Novel assay for determining infectious virus titer

Abstract

[00121] The present invention relates to a label-free, high throughput method for determining an infectious virus titer in an aqueous sample comprising infectious virus, wherein the method comprises detecting the kinetics of morphological cell alteration, such as cell rounding, following viral infection, determining a parameter defining the curve of the morphological cell alteration as a function of time post infection and determining the infectious virus titer based on a standard curve of the infectious virus at a known infectious titer.

Description

NOVEL ASSAY FOR DETERMINING INFECTIOUS VIRUS TITERFIELD OF THE INVENTION

[0001] The present invention relates to a label-free, high throughput method for determining an infectious virus titer in an aqueous sample comprising infectious virus, wherein the method comprises detecting the kinetics of at least one morphological cell alteration, such as cell rounding, following viral infection, determining a parameter defining the curve of the at least one morphological cell alteration as a function of time post infection and determining the infectious virus titer based on a standard curve of the infectious virus at a known infectious titer.BACKGROUND OF THE INVENTION

[0002] In addition to classical vaccination approaches virus-based therapeutics are gaining importance. In addition to, e.g., adeno-associated virus (AAV) based gene therapy and others, the importance of vesicular stomatitis virus (VSV) is increasing with rVSV-ZEBOV recently being approved as a vaccine against Ebola (Piszczatoski, C.R. and Gums, J.G., 2020, J. Pharm. Technol. 36, 243-250; Saphire, E.O., 2020, Cell 181 , 6) and different oncolytic approaches, such as VSV-IFNp-NIS (Merchan, J.R. et al., 2020, J. Clin. Oncol., 38, 3090) and VSV-GP (Muik, A. et al., 2014, Cancer Res 74, 3567-3578).

[0003] Virus infectivity is a critical quality attribute and often the dose-defining parameter for viral therapeutics. A fast, precise, and accurate method to measure infectious virus titers is therefore one of the key requirements to develop and manufacture such virus-based drugs and finally to ensure patients’ safety. Physico-chemical analytical methods such as nanoparticle tracking analysis (NTA), dynamic light scattering (DLS), PCR-based methods or enzyme-linked immunosorbent assays (ELISA) are usually fast and easy to apply and can be used to determine concentrations of viral particles or genomic titers. These assays, however, cannot be used to determine viral infectivity as they do not reflect viral functionality and fitness in the complex interplay that allows the virus to enter its target cell, replicate, and spread its progeny. Usage of cell-based assays is therefore the gold standard to determine infectious virus titers. Two classical and well-used examples of these cell-based infectivity assays are the plaque assay and the TCID50 assay.

[0004] In the plaque assay, adherently growing target cells, that are typically seeded in 12- or 24- well plates, are infected with different virus doses. After few hours in which the virus enters the cells, an overlay medium with high viscosity is added to the wells, which prevents newly formed progeny virus to float through the culture media and infect distant cells. Virus spread therefore only happens from the initially infected target cell to directly neighboring cells. This cell-cell-spread results in a so-called plaque, describing a roughly round accumulation of infected cells, that resulted from a single infection event. When titrating viruses with a non-lytic replication cycle, the plaques need to be visualized by virus-specific staining. In contrast, lytic virus spread manifests itself in plaques, that are formed by lysed cells. In either case, the number of plaques per well needs to be counted and can finally be converted into an infectious unit expressed as plaqueforming unit (pfu) per volume.

[0005] Also, the tissue culture infectious dose 50% (TCID50) assay is performed on adherently growing target cells, but typically in 96-well plates. In this end-point dilution assay, serial dilutions of the virus material are added to the culture wells in multiple replicates. Viral replication and spreading infection throughout the well is allowed to happen freely for a sufficient duration, after which infected wells can be unambiguously discriminated from uninfected wells. The TCID50 is defined as the sample dilution factor, at which half of the inoculated wells are infected, whereas the other half of the wells show no sign of viral replication. After classifying all wells of the inoculated cell culture plate for presence or absence of viral replication, the infectious virus titer, expressed as TCID50 per virus volume, can be calculated using the mathematical models of Spearman / Kaerber or Reed / Muench (Karber, G., 1931 , Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche, Naunyn-Schmiedebergs Arch, fur Exp. Pathol. Pharmakoi. 162, 480-483; Reed, L. J. and Muench, H., 1938, A simple method of estimating fifty per cent endpoints, Am. J. Epidemiology 27, 493-497; Spearman, C., 1908, The method of ‘right and wrong cases’ (‘constant stimuli’) without Gauss’s formulae, Br. J. Prychol., 2(3), 227-242). This assay is frequently used for determining viral titers, such as in EP 1 217 891 B1.

[0006] Infectivity assays, such as TCID50 or plaque assays, are therefore very laborious, have low precision, require a large number of infected wells (due to replicates and end-point dilution) and manual read-out (such as using a binocular) and are time-consuming as they need up to several weeks. Thus, these assays are typically not suitable for automated read-out and / or high throughput screening of different parameters and conditions. Recently, it has been reported that hands-on time for plaque assays is reduced by automated plaque counting of fluorescently labeled infected cells (Masci, A.L. et al., 2019, Mol. Ther. - Methods Clin. Dev. 14, 270-274). ForVSV it has been reported that TCID50 determines the virus titer more robustly than the plaque assay (Elahi, S.M. et al., 2019, J. Virol. Methods 275, 113758).

[0007] We have previously shown that the time-consuming step of evaluating infection in all inoculated wells of TCID50 assay plates can be automated with algorithm-based analyses of label-free images (Hochdorfer D. et al., 2022, J. Virol. Methods 299, 114318). Nevertheless, the classical cell-based end-point infectivity assays still suffer from low sample throughput and limited precision, which often needs to be overcome by an increased number of assay repetitions. This creates an analytical bottleneck limiting the quality and efficiency of studies, especially those with high sample numbers.

[0008] Thus, there remains a need for a faster, less laborious (less hands-on time = less researchers required) robust and precise assay for determining infectious viral titer that allows for more advanced, faster and higher throughput measurements with increased precision in order to improve process development for viral therapeutics, such as VSV-GP.

[0009] To overcome this bottleneck, we developed a novel setup of cell-based infectivity assays, in which infectious titers can be determined with high throughput based on kinetic monitoring of infection-induced morphological cellular changes.SUMMARY OF THE INVENTION

[0010] The present invention provides a semi-automated label-free viral kinetic infectivity assay with an automated evaluation based on a morphological cell alteration, such as cell rounding, analysed in bright-field images from a microtiter plate reader. The kinetic infectivity assay determining cell rounding (also referred to as RT50 (rounding time 50 %) assay) determines infectious virus titers, such as infectious virus titers of VSV-GP and variants thereof. The kinetic infectivity assay (or kinetic infectious virus titer assay (KIT assay)) is a cell-based method in which microscopic (bright-field) images of infected cells are acquired at different times after infection. An algorithm determines, e.g., cell rounding, which occurs early after infection and correlates with the virus dose used for infection. The RT50 values (time after infection at which the half-maximal extent of cell rounding is reached) or another kinetic parameter of different doses of a well-defined standard virus material are used to generate a standard curve. Based on this standard curve the titers of the virus (e.g., VSV-GP) samples can be interpolated.

[0011] The present invention relates to a method for determining the infectious virus titer in an aqueous sample comprising infectious virus, comprising (a) providing a plurality of aqueous samples comprising infectious virus; (b) providing at least one multiwell plate comprising 96 or more wells; (c) seeding adherent cells susceptible for infection with said virus at a set cell seeding density (CSD) into the wells of the at least one multiwell plate and culturing the adherent cells; and (d) analysing the plurality of aqueous samples comprising: (di) infecting said adherent cells of step (c) by contacting the adherent cells with each of the plurality of aqueous samples of step (a) in a separate well or replicates thereof; (dii) detecting the kinetics of at least one morphological cell alteration after infection in each well over time, preferably wherein the at least one morphological cell alteration comprises cell rounding; (diii) determining for each well or replicates thereof a parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i), wherein said parameter is preferably time post infection until % maximal extent of the at least one morphological cell alteration is achieved (time post infection % maximal extent of morphological cell alteration) or the area under the curve (AUC); and (div) determining the infectious virus titer for each of the plurality of aqueous samples analysed by comparing said parameter determined for each well or replicates thereof with the same parameter in a standard curve of a set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer, wherein the standard curve correlates the parameter with the infectious virus titer.

[0012] In certain embodiments the known infectious virus titer of the reference standard is provided as tissue culture infectious dose 50 (TCID50), multiplicity of infection (MOI) or plaque forming unit (PFU). The method according to the invention is further a label-free method.

[0013] In certain embodiments of the method, each of the plurality of aqueous samples comprising infectious virus of step (a) are analyzed in at least one assay dilution, preferably wherein (i) each of the plurality of aqueous samples are analysed in no more than three assay dilutions, more preferably at a single assay dilution; and / or (ii) wherein said at least one assay dilution of the plurality of aqueous samples has an infectious virus titer within the range of the standard curve. In a preferred embodiment step (di) comprises infecting said adherent cells of step (c) by contacting the adherent cells with (i) each of the plurality of aqueous samples of step (a) in replicate wells, preferably triplicates; and / or (ii) each of the set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer in replicate wells, preferably triplicates.

[0014] Further, detecting the kinetics of at least one morphological cell alteration after infection in each well over time in step (dii) comprises automated image acquisition and analysis. The automated image analysis may comprise bright-field image acquisition, dark-field image acquisition or phase contrast image acquisition, preferably the automated image analysis comprises bright-field image acquisition, more preferably high contrast brightfield image acquisition.

[0015] In a preferred embodiment, the at least one morphological cell alteration comprise cell rounding, preferably wherein the at least one morphological cell alteration comprise cell rounding and the detection comprises automated image acquisition and automated image analysis comprising counting cells with circular shape per well and / or per total number of analyzed cells. In certain embodiments the at least one morphological cell alteration after infection in step (dii) comprises (i) cell rounding and (ii) a further morphological cell alteration.

