|Home | About | Journals | Submit | Contact Us | Français|
Transmissible spongiform encephalopathies (TSE), including bovine spongiform encephalopathy (BSE), are fatal neurodegenerative disorders in humans and animals. BSE appears to have spread to cattle through the consumption of feed contaminated with BSE/scrapie agents. In the case of an oral infection, the agents have to cross the gut-epithelial barrier. We recently established a bovine intestinal epithelial cell line (BIE cells) that can differentiate into the M cell type in vitro after lymphocytic stimulation (K. Miyazawa, T. Hondo, T. Kanaya, S. Tanaka, I. Takakura, W. Itani, M. T. Rose, H. Kitazawa, T. Yamaguchi, and H. Aso, Histochem. Cell Biol. 133:125-134, 2010). In this study, we evaluated the role of M cells in the intestinal invasion of the murine-adapted BSE (mBSE) agent using our in vitro bovine intestinal epithelial model. We demonstrate here that M cell-differentiated BIE cells are able to transport the mBSE agent without inactivation at least 30-fold more efficiently than undifferentiated BIE cells in our in vitro model. As M cells in the follicle-associated epithelium are known to have a high ability to transport a variety of macromolecules, viruses, and bacteria from gut lumen to mucosal immune cells, our results indicate the possibility that bovine M cells are able to deliver agents of TSE, not just the mBSE agent.
Transmissible spongiform encephalopathies (TSE) or prion diseases, including human Creutzfeldt-Jakob disease (CJD) and endemic sheep scrapie, are fatal neurodegenerative diseases. The host cellular prion protein (PrPC), which is thought to have neuroprotective function, is expressed in both humans and a range of other animal species (36), and PrPC expression is essential for TSE disease susceptibility (7). The prion hypothesis suggests that infectious abnormally folded prion protein (PrPSc) is the primary or sole composition of the infectious agent of TSE (known as the prion). However, the molecular composition of PrPSc remains speculative and unclear. It is well known that the detergent-insoluble and relatively proteinase K (PK)-resistant prion protein (PrP-res) is detectable in many kinds of TSE-infected tissues, including the brain. Although some studies have revealed that PrP-res does not correlate with infectivity levels in animal tissues as well as in subcellular fractions (37, 40), PrP-res is a useful surrogate marker for TSE infection.
Bovine spongiform encephalopathy (BSE) is a TSE of cattle. The first case of BSE in the world was found in the United Kingdom in 1986 (41), and it spread to continental Europe, North America, and Japan. At present, BSE is a threat to human health because of the appearance of BSE-linked variant Creutzfeldt-Jakob disease (vCJD). The cattle BSE agent appears to spread to the cattle population through the consumption of rendered meat and bone meal contaminated with BSE-infected brain or spinal cord (32). Likewise, the transmission of vCJD to humans is likely to have occurred following the consumption of BSE-contaminated food (6, 13, 45). In cases of oral transmissions such as BSE and vCJD, TSE agents first have to cross the gut epithelium, but the exact mechanisms for intestinal invasion still are unknown.
Intestinal epithelial cells are bound to each other by tight junctions. This close-packed structure forms a highly selective barrier for macromolecules and limits the access of pathogenic bacteria to the underlying host tissues (43). Gut epithelia are composed of two different epithelial types. One is the villous epithelium, and the other is the follicle-associated epithelium (FAE), which overlies gut-associated lymphoid tissues (GALTs) such as Peyer's patches. The FAE is considerably different from the surrounding villous epithelium, in that it contains membranous (M) cells. Because M cells have a high capacity for the transcytosis of a wide range of macromolecules, viruses, and microorganisms, they are specialized epithelial cells and act as an antigen sampling system from the gut lumen (28). M cells are, however, exploited by some pathogenic microorganisms and viruses as the entry site to invade the body (20, 29). In fact, some experiments have proposed that M cells transport TSE agents (12) and that Peyer's patches including the FAE are associated with TSE disease susceptibility (35). In contrast, some authors have suggested the M cell-independent pathway as the main transport route of TSE agents across the intestinal epithelium (16, 23, 27). The intestinal cell types involved in the transport of TSE agents therefore are still a matter of controversy at this stage.
