Effectively every known eukaryotic lineage possesses mitochondria, or their descendants (which often are degenerate relative to mitochondria, i.e., mitochondria sensu stricto). As such, subsequent endosymbiont capture can be described as a serial affair (as opposed to parallel), where one or more additional endosymbiotic organism is acquired, by the host organism, in a step-wise manner. The most common consequence of serial endosymbiosis is the existence of plastids, in addition to mitochondria, in the cytoplasms of photosynthetic eukaryotes. Many additional examples of serial endosymbiosis exist, however. Note, though, that as with any horizontal gene transfer event, or for that matter any modification of an organism's genotype, the resulting organism (or organisms in the case of symbioses) are subject to selection, and presumably most new "experiments" do not survive. Furthermore, not all products of endosymbiont capture are equally numerous or diverse. Thus, while there are extensive examples of serial endosymbiosis, none will be as familiar to you as are the green plants which, of course, form the basis of terrestrial food chains.


The most familiar product of serial endosymbiosis are plastids, which represent a category of endosymbionts, the most familiarly of which is the chloroplast. Chloroplasts actually are one of a number of phenotypic forms that the plant plastid can differentiate into (Thomson and Whatley, 1980) . Other forms include chromoplasts, amyloplasts, etioplasts, leucoplasts, etc. The chloroplast found in green plants is descended from a once free-living cyanobacterium. The acquisition of this chloroplast resulted in the generation of green algae, a eukaryotic algae that grows predominantly in shallow waters and among which the most recent algal ancestor to the green plants is found. Other eukaryotic algae also have plastids that appear to be descendants from this same endosymbiotic event, though the route taken be these other algae towards acquiring photosynthesis is more complicated than that of green algae, as I consider in the immediately following subsection. With this complication as a caveat, nonetheless plastids appear to be monophyletic, though not all members of plastid-bearing clades are necessarily photosynthetic. In terms of serial endosymbiosis, the plastid presumably was acquired after the acquisition of the mitochondrion.

Table: Types of Plastids (Phenotypic Variants) as found in Plants*.

Amyloplasts Plastids with starch synthesis and storage functions
Chloroplasts Plastids employed predominantly for photosynthesis functions
Chromoplasts Plastids with pigmentation and storage functions
Elaioplasts Plastids used for fat storage
Etioplasts Chloroplasts of plants that have been grown in the dark and which can give leaves a characteristic yellow color
Leucoplasts Plastids that lack pigment and which are found in roots and other tissues that are not sun exposed
Proplastids Undifferentiated plastids that can differentiate into the various plastid forms described above

*See http://encyclopedia.laborlawtalk.com/Plastid and note that this table is for your information rather than something that you need to learn.

Secondary and Tertiary Endosymbiosis

Chloroplasts appear to be monophyletic, at least in terms of the original endosymbiotic event (McFadden and van Dooren, 2004) . Nonetheless, chloroplast acquisition by eukaryotic organisms appears to have occurred multiple times over the course of evolution. The apparent contradiction in this statement is resolved by every acquisition, except the first (the primary endosymbiotic event), representing secondary or tertiary endosymbioses. That is, acquisition by a eukaryotic organism of a eukaryotic organism that in turn serves as a chloroplast. The result is a cyanobacterium endosymbiont (plastid) of a eukaryotic alga, where that eukaryotic alga serve as chloroplasts for what, as a consequence, has now also become a eukaryotic alga: a cell living within a cell which in turn lives within a cell. But that's only secondary endosymbiosis. In tertiary endosymbiosis the same scheme is taken one step further, with the cell within a cell within a cell itself a chloroplast living within yet another cell, which as a consequence is now photosynthetic and therefore represents yet another lineage of eukaryotic algae. Another way of viewing these layers of endosymbiosis is that while the original plastid is monophyletic among these various algal lineages, the algae themselves are polyphyletic, and legitimately so! A summary figure of these sorts of chains of endosymbioses is presented at the end of this chapter.

Keep in mind that secondary endosymbioses, while distinct from serial endosymbioses, in practice also represent examples of serial endosymbiosis. That is, while the layer of endosymbionts, one inside of the other, represents secondary (or tertiary) endosymbiosis, the acquisition of these eukaryotic cells by an already mitochondrion-containing cell makes these serial endosymbiotic events as well.