[0016] The adherent cells susceptible for infection with said virus allows viral entry and viral gene expression, optionally also replication of said virus, preferably the virus is a cytopathic virus. The adherent cells may be, e.g., a rodent cell, a monkey cell, an avian cell, a sheep cell or a human cell. Preferably, the adherent cell is selected from the group consisting of BHK-21 cells, Vero cells, HEK293 cells, HEK293T cells, DF-1 cells, LMH cells, HeLa cells, A549 cells, MA-104 cells, MDOK cells, duck embryonic fibroblasts (DEF cells), Huh7.5 cells and COS7 cells. Further, the virus is preferably selected from the group consisting of Vesicular stomatitis virus (VSV), Newcastle disease virus (NDV), Orf-Virus (ORFV, parapoxvirus ovis), Egg drop syndrome virus (EDSV; Duck Atadenovirus A), human cytomegalovirus (HCMV), ZIKA virus (ZIKV), vaccinia virus, herpes simplex virus (HSV) and derivatives thereof. In certain embodiments, the adherent cell is a BHK-21 cell or a HEK293T cell and the virus is VSV or a derivative thereof, such as VSV- GP. In certain embodiments, the adherent cell is an avian cell, preferably a DF-1 cell, and the virus is NDV or a derivative thereof. In certain embodiments, the adherent cell is a sheep or human cell, preferably a MDOK cell or a human cell, such as a HeLa cell or a A549 cell, and the virus is a ORFV or a derivative thereof.

[0017] In certain embodiments the parameter is time post infection % maximal extent of morphological cell alteration and time post infection % maximal extent of morphological cell alteration is calculated based on a non-linear fit, preferably a non-linear kinetic fit model. In certain embodiments the standard curve correlating the parameter (e.g., time post infection % maximal extent of morphological cell alteration or AUG) with the infectious virus titer is reflected by a nonlinear fit.

[0018] The plurality of aqueous samples are preferably analyzed in at least one assay dilution. In certain embodiments, the at least one assay dilution of the plurality of aqueous samples and / or the reference standard are prepared with a pipetting robot and / or infecting the adherent cells in step (di) is performed with a pipetting robot.

[0019] In a further aspect, the invention relates to a composition or an aqueous sample comprising an infectious virus at a titer determined by the method according to the invention.DESCRIPTION OF THE FIGURES

[0020] Figure 1 : (A) Process flow chart for the kinetic infectivity assay for VSV. (B) Exemplary layout of a kinetic infectivity assay plate. A deep well plate is used to prepare assay dilutions for 24 samples (1-24) and the assay control, a standard curve and negative control. From the deepwell plate the assay dilutions of the samples, standard and controls are transferred to the assay plate containing BHK-21 cells in triplicates with alternating positioning, as indicated. S = standard (serial dilution of standard material), m = mock (uninfected control cells), AC = assay control (positive control), 1-24 (assay dilutions of samples).

[0021] Figure 2: Representative images of BHK-21 cells infected with different doses (MOI = multiplicity of infection) of VSV-GP. Images were acquired at 12 hours post infection. Bright field images are shown in the left panel demonstrating a comparable cell density. In the right panel cells that are classified by the algorithm as rounded are highlighted by white frames, resulting in an overall brighter appearance of the images with increasing dose, demonstrating that cell rounding increases with higher doses.

[0022] Figure 3: Data analysis workstream to convert kinetic values of % rounded cells into an infectious virus titer. Top to middle graph: Based on a kinetic fit model applied to Rounded Cells (%) plotted against hours post infection (h.p.i), RT50 (rounding time 50%) values are determined and plotted against the applied MOIs to generate a standard curve. Bottom graph: Unknown MOIs of samples (square) and assay control (AC, triangle) are interpolated from the standard curve based on the RT50 values determined for samples and assay control. Bottom: Back-calculation of infectious titers (TCID50 / ml) based on interpolated MOIs.

[0023] Figure 4: Kinetics of cells rounding in BHK-21 cells infected at different MOIs with VSV- GP. Each point represents the mean percentage of rounded cells in eight replicate wells.

[0024] Figure 5: Extended standard curve with RT50 values (time post infection until half-maximal cell rounding is achieved) determined via kinetic fit model from BHK-21 cells infected in eightreplicate wells for each of the indicated MOIs. RT50 values were calculated based on the data shown in Figure 4.

[0025] Figure 6: Comparison of RT50 standard curves from routine assays performed by different operators over a course of three months.

[0026] Figure 7: Intra-plate precision of the kinetic infectivity assay. Kinetic infectivity assay was used to determine infectious virus titers of 65 different VSV-GP samples. Each sample was tested twice in triplicate wells on the same plate. The two measured titers (black dots) as well as their mean value (grey line) are shown.

[0027] Figure 8: Inter-plate precision of the kinetic infectivity assay. Infectious titers (black dots) of the internal assay control determined in different routine assays are shown. The assay control is handled exactly like a sample. The solid line represents a best-fit line through all datapoints. The dotted lines show the 95 % confidence bands of the best-fit line.

[0028] Figure 9: Comparison of assay variability between TCID50 and kinetic infectivity assays. Box-plots show the median value as well as the first (lower end of the box) and third quartile (upper end of the box), containing 50 % off all data points. Whiskers indicate the 2.5 % / 97.5 % percentile. 296 measurements over more than two years are shown for the TCID50 assay. 16 measurements over a course of about three months are shown for the kinetic infectivity assay. The dotted lines indicate the ± 0.5 Iog10 values of the TCID50 assay mean, representing assay limits proposed by ICH guidelines.

[0029] Figure 10: Kinetic infectivity assay performed with VSV-GP and VSV-GP-Cargo. A) Kinetics of cell rounding in BHK-21 cells infected at different MOIs with either variant of VSV-GP. B) Correlation between the MOI used for infection and RT50 values determined from the kinetic curves shown in A.

[0030] Figure 11 : Cell rounding is specific to viral entry and gene expression. (A) Cell rounding kinetics determined in BHK-21 cells infected with VSV-GP or a variant thereof with a mutated M protein (VSV-GP AM51) at an MOI of 1. Where indicated, viral entry was inhibited by treatment with a neutralizing antibody (NAb) (10 pg / ml) or viral RNA transcription was inhibited by treatment with the polymerase inhibitor Vesiculopolin A (VPI A) (7.5 pM). Each point represents the mean of three wells. (B) BHK-21 cells were infected at an MOI of 0.32 with a VSV-GP variant expressing the red fluorescent Katushka protein. Cell rounding was determined in bright-field images. In the same field of view, fluorescence was determined with an excitation / emission wavelength of 584±20 nm / 625±20 nm. Rounded cells (filled squares) and Katushka positive cells (filled dots)are both expressed as percentage of the total number of cells in the bright field image. Each point represents the mean of three wells. A second plate was infected and incubated in parallel and used to harvest supernatants at 0, 8, 12, 18 and 24 h.p.i. Genomic virus titers (filled triangles) were determined by qRT-PCR to monitor viral replication and are expressed as genomic copies (GC) / ml. A non-linear 4 parameter logistic model was applied to fit the curve. (C) BHK-21 cells were infected at an MOI of 10 and brightfield and fluorescence images were acquired and analysed as described for (B).

[0031] Figure 12: The kinetic infectivity assay is stability-indicating. (A) Forced sample degradation by the indicated stress conditions (45°C for 10, 30, 120 and 240 min or UV irradiation using 3.9, 39 or 397 mJ / cm2) resulted in a delay of time-dependent rounding of BHK-21 cells infected with VSV-GP. Equal sample volumes were used for all infections, which correspond to a MOI of 1.78 for the untreated sample. (B) The kinetic infectivity assay was used to determine infectious virus titers of the samples shown in (A) relative to a reference standard. At the highest dose of UV-irradiation no virus-dependent cell rounding occurred (curve indistinguishable from mock) and the infectious virus titer is thus below limit of quantitation (LOQ).

[0032] Figure 13: Comparison of performance characteristics of different infectivity assays. The hands-on time and number of processed multiwell plates represent values for measurement of 24 samples. Variability of the TCID50 assays refers to a single determination and can be reduced by averaging result from multiple measurements.

[0033] Figure 14: Kinetic infectivity assay performed with Newcastle disease virus (NDV) in chicken embryo fibroblasts (DF-1). (A) Kinetics of cell rounding in DF-1 cells infected at different MOIs with NDV. Error bars show the standard deviation between triplicate wells. (B) Area under the curve (AUC) was calculated from the curves shown in A. AUC of the uninfected mock control was subtracted as background from the AUC of all other curves. A non-linear 4 parameter logistic fit was used to model the dose-response curve. Errors bars show the standard deviation between AUC calculated from triplicate wells.

[0034] Figure 15: Application of the kinetic infectivity assay to determine infectious titers of Newcastle disease virus (NDV) under optimized cell-growth conditions. (A) Time-dependent cell rounding after infection of DF-1 chicken fibroblasts with NDV at the indicated MOI. Each point represents the mean of three wells. (B) RT50 values of the infections shown in (A). A sigmoidal four parameter logistic fit was applied to the RT50 values to generate a standard curve. RT50values and the interpolated MOIs of three samples (M1-M3) tested at four different concentrations are depicted on the standard curve. Refer to Table 2 for the tabulated results of the experiment.

[0035] Figure 16: Kinetic infectivity assay performed with parapoxvirus ovis (ORFV) in HeLa cells. (A) Time-dependent cell rounding after infection of HeLa cells with ORFV at the indicated MOI. Each point represents the mean of three wells. (B) Area under the curve (AUG) was calculated for each of the curves shown in (A) and plotted against the respective MOI used for infection. A non-linear 4 parameter logistic fit was used to model the dose-response curve. Error bars show the standard deviation between AUG calculated from triplicate wells.

[0036] Figure 17: Kinetic infectivity assay performed with parapoxvirus ovis (ORFV) in A549 cells. (A) Time-dependent cell rounding after infection of A549 cells with ORFV at the indicated MOI. Each point represents the mean of three wells. (B) Area under the curve (AUG) was calculated for each of the curves shown in (A) and plotted against the respective MOI used for infection. A non-linear 4 parameter logistic fit was used to model the dose-response curve. Error bars show the standard deviation between AUG calculated from triplicate wells.DETAILED DESCRIPTION OF THE INVENTION

[0037] Virus infectivity is a critical quality attribute and often the dose-defining parameter for viral therapeutics. The present invention provides a fast, precise and accurate high throughput infectivity assay for determining infectious virus titer in sample during development and market supply of therapeutic virus-based biologies. The assay is a method that is based on kinetic monitoring of virus-induced morphological cell alterations that occur early after infection and correlates with the virus dose used for infection. Since these morphological cell alterations occur early during the viral infection cycle the method is faster compared to conventional methods for infectivity assays (such as TCID50 and plaque assay). Moreover, detection of the morphological cell alterations, including image acquisition and analysis, is automated and most pipetting steps (seeding, dilution and / or infection) can be performed using a pipetting robot, which significantly reduces the hands-on time and turnaround time needed to perform the assay and standardizes the assay.

