New Adeno-Associated Virus Strategies to Support Momentum in the Clinic

Abstract

A t what seems to be an ever-increasing pace, promising news regarding clinical results using investigational gene therapy products is emerging, including several exciting advances with recombinant adeno-associated virus (AAV). Much like the preceding development of monoclonal antibodies, evolution in therapeutic product design, manufacturing, and characterization builds on and is guided by lessons learned in the pioneering first-generation clinical studies. Optimizing vector design and manufacturing methods for recombinant AAV to make vectors of consistently high safety, purity, and potency, and in sufficient quantities to support the investigational product clinical stages of AAV development with a reasonable “scale-up” pathway to meet the large-scale demands of for prospective commercial success, remains a challenge and important objective. Current vector design and manufacturing challenges include the following three areas, reflecting important aspects of investigational product chemistry, manufacturing, and controls: (1) vector design—optimizing vectors that achieve both efficient payload delivery to target tissues and overcome immunological barriers; (2) manufacturing process and capacity—generating and purifying sufficient amounts of vectors meeting the high standards required for clinical use; and (3) vector characterization—developing and validating robust methods, including critical dose-determining assays, to ensure product safety and consistency of performance, and enabling valid comparison of results among different groups involved in translational research studies. Five articles in this issue of Human Gene Therapy represent important progressive steps in the establishment of recombinant AAV as a successful therapeutic product platform.

Nathwani and colleagues (2010) described promising preliminary clinical results for a recombinant AAV trial in subjects with severe hemophilia B. Building on the experiences and results reported in an earlier AAV2-FIX trial (Manno et al., 2006), this team is using an AAV8 capsid for gene delivery and a self-complementary (sc) factor IX transgene configuration, features predicted to reduce the vector dose needed to achieve efficacy. In this issue of Human Gene Therapy Allay and colleagues (2011) describe in detail the vector-manufacturing process methods and product characterization results for the rAAV8-scFIX vector. The high detail provided is important, and will enable thorough comparison with results obtained in other trials. Transfection of adherent HEK293 cells was used for vector generation, an approach of recognized limited scalability. The vector purification method includes multiple chromatography steps, the basis for a fully scalable purification process. Vector purification did not include a step to remove empty capsids, and as a result the clinical product contains substantial amounts of this product-related impurity. It will be important to determine whether the clinical results obtained support the notion that this more efficient vector helps to avoid the host immune responses that limited efficacy in the previous AAV2-FIX trial. Removal of empty capsids appears to remain an option to reduce capsid protein content to further enhance the therapeutic window. Another important finding is reported regarding titering methods, namely that quantitative PCR (qPCR) underestimates titers 5- to 10-fold relative to titers determined by dot-blot and spectrophotometric methods, likely because of inefficient binding of qPCR primers to scDNA templates. Improved accuracy in vector characterization methods is important, and enables more meaningful comparisons between related clinical studies.

Production of sufficient quantities of recombinant AAV remains a major bottleneck for many promising clinical applications. Although transient transfection approaches for vector generation have certain advantages in the early stages of translational research, for later stages and large-scale vector production “producer” cell lines offer clear benefits. Many such promising producer cell culture systems have been developed, generally involving a cell line containing a subset of requisite genes, with the remaining complement of genes required to initiate AAV biosynthesis introduced by infection with a wild-type or recombinant helper virus. One concern with many scalable, helper virus-dependent methods is the need to use potentially pathogenic wild-type helper viruses in the manufacturing process, requiring implementation and validation of rigorous viral clearance steps during purification. In this issue, Yuan and colleagues (2011) report improvements and simplifications to previously reported HEK293 based producer cell lines (Qiao et al., 2002). In this system AAV vector generation can be supported using a replication-deficient E1-deleted adenovirus, because HEK293 cells already contain the adenoviral E1 gene, resulting in highly efficient production of AAV vectors (>5 × 10¹³ vector genomes per CF10 Cell Factory). This new technology further broadens options for large-scale vector production in a safe and scalable manner.

Novel strategies are required to enhance transduction efficiency of single-stranded AAV (ssAAV) vectors that are too large to fit into a self-complementary vector genome. In previous publications, Srivastava and colleagues reported that the phosphorylated form of the host cell protein FKBP52 (FK506-binding protein) interacts with the AAV inverted terminal repeat (ITR) D-sequence and strongly inhibits viral second-strand synthesis. When present, protein phosphatase-5 (PP5) can reverse this inhibition by dephosphorylating FKBP52 and thereby enhance transduction by ssAAV vectors (Jayandharan et al., 2008). This strategy can reduce vector doses and thereby effectively increase manufacturing capacity by enabling administration of smaller doses per subject. In the current issue of Human Gene Therapy, Ma and colleagues describe a novel strategy to manufacture a “compound” investigational product, composed of the vector expressing a single-stranded therapeutic transgene of interest together with a second, self-complementary vector expressing the PP5 gene, in a single manufacturing process. They performed “quadruple” transfection using standard AAV packaging and adenovirus helper plasmids, and included both ITR-flanked vector plasmids together in the same transfection mixture. The molar ratios of the vector plasmids were adjusted to obtain the desired target ratio of the respective vectors. The authors report that PP5 vector produced and copurified at a level corresponding to 10% of total vector was sufficient to achieve a 5- to 10-fold increase in transduction efficiency using an AAV2-ssEGFP model vector, an enhancement documented in vitro in cultured human cells (HEK293) and in vivo in mice. Although concurrent production and purification leading to a two-vector compound product will introduce the need for additional vector characterization and some new quality control tests, the coproduction approach should offer significantly greater cost benefits in comparison with admixing and recharacterizing two separately manufactured lots.

