Monday, March 1, 2010

Fish, and plankton, and sea greens, and termites from the sea!

shipworm damage; photo CC flickr user alaskaent

Wood makes for a pretty tough meal. To start with, cellulose -- wood’s primary constituent -- is indigestible by most organisms. If that weren’t bad enough, the cellulose itself is packed in a dense matrix of other intransigent biopolymers like hemicellulose, pectin, and lignin, making it structurally difficult even to get enzymes in and working. And to top it all off, wood contains very little nitrogen -- about an order of magnitude less than most organisms need to grow, relative to carbon.

With such a difficult hurdle to overcome, it’s pretty strange that one of the world’s champion wood-eating animals lives in an environment where no wood grows (the ocean) and belongs to a class that has never even evolved teeth (the bivalves). Thus the somewhat unfortunately named shipworms (family Teredinidae) are something of an enigma. 

Of course, if you’ve ever walked along a beach at low tide, you know there’s plenty of wood in the sea. And if you’ve ever collected some of that driftwood and made a fire, you can attest to the amount of energy stored within, if you can only get to it. Much to the consternation of boat owners, the Teredinidae are excellent at getting to it. Although they’re actually a type of clam, the Teredinids do look a lot like worms, with greatly reduced valves and elongate siphons. They use a modified foot to slowly bore into exposed wood, ingesting the resulting shavings. 

Their secret, of course, is a partnership with mutualistic bacterial symbionts. Much like cows, rabbits, and termites, shipworms host microbes that seem to be both the source of the cellulytic capabilities and much of their nitrogen.

These microbes were first detected via electron microscopy of an organ called the Gland of Deshayes (Popham and Dickson, 1973), which runs along the gill and is unique to the Teredinids and had previously been speculated to have some sort of digestive role. Soon thereafter, extractions from shipworms were shown to have both cellulytic and nitrogen-fixing capabilities (Carpenter and Culliney, 1975), and eventually a strain of N2-fixing, cellulose-digesting bacteria was isolated from the Gland of Deshayes in Woods Hole’s laboratory population of the shipworm Lyrodus pedicellatus (Waterbury et al, 1983). Repeated culturing attempts demonstrated that microbes of this type could be cultured from many species of shipworms. In 1991, Dan Distel, along with Ed DeLong and John Waterbury, used newly-developed molecular techniques to demonstrate that this bacterium did, in fact, reside in the shipworm gill, and was not just a chance environmental contaminant. Eventually, this bacterium would be formally described asTeredinibacter turnerae, named after Ruth Turner, an expert on shipworms and one of the first tenured female faculty at Harvard (Distel et al, 2002). The case seemed largely closed -- shipworms formed an association with a specific lineage of gamma-proteobacteria to inhabit a xylophagous (wood-eating) marine niche. 

But, as is almost always the case, the microbial world proves to be much more complicated upon closer inspection. Using a PCR- and cloning-based approach, Sipe et al (2000) found evidence of a distinct proteobacterial symbiont inhabiting the gills of the Pacific shipworm Bankia setacea. This microbe was 95% identical at the ribosomal 16S gene to T. turnerae, and fluorescent in-situ hybridization (FISH) using specific probes demonstrated that it could be found both in Bankia gills and in surface-sterilized ova. Localization of symbionts to ova suggests a vertical transmission through the germ line, a transmission strategy found in many other endosymbioses, such as aphids and vesicomyid clams. However, vertical transmission also tends to accelerate mutation rate  in the symbiont genome, leading to quick sequence divergence between symbionts of different host species; as this is not seen in T. turnerae, for which identical 16S sequences have been recovered from multiple host species, a horizontal mode of transmission (i.e. acquired from the environment each generation) has always been assumed. How could two relatively closely-related host species, with similar lifestyles and relatively closely related symbionts, have such different lifestyles?

Two years later, Dan Distel went back to the Woods Hole population of L. pedicellatus shipworms for some clues. Sipe et al. had noted that FISH using the Bankia-symbiont-specific probe was dimmer than with a probe designed to hybridize generally to shipworm symbionts, but attributed this difference to endogenous differences in hybridization efficiency of the two probes. An alternative explanation would be the presence of additional strains of symbionts that had gone undetected in their clone screen; and in fact, this is what Distel et al. (2002) found in L. pedicellatus.

