Present address: Yongjie Wang, Dienstleistungszentrum Ländlicher Raum (DLR Rheinpfalz), Abteilung Phytomedizin, Neustadt an der Weinstraße, Germany.
Bacterial symbionts in the hepatopancreas of isopods: diversity and environmental transmission
Article first published online: 16 MAY 2007
FEMS Microbiology Ecology
Volume 61, Issue 1, pages 141–152, July 2007
How to Cite
Wang, Y., Brune, A. and Zimmer, M. (2007), Bacterial symbionts in the hepatopancreas of isopods: diversity and environmental transmission. FEMS Microbiology Ecology, 61: 141–152. doi: 10.1111/j.1574-6941.2007.00329.x
Editor: Christoph Tebbe
- Issue published online: 16 MAY 2007
- Article first published online: 16 MAY 2007
- Received 20 November 2006; revised 6 March 2007; accepted 9 March 2007.First published online 16 May 2007.
- symbiotic bacteria;
The midgut glands (hepatopancreas) of terrestrial isopods contain bacterial symbionts. We analysed the phylogenetic diversity of hepatopancreatic bacteria in isopod species from various suborders colonizing marine, semiterrestrial, terrestrial and freshwater habitats. Hepatopancreatic bacteria were absent in the marine isopod Idotea balthica (Valvifera). The symbiotic bacteria present in the midgut glands of the freshwater isopod Asellus aquaticus (Asellota) were closely related to members of the proteobacterial genera Rhodobacter, Burkholderia, Aeromonas or Rickettsiella, but differed markedly between populations. By contrast, species of the suborder Oniscidea were consistently colonized by the same phylotypes of hepatopancreatic bacteria. While symbionts in the semiterrestrial isopod Ligia oceanica (Oniscidea) were close relatives of Pseudomonas sp. (Gammaproteobacteria), individuals of the terrestrial isopod Oniscus asellus (Oniscidea) harboured either ‘Candidatus Hepatoplasma crinochetorum’ (Mollicutes) or ‘Candidatus Hepatincola porcellionum’ (Rickettsiales), previously described as symbionts of another terrestrial isopod, Porcellio scaber. These two uncultivated bacterial taxa were consistently present in each population of six and three different species of terrestrial isopods, respectively, collected in different geographical locations. However, infection rates of individuals within a population ranged between 10% and 100%, rendering vertical transmission unlikely. Rather, feeding experiments suggest that ‘Candidatus Hepatoplasma crinochetorum’ is environmentally transmitted to the progeny.
Isopods (Crustacea: Isopoda) originated from the marine environment and have successfully colonized terrestrial habitats. It is assumed that the terrestrial suborder Oniscidea shares a common ancestor with the marine suborders Valvifera and Sphaeromatidea (Zimmer, 2002). The midgut glands (hepatopancreas) of the terrestrial isopods Porcellio dilatatus (Donadey & Besse, 1972), Porcellio scaber (Wood & Griffith, 1988; Hames & Hopkin, 1989; Zimmer & Topp, 1998a, b; Zimmer, 1999; Wang et al., 2004a, b), Oniscus asellus (Hopkin & Martin, 1982; Wood & Griffith, 1988; Hames & Hopkin, 1989) (both Oniscidea: Crinocheta), and of the semiterrestrial species Ligia pallasii (Oniscidea: Diplocheta) (Zimmer et al., 2001) carry symbiotic bacteria. The hepatopancreas secretes digestive fluids into the hindgut where digestion takes place and is involved in the resorption of digestively released nutrients (Zimmer, 2002). One of the most interesting aspects of bacterial symbionts inside the hepatopancreatic lumen is their proposed contribution to digestive processes of the isopods, i.e. the hydrolysis of cellulose (Zimmer & Topp, 1998a, b; Zimmer et al., 2002) and the oxidation of phenolics (Zimmer, 1999; Zimmer et al., 2002) and lignins (Zimmer & Topp, 1998b; Zimmer et al., 2002). In contrast to macroalgal food sources of marine isopods, detrital food sources of terrestrial (and freshwater) isopods contain little nutrients, but are rich in cellulose and phenolics. Thus, hepatopancreatic bacteria may have facilitated the evolutionary colonization of terrestrial habitats by Oniscidea and their utilization of terrestrial food sources (Zimmer & Topp, 1998b; Zimmer et al., 2001; Zimmer & Bartholmé, 2003).
Either hepatopancreatic bacteria were acquired by the isopods simultaneously with numerous adaptations that allowed colonization of land or the acquisition of hepatopancreatic bacteria was a predisposition to the colonization of land. In the former case, these bacteria would be lacking in marine isopods; in the latter case, hepatopancreatic bacteria would be present in marine species (Zimmer et al., 2002). In contrast to their terrestrial relatives, those four marine isopod species of the suborders Valvifera and Sphaeromatidea tested thus far do not harbour bacteria in their midgut glands (Zimmer et al., 2001, 2002; S. Fraune & M. Zimmer, unpublished data), even though they live in a benthic environment that is rich in potential bacterial inoculants. Thus, it has been discussed whether these marine isopods either are effective in protecting themselves from bacterial inoculation or lack the capability of hosting bacteria in their hepatopancreas (Zimmer, 2002). However, this is contradicted by the observation of hepatopancreatic bacteria in Asellus aquaticus (Zimmer & Bartholmé, 2003), a freshwater representative of Asellota, an isopod suborder common in both marine and freshwater systems. Asellus aquaticus, which is one of the most common detritivorous crustaceans in temperate lentic freshwaters, feeds mainly on decaying plant leaf litter like its terrestrial relatives (Isopoda: Oniscidea).
Recently, we reported that two types of phylogenetically distinct hepatopancreatic symbionts –‘Candidatus Hepatincola porcellionum’, curved rod-shaped bacteria related to the Rickettsiales (Alphaproteobacteria), and ‘Candidatus Hepatoplasma crinochetorum’, sphere-shaped Mycoplasma-like bacteria (Mollicutes) – colonize the epithelial brush border of the hepatopancreas in the terrestrial P. scaber (Wang et al., 2004a, b). However, little is known about the identity of bacterial symbionts in other terrestrial isopod species, owing to the failure in isolating and cultivating these bacteria in previous studies. Yet, this knowledge would be valuable in shedding light on the role of bacterial symbionts during the colonization of land by arthropods with marine ancestors.