Products of secondary endosymbiosis include the brown and golden algae. Included among the brown algae are the multicellular kelp, which are often rather large organisms, even huge. The chloroplasts of diatoms, too, are products of secondary endosymbiosis. In each case, it appears that it is red algae – a product of only primary endosymbiosis – that was acquired as the chloroplast. Euglenoids, which are unusual in that they can display mixotrophic metabolisms, that is, gaining their energy not just from photons but from organic molecules as well, possess chloroplasts that are green algae. Since Euglenoids are either phagocytic or derived from phagocytic forms, we can speculate that it was this mechanism that was involved in chloroplast acquisition, with the mixotrophic state perhaps equivalent to an intermediate stage potentially seen in the course of evolution of many eukaryotic algae. Certain dinoflagellates, in turn, appear to possess chloroplasts that are products of tertiary endosymbiosis, or at least are chloroplasts surrounded by four membranes versus the three seen with secondary endosymbiosis or the two observed with both cyanobacteria and plastids (in each case ignoring in this count the plasma membrane of the free living algal host, i.e., the all-encompassing cell membrane surrounding the cytoplasm containing all of these various endosymbionts).

The bottom line in all of these complications are that plastid acquisition has happened multiple times, creating otherwise unrelated algal lineages that, nonetheless, seem for the most part to possess a plastid that is monophyletic among some approximation of all of them. There is, in other words, something special about chloroplasts but at the same time nothing particularly special about chloroplast acquisition.

Other Endosymbioses

Besides mitochondria and plastids, however obtained, there exist a great diversity of additional endosymbioses. Many of these have been observed in animals, especially insects (Hoffmeister and Martin, 2003) , though classically it has been nitrogen fixing in plants that are the best known additional endosymbioses. The latter involves an uptake of rhizobia bacteria from the soil, where the bacteria are actually transferred into the cytoplasm of plant cells, which then form into the root nodule. This mechanism of acquisition differs from the other endosymbionts we have thus-far discussed in that it both does not involve vertical inheritance and it does not even involve a significant degree of genetic linkage between host and endosymbiont. This lack of co-vertical inheritance might be a consequence simply of the distance between plant roots and plant flowers, where while it may be possible for seeds to carry associated bacteria, it is difficult for those bacteria to make their way from roots to seeds in order to be so carried.

Among insect-associated endosymbionts, the best known is Buchnera. Buchnera, like many insect endosymbionts, is involved in compensating for the nutrient deficiencies in food sources. That is, aphids are specialized for feeding on the sap of plants, which is deficient in various amino acids, many of which are essential for aphids (i.e., they can't synthesize the missing amino acids themselves, which, from the plant's perspective, is the point of not providing them). However, Buchnera is capable of synthesizing these amino acids, and supplying them to their aphid host. Thus, because of their Buchnera endosymbiont, the aphids are capable of surviving on what otherwise is a low-quality nutrient source. Interestingly, while the aphids obtain various amino acids particularly from Buchnera, Buchnera at the same time is dependent on its host for the amino acids that it cannot synthesize. Befitting such high levels of metabolic integration and dependence, the Buchnera endosymbiont is transmitted vertically and, like mitochondria, maternally. Bacteria in addition to Buchnera play similar roles in the metabolism of other insects which rely on similarly nutrient-lacking food sources.

Mealybugs are another insect which has been shown to possess endosymbiotic bacteria. These endosymbionts are doubly unusual. First, they appear to represent a secondary endosymbiotic event in animals. Second, the primary endosymbiotic event appears to be the acquisition of a bacterium endosymbiont by a bacterium! Thus, a bacterium is found within a second bacterium that in turn is found within the mealybug cytoplasm. To the extent that bacteria do not possess mechanisms of phagocytosis, this example would seem to support the idea that endosymbiosis can occur absent phagocytosis and therefore might even predate the invention of phagocytotic mechanisms. Consistently, Lake (2009) argues that Gram-negative bacteria themselves could be products of an endosymbiotic merger of a clostridium and an actinobacterium.

Speculative Endosymbiotic Acquisition

There exist a number of papers that take the microbial symbiosis ball and run with it, attempting to explain especially various features of eukaryotic cells as having bacterial origins. Li and Wu (2005) , for example, argue that the origin of eukaryotic flagella is a potential endosymbiotic event. In general these scenarios are fascinating. To what extent they represent reasonable extensions of the endosymbiosis hypothesis for the origins of mitochondria and plastids, however, remains an open question. The reason for difficulty, especially, seems to a consequence of the absence of remnants of bacterial DNA within the organelles in question, i.e., as is found in both mitochondria and chloroplasts as well as the various insect endosymbionts described above. Thus, either these various organelles were once free-living bacteria but now lack even genomes, or they lack genomes because they never existed as free-living organisms.