Recently, we succeeded in the establishment of a bovine intestinal epithelial cell line (BIE cells) and the development of an in vitro bovine M cell model by coculture with murine intestinal lymphocytes or the supernatant of bovine peripheral blood mononuclear cells (PBMC) stimulated by interleukin 2 (IL-2) (25). In this study, we investigate whether M cells can transport the murine-adapted BSE (mBSE) agent using BIE cells. We demonstrate here that M cell-differentiated BIE cells are able to deliver mBSE agents at least 30-fold more efficiently than undifferentiated BIE cells, although a small number of the mBSE agents pass through undifferentiated BIE cells. Our findings thus provide an insight into the uptake mechanisms of TSE agents, including the cattle BSE agent from the gut lumen.
Yellow-green fluorescent microspheres (1.0 μm diameter) were purchased from Molecular Probes (Carlsbad, CA). A commercial goat antibody against C-terminal PrP peptides (M20) and horseradish peroxidase (HRP)-conjugated donkey anti-goat IgG antibody were purchased from Santa Cruz Biotech (Santa Cruz, CA).
ddY and C57BL/6 mice (SLC Japan, Shizuoka, Japan) for these experiments were fed under specific-pathogen-free conditions and used for the bioassay to evaluate the infectious titer. Experiments involving mouse-adapted BSE agent inoculation were performed in a biohazard prevention area (P3). The experiments were permitted by the Animal Care and Use Committee of Nagasaki University and the Tohoku University Environmental & Safety Committee and were conducted in accordance with the Guidelines for Animals Experimentation of Nagasaki University and Tohoku University.
We used murine-adapted BSE (mBSE) in this study. mBSE-infected brain was taken from ddY mice at the terminal stage of the disease after the third passage of original BSE by intracerebral inoculation (IC), and then the brains were frozen at −80°C. Pooled brain was homogenized to 10% (wt/vol) in cold phosphate-buffered saline (PBS). These 10% mBSE brain homogenates (mBSE-BH) were kept at −80°C until use.
The bovine intestinal epithelial cell line (BIE cells) was maintained in Dulbecco's modified Eagle medium (DMEM; GIBCO, Grand Island, NY) containing 10% heat-inactivated fetal bovine serum (FBS) and penicillin-streptomycin until it was differentiated into the M cell type and used for further studies.
To induce M cell differentiation, we seeded BIE cells onto each well of 12-transwell inserts (9 × 105 cells/well, 3.0-μm-pore-size snapwell filter polycarbonate membrane; Corning, NY) for coculture with murine small-intestinal lymphocytes (cocultures) or monoculture without lymphocytes (monocultures). After 3 days of incubation each well was inverted, and a silicon tube was attached to the bottom of each transwell insert. Following this, isolated small-intestinal lymphocytes (1 × 107 cells) were added inside the attached silicon tube of each transwell insert in a volume of 0.5 ml. In the case of the monocultured BIE cells without small-intestinal lymphocytes, 0.5 ml of 10% FBS-DMEM was placed in inside the attached silicon tubes. After 24 h of incubation, the silicon tube in each transwell insert was removed, and all of the transwell inserts were placed in 12-well plates (19, 25). To evaluate the transcytosis activity of monocultures and cocultures, 1.35 × 108 yellow-green fluorescent microspheres were added to the apical surface of monocultures and cocultures (four wells each). After 9 h of incubation, the basal medium of the monocultures and cocultures were collected, and the transported microspheres were quantified using a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA).
BIE cells were cultured on transwell inserts for 3 days and then left untreated or treated with the IL-2-stimulated PBMC culture supernatant for another 3 days. At 9 h after the addition of fluorescein isothiocyanate (FITC)-conjugated microspheres, BIE cells on transwell inserts were washed with cold PBS, fixed in methanol-acetone for 7 min at −20°C, counterstained with propidium iodide for 5 min, and then examined by a confocal laser microscope (MRC-1024; Bio-Rad, Richmond, CA). BIE cells were washed with 0.1 M phosphate buffer (PB, pH 7.4) once and then fixed with 2.5% glutaraldehyde in PB for 1 h. Following three washes with PB, the cells were dehydrated by passage through graded dilutions of ethanol and substituted with t-butylalcohol. Cells were freeze-dried, coated with platinum-palladium, and observed by scanning electron microscopy (SEM) (S4200; Hitachi, Tokyo, Japan).
