Virus-directed enzyme prodrug therapy
Adenoviruses are particularly suited as therapeutic tool as they are not integrated into the host genome and possess a variable load capacity for gene transfer which is about 2 kb for wild type adenovirus and can be expanded stepwise up to 11 kb by deletion of certain viral genes (Bett et al., 1993; Amalfitano et al., 1998). The first report of using viruses for treating diseases was by Rogers et al. (1973) who tried to induce arginase activity with the Shope papilloma virus (or cottontail rabbit papilloma virus) in tissue culture cells from an argininaemic patient—albeit without success. The VDEPT has been used among others for selective activation of prodrugs in tumor tissues by exogenous enzymes for cancer therapy. As the name suggests, in this kind of GDEPT a virus delivers the gene for a non-endogenous enzyme to a target tissue via a viral vector. Viral vectors employed for VDEPT/GDEPT include retroviruses, adenoviruses, e.g. (recombinant) adeno-associated virus (serotype 2, AAV-2), herpes simplex virus and lentiviruses of which the two first ones are employed most often and two-thirds of gene therapy clinical trials address cancer diseases, followed by cardiovascular and monogenic diseases. Most viral vector systems used in gene therapy are engineered; a deletion of many viral sequences aims rendering these gene delivery vehicles replication defective and limiting the production of potentially harmful viral antigens resulting in unwanted side effects; an alternative are so-called conditionally replicating adenoviruses (see below). For more information including methods of efficient cell targeting as a key to enhancing therapeutic effect, and the different types of promoters available for regulating transgene expression, see the review published by Bouard et al. (2009).
Examples of application are among others a phase I/II clinical trial in localized prostate cancer (using direct intraprostatic injection) transgene expression of a virus encoding bacterial nitroreductase (Patel et al., 2009) and a directed enzyme prodrug therapy for treating patients with resectable liver cancer (Palmer et al., 2004). In both investigations, a non-replicating Ad5 adenovirus Vector (CTL102) was used together with CB1954 as prodrug. Szewczuk et al. (2017) investigated the cytotoxicity of N02-groups bearing benzimidazole derivative-prodrugs with A549 (non-small-cell lung cancer) cell line. These cells were transfected with adenovirus transferring E. coli nfsB genes with the gene product being FMN nitroreductase (see above and also the previous Sections), too. The prodrug substrates become cytotoxic by reducing the -N02 moieties to -NH(OH) residues (Vass et al., 2009). They observed among others that all tested benzimidazole derivatives reduced the expression of the anti- apoptotic protein BCL2 but significantly stimulated the BAX gene producing the pro-apoptotic BAX protein under hypoxia as well as normoxia conditions. However, in hypoxia apoptosis was induced earlier and with distinctly higher efficiency.
Modular recombinant adeno-associated virus (rAAV) vectors for cellular VDEPT were developed by Hagen et al. (2014). They made use of the fact that the human epidermal growth factor receptor (EGFR) is expressed on tumor surfaces and armed rAAV-2 capsids with two published modular targeting binding proteins DARPin or Affibody molecules (see, e.g., Friedman et al., 2008) which are genetically engineered antibody mimetic proteins with high-specificity in the nanomolar range for the cancer cell-surface marker EGFR. They also included an affinity tag (His6-tag) in a surface-exposed loop for affinity purification of the rAAV capsids, e.g., to separate them from unmodified capsids potentially causing negative side effects. The particles were equipped with genes coding for the enzymes thymidine kinase or cytosine deaminase converting the prodrugs ganciclovir and 5- fluoro-cytosine, respectively; the system cytosine deaminase/5- fluorocytosine is of advantage in so far that the resulting drug 5-fluorouracil can pass the membrane thereby eliciting the above mentioned killing bystander effect. These modularly assembled recombinant viral vectors were shown to effectively induce apoptosis to, e.g., EGFR-overexpressing A431 (epidermoid carcinoma) cells. Furthermore, this approach might be applied for targeting a variety of other cell surface markers presupposed the guiding molecules are exchanged.
