FEMS Microbiology Ecology, 93, 2017, fix025
doi: 10.1093/femsec/fix025
Advance Access Publication Date: 23 February 2017
Research Article
RESEARCH ARTICLE
Biparental transmission of Verminephrobacter
symbionts in the earthworm Aporrectodea tuberculata
(Lumbricidae)
Laura-Carlota Paz1 , Andreas Schramm1,∗ and Marie Braad Lund1,2
1
Section for Microbiology, Department of Bioscience, Aarhus University, Ny Munkegade 116, Building 1540,
8000 Aarhus-C, Denmark and 2 Aarhus Institute of Advanced Studies, Aarhus University, Høegh-Guldbergs
Gade 6B, 8000 Aarhus-C, Denmark
∗
Corresponding author: Section for Microbiology, Department of Bioscience, Aarhus University, Ny Munkegade 116, Building 1540, 8000 Aarhus-C,
Denmark. Tel: +45 6020 26 59; E-mail: [email protected]
One sentence summary: Verminephrobacter, the nephridial symbionts of earthworms, are transmitted from both parents to the next host generation,
which results in mixing of symbiont strains and may explain symbiont genome evolution and population ecology.
Editor: Julie Olson
ABSTRACT
Most lumbricid earthworms harbor species-specific Verminephrobacter symbionts in their excretory organs (nephridia).
These symbionts are vertically transmitted via the cocoon, where they colonize the embryos. Despite cospeciation
for >100 million years with their hosts, Verminephrobacter lack genome reduction and AT bias typical of evolutionary
old, vertically transmitted symbionts, caused by recurring bottlenecks. We hypothesized that biparental symbiont
transmission into the cocoon enabled genetic mixing and relieved the bottleneck, and tested biparental transmission
experimentally for V. aporrectodeae subsp. tuberculata, the specific symbiont of the earthworm Aporrectodea tuberculata,
for which aposymbiotic worm lines are available. Virgin symbiotic and aposymbiotic adult worms were tagged, mated
in pairs, separated before the start of cocoon production and their offspring assessed for Verminephrobacter. Specific PCR
detected the symbionts in 41.5% of 188 juveniles produced by 20 aposymbiotic worms; fluorescence in situ hybridization
showed a patchy but successful colonization of their nephridia. Symbionts were present in the mucus but absent in feed,
soil, and spermatophora/nephridia of the aposymbiotic partner, suggesting symbiont transfer via mucus during mating.
These results are consistent with the hypothesis that genome evolution in Verminephrobacter is distinct from other vertically transmitted symbionts due to genetic mixing during transmission, partially facilitated by biparental transmission.
Keywords: Verminephrobacter–earthworm symbiosis; symbiont transmission; vertical parental transmission; mucus; genome
evolution
INTRODUCTION
Animal–microbe symbioses are ubiquitous in nature, with bacterial symbionts providing diverse services to their hosts, such
as essential nutrients, protection or waste recycling (reviewed
e.g. in Gilbert, Sapp, and Tauber 2012; McFall-Ngai et al. 2013).
One of the key factors for the establishment and evolution of
symbioses is the mode of symbiont transmission (Bright and
Bulgheresi 2010; Ebert 2013; Salem et al. 2015): vertical transmission (VT), i.e. from parent to offspring, results over time in
cospeciation between host and symbiont (Baumann 2005; Dale
and Moran 2006; Moran et al. 2008; Bright and Bulgheresi 2010),
Received: 5 November 2016; Accepted: 21 February 2017
C FEMS 2017. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Microbiology Ecology, 2017, Vol. 93, No. 5
while the recurrent population bottlenecks during each transmission lead to enhanced genetic drift, accelerated substitution rates and ultimately genome degradation of the symbiont
(Wernegreen 2002; Moran, McCutcheon and Nakabachi 2008;
Moya et al. 2008; Toft and Andersson 2010; McCutcheon and von
Dohlen 2011; Bennett and Moran 2015). Horizontal transmission
(HT) in contrast, where the symbiont in each host generation is
acquired from the environment, facilitates genetic mixing and
host switching (Bright and Bulgheresi 2010; Ebert 2013); thus, HT
symbionts rarely show cospeciation with their hosts, and their
genomes are not or hardly reduced in size (Salem et al. 2015).
Exclusive VT or HT is rare (Moran, McCutcheon and
Nakabachi 2008; Bright and Bulgheresi 2010), and most symbionts are presumably transmitted by a mixture of VT and HT in
so-called mixed-mode transmissions (MMT) (Ebert 2013). MMT
can enhance the efficiency of symbiont transmission (Kaltz and
Koella 2003) and help evade the disadvantages of strict VT or HT
(Bright and Bulgheresi 2010; Byler et al. 2013; Ebert 2013). Even
low rates of HT in otherwise VT-dominated symbioses (Stewart,
Young and Cavanaugh 2008; Brandvain, Goodnight and Wade
2011; Sipkema et al. 2015) may prevent symbiont genome degradation (Ebert 2013), e.g. by facilitating homologous recombination events (Mouton et al. 2012; Ros et al. 2012).