We also applied 1% mBSE-BH, which was diluted with 10% FBS-DMEM from 10% BSE-BH, to the apical surface of monocultures and cocultures (eight wells each) and maintained both cell cultures for 24 h at 37°C in parallel. After 24 h of incubation, basal media (1.5 ml for 8 wells = 12 ml) were collected from each cell culture and stored at −80°C until use. We performed each experiment three times.
The infectivity titers of basal medium from the monocultures and cocultures were evaluated by mouse endpoint assay with ddY mice. We used 100-fold-concentrated basal medium of both cell cultures as the inoculum in the mouse bioassay. Briefly, we centrifuged 11 ml of basal medium at 20,000 × g for 1 h at 4°C. After centrifugation, we aspirated the supernatant and resuspended the pellet in 110 μl of serum-free medium and sonicated this 10 times for 10 s. Twenty μl of each concentrated basal medium was inoculated into the left side of a mouse brain. The inoculated mice were observed until 400 days after inoculation. The dose of mouse brain causing 50% mortality (LD50) was determined using the Reed and Muench formula. The infectious titers of 1% mBSE-BH and concentrated basal medium from monocultures and cocultures were calculated according to the relationship y = 13.023 − 0.0482x, where y is the log LD50 and x is the incubation period (time interval from inoculation to death) based on the standard incubation times of serial dilutions of mBSE-BH (see Fig. S1 in the supplemental material). All infectious titers are shown as LD50/ml.
As the mouse neuronal cell line GT1-7 shows a high susceptibility to various TSE strains (2, 3, 30) and accumulates the detectable PrP-res after several passages, we also used GT1-7 for the simple evaluation of the infectivity of basal medium of both monocultures and cocultures. GT1-7 cells were grown on six-well plates at 2 × 105 cells/well. After 2 days of incubation, cells were washed with medium and incubated with 1 ml of the basal medium collected from monocultures and cocultures for 24 h in parallel. An additional 1 ml of plain medium was added to each well after 24 h of incubation. Cells were incubated for 1 more day and seeded into a 25-cm2 flask for the first passage (P1). For passage, cells were seeded at 1:3 and maintained until P9. We collected cells at P7, P8, and P9 for the detection of PrP-res.
Confluent cells were washed with PBS and then lysed with NP-40-DOC lysis buffer (50 mM Tris-HCl [pH 7.5] containing 150 mM NaCl, 0.5% NP-40, 0.5% sodium deoxycholate [DOC], and 10 mM EDTA). After 1 min of 10,000 × g centrifugation at 4°C, the supernatant was collected, and its total protein concentration was measured by the bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Waltham, MA). The protein concentration of each cell lysate was adjusted to 2 mg/ml. To detect PrP-res, the samples were digested with 20 μg/ml of proteinase K (PK) at 37°C for 30 min. The samples were centrifuged at 20,000 × g for 1 h at 4°C. Following this, the pellet was resuspended in sodium dodecyl sulfate (SDS) loading buffer (50 mM Tris-HCl [pH 6.8] containing 5% glycerol, 1.6% SDS, and 100 mM dithiothreitol) and loaded onto a 12% polyacrylamide gel after being boiled for 5 min. The proteins were transferred onto an Immobilon-P membrane (Millipore, Billerica, MA) in transfer buffer containing 20% methanol at 300 mA for 70 min, and then the membrane was blocked in 5% nonfat dry milk in TBST (100 mM Tris-HCl [pH 7.8], 100 mM NaCl, and 0.1% Tween 20) for 1 h at room temperature. After being blocked, the blot was incubated with goat anti-PrP antibodies (M20) for 1 h at room temperature. Following three TBST washing cycles, the blot was incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. After three TBST washing cycles, immunoreactive bands were visualized by an enhanced chemiluminescence system (ECL plus; Amersham Pharmacia Biotech, Piscataway, NJ). The equivalent of 50 μg of total protein in SDS loading buffer was subjected to 12% polyacrylamide gel electrophoresis for the detection of total PrP.
