A method for determining the concentration of residual vero cell genomic DNA in a sample containing virus obtained from production in vero cells

Abstract

The present invention relates to a method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells, and quality methods for determining the quality of in-process control (IPC) and end product control (EPC) samples in biopharmaceutical manufacturing processes.

Description

[0001] A method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells

[0002] Co-filed sequence listing

[0003] The present specification makes reference to a sequence listing (submitted electronically on the same date as the present application). The entire contents of the Sequence Listing are incorporated herein by reference.

[0004] Technical Field of the Invention

[0005] The present invention relates to a method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells, and quality control methods for determining the quality of in-process control (I PC) and end product control (EPC) samples in biopharmaceutical manufacturing processes.

[0006] Background of the Invention

[0007] Residual DNA (rDNA) is defined as the sum total of deoxyribonucleic acid (DNA) and fragments present in biological samples derived from host cells during expression, sometimes it is also referred to as residual host cell DNA. The potential risks associated with rDNA are infectivity (through virus such as HIV), oncogenicity (through oncogenes such as Ras), immunogenicity (through CpG rich sequences from bacteria), and mutagenesis (through transposons, retrotransposons, and DNA recombination). The potential risks associated with the presence of rDNA in products developed for human use, and a regulatory requirement to confirm its clearance below specified levels in bioprocesses, is the primary reason for rDNA levels in final drug substances are closely monitored. Test methods need to be accurate, sensitive, and quantitative to assure DNA is cleared to specified levels.

[0008] The establishment of Vero cells as an experimental tool for studies ranging from virology to toxicology started in 1962 with the extraction of the cells from the Female African Green Monkey kidney. It was later discovered that the Vero cell line is a continuous cell line, meaning that it can be passaged over a significant period of time without losing its growth characteristics. Importantly, this is maintained without acquisition of tumorigenic functions as compared to other primary cell lines with limited passage numbers, allowing for several sublines to be derived and cell banking. Additionally, it was reported that Vero cells have a deficiency in interferon expression, translating in an impaired antiviral response to infection which make this cell line susceptible to infection by a wide range of viruses. Even infected at high multiplicity of infection, Vera cells do not show an active cell death pathway generally triggered by interferon expression. Consequently, WHO regulatory authorities recommended the use of Vera cell line as the first continuous cell line for human viral vaccines manufacturing.

[0009] Cao et al., Human Vaccin Immunother 9 (2013), 413-419 report the development of a Vera cell DNA reference standard for residual DNA measurement. The assay is a dot blot hybridization. The Vera cell genomic DNA was isolated prior to the dot blot hybridization.

[0010] A commercial residual DNA kit for determining the amount of Vero DNA is available from ThermoFisher (resDNASEQ quantitative Vero DNA kit). The target gene is not known. The assay is based on quantitative Real-time PCR (qPCR). The test procedure requires that the Vero DNA is purified before the assay.

[0011] Palacios et al., Open Medicinal Chem Journal 11 (2017), 66-80 describe the development of a qPCR method based on SYBR Green chemistry using highly repetitive DNA (Alu) and a house-keeping (beta-actin) as target sequences. The DNA was extracted prior to qPCR.

[0012] Andre et al., Biologicals 44 (2016), 139-149 relate to the development of a qPCR method for determining Vero DNA wherein the ribosomal 18S RNA gene is targeted.

[0013] Vernay et al., J Virol Methods 268 (2019), 9-16 describes a qPCR method for determining residual Vero genomic DNA in viral vaccines produced in Vero cells. This method requires a step of DNA extraction prior to the qPCR reaction. No restriction enzyme treatment was performed before the qPCR reaction.

[0014] Kinloch et al., J. Mol. Diagnostics 23 (2021), 907-919 relates to the determination of SARS-Cov2 RNA by RT-droplet digital PCR (RT-ddPCR). It has been found that the RT-ddPCR sensitivity is comparable to real-time RT-PCR (RT-qPCR). Park et al., J Microbiol Biotechnol 31 (2021), 358-367 relates to the determination of SARS-Cov2 RNA by RT-droplet digital PCR (RT-ddPCR). It does not describe a restriction enzyme treatment of the sample before the ddPCR reaction.

[0015] Azizi et al., J. Biotechnol 168(2013), 382-387 relates to the determination of HSV-1 UL5 and UL29 gene copy numbers in a Vero cell line transgenic for both HSV-1 genes by ddPCR. Before the ddPCR reaction the genomic DNA was extracted. Restriction enzyme treatment was performed, wherein the restriction enzyme Ncol was selected to cut outside of the UL5 and UL29 genes but within the integrating plasmids.

[0016] In view thereof, there is a need for a sensitive, specific and robust assay for determining the concentration of Vero cell genomic DNA in in-process samples or end product control samples in biopharmaceutical manufacturing, in particular virus containing samples, wherein the assay does not require a prior extraction of DNA from the sample.

[0017] Summary of the Invention

[0018] The technical problems underlying the invention are solved by the provision of the subject-matter as defined in the claims.

[0019] According to a first aspect, the present invention provides a digital PCR (dPCR) method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells comprising the steps of a) preparing serially diluted samples without prior DNA extraction; b) contacting the serially diluted samples of step a) separately with a composition comprising:

[0020] (i) a set of primers suitable for amplification of an alpha satellite target sequence, wherein the alpha satellite target sequence comprises the nucleotide sequence set forth in SEQ ID NO:

[0021] 1 ,

[0022] (ii) a labeled probe specific for the alpha satellite target sequence,

[0023] (iii) a DNA polymerase, and nucleotides, and

[0024] (iv) a restriction enzyme, c) incubating the mixture of step b) under conditions allowing a restriction enzyme digest of the mixture; d) partitioning the digested mixture of step b); e) PCR-amplifying the alpha satellite target sequence in the partition of step d); f) determining the binding of said labeled probe to the PCR product and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample. According to a second aspect provided, a quality control method for determining the amount of residual Vero cell genomic DNA in a sample from a manufacturing process of a flavivirus vaccine comprising performing the method according to the invention on an in-process control (IPC) sample or an end product control (EPC) sample of said process.

[0025] According to a third aspect is provided, a combination of quality control methods for vaccines containing live, attenuated dengue virus comprising performing the method according to the invention and at least one further method selected from the group consisting of immunofocus assay, identity assay, attenuation check, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s).

[0026] The dPCR method which is preferably a nano-plate based dPCR method, targeting simian alpha satellite DNA for quantifying of Vero genomic DNA in a sample containing virus obtained from production in Vero cells has been established with the positive effect that the method does not require prior DNA extraction from the sample.

[0027] Further, the dPCR method according to the invention is characterized by a limit of detection (LCD) of 1 pg / mL of DNA (corresponding to about 13 to 15 copies of alpha-satellite gene per pL of the reaction) in the presence of restriction enzyme in the mastermix. which can reliably be detected. The dPCR method according to the invention has a lower limit of quantification (LLOQ) of equal to or less than 0.001 ng / mL (=1 pg / mL), such as in particular 0.0003 ng / mL (= 0.3 pg / mL) which is significantly lower compared to other prior art assays for Vero genomic DNA quantitation.

[0028] Brief Description of the Drawings

[0029] Figure 1 shows the transfer to Assay Plate from Standard PCR Plate.

[0030] Figure 2 provides a representative Nanoplate Layout that may be modified to suit the assay design.

[0031] Figure 3 shows (A) 1 D- scatterplots of eight 10-fold dilutions of Vero DNA without (MM1 , wells A1- C2) or with (MM2, wells C3-F1) treatment of restriction enzyme (Xbal) digestion priorto amplification of the target in dPCR. (B) 1 D- scatterplots of three 10-fold dilutions of Human Embryonic Kidney - 293 (HEK-293) cell DNA without (MM1 , wells G1-G3) or with (MM2, wells H1-H3) treatment of restriction enzyme (Xba-I). Wells F2 and F3 contained water as a NTC. The horizontal lines represent the threshold set for separating positive from negative populations.

[0032] Figure 4 shows a Flow-chart to prepare samples NE = No extraction; Ext. = Extraction. TDV-2 Stab. Neat refers to a stabilized TDV-3 sample which is undiluted.

[0033] Figure 5 shows 1 D-Scatterplots of Neat and 10-fold Dilutions of TDV-3 D7 Post-stabilization Harvest Sample Without Extraction (Wells A1-B1) and With Extraction (Wells B2-C2). The horizontal lines represent the threshold set for separating positive from negative populations.

[0034] Figure 6 shows 1 D-scatterplots of 10-fold diluted TDV-3 post-stabilization sample with different combinations of treatment like extraction or non-extraction of DNA by magnetic beads, spiked or non-spiked stabilized sample, and its dilutions with known amount of Vero DNA.