This time, rather than simply amplifying and sequencing the symbiont they expected to find (T. turnerae), Distel et al. sequenced 43 cloned 16S sequences amplified with ‘universal’ bacterial primers. Only one of those sequences turned out to be identical to T. turnerae. The great majority of the rest were, like the Bankia symbiont, closely related but not identical to T. turnerae; these fell into three groups, with T. turnerae representing the fourth. Equally surprisingly, FISH showed that the four symbiont genotypes segregated nonrandomly into different bacteriocytes within the host gill: types 2 and 4 were sometimes found within the same host cell, as were types 1 and 3; but never any other combination. It was becoming clear that the shipworm-bacterial association, which had long been thought to be between the host animal and a monoculture of a single bacterial type, was actually a much more diverse system. 

That diversity was driven home when Yvette Luyten et al (2007) performed a more exhaustive and quantitative survey of symbionts from a number of L. pedicellatus individuals from the Woods Hole population. Their survey returned 18 individual ribotypes, clustered into 5 phylogenetic clades. Using two quantitative techniques, they were also able to describe the abundances of these 5 types in a number of individual worms. As in the previous study, the ribotype most closely allied to the sequenced T. turnerae was present only in very low abundances, with the bulk of the symbiont population being composed of bacteria from two related clades. Somewhat surprisingly, symbionts from the two dominant types were rarely found in abundance together in the same individual; in contrast, the low-abundance T. turnerae type was found in almost every individual surveyed. 

So what’s going on here? It seems clear that there are non-random differences between the various symbionts strains found in L. pedicellatus, which may reflect ecological interactions between the symbionts (the various strains inhabit different niches, or in the case of the non-overlapping strains, compete for the same niche), differences in relationship between host and symbiont (different host genotypes have constitutively different affinities for different symbiont strains), or simply represent historical contingency (some symbiont strains are transmitted vertically, and different populations simply represent the different histories of that host lineage). It is still unclear whether the different symbiont types found in L. pedicellatus have different transmission modes, as one might hypothesize based on comparison with the Sipe et al (2000) study on Bankia. Furthermore, while intra-host symbiont diversity allies the shipworm mutualism aesthetically to the diversity found in coral symbioses, the question of whether, as in the coral system, there are physiological and ecological consequences to symbiont type remains unaddressed in the literature. Key elements of the partnership’s physiology, such as how nutrients are transmitted between partners, or even how the cellulytic capabilities of an intracellular microbe in the gill are utilized to degrade wood in the gut, remain unknown.

The great genomic revolution in biology offers some hope here. The diversity of microbial symbionts of L. pedicellatus, by virtue of having one of their number in pure culture, offer an exciting opportunity for the kinds of genomic analyses that have been so useful in studies of insect nutritional endosymbionts. T. turnerae was recently sequenced (Yang et al, 2009), and the genomic analysis offers some intriguing insight into the microbe’s life history. Unlike the aforementioned vertically-transmitted aphid symbionts, the T. turnerae genome shows none of the hallmarks of obligate transovarial endosymbiosis (genome reduction, decrease in %G+C, loss of mobile genetic elements); in fact, the genome contains a number of functional antiviral CRISPR loci, transposable elements, and phage sequences, lending support to the idea that this strain at least maintains part of its life-cycle ex host. An impressively large proportion of the genome also appears to be devoted to secondary metabolite synthesis, which may be involved in communication with the host (as in rhizobial nod factors) or microbial interactions. Is it possible that T. turnerae, far from being a shipworm friend, is actually a recidivist parasite, taking more than it gives?

Although we still know relatively little about the shipworm symbiosis, what we do know is tantalizing. Does T. turnerae represent an ‘ancestral’ symbiont type that never took the plunge to vertical transmission? How are niches divided among the other symbiont types? What genomic changes are associated with their different roles? Most of what is known about symbiont diversity is from the Woods Hole laboratory population of L. pedicellatus; what are the symbiont populations like in wild L. pedicellatus and its relatives? Why are shipworm symbionts housed intracellularly in the gill, rather than extracellularly in the gut like most other known xylophagous mutualisms? 

