Introduction

Mosquito cell lines offer unique pathogen susceptibilities making them advantageous for medical studies (Pudney et al. 1982; Mitsuhashi 1994). Notably, mosquitoes include many species known to be principal vectors that facilitate pathogenic viral transmission. Over the past several years, mosquito cell culture has significantly progressed with respect to virological research (Kuwata et al. 2012). However, the variety of mosquito species from which the available cell lines are derived is limited, depending upon the specific vector–virus association. C6/36 cells cloned from Singh’s line from the Aedes albopictus mosquito (Igarashi 1978) have been used for experiments with viruses, such as the Japanese encephalitis virus (JEV), which are mainly transmitted by the Culex mosquitoes. This limited variety of mosquito cell lines is partly attributed to the loss of cryopreserved lines. Some of these lost lines have since been recovered through the establishment of new lines, especially from Culex (Athawale et al. 2002; Segura et al. 2012; Kuwata et al. 2012). The mosquito Armigeres subalbatus (Coquillett) has been recognized as one of the synanthropic and medically important species in Asia (Tanaka et al. 1979; Muslim et al. 2013) and is also a physiological model species because of its distinct innate immune system (Sasaki et al. 2010; Aliota et al. 2010). Thus, an Armigeres cell line would play a significant role in several studies concerning the biology and virology of Armigeres mosquito because no cell line from this species has ever been reported (Mitsuhashi 2002) although an attempt to culture A. subalbatus was noted (Hink 1976).

In this study, we attempted to culture A. subalbatus mosquito cells to establish a cell line from the species. However, establishment of mosquito cell lines still depends on the spontaneous growth of primary cultured cells without an efficient derivation method such as the expression of oncogenes in Drosophila (Simcox et al. 2008). Since the establishment of earlier mosquito lines from the genus Aedes (Grace 1966; Singh 1967) using Grace’s (Grace 1962) and MM (Mitsuhashi and Maramorosch 1964) media, modification of the media components has facilitated the establishment of cell lines from species belonging to other genera such as Psorophora (Bello et al. 2001), Culex (Kitamura 1970), as well as Anopheles (Schneider 1969). Accordingly, we initiated mosquito cultures in media that have been successfully used for culturing various genera of mosquitoes because the genus Armigeres is taxonomically different from any genera known from which cell lines have been established (Tanaka 2005; Reinert et al. 2009). After establishing the mosquito cell lines, we performed basic characterization of their morphology, growth kinetics, karyotyping, and authentication, and further analyzed their viral susceptibility.

Materials and Methods

Culture media.

We prepared four types of media specifically developed for insect cell culture: MM (Mitsuhashi and Maramorosch 1964), VP-12 (adjusted to pH 6.9) (Varma and Pudney 1969), Schneider’s (Sigma-Aldrich Japan, Tokyo, Japan) (Schneider 1972), and MGM450 (adjusted to pH 6.3) (Mitsuhashi and Inoue 1988). The mosquito cells were cultured separately in each media and in mixtures of two types of media; thus, a total of ten types of media were tested to select the most suitable for the primary cultures. All the media were supplemented with 10% heat-inactivated (56°C, 30 min) fetal bovine serum (FBS), which did not contain any antibiotics.

Mosquito and primary culture.

A laboratory colony of A. subalbatus established by the US Army 406th Medical General Laboratory was maintained in a controlled insectary at 25°C in 50–70% relative humidity and a light/dark photoperiod of 16/8 h. The eggs deposited on the filter paper were surface-sterilized three times by immersion in 70% ethanol for 3 min and subsequently dried on autoclaved filter paper. The sterilized eggs were then transferred to a new 35-mm plastic dish (Falcon, Oxnard, CA) containing 2 mL of culture medium and allowed to hatch. The hatched neonate larvae were collected in 1.5 mL centrifuge tubes (Greiner Bio-One Co Ltd., Tokyo, Japan), placed on ice, and then washed three times with fresh medium without floating the larvae that settle at the bottom. The sterile larvae were mixed with a drop of medium (80 μL volume, ten larvae per drop) sitting on a new 35-mm surface-modified tissue culture dish (Primaria; Falcon). The larvae were then cut into pieces with a razor blade, and the head portions were removed to avoid their feeding on the other larval pieces. All cultures were maintained by placing a drop of sterile water at the margins of the dish to prevent desiccation of the culture medium, sealed with Parafilm (American National Can, Greenwich, CT), and kept at 25°C. Primary cultures were initiated under the same conditions in triplicate.