Another mode of transmission, which allows for homologous
recombination and thus may help preventing genome degradation in VT symbioses, is biparental transmission: here, VT from
both parents to the same offspring leads to mixing of symbiont
populations. Its impact on symbiont genome evolution may be
comparable to MMT but direct experimental proof of biparental
transmission is rare (Moran and Dunbar 2006; Damiani et al.
2008; Watanabe et al. 2014; De Vooght et al. 2015). Here we use an
experimentally tractable model system, the Verminephrobacter–
earthworm symbiosis (Lund, Kjeldsen and Schramm 2014), to
address the link between a biparental transmission mode and
symbiont genome evolution.
Verminephrobacter are the beneficial, extracellular, speciesspecific, nephridial symbionts of lumbricid earthworms
(Oligochaeta: Lumbricidae) (Schramm et al. 2003; Lund et al.
2010a,b; Davidson, Powell and James 2013). They are vertically
transmitted via deposition into the cocoon, where they colonize
the developing embryo as shown for Eisenia fetida by Davidson
and Stahl (2006, 2008), and they display a pattern of cospeciation
with their hosts since the origin of lumbricids about 100 million
years ago (Lund et al. 2010a; Møller et al. 2015).
Despite this evolutionarily old association, Verminephrobacter
lack the genome degradation and size reduction typical of VT
symbionts of similar age (Moran, McCutcheon and Nakabachi
2008; Moran, McLaughlin and Sorek 2009; Toft and Andersson
2010). Instead, they show only slightly relaxed purifying selection but extensive genome shuffling (Kjeldsen et al. 2012; Lund,
Kjeldsen and Schramm 2014). Horizontal gene transfer and homologous recombination have been suggested as explanation
(Kjeldsen et al. 2012; Lund, Kjeldsen and Schramm 2014) and, facilitated by the ability of Verminephrobacter to take up free DNA
(Davidson et al. 2014), may take place within the mixed bacterial community of the cocoon (Zachmann and Molina 1993;
Davidson, Powell and Stahl 2010) or, for some multisymbiotic
earthworm species (Davidson and Stahl 2008; Lund et al. 2010a;
Davidson, Powell and James 2013), in the nephridia.
Biparental transmission of Verminephrobacter would further
increase the opportunity of homologous recombination by the
mixing of different symbiont strains present in a host population (Kjeldsen et al. 2012; Viana et al. 2016). The objective of this study was therefore to test the hypothesis that
Figure 1. Schematic outline of the experimental setup for biparental transfer
of Verminephrobacter in A tuberculata. Aposymbiotic and symbiotic adults were
mated and separated for reproduction, cocoons were collected and stored until
hatching, and the resulting juveniles screened for Verminephrobacter by specific
PCR and FISH.
Verminephrobacter are biparentally transmitted to the next host
generation. To this end, we utilized an earlier established
symbiont-free (aposymbiotic) line of the garden earthworm
Aporrectodea tuberculata (Lund et al. 2010b). Aporrectodea tuberculata are strictly sexually reproducing, non self-fertilizing
hermaphrodites (Sims and Gerard 1985), which carry Verminephrobacter aporrectodeae subsp. tuberculatae as their sole
nephridial symbiont (Lund et al. 2010a). Virgin aposymbiotic
A. tuberculata were mated with symbiotic worms, and the offspring of the aposymbiotic parents was then screened for Verminephrobacter by specific PCR and FISH. Possible transmission
routes were investigated by analyzing the nephridia, spermatheca, mucus and bedding soil of the aposymbiotic parents for
Verminephrobacter (Fig. 1).
MATERIAL AND METHODS
Earthworm cultures
The aposymbiotic culture of Aporrectodea tuberculata had been
established in a previous study by antibiotics treatment of the
cocoons, resulting in Verminephrobacter-free offspring (Lund et al.
2010b). Aposymbiotic and symbiotic cultures were propagated
in the laboratory as previously described (Lund et al. 2010b). All
bedding soil and feed (soil-to-cow dung mix, 3:1) used during
the experiment was dried at 80˚C for 24 h and rehydrated at the
time of use. Newly hatched juveniles were grown in groups of
five to six symbiotic or aposymbiotic worms, respectively. After
3 months, individual juveniles were transferred to separate containers with ∼100 g soil each and fed biweekly until they were
Paz et al.
sexually mature, as indicated by a fully developed clitellum. Cocoon production was not observed.
Mating experiment
One week before the mating experiment, seven symbiotic
worms were marked by injection with Visible Implant Elastomer (Northwest Marine Technology) into the body cavity
of the worm tail (Butt and Lowe 2007); this facilitated visual
differentiation of symbiotic and aposymbiotic worms. Each
marked worm was paired with an aposymbiotic worm of similar weight. Pairs were transferred into 350 g soil with 10 g
feed for a mating period of one week. The symbiotic worms
were then re-mated twice with new aposymbiotic worms to
achieve a higher number of juveniles from aposymbiotic parents
(n = 20; one worm deceased) for analysis. After mating, the
27 worms were placed individually on humid filter paper for 2
days to empty their guts. Single worms were then transferred
into 400 g soil with 10 g feed for cocoon production. Every third
week, soil and feed was replaced and cocoons (up to 10 per
worm, total n = 249) were collected. Cocoons were allowed to
hatch in petri dishes with humid filter paper (exchanged weekly)
at 15◦ C. Hatchlings were collected twice a week, rinsed with water, and one-third to half of their body length (50–90 segments)
was cut off. This tail part was frozen in liquid nitrogen and
stored at –20˚C until DNA extraction. Front ends of hatchlings
from the same parent were pooled, transferred into 100 g soil
with 5 g feed and grown up for later FISH analysis. When these
juveniles where big enough for dissection, they were rinsed with
water, the tail was cut off about 20 segments above the visible constriction of the regenerated part and transferred to 50%
ethanol for immobilization, while the front end was kept alive.
