WO2024088585A1 - Method for generation of tumor-adapted oncolytic viruses - Google Patents

Abstract

The present invention relates to a method for generating personalized oncolytic RNA viruses, improving known oncolytic RNA viruses, or circumventing the resistance of tumor cells against such RNA viruses. The method comprises performing of a minimalistic number of passages in e.g. a tumor cell line or in patient tissue extracted from a tumor and then identifying mutations occurred during the early passages. The method further comprises the generation of a cDNA clone comprising at least one, most or all mutations identified and propagating a correspondent oncolytic RNA virus, which carries at least one, most or all mutations of the RNA virus adapted in the method for use in a tumor treatment.

Description

Method for generation of tumor-adapted oncolytic viruses

The present invention relates to a method of generation of personalized oncolytic RNA viruses which improves known oncolytic RNA viruses, circumvent the resistance of tumor cells against the infection and increases the lysis by RNA oncolytic viruses, provides cDNA clones of the improved and/or personalised oncolytic RNA viruses as well as oncolytic RNA viruses harvested from cells transfected with said cDNA of the oncolytic RNA virus. Furtherthe present invention provides improved or personalised viruses for use in the treatment of cancer.

Background

Advances in genetic engineering and the development of new methods have led to significant advances in the field of oncolytic virotherapy in recent years. Some oncolytic viruses (OV) are already available as therapeutics such as Oncorine®, T-Vec or CAVATAK forthe treatment of cancer or are in late-stage clinical testing (Sanjuan and Grdzelishvili 2015). With the help of genetic engineering methods, an armed oncolytic virus can be generated by introducing transgenes such as GM-CSF, which trigger an increased adaptive immune response. OV attenuation can also be achieved by deleting virulence genes (Sanjuan and Grdzelishvili 2015; Harrington et al. 2019). However, the approach of targeted manipulation of OVs through genetic engineering is limited in part by the complexity and variety of virus-host interactions. This is a particular challenge in the context of virus-tumor cell interactions, as tumor cells can vary greatly between cancer types, and between and within patients (Burrell et al. 2013).

Adaptation of viruses using directed evolution offers an alternative approach to obtain new OVs, improve existing OVs, or overcome tumor cell resistance to infection and lysis by OVs without precise knowledge of the underlying molecular properties and signaling pathways (Sanjuan and Grdzelishvili 2015; Bauzon and Hermiston 2012).

Directed evolution is a useful tool to improve the fitness and selectivity of oncolytic viruses without making genetic modifications. Adaptation is not limited by the complexity of virus-host interactions since oncolytic viruses are naturally adapted by serial passage. The basis of the ability to adapt is the high error rate of the viruses during genome replication. This leads to the formation of new quasispecies, i.e., a mixture of genetically different viruses. During the adaptation process, one virus species of the quasispecies then prevails and makes up the main population. This is a typical phenomenon, particularly in the case of RNA viruses. The selectivity or oncolytic ability improved by adaptation can therefore be attributed to accidentally acquired mutations. However, it is not possible to predict which mutation will occur.

For adaptation by directed evolution in vitro, the oncolytic virus is passaged several times e.g. on a tumor cell line. In each passage, the resistant tumor cell line is infected, the supernatant containing newly generated viruses is taken after a certain time and is given to fresh tumor cells. This is repeated until an adapted virus population has developed. Through the passaging, an adapted oncolytic virus is created, which e.g., has improved replication and increased toxicity compared to the parent virus (see Figure 1). Mutations responsible for the adapted phenotype can then be identified by sequencing the viral genome (Sanjuan and Grdzelishvili 2015; Zainutdinov et al. 2019; Bauzon and Hermiston 2012).

Since DNA viruses have a lower mutation rate than RNA viruses due to the proofreading function of the polymerase, a higher number of passages is necessary for the adaptation of DNA viruses. Due to the higher mutation rate, RNA viruses can therefore be adapted more quickly through directed evolution (Sanjuan and Grdzelishvili 2015; Zainutdinov et al. 2019). The mutation rate can additionally be increased by using chemical mutagens or genetically modified polymerases. The analysis of independently adapted virus populations makes it possible to differentiate between driver mutations that are largely responsible for adaptation and random mutations caused by genetic drift. Further, the adaptation and specialization of the OV to a specific tumor cell lineage can lead to reduced fitness in other cell types (Sanjuan and Grdzelishvili 2015; Zainutdinov et al. 2019; Bauzon and Hermiston 2012).

It was known that in Sanger sequencing at early passages during such an adaptation process double peaks, indicating that the virus population consists of different, at least two virus genotypes, were found, one representing the genotype of the parental virus one representing a new viral genotype. In the course of further passaging, the virus population corresponding to the signal of the parental genotype disappears and the signal of the new viral variant genotype is becoming more and more homogeneous. Seegers et al. (2020) observed this effect in the adaptation of a vesicular stomatitis virus (VSV - Vesicular stomatitis virus) to pancreatic carcinoma cells. For E238K, a double signal can be seen from passage 10 to passage 30. A homogeneous virus population having the mutation E238Kwas only present after 33 passages (Figure 2). It is known that directed evolution has the potential to improve oncolytic viruses in terms of their replication behaviour and their oncolytic effectivity. Thus, one would assume that the potential of oncolytic viruses for a tumor treatment would be prioritised.

