SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The earthworm Eisenia fetida harbours Verminephrobacter eiseniae within their excretory nephridia. This symbiont is transferred from the parent into the egg capsules where the cells are acquired by the developing earthworm in a series of recruitment steps. Previous studies defined V. eiseniae as the most abundant cell type in the egg capsules, leaving approximately 30% of the bacteria unidentified and of unknown origin. The study presented here used terminal restriction fragment length polymorphism analysis together with cloning and sequencing of 16S rRNA genes to define the composition of the bacterial consortium in E. fetida egg capsules from early to late development. Newly formed capsules of E. fetida contained three bacterial types, a novel Microbacteriaceae member, a Flexibacteriaceae member and the previously described V. eiseniae. Fluorescent in situ hybridization (FISH) using specific and general rRNA probes demonstrated that the bacteria are abundant during early development, colonize the embryo and appear in the adult nephridia. As the capsules mature, Herbaspirillum spp. become abundant although they were not detected within the adult worm. These divergent taxa could serve distinct functions in both the adult earthworm and in the egg capsule to influence the competitive ability of earthworms within the soil community.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Earthworms, members of the Annelida, are conspicuous members of soil communities known for processing large amounts of soil and altering the organic content through the action of bacteria within the gut that primarily originate from the environment (Edwards, 2004; Drake and Horn, 2007). Members of the lumbricid earthworms also harbour specific bacteria within the excretory nephridia belonging to the recently described genus Verminephrobacter (Schramm et al., 2003; Pinel et al., 2008). Evidence indicates Verminephrobacter spp. occur throughout the Lumbricidae and are species-specific symbionts (Schramm et al., 2003; Lund et al., 2009; S.K. Davidson, R.J. Powell, D.A. Stahl and S.W. James, unpublished). Previous studies suggested that the earthworm nephridia primarily harboured Verminephrobacter spp. New evidence presented here reveals a more complex symbiosis in the composting earthworm Eisenia fetida involving three bacterial symbionts from divergent classes of bacteria.

The Verminephrobacter eiseniae symbiont is transmitted vertically to the next E. fetida generation by deposition into the egg capsules and the juveniles emerge with their nephridia fully colonized (Davidson and Stahl, 2006). After mating, eggs, sperm, albumin and bacteria are expelled through pores into a pre-capsule that forms around the external surface of the clitellum (collar). The pre-capsule then slides off the anterior of the worm and the chitin shell hardens (Grove and Cowley, 1926; Davidson and Stahl, 2006). The embryos may take several months to mature, depending on the species, protected from the complex and competitive soil community by a thin chitin shell. Within freshly deposited E. fetida egg capsules, previous studies determined V. eiseniae to make up the major portion of the bacterial population (60–70%) leaving the remainder of the population uncharacterized (Davidson and Stahl, 2006).

During the course of embryonic development, V. eiseniae cells are recruited to the nascent nephridia beginning about day 8–10 in a series of steps involving recruitment, migration and colonization of the correct nephridial tissue (Davidson and Stahl, 2008). Non-Verminephrobacter cells were observed to occur on the embryos surface, and increase in number and diversity within the developing gut. These other bacterial types could either be entrained from the soil, and incidental inhabitants, or function as consistently associated symbionts. In this study we determined the source, composition and changes in relative abundance of the major bacterial populations of the earthworm egg capsule using a combination of terminal restriction fragment length polymorphism (T-RFLP) analysis of the 16S rRNA gene (Avaniss-Aghajani et al., 1994; 1996; Liu et al., 1997),cloning and sequencing, and fluorescent in situ hybridization (FISH). The bedding, egg capsules and the adults were examined to determine the identity and likely source of the capsule bacteria. Within the egg capsules, the bacterial community was evaluated at initial formation, mid-embryonic development and near hatching to characterize changes in the populations.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Egg capsule bacterial community composition

Egg capsules are deposited by multiple worms in the bedding and collected for analysis. To evaluate whether bacteria in capsules originated from the bedding or the earthworm, and if changes in bedding bacterial profiles were reflected in the egg capsule community, a set of capsules and bedding were sampled and evaluated from bins that had been worked by worms for 6 days and 20 days. Healthy E. fetida egg capsules were collected, and DNA extracted at developmental stages 0–1 day (T1), 8–10 days (T2) and 17–18 days (T3) (Fig. 1). T-RFLP peak patterns and relative abundances of replicate PCR reactions of single DNA samples were highly consistent (varied only a fraction of a per cent, n = 3). The T-RFLP profiles obtained from egg capsules collected from either 6-day-old or 20-day-old bedding were not different (same peaks detected in similar ratios) and therefore grouped together in the final analysis.

image

Figure 1. Developmental time-course and stages sampled in this study for E. fetida. Time 0 is the time of deposition of the capsule into the soil. T1 (day 0–1), T2 (day 8–10) and T3 (day 17–18) are early, middle and late development samples taken for microbial analysis.

Download figure to PowerPoint

The T-RFLP pattern obtained from T1 capsules (no visible embryos) contained two dominant peaks and one minor peak (Fig. 2; Table 1). These fragments were identified in the clone libraries described below as V. eiseniae (66 bp), Herbaspirillum sp. (221) and a Microbacteriaceae member (228 bp) (Fig. 2; Table 1). The terminal fragment pattern for T2 and T3 egg capsules were significantly different from the T1 day patterns. A notable change was an increase in the initially minor Herbaspirillum sp. peak that became the dominant peak during mid- to late development. The Microbacteriaceae peak remained consistently around 25%, whereas the V. eiseniae signature decreased in relative abundance throughout development, comprising only about 20% of the total community signal near completion of development. The T-RFLP pattern in late development (T3), as embryos reach maturity, was more complex and variable due to the appearance of additional bacterial peaks, including an abundant 292 bp fragment. Fewer samples were analysed in later stages and the consistency of bacteria that appear near the end of the development has not been thoroughly evaluated.

image

Figure 2. Representative T-RFLP traces for early, middle and late development, and the earthworm bedding. Fragment peaks: 1, 66 bp fragment, Verminephrobacter eiseniae; 2, 221 bp, Herbaspirillum spp.; 3, 228 bp, a novel Microbacteriaceae member; 4, 292 bp, Sphingomonas sp. X-axis, fluorescence units; y-axis, time (fragment size).

