Today for a change a movie review : The remake of the 1967 classic Le bal des vampires by Roman Polanski. Starring as Professor Abronsius is Géraldine Laloux from the de Duve Institute at Louvain University, Belgium and co-starring as Alfred, the assistant, is her postdoc Yoann Santin. Both also co-wrote the script and directed the film assisted by Adrià Sogues, Yvann Bourigault, and Han Remaut. They had casted Caulobacter crescentus as the innkeeper's pretty daughter Sarah, and as the villain Count von Krolock none other than Bdellovibrio exovorus.
Having bdellovibrios play the role of villains in microbiological vampire films is not new. Bdellovibrio bacteriovorus, which is usually cast in this role, was first described four years before Roman Polanski released his movie in 1967. Briefly, it attacks its Gram‑negative bacterial victims, slips into their periplasm, and then sucks them dry as it grows and propagates (more on its predatory lifestyle in the last section). Susan Koval & Sandra Hynes (1991) describe the isolation of a "predacious vibrio that possessed a single, polar sheathed flagellum", and continue that they "did not observe any intraperiplasmic growth of strain JSS on C. crescentus, an essential character for Bdellovibrio species and [shall] refer to this isolate as Bdellovibrio-like." This isolate was baptized Bdellovibrio exovorus JSST by Koval et al. (2013) and characterized as a predatory epibiont of C. crescentus that remains attached to the outside of its host − better: victim − during growth and subsequent binary fission.
All previous studies of B. bacteriovorus, B. exovorus and other Bdellovibrio-like organisms (BALOs) have failed to clarify why these predators have very different host ranges among Gram‑negative bacteria. Related to this is the question of how they actually attach to their victims, that is, what their receptors are. The S-layer, a porous self-assembling paracrystalline protein lattice that covers the whole cell and which many bacterial and most archaeal species possess, was thought to be a protective coating that keeps vampires at bay. But this has not been proven. I have gone into such detail here to make it clear that new studies often build on decades of earlier work. Santin et al. (2023) were aware of this and mentioned it, of course, when they drafted the screenplay for their remake of Le bal des vampires.
B. exovorus takes center stage Every moviegoer expects a vampire movie to include several scenes in which they can witness a bitten victim turning paler and paler. B. exovorus does not disappoint here, and you see in Figure 1 how tiny hungry vampires attack and attach to the larger, happily growing Caulobacter cells. As in the other images in the study by Santin et al. (2023), the Caulobacters are mostly predivisonal cells, indicating normal growth with the signature stalk at one pole of the mother cell and, on the other side of a visible constriction, the to‑be daughter cell ("swarmer cell") lacking visible flagella at the other pole (like pili, flagella are too thin to be visible under the light microscope). Once attached, B. exovorus sucks out a Caulobacter cell in about 4 hours.
In the upper panel of Figure 1 (phase), the Caulobacter cells successively turn pale ("ghosts"), and the lower panel shows that a fluorescence-labeled protein completely bleaches out of the cytoplasm of the Caulobacter victims without (re)appearing in the firmly attached B. exovorus cells. This indicates that this protein, mCherry, is degraded and only oligopeptides or amino acids are absorbed by the vampire. In a similar experiment, DAPI‑labeled DNA in the cytoplasma of a "bitten" Caulobacter cell disappeared even a tad faster than its proteinaceous content. You will notice that the pale Caulobacter ghosts (Figure 1, top panel) retain their typical crescent shape. When the researchers labelled the prey peptidoglycan (PG) with the fluorescent D-amino acid HADA prior to the addition of B. exovorus, they found that ghost cells were stained entirely, like uninfected cells, indicating that predation only marginally disrupts the victim's cell wall.
During shooting the hunting scenes in their remake, Yoann Santin and his coworkers also casted several Alphaproteobacteria from the wider Caulobacter family in the role of the victim and found that some, which I am not listing here, are also "vampirized." It looks as if the host range of B. exovorus is not as limited as previously assumed. And while they were still busy with the hunting scenes they surprisingly found that the reference strain C. crescentus CB15N, which expresses the rsaA gene forming a functional S‑layer, was similarly predated as the S layer-deficient CB15N ΔrsaA mutant strain. Thus, the S-layer does not seem to protect against vampires very efficiently, at least not against B. exovorus. Of note: the 27 Å‑wide pores of the Caulobacter S-layer appear to be wide enough to allow, for example, pili (⌀ ~10 Å) to pass through and reach the outer membrane (OM).
The vampire's teeth Santin et al. (2023) caught the vampires grabbing their prey in flagrante, (see movies below), but light microscopy cannot provide details of the attachment process, which molecular structures are involved. However, they got a closer look at its teeth using cryo‑EM (Figure 2). You see in the magnification (Figure 2, ii) that (1.) Caulobacter's S-layer is penetrated, (2.) B. exovorus's OM is drawn so close to the OM of Caulobacter that both appear fused, and (3.) Caulobacter's inner membrane (IM) shows signs of progressive fragmentation. Apparently, B. exovorus uses this structure − the authors call it a "feeding" junction − to gain access to the cytoplasmic contents of its victim. This happens very quickly when secreted lipolytic enzymes fragment its IM, and secreted proteases and nucleases turn the cytoplasm into a smoothie.