Initially, we successfully induced BIE cells to differentiate into the M cell type, monitored by determining the active transportation of microspheres into the basal compartment. This confirms our earlier findings (19, 25). We detected a large number of microspheres in the basal compartment of cocultures with intestinal lymphocytes after 9 h of incubation. On the other hand, we detected few microspheres in the basal medium of the monocultures. Cocultures containing M cell-differentiated BIE cells had 170-fold transcytosis ability compared to that of the monocultures (Fig. (Fig.1A).1A). The uptake and internalization of microspheres in BIE cells was observed using a confocal laser microscope and scanning electron microscope (SEM). Many attached and incorporated microspheres were observed in cocultured BIE cells (Fig. (Fig.1C).1C). The x-z section detected microspheres around and under the nuclei of BIE cells in cocultures (Fig. (Fig.1E),1E), and it indicated the possibility that microspheres were incorporated into the cytoplasm of cocultured BIE cells. This level of incorporation also was observed in BIE cells cocultured with intestinal lymphocytes. In contrast, only a few microspheres were observed in monocultured BIE cells (Fig. 1B and D). SEM analysis revealed that BIE cells monocultured for 3 days developed short and irregular microvilli-like structures (Fig. (Fig.1F).1F). After M cell differentiation, the microvilli-like structures disappeared from the surface of the BIE cells, which put on a smooth appearance coincident with an enhanced ability to incorporate the microspheres (arrows in Fig. Fig.1G).1G). These data indicate that BIE cells are able to differentiate into M cells following stimulation, undergoing ultrastructural changes on the cell surface similar to those seen in vivo in M cells.
We next investigated whether M cells were able to transcytose the BSE agents. After the induction of M cell differentiation, 1% mBSE-BH was added to the apical surface of undifferentiated (monocultures) and M cell-differentiated BIE cells (cocultures). Because GT1-7 cells are able to replicate a variety of murine-adapted TSEs and to accumulate PrP-res as a surrogate marker for infection (31), we collected the basal medium of the mono- and cocultures and then evaluated their infectivity using GT1-7 cells. GT1-7 cells exposed to the cocultured basal medium produced PrP-res (Fig. (Fig.2A,2A, lane GT/co-supe). In contrast, GT1-7 cells exposed to the monocultured basal medium failed to produce PrP-res (Fig. (Fig.2A,2A, lane GT/Mono-supe). We tried to detect PrP-res in GT1-7 cells exposed to each basal medium between P7 and P9. We were able to detect PrP-res in GT1-7 cells exposed to the cocultured basal medium at every passage number, and the accumulation of PrP-res in GT1-7 cells tended to be higher as the passage number increased (Fig. (Fig.2B,2B, lanes GT/Co-supe). Nevertheless, no PrP-res bands were detectable between P7 and P9 in GT1-7 cells exposed to the monocultured basal medium (Fig. (Fig.2B,2B, lanes GT/Mono-supe). These data indicate that M cell-differentiated BIE cells have the ability to transport the mBSE agent.
To evaluate the infectivity of both cultures more accurately, we carried out a mouse endpoint assay. As shown in Fig. Fig.3,3, there were quite substantial differences in survival time between the mice inoculated with the cocultured basal medium and the monocultured basal medium. The mean number of days survived by mice inoculated with the cocultured basal medium was greater than that of mice inoculated with 1% mBSE-BH. Nevertheless, mice inoculated with the monocultured basal medium survived longer than mice inoculated with the cocultured basal medium by 40 to 70 days. There was a significant difference in survival time between mice inoculated with monocultured and cocultured basal medium (P < 0.001). However, all mice inoculated with monocultured basal medium developed clinical signs and died.
As shown in Table Table1,1, we calculated the infectious titer (LD50/ml) of each experimental group based on the standard incubation times of serial dilutions of mBSE-BH (see Fig. S1 in the supplemental material). According to this formula, the maximum infectious titer of monocultured basal medium samples was estimated to be 101.8 LD50/ml. On the other hand, the minimum infectious titer of cocultured basal medium samples was estimated to be 103.3 LD50/ml. Therefore, cocultured basal medium is at least 30-fold more infectious than monocultured basal medium. These results imply that M cell-differentiated BIE cells transport mBSE agent to the basal compartment more efficiently than undifferentiated BIE cells. However, undifferentiated BIE cells also can transport the mBSE agent. Our results indicate the possibility, therefore, that both enterocytes and M cells transport TSE agents from gut lumen in cattle.