The regenerated tail below the constriction was cut off and discarded. The remaining part was dissected to remove the gut, and
body wall pieces were fixed in 4% (w/v) paraformaldehyde for
2–3 h at 4◦ C, washed twice in phosphate-buffered saline (PBS)
and stored in 70% ethanol in PBS at –20◦ C. Parent worms
were dissected, fixed and stored in the same way, and presence/absence of Verminephrobacter aporrectodeae was confirmed
by FISH (see below).
3
the worms were killed, dissected, their guts removed and then
the worms were cut into halves longitudinally along the dorsal/ventral line. One half was fixed for FISH as described above,
while from the other half four spermatheca and 56–70 nephridia
per worm (from the anterior end) were dissected and stored at
–20◦ C for DNA extraction.
DNA extraction, PCR and sequencing
Tail, spermatheca, nephridial samples, and freeze-dried mucus were digested in proteinase K and digestion buffer [50 mM
Tris-HCl, 1 mM EDTA, 0.5% SDS, 10 μL proteinase K (>800 U
μL−1 )] for 2–3 h at 55◦ C on a shaker. The digest was transferred to EconoSpin Columns (Epoch Life Sciences, TX, USA),
washed twice with washing buffer (2 mM Tris-HCl, 20 mM NaCl,
80% ethanol), eluted with 100 μL dH2 O, and stored at –20◦ C.
R
Freeze-dried soil samples were extracted using the PowerLyzer
R
Power Soil
DNA Isolation Kit (MoBio Qiagen). The 16S rRNA
gene of V. aporrectodeae was detected by nested PCR: 30 cycles of
bacteria-specific PCR (primers Eub26F; Hicks, Amann and Stahl
(1992) and 1492R; Loy et al. (2002), annealing, 57◦ C) were followed by 30 cycles of a newly developed V. aporrectodeae-specific
PCR (primers LSB145F; Schweitzer et al. (2001) and Apo1034R
(5 -TCTTTCGAGCACCTCTCCAT, this study); annealing, 64◦ C; PCR
product length, 889 bp). Primer Apo1034R was designed based on
public 16S rRNA gene sequences of V. aporrectodeae. All PCR reactions were run with TaKaRa Ex Taq Polymerase (5 U μL−1 ), 10×
Ex Taq Buffer (containing 20 mM MgCl2 ), dNTP Mixture (2.5 mM
each dNTP) and 0.2 μM of each primer. Sensitivity of the nested
PCR assay was determined in serial dilutions of genomic DNA of
V. aporrectodeae At4 in DNA extracts from a symbiont-free worm
tail or from symbiont-free soil, respectively. Given a genome size
of 4.7 MB, a single rrn operon (Kjeldsen et al. 2012) and a DNA extraction efficiency of 10%, the detection limit for V. aporrectodeae
was 10 cells in worm tails, while it was 50 cells in soil. Specificity of the PCR for V. aporrectodeae was confirmed by sequencing 12 randomly selected PCR products; these sequences were all
>99.87% identical to the 16S rRNA gene of the V. aporrectodeaetype strain At4 (FJ214186).
FISH
Assessment of symbiont transmission routes
Five pairs of symbiotic and aposymbiotic worms were mated as
described above. After separating the couples, three aposymbiotic and three symbiotic worms were kept individually on humid
filter paper, while the remaining four worms were kept individually in soil. Mucus and soil was collected immediately after separation (t1 ), and after 6 (t2 ) and 14 days (t3 ) in isolation. Mucus
excretion was induced by submitting individual worms to 5%–
10% ethanol. Worms were then transferred to tap water, where
the mucus solidified; mucus was collected from the worm surface with an autoclaved wooden stick, and either freeze-dried
(Edwards Micro Modulyo, Thermo) for 3 days and stored at –
20◦ C for DNA extraction, or fixed for subsequent FISH analysis as
described above.
Additional samples were collected from (i) the prepared feed,
(ii) the bedding before the worms were added for mating and
(iii) the bedding directly after mating. Soil of the symbiotic
stock culture, which was densely populated by symbiotic A.
tuberculata, served as positive control for successful PCR amplification of V. aporrectodea 16S rRNA genes (see below) from
soil. All soil samples were freeze-dried for DNA extraction or
fixed for FISH as described above. Two weeks after mating,
Verminephrobacter aporrectodeae subsp. tuberculatae was specifically detected by FISH with the newly designed probe At112
(5 -CCGATGTATTACTCACCC-3 ). FISH conditions were optimized
with pure cultures of V. aporrectodeae subsp. tuberculatae and V.
aporrectodeae subsp. caliginosae (Lund et al. 2012). Specific and
sensitive hybridization required equimolar amounts of an unlabeled helper probe H89 (5 -GTTCGCCACTCGTCAGCG-3 ) and
35% formamide in the standard hybridization buffer (Pernthaler
et al. 2001). Probes EUB338,-II,-III (Daims et al. 1999) and NON338
(Manz et al. 1992) were used as positive and negative controls,
respectively. All probes were synthesized and labeled with the
fluorescent dyes Cy3 or Cy5 by Biomers (biomers.net, Germany).
