It's About Time to Consider Osedax

by Christoph
 

Figure 1. A Female O. frankpressi  dis­sec­ted from a whale bone, showing greenish root tissue and white ovaries. The plume-like gills connect to a muscular trunk, which can be withdrawn into a trans­parent tube when the worm is disturbed. Credit: Greg Rouse. Source, B Schematic of Osedax individual showing its com­part­mentalization. blue: basal trunk; yel­low: anterior ovisac; red: outer ovisac sheath; orange: inner ovisac tissue; green: posterior ovisac/ root. Source, Picture on main page: A squadron of Osedax worms chowing down on a whale bone. Credit: N.D. Higgs. Source

The Great Plate Count Anomaly, a recent post by fellow blogger Gemma Reguera, considered why so few of the bacteria and archaea that are readily observed under the microscope can be isolated and cultivated in the lab. Regardless of the gains achieved by the methodological tricks mentioned therein (and in the comments!), the culturing of isolated bacteria and archaea for in-depth studies in the lab still has to tackle the problem that many of them, maybe the majority, live in prokaryotic consortia or in symbioses with eukaryotes (a STC post from 2010 gives some clues). This 'long-read' here — three chapters, three detours, 3,360 words — is about a marine worm/bacterium symbiosis. Oh no, not the shipworms again, which are actually clams harboring symbionts in their gills, but true worms (Annelida) with symbionts in their 'roots' (it will become clear soon why borrowing a term from plant anatomy is fitting).
 

Part 1: Zoology   Way back in 2004, Rouse, Goffredi & Vrijenhoek published their finding, during deep-sea explorations, of large colonies of two mouthless and gutless worm species on the bones of a gray whale carcass at a depth of ~3 km in Monterey Bay, CA. They named the two Osedax rubiplumus (Fig. 3 A) and Osedax frankpressi (Fig. 1 A), respectively, and found, by phylogenetic analysis of the COX mitochondrial marker, that both had diverged from a common ancestor ~40 Ma (million years) ago. The svelte female O. rubi­plu­mus worms — named for their ~2 cm long, feather boa-like reddish gills — can reach body lengths of ~4 cm while having a XXS waist of ~3 mm. Despite their obvious elegance, Osedax (Latin for "bone eater") became instantly notorious on the web as "the zombie worms" (~1.2 x 10⁶ Google hits). That's clearly a misnomer as not only vodoo adepts know that zombies slurp brains, not bones.

Figure 2. Proposed model of acid secretion for bone dissolution in female Osedax. A O₂ is absorbed from sea water across the palps (=gills) and transported in blood and coelomic fluid to roots, where CO₂ is produced by aerobic respiration. H⁺ and HCO₃^– ions are produced and secreted or absorbed by the cellular mechanisms depicted in B. The secreted H⁺ dissolve bone hydroxyapatite, while the absorbed HCO₃^- is transported in blood and coelomic fluids to the palps/gills and secreted into sea water. B Proposed cellular mechanism for H⁺ secretion and HCO₃^– absorption in root epithelial cells. Metabolic CO₂ is hydrated by intracellular carbonic anhydrases (CA). H⁺ ions are secreted directly onto the bone by apical Vacuolar-H⁺-ATPases (VHA), dissolving bone hydroxyapatite and releasing trapped collagen and lipids that are absorbed and transported to symbiotic bacteria located in sub-epidermal bacteriocytes in the roots. HCO₃^– ions are absorbed and transported to the palps/gills for secretion. Source

