If the chlamydial progenitor was a stranger in a strange land, then today's Chlamydiaceae are the occupier of a foreign country, and the boundary of the occupied territory is not readily defined. If the inclusion membrane is the product of chlamydial activity, the interface between host and parasite may be at that membrane and not at the outer membrane of the chlamydial cell itself. And if chlamydiae specify the nature of the inclusion membrane, perhaps they also specify the composition of the intra-inclusion fluid as well. One begins to wonder if the chlamydial inclusion exercises some of the functions of a multicellular organism and thereby reaps some of the benefits of multicellularity. On the basis of their experience with cellular slime molds and myxobacteria, Dworkin (1997); Bonner (2000); Shimkets (2000); and Kaiser (2001) have suggested minimum requirements for the simplest sort of multicellular organism. Inclusions definitely contain specialized cell types, and by shameless stretching of their definitions, the chlamydial inclusion might meet the requirements for self - recognition and spatially - oriented morphogenesis, but there is no evidence for communication between individual chlamydial cells, even when chlamydial inclusions are packed with EBs and RBs. Chlamydiae communicate with their hosts and vice versa, but they do not seem to ‘talk’ to each other. The inclusion is a likely place for operation of quorum sensing (Miller and Bassler, 2001), but I am not aware that it has ever been demonstrated. As emphasized by Shapiro (1997), cell - to - cell communication is an absolute requirement for multicellularity. The inclusion really does not fit the paradigm for a simple intracellular organism, but are there still enough intimations of multicellularity to make the inclusion a unit of evolutionary selection acting independently of selection on individual chlamydial cells?

There must have been strong selection for multicellular plants and animals - there are so many of them! It is likewise clear that selective forces have acted to restrict chlamydial multiplication within the bounds of the inclusion. Of all known Chlamydiales, only some (but not all) Neochlamydia reproduce outside of inclusions (Part I). In a broader context, Goebel and Gross (2001) have pointed out that intracellular parasites in general are much more likely to reside in membrane-bound vacuoles than directly in the cytoplasm. Bacteria that normally inhabit cytoplasmic vacuoles failed to multiply when micro-injected into the host cytosol (Goetz et al., 2001). Production of specialized cell types is one of the selective advantages attributed to multicellular organisms, and chlamydiae surely do that. Another frequently cited multicellular advantage is protection from environmental insults. The failure of RBs to survive, let alone multiply, outside the inclusion (with the one noted exception) is the best evidence that the inclusion serves this purpose. The highly selective inclusion membrane must protect against the effectors of innate and combinatorial immunity, although data directly demonstrating such protection are lacking. Simple multicellular organism are said to have the advantage of social or cooperative feeding. The inclusion may offer such an advantage. Now that it is almost certain that the hemispheric projections on the chlamydial cell surface represent the needle complexes of the Type Ill secretion apparatus (see Type Ill secretion systems) and not transport conduits, the mechanisms by which the many nutrients needed by chlamydiae pass from the host cell cytoplasm through the inclusion membrane and eventually to individual chlamydial cells remain as problematical as ever. However, it is likely that many metabolites are transported through the inclusion membrane by specific chlamydia - derived mechanisms. Thus, the fluid content of the inclusion may be at least in part chlamydia - determined. Selection may have operated to make the intra - inclusion fluid better able to support chlamydial multiplication and differentiation than the host cytosol. With some imagination this can be regarded as communal feeding.

Returning to the question of whether the inclusion is a unit of evolutionary selection, no definite answer is forthcoming. However, if it were to be ‘yes’, then that would help explain the evolution of the complex relationship between the host cell, the inclusion membrane, and the chlamydial cell itself.
[JWM]