Virus-mediated treatment of familial lipoprotein lipase deficiency
Lipoprotein lipase (LPL) hydrolyzes triglycerides (TG) in lipoproteins (found in lipoprotein particles such as chylomicrons and very low-density lipoproteins) into two free fatty acids and one monoacylglycerol molecule. The enzyme requires apolipoprotein C2 as cofactor for activation and is present among others in heart, adipose, and skeletal muscle tissue. The molecular mass of the glycoprotein LPL is between 55 and 58 kDa of which 12% are carbohydrates. Lipoprotein lipase provokes clearance of triglyceride-rich lipoproteins from the blood. Deficiencies or defects in LPL cause serious hypertriglyceridemia and susceptibility to chronic, life-threatening pancreatitis. Lipoprotein lipase deficiency, LPLD, is an ultra-rare autosomal recessive lipid disorder that results from a mutation in the gene encoding LPL lipoprotein lipase (Burnett et al., 2017). It is estimated that only 1 or two out of 1.000 000 people are affected.
In July 2012, the European Medicines Agency (EMA; see also Vol. 4, Chapter 1) recommended Alipogene tiparvovec (marketed under the trade name Glybera by Amsterdam-based uniQure N.V., a gene therapy company), a gene therapy treatment of LPLD, after it has been reviewed by the EMA's Committee for Advanced Therapeutics (CAT) and recommended to the EMA Committee on Human Medicinal Products (CHMP); the recommendation was finally confirmed by the European Commission in November 2012 (Richards, 2012). In the USA, the first gene therapies were approved by the FDA in 2017.
The Glybera therapy designed as a one-time treatment is based on adeno-associated virus that delivers a functional copy of the LPL gene to skeletal muscle via intramuscular injection. The adenoviral vector serves as a carrier for the S447X variant (carried by 20% of the human population) of LPL. This mutation reduces the full-length LPL by one residue by substituting the codon for a terminal serine by a stop codon. S447X is a gain- of-function mutation; it not only decreased plasma TG but also increases high-density lipoprotein cholesterol (HDL-C) and reduces coronary artery disease risks (Ranganathan et al., 2012). In Glybera 96% of the AAV genome was replaced with the LPL gene, a promotor, and regulatory elements from other viruses, constituting the therapeutic DNA (Richards, 2012). Altogether alipogene tiparvovec with multiple rounds of applications faced rocky path to approval, last not least because the numbers of patients suffering from LPLD were too small to reach statistical significance (Libby and Wang, 2014, and literature cited therein).
With 1.000 000 + per treatment Alipogene tiparvovec was the most expensive therapy world-wide. Rather recently the management of UniQure decided not to ask European authorities to renew the gene therapy's marketing authorization when it expires in October 2017 due to its extremely limited demand since its approval in Europe in 2012 (Sagonowsk, 2017).
Conditionally replicating adenoviruses for cancer therapy
Conditionally replicating adenoviruses (CRA; also known as oncolytic adenoviruses) have an advantage over replication- deficient adenovirus vectors as a transgene delivery system because they selectively replicate in tumor cells and lyses them and amplify the expression of therapeutic genes in the tumor microenvironment. Their generation is achieved by deleting viral elements that are necessary for virus replication in normal cells but dispensable in tumor cells (Jounaidi et al., 2007). Such an element may be the E1B gene that encodes a 55-kilodalton protein inactivating the cellular tumor suppressor protein p53. The p53 protein is a nuclear transcription factor which can activate DNA repair proteins or initiate apoptosis in case of irreparable DNA damage and functions as a tumor suppressor; more than 50% of human cancers carry loss of function mutations in p53 gene (Ozaki and Nakagawara, 2011). An example of a CRA is Onyx-015 (an adenovirus hybrid of serotypes 2 and 5) were the ElB-55kDa gene has been deleted. As shown for the first time by Bischoff et al. (1996), this mutant adenovirus can replicate in p53-deficient human tumor cells and lyse them but cells with functional p53 remain unaffected. They further demonstrated that an injection of this mutant virus into p53-deficient human cervical carcinoma cells grown in nude mice resulted in significant tumor size reduction and regression. The fact that Onyx-015 does not replicate in normal human mammary epithelial cells and microvascular endothelial cells was confirmed by other groups, but it has been meanwhile also found that a variety of tumor cell lines with wild-type p53 allow efficient replication of Onyx-015 (Ries and Korn, 2002). An alternative approach to generate oncolytic viruses is by directed evolution. The success depends on a diverse starting pool based on some mutated serotypes up to entire viral classes and is further improved by the ability of serotypes to undergo recombination, and on selective pressure achieved by passaging them under conditions that they are expected to encounter in human cancer microenvironments using a variety of different cancer cell lines. As a result, the chimeric ColoAdl virus was isolated from viruses that evolved under this selective pressure as the first nonAd5-based oncolytic adenovirus. Its efficacy was tested in vivo in a colon cancer liver metastasis xenograft model and was found to be more potent and selective than, e.g., Onyx-015. The ColoAdl virus could be also armed with exogenous genes by incorporating them into the viral genome and accounting for the complexity of treatment of solid tumors, in this case colon cancer (Kuhn et al., 2008; Bauzon and Hermiston, 2012).