In favour of this assumption it was shown, that e.g. the chimeric oncolytic adenovirus ColoAdl, which is already in clinical trials for the treatment of CRC, was generated using directed evolution on CRC cells over 20 passages, which corresponds to typically 20 weeks. This enabled a significantly increased replication and a higher selectivity for colorectal carcinoma cells to be achieved (Kuhn et al. 2008). Also, Gao et al. was able to increase the viral replication of a recombinant VSV by three log levels in 15 passages by serial passaging on D2F2/E2 cells. The sequencing of the adapted VSV revealed two mutations whose relevance was confirmed in mutagenesis studies by the vector-based production of recombinant VSV (Gao et al. 2006). Seegers et al. was able to generate novel oncolytic VSV with an improved ability to replicate in previously virus-resistant pancreatic adenocarcinoma cell lines using the directed evolution approach and passaging over 32 passages, which corresponds to 32 weeks or 8 months (Seegers et al. 2020).

During the adaptation of coxsackieviruses by directed evolution, mutations in the capsid proteins VP1 - VP4 have been identified as the cause of the adaptation. These mutations resulted in changes in the receptor tropism and also enabled the binding of new receptors. For example, Svyatchenko et al. (2017) was able to adapt a coxsackievirus B6 to the partially resistant tumor cell lines RD, A431 and MCF7 using 15 passages. A subsequent sequencing of the adapted viruses showed mutations in the VP1, VP2 and VP4 proteins. The adaptation of the coxsackievirus A21 to ICAM- negative rhabdomyosarcoma cells (RD cells) was also made possible by two mutations in the VP3 capsid protein.

In summary, the number of serial passages in direct evolution approaches was typically between 15 and 50 passages, the exact number is highly dependent on the virus and the selection pressure (Sanjuan and Grdzelishvili 2015). This means that even with a high potential for therapeutic approaches the time span to personalise an oncologic virus is to long for tumor patients to survive until a personalised therapy would be available.

Unfortunately, it is often found that primary tumors or tumor cells are resistant to infection with oncolytic viruses due to e.g. lack of expression of the viral receptors on the tumor cell surface. Consequently, a patient bearing such a tumor which lacks viral receptors, is resistant and cannot effectively be treated with an oncolytic virus. While it is known that in general the resistance of the tumor cells to oncolytic viruses can be overcome by adaptation of a virus to such tumor cells, in reality the adaption process is too slow and often might take up to several months as mentioned above.

This is why this technique has been sparely studied in the context of virotherapy in cancer. Cancer patients with a short life expectation often do not have this time, so developing tumor-adapted personalized oncolytic viruses is not considered an option for them.

Therefore, to make use of the personalized oncolytic viruses for the patient’s benefit there is an urgent need for a method of adaptation of the oncolytic viruses to particular patient derived tumor cells which would allow the creation of such personalized viruses in a reasonable time span.

Description of the invention

The present invention provides the following solution to the above problem.

Provided is a method for generating personalized oncolytic viruses, and thereby improving known oncolytic viruses, and/or circumventing the resistance oftumor cells against the infection and lysis by such oncolytic viruses, comprising: a) infecting in vitro patient-derived tumor cells or a tumor cell line closely related to a patient’s tumor with an oncolytic virus, here addressed as parent virus or founder virus. b) performing multiple passages of the virus shedding from the infected tumor cells or the tumor cell line to a fresh culture of such cells until at least one adapted RNA virus variant is achieved and/or detectable; c) identifying and/or sequencing at least one, most or all mutations of the at least one adapted RNA virus variant detected at the adaptation stage of step b), said adaptation stage comprising a mixed population of genetically different RNA virus mutants; d) generating by molecular cloning techniques a modified cDNA clone of parent RNA virus comprising one, most or all mutations identified in step c).

“Oncolytic viruses” in the context of the present invention are viruses capable of infecting and lysing (break down) cancer cells but not normal cells. The term “parent, parental or founder virus” defines in the context of this disclosure the starting material of the selected oncolytic virus, which is then used as the starting point for the direct evolution. This parent virus has an interesting or assumed-advantageous therapeutic profile in the light of the specific needs of a tumor patient. Accordingly, one would choose for the treatment of colon cancer an enterogastric oncolytic virus as parent virus, and for the treatment of a brain tu or another type of oncolytic viruses with a different tissue specificity. In any case the parent virus is a known and pre-described virus strain, which ideally was already used in humans and/or has a defined safety profile.

While the method of the invention is applicable to different viruses, it may preferably be applied to RNA viruses due to the higher mutation rate RNA viruses. However, also DNA viruses undergo comparable modification during viral passages and are thus, also a starting point for the described method.

In the present specification the method is explained using a coxsackievirus strains as parent virus.

In the context of the present invention, “adapted (RNA) virus variant” means a mutated (RNA) virus, with one or several mutations compared to the parent virus, wherein the mutations lead to an adapted phenotype allowing this adapted strain e.g. to circumvent the resistance of particular tumor cells or to replicate much faster in particular tumor cells compared with the parent (RNA) virus.

A “personalized oncolytic RNA virus” in the context of the present invention means a virus strain generated following the method of the invention and specifically adapted to particular tumor cells isolated from a patient, or adapted to a particular tumor cell line closely related to a patient’s tumor; One such example would be the colon cancer cell line e,g, Colo320).

In the context of the present invention, “improving known oncolytic RNA viruses” means obtaining oncolytic RNA viruses using the method of the present invention which demonstrate an increased oncolytic activity and/or replication rate compared to the known oncolytic RNA viruses.