Download figure to PowerPoint

Table 1.  Summary of relative peak abundance (%) in capsule T-RFLP profiles.
Peak identityT1 (n = 9)T2 (n = 5)T3 (n = 3)Bedding (n = 4a)
  • a.

    Total average of 12 PCR runs (3 each sample).

  • *,**t-test, significant difference; **T1 vs. T2, P < 0.01; *T2 vs. T3, P < 0.05.

  • n = number of DNA samples. Each DNA sample is extracted from three pooled capsules.

Verminephrobacter (66 bp)60 ± 1235 ± 11**20 ± 41.5 ± 0.2
Microbacteriaceae (228 bp)25 ± 630 ± 1125 ± 37.3 ± 3.1
Herbaspirillum spp. (221 bp)5 ± 620 ± 5**41 ± 3*3.4 ± 1.8
Unk peak 29204 ± 47 ± 41.3 ± 0.3
Other15 ± 1013 ± 108 ± 186.5

The 16S rRNA gene T-RFLP profiles from the bedding samples were highly complex, and not analysed in detail. A gross comparison of these data to the patterns observed in the egg capsules established that the T-RFLP peaks most abundant in the egg capsules were only minor components of the bedding population (Fig. 2; Table 1). This indicated that the major members of the egg capsule population were not passively entrained during formation, but either migrate from the bedding to the worm and into the mucus during capsule formation or originate from the worm. The V. eiseniae fragment appeared in only 50% of the bedding amplifications (n = 12) at ∼1% of the total population. The other egg capsule 16S rRNA gene fragments appeared at low levels in all the samples. The bedding worked by worms for only 6 days (19–27 fragments; average 23) had fewer fragments than the 20 day bedding (30–39 fragments; average 33).

Community composition by 16S rRNA gene clone library analysis

The composition of egg capsule communities at different developmental stages was confirmed by construction of two sets of 16S rRNA gene clone libraries for two different colonies of E. fetida, an established colony (2004) and a new shipment of E. fetida received in 2007. The results are summarized in Table 2. The diversity and identity of the clones in the libraries were consistent with the T-RFLP profiles. Clone libraries from the 2004 and 2007 colonies contained V. eiseniae, Microbacteriaceae and Flexibacter sequences matching 98–99% identity and the proportion of Herbasprillum spp. clones increased in libraries from T3 capsules relative to T1 (Table 2; two main phylotypes, Fig. 3C). Rare clones included a Flexibacter (11 clones total) member shown by other methods to be a common member of the capsule community. Both the 2004 and 2007 T3 (21 representatives were fully sequenced out of 152 clones) included several well-represented clones, notably Herbaspirillum spp. (75), Sphingomonas sp. (22) and Enterobacter spp. (14) (Tables 2 and 3). Based on 16S rRNA gene sequence, the Sphingomonas sp. likely represents the 292 peak that increased at the end of development although this was not confirmed by T-RFLP of the clone. A few taxa would not have appeared in the T-RFLP traces using the HaeIII digest (Flexibacter, Stenotrophomonas and a few of the Enterobacter spp.) because the resulting terminal fragments are less than 50 base pairs, but all others would have been detected if signals were above a threshold of 0.5%.

Table 2.  Proportion (%) of taxa in 16S rRNA gene clone libraries from E. fetida egg capsules.
Taxonomic affiliationPhylotypes; accession No.T1aT3T3
  • a.

    A total of 397 clones were screened: T1 (2007; 137), T1–T2 mixed age (2003; 60), T2 (2004; 44), T3 (2004; 77: 2007; 91). Only unique clones (45) were fully sequenced due to similarity of clones among the libraries.

Verminephrobacter eiseniae1; GU201577–803146
Microbacteriaceae1; GU201566–7048922
Flexibacter-like1; GU201552–55813
Herbaspirillum2; GU201556–6323949
SphingomonasGU201573–751189
All other (mixed) 102911
imageimage

Figure 3. Phylogenetic relationships of E. fetida egg capsule bacteria based on 1460 bp of the 16S rRNA gene sequence. (A) Actinobacteria, (B) Flexibacteriaceae, (C) Herbaspirillum-related. These are consensus topologies from maximum likelihood (ML), 1000× bootstrap resampling analyses. Branch lengths were determined from ML tree matching consensus topology. Nodes with < 50% support were collapsed. *MP bootstrap value also presented at nodes with disagreement between ML and MP support.

Table 3.  Bacteria found within E. fetida egg capsule clone libraries (excluding symbionts).
Clone name (capsule stage, year, GenBank accession No.)Taxonomic affiliationNext relative (sequence similarity)
EfT304_D09 GU201540Acidobacteria; AcidobacteriaceaeTerriglobus sp. TAA 48 AY587229 (98)
EfT304_E09 GU201541  
EfT107_F12 GU201542  
EfT107_A12 GU201543 Uncultured AY211077 (96)
EfT307_H10 GU201575Alphaproteobacteria; SphingomonadaceaeSphingomonas yabuuchiae AB071955 (98)
EfT304_C03 GU201573 Sphingomonas sp. SKJH-30 AY749436 (99)
EfT107_H08 GU201574  
EfT107_D08 GU201544Betaproteobacteria; AlcaligenaceaeBordetella petrii AM902716 (97)
EfT307_A05 GU201545  
EfT304_A08 GU201564OxalobacteraceaeJanthinobacterium sp. IC161 AB196254 (98)
EfT304_F06 GU201549Gammaproteobacteria; EnterobacteriaceaeEnterobacter sp. WAB-1926 AM184265 (99)
EfT304_G03 GU201548  
EfT307_D05 GU201565 Klebsiella ornithinolytica KOU78182 (99)
EfT307_C07 GU201572PseudomonadaceaePseudomonas lutea EU184082 (98)
EfT304_H09 GU201576XanthomonadaceaeStenotrophomonas maltophilia EU034540 (99)
EfT107_A11 GU201547Bacteroidetes;ChitinophagaceaeChitinophaga pinensis CP001699 (96)
EfT107_C10 GU201546 C. arvensicola AM237314 (99)
EfT107_F07 GU201551 Flavisolibacter ginsengiterrae AB267476 (92); FJ263932 (96)
EfT107_B07 GU201550  
EfT107_D05 GU201571PlanctomycetesNo close matches