Results reported in previous studies using AAV have documented the occurrence of cytotoxic T lymphocyte directed to AAV capsid protein, and transient immune modulation may be beneficial for some clinical applications. In this issue of Human Gene Therapy, Montenegro-Miranda and colleagues report the differential effect of mycophenolate mofetil (MMF) on the transduction efficiency of ssAAV and scAAV vectors. This team is developing an AAV1 vector for delivery of the gene encoding uridine diphosphoglucuronosyltransferase isozyme 1A1 (UGT1A1), a deficiency of which results in Crigler–Najjar syndrome, a recessive inherited liver disorder. Their data obtained using the Gunn rat model as well as in cell culture (HEK293T cells) show that MMF at therapeutically relevant concentrations, although effective at delaying and reducing relevant immune responses, also inhibits transduction by ssAAV but not scAAV vectors. This inhibition is proposed to occur via depletion of the intracellular guanosine pool by mycophenolic acid (MPA), the active metabolite of MMF, preventing the second-strand synthesis required for efficient ssAAV transduction. Mingozzi and colleagues previously reported no inhibition of ssAAV carrying the human coagulation factor IX gene by MMF in nonhuman primates (Mingozzi et al., 2007), and this difference may be attributable to the higher systemic MPA levels in the Gunn rat because of its impaired metabolism. These results emphasize the importance of assessing potential interference with AAV vectors by pharmacological agents that may be coadministered in gene transfer protocols.

Investigational product impurities, especially trace nonvector DNA, may adversely affect long-term transgene expression using recombinant AAV. Several laboratories have reported that small amounts of residual cap DNA are present even in highly purified AAV. Although evaluation of a clinical-grade AAV2 vector used in a hemophilia B clinical trial failed to detect transcription of cap DNA by sensitive reverse transcription qPCR methods (Hauck et al., 2009), for any clinical applications in which vector doses can easily exceed 1 trillion particles, it remains a challenge to achieve and validate complete absence of cap-expressing particles. A theoretical risk remains that even trace levels (e.g., 10^–10) of “replication-competent” AAV present in clinical-grade vectors might be sufficient to initiate anti-capsid immune responses, and/or to propagate as a result of in vivo adventitious infection by an AAV helper virus. In this issue of Human Gene Therapy, Lu and colleagues describe a novel approach to eliminate trace residual cap DNA expression by insertion of liver and hematopoietic tissue-specific microRNA (miR) binding cassettes beside cap in the production plasmid used for vector manufacture. They report that vector yield and functional activity were unaffected by inclusion of the miR binding sequences in the production plasmid. Importantly, they found that cap expression was dramatically reduced by the presence of the miR binding cassettes when using a CMV-Cap construct. This approach permanently “marks” trace AAV particles containing expression-competent cap DNA for degradation by an existing host biochemical pathway, and when used in conjunction with other strategies to eliminate replication-competent AAV, provides an important approach to further enhance the safety and efficacy of clinical vectors.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Allay

J.A.

, Sleep

S.E.

, Long

et al. 2011. Good manufacturing practice production of self-complementary serotype 8 adeno-associated viral vector for a hemophilia B clinical trial. Hum. Gene Ther., 22:595–604.

Hauck

, Murphy

S.L.

, Smith

P.H.

et al. 2009. Undetectable transcription of cap in a clinical AAV vector: Implications for pre-formed capsid in immune responses. Mol. Ther., 17:144–152.

Jayandharan

G.R.

, Zhong

, Li

et al. 2008. Strategies for improving the transduction efficiency of single-stranded adeno-associated virus vectors in vitro and in vivo. Gene Ther., 15:1287–1293.

, Qu

, Yang

et al. 2011. Systemic elimination of de novo capsid protein synthesis from replication-competent AAV contamination in the liver. Hum. Gene Ther., 22:625–632.

Manno

C.S.

, Pierce

G.F.

, Arruda

et al. 2006. Successful transduction of liver in hemophilia by AAV-factor IX and limitations imposed by the host immune response. Nat. Med., 12:342–347.

Mingozzi

, Hasbrouck

N.C.

, Basner-Tschakarjan

et al. 2007. Modulation of tolerance to the transgene product in a nonhuman primate model of AAV-mediated gene transfer to liver. Blood, 110:2334–2341.

Montenegro-Miranda

, Ten Bloemendaal

, Kunne

et al. 2011. Mycophenolate mofetil impairs transduction of single-stranded adeno-associated viral vectors. Hum. Gene Ther., 22:605–612.

Nathwani

, Tuddenham

, Rosales

et al. 2010. Early clinical trial results following administration of a low dose of a novel self complementary adeno associated viral vector encoding human factor IX in two subjects with severe hemophilia B. Blood, 116:114.

Qiao

, Wang

, Zhu

et al. 2002. A novel gene expression control system and its use in stable, high-titer 293 cell-based adeno-associated virus packaging cell lines. J. Virol., 76:13015–13027.

10.

Srivastava

, Ma

, Li

et al. 2011. A simple method to increase the transduction efficiency of single-stranded adeno-associated virus vectors in vitro and in vivo. Hum. Gene Ther., 22:633–640.

11.

Yuan

, Qiao

, Hu

et al. 2011. A versatile adeno-associated virus vector producer cell line method for scalable vector production of different serotypes. Hum Gene Ther., 22:613–624.