Since then, some 30+ Osedax species/strains have been sampled from whale bones across all the oceans and, in several studies, from any depth wanted by intentionally exposing bones obtained from stranded whales as 'bait' at particular locations on the seafloor. Osedax muco­flo­ris, for example, was isolated this way after placing 'bait' whale bones at a depth of 125 m in the North Sea, At­lan­tic (see its portrait in a stunning video clip). Osedax thus appears to be a ubiquitous and rather 'speciose' (in pe­des­trian talk: species-rich) genus. Phylogenetic ana­lyses place Osedax within the worm family Siboglinidae, whose members are devoid of a differentiated digestive system. However, unlike other siboglinids, female Osedax lack a trophosome (a specialized organ housing chemo­syn­the­tic bacterial symbionts) and possess instead a bul­bous posterior ovisac covered by a sheath of greenish tissue that branches into a vascularized 'root' system that invades the bone marrow (Fig. 1 B). The ovisac and root systems of O. frankpressi and O. rubiplumus revealed numerous bacteriocytes (diameter 20 − 50 µm) with several vacuoles, each containing up to five large pleo­morphic, rod-shaped bacteria, likely symbionts (2.0 x 6.0 µm diameter in  O. frankpressi, 3.0 x 8.0 µm in O. rubi­plumus).

How to banquet on bones and how to get right into them in the first place? These questions were addressed in a recent study suggesting that the Osedax root epithelium absorbs bone cholesterol and collagen, which are meta­bolized by the symbiotic bacteria who, in turn, provide Osedax with nutrients — or serve as nutrients themselves, who knows. Osedax roots express high amounts of vacuolar-H⁺-ATPase (VHA), which is located in the apical membrane and in cytoplasmic vesicles of root and ovisac epithelial cells (Fig. 2). The enzyme carbonic anhydrase (CA), which catalyzes the hydration of CO₂ into H⁺ and HCO₃^–, is also expressed in roots and throughout Osedax's body. This suggests Osedax roots have massive acid-secreting capacity via VHA, fueled by H⁺ derived from the CA-catalysed hydration of CO₂ produced by aerobic metabolism, enough to solubilize bone's hydroxyapatite lattice and release the collagen and lipids contained within. The striking size of the prominent Osedax gills, or more precisely: their surface area, certainly points to an active aerobic metabolism. Apparently, the Osedax have completely 'externalized' all of their stomach, guts, and lungs.

Figure 3. O. rubiplumus males and females. A Adult female on bone photographed in an aqua­rium immediately after collecting. The plume has contracted from the normal con­dition in situ and has retracted slightly into the transparent tube surrounding the trunk. A harem of microscopic males lies next to her trunk in the lumen of the tube (arrow). B A harem of males attached to the transparent tube after removal of the female. C The anterior trunk of a female, showing a harem of males lying adjacent to her oviduct. D Four males from a single harem, illustrating the extent of size variation among males. The smallest male is still full of yolk, but the midsized specimens have optically refringent yolk granules and spermatids. The largest male has no obvious yolk granules. Source

And what about the males? Only female Osedax have been mentioned so far because in the reproductive cycle males play a — literally — minute role. Males are recrui­ted by mature sessile females from the larval pool through a mechanism biologists termed 'environmental sex determination' (ESP) and which is not uncommon among sedentary deep-sea animals. However, the actual 'trigger' substance(s) for ESP in Osedax is not yet known. 'Masculinized' larvae (0.4 x 1.1 mm) retain morphological traits typical of siboglinid trochophore, a distinct larval stage, throughout their short lifes (Fig. 3). These dwarf males accumulate in large numbers — up to 100 — in the gelatinous sheet that covers the female Osedax's body and oviduct. They don't seem to parasitize the females, but it is not known whether they are actively fed either. Females are strikingly prolific and produce eggs con­ti­nu­ous­ly at high rates. Therefore, the simplest hypothesis is that males deplete their yolk reserves to manu­fac­ture sperm and then die, while new males are con­ti­nu­ous­ly recruited from the larval pool. Yet — male readers rejoice! — an Osedax species was discovered just recently with males and females having much the same size and body differentiation (Fig. 4). Because the authors ob­ser­ved that these males can extend their body to ten-times its contracted state to reach out for females, this species was given the telling name Osedax priapus. Think of Linnaeus who was also all but shy about naming various species and their anatomical parts with references to genitalia. Phylogenetic analyses revealed O. priapus to be a derived species, and, therefore, the absence of dwarf males represents a 'character reversal' (i.e., the re-establishment of an ancestral character state) for this genus.
 

