by Christoph Weigel
Intro Imagine you equip your standard E. coli cells with a few additional genes, conveniently placed on plasmids and under control of inducible promoters. You grow the cells and induce the expression of these additional genes. Not long after, the cytoplasm of the cells turns reddish, when looked at under the fluorescence microscope, due to expression of a fluorescent protein, mCherry, encoded by one of the 'few additional genes'. Conspiciously, most of the 0.5 x 2 µm-sized cells are by now stuffed with single large "inclusion bodies" that, when examined under the electron microscope, turn out to be proteinaceous tapes tightly rolled‑up to barrel-shaped hollow cylinders with a diameter of ~0.5 µm and a width of ~0.5 µm (Figure 1). That's what expression of repA – repD, the other four 'additional genes', brought about. Next, you treat the cells with lysozyme that breaks down their cell walls. Such wall‑less cells, called spheroplasts, need to be osmotically stabilized to keep them from bursting, therefore you add sucrose to the cell suspension. When you now lower the pH of the medium, the barrel-shaped coils un-roll in a blink, from the inside, telescope fashion, and form tubes ~165 nm in diameter and up to 20 μm(!) long. ZAP! (you can watch a movie of this zap!-ing here). These long tubes deform and puncture the spheroplasts' membranes – maybe twice actually, due to their lengths – thus mixing the contents of the bacteria with the medium: the mCherry color leaks out (Figure 1). Among biologists, these veritable stilettos are commonly known as R-bodies, whose study goes back to 1938 when Tracy Sonneborn first described "killer" strains of the ciliate . This story, how researchers found out that the "killer" trait of Paramecium aurelia is due do the ability of its kappa particle – one of its several bacterial endosymbionts – to produce these R-bodies that under appropriate conditions virtually stab their victims is worth a separate post, today it's just looking at R-bodies.
Picture #38 Electron micrographs of coiled and extended R-bodies within bacterial cells were known since the mid-1960s but Quackenbush and Burbach were, to my knowledge, the first to show in their paper from 1983 electron micrographs of purified type 51 R-bodies (Figure 2). The enormous differences in size of the compact, contracted form (panel a) and the partially (b) or fully un-rolled, extended form (c) likely made taking these pictures a hell of a job for the electron microscopist (Note that in panel a the magnification is ~3x higher than in panel c). But despite the rather poor quality of the printed figure (and even worse in the PDF version due to poor scanning) the dimensions of both forms can be measured and some structural details are readily seen: the tightly coiled-up 'tape' of the contracted form with distinct layers (a), and un-rolling of the tape from the inside in the partially extended form (b). These R-bodies were purified from E. coli cells that had been transformed by a recombinant plasmid carrying a 2.8 kb DNA fragment from a much larger plasmid of the Caedibacter taeniospiralis 47 endosymbiont of a Paramecium tetraurelia strain. The cloned fragment codes for four small proteins RebA – RebD with molecular weights of 11.6, 11, 6, and 8.8 kD, respectively. Only the two larger ones finally form the R-body while the function of the two smaller ones is unknown. But because their expression is required for R-body formation they might be involved in stitching together the larger ones to assemble the proteinaceous 'tape'. Although not a scientific expression, "stitching" is appropriate here because the protein tape must contain several thousands of RebA and RebB monomers to obtain a final tape lengths of ~20 µm, and – a nightmare for the biochemist! – these tapes are highly resistant to denaturing conditions, treatments with "in combinations of 8 M urea, 1 to 10% SDS, and up to 5% 2‑mercaptoethanol or dithiothreitol. Incubations were carried out at 100°C for up to 1 h. Additionally, incubation at 37°C for 2 h in 8.6 M guanidine hydrochloride or 6 M guanidine thiocyanate, followed by dialysis and standard treatment with the final denaturing and reducing sample buffer, did not result in greater R-body dissociation. Also, boiling the R-bodies for 5 min in dilute hydrochloric acid (pH 1.8) had no noticeable effect on dissociation. The unusual stability of these structures suggests that covalent bonds other than disulfide linkages are involved in the polymerization of subunit polypeptides". Intriguingly, un-rolling and re-winding of the R-body coil can be repeated more than 200 times by just changing the pH of the suspension, no energy input required. Apparently, a pH‑dependent conformational switch between looped and α‑helical structure of a small stretch of ~20 amino acids in the C-terminal part of RebA (114 amino acids) is sufficient to drastically change the structure of the entire multimer. Although anything but a common theme in biology, such pH‑dependent larger-scale conformational rearrangements in proteins are known from hemagglutinins of eukaryotic viruses.
