Patent Application
20110250188
The present invention relates to recombinant viruses derived from the vesicular stomatitis (VSV) and viral vectors which comprise a gene coding for the glycoprotein GP of the lymphocyte choriomeningitis virus (LCMV), packaging cells which produce the GP-LCMV-pseudotyped VSV virions, and the use of said virions and packaging cells for the preparation of a pharmaceutical composition for the therapy of solid tumors.
Malignant gliomas, the largest group of primary intracranial brain tumors, represent a therapeutic problem which is not solved yet. Although the knowledge of the biology of these tumors has grown due to intense basic research, clinical progress and prognosis are still very poor.
Malignant gliomas are tumors of neuroepithelial origin and are cytologically divided in ependimomas, oligodendrogliomas, oligoastrocytomas, astrocytomas and glioblastomas. With a proportion of more than 60%, diffuse-infiltrating astrocytomas (WHO Grade II-IV) represent the largest group of intracranial tumors. The WHO classification revised in 2001 is largely established as grading scheme for astrocytomas. According to the WHO, brain tumors are allocated by means of histological criteria to 4 malignancy grades (Kleihus and Cavenee, 2000). The prognosis for diffuse-infiltrating astrocytomas is generally poor. The prognosis depends, on the one hand, on the malignancy grade and, on the other, on the localization of the tumor and the therapy procedure. The average survival rate for patients with an astrocytoma WHO grade II is more than 5 years, with an astrocytoma WHO grade III, 2 to 5 years, and with an astrocytoma-glioblastoma of the WHO grade IV (=glioblastoma), less than 1 year.
The molecular pathogenesis of tumors is a complex process and is based on mutations of different genes which are responsible for the control of the cell cycle. Mutations in the tumor suppressor gene p53 are the most frequently found alterations in human tumors and are also responsible for the development of low-grade astrocytomas as well as for the progression to the secondary glioblastoma. However, primary developed glioblastomas very rarely have p53 mutations. A further gene, which indicates a malignant tendency of diffuse astrocytomas, is suspected to be on the long arm of chromosome 19. Further genes which are frequently altered in case of glioblastomas are the oncogenes MDM2 and MDM4 and also the tumor suppressor gene p14ARF, which are involved in the p53-dependent control of the cell cycle (Dai and Holland, 2001). An amplification of the EGF receptor gene is observed in 30-40% of the primary glioblastomas and is therefore the most frequently amplified oncogene in this tumor group (Holland, 2001). The majority of malignant gliomas responds poorly to a chemo- or radiotherapy. It is assumed that the reason for this are mutations of cell-cycle-associated genes which are also involved in the regulation of the apoptosis.
More effective therapy methods for malignant gliomas are urgently needed because from the existing therapy methods such as chemo- or radiotherapy, no significant improvements for the prognosis of the disease are to be expected. In contrast, the gene therapy of the glioblastoma offers promising possibilities which need to be exploited. A plurality of different, very effective genes was developed for this purpose. For most of them, data from experiments on animals are available (Shir and Levitzki, 2001). These therapeutic genes can be allocated to four different active principles:
(i) The gene product of so-called suicide genes converts precursors in cytotoxic molecules. An example is the thymidine kinase (TK) of the herpes simplex virus (HSV) in connection with a dosage of ganciclovir. A particular advantage is that the toxic ganciclovir triphosphate diffuses into adjacent cells, whereby a bystander effect takes place. In the last years, the activity of this enzyme was further increased. HSV TK is currently a very efficient possibility to eliminate tumor cells as well as implanted vector producing cells.
(ii) The expression of immunostimulatory cytokines such as the IL-4 can stimulate the natural defense against tumor cells.
(iii) The secretion of anti-angiogenetic proteins such as the endostatin results in a lack of blood vessels and therefore in a lack of nutrient supply in the metabolically very active tumor tissue. The tumor is virtually “starving”.
(iv) Finally, a series of genes was described which engage into the signal transduction or the cell cycle of the tumor cell in order to inhibit the uncontrolled growth of these cells. However, the possibilities of use of these genes in the clinic are limited because these genes act only in the gene-modified cell itself and do not have the bystander effect as the first-mentioned active principles. This means that in order to achieve a therapeutic effect, virtually all malignant cells have to be genetically modified which is hopeless even with ideal vector systems.
The most important prerequisite for a successful gene therapy of the glioblastoma is provided by the multitude of existing, very effective principles of action. However, a problem which is not solved yet is the inefficient gene transfer and a poor expression of the therapeutic gene in the target cells. This is also the reason why, despite the multitude of efficient therapeutic genes, the gene therapy of glioblastoma failed in the clinic.
The advantage of the viral gene transfer over physicochemical transfection methods is the higher gene transfer rate and the long term expression of the genes because the viruses have developed particularly efficient mechanisms to introduce their genome into cells and to express it. In particular replication-competent viruses such as, amongst others, herpes simplex virus (HSV), adenoviruses (Ad), Newcastle Disease Virus (NDV) and the vesicular stomatitis virus (VSV) are currently used as oncolytic viruses (OV). For an optimal virotherapy of glioblastoma, the OV should have the following features:
(i) They should have a tumor-specific tropism, whereby virus replication and cell lysis remains limited to the tumor tissue. This property can be enhanced by modification of the viral envelope or by using tumor-specific promoters. Since the assumption is that only a small portion of the glioma cells divide during the treatment, viruses which infect resting cells as well as proliferating cells are of advantage.