Figure 4. Close-up view of a male O. priapus. Scale bar 0.3 mm. Credit G. Rouse. Source

Detour #1   Whale carcasses are so immense (~40 t for a grey whale!) that, as they decay, they can be studied as a distinct habitat for a plethora of pro- and eukaryotes involved in a clearly observable succession of "de­com­po­sers" (don't miss this playful animation from the Smith­sonian, as a primer). The Osedax enter this succession late when just the bare bones are left. It takes a fully developed Osedax colony probably a decade to consume an entire whale skeleton (see picture on STC's main page for this post). But it's a finite resource and an extremely rare one, too. Even before excessive human whaling during the last three centuries markedly depleted the oceans of the larger species — not to mention the ocean's depletion of sea turtles, sharks, codfish et cetera, by the way — whale carcasses scattered across the vastness of the oceans were scarce as oases in the Sahara desert, only more so.

As adult femal Osedax are sessile and no longer wander, it's up to the fertilized eggs or the larvae to reach out for new food resources. Metabolites diffusing from a decomposing whale carcasse at the ocean's bottom could potentially serve as attractants for Osedax larvae as well as their to-be symbionts in the pool of free-living bacteria (see part 2). Clearly, such metabolites would be ra­pid­ly diluted below any reasonable level of detection by underwater currents. Using a metabolite gradient as guide to the closest 'oasis' would require, in addition, active swimming against this gradient, i.e., against the prevailing current — a task likely too energy-consuming under conditions of deficient food supply. This would hit particularly hard on Osedax larvae that paddle around using their cilia while consuming their yolk in so doing. In fact, O. japonicus larvae in the tro­cho­pho­re stage survive for no longer than 10 d, at least under lab conditions. As velocities of up to  20 − 40 cm/sec have been measured for some horizontal water currents at the sea bottom, such strong currents could — neglecting convection — transport Osedax offspring some 300+ km within 10 d. Conservative estimates by Smith & Baco suggest that within the North Pacific gray-whale range, whale falls occur annually with an average nearest neighbor distance of <16 km, accounting for approximately one carcass every 5 − 20 km². It thus seems just about feasible for Osedax trochophore larvae to ride the currents to a distant food source. From the ocean wide distribution of the various Osedax species, in shallow and deep waters alike, we know that they cope with these tasks, but we simply don't know exactly how.
 

Part 2: Microbiology   To identify the symbionts in the bacteriocytes, Shana Goffredi and co­workers performed 16S-rRNA phylotyping on dissected O. rubiplumus and O. frankpressi samples. In ovisac and root tissues, the tissues particularly enriched for bacteriocytes (Fig. 1 B), they found two predominant ribotypes (=90% identity) both of which cluster with Oceanospirillales (facultative anaerobic heterotrophs and members of the Gammaproteobacteria) in phylogenetic analyses. The Oceanospirillales-like ribotypes differed slightly between the O. rubiplumus and O. frankpressi samples, indicating that the sym­bionts of both worms, although closely related, were not identical. Besides these two predominant ribotypes, they found lesser numbers of ribotypes clustering with Epsilonproteobacteria (4 − 9%), Cytophaga-Bacteroides bacteria (0 − 4%) and Fusobacteria (2 − 4%). But the ribotypes of these three groups were also found in Osedax-free bone tissue samples, thus they might represent unspecific epibionts rather than true symbionts (or simply contaminations from the sample preparation step, as the authors note). The surprise was that the Osedax sym­biont phylotypes differed greatly from the sulfide- or methane-oxidizing chemoautotrophic sym­bionts found in other siboglinid worms, and which are in many cases members of the Thiotrichales branch of the Gammaproteobacteria; Thiotrichales are, at best, weekly related to the oce­ano­spi­ril­la­les.

