A validated and standardized pseudotyped microneutralization assay as a safe and powerful tool to measure LASSA virus neutralising antibodies for vaccine development and comparison

2.1 Cell line maintenance

For the production and titration of Lassa virus pseudotypes, as well as for pseudotype-based microneutralization assay, human embryonic kidney 293T/17 (HEK293T/17) (ATCC, cat number CRL-11268) are maintained in complete medium Dulbecco’s modified essential medium (DMEM) high glucose, pyruvate (Gibco, Catalog number: 11995073). The medium is supplemented with 10% FBS South America Origin EU approved Heat Inactivated (Euroclone, Catalog number: ECS5000LH) and 1% penicillin–streptomycin (Pen-Strep) (Merk, cat number: P4333). The cells are kept in incubator at 37°C with 5% CO₂.

2.2 Lassa virus and lentiviral vector packaging plasmids

The gene of Lassa virus (LASV) (strain Josiah) glycoprotein is synthesized and cloned into plasmid expression vector phCMV3 (Sino Biological, cat number: VG40079-UT). The p8.91 lentiviral packaging plasmid expressing gag-pol, and the pCSFLW plasmid containing a firefly luciferase reporter are provide by Prof. Nigel Temperton (Viral Pseudotype Unit-VPU, University of Kent Medway School of Pharmacy).

2.3 Pseudotype production and harvest

For LASV pseudotype production, approximately 4.5×10⁶ HEK293T/17 cells. 24h hours post seeding, cells are transfected with a plasmid DNA mixture composed of: 3 μg of p8.91, 4.5 μg of pCSFLW and 3 μg of LASV-G using 24 μl of EndoFectin™-Lenti (GeneCopoeia, Catalog number: EF001) in 500 μl reduced serum medium Opti-MEM 1X (Gibco, Catalog number: 31985-070). At least 6 hours post-transfection 8 ml of fresh DMEM is replaced to the plates. 72 hours post-transfection LASV pseudotype particles are collected, centrifuged, filtered through 0.45μm syringe-driven membrane filters and stored at -80°C.

2.4 Pseudotype titration assay

Pseudotype titration is performed in 96 well plate format: 50 μL of DMEM are added to every well except raw A where 100 μL of LASV pseudotype particles are added. Serial two-fold dilutions are performed from raw A to G, and 1×10⁵ HEK293T/17 cells are added to each well. After 48 hours, 50 μL of Bright-Glo Luciferase Assay System (Promega, catalog number: E2620) are dispensed in each well and placed in a microplate shaker for 5 minutes at 400 rpm (Fisherbrand Catalog number: 88861024). Plate readout is performed with luminometer (Perkin Elmer; Victor Nivo) following parameter: Luminescence Endpoint, counts single label, 700 nm, 1000 ms measurement time.

2.5 Pseudotype based microneutralization assay

Pseudotype-based microneutralization (PBNA) is performed in 96 well plate format. LASV pseudotype working solution corresponds to 1×10⁸ relative luminescence unit (RLU) divided by the pseudotype titre expressed in RLU/mL. 90 μL of DMEM are added to column one and 50 μL are added to the other wells. To column 1 are added 10 μL of serum samples (1st International Reference Panel for anti-Lassa fever virus antibodies 21/332 NIBSC containing 20/228, 20/204, 20/246, 20/222 and 20/248) at the proper dilution and immunoglobulins depleted serum (BBI Solutions, Catalog number: SF505-2) as used as negative control. Serial two-fold dilutions are performed from column 1 to 10. Subsequently, 50 μL of pseudotype working solution are added to each well up to column 10, and to column 12 from raw A to H (pseudotype control wells). The plate containing the mix pseudotypes-serum is incubated for 1 hour at 37°C with 5% CO₂, afterwards 1×10⁵ HEK293T/17 cells are added to each well. After 48 hours, 50 μL of Bright-Glo Luciferase Assay System are dispensed in each well and the plate is shake for 5 minutes at 400 rpm in a dark environment (Fisherbrand, Catalog number: 88861024). Plate readout is performed with a luminometer (Perkin Elmer; Victor Nivo).

