Talmudic Question #24

Why didn't mitochondria and chloroplasts cede all of their DNA to the nucleus, as did most hydrogenosomes and certain other organelles? 

Comments

[Dorothy Lewis]

I dont understand why we are to answer the question about invertebrates and innate immune systems under this question.

However, to generate immune cells is very wasteful--most of the T and B cells that are produced die--so there is likely an evolutionary advantage to undergo such a wasteful process. And that advantage is memory, we and other vertebrates tend to reach reproductive age after a relatively long adaptive, growing period and we need to remember what we have already dealt with. Invertebrates can live a long time--ie look at that lobster!, but to reach reproductive age is not likely to require as much time for defense against pathogens. The first evidence for DNA rearrangement is found as the cartilagenous fish change to the bony fishes, so that may be the time at which the benefits of memory responses for the host kicked it evolutionarily speaking.

[elio]

Jesse,

Many thanks for such an erudite treatment. It made the question even more interesting, pointing out so many possibilities. It sounds like it will be around for a while.

Elio

[Jesse Woodson]

Hello,

I am an ex-microbiologist now working with the flowering plant Arabidopsis as a postdoc at the Salk Institute and have recently been interested in this very question.

So far, there has been no single satisfactory explanation to why mitochondria and chloroplasts contain their own genomes, but since hundreds of different proteins encoded in the nucleus are required to maintain a genetic system in an organelle encoding between 3-200 proteins, it seems reasonable to conclude that complete transfer of the organelle genome is either impossible, or that there is a selective advantage to it. The “it has not been long enough” hypothesis seems unlikely to me as organelle to nucleus DNA transfer is a constant process that occurs at a very high frequency and there is conservation among exactly which organelle genes are left in the mitochondria and chloroplasts. This suggests that the organelle genome we see today has probably been relatively constant for a very long time.

So maybe moving certain genes is impossible. The genes found in organelles usually includes those for the genetic system (ribosome subunits, rRNA genes, tRNA genes, and a bacteria-like polymerase in chloroplasts) and those encoding the proteins for photosynthesis and the electron transport chain in chloroplasts and mitochondria respectively. Many of these are integral membrane proteins that are very hydrophobic and it is possible that they cannot be efficiently imported from the cytoplasm into the organelles or that their presence in the cytoplasm may be toxic. This is contradicted by experiments that have successfully imported chloroplast-encoded photosynthetic proteins by inserting their genes in the nucleus and encoding a fused chloroplast-targeting peptide. Also, this hypothesis does not explain why only certain proteins are encoded in organelles as the most hydrophobic of the photosynthesis proteins, the chlorophyll-binding proteins, are encoded in the nucleus. Lastly, the large subunit or RUBISCO is always encoded in the chloroplast and it is completely soluble.

One of the other posts mentioned the CORR (co-location by redox regulation) hypothesis, which suggests that encoding photosynthetic or respiratory proteins on site allows for rapid regulation. Moreover, the organelle does not have to send a signal to the nucleus for more protein. This would not only be slower, but the nucleus would have no way of knowing which organelle made the request if there are several mitochondria or chloroplasts. While this hypothesis may sound more likely than the previous one, there is not much evidence supporting it. There are only two cases I know of where the redox state of an organelle regulates gene expression. In one case, the redox state of land plant chloroplasts, which fluctuates due to light quality/quantity regulates the transcription of the chloroplast-encoded light harvesting centers to increase photosynthetic efficiency. In the second example, nuclear-encoded post-transcriptional regulators in green algae respond to the redox state of the organelle and bind to or release mRNA transcripts. There are no examples or this type of regulation in mitochondria. In the end, both of these hypotheses may help explain why organelle genomes are retained as they are not mutually exclusive.

If you look through the literature three are still other hypothesis that people have put forward, but many are falling out of favor as we sequence more organelle genomes. For instance, one early hypothesis suggested that the alternate genetic code in mitochondria and the rare codon usage in chloroplasts made it more difficult to move genes to the nucleus. This, however, cannot explain why only certain genes have been retained and now we know that these alternative genetic systems evolved secondarily to the organelle. Thus, this may actually be a clever mechanism to ensure genes are NOT lost the nucleus.