[0038] The present invention provides a method for determining the infectious virus titer in an aqueous sample comprising infectious virus, the method comprising (a) providing a plurality of aqueous samples comprising infectious virus; (b) providing at least one multiwell plate comprising96 or more wells; (c) seeding adherent cells susceptible for infection with said virus at a set cell seeding density (CSD) into the wells of the at least one multiwell plate and culturing the adherent cells; and (d) analysing the plurality of aqueous samples comprising: (di) infecting said adherent cells of step (c) by contacting the adherent cells with each of the plurality of aqueous samples of step (a) in a separate well (or in replicates thereof); (dii) detecting the kinetics of at least one morphological cell alteration after infection in each well over time; (diii) determining for each well (or replicates thereof) a parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i), wherein said parameter is preferably time post infection until % maximal extent of the at least one morphological cell alteration is achieved (time post infection % maximal extent of morphological cell alteration) or the area under the curve (AUC); and (div) determining the infectious virus titer for each of the plurality of aqueous samples analysed by comparing said parameter determined for each well (or replicates thereof) with the same parameter in a standard curve of a set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer, wherein the standard curve correlates the parameter with the infectious virus titer.

[0039] Step (dii) comprises detecting the kinetics of at least one morphological cell alteration after infection in each well over time. The person skilled in the art will understand that the at least one morphological cell alteration after infection is at least one virus-induced morphological cell alteration after infection. Typically, one morphological cell alteration after infection is detected, however, theoretically also two or three different morphological cell alterations may be detected in parallel for each well over time. The detection step comprises image acquisition over a time course post infection and image analysis of the percentage of morphologically altered cells in each well at a specific point of time after infection. The outcome of the detection step is the percentage of morphologically altered cells (percent of total number of analyzed cells) in each well at a specific point of time after infection and detected over time and hence the kinetics of the percentage of morphologically altered cells in each well over the time course post infection. The term “over time” as used herein means over a time course post infection. This means that the kinetics of the at least one morphological alteration after in infection is detected in living cells. Moreover, the term “detecting the kinetics” as used herein means detection of the at least one morphological cell alteration over a time course post infection and preferably includes detection at 10 or more, preferably 15 or more, more preferably 20 or more different time points over a time course post infection.

[0040] Suitable morphological cell alterations after infection include, without being limited thereto, cell rounding (circularity), granulation, texture, area per cell and / or cytoplasmic vacuolation, preferably cell rounding and / or granulation and / or area per cell, more preferably cell rounding. The term “morphological cell alteration” as used herein refers to a virus-induced morphological cell alteration, preferably an early virus-induced morphological alteration of individual cells, which correlates with the virus titer used for infection. The morphology of the cell includes its shape and appearance and the morphological cell alteration is detectable by label-free morphological analysis, preferably automated, algorithm-based morphological analysis. It does not include virus specific staining or metabolic cell viability assays. The advantage of detecting early morphological cell alterations after viral infection is a reduced assay duration. The advantage of detecting morphological alterations of individual cells is that it allows the analysis of the percentage of morphologically altered cells in each well, i.e., a value relative to the total number of cells analysed, resulting in a higher assay accuracy because small variations in cell density per well do not affect the percentage. Thus, in certain embodiments, the method comprises detecting the percentage of morphologically altered cells in each well and / or detecting at least one early morphological cell alterations. In certain embodiments, the at least one morphological cell alteration is not a visible cytopathic effect (CPE). The term “automated” as used herein means the use of technologies that reduces human intervention in a process, such as image acquisition at a pre-set time interval by the image reader and / or algorithm-based analysis of images for a certain parameter, such as a morphological cell alteration, including without being limited thereto, cell rounding. Preferably, the at least one morphological cell alteration comprises cell rounding. In certain preferred embodiments the morphological cell alteration is cell rounding. In alternative embodiments the at least one morphological cell alteration after infection comprises (i) cell rounding and (ii) a further morphological cell alteration, such as granulation. Detection of the kinetics of the at least one morphological cell alteration after infection and hence the method according to the invention is label-free. Label-free detection reduces the pipetting steps, which reduces hands-on time and increases accuracy due to reduced variations. Thus, although label- free detection is advantageous, the method may be further combined with detection methods requiring a label, such as a fluorescent label. The hand-on time is the sum of time required to complete cell seeding, sample preparation, infection and device loading.

[0041] The term “cell rounding” as used herein refers to abundance of rounded cells that increases following viral infection. The kinetic of cell rounding can be detected over time following infection (time course post infection). Cell rounding is observed already very early after infectionwith most viruses, such as VSV and NDV. It is most likely the result of virus-induced subversion of the cellular cytoskeleton and / or represents the beginning of infection-dependent cytopathic effects (CPE). An image analysis tool identifies all cells within a pre-defined area of the well and then determines whether the cell is rounded or not. A cell may be defined as rounded, if the ratio of its smallest diameter divided by its largest diameter is greater than 0.3. This cutoff value was determined for VSV using BHK-21 cells and might have to be adapted if other virus / cell combinations are used in the assay. The outcome of the image analysis is the percentage of rounded cells in each well at a specific point of time after infection. Since the image acquisition and analysis happens over a time course post infection, the detection step provides the kinetics of cell rounding post infection.

[0042] In certain embodiments detecting the kinetics of at least one morphological cell alteration after infection in each well over time in step (dii) comprises (i) detecting for about 24 h to about 72 h post-infection (e.g., 24 h to about 50 h post-infection), and / or (ii) starting detecting immediately after infection, and / or (iii) detecting the at least one morphological cell alteration in regular intervals, preferably in regular intervals of about every hour or less. The length of the time over which the kinetics of the at least one morphological cell alteration is detected depends on the virus. For VSV the time course post infection is typically about 24 to about 30 hours, while for NDV the time course post infection is typically about 40 to about 80 hours (e.g., about 50 h to about 72 h). Preferably detection starts immediately after infection, i.e., between 0 and 1 hour post infection (h.p.i), more preferably within 30 min post infection, even more preferably within 15 min post infection and detection (image acquisition and analysis) is at regular intervals, i.e., about every hour or less. Detecting the kinetics of at least one morphological cell alteration (e.g., cell rounding) after infection in each well over time in step (dii) comprises image acquisition and analysis, preferably automated image acquisition and analysis. Automated image analysis means algorithm-based analysis. More preferably the at least one morphological cell alteration comprises or is cell rounding and detecting the kinetics of at least one morphological cell alteration after infection in each well over time comprises determining the percent of rounded cells, i.e., counting cells with circular shape per well and / or per total number of analyzed cells, preferably per total number of analysed cells (percent circular shape of total number of cells analysed). Software for label-free cell morphology analysis to be used in combination with an imaging reader are known in the art and commercially available and include Gen5 IVD software (Agilent BioTek) or I ncucyte® Advanced Label-Free Classification Analysis Software Module.

[0043] The image analysis may comprise bright-field image acquisition, dark-field image acquisition or phase contrast image acquisition, preferably bright-field image acquisition, more preferably high contrast brightfield image acquisition. The person skilled in the art will understand that the image analysis is live cell analysis. Preferably, the analysis system allows parallel processing of multiple multiwell plates (such as up to 8 or more multiwell plates).

[0044] Step (diii) of the method according to the invention comprises determining for each well (or replicates thereof) a parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i). Preferably, said parameter is time post infection until % maximal extent of the at least one morphological cell alteration is achieved (time post infection % maximal extent of morphological cell alteration) or the area under the curve (AUG), more preferably said parameter is time post infection % maximal extent of morphological cell alteration. The % maximal extent is a pre-defined % of the detected maximal extent of the at least one morphological cell alteration, such as 50% maximal extent of the at least one morphological cell alteration. The “% maximal extent” may also be referred to as “% maximal response”, such as for 50 % a “half maximal response of the at least one morphological cell alteration” or “half maximal response of cell rounding”. While the parameter time post infection % maximal extent of morphological cell alteration is typically determined as time post infection of half maximal extent (50% maximal extent) of morphological cell alteration, theoretically, the time post infection of, e.g., 40% maximal extent of morphological cell alteration may be determined and would provide similar results. The % maximal extent should be in the range of 20-80%, preferably 30-70%, more preferably 40-60% even more preferably about 50%. It should be chosen to be within the linear range of the sigmoidal curves of morphological cell alteration as a function of time post infection. In case the at least one morphological cell alteration is cell rounding the parameter is preferably time post infection % maximal extent of cell rounding (e.g., t.p.i half maximal extent of cell rounding). The time post infection half (50%) maximal extent of cell rounding is also referred to as “rounding time 50%” or “RT50” herein. The person skilled in the art will understand (based on the fit model used for calculation) that the maximal extent of morphological cell alteration is the maximal response over baseline of morphological cell alteration per well (e.g., rounded cells per well). Also, the maximal extent of the morphological cell alteration can depend on the virus dose used for infection and can therefore vary between different MOIs.

[0045] The curves of morphological cell alteration (e.g., cell rounding) as a function of time post infection describes a baseline morphological cell alteration, such as cell rounding (cell roundingprior to infection or in mock infected cells), followed by a strong increase in morphological cell alteration, such as in cell rounding, until a dose-dependent or dose-independent peak level is reached (plateau) and hence usually forms a sigmoidal curve. In case the time post infection % maximal extent of morphological alteration cannot be determined, e.g., because the curve is not a sigmoidal curve, due to a lack of the strong increase in morphological cell alteration (i.e., cell rounding) and / or the curve shapes are heterogeneous, the parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i) may be the area under the curve (AUC) or another kinetic parameter. Thus, in certain embodiments the parameter is the AUC. The term “area under the curve” or “AUC” as used herein is the definite integral of the % morphologically altered cells as a function of time post infection for each well.

[0046] In some embodiments, the parameter defining the curve of the at least one morphological cell alteration as a function of time post infection, such as time post infection % maximal extent of morphological cell alteration, is calculated based on a non-linear fit or a parallel line assay, more preferably based on a non-linear fit, even more preferably a non-linear kinetic fit model.

[0047] The term “parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i)” also referred to as “parameter” herein, refers to any (single) parameter defining the curve that correlates with the applied virus dose (titer) and wherein a reference standard of known infection titers provides a standard curve. Preferably, the standard curve correlating the parameter (e.g., time post infection % maximal extent of morphological cell alteration or AUC) with the infectious virus titer is reflected by a non-linear fit or a parallel line assay, more preferably a non-linear fit model. The parameter allows comparing the curves obtained for different samples and the reference standard and to apply the same kinetic fit mode. In certain embodiments the standard curve correlating the parameter (e.g., time post infection % maximal extent of morphological cell alteration or AUC) with the infectious virus titer is reflected by a non-linear fit or a parallel line assay, preferably based on a non-linear fit. In certain embodiments the parameter is time post infection % maximal extent of morphological cell alteration and time post infection % maximal extent of morphological cell alteration is calculated based on a non-linear fit, or a parallel line assay, preferably based on a non-linear fit, more preferably a non-linear kinetic fit model.