Establishment of cell lines.

One week after initiation of primary cultures, 1.5 mL of each medium was added to the samples after removal of the water drop. The medium (0.2–0.5 mL) was exchanged or added regularly. During this process, both migrations and divisions of the cells in each sample were observed by phase contrast microscopy (IX71, Olympus). When the cells reached confluency, they were suspended by pipetting and subcultured in 12.5-cm2 culture flasks (3018, Falcon). Thereafter, the cells were subcultured at intervals of approximately 2 wk, with a split ratio of 1:1 to 1:2, leading to the establishment of two cell lines designated as Ar-3 and Ar-13.

Characterization of the cell lines.

The morphology of the Ar-3 and Ar-13 cells was observed by phase contrast microscopy. For construction of growth curves, the initial cell density was set at 4.0 × 105 cells/mL, and the number of cells was counted in triplicate every 2 d using a Thoma’s hemocytometer. The growth curves of the Ar-3 and Ar-13 cell lines were analyzed at the 100th and 70th passages, respectively. Next, the karyotypes of the Ar-3 and Ar-13 cells were analyzed at the 100th and 88th passages, respectively, as previously described (Mitsuhashi 2002). Random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR) ( Mitsuhashi 2002) was then performed on the following samples: Ar-3 and Ar-13; adult A. subalbatus mosquitoes; the three mosquito cell lines C6/36 (Health Science Research Resources Bank, Osaka, Japan), NIAS-AeAl-2 (Mitsuhashi 1981) from A. albopictus, and NIID-CTR from Culex tritaeniorhynchus (Kuwata et al. 2012); the dipteran S2 cell line from Drosophila melanogaster (Schneider 1972); two lepidopteran cell lines, Sf21 from Spodoptera frugiperda (Vaughn, et al. 1977) and GaMe-LF1 from Galleria mellonella (Eguchi and Iwabuchi 2006); and the coleopteran cell line PC-1 from Plagionotus christophi (Hoshino et al. 2009a). DNA was extracted from the cell pellets or mosquito legs using the DNeasy blood and tissue kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. The eluted DNA solution was washed and concentrated using the ethanol precipitation method. Both the concentration and the purity of the samples were confirmed using Nanodrop (Thermo Fisher Scientific, Wilmington, DE). The ratio of absorbance at 260 and 280 nm (260/280) ranges from 1.61 to 1.86. The samples were stored at −80°C until further use. PCR was conducted using Quick Taq HS Dyemix (Toyobo Co. Ltd., Tokyo, Japan). Two 12-mer oligonucleotides (5′-CCGCAGTTAGAT-3′ and 5′-ACTGGCCGAGGG-3′) were used for RAPD-PCR analysis. PCR was performed in duplicate, according to the manufacturer’s instructions using 50 ng DNA in a 50-μL reaction volume under the following PCR conditions: 2 min at 94°C, followed by 45 cycles of 30 s at 94°C, 30 s at 42°C, and 2 min at 68°C. The amplicons were electrophoresed in 1.5% agarose gel and stained with ethidium bromide, followed by comparative analysis of the images captured under UV light.

Viral susceptibility.

