|Home | About | Journals | Submit | Contact Us | Français|
The European medicinal leech, Hirudo verbana, harbors simple microbial communities in the digestive tract and bladder. The colonization history, infection frequency, and growth dynamics of symbionts through host embryogenesis are described using diagnostic PCR and quantitative PCR. Symbiont species displayed diversity in temporal establishment and proliferation through leech development.
The hermaphroditic European medicinal leech (Hirudo spp.) is one of the most extensively examined animal models in neurophysiological, developmental, and behavioral studies (14) and has recently been used as a naturally occurring simple model for beneficial symbioses (5, 13). A fundamental question of microbial symbioses is how to determine the transmission mode of the symbionts between generations. Hirudo verbana reproduces by depositing eggs, which are surrounded by a cocoon. The cocoon is secreted from glandular cells of the parental mouth and usually contains 5 to 25 eggs. Each individual egg is encased by a self-enclosed vitelline membrane, referred to as the larval sac, and is bathed in a nutritious albumenous fluid (14). Complete embryonic development occurs within the cocoon and is composed of two distinct life stages, cryptolarva and juvenile. The cryptolarva transitions into a juvenile approximately midway into embryogenesis. The temporal acquisition of morphological attributes during embryonic development have been well described (3, 12, 16) (Fig. (Fig.11).
The medicinal leech houses distinct microbial communities within its digestive tract and secretory bladders. Culturing and culture-independent profiling of the European medicinal leech, H. verbana, through fluorescence in situ hybridization, study of 16S rRNA gene clone libraries, and terminal restric-tion length polymorphism, revealed a simple and stable microbial community within the adult midgut (2, 4, 7, 8, 18). The gammaproteobacterium Aeromonas veronii and a member of the Bacteroidetes, Rikenella, were identified as consistent and dominant extracellular residents of the medicinal leech crop and intestinum. An early culture-based study detected a bacterium that is now considered to be A. veronii in the cocoon albumen and in young leeches after hatching (1). In previous electron microscopy work investigating the embryonic development of the bladders, intracellular bacteria were detected within the bladder wall and extracellular bacteria within the lumen (2, 16, 17). A recent study revealed that this microbiota is organized in distinct layers and is composed of the deltaproteobacterium Bdellovibrio, betaproteobacteria Comamonas and Sterolibacterium, members of the Bacteroidetes, Sphingobacterium and Niabella, and alphaproteobacterium Ochrobactrum spp. (10). Although the microbial constituents of the adult H. verbana digestive tract have been previously characterized, the succession, inoculum sizes, frequency of infection, and growth dynamics of these symbiont species during embryogenesis remain to be described.
Putative functional roles for the crop/intestinum symbionts of the leech host include aiding in digestion, provisioning essential nutrients to the host, which are lacking in the blood meal (14), and preventing the establishment of foreign microbiota (1, 15). The symbionts localized in the bladders are suspected to play a role in the recycling of host metabolic waste into ammonia (10). The digestive tract symbionts may also display nutritional syntrophy, and possibly, A. veronii primes the host's digestive tract to enable the establishment and persistence of the obligate anaerobic Rikenella-like bacterium, thereby contributing to the selection of the future microbiota (reviewed in reference 13). This paper details the microbial colonization patterns relative to H. verbana embryogenesis using a combination of species-specific diagnostic PCR and quantitative PCR (qPCR) analyses.
If the symbionts are vertically transmitted in sufficient numbers, diagnostic PCR can be used to describe the microbial composition within the albumen bathing the embryos inside the cocoon. A breeding colony for H. verbana at the University of Connecticut was maintained at 25 ± 1°C with a 14 h-10 h light-dark cycle (12). Collection tanks contained autoclaved Instant Ocean H2O (Carolina Biological Supplies, Burlington, NC) and autoclaved peat moss bedding. Leeches received fresh, sterile sheep blood every 30 to 45 days. Leeches were mated, and observations for deposited cocoons were performed daily. Cocoons, with known deposition times, were maintained within petri dishes on sterile peat moss moistened with autoclaved Instant Ocean H2O at 20 ± 1°C. At 20 ± 1°C ambient temperature, complete embryogenesis (spanning from cocoon deposition to the emergence of adult-like juveniles) requires approximately 30 days (13; this study). Total DNA was extracted from cocoon albumen using the Holmes-Bonner method (6) and resuspended in 1× TE (10 mM Tris-1 mM EDTA, pH 8.0).