2.6 Statistical analysis

Calculation of Pseudotype titres were estimated by means of Excel^TM software; the pseudotype titres obtained at each point in a range of dilution points were expressed as RLU/mL, and the arithmetic mean was calculated. For the analyses of pseudotype-based neutralisation assays, the GraphPad Prism version 8.4 package was used (GraphPad Software, GraphPad, 2365 Northside Dr, Suite 560, San Diego, CA 92108, USA). The titres were firstly normalised based on pseudotype only mean value and cell only mean value, and IC₅₀ values were calculated by a non-linear regression model (log (inhibitor) vs. normalised response-variable slope) analysis. Titres were subsequently expressed as the range of dilution in which the IC₅₀ value lay.²⁷

2.7 Relative accuracy

The assessment of the accuracy of a test is conducted by analyzing the data acquired for the evaluation of Dilutional Linearity (GMTs observed). According to the ICH guideline Q2 (European Medicines Agency. ICH Q2(R2) Validation of analytical procedures - Scientific guideline. 2022²⁸), there are two ways to test the accuracy of a test: by using a conventional true value or an accepted reference value as the expected value of GMT. The expected GMTs are determined by calculating the geometric mean of the results obtained from the neat sample divided by the corresponding factor of the two-fold serial dilution. To evaluate the Relative Accuracy, the percentage of recovery on the GMT of the observed values and the expected (true) titre is calculated by applying the following formula:

2.8 Dilutional linearity

Linearity is assessed by testing the 20/228, the 20/222 and the 20/204 samples from NIBSC reference panel for anti-Lassa fever virus in a two-fold dilution scheme, starting from a dilution of 1:20 to 1:2560 dilution The sample dilutions were tested in two repetitions per plate, and in two different plates by two different operators, on two different days. For each dilution, the geometric mean titre (GMT) is calculated. A linear regression analysis of logarithm base 2 of serum dilution versus logarithm base 2 of GMTs using the least squares method over the entire dilution range is performed and the slope of regression line and the coefficient of determination (R²) is calculated.

2.9 Precision

Precision is used to define the amount of scatter between several measurements obtained from multiple testing of the same sample under the standardized conditions. The Precision parameters are evaluated by the Percent Geometric Coefficient of Variation (%GCV) determined as

The validation process considers four crucial aspects of precision that are essential to evaluate the reliability of laboratory testing.

These aspects include intra-day variation, which measures the expected variation within each day under identical operating conditions.

Additionally, intra-operator variation is evaluated, which measures the expected variation within each operator under identical operating conditions.

Intermediate precision, also known as inter-test variation, is considered to account for the variations and random events that may occur within the laboratory due to different environmental conditions, operators, equipment, or days.

A further aspect is format variability (FV), which evaluates the expected variation between the average titer results obtained from multiple replicates in routine testing and considers two independent analyses (k = 2), each consisting of one replicate (n = 1). The FV is determined by the following equation:

2.10 Specificity

An assay is defined as specific when it can, in a non-ambiguous manner, detect the analyte in the presence of other components. Here, specificity was assessed by testing LASV pseudotype against the second International Standard for anti-SARS-CoV-2 immunoglobulin (NIBSC code: 21/340), and against the first WHO International Reference Panel of Anti-Ebola virus (EBOV) Convalescent Plasmas (NIBSC code: 16/344) whose individual panel members are reported with the following NIBSC codes: 15/280 (ARC), 15/282 (NHSBT), 15/284 (NOR), 15/286 (INMI), 15/288 (negative human plasma). As negative control the immunoglobulin’s depleted serum was used (“Pseudotype based microneutralization assay” paragraph).

To assess the Specificity parameter the ratio between sera and no specific NIBSC, and between sera and negative samples is calculated.