The authors looked for possible trophic interactions between the symbionts and their worm hosts. With respect to nutrients, whale bones are mainly composed of lipids such as cholesterol and pro­teins such collagen. They analyzed lipid metabolites (such as FAMEs) in Odedax tissue samples and found their patterns to be highly similar for root/ovisac tissue and bacteriocyte-free crown/trunk samples. However, both kinds of samples differed completely from whale bone samples. The authors concluded: "the worm is getting bacterial fatty acids from the symbionts rather than from free-living microbes in the environment because there is an enormous dif­fer­en­ce in the diversity and make-up of bone versus symbiont fatty acids". Complementing the lipid analyses, Goffredi et al. found significantly more collagenase activity in root/ovisac tissue than in crown/trunk tissue. Similarly, zinc, the metal at the active-site in most collagenases, was found to be highly enriched in root/ovisac tissue as compared to crown/trunk tissue. But since col­la­gen­ase(s) might also be expressed by the host, a clear indication for symbiont-dependence of the Osedax host on symbiont col­la­ge­na­se(s) could not be obtained nor could cooperation with reci­procal benefits be excluded.

In a follow-up study, the same researchers found that fertilized eggs are consistently symbiont-free, suggesting that the Oceanospirillum symbionts are not transmitted maternally — as are the Buchnera endosymbionts of aphids — but aquired directly from the bone substrate by settling female larvae or juveniles. Also sequencing of the genomes of the two symbionts, Rs1 and Rs2, revealed that this Osedax/Oceanospirillum symbiosis is more of a loose partnership than an intimate relationship as there are no signs of genome reduction in the symbionts. The genome size of both symbionts, ~4.5 Mb, is typical for free-living Oceanospirillales (e.g., Neptunomonas japonica 4.3 Mb). And just to give you a 'flavor' of their similarity: the Rs1 and Rs2 DnaA proteins are 92% identical — exactly the value you find for the DnaA proteins of E. coli and Y. pseudo­tuber­cu­losis, while the DnaAs of E. coli and S. enterica are 96% identical. Since Escherichia, Yersinia, and Salmonella are considered separate genera within the Enterobacteriaceae family, the same dis­tinction should apply for the Osedax symbionts within the Oceanospirillales family. Such phy­lo­genetic intricacies can't reveal any secrets of this worm/bacterium symbiosis, but they underpin that what drives the symbiosis are not species-specific trait but rather traits common to the entire Oceanospirillales family.

Figure 5. Oxygen concentration profile in sea­water. Oxygen values are in micromoles (µmol) per kg of seawater; one micromole = 32 µg of oxygen. Source

How might the symbionts profit from living inside Ose­dax bacteriocytes? A better oxygen supply perhaps? Unlike what you might guess, sea­water at great depths is not depleted of oxygen. In most regions of the oceans it's far from anoxic and even more oxygen-rich than sea­water at depths of ~0.5 − 1 km (Fig. 5). It is a com­pa­ra­tively cozy habitat for facultative anaerobic heterotrophs such as the Oceanospirillales, despite temperatures being low (2 − 4°C) and nutrients scarce. While the upper 'oxy­gen-minimum zones' (OMZ) that occur in the oceans worldwide are oxygen depleted by the decay of organic matter from highly productive surface waters, constant oxygen supply at the sea bottom is apparently provided by currents of cold, oxygen-rich water moving from the poles towards the equator as part of the huge thermohaline circulation. However, occupation of the highly oxygenized Osedax tissue (see part 1) might still be attractive to the symbionts for boosting their metabolism, and thence their proliferation. But can they 'escape' from their host's bacteriocytes? That's not known.