Nephridia were hybridized still attached to the body wall, while
spermathecae were picked individually and squeezed onto multiwell slides (Paul Marienfeld, Germany) and, just as soil and mucus samples, immobilized by drying. Hybridizations were carried
out following published protocols (Pernthaler et al. 2001). Hybridized specimens were mounted in a Vectashield (Vector Laboratories Inc., Burlingame, CA)—Citifluor (Citifluor Ltd, London,
UK) mix (3:1) containing 4 ,6-diamidino-2-phenylindole (DAPI;
1 μg mL−1 ), and were imaged with an Axiovert 200 M epifluorescence microscope (Carl Zeiss, Jena).
4
FEMS Microbiology Ecology, 2017, Vol. 93, No. 5
Figure 2. Aporrectodea tuberculata juveniles colonized with Verminephrobacter when produced by symbiotic (red) or aposymbiotic (blue) parents. Dark bars, Verminephrobacter positive; light bars, Verminephrobacter negative. Every symbiotic parent worm (A–G; n = 7) was consecutively mated with two to three aposymbiotic worms (1–20),
i.e. symbiotic worm A was mated with aposymbiotic worm 1, then 2, then 3, etc.
Statistical analysis
All statistical analyses were performed in R version 3.2.1
(R-Development-Core-Team, 2015) with RStudio interface version 0.99.902 with inbuilt statistical methods (Wilcoxon test,
ANOVA and Pearson’s product-moment correlation).
RESULTS
Verminephrobacter are transmitted to the offspring
of aposymbiotic parents
All 27 worms used in the mating experiment started cocoon production within 3 weeks after mating, with the majority of cocoons produced 20–44 days after mating. Within 3 months after mating, 10 cocoons could be collected per parent, except for
three symbiotic and four aposymbiotic worms that produced
only 3–9 cocoons each. Specific PCR of the juveniles showed
that 10%–80% of the offspring produced by the 20 aposymbiotic worms contained Verminephrobacter (Fig. 2). This indicates that the symbiotic parent had transferred Verminephrobacter to its aposymbiotic partner during mating, which then deposited the symbionts into some if its cocoons, resulting in
colonization of the developing worm. Overall, 41.5% of the
188 juveniles produced by aposymbiotic parents were symbiotic (Fig. 2). The frequency of symbiotic juveniles was independent of the number of mating events, i.e. the aposymbiotic
parent mated first did not produce a higher fraction of symbiotic offspring compared to the parent mated last (ANOVA,
p = 0.522; Fig. 2).
The frequency of symbiotic juveniles was also independent
of the time between mating and cocoon production, i.e. juveniles
from cocoons produced directly after mating were not more frequently colonized than juveniles from cocoons produced at the
end of the experiment (Fig. 3A). Apparently, Verminephrobacter
can be stored for at least 3 months and consistently deposited
into the cocoons by the aposymbiotic parent. There was no correlation between the time that the juveniles had spent in the cocoon prior to hatching and the frequency of symbiont colonization; symbiotic and aposymbiotic offspring took almost identical times from cocoon deposition to hatching of the juvenile, on
average 62 and 63 days, respectively (Fig. 3B).
FISH confirmed that the symbiotic parents were indeed fully
colonized with Verminephrobacter and that aposymbiotic parents
indeed were symbiont free. FISH on seven juveniles originating from the two aposymbiotic adults with the highest number
of symbiotic offspring revealed that, although Verminephrobacter was detected in all juveniles, not all nephridia were colonized (Fig. 4). In four worms, only 13%–26% of the analyzed
nephridia (n = 61–98 per specimen) contained Verminephrobacter;
nephridial colonization was patchy and did not follow a specific
pattern. The remaining three worms were fully colonized. Finally, specific PCR on 70 juveniles produced by the seven symbiotic parents (intended as positive control for symbiont transmission) showed, to our surprise, that only 54.3% of this ‘symbiotic’
offspring in fact contained Verminephrobacter (Fig. 2). Symbiont
transmission efficiency among individual parents was highly
variable (0%–100%). FISH on 12 of these juveniles overall confirmed the PCR data and showed both absence of Verminephrobacter, colonization of all nephridia and, as described above for
the offspring of the aposymbiotic parents, patchy colonization
(n = 20–30 nephridia per specimen). However, FISH detected
Verminephrobacter in slightly more juveniles than the PCR approach. One possible explanation for the higher sensitivity of
the FISH compared to the PCR approach is that earthworm embryos are colonized head to tail as each worm segment matures
(Davidson and Stahl 2008), and it is unknown if the colonization
of tail nephridia is complete. Since PCR was done on tails, while
FISH was done on middle segments, the PCR may have underestimated the number of symbiotic offspring, and colonization
Paz et al.