Ar-3 cultured in VP12 medium was selected for large-scale culture based on the medium cost. The following four inoculated flaviviruses were selected: dengue virus (DENV); JEV, isolated from wild-caught A. subalbatus (Weng et al. 1999; Chen et al. 2000; Feng et al. 2012; Liu et al. 2013); and Culex flavivirus (CXFV) and Aedes flavivirus (AEFV), insect flaviviruses associated with the Culex and Aedes mosquitoes, respectively (Hoshino et al. 2007, 2009b), but never isolated from the Armigeres species. To test the viral susceptibility of Ar-3 cells, the following four flaviviral strains were prepared from our virus stocks: JEV strain Mie/41/2002 (Nerome et al. 2007; GenBank accession no. AB241119), DENV strain NIID02-20 (Tajima et al. 2006; GenBank accession no. AB178040), CXFV strain NIID-21 (Hoshino et al. 2007; GenBank accession no. AB377213), and AEFV strain Narita-21 (Hoshino et al. 2009b; GenBank accession no. AB488408). The copy number of insect flaviviral genomic RNA in the stock solution was determined by quantitative reverse-transcription PCR (RT-PCR) using primers specific for CXFV (Isawa et al. 2012) and AEFV (sequence position 156–363 in strain Narita-21). To prepare the viral RNA standards, in vitro-transcribed RNAs from CXFV and AEFV cDNA were generated according to a previously described method (Isawa et al. 2012). The resultant transcripts were purified using the RNeasy MinElute Cleanup kit (QIAGEN), quantified spectrophotometrically using Nanodrop (Thermo Fisher Scientific) to determine the copy number of the RNA standard, and then used as a template for standard curve analysis. Each viral stock solution in VP12 medium (1 mL) containing 2% FBS was inoculated onto 80% confluent cell monolayers of Ar-3 cells seeded in 25-cm2 flasks (3014, Falcon) at 102–103 plaque-forming units (PFU)/mL for JEV and DENV (titered with standard plaque assay method described by Tajima et al. 2010) or 102–103 RNA copies/mL for CXFV and AEFV. The inoculum was removed after 2 h of viral adsorption. Finally, the flasks were filled with 5 mL fresh medium containing 2% FBS after washing three times with Dulbecco’s phosphate-buffered saline (D8662, Sigma-Aldrich Japan). The culture medium was sampled on days 0, 1, 2, 3, 5, and 7 after the viral inoculation. RNA was extracted from the samples using QIAamp Viral RNA Mini Kit (QIAGEN) according to the manufacturer’s instructions. Viral detection was conducted by RT-PCR using PrimeScript One Step RT-PCR Kit Ver.2 (Takara Bio., Shiga, Japan). PCR was performed according to the manufacturer’s instructions under the following conditions: 30 min at 50°C for reverse transcription, 2 min at 94°C, followed by 30 cycles of 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C. The primers used were specific for each viral envelope region as follows: JEV-E31F, 5′- TTC ATA GAA GGA GCC AGT GGA GC -3′ and JEV-E287R, 5′- AAG CCT TGT TTG CAC AC -3′ for JEV; DEN-F, 5′-AAG GAC TAG AGG TTA KAG GAG ACC C-3′ and DEN-R, 5′-GGC GYT CTG TGC CTG GAW TGA TG-3′ for DENV (Callahan et al. 2001); CXFV-E417F, 5′- CTC AGT GGA TGA CGT CCA GCA ACT C -3′ and CXFV-E963R, 5′- TCC AGG TCC GAA CAC CAT CTT CGT C -3′ for CXFV; and AEFV-E156F, 5′- AAC GTT GCT AAC ATC ATG TGA GAT -3′ and AEFV-E363R, 5′- ATT TAC CAT GGT CAG GCA TTG GA -3′ for AEFV. The amplified products were separated on 2% agarose gels and visualized with ethidium bromide. For the infective virus quantification of JEV and DENV, small aliquots of the medium were recovered on days 1, 3, and 7 after the viral inoculation and titrated by a standard plaque assay on Vero cells described by Tajima et al. 2010. Briefly, Vero cells were fixed with 3.7% (v/v) formaldehyde solution in PBS for 1 h, the methylcellulose overlay was removed, and the cells were stained with methylene blue solution.

Results

Primary culture and establishment of cell lines.

Cell migration was significantly induced in the primary cultures when VP12, MM + MGM450, and Schneider’s media were used (Fig. 1A ). Three independent cultures grown in each of these three media were found to show enough growth for subculturing within 2–3 mo and were then transferred to new flasks. During subculturing, cell growth continued with frequent appearance of vacuolated cells in the earlier phase (Fig. 1B, C ). However, as the passage number increased, the cells became uniformly adherent (Fig. 1D ). Finally, two cell lines grown in VP12 and MM + MGM450 media and stably maintained without any crisis period were selected and designated as Ar-3 and Ar-13, respectively. These lines have now been subcultured for more than 200 passages over a period of 7 yr. Both cell lines formed many compact colonies with high densities, even if dissociated during passaging (Fig. 2A, B ). Unfortunately, all the cell lines grown in Schneider’s medium were lost during subculturing due to severe crystallization of the medium. Among the other media samples, some cell migration was observed in the lines grown in VP12 + MM [Tra medium in Kuno (1980)], VP12 + MGM450, and VP12 + Schneider’s media, while MM medium promoted the formation of several hollow spherical vesicles. However, none of these media enabled continuous cell growth.

Figure 1
figure 1

Phase contrast micrographs of A. subalbatus cells in culture. A Representative image of cell migration from an explant (ex) of a neonate larval piece. B Surviving and vacuolated cells (arrows) observed at 3 mo after the initiation of primary culture. Lines on the surface of the culture dish seen in A and B are scars made by razors used to cut the neonate larvae. Cells growing after the seventh and eighth passages and vacuolated cells (arrows) in the C Ar-3 and D Ar-13 cell lines. Bars indicate 50 μm.