Zhang et al. (2003) published a novel concept which they termed Gene-ViroTherapy strategy that couples the lytic capability of adenovirus with the capacity to deliver therapeutic gene, and for which they used the adenovirus ZD55 derived from Ad5 that does not only bear the E1B 55-kD gene deletion similar to ONYX-01 but also a clone site for easy insertion of a foreign gene for which they used the cytosine deaminase (CD) gene. An application of the ZD55-CD construct in combination with 5-fluorouracil proved significant antitumoral efficacy in nude mice with subcutaneous human colon cancer.
A more recent alternative application example has been provided by Liang et al. (2017). They also employed a E1B-55 gene-deleted ZD55 adenovirus that was further engineered to express the interleukin (IL) gene IL-24. IL-24 belongs to the IL-10 cytokine family and is an immunomodulatory cytokine and in addition displays broad cancer-specific suppressor effects. These include an inhibition of angiogenesis, sensitization to chemotherapy, and induction of cancer-specific apoptosis; it has no effect on normal cells, and lacks of significant side effects (Persaud et al., 2016). They combined this adenovirus expressing IL-24 with Temozolomide (TMZ), an imidazotetrazine derivative that can reach the central nervous system in therapeutic concentrations; it is spontaneously converted at physiological pH to the active metabolite (3- methyltriazen-l-yl) imidazole- 4-carboxamide. The therapeutic benefit of this oral chemotherapeutic drug relies on its ability to alkylate/methylate DNA mainly at the N-7 or 0-6 positions of guanine residues resulting in DNA damage and death of tumor cells. Its clinical activity may be limited by 0-6-alkylguanine-DNA alkyltransferase, a DNA repair protein which removes О-6-alkylguanine adducts in DNA (Zhu et al., 2014, and literature cited therein). As found by infecting A375 and M14 melanoma cells with ZD55-IL-24 in the presence or absence of TMZ, ZD55-IL-24 and TMZ upregulate p53 expression, whereas TMZ alone induced only a weak expression. This was also observed in connection with protein expression levels of the proapoptotic proteins В-cell lymphoma-2 (Bcl-2)-like protein 4, and reduced the levels of the antiapoptotic proteins Bcl-2. The BCL-2 family proteins are key mediators of the apoptotic response to targeted anti-cancer therapeutics (Hata et al., 2015). As a result, the combination of ZD-55-IL-24 and TMZ was more effective at destroying melanoma cells in vitro than either treatment alone (Liang et al., 2017). For a review of oncolytic viruses in cancer treatment, see Lawler et al. (2017).
Chimeric antigen receptor (CAR) T cell therapy and recently FDA-and EMA-approved gene therapies
Advances in molecular biology have led to the identification of chimeric antigen receptors (CAR) as a potential target for reconstructing/inserting new epitopes on the receptor region resulting in a new kind of immunotherapy, the CAR T-cell therapy, which can lead to T-cell activation via antibody-like recognition. T cells, termed so because they mainly mature in the thymus from thymocytes, are a subtype of white blood cells. In contrast to other lymphocytes (B cells, natural killer cells) they are decorated on their surface with a receptor, the T-cell receptor (TCR). The majority of human T cells are part of the adaptive immune system and protect against various pathogens after being converted to activated effector T cells following encounter with antigen-presenting cells (APCs: e.g., macrophages, dendritic cells). Among these activated effector T cells are cytotoxic T cells, which kill their target cells by apoptosis. For more details concerning the properties of T cells, see Smith-Garvin et al. (2009).