(ii) With respect to safety-relevant aspects, viruses with high genetic stability and a low toxicity outside of the tumor tissue are particularly suitable for clinical use. This allows a high virus titer and a purification of the vectors under GMP conditions. Ideally, the OV should be apathogenic for humans and should have a low infection rate among the population. An already existing immunity would result in a premature neutralization of the virus and thus would not allow an efficient therapy.
Prominent examples for oncolytic viruses in the therapy of glioblastomas are the attenuated HSV variants G207 and 1716 and the adenovirus ONYX-015. A serious objection against the use of oncolytic HSV for the treatment of CNS tumors is its high level replication in normal brain cells which can result in a life threatening encephalitis. Moreover, besides potential persistence, there is the possibility of reactivation of latent HSV. The HSV variant 1716, which was generated by deleting a plurality of genes, selectively replicates in rapidly proliferating cells of the CNS but not in postmitotic neurons. Thereby, the neurotoxicity of HSV was significantly reduced. The treatment of experimental gliomas in the rat and in the mouse with HSV 1716 resulted in selective destruction of tumor cells while surrounding brain tissue remained undamaged. ONYX-015 is a further oncolytic virus which was developed for the glioblastoma therapy. Through a deletion in the E1B gene, this adenovirus is intended to selectively lyse cells with defective p53.
Oncolytic HSV as well as ONYX-015 were already clinically tested for the treatment of gliomas. Both attenuated oncolytic viruses showed a sufficient safety in clinical phase I/II studies. Independently of whether the viruses were injected intratumorally or into the resection cavity, the treatment was well tolerated and no serious side effects were observed. However, the cytolytic effects were only of transient nature, always followed by a recurrence (Cutter et al., 2206). Since normally the proliferation rate of gliomas exceeds the amplification rate and thus the spreading wave of viruses, the destruction of gliomas only by oncolysis is questionable. In fact, for an efficient treatment, the combination of a plurality of active principles is required. By using suicide genes or immunomodulatory genes in OV, a synergistic effect was demonstrated in preclinical studies (Tyminski et al., 2005; Fukuhara et al., 2005).
Besides the aforementioned DNA viruses, oncolytic RNA viruses are also under development. VSV is an enveloped negative-strand virus, the host spectrum of which comprises rodents and livestock. Infections of humans are rare and are mostly asymptomatic. Due to the very low seroprevalence among the population, an impairment of the therapy efficiency by VSV-neutralizing antibodies is not to be expected. The infection and the cytoplasmatic replication of VSV take place independently of the cell cycle so that actively dividing cells and resting cells are equally infected. The efficient and preferential lysis of neoplastic cells by VSV is related to the mostly defective interferon signaling pathway and the accompanying viral replication in these cells (Wollmann et all., 2007). It was also demonstrated that independent of the cellular immune response, tumor cells with defects in the genes Myc, Ras or p53 also support the reproduction of VSV (Barber, 2004). The tumor specificity of VSV was further optimized in the last years by the preparation of recombinant viruses. The main focus here is on variants with mutation in the M protein (ΔM51). This variant is not able to prevent the interferon response in healthy cells, whereby virus replication in such cells is suppressed. In tumor cells with defective IFN response, the virus can replicate and thus be selectively oncolytically active. Even after systemic application, VSVΔM51 showed a secure and efficient oncolysis of human gliomas in the mouse model (Lun et al., 2006).
The application of viral vectors directly into the brain requires a high selectivity for tumor cells and is only possible in relatively small volumes. Thus, an efficient gene transfer can only be achieved with highly concentrated vector preparations (>108/ml) which have a strongly developed tropism for glioma cells. In case of gamma-retroviral and lentiviral vectors, vector tropism and vector stability can be influenced by integrating a non-retroviral envelope protein. In many cases, the retroviral envelope protein is replaced with the more stable G-protein of VSV. A problem of these so-called pseudotyped vectors is that VSV-G is cell-toxic, namely in the producer cells as well as for the surrounding healthy tissue, which previously stood in the way of a widespread use of such VSV-G pseudotypes in the clinic.
The inventors have previously developed a retroviral vector type which allows for an efficient gene transfer into glial cells of the CNS. In this new vector type, the glycoprotein GP of the lymphocyte choriomeningitis virus (LCMV) serves as viral envelope protein (Miletic et al., 1999; Beyer et al., 2002; EP 1 006 196). In comparative in vitro and in vivo tropism studies it was demonstrated that LCMV-GP pseudotypes transduce preferentially glioma cells (Miletic et al., 2004; Miletic et al., 2007). In contrast to that, VSV-G pseudotypes preferentially transduced neurons; whereas gene transfer in glioma cells was less efficient than with LCMV-GP pseudotypes. Also individual infiltrating tumor cells were efficiently transduced by LCMV-GP pseudotypes. In the rat glioma model, 90% of the rats were healed by intratumoral injection of LCMV-GP-pseudotyped lentiviral vectors (Miletic et al.; 2007b). The used vectors coded for the thymidine kinase (TK) of the herpes simplex virus, with the effect that in the transduced cells as well as in the surrounding cells, ganciclovir was converted into a cell-toxic triphosphate compound.