3.1 Lassa envelope glycoprotein pseudotyped lentivirus particles

To develop a safe and robust pseudotype neutralisation assay for LASV, phCMV3-GP LASV (strain Josiah) was co-transfected with p.8.91 gag-pol plasmid and pCSFLW plasmid containing a firefly luciferase reporter as shown in Figure 1A. 72 h post-transfection, cell culture supernatants containing pseudotypes particles (PV) were harvested and their transduction efficiency was assessed by measuring titres expressed as RLU/ml in HEK293/17 cells. As shown in Figure 1B, Lassa GP showed a pseudotype mean titre/ml of 4.98e+9 RLU/ml.

Figure 1. Schematic representation describing PV production and titration.

(A) Packaging construct gag pol, reporter vector and pCMV3-GP LASV (strain Josiah) expression plasmids are co-transfected into HEK293/17 cells. Pseudotype proteins are packaged by the cell and budding occurs at the cell membrane to yield pseudotypes bearing desired glycoproteins and incorporated reporter. Culture supernatant was collected 72 hours after transfection and filtered through a 0.45 um filter. (B) Target HEK293/17 cells were transduced with LASV GP-pseudotyped lentiviral particles and the infectivity measured as relative luminescent units (RLU/mL). The grey column indicated the cellular control (CC CTR) as comparison.

3.2 In vitro inhibition of LASV pseudotypes by antisera

To validate this pseudotype neutralisation assay for LASV, we performed a series of validation experiments using the MHRA Lassa panel 21/332 (1st International Reference Panel for anti-Lassa fever virus antibodies 21/332 NIBSC ).

This panel consisted of 5 sera (20/228, 20/204, 20/222, 20/246 and 20/248) and was established by an expert committee on biological standardisation in 2021. First, the IC50 of all 5 sera in the panel were tested against LASV GP pseudotypes/plate and 20/228 was selected as high control, 20/204 as medium and 20/222 as low control, Figure 2.

Figure 2. In vitro inhibition of pseudotypes by Lassa panel 21/332 antisera.

(A) Neutralisation curves obtained from WHO Lassa panel. Reference antisera were used at starting dilution of 1:20 while LASV PV diluted to 1.000.000 RFU dose input as determined previously from titration. Each points represents the mean and standard deviation of two replicates per dilution repeated for three independent experiments. (B) Half-maximal inhibitory dilution of reference antisera as calculated from “log (inhibitor) vs. normalized response—Variable slope” from “Dose-response-Inhibition using Graph pad software. The values are previously normalized on only mean value of PV value and cell-only mean value.

3.3 Relative accuracy

To evaluate the relative accuracy of the assay, each serum was tested in 2 different analytical sessions performed by 2 operators with 2 replicates per session. The assay is considered to have acceptable relative accuracy if the percentage of recovery on the GM of the reportable values and the expected is within 50% - 200%. The assay is accurate from 1 to 64 dilutions for 20/228 standard. In addiction the table below, Table 1, also shows that the RAs at the 1:16 dilution for 20/204 and the 1:4 dilution for 20/222 are just below 50% but can be considered within the range of accurate dilutions as they are between multiple accurate dilutions.

3.4 Precision

Precision parameters are usually evaluated through the variance, standard deviation, or coefficient of variation of a series of measurements, Figure 3.

Figure 3. Precision representation for all dilution for 20/228 high, 20/204 middle and 20/222 low serum.

Visual plot of the precision results obtained from each analysed sample. The x-axis shows the different dilutions, while the y-axis shows the IC50 value. Each dot represents one replicate, performed in a different day or by a different operator. Straight lines indicate the mean of the 4 values, coupled with the SD of the replicates.

3.4.1 Intra-Day variation

The two geometric means between results from same day but different operator (D1-O1 and D1-O2, D2-O1 and D2-O2) are calculated for each dilution. The Intra-Day variation is determined by the %GCV of the two geometric means.

The assay is considered to have acceptable Intra-Day variation if all samples have a %GCV less or equal than 50%. The results shown in Table 2 indicate that the assay has acceptable Intra-Day variation from 1:20 (neat) to 1:2560 for 20/228, 20/224 and 20/222.

The two geometric means between results from same operator but different day (D1-O1 and D2-O1, D1-O2 and D2-O2) are calculated for each dilution. The Intra-Operator variation is determined by the %GCV of the two geometric means.