Phylogenetic analysis of cloned sequences

The phylogenetic relationships of the most abundant taxa based on 16S rRNA sequences are presented in Fig. 3, with the exception of V. eiseniae, which have been presented previously (Schramm et al., 2003; Pinel et al., 2008; Lund et al., 2009). The capsule Actinobacteria member falls within the Microbacteriaceae with weak support for grouping most closely to Agromyces (Fig. 3A), but not clearly within characterized genera. The Flexibacteriaceae member clearly falls within this family, but without strong support for placement within a described genus (Fig. 3B). The Herbaspirillum spp. clones form two clusters, one within the Herbaspirillum genus, the other is affiliated with Herbaspirillum, without strong support for placement within Herbaspirillum among closely related Betaproteobacteria (Fig. 3C). Other sequences from the later developmental stages were members of the Proteobacteria (Table 3). These sequence data have been submitted to the GenBank database under Accession No. GU201540–GU201580.

Fluorescent in situ hybridization (FISH)

Probes specific for the Flexibacter member and Actinobacteria labelled bacteria in the ampulla of the nephridia of adult and juvenile E. fetida (Figs 4 and 5), and in the egg capsule albumin (not shown). A probe designed to bind the Flexibacter-like symbiont 16S rRNA labelled long filamentous cells extending from host cell surface, between the V. eiseniae cells into the lumen of the ampulla (4–8 µm long, 0.25 µm wide). These cells were observed in both archived freezer samples (fixed in 2003/2004) and newly acquired specimens, confirming that they were missed by earlier studies using FISH and 16S rRNA gene cloning. The filamentous cells were limited to a specific section of the ampulla and only the V. eiseniae were observed in the exit tubule of the ampulla (Fig. 4C). The Actinobacteria cells were not closely associated with the host cell surfaces (TEM cross-sections, not shown); far fewer cells were observed, and not clearly present in all nephridia within a single worm.

image

Figure 4. Bacterial cells within the nephridia. A. Diagram illustrating the location and structure of the earthworm nephridia (holonephridia). V, ventral; D, dorsal. Bacteria are located in the ampulla of the second nephridial loop. B–F. Laser scanning confocal microscopy (LSCM) images of bacteria within ampulla labelled by FISH with probes specific for Verminephrobacter (LSB 145, red; B–E), Flexibacter-like bacteria (Flexi 145, green; C–E) and Actinobacteria (HGC 69a, yellow; F). (B) Low-magnification image of ampulla counter-stained with Cell Tracker™ Green BODIPY® (CT) (Invitrogen™ Molecular Probes®); red, V. eiseniae cells.

Download figure to PowerPoint

image

Figure 5. Bacteria on the surface and colonizing the nephridia of developing E. fetida embryos. A–C. LSCM images of bacteria labelled by FISH with probes specific for Verminephrobacter (LSB 145, green; A, B), Flexibacter-like bacteria (Flexi 145, red; A, B) and Actinobacteria (HGC 69a, green; C). (A) Yellow arrow, co-aggregates of V. eiseniae and the Flexibacter-like symbionts on the surface of the embryo; fluorescence of both probes displays as yellow. White arrows, ampulla of nephrida colonized by both V. eiseniae and Flexibacter; white arrow head, ampulla (red) colonized by Flexibacter alone. Verminephrobacter eiseniae cells in the colonization canals are visible as green lines. Inset, enlargement of an aggregate. (B) Detail of nephridial ampulla as colonization progresses. (C) Arrows indicate location of Actinobacteria in pores; insets, magnification of surface (left) and cross-section of pore (right). D. SEM of bacterial cells on the surface of an embryo; arrows indicate three distinct morphologies. E. TEM of bacteria in the colonization canal entrance.

Download figure to PowerPoint

The timing of the Flexibacter and the Actinobacteria migration into the embryo in relation to Verminephrobacter recruitment was confirmed by FISH using specific probes to detect these cells during development as described in Davidson and Stahl (2008). The Actinobacteria partially migrates, and remains as a cell plug in the entrance to the nephridia (Fig. 5C, insets). The Flexibacter and V. eiseniae tend to aggregate together in the gut lumen, on the surface of embryos, and at the entrance to the infection canal (Fig. 5A). During the colonization process, the Flexibacter-like cells were often, but not always, found present in the ampulla prior to the V. eiseniae. The V. eiseniae cells were in turn observed to be much more abundant in the canal, although both cell types were present (Fig. 5A). However, simultaneous colonization was also observed and both are present after full colonization of the nephridia (Fig. 5A and B). By SEM and TEM, all three cell morphologies are found to associate with the surface of the embryos, and appear in the colonization canal entrance (Fig. 5D and E).

Detection of bacteria in E. fetida organs and body wall

After detecting the novel Actinobacteria within the capsules, but only sporadically within the nephridia, a survey by PCR was completed to locate these bacteria elsewhere within the adult worm. Using a specific PCR primer pair for the 16S rRNA gene, DNA extracted from E. fetida body wall, nephridia and internal reproductive organs were screened for the presence of the Actinobacteria. Body wall samples from the entire length of the worm (anterior, clitellum, middle and posterior sections) as well as the nephridia were positive for the Actinobacterium discovered in the egg capsules. The reproductive organs were negative. The cloned and sequenced products confirmed the identity (∼99%) of the actinobacterium 16S rRNA gene (1140 bp) from the body wall (n = 34), egg capsule (T1; n = 32) and nephridia (n = 34). A sequence variant was detected in the nephridia (15% of the library) but not in egg capsule or body wall or subsequent libraries (n = 2). Complete sequences were determined from representative clones of the body wall (2), nephridia (3) and capsule (11). Accession numbers GU201581–GU201596.