The selective gene transfer into gliomas, but also an efficient vector distribution within the entire tumor are decisive for the success of the therapy. Therefore, the inventors have developed tumor-infiltrating packaging cells which are supposed to release pseudotyped vectors within the entire tumor (WO 2006/008074). A promising cell type with migratory capabilities is the multipotent, adult progenitor cell which can be isolated from the bone marrow (Jiang et al., 2002). In transplantation experiments, the migration behavior of these cells was examined using the rat glioma model. It was demonstrated that the progenitor cells efficiently penetrated the tumor mass, but did not infiltrate the surrounding healthy brain tissue (Fischer et al., 2007). Conventional cell lines such as 3T3 mouse fibroblasts, Rat-1 rat fibroblasts or human 293T did not show any tumor infiltration. Rather, these cells were locally limited to the injection site or the vicinity and did not show glioma-specific migration.
In a further study, the inventors examined the therapeutic efficacy of TK-expressing progenitor cells in the rat glioma model. The result of this study was that in 70% of the rats the destruction of the tumor was solely due to the bystander effect between progenitor cell and tumor cell (Miletic et al., 2007). Moreover, the intratumoral localization of the progenitor cells by means of imaging methods was confirmed. Histological slices of the brain of treated, symptom-free rats had a cavity with a pronounced scar tissue at the location of the tumor; this indicates that the tumor in the animals was successfully eradicated by the gene therapy. Thanks to the strong expansion potential of these progenitor cells, a genetic modification with subsequent selection of individual packaging cell clones is possible. A progenitor-based packaging cell for gamma-retroviral LCMV-GP pseudotypes was developed. These packaging cells continuously produced retroviral vectors with a titer of 1-7×10E3 TU/ml. The titers remained stable over several weeks and after repeated freezing and thawing the cells (Fischer et al., 2007).
The inefficient gene transfer in vivo and not the lack of therapeutically effective genes currently hinders successful gene therapy of the glioblastoma. The previously known vectors for gene therapy and oncolytic virotherapy of gliomas are not optimal for various reasons. The efficiency, specificity and safety of previous gene transfer methods are to be increased to such an extent that a therapeutically effective gene transfer in patients is possible.
It is therefore the object of the present invention to develop a highly potent oncolytic viral gene transfer system for therapeutic genes for the therapy of highly malignant brain tumors such as gliomas and other solid tumors.
The subject matter of the invention is therefore a vesicular stomatitis virus (VSV) pseudotype vector which comprises a gene coding for a glycoprotein GP of the lymphocyte choriomeningitis virus (LCMV) and no functional gene coding for envelope protein G of the VSV. LCMV-GP can be the glycoprotein GP-1 or GP-2 of the LMCV.
According to a further embodiment, in the VSV-LCMV-GP pseudotype vector, the envelope protein G of the VSV is replaced by GP of the LCMV (LCMV-GP).
According to a further embodiment, the vector lacks at least one gene selected from the group of n, l, p and m genes coding for proteins N, L, P and M of the VSV.
According to a further embodiment, the M protein of the VSV comprises mutations which reduce the cytopathogenicity of the VSV. Examples of mutations resulting in reduced cytopathogenicity of VSV are amino acid exchanges in the 37PSAP40 region of the M protein as well as mutations M33A, M51A, V221F, S226R or their combination.
According to a further embodiment, the VSV-LCMV-GP pseudotype vector comprises at least one therapeutically applicable transgene. The transgene can be a suicide gene or an immunostimulatory gene. Examples of suicide genes are genes which code for thymidine kinase of the herpes simplex virus (HSV-TK), cytosine deaminase, FKBP-FAS or FKBP-caspase 9. Examples of immunostimulatory genes are genes which code for cytokines IL-2, IL-4, IL-12, neutralizing anti-TGFbeta, or Flt3L.
According to a further embodiment, the VSV-LCMV-GP pseudotype vector comprises a marker gene. The marker gene can be LacZ, an antibiotic resistance gene or a gene coding for fluorescence protein (GFP, RFP, GGP etc.).
The subject matter of the invention is further a VSV-LCMV-GP pseudotype vector system which comprises at least two complementary replicating (cr) VSV vectors, wherein one vector of the vector system comprises a gene gp coding for LCMV-GP, wherein the vector system further comprises genes n, l, p and m coding for proteins N, L, P and M of the VSV and no functional gene coding for envelope protein G of the VSV, wherein each vector of the vector system lacks one of the genes (“complementary gene”) gp, n, l, p and m, and wherein the lacking gene is present on any other vector of the vector system. The genes m and gp are preferred complementary genes.
According to a further embodiment, the M protein of the VSV-LCMV-GP pseudotype vector system comprises mutations which reduce the cytopathogenicity of the VSV, as illustrated above.
According to a further embodiment, the VSV-LCMV-GP pseudotype vector system comprises at least one therapeutically applicable transgene and/or a marker gene, as illustrated above. The transgene/marker gene can be located on any vector of the vector system.
The subject matter of the invention is further a VSV virion pseudotyped with LCMV-GP which comprises a GP protein of the LCMV as envelope protein.
The subject matter of the invention is further a virus producing cell which produces a VSV virion pseudotyped with LCMV-GP.
According to a further embodiment, the virus producing cell is an adult stem cell. The adult stem cell can be a multipotent adult progenitor cell (MAPC), a neuronal stem cell (NSC), a mesenchymal stem cell (MSC) or a BM-TIC cell (bone marrow derived tumor infiltrating cell) derived from MSC.
According to a further embodiment, the virus producing cell comprises one or more expression cassettes for the expression of genes selected from the group consisting of the genes n, l, p and m coding for proteins N, L, P and M of the VSV, respectively, and a gene gp coding for LCMV-GP glycoprotein.