[0035] Figure 7 shows 1 D-Scatterplots of 10-fold Diluted TDV-3 D7 Post-AEX (AEX: anion exchange) Harvest Sample with Different Combinations of Treatments.

[0036] Figure 8 shows 1 D-Scatterplots of 10-fold Diluted TDV-3 D7 BDS (BDS: bulk drug substance) Harvest Sample with Different Combinations of Treatments.

[0037] Figure 9 shows 10-fold serial dilutions of Vero DNA (10 ng / mL to 0.0001 ng / mL) showed linear response in dPCR using a-SAT primers and FAM-probe.

[0038] Detailed Description of the Invention

[0039] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention pertains. Although, any processes and materials similar or equivalent to those described herein can be used in practice for testing of the present invention, the preferred materials and processes are described herein. In describing and claiming the present invention, the following terminology will be used in accordance with the definitions set out below.

[0040] Where the term “comprise” or “comprising” is used in the present description and claims, it does not exclude other elements or steps. For the purpose of the present invention, the term “consisting of’ is considered to be an optional embodiment of the term “comprising”. If hereinafter a group is defined to comprise at least a certain number of embodiments, this is also to be understood to disclose a group which optionally consists only of these embodiments. Where an indefinite or a definite article is used when referring to a singular noun e.g. “a” or “an”, “the”, this includes a plural form of that noun unless specifically stated.

[0041] Vice versa, when the plural form of a noun is used it refers also to the singular form.

[0042] Furthermore, the terms first, second, third or (a), (b), (c) and the like in the description and in the claims are used for distinguishing between similar elements and not necessarily for describing a sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein unless context clearly indicates otherwise.

[0043] In the context of the present invention any numerical value indicated is typically associated with an interval of accuracy that the person skilled in the art will understand to still ensure the technical effect of the feature in question. As used herein, the deviation from the indicated numerical value is in the range of ± 10%, and preferably of ± 5%. The aforementioned deviation from the indicated numerical interval of ± 10%, and preferably of ± 5% is also indicated by the terms “about” and “approximately” used herein with respect to a numerical value.

[0044] According to the first aspect of the present invention, a method the present invention provides a digital PCR (dPCR) method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells comprising the steps of a) preparing serially diluted samples without prior DNA extraction; b) contacting the serially diluted samples of step a) separately with a composition comprising:

[0045] (i) a set of primers suitable for amplification of an alpha satellite target sequence, wherein the alpha satellite target sequence comprises the nucleotide sequence set forth in SEQ ID NO:

[0046] 1 ,

[0047] (ii) a labeled probe specific for the alpha satellite target sequence,

[0048] (iii) a DNA polymerase, and nucleotides, and

[0049] (iv) a restriction enzyme, c) incubating the mixture of step b) under conditions allowing a restriction enzyme digest of the mixture; d) partitioning the digested mixture of step b); e) PCR-amplifying the alpha satellite target sequence in the partition of step d); f) determining the binding of said labeled probe to the PCR product and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample. “virus” herein means any virus which can be produced in Vero cells. The virus includes from hepatitis virus, influenza virus and a flavivirus.

[0050] The virus may further include virus containing vaccines against Ebolavirus disease, influenza, Japanese encephalitis, poliomyelitis, rabies, rotavirus, smallpox and virus containing vaccines to control the current COVID-19 pandemic. It also includes flaviviruses.

[0051] The virus may be used as an inactivated virus or live, attenuated virus for vaccine purposes.

[0052] “An attenuated virus” herein means that the virus has a reduced replication capacity and / or a reduced infectivity in host cells compared to wild-type virus. The replication capacity and / or infectivity may be determined in vitro in suitable cell systems or in vivo in suitable animal models. Attenuation may be achieved by serial passaging of the virus in a foreign host such as in tissue culture, embryonated eggs or live animals. Alternatively, attenuation may be performed by chemical agents.

[0053] “Live attenuated” viruses are important for use in viral vaccines, since the live attenuated virus generates a stronger immune response compared to an inactivated virus. As used herein, the term “live attenuated virus” can refer to e.g. a chimeric virus or a non-chimeric virus. The flavivirus- containing sample may be a flavivirus composition.

[0054] The live attenuated virus may be an RNA virus or a DNA virus. Suitable live attenuated RNA viruses may be selected from dengue virus, poliovirus, rubella virus, measles virus, yellow fever virus, mumps virus, zika virus, and influenza virus. Preferably, the live attenuated RNA virus is dengue virus or a zika virus, such as dengue virus. The live attenuated RNA virus may be a chimeric virus. Suitable live attenuated DNA viruses include varicella zoster virus, vaccinia virus, smallpox virus, Herpes simplex virus and chikungunya virus. The live attenuated virus may differ from the wild-type virus in one or more mutations in the nucleic acid sequence of the virus. For example, for one tetravalent dengue vaccine (TDV) it is known that attenuation of neurovirulence in newborn mice is determined by mutations 5’NCR-57 C->U, NS1-53 Gly->Asp (nt-2579 G->A) and NS3-250 Glu->Val (nt-5270 A->U). The above nucleotide sequence numbering relates to the nucleotide sequence of the live attenuated Dengue virus serotype 2 (TDV-2) set forth in SEQ ID NO:2.

[0055] The live, attenuated flavivirus can be a flavivirus in which all components are derived from the same flavivirus serotype, or it can be a chimeric flavivirus having parts from two or more flavivirus serotypes. The live, attenuated dengue virus can be a dengue virus in which all components are derived from the same dengue serotype, or it can be a chimeric dengue virus having parts from two or more dengue serotypes. A "virus strain" and in particular a "dengue virus strain" is a genetic subtype of a virus, in particular of a dengue virus, which is characterized by a specific nucleic acid sequence. A dengue serotype may comprise different strains with different nucleic acid sequences which have the same cell surface antigens and are therefore recognized by the same antibodies. A dengue virus strain can be a dengue virus in which all components are derived from the same dengue serotype, or it can be a chimeric dengue virus having parts from two or more dengue serotypes.

[0056] A “chimeric virus” or "chimeric strain" or “chimeric virus strain” in general comprises parts from at least two different viruses. For example, a chimeric virus can comprise the prM and E proteins of dengue virus and the other proteins from another flavivirus. The chimeric virus can comprise the prM and E proteins of dengue virus and the other proteins from another flavivirus such as yellow fever virus, Zika virus, West Nile virus, Japanese encephalitis virus, St. Louis encephalitis virus and tick- borne encephalitis virus. The chimeric virus can comprise the prM and E proteins of dengue virus and the other proteins from yellow fever virus strain YF-17D. Such chimeric viruses are present in the commercial product Dengvaxia® and are described in, e.g., WO 98 / 37911 , WO 03 / 101397, WO 2007 / 021672, WO 2008 / 007021 , WO 2008 / 047023 and WO 2008 / 065315.

[0057] A “chimeric dengue virus” or "chimeric dengue serotype strain" or “chimeric dengue strain” preferably comprises parts from at least two different dengue serotypes, i.e. a dengue-dengue chimera. Such chimeric dengue viruses are described in WO 01 / 060847 A2, WO 2014 / 150939 A2 and WO 2017 / 179017 A1 .

[0058] As used herein, “TDV” refers to a tetravalent live attenuated dengue vaccine that comprises a mixture of the four live attenuated dengue virus strains TDV-1 , TDV-2, TDV-3 and TDV-4 expressing surface antigens from the four dengue serotypes DENV-1 (dengue serotype 1), DENV-2 (dengue serotype 2), DENV-3 (dengue serotype 3) and DENV-4 (dengue serotype 4), respectively.

[0059] As used herein, “TDV-2” refers to a molecularly characterized and cloned dengue serotype 2 strain derived from the live attenuated DEN-2 PDK-53 virus strain. The PDK-53 strain is described for example in Bhamarapravati et al. (1987) Bulletin of the World Health Organization 65(2): 189-195. In one embodiment, the TDV-2 strain served as a backbone for the chimeric TDV-1 , TDV-3 and TDV-4 strains into which parts from the DENV-1 , DENV-3 and DENV-4 strains were introduced.

[0060] As used herein, a “TDV-1” refers to a dengue virus chimeric construct which comprises parts from both DENV-2 and DENV-1. Preferably, in TDV-1 the prM and E proteins from DENV-1 replace the prM and E proteins from DENV-2. As used herein, “TDV-3” refers to a dengue virus chimeric construct which comprises parts from both DENV-2 and DENV-3. Preferably, in TDV-3 the prM and E proteins from DENV-3 replace the prM and E proteins from DENV-2.

[0061] As used herein, “TDV-4” refers to a dengue virus chimeric construct which comprises parts from both DENV-2 and DENV-4. Preferably, in TDV-4 the prM and E proteins from DENV-4 replace the prM and E proteins from DENV-2.

[0062] In embodiments where the flavivirus composition comprises more than one live, attenuated virus, the composition in a preferred embodiment comprises at least one chimeric live attenuated flavivirus.