The prevalence of leech symbionts was determined by using a species-specific PCR amplification assay. Species-specific primers were chosen through the alignment of the previously characterized microbiota of the adult H. verbana crop and bladders (10, 18) using ClustalX in the Vector NTI software package (Invitrogen, Carlsbad, CA). Symbiont-specific oligonucleotides and amplicon sizes are summarized in Table Table1.1. The identities of the amplicons were confirmed through DNA sequencing. PCR amplification profiles were as follows: 95°C for 3 min, followed by 30 cycles of 95°C for 30 s, 55°C (except that a 46°C annealing temperature was used for Ochrobactrum detection) for 30 s, and 72°C for 1 min 30 s, with a final 5-min extension period at 72°C. For all amplifications, we used 0.1 μl AmpliTaq (Applied Biosystems, Foster City, CA), 0.2 μl of each primer at a concentration of 10 μm, and 1 μl of DNA template (≥1 ng) in a 20-μl reaction. The host cytoplasmic actin gene (Hirudo Act1, NCBI accession no. DQ333328) served as a control for input DNA quality. A minimum of three individuals per developmental time period (extracted from different cocoons) was analyzed. All negative samples were subjected to a parallel PCR amplification to confirm the absence of symbionts. Furthermore, a portion of the negative samples was also reamplified in a nested PCR approach to further support their lack of presence. Negative controls were included in each set of amplification reactions. The amplification products were analyzed by agarose gel electrophoresis and visualized with Kodak 1D image analysis software. DNA sequencing was performed at the West Virginia University Department of Biology Genomics Center on an ABI 3130xl analyzer (Applied Biosystems, Foster City, CA) using a 3.1 BigDye protocol (Applied Biosystems).
Within the cocoons, an inoculum consisting of A. veronii and Niabella, Ochrobactrum, Comamonas, and Sterolibacterium spp. was detectable at ≤24 h postdeposition (PD) (Fig. (Fig.2).2). Amplicons had ≥99% nucleotide identities to the 16S rRNA sequences of the microbial species previously reported within adults (10, 18). The presence of these symbiont species, immediately following cocoon deposition, strongly supports their vertical transmission. The transmission could be mediated by containing bacteria intracellularly within the developing embryo, extracellularly in the larval sac of each individual, or within the nutritious albumen surrounding the cryptolarvae within the cocoon. The remaining three symbiont species examined, Rikenella, Bdellovibrio, and Sphingobacterium spp., were not detected at this early time point but could be present within the cocoon below the level of detection.
The temporal pattern of symbiont presence was evaluated in cocoons at 6, 17, 21, and 25 days PD. The cocoons were surface sterilized using 10% bleach, followed by 70% ethyl alcohol washes. The spongy cocoon outer layer was removed using a sterile razor blade, and developing immature leeches (known as cryptolarvae and juveniles depending on days PD [Fig. [Fig.1])1]) were surgically removed and dipped in sterile TE buffer prior to DNA isolation to minimize any contamination arising from the cocoon. The examination of symbiont prevalence within individual hosts (Fig. (Fig.2)2) was first performed with cryptolarvae at 6 days PD, as this is the earliest point in development when single embryos can be differentiated and dissected from the cocoon. From 6 days PD and beyond, A. veronii and Niabella, Comamonas, and Ochrobactrum spp. were continually detected in the majority of individuals, suggesting either the incorporation of a founding seed population within the embryos (rather than within the cocoon albumen), perhaps through a transovarial or paternal route, or the invasion of cryptolarvae from the surrounding albumen early in embryogenesis. The consistent detection of these microbial species throughout development supports a high degree of coordination in their life history with the leech host and raises intriguing questions about the extent of coevolution. The prevalence of Sphingobacterium, Sterolibacterium, and Bdellovibrio spp. varied among the sampled embryonic time points, appearing only transiently during embryogenesis. Interestingly, Ochrobactrum, Comamonas, Bdellovibrio, and Sphingobacterium species were consistently detected in adult bladders, while Sterolibacterium and Niabella species were observed at a lower frequency (10). Our results suggest modifications in the composition of the symbiont populations during the transition from the embryonic to the adult life stage that may arise due to a variety of factors, including alterations in the external environment (e.g., adult leeches and cocoons are maintained at different ambient temperatures) or host physicochemical modifications (e.g., related to dietary changes from albumen to blood). An alternative explanation is that different bladders in the adult harbor subsets of the symbionts, while our sampling scheme of the embryos included all of the bladders.
Based on the inability to detect Comamonas and A. veronii in some and Sterolibacterium in all of the cryptolarvae at 6 days PD, but the detection of these organisms within cocoons at ≤24 h PD (i.e., in total cocoon fluid, consisting of albumen and embryos), we hypothesize that these symbionts reside within the albumenous fluid rather than within individual embryos at the onset of cryptolarval development. In contrast, Ochrobactrum spp. were consistently detected in all of the cryptolarvae. Given the intracellular location of the Ochrobactrum spp. and the detection of intracellular bacteria in the embryonic nephridia, we predict that this bacterium is transmitted directly to the embryo. The frequency of infection with the dominant embryonic symbionts (i.e., A. veronii and Niabella) increased with the progression of host development.
The final symbiont to be detected was Rikenella, which could be amplified only with species-specific diagnostic PCR from juveniles 2 weeks following their emergence from cocoons but prior to receiving a blood meal. Interestingly, the last 8 to 10 days of cocoon development are devoted mostly to the morphogenesis of the gut epithelium (12) (Fig. (Fig.1).1). Consistent with previous studies, such as that of the colonization of the human intestinal microbiota during the first year of life (11), the earliest microbial colonizers are typically characterized as facultative anaerobes (e.g., A. veronii and Niabella and Ochrobactrum spp.), whereas the last of the dominant symbionts is likely an obligate anaerobe (Rikenella). Two weeks following the emergence of the cocoon by juvenile leeches, 100% of all individuals were colonized with the adult-like microbiota (A. veronii and Rikenella, Ochrobactrum, Niabella, Comamonas, and Sphingobacterium spp.).