According to a further embodiment, the virus producing cell further comprises a gene transfer vector for packaging into a VSV virion pseudotyped with LCMV-GP.
According to a further embodiment, the gene transfer vector comprises a therapeutically applicable transgene and/or a marker gene, as illustrated above.
The subject matter of the invention is further an in vitro method for transferring a transgene into cell, in which method the cell is transduced with a VSV virion pseudotyped with a LCMV-GP, wherein the virion comprises a transgene.
The subject matter of the invention is further an in vitro method for transferring a transgene into a cell, in which method the cell is contacted with a virus producing cell which produces a VSV virion pseudotyped with LCMV-GP, wherein the virion comprises a transgene.
It is preferred that the cell is a tumor cell, for example, a glioma cell.
The subject matter of the invention is further the use of a VSV-LCMV-GP pseudotype vector or a VSV-LCMV-GP pseudotype vector system of the invention for the preparation of a pharmaceutical composition for the therapy of a solid tumor.
The subject matter of the invention is the use of a VSV virion pseudotyped with LCMV-GP for the preparation of a pharmaceutical composition for the therapy of a solid tumor.
The subject matter of the invention is further the use of a virus producing cell of the invention for the preparation of a pharmaceutical composition for the therapy of a solid tumor. It is preferred that the tumor is a brain tumor, in particular a glioma.
According to a further embodiment, at least two virus producing cells are used for the preparation of a pharmaceutical composition for the therapy of a solid tumor, wherein a first virus producing cell comprises a first vector of the VSV-LCMV-GP pseudotype vector system and a second packaging cell comprises a second vector of the VSV-LCMV-GP pseudotype vector system.
Furthermore, subject matter of the invention is a pharmaceutical composition which comprises a VSV-LMCV-GP pseudotype vector, a VSV-LCMV-GP pseudotype vector system, a VSV virion pseudotyped with LCMV-GP, or a virus producing cell which produces a VSV virion pseudotyped with LCMV-GP. The composition can also comprise suitable auxiliary substances and/or carriers.
The invention relates to recombinant vesicular stomatitis viruses (VSV) and VSV vectors. The VSV genome comprises five genes l, m, n, p and g which code for the proteins L, M, N, P and G and which are essential for the reproduction of the virus. N is a nucleoprotein which packages the VSV genomic RNA; the VSV genome can only be replicated as RNA-protein complex. L and P together form a polymerase complex which replicates the VSV genome and transcribes the VSV mRNA. M is a matrix protein which forms a kind of kit between lipid envelope and nucleocapsid and is important for particle sprouting at the cell membrane. G is the envelope protein which is incorporated in the viral envelope and is essential for the infectivity of the virus.
The subject matter of the invention is a VSV-LCMV-GP pseudotype vector and a VSV virion pseudotyped with LCMV-GP. The vector of the invention comprises a gene gp coding for the GP protein instead of gene g coding for the G protein of the VSV. Accordingly, the virus/virion of the invention comprises a LCMV-GP protein as envelope protein. The GP protein can be GP1 or GP2. The invention includes GP proteins from different LCMV strains. In particular, LCMV-GP variants can be derived from LCMV wild-type or LCMV strains LCMV-WE, LCMV-WE-HPI, LCMV-WE-HPIopt (WO 2006/008074).
Vectors which are based on the virus of the vesicular stomatitis have advantages over retroviral vectors:
Taken together, the VSV vectors have a particularly high therapeutic potential.
Another advantage of the VSV-LCMV-GP pseudotype vector according to the invention is a considerably reduced toxicity of the VSV viruses pseudotyped with LCMV-GP against healthy brain cells, i.e., neurons. The neurotoxicity of the VSV is attributed to the G protein of the VSV (Shinosaki et al., 2005). Actually, VSV-G pseudotypes preferentially infect normal neurons. Since neurotropism is a dose-limiting factor in all applications of oncolytic VSV, the use of the vector according to the invention is of great advantage for all tumors.
In addition, the VSV-LCMV-GP pseudotype vector has an increased specificity for brain tumor cells which goes back to the glioma-specific tropism of the used LCMV-GP glycoprotein and the selective transcription/replication of VSV in tumor cells. In contrast to VSV-G, the LCMV-GP envelope protein has a particular tropism for glioblastoma cells. However, the vector according to the invention can successfully be used in other tumors outside the CNS because there too the neurotoxicity of VSV, in particular in case of systemic application, is a dose-limiting factor.
Vectors pseudotyped with LCMV-GP have three crucial characteristics which non-pseudotyped vectors do not have, namely:
The invention includes replication-deficient as well as replication-competent viruses. The latter have an additional advantage because a high transduction rate can be achieved by using viruses capable of reproduction. In this respect, replicable oncolytic viral vectors are more efficient than replication-incompetent vectors.
To increase the safety during the use of replicable viruses in therapeutic uses, a vector system is provided which ensures that replication, oncolysis and the production of VSV viruses takes place only in cells which are infected by at least two replication-deficient mutually complementing vectors.
The subject matter of the invention is therefore a VSV-LCMV-GP pseudotype vector system which comprises at least two complementary VSV vectors. Such complementary replicating (cr) VSV vectors can spread to a limited extent within the tumor, which increases the efficiency of the gene transfer and the oncolysis. Thus, the vector system according to the invention allows for the preparation of oncolytic VSV-LCMV-GP pseudotype vectors with limited reproduction capability for the gene transfer in gliomas and other tumors.
