Greater complexity can be viewed as occurring in multiple stages, including in terms of the evolution of intracellular membrane compartments (i.e., the endomembrane system in eukaryotes, which includes the nucleus), endosymbiosis (which is typically associated with but not limited to eukaryotes), gathering into colonies, achieving cellular differentiation, and acquisition of microbiomes. The most primitive form of differentiation is the separation of soma from germ lines.
Acquisition of endosymbionts was a key step in the evolution of eukaryotes. It provided the pre-endosymbiont ancestors with new metabolic pathways, while at the same time gave these cells a chemical modularity not found in bacteria. That is, following endosymbiont acquisition the resulting organisms possessed at least two membrane types: that associated with the plasma membrane, which could become specialized for resource acquisition, and that associated with the endosymbiont, which could become specialized for other functions, such as energy transduction for mitochondria and chloroplasts.
An interesting question is whether these acquisitions predated the formation of the endomembrane system. That is, were endosymbionts acquired initially through engulfment, which seemingly would require some degree of endomembrane functioning? Or, instead, did proto-eukaryotes acquire endosymbionts, such as mitochondria, perhaps through infection by an otherwise intracellularly pathogenic bacterium, and then only subsequently evolved an endomembrane system, including associated nucleus? Thus, endosymbionts represent an important step in terms of membrane-function differentiation, but whether endosymbionts were acquired following the evolution of engulfment or were acquired initially as otherwise benign intracellular bacterial pathogens are open questions. Note, though, that it is perhaps more likely that chloroplasts were obtained via engulfment assuming that fewer cyanobacteria than heterotrophic bacteria exist as intracellular pathogens.
Within cells, greater complexity is seen especially in the transition from prokaryotic cell types to eukaryotic ones, which also was accompanied by greater cell size. A large part of this complexity is seen with intracellular partitioning, as effected by the endomembrane system. The endomembrane system consists of such organelles as the endoplasmic reticulum, Golgi apparatus, various vacuoles, lysosomes, and the nucleus. Each distinct member plays a different role within the cell, with the compartments they create supporting different chemical reactions – i.e., different metabolic processes – from those that take place in the cytoplasm. The nucleus, of course, separates chromosomes from the cytoplasm, and is the defining feature distinguishing eukaryotes from prokaryotes.
While it is clear that the endomembrane system serves roles that are basic to the functioning of eukaryotic cells, these roles do not necessarily represent the initial motivation for the development of internal membranes. What then motivated the evolution of the endomembrane system? Perhaps three explanations may be put forth: (i) the endomembrane system served to increase cell surface or at least membrane area, (ii) the endomembrane system was established as a means of separating chemical reactions, or (iii) the endomembrane system evolved in conjunction with the evolution of the nucleus. To this I add a fourth that I won't otherwise discuss and that is that engulfment of extra-cytoplasmic materials could have predated eukaryotes since the process of engulfment appears to occur in at least some bacteria in the course of endospore formation plus seemingly complex membrane morphologies can be associated with prokaryotic microbial consortia (Lake, 2009) .
Membranes transecting cytoplasms, either fully or partially, are not unique to eukaryotes. Indeed, invaginations, representing partial cytoplasmic transections, are seen in various prokaryotes as a means of increasing plasma-membrane area. These invaginations are seen most familiarly as the cristae of mitochondria, but also are found in the plasma membranes of purple photosynthetic bacteria as well as the plasma membrane of the very large Gram-positive bacterium, Epulopiscium fishelsoni. Note that the increase in surface area of cristae of mitochondria, and perhaps the invaginations of purple bacteria as well, occurs not for the sake of increasing the cell's surface-to-volume ratio but instead as a means of accommodating additional membrane-associated chemistry. Thus, at least in principle, endomembrane evolution need not have been solely for the sake of providing more membrane area for the sole utility of increasing resource-absorptive capacity.
Cytoplasmic extensions are presumably a more reasonable means of providing greater cell surface area than invaginations, a strategy employed not just by various eukaryotic cells (especially phagocytic ones) but also can been seen among certain marine prokaryotes. A logical basis for that argument would be that concentration gradients are more easily dispersed from the surface of exvaginations than they are from the surface of invaginations. Thus, cytoplasmic extensions presumably are a more effective means of increasing surface area to effect greater rates of movement across membranes. A counter argument would be that cytoplasmic extensions require greater degrees of within-cell movement of materials to take full advantage of overall greater surface areas, i.e., distances to the center of a cell are greater given cytoplasmic extensions versus invaginations. The basic point therefore should be that evolving plasma-membrane invaginations as a means of increasing the surface-to-volume ratios of cells is neither the only solution to this problem of resource acquisition nor necessarily the only utility associated with developing invaginations. For considerable discussion of the evolution of cell shapes, see Young (2006; 2007) .