As with most good biological systems, every shipworm answer has raised a whole host of interesting new questions. And they’re not all purely academic: shipworms and their microbial accomplices cause billions of dollars of damage to ships, docks, and other marine structures every year. On the flipside, we’d really like to know how to harness all that cellulose laying around for our own nefarious energy use. 

Not that the academic questions aren’t fascinating. They are. Extremely so. 

Hmm... Postdoc? 

Next Week on Mutualism Mondays: CARPENTER ANTS!
Major questions:
  • How are different symbionts transmitted between shipworm generations?
  • Do different symbiont strains have different functional roles? Are all of them mutualists, or could some be parasitic?
  • Do the genomes of different symbiont strains reflect these differences in function and transmission?
  • What influences the population dynamics of symbionts within and among shipworms?
  • What is the symbiont diversity in natural shipworm populations, and in different species?
  • How did this interaction evolve? 
  • What is the ‘currency of exchange’ between the shipworms and their symbionts? How exactly do they help to digest cellulose?
Annotated Bibliography:

Bartsch. More about shipworms. Science (1931) vol. 73 (1894) pp. 418-420

Hilarious response from an American taxonomist (Paul Bartsch) to a counterpart speaking to the British AAS re: Bartsch’s classification of a number of Western Atlantic shipworm species as sp. nov. The Brit was a lumper, and Bartsch was a splitter, getting in a number of good hits against soft body morphologists as well.

Also has tragically, hilariously racist passage: “Personally, I do not see that it makes any difference whether there is one species or a thousand species of shipworms. Shipworms, except where cultivated for food, as in Siam, are like the Indians of old, all bad, and undesirable.” 
CARPENTER and CULLINEY. NITROGEN-FIXATION IN MARINE SHIPWORMS. Science (1975) vol. 187 (4176) pp. 551-552

First paper demonstrating nitrogen fixation in shipworms. Used acetylene reduction assay using extracts from shipworms of several species and stages of development. Juveniles showed higher rates of fixation.

Also reported culturing N
2-fixing bacteria from the cecum, but not the Gland of Deshayes. 

First paper properly describing the nitrogen-fixing and cellulose-degrading symbionts from the Gland of Deshayes. Suppose that community in Deshayes’s gland is essentially a monoculture of the isolated, gram-negative bacterium. Isolated from six species of shipworms.

Note that Capenter and Culliney never went on to describe the spirillum-shaped microbe cultured from the cecum, and suppose that it must have been a free-living bug that got eaten.

Dan, Ed, and John sequence 800bp of 16S from the cultured strains from Waterbury et al (1983), and then use FISH to demonstrate that they were localized to vesicle-looking things in the gill filaments. Note that they are gram-negative gamma-proteobacteria. 
Distel and Roberts. Bacterial endosymbionts in the gills of the deep-sea wood-boring bivalves Xylophaga atlantica and Xylophaga washingtona. Biol Bull (1997) vol. 192 (2) pp. 253-261

Identify microbial symionts in another family of bivalves. Suggest that gill symbionts of Teredinids are intracellular, and thus there must be a way for them to excrete their cellulases extracellularly to function as nutritional partners. 
Sipe et al. Bacterial symbiont transmission in the wood-boring shipworm Bankia setacea (Bivalvia : Teredinidae). Applied and Environmental Microbiology (2000) vol. 66 (4) pp. 1685-1691

Used PCR cloning and FISH to identify symbionts in a West Coast Teredinid. Amplified 50 clones, had same restriction pattern with HaeIII; about 4% different to the ‘usual’ teredinid symbiont. Were able to amplify from eggs and ovaries, as well, suggesting vertical transmission.
Distel et al. Coexistence of multiple proteobacterial endosymbionts in the gills of the wood-boring bivalve Lyrodus pedicellatus (Bivalvia : Teredinidae). Applied and Environmental Microbiology (2002) vol. 68 (12) pp. 6292-6299

Describes multiple 16S types within a single shipworm.

Cloned and sequenced a bunch of sequences from shipworm L. pedicellatus. Found four types, one of which was identical to the cultured sequence. But that was the lowest abundance (1/50 clones). FISH staining revealed interesting patterns, where types 4 and 1 were mutually exclusive, 3 occurred in lower abundancean many of the cells that 1 occurred in, but not all. 2 occasionally co-occurred with 4.