One last thought. Some mitochondria that do not perform respiration have lost their genomes. All chloroplasts, however, have genomes. Even those that do not do photosynthesis. Because heme biosynthesis occurs in chloroplasts (in organisms that have chloroplasts, otherwise it occurs in the mitochondria) and requires the chloroplast tRNA-GLU as the starting precursor (which is always encoded in the chloroplast for unknown reasons), it has recently been hypothesized that chloroplasts must retain genomes in order to synthesize heme for the rest of the cell.

Thank you for reading this. I hope it has piqued your interest in eukaryotic biology.

[elio]

Ole Kristian,

The paper by Race, Herrmann, and Martin that you refer to is indeed most enlightening. Thanks for pointing it out.

Elio

[Ole Kristian Tørresen]

I've just read an article about just this. The conclusion in that was: "Redox control of organelle gene expression would permit a rapid response, via a short, simple signalling pathway, to changes in the physical enviroment, allowing the cell (and the organelle) to protect itself from harm stemming from its own electron transport chain." (Trends in Genetics. Volume 15, Issue 9, 1 September 1999, Pages 364-370) http://dx.doi.org/10.1016/S0168-9525(99)01766-7

[Mark O. Martin]

Paul, you rock the (microbial) house. Elio, you and Merry keep producing great questions!

I was just teaching about establishments of symbioses, and brought up the work of Jeon and his coworkers---how a bacterium can "integrate" into a macrobe very quickly. I don't think anyone continues to work in this system, but I would LOVE to know if any of the bacterial genome has migrated into the amoebae's nuclei, and vice versa. Can we set up conditions in the laboratory to influence that process?

A perfect model? Not at all. But a place to start....

[elio]

Paul, thanks for your insightful observations. They should stimulate some responses. This is not my field, but the proportion of the mitochondrial genome that transferred to the nucleus is variable. See http://www.blackwell-synergy.com/doi/pdf/10.1111/j.1469-8137.2006.01821.x

[Paul Orwin]

So, based on the question and a quick read of this reference (Trends Microbiol. 2005 Oct;13(10):457-9. The missing link between hydrogenosomes and mitochondria. Martin W) It would appear that hydrogenosomes are anaerobic mitochondria, based on functional and structural analysis as well as DNA from the sole known genome carrying example (from Nyctotherus) that retain DNA making them unmistakably mitochondrial in origin. The H2 production feeds anaerobic methanogens (Archaea, see TQ previous!).
And thus I learn something new from reading this blog (hooray!)
So, the relatively straightforward answer is that the electron transport - Proton gradient - ATP synthase mechanism requires the genes to be transcribed and translated locally (at least some of them).

It is interesting to consider why the respiratory chain genes couldn't be moved to the nuclear DNA, but is it just a case of leaving well enough alone? Since the cell can't survive very well without making ATP from a proton gradient, perhaps they just don't mess with it?
One potential (selection based)answer is that cells need to have "maintenance energy", which perhaps can't wait until the parts are synthesized and transported from the nucleus to the ER to the mitochondrion or chloroplast. So the cells with ready made ready to go power plants do better than the ones that have to make them work by expressing the genes anew from the nucleus. Of course, one can imagine a scenario where the fully formed Mitochondrion is passed on even if the gene expression happens only in the nucleus.

Another possibility is that it will be lost to the nucleus, given enough time. It might be the case that there is no advantage to one state over the other, but allele frequencies are such that translocation events get lost in the reproductive gene shuffle. If this was the case, we might expect to find parts of the mitochondrial genome present in nuclear DNA - is that the case?

A final possibility is the molecular biology based case - the DNA/RNA level differences are now too great between miDNA and nucDNA, so the control of gene expression gap is just too wide, so now (1.5 by after the initial endosymbiosis) this just can't happen. I don't know if that sort of thing has been tested - surely you can express eukaryotic genes (sans introns) in bacteria, but how well can you express bacterial genes in nuclear DNA? Especially without providing a strong eukaryotic promoter recombinantly? I'm guessing not well at all! So that seems like a pretty high hurdle. Doesn't answer the question of why it didn't happen 1.5 bya, though.

One final thought - it does seem that these possibilities yield testable hypotheses, that can be looked at genomically. I haven't scoured the literature, but I wonder if there is any info in eukaryotic genomes that can shed light?