by Christoph
I like to imagine that microscopists in the late 19th century who looked at water samples were happily surprised when they saw organized structures of cell clusters in the whole tangle of protists and bacteria, which they then liked to call “rosettes” because these were structures known from botany. And on closer inspection, it turned out that some of them were actually bacterial rosettes.
Of course, it could have been fun to research when the term "rosette" first appeared in the scientific literature, but I found it more interesting to understand how bacteria craft these rosette structures, which bacteria are known to do this, and when, that is, under which growth conditions. Here are two (and a half) examples, all of which go back to research from the early days of bacteriology.
Planctomyces bekefii rosettes
That Planctomyces bekefii (phylum Planctomycetota) forms rosettes has long been known: according to Fuerst (1995), the Hungarian researcher Gimesi found them in a pond in Budapest and initially thought it was a planktonic fungus (its conidiophores, to be precise). In these rosettes and those of the closely related Planctomyces gracilis, few or up to ~30 ovoid cells are connected to their center via rigid stalks (Figure 5). P. bekefii cells mostly divide by budding rather than binary fission, as is common among planctomycetes (see here and here in STC). You can spot a few buds in Figure 2A.
(click to enlarge)
Figure 5. (A) Phase-contrast micrograph of a rosette of P. bekefii, showing non-cellular stalks radiating towards a central phase-light granule from spherical cells at the rosette circumference, some of which possess buds. Micrograph taken during collaborative research of the author and J. T. Staley (Dept. of Microbiology, University of Washington). Bar,10 µm. Source. (B) Pg, a rosette of P. gracilis with at least 11 visible filaments ; Pb, a small, three-membered rosette of P. bekefii, the type species of the genus Planctomyces and morphotype la of the Blastocaulis-Planctomyces group of budding and nonprosthecately appendaged bacteria. Bar, 5.0 μm (Not visible in this cutout). Source
This peculiarity of the planctomycetes makes them easily distinguishable from the superficially similar rosettes of Caulobacter & company, which propagate indiscriminately by binary fission (not Hyphomicrobium though, see part 3). Also, the rosettes of Caulobacterales and Planctomycetes can be easily distinguished under the microscope by their stalks: while the typical stalk of the Caulobacterales is highly flexible, that of several Planctomyces species is non‑prosthecate, multifibrillar and pretty rigid (⌀ ~0.3 µm), but thinner ones are known from P. maris (⌀ ~0.1 µm). Sometimes, to the delight of the microscopist, they are also encrusted with iron and manganese oxide in samples from native freshwater habitats. "Non-prosthecate" means that the stalks of the Planctomycetes are not cytoplasm-containing extrusions of the outer membrane, as in the case of the Caulobacterales.
How the P. bekefii stalks are connected/attached to the cells is not known, nor is it known how they are synthesized, i.e., how they are extended during growth, and how they connect to the center of a rosette (a micro-particle or another stalk tip?). It is known that many Planctomycetes can grow on surfaces as a biofilm, but whether this also applies to P. bekefii can only be clarified if these organisms can be studied in the lab during all phases of their life cycle.
Nevskia ramosa rosettes
You may never have heard of Nevskia, but they are not that rare: "During a calm and sunny weather period 430,000 Nevskia-like bacteria per mL were found in surface samples" from a (freshwater) ditch near their institute in Oldenburg, Germany by Pladdies et al. (2004). "Surface samples" indicates that Nevskia is preferentially found in the neuston, that is, the company of microbes that thrive in and as the 1–50 µm-thin film at the liquid/air interface (in the lab, we would call it a "pellicle"). Nevskia ramosa is a genus in the order Nevskiales and related to the Xanthomonadales in the so far poorly resolved "basal branches" in the (slightly outdated) phylogenetic tree of the Gammaproteobacteria (see here).