The assay is considered to have acceptable Intra-Operator variation if all samples have a %GCV less or equal than 50%. The assay has also acceptable Intra-Operator variation from 1:20 (neat) to 1:2560 for 20/228, 20/224 and 20/222 (Figure 3, Table 3).

3.4.2 Intra-Operator variation

The two geometric means between results from same operator but different day (D1-O1 and D2-O1, D1-O2 and D2-O2) are calculated for each dilution. The Intra-Operator variation is determined by the %GCV of the two geometric means.

The assay is considered to have acceptable Intra-Operator variation if all samples have a %GCV less or equal than 50%. The assay has also acceptable Intra-Operator variation from 1:20 (neat) to 1:2560 for 20/228, 20/224 and 20/222 (Figure 3, Table 3).

3.4.3 Intermediate precision

The Intermediate Precision (IP) is assessed by calculating the %GCV of the four determinations for each dilution. The assay is deemed to be precise if all estimates of intermediate precision have %GCV less or equal than 50%. The results reported in Table 4 demonstrate that the assay is precise from 1:20 (neat) to 1:2560 for high, medium, and low serum.

3.4.4 Format variability

The Format Variability (FV) measures the variability of a bioassay and is expressed as %GCV. The FV has a strong relationship with Critical Fold Difference (CFD) parameter given by $\mathit{CFD}={(1+\mathit{FV})}^{q_{\alpha }}$, where qα is the quantile from the t-distribution with α = 0.01 and infinite degrees of freedom, which is well and properly approximated by the standard normal percentile zα = 2.326348. CFD is set to 4, and therefore the assay will be considered to have acceptable FV if all samples met %GCV ≤ 81.5%. The results shown in Table 5 indicate that the assay has acceptable FV from 1:20 (neat) to 1:2560 dilution for all control sera.

3.5 Linearity

Dilution linearity is a critical evaluation conducted to ensure that a sample with a high concentration can be diluted to a concentration within the working range and still provide a reliable result. This test determines the linearity of the dose-response of the analyte, which is fundamental to ensuring that sample dilution does not affect accuracy and precision. Acceptable linearity of the test is achieved when the absolute value of the slope falls within the range of 0.7 to 1.3. The absolute slope values of samples 20/228, 20/204, and 20/222 were 1.1, 1.0, and 0.8, respectively as showed in Figure 4 and Figure 5.

Figure 4. Linear regression lines for 20/228 high, 20/204 middle and 20/222.

The proposed graph show linear regression slope for 20/228 high, 20/204 medium and 20/222 Lassa serum. Logarithm (base 2) of the dilution is indicated on the x-axis and the logarithm (base 2) of the obtained GMT on the y-axis.

Figure 5. In-vitro inhibition of Lassa panel 21/332 (20/228, 20/224 and 20/222) antisera at different serum serial starting dilution.

Dots represent % neutralisation average between the 4 replicates, error bars indicate standard deviation. Neutralization curves of serially diluted two-fold from a starting dilution of 1:20 to 1:2560. For all plots, each point represents the mean of two replicates per dilution.

3.6 Specificity

Specificity is the ability of the assay to detect and differentiate the analyte of interest.^23,24,27 To determine the specificity of the assay, all 5 sera from the Lassa panel, positive versus heterologous viruses and negative donors were tested. As visualised in Figure 6 and Table 6, 20/228 showed a fold difference of 56 versus heterologous virus serum and negative donors; 20/204 a fold difference of 37; 20/222 a fold difference of 16; 20/246 a fold difference of 8 and 20/248 a fold difference of 5. All 5 components of the panel are specific for this assay.

Figure 6. Specificity results for the full Lassa panel: 20/228, 20/204, 20/222, 20/246 and 20/202.

All 5 sera from the Lassa panel, positive versus heterologous viruses and negative donors were tested against LASV pseudoviruses as indicated in x axes. Each symbol represents a single IC50 value for the different sera, Each serum was tested in duplicated in two different days.

The assay will be considered specific if Lassa serum have at least a four-fold difference (i.e., ≥ 4-fold) to the heterologous panel and negative panel.