The concept of CAR goes back to Gross et al. (1989) who constructed the first chimeric T cell receptor gene, demonstrating the principle of genetically redirecting cytotoxic T cells to tumor cells. This cTCR was expressed on the surface of cytotoxic T lymphocytes, recognized tumor-associated surface antigens and triggered T-cell activation independent of the expression of the antigen peptides-loaded major histocompatibility complex (MHC) molecules (cell surface proteins essential for the immune system to recognize foreign molecules) as required in case of the T cell receptors (TCRs) that recognize antigens in the context of MHC. The MHC is a set of cell surface proteins that in connection with the adaptive immune system binds pathogen-derived antigens to present them on the cell surface for T-cell recognition. An engineering of blood T lymphocytes with a recombinant T-cell receptor of defined specificity can recognize related MHCs of a tumor-associated antigen (TAA). Such genetically engineered T cells have been reported to mediate cancer regression in human patients with metastatic melanoma; tumor recognition was conferred by a T-cell receptor encoding retrovirus (Morgan etal.,2006).
Whereas the TCR is a heterodimeric protein, the chimeric antigen receptor (CAR also termed "T-bodie”) consists of a single polypeptide chain. A CAR receptor is characterized by three main components. The extracellular domain contains the scFv region, a chimeric protein made up of the light and heavy chains of immunoglobins (VH and VL immunoglobulin homologs). This CAR antigen binding domain is linked via an extracellular hinge domain to the hydrophobic transmembrane domain that is responsible for the stability of the receptor and influences the immunogenicity. The intracellular domain contains the CD (Cluster of Differentiation) 3 zeta chain, a glycoprotein that plays an important role in coupling antigen recognition to a variety of intracellular signal-transduction pathways. The intracellular tails of the £ chain contain three immunoreceptor tyrosine-based activation motifs (ITAMs); their phosphorylation induces the downstream signaling pathway for T-cell activation. This construct is known as first-generation CAR. It should be reminded here that the cognate antigen must be on the cancer cell surface to trigger CAR T-cell activation which means that intracellular antigens are not recognized.
Higher generation CARs are additionally equipped with costimulatory molecules such as CD27, CD 28 or CD 137. Costimulation is essential for T cell proliferation, differentiation, and survival and determines significantly the result of a T cell's encounter with an antigen. Co-stimulation signals are generated from the interaction of receptors on the T cells' surface with ligands on antigen-presenting cells. T cell stimulation by CD28 (the receptor for CD80 (B7.1) and CD86 (B7.2) proteins) is among others involved in the production of various interleukins (e.g., IL-2, IL-6). CD80 expression is upregulated in antigen presenting cells (APCs) via Toll-like receptors, whereas CD86 expression on APCs is constitutive. However, CD80 and CD 86 are missing in many cancer cells with the consequence to fail to respond to their specific antigen (T cell anergy) so that co-stimulation is indispensable for full T-cell activation which is achieved by combining the intracellular signaling domain of CD28 to CD3£ in one polypeptide chain of the same second- generation CAR. In addition, costimulatory molecules like CD 137, a member of the tumor necrosis factor receptor (TNFR) superfamily family expressed on activated CD8+ T cells can be integrated in first- or second-generation CARs to give a third- generation one. Altogether these CAR-modifications serve to preserve survival and prolong polyclonal expansion of engineered T cells contributing to an increased amplification which results in prolonged T-cell persistence and an improved anti-tumor attack (Chmielewski et al., 2013). According to Sommermeyer et al. (2016) chimeric antigen receptor-modified T cells derived from defined CD8+ and CD4+ (T helper cells) subsets confer superior antitumor reactivity in vivo.
