For the production of CHIKV replicon particles a split helper system should be applied since it has been described that providing the structural proteins via two separate helper constructs did not result in detectable incidental production of infectious particles due to recombination 6 7 .

A CHIKV replicon was established, in which the structural genes were replaced by a reporter gene (Figure 1A). For easy readout, Gluc was chosen as reporter protein, since it is secreted into the supernatant and therefore allows easy readout without the need of preparing cell lysates. Furthermore, two helper plasmids were established expressing either the CHIKV capsid protein C (pChikHelper-C) or the CHIKV envelope proteins p62-6K-E1 (pChikHelper-E). The helper RNAs contain the 5’ and 3’ CHIKV replication signals and a subgenomic promoter followed by either the capsid gene or the remaining structural proteins, respectively (Figure 1A). The helper RNAs as well as the CHIKV replicon RNA can be produced in vitro from plasmids, which contain these sequences under control of an SP6 promoter.

To prove efficient secretion of Gluc from the replicon in vitro transcribed replicon RNA was electroporated into baby hamster kidney-21 (BHK-21) cells. After electroporation, increasing amounts of Gluc could be detected over time in the supernatant demonstrating the efficient replication and secretion of Gluc of the established CHIKV replicon (Figure 1B).

For production of VRPs the CHIKV Gluc replicon was coelectroporated with the two helper RNAs into BHK-21 cells. It has been observed that VRP production is more efficient at temperatures lower than 37°C. Hence, to compare VRP production at different temperatures, coelectroporated cells were incubated at either 37°C or 32°C and the activity of secreted Gluc was measured in Relative Light Units (RLUs). Whereas at 37°C Gluc secretion increased over time until 48 h post electroporation before dropping down again, Gluc levels remained lower and more constant at lower temperature (Figure 2A). Interestingly, when the supernatants were passaged once on BHK cells to analyze for produced VRPs it turned out that VRPs had been released to higher and more constant levels for the experiment performed at 32°C (Figure 2B). This suggests that although Gluc was more efficiently secreted at 37°C the VRP production was indeed more stable at lower temperature. Hence, for optimized VRP production, the coelectroporated cells were incubated at 32°C and VRPs were harvested at 36 h post electroporation. Harvested VRPs were purified via a sucrose cushion. As determined by CHIKV real-time PCR, the 30-fold concentrated VRP stocks contained around 1 × 10⁹ viral RNA copies/ml.

For comparable analyses same test conditions need to be applied. These include among others parameters like the amount of VRPs used in the NT assay as well as the readout time or the plate format used. Even though the VRP concentration in established stocks can be determined via real-time PCR to calculate defined infection rates it might be desirable to use the same VRP batch for as many NT assays as possible. NT assays using different amounts of VRPs corresponding to multiplicities of infection (MOIs) of 5, 0.5 and 0.05 calculated based on the determined RNA copies in the VRP stock were evaluated. Another reason for testing also lower MOIs was that this better reflects the conditions used in classical plaque neutralization assays. There, a low MOI needs to be chosen to be able to count single plaques for readout. Evaluation of different MOIs was first carried out using the neutralizing monoclonal antibody (mAb) D3.62 directed against CHIKV E2 (T. Couderc and M. Lecuit, unpublished data). The respective experiments were performed in either a 24- or 96-well format and readout was done at 6 and 24 h post VRP infection (Figure 3). To evaluate the neutralization activity, the infectivity rates measured via Gluc secretion retained after VRP incubation with the respective antibody dilution were determined in percent compared to the VRP only control.

As shown in Figure 3, stepwise neutralization was observed after applying the monoclonal antibody in different dilutions. In the 24-well format the most consistent data were obtained when VRPs were used at MOI 5 or MOI 0.5. Decreasing the MOI to 0.05 increased the deviation from the mean. For the 96-well plate format MOI 5 yielded the most reliable data whereas the number of outliers increased at lower MOI. The least reliable data were obtained in the 96-well format using an MOI of 0.05. Comparing the data received after evaluating the NT assay at 6 h post infection (p.i.) versus 24 h p.i. revealed similar results indicating that the readout of the Gluc VRP assay readily can be performed within one day.

Similar results were obtained when the same set of conditions was tested using a human serum from a CHIKV patient (1753/06) (Figure 4). Using immunofluorescence analysis, the serum was confirmed to be positive against CHIKV (see also Figure 5A).