[0063] More preferably, the composition comprises one chimeric flavivirus and three non-chimeric flaviviruses. For example, the composition can comprise live, attenuated DENV-1 , DENV-2 and DENV-4 viruses and a chimeric DENV-2 / DENV-4 chimeric virus. In an alternative preferred embodiment, the composition comprises at least two chimeric live, attenuated viruses and at least one non-chimeric flavivirus. More preferably, the composition comprises a live, attenuated DENV-2 virus, a chimeric DENV-2 / DENV-1 virus, a chimeric DENV-2 / DENV-3 virus and a chimeric DENV- 2 / DEN-4 virus, most preferably e.g. the commercially available vaccine under trademark “Qdenga©”.

[0064] “Vero cell” The Vera cell is originally derived from the kidney of an African Green monkey in the 1962 by Yasamura and Kawakita at the Chiba University in Japan. It was first used for human vaccines in the early 1980s for the production of viruses by Montagnon and colleagues at the Institut Merieux in Lyon, France.

[0065] The term “Vero cell” herein includes all (including commercially available) Vero cell sublines such as Vero76, Vero6, Vero C1008, Vero / hSLAM, Vero (AC-free), Vero-Hektor, MA104 and B95a (all of them available from ECACC). It also includes genetically modified variants of the original Vero cell line, preferably knockout (KO) variants with knockout in Vero target genes (Hoeksema et al., Vaccine 36 (2018), 2093-2103.

[0066] “Virus-containing sample obtained from production in Vero cells”

[0067] As outlined above, the production of virus in Vero cells is well known to the skilled person. Preferably, the production is suitable for biopharmaceutical industry. More preferably, the production is large-scale. The sample may be an in-process control (IPC) sample or an end product control (EPC) sample of the manufacturing process.

[0068] The in-process control may be taken before or after a purification step during the manufacture. Purification steps may include ion exchange, filtration, centrifugation, precipitation. The end-process control sample may contain at least one pharmaceutically acceptable excipient. Pharmaceutically acceptable excipients are known to the skilled person.

[0069] Suitable excipients include stabilizers, detergents, sugars, and buffer salts.

[0070] Preferably, the pharmaceutically acceptable excipients comprise F127, trehalose and / or human serum albumin.

[0071] In one preferred embodiment the sample is a monovalent drug substance or a tetravalent drug product in the case of production of flavivirus vaccines on the basis of live, attenuated viruses.

[0072] “alpha satellite sequence” Alpha satellite DNA is the most abundant satellite DNA in Primates and is found both at the site of centromere attachment and in neighboring heterochromatic regions, referred to as pericentromeric regions. Alpha satellite DNA was originally isolated as a highly repetitive component of the Chlorocebus aethiops (also called African green monkey) genome; homologous repeats were then described throughout the Primate order including apes, Old World and New World monkeys.

[0073] Alpha satellite target sequence herein refers to a segment of the above alpha satellite DNA. Preferably, the alpha satellite sequence comprises the nucleotide sequence consisting of SEQ ID NO: 1. More preferably, the alpha-satellite sequence comprises SEQ ID NO: 1. a) preparing serially diluted samples without prior DNA extraction

[0074] In a preferred embodiment serial dilutions of the above samples may be 1 :10; 1 :100; 1 :1000; 1 :10.000 etc.

[0075] “without prior DNA extraction” herein means that no step of extraction of the DNA by e.g. precipitation of the DNA from the sample is performed. Methods for precipitation which are not used by the present invention include organic solvent-based DNA extraction such as phenol / chloroform or isoamyl alcohol; inorganic solvent-based DNA extraction such as Proteinase K DNA extraction, salting out method, sodium iodide precipitation, SDS DNA extraction, CTAB DNA extraction, silica- gel-based techniques; physical or solid-based DNA extraction methods such as paper DNA extraction or magnetic bead DNA extraction. b) contacting the serially diluted samples of step a) separately with a composition comprising:

[0076] (i) a set of primers suitable for amplification of an alpha satellite target sequence, wherein the alpha satellite target sequence comprises the nucleotide sequence set forth in SEQ ID NO: 1, (ii) a labeled probe specific for the alpha satellite target sequence,

[0077] (iii) a DNA polymerase, and nucleotides, and

[0078] (iv) a restriction enzyme,

[0079] The skilled person is aware of software for generating suitable forward primers, reverse primers and probes for a given sequence. Preferably, the forward primer comprises the nucleotide sequence of SEQ ID NO: 2 or a variant thereof having at least 80 % sequence identity; and the reverse primer comprises the nucleotide sequence of SEQ ID NO: 3 or a variant thereof having at least 80 % sequence identity. More preferably, the sequence variant has a sequence identity of at least 90%, even more preferred of at least 95 % sequence identity and in particular of at least 97 %. Algorithms for the determination of sequence identity are known to the skilled person. Preferably, the entire sequence of SEQ ID NO: 2 or SEQ ID NO: 3 is used for sequence comparison.

[0080] The probe is a labelled probe. The label may be a colorimetric, luminescent or fluorescent label. The label may be introduced by covalent or non-covalent labelling. In addition, fluorescent nucleoside analogs may be used. External labeling with classical fluorophores, such as rhodamine, cyanine, and fluorescein dyes, linked to the nucleoside via a flexible spacer arm is very common.

[0081] The probe may be a Taqman probe having at the 5' end a fluorescent label and at the 3’ end a quencher. When the Taqman probe binds to the target sequence, the probe fluoresces. During amplification of the target sequence with the DNA polymerase, the probe is digested and the fluorescence disappears.

[0082] Preferably, the final concentration of the forward primer is about 0.2 pM to about 1 .2 pM, more preferably about 0.7 pM to about 0.9 pM;

[0083] In another preferred embodiment, the final concentration of the reverse primer is about 0.2 pM to about 1 .2 pM, more preferably about 0.7 pM to about 0.9 pM.

[0084] In another preferred embodiment, the final concentration of the probe is about 0.2 pM to about 1 .2 pM, preferably about 0.3 pM to about 0.5 pM.

[0085] The polymerase is a thermostable polymerase useful in PCR reactions such as the Taq polymerase or variants thereof. The nucleotides are mixtures of adenosine, guanosine, thymidine and cytosine.

[0086] “restriction enzyme” herein includes any known restriction enzyme. Traditionally, four types of restriction enzymes are recognized, designated I, II, III, and IV, which differ primarily in structure, cleavage site, specificity, and cofactors. Types I and III enzymes are similar in that both restriction and methylase activities are carried out by one large enzyme complex, in contrast to the type II system, in which the restriction enzyme is independent of its methylase. Type-ll restriction enzymes also differ from types I and III in that they cleave DNA at specific sites within the recognition site; the others cleave DNA randomly, sometimes hundreds of bases from the recognition sequence. Several thousand type II restriction enzymes have been identified from a variety of bacterial species. These enzymes recognize a few hundred distinct sequences, generally four to eight bases in length. Type IV restriction enzymes cleave only methylated DNA and show weak sequence specificity. Preferably, the restriction enzyme is selected from Xbal, Hindi 11 or a genetically engineered variant thereof, more preferably the restriction enzyme is a genetically engineered Hindlll variant having reduced star activity such as Hindi I l-H F (New England Biolabs). Star activity means that under non-standard reaction conditions, some restriction enzymes are capable of cleaving sequences which are similar, but not identical, to their defined recognition sequence. It has been found by the inventors that by the use of the selected restriction enzyme the dPCR method shows a high accuracy and sensitivity. c) incubating the mixture of step b) under conditions allowing a restriction enzyme digest of the mixture

[0087] The incubation can be performed before or during step d). d) partitioning the digested mixture of step b)

[0088] The digital PCR (dPCR) method may be carried out in the form of a droplet digital PCR (ddPCR) or using a microfluidic device such as a nanoplate (hereinafter: nanoplate-based dPCR). Alternatively, the dPCR method may be carried out as a chip-based digital PCR (cdPCR), microwell chip-based dPCR, a crystal digital PCR, a chip-in-a-tube digital PCR or a semiconductor chip-based dPCR. Preferably, the dPCR method is carried out using a microfluidic device, more preferably as a nanoplate-based dPCR.

[0089] Devices for carrying out a droplet digital PCR method are commercially available e.g. from Bio-Rad. For example, a QX ONE, QX200 or QX600 ddPCR system may be used.

[0090] Droplet Digital PCR technology is a digital PCR method utilizing a water-oil emulsion droplet system. A sample is fractionated into 20,000 droplets, and PCR amplification of the template molecules occurs in each individual droplet. Droplets are formed in a water-oil emulsion to form the partitions that separate the template DNA molecules. The droplets serve essentially the same function as individual test tubes or wells in a plate in which the PCR reaction takes place, albeit in a much smaller format. The massive sample partitioning is a key aspect of the ddPCR technique.

