Adenovirus: Teaching an Old Dog New Tricks

A denovirus was the first DNA virus to enter rigorous therapeutic development, largely because of its well-defined biology, its genetic stability, and its ease of large-scale production. Over the years, the virus has seen many aspects of development, ranging from its deployment as a vector for transgene delivery and supplementation and vaccination through to its use as an oncolytic agent. These developments have encompassed the use of wild-type adenoviruses for cancer therapy and as vaccines, oncolytic viruses with subtle deletions that are intended for functional complementation by the tumor phenotype or that are reliant on tumor-associated transcription factors, E1- and E1,E3-deleted viruses for transgene delivery, through to high-capacity helper-dependent adenoviruses that are intended to deliver very large genetic cargoes with minimal provocation of antiviral immunity. Because adenovirus occurs (in humans) as 57 known serotypes grouped as 7 species (A–G), these diverse technological approaches can give rise to a vast range of therapeutic candidate viruses. It is, therefore, not surprising that adenovirus continues to occupy center stage in gene therapy engineering, and this issue of Human Gene Therapy presents a range of papers that illustrate the innovative approaches being pursued to maximize the utility of this versatile platform.

It is widely accepted that poor delivery represents a limiting factor for most forms of in vivo gene therapy, particularly extrahepatic gene therapy, the only exceptions being where therapeutic viruses can be injected directly into the target site or where they can be introduced into target cells ex vivo. Delivery is very much a pharmaceutical concept, and optimizing delivery relies on measuring and improving the distribution and bioavailability of viral particles, as distinct from just measuring transduction, as that does not take account of unsuccessful delivery events. Accordingly, optimizing delivery requires elucidation of “particle kinetics,” and it is refreshing when a study reports efficient recovery of the number of viral particles that were initially injected. This is usually performed by quantitative PCR to measure viral genomes, and is generally possible only in very short time frames, as viral particles are rapidly catabolized in vivo.

In this volume, Zhang and colleagues have prepared Ad5 with modified hexon protein, to prevent binding to factor X, leading to less uptake into hepatocytes in vivo and decreased toxicity. This innovative approach to decreasing hepatocyte toxicity is superior to the use of intracellular controls such as engineering the virus to include binding sites for hepatocyte-specific microRNA, because it decreases the hepatocyte capture of viral particles and should have a positive effect on particle kinetics. This decreased hepatic toxicity is a definite advantage in mice, although we are still awaiting clinical data showing that the equivalent hepatic toxicity is a problem in humans. In addition, preventing capture of viral particles by Kupffer cells is likely to synergize with ablation of hepatocyte infection in extending plasma circulation of viral particles.

Overcoming poor infection rates is an unusual challenge because adenovirus is rather promiscuous in its tropism. Nevertheless, to improve infection of refractory cells, Wang and colleagues have used a cationic liposome preparation to deliver genomes and thereby overcome the deficiencies of the coxsackievirus–adenovirus receptor (CAR) to which they attribute poor infection of the virus.

Acute toxicity and rapid elimination of adenoviral particles after intravenous delivery are due to its interaction with the innate immune system, and the paper by Suzuki and colleagues shows that the cytosolic sensor protein NOD2 (nucleotide-binding and oligomerization domain-2) plays an important role in regulating the immediate response to adenoviral particles. It follows that NOD2 inhibitors might decrease acute inflammatory responses to adenovirus, and simultaneously perhaps increase circulation time.

The adaptive immune response to viral particles and transgene products is also a major challenge, particularly for the use of adenovirus for long-term expression of transgenes. The observation by Seregin and colleagues that incorporating the complement inhibitor DAF (decay-accelerating factor) into the viral capsid can decrease immune provocation by both viral capsid proteins and transgenes may provide important insights into how viral biocompatibility can be enhanced. It may be possible to use this, or related, approaches both to decrease toxicity and to improve persistence of the vector and its cargo.

Oncolytic virotherapy represents perhaps the most exciting avenue of adenovirus development; however, the absence of representative animal models has continually frustrated the study of oncolytic activity in the presence of a functioning immune system. The demonstration by Weaver and colleagues that A20 murine B cell lymphoma cells will support adenoviral oncolysis provides an interesting way forward, and their observation that wild-type group D adenovirus (Ad26 and Ad48) are the most effective virotherapy agent against these cells in vivo provides useful insights into the possible activity of serotype D adenoviruses in B cell lymphomas.

Last, there is considerable discussion of arming viruses with antiangiogenic agents. The paper by Tysome and colleagues shows that the oncolytic Lister vaccine strain of vaccinia is potentiated by encoding a fusion protein of endostatin and angiostatin, showing significantly greater anticancer potency than the unarmed parental virus (and also compared with the early oncolytic adenovirus, ONYX-015).

Smith and colleagues provide an overview of factors limiting the utility of oncolytic adenovirus, and they identify several delivery issues including the presence of interstitial cells and stroma as likely to prevent virus spread through tumor nodules. A range of approaches are being developed worldwide to address the issues identified in this review, and we await the outcomes of several preclinical and clinical studies to identify which routes represent the optimal ways forward.