The principle of the vector system according to the invention is that each vector of the system lacks one of the essential genes m, n, l and p of the VSV or gp of the LCMV which, however, is present on any other vector of the system. The gene gp coding for LCMV-GP as well as possible additional genes such as therapy genes and/or marker genes can be present on any vector of the system.
Different variants of the vector system according to the invention are possible. For example, the vector system can consist of two vectors as illustrated in
Up to now, different concepts for the therapy of the glioblastoma have been pursued: (i) the transfer of suicide genes, (ii) the elimination of the glioblastoma-related immunosuppression using immunostimulatory genes and by means of cytokine-based immunotherapy, (iii) the transfer of factors which counteract the tumor-induced angiogenesis, (iv) the use of cell cycle modulators and (v) the induction of apoptosis. For the purposes of the invention, the first three approaches (i-iii) are to be primarily considered.
In addition to their inherent oncolytic properties, the VSV-LCMV-GP pseudotype vectors and vector systems based thereon can further be improved by introducing suicide genes or/and immunostimulatory genes. Examples of suicide proteins are thymidine kinase of the herpes simplex virus (HSV-TK), cytosine deaminase, FKBP-FAS, FKBP-caspase9. Examples of immunostimulatory proteins are cytokines such as IL-2, IL-4, IL-12 and Flt3L, neutralizing anti-TGFbeta. Insertions of genes with a size of up to 4.5 kb are tolerated by the VSV genome so that it would even be possible to combine two or more therapeutic genes (e.g. HSV-TK+cytokine) in one VSV genome. In addition, the gene products exert their action also on non-infected tumor cells by means of a so-called bystander effect.
Injected viral vectors penetrate from the injection site only a few millimeters into the tumor tissue. For optimizing the vector distribution and for the targeted destruction of the tumor, tumor-infiltrating virus producing cells are used which specifically migrate in the tumor thereby releasing the viruses at the site remote from the injection site.
Virus producing cells in the meaning of the invention include classical packaging cells for the production of virions from non-replicable vectors as well as producer cells for the production of virions from vectors capable of reproduction. Packaging cells usually comprise one or more plasmids for the expression of essential genes which lack in the respective vector to be packaged and/or are necessary for the production of virions.
In previous studies, packaging cells were used for transferring viral vectors; however, this involved mainly fibroblasts which do not migrate within the tumor (Short et al., 1990, Culver et al., 1992). In contrast, adult stem cells, in particular neuronal (NSC) and mesenchymal stem cells (MSC) have a high migratory potential. They remain confined to the tumor tissue, whereby a very efficient but also specific gene transfer into the tumor tissue is achieved. However, these stem cells have limited passage capacity in vitro.
A subpopulation of adult mesenchymal stem cells, so-called BM-TIC (bone marrow derived tumor infiltrating cells) infiltrate, after injection into experimentally induced gliomas, the entire tumor and, in addition, track individual tumor cells remote from the tumor mass (Miletic et al., 2007). BM-TIC are isolated from adult bone marrow, have a high expansion potential and can be used as migrating producer cells for MLV (Fischer et al., 2007) and VSV vectors.
The subject matter of the invention is thus virus producing cells which produce oncolytic VSV-LCMV-GP pseudotype vectors. In particular, these are tumor-infiltrating producer cells which release the said vectors during their migration within the tumor. Preferred cells are adult stem cells, in particular neuronal (NSC) and mesenchymal stem cells (MSC). Particularly preferred cells are BM-TIC cells derived from MSC.
An obstacle for the preparation of a rhabdoviral virus producing cell is the cytopathogenicity of the proteins M and G. Since the G protein of the VSV is replaced by the glioma-specific and non-cytotoxic glycoprotein of the LCMV in the VSV-LCMV-GP pseudotype vectors of the invention, only the M protein remains a problem. For reducing the toxicity of the M protein in the VSV producing cell line, a non-cytopathogenic variant of the M protein can be used.
Therefore, the virus producing cells of the invention and hence also the VSV-LCMV-GP pseudotype vectors produced by said cells may comprise a gene coding for a mutated M protein. This vector variant is selectively oncolytic for tumor cells, whereas it is not toxic for healthy cells. M variants with amino acid exchange in the 37PSAP40 region of the M protein or with single (M51R) or multiple (V221F and S226R; M33A and M51A) mutations outside of the PSAP region of the M protein are preferred. An M protein with mutations M33A, M51R, V22F and S226R is particularly preferred. In order to ensure an efficient virus production in packaging cells, the M variant can be stably transfected with a viral interferon antagonist.
In addition, subject matter of the invention is an in vitro method for gene transfer, wherein a LCMV-GP-VSV pseudotype vector or LCMV-GP-VSV pseudotype vector system comprising a transgene is introduced into a cell either directly or by means of virus producing cells (packaging cells) according to the invention. If a cr vector system with at least two vectors is used, at least two packaging cells are used, wherein each of the cells produces one of the (replication-incompetent) cr vectors. The production of VSV viruses takes place only in cells which are infected with all vectors of the cr vector system and hence comprise all essential viral genes.
In addition, the invention relates to the use of vectors and virus producing cells according to the invention as drugs in therapeutic methods. In particular, the vectors and virus producing cells according to the invention are used for the therapy of solid tumors. The therapeutic effect is caused by the oncolytic properties of the recombinant vectors and viruses as well as by the use of therapeutic genes.