For oral transmission, the TSE agents first must cross the intestinal epithelium to successfully infect the body. However, the intestinal cell types involved and the mechanisms associated with the uptake of these agents remain unclear. We report here that bovine intestinal epitheliocytes (BIE cells) cocultured with murine intestinal lymphocytes (cocultures; M cell-differentiated BIE cells) demonstrated a high capacity for the transcytosis of the mBSE agent compared to that of BIE cells without lymphocytes (monocultures; undifferentiated BIE cells). It is conceivable, however, that the mBSE agent has some properties that are different from those of the original cattle BSE agent. This is because of the initial passage from primary host to the mouse. Additionally, in ruminants, orally acquired TSE agents, including the BSE agent, are recycled between the oral cavity and the rumen through the process of rumination. Following this, the TSE agents will pass through the omasum, reticulum, and abomasum. Finally, TSE agents reach the intestine and are incorporated into differentiated enterocytes and M cells. During these alimentary steps, many complicated factors, including the bacterial fermentation and digestive enzymes, interact with the TSE agents and degrade most PrP-res (16). As these factors might affect the conformational structure and the infectivity of TSE agents, the agents under such circumstances might be different from those of the brain homogenate used in our in vitro model.
M cells have been reported as a candidate for the mucosal gate for the scrapie agent following experiments using the human colon adenocarcinoma Caco-2 cell culture model (12). These findings support our results that bovine M cells are likely to sample and deliver TSE agents to the basolateral side of the gut epithelium. Prinz et al. demonstrated that susceptibility to TSE infection following an oral challenge correlates with the number of Peyer's patches (PPs). They have estimated that the number of M cells in B cell-deficient (PP reduced and atrophied) mice is approximately 150 to 840 times less than that in wild-type mice and have shown that B cell-deficient mice are resistant to scrapie infection (35). In some rodent and sheep experiments, immunolabeling for PrP-res is observed in epithelial cells of the FAE (4, 10). These findings also suggest that the transcytosis of TSE agents occurs via the FAE-containing M cells. M cells in vivo display a unique pocket-like structure (24). The pocket of M cells contains B and T lymphocytes along with a small number of macrophages (18), and the dome region underlying the FAE is also rich in dendritic cells (DCs), macrophages, and lymphocytes (29). We previously reported that PrPC was expressed in myeloid lineage cells such as myeloid DCs and macrophages in bovine Peyer's patches, and that some of these PrPC-expressing immune cells were located in the dome region just under the FAE (26). Current reports indicate that DCs and macrophages acquire TSE agents during the process of infection. Further, it is thought that macrophages degrade PrP-res and clear infectivity (5, 22, 38). Unlike macrophages, however, some DCs can capture and retain TSE agents in the infectious form (15). M cells and PrPC-expressing DCs in bovine Peyer's patches seem to be involved in the TSE agent transportation. Numerous M cells are also located in the ruminant nasopharyngeal region (42), and the experimental transmission of the scrapie agent has been reported via a nasal route in sheep (9). Indeed, there is the possibility that nasal M cells take up TSE agents in cattle in vivo. However, the accumulation of PrP-res has not been reported in lymphoid tissues around the head and neck in BSE-infected cattle. On the other hand, infectivity has been detectable in retinal, optic, and facial nerves of naturally BSE-infected cows (8). An oral BSE challenge and sequential kill study of cattle suggests a neuronal rather than a lymphoreticular progression of BSE agent to the brain (14). This fact may explain the absence of a detectable level of PrP-res accumulation in nasal associated lymphoid tissues containing M cells in BSE-infected cattle. The neuroinvasion of BSE agents actually occurs in the absence of detectable PrP-res accumulations in lymphoreticular tissues in some BSE-infected sheep (17). It is conceivable, therefore that BSE-infected cattle, BSE-infected sheep, and scrapie-infected sheep possess common mechanisms of neuroinvasion that do not involve the lymphoreticular system. In contrast to cattle infected with BSE, however, most BSE-infected sheep show PrP-res accumulations in the lymphoreticular tissues as well as scrapie-infected sheep (11, 17). The infectivity also has been detectable within the entire lymphoreticular system of BSE- or scrapie-infected sheep (1, 34, 44). Thus, the magnitude of lymphoid tissue involvement can vary considerably depending on the host and the TSE agent, and the accumulation of PrP-res in lymphoid tissues originally is limited in BSE-infected cattle (21). It is reasonable to consider that the lymphoreticular system in sheep is positively involved in the replication and the neuroinvasion of TSE agents, in contrast to cattle.