5
Figure 4. FISH image of four nephridia attached to the body wall of a
juvenile originating from an aposymbiotic parent. Ampullas are indicated by
arrows. Double hybridization with the bacterial probe EUB-Cy5 (displayed in
magenta) and the Verminephrobacter-specific probe LSB145-Cy3 (displayed in yellow) renders Verminephrobacter red in nephridium 4, while nephridia 1–3 are not
colonized. Worm nuclei appear blue from DAPI staining; worm tissue (including nephridial tubes) shows yellow-greenish autofluorescence. The bright green
blurs in the upper part are the highly autofluorescent setae of the worm, indicating the middle of a segment. Scale bar, 100 μm.
Figure 3. (A) Proportion of Verminephrobacter-positive juveniles produced by
aposymbiotic parents. Cocoons were collected for 3 months after mating. The
size of the circles reflects the total number of juveniles collected at each
time point. There was no correlation between collection date and percentage
of Verminephrobacter-positive juveniles (Pearson’s product-moment correlation;
p-value = 0.7537). (B) Hatching time (from cocoon collection to hatching) of
Verminephrobacter-positive and negative juveniles. Boxplots show median and
first quartile, and the whiskers the third quartile. Hatching time was not significantly different between Verminephrobacter-positive and negative juveniles
(Wilcoxon; p-value = 0.284).
frequencies reported in Fig. 2 should be viewed as minimum estimates.
Mucus as possible transmission route?
To exclude HT of contaminating Verminephrobacter from soil or
feed, soil samples from different steps of the mating experiment
and prepared feed were screened by specific PCR. All replicate
soil and feed samples (n = 2–5 per sample type) proved negative for Verminephrobacter, except for two out of five samples of
the soil used for mating. Since one symbiotic worm had been
present in this soil for 1 week, shedding Verminephrobacter, the
detection of symbionts in these samples is not surprising. Most
importantly, Verminephrobacter was not detected in the soil in
which the aposymbiotic worms produced their cocoons, neither
after 6 (t1) nor after 14 days (t2).
If the symbionts are transferred from the symbiotic parent
to the aposymbiotic parent during mating, and then are deposited into the cocoons of the aposymbiotic worms over a period of 3 months, they must be stored and possibly propagated
by the aposymbiotic worm. Three potential sites of storage,
and thus ultimately routes of transmission, were tested by both
specific PCR and FISH in the aposymbiotic parents: spermatheca, nephridia and mucus. Spermatheca (n = 10) were always
negative for Verminephrobacter, as were the nephridia (n = 150).
Mucus could not be investigated by FISH due to its very strong
autofluorescence. Specific PCR on mucus samples was always
negative for worms that had been kept on filter paper after mating (n = 3; tested at t1, t2 and t3). In contrast, Verminephrobacter
was detected in mucus of worms kept in soil, at t1 (2 out of 2
specimens positive), t2 (1 out of 2) and t3 (2 out of 2).
DISCUSSION
Biparental transmission
The key result of our study is the production of symbiotic
progeny by aposymbiotic worms after mating a symbiotic partner, which is evidence of paternal transmission: earthworms
are hermaphrodites that exchange sperm during mating; after
separation, each partner deposits egg, stored sperm and symbionts into the newly produced cocoons, where the colonization
takes place (Davidson and Stahl 2006, 2008). Symbiont transfer
into cocoons via gut content or soil was excluded, as worms
had emptied their guts before cocoon production started, and
soil samples always tested negative for Verminephrobacter. Thus,
symbionts must have been transferred between partners during mating, stored for up to 3 months (Fig. 3A) by the aposymbiotic parent, and then deposited into the cocoon. In nature, the
cocoon-producing parent will also deposit its own symbionts, resulting in biparental transmission. Nephridial colonization was
often incomplete and patchy in our experiment (Fig. 4), possibly due to the low number of symbionts deposited into the cocoon compared to natural biparental transmission. Since every
nephridium has to be infected separately (Davidson and Stahl
2008), low symbiont starting populations in the cocoons can impede full colonization, as demonstrated by restricting symbiont
numbers with low-dose antibiotics treatment in Eisenia fetida
(Davidson and Stahl 2006).
Similar sexual transmission of male-borne symbionts during mating has only been reported for facultative symbionts of
aphids (Moran and Dunbar 2006), Asaia symbionts of mosquitos
(Damiani et al. 2008), the secondary tsetse fly symbiont Sodalis
glossinidius (De Vooght et al. 2015) and nuclear-targeting Rickettsia
6
FEMS Microbiology Ecology, 2017, Vol. 93, No. 5
in the leafhopper Nephotettix cincticeps (Watanabe et al. 2014).
In these cases, the symbionts are associated with the male reproductive organs (Moran and Dunbar 2006) or spermatophores
(De Vooght et al. 2015), or even transferred in the sperm head
(Watanabe et al. 2014). In contrast, Verminephrobacter was not
detected in the spermatheca of aposymbiotic earthworms after mating and did not re-colonize the nephridia of the mature
aposymbiotic worms, indicating a different route of symbiont
transfer. Aeromonas veronii, the VT gut symbiont of the medicinal leech Hirudo verbena, can be horizontally transmitted via
the leech’s mucus (Ott et al. 2015, 2016), in which it proliferates
synchronously with mucus shedding frequency (Ott et al. 2016).
