Coxsackievirus B3-Its Potential as an Oncolytic Virus

Abstract

Oncolytic virotherapy represents one of the most advanced strategies to treat otherwise untreatable types of cancer. Despite encouraging developments in recent years, the limited fraction of patients responding to therapy has demonstrated the need to search for new suitable viruses. Coxsackievirus B3 (CVB3) is a promising novel candidate with particularly valuable features. Its entry receptor, the coxsackievirus and adenovirus receptor (CAR), and heparan sulfate, which is used for cellular entry by some CVB3 variants, are highly expressed on various cancer types. Consequently, CVB3 has broad anti-tumor activity, as shown in various xenograft and syngeneic mouse tumor models. In addition to direct tumor cell killing the virus induces a strong immune response against the tumor, which contributes to a substantial increase in the efficiency of the treatment. The toxicity of oncolytic CVB3 in healthy tissues is variable and depends on the virus strain. It can be abrogated by genetic engineering the virus with target sites of microRNAs. In this review, we present an overview of the current status of the development of CVB3 as an oncolytic virus and outline which steps still need to be accomplished to develop CVB3 as a therapeutic agent for clinical use in cancer treatment.

Keywords: Coxsackievirus B3; cancer; miR; microRNA; oncolytic virus; virus adaptation.

Figure 1

Structure of the CVB3 capsid. (A) A schematic model of the icosahedral CVB3 capsid structure composed of 60 asymmetric units, each composed of one VP1 (green), VP2 (yellow), VP3 (blue) and VP4. VP1, VP2 and VP3 form the capsid surface, while VP4 is located on the inside surface. The red outlined area depicts a pentamer, the white outlined triangle depicts an asymmetric unit. (B) Upper panel: The capsid surface structure colored radially. The radial surface structure was calculated and modelled with the bioinformatic software UCSF ChimeraX [23] based on the structural data of CVB3 RD from the RCSB protein databank (accession no. 4GB3). Lower panel: A magnified section of the asymmetric unit with the arrows highlighting the puff region and the canyon.

Figure 2

Structure of the CVB3 genome and the functions of the viral proteins. The ~7.5 kb viral genome comprises a 5′-UTR with the cloverleaf (CL) structure and the internal ribosomal entry site (IRES), modelled according to Bailey et al. [40], an open reading frame and a 3′-UTR with a poly(A) tail. The small VPg protein binds to the 5′-end of the genome. The open reading frame is translated into a polyprotein which is autocatalytically processed into the structural (green) and non-structural (red and yellow) proteins. During the processing of the non-structural proteins, three precursor proteins (2BC, 3AB, 3CD) with distinct functions in the viral life cycle are formed, which are processed further into the seven non-structural proteins (2A, 2B, 2C, 3A, 3B, 3C, 3D).

Figure 3

Amino acids of the CVB3 capsid involved in CAR and heparan sulfate binding. Upper image: The capsid is shown along with the tertiary structure of the capsid proteins VP1 (green), VP2 (yellow), VP3 (blue) and VP4 (red). The triangle outlines one asymmetric unit. The structure is modelled with the bioinfomatic software PyMOL (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC) based on the structural data of CVB3 RD from RCSB protein databank (accession no. 4GB3). Lower left magnified image: the amino acids involved in CAR binding (according to He et al. [86] and Organtini et al. [26]) are shown in orange. Lower right magnified image: the amino acids within one asymmetric unit of CVB3 PD strain (NCBI accession no. AF231765.1) that differ to the CVB3 Nancy strain (NCBI accession no. JX312064.1) are shown in magenta. Their position within VP1 is indicated in the table. The amino acids which differ in the sequence of PD compared to Nancy are involved in binding of the virus to N- and 6-O-sulfated heparan sulfates.

Figure 4

Interaction of CVB3 with cellular receptors. The prototype CVB3 strain Nancy uses CAR for cell entry, whereas the CVB3 variant PD can infect cells via N- and 6-O-sulfated heparan sulfates and CAR. Upper panel: CAR-positive cells can be infected with PD and the Nancy strain. Lower panel: CAR-negative cells cannot be infected with the Nancy strain, but with PD when N- and 6-O-sulfated heparan sulfates are expressed on the cell surface.

Figure 5

Oncolytic CVB3 strains-origin, oncolytic efficacy and side effects. The Nancy strain of CVB3 was isolated in New Haven, CT, USA in 1948 from a 3-year-old patient called Nancy [99]. The Nancy strain induces pancreatitis and myocarditis and has been shown to be lytic in colorectal and lung cancer [20,92,103,104,105]. The PD variant arose after 13 passages of persistent CVB3 Nancy p strain in human fibroblasts (HuFi) [102]. The PD strain is lytic in HuFi and colorectal carcinoma [92,102]. The strain 31-1-93 was isolated after passaging the PD strain through the murine heart [106]. It shows oncolytic activity in colorectal carcinoma and induces severe pancreatitis and myocarditis in mice [77,92]. CV-B3/2035A strain was isolated in Xiamen, PRC in 2008 from a patient’s throat swab, who presented with light symptoms of hand, foot and mouse disease. The CV-B3/2035A strain has oncolytic potential in endometrial carcinoma [91]. The strain H3 is a cardiotropic strain of CVB3 Nancy, which has oncolytic activity in colorectal cancer but induces severe pancreatitis and myocarditis [74,100,107].

Figure 6

MiR-regulated oncolytic CVB3. A schematic representation of the CVB3 viral genome depicting the insertion sites and copy number of miR-TS used in publications to date. CVB3 genome sites which did not tolerate insertion of miR-TS are shown by black arrows with red crosses through them.