In the context of the present invention, “circumventing the resistance of tumor cells against the infection” means that applying the method of the invention to the known oncolytic RNA viruses and identifying most or all mutations from at least one, most or all substrains or variants in the population, and then incorporating these mutations by molecular cloning techniques into a cDNA clone of the parent virus thereby generating “modified cDNA clone” of a new oncolytic RNA virus. This adapted and new oncolytic RNA virus is then capable to infect and lyse particular tumor cells with a much higher effectivity and thus improve the therapeutic expectations.

Specifically, the method of the present invention resulting in adaptation is based on the following steps:

Step a): Using a parent virus for infecting patient-derived tumor cells or a tumor cell line, which is closely related to the patient-derived tumor cell under suitable culture conditions. The infected cells are then propagated under suitable conditions until either the newly generated viruses are released into the supernatant, or until the infected cells become lytic. In both cases the newly generated viruses are isolated from the supernatant or the culture medium and used to newly infect either the patient-derived tumor cells and/or the closely related tumor cell line in the next so-called passage.

Step b): Passaging the newly generated virus. While it depends on the virus, it can be expected that each passage typically takes 2-5, if not 7 days. During the one, two, three, four, five, six, seven, eight, nine, ten or more passages the newly generated viruses evolve and thereby collect mutations and genomic modifications, which lead to an adaptation. Thus, this step is also addressed as adaptation stage. The viruses in the adaptation stage form a mixed population, a quasispecies, with at least one or multiple changes and/or mutations in the genome of these newly evolving viruses in the mixed population.

Step c): Identifying mutations which are early on visible through sequencing of the genome or genomes of the quasispecies before an adapted virus has fully established itself as a new strain in a homogenous virus population. It is not known in the initial phase whether all occurred mutations that are found by sequencing are on the same virus genome or on different virus genomes. Regardless, as it is assumed that the mutations even if they are found in different viral genomes could act cooperatively improving the viral survival characteristics, such as speed of replication or tumor cytotoxicity of the whole population, the method of the invention intents to identify and subsequently use at least one, preferably most or all mutations identified in the quasispecies of the evolving virus population.

Step d): By molecular cloning techniques artificially combining at least one, preferably most or all occurring and/or identified mutations in one cDNA clone of the parent virus (e.g., by inserting the mutations into the genome of the parent or founder virus in a targeted manner using genetic engineering), a modified cDNA clone of an adapted virus is obtained. The advantage of the present invention is that step c) is taken at a very early point in time in the adaptation process, i.e., at a point in time when the adapted virus population is still a mixed population of genetically differentviruses i.e. a population comprising many different virus variants. In contrast to the conventional approach, the method of invention does not wait until the virus variant has fully established itself in the population and has overgrown other potential variants, but the method of the invention uses the virus at the mixed population stage (Fig. 3).

For the adaptation of the virus to resistant tumor cells, the virus is serially adapted over several passages until at least one improved phenotype is identifiable or determined which may be e.g. an increased tumor cell lysis or an improved and faster viral replication. Then the sequencing takes place using e.g., Sanger technology, to identify at least one, most or all mutations in the newly generated mixed population.

While the availability of a plasmid with the cDNA of the parental viral genome is helpful for accelerating step d) of the inventive method, it is not essential and the skilled practitioner can isolate the genomic material of the parent virus and generate a vector typically in the format of a plasmid, which is then the starting point for the molecular modification according to step d). Such a plasmid is available for the CVB3 strain PD used in the examples as parent virus (SEQ ID No.: 9).

Ultimately, with the method of the invention, one can quickly insert all mutations occurring during the adaptation stage into the cDNA of the parental virus genome. The thereby newly generated modified cDNA clone, comprising at least one, preferably most or all mutations identified, can then be used to produce the adapted personalised viruses suitable for an improved treatment of cancer. For this a production cell line is transfected with the modified cDNA clone under suitable conditions as described in step e) and then is the personalised adapted virus harvested from the culture medium or the supernatant as described in step f). Optionally, there may be the need to purify the virus harvested.

According to one aspect, the oncolytic RNA virus is a picornavirus. More preferably, picornavirus is a coxsackievirus. Even more preferably, coxsackievirus is from B group coxsackieviruses, preferably B3 strain coxsackievirus. The “B3 strain coxsackievirus” according to the present invention may be any B3 group coxsackievirus including known and classified B3 group coxsackieviruses and yet to be classified coxsackie prototype and clinically isolated viruses. It may be naturally occurring or a modified form thereof. The B3 group coxsackievirus naturally occurring when it is isolated from a patient and has not been intentionally modified in the laboratory for instance the B3 group coxsackievirus may be obtained directly from a human patient.

In contrast to this, the B3 coxsackievirus variant PD-H is a previously described laboratory strain (Hazini et al. 2021), which has been modified for safety reasons and thus shows optimised properties for use as therapeutic virus or as oncolytic virus. It is more efficient in lysing tumor cell lines than other CVB3 strains since it is able to infect such tumor cells without the presence of coxsackievirus-adenovirus-receptor (CAR) on the cell surface. This capability derives from several mutations in the VP proteins. Due to its attenuated phenotype, PD-H is additionally safe for intratumoral use and it has been shown that unspecific organ infections only occur in exceptional cases due to its mutations. This safety profile has been a prior topic, which has been successfully solved from the inventors and co-laborators by inserting microRNA target sequences into the CVB3 genome (Pryshliak et al. 2019).