Full size image
Figure 2
figure 2

Cellular morphology of the established A. subalbatus cell lines. Both A Ar-3 and B Ar-13 cell lines contain adherent cells (appearing black due to phase contrast microscopy) and compact colonies (cc) with a patch-like distribution. Bar indicates 50 μm.

Full size image

Cell growth, karyotypic analysis, and authentication by RAPD-PCR.

The split ratio during subculturing was fixed at 1:5 after passage 100. Both the Ar-3 and Ar-13 cell lines showed similar growth curves consisting of a clear but flat lag phase, an exponential phase that slowly increased for 6–7 d, and a confluent plateau phase that remained stable for at least 4 d (Fig. 3). The population doubling time (PDT) at the most rapid growth period during the exponential phase was calculated to be 2.7 and 3.0 d for the Ar-3 and Ar-13 lines, respectively. The chromosome number was counted in the Ar-13 line and determined to be 6 (Fig. 4), which corresponds to that of diploid (2n = 6) in the species of origin, A. subalbatus (Clements 1992; Marquardt 2004), with N = 36 and N = 30 for Ar-3 and Ar-13 cells, respectively. RAPD-PCR was performed to further confirm that both Ar cell lines are novel. The electrophoresis band patterns obtained after RAPD-PCR on the DNA extracted from Ar-3, Ar-13, A. subalbatus, and seven additional insect cell lines were compared. The band pattern was confirmed by each of the two primers to correspond to Ar-3, Ar-13, and A. subalbatus, although some differences in the intensity of the bands were detected (Fig. 5). Both the Ar-3 and Ar-13 lines showed band patterns that were distinct from those of the other seven insect cell lines. In contrast, the two cell lines that originated from the same species (C6/36 and AeAl-2) did not show the same band patterns with each of the two primers. There was no difference between the resultant band patterns of RAPD-PCR performed in duplicate; one representative electrophoretic image for each primer is shown in Fig. 5.

Figure 3
figure 3

Growth curves of the A. subalbatus cell lines. A Ar-3 cells at the 100th passage. B Ar-13 cells at the 70th passage. Each point is the mean of three independent cultures. Bars represent standard deviation.

Full size image
Figure 4
figure 4

Micrograph of the chromosomes from Ar-13 cells stained with Giemsa (arrows). The number of chromosomes is 6, consistent with that of the diploid cells of the species of origin.

Full size image
Figure 5
figure 5

Electrophoretic images of random amplified polymorphic DNA (RAPD) profiles obtained by polymerase chain reaction (PCR) with two different 12-mer oligonucleotides (5′-CCGCAGTTAGAT-3′; 5′-ACTGGCCGAGGG-3′). Marker molecular marker; the numbers on the left represent the molecular size (bp). A. sabalbatus mosquito, the original source of the new cell lines; Ar-3 the new cell line from A. subalbatus, cultured in VP12 medium; Ar-13 the new cell line from A. subalbatus, cultured in MM + MGM450 medium. NIAS-AeAl-2, C6/36 cell line from A. albopictus mosquito; NIID-CTR cell line from C. tritaeniorhynchus mosquito; S2 dipteran D. melanogaster cell line; Sf21 lepidopteran S. frugiperda cell line; GaMe-LF1 lepidopteran G. mellonella cell line; and PC-1 coleopteran P. christophi cell line. The band patterns were confirmed to correspond to Ar-3, Ar-13, and A. subalbatus with some differences in band intensity. Both Ar lines showed band patterns that were distinct from those of the seven other insect cell lines.

Full size image

Susceptibility to four flaviviruses.

Viral susceptibility was demonstrated through molecular analysis of the supernatant of the virus-inoculated Ar-3 cells using RT-PCR (Fig. 6). The specific band at the expected amplicon size for JEV was detected from 3 d post-inoculation (p.i.), and intense staining of the band was observed 5 d p.i. In contrast, DENV was not detected until 15 d p.i. (data not shown). The insect flaviviruses CXFV and AEFV were only faintly detected at 3 d p.i. and then increased in staining intensity as the days p.i. progressed. Negative results for all viruses during the early phase (0–2 d p.i.) indicate that the remnant of the inoculum was low enough to test the susceptibility of cells to each virus. We occasionally observed cytopathic effect (CPE) in the JEV-inoculated cells 7 d p.i., whereas no CPEs in the CXFV-, and AEFV-inoculated cells, even after 7 d p.i. (data not shown). The titers of JEV and DENV in Ar-3 cells were determined by a standard plaque assay. JEV was detected from 3 d p.i. (8.67 × 103 PFU/mL) and then reached the high titer of 3.80 × 108 PFU/mL 7 d p.i., but DENV did not increase, which is in accordance with the RT-PCR results.