A qPCR assay was used to examine symbiont inocula within cocoons, as well as the relative growth dynamics of leech symbionts through embryonic development. We chose to analyze the proliferation of A. veronii and Ochrobactrum and Niabella spp., as these are the more prevalent symbionts, through host embryogenesis. qPCR was performed on cocoon albumen at ≤24 h PD, and individuals were extracted at 6, 17, 21, and 25 days PD using an iCycler iQ real-time PCR detection system (Bio-Rad, Hercules, CA). For each corresponding time point, measurements were collected from at least three samples. A. veronii gyrB (NCBI accession no. AY299320) and species-specific portions of the 16S rRNA gene were used for Niabella and Ochrobactrum species quantification, respectively. The H. verbana cytoplasmic actin gene (Hirudo Act1) was used for normalization of samples. The fluorometric intensity of SYBR green I dye (Bio-Rad, Hercules, CA) was used for gene quantification. The PCR amplification parameters, reagents, primer design criteria, and sequences used for quantification are summarized in Table Table22.
Symbiont density is defined as the relative ratio of symbiont genes to host cell number. Internal standard curves were generated for each primer combination using the four above-mentioned amplicons, each cloned into pGEM-T vector (Promega, Madison, WI). Density estimates were obtained by comparison to a standard curve using Bio-Rad iCycler iQ multicolor real-time PCR data analysis software. Quantitative measurements were carried out using 96-well plates in duplicate. Replicates were averaged for each sample prior to the construction of relative copy number ratios. Negative controls were included in all amplification reactions. Values are represented as means (± standard errors of the means).
Data were analyzed using JMP 7.0 software (SAS Institute, Cary, NC). Analysis of variance and post hoc pairwise comparison of the means were performed to determine whether densities significantly differed between time points. The normality of density distributions was determined with a goodness-of-fit test. Symbiont densities were log transformed to satisfy normality.
The time after cocoon deposition was shown to significantly affect the density of the examined symbiont species (F = 56.1, P < 0.0001), suggesting oscillating population dynamics through host embryogenesis. At ≤24 h PD, the three symbiotic species are found at different abundances within the cocoon albumen, with A. veronii at a significantly lower density than the comparable abundances of Ochrobactrum and Niabella spp. (F = 9.38, P < 0.0001; results not shown). These differences in inoculum sizes within an individual cocoon may be a factor impacting the frequency of infected individuals or reflect a difference in the transmission of bladder symbionts rather than crop symbionts.
Correspondingly, these three symbiont species also displayed diverse growth patterns within individual cryptolarvae and juvenile leeches (Fig. (Fig.3).3). The A. veronii and Niabella are found at higher concentrations during the onset of embryogenesis that decreases toward midembryogenesis, although statistical significance was obtained only for the former symbiont. This initial decrease is likely due to the more rapid increase in the host actin signal, as the leech cells proliferate during development. In contrast to the increase in density of Ochrobactrum and Niabella spp. during late embryonic development (i.e., juveniles at 21 versus 25 days PD), A. veronii remained stagnant. Typically, A. veronii is harbored at low levels within starved adult leeches (sometimes below the limit of detection), with its abundance rapidly increasing following the consumption of a blood meal (9).
Intuitively, one assumes that extracellular digestive tract symbionts are acquired horizontally after hatching or birth. We present evidence that supports the vertical transmission of A. veronii to its leech host. Our experiments reveal the dynamic nature of symbiont invasion through host development by microbial species in terms of inoculum size, time of establishment, and frequency of infection even by a relatively simple microbiota. Synchronization of host development with the establishment of particular members of the microbial community is likely important for niche accessibility, particularly given the microbial stratification observed within adult hosts (10). The microbial stratification also suggests the presence of intimate interactions, not only host-microbe but also microbe-microbe mediated, which will affect the selective process, perhaps by promoting the recruitment and stabilization of heterospecifics. Competition or, contrastingly, cooperation between host-associated microbes could affect density, virulence, or transmission modes of cohabitating microorganisms. Future studies will explore the potential role of symbiont species in promoting host establishment by other microbial species.
We thank Yoshitomo Kikuchi for sharing unpublished sequence data on the microbial community of the adult H. verbana bladder and for his critical review of the manuscript. We also thank members of the French lab (UCSD), particularly Joyce Murphy, for their guidance in leech breeding. We are grateful to Adam Silver for his suggestions on the manuscript and to Zach Fowler for his assistance with the statistical analysis.
This research was supported by the NSF CAREER grant MCB 0448052 and the University of Connecticut Research Foundation grant to J.G. and by the West Virginia University Research Foundation grant to R.V.M.R.
Published ahead of print on 31 July 2009.