The detection of Verminephrobacter in the mucus of aposymbiotic worms after mating suggests a similar mechanism for
the earthworm symbionts. Stable colonization frequencies over
>3 months (Fig. 3A) further suggest that Verminephrobacter was
actively growing in the mucus. Since Verminephrobacter was only
present on worms kept in soil, not on filter paper, the otherwise
nutrient-rich mucus (Jamieson 1981) may lack certain growth
factors present in soil; alternatively, stressful conditions on filter
paper may have induced a mucus shedding frequency too high
for Verminephrobacter to keep up proliferation, leading to symbiont loss before deposition into the cocoon. Stressful conditions
on filter paper may also have increased the secretion of antibacterial peptides into the mucus (as reported for earthworms of
the genus Eisenia, Lumbricus and Pheretima; Valembois, Roch and
Lassegues 1986; Bilej et al. 2010) to a level not tolerated by the
symbionts.
The production of aposymbiotic progeny by the symbiotic
parent (Fig. 2) indicates imperfect maternal transmission (Fig. 2)
and confirms earlier data for lab cultures of E. fetida (Davidson
and Stahl 2006). In contrast, wild-caught individuals of Aporrectodea tuberculata, E. fetida or E. andrei have never been found without Verminephrobacter (Lund et al. 2010a; Davidson, Powell and
James 2013; Viana et al. 2016), which is evidence that the beneficial effect of Verminephrobacter on host fitness, i.e. earlier sexual maturity and higher offspring numbers (Lund et al. 2010b),
is strong enough for selection to remove aposymbiotic earthworms in nature.
Consequences for symbiont evolution and ecology
Genome degradation as consequence of enhanced genetic drift
due to transmission bottlenecks is common among evolutionary old symbionts with maternal VT (Wernegreen 2002; Moran,
McCutcheon and Nakabachi 2008; Moya et al. 2008; Toft and
Andersson 2010). Biparental VT can relieve this bottleneck
(Moran and Dunbar 2006); indeed, Sodalis glossinidius and Serratia symbiotica both show biparental VT (Moran and Dunbar
2006; De Vooght et al. 2015) and only moderate genome reduction
and AT bias, with chromosomal rearrangements still ongoing and mobile elements present (Toh et al. 2006; ManzanoMarı́n and Latorre 2016). The leech symbiont A. veronii on the
other hand, which shows not just biparental VT but MMT (Ott
et al. 2015, 2016), has completely evaded genome degradation
and AT bias. Verminephrobacter finally takes an intermediate position, without genome reduction and AT bias but with accelerated evolutionary rates, many mobile elements and massive
chromosome rearrangements (Kjeldsen et al. 2012; Lund, Kjeldsen and Schramm 2014). This is likely a consequence of biparental VT-enabled homologous recombination combined with
occasional horizontal gene acquisition by natural transformation from e.g. soil bacteria in the cocoon (Davidson et al. 2014).
Despite this opportunity for genetic mixing, Verminephrobacter
still shows a fairly strong cospeciation pattern with its host
(Lund et al. 2010a; Møller, Lund and Schramm 2015). Although
biparental VT allows for homologous recombination, it does not
cross species borders, and since the ability of Verminephrobacter to thrive in soil outside its host appears rather limited (Pinel
2009; this study), HT and host switching must be quite rare
events.
Last not least, our results are consistent with a recent study
of Verminephrobacter population structure in wild E. fetida, E. andrei and A. tuberculata populations (Viana et al. 2016). The highdensity, fast reproducing Eisenia had homogeneous symbiont
populations across the host population, while the low-density,
slow reproducing A. tuberculata had discrete symbiont populations in each host individual (Viana et al. 2016). This pattern can
be explained by more frequent symbiont mixing in dense worm
populations by biparental or even multiparental VT, which results in homogenization of the symbiont population.
CONCLUSION
Our study presents strong evidence for bi-parental transmission of Verminephrobacter in A. tuberculata, which starts by symbiont exchange between parents during mating. Our results
further indicate that the symbionts are stored and proliferated for >3 months in mucus on the worms’ surface, from
where they are deposited into the cocoon for colonization of
the worm embryo. Biparental transmission can partly explain
the lack of genome degradation in this evolutionarily old, VT
symbiosis, and contributes to symbiont population structure
in dense, fast-reproducing earthworm populations by homogenization the symbiont population.
ACKNOWLEDGEMENTS
We thank Britta Poulsen, Susanne Nielsen, and Annemarie
Højmark for excellent technical assistance, and Martin Holmstrup and Flávia Viana for helpful discussions and advice on
worm culturing.
FUNDING
This work was supported by The Danish Council for Independent Research |Natural Sciences (FNU) and a fellowship of the
Aarhus Institute for Advanced Studies (AIAS) to MBL.
Conflict of interest. None declared.