Amplification of total 16S rRNA from body wall was completed with the eubacterial primers GM3 and GM4 (Muyzer et al., 1995), and 51 clones sequenced and identified as Actinobacteria (11), V. eiseniae (26) and Flexibacter (1) symbionts. Other sequences were related to Paenibacillus (4), Propionibacterium (3), an Enterococcus, an Alphaproteobacteria and an Agromyces. These body wall samples were very carefully dissected to avoid nephridia. Herbaspirillum were not detected either by FISH or by PCR in nephridia or body wall of adult worms. Further PCR detection assays with specific probes failed to amplify the Flexibacter-like genes from the body wall DNA.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

This study revealed that embryos of the common epigeic earthworm, E. fetida, develop in intimate association with a microbial assemblage consisting of four major bacterial taxa: (i) V. eiseniae (Betaproteobacteria), (ii) a novel Microbacteriaceae member (Actinobacteria), (iii) a novel Flexibacteriaceae member (Bacteroidetes) and (iv) Herbaspirillum spp. (Betaproteobacteria). The presence of V. eiseniae, Microbacteriaceae and Flexibacter-like bacteria in both capsules and adult earthworms confirmed that they are regularly associated symbionts of E. fetida, and transmitted to the next generation. The Flexibacteriaceae are members of the Bacteroidetes which contains members of gut systems, and leech bladders, but the only host-associated Flexibacter members found in the databases related (not closely) to the earthworm symbionts were reported as sponge bacteria and their activities have not been characterized (Enticknap et al., 2006; Thiel et al., 2007; Kikuchi et al., 2009). In previous studies, the Flexibacter-like cells were not detected in adult nephridia using general eubacterial probes (Eub338), were only minor components of 16S rRNA gene clone libraries and absent from T-RFLP traces. Cells matching their unique morphology were observed in egg capsule albumin in our earlier studies using cell counts by FISH (Davidson and Stahl, 2008). Detection failure is now attributed to probe and primer bias. New primers and probes (designed using clone sequences in T3 capsule libraries) established that they are stable and abundant inhabitants of the earthworm egg capsules and nephridia.

Because direct observation of bacterial migration from the parent into the capsule has not yet been made, and the Flexibacter and Actinobacteria members do occur at low levels in the bedding, there remains a possibility that these bacteria are recruited from the bedding during capsule formation. Our earlier evaluation of newly formed capsules by FISH showed high densities of cells. The T-RFLP traces together with clone libraries confirmed this population is dominated by at least two of three symbionts found also in the nephridia. Because the egg capsules become sealed against bacterial entrance soon after deposition into the soil, the most direct route to obtain the high density observed (109 ml−1) in the short amount of time (hours) is by transmission from the adult. Both the Flexibacter and the V. eiseniae are abundant in the nephridia with the Actinobacterium also present but less numerous. Observations by FISH of the nephridia in the clitellum region of another lumbricid species revealed that bacteria had migrated from the nephridia ampulla into the bladders (S.K.D. Davidson, unpubl. data). The clitellum is responsible for the secretions of materials into the capsule. The migration of bacteria from the ampulla to the bladder strongly suggests that cells migrate out through the nephridial pore into the capsules during reproduction. This remains to be confirmed for E. fetida in the lab by direct observation, but available data are very suggestive that V. eiseniae cells are transmitted from the parent during capsule formation, increasing the likelihood that the other symbionts are simultaneously transmitted (Davidson and Stahl, 2006).

From the data presented here, we conclude that the Herbaspirillum population (represented by two sequence types) originated from the bedding since the 16S rRNA gene signal occurred at relatively low levels in both the bedding and newly formed capsules, and was not detected in either body wall or nephridia. Although the Herbasprillum might have originated from an undetected source in the worm, it is more likely that they were entrained from the bedding during capsule formation, and subsequently proliferated in the albumin. There is evidence for inhibitory activity within the egg capsule and it is possible the Herbaspirillum strains that appear are resistant to these inhibitory substances (Valembois et al., 1986; S.K.D. Davidson, unpubl. data).

The temporal overview of community composition provided by T-RFLP traces revealed a low diversity was maintained throughout development, but the relative abundance of Herbaspirillum spp. increased as the V. eiseniae symbiont decreased. The relative abundances of the Actinobacteria remained approximately the same. Although T-RFLP analysis and clone libraries only describe relative abundances, and do not distinguish proliferation or loss, based on observations of bacterial colonization of the embryo over the course of development (Davidson and Stahl, 2008) we speculate that these fluctuations are a consequence of embryonic development. As they grow, the earthworm embryos consume the albumin containing bacterial cells, reducing albumin volume. Bacterial cells are predominately in the albumin when development begins. In mid-development, each embryo begins to recruit V. eiseniae and Flexibacter-like symbionts from the albumin into the nephridia, and this process is completed a few days before hatching (Davidson and Stahl, 2008). Meanwhile, bacteria in the developing gut lumen increase in number and shift composition so that V. eiseniae is a minor component at hatching (Davidson and Stahl, 2008). Thus, the relative reduction of the V. eiseniae signal corresponds to a loss of albumin during development, such that this symbiont is restricted primarily to the nephridia by the time of hatching. A correlation between the increase in the number of bacterial cells in the gut lumen and the increase in relative abundance of Herbaspirillum spp. signal suggests Herbaspirillum may proliferate here, but the specific composition of this gut population remains to be explored.

Changes in abundance of the Flexibacter-like symbiont could not be determined by the data presented here, but these cells are abundant in the capsule albumin, clearly interacting with the embryos and with Verminephrobacter. In the mature nephridia, the Flexibacter together with V. eiseniae form a two-layered population with V. eiseniae cells attached to the epithelial surface and the Flexibacter-like cells primarily in the lumen extending between V. eiseniae cells to attach to the host surface. The Actinobacteria member is found within the lumen of the ampulla, at lower abundance and inconsistently, adding an additional layer to the consortium. Although comparative studies suggest that nephridial consortia are common among earthworms (S.K. Davidson, R.J. Powell and S.W. James, unpublished), we have yet to ascribe definitive functions. It is possible the symbionts work in concert to assist with waste processing and nitrogen conservation.

