 25 thousand carbon atoms in mass.
They all have a characteristic “N terminus” of variable mass from 3 to 30 kDA. They all have a characteristic “Tim44 domain” of mass around 20 kDa. (A protein N terminus is just one end of a protein where a free amino group (NH2) resides. By “free” it is meant that no other amino acid is connected to it. The other end of a protein has a “C terminus”. Here resides a fee carboxyl group (COOH)).

The “characteristic Tim44 domain” binds fatty organic molecules called lipids (which make up the main part of membranes) and independent research had shown that it probably binds to the underside of membranes in mitochodria, in part by electrostatic interactions and in part by partial penetration of the membrane:-

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC17703/
(This paper is online)

From studies of human and yeast Tim44, it’s found that the domain contains a deep hydrophobic (water repelling) pocket which may be crucial in lipid binding.

So do alphaproteobacteria (the precursors to mitochondria) contain homologues to Tim44? Yes, studies to date show that they do. They contain proteins, which the authors dub “TimA”, which share more than just skin deep sequence similarity with Tim44. TimA also seems to have that deep hydrophobic pocket, and structural modeling of TimA shows characteristics of the Tim44 family. For the modeling results you need to look up the supporting information - the first figure at:-

http://www.pnas.org/content/suppl/2009/08/26/0908264106.DCSupplemental/0908264106SI.pdf
(this supporting information is also online)

Tim14/Pam18 also has its homologue in alphaproteobacteria studied to date. This homologue they dub “TimB”. They describe TimB in relation to Caulobacter crescentus a tiny bacterium that lives in streams and lakes. It too maintains sequence similarity to Tim14/Pam18. It has the same kind of domain structure (where domains are specific regions of proteins having different functions). Like Tim14/Pam18 it has an N terminal segment that can hook across a membrane. It has a C terminal segment called a “J domain” containing 3 rather than the normal 4 helicies. It has a “signature” of amino acid residues, HPD-X-GCS, where “X” is a single residue that varies across species. The other residues don’t vary.

All this can be seen at the second figure in the above mentioned supporting information.

To date then, the researchers found what they think are homologues to the Tim44 and Tim14/Pam18 proteins that make up TIM23, homologues that look to reside in all alphaproteobacteria from which the mitochondria arose. These homologues have similarities that go more than skin deep. Not only were there there sequence similarities, but there were structural similarities enough to suggest that Tim44 and TimA are related and that Tim14/Pam18 and TimB are related.

The above reads somewhat like a shopping list and there is not much that can be done about this. It’s their evidence showing relatedness between the proteins making up the machine, TIM23 and various proteins existing in extant bacteria, but bacteria from which mitochondria originally arose.

So the researchers the had found two proteins, TimA and TimB that looked very much like Tim44 and Tim14/Pam18 of the TIM23 complex? Well now they wanted to see if the A and B had characteristics that closely matched some important properties of Tim44 and Tim14. This would make their ‘jump’ to the TIM23 system much easier to explain. One important characteristic of Tim44 and Tim14 is that they are proteins that are associated with membranes and that they have a particular orientation with respect to those membranes.

So they set to work with TimA and TimB and outer and inner membranes.

Antibodies are proteins that protect us by hooking onto foreign substances, allowing the immune system to then identify and neutralize them. They are proteins very specific to the substances they hook on to. Because of this specificity, antibodies are very useful for identifying proteins and if necessary helping to remove them from the environment they are in.

Clements et al made antibodies against TimA and TimB and developed a method of removing and separating the outer and inner membranes from C. crescentus. From here they were able to show that both proteins are associated with the inner membrane, not the outer membrane, just like Tim44 and Tim14 (and TIM23) are inner membrane complexes.

The next question was, which way did TimA and TimB point with respect to the inner membrane? Did they reside in it, under it or on top of it? Using a different technique this time, they were able to show that the proteins were not exposed on the outer side of the inner membrane, but rather that they were in the inner membrane and facing the inside - towards the cytoplasm enclosed by that membrane.

So already they were building up a picture of the two proteins having properties similar to Tim44 and Tim14.

Next, by using antibodies to precipitate out the proteins, they were able to show that TimA and TimB do not associate with each other on the inner membrane and from there to show that two other proteins named FtsH and HflC also associate with TimA, while no other proteins associate with TimB.

Mitochondria have proteins called AAA-proteins and FtsH belongs to that family* of proteins. Mitochondria also have proteins called prohibitins, and HflC belongs to that family. Neither of these two new proteins associate with Tim44 although clearly they do with TimA - the molecule that has similar characteristics to Tim44 (in sequence and that deep hydrophobic pocket which is probably crucial to lipid binding). What do these two molecules do? Well other research (cited in the paper) had shown that HflC and FtsH actually associate in cellular membranes with HflC regulating FtsH and forming a membrane garbage collection duo - they degrade unwanted membrane proteins.

And so the authors think that here, TimA forms an association with HflC and FtsH for a similar purpose.