While membrane invaginations can increase the amount of membrane-associated chemical reactions that a given cell can effect, it is not the solution for establishing multiple, especially otherwise incompatible chemistries within the same cell. The latter may be achieved by having membranes that are separated from the plasma membrane, such as membrane-enclosed compartments. A first step towards achieving this state, however, may involve a simpler segregation of otherwise somewhat compatible plasma-membrane associated reactions such that the efficiency of those reactions may be increased. Starting with this separation, a relatively simple means of achieving greater efficiency could involve the creation of greater concentration gradients across membranes by pumping materials into membrane-enclosed organelles rather than across the plasma membrane into the extracellular space.
The latter mechanism is seen in chloroplasts, as well as free-living cyanobacteria. In these prokaryotic organisms the thylakoid membranes represent bona fide membrane-bound organelles which may be viewed as at least analogous to the membrane-bound organelles of the eukaryotic endomembrane system. Presumably homologous to proton pumping by heterotrophic bacteria such as E. coli and also mitochondria across their inner membranes, protons are pumped by these cyanobacteria across thylakoid membranes into what topologically is an equivalent location. That is, the lumen of the endomembrane system and the outside of a cell are equivalently distinct from the same cell's cytoplasm; in addition, one can be converted to the other, endomembrane into plasma membrane and vice versa, through invagination of the plasma membrane or the fusing of a vesicle with the plasma membrane, which are endocytosis and exocytosis, respectively. By pumping proteins into these vesicles as formed by the thylakoid membrane, a much steeper protein gradient (proton motive force) may be established via the movement of fewer protons.
This chemistry should be viewed as a relatively simple utility associated with transitioning from plasma-membrane invaginations which are continuous with the plasma membrane to vesicles which, at least physically, are not. That specific utility, however, is not necessarily a description of the specific chemistry motivating formation of the endomembrane systems of eukaryotes. Nonetheless, once separation between compartments has occurred, a next step in endomembrane evolution could be a concentration of different proteins in different membranes and thereby the generation of separate membranes and compartments with different functions.
An alternative route toward the evolution of endomembranes is via phagocytosis. Phagocytosis involves the invagination of the plasmas membrane until a vesicle pinches off, i.e., it is a form of endocytosis. Phagocytosis can be accomplished if a cell does not possess a cell wall, since cell walls otherwise would separate plasma membranes from to-be-engulfed food particles (though, it should be noted, this separation would be less absolute given instead the pinocytosis of extracellular fluids). Absent cell walls, then phagocytosis is an obvious advantage of having an endomembrane system since it allows for an increase in the membrane area found surrounding nutrient sources as they digest, i.e., across which nutrient absorption ultimately can take place. In addition, phagocytosis is arguably an advantage particularly in wet environments where extracellular digestion can be disadvantageous due to the propensity for exoenzymes to diffuse away (indeed, phagocytosis could be viewed as a means of protecting the public good/extended phenotype of digestive enzymes from exploitation by cheaters).
Thus, an organism which could develop phagocytic mechanisms would avail itself to resources, e.g., both dead and living planktonic bacteria, which otherwise are not easily or effectively exploited while they remain within the water column. Alternatively, the extracellular matrix of bacteria, including perhaps especially as found in biofilms, can retain exoenzymes within the immediate bacterial vicinity (Flemming and Wingender, 2010) , though this route towards substrate digestion presumably is neither a substitute for nor the equivalent of the digestion that can be effected via phagocytosis by protists.
It is possible to envisage an intermediate state to full-blown phagocytosis, one that involves plasma-membrane invagination but not also the familiar pinching off. Here, a larger, e.g., proto-amoeba-like cell might nestle against a substrate, rather than fully engulfing a particle, and release hydrolytic exoenzymes that are less free to disuse away from the organism due to this nestling. Indeed, it is relatively easy to envisage a progression where (i) protist docking against a substrate could allow for the acquisition of free floating nutrients that are associated with that substrate (e.g., to steal the products of extracellular digestion by bacteria), (ii) this could be followed evolutionarily by exoenzyme secretion by these same protists but without associated invagination, (iii) partial invagination against a substrate could then allow for better concentrating of exoenzymes upon the substrate (representing essentially a sealing mechanism around the edges of the interface), and (iii) development of further invagination might occur until engulfing of smaller particles is achieved (e.g., should protist-released hydrolytic enzymes result in the breaking off of aspects of surfaces such as of biofilms). Note that packaging of hydrolytic enzymes into preformed vesicles (lysosomes) could have evolved prior to the evolution of the resulting food vacuoles since secreted enzymes require delivery to the plasma membrane for extracellular digestion, though fine tuning of intracellular digestion would require fusion with the food vacuoles (for phagolysosome formation) rather than with the plasma membrane.
This scenario, in its progression, is basically an elaboration of the increased membrane area and separation of chemistries hypotheses provided above. It also is equivalent to how one might envisage the evolution of engulfment of foods by multicellular organisms (i.e., pancaking against a substrate preceding the evolution of intake of a substrate into a more enclosed compartment). Note also that the idea of prokaryotes achieving such membrane flexibility is not entirely inconsistent with modern functions since a number of bacteria exist that lack cell walls plus even phagocytosis-like mechanisms exist in bacteria, i.e., such as those involved in endospore formation, as previously noted .