Similar to the earthworm, the cocoons of the medicinal leech (Hirudo verbana) contain the bacterial symbionts that occur in the adults. The adult leech harbours distinct bacteria in the gut (crop, intestinum) and nephridial bladders (Kikuchi et al., 2009; Rio et al., 2009). Both bladder and gut symbionts occur in the cocoon where the embryos develop with the exception of Rikenella. A study examining the presence of symbionts within the leech embryos demonstrated that abundances change over the course of development (Rio et al., 2009). Four out of six identified leech bladder symbionts are from the same orders and families as the earthworm symbionts, the Comamonadaceae and Bacteroidetes, but are not closely related to the E. fetida bacteria.

Multimember low-diversity bacterial consortia associated with animal hosts are common (shipworms, termites, aphids), and include earthworm relatives, the gutless oligochaetes that colonize marine sediments and the leeches (Dubilier et al., 1995; Goffredi et al., 2005; 2007; Laufer et al., 2008; Kikuchi et al., 2009). Within these consortia, members may have distinct functions that complement each other and complete the requirements of their host (Distel et al., 2002; Distel, 2003; Moran et al., 2005; Woyke et al., 2006). The three bacterial symbiont taxa discovered in E. fetida, members of Betaproteobacteria, Actinobacteria and Bacteroidetes, are from widely divergent groups suggesting each serves distinct functions that may change over the course of the earthworm life cycle. In the egg capsule, which remains in the soil for extended periods of time, the bacterial activity and interactions between bacteria and the developing embryo are likely very different than in the mature association within the nephridia, where the cells form an intimate association with a particular tissue type.

Among invertebrates, there are many examples of communities of bacteria associated with eggs, shown or suspected to be, for chemical protection of the eggs against microbial overgrowth or predation (Gil-Turnes et al., 1989; Gil-Turnes and Fenical, 1992; Benkendorff et al., 2001; Lindquist, 2002; Kaltenpoth et al., 2005; Pichon et al., 2005). Actinobacteria occur as symbionts in a wide variety of eukaryotic hosts from sponges to insects, often producing defensive antifungal or antimicrobial substances (Gil-Turnes et al., 1989; Gil-Turnes and Fenical, 1992; Cafaro and Currie, 2005; Hill et al., 2005; Montalvo et al., 2005; Kaltenpoth et al., 2006; Zhang et al., 2007). The maintenance of the Microbacteriaceae population in the E. fetida capsules at approximately the same relative abundance throughout development suggests this organism is necessary at a certain level for the success of the embryos. We hypothesize that the bacteria associated with the egg capsules protect the earthworm embryos during development. Earlier studies suggested the albumin suppressed bacterial growth (Day, 1950; Valembois et al., 1986). The selective growth of Herbaspirillum and Sphingomonas spp. in late development provides additional evidence for selective suppression of bacteria, possibly provided by the symbionts present. Given the diversity of fungi, bacteria and potential predators in the soil community, an egg capsule-conferred chemical defence of earthworm egg capsules is very plausible.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Laboratory rearing of E. fetida

Adult E. fetida (60 worms per 1 l bedding; Yelm Earthworm and Castings Farm, Yelm, WA) were maintained at 21–23°C, in autoclaved hydrated (w/diH2O) coir (mulched coconut husks, Coconut Palm Resources, Hillsboro, OR), 5 g of oatmeal and 10 g of coffee grounds. Bedding was changed every 7–10 days. For the treatments in this study, two sets of beds were sampled to determine if microbial profiles differed with the length of time bedding was worked by E. fetida. One treatment (two bins) was sampled after 6 days of exposure to worms, and the second treatment (two bins) was sampled after 20 days, for a total of four bins sampled.

Selection, DNA extraction and fixation of E. fetida embryonic stages and adult tissue

Newly deposited capsules (0–1 day old; T1) were removed daily from each bin and examined by dissection microscope to confirm lack of developed embryos. Capsules were maintained in Petri dishes on Whatman 1 filters moistened with distilled water, and apparently healthy developmental stages were collected for DNA extraction at T1, T2 and T3 (nearly mature; Fig. 1). DNA was extracted from the pooled contents of three capsules (DNeasy Blood and Tissue kit, QIAGEN, Valencia, CA), and replicate (n = 5–10) samples were extracted for each developmental stage. From each earthworm bin, DNA from bedding (0.5 g) was extracted using the Bio 101 FastDNA® Kit for Soil (Bio 101, Vista, CA) following manufacturer's instructions. Capsule contents for FISH were fixed according to Davidson and Stahl (2008).

For DNA samples and fixation for FISH, adult E. fetida earthworms were washed in sterile distilled water, anaesthetized in sterile-filtered water with ∼10% ethanol, and opened along the dorsal midline to expose internal organs. The reproductive organs, ovaries, testes, vas deference and seminal vesicles, were removed and DNA extracted from 25 mg of each as described above. DNA was also extracted from a portion of the body wall alone (no organs), and nephridia alone along the length of the worm for four samples per worm: (i) anterior to clitellum, (ii) the clitellum, (iii) posterior to clitellum and (iv) near the last segment.

PCR and T-RFLP analysis

Terminal restriction fragment length polymorphism (T-RFLP) profiles of the microbial community are based on the differences in the length of restriction fragments derived from unique 16S rRNA sequence types (Avaniss-Aghajani et al., 1994; 1996; Liu et al., 1997). Procedures for the amplification of the 16S rRNA gene were the same for both the T-RFLP analyses and for cloning and sequence analyses of the full 16S rRNA gene, with the exception of the reverse primer. PCR reactions contained 1 µl of DNA directly from the extraction eluant, or diluted as needed (10–20× for soil samples) to obtain amplification with the following reaction mix (20 µl): 2 µl of 10× Fermentas buffer, 2 µl of 25 mM MgCl2, 2 µl of 4% BSA, 0.4 µl of 10 µM primer, 0.4 µl each of 10 mM dNTP, 0.1 µl (5 U µl−1) of Taq polymerase (MBI Fermentas, Hanover, MD), and thermocycles run as (MJ Research, model PTC-100): 94°C, 3 min, then 25 cycles of 94°C, 30 s; 52°C, 30 s; 72°C, 40 s with a final extension 72°C, 5 min.

