In vitro cultivation of shrimp Taura syndrome virus (TSV) in a C6/36 mosquito cell line

Taura syndrome virus (TSV) is a single-stranded RNA (ssRNA) virus of positive sense that has been classified in the family Dicistroviridae near the genus Cripavirus (Christian, Carstens, Domier, Johnson, Nakashima, Scotti & van der Wilk 2005). It is the causative agent of Taura syndrome, a disease that was first described in Eduador in 1992 as the cause of high mortality in cultivated Penaeus (Litopenaeus) vannamei (Hasson, Lightner, Poulos, Redman, White, Brock & Bonami 1995). Since its discovery, much work has been performed on its characterization, detection and control (Brock, Gose, Lightner & Hasson 1997; Lightner 1999) and domesticated stocks of P. vannamei have been developed that are highly tolerant to it (Moss, Doyle & Lightner 2005). Laboratory challenge studies with TSV and other shrimp viruses are usually dependent on production of viral inoculum from infected shrimp, because no immortal cell lines for shrimp or other crustaceans have ever been developed (Crane & Benzie 2000). As an alternative to an immortal crustacean cell line, Sudhakaran, Parameswaran & Sahul Hameed (2007) successfully challenged C6/36 cells with Macrobrachium rosenbergii nodavirus (MrNV) and extra small virus (XSV). They also demonstrated that C6/36 cells could be used to isolate MrNV and XSV from RT-PCR positive insect carriers (Sudhakaran, Haribabu, Rajesh Kumar, Sarathi, Ishaq Ahmed, Sarath Babu, Venkatesan & Sahul Hameed 2008). In addition, it was shown that C6/36 mosquito cells challenged with white spot syndrome virus (WSSV) and yellow head virus (YHV) of shrimp followed by serial split-passage of whole cells could lead to insect cell cultures in which 100% of the cells persistently expressed shrimp viral antigens (Sriton, Khantong, Gangnonngiw, Sriurairatana, Ubol & Flegel 2009). The infected insect cells did not produce free shrimp virus particles in the surpernatant culture solution, and it was proposed that viral genomic material was maintained in the cultures by replication and passage from mother to daughter cells during cell cultivation. Because the insect cells remained 100% antigen positive for more than 100 passages, it was reasoned that viral genomic material must have replicated to support antigen production. In a subsequent publication (Gangnonngiw, Kanthong & Flegel 2010), it was shown that C6/36 cells persistently expressing YHV antigens could serve as inoculum to infect shrimp with YHV. Insect cells from passage 5 caused shrimp mortality, while those from passage 15 did not, even though the challenged shrimp were positive for YHV by immunohistochemical analysis. Here, we describe the successful production of C6/36 cells that persistently expressed TSV antigens as previously described for WSSV and YHV. We also report positive RT-PCR results for TSV in the immunopositive insect cells and the successful infection of Penaeus vannamei with homogenates of TSV-immunopositive insect cells from passage 15. For persistent infections with TSV, C6/36 mosquito cells (a single cell-type clone obtained from the American Type Culture Collection, catalogue number CRL-1660) were maintained in supplemented Leibovitz's (L-15) medium as previously described (Sriton et al. 2009). A TSV viral stock solution was prepared from infected Penaeus (Litopenaeus) vannamei haemolymph according to Boonyaratpalin et al. (1993). The TSV-infected shrimp were obtained from a farm outbreak in Samutsakorn Province, Thailand. The TSV stock was diluted to 1:100 with the supplemented L-15 medium, and 2 mL was used to challenge C6/36 monolayers in a six-well culture plate at 28 °C, after the existing culture medium had been removed. The cells were exposed for 2 h after which the inoculum solution was removed and replaced with 2 mL fresh L-15 medium. After 6 days, the challenged culture was examined for evidence of cytopathic effects (Sriton et al. 2009) and for evidence of TSV antigens by immunohistochemistry. Thereafter, the culture supernatant was removed and 2 mL fresh L-15 medium was added. The cells were suspended in the fresh medium by gentle knocking. Then, 1/3 of the suspension was transferred to a fresh culture well (i.e. 1/3 split) to which L-15 medium was added to make up a total of 2 mL. After incubation for 2 days (full confluence) the process of decanting, re-suspension, split and transfer were repeated continuously, as previously described (Sriton et al. 2009). Tests were performed in triplicate and naïve C6/36 cells treated in the same manner were used as controls. It is imperative to understand that the cells were being transferred, not the medium. Immunohistochemical analysis using confocal microscopy (Chaivisuthangkura, Tejangkura, Rukpratanporn, Longyant, Sithigorngul & Sithigorngul 2006) showed that challenged cultures of C6/36 mosquito cells became 100% positive for TSV within 10 passages (Fig. 1). Fluorescent intensity at passage 1 was somewhat less than that at later passages and the character of the fluorescence changed (Fig. 2), suggesting that the cells accommodated TSV in an adaptive manner. The cultures were continuously split-passaged for up to 22 passages at which time they