[0048] For example, the determining step (diii) for detecting cell rounding as morphological cell alteration and determining rounding time 50% as parameter may be performed as follows: The outcome of the image analysis is the percentage of rounded cells in each well at a specific point of time after infection. If these two factors are plotted against each other with time post infectionon the x-axis and % rounded cells on the y-axis, this results in a sigmoidal curve, which is shifted along the x-axis (representing the time post-infection) depending on the virus dose used for infection. A kinetic curve fit modelas described in more detail in the examples is used to determine the rounding time 50% (RT50) value, i.e., the time after infection (h.p.i) at which the half-maximal cell-rounding is reached, from each of the sigmoidal curves.

[0049] Step (div) of the method according to the invention comprises determining the infectious virus titer for each of the plurality of aqueous samples analysed by comparing said parameter determined for each well (or replicates thereof) with the same parameter in a standard curve of a set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer, wherein the standard curve correlates the parameter with the infectious virus titer.

[0050] The person skilled in the art will understand that the infectious virus of the reference standard comprises the same or at least a closely related virus, preferably the same virus, as in the aqueous samples. Closely related means the basically same virus differing, e.g., only in transgene expression, or basically the same virus differing, e.g., in comprising a mutation, and testing the effect of the mutation on the production of infectious virus.

[0051] Standard curve correlates the parameter with the infectious virus titer, wherein the parameter may be, e.g., time post infection until % maximal extent of the at least one morphological cell alteration is achieved or the AUC, and is generated using a set of serial dilutions of said infectious virus at a known infectious titer. Wherein the known infectious titer may be determined using a classical assay, such as a TCID50 assay or a plaque assay, or may be determined using the method according to the invention. The known infectious virus titer of the reference standard may be provided as tissue culture infectious dose 50 (TCID50), multiplicity of infection (MOI) or plaque forming unit (PFU). The person skilled in the art will understand that the infections virus titer determined by the method according to the invention has the same unit, i.e., tissue culture infectious dose 50 (TCID50), multiplicity of infection (MOI) or plaque forming unit (PFU), respectively, as the known infectious titer of the reference standard.

[0052] The set of dilutions of the reference standard comprises at least three dilutions, preferably at least 4 dilutions, more preferably at least 5 dilutions; (ii) is a set of serial dilutions of thereference standard, preferably a set of 1.5 to 4-fold serial dilutions, more preferably a set of 2 to 3.5 fold serial dilutions, of the reference standard; and / or (iii) comprises an infectious virus titer range of at least 1.5 log-io, preferably of 1.5 to 4 log-io steps. The reference standard may be analysed on each of the at least one multiwell plates or in case more than one multiwell plates are analysed simultaneously on at least one of the more than one multiwell plates. Theoretically, the reference standard can also be analysed separately, as long as cell culturing and infection parameters are standardized with neglectable variations.

[0053] Detecting the kinetics of the at least one morphological cell alteration (step dii), determining for each well the parameter (step diii) and determining the infectious titer (step div) do not require a label (label-free), such as a fluorescent label. The method according to the invention is therefore a label-free method and uses label-free, automated and algorithm-based analysis of a morphological cell alteration. It does therefore not require additional handling steps for staining.

[0054] The method according to the invention comprises comparing said parameter determined for each well (or replicates thereof) with the same parameter in a standard curve of a set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer, wherein the standard curve correlates the parameter with the infectious virus titer. Thus, in certain embodiments the method comprises comparing the determined time post infection % maximal extent of morphological cell alteration, preferably rounded cells, with the time post infection % maximal extent of morphological cell alteration in a standard curve. Alternatively, the method comprises comparing the determined AUG with the AUG in a standard curve. The standard curve then correlates the parameter, such as the time post infection % maximal extent of morphological cell alteration or the AUG with the infectious virus titer.

[0055] The method according to the invention is an in vitro method. The method according to the invention further comprises kinetic monitoring (detecting) of at least one morphological cell alteration. This means it is not an end-point assay determining the morphological alteration at one time point, but rather monitors (detects) the at least one morphological cell alteration over a time course post-infection. Moreover, the method according to the invention does not require end-point dilution of the plurality of aqueous samples. In contrast one advantage of the novel method is that the sample can be analysed in a single dilution (assay dilution) or if required two or three dilutions, preferably in a single dilution, which significantly reduces the number of multiwell plates and the hands-on time needed for the method.

[0056] The aqueous samples are typically analysed as a diluted sample (assay dilution). In certain embodiments each of the plurality of aqueous samples comprising infectious virus of step (a) are diluted to at least one assay dilution used for infecting said adherent cells in step (di). In certain embodiments the at least one assay dilution of the plurality of aqueous samples has an infectious virus titer within the range of the standard curve. Preferably no more than three, preferably no more than two, more preferably a single assay dilution of each of the plurality aqueous sample(s) comprising virus is used for infecting said adherent cells in step (di). The person skilled in the art will understand that only a single assay dilution is required if said assay dilution is within the range of the standard curve.

[0057] Step (a) comprises providing a plurality of aqueous samples comprising infectious virus. All samples, including the aqueous samples comprising infectious virus and the set of dilutions of a reference standard comprising said infectious virus, are preferably analysed in replicates, preferably duplicates, more preferably triplicates. Optionally also a negative control (i.e. a sample not comprising infectious virus; mock control) and / or a positive control (assay control) is included, which is also preferably analysed in replicates similar to the aqueous sample comprising infectious virus and / or the set of dilutions of a reference standard comprising said infectious virus. In certain embodiments, step (di) comprises infecting said adherent cells of step (c) by contacting the adherent cells with (i) each of the plurality of aqueous samples of step (a) in separate replicate wells, preferably duplicates, more preferably triplicates; and / or (ii) each of the set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer in replicate wells, preferably duplicates, more preferably triplicates. If replicates are used, the detecting in step (di) is performed in each well over time, while the determining of the parameter in step (dii) and the determining of the infectious virus titer for each of the plurality of aqueous samples may be for each well or for the replicates thereof, such as on the mean or median of the replicates.

[0058] Thus, the novel method allows analysing a plurality of aqueous samples comprising infectious virus simultaneously. In certain embodiments, the method is adapted to determine the infectious virus titer in a plurality of aqueous samples comprising infectious virus in parallel on the same plate, such as 5 or more, 10 or more, 20 or more or 24 or more aqueous samples comprising infectious virus in at least triplicates per 96 wells of said at least one multiwell plate.

[0059] The aqueous sample comprising infectious virus may be any aqueous sample comprising infectious virus. Preferably the aqueous sample is an ex vivo sample. More preferably, the aqueous sample is derived from cell culture production and may be a process development sample or an in process control sample, i.e., a harvested cell culture supernatant or cell lysate, asample taken during purification or a virus-based therapeutic sample, such as a drug substance or drug product.

[0060] The at least one assay dilution of the plurality of aqueous samples and / or the set of dilutions of the reference standard are preferably prepared with a pipetting robot.

[0061] Step (b) comprises providing at least one multiwell plate comprising 96 or more wells. Typically, multiwell plates are provided in multiples of 96 wells, such as 192 well plates or 384 well plates. A multiwell plate may also be referred to as microtiter plate herein. The assay is performed in standard multiwell plates, which are among the standard consumables in molecular biology or virology laboratories. The use of multiwell plates allows analysing aqueous samples in high throughput mode. The term “high throughput method” as used herein refers to a method that allows automated or semi-automated testing of a large number of samples (e.g., 24, 48, 72, 96, 120, 144, 168, 192 or more) in parallel. In addition to the use of multiwell plates, such as 96-well plates or preferably even plates comprising multiples of 96 wells, it further involves a short turnaround time (total assay duration) and more importantly a reasonable hands-on time per sample, multiwell plate and / or multiples of multiwell plates.

[0062] Step (c) comprises seeding adherent cells susceptible for infection with said virus at a set cell seeding density (CSD) into the wells of the at least one multiwell plate and culturing the adherent cells. A set cell seeding density (CSD) refers to a predetermined cell seeding density. Preferably the set cell seeding density is a constant seeding cell density, i.e. , identical in all wells.

[0063] Culturing the adherent cells comprises culturing the cells under appropriate conditions for growth and maintenance and for a period of time that allows the adherent cells to settle and adhere to the bottom of the multiwell plate. The plates may be coated and are typically standard multiwell plate suitable for cell culture. In certain embodiments the method comprises that the adherent cells in step (c) are cultured (i) for about 6 hours to about 30 hours prior to infection in step (di), preferably for 12-28 hour prior to infection in step (di); and / or (ii) until about 30% to about 80 % confluent prior to infection in step (di), preferably until about 40 to about 70% confluent prior to infection in step (di). The optimal cell seeding density may depend on the adherent cell. In certain embodiments the adherent cells susceptible for infection for said virus are seeded at a constant set cell seeding density into the wells of the at least one multiwell plate, and / or the cell seeding density is within a range corresponding to from about 0.5 x 104cells / well to about 2.0 x 104cells / well in a 96-well plate, more preferably from about 1.5 x 104cells / well to about 2.0 x 104cells / well in a 96-well plate or the corresponding seeding density in other multiwell plates.

[0064] Step (di) comprises infecting said adherent cells of step (c) by contacting the adherent cells with each of the plurality of aqueous samples of step (a) in a separate well (or in replicates thereof). Typically, each of the plurality of aqueous samples is analyzed in at least one assay dilution. In certain embodiments the method comprises that each of the plurality of aqueous samples comprising infectious virus of step (a) are diluted to at least one assay dilution used for infecting said adherent cells in step (di), wherein no more than three, preferably no more than two, more preferably a single assay dilution of each of the plurality aqueous sample(s) comprising virus is used for infecting said adherent cells in step (di). Preferably infecting the adherent cells in step (di) is performed with a pipetting robot.

[0065] The invention further relates to a composition or an aqueous sample comprising an infectious virus at a titer determined by the method according to the invention.

[0066] Generally, detection of the morphological cell alterations, including image acquisition and analysis, is automated and most pipetting steps (seeding, dilution and / or infection) can be performed using a pipetting robot, which significantly reduces the hands-on time needed to perform the assay and standardizes the assay.

[0067] The method according to the invention is suitable for any virus in combination with an adherent cell susceptible for infection with said virus. Infection of the adherent cell susceptible for infection with said virus typically results in at least one detectable morphological cell alteration. An adherent cell susceptible for infection with said virus means that it allows at least viral entry followed by viral gene expression. In addition, it may also allow replication of said virus. In certain embodiments, the virus is a cytopathic virus. Suitable viruses include, without being limited thereto, an oncolytic virus, a vaccine virus or a gene therapeutic virus (viral vector for gene therapy). An oncolytic virus may further comprise transgenes, such as suicidal genes, genes that suppress angiogenesis (e.g., angiostatin gene or endostatin gene) or genes that support concomitant treatment, such as sodium-iodide symporter (NIS) gene (which results in iodine accumulation in combination with radioiodine therapy)).

