SUBJECTS, MATERIALS AND METHODS

Serum samples, ELISA and RT-PCR - We analysed the samples using a neutralisation test; we used serum samples from 37 patients from the city of Ribeirão Preto, state of São Paulo, Brazil with HCPS previously confirmed through positive IgM or a high IgG titre in ELISAs using a recombinant N protein from Araraquara virus (ARAQV) as the antigen (^{Figueiredo et al. 2008}, ^2009a). Additionally, in 27 of the samples, the ARAQV genome was amplified using RT-PCR (^{Moreli et al. 2004}). An infection by ARAQV was confirmed in these patients through sequencing the amplicons containing 264 nucleotides from the viral S segment.

Virus culture - The Rio Mamoré virus (RIOMV) strain HTN-0007 was kindly provided by Dr Robert E Shope from the University of Texas Medical Branch, at Galveston. The virus was grown in Vero-E6 cells (African green monkey kidney) and maintained in Eagle’s minimum essential medium (EMEM) supplemented with 10% heat-inactivated foetal bovine serum, 50 mg/mL of gentamicin and 2 mg/mL of amphotericin B (Vitrocell, Brazil). The cells infected with RIOMV were cultivated for 14 days at 37°C with 5% CO₂. Tissue culture medium obtained from flasks containing infected cells was aliquoted as a viral stock and stored at -70ºC.

Plaque (microplaque) assay - The plaque assays (microplaques) were performed after the Vero-E6 cell colonies were incubated for 48 h at 2 × 10⁵ cells per well in 24-well plates. We added 150 µL per plate well of decimal serial dilutions (10^-1-10^-8) using the RIOMV stock to the Vero-E6 colonies in duplicate. The microplates were allowed to adsorb for 1 h at 37°C with 5% CO₂ and were gently rocked every 15 min. After adsorption, each well was overlaid with 1 mL of a medium-viscosity medium containing 1.5% of carboxymethylcellulose sodium (Sigma-Aldrich, Germany) in EMEM supplemented with 1.5% of foetal calf serum and 0.3-1.5% of dimethyl sulfoxide (DMSO) (Sigma-Aldrich). The plates were incubated for 13 days at 37°C with 5% CO₂. After incubation, the cell monolayers were rinsed and fixed with a 10 % buffered formalin solution for 1 h at room temperature and under constant agitation. Uninfected cells were fixed and used as negative controls. The plates infected with RIOMV were handled in a BSL-3 laboratory. The fixed cell monolayers were stained using a naphthalene black solution (0.1% naphthol blue-black, 6% glacial acetic acid and 1.36% sodium acetate) under constant agitation for 1 h.

Finally, the microplaques in the cell monolayers were counted under a microscope. The RIOMV stock titre was calculated as plaque-forming units (PFUs) per millilitre using the following formula: viral titre (PFU/mL) = number of microplaques x dilution factor/volume of virus or virus-serum-mixture per well.

Plaque (microplaque) reduction neutralisation test - Heat-inactivated serum specimens (56ºC for 30 min) were diluted two-fold (1:10-1:160) with EMEM and 250 µL of each serum dilution was mixed to the same volume of a solution containing 100 PFU of RIOMV. We also mixed serum from healthy individuals who were negative for hantavirus antibodies with the virus, which was used as a negative control in the assay. Tubes containing the mixtures were incubated for 1 h at 37°C to allow the neutralising antibodies to bind with the virus. Thereafter, 250 µL of the virus-serum mixture was added to the wells in duplicate and the plate was incubated for 1 h. Next, the virus-serum mixtures were removed and the wells were rinsed in phosphate-buffered saline (PBS). Two wells containing only culture medium were used as negative controls and two other wells were inoculated with 100 PFU of RIOMV as positive controls. The plates were incubated at 37°C with 5% CO₂ and gently rocked every 15 min for 1 h. Thereafter, each well was overlaid with 1 mL of the semi-solid overlay medium used for the microplaque assay and the cell monolayers were incubated for 13 days at 37°C in a CO₂ incubator, as mentioned above. Next, the cell monolayers were stained with naphthalene black, as described above. A neutralising antibody titre was defined as the maximum serum dilution that inhibited 50% of microplaque formation by RIOMV in the cell monolayer (^{Yu et al. 2013}).

RESULTS

Virus microplaques were not observed either without using DMSO or using 0.3% of DMSO in the cell culture medium. Using 0.6% and 0.8% of DMSO, the microplaques in the cell monolayer were visible 13 days after infection. Using 1% DMSO, we observed increased destruction of the cell monolayer and the microplaques became impossible to distinguish. Concentrations of 1.5% DMSO or higher promoted considerable destruction of the cell monolayer, likely due to a toxic effect. Based on these results, we used 0.6% as the DMSO concentration in the medium-viscosity overlay medium.

Under the microscope, microplaques with an average size of 328 µm (± 94 µm) and an irregular border were used to determine the RIOMV titre of our stock as 6.6 × 10⁵ PFU/mL (Figure).

Microplaques formed in Vero-E6 cell infected by Rio Mamoré virus with addition of dimethyl sulfoxide 0.6% in medium-viscosity overlay medium. Shown in 40X (A) and 1,000X (B). 