Until December 2015, more than 200 protocols related to adoptive T cell therapy (АТС) in human cancers were registered by the U.S. National Library of Medicine, and about 40% of these deal with the use of CAR T cells of which 65% are studied in trials for hematological malignancies which include relapsed В-cell acute lymphoblastic leukemia (B-ALL), chronic lymphocytic leukemia (CLL), and В-cell non-Hodgkin lymphoma (B-NHL) (Park et al., 2016). The most commonly targeted antigen for treating these diseases with CAR-Ts is the B-lymphocyte antigen CD 19. CAR-T cell therapy for solid tumors is still in its early stages. Among the targeted biomarkers are mesothelin, a protein overexpressed in several cancers, or human epidermal growth factor receptor family members, overexpressed in several solid tumors (breast, ovarian, bladder, pancreatic, non-small-cell lung cancer, etc.). For reviews related to CAR-T cell therapy for solid tumors, see Newick et al. (2017), Yong et al. (2017) or Almasbak et al. (2016). More general reviews have been published, e.g., by Abate-Daga and Davila (2016) and Smith et al. (2016) and by the National Cancer Institute at the National Institutes of Health (2017).
A CAR-T cell therapy starts with removing white blood cells including the T cells from the patient's blood by apheresis. Then T cells are activated and genetically modified with a CAR gene in the lab which will lead to an expression of CAR protein on the T-cell surface that recognizes certain cancer biomarkers. CAR transduction is achieved by using a virus (retrovirus or lentivirus) or by electroporation. Finally the patient receives an infusion with the modified T-cells. IL-2 is known for its potent T cell growth-inducing functions in vitro and has been used for the culture and expansion of various T cell products including CAR T cells (Zhang et al., 2018).
The most often observed side-effect of CAR-T cell therapy after infusion is the cytokine release syndrome (CRS), an immune activation resulting in elevated inflammatory cytokines ranging from mild to severe CRS with the latter being potentially life threatening. If the targets of CAR-T cells are not restricted to antigens on tumor cells but also recognize antigens on normal cells, toxicity may result from this so-called "on-target/off-tumor” effect. One strategy to overcome this complication is to arm the CARs with additional suicide genes, e.g., the herpes simplex virus thymidine kinase (HSV-TK) gene (see Section 6.3.2) the activation of which in case of adverse events would destroy CAR T cells specifically and permanently. Alternatively the specificity of CARs may be enhanced by constructing them in a way that the recognition of two antigens is required for being activated (Newick et al., 2016, and literature cited therein). Furthermore, murine mAb-derived antigen-recognition domains in CAR-T cells may result in their rejection due to the immunogenicity of foreign protein (Bonifant et al., 2016). Improvements in developing a new generation of CAR-T cell therapies with increased safety and efficacy have been reviewed by Li and Zhao (2017).
Rather recently the US FDA approved the first three gene therapies, one is for young people up to 25 years of age with В cell acute lymphoblastic leukemia (termed Kymriah), approved in August 2017 and the other one (Yescarta), a CD19- directed genetically modified autologous T cell immunotherapy for adults with non-Hodgkin lymphoma, approved in October 2017. According to Novartis, Kymriah costs $ 475,000 for a onetime treatment, and the prize for Yescarta made by Gilead Sciences is around $ 373,000. Yescarta and Kymriah both are CAR-T-Cell therapeutic agents. The third one, Luxturna (Spark Therapeutics) approved in December 2017, is an adeno-associated virus vector-based gene therapy indicated for the treatment of patients born with retinal dystrophy, a rare condition that destroys cells in the retina required for vision; the therapy relies on a genetically modified virus that carries the gene for the protein RPE65 that is essential for restoring the active light- sensitive receptor protein rhodopsin. It is speculated that the costs for one treatment will amount to at least $ 1 million (Stein, 2017; Hamers, 2017). In August 2018, Tisagenlecleucel (Kymriah®) has been approved by EMA for patients up to age 25 years with В-cell acute lymphoblastic leukemia, and for adults with diffuse large В-cell lymphoma (DLBCL). In addition axicabtagene ciloleucel (Yescarta®) has also been granted marketing authorization by the European Commission in patients with relapsed or refractory DLBCL, as well as for patients with primary mediastinal large В-cell lymphoma (for details/conditions, see, e.g., Williams, 2018).