Figure 6. A Effect of temperature on the growth rate of P. piezophila YC-1 under 60 MPa, equi­va­lent to the pressure at a depth of 6,000 m. B Growth rate of P. piezophila YC-1 under diffe­rent pressures at 8°C. C Scanning electron mi­cro­graph of P. piezophila YC-1 cells cultivated in marine broth 2216 at 8°C under pressures of 20 MPa, D ...of 60 MPa (atmospheric pressure at sea level = ~0.1 MPa). Bars, 2 µm. Source

If a symbiosis is difficult to study in the lab — and that's indeed the rule rather than the exception — one might try to assess gains and losses for the bacterial partner by studying the lifestyle of free-living siblings or close re­la­tives. A free-living Oceanospirillum, a close relative of the Osedax symbionts, was recently found at 6 km depth in the Puerto Rico trench and, given some ingenuity, could be cultivated in the lab. The appropriately named Pro­fun­di­mo­nas piezophila (piezophilic: synonymous with 'baro­philic' and meaning pressure-loving) is well adapted to a life in the Atlantic's abysses as it grows best — although still very slowly compared to enterobacteria — at re­fri­ge­ra­tor temperatures and cannot grow at pressures lower than ~20 MPa (Fig. 6 A,B). Like the Osedax symbionts, P. piezophila is a facultative anaerobic heterotroph able to metabolize hydrocarbons. Interestingly, P. piezophila cells show different morphologies depending on the pressure applied during growth (Fig. 6 C,D), which may be related to the unusually high amounts of mono-unsaturated fatty acids (MUFAs) in their membranes needed to keep them sufficiently fluid under high pressure — a particularly telling example of bacterial adaptability. Compared to their free-living cousins, the Osedax symbionts might profit from the H⁺-driven solubilization of the rich lipid (carbon) and collagen (carbon, nitrogen) resources in the whale bones by their host (Fig. 2), albeit at the price of having to share the bounty.

The above two paragraphs can't satisfactory explain the perks for the Oceanospirillales in their symbiosis with Osedax. But they point to a possible explanation for why Osedax species/strains did not recruit any of the various Oceanospirillales to permanently populate their bacteriocytes. The Osedax colonize not only whale bones but other marine vertebrate bones as well, although with less voracity, and they do that at greatly varying sea depths. It might actually be more ad­van­ta­geous to hire symbionts under fixed-term contracts rather than give them tenure, recruiting them each time from the pool of Oceanospirillales thriving at each new locale and thus already well adapted to the particular local conditions (type of food source, pressure, availability of oxy­gen, etc.). One would like to know, though, how Osedax distinguish between suitable Oce­ano­spi­ril­lales present on carcasses and inept ones. Perhaps they 'evaluate' the Oceanospirillum's col­la­ge­na­ses that are usually either membrane-bound or secreted. These could prepare an appetizing microenvironment for a juvenile Osedax in search of a richly laid table on a bone where it will then dine for the rest of its life. That's mere speculation, however. But one would also like to know if — and if yes, then how — the Osedax lure "their" Oceanospirillales into symbiosis or whether they simply grab them.
 

Detour #2   Under the microscope, Goffredi et al. observed what appeared to be actively dividing symbionts in several secondary vacuoles within bacteriocytes. This reminded me that, in the 16S rRNA-based phylogeny, the Oceanospirillales are a sister clade of the Pseudomonadales and might, therefore, share common features in their regulation of chromosome replication. From this viewpoint, (some) molecular geneticists were always puzzled by the observation of dual replication origins in P. aeruginosa and most other pseudomonads (see my earlier post). This 'indecisiveness' is apparently in the process of being settled among the Oceanospirillales: while Marinomonas posidonica has a dual oriC the same as P. aeruginosa  — one upstream of the dnaA gene, the other at a distance of ~6 kb, upstream of the gidA gene — Hahella chejuensis has only the single oriC located upstream of dnaA. In the Osedax symbionts Rs1 and Rs2, however, a single oriC is present, located upstream of gidA, as in all later branches of the Gammaproteobacteria (author's own observation). That's not quite the traditional way to map phylogeny yet fun.
 

