According to the endosymbiotic theory, eukaryotic organelles (mitochondria, chloroplasts) are the remains of certain bacteria that established intimate associations with eukaryote ancestors. This theory is widely supported by biochemical, genetic and proteomic evidences. A take-home message may be: a bacterium became an endosymbiont, then degenerated, and voilà, turned into an organelle. But that's an oversimplification. A full story should include an active participation of the host, some horizontal gene transfer from different microbial sources and, most likely, the involvement of other endosymbiotic partners that became extinct long time ago.
The origin of mitochondria, through the association between a eukaryotic ancestor and an alpha-proteobacterium (perhaps similar to Rickettsia) was a crucial, early event: thusfar, all eukaryotes examined possess mitochondria or mitochondria-related compartments, or at least some genes derived from such an organelle. About 1.5 billion years ago, a cyanobacterium joined the partnership to give rise to plastids in a common ancestor of red and green algae, land plants and glaucophytes (together the Archaeplastida). Although the identity of the cyanobacterial symbiont is far from clear, recent studies support the hypothesis that it might be similar to contemporary Nostoc or Anabaena. Many of these filamentous, nitrogen-fixing cyanobacteria are endosymbionts, which generally provide nitrogen to their hosts (this agrees with suggestions that nitrogen fixed by the endosymbiont might have played an important role during the origin of plastids). These organelles were later spread through secondary endosymbioses, that is, plastid-containing eukaryotes were engulfed by other eukaryotes; this process has occurred several times and has produced an astonishing diversity of photosynthetic organisms.
For a bacterium to become a permanent endosymbiont and, eventually, an organelle, some new functions had to be developed, or re-adapted from old ones. Firstly, a controlled exchange of metabolites ("food", waste products, ions) had to be established between the host cell and the endosymbiont. Additionally, the two partners had somehow to coordinate cell division, to assure that at least one endosymbiont cell was included in each daughter cell when the host reproduced.
Gene loss and gene transfer to the nucleus
On the other side, long-term endosymbionts could get rid of many functions that were required by free-living bacteria, and many unnecessary genes were lost. In general, genomes of mitochondria and plastids are about 10-20 times smaller than genomes of present-day Rickettsia and free-living cyanobacteria, respectively. Actually, the number of protein-coding genes in organellar genomes is very small (3-70 for mitochondria, 50-200 for chloroplasts). Does this mean that only 3 (or 50) proteins are needed for a mitochondrion (or chloroplast) to function? Of course not — these organelles are complicated molecular machines, requiring the coordinated action of several hundreds or thousands of different proteins. However, most of these proteins are not coded in the organelle genome but in the nuclear DNA. That is, many bacterial genes were lost, and most of the remaining genes were transferred to the host genome. Apparently, the nucleus is a safer environment for DNA: deleterious mutations may accumulate at a faster pace in organelles, due to a higher mutation rate caused by oxygen free-radicals, in addition to the non-recombining nature and small population size of organelle genomes. Then, it is not clear why mitochondria and chloroplasts retain a genome at all. Perhaps each type of organelle needs to keep a minimum set of key genes that are subject to a fast and tight regulation depending on local (organelle) conditions; or the encoded proteins cannot be efficiently transported into the organelle. On the other hand, some organelles do exist that appear to lack a genome — the hydrogenosomes, which might be derived from mitochondria.
Once the endosymbiont genes landed on the nuclear genome, they had to suffer some modifications to be properly expressed in the new location. Additionally, nuclear genes are, by default, translated into proteins in the host cytoplasm — if these proteins need to function in the endosymbiont cell, they have to be properly targeted and directed to the right compartment. This required the development, or re-adaptation, of an efficient system for protein targeting and translocation.
With a little help from some friends
Interestingly, both cellular partners were active players in organelle establishment: all key functions required for endosymbiosis were driven by combinations of endosymbiont- and host-derived proteins. Organelle proteomes are a potpourri of proteins with diverse origins: a surprisingly large pool was derived from the host, while other proteins were provided by multiple symbionts and/or horizontal gene transfer. A study of the yeast mitochondrial proteome showed a predominant eukaryotic origin (about 50%), and only about 15% of the proteins came from alpha-proteobacteria; the remainder had other prokaryotic origin. In the case of the plastid proteome, most proteins (1,000-1,500) are derived from cyanobacteria, and several hundreds are of eukaryotic origin.