Nevertheless, there are certain colorectal tumor cell lines e.g., Colo320 and LS174T (see Figure 6) that are still resistant to lysis by CVB3 viruses possessing the specific features of PD-H.

In another aspect of the invention, the method of the invention uses the multiplicity of infection (MOI) in step a) which is less than 1. For a successful adaptation of an oncolytic virus by directed evolution, the population size used at the beginning has a significant influence on the adaptation, since gene drift predominates if the MOI is too low and on the other hand if the MOI is too high lower fitness mutants arise, because cells can be infected by multiple virus particles. An MOI of less than 1 is thus preferred, further preferred is a MOI of 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1 or even less infectious units for in vitro adaptation according to the invention.

In another aspect of the invention, in the method of invention, the number of passages in step b) is between 2 and 15, 2 to 14, 3 to 14, 4 to 13, 5 to 11, 6 to 10, 5 to 9 or 7 to 9. The minimal number of passages is enough for at least one or some mutated variants to evolve, but at the same time the newly evolved variants have not yet overgrown other mutants. Thus, the mutations or the mutated variants are still multiple in the population (also considered as quasispecies) and none of them has become dominant.

The method thus allows to reduce the number of passages dramatically and still is able to identify at least one or several adapted virus variants which then are used to generate first the modified cDNA clone and then a patient adapted oncolytic RNA virus. In comparison to this the known methods wait until a homogeneous virus population is evolved and do not isolate a virus from the quasispecies.

The term “quasispecies” in the context of virology defines a group of virus variants developing in parallel and in the same host as an immune-evasion response. If isolated from the host such quasispecies comprise many different mutations and mutants or many mutated variants.

Additionally, in the above-described method of the state of the art, where an isolation of a clone takes place only if a homogenous population has been developed, such isolated clones have only particular mutations because not all RNA genomes evolved in a similar way, and many may not have come so far as to become a dominant species or optimally adapted.

In the method of the present invention the most or all mutations are already identified in the population at the stage of the quasispecies and thus at significantly lower number of passages. The at least one, most or all mutations are then used to generate much faster a personalized and adapted oncolytic RNA virus. Moreover, since all identified mutations are included by molecular techniques in one cDNA clone, which is then used to generate one homogenous population of the adapted RNA virus, the advantages being e.g. improved characteristics such as a faster replication cycle or an improved tumor toxicity, of all occurred mutations in the population are combined in one virus to the patient’s benefit.

In another aspect of the invention, the modified cDNA clone comprising at least one, most or all mutations identified in step c) is generated using recombinant techniques. For example, the recombinant technique is In-Fusion technique (Fig. 5). This technique uses enzymes which degrades two PCR fragments with 15 bp overlapping ends at the 3' end followed by annealing the complementary ends. This technology allows quick insertion of at least one, most or all identified mutations into the cDNA of a parental virus, which can then be used to generate the adapted virus and thus saves valuable time pivotal for a success of the treatment.

In brief, the In-Fusion cloning technique (Zhu et al. 2007) was used to generate according to the method of the invention the modified cDNA clones with defined mutations. For this purpose, mutagenesis primers were first designed, with the help of which cDNA clones were generated that carried one or more specific mutations. Two mutagenesis primers PD_749-M_F (SEQ ID No.: 7) and PD_749-M_R (SEQ ID No.: 8) with their corresponding binding sites in the CVB3 genome are shown in Figure 4. Forthe mutagenesis of guanine (G) to adenine (A) at nucleotide 749 in the CVB3 genome, there is a mismatch at this point in the primer (highlighted), so that the PCR produces a genome fragment with an A instead of a G at nucleotide position 749. The mutagenesis primers each overlap by 15 bp, so that the fragments could later be joined using in-fusion technology. Alternatively, established molecular cloning and/or ligation technology can be used for inserting the mutations into the cDNA clone of interest.

In another aspect of the invention, the method further comprises an additional step e) wherein one or more RNA motives capable of controlling or limiting the expression of the virus in various nontumor tissues are additionally inserted into the cDNA clone of the oncolytic parental virus. Such RNA motives are selected from the group comprising one or more shRNAs, artificial microRNAs, microRNAs, tissue specific miRNAs target sequences, and other transgenes. Small RNA molecules like tissue specific miRNAs are known to induce degradation of mRNA bearing targeting sequences for these miRNAs. By introducing specific miRNA targeting sequences, expression of oncolytic viruses in tissues other than tumor tissue is hampered or avoided, thereby improving the safety features and eventually reducing damages of other tissues.

According to further embodiments it is desirable to incorporate additional transgenes into the cDNA clone of the newly generated oncolytic virus. Such transgenes may be selected from the group of interleukins, preferably, the interleukin is neoleukin-2/15. The introduction of transgenes also improves the safety features and/or the immunological visibility of the virus infected tumor cells, thereby strengthening the immune response of the patient towards to tumor eradication.

In another aspect of the invention, step a) of the method of the invention further comprises the addition of chemical mutagens and/or genetically modified polymerases mutation. The increased occurrence of mutations observed after the treatment with chemical mutagens or using polymerases with reduced correction function allows to reduce the number of passages and thus reduces the time until a personalized virus is obtained.

Additionally, the exposure of the early infection step a) and the passages during the adaption stage in step b) with chemical mutagens and/or the addition of genetically modified polymerases inducing even more mutation is particularly useful, if more stable parental viruses or DNA viruses shall be adapted in the method of the invention.

