Tales of Mystery and Imagination (1/2)

by Daniel P. Haeusser
 

The Story so far

Over the last ten-plus years of this blog's existence, one cluster of branches on the bacterial phy­lo­ge­ne­tic tree in particular has presented us with fascinating mysteries regarding basic, and seem­ing­ly unique, cellular biology. Many aspects of the species on these branches remain enigmatic, but what details researchers have been able to coax from them continue to induce sharp turns in our microbial assumptions and understanding: conceptual angles that mimic the bumpy, uneven cellular surface of some of those species under discussion (Figure 1).
 

Figure 1. Members of the Planc­to­my­ce­tes–Ver­ru­co­mi­cro­bia–Chla­my­diae super­phylum (PVC) arose from a common Gram-negative ancestor and presently display variations in cell shape and cytokinesis. Like present-day E. coli, the Gram-negative ancestor contained a pe­ri­plas­mic peptidoglycan cell wall (dashes), employed an MreB cytoskeleton (not shown) to elongate, and used an FtsZ cytoskeleton to constrict (mid­cell black dot). Recent evidence indicates that Planctomycetes have retained peptidoglycan cell walls. Some members have also retained FtsZ-based cytokinesis (not shown), but many have lost FtsZ and now divide by budding, even with cell walls. The Verrucomicrobia have kept cell wall and division components of their evo­lu­tio­nary ancestor, but often have outer mem­brane 'wart-like' extensions that are not well characterized. Some Chlamydiae have retained cell wall all along their length (not shown), but pathogenic species have very low amounts of pep­tido­glycan synthesis, limited to the site of di­vision during cytokinesis. Additionally for un­known reasons, their cytokinesis has evolved to occur slightly asymmetrically (exaggerated in the illustration for emphasis). The re­cent re­ve­la­tions of peptidoglycan being present through­­out the PVC superphylum leave the Mollicutes (illustrated here with the representative My­co­plas­ma mobile ) as the only main group of bac­te­ria that seem to have evolved to lack pep­ti­do­gly­can and to have also lost FtsZ (present in their Gram-positive ancestor). Despite this loss, Mollicute species like M. mobile have complex shapes and uncharacterized cytoskeletons. I­llu­stration by Allyson Corigliano (Canisius College) based on the summary figure of the Frontiers in Microbiology review source

That cluster of branches (if you haven't twigged it yet) would be the Planctomycetes–Verrucomicrobia–Chla­my­di­ae superphylum (PVC). If you fondly recall our past vi­sits with these microbes, you could skip on to the next section. But for those wanting a little context before delv­ing into new developments on the PVC front, a recap and introduction:
 

Though mentioned on occasion, we haven't really fea­tur­ed anything on the Chylamydiae branch, a phylum of ob­ligate intracellular bacteria that includes, but is not li­mit­ed to, the well-known pathogenic Chlamydia family. As is common for obligate intracellular organisms, these bac­te­ria have reduced genomes (which will have importance discussed further below). But they are particularly no­table for their cellular development into an elementary body, a non-replicating form in the nanometer-diameter range that can transmit between host cells via the ex­tra­cel­lu­lar milieu. Though often compared to a spore, it might be more appropriate to view the elementary body as a ‘virion’ mode of behavior for the Chlamydiae phylum species.
 

One of my first posts for this blog featured the discovery of tubulins in a Prosthecobacter member of the Ver­ru­co­mi­cro­bia, a phylum that Elio mentioned in a 'whiff of ta­xo­no­my' post that speculated on their eponymously war­ty appearance. To my knowledge, the functions of the prosthecae and the warty bumps of the various mem­bers of this phylum are still unverified, though in theory they might be expected to function in surface adhesion, as in the prosthecate Alphaproteobacteria, and/or to pro­vide a greater membrane surface area.
 

A decade ago, a group of grad students wrote on the "far out" Planctomycetes, a phylum known for tantalizingly eukaryotic-like features, such as extensive en­do­mem­brane formation, including the apparently membrane- bound bacterial organelle housing the complex res­pon­sible for anaerobic oxidation of ammonium. Interest in the Planctomycetes particularly surged with the dis­co­ve­ry of membrane compartmentalization around DNA in Gemmata obscuriglobus. However, as cov­ered here by another graduate student contributor, this purported prokaryotic 'nucleus' seems to be formed by a bilayer continuous with, not independent of, the cell membrane. The last word is not in on this topic, as the purported 'nuclear membrane' contains nuclear-pore-like structures.
 

An Aside on Planctomycete Chromosome Replication

What can you do if you don't want to wait patiently for more detailed studies of the enigmatic mem­brane compartmentalization around the nucleoid in Gemmata obscuriglobus ? Ask a co-blogger ! My colleague Christoph had already contributed insights from bioinformatics to chromosome re­plication of the CPR bacteria. So here is what he said.
 