[0091] The Droplet Digital PCR System partitions nucleic acid samples into thousands of nanoliter-sized droplets, and PCR amplification is carried out within each droplet. Following PCR, each droplet is analyzed or read to determine the fraction of PCR-positive droplets in the original sample. These data are then analyzed using Poisson statistics to determine the target DNA template concentration in the original sample.

[0092] If the dPCR method is a ddPCR method, the digested mixture of step b) will be partitioned into nanoliter-sized droplets and PCR amplification is carried out within each droplet.

[0093] Devices for carrying out the dPCR method using a microfluidic device, in particular nanoplate-based dPCR are commercially available e.g. from Qiagen. Preferably, the device is the QIAcuity system. For the QIAcuity system 24- and 96-well plates with 8500 or 26000 partitions per well are available.

[0094] If the dPCR method is a nanoplate-based dPCR, the digested mixture of step b) applied in a well of the nanoplate is partitioned into partitions of the nanoplate, wherein each well is preferably partitioned into at least 50 partitions, more preferably, each well is partitioned into at least 8500 or at least 26000 partitions. e) PCR-amplifying the alpha satellite target sequence in the partition of step d)

[0095] The PCR amplification is achieved by thermocycling. Preferably, an initial heat activation step for about 1 to 3 min at about 90 to 98°C is used. In another preferred embodiment the denaturation step is conducted for about 10 to 20s at about 90 to 98°C. In another preferred embodiment the annealing / extension step is carried out for about 20 to 40s at about 50 to 70°C.

[0096] The following conditions are more preferably used: i) initial heat activation for about 2 minutes at 95°C; and ii) about 15s at 95°C (denaturation) followed by about 30s at 60°C (combined annealing / extension) over about 30 to about 50 cycles, preferably 40 cycles. f) determining the binding of said labeled probe to the PCR product and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample

[0097] In a preferred embodiment step f) comprises conversion of the alpha satellite copy number / pl sample to ng Vero DNA / ml sample by using the formula:

[0098] — = about 1x105to about 3x105(Copies mean)(dilution factor)

[0099] In a preferred embodiment the formula is: — = (Copies mean)(1 .4993 x 10-5)(dilution factor) . According to another aspect, the present invention provides a quality control method for determining the amount of residual Vero cell genomic DNA in a sample from a manufacturing process of a flavivirus vaccine comprising performing the method according to the invention on an in-process control (IPC) sample or an end product control (EPC) sample of said process.

[0100] According to a further aspect, the present invention provides a combination of quality control methods for vaccines containing live, attenuated dengue virus comprising performing the quality control method according to the invention and at least one further method selected from the group consisting of immunofocus assay, identity assay, attenuation check, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least two further method selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least three further method selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least four further method selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least five further method selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least six further method selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least seven further methods selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least eight further methods selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and at least nine further methods selected from the group consisting of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s). In an embodiment, the quality control method comprises or consists of the method of the first aspect of the invention and all of immunofocus assay, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s).

[0101] In another preferred such embodiment, the excipients are a combination of trehalose, at least one poloxamer, urea, arginine hydrochloride, tromethamine and human serum albumin. More preferably, the excipients are a combination of trehalose, poloxamer, urea, arginine hydrochloride, tromethamine, human serum albumin, chloride salts and phosphate salts. Preferably, the chloride salts comprise or consist of sodium chloride.

[0102] Item List

[0103] 1 . A digital PCR (dPCR) method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells comprising the steps of a) preparing serially diluted samples without prior DNA extraction; b) contacting the serially diluted samples of step a) separately with a composition comprising:

[0104] (i) a set of primers suitable for amplification of an alpha satellite target sequence, wherein the alpha satellite target sequence comprises the nucleotide sequence set forth in SEQ ID NO: 1 ,

[0105] (ii) a labeled probe specific for the alpha satellite target sequence,

[0106] (iii) a DNA polymerase, and nucleotides, and

[0107] (iv) a restriction enzyme, c) incubating the mixture of step b) under conditions allowing a restriction enzyme digest of the mixture; d) partitioning the digested mixture of step b); e) PCR-amplitying the alpha satellite target sequence in the partition of step d); f) determining the binding of said labeled probe to the PCR product and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample.

[0108] 2. The dPCR method of item 1 , wherein the alpha satellite target sequence consists of the nucleotide sequence set forth in SEQ ID NO: 1 .

[0109] 3. The dPCR method of item 1 or 2, wherein the forward primer has the nucleotide sequence set forth in SEQ ID NO: 2 or a variant thereof having at least 80 % sequence identity; and the reverse primer has the nucleotide sequence set forth in SEQ ID NO: 3 or a variant thereof having at least 80 % sequence identity.

[0110] 4. The dPCR method of any one of items 1 to 3, wherein the labeled probe has the nucleotide sequence set forth in SEQ ID NO: 4 and preferably the probe is a Taqman probe.

[0111] 5. The dPCR method of any one of items 1 to 4, wherein the restriction enzyme is selected from Hindlll or a genetically engineered variant thereof, preferably the restriction enzyme is a genetically engineered Hindlll variant having reduced star activity.

[0112] 6. The dPCR method of any one of items 1 to 5, wherein the dPCR method is carried out as a droplet dPCR method or as a microfluidic device-based dPCR method, preferably the dPCR method is a nanoplate-based dPCR method.

[0113] 7. The dPCR method of any one of items 1 to 6, wherein the dPCR method comprises d) partitioning the digested mixture of step b) into at least 50 partitions of a well of a nanoplate; e) PCR-amplifying the alpha satellite target sequence in the at least 50 partitions; f) determining the binding of said labeled probe to the PCR product in the at least 50 partitions and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample.

[0114] 8. The dPCR method of any one of items 1 to 7, wherein the digested mixture of step b) is partitioned into 8500 or 26000 partitions per well of the nanoplate, preferably 26000 partitions per well.

[0115] 9. The dPCR method of any one of items 1 to 8, wherein

[0116] (i) the final concentration of the forward primer is about 0.2 pM to about 1 .2 pM, preferably about 0.7 pM to about 0.9 pM;

[0117] (ii) the final concentration of the reverse primer is about 0.2 pM to about 1.2 pM, preferably about 0.7 pM to about 0.9 pM; and

[0118] (iii) the final concentration of the probe is about 0.2 pM to about 1 .2 pM, preferably about 0.3 pM to about 0.5 pM. 10. The dPCR method of any one of items 1 to 9, wherein the PCR-amplifying step e) comprises i) initial heat activation for about 2 minutes at 95°C; and ii) about 15s at 95°C (denaturation) followed by about 30s at 60°C (combined annealing / extension) over about 30 to about 50 cycles, preferably 40 cycles.

[0119] 11 . The dPCR method of any one of items 1 to 10, wherein the calculation step f) comprises conversion of the alpha satellite copy number / pl sample to ng Vero DNA / ml sample by using the formula: ^■ = about 1x10’5to about 3x10’5(Copies mean)(dilution factor).

[0120] 12. The dPCR method of item 11 , wherein the formula is:

[0121] = (Copies mean)(1.4993 x 10-5)(dilution factor).

[0122] 13. The dPCR method of any one of items 1 to 12, wherein the virus is selected from hepatitis virus, influenza virus and a flavivirus.

[0123] 14. The dPCR method of item 13, wherein the flavivirus is selected from dengue virus, yellow fever virus, Zika virus, West Nile virus, Japanese encephalitis virus, St. Louis encephalitis virus and tick-borne encephalitis virus and combinations thereof, preferably the flavivirus is dengue virus.

[0124] 15. The dPCR method of any one of items 13 or 14, wherein the flavivirus is a live attenuated or a chimeric virus flavivirus, preferably, the live attenuated flavivirus comprises one or more of live attenuated dengue viruses, more preferably, the live attenuated flavivirus comprises one or more of TDV-1 , TDV-2, TDV-3 and TDV-4.

[0125] 16. The dPCR method of any one of items 1 to 15, wherein the sample containing flavivirus produced in Vero cells is an in-process monovalent drug substance, a bulk drug substance or a final tetravalent drug product.

[0126] 17. The dPCR method of any one of items 1 to 16, wherein the sample containing flavivirus produced in Vero cells comprises one or more pharmaceutically acceptable excipients, preferably the sample comprises F127, trehalose and / or human serum albumin.

[0127] 18. A quality control method for determining the amount of residual Vero cell genomic DNA in a sample from a manufacturing process of a flavivirus vaccine comprising performing the method of any one of items 1 to 17 on an in-process control (I PC) sample or an end product control (EPC) sample of said process. 19. A combination of quality control methods for vaccines containing live, attenuated dengue virus comprising performing the method according to item 18 and at least one further method selected from the group consisting of immunofocus assay, identity assay, attenuation check, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s).