Generally, 4 and 2 were symmetrically distributed in dorsal and medial regions, adjoining the central axis. 1 and 3 were more widely distributed, in more lateral and ventral areas.
Distel et al. Teredinibacter turnerae gen. nov., sp nov., a dinitrogen-fixing, cellulolytic, endosymbiotic gamma-proteobacterium isolated from the gills of wood-boring molluscs (Bivalvia : Teredinidae). Int J Syst Evol Micr (2002) vol. 52 pp. 2261-2269

Formal description of bacterium isolated from L. pedicellatus by Turner et al. 
Luyten et al. Extensive variation in intracellular symbiont community composition among members of a single population of the wood-boring bivalve Lyrodus pedicellatus (Bivalvia : teredinidae). Applied and Environmental Microbiology (2006) vol. 72 (1) pp. 412-417

Used quantitative methods (SSCP [aka CDCE] and QPCR) to assess the diversity of symbionts in a number of L. pedicellatus from the Woods Hole culture line. Found five clades of ribotypes (most but not all have short internal branches), with a strongly skewed distribution -- 100 of 163 clones were ‘P3’ type; T. turnerae groups with a much less derived clade with a really long branch compared to the others, and which was only in 2 clones. Most other ribotypes were very low abundance (1-2 clones).

Phylotypes P1 and P3 were pretty much mutually exclusive dominant types. P4 was in low anundance in almost all; method probably can’t detect members < 1%, though.  P1 and P3 were observed in Distel Beaudoin and Morrill 2002 to co-occur in bacteriocytes, and thus may compete for same space. Note that free-living bacterioplankton and sediment communities have similar compositions.

Note that all phylotypes are from same clade (unbroken by other, environmental samples) and thus may represent a single evolutionary colonization of the worms... Why the subsequent diversification? 
Lechene et al. Quantitative imaging of nitrogen fixation by individual bacteria within animal cells. Science (2007) vol. 317 (5844) pp. 1563-1566

Use MIMS to show nitrogen fixation in symbiont cells within a L. pedicellatus gill filament after being exposed to
15N-enriched N2 for 8 days. Also demonstrated transport of fixed nitrogen to non-symbiont-bearing shipworm tissues.
Yang et al. The Complete Genome of Teredinibacter turnerae T7901: An Intracellular Endosymbiont of Marine Wood-Boring Bivalves (Shipworms). PLoS ONE (2009) vol. 4 (7) pp. e6085

Sequenced the genome of T. turnerae. Found lots of cellulases! Also a weird insertion of ORFs into the ITS1 of a ribosomal operon. Unusually, also a lot of enzymes with multiple catalytic domains. May serve to overcome the challenge of multiple necessary catalytic reactions in close physical proximity when chewing up wood.

Sequencing suggests Type 2 secretion system is important -- 20% of predicted proteins have N-terminal signal peptides which would lead to secretion. T3 and T4SSs are absent, though. Has T6SS, which may be important in translating proteins to host cells.

Retains cellobiose phyosphotransferase genes, suggesting that there may be a way that cellulose is making it up to the gills. OR that they’re maintaining on cellulose during an extracellular portion of the lifecycle.

Nif cluster probably acquired horizontally from a Pseudomonas.

Lots of big secondary metabolite domains -- 7% of genome! Not homologous to S. degredans, a closely related sequenced free-living bug; hypothesized to be laterally acquired. Maybe involved with microbial warfare with other symbionts, or signaling with the host?

Lots of CRISPR sequences, phage, and transposable elements. 


kw said...

Nice post, Jon! You've done a great job of making the content accessible to nonspecialists—like me!

Jon Sanders said...

Thanks! Let's call it 'preparing for two possible careers at the same time.' :D

Anonymous said...

good points and the details are more precise than elsewhere, thanks.

- Thomas

Anonymous said...

hello there owner found your site via search engine but it was hard to find and I see you could have more visitors because there are not so many comments yet. I have discovered site which offer to dramatically increase traffic to your site they claim they managed to get close to 4000 visitors/day using their services you could also get lot more targeted traffic from search engines as you have now. I used their services and got significantly more visitors to my website. Hope this helps :) They offer best services to increase website traffic at this website