Whereas the human CHIKV patient serum resulted in dose dependent neutralization, a negative serum from a healthy person did not show any neutralization activity (Figure 4). Again, MOI 5 and MOI 0.5 turned out to be the most applicable conditions in the 24-well format and MOI 5 yielded the most consistent results in the 96-well format (Figure 4).

To further evaluate the applicability of the established VRP based NT assay comparison to a classical plaque neutralization assay based on infectious CHIKV was drawn. The respective analysis was done using again the neutralizing monoclonal antibody D3.62, the patient serum 1753/06, as well as two further sera (662/06 and 575/06) from CHIKV patients diseased during the 2006 epidemic in the Indian Ocean region 30 . All sera as well as the monoclonal antibody were confirmed to be reactive against CHIKV using indirect immunofluorescence (Figure 5A). The plaque neutralization assay was performed in a 6-well plate format using 100 plaque forming units (PFU) per well (Figure 5B), whereas for the VRP based NT assay both the 24-well and 96-well format was applied using MOI 5 (Figure 5C). The neutralization activity of a serum is usually judged by determining the serum dilution, which results in 50 or 80% reduction of infectivity (NT₅₀ or NT₈₀ titer, respectively). When using different methods of NT assays, comparison of absolute NT₅₀ values is limited due to different test conditions. However, the order of neutralizing activity among antisera tested should be the same. Table 1 provides details on the predicted NT₅₀ values using the data sets depicted in Figure 5B and 5C. Overall, the NT₅₀ titers determined via the classical plaque neutralization assay were lower than those observed in the VRP assays (Table 1). Nevertheless, the same order with regard to neutralizing activity of the samples was obtained for all NT assay types, with the monoclonal antibody neutralizing the best and serum 575/06 neutralizing the least (Table 1). This comparison was supported by narrow 95% confidence intervals (CI) determined for each serum (Table 1). In the 96-well VRP assay format, all 95% CI overlapped partially. In contrast, the 95% CI overlapped marginally only for two sera in the plaque assay, and no overlap at all was observed in the 24-well VRP assay format. Hence, the CHIKV VRP based NT assay represents a useful alternative to CHIKV NT assays involving infectious virus and in addition allows a fast and easy analysis of neutralizing antibodies in serum samples that can be performed within one day.

a Neutralization assay performed with virus particles (plaque assay).

b Neutralization assay performed with virus replicon particles (Gluc assay).

c NT₅₀ values were determined via probit analysis using the data set of Figure 5B.

d NT₅₀ values were determined via probit analysis using the data sets of Figure 5C.

e 95% confidence interval.

Baby hamster kidney-21 (BHK-21) cells were maintained in Glasgow´s Minimum Essential Medium (GMEM) supplemented with 5% fetal bovine serum (FBS), 1% L-glutamine, 10% tryptose phosphate broth, 20 mM HEPES pH 7.2, 100 U/ml penicillin and 0.1 mg/ml streptomycin at 37°C and 5% CO₂.

A stock of recombinant wild-type CHIKV (CHIKV-LR2006 OPY1, 36 ) was produced on BHK-21 cells. Cells were infected at an MOI of 0.1 for 1 h and virus harvested from the supernatant 2 days p.i. was stored at −80°C. Experiments with infectious CHIKV were performed in a biosafety level 3 laboratory.

Monoclonal antibody D3.62 (35 mg/ml) is directed against CHIKV E2 protein (T. Couderc and M. Lecuit, unpublished data). Human sera containing neutralizing CHIKV antibodies were obtained from patients returning to Europe from the Indian Ocean region in 2006 and have been described previously 30 . Briefly, sera from patient 575/06 returned from the Seychelles, patient 662/06 from La Réunion and patient 1753/06 from Mauritius. Monoclonal antibody Z2G2 recognizing CHIKV capsid protein was kindly provided by Petra Emmerich (Bernhard Nocht Institute, Hamburg, Germany).

BHK-21 cells were cultured on glass coverslips and infected with CHIKV-LR2006 OPY1 at an MOI of 0.5. At 24 h after infection, cells were fixed with ice-cold methanol / acetone (1:1) and air-dried. Serum samples or monoclonal antibodies were diluted 1:5000 in PBS and incubated for 1 h at 37°C on fixed cells. Serum antibodies were detected by Alexa 488-labeled goat anti-human IgG (Jackson ImmunoResearch, 1:500) and monoclonal antibodies by cyanine 3-conjugated goat anti-mouse IgG (Jackson ImmunoResearch, 1:200) using fluorescence microscopy (Axiovert 40 microscope, Zeiss). Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI).