For T-RFLP profile generation (Liu et al., 1997), a 500 bp piece of the 16S rRNA gene was amplified using GM3-FAM (6-carboxy fluorescein) and 519 R (Lane, 1990; Muyzer et al., 1995), cut by restriction enzyme HaeIII, then run on a sequencer to determine the size and relative abundance of FAM-labelled terminal fragments. The HaeIII restriction enzyme was selected after testing four enzymes (MspI, HhaI, AluI and HaeIII) for restriction patterns of the bacterial 16S rRNA genes known to be present in the capsules. Three PCR reactions were run for each DNA sample and analysed separately to evaluate variation between reactions. The PCR products were quantified on a Turner biosystems fluorometer using Hoechst dye, then 5–10 µl was digested with HaeIII overnight, precipitated, rinsed with 70% ethanol and resuspended in 10 µl of water and the concentration estimated assuming a 70% recovery rate. To optimize the fragment peak signal, a range of 10–20 fmol replicates (n = 3) per sample were loaded into 96-well plates for analysis on a ABI 3100 capillary DNA sequencer (Applied Biosystems); ABI Genescan® software was used to determine fragment sizes based on internal lane standards. A customized DAx® (Van Mierlo Software Consultancy) software package was used to determine total fluorescent signal and the relative contribution of each peak to the total (%; relative abundance). Profiles with either saturated peaks or insufficient signal were omitted from analyses. For each time point the average abundance and standard deviations for each peak were determined from the total set of PCR reactions; however, sample number (n) is expressed as the number of DNA samples examined, not the number of PCR reactions.

Clone libraries and T-RFLP peak identification

To identify bacterial taxa corresponding to T-RFLP peaks and detect bacteria missed by T-RFLP a total of seven clone libraries (2003, 2004, 2007) were constructed, screened and representatives sequenced. All clone libraries were constructed from 16S rRNA gene amplifications from egg capsule DNA using primer pair GM3 (S-D-Bacteria-8-b-S-16)/GM4 (S-*-Univ-1492-b-A-16; names as in Alm et al., 1996) then cloned with Invitrogen, TOPO-TA cloning kit according to manufacturer's protocol. Libraries obtained in 2003/2004 (T1, T2, T3 capsules from an E. fetida colony established in the lab for over a year) were screened by restriction fragment length polymorphism (RFLP) using restriction enzymes HaeIII and MspI separately, following manufacturer's guidelines (New England Biolabs, Beverly, MA). Both strands of unique clones were fully sequenced (∼1460 bp) on an ABI 3730XL high-throughout capillary DNA Analyser according to manufacturer's recommendations (Molecular Dynamics, Sunnyvale, CA) using primers GM3, GM4 and selected internal 16S rRNA primers.

Three additional libraries were constructed (two T1 – 137 clones; one T3 – 91 clones) from a new shipment of E. fetida (2007) raised in manure compost at a local earthworm grower (Yelm, WA) to confirm consistency of the capsule bacterial community. These 2007 libraries were screened by partial sequencing using a high-throughput sequencing facility to obtain around 850 bp single strand clone sequence, then representatives of clone families (98% identity) were fully sequenced. Sequences were assembled using Sequencher™ (Gene Codes Corp.).

A capsule library (3–10 day, 60 clones, 2003) was used to identify and confirm bacterial 16S rRNA clones corresponding to terminal fragments observed in the T-RFLP analysis. Both predicted restriction patterns based on gene sequence, and T-RFLP peak analysis performed as described above on individual clones, confirmed terminal fragment correspondence to bacterial taxa found in the libraries.

Phylogenetic analysis of cloned sequences

Full sequences were aligned using the ARB software package (http://www.arb-home.de), placed into the greater bacterial 16S rRNA phylogeny within the SILVA comprehensive rRNA database http://www.arb-silva.de, and both closely related taxa and taxa estimated to be just outside of the phylogroup were selected for the pylogenetic analysis. A range of taxa were selected to best estimate affiliation of unknown sequences with known genera. The alignment was exported and analysed in Phyllip 3.66 to generate maximum parsimony (MP) and maximum likelihood (ML) consensus trees, with branch topology confidence determined by 1000 bootstrap resampling. A taxon estimated to fall just outside the presumed group was selected as an outgroup. The best-supported topologies are presented (Fig. 3) with ML bootstrap values presented and nodes with disagreement in confidence levels between the ML and MP analyses are noted.

Fluorescent in situ hybridization (FISH)

Bacterial cells within the capsules, the embryos and nephridia were detected by in situ hybridization using a series of fluor-labelled oligonucleotide probes (Cy3, Cy5 or fluorescein), then viewed with a Zeiss Pascal laser scanning confocal microscope to obtain three-dimensional reconstructions. The LSB 145 (S-G-Acidov-0145-a-A-18) probe identifies a subset of the Acidovorax group that includes the earthworm bacterial symbiotic strains (Schweitzer et al., 2001; Schramm et al., 2003). Actinobacterial 23S rRNA probe HGC 69a targets most Actinobacteria members (Roller et al., 1994). EUB 338 I, II and III probes will bind most members of the bacterial domain (Amann et al., 1990). A probe specific for the detected Flexibacteriaceae member (S-G-Flexibact-0145-a-A-18; 5′-GTTTCCCCGAGCTATRCC-3′) and Herbasprillum spp. (S-G-Herbasp-0145-a-A-21; 5′-CTTTCGACTAGTTATCCCCCA-3′) were designed, tested for specificity using a range of temperatures, and used to detect these bacteria within the nephridia and capsules. These two probes target the same site as the LSB 145 probe and are specific at 46°C, 35% formamide concentration under buffer conditions noted in Pernthaler and colleagues (2001).