Detour #3   While the Osedax symbionts assist their host in metabolizing lipids and collagen, other members of the Oceanospirillales family — like the cultivatable Neptunomonas naph­tho­vo­rans isolated from a contaminated site in shallow water sediment in Puget Sound, WA and a 'first cousin' of the Osedax symbionts — can sustain growth even on complex polycyclic aromatic hy­dro­carbons (PAH). No wonder Oceanospirillales were recruited to the microbial cleaning squad that tacitly eliminates the huge hydrocarbon plume drifting in the Gulf of Mexico since the Deep­water Horizon oil spill in 2010. Complete clean-up will still take a while, though.
 

Part 3: Paleobiology   For some years, researchers assumed that the Osedax split from their Siboglinidae relatives and further diversified at least ~45 Ma ago during the Eocene — possibly coincident with the origins of large Archeoceti cetaceans, the ancestors of the extant balean whales — or as much as ~125 Ma ago in the Cretaceous, when they could have lived on and from the bones of large marine Mesozoic reptiles. A narrower time frame was impossible to achieve by phylogenetic studies, given the inherent problems with calibrating 'molecular clocks'.

Figure 7. A CT reconstruction of Plesiosaur bone (semi-transparent) with two Osedax borings reconstructed in orange. B,D Boring-1 closeup in situ and digitally dissected. C,E Boring-2 closeup in situ and digitally dissected. Scale bar is 1 cm and scale meshes have spacing of 1 mm. (See D + E by clicking on the picture). Source

Enter Danise & Higgs with their clever adaptation and refinement of an already successful approach previously used to identify Osedax-type bore holes in fossilized bones — sort of "negative fossils" — by comparing them to Osedax bore holes in present-day samples. Their method: micro-computed tomography (Micro-CT), which quantitatively determines the morphology of sub-surface bone structures and generates their three-dimensional reconstruction. Their studied objects: ~100 Ma old fossils including a Plesiosaur humerus (Fig. 7) and two bone fragments from a marine turtle of the Chelonoidea family, from a rib and costal plate (part of the carapace), respectively. It is worth mentioning that they looked meticulously for cavities with single bore holes as only these reflect the shape of the root system of individual Osedax animals. An aside: similarly, virologists and phage aficionados prefer to study well-separated, single pla­ques rather than look at agar plates showing con­flu­ent lysis. The intact borings identified in the Plesiosaur and sea turtle fossils were relatively small but their sizes and morphologies were well within the range of known Osedax borings. In modern Osedax, the morphology of the borings by individuals from a single species are consistent within the same bone (but differ between bone types), which suggested to them that each of the fossil bones had been colonized by a single species.

Since the marine Plesiosauria were among the victims of the Cretaceous mass extinction, the Ose­dax family had apparently already established their particular lifestyle in these early times and relied on a diet of Chelonioidea carcasses (sea turtles, a family that escaped the mass extinction) and smaller bony fish corpses until the arrival of the cetaceans much later. "The increasing evi­dence for Osedax throughout the oceans past and present, combined with their propensity to rapidly consume a wide range of vertebrate skeletons, suggests that Osedax may have had a sig­ni­fi­cant negative effect on the preservation of marine vertebrate skeletons in the fossil record", Silvia Danise said in a press statement, and continued: "by destroying vertebrate skeletons before they could be buried Osedax may be responsible for the loss of data on marine vertebrate anatomy and carcass-fall communities on a global scale. The true extent of this 'Osedax effect', previously hypothesized only for the Cenozoic, now needs to be assessed for Cretaceous marine vertebrates".
 

Encore   I wonder — now drifting slightly into STC's Talmudic Questions department — whether members of the Osedax family were among the passengers of Noah's Ark, together with Oceano­spirillum stowaways. If so, it wouldn't be that surprising after all that creationists can easily deride biologists for so often failing to present fossils of presumed evolutionary intermediates (Thanks to Merry Youle for this twist).