Chlamydiae and the origin of plastids
Recent sudies have shed new light on the origin of plastids, suggesting the involvement of another group of bacteria, the chlamydiae. Present-day chlamydiae live as obligate intracellular parasites in eukaryotic cells. Many chlamydiae are pathogens of humans and animals, while others are found as endosymbionts in amoebae and insects; thusfar, they have not been found in plants or other plastid-containing eukaryotes. Nevertheless, it was known for some time that an unexpected number of chlamydial genes were most similar to plant homologs, most of which contained a plastid-targeting signal. This observation prompted several hypotheses, including an ancestral evolutionary relationship between chlamydiae and the cyanobacterial donor of the plastid, and horizontal gene transfer between the two groups. It was also suggested that plants might have acquired these genes through intermediate vectors such as infected insects.
A phylogenomic analysis of 17 Archaeplastida species has shown that at least 55 genes were transferred between chlamydiae and primary photosynthetic eukaryotes. Two thirds of the genes appear to encode plastid proteins, while the rest may have other cellular functions (including mitochondrial functions). The data provided strong evidence for a long-term symbiotic association between chlamydia-like cells and the Archaeplastida ancestor. It is possible that both cyanobacterial and chlamydial endosymbionts may have co-existed in the cell. It is not clear if the chlamydial association consisted of a recurrent infection by one or more specific endoparasites (as occurs in contemporary chlamydiae and their eukaryotic hosts), or an organelle-like endosymbiosis. While the cyanobacterium provided the photosynthetic function and was retained as an organelle, the chlamydiae supplied a number of key genes, including those coding for some metabolite transporters used by the plastid (especially the ADP/ATP translocator). This facilitated the communications between the cyanobacterium and the eukaryotic cell. However, the chlamydial partner was eventually eliminated by the host, once it became an unnecessary burden.
Endosymbionts or organelles?
The establishment of plastids from cyanobacteria probably occurred only once in a common ancestor of Archaeplastida; this was thought to be the only example of primary endosymbiosis leading to a photosynthetic eukaryote. However, another independent case of primary endosymbiosis is now known: Paulinella chromatophora is an autotrophic amoeba that contains two photosynthetic bodies that appear to have been acquired much more recently than the plastids of all other eukaryotes by the uptake of a Synechococcus-like cyanobacterium. Other examples of extreme bacterial endosymbionts, or pseudo-organelles, include Buchnera (in aphid insects) and Carsonella (with a tiny 160-kilobase genome, in psyllid insects). Additionally, the transfer of a complete genome from an endosymbiont to the host chromosome has been well documented for Wolbachia (in several insects).
Nature is unfinished...
Scholar articles (open access):
- Huang, J., Gogarten, J. (2007). Did an ancient chlamydial endosymbiosis facilitate the establishment of primary plastids? Genome Biology, 8(6), R99. DOI: 10.1186/gb-2007-8-6-r99
- Moustafa, A., Reyes-Prieto, A., Bhattacharya, D., DeSalle, R. (2008). Chlamydiae has contributed at least 55 genes to Plantae with predominantly plastid functions. PLoS ONE, 3(5), e2205. DOI: 10.1371/journal.pone.0002205
Related links:
Image credits:
- Unnamed artwork of a plant cell: Photo by R.W.W. (Ambra Galassi) on Flickr.
- Chlamydiae and the origin of plastids: Figure 4 from Moustafa A, Reyes-Prieto A, Bhattacharya D (2008) Chlamydiae Has Contributed at Least 55 Genes to Plantae with Predominantly Plastid Functions. PLoS ONE 3(5): e2205.
Read the rest of the article >>>
It seems that all known chlamydiae are obligate intracellular symbionts -- they can only reproduce inside eukaryotic cells, and remain metabolically inactive outside of their hosts (a virus-like lifestyle). Chlamydiae can infect different kinds of animals (mammals, birds, fishes, arthropods, crustaceans) and unicellular eukaryotes (such as environmental amoebae). Remarkably, chlamydiae have never been found in plants or in other plastid-containing organisms (red and green algae, plants and glaucophytes, together known as Plantae or Archaeplastida).