(Click to enlarge)
Figure 2. A Schematic re­pre­sen­ta­tion of pre­dicted oriC structures for 5 Planctomycetes. Red half­-circles: pre­dicted DnaA-binding sites; orange ellipses: predicted DNA-un­winding elements (DUE); blue ellipses: DnaA ssDNA-binding motif (DnaA-trio); gray boxes: flanking genes. The oriCs of B. subtilis and E. coli are shown for comparison. C. Weigel, unpublished. B Domain struc­ture of the E. coli DnaA protein. Source

The 15 genomes of the Planctomycetes that have been sequenced so far are rather large (3 – 10 Mb, in most cases approximately twice the size of the E. coli genome) and they are all single circular chromosomes. For five of them, including G. obscuriglobus, the replication origin oriC can be safely predicted to localize in the ~300 – 500 bp long intergenic region between the dnaA-2 and dnaN genes (Figure 2A). This genomic location for oriC is quite common for bacteria from all known phyla but this is not a strict rule. The best-studied model, E. coli, being a pro­minent exception. The oriC structure of these five Planc­to­my­ce­tes is fairly similar to the known replication ori­gins of B. subtilis and E. coli (shown in Figure 2A for an easier comparison). They contain a prominent DNA-un­winding element (DUE), that is, a short stretch of ~50 bp that becomes easily single-stranded when negatively supercoiled. Also, they all contain a 'DnaA-trio' motif flank­ing the DUE, which has recently been shown to bind DnaA to single-stranded DNA during the initiation step of chromosome replication in B. subtilis. Three of the Planctomycetes oriCs lack a hallmark of all known oriCs: a DnaA‑binding site in reverse orientation flanking the DUE (marked 'R1_{E. coli'} in Figure 2A). This R1 DnaA‑binding site is thought to be important for loading of the he­li­case during initiation of chromosome replication, and the mechanism might therefore be different in the Planctomycetes. This latter suspicion is reinforced by the observation that the Planc­to­my­ce­tes genomes contain two dnaA genes (encoding the replication initiator protein), and one of these genes, dnaA-2, lacks the N-terminal domain 1 known to be important for interaction with the he­li­case, while the dnaA-1 genes encode full-length DnaA orthologs (Figure 2B). Note that two dnaA genes are very rarely found in bacterial genomes, the Chlamydiae and one branch of the Del­ta­pro­teo­bac­teria, the Desulfovibrionales, being exceptions but in both cases the twin dnaA genes are full-length orthologs. So, taken together, it is safe to predict that chromosome replication in the Planctomycetes is not all that different from other bacteria for the initiation step, but the helicase-loading step certainly requires more detailed analyses.
 

The chromosome segregation machinery in the Planctomycetes shows common features in that all genomes encode orthologs of the ParA and ParB proteins, and the ParABS system is therefore probably operative. Also, all Planctomycetes encode orthologs of the FtsK DNA translocase. How­ever, the Planctomycetes FtsKs lack the N-terminal membrane-anchoring domain of E. coli FtsK and also the distribution of KOPS motifs (=FtsK-binding sites) is not skewed (=gradient from oriC to the terminus) as in the E. coli genome but more random, for example, in the genome of Pirellula staleyi DSM 6068. This indicates that, together with the lack of FtsZ in the Planctomycetes (see above), the late steps of chromosome segregation and subsequent cell division are certainly quite different from your garden-variety rod-shaped bacteria like B. subtilis and E. coli.
 

So then, aside from having odd shapes or containing eukaryotic-like factors and endomembrane structures, what makes the PVC superphylum so uniquely bizarre for bacteria?
 

Division and Growth Nonconformists?

With a few exceptions, bacteria contain cell walls, predominantly composed of peptidoglycan. Most characterized bacteria divide by binary fission, and even among those that use other mechanisms, most still utilize the prokaryotic tubulin FtsZ in the cytokinetic process. Among the bacterial non­con­formists who seemed to abide without cell wall and/or divide independently of FtsZ, many seemed to fall within the PVC superphylum. A recent review in Frontiers in Microbiology presents an evolutionary perspective on some long-standing mysteries regarding PVC superphylum cell di­vision and the historic lack of observed cell wall synthesis in PVC phyla.
 

It is not particularly surprising to see the two nonconformities (in cell wall and in cell division) go hand-in-hand. If a bacterial species has evolved to have a cell wall, it is necessary to produce and turnover that structure during growth, including for the growth that must accompany cytokinesis. Indeed, for some bacteria, divisional growth is their only means of growth. The major roles of FtsZ are in providing a force for membrane constriction and a foundation for the localization and ac­ti­vity of cell wall synthesis enzymes, as two recent papers (here and here) demonstrate. Thus, were a species no longer require peptidoglycan structure for its fitness, its division ma­chi­ne­ry could be theoretically become FtsZ independent. That is, if another means of membrane con­striction were available. Conversely, were a species to develop an alternate means of division and lose ftsZ, it would require some other factor for guiding wall synthesis during division, or be freed to lose that cell wall structure.
 

For the nonconformist members of the PVC superphylum, how did they evolve their alternate di­vision strategies or evolve the ability to do away with peptidoglycan structure? And how much are the two linked? Before answering that question, however, we have to question the underlying as­sumption that several PVC superphylum members lack peptidoglycan. This assumption hasn't been based so much on scientific ob­ser­va­tion, as much as on the historical lack of scientific ob­ser­va­tion of peptidoglycan cell wall structure in certain PVC superphylum species. Recent data now indicates that more PVC members contain cell wall than assumed. So stay tuned for a second part of this post! I will discuss then what the past un­derstanding of cell wall conservation in bacteria was, and how it has changed...
 

Front page: Cryo-electron tomography micro­graph and reconstructed, averaged to­mo­graph­ic slice subframes of Planc­to­my­ces lim­no­philus cells. Source