In another aspect of the invention a cDNA clone of the oncolytic RNA virus modified by the method of the invention is provided. This modified cDNA clone comprises at least one but preferably most or all mutations identified in the method of the invention. One striking advantage of the invention is the provision of such modified cDNA clone comprising at least one, most or all mutations identifiable in an adaptation process according to the method of the invention.

Such modified cDNA clone is easily propagated and very stable in storage. If needed it leads quickly by standard transfection techniques in suitable cells to the production and generation of the highly adapted RNA virus population, i.e. the personalized RNA virus. The highly adapted RNA virus, i.e. the personalized RNA virus can then be isolated from the supernatant or culture medium of the transfected cells and can either be directly used as a medicament for the treatment of a tumor in a patient or can be purified by standard purification methods and then prepared as a pharmaceutical composition comprising beside the purified virus also pharmaceutical acceptable carriers or diluents.

In another aspect the invention provides thus the nucleic acid sequence of the modified cDNA clone obtained by the method of the present invention, which does incorporate the at least one, most or all mutations identified in the adaption process. This nucleic acid sequence can be directly used in a treatment or can be translated into e.g. mRNAfor use in a treatment.

Yet another aspect of the invention is the provision of the oncolytic virus harvested from cells transfected with the modified cDNA clone generated in the method of the invention. Since the oncolytic virus is adapted to tumor cells of a patient, it is thus personalized. Such a personalized virus can be used for the treatment of the cancer of a particular patient.

The personalized RNA virus according to the invention can be used for the treatment of the cancer of a particular patient but also can be used forthe treatment of other patients as long as the tumor to be treated in this other patient shows similar characteristics, such similar a cellular or tumor markers and/or similar cell determinants.

In yet another aspect of the invention the modified cDNA clone of the oncolytic RNA virus of the invention or the oncolytic RNA virus made from the modified cDNA clone of the invention is for use in the treatment of cancer. Preferably, the cancer to be treated is colorectal cancer.

Advantages and improvements over the prior art

The combination of natural selection and early production of recombinant viruses is saving time in order to produce an adapted virus, which benefits the patient in personalized therapy. According to the invention an adapted virus is available more quickly than an adapted virus obtained by the known method which requires much more passages and consequently more time that cancer patients often do not have.

The most interesting advantage of the invention is the step of isolating the mutations at a very early point in time in the adaptation process, i.e. , at a point in time when the adapted virus population is still a mixed population or a so-called quasispecies of genetically different viruses and variants. The variations of mutations and the completeness of potentially interesting mutations is broadest and not yet reduced by growth advantages of individual variants.

This leads to the most advantageous cDNA clone, comprising most or all identifiable mutations for the virus adaptation. Such cDNA clone is stable and can very well and fast be propagated and/or retransformed easily into fresh and unlimited batches of adapted oncolytic RNA virus for direct treatment of one or several tumor patients.

Brief description of the drawings

Figure 1: Principle of directed evolution. A parental oncolytic virus is passaged on target tumor cells multiple times under defined selection pressure. During adaptation, mutations occur in the viral genome. Improved replication and toxicity can be demonstrated by subsequent analysis of viral fitness. Graphic edited after Sanjuan and Grdzelishvili 2015.

Figure 2: Chronological progression of the VSV-G (Vesicular stomatitis virus) mutation E238K during adaptation is presented as a comparative example illustrating the need of elongated passaging. VSV-p53wt was serially passaged on SUIT-2 cells with a MOI of 0.1. The viral genome was sequenced after passages 10, 20, 26, 30 and 33. A double signal is visible from passage 10 to passage 30 (boxes). Graphic edited after Seegers et al. 2020. The black peak shows the nucleotide guanidine, which occurs at this point in the genome of the parent virus. Grey shows adenine, which only becomes visible after 10 passages. It shows the adapted virus. From passage 10-30 guanine and adenine appear in parallel. This means that the virus population consists of two species. Only in passage 33 does adenine appear. The adapted virus has prevailed, the parent virus has disappeared.

Figure 3 Scheme of the adaptation method. Conventional methods and those used by us are presented in a comparative manner. The new approach reduces the time required for adaptation.

Figure 4: Mutagenesis primers PD_749-M_F (SEQ ID No.: 7) and PD_749-M_R (SEQ ID No.: 8) with their binding sites in the CVB3 genome (extract from SnapGene). Shown are the mutagenesis primers (framed) for mutagenesis from guanine to adenine at nucleotide 749 (highlighted) in the VP4 protein of the CVB3 genome. The binding sites of the primers overlap by 15 bp.

Figure 5: Schematic representation of the In-Fusion assembly reaction (image created from Takara Bio USA). Two PCR fragments with 15 bp overlapping ends (light grey) (A) are degraded at the 3' end by the In-Fusion enzyme (IF) (B, C). This is followed by annealing of the complementary ends (D).

Figure 6: CVB3 strains have differential oncolytic potency against colorectal carcinoma cells in vitro. The cell killing assay was performed with the CVB3 strains Nancy, H3, PD-0 and 31-1-93. HeLa cells as well as DLD1, Colo680h, Colo205, Colo320 and LS174T colon carcinoma cells were infected with either a MOI of 1, 10 or 100. Cell viability was determined 24 h, 48 h and 72 h after infection by crystal violet staining

Figure 7: Analysis of replication and oncolytic activity of PD-H adapted to Colo320 cells.