[0068] In certain embodiments, the virus is selected from the group consisting of Vesicular stomatitis virus (VSV), Newcastle disease virus (NDV), Orf-Virus (ORFV, parapoxvirus ovis), Egg drop syndrome virus (EDSV; Duck Atadenovirus A), human cytomegalovirus (HCMV), Zl KA virus (ZIKV), vaccinia virus, herpes simplex virus (HSV), influenza virus and derivatives thereof.

[0069] The adherent cells may be a rodent cell, a monkey cell, an avian cell, a sheep cell or a human cell, preferably the adherent cell is selected from the group consisting of BHK-21 cells,Vero cells, HEK293 cells, HEK293T cells, avian fibroblast cells (e.g., DF-1 cells or duck embryo fibroblasts (DEF)), avian hepatocytes (e.g., LMH cells), HeLa cells, A549 cells, MA-104 cells, MDOK cells, Huh7.5 cells and COS7 cells.

[0070] Adherent cells susceptible for infection with a virus of interest are known in the art and the skilled person would know the cells to use in an assay to determine the infectious titer for a certain infectious virus. For example, adherent cells susceptible for infection with VSV (and recombinant derivatives thereof, such as VSV-GP) are BHK-21 cells or HEK293T cells, adherent cells for infection with NDV are avian fibroblast cells, such as DF-1 cells and adherent cells for infection with ORFV are sheep cells (e.g., MDOK) or human cells (e.g., HeLa or A549). Infection with VSV resulting in cytopathic effects has also been observed in HCT116, HCT115, A549, HEK293, DU- 145, MCF-7, T47-D, ACHN, 786-0, OVCAR-8, NCI / ADR-RES, M14, U-251 , SF-295, LoVo, CT26.CL25, Vero, J774 and B16 cells, which are therefore also suitable adherent cells susceptible for infection with VSV.

[0071] VSV-GP is a vesicular stomatitis virus with the surface glycoprotein of the lymphocytic choriomeningitis virus. More specifically, it is a genetically modified version of the vesicular stomatitis virus (VSV) in which the original G protein was replaced by the envelope glycoprotein (GP) of lymphocytic choriomeningitis virus (LCMV). Additional cargo variants of VSV-GP have been generated, which encode transgenes with specific functions supporting the therapeutic properties when used, e.g., as oncolytic agent. Due to frequently defective antiviral immune responses, malignant cells are especially prone to infection with VSV-GP and infection leads to cell death and lysis. VSV-GP is therefore an oncolytic virus.

[0072] Although this assay workflow was optimized for VSV-GP, it can be easily transferred to other viruses, such as NDV or ORFV, as demonstrated in the Examples, particularly to cytopathic viruses.EXAMPLESMaterial

[0073] The following equipment was used for the kinetic infectivity assay: Pipetting robot for full workflow automation (Assist Plus; Integra), electronic pipet (VIAFLOW, Integra), cell counter (NucleoCounter® NC-200TM, chemometec), clinical imaging multimode reader (Cytation 5 cell imaging multimode reader, BioTek) and automated incubator (BioSpa, Agilent BioTek). The cellimaging multimode reader comprises automated microscopy and conventional microplate detection in a configurable upgradable platform for laboratory use. This microscopy module offers up to 60x magnification in fluorescence, brightfield, high contrast brightfield, color brightfield and phase contrast to address many applications and workflows. The multimode detection modules include filter- and monochromator-based fluorescent detection, luminescence, and UV-Vis absorbance detection.

[0074] Other materials: Glasgow’s minimal essential medium (GMEM; Gibco / Thermo Fisher, Cat. No. 21710082), fetal calf serum (FCS, Gibco / Thermo Fisher, Cat. No. 10500064), Tryptose phosphate broth (Gibco / Thermo Fisher, Cat. No. 18050039), Dulbeccos’ phosphate buffered saline (DPBS, no calcium, no magnesium, Gibco / Thermo Fisher, Cat. No. 14190144), TrypLE™ select enzyme (1X) no phenol red (Gibco / Thermo Fischer; Cat. No. 1256311), DMEM (Gibco #11960) and Ultroser G (Sartorius).Kinetic infectivity assay principle used to determine infectious virus titers of VSV-GP

[0075] VSV-GP is a genetically modified variant of the vesicular stomatitis virus (VSV) in which the gene encoding the surface glycoprotein G was replaced by the gene encoding the surface glycoprotein GP of lymphocytic choriomeningitis virus (LCMV) (Muik, A. et al., 2014, Cancer Res 74, 3567-3578).

[0076] For the kinetic infectivity assay, BHK-21 cells were seeded in 96-well plates. Per well 104cells were seeded in 100 pl cell culture medium using Glasgow’s minimal essential medium (GMEM) complete medium (complete, 8.7% FCS, 4.3% tryptose phosphate broth). 24±6 hours after seeding, cells were infected in triplicates with six different doses of 1 - log-io serially diluted standard virus sample, with known infectious virus titer, using GMEM as diluent. The doses of the standard curve were expressed as multiplicity of infection (MOI), defied as TCID50 per cell. The range of the standard curve covers for example MOI 10 - MOI 0.032 as shown in Table 1. Uninfected control wells (medium only) and an assay control (virus sample with known titer included on every assay plate for trending purposes) were as well added into triplicate wells. On every assay plate up to 24 samples with unknown titers can be titrated. Each sample was applied in triplicates at a single assay dilution (in GMEM), which is calculated based on the expected titer and aimed at the middle of the range of the standard curve (e.g., a target MOI of about 0.562 for the exemplified standard curve used herein). If the titer of the sample is completely unknown, different assay dilutions (such as 2 or 3) can be applied for a sample to increase the chance tomeet the range of the standard. The workflow of the assay is illustrated in Figure 1A. An exemplary plate layout is shown in Figure 1 B.Table 1 Exemplary range for virus samples used as standard in the kinetic infectivity assay

[0077] Immediately after infection, the plate (or multiple plates) was placed into an automated incubator (BioSpa) at 37°C and 5% CO2 and automated image acquisition was performed using the automated Cytation5 imaging reader. The Cytation5 imaging reader acquires a bright field image of each well of the 96-well plate every hour for a total duration of 24-32 hours and in combination with the BioSpa allows parallel reading of multiple plates (such as up to 8 microplates). The image analysis tool of the built-in Gen5 software identifies all cells within a predefined area of the well and then determines whether the cell is rounded or not (Figure 2). Cell rounding is observed already very early after infection with VSV-GP and most likely is the result of virus-induced subversion of the cellular cytoskeleton (Taylor M. et al., 2011. Subversion of the actin cytoskeleton during viral infection. Nat. Rev. Microbiol. 9, 427-439) and / or represents the beginning of infection-dependent cytopathic effects (CPE). The baseline of rounded cells, which is shared by uninfected cells (mock) and infected cells, prior to onset of infection, depends on the definition of a rounded cell. For our assay, a cell is defined as rounded, if the ratio of its smallest diameter divided by its largest diameter is greater than 0.3. This cutoff value might have to be adapted if other virus / cell combinations are used in the assay. A slight increase in rounded cells was also observed in mock infected wells, which is likely a consequence of growing cell density. This is not critical, because its extent and time-dependent progression never reached the levels of infection-induced cell rounding.

[0078] The outcome of the image analysis is the percentage of rounded cells in each well at a specific point of time after infection. If these two factors are plotted against each other with time post infection on the x-axis and % rounded cells on the y-axis, this results in a sigmoidal curve (Figure 3). The following kinetic curve fit model is used to determine the rounding time 50%(RT50) value, i.e. the time after infection (h.p.i) at which the half-maximal cell-rounding is reached, from each of the sigmoidal curves.Kinetic curve fit model:In which:LA = lower asymptote. Value is defined by the baseline of rounded cells per well.UA = upper asymptote. Value is the maximum % rounded cells, that is reached based on the defined image analysis parameters. k = slope factor at the inflection point of the curve.RT50 = time after infection at which half-maximal extent of cell rounding is reached.Inflection point of the curve.

[0079] The RT50 values of the different doses of the standard sample are plotted against the corresponding known MOIs, resulting in a standard curve (Figure 3). Based on the determined RT50 values of the assay control and the samples with unknown titer, their MOIs can be interpolated from this standard curve. The interpolated MOIs, together with the known number of cells seeded per well and the volume of the samples (defined by the assay-dilution) that was added per well, is then used to calculate the infectious titer (Figure 3) using the following calculation.Calculation of sample infectious titers:

[0080] The calculated infectious virus titer has the same unit in which infectivity of the standard sample is expressed, such as TCID50 in the present case.Example 1 : Evaluation of the kinetic infectivity assay to determine VSV-GP infectious titers

[0081] Quantification of rounded cells in kinetically acquired images of BHK-21 cells infected with VSV-GP at different multiplicities of infection (MOI = TCID50 per cell) shows that infectiondependent cell rounding occurs within hours after infection and in a dose-dependent manner (Figure 4).

[0082] Dose-dependent cell rounding can be observed over a broad range of MOIs spanning 5 logio. However, the variability of measured morphological changes increases towards both ends of this concentration range. For the very high MOIs, this variability is explained by the fact that cell rounding occurs so fast, that the base-level of rounded cells, which is also present in uninfected wells, is not covered by the images. The lower asymptote of the usually sigmoidal curves of rounding kinetics is therefore difficult to predict. However, it is obvious that on average the baseline of rounded cells must be equal in all wells right at the beginning of the infection process, before the virus is able to induce any morphological changes, independent of the applied concentration.

[0083] Therefore, we chose to apply a kinetic fit model with a shared lower asymptote for all MOIs. This allows precise calculation of the time post infection, at which half-maximal cell rounding is achieved (RT50), even at the highest tested MOIs (Figure 5). In contrast, the high variability at the low virus concentrations is rather due to biological reasons, which cannot be easily overcome. At an MOI of 0.001 , for example, only ten TCID50 (infectious units) are added per well, resulting in a very low number of initial infection events. The efficacy of production and spread of progeny viruses from these few cells, representing a variable process per se, has therefore a huge influence on the subsequent course of the rounding observed throughout the well. Statistical analysis of several experiments, in which eight replicate wells were infected per concentration, showed that a MOI range of 100-0.01 can be used for robust determination of RT50 values for VSV-GP (Figure 5). Within this MOI range also triplicate wells yield sufficient precision for accurate titer determinations.