[0128] The invention is further described in the following examples which are solely for the purpose of illustrating specific embodiments of the invention, and are also not to be construed as limiting the scope of the invention in any way.

[0129] Examples

[0130] 1.1 General Materials, Equipment, and Reagents

[0131] *or equivalent

[0132] *or equivalent

[0133] Table of primer and probe sequences

[0134] * Or Equivalent

[0135] 6 FAM: 6-carboxyfluorescein

[0136] MGB: minor groove binder

[0137] NFQ: nonfluorescent quencher Example 1 : Procedure of dPCR assay

[0138] Preparation of Test Material (samples), Primers-probe Mix, Positive and Non-template Controls

[0139] The samples such as post AEX (IPC) and BDS (EPC) will not need any pre-dilutions before testing in this procedure because they have already been through the rigorous purification steps during TDV manufacturing process. The following table provides a summary of suggested dilution factors based on sample classification. Any dilutions more concentrated than 1 :10,000 will saturate wells and produce unreliable results. Preparation of serial dilutions of in-house prepared Vero PC DNA Stock

[0140] (Stock concentration = lOOng / pL (i.e., 100,000pg / pL); Volume = 20pL / vial)

[0141] Reaction Setup Table 1 Reaction Setup

[0142] RE, Restriction Enzyme. fVolume may vary, depending on concentration of the primer / probe mix used. t RNase-free water amount is variable and depends on amount of 10X primer-probe mix. *Nanoplate 26K, 24-well with a high dynamic range of logs for highest precision The reaction mix is vortexed for 5-10 seconds to ensure proper mixing of the components.

[0143] 35 pL of the reaction mix are dispensed, which contains all components except the template, into the wells of any three adjacent columns of a standard PCR plate (a plate with V-shaped wells preferred). Then, 5 pL template DNA is added into each appropriate well that contains the reaction mix. Pipette up and down 5-8 times using a 100 pL pipette set at 30 pL to ensure solution homogeneity. All content (40 pL) from each well of the standard PCR plate is transferred to the wells of the nanoplate as shown in Figure 1. Figure 2 provides a representative Nanoplate Layout that may be modified to suit the assay design.

[0144] Thermal Cycling Conditions

[0145] Table 2 Cycling Conditions

[0146] Data Analysis and Interpretations

[0147] Conversion of copies / pL to ng / mL: To determine the final concentration of Vero res HC-DNA in the TDV IPC and EPC in ng / mL, the concentration values in copies / pL will be converted into amount of DNA by Avogadro’s number formula using the following equation for each sample type, stabilized & post AEX (IPC) and BDS (EPC). Important parameter values are provided in the table below.

[0148] Example to calculate res HC-DNA (in ng / mL) in original TDV stabilized sample:

[0149] = (Copies mean)(l.4993 x 105) (dilution factor)

[0150] AS, a-Satellite; CF, conversion factor. Example 1A: Alternative Procedure of dPCR assay

[0151] The preparation of test materials and serial dilutions correspond to the description of Example 1 . The Reaction set-up has been changed as follows:

[0152] Table 3 Reaction Setup

[0153] Volume / reaction (Nanoplate i . . ..

[0154] Components 26K, 24-well*) Final concentration

[0155] 4X Probe PCR Master Mix 12 pL 1x g 0-8 pM forward primer, 0.8 pM

[0156] 10X primer-probe mix

[0157] Ter, 0.2 pM probe

[0158] RE Hindlll-HF 1 pL 14 U / pL

[0159] RNase-free water 24.2 pL$ NA

[0160] Template DNA (added at step 6.2.2 4)

[0161] Total reaction volume 4

[0162] RE, Restriction Enzyme.

[0163] ■fVolume may vary, depending on concentration of the primer / probe mix used. t RNase-free water amount is variable and depends on amount of 10X primer-probe mix. *Nanoplate 26K, 24-well with a high dynamic range of logs for highest precision

[0164] The reaction mix is vortexed for 5-10 seconds to ensure proper mixing of the components.

[0165] 42 pL of the reaction mix are dispensed, which contains all components except the template, into the wells of any three adjacent columns of a standard PCR plate (a plate with V-shaped wells preferred). Then, 6 pL template DNA is added into each appropriate well that contains the reaction mix. Pipette up and down 5-8 times using a 100 pL pipette set at 30 pL to ensure solution homogeneity.

[0166] (40pL out of 48pL of total volume) from each well of the standard PCR plate is transferred to the wells of the nanoplate.

[0167] The thermocycling conditions and data analysis correspond to the conditions and analysis described for Example 1.

[0168] Example 2: Testing of Samples

[0169] Using optimized a-SAT primers / probe concentration and thermal cycler conditions from the last qPCR assay, the dilutions of new Vero DNA and HEK-293 DNA samples were tested in both qPCR as well as dPCR assay. Microbial DNA-free water (Qiagen) was used as a non-template control sample. Two separate mastermixes, one without RE Xbal and other with RE Xbal, were prepared to test serially diluted DNA samples. For qPCR, T aqMan Advanced Fast master-mix was used with optimized thermal cycler conditions on Quantstudio 7 Flex. For dPCR assay, Probe PCR (4X) kit from Qiagen was used with recommended concentrations of primers / probe for 24 well, 26K Nanoplate in the master-mix. The thermal cycler conditions were followed as per the Qiagen QIAcuity Four protocol recommendations. Briefly, the reaction began with standard priming profile of the samples in reaction nanoplate, followed by cycling conditions which included PCR initial heat activation for 2 minutes at 95°C, 40 cycles of 15 seconds denaturation at 95°C, and combined annealing / extension for 30 seconds at 60°C.

[0170] A comparison between qPCR and dPCR showed great correlation in terms of dilutional linearity and primer specificity between two methods when data from serially diluted Vero DNA was compared. The first two dilutions of standard (5 ng and 0.5 ng DNA input per reaction) in dPCR reached the highest detection limit of the dPCR instrument and could not be quantitated. Data shown below in Table 3.

[0171] Table 3 Raw data from qPCR and dPCR using dilutions of in-house prepared Vero DNA

[0172] (without RE Xbal in the MM)

[0173] The first two dilutions of Vero DNA were not detected as they were beyond the upper detection limit of the assay. To successfully quantitate positive partitions, negative partitions are needed to set up the threshold of detection. Saturation, therefore, prevents calculation of estimation of copy number in the sample. The signal in the form of copies / pL only started to appear at STD3 (3rd dilution, 0.05ng / reaction) of the control Vero PC DNA. There was a saturation warning for the following wells: A1 , A2, A3, B1 , B2, B3, C3, D1 , D2, D3, E1 , E2, H3.

[0174] The concentration calculated by dPCR machine in STD3 dilution was slightly different for mastermix that contained Xbal restriction enzyme (MM2) versus the non restriction enzyme containing mastermix (MM1) (i.e., 1627.4 (MM1) vs 2195.10 (MM2)). The concentrations calculated for STD4 (well B1 , 178.60) and STD5 (B2, 15.99) were slightly higher for MM1 than the concentrations in STD4 (well D3, 163.00) and STD5 (well E1 , 12.66) for MM2. There was no detection of non-specific DNA (HEK-293) at 5000 ng / mL in both mastermixes, and there was no signal detected in NTC (water, non-template control) wells using MM1 . Each Vero DNA dilution, HEK-293, and NTC were tested in single reactions whose 1 D-scatterplots are provided in Figure 3. The concentrations were calculated by the Poisson distribution rule.

[0175] Many other restriction enzymes (RE) such as EcoRI, BamHI, Alul, Hindlll-HF, Xmnl were also tested in subsequent experiments to achieve accurate and precise detection efficiency and improved quantitation of the DNA; however, all except Alul and Hindlll-HF failed to provide significantly different results than those observed with RE Xbal. The mastermix containing 1 L of (100,000 U / mL) Hindlll- HF (NEB R3101T) per 40 pL reaction has been able to show better quantitation of DNA in TDV-3, Day 7 (D7) stabilized (at 10-4dilution), post AEX (neat) and BDS (neat) samples. It would be recommended to add RE Hindlll-HF in the mastermix when analyzing TDV IPC and EPC samples at their recommended dilutions. Vero DNA can also be used as an assay positive control (PC) with RE Hindlll- HF in the mastermix.

[0176] 1.2 Matrix Interference testing - test use of assay

[0177] During manufacturing of vaccines, the presence of residual HC-DNA can be quantitated in samples collected from different stages of the manufacturing, for example for testing stabilized sample, post AEX (Anion Exchange Chromatography) and BDS (Bulk Drug Substance). The composition of the buffers in these samples will be different at different stages of the downstream process. It is critical to evaluate the overall impact of buffer matrix on the performance of the assay. Using TDV-3 as a model, samples from each of these stages harvested on day 7 post propagation in Vero cells were evaluated in this exercise.