The construction of full-length infectious cDNA of CHIKV-LR2006 OPY1 is described elsewhere 36 . To obtain plasmid pChikRepl containing the cDNA of a CHIKV replicon, the region of the infectious cDNA clone corresponding to the coding sequence of CHIKV structural proteins (nucleotides 7565–11310 of CHIKV-LR2006 OPY1) was replaced with sequence 5’

CCTAGG

GTTTAAAC

CCTAGG

GTTTAAAC

CCTAGG

GTTTAAAC

Replicon and helper DNA templates were linearized with NotI and in vitro transcribed using the mMESSAGE mMACHINE SP6 Kit (Ambion). For recovery of VRPs 1 μg of each, the replicon and helper RNAs, were coelectroporated into 1 × 10⁶ BHK-21 cells using BHK-21 preset protocol of Gene Pulser Xcell (Bio-Rad). Cells were seeded into 25 cm² flasks and incubated at 32°C for 36 h. For large-scale VRP production supernatants of six electroporations were pooled for purification. After removing detached cells and cell debris by clarifying for 30 min at 4000 g and 4°C, the supernatants were passed through a 0.45 μm filter and applied to a 20% sucrose cushion followed by centrifugation for 90 min at 25000 rpm and 4°C (SW 32 Ti rotor, Beckman Coulter). Pellets were resuspended in TNE buffer (50 mM Tris–HCl, pH 7.4, 100 mM NaCl, 0.5 mM EDTA) over night at 4°C before passing through a 0.22 μm filter. VRPs were stored at −80°C.

RNA from VRPs in the cell culture supernatant or after sucrose cushion purification was extracted using NucleoSpin RNA Virus Kit (Macherey-Nagel). The isolated RNA was detected by real-time reverse transcription-PCR (RT-PCR) using the SuperScript III One-Step RT-PCR System with Platinum Taq DNA polymerase (Invitrogen). For detection of CHIKV RNA, the 25 μl reaction contained 3 μl of RNA, 1x Reaction Mix, 0.5 μg BSA, 0.5 μl SS III RT / Platinum Taq Mix, 0.6 μM of primer CHIKSI, 0.6 μM of primer CHIKASI and 0.2 μM of the CHIKP probe 30 . Thermocycling was performed on a LightCycler 480 (Roche) programmed for: 30 min at 50°C for reverse transcription, 2 min at 94°C to activate the Taq polymerase and 45 PCR amplification cycles of 15 sec at 94°C, 30 sec at 58°C and 30 sec at 72°C. Photometrically quantified in vitro-RNA transcripts of the target regions were used in the PCR to generate a standard curve for viral RNA quantification.

Gluc was measured from the supernatant of infected cells using Renilla Luciferase Assay System (Promega) according to the manufacturer’s instructions. Luciferase activity was measured in RLUs. Measurement was performed automated in a Synergy 2 microplate reader (BioTek) using polystyrol microplates (Greiner Bio-one).

The day before infection, 24- or 96-well plates were seeded with 1 × 10⁵ or 2 × 10⁴ BHK-21 cells per well, respectively. VRPs were applied in the NT assay at MOI 5, 0.5 or 0.05 (calculated based on VRP RNA copies/ml). All dilutions were performed using GMEM supplemented with 1% FBS. VRP dilutions were incubated with serially diluted antibody or human serum for 1 h at 37°C before adding the mixture to a monolayer of BHK-21 cells in either a 24-well plate (total volume per well 200 μl) or a 96-well plate (total volume per well 40 μl). After incubation for 1 h at 37°C the inoculum was removed, cells were washed once with PBS and medium was added. Supernatants for Gluc measurement were taken at 6 h and 24 h p.i. Neutralization potency was determined as percentage of measured Gluc activity compared to Gluc readout after VRP application without antibody/serum.

About 100 PFU of infectious wild-type virus were incubated with serially diluted antibody or human serum for 1 h at 37°C, added to a monolayer of BHK-21 cells in a 6-well plate (total volume per well 600 μl) and incubated at 37°C for 1 h. Subsequently, the inoculum was replaced by an overlay containing 0.6% agarose in MEM. At 48 h p.i. cells were fixed with 7% formaldehyde and plaques were visualized using crystal violet staining (1% crystal violet in 50% ethanol). Neutralization potency was determined as percentage of plaque titers compared to plaque titers after virus infection without antibody/serum.

Probit analysis for determination of NT₅₀ values was done with the SPSSV21 software package (IBM, Ehningen, Germany).