Figure 6
figure 6

RT-PCR detection of viral RNAs from the cultured medium of Ar-3 cells at 0, 1, 2, 3, 5, and 7 d after viral inoculation. The amplified products were separated on 2% agarose gels and visualized with ethidium bromide, and the portion of the expected amplicon size of each virus is shown. JEV Japanese encephalitis virus, DENV dengue virus, CXFV Culex flavivirus, AEFV Aedes flavivirus. The strain name appears after each abbreviated virus name.

Full size image

Discussion

In this study, we initiated primary cultures and established two cell lines from the mosquito A. subalbatus. In the primary culture, cell migration occurred in three different media, among which VP12 is unequivocally one of the best for culturing a wide range of mosquito species (Varma and Pudney 1969; Kuwata et al. 2012), including Armigeres, as confirmed by this study. The MM medium also promoted cell migration, as reported for many aedine mosquito cell lines (Hink 1976); when mixed with the MGM450 medium, this combination efficiently established the Armigeres mosquito cell line. Although these two media differ considerably in component richness and Na/K ratio (Mitsuhashi 1989), it is possible that mixing the media may offer some favorable culture conditions as observed previously (Kuno 1980; Bello et al. 2001). When grown in Schneider’s medium, the primary cells themselves possibly changed salt concentrations to form crystals, as the frequently formed vesicles are known to function as ion-exchange cell layers, similar to renal cells (Lynn et al. 1998). Thus, by employing combinations of different media, the optimal culture conditions were obtained, which likely facilitated continuous growth by inhibiting the vesicle formation observed in MM medium alone.

Despite being cultured in different media, the two established cell lines (Ar-3 and Ar-13) were composed of similar types of cells that were adherent and karyotypically normal. Both cell lines formed distinct, compact colonies reminiscent of those formed in mammalian embryonic stem cell lines (Englund et al. 2010; Aran et al. 2010). Similar colonies have been observed in other mosquito cell lines, including AeAl-2 (Mitsuhashi 1981) and ATC-10 (Singh 1967). Therefore, this characteristic may arise in cell lines of mosquito neonate larval tissue origin. RAPD-PCR results confirmed that both Ar lines originated from A. subalbatus; thus, these cell lines are the first available in vitro cell models for studying the biology of this species. Comparative analysis of RAPD-PCR results demonstrates slight differences between Ar-3, Ar-13, and A. subalbatus, suggesting that genetic alterations may have been generated. Cryopreservation at −80°C in 12% dimethyl sulfoxide for at least 1 yr was confirmed in both cell lines at around the 100th passages (data not shown), allowing storage without accumulation of the genetic and karyotypic changes expected a priori in cultured cells.

The Ar-3 cell line is susceptible to JEV but relatively insensitive to DENV, as demonstrated by the viral isolation results from wild-caught A. subalbatus mosquitoes (e.g., Weng et al. 1999). Additionally, the viral susceptibilities of this cell line did not tend to be accompanied by CPEs as in many mosquito cell lines (e.g., Peleg 1968; Sudeep et al. 2009). Ar-3 is also susceptible to insect flaviviruses with no signs of CPE, suggesting the possibility of persistent infection by insect flaviviruses. Insect viruses known to be harbored in mosquito cell lines include the first reported insect flavivirus, a cell fusing agent isolated from Peleg’s cell line from Aedes aegypti (Stollar and Thomas 1975), other insect flaviviruses, and other insect viruses (Scott et al. 2010). Accordingly, the Ar cell lines should be examined for persistent viral infections that could potentially be related to derivation of a cell line that generates CPE-like vacuolation in early subcultures. In general, viral susceptibility in vitro is restricted by the host range of the virus (Kuno 2007); therefore, Ar-3 will be extremely valuable for the detection of Armigeres-associated viruses such as Armigeres totivirus (Zhai et al. 2010). The Ar-3 line is expected to improve efficient isolation of viruses from A. subalbatus mosquitoes, as C6/36 cells exhibit serious cellular damage, including cell death, in almost all cases of inoculation with A. subalbatus mosquito homogenates. Thus, the two new cell lines established from A. subalbatus in this study are cellular models for this species and may serve as beneficial tools for the propagation of viruses associated with the Armigeres mosquito.