Many terrestrial arthropods have evolved elegant transmission mechanisms to ensure the inoculation of their progeny with specific bacterial symbionts, involving three different modes of transmission (Cary & Giovannoni, 1993): (1) vertical transmission from (female) parent to the offspring (Aksoy, 2003), (2) horizontal transmission between syntopic hosts and (3) environmental transmission, where the new host generation takes up its symbionts from the environment (McFall-Ngai, 2002). Theory predicts that mutualistic symbioses will evolve under conditions of vertical transmission of the symbiont from parent to offspring (Herre et al., 1999) as the transmission of symbionts is critical in mutualistic symbioses, both for obligate symbionts and for their host. It is only through vertical transmission that a host can ensure that none of its progeny will stay aposymbiotic. Yet, many mutualists rely on horizontal transmission (discussed in Wilkinson & Sherratt, 2001), and horizontal transfer of symbionts is the rule in sexually reproducing animals with symbionts that are not harboured inside the reproductive tract (Douglas, 1995). Thus, knowing the mechanism of symbiont transmission in isopods is crucial to our understanding of isopod–symbiont interactions.
In the present study, we (1) determine the phylogenetic affiliations of symbiotic bacteria in the hepatopancreas of selected isopod species from different habitats (aquatic, semiterrestrial, terrestrial); (2) survey the distribution of the symbionts among geographically distinct isopod populations; and (3) investigate the mode of symbiont transmission.
Materials and methods
Collection and culture of isopods
Idotea balthica Pallas 1772 (Valvifera: Idoteidae) was collected from seaweed at the Falckenstein beach near Kiel, Germany, kept in plastic containers filled with artificial seawater, and fed on brown algae (Fucus spp.). Ligia oceanica Brandt 1833 (Oniscidea: Ligiidae) was collected beneath rocks at the Falckenstein lighthouse near Kiel, kept in plastic containers with a bottom of 1 cm of wet sand, and fed on brown algae (Fucus spp.). Oniscus asellus Linnaeus 1758 (Oniscidea: Oniscidae) and Porcellio scaber Latreille 1804 (Oniscidea: Porcellionidae) were collected from beneath decaying wood in the botanical garden of the Christian-Albrechts-Universität of Kiel, kept in plastic containers with a bottom of 1 cm of moist plaster, and fed with decaying leaf litter. Asellus aquaticus (Linnaeus 1758) (Asellota: Asellidae) was collected in various ponds in the vicinity of Kiel, and in Lake Neuwühren near Plön, about 20 km south-east of Kiel; they were kept in aerated tap water and fed with mixed leaf litter taken from the field. Isopod cultures were maintained at 12°C.
For the survey of the geographical distribution of bacterial symbionts in different terrestrial isopod species, an area of about 10 m2 at each collection site was screened for isopods living underneath decaying detritus.
Gravid females of P. scaber were divided into treatment and control groups. For the treatment group, isopods were washed five times in autoclaved water, surface-sterilized with 70% ethanol and UV light (λ=254 nm) for 10 s and 5 min, respectively, and washed again five times in water. Subsequently, they were cultivated individually in sterile Petri dishes (85 mm diameter) with sterile filter paper (80 mm) covering the bottom. The filter paper was wetted using sterile-filtered supernatant of an aqueous soil suspension and was renewed every 2 days during the first 2 weeks of the experiment. Leaf litter was surface-sterilized by means of combined ethanol and UV treatment (λ=254 nm) before it was offered as food. Gravid females in the control group were also kept individually in Petri dishes, but were not subjected to surface sterilization, and were fed nonsterilized leaf litter. Unfiltered supernatant (see above) was used to moisten the filter papers. Isopods were maintained at 15°C, 16/8 h light–dark. Isopod hatchlings (mancae) were harvested just before hatching, and juveniles were collected at day 1 (stage 1) and at day 7 (stage 2) after their release from their mother's brood pouch and the immediate removal of the mother upon offspring hatching. For each group, 30 embryos or 40 juveniles were obtained from each of the 10 mothers. Mother isopods were dissected immediately after obtaining their embryos or juveniles.
Adult isopods (both males and females) were washed five times in autoclaved water, surface-sterilized with 70% ethanol and UV light (λ=254 nm) for 1 and 5 min, respectively, and washed again five times in water. After blotting their bodies dry, the midgut glands were dissected and stored in autoclaved reaction tubes filled with 1 mL acetone. Total DNA was extracted and purified using a bead-beating protocol (Friedrich et al., 2001). DNA was recovered through isopropanol precipitation, then rinsed in 70% ethanol (v/v, −20°C), subsequently resuspended in PCR water (Sigma), and stored at −20°C before usage.
PCR, sequencing and sequence analysis
The PCR primers applied in this study are listed in Table 1. PCR amplification, cloning, and restriction fragment length polymorphism (RFLP) analysis of 16S rRNA genes followed the protocols described in Wang et al. (2004a, b). Diagnostic PCR with symbiont-specific primers used a 200-μL PCR tube with a 50-μL reaction volume containing 200 μM each dNTP, 5.0 μL of a 10 × reaction buffer (Eppendorf), 25 pmol of each primer, 1.0 U of Taq DNA polymerase (Eppendorf) and 1 μL of template DNA. Samples were amplified with a DNA thermal cycler (Eppendorf) under the following conditions: initial denaturation at 94°C for 3 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s and extension at 72°C for 1 min, followed by a final elongation for 7 min at 72°C.
|Primer||OPD designation*||Primer sequence (5′–3′)||Reference or source|
|27f||S-D-Bact-0007-a-S-21||CAGAGTTTGATCCTGGCTCAG||Weisburg et al. (1991)|
|63f||S-D-Bact-0043-a-S-21||CAGGCCTAACACATGCAAGTC||Marchesi et al. (1998)|
|533f||S-D-Bact-0515-a-S-19||GTGCCAGCAGCCGCGGTAA||Henckel et al. (1999)|
|907r||S-D-Bact-0907-a-A-15||AATTCCTTTGAGTTT||Henckel et al. (1999)|
|1387r||S-D-Bact-1387-a-A-18||GGGCGGWGTGTACAAGGC||Marchesi et al. (1998)|
|1492r||S-D-Bact-1492-a-A-22||TACGG(C/T)TACCTTGTTACGACTT||Weisburg et al. (1991)|
Both strands of the inserts were sequenced using primers M13 forward, M13 reverse, and 533f/907r on an ABI sequencer by GATC (http://www.gatc.de). Sequences were assembled and compared with those in public databases using blast (Altschul et al., 1997). Closely related sequences were retrieved and added to the alignment. Only sequences with more than 1200 nucleotides were used for the alignment. Multiple alignments were made using the clustal w program (Thompson et al., 1994). Alignments were always manually checked. Phylogenetic analysis utilized the maximum-parsimony, neighbour-joining and minimum evolution algorithms as implemented in mega 3.1 (Kumar et al., 2004).