Fluorescent in situ hybridization was performed according to a published protocol under stringent conditions for the probes used (Pernthaler et al., 2001; Davidson and Stahl, 2008). Probe-conferred fluorescence was imaged using appropriate wavelength excitation on a Zeiss LSM Pascal laser scanning confocal microscope (LSCM; Carl Zeiss, Jena, Germany).

Amplification of Microbacteriaceae member 16S rRNA genes

A PCR assay was used to localize the actinobacterial symbiont within the worm tissues using specific primers, EfA 243F (S-F-Actino-0243; 5′-GGATGGACTCGCGGCCTA) and EfA 1378R (S-F-Actino-1378; 5′-CGGTGTGTACAAGGCCCGGSAACG) modified from original Actinobacteria primers F243 and R1378 from Heuer and colleagues (1997). This pair amplifies a fragment 1175 bp long from a subset of Microbacteriaceae, including the sequenced E. fetida capsule bacteria. DNA was extracted from adult E. fetida tissue and egg capsules as described above. PCR was conducted as described with a 60°C annealing temperature, and products were cloned, sequenced and compared with sequences obtained from the egg capsules.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

This funding was supported by NSF IOB 0345049 and DEB 0516520. We thank Stephen MacFarlane for electron microscopy, and Fred An, Erin Lockert, and undergraduate researchers Ruth Go, Wesley Tang and Rebecca Lewis for assistance with obtaining molecular data. The comments of reviewers were very helpful in improving this manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Alm, E.W., Oerther, D.B., Larsen, N., Stahl, D.A., and Raskin, L. (1996) The oligonucleotide probe database. Appl Environ Microbiol 62: 35573559.
  • Amann, R.I., Binder, B.J., Olson, R.J., Chisholm, S.W., Devereux, R., and Stahl, D.A. (1990) Combination of 16s Ribosomal-RNA-targeted oligonucleotide probes with flow-cytometry for analyzing mixed microbial-populations. Appl Environ Microbiol 56: 19191925.
  • Avaniss-Aghajani, E., Jones, K., Chapman, D., and Brunk, C. (1994) A molecular technique for identification of bacteria using small subunit ribosomal RNA sequences. Biotechniques 17: 144149.
  • Avaniss-Aghajani, E.K., Jones, K., Holtzman, A., Aronson, T., Glover, N., Boian, M., et al. (1996) Molecular technique for rapid identification of mycobacteria. J Clin Microbiol 34: 98102.
  • Benkendorff, K., Davis, A.R., and Bremner, J. (2001) Chemical defense in the egg masses of benthic invertebrates: an assessment of antibacterial activity in 39 mollusks and 4 polychaetes. J Invertebr Pathol 78: 109118.
  • Cafaro, M.J., and Currie, C.R. (2005) Phylogenetic analysis of mutualistic filamentous bacteria associated with fungus-growing ants. Can J Microbiol 51: 441446.
  • Davidson, S.K., and Stahl, D.A. (2006) Transmission of nephridial bacteria of the earthworm Eisenia fetida. Appl Environ Microbiol 72: 769775.
  • Davidson, S.K., and Stahl, D.A. (2008) Selective recruitment of bacteria during embryogenesis of an earthworm. ISME J 2: 510518.
  • Day, G.M. (1950) The influence of earthworms on soil microorganisms. Soil Sci 69: 175184.
  • Distel, D.L. (2003) The biology of marine wood boring bivalves and their bacterial endosymbionts. In Wood Deterioration and Preservation. Goodell, B., Nicholas, D.D., and Schultz, T.P. (eds). Washington, DC, USA: American Chemical Society Press, pp. 253271.
  • Distel, D.L., Beaudoin, D.J., and Morrill, W. (2002) Coexistence of multiple proteobacterial endosymbionts in the gills of the wood-boring bivalve Lyrodus pedicellatus (Bivalvia: Teredinidae). Appl Environ Microbiol 68: 62926299.
  • Drake, H.L., and Horn, M.A. (2007) As the worm turns: the earthworm gut as a transient habitat for soil microbial biomes. Annu Rev Microbiol 61: 169189.
  • Dubilier, N., Giere, O., Distel, D.L., and Cavanaugh, C.M. (1995) Characterization of chemoautotrophic bacterial symbionts in a gutless marine worm (Oligochaeta, Annelida) by phylogenetic 16s ribosomal-RNA sequence-analysis and in-situ hybridization. Appl Environ Microbiol 61: 23462350.
  • Edwards, C.A. (ed.) (2004) Earthworm Ecology, 2nd edn. New York, USA: CRC Press LLC.
  • Enticknap, J.J., Kelly, M., Peraud, O., and Hill, R.T. (2006) Characterization of a culturable alphaproteobacterial symbiont common to many marine sponges and evidence for vertical transmission via sponge larvae. Appl Environ Microbiol 72: 37243732.
  • Gil-Turnes, M.S., and Fenical, W. (1992) Embryos of Homarus americanus are protected by epibiotic bacteria. Biol Bull 182: 105108.
  • Gil-Turnes, M.S., Hay, M.E., and Fenical, W. (1989) Symbiotic marine bacteria chemically defend crustacean embryos. Science 246: 116118.
  • Goffredi, S.K., Orphan, V.J., Rouse, G.W., Jahnke, L., Embaye, T., Turk, K., et al. (2005) Evolutionary innovation: a bone-eating marine symbiosis. Environ Microbiol 7: 13691378.
  • Goffredi, S.K., Johnson, S.B., and Vrijenhoek, R.C. (2007) Genetic diversity and potential function of microbial symbionts associated with newly discovered species of Osedax polychaete worms. Appl Environ Microbiol 73: 23142323.