Solid tumors can be brain tumor, liver tumor, in particular hepatocellular carcinoma, lung tumor, in particular bronchial carcinoma, and intestinal tumor, in particular colon carcinoma. Preferred tumors are brain tumors, for example a glioma, in particular ependymoma, oligodendroglioma, oligoastrocytoma, astrocytoma, glioblastoma, or a medulloblastoma.
The subject matter of the invention is further a pharmaceutical composition which comprises the vector, the virion, the virus producing cell of the invention and optionally additives such as carrier and auxiliary substances.
In order to increase the viral oncolysis and the transfer efficiency of the therapeutic genes, tumor-infiltrating virus producing cells which continuously release vectors are formulated for direct implantation into the tumors.
The invention is illustrated by means of the following examples.
Defective VSV vectors which code GFP but no viral envelope protein (VSVgfp-ncp-ΔG) were transduced into BHK-21 and vero cells which in turn stably express the GP of LCMV. 1×10E5 cells (BHK-21, -GP, vero, -GP) were sown per 24-well plate cavity and four hours later transduced at MOI=5 with VSVgfp-ncp-ΔG vectors. 24 hours later, the culture supernatant was collected and the titer in BHK-21 was determined by means of FACS analysis of the GFP expression.
Since the pseudotypes were intended for the gene transfer in gliomas, the transduction efficiency of the vectors was checked for different glioma cell lines. Two human glioma cell lines (U87, G44) and a rat glioma cell line (9L) were tested. BHK-21 were used again as a control. 1×10E5 cells per 24-well plate cavity were seeded and four hours later transduced at MOI=0.3 (titration on BHK-21) with VSV-LCMV-GP pseudotype vectors. 24 hours later, the percentage of GFP-expressed cells was determined by FACS and the titer was calculated therefrom.
The inventors have already demonstrated that stem cells which were isolated according to the protocol of C. Verfaillie (Jiang et al., 2002), have tumor-infiltrating properties. Because only their migration potential but not their differentiation potential was examined, these cells were called BM-TIC (bone marrow derived tumor infiltrating cells). For testing the packaging capability of BM-TIC for VSV vectors, LCMV-GP-expressing BM-TIC were transduced with the defective VSV-GFP vector. 1×10E5 cells (BM-TIC, -GP, BHK-21) were seeded per 24-well plate cavity and four hours later transduced at MOI=5 with VSVgfp-ncp-ΔG vectors. 24 hours later, the culture supernatant was collected and titrated on BHK-21 cells.
A high transduction rate can be achieved with viruses capable of reproduction. In order to increase the gene transfer rate in gliomas and at the same time to ensure a high degree of safety, a VSV pseudotype vector was established which is replication-competent to a limited extent. Systems with complementary defective viruses have been described for retroviruses (Trajecevski et al., 2004) and different flaviviruses (Riepl & Mandl, 2007). The principle of such systems is to distribute the viral genome between two incomplete replicons. Infectious viruses are only produced if one and the same cell is infected with both vectors and only through this co-infection comprises all the necessary components for packaging the virus genome.
While only the genomic RNA of positive-strand RNA viruses is infectious, negative-strand RNA viruses such as the VSV need in addition to the viral RNA genome at least viral nucleoprotein (N) and viral polymerase (L) in order to generate an infectious virion from cloned viral cDNAs. The system for preparing recombinant VSV is based on the cytoplasmatic expression of viral (+) RNA and the viral proteins N, P and L using the T7-RNA polymerase which is constitutively expressed in BSR T7/5 cells. In order to prepare the desired vectors, genomic constructs together with the expression plasmids for N, P and L and additionally the expression plasmid for VSV-G in the case of Vector b) (
After characterization of the individual vectors with respect to their titer and their transduction efficiency, the spread of the vectors in vitro is examined. For this, BHK-21 cells are co-infected with both vectors and the replication and thus the generation of new vectors by serial passage of the culture supernatant on native BHK-21 is examined. These experiments are repeated with different glioblastoma cell lines (G44, G62, U87) which are cultivated as monolayer as well as spheroids.
Tumor spheroids are three-dimensional, organoid cell groups which reflect the nature and heterogeneity of tumors better than monolayer cultures. The actively dividing tumor cells are located in a loose formation at the edge of the cell aggregate, whereas cells located deeper inside no longer divide and a formation of necrotic areas takes place here (Carlson et al., 1984). Experiments with spheroids provide information about intratumoral transduction efficiency, the spreading kinetics and the oncolytic effect of the cr VSV-LCMV-GP pseudotype vector system. These spheroids thus allow an initial optimization of the system without the need of conducting experiments in animals. A possible toxicity of the cr LCMV-GP-pseudotyped VSV vectors, the therapeutic effectiveness of the oncolysis and the presence of the remaining viruses after tumor elimination are examined in the animal model (see example 6).
The viral RNA polymerase has no recombinase activity and there are also no cellular recombinases in the cytoplasm so that the recombination between two negative-stranded RNA viruses can only take place by a template switching of the RNA polymerase. However, this is an extremely rare event (Finke & Conzelmann, 2005; Spann et al., 2003). The generation of a recombinant, replicable VSV represents a safety risk and is examined. For this, viral RNA is isolated at different times from the culture supernatant of co-infected cells and examined with a transgene-specific (e.g. gfp) probe in Northern blots. If RNA species with “excess length” are detectable, the viruses from the supernatant are plaque-purified and examined by means of standard methods with respect to their infectivity and are subsequently characterized by means of molecular biology.