REFERENCES
Baumann P. Biology of bacteriocyte-associated endosymbionts
of plant sap-sucking insects. Annu Rev Microbiol 2005;59:
155–89.
Bennett GM, Moran NA. Heritable symbiosis: the advantages and
perils of an evolutionary rabbit hole. P Natl Acad Sci USA
2015;112:10169–79.
Bilej M, Procházková P, Šilerová M et al. Earthworm Immunity.
In: Söderhäll K (ed.). Invertebrate Immunity. New York: Landes
Bioscience and Springer Science+Business Media, LLC. 2010,
66–78.
Brandvain Y, Goodnight C, Wade MJ. Horizontal transmission
rapidly erodes disequilibria between organelle and symbiont
genomes. Genetics 2011;189:397–404.
Bright M, Bulgheresi S. A complex journey: transmission of microbial symbionts. Nat Rev Microbiol 2010;8:218–30.
Butt KR, Lowe CN. A viable technique for tagging earthworms
using visible implant elastomer. Appl Soil Ecol 2007;35:454–7.
Paz et al.
Byler KA, Carmi-Veal M, Fine M et al. Multiple symbiont acquisition strategies as an adaptive mechanism in the coral Stylophora pistillata. PLoS One 2013;8:e59596.
Daims H, Brühl A, Amann R et al. The Domain-specific probe
EUB338 is Insufficient for the detection of all Bacteria: development and evaluation of a more comprehensive probe set.
Syst Appl Microbiol 1999;22:434–44.
Dale C, Moran NA. Molecular Interactions between bacterial
symbionts and their hosts. Cell 2006;126:453–65.
Damiani C, Ricci I, Crotti E et al. Paternal transmission of symbiotic bacteria in malaria vectors. Curr Biol 2008;18:1087–8.
Davidson SK, Dulla GF, Go RA et al. Earthworm symbiont
Verminephrobacter eiseniae mediates natural transformation
within host egg capsules using type IV pili. Front Microbiol
2014;5:1–10.
Davidson SK, Powell R, James S. A global survey of the
bacteria within earthworm nephridia. Mol Phylogenet Evol
2013;67:188–200.
Davidson SK, Powell RJ, Stahl DA. Transmission of a bacterial
consortium in Eisenia fetida egg capsules. Environ Microbiol
2010;12:2277–88.
Davidson SK, Stahl DA. Transmission of nephridial bacteria of
the earthworm Eisenia fetida. Appl Environ Microb 2006;72:
769–75.
Davidson SK, Stahl DA. Selective recruitment of bacteria during
embryogenesis of an earthworm. ISME J 2008;2:510–8.
De Vooght L, Caljon G, Van Hees J et al. Paternal transmission of
a secondary symbiont during mating in the viviparous tsetse
fly. Mol Biol Evol 2015;32:1977–80.
Ebert D. The epidemiology and evolution of symbionts with
mixed-mode transmission. Annu Rev Ecol Evol S 2013;44:
623–43.
Gilbert SF, Sapp J, Tauber AI. A symbiotic view of life: we have
never been individuals. Q Rev Biol 2012;87:325–41.
Hicks RE, Amann RI, Stahl DA. Dual staining of natural bacterioplankton with 4 ’, 6-diamidino-2-phenylindole and
fluorescent oligonucleotide probes targeting kingdomlevel 16S rRNA sequences. Appl Environ Microb 1992;
58:2158–63.
Jamieson BGM. The Ultrastructure of the Oligochaeta. London: Academic Press Inc, 1981.
Kaltz O, Koella JC. Host growth conditions regulate the plasticity
of horizontal and vertical transmission in Holospora undulata,
a bacterial parasite of the protozoan Paramecium caudatum.
Evolution 2003;57:1535–42.
Kjeldsen KU, Bataillon T, Pinel N et al. Purifying selection and
molecular adaptation in the genome of Verminephrobacter,
the heritable symbiotic bacteria of earthworms. Genome Biol
Evol 2012;4:307–15.
Loy A, Lehner A, Lee N et al. Oligonucleotide microarray for
16S rRNA gene-based detection of all recognized lineages of
sulfate-reducing prokaryotes in the environment. Appl Environ Microb 2002;68:5064–81.
Lund MB, Davidson SK, Holmstrup M et al. Diversity and host
specificity of the Verminephrobacter-earthworm symbiosis.
Environ Microbiol 2010a;12:2142–51.
Lund MB, Holmstrup M, Lomstein BA et al. Beneficial effect
of Verminephrobacter nephridial symbionts on the fitness of
the earthworm Aporrectodea tuberculata. Appl Environ Microb
2010b;76:4738–43.
Lund MB, Kjeldsen KU, Schramm A. The earthwormVerminephrobacter symbiosis: an emerging experimental
system to study extracellular symbiosis. Front Microbiol
2014;5:1–7.
7
Lund MB, Schätzle S, Schramm A et al. Verminephrobacter aporrectodeae sp. nov. subsp. tuberculatae and subsp. caliginosae,
the specific nephridial symbionts of the earthworms Aporrectodea tuberculata and A. caliginosa. Anton Leeuw 2012;101:
507–14.
McCutcheon JP, von Dohlen CD. An interdependent metabolic
patchwork in the nested symbiosis of mealybugs. Curr Biol
2011;21:1366–72.
