For successful transgene expression, the gene of interest must fit within the packaging capacity of the vector to enable transport to the target cells.^12,14 For rAAV, this packaging capacity is ≈5.0 kb.^12,14 If a transgene exceeds this size, the complementary DNA (cDNA) may need to be altered to overcome this barrier.¹¹ Additionally, transgene expression can potentially be improved through the optimization of the transgene itself (e.g. via codon optimization).³ For hemophilia gene therapy, there are various considerations when selecting and optimizing F8 or F9 transgenes.¹¹

Hemophilia A - F8 gene size

FVIII is encoded by the F8 gene, which is ≈7 kb in size. This far exceeds the limited packaging capacity of rAAV, and therefore, the F8 cDNA must be reduced in size in order to be packaged within rAAV.¹¹

The structure of the F8 gene is well documented and is known to comprise three homologous A domains, two homologous C domains, and a unique B domain that constitutes 38% of the primary cDNA.¹⁵ The A and C domains are essential for the procoagulant activity of FVIII; however, deletion of the B domain does not appear to impair protein function.¹⁵ While data have suggested that complete deletion of the B domain may impair post-translational trafficking and secretion,^2,16 constructs exploring partial deletion of the B domain have been explored.² Partial deletion of the F8 B domain has been shown to lead to an increase in mRNA levels and an increase in secreted FVIII protein compared with wild-type F8.² Recombinant B-domain-deleted (BDD)–FVIII is widely used as a replacement factor for the treatment of people with hemophilia A,² and adopting a similar approach of using BDD-F8 cDNA for gene therapy can address the packaging restrictions of rAAV.¹¹

Hemophilia B - F9 gain-of-function mutation

FIX is encoded by the F9 gene, which is 1.6 kb in size.¹¹ This small gene can be packaged into the rAAV expression cassette without any modifications to its size.^2,11

Optimization of F9 for gene therapy has focused primarily on increasing gene expression following delivery of the transgene to the target cells. This can be achieved by incorporating a gain-of-function point mutation in the F9 gene. The naturally occurring gain-of-function point mutation (R338L), named F9-Padua, has been shown to lead to an increase in FIX activity versus wild-type FIX.^16,17 Use of F9-Padua cDNA is being investigated for hemophilia B gene therapies.¹⁶

Codon optimization of F8 and F9 gene

Transgene codon optimization is an approach used to potentially maximize the expression and therapeutic potential of a particular transgene.²¹ Most amino acids are encoded by more than one codon, with different organisms and cell types showing bias towards certain codons.²² The goal of codon optimization is to match codon usage in the transgene with the abundance of transfer RNA for each codon in a particular cell type²¹ through the use of synonymous codon substitutions.²² This, in turn, may increase the rate and efficiency of translation by using more abundant codons.²²

Expression of the F8 and F9 transgenes can be further improved through codon optimization, which is now standard practice when expression cassettes and vectors are being developed for clinical studies.³ Several different codon-optimized F8 transgenes have been demonstrated to improve expression levels of FVIII.^2,16 For F9 transgenes, codon optimization typically results in an increase in expression.¹⁶ However, data have shown that codon optimization of F9 may impact protein conformation and lead to different translation kinetics compared with wild-type FIX.²² These changes can be unpredictable and require further investigation.²²

In addition to optimizing the transgene, the expression cassette can be further enhanced through consideration of the promoter and enhancer elements.⁴

Promoters

Promoters are essential for transgene expression,¹⁰ and can be ubiquitous or tissue-specific.⁴

Ubiquitous promoters, such as elongation factor 1α-subunit and immediate-early cytomegalovirus (CMV), can be used to promote expression across many tissue types. The level of expression can vary between tissues and cell types.⁴

Tissue- or cell-specific promoters allow for greater control of transgene expression.⁴ Hemophilia gene therapy aims to establish F8 or F9 transgene expression in liver hepatocyte cells.¹¹ By using promoter sequences from proteins naturally synthesized in hepatocytes, such as α1 antitrypsin or thyroxine-binding globulin, the expression of F8 and F9 can be targeted to the liver with minimal expression in other tissues, therefore reducing the likelihood of a transgene-induced immune response.^3,4,6 Additionally, natural endogenous promoters of the therapeutic gene offer the possibility of directing expression in relevant cells, at the appropriate level and time.²³

Modification of strong, large promoters is also under investigation, with small cis elements possibly replacing long promoter sequences.³

Enhancer

Enhancers support the activity of promoters by providing binding sites for regulatory proteins.⁷ An in vitro study of enhancer / promoter combinations for optimal FVIII expression in nonviral vectors found that the combination of chimera enhancer HCR-1 / HCR-2 and CMV promoter led to an increase in factor expression compared with only the promoter alone.²⁴ ​​​​

Introns can also be used to optimize transgene expression. In a study optimizing rAAV vectors for liver-directed gene therapy, it was shown that placing an intron between the promoter and transgene resulted in higher transgene expression than if no intron was included, with this effect being most apparent with the intron from minute virus of mice (MVM, from the parvovirus family of viruses).^4,25,26

An additional consideration relates to the biology of the single-stranded AAV-delivered transgenes. After delivery to the nucleus, the single-stranded transgene needs to be converted into a double-stranded transgene before expression can occur.²⁷ This rate-limiting step to gene expression can be bypassed using self-complementary (sc) DNA vectors, referred to as scAAV.²⁷ scAAV contains both the coding and complementary sequence of the transgene expression cassette.²⁷

scAAV vectors have been demonstrated to be more potent than single-stranded vectors and are currently being investigated for the delivery of F9 in some clinical trials.^2,28 Because of the reduced packaging capacity of scAAV, this approach is not currently being investigated for the delivery of F8.^2,27