(click to enlarge)
Figure 6. Phase-contrast photomicrographs of Nevskia ramosa. (E) An enrichment of flat rosettes of Nevskia-like cells from stagnant water in a watering can. (F) Geometrical patterns often formed on agar plates. (G) Safranine-stained flat rosettes of N. ramosa Soe1 (detail in the Frontispiece). (H) Pure culture of N. ramosa strain OL1. Bars, 25 µm (A to D and F to H) and 10 µm (E). Source
Babenzien (1967) wrote: "Young motile cells develop submersed, then adsorb to the water surface, lose the polar flagellum, and form a hyaline slime stalk on the concave side of the cell. When a cell multiplies by binary fission, branching of the stalk occurs. The resulting flat rosette can reach a size of 70 µm in diameter" (see images here from H. Cypionka's Mikrobiologischer Garten (eng.)). In Figure 6, you see that the formation of rosettes by Nevskia – first described, mind you, in 1892 (ref. 7) – is very efficient. But in contrast to the formation of rosettes by Phaeobacter (see part 1) or Planctomyces (this part) or Caulobacter (part 3) there is 1. no approximate final size of the rosettes (number of individual cells per rosette), and 2. the individual cells are not connected to each other through "fibrils," but share an exopolysaccharide (EPS) sheath. The EPS consists mainly of rhamnose, with small amounts of glucose and mannose, but it is not known how the nicely visible branches come about, whether this is a physico-chemical process or whether the Nevskia actively influence it.
No, 'Tubercle bacilli' do not make rosettes!
In my (laptop) folder of microbial rosette images, I stumbled across this one in Figure 7. Stupidly, I had forgotten to note the source. At a cursory glance, I thought it was an obscure older drawing of a choanoflagellate rosette and had added it to this folder. It seemed suspicious to me, though, that, unlike in other images of choanflagellates, the flagella did not appear to be arranged peripherally on the outside (see here for an example from Kayley Hake), but rather on the inside. A Google reverse image search was surprisingly unsuccessful. But a perk of communicating via “social media” is that you can upload a picture and ask: "Does anyone know of this picture and where it comes from?" (on mastodon, on Bluesky, and on 𝕏). Less than 24 hours later, a positive response from mastodon user benetwict popped-up in my timeline (thanks a lot, again!). The image is reproduced on plate 48 as number VI in the monograph (stand-alone scholarly book in librarian lingo):
(click to enlarge)
Figure 7. See text for details. Source
K. B. Lehmann, and R. O. Neumann. Atlas and essentials of bacteriology. William Wood and Company, New York, 1897
A digital version of the book is stored in the Internet Archive that you can read here or download as PDF.
The legend to the image says that it shows 'Tubercle bacilli' (Mycobacterium tuberculosis) lining the cells of a cavity of a "cheesy" (pus‑filled?) bronchial gland of a tuberculosis patient. So, what I mistook as flagella are actually bacteria, and the suspicious rosette are lung epithelial cells. Case solved. The authors further mention that the image is a copy of Pl. II., 9. from a famous publication by Robert Koch, Die Ätiologie der Tuberkulose, which I found as chapter in a digitized version of the Mittheilungen aus dem kaiserlichen Gesundheitsamte (1884). The comparison of Figure 7 with the original Taf. II., 9. reminded me that at the end of the 19th century, "copying" quite literally meant "re-drawing" (with a whiff of artistic license), for technical reasons. Today, we tend to think of hi‑res/low-loss photocopies, or at worst, some AI‑generated nonsense.
Stay tuned for part 3, in which I will turn to rosette formation in Caulobacter, in Escherichia coli (inevitably), and in multicellular magnetotactic bacteria (MMB) that were first mentioned in Elio's piece Could We Have Started Out as Magnetotactic Bacteria? back in 2007.
I gladly mention that Heribert Cypionka, who led most of the research on Nevskia mentioned here, showed me how incredibly fast the huge Achromatium oxaliferum bacteria (up to 30×125 µm in size) can move during a lab visit of the Lake Stechlin branch of the IGB in 2017.
Do you want to comment on this post? We would be happy about it! Please comment on Mastodon, Bluesky, or on 𝕏 (formerly Twitter).
Comments