Although we failed to estimate the infectious titers in two of the three monocultured cell experiments because one of their incubation periods was very close to the detection limit and the other was out of the detection range, we nevertheless observed that undifferentiated BIE cells transported the mBSE agent. The other intestinal cell types, such as absorptive epithelial cells in villous epithelium, might participate in the transcytosis of TSE agents as well as M cells in a bovine in vivo situation. One group has reported that the treatment of sporadic CJD brain homogenate with digestive enzymes generates a C-terminal fragment of PrP similar to the PK-resistant core of PrPSc, and that these fragments can form complexes with ferritin and are transported by Caco-2 cells (23). Another group has reported that PrPSc from BSE brain homogenate is transported transepithelially by Caco-2 cells via 37-kDa/67-kDa laminin receptors (27). These in vitro studies support the M cell-independent pathways of the TSE agent. Caco-2 cells have some properties of intestinal absorptive cells. Undifferentiated BIE cells still retain characteristics of relatively undifferentiated intestinal epithelial cells, because they do not possess packed microvilli and alkaline phosphatase activity (25). There might be some differences in the differentiation stage between Caco-2 cells and BIE cells. Less transepithelial transport of mBSE agent by undifferentiated BIE cells might be due to the relatively undifferentiated status of this cell line. Since it is still unclear what titer of infectivity was transported across the basolateral membrane of Coco-2 cells, further studies are required to confirm the transport ratio of TSE infectivity by absorptive epithelial cells.
There are two additional in vivo studies on the transportation of TSE agents across the absorptive epithelium in mice (33) and sheep (16) by oral exposure with sheep scrapie brain and an intestinal loop assay. These authors analyzed the immunoreactivity for PrP-res or disease-associated forms of PrP (PrPd), which mean tissue forms of protease-sensitive and protease-resistant PrPSc detected by immunohistochemistry. Although alimentary fluid degraded most PrP-res and PrPd, some PrPd deposits were observed in absorptive epithelial cells by immunohistochemistry. We consider that these PrPd deposits might represent the uptake of digested smaller particle size of PrPSc revealed by the study using flow field-flow fractionation (39). Delivering a small number of mBSE agents by undifferentiated BIE cells might reflect the transcytosis of smaller PrPSc particles. The relationship between the infectivity and the size of various PrPSc aggregations has been analyzed and revealed that infectivity peaks at a particle size of 17 to 27 nm (3 × 105 to 6 × 105 Da) (39) or 25 to 30 nm (106 to 107 Da) (40). With respect to transcytosis, this size of the particle is not too large for M cell-differentiated BIE cells, but it might be too large for undifferentiated BIE cells. As smaller particles of PrPSc aggregates still show considerable infectivity, the transcytosis of different particle sizes of PrPSc aggregations could explain the difference in transported infectivity between M cell-differentiated and undifferentiated BIE cells. It is possible that the degraded smaller particle of PrPSc is a major component of TSE agents in the in vivo situation, because agents must transit through the gastrointestinal tract with many kinds of digestive enzymes, unlike the mBSE-BH used in our study. Indeed, we showed that M cell-differentiated BIE cells transported TSE agent efficiently in our model, but we do not rule out the possibility that enterocytes play a more important role in the uptake of TSE agents in living cattle.
In conclusion, results obtained here demonstrate that M cell-differentiated BIE cells show at least 30-fold-higher transepithelial transport of the mBSE agent than undifferentiated BIE cells. Accordingly, though we do not exclude the importance of M cell-independent pathways in the in vivo situation, FAE containing M cells could act as a possible entry site of TSE agents, including the BSE agent into cattle following oral exposure. In addition, we suggest that undifferentiated BIE cells and the in vitro M cell model are useful for the analysis of mechanisms of the transcytosis of TSE agents by bovine intestinal epithelial cells.
This research was supported by a Grant-in-Aid for Scientific Research (21380170) from the Ministry of Education, Culture, Sports, Science and Technology, BSE Control Project from the Ministry of Agriculture, Forestry and Fisheries, and Postdoctoral Fellowships for Research Abroad from Japan Society for the Promotion of Science (JSPS).
Published ahead of print on 22 September 2010.
†Supplemental material for this article may be found at http://jvi.asm.org/.
#The authors have paid a fee to allow immediate free access to this article.