  • Grove, A.J., and Cowley, L.F. (1926) Memoirs: on the reproductive processes of the brandling worm, Eisenia fetida. (Sav.). Q J Microsc Sci 70: 559581.
  • Heuer, H., Krsek, M., Baker, P., Smalla, K., and Wellington, E.M.H. (1997) Analysis of Actinomycete communities by specific amplification of genes encoding 16S rRNA and gel-electrophoretic separation in denaturing gradients. Appl Environ Microbiol 63: 32333241.
  • Hill, M.S., Lopez, N.A., and Young, K.A. (2005) Anti-predator defenses in western North Atlantic sponges with evidence of enhanced defense through interactions between spicules and chemicals. Mar Ecol Prog Ser 291: 93102.
  • Lund, M.B., Davidson, S.K., Holmstrup, M., James, S., Kjeldsen, K.U., Stahl, D.A., and Schramm, A. (2009) Diversity and host-specificity of the Verminephrobacter-earthworm symbiosis. Environ Microbiol (in press): doi: 10.1111/j.1462-2920.2009.02084.x.
  • Kaltenpoth, M., Gottler, W., Herzner, G., and Strohm, E. (2005) Symbiotic bacteria protect wasp larvae from fungal infestation. Curr Biol 15: 475479.
  • Kaltenpoth, M., Goettler, W., Dale, C., Stubblefield, J.W., Herzner, G., Roeser-Mueller, K., and Strohm, E. (2006) Candidatus Streptomyces philanthi’, an endosymbiotic streptomycete in the antennae of Philanthus digger wasps. Int J Syst Evol Microbiol 56: 14031411.
  • Kikuchi, Y., Bomar, L., and Graf, J. (2009) Stratified bacterial community in the bladder of the medicinal leech, Hirudo verbana. Environ Microbiol 11: 27582770.
  • Lane, D.S. (1990) 16S and 23S rRNA sequencing. In Nucleic Acid Techniques in Bacterial Systematics. Stackebrandt, E., and Goodfellow, M. (eds). New York, USA: John Wiley, pp. 115148.
  • Laufer, A., Siddall, M., and Graf, J. (2008) Characterization of the digestive-tract microbiota of Hirudo orientalis, a European medicinal leech. Appl Environ Microbiol 74: 61516154.
  • Lindquist, N. (2002) Chemical defense of early life stages of benthic marine invertebrates. J Chem Ecol 28: 19872000.
  • Liu, W.-T., Marsh, T.L., Cheng, H., and Forney, L.J. (1997) Characterization of microbial diversity by determining terminal restriction fragment length polymorphisms of genes encoding 16S rRNA. Appl Environ Microbiol 63: 45164522.
  • Montalvo, N.F., Mohamed, N.M., Enticknap, J.J., and Hill, R.T. (2005) Novel actinobacteria from marine sponges. Antonie Van Leeuwenhoek (Int J Gen Mol Microbiol) 87: 2936.
  • Moran, N.A., Tran, P., and Gerardo, N.M. (2005) Symbiosis and insect diversification: an ancient symbiont of sap-feeding insects from the bacterial phylum Bacteroidetes. Appl Environ Microbiol 71: 88028810.
  • Muyzer, G., Teske, A., Wirsen, C.O., and Jannasch, H.W. (1995) Phylogenetic-relationships of thiomicrospira species and their identification in deep-sea hydrothermal vent samples by denaturing gradient gel-electrophoresis of 16s rDNA fragments. Arch Microbiol 164: 165172.
  • Pernthaler, J., Glöckner, F.O., Schönhuber, W., and Amann, R. (2001) Fluorescence in situ hybridization (FISH) with rRNA-targeted oligonucleotide probes. In Methods in Microbiology. Paul, J. (ed.). San Diego, CA, USA: Academic Press, pp. 207226.
  • Pichon, D., Gaia, V., Norman, M.D., and Boucher-Rodoni, R. (2005) Phylogenetic diversity of epibiotic bacteria in the accessory nidamental glands of squids (Cephalopoda: Loliginidae and Idiosepiidae). Mar Biol 147: 13231332.
  • Pinel, N., Davidson, S.K., and Stahl, D.A. (2008) Verminephrobacter eiseniae gen. nov., sp. nov., a nephridial symbiont of the earthworm Eisenia fetida (Savigny). Int J Syst Evol Microbiol 58: 21472157.
  • Rio, R.V., Maltz, M., McCormick, B., Reiss, A., and Graf, J. (2009) Symbiont succession during the embryonic development of the European medicinal leech, Hirudo verbana. Appl Environ Microbiol 75: 68906895.
  • Roller, C., Wagner, M., Amann, R., Ludwig, W., and Schleifer, K.H. (1994) In-situ probing of gram-positive bacteria with high DNA G+C content using 235-ribosomal-RNA-targeted oligonucleotides. Microbiology 140: 28492858.
  • Schramm, A., Davidson, S.K., Dodsworth, J.A., Drake, H.L., Stahl, D.A., and Dubilier, N. (2003) Acidovorax-like symbionts in the nephridia of earthworms. Environ Microbiol 5: 804809.
  • Schweitzer, B., Huber, I., Amann, R., Ludwig, W., and Simon, M. (2001) Alpha- and beta-Proteobacteria control the consumption and release of amino acids on lake snow aggregates. Appl Environ Microbiol 67: 632645.
  • Thiel, V., Neulinger, S.C., Staufenberger, T., Schmaljohann, R., and Imhoff, J.F. (2007) Spatial distribution of sponge-associated bacteria in the Mediterranean sponge Tethya aurantium. FEMS Microbiol Ecol 59: 4763.
  • Valembois, P., Roch, P., and Lassegues, M. (1986) Antibacterial molecules in annelids. In Immunity in Invertebrates. Brehelin, M. (ed.). Berlin, Germany: Springer-Verlag, pp. 7493.
  • Woyke, T., Teeling, H., Ivanova, N.N., Huntemann, M., Richter, M., Gloeckner, F.O., et al. (2006) Symbiosis insights through metagenomic analysis of a microbial consortium. Nature 443: 950955.
  • Zhang, M.M., Poulsen, M., and Currie, C.R. (2007) Symbiont recognition of mutualistic bacteria by Acromyrmex leaf-cutting ants. ISME J 1: 313320.