[0178] These samples were subjected to either DNA extraction or no extraction process based on magnetic beads using Thermo’s nucleic acid extraction kit. The rationale is that because of the small volume and extensive dilution needed to run dPCR, any inhibitory molecules should not interfere with the PCR process. Furthermore, being an end-point readout, even if PCR efficiency may differfrom one partition to the other, the binary classification of them overcomes any local variability. Extraction could, therefore, become optional. Some samples were also tested post spiking with a known amount of Vero stock DNA in the sample.

[0179] The purpose of testing these samples was to determine if it would be critical to isolate DNA from them priorto quantitation in dPCR, if a spiked amount of Vero DNA could be recovered post extraction, and finally what is the limit of detection of DNA in these samples.

[0180] 1.2.1 TDV-3 D7 Stabilized Harvest

[0181] The test article was diluted with 1X PBS buffer to make three serial 10-fold dilutions, from 1 / 10 to 1 / 1000. Two separate sets of aliquots (2 aliquots in each set, with 100 pL per aliquot) of undiluted (neat) and each of the three dilutions were made using 1 .5 mL Eppendorf tubes. The firstset of aliquots for each sample, neat or diluted, were stored at 4°C until needed for dPCR assay the next morning. The second set of sample aliquots was subjected to DNA isolation with magnetic beads using Thermo’s nucleic acid extraction kit and reagents. As a control, 1X PBS bufferwas spiked with known amount of Vero DNA at two levels (10 pL of DIL3 (0.1 ng) or 10pL of DIL6 (0.1 pg) into 100 pL of 1X PBS). Figure 4 provides a flow-chart showing the hierarchy of sample preparation process.

[0182] No detection was observed by dPCR in any sample (neat or diluted, extracted or not extracted), except spiked PBS. The signal for green FAM channel reached the saturation in sample wells. PBS spiked with STD DIL3 of Vero DNA displayed 205 copies / pL, close to 179 copies / pL in STD DIL4. An Input as low as of 5 fg (5 pL of STD DIL6 in 40 pL dPCR reaction) in dPCR reaction can be recovered and detected which is equivalent to 0.001 ng / mL of the original DNA stock sample.

[0183] The same set of TDV-3 samples (extracted or not) were also analyzed in qPCR assay to investigate if the lack of detection in dPCR is due to an impact of matrix inhibition or the absence of quantifiable DNA. qPCR results confirmed the presence of DNA in all samples except non-extracted neat samples which inhibited the detection of signal. The assay was able to detect and quantitate DNA amount in extracted neat sample with Ct (mean) value of 7.76. Ct (mean) values of all dilutions (1 / 10-1 / 1000) revealed at least 3X difference in quantitation of DNA when compared extracted vs non extracted samples. These findings clearly indicate benefits of DNA purification of TDV-3 day 7 (D7) stabilized sample to remove possible inhibitors in qPCR reactions. The repeat dPCR assay on a different aliquot of TDV-3 D7 Stabilized harvest sample (neat and diluted up to 1 / 1000) continued to show no valid outcome (copies / pL) due to signal exceeding (saturation) 24-well 26K Nanoplate’s upper limit of detection. The data table from this repeated dPCR is not shown in this report, but 1 D- scatterplots are provided in Figure 5 below.

[0184] After observing matrix interference effect (saturation due to high DNA contents) of neat and diluted (up to 1 / 1000) TDV-3 D7 Stabilized harvest sample in dPCR assay, the samples were also tested in qPCR assay to confirm inhibition in another platform. In a separate experiment, a new vial of TDV-3 D7 Stabilized sample freshly taken from <65°C storage was also tested neat and diluted (1 / 10-1 / 1000) to confirm the results of first dPCR assay.

[0185] After observing continuous inhibition of detectable signal in dPCR at 1 / 1000 dilution, TDV-3 stabilized sample was further diluted up to 10'8to help eliminate matrix effect in dPCR assay. Three sets of 100 pL aliquots for each of 10-3, 1 O'4, 10'6, and 10'8dilutions were made with 1X PBS. The first set (1 O'3, 10'4, 10-6, and 10'8) was used directly in dPCR reaction without spiking with in-house prepared Vero DNA or purifying nucleic acids from it using magnetic beads method. The second set (1 O'4, 10'6, and 10'8) was not spiked with Vero DNA, but it was subjected to nucleic acid purification prior to adding into dPCR reaction plate. Third and last set (1 O'4, 10'6, and 10'8) was spiked with 10 pL (equals to 100 pg) of DIL3 ofVero DNA followed by purification of nucleic acid purification priordPCR reaction. Spiked 1X PBS at two different levels and positive & negative controls were also included in this dPCR assay. Table 4 provides dPCR data on higher dilutions of TDV-3 D7 stabilized harvest sample. TDV-3 D7 stabilized harvest when diluted at least 1 / 10000 (10'4) can be detected (1095.5 copies / pL of reaction) without needing to extract the DNA. The testing at lower dilution (10'3) still showed no quantitation (only 1 positive partition) due to the high amount of DNA present in the sample causing the saturation of the wells, with or without extraction of nucleic acids. The samples at dilutions 10'4, 10'6, and 10'8showed better concentration (>2 times) when DNA was extracted prior to testing. The quantitation of DNA concentration in non-extracted TDV-3 D7 stabilized harvest sample vs extracted at 10'4dilution could be lower, but it can still be measured and meet the requirement of detection limit of 35 ng / dose for drug substance.

[0186] Table 4 dPCR data on higher dilutions of TDV-3 D7 stabilized harvest

[0187] Sp = Sample

[0188] Saturation Warning: During imaging the signal reached the saturation for the following wells and channels:

[0189] Channel: Green Wells: A1, A3

[0190] Figure 6 shows 1 D-scatterplots of 10-fold diluted TDV-3 post-stabilization sample with different combinations of treatment like extraction or non-extraction of DNA by magnetic beads, spiked or nonspiked stabilized sample, and its dilutions with known amount of Vero DNA. Assay controls are spiked 1X PBS, DIL4, and DIL7 of Vero DNA as positive control (PC) and nuclease-free water as a negative control (NC).

[0191] 1.2.2 TDV-3 D7 Post AEX Harvest

[0192] The test article was diluted with 1X PBS buffer to make three serial 10-fold dilutions, from 10'1to 10'3. Like the sample preparation flow chart of Stabilized sample, two separate sets of aliquots (2 aliquots in each set, with 100 pL per aliquot) of undiluted (neat) and each of the three dilutions were made using 1 .5 mL Eppendorf tubes. The first set of aliquots for each sample, neat or dilution, were stored at 4°C till needed for dPCR assay next day. The second set of aliquots of samples was subjected to DNA extraction with magnetic beads using Thermo’s nucleic acid extraction kit and its reagents. As a control, 1X PBS buffer was spiked with known amount of Vero DNA at two levels (10 pL of DIL3 (0.1 ng) or 10 pL of DIL6 (0.1 pg) into 100 pL of 1X PBS).

[0193] Detection by dPCR was observed in all samples. No sample reached the saturation for green FAM channel. Based on historical qPCR data, the amount of res HC-DNA present in AEX sample is expected to be below 0.01 ng / mL. Post purification, a reaction with 100pg input of spiked Vero DNA (10pL of DIL3) in neat AEX sample (Nanoplate well C3) measured 190 copies / pL of the reaction - approximately 6X more than the 33 copies / pL measured in non-spiked purified neat AEX sample (Nanoplate well B2). The concentrations of DNA copies / pL in non-purified and purified neat TDV-3 D7 AEX sample were 21 and 33, respectively. There is a <2X (-1.57X) difference in concentration (copies / pL) obtained in extracted versus non-extracted neat AEX samples. Extracted samples displayed higher copy numbers at each dilution than those for non-extracted samples (i.e., 33 vs 21 copies / pL in neat AEX sample). Since the expected amount of res HC-DNA in AEX is very low, it is possible to avoid the purification and measure the genomic DNA copies from a neat AEX sample and then multiple by 1.57 to determine final copies / pL. This multiplication factor may or may not be accurate due to the small sample size evaluated here, and it may change with more data collected in the future. Both high- and low-level spiked DNA amount was successfully recovered post extraction as it is apparent by comparing concentrations of non-spiked versus spiked samples.

[0194] PBS spiked with STD DIL3 (100pg in 100pL PBS pre-extraction) post extraction displayed 157 copies / pL (Nanoplate well F2) and close to 179 copies / pL in STD DIL4 with same input in the dPCR reaction. An Input as low as 5 fg (5pL of STD DIL6 in 40 pL dPCR reaction) in a dPCR reaction (nanoplate well F3, 8 copies / uL) can be recovered and detected.

[0195] Figure 7 shows 1 D-scatterplots of neat and 10-fold diluted TDV-3 Post AEX sample with different combinations of treatment, such as extraction or non-extraction of DNA by magnetic beads, spiked or non-spiked AEX sample, and its dilutions with known amount of Vero DNA. Assay controls are spiked 1X PBS, DIL3 and DIL5 dilutions of Vero DNA are positive controls (PC) and nuclease-free water is a negative control (NC).

[0196] 1.2.3 TDV-3 D7 BPS

[0197] The test article was diluted with 1X PBS buffer to make three serial 10-fold dilutions, from 10'1to 10'3. Like the sample preparation flow chart of stabilized sample, two separate sets of aliquots (2 aliquots in each set, with 100 pL per aliquot) of undiluted (neat) and each of the three dilutions were prepared using 1 .5 mL Eppendorf tubes. The first set of aliquots for each sample, neat or diluted, were stored at 4°C till needed for dPCR assay next day. The second set of aliquots of samples was subjected to DNA extraction with magnetic beads using Thermo’s nucleic acid extraction kit and its reagents. As a control, 1X PBS buffer was spiked with known amount of Vero DNA at two levels (10 pL of DIL3 (0.1 ng) or 10 pL of DIL6 (0.1 pg) into 100 pL of 1X PBS). The detection was observed in all samples. No sample reached the saturation for green FAM channel. Based on historical qPCR data, the amount of residual HC-DNA present in BDS sample is expected to be below 0.01 ng / mL. Post purification, a reaction with 100pg input of spiked Vero DNA (10pL of DIL3) in neat BDS sample (nanoplate well F2) measured 518 copies per L of the reaction which is about 1 ,3X more than the 399 copies / L measured in non-spiked purified neat BDS sample (nanoplate well E1). The concentrations of DNA copies / pL in non-purified and purified neatTDV-3 D7 BDS sample were 244 and 399, respectively. There is a <2X (~1.63X to be precise) difference in concentration (copies / pL) obtained in extracted versus non-extracted neat BDS samples. Extracted samples displayed higher copy numbers at each dilution than the ones for non-extracted samples (i.e., 399 vs 244 copies / pL in neat BDS sample). Since the expected amount of res HC-DNA in BDS is very low, it is possible to avoid the purification and measure the genomic DNA copies from a neat BDS sample and then multiply by 1 .63 to determine final copies / pL. Both high- and low-level spiked DNA amount was successfully recovered post extraction as is apparent by comparing concentrations of non-spiked versus spiked samples.

[0198] PBS spiked with STD DIL3 (100pg in 100pL PBS pre-extraction) post extraction displayed 130 copies / pL (Nanoplate well G3) and close to 179 copies / pL in STD DIL4 with same input in the dPCR reaction. An Input as low as of 5 fg (5 pL of STD DIL6 in 40 pL dPCR reaction) in a dPCR reaction (nanoplate well H1 , 7 copies / uL) can be recovered and detected.

[0199] Figure 8 provides 1 D-scatterplots of neat and 10-fold diluted TDV-3 Post BDS sample with different combinations of treatment like extraction or non-extraction of DNA by magnetic beads, spiked or nonspiked AEX sample, and its dilutions with known amount of Vero DNA. Assay controls are spiked 1X PBS, DIL3 and DIL5 dilutions of Vero DNA are positive controls (PC) and nuclease-free water is a negative control (NC).

[0200] 1.2.4 BDS Buffer

[0201] BDS buffer was prepared with the same ratio of each of the components that simulate the buffer in which the final TDV BDS resides in. Neat and 10-fold diluted buffer was spiked with two levels of known amounts (10pL of DIL3 (0.1 ng) or DIL6 (0.1 pg) mixed into 100pL of buffer) of Vero DNA. One set of spiked neat and diluted buffers were subjected to nucleic acid purification prior to testing in dPCR. The other set was kept at 2-8°C until testing with dPCR along with extracted samples. Spiked Vero DNA into 1X PBS at two different concentrations and positive & negative controls were also included in this dPCR experiment.

[0202] In parallel, these same samples were also tested in qPCR assay to confirm sensitivity and detection level in two different platforms. Data from the qPCR assay shown in Table 5 suggest no major impact of BDS buffer matrix on the detection of spiked DNA (both spiking levels), as there is not much difference in Ct mean values and quantity amongst dilutions of BDS buffers that were spiked with same amount of DNA. Extraction does not make a big difference, except there was about 1 Ct mean or less difference between non-extracted (20.63, Ct mean of last three dilutions) and extracted (19.63, Ct mean of last three dilutions) samples spiked with higher amount (DIL3, 0.1 ng) of DNA.

[0203] Table 6 includes dPCR assay results showing no real matrix effect of BDS buffer in detection of spiked DNA, as there is not much difference in copies / pL (4608.6 (neat), 4568.1 (1 / 10), 4668.9 (1 / 100), and 4556.8 (1 / 1000)) and quantity amongst dilutions of BDS buffers that were spiked with same amount (DIL3, 0.1 ng) of Vero DNA. The mean, standard deviation, and coefficient of variation (CV) of copies / pL was 4600.4, 43.88 and 0.95%, respectively. The overall concentration (copies / pL) values for all the samples seem to be higher than normally observed values for samples spiked with DIL3 (0.1 ng) of Vero DNA. Even though the numbers are higher, they are consistent throughout the plate. The data show complete recovery of spiked DNA regardless of spiking into undiluted or diluted BDS buffer. Not all samples with DIL6 (0.1pg) showed signal in terms of copies / pL. Non-extracted BDS buffer spiked with DIL3 (0.1 npg) showed even better concentration than extracted samples. Table 6 dPCR Data on BDS Buffer Spiked with Known Amount of Vero DNA

[0204] Sp = Sample

[0205] Saturation warning: During imaging the signal reached the saturation for the following wells and channels:

[0206] Wells: A1 , A2, A3, B1 , B2, B3, C1 , C2, D1 , D2, D3, E1 , E2, E3, F1 , F2, F3, G1 , G2, G3, H1 , H2, H3

[0207] 2.0 Method Performance Testing

[0208] 2.1 Limit of Detection

[0209] Limit of detection (LOD) describes the smallest concentration (in this case, copies / pL) of an analyte that can be detected by an analytical method. Using optimized thermal cycler conditions and 4X QIAcuity Probe PCR mastermixes, with and without RE Hindi ll-HF and Alul, 10-fold serial dilutions of Vero DNA, starting from DIL4 (DNA diluted to 1 / 105at expected concentration of 1 ng / mL) to DIL7 (DNA diluted to 1 / 108at expected concentration of 1 pg / mL) were tested. The data show that 1 pg / mL of DNA in DIL7 (equivalent to 5fg / 40pL reaction) can reliably be detected (>13-15 copies of a-SAT gene per pL of the reaction) in the presence of restriction enzyme (RE) in the mastermix. Table 7 provides a summary of the results.

[0210] Table 7 Detection Limit of Vero DNA by dPCR using Known Amount of Vero DNA and Optimized Assay Conditions

[0211] Dilutions of pg / 5pL / Cone. copies / pL Cone. copies / pL Cone. copies / pL

[0212] Vero DNA 40pL reaction (No RE) (Hindlll-HF) (Alul)

[0213] DIL5 0.1 ng 0.5pg 46 2109 2202

[0214] DIL6 0.01 ng 0.05pg 5 139 209

[0215] DIL7 0.001 ng 0.005pg 1 13 15

[0216] Cone, concentration; RE, restriction enzyme.

[0217] 2.2 Quantitation Limit

[0218] 2.2.1 Upper Limit Of Quantification (ULOQ)

[0219] The upper limit of quantification (ULOQ) is the maximum concentration of analyte that can be reliably and reproducibly measured by a given instrument. 24-well 26K Nanoplates offer a maximum detection of 22,000 - 130,000 copies per 40pL reaction (equivalent to 550 - 3250 copies / pL). Only 24.3 pL of 40 pL will be used for analysis and the determination of copies / pL. Based on the data from serially diluted Vero DNA, the ULOQ was determined to be 0.1 ng / mL in DIL5 (equivalent to 0.5pg / 40pL reaction) of Vero DNA which has precisely given approximately 2200 copies / pL reaction volume (see Table 7). DIL4 (1 ng / mL) produces inconsistent copies / pL, exceeds the upper limit of measurement of the 24-well 26K nanoplate, and saturates the well for FAM channel; DIL4 cannot be considered as a ULOQ. Subsequently, DIL4 does not provide an appropriate distribution of positive and negative partitions to allow for accurate concentration (copies / pL) determination through Poisson statistics. 2.2.2 Lower Limit Of Quantification (LLOQ)

[0220] The lower limit of quantification (LLOQ) is the lowest concentration of an analyte which can be measured with high precision and accuracy. Based on the data and analysis, the LLOQ was determined to be 0.01 ng / mL in DIL6 (equivalent to 0.05pg / 40pL reaction volume) of Vero DNA, resulting in approximately 139-209 copies / pL reaction volume.

[0221] In a separate experiment the LLOQ was determined as follows:

[0222] Sample Preparation

[0223] All samples evaluated for Accuracy, Precision, and Linearity were used to evaluate the Range, and thus were used to evaluate the lower limit of quantitation (Samples 1 & 3). As all EPC test samples were tested neat (undiluted), Sample 3 was prepared with the same BDS matrix as EPC samples to estimate LLOQ (refer to Table 12 for LLOQ requirements for IPC and EPC samples). Refer to VAL-536292 and Table 2 for details of Sample 1 and 3 preparations.

[0224] Results. Requirement and Evaluation

[0225] The Lower Limit of Quantitation (LLOQ) corresponded to the highest dilution at which linearity, accuracy, and precision were acceptable, based on the quantitation limits established in Analytical Target Profile REC-380338 for TDV-3 Stabilization (0.015ng / mL) and PC DNA in BDS Matrix (0.0015ng / mL) samples. The results for LLOQ for Samples 1 and 3 are presented below. For Sample 1 , TDV3 / IPC / Stabilized Harvest diluted into Stab / AEX buffer at 10-4included the second highest dilution (1 .25x10®) with a Geometric Mean dPCR = 19.95 copies / pL, and DNA amount by dilution = 3.00x10-4ng / mL. For Sample 3, Vero PC DNA diluted into BDS buffer at 106, linearity, precision, and accuracy were acceptable (%CV = 22.6) at the third highest dilution (2.5x107). At this dilution, the Geometric Mean dPCR = 75.86 copies / pL, and the DNA amount by dilution = 1 .15x10-3ng / mL.

[0226] 2.3 Linearity

[0227] Eight 10-fold dilutions of in-house prepared Vero DNA were tested in dPCR using Qiagen’s 4X PCR Probe master mix with a-SAT primers / probe set. The first two dilutions (5 ng and 0.5 ng DNA per 40|jL reaction) were out of the higher detection limit of QIAcuity instrument and reached the saturation points of the channel. Valid positive and negative partitions for the last six dilutions (from 0.05 ng / reaction to 0.5 fg / reaction) were measured and concentrations in copies / pL using Poisson’s distribution equation were calculated. The linear response (R2= 0.9893) of Vero DNA copies / pL to Vero DNA serial dilutions in master mix with a-SAT primers / probe set is presented in Figure 9.

[0228] 2.4 Range

[0229] Eight successive 10-fold dilutions of Vero DNA were tested in dPCR using PCR Probe master mix with a-SAT primers / probe set. The first two dilutions (5 ng and 0.5 ng DNA per 40pL reaction volume) were beyond the higher detection limit of QIAcuity instrument and reached the saturation points of the channel. Valid positive and negative partitions for the last six dilutions (from 0.05 ng / reaction to 0.5 fg / reaction) were measured and concentrations in copies / pL using Poisson’s distribution equation were calculated. These last six dilutions can be used to determine the range of detection of target sequence in a sample. The upper and lower values of the detection range are 0.05 ng / reaction and 0.5 fg / reaction, respectively, and represent where the assay can detect at least one or less positive partition with concentration of about 0.055 copies / pL.

[0230] 2.5 Specificity

[0231] Specificity was tested using human DNA (ATCC) isolated from HEK293 cells. The reaction mix contained master-mix components of Qiagen’s 4X Probe PCR kit and a-SAT primers and FAM-probe.

Claims

C l a i m s1 . A digital PCR (dPCR) method for determining the concentration of residual Vero cell genomic DNA in a sample containing virus obtained from production in Vero cells comprising the steps of a) preparing serially diluted samples without prior DNA extraction; b) contacting the serially diluted samples of step a) separately with a composition comprising:(i) a set of primers suitable for amplification of an alpha satellite target sequence, wherein the alpha satellite target sequence comprises the nucleotide sequence set forth in SEQ ID NO: 1 ,(ii) a labeled probe specific for the alpha satellite target sequence,(iii) a DNA polymerase, and nucleotides, and(iv) a restriction enzyme, c) incubating the mixture of step b) under conditions allowing a restriction enzyme digest of the mixture; d) partitioning the digested mixture of step b); e) PCR-amplifying the alpha satellite target sequence in the partition of step d) ; f) determining the binding of said labeled probe to the PCR product and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample.

2. The dPCR method of claim 1 , wherein the alpha satellite target sequence consists of the nucleotide sequence set forth in SEQ ID NO: 1.

3. The dPCR method of claim 1 or 2, wherein the forward primer has the nucleotide sequence set forth in SEQ ID NO: 2 or a variant thereof having at least 80 % sequence identity; and the reverse primer has the nucleotide sequence set forth in SEQ ID NO: 3 or a variant thereof having at least 80 % sequence identity4. The dPCR method of any one of claims 1 to 3, wherein the labeled probe has the nucleotide sequence set forth in SEQ ID NO: 4 and preferably the probe is a Taqman probe.

5. The dPCR method of any one of claims 1 to 4, wherein the restriction enzyme is selected from Hind 111 or a genetically engineered variant thereof, preferably the restriction enzyme is a genetically engineered Hindi 11 variant having reduced star activity6. The dPCR method of any one of claims 1 to 5, wherein the dPCR method is carried out as a droplet dPCR method or as a microfluidic device-based dPCR method, preferably the dPCR method is a nanoplate-based dPCR method.7 The dPCR method of any one of claims 1 to 6, wherein the dPCR method comprises d) partitioning the digested mixture of step b) into at least 50 partitions of a well of a nanoplate; e) PCR-amplifying the alpha satellite target sequence in the at least 50 partitions; f) determining the binding of said labeled probe to the PCR product in the at least 50 partitions and calculating therefrom the concentration of residual Vero cell genomic DNA in the virus-containing sample.

8. The dPCR method of any one of claims 1 to 7, wherein the digested mixture of step b) is partitioned into 8500 or 26000 partitions per well of the nanoplate, preferably 26000 partitions per well.

9. The dPCR method of any one of claims 1 to 8, wherein(i) the final concentration of the forward primer is about 0.2 pM to about 1.2 pM, preferably about 0 7 pM to about 0 9 pM;(ii) the final concentration of the reverse primer is about 0.2 pM to about 1.2 pM, preferably about 0.7 pM to about 0.9 pM; and(Hi) the final concentration of the probe is about 0 2 pM to about 1.2 pM, preferably about 0.3 pM to about 0.5 pM.10 The dPCR method of any one of claims 1 to 9, wherein the PCR-amplifying step e) comprises i) initial heat activation for about 2 minutes at 95°C; and ii) about 15s at 95°C (denaturation) followed by about 30s at 60°C (combined annealing / extension) over about 30 to about 50 cycles, preferably 40 cycles.11 The dPCR method of any one of claims 1 to 10, wherein the calculation step f) comprises conversion of the alpha satellite copy number / pl sample to ng Vero DNA / ml sample by using the formula:— = about 1x10-5to about 3x105(Copies mean)(dilution factor)12 The dPCR method of claim 11 , wherein the formula is:— = (Copies mean)(1 .4993 x 10-5)(dilution factor).13 The dPCR method of any one of claims 1 to 12, wherein the virus is selected from hepatitis virus, influenza virus and a flavivirus.14 The dPCR method of claim 13, wherein the flavivirus is selected from dengue virus, yellow fever virus, Zika virus, West Nile virus, Japanese encephalitis virus, St Louis encephalitis virus and tick-borne encephalitis virus and combinations thereof, preferably the flavivirus is dengue virus.15 The dPCR method of any one of claims 13 or 14, wherein the flavivirus is a live attenuated or a chimeric virus flavivirus, preferably, the live attenuated flavivirus comprises one or more of live attenuated dengue viruses, more preferably, the live attenuated flavivirus comprises one or more of TDV-1 , TDV-2, TDV-3 and TDV-4.16 The dPCR method of any one of claims 1 to 15, wherein the sample containing flavivirus produced in Vero cells is an in-process monovalent drug substance, a bulk drug substance or a final tetravalent drug product.17 The dPCR method of any one of claims 1 to 16, wherein the sample containing flavivirus produced in Vero cells comprises one or more pharmaceutically acceptable excipients, preferably the sample comprises F127, trehalose and / or human serum albumin.18 A quality control method for determining the amount of residual Vero cell genomic DNA in a sample from a manufacturing process of a flavivirus vaccine comprising performing the method of any one of claims 1 to 17 on an in-process control (IPC) sample or an end product control (EPC) sample of said process.19 A combination of quality control methods for vaccines containing live, attenuated dengue virus comprising performing the method according to claim 18 and at least one further method selected from the group consisting of immunofocus assay, identity assay, attenuation check, visual inspection, determination of reconstitution time or resuspendability of lyophilisates, sterility test, test for bacterial endotoxins, determination of pH, colorimetric determination of water, determination of osmolality, determination of the content of one or more excipient(s)

Images