[0084] For a method used to determine infectious virus titers, an assay range of ~2.5 log is mostly sufficient. Therefore, an MOI range of 10-0.032 with half logarithmic serial dilution steps was typically chosen for the standard curve, if not stated otherwise.

[0085] Multiple assays, performed by different operators over a course of three months, revealed that VSV-GP-dependent cell rounding and the resulting RT50 values showed a robust correlation with the applied virus dose, reflected by correlation coefficients (R2) between 0.966 and 1 (Figure 6). The differences in the curve slopes are most likely caused by differences in cell culture parameters, such as culture duration prior to seeding or passage number. Design of experiment (DoE) studies were undertaken to further optimize and harmonize these parameters. Since these differences affect the standard curve and the measured samples to similar extent, the slightly different cell rounding kinetics between different experiments, do not affect the determination of unknown titers relative to the standard curve and the assay hence provides accurate results. Theresults further demonstrate that optimization and harmonization of these parameters will most likely improve inter-assay comparability and hence suggest that under optimized conditions a standard curve may not be necessary for each assay run or on each multiwell plate.

[0086] To measure infectious virus titers with the kinetic infectivity assay, samples are typically applied in triplicate wells at a single assay dilution. In order to experimentally determine the intraplate precision of the assay, two independent assay dilutions were prepared per sample and each of these assay dilutions was used to infect triplicate wells within the same 96-well plate. This was done for a total of 65 samples, representing different in-process controls from upstream and downstream development as well as final drug substance, covering a very broad range of different titers. Independent of the absolute value, the kinetic infectivity assay precisely determined the sample titers with a mean coefficient of variation (CV %) of 11.7 % between the duplicate measurements (Figure 7).

[0087] Using the typical assay plate setup (Figure 1 B), the maximum number of samples, which can be measured in triplicates on a single plate is 24. For studies with larger sample numbers, it is necessary to split the samples to different assay plates. The assay variability between samples measured on different plates is therefore the better method performance indicator. To determine the inter-plate precision of the kinetic infectivity assay, infectious virus titers of an internal assay control, measured by different operators on different plates over a course of three months were compared (Figure 8). The determined titers ranged from 3.27E+07 to 5.69E+07 TCID50 / ml, representing a maximum deviation of 27 % from the mean value of all measurements. The coefficient of variation (CV %) for the titers measured throughout different experiments was 14.1 %.

[0088] The data therefore confirm that the kinetic infectivity assay is a very accurate, precise and reproducible assay to determine the infectious virus titer in a sample in a short period of time by measuring the kinetics of infection induced morphological changes in cells. While cell rounding was used, other infection induced morphological changes that can be detected by kinetic image acquisition are likewise envisaged.

[0089] Following completion of the DoE cell culture conditions have been slightly adapted and applied in Example 4 (Figures 11 and 12), such as fetal calf serum (FCS) concentrations (using GMEM supplemented with 6.1 % FCS and 4.3% tryptose phosphate broth) and seeding cell density (18,000 cells in 100 ml seeded 22±4h before infection), as well as shifting the reference standard curve towards higher MOIs (six MOIs with 1 log-io increment, resulting in a 2.5 log-ioassay range (MOI 31.6 - MOI 0.1). To broaden the assay range, also lower MOIs can be used, but the duration of image acquisition might have to be prolonged to ensure that all curves reach the constant upper asymptote of rounded cells. For all samples and the assay control, the target MOI used for infection was 1.78, which lies in the middle of the standard curve and allows valid measurements within the working range of the assay even if the assumed sample titer differs ±1 logio from the actual titer. In ~60 assays performed under optimized conditions, the R2of all standard curves was between 0.99 and 1. Yet, unless explicitly stated the initial conditions have been used throughout the experiments. Also, in a larger dataset using the optimized conditions a coefficient of variation of <20% (CV = 17%) was confirmed.

[0090] As the kinetic infectivity assay is a relative assay, the determined infectious titer of the samples will always be expressed in the same unit as results from the assay used to characterize the reference standard. With the described setup of the kinetic infectivity assay, infectious titers of up to 24 samples can be determined on a single 96-well plate. In routine use of the assay, turnaround time for completion of a full plate is less than 45 minutes and results are available already 24 h after infection.Example 2: Comparison of the kinetic infectivity assay with a TCID50 assay to determine VSV-GP infectious titers

[0091] In the field of virology, it is well accepted that the classical titration methods used to determine infectious virus titers suffer from high inherent variability, resulting from a combination of different factors: First, all infectivity assays are cell-based assays, in which the cells as biological system pose a source of day-to-day variation, which can hardly be avoided. This kind of variation affects especially absolute assays, in which the results are not determined relative to a standard sample. Second, as virus samples are often highly concentrated, serval pipetting and dilution steps are required to bring the sample concentrations into the working range of the assay. A robust and reproducible execution of the pipetting is therefore crucial to minimize the technical assay variation. Third, in end-point dilution assays, such as the TCID50 assay, cell culture wells are inoculated with virus samples that were diluted to a point at which the inoculated wells will receive only a single (in optimal case) or no infectious virus particle. To read out this kind of assay, inoculated wells are scored in a binary way as positive or negative for infection. The turning point from infected to uninfected wells follows the Poisson distribution and is diffuse per se. Additionally,the extremely low virus concentrations are over proportionately affected by technical and biological variability, adding another level of uncertainty to the TCID50 results.

[0092] The relatively high variability of classical virus titration methods is also considered in the ICH Guideline Q5A(R2), which states that for infectivity assays “the 95% confidence limits for results of within-assay variation normally should be on the order of ±0.5 log-io of the mean” (ICH, 2022; ICH Guideline Q5A(R2) on viral safety evaluation of biotechnology products derived from cell lines of human or animal origin - Scientific guideline, European Medicines Agency). This is a large range covering values from 31.6 % to 316.2 % of the mean value. In the past the inventors optimized and automated the TCID50 assay for the determination of VSV-GP infectious titers in BHK-21 cells (Hochdorfer D. et al., 2022, J Virol Methods 299, 114318). Figure 9 shows the infectious titers for an assay control, determined in 269 TCID50 measurements (left) over more than two years using the optimized and automated TCID50 assay (which has a comparable CV and 95% Cl to the classical manual TCID50 assay and is only optimized for the read-out and reduced hands-on time). The measurements have a CV of 43.0 % and 95 % of all values lie between 49 % and 237 % of the mean value. The ±0.5 log-io acceptance limits set by the ICH guideline (depicted as dotted lines in Figure 9) are therefore met. However, with a CV of 14.1 %, and 95% of all values ranging between 72 % and 126 % of the mean value, the kinetic infectivity assay offers a superior precision compared to the TCID50 assay (Figure 9).

[0093] Statistical analyses showed that by increasing the number of assay repetitions, the TCID50 assay can achieve the precision of the kinetic infectivity assay. However, approximately six TCID50 measurements would be required for each sample to catch up in assay precision. This in turn would further increase the workload and decrease the throughput of the TCID50 assay. To measure infectious titers of 24 samples with similar precision, would require about 20 h hands-on time and processing of 144 96-well plates for the semi-automated TCID50 assay as described in Hochdorfer et al., 2022 (and even almost 4-times more hands-on time when using manual TCID50 assay) compared to only 2 h hands-on time and a single 96-well plate for the kinetic infectivity assay. The kinetic infectivity assay therefore drastically decreases the material demand and increases the analytical throughput > 10-fold compared to the TCID50 assay.Example 3: Applicability of the kinetic infectivity assay principle to different VSV variants

[0094] In order to demonstrate the versatility of the kinetic infectivity assay in addition to VSV-GP a genetic variant of this oncolytic virus, VSV-GP-Cargo, carrying a therapeutically activetransgene was tested. VSV-GP-Cargo expresses a therapeutic fusion protein in infected cells. The same kinetic infectivity assay procedure was applied to VSV-GP and VSV-GP-Cargo to verify that the assay can be used to determine infectious virus titers of different VSV-GP variants. VSV- GP-Cargo induced dose-dependent rounding of BHK-21 cells a few hours after infection similar to VSV-GP (Figure 10A). While the onset of cell rounding was almost indistinguishable between the two virus variants, the kinetics of cell rounding were slower at later stages of the VSV-GP- Cargo infection cycle. This is reflected by slightly higher RT50 values for the Cargo-expressing variant of VSV-GP (Figure 10B). These differences in cell rounding kinetics affect the standard curve and samples to a similar extent and the kinetic infectivity assay can therefore be used without further adaptation to determine infectious virus titers of both tested VSV-GP variants.

[0095] The data confirm that the kinetic infectivity assay can be applied for VSV variants and most likely also other oncolytic viruses. Moreover, although the assay has so far only been optimized and established for BHK-21 cells, it has been shown for about 20 other adherent cell lines (HCT116, HCT115, A549, HEK293, DU-145, MCF-7, T47-D, ACHN, 786-0, OVCAR-8, NCI / ADR-RES, M14, U-251 , SF-295, LoVo, CT26.CL25, Vero, J774, B16) that VSV-GP infection induces at least a cytopathic effect. It is therefore highly likely that the majority of these cells is also suitable for use in this assay.Example 4: Specificity and sensitivity of the observed morphological alterations for VSV- GP infection

[0096] These experiments were performed under optimized conditions as described above. To confirm specificity of the observed morphological alterations for VSV-GP infection, cell rounding was tested under conditions in which viral replication was inhibited at different stages. In the presence of a neutralizing antibody (NAb), which binds to the viral glycoprotein (GP) and inhibits viral entry, no VSV-GP-dependent rounding of BHK-21 cells could be observed (Figure 11 A). Similarly, inhibition of viral RNA transcription by the polymerase inhibitor Vesiculopolin A (VPI A) (Ogino, M. et al., Viruses 2019, 11 , 856, doi: 10.3390 / v11090856) fully prevented virus-induced cell rounding (Figure 11A).

[0097] The delayed cell rounding kinetics in cells infected with the genetically modified VSV-GP AM51 further indicates that the viral M protein contributes to cell rounding (Figure 11 A). The AM51 mutation reduces the association of the viral M protein with nuclear core complexes and its ability to inhibit nucleocytoplasmic transport, which might be related to M protein mediated cytotoxicity.

[0098] The connection between progression of viral gene expression and cell rounding was further assessed in cells infected with a VSV-GP variant encoding the red fluorescent Katushka protein. The increase in the proportion of Katushka-positive cells was detectable ~6 hours after the onset of cell rounding. However, the dependence of cell rounding on viral gene expression (Figure 11 A), together with the finding that the exponential viral growth phase preceded detectability of Katushka (Figure 11 B) strongly suggests that this is rather a sensitivity issue than the actual sequence of biological events. Importantly, cell rounding and Katushka expression, which reflect viral protein production, show similar curve progression (Figure 11 B), independent of the MOI used for infection (data not shown). While at a higher MOI all kinetics shift towards earlier timepoints, the temporal relation between cell rounding, the emergence of Katushka positive cells and release of progeny virions, remains unaffected (Figure 11C). Altogether, these findings show that entry and viral gene expression are a prerequisite for virus-induced cell rounding, which reflects viral growth kinetics in infected cells.