Such "feeding" junctions do not show any conspicuous structures in the cryo-EM images, but have a remarkably consistent diameter of 165±20 nm (48 cells) in two experiments with C. crescentus CB15N (S‑layer) and of 164±22 nm (12 cells) in the S layer-deficient CB15N ΔrsaA mutant. What seems interesting, but does not lead to learning more about their molecular composition, is that the vampire attacks its victim over the entire surface of the predivisional cells − including the stalk, although less frequently − and only avoids the flagellated pole of the (future) swarmer cell. It is known that this pole ("polar microdomain") is heavily crowded internally − see here its modeling by David Goodsell and Keren Lasker − but it is unknown whether the cell wall here is depleted of any (suspected) receptor(s) for the attachment of B. exovorus.
The astounding vampire triplets Contrary to the conclusion of Koval et al. (2013), B. exovorus does not appear to reproduce by binary fission. Santin et al. (2023) found that predivisional B. exovorus cells, short filaments, mostly displayed two constriction sites, indicating that these vampires rely on a non-binary mode of cell division (Figure 3, B). The majority of B. exovorus cells produced three daughter cells (76%), while two progenies were rarer (24%), and four progenies were only occasionally observed in an (artificially) elongated Caulobacter (popZ::Ω) host.
Time lapse observation with short time intervals (Figure 3, A) revealed that B. exovorus initiates cell constriction during growth, apparently well coordinated with cell wall synthesis for cell elongation. Release of progeny cells is sequential, where the outermost (distal) cell leaves first and the prey-attached (proximal) cell leaves last (marked 1−3 in white, in Figure 3, A). Two other interesting details can be seen very clearly in Figure 3, A. First, after being "bitten" by B. exovorus, the predivisonal Caulobacter cell instantly halts its normal cell cycle progression, that is, it "freezes" before division, and successively (suckessively?) turns into a "ghost". Second, and this can only be seen in the bottom panel (mCherry), the residual fluorescence that remains in the ghost after the vampire has detached leaks out (min 207−209). This may indicate that when the vampire detaches from its victim, it plucks a piece out of the victim's OM and takes it with it, leaving an unsealed scar.
The B. exovorus life cycle Based on the above, Santin et al. (2023) describe the life cycle of Bdellovibrio exovorus in its main steps in Figure 4. While there are some similarities, there are still notable differences from the life cycle of the closely related, better-studied Bdellovibrio bacteriovorus (see here in STC, and here for a diagram). Both bdellos actively hunt by swimming, propelled by flagella, and attach with pili to prey cells they encounter to sites opposite to the flagellated pole (Figure 4, 1.). Once firmly attached to a prey cell, B. bacteriovorus slips through the outer membrane (OM) into the victim's periplasm, while B. exovorus forms a tight junction between its own OM and the prey OM (Figure 4, 2.).
The secretion of specific enzymes enables B. exovorus to disintegrate the inner membrane (IM) and to digest prey contents in situ (Figure 4, 3.). As the attached predator cell grows, sequential constriction sites become visible along the filament (Figure 4, 4.). Once growth is complete, the filament sequentially releases the outermost and then the second progeny (Figure 4, 5.+6.). When finally detaching from the sucked-out prey cell, which remains behind as a "ghost" while retaining its shape, B. exovorus may pluck out part of its victim's OM and leave an open scar (Figure 4, 7.).
During the shooting of the movie Le bal des vampires, various rehearsal shots were taken, of course. The directors kindly provided two short takes as time-lapse. Can you follow the successive steps of the diagram in Figure 4 in these clips? I apologize for the lousy, pixelated quality of the clips, as I had to compress the large video files to show them in this size, 400×400 pxl, on STC's website. However, all the essential details, for example the Caulobacter stalks, can be seen, and flagella are not/hardly visible under the light microscope anyway.
Unlike B. exovorus, B. bacteriovorus modifies the cell wall of its victim and grows within the periplasm ("bdelloblast") while sucking-up the cytoplasm through the shrinking but intact IM. Within the belloblast, B. bacteriovorus grows as a filament with regular constrictions, the number of which − and thus the number of equally-sized progeny cells later − is largely determined by the size of the host cell, while the growth rate depends on the nutritional status of the host cell. Finally, the filament divides (non-binary!) into up to a dozen or so mature, that is, already flagellated progeny cells and the bdelloblast is popped to release them.
Isn't it fascinating to learn that two organisms with the same lifestyle and a very similar genetic makeup arrive at quite different solutions in detail to successfully practice vampirism? Exactly what one finds again and again in almost all of biology: distinctly different molecular, mechanistic, and regulatory solutions for the same pathway among organisms with similar genetic makeup.