Migrating VSV-LCMV-GP producer cells are established for cr VSV vectors and for replication-competent VSV-LCMV-GP vectors.
The VSV M protein has two essential tasks within the VSV cycle. On the one hand, it is essential as structural component of the virion for the assembly and the sprouting of the viruses. On the other hand, it contributes significantly to the viral pathogenesis through its “host shut-off” activity directed to the host protein synthesis and through induction of the apoptosis. In this connection, an important function of M is the inhibition of the nucleo-cytoplasmatic transport of IFN mRNA. Thereby, the first defense mechanism of the cell against viral infection is suppressed.
Due to this strong cytotoxic effect, it is not possible to express the M protein in cell lines on a long-term basis. However, a plurality of mutations in the M protein is known which result in the attenuation of the virus. Irie and coworkers demonstrated that by amino acid exchanges in the so-called 37PSAP40 region of the M protein, a VSV mutant with greatly reduced cytopathogenicity was created (Irie et al., 2007). Furthermore, as shown in mouse experiments, M variants with single (M51R) or multiple (V221F and S226R; M33A and M51A) mutations outside of the PSAP region are also greatly attenuated (Desforges et al., 2001; Jayakar and Whitt, 2002). These viruses are not able to suppress the release of IFN-a/β and accordingly induce an antiviral status which protects the animals against an infection. However, in tumors which often have defects in the IFN system, these viruses can reproduce and lyse the tumor. For safety reasons, such attenuated variants are of particular importance for the gene therapy because they replicate preferentially in tumor cells.
Also, the inventors have cloned a non-cytopathogenic variant of the M protein (Mncp) which comprises the mentioned mutations M33A, M51R, V221F and S226R and is no longer able to suppress the synthesis of IFN-β. This variant replicates in a similar efficient manner as the parent virus on IFN-incompetent cells, but is attenuated on IFN-competent cells. To ensure efficient virus production in BM-TIC, the cells are stably transfected with a viral interferon antagonist (Haller et al., 2006).
Tumor cells are primarily transduced due to the glioma-specific tropism of the VSV-LCMV-GP pseudotypes. Therefore, it is possible to also work with M variants which still show certain cytopathogenicity. In order to keep the toxic effects in BM-TIC low, the antiviral MxA protein should be expressed in the cells via a Tet-inducible promoter. Retroviral vectors with Tet-inducible promoter are known (Löw et al., 2006). The Tet-responsive element (tTA) is also introduced into BM-TIC via a retroviral vector. MxA is an interferon-induced GTPase with antiviral activity towards RNA viruses (Pavlovic et al., 1992; Stäheli and Pavlovic, 1991). In the cytoplasm it is present in the form of large oligomers, is not toxic and inhibits the replication of VSV by a hitherto unknown mechanism. LCMV-GP expressing BM-TIC are transduced with the Tet-regulated MxA and the tTA vector and subsequently infected with the VSV vectors. As long as the antibiotic is present in the medium, the cells are resistant against a productive infection with VSV (LCMV-GP) due to the presence of MxA. Only during the in vivo application, the antibiotic is removed and a viral replication and production of VSV (LCMV-GP) gradually begins. Since MxA with a half-life of two days is very stable, the mutant MxA (L612K) can alternatively be used. MxA (L612K) inhibits the infection of VSV with the same effectiveness as the wild-type protein, but is not able to form oligomers. Hereby it is destabilized and has a short half-life of only two hours (Janzen et al., 2000).
Both producer cells, BM-TIC which produce vectors with non-cytopathogenic M as well as BM-TIC which are protected against VSV-related lysis through MxA expression, are characterized as follows:
A suicide gene can be used as therapeutic gene according to the invention. An example of such a gene is the thymidine kinase of the herpes simplex virus (HSV-TK) with the most experience in the clinic. Through the constitutive expression of the HSV-TK, packaging cells and transduced cells can be reliably eliminated at the end of the therapy so that no danger can arise from mitotically active packaging cells.
For the immunotherapy of malignant gliomas, i.e., for strengthening the poorly developed local immune response in the brain, immunomodulatory genes can be used. Examples of such genes are cytokines, in particular IL-12 and Flt3L. After antigen stimulation, IL-12 is naturally secreted by antigen-presenting cells such as dendritic cells, macrophages and B cells and mediates a strong Th1-based anti-tumor immune response. It enhances the growth and the differentiation of naive T cells into Th1 cells and thus counteracts a glioma-induced suppression of the T cell proliferation and IFN-gamma production. IL-12 enhances the cytotoxic activity of natural killer and CD8+ cells and has in addition an anti-angiogenic activity (Majewski et al., 1996). In a plurality of animal models, the effectiveness of an IL-12-mediated immune response against gliomas has already been demonstrated in an impressive manner (Toda et al., 1998; Kikuchi et al., 1999). The cytokine Flt3L is essential for the proliferation and differentiation of antigen-presented dendritic cells and thus seems to be particularly well suited for the purposes of a gene therapy. Sumia and colleagues observed in the rat/mouse a strong local influx of dendritic cells after i.c. transduction of gliomas with the Flt3L gene. 70% of the treated rats showed tumor regression and survived.
The combination of a cytokine with a suicide gene (TK) does not only offer maximum safety, but also allows a successful therapy with low transduction efficiency due to the bystander effect. This approach combines three active principles, namely viral oncolysis, suicide gene-induced apoptosis and local immunostimulation in order to increase the efficiency of the gene therapy of the glioblastoma.