[0099] The kinetic infectivity assay was used to measure VSV-GP subjected to increased temperature or UV irradiation. The forced sample degradation manifests itself in a delayed rounding of infected BHK-21 cells (Figure 12A). This reflects a loss of infectious virus titers, the extent of which correlates with the applied stress level (Figure 12B). These findings confirm that the KIT assay is a stability-indicating method, which can resolve also small differences in virus infectivity.

[0100] In conclusion, a high-throughput method to determine infectious virus titers based on automatic quantification of infection-induced cell rounding in kinetically acquired bright-field images has been developed. The kinetic infectivity assay (or kinetic infectious titer (KIT) assay) is characterized by its straightforward principle, low running costs, short time to results and superior precision despite a >10-fold increased throughput compared to the traditional TCID50 assay (Figure 13). The TCID50 assay has been widely utilized in various virological disciplines for many years due to its simplicity and the independence of special instrumentation. However, as an endpoint dilution assay it requires every sample to be diluted to a point at which infection is absent in the inoculated culture wells or caused by only a single (or few) infectious virus particle. Furthermore, the assay design includes infection in multiple replicates for each dilution step to calculate the TCID50, describing the sample dilution at which 50 % of the inoculated wells shows signs of infection. This high number of pipetting steps and the laborious readout procedure, which often depends on manual microscopic evaluation of infection-induced cytopathic effects (CPE), limit the sample throughput. Furthermore, the need to wait until also a single initial infection eventresults in an infection outcome, which by manual microscopic evaluation can be unambiguously identified as such, keeps the time to result long. As previously shown, automation of the sample dilution procedure combined with automated image-based CPE evaluation reduces the hands-on time ~4-fold (Hochdorfer D. et al., 2022, J. Virol. Methods 299, 114318). The present kinetic infectivity assay monitors the morphological cell alteration cell rounding, which allows an early readout based on bright-field images without additional staining steps early after infection and is likely the result of virus-induced subversion of the actin cytoskeleton and / or represents the onset of CPE. As this is true for different types of viruses, the assay principle is further likely to be applicable to a large variety of viruses.

[0101] The algorithm-based image analysis allows reliable assessment of subtle infection-induced cellular changes, much earlier than this would be possible by eye. Implementation of this readout procedure could therefore also be an option for traditional infectivity assays, where presence / absence of CPE or number of plaques is determined as at single timepoint. Several examples of using artificial intelligence (Al) to detect CPEs in bright field images have been presented previously. However, it is often not fully understood, on which properties the Al bases its decision. Using these sophisticated software tools in infectivity assay therefore entails the challenge of robust validation, quality control, and regulatory compliance, especially when applying the assay under GMP-conditions. This is more straightforward if simple and transparent readout parameters, such as cell rounding, are used.

[0102] In addition to its user-friendliness, the major benefit of the kinetic infectivity assay lies within the principle of kinetic measurements. The curve fit of the time-dependent progression of cell rounding takes into consideration the values obtained over the complete assay duration. Thus, a curve parameter such as the RT50 has a much higher accuracy and precision in comparison to a measurement at a single timepoint. This makes the measurement robust, less sensitive to outliers and yields precise results.

[0103] This assay has been successfully up-scaled to 384-well plates without major adaptations (data not shown). If higher assay precision is required in the kinetic infection assay, this could therefore be achieved by increasing the number of replicates, with only a minor reduction of sample throughput. Due to the kinetic measurement, it is also not necessary to define a fixed timepoint for the readout as images can simply be acquired until all curves reach the upper asymptote, represented by stable percentage of rounded cells. This results in a broad assay range and in most cases abolishes the need to test samples at different concentrations, which has a positive effect on number of samples which can be measured in parallel. As prolongedmeasurements negatively affect the time to results, a reasonable and practical assay range and thus duration for image acquisition should be chosen. Altogether, the presented design of the kinetic infectivity assay allows to measure up to 24 samples on a single 96-well plate within 24 hours, while keeping the variability below a CV of 20 %.

[0104] As the kinetics of cell rounding reflect productive viral replication, the kinetic infectivity assay is not only able to measure differences in concentration, but also in viral functionality. The kit assay is thus a stability-indicating method. Due to the determination of sample titers relative to a reference standard, the influence of day-to-day variability, which is often caused by cultured cells, is minimized, what makes the kinetic infectivity assay a powerful tool for long-term stability studies. Moreover, it is expected that due to the kinetic readout the kinetic infectivity assay allows to detect differences in viral fitness, which are reflected by slower infection and replication kinetics, which would be neglected by traditional infectivity assays, which depend on a single endpoint readout.Example 5: Applicability of the kinetic infectivity assay principle to Newcastle disease virus (NDV)

[0105] Morphological changes of host cells are a common phenomenon upon infection with various types of viruses. This suggests that the principle of measuring the kinetics of cell rounding (or other morphological changes of a host cell) to determine infectious titers is not limited to VSV or VSV-GP but also holds promise for unrelated virus-cell combinations. To confirm this, we used the kinetic infectivity assay principle to determine titers of Newcastle disease virus (NDV) based on rounding of DF-1 chicken fibroblasts.

[0106] It would be a significant improvement for various disciplines in the field of virology, if infectious titers of other virus types could also be determined using kinetic microscopy. As a first model virus to test the applicability of the kinetic infectivity assay principle, Newcastle disease virus (NDV) was chosen. NDV is an economically relevant virus from the family of Paramyxoviridae, causing Newcastle disease in infected birds. Measurement of NDV titers is for example important during the production of virus-based vaccines. When applying the kinetic infectivity assay setup without further adaption, NDV also induces cell rounding in infected BHK- 21 cells. The extent of cell rounding was lower, and the morphological changes appeared later compared to VSV-GP and were found as not being optimal to serve as reasonable readout parameter for routine use (data not shown). Because NDV is an avian virus, viral infection and / orreplication was expected to be more efficient in bird cells and thus chicken fibroblast cells (DF-1) were used for further experiments.

[0107] In an initial experiment, uninfected (mock) DF-1 cells showed an increase in cell rounding until 20 h, followed by a decline back to baseline within the monitored time frame (Figure 14A). High NDV doses strongly increased cell rounding until a dose-dependent peak level was reached. Lower virus doses (MOI < 10) only affected the baseline, which the proportion of rounded cells adopted during the phase of decline. These dose-dependent but heterogeneous curve shapes did not allow to apply the same kinetic fit model used to determine RT50 values in VSV- GP infected cells. As a convenient alternative, the area under the curve (AUG) was calculated (as an alternative parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i)). The AUG clearly correlates with the applied NDV dose, resulting in a sigmoidal curve shape (Figure 14B), which is precisely reflected by a 4 parameter logistic fit model and was used to interpolate infectious virus titers of NDV samples in this experiment.

[0108] Although it was necessary to adapt the data analysis for NDV in this experiment, this example clearly shows that infection-induced cell rounding determined from kinetically acquired images is a simple and powerful method to determine infectious titers also of different virus types. Based on these results it was considered likely that the NDV assay can be further improved by using host cells for infection and / or culture conditions that are optimized for NDV infection in vitro. It is therefore intriguing to speculate that with minor adaptations, the kinetic infectivity assay principle can be applied to a majority of viruses, which induce morphological changes in infected cells.

[0109] This was confirmed in an optimized experiment for which DF-1 cells (Elabscience #CL-0279) were cultured and infected under serum-free medium conditions. Without being bound by theory, serum-free conditions seem to facilitate viral replication and spread and / or slow cell growth and lead to a more rapid increase of rounded cells normalized to the total cell number, particularly for low MOIs. DF-1 cells were cultured in DMEM supplemented with 2% Ultroser G (to replace fetal calf serum in the initial experiment), 4 mM L-Glutamine and 1 mM sodium pyruvate. The overall infection procedure was the same for determining the titer of NDV in DF-1 chicken fibroblast as described for VSV-GP in BHK-21 cells. Only minor differences were implemented. For example, a seeding density of 15,000 cells / well and different MOIs used for infection were used and imaging of all wells in regular intervals of 2 h for NDV (vs. 1 h for VSV- GP) for > 70 h during incubation was performed.

[0110] As observed with VSV-GP in BHK-21 cells, under optimized conditions NDV infection induced dose- and time-dependent rounding of DF-1 cells (Figure 15A). While all curves start at a common basal value of rounded cells, the height of the upper asymptote depends on the applied MOI. It is not surprising that the relation between virus dose and cell rounding differs from the observations made with VSV-GP. Factors contributing to these characteristics include the apparently less effective spread and induction of cell death by NDV, allowing a certain level of proliferation of regular shaped DF-1 cells, which limit the proportion of rounded cells, especially at lower MOIs. Yet, unlike in the initial experiment these dose-dependent curve shapes are more homogenous and the sigmoidal curve shapes allow to apply a kinetic fit model to determine RT50 values. The RT50 values of the different doses of the reference standard are plotted against the corresponding known MOIs, resulting in a standard curve. For VSV-GP in BHK-21 cells this standard curve is reflected by a semi-log line, whereas a sigmoidal four parameter logistic fit model was applied for NDV in DF-1 cells (Figure 15B). As for VSV-GP, based on the determined RT50 value of the samples with unknown titer, their MOIs can be interpolated from this standard curve. The interpolated MOIs, together with the known number of cells seeded per well and the volume of the sample that was added per well, is then used to calculate the infectious titer.

[0111] Precision and accuracy of the measurements for NDV were evaluated by interpolating MOIs of three NDV samples (M 1-M 3) at four different concentrations (Table 2).Table 2: Results of precision and accuracy for the kinetic infectivity assay for determination of NDV titers. Recovery of the actual sample titer (3.80E+08 TCID50 / ml) and variability expressed as coefficient of variation (CV) are shown.

[0112] The mean CV of 13 % and 107 % recovery of the actual sample titer highlight that the kinetic infectivity assay does not only allow precise and accurate determination of VSV-GP titers, but this assay principle can also be applied to completely unrelated virus-cell-combinations, for which titers can be determined with the same high throughput.