Always involved, of course: phages The plasmid-driven production of type 51 R-bodies by the Caedibacter endosymbiont for the benefit for its ciliate host Paramecium is such an intricate system and has evolved to such an extent that is offers hardly any clue on how it came into existence in the first place. Suspiciously, however, several other R-body-producing Caedibacter endosymbionts of paramecia lack plasmids yet electron micrograms of their R-bodies show them, more often than not, surrounded by and attached to capsomer-like structures and capsids of helical or icosahedral phages, some of the latter with tails (Figures 3+4). Phages of endosymbionts are intriguing in their own right but it has not yet been demonstrated that they carry the genes required for R-body production. Yet, if indeed such (pro)phages were responsible for R-body production, it would not be too far fetched to speculate that R-bodies might have been evolved as a means for the phages to "leave" their host. Stabbing him with a stiletto to this end seems costly and unwieldy compared to the elegant – but equally brutal – method other phages employ: punching holes in their hosts' membranes and cell walls using holins and lysins, respectively.
Outro Venerable research objects *), R-bodies may nevertheless have a promising future in applications, medical applications to be precise. Polka and Silver speculate, for example, that not so far from now R-bodies could be packed into epitope-tagged artificial vesicles together with cytotoxins as cargo, to be delivered by the bloodstream specifically to cancer cells for endocytosis. Once inside the target cell within a food vacuole that acidifies over time, this lowering of the pH would trigger un-rolling of the R-body into its stiletto form. Piercing the membranes of both, of the vesicles and concomitantly of the vacuole, would release the cytotoxin cargo directly into the cytoplasm of the cancer cell; no 'collateral damage' as with chemotherapeutics usually occurs. These researches have already characterized mutant R-bodies that un-roll at different pH values, thus providing a solution to problems that could arise from targeting less acidic vacuoles, e.g., phagosomes of certain types of macrophages. Turning a stiletto into a micro-scale scalpel, or lancet, remains a formidable task, but once achieved it could challenge micro-electro-mechanical systems (MEMs), the darlings of nanotechnology. There's a caveat, however: biochemists would have to find out how to improve the biodegradability of R-bodies – think of the known nasty impact on cells of carbon nanotubes and asbestos fibers, the finest of which are within the size range of the R stilettos.
*) and also venerable guests in STC ! Elio pointed out to me that R-bodies were first mentioned in the 16th post of this blog, back in 2007. The R-bodies mentioned there are made by epibionts of the ciliate Euplotidium, and which belong, as judged by their 16S rRNA signature, to the Verrumicrobia, while the R-body producing Caedibacter endosymbionts of Paramecium belong to the very distantly related Gammaproteobacteria. It is presently not known to which extent these R-bodies are related – if at all – despite their structural similarities.
Christoph Weigel is lecturer at the Life Science Engineering faculty of HTW, Berlin’s University for Applied Sciences and an Associate Blogger for STC
Polka JK, Silver PA (2016). A Tunable Protein Piston That Breaks Membranes to Release Encapsulated Cargo. ACS Synth Biol, 5 (4), 303 – 311 PMID 26814170 (Free Article)
Pond FR, Gibson I, Lalucat J, Quackenbush RL (1989). R-body-producing bacteria. Microbiol Rev, 53 (1), 25 – 67 PMID 2651865 (Free Article)