McFall-Ngai M, Hadfield MG, Bosch TCG et al. Animals in a bacterial world, a new imperative for the life sciences. P Natl Acad
Sci USA 2013;110:3229–36.
Manz W, Amann R, Ludwig W et al. Phylogenetic oligodeoxynucleotide probes for the major subclasses of proteobacteria:
problems and solutions. Syst Appl Microbiol 1992;15:593–600.
Manzano-Marı́n A, Latorre A. Snapshots of a shrinking partner: genome reduction in Serratia symbiotica. Nat Sci Rep
2016;6:32590.
Moran NA, Dunbar HE. Sexual acquisition of beneficial symbionts in aphids. P Natl Acad Sci USA 2006;103:12803–6.
Moran NA, McCutcheon JP, Nakabachi A. Genomics and evolution of heritable bacterial symbionts. Annu Rev Genet
2008;42:165–90.
Moran NA, McLaughlin HJ, Sorek R. The dynamics and time scale
of ongoing genomic erosion in symbiotic bacteria. Science
2009;323:379–82.
Mouton L, Thierry M, Henri H et al. Evidence of diversity and recombination in Arsenophonus symbionts of the Bemisia tabaci
species complex. BMC Microbiol 2012;12:S10.
Moya A, Peretó J, Gil R et al. Learning how to live together: genomic insights into prokaryote-animal symbioses. Nat Rev
Genet 2008;9:218–29.
Møller P, Lund MB, Schramm A. Evolution of the tripartite symbiosis between earthworms, Verminephrobacter
and Flexibacter-like bacteria. Front Microbiol 2015;6, DOI:
10.3389/fmicb.2015.00529.
Ott BM, Dacks AM, Ryan KJ et al. A tale of transmission:
Aeromonas veronii activity within leech-exuded mucus. Appl
Environ Microb 2016;82:2644–55.
Ott BM, Rickards A, Gehrke L et al. Characterization of shed
medicinal leech mucus reveals a diverse microbiota. Front Microbiol 2015;5:757.
Pernthaler J, Glöckner FO, Schönhuber W et al. fluorescence
in situ hybridization with rRNA-targeted oligonucleotide
probes. In: Paul J (ed.). Methods in Microbiology: Marine Microbiology, vol. 30. Academic Press Ltd, London, UK, 2001.
Pinel N. Physiological and genomic insight into the biology
of Verminephrobacter eiseniae, a bacterial symbiont of the
earthworm Eisenia fetida. 2009. PhD Thesis. University of
Washington, Seattle, WA, USA.
Ros VID, Fleming VM, Feil EJ et al. Diversity and recombination
in Wolbachia and Cardinium from Bryobia spider mites. BMC
Microbiol 2012;12:S13.
Salem H, Florez L, Gerardo N et al. An out-of-body experience: the
extracellular dimension for the transmission of mutualistic
bacteria in insects. P Roy Soc B-Biol Sci 2015;282:20142957.
Schramm A, Davidson SK, Dodsworth JA et al. Acidovorax-like
symbionts in the nephridia of earthworms. Environ Microbiol
2003;5:804–9.
Schweitzer B, Huber I, Amann R et al. Alpha- and betaproteobacteria control the consumption and release of
amino acids on lake snow aggregates. Appl Environ Microb
2001;67:632–45.
Sims RW, Gerard BM. Earthworms, Synopses of the British Fauna No.
31. Leiden: E.J. Brill/Dr. W. Backhuys, 1985.
8
FEMS Microbiology Ecology, 2017, Vol. 93, No. 5
Sipkema D, de Caralt S, Morillo JA et al. Similar spongeassociated bacteria can be acquired via both vertical and horizontal transmission. Environ Microbiol 2015;17:3807–21.
Stewart FJ, Young CR, Cavanaugh CM. Lateral symbiont acquisition in a maternally transmitted chemosynthetic clam endosymbiosis. Mol Biol Evol 2008;25:673–87.
Toft C, Andersson SGE. Evolutionary microbial genomics:
insights into bacterial host adaptation. Nat Rev Genet
2010;11:465–75.
Toh H, Weiss BL, Perkin SAH et al. Massive genome erosion and
functional adaptations provide insights into the symbiotic
lifestyle of Sodalis glossinidius in the tsetse host. Genome Res
2006;16:149–56.
Valembois P, Roch P, Lassegues M. Antibacterial Molecules in
Annelids. In: Brehèlin M (ed.). Immunity in Invertebrates, New
York: Springer, 1986.
Viana F, Jensen CE, Macey M et al. Earthworm ecology affects the
population structure of their Verminephrobacter symbionts.
Syst Appl Microbiol 2016;39:170–2.
Watanabe K, Yukuhiro F, Matsuura Y et al. Intrasperm vertical
symbiont transmission. P Natl Acad Sci USA 2014;111:7433–7.
Wernegreen JJ. Genome evolution in bacterial endosymbionts of
insects. Nat Rev Genet 2002;3:850–61.
Zachmann JE, Molina JAE. Presence of culturable bacteria in cocoons of the earthworm Eisenia fetida. Appl Environ Microb
1993;59:1904–10.