A: Virus titers of PD-H during serial passaging on Colo320 cells. For this purpose, the cells were infected with a MOI of 0.1 and incubated for 72 h, the titers were determined after each passage using a plaque assay. Up to passage 7, an increase in replication could be determined. B: Cell viability of Colo320 cells 72 h after infection with adapted CVB3 PD-H. A further passaging up to passage 10 did not result in any further increase in viral replication. The cell viability was analyzed using the XTT assay and normalized to the values of uninfected cells. XTT assays were performed for passages 0, 5 and 10. Shown are the mean values with SEM from three independent experiments (n=3).

Figure 8: Schematic representation of the PD-H genome. The schematic representation shows the CVB3 genome with the 11 viral proteins as well as 5' and 3' UTR. The last or first nucleotide of the adjacent viral protein is shown above the representation. The arrows mark the five nucleotides within the genome at which mutations have occurred as a result of adaptation to Colo320 cells. The mutations that could already be identified in passage s are nt 2529 and nt 3046.

Figure 9: Cut-outs from the chromatograms at the mutated sites of the PD-H genome. Shown are cut-outs of the chromatograms of the sequencing of PD-0, PD-5 and PD-10 at the mutated sites (only sites where amino acid exchange occurs) within the CVB3 genome. A double signal was identified in passage 5 for the mutations at nucleotides 2529 and 3046, which was also visible in the sequencing of P10. The mutations at nucleotides 749, 3561 and 3821 were first visible in the sequencing of passage 10. The mutations at nucleotides 749 and 3561 showed a double signal, whereas the sequencing at nucleotide 3821 only detected a signal for guanine instead of adenine. The mutated nucleotide is highlighted with a black box. nt = nucleotide. AA = amino acid.

Figure 10: Growth curves of PD-H and adapted or engineered PD-H variants on Colo320 cells. Colo320 cells were infected with an MOI 0.1. Viral titer were determined by plaque assay after 0, 4, 8, 24, 48 and 72 hours. The PD variants produced show different growth curves. Shown is the mean value with SEM of three independent experiments (n=3).

Figure 11: Cell viability of Colo320 cells 24 h (top) and 48 h (bottom) after infection with CVB3 PD-H and adapted or engineered PD-H variants. The Colo320 cells were infected with an MOI of 0.1 and MOI of 1. 24 h and 48 h p. i., the cell viability was analyzed using the XTT assay and normalized to the values of uninfected cells (MOCK). Significance was compared to cell viability of PD-10 infected Colo320 cells. Shown is the mean value with SEM of four independent measurements (n=4). p. i. = post infection.

Figure 12: Cell viability of different colorectal tumor cell lines after infection with PD-SK and the parental virus PD-H. The tumor cells were infected with a MOI of 0.1 and MOI of 1. 24 h and 48 h p. i. the cell viability was analyzed using the XTT assay and normalized to the values of uninfected cells (MOCK). Shown is the mean value with SEM of three independent measurements (n=3).

Figure 13: Tumor volume after treatment with PD-H-375TS and PD-SK-375TS. Two tumors were established by injecting each 5xl0⁶ Colo320 cells into the right and left flank of nude mice. One of the tumors was injected with 3><10⁶ PFU of the viruses when tumor volume was about 50 mm².Tumor volume was measured every 2 days.

Figure 14: Virus titers and viral genome copies in the tumors and organs 13 days after virus injection.

Figure 15: Dissected Colo320 tumors 13 days after virus injection. * denotes the injected tumor.

Examples

Example 1: Adaptation of CVB3 strain PD-H to the Colo320 cell line (for reproducibility)

To overcome resistance to CVB3 infection, the CVB3 strain PD-H was previously adapted to the colorectal cell line Colo320 (e.g. COLO-320, DSMZ no.: ACC 144) by directed evolution. For this adaptation of CVB3 strain PD-H, the virus was serially passaged over 10 passages by repeated infection with a MOI of 0.1 on Colo320 cells. The replication ability of the adapted viruses was determined after each passage by determining the titer achieved using a plaque assay (FIG. 7A). To assess the oncolytic activity of the virus, a cell viability assay was performed, and cell viability was compared with the parental virus PD-H at passage 0 (P0) after infection, with adapted PD-H after passage 5 (P5) or passage 10 (PIO) (Figure 7B).

Through the serial passaging of PD-H, a significant increase in the titer by a little more than one log level and thus a significant increase in replication could be achieved after just 4 passages. After 7 passages, no further increase in replication could be observed. No significant improvement in oncolytic activity could be measured for adapted PD-H at P5. Only a further adaptation over 5 passages showed a significantly increased oncolysis of adapted PD-H. Cell viability after infection of adapted PD-H at PIO with a MOI of 1 decreased by about a third compared to parental PD-H.

In summary, serial passaging of PD-H on Colo320 cells led to a significant increase in replication after just 5 passages and a significant increase in oncolytic activity after 10 passages, so that an adaptation of PD-H through serial passaging on Colo320 cells can be assumed.

Example 2: Study of adaptation mechanisms

In order to investigate the adaptation mechanisms, the viral genome of the adapted viruses of passages 5 and 10 was sequenced in order to determine mutations acquired through the adaptation. 5 mutations were identified: one in the VP4 protein, two in the VP-1 protein, one in the 2A and one in the 2B protein (compare Figures 8 and 9).