Denaturing gradient gel electrophoresis (DGGE)
Denaturing gels were prepared using two 6.5% (w/v) acrylamide gel stock solutions [37.5 : 1, acrylamide-N,N″-methylenebisacrylamide in 1 × Tris-acetate-EDTA (TAE) (containing 40 mM Tris base, 20 mM sodium acetate and 1 mM EDTA-Na2)] with 0 and 80% denaturant [100% denaturant contained 7 M urea and 40% (v/v) formamide in 1 × TAE], respectively. A linear 40–80% denaturant gradient was prepared according to the method of Myers et al. (1987).
For PCR amplification of bacterial DNA prior to DGGE, a GC clamp was attached at the 5′ end of the Bacteria-specific reverse primer 907r (Henckel et al., 1999). After pre-electrophoresis for 30 min, PCR products were loaded onto the gel, which was run at 60°C and 60 V for 15 h. The gel was stained with ethidium bromide (0.5 μg mL−1) for 15 min and photographed under UV light with a digital camera.
Dissected and ruptured hepatopancreas preparations, kept in 500 μL phosphate-buffered saline (PBS; see above), were homogenized by repeated passage through a pipette tip. After gravity sedimentation of the tissue shreds (5 min), 450 μL of supernatant was carefully transferred to a new sterile 1.5-mL tube containing 50 μL of 40% (w/v) formaldehyde. After fixation at 4°C for 15 h, the samples were centrifuged at 10 000 g for 5 min and washed three times with 500 μL PBS. The final pellets were dissolved in a mixture of 250 μL PBS and 250 μL of 97% (v/v) ethanol and stored at −21°C.
For fluorescence hybridization, hepatopancreas samples were filtered onto polycarbonate filters (0.2-μm pore size) and dried at 46°C for 30 min. Samples were stained with 4′,6-diamidino-2-phenylindole (DAPI) and hybridized with specific fluorescently labelled oligonucleotides (Table 2) as described by Wagner et al. (1993); negative controls with an EUB338 antisense probe (Wallner et al., 1993) were used to exclude nonspecific probe binding. All probes were synthesized and 5′-labelled with the fluorescent cyanine dye Cy3 or with 5(6)-carboxyfluorescein-N-hydroxysuccinimide ester by Thermo Hybaid (http://www.interactiva.de). Samples were covered with Citifluor (Citifluor Ltd, London) and examined at 1000-fold magnification with a Zeiss Axiophot epifluorescence microscope using filter sets for DAPI, Cy3 and fluorescein. Images were recorded with a digital camera (Olympus).
|Fluorescence- labelled probes*||OPD designation†||Probe Sequence (5′–3′)||Target site‡ (rRNA position)||Formamide (%)§||Reference|
|EUB338||S-D-Bact-0338-a-A-18||GCTGCCTCCCGTAGGAGT||16S (338–355)||0–50||Amann et al. (1990)|
|NONEUB||S-D-Bact-0338-a-S-18||ACTCCTACGGGAGGCAGC||Wallner et al. (1993)|
|ALF1B||S-Sc-αProt-0019-a-A-17||CGTTCGYTCTGAGCCAG||16S (19–35)||20||Manz et al. (1992)|
|GAM42a||L-C-γProt-1027-a-A-17 (with competitor BET42a)||GCCTTCCCACATCGTTT GCCTTCCCACTTCGTTT||23S (1027–1043)||35||Manz et al. (1992)|
|PSM G440¶||S-G-Pseu-0440-a-A-15||CCTTCCTCCCAACTT||16S (440-454)||35||Braun-Howland et al. (1993)|
|PsSym352||S-G-Hepa-0352-a-A-21||GTGAAAAATTCCCTACTGCTG||16S (352–373)||0–20||Wang et al. (2004a, b)|
|PsSym120||S-G-Hepa-0120-a-A-18||AGCCAAATTCCCACGTGT||16S (120–138)||20||Wang et al. (2004a, b)|
For in situ hybridization, thin sections (8 μm) of paraffin-embedded midgut glands, embryos and juveniles were prepared and hybridized as previously described (Wang et al., 2004a, b).
Hepatopancreatic bacteria in isopods from different habitats
Hepatopancreatic bacteria were detected in midgut gland homogenates of almost all isopod species tested by fluorescence microscopy and whole-cell hybridization with the fluorescence-labelled oligonucleotide probe EUB338, and 16S rRNA genes were successfully amplified from DNA extracts of the midgut glands. The only exception was the marine isopod Idotea balthica, in which no bacterial cells were observed in DAPI-stained preparations of the 30 individuals tested, and no PCR products were obtained with bacterial primers.