First, Il-12 and Flt3L are cloned into the VSV vectors. IL-12 is species-specifically active so that the murine IL-12 has to be used for testing in the rat glioma model. In contrast, more than 70% of human and murine Flt3L are homologous and are cross-reactive in both species. The immunostimulatory genes are cloned into the position of the p gene, whereas TKgfp or gfp on the second vector replaces the VSV-G gene. The following gene combinations are cloned into the two vectors of the cr VSV-LCMV-GP pseudotype vector system (
1) vector a=IL12 or Flt3L+vector b=TKgfp (or TK)
2) vector a=IL12 or Flt3L+vector b=gfp
3) vector a=IL12 or Flt3L+vector b=empty
4) vector a=empty+vector b=TKgfp (or TK)
5) vector a=empty+vector b=gfp
6) vector a=empty+vector b=empty
After cloning, glioma cell lines and BM-TIC are transduced and the concentration of the cytokine contained in the supernatant is verified by ELISA. For the production of physiologically relevant amounts the vectors are concentrated via ultracentrifugation. The vectors are first used for the transduction of spheroids and their competence to bystander killing is examined in this system. Afterwards, the vector supernatants together with the prepared packaging cells are sent to Bergen to our cooperation partner. There, the in vivo testing in the syngeneic rat glioma model takes place. The survival rate of the treated rats after ganciclovir application is evaluated. At the same time, the therapeutic effectiveness of the immunotherapy in the absence of TK is examined. The vector combination 6) provides information about the effectiveness of the oncolysis. Tumor size and thus the therapy success are monitored via MRI and PET (Miletic et al., 2007). The effect of the immunomodulatory genes is evaluated by means of immunohistochemical methods for detection of infiltrate cells. As described in the following paragraph, the therapeutic approach is also examined in the human glioma model.
The recombinant VSV vectors are first characterized in vitro and are subsequently investigated in the animal model for their effectiveness. For this purpose, two particularly meaningful rat models are used.
A syngeneic 9L rat glioma model is used for the examination of immunological aspects and the carrying out of immunotherapeutic methods. For this purpose, glioma cell lines are implanted into the brains of Fisher rats which subsequently develop tumors within a few days. 9L gliosarcoma cells which were isolated from Fisher rats served as tumor cell lines. Various vectors or cell lines are then stereotactically injected into the induced tumors. This rat model has the following characteristics:
A human glioma model in nude rats is used for testing the oncolytic activity of the vectors and cells according to the invention towards human glioma cells. The rat xenograft model simulates various growth characteristics of human gliomas. Not a defined glioblastoma cell line but primary tumor material from patients is used here for establishing the tumor, which allows to simulate the heterogeneity within human gliomas (Sakariassen et al., 2006). By the serial passage of human gliomas in nude rats, different phases of the malignant tumor development can be mapped and the therapy success can be examined. The human glioma model is particularly well suited for the examination of therapeutic approaches because it has the following characteristics:
Two categories of tumors can be distinguished:
The human glioma model is very well suited for examining suicide gene therapy as well as the viral oncolysis. However, up to now there are no animal models for the human glioblastoma which can be used for the analysis of the immunotherapeutic effectiveness of gene therapeutic methods.
The success of therapy is analyzed and evaluated by means of imaging methods and finally with histological techniques. First, the mode of action of the complementary replicating VSV (LCMV-GP) pseudotypes is examined by intratumoral injection of concentrated vector supernatants in the rat glioma model. The transduction efficiency and the distribution of the gfp-coding vectors within the tumors are examined after a few days by means of a fluorescence microscopic analysis of frozen sections. Furthermore, oncolysis and the invasion of immune cells are immunohistochemically evaluated. As a supplement, the effectiveness is also examined in the animal model for the human glioblastoma.
In the same model, the migration capability of the BM-TIC producer cells and the infectivity of the produced VSV vectors are examined. For this purpose, the cells are stained with a green dye whereas the released vectors code for rfp. This approach allows a simple fluorescence-based differentiation between BM-TIC and transduced tumor cells.
Finally, the therapeutic efficiency of the individual transgene-coding pseudotypes is compared. For this purpose, the packaging cells and also the concentrated vector supernatants are injected into established 9L tumors. The vectors comprise an immunostimulatory gene in combination with TK. Ganciclovir administration begins a few days after the stereotactical application of the vectors or the packaging cells. The therapeutic efficiency of the immunomodulatory component of the vector is evaluated in animals that receive vectors containing only TK. Upon completion of the treatment, the effectiveness of the individual therapy concepts is evaluated based on the tumor size with imaging methods (MRI; PET). A portion of the brains of rats treated with immunotherapy is used for characterizing the immune status. The expression of MHC molecules and the distribution of T cells, granulocytes, macrophages and microglial cells are examined on the frozen sections by immunohistochemical methods. In addition, the quantity and phenotypical composition of the i.c. infiltrate cells are analyzed by flow cytometry. The formulation with the best therapeutic effect is adapted for the treatment of human gliomas. Such a prototype could be transferred directly into preclinical safety studies and, after establishing the respective methods for GMP production, into clinical studies.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2008 050 860.8 | Oct 2008 | DE | national |
| PCT/EP2009/007230 | Oct 2009 | EP | regional |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP09/07230 | 10/8/2009 | WO | 00 | 6/29/2011 |