Neutralising antibodies against RIOMV were observed in 16.21% (6/37) of the patient sera. Neutralising antibody titres ranged from 10-80. The titre of the neutralising antibodies could not be correlated with evolution (survival or death) of the patients, as shown in Table. Additionally, the neutralising antibody titres could not be correlated with the titres obtained using IgG-ELISA and the N recombinant antibody of ARAQV as an antigen. For example, the serum from one fatal HCPS case (patient 9) (Table) exhibited a 6,400 titre in the IgG ELISA, but exhibited only a 40 titre in the neutralisation test.

DISCUSSION

In cell culture, hantavirus infection typically yields a slow cytopathic effect that also reduces plaque production (^{Markotic et al. 2003}). In the present study, we adapted a treatment for virus-infected cells using DMSO; this treatment has been indicated as capable of increasing the size and number of plaques in virus-infected cell monolayers (^{Hughes 1993}). Therefore, we found that adding 0.6% of DMSO to the medium-viscosity overlay enhances microplaque formation in Vero-E6 cell cultures infected by hantavirus (RIOMV), which facilitates clear visualisation of these microplaques under the microscope.

It is a known difficulty to adapt wild hantavirus to cell culture. In certain cases, the virus is lost after three or four passages, which occurred using ARAQV in our laboratory (^{Kallio et al. 2006}). Thus, we used RIOMV, which is a virus that is well-adapted to Vero-E6 cell culture for use in the plaque (microplaque) reduction neutralisation test.

RIOMV was first described in Bolivia in 1996 and was associated with the Oligoryzomys microtis rodent (^{Bharadwaj et al. 1997}). RIOMV was also isolated from O. microtis in Peru and two strains of the virus were observed in the Brazilian Amazon and infected the same rodent as well as Oligoryzomys sp. RIOMV-producing HCPS has recently been described in Peru (2 cases) and in French Guiana (1 case produced by the variant Maripa virus) (^{Figueiredo et al. 2014}). A human fatal disease involving RIOMV was recently reported in Brazil (^{Carvalho de Oliveira et al. 2014}).

ARAQV was first observed in a fatal HCPS case from the city of Araraquara in 1995. ARAQV is harboured by Necromys lasiurus, which is a rodent that lives in regions of Brazilian Cerrado (a savannah-like ecosystem) (^{Suzuki et al. 2004}). ARAQV is the most virulent hantavirus in Brazil and causes HCPS with an approximately 50% case-fatality rate in the southeastern region as well as in the central plateau (^{Figueiredo et al. 2014}). Serum samples analysed in the present study were collected from patients that were likely infected by ARAQV.

To detect hantavirus-neutralising antibodies, we developed a test that considers the ability of serum to reduce the number of microplaques by 50%. Other authors have successfully reduced plaques against hantaviruses by 50% (^{Yu et al. 2013}). Hantavirus-neutralising antibodies can be observed earlier than 10 days after infection and are commonly present at the onset of a hantavirus disease (^{Horling et al. 1992}). Using our microplaque reduction neutralisation test for RIOMV, we identified these antibodies in 16.2% of the tested sera. The low level of positivity in samples that were previously diagnosed through ARAQV genome amplification using RT-PCR (73%) and by IgG or IgM-ELISAs using ARAQV (all samples) was likely due to ARAQV antibodies cross-reacting with RIOMV; this inconsistency was only present in six samples (^{Figueiredo et al. 2009b}).

As previously demonstrated, ARAQV induces a mixed T-helper (Th)1/Th2 strong immune response during the course of HCPS and the magnitude of the Th1 response effector cytokines is correlated with disease severity (^{Borges et al. 2006}). However, protection markers that clearly relate to disease survival are not well-known. Previous work has shown that a neutralising antibody immune response confers protection from disease severity by the Andes virus in animals (^{Custer et al. 2003}). Previous work has also shown that high neutralising antibody titres correlate with less severe HCPS cases (^{Bharadwaj et al. 2000}). Neutralising antibodies specifically bind to certain extracellular hantavirus protein epitopes and impair infection of its target cells, which also reduces the level of free virus (^{Borges et al. 2006}, ^{Easterbrook & Klein 2008}, ^{Kruger et al. 2011)}. Thus, the level of hantavirus-neutralising antibodies may suggest HCPS prognoses in addition to discerning the infecting virus. However, we did not observe a correlation between patient prognosis and neutralising antibody titres in the present study and fatal cases showed the same neutralising antibody titre as the survivors. The lower number of studied cases may have influenced our results. Moreover, we could not obtain information on the infecting virus because patients were infected by ARAQV and the virus used in the neutralisation test was RIOMV.

In our study, low neutralising serum titres were observed in all six positive sera (a titre of 80 or less). This result is consistent with previously observed titres in survivors of hantavirus infections with the Andes and Sin Nombre viruses, where high neutralising-antibody titres against the hantavirus predominant in each region were observed and titres lower than 100 were observed through cross-reaction neutralisation for other species of hantavirus (^{Valdivieso et al. 2006}). Furthermore, neutralising antibodies compose a small proportion of the antibodies against hantavirus in HCPS patients. Moreover, RIOMV belongs to a different species of ARAQV than the virus that infected our patients.

Herein, we showed the development and evaluation of a microplaque reduction neutralisation test for RIOMV that can detect and quantify hantavirus-neutralising antibodies in patient sera. This assay could be used to diagnose infection, in pathogenesis studies and to analyse antiviral drugs. Additional studies using ARAQV or another genotype adapted to cell culture are necessary to detect neutralising antibodies against hantaviruses that cause HCPS in Brazil and in other South American countries (^{Figueiredo et al. 2009b}, ^{Souza et al. 2011}, ²⁰¹²).