[0113] As explained above, the sigmoidal curve shape observed for all MOIs under optimized conditions (serum-free conditions) allows to apply a sigmoidal four parameter logistic fit-model, which yields more precise and accurate results than the former AUC calculation and allows more reliable interpolation of MOIs to back-calculate infectious titers of the sample. Yet, the results from the initial conditions (Figure 14) have demonstrated that in principle this is also possible using the AUC analysis.Example 6: Applicability of the kinetic infectivity assay for determining the infectious titer of different viruses

[0114] In further experiments, the kinetic infectivity assay, which is based on quantification of infection-induced morphological cell alterations in kinetically acquired bright-field images, is used to determine infectious titers of different types of viruses, such as parapoxvirus ovis (Orf- Virus, ORFV) or Egg drop syndrome virus (EDSV; Duck Atadenovirus A). ORFV is an enveloped DNA virus, which causes skin lesions primarily in sheep, but can also occur on the hands of humans. EDSV is a non-enveloped DNA virus which infects poultry, causing a decrease in egg production and quality.

[0115] To identify the optimal combination of virus and cell, different adherent host cells were used for an initial screening. For ORFV these include BHK-21 , Vero, A549 and HeLa cells. Other potentially suitable host cells for ORFV include HEK293, MA104 and MDOK cells. The cells were seeded into 96-well plates. After an incubation time of 24 hours, during which the cells adhered to the bottom of the plate, cells were infected in technical triplicates at different multiplicities of infection (MOI 10 - 0.032 in 1 -logio serial dilution steps for HeLa cells or MOI 5 - 0.0512 in 2.5-fold serial dilution steps for A549 cells). Immediately after infection, the 96-well plates were placed into an automated incubator (BioSpa), which transfers the plates into the automated imaging device (Cytation5), by which bright-field images were acquired in 1 h intervals for a total duration of 48 h. Based on the acquired bright-field images, the proportion of rounded cells in relation to the total cell number was determined.

[0116] In the BHK-21 and Vero cells no robust ORFV-dependent cell rounding was observed. In contrast, ORFV infection of HeLa (Figure 16A) and A549 (Figure 17A) cells induced a time-dependent increase in rounded cells. The extent and kinetics of this increased cell rounding correlated with the MOI used for infection. These dose-dependent but heterogeneous curve progressions did not allow to apply the same kinetic fit model used to determine RT50 values inVSV-GP infected cells. As a convenient alternative, the area under the curve (AUG) was calculated (as an alternative parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i)). For HeLa (Figure 16B) and A549 (Figure 17B) cells, plotting the AUG against the applied MOI resulted in a sigmoidal curve, which can be fit by a four parameter logistic fit model. As shown before, this fit can be used as a standard curve, based on which MOIs of ORFV samples could be interpolated to finally determine their infectious virus titers.

[0117] Although slight adaptations regarding data analysis may be necessary, depending on specific biological properties of the virus and the chosen host cells, this example confirms that infection-induced cell rounding determined from kinetically acquired label-free images is a simple and powerful method to determine infectious titers that can be adapted and applied to various types of viruses.

[0118] For EDSV adherent host cells that are used for an initial screening include BHK- 21 , Vero, HEK293, LMH, DF-1 and duck embryonic fibroblasts (DEF). To perform the kinetic infectivity assay, cells are seeded into 96-well plates. After sufficient time (typically ~24 hours) during which the seeded cells can adhere to the bottom of the 96-well plate, the cells are infected in technical replicates (e.g. triplicate wells) with different concentrations of either virus. Virus concentration can be expressed as multiplicity of infection (MOI), describing the number of infectious virus particle added per cell. The MOIs used for the initial screening cover a broad range (e.g. MOI 100 - 0.001 in 1 -logio serial dilution steps) to identify the assay working range for each virus-cell combination. Immediately after infection, the 96-well plates are placed into an automated incubator (e.g. BioSpa), which transfers the plates into the automated imaging device (e.g. Cytation5), by which bright-field images are acquired in regular time intervals (e.g. 1 h). The total required imaging duration depends on cell- as well as virus-biology and should be chosen long enough (e.g. 72 h) so that infection with several or all MOIs induces the maximum level of morphological cell alteration.

[0119] After finishing the measurement, the determined morphological parameter (e.g. proportion of rounded cells) is plotted for each MOI against the time after infection. A fit model is applied, describing the observed kinetics of infection-induced morphological change. A suitable factor (e.g. RT50), is extracted from the fit models and plotted against the virus dose to determine the mathematical relation between these two factors. This relation can be used to determine sample titers relative to a reference standard for EDSV as well as other viruses.

[0120] In conclusion, the engineered oncolytic virus VSV-GP has been used as a model for the initial assay development of the kinetic infectivity assay. However, it is likely that the KIT assay can also be applied and might thus be of interest for the growing number of other VSV- based therapeutics or vaccines. Importantly, the kinetic infectivity assay was further shown to allow determining infectious titers of unrelated viruses, such as Newcastle disease virus (NDV) and ORFV and is further under development for EDSV. While the process of image acquisition and analysis was / is identical to VSV-GP, specific replication characteristics of NDV in DF-1 cells and ORFV in HeLa and A549 cells made it necessary to prolong the overall assay duration and to slightly modify the fit models for generation and analysis of the standard curve. Such biological differences regarding the growth rate of host cells, effectiveness of virus spread, or the strength of infection induced cytopathicity, require slight adaptations in the analysis procedure for specific virus-cell combinations. With the wide range of conventional or Al-based image analysis tools, suitable infection-specific morphological measurement parameter with strong time- and dosedependency are expected to be available for most types of viruses. This makes the kinetic infectivity assay a versatile tool.

Claims

CLAIMS1 . A method for determining the infectious virus titer in an aqueous sample comprising infectious virus, comprising(a) providing a plurality of aqueous samples comprising infectious virus;(b) providing at least one multiwell plate comprising 96 or more wells;(c) seeding adherent cells susceptible for infection with said virus at a set cell seeding density (CSD) into the wells of the at least one multiwell plate and culturing the adherent cells; and(d) analysing the plurality of aqueous samples comprising:(di) infecting said adherent cells of step (c) by contacting the adherent cells with each of the plurality of aqueous samples of step (a) in a separate well or replicates thereof;(dii) detecting the kinetics of at least one morphological cell alteration after infection in each well over time, preferably wherein the at least one morphological cell alteration comprises cell rounding;(diii) determining for each well or replicates thereof a parameter defining the curve of the at least one morphological cell alteration as a function of time post infection (t.p.i), wherein said parameter is preferably time post infection until % maximal extent of the at least one morphological cell alteration is achieved (time post infection % maximal extent of morphological cell alteration) or the area under the curve (AUC); and(div) determining the infectious virus titer for each of the plurality of aqueous samples analysed by comparing said parameter determined for each well or replicates thereof with the same parameter in a standard curve of a set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer, wherein the standard curve correlates the parameter with the infectious virus titer.

2. The method of claim 1 , wherein the known infectious virus titer of the reference standard is provided as tissue culture infectious dose 50 (TCID50), multiplicity of infection (MOI) or plaque forming unit (PFU) and / or wherein the method is a label-free method.

3. The method of any one of the preceding claims, wherein each of the plurality of aqueous samples comprising infectious virus of step (a) are analyzed in at least one assay dilution, preferably wherein (i) each of the plurality of aqueous samples are analysed in no more thanthree assay dilutions, more preferably at a single assay dilution; and / or (ii) wherein said at least one assay dilution of the plurality of aqueous samples has an infectious virus titer within the range of the standard curve.

4. The method of any one of the preceding claims, wherein step (di) comprises infecting said adherent cells of step (c) by contacting the adherent cells with (i) each of the plurality of aqueous samples of step (a) in replicate wells, preferably triplicates; and / or (ii) each of the set of dilutions of a reference standard comprising said infectious virus at a known infectious virus titer in replicate wells, preferably triplicates.

5. The method of any one of the preceding claims, wherein detecting the kinetics of at least one morphological cell alteration after infection in each well over time in step (dii) comprises automated image acquisition and analysis.

6. The method of claim 5, wherein the automated image analysis comprises bright-field image acquisition, dark-field image acquisition or phase contrast image acquisition, preferably bright-field image acquisition, more preferably high contrast brightfield image acquisition.

7. The method of any one of the preceding claims, wherein the at least one morphological cell alteration comprise cell rounding, preferably wherein the at least one morphological cell alteration comprise cell rounding and the detection comprises automated image acquisition and automated image analysis comprising counting cells with circular shape per well and / or per total number of analyzed cells.

8. The method of any one of the preceding claims, wherein the adherent cells susceptible for infection with said virus allow viral entry and viral gene expression of said virus, preferably wherein the virus is a cytopathic virus.

9. The method of any one of the preceding claims, wherein the adherent cells is a rodent cell, a monkey cell, an avian cell, a sheep cell or a human cell, preferably the adherent cell is selected from the group consisting of BHK-21 cells, Vero cells, HEK293 cells, HEK293T cells, DF-1 cells, LMH cells, HeLa cells, A549 cells, MA-104 cells, MDOK cells, DEF cells, Huh7.5 cells and COS7 cells.

10. The method of any one of the preceding claims, wherein the virus is selected from the group consisting of vesicular stomatitis virus (VSV), Newcastle disease virus (NDV), Orf-Virus (ORFV, parapoxvirus ovis), Egg drop syndrome virus (EDSV; Duck Atadenovirus A), humancytomegalovirus (HCMV), ZIKA virus (ZIKV), vaccinia virus, herpes simplex virus (HSV) and derivatives thereof.

11. The method of any one of the preceding claims, wherein (a) the adherent cell is a BHK-21 cell or a HEK293T cell and the virus is VSV or a derivative thereof, such as VSV-GP; (b) the adherent cell is an avian cell, preferably a DF-1 cell, and the virus is NDV or a derivative thereof; or (c) the adherent cell is a sheep or human cell, preferably a MDOK cell or human cell selected from the group consisting of HeLa cell and A549 cell, and the virus is a ORFV or a derivative thereof.

12. The method of any one of the preceding claims, wherein the at least one morphological cell alteration after infection in step (dii) comprises (i) cell rounding and (ii) a further morphological cell alteration.

13. The method of any one of the preceding claims, wherein (i) the standard curve correlating the parameter with the infectious virus titer is reflected by a non-linear fit; and / or (ii) the parameter is time post infection % maximal extent of morphological cell alteration and is calculated based on a non-linear fit, preferably a non-linear kinetic fit.

14. The method of any one of the preceding claims, wherein the plurality of aqueous samples are analyzed in at least one assay dilution and preferably wherein (i) the at least one assay dilution of the plurality of aqueous samples and / or the reference standard are prepared with a pipetting robot and / or (ii) infecting the adherent cells in step (di) is performed with a pipetting robot.

15. A composition or an aqueous sample comprising an infectious virus at a titer determined by the method of any one of claims 1-14.

Images