still remained strongly immunopositive for TSV. Despite the intense fluorescence, even in late-passage cultures, morphology of the cells by phase-contrast microscopy never differed from that of naïve, unchallenged cells, as previously reported for WSSV and YHV in C6/36 cells (Sriton et al. 2009; Gangnonngiw et al. 2010). Without any propagation and transfer of the viral genomic material, the number of immunopositive cells at split-passage 10 would have been 1 in 310 (c. 1 in 1.4 × 107 cells), and by passage, 22 it would have been 1 in 322 (c. 1 in 3.1 × 1010 cells). Thus, the fact that the cells remained 100% immunopositive throughout meant that the viral nucleic acid must have replicated to support continuous antigen production, as previously proposed (Sriton et al. 2009), even though the nature of its maintenance in the passaged cells has not been clarified. Confocal microscope photographs of immunofluorescence test for Taura syndrome virus (TSV) in C6/36 at passages 1 and 9. (a) Cells from passage 1 (P1) after incubation for 2 days following split from the culture (P0) originally exposed to TSV. (b) Cells from passage 9 after incubation for 2 days. Note that the fluorescence is generally more intense and somewhat different in texture than that at passage 1. (c) Negative control cells not exposed to TSV. Positive fluorescence for anti-VP1 of TSV is red, and the blue signal shows To-Pro-3 iodide staining of DNA in nuclei. Transmission electron micrographs of C6/36 cells persistently infected with Taura syndrome virus (TSV). (a) Low-magnification micrograph showing cells with unusual cytoplasmic inclusions (arrow). (b) High-magnification micrograph of the area marked by arrow in (a) showing unusual structures that resemble virogenic material containing virus-like particles similar to those reported for TSV in shrimp. Taura syndrome virus-infected cells from passage 19 were fixed in situ, embedded, sectioned for electron microscopy as previously described (Chien, Hsu, Sheng, Jung, Shu, Nan, Tien, Ming & Shih 1999) and stained with 2% uranyl acetate plus lead citrate for viewing with a H7100 Hitachi electron microscope at 100 kV. Electron micrographs revealed areas with viral-like particles together with vesicular structures in the cytoplasm of the immunopositive C6/36 cells. Such structures are not seen in normal C6/36 cells (Sriton et al. 2009). The viral-like particles resembled those seen in virogenic stomata of shrimp cells infected with TSV (Chien et al. 1999) (Fig. 2). With respect to abnormal vesicular structures, our results resembled those previously reported for YHV and WSSV infections in C6/36 cells by Sriton et al. (2009). However, they differed in that no recognizable structures of YHV or WSSV viral particles were seen in immunopositive C6/36 cells in the previous report (Sriton et al. 2009). Despite the presence of the viral-like structures in the cells immunopositive for TSV, the supernatant solution of the culture medium from such cells could not be used to successfully infect naïve C6/36 cells. This result was in accord with the previous report by Sriton et al. (2009), indicating that insect cells infected with YHV and WSSV were unable to release free, infectious virions into the culture medium, even though viral genomic material replicated and was transmitted to daughter cells (probably during cell division). Specific detection of TSV by nested RT-PCR was achieved at passage 15 using an IQ 2000 test kit (Farming Intelligene) (Fig. 3). According to the kit manual, the result indicated a light infection level equivalent to not <20 and not more than 200 copies of the TSV target in the reaction vial. Cloning, sequencing and comparison of the nucleotide sequence of the 284-bp amplicon from insect cells revealed 99% identity with the corresponding TSV sequences at GenBank (Accession number: AY355311). In contrast, a 2 μL sample template from a 50 μL RNA extract solution prepared from 700 μL of culture supernatant solution following the protocol of TRI Reagent (Molecular Research Center Inc.) gave a negative result, suggesting that TSV was absent or not present in sufficient quantities to give a positive test kit result (i.e. <20 copies of TSV target in the 2 μL of template equivalent to 28 mL of culture supernatant used for RNA extraction). As described for the immunohistochemical assays earlier, without viral genome replication, the original TSV inoculum would have been diluted by 315 = 1.4 × 107 times by passage 15 yielding <1 cell in 14 million positive for TSV. Because the cells were 100% immunopositive for TSV, we must assume at least one TSV genome copy in each cell to account for viral antigen production. Because c. 106 cells were used for RNA extraction, we could expect (with perfect RNA extraction), 106 TSV copies in the total RNA extract of 50 μL (final volume) and 40 000 copies in the 2 μL template used for the RT-PCR reaction. This contrasts with the actual RT-PCR result indicating only 20–200 copies in the 2 μL template. We cannot account for this apparent discrepancy, but our conclusion is that the copy number of TSV per C6/36 cell was very low, despite the strong antigen reaction seen. The same phenomenon was seen with YHV in