At this stage, the evidence pointed to one of the important protein sub-units of TIM23, namely Tim44 having its origin in a TimA:HflC:FtsH membrane bound protein degradation system.

The researchers had already presented a lot of evidence linking TimB to Tim14 - the protein sequences, the number of helices, the end terminals and the special “HPD-X-GGS” sequence in the J domain.

What they do next is see how easy it would be to convert TimB into Tim14. When Tim14 activates it pulls in a neighboring Tim16 protein subunit by forming a hydrogen bond with it. (A hydrogen bond is a special kind of bond that occurs when a hydrogen atom is attracted to another atom by virtue of the fact that the other atom tends to keep the joint electron cloud to itself - leaving the hydrogen atom with a slight positive charge and the other atom with a slight negative charge.)

This hydrogen bond is formed via one amino acid residue with exists on both Tim14 and Tim16. The residue is asparagine and it is just at one location that the relevant hydrogen bond is made through the presence of this residue on both subunits.

This residue however, does not occur on TimB.

So Clements and her co-workers isolated TimB from an alphaproteobacterium named Parvularcula bemudensis. This critter has a TimB which appears to be most closely related to the Tim14 of yeast. (Remember, alphaproteobacterium is thought to have given rise to mitochondria and yeast is an organism having mitochondria.)

They then tweaked the TimB of P. bemudensis to give it a point mutation at position 139. They replaced its alanine residue with asparagine. Using genetic engineering techniques they placed the mutated TimB into yeast and showed that these cells were viable.

That is, potentially all that was required was one mutation to change TimB into a viable subunit of the TIM23 complex.

At this point the researchers concluded the experimental portion of their paper and began the discussion.

The discussion

This is the part where the authors pull the research together and explain what it all means, from their perspective.

The paper was to demonstrate the plausibility of the notion of a molecular machine complex being able to evolve in a step by step process from preexisting parts, where those parts themselves were viable entities undertaking different functions elsewhere in the cell hosting them.

The paper was also written to take a heavy swipe at intelligent design and its notion of “irreducible complexity”.

The process of using preexisting parts is known as co-option and was first mooted by Darwin 150 years ago. It wasn’t until the late 1970s that the idea was put on a firm footing by Fancois Jacob in his influential paper “Evolution and Tinkering”. Here is a link to an online print of that paper:-

http://www.gvsu.edu/cms3/assets/6D2549F6-ED41-142A-2D7251DEDEE796B4/Evolution%20and%20tinkering.pdf

Here is another online paper discussing the concept:-

http://www.springerlink.com/content/hlx8711165241v29/fulltext.pdf

(The abstract at the front is worth a read, if you only want a quick look.)

What the authors did in their research, which is the subject of my post, was to show that the most important parts of the mitochondrial TIM23 protein transport machine were probably already there inside the cell of the bacterium from which the mitochondria arose. And they showed that it was conceivable that these components could come together in a step wise fashion to form a primitive transport machine which could be subject to further refinement. Thus no fish to monkey in one go transitions were required.

1. The Tim23 transport channel (the ‘hole’ in the membrane for want of a better word) came from a LivH-type amino acid transporter. Here they cited other research demonstrating this. They note that the requirement here would have been for the transporter to have been modified slightly to accept peptides or strings of amino acids, rather than simply individual acids.

2. The Tim44 component of the complex derived from a newly discovered protein TimA. A few point mutations at a critical region would allow TimA to interact with LivH providing a docking point for a common protein Hsp70 and Hsp70 is the direct homologue of the Hsp70 protein which acts as the motor to pull a protein through the TIM23 complex.

3. The Tim14 component which helped drive the Hsp70 motor came from another newly discovered protein, TimB and that TimB only needed a minor modification to allow it to function in association with the TIM23 complex.

And not only were these components already lying around doing other things in the cell, but that they were in some ways primed to make the jump. For example, TimA and TimB were carrying out different functions to each other and to Tim44 and Tim14, but they were nevertheless inner membrane proteins.

The authors note that molecular machines have been described as irreducibly complex, and they ask if their machine, TIM23, could function with just a single component present and the other components missing.

Unfortunately no organism has been found with the LivH/Tim23 channel present and with Tim44 and Tim14 absent. However, the outer membrane machine, TOM, did provide them with their “proof of principle”.

It too should be irreducibly complex and is made up of three essential sub components - Tom40 (the channel) and Tom22 and Tom7. TOM has been very well studied in yeast and other fungi and there it has been found that Tom5 and Tom6 also make up the core machinery. Yet in one family of microorganisms, the microsporidia, the Tom22, Tom5, Tom6 and Tom7 proteins are missing, leaving only the Tom40 channel subunit to do the membrane transfers. And it does.

As the authors explain:-

The simplified TOM complex in microsporidians provides an excellent example of how the first, simple protein import machines might have functioned.

Anyway, the character count is up, so its time to stop.




Regards, Roland

[lee_merrill]

Good explanation! Lee is thinking about cranking up his RolandDisputer (model ZRX-21) to make a point or two in reply.

Blessings,
Lee