Hepatopancreatic bacteria in the semiterrestrial Ligia oceanica
In the semiterrestrial isopod Ligia oceanica, the hepatopancreatic bacteria were straight, rod-shaped cells (0.8 μm wide and 1.8–2.5 μm long: Fig. 1a–c) of uniform appearance. The two ribotypes in the corresponding clone library differed only slightly in their RFLP patterns. Phylogenetic analysis of representative clones showed that they were closely related to each other (99% sequence identity) and represented a lineage within the genus Pseudomonas (Table 3). Cell densities ranged from 0.3 × 107 to 6.0 × 107 cells per animal (n=10). All bacteria in the hepatopancreas of L. oceanica (DAPI-stained: Fig. 1a–c) exhibited strong hybridization signals with probes EUB338, GAM42a (with BET42a as competitor) and PSM G440, specific for all Bacteria, only Gammaproteobacteria and members of the genus Pseudomonas, respectively, confirming that the cloned 16S rRNA genes originated from the hepatopancreatic symbionts and that the latter form a homogeneous population of Pseudomonas sp.
|Isopod species||Clones tested||Clones in ribotype||Representative clone ID||Sequence affiliation||Sequence similarity (%)|
|Asellus aquaticus*† (Asellota: Asellidae) (Plön)||15||8||Aa-12||Alphaproteobacteria||Rhodobacter sp. (AY584573)||98.1|
|7||Aa-3||Gammaproteobacteria||Aeromonas sobria (X74683)||99.9|
|Asellus aquaticus*† (Asellota: Asellidae) (Kiel)||37||15||Aa-7||Gammaproteobacteria||Uncultured Gammaproteobacterium (AY947958)||98.8|
|13||Aa-10||Gammaproteobacteria||Uncultured Gammaproteobacterium (AY947958)||98.9|
|9||Aa-11||Betaproteobacteria||Burkholderia sp. N2P5 (U37342)||99.9|
|Ligia oceanica‡§ (Oniscidea: Ligiidae) (Kiel)||33||16||Lo-6||Gammaproteobacteria||Pseudomonas sp. (AY958857)||99.7|
|17||Lo-8||Gammaproteobacteria||Pseudomonas sp. (AY958857)||99.5|
|Oniscus asellus†¶ (Oniscidea: Oniscidae) (Kiel)||12||6||Oa-2||Mollicutes||‘Candidatus Hepatoplasma crinochetorum’ (AY500250)||99.2|
|4||Oa-12||Mollicutes||‘Candidatus Hepatoplasma crinochetorum’ (AY500250)||98.7|
|2||Oa-13||Mollicutes||‘Candidatus Hepatoplasma crinochetorum’ (AY500250)||98.7|
Hepatopancreatic bacteria in Oniscus asellus and other terrestrial isopods
Individuals of the terrestrial isopod Oniscus asellus contained either spherical cells (0.5–0.8 μm in diameter) or curved rods (0.5 μm wide and 1.5–3.8 μm long), which morphologically resembled the two previously described bacterial symbionts in the hepatopancreas of the closely related isopod species Porcellio scaber (Wang et al., 2004a, b). Cell densities ranged between 0.6 × 108 and 9.3 × 108 cells per animal (n=5).
Accordingly, DGGE of PCR products obtained with the same primer combination (533f/907r) yielded a single DNA band for L. oceanica, whereas two different bands were obtained for different individuals of O. asellus. The positions of the bands in the acrylamide gel obtained for O. asellus were identical to those obtained with clones of ‘Candidatus Hepatincola porcellionum’ and ‘Candidatus Hepatoplasma crinochetorum’ present in P. scaber (Wang et al., 2004a, b). These bacterial species seem to mutually exclude each other, as we never found an individual isopod that harboured both symbiont in its midgut glands.
A clone library of 16S rRNA genes from O. asellus comprised three ribotypes that differed only slightly in their RFLP patterns. Representative clones showed high sequence similarity among each other (98.1–99.6%). They were closely related to the sphere-shaped symbiont ‘Candidatus Hepatoplasma crinochetorum’ in the hepatopancreas of P. scaber (Table 3).
The specific probe PsSym352, previously designed to detect ‘Candidatus Hepatoplasma crinochetorum’ in P. scaber (Wang et al., 2004b), also matched the target sequence of all three cloned 16S rRNA genes from O. asellus (Fig. 1d–f). All of the DAPI-stained, sphere-shaped cells showed specific hybridization with both probe EUB338 and PsSym352, supporting the close relationship between the Mycoplasma-like symbionts found in O. asellus and P. scaber. The curved rod-shaped hepatopancreatic bacteria found in O. asellus (Fig. 1g–i) hybridized with the Bacteria-specific probe (EUB338), the Alphaproteobacteria-specific probe (ALF1B) and the probe PsSym120 designed to detect specifically ‘Candidatus Hepatincola porcellionum’ in P. scaber (Wang et al., 2004a), which suggested that the second symbiont, which was not represented in the clone library of O. asellus (see above), belongs to this rickettsial lineage.
Using the specific FISH probes, ‘Candidatus Hepatoplasma crinochetorum’ was detected not only in O. asellus and P. scaber populations collected in different locations (Table 4), but also in the terrestrial isopod species Philoscia muscorum (Scopoli 1763), Armadillidium vulgare (Latreille 1804), Trachelipus rathkii (Brandt 1833) and Alloniscus perconvexus Dana 1854, whereas ‘Candidatus Hepatincola porcellionum’ was found only in T. rathkii. Aposymbiotic individuals were present in all species (Table 4). Morphologically different, rod-shaped bacteria, not hybridizing with the probes for ‘Candidatus Hepatoplasma crinochetorum’ and ‘Candidatus Hepatincola porcellionum’, were observed in the hepatopancreas of 6.3% of A. perconvexus.
|Isopod species||Geographical origin||Habitat||N||Prevalence* (%)||Morphotype†||Frequency (%)‡|
|Porcellio scaber||Kiel||Woodland||22||100.0||S, C||59.1||40.9|
|Kiel||Artificial salt marsh||70||100.0||S, C||92.8||7.2|
|Köln, Germany||Woodland||9||100.0||S, C||88.9||11.1|
|Köln, Germany||Grassland||15||100.0||S, C||85.7||14.3|
|Vancouver Island, Canada (BC)||Grassland||24||25.0||S, C||20.8||4.2|
|Haines Island, Canada (BC)||Coastal grassland||37||13.5||S, C||8.1||5.4|
|Oniscus asellus||Kiel||Woodland||35||68.6||S, C, R||45.7||8.6|
|Köln, Germany||Woodland||14||71.4||S, C||28.5||42.9|
|Köln, Germany||Grassland||11||100.0||S, C||80.0||20.0|
|Trachelipus rathkii||Kiel||Grassland||8||80.0||S, C||60.0||20.0|
|Alloniscus perconvexus||Vancouver Island, Canada (BC)||Sandy beach||16||25.0||S, R||18.8||0|
Hepatopancreatic bacteria in the freshwater isopods Asellus aquaticus
The hepatopancreatic community in the freshwater isopod Asellus aquaticus was more complex than that in the other isopod species. Microscopic inspection had already shown that bacterial morphotypes in the midgut glands differed between both individuals and populations. Short rod-shaped cells (0.5 μm wide and 1.0–1.3 μm long) with densities ranging from 1.2 × 106 to 2.9 × 106 cells per animal (n=6) and long rod-shaped cells (0.7 μm wide and 1.8–2.7 μm long) with densities ranging between 0.3 × 106 and 1.7 × 106 cells per animal (n=4) were observed in individuals collected in Plön. Sphere-shaped cells (0.3–0.7 μm wide) with densities ranging from 1.6 × 106 to 1.9 × 106 cells per animal (n=2) and short rod-shaped cells in chains (0.4 μm wide and 0.4–0.6 μm long) with densities of 0.7 × 106 cells per animal (n=1) were observed in different individuals collected near Kiel.