insect cells (Sriton et al. 2009; Gangnonngiw et al. 2010). Nested RT-PCR detection of Taura syndrome virus (TSV) in persistently infected C6/36 cells at passage 15. Lane M: molecular marker; lane 1: positive control; lane 2: negative control; lanes 3 and 4: amplicons from TSV-infected C6/36 5 × 106 cells. Two amplicon bands (284 and 680 bp) indicate light TSV infections according to the kit manual. The band at 680 bp is the kit internal control to verify the integrity of the shrimp RNA template and it appears only with light TSV infections or with uninfected shrimp. A preliminary test using injection of homogenates of 1, 2 or 3 million TSV-immunopositive insect cells from passage 3 into experimental American whiteleg shrimp Penaeus (Litopenaeus) vannamei revealed the presence of immunopositive haemocytes at 24 and 48 h post-injection, and a trend for increasing numbers of immunopositive non-granular cells (Fig. 4) with time post-injection. Because granular cells in penaeid shrimp are phagocytic and non-granular cells are not, the results suggested that TSV was replicating and spreading in the non-granular cells and that the fluorescence seen was not simply the result of phagocytosis of injected antigens. In a second test, P. vannamei was injected with concentrated culture supernatant solution from passage 15 but this did not result in shrimp mortality or in the detection of TSV in the challenged shrimp by either RT-PCR or immuhohistochemical analysis of haemocytes. However, injection of homogenates of 1 million TSV-immunopositive C6/36 cells from passage 15 led to a high percentage of haemocytes immunopositive for TSV using confocal microscopy at 8-days post-challenge (Fig. 5b). By contrast, no infected haemocytes were detected in shrimp challenged with supernatant derived from homogenates of naïve cells (Fig. 5a). The shrimp were positive for TSV by immunohistochemistry and were negative for TSV by RT-PCR, indicating that the infections were very light. The shrimp with haemocytes immunopositive for TSV showed no mortality and no histopathology typical of TSV infections, possibly because they were genetically selected for tolerance to TSV. As we did not have ready access to shrimp stocks highly susceptible to TSV, we could not assess the virulence of TSV from the cultured insect cells. Confocal microscope photomicrographs showing Taura syndrome virus (TSV) immunofluorescence in haemocytes of shrimp challenged in a preliminary test with a homomogenate of 3 million C6/36 cells immunopositive for TSV. (a) Haemocyte preparation at 24-h post-challenge with phase setting to identify granular cells and several non-granular cells. (b) Same field as in (a) but showing immunofluorescence for TSV antigen in one granular cell and several non-granular cells, most with light immunofluorescence. (c) Haemocyte preparation at 48-h post-challenge with phase setting showing absence of granular cells. (d) Same field as in (c) showing several immunopositive non-granular cells and proportionally more with strong immunofluorescence than at 24 h. Confocal microscope photomicrographs showing Taura syndrome virus (TSV) immunofluorescence in haemocytes of shrimp at 8-days post-challenge with a homogenate of 1 million C6/36 cells immunopositive for TSV. (a) Phase-contrast image showing the absence of granular cells in the preparation. (b) Positive red fluorescence for anti-VP1 of TSV in the cytoplasm of most of the haemocytes. The immunofluorescence is somewhat less intense and more diffuse than that shown for 48 h in Fig. 4. No immunofluorescence was seen in haemocytes of shrimp injected with homogenates of naïve C6/36 cells. Recent work with shrimp YHV cultivated in C6/36 cells and then used to challenge shrimp showed that shrimp challenged with cell homogenates from passage 5 died and showed typical YHV pathology, while those from passage 15 did not, even though they did become infected with YHV (Gangnonngiw et al. 2010). It was suggested that the virus had become attenuated after 15 passages in insect cells. Thus, it is possible that TSV from passage 15 in C6/36 cells may also have become attenuated and that cells from lower passages of C6/36 cells infected with TSV might cause Taura syndrome in susceptible lines of P. vannamei. In summary, this work revealed that TSV genomic material could replicate successfully and be perpetually transmitted in continuously passaged mosquito cells that showed no gross signs of infection. Although there was no evidence for the release of infectious viral particles into the culture supernatant solution, homogenates of whole cells from passage 15 were capable of producing very light TSV infections in whiteleg shrimp. Because the TSV-infected insect cells could be stored at −80 °C and revived at will, this method may be useful for preservation and cultivation of TSV and for in vitro studies on viral gene expression, host–viral interaction and TSV gene knockdown. This work was supported by the Office of the Higher Education Commission and Mahidol University under the National Research Universities Initiative. The authors would like to thank T.W. Flegel with assistance in conceiving and designing the study and with preparation of the manuscript.