The double signals in the Sanger sequencing show an inhomogeneous virus population consisting of different subpopulations. Based on the literature, it can be assumed that a homogeneous virus population would develop through longer serial passaging (Seegers et al. 2020), but this method takes a lot of time. In addition, there is no guarantee that new virus subpopulations will not emerge, so that a clearly defined virus population will never be reached.

Example 3: Generation of a homogeneous virus population

In order to decrease the number of passages and to generate a homogeneous virus population more quickly, the mutations detected during the adaptation were inserted in their entirety into the cDNA clone of PD-H, a plasmid containing the genome of the PD-H virus in the form of a cDNA (SEQ ID NO: 9), using In-Fusion technology. In addition, six modified cDNA clones with different combinations of mutations were generated, which served to clarify the meaning of the individual mutations (cf. Table 1).

Table 1: Overview of the seven generated cDNA clones. The amino acid that deviates from the amino acid sequence of the original PD-H (founder or parent virus) are printed in thick and marked with an *. nt - nucleotide, aa - amino acid.

These seven virus mutants were generated by transfecting the newly generated plasmids (modified cDNA clones) into HEK293T cells. Then their replication behaviour and oncolytic activity on Colo320 were investigated. A similar replication behaviour could be observed for two virus variants (PD-K2B and PD-K2AB) as for the adapted virus PD-10. Surprisingly, the clones PD-SK and PD-K1-2AB were even able to surpass the replication behaviour of the adapted virus PD-10 (cf. FIG. 10). In terms of cytotoxicity, PD-SK had as well the strongest effect of all the viruses compared (see FIG. 11).

In summary, it is shown that the PD-SK virus, which contains all 5 mutations, was even superior to all other generated variants and to the adapted virus from passage 10 (PIO) with regard to its replication and cytotoxicity.

Example 4:

Study of the oncolytic activity of the adapted viruses on further colorectal tumor cells

It was further investigated whether the improved cytotoxicity and replication of SK-PD in colorectal Colo320 tumor cells has also occured in other colorectal tumor cells. The results are shown in Figure 12. PD-SK showed poorer cytotoxicity in all tumor cell lines examined. This underlines that the adaptation of the virus was cell-specific and therefore only affected the cytotoxic activity in Colo320 tumor cells.

Example 5: In vivo experiments In order to test the effectiveness of the generated virus PD-SK /7? vivo, an animal experiment was carried out. To increase the safety of the virus, the target sequence of the miR-375 specifically expressed in pancreas was inserted into the viral genome. Binding of miR-375 to the specific binding sites results in destroying mRNA bearing the target sequence and thus prevent the expression of the corresponding gene. Although PD-H is safe in vivo after intratumoral injection and no virus-induced side effects occur, in order to avoid replication of PD-SK in pancreas, miR-375 target sequences were inserted into the viral genome. The inventor was previously also involved in showing that this can completely block otherwise fatal CVB3 infections (Hazini et al. 2021).

Additionally, the inventors were able to successfully show that PD-SK-375TS inhibits tumor growth in Colo320 tumors, while injection of the control virus PD-H-375TS had no effect on tumor growth (cf. Figures 13 and 15). No virus was found in the other organs of the mice, so intratumoral injection of PD-SK-375TS can be considered quite safe (see Figure 14).

References

Bauzon, M.; Hermiston, T. W. (2012): Oncolytic viruses: the power of directed evolution. In: Adv Virol. 2012 (586389), S. 586389.

Burrell, R. A.; McGranahan, N.; Bartek, J.; Swanton, C. (2013): The causes and consequences of genetic heterogeneity in cancer evolution. In: Nature. 501 (7467), S. 338-345.

Gao, Y.; Whitaker-Dowling, P.; Watkins, S. C.; Griffin, J. A.; Bergman, I. (2006): Rapid adaptation of a recombinant vesicular stomatitis virus to a targeted cell line. In: J Virol. 80 (17), S. 8603-8612.

Harrington, K.; Freeman, D. J.; Kelly, B.; Harper, J.; Soria, J. C. (2019): Optimizing oncolytic virotherapy in cancer treatment. In: Nat Rev Drug Discov. 18 (9), S. 689-706.

Hazini, A.; Pryshliak, M.; Bruckner, V.; Klingel, K.; Sauter, M.; Pinkert, S. et al. (2018): Heparan sulfate binding coxsackievirus B3 strain PD: a novel avirulent oncolytic agent against human colorectal carcinoma. In: Hum Gene Then 29 (11), S. 1301-1314.

Hazini, Ahmet; Dieringer, Babette; Pryshliak, Markian; Knoch, Klaus-Peter; Heimann, Lisanne; Tolksdorf, Beatrice et al. (2021): miR-375- and miR-l-Regulated Coxsackievirus B3 Has No Pancreas and Heart Toxicity But Strong Antitumor Efficiency in Colorectal Carcinomas. In: Human Gene Therapy 32 (3-4), S. 216-230. DOI: 10.1089/hum.2020.228.

Kuhn, I.; Harden, P.; Bauzon, M.; Chartier, C.; Nye, J.; Thorne, S. et al. (2008): Directed evolution generates a novel oncolytic virus for the treatment of colon cancer. In: PLoS One. 3 (9), e2409.

Pryshliak, M.; Hazini, A.; Knoch, K.; Dieringer, B.; Tolksdorf, B.; Solimena, M. et al. (2019): MiR-375- mediated suppression of engineered coxsackievirus B3 in pancreatic cells. In: FEBS Lett. 594 (4), S. 763-775.

