Life persists. Plants, animals, and micro organisms have all worked out ways of ‘toughing out’ the hard times and waiting for the good ones (Levin, 1997). Birds migrate, bears hibernate, bacteria sporulate. So it is not surprizing that the Chlamydiaceae produce low-grade, long-lasting infections in their natural hosts. This behavior, usually referred to as persistent infection, is the subject of intense current interest because of its possible involvement in human disease. Persistent infections have been discussed at length in many excellent reviews (Beatty et al., 1994; Brunham, 1999; Darville, 2000; Nataro et al., 2000; Ward, 1999). Early investigations were summarized by Moulder (1991). Here I will concentrate on a few evolutionary questions. The ability to engender persistent infections is so widespread among Chlamydiaceae that it was in all probability a capability already possessed by the familial LCA. Chlamydial persistence comes in many guises and no one model fits them all. However, since chlamydiae are closely related organisms with pared - down genomes, a large number of completely different persistence mechanisms should not be expected. Let us hope that, when regulation of chlamydial multiplication in both acute and persistent infections is better understood, there will emerge a general underlying mechanism whose variations will account for the many faces of persistence.

Has the ability of chlamydiae to cause persistent infection been shaped by evolutionary selection? Is the persistent state merely the "normal" developmental cycle in stasis or is it a programmed survival response engrained in the genome?

It is generally assumed that persistent infection is adaptive, that it is of benefit to chlamydiae. In evolutionary terms, is the organism that can go into a persistent state a fitter organism than one that cannot? This is a hard question to answer. There is no doubt that persistence occurs and that it affects the course of chlamydial disease, but how is one to tell if a highly successful pathogen would be less successful in the absence of persistence? If a chlamydial population in the persistent state never produces progeny or if its progeny never reaches a new host, then that population, however much it may affect the well-being of the infected individual, has come to a dead end. Selection will have a chance to operate only if new hosts are colonized. It is easy to see how this could happen if reactivation of a persistent infection occurs in the originally infected mucosa of genital tract, eye, or lung. If reactivation occurs in deep sites like artery, brain or synovium, infection of new hosts with EBs originating in these sites will be much harder. Reactivation of human persistent infections and colonization of new hosts is unproven, but if persistence is indeed an adaptive response, if genotypes better fitted to the persistent mode have been selected and transmitted to new hosts, then new questions present themselves. When and where did selection occur? What phenotypic traits were selected and what is their genetic basis? Is there a distinctive persistent phenotype? What benefits accrue to chlamydiae from persistence?

It is hard to study persistence in laboratory animals, let alone people, so cell culture models have been used extensively. An inescapable difficulty with such models is that the chlamydial strains employed have almost always been maintained by serial passage in cell culture, a procedure that selects for rapid production of maximum numbers of infectious progeny, not for the production of persistence. As with host range, it may be that some of the determinants of natural persistent infections are not reproduced in cell culture. Nevertheless, much has been learned about persistent infection in cell culture models, and a general picture of the course of events has been obtained. The persistent state is initiated by interruption of the "normal" developmental cycle by a number of different conditions and agents, among them being antibiotics, nutrient deprivation, and immune factors, gamma interferon in particular. What these precipitators of persistence have in common is that they all produce stress. RBs cease to divide and are transformed into large, morphologically aberrant forms. Specific markers for the persistent state have not been found, but chlamydial gene expression is altered in persistent infections as compared to active infections. Genes for DNA synthesis are expressed, but genes for cell division are not (Mathews et al., 2001; Gerard et al., 2001; Byrne et al., 2001). When after days, weeks, or even months the stressful condition is removed, the aberrant RBs may reorganize and eventually produce infectious EBs. It is not clear what fraction of a non-dividing population reorganizes to form infectious progeny. Is it the entire population or just rare individuals? Because it is unlikely that a quiescent chlamydial cell would survive indefinitely without dividing, there must be mechanisms for keeping persistent infections going. A few EBs may be produced continuously or "non-dividing" chlamydiae may actually multiply at slow rates.

The cell culture models are consistent with the operation of selection but they do not prove it. If only rare individuals in a non-dividing population are reactivated when stress is removed, then selection for persistence could have occurred, but the question remains as to whether the reactivated individuals are actually different. Identification of one or more chlamydial genes that are expressed only in the peristent state would be evidence for selection. There is also the possibility of persistence being favored by loss of genes. C. trachomatis strains with mutated incA genes form inclusions that have no IncA in their membranes and that do not fuse with each other. Such strains produce subclinical infections more often than normally fusogenic strains (Geisler et al., 2001). It could also be that persistent states are brought about by down-regulation of genes essential for progression from RBs to EBs and that there are no unique single markers, only unique patterns of gene expression.

There is still no convincing evidence that persistent infections in people are like the ones studied in cell culture. Non-dividing forms have not been demonstrated, but there are reports of DNA and antigens in the absence of recoverable infectivity. I am not sure just what sort of evidence would be convincing. Because non-dividing, morphologically aberrant RBs are so prominent in cell culture models, the elusive persistent phenotype in human infections is often visualized as something similar. But there are other possibilities. How about a developmentally challenged EB? After all, EBs are the chlamydial equivalent of stationary phase cells and spores in other bacteria. Could EBs blocked in one of the early steps in the EB to RB transformation persist in a state of arrested development and then later resume their progression to RBs and eventually a new generation of infectious EBs?

Cell culture studies have suggested several ways in which chlamydial populations that enter into persistent states might have survival advantages over those that do not. Persistent populations might be more resistant to antibiotics. They might be less susceptible to immune factors in general and to the inducible nitric oxide synthetase - inducing and tryptophan - depleting effects of interferon gamma in particular. They might more effectively delay the onset of apoptosis, and their prolonged presence in hosts might give opportunities for wider dissemination of infection. Any and all of these potential benefits could conceivably be the basis for selection of persistence-promoting genes.

If entering into a persistent state is an adaptive trait possessed by the common ancestor of Chlamydiaceae, then as it radiated to different hosts and produced the several modern species, mechanisms of persistence should have diverged. Differences among species have been observed in both laboratory animals and cell cultures. Some may have resulted from the way persistence was observed and measured, but others appear to be real. For example, C. trachomatis is more susceptible to growth inhibition by gamma interferon than is C. muridarum (Perry et al., 1999). As persistence is studied in a wider range of chlamydial populations and as more complete genome sequences are established, a truly evolutionary approach to persistence will become possible.

The host remains a largely unexplored term in the equation that defines chlamydial persistence. That inbred mouse strains differ in susceptibility to infection with both Chlamydia and Chlamydophila spp. is well known and evidence is accumulating for a definite genetic element in human infections as well (Ward,1999) [see: genotypic determinants of human disease]. However, host genetic factors that specifically impinge on initiation, duration, and activation of persistent infections have not been identified.
[JWM] Dec 2001