A clone library of 16S rRNA genes representing hepatopancreatic bacteria of a pooled population of eight individuals collected in Kiel contained three different ribotypes (Table 3). Sequencing of representative clones showed that 76% of the clones in the clone library were virtually identical in sequence (99.7% similarity) and were related (>96% similarity) to Rickettsiella grylli (Vago & Martoja, 1963; Roux et al., 1997), whereas the rest were virtually identical to a free-living Burkholderia species in the Burkholderia glathei subgroup (Table 3).
For Asellus aquaticus collected in Lake Neuwühren (Plön), 16S rRNA genes were successfully amplified only with the alternative primer pair 63f/1387r (Table 1). The corresponding clone library, representing hepatopancreatic bacteria of a pooled population of 15 individuals, contained two different ribotypes closely related either to Aeromonas sobria or to Rhodobacter species (Table 3).
Transmission of symbionts in the terrestrial Porcellio scaber
Whereas all mother isopods were screened positive for ‘Candidatus Hepatoplasma’, no symbiotic bacteria were detected by DAPI/FISH staining in the homogenates of 100 P. scaber embryos obtained from 10 mothers, either in the sterile treatment or the control group. The absence of hepatopancreatic bacteria was corroborated by the lack of PCR products from DNA extracts of embryos (n=30 each from 10 mothers from each group) from both control and treatment groups as well as of juveniles from the treatment group. By contrast, an amplicon of 1162 bp was obtained from DNA extracts of juveniles from the control group, indicating the presence of ‘Candidatus Hepatoplasma crinochetorum’.
Data from whole-cell hybridization and PCR were further confirmed by the results obtained with thin sections of whole embryos and juveniles. In situ hybridization with probes EUB338 and PsSym352 did not detect any bacteria in embryos from both control and treatment groups or in juveniles of the treatment group, whereas hybridization signal with both probes was observed in the hepatopancreas of juveniles from the control group. The prevalence of bacterial symbionts in juveniles of the control group, however, was low (juvenile stage 1, <20%; stage 2, <30%) and varied between different cohorts (juvenile stage 1, 0–20%; stage 2, 0–30%).
The results of this study support the hypothesis that hepatopancreatic bacterial symbionts of isopods have been acquired simultaneously with numerous adaptations that allowed the colonization of land, as discussed by Zimmer et al. (2002). No bacteria or bacterial DNA were found in the hepatopancreas of the marine Idotea balthica (Valvifera: Idoteidae). Having thoroughly screened a total of four species of the marine suborders Valvifera and Sphaeromatidea (Zimmer et al., 2001, 2002; S. Fraune & M. Zimmer, unpublished data), we suggest that these marine isopods – although living in a microbe-rich environment with a high risk of bacterial infection – do not harbour hepatopancreatic bacteria. All other isopod species investigated contained bacteria in their midgut glands. The presence and identity of hepatopancreatic bacteria coincided nicely with the phylogenetic position of their isopod hosts (Fig. 2).
The reasons for the exclusion of bacteria from the hepatopancreas of marine isopods are not clear. Sleeter et al. (1978) proposed the production of antibiotics in the gut of the wood-boring Limnoria tripunctata, simply because they could not find bacteria inside the gut lumen [although Zachary & Colwell (1979) found gut microorganisms in this species]. It is possible that antibiotics of endogenous or external origin prevent the colonization of the midgut glands by bacteria in marine isopods. Hellio et al. (2000) found antibiotic agents in nine of 16 species of marine macroalgae, and such compounds might still be active in the guts of their consumers (namely isopods or amphipods), preventing bacterial colonization of the hepatopancreas. In such a scenario, the acquisition of hepatopancreatic symbionts in other isopod species (as shown herein) would simply be the consequence of an evolutionary change in feeding habit, replacing a diet of fresh algae that is potentially rich in antibiotics (Hellio et al., 2000) with detritus or microbial epibionts, lacking the antibiotic agents of macroalgal origin. The former is the major food source of Idotea balthica and other Valvifera and Sphaeromatidea, while the latter is the preferred food source of Asellus aquaticus (see below), Ligia oceanica and other oniscid isopods.
Plante et al. (1990), however, proffer a different explanation, emphasizing the importance of the microhabitat in being a successful host to microbial endosymbionts. While the gut lumen of marine invertebrates exhibits conditions virtually identical with the external environment in terms of ionic and osmotic levels, this is not the case in freshwater invertebrates. The gut of terrestrial invertebrates provides an aquatic habitat in a terrestrial environment. Thus, the significance of microbial gut symbionts can be explained from different points of view: (1) for marine microorganisms, it is not advantageous to colonize the gut of an invertebrate instead of living in sea water, except for being protected from other consumers, whereas the gut of a terrestrial invertebrate grants a more favourable environment than soil or leaf litter; and (2) for terrestrial invertebrates, gut symbionts in a stable environment are more favourable partners to rely on in terms of facilitation in a hostile environment than soil- or litter-colonizing bacteria that inhabit a variable milieu, while this difference is thought to be insignificant for marine invertebrates (Harris, 1993).
Our present results, together with a previous study by Zimmer & Bartholmé (2003), clearly prove the presence of bacterial symbionts in the hepatopancreas of Asellus aquaticus. Considering the similar nutritive ecology of freshwater and terrestrial isopods, both being detritivores that feed on leaf litter, the presence of bacteria in the midgut glands may reflect a convergent adaptation to the same food source in different taxonomic branches of isopods (cf. Zimmer & Bartholmé, 2003). On the other hand, Asellota are closely related to, and may even be the sister group of, Oniscidea (cf. Fig. 2). Thus, we may propose a common marine ancestor of these taxa to have acquired the ability to harbour bacterial symbionts in the midgut glands by feeding on food sources other than macroalgae (see above). This evolutionary achievement of hepatopancreatic symbionts may then have been a prerequisite for the utilization of terrestrial food sources (Zimmer et al., 2001), be it in freshwater or terrestrial habitats. However, hepatopancreatic bacteria in A. aquaticus are not only phylogenetically diverse (two or three phylotypes from different proteobacterial lineages), but also differ between isopod populations collected in geographically close sites (Plön and Kiel, Germany). Moreover, the symbionts we detected in A. aquaticus are not related to those of terrestrial isopods (Wang et al., 2004a, b), indicating that the midgut glands of terrestrial and freshwater isopods harbour different bacterial species with presumably different relationships to their hosts.
In contrast to A. aquaticus, the number of potential hepatopancreatic symbionts appears to be limited to a few bacterial species in Oniscidea. We could unambiguously show that various oniscid isopod species of the Crinocheta (see Fig. 2) from different regions and habitats consistently harboured one of two different bacterial symbionts. ‘Candidatus Hepatoplasma crinochetorum’ and ‘Candidatus Hepatincola porcellionum’ were present in a total of six and three crinochete isopod species from European and North American locations, respectively. This strongly indicates that symbiotic associations with these two bacterial symbionts are ubiquitous among the Crinocheta (for the semiterrestrial genus Ligia, see below). The high incidence (up to 90%) of aposymbiotic individuals, however, indicates a facultative rather than an obligate association between hepatopancreatic symbionts and isopods.
In contrast to the crinochete isopod species tested in the present study, the semiterrestrial Ligia oceanica harboured symbionts of the genus Pseudomonas. Rod-shaped bacteria were observed also in some individuals of Porcellio scaber, Oniscus asellus and Alloniscus perconvexus (present study; Hopkin & Martin, 1982; Wood & Griffith, 1988; Ullrich et al., 1991; Zimmer et al., 2002); based on FISH, we tentatively classified them as members of the genus Pseudomonas (S. Fraune & M. Zimmer, unpublished data).
From our present results on bacterial symbionts in Oniscidea, we conclude that the colonization of supralittoral habitats by the common ancestor of all terrestrial isopods was accompanied by the acquisition of bacterial symbionts. Although a more detailed assay would be required to decide upon this issue unambiguously, we propose that these initial hepatopancreatic symbionts were ancestors of Pseudomonas sp. that are still harboured in the semiterrestrial genus Ligia, representing an evolutionary prototype of Oniscidea (Schmalfuss, 1978; Carefoot & Taylor, 1995), as well as presumably in at least some crinochete species (see above). These initial symbionts of Oniscidea seem to have been supplemented during the subsequent phylogenetic radiation of terrestrial isopods by new symbionts (‘Candidatus Hepatoplasma’ and ‘Candidatus Hepatincola’), which are currently present in various species of crinochete terrestrial isopods (cf. Fig. 2).
Such an evolutionary takeover of a host has recently been demonstrated experimentally in aphids (Hemiptera: Aphidina), where the primary symbiont, Buchnera aphidicola, is being replaced by secondary symbionts (Gammaproteobacteria) (Koga et al., 2003). Similarly, tsetse flies (Glossina spp.; Diptera: Glossinidae) may lodge a commensal, Sodalis glossinidius (Gammaproteobacteria), as secondary symbiont that belongs to the same family as the primary symbiont, Wigglesworthia glossinidia (Dale & Maudlin, 1999). In dryophtherid beetles (Coleoptera: Curculionoidea) even three lineages of Enterobactericeae exist that colonized their host in independent evolutionary steps (cf. Nardon et al., 2003).
Applying a model by Doebeli & Dieckmann (2000), we hypothesize that either the initial acquisition of Pseudomonas sp. or the subsequent symbiotic takeover may have led to the branching and later speciation within the Oniscidea, as has been discussed by Buckling & Rainey (2002) in broader context. Leonardo & Muiru (2003) recently presented evidence for nutritional specialization being determined by the identity of facultative bacterial symbionts in aphids.
Hepatopancreatic bacteria in both terrestrial (cf. Zimmer, 2002) and aquatic isopods (Zimmer & Bartholmé, 2003) are thought to contribute to the utilization of low-quality food sources. Notably, members of the proteobacterial genera Burkholderia, Aeromonas and Pseudomonas, close relatives of the symbionts of A. aquaticus and L. oceanica, are capable of degrading plant polymers, phenolic compounds or polycyclic aromatic hydrocarbons. Recently, a symbiotic Pseudomonas sp. in the gut of Tetraponera ants (van Borm et al., 2002) was found to be involved in nitrogen fixing. Although nitrogen is abundant in fresh seaweeds, it is significantly reduced in decaying seaweed detritus, one of the major food sources of Ligia, and nitrogen-fixing hepatopancreatic symbionts may have helped the semiterrestrial L. oceanica and its evolutionary kin in colonizing terrestrial habitats. In the case of the symbionts of terrestrial isopods, however, the next relatives (distantly related members of the Entomoplasmatales and Ricketsiales) do not offer any clues with respect to their possible function in the symbiosis (Wang et al., 2004a, b). The nature of these associations and their significance for their hosts remain to be elucidated.
Transmission of symbionts
The hepatopancreatic symbionts of terrestrial isopods appear to be transmitted either horizontally among syntopic conspecifics or through the environment. Symbiotic bacteria were not detected in soils and leaf litter samples by FISH (data not shown), but the number of symbionts in the environment – and that of the symbionts necessary successfully to establish the symbiosis in the juvenile host – may be rather low.
In the offspring of P. scaber raised in the laboratory under sterile conditions, neither embryos nor juveniles harboured any hepatopancreatic bacteria. However, if kept under nonsterile conditions, bacteria were detected in the midgut glands of juveniles 1 week after hatching. Since we did not surface-sterilize eggs in the brood pouch, we cannot exclude maternal smearing of eggs either inside, or prior to their transport into, the brood pouch. However, if this played a role in symbiont transmission, we would expect embryos and juveniles under sterile conditions to be infected with symbionts, too. As clearly shown by the present results, this is not the case.
The transmission of symbionts is critical in mutualistic symbioses, both for obligatory symbionts and for their host. It is only through vertical transmission that a host can ensure that none of its progeny will stay aposymbiotic, and many obligate symbioses are based on this mechanism. Nevertheless, many mutualists still rely on horizontal transmission (discussed in Wilkinson & Sherratt, 2001), and horizontal transfer of symbionts seems to be the rule in sexually reproducing animals with symbionts that are not harboured inside the reproductive tract (Douglas, 1995): larvae of pony fish (Leiognathus nuchalis; Perciformes: Leiognathidae) are aposymbiotic when they hatch and become infected by luminescent bacteria at an age of at least 45 days (Wada et al., 1999). Similarly, aposymbiotic juveniles of lucinid mussels acquire sulfide-oxidizing intracellular gill symbionts (Gammaproteobacteria) horizontally.
Given that competition between different symbionts may exert negative effects on the host (discussed in Koga et al., 2003), it seems advantageous to the host to keep the diversity of symbionts low. Again, the best way to achieve this aim is controlled vertical transmission, but even those systems dependent on environmental transmission have developed strategies to master this problem. The flatworm Convoluta roscoffensis harbours endosymbiotic algae of the genus Tetraselmis, but host individuals differ in what species of Tetraselmis they accommodate. Aposymbiotic juveniles take up algae while feeding and reject all but one species (Douglas, 1980). A similar transmission mechanism, with bacteria being ingested along with the food, has been proposed for the sponge Halichondria panicae and its symbionts of the genus Rhodobacter (Alphaproteobacteria) (Althoff et al., 1998; discussed in Stackebrandt & Pukall, 1999; Müller, 1999).
Horizontal transfer of symbionts via faeces or litter, as suggested by our findings in P. scaber (see above), seems to be inefficient, since it apparently does not ensure that all offspring gain access to (the correct) symbionts, indicated by the often low prevalence of symbionts in the hosts tested herein. Nonetheless, despite the lack of detection of the symbionts outside of their hepatopancreatic habitat, ‘Candidatus Hepatincola’ and ‘Candidatus Hepatoplasma’ are present in 10–100% of individuals tested, suggesting that (1) they are common enough in the environment to ensure frequent infection of hosts, and (2) at least one of the symbiotic partners invests in finding, or being found by, an associate. Since the prevalence of symbiotic bacteria is mainly affected by both transmission rate and fitness effects for the host, and transmission rate is potentially low owing to the environmental transmission route, we assume positive effects of this symbiont to its host, but this remains to be tested.
This study was supported by the Deutsche Forschungsgemeinschaft, grant Zi 633/3-1+2. We thank Ulrich Stingl (Konstanz) for valuable discussion, Malte Mews (Kiel) for collecting and culturing I. balthica and A. aquaticus, and Karen A. Brune for proofreading the manuscript.
- 2003) Control of tsetse flies and trypanosomes using molecular genetics. Vet Parasitol 115: 125–145. (
- 1996) The oligonucleotide probe database. Appl Environ Microbiol 62: 3557–3559. , , , & (
- 1998) Evidence for a symbiosis between bacteria of the genus Rhodobacter and the marine sponge Halichondria panicea: harbor also for putatively toxic bacteria? Mar Biol 130: 529–536. , , , & (
- 1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402. , , , , , & (
- 1990) Fluorescent-oligonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology. J Bacteriol 172: 762–770. , & (
- 1993) Use of a simplified cell blot technique and 16S rRNA-directed probes for identification of common environmental isolates. Appl Environ Microbiol 59: 3219–3224. , & (
- 1981) Gene organization and primary structure of a ribosomal RNA operon from Escherichia coli. J Mol Biol 148: 107–127. , , & (
- 1991) A phylogenetic analysis of the isopoda with some classificatory recommendations. Mem Queensland Mus 31: 143–204. & (
- 2002) The role of parasites in sympatric and allopatric host diversification. Nature 420: 496–499. & (
- 1995) Ligia: a prototypal terrestrial isopod. Crustacean Issues 9: Terrestrial Isopod Biology (AlikhanMA, ed), pp. 47–60. Balkema, Rotterdam. & (
- 1993) Transovarial inheritance of endosymbiotic bacteria in clams inhabiting deep-sea hydrothermal vents and cold seeps. Proc Natl Acad Sci USA 90: 5695–5699. & (
- 1999) Sodalis gen. nov. and Sodalis glossinidius sp. nov., a microaerophilic secondary endosymbiont of the tsetse fly Glossina morsitans morsitans. Int J Syst Bacteriol 49: 267–275. & (
- 2000) Evolutionary branching and sympatric speciation caused by different types of ecological interactions. Am Nat 156: S77–S101. & (
- 1972) Étude histologique, ultrastructurale et expérimentale des caecums digestifs de Porcellio dilatatus et Ligia oceanica (Crustacea, Isopoda). Tethys 4: 145–162. & (
- 1980) Establishment of the symbiosis in Convoluta roscoffensis. J Mar Biol Assoc UK 63: 419–434. (
- 1995) The ecology of symbiotic micro-organisms. Adv Ecol Res 26: 69–103. (
- 2001) Axial differences in community structure of crenarchaeota and euryarchaeota in the highly compartmentalized gut of the soil-feeding termite Cubitermes orthognathus. Appl Environ Microbiol 67: 4880–4890. , , & (
- 1989) The structure and function of the digestive system of terrestrial isopods. J Zool Lond 217: 599–627. & (
- 1993) The presence, nature, and role of gut microflora in aquatic invertebrates: a synthesis. Microb Ecol 25: 195–231. (
- 2000) Inhibition of the development of microorganisms (bacteria and fungi) by extracts of marine algae from Brittany, France. Appl Microbiol Biotechnol 54: 543–549. , , , & (
- 1999) Molecular analyses of the methane-oxidizing microbial community in rice field soil by targeting the genes of the 16S rRNA, particulate methane monooxygenase, and methanol dehydrogenase. Appl Environ Microbiol 65: 1980–1990. , & (
- 1999) The evolution of mutualisms: exploring the paths between conflict and cooperation. Trends Ecol Evol 14: 49–53. , , & (
- 1982) The distribution of zinc, cadmium, lead and copper within the hepatopancreas of a woodlouse. Tissue Cell 14: 703–715. & (
- 2003) Changing partners in an obligate symbiosis: a facultative endosymbiont can compensate for loss of the essential endosymbiont Buchnera in an aphid. Proc Roy Soc Lond B 270: 2543–2550. , & (
- 2004) MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinform 5: 150–163. , & (
- 2003) Facultative symbionts are associated with host plant specialization in pea aphid populations. Proc Roy Soc Lond B 270: S209–S212. & (
- 1992) Phylogenetic oligodeoxynucleotide probes for the major subclasses of Proteobacteria: problems and solutions. System Appl Microbiol 15: 593–600. , , , & (
- 1998) Design and evaluation of useful bacterium-specific PCR primers that amplify genes coding for bacterial 16S rRNA. Appl Environ Microbiol 64: 795–799. , , , , , & (
- 2002) Unseen forces: the influence of bacteria on animal development. Dev Biol 242: 1–14. (
- 1999) 16S rRNA sequences as diagnostic tools to elucidate potential symbiotic relationships between bacteria and the marine sponge Halichondria panicea: reply to Stackebrandt and Pukall (Mar Biol 133: 159–161). Mar Biol 134: 213–215. (
- 1987) Detection and localization of single base changes by denaturing gradient gel electrophoresis. Methods Enzymol 155: 501–527. , & (
- 2003) Symbiosis in the dryophthoridae weevils (Coleoptera, Cuculionoidea): morphological variability of symbiotic intracellular bacteria. Symbiosis 34: 231–251. , , , & (
- 1990) Digestive associations between marine invertebrates and bacteria. Annu Rev Ecol Syst 21: 93–127. , & (
- 1997) Reassessment of the taxonomic position of Rickettsiella grylli. Int J Syst Bacteriol 47: 1255–1257. , , & (
- 1978) Ligia simoni: a model for the evolution of terrestrial isopods. Stuttg Beitr Natkde 317: 1–5. (
- 1989) Phylogenetics in Oniscidea. Monit Zool Ital (NS) Monogr 4: 3–27. (
- 2001) Morphology and evolution of respiratory structures in the pleopod exopodites of terrestrial isopoda (Crustacea, Isopoda, Oniscidea). Acta Zool 82: 315–330. & (
- 1978) Relationships between marine microorganisms and the wood-boring isopod Limnoria tripunctata. Mar Biol 45: 329–336. , , & (
- 1999) Deriving taxonomic decisions from 16S rDNA: a case study–Response to Althoff et al. (Mar Biol 130: 529–536). Mar Biol 130: 529–536. & (
- 1999) Contribution a la connaissance de la phylognénie des isopoda (Crustacea). Vie et Milieu 49: 163–176. & (
- 1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680. , & (
- 1991) Bacteria on the food, in the intestine and on the faeces of the woodlouse Oniscus asellus (Crustacea, Isopoda). Pedobiologia 34: 41–51. , & (
- 1963) Rickettsiosis in the Gryllidae (Orthoptera). C R Hebd Seances Acad Sci 256: 1045–1047. & (
- 2002) Tetraponera ants have gut symbionts related to nitrogen-fixing root-nodule bacteria. Proc Biol Sci 269: 2023–2027. , , & (
- 1999) Transfer of symbiotic luminous bacteria from parental Leiognathus nuchalis to their offspring. Mar Biol 135: 683–687. , , & (
- 1989) Evolution und phylogenetisches System der Isopoda. Zoologica 47: 1–226. (
- 1993) Probing activated sludge with oligonucleotides specific for proteobacteria: inadequacy of culture-dependent methods for describing microbial community structure. Appl Environ Microbiol 59: 1520–1525. , , & (
- 1993) Optimizing fluorescent in situ hybridization with rRNA-targeted oligonucleotide probes for flow cytometric identification of microorganisms. Cytometry 14: 136–143. , & (
- 2004a) ‘Candidatus Hepatoplasma crinochetorum,’ a new, stalk-forming lineage of Mollicutes colonizing the midgut glands of a terrestrial isopod. Appl Environ Microbiol 70: 6166–6172. , , , , & (
- 2004b) ‘Candidatus Hepatincola porcellionum’ gen. nov., sp. nov., a new, stalk-forming lineage of Rickettsiales colonizing the midgut glands of a terrestrial isopod. Arch Microbiol 181: 299–304. , , , & (
- 1991) 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173: 697–703. , , & (
- 2001) Horizontally acquired mutualisms, an unsolved problem in ecology? Oikos 92: 377–384. & (
- 1988) Bacteria associated with the hepatopancreas of the woodlice Oniscus asellus and Porcellio scaber (Crustacea, Isopoda). Pedobiologia 31: 89–94. & (
- 1979) Gut-associated microflora of Limnoria tripunctata in marine creosote-treated wood pilings. Nature 282: 716–717. & (
- 1999) The fate and effects of ingested hydrolysable tannins in Porcellio scaber. J Chem Ecol 25: 611–628. (
- 2002) Nutrition in terrestrial isopods (Isopoda: Oniscidea): an evolutionary-ecological approach. Biol Rev 77: 455–493. (
- 2003) Bacterial endosymbionts in Asellus aquaticus (Isopoda) and Gammarus pulex (Amphipoda), and their contribution to digestion. Limnol Oceanogr 48: 2208–2213. & (
- 1998a) Microorganisms and cellulose digestion in the gut of Porcellio scaber (Isopoda: Oniscidea). J Chem Ecol 24: 1395–1408. & (
- 1998b) Nutritional biology of terrestrial isopods (Isopoda: Oniscidea): copper revisited. Isr J Zool 44: 453–462. & (
- 2001) Hepatopancreatic endosymbionts in coastal isopods (Crustacea: Isopoda), and their contribution to digestion. Mar Biol 138: 955–963. , , , , , & (
- 2002) Cellulose digestion and phenol oxidation in coastal isopods (Custacea: Isopoda). Mar Biol 140: 1207–1213. , , , , & (