Sanjuan, R.; Grdzelishvili, V. Z. (2015): Evolution of oncolytic viruses. In: CurrOpin Virol. 13, S. 1-5. Seegers, Sara L.; Frasier, Connor; Greene, Sarah; Nesmelova, Irina V.; Grdzelishvili, Valery Z. (2020): Experimental Evolution Generates Novel Oncolytic Vesicular Stomatitis Viruses with Improved Replication in Virus-Resistant Pancreatic Cancer Cells. In: J Virol. 94 (3). DOI: 10.1128/ J VI .01643-19.

Svyatchenko, Victor A.; Ternovoy, Vladimir A.; Kiselev, Nikolai N.; Demina, Anna V.; Loktev, Valery B.; Netesov, Sergey V.; Chumakov, Peter M. (2017): Bioselection of coxsackievirus B6 strain variants with altered tropism to human cancer cell lines. In: Arch Virol. 162 (11), S. 3355-3362. DOI: 10.1007/S00705-017-3492-0.

Takara Bio USA: In-Fusion Cloning mechanism. Aufgerufen am 13.01.2021. Online verfugbar unter https://www.takarabio.com/learning-centers/cloning.

Zainutdinov, Sergei S.; Kochneva, Galina V.; Netesov, Sergei V.; Chumakov, Peter M.; Matveeva, Olga 0. (2019): Directed evolution as a tool for the selection of oncolytic RNA viruses with desired phenotypes. In: Oncolytic Virother. 8, S. 9-26. DOI: 10.2147/OV.S176523

Baogong Zhu; Guifang Cai; Emily 0 Hall; Gordon J Freeman (2007): In-fusion assembly: seamless engineering of multidomain fusion proteins, modular vectors, and mutations. In: Biotechniques 43(3):354-9. DOI: 10.2144/000112536.

Hazini A.; Dieringer B.; Klingel K.; Pryshliak M.; Geisler A.; Kobelt D.; Daberkow O.; Kurreck J.; Linthout S.V.; Fechner H. (2021) Application Route and Immune Status of the Host Determine Safety and Oncolytic Activity of Oncolytic Coxsackievirus B3 Variant PD-H. Viruses. 13(10):1918. DOI: 10.3390/V13101918.

Claims

Patent Claims

1. A method for generating personalized oncolytic RNA viruses, comprising the steps: a. infecting in vitro patient-derived tumor cells or a tumor cell line closely related to a patient’s tumor with an RNA virus, addressed as parent virus or founder virus; b. performing multiple passages of the RNA virus by separating the replicated RNA viruses from the infected patient tumor cells or the tumor cell line and passaging it to a fresh culture of said cells until at least one adapted RNA virus variant is achieved and/or detectable; c. identifying and/or sequencing at least one, most or all mutations of the at least one adapted RNAvirus variant detected at the adaptation stage of step b), said adaptation stage comprising a mixed population of genetically different RNAvirus mutants; d. generating by molecular cloning techniques a modified cDNA clone, which incorporates at least one, most or all mutations identified in step c); e. transfecting the modified cDNA clone of step d.) into a suitable cell line for the production of personalized RNAvirus and f. isolating and optionally purifying the personalized RNAvirus.

2. The method according to claim 1, wherein the oncolytic RNAvirus is a picornavirus.

3. The method according to claims 1 or 2, wherein the picornavirus is a Coxsackievirus strain B, preferably B3 strain.

4. The method according to any of claims 1 - 3, wherein multiplicity of infection (MOI) in step a) is less than 1.

5. The method according to any of claims 1- 4, wherein the number of passages in step b) is between 2 and 15.

6. The method according to any of claims 1- 5, wherein cDNA clone comprising all mutations identified in step c) is generated using a recombinant technique, optionally In-Fusion technique.

7. The method according to any of claims 1- 6, further comprising a step e) inserting into the cDNA clone of the oncolytic parent virus one or more expression controlling RNA motives selected from the group comprising: shRNAs, microRNAs, artificial microRNAs, tissue specific miRNAs target sequences, and optionally inserting one or more transgenes.

8. The method according to claims 7, wherein the transgene is an interleukin, preferably neoleukin-2/15.

9. The method according to any of claims 1- 8, wherein step a) further comprises the addition of chemical mutagens and/or the addition of genetically modified polymerases.

10. A modified cDNA clone of the oncolytic RNA virus produced in step d.) of the method accordingto any one of claims 1 - 9.

11. Nucleic acid sequence of the modified cDNA clone accordingto claim 10.

12. Oncolytic RNA virus harvested from cells transfected with the modified cDNA clone of the oncolytic RNA virus accordingto claim 10.

13. The modified cDNA clone of the oncolytic RNA virus accordingto claim 10, nucleic acid sequence accordingto claim 11 or the oncolytic RNA virus accordingto claim 12 for use in the treatment of cancer.

14. The modified cDNA clone of the oncolytic RNA virus, nucleic acid sequence or the oncolytic RNA virus for use accordingto claim 13, wherein the cancer is colorectal cancer.

15. Pharmaceutical composition comprising the modified cDNA clone of the oncolytic RNA virus according to claim 10, nucleic acid sequence according to claim 11 and/or the oncolytic RNA virus